Energy Technology 2023: Carbon Dioxide Management and Other Technologies 3031226372, 9783031226373

Table of contents :
Preface
Contents
About the Editors
Part I Renewable Energy and Combustion Technologies
1 Analysis of Environmental Impact of Vertical Axis Wind Turbine Using Circular Economy Approach
2 Corrosion and Erosion Protection to Accelerate Deployment of Sustainable Biomass
3 Development of Indium-Tin Oxide Thin Films on PAMAM Dendrimer Layers for Perovskite Solar Cells Application
4 DFT Study of CuS-ZnS Heterostructures
5 Effect of H₂ Enrichment on CO/N₂/H₂-air Turbulence Partial Premixed Flame Combustion Characteristics
Part II Energy Efficiency, Decarbonization and CO₂ Management
6 CO₂ Mineralization and Critical Battery Metals Recovery from Olivine and Nickel Laterites
7 Decarbonization Pathways for an Aluminum Rolling Mill and Downstream Processes
8 Rethinking the Decomposition of Refractory Lithium Aluminosilicates: Opportunities for Energy-Efficient Li Recovery from LCT Pegmatites
9 Energy-Saving Green Technologies in the Mining and Mineral Processing Industry
10 Extraction of Valuable Metals from Luanshya Copper Smelting Slag with Minimal Waste Generation
11 Carbon Footprint Assessment of Waste PCB Recycling Through Black Copper Smelting in Australia
12 Screening High-Entropy Alloys for Carbon Dioxide Reduction Reaction Using Alchemical Perturbation Density Functional Theory
Part III Thermal Management, Environmental and Energy Technologies
13 Novel Thermal Conductivity Measurement Technique Utilizing a Transient Multilayer Analytical Model of a Line Heat Source Probe for Extreme Environments
14 The Effect of Reduced Flue Gas Suctioning on Superstructure and Gas Temperatures
15 Assessing the Environmental Footprints of Gold Production in Nevada
16 Polymeric Composite Dense Membranes Applied for the Flue Gas Treatment
17 Molten Salt Mg-Air Battery Improvement and Recharging
18 Superconductor Busbar Systems in the Light of Increased Energy Costs
19 Critical Metals for Clean Energy: Extraction of Rare Earth Elements from Coal Ash
Part IV Energy Technologies
20 Investigation of Slag and Condensate from the Charge Top in a FeSi75 Furnace
21 Lithium Extraction from Natural Resources to Meet the High Demand in EV and Energy Storage
Part V Poster Session
22 Hydrogen Storage Properties of Graphitic Carbon Nitride Nanotube Synthesized by Mix-Grind Technique
23 Study on Preparation and Electrocatalytic Performance of Self-supported Carbon Transition Metal Catalysts
24 Modification and Evaluation of Energy Saving and Consumption for Reduction Technology of 500 t/d Beckenbach Annular Lime Kiln
25 Research on the Gasification Characteristic of Cokes of BIOC-HPC Extracted from the Mixture of Low-Rank Coal and Biomass
26 Thermodynamic Examination of Selected Phases in the Ag–Co–Sn–S System at  T < 600 K by the Solid-State EMF Method
Author Index
Subject Index

Citation preview

Energy Technology 2023 Carbon Dioxide Management and Other Technologies

EDITORS

Shafiq Alam Donna Post Guillen Fiseha Tesfaye Lei Zhang Susanna A. C. Hockaday Neale R. Neelameggham Hong Peng Nawshad Haque Yan Liu

The Minerals, Metals & Materials Series

Shafiq Alam · Donna Post Guillen · Fiseha Tesfaye · Lei Zhang · Susanna A. C. Hockaday · Neale R. Neelameggham · Hong Peng · Nawshad Haque · Yan Liu Editors

Energy Technology 2023 Carbon Dioxide Management and Other Technologies

Editors Shafiq Alam University of Saskatchewan Saskatoon, SK, Canada

Donna Post Guillen Idaho National Laboratory Idaho Falls, ID, USA

Fiseha Tesfaye Metso Outotec Metals Oy Espoo, Finland

Lei Zhang University of Alaska Fairbanks Fairbanks, AK, USA

Susanna A. C. Hockaday Curtin University Perth, WA, Australia

Neale R. Neelameggham IND LLC South Jordan, UT, USA

Hong Peng University of Queensland Brisbane, QLD, Australia

Nawshad Haque Commonwealth Scientific and Industrial Research Organisation (CSIRO) Clayton, VIC, Australia

Yan Liu Northeastern University Shenyang, China

ISSN 2367-1181 ISSN 2367-1696 (electronic) The Minerals, Metals & Materials Series ISBN 978-3-031-22637-3 ISBN 978-3-031-22638-0 (eBook) https://doi.org/10.1007/978-3-031-22638-0 © The Minerals, Metals & Materials Society 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

It is an honor to present the Energy Technology 2023: Carbon Dioxide Management and Other Technologies proceedings volume that contains peer-reviewed papers presented at the Energy Technologies and CO2 Management Symposium organized in conjunction with the TMS 2023 Annual Meeting & Exhibition in San Diego, California, USA. This symposium was organized by the TMS Energy Committee, which is part of the TMS Extraction & Processing Division and Light Metals Division. Industrial activities, economic development, the environment, and public welfare are dependent on clean and sustainable energy. Taking into consideration the goal of NetZero, researchers in both academia and industry as well as policymakers are now putting tremendous efforts into the generation, storage, and applications of clean energy. Therefore, the Energy Technologies and CO2 Management Symposium was open to participants from academia, industry, and government sectors and focused on new and efficient energy technologies including innovative ore beneficiation, smelting technologies, recycling and waste heat recovery, and emerging novel energy solutions. This proceedings volume reflects an incredible effort by all the authors and their organizations in conducting their research and preparing the manuscripts that are published in this book, which contains research and development papers on mature and new technological aspects of sustainable energy ecosystems, materials for energy storage, life cycle assessment of energy systems, energy-efficient technologies in extractive metallurgy as well as processes and devices that improve energy efficiency, reduce thermal emissions, and reduce carbon dioxide and other greenhouse gas emissions. It is the editors’ sincere hope that this proceedings volume will remain a valuable record of the 2023 Energy Technologies and CO2 Management Symposium and that it will serve as a reference for materials scientists and engineers as well as metallurgists for exploring innovative energy technologies, novel energy materials processing, and carbon sequestration techniques. The production of the proceedings volume was a major undertaking, and many individuals were involved over the course of several months. The engagement of TMS

v

vi

Preface

and committee members to chair sessions and review manuscripts made this symposium and its proceedings possible. The editors would like to extend their sincere appreciation to the authors for their contributions and acknowledge the invaluable support from the TMS staff members in the production of this proceedings volume. Shafiq Alam, Ph.D., P.Eng. Lead Organizer

Energy Technology 2023: Carbon Dioxide Management and Other Technologies Editors Shafiq Alam, University of Saskatchewan, Canada Donna Post Guillen, Idaho National Laboratory, USA Fiseha Tesfaye, Metso Outotec Metals Oy, Finland Lei Zhang, University of Alaska Fairbanks, USA Susanna A. C. Hockaday, Curtin University, Australia Neale R. Neelameggham, IND LLC, USA Hong Peng, University of Queensland, Australia Nawshad Haque, CSIRO, Australia Yan Liu, Northeastern University, China

Contents

Part I

Renewable Energy and Combustion Technologies

Analysis of Environmental Impact of Vertical Axis Wind Turbine Using Circular Economy Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Satyendra Dayalu, Shalini Verma, Akshoy Ranjan Paul, and Nawshad Haque Corrosion and Erosion Protection to Accelerate Deployment of Sustainable Biomass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patrick Shower, Scott Weaver, Voramon Dheeradhada, Aida Amroussia, Michael Pagan, Patrick Brennan, Martin Morra, Bruce Pint, Suresh Babu, Phil Gilston, Steve Lombardo, Tamara Russell, and Anteneh Kebbede Development of Indium-Tin Oxide Thin Films on PAMAM Dendrimer Layers for Perovskite Solar Cells Application . . . . . . . . . . . . . . Firdos Ali, Alecsander D. Mshar, Ka Ming Law, Xiao Li, A. J. Hauser, Shanlin Pan, Dawen Li, and Subhadra Gupta DFT Study of CuS-ZnS Heterostructures . . . . . . . . . . . . . . . . . . . . . . . . . . . . Louis Oppong-Antwi and Judy N. Hart Effect of H2 Enrichment on CO/N2 /H2 -air Turbulence Partial Premixed Flame Combustion Characteristics . . . . . . . . . . . . . . . . . . . . . . . . Fan Yang, Qingguo Xue, Haibin Zuo, Binbin Lv, Yu Liu, and Jingsong Wang Part II

3

15

27

39

51

Energy Efficiency, Decarbonization and CO2 Management

CO2 Mineralization and Critical Battery Metals Recovery from Olivine and Nickel Laterites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fei Wang and David Dreisinger

63

vii

viii

Contents

Decarbonization Pathways for an Aluminum Rolling Mill and Downstream Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alexander Wimmer

75

Rethinking the Decomposition of Refractory Lithium Aluminosilicates: Opportunities for Energy-Efficient Li Recovery from LCT Pegmatites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Joanne Gamage McEvoy, Yves Thibault, Nail R. Zagrtdenov, and Dominique Duguay

81

Energy-Saving Green Technologies in the Mining and Mineral Processing Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shafiq Alam

89

Extraction of Valuable Metals from Luanshya Copper Smelting Slag with Minimal Waste Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yaki Chiyokoma Namiluko, Yotamu Rainford Stephen Hara, Agabu Shane, Makwenda Thelma Ngomba, Ireen Musukwa, Alexander Old, Ronald Hara, Rainford Hara, and Stephen Parirenyatwa

97

Carbon Footprint Assessment of Waste PCB Recycling Through Black Copper Smelting in Australia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 A. Q. Mairizal, A. Y. Sembada, K. M. Tse, N. Haque, and M. A. Rhamdhani Screening High-Entropy Alloys for Carbon Dioxide Reduction Reaction Using Alchemical Perturbation Density Functional Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Mohamed Hendy, Okan K. Orhan, Homin Shin, Ali Malek, and Mauricio Ponga Part III Thermal Management, Environmental and Energy Technologies Novel Thermal Conductivity Measurement Technique Utilizing a Transient Multilayer Analytical Model of a Line Heat Source Probe for Extreme Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Katelyn Wada, Austin Fleming, and David Estrada The Effect of Reduced Flue Gas Suctioning on Superstructure and Gas Temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Brandon Velasquez, Sarah DiBenedetto, Yonatan A. Tesfahunegn, Maria Gudjonsdottir, and Gudrun Saevarsdottir Assessing the Environmental Footprints of Gold Production in Nevada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Saeede Kadivar and Ehsan Vahidi

Contents

ix

Polymeric Composite Dense Membranes Applied for the Flue Gas Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Dragutin Nedeljkovic Molten Salt Mg-Air Battery Improvement and Recharging . . . . . . . . . . . . 171 Mahya Shahabi, Nicholas Masse, Amanda Lota, Lucien Wallace, Heath Bastow, and Adam Powell Superconductor Busbar Systems in the Light of Increased Energy Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Wolfgang Reiser, Till Reek, Claus Hanebeck, and Peter Abrell Critical Metals for Clean Energy: Extraction of Rare Earth Elements from Coal Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 Sara Penney and Shafiq Alam Part IV Energy Technologies Investigation of Slag and Condensate from the Charge Top in a FeSi75 Furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 M. B. Folstad, K. F. Jusnes, and M. Tangstad Lithium Extraction from Natural Resources to Meet the High Demand in EV and Energy Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 Valan Namq and Shafiq Alam Part V

Poster Session

Hydrogen Storage Properties of Graphitic Carbon Nitride Nanotube Synthesized by Mix-Grind Technique . . . . . . . . . . . . . . . . . . . . . . 223 Barton Arkhurst, Ruiran Guo, Ghazaleh Bahman Rokh, and Sammy Lap Ip Chan Study on Preparation and Electrocatalytic Performance of Self-supported Carbon Transition Metal Catalysts . . . . . . . . . . . . . . . . . 233 Ze Yang, Yanfang Huang, Guihong Han, Bingbing Liu, and Shengpeng Su Modification and Evaluation of Energy Saving and Consumption for Reduction Technology of 500 t/d Beckenbach Annular Lime Kiln . . . 243 Yapeng Zhang, Wen Pan, Zhenping Miao, Jianbo Zhu, Shaoguo Chen, Huaiying Ma, and Zhixing Zhao Research on the Gasification Characteristic of Cokes of BIOC-HPC Extracted from the Mixture of Low-Rank Coal and Biomass . . . . . . . . . . 251 Jun Zhao and Xueya Wang

x

Contents

Thermodynamic Examination of Selected Phases in the Ag–Co–Sn–S System at T < 600 K by the Solid-State EMF Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 Mykola Moroz, Fiseha Tesfaye, Pavlo Demchenko, Myroslava Prokhorenko, Oksana Mysina, Lyudmyla Soliak, Daniel Lindberg, Oleksandr Reshetnyak, and Leena Hupa Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275

About the Editors

Shafiq Alam is an Associate Professor at the University of Saskatchewan, Canada. He is an expert in the area of mining and mineral processing with profound experience in industrial operations, management, engineering, design, consulting, teaching, research, and professional services. As a productive researcher, Dr. Alam has secured 2 patents and has produced over 175 publications. He is the lead/co-editor of 13 books, and an editorial board member of two mining and mineral processing journals named Minerals (an Open Access Journal by MDPI) and the International Journal of Mining, Materials and Metallurgical Engineering. He is the winner of the 2015 Technology Award from the Extraction & Processing Division of The Minerals, Metals & Materials Society (TMS), USA. With extensive relevant industry experience as a registered professional engineer, Dr. Alam has worked on projects with many different mining industries. He is an Executive Committee Member of the Hydrometallurgy Section of the Canadian Institute of Mining, Metallurgy and Petroleum (CIM). During 2015–2017, he served as the Chair of the Hydrometallurgy and Electrometallurgy Committee of the Extraction & Processing Division (EPD) of TMS. Currently, he is the Secretary of the Recycling and Environmental Technologies Committee of TMS and is serving on the TMS-EPD Awards Committee. He is a lead/co-organizer of at least 17 symposia at international conferences through CIM and TMS. Dr. Alam is one of the founding organizers of the Rare Metal Extraction & Processing Symposium at TMS and since 2014, he has been involved xi

xii

About the Editors

in organizing this symposium every year with great success. In the past, he was involved in organizing the International Nickel-Cobalt 2013 Symposium and TMS 2017 Honorary Symposium on applications of Process Engineering Principles in Materials Processing, Energy and Environmental Technologies. Dr. Alam was also a co-organizer of the 9th International Symposium on Lead and Zinc Processing (PbZn 2020), and the 2022 Energy Technologies and CO2 Management Symposium. In addition to leading the 2023 Energy Technologies and CO2 Management Symposium, he is involved in co-organizing the 2023 Rare Metal Extraction and Processing Symposium, co-located with the TMS 2023 Annual Meeting and Exhibition in San Diego, California. He is also a co-organizer of the PbZn 2023 conference in China and the Pressure Hydrometallurgy 2023 conference in Toronto, Canada. Donna Post Guillen is the Group Lead for Modeling and Simulation in the Materials Science and Engineering Department at the Idaho National Laboratory. Dr. Guillen has over 35 years of research engineering experience and has served as principal investigator/technical lead for numerous multidisciplinary projects encompassing waste heat recovery, combustion, heat exchangers, power conversion systems, nuclear reactor fuels and materials experiments, waste vitrification, and advanced manufacturing. Her core area of expertise is computational modeling of energy systems, materials, and thermal fluid systems. She is experienced with X-ray and neutron beamline experiments, computational methods, tools and software for data analysis, visualization, application development, machine learning and informatics, numerical simulation, and design optimization. As Principal Investigator/Technical Lead for the DOE Nuclear Science User Facility Program, she has engaged in irradiation testing of new materials and performed thermal analysis for nuclear reactor experiments. She actively mentors students, serves in a leadership capacity as well as routinely chairs and organizes technical meetings for professional societies, provides subject matter reviews for proposals and technical manuscripts, has published over 100 papers and received three Best

About the Editors

xiii

Paper awards, authored numerous technical reports and journal articles, and has written/edited several books. Fiseha Tesfaye is a Process Metallurgist at Metso Outotec Metals Oy and Adjunct Professor in the Laboratory of Molecular Science and Engineering at Åbo Akademi University. He received his M.Sc. degree in materials processing technology and Ph.D. degree in metallurgy from Aalto University. With expertise in metallurgical thermodynamics, Dr. Tesfaye’s research activities are focused mainly on the thermodynamic characterization of inorganic materials as well as rigorous theoretical and experimental investigations for promoting sustainable production of metals and renewable energy. In 2018, Dr. Tesfaye was also appointed as a visiting research scientist in Seoul National University, South Korea. Dr. Tesfaye is an active member of The Minerals, Metals & Materials Society (TMS), and is a winner of the 2018 TMS Young Leaders Professional Development Award. He serves as a subject editor for different journals including JOM, the member journal of TMS, and Energies, MDPI. In addition, he has edited several scientific research books. His personal research achievements include remarkable improvement of experimental research applying the solid-state EMF technique for thermodynamic investigations of inorganic materials, as well as a noticeable contribution to promoting the transition toward the circular economy. In his research areas, Dr. Tesfaye has published over 75 peer-reviewed articles. Lei Zhang is an Associate Professor in the Department of Mechanical Engineering at the University of Alaska Fairbanks (UAF). Prior to joining UAF, she worked as a postdoctoral associate in the Department of Chemical and Biomolecular Engineering at the University of Pennsylvania. Dr. Zhang obtained her Ph.D. in Materials Science and Engineering from Michigan Technological University in 2011, and her M.S. and B.E. in Materials Science and Engineering from China University of Mining and Technology, Beijing, China, in 2008 and 2005, respectively. Her current research mainly focuses on the synthesis of metal-organic frameworks (MOFs) and MOF-based nanocomposites, and the manipulation

xiv

About the Editors

of their properties and applications in gas storage, separation, and water treatment. She is also working on the development and characterization of anti-corrosion coatings on metallic alloys for aerospace and biomedical applications. Dr. Zhang has served on the TMS Energy Committee since 2014, including the Vice-Chair role in 2018–2020 and the Chair role in 2020–2022. She also served on a Best Paper Award Subcommittee of the committee. She has served as a frequent organizer and session chair of TMS Annual Meeting symposia (2015– present). She was the recipient of the 2015 TMS Young Leaders Professional Development Award. Susanna A. C. Hockaday has 18 years of pyrometallurgical research experience in the non-ferrous industry. She joined Mintek in 2002 after obtaining her B. Chem. Eng. (Minerals Processing specialization) and M.Sc. in Extractive Metallurgy at the University of Stellenbosch in South Africa. During 2002 to 2010 she worked in the commercial projects group on various projects including the recovery of precious metals in liquid iron and the smelting of ores to produce design specifications of an industrial ferrochrome DC arc furnace. From 2011 to late 2015 she took a break from work and had two delightful children, now aged 11 and 8. From 2015 till 2021, she has been involved in research of new technologies for titanium metal production, chlorination of titanium dioxides in a fluidized bed, and the application of renewable energy in minerals processing. Since 2016 she has been enrolled at the University of Stellenbosch as a part-time Ph.D. student with the working title of “Solar thermal treatment of Manganese Ores”. She acted as a work package leader responsible for e2 million of research toward advancement of solar thermal process heating technology in manganese ferroalloy production for the PRÉMA project. The PRÉMA project is funded by the European Union’s Horizon 2020 Research and Innovation Programme under Grant Agreement No 820561 and is part of the SPIRE group of projects. Mrs. Hockaday was co-ordinator and the main author for Mintek’s Roadmap for Solar Thermal Applications in Minerals Processing (STAMP). Mrs. Hockaday is a member of the South African Institute of Mining and

About the Editors

xv

Metallurgy (SAIMM) and part of the SAIMM technical programme committee. She was also head of the organizing committee for the SAIMM Colloquium on Renewable Energy Solutions for Energy Intensive Industry in 2020 and the Renewable Energy Solutions for Energy Intensive Industry Conference in 2021. Mrs. Hockaday resigned from Mintek in June 2021 to move with her family to Perth, Australia. She has founded GamAesa as vehicle to continue her support of process innovation and renewable energy integrating in minerals processing. Neale R. Neelameggham IND LLC, is involved in international technology and management licensing for metals and chemicals, thiometallurgy, energy technologies, Agricoal, lithium-ion battery, energy efficient low cost OrangeH2, Netzero sooner with Maroon gas and Pink hydrogen, rare earth oxides, etc. He has more than 38 years of expertise in magnesium production and was involved in the process development of its startup company NL Magnesium to the present US Magnesium LLC, UT until 2011, during which he was instrumental in process development from the solar ponds to magnesium metal foundry. His expertise includes competitive magnesium processes worldwide and related trade cases. In 2016, Dr. Neelameggham and Brian Davis authored the ICE-JNME awardwinning paper “Twenty-First Century Global Anthropogenic Warming Convective Model.” He is working on Agricoal® to greening arid soils, and at present energy efficient Orange hydrogen, and methane abatement. He authored the ebook The Return of Manmade CO2 to Earth: Ecochemistry. Dr. Neelameggham holds 16 patents and applications and has published several technical papers. He has served in the Magnesium Committee of the TMS Light Metals Division (LMD) since its inception in 2000, chaired in 2005, and since 2007 has been a permanent advisor for the Magnesium Technology Symposium. He has been a member of the Reactive Metals Committee, Recycling Committee, Titanium Committee, and Program Committee for LMD and LMD council. Dr. Neelameggham was the Inaugural Chair, when in 2008, LMD and the TMS Extraction and Processing Division (EPD) created the Energy Committee and has been a Co-Editor of the

xvi

About the Editors

Energy Technology Symposium through the present. He received the LMD Distinguished Service Award in 2010. As Chair of the Hydrometallurgy and Electrometallurgy Committee, he initiated the Rare Metal Technology Symposium in 2014 and has been a co-organizer to the present. He organized the 2018 TMS Symposium on Stored Renewable Energy in Coal. Hong Peng is currently the Senior Research Fellow at the School of Chemical Engineering in the University of Queensland (UQ), Australia. He obtained a bachelor’s degree in Minerals Engineering and a master’s degree in Microbiology at Central South University, China followed by a Ph.D. in Chemical Engineering at UQ. Before joining UQ, Dr. Peng had the experience as a chemical engineer in its Newcastle Technology Research Center and Olympic Dam at BHP Billiton. He was the recipient of the 2020 TMS Young Leaders Professional Development Award. Dr. Peng’s research focuses on the fundamental aspects of mineral processing, interfacial colloid science, crystal kinetics, and precipitation as well as molecular dynamics simulation. These projects are of interest to the nanobubbles, mine tailings, zeolite, clay minerals, and metal resource recovery. Nawshad Haque is a Principal Research Scientist at the Commonwealth Scientific and Industrial Research Organisation (CSIRO) Australia and leads the Electrochemical Energy Systems Team in the Energy Technologies Program. He manages a team of around 13 researchers, 2 post-docs, and 8 Ph.D. students. The Team is developing low-cost electrolyzers and other associated technologies for hydrogen energy systems. Dr. Haque has over 20 years R&D experience since having a Ph.D. in Chemical Engineering from the University of Sydney. He has played a key role in the development of CSIRO’s technology evaluation capabilities including flowsheeting, techno-economic and life cycle assessment methodologies using various databases, tools, and software to aid in decision. Dr. Haque has a strong interest in the evaluation of technologies for decarbonization of the energy, mining, mineral processing, and metallurgical industries. In his current role, Dr.

About the Editors

xvii

Haque has initiated several large international multiparty collaborative research relationships with governments and industries leading to the establishment of multi-million-dollar projects related to various technologies for science capacity development. Dr. Haque is the Leader for the CSIRO BCSIR Bangladesh and RMIT University long-term collaboration program for scientific research capability development in Bangladesh. Dr. Haque has co-authored more than 100 publications attracting numerous citations. Dr. Haque is currently an elected Fellow of the Australian Institute of Energy, Australasian Institute of Mining and Metallurgy, and a member of TMS. He has assisted several journals including Minerals, Journal of Cleaner Production, Drying Technology, and is a regular reviewer of International Journal of Hydrogen Energy. He has served as a Director of the Board of Australian Life Cycle Assessment Society for 10 years. Dr Haque has co-supervised CSIRO sponsored 11 Ph.D. students and over 54 vacation scholarship projects to completion, many on technoeconomic and life cycle assessment. He is an adjunct academic at Swinburne University, Monash University, and RMIT University in Australia. Yan Liu is currently a professor at the School of Metallurgy at Northeastern University, China. She obtained her bachelor’s degree, master’s degree, and doctorate in Thermal Energy and Dynamic Engineering at Northeastern University, China. She is mainly engaged in the fields of nonferrous metallurgy, metallurgical reaction engineering, refinement and dispersion of bubbles in chemical metallurgy process, design of chemical metallurgy reactor, and physical and numerical simulation of reactor. Also, she has built a physical and numerical simulation platform. She has done a lot of effective work in CO2 capture and reactor simulation. Dr. Liu has participated in the TMS Annual Meeting almost every year for decades, and her team has won several awards in TMS conference papers.

Part I

Renewable Energy and Combustion Technologies

Analysis of Environmental Impact of Vertical Axis Wind Turbine Using Circular Economy Approach Satyendra Dayalu, Shalini Verma, Akshoy Ranjan Paul, and Nawshad Haque

Abstract The need for energy is constantly growing due to economic, population growth, and technological advancement. In India, coal is the largest contributor with about 75% in electricity generation in 2019, but coal combustion produces higher emissions among the fossil fuels and other non-renewable energy sources. Thus, the contribution of renewables (especially wind) is gaining importance in recent years due to its easy availability and low carbon emission. The wind turbine does not impact the environment in its operational phase; however, raw material extraction, production, transportation, and decommissioning affect the environment through harmful emissions. This study aims to integrate the circular economy in the life cycle of Vertical Axis Wind Turbine (VAWT). The study presents the assessment of environmental impact for baseline case and circular economy scenario (recycling) of materials used in VAWT production. The result suggests that aluminium has a higher contribution to environmental impact for each impact category CO2 , NOx , SO2 , and PM2.5 . Among different impact categories, global warming potential highly impacts the environment. The capacity factor is a key parameter in reducing the impact on the environment using a VAWT. The recycling scenario of 50% and 90% reduces the environmental impact by 25.7% and 46.3%, respectively. Thus, the integration of circular economy with VAWT is likely to be a sustainable transition with reduced emissions. Keywords Circular economy · Renewable energy · Vertical axis wind turbine · Environment impact · Pollutant emissions S. Dayalu · S. Verma (B) · A. R. Paul (B) Department of Applied Mechanics, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, Uttar Pradesh 211004, India e-mail: [email protected] A. R. Paul e-mail: [email protected] N. Haque Commonwealth Scientific and Industrial Research Organization (CSIRO) Energy, Private Bag 10, Clayton South, VIC 3169, Australia e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_1

3

4

S. Dayalu et al.

Abbreviations VAWT HAWT CE EoL LCA CF

Vertical axis wind turbine Horizontal axis wind turbine Circular economy End-of-life Life cycle assessment Capacity factor

Introduction The world population is increasing, and consequently, the energy demand is on the rise. Generally, non-renewable energy resources, i.e., fossil fuels, are used for electricity generation but it adversely affects the environment and also resources are limited in nature. In order to avoid environmental impact and scarcity, there are numerous alternative sources of energy like wind energy, solar Energy, geothermal Energy, hydro-power, ocean energy, and biomass energy. Among these, India stands at 4th rank and 5th rank in the sector of wind Energy and solar energy, respectively, in the world in terms of capacity [1]. India has set a target of 175 GW capacity from all types of renewable energy sources by 2022 in which onshore and offshore wind power generation capacity is 60 GW and 5 GW respectively [2]. The next goal for India is even more ambitious, as the target of 450 GW by 2030 in which 140 GW would come solely from wind-based generation [3]. According to the Global Wind Report 2022, India is slow to meet the objectives of net zero emissions by 2050, for wind energy alone we need to install four times more than the current capacity of wind power generation to remain on a realizable net-zero path [5]. Wind turbines are used to harvest wind energy and are categorised based on the orientation of the axis of the rotor, thus deriving its name as ‘Horizontal Axis’ Wind Turbine (HAWT), which are very popular in large-scale projects and ‘Vertical Axis’ Wind Turbine (VAWT) which are currently limited in application. The popularity of HAWT in large-scale projects is an undisputed fact but varied factors are responsible for the surge in popularity of VAWT which include robust mechanical and compact design leading to building integration on small-scale projects and requiring lesser maintenance, omni-directional nature of rotor, well suited for the built-in environment and lower cut-in speed. It is a fact that wind energy is a renewable source of energy but this does not imply that it has a ‘zero’ carbon footprint. Life cycle assessment (LCA) as a tool for analysis has been performed to measure the environmental impact of HAWT, installed over different locations [6–9]. A few researchers have explored the LCA for VAWTs. Rashedi et al. [10] presented a life cycle assessment of VAWT along with onshore and offshore HAWT. The result suggested that VAWT is more sustainable compared to HAWT. This study also found that copper is the highest contributor to environmental impact and replacing

Analysis of Environmental Impact of Vertical Axis Wind Turbine Using …

5

it with aluminium can reduce the environmental impact by 7% for modified VAWT. Thus, an optimal material selection is necessary for a sustainable environment [11] performed LCA by considering it in terms of embodied energy and environmental impact including emissions for roof-mounted VAWT and HAWT located at different locations in Thailand. The results found that reuse scenarios and proper material selection could reduce the embodied energy and environmental impact, [12] analysed the environmental impact of 1 and 3 kW VAWTs installed in the Italian region. It is observed from the result that the wind parameter influences the environmental impact and the highest impact contribution is from the manufacturing phase compared to maintenance and transportation. The LCA analysis performed by [13] suggested that appropriate site selection, recycling of the metals, and roof-mounted VAWTs could render VAWT an environmental friendly technology. The literature review found that the components of the wind turbine are manufactured from different materials which affect the environment during decommissioning of the wind turbine; however, it does not cause any emission during electricity production. The circular economy (CE) idea aims to keep resources, goods, and services in use for as long as needed. A circular economy minimises resources used, remanufactures the products, and recycles ‘waste’ as a resource to create new products. Recycling materials can be a significant factor in reducing the rate of climate change. The CE can advance social justice, enhance economics, and save the environment when it is thoughtfully and broadly constructed. As mentioned, in the technical cycle, products are tracked on 4R (reuse, repair, remanufacture, and recycle) steps. The CE is used in different applications of engineering like heat pumps and gas boilers [14] and biomass [15]. The circular process employing LCA for the building sector and renewable energy systems was examined by [16]. The outcome shows the growing necessity for using circular processes for recycle/reuse materials in both the building and energy sectors. Ralph et al. [17] describe the design standard of renewable energy technology associated with CE approaches and equilibrium between design for internationalized technology along with EOL (End of Life) products. CE inherently closes the loop between energy expense and used resources. A comparison of various established methods and cordiality criteria had been done to identify gaps in CE strategies, the research of this paper is restricted by its exploratory and conceptual approach substituting empirical evidence together with renewable energy designers. Hao et al. [18] used CE for the recovery of Carbon Fibre Reinforced Polymer (CFRP) from EOL blades using pyrolysis. Jensen and Skelton [19] explained about challenges in CE and different possibilities of recycle and recovery. They suggested that materials from end-of-life blades can be used for various secondary applications. Jensen [20] suggested that the materials used to make the turbines account for 70 to 80% of the environmental impact. The amount of energy saved is expected to be around 81 TJ. Approximately 7351 tonnes of CO2 have been saved due to the recycling of wind turbine parts. Rentizelas et al. [21] contributed towards exploring the feasibility of mechanical recycling in CE for reusing blades. Therefore, the idea of the CE can be effectively used to reduce net carbon emissions, although limited literature is reported on the CE for VAWTs. The built-up region in India has increased by 176.3% between 1975 and 2014. Thus, installation

6

S. Dayalu et al.

of VAWT in rural/urban areas and highways has increased currently. The utilisation of the CE strategy for VAWT is a novel aspect of the research. The paper’s goal is to assess the environmental impact of using LCA. The paper emphasizes the effects of recycling the VAWT resources and reuse, remanufacture concepts are omitted. This paper discusses the methodology for LCA, followed by life cycle inventory and impact analysis using recycling. Finally, the paper concludes in the further section.

Methodology In this study, the environmental impact assessment is performed through the CE approach in which recycling is considered. The environmental impact assessment is done using LCA. The LCA is executed using the SimaPro software. In this analysis, we first establish the purpose and scope of the study, assemble the life cycle inventory (LCI), and then carry out the life cycle impact assessment following the LCA framework as mentioned in ISO 14040 guidelines International Standard Organization, [22].

Objective and Scope Definition The objective of the study is to assess the environmental impact of VAWT using a CE method. The VAWT is H-Darrieus five-bladed VAWT with wheel dia. 1.38 m, turbine height 1 m, and capacity 500 W. This VAWT is manufactured by IYSERT in India [23]. The components of this VAWT are the blade, hub, tower, connection rod, and shaft. The functional unit and system boundary definition are also included in this phase: (i) Functional unit Electricity production is the purpose of a wind power system. The functional unit is 1 kWh of electricity generated by each energy system throughout its intended life-times of services. The research is carried out under the assumption that VAWT has a 20-year design lifespan as standard. (ii) System boundary It consists of an assessment of wind turbine manufacturing steps like extraction of raw material, cleaning, and segregation, production, transportation, installation, maintenance, decommissioning, and disposal are all considered for the evaluation. Figure 1 displays the system boundary of VAWT. At first, raw materials are generated or extracted, then moved to the location where wind turbines are manufactured. In between cleaning process would be done by using mechanical, chemical, and thermal stages. At the production site, materials that have been transported are used to create parts of wind turbines like the rotor hub,

Analysis of Environmental Impact of Vertical Axis Wind Turbine Using …

7

blade, and tower. Pollutants are emitted, and energy is used up throughout this production phase. To the location of installation, the manufactured pieces are transported. The wind turbine’s foundation is built at the installation site, then it is put into place. Additionally, the stages of operation and maintenance involve the production of electricity, routine maintenance, and replacement of energyintensive equipment. After the wind turbine has served its purpose and is used to its EOL, it is decommissioned. The end-of-life scenario is then categorised into four parts such as recycle, reuse, remanufacture, and landfill. Changes in how we produce, use, consume, (and reuse), and manage waste are therefore necessary as we move toward CE. CE practices include recycling and recovery. This is the step where used materials are treated to reuse them in another system. Remanufacturing is essential where exhausted/spoiled, or EOL products are brought in, and product life extension involves an engineering step of refurbishing products for an extended life-time. Implicitly, these practices are dependent upon technological advances and variations in processing. From the viewpoint of CE, it is important to close the material loop to achieve a satisfactory outcome. This is largely dependent on the quality of the recovered material and the process used.

Fig. 1 System boundary of VAWT including circular approach

8

S. Dayalu et al.

The CE approach of closing the system boundary loop is followed by transporting the recyclable and remanufactured components to the location of manufacturing, reusable parts to the construction site, and throwaway parts to the landfill.

Life Cycle Inventory The material data needed to estimate the emissions from a VAWT is collected from the references. The net weight of the selected VAWT is 25 kg [23]. A H-Darrieus type VAWT is presented in Fig. 2, and the components are blades, hub, connection rod, shaft, and tower. The most used material in a VAWT is aluminium and steel and other materials have a less share. In a wind turbine, the steel percentage is used as 66–79% [24]. The quantity of the materials used in the manufacturing of these components is presented in Table 1, and it presents the specification and material used in a VAWT. Steel and aluminium are only used for the study as it has a higher share contribution among other materials.

Fig. 2 Components and materials contribution in H-Darrieus VAWT

Table 1 Specification and material used in a VAWT Parameter

Value

Materials

Quantity (kg)

Generated power

500 W

Steel for the hub, tower, shaft, connection rod

18.75

Aluminium for blade

6

Analysis of Environmental Impact of Vertical Axis Wind Turbine Using …

9

Assumptions The life cycle emissions are calculated based on the following assumptions: • In this study, the materials like copper having lower contribution have been neglected. The lubricant and permanent magnets are neglected in the study due to less share and negligible effect. • This study only includes the wind turbine and neglects the inverter and cables. The electricity losses are assumed negligible. • The life-time of VAWT is assumed as 20 years. • The emissions and costs associated with decommissioning (e.g., removing and transporting) old turbine components to a recycling plant are not taken into account.

Results and Discussion Emission Intensity Based on the life cycle inventory, the results for the life cycle impact assessment are calculated in SimaPro. The emission intensity is denoted by kg eq./kWh. The electricity generation for a 20-year life-time is calculated by Eqs. (1) and (2) and considering 20.5% capacity factor (CF) for the Rajasthan region [25]. AEP = C × CF × 24 × 365

(1)

E lifetime = N × AEP

(2)

where AEP = Annual energy Production in kWh, C = Capacity in kW, CF = Capacity factor (20.5% for Rajasthan), E life-time = Electricity generation for life cycle of VAWT, N = Life-time of VAWT, in years. The emission intensities are presented in Table 2. The impact categories for the analysis include global warming potential (CO2 ), acidification potential (SO2 ), photochemical oxidant potential (NOx ), and particulate matter (PM2.5 ). Among different impact categories, global warming potential has a higher contribution to environmental impact and it quantifies as 0.00952 kg CO2 -eq./kWh. Aluminium has a higher contribution to environmental impact about 75.7%, 72.9%, 82.1%, and 82.2% for CO2 , NOX , SO2 , and PM2.5 respectively as shown in Fig. 3.

10

S. Dayalu et al.

Table 2 Total life cycle environmental impact of VAWT per unit of electricity generation for a 20-year life cycle Emissions (× 10−3 kg eq./kWh)

CO2

9.52

NOX

0.029

SO2

0.046

PM2.5

0.014

% Contribution

Pollutant Emissions

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% CO₂

NOₓ

SO₂

PM₂.₅

Impact Categories Al

Steel

Fig. 3 Contribution of materials to different environmental emissions

Effect of the Capacity Factor on the Environment The CF for wind turbines varies between 20 and 40% [26]. In this study, variation of capacity factor is also explored to find out the environmental impact. The study is performed for 20.5 and 40% CF. Thus, for both scenarios, the environmental impact is presented in Fig. 4 and it represents that the CF has a greater influence on the environmental impact. The emission has increased by 48.75% increasing CF from 20.5 to 40%.

Effect of Recycling Scenario on Emissions The environmental impact is estimated in terms of CO2 , NOx , SO2 , and PM2.5 . At the end-of-life stage of VAWT, recyclable materials are circulated for recycling. In the current study, 50 and 90% recycling capacity are considered. Figure 5 represents that CO2 emission highly impacts the environment and there is a reduction of 25.7% and 46.3% CO2 -eq emissions from a VAWT for 50% and 90% recycling of the materials respectively. The recycling scenario considers the electricity mix for aluminium

Total Emission (×10 3 kg-eq/kWh)

Analysis of Environmental Impact of Vertical Axis Wind Turbine Using …

11

0.05

10 9 8 7 6 5 4 3 2 1 0

0.04 0.03 0.02 0.01 0 NOₓ

SO

NOₓ

CO

PM . PM .

SO

Pollutant Emissions CF=40%

CF=20.5%

Fig. 4 Environmental impact on the variation of capacity factor

Total Emission (×10 3 kg-eq/kWh)

smelting but omits the transportation due to data unavailability. The eco-system for each unit of electricity generation through VAWT is 0.00952 kg CO2 -eq.; however, after recycling this is 0.00707 kg CO2 -eq. and 0.00511 kg CO2 -eq. 10 9 8 7 6 5 4 3 2 1 0

0.05 0.04 0.03 0.02 0.01 0 NOₓ

CO

NOₓ

SO

SO

PM .

PM .

Pollutant Emissions Without Recycling

50% Recycling

Fig. 5 Emissions from VAWT for recycling scenario

90% recycling

12

S. Dayalu et al.

Conclusions This study mainly recommends careful consideration of the implementation of VAWT as an environmentally safe technology for generating power. This LCA study concludes that the recycling of the material might improve VAWT more environmentally beneficial. The CE approach using LCA is used to assess the environmental impact of the specific model of 500 W VAWT as the functional unit of 1 kWh electricity sent out. In this approach, recycling (50 and 90%) is considered for this study. Electricity generation is estimated using 20.5 and 40% CF. Aluminium has a higher environmental impact compared to steel for all the selected impact categories. The exploration of the capacity factor concludes that environmental degradation increases on declining CF. The recycling approach reduces the environmental impact of VAWT by 25.7% and 46.3% for 50% and 90% recycling respectively. The future study suggests for analysis of the environmental impact of higher capacity and improved technology VAWT as well as techno-economic analysis.

References 1. MNRE (2018) Year end review 2018. Ministry of New and Renewable Energy. Available on: https://pib.gov.in/PressReleaseIframePage.aspx?PRID=1555373. Accessed 23 Aug 2022 2. NITI AYOG, Report of the Expert Group on 175 GW RE by 2022. National Institution for Transforming India. Available on: https://www.niti.gov.in/sites/default/files/energy/175-GWRenewable-Energy.pdf 3. 2030 Renewable Energy Target: Panel to be set up soon for Mission 500GW’ COP 26 Summit at Glasgow. Available on: https://economictimes.indiatimes.com/industry/renewables/2030renewable-energy-target-panel-to-be-set-up-soon-for-mission500gw/articleshow/88267104. cms?utm_source=contentofinterest&utm_medium=text&utm_campaign=cppst. Accessed 23 Aug 2022 4. Global Wind Report (2022) Available on: https://gwec.net/wp-content/uploads/2022/04/Ann ual-Wind-Report-2022_screen_final_April.pdf. Accessed 23 Aug 2022 5. Global Wind Report (2022). https://gwec.net/wp-content/uploads/2022/03/GWEC-GLOBALWIND-REPORT-2022.pdf 6. Verma S, Paul AR, Haque N (2022) Selected environmental impact indicators assessment of wind energy in India using a life cycle assessment. Energies 15:3944. https://doi.org/10.3390/ en15113944 7. Schreiber A, Marx J, Zapp P (2019) Comparative life cycle assessment of electricity generation by different wind turbine types. J Clean Prod 233:561–572. https://doi.org/10.1016/j.jclepro. 2019.06.058 8. Gomaa MR, Rezk H, Mustafa RJ, Al-Dhaifallah M (2019) Evaluating the environmental impacts and energy performance of a wind farm system utilizing the life-cycle assessment method: a practical case study. Energies 12:3263. https://doi.org/10.3390/en12173263 9. Gkantou M, Rebelo C, Baniotopoulos C (2020) Life cycle assessment of tall onshore hybrid steel wind turbine towers. Energies 13:3950. https://doi.org/10.3390/en13153950 10. Rashedi A, Sridhar I, Tseng KJ (2013) Life cycle assessment of 50 MW wind firms and strategies for impact reduction. Renew Sustain Energy Rev 21:89–101. https://doi.org/10.1016/ j.rser.2012.12.045

Analysis of Environmental Impact of Vertical Axis Wind Turbine Using …

13

11. Uddin MS, Kumar S (2014) Energy, emissions and environmental impact analysis of wind turbine using life cycle assessment technique. J Clean Prod 69:153–164. https://doi.org/10. 1016/j.jclepro.2014.01.073 12. Lombardi L, Mendecka B, Carnevale E, Stanek W (2018) Environmental impacts of electricity production of micro wind turbines with vertical axis. Renew Energy 128:553–564. https://doi. org/10.1016/j.renene.2017.07.010 13. Kouloumpis V, Sobolewski RA, Yan X (2020) Performance and life cycle assessment of a small scale vertical axis wind turbine. J Clean Prod 247:119520. https://doi.org/10.1016/j.jcl epro.2019.119520 14. Sevindik S, Spataru C, Domenech AT, Bleischwitz R (2021) A comparative environmental assessment of heat pumps and gas boilers towards a circular economy in the UK. Energies 14(11):3027. https://doi.org/10.3390/en14113027 15. Teigiserova DA, Hamelin L, Tiruta-Barna L, Ahmadi A, Thomsen M (2022) Circular /bio economy: life cycle assessment of scaled-up cascading production from orange peel waste under current and future electricity mixes. Sci Total Environ 812:152574. https://doi.org/10. 1016/j.scitotenv.2021.152574 16. Papadaki D, Nikolaou DA, Assimakopoulos MN (2022) Circular environmental impact of recycled building materials and residential renewable energy. Sustainability 14(7):4039. https:// doi.org/10.3390/su14074039 17. Ralph N (2021) A conceptual merging of circular economy, regrowth and conviviality design approaches applied to renewable energy technology. J Clean Prod 319:128549. https://doi.org/ 10.1016/j.jclepro.2021.128549 18. Hao S, Kuah AT, Rudd CD, Wong KH, Lai NYG, Mao J, Liu X (2020) A circular economy approach to green energy: wind turbine, waste, and material recovery. Sci Total Environ 702:135054. https://doi.org/10.1016/j.scitotenv.2019.135054 19. Jensen JP, Skelton K (2018) Wind turbine blade recycling: experiences, challenges and possibilities in a circular economy. Renew Sustain Energy Rev 97:165–176. https://doi.org/10.1016/ j.rser.2018.08.041 20. Jensen JP (2019) Evaluating the environmental impacts of recycling wind turbines. Wind Energy 22(2):316–326. https://doi.org/10.1002/we.2287 21. Rentizelas A, Trivyza N, Oswald S, Siegl S (2022) Reverse supply network design for circular economy pathways of wind turbine blades in Europe. Int J Prod Res 60(6):1795–1814. https:// doi.org/10.1080/00207543.2020.1870016 22. ISO 14044 (2006) Environmental management—life cycle assessment—requirements and guidelines. International Organisation for Standardisation (ISO), Geneva, Switzerland 23. IYSERT available on: https://iysertenergy.com/IYSERT%20VERTICAL%20AXIS%20W IND%20TURBINE.pdf 24. https://www.usgs.gov/faqs/what-materials-are-used-make-windturbines#:~:text=According% 20to%20a%20report%20from,aluminum%20(0%2D2%25). https://www.nrel.gov/docs/fy1 7osti/66861.pdf 25. ICF International (2014) Capacity value of wind generation in India—an assessment. ICF, Fairfax. VA, USA 26. Power W (2006) Capacity factor, intermittency, and what happens when the wind doesn’t blow. University of Massachusetts, Renewable Energy Research Laboratory

Corrosion and Erosion Protection to Accelerate Deployment of Sustainable Biomass Patrick Shower, Scott Weaver, Voramon Dheeradhada, Aida Amroussia, Michael Pagan, Patrick Brennan, Martin Morra, Bruce Pint, Suresh Babu, Phil Gilston, Steve Lombardo, Tamara Russell, and Anteneh Kebbede

Abstract Thermochemical processing of sustainable biomass paired with carbon capture and storage has the potential to provide 1/7th of the emission mitigations necessary for the world to meet net-zero targets by 2050 while also providing carbon-negative fuel, electricity, and economic development. This work has developed coating solutions to mitigate the hot corrosion that occurs when biomass is processed in boilers, gasifiers, and other thermal conversion equipment, in addition to coating solutions to mitigate solid particle erosion that occurs when steam turbines are used for aggressive load following. Analysis of 66 hot corrosion coatings and 75 solid particle erosion coatings reveals unique mechanisms that enable significantly improved performance relative to conventional coatings used today without increasing material cost. Implications of this work to fuel flexibility, process efficiency, and lessons learned utilizing ICME will also be discussed. This material is based upon work supported by the Department of Energy under Award Number DE-FE0031911. Keywords High temperature corrosion · Solid particle erosion · Coating · Biomass

P. Shower (B) · S. Weaver · V. Dheeradhada · A. Amroussia · P. Brennan · M. Morra · A. Kebbede General Electric Global Research, Niskayuna, NY, USA e-mail: [email protected] M. Pagan · S. Babu University of Tennessee, Knoxville, TN, USA B. Pint Oak Ridge National Laboratory, Oak Ridge, TN, USA P. Gilston · S. Lombardo General Electric Power Portfolio, Windsor, CT, USA T. Russell General Electric Gas Power, Schenectady, NY, USA © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_2

15

16

P. Shower et al.

Introduction Based on outage data and feedback provided by GE customers, the reliability of thermal power plants in the current fleet is primarily determined by the onset of leakage in boiler tubing (especially the superheater and reheater sections) and the accumulation of solid particle erosion damage in the HP turbine (especially on inlet nozzles and early stage blades). Coatings for both boiler tubing and HP turbine blades have been developed, but are not able to meet the fiscal and technical requirements of the current fleet [1]. In order to provide technical context to this work, the operational challenges being solved by these novel corrosion and erosion coatings are expanded upon, along with the potential benefit that each technology represents.

State of the Art in Boiler Tubing Performance and Weld Overlay of Alloy 72 In the field, stress corrosion cracking and intergranular corrosion at the outer diameter of austenitic stainless steel tubing lead to leakage and limits time between scheduled boiler outages to 18–24 months [1]. This is aggravated by thermal cycling in partial loading conditions. The current state of the art for enhancing the reliability of superheater and reheater tubing in boilers is the application of Alloy 72 (Ni–Cr alloy) via weld overlay. This coating extends the life of tubing components by the formation of a chrome oxide protective layer but is only used in limited amounts due to its cost, which is driven by the cost of Ni-based feedstock [1]. To coat boiler tubing with Alloy 72 costs approximately 1.5 times what the tubing itself costs, meaning that coated tubing is 2.5 times as expensive as uncoated tubing. In practical terms, Alloy 72 has not mitigated the problem of tubing leakage. This project will deliver a smarter, more effective coating solution with multiple protection mechanisms rather than the single mechanism used today.

Potential Benefits of FeCrAlY-Derived Weld Overlay for Boiler Tubing Recently, FeCrAlY alloys capable of withstanding nuclear accident conditions have been developed at GE Research (GER) and show immediate chemical passivation and minimal weight change under high temperature oxidizing conditions [1, 2]. In addition to a chrome oxide layer which is stable to 1100 °C, the FeCrAlY coating will be capable of forming aluminum oxide protective layer which is stable to 1300 °C [3–8]. Work by Pint et al. as well as Special Metals has shown that the addition of Al to Cr-containing alloys like Alloy 72 results in enhanced resistance to corrosive attack in fireside conditions (as well as steam) [9, 10].

Corrosion and Erosion Protection to Accelerate Deployment …

17

Table 1 Comparison of Alloy 72 with FeCrAlY Critical to quality characteristic

Alloy 72 (NiCr)

Advanced weld overlay (FeCrAlY)

Cost of primary element

13.05 $/kg [11]

0.24 $/kg [11]

Maximum temperature for oxide stability in steam

1100 °C [3]

1300 °C [3]

Scaling temperature in air

1150 °C [12]

1425 °C [7]

However, there are technical risks involved with attempting to transfer to oxidization and corrosion performance of FeCrAlY alloys to tubing via a weld overlay. These include a chemistry shift during welding, defect formation, or susceptibility of attack to alkaline deposits, chlorine, or other chemical agents on which these alloys have not yet been tested. Potential secondary benefits of developing a low- or no-Ni overlay include cost effectiveness and the potential for greater compatibility with austenitic or ferritic steel tubing in terms of CTE and general chemical compatibility. These factors may allow a low- to no-Ni overlay be applied to a greater surface area of tubing for the same cost. A comparison of Alloy 72 with FeCrAlY is shown in Table 1.

State of the Art for High Pressure Steam Turbine Blades and Erosion Resistant Coatings High pressure (HP) steam turbine nozzles and blades accumulate structural damage and require an outage approximately every 8 years, which necessitates a complete shutdown of the plant. It would be beneficial to reduce the rate of solid particle erosion damage and increase steam inlet temperature towards supercritical steam capacity (though the limitations of other steam path components would also need to be addressed before this goal is realized) [1]. Today, it is common to apply a titanium nitride (TiN) coating is applied via cathodic arc deposition. However, experience in the field has shown that TiN is not adequate to prevent solid particle erosion and is often worn through upon inspection. The alternative approach to mitigate erosion is to apply a protective cermet via thermal spray. Generally, such coatings adequately resist erosion, but are much thicker than cathodic arc coating and have increased surface roughness, therefore leading to increased turbulence in the HP turbine and imbuing a performance debit [1, 8].

18

P. Shower et al.

Potential Benefits of Layered IPD Coatings for HP Turbine Blades Based on the performance needs of HP steam turbines, it would be beneficial to apply a coating that can adequately resist solid particle erosion, produce a passivating oxide layer in steam at increased inlet steam temperatures above, and not interfere with the aerodynamic performance of the blades. It is difficult to identify a single composition that could achieve all three objectives, but it may be possible to combine the beneficial properties from multiple compositions. For instance, a Cr-containing composition that can resist oxidation attack in steam up to 1100 °C [3, 4, 7] and a high strength ceramic [1] (or intermetallic [13]) with an erosion resistance greater than TiN, applied in a very conformal coating via IPD. Recent work has shown that by depositing alternating layers of Cr with a hard material, one can in fact achieve a coating with the best properties of both compositions. Brachet et al. [13] demonstrated that while a Cr coating could reduce the rate mass change due to oxidation in steam by 88% relative to the baseline, a LIPD coating could reduce the mass change rate by 84%, retaining almost all the benefits of the Cr layer while adding the erosion resistance of the hard layer and the inherent toughness of a laminar structure.

Non-coating Approaches to Mitigate Erosion and Corrosion Any novel coating technologies for erosion and corrosion resistance will be competing with existing technologies. The value proposition with relative to existing technologies is highlighted in Sections “Corrosion Coating Results” and “Erosion Coating Results”. To date, there is no competing surface modification technology that combines cost reduction and reliability improvement targeted for boiler tubing or the combination of erosion and steam oxidation resistance targeted for HP turbine blades. Non-surface technologies to improve the reliability of HP turbine blades include geometric optimization (which has been largely achieved) and installation of bypass infrastructure to remove solid particles from the steam path before it enters the HP Steam Turbine (used in Europe but costly to implement in the US). Sandvik offers composite tubing with a load bearing inner layer and a corrosion resistant outer layer. However, corrosion resistance is based on Cr content (25%) alone, and production of the tubing requires additional, costly steps of co-extrusion, cold pilgering, and non-destructive testing [14].

Corrosion Coating Results This program aims to design a drop-in wire feedstock composition for weld overlay coatings that demonstrates low raw material cost while maintaining the weldability

Corrosion and Erosion Protection to Accelerate Deployment …

19

and corrosion resistance in biomass combustion environments as conventional alloys like “FM72 (referred to hereafter as Alloy 72)” and Inconel 625. Cost reduction is achieved principally by eliminating Ni and Co. Weldability is evaluated using physics-based models of solidification as well as ASME Section IX transverse bend testing. Corrosion resistance is measured in a laboratory setting with a controlled temperature, time, process gas, and synthetic biomass ash based on the superheater of a thermal utility plant. A total of 66 alloy compositions have been cast. These include 22 commercial alloys utilized for their corrosion resistance from the Co-based, Ni-based, and Austenitic Stainless Steel alloy families. Eleven compositions are “model alloys” with incremental amounts of Fe, Ni, Cr, and Al to benefit the team’s mechanistic understanding. Nineteen compositions were designed based on initial literature review, computational thermodynamics, and prior work of team members at GE, UTK, and ORNL. Eighteen alloys were designed once previous samples had been analyzed with machine learning and advanced characterization, 4 of which were directly suggested by machine learning algorithms as compositions that would optimize between hot corrosion resistance and the hot cracking index used to computationally screen for weldability. These alloys have been tested in a laboratory hot corrosion environment and subsequently analyzed for microstructural evolution and metal loss. Metal loss results are measured as the greatest depth of damage to the sample from internal or external oxidation or sulfidation. In terms of field service, this is the amount of material one would have to remove to get down to “white metal” as is typically called for during an outage. The cost of the alloy in terms of raw material feedstock was also estimated by simply using a weighted sum of elemental costs. Figure 1 shows that alloys developed on this program (denoted “Gen 2” and “Gen 3” alloys) were able to achieve both reduced corrosion rate and reduced material costs relative to commercially available alloys. The most successful alloy demonstrated a 76% reduction in raw material cost and a 98% reduction in metal loss as compared to Alloy 72 over a 500 h period at 700 °C.

Corrosion Coating Discussion Mechanisms that are correlated to the improved performance observed in labscale testing include multi-level oxides for external corrosion protection, gettering elements for internal oxidation/sulfidation protection, and phases that will be detrimental for weldability if they appear at scale.

20

P. Shower et al.

Fig. 1 A summary plot of the performance of alloys developed on this program. Alloys plotted closer to the origin have lower raw material cost and lower corrosion rate and are therefore considered more successful in achieving the program’s goals. “Weld development needed” was taken from quantitative Schiele simulations that predicted the risk of hot cracking upon welding

Multi-Level Oxides The majority of candidate compositions that outperformed Alloy 72 demonstrated a complex oxide structure. Whereas the internal oxide and non-protective external oxide of Alloy 72 are composed almost entirely of Chromia (Fig. 2), the protective oxide of most alloys designed on this program contains multiple and complimentary layers (Fig. 3).

Gettering of Internal Oxygen and Sulfur Another mechanism that is common to most of the best-performing alloys examined in the program is their response to oxygen and sulfur ingress. Such ingress is an enabling aspect of hot corrosion propagation, so arresting it near the metal surface is beneficial. This is achieved by introducing alloying elements that are more reactive to oxygen and sulfur than chrome and are also more thermodynamically stable once oxidized/sulfurized (Fig. 4). Commercial alloys such as Alloy 72 and Inconel 622 tend to form internal chrome sulfides. These particles both reduce the

Corrosion and Erosion Protection to Accelerate Deployment …

Internal Oxidaon Aack Connues ↓

21

Chrome

Fig. 2 Internal and external oxidation in Alloy 72 comprises non-protective Chromia. Images (backscatter electron micrograph on left and energy dispersion spectroscopy image on right) are at the same scale as the micrograph in Fig. 3

Fig. 3 A representative multi-layer oxide in a candidate alloy that has proven protective in hot corrosion testing at 700 °C. The backscatter electron micrograph (left) is at the same scale as the images in Fig. 2 to highlight how compact it is by comparison

22

P. Shower et al.

Fig. 4 Example of internal oxidation and sulfurization mitigated by non-Cr getter elements in candidate alloy

amount of chrome available to form a protective external oxide and readily release their contained sulfur as the corrosion front advances, thereby continuing the cyclic mechanism of hot corrosion attack.

Hot Cracking Although the representative candidate alloys that were tested according to ASME Section IX all passed their bead-on-plate transverse bend tests for weldability, microscopic inspection revealed modest hot cracking which could pose challenges in processing if not mitigated. In order to mitigate this mechanism in future alloys, Schiele simulations were used to identify the low-temperature-solidifying phases that might be the cause. The hot cracking was then examined using energy dispersive spectroscopy which was able to confirm the responsible phase (shown as rich in “Element G” in Fig. 5). This clear identification was then leveraged to confidently remove “Element G” from the next generation of candidate composition.

Erosion Coating Results This program aims to develop thin, conformal coatings, applicable to steam turbine blades, that are more resistant to solid particle erosion (resulting from load following with a conventional thermal utility plant) than today’s Physical Vapor Deposited (PVD) TiN coatings while being more aerodynamically efficient than today’s thermally sprayed cermet DiamondTuff™ coatings. To be rapidly deployable, these coatings are limited to current PVD methodology/production hardware. These requirements and current technological options for addressing them are summarized in Table 2.

Corrosion and Erosion Protection to Accelerate Deployment …

23

Iron

Element G

Nickel

Fig. 5 A hot crack present in a weld bead fabricated from a candidate alloy imaged with backscatter electron microscopy and energy dispersive spectroscopy. This examination, coupled with thermodynamic calculations, indicated a detrimental phase rich in “Element G” was most likely responsible for hot cracks Table 2 Current options for preventing solid particle erosion in steam turbines Erosion protection strategy

Coating thickness

Adequate service life

Minimal aerodynamic debit

Rapid implementation

Steam path redesign

N/A

✔

✔

–

✔

–

✔

Thermal 150–250 µm spray cermet PVD TiN

3–10 µm

–

✔

✔

Novel PVD coatings

10–30 µm

✔

✔

✔

24

P. Shower et al.

Fig. 6 Pre- and post-steam erosion resistance of several development coatings demonstrating that the post-steam erosion resistance target for the program was greatly exceeded

The performance of novel PVD coatings on coupons in terms of “adequate service life” is quantified using a combination of erosion and steam exposure. A total of 75 coatings have been evaluated. The most successful coating shows a 13.8 × improvement in post-steam erosion resistance compared to PVD TiN that is currently deployed in this application, as provided by a vendor. At the outset of the program, only a 2.7 × improvement was targeted. A representative sampling of pre- and post-steam erosion resistance for these developmental coatings is given in Fig. 6.

Erosion Coating Discussion A patent application has been filed for the compositional and process improvements that resulted in this 13.8 × improvement in post-steam erosion resistance over technical coatings. However, it can be stated that the following strategies were leveraged in order to achieve the result: • • • •

Utilize both dopants and microstructure control High nanohardness phases Internal stress modulation Minimal oxidation kinetics.

Corrosion and Erosion Protection to Accelerate Deployment …

25

References 1. Technical communication with GE Steam Power, GE Gas Power, and GE Research Engineers (2020) 2. GEP (2020) Steam turbines. https://www.ge.com/power/steam/steam-turbines 3. Rebak RB (2017) Iron-chrome-aluminum alloy cladding for increasing safety in nuclear power plants. EPJ Nuclear Sci Technol 3:34 4. Rebak RB, Gupta VK, Larsen M (2018) Oxidation characteristics of two FeCrAl alloys in air and steam from 800°C to 1300°C. JOM 70(8):1484–1492 5. Yamamoto Y et al. (2008) Development of alumina-forming austenitic stainless steels. In: International technical conference on coal utilization and fuel systems, Clearwater, FL 6. Muralidharan G, Yamamoto Y et al. (2008) Development of cast alumina-forming austenitic stainless steel alloys for use in high temperature process environments. OSTI.gov 7. AB T (2012) Chemical composition and properties of FeCrAl alloys 8. UNR Commission (ed) (1975) Guidance for avoiding intergranular corrosion and stress corrosion in austenitic stainless steel components of fuel reprocessing plants, Washington, D.C. 9. Zhang R, Kiser S, Baker B (2010) Nickel alloy weld overlays improve the life of power generation boiler tubing. In: Stainless steel conference 10. Pint B (2014) The use of model alloys to study the effect of alloy composition on steam and fireside corrosion. Corrosion, San Antonio, TX 11. Daily Metal Prices. https://www.dailymetalprice.com/metalprices.php?c=ni&u=kg&d=5 12. Cheng W et al. (2014) Characteristics of oxide scale formed on ferritic stainless steels in simulated reheating atmosphere, University of Wollongong Research Online 13. Brachet J-C et al. (2019) Early studies on Cr-Coated Zircaloy-4 as enhanced accident tolerant nuclear fuel claddings for light water reactors. J Nucl Mater 517:268–285 14. Sandvik 310/T22 (2012) Composite superheater tube produce catalog 15. GESP (2019) Steam power product catalog 16. Gupta G et al. ( 2017) Critical reliability analysis of superheater tubes of coal based boiler 17. Nava J (2007) Evaluation of Amstar 888 coating. Alstom Progress Report 18. AW Society (2016) Standard methods for mechanical testing of welds. Gapped bead on plate test

Development of Indium-Tin Oxide Thin Films on PAMAM Dendrimer Layers for Perovskite Solar Cells Application Firdos Ali, Alecsander D. Mshar, Ka Ming Law, Xiao Li, A. J. Hauser, Shanlin Pan, Dawen Li, and Subhadra Gupta

Abstract Despite the dramatic progress that has been made in the power-conversion efficiency (PCE) of perovskite solar cells (PVSCs), there are still many obstacles to be overcome before these devices can be economically competitive in the photovoltaics market. One of the major hurdles in the commercialization of PVSCs is low stability, which severely limits the effective lifetime of the devices. One of the approaches to achieving higher stability and lifetime of PVSCs is improvement of PVSC film quality. In this paper, we have employed a PAMAM dendrimer layer to reduce the surface roughness of sputter-deposited indium-tin oxide (ITO) films, which were then used in the fabrication of PVSCs. A PAMAM-8 dendrimer layer was deposited by dip-coating the substrates in 25 mL of a 1 μM PAMAM-8 ethanol solution before ITO deposition. X-ray refractivity (XRR) was used to verify the PAMAM layer on the substrate. ITO films of 150 nm thickness were then deposited onto the PAMAM layer using DC magnetron reactive sputtering. The surface roughness, sheet resistance, and transmissivity of the ITO films were optimized by varying the parameters of the sputtering process. Atomic force microscopy (AFM) was used to measure the surface roughness of the ITO films with and without PAMAM dendrimer layer. A root-mean-square (RMS) film roughness of 1.6 nm, sheet resistance of 21 /ϒ, and transmissivity of > 91% at a wavelength of 400–700 nm were obtained after optimization.

F. Ali (B) · S. Gupta Department of Metallurgical and Materials Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA e-mail: [email protected] A. D. Mshar · D. Li Department of Electrical and Computer Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA K. M. Law · A. J. Hauser Department of Physics and Astronomy, The University of Alabama, Tuscaloosa, AL 35487, USA X. Li · S. Pan Department of Chemistry, The University of Alabama, Tuscaloosa, AL 35487, USA © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_3

27

28

F. Ali et al.

Keywords ITO · G8 PAMAM dendrimer · Sputtering · Sheet resistance · Transmittance

Introduction Perovskite solar cells (PVSCs) have recently garnered significant attention in many research communities. Current single-junction PVSCs rival the power-conversion efficiency (PCE) of Si cell designs and are generally cheaper and easier to fabricate than similarly performing Si cells [1–6]. One of the major roadblocks to further adoption of PVSCs is their low stability compared to Si cells—PVSCs do not perform well in ambient conditions and quickly deteriorate [7–9]. Since PVSCs are composed of several layers of different materials that are sequentially deposited, the performance and stability of these devices are heavily influenced by the quality of these layers and their interfaces. In this paper, we have developed processes to improve the film quality of sputter-deposited indium-tin oxide (ITO) for the synthesis of efficient PVSCs. ITO is a transparent, conductive oxide that is widely used in many different devices. ITO is commonly used in PVSCs as a transparent transport layer. As a transport layer, the ITO film must have both high transparency and high conductivity to reduce parasitic loss in the device. The improved surface roughness of these ITO thin films will (1) improve the PCE of solar cell devices that utilize this material and (2) improve the interfacial structure of layered devices, contributing to better overall stability. In this paper, we have optimized ITO thin films by three metrics: resistivity, transmissivity, and surface roughness. These film properties were optimized by varying the parameters of the film deposition technique and post-deposition treatment. The ITO films were deposited by reactive DC magnetron sputtering using a Sputtered Films, Inc. Planetary Sputtering System. The deposition parameters of sputter power, sputter time (film thickness), reactive gas flow rate, and post-deposition annealing time were all varied to produce an optimal ITO thin film for use in PVSCs. In addition to the deposition parameters, a pre-deposition PAMAM dendrimer layer was also employed to further reduce the surface roughness of the ITO films [10]. The ITO films were then characterized by several different methods: a 4-point probe was used to measure the sheet resistance of the films; a profilometer and ellipsometer were used to measure the film thickness; an atomic force microscope (AFM) was used to measure the film surface roughness; and an X-ray diffractometer was used to perform X-ray reflectivity (XRR) measurements in order to detect the presence of the PAMAM dendrimer layer.

Experimental Eighth generation (G8) PAMAM amine-terminated dendrimer (Dendritech, Midland, MI) in a 5% w/w methanol solution was used without any additional processing.

Development of Indium-Tin Oxide Thin Films on PAMAM Dendrimer …

29

Ethanol (Purity > 99.99) (Sigma-Aldrich Inc, St. Louis, MO) was used to prepare a 1 μM solution of the G8 amine-terminated dendrimer. 25 mL of the 1 μM dendrimer solution was used to fabricate a dendrimer layer on glass substrates with the dimensions of 25 × 25 × 1 mm [11]. The glass substrates were fully submersed in an ethanol (EtOH) and potassium hydroxide (KOH) solution for 1 h to remove organic and inorganic impurities. The substrates were then rinsed with de-ionized (DI) water (18 M) and isopropanol and dried under a steady stream of nitrogen gas. The substrates were subsequently cleaned using a UV/O3 cleaner (UV/O3 -Cleaner model 42, Jelight Company, Irvine, CA) for 1 min to remove any adsorbed hydrocarbon contaminants. Finally, the substrates were fully submerged in the 1 μM PAMAM dendrimer ethanol solution for over 15 h. After removing the glass substrates from the solution, they were thoroughly rinsed in pure ethanol and then dried under a steady stream of nitrogen gas. The final samples were stored in a desiccator to prevent decomposition of the dendrimer layer. All the steps described above are shown in Fig. 1. Sputter deposition of the ITO films was done using a sputter up SFI Shamrock Planetary Sputtering System in the Microfabrication Facility at the University of Alabama. The sputter target used was a 95% dense pressed powder compact of In2 O3 :SnO2 90:10 wt.%. First, the following ITO sputter deposition conditions were optimized with respect to resistivity and transmissivity: 200–700 W DC sputter power and 0–2.0 sccm oxygen flow with a total (argon + oxygen) flow rate of 50 sccm. The ITO sputter depositions were all done at a 2.71 mTorr process pressure. Depositions over a similar range of conditions were carried out on both silicon wafers and glass substrates in the same batch. The glass substrates and silicon wafers were first rinsed

Fig. 1 Flow chart of the PAMAM layer deposition process

30

F. Ali et al.

with 100% isopropanol to remove any unbound impurities and then dried in a stream of nitrogen gas. Transmission measurements were performed using UV–VIS spectrophotometers over a range of 400–700 nm wavelength. Baseline corrections were performed using a blank glass substrate, and the scan rate was 600 nm/min. Sheet resistance measurements were performed using four-point probes at two different facilities. Thickness measurements were done using a Veeco Dektak profilometer. Surface roughness measurements were done using a Digital Instruments Dimension 3100 AFM in tapping-mode. XRR measurements were done using a Rigaku SmartLab X-ray diffractometer in parallel-beam mode.

Results and Discussion Reactive Gas Flow Rate It has been shown that the sheet resistance and transmittance of ITO films are directly related to the oxygen flow rate during the sputtering process [12]. ITO thin films were sputtered at different oxygen flow rate percentages to optimize for resistance and transmittance. Figure 2 shows the sheet resistance of several 150 nm-thick ITO films deposited at various oxygen flow rates, at a power of 700 W and a total gas flow rate of 50 sccm at room temperature. As oxygen flow rate is increased from 0%, the sheet resistance first decreased to 54 /ϒ at an oxygen flow rate of 1.0% and then increased with an increasing oxygen flow rate until saturating at around 83–85 /ϒ near 2.0%. The above supports the fact that, since the conductivity of ITO films strongly depends on charge carrier concentration such as Sn4+ and oxygen vacancies on the substitution sites of In2 O3 , the conductivity of an ITO film can be increased by depositing the films in a low oxygen environment [13]. Figure 3 shows the transmittance of several ITO films over a wavelength range of 400–700 nm. As the flow of oxygen increases, transmittance of the films also increases. Similar behavior was observed with other oxygen flow percentages which is not shown in Fig. 3.

Film Thickness Figure 4 shows how the sheet resistance varies with ITO film thickness. Initially, the sheet resistance decreases as the film thickness increases. As the film thickness increases beyond 150 nm, the sheet resistance starts to increase. It is known that the transmissivity of the ITO films strongly depends on film thickness—thicker films have lower optical transmission due to increased optical path and optical scattering. This is demonstrated in Fig. 5, which shows the decrease of optical transmittivity as thickness increases.

Development of Indium-Tin Oxide Thin Films on PAMAM Dendrimer …

31

Fig. 2 ITO film room temperature sheet resistance with increasing oxygen flow rate. Sheet resistance reaches a minimum value at 1.0% flow rate

Fig. 3 ITO film optical transmission with increasing oxygen flow rate

Annealing Conditions Figure 6 shows how the sheet resistance varies with annealing conditions. Three samples were deposited simultaneously with identical parameters: 700 W deposition power, 15 min deposition time, and a 1% oxygen flow rate. All anneals were done for 5 min at 180 °C. The first sample shows a sheet resistance of 54 /ϒ without

32

F. Ali et al.

Fig. 4 ITO film room temperature sheet resistance with increasing film thickness

Fig. 5 ITO film optical transmission with increasing film thickness

annealing. The second sample went through both pre- and post-deposition annealing, which reduced the sheet resistance from 54 to 29 /ϒ. The final sample was only post-annealed, which showed the lowest sheet resistance of 21 /ϒ. We postulate that the increase in temperature immediately before the deposition increased residual stress due to a mismatch in thermal expansion coefficients between the substrate and the films. Figure 7 shows the sheet resistance of the films as the post-annealing time was changed. The sample that was annealed for 5 min yields the lowest sheet resistance

Development of Indium-Tin Oxide Thin Films on PAMAM Dendrimer …

33

Fig. 6 ITO film room temperature sheet resistance of samples that underwent different annealing conditions

of 21 /ϒ. As the annealing time is increased beyond 5 min, the sheet resistance of the film begins to increase. The data suggests that post-annealing times lower than 5 min may not have been sufficient for good crystallinity, hence leading to poor film quality and increased sheet resistance. In addition, Fig. 8 shows the transmittance of the ITO films at wavelengths between 400 and 700 nm for films with different post-annealing durations. The sample annealed for 5 min at 180 °C shows the best transmittance.

Deposition Power The DC sputtering power was varied between 200 and 700 Wfor the ITO deposition. An optimum ITO film thickness of 150 nm was chosen from the results shown above. The final thickness of the deposited films was kept constant by varying the deposition time according to the deposition power. The relationship between sheet resistance and sputtering power is shown in Fig. 9, which depicts an overall downwardsloped trend, with the lowest sheet resistance observed being 21 /ϒ at a deposition power of 700 W. This may be attributed to the higher deposition rate and ion-energy during the deposition process, both of which likely increased the film uniformity and crystallinity.

34

F. Ali et al.

Fig. 7 ITO film room temperature sheet resistance with increasing post-annealing time

Fig. 8 ITO film optical transmission spectra at different post-annealing times

PAMAM Dendrimer Layer ITO thin films were deposited on a G8 PAMAM dendrimer layer on a glass substrate. The successful deposition of the dendrimer layer was verified by XRR, as shown in Fig. 10. The layer thickness was calculated as follows:

Development of Indium-Tin Oxide Thin Films on PAMAM Dendrimer …

35

Fig. 9 ITO film room temperature sheet resistance with increasing sputtering power

 2D sin2 (θm ) − sin2 (θc ) = mλ The critical angle was identified to be at 0.46°, along with three thickness fringes at 0.68°, 1.11°, and 2.14°. The presence of XRR thickness fringes indicates two abrupt interfaces, which we attribute to the substrate/dendrimer layer interface and the layer surface. If this were the case, then the angular spacing between each adjacent pair of fringes would give the distance between the interfaces, i.e. layer thickness,

Fig. 10 X-ray reflectivity spectrum of PAMAM layer on glass substrate

36

F. Ali et al.

Fig. 11 AFM image of ITO thin film on PAMAM dendrimer layer (left) and without PAMAM dendrimer layer (right). The film deposited on PAMAM dendrimer layer has an RMS surface roughness of 1.6 nm, while the film deposited on bare glass substrate has a roughness of 3.2 nm

which is approximately 4 nm. While 3 is a small number of fringes, note that the number of fringes is not only limited by the film layer interface quality, but also by the instrument noise level. A thicker dendrimer layer would have shown more fringes. Figure 11 shows AFM images of ITO thin films with and without an underlying dendrimer layer. These measurements show that both films generally have no height variations larger than 20 nm [14]. The RMS surface roughness of the ITO sample deposited on the dendrimer is 1.6 nm, while the surface roughness of the sample without a dendrimer layer is 3.2 nm. The ITO film deposited on a dendrimer layer demonstrated marginally lower surface roughness than the ITO film that was deposited directly onto the glass substrate. This difference may be attributed to the layer’s ability to facilitate and form nucleation sites for the deposited ITO film. This leads to finer grains and a smoother film, though the improvement in surface roughness is not significant. Figure 12 shows that the sheet resistance of an ITO film is slightly higher when deposited on a dendrimer layer-coated glass substrate, since the PAMAM dendrimer is insulative. The sample prepared with a dendrimer layer shows slightly higher transmission when compared with the film without a dendrimer layer. This enhancement in transmission is attributed to the index-matching provided by the PAMAM dendrimer layer.

Conclusion The sputter deposition process of ITO films was optimized for use in the fabrication of efficient, stable PVSCs. The film qualities of sheet resistance, transmission, and

Development of Indium-Tin Oxide Thin Films on PAMAM Dendrimer …

37

Fig. 12 Room temperature sheet resistance (left) and optical transmission (right) comparison of ITO films with and without PAMAM layer

surface roughness were improved by varying the deposition conditions and introducing other film and substrate treatments. A G8 PAMAM dendrimer layer was used as interfacial layer to improve the surface roughness of the ITO film. The surface roughness of these films was decreased from 3.2 nm without a dendrimer layer to 1.6 nm with a dendrimer layer. By varying the sputter deposition parameters and annealing treatments, a sheet resistance of 21 /ϒ was achieved. In addition to low resistivity, a high transmissivity above 91% was achieved in the wavelength range of 400–700 nm. Acknowledgements We acknowledge NSF ECCS-2053954 for support. The authors would like to acknowledge Dr. Shane Street for valuable advice. Conflict of Interest The authors have no conflicts of interest to disclose.

References 1. Best research-cell efficiency chart. In: NREL.gov. https://www.nrel.gov/pv/cell-efficiency. html. Accessed 4 Oct 2022 2. Zhao Y, Zhu P, Wang M et al. (2020) A polymerization-assisted grain growth strategy for efficient and stable perovskite solar cells. Adv Mater 32:1907769 3. Jung EH, Jeon NJ, Park EY, Moon CS, Shin TJ, Yang T-Y, Noh JH, Seo J (2019) Efficient, stable and scalable perovskite solar cells using poly(3-hexylthiophene). Nature 567:511–515 4. Yoo JJ, Wieghold S, Sponseller MC et al. (2019) An interface stabilized perovskite solar cell with high stabilized efficiency and low voltage loss. Energy Environ Sci 12:2192–2199 5. Zhu H, Liu Y, Eickemeyer FT et al. (2020) Tailored amphiphilic molecular mitigators for stable perovskite solar cells with 23.5% efficiency. Adv Mater 32:1907757 6. Min H, Kim M, Lee S-U, Kim H, Kim G, Choi K, Lee JH, Seok SI (2019) Efficient, stable solar cells by using inherent bandgap of α-phase formamidinium lead iodide. Science 366:749–753

38

F. Ali et al.

7. Zhang Y, Hu X, Chen L, Huang Z, Fu Q, Liu Y, Zhang L, Chen Y (2016) Flexible, hole transporting layer-free and stable CH 3 NH 3 PBI 3/PC 61 BM planar heterojunction perovskite solar cells. Org Electron 30:281–288 8. Snaith HJ, Abate A, Ball JM, Eperon GE, Leijtens T, Noel NK, Stranks SD, Wang JT-W, Wojciechowski K, Zhang W (2014) Anomalous hysteresis in perovskite solar cells. J Phys Chem Lett 5:1511–1515 9. Gupta V, Lucarelli G, Castro-Hermosa S, Brown T, Ottavi M (2020) Investigation of hysteresis in hole transport layer free metal halide perovskites cells under dark conditions. Nanotechnology 31:445201 10. Gupta S, Ada E (2005) Optimization of process parameters to achieve high quality as-deposited indium-tin oxide films for display applications. J Vac Sci Technol A Vac Surf Films 23:1173– 1179 11. Thunuguntla R (2006) Effect of dendrimer mediation and rapid thermal annealing on indium-tin oxide thin films. Thesis 12. Sohn MH, Kim D, Kim SJ, Paik NW, Gupta S (2003) Super-smooth indium–tin oxide thin films by negative sputter ion beam technology. J Vac Sci Technol A Vac Surf Films 21:1347–1350 13. Wong FL, Fung MK, Tong SW, Lee CS, Lee ST (2004) Flexible organic light-emitting device based on magnetron sputtered indium-tin-oxide on plastic substrate. Thin Solid Films 466:225– 230 14. Chu JB, Huang SM, Zhu HB, Xu XB, Sun Z, Chen YW, Huang FQ (2008) Preparation of indium tin oxide thin films without external heating for application in solar cells. J Non-Cryst Solids 354:5480–5484

DFT Study of CuS-ZnS Heterostructures Louis Oppong-Antwi and Judy N. Hart

Abstract Heterostructure and solid solution formation provide new approaches to tuning the optoelectronic properties of semiconductor materials. In this work, density functional theory (DFT) calculations are used to systematically study the structural and optoelectronic properties of CuS–ZnS heterostructures as a function of ZnS layer thickness. The results show that varying the thickness of ZnS influences the gap between the Highest Occupied Molecular Orbital (HOMO) and Lowest Unoccupied Molecular Orbital (LUMO) of the structure. Also, the electronic properties of the CuS–ZnS heterostructures are sensitive to changes in the bonding environment at the interface, with particular thicknesses of ZnS corresponding to interfacial arrangements that give lower formation energies and HOMO-LUMO gaps than other structures. Based on the results, CuS–ZnS heterostructures can be expected to have tunable optoelectronic properties with HOMO-LUMO gaps in the energy range of visible light. Keywords Heterostructures · DFT · CuS · ZnS

Introduction Semiconductors are important for many optoelectronics and clean-energy applications, such as solar cells, photoelectrodes for photoelectrochemical water splitting and CO2 reduction, UV light emitting devices, and photothermal ablation devices for cancer [1–8]. However, many individual semiconductors have significant disadvantages that limit their performance in these applications, particularly those involving conversion of solar energy, such as wide band gaps that reduce utilization of sunlight, indirect band gaps, poor stability, and poor charge carrier mobility [9–12]. Combining two semiconductors to form a solid solution is one approach that has been used for L. Oppong-Antwi · J. N. Hart (B) School of Materials Science and Engineering, UNSW Sydney, Kensington, NSW 2052, Australia e-mail: [email protected] L. Oppong-Antwi e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_4

39

40

L. Oppong-Antwi and J. N. Hart

reducing band gaps and thus increasing visible-light absorption, as well as improving stability [13–15]. An example is solid solutions of ZnS and GaP, which have lower band gaps than the individual constituents on their own [16]. Similarly, Zn1−x Cdx S [17] and (ZnO)1−x (GaN)x [18] solid solutions have also been shown to demonstrate efficient visible-light photocatalytic activity for H2 evolution. Combining two compounds in the form of a solid solution influences the photocatalytic activity mostly by changing properties such as the band gap energy, the valence and conduction band widths, and the redox potentials of the conduction and valence bands. For example, for Zn1−x Cdx S, a high Zn content (i.e., Zn0.80 Cd0.20 S) was found to give the best performance due to the Cd reducing the band gap while the high Zn content maintains the relatively negative conduction band, giving more efficient hydrogen generation and better stability than pure CdS [19]. Another approach is heterostructure formation of a p-type and n-type semiconductor (and/or formation of type-II heterostructures) where an electric field or band offset across the interface helps to promote charge separation and thus hinder electron–hole recombination and increase efficiency [20–22]. Feng et al. showed that CuS–ZnS hetero-nanowires have enhanced visible-light photocatalytic activity compared with that of their individual counterparts, and this was attributed to the p–n heterojunction, as well as the larger specific surface area of the synthesized structures [23]. In addition to promoting charge separation, the unique bonding environments and corresponding electronic states at interfaces have also been found to allow enhanced visible-light absorption, for example in ZnS–GaP multilayer structures [24], making heterostructure formation and interface engineering promising approaches for tuning optoelectronic properties and overcoming deficiencies in the properties of individual semiconductors. In this work, we use density functional theory (DFT) calculations to investigate the optoelectronic properties of CuS–ZnS heterostructures, particularly how the thickness of each component in the structure affects the overall properties. This system has the advantage of being an all-sulfide system, thus giving an opportunity to better understand the optoelectronic properties of a non-oxide system.

Computational Method First principles DFT calculations of CuS–ZnS multilayer structures were carried out with the CRYSTAL17 code [25] using the B3LYP hybrid method [26, 27]. Previously published basis sets were used, with a 86-4111(d41) basis set for Cu [28], 86-311G basis set for S [29], and TZVP basis set for Zn [30]. To begin, calculations were done for bulk CuS and ZnS. The lattice parameters of the optimized CuS structure were a = b = 3.919 Å, c = 16.814 Å, in good agreement with the experimental values (a = b = 3.972 Å, c = 16.972 Å) [31, 32]. The crystal structure of CuS is made up of layers of planar CuS3 triangles alternating with layers of CuS4 tetrahedra connected by S−S bonds. Here, the Cu and S in the planar layer of CuS3 triangles are denoted Cu1 and S*, while Cu2 is used for Cu in the tetrahedra

DFT Study of CuS-ZnS Heterostructures

41

Fig. 1 Schematic illustration of generation of the six-atomic layer CuS slab (C-6)

layers, and S1 and S2 for the atoms in the S−S bond. ZnS has a zinc blende crystal structure at room temperature. After optimization, the calculated lattice parameter of bulk ZnS is 5.47 Å, similar to the experimental value (5.41 Å) [33]. Heterostructures of CuS and ZnS were made by firstly cutting slabs from both structures, along the [001] direction for hexagonal CuS and the [111] direction for zinc blende ZnS. Along the [001] direction of CuS, a slab thickness of six atomic layers was used, denoted as C-6. As shown in Fig. 1, this slab was made by cleaving the bulk CuS at the Cu-S bond parallel to the c-axis as this bond has been shown to be the weakest bond of the structure along the [001] direction [34]. To avoid polarization effects, the top and bottom surfaces of the slab were kept the same (i.e., a plane of CuS3 triangles). Keeping the thickness of this C-6 slab of CuS fixed, it was combined with ZnS slabs cut along the [111] direction with thicknesses ranging from two atomic layers (i.e., one Zn-S unit) to ten atomic layers in increments of two, thus forming five CuS–ZnS heterostructures denoted Z2C-6, Z4C-6, Z6C-6, Z8C-6, and Z10C-6, respectively. The structures formed were 3D periodic structures with no vacuum regions. After construction, these multilayer structures were fully optimized relaxing both atomic positions and lattice parameters. The formation energies of the heterostructures, E f , were calculated to study the stability of the structures. Generally, the lower the formation energy, the more stable the system. The formation energies of the heterostructures were calculated by: E f = E(heterostructure) − n E(CuS) − m E(ZnS)

(1)

where E(heterostructure) is the total energy of the heterostructure, and E(CuS) and E(ZnS) denote the total energies of bulk CuS and ZnS per formula unit, respectively, and n and m are the number of Cu and Zn atoms in the heterostructure, respectively.

42

L. Oppong-Antwi and J. N. Hart

Results and Discussion The calculated partial density of states (PDOS) of bulk CuS is shown in Fig. 2a. The PDOS shows that the top of the valence band (VB) is mainly composed of Cu 3d and S 3p states with the largest contribution from the Cu2 atoms. At energy levels above the valence band (~ 2 eV), a mid-gap state is formed by the sulfur atoms in the S–S bond, with an equal contribution from both S atoms. Above the mid-gap state, the conduction band (CB) is formed predominantly by Cu2 atoms with a hybridized contribution from all other atoms in the structure, although less from S*. The bulk structure has no calculated band gap after optimization (i.e., the Fermi level lies within the valence band, consistent with previous reports [34]); hence, the electronic properties of CuS were analyzed in terms of the gap between the Highest Occupied Molecular Orbital (HOMO, i.e., the VB) and the Lowest Unoccupied Molecular Orbital (LUMO, i.e., the mid-gap state formed by the sulfur atoms in the S–S bond) as has been done in previous reports on the electronic properties of CuS [35]. The HOMO-LUMO gap is found to be 0.85 eV, consistent with previous work [35]. This calculated HOMO-LUMO gap value is low compared to experimentally measured band gaps of CuS (1.8–2.5 eV) [36]. However, we note that the energy difference between the Fermi energy and the LUMO (2.10 eV) is similar to experimental values of the band gap. From the calculated electronic PDOS shown in Fig. 2b, ZnS has a band gap of 3.63 eV, similar to band gaps recorded experimentally (3.66 eV) [37]. The optimized heterostructures of Z2C-6 and Z10C-6, as representative examples, are shown in Fig. 3. The formation energies of the heterostructures are influenced by the thickness of the ZnS layer (Table 1). As the ZnS thickness increases from two to six layers, the formation energy progressively increases, and it then drops to a lower formation energy value when the ZnS thickness is eight layers, before increasing again (though

Fig. 2 Partial densities of states for bulk a CuS and b ZnS. In a, the Fermi level (indicated by the grey vertical line) is set to 0 eV, while in b, the top of the valence band is set to 0 eV

DFT Study of CuS-ZnS Heterostructures

43

Fig. 3 Optimized heterostructures of a Z2C-6 and b Z10C-6. The black lines show the boundaries of the simulation cells

Table 1 Formation energy, HOMO-LUMO gap, and interface bond lengths of the CuS–ZnS heterostructures Heterostructure

Formation energy (kJ/mol)

HOMO-LUMO gap (eV)

Bond lengths at interface (Å) Cu-S(Z)

Zn-S(C)

Z2C-6

56.4

1.39

2.44

2.53

Z4C-6

61.3

1.35

2.40

2.46

Z6C-6

61.4

1.33

2.39

2.44

Z8C-6

51.7

0.65

2.34

2.41

Z10C-6

54.2

1.19

2.38

2.42

remaining lower than that of Z2C-6) when the ZnS thickness is increased to ten layers. These changes in formation energy could be due to changes in the resulting structure, particularly changes in bond lengths and angles, and distortion in the structures. As seen in Table 1, when the ZnS thickness is increased from two to six layers, the bond length between Cu at the interface and an S atom in the ZnS layer (denoted S(Z)), as well as the bond length between Zn at the interface and an S atom in CuS (denoted S(C)), reduces, corresponding to an increase in formation energy. However, Z8C-6 and Z10C-6 behave differently with both having shorter interfacial bond lengths but lower formation energies than the other structures. In addition to changes in bond lengths, the bond angle at the interface also seems to be significant, particularly for explaining the low formation energy of Z8C-6. The bond angle at the interface between Cu, S(Z), and Zn shows a gradual increase from 106.5° to 107.5° and 108.0° for Z2, Z4, and Z6C-6, respectively, but then a significant decrease to 102.4° for Z8C-6, before returning to a value of 107.9° for Z10C-6, more consistent with the structures with thinner ZnS layers. Thus Z8C-6 has the lowest bond angle as well as the lowest interfacial bond lengths of any structure. This anomalous behavior of Z8C-6, in terms of bond lengths, angles, and formation energy, could be due to the ABCABC stacking sequence of the ZnS structure; as the ZnS thickness increases, whether the layers interfacing with CuS are of ‘A’, ‘B’, or ‘C’ type change and it may be that the CuS-ABCA-SCu sequence formed for Z8C-6 results in a greater extent of relaxation and rearrangement at the interface compared to Z4, Z6, and Z10, resulting in the lower formation energy (Fig. 4). Although corresponding interfaces

44

L. Oppong-Antwi and J. N. Hart

Fig. 4 a Z6C-6, b Z8C-6, and c Z10C-6 showing the distorted behavior parallel to the c-axis of the structure of Z8C-6. The black lines show the boundaries of the simulation cells

with ‘A’ layers would also occur for the Z2C-6 structure (i.e., CuS–A–SCu), the same interfacial arrangement with low bond lengths and angles is not seen as for Z8C-6, possibly due to the more flexible nature of the thinner layer of ZnS in this case. The PDOS of all the heterostructures was calculated (Fig. 5a–e) with the results showing that all the structures have the top of the VB formed by Cu atoms at the interface bonded to an S atom in the ZnS layer. Also, as for bulk CuS, states from the S atoms in the CuS layer are seen at the VB maximum for all the structures but their contribution is small compared to the Cu atoms. The mid-gap state is still formed by atoms in the S–S bond in the center of the CuS layer, as was seen for bulk CuS, but when the ZnS layer has a thickness of two atomic layers, the Zn atoms also contribute states at lower energies of the mid-gap state. This is also seen for Z8C-6 and Z10C-6. The bottom of the CB is formed mainly by Zn atoms at the interface for all structures. Since it is the energy difference between the VB maximum and the mid-gap state that determines the HOMO-LUMO gap, it may be expected that changes in the HOMO-LUMO gap for heterostructures with different ZnS layer thickness (Table 1) will result mainly from changes in the VB edge and/or mid-gap state. As the number of ZnS layers increases from two to six, the VB edge shifts only a small amount to lower energies relative to the Fermi level (change in VB maximum from Z2 to Z6 is 0.04 eV), but the lower edge of the mid-gap state shifts significantly to lower energies (energy difference between the mid-gap edge of Z2C-6 and Z6C-6 is 0.12 eV). Thus, it is mainly the shift in the mid-gap state that causes the HOMO-LUMO gap to reduce with increasing thickness of the ZnS layer. Similarly, for Z10C-6, the VB edge is at a similar energy to the Z2–Z6 structures, but the mid-gap state is shifted

DFT Study of CuS-ZnS Heterostructures

45

Fig. 5 Partial densities of states of a Z2C-6, b Z4C-6, c Z6C-6, d Z8C-6, e Z10C-6. f Mulliken charge magnitude of S in the ZnS layer, S(Z), that is bonded to Cu at interface HOMO-LUMO gap as a function. In all PDOS, the Fermi level (indicated by the grey vertical line) is set to 0 eV

46

L. Oppong-Antwi and J. N. Hart

to a significantly lower energy (by 0.08 eV relative to Z6C-6), corresponding to the reduced HOMO-LUMO gap. However, the most significant changes in the PDOS and the energies of the band edges are seen for Z8C-6. For all structures, states from the S atoms in ZnS bonded to Cu at the interface increasingly contribute at the VB maximum as the number of ZnS layers increases, but this effect is most significant for Z8C-6. The increased contribution of these S states to the VB maximum in Z8C-6 corresponds to spreading of the VB states to energies further above the Fermi level, by ~ 0.4 eV relative to the other structures. The mid-gap state also shifts to lower energies (difference in energy of the mid-gap edge between Z6C-6 and Z8C-6 is 0.21 eV), thus resulting in the much narrower HOMO-LUMO gap of Z8C-6 compared with the other structures (Table 1). At least for Z8C-6, it is evident from the PDOS that it is the changes in the S(Z) bonded to Cu at the interface that predominantly influence the HOMO-LUMO gap; thus, changes in the S(Z) charge as a function of ZnS layer thickness are investigated. As seen in Fig. 5f, the S(Z) charge tends to become smaller in magnitude as the ZnS layer thickness increases. For Z2, Z4, Z6, and Z10, the change in charge is relatively small, corresponding to small changes in the HOMO-LUMO gap (Table 1). However, for Z8C-6, there is a much larger change in the S(Z) charge and a correspondingly much larger decrease in the HOMO-LUMO gap. This decreased S(Z) charge in Z8C6 is consistent with the increased S(Z) contribution to unoccupied VB states above the Fermi level in the PDOS (Fig. 5d). Also, bond lengths at the interfaces tend to affect the HOMO-LUMO gap, with a decrease in the bond lengths of Cu-S(Z) and Zn-S(C) corresponding to a decrease in the HOMO-LUMO gap (Fig. 6). A similar effect is also seen for the S–S bond lengths. In particular, the Cu-S(Z) bond length is significantly shorter in Z8C-6 than in any of the other structures, corresponding to the larger contribution from the S(Z) atom in this bond to the VB states and the significantly smaller HOMO-LUMO gap.

Fig. 6 HOMO-LUMO gap of CuS-ZnS heterostructures with different ZnS layer thickness as a function of interfacial bond lengths for a Zn-S(C) bond and b Cu-S(Z) bond

DFT Study of CuS-ZnS Heterostructures

47

Thus, the interatomic bonds at the interfaces between the layers play a major role in determining the valence and conduction band energies and thus the HOMO-LUMO gap. This could be due to changes in the bond lengths and corresponding atomic charges and, particularly for Z8C-6, to changes in the bonds between Cu atoms and S(Z) atoms at the interface (which are slightly distorted parallel to the c-axis, Fig. 4). Changes in these parameters are in turn related to the ZnS layer thickness and which layer in the ZnS stacking sequence is bonded to CuS.

Conclusions The optoelectronic properties of CuS-ZnS heterostructures have been studied by DFT calculations. Formation energies and HOMO-LUMO gaps depend on the thickness of the ZnS layer in the heterostructure, and changes in these properties can be related to interfacial bond lengths and angles, as well as atomic charges. The HOMO-LUMO gap generally decreases with increasing ZnS thickness. At a ZnS thickness of eight atomic layers, the formation energy is at its lowest compared to all other structures, including those with both thicker and thinner ZnS layers, indicating a more stable structure, and this ZnS thickness also gives the narrowest HOMO-LUMO gap. The lower formation energy of this structure is correlated with changes in bond lengths and angles, while the lower HOMO-LUMO gap can be attributed to an increased contribution from the S atom (in ZnS bonded to Cu) at the interface to the unoccupied states at the VB maximum. Thus, it seems there are particular thicknesses of ZnS that result in specific interfacial atomic arrangements that correspond to low formation energies and HOMO-LUMO gaps compared with other structures. The changes in optoelectronic properties observed imply that CuS–ZnS heterostructures have HOMO-LUMO gaps that can be tuned across the range of visible light and thus have great potential for applications in optoelectronic devices. Acknowledgements This research was undertaken with the assistance of computational resources provided by the Australian Government through the National Computational Infrastructure (NCI) under the National Computational Merit Allocation Scheme. This research was supported by the Scientia Scholarship Scheme from UNSW Sydney.

48

L. Oppong-Antwi and J. N. Hart

References 1. Masuko K, Shigematsu M, Hashiguchi T, Fujishima D, Kai M, Yoshimura N, Yamaguchi T, Ichihashi Y, Mishima T, Matsubara N, Yamanishi T, Takahama T, Taguchi M, Maruyama E, Okamoto S (2014) Achievement of more than 25% conversion efficiency with crystalline silicon heterojunction solar cell. IEEE J Photovoltaics 4:1433–1435. https://doi.org/10.1109/ JPHOTOV.2014.2352151 2. Widyastuti E, Hsu JL, Lee YC (2022) Insight on photocatalytic and photoinduced antimicrobial properties of ZnO thin films deposited by HiPIMS through thermal oxidation. Nanomaterials 12. https://doi.org/10.3390/nano12030463 3. Muthukrishnan S, Subramaniam V, Mahalingam T, Helen SJ, Sumathi P (2017) Improved properties of spray pyrolysed CuO nanocrystalline thin films. J Mater Sci Mater Electron 28:4211–4218. https://doi.org/10.1007/s10854-016-6043-7 4. Tian Q, Jiang F, Zou R, Liu Q, Chen Z, Zhu M, Yang S, Wang J, Wang J, Hu J (2011) Hydrophilic Cu 9 S5 nanocrystals: a photothermal agent with a 25.7% heat conversion efficiency for photothermal ablation of cancer cells in vivo. ACS Nano 5:9761–9771. https://doi.org/10. 1021/nn203293t 5. Xiang Q, Cheng F, Lang D (2016) Hierarchical layered WS2/graphene-modified CdS nanorods for efficient photocatalytic hydrogen evolution. Chemsuschem 9:996–1002. https://doi.org/10. 1002/cssc.201501702 6. Gao Q, Wu X, Huang T (2021) Greatly improved NIR shielding performance of CuS nanocrystals by gallium doping for energy efficient window. Ceram Int 47:23827–23833. https://doi. org/10.1016/j.ceramint.2021.05.090 7. Kurnia F, Seidel J, Hart JN, Valanoor N (2021) Optical tuning of resistance switching in polycrystalline gallium phosphide thin films. J Phys Chem Lett 12:2327–2333. https://doi.org/ 10.1021/acs.jpclett.1c00163 8. Luo J, Ma L, He T, Ng CF, Wang S, Sun H, Fan HJ (2012) TiO 2 /(CdS, CdSe, CdSeS) nanorod heterostructures and photoelectrochemical properties. J Phys Chem C 116:11956– 11963. https://doi.org/10.1021/jp3031754 9. Iqbal S, Bahadur A, Anwer S, Ali S, Saeed R, Muhammad Irfan H, Li M, Javed MR, Shoaib M (2020) Shape and phase-controlled synthesis of specially designed 2D morphologies of L-cysteine surface capped covellite (CuS) and chalcocite (Cu2 S) with excellent photocatalytic properties in the visible spectrum. Appl Surf Sci 526:146691. https://doi.org/10.1016/j.apsusc. 2020.146691 10. Standing A, Assali S, Gao L, Verheijen MA, Van Dam D, Cui Y, Notten PHL, Haverkort JEM, Bakkers EPAM (2015) Efficient water reduction with gallium phosphide nanowires. Nat Commun 6:1–7. https://doi.org/10.1038/ncomms8824 11. Heidari G, Rabani M, Ramezanzadeh B (2017) Application of CuS–ZnS PN junction for photoelectrochemical water splitting. Int J Hydrogen Energy 42:9545–9552. https://doi.org/ 10.1016/j.ijhydene.2017.01.176 12. Wang X, Peng WC, Li XY (2014) Photocatalytic hydrogen generation with simultaneous organic degradation by composite CdS-ZnS nanoparticles under visible light. Int J Hydrogen Energy 39:13454–13461. https://doi.org/10.1016/j.ijhydene.2014.04.034 13. Hart JN, Cutini M, Allan NL (2014) Band gap modification of ZnO and ZnS through solid solution formation for applications in photocatalysis. Energy Proc 60:32–36. https://doi.org/ 10.1016/j.egypro.2014.12.338 14. Xing C, Zhang Y, Yan W, Guo L (2006) Band structure-controlled solid solution of Cd1-x Znx S photocatalyst for hydrogen production by water splitting. Int J Hydrogen Energy 31:2018–2024. https://doi.org/10.1016/j.ijhydene.2006.02.003 15. Kazuhiko Maeda KT, Daling Lu NS, Takata T, Yasunobu Inoue KD Photocatalyst relaesing hydrogen form water. Nature (n.d.) 9 16. Musavigharavi P, Kurnia F, Xie L, Park CKY, Ng YH, He J, Hart JN, Valanoor N (2021) ZnS-GaP solid solution thin films with enhanced visible-light photocurrent. ACS Appl Energy Mater 4:10756–10761. https://doi.org/10.1021/acsaem.1c01733

DFT Study of CuS-ZnS Heterostructures

49

17. Mei Z, Zhang M, Schneider J, Wang W, Zhang N, Su Y, Chen B, Wang S, Rogach AL, Pan F (2017) Hexagonal Zn1-x Cdx S (0.2 ≤ X ≤ 1) solid solution photocatalysts for H2 generation from water. Catal Sci Technol 7:982–987. https://doi.org/10.1039/c6cy02572b 18. Maeda K, Takata T, Hara M, Saito N, Inoue Y, Kobayashi H, Domen K (2005) GaN:ZnO solid solution as a photocatalyst for visible-light-driven overall water splitting. J Am Chem Soc 127:8286–8287. https://doi.org/10.1021/ja0518777 19. Wang DH, Wang L, Xu AW (2012) Room-temperature synthesis of Zn0.80 Cd0.20 S solid solution with a high visible-light photocatalytic activity for hydrogen evolution. Nanoscale 4:2046– 2053. https://doi.org/10.1039/c2nr11972b 20. Gnanasekaran L, Jalil AA, Kumar S, Rajendran S, Gracia F, Soto-Moscoso M, Habila MA, Saravanakumar K (2022) Nanoflower shaped NiO/CeO2 p-n junction material for the degradation of pollutant under visible light. Mater Lett 317:132122. https://doi.org/10.1016/j.matlet. 2022.132122 21. Wei L, Shifu C, Huaye Z, Xiaoling Y (2011) Preparation, characterisation of p-n heterojunction photocatalyst CuBi2 O4 /Bi2 WO6 and its photocatalytic activities. J Exp Nanosci 6:102–120. https://doi.org/10.1080/17458081003770295 22. Wang Y, Wang Q, Zhan X, Wang F, Safdar M, He J (2013) Visible light driven type II heterostructures and their enhanced photocatalysis properties: a review. Nanoscale 5:8326–8339. https:// doi.org/10.1039/c3nr01577g 23. Feng C, Meng X, Song X, Feng X, Zhao Y, Liu G (2016) Controllable synthesis of hierarchical CuS/ZnS hetero-nanowires as high-performance visible-light photocatalysts. RSC Adv 6:110266–110273. https://doi.org/10.1039/c6ra20306j 24. Park CK, Gharavi PSM, Kurnia F, Zhang Q, Toe CY, Al-Farsi M, Allan NL, Yao Y, Xie L, He J, Ng YH, Valanoor N, Hart JN (2019) GaP-ZnS multilayer films: visible-light photoelectrodes by interface engineering. J Phys Chem C 123:3336–3342. https://doi.org/10.1021/acs.jpcc.8b1 0797 25. Dovesi R, Erba A, Orlando R, Zicovich-Wilson CM, Civalleri B, Maschio L, Rérat M, Casassa S, Baima J, Salustro S, Kirtman B (2018) Quantum-mechanical condensed matter simulations with CRYSTAL. Wiley Interdiscip Rev Comput Mol Sci 8:1–36. https://doi.org/10.1002/wcms. 1360 26. Stephen PJ, Devlin FJ, Chabalowski CF, Frisch MJ (1994) Ab initio calculation of vibrational absorption. J Phys Chem 98:11623–11627 27. Becke AD (1993) Density-functional thermochemistry. III. The role of exact exchange. J Chem Phys 98:5648–5656 28. Doll K, Harrison NM (2000) Chlorine adsorption on the Cu(111) surface. Chem Phys Lett 317:282–289. https://doi.org/10.1016/S0009-2614(99)01362-7 29. Azavant P, Lichanot A, Rerat M, Pisani C (1994) Ab initio Hartree-Fock study of lithium and sodium sulfides: electronic and scattering properties. Acta Crystallogr Sect B 50:279–290. https://doi.org/10.1107/S0108768193013849 30. Peintinger MF, Oliveira DV, Bredow T (2013) Consistent Gaussian basis sets of triple-zeta valence with polarization quality for solid-state calculations. J Comput Chem 34:451–459. https://doi.org/10.1002/jcc.23153 31. Subramanyam K, Sreelekha N, Amaranatha Reddy D, Murali G, Rahul Varma K, Vijayalakshmi RP (2017) Chemical synthesis, structural, optical, magnetic characteristics and enhanced visible light active photocatalysis of Ni doped CuS nanoparticles. Solid State Sci 65:68–78. https:// doi.org/10.1016/j.solidstatesciences.2017.01.008 32. Roberts HS, Ksanda CJ (1929) The crystal structure of covellite. Am J Sci s5–17:489–503. https://doi.org/10.2475/ajs.s5-17.102.489 33. Yim WM (1969) Solid solutions in the pseudobinary (III-V)-(II-VI) systems and their optical energy gaps. J Appl Phys 40:2617–2623. https://doi.org/10.1063/1.1658043 34. Conejeros S, Moreira IDPR, Alemany P, Canadell E (2014) Nature of holes, oxidation states, and hypervalency in covellite (CuS). Inorg Chem 53:12402–12406. https://doi.org/10.1021/ ic502436a

50

L. Oppong-Antwi and J. N. Hart

35. SenthilKannan K, Sivaramakrishnan V, Kalaipoonguzhali V, Chinnadurai M, Kannan S (2020) Electronic transport, HOMO-LUMO and computational studies of CuS monowire for nano device fabrication by DFT approach. Mater Today Proc 33:2746–2749. https://doi.org/10.1016/ j.matpr.2020.01.574 36. Wang X, He Y, Hu Y, Jin G, Jiang B, Huang Y (2018) Photothermal-conversion-enhanced photocatalytic activity of flower-like CuS superparticles under solar light irradiation. Sol Energy 170:586–593. https://doi.org/10.1016/j.solener.2018.06.022 37. Kurnia F, Ng YH, Tang Y, Amal R, Valanoor N, Hart JN (2016) ZnS thin films for visiblelight active photoelectrodes: effect of film morphology and crystal structure. Cryst Growth Des 16:2461–2465. https://doi.org/10.1021/acs.cgd.5b01590

Effect of H2 Enrichment on CO/N2 /H2 -air Turbulence Partial Premixed Flame Combustion Characteristics Fan Yang, Qingguo Xue, Haibin Zuo, Binbin Lv, Yu Liu, and Jingsong Wang

Abstract To investigate the effect of H2 enrichment on the combustion characteristics of H2 /CO/N2 partial premixed flames, experiments were conducted using tunable diode laser absorption spectroscopy (TDLAS) and an infrared gas analyzer. The changes in the H2 content (contents of 0, 16, 32, 48, and 65%) on the flame structure, flame temperature, and concentration of gases (CO, CO2 ) in the flame at a fixed air–fuel ratio of 3.2 were investigated. As the H2 content increased, the flame length decreased, high-temperature zone moved forward, temperature increased, CO2 content first increased and then decreased, and CO content decreased. The addition of H2 facilitated the formation of OH radicals, causing the oxidation reaction of CO to proceed faster than in the presence of only O2 and O. Keywords Flame temperature · Flame length · CO concentration · TDLAS

Introduction Blast furnace gas (CO/CO2 /N2 /H2 ) produced by blast furnace ironmaking is a low calorific value gas mixture that can be recovered into the blast furnace through the tuyere after CO2 removal. Blast furnace gas is an important secondary energy source for iron and steel enterprises and can be used as fuel in hot blast furnaces, blast furnaces, and power plants. Different blast furnace gas compositions affect the gas combustion behavior as reflected in the flame temperature, flame length, and gas composition produced [1–3]. Therefore, simulating the combustion of blast furnace gas components by mixing multiple gases at a certain ratio is an effective method for studying the flame temperature and component concentration during the combustion process [4]. Tunable diode laser absorption spectroscopy (TDLAS) has been widely used in recent years to study the combustion behavior of various atmospheres in industrial F. Yang · Q. Xue · H. Zuo · B. Lv · Y. Liu · J. Wang (B) State Key Laboratory of Advanced Metallurgy, University of Science and Technology Beijing, Beijing 100083, China e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_5

51

52

F. Yang et al.

environments under harsh ambient conditions. This non-contact spectroscopy technique simultaneously measures the flame temperature and component concentration changes in a flame. Zhou et al. [5] compared two tunable diode laser temperature sensors to determine the gas temperature and its fluctuations. Kamimoto et al. [6] discussed the high-temperature field application of a two-dimensional (2D) temperature measurement technique using CT-TDLAS. Ma et al. [7] reported in situ measurements of non-uniform temperature, H2 O, and CO2 concentration distributions in a premixed methane-air laminar flame using TDLAS. TDLAS has been widely used for temperature measurement in harsh ambient conditions. The combustible gas component of blast furnace gas (CO) accounts for 20–30%, and its combustion generates a large amount of CO2 , which is a greenhouse gas. To reduce CO2 emissions, CO alternatives are required. H2 has been widely studied as an efficient and clean energy source that can replace CO; its combustion product is H2 O, which has no environmental impact. The effects of H2 and H2 O addition on the flame structure and NOx emissions from methane-air counter flow diffusion flames were investigated by Park et al. [8]. Shy et al. [9] studied the effects of H2 and CO2 addition, equivalence ratio, and turbulent strain on the turbulent velocity of lean premixed methane combustion. Babak and Sadegh [10] studied the effect of H2 on the non-premixed turbulent combustion of C3 H8 -H2 -N2 mixtures. Jie et al. [11] studied the effect of H2 and CO addition on the flame properties of laminar methane flow. However, limited research has been conducted on the combustion behavior of gas mixtures with H2 instead of CO in blast furnace gas. In this study, we investigated the effect of partial CO replacement by H2 on the combustion characteristics of a H2 /CO/N2 (stripped CO2 ) partially premixed flame at a fixed air–fuel ratio of 3.2. The flame temperature, length, and component concentration were investigated using tunable diode laser absorption spectroscopy (TDLAS) and an infrared gas analyzer to investigate the effect of H2 on the flame combustion of the blast furnace gas and provide theoretical support for hydrogen-rich blast furnace smelting.

Experimental The TDLAS Combustion Diagnostic System The TDLAS-based combustion diagnostic system consists of a gas supply, burner, combustion chamber, displacement platform, and measurement device, as shown in Fig. 1. This equipment has been previously described [4]. The flame shapes and lengths are captured by high-precision industrial cameras. The flame length was defined as the average visible flame length obtained from 50 captured flame photographs. The flame temperature and CO2 concentration in the flame along the axial centerline were detected in real-time using a laser, and the CO concentration at the same position was detected using an infrared gas analyzer. Spectral lines with

Effect of H2 Enrichment on CO/N2 /H2 -air Turbulence Partial Premixed …

53

Fig. 1 Schematic of TDLAS combustion diagnostic system [4]

wavelengths of 1996 and 2004 nm were selected experimentally as light beams to detect the absorption of CO2 , similar to previous studies [12, 13].

Experimental Conditions The experiments simulated the combustion of blast furnace gas in an industrial turbulent partial premix burner, as shown in Fig. 2. A mixture of H2 and CO was used as the fuel under dilute combustion conditions, which are listed in Table 1. Combustion case 2 was taken from the percentage of blast furnace gas components in a steel plant (after simplified treatment) and constituted the basis for the design of the remaining four experimental conditions. Under all experimental conditions, the fixed air–fuel ratio was 3.2 with H2 replacing CO, thus changing the volume percentage of H2 . The volume percentage of H2 was defined as RH2 = Q H2 /Q fuel , Q fuel = Q H2 /Q CO . The flame length, axial centerline flame temperature, and concentration changes of the components (CO and CO2 ) in the flame were investigated.

Results and Discussion Flame Length The flame morphology and length after the replacement of CO with H2 in the fuel are shown in Fig. 3. As shown in Fig. 3b, the flame length decreases almost linearly

54

F. Yang et al.

Fig. 2 Schematic diagram of the burner

Table 1 Experimental conditions Cases

Q fuel

Q N2

Q H2

Q CO

RH2

Q air

L/min

L/min

L/min

L/min

%

L/min

1

62

10

0

52

00

200

2

62

10

10

42

16

200

3

62

10

20

32

32

200

4

62

10

30

22

48

200

5

62

10

40

12

64

200

and the width decreases as the H2 content in the fuel increases with the same fuel flow rate and burner structure. The rapid combustion of hydrogen leads to a flame propagation speed of 265–325 cm/s. Because of the small size of the H2 molecules and their high diffusion rate, hydrogen easily reacts with oxygen. The increase in the percentage of H2 intensifies the reaction H2 + O2 = H2 O, which creates a chain reaction that is intensified by the large production of its intermediate products (OH, H, O, and other free radicals). This results in the flame length and width decrease [14].

Flame Temperature Figure 4 shows the flame axial center temperature distribution after H2 substitution for CO for different H2 percentages in the fuel. From Fig. 4, it can be found that the flame temperature increases as the H2 content increases. This may be explained by the calorific value of H2 being slightly greater than the calorific value of CO, the fast reaction rate, and resulting fast energy release per unit of time.

Effect of H2 Enrichment on CO/N2 /H2 -air Turbulence Partial Premixed …

Flame length (mm)

260

55

(b)

H2

243

240

241 225

220

217 205

200

180

0%

16%

32%

48%

64%

Fig. 3 Flame morphology and length: a flame morphology; b flame length

1350

Flame temperature/K

1300 1250 1200

H2-0%

1150

H2-16%

1100

H2-48%

H2-32%

H2-64%

1050 1000 0

28

56

84

112

140

168

196

Axial position along the flame centerline(mm) Fig. 4 Axial center flame temperature at different H2 percentage

In addition, along the axial centerline flame exit to 70 mm, the flame temperature slightly increases at a moderate rate, followed by a rapid decrease. The points indicated by the arrows in Fig. 4 are the corresponding maximum flame temperatures under the five conditions, which are 1291, 1306, 1317, 1328, and 1341 K. It can be observed that the corresponding position of the highest temperature moves along the

56

F. Yang et al.

axial centerline towards the flame outlet. This indicates the existence of a reaction zone for fuel combustion that shrinks with increasing H2 content.

Component Concentration The changes in the concentrations of the components of the combustion products, CO and CO2 , are shown in Figs. 5 and 6. In this study, CO and CO2 concentrations in the flame refer to the volume percentages at different positions along the axial direction. As can be seen in Fig. 5, when the H2 contents are 0 and 16%, the CO concentration increases slightly and then decreases rapidly. In contrast, when the H2 contents are 32, 48, and 64%, the CO concentration decreases without any increasing trend. In addition, Fig. 5 also shows that the overall CO concentration decreases as the percentage of H2 increases. When the percentage of H2 is 64%, the initial concentration of CO is just 9.5%. CO is derived from CO in fuel, and its consumption is by reaction with O2 and OH radicals with reaction equations such as: 2CO + O2 → 2CO2 (1) and CO + OH → CO2 + H (2). When the hydrogen content is small, CO reacts with O2 as the main reaction, and its reaction rate is slow (Carbon monoxide reacts slowly with oxygen molecules and reacts quickly with particles such as oxygen O atoms or free radicals.) [12], which is accompanied by the reaction of CO + O → CO2 . Therefore, the CO concentration at the outlet varies slightly at 0, 16, and 32% of H2 , and is approximately 20%. At 42 mm from the flame exit, the flame rolled up the surrounding air and the CO concentration decreases rapidly. When the hydrogen 25

CO concentration (v%)

H2 -0% 20

H2 -16%

15

H2 -48%

H2 -32% H2 -64%

10

5

0 0

28

56

84

112

140

168

Axial position along the flame centerline(mm) Fig. 5 Variation of CO concentration along the axial centerline

196

Effect of H2 Enrichment on CO/N2 /H2 -air Turbulence Partial Premixed …

57

CO2 concentration (v%)

35 H2 -0%

30

H2 -16%

H2 -32%

H2 -48%

H2 -64%

25 20 15 10 5 0 0

28

56

84

112

140

168

196

Axial position along the flame centerline(mm) Fig. 6 Variation of CO2 concentration along the axial centerline

content is high, numerous OH radicals are produced, increasing the reaction rate of Eq. (2) (Faster reaction rate of CO with OH radicals) [13], thus decreasing the CO concentration rapidly at the outlet. Figure 6 shows the variation in the concentration of CO2 , which first increases and then decreases along the axial position. As the H2 concentration increases, the overall CO2 concentration decreases. The CO2 concentration increases first because of the simultaneous presence of the reactions (1) and (2) and the O2 in the surrounding environment being consumed by the flame, thus increasing the CO2 concentration. At 112 mm from the exit, the CO2 concentration decreases due to the dilution of the environment. In addition, when the percentage of H2 is 48 and 64%, the CO2 concentration increases slowly until 112 mm and then slightly decreases or remains constant. This indicates that the H2 combustion reaction occurs preferentially to the CO combustion reaction throughout the combustion process, and when the H2 content is large, CO + OH → CO2 + H is the main reaction for CO2 production.

Conclusions As a secondary energy source, the recovery and use of blast furnace gas are of great importance. Studying the combustion reaction of H2 instead of CO can provide theoretical support for hydrogen-rich smelting. The variation in the combustion characteristics of the CO/N2 /H2 -air turbulent partially premixed flame was studied using the TDLAS technique. We found that, as the H2 concentration increases, the flame length decreases, width decreases, and temperature increases; the highest flame temperature moves towards the flame exit, and the reaction zone is slightly reduced. Moreover, CO

58

F. Yang et al.

concentration decreases along the axial centerline and then levels off, and the overall CO concentration decreases as the percentage of H2 increases; CO2 concentration decreases overall as the H2 content increases, and along the axial centerline, CO concentration first increases and then decreases or plateaus. In addition, an increase in H2 concentration increases the concentration of OH radicals, which accelerates the chain reaction and plays a significant role in the combustion reaction. Acknowledgements The authors gratefully acknowledge the financial support of the National Natural Science Foundation of China (No. U1960205, No. 51804024).

References 1. Wang YK, Lei XM, Deng L, Wang XX, Wang CA, Che DF (2014) A review on utilization of combustible waste gas(I):blast furnace gas, converter gas and coke oven gas. Ther Power Gener 43(7):1–14 2. Hou SS, Chen CH, Chang CY, Wu CW, Ou JJ, Lin TH (2011) Firing blast furnace gas without support fuel in steel mill boilers. Energy Convers Manage 52(7):2758–2767 3. Lin H, Jin HG, Gao L, Zhang N (2014) A polygeneration system for methanol and power production based on coke oven gas and coal gas with CO2 recovery. Energy 74(C):174–180 4. Liu Y, Wang JS, Xue QG et al. (2020) Investigation of H2 addition effects on CO/CO2 /H2 -air flames by a combustion diagnostic system based on TDLAS. In: Energy technology 2020: recycling, carbon dioxide management, and other technologies. The Minerals, Metals & Materials Series. Springer, Cham 5. Zhou X, Jeffries JB, Hanson RK, Li G, Gutmark EJ (2007) Wavelength-scanned tunable diode laser temperature measurements in a model gas turbine combustor. AIAA J 45(2):420–425 6. Kamimoto T, Deguchi Y, Kiyota Y (2015) High temperature field application of two dimensional temperature measurement technology using CT tunable diode laser absorption spectroscopy. Flow Meas Instrum 46:51–57 7. Ma LH, Lau LY, Ren W (2017) Non-uniform temperature and species concentration measurements in a laminar flame using multi-band infrared absorption spectroscopy. Appl Phys B Lasers O 83:576–587 8. Park J, Sang IK, Yun JH (2007) Addition effects of H2 and H2 O on flame structure and pollutant emissions in methane-air diffusion flame. Energy Fuels 21(6):3216–3224 9. Shy SS, Chen YC, Yang CH, Liu CC, Huang CM (2008) Effects of H2 or CO2 addition, equivalence ratio, and turbulent straining on turbulent burning velocities for lean premixed methane combustion. Combust Flame 153(4):510–524 10. Babak K, Sadegh T (2013) A numerical study on the effects of H2 addition in non-premixed turbulent combustion of C3H8–H2 –N2 mixture using a steady flamelet approach. Int J Hydrogen Energ 38(23):9918–9927 11. Jie L, Xin Z, Tao W (2015) Numerical study of the chemical, thermal and diffusion effects of H2 and CO addition on the laminar flame speeds of methane–air mixture. Int J Hydrogen Energ 40:8475–8483 12. Liu Y, Xue QG, Zuo HB, She XF, Wang JS (2021) Experimental study of H2 and/or N2 addition effects on CO/CO2 -air flames using a combustion diagnostic system. J Therm Sci 30(4):1268–1277

Effect of H2 Enrichment on CO/N2 /H2 -air Turbulence Partial Premixed …

59

13. Liu Y, Xue QG, Zuo HB, Yang F, Wang JS (2021) Effects of CO2 and N2 dilution on the combustion characteristics of H2 /CO mixture in a turbulent, partially premixed burner. ACS Omega 6:15651–15662 14. Duan XL, Wang ZM, Wang LJ, Qiu XQ (2003) Influence of hydrogen content in gas on flame property. Gas Heat 23(11):655–658

Part II

Energy Efficiency, Decarbonization and CO2 Management

CO2 Mineralization and Critical Battery Metals Recovery from Olivine and Nickel Laterites Fei Wang and David Dreisinger

Abstract The global clean energy transition requires CO2 emission reduction with a concurrent increase in the global supply of critical battery metals. A process has been developed at the lab scale. The hydrometallurgical process achieves CO2 mineralization and selective battery metal recovery from olivine and laterites. The natural minerals are processed at a modest temperature with a carbon dioxide pressure in a sodium bicarbonate solution containing soluble ligands enabling nickel and cobalt extraction. Iron and magnesium react with CO2 gas to form stable mineral carbonates for carbon dioxide sequestration. The leached nickel and cobalt are recovered by sulfide precipitation as high-value sulfides. The corresponding barren solution is recycled with no decrease in performance. The process consumes carbon dioxide and a source of sulfide. No additional acid or base is consumed in this novel process. Therefore, this work can potentially make significant contributions to the enhanced production of critical battery metals with enhanced CO2 storage and to the clean energy transition. Keywords CO2 mineralization · Critical battery metal recovery · Clean energy transition · Olivine and laterites · Selective nickel and cobalt extraction

Introduction The clean energy transition is necessary to achieve sustainable development [1, 2]. To achieve the clean energy transition, the global supply of critical battery metals must be increased, including nickel and cobalt [3]. Nickel and cobalt account for more than 70% and 10%, respectively, of the most popular nickel-manganese-cobalt (NMC) lithium battery cathodes and may further increase in the future [4–6]. Cano F. Wang (B) · D. Dreisinger Department of Materials Engineering, The University of British Columbia, Vancouver, BC V6T 1Z4, Canada e-mail: [email protected] F. Wang Department of Mining, Metallurgy and Materials Engineering, Laval University, Quebec City, QC G1V 0A6, Canada © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_6

63

64

F. Wang and D. Dreisinger

et al. [7] predicted that the share of electrical vehicles (EVs) may increase and reach over 85% of the vehicle market by 2060. An enhanced supply of nickel and cobalt would be essential for the development of EVs and the clean energy transition. The global supply of nickel will be mainly from laterites since nickel in laterite resources accounts for around 70% of the global nickel reserves [8]. It is important to achieve clean production of nickel from laterites as the origin of the clean energy transition. Dreisinger and Wang [9] have developed the innovative process, Concerted Mineral Carbonation and Selective Leaching from Laterites (CMCSL). This process utilizes the CO2 mineralization process to selectively leach nickel and cobalt from laterites and thus the CO2 emission reduction can be achieved for sustainable production. In addition, this CMCSL process is suitable for all layers of laterites including limonite laterite and saprolite laterite and thus the capital cost for the production can be significantly reduced. Therefore, it will play an important role in future sustainable production of nickel and cobalt. In addition, an enhanced supply of nickel and cobalt can be supplemented from waste tailings and wastes [10–13]. Ultramafic minerals, especially olivine, are usually considered as gangue minerals and discarded as tailings but contain residual nickel and cobalt which cannot be recovered through traditional methods [10]. Wang et al. [12, 13] have confirmed that it is possible to convert nickel and cobalt from olivine and tailings to nickel/cobalt sulfides by utilizing the CO2 mineralization process under a CO2 –H2 S gas mixture. Wang et al. [14] have further confirmed the possibility of selective nickel and cobalt leaching from olivine with simultaneous CO2 mineralization using a complexing ligand edetate disodium salt (EDTA). Since the amounts of waste tailings in various mining industries are enormous and will continue to increase, recovery of nickel and cobalt from the tailings with simultaneous CO2 mineralization for permanent storage is also significant to meet the increasing demand of nickel and cobalt for the clean energy transition. This work introduces CO2 mineralization and selective nickel and cobalt recovery from olivine, limonite laterite, and saprolite laterite by using a complexing ligand trisodium nitrilotriacetate (NTA). The flow of nickel and cobalt during the process and the phase transformation of minerals are summarized in this work.

Materials and Methods Materials A pure olivine sample representing waste tailings and a limonite laterite and a saprolite laterite representing different layers of laterites were used in this work. The chemical composition of the three samples is shown in our previous work [15]. The olivine sample contained 0.22% nickel and 0.0055% cobalt which represent residual nickel and cobalt in tailings. The limonite laterite and saprolite laterite contained 1.16% nickel and 0.041% cobalt and 2.28% nickel and 0.063% cobalt, respectively.

CO2 Mineralization and Critical Battery Metals Recovery from Olivine …

65

The samples also contained significant contents of magnesium, iron, and calcium which were used to sequester CO2 through CO2 mineralization as stable mineral carbonates. The mineral compositions of the samples are shown in our previous work [15]. The olivine sample comprised dominant olivine followed by a little fraction of enstatite and serpentine; the limonite laterite consisted of goethite, montmorillonite, and quartz, followed by serpentine, olivine, enstatite, and talc; and the saprolite laterite mainly contained serpentine, montmorillonite, and quartz, followed by olivine, enstatite, goethite, and talc. In addition to the mineral samples, sodium bicarbonate, trisodium NTA and pure CO2 gas were also used for the CO2 mineralization and selective leaching process. The recovery of nickel and cobalt from an aqueous solution utilized a gas mixture containing 5% H2 S and 95% CO2 for sulfide precipitation.

Methods The CO2 mineralization with simultaneous nickel and cobalt leaching tests was carried out in a 600 mL stainless steel autoclave (Parr Instrument, Model 5103) at 175 °C, 34.5 bar CO2 pressure P(CO2 ), 1.5 molal sodium bicarbonate, and pulp density of 5 wt.% and 0.5 wt.% for the olivine and the laterite samples, respectively. The slurry was then simply filtered to obtain the pregnant leach solution (PLS) containing nickel- and cobalt-complex ions and sodium bicarbonate and the residue for storage as mineral carbonates. Prior to the reaction of CO2 mineralization and leaching, the laterite samples may need calcination/reduction by a rotary electric furnace (Nabertherm RSRB 80-750/11, Germany) at 700 °C and 300 mL/min gas flow containing 5% CO and 95% N2 for 1 h to convert hydrated silicate minerals including serpentine to reactive olivine and ferric oxide (including goethite) to reactive ferrous oxide wustite. The leach solution was subsequently precipitated at 120 °C and 30.6 bar of the total pressure of gas mixture containing 5% H2 S and 95% CO2 for recovery of nickel and cobalt as high-value nickel– and cobalt–sulfide concentrates. The nickel-poor solution after sulfide precipitate was then recycled back to the CO2 mineralization and selective leaching step without supplemental addition of sodium bicarbonate or NTA. The metal extraction/recovery efficiency was determined by the leached metal content in an aqueous solution through inductively coupled plasma atomic emission spectroscopy (ICP-AES) analysis. The CO2 mineralization efficiency was determined by the amount of sequestered CO2 in residues as mineral carbonates through LECO total inorganic analysis (LECO CS230). The residues were also analyzed by X-ray diffraction to determine the change in mineral compositions and to identify the formation of mineral carbonates.

66

F. Wang and D. Dreisinger

Results and Discussion CO2 Mineralization and Simultaneous Critical Battery Metal Extraction Olivine The CO2 mineralization and simultaneous critical battery metal extraction from the olivine samples are shown in Fig. 1. With the increase in reaction time, both CO2 mineralization and the extraction efficiencies of nickel and cobalt gradually increased with an identical increasing trend. Within the first 5 h, the nickel and cobalt extraction and CO2 mineralization significantly increased to 91%, 89%, and 84%, respectively, as shown in Fig. 1a. Subsequently, the nickel and cobalt and CO2 mineralization further slightly increased to 96%, 97%, and 90% at 8 h, respectively. In contrast, the iron and magnesium extraction kept decreasing and low to 2% and 0.7%, respectively. The simultaneous CO2 mineralization and nickel and cobalt extraction from olivine is a highly selective process. Considering the contents of nickel and cobalt and CO2 mineralization capacity based on iron, magnesium, and calcium, the amount of extracted nickel and cobalt and sequestered CO2 can reach 2.0 kg Ni, 0.049 kg Co, and 461 kg CO2 per tonne olivine in 5 h (Fig. 1b). This amount can further increase to 2.1 kg Ni, 0.053 kg Co, and 497 kg CO2 per tonne olivine in 8 h, although the reaction time within 5 h is more reasonable. According to the change in mineral composition of the olivine sample (Fig. 1c), olivine has been completely reacted within 8 h and magnesite-siderite (magnesium-iron-carbonates) were generated, while enstatite and serpentine remained unreacted. Our previous studies [16, 17] have confirmed that olivine is the most reactive mineral and serpentine (and enstatite) without heat pretreatment is not reactive. With the reaction of olivine and the formation of mineral carbonates, the silica converted to amorphous silica [18, 19] and thus was not shown in the X-ray diffraction spectra. The work on the olivine represents the fundamental study of CO2 mineralization and simultaneous critical metal extractions. The tailings from flotation plants having significant amounts of olivine may be also suitable for this work.

Limonite Laterite With inspiration from the fundamental work on the olivine sample, the CO2 mineralization and simultaneous nickel and cobalt from laterites would be more attractive since the laterites contained more than 1% nickel much higher than that in the olivine. As shown in Fig. 2a, the raw limonite laterite was not suitable for direct CO2 mineralization and selective leaching, with lower than 51% of nickel and cobalt extraction and less than 2% CO2 mineralization. The goethite, hydrate silicates including montmorillonite, serpentine, and talc were not reactive during the CO2 mineralization process.

CO2 Mineralization and Critical Battery Metals Recovery from Olivine …

67

Fig. 1 CO2 mineralization and simultaneous critical battery metal extraction from olivine: a metal extraction and CO2 mineralization efficiency (originated from [15]); b amount of extracted critical battery metals and sequestered CO2 ; and c X-ray diffraction spectra of reacted olivine residue

The heat pretreatment through calcination/reduction was required. After the pretreatment step, the nickel and cobalt extraction increased to 87% and 81% within 2 h, respectively, together with 45% CO2 mineralization efficiency. The iron and magnesium extraction were very low at 11% and 1.8%, respectively. It is still a highly selective extraction process. Correspondingly, 10 kg Ni, 0.33 kg Co, and 131 kg CO2 per tonne the limonite laterite (Fig. 2b) were selectively extracted and sequestered. The mineral composition of the reacted limonite residue (Fig. 2c) shows that the mineral carbonates, magnesite, and siderite, have become the dominant composition, followed by quartz. The olivine (the original mineral olivine and the generated olivine during pretreatment step) has nearly completed the reaction with only a little fraction remained in the residue, while the enstatite remained unreactive. Nevertheless, the limonite laterite is also suitable for simultaneous CO2 mineralization and selective nickel and cobalt extraction, although heat pretreatment is required.

Saprolite Laterite Saprolite laterite is usually at the lower layer in a laterite deposit and contains higher nickel than limonite laterite. Similar to the limonite laterite sample, the saprolite

68

F. Wang and D. Dreisinger

Fig. 2 CO2 mineralization and simultaneous critical battery metal extraction from limonite laterite: a metal extraction and CO2 mineralization efficiency; b amount of extracted critical battery metals and sequestered CO2 ; and c X-ray diffraction spectra of reacted Limonite residue

laterite sample is also suitable for CO2 mineralization and concurrent nickel and cobalt extraction with a heat pretreatment step of calcination/reduction, as shown in Fig. 3. Without the pretreatment step, the extraction of nickel and cobalt and CO2 mineralization efficiency was less than 59% and 3% (Fig. 3a), respectively. In contrast, the pretreatment step can increase the CO2 mineralization efficiency and nickel and cobalt extraction to 47%, 94%, and 92% within 2 h, respectively. The iron and magnesium extractions were still lower than 10% and 1.4%, respectively. Correspondingly, the amounts for the extraction and carbon sequestration have increased to 21 kg Ni, 0.58 kg Co, and 171 kg CO2 per tonne saprolite laterite (Fig. 3b), respectively. The X-ray diffraction spectra of the residue (Fig. 3c) also show that the magnesite-siderite has become the dominant mineral, followed by quartz and the residual olivine fraction and the unreactive enstatite. As known from the work of the olivine sample, amorphous silica would be also generated.

CO2 Mineralization and Critical Battery Metals Recovery from Olivine …

69

Fig. 3 CO2 mineralization and simultaneous critical battery metal extraction from saprolite laterite: a metal extraction and CO2 mineralization efficiency; b amount of extracted critical battery metals and sequestered CO2 ; and c X-ray diffraction spectra of reacted Saprolite residue

Critical Battery Metal Recovery and Solution Recycling Critical Battery Metal Recovery by Sulfide Precipitation The extracted nickel and cobalt in the aqueous solution were in the form of (Ni,Co)NTA− complex ions and needed a further recovery step. Sulfide precipitation is one of the suitable recovery methods and has been widely used in the laterites industry [8]. The leachate from the pretreated saprolite laterite was tested for critical metal recovery through sulfide precipitation using a gas mixture of 5% H2 S and 95% CO2 as shown in Fig. 4. Nearly all the nickel, cobalt, and iron were precipitated as high-value nickel sulfide concentrate while magnesium remained untouched in the aqueous solution. The sulfide concentrate having high nickel content with a molar ratio of nickel/iron = 1.5 can be directly saleable in the market.

Solution Recycling After the sulfide precipitation for nickel and cobalt recovery, the nickel-poor barren solution needed to be recyclable to minimize the operation costs. Figure 5 compares

70

F. Wang and D. Dreisinger

Fig. 4 Critical battery metal recovery from pregnant leach solution of pretreated saprolite laterite through sulfide precipitation

the metal extraction between using the fresh solution containing sodium bicarbonate and NTA and using the recycled solution without any supplementary chemicals. There was no obvious difference in the metal extraction of nickel and iron (82% and 11% in 1 h, respectively) during the CO2 mineralization and selective extraction from the pretreated saprolite laterite. The recycled solution for the selective leaching step can obtain even lower magnesium extraction efficiency, only 0.4% compared to 2.8% using fresh solution because magnesium was not removed during the sulfide precipitation step and the CO2 -sodium bicarbonate-NTA system during the CO2 mineralization step maintained the magnesium concentration level (less than 30 mg/L). The success in the solution recycling can also indicate that there was no consumption of NTA and sodium bicarbonate throughout the whole process including the CO2 mineralization with leaching and subsequent sulfide precipitation steps.

Phase Transformation Summary Mineral Phase Transformation The phase transformation of minerals throughout the pretreatment and CO2 mineralization with simultaneous metal extractions from the olivine, limonite laterite, and saprolite laterite has been summarized in Fig. 6. For the olivine sample, the pretreatment step was not needed. The mineral olivine converted to magnesite-siderite and amorphous silica during the CO2 mineralization and selective leaching step, while enstatite and serpentine remained unreacted. The laterites including limonite laterite and saprolite laterite needed pretreatment through calcination/reduction with a gas

CO2 Mineralization and Critical Battery Metals Recovery from Olivine …

71

Fig. 5 Comparison of fresh leach solution and recycled barren solution after sulfide precipitation for CO2 mineralization and selective metal extraction from pretreated saprolite laterite

mixture of CO–N2 or CO–CO2 . The hydrate silicate minerals including serpentine, montmorillonite, and talc converted to reactive olivine. The ferric oxide goethite converted reactive wustite under a reductive atmosphere. During the CO2 mineralization and simultaneous leaching step, the minerals of olivine and wustite converted to magnesite-siderite and amorphous silica with a little fraction of olivine remained in the residue. Enstatite and quartz remained unreacted in the residue.

Critical Battery Metal Phase Transformation In addition to the change in mineral compositions and the storage of CO2 into mineral carbonates, the recovery of critical battery metals is also important and may determine the feasibility of economics [11]. The phase transformation of critical battery metals (nickel and cobalt) throughout the pretreatment, CO2 mineralization and selective leaching, and sulfide precipitation was summarized in Fig. 7. For the olivine sample, nickel and cobalt were in the crystal structure of mineral olivine and then were extracted by the complexing ligand NTA and sodium bicarbonate during the CO2 mineralization and simultaneous leaching step as (Ni, Co)NTA− complex ions in aqueous solution. The nickel and cobalt in (Ni, Co)NTA− were then recovered by H2 S during the sulfide precipitation step as a high-value nickel sulfide concentrate (Ni, Co)S while H2 NTA− and bicarbonate ions remained in the aqueous solution for recycling back to CO2 mineralization and simultaneous step. For the limonite laterite sample, nickel and cobalt were mainly in goethite and montmorillonite followed by olivine and serpentine. With the pretreatment, nickel and cobalt in montmorillonite and serpentine converted to olivine, while the nickel and cobalt in goethite converted to wustite. During the CO2 mineralization and simultaneous leaching, nickel and cobalt in olivine and wustite were extracted by NTA into aqueous solution as (Ni,

72

F. Wang and D. Dreisinger

Fig. 6 Phase transformation of minerals throughout pretreatment, CO2 mineralization, and simultaneous leaching

Co)NTA− complex ions, followed by recovery as (Ni, Co)S concentrate during the sulfide precipitation. For the saprolite laterite, the situation was similar except that the nickel and cobalt were mainly in montmorillonite and serpentine followed by in goethite and olivine. The whole process for the nickel and cobalt recovery is simple and short and thus can be potentially applicable in the laterites industry and waste tailings.

Conclusion CO2 mineralization and critical battery metals recovery from olivine and nickel laterites are discussed in this work. Simultaneous CO2 mineralization and selective nickel and cobalt extraction from the various olivine, limonite, and saprolite samples can be achieved in one step by using the CO2 -sodium bicarbonate-NTA system. More than 90% nickel and cobalt extractions can be obtained together with more than 50% CO2 mineralization efficiency while iron and magnesium extractions were less than 10% and 2%, respectively. The extracted nickel and cobalt can be further recovered by sulfide precipitation as high-value nickel–cobalt-sulfide concentrates and the nickel-poor barren solution can be recycled back to CO2 mineralization

CO2 Mineralization and Critical Battery Metals Recovery from Olivine …

73

Fig. 7 Phase transformation of critical battery metals throughout pretreatment, CO2 mineralization and simultaneous leaching, and metal recovery

and selective leaching step with comparable results to fresh solution. The NTA and sodium bicarbonate were not consumed and can be recyclable throughout the process. A pretreatment by calcination/reduction is required for limonite and saprolite laterites to convert hydrate silicate minerals and ferric oxide to reactive olivine and wustite. The phase transformations of minerals and critical metals throughout the process are also summarized. This process is important for the sustainable production of nickel and cobalt from nickel laterites and waste tailings containing olivine and can make a significant contribution to clean energy transition. Acknowledgements The authors thank Mitacs Accelerate and LeadFX Inc. (IT26205) for financial support. Sibelco Europe is also thanked for the natural olivine sample.

References 1. International Energy Agency I (2016) Technology roadmap—low-carbon transition in the cement industry. World Business Council For Sustainable Development (WBCSD). https:// doi.org/10.1007/springerreference_7300 2. Niass T, Aramco S, Arabia S, Kislear J (2017) Mission innovation: accelerating the clean energy revolution, carbon capture innovation challenge, report of the carbon capture, utilization, and storage workshop 3. Watari T et al. (2019) Total material requirement for the global energy transition to 2050: a focus on transport and electricity. Resour Conserv Recycl 148:91–103

74

F. Wang and D. Dreisinger

4. Patel P (2020) Lithium-ion batteries go cobalt-free. C&EN Glob Enterp 98:9–9 5. Li W, Lee S, Manthiram A (2020) High-nickel NMA: a cobalt-free alternative to NMC and NCA cathodes for lithium-ion batteries. Adv Mater 32 6. Shafiei Sabet P, Sauer DU (2019) Separation of predominant processes in electrochemical impedance spectra of lithium-ion batteries with nickel-manganese-cobalt cathodes. J Power Sources 425:121–129 7. Cano ZP et al. (2018) Batteries and fuel cells for emerging electric vehicle markets. Nat Energy 3:279–289 8. Crundwell F, Moats M, Robinson T, Ramachandran V, Davenport W (2011) Extractive metallurgy of nickel and cobalt. Elsvier. https://doi.org/10.1016/C2009-0-63541-8 9. Dreisinger D, Wang F (2022) Concerted mineral carbonation and selective leaching from laterites. WO/2022/077112 10. Wang F, Dreisinger DB, Jarvis M, Hitchins T (2018) The technology of CO2 sequestration by mineral carbonation: current status and future prospects. Can Metall Q 57:46–58 11. Wang F, Dreisinger D (2022) Status of CO2 mineralization and its utilization prospects. Miner Miner Mater 1:4 12. Wang F, Dreisinger D, Jarvis M, Hitchins T, Trytten L (2021) CO2 mineralization and concurrent utilization for nickel conversion from nickel silicates to nickel sulfides. Chem Eng J 406:126761 13. Wang F, Dreisinger D, Barr G, Martin C (2022) Utilization of copper nickel sulfide mine tailings for CO2 sequestration and enhanced nickel sulfidization. In: REWAS 2022: developing tomorrow’s technical cycles, vol 1. Springer, pp 227–239. https://doi.org/10.1007/978-3-03092563-5_24 14. Wang F, Dreisinger D (2022) Carbon mineralization with concurrent critical metal recovery from olivine. Proc Natl Acad Sci USA 119:1–7 15. Wang F, Dreisinger D (2022) An integrated process of CO2 mineralization and selective nickel and cobalt recovery from olivine and laterites. Chem Eng J. https://doi.org/10.1016/j.cej.2022. 139002 16. Wang F, Dreisinger D, Jarvis M, Hitchins T (2021) Kinetic evaluation of mineral carbonation of natural silicate samples. Chem Eng J 404:126522 17. Wang F, Dreisinger D, Jarvis M, Trytten L, Hitchins T (2021) Application and optimization of a quantified kinetic formula to mineral carbonation of natural silicate samples. Miner Eng 161:106712 18. Wang F, Dreisinger D, Jarvis M, Hitchins T (2019) Kinetics and mechanism of mineral carbonation of olivine for CO2 sequestration. Miner Eng 131:185–197 19. Wang F, Dreisinger D, Jarvis M, Hitchins T, Dyson D (2019) Quantifying kinetics of mineralization of carbon dioxide by olivine under moderate conditions. Chem Eng J 360:452–463

Decarbonization Pathways for an Aluminum Rolling Mill and Downstream Processes Alexander Wimmer

Abstract In this work the corporate carbon footprint (CCF) of Constantia Teich, the biggest plant for aluminum-based flexible packaging in Europe, was analyzed. Based on the CCF together with an in-depth analysis of processes the product carbon footprint (PCF) was calculated. Based on the CCF and PCF calculations a strategy for net zero CO2 emissions in 2040 was developed. Keywords Carbon footprint · Scope 1 · Scope 2 · Aluminium · Flexible packaging

Introduction Europe wants to become the first climate-neutral continent in 2050 [1]. Some countries like Austria want to reach this ambitious goal already in 2040 [2]. Although this goal sounds simple and plausible, the way to achieve it is still largely unknown and there is an almost unmanageable number of scenarios to reach this goal [3]. Many industries focus on carbon capture and utilization (CCU), while it is known from the literature that this can be only a small part of the solution [4]. Subsequently, this paper is focusing on the prevention of CO2 emissions. Constantia Teich produces more than 70,000 tons, corresponding to 1 billion square meters of flexible, aluminum-based packaging per year. Constantia Teich has its own aluminum foil rolling mill and its own research and development competence center for aluminum and functional coatings. In addition to a highly efficient basic finishing facility (for extrusion coating, laminating, and lacquering of aluminum foils), the site has its own lacquer production facility to implement recipes developed in-house. In 3–5 shift operation, materials are rolled, extruded, lacquered, laminated, printed, cut, deep-drawn, embossed, and die-cut in a vertically integrated production from aluminum foil to the final product. This paper is focusing on internal emissions, which means Scope 1 and 2 emissions [5].

A. Wimmer (B) Constantia Teich GmbH, Mühlhofen 4, 3205 Weinburg, Austria e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_7

75

76

A. Wimmer

Methods and Results Due to the European emission trading system (ETS) the corporate carbon footprint (CCF) is already known and was 32,000 tons in 2021 (Fig. 1). Main drivers were the consumption of pipeline natural gas (PNG) and the consumption and subsequent regenerative thermal oxidation (RTO) of solvents (from lacquering and printing) (Fig. 2). Since 2019 electricity is carbon neutral, resulting in zero Scope 2 emissions. Based on this number the carbon footprint of each division (rolling mills, basic converting, printing, and finishing) and each department (Fig. 3) was calculated. Subsequently, the specific carbon footprint per square meter foil (g/m2 ) was analyzed (Fig. 4) to be able to calculate a product carbon footprint (PCF).

Fig. 1 Annual corporate carbon footprint (CCF) of Constantia Teich from 2015–2021. Since 2019 electricity is carbon neutral, resulting in zero Scope 2 emissions

Fig. 2 Monthly corporate carbon footprint (CCF) of Constantia Teich for January–June 2022

Decarbonization Pathways for an Aluminum Rolling Mill …

77

Fig. 3 Monthly corporate carbon footprint (CCF) of different departments for January–June 2022

Fig. 4 Monthly, specific carbon footprint (gram CO2 per square meter foil, g/m2 ) of different departments for January–June 2022. The decreasing trend of departments reel cutting, embossing, die cutting, and deep drawing can be explained by the fact that room heating is the main CO2 driver

Discussion Figures 1 and 2 are illustrating that Constantia Teich has no Scope 2 emissions and only Scope 1 emissions, mainly based on the consumption of natural gas and solvents for lacquering. In the case of solvents, fossil solvents should be replaced by biogenic solvents in the future and concepts have to be developed to keep solvents as long as possible in a closed loop, e.g., by distilling (Fig. 5). To cushion cost increases, a concept was developed to distinguish between the combustion of fossil and biogenic energy sources by C-14 method, thus reducing the need to purchase certificates in the emission trading system (ETS).

78

A. Wimmer

Fig. 5 Annual corporate carbon footprint (CCF) of Constantia Teich from 2015–2021 and measures to reduce the carbon footprint to net zero in 2040

For natural gas, a concept must be developed that significantly reduces natural gas consumption. On the one hand, natural gas can be replaced by biomethane, on the other hand existing natural gas burners can be upgraded to dual fuel burners, allowing the combustion of biogasoline (FAME, fatty acid methyl ester), which has been also analyzed in other projects [6]. Moreover, process heat can be produced by a biomass cogeneration plant. In the case of gaseous biofuel (biomethane), existing infrastructure can be used, however, in the case of on-site production, a biogas plant requires an investment in the order of 1 Mio USD per 10 GWh annual biomethane production; furthermore purification of the raw gas (removal of CO2 , sulfur, etc.) requires extensive plant equipment. Liquid biofuels (biodiesel) require a replacement or upgrade of the existing burner infrastructure. In the case of on-site production, investment is about 50% compared to on-site gaseous biofuel production. Finally, in the case of solid biofuel, heat production has to be centralized and heat is then distributed to the production machines. Investment for on-site production (biomass cogeneration plant) is in the same order of magnitude as on-site liquid biofuel production, however, the heat distribution network is in the same order of magnitude as the cogeneration plant itself. In conclusion, natural gas can be replaced either by solid, liquid, or gaseous biofuel (Fig. 5). With the above-mentioned optimizations in the area of solvents and natural gas, 2/3 of our emissions can be saved. A further 10% is to be reduced through efficiency improvements. Finally, as also shown by the CO2 strategies of other operating companies [7], compensation measures are necessary for the remaining 20% of emissions.

Conclusion In this paper, a comprehensive in-depth analysis of Scope 1 and Scope 2 emissions was carried out. Based on this analysis, it is possible to analyze the carbon footprint of the entire factory, individual departments, and even individual products on a

Decarbonization Pathways for an Aluminum Rolling Mill …

79

monthly basis. On the one hand, this serves to support the customer in the selection of product variants; on the other hand, both individual machines and departments can be compared on a monthly basis and deviations can be examined and energy waste can be reduced. A strategy has been presented to reduce CO2 emissions to net zero in 2040; this makes our products much more attractive to customers and reduces the financial risk of fluctuations in the emissions trading system [8]. As a next step, Scope 3 emissions will also be included in our analysis in order to be able to examine the CO2 emissions over the entire life cycle of our products [9]. Regarding raw materials, concepts were developed which, on the one hand, allow a certain recycling content in the raw material and, on the other hand, a strategy will be developed to make selected products 100% recyclable through measures such as mono-material composites or recycling-tolerant alloys.

References 1. European Parliament. https://www.europarl.europa.eu/news/de/press-room/20210419IPR0 2302. Accessed 31 Aug 2022 2. Austrian ministry for climate protection. https://www.bmk.gv.at/themen/klima_umwelt/agenda 2030/bericht-2020/nachhaltigkeit.html. Accessed 31 Aug 2022 3. Clora F, Yu W (2022) GHG emissions, trade balance, and carbon leakage: Insights from modeling thirty-one European decarbonization pathways towards 2050. Energy Econ 113:106240 4. Kleijne K, Hanssen SV, van Dinteren L, Huijbregts MAJ, van Zelm R, Coninck H (2022) Limits to Paris compatibility of CO2 capture and utilization. One Earth 5(2):168–185 5. Anquetin T, Coqueret G, Tavin B, Welgryn L (2022) Scopes of carbon emissions and their impact on green portfolios. Econ Model 115:105951 6. Djuri´c Ili´c D, Ödlund L (2022) Integration of biofuel and DH production—possibilities, potential and trade-off situations: a review. Fuel 320:123863 7. https://www.handelsblatt.com/unternehmen/industrie/co2-reduktion-bosch-will-komplett-kli maneutral-arbeiten-und-laesst-sich-das-eine-milliarde-euro-kosten/24321480.html. Accessed 31 Aug 2022 8. Flora M, Vargiolu T (2020) Price dynamics in the European Union emissions trading system and evaluation of its ability to boost emission-related investment decisions. Eur J Oper Res 280(1):383–394 9. Schulman DJ, Bateman AH, Greene S (2021) Supply chains (Scope 3) toward sustainable food systems: an analysis of food and beverage processing corporate greenhouse gas emissions disclosure. Clean Prod Lett 1:100002

Rethinking the Decomposition of Refractory Lithium Aluminosilicates: Opportunities for Energy-Efficient Li Recovery from LCT Pegmatites Joanne Gamage McEvoy, Yves Thibault, Nail R. Zagrtdenov, and Dominique Duguay Abstract Extracting lithium from Li–Cs–Ta (LCT) pegmatites is highly energyintensive, involving an initial heat treatment at temperatures exceeding 1000 °C (decrepitation) to induce a transition in the refractory aluminosilicate carrier minerals (α-spodumene, petalite) to a more reactive phase (β-spodumene), allowing the breakdown of the crystalline structure and lithium release in a subsequent acid baking stage (~250–300 °C). This study focuses on investigating alternative, energy-efficient approaches to lithium recovery from LCT pegmatites. Our results indicate that there exist optimal conditions that induce effective alkali exchange directly from the primary phases, attainable at temperatures < 450 °C in under 30 min, creating opportunities for a viable single-stage mineral decomposition process without the need for a decrepitation step. The nature of the reactive process will be discussed, where exchanged alkali and other impurities are partitioned in stable solid phases, allowing improved purity of the pregnant leach solution, therefore minimizing downstream challenges associated with battery precursor production (Li2 CO3 , LiOH·H2 O). Keywords Lithium · Alkali-based decomposition · Spodumene

J. Gamage McEvoy (B) · Y. Thibault · N. R. Zagrtdenov · D. Duguay Natural Resources Canada, CanmetMINING, 555 Booth Street, Ottawa, ON K1A0G1, Canada e-mail: [email protected] Y. Thibault e-mail: [email protected] N. R. Zagrtdenov e-mail: [email protected] D. Duguay e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_8

81

82

J. Gamage McEvoy et al.

Introduction Global efforts to electrify transportation are a major driver for growth in lithium demand, and supply chain diversification is a priority. Numerous occurrences of Li–Cs–Ta (LCT) pegmatites in Canada represent an important mineral-based source of lithium, where α-spodumene and petalite are the main carrier phases. However, due to their refractory nature, conventional recovery approaches are highly energyintensive, involving an initial heat treatment at temperatures exceeding 1000 °C (‘decrepitation’; e.g., [1–3]) to induce a transition to a more reactive phase (βspodumene), allowing the breakdown of the crystalline structure and lithium release in a subsequent acid baking stage at around ~ 250–300 °C (e.g., [3–5]). This paper summarizes some recent results from a study focusing on the investigation of alternative energy-efficient strategies to directly exchange lithium from the primary minerals hosted in LCT pegmatites at low temperatures without the need for a decrepitation step.

Materials and Methods Materials The starting material used is a spodumene-quartz intergrowth (SQI) sample from the Tanco LCT pegmatite, Bernic Lake, Manitoba. Characterization by X-ray diffraction (XRD) analysis and automated quantitative mineralogy (Fig. 1) confirm that α-spodumene and quartz are the only significant phases present, in proportions consistent with the inferred natural conversion of petalite (LiAlSi4 O10 petalite → LiAlSi2 O6 spodumene + SiO2 quartz ) during progressive crystallization of the pegmatite [6]. The spodumene chemical composition, as determined by electron probe X-ray microanalysis (EPMA) using wavelength-dispersive spectrometry (WDS) and laserablation inductively coupled plasma mass spectrometry (LA-ICP-MS), is close to the ideal endmember, with only sodium (~500 ppm Na2 O), manganese (~200 ppm MnO), and iron (~200 ppm FeO) as trace elements of significant concentration (≥ 100 ppm; Table 1).

Alkali-Lithium Exchange Experiments The alkali-lithium exchange experiments were performed in the NaOH–KOH system using equimolar amounts of the endmembers, which, due to the presence of a deep eutectic, melt at a temperature as low as 170 °C [7, 8]. The feed consisted of solid SQI blocks, polished on two parallel faces to create a reference exchange interface, that were loaded in a Ni crucible (Fig. 2a). After adding the hydroxide reagent, the

Rethinking the Decomposition of Refractory Lithium Aluminosilicates: …

83

b)

intensity (a.u.)

a)

α-spodumene quartz

5 mm

α-spodumene: 64 wt%

quartz: 36 wt%

2θ (Cu Kα)

Fig. 1 General characteristics of the SQI sample used for this study. a Powder-XRD pattern. Simulated ICDD reference patterns for α-spodumene (00-033-0786) and quartz (00-046-1045) are shown for comparison. b Modal distribution and textural relationship as determined by automated quantitative mineralogy based on energy-dispersive spectrometry and backscattered electron (BSE) imaging

Table 1 Chemical composition of α-spodumene in the Bernic Lake SQI. Si, Al, and Na were determined by WDS-EPMA; Li, Mn, and Fe were measured by LA-ICP-MS α-spodumene, Bernic Lake N: 57 SiO2

Wt.%

1σ

64.87

0.29

Al2 O3

27.20

0.11

Li2 O

7.84

0.05

Na2 O

0.05

0.01

MnO

0.02

0.01

FeO

0.02

0.01

Total

99.99

–

crucible was covered (Fig. 2b) and transferred to a furnace pre-heated to 300 °C. The temperature was then increased at a rate of 6.7 °C min−1 and, at peak temperature, the crucible was removed and the product was rapidly quenched on a PTFE membrane (splat quench; Fig. 2c). To obtain information on the textural nature and elemental distribution within the exchanged layer, the recovered SQI block was polished normal to the reference reaction front using methanol as a lubricant in order to dissolve the hydroxides but preserve all other water-soluble phases.

84

J. Gamage McEvoy et al.

a)

b) SQI block

1 cm

c)

PTFE membrane reacted SQI block

(Na0.5K0.5)OH melt

quenched melt

SQI block Ni crucible

Ni crucible

1 cm

splat quench

Fig. 2 a Typical doubly-polished SQI block, shown in Ni crucible, used for decomposition experiments. b Schematic of SQI block and alkali reagent loaded in covered Ni crucible. c Recovery of partially reacted SQI block through splat quench on PTFE membrane

Results and Discussion Figure 3 shows the growth of the exchanged layer for dynamic heating experiments from 300 °C to peak temperature ranging from 400 up to 500 °C. Through such dynamic heating alkali-exchange experiments, the identification of a thin reacted layer occurred at around 400 °C, reached after 15 min, with no significant growth up to 425 °C. However, a very rapid increase in the exchange rate was initiated in a narrow window between 425 and 450 °C in less than 4 min followed by moderate growth of the exchanged layer up to 500 °C. From these observations, we estimate that for grains of a few millimeters (Fig. 3 inset), complete spodumene decomposition can be achieved in less than 30 min. Quantitative elemental distribution across partially-reacted spodumene was obtained by WDS-EPMA (Si, Al, K, Na) and LA-ICP-MS (Li), and the maps shown are expressed as the weight percent of the oxides (Fig. 4). The observed depletion in lithium of almost an order of magnitude in the reacted layer indicates very efficient 400oC

exchanged layer

BSE

425oC

exchanged layer

BSE

quartz

α-spodumene

450oC

BSE

quartz

α-spodumene

500oC

100 μm BSE

exchanged layer

exchanged layer 100 μm

α-spodumene

100 μm

100 μm

α-spodumene

100 μm

Fig. 3 Temperature–time path (left) and corresponding BSE images (right) showing the evolution of the exchanged layer as a function of peak temperature. The BSE image in the small inset emphasizes the extent of reactivity for a mm-size spodumene crystal

Rethinking the Decomposition of Refractory Lithium Aluminosilicates: …

85

exchange. Interestingly, there is strong partitioning between the alkalis with potassium preferentially exchanging with lithium in the form of an aluminosilicate and sodium mainly interacting with quartz to form sodium silicates.

BSE

wt%

WDS

100 90 80 70 60 50 40 30 20 10 0

exchanged layer quartz

α-spodumene

WDS

200 μm

SiO2 wt% 30

wt%

WDS

30

25

25

20

20

15

15

10

10

5

Al2O3 WDS

0 wt% 20

5

K2O LA-ICP-MS

0

lithium depletion

wt% 9 8 7

15

6 5

10

4 3

5

Na2O

0

2

Li2O

1 0

Fig. 4 BSE image and quantitative elemental distribution obtained by WDS-EPMA and LA-ICPMS on a polished section across the exchanged layer formed on a partially reacted SQI block. The elemental abundances are expressed as wt.% of the oxides

86

J. Gamage McEvoy et al.

stage #1: heat treatment: > 1000oC α→β transition ≈ 30 mins

single stage: 300 → 450oC ≤25 min

stage #2: β-spodumene decomposition in concentrated H2SO4 at 250 oC, ≈ 60 mins

Fig. 5 Comparison of lithium aluminosilicate decomposition pathways using the conventional process (red) and the proposed single-stage approach (blue), respectively

Conclusions Following specific temperature–time pathways within a narrow window between 400 and 500 °C, the decomposition of millimeter-sized α-spodumene crystals with excellent Li fractionation from Na and K can be achieved in less than 30 min. As illustrated in Fig. 5, compared to the conventional process, which involves a high-temperature decrepitation step (> 1000 °C) followed by an acid-baking treatment (~250–300 °C), the proposed low-temperature single-stage approach can lead to major reductions in energy consumption. Current efforts include optimization of the temperature path as well as the hydroxide reagent composition and concentration. Selective dissolution processes to improve the purity of the recovered lithium hydroxide are also being explored. Acknowledgements Financial support for this project was provided by Natural Resources Canada through a special fund for the Critical Minerals R&D Program. The authors would like to thank Derek Smith (CanmetMINING) for support with the XRD characterization, as well as Simon Jackson and Duane Petts (Geological Survey of Canada) for performing the LA-ICP-MS analyses.

Rethinking the Decomposition of Refractory Lithium Aluminosilicates: …

87

References 1. Choubey PK, Kim M-S, Srivastava RR, Lee J-C, Lee J-Y (2016) Advance review on the exploitation of the prominent energy-storage element: lithium. Part 1: from mineral and brine resources. Miner Eng 89:119–137 2. Li H, Eksteen J, Kuang G (2019) Recovery of lithium from mineral resources: State-of-the-art and perspectives—a review. Hydrometallurgy 189:105129 3. Ellestad RB, Leute KM (1950) Method of extracting lithium values from spodumene ores. US Patent 2516109. 25 July 1950 4. Lajoie-Leroux F, Dessemond C, Soucy G, Laroche N, Magnan J-F (2018) Impact of the impurities on lithium extraction from β-spodumene in the sulfuric acid process. Miner Eng 129:1–8 5. Yelantontsev D, Mukhachev A (2021) Processing of lithium ores: industrial technologies and case studies—a review. Hydrometallurgy 201:105578 ˇ 6. Cerný P, Ferguson RB (1972) The Tanco pegmatite Bernic Lake, Manitoba. IV. Petalite and spodumene relations. Can Miner 11:660–678 7. Abdelkader AM, Daher A, El-Kashef E (2008) Novel decomposition method for zircon. J Alloys Compd 460:577–580 8. Thibault Y, Gamage McEvoy J, Duguay D (2020) Optimizing Zr and REE recovery from zircon through a better understanding of the mechanisms governing its decomposition in alkali media. In: Azimi G, Forsberg K, Ouchi T, Kim H, Alam S, Baba A (eds) Rare metal technology 2020. The minerals, metals and materials series. Springer, Cham. https://link.springer.com/chapter/10. 1007/978-3-030-36758-9_11

Energy-Saving Green Technologies in the Mining and Mineral Processing Industry Shafiq Alam

Abstract Pyrometallurgy and hydrometallurgy are the two main processes to recover valuable metals from ore and secondary resources. While pyrometallurgy uses high temperatures to extract metals, in hydrometallurgy, an aqueous solution is used to extract metals from mineral resources. As a result, hydrometallurgical techniques are considered to be energy-saving technologies in the mining and mineral processing industry. This paper will address how energy can be reduced to extract metals in the mining and mineral processing industries with an emphasis on the extraction of base metals such as zinc and copper. Our novel techniques can even reduce more energy than what is currently consumed in the conventional hydrometallurgical copper electrowinning processes. Those energy-saving and sustainable green technologies would help meet the NetZero target in the mining and mineral processing industries. Keywords Energy · Mining · Mineral processing · Recycling/urban mining · Sustainability · GHG · NetZero

Introduction Mining and mineral processing industries consume huge energy that comes either from burning fossil fuels or by taking electricity from the grid. Therefore, those industries are directly and/or indirectly releasing a large volume of greenhouse gases (GHG) into the environment. There are many ways to save energy in mineral processing industries. Some are described below:

S. Alam (B) Department of Chemical and Biological Engineering, University of Saskatchewan, 57 Campus Drive, Saskatoon, SK S7N 5A9, Canada e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_9

89

90

S. Alam

Pyrometallurgy Versus Hydrometallurgy Metals are extracted from their resources, such as ore or waste materials, by physical separation of gangue materials followed by either pyrometallurgical or hydrometallurgical treatment. Pyrometallurgy is the thermal treatment process of minerals that involves the application of heat in the form of combustion of fuels or from electrical heat to liquefy the crushed ore/concentrate in a furnace. The most important unit operations of pyrometallurgy are calcination, roasting, smelting, and electrorefining which require high temperatures of over 1000 °C in different units, hence this process is energy intensive. On the other hand, hydrometallurgy employs the use of aqueous solutions (liquid solvents) to separate desired metal(s) from ores/concentrates, and recycled or residual materials. This process involves leaching to dissolve the desired metal(s) by acid/base, purification to remove the impurities from the solution of dissolved metals, and electrowinning to get pure metal in solid form. Other than normal operations of pumps and agitators, the only energy consumption is in the process of electrowinning, and in some cases during high-pressure leaching at a temperature of around 200 °C. As a result, the hydrometallurgical treatment is an energy-saving process. According to Habashi [1], pyrometallurgy is as ancient as our civilization when ancient people started to use fire. Many old pyrometallurgical plants around the world are still in operation; however, it is increasingly difficult to permit new installations because not only it is expensive to install but also it is an energy-intensive process and releases greenhouse gases to the environment. Although hydrometallurgy originated in the sixteenth century, this process was not used until the end of the nineteenth century. The development of robust solution purification methods such as ion exchange, solvent extraction, and other processes has led to an extremely broad range of applications of hydrometallurgy, now used to produce more than 70 metallic elements [2].

Primary Versus Secondary Feedstocks 1. Metals Recovery from Primary Feedstock Metals can be produced from both primary and secondary feedstocks. Primary feedstock comes from mines as ore that contains only a few percentages (~3%) of valuable metal(s), and the rest (~97%) are commercially undesirable/worthless materials called gangue. Those gangue materials need to be separated, at least partially, before feeding this ore into the process plant. In the mining industry, after the removal of overburden from the earth’s surface, ores are extracted by progressive removal of ore from the crust using heavy machinery. Sometimes the use of explosives is necessary to either extract or access an ore body. Those big chunks (~1.5 m) of ores are crushed and ground to about

Energy-Saving Green Technologies in the Mining and Mineral …

91

40 µm to liberate valuable minerals from the rocks following the physical separation processes to separate gangue materials, hence making them concentrate that is fed to the pyro/hydro-metallurgical process plants for the production of pure metal(s). Figure 1 shows how metals are produced from ore after mining. Physical separation of valuable minerals from gangue materials is called beneficiation which includes milling (crushing and grinding) and flotation. This part of mineral processing is very costly as milling is an energy-intensive process. Table 1 shows a typical relative cost of beneficiating an ore. 2. Metals Recovery from Secondary Feedstock Secondary feedstock contains raw materials from waste products sent to landfill, such as spent appliances, electronic and electrical waste (e-waste), automobile scraps, waste batteries, spent catalysts, etc. The amount of metal content in these wastes is

ORE

Beneficiation (Physical Separation)

CONCENTRATE

Crushing Grinding Flotation Dewatering Drying

GANGUE

ORE

Extraction Process (Pyro- or HydroMetallurgy)

METAL

TAILINGS

METAL

Fig. 1 Ore to metal (pictures collected from the public domain [3])

Table 1 Typical cost of beneficiating an ore [3]

Operations

Operating cost (%)

Crushing

5–20

Grinding

25–75

Flotation

25–45

Dewatering and drying

10–20

Other operations

5–10

92

WASTE

S. Alam

Physical Separation

NON-METALLIC COMPONENTS

METALLIC PARTS

Extraction Process (Pyro- or HydroMetallurgy)

METAL

TAILINGS Shredding Sorting Magnetic Separation

Fig. 2 Metal production from waste materials

much higher than that in its ore. For example, the waste printed circuit board (PCB) of mobile phones contains 10 times higher precious metals than those of rich-content ores [4–6]. While copper ore usually contains about 0.5–2% copper [7], the PCB itself contains 10–25% copper depending on the source and type of the circuit board [8]. Unlike virgin ore, there are no or very fewer gangue minerals in the secondary resources, hence there is no need for an energy-intensive beneficiation process, and the flowsheet is simplified as shown in Fig. 2. Therefore, metals recovery and recycling from waste materials are now a growing trend for economic benefit because the capital and operating costs are much lower for the recycling of metals from scrap rather than production from virgin ore. Also, recycling of waste materials helps reduce greenhouse gases (GHG) as the wastes in landfills break down and decompose producing methane and CO2 . Since the flow of those waste streams is coming from residential and commercial sectors in municipal areas, they are also called urban mining. Recycling of waste materials saves and conserves primary resources. For a million cell phones, we can recover gold weighing 34 kg, silver of 350 kg, copper of 16,000 kg, and palladium of 15 kg [9]. Also, recycling reduces energy, for example, for a million laptops we recycle, we can save the equivalent of electric power capable of running 3657 households for one year [9]. Elsa Olivetti mentioned that up to 90% less energy can be required for the production of aluminum from its waste materials [10]. In our past research, an innovative process was developed to recycle iron and zinc from automobile scraps to manufacture newer steel products and zinc metal that were described in the following sections.

Process Options Innovative research and process options can help reduce energy consumption in mining and mineral processing industries. As mentioned earlier, even though

Energy-Saving Green Technologies in the Mining and Mineral …

93

hydrometallurgy is an energy-saving technology when compared with pyrometallurgy, researchers can make it more sustainable with their innovative process design. In the hydrometallurgical process, desired metals are dissolved in solution using low concentrated acid/base. This method is called leaching. After solid–liquid separation, the metal-containing solution called pregnant leach solution (PLS) is purified and it is then sent to the electrowinning section where electricity is used to get the solid form of metals for marketing. Figure 3 shows a conventional copper (Cu) electrowinning cell, where copper is produced from the PLS. In this process, ionic copper (Cu2+ ) in the solution is electrodeposited as metallic copper (Cuo ) on the cathode taking two electrons from the circuit, and a counter-reaction has taken place at the anode to hydrolyze the water. In the industry, hundreds of such electrowinning cells are used side by side in the tank house to produce bulk quantities of copper at a time as shown in Fig. 4. As mentioned earlier, the electrowinning process consumes the most energy in the hydrometallurgical technique, this is because of the current practice of the anodic reaction to hydrolyze the water. However, our innovative research on an alternative anode reaction technology (AART) showed a significant reduction in energy consumption in the electrowinning process that was described in the following sections. Voltage Source

Fig. 3 Copper electrowinning cell with reactions

Electrons Current

+ Anode

Electrolyte

SO42-

Cu2+ Cations

CuSO4 + 2e − ⎯ ⎯→ Cu 0 + SO4 −

+

Anions

−2

H2 O ⎯ ⎯→ 2e + 2 H + 1 / 2O2↑ SO 4

−2

+

+ 2H ⎯ ⎯→ H 2 SO4

Cathode

Cathode Anode hydrolysis of water Anode acid creation

⎯→ H 2 SO4 + 1 / 2O2↑ Overall reaction at anode C uSO 4 + H 2 O ⎯

94

S. Alam

Fig. 4 Copper electrowinning tank house

Zinc Recovery from Secondary Resources The recycling of steel leads to substantial energy savings and significant reductions in CO2 emissions. In 2019, U.S. steel consumption was 100 million tons [11]. The global market for steel scrap is estimated at 574.5 million tons in the year 2020, which is projected to reach 748.2 million tons by 2026. About 80% of the world’s zinc (Zn) is used for galvanizing steel, which is used mostly for making car bodies. Steel scrap contains about 3% Zn on it, and recycling of 574.5 million tons of galvanized steel scrap per year gives 17.24 million tons of zinc and will be 22.45 million tons of zinc per year by 2026 with increased recycling of steel scrap. In our previous work, we developed a hydrometallurgical process to recycle iron and zinc from steel scrap, which is a secondary resource that comes from automobile junkyards [12]. It was found that the energy consumption to produce 17.24 million tons of zinc from secondary resources using hydrometallurgical processes is 0.32 × 109 GJ, which is about one-third less than the production of the same amount of zinc from primary resources that consumes 1.05 × 109 GJ (Calculated based on 61 GJ/t of zinc production from low-grade zinc ore [13, 14]).

Energy-Saving Green Technologies in the Mining and Mineral …

95

Copper Recovery Using Novel Process Options In our innovative research, a novel process has been developed using AART technology to recover copper from the printed circuit board (PCB) from e-wastes. In that process, copper from e-wastes is leached in an alkaline solution followed by purifications and energy-saving monovalent copper electrowinning processes [8]. In that process, the power consumption during the electrowinning of Cu+ was found to be 1300 and 500 kWh/ton in the sulfate and chloride systems, respectively, which are much lower than that of conventional Cu2+ electrowinning, i.e., 2000–2200 kWh/ton [15]. This is because in the anode Cu+ is oxidized to Cu2+ rather than water hydrolysis, which is used in the conventional process that takes much higher energy to decompose. Also, at the cathode, Cu+ is electrodeposited taking one electron rather than two as shown earlier in Fig. 3 for Cu2+ .

Conclusion Sustainable processes have been developed for zinc and copper recovery from secondary resources using hydrometallurgical techniques. Both processes showed low power consumption for the overall process as well as in the electrowinning cell. It also decreases greenhouse gas production. Our sustainable processes can help meet the goal of NetZero.

References 1. Habashi F (2013) Fire and the art of metals: a short history of pyrometallurgy, pp 165–171. https://doi.org/10.1179/037195505X63358 2. https://www.britannica.com/technology/hydrometallurgy. Accessed 2 Sept 2022 3. Schwartz L (2010) GeoE 498 Introduction to mining and mineral processing engineering lecture. University of Saskatchewan, Canada 4. Li J, Lu H, Guo J, Xu Z, Zhou Y (2007) Recycle technology for recovering recourses and products from waste printed circuit boards. Environ Sci Technol 41:1995–2000 5. Guo J, Rao Q, Xu Z (2008) Application of glass-nonmetals of waste printed circuit boards to produce phenolic moulding compound. J Hazard Mater 153:728–734 6. He W, Li G, Ma X, Wang H, Huang J, Xu M, Huang C (2006) WEEE recovery strategies and the WEEE treatment status in China. J Hazard Mater B136:502–512 7. https://www.sciencedirect.com/topics/earth-and-planetary-sciences/copper-ore. Accessed 2 Sep 2022 8. Alam MS, Tanaka M, Koyama K, Oishi T, Lee J-C (2007) Electrolyte purification in energysaving monovalent copper electrowinning processes. Hydrometallurgy 87:36–44 9. https://www.conserve-energy-future.com/e-waste-recycling-process.php#:~:text=E%2Dw aste%20recycling%20is%20the,pollution%20impacts%20of%20e%2Dwaste. Accessed 12 Sep 2022 10. Olivetti E (2022) Department of Materials Science and Engineering, MIT. https://olivetti.mit. edu/. Accessed 4 Oct 2022

96

S. Alam

11. https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-iron-steel-scrap.pdf. Accessed 12 Sep 2022 12. Lakshmanan VI, Sridhar R, Alam MS (2005) Recovery of zinc from galvanized coatings. Canadian Patent Serial No. CA 2347552, Issued on 19 Apr 2005 13. https://www.researchgate.net/figure/Energy-requirements-in-GJ-t-for-copper-zinc-and-goldproduction-as-a-function-of-the_fig2_309731859. Accessed 12 Sep 2022 14. Energy and environmental profile of the U.S. mining industry. https://www.energy.gov/sites/ prod/files/2013/11/f4/lead_zinc.pdf. Accessed 12 Sep 2022 15. Prasad MS, Kenyen VP, Assar DN (1992) Development of SX–EW process for copper recovery—an overview. Miner Process Extr Metall Rev 8:95–118

Extraction of Valuable Metals from Luanshya Copper Smelting Slag with Minimal Waste Generation Yaki Chiyokoma Namiluko, Yotamu Rainford Stephen Hara, Agabu Shane, Makwenda Thelma Ngomba, Ireen Musukwa, Alexander Old, Ronald Hara, Rainford Hara, and Stephen Parirenyatwa

Abstract An efficient method of separating valuable elements (copper, cobalt, chromium, and sulphur) from copper smelting slag of Luanshya district of the Copperbelt province in Zambia has been developed. The as-received slag material was characterised via scanning electron microscope. The valuable elements were separated through a combination of magnetic separation, flotation, and gravity separation steps. Magnetic separation of the as-received material separates cobalt/iron and copper/sulphur/chromium rich fractions due to differences in magnetic properties. Residual copper in the magnetic fraction was upgraded to more 25 weight% via flotation. By comparison, flotation of the non-magnetic fraction yielded low grade copper concentrate due to high presence of sulphur. Chromium was upgraded by a factor of more than 5 when the non-magnetic fraction was subjected to gravity concentration. The effect of particle size was studied during magnetic separation of feed material. Keywords Cobalt · Luanshya chromium · Slag · Zambia · Waste

Introduction Copper smelting slag from Luanshya district of the Copperbelt province of Zambia has significant amounts of copper, cobalt, chromium, and elemental sulphur. Copper grade varies between 0.6 and 3 weight%, and it was lost to the slag due to inefficiency of the process. On the other hand, cobalt was lost to the slag as a result of oxidising [1] environment during copper smelting and converting processes. The grade of cobalt is 0.1–0.4 weight%. Chemical analysis of various samples from Luanshya copper smelting slag has revealed high presence of chromium (more than 0.5 weight%). Chromium reported to the slag through two methods; (i) chemical attack between furnace lining (chrome–magnesite bricks) and molten slag and (ii) concentrate feed Y. C. Namiluko · Y. R. S. Hara (B) · A. Shane · M. T. Ngomba · I. Musukwa · A. Old · R. Hara · R. Hara · S. Parirenyatwa Copperbelt University, Kitwe, Zambia e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_10

97

98

Y. C. Namiluko et al.

material. Surprisingly, elemental sulphur is very high in Luanshya slag and this was dumped as fines from another unknown process. Considering the high demand of copper and cobalt worldwide today, it is necessary to recover these elements from copper smelting slags. Even though chromium metal is one of the important element in steel industry, it is very toxic [2] especially in the hexavalent state [3]. Trivalent chromium is generally considered safe but the danger is that it can oxidise to hexavalent state over a period of time. Bearing in mind both the demand of chromium in the steel industry and high levels of this element in Luanshya slag, it is necessary to investigate a method that can recovery chromium alongside copper and cobalt.

Processing of Smelting Slag About 50–70 weight% of the copper in slag exist in sulphide (matte) form while the rest is in the silicate form as copper was lost due to mechanical entrainment and oxidation, respectively [1, 4]. Due to the fact a majority of copper in the slag is in the matte form, copper slags in Zambia are processed via froth flotation [4, 5]. However, Kafre Enterprises run pilot test works to float Luanshya slag at a throughput rate of 200 tons per day. The following challenges were encountered: • Recovery of copper was very poor (less than 50%) as most of the copper was lost to the tailings stream. • The grade of copper in the concentrate was low (8–15 weight% copper) even after using more selective collectors. The poor recovery of copper was attributed to the recovery of elemental sulphur into the concentrate. • Both cobalt and chromium were lost to the tailings to the extent that the tailings had 0.8–1 weight% chromium. Iron and cobalt are magnetic metals [6, 7] while sulphur and copper and nonmagnetic. As such magnetic separation of the as-received material may permit beneficiation of cobalt and iron [8] from the slag material.

Experiment Material Material used was obtained from Luanshya copper smelting slag. The material was collected from 100 different locations in order to obtain a representative sample. The as-received material was homogenised by crushing down to particle size of less than 2 mm. Chemical composition showing the major constituents of the as-received sample is shown in Table 1. It can be observed from Table 1 that the material contains

Extraction of Valuable Metals from Luanshya Copper Smelting Slag …

99

Table 1 Chemical analysis in weight% of the as-received Luanshya slag TCu

TCo

Fe

S

Cr

SiO2

Al2 O3

CaO

MgO

K2 O

MnO

2.11

0.17

18.13

6.26

0.86

40.69

8.19

9.76

0.10

2.36

0.18

Table 2 Summary of chemical dosing during froth flotation

Reagents

Kg/ton

G/ton

Sodium hydrogen sulphide

0.88

–

Sodium ethyl xanthate (SEX)

0.36

–

Sodium amyl xanthate (SAX)

0.23

–

Pine oil

0.15

–

Senkol

–

0.09

Starch

0.1

–

2.11 weight% total copper (TCu). Sulphur content is higher (6.26 weight%) than any other copper smelting slag that has been reported [9–14], and this is due to presence of fine elemental sulphur. The major gangue constituents are SiO2 , CaO, Al2 O3 , and Fe, and this typical of any other copper smelting slags [11, 15]. The results in Table 1 further show that the slag has higher content of chromium.

Froth Flotation Test Works The purpose of froth flotation test works was to recover copper as a concentrated. The as-received material was milled to particle size of 60% pass through 75 microns sieve. The material was floated using Metso flotation machine at a pulp density of 33%. As part of the material is oxide, a sulphidiser (sodium hydrogen sulphide) was used during froth flotation. The summary of reagent dosing is shown in Table 2.

Gravity Separation The tailings from froth flotation were subjected to spiral concentration. Spiral concentration was carried out at pulp density of 20%. The rougher concentrate and tailings were cleaned and scavenged, respectively, twice so as to maximise the recovery of valuable elements.

100

Y. C. Namiluko et al.

Magnetic Separation Due to the presence of cobalt and iron, copper smelting slags are magnetic. Therefore, magnetic separation was carried for the purpose of separating magnetic from nonmagnetic fraction. Magnetic separation was carried at different particle sizes but at strength of 3000 gauss. Three streams were produced namely magnetic, weakly magnetic, and non-magnetic fractions.

Chemical Analysis For chemical analysis, a representative portion of the sample was digested in acid and analysed by titration, gravimetric, and atomic absorption spectrometer (Perkin Elmer Analysis 300). The phases in the samples were characterised using scanning electron microscope. The as-received sample was cold mounted in epoxy resin. The mounted sample was ground and polished down to less than 1 µm. The sample was scanned under backscattered electron imaging in which the heavy and light phases (particles) appear bright and dark, respectively. The individual phases were quantified via SEM-EDX point analysis.

Results and Discussion Scanning electron microscope image under backscattered electron imaging for the asreceived sample is shown in Fig. 1. Several phases can be observed in Fig. 1. Energy dispersive X-ray point analysis showed that cobalt is predominantly contained in iron–silicate phase (fayalite). On the other hand, copper as matte (Cu5 FeS4 and Cu2 S) and silicate phase. Some of the copper matte particles are free while some are dispersed in the iron silicate phase as shown in area E of Fig. 1b.

Magnetic Separation of Feed Material The results for magnetic separation at different particle sizes are shown in Fig. 2 from which the following important observations can be made: 1. The recovery of cobalt in the concentrate is higher than that of copper and chromium, and this is expected as copper and cobalt are non-magnetic. 2. The recovery of cobalt into the concentrate increases continually with increase in particle size while the recoveries of copper and chromium increase with increase in particle size and reach maximum at particle size of 1.18 mm.

Extraction of Valuable Metals from Luanshya Copper Smelting Slag …

101

Fig. 1 Scanning electron microscope under backscattered electron imaging for the as-received sample 1.8

110

(a)

1.6

Co

% Recovery

90 80 70

Cr

60

(b)

1.2 1.0 0.8 0.6

50 40

Cu

1.4

% Grade

100

0.4

Cu

0.2

30 20 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5

Particle Size (mm)

Cr

Co 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5

Particle Size (mm)

Fig. 2 Plot of a % recovery and b % grade in the concentrate against particle size

3. The grade of copper in the concentrate is higher than that of cobalt and chromium. 4. The grades of copper, cobalt and chromium are highest at particle size of 1.18 mm. Based on the results in Fig. 2, it can be concluded that the optimum particle size for magnetic separation is 1.18 mm. Summary of results for the magnetic separation of the ground as-received material at particle size of 2.36 and 5.00 mm is presented in Table 3. The results in Table 3 clearly show that cobalt and iron are mostly recovered and hence concentrated in the magnetic fraction. On the other hand, chromium and sulphur are concentrated in the non-magnetic fraction as they are non-magnetic elements.

102

Y. C. Namiluko et al.

Table 3 Summary of the magnetic separation results at particle size of 2.36 and 5.00 mm Particle size (mm)

Fraction

−5.00

Magnetic

445.53

Non-magnetic

376.48

Total

822.01

Magnetic

400.65

Non-magnetic

394.21

Total

795.43

− 2.36

Weight (g)

%Weight

Grade (weight%) Cu

Co

Fe

Cr

54.20

1.21

0.3

28.824

0.34

0.86

45.80

2.46

0.01

10.678

1.17

12.81

–

–

50.69

1.63

0.29

28.868

0.36

0.32

49.54

2.43

0.01

11.315

1.19

12.49

–

–

–

–

–

–

–

–

S

–

–

Froth Flotation of Magnetic and Non-Magnetic Fractions Froth flotation was carried out separately, on the magnetic and non-magnetic fraction and the results presented in Tables 4 and 5, respectively. It can be observed in Table 4 that copper was upgraded to 26.16 weight% in the concentrate with a cumulative recovery (Cum Rec.) of 60.57 weight%. It is worth noting that the concentrate grade that was obtained from froth flotation of magnetic fraction is higher than what is obtained without magnetic fraction. Even though the grade of chromium is nearly the same in all fractions, 87.73 weight% reported to the tailings stream. By comparison, the grade of cobalt was lowest in the concentrate (Conc.) stream. The results in Table 4 also show that grade of sulphur is highest and lowest in the concentrate and final tailings (FT) streams, respectively. Results that were obtained after froth flotation of the non-magnetic fraction are shown in Table 5, and it can be noted that the grade of copper in the concentrate is 10.66 weight%. This grade is similar to what was obtained when Luanshya is floated at commercial scale without prior magnetic separation. Sulphur is upgraded Table 4 Froth flotation results of the magnetic fraction Fraction

Wt (g)

%Wt

Grade (%Weight)

% Rec

Cu

Cr

Co

S

Cu

% Cum Rec Cr

Co

Conc

7.16

1.86

26.16

0.28

0.19

6.47

32.39

1.78

1.44

RCT

3.73

0.97

14.75

0.30

0.25

3.80

9.51

1.00

0.95

CT 2

3.4

0.88

5.66

0.29

0.27

2.48

3.33

0.88

0.96

CT

14.55

3.78

2.01

0.26

0.27

1.47

5.06

3.39

4.03

SCV 1

12.57

3.27

2.65

0.24

0.27

1.96

5.75

2.63

3.48

SCV 2

13.37

3.48

1.97

0.22

0.28

1.29

4.53

2.58

3.88

FT

329.85

85.76

0.69

0.30

0.25

0.75

39.43

87.73

85.27

Total

384.63

1.32

0.29

0.25

0.98

–

–

–

Cu

Cr

60.57

12.27

–

–

–

–

Extraction of Valuable Metals from Luanshya Copper Smelting Slag …

103

Table 5 Froth flotation results of the non-magnetic fraction Fraction

Wt (g)

% Wt

Grade (%Weight)

% Rec

Cu

Cr

Co

S

Cu

% Cum Rec Cr

Co

Conc

11.55

2.96

10.66

0.37

0.03

14.21

14.66

1.66

1.09

RCT

13.55

3.47

9.57

0.43

0.06

9.97

15.44

2.31

2.24

CT 2

12.75

3.27

7.41

0.51

0.09

3.85

11.25

2.54

3.25

CT

18.8

4.82

4.17

0.78

0.09

2.70

9.34

5.72

4.42

SCV 1

14.23

3.65

4.43

0.65

0.08

3.16

7.50

3.64

2.96

SCV 2

12.52

3.21

4.18

0.69

0.07

2.29

6.23

3.40

2.35

SCV 3

2.35

0.60

0.97

1.33

0.04

1.40

0.27

1.23

0.25

FT

304.27

78.01

0.98

0.67

0.10

0.87

35.30

79.50

83.44

Total

390.02

1.89

–

–

–

–

–

–

Cu

Cr

64.7

1.66

–

–

–

–

to 14.66 weight% in the concentrate. The higher grade of sulphur in the concentrate might attribute to low grade of copper in the concentrate. On the other hand, chromium was upgraded more in the scavenger 3 and final tailings.

Gravity Separation of Non-magnetic Fraction Froth flotation results have shown non-magnetic fraction does not respond well to froth flotation owing to high content of sulphur. Therefore, non-magnetic fraction was subjected to gravity separation (table shaking and spiral concentration). The summary of results for table shaking is shown in Table 6. It can be observed from Table 5 that two concentrates namely conc 1 and 2 were obtained. Conc 1 has the highest content of chromium (8.64 weight%). Conc 2 has 1.67 and 0.91 weight% copper and sulphur, respectively. As a result, conc 2 may be taken for flotation. Scavenger (Scav) has 73.34 weight% sulphur. The final tails have 0.34 weight% chromium. Based on the results in Table 6, it can be concluded that table shaking of the non-magnetic fraction may permit separation of chromium and sulphur rich separate streams which may be taken for further processing.

Process Flowsheet The proposed process flowsheet for processing Luanshya copper smelting slag is shown in Fig. 3. The material is first treated by magnetic separation followed by froth flotation and then spiral concentration. As cobalt concentrates together with iron, minimal waste will be generated if the material is processed according to the proposed process flowsheet in Fig. 3.

104

Y. C. Namiluko et al.

Table 6 Table shaking of the non-magnetic fraction Stream

Weight (g)

%Cu

%Co

%Cr

Conc 1

2.715

4.38

0.12

8.64

Conc 2 Cleaner tails Middlings Scav tails Final tails

%S 1.07

114.17

1.67

0.03

0.91

0.91

26.34

1.36

0.00

0.10

32.59

7.19

2.24

0.06

0.35

13.27

2.475

1.50

0.04

0.13

73.34

3.61

0.07

0.34

2.03

20.75

Fig. 3 Proposed process flowsheet for processing Luanshya copper smelting slag

Conclusions and Recommendations Conclusions Based on the laboratory test works carried out, the following conclusions were made. 1. Magnetic separation of the crushed feed material revealed the following: a. Nearly all cobalt and most of iron are recovered to the magnetic fraction. Cobalt was upgraded by factor of 2. b. 50–60% of the copper is recovered to the magnetic fraction under optimised particle size. c. A majority of chromium and sulphur are recovered to the non-magnetic fraction. 1. Froth flotation of the magnetic fraction yields a higher grade copper concentrate (> 25 weight% copper) which is suitable for smelting. On the other hand, the grade of copper concentrate was low after flotation of the non-magnetic fraction.

Extraction of Valuable Metals from Luanshya Copper Smelting Slag …

105

2. Sulphur and chromium rich separate streams or concentrates were produced after table shaking of the non-magnetic fraction of the feed material.

Recommendations From the test works carried out, it is recommended that Luanshya copper smelting slag should be processed via a combination of magnetic fraction, froth flotation, and gravity separation.

References 1. Davenport WGL et al (2002) Extractive metallurgy of copper, 4th edn. Chemical, Petrochemical and Process, Elsevier, pp 1–452 2. Laxmi V, Kaushik G (2020) toxicity of hexavalent chromium in environment, health threats, and its bioremediation and detoxification from tannery wastewater for environmental safety. In: Saxena G, Bharagava R (eds) Bioremediation of industrial waste for environmental safety 3. Shrivastava R et al. (2002) Effects of chromium on the immune system. FEMS Immunol Med Microbiol 34(1):1–7 4. Yotamu RS, Hara AJ (2016) Carbothermic processing of copper–cobalt mineral sulphide concentrates and slag waste for the extraction of metallic values. CRC Press, Boca Raton 5. Das BMB, Angadi S, Pradhan SK, Prakash S, Mohanty J (2009) Characterization and recovery of copper values from discarded slag. Waste Manag Res 6:561–567 6. Betteridge W (1982) Cobalt and its alloys, 1st edn. Industrial metals. Ellis Horwood Limited, Chichester 7. Anon (1960) Cobalt monograph, vol 515. Centre d’Information du Cobalt, Brussels, Belgium 8. Wang J-P, Erdenebold U (2020) A study on reduction of copper smelting slag by carbon for recycling into metal values and cement raw material. Sustainability 12(4):1421 9. Altundoˇgan HS, Tümen F (1997) Metal recovery from copper converter slag by roasting with ferric sulphate. Hydrometallurgy 44(1–2):261–267 10. Deng T, Ling Y (2007) Processing of copper converter slag for metal reclamation. Part I: extraction and recovery of copper and cobalt. Waste Manage Res 25(5):440–448 11. Zhai X-J et al. (2011) Recovery of cobalt from converter slag of Chambishi copper smelter using reduction smelting process. Trans Nonferrous Met Soc China 21(9):2117–2121 12. Sukla LB, Panda SC, Jena PK (1986) Recovery of cobalt, nickel and copper from converter slag through roasting with ammonium sulphate and sulphuric acid. Hydrometallurgy 16(2):153–165 13. Arslan C, Arslan F (2002) Recovery of copper, cobalt, and zinc from copper smelter and converter slags. Hydrometallurgy 67(1):1–7 14. Tümen F, Bailey N (1990) Recovery of metal values from copper smelter slags by roasting with pyrite. Hydrometallurgy 25(3):317–328 15. Ettler VE et al. (2022) Cobalt-bearing copper slags from Luanshya (Zambian Copperbelt): mineralogy, geochemistry and potential recovery of critical metals. J Geochem Explor 237

Carbon Footprint Assessment of Waste PCB Recycling Through Black Copper Smelting in Australia A. Q. Mairizal, A. Y. Sembada, K. M. Tse, N. Haque, and M. A. Rhamdhani

Abstract Electronic waste (e-waste) is one of the fastest growing waste streams in Australia. The high potential value of precious metals in e-waste is a factor in the development of an e-waste recycling process facility in Australia. Preliminary environmental impact analysis using net carbon footprint as an indicator for developing comprehensive waste PCB processing facilities in Australia has been carried out and presented in this paper. The paper analyses the current situation of e-waste management (focused on waste PCB) based on three different scenarios: (1) recycling of waste PCB in a small-scale facility, (2) recycling of waste PCB integrated with industry, and (3) recycling of waste PCB in a centralized and large recycling facility. The total carbon footprints of these scenarios were estimated to be in the range of 1.96–3.76 (kg CO2 -eq/kg Cu). The transport of key materials to the plant was found to only contribute a little to the overall carbon emission. Reduction of carbon emission by 18–31% was estimated when renewable energy sources were used for supplying the electricity for the process. Keywords Carbon footprint · Black copper smelting · Recycling · Waste PCB · Australia

A. Q. Mairizal (B) · K. M. Tse · M. A. Rhamdhani Department of Mechanical and Product Design Engineering, School of Engineering, Swinburne University of Technology, Melbourne, VIC 3122, Australia e-mail: [email protected] M. A. Rhamdhani e-mail: [email protected] A. Y. Sembada Department of Management and Marketing, School of Business, Swinburne University of Technology, Melbourne, VIC 3122, Australia N. Haque CSIRO Energy, Research Way, Clayton, VIC 3168, Australia © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_11

107

108

A. Q. Mairizal et al.

Introduction Electronic waste (e-waste) is one of the fastest growing waste streams in Australia. The rising consumption on electrical and electronic equipment combined with technological innovations has resulted in a large volume of e-waste in the country. It was estimated that the e-waste generation in Australia has reached up to 21.3 kg per capita, making Australia as one of the highest e-waste generation per capita in the world [1]. It has also been estimated that the amount of e-waste generation in Australia will continue to increase up to 3.5 million tons (Mt) in 2047 with the increase rate of 3–5% annually [2]. This high volume of e-waste generation in Australia may put a significant pressure on e-waste management system to prevent the escalating problems from e-waste in the future. E-waste contains a wide range of materials from metals, plastic, and other substances. E-waste can contain up to 60 elements [3], including the base metals (BMs) of Cu, Pb, Al, Ni, precious metals (PMs) Au, Ag, Pd, Pt, rare earth elements (REE) Nd, Gd, Ce, and Dy, and hazardous (halogens, Hg, and Cd). Due to the presence of hazardous materials, disposing of e-waste in a landfill can cause severe consequences on the human health and environment. However, e-waste also contains valuable metals hence attract its recycling to recover the wealth. These valuable metals on e-waste mostly located in waste PCB (printed circuit boards). The high potential of metal recovery from e-waste in Australia has attracted research that focus on quantification of e-waste generation potential in Australia [2] and e-waste recycling technologies. The high content of copper in waste PCB and the compatible chemistry of the precious metals (Ag, Au) with copper makes waste PCB recycling suitable to be embedded with copper smelting processes. One of the most common routes in industrial practice for waste PCB recycling process is through black copper smelting that is commonly used for recycling copper scrap [4]. However, the metal recovery of waste PCB also requires resource and energy consumption for the process. Most of the current studies on e-waste processing technologies are more focused on metal recovery rates, based on economic perspective [5, 6], and more fundamental on the distribution of the valuable elements in the conditions relevant to industrial process [7, 8]. In order to evaluate the impact of waste PCB processing on the environment, the current study provides a generic carbon footprint assessment of the recovery of copper, gold, and silver from waste PCB during its recycling through black copper smelting. Various scenarios of waste PCB recycling in an Australian setting were simulated to investigate their overall carbon footprint emission. Three scenarios were investigated: (1) recycling of waste PCB in a small-scale facility; (2) recycling of waste PCB integrated with industry; and (3) recycling of waste PCB in a centralized and large waste facility. Scenarios with electricity sourced from different renewable energies were also compared and analyzed in this study. Based on the analysis of the recycling process, the energy consumption, material transport, and carbon emission

Carbon Footprint Assessment of Waste PCB Recycling Through Black …

109

of each copper production, the carbon footprints of waste PCB recycling process in Australia were established.

Process Description In this study, a pyrometallurgical route of the secondary copper smelting process, often known as the black copper smelting process, was selected for the carbon footprint assessment. The definition of secondary copper is based on the input feed material of the process, which mostly comes from a secondary source such as copper scrap and waste PCB. In industrial practice, waste PCB can be mixed with other base metal scraps such as copper, lead, and nickel for the metals recovery process. The primary input of the process is copper scrap, waste PCB, flux, metallurgical coke, and air. The waste PCB recycling through the black copper smelting process consists of four stages including reduction, oxidation, electrorefining, and precious metal refining process. The first stage is the reduction process in which the input materials are added to a reduction furnace and treated under a reducing condition at 1300 °C. The product of the reduction process is also known as black copper. The black copper is transferred to the next process of oxidation, where the impurities such as Zn, Sn, and Pb are separated from liquid copper and removed as gas phases or slag. It should be noted that in industrial practice, the reduction and oxidation processes can take place in a single reactor or two separate reactors. The anode liquid copper from the oxidation process is then purified through an electrorefining process to produce a higher purity of cathode copper up to 99.9%. The slime, which contains precious metals, as a by-product of the electrorefining process, is collected and treated to recover precious metals such as gold and silver. The process flowsheet and mass balance of the waste PCB recycling through black copper smelting can be found in Fig. 1. The mass balance equation of black copper smelting process is adopted from [5].

Methodology Goal Definition The goal of the study is to assess the environmental impact of waste PCB recycling through black copper smelting process in the context of Australia. The overall carbon footprint of different scenarios of waste PCB recycling in Australia was calculated. The procedure used to quantify the carbon footprint in this study was based on the Australian National Greenhouse Account Factor [9]. A wide range of Life Cycle Inventory (LCI) data for the analysis was taken from the literature and Ecoinvent database.

110

A. Q. Mairizal et al.

Fig. 1 Process flowsheet and mass balance of waste PCB recycling via black copper smelting

System Boundaries and Scope of the Study The system boundary and the scope of this study include transportation, mechanical treatment, material input, black copper smelting process, and refining process, as shown in Fig. 2; and the overall process is classified into six stages. All the related input–output of the materials and energy for each stage were included to determine the entire system’s environmental impact. According to the Australian National Greenhouse Account Factor [9], the carbon emissions of metal processing plants can be estimated using a mass balance approach. This approach involves combining the total quantities of carbon entering, leaving, and consumed during the process, as well as the carbon emission released during the combustion process.

Functional Unit The functional unit used in the present study was defined per one kg of copper produced (kg Cu). The quantification of CO2 emission in this study was expressed in terms of kg CO2 -eq per kg of copper produced (kg CO2 -eq/kg Cu).

Carbon Footprint Assessment of Waste PCB Recycling Through Black …

111

Fig. 2 System boundary and scope in this study

Scenario and Case Description In this study, the waste PCB recycling was considered based on the three scenarios presented in more detail in Table 1. These scenarios were considered to support a recent work by the authors [10] which focuses only on the techno-economic analyses of waste PCB recycling with the same scenarios.

Carbon Footprint Assessment To assess the carbon footprint of waste PCB recycling process, the associated carbon input and carbon output flow into and out of the process were calculated using the following steps and equations: 1. Direct emission from fuel combustion in tonnes during one year of activity as follows:  Q i × ADw × EFw,i × GWPi (1) Ow = i

Ow represents the total CO2 equivalent produced by fuel combustion. Q i is the quantity of fuel type I (tonne). ADw is the energy content of fuel (GJ/tonne). EFw,i is the emission factor of greenhouse gas i for fuel combustion for the processes provided in Table 2. GWPi is the 100-year global warming of greenhouse gas i. It was assumed that the fuel was fully combusted. 2. Direct emission from carbon dioxide released during one year of operation based on mass balance presented in Fig. 1. 3. Indirect emission from electricity consumption during one year of operation. The greenhouse gas (GHG) emission model generated as follows:

112

A. Q. Mairizal et al.

Table 1 Scenarios considered in the current study Name

Scenario 1

Plant capacity

10,000 tonnes per year 10,000 tonnes per year 100,000 tonnes per (tpy) (tpy) year (tpy)

Scenario 2

Scenario 3

Plant location

Sydney, NSW

Sydney, NSW

Sydney, NSW

Material input

1. Copper scrap (48 wt%) 2. Waste PCB (48 wt%) 3. Flux/slag (3.4 wt%) 4. Coke (0.6 wt%) 5. Air

1. Copper scrap (48 wt%) 2. Waste PCB (48 wt%) 3. Flux/slag (3.4 wt%) 4. Coke (0.6 wt%) 5. Air

1. Copper scrap (92 wt%) 2. Waste PCB (4 wt%) 3. Flux/slag (3.4 wt%) 4. Coke (0.6 wt%) 5. Air

Scope and boundary 1. Material transport to Sydney, NSW 2. Mechanical treatment 3. Black copper smelting process 4. Electrorefining process 5. Precious metal refining process

1. Material transport to Sydney, NSW 2. Mechanical treatment 3. Black copper smelting process 4. Electrorefining process 5. Material transport to existing large smelter, e.g. at Olympic Dam

1. Material transport to Sydney, NSW 2. Mechanical treatment 3. Black copper smelting process 4. Electrorefining process 5. Precious metal refining process

Products

Cu, Au, Ag

Black Cu

Cu, Au, Ag

Functional unit

1 kg Cu

1 kg Cu

1 kg Cu

Database

Ecoinvent (SimaPro)

Ecoinvent (SimaPro)

Ecoinvent (SimaPro)

Carbon footprint

(kg CO2 -eq/kg Cu)

(kg CO2 -eq/kg Cu)

(kg CO2 -eq/kg Cu)

Table 2 Emission factors used for the calculation in the current study Name

Energy content factor (GJ/t)

Emission factor (EF) CO2 -e (kg CO2 /GJ)

References

Coke

27

107

[9]

Waste PCB Polyethylene (15 wt%)

26.3

82

[9]

Polypropylene (5 wt%)

26.3

82

[9]

Op =



 E mp + E sp + E ep + E pp × EF p,i × GWPi

(2)

i

where O p represents the total CO2 emission equivalent based on electricity consumption during recycling process. E mp + E sp + E ep + E pp represents the electricity consumption during mechanical process, black copper smelting process, electrorefining process, and precious metal refining process respectively (kWh)

Carbon Footprint Assessment of Waste PCB Recycling Through Black …

113

Table 3 Energy intensity for major processes Unit process

Power consumption

References

Mechanical treatment (E mp )

300 kWh/t

[5]

Electric slag cleaning furnace (E sp )

50 kWh/t slag

[11]

Electrorefining (E ep )

400 kWh/t yield

[11]

Precious metal refining (E pp )

400 kWh/t yield

[11]

Table 4 Emission factors for electricity consumption used in the current study Type of electricity

Emission factor (kg CO2 -e/kWh)

References

Electricity from grid (NSW)

0.81

[9]

Electricity from solar PV

0.0168

[12]

Electricity from wind power

0.0045

[12]

Electricity from hydrogen based through SMR

0.345

[13]

Electricity from hydrogen through solar PV electrolysis

0.075

[14]

Electricity from hydrogen through wind electrolysis

0.029

[14]

that is provided in Table 3. EF p,i represents the greenhouse gas emission equivalent factor (kg CO2 -e/kWh) provided in Table 4. GWPi is the 100-year global warming of greenhouse gas i. 4. Energy consumption during material transportation Ot =



m t × dt × EFt,i × GWPi

(3)

i

OT represents the total CO2 emission during material transportation. m t is the material quantity (tonne), dt is the transportation distance (km), and f is the energy consumption per t km. EFt,i is the emission equivalent factor of greenhouse gas for transportation type i (kg CO2 -e/tonne km) provided in Table 5. GWPi is the 100-year global warming of greenhouse gas i. Table 6 provides data for m t and dt . Table 5 Transportation emission factor Type

Emission factor

Unit

References

Truck

0.0933

kg CO2 /tonne km

[12]

Train

0.0082

kg CO2 /tonne km

[12]

Ship

0.0402

kg CO2 /tonne km

[12]

114

A. Q. Mairizal et al.

Table 6 Data assumption for material transportation Location

Destination

Distance (km)

Material (tonnes)

Transportation mode

Transportation waste PCB (All scenarios) Perth (WA)

Sydney (NSW)

4352

245

Train

Adelaide (SA)

Sydney (NSW)

1371

280

Train

Melbourne (VIC)

Sydney (NSW)

953

947

Train

Hobart (TAS)

Sydney (NSW)

1310

56

Ship

Canberra (ACT)

Sydney (NSW)

247

49

Truck

Darwin (NT)

Sydney (NSW)

4920

13

Sydney (NSW)

Sydney (NSW)

50

1269

Truck

Brisbane (QLD)

Sydney (NSW)

1440

589

Ship

50

4800

Truck

1371

4128

Train

541

4128

Truck

Ship

Transportation copper scrap (Scenario 1) Sydney (NSW)

Sydney (NSW)

Transportation black copper to industry (Scenario 2) Sydney (NSW)

Adelaide (SA)

Adelaide (SA)

Olympic Dam (SA)

Transportation copper scrap (Scenario 3) Sydney (NSW)

Sydney (NSW)

Melbourne (VIC)

Melbourne (VIC)

50

30,000

Truck

953

30,000

Train

Brisbane (QLD)

Sydney (NSW)

1440

30,000

Ship

Canberra (ACT)

Sydney (NSW)

247

1200

Truck

Results and Discussion Carbon Footprint of Waste PCB Recycling Scenarios The results of the carbon footprint calculation are presented in Table 7. It can be seen from Table 7 that the total carbon emission for waste PCB recycling for scenario 1, scenario 2, and scenario 3 were approximately 15,543 tonnes CO2 -eq/year, 14,638 tonnes CO2 -eq/year, and 133,411 tonnes CO2 -eq/year respectively. The total yield copper production based on the mass balance in this study were found to be 4328 kg Cu, 4128 kg black Cu and 68,077 kg Cu for scenario 1, scenario 2, and scenario 3 respectively. As a result, the total carbon footprint for each scenario found in the study was 3.76 kg CO2 -eq/kg Cu, 3.55 kg CO2 -eq/kg Cu and 1.96 kg CO2 -eq/kg Cu, respectively. Comparing the small-scale waste PCB recycling facility to produce Cu, Ag, Au (scenario 1) or to produce intermediate product of black copper (scenario 2), the total carbon emission was not significantly different (3.76 vs. 3.55 kg CO2 -eq/kg Cu). The lower carbon footprint of the scenario 2 was due to the fact that electrorefining and precious metal refining were not part of the overall process, rather the process stopped

Carbon Footprint Assessment of Waste PCB Recycling Through Black …

115

Table 7 Calculation results of GHG emission of waste PCB recycling scenarios Stages Transportation stage

Categories

Items

GHG emission (tonnes CO2 -eq) Scenario 1

Scenario 2

Scenario 3

Road, rail, and water transportation

Truck

29

238

175

Train

19

66

254

Ship

40

40

1776

Mechanical treatment

Power consumption

Electricity

2333

2333

23,328

Material input stage

Feed material

Cu scrap

513

513

9758

Waste PCB

3459

3459

3459

Flux/slag

14

14

143

Coke

263

263

2631

Reduction and oxidation process

CO2 release slag cleaning

CO2 electricity

6319 1392

6319 1392

66,360 13,921

Electrorefining process

Power consumption

Electricity

1159

0

11,594

Precious metal refining process

Power consumption

Electricity

1

0

11

Total carbon emission per year (tonnes CO2 -eq/year)

15,543

14,638

133,411

Total carbon emission per input material (CO2 -eq/kg input material)

1.56

1.46

1.33

Total carbon emission per kg Cu (kg CO2 -eq/kg Cu)

3.76

3.55

1.96

up to the production of black copper (containing the precious metals) which can then be sold as added-value product. It should be noted that the economic returns of scenario 1 are very high compared to the scenario 2 [10]. Scenario 3 resulted in total carbon emission of 1.96 kg CO2 -eq/kg Cu, which is the lowest compared to the scenarios 1 and 2. This mainly due to the fact that the ratio of waste PCB to the copper scrap was quite small in the case of scenario 3. Having high waste-PCB concentration in the feed (in scenarios 1 and 2) means that more C (from the plastic/polyethylene/polypropylene) involved in the process. Figure 3a shows the calculated total carbon emission in each stage in the overall process. In all scenarios, it can be seen from the graph that the reduction and oxidation processes contribute the highest carbon emission during the waste PCB recycling, i.e. contribute up to 50–60% of carbon emission. In scenarios 1 and 2, the second highest carbon emissions were from the material input (mainly from the waste PCB). The carbon emission was in range of 27–29%. Scenario 3 was found to have the least carbon emission from the material input compared to the other scenarios. This was mainly from the high carbon content in the waste PCB that contains polyethylene (15%) and polypropylene (5%) that can act as fuel and reductant during the process [4]. The ratio of copper scrap to waste PCB in the feed material entering the black copper smelting for scenario 3 was set to be 96 wt%: 4 wt%; while for scenario

116

A. Q. Mairizal et al.

Fig. 3 a Total carbon emission in each stage in the process (%); b carbon emission for scenario 1 based on various electricity sources; c carbon emission for scenario 2 based on various electricity sources; d carbon emission for scenario 3 based on various electricity sources

1 and 2 the ratio of waste PCB to copper scrap entering the process was 1 to 1 (50 wt%: 50 wt%). The carbon emission during mechanical treatment was calculated to contribute up to 15%, followed by electrorefining process ~ 7%, transportation ~ 0.57%, and precious metal refining ~ 0.01%. The breakdown of carbon emission in each stage can also be found in Table 7. The results also show that transportation of the key materials into the plant had a low contribution on the overall total carbon footprint of waste PCB recycling process.

Carbon Emission Based on Renewable Energy Scenarios Figure 3b–d show calculated carbon emission under different renewable energy scenarios. In this study, the assessment was carried out on the electricity sourced from solar PV, wind power, and hydrogen through SMR (Steam Methane Reforming), solar PV electrolysis, and wind power electrolysis. It should be noted that the current Australian electricity grids are still mainly coming from coal sources. As can be seen from the Figure, there were significant reductions of carbon emission for all scenarios when the electricity was sourced from renewable energies. In scenario 1,

Carbon Footprint Assessment of Waste PCB Recycling Through Black …

117

up to 15.5 kilo tonnes CO2 -eq/year of emission was calculated when electricity from the grid was used. The carbon emission was decreased to 10.68–10.75 kilo tonnes CO2 -eq/year when the electricity from solar PV and wind power were used, respectively (approximately 31% decrease in the overall carbon emission). In addition, when the electricity was generated from hydrogen (based on solar PV electrolysis and wind power electrolysis), the carbon emission was reduced to 10.8–11.1 kilo tonnes CO2 -eq/year. Alternatively, the electricity might be produced from hydrogen based on SMR process. As the SMR uses natural gas as the main feedstock, the process still releases GHG emission into the environment, resulting in a greater emission factor compared to solar PV-based and wind power-based. Nevertheless, a significant reduction of carbon emission was still achieved when the electricity from hydrogen based on SMR process was used, with total emission of 12.7 kilo tonnes CO2 -eq/year, which represents an 18% of reduction in carbon emission. In all scenarios investigated in this study, the source of electricity can greatly impact the overall carbon emission during waste PCB recycling process. This study found that the overall carbon emission from electricity based on solar PV and wind power is quite comparable. A slightly higher carbon emission was found for the electricity from hydrogen based on solar PV and wind power. In general, switching the electricity sources from grid to alternative energy sources may reduce the total carbon emission from 18 to 31%.

Conclusions In this study, the carbon footprints of waste PCB recycling through black copper smelting in an Australian context were estimated. Three scenarios were investigated: (1) recycling of waste PCB in a small-scale facility, (2) recycling of waste PCB integrated with industry, and (3) recycling of waste PCB in a centralized and large waste PCB recycling facility. The results show that the total carbon footprints ranged from 1.96 to 3.76 (kg CO2 -eq/kg Cu). The highest CO2 emission from the entire recycling process was found to be during the smelting (reduction and oxidation process). To potentially reduce the overall carbon emission, renewable energy sources can be used to supply electricity. This can reduce between 18 and 31% of the overall total carbon emission. The study also found that the emission from transportation has a small contribution to the overall carbon emission. Acknowledgements The authors thank and acknowledge the financial support from Swinburne Growth-SUPRA (Swinburne University Postgraduate Research Award) Ph.D. scholarship. Conflict of Interest The authors declare that they have no conflict of interest.

118

A. Q. Mairizal et al.

References 1. Forti V, Baldé C, Kuehr R, Bel G (2020) The global E-waste monitor 2020: quantities, flows and the circular economy potential. United Nations University (UNU)/United Nations Institute for Training and Research (UNITAR)—Co-hosted SCYCLE Programme, International Telecommunication Union (ITU) and International Solid Waste Association (ISWA), Bonn/Geneva/Rotterdam 2. Islam MT, Huda N (2019) E-waste in Australia: generation estimation and untapped material recovery and revenue potential. J Clean Prod 237:117787 3. Kaya M (2019) Electronic waste and printed circuit board recycling. Springer, Berlin 4. Khaliq A, Rhamdhani MA, Brooks G, Masood S (2014) Metal extraction processes for electronic waste and existing industrial routes: a review and Australian perspective. Resources 3:152–179 5. Ghodrat M, Rhamdhani MA, Brooks G, Masood S, Corder G (2016) Techno economic analysis of electronic waste processing through black copper smelting route. J Clean Prod 126:178–190 6. Zhu X, Lane R, Werner T (2017) Modelling in-use stocks and spatial distributions of household electronic devices and their contained metals based on household survey data. Resour Conserv Recycl 120:27–37 7. Shuva M (2017) Analysis of thermodynamic behaviour of valuable elements and slag structure during E-waste processing through copper smelting. Swinburne University of Technology, Melbourne 8. Hasan M, Rhamdhani MA, Brooks G (2022) Thermodynamics of gallium (Ga) at black copper smelting conditions relevant to E-waste processing. Metall Mater Trans B 53(5):3136–3146 9. Commonwealth of Australia (2021) Australian national greenhouse accounts. Australian Government Department of Industry, Science, Energy and Resources 10. Mairizal AQ, Sembada YA, Tse KM, Haque N, Rhamdhani MA (2022) Techno-economic analysis of waste PCB recycling in Australia. Submitted to Resources, Conservation & Recycling 11. Schlesinger M, King M, Sole K, Davenport W (2011) Extractive metallurgy of copper. Elsevier, Amsterdam 12. Wernet G, Bauer C, Steubing B, Reinhard J, Moreno-Ruiz E, Weidema B (2016) The ecoinvent database version 3 (part I): overview and methodology. Int J Life Cycle Assess 21(9):1218–1230 13. Bareiß K, Rua C, Möckl M, Hamacher T (2019) Life cycle assessment of hydrogen from proton exchange membrane water electrolysis in future energy systems. Appl Energy 237:862–872 14. Ozbilen A, Dincer I, Rosen M (2013) Comparative environmental impact and efficiency assessment of selected hydrogen production methods. Environ Impact Assess Rev 42:1–9

Screening High-Entropy Alloys for Carbon Dioxide Reduction Reaction Using Alchemical Perturbation Density Functional Theory Mohamed Hendy, Okan K. Orhan, Homin Shin, Ali Malek, and Mauricio Ponga

Abstract The carbon dioxide reduction reaction (CO2 -RR) has the potential to transform the production of carbon-based fuels to a closed carbon cycle with no net carbon emission. Recently, high-entropy alloys (HEAs) have shown remarkable catalytic performance for CO2 -RR. The most challenging aspect about investigating HEA for CO2 -RR stems from its inherent surface complexity. To tackle this issue, robust approaches to efficiently screen the configurational space of catalytic HEA materials need to be developed along with an efficient method to navigate the configuration space of HEA alchemical perturbation density functional theory (APDFT). A key advantage of APDFT is that a single density functional theory (DFT) calculation of the adsorbate’s binding energy (BE) can be used to predict many hypothetical catalysts surface structures’ BE at a negligible additional computational cost. This characteristic makes APDFT an appealing technique to explore the configurational space of catalytic HEAs at significantly less computational cost compared to con-

M. Hendy · O. K. Orhan · M. Ponga (B) Department of Mechanical Engineering, University of British Columbia, 2054-6250 Applied Science Lane, Vancouver, BC V6T 1Z4, Canada e-mail: [email protected] M. Hendy e-mail: [email protected] O. K. Orhan e-mail: [email protected] H. Shin Security and Disruptive Technologies Research Centre, National Research Council Canada, Ottawa, ON K1A 0R6, Canada e-mail: [email protected] A. Malek Energy, Mining and Environment Research, National Research Council Canada, Vancouver, BC V6T 1W5, Canada e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_12

119

120

M. Hendy et al.

ventional DFT. Here we investigate the accuracy of using APDFT to screen HEAs for catalytic applications. Keywords Catalysis · High-entropy alloys · Alchemical perturbation · Density functional theory

Introduction The carbon dioxide reduction reaction (CO2 -RR) is an important step to reach a carbon-neutral economy [1]. This cannot be achieved without discovering new catalytic materials. One type of catalytic materials which gained high interest as a potential catalyst is high-entropy alloys (HEAs). HEAs have shown promising results for catalytic application. For instance, Ni20 Fe20 Mo10 Co35 Cr15 HEA has been reported to have better catalytic activities and electrochemical stability for the hydrogen evolution reaction compared to Pt [2]. Another example is Ir0.19 Os0.22 Re0.21 Rh0.20 Ru0.19 HEA which showed remarkable electrocatalytic activity in methanol oxidation [3]. Moreover, a nanocrystalline AuAgPtPdCu HEA was used to reduce carbon dioxide (CO2 ) reaction to gaseous hydrocarbons. AuAgPtPdCu showed about 100% CO2 reduction to gaseous products at a low voltage [4, 5]. The ability to discover new HEAs for catalytic application is hindered by the vast configuration space resulting from the ability to mix different elements [6]. To tackle this issue, robust approaches to efficiently screen the configurational space of catalytic HEA materials need to be developed. One method to explore the configuration space is machine learning (ML). ML approaches have been recently used for catalysis prediction. However, a significant drawback of ML is that large data sets are needed for training and validating the ML models. Besides, ML models are ill-placed to extrapolate catalysis reactions and energies to new systems [7]. An efficient method to navigate the configuration space is alchemical perturbation density functional theory (APDFT) [8, 9]. A key advantage of APDFT is that a single density functional theory (DFT) calculation of the adsorbate’s binding energy (BE) is used to predict many hypothetical catalysts surface structures and BE at a negligible additional computational cost. This characteristic makes computational alchemy an appealing technique to explore the configurational space of catalytic HEAs. APDFT showed high-accuracy results for materials with no bandgap. On the other hand, it is less accurate for semi-conductor systems with a higher bandgap [10]. It has been reported that alchemy qualitatively predicts the lattice stability, equilibrium volumes, and bulk modulus for Zr, Nb, Mo, Tc, Ru, Rh, Pd, and Ag [11]. Carbides, nitrides, and oxides catalysts were studied using computational alchemy. It has been reported that APDFT yields reasonable accuracy when up to six atoms were transmuted for H and OH adsorbate on TiC, TiO, and TiN [10]. The BE of oxygen to the Pd cluster of 79 atoms was studied as a reference state. Then up to 72 transmutations from Pd to Rh and Ag were introduced [12]. The APDFT prediction for the BE reasonably indicated the ability of APDFT to predict BE for hybrid systems

Screening High-Entropy Alloys for Carbon Dioxide Reduction . . .

121

with reasonable accuracy. It is worth mentioning that the transmutations involved in this study were isoelectronic with Z = 1, with minor differences between the reference atom and transmuted atoms. The accuracy of APDFT decreases when the alchemical derivative is high [7]. For instance, when transmutations occur close to the adsorption site, when many atoms are transmuted, or when transmutations involve elements with Z > 1 compared to the reference system [13], where Z > 1 is the difference in atomic numbers between reference and new states. It is worth mentioning that spin polarization effects have not been included in alchemy [11]. Hence it is limited to non-magnetized systems. The aim of this paper is to explore if we can extend the use of APDFT to study the catalysis of HEA by evaluating the accuracy of applying APDFT to predict the binding energy on HEA surface, which can be used for CO2 -RR.

Theoretical and Computational Methodology In this section, we present the theoretical basis of APDFT and the computational details of the Kohn-Sham DFT (KS-DFT) calculations.

APDFT Theoretical Background The thermodynamics of catalytic reaction depends on the reaction intermediate binding energies (BEs) calculated according to the following equation BE = E site + E ads − E ads-site ,

(1)

where E site is the energy of a system with a given surface (e.g., [111] face-centered cubic (fcc) surface), E ads is the energy of the adsorbate, and E ads-site is the energy of a system consisting of an adsorbate near the surface. This means that for brute force DFT calculations, two different sets of simulations are required to explore each possible different binding site for a certain adsorbate which, in turn, makes the exploration of the configurational space of HEAs practically impossible. The general idea behind APDFT relies on introducing small perturbations in the atomic charges from a reference system. Let λ denotes the reaction path, and the reference system is labeled with λ = 0, while the new system is labeled with λ = 1. Then, using the thermodynamic cycle between the reference and new system and approximating the result as a Taylor series expansion, the BE of the new system E(λ = 1) can be related to the binding energy of the reference system E 0 = E(λ = 0) as

122

M. Hendy et al.

1 E(λ = 1) = E 0 + ∂λ E 0 λ + ∂λ2 E 0 λ2 + H.O.T., 2

(2)

where λ = 1, ∂λ E 0 denotes the derivative of E 0 with respect to λ, and H.O.T. denotes the higher-order term. It is worth mentioning that E 0 is the BE of the reference site calculated from Eq. 1. The predicted BE of the new system predicted with Eq. 2 equals the BE of the reference system from DFT calculations plus some correction term from the Taylor series expansion named alchemical derivatives. Considering only the first order derivative, we obtain ∂λ E 0 =

 I

μn (R I )∂λ N I −



FI · ∂λ R I + μe ∂λ Ne .

(3)

I

The alchemical derivative, ∂λ E 0 , consists of three terms. The first term accounts for the nuclear chemical potential gradient μn (R I ) due to variation in nuclear charge N I at position R I from the reference to the new structure. μn (R I ) represents the difference in atomic electrostatic potential between the pristine surface and the surface with adsorbate in the reference structures. Hence, this can be obtained with DFT reference calculations. ∂λ N I is another array of the difference in atomic charges between the reference and the new structure. Generally, this term is fundamental as alchemy relies on transmuting atoms which means different μn (R I ) and ∂λ N I . The second term accounts for energy gradients due to the difference in forces FI on atoms due to changes in atomic positions R I from reference to the new structure. It has been shown that equilibrium volume and bulk modulus predicted by alchemy for mixtures of Rh, Pd, and Ag are in qualitative agreement with full DFT calculations [11], justifying neglecting FI . The third term in Eq. 3 accounts for electronic chemical potential gradient μe due to variation in total number of electrons Ne . This term can be neglected if the total number of electrons in the reference and the new state is the same, which is the case for HEA. In the current study, isoelectronic transmutation can be used to ignore this term [7] by swapping atoms.

Computational Details of DFT Calculations The Vienna Ab-Initio Simulation Package (VASP) [14, 15] was used to conduct the Kohn-Sham DFT calculations. The projector augmented wave method (PAW) was employed with ultra-soft pseudo-potentials. The generalized gradient approximation (GGA) with the revised Perdew-Burke- Ernzerhof (RPBE) [16] exchange-correlation potential was chosen. A plane-wave basis set with an energy cut-off of 700 eV was used. The force on each ion for the ionic relaxation calculations was less than 0.05 eV/Å.

Screening High-Entropy Alloys for Carbon Dioxide Reduction . . .

123

The lattice parameter of the CuPdAgPtAu HEA was taken as the average of the lattice parameter of each element in bulk. The value of the lattice parameter was 4.012 Å. The initial supercell was created to represent a [111]-FCC surface with stacking of ‘ABC’. The supercell was 3 × 3 of 5 layers consisting of 45 atoms of equi-molar fraction for each element. A vacuum of 18 Å was inserted in the direction perpendicular to the surface to avoid surface-surface interaction. A Monkhorst-Pack k-points of 6 × 6 × 1 were used, where the gamma point was used along the [111] direction. We first performed three calculations of HEA slab, O2 molecule, and a slab with O atom. These calculations represent the reference system calculations which all the APDFT calculations were calculated from at no computational cost. Then we performed another set of DFT calculations to compare with the APDFT calculations in order to assess the APDFT calculations’ accuracy. The initial HEA supercell was relaxed with all atoms allowed to move. Then the O atom was added at the bridge binding site, while keeping the HEA atoms fixed. The O atom was allowed to move in the vertical direction on the surface.

Results In order to assess the accuracy of APDFT in predicting the BE of CuPdAgPtAu HEA without additional computational cost, we chose an atom on the BE of bridge binding site and another atom on the surface away from the binding site. The atom at the surface was transmuted (swapped) with all the other atoms in the supercell with a difference in atomic number Z = 1. More specifically Au with Pt. We limited the swapping with Z = 1 because the previous studies showed that transmutations (changing atoms) with Z much larger than 1 can lead to large error [17]. The swaps were classified into two categories which are swaps involving atoms near the binding site and away from the binding site. The results of swapping Au with Pt are shown in Fig. 1. It was observed that the APDFT results when performing swapping of element on the surface but not directly on the binding site agrees better with the DFT predictions for the bridge site. We have calculated the mean average error (MAE) for the Pt/Au swapping Fig. 2. The MAE for swapping Pt/Au atoms on the surface but away from the binding site for the adsorbate at the bridge site was 0.009 eV. The small MAE value shows that APDFT can reasonably predict the BE for the bridge binding site. For the case of swapping atoms near the binding site, the MAE is much higher. Swapping Pt/Au resulted in MAE of 0.22 eV for bridge site. Hence the errors when swapping close to the binding site are high.

124

M. Hendy et al.

Fig. 1 Binding energy (BE) in eV for swapping Au with Pt away (blue) and close (orange) from the binding site for bridge site

Fig. 2 Mean average error for atoms swapped away and close from the binding site for bridge site

Conclusion We have assessed the accuracy of using alchemical perturbation density functional theory (APDFT) in predicting the binding energy (BE) of oxygen (O) on CuPdAgPtAu HEA surface for bridge site. APDFT could predict the BE of O on CuPdAgPtAu HEA surface with a high accuracy for the atoms not directly on the binding site (away from the binding site). The mean average error (MAE) for APDFT away from the binding site was less than 0.02 eV which is in a good agreement with the DFT

Screening High-Entropy Alloys for Carbon Dioxide Reduction . . .

125

calculations. On the other hand, for the atoms close to the binding site, the results of APDFT were less accurate with a MAE of 0.22 eV for Pt/Au swapping. Hence, APDFT has shown to predict accurately the BE for HEA when swapping element on the surface away from the binding bridge site, while showing less accuracy close to the binding site.

References 1. Kim C, Dionigi F, Beermann V, Wang X, Müller T, Strasser P (2019) Alloy nanocatalysts for the electrochemical oxygen reduction (ORR) and the direct electrochemical carbon dioxide reduction reaction (CO2RR). Adv Mater 31(31):1805617 2. Zhang G, Ming K, Kang J, Huang Q, Zhang Z, Zheng X, Bi X (2018) High entropy alloy as a highly active and stable electrocatalyst for hydrogen evolution reaction. Electrochim Acta 279:19–23 3. Yusenko KV, Riva S, Carvalho PA, Yusenko MV, Arnaboldi S, Sukhikh AS, Hanand M, Gromilov SA (2017) First hexagonal close packed high-entropy alloy with outstanding stability under extreme conditions and electrocatalytic activity for methanol oxidation. Scr Mater 138:22–27 4. Pedersen JK, Batchelor TA, Bagger A, Rossmeisl J (2020) High-entropy alloys as catalysts for the CO2 and CO reduction reactions. ACS Catal 10(3):2169–2176 5. Nellaiappan S, Katiyar NK, Kumar R, Parui A, Malviya KD, Pradeep K, Singh AK, Sharma S, Tiwary CS, Biswas K (2020) Highentropy alloys as catalysts for the CO2 and CO reduction reactions: experimental realization. ACS Catal 10(6):3658–3663 6. Banko L, Krysiak OA, Pedersen JK, Xiao B, Savan A, Löffler T, Baha S, Rossmeisl J, Schuhmann W, Ludwig A (2022) Unravelling composition–activity–stability trends in high entropy alloy electrocatalysts by using a data-guided combinatorial synthesis strategy and computational modeling. Adv Energy Mater 2103312 7. Griego CD, Kitchin JR, Keith JA (2021) Acceleration of catalyst discovery with easy, fast, and reproducible computational alchemy. Int J Quantum Chem 121(1):26380 8. Von Lilienfeld OA, Tuckerman M (2007) Alchemical variations of intermolecular energies according to molecular grand-canonical ensemble density functional theory. J Chem Theor Comput 3(3):1083–1090 9. Von Lilienfeld OA, Tuckerman ME (2006) Molecular grand-canonical ensemble density functional theory and exploration of chemical space. J Chem Phys 125(15):154104 10. Griego CD, Saravanan K, Keith JA (2019) Benchmarking computational alchemy for carbide, nitride, and oxide catalysts. Adv Theor Simul 2(4):1800142 11. To Baben M, Achenbach J, Von Lilienfeld O (2016) Guiding ab initio calculations by alchemical derivatives. J Chem Phys 144(10):104103 12. Sheppard D, Henkelman G, von Lilienfeld OA (2010) Alchemical derivatives of reaction energetics. J Chem Phys 133(8):084104 13. Saravanan K, Kitchin JR, Von Lilienfeld OA, Keith JA (2017) Alchemical predictions for computational catalysis: potential and limitations. J Phys Chem Lett 8(20):5002–5007 14. Kresse G, Hafner J (1994) Ab initio molecular-dynamics simulation of the liquid-metal– amorphous-semiconductor transition in germanium. Phys Rev B 49:14251–14269. https://doi. org/10.1103/PhysRevB.49.14251 15. Kresse G, Furthmüller J (1996) Efficient iterative schemes for ab initio totalenergy calculations using a plane-wave basis set. Phys Rev B 54:11169– 11186. https://doi.org/10.1103/PhysRevB. 54.11169 16. Hammer B, Hansen LB, Nørskov JK (1999) Improved adsorption energetics within densityfunctional theory using revised Perdew-Burke-Ernzerhof functionals. Phys Rev B 59:7413– 7421. https://doi.org/10.1103/PhysRevB.59.7413

126

M. Hendy et al.

17. Saravanan K, Kitchin JR, von Lilienfeld OA, Keith JA (2017) Alchemical predictions for computational catalysis: potential and limitations. J Phys Chem Lett 8(20):5002–5007. https:// doi.org/10.1021/acs.jpclett.7b01974. PMID: 28938798

Part III

Thermal Management, Environmental and Energy Technologies

Novel Thermal Conductivity Measurement Technique Utilizing a Transient Multilayer Analytical Model of a Line Heat Source Probe for Extreme Environments Katelyn Wada, Austin Fleming, and David Estrada Abstract Advancements in thermal properties analysis are crucial for continual improvement of existing and next generation reactors, space exploration, and environmental safety. Extreme environments pose a great hurdle for instrumentation to measure real time thermal properties due to the extreme temperatures, high radiation, and variable electromagnetic environments. Nevertheless, measurement systems are tremendously important for the design, performance, and safety considerations of nuclear fuels, spacecraft, and deep sea/deep earth drilling. Thermal properties may change significantly in these environments creating challenging problems for temperature and thermal conductivity measurement systems. A recent focus has surrounded improvements in such systems for accurate determination of temperature and thermal properties to increase efficiencies, reduce costs, calibrate models, and tackle problems previously unfulfilled. Here we report on the thermal quadrupoles method to develop analytical models, which have been verified using multiphysics finite element analysis, for thermal conductivity measurements conducted with a line heat source probe. A novel measurement technique was developed to monitor the temperature rise of the sample via the temperature dependent resistance of the probe’s heater wire. This innovative approach provides a feasible method for extracting thermal conductivity in extreme environments. Keywords Thermal conductivity · Line heat source · Thermal quadrupoles

K. Wada · D. Estrada (B) Micron School of Materials Science and Engineering, Boise State University, Boise, ID 83725, USA e-mail: [email protected] A. Fleming · D. Estrada Idaho National Laboratory, Idaho Falls, ID 83415, USA D. Estrada Center for Advanced Energy Studies, Boise State University, Boise, ID 83725, USA © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_13

129

130

K. Wada et al.

Introduction To ensure optimal operation for existing and future reactors, secure health and safety regulations on spacecraft, and proper discerning of critical properties during deep earth/sea drilling, accurate knowledge of thermal properties is crucial. In these extreme systems, temperature determination is a primary property of interest, though such harsh environments pose great problems for contemporary measurement techniques. Material degradation due to radiation and extremely high temperatures makes materials selection and system engineering quite difficult. Electromagnetically noisy environments also cause unfavorable disruptions in measurements. For nuclear reactors, current improvement efforts in temperature and thermal conductivity determination are of great concern to enable in-pile measurement systems. Thermal conductivity degradation is a key limiter of fuel performance and lifetime. Efficiencies of energy conversion can be calculated knowing the rate of heat transfer by conduction through the material. Knowing thermal conductivity can also help ascertain possible risks associated with selecting certain fuels. The standard out-of-pile temperature and thermal conductivity measurements require testing facilities and expensive removal of the irradiated fuel. This method leads to a complete loss of information during irradiation where significant changes in thermal conductivity occur. Conducting inpile measurements would allow a better understanding of how fuels behave under irradiation, alleviate time necessary for out-of-pile measurements, and reduce costs [1–5]. Another extreme environment that is becoming continuously more involved in human affairs is space, particularly low earth orbit [6, 7]. Spacecraft must be equipped with highly accurate temperature and thermal conductivity measuring devices for the purpose of health and safety of their inhabitants, as well as, for the integrity of the spacecraft itself. Securing the viability of space travel and longevity of space faring equipment is a necessary step for the future of space exploration. This environment poses many problems including materials degradation due to radiation, large temperature gradients, complicated wiring schemes due to a large number of required measurements, and minimal payload mass budget requirements for small spacecraft [8–10]. Another type of extreme environment worth considering is deep sea/deep earth drilling. For such systems, the physical and temperature requirements for probes and sensors are extremely harsh [11, 12]. Deep sea/earth drilling is needed for environmental safety precautions, material extraction, and geothermal exploration. Increasing the accuracy and extrapolating more detailed temperature profiles can minimize environmental impact and reduce costs associated with needing a large number of poorer quality temperature gradient wells [13, 14]. This report details a novel measurement technique and approach for measuring temperature and thermal conductivity for such applications.

Novel Thermal Conductivity Measurement Technique Utilizing …

131

Fig. 1 Single wire probe design. a Cross-sectional view, b image of probe ends, and c length view, where rsheath is the sheath radius, rwires is the radius of the effective wire layer, rinsulation is the insulation radius, rsample is the sample radius, Q˙ is heat flux, and T1 , T2 , and T3 are temperature solutions at the layer boundaries

Probe Geometry The device geometry was selected to accommodate size limitations in complex systems and provide a more simplified symmetry for modeling purposes. The single wire geometry is depicted in Fig. 1. This geometry consists of a 0.1 mm diameter platinum wire surrounded by magnesium oxide insulation and an Inconel 600 sheath. This probe is 1 mm in diameter and 12 in in length. This length is not particularly necessary; however, a sufficient length should be chosen to eliminate end effects and ensure a linear heat source [15]. The probe is embedded in a sample; in this case PTFE (Teflon), 10 mm in diameter, was chosen as a viable test sample.

Theory By measuring the resistance change through the temperature dependent voltage across our linear heat source, we aim to use the resistance of our Pt wire as a thermometer for thermal conductivity extractions. This technique utilizes established theories similar to platinum resistance thermometry and the 3-omega technique [16– 19]. The measured voltage can be converted to temperature via the equations below. Resistance is a temperature dependent quantity which can be approximated as shown in Eq. (1)

132

K. Wada et al.

R(T ) = R0 + R  T

(1)

where R0 is the baseline measurement for resistance and R  is the temperature sensitivity of resistance [17]. The measured voltage is given by Ohm’s Law V = I R(T )

(2)

where the current, I , can be represented with Eq. (3). I = A cos(ωt)

(3)

The resulting equation is given as Eq. (4). This equation allows temperature extraction from the measured voltage of the heater wire and can thus provide a temperature profile from the measurement.   V (T ) = A cos(ωt) R0 + R  T

(4)

Thermal Model A real solution to the heat flow equation can be constructed utilizing all of the thermal properties of the probe layers (wire, insulation, and sheath). Alternatively, an effective property calculation can be done to model the entire probe as a single layer. To derive an effective thermal conductivity for the entirety of the probe including all its components, we start with a one-dimensional Fourier’s Law of heat flow per unit ˙ in cylindrical geometry time ( Q)   T Q˙ = 2π Lk r

(5)

where L is the length of the entire probe, k is thermal conductivity, and T is change r in temperature over change in radial distance. Using a log mean cross-sectional area, a more accurate equation can be achieved with the cylindrical geometry. 2π Lkeff T   Q˙ = ln rrsheath wire

(6)

A similar expression may be obtained for the insulator (ins) and sheath layers. Q˙ =

 ln

2π LT rins rwire

kins





+

ln

rsheath rins

ksheath



(7)

Novel Thermal Conductivity Measurement Technique Utilizing …

133

Finally, setting Eqs. (6) and (7) equal to each other, the probe’s effective thermal conductivity can be calculated using the heat flow resistance network (Eq. 8) that takes the resistances of each layer into account. Here, we must assume there is no heat generation and that heat flow is purely one dimensional [3]. 

keff =



rsheath rwire    r ln sheath r

ln  ln

rins rwire

+

kins

(8)

ins

ksheath

This equation represents the concentric cylindrical layers of the heater wire, insulation, and sheath of the probe using a log mean cross-sectional area approach while designating the heater wire as the only heat generating element [20]. We also must use an averaged thermal mass approach for determining the effective thermal mass of the probe (Eq. 9). For accurate measurements, the actual and effective thermal masses of the probe must be the same [3]. For calculations involving thermocouple elements, averaging of both materials’ thermal conductivities and thermal diffusivities (α) is necessary.

 ρc p eff =

k k = ρc p → α = α ρc p     Awire ρc p wire + Ains ρc p ins + Ash ρc p sheath

(9)

Aprobe

Here ρ is density, c p is the specific heat capacity at constant pressure, and A is the cross-sectional area. Using the effective property calculations, and utilizing a thermal quadrupoles approach, an analytical model can be generated in the form depicted in Eq. (10) [21].

θ1 ϕ1



=

A1 B1 C 1 D1



1 Rth 0 1



A4 B4 C 4 D4



10 h1



θ3 ϕ3

 (10)

where θ is the Laplace temperature, ϕ is the Laplace heat flux, h is the convection coefficient, index 1 corresponds to the effective probe layer, and index 4 is the sample layer. In addition, a thermal contact resistance, Rth , layer can be added between layers. In this case, it is added between the probe and sample layers; however, we found the thermal contact resistance to be negligible in our experiments. This value could be added to compensate for discrepancies in the probe’s insulator thermal properties due to fabrication related alterations. Adding layers is quite simple and merely requires the superposition of additional layers in radial order. Modeling each of the probe’s layers explicitly can be achieved with Eq. (11).

θ1 ϕ1



=

A1 B1 C 1 D1



A2 B2 C 2 D2



A3 B3 C 3 D3



1 Rth 0 1



A4 B4 C 4 D4



10 h1



θ3 ϕ3

 (11)

134

K. Wada et al.

Table 1 Coefficients for the material matrices in Eqs. (10) and (11) Sample layer √ p/αi , q2,i+1 = ri+1 p/αi       Ai = q2,i I0 q1,i K 1 q2,i + I1 q2,i K 0 q1,i     Bi = 2π1k L [I0 q2,i K 0 q1,i − I0 q1,i K 0 q2,i

q1,i = ri

√

Ci =       2π k Lq1,i q2,i I1 q2,i K 1 q1,i − I1 q1,i K 1 q2,i     Di = q1,i [I0 q2,i K 1 q1,i + I1 q1,i K 0 q2,i

Single wire, insulation, and sheath layers Average temperature solution √ qi = ri p/αi Ai = 1 Bi =

I0 (qi ) 1 2π k L qi I1 (qi )

−

1 ρcπri2 L p

Ci = ρcπri2 L p Di =

qi I0 (qi ) 2 I1 (qi )

α thermal diffusivity, p Laplace parameter, r radius, k thermal conductivity, L length, I and K modified Bessel functions, ρ density, c specific heat capacity

In this case, indices 1, 2, and 3 correspond to the wire, insulation, and sheath layers respectively. The coefficients for the material matrices are shown in Table 1.

Experimental Methods Experimental measurements were conducted utilizing a lock-in amplifier, set to A-B differential input. A power source provides a DC signal to heat the probe while a small AC signal is superimposed using a function generator. The lock-in is able to lock onto that small AC signal allowing resolute measurements in potentially noisy environments. The probe is suspended in air to allow natural convection as shown in Fig. 2a. A four-wire configuration was used to mitigate noise from the lead resistances (Fig. 2b) [22, 23]. In this case, the lock-in measures across the R3 resistor seen in Fig. 2c. This circuit was employed to increase the signal to noise ratio of the probe by allowing a larger DC current to be used while protecting the lock-in with a voltage divider.

Results and Discussion To validate our results, we compared experimental measurements of Teflon (PTFE) cylinders with a 10 mm diameter using our line heat source method described above to our analytical model and a multiphysics finite element model. PTFE is known to have a thermal conductivity of 0.25 W m−1 K−1 . The extracted temperature coefficient of resistance for our platinum wire was 0.00693 K−1 , which corresponds to a voltage sensitivity of 6.63 μV K−1 . By applying current of 0.36 A, Joule heating was induced in the probe resulting in a temperature increase of 3 K over 10 s. This created the necessary temperature gradient for heat flow. Using our analytical and finite element

Novel Thermal Conductivity Measurement Technique Utilizing …

135

Fig. 2 a Line source probe experimental setup. b Diagram of the four-point electronic configuration used to mitigate the influence of resistance in the lead extensions. c Circuit diagram generated in LT Spice to increase signal to noise ratio of the probe

framework, we were able to compare our models to the experimental data by fitting the thermal conductivity of the Teflon samples. COMSOL Multiphysics was used to validate the analytical model, as shown in Fig. 3. The effective and real models for both the analytical and COMSOL results match well. This validates the effective property calculations as well as the analytical models themselves. Experimental results are also shown and match the models relatively well until around 10 s. After 10 s there is a deviation in the measurement from the models. This is potentially due to the fragility of the 0.1 mm diameter platinum wire or the probe stage/setup perturbing the measurement. Constructing a more robust probe with easy-to-handle wire attachments would likely provide better results. By comparison to traditional line heat source method, our approach provides a more compact and simple design. Traditionally, the line heat source method completely neglects the probe’s thermal properties as it is assumed to be surrounded by an infinite medium. Hence, the measurement is limited for small samples and must start at the appropriate time to accurately measure finite samples [2]. The probe itself is also more complex including both a heater wire and thermocouple to complete measurements [3]. Our single wire probe offers a smaller diameter which provides minimal intrusion into the sample and less impact on convection effects at the sample boundary. It also eliminates issues with cross-talk between multiple thermal and electrical elements within the probe, particularly at high temperatures where dielectric insulators between such elements begin to allow high amounts of leakage currents [24–26].

136

K. Wada et al.

Fig. 3 a Real and effective analytical models are depicted as lines, COMSOL Multiphysics models are depicted as squares, and an experimental measurement is shown as black dots. b COMSOL temperature plot at 10 s of the real/explicit model

Conclusion and Summary To summarize, a novel thermal conductivity probe and measurement technique has been developed. Two different analytical models were created that have both been verified against finite element analysis and experiment. This new technique has the potential to measure thermal conductivity of materials in extreme environments and can be catered to fit different application needs. We measured 10 mm PTFE pellets using a single wire probe which acted as a heating and sensing element. The probe had a temperature coefficient of resistance of 0.00693 K−1 which translated to a voltage sensitivity of 6.63 μV K−1 . With our methods, we demonstrate an effective approach for thermal conductivity measurements of small and finite samples with potential applications in extreme environments. The single wire probe was, however, difficult to implement due to the fragile nature of the thin Pt wire. Further improvements to the physical resilience of the probe itself would be highly beneficial moving forward. Conflict of Interest The authors declare that they have no conflict of interest. Funding This work was prepared as an account of work sponsored by the U.S. Department of Energy, Office of Nuclear Energy Advanced Sensors and Instrumentation program under DOE Contract DE- AC07- 05ID14517. Neither the U.S. Government nor any agency thereof, nor any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness, of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. References herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the U.S. Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the U.S. Government or any agency thereof. Additionally, this material is based upon work supported under a University Nuclear Leadership Program Graduate Fellowship through the Department of Energy, Office of Nuclear Energy.

Novel Thermal Conductivity Measurement Technique Utilizing …

137

References 1. Smith JA, Kotter D, Garrett SL, Ali RA (2012) Sensors synergistic with nature for in-pile nuclear reactor measurements. In: FIIW 2012—2012 future of instrumentation international workshop proceedings, pp 67–69. http://doi.org/10.1109/FIIW.2012.6378330 2. Daw JE, Rempe JL, Knudson DL (2012) Hot wire needle probe for in-reactor thermal conductivity measurement. IEEE Sens J 12(8):2554–2560. https://doi.org/10.1109/JSEN.2012.219 5307 3. Hollar C, Fleming A, Davis K, Budwig R, Jensen C, Estrada D (2019) A parametric study for in-pile use of the thermal conductivity needle probe using a transient, multilayered analytical model. Int J Therm Sci 145:106028. http://doi.org/10.1016/j.ijthermalsci.2019.106028 4. Fox B, Ban H, Rempe JL, Daw JE, Knudson DL, Condie KG (2009) In-pile thermal conductivity measurement method for nuclear fuels. In: Proceedings of the 30th international thermal conductivity conference and the 18th international thermal expansion symposium, thermal conductivity 30/thermal expansion 18, October 2010 5. Ronchi C, Sheindlin M, Staicu D, Kinoshita M (2004) Effect of burn-up on the thermal conductivity of uranium dioxide up to 100.000 MWdt−1 . J Nucl Mater 327(1):58–76. https://doi.org/ 10.1016/j.jnucmat.2004.01.018 6. Wertz JR, Everett DF, Puschell JJ (eds) (2011) Space mission engineering: the new SMAD. Microcosm Press, pp 127–149 7. Badhwar GD (1997) The radiation environment in low-earth orbit, vol 148, no 5. Radiation Research Society, pp S3–S10 8. Paschalidis N (1998) A remote I/O (RIO) smart sensor analog-digital chip for next generation spacecraft, pp 1–11 9. Paschalidis NP (1999) A smart sensor integrated circuit for NASA’s new millennium spacecraft. In: Proceedings of the IEEE international conference on electronics, circuits, and systems, vol 3, pp 1787–1790. http://doi.org/10.1109/ICECS.1999.814554 10. Barth JL (2003) Space and atmospheric environments: from low earth orbits to deep space. European Space Agency, (Special Publication) ESA SP, no 540, pp 17–30. http://doi.org/10. 1007/1-4020-2595-5_2 11. Wang H, Ge Y, Shi L (2017) Technologies in deep and ultra-deep well drilling: present status, challenges and future trend in the 13th five-year plan period (2016–2020). Nat Gas Ind B 4(5):319–326. https://doi.org/10.1016/j.ngib.2017.09.001 12. Kelessidis VC (2009) Drilling fluid challenges for oil-well deep drilling. In: 9th international multidisciplinary scientific geoconference and EXPO—modern management of mine producing, geology and environmental protection, SGEM 2009, October, vol 1, pp 595–604 13. Coolbaugh MF, Sladek C, Faulds JE, Zehner RE, Oppliger GL (2007) Use of rapid temperature measurements at a 2-meter depth to augment deeper temperature gradient drilling. In: Proceedings, thirty-second workshop on geothermal reservoir engineering 14. Stegemeier GL, van Meurs P, van Egmond CFH (1986) Mini-well temperature profiling process, p 4616705 15. Kömle NI et al (2011) In situ methods for measuring thermal properties and heat flux on planetary bodies. Planet Space Sci 59(8):639–660. https://doi.org/10.1016/j.pss.2011.03.004 16. Cahill DG, Goodson K, Majumdar A (2002) Thermometry and thermal transport in micro/nanoscale solid-state devices and structures. J Heat Transf 124(2):223–241. https://doi. org/10.1115/1.1454111 17. Trivedi R, Mathur G, Mathur A (2011) A survey on platinum temperature sensor, pp 13–14 18. Sittler FC, Toy AC (1988) Thin film platinum resistance thermometer with high temperature diffusion barrier, p 4791398 19. Cahill DG (1990) Thermal conductivity measurement from 30 to 750 K: the 3ω method. Rev Sci Instrum 61(2):802–808. https://doi.org/10.1063/1.1141498 20. Department of Energy (1992) Thermodynamics, heat transfer, and fluid flow, vol 2 21. Maillet D, Batsale J-C, Andre’ S, Degiovanni A (2000) Thermal quadrupoles, solving the heat equation through integral transforms.

138

K. Wada et al.

22. Cahill DG, Pohl RO (1987) Thermal conductivity of amorphous solids above the plateau. Phys Rev B 35(8):4067–4073. https://doi.org/10.1103/PhysRevB.35.4067 23. Worledge D (2005) Reduction of positional errors in a four point probe resistance measurement 24. Rempe JL, Knudson DL, Daw JE, Condie KG, Curtis Wilkins S (2009) New sensors for in-pile temperature detection at the advanced test reactor national scientific user facility. Nucl Technol 175(3):681–691. http://doi.org/10.13182/NT11-A12515 25. Nakamura Y, Sorimachi M, Kimura H, Matsuyama S (2011) Radiation resistant composition, wire and cable, p 8076408 26. Afzal FA, Giutronich JE (1973) Measurement of electrical conductivity of magnesium oxide single crystal at high temperature using a solar furnace. Solar Energy 15(2). http://doi.org/10. 1016/0038-092X(73)90122-9

The Effect of Reduced Flue Gas Suctioning on Superstructure and Gas Temperatures Brandon Velasquez, Sarah DiBenedetto, Yonatan A. Tesfahunegn, Maria Gudjonsdottir, and Gudrun Saevarsdottir

Abstract Reducing CO2 emissions from aluminum smelters is of great interest to reach the goal of carbon neutrality. One possible approach is to implement carbon capture and sequestration techniques (CCS). This is already being done within the geothermal sector in Iceland, where captured CO2 is sequestered through the Carbfix method of mineralization. Under the current smelter operation, the CO2 concentration in the exhaust gas is below 1%, which is too low for conventional up-concentration technology, but by adjusting the draft rate, the concentration can increase to the required 4% or higher. In order to determine the feasibility of retrofitting this method into existing smelters, a CFD model has been developed to predict the effects that the draft rate modifications would cause within the system. In this paper, the results from CFD modeling of the flue gas and superstructure of the cell are used to predict changes in flow and thermal conditions. Outlet temperature values are determined for the air passing through the system as well as the surface temperatures of the anode cover material (ACM), hood cover, and anode rods. Keywords Aluminum smelting · Modeling and simulation (CFD) · Sustainability · Environmental effects

Introduction The aluminum industry has a large presence in Iceland, consuming around 66% of the total electricity production [1] as well as generating 1.4 million tonnes of CO2eq /year [2]. Due to the country’s use of renewable energy sources for 100% of the electricity production, the main contributor to CO2 emissions is the electrolysis process to produce the metal. This is why methods aiding in mitigating or eliminating these emissions have been sought in recent years. One of the proposed methods is to capture CO2 from the flue gas generated in the smelting process and sequester it through mineralization. This is currently employed in the geothermal energy sector B. Velasquez (B) · S. DiBenedetto · Y. A. Tesfahunegn · M. Gudjonsdottir · G. Saevarsdottir Department of Engineering, Reykjavik University, Menntavegur 1, 101 Reykjavik, Iceland e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_14

139

140

B. Velasquez et al.

in Iceland where carbon emissions from a geothermal power plant are captured and dissolved in water. The CO2 -rich fluid is then injected into suitable ground, where it eventually forms into basalts and peridotites [3, 4]. This method, called Carbfix, was developed by scientists at Reykjavik University and the University of Iceland. For the aluminum smelters to use this existing method, there is a need to upconcentrate the flue gas emitted from the aluminum reduction cells. Under current operations, the volume concentration of CO2 hovers around 1%, which is too low. A concentration of 4% or higher is needed for traditional amine absorption to be feasible for up-concentration. For direct application of the Carbfix method, a concentration of at least 10% is needed [4]. In either case, a change to the current setup employed by the smelters is required, such as a draft rate reduction [5]. This, of course, would affect the heat transfer rate from the cell, since the flow of air through the system carries away over half of all the waste heat of the cell [6]. Previous work has been completed concerning draft rate reduction within aluminum reduction cells, pertaining to a desired increase in thermal quality of waste heat that could then be recovered in an additional heat exchange process [7]. There are two sources of heat in the upper section of the reduction cell. First, the anode assembly, which heats up due to the passage of electric current, Joule heating. It has been shown in Taylor et al. [8] that under normal ventilation conditions, convective heat transfer has a larger influence on the heat loss than radiative heat transfer. Second, the main source of heat in the upper cell is heat flux through the anode cover material, which is a layer of separation between the molten electrolyte and the air under the cell hood. Thermal circuit modeling has been done on this section of the system and shown that the thermal resistance existing therein plays an important role by acting as insulation, but also allowing enough heat to leave in order to maintain the desired temperature in the bath [9]. Both of these heat sources are in direct contact with the air coming into the system; therefore, it is apparent that changes in draft rate will have notable effects. In this work, a computational fluid dynamics (CFD) model has been developed to study the effect of draft rate on thermal and flow conditions in the existing reduction cell design. The model focuses on the simulation of the airflow in and around the reduction cell and the solid surfaces in contact during heat transfer from the cell. These results are used to predict how a reduction in draft rate would affect the system.

Computational Model In this section, descriptions are given for the mathematical modeling, the cell geometry, mesh, boundary conditions, and grid convergence study.

The Effect of Reduced Flue Gas Suctioning on Superstructure and Gas …

141

Mathematical Modeling The reduction cell CFD model was developed in ANSYS Fluent [10] to solve the conservation equations for mass, momentum, and energy. These were solved under a steady-state, pressure-based solver with gravity effects enabled. In order to simulate turbulence in the system, the Reynolds-Averaged Navier–Stokes (RANS) equations with the Shear Stress Transport (SST) k-ω turbulence model were used. Due to an expected significant temperature difference between the ambient air and the surface temperatures within the cavity between the reduction cell hood and the anode cover material, the natural convection occurring within the system would be strong, which is better represented by the k-ω model [11, 12]. Due to the variable temperature, the density was assumed to follow the ideal gas law, ρ=

P RT

(1)

where P is the pressure, R is the specific gas constant, and T is the air temperature. Second-order upwind discretization schemes were used for all variables with default settings for the relaxation factors. Advanced solution controls were also implemented in order to aid with the stability of the results. The termination value for the multigrid solver was set to 0.01 for the flow, turbulent kinetic energy, specific dissipation rate, and energy. Additionally, the cycle type was set to W-cycle, with the bi-conjugate gradient stabilized method (BCGSTAB) enabled in order to help to improve convergence.

Simplifications of Model Geometry The geometry of the aluminum cell was simplified from the original schematic for compliance with the modeling tool. Since the cavity between the reduction cell hood and the anode cover material is the main area of interest in the model, many details relating to the other portions of the actual geometry were significantly simplified, such as surface features and ancillary equipment. To speed up the computation time, a few additional simplifications were made to the model as well: 1. Each electrolysis cell contains 20 anodes, in two rows of 10. There are covers lining either side of the cell hood with gaps between them acting as inlets for the cooling air to enter, aligned with each anode in the cell. All the air that enters the hood is pulled towards an outlet pipe at the top of one end of the domain. Since the heat flux from the ACM was considered uniform and all of the anode rods are identical in shape, size, and Joule heating amount, then there is a natural symmetry plane centered along the length of the system, which permits using half of the domain for the simulation, reducing the computational resource need (see Fig. 1).

142

B. Velasquez et al.

Fig. 1 Front (left) and isometric (right) views of flow domain of interest for the CFD model

2. In order to improve the stability of the simulation, the fluid domain was extended beyond the superstructure of the hood to a distance halfway across the walkway between cells and to the top edge of the upper duct, above the anode rods. This helps to avoid issues involving improper flow development after initialization and to see the induced convection currents in the air outside the cell. Therefore, two fluid domains were used, coming into contact only at the gaps in the covers along the hood edges. To reduce the computation time for the simulation, the mesh size for the air in the outer domain was coarser than that of the inner air domain.

Meshing Setup The mesh used in the simulations presented was designed to be better suited to the geometry at hand. The mesh sizing was decreased significantly in the smaller areas of the model, such as the gaps between the covers of the cavity, due to the smaller flow area between the inner and outer fluid domains. Inflation layers were added to these areas due to the points of contact between the two fluid domains and the increased flow velocity. In order to receive more useful thermal results, inflation layers were also imposed on the solid–fluid boundaries, i.e. between the anode rods and air, as well as the floor of the cavity and the inner air domain, since the majority of heat transfer in the system occurs along those interfaces. A maximum of five inflation layers, with a growth rate of 1.2, were used for all instances of inflation in this model. Mesh generation was performed initially utilizing tetrahedral cells. A grid independence study was performed using a coarse and a fine mesh. The coarse regime sizing was defined as two times larger than the medium mesh sizing, while the fine regime sizing was defined as half of the medium mesh sizing. In order to speed up

The Effect of Reduced Flue Gas Suctioning on Superstructure and Gas …

143

Table 1 Mesh details for each regime sizing Refinement

Course

Medium

Fine

Outer air domain (mm)

200

100

50

Inner air domain (mm)

150

75

37.5

Door gap surface (mm)

4

2

1

Tetrahedral mesh

Anode rod (mm)

60

30

15

Rod–air interface (mm)

60

30

15

ACM surface (mm)

100

50

25

Nodes

7.34E05

1.96E06

6.53E06

Elements

3.72E06

9.69E06

30.81E06

Polyhedral mesh Nodes

4.63E06

12.44E06

40.91E06

Elements

1.13E06

3.40E06

12.90E06

Average run time (h)

3–4

7–8

24–28

the simulation time without compromising reliability, the completed meshes were imported into ANSYS Fluent. They were then converted into polyhedral meshes. This eases the computation time due to increasing the number of nodes present and decreasing the number of elements. Specific mesh details can be seen in Table 1.

Boundary Conditions The inlet condition was defined as an atmospheric pressure boundary with a temperature of 303 K and was applied to the edges of the outer air domain (see Fig. 1). A velocity inlet was also provided from the flow of air coming from the basement located under the pot, originating from outside the potroom that flows past heat exchange fins located along the sides of the pot. A velocity of 10 kph (2.78 m/s) was used, directed parallel to the door covers, at a temperature of 313 K. At the outlet duct, a varied negative gauge pressure (26.875, 430, and 6880 Pag) was provided in order to adjust for different draft rates through the hood cavity (one-quarter flow, normal flow, and four-times flow, respectively). Backflow was also prevented in the outlet duct. Turbulence at the inlets was limited to 1% intensity, with turbulent viscosity ratios of 1. Boundary conditions relating to the heat transfer within the system were also defined. The main source term was defined as a heat flux directed from the bottom surface of the domain, bordering the inner air domain. This represents the heat passing through the ACM and was set to a value of 2000 W/m2 [13]. Additionally, a source term was provided to represent the contribution of heat that would be generated through the Joule heating of the rods, without modeling the electrical portions of

144

B. Velasquez et al.

Table 2 CFD result comparison between mesh regimes at the normal draft rate (430 Pag) Parameter

Coarse

Mass flow (kg/s)

Medium

Fine

Relative error % Relative error % Coarse/medium Medium/fine

1.3066

1.2990

1.3767 0.5835

5.9795

24.3256

24.3220

25.1646 0.0148

3.4645

Outlet temperature (K) 364.9989 367.2891 364.7241 0.6236

0.6984

Hood temperature (K)

0.4829

Outlet velocity (m/s)

334.6757 333.9749 332.3621 0.2098

the system, with a value of 7500 W/m3 [13]. The interface of the inner and outer air domains was automatically thermally coupled within Fluent. All of the solid– fluid interfaces were automatically defined as thermally coupled walls, and all of the shell–air interfaces were manually defined as coupled walls. Except for the symmetry plane, all other surfaces along the outside of the system were defined as convective boundaries with heat transfer coefficients of 10 W/m2 K with free stream temperatures of 303 K. Lastly, the temperature of the base of the anode rod stubs was set to 723 K [13].

Grid Independence The grid independence study performed can verify the simulation results by observing a few different metrics. A comparison of results for mass flow values through the system, the velocity and temperature observed at the outlet, and the surface temperature of the hood using different mesh sizes at a fixed, normal draft rate can be seen in Table 2. The values do not vary considerably from case to case, which is a good indicator that continuity and energy are consistent across the different meshing regimes. Only one mesh is needed for further analysis, with the results validated for all meshes. The medium mesh was selected for use in the rest of this work.

Results and Discussion As expected, a modification of the draft rate through the cavity results change in temperature of the solid surfaces and the air flowing through. However, the magnitude of the change was considerably more noticeable for a slow draft rate compared to a fast draft rate. As shown in Fig. 2, representing the outlet airflow temperature and hood temperature, respectively, there is a dramatic rise in temperature for the cases involving the slow draft rate. Although a temperature jump to 443 K occurs in the air when operating with the slow draft rate, compared to the 367 K in the normal draft rate case, this does not

The Effect of Reduced Flue Gas Suctioning on Superstructure and Gas …

145

Fig. 2 Varied draft rates using the medium mesh, results for a the average temperature of the airflow at the outlet pipe and b the average temperature of the hood surface

cause it to exceed the limits of the proposed carbon capture system, which passes the process gas through a heat recovery process prior to capture. Test cases have worked successfully in previous work, simulated in Aspen Plus, for temperatures up to 638 K, while under an approximately ten times reduction in draft rate [5]. Regarding the hood temperature, a surface around 365 K exposed to the potroom is not ideal, as it poses safety concerns for the workers. However, modern personal protective equipment (PPE) can allow workers to be safe, as it is not dramatically hotter than the existing conditions, 330–350 K, under the normal draft rate. The temperatures of the internal surfaces of the hood cavity are also affected by the change in draft rate since they are cooled by the ambient air flowing into the hood from the potroom. These surfaces include the anode rods and the ACM at the base of the domain. As can be seen in Fig. 3, the temperatures of the anode rods increase overall, with higher temperatures occurring farther up the rods as the draft rate decreases. Average temperature values for the rods being 631, 590, and 525 K for the slow, normal, and fast draft rates, respectively. Similarly, Fig. 4 shows the temperatures of the ACM, where average temperature values of the ACM were 797, 646, and 444 K for the slow, normal, and fast draft rates, respectively. All of the results indicate a large impact of the draft rate on the temperatures in and around the hood. The change in temperature shows the importance of convection in the system, the main heat transfer mechanism utilized to remove the generated heat from the cell. This will be the focus of improving the feasibility of a draft rate reduction being implemented in a real-world system. If the current design specifications are not adequate for the higher temperatures that will be encountered, then retrofitting will be the best course of action.

Conclusion The model has been shown to be grid independent, which is a fundamental first step in developing any CFD model. It is able to show realistic effects in the hood cavity in terms of temperature variations and flow conditions. It is limited due to

146

B. Velasquez et al.

(a)

(b)

(c)

Fig. 3 Temperature contours of an anode rod for a slow, b normal, and c fast draft rates

(a)

(b)

(c)

Fig. 4 Temperature contours of the ACM for a slow, b normal, and c fast draft rates

assumptions regarding the heat flow source terms, as they would ideally be defined using the components of the system as a whole, but future iterations of the model will include these components, such as electrical effects and magnetohydrodynamics. Additionally, the model will be implemented on the lower half of the cell as well (i.e. the electrolyte bath, ledge, anode blocks, and cathodes). Conflict of Interest The authors declare that they have no conflict of interest.

The Effect of Reduced Flue Gas Suctioning on Superstructure and Gas …

147

References 1. Power Intensive Industries (2014) Orkustofnun [Online]. Available: https://nea.is/hydro-power/ power-intensive-industries/nr/70. Accessed 3 Aug 2022 2. Keller N, Stefani M, Einarsdóttir SR, Helgadóttir ÁK, Helgason R, Ásgeirsson BU, Helgadóttir D, Helgadóttir IR, Tinganelli L, Brink SH, Snorrason A, Þórsson J (2022) National inventory report. The Environment Agency of Iceland, Reykjavik 3. Gislason SR, Wolf-Boenisch D, Stefansson A, Oelkers EH, Gunnlaugsson E (2010) Mineral sequestration of carbon dioxide in basalt: a pre-injection overview of the carbfix project. Int J Greenhouse Gas Control 4:537–545 4. Snæbjörnsdóttir SÓ, Sigfússon B, Marieni C, Goldberg D, Gislason SR, Oelkers EH (2020) Carbon dioxide storage through mineral carbonation. Nat Rev Earth Environ 1:90–102 5. Mathisen A, Sørensen H, Eldr N, Eldrup N, Skagestad R, Melaaen M (2014) Cost optimised CO2 capture from aluminum production. Energy Procedia 51:184–190 6. Shen XC, Hyland M, Welch B (2008) Top heat loss in Hall-Heroult cells. In: TMS light metals, pp 501–504 7. Zhao R, Gosselin L, Fafard M, Ziegler DP (2013) Reduced ventilation of upper part of aluminum smelting pot: potential benefits, drawbacks, and design modifications. In: TMS light metals 8. Taylor MP, Johnson GL, Andrews EW, Welch BJ (2004) The impact of anode cover control and anode assembly design on reduction cell performance. In: Light metals 2004, pp 199–206 9. Taylor MP (2007) Anode cover material science, practice, and future needs. In: Proceedings of 9th Australasian aluminum smelting technology conference 10. ANSYS Fluent (2021) Release 21.2. ANSYS, Inc., Canonsburg, PA 11. Zhai Z, Zhang Z, Zhang W, Chen Q (2007) Evaluation of various turbulence models in predicting airflow and turbulence in enclosed environments by CFD: part-1: summary of prevalent turbulence models. HVAC&R Res 13(6):853–870 12. Zhang Z, Zhang W, Zhai Z, Chen Q (2007) Evaluation of various turbulence models in predicting airflow and turbulence in enclosed environments by CFD: part-2: comparison with experimental data from literature. HVAC&R Res 13(6) 13. Zhao R (2015) Analysis, simulation and optimization of ventilation of aluminum. Laval University, Québec

Assessing the Environmental Footprints of Gold Production in Nevada Saeede Kadivar and Ehsan Vahidi

Abstract Gold has always been regarded as a valuable commodity in high demand throughout history. Nevada is the leading producer of gold in the United States, and gold mining contributes to a significant part of Nevada’s economy. Depending on the types of gold ore and its mineralogical characteristics, several beneficiation and extraction methods are available for gold. Gold can be beneficiated by heap leaching, flotation, roasting, autoclave, or a combination of these techniques. Regardless of the rigorous environmental management standards, different gold processing routes can impact the ecosystem and human health since gold mining is a significant source of hazardous chemicals. In this study, a life cycle assessment (LCA) was conducted to evaluate the environmental performance of four main processes. Using the TRACI method, categories of ozone depletion, global warming, smog, acidification, eutrophication, carcinogenics, non-carcinogenics, respiratory effects, ecotoxicity, and fossil fuel depletion were evaluated for the processes that occurred for gold recovery. Keywords Gold production · Gold processing · Life cycle assessment (LCA)

Introduction Gold has always been regarded as a valuable commodity that has had a global appeal to humans [1, 2]. It is currently the favored material for a variety of applications, including jewelry, electronic devices, and medical diagnostic devices. In contrast to other highly conductive metals like copper and silver, gold does not tarnish or corrode, and this characteristic, along with the fact that it is an incredibly effective electric conductor, makes it a very often used material in circuit boards and many electronic applications [1, 3]. In 2016, Nevada produced over 5.4 million ounces of gold accounting for roughly 75% of all annual U.S. gold production. Some of the world’s major mining companies, including Newmont Mining, Barrick Gold, and Kinross Gold, operate gold mines in the United States [4]. Several extraction S. Kadivar · E. Vahidi (B) University of Nevada Reno, Reno, NV, USA e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_15

149

150

S. Kadivar and E. Vahidi

techniques and processing routes for gold recovery are available, depending on the types of gold ore and its mineralogical characteristics [5]. Gold ores are classified into non-refractory, refractory, and double refractory [5–7]. Non-refractory ore is regarded as the one that is easy to treat with conventional cyanide leaching; however, refractory and double refractory ores are not easily treated by cyanidation because gold is locked within the sulfide minerals in refractory ores [1]. Furthermore, cyanide leaching for gold recovery from double refractory ores is not feasible because gold is encapsulated within both sulfide and carbonaceous materials [7]. Therefore, appropriate pretreatment methods are required before leaching to remove these impurities. These pretreatment methods include flotation, autoclave or pressure oxidation, biooxidation, and roasting [8–10]. A general flow diagram of all different processing routes for gold recovery is shown in Fig. 1. Heap leaching is the most effective method for precious metal recovery from low-grade oxide ores containing 0.5–1.5 g Au per tonne of ore due to its low capital cost compared to other methods [11]. Pressure oxidation extracts gold by oxidizing sulfides into sulfates, allowing gold to be easily leached by cyanidation [8]. Roasting is mainly used for treating double refractory ores to remove carbon and sulfide at a high temperature in the presence of oxygen [12]. The rate of refractory gold ore processing has increased significantly in recent decades, necessitating the development of effective and sustainable technologies that are economically beneficial, environmentally friendly, and safe. In this regard, oxidative roasting is the most common method for treating double refractory ores worldwide [13, 14]. However, it is also accompanied by gaseous emissions of toxic substances such as particulate solids, SO2 , CO, NOx emissions, and arsenic compounds [15]. In addition, global greenhouse gas (GHG) emissions from the gold mining industry have been estimated to exceed 100 million tonnes of CO2 annually. However, there is remarkably little information about country contributions, potential emissions from each step, processing routes of different types of gold ore, and potential solutions for emissions reduction [16]. There are also numerous questions concerning whether gold mining is environmentally sustainable or not, but regardless of how strong and rigorous the environmental management standards are, the reality is that gold mining is an energy-intensive industry [3]. In some communities, such as Australia, contamination from the processing gold production industry has become a major threat to public health and the environment, and it has been reported that Australia’s GHG emissions production was 533.7 million tonnes of carbon dioxide in 2018 [17]. Globally, there are numerous environmental and public health risks associated with gold mining due to ecosystem degradation, vegetation loss, and soil exploitation. In addition, the health of individuals and species is seriously threatened by toxic materials released during the gold mining process, such as cyanide and arsenic compounds [1, 17]. Keeping the negative impacts of gold mining in mind, focusing only on gold production value means ignoring significant sources of negative environmental impacts including GHG emissions generated by electricity consumption and hazardous chemicals, whose use in gold production necessitates expensive and complicated pollution treatment steps [1]. Thus, gold mining in Nevada is recognized as a crucial part of Nevada’s economy [18] and should undergo intense pressure to reduce its environmental footprint throughout all the processing stages.

Assessing the Environmental Footprints of Gold Production in Nevada

151

Fig. 1 General flow diagram of gold recovery for different processing routes

This issue can be addressed by conducting a systematic evaluation of the environmental impacts generated by gold production [1, 3, 16, 19]. There have been a few studies on the environmental burdens of gold mining, but none of them have focused on gold production in the United States, particularly in Nevada. Life cycle assessment (LCA) is an effective tool for evaluating the environmental impacts of a given product [20]. LCA has been extensively applied for evaluating the environmental impacts generated by the metallurgy industry, such as aluminum, steel, zinc, and lead [3]. However, there is little information about the environmental footprints of gold production particularly in Nevada. Thus, picking a mine operated by Nevada Gold Mines allowed us to extrapolate the environmental footprints of gold production with as little error as possible. Therefore, our main goal in this study is to conduct an exact assessment of gold recovery from a double refractory ore using the roasting method.

Material and Methods Life Cycle Assessment (LCA) In this study, LCA methodology was conducted for the environmental evaluation of gold production including four main steps: goal and scope definition, life cycle inventory (LCI), life cycle impact assessment (LCIA), and interpretation.

152

S. Kadivar and E. Vahidi

Fig. 2 System boundary for the LCA study of roasting process

Goal and Scope The main goal of this study is to define the environmental footprints of 1 kg gold production through roasting. Figure 2 shows a simplified system boundary of the roasting process for producing 1 kg of gold. The roasting facility at Goldstrike Mine in Nevada was developed to process 12,000 tonnes of ore per day of double refractory ore. As shown in Fig. 2, the process consists of crushing, dry grinding, roasting, and gaseous treatment circuit from roasting, quenching, neutralization, and carbon in leach followed by carbon stripping and electrowinning [21].

Life Cycle Inventory (LCI) LCI of gold production from the roasting process is listed in Table 1. All the input material, energy consumption, and emissions were based on the functional unit which was 1 kg of gold production. All the data are collected from the Goldstrike Mine Technical Report which is located in the Carlin Complex in Nevada. It is noteworthy that there was a significant lack of data for the process parameters of double refractory gold roasting. Thus, due to little public information in technical reports, some of the input materials are collected from feasibility studies and research reports on the Goldstrike Mine roaster [22]. All the major input materials, energy, and emissions for the roasting process are listed in Table 1. Roasting is also accompanied by gaseous emissions of SO2 , CO, particulate, mercury, and NOx .

Assessing the Environmental Footprints of Gold Production in Nevada

153

Table 1 Life cycle inventory for producing 1 kg of gold from double refractory ores in roasting [21, 22] Stage

Input

Value

Unit

Crushing

Electricity

434.00

kWh

Grinding

Electricity

6678.67

kWh

Grinding media

222.62

kg

Fuel

0.34

m3

Oxygen

7791.78

kg

Fuel

78.92

m3

Quenching

Water

89.04

m3

Neutralization

Lime

890.48

kg

Electricity

99.73

kWh

Electricity

4.45

kWh

Recycled water

545.87

m3

Electricity

299.21

kWh

Water

0.053

m3

Sodium cyanide

89.04

kg

Activated carbon

8.91

kg

Sodium hydroxide

27.9

kg

Hydrochloric acid

11.31

kg

Electricity

3.1

kWh

Sulfuric acid

15.00

kg

Particulate

0.42

kg

Mercury

0.014

kg

Carbon monoxide

5.35

kg

Sulfur dioxide

2.86

kg

Nitrogen oxide

2.88

kg

Roasting

Thickener (dewatering) Carbon in leach (CIL)

Carbon stripping Electrowinning Output Emissions

Assumptions The challenge in this LCA study was the lack of available industrial information about the recovery method of the double refractory ore using the roasting process. Thus, the following assumptions were considered in this study for conducting the LCA study: • • • •

Underground mining Ore grade: 7 g Au/t Pre-treatment method: roasting Extraction process: carbon in leach (CIL)

154

S. Kadivar and E. Vahidi

• Solid to liquid ratio in quenching, thickener, and CIL tanks were 20%, 48%, and 42% • Off-gas treatment is outside of the scope of this study • Doré contains 99% gold. Life Cycle Impact Assessment (LCIA) The LCIA using the TRACI (Tool for reduction and assessment of chemicals and other environmental impacts) method was performed for the environmental impact assessment in this study. The categories of ozone depletion, global warming, smog, acidification, eutrophication, carcinogenics, non-carcinogenic, respiratory effects, ecotoxicity, and fossil fuel depletion were evaluated for the processes that occurred at the Carlin Complex.

Results and Discussion Figure 3 shows the contribution of different stages for processing double refractory ore to produce 1 kg of gold. Based on the results, it can be concluded that grinding and roasting had higher environmental impacts in all categories compared to other stages in the roasting facility. For instance, the grinding stage contributed 35.6% to the total global warming category, while it was 56.9% for the roasting stage. Since the dry grinding method is used at the Goldstrike Mine facility, fuel (natural gas) is used to dry the ore before the roaster [12, 21]. High values of electricity consumption at the grinding stage caused a higher amount of greenhouse gas emissions with a value of 3.75 × 103 kg CO2 eq. Grinding steel media used to grind the ore to the target particle size [22] also showed a significant impact on global warming, producing 1.44 × 103 kg CO2 eq. For the global warming, acidification and smog categories in the roasting stage, fuel and oxygen consumption were the dominant contributors, while electricity consumption for grinding was the major contributor to human toxicity and ecotoxicity. The use of a roaster enables the oxidation of carbon and sulfide compounds in the ore and liberates the gold from the minerals [10, 15]. During the oxidation in the roasting stage, a gas stream of heavy particulate matter, SO2 , CO, and NOx, and significant quantities of mercury have produced that need a cleaning system before releasing to the atmosphere. Furthermore, despite the upstream roaster’s off-gas treatment procedures, air emissions show a significant influence on the environment. It is also noteworthy that using roasting as the pre-treatment for double refractory ore before cyanide leaching had a significant contribution to direct emissions of sulfur dioxide and carbon monoxide to the air [3], contributing to acidification and global warming categories, respectively. In addition, crushing and CIL stage showed smaller contributions being the largest in ecotoxicity and ozone depletion impact categories. Compared to the environmental impacts

Assessing the Environmental Footprints of Gold Production in Nevada

155

in the roasting and grinding stages, contributions of other stages were insignificant resulting in smaller impacts that are not visible in the contribution figure. Table 2 presents the total environmental impacts of all stages in producing 1 kg of gold production using the TRACI method. One kg of gold production produces a high environmental impact in terms of CO2 emissions. It is also shown that ecotoxicity and global warming with values of 5.74 × 104 CTUe and 1.45 × 104 kg CO2 eq were the dominant contributors compared to other impact categories.

Fig. 3 Contributions of stage emissions and input energy and materials at roaster facility using TRACI

Table 2 Total life cycle impact assessment results for producing 1 kg of gold using the TRACI method TRACI/Input

Total

Acidification [kg SO2 -Equiv.]

6.11E+01

Ecotoxicity [CTUe]

5.75E+04

Eutrophication [kg N-Equiv.]

4.12E+01

Global warming [kg CO2 -Equiv.]

1.45E+04

Respiratory effects [kg PM2,5-Equiv.]

2.74E+01

Carcinogenics [CTUh]

3.50E−03

Non-carcinogenics [CTUh]

3.06E−03

Ozone depletion [kg CFC 11-Equiv.]

9.04E−04

Fossil fuel depletion [MJ surplus energy]

1.17E+04

Smog [kg O3 -Equiv.]

7.16E+02

156

S. Kadivar and E. Vahidi

Conclusion This study was focused on the life cycle assessment of gold production from double refractory ore at Goldstrike Mine located in Nevada. All the inventory data and environmental impact category results were based on 1 kg of gold production which was considered our system’s functional unit. The LCIA results showed that the grinding and roasting stages were the dominant contributors to all the impact categories. For instance, for the GHG emissions, due to the significant amount of electricity and fuel consumption given in the input inventory, roasting and grinding had potential effects on global warming, resulting in 8.26 × 103 kg CO2 eq and 5.19 × 103 kg CO2 eq, respectively. Higher CO2 emissions released by the grinding stage show that grinding deserves more attention towards using more green electricity sources to achieve the goal of GHG emissions reduction.

References 1. Norgate T, Haque N (2012) Using life cycle assessment to evaluate some environmental impacts of gold production. J Clean Prod 29–30:53–63 2. Sun WY et al (2019) Life cycle assessment of lead production in China. Mater Sci Forum 944:1123–1129 3. Chen W et al (2018) Life cycle assessment of gold production in China. J Clean Prod 179:143– 150 4. U.S. Geological Survey (2021) Mineral commodity summaries for gold 5. Marsden J, House I (2006) The chemistry of gold extraction. Society of Mining, Metallurgy and Exploration 6. Stenebraten JF et al (2000) Characterization of Goldstrike ore carbonaceous material, vol 1. Society of Mining, Metallurgy and Exploration, pp 7–15 7. Thomas KG et al (2002) Barrick Gold-autoclaving and roasting of refractory ores. Society of Mining, Metallurgy and Exploration 8. Thomas KG, Pearson MS (2016) Pressure oxidation overview in gold ore processing. Elsevier, Amsterdam, pp 341–358 9. Miller P, Brown A (2005) Bacterial oxidation of refractory gold concentrates. Dev Mineral Process 15:371–402 10. Warnica D et al (2002) Design of Barrick Goldstrike’s two-stage roaster. Society of Mining, Metallurgy and Exploration 11. Kappes DW (2002) Precious metal heap leach design and practice. In: Proceedings of the mineral processing plant design, practice, and control, vol 1, pp 1606–1630 12. Buckingham L et al (2001) Barrick Goldstrike roaster facility dry grinding and leaching. Society of Mining, Metallurgy and Exploration, pp 01–141 13. Fernandez RR et al (2000) Process for treating refractory gold ores by roasting under oxidizing conditions. Society of Mining, Metallurgy and Exploration, vol 17, pp 1–6 14. Afenya PM (1991) Treatment of carbonaceous refractory gold ores. Miner Eng 4:1043–1055 15. Cole A et al (2001) Barrick Goldstrike roaster-roasting and gas handling 16. Ulrich S et al (2022) Gold mining greenhouse gas emissions, abatement measures, and the impact of a carbon price. J Clean Prod 340:130851 17. Ulrich S et al (2020) Greenhouse gas emissions and production cost footprints in Australian gold mines. J Clean Prod 267:122118 18. Barrick Gold Corporation sustainability report (2020). The gold standard in sustainability

Assessing the Environmental Footprints of Gold Production in Nevada

157

19. Awuah-Offei K, Adekpedjou A (2011) Application of life cycle assessment in the mining industry. Int J Life Cycle Assess 16:82–89 20. Vahidi E, Zhao F (2017) Environmental life cycle assessment on the separation of rare earth oxides through solvent extraction. J Environ Manage 203:255–263 21. Thomas KG, Buckingham L (1997) Dry grinding at barrack Goldstrike’s roaster facility. International autogenous and semi-autogenous grinding technology 22. Nevada gold mines carlin complex technical report (2020). Report for NI 43-101

Polymeric Composite Dense Membranes Applied for the Flue Gas Treatment Dragutin Nedeljkovic

Abstract Control and reduction of the amount of carbon-dioxide have emerged as one of the main tasks and problems in various fields of industry and production. Therefore, various techniques for treatment of waste gases have been developed. Membrane technology for flue gas treatment emerged as one of the most promising processes for this purpose. Membrane procedures with different types of membranes have huge advantages in comparison with conventional methods. Treatment of gases with various amounts of carbon-dioxides was tested. Different types of membranes are discussed in this paper, their advantages and disadvantages are described, and some basic properties and mechanism of work are presented. Basic types of membranes (polymeric, carbon, and inorganic) were described. Within each category, properties can be bit fine-tuned by using two different materials of the same type for the synthesis. As a separate category, mixed matrix membrane that combines good properties of polymer matrix and inorganic dispersed phase have shown better properties in comparison to any other category of membrane is described. Main disadvantages of any type of membranes are sensitivity to heating and cooling cycles and fouling by condensable components, mainly water. Keywords Flue gas treatment · Dense membranes · Mixed matrix membranes · Carbon capture · Inorganic membrane · Carbon membrane

Introduction Extensive use of the fossil fuel in the recent decades came as a consequence of the rapid development of the industry, and therefore high demand for the energy. At the current stage of the development of the engineering and technology, renewable energy sources (wind, tide, solar) still cannot provide supply that would be reliable over broader period of time at the larger scale. Together with the increased demand for the energy, the environmental concerns came into the consideration. The fossil fuel D. Nedeljkovic (B) College of Engineering and Technology, American University of the Middle East, Egaila, 54200 Kuwait, Kuwait e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_16

159

160

D. Nedeljkovic

produces energy by the combustion, which result in the flue gases. Any impurities present in the fuel would affect the atmosphere and have the negative impact to the human health and the environment. Therefore, two main directions in pollutant removal from the fuel were either to reduce the amount of the pollutant in the fuel itself, or to remove hazardous product of combustion from the flue gases. The main products of the combustion process that are the most harmful for the environment are carbon-dioxide (CO2 ), oxides of nitrogen (usually presented as NOx ), and sulphurdioxide (SO2 ). Although the absolute effect of carbon-dioxide is not as harmful as effect of other two pollutants, its huge excess as product makes it the main target for removal in flue gas treatment. Since the beginning of the research and development of the flue gas treatment, main directions of the research were either to the physical or chemical adsorption of carbon-dioxide [1]. Each of those processes had significant disadvantages. Chemical adsorption would lead to transfer of pollution from one phase to another, at the same time requiring significant amounts of relatively expensive adsorbent which could not be recovered (or could be at very high cost) [2]. On the other hand, physical adsorption process requires a lot of energy for the adsorbent recovery as it must be heated to the appropriate temperature with additional problem of storage of released carbon-dioxide. Cryoscopic methods (methods that are based on cooling of carbondioxide to the low temperatures and its transfer to liquid state) also require huge amounts of energy due to relatively low critical temperature for carbon-dioxide. Membrane processes are relatively simple and efficient, and captured carbon-dioxide can be directly stored which make them very attractive for research and application in the field of flue gas treatment and carbon-dioxide separation [3]. Ideal membrane system would completely allow one component to pass through it (for that component it would be permeable membrane), while it would completely block the passage of the other component(s) through it [that membrane would be impermeable for such component(s)]. Although no real membrane could have those ideal properties, those methods can still give reasonably good separation properties at relatively low costs [4, 5]. Two main criteria are used to estimate the membrane performance: (i) permeability—the amount of the desired gas that can pass through the membrane and (ii) selectivity—ratio of permeability of the desired gas versus permeability of all other gas(es) present in the mixture. According to the permeation mechanism, membranes could be classified in one of five groups [6, 7]: (i) Knudsen diffusion— molecules pass the channel driven by their thermal movements, and channels are sufficiently small in diameter to prevent bulk diffusion; (ii) molecular sieving—only molecules with kinetic diameter lower than the diameter of the membrane channel can pass through the channel; (iii) solution-diffusion mechanism—membrane contains no pores, and molecules are dissolved in the bulk of the membrane and then diffuse through it; (iv) surface diffusion—Diffusion occurs along the surface of the channel, and the diffusion rate is determined by different strength of interactions between molecules and the wall of the surface; (v) capillary condensation—similar to surface diffusion, but it happens at very low pressure which may cause condensation, so molecules of the condensed component are moving by the capillary force through very narrow channels. The main focus of this review is on the third type of the

Polymeric Composite Dense Membranes Applied for the Flue Gas …

161

membranes (solution-diffusion mechanism). As those membranes contain no pores, they are called dense membranes. Mechanism of separation of those membranes is that one of the components of the gas mixture is adsorbed at the surface of the membrane. Adsorbed molecule then diffuses through the thickness of the membrane, and on the opposite side is being desorbed. The side at which adsorption occurs is called feed side or retention side, while the side of the membrane where desorption occurs is called permeation side. The driving force for diffusion is pressure and/or concentration gradient through the thickness of the membrane. The first proposals of idea to use membrane for separation of gases are dated as early as middle of XIX century [8]. The first non-homogenous, anisotropic membrane that was based on the polymer matrix was synthesized in 1961 [9, 10]. Major breakthrough in gas separation came in 1977 when the first membrane that was successfully applied for separation of the gas from the mixture (system was used for the recovery of hydrogen) [11]. Since that time, new generations of material suitable for separation were synthesized with significant improvements in efficiency, durability, speed of separation, and price [12]. Initial research efforts in carbon-dioxide separation were directed towards the treatment of the natural gas [13]. Carbon-dioxide was removed from the natural gas in order to increase its efficiency as energy or heat source and to increase the efficiency of combustion process [14]. Composite membranes based on polymer matrix are in general non-porous, so their separation properties are based on different solubility of gases on the surface of the membrane and different diffusion coefficients through the thickness of the membrane. Permeability of any gas through the membrane can be defined as: P =S·D In this equation, P is the permeability of the membrane for the specific gas in unit cm3 (STP) cm−2 s−1 cmHg−1 ; D is diffusivity coefficient in cm2 s−1 , and S is solubility coefficient in cm3 (STP) cmHg−1 . Common unit for permeability in the membrane community is Barrer, where permeability of 1 Barrer corresponds to permeability of 10–10 cm3 (STP) cm−2 s−1 cmHg−1 . Physical meaning of this unit is that it represents permeability of 1 cm3 of the ideal gas at STP that is permeated through 1 cm thick membrane of surface area of 1 cm2 driven by the driving force of 1 cmHg of pressure gradient in 1 s, multiplied by factor 10–10 . Permeability of the membrane can be calculated by: Q P = l Ap where P is permeability of membrane, l is its thickness, Q is permeation rate through membrane, and p is pressure gradient between feed and permeation side that acts as driving force. Selectivity of membrane is usually expressed as “selectivity of A versus B” and is defined by:

162

D. Nedeljkovic

α A/B =

PA PB

where α A/B is selectivity and PA and PB are permeability of gases A and B respectively. All membranes that are based on polymers (no matter if they are made of pristine polymer or from composite material based on polymer matrix) can be classified as (i) rubbery membranes (operating temperature is above the glass transition temperature of the polymer) or (ii) glassy membranes (operating temperature is below the glass transition temperature of polymer) [15]. Glassy membranes operate at relatively lower temperatures, polymer chains are in more rigid form, and they may need very long time to reach thermodynamic equilibrium. The way the chains are packed in them means that microvoids are present in the bulk of the membrane. Because of that adsorption in glassy membranes happens as the combination of solution and diffusion through the thickness of the membrane and Langmuir adsorption: C = KD f +

C H

CH − CH

The first term in this equation describes solution and diffusion through polymer matrix that is the described by Henry’s Law. K D is Henry’s constant, and f is the partial pressure of gas (or fugacity if real gas conditions must be applied). The second term of the equation describes contribution of the Langmuir’s adsorption in the bulk to the overall uptake of the gas. Langmuir’s relationship described by C H and C H is maximum possible amount of gas that can be adsorbed under given conditions. The main challenge in providing sufficient amount of energy for gas separation is to provide sufficient pressure difference that would provide good driving force for permeation, but that will not mechanically affect the membrane. The most suitable (both technically and financially) way to obtain it is by compression of the feed side [16]. Efficiency of the membranes highly depends on concentration of gas that is being treated. If the concentration of carbon-dioxide is 10%, selectivity of carbondioxide versus nitrogen must be at least 120 in order to obtain membrane that can be comparable to adsorption processes. This requirement somewhat decreases with increase of concentration of carbon-dioxide in treated gas, and at the molar concentration of 20%, it should be around 60 [17]. Current state of the art in this field is that under some circumstances, those values could be achieved [18]. As in general, membrane will have high permeability for all components of the mixture, this means that higher flux for carbon-dioxide would lead to high flux of nitrogen (or oxygen, methane…). So, proper membrane design must include reasonable trade-off and optimization between two inversely proportional effects (selectivity and permeability). This also means that separation membranes will have maximum of performance where any increase of permeability would lead to decrease in selectivity and vice versa. This relation is usually presented as a straight line in logarithmic graph with permeability on x-axis and selectivity on y-axis. This graph is known as Robeson’s plot, and overcoming the limitation of Robeson’s line is one of the biggest challenge that membrane community is currently facing [19]. Besides

Polymeric Composite Dense Membranes Applied for the Flue Gas …

163

permeability and selectivity, suitable membrane must be thermally, chemically, and mechanically stable, be resistant to any kind of decay, and keep its properties over desired number of cycles of exposure to high and low pressure, keeping the price of production and exploitation as low as possible. In this paper, brief presentation of current achievements in the field of dense membranes will be described and discussed.

Current State of the Membrane Technology Polymeric Membranes Membranes made of pristine, natural polymers (e.g. cellulose) were first membranes that were used in praxis for separation processes. As a rule, polymeric membranes have thick and dense layer near the surface supported by less dense non-porous bulk. As working temperatures of the carbon-dioxide separation are (in best case) close to the decomposition temperature of many polymers, most of polymeric membranes are very susceptible on repeated cycles of heating and cooling and exposure to hot steam. Improvement of membranes performance is mainly directed in two directions: (i) increasing diffusion of carbon-dioxide through polymer membrane—this can be achieved by increasing the free volume of the polymer itself (introduction of bulky side groups; decreasing the crystallinity; casting and annealing methods and conditions…) [6]; (ii) increasing solubility of carbon-dioxide in polymers—this is mainly achieved by changes in polymer composition and introduction of groups that can interact with CO2 molecules. Naturally, the best performance will be achieved for materials that would combine both of effects (polyamides, polyacrylates, polyaniline, polypyrrolones…) [20–23]. However, two groups of polymers have shown the best results regarding permeability and selectivity in the category of membranes made of pure polymers: polysulfones and polyimides, with the latter ones being better [24, 25]. Most of the polymers in both of categories can be modified by various side chains or functional groups (mainly by changes in free volume, membrane density and CO2 solubility), and most of different patents and research reports are dealing with effects of different groups to performance and durability of the membrane. Poly(ethylene-oxide) (PEO) is also commonly used as matrix phase as its Hansen solubility coefficient is close to the Hansen coefficient of carbon-dioxide which guarantees strong interactions between carbon-dioxide molecules and polymer chains and high solubility of carbon-dioxide in PEO [26]. Combination of different types of the polymers (either as a blend or as a copolymer) can be understood as a transition between pure polymeric membranes and membranes made of “pure” polymer. General structure of copolymer used for this purpose is that bulk of the membrane is made of soft, flexible rubbery phase and hard, glassy phase as a dispersed phase. In this structure, hard, dispersed phase provides selectivity, while rubbery phase provides good permeability and gives sufficient mechanical stability [27]. So, the main challenge in polymer membrane design

164

D. Nedeljkovic

is to provide that selectivity and permeability would place the membrane above the Robeson’s line. One possible solution is to form thin rods of one polymer randomly distributed in the bulk of another. Properties of those rods are analogous to “classical” molecular sieves [28]. If two types of polymers are not sufficiently compatible to each other, compatibilizing additive must be introduced in order to provide miscibility and good contact between two phases. For example, poly(phenylene oxide) (PPO) has been known for its application for gas separation. If Nylon 6 is added, permeability would decrease with only negligible increase in selectivity. However, if poly(styrene-co-maleic anhydride) (PSMA) is also present in blend, selectivity would be doubled with acceptable decrease in permeability. This decrease in permeability occurs because PSMA prevents phase separation effectively reducing free volume. However, if the amount of additive is too high, it might precipitate which would lead to total mechanical failure of the membrane. Another option is construction of the membrane in the shape of hollow fibers with the best results obtained for annealing of fibers at pressure of 50 kPa, temperature between 100 and 250 °C during time period between 6 and 30 h [29]. Poly(benzimidazole) improves mechanical properties of the membrane as it is flexible and rubbery, indirectly improving permeation properties as well. Cross-linking increases hardness by increase in yield stress reaching carbon-dioxide permeability of 7.9 Barrer and selectivity versus nitrogen of 27 (for comparison, permeability for carbon-dioxide of non-cross-linked poly(benzimidazole) is 0.3 Barrer, and selectivity versus nitrogen is 18). Some aromatic polyimide-based membranes show very good permeation properties at relatively low temperatures. For those membranes, increase of solubility of carbon-dioxide in a bulk overcomes the effect of loss of diffusivity at very low temperatures, resulting in overall improvement in permeability. However, none of those systems obtained properties above Robeson’s line. Presence of various polyketones (polyether, polyaryl, polyimide) can extend exploitation life of the membrane by decreasing plasticization effect of hydrocarbons that are commonly present in treated gases. Presence of polyketones can also increase the durability and resistance to the elevate temperatures and presence of condensable impurities. The best results in the category of polymer membranes were obtained for mixtures of polyamides and polyimides. Those membranes have higher mechanical and chemical resistance in comparison to their counterparts made of pure polymer, without significant losses in permeability and selectivity [30].

Carbon Membranes Carbon membranes can be classified as type of the polymeric membranes that exclusively work as molecular sieve. In comparison to purely polymeric membranes, carbon membranes have better selectivity and permeability, longer durability [31]. Because of those properties, this type of membrane is usually present in chemical reactors and catalytic systems. The main challenge in design and synthesis of carbon

Polymeric Composite Dense Membranes Applied for the Flue Gas …

165

type membranes is proper formation of suitably thin layer. Increase in thickness of layer decreases flux under same permeability. Although carbon membranes are chemically stable, constant exposure to oxygen at the elevated temperatures can cause oxidation and formation of microsized cracks in the structure. This would further lead to accumulation of water (that is commonly present in working environment) and rapid decay in membrane performance. This can be prevented by chemical, thermal, or microwave surface treatment. Another disadvantage of carbon membranes (especially in comparison with usually flexible and soft polymeric membranes) is that they are usually very brittle so their shaping can be very tedious and challenging. If another material is used as a carrier, differences in thermal expansion coefficients between two different materials may cause formation of cracks and mechanical failure of the membrane. As in the case of other types of membranes, properties could be improved by adding various additives that would improve mechanical properties while keeping permeability and selectivity acceptable. Usual additives to carbon membranes are carbides, nitrides, and oxides of various transition metals (Ti, Fe, Zn, Cr…) [32]. Carbon membranes are usually made by pyrolysis of polymeric materials in the absence of oxygen. Resulting membrane is resistant to high temperatures, and separation occurs as a combination of molecular sieves effect and surface diffusion. Properties of carbon membranes are mainly determined by the type of polymeric precursor. The most common precursors are resins and polymer surfactants.

Inorganic Membranes Inorganic membranes are membranes that are based on non-organic, non-polymeric membranes and that can be used for flue gas treatment. Similar to other types of membranes, inorganic membranes could be classified as porous or non-porous. Non-porous membranes are usually based on various transition metals and are applied in hydrogen recovery when separation could be combined with storage [33]. Porous inorganic membranes have significantly lower selectivity, but their price is lower as well. The main advantage of this type of membranes for waste gas treatment is that effectively there is no upper limit in working temperature as those membranes can withstand very high temperatures. According to type of the material that is used for membrane, inorganic membranes could be classified as either zeolite (ones that contain alumosilicate material) and non-zeolite (ones that do not contain alumosilicates) [34]. Frameworks of zeolite materials are usually arranged in well-defined systems with regular network and channels and cages. Those channels act as molecular sieves and their diameter compared to kinetic diameter of gas mixture constituents. As hydrogen is very small, relatively rounded molecule that does not contain any polarity, separation based on molecular sieving is not possible. Therefore, the only possibility for improving properties is functionalization of pores. Most of inorganic membranes are based on ceramic support combined by separation layer that consists of oxides of metals of groups 2, 3, or 4 of Periodic Table [35]. For example, if BaTiO3 layer is deposited on Al2 O3 , with pores that are 5 nm

166

D. Nedeljkovic

in diameter, selectivity of CO2 versus H2 is 3.1, and if pore diameter is decreased to 1 nm, selectivity of the same system increases to 18.4. Permeability of carbondioxide for those systems is about 0.02 Barrer, which puts this system very close to the upper Robeson’s line, making it applicable in real-life systems [36]. In case of zeolite-based membranes, good separation properties for carbon-dioxide from other hydrocarbons are about 0.3 nm, and for separation of CO2 from N2 or CH4 is about 0.4 nm. Combination of those two materials can combine exclusion based on size of molecules (framework, molecular level) with pore diffusion (pores, macroscopic level). The main challenge in this approach is to provide proper positioning of zeolite particles within the pores. If this condition is not fulfilled, voids are present between particles and walls and bulk diffusion would occur, effectively eliminating any potential selectivity [30].

Mixed Matrix Membranes Mixed matrix membranes are relatively new approach in the field of waste gas treatment. In general, any membrane that consists of matrix phase and dispersed phase that belong do different types of materials can be classified as mixed matrix membrane. In praxis, the most common (practically, the only feasible construction) is system in which zeolite powders are dispersed in polymer matrix. This type of membrane is the only one that combines three different effects that are acting in a synergy: (i) Zeolite powders are acting as molecular sieves with their openings and channels; (ii) structure of polymer matrix can be disrupted leaving microcavities that increase permeability; however, this can cause rapid drop in selectivity; and (iii) carbondioxide is being dissolved in polymer matrix and diffuses through the thickness of membrane. This approach that takes into account contributions of various effects effectively overcomes limitations of only one type of material, and mixed matrix membranes routinely have performance above the Robeson’s line. Performance of mixed matrix membrane could be evaluated by following equations: l2 6θ   V p l p p2 − p p1   P = D·S= p +p A RT t p f − ( p2 2 p1 ) D=

where S is solubility, D is diffusivity, P is permeability, V p is the permeate volume, l is the thickness of the membrane, R is the universal gas constant, t is for the time required for permeate pressure to increase from value pp1 to value pp2 , pf is feed pressure, θ is time required for pressure on the permeate side to reach constant value. Matrix phase of mixed matrix membrane is usually block-co-polymer that contains one block that provides mechanical stability and another block that is

Polymeric Composite Dense Membranes Applied for the Flue Gas …

167

responsible for diffusion and solubility. Various commercial polymers could be used for this purpose, for example poly(amide-b-ether)-b-poly(ethylene oxide) or poly(ethylene glycol)-co-poly(butylene terephthalate) [18]. Powder that is used for dispersed phase should consist of sufficiently small particles, to keep thickness of membrane below 300 µm. Various zeolite powders fulfill this requirement, and the best results were obtained with powders that contain 3-dimensional pores (IWS or FAU) as the orientation of the pores in that case plays no role in gas transport [37]. As mixed matrix membranes contain highly non-polar long polymer chains and very polar zeolite particles, proper additive that would provide good mixing must be added. The best solution for the additive is n-tetradecane trimethyl ammonium bromide (n-C14-TMABr) that contains long, non-polar, hydrophobic chain as well as highly polar nitrogen-bromine bond. Presence of additive prevents formation of voids, thus preventing bulk diffusion and loss of selectivity. Another type of surface agent that may improve performance is either inorganic ionic liquids or organic amines. In both of cases, powder is treated with agent with certain basic properties, which increases the affinity towards acidic carbon-dioxide. Mixed matrix membranes based on polymer matrix and zeolite powder can reach permeability of carbon-dioxide of 120 Barrer and selectivity versus nitrogen as high as 60. Alternative to zeolite particles may be crystalline silica, which would improve selectivity and permeability in a similar way as zeolites. Contribution of silica to the permeability and selectivity will not be as strong as in the case of zeolites particles, but silica is in general more compatible with polymer matrix, so the probability of formation if microcavities will be lower. Surface diffusion can be further improved by addition of activated carbon to silica particles [38]. Carbon particles are between 1 and 5 µm in diameter, and their presence increases permeability by 70% and selectivity by 100% compared to silica without additive [39]. Silica can also be incorporated in polysulfone matrix improving its performance. Pure polysulfone is not suitable for membrane synthesis as its properties put it relatively low on Robeson’s graph. However, addition of mesoporous silica significantly improves both permeability and solubility effectively positioning this membrane close to the Robeson’s line. Increased fraction of silica creates more microvoids in membrane structure increasing permeability of carbon-dioxide and decreasing the selectivity. It should be noted that effect of increased permeability is much stronger that the effect of decrease in selectivity which means that overall properties of membrane are improved. Opposite effect can be obtained by treating silica particles with organic amines. In that case, number and size of microvoids are reduced, so permeability is decreased, but selectivity is increased due to acid–base reaction of carbon-dioxide with amines. As in the previous case, positive effect of selectivity overcomes the negative effect of permeability. In both of mentioned cases, amount of amine or silica must be carefully determined and controlled in order to prevent mechanical failure of membrane. In ideal case, structure of mixed matrix membrane is converted to asymmetric membrane. The main advantage of asymmetric membrane is that separation occurs at very thin active layer supported by thicker, bulkier carrier. As the active layer is very thin, separation occurs at very high flux. Another advantage of asymmetric membrane is that it can be easily shaped in hollow-fiber shape that

168

D. Nedeljkovic

is very suitable for industrial application as separation equipment take significantly smaller volume in comparison to flat membranes. The main disadvantage of any kind of membranes based on molecular sieves is that accumulates condensable impurities over the time of exploitation. Water is normal product of hydrocarbon combustion and can be condensed during various cycles of heating and cooling when flue gases are treated. Those condensed molecules can accommodate very small space of the zeolite framework effectively blocking the passage of carbon-dioxide molecule through zeolite particle.

Conclusion and Future Perspectives As it is discussed and described, suitable membrane for carbon-dioxide separation should combine various contributions of all available materials that are suitable for membranes. Combination of materials can be polymer–polymer blend, co-polymer, or polymer-zeolite composite material. Therefore, first direction for future research would be to synthesize more membranes that would have permeability-solubility combination above the Robeson’s line. On the other hand, other direction would be to prepare material that would be sufficiently cheap, and that would keep its properties over extended period of time and number of repeated cycles. So, in the future we can expect reports about materials (both composition and synthesis procedures) with improved chemical, thermal, and mechanical strength. Other research interest may be avoiding plasticization due to the presence of condensable matter (most commonly water, but some hydrocarbons may also be present). Currently, very few data are available about effect of gases that are present in waste gases in low concentration, but that are very reactive and harmful (NH3 , CO, H2 S, SOx …). Having all aforementioned in mind, it may be concluded that carbon-dioxide separation is promising field for further development and that researchers from various fields of engineering will get engaged into it.

References 1. Aaron D, Tsouris C (2005) Separation of CO2 from flue gas: a review. Sep Sci Technol 40:321– 348 2. Rao AB, Rubin ES (2002) A technical, economic, and environmental assessment of aminebased CO2 capture technology for power plant greenhouse gas control. Environ Sci Technol 36:4467–4475 3. Basu A, Akhtar J, Rahman MH, Islam MR (2004) A review of separation of gases using membrane systems. Pet Sci Technol 22:1343–1368 4. Baker RW (2002) Future directions of membrane gas separation technology. Ind Eng Chem Res 41:1393–1411 5. Powell C, Qiao G (2006) Polymeric CO2 /N2 gas separation membranes for the capture of carbon dioxide from power plant flue gases. J Membr Sci 279

Polymeric Composite Dense Membranes Applied for the Flue Gas …

169

6. Paul DR (2017) Polymeric gas separation membranes. CRC Press, Boca Raton. http://doi.org/ 10.1201/9781351075886 7. Dhorali G, Reddy BSR (2012) Cost effective poly(urethane-imide)-POSS membranes for environmental and energy-related processes. Clean Technol Environ Policy 15 8. Graham T (1866) LV. On the absorption and dialytic separation of gases by colloid septa. Lond Edinb Dublin Philos Mag J Sci 32:401–420 9. US3133132.pdf 10. Li D, Wang H (2010) Recent developments in reverse osmosis desalination membranes. J Mater Chem 20:4551–4566 11. Sridhar S, Bee S, Bhargava S (2014) Membrane-based gas separation: principle, applications and future potential. Chem Eng Dig 1(1):1–25 12. Spillman R (1995) Chapter 13 economics of gas separation membrane processes. In: Noble RD, Stern SA (eds) Membrane science and technology, vol 2. Elsevier, Amsterdam, pp 589–667 13. Mazur WH, Chan MC (1982) Membranes for natural gas sweetening and CO2 enrichment. Chem Eng Prog U.S. 78:10 14. Coady AB, Davis JA (1982) Separation techniques: CO2 recovery by gas permeation. Chem Eng Prog U.S. 78:10 15. Yampolskii Y (2012) Polymeric gas separation membranes. Macromolecules 45:3298–3311 16. Kanehashi S, Nagai K (2005) Analysis of dual-mode model parameters for gas sorption in glassy polymers. J Membr Sci 253:117–138 17. Favre E (2007) Carbon dioxide recovery from post-combustion processes: can gas permeation membranes compete with absorption? J Membr Sci 294:50–59 18. Nedeljkovic D (2021) Homogenization of the dense composite membranes for carbon dioxide separation. In: Baba AA et al (eds) Energy technology. Springer International Publishing, Cham, pp 51–60 19. Robeson LM (1991) Correlation of separation factor versus permeability for polymeric membranes. J Membr Sci 62:165–185 20. Du N, Park HB, Dal-Cin MM, Guiver MD (2012) Advances in high permeability polymeric membrane materials for CO2 separations. Energy Environ Sci 5:7306–7322 21. Mustafa J, Farhan M, Hussain M (2018) CO2 separation from flue gases using different types of membranes. J Membr Sci Technol 6:1–7 22. Gerhard I (2001) Polyaniline membrane, for use in gas separation, pervaporation and liquid phase separation, has layer(s) of polyorganosiloxane 23. Koros WJ, Walker DRB (1993) Polyamides and polypyrrolones for fluid separation membranes 24. Makino H, Kusuki Y, Harada T, Shimazaki H, Isida T (1985) Aromatic polyimide composite separating membrane 25. Mushtaq A, Mukhtar H, Shariff AM (2014) Gas permeability study of amine-polymer blend membranes to separate carbondioxide from natural gas. Appl Mech Mater 625:704–708 26. Akhtar FH et al (2019) Scalable synthesis of amphiphilic copolymers for CO2 − and waterselective membranes: effect of copolymer composition and chain length. Macromolecules 52:6213–6226 27. Rybak A, Grzywna Z, Sysel P (2013) Mixed matrix membranes composed of various polymer matrices and magnetic powder for air separation. Sep Purif Technol 118:424–431 28. Ahn J et al (2010) Gas transport behavior of mixed-matrix membranes composed of silica nanoparticles in a polymer of intrinsic microporosity (PIM-1). J Membr Sci 346:280–287 29. UV and thermally treated polymeric membranes. Patent US-9492785-B2. PubChem. https:// pubchem.ncbi.nlm.nih.gov/patent/US-9492785-B2 30. Rahimalimamaghani A, Pacheco Tanaka DA, Llosa Tanco MA, Neira D’Angelo MF, Gallucci F (2022) Ultra-selective CMSMs derived from resorcinol-formaldehyde resin for CO2 separation. Membranes 12:847 31. CO2 selective carbon tubular membrane: the effect of stabilization temperature on BTDATDI/MDI P84 co-polyimide (2018). Int J Eng 31 32. Metal oxide nanoparticle filled polymers. Patent US-7510595-B2

170

D. Nedeljkovic

33. Kayvani Fard A et al (2018) Inorganic membranes: preparation and application for water treatment and desalination. Materials 11:74 34. Price GD (1989) A. Dyer 1988. An introduction to zeolite molecular sieves. Wiley, Chichester, New York, Brisbane, Toronto, Singapore. Geol Mag 126:446–447 35. US patent: US-9321014-B2. Hollow fiber membrane with compatible reinforcements | Unified patents 36. Liguori S (2018) Chapter 14 design considerations for postcombustion CO2 capture with membranes 37. Nedeljkovi´c DM et al. The application of zeolite nanopowder for the construction of the dense composite polymer membranes for carbon dioxide separation 38. Mixed matrix membrane for separation of gases. Patent WO-03039728-A3. PubChem 39. Ghanbari H, Saxen H, Grossmann I (2013) Optimal design and operation of a steel plant integrated with a polygeneration system. AIChE J

Molten Salt Mg-Air Battery Improvement and Recharging Mahya Shahabi, Nicholas Masse, Amanda Lota, Lucien Wallace, Heath Bastow, and Adam Powell

Abstract Decarbonization of the shipping industry along with other long-haul transportation is among the toughest and most important challenges to resolve toward greenhouse emission elimination. A molten salt magnesium-air battery shows promise toward meeting this challenge. This work presents the improvement achieved in long-run battery performance through solidifying the MgO reaction product from the molten salt electrolyte by using a cold finger. Moreover, preliminary result on Mg reduction at the battery anode is discussed which leads to the recharging of these batteries. Finally, a battery design in a 20-foot shipping container could deliver 60–90 MWh of energy, which is 15–22 times the energy of containerized lithiumion batteries, at a fraction of the cost. Disadvantages include the inability to scale down due to high-temperature operation, and lower round-trip efficiency than Li-ion batteries. Keywords Metal-air battery · Directional solidification · Oxygen reduction · Zero-emission shipping

Introduction Current Energy Storage Technology An energy storage system usually comprises mainly batteries, control and power conditioning system, and the rest of the plant [1]. The batteries, which are the major components, are made of stacked cells in which the chemical energy is converted to electrical energy and vice versa [1]. The desired battery voltage and current levels are obtained by electrically connecting the cells in parallel and series [1]. Energy and power capacity are the main features for rating different batteries [1]. Roundtrip efficiency, cycle life, operating temperature, depth of discharge (the extent to which M. Shahabi · N. Masse · A. Lota · L. Wallace · H. Bastow · A. Powell (B) Worcester Polytechnic Institute, Worcester, MA, USA e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_17

171

172

M. Shahabi et al.

a battery can discharge less than %100), self-discharge, and energy density are other important features describing a battery [2, 3]. Researchers now study multiple battery technologies, some of which are commercially available, and some still in the experimental stage [4]. Current power systems batteries applications are deep cycle batteries with energy capacity ranging from 17 to 40 MWh and an efficiency of 70–80% [1]. Some of the most widely used batteries in this category along with other energy sources are listed below: • Lithium-ion (Li-ion): These batteries have lithiated metal oxide for cathode and graphitic carbon with a layer structure for anode [5]. The electrolyte is lithium salts dissolved in organic carbonates [5]. Li-ion batteries with high energy density (up to 705 Wh/L) and power density (up to 10,000 W/L) are widely used in power sources for cell phones, laptops, and other portable devices [6]. The acceptable temperature region for these batteries is −20 to 60 °C [6] with optimal range of 15–35 °C [6]. • Lead acid: These batteries comprise a cathode electrode of lead dioxide and an anode electrode of sponge lead separated by a micro-porous material. Anode and cathode pairs are immersed in an aqueous sulfuric acid electrolyte. • Sodium sulfur (NaS): The cathode in these batteries consists of molten sulfur, and the anode is molten sodium separated by a solid beta alumina ceramic electrolyte. The operating temperature for these batteries is about 300 °C to keep the electrode materials in a molten state [7]. • Metal air: The cathodes in these batteries are air electrodes often made of porous carbon structure, metal mesh covered with proper catalysts [8]. The anodes are commonly available metals with high energy density such as aluminum [9], zinc [10], iron[11], or magnesium [12] that release electrons when oxidized. The electrolytes are often a good hydroxide (OH–) ion conductor in liquid form, a saturated solid polymer membrane [1], or in the case of Mg-air molten chloride salt [12]. • Ammonia: Ammonia is a good energy vector with 22.5 MJ/kg specific energy at a higher heating value (HHV) and 11.5 MJ/L energy density for raw ammonia. It has drawn a lot of attention as its only product is water from either combustion or fuel cells [13]. However, ammonia’s toxicity is the major drawback of this technology [14]. Ammonia has much lower cost to store, and deliver when compared to hydrogen [6]. • Hydrogen: Hydrogen is a non-toxic, clean energy carrier and has higher specific energy than gasoline [15]. With 8.491 MJ/L energy density [13] and 141.9 MJ/Kg specific energy [16], hydrogen has good advantages as a clean energy media; however, storage of hydrogen in liquid form is difficult and requires very low temperatures and/or high pressures [16]. Mg metal is potentially a viable zero-emission fuel with higher energy/volume than hydrocarbons and five times the energy/mass of iron as shown in Figure 1 [17]. It is also abundant with 2 × 1015 tons of it in the oceans, and non-toxic [12]. Comparing Mg-air battery to other metal-air batteries, although the energy density of similar Si-air and Al-air is in principle higher than Mg-air, the volatility of SiCl4 , which boils at −60 °C, and toxicity of AlCl3 make Mg a more suitable material.

Molten Salt Mg-Air Battery Improvement and Recharging

173

Energy density [kW.h/L]

45

Data from Powell 2022

B

40 35

Be

30 25

Ti

20

Al Si Mg with MAB Efficiency Advantage

Fe

15

Mg Disel Zn 10 Gasoline Coal LNG 5 DOE H2 Ammonia Li Biomass CNG 0 Li Battery 0

5

10

15

Liquified H2 Compressed H2 Gas 20

25

30

35

Specific Energy[kW.h/kg]

Fig. 1 Energy density and specific energy comparison of metals and fuels

Current Technologies with Their Criteria Performance Pressure to reduce shipping’s GHG footprint has risen sharply and keeps growing [18]. The International Maritime Organization (IMO) is the most influential regulator in the shipping industry and requires them to reduce their shipping emissions [18]. A fully electric cruise ship, Yangtze River Three Gorges 1, started its first voyage in March 2022. It has a 7.5 MWh battery capacity which is equivalent to over 100 EVs. Using this ship will result in 1660 fewer tons of toxic emissions each year; however, the current range for each charge of this ship is very limited at about 100 km [19]. Another current fully electric ship, Danish ferry Ellen, is an all-electric ferry with a battery of 4.3 MWh capacity and a travel distance of 40 km. This is also the first ferry with no emergency diesel generator on board [20]. With two high-end technologies at the same time, Yara Birkeland will be the world’s first fully electric and autonomous container vessel producing zero emissions [21]. With a 6.8 MWh battery, Yara Birkeland will sail between Herøya and Brevik (only 13 km) carrying chemicals and fertilizer [22]. MV Ampere is an electric car ferry currently operating in Norway. It has two onboard 450 kW electric motors, powered by lithium-ion batteries and an overall output of 1 MWh traveling 7 km in Norway [23]. The current challenge with fully electric ships is the battery’s energy density and consequently short travel distances. The Mg-air battery is designed to solve the toughest transportation decarbonization challenges, by efficiently electrifying of long-distance shipping.

174

M. Shahabi et al.

Overview This study discusses the recent improvement of the molten salt Mg-air battery [11]. Using a cold finger in the battery setup will freeze out the MgO from the molten salt which makes the long-term operation of these batteries possible. A rough battery design in a 20-foot shipping container illustrates the opportunity for very high energy density using this technology, and a cost model of these batteries shows their potential for very low cost.

Materials and Manufacturing Mg-Air Cold Finger Experiment The earlier high-temperature experiments were performed as a proof of concept to measure the open-circuit voltage (OCV) and voltage-current characteristics of the Mg-air battery [12]. In a more recent experiment, the effect of the cold finger on filtering out the MgO from the molten salt electrolyte was studied. To start the experiment, the electrolyte salt (MgCl2 20 wt%, NaCl 32 wt%, and KCl 48 wt%, eutectic temperature 403 °C [24]) was added to the crucible and placed in a furnace of 300 °C to prebake and eliminate its excess moisture content. The crucible was then put in the experimental cage as shown in Fig. 2. Next, the cathode which consisted of a stainless-steel tube with porous nickel mesh at the bottom was set up to stay above the crucible. The nickel mesh has 2 cm2 of exposed area. Similarly, the anode which was a stainless-steel threaded rod with a cylinder shape Mg with 2.5 cm diameter and 5 cm height screwed to the bottom was set on top of the crucible. The cold finger consists of two tubes nested within one another. The inner tube is a 304 stainless-steel tube with a 0.125 outer diameter (OD) and was used for injecting the compressed air. The outer tube was a stainless-steel tube with 3/8 OD with a close-ended Yor-Lok connection. After isolating the crucible from the cage and connecting the thermocouples to the system, the experimental cage was inserted into the furnace and the system was sealed to minimize oxygen interaction during temperature ramp-up. Before starting the heat ramp-up, the oxygen partial pressure was reduced to below 0.05 bar by injecting argon gas into the furnace. First, the temperature was set to 250 °C at 10 °C/min and held for 60 min in order to remove water with minimal oxide formation. Then, the temperature was increased to 550 °C at the same rate. The battery was ready to perform once the salt was fully molten. Next, the cathode’s air flow of 1 standard liter/minute was started and inserted into the molten salt bath along with the anode. OCV and voltage versus current data were recorded using a potentiometer to sweep resistance from 0 to 25 . After 1.5 h of battery operation and data logging, the cold finger flow was started and it was inserted into the molten salt. Due to the cool-down effect of the cold finger, the MgO will most likely deposit on the cold finger and filter out from the molten salt, reducing the

Molten Salt Mg-Air Battery Improvement and Recharging

175

Fig. 2 Experimental setup using cold finger

MgO concentration in the salt. When ending the test, all anode, cathode, and cold finger leads were drawn out of the bath, the cathode and cold finger flows were shut off and lastly, and the furnace was turned off.

Results and Discussion Experimental Results To identify the cold finger effect on Mg-air battery performance, a study similar to the one from [12] was performed and a voltage-current plot was studied. The electrical circuit consists of the battery, a 0.02  shaft resistor, a 0–25  potentiometer, and a voltmeter to measure the battery voltage. At the beginning of the test, the OCV was 1.58 ± 0.01 V. Starting 10 min into the test four resistance sweep test was performed with a 10-min interval between the first three tests and 40 min interval for the last one. In these tests, the resistor was decreased from 25 to 0  and then back to 25 . As shown in Fig. 3, the voltage for different currents decreases as the battery is used which can be due to the increasing composition of MgO in the molten electrolyte. After 2.5 h of the battery running the cold finger cooling down flow was started at 1000 SCCM, and the cold finger was inserted into the molten electrolyte. 5 min after cold finger initiation, the OCV started to increase from 0.55 to 1.35 V. The cold finger flow was then increased which resulted in a further increase of OCV to 1.39 V

176

M. Shahabi et al.

Fig. 3 Voltage-current plot from experiment

Fig. 4 a Experimental OCV. b Anode and cold finger after experiment

as shown in Fig. 4a. Figure 4b shows the Mg and cold finger after the experiments. It is visible that a layer of oxides was formed on the cold finger in this process.

Model of Full-Scale Mg-Air Battery Performance and Costs As discussed in [12], at maximum power, the current density of a single lab scale Mg-air battery is 2.3 A/cm2 at 100% transformation of one electron per atom of Mg. According to the theoretical battery model, the OCV can be estimated as 2.6 V, and

Molten Salt Mg-Air Battery Improvement and Recharging

177

Fig. 5 Full scale battery using 20-foot shipping container

at this current density, the model estimates operating voltage of 1.3 V. This will result in a power density of 2.93 W/cm2 . Considering a single anode–cathode pair from the 20-foot full-scale battery, as shown in Fig. 5, the Mg slab’s dimensions are 1.5 × 2 × 0.45 m3 resulting in a 30,000 cm2 surface area on the side. The power of a single cell is then estimated as 87.96 kW with an efficiency of 40.89% operating over 75 h. 10 anode and cathode pairs connected in series will create a full-size 20-foot battery as shown in Fig. 5. At 40% efficiency, the theoretical current density of this cell is 2.3 A/cm2 and the cell voltage is 1.3 V, creating a power density of 2.9 W/cm2 . With 10 cells in a stack, the overall power in a container ship is 880 kW. With 75.1 h operating time, the capacity of the cell is approximately as 66 MWh. Similarly, at 0.2 A/cm2 the battery achieves 2.5 V, 0.5 W/cm2 power density, 81% efficiency, 151 kW power, and 130 MWh capacity—assuming perfect current efficiency and oxide removal from the bath. Based on the experimental results [12], the MAB showed 0.4 W/cm2 power density or 109 KW power on the full scale. Voltage-current profile has shown 50–80% of theoretical voltage, likely due to dissolved oxide in the molten salt. If this performance holds for a larger battery, this will lead to 60–90 MWh energy delivery from the 20-foot container battery. The biggest cost contributors to Mg-air full-scale batteries are nickel cathode material, in the worst-case scenario. The overall cost of these cells is $315–365 k for 60–90 MWh.

178

M. Shahabi et al.

Summary and Conclusions Key Findings • With 60–90 MWh battery capacity, the Mg-air battery has at least 8 times the battery capacity of other shipping batteries making transpacific zero emission cruising possible. • The cold finger demonstrates that long-term battery operation is possible by selective solidification of MgO to maintain MgO composition in the molten salt. • The MgO formed on the cold finger can be used for Mg electrolysis and recharging of the battery. • This molten salt Mg-air battery concept shows potential for much higher energy density at comparable cost versus other metal-air chemistries such as iron-air [25].

References 1. Divya KC, Østergaard J (2009) Battery energy storage technology for power systems—an overview. Electr Power Syst Res 79(4):511–520. https://doi.org/10.1016/j.epsr.2008.09.017 2. High-temperature batteries 3. Selman JR, Steuenberg RK, Barghusen JJ, Howard WG (1976) Proceedings of the symposium and workshop on advanced battery research and design, 22–24 Mar 1976 [Held at ANL]. Argonne National Lab., III (USA), ANL-76-8; CONF-760312-, Jan 1976. Accessed: 15 Sept 2022 [Online]. Available: https://www.osti.gov/biblio/7339886#page=91 4. Martins LS, Guimarães LF, Botelho AB Jr, Tenório JAS, Espinosa DCR (2021) Electric car battery: an overview on global demand, recycling and future approaches towards sustainability. J Environ Manag 295, p 113091, Oct 2021. https://doi.org/10.1016/j.jenvman.2021.113091 5. Nitta N, Wu F, Lee JT, Yushin G (2015) Li-ion battery materials: present and future. Mater Today 18(5):252–264. https://doi.org/10.1016/j.mattod.2014.10.040 6. Ma S et al (2018) Temperature effect and thermal impact in lithium-ion batteries: a review. Progr Nat Sci Mater Int 28(6):653–666. https://doi.org/10.1016/j.pnsc.2018.11.002 7. Oshima T, Kajita M, Okuno A (2005) Development of sodium-sulfur batteries. Int J Appl Ceram Technol 1(3):269–276. https://doi.org/10.1111/j.1744-7402.2004.tb00179.x 8. Drenckhahn WW et al (2013) A novel high temperature metal—air battery. ECS Trans 50(45):125–135. https://doi.org/10.1149/05045.0125ecst 9. Yang S (2002) Design and analysis of aluminum/air battery system for electric vehicles. J Power Sources 112(1):162–173. https://doi.org/10.1016/S0378-7753(02)00370-1 10. Toussaint G, Stevens P, Akrour L, Rouget R, Fourgeot F (2010) Development of a rechargeable zinc-air battery. ECS Trans 28(32):25–34. https://doi.org/10.1149/1.3507924 11. McKerracher RD, Ponce de Leon C, Wills RGA, Shah AA, Walsh FC (2015) A review of the iron-air secondary battery for energy storage. ChemPlusChem 80(2):323–335. https://doi.org/ 10.1002/cplu.201402238 12. Shahabi M, Masse N, Sun H, Wallace L, Powell A, Zhong Y (2022) Design of a molten salt metal-air battery with high-energy density. In: Tesfaye F, Zhang L, Guillen DP, Sun Z, Baba AA, Neelameggham NR, Zhang M, Verhulst DE, Alam S (eds) REWAS 2022: energy technologies and CO2 management, vol II. Springer International Publishing, Cham, pp 47–57. https://doi. org/10.1007/978-3-030-92559-8_6

Molten Salt Mg-Air Battery Improvement and Recharging

179

13. Lan R, Tao S (2014) Ammonia as a suitable fuel for fuel cells. Front Energy Res 2. https://doi. org/10.3389/fenrg.2014.00035 14. Reich CM, Kaiser A, Irvine JTS (2001) Niobia based rutile materials as SOFC anodes. Fuel Cells 1(3–4):249–255. https://doi.org/10.1002/1615-6854(200112)1:3/4%3c249::AIDFUCE249%3e3.0.CO;2-A 15. Polat T, Yalçin AH, Sahin HM (2012) Study on processes of hydrogen production based nuclear energy, p 21 16. Midilli A, Ay M, Dincer I, Rosen MA (2005) On hydrogen and hydrogen energy strategies I: current status and needs. Fuel Energy Abstr 46(4):222. https://doi.org/10.1016/S0140-670 1(05)81477-8 17. Powell A (2022) A magnesium clean energy ecosystem vision. In: Maier P, Barela S, Miller VM, Neelameggham NR (eds) Magnesium technology 2022. Springer International Publishing, Cham, pp 121–126. https://doi.org/10.1007/978-3-030-92533-8_20 18. DNV_Maritime_Forecast_2050_2021-Web.pdf 19. Doll S (2022) ‘World’s largest electric cruise ship’ makes maiden voyage in China with a whopping 7,500 kWh in battery power. Electrek, 31 Mar 2022. https://electrek.co/2022/03/31/ worlds-largest-electric-cruise-ship-makes-maiden-voyage-in-china-with-a-whopping-7500kwh-in-battery-power/. Accessed 15 Sept 2022 20. Batteries powering new cruise ship AND e-ferry—Plugboats, 03 July 2019. https://plugboats. com/batteries-powering-new-cruise-ship-and-e-ferry/. Accessed 15 Sept 2022 21. Yara Birkeland | Yara International, Yara None, 12 Nov 2021. https://www.yara.com/newsand-media/press-kits/yara-birkeland-press-kit/. Accessed 15 Sept 2022 22. Ship YARA BIRKELAND (Container Ship) Registered in Norway—Vessel details, current position and Voyage information—IMO 9865049, MMSI 257646000, Call Sign LFQX | AIS Marine Traffic. MarineTraffic.com. https://www.marinetraffic.com/en/ais/details/ships/shi pid:6407606/mmsi:257646000/imo:9865049/vessel:YARABIRKELAND. Accessed16 Sept 2022. 23. Ampere Electric-Powered Ferry (2022) Ship technology. https://www.ship-technology.com/ projects/norled-zerocat-electric-powered-ferry/. Accessed 15 Sept 2022 24. Xu X et al (2018) Experimental test of properties of KCl–MgCl2 eutectic molten salt for heat transfer and thermal storage fluid in concentrated solar power systems. J Sol Energy Eng 140(5):051011. https://doi.org/10.1115/1.4040065 25. Wood DL, Li J, Daniel C (2015) Prospects for reducing the processing cost of lithium ion batteries. J Power Sources 275:234–242. https://doi.org/10.1016/j.jpowsour.2014.11.019

Superconductor Busbar Systems in the Light of Increased Energy Costs Wolfgang Reiser, Till Reek, Claus Hanebeck, and Peter Abrell

Abstract The Russian invasion of Ukraine led to a shortage of commodities and increased energy costs. In this situation, it is important to realize that every MWh saved does not need to be generated. Direct current transmission with superconductors takes place without electrical resistance and without line losses. The potential use cases in the aluminum industry presented during TMS 2021 will be highlighted and re-evaluated with current energy costs. The relation between conventional and the innovative superconductor technology is changing. Technologies previously considered unfeasible gain importance. In this paper, different variants of superconductors and refrigeration technologies are presented with respect to investment and operating costs. Alternative variants are described technically and evaluated economically. Investment costs are indicated as well as costs for maintenance, operation, and total cost of ownership for various timelines. A special chapter will describe the advantages of a combined pipeline for the transfer of liquid hydrogen and superconductor electricity. Keywords Superconductor · Busbar · Efficiency · Cost · CO2 reduction · Hydrogen

Shortage and Increase of Energy Costs With the Climate Change Act of 2021, the German federal government has again significantly raised its climate protection targets: • Total emission reduction of 65% compared to 1990. • Greenhouse-gas (GHG) neutrality in 2045. Effective climate protection is of course a global endeavor demanding longterm, internationally comparable goals and stronger global, or at least pan-European, governance. W. Reiser · T. Reek · C. Hanebeck (B) · P. Abrell Vision Electric Super Conductors GmbH, Kaiserslautern, Germany e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_18

181

182

W. Reiser et al.

In addition, future technological developments, for example in technologies such as hydrogen or superconductors can lead to other long-term technology paths. Europe and especially Germany must undertake the greatest transformation in its post-war history. Legally mandated GHG neutrality by 2045 requires a fundamental restructuring of the German energy system and large parts of industry. To achieve the legally set climate targets, Europe needs to largely eliminate fossil technologies as soon as possible. Coal-fired power generation as the main source for GHG emissions must fall much faster than previously planned. Renewable electricity generation needs to expand to the limits of its capacity. To achieve the mandated climate target, coal-fired power generation will phase out in Germany by 2030 under the assumptions made, in parallel to the end of nuclear energy by 2022. Taking these capacities off the grid while maintaining a secure electricity supply will require the addition of more than 40 GW of new (“H2-ready”) gas-fired power plants. This plan has now failed due to the Russian war in Ukraine. The unlimited supply of energy on lowest cost levels has stopped in 2022. Energy, especially natural gas, is not readily available anymore and has become a limited and costly commodity. The dramatic increase of electricity costs has two components: • The shortage of natural gas • The merit order effect. Prices on the gas and electricity markets have reached unprecedented heights. A great shortage is present on the global gas market. Gas-fired plants produce a relatively small proportion of total electrical power, so why are electricity costs increasing enormously? The reason is price setting mechanism of the merit order principle. Merit order is a system used to define the market price for a set quantity of goods. The name implicates that the order in which demand is divided between producers depends upon some kind of merit. In the electrical energy market, goods are defined by a quantity of electrical energy needed. Producers are represented by different types of energy generation, such as renewables, nuclear, coal, and gas. Since electrical energy is a perfectly homogenous commodity, merit is defined by the cost of producing energy (short run marginal costs). In contrast to conventional electricity generation, renewable energies need no raw material to function, therefore having the lowest production costs. Following renewables is nuclear energy, trailed by coal and finally gas. The final price for the demanded quantity energy is determined by the most expensive energy source needed to fulfil the demand. For example, in a certain region, there is a total demand of 100 MWh. The local power plants can produce energy as follows (Fig. 1). Scenario 1 requires all four power plant types to fulfil 100 MWh, therefore setting the price to the most expensive power plant used, e.g. in case gas costs in this chart increase, all other energy sources will be paid the same price. On the other hand, scenario 2 does not need to utilize the most expensive power plant, since it has increased renewable capacities.

Superconductor Busbar Systems in the Light of Increased Energy Costs

183

Fig. 1 Merit order principle

This price reduction of electricity, due to increased supply of renewable energy, is called the merit order effect. Recently, policy makers in Germany have been implementing key steps in the Energiewende (Germany’s energy transition away from non-renewable fuels). Measures include the abandonment of nuclear energy and a steady decline in coal power plants. Neglecting these two energy types leaves only renewables and gas as power options. Taking the merit order into account, this means that the electricity price is fixed at the marginal cost of gas-powered power plants, i.e., natural gas, if renewables cannot fully cover the electricity demand, which they cannot at the moment. Since January, the price has more than doubled, from around 4 to 9 USD/MMBtu, a 14 year high. The lack of power generation alternatives in Germany shows that the merit order concept needs multiple substitute producers to be effective and that Germany is in a volatile and expensive energy situation, by reducing two prominent energy sectors. Not only Germany or Europe is confronted with the energy shortage. On the long run most countries and regions will feel the new energy situation. Low availability and increased costs on energy are affecting industries with high specific energy demand to a particular extent (Fig. 2). The increase of electricity costs has significant influence on potential investment opportunities for superconductor applications. Investment costs are increasing but at

184

W. Reiser et al.

Fig. 2 Cost of natural gas development in 2022 [1]

a much lower rate than electricity costs. Electrical transmission losses play a more important role now, than during years of low costs a few years ago. Superconductors can help to eliminate a large portion of transmission losses. Every MWh that is not lost in transit does not need to be generated and does not consume natural gas. Transport losses of electricity are directly related to the electrical current. To avoid losses is more important on high current applications. Aluminium plants are operated the highest currents of all industrial processes and in large quantities worldwide. Superconductors will not substitute copper and aluminum conductors in all cases but will bring better efficiencies for high current applications. The following case studies will illustrate conditions where superconductors are beneficial, while also showcasing areas in which superconductors will not improve the present situation.

Potential Use Cases in the Aluminum Industry Superconductor busbars have reached industrial readiness. Superconductors are conducting direct current with extremely high densities of more than 50 kA/cm2 with zero losses. Due to their low space requirements, high current carrying capacity, and utmost efficiency, superconductor busbars will find use in the aluminum industry. The paper presented at TMS 2021 [2] introduces the technical basis of superconductors and portrays technical and economic advantages. Four out of five different applications are discussed in this paper:

Superconductor Busbar Systems in the Light of Increased Energy Costs

185

Fig. 3 Four use cases for aluminum industry

• • • •

Main busbars between rectifiers and potroom (collector bars) Interconnection busbars between potrooms and standby rectifiers Magnetic field compensation with minimal space and power requirements DC connection between potline and power plant or grid connection point (Fig. 3).

The major portion of operating costs (OPEX) are directly related to the costs of electricity. Repair and maintenance have little impact on the total operating costs and are neglected going forward. In September 2022 the fluctuating stock exchange price for electricity increased to factor 10 against 2020 costs of 50 e/MWh for industry use. For a realistic comparison the average cost increase is estimated at a factor of 2. In the comparison calculation we will use 100 e/MWh for energy costs against 50 e/MWh some years ago. Superconductors carry very high currents below their critical temperature. The current carrying capacity increases with lower temperatures. The increase between 70 and 20 K operating temperature leads to an increase of the superconductor current between factors 3 and 4. In other words: The reduction of the operating temperature from 70 to 20 K reduces the necessary amount of superconductor tapes by the same factor and therefore decreases the investment costs for the superconductor material (Fig. 4). On the other hand, the reduction of superconductor operating temperature leads to higher investment and higher operating costs of the cooling system. The cooling machine efficiency is dramatically reduced. At 70 K the cooling efficiency is about 5%. At 20 K this is reduced to between 1 and 2%. A cooling power of 1 kW at 70 K requires 20 kW electricity whereas the same power at 20 K needs an electrical power between 50 and 100 kW. The reduction of the operating temperature is reasonable as long as the superconductor savings are higher than the total costs for cooling over a given period of years. In the following four calculation spreadsheets cases out of the TMS2021 are used. In addition to the already known results two columns are added with modifications:

186

W. Reiser et al.

Fig. 4 Typical critical current of Theva TPL2100 wire as a function of the temperature. Boiling temperatures at 1 bar of various cooling agents are marked in red

In Modification A the electricity costs including the costs for cooling superconductors are raised from 50 to 100 e/MWh. Reflecting the present increase of project costs CAPEX is assumed to increase by 10–15%. In Modification B technical changes lead to different results: B1 considers the reduction of the operating temperatures, from 70 to 20 K and its effect on the OPEX. As explained earlier this leads to an increased current carrying capacity of the single superconductor tapes or decreases the number of superconductors deployed. B2 reflects the impact on investment costs due to the change in temperature and the resulting change of superconductors required. Lower operating temperatures lead to higher cost of cooling equipment but on the other side lower superconductor costs. The break-even between both opposite effects depends mainly on operating current and length of the connection and is different case by case.

Four Use Cases with Modifications Five superconducting use cases for Aluminium industry have been presented at TMS2021 under the title Superconductor Busbars—High Benefits for Aluminium Plants [2]. Four out of these five cases were explained in detail with tables listing main technical data, CAPEX, OPEX, and Total Cost of Ownership (TCO) over different

Superconductor Busbar Systems in the Light of Increased Energy Costs

187

time periods. In the following reviews, not all lines of [2] are displayed. For the case descriptions and survey on all lines, we refer to Ref. [2]. Two modifications are added to the following tables. Only energy costs are changed in Modification A from 50 to 100 e/MWh. This has an effect on the difference between the OPEX of aluminum bars and the superconducting system which influences payback period and total cost of ownership (TCO). Modification B contains the forementioned increased energy costs plus technical changes. The modifications B influence OPEX by reducing the operating temperature to 20 K and CAPEX due to increased investment in cooling equipment. Looking to the use cases the commercial effects of these modifications are shown. In all cases the use of superconductors is reducing GHG emissions by reducing line losses during electricity transmission. Case 1: Main Busbars Between Rectifiers and Potroom (Collector Bars) The substitution of collector bars at an operating current of 400 kA leads to several positive effects: reduced GHG emissions, reduced space requirements, and reduced OPEX. Higher CAPEX is compensated at 50 e/MWh within 9.6 years. With doubles energy costs double the payback period is cut down. Operating at lower temperatures the positive effect of less superconducting material outweighs increased costs for the cooling equipment. CAPEX in Modification B is decreased by 25%, and the payback period reduces to 3.3 years acceptable to most plants (Tables 1 and 2). Case 2: Interconnection Busbars Between Potrooms and Standby Rectifiers The increase of energy costs shows that the payback period is cut to half, the same result as case 1. In this case technical modification will not improve the situation. It is better to stay at the liquid nitrogen temperature level of 70 K compared to reducing the operating temperature to 20 K. The costs of additional cooling equipment are much greater than the savings generated from less superconductor material. This demonstrates that the 20 K temperature level provides advantages on high operating currents only. Case 3: Magnetic Field Compensation (MFC) In case 3 the operating current is in the same range as case 2. The same logic applies. The total investment in superconducting MFC is paid back in less than 3 years. At today’s energy rates superconducting MFC generate savings every year after the 3-year payback period. Space savings—superconductor busbars are much smaller than conventional ones—and safety features—superconductor busbars are entirely enclosed and personnel safe—are additional advantages of superconducting MFC (Table 3). Case 4: DC Connection A superconductor DC connection on the voltage level of the potline avoids multiple energy conversion steps which will reduce GHG emission due to reduced losses.

188

W. Reiser et al.

Table 1 Case 1: results on modifications at collector bars Case 1: collector bars Al busbar Nominal current

SC busbar

Modification A

Modification B

Energy costs—no techn. changes

Energy costs + techn. changes

Al busbar

SC busbar

SC busbar

400

kA

Overall length 415 Busbar power losses

3557

Electr. power SC cooling

m 120

3557

820 17,000

7000

120

120

kW

820

870

B1

kW

19,000

14,300

B2

ke

100

A1

CAPEX

6200

Energy costs

50

Operating energy

30,700

8100

30,700

8100

8550

MWh/year

OPEX

1535

405

3070

810

855

ke/year

Payback period

Base

9.6

Base

5.3

3.3

Years

ke

100

e/MWh

Total cost of ownership @ 10 years

21,550

21,050

37,700

27,100

22,850

@ 25 years

44,575

27,125

83,750

39,250

35,674

ke

@ 40 years

67,600

33,200

129,800

51,400

48,498

ke

Additional environmental issues may also give reasons for the installation of a superconductor DC connection. With present energy costs both modifications show the same payback period of below two years. The payback periods will change depending on connection lengths (Table 4).

Advantages of a Combined Pipeline for Transfer of Liquid Hydrogen and Superconductor Electricity Intercontinental and longdistance transport of hydrogen by ship or by pipeline requires the conversion of gaseous hydrogen (GH2 ) into liquid hydrogen (LH2 ). The high amount of energy needed for liquefaction is often seen as a disadvantage, but it turns out to be advantageous and highly efficient in the overall system: abundant renewable energy at the point of origin generates and cools the hydrogen into its liquid state. The energy and low temperature of the liquid hydrogen can then be utilized at the destination where energy is not as readily available, e.g. air conditioning or plant cooling applications.

Superconductor Busbar Systems in the Light of Increased Energy Costs

189

Table 2 Case 2: results on interconnection busbars Case 2: interconnection busbars Al busbar

SC busbar

Modification A

Modification B

Energy costs—no techn. changes

Energy costs + techn. changes

Al busbar

SC busbar

SC busbar

Nominal current 50

kA

Overall length for 2 phases

1000

m

Busbar power losses

1071

Electr. power SC cooling

15

1071

141 5680

2150

15

15

kW

141

280

B1

kW

6500

7530

B2

ke

100

A1

e/MWh

CAPEX

1920

Energy costs

50

Operating energy

9300

1300

9300

1300

2550

MWh/year

OPEX

465

65

930

130

255

ke/year

Payback period

Base

9.4

Base

5.4

8.0

Years

ke

100

Total cost of ownership @ 10 years

6570

6330

11,450

7800

10,080

@ 25 years

13,545

7305

25,400

9750

13,905

ke

@ 40 years

20,520

8280

39,350

11,700

17,730

ke

Furthermore, achieving low temperatures is not only helpful for the pipeline transport of hydrogen to high-demand regions. The cryogenic state of LH2 can be used in combination with superconductors to aid in the energy transition. This allows the necessary increased electricity transport between regions, the reduction of space usage for combined LH2 -pipelines and electric lines and the availability of a suitable fuel for the decarbonization of all different kinds of applications. Synergies arise with usage of LH2 as an energy carrier. Together with superconductors LH2 can facilitate long-range electricity transport with highest efficiency and neglectable losses. LH2 produced with renewable wind and solar energy is the green substitute for natural gas and the only long-term solution for energy intensive industry. LH2 is not only an energy carrier but the base commodity for a number of industrial processes. Chemical industry will use LH2 as the GHG-free base for chemical products, which are based on natural gas and oil today. Direct reduction of iron ore with hydrogen will turn the steel production into a green industry. LH2 can be used in heat processes, not only in secondary aluminum industry but in all melting and heating processes. The best method to transport hydrogen is as a liquid on a temperature level of 20 K in a cryostat pipeline. This dramatically reduces safety requirements. Due to

190

W. Reiser et al.

Table 3 Case 3: results on magnetic field compensation (MFC) Case 3: MFC

Al busbar

SC busbar

Modification A

Modification B

Energy costs—no techn. changes

Energy costs + techn. changes

Al busbar

SC busbar

SC busbar

Nominal current

40

kA

Overall length for 2 phases

500

m

Busbar power losses

429

Electr. power SC cooling

8

429

43 2200

1400

8

8

kW

43

113

B1

kW

2450

4150

B2

ke

100

A1

e/MWh

CAPEX

1250

Energy costs

50

Operating energy

3700

400

3700

400

1050

MWh/year

OPEX

185

20

370

40

105

ke/year

Payback period

Base

5.8

Base

3.2

10.4

Years

100

Total cost of ownership @ 10 years

3100

2400

5100

2850

5200

ke

@ 25 years

5875

2700

10,650

3450

6775

ke

@ 40 years

8650

3000

16,200

4050

8350

ke

the high energy density of LH2 the pipeline size is small, and the pipe wall thickness and weight are much lower than for a high-pressure gas pipeline. In addition to the transfer of hydrogen, superconductor cables can be placed inside the LH2 pipeline for a lossless transmission of electricity over long distances. Europe will not be able to fulfill the demand of green hydrogen by itself. Europe already entered into import agreements to import LH2 by ship from remote places like Africa, South America, or even Australia. It would make a lot of sense to transport LH2 from the North Sea or Atlantic ports to the industry centers of Europe together with electricity generated from offshore wind farms. Thus 10 GW hydrogen and 10 GW electricity can be transferred to the industry centers in the center of Germany or even into the center of Europe. Superconductors inside LH2 pipelines are the most favorable option for the transport of electrical power, thanks to lower CAPEX and neglectable OPEX. Presently this system is under demonstration at an R&D project funded by the German Federal Ministry of Research [3].

Superconductor Busbar Systems in the Light of Increased Energy Costs

191

Table 4 Case 4: results on DC connection between potline and power plant/grid connection point Case 4: DC versus AC connection Al busbar Nominal power

Modification A

Modification B

Energy costs—no techn. changes

Energy costs + techn. changes

SC Al busbar busbar

SC busbar

SC busbar

400

MVA

System Transfo/OH SC Transfo/OH SC configuration line/Transfo busbar line/Transfo busbar

SC busbar

Nominal voltage

AC 110

DC 1.25

AC 110

DC 1.25

DC 1.25

No of phases 3

2

3

2

2

Nominal current

2.1

160

2.1

160

2.1

Length of connection

1500

Line power losses

3430

Electr. power SC cooling

kV

kA m

96

3430

696 22,100 20,000

96

96

kW

696

900

B1

kW

24,400

23,700

B2

ke

100

A1

e/MWh

CAPEX

18,000

Energy costs

50

Operating energy

29,600

6800

29,600

6800

8600

MWh/year

OPEX

1480

340

2960

680

860

ke/year

Payback period

Base

3.6

Base

1.9

1.8

Years

100

Total cost of ownership @ 10 years

32,800

25,500 49,600

31,200

32,300

ke

@ 25 years

55,000

30,600 94,000

41,400

45,200

ke

@ 40 years

77,200

35,700 138,400

51,600

58,100

ke

References 1. Trading Economics (2022) Natural gas. https://tradingeconomics.com/commodity/natural-gas 2. Reiser W, Reek T, Räch C, Kreuter D (2021) Superconductor busbars—high benefits for aluminium plants. In: Light metals 2021, the minerals, metals and materials series book series (MMMS). https://link.springer.com/chapter/10.1007/978-3-030-65396-5_52 3. Wasserstoff-Leitprojekt TransHyDE (2022) Project paper—internet link comes till end of 2022

Critical Metals for Clean Energy: Extraction of Rare Earth Elements from Coal Ash Sara Penney and Shafiq Alam

Abstract As the global community strives to reduce greenhouse gas emissions to combat climate change, green technologies are becoming a top priority, in turn increasing the demand for rare earth elements (REEs) as they are crucial in the production of clean energy. This increase in demand has led to the search for alternative sources of REEs, hence the recovery from fly and bottom ash, a waste product from the burning of coal for energy. This paper will discuss the extraction of REEs from the coal ash of SaskPower’s Poplar River Thermal Power Plants. Keywords Mineral processing · Hydrometallurgy · Rare earth elements (REEs) · Clean energy · Coal ash · Process optimization

Introduction In today’s society, rare earth elements (REEs) are increasingly more important. These elements include the 15 lanthanides (lanthanum, cerium, praseodymium, neodymium, promethium, samarium, europium, gadolinium, terbium, dysprosium, holmium, erbium, thulium, ytterbium, and lutetium) as well as yttrium and scandium (Fig. 1). They are classified together due to their unique and similar chemical and physical characteristics such as electronic, optical, and magnetic capabilities [1]. Furthermore, yttrium, neodymium, europium, terbium, and dysprosium are considered to be critical rare earth elements [2] as these are the most widely sought-after elements and have more practical uses than the others. They are also referred to as rare earth metals (REMs) given that all of the elements listed are metals. The increasing importance of REEs is primarily due to their specialized use in high-tech products such as smartphones, electric cars, permanent magnets, and green S. Penney · S. Alam (B) Department of Chemical and Biological Engineering, University of Saskatchewan, 57 Campus Drive, Saskatoon, SK S7N 5A9, Canada e-mail: [email protected] S. Penney BRAYA Renewable Fuels, 1 Refinery Rd, Box 40, Come By Chance, NL A0B 1N0, Canada © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_19

193

194

S. Penney and S. Alam

Fig. 1 Elements referred to as rare earths [1]

energy technologies [3] as currently there are no alternatives to using these metals. These 17 elements can be difficult to extract from source ore due to their high stability and low concentrations [4], this is why it is essential to start recovering them from secondary sources such as coal and fly ash. While the concentrations in these products can be low, they are available as a by-product in vast quantities. Fly and bottom ash are often utilized for other purposes, such as in concrete, still, only approximately 25% of the ash is used [5], and the rest is disposed of as waste. Furthermore, fly ash and bottom ash are by-products that can be environmentally harmful when kept in long-term storage such as tailings ponds, by selling it the volume being stored can be vastly reduced and the elements recovered for use in industry. This paper will deal with the recovery of REEs from coal ash using hydrometallurgical techniques.

Methodology and Results In this process, REEs were leached from fly and bottom ash using different leaching agents. Before leaching, the ash samples collected from a coal-fired power plant were characterized to determine the content of REEs for the viability of extraction. Four samples of coal ash were provided (Fly Ash Unit 1, Fly Ash Unit 2, Bottom Ash Unit 1, Bottom Ash Unit 2), and content was determined using an inductively

Critical Metals for Clean Energy: Extraction of Rare Earth Elements …

195

coupled plasma mass spectrometry (ICP-MS). Overall, the content of REEs was fairly consistent across all of the samples. In this research, four different leaching agents were used to dissolve rare earth metals from coal ash. They are sulfuric acid (H2 SO4 ), hydrochloric acid (HCl), nitric acid (HNO3 ), and sodium hydroxide (NaOH). A 3 g sample was dissolved in 100 mL (30 g/L) of the respective acid/base and stirred constantly for 8 h at room temperature. The leaching agents were also tested at two different molarities (1 and 6 M) to determine which range of molarity works better. Figure 2 shows that the highest recovery is that of the 1 M nitric acid where about 95% overall REEs extraction was achieved. The sulfuric acid and sodium hydroxide did not give any results. It is possible that the leaching was not zero but the ICP was unable to detect amounts less than 0.01 ppm of the individual elements within the samples provided. The lower molarity acids gave better results, and this also supports the claim by Mayfield and Lewis [6]. There are numerous parameters that can be manipulated to achieve optimal leaching of REEs into solution: the molarity (pH), the solid–liquid ratio (S/L ratio), and the temperature. The previous 8-h test results of the 30 g/L 1 M acid already recovered just less than 95% of the initial REE content of the fly ash; therefore, the parameters were modified around these parameters within range to see if there is a better combination. The temperature was kept at room temperature given the already high results at room temperature, and it is not profitable to increase the temperature of the operation. The leaching time was 24 h. To optimize the process, Fly Ash Unit 1 and Bottom Ash Unit 1 were tested with molarity varied at 1, 1.5, 2, 3, and 6. Each molarity was tested at solid–liquid ratios of 30 g/L and 50 g/L as shown in Figs. 3 and 4, respectively. Bottom ash was not tested at 50 g/L due to the negative results of the fly ash. OVERALL 8 HR %RECOVERY 94.89%

100.00% 90.00% 80.00%

81.64% 74.80%

74.91%

70.00% 60.00% 50.00% 40.00% 30.00%

21.39%

20.00% 10.00%

0.00%

0.00% 30g/L 1M HCl 30g/L 6M HCl 30g/L 1M 20C 20C HNO3 20C

Fig. 2 Results of leaching tests

30g/L 6M HNO3 20C

30g/L 1M H2SO4 20C

0.00% 30g/L 6M H2SO4 20C

30g/L 6M NaOH 20C

196

S. Penney and S. Alam

24 Hour Precent Recovery - 30g/L 100

93.5

91.52

94.78

96.83

97.82

% Recovery of REEs

90 80

77.19

78.28

3M HNO3 30g/L

6M HNO3 30g/L

66.11

70 60 50 40 30 20 10 0 1M HNO3 30g/L

1.5M HNO3 30g/L

Fly Ash Unit 1

2M HNO3 30g/L

Bottom Ash Unit 1

Fig. 3 Percent recovery of REEs at various molarity HNO3 and constant solid–liquid ratio of 30 g/L

24 Hour Precent Recovery - 50g/L 100

% Recovery of REEs

90 80 70

78.43 69.18

75.37

60 50 40 30 20 10 0 1M HNO3 50g/L

3M HNO3 50g/L

6M HNO3 50g/L

Fly Ash Unit 1

Fig. 4 Percent recovery of REEs at various molarity of HNO3 and constant solid–liquid ratio of 50 g/L

As shown in Fig. 3, overall the best leaching conditions for the fly ash were the 2 M HNO3 at 30 g/L where 98% recovery was achieved. The bottom ash achieved optimal recovery at a slightly lower molarity of 1.5, which is 97%. For the kinetic test, the ash sample was leached in nitric acid and stirred for 36 h with periodic sampling to determine the percent recovery. Figure 5 shows the results of kinetic studies, where it was found that the equilibrium for REE extraction was reached after 12 h.

Critical Metals for Clean Energy: Extraction of Rare Earth Elements …

197

Determination of Equilibrium Time

% Recovery of REEs

120.00% 100.00% 80.00% 60.00% 40.00% 20.00% 0.00% 3

5

8

12

24

36

Time (Hours)

Fig. 5 Kinetic test

Conclusion The extraction of REEs from coal ash by-products has been carried out in batchwise leaching tests. Operating parameters were optimized to get the best leaching conditions. It was found that 98% of the REE content of the ash was recovered by leaching with 2 M nitric acid.

References 1. Awais M (2016) What are rare earth elements, REEs? Retrieved from https://etn-demeter.eu/ what-are-rare-earth-elements-rees/ 2. Kalvig P (2014) Forecasting future demand and supply of rare earth elements (REE) with some focus on issues related to recycling. Presentation 3. Binnemans K, Pontikes Y, Jones P, Van Gerven T, Blanpain B (2013) Recovery of rare earths from industrial waste residues: a concise review 4. United States Department of Energy (2017) Report on rare earth elements from coal and coal byproducts 5. Franus W, Wiatros-Motyka M, Wdowin M (2015) Coal fly ash as a resource for rare earth elements. Environ Sci Pollut Res 22(12):9464–9474. https://doi.org/10.1007/s11356-0154111-9 6. Mayfield D, Lewis A (2013) Environmental review of coal ash as a resource for rare earth and strategic elements. In: World of coal ash (WOCA) conference

Part IV

Energy Technologies

Investigation of Slag and Condensate from the Charge Top in a FeSi75 Furnace M. B. Folstad, K. F. Jusnes, and M. Tangstad

Abstract Metallurgical silicon/ferrosilicon is produced industrially in submerged arc furnaces by carbothermic reduction of quartz. In addition to raw materials, oxide impurities are present in the furnace. Accumulated slag is typically found along the furnace walls towards the charge top, as well as the furnace bottom and taphole. During optimal tapping, the slag will follow the alloy. Accumulated slag in the furnace will affect the furnace operation. Large amounts of accumulated slag will have a negative effect on the operation and lead to more CO2 emissions. A partly melted charge at the surface from a FeSi75 furnace has been collected and analyzed. All samples contained slag and/or condensate, carbon, and/or unreacted SiO2 . The slag samples contained SiO2 , in addition to oxides of Fe, Na, K, Mg, Al, and Ca. Slag at the charge surface may be due to sufficiently high temperature to produce slag, which again can affect how the materials distribute in the furnace. Keywords Si production · Charge surface · Slag · Condensation · CO2 emission

Introduction Metallurgical silicon and ferrosilicon are produced industrially by carbothermic reduction. The process takes place inside a submerged arc furnace with a diameter varying from 5 to 11 m and a height from 3 to 5 m. Quartz (SiO2 ), carbon materials, and possibly iron sources are added as raw materials. Reaction 1 gives the overall mass balance, but the actual process is more complex. SiO2 + 2C = Si + 2COg

(1)

M. B. Folstad (B) · M. Tangstad Norwegian University of Science and Technology, Trondheim, Norway e-mail: [email protected] K. F. Jusnes Finnfjord AS, Finnsnes, Norway © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_20

201

202

M. B. Folstad et al.

The raw materials are added at the top of the furnace and will experience a steep temperature increase as it meets a charge surface that holds a temperature around 1300 °C [1, 2]. The Si and FeSi furnaces are divided into a low temperature zone that includes the higher parts of the furnace and along the furnace walls, and a high temperature zone in the lower parts of the furnace. In the furnace, the raw materials are heated by electricity through three carbon electrodes. There are also several oxides present in the furnace, slag. These originate from the impurities in the raw materials and from limestone, which many companies add together with the raw materials. The amount of accumulated slag and condensate along the furnace walls up to the charge top is an important factor in the productivity of the furnace, and large variations between different furnaces are found during several furnace excavations over the last fifteen years. Figure 1 illustrates three different situations with various amounts of slag and condensate in an operating Si furnace. Situation a has almost no slag present along the furnace wall and has a wide route for the raw materials to descend and react down through the furnace. Situation b is more average, and the condensate and slag will affect the furnace operation. Slag that is pushed up in the furnace will start to solidify as the temperature decreases in the low temperature zone. A too thick layer of slag in the charge will give a narrower reaction route and hinder the raw materials from descending and reacting further down in the furnace. Situation c illustrates an accumulated situation; there is almost no room for reacting materials to descend. An accumulated furnace produces less Si/FeSi and more gas escapes in the off-gas system, which again gives an increased CO2 emission. The slag present in the lower parts of the Si/FeSi furnace mainly consists of the SiO2 –CaO–Al2 O3 system. In the charge it is expected to find unreacted charge, condensate/slag, and SiC. The borderline between slag and SiO2 is in this study set to 90% SiO2 . Higher content of SiO2 in the slag in the higher parts of the furnace is expected. The reduction rate of SiO2 from slag is strongly dependent on temperature [3, 4]. Increasing temperature increases the SiO2 reduction rate. The SiO2 content affects several properties of the slag, for instance the viscosity, and the slag viscosity has a great impact on how the materials move through the furnace. A lower viscosity

Fig. 1 Schematic figures illustrating the interior of an operating Si furnace. The left situation a shows an ideal situation with almost no condensate/slag along the furnace walls. Situation b illustrates an average situation with condensate/slag that affects the furnace operation. The last situation c shows an accumulated furnace

Investigation of Slag and Condensate from the Charge Top in a FeSi75 …

203

ensures a good flow and good drainage of the furnace. Condensates found in the charge are generated from the gas mixture of mainly SiO and CO according to reactions 2 and 3 [5, 6]. SiO and CO gas are created in the high temperature zone and ascend to the low temperature zone. 2SiOg → Sis,l + SiO2s,l

(2)

3SiOg + COg → 2SiO2s,l + SiCs

(3)

Condensates will stick firmly to any surface and create a crust. The crust in the charge area is referred to as the stoking crust. Two types of condensate are found through industrial and pilot scale excavations of Si and FeSi [7–9]: white condensate from reaction 2 and brown condensate from reaction 1. Broggi [10] characterized the product of reaction 1 and found that the brown condensate consists of a mixture of Si spheres embedded in a SiO2 matrix and that the compound generates in the temperature range from 1400 to 1780 °C. The white condensate consists of a SiO2 matrix with SiC embedded in them. Identification of materials at the charge surface is valuable to help explain the furnace operation. When the operation is in good condition, only heated raw materials will be present at the charge top. When deviation from this situation occur, the materials can give knowledge on reactions in the furnace. The distribution of raw materials in the charge has a great impact on the furnace operation. Sampling during operation (or short operational stops) can be related to the furnace condition during operation. Samples collected during operation are also more economic, efficient and exact compared to samples taken during excavations.

Experimental Samples from two different FeSi75 furnaces, O and R, were collected from the charge surface. The furnaces have a maximum power of 42 MW and 45 MW, respectively, and the sampling was done in a period of 6 months. The samples were dragged out of the furnace with a stoking car by furnace operators when the furnace operation was reduced or during a planned maintenance stop. An overview of the samples is listed in Table 1. Some of the samples have also been used in another study by Jusnes et al. [11]. Most of the samples were large and heterogenous, as can be seen in Fig. 2, and were divided into several pieces. Different areas were chosen and in some cases named based on visual observations. All samples were analyzed using an electron probe microanalyzer (EPMA), and the slag composition was found using wavelength dispersive spectrometers (WDS).

204

M. B. Folstad et al.

Table 1 Samples collected from the charge top. Some of the samples, CPR1809a, CPR1809b, CPR1809c, and CPR0211 are also used in another study [11] Sample CPR1809a CPR1809b

Comment Same time of sampling, different location

Charge top Viscous matter between two electrodes

CPR1809c

0.5 m from crust edge

CPR0211

Viscous matter

CPO1101

Large sample. Viscous and hard charge surface. Divided into S (slag), BK (brown condensate), M (metallic), K (quartz), P (porous), U (unknown)

GR1602

Viscous matter, glued together

CPR2502

Large sample. Divided into k (quartz), m (metallic), y (edge), a, b

CPR1903

Large areas with brown condensate

Fig. 2 Sample CPR1809a, b, c and CPO1101, S, and BK. The charge surface samples were heterogenous, and several samples were analyzed based on visual differences

Results and Discussion Partly melted samples from the charge surface are often heterogenous and mainly contain raw materials SiO2 and C, condensate, slag, and some SiC. Some of the samples were collected during troubling conditions. The charge surface materials are in those situations often sticky and viscous and appear as “glued” together. In

Investigation of Slag and Condensate from the Charge Top in a FeSi75 …

205

Fig. 3 Sample GR1602, which shows partly melted SiO2 . In the sample are also some FeSi metal droplets in the SiO2 matrix. A thin red FeO layer is covering the sample

Fig. 3 it can be observed partly melted SiO2 together with droplets of FeSi and a thin layer of FeO. The charge surface is expected to have a temperature of around 1300 °C [1, 2], but the temperature increases rapidly as the materials descend inside the furnace. SiO2 melts around 1700 °C, and with a higher temperature at the top of the charge the SiO2 will melt higher up in the furnace. It is preferable that SiO2 melts as deep in the furnace as possible. If the SiO2 starts to soften at a higher position it might cause a more compact charge and the gas permeability decreases. A sufficient good permeability is necessary to capture ascending SiO and CO gas to produce SiC and SiO2 according to reactions 1 and 2. A compact charge increases the SiO gas loss and hence increases the amount of CO gas. More SiO gas in the off-gas system will give an increased amount of microsilica.

Brown and White Condensate from Charge Surface In most samples, areas of SiO2 covered with a thin layer of brown condensate were found. Picture and EPMA image of the typical condensate can be seen in Fig. 4. The condensate consists of a SiO2 matrix with small Si spheres, which is the same as Broggi [10] and Jusnes et al. [11] found in their samples. Materials from the charge area have earlier been analyzed during excavations [7–9, 12]. Glassy condensates with white, brown, and in some cases green colors are often found covering and holding together partly melted raw materials. The presence of Si on EPMA images has confirmed condensation reaction 1. White and green condensates are often mostly

206

M. B. Folstad et al.

Fig. 4 Sample CPR1809b, a typical brown condensate. The brown condensation layer is Si spheres embedded in a SiO2 matrix, shown in the right EPMA image

SiO2 mixed with SiC. Most of the condensate in this study was brown, but some areas included some white condensate, as shown in Fig. 5. Brown and white stripes can be observed in the left picture. The different colors are also seen as different zones in the right EPMA image. Areas with Si spheres embedded in the SiO2 matrix are from the brown-colored area, while the white area is SiO2 and SiC from condensation reaction 2. From equilibrium calculations and temperature measurements, Broggi [10] found that Si–SiO2 condensate develops from 1400 to 1780 °C. The existence of brown condensate and FeSi metal in the charge surface might indicate that the temperature at the charge surface has been sufficiently high to run the condensation reaction from SiO gas and to produce metal. A too high temperature on the charge surface will have a negative consequence on the furnace operation. The reaction zones in the furnace might change and again decrease the temperature in the high temperature zone in the furnace. A temperature above 1811 °C in addition to a SiO pressure of a minimum of 0.7 bar is necessary to produce Si metal according to reaction 4. SiC + SiOg = 2Si + COg

(4)

Initial Slag Formation In Si and FeSi furnaces it is expected to find accumulated oxide impurities in form of slag. Raw materials are used that contain various amounts of impurities. In addition, some furnaces add calcium in form of limestone to the process. As the temperature increases, most of the impurity oxides will be reduced from the SiO2 . During normal conditions with a charge top temperature around 1300 °C, it is expected melting of impurities in the raw materials and initial slag formation. In this study, the borderline

Investigation of Slag and Condensate from the Charge Top in a FeSi75 …

207

Fig. 5 Sample CPO1101_P2, brown and white condensate. The right image is the EPMA image of the same sample. The brown areas are SiO2 + Si, while the white areas are SiO2 + SiC

between slag and SiO2 is defined as 90wt% SiO2 . Initial slag formation from SiO2 can be seen in Figs. 6 and 7. From the EPMA image in Fig. 6, it can be observed a white slag phase accumulated at the grain boundaries and in the cracks in the SiO2 , in addition to a bright grey slag phase next to the melted SiO2 . A SiO2 -rich slag can also be seen as the brighter grey-colored phase in the SiO2 matrix in Fig. 7. Table 2 lists the composition given in weight percent for the slags found in this study. The slag in CPO1101_P.1 is 79 wt% SiO2 and appears homogeneous. Earlier studies have found that impurities may lower the melting temperature [13, 14]. Contaminants may break up the SiO2 network system and create bindings between dislocated cores. The SiO2 -rich slag is therefore most probably the initial formation of slag from SiO2 areas with increased content of impurities. The slag in the charge surface mostly contains SiO2 in addition to FeO, Al2 O3 , and traces of alkali oxides Na, K, Mn, Mg, and Ca, which are common impurities in the SiO2 . The existence of FeO in the slag is expected since iron oxide pellets are added together with the raw materials. As the temperature increases, the alkali oxides and the FeO will be reduced from the slag. The slag in the lower parts of the furnace mainly consists of the SiO2 –CaO–Al2 O3 system. As increasing temperature increases the reduction rate of SiO2 from SiO2 –CaO–Al2 O3 slag it is expected a higher amount of SiO2 in the slag phase in the higher, low temperature parts of the furnace. By summarizing the basic oxides and fitting the slag system into the SiO2 –FeO–Al2 O3 or SiO2 –CaO–Al2 O3 system the liquidus temperatures can be found, given in the right column in Table 2. The liquidus temperature is the temperature at which the slag system is all liquid and gives an indication of the charge surface temperature. It can be noticed that the slags in the grain boundaries/cracks in the SiO2 have a liquidus temperature around the expected charge surface temperature. An increased SiO2 content to >80 wt% increases the liquidus temperature considerably to >1600 °C,

208

M. B. Folstad et al.

Fig. 6 Initial slag formation in sample CPR1809a.2. Slag accumulates in the grain boundaries and cracks in the SiO2 , shown as the white phase. An area with slag is also found together with the melted SiO2 , shown as brighter grey color

Fig. 7 Initial slag formation in sample CPO1101_P.1. The slag is the brighter grey-colored phase in the darker grey SiO2

which is higher than the charge temperature around 1300 °C. The existence of this slag composition means again either that the charge surface temperature is sufficiently high to produce this slag composition, or the slag has been pushed upwards from deeper in the furnace. Softened/melted SiO2 and SiO2 rich slag around its liquidus temperature are viscous, which gives a sticky and compact charge surface. Several of the samples were reported as viscous during sampling.

51

83

43

CPR1809a.2.2 (in between SiO2 )

CPR1809c.1.1

CPR1809c.1.2 (in between SiO2 )

81

81

86

CPR2502b

CPR2502a1

CPR2502a2

0

0

0

0

1.7

0

2.7

1.9

1.7

2.3

0.5

1.0

0.8

0.5

0.2

0.4

0.2

0.1

0.1

0.3

Na2 O

0

0

0

0

5.8

9.3

3.8

2.9

6.0

7.3

2.2

2.4

2.2

1.7

0.3

0.9

0.5

0.4

0.1

0.4

K2 O

0

0

0

0

0

0

0

0

0

0

0.1

0

0.1

0

0

0

0.1

0

0.1

0

MnO

Compositions are given in wt% a Analyses are only valid to know which oxides are in the sample, not the amount

55

CPO1101.M2

CPR2502m

80.9

CPO1101.M1

79.1

85.3

CPO1101.S2

83.2

86.2

CPO1101.S1

CPO1101.U

80

CPR0211.3.2 (in between SiO2 )

CPO1101.P1

72

50

CPR0211.3.1

87

84

CPR1809a.2.1

73

54

CPR1809a.1.2 (in between SiO2 )

CPR1809c.2.2 (in between SiO2 )

64

CPR1809a.1.1

CPR1809c.2.1

SiO2

Sample

0

0

0

2

1.5

0

3.7

2.5

6.0

1.6

3.2

2.0

1.7

0.7

0.1

0.2

1.1

0.3

0.7

0.4

MgO

Table 2 Composition of the main slag phases found in the charge samples

5

6

6

11

1.9

0

1.0

2.3

0

1.2

1.4

1.2

1.6

0.6

2.4

0.7

5.1

1.6

2.9

1.7

CaO

0

0

0

0

2.0

0.2

0.4

0.1

0

5.5

42

21

15

5.5

43

10.3

24

8

24

11.7

FeO

9

13

13

32

3.5

11.4

0.4

1.7

0

2.6

0.6

0.4

6.1

3.4

9.6

4

18

5.7

18

21.1

Al2 O3

0

0

0

0

0.4

0

SO3 :15.0

SO3 :2.5

0

0

0

0

0

0.1

0.4

0.2

0.2

0.1

0.6

0.4

TiO2

0

0

0

0

0

0

0

0

0

0

0

0

0.1

0

0

0.1

0.3

0

0.1

0

BaO

100

100

100

100

100

100

107a

100

100

100

100

100

100

100

100

100

100

100

100

100

Total

1550

1480

1480

1570

1650

1500

1700

1660

1560

1200

1650

1620

1250

1640

1310

1610

1350

1490

Tliquidus

Investigation of Slag and Condensate from the Charge Top in a FeSi75 … 209

210

M. B. Folstad et al.

Conclusion Samples from two different FeSi75 furnaces were collected from the charge surface and analyzed. It was found that these mostly contain partly melted raw materials SiO2 and C, in addition to condensate, slag and some FeSi and SiC. The results can be summarized as follows: • FeSi metal droplets were found in several charge surface samples • Condensation found at the charge surface is most often brown-colored, but some samples also had white condensate. Brown condensate was found to be Si spheres embedded in a SiO2 matrix. In the samples with white condensate, SiO2 was mixed with SiC. • Slag formation at the charge surface initiates at the SiO2 grain boundaries and inside cracks in the SiO2 . The slag contains mainly SiO2 (~50–90 wt%), FeO, Al2 O3 , and in some cases CaO or K2 O. Most of the slags contained traces of alkali oxides Na, Mn, K, Mg, Ca, and TiO2 . One sample also had SO3 . • The slag in the grain boundaries and cracks in the SiO2 had a lower wt% of SiO2 and a lower liquidus temperature than the slag next to melted SiO2 . • The existence of FeSi metal, brown condensate, and slag with liquidus temperature >1600 °C indicates that the temperature has been higher than expected at the charge surface, or that material has been pushed upwards from the high temperature zone. Acknowledgements This paper is published with the permission of Finnfjord AS. The authors thank Finnfjord AS and especially the tapping operators who collected the samples, the Norwegian Ferroalloy Research Association (FFF), and the Norwegian Research Council (NRC). We are grateful for their financial support through the KBN project Controlled Tapping, project no. 267621.

References 1. Johansen ST, Tveit H, Grådahl S, Valderhaug A, Byberg JÅ (1998) Environmental aspects of ferro-silicon furnace operations—an investigation of waste gas dynamics. Presented at the INFACON VIII, Beijing 2. Ksiazek M, Grådahl S, Rotevant EA, Wittgens B (2016) Capturing and condensation of SiO gas from industrial Si furnace. In: Advances in molten slags, fluxes, and salts: proceedings of the 10th international conference on molten slags, fluxes, and salts. Wiley, pp 1153–1160 3. Fulton J, Chipman J (1959) Kinetic factors in the reduction of silica from blast-furnace type slags. Trans Am Inst Min Metall Eng 215:888–891 4. Sun H, Mori K, Pehlke RD, Reduction rate of SiO2 in slag by carbon-saturated iron. Metall Trans B, 8 5. Vangskåsen J, Metal-producing mechanisms in the carbothermic silicon process. Master thesis. Norwegian University of Science and Technology, Trondheim 6. Li F (2017) SiC production using SiO2 and C agglomerates. Ph.D. thesis. Norwegian University of Science and Technology, Trondheim. Accessed: 13 Sept 2022 [Online]. Available: https:// ntnuopen.ntnu.no/ntnu-xmlui/handle/11250/2465202

Investigation of Slag and Condensate from the Charge Top in a FeSi75 …

211

7. Tangstad M, Ksiazek M, Andersen JE (2014) Zones and materials in the Si furnace. In: Silicon for the chemical and solar industry XII 8. Tranell G, Andersson M, Ringdalen E, Ostrovski O, Steinmo JJ (2010) Reaction zones in a FeSi75 furnace—results from an industrial excavation, p 8 9. Ksiazek M, Tangstad M, Ringdalen E (2013) Five furnaces five different stories. In: Silicon for the chemical and solar industry XIII, pp 33–42 10. Broggi A, Tangstad M, Ringdalen E (2019) Characterization of a Si–SiO2 mixture generated from SiO(g) and CO(g). Metall Mater Trans B 50(6):2667–2680. https://doi.org/10.1007/s11 663-019-01678-x 11. Jusnes KF, Hjelmseth R, Folstad MB, Ditlefsen NS, Tangstad M (2021) Investigation of slag compositions and possible relation to furnace operation of a FeSi75 furnace. Presented at the INFACON, Trondheim 12. Ksiazek M, Tangstad M, Ringdalen E (2018) The rapid Si-furnace excavation—an unique chance to investigate the interior of the furnace. Presented at the INFACON XV, Cape Town 13. Ringdalen E, Tangstad M (2016) Softening and melting of SiO2 , an important parameter for reactions with quartz in Si production. In: Advances in molten slags, fluxes, and salts: proceedings of the 10th international conference on molten slags, fluxes and salts 2016, Cham, pp 43–51. https://doi.org/10.1007/978-3-319-48769-4_4 14. Ringdalen E, Tveit H, Bao S, Nordnes E (2019) Melting properties of quartz and their effect on industrial Si and FeSi production. Presented at the non-ferrous metals and minerals, Krasnoyarsk

Lithium Extraction from Natural Resources to Meet the High Demand in EV and Energy Storage Valan Namq and Shafiq Alam

Abstract Electrification of vehicles will increase lithium demand drastically in the next decades, as the demand increases the supply remains constant. Lithium is mainly produced from hard rock spodumene mining and high concentration brines in South America but most of the lithium reserves are in the low lithium concentration continental brine. The exploitation of this reserve would make the industry easily meet the demand. But current technology has limitations to extract lithium from lowconcentration brine. This paper describes most of the applicable methods alongside with mass balance sheet for concentrating lithium from the brine into lithium chloride solution, then purifying and crystallization of lithium solution to lithium carbonate salt. Keywords Lithium · Energy storage · Clean energy · EV · Sustainability · GHG · NetZero

Introduction Technology advancement and the growing demand for renewable energy resources have increased recently. The need for materials that are used in renewable resources has risen significantly. Among these materials, lithium is one important element that is used in lithium-ion batteries, ceramics, lubricants, metal additives, and many more applications. A lithium-ion (Li-ion) battery is an advanced battery technology that uses lithium ions as a key component of its electrochemical cell. This is one of the most popular forms of energy storage in the world, accounting for 85.6% of deployed energy storage systems in 2015 [1].

V. Namq · S. Alam (B) Department of Chemical and Biological Engineering, University of Saskatchewan, 57 Campus Drive, Saskatoon, SK S7N 5A9, Canada e-mail: [email protected] V. Namq Prairie Lithium Inc., 2010—11th Ave, Regina, SK S4P 0J3, Canada © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_21

213

214

V. Namq and S. Alam

Li-ion batteries have several advantages over other high-quality rechargeable batteries, such as nickel–cadmium (Ni–Cd) or nickel–metal-hydride (Ni–MH), because of their highest energy densities. Today’s Li-ion batteries have an estimated energy density of 250 Wh/kg [2]. Figure 1 shows the specific energy density and volumetric energy density of various battery types. In addition, Li-ion battery cells can deliver up to 3.6 V, 3 times higher than that of Ni–Cd or Ni–MH batteries [3]. Therefore, Li-ion batteries have replaced Ni–Cd batteries as the market leader in portable electronic devices, such as smartphones and laptops. From a clean energy perspective, Li-ion batteries are used in electric vehicles (EVs). Currently, among others, Nissan Leaf and the Tesla Model S both use Li-ion batteries as their primary fuel source. The demand for Li-ion batteries is projected to increase tenfold from 2020 to 2030, because of the growing demand for EVs. The electric vehicle batteries accounted for 34% of lithium demand in 2020 which translates to 0.4 Metric tons (Mt) of lithium carbonate equivalents (LCE), which is forecasted to increase to 75% in 2030 based on a projection from Bloomberg New Energy Finance that suggests nearly 40 million EV by 2030 [4, 5]. As a result, total lithium demand will increase significantly in near future. An estimated amount of around 160 g of Li metal is required per kWh of battery power, which equals about 850 g of LCE in a battery for one kWh of power [6]. This means a typical EV (with around 50 kWh battery capacity) will require around 40 kg of LCE. Therefore, a 2000 GWh battery demand by 2030 (based on 40 million vehicles at 50 kWh/vehicle) would imply a demand of 1.7 Mt LCE, to which 0.3 Mt of other demand must be added, yielding a total of 2 Mt LCE demand by 2030, which is a five-fold increase from 2020 levels. Fig. 1 Specific energy density and volumetric energy density of different battery types [2]

Lithium Extraction from Natural Resources to Meet the High Demand …

215

Different forms of lithium are sold in the market, such as lithium carbonate (Li2 CO3 ), lithium hydroxide (LiOH), lithium concentrate, lithium metal, lithium chloride, butyllithium, and others. Lithium carbonate is the most popular compound for many applications. Total worldwide lithium production in 2020 was 82,000 Mt or 436,000 Mt of LCE [7]. Lithium is produced from brine or hard rock ore (spodumene). Around 40% of global lithium carbonate production is from brine-based resources, which are enhanced by ore-based production, both of which are expanding rapidly. Australia and Chile dominate today’s mining, but new mines are being developed in many countries across the world. Argentina, Bolivia, and Chile account for most of the global production of lithium from brine resources (salars). In Nevada, American Lithium and Noram Ventures are among a number of companies exploring the feasibility of extracting lithium from clay. Energy consumption to produce lithium from ore is considerably higher than for brine-based production. According to Kelly et al. [8], brine-based production releases about 5 tons of CO2 to the environment to produce one ton of lithium hydroxide, whereas hard rock-based production is nearly three times as CO2 intensive as brinebased production. They also concluded that lithium hydroxide products made from spodumene resources are almost seven times more CO2 intense than the production of lithium carbonate from brine, which is because of the energy required for mining and processing.

Lithium Extraction from Natural Resources In the hard rock mining process, ore such as spodumene is processed by crushing/grinding, calcination, and roasting followed by acid or alkaline leaching, solution purification, and precipitation to get lithium salt [9]. However, as mentioned earlier, lithium production from ores is energy intensive and this process releases more greenhouse gases (GHG). Conventionally lithium production from brine includes solar evaporation followed by contaminants precipitation by increasing pH, then polishing process using solvent extraction or ion exchange process, and finally lithium salt precipitation and drying. This method is most successful economically in high lithium concentration brines; however, low-concentration lithium brines are challenging in terms of lithium extraction as these brines contain a significant amount of other metals that are extracted with lithium such as sodium, magnesium, calcium, and potassium, and they are difficult to separate from lithium salt. Various approaches have been investigated to extract lithium from brine with mixed metal ions; however, lithium extraction from brine with magnesium content is very difficult because they have similar ionic characteristics in their solution. Lab results showed that brine with lithium to magnesium ratio of 6 is possible to separate using membrane; however, divalent magnesium ions plug the membrane pores [10]. Therefore, the need for an efficient extraction approach to extract lithium from low lithium concentration

216

V. Namq and S. Alam

brines is an ongoing effort. In this work, a model has been developed for concentrating lithium from low-concentration brine, which is presented in the following sections.

Lithium Extraction from Brine The process of lithium extraction from brine consists of two main stages which are extraction of lithium from brine and concentration of lithium in solution, and purification of concentrated lithium-containing solution and producing lithium salt either as lithium carbonate or lithium hydroxide. Since brine contains a low concentration of lithium, a model has been developed to simulate concentrating lithium from the brine into hydrochloric acid (HCl) solution by a factor of 100. The initial lithium concentration in brine is assumed to be 95 ppm. As shown in Fig. 2, brine was passed through an ion exchange column packed with permanganate-based resin (HMnO4 ) where lithium was adsorbed replacing proton (H+ ) on the resin. After adsorption, lithium was eluted as LiCl in concentrated form from the loaded resin (LiMnO4 ) by passing concentrated HCl through the column. This process also regenerated resign back to HMnO4 form taking H+ from HCl. To develop a process flowsheet, mass balance calculation has been done using elemental composition. The produced lithium chloride solution undergoes different processes to purify and crystallize as a lithium salt. Since lithium chloride is acidic, as shown in Fig. 3, the concentrated lithium chloride from Fig. 2 is adjusted for pH using sodium hydroxide (NaOH) where lithium chloride is converted to lithium hydroxide (LiOH). During pH adjustment, metal contaminants such as boron, silicon, calcium, and magnesium are precipitated and the purified lithium hydroxide solution is separated by decantation. This high-purity lithium hydroxide stream enters the carbonation process where sodium carbonate (Na2 CO3 ) is added to the solution to precipitate lithium as lithium carbonate (Li2 CO3 ). Then, the precipitate is separated from the liquid and sent to the final stage for drying and crystallization using a rotary drum dryer.

Conclusion Lithium is critical for energy storage, which is commonly used in rechargeable batteries for laptops, cellular phones, and electric vehicles (EV) as well as in ceramics and glass. A model of sustainable process flowsheet has been developed using hydrometallurgical techniques to produce lithium as Li2 CO3 salt from the low concentrated brine solution. Production of lithium from brine-based resources is an economical and energy-saving process. It also reduces the greenhouse gasses that can help meet the target of NetZero.

217

Fig. 2 Separating and concentrating lithium from brine-based resources

Lithium Extraction from Natural Resources to Meet the High Demand …

Fig. 3 Purification of lithium chloride solution and conversion to lithium carbonate

218 V. Namq and S. Alam

Lithium Extraction from Natural Resources to Meet the High Demand …

219

References 1. Assad MEH, Khosravi A, Malekan M, Rosen MA, Nazari MA (2021) Energy storage. In: Design and performance optimization of renewable energy systems, pp 205–219 (Chapter 14) 2. Houser K (2021) The world’s first sodium-ion battery for EVs has arrived. www.freethink. com/technology/sodium-ion-battery 3. The Clean Energy Institute (CEI) at the University of Washington. https://www.cei.washin gton.edu/education/science-of-solar/battery-technology/. Accessed on 12 Sept 2022 4. BNEF (Bloomberg New Energy Finance) (2020) Long-term electric vehicle outlook 2020. www.bnef.com. BNEF (2021) Long-term electric vehicle outlook 2021. www.bnef.com 5. Gielen D, Lyons M (2022) Critical materials for the energy transition: lithium. International Renewable Energy Agency, Abu Dhabi 6. Martin P (2017) How much lithium is in a Li-ion vehicle battery? www.linkedin.com/pulse/ how-muchlithium-li-ion-vehicle-battery-paul-martin/ 7. USGS (US Geological Survey) (2021) Lithium. https://pubs.usgs.gov/periodicals/mcs2021/ mcs2021-lithium.pdf 8. Kelly J, Wang M, Dai Q, Winjobi O (2021) Energy, greenhouse gas, and water life cycle analysis of lithium carbonate and lithium hydroxide monohydrate from brine and ore resources and their use in lithium ion battery cathodes and lithium ion batteries. Resour Conserv Recycl 174 9. Luong VT, JunKang D, WoongAn J, AnhDao D, Kim MJ, Tran T (2014) Iron sulphate roasting for extraction of lithium from lepidolite. Hydrometallurgy 141:8–16 10. Liu X, Zhong M, Chen X, Zhao Z (2018) Separating lithium and magnesium in brine by aluminum-based materials. Hydrometallurgy 176:73–77

Part V

Poster Session

Hydrogen Storage Properties of Graphitic Carbon Nitride Nanotube Synthesized by Mix-Grind Technique Barton Arkhurst, Ruiran Guo, Ghazaleh Bahman Rokh, and Sammy Lap Ip Chan

Abstract Hydrogen has been touted as the fuel to potentially replace fossils; however, the bottleneck towards its acceptance is its storage and transportation in a manner considered practical and safe. For mobile applications, hydrogen has the potential to be used in fuel cell powered cars as a clean fuel; however, its realization rests on its efficient mode of storage. Presently, the technique of storing hydrogen in pressurized tanks at 700 bar has safety concerns. In this work, we report a mix-grind technique of synthesizing graphitic carbon nitride (g-C3 N4 ) nanotubes as hydrogen storage materials. The morphology, crystal structure, surface, and hydrogen storage properties of the samples were analyzed. Results showed that nanotubes of high specific surface area of about 114.21 m2 /g can be produced. The measured hydrogen storage capacity of the nanotubes was around 0.67 wt.% at 37 bar, at room temperature. The hydrogen storage capacity is predicted to reach 3.3 wt.% at 100 bar pressure, at room temperature. This study provides a facile approach in producing large scale g-C3 N4 nanotubes for applications such as hydrogen storage, photocatalysis, electrochemical systems, and metal free catalyst used to produce hydrogen via water splitting. Keywords Hydrogen storage · Nanotubes · Specific surface area · Adsorption · Desorption

Introduction Human lives as we know can have serious consequences without energy which shows the importance of energy to life on earth. Every sphere of task demands almost the use of energy with the bulk of this energy coming from natural gas, coal, and petroleum B. Arkhurst · R. Guo · G. B. Rokh · S. L. I. Chan (B) School of Materials Science and Engineering, University of New South Wales, Sydney, NSW 2052, Australia e-mail: [email protected] S. L. I. Chan Department of Chemical and Materials Engineering, National Central University, Zhongli 320317, Taiwan © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_22

223

224

B. Arkhurst et al.

[1, 2]. The desire by the human race to advance technologically, scientifically, and in the area of livelihood has placed enormous pressure on the globe’s energy reserves and has resulted in its depletion at a harrowing rate [3, 4]. In view of this rate of depletion of the energy reserves, natural gas, petroleum, and coal are expected to be completely exhausted by the next 60, 40, and 156 years [5]. In effect, this threatens the energy security of world economies for the foreseeable future. Apart from the dire energy security that comes with the rapid depletion of energy reserves, global warming is another issue that is of major concern when it comes to fossils. This has necessitated the need to find alternative forms of energy, which is clean, cheaper, environmentally friendly, and sustainable in the long term. Hydrogen has been tipped to be the best alternative source of energy since it is clean and sustainable in the long term [6–9] thus hydrogen economy describes a scenario where hydrogen fuel is the main energy carrier [10]. The main hold-up of the hydrogen economy from being realized is the effective and safe storage and transportation of hydrogen. For automobile applications, hydrogen presently is stored in pressurized tanks at 700 bar which poses a safety concern. As such the search for a material which can store a lot of hydrogen at a lower pressure has been on for the last decade. Nanostructured materials have been envisaged as the potential breakthrough materials. Graphitic carbon nitride (g-C3 N4 ) nanotubes have recently received much notice because of their distinctive tube structure [11, 12] and have been reported to satisfy by first principles calculation [13] the US department of energy capacity target of 5.45 wt.% required for onboard systems. g-C3 N4 nitride are normally synthesized via the use of template approach which uses chemical vapor deposition techniques into producing the nanotubes which are normally characterized by high cost, toxicity, and environmentally unfriendly substances. There have been several attempts recently to synthesize g-C3 N4 nanotubes without the use of templates [14–16]. Wang et al. [17] synthesized holey g-C3 N4 nanotubes using a one-step method by mixing melamine and cyanuric acid in specific ratio, and a nanotube with specific surface area of 50.7 m2 /g was obtained which showed enhanced photocatalytic H2 production. In this work, we report a mix-grind technique of synthesizing g-C3 N4 nanotubes. The morphology, crystal structure, and specific surface area of the sample are analyzed. Moreover, the hydrogen storage properties of the sample are investigated.

Experimental Procedure Synthesis of g-C3 N4 Nanotubes The g-C3 N4 nanotubes were prepared using a mixing-grinding technique. In this technique, first, 1 g of melamine powders (≥ 99%) was added to 5 g of cyanuric acid powders (≥ 98%). The mixture was then ground for 30 min in an agate mortar. The ground white powdery substance was then calcinated in a muffle furnace at a

Hydrogen Storage Properties of Graphitic Carbon Nitride Nanotube …

225

temperature of 550 °C at a heating rate of 5 °C/min for 4 h in the presence of N2 at a flowrate of 70 mL/min. This sample was named as CNT . A bulk g-C3 N4 designated as CN was prepared using the same technique without any addition of cyanuric acid.

Sample Characterizations The X-ray diffraction (XRD) patterns of the samples were obtained using a powder designed MPD XRD instrument with operating parameters being a current of 40 mA, voltage of 45 kV, CuKα radiation and λ of 1.5406 Å. The sample morphologies were analysed using a scanning electron microscopy (SEM) (FEI Nova nanoSEM 230) and transmission electron microscopy (TEM) (JEOL TECNAI). The operating parameters of the SEM were a spot size of 3, beam voltage of 15 kV, and 5 mm working distance. For the TEM analyses, the operating parameters were a voltage of 80–200 kV, spot size of 6, and a beam current of 500 mA. X-ray photoelectron spectroscopy (XPS) analyses of the samples were carried out using an X-ray photoelectron spectrometer (Model: Thermo Escalab 250i) with Al Kα monochromatic source (1486.68 eV) and 120 W power. Fourier transform infrared (FTIR) spectrometry analyses were carried out using FTIR spectrometer (Model: Perkin Elmer Spectrum 2) within the range 450–4000 cm−1 at a resolution of 4 cm−1 in ATR mode. Brunnauer-Emmett-Teller (BET) analyses method and Barrett-JoynerHalenda (BJH) adsorption method were used to determine the specific surface area and internal pore size distribution, respectively via N2 adsorption and desorption isotherms of the samples using a porosity analyser (TriStar 3000) at 77 K. A degassing of the samples for 3 h at 150 °C in vacuum was performed to get rid of dissolved gases and moisture before the BET and BJH tests. Hydrogen adsorption and desorption and the respective kinetics were performed using a high-pressure Sievert instrument at a temperature 25 °C and pressure range 0–3.7 MPa. The samples prior to the test were degassed at 150 °C to get rid of dissolved gases and moisture.

Results and Discussion Microstructure Analyses The morphology of the sample was analyzed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Figure 1a displays the SEM image of the bulk CN sample. It shows a bulky irregular lump-like structure with a glossy surface. Figure 1b shows the SEM image of the CNT sample which shows nanotubes of a multi-wall thickness with open ends. Interestingly, the nanotubes were an aggregate of nanotubes forming a web of nanotube structure. The TEM image in Fig. 1c

226

B. Arkhurst et al.

(a)

(b)

(c)

Fig. 1 SEM images of the sample a CN, b CNT . TEM image and c CNT

shows that the nanotube has a wall thickness of approximately 29 nm and tube diameter of around 128 nm. Interestingly, the nanotubes were an aggregate of nanotubes forming a web of nanotube structure. This method of producing g-C3 N4 nanotubes was similar to the one proposed by Wang et al. [17] except that in this method, after the mixture of the cyanuric acid and the melamine, the mixture was ground for 30 min. The average thickness and diameter of nanotubes produced were approximately 15 and 150 nm, respectively compared to this work that had an average thickness and tube diameter of around 29 and 129 nm, respectively. This implies that this method produces a much smaller nanotube diameter which is desired when it comes to hydrogen storage.

X-Ray Diffractometric Analyses Figure 2 displays the XRD pattern of sample CNT . Two pronounced peaks were observed at 13.1° and 27.70° which corresponded to (100) and (002) peak positions. Similar peaks have been reported for g-C3 N4 materials reported in literature [18–21]. In-plane repeating structure of the tri-s-triazine is associated with the (100) peak, and the (002) peak characterizes the g-C3 N4 nanotubes interplanar stacking.

Hydrogen Storage Properties of Graphitic Carbon Nitride Nanotube …

227

Fig. 2 XRD patterns of sample CNT

FTIR Analyses The chemical structure of the g-C3 N4 nanotubes was investigated with Fourier transform infrared (FTIR) technique. Figure 3 displays the FTIR spectra of the CNT and the bulk CN samples. Both samples showed similar infrared (IR) bands indicating similar chemical structures. The vibration of tri-s-triazine heterocycles corresponds to the sharp peak observed at 800 cm−1 . The peaks within the range 2900–3600 cm−1 are attributed to the stretching vibrations of the N–H bonds. The observed peaks within the range 1200–1650 cm−1 correspond to the heterocyclic units’ vibrations of the carbon nitride. The IR bands reported in this work are in tune with IR bands reported in literature for g-C3 N4 materials [21–23].

Nitrogen Adsorption/Desorption Analyses The surface properties of the g-C3 N4 nanotubes such as surface area, pore size distribution, and volume of pores were investigated by nitrogen adsorption and desorption isotherms. Figure 4a displays the N2 adsorption/desorption isotherms of the samples CNT and CN. The CNT sample exhibited a hysteresis loop of the H4 type. This H4 loop type is typical of mesoporous materials and is attributed to a collection of slit-shaped pores. Figure 4b shows the pore size distribution. The CNT and the CN samples both showed wide distribution of pores for adsorption compared to desorption. Table 1 displays the average BET surface area, pore size, and pore volume for the CNT and CN samples. The average BET surface area of the CNT sample was approximately 114.21 m2 /g which was about four times the bulk g-C3 N4 sample CN which had a BET surface area of around 28.56 m2 /g. Moreover, the CNT sample

228

B. Arkhurst et al.

Fig. 3 FTIR spectra of CNT and CN

exhibited smaller pore size and higher pore volume in comparison with the bulk CN sample. The average specific surface area of the g-C3 N4 nanotubes produced in this study was approximately 114 m2 /g which was about two times the specific surface area reported in Wang’s [17] work. This implies that this technique presents a potential to produce high-specific surface area g-C3 N4 nanotubes which is desirable in applications such as hydrogen storage, photocatalysis, electrochemical systems, and metal free catalyst used for the production of hydrogen via water splitting.

Hydrogen Storage Analyses The hydrogen storage capacity of the CNT sample was measured by the Sieverts technique. Figure 5a shows the pressure composition isotherm (PCI) curve for the CNT sample at 25 °C within the pressure range of 0–3.7 MPa. From Fig. 5, a hydrogen storage capacity of 0.67 wt.% was obtained. The reversible hydrogen storage capacity was thus approximately 0.47 wt.%. The retained hydrogen after desorption was approximately 0.2 wt.%. This retained hydrogen means that hydrogen was not fully desorbed at 25 °C. The hydrogen storage capacity was measured up to 3.7 MPa due to the limitation of the PCI instrument. However, it could be seen from the PCI curves that the hydrogen storage capacity increases as the pressure increases, and the hydrogen storage capacity is predicted to around 3.3 wt.% at 10 MPa which is lower than the hydrogen storage capacity target of 5.5 wt.% set by the US department

Hydrogen Storage Properties of Graphitic Carbon Nitride Nanotube …

229

Fig. 4 a N2 adsorption and desorption isotherms and b pore size distribution curves for samples CNT and CN

Table 1 Summary of the average pore size, pore volume, and BET surface area

Sample designation

Average pore size (nm)

Pore volume (cm3 /g)

BET surface area (m2 /g)

CNT

11.07

0.43

114.21

CN

19.22

0.10

28.56

of energy (DOE) for hydrogen storage materials for onboard systems. However, according to first principle calculations by Koh [13], the theoretical hydrogen storage capacity of g-C3 N4 nanotubes is around 5.45 wt.% which meets the requirement set by the US DOE. The nanotube diameter used in the calculation by Koh [13] was in the range 0.8–2.3 nm far lower than the results of the nanotube diameter reported in this study. This implies that the storage capacity of the g-C3 N4 synthesised using this method could be improved further by optimizing processing parameters.

Conclusion In this work, a mix-grind technique was proposed in synthesizing g-C3 N4 nanotubes. Porous nanotubes of 29 nm and 129 nm thickness and diameter, respectively, were successfully produced with a large specific surface area of around 114 m2 /g. The measured hydrogen storage capacity of the nanotubes was around 0.67 wt.% at 25 °C

230

B. Arkhurst et al.

Fig. 5 a Hydrogen adsorption and desorption isotherms at 25 °C of sample CNT and b prediction of the hydrogen storage capacity at 10 MPa through extrapolation

at a pressure 3.7 MPa with the hydrogen storage capacity predicted to reach 3.3 MPa at 10 MPa. This technique is suitable for large scale production of g-C3 N4 nanotubes for applications such as hydrogen storage, photocatalysis, electrochemical systems, and metal free catalyst used for the production of hydrogen via water splitting. Conflict of Interest The authors declare that they have no conflict of interest.

References 1. Rusman NAA, Dahari M (2016) A review on the current progress of metal hydrides material for solid-state hydrogen storage applications. Int J Hydrogen Energy 41:12108–12126 2. Sun Y, Shen C, Lai Q, Liu W, Wang D-W, Aguey-Zinsou K-D (2018) Tailoring magnesiumbased materials for hydrogen storage through synthesis: current state of the art. Energy Storage Mater 10:68–98 3. Sahaym U, Norton MG (2008) Advances in the application of nanotechnology in enabling a ‘hydrogen economy.’ J Mater Sci 43:5395–5429 4. Durbin DJ, Malardier-Jugroot C (2013) Review of hydrogen storage techniques for on board vehicle applications. Int J Hydrogen Energy 38:14595–14617 5. Midilli A, Ay M, Dincer I, Rosen M (2005) Hydrogen and hydrogen energy strategies I: current status and needs. Renew Sustain Energy Rev 9:255–271 6. Zhang B, Wu Y (2017) Recent advances in improving performances of the lightweight complex hydrides Li-Mg-N-H system. Prog Nat Sci: Mater Int 27:21–33 7. Johnson SR, Anderson PA, Edwards PP, Gameson I, Prendergast JW, Al-Mamouri M, Book D, Harris IR, Speight JD, Walton A (2005) Chemical activation of MgH2 ; a new route to superior hydrogen storage materials. Chem Commun 22:2823–2825 8. Iordache I, Schitea D, Gheorghe AV, Iordache M (2014) Hydrogen underground storage in Romania, potential directions of development, stakeholders and general aspects. Int J Hydrogen Energy 39:11071–11081 9. Deveci M (2018) Site selection for hydrogen underground storage using interval type-2 hesitant fuzzy sets. Int J Hydrogen Energy 43:9353–9368

Hydrogen Storage Properties of Graphitic Carbon Nitride Nanotube …

231

10. Abe JO, Popoola API, Ajenifuja E, Popoola OM (2019) Hydrogen energy, economy and storage: review and recommendation. Int J Hydrogen Energy 44:15072–15086 11. Wang AJ, Li H, Huang H, Qian ZS, Feng JJ (2016) Fluorescent graphene-like carbon nitrides: synthesis, properties and applications. J Mater Chem C 4:8146–8160 12. Kessler FK, Zheng Y, Schwarz D, Merschjann C, Schnick W, Wang X, Bojdys MJ (2017) Functional carbon nitride materials-design strategies for electrochemical devices. Nat Rev 2:17030 13. Koh G, Zhang YW, Pan H (2012) First principles study on hydrogen storage by graphitic carbon nitride nanotubes. Int J Hydrogen Energy 37:4170–4178 14. Gao J, Zhou Y, Li Z, Yan S, Wang N, Zou Z (2012) High-yield synthesis of millimetrelong, semiconducting carbon nitride nanotubes with intense photoluminescence emission and reproducible photoconductivity. Nanoscale 4:3687–3692 15. Wang S, Li C, Wang T, Zhang P, Li A, Gong J (2014) Controllable synthesis of nanotube-type graphitic C3 N4 and their visible-light photocatalytic and fluorescent properties. J Mater Chem A 2:2885 16. Guo S, Deng Z, Li M, Jiang B, Tian C, Pan Q, Fu H (2010) Phosphorus-doped carbon nitride tubes with a layered micronanostructure for enhanced visible-light photocatalytic hydrogen evolution. Angew Chem Int Ed 55:1830−1834 17. Wang X, Zhou C, Shi R, Liu Q, Waterhouse GI, Wu L, Tung C, Zhang T (2019) Supramolecular precursor strategy for the synthesis of holey graphitic carbon nitride nanotubes with enhanced photocatalytic hydrogen evolution performance. Nano Res 12(2019):2385–2389 18. Zhou C, Shi R, Shang L, Wu LZ, Tung CH, Zhang T (2018) Template-free large-scale synthesis of g-C3 N4 microtubes for enhanced visible light-driven photocatalytic H2 production. Nano Res 11:3462–3468 19. Jordan T, Fechler N, Xu J, Brenner TJ, Antonietti M, Shalom M (2015) “Caffeine doping” of carbon/nitrogen-based organic catalysts: caffeine as a supramolecular edge modifier for the synthesis of photoactive carbon nitride tubes. ChemCatChem 7:2826–2830 20. Wu X, Fan H, Wang W, Lei L, Chang X (2021) Ordered and ultralong graphitic carbon nitride nanotubes obtained via in-air CVD for enhanced photocatalytic hydrogen evolution. ACS Appl Energy Mater 4:13263–13271 21. Shalom M, Inal S, Fettkenhauer C, Neher D, Antonietti M (2013) Improving carbon nitride photocatalysis by supramolecular preorganization of monomers. J Am Chem Soc 135:7118– 7121 22. Niu P, Zhang L, Liu G, Cheng HM (2012) Graphene-like carbon nitride nanosheets for improved photocatalytic activities. Adv Funct Mater 22:4763–4770 23. Zhang Y, Mori T, Ye J (2012) Polymeric carbon nitrides: semiconducting properties and emerging applications in photocatalysis and photoelectrochemical energy conversion. Sci Adv Mater 4:282–291

Study on Preparation and Electrocatalytic Performance of Self-supported Carbon Transition Metal Catalysts Ze Yang, Yanfang Huang, Guihong Han, Bingbing Liu, and Shengpeng Su

Abstract Ni/CP and FeO/CP oxygen evolution electrocatalysts were prepared by one-step electrodeposition on carbon paper (CP). X-ray diffraction (XRD) and scanning electron microscope (SEM) were used to characterize the structure and morphology of catalysts. Compared with CP before deposition, CP was completely covered by Ni and FeO. The electrochemical performances of Ni/CP and FeO/CP were measured by cyclic voltammetric curves (CV), anodic polarization curve, and electrochemical impedance spectra (EIS). The results show that the overpotential of CP, Ni/CP, and FeO/CP was 776 mV, 854 mV and 788 mV respectively at the current density of 50 mA cm−2 , and the double layer capacitors were 43.30 mF cm−2 , 55.65 mF cm−2 , and 99.20 mF cm−2 , respectively. Ni/CP and FeO/CP can carry more charge and have better electrocatalytic performance of oxygen evolution; consequently they have the potential to be used as new zinc electrowinning anode materials. Keywords Electrodeposition · Oxygen evolution · Overpotential · Electrocatalytic

Introduction Electrowinning is one of the most important unit operations in zinc hydrometallurgy industry. In actual production, zinc electrowinning consumes a lot of energy and increases the total cost of production. Considering that the global fossil energy consumption is increasing, reducing the electric energy consumption in the process of zinc electrowinning has become an urgent task. The main components of zinc electrodeposition electrolyte are zinc sulfate, sulfuric acid, and water, and the reaction formula of zinc electrowinning is shown below. The cathodic reaction is shown in Eq. (1) Zn2+ + 2e− → Zn ↓

(1)

Z. Yang · Y. Huang · G. Han (B) · B. Liu · S. Su School of Chemical Engineering, Zhengzhou University, Zhengzhou 450001, PR China e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_23

233

234

Z. Yang et al.

the anode reaction is shown in Eq. (2), H2 O − 2e− → 21 O2 ↑ +2H+

(2)

the total reaction is shown in Eq. (3). Zn2+ + H2 O = Zn ↓ + 21 O2 ↑ +2H+

(3)

In the current zinc electrowinning process, the traditional industrial anode materials mainly include lead and lead alloy anode, titanium anode, graphite inert anode, ceramic inert anode, and plastic inert anode [1]. Although the conventional lead alloy anode Pb-Ag (0.8–1 wt%) has superior stability in acidic solution, it still has some shortcomings, such as high OER overpotentialand low mechanical strength, and lead is easy to dissolve and reduce the purity of zinc cathode [2–4]. The anodic reaction of zinc electrowinning is similar to the anodic OER of water decomposition. Therefore, it is an effective method to reduce the OER overpotential by introducing oxygen evolution catalyst into metal electrodeposition process [5, 6]. The research of electrocatalysts for oxygen evolution is mainly focused on metal and non-metal. The typical catalysts for oxygen evolution are precious metals Ru, Ir, Pt, and their compounds [7–9]. Transition metal-based catalysts are widely used as electrocatalysts for oxygen evolution because of their favorable electron transport effect. Electrodeposition is a traditional preparation technology of functional materials, which is widely used because of its simple preparation process and pure products. In the preparation of functional materials, the main forms include traditional powder and self-supporting preparation [10–12]. Song et al. prepared a well-behaved porous Ni– P oxygen evolution catalyst by electrodeposition with a Tafel slope of 323 mV dec−1 at a current density of 100 mA cm−2 [13]. Ni/CP and FeO/CP self-supported oxygen evolution electrocatalysts were prepared by one-step electrodeposition on carbon paper (CP) with stable spatial structure by constant current density method. The physical properties of materials were characterized. At the same time, its electrochemical performance in zinc electrodeposition system was investigated.

Experimental Material Conductive carbon paper (Ningbo Weitai Energy Technology Co., LTD.), concentrated hydrochloric acid (AR, Luoyang Haohua Chemical Reagent Co., LTD.), sulfuric acid (AR, Luoyang Haohua Chemical Reagent Co., LTD.), NiSO4 ·6H2 O (AR, Shanghai Macklin Biochemical Technology Co., LTD.), FeCl3 ·6H2 O (AR,

Study on Preparation and Electrocatalytic Performance …

235

Shanghai Macklin Biochemical Technology Co., LTD.), zinc sulfate (AR, Tianjin Yongda Chemical Reagent Co., LTD.) were prepared for electrodeposition. The reagents used in this study were of technical purity. Deionized water was homemade and used for all steps.

Preparation of Ni/CP and FeO/CP Carbon-based transition metal catalyst for oxygen evolution reaction was prepared by one-step electrodeposition by constant current density method. The transition metals involved in this study are Ni and FeO respectively. Firstly, the conductive carbon paper (CP) was cut with a working area of 2.5 * 3.5 cm2 . Then, the conductive carbon paper was ultrasonically treated for 600 s in 1 mol L−1 hydrochloric acid solution to increase its hydrophilicity. Next, the acid-treated conductive carbon paper is rinsed three times with deionized water to remove residual acid from the surface. It was then dried for 20 min at 65 °C. Finally, the pretreated carbon paper was weighed for use. The pretreated conductive carbon paper was immersed in 80 mL electrodeposition liquid containing transition metal ions at a concentration of 0.1 mol L−1 , and the electrodeposition liquid was fully infiltrated into the substrate by ultrasonic 600 s at 25 °C. The transition metal salt solutions used in this process are NiSO4 and FeCl3 respectively. Then, a three-electrode system was used for cathodic electrodeposition on an electrochemical workstation. The working electrode, counter electrode, and reference electrode were conductive carbon paper, carbon rod, and saturated calomel electrode respectively. The electrodepositions of Ni and FeO on CP were under a current density of − 20 mA cm−2 for 1800 s. Finally the working electrode orientation was switched to repeat the above steps on the back of the electrode. The purpose of double-sided electrodeposition is to make the electrode surface deposition uniform. The comparative samples were dried at 25 °C.

Characterization of Material Scanning electron microscopy (SEM, TESCAN, MIRA 3 LMH/LMU) was used to detect the morphology of the samples, and the element distribution on the electrode surface was obtained by energy chromatograph (OXFORD Xplore, EDS). The physical properties of related materials were determined by X-ray diffractometer (XRD, Dutch PANalytical X-ray diffractometer), excitation voltage was 45 kV, excitation current was 30 mA, 2θ range was 5–90°, and scanning step was 0.02°. X-ray diffraction (XRD) patterns of the samples were obtained by radiation.

236

Z. Yang et al.

Electrochemical Measurement of Material The electrochemical performance of material was measured with a three-electrode by using electrochemical workstation (Metrohm Autolab, Swiss, PGSTAT 302N). The working electrode was 2.5 * 3.5 cm2 Ni/CP and FeO/CP, the counter electrode was graphite electrode, and the reference electrode was saturated calomel electrode. The electrolytes were 50 g/L Zn2+ and 150 g/L H2 SO4 solution. Linear sweep voltammetry (LSV) measurements were performed with a scan rate of 10 mV s−1 , and the scan range is between 0.4 and 2.5 V. Electrochemical impedance spectroscopy (EIS) techniques were measured between 0.1 Hz and 100 kHz, and the amplitude is 10 mV.

Results and Discussion Characterization of Ni/CP and FeO/CP The SEM determination results are shown in Fig. 1. It can be seen from Fig. 1b that Ni completely covers the surface of CP and has a high contrast in its photos, which proves that it has a good electron reflection effect. Meanwhile, it indicated that the surface conductivity of Ni/CP is superior and conducive for charge transmission. It can be seen from the high magnification (partial magnification) that the Ni deposits are tiny particles with a single particle size of about 10–20 nm. The nanometer particle size increases the specific surface area of the catalyst, and it provides a large number of active sites for the reaction. As shown in Fig. 1f, FeO was a micron sheet composed with a large number of nanoparticles, which was deposited on the substrate, increasing the number of atoms of the catalyst and providing a large number of active sites for the reaction.

Fig. 1 a–c Low magnification, d–f high magnification scanning electron microscopy images (SEM) of contrast CP, Ni/CP, and FeO/CP

Study on Preparation and Electrocatalytic Performance …

237

The SEM–EDS images of materials are shown in Fig. 2. The results of EDS surface scanning showed that Ni/CP catalyst was composed of C, O, and Ni, while FeO/CP catalyst was composed of C, O, and Fe. The doping of transition metal elements can improve the catalytic performance of the catalyst and provide more active sites for electrode reaction due to the participation of different elements in the reaction process. The XRD diffractograms of the materials are shown in Fig. 3; it can be seen from Fig. 3a that the Ni/CP catalyst has sharp narrow peaks at 44.5°, 51.8°, and 76.4°, corresponding to Ni (JCPDS#87-0712). As shown in Fig. 3b, the characteristic peaks of graphite at 42.5° and 44.6° disappeared after Ni was deposited on the surface, which may be due to the influence of Ni in the deposition process. It can be seen that, compared with CP, FeO/CP catalyst has sharp small peaks at 36.0°, 41.9°, 60.7°, 76.5°, and 91.3°, corresponding to FeO (JCPDS#06-0615).

Fig. 2 Energy dispersion X-ray spectrum (EDS) mapping scanning distribution of a Ni/CP, b FeO/CP catalyst

Fig. 3 a Wide range, b enlarged X-ray diffraction patterns (XRD) of Ni/CP, FeO/CP catalyst, and the CP

238

Z. Yang et al.

Electrochemical Text   RT According to Nernst equation: E = − G = Eθ −  ln J , J on behalf of each nF nF reaction component activity product, namely J = B a vBB , when reference electrodes are SCE. The conversion relationship between standard potential and referθ = ence electrode potential at different temperatures follows the equation: E SCE (0.2415 − 0.000761(T − 298.15))V , therefore, different temperatures, Relative to the electrode reaction potential of reversible hydrogen electrode (RHE) reaction  θ V [14]. The overpotential (η) of OER potential for E = E SCE + 0.0591 p H + E SCE is matched with this formula: η = (E RHE − 1.23)V , where 1.23 V is the standard potential for oxygen evolution reaction [15]. Figure 4a and b shows the anodic polarization curve and Tafel slope fitting results of CP, Ni/CP, and FeO/CP in 50 g/L Zn2+ + 150 g/L H2 SO4 solution, respectively. As shown in Fig. 4a and Table 1, when the current density passing through the electrode surface was 50 mA cm−2 , CP showed a lower OER overpotential, and the overpotential of Ni/CP and FeO/CP catalysts was 854 mV and 788 mV, respectively, which was different from the trend in 150 g/L H2 SO4 solution system [16]. The Tafel slopes of CP, Ni/CP, and FeO/CP materials are close, which indicated that the materials have similar electrocatalytic activities in zinc electrowinning system.

Fig. 4 a Anode polarization curve, b Tafel slope of Ni/CP, FeO/CP catalyst, contrast CP, and pure Pb anode in 50 g/L Zn2+ + 150 g/L H2 SO4 solution

Table 1 Comparison of OER of Ni/CP, FeO/CP catalyst, contrast CP, and pure Pb anode in 50 g/L Zn2+ + 150 g/L H2 SO4 solution Anode material

Overpotential (mV@50 mA cm−2 )

Tafel slope (mV dec−1 )

CP

776

37

Ni/CP

854

35

FeO/CP

788

36

Pure Pb

1030

27

Study on Preparation and Electrocatalytic Performance …

239

Figure 5a–c shows cyclic voltammetry curves of CP, Ni/CP, and FeO/CP at different sweeping speeds, respectively. The anodic current density and cathodic current density in each non-Faraday region were averaged numerically, and the relationship between the current density and the sweep speed is shown in Fig. 5d. The slope of the fitted line is the double-layer capacitance, and the electrocatalytic performance of OER catalyst can be further studied according to the capacitance of double electric layer. Further, as shown in Fig. 5d, in the non-Faraday region (1.12 V (vs. RHE)), the double capacitance of CP, Ni/CP, and FeO/CP is 43.30, 55.65, and 99.20 mF cm−2 , respectively. The larger the double layer capacitance is, the larger the charge can be carried by the catalyst in the electrochemical reaction. Ni/CP and FeO/CP show good electrocatalytic activity in zinc electrowinning system. Cell voltage curves of Ni/CP, FeO/CP catalyst, and pure Pb anode during zinc electrowinning process are shown in Fig. 6. Compared with the pure Pb anode, the cell voltage corresponding to Ni/CP is lower. In the testing process, the cell voltage of the pure Pb is corrugated, corresponding to the dissolution—generation of β-PbO2 of Pb surface. The cell voltage of Ni/CP anode tends to be stable after rising and finally stabilizes at 2.80 V. The cell voltage of FeO/CP anode increases and eventually stabilizes at 2.82 V. With the progress of electrowinning, the cell voltage curve tends

Fig. 5 Cycle voltammetry curves of a CP, b Ni/CP, and c FeO/CP in 50 g/L Zn2+ + 150 g/L H2 SO4 solution at different sweep rates for 20, 40, 60, 80, 100 mV s−1 , d scanning rate-current density relationship at 1.12 V (vs. RHE) and the fitting diagram of the double layer capacitance (C dl )

240

Z. Yang et al.

Fig. 6 Cell voltage curves of Ni/CP, FeO/CP catalyst, and pure Pb anode during zinc electrowinning process

to be smooth without drastic fluctuation. The results show that Ni/CP and FeO/CP have better OER catalytic performance than Pb anode, and FeO/CP shows better stability.

Conclusion In this study, self-supported carbon transition metal catalysts are prepared by one-step electrodeposition on carbon paper (CP). XRD and SEM results show that the electrode surface structure remains stable before and after the zinc electrowinning without disintegration. The electrochemical performances show that when the temperature is 35 °C, the double capacitance of FeO/CP can be maintained at 99.20 mF cm−2 in the non-Faraday region (1.12 V (vs. RHE)), and the electrode can carry more charge in the process of zinc electrowinning. Meanwhile, during the 3600 s zinc electrowinning process, the cell voltage of FeO/CP shows excellent stability. Further, the OER catalytic performance of FeO/CP electrode is better than Ni/CP, and it has the potential to be used as a new zinc electrowinning carbon-based anode. This study provides a possible direction for the development of anode materials for zinc electrowinning. Acknowledgements This work was financially supported by the Original Exploration Project of China (52150079), the Natural Science Foundation of China (U2004215, No. 51974280 and No. 51774252), and the Educational Commission Fund of Henan Province of China (No. 20HASTIT012, No. 18A450001, and No. 17A450001).

Study on Preparation and Electrocatalytic Performance …

241

References 1. Ye W, Xu F, Jiang L (2021) Lead release kinetics and film transformation of Pb-MnO2 precoated anode in long-term zinc electrowinning. J Hazard Mater 408:124931 2. Yang HT, Liu HR, Guo ZC (2013) Electrochemical behaviors of Pb–0.3%Ag-0.06%Ca rolled alloy anode during and after zinc electrowinning—Tafel investigations. Appl Mech Mater 401:779–782 3. Petrova M, Noncheva Z, Dobrev T (1996) Investigation of the processes of obtaining plastic treatment and electrochemical behaviour of lead alloys in their capacity as anodes during the electroextraction of zinc I. Behaviour of Pb-Ag, Pb-Ca and PB-Ag-Ca alloys. Hydrometallurgy 40(3):293–318 4. Felder A, Prengaman RD (2006) Lead alloys for permanent anodes in the nonferrous metals industry. JOM 58(10):28–31 5. Liang Q, Brocks G, Bieberle-Hütter A (2021) Oxygen evolution reaction (OER) mechanism under alkaline and acidic conditions. J Phys: Energy 3(2):026001 6. Liyanage DR, Li D, Cheek QB (2017) Synthesis and oxygen evolution reaction (OER) catalytic performance of Ni2−x Rux P nanocrystals: enhancing activity by dilution of the noble metal. J Mater Chem A 5(33):17609–17618 7. Wang J, Han L, Huang B (2019) Amorphization activated ruthenium-tellurium nanorods for efficient water splitting. Nat Commun 10(1):1–11 8. Reier T, Oezaslan M, Strasser P (2012) Electrocatalytic oxygen evolution reaction (OER) on Ru, Ir, and Pt catalysts: a comparative study of nanoparticles and bulk materials. ACS Catal 2(8):1765–1772 9. Seitz LC, Dickens CF, Nishio K (2016) A highly active and stable IrOx /SrIrO3 catalyst for the oxygen evolution reaction. Science 353(6303):1011–1014 10. Dem’yanets LN, Ivanov-Shits AK, Kireev VV (2004) Electric-field effect on crystallization in the Li3 PO4 –Li4 GeO4 –Li2 MoO4 –LiF system. Inorg Mater 40(8):874–877 11. Sieradzki K, Brankovic DN (1999) Electrochemical defect-mediated thin-film growth. Science 284(5411):138–141 12. Budevski E, Staikov G, Lorenz WJ (2000) Electrocrystallization: nucleation and growth phenomena. Electrochim Acta 45(15–16):2559–2574 13. Song D, Hong D, Kwon YK (2020) Highly porous Ni–P electrode synthesized by an ultrafast electrodeposition process for efficient overall water electrolysis. J Mater Chem A 8(24):12069– 12079 14. Zhang W, Robichaud M, Ghali E (2016) Electrochemical behavior of mesh and plate oxide coated anodes during zinc electrowinning. Trans Nonferrous Met Soc China 26(2):589–598 15. Walter MG, Warren EL, McKone JR (2010) Solar water splitting cells. Chem Rev 110(11):6446–6473 16. Negahdar L, Zeng F, Palkovits S (2019) Mechanistic aspects of the electrocatalytic oxygen evolution reaction over Ni−Co oxides. ChemElectroChem 6(22):5588–5595

Modification and Evaluation of Energy Saving and Consumption for Reduction Technology of 500 t/d Beckenbach Annular Lime Kiln Yapeng Zhang, Wen Pan, Zhenping Miao, Jianbo Zhu, Shaoguo Chen, Huaiying Ma, and Zhixing Zhao Abstract The ejector system of annular lime kiln was modified for energy saving and pollution reduction. Under the principle of high air pressure and speed, the driving air flow and movement system was improved to achieve high ejector pressure and strong siphon effect. The technical transformation has broken the traditional cognition that two driving fans must be applied at full load in the industry. A single fan can meet the full production. The power and fuel consumption potentially reduced, and the output of quicklime increased. After renovation works, about 125 m3 /h of coke oven gas and 250 m3 /h of converter gases were saved. The temperature of the combustion chamber increased for 20 °C after modification. Keywords Beckenbach annular lime kiln · Ejector · Siphonic effect · Energy-saving and cost-reducing

Introduction QuickLime is an indispensable and important auxiliary raw material in the process of iron and steel production [1], which has a significant impact on the key processes such as hot metal pretreatment, steel making, and steel refining. Sleeve kiln technology [2– 4] is one of the advanced quicklime production technologies. Shougang Changgang also uses the Beckenbach annular sleeve shaft kiln [5] (output: 500 t/d) to produce quicklime. One of the significant technological characteristics of the annular sleeve kiln is the counter current calcination and parallel current calcination simultaneously in the same kiln chamber. Parallel current calcination is the most prominent feature Y. Zhang (B) · W. Pan · S. Chen · H. Ma · Z. Zhao Beijing Key Laboratory of Green Recyclable Process for Iron & Steel Production Technology, Beijing 100043, PR China e-mail: [email protected] Research Institute of Iron & Steel, Shougang Group Co., LTD Research Institute of Technology, Beijing 100043, PR China Z. Miao · J. Zhu Shougang Changgang Steel & Iron Co., LTD., Changzhi 046031, PR China © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_24

243

244

Y. Zhang et al.

of the calcination process of the annular sleeve kiln. To a large extent, it determines that the sleeve kiln has the advantage of low energy consumption. The driving air of the ejector realizes the circulation of hot air in the kiln. The circulating gas of the annular sleeve kiln is the key to the whole calcination process and the energy-saving control of the sleeve kiln [6, 7]. The formation of the circulating gas is affected by two forces: one is the upward pumping force generated by the high-temperature exhaust gas fan, the other is the upward pumping force generated by the ejector in the lower inner sleeve driven by the high-pressure air generated with the driving fan, and the downward pumping force in the parallel current calcination belt. The pumping force in the counter current calcination zone is completely generated by the ejector system. Before the technical transformation of the sleeve kiln in flux plant of Shougang Changgang group, the ejector system was driven by two 132 kW motors. From June 2017 to March 2018, following the principle of high air pressure and strong air speed, the driving air system was improved to increase the injection pressure, improve the strong negative pressure siphon effect, reduce one 132 kW motor, and change the frequency value of the other motor from 50 to 40 Hz. In addition, due to the reduction of the cold air volume during the heat exchange process, the adhesive blockage caused by the mixing of dust and moisture in the tube bundle of the heat exchanger is effectively controlled, and the cleaning interval of the heat exchanger is extended, from 12 h per month to 12 h every two months, thus increasing the production and stabilizing the production. Through the above transformation, the recognition that two driving fans must be started in the industry to achieve full production is broken, and the technical innovation that a single fan can meet full production is realized.

Methods of Technical Transformation The 500 t/d Backbench annular sleeve shaft kiln of Shougang Changgang has been using the previous traditional double drive fan configuration. It was found that there were some problems in the ejector technology, mainly including unreasonable ejector structure, large diameter, low injection efficiency and serious on-site blockage. At the same time, it was easy to cause blockage of the heat exchanger tube bundle and forced to stop production for maintenance. The maintenance period was 20–25 days, which seriously affected the normal operation of the kiln. The ejector of sleeve kiln must be reformed in order to change this situation.

Modification and Evaluation of Energy Saving and Consumption …

245

Analysis of Transformation Scheme It is one of the biggest process characteristics of the annular sleeve kiln to realize counter current and parallel current calcination in the same kiln chamber. The circulating gas running in the kiln promotes the parallel current calcination, which is also the key factor to form the parallel flow belt in the sleeve kiln process. Circulating gas refers to the fuel and combustion supporting air entering from the lower burner. After the counter current calcination belt at the lower part of the lower combustion chamber fully reacts with the limestone raw material, it enters the upper section of the lower inner cylinder together with the cooling air entering from the cooling belt at the bottom of the kiln, and then flows into the injector through the circulating gas channel in the upper arch bridge. After mixing with the hot air in the injector, it is injected into the lower combustion chamber together. The quality of lime products is directly affected by the operation of circulating gas, which is mainly formed by the driving air through the ejector. The main functions of the driving air are as follows: first, it provides strong power for the formation of circulating gas and is the original power for the formation of the parallel flow calcination belt, so that the driving air at the lower arch bridge position and the suction of the exhaust gas fan in the kiln reach a relative balance, which is an important factor determining the position of the calcination belt; second, its air volume participates in the combustion in the kiln as secondary air. The actual air volume for production is provided by two roots fans, one for full pressure startup and one for variable frequency speed regulation. The driving air generated by the driving fan enters the heat exchanger, and after heat exchange, it changes from cold air to hot air, flows into the upper ring pipe, and then flows into the ejector injection pipe. The high-speed air flow also generates siphon effect at the nozzle position, driving the gas flow around the injection pipe (circulating gas channel). At this time, the gas inside and outside the nozzle enters the lower combustion chamber together to participate in combustion. The driving air injected into the lower combustion chamber makes the high-temperature gas in the lower combustion chamber enter the annular calcination space in the kiln, and most of the gas flows downward (in the same direction as the material flow) to form a cocurrent calcination zone. The co-flow belt gas descends to the inlet of the circulating gas of the lower inner cylinder, enters the lower inner cylinder, flows through the circulating gas channel, reaches the ejector and the driving air meet at the nozzle of the injection pipe, continues to flow, and thus circulates. If the nozzle diameter of the ejector was too big, it will cause the diameter of the air column to be too large, and the large air volume is not conducive to the suction of the circulating gas; if the diameter of the nozzle is too small, the diameter of the air column is too small, and the vacuum state cannot be formed and the circulating gas cannot be drawn. As there is no precedent to learn from, a new type of ejector for the sleeve kiln was designed according to the actual experience of the production of the sleeve kiln and the existing technology in the contract industry, so as to improve the air flow structure of the annular sleeve kiln, achieve the high-efficiency spraying effect, and

246

Y. Zhang et al.

Fig. 1 Schematic diagram of reconstruction of intercepting cyclone distributor of ejector

realize the technical innovation that a single drive fan can meet the full production of the sleeve kiln. Referring to the process principle of “high air pressure and low air volume” in the combustion theory of other kiln types (such as rotary kiln burners) for calcining lime, a new reconstruction of intercepting cyclone distributor of ejector is designed. Figure 1 shows the schematic diagram of reconstruction of intercepting cyclone distributor of ejector. By adjusting the screw rod to drive the intercepting cyclone distributor to move up and down in the axial direction in the injection pipe, the ventilation area of the ejector pipe can be adjusted online to control of the driving air volume, pressure, and speed. At the same time, driven by the swirling high air speed, a larger strong negative pressure zone is formed, which has a stronger siphon effect on the high-temperature circulating gas. A large amount of high-temperature circulating gas participates in combustion support and reduces gas consumption.

Manufacture and Application of Intercepting Cyclone Distributor (1) Design and type selection After the demonstration, three structural types are selected and carried out in sequence: first, the diameter is reduced and a ring is set in the original nozzle. If the desired effect is not achieved, the cross-sectional shape of the nozzle shall be improved. The inner diameter of the nozzle shall be unchanged, and a

Modification and Evaluation of Energy Saving and Consumption …

247

cylindrical inner core shall be added at the center of the nozzle. If the desired effect is not achieved, the cross-sectional shape of the nozzle shall further be improved and a conical inner core shall be added at the center. (2) Suitable materials selection In order to reduce the production cost, a set of intercepting cyclone distributor is trial manufactured with ordinary carbon steel, and then the material requirements are optimized according to the actual working place and working conditions. If the preferred material is not feasible, the high-temperature ablation and high-speed air flow scouring at the working place shall be mainly considered, so the stainless steel material shall be used. (3) Fabrication and adjustment First, the prefabricated intercepting distributor shall be processed according to the predetermined design with the general manufacturing process. And the injection pipe shall be disassembled and taken out with the help of low maintenance opportunity. The new intercepting distributor shall be directly installed into the inside from the nozzle of the injection pipe, the ejector shall be fixed in place, and then put into production. (4) Industrial trial According to the industrial trial situation, the intercepting cyclone distributor was adjusted and improved. After the determination of the manufacturing material and structural type, the most reasonable structural parameters were optimized according to the test process.

Comparison of Indexes Before and After Transformation Through a large number of trials, it is basically determined that the intercepting cyclone distributor should be made of stainless steel, and the structural type should be conical core. After using the intercepting cyclone distributor, the function of the ejector is obviously strengthened. The structure of the new ejector is simple and compact, which is convenient for control and operation, and various technical and economic indexes of the sleeve kiln have been improved. (1) Power consumption is reduced. Since the traditional configuration of using two driving fans at the same time for the sleeve kiln is broken, the full production of the sleeve kiln is realized by using a single drive, thus reducing the power consumption (see Table 1). Before the transformation, the power consumption is 44–45 kWh/tlime in the normal production of the sleeve kiln; after the transformation, the power consumption is 39–40 kWh/tlime , and the average power consumption reduced about 5 kWh/tlime . (2) Reduce fuel consumption. The indexes such as driving air volume and gas consumption before and after the transformation are shown in Table 2. From the changes, it can be seen that the new ejector changes the sleeve kiln from the

248

Y. Zhang et al.

Table 1 Electric energy consumption before and after transformation of ejector Before/after

Average output (tlime /d)

Average power consumption (kWh/tlime )

Before

478

44

After

481

39

Table 2 Driving air volume and gas consumption before and after transformation of ejector Parameters

Before

After

Output (tlime /d)

470

480

490

470

480

490

Temperature of driving air (°C)

800

805

810

785

790

795

Pressure of driving air (kPa)

41

41

41

41–43

41–43

41–43

Volume of driving air (m3 /h)

7900

7800

7800

4800

4800

4800

Converter gas consumption (m3 /t)

440

455

465

420

435

450

45

50

55

40

45

50

Coke oven gas consumption

(m3 /t)

dual machine configuration to the single machine configuration. Due to the full play of its role, the total amount of driving air is reduced, and the amount of gas and air involved in combustion is reduced (i.e. low nitrogen combustion), so the N and O compounds produced by combustion are naturally reduced. The content of NOx emitted into the air is also reduced correspondingly, and the average daily NOx emission can be reduced by about 5000 m3 , which has made contributions to environmental protection. (3) Increase the combustion chamber temperature. After the new ejector is used in the sleeve kiln, the driving air volume is greatly reduced. On the one hand, the heat loss caused by a large amount of low-temperature air entering the combustion chamber is reduced. And the participation of a large amount of high-temperature circulating gas in the combustion is ensured. At the same time, the combustion air is better distributed and utilized to improve the temperature of the combustion chamber. Under the same gas (converter gas 8000 m3 /h) consumption, the temperature of combustion chamber before transformation is about 1190 °C and after transformation is 1210 °C. (4) Effectively solve the blockage of heat exchanger. Figure 2 shows the scaling state of the flow tube bundle of the heat exchanger before and after the transformation, Fig. 2a shows the internal dirt state of the heat exchanger after 25 days of operation before the transformation, and Fig. 2b shows the internal dirt state of the heat exchanger after 60 days of operation after the transformation. During the cold and hot air exchange process of the drive air through the heat exchanger, the flow speed of the drive air is slowed down after the transformation, which effectively controls the phenomenon of sticking and blocking on the inner wall of the convection tube bundle due to excessive temperature difference, thus prolonging the period of keeping the tube bundle of the heat exchanger unblocked. The

Modification and Evaluation of Energy Saving and Consumption …

249

Fig. 2 Scaling state diagram of flow tube bundle of heat exchanger before and after transformation

cleaning period is extended from the original 25 to 120 days, which greatly saves the shutdown and maintenance time (8 days for the whole year).

Conclusions (1) By improving the driving air system, increasing the injection pressure and enhancing the strong negative pressure siphon effect, the full production of a single driving fan of the sleeve kiln can be realized, and the power consumption can be greatly reduced, breaking the recognition that two driving fans must be started to produce at full in the industry. (2) After the new ejector is used in the sleeve kiln, the driving air volume is greatly reduced. On the one hand, the heat loss caused by a large amount of lowtemperature air entering the combustion chamber is reduced, and the participation of a large amount of high-temperature circulating gas in the combustion is ensured. At the same time, the combustion air is better distributed and utilized, so as to improve the temperature of the combustion chamber. Under the same gas (converter gas 8000 m3 /h) consumption, the temperature of combustion chamber before transformation is about 1190 °C and after transformation is 1210 °C. (3) After the technical transformation, the adhesive blockage caused by the mixing of dust and moisture in the tube bundle in the heat exchanger can be effectively controlled, and the cleaning interval of the heat exchanger can be extended from 12 h per month to 12 h every two months, thus increasing the output and stabilizing the production.

250

Y. Zhang et al.

Acknowledgements The authors are grateful to the financial support of National Key R&D Program of China (2017YFB0304300 and 2017YFB0304302).

References 1. Chu J, Gao S (2009) Technical manual for metallurgical lime production. Metallurgical Industry Press, Beijing 2. Wang H, Guo W, Jie T (2016) Technical development report of sleeve lime kiln in 2015. Fire Refract Lime 41(03):1–13 3. Zhao Y (2012) Process design of 500 t/d sleeve lime kiln. Ind Furnace 34(05):23–25 4. Fu T (2009) Technical progress and development of annular sleeve kiln. China Iron Steel Ind 09:24–26 5. Zhou H, Shougang (2013) Metallurgical lime technology innovation and engineering practice. In: Proceedings of the 9th China iron and steel annual conference, China Metal Society, Beijing, pp 7–8 6. Liang Y, Shan W, Qi X (2014) Research and optimization of sleeve lime kiln technology in Jigang. Mech Eng 2014(02):57–58 7. Zou L, Yan D, Tu S (2009) Process characteristics of producing active lime by gas fired sleeve lime shaft kiln. Ind Furnace 31(01):13–15

Research on the Gasification Characteristic of Cokes of BIOC-HPC Extracted from the Mixture of Low-Rank Coal and Biomass Jun Zhao and Xueya Wang

Abstract The aim of this paper is to provide theoretical guidelines for biomass and low-rank coal in coking and promoting the development of low carbon ironmaking; the modified coal (BIOC-HPC) was produced, and its performance as an additional component in coking process was also investigated through co-thermal extraction method. The effect of the addition amount of BIOC-HPC extracted from 80 wt.% low-rank coal and 20 wt.% biomass on the gasification performance of BIOC-HPC coke was investigated. The results showed that with the increase of BIOC-HPC content, the coking yield decreased after high-temperature carbonization, from 70.45 to 68.42 wt.%, while the ash content of coke is decreased from 11.82 to 10.05 wt%. The gasification reaction rate of BIOC-HPC coke gradually increases with the increase of BIOC-HPC ratio and the gasification temperature. The optimal ratio of BIOC-HPC in coking is in the range of 10–20 wt.%; the pore size is relatively uniform, about 10–19 nm. Keywords Low-rank coal · BIOC-HPC · Biomass · Gasification

J. Zhao Jiangsu Key Laboratory of Coal-Based Greenhouse Gas Control and Utilization, China University of Mining and Technology (CUMT), Xuzhou 221008, PR China Faculty of Safety Engineering, China University of Mining and Technology (CUMT), Xuzhou 221116, PR China State Key Laboratory of Advanced Metallurgy, University of Science and Technology Beijing, Beijing 100083, PR China X. Wang (B) School of Chemistry and Biological Engineering, University of Science and Technology Beijing, Beijing 100083, PR China e-mail: [email protected] © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_25

251

252

J. Zhao and X. Wang

Introduction China’s energy structure is characterized by rich coal, poor oil, and low gas. Coal accounts for 60–65% of China’s primary energy consumption. Clean and efficient use of coal is still one of the most important energy strategies in China [1–3]. Coking coal is a scarce resource worldwide, and it is also an important strategic resource of China [4]. Among the world’s coal resources, coking coal is less than 10% of the total amount. Of the total 1.34 trillion tons of coking coal resources, fat coal, coking coal, and lean coal account for about 50%, and economically recoverable reserves are only 350–400 billion tons, of which only 60 billion tons of high-quality coking coal resources with low ash and low sulfur [5]. In recent years, the national coke demand has continued to be at a high level. However, due to the scarcity of high-quality fat coal and main coking coal resources, the overall quality of coke has declined, and the price of high-quality coke has increased year by year. This has led to the need for a large amount of imported coking coal in the country, which is unpredictable in the international trade environment. Under this circumstances, it has become a hidden danger for the future development of the coke industry [6– 8]. Therefore, increasing the proportion of non-coking coal, reducing coal blending costs, and improving coke quality have become one of the effective ways to solve the problem of shortage of coking coal resources [9]. Research and practice have shown [10] that solvent-modified coal can replace about 5–10% of the coking coal consumed in the coking process, which can effectively reduce the cost of coking production, improve the metallurgical properties of coke, improve the energy supply of steel production enterprises problems, improve the economic efficiency of enterprises, and reduce environmental pressure. China’s low-rank coal resources have the characters of high ash, high water, poor caking property, and low calorific value [11]. It is difficult to be directly used for coking and coal liquefaction and gasification, resulting in waste of low-rank coal resources [12]. Therefore, many scholars have focus on the low-rank coal upgrading technology, such as drying dehydration modification technology, molding modification technology, acid–base upgrading technology, and solvent extraction technology [11–14], in order to produce modified coal with no ash and excellent caking properties. Among these low-rank coal modification technology reported, solvent extraction technology is widely used in the modification of low-rank coal, because it has the features of simplicity of operations, reusing of solvent, and environmentfriendly. Fischer [13] conducted solvothermal extraction of coal and separated a cohesive component, and found that the residual coal lost its cohesiveness after thermal extraction. It was found that the cohesiveness was enhanced by adding this component to weakly caking coal. At present, many researchers reported on solvent thermal extraction of low-rank coal mainly involve solvent extraction process, reaction mechanism, and high-value utilization [11–14]. Renganathan et al. [14] used a mixed solvent of N-methylpyrrolidone (NMP) and carbon disulfide (CS2 ) to perform thermal extraction of high-sulfur coal, results showed that the extraction rate is as high as 74 wt.%, the ash content is less than 0.7 wt.%, and the modified coal has

Research on the Gasification Characteristic of Cokes of BIOC-HPC …

253

higher fluidity and caking properties. Therefore, ash and harmful elements in lowrank coal can be removed by the solvent extraction method, improving fluidity and thermoplasticity of low-rank coal, and meeting the requirements of coking coal. By using thermal extraction method, it can not only improve the utilization efficiency of low-rank coal resources, but also reduce the consumption of high-quality coking coal, reduce coking production costs, and realize the green application of low-rank coal resources. In this paper, modified coal is extracted from the mixture of low-rank coal and sawdust residue by using N-methyl-2-pyrrolidinone (NMP) under high temperature and pressure. According to the previous study [11, 12], it can be found that modified coal, which is also called hypercoal (HPC) in some researches [12, 13], showed the features of excellent caking property and fluidity and ash-free. Therefore, the modified coal in this paper was called BIOC-HPC in order to define this is the product of the co-thermal extraction of biomass and low-rank coal. Considering that thermogravimetric analysis is one of the most effective methods to describe the gasification performance of coke, it has been widely used in the study of coke gasification characteristics by many researchers [11–14]. Therefore, the gasification characteristics of BIOC-HPC coke formed by the mixing of BIOC-HPC and blending coals in different proportions were studied by thermogravimetric analysis. The aim of this paper is to provide theoretical guidelines for biomass and low-rank coal in coking and promoting the development of low carbon ironmaking.

Experimental Materials A sub-bituminous coal (KL) and coking coal (SH) were used in this study. The biomass was obtained from the sawdust of China fir residue which is the most common type of biomass in Northern China. The coal samples and the biomasses were dried at 80 °C for 12 h in a vacuum and then ground until they were less than 200 mesh (74 µm). The samples were deposited into sealed bags until they were used in the experiments. The polar solvents including N-methyl-2-pyrrolidinone (NMP) (AR, > 99.0%) and ethanol were provided by Aladdin (China). The proximate and ultimate analysis of coal samples were performed according to the methods recommended by the National Standard of China GB/T476-2001 and GB/T212-2008, respectively.

254

J. Zhao and X. Wang

The Preparation of BIOC-HPC and BIOC-HPC Coke Figure 1 represents the overview of BIOC-HPC and coke preparation. The previously published work gave the utmost experimental guidance and optimum processing parameters [11, 12]. Thus, the ratio of biomass and coal and the volume of polar solvent NMP were fixed at 1:4 and 400 mL, respectively. And the corresponding BIOC-HPC was prepared under the optimal reaction conditions. The BIOC-HPC was blended with GD and SH according to the standard mixing ratios listed in Table 1. The blended coals, approximately 3 g, were then poured into a graphite crucible and loaded in the furnace. Subsequently, the blend coals were carbonized under a load of 6 kPa induced by an iron rod at 1000 °C for 2 h in a nitrogen atmosphere. After that, the furnace was cooled to room temperature under the protection of nitrogen, and the coke yield was calculated by the weight loss. The BIOC-HPC cokes prepared by different proportions of BIOC-HPC were labeled 0%-BIOC-HPC, 5%-BIOC-HPC…20%-BIOC-HPC, respectively.

Results and Discussion Physical and Chemical Property of Raw Samples The proximate analysis and coking yield of coke made from BIOC-HPC under different ratios are listed in Table 2. It could be found that with the increase of BIOC-HPC content, the coking yield decreased after high-temperature carbonization, from 70.45 to 68.42 wt.%. The main reason is that the high volatile content of BIOC-HPC escaped during the coke preparation. Due to the ultimate low ash content of BIOC-HPC, the ash content of coke is decreased from 11.82 to 10.05 wt.% with the increase of BIOC-HPC content. It suggests that adding BIOC-HPC effectively decreased the ash content of coke and improves its quality. When using BIOC-HPC coke in the blast furnace, it was conducive to reducing the slag volume and prolonging the service life of blast furnace.

The Gasification Rate of BIOC-HPC Coke The relationship between weight loss rate and time of BIOC-HPC coke under different gasification temperatures is shown in Fig. 2. By comparing the gasification weight loss curves of BIOC-HPC coke with different ratios, it can be found that with the increase of BIOC-HPC ratio, the gasification reaction rate of BIOC-HPC coke gradually increases. Compared with the conditions of 1100 and 1200 °C, the gasification rate of coke increases greatly with the increase of temperature, and the gasification rate of coke prepared by adding BIOC-HPC with a proportion of 20%

Fig. 1 Schematic diagram of BIOC-HPC and coke preparation and gasification

Research on the Gasification Characteristic of Cokes of BIOC-HPC … 255

0%-BIOC-HPC

95

5

0

Sample

SH

GD

BIOC-HPC

5

5

90

5%-BIOC-HPC

Table 1 Mixing ratios of the coal used in the preparation of cokes (wt.%)

10

5

85

10%-BIOC-HPC

15

5

80

15%-BIOC-HPC

20

5

75

20%-BIOC-HPC

256 J. Zhao and X. Wang

Research on the Gasification Characteristic of Cokes of BIOC-HPC …

257

Table 2 Coke yield and proximate analysis of BIOC-HPC coke (db, wt.%) Sample

V

A

FC

Coke yield

0%-BIOC-HPC

4.32

11.82

83.86

70.45

5%-BIOC-HPC

4.19

11.49

84.32

69.55

10%-BIOC-HPC

5.56

11.18

83.26

69.05

15%-BIOC-HPC

5.35

11.06

83.60

68.91

20%-BIOC-HPC

4.74

10.05

85.22

68.42

25

0% 5% 10% 15% 20%

40 30

15

X %

X %

20

50

1100

BIOC-HPC BIOC-HPC BIOC-HPC BIOC-HPC BIOC-HPC

10

1200

BIOC-HPC BIOC-HPC BIOC-HPC BIOC-HPC BIOC-HPC

20 10

5 0

0% 5% 10% 15% 20%

0 0

5

10

15

Time (min)

20

25

30

0

10

20

Time (min)

30

Fig. 2 Relationship between weight loss rate and time of BIOC-HPC coke under different gasification temperatures

increases from 21 to 39%. Considering that the gasification rate of coke through direct reduction reaction in blast furnace is 25–30%. The gasification behavior of coke at 1100 °C is investigated, which is closer to the gasification rate of coke in blast furnace, and the carbonaceous structure of coke is destroyed. It is maintained within a certain limit, which is convenient for later analysis and detection.

The Weight Loss Rate of BIOC-HPC Coke The weight loss rate of BIOC-HPC coke gasification for 30 min is shown in Fig. 3. It can be seen that compared with the coking coal coke (C1), the gasification rate of the coke after adding BIOC-HPC is significantly improved. When the addition ratio of BIOC-HPC is less than 10%, the gasification weight loss rate of the coke increases significantly. Continuing to increase the proportion of BIOC-HPC, the increase in gasification weight loss rate is significantly reduced. Therefore, the addition of 10% of BIOC-HPC is a turning point in the change of coke gasification reaction rate.

258

J. Zhao and X. Wang 50

1100 1200

40

X

30

20

10

0

0%BIOC-HPC

5%BIOC-HPC

10%BIOC-HPC 15%BIOC-HPC 20%BIOC-HPC

Temperature䠄䉝䠅

Fig. 3 Weight loss rate of BIOC-HPC coke gasification for 30 min

The Pore Structure of BIOC-HPC Coke The macroscopic morphologies of coke prepared with different BIOC-HPC ratios after reaction are shown in Fig. 4. At 1100 °C, after 30 min of gasification reaction, a certain amount of white particles appeared on the surface of the crucible coke, and with the increase of the BIOC-HPC ratio, the area of the white particles became larger, and the white particles were the remaining ash after the reaction, which indicated that the consumption of activated carbon on the surface of the coke became larger, and the remaining coke ash after the reaction was deposited on the surface of the coke. Compared with the unreacted coke, the surface of the reacted coke is rougher. With the consumption of the surface carbon matrix, the diameter of the crucible coke decreases significantly. When the addition ratio is 20%, the diameter is the smallest. Compared with the change of coke cross section, as the reaction progresses, the coke gradually appears holes and grooves from the initial relatively flat surface, but the main carbon matrix remains intact, indicating that the gasification reaction mainly occurs in the surface layer. Compared with the unreacted coke, the cross section of the coke has a certain shrinkage, which is mainly related to the consumption of the gasification reaction and the escape of some residual volatiles from the coke pore wall at high temperature. Combined with the effect of BIOC-HPC addition ratio on coke gasification rate, it can be found that the gasification reactivity of BIOC-HPC coke is significantly improved with the increase of BIOC-HPC ratio. In order to further analyze the pore structure of the BIOC-HPC coke, nitrogen adsorption was used to characterize the pore distribution and specific surface area. The BET analysis results of the BIOC-HPC coke before the reaction are shown in

Research on the Gasification Characteristic of Cokes of BIOC-HPC …

259

Fig. 4 Macrostructure images of the BIOC-HPC coke surfaces and cross sections after gasification

Table 3 Specific surface areas and pore parameters of the BIOC-HPC coke Sample

Specific surface area (m2 /g)

Pore volume (cm3 /g)

Pore diameter (nm)

0% BIOC-HPC

0.6006

0.001869

12.6387

5% BIOC-HPC

0.6143

0.001433

11.2043

10% BIOC-HPC

0.6725

0.001203

10.1808

15% BIOC-HPC

0.7564

0.001763

14.5687

20% BIOC-HPC

0.7884

0.00189

19.4586

Table 3. It can be found that with the increase of the content of BIOC-HPC, the specific surface area shows a gradually increasing trend. The main reason is that the BIOC-HPC has good cohesiveness. The inert components adhere to form porous coke, thereby increasing the specific surface area of BIOC-HPC. When the ratio of BIOC-HPC is 20%, the specific surface area reaches a maximum value. The main reason is that due to its high volatile content, the BIOC-HPC causes a large amount of gas to escape during the carbonization process, thereby destroying the formation of colloids. The coke forming ability becomes poor, and a large amount of foam coke is formed. At the same time, a large amount of volatilization was analyzed, which led to the change of the pore size and pore volume of the micropores, and a large number of string holes appeared. From the pore volume and pore size, it can be found that when the ratio is less than 20%, the pore size is relatively uniform, about 10–19 nm.

Conclusions Based on the experimental observations in this work, the following conclusions are drawn:

260

J. Zhao and X. Wang

1. With the increase of BIOC-HPC content, the coking yield decreased after hightemperature carbonization, from 70.45 to 68.42 wt.%, while the ash content of coke is decreased from 11.82 to 10.05 wt.%, 2. With the increase of BIOC-HPC ratio and the gasification temperature, the gasification reaction rate of BIOC-HPC coke gradually increases, and 3. The optimal ratio of BIOC-HPC in coking is in the range of 10–20 wt.%; the pore size is relatively uniform, about 10–19 nm. Acknowledgements This work was financially supported by the National Natural Science Foundation of China (Nos. 51574023, 52074029, and 52174295), Fundamental Research Funds for the Central Universities (FRF-MP-20-018 and 2022QN1012), Outstanding Postdoctoral Program of Jiangsu Province (13013799), and State Key Laboratory of Advanced Metallurgy open funding (KF21-05). Conflict of Interest The authors declare that they have no conflict of interest.

References 1. Xiong G, Li Y, Jin L et al (2015) In situ FT-IR spectroscopic studies on thermal decomposition of the weak covalent bonds of brown coal. J Anal Appl Pyrolysis 83 2. Kashimura N, Takanohashi T, Saito I (2006) Upgrading the solvent used for the thermal extraction of sub-bituminous coal. Energy Fuel 2063 3. Shui H, Zhao W, Shan C et al (2014) Caking and coking properties of the thermal dissolution soluble fraction of a fat coal. Fuel Process Technol 64 4. Takanohashi T, Shishido T, Saito I (2008) Effects of hypercoal addition on coke strength and thermoplasticity of coal blends. Enegy Fuel 1779 5. Rahman M, Samanta A, Gupta R (2013) Production and characterization of ash-free coal from low-rank Canadian coal by solvent extraction. Fuel Process Technol 88 6. Masaki K, Kashimura N, Takanohashi T et al (2005) Effect of pretreatment with carbonic acid on “HyperCoal” (ash-free coal) production from low-rank coals. Energy Fuels 2021 7. Wang YN, Wei XY, Li ZK et al (2017) Extraction and thermal dissolution of Piliqing subbituminous coal. Fuel 282 8. Pan CX, Liu HL, Liu Q et al (2017) Oxidative depolymerization of Shenfu subbituminous coal and its thermal dissolution insoluble fraction. Fuel Process Technol 168 9. Iino M, Takanohashi T, Ohsuga H et al (1988) Extraction of coals with CS2-N-methyl-2pyrrolidinone mixed solvent at room temperature: effect of coal rank and synergism of the mixed solvent. Fuel 1639 10. Do Kim S, Woo KJ, Jeong SK et al (2008) Production of low ash coal by thermal extraction with N-methyl-2-pyrrolidinone, Korean J Chem Eng 758 11. Zhao J, Mangi HN, Zhang Z et al (2022) The structural characteristics and gasification performance of cokes of modified coal extracted from the mixture of low-rank coal and biomass. Energy 124864 12. Zhao J, Zuo H, Wang G et al (2019) Improving the coke property through adding HPC extracted from the mixture of low-rank coal and biomass. Energy Fuels 1802 13. Fischer JR (1983) Solvent extraction of coal. Academic, p 173 14. Renganathan K, Zondlo JW, Mintz EA et al (1988) Preparation of an ultra-low ash coal extract under mild conditions. Fuel Process Technol 18:273

Thermodynamic Examination of Selected Phases in the Ag–Co–Sn–S System at T < 600 K by the Solid-State EMF Method Mykola Moroz, Fiseha Tesfaye, Pavlo Demchenko, Myroslava Prokhorenko, Oksana Mysina, Lyudmyla Soliak, Daniel Lindberg, Oleksandr Reshetnyak, and Leena Hupa Abstract Phase equilibria in the part SnS–SnS2 –CoS2 –CoS–Ag2 CoS2 –SnS of the Ag–Co–Sn–S system at T < 600 K were investigated by the modified solid-state electromotive force (EMF) method. The position of established phase regions vs point of Ag was used to express the overall potential-forming reactions. The reactions were performed in the positive electrodes of the electrochemical cells (ECCs). The positive electrodes of ECCs were prepared from carefully mixed non-equilibrium powder mixtures of Ag, Ag2 S, CoS2 , Co3 S4 , Co9 S8 , SnS2 , Sn2 S3 , and SnS. Synthesis of the equilibrium set of phases in the positive electrodes of the ECCs was facilitated by Ag+ ions that shifted from the left electrode and acted as the small nucleation centers of formation of the compounds. Linear dependences of the EMF of the ECCs on temperature were used for calculating standard Gibbs energies, enthalpies, and entropies of formations of the compounds CoS, AgCoS2 , Ag2 CoS2 , Ag2 CoSnS4 , and Ag2 CoSn3 S8 . The thermodynamic data obtained in the present study were compared and analyzed in detail. M. Moroz (B) · O. Mysina · L. Soliak Department of Chemistry and Physics, National University of Water and Environmental Engineering, Rivne 33028, Ukraine e-mail: [email protected] F. Tesfaye · L. Hupa Johan Gadolin Process Chemistry Centre, Åbo Akademi University, 20500 Turku, Finland F. Tesfaye Metso Outotec Finland Oy, 02231 Espoo, Finland P. Demchenko Department of Inorganic Chemistry, Ivan Franko National University of Lviv, Lviv 79005, Ukraine M. Prokhorenko Department of Cartography and Geospatial Modeling, Lviv Polytechnic National University, Lviv 79013, Ukraine D. Lindberg Department of Chemical and Metallurgical Engineering, Aalto University, 02150 Espoo, Finland O. Reshetnyak Department of Physical and Colloid Chemistry, Ivan Franko National University of Lviv, Lviv 79005, Ukraine © The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0_26

261

262

M. Moroz et al.

Keywords Ag-based compounds · Phase equilibria · Thermodynamic properties · EMF method · Gibbs energy

Introduction The phase composition of the equilibrium T–x space of a great number of complex inorganic systems, formed with the participation of chalcogenides and halides of transition metals, below the conventional temperature of 600 K, remains unknown. The reason for this is the insufficient energy of the thermal motion of the atoms and ions for the formation of the thermodynamically stable set of phases. The action of such external factors as long-term annealing during temperature and pressure variations is ineffective in many cases. The possibility of overcoming such kinetic obstacles was established in [1–3]. For this purpose, the Ag+ ions were used as the small nucleation centers for the formation of an equilibrium set of phases. The literature data about compounds of Ag–Co–Sn–S system are very scarce and relate mainly to the phase composition and thermodynamic properties of binary compounds of the Co–S cross-section, in particular CoS2 , Co3 S4 , Co1–x S, Co9 S8 , and Co4 S3 [4, 5]. Only one article by Padiou et al. [6] provide the information about the crystal structure, theoretical and experimental density, and Mossbauer effect of the Ag2 CoSn3 S8 compound. Spinel-type compounds of the formula composition A2 MSn3 S8 (A = Li, Cu, Ag; M = Mn, Fe, Co, Ni) are promising materials for use in solar cell and thermoelectric applications [7, 8]. Furthermore, the complex Ag-based phases are also expected to be good superionic conductors [9]. The purpose of this work was to establish the equilibrium phase composition of the Ag–Co–Sn–S system in the SnS–SnS2 –CoS2 –CoS–Ag2 CoS2 –SnS part below 600 K and to calculate the values of standard thermodynamic functions (Gibbs energy, enthalpy, and entropy) of selected compounds by the EMF method. This method with solid and liquid electrolyte was successfully used for thermodynamic study of different multicomponent systems [10–12].

Experimental In this study, the high-purity (> 99.9 wt%) silver and binary compounds Ag2 S, AgBr, CoS2 , Co3 S4 , Co9 S8 , GeS2 , SnS2 , Sn2 S3 , and SnS were used. Crushed to a particle size of ~ 5 μm, polycrystalline samples of the binary compounds were used for Xray diffraction data and preparation of positive electrodes of electrochemical cells (ECCs). An STOE STADI P diffractometer equipped with a linear position-sensitive detector PSD, in a Guinier geometry (transmission mode, CuKα1 radiation, a bent Ge(111) monochromator, and 2θ/ω scan mode), was used to establish the phase composition of the samples. The following programs STOE WinXPOW (version

Thermodynamic Examination of Selected Phases in the Ag–Co–Sn–S …

263

3.03) [13], PowderCell (version 2.4) [14], FullProf.2k (version 5.60) [15] as well as databases [16, 17] were used for X-ray diffraction data. Synthesis of a thermodynamical equilibrium set of compounds below 600 K from the phase non-equilibrium mixtures of silver and binary compounds and subsequent EMF (E) measurements were performed in ECCs type (A):    (−)C|Ag|SER Ag+ PE|C(+),

(A)

where C is the graphite (inert electrode), Ag is the left (negative) electrode, SE is the solid-state electrolyte (Ag3 GeS3 Br glass [18]), PE is the right (positive) electrode, R(Ag+ ) is the buffer region of PE that contacts with SE. The ratios of components of the PE of ECCs were determined from the equations of potential-forming reactions in respective phase regions. Components of the ECCs were pressed at 108 Pa through a 2 mm diameter hole arranged in a fluoroplastic matrix up to density ρ = (0.93 ± 0.02) · ρ 0 , where ρ 0 is the experimentally determined density of cast samples [19, 20]. The process of forming the thermodynamically stable set of phases from a non-equilibrium phase mixture of finely dispersed components was carried out in the R(Ag+ ) region. The Ag+ ions acted as small nucleation centers for stable phases [2]. The experiments were performed in a horizontal resistance furnace, similar to that described in [21]. As the protection atmosphere, we used a flow of highly purified (99.99 volume fraction) Ar (g) at P = 1.2 × 105 Pa. The gas flow of Ar at the rate of 10–5 m3 min–1 had direction from the left to the right electrodes of the ECCs. The temperature was maintained with an accuracy of ± 0.5 K. The EMF of the cells were measured using high-resistance (input impedance of > 1012 ) the Picotest M3500A universal digital multimeter. The equilibrium in the ECCs at each temperature was achieved within 2 h. During equilibrium the EMF values were constant, or their variations were not exceeding ± 0.2 mV [22, 23].

Results and Discussion The phase equilibria of the Ag–Co–Sn–S system in the part SnS–SnS2 –CoS2 –CoS– Ag2 CoS2 –SnS are shown in Fig. 1. The mentioned area was divided into separate phase regions according to the data reported in [24–28] and by the EMF and X-ray diffraction methods presented in this study. According to the X-ray diffraction results, the sample with the nominal composition Ag2 CoSn3 S8 can be described as a practically single-phase with the reflection lines for the compound Ag2 CoSn3 S8 (AgCo0.5 Sn1.5 S4 ) [6]: space group I41 /a, with unit-cell parameters a = 7.487 and c = 10.592 Å, Fig. 2. The cooled melts of the nominal formula compositions AgCoS2 and Ag2 CoS2 contained characteristic peaks for the Ag2 S, Co3 S4 , and CoS2 compounds, Fig. 3. The cooled melt of the Ag2 CoSnS4 composition consists a mixture of the Co3 S4 , Ag8 SnS6 compounds, and adulterant of the Co1–x S phase. Solid-state vacuum homogenizing annealing at 600 K

264

M. Moroz et al.

Fig. 1 Phase equilibria of the Ag–Co–Sn–S system in the part SnS–SnS2 –CoS2 –CoS–Ag2 CoS2 – SnS below 600 K

did not change the phase composition of the specified cooled melts. The combination of binary and ternary Ag8 SnS6 compounds in these alloys is considered a metastable state of the samples for kinetic reasons.

Fig. 2 X-ray powder diffraction pattern of the sample with nominal composition Ag2 CoSn3 S8 . Compositions of the sample and identified phase (with space group indicated) are shown in the upper right corner

Thermodynamic Examination of Selected Phases in the Ag–Co–Sn–S …

265

Fig. 3 X-ray powder diffraction pattern of the sample with nominal composition AgCoS2 . Compositions of the sample and identified phases (with space group indicated) are shown in the upper right corner

In the SnS–SnS2 –CoS2 –CoS–Ag2 CoS2 –SnS phase area, the following two-, three-, and four-phase regions were established: CoS2 –AgCoS2 (I), AgCoS2 –Ag2 CoS2 (II), SnS2 –AgCoS2 –Ag2 CoSn3 S8 (III), Ag2 CoSn3 S8 –AgCoS2 – Ag2 CoSnS4 (IV), Sn2 S3 –CoS–AgCoSn3 S8 –SnS (V), Ag2 CoSnS4 –CoS–AgCoS2 – SnS (VI), and SnS2 –CoS–Ag2 CoSn3 S8 –Sn2 S3 (VII). The correctness of the presented division was confirmed by the following calculations of the thermodynamic quantities of the compounds in different phase regions. In accordance with Fig. 1, the electrochemical processes of synthesis of compounds in the PE of ECCs of the phase regions (I)–(VII) can be expressed by the overall cell reactions (R1)–(R7), respectively: Ag + CoS2 = AgCoS2 ,

(R1)

Ag + AgCoS2 = Ag2 CoS2 ,

(R2)

Ag + 3SnS2 + AgCoS2 = Ag2 CoSn3 S8 ,

(R3)

2Ag + Ag2 CoSn3 S8 + 2AgCoS2 = 3Ag2 CoSnS4 ,

(R4)

2Ag + 4Sn2 S3 + CoS = Ag2 CoSn3 S8 + 5SnS,

(R5)

2Ag + Ag2 CoSnS4 + CoS = 2Ag2 CoS2 + SnS,

(R6)

266

M. Moroz et al.

2Ag + 5SnS2 + CoS = Ag2 CoSn3 S8 + Sn2 S3 .

(R7)

Based on the reactions (R1)–(R7), the compositions of the phase non-equilibrium mixtures of the dispersed silver and binary silver, cobalt, and tin sulfides in the PE of ECCs of the phase regions (I)–(VII) were determined. The process of forming the thermodynamically stable set of phases from phase non-equilibrium mixture of components for the participation of Ag+ ions as a catalyst ends in 48 h at 600 K. The criterion for attaining phase equilibria in the R(Ag+ ) region of PE is the reproducibility of the E versus T relations of ECCs during the heating–cooling cycles. The dependences E versus T of ECCs are shown in Fig. 4. Linear equations for E(T ), calculated by applying the methodology described in [29, 30], are listed in Table 1. In these equations, the term after sign ‘±’ is the expanded uncertainty for E with 95% level of confidence. The Gibbs energies, enthalpies, and entropies of reactions (R1)–(R7) can be calculated by applying the thermodynamic Eqs. (1)–(3): r G = −z F E,

(1)

r H = −z F[E − (dE/dT )T ],

(2)

r S = z F(dE/dT )

(3)

where z is the number of electrons involved in the reactions (R1)–(R7), F = 96,485.33289 C mol–1 is Faraday’s constant, and E is the EMF of ECCs.

Fig. 4 Dependences E versus T of the ECCs with PE of the phase regions (I)–(VII)

Thermodynamic Examination of Selected Phases in the Ag–Co–Sn–S …

267

Table 1 EMF versus temperature relations for the type (A) ECCs in the (I)–(VII) phase regions of the Ag–Co–Sn–S system: a is the intercept, b is the slope, k St is the Student’s coefficient, u 2E and u 2b are statistical dispersions of the values E and b, n is the number of experimental pairs Ti and E i , and T is the average temperature in the measured range     u 2E 2 T −T 2 Phase region E = a + bT ± k St + u b n (I)



1.1910−3 E = 54.77 + 230.4910−3 T ± 2.16 + 2.7110−7 (T − 479.72)2 13

(II)



6.9010−4 E = 69.12 + 153.1010−3 T ± 2.16 + 1.5710−7 (T − 489.55)2 13

(III)



3.1010−3 E = 167.63 + 193.5910−3 T ± 2.16 + 7.0210−7 (T − 479.72)2 13

(IV)



2.8210−3 E = 68.90 + 342.2710−3 T ± 2.16 + 6.4010−7 (T − 479.72)2 13

(V)



1.0410−2 E = 131.60 + 273.4610−3 T ± 2.16 + 2.3610−6 (T − 479.72)2 13

(VI)



1.2010−2 −6 (T − 484.64)2 E = 100.75 + 163.7110−3 T ± 2.16 + 2.7210 13

(VII)

E=

144.51 + 297.9110−3 T



1.1510−2 ± 2.16 + 2.6210−6 (T − 479.72)2 13

The values of thermodynamic functions of the reactions (R1)–(R7) calculated by Eqs. (1)–(3) at 298 K and p = 105 Pa are listed in Table 2. The Gibbs energy, enthalpy, and entropy of reaction (R1) are related to the Gibbs energy, enthalpy, and entropy of the AgCoS2 compound and pure Ag by Eqs. (4)–(6): r(R1) G ◦ = f G ◦AgCoS2 − f G ◦CoS2 ,

(4)

Table 2 Values of standard thermodynamic functions of the reactions (R1)–(R7)a Reaction

−r G ◦ (kJ mol–1 )

−r H ◦ (kJ mol–1 )

r S ◦ [J (mol K)−1 ]

(R1)

11.91 ± 0.02

5.28 ± 0.05

22.24 ± 0.11

(R2)

11.07 ± 0.02

6.67 ± 0.04

14.77 ± 0.08

(R3)

21.74 ± 0.03

16.17 ± 0.08

18.68 ± 0.17

(R4)

32.98 ± 0.06

13.30 ± 0.16

66.05 ± 0.33

(R5)

41.12 ± 0.12

25.39 ± 0.31

52.77 ± 0.64

(R6)

28.86 ± 0.13

19.44 ± 0.33

31.59 ± 0.69

(R7)

45.02 ± 0.12

27.89 ± 0.32

57.49 ± 0.67

a

Uncertainties for r

G◦,

r

H ◦,

and r

S◦

are standard uncertainties

268

M. Moroz et al. ◦ ◦ r(R1) H ◦ = f HAgCoS − f HCoS , 2 2

(5)

◦ ◦ ◦ r(R1) S ◦ = SAgCoS − SAg − SCoS . 2 2

(6)

It follows from Eqs. (4)–(6) that: f G ◦AgCoS2 = f G ◦CoS2 + r(R1) G ◦ ,

(7)

◦ ◦ f HAgCoS = f HCoS + r(R1) H ◦ , 2 2

(8)

◦ ◦ ◦ SAgCoS = SAg + SCoS + r(R1) S ◦ . 2 2

(9)

Similarly, the corresponding equations to determine f G ◦ , f H ◦ , and S ◦ of the Ag2 CoS2 , Ag2 CoSn3 S8 , Ag2 CoSnS4 , CoS, CoS, and Ag2 CoSn3 S8 compounds in the phase regions (II)–(VII) can be written based on reactions (R2)–(R7), respectively. For the first time, the standard thermodynamic quantities for selected compounds of the Ag–Co–Sn–S system were calculated by using Eqs. (7)–(9) and the thermodynamic data of the pure substances Ag, Co, Sn, S, CoS2 , SnS, SnS2 , and Sn2 S3 [31]. A comparative summary of the calculated values and the available literature data are listed in Table 3. The temperature dependences of the Gibbs energies of formation of the binary, ternary, and quaternary compounds in the respective phase regions are described by Eqs. (10)–(16):   f G CoS,(V) / kJ mol−1 = −(95.2 ± 3.3) + (15.4 ± 1.1)10−3 T /K,

(10)

  f G CoS,(VI) / kJ mol−1 = −(96.9 ± 3.7) + (11.6 ± 0.9)10−3 T /K,

(11)

  f G AgCoS2 ,(I) / kJ mol−1 = −(158.4 ± 3.9) + (2.9 ± 0.1)10−3 T /K,

(12)

  f G Ag2 CoS2 ,(II) / kJ mol−1 = −(165.1 ± 4.5) − (11.9 ± 0.5)10−3 T /K,

(13)

  f G Ag2 CoSnS4 ,(IV) / kJ mol−1 = −(321.8 ± 6.3) + (2.5 ± 0.1)10−3 T /K,

(14)

  f G Ag2 CoSn3 S8 ,(III) / kJ mol−1 = −(635.2 ± 10.2) + (67.8 ± 2.1)10−3 T /K, (15)   f G Ag2 CoSn3 S8 ,(VII) / kJ mol−1 = −(627.3 ± 10.6) + (63.1 ± 2.0)10−3 T /K. (16)

Thermodynamic Examination of Selected Phases in the Ag–Co–Sn–S …

269

Table 3 Values of standard thermodynamic functions of selected compounds of the Ag–Co–Sn–S system at 298 Ka Phase

Phase region

−f G ◦ (kJ mol−1 )

−f H ◦ (kJ mol−1 )

S ◦ [J (mol K)−1 ] References

Ag

–

0

0

42.677

[31]

Co

–

0

0

30.041

[31]

Sn

–

0

0

51.195

[31]

S

–

0

0

32.056

[31]

CoS2

–

145.645

153.134

69.036

[31]

SnS

–

106.079

107.947

76.986

[31]

SnS2

–

145.246

153.553

87.446

[31]

Sn2 S3

–

253.417

263.592

164.431

[31]

AgCoS2

(I)

157.6 ± 2.4

158.4 ± 3.9

134.0 ± 6.2

This work

Ag2 CoS2

(II)

168.6 ± 2.8

165.1 ± 4.5

191.4 ± 8.7

This work

Ag2 CoSn3 S8

(III)

615.0 ± 8.9

635.2 ± 10.2

457.6 ± 14.1

This work

Ag2 CoSnS4

(IV)

321.0 ± 5.4

321.8 ± 6.3

292.3 ± 4.9

This work

CoS

(V)

90.6 ± 1.8

95.2 ± 3.3

46.7 ± 3.2

This work

CoS

(VI)

93.4 ± 2.0

96.9 ± 3.7

50.5 ± 3.6

This work

Ag2 CoSn3 S8

(VII)

608.5 ± 9.1

627.3 ± 10.6

462.4 ± 14.7

This work

a

Uncertainties for f G ◦ , f H ◦ , and S ◦ are standard uncertainties

It follows from Table 3 that calculated values of the Gibbs energy of the CoS in the phase regions (V) and (VI) as well as Ag2 CoSn3 S8 compound in the (III) and (VII) are convergence within the experiment errors (the relative differences are ~ 3% and ~ 1%, respectively). It validates: (1) phase composition and division of the equilibrium concentration space of the Ag–Co–Sn–S systems in the part SnS–SnS2 –CoS2 –CoS–Ag2 CoS2 –SnS; (2) calculated values of the thermodynamic quantities of the compounds; (3) reliability of the literature values of the thermodynamic properties of the CoS2 , SnS, SnS2 , and Sn2 S3 compounds; and (4) negligible homogeneity region of the CoS and Ag2 CoSn3 S8 compounds.

Conclusions Phase equilibria in the part SnS–SnS2 –CoS2 –CoS–Ag2 CoS2 –SnS of the Ag–Co– Sn–S system below 600 K were investigated using the EMF method. The formation of the ternary AgCoS2 , Ag2 CoS2 , and quaternary Ag2 CoSnS4 compounds has been established for the first time. The synthesis of the equilibrium set of phases, including mentioned compounds, was carried out in the R(Ag+ ) region of the

270

M. Moroz et al.

electrochemical cells. The Ag+ ions act as the small nucleation centers and catalysts for forming the equilibrium set of phases. The measured EMF values of the electrochemical cells have been applied to calculate the standard Gibbs energies, enthalpies, and entropies of formations of the CoS, AgCoS2 , Ag2 CoS2 , Ag2 CoSnS4 , and Ag2 CoSn3 S8 compounds. The correctness of the proposed division of the Ag– Co–Sn–S system into separate phase regions and the determined thermodynamic quantities of the binary, ternary, and quaternary compounds are confirmed with the values of Gibbs energies of the CoS and Ag2 CoSn3 S8 compounds calculated in two different phase regions. Acknowledgements This research was supported by the national projects of the Ministry of Education and Science of Ukraine: “Scientific and experimental basis for the production of composite oxide, chalcogenide materials with extended service life” (No. 0121U109620) and “Synthesis, physicochemical and thermodynamic properties of nano sized and nanostructured materials for electrochemical systems” (No. 0120U102184). This work was partly funded by the Swedish Cultural Foundation in Finland under the project “Innovative e-waste recycling processes for greener and more efficient recoveries of critical metals and energy” at Åbo Akademi University. Conflict of Interest The authors declare that they have no conflict of interest.

References 1. Moroz M, Tesfaye F, Demchenko P et al (2020) Solid-state electrochemical synthesis and thermodynamic properties of selected compounds in the Ag–Fe–Pb–Se system. Solid State Sci 107:106344(1)–(9). https://doi.org/10.1016/j.solidstatesciences.2020.106344 2. Moroz M, Tesfaye F, Demchenko P et al (2021) Non-activation synthesis and thermodynamic properties of ternary compounds of the Ag–Te–Br system. Thermochim Acta 698:178862(1)– (7). https://doi.org/10.1016/j.tca.2021.178862 3. Moroz M, Tesfaye F, Demchenko P et al (2021) The equilibrium phase formation and thermodynamic properties of functional tellurides in the Ag–Fe–Ge–Te system. Energies 14:1314(1)–(15). https://doi.org/10.3390/en14051314 4. Masset PJ, Guidotti RA (2008) Thermal activated (“thermal”) battery technology. J Power Sources 178:456–466. https://doi.org/10.1016/j.jpowsour.2007.11.073 5. Chen YO, Chang YA (1978) Thermodynamics and phase relationships of transition metalsulfur systems: I. The cobalt-sulfur system. Metall Trans B 9:61–67. https://doi.org/10.1007/ BF02822672 6. Padiou J, Jumas JC, Ribes M (1981) Sur une nouvelle famille de composes quaternaires MM 0.5Sn1.5S4 (M=Cu, Ag; M = Mn, Fe Co, Ni) de type spinelle: caracterisation par spectroscopie Mossbauer de 119Sn et proprietes magnetiques. Rev Chim Miner 18:33–42 7. Heppke EM, Mahadevan S, Lerch M (2020) New compounds of the Li2 MSn3 S8 type. Z Naturforsch B 75:625–631. https://doi.org/10.1515/znb-2020-0050 8. Quintero MA, Hao S, Patel SV et al (2021) Lithium thiostannate spinels: air-stable cubic semiconductors. Chem Mater 33:2080–2089. https://doi.org/10.1021/acs.chemmater.0c04651 9. Studenyak IP, Pogodin AI, Studenyak VI et al (2020) Electrical properties of copper- and silver-containing superionic (Cu1−x Agx )7 SiS5 I mixed crystals with argyrodite structure. Solid State Ionics 345:115183. https://doi.org/10.1016/j.ssi.2019.115183 10. Vassiliev VP, Lysenko VA, Bros JP (2019) Thermodynamic study of the Ag-In-Sn system by the EMF method. J Alloys Compd 790:370–376. https://doi.org/10.1016/j.jallcom.2019.03.016

Thermodynamic Examination of Selected Phases in the Ag–Co–Sn–S …

271

11. Babanly NB, Orujlu EN, Imamaliyeva SZ et al (2019) Thermodynamic investigation of silverthallium tellurides by EMF method with solid electrolyte Ag4 RbI5 . J Chem Thermodyn 128:78– 86. https://doi.org/10.1016/j.jct.2018.08.012 12. Imamaliyeva SZ, Musayeva SS, Babanly DM et al (2019) Determination of the thermodynamic functions of bismuth chalcoiodides by EMF method with morpholinium formate as electrolyte. Thermochim Acta 679:178319(1)–(7). https://doi.org/10.1016/j.tca.2019.178319 13. Diffractometer Stoe WinXPOW, version 303, Stoe & Cie GmbH, Darmstadt (2010) 14. Kraus W, Nolze G (1996) POWDER CELL—a program for the representation and manipulation of crystal structures and calculation of the resulting X-ray powder patterns. J Appl Crystallogr 29:301–303. https://doi.org/10.1107/S0021889895014920 15. Rodriguez-Carvajal J (2001) Recent developments of the program FULLPROF, IUCr Commission on Powder Diffraction. Newsletter 26:12–19 16. Downs RT, Hall-Wallace M (2003) The American Mineralogist crystal structure database. Am Mineral 88:247–250 17. Villars P, Cenzual K (2014) Pearson’s crystal data: crystal structure database for inorganic compounds, release 2014/15. ASM International, Materials Park, OH, USA 18. Moroz MV, Prokhorenko MV, Prokhorenko SV (2015) Determination of thermodynamic properties of Ag3 SBr superionic phase using EMF technique. Russ J Electrochem 51:886–889. https://doi.org/10.1134/S1023193515090098 19. Prokhorenko MV, Moroz MV, Demchenko PYu (2015) Measuring the thermodynamic properties of saturated solid solutions in the Ag2 Te-Bi-Bi2 Te3 system by the electromotive force method. Russ J Phys Chem A 89:1330–1334. https://doi.org/10.1134/S0036024415080269 20. Moroz MV, Demchenko PYu, Prokhorenko MV, Reshetnyak OV (2017) Thermodynamic properties of saturated solid solutions of the phases Ag2 PbGeS4 , Ag0.5 Pb1.75 GeS4 and Ag6.72 Pb0.16 Ge0.84 S5.20 of the Ag-Pb-Ge-S system determined by EMF method. J Phase Equilib Diffus 38:426–433. https://doi.org/10.1007/s11669-017-0563-6 21. Tesfaye F, Taskinen P (2014) Electrochemical study of the thermodynamic properties of matildite (β-AgBiS2 ) in different temperature and compositional ranges. J Solid State Electrochem 18:1683–1694. https://doi.org/10.1007/s10008-014-2395-1 22. Moroz MV, Prokhorenko MV, Rudyk BP (2014) Thermodynamic properties of phases of the Ag-Ge-Te system. Russ J Electrochem 50:1177–1181. https://doi.org/10.1134/S10231935141 20039 23. Moroz M, Tesfaye F, Demchenko P et al (2018) Phase equilibria and thermodynamics of selected compounds in the Ag–Fe–Sn–S system. J Electron Mater 47:5433–5442. https://doi. org/10.1007/s11664-018-6430-3 24. Fedorchuk AO, Zhbankov OYe, Lakshminarayana G et al (2012) Synthesis and spectral features of Ag2 SnS3 crystals. Mater Chem Phys 135:249–253. https://doi.org/10.1016/j.matchemphys. 2012.04.025 25. Aliyeva ZM, Bagheri SM, Aliev ZS et al (2014) The phase equilibria in the Ag2 S–Ag8 GeS6 – Ag8 SnS6 system. J Alloys Compd 611:395–400. https://doi.org/10.1016/j.jallcom.2014.05.112 26. Ghrib T, Al-Otaibi AL, Almessiere MA et al (2015) High thermoelectric figure of merit of Ag8 SnS6 component prepared by electrodeposition technique. Chinese Phys Lett 32:127402(1)–(5). https://doi.org/10.1088/0256-307X/32/12/127402 27. Moroz M, Tesfaye F, Demchenko P et al (2018) Thermodynamic properties of magnetic semiconductors Ag2 FeSn3 S8 and Ag2 FeSnS4 determined by the EMF method. In: Lambotte G, Lee J, Allanore A, Wagstaff S (eds) Materials processing fundamentals 2018. Springer, Cham, pp 87–98. https://doi.org/10.1007/978-3-319-72131-6_8 28. Moroz M, Reshetnyak O, Demchenko P et al (2020) Thermodynamic properties of silvercontaining compounds of the Ag–Fe–Sn–S system obtained by low-temperature solid-state synthesis. Ukr Chem J 86:34–50. https://doi.org/10.33609/2708-129X.86.11.2020.34-50 [in Ukrainian] 29. Hasanova GS, Aghazade AI, Imamaliyeva SZ et al (2021) Refinement of the phase diagram of the Bi-Te system and the thermodynamic properties of lower bismuth tellurides. JOM 73:1511– 1521. https://doi.org/10.1007/s11837-021-04621-1

272

M. Moroz et al.

30. Hasanova GS, Aghazade AI, Babanly DM et al (2022) Experimental study of the phase relations and thermodynamic properties of Bi-Se system. J Therm Anal Calorim 147:6403–6414. https:// doi.org/10.1007/s10973-021-10975-0 31. Barin I (1995) Thermochemical data of pure substances. VCH, Weinheim

Author Index

A Abrell, Peter, 181 Alam, Shafiq, 89, 193, 213 Ali, Firdos, 27 Amroussia, Aida, 15 Arkhurst, Barton, 223

G Gamage McEvoy, Joanne, 81 Gilston, Phil, 15 Gudjonsdottir, Maria, 139 Guo, Ruiran, 223 Gupta, Subhadra, 27

B Babu, Suresh, 15 Bastow, Heath, 171 Brennan, Patrick, 15

H Hanebeck, Claus, 181 Han, Guihong, 233 Haque, Nawshad, 3, 107 Hara, Rainford, 97 Hara, Ronald, 97 Hara, Yotamu Rainford Stephen, 97 Hart, Judy N., 39 Hauser, A. J., 27 Hendy, Mohamed, 119 Huang, Yanfang, 233 Hupa, Leena, 261

C Chan, Sammy Lap Ip, 223 Chen, Shaoguo, 243

D Dayalu, Satyendra, 3 Demchenko, Pavlo, 261 Dheeradhada, Voramon, 15 DiBenedetto, Sarah, 139 Dreisinger, David, 63 Duguay, Dominique, 81

E Estrada, David, 129

F Fleming, Austin, 129 Folstad, M. B., 201

J Jusnes, K. F., 201

K Kadivar, Saeede, 149 Kebbede, Anteneh, 15

L Law, Ka Ming, 27 Li, Dawen, 27 Lindberg, Daniel, 261 Liu, Bingbing, 233 Liu, Yu, 51

© The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0

273

274 Li, Xiao, 27 Lombardo, Steve, 15 Lota, Amanda, 171 Lv, Binbin, 51

M Ma, Huaiying, 243 Mairizal, A. Q., 107 Malek, Ali, 119 Masse, Nicholas, 171 Miao, Zhenping, 243 Moroz, Mykola, 261 Morra, Martin, 15 Mshar, Alecsander D., 27 Musukwa, Ireen, 97 Mysina, Oksana, 261

N Namiluko, Yaki Chiyokoma, 97 Namq, Valan, 213 Nedeljkovic, Dragutin, 159 Ngomba, Makwenda Thelma, 97

O Old, Alexander, 97 Oppong-Antwi, Louis, 39 Orhan, Okan K., 119

P Pagan, Michael, 15 Pan, Shanlin, 27 Pan, Wen, 243 Parirenyatwa, Stephen, 97 Paul, Akshoy Ranjan, 3 Penney, Sara, 193 Pint, Bruce, 15 Ponga, Mauricio, 119 Powell, Adam, 171 Prokhorenko, Myroslava, 261

R Reek, Till, 181 Reiser, Wolfgang, 181 Reshetnyak, Oleksandr, 261 Rhamdhani, M. A., 107 Rokh, Ghazaleh Bahman, 223 Russell, Tamara, 15

Author Index S Saevarsdottir, Gudrun, 139 Sembada, A. Y., 107 Shahabi, Mahya, 171 Shane, Agabu, 97 Shin, Homin, 119 Shower, Patrick, 15 Soliak, Lyudmyla, 261 Su, Shengpeng, 233

T Tangstad, M., 201 Tesfahunegn, Yonatan A., 139 Tesfaye, Fiseha, 261 Thibault, Yves, 81 Tse, K. M., 107

V Vahidi, Ehsan, 149 Velasquez, Brandon, 139 Verma, Shalini, 3

W Wada, Katelyn, 129 Wallace, Lucien, 171 Wang, Fei, 63 Wang, Jingsong, 51 Wang, Xueya, 251 Weaver, Scott, 15 Wimmer, Alexander, 75

X Xue, Qingguo, 51

Y Yang, Fan, 51 Yang, Ze, 233

Z Zagrtdenov, Nail R., 81 Zhang, Yapeng, 243 Zhao, Jun, 251 Zhao, Zhixing, 243 Zhu, Jianbo, 243 Zuo, Haibin, 51

Subject Index

A Adsorption, 121, 160–162, 216, 225, 227, 229, 230, 258 Ag-based compounds, 262 Alchemical perturbation, 119, 120, 124 Alkali-based decomposition, 81 Aluminium, 3, 5, 8, 9, 12, 184, 186 Aluminium smelting, 11 Australia, 107–109, 150, 190, 215

B Beckenbach annular lime kiln, 243 BIOC-HPC, 251, 253–260 Biomass, 4, 5, 15, 19, 78, 251, 253, 254 Black copper smelting, 108–110, 112, 115, 117 Busbar, 184–191

C Carbon capture, 15, 75, 139, 145 Carbon footprint, 4, 75–78, 107–109, 111, 112, 114, 116, 117 Carbon membrane, 164, 165 Catalysis, 120, 121 Charge surface, 201–208, 210 Circular economy, 3–5 Clean energy, 52, 172, 193, 214 Clean energy transition, 63, 64, 73 CO2 emission, 63, 64, 110, 112, 113, 117, 202 CO2 mineralization, 63–73 CO2 reduction, 39 Coal ash, 193–195, 197 Cobalt, 63–69, 71–73, 97, 98, 100–104, 266 CO concentration, 52, 56–58

Condensation, 160, 205, 206, 210 Cost, 9, 15–20, 64, 69, 77, 91, 92, 119, 120, 123, 129, 130, 150, 160, 171, 172, 174, 176, 178, 181–191, 224, 233, 247, 252, 253 Critical battery metal recovery, 69, 70 CuS, 39–47, 93, 95, 102–104, 107, 108, 110, 112, 114, 115, 117, 262

D Dense membranes, 161, 163 Density Functional Theory (DFT), 39, 40, 47, 119–124 Desorption, 161, 225, 227–230 Directional solidification, 178

E Efficiency, 15, 27, 28, 40, 65–70, 72, 78, 129, 130, 161, 162, 171, 172, 177, 184, 185, 189, 244, 245, 252, 253 Ejector, 243–249 Electric vehicles (EV), 64, 173, 214, 216 Electrocatalytic, 233, 238, 239 Electrodeposition, 233–235, 240 EMF method, 262, 269 Energy, 3–7, 9, 15, 21–23, 33, 39–44, 46, 47, 51, 54, 57, 77, 79, 81–83, 86, 89–95, 100, 108, 110–113, 117, 119–122, 124, 130, 139, 141, 144, 150, 152, 155, 159–162, 171–174, 177, 178, 181–183, 185, 187–191, 193, 194, 214–216, 223, 224, 229, 233–235, 237, 243, 244, 248, 252, 262, 270

© The Minerals, Metals & Materials Society 2023 S. Alam et al. (eds.), Energy Technology 2023, The Minerals, Metals & Materials Series, https://doi.org/10.1007/978-3-031-22638-0

275

276 Energy-saving and cost-reducing, 89, 90, 93, 95, 216, 243 Energy storage, 171, 213, 216 Environment impact, 3–6, 9–12, 52, 107, 109, 110, 130, 150, 151, 154–156 Environmental effects, 139 Electric Vehicles (EV), 214, 216

F Flame length, 51–55, 57 Flame temperature, 51–55, 57 Flexible packaging, 75 Flue gas treatment, 159, 160, 165

G Gasification, 251–255, 257–260 Gibbs energy, 261, 262, 266–270 Gold processing, 149 Gold production, 149–152, 155, 156 G8 PAMAM dendrimer, 34, 37 Greenhouse gas (GHG), 52, 89, 90, 92, 95, 111, 113, 115, 117, 150, 154, 156, 173, 181, 182, 187, 189, 193, 215

H Heterostructures, 39–44, 46, 47 High-entropy alloys, 119–121 Hydrogen, 40, 52, 54, 56, 57, 99, 113, 116, 117, 120, 161, 165, 172, 181, 182, 188–190, 223–225, 228, 230, 238 Hydrogen storage, 223, 224, 226, 228–230 Hydrometallurgy, 89, 90, 93, 233

I Indium Tin Oxide (ITO), 27–37 Inorganic membrane, 165

L Life Cycle Assessment (LCA), 4–6, 12, 149, 151–153, 156 Line heat source, 129, 134 Lithium, 63, 81, 82, 84–86, 171–173, 213–218 Low-rank coal, 251–253 Luanshya chromium, 98

M Metal-air battery, 172

Subject Index Mineral processing, 89, 91, 92 Mining, 64, 89–92, 149–151, 153, 213, 215 Mixed-matrix membranes, 159, 166, 167 Modeling and simulation (CFD), 139, 140

N Nanotubes, 223–230 NetZero, 89, 95, 216

O Olivine and laterites, 63 Overpotential, 233, 234, 238 Oxygen evolution, 233–235, 238 Oxygen reduction, 174

P Phase equilibria, 261, 263, 264, 266, 269 Pollutant emissions, 10 Process optimization, 195

R Rare Earth Elements (REEs), 108, 193–197 Recycling, 3, 5–7, 9–12, 69–71, 79, 92, 94, 107–112, 114–117, 270 Recycling/Urban mining, 92 Renewable energy, 3–5, 107, 108, 116, 117, 139, 159, 182, 183, 188, 213

S Scope 1, 75, 77, 78 Scope 2, 76–78 Selective nickel and cobalt extraction, 67, 72 Sheet resistance, 27, 28, 30–37 Siphonic effect, 243–246, 249 Si production, 201 Slag, 97–100, 103–105, 109, 112, 113, 115, 201–204, 206–210, 254 Specific surface area, 40, 223–225, 228, 236, 258, 259 Spodumene, 81–84, 86, 213, 215 Sputtering, 27–30, 33, 35 Superconductor, 181–190 Sustainability, 89, 139, 213

T Thermal conductivity, 129–136 Thermal quadrupoles, 129, 133

Subject Index Thermodynamic properties, 262, 269, 270 Transmittance, 30, 33 Tunable Diode Laser Absorption Spectroscopy (TDLAS), 51–53, 57

V Vertical Axis Wind Turbine, 3, 4

277 W Waste, 5, 7, 64, 72, 73, 79, 90–92, 95, 103, 107, 108, 115, 140, 159, 165, 166, 168, 193, 194, 252, 270 Waste PCB, 107–112, 114–117

Z Zambia, 97, 98 Zero-emission shipping, 178 ZnS, 39–47, 94, 109, 165