Computers in Flavor and Fragrance Research 9780841208612, 9780841210899, 0-8412-0861-1

Content: Application of microcomputer technology in flavor research : sensory evaluation / Mark R. McLellan -- MAECIS, a

546 31 2MB

English Pages 158 Year 1984

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Computers in Flavor and Fragrance Research
 9780841208612, 9780841210899, 0-8412-0861-1

Citation preview

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.fw001

Computers in Flavor and Fragrance Research

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.fw001

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

ACS SYMPOSIUM SERIES

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.fw001

Computers in Flavor and Fragrance Research Craig B. Warren, EDITOR John P. Walradt, EDITOR International Flavors and Fragrances

Based on a symposium sponsored by the Division of Agricultural and Food Chemistry at the 186th Meeting of the American Chemical Society, Washington, D.C., August 28-September 2, 1983

American Chemical Society, Washington, D.C. 1984

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

261

Library of Congress Cataloging in Publication Data

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.fw001

Computers in flavor and fragrance research. (ACS symposium series, ISSN 0097-6156; 261) "Based on a symposium sponsored by the Division of Agricultural and Food Chemistry at the 186th Meeting of the American Chemical Society, Washington, D.C., August 28-September 2, 1983." Includes bibliographies and indexes. 1. Food—Odor—Analysis—Data processing— Congresses. 2. Flavor—Analysis—Data processing— Congresses. I. Warren, Craig B., 1939. II. Walradt, John P., 1942.III. American Chemical Society. Division of Agricultural and Food Chemistry. IV. American Chemical Society. Meeting (186th: 1983: Washington, D C . ) V. Series. TP372.5.C66 1984 ISBN 0-8412-0861-1

644'.07

84-14607

Copyright © 1984 American Chemical Society All Rights Reserved. The appearance of the code at the bottom of the first page of each chapter in this volume indicates the copyright owner's consent that reprographic copies of the chapter may be made for personal or internal use or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc., 21 Congress Street, Salem, M A 01970, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to copying or transmission by any means—graphic or electronic—for any other purpose, such as for general distribution, for advertising or promotional purposes, for creating a new collective work, for resale, or for information storage and retrieval systems. The copying fee for each chapter is indicated in the code at the bottom of the first page of the chapter. The citation of trade names and/or names of manufacturers in this publication is not to be construed as an endorsement or as approval by A C S of the commercial products or services referenced herein; nor should the mere reference herein to any drawing, specification, chemical process, or other data be regarded as a license or as a conveyance of any right or permission, to the holder, reader, or any other person or corporation, to manufacture, reproduce, use, or sell any patented invention or copyrighted work that may in any way be related thereto. Registered names, trademarks, etc., used in this publication, even without specific indication thereof, are not to be considered unprotected by law. PRINTED IN THE UNITED STATES

OF

AMERICA

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

ACS Symposium Series Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.fw001

M . Joan Comstock, Series Editor Advisory Board Robert Baker U.S. Geological Survey

Geoffrey D . Parfitt Carnegie-Mellon University

M a r t i n L. Gorbaty Exxon Research and Engineering C o .

Theodore Provder Glidden Coatings and Resins

Herbert D . Kaesz University of California—Los Angeles

James C . Randall Phillips Petroleum Company

R u d o l p h J . Marcus Office of Naval Research

Charles N . Satterfield Massachusetts Institute of Technology

M a r v i n Margoshes Technicon Instruments Corporation

Dennis Schuetzle Ford Motor Company Research Laboratory

D o n a l d E. M o r e l a n d U S D A , Agricultural Research Service W. H . N o r t o n J. T. Baker Chemical Company Robert O r y U S D A , Southern Regional Research Center

Davis L . Temple, Jr. Mead Johnson Charles S. Tuesday General Motors Research Laboratory C . Grant Willson I B M Research Department

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

FOREWORD T h e A C S SYMPOSIUM SERIES was founded in 1974

to provide

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.fw001

a medium for publishing symposia quickly in book form. T h e format of the Series parallels that of the continuing ADVANCES IN CHEMISTRY SERIES except that in order to save time the papers are not typeset but are reproduced as they are submitted by the authors in camera-ready form. Papers are reviewed under the supervision of the Editors with the assistance of the Series Advisory Board and are selected to maintain the integrity of the symposia; however, verbatim reproductions of previously published papers are not accepted. Both reviews and reports of research are acceptable since symposia may embrace both types of presentation.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

PREFACE WE

A R E D R O W N I N G IN I N F O R M A T I O N but starved for knowledge." This

statement by John Naisbitt in his book Megatrends probably best sums up the prevailing theme of the symposium that underlies this book, that is, the use of the computer to convert complex and elusive information into useful knowledge.

O u r book,

Computers

in Flavor

and

Fragrance

Research,

attempts to provide a general view of this impact. The book may be viewed

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.pr001

as comprising three sections. The first section, Chapters 1 to 6, deals with the stand-alone use of computers for information handling and number crunching.

Chapters 7 through

10 describe the synergistic combination of a

dedicated computer and an analytical instrument to create a hybrid analytical technique. Finally, Chapter 11 suggests that we all had better start thinking about the use of robotics as a tool for automating routine laboratory operations. We chose a general view because the impact of computers on flavor and fragrance research is not limited to a particular area. The advent of the microprocessor has made powerful, inexpensive microcomputers available to the analytical chemist and the sensory scientist alike. These people have connected them to their machines, used them to control robots, and placed them in their sensory evaluation booths. The successful development of inexpensive memory and very fast central processing units, on the other hand, has made very powerful minicomputers available to the computational chemist and the information scientist. These researchers now routinely use the computer to design new functional molecules, design new products, and keep track of huge collections of molecules and associated data. Although the computer applications presented in this book may appear to be slanted toward the flavor and fragrance field, it should be emphasized that this slant only represents a particular end use of a molecule. The computer applications are valid for all molecules; hence, this book should be of general interest to any researcher whose job it is to develop new molecules or products based on molecules. C R A I G B. W A R R E N J O H N P. W A L R A D T

International Flavors and Fragrances 1515 Highway 36 Union Beach, New Jersey March 1984 IX

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

1 Application of Microcomputer Technology in Flavor Research Sensory Evaluation M A R K R. McLELLAN

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch001

Cornell University, Department of Food Science and Technology, New York State Agricultural Experiment Station, Geneva, N Y 14456

The advances i n m i c r o e l e c t r o n i c s have b r o u g h t us the development of powerful microcomputers. These computers can be used i n various areas of research i n c l u d i n g f l a v o r research. A review of the technology is given with specific notes on microcomputer development, design, and languages. A summary of f i n d i n g s from the d e s i g n and use o f the m i c r o c o m p u t e r as a d a t a collection tool i n sensory a n a l y s i s i s reported. Evaluations of t h i s as w e l l as other p o t e n t i a l areas of a p p l i c a t i o n i s i n c l u d e d and d i s c u s s e d as related to sensory analysis.

The microcomputer l i k e any other t o o l needs t o be d e v e l o p e d and s t u d i e d f o r i t to p e r f o r m r e l i a b l y and c o n s i s t e n t l y . Like few other tools the p o s s i b l e a p p l i c a t i o n s of microcomputers has o n l y begun to be d i s c o v e r e d . In the f i e l d of computers, the pace of development has been phenomenally r a p i d . I t has been r o u g h l y akin to going from the Wright B r o t h e r s f i r s t f l i g h t to the space s h u t t l e i n l i t t l e over ten years. Few other i n d u s t r i e s have ever matched t h i s r e l a t i v e s c a l e of development. I t i s certainly reasonable to assume t h a t , w i t h the c u r r e n t advances i n l a r g e s c a l e and u l t r a l a r g e s c a l e i n t e g r a t i o n and developments w i t h i n the semiconductor industry i n g e n e r a l , computers w i l l e v e n t u a l l y a f f e c t a l l aspects of human l i f e . 1

The d e f i n i t i o n of a microcomputer Computers i n g e n e r a l have t r a d i t i o n a l l y been d i s t i n g u i s h e d by a number of c h a r a c t e r i s t i c s . The two f o r e m o s t c h a r a c t e r i s t i c s g e n e r a l l y c i t e d a r e , the word s i z e which the computer uses and the speed of operations performed by the computer. In any computer, the most fundamental u n i t of information i s considered the b i n a r y d i g i t , o t h e r w i s e known as a " b i t " . In 0097-6156/84/ 0261 -0001 $06.00/0 © 1984 American Chemical Society

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch001

2

COMPUTERS IN FLAVOR A N D F R A G R A N C E RESEARCH

s c r i p t a b i t can be r e p r e s e n t e d as e i t h e r a n u m e r i c a l 1 or numerical 0. However, i n the a c t u a l c i r c u i t r y of the computer, t h i s i n f o r m a t i o n i s represented as the presence of a voltage or the absence of a v o l t a g e . T y p i c a l l y , the p r e s e n c e of a v o l t a g e i s r e p r e s e n t e d a t a +5 v o l t l e v e l . The absence of a voltage i s g e n e r a l l y considered ground p o t e n t i a l or 0 v o l t s . The p r i m a r y u n i t o f i n f o r m a t i o n i n a computer i s a group of b i t s r e f e r r e d to as a word. An 8 - b i t word i s c a l l e d a " b y t e " and t h i s can be c o n s i d e r e d a u n i v e r s a l l y used u n i t i n the computer i n d u s t r y . In the d e s i g n o f computers, any s i z e of word can be u t i l i z e d ; however, t y p i c a l l y , word s i z e s vary i n m u l t i p l e s of 8 b i t s . The speed of the computer system i s g e n e r a l l y c i t e d i n terms of the number of machine operations that are capable of o c c u r r i n g w i t h i n a c e r t a i n p e r i o d of time. Many times t h i s i s l i s t e d as the number of c y c l e s per second ( h e r t z ) that the c e n t r a l processor u n i t can operate a t . T r a d i t i o n a l l y , a main frame computer was one that operated using 32 or 64 b i t words. A minicomputer was one that u t i l i z e d a 16 b i t word and the microcomputer was one that used a 4 or 8 b i t word. However, these d i s t i n c t i o n s i n c l a s s i f i c a t i o n s a r e v e r y q u i c k l y d i s a p p e a r i n g w i t h the f u r t h e r m i n i a t u r i z a t i o n of computers, the i n c r e a s i n g speed of microprocessors, and s p e c i f i c a l l y the d e s i g n o f more p o w e r f u l m i c r o and mini computers. Many of the new microcomputers a r e f a r more p o w e r f u l and c a p a b l e t h a n what j u s t a few y e a r s ago was c o n s i d e r e d a l a r g e main frame system. The b i r t h o f t h e microcomputer can a c t u a l l y be traced back to the development of the t r a n s i s t o r . With e a r l y e l e c t r o n i c d e v i c e s and computers the system of s t o r i n g d i g i t a l information was based on the use of vacuum t u b e s . These were cumbersome, expensive, and used a tremendous amount of power. They were much f a s t e r than r e l a y s , however, they were c o n s i d e r a b l y s l o w e r t h a n a n y t h i n g produced under today's standards. With the development of the t r a n s i s t o r a r e v o l u t i o n i n the design of computer systems u s h e r e d i n t h e time when systems would become smaller and more capable and l e s s c o s t l y . The t r a n s i s t o r was f a s t e r t h a n i t s p r e d e c e s s o r , the vacuum t u b e , r e q u i r e d less power and was much cheaper to develop and produce than the vacuum tube technology. As development progressed s t a r t i n g i n the l a s t 1950*8 and on i n t o the 7 0 s , a new technology d e v e l o p e d which a l l o w e d us t o b u i l d s o l i d s t a t e d e v i c e s c o n t a i n i n g s e v e r a l components a l l on the same s i l i c o n c h i p . The term f o r t h i s t e c h n o l o g y i s i n t e g r a t e d e l e c t r o n i c c i r c u i t r y ( 1 ) . The i n t e g r a t e d c i r c u i t components e s s e n t i a l l y became the f o u n d a t i o n f o r the p o c k e t c a l c u l a t o r and e v o l v e d i n t o what i s known t o d a y as t h e microcomputer. This e l e c t r o n i c development has r a p i d l y expanded t o a p o i n t where we now see the use of large s c a l e i n t e g r a t i o n and the development of u l t r a - l a r g e s c a l e i n t e g r a t i o n . This i s l e a d i n g t o h i g h l y s o p h i s t i c a t e d equipment a v a i l a b l e f o r use by f

f

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

1.

McLELLAN

Sensory Evaluation

1

3

t o d a y s sensory a n a l y s t ; a l l one needs t o l e a r n i s how t o communicate a p p r o p r i a t e l y with t h i s new technology.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch001

Communications with a

microcomputer

There a r e a large number of languages that are a v a i l a b l e f o r use on microcomputers. A computer "understands" o n l y one l a n g u a g e , machine language - binary code. A l l languages, not b i n a r y , must be i n t e r p r e t e d or c o m p i l e d down t o machine l a n g u a g e , so t h a t computers can understand the i n s t r u c t i o n s . One could use binary to communicate with the computer, however, i t would be tremend o u s l y burdensome. A more usual means of communicating with the system would be u s i n g the more s o p h i s t i c a t e d o r h i g h l e v e l languages. BASIC i s p r o b a b l y one of the most p o p u l a r of a l l computer languages. I t can a l s o be c o n s i d e r e d one of the most versatile. BASIC i s an acronym that stands f o r "Beginners A l l purpose Symbolic I n s t r u c t i o n a l Code". There i s a v e r s i o n of BASIC a v a i l a b l e f o r most every computer system that i s marketed today. BASIC does have some drawbacks caused by i t s i n h e r e n t l a c k of s t r u c t u r e . I t i s o f t e n s a i d that BASIC programmers have too much freedom to be able to jump around and the r e s u l t of t h i s type of freedom i s the l a c k of consistency and tremendous comp l e x i t y i n the of d e v e l o p m e n t of s o p h i s t i c a t e d programs. FORTRAN, an acronym which stands f o r FORmular TRANslation, i s a language which can be found on many systems. This was one of the f i r s t high l e v e l languages to achieve a b a s i c s t a n d a r d i z a t i o n and wide a c c e p t a n c e . I t was m a i n l y d e s i g n e d f o r s c i e n t i f i c and m a t h e m a t i c a l use. In g e n e r a l , the e a r l y microcomputers s u p p l i e d d i d not have the memory c a p a c i t y t o r u n FORTRAN programs and t h e r e f o r e had to do with BASIC. With the development of l a r g e r memories and a b i l i t y to run compiled l a n g u a g e s , i t became poss i b l e to run FORTRAN on the smaller systems. A f t e r the w i d e s p r e a d u s e o f BASIC and FORTRAN t h e r e d e v e l o p e d a r e c o g n i z e d need to have a v a i l a b l e s t r u c t u r e d languages. A s t r u c t u r e d language i s one based on a h i e r a r c h y of o p e r a t i o n s s t a r t i n g w i t h the g e n e r a l and p r o c e e d i n g t o the specific. There has been a movement i n computer s c i e n c e towards t h e s e languages t h a t f u l l y u t i l i z e data s t r u c t u r e s and r e q u i r e users to be very s p e c i f i c i n t h e i r d e c l a r a t i o n o f a l l c h a r a c t e r i s t i c s f o r a program. P a s c a l , PL/1, C, and Ada are among the languages of t h i s type and they a r e q u i c k l y r e p l a c i n g o l d e r languages f o r both s c i e n t i f i c and business computing. Structured languages, e s p e c i a l l y P a s c a l , are becoming major i n s t r u c t i o n a l languages i n computer s c i e n c e d e p a r t m e n t s . We can c e r t a i n l y expect to see these languages encroach on the domain of FORTRAN, BASIC and C0BAL as b e i n g the most important language a v a i l a b l e f o r use. The U. S. Department of Defense has decided that Ada s h a l l be the language required on a l l programming a p p l i c a t i o n s d e a l i n g w i t h the Army, Navy and A i r F o r c e i n the near f u t u r e . Ada was

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

4

COMPUTERS IN FLAVOR AND FRAGRANCE RESEARCH

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch001

e s s e n t i a l l y developed as a d e r i v a t i v e of Pascal and i t was named a f t e r the f i r s t computer programmer, Lady Ada Augusta Byron, daughter of the poet Lord Byron. Although we have yet to see any s e r i o u s m a r k e t i n g e f f o r t , of a f u l l y implemented microcomputer based Ada; we s h o u l d c e r t a i n l y e x p e c t t o have t h i s l a n g u a g e a v a i l a b l e on most systems i n the not too f a r f u t u r e . FORTH can be considered independent of the p r e v i o u s l y d i s ­ cussed language or what one might consider a Fourth generation of languages. T h i s language i s b o t h a language and a computer o p e r a t i n g s y s t e m i n one. I t i s h a i l e d as a g i a n t l e a p i n f l e x i b i l i t y and c a p a b i l i t y ( 2 ) . FORTH was c r e a t e d by one man, C h a r l e s H. More and was f i r s t used to c o n t r o l the telescope a t Kipp's Peak Observatory. I t has since developed a f o l l o w i n g o f programmers and i s supported as a good process c o n t r o l language. Current A p p l i c a t i o n s i n Sensory A n a l y s i s Microcomputers as a n a l y s i s

tools

When we speak i n terms of a n a l y s i s t o o l s f o r sensory e v a l u a t i o n , g e n e r a l l y we a r e t a l k i n g about the d a t a a n a l y s i s done u s i n g various s t a t i s t i c a l methodologies. In order to a c c o m p l i s h t h i s type of t a s k one must e i t h e r turn to a preprogrammed s t a t i s t i c a l packages or e l s e do s p e c i f i c s t a t i s t i c a l a n a l y s i s w i t h i n ones own programming. There a r e now a v a i l a b l e some s m a l l s t a t i s t i c a l packages f o r m i c r o c o m p u t e r s . However, p r e d o m i n a n t l y t h e s e s t a t i s t i c a l packages have been w r i t t e n by i n d i v i d u a l s i n the f i e l d and not by major s t a t i s t i c a l s o f t w a r e v e n d o r s . With the development of systems s u p p o r t i n g l a r g e amounts of memory and s o p h i s t i c a t e d support software; the development of s e r i o u s and extensive s t a t i s t i c a l a n a l y s i s t o o l s f o r microcomputers should be f a i r l y s t r a i g h t f o r w a r d . We are now s e e i n g a v e r y few s p e c i f i c s o f t w a r e packages designed f o r the sensory a n a l y s i s s c i e n t i s t to help him/her evaluate data from sensory e v a l u a t i o n s . An example of t h i s i s s o f t w a r e b e i n g d e v e l o p e d by A & Ν A s s o c i a t e s (Lansing, M i c h i g a n ) f o r v e r y s p e c i f i c s e n s o r y a n a l y s i s t e s t s i n c l u d i n g D i f f e r e n c e T e s t s , ANOVA, e t c . As we see extended development i n t h i s a r e a , we w i l l f i n d s o f t w a r e becoming more a v a i l a b l e and v e r y s p e c i a l i z e d i n t h e a r e a of support f o r the sensory a n a l y s t . In t h i s l a b o r a t o r y we a r e c u r r e n t l y i n the p r o c e s s of d e v e l o p i n g and s t u d y i n g an a p p l i c a t i o n of r a n k a n a l y s i s on microcomputers. T h i s a n a l y s i s i s s p e c i f i c a l l y designed f o r use i n the e v a l u a t i o n of m u l t i p l e r e c i p e s or m u l t i p l e sample l o t s f o r determination of p o s s i b l e rank d i f f e r e n c e s . The r e s u l t i n g output from our software package w i l l c o n s i s t of an organized l i s t i n g of j u d g e s ' r a n k i n g s , rank t o t a l s , rank order of samples based upon rank t o t a l s and s i g n i f i c a n t d i f f e r e n c e s where noted. Once f u l l y d e v e l o p e d t h i s p i e c e of software would be a tremendous asset to those involved i n a comparison of m u l t i p l e l o t s and should a s s i s t

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

1.

McLELLAN

Sensory Evaluation

5

the r e s e a r c h e r i n h i s e v a l u a t i o n o f b l e n d - t o - b l e n d v a r i a t i o n s and/or raw product v a r i e t a l d i f f e r e n c e s . We hope to see f u r t h e r developments i n t h i s a r e a of s p e c i f i c software f o r support of data a n a l y s i s techniques i n sensory a n a l y s i s .

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch001

Microcomputers i n data a c q u i s i t i o n In g e n e r a l , we can speak of two s p e c i f i c types of o p e r a t i o n s u s i n g m i c r o c o m p u t e r s i n s e n s o r y a n a l y s i s as d a t a a c q u i s i t i o n t o o l s . The f i r s t would be i n a passive mode, that i s the system would be used to c o l l e c t data as entered a f t e r the a c t u a l evaluat i o n has t a k e n p l a c e and the d a t a has been c o l l e c t e d on some o t h e r f o r m , f o r example, a paper b a l l o t . In t h i s case the m i c r o c o m p u t e r i s b e i n g used as a s e c o n d a r y s o u r c e o f d a t a a c q u i s i t i o n . The primary one being the paper b a l l o t . The second mode of o p e r a t i o n would be using the microcomputer as the primary and s o l e data a c q u i s i t i o n medium. We would then be r e l y i n g upon the computer to handle not only the c o l l a t i o n and s t o r a g e of the data but a l s o the i n t e r a c t i v e means by which the a n a l y s t c o l l e c t s data from the p a n e l i s t s . The use of the microcomputer i n a passive mode can be done using any number of techniques. For example, the most common use i n t h i s type of secondary data c o l l e c t i o n phase would be manually punching i n the data i n t o the microcomputer. For a l l i n t e n t s and p u r p o s e s t h i s i s e s s e n t i a l l y equivalent to using the system as a data management system or a data a n a l y s i s system. A second form of t h i s p a s s i v e d a t a a c q u i s i t i o n would be use of the d i g i t i z e r technology. A d i g i t i z e r i s a system where d a t a i s c o l l e c t e d u s i n g a paper b a l l o t t y p e method and the microcomputer i s connected to the d i g i t i z e r which i s then used to reproduce the d a t a from the paper b a l l o t i n t o a d i g i t a l format as c o l l e c t e d by the digitizer. The d i g i t i z e r i t s e l f i s e s s e n t i a l l y an e l e c t r o n i c s t y l u s connected to s p e c i a l i z e d g r i d network. The g r i d network acts as a p l a t f o r m where the paper b a l l o t i s p l a c e d , then the stylus i s u s e d to s i m p l y touch the paper o v e r each of the premarked points and the g r i d network underneath the paper i d e n t i f i e s where the s t y l u s i s t o u c h i n g and what that p a r t i c u l a r d i g i t i z e d x, y p o i n t i s . A s i g n i f i c a n t amount of s o f t w a r e i s r e q u i r e d t o d r i v e t h i s type of system and to i d e n t i f y the a c t u a l data p o i n t s f o r what they are. A n o t h e r type of p a s s i v e system has been suggested using automated c a r d r e a d i n g d e v i c e s ( 3 ) . In t h i s case a c a r d i s d e s i g n e d t o emulate as c l o s e as p o s s i b l e the type of sensory e v a l u a t i o n s c a l e r e q u i r e d f o r the p a r t i c u l a r s t u d y . The s c a l e must be denoted i n the form of i n d i v i d u a l marks. Each mark can be read, by the microcomputer using the c a r d r e a d e r . The t e c h nique i s very quick. I t can be automated f a i r l y e a s i l y and i t has been u t i l i z e d f o r a number of d i f f e r e n t types o f s c a l e s and tests. A l t h o u g h we see i t b e i n g a p p l i e d i n e v e r y t h i n g from d i f f e r e n c e t e s t s to q u a n t i t a t i v e d e s c r i p t i v e a n a l y s i s , t h e r e may

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch001

6

COMPUTERS IN FLAVOR AND FRAGRANCE

RESEARCH

be some b i a s e s and problems b u i l t i n when u s i n g t h i s type of marking system f o r a continuous and semi-continuous s c a l e such as t h o s e used f o r the q u a n t i t a t i v e d e s c r i p t i v e a n a l y s i s . The s c a l e emulation i s done by having a whole s e r i e s of c i r c l e s forming the continuous l i n e . The p a r t i c u l a r c i r c l e of i n t e r e s t r e f l e c t i n g the p o i n t of i n t e r e s t on the s c a l e must be darkened by a p e n c i l and then that point i s read by the card reader. Biases may a r i s e by having d i s c r e t e p o i n t s on a l i n e as opposed t o a f u l l y continuous s c a l e or l i n e . The second form of u s i n g the m i c r o c o m p u t e r as a d a t a a c q u i s i t i o n t o o l takes the form of an a c t i v e or what we c a l l " r e a l time" data a c q u i s i t i o n . By r e a l - t i m e we mean use of the computer for d a t a c o l l e c t i o n d u r i n g t h e a c t u a l e v a l u a t i o n by the panelists. The p a n e l i s t s would t h e n i n t e r a c t u s i n g v a r i o u s methods with the computer g i v i n g r e s u l t s to the computer based on sensory perceptions of the samples. Thus f a r there have been two proposed systems f o r t h i s type of i n t e r a c t i v e use. The two modes a v a i l a b l e f o r t h i s r e a l time d a t a a c q u i s i t i o n would be 1) a modified d i g i t i z i n g technique and 2) a d i r e c t data entry technique. In the d i g i t i z i n g system the d i g i t i z e r would be b u i l t i n t o a sensory panel booth. The p a r t i c u l a r b a l l o t would s t i l l be made up i n the form of a paper b a l l o t and p l a c e d on the e l e c t r o n i c g r i d p a l l e t and t h e n the p a n e l i s t would e v a l u a t e h i s or h e r samples and d i r e c t l y mark the paper b a l l o t using the e l e c t r o n i c s t y l u s that i s located i n the booth. Concurrently the computer would m o n i t o r the data input from the d i g i t i z e r and c o l l e c t the appropriate responses. T h i s a p p l i c a t i o n appears t o be v e r y successful. There might be considered a number of drawbacks to i t i n that i t r e q u i r e s a f a i r l y heavy i n v e s t m e n t i n hardware, t h a t b e i n g the d i g i t i z i n g equipment as w e l l as a s o p h i s t i c a t e d enough system to c o n t r o l t h i s hardware. I t s t i l l i n v o l v e s the use of paper b a l l o t s , however, i n general i t i s a v a s t improvement over t r a d i t i o n a l techniques of d a t a c o l l e c t i o n i n s e n s o r y evaluation. Under the second type of r e a l time data a c q u i s i t i o n , we have d i r e c t e n t r y methods where e s s e n t i a l l y you are i n c o r p o r a t i n g an a c t u a l terminal i n t o the sensory e v a l u a t i o n booth i t s e l f . The e l e c t r o n i c t e r m i n a l would i n c l u d e a data entry keyboard as w e l l as a d i s p l a y screen f o r r e c e i v i n g r e s u l t s and prompts from the m i c r o c o m p u t e r ( 4 ) . T h i s t e c h n i q u e has been developed i n our l a b o r a t o r i e s and we f e e l t h a t i t h o l d s tremendous promise i n terms of the v e r s a t i l i t y and the v i a b i l i t y f o r expanding i n t o d i f f e r e n t areas of sensory e v a l u a t i o n . Under t h i s t y p e of t e c h n i q u e , the system f l a s h e d on the d i s p l a y the various questions and s c a l e s f o r use by the p a n e l i s t to e n t e r h i s d a t a on. In the prototype that we developed, the software must be programmed f o r the p a r t i c u l a r type of t e s t ahead of t i m e . A p a n e l i s t entered the booth with the samples already

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch001

1.

McLELLAN

Sensory Evaluation

7

i n place and logged onto the computer system. The l o g on p r o c e dure i s f a i r l y s t r a i g h t f o r w a r d where the p a n e l i s t enters h i s or h e r own name and t h e m i c r o c o m p u t e r would come back w i t h t h e response to the person a f t e r having checked a small data base f o r p a n e l i s t name v e r i f i c a t i o n . The m i c r o c o m p u t e r then randomized the order of the samples that were i n f r o n t of the p a n e l i s t . The p a n e l i s t then responded t o t h e q u e s t i o n s f o r each sample and upon f i n i s h i n g the questions f o r each of the samples the p a n e l i s t was t o l d that he was done. The data was then stored i n a s p e c i a l f i l e f o r t h e purpose o f t r a n s f e r t o another computer f o r data a n a l y s i s or stored f o r a l a t e r a n a l y s i s on t h e same system c o l l e c t i n g the data. The time savings alone based upon t h i s system i s enough t o warrant f u r t h e r i n v e s t i g a t i o n . Based upon a t y p i c a l q u a n t i t a t i v e d e s c r i p t i v e a n a l y s i s panel c o n s i s t i n g of 18 panels each e v a l u a t ing 3 samples w i t h 20 q u e s t i o n s the time savings over s t r i c t l y manual methods was approximately 4-5 hours. Compared to d i g i t i z ing t h e raw data the time savings was approximately an hour. We f e e l that t h i s type of d i r e c t entry system w i l l o f f e r a tremendous advantage over most o t h e r systems i n terms of c a p a b i l i t y , v e r s a t i l i t y and the development o f new a p p l i c a t i o n s as w e l l as extensive time savings. Future A p p l i c a t i o n s The f u t u r e a p p l i c a t i o n s can g e n e r a l l y b r e a k down i n t o two specific categories. The f i r s t i s d a t a a n a l y s i s by m i c r o c o m p u t e r s . With the increased power of the new systems and extended memory c a p a b i l i t i e s ; we w i l l see t h e development and a p p l i c a t i o n of v e r s i o n s of the t r a d i t i o n a l l a r g e s t a t i s t i c a l packages a v a i l a b l e f o r use on t h e s m a l l s y s t e m s . With the d e v e l o p m e n t o f t h e s e p a c k a g e s we w i l l h a v e a tremendous c a p a b i l i t y of doing very extensive data a n a l y s i s using microcomp u t e r systems. Desk top computers w i l l not be simply be thought of as a personal data base type machine b u t r a t h e r as a major s t a t i s t i c a l t o o l t h a t can h a n d l e tremendous q u a n t i t i e s of data doing h i g h l y s p e c i f i c and h i g h l y s o p h i s t i c a t e d data a n a l y s i s and interpretation. The second area of p r o g r e s s i v e development w i l l be i n t h e a r e a of v e r y s p e c i a l i z e d data c o l l e c t i o n mediums. We foresee a major extension i n the area of d i r e c t data i n p u t u s i n g i n t e r a c t i v e v i d e o terminals b u i l t w i t h i n the sensory e v a l u a t i o n booths. Although t h i s w i l l take a f a i r l y e x t e n s i v e development i n t h e a r e a of software, the gains to be reaped by the development w i l l f a r exceed the development c o s t s i t s e l f . A l a r g e a r e a of r e s e a r c h has y e t t o be c o n d u c t e d c o n c e r n i n g t h e a p p l i c a t i o n of microcomputers i n sensory a n a l y s i s as a d a t a c o l l e c t i o n t o o l . S t u d i e s o f i n h e r e n t b i a s , e f f e c t s on response freedom e t c . w i l l have to be undertaken i n order to p r o p e r l y evaluate the computer as a data a c q u i s i t i o n t o o l . Biases based on the way the software

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

8

C O M P U T E R S IN F L A V O R A N D F R A G R A N C E

RESEARCH

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch001

i s designed c o u l d be v e r y e a s i l y i n c o r p o r a t e d i n t o a s e n s o r y p r o j e c t w i t h o u t knowledge by the p r i n c i p l e i n v e s t i g a t o r . These type of things should be s t u d i e d and the p e o p l e i n v o l v e d i n t h e development of t h i s type of software made aware of i t . We a l s o f e e l t h e r e i s room f o r e x t e n s i v e t e s t i n g and r e search i n new areas a s s o c i a t e d with t h i s type of technology. For example, time dependent measurements may be important f o r various types of sensory e v a l u a t i o n . P r e v i o u s l y , to achieve time dependent measurements was a tremendously exhaustive chore and may n o t have been able to be made using normal t a s t e panel procedures and environments. With t h i s type of a p p l i c a t i o n of microcomputers i n s e n s o r y e v a l u a t i o n we foresee the use of time dependent measurements as a new area of p o s s i b l e development. Literature Cited 1. 2. 3. 4.

Waite, M.; Pardu, M. "Microcomputer Primer"; Howard W. Sams & Co., Inc.; I n d i a n a p o l i s , Indiana, 1977. Dessy, R. E.; S t a r l i n g , M. K. Am. Lab. 1980, 12(2), 21. Van de V o o r t , F. R.; Miller, H.; A s t o n , G. C. Can. I n s t . Food Sci. Tech. J . 1981, 14(3), 173. McLellan, M. R. Food Tech. 1983, 31(1), 97.

R E C E I V E D A p r i l 27, 1984

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

2 MAECIS A Computer System for Handling and Analysis of Flavor and Fragrance Molecules WILLIAM E. BRUGGER

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

International Flavors and Fragrances, 1515 Highway 36, Union Beach, NJ 07735

MAECIS is a computer program that combines a chemical registry system, a chemical information handling system, and various molecular manipulation c a p a b i l i t i e s into a single program that is available to the research staff at IFF on an interactive basis. MAECIS provides chemical information retrieval using standard text searching procedures as well as substructure and structure searching. MAECIS also allows users to calculate various molecular properties and a three-dimensional model of any chemical in the system as well as new chemicals yet to be made. MAECIS is designed to be used by the chemist in an interactive manner without the need of a computer expert. The e f f i c i e n t and proper handling of information are major concerns of a l l companies doing business in today's world. The a b i l i t y to retrieve the type of information that allows one to work smarter is a primary driving force behind laboratory and o f f i c e automation. In companies that deal with chemical processes, information handling systems are more complex due to the need to associate chemical structures with other information. The problem becomes even more complex in the flavors and fragrance industry where i t is common to have many chemicals present in a single product as well as one chemical appearing in a number of products. The combination of the above situation with the pract i c e of using t r i v i a l names and more than one numbering system for identical products, results in a major chemical information handling problem that only a computer system can solve. Our solution to this problem was the development of an interactive computer system that can be used by a l l professionals in the company to locate and manipulate chemical information. This system i s called MAECIS, an acronymn for Management and Analysis Executive for Chemical Information and Structures. As the name implies, MAECIS is both a chemical information storage system and a chemical structure analysis system. 0097-6156/ 84/ 0261 -0009506.00/ 0 © 1984 American Chemical Society

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

10

COMPUTERS IN FLAVOR A N D FRAGRANCE RESEARCH

MAECIS contains five major sections that are integrated into a single system. At i t s center is a database management system that handles the storage and updating of data and chemical structures. Associated with this section is an information retrieval section that allows one to search the database for specific information composed of any combination of data and/or substructures. Once information is retrieved, i t can be displayed on a computer terminal or sent to a printer for hardcopy output using the display section of MAECIS which includes the a b i l i t y to display chemical structures in various formats. The final two sections of MAECIS allow for the manipulation of chemical structures and the calculation of various molecular properties that are related to chemical structures. The majority of MAECIS is written in FORTRAN with a few assembler routines for database management. The program i s currently running on a VAX 11/780 computer system and uses Tekronix 4025 graphics terminals as the primary input/output devices. MAECIS is a command oriented program that has a b u i l t in prompt mechanism to assist the user with commands. It also has a HELP section that provides the user with detailed online instructions. DATABASE MANAGEMENT The database section of MAECIS gives the user access to a large public database or allows him to create a private one. The public database is available for use by any authorized person in a read only mode. The actual maintenance of the data is done by our Technical Information staff who are responsible for the accuracy of the data. To allow a chemist to use the tools a v a i l able in MAECIS for experimental purposes, a private system is provided. Each chemist can have one or more private f i l e s containing "experimental" molecules which can be manipulated in the same manner as those in the public database. The public database consists of a number of indexed sequent i a l f i l e s that contain the data. At the center of the database i s the master f i l e which contains an unique International Flavors & Fragrances (IFF) registry number, the Chemical Abstract Service's registry number ( i f one has been assigned), the molecular weight, various dates associated with discovery and entry of the data, the source of the structure, gas chromatographic retention times, and a series of flags that indicate the a v a i l a b i l i t y of various analytical data. Other f i l e s in the public database contain chemical names, company code numbers, organoleptic evaluation data, and chemical structure data. The information in these data f i l e s are related to the master f i l e by means of the unique registry number that is assigned to each molecule when entered into the system. The public system is setup to allow the entry of a molecule only once. Thus, MAECIS serves as our chemical registry system.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

2.

BRUGGER

MAECIS: Molecular Analysis

11

The private database system in MAECIS, on the other hand, i s much simpler. It stores only a number and name assigned by the user along with the associated chemical structure. There i s no attempt to register the compounds. The owner of a private f i l e has f u l l read and write control over his data. Along with the public and private databases, there are special public and private f i l e s for storing substructures and structure backbones. Backbones are structure elements, e;. £ . , a benzene ring, used to construct more complex chemical structures. Like the public database, the public substructures and backbones can be changed only by authorized people, but the private f i l e s are t o t a l l y controlled by the individual user. The entry of chemical structures into MAECIS is done using a new version of UDRAW (1) that has been developed for the Tektronix 4025 graphics terminal. The structure is sketched onto the terminal's screen in a two-dimensional graphical representation that a l l chemist's are familiar with. As the chemical structure i s sketched, the computer system is storing i t s connection table and two-dimensional coordinates. Upon completion of the input, the system checks for any unusual valences or other obvious mistakes in the chemical diagram. If the structure is going into the public database, a search is made of a l l public structures to make sure that the new structure is unique. If no match is found, the system assigns a new IFF registry number and the user i s presented a data form on the CRT screen and asked to enter associated information. The modification of data already on f i l e i s accomplished in a somewhat similar manner. The information in the database is f i r s t retrieved through one of the query methods and then altered. In the case of invalid data or an unplausable chemical structure, MAECIS also contains a delete function. MAECIS creates and maintains for the user a l l of the necessary crossreferences between the associated data f i l e s . The chemical connection table(2) in MAECIS stores a l l the molecular structure information such as atomic valences, atomic symbols, atomic numbers, formal charges, stero-chemical data, and ring bonding information. Along with the connection table, twodimensional and three-dimensional coordinates obtained from molecular modeling, are stored for each structure. The storage of three-dimensional coordinates gives the user rapid access to a three-dimensional representation of a public structure without having to go through the molecular modeling procedure. To obtain the three-dimensional coordinates a l l structures in the public system are automatically modeled in a batch mode and then reviewed for unusual conformations.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

12

COMPUTERS IN FLAVOR A N D FRAGRANCE RESEARCH

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

INFORMATION RETRIEVAL AND DISPLAY The information retrieval in MAECIS is accomplished using one of three available commands: SHOW, FIND, or SEARCH. The SHOW command is the simplest one to use and requires only a code number or registry number. It allows the user to retrieve a l l chemical structures and associated information stored under a particular code number. In most cases this f u l f i l l s the user's needs. The FIND command is used for complex searches involving various combinations of multiple data f i e l d s . The command also handles substructure searching. Queries such as "find a l l the structures with a molecular weight between 200 and 250 containing an ester substructure" are handled by the FIND command. F i n a l l y , the SEARCH command is used for chemical structure searches. This search takes only seconds and allows the chemist to determine i f a particular molecule is already in the database. The result of any retrieval command in MAECIS is a set of chemical structures which is referred to as the "current set". If a second retrieval is made, the current set is moved into another internal storage area and is referred to as the "prior set". In this way, MAECIS automatically keeps track of at least two retrievals for the user. However, i t is often necessary to establish several sets of molecules during a single computer session. In MAECIS, any number of sets can be retained using a SAVE command which allows the user to name and save any set of molecules created by a retrieval command. Any stored set in MAECIS can also be referenced in a subsequent FIND command. Thus, new sets can be generated from old sets through the find command's boolean operations. Sets retained with the SAVE command are in temporary memory. To make them permanent, the STORE and RESTORE commands are available to move sets to and from permanent disk f i l e s . Given a set of molecules, the REVIEW command is used to display the chemical structure and associated information of any member of the set. The review mode allows the user to both page through the current set of molecules, one after another, as well as to jump to a specific entry within the set. The chemical structure currently displayed is referenced by MAECIS as the "current" structure. The user can also designate a second member of the set as the "alternate" structure. This allows for comparison of structures which w i l l be described in the section on chemical structure manipulation. The REVIEW command draws the chemical structure on one part of the screen, and then writes the associated information on the remainder. To obtain a f u l l screen view of a chemical structure, the DRAW command is used. Options with the DRAW command include atom numbered structures, space-filled diagrams with or without imbedded stick diagrams, stero pairs, and a special box view where the top, side, and front view of the structure are shown at the same time (see Figure 1). The draw routine can use either

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

2. BRUGGER

MAECIS: Molecular Analysis

Figure 1. Example o f the BOX o p t i o n o f the DRAVT command using galaxolide (1,3,^,6,7,8-hexahydro-U,6,6,7,8,8-hexamethylcyclopenta g -2-"benzopyran) as an example.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

13

14

COMPUTERS IN FLAVOR A N D F R A G R A N C E RESEARCH

two-dimensional display coordinates or the three-dimensional coordinates obtained from molecular modeling. Also, the draw command allows the alternate structure to be drawn with the current structure for a one-to-one comparisons.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

MANIPULATION AND CALCULATIONS Although the two-dimensional representation of a chemical structure i s sufficient in chemical information management, the laboratory chemist looking for s i m i l a r i t i e s between a group of chemicals requires additional information obtainable from the three-dimensional representation of the chemical. In MAECIS, a l l chemicals in public data f i l e s are modeled into a low strain conformation (3) and the three-dimensional coordinates are stored with the connection table. When a molecule is retrieved, the user can specify to use the modeled rather than the display coordinates. To assist in the viewing of the three-dimensional structure, the structure can be rotated about any axis or two atoms can be aligned with any screen axis. The DRAW command is then used to view the molecule in i t s new orientation. In some cases, the chemist may want to move atoms around and then remodel. MAECIS provides these capabilities as w e l l . Hyrdrogens can be added to the structure or removed. If chiral centers are present in the molecule, the mirror image of the molecule can be obtained with the MIRROR command. In a l l there are over 10 commands with various options to manipulate three-dimensional chemical structures. One of the more powerful manipulation commands in MAECIS is COMPARE which allows the current molecule to be superimposed upon the alternate molecule. This is accomplished in the following steps. F i r s t , the user specifies the atom pairs in the two structures which are to be overlapped. At least three pairs of atoms must be specified. The program then performs a nonlinear least-squares calculation to minimize the distance between these atom pairs. F i n a l l y , the user specifies that the molecules are to be drawn superimposed upon each other. The superimposed molecules can be drawn with any of the options available with the DRAW command. An i l l u s t r a t i o n of this i s contained in Figure 2 where two musk odorants are compared. Thus, a chemist can obtain an idea of the structural s i m i l a r i t y of any two molecules. MAECIS also contains a molecular conformation analysis system (4). This system allows the user to generate a l l possible conformations of the current molecule over a series of single bond rotations. Energy contour maps can be obtained for the various conformations and this allows for the selection of low energy conformations for further manipulation or calculations. In addition to the manipulation commands, several molecular properties can be calculated for a molecule. The molecular weight is calculated using the exact mass (5) of the most abun-

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

2. BRUGGER

15

MAECIS: Molecular Analysis

Figure 2. I l l u s t r a t i o n o f the r e s u l t o f a COMPARE operation using galaxolide (1,3,^ ,6,7,8-hexahydro-U,6,6,7,8,8-hexamethylcyclopenta g -2-benzopyran) and Musk-89 (3 h,6,7,8,9-hexahydroh,6,6,9,9-pentamethyl ΙΗ-naptho 2,3-C pyran) as the example mole­ cules . 9

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

16

COMPUTERS IN FLAVOR AND F R A G R A N C E RESEARCH

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

dant isotope of each atom in the structure. The molecular volume is calculated using van der Waal radii (6) and either standard or modeled bond distances. Normal, bond, and valence connectivity indicies (7) are calculated from the connection table. Also using the connection table, the LogP octanol/water partition coefficients are calculated (8). F i n a l l y , the principal moments of the molecule can be calculated using the three-dimensional coordinates in conjunction with the connection table data (9). Since these parameters can be used for doing structure a c t i v i t y analyses, MAECIS provides a method to calculate these parameters for a set of molecules and then to dump them to a computer f i l e for subsequent access by other systems. SUMMARY Currently, MAECIS is at the center of our technical information database and contains over 19,000 unique chemical structures. It is used daily by our research staff to locate information on specific products or chemicals. The database structure in MAECIS serves as a bridge between the analytical and organic chemist doing research on specific molecules and those concerned with new products which are usually mixtures of chemicals. Since i t s introduction over three years ago, MAECIS has enjoyed wide use for information r e t r i e v a l . However, the computational parts of MAECIS are s t i l l largely unexplored and under u t i l i z e d . In time this w i l l change as our chemist spend less time looking for information and more time using information. MAECIS represents the first-generation of chemical informations systems along with many others in various corporations (MMMS at Merck (10); TRIBBLE at duPont (11), MOLY at Rohm and Haas (12), COUSIN at Upjohn (13), and M0L0CH-2 at Searle (14)). Although a large amount of effort has gone into MAECIS, i t has more than paid for i t s e l f by increasing the productivity of the professional who use i t . In the future, new systems w i l l be developed which easily surpass MAECIS in ease of use and s a t i s faction. However, MAECIS has allowed us to take the f i r s t step into computational chemistry and information management.

Literature

1. 2. 3.

Cited

Brugger, W. E. and Jurs, P. Anal. Chem. 1975, 47, 781. Lynch, M. F.; Harrison, J. M.; Town, W. G. "Computer Handling of Chemical Structure Information", Macdonald: London, 1971. Wipke, W. T.; Dyott, T. M.; Verbal i s , J. G., Abstract, 161st National Meeting, American Chemical Society, Los Angeles, CA, March, 1971.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

2.

BRUGGER

4. 5. 6. 7. 8. 9. 10. Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch002

11. 12. 13. 14.

MAECIS: Molecular Analysis

17

Hopfinger, J . "Conformational Properties of Macromolecules", Academic Press: New York, 1973. Wapstra, A. H. and Bos, Κ., Atomic Data and Nuclear Data Tables 1977, 19(3), 177. Bondi, A. J. J. Phys. Chem. 1964, 68, 441. Randic, M. J. Am. Chem. Soc. 1975, 97, 6609. Chou, J. T. and Jurs, P. C. J. Chem. Inf. Comp. S c i . 1979, 19, 172. Stuper, A. J.; Brugger, W. E.; Jurs, P. C. "Computer Assisted Studies of Chemical Structure and Biological Function", Wiley Interscience: New York, 1979. Gund, P.; Andose, J. D.; Rhodes, J. B.; Smith, G. M. Science 1980, 208, 1425. Eaton, D. F. and Pensak, D. A. J. Am. Chem. Soc. 1982, 22, 28. Dyott, T. M.; Stuper, A. J.; Zander, G. S. J. Chem. Inf. Comp. S c i . 1980, 20, 28. Howe, W. J. and Hagadone, T. R. J. Chem. Inf. Comp. S c i . 1982, 22, 28. Hodges, D.; Nordby, D.; Marshall, G., Abstract of the 169th National Meeting of the American Chemical Society, Philadelphia, PA, April, 1975.

RECEIVED May 3, 1984

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

3 Computer-Based Molecular Design of Artificial Flavoring Agents A. J. HOPFINGER and D. E. WALTERS

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

Searle Research and Development, Department of Medicinal Chemistry, Skokie, IL 60077

Computational chemistry methodology is f i n d i n g i n c r e a s i n g a p p l i c a t i o n t o the design of new f l a v o r i n g agents. This chapter surveys s e v e r a l u s e f u l techniques: l i n e a r free energy r e l a t i o n s h i p s , q u a n t i t a t i v e s t r u c t u r e - a c t i v i t y r e l a t i o n s h i p s , conformational a n a l y s i s , e l e c t r o n i c s t r u c ­ t u r e c a l c u l a t i o n s , and statistical methods. A p p l i c a t i o n s t o the study of artificial sweeteners are described. I m p l i c i t to research i n the biochemical sciences, of which the food and f l a v o r i n d u s t r i e s are members, i s that b i o l o g i c a l response, BR, i s a f u n c t i o n of chemical s t r u c t u r e , CS. The goal i s to f i n d that CS (compound) which maximizes a d e s i r e d BR and, simultaneously, min­ imizes other BRs, e s p e c i a l l y those d i r e c t l y r e l a t i n g to t o x i c i t y . In p r a c t i c e , t h i s i n v o l v e s the synthesis and t e s t i n g of new chemical e n t i t i e s , and, u n f o r t u n a t e l y , i s governed to a l a r g e extent by chance. A p p l i c a t i o n of computer-based molecular modeling i s i n t e n d ­ ed to reduce the chance component, and, thereby, to enhance design capabilities. The design f u n c t i o n r e s t s upon r e c o g n i t i o n and implementation of the r e l a t i o n s h i p s shown i n Figure 1. In part A we f o r m a l i z e that not only do a l l BRs map i n t o (depend upon) CS, but so do a l l p h y s i cochemical p r o p e r t i e s , PPs. Therefore, the CS can, from f u n c t i o n a l mapping, be considered the common node between the BRs and the PPs. Consequently, a unique f u n c t i o n , f , must e x i s t between each BR and the set of PP. T h i s i s defined in°part Β of Figure 1. The key assumption i n molecular design i s that i f a p a r t i c u l a r BR i s of i n t e r e s t , say BR^, then the corresponding f . holds f o r any CS. T h i s i s expressed i n part C of Figure 1 where CS denotes an a r b i t r a r y mth chemical s t r u c t u r e . The r e l i a b i l i t y of t h i s assump­ t i o n increases as the s t r u c t u r a l homology of the set of CS increases and, a l s o , as the mechanisms of a c t i o n to r e a l i z e BR^ over the [CS ] converge to commonality. This i s a formal explanation why q u a n t i t a t i v e s t r u c t u r e a c t i v i t y r e l a t i o n s h i p , QSAR, s t u d i e s are more s u c c e s s f u l when a p p l i e d to e x p l a i n i n v i t r o a c t i v i t y i n a set of con­ generic analogs, as opposed to i n v i v o a c t i v i t y f o r a set of s t r u c t u r a l l y d i v e r s e compounds. m

0097-6156/ 84/ 0261 -0019S06.00/ 0 © 1984 American Chemical Society

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

20

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

COMPUTERS IN FLAVOR AND F R A G R A N C E RESEARCH

BRl»fl(PPlrPP2r

fPPk)

BR2-f2(PPlfPP2f

fPPk)

;

BRn-f (PPl»PP2»~-PPk) n

C

(BR ) =(f.(PPi,PP , 1

m

2

PPk))in

Figure 1. The i n t e r r e l a t i o n s h i p s between b i o l o g i c a l response (BR), chemical s t r u c t u r e (CS), and physicochemical p r o p e r t i e s (PP) that l a y the conceptual b a s i s f o r molecular design.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

3.

HOPFINGER AND WALTERS

21

Artificial Flavoring Agents

Methods

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

1.

A d d i t i v e Property Models The simplest r e p r e s e n t a t i o n of a molecule, with respect to computing physicochemical p r o p e r t i e s , i s to assume the property to be the sum of the property values of the i n d i v i d u a l c o n s t i t u e n t atoms, or groups of atoms, Extensive data bases (1,2) of atomic and group (fragment) property values have been compiled to f a c i l i t a t e implementation of t h i s model. The most notable physicochemical p r o p e r t i e s employed i n QSARs using an a d d i t i v e property model are: 1. l o g P, the water-octanol p a r t i t i o n c o e f f i c i e n t , ( 3 ) 2. σ, the Hammett constant,(4) 3. MR, molecular r e f r a c t i v i t y index,(5) 4. pk , i o n i z a t i o n constant,(6) and 5· E , the T a f t s t e r i c constant. (2) a

s

Log Ρ and MR are considered thermodynamic d e s c r i p t o r s , pK a combined thermodynamic and e l e c t r o n i c index, and σ an e l e c t r o n i c property index. Ε i s designed to account f o r s t e r i c e f f e c t s . C o r r e c t i o n s f o r n o n - a d d i t i v i t y , based upon the chemical bonding topology, have been suggested and used. These i n c l u d e p r o x i m i t y , bond type, r i n g , and group shape c o r r e c t i o n f e a t u r e s . (8-10) Molecular c o n n e c t i v i t y , ( 1 1 ) which i s based upon graph theory, (12) i s an e m p i r i c a l a l t e r n a t i v e to an a d d i t i v e model employing physicochemical p r o p e r t i e s . D e s c r i p t o r s d e r i v e d from the molecular connection t a b l e (chemical bonding topology) employing mathematical f u n c t i o n s are considered as p o t e n t i a l a c t i v i t y c o r r e l a t e s . As such, t h i s approach i s completely mathematical and has no obvious physicochemical b a s i s . I t s strength i s that c o r r e l a t i o n i n d i c e s can always be generated (provided one knows how to a s s i g n i n t r i n s i c r e l a t i v e weights to i n d i v i d u a l atom types). a

2.

Hansch A n a l y s i s The most s u c c e s s f u l , and the most o f t e n used, method to c o n s t r u c t a QSAR i s that of Hansch.(13) T h i s method employs multi-dimensional l i n e a r r e g r e s s i o n a n a l y s i s to c o r r e l a t e s t r u c t u r e to a c t i v i t y i n a c h e m i c a l l y congeneric set of compounds. The s t r u c t u r a l f e a t u r e s have been t r a d i t i o n a l l y d e r i v e d from a d d i t i v e property models. However, recent a p p l i c a t i o n s of Hansch A n a l y s i s recognize any and a l l molecular d e s c r i p t o r s as p o t e n t i a l c o r r e l a t e s to a c t i v i t y . ( 1 4 , 1 5 ) Prominent among t h i s l i n e of t h i n k i n g i s Hansch h i m s e l f who now f r e e l y uses i n d i c a t o r v a r i a b l e s as c o r r e l a t e s . ( 1 6 ) An i n d i c a t o r v a r i a b l e has a value of 1 i f some user-defined property i s present i n a compound and a value of zero i f the property i s absent. I t i s important to p o i n t out that Hansch A n a l y s i s i s based upon a b i o l o g i c a l a c t i o n model. By d e r i v i n g the general QSAR equation a s s o c i a t e d with the

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

22

COMPUTERS IN FLAVOR AND FRAGRANCE

RESEARCH

a c t i o n model, i t i s p o s s i b l e to c o n c e p t u a l l y j u s t i f y the general usage of any molecular d e s c r i p t o r i n a c o r r e l a t i o n analysis.(17)

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

3.

Conformational A n a l y s i s A drawback to a d d i t i v e property models of molecular s t r u c t u r e i s that 3-dimensional molecular p r o p e r t i e s cannot be determined. However, i t i s c l e a r that conformation, or more g e n e r a l l y , molecular shape, can be an important f a c t o r i n expression o f b i o l o g i c a l a c t i v i t y . Thus, there i s a need to be a b l e to determine conformational/shape p r o p e r t i e s of molecules and to use these p r o p e r t i e s to develop a QSAR. There are two p r i n c i p a l methods of performing computational conformational analyses: Molecular mechanics(18-19) and quantum mechanics.(20) Molecular mechanics considères a molecule as a set of b a l l s (the atoms), coated with s t i c k i n g paste, connected by a s e t of s p r i n g s (the bonds) according to a p r e s c r i b e d s e t of valence angles. Quantum mechanics views the molecule as a s e t of n u c l e i i n space with e l e c t r o n s moving about the n u c l e i . The goal of a conformational a n a l y s i s i s to minimize the energy of the molecule as a f u n c t i o n of i t s geometry. The energy minimization can i n v o l v e d i f f e r e n t types and numbers of geometric degrees of freedom which i n c l u d e bond l e n g t h s , bond angles, and t o r s i o n a l bond r o t a t i o n angles. An o f t e n employed approximation, e s p e c i a l l y f o r l a r g e molecules, i s to hold the valence geometry constant and minimize the energy as a f u n c t i o n of only t o r s i o n a l bond r o t a t i o n s . The unproven, but reasonable, assumption i m p l i c i t to SARd i r e c t e d conformational s t u d i e s , both experimental and t h e o r e t i c a l , i s that one of the s t a b l e i n t r a m o l e c u l a r conformers i s the " a c t i v e " conformation. A d i f f i c u l t y to applying conformational data i n q u a n t i t a t i v e drug design i s s e l e c t i o n of conformational f e a t u r e s f o r QSAR development. Moreover, molecular shape p r o p e r t i e s a r e p r e f e r a b l e f e a t u r e s to have a v a i l a b l e i n design s t u d i e s . Conformation i s a component of shape. The p r o p e r t i e s of the atoms, most notably t h e i r " s i z e s , " comprise an a d d i t i o n a l set of f a c t o r s needed t o s p e c i f y molecular shape. Conformational f e a t u r e s have been used i n some s t r u c t u r e a c t i v i t y s t u d i e s . Some examples a r e : (1) an i n t e r a t o m i c d i s t a n c e w i t h i n a molecule,(21,22) (2) a s e t of i n t e r a t o m i c d i s t a n c e s w i t h i n a molecule,(23,24) (3) a s e t of atomic coordinates w i t h i n a molecule,(25,26) (4) a s e t of c r i t i c a l i n t e r m o l e c u l a r b i n d i n g distances.(27) Molecular shape p r o p e r t i e s , d e r i v e d from conformational i n v e s t i g a t i o n s , have a l s o been used to r a t i o n a l i z e commonality and d i v e r s i t y i n b i o l o g i c a l a c t i o n . These shape p r o p e r t i e s include: (1) molecular volume,(28) (2) molecular s u r f a c e area,(29)

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

3.

HOPFINGER AND WALTERS

Artificial Flavoring Agents

23

(3) s p a t i a l p o t e n t i a l surfaces of a molecule with respect to a t e s t species.(30,31) A very powerful t o o l of v i s u a l i z i n g three-dimensional molecular p r o p e r t i e s , i n c l u d i n g p o t e n t i a l s u r f a c e s , i s computer graphics.(32) Computer graphics i s p a r t i c u l a r l y u s e f u l i n the q u a l i t a t i v e comparison of two or more molecules. A general theory of q u a n t i t a t i v e l y comparing molecular shapes using common overlap s t e r i c volume(33-36) and, more r e c e n t l y , d e s c r i p t o r s derived from superimposed molecular p o t e n t i a l energy f i e l d s of p a i r s of molecules(37) has been derived and t e s t e d . T h i s theory allows a "marriage" between Hansch a n a l y s i s and conformational a n a l y s i s .

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

4.

E l e c t r o n i c Structure Calculations Molecular o r b i t a l theory provides e l e c t r o n i c , as w e l l as conformational, data f o r i n c l u s i o n i n QSAR development. E l e c t r o n i c p r o p e r t i e s of a molecule might, i n p a r t , c o n t r o l b i o l o g i c a l a c t i v i t y when chemical r e a c t i o n s are p a r t of the mechanism of b i o l o g i c a l a c t i o n . E l e c t r o n i c i n d i c e s should a l s o be considered i n modeling p h y s i c a l i n t e r a c t i o n s i n order to provide charge d i s t r i b u t i o n data f o r estimating binding energies. E l e c t r o n i c i n d i c e s o f t e n considered i n structure«-activity studies include; a) atomic charge d e n s i t i e s , b) bond, group, and/or molecular d i p o l e moments, c) o r b i t a l energy l e v e l s , e s p e c i a l l y the highest occupied molecular o r b i t a l , HOMO, lowest unoccupied molecular o r b i t a l , LUMO, and t h e i r d i f f e r e n c e , and d) o r b i t a l wavefunction c o e f f i c i e n t s . Several d i f f e r e n t molecular o r b i t a l methods have been used i n SAR i n v e s t i g a t i o n s . These i n c l u d e simple Huckel theory, HT,(38) extended Huckel theory, EHT,(39) CNDO,(40) NDD0,(41) MINDO/3,(42) and PCILO.(43) 5.

S t a t i s t i c a l Methods Multidimensional l i n e a r r e g r e s s i o n a n a l y s i s i s the most o f t e n employed s t a t i s t i c a l method f o r QSARs. T h i s p o p u l a r i t y i s coupled t o the acceptance o f the Hansch method f o r QSAR analyses. The techniques and p i t f a l l s of r e g r e s s i o n a n a l y s i s have been w e l l described.(44,45) Other s t a t i s t i c a l methods employed i n q u a n t i t a t i v e molecular design i n v e s t i g a t i o n s i n c l u d e d i s c r i m i n a n t a n a l y s i s , c l u s t e r a n a l y s i s , m u l t i p l e f a c t o r a n a l y s i s , and p a t t e r n r e c o g n i t i o n procedures.(46-48) P a t t e r n r e c o g n i t i o n may prove p a r t i c u l a r l y u s e f u l when the design o b j e c t i v e i s a complicated p r o f i l e of b i o l o g i c a l a c t i v i t i e s , as opposed to only a maximized potency and minimized t o x i c i t y . Gottman(49) and Schiffman(50) have used multidimensional s c a l i n g methods to e s t a b l i s h i n t e r r e l a t i o n s h i p s between p s y c h o l o g i c a l sensory p r o f i l e s and physicochemical p r o p e r t i e s of f l a v o r agents, e s p e c i a l l y sweeteners.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

COMPUTERS IN FLAVOR AND F R A G R A N C E RESEARCH

24 Case Study A p p l i c a t i o n s

L - A s p a r t y l Dipeptide Sweeteners Iwamura(51) has i n v e s t i g a t e d the structure-sweetness r e l a t i o n s h i p i n four c l a s s e s of L - a s p a r t y l d i p e p t i d e s using l i n e a r f r e e energy d e s c r i p t o r s and multi-dimensional r e g r e s s i o n a n a l y s i s . In essence, the Hansch methodology was employed. The four c l a s s e s of compounds are: « 1) L - a s p a r t i c a c i d amides, L-Asp-NHC H(R )C H ( R ) R Ί

?

1

N=66

2

2

(VII)

1 2 2) L - a s p a r t y l a m i n o e t h y l e s t e r s , L-Asp-NHC H(R )C H(R )OCOR !

1

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

N=61

2

2

(VIII) 1

2 l

3) L-aspartylaminopropionates, L-Asp-NHC H(R )C H(R )COOR 1

N=31

2

2

(IX)

4) L-aspartylaminoacetates, L-Asp-NHC"^ ( R ^ C O C ^ N=59

(X)

Ν i s the number of compounds i n each c l a s s . QSARs were generated f o r each of the four data bases. A r e p r e s e n t a t i v e multi-dimensional l i n e a r r e g r e s s i o n equation i s that developed f o r the L - a s p a r t y l a m i n o e t h y l e s t e r s (see F i g u r e 2 ) : 2

log(SP)=0.67o*(±0.48)+3.36L (±1.03)-0.29L (±0.08) 2

2

2

+4.18W (±0.88)-0.85W (±0.18)1

1

0.531^10.18)-11.33 n=51

(1) r=0.88

In equation (1) the L^ parameter (9) expresses the length of s u b s t i t u e n t R. to the r e s t of the molecule. W. i s the width upward of R^ when one views i t from the connecting end along the bond a x i s d e f i n i n g L.. The e l e c t r o n i c parameter, σ*, was estimated f o r the s t r u c t u r e s u b s t i t u t e d on the common a s p a r t y l amino moeity, so that the e l e c t r o n i c e f f e c t i s d i r e c t e d to the peptide bond. Ten compounds of the o r i g i n a l 61 do not f i t equation (1); hence n=51, not 61. The a n a l y s i s , i n composite over the four c l a s s e s of L - a s p a r t y l d i p e p t i d e s suggests that the electron-withdrawing e f f e c t of s u b s t i t u e n t s d i r e c t e d to the peptide bond, and the s t e r i c dimensions of the molecules, are important i n e l i c i t i n g the sweet t a s t e . The values o f the r e g r e s s i o n c o e f f i c i e n t s of the σ* term i n the QSAR equations f o r L - a s p a r t i c a c i d amides, L - a s p a r t y l a m i n o e t h y l e s t e r s , and L-aspartylaminopropionates a l l

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

3.

HOPFINGER AND WALTERS

Artificial Flavoring Agents

25

are approximately 0.7 suggesting these three c l a s s e s of d i p e p t i d e sweeteners i n t e r a c t i n a common manner at the r e c e p t o r . However, the σ* r e g r e s s i o n c o e f f i c i e n t f o r the L-aspartylaminoacetates QSAR i s approximately 1.5. T h i s v a l u e , along with an examination of the optimum s t e r i c parameter values i n the QSARs, suggest to Iwamura, that the L-aspartylaminoacetates can b e t t e r f i t to the receptor than the other three c l a s s e s of compounds. The QSARs developed by Iwamura provide c o n c i s e r e c i p e s f o r designing d i p e p t i d e sweeteners. However, the uniqueness of an LFE-based QSAR i s always an i s s u e . For t h i s case the s i n g u l a r i t y of the QSARs with respect to the s t e r i c parameters L^, L^ and i s suspect. The values of L^ and can be shown to c o r r e l a t e to the l i p o p h i l i c i t y of R^ as measured by l o g Ρ f o r 38 compounds with a c o r r e l a t i o n c o e f f i c i e n t of .943. Also c o r r e l a t e s with log Ρ f o r 40 choices of R^ with a c o r r e l a t i o n c o e f f i c i e n t of .908. O v e r a l l , i t i s p o s s i b l e to express l o g (SP) as;

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

9

log(SP)=f(σ*, l o g P ( R ) , l o g P ( R ) ) 1

n=36

r=0.87

(2)

2

s=0.24

The p r a c t i c a l advantage of equation (1) over equation (2) i s that r e l i a b l e estimates of l o g Ρ f o r choices of R^ and R are a v a i l a b l e f o r only 36 of the 51 compounds f o r which L^, and L can be measured. Nevertheless, care needs always to be taken i n e s t a b l i s h i n g the extent of c o r r e l a t i o n uniqueness among LFE d e s c r i p t o r s . T h i s i s normally achieved by c o n s t r u c t i n g a c o r r e l a t i o n matrix among a l l d e s c r i p t o r s considered. 2

2

Extension of the Shallenberger AH...Β Sweetener Model Shallenberger and coworkers(52) proposed what i s probably the most c o n s i s t e n t requirement i n sweeteners, namely the A-H...B hydrogen-bond model. An example of t h i s model, as a p p l i e d by Hopfinger and Jabloner,(53) i s shown i n F i g u r e 3 f o r the a r t i f i c i a l sweetener P-4000 (4000X s u c r o s e ) . In t h i s model the A-H group provides a proton donor f o r a weak i n t r a m o l e c u l a r hydrogen bond to an acceptor Β. Β can be any of a wide range of e l e c t r o n e g a t i v e groups. The d i s t a n c e between the proton and Β i s 2.5-4.OA, a range of values l a r g e r than normally a s s o c i a t e d with hydrogen bonds (1.7-1.9Â). The problem with the A - H — Β model i s that many organic compounds, which are not sweet, c o n t a i n such a grouping. Further, many sweeteners c o n t a i n m u l t i p l e A-H and/or Β groups which makes i t ambiguous i n a s s i g n i n g the r e l e v a n t A-H...Β system. For example, the a r t i f i c i a l sweetener aspartame contains three p o s s i b l e A-H s i t e s and s i x p o s s i b l e Β atoms, see F i g u r e 4. Thus, the u t i l i t y of using the A-H...Β model i n a design mode i s l i m i t e d . A hydro­ phobic s i t e " l o c a t e d " 5-6A from the A-H...Β system has been suggested(54) as an a d d i t i o n a l c o n s t r a i n i n g f e a t u r e necessary f o r sweeteners. Again a wide range of non-sweet organic compounds meet t h i s a d d i t i o n a l requirement.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

26

COMPUTERS IN FLAVOR A N D F R A G R A N C E RESEARCH

H

O H Η , η ι ι L. H NC°(L)C—NC (D/L)|;OCO *-=-» I

a

2

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

COOH /

Figure 2. K^ R , 9

2

R

2

+

General s t r u c t u r e o f the L-aspartylaminoethylesters L , and σ* d e f i n e d . 2

"H" Η. (2.5-4.0 Â)

Η \

/

c\

i

""θ—C~H-

λ

3

7

/

0

Figure 3.

P-UOOO with the A-H

F i g u r e U. M u l t i p l e A-H aspartame.

...

Β system d e f i n e d .

( e l l i p s e s ) and Β (squares)

sites in

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

with

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

3.

HOPFINGER A N D WALTERS

27

Artificial Flavoring Agents

Therefore, the question a r i s e s i f there a r e a d d i t i o n a l p r o p e r t i e s r e l a t e d to the A-H...Β system that might be c h a r a c t e r i s t i c of sweeteners. Hopfinger and Jabloner(53) were s t r u c k by the unusual geometry of many A-H.. Β systems i n sweeteners. The geometry i n F i g u r e 3 i s exemplary. One could e n v i s i o n the formation o f bimodal hydrogen bonding i n v o l v i n g both Η and Β with i n t e r m o l e c u l a r , r e c e p t o r - s i t e acceptors and donors r e s p e c t i v e l y . These workers proceeded to c a r r y out t h e o r e t i c a l confor­ mational analyses of a s e t of sweeteners, many of which have m u l t i p l e A-H and/or Β s i t e s . The s e t of compounds analyzed a r e reported i n Table I along with the sweet potency r e l a t i v e to sucrose. A f i x e d valence geometry molecular mechanics f o r c e f i e l d was used i n the conformational analyses. The conforma­ t i o n a l search s t r a t e g y was as f o l l o w s : 1. Generate a l l i n i t i a l conformer s t a t e s having at l e a s t one A-H...Β system i n which the H...B d i s t a n c e i s i n the 2.5-5.OA range. 2. Minimize the energy of each i n i t i a l conformer as a f u n c t i o n of the allowed t o r s i o n a l degrees o f freedom. 3. A f t e r t o t a l energy minimization r e c o r d / c a l c u l a t e the proper­ t i e s of the A-H...Β hydrogen bond i n c l u d i n g the bond energy. The most i n t e r e s t i n g o b s e r v a t i o n of t h i s a n a l y s i s i s that each compound i n the data base can e x i s t i n a t l e a s t one s t a b l e conformer f o r which the A-H...Β hydrogen bonding energy E(HB) i s i n the range of -1.5 to -2.5 kcal/mole. F u r t h e r , most potent sweeteners have an E(HB) near -2.0 kcal/mole. The set of E(HB) are reported as p a r t o f Table I, and a histogram p l o t of sweet potency versus E(HB) f o r the compounds of Table I i s shown i n F i g u r e 5. Organic compounds not sweet, but c o n t a i n i n g an A-H...Β system, have a range i n E(HB) v a l u e s of -0.6 to -9.0 kcal/mole. Thus, i t i s tempting to p o s t u l a t e that an a d d i t i o n a l necessary, but perhaps not s u f f i c i e n t , c o n d i t i o n f o r sweetness i s that the A-H..,Β system have an E(HB) i n the -1.5 to -2.5 kcal/mole range. One p o s s i b l e i n t e r p r e t a t i o n of A-H...Β e n e r g e t i c s i s that E(HB) s weaker than -1.5 to -2.5 kcal/mole may not be able to s p a t i a l l y o r i e n t the A-H...Β groups p r o p e r l y while E(HB) s stronger than -2.5 kcal/mole may be too great to allow the receptor B...A-H system to compete with the i n t e r m o l e c u l a r behavior. An obvious questions which a r i s e s i s what f a c t o r s a r e r e s p o n s i b l e f o r E(HB)? A general A-H...Β system i s shown i n F i g u r e 6. The energy E(HB) i s p r o p o r t i o n a l as, f

f

E(HB) OC

ε

f(Θ(ΑΗΒ)) d, ΉΒ

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

(3)

COMPUTERS IN FLAVOR A N D F R A G R A N C E RESEARCH

28

TABLE I

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

R e l a t i v e sweetness, RS, and hydrogen bond energy, E(HB) of an AH.,.Β system i n a s e t of sweeteners. Number 1 2 3 4 5 6 7

Compound

RS

Ε (HB) kcal/mole

P-4000 APM phyllodulcin (D)-tryptophan (D)-6-Cl tryptophan (D)-serine-O-n-C^H (D)-threonine-O H

4000 180 250 35 1600 320 150

-1, 95 -1, 83 -1, 80 -1.66 -1.85 -1. 72 -1. 75

3000 100 0 3000 25 340 225 130 260

-2.13 -1.92 -1.49 -2, 01 -1, 50 -1, 79 -1, 65 -1, 55 -1, 60

H 7

R NHCH(C=0)NH 1

(CH )nC0 H 2

8 9 10 11 12 13 14 15 16

2

CÔCF 1 COCF^ 1 COCH^ 1 COCF^ 2 cyclamate neohesperidin DHC perillartine acesulfam saccharin

.2 CN NO NO^ CN

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

3.

HOPFINGER A N D WALTERS

Artificial Flavoring Agents

Ε (HB)

29

kcal/mole

Figure 5· A histogram o f RS vs. E ( H B ) f o r the s e t o f sweeteners reported i n Table I .

Figure 6.

General s t r u c t u r e o f a hydrogen bond.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

30

COMPUTERS IN FLAVOR A N D FRAGRANCE

RESEARCH

Q and Q are the p a r t i a l atomic charges on the proton and Β atoms, r e s p e c t i v e l y , d ^ i s the d i s t a n c e between the Η and Β atoms and f(0(AHB)) i s a s c a l i n g f u n c t i o n between 0 and 1 depending upon the bond angle Θ(ΑΗΒ). The f i x e d molecular d i e l e c t r i c i s expressed as ε. Equation (3) i n d i c a t e s that four v a r i a b l e s c o n t r o l E(HB) and that r e a l i z a t i o n of a small range i n values l i k e -1.5 to -2.5 can be due to competitive c o n t r i b u t i o n s over the four v a r i a b l e s . Nevertheless, i t i s q u i t e i n t e r e s t i n g to observe that these four s t r u c t u r a l v a r i a b l e s i n t e r p l a y to e l i c i t a very t i g h t range i n E(HB) f o r the set of sweeteners considered. fi

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

Summary and Prospects The pharmaceutical i n d u s t r y has pioneered i n the a p p l i c a t i o n of computer-assisted drug design methods i n product r e s e a r c h . To a s i g n i f i c a n t degree t h i s i s a consequence of the d i r e c t use of computational chemistry i n enhancing the e f f i c i e n c y of the chemical lead o p t i m i z a t i o n process. For the f l a v o r i n d u s t r y to take s i g n i f i c a n t advantage of q u a n t i t a t i v e molecular design, more s t r i n g e n t measures of b i o l o g i c a l responses are necessary. The a r t i f i c i a l sweetener area represents an example i n f l a v o r chemistry where high q u a l i t y measures of sweet potency have made i t p o s s i b l e f o r QSAR techniques to be s u c c e s s f u l l y a p p l i e d . A l t e r n a t e a r t i f i c i a l f l a v o r i n g areas could be w e l l served by f o l l o w i n g the sweetener example. A great amount of methodology i n molecular design i s now a v a i l a b l e f o r a r t i f i c i a l f l a v o r i n g QSAR a p p l i c a t i o n s . Much of the methodology i s formatted i n easy to use software packages. Thus, computer-assisted molecular design can be viewed as a ready to use a n a l y t i c a l t o o l by f l a v o r chemists. The major b a r r i e r to using these t o o l s i s the l a c k of f a m i l i a r i t y and understanding f l a v o r chemists have with the a v a i l a b l e software. Literature Cited 1.

C. Hansch and A. Leo, "Substituent Constants f o r C o r r e l a t i o n A n a l y s i s in Chemistry and B i o l o g y , " W i l e y - I n t e r s c i e n c e s , New York, 1979. 2. R. R. Rekker, "The Hydrophobic Fragmental Constant," Pharmacochemistry L i b r a r y , V o l . 1, E l s e v i e r , New York, 1977. 3. C. Hansch and J . M. C l a y t o n , J. Pharm. Sci. 1973, 6 2 , 1. 4. M. Charton, Chemtech 1974, 502; 1975, 245. 5. C. Hansch, S. H. Unger, and A. B . Forsythe J . Med. Chem.

1973, 6. 7. 8.

16, 1217.

P. S e i l e r , Eur. J . Med. Chem. 1974, 9 , 663. A. J. Verloop, in "Drug Design," V o l . 3 , E. J. A r i e n s , Ed., Academic P r e s s , New York, 1972, p. 133. A. Leo, P. Y. C. Jow, C. Silipo, and C. Hansch J . Med. Chem.

1975,

18, 865.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

3. 9.

10.

HOPFINGER A N D WALTERS

Artificial Flavoring Agents

A. Verloop, W. Hoogenstraaton, and J. T i p k e r , in "Drug Design," V o l . 7, E. J. A r i e n s , Ed., Academic P r e s s , New York, 1976, p. 165. K. Bowden and K. R. H. Woolridge Biochem. Pharm. 1973, 2 2 ,

1015. 11. 12. 13. 14. 15.

L. B. K i e r and L. H. Hall, "Molecular C o n n e c t i v i t y i n Chem­ istry and Drug Research," Academic P r e s s , New York, 1976. R. Wilson, "Introduction t o Graph Theory," Academic P r e s s , New York, 1972. C. Hansch, A. R. Steward, J. Iwasa, and E. W. Deutsch Mol. Pharmacol. 1965, 1, 205. C. Silipo and C. Hansch J . Am. Chem. Soc. 1975, 97, 6849. Y. C. M a r t i n , T. M. Bustard, and K. R. Lynn J . Med. Chem.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

1973, 16, 1089. 16.

27. 28. 29. 30. 31. 32.

C. Hansch, C. Silipo, and E. E. S t e l l e r J . Pharm. Sci. 1975, 64, 1186. A. J. Hopfinger, "Intermolecular I n t e r a c t i o n s and Biomolecular O r g a n i z a t i o n , " W i l e y - I n t e r s c i e n c e , New Yor,, 1977. A. J. Hopfinger, "Conformational P r o p e r t i e s o f Macromolecules," Academic P r e s s , New York, 1973. D. B. Boyd and K. B. Lipkowitz J. Chem. Ed. 1982, 5 9 , 269. G. A. Segal, Ed. "Methods o f E l e c t r o n i c S t r u c t u r e C a l c u l a t i o n s , Parts A and B," Plenum P r e s s , New York, 1977. L. B. K i e r J . Pharmacol. Exp. Ther. 1968, 164, 75N. S. Ham i n "Molecular and Quantum Pharmacology," E. D. Bergmann and B. Pullman, Eds., Reidel-Dordrecht, H o l l a n d , 1974, p. 261. D. C. Rohrer, D. S. F u l l e r t o n , A. H. L. From, and K. Ahmed i n "Computer A s s i s t e d Drug Design," E. C. Olson and R. E. C h r i s t o f f e r s e n , Eds., ACS Monograph S e r i e s 112, Washington, D.C., 1979, p. 259. L. B. K i e r in "Fundamental Concepts in Drug-Receptor I n t e r a c t i o n s , " J . Danielli, J . Moran, and D. T r i g g l e , Eds., Academic P r e s s , New York, 1970, p. 15. H. J. R. Weintraub and A. J. Hopfinger J. Theor. Biol. 1973, 41, 53. H. Weinstein, Β. Z. A p f e l d e r , S. Cohen, S. Maayani, and M. Sokolovsky in "Conformation o f B i o l o g i c a l Molecules and Polymers," E. D. Bergmann and B. Pullman, Eds., Academic Press, New York, 1973, p. 531. G. M. Crippen J. Med. Chem. 1979, 2 2 , 988. K. Yamamoto J . Biochem. 1974, 76, 385. D. Pensak, Unpublished work, 1978. H. Weinstein I n t . J. Quantum Chem. 1975, 2, 59. J . B a r t l e t t and H. Weinstein Chem. Phys. L e t t . 1975, 30, 441. N. L. Max, D. MAlhotra and A. J. Hopfinger Computers and Chem.

33. 34.

A. J. Hopfinger J. Am. Chem. Soc. 1980, 120, 7196. C. B a t t e r s h e l l , D. Malhotra, and A. J. Hopfinger J. Med. Chem.

17. 18. 19. 20. 21. 22.

23.

24.

25. 26.

1981, 5 , 19. 1981, 24, 812.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

32

35. 36. 37. 38. 39. 40. 41. 42.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch003

43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

C O M P U T E R S IN F L A V O R A N D F R A G R A N C E R E S E A R C H

A . J. Hopfinger J. Med. Chem. 1981, 2 4 , 818. A . J. Hopfinger and R. Potenzone, Jr. M o l . Pharmacol. 1982, 21, 187. A . J. Hopfinger J. Med. Chem. 1983, 26, 990. A . S t r e i t w e i s e r , "Molecular O r b i t a l Theory f o r Organic Chemists," W i l e y , New Y o r k , 1961. R. Hoffman J. Chem. Phys. 1963, 39, 1397. J. A . P o p l e , D. P . S a n t r y , and G. A . Segal J. Chem. Phys. 1965, 43, 5129. O. K i k u c h i Bull. Chem. Soc. Japan 1977, 50, 593. R. C. Bingham, M. J. S. Dewar, and D. H . Lo J. Am. Chem. Soc. 1975, 97, 1302. J. L . C o u b e i l s , P h . C o u r r i e r e , and B. Pullman C. R. Acad. Sci., P a r i s 1971, 272, 1813. P . J. Goodford in "Advances i n Pharmacology and Chemotherapeutics," Academic P r e s s , New Y o r k , 1973, p . 52. J. G. T o p l i s s and R. J. C o s t e l l o J. Med. Chem. 1972, 15, 1066. Β. R. Kowalski and C. F . Bender J. Am. Chem. Soc. 1973, 95., 586. R. J. Matthews J. Am. Chem. Soc. 1975, 97, 935. C. L . Perrin Science 1973, 183, 551. L . Bottman Psychometrika 1968, 33, 469. S. S. Schiffman Science 1974, 185, 112. H . Iwamura J. Med. Chem. 1981, 24, 572. R. S. Shallenberger and T. C. Acree Nature (London) 1967, 216, 480. A . J. Hopfinger and H. J a b l o n e r in "The Q u a l i t y of Foods and Beverages," G. Charalambous and G. Inglett, E d s . , Academic P r e s s , New Y o r k , 1981, p . 83. L . B . K i e r J. Pharm. Sci. 1972, 61, 1394.

RECEIVED May

4,

1984

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

4 Mathematical Approaches for Quantitative Design of Odorants and Tastants SUSAN S. SCHIFFMAN

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

Duke Medical Center, Department of Psychiatry, Durham, N C 27710

Odor and taste q u a l i t y can be mapped by multidimensional scaling (MDS) techniques. Physicochemical parameters can be related to these maps by a variety o f mathematical methods including multiple regression, canonical correlation, and partial l e a s t squares. These approaches t o s t u d y i n g QSAR ( q u a n t i t a t i v e structure-activity relationships) in the c h e m i c a l senses, along w i t h procedures developed by the p h a r m a c e u t i c a l i n d u s t r y , may ultimately be u s e f u l in d e s i g n i n g f l a v o r compounds by computer.

I t i s n o t y e t p o s s i b l e t o design a m o l e c u l e w i t h s p e c i f i c odor (or t a s t e ) c h a r a c t e r i s t i c s because t h e r e l a t i o n s between sensory p r o p e r t i e s o f f l a v o r compounds and t h e i r m o l e c u l a r p r o p e r t i e s a r e not w e l l understood. As a consequence, t h e development o f compounds w i t h d e s i r e d f l a v o r q u a l i t i e s h a s h a d t o r e l y o n r e l a t i v e l y t e d i o u s s y n t h e t i c approaches. Recent advances, however, i n computer-based methods developed by t h e p h a r m a c e u t i c a l i n d u s t r y t o s t u d y QSAR ( q u a n t i t a t i v e s t r u c t u r e a c t i v i t y r e l a t i o n s h i p s ) may u l t i m a t e l y be h e l p f u l i n t h e r a t i o n a l d e s i g n o f new f l a v o r - s t r u c t u r e s w i t h p r e d i c t a b l e s e n s o r y a t t r i b u t e s . R e s u l t s f r o m QSAR s t u d i e s may a l s o p r o v i d e i n s i g h t i n t o t h e mechanism o f t h e m o l e c u l e - r e c e p t o r i n t e r a c t i o n . QSAR s t u d i e s o f f l a v o r m o l e c u l e s r e q u i r e t w o t y p e s o f d a t a as i n p u t : 1. Q u a n t i t a t i v e measures obtained i n psychophysical experiments that describe the sensory p r o p e r t i e s o f the molecules. 2. Physicochemical parameters that provide a d e s c r i p t i o n of t h e m o l e c u l a r p r o p e r t i e s r e l e v a n t t o t h e f l a v o r properties. U n d e r s t a n d i n g t h e r e l a t i o n s b e t w e e n t h e p s y c h o p h y s i c a l and p h y s i c o c h e m i c a l p r o p e r t i e s c a n be a c h i e v e d b y a v a r i e t y o f m a t h e m a t i c a l methods d e s c r i b e d i n t h i s paper. 0097-6156/ 84/ 0261 -0033506.00/0 © 1984 American Chemical Society

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

34

COMPUTERS IN FLAVOR AND

Obtaining q u a n t i t a t i v e sensory data. In the past, the q u a n t i f i c a t i o n o f o d o r (and t a s t e ) q u a l i t y has r e l i e d upon r a t i n g s on v e r b a l d e s c r i p t o r s . T h i s a p p r o a c h has s e v e r a l d r a w b a c k s i n t h a t : (a) i t i s h i g h i n e x p e r i m e n t e r c o n t a m i n a t i o n b e c a u s e a p r i o r i a s s u m p t i o n s m u s t be made a b o u t t h e r e l e v a n t a t t r i b u t e s and (b) t h e r e i s c o n s i d e r a b l e v a r i a b i l i t y among s u b j e c t s because of i n d i v i d u a l d i f f e r e n c e s i n i n t e r p r e t a t i o n of t h e m e a n i n g s and r e l a t i v e i m p o r t a n c e o f t h e w o r d s . The r e c e n t l y d e v e l o p e d m e t h o d o l o g y o f m u l t i d i m e n s i o n a l s c a l i n g (MDS) o v e r c o m e s t h e s e d r a w b a c k s by u s i n g d i r e c t m e a s u r e s o f s i m i l a r i t y as i n p u t (1). i s i m p l y a m a t h e m a t i c a l t o o l t h a t uses e x p e r i m e n t a l j u d g m e n t s o f p e r c e i v e d f l a v o r s i m i l a r i t y t o d e r i v e a map that r e p r e s e n t s t h e t a s t e o r s m e l l q u a l i t i e s o f s p e c i f i c m o l e c u l e s as p o i n t s i n a space. Compounds j u d g e d s i m i l a r i n t a s t e ( o r o d o r ) s e n s a t i o n a r e a r r a n g e d by MDS p r o c e d u r e s c l o s e t o one a n o t h e r i n a r e s u l t a n t s p a t i a l map; t h o s e d i s s i m i l a r a r e l o c a t e d f a r f r o m one a n o t h e r . An e x a m p l e f r o m g e o g r a p h y i l l u s t r a t e s how MDS p r o c e d u r e s work. Maps o f t h e U n i t e d S t a t e s f r e q u e n t l y g i v e t a b l e s o f i n t e r c i t y d i s t a n c e s among m a j o r c i t i e s . I f MDS w e r e a p p l i e d t o a l l c o m b i n a t i o n s of d i s t a n c e s between these c i t i e s , the a n a l y s i s w o u l d a c c u r a t e l y r e c o v e r t h e u n d e r l y i n g s t r u c t u r e , t h a t i s , a map o f t h e c i t i e s i n p r o p e r r e l a t i o n s h i p t o one a n o t h e r . The d i f f e r e n c e b e t w e e n g e o g r a p h i c a l and f l a v o r s p a c e s i s s i m p l y t h a t we don't know a p r i o r i w h a t f l a v o r s p a c e s l o o k l i k e . T h e r e a r e many ways t o o b t a i n d i s t a n c e - l i k e m e a s u r e s among f l a v o r compounds. F i r s t , s t i m u l i a r e g e n e r a l l y e q u a t e d i n s u b j e c t i v e i n t e n s i t y so t h a t j u d g m e n t s a r e b a s e d on q u a l i t y r a t h e r t h a n i n t e n s i t y . O d o r a n t s a r e d i l u t e d i n an o d o r l e s s g r a d e o f d i e t h y l p h t h a l a t e and t a s t a n t s , i n d e i o n i z e d w a t e r . Then s u b j e c t s r a t e a l l the n ( n - l ) / 2 p o s s i b l e p a i r s f o r a set of η s t i m u l i a l o n g an u n d i f f e r e n t i a t e d 5" l i n e : M D S

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

FRAGRANCE RESEARCH

exact same

8

most different

S i m i l a r i t y m e a s u r e s c a n a l s o be o b t a i n e d u s i n g a 10 p o i n t s c a l e (e.g., f r o m 1 t o 10) o r by o b t a i n i n g c o n f u s a b i l i t i e s among triads. Odorants (or t a s t a n t s ) confused most f r e q u e n t l y are c o n s i d e r e d more s i m i l a r . M i s s i n g d a t a d e s i g n s c a n be u s e d t o r e d u c e t h e number o f a c t u a l s t i m u l u s p r e s e n t a t i o n s . The s i m i l a r i t y j u d g m e n t s a r e t h e n a n a l y z e d by one o f a number o f s p e c i a l l y d e s i g n e d c o m p u t e r p r o g r a m s t h a t have b e e n d e t a i l e d by S c h i f f m a n , R e y n o l d s and Young (1). Some p r o g r a m s , s u c h as INDSCAL, ALSCAL, and MULT IS CALE, n o t o n l y p r o v i d e m u l t i d i m e n s i o n a l a r r a n g e m e n t s o f m o l e c u l e s b a s e d on t h e i r f l a v o r s i m i l a r i t y but p r o v i d e q u a n t i t a t i v e measures t h a t d e l i n e a t e the i n d i v i d u a l d i f f e r e n c e s i n r e s p o n s e f o r e a c h s u b j e c t as w e l l . The

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

4.

SCHIFFMAN

35

Design of Odorants and Tastants

maps o b t a i n e d by MDS c a n be i n t e r p r e t e d by a v a r i e t y o f s t a t i s t i c a l techniques, including m u l t i p l e regression, canonical c o r r e l a t i o n , p a r t i a l l e a s t s q u a r e s , and a v a r i e t y o f o t h e r methods t o d e t e r m i n e those p r o p e r t i e s o f m o l e c u l e s r e s p o n s i b l e f o r t h e i r f l a v o r p r o p e r t i e s ; these approaches a r e d e s c r i b e d below. O b t a i n i n g r e l e v a n t p h y s i c o c h e m i c a l parameters. The c h o i c e o f p h y s i c o c h e m i c a l p a r a m e t e r s t o r e l a t e t o MDS s p a c e s i s c r u c i a l i f t h e p r o p e r t i e s f o u n d t o be m a t h e m a t i c a l l y i m p o r t a n t a r e i n d e e d appropriate chemical p r e d i c t o r s f o r future design of molecules w i t h d e s i r e d f l a v o r p r o p e r t i e s . U n f o r t u n a t e l y , we o f t e n have no i d e a what p h y s i c o c h e m i c a l p r o p e r t i e s a r e i n d e e d i m p o r t a n t , a l t h o u g h many o f t h e p a r a m e t e r s d e s c r i b e d i n t h e e x a m p l e s b e l o w as w e l l a s t h o s e g i v e n i n T a b l e I ( s e e 2* 2) probably relevant. H y d r o p h o b i c b o n d i n g , c h a r a c t e r i z e d by t h e o i l - w a t e r p a r t i t i o n c o e f f i c i e n t , has been i m p l i c a t e d i n i n t e n s e sweetness, p e r c e p t i o n o f b i t t e r n e s s , as w e l l a s o d o r i n t e n s i t y . H o w e v e r , i t w i l l be shown i n S t u d y 2 b e l o w t h a t when t h e c o n c e n t r a t i o n s o f p y r i d y l ketones a r e a d j u s t e d t o equal s u b j e c t i v e i n t e n s i t i e s , the p a r t i t i o n c o e f f i c i e n t i s not p a r t i c u l a r l y important i n discriminating qualitative differences. Short range e l e c t r i c a l p r o p e r t i e s such as d i s p e r s i o n bonding a r e probably an i m p o r t a n t s o u r c e o f s p e c i f i c i t y and may e x p l a i n why d i f f e r e n t o d o r q u a l i t i e s a r e produced by i d e n t i c a l s u b s t i t u e n t s a t d i f f e r e n t p o s i t i o n s of a molecule. D i s p e r s i o n bonding i s r e l a t e d t o t h e p o l a r i z a b i l i t y o f m o l e c u l e s and i n v e r s e l y t o t h e s i x t h p o w e r o f t h e i r s e p a r a t i o n . Longer range e l e c t r i c a l f o r c e s , whether i o n i o n , i o n - d i p o l e , o r d i p o l e - d i p o l e a r e a l s o l i k e l y t o be s i g n i f i c a n t . Hydrogen bonding has been i m p l i c a t e d i n sweet perception. C h a r g e t r a n s f e r e f f e c t s s u c h a s e l e c t r o n d o n a t i o n by e t h e r s and many a r o m a t i c compounds a s w e l l a s e l e c t r o n a c c e p t a n c e of m e r c a p t a n s i s p r o b a b l y i m p o r t a n t . S t e r i c r e p u l s i o n , though n o t b o n d i n g b u t n e v e r t h e l e s s a n o n - c o v a l e n t i n t e r a c t i o n , may a l s o contribute to specificity.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

a

r

e

Examples Four s t u d i e s a r e d e s c r i b e d here t h a t r e l a t e p h y s i c o c h e m i c a l p r o p e r t i e s t o o d o r q u a l i t y as d e f i n e d by maps d e r i v e d by m u l t i d i m e n s i o n a l s c a l i n g procedures. The m a t h e m a t i c a l procedures u s e d t o r e l a t e t h e p h y s i c o c h e m i c a l p r o p e r t i e s t o t h e maps a r e d i s c u s s e d as w e l l . S t u d y 1: Broad range o f odorants. A group o f odorants t h a t v a r i e d w i d e l y i n q u a l i t y and s t r u c t u r e w e r e a r r a n g e d o n t h e b a s i s o f o d o r s i m i l a r i t y i n a t w o d i m e n s i o n a l s p a c e shown i n F i g u r e 1 (4). The s p a c e d e r i v e d i s t w o - d i m e n s i o n a l w i t h a n a f f e c t i v e l y r a t h e r p l e a s a n t subset f a l l i n g t o t h e l e f t and a r a t h e r

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984. ]

S octanol )

Signa values or charge from molecular o r b i t a l calculations

λ

E l e c t r o s t a t i c bonding Interactions between charges : ion-ion, ion-dipole, and dipole-dipole (There i s more s p e c i f i c i t y f o r dipole-dipole interactions than those involving ions.)

[drug]water

[ d r u

Molar r e f r a c t i v i t y χ i o n i z a t i o n potential or molar r e f r a c t i v i t y alone, Hildebrand's molar a t t r a c t i o n constant, or parachor

(where Ρ *=

Log Ρ

Physical properties correlated with strength

Dispersion bonding London forces, induced dipole-induced dipole bond (Although a molecule may not have a permanent dipole, an instantaneous dipole can r e s u l t from v i b r a t i o n of electrons r e l a t i v e to the nucleus and ultimately induce a dipole i n a neighbouring molecule. Ehergy of interaction f a l l s o f f at 1/R .)

Hydrophobic bonding (A bond formed because water molecules tend to associate with one another rather than with non-polar molecules. The energy involved i n 'Squeezing out" water i s equivalent to a bond.)

Interaction

Table I. Relation o f parameters t o nonroovalent bonding (adapted from (2), see (3))

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984. Signa values or E(lemo) and E(homo) from molecular o r b i t a l calculations

van der Waals' r a d i i o r Taft Es values

S t e r i c repulsion (Intermolecular forces can be repulsive as w e l l as a t t r a c t i v e i n nature. Two molecules ultimately reach a minimi m distance between then as they approach one another, and t h i s distance i s the sum of the van der Waale' r a d i i of the interacting groups. Hence, the van der Waals' radius i s considered a measure of the e f f e c t i v e s i z e of an atom i n noncovalent interactions. Van der Waals' radius i s correlated w i t h another measure o f s t e r i c repulsion, Es.)

Sane as f o r e l e c t r o s t a t i c bonding; only f o r the atom involved i n the hydrogen bond

Charge transfer bonding (When a molecule w i t h a good electron donor comes i n contact w i t h a molecule which i s a good electron acceptor, the donor may transfer sane of i t s charge t o the acceptor.)

Hydrogen bonding (As a very small hydrogen atom i s the bridge between two electrogegative atoms, the interacting molecules can approach s u f f i c i e n t l y close t o each other t o produce an a t t r a c t i o n strong enough t o be considered a bond, rather than j u s t another dipole-dipole interaction. )

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

n-Hexyl acetate

Benzothiazole

• n-Caproic acid

J

Hydrogen sulfide

• Skatole

Cyclohexene · •Pyridine

Cyclopentene

•Propionic acid • n- Butyric acid ^Acetic acid Thiophene

•Allyl disulfide

F i g u r e 1. Two-dimensional space f o r a broad range o f odorants. Compounds l o c a t e d c l o s e t o one another have s i m i l a r odor q u a l i t y (from Réf. U).

Citronellol-*

T

alpha erpineol ornyl acetate Butyl alcohol Cyclohexane Acetone Benzene Cyclopentone Butyl alcohol , 'Carbon tetrachloride W

η-Propyl alcohol 'éapnc acid /

•Limonene Guajocol •Turpentine

•Phenol

Menthol •Hydroxycitronellal • Vanillin n-Decyl alcohol i^Geraniol Ethyl alcohol; • •n-Undecyl alcohol •Aldehyde C-14 * f n-Nonyl alcohol Nitrobenzene · φ ^ Dlphenyl methane Citral

Methyl salicylate é · θ η η ο Γ η ι ς aldehyde |6enz aldehyde

•S-8001

• Eugenol

II

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

70

η χ

70

>

Ζ Ο m το m C/3 m

>

70

Ο

70 >

α -π

>

ι

c/5

70

*Ό C H m

η ο

4.

SCHIFFMAN

Design of Odorants and Tastants

39

unpleasant group t o the r i g h t . B e c a u s e no s i n g l e f a c t o r ( s u c h as m o l e c u l a r w e i g h t , number o f d o u b l e bonds, o r d i p o l e moment) c o u l d i n d i v i d u a l l y a c c o u n t f o r q u a l i t y , S c h i f f m a n (4) w e i g h t e d a s e r i e s o f v a r i a b l e s shown i n F i g u r e 2 and was a b l e t o r e c a p t u r e 84% o f t h e i n t e r s t i m u l u s d i s t a n c e s i n a two d i m e n s i o n a l space. The m a t h e m a t i c a l p r o c e d u r e u s e d t o m a x i m i z e t h e c o n f i g u r a t i o n a l s i m i l a r i t y i n F i g u r e 1 w i t h a s p a c e g e n e r a t e d by t h e w e i g h t e d p h y s i c o c h e m i c a l p a r a m e t e r s i s b a s e d on a l e a s t squares method i n w h i c h the b a s i c m a t r i x e q u a t i o n s a r e :

« p + κ

f Ρ - DQ Ρ DQ + Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

β

Β

w h e r e : η i s t h e t o t a l number o f s t i m u l i , and Ρ i s a n n ( n - l ) / 2 c o l u m n v e c t o r whose e l e m e n t s p — r e p r e s e n t a l l t h e i n t e r s t i m u l u s d i s t a n c e s b e t w e e n s t i m u l u s i and s t i m u l u s j . The p r o x i m i t y m e a s u r e s b a s e d on w e i g h t e d p h y s i c o c h e m i c a l p a r a m e t e r s a r e g i v e n by P, an n ( n - l ) / 2 c o l u m n v e c t o r . D i s an [ n ( n - l ) / 2 ] by k s c a l a r d i s t a n c e m a t r i x whose e l e m e n t s d r e the squared d i f f e r e n c e s b e t w e e n s t i m u l u s i and J f o r e a c h k p h y s i c o c h e m i c a l p a r a m e t e r . The w e i g h t s f o r k p h y s i c o c h e m i c a l p a r a m e t e r s a r e r e p r e s e n t e d by Q, a k e l e m e n t c o l u m n v e c t o r o f w e i g h t s . Ε i s an n ( n - l ) / 2 column v e c t o r r e p r e s e n t i n g the e r r o r between the s u b j e c t i v e p r o x i m i t i e s and t h e p r o x i m i t i e s b a s e d on p h y s i c o c h e m i c a l measures. The e r r o r t o be m i n i m i z e d i s : a

l e a d i n g t o the l e a s t squares

solution

Q «= ( D ' D ) "

1

D'P

T h i s m e t h o d o l o g y c a n be h e l p f u l i n r e l a t i n g s t r i c t q u a n t i t a t i v e measures of o l f a c t o r y q u a l i t y w i t h q u a n t i t a t i v e p h y s i c o c h e m i c a l m e a s u r e s . However, t h e r e s u l t s f o r t h i s s p e c i f i c s t u d y s u g g e s t t h a t t h e r a n g e of m o l e c u l a r s t r u c t u r e s and o l f a c t o r y q u a l i t y are too broad t o d e r i v e p r a c t i c a l i n f o r m a t i o n for the r a t i o n a l design of molecules w i t h s p e c i f i c f l a v o r qualities. S t u d y 2: P y r i d y l ketones. A n a r r o w e r r a n g e o f compounds, a g r o u p o f p y r i d y l k e t o n e s , was e x a m i n e d by S o u t h w i c k and S c h i f f m a n (5). Two-dimensional c r o s s - s e c t i o n s through the t h r e e d i m e n s i o n a l s p a c e a c h i e v e d by t h e MDS a r e shown i n F i g u r e s 3a and 3b. The f i r e t d i m e n s i o n s e p a r a t e s t h e t h r e e 2 - p y r i d y l k e t o n e s t h a t h a v e p o p c o r n - n u t t y aromas f r o m t h e s i x o t h e r compounds t h a t are green-vegetable i n character.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

C O M P U T E R S IN F L A V O R A N D F R A G R A N C E R E S E A R C H

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

40

MEAN RAMAN INTENSITY

(cm-')

Figure 2. A s e r i e s o f v a r i a b l e s and t h e i r weights that were used t o account f o r the MDS space i n F i g u r e 1.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

4.

SCHIFFMAN

41

Design of Odorants and Tastants

0933 Ν

\C-CH CH 2

3



0 s /CH CH N

c

2

3



N Ν

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004



N

C-CH

\

3

^

C_CH2CH3

·

t Ν

C-CH CH CH 2

2

3

-0.933 -1.200 -1.200

1

1 -0.800

1

1 -0.400

1

1

1 0.400

1

1 0800

1

1 1.200

1.200 C\

/

CH CH 2

3

0.933 ο •ι

0667 ^ /

0.400

N

C

y /

_

C

3

H

\

/ /

^C"CH CH3 2

4 *

0.133 _

N

-0 133

C-CH CH CH 2

2

3

°vW

C H

3

Φ ν ° — " Γ

-0.400 % Ν

-0.667

C-CH CH 2

J

3

N^'-CH

N

θ \

/

CH CH CH 2

2

2

2

3

ψ

3

é

-0933 -1.200 -1.200

^ ^ ^ - ^ ^ ^C-CH CH CH

\ 3

1

1

-0.800

1

1

-0400

1

1

1

0.400

1

1

0.800

1

1

1.200

Figure 3. Two-dimensional c r o s s - s e c t i o n s through a t h r e e dimensional space achieved from odor s i m i l a r i t y data among p y r i d y l ketones (from Ref. 5 ) .

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

42

COMPUTERS IN FLAVOR A N D F R A G R A N C E RESEARCH

Two m e t h o d s w e r e u s e d t o r e l a t e a s e r i e s o f p h y s i c o c h e m i c a l d e s c r i p t o r s t o t h e s p a c e i n F i g u r e 3, a v e c t o r m o d e l a n d a n i d e a l p o i n t m o d e l ( s e e i , 6 ) . The v e c t o r m o d e l assumes t h a t t h e r e i s a d i r e c t i o n through t h e space that corresponds t o i n c r e a s i n g amounts o f a c h e m i c a l d e s c r i p t o r . The d i r e c t i o n o f t h e v e c t o r c a n be f o u n d b y l i n e a r r e g r e s s i o n t e c h n i q u e s i n w h i c h a physicochemical v a r i a b l e i s regressed over the coordinates o f the c o n f i g u r a t i o n . Weighted combinations o f t h e c o n f i g u r a t i o n c o o r d i n a t e s a r e sought t h a t best e x p l a i n the v a r i a b l e . When y^ i s t h e v a l u e o f t h e p h y s i c o c h e m i c a l p a r a m e t e r f o r stimulus i , then

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

h

b

- o

+

b

(

l *il>

+ · · · • b (x r

i r

) - b

Q

• Σ

b a=l

a

(x

i a

)

w h e r e y^ i s t h e v a l u e o f t h e v a r i a b l e y ^ i f t h e l i n e a r r e l a t i o n s h i p w e r e p e r f e c t ; b^, b2» e t c . a r e t h e r e g r e s s i o n weights; b i s t h e i n t e r c e p t ; x^ i s t h e coordinate f o r the i s t i m u l u s on the r d i m e n s i o n . The m o s t common m e t h o d t o s o l v e f o r the values o f b i s c a l l e d l e a s t squares r e g r e s s i o n i n which the c o e f f i c i e n t s a r e chosen t o m a x i m i z e G where: Q

G « Σ i-1

(y. -

The i d e a l p o i n t m o d e l i s u s e f u l when a p o i n t i n t h e s p a c e c a n be f o u n d t h a t i s m o s t l i k e t h e p h y s i c o c h e m i c a l p a r a m e t e r . Thus, t h e i d e a l p o i n t i s t h e h y p o t h e t i c a l s t i m u l u s , i f i t e x i s t e d , t h a t w o u l d c o n t a i n t h e maximum amount o f t h e p h y s i c o c h e m i c a l a t t r i b u t e . The a t t r i b u t e r e a c h e s i t s maximum a t the i d e a l p o i n t and f a l l s o f f i n a l l d i r e c t i o n s as t h e square o f t h e d i s t a n c e f r o m t h e i d e a l p o i n t . The i d e a l p o i n t i s l o c a t e d i n an MDS s p a c e by a s p e c i a l k i n d o f r e g r e s s i o n p r o p o s e d by C a r r o l l (6) t h a t c o r r e l a t e s t h e p h y s i c o c h e m i c a l a t t r i b u t e v a l u e s w i t h t h e s t i m u l u s c o o r d i n a t e s a n d a dummy v a r i a b l e c o n s t r u c t e d f r o m t h e sums o f s q u a r e s o f t h e c o o r d i n a t e s f o r e a c h p o i n t :

*i -

b

o

+

ί a a«l b

+

b

r l +

a»l

The r e s u l t s o f v e c t o r a n d i d e a l p o i n t a n a l y s e s a r e g i v e n i n T a b l e s I I a n d I I I r e s p e c t i v e l y a s a n a l y z e d b y t h e p r o g r a m PREFMAP ( 6 , 1 ) . Two kindβ o f i d e a l p o i n t s w e r e f o u n d , a p o s i t i v e i d e a l point f o r which the relevance o f a d e s c r i p t o r decreases w i t h i n c r e a s i n g d i s t a n c e o f t h e s t i m u l i f r o m t h e i d e a l p o i n t and a negative i d e a l point, f o r which therelevance increases w i t h i n c r e a s i n g d i s t a n c e f r o m t h e i d e a l p o i n t . The c o r r e l a t i o n s f o r t h e v e c t o r a n d i d e a l p o i n t a n a l y s e s a r e g i v e n i n T a b l e IV. I t

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

4.

SCHIFFMAN

Design of Odorants and Tastants

43

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

can be s e e n t h a t o d o r q u a l i t y i s h i g h l y c o r r e l a t e d f o r b o t h v e c t o r and i d e a l p o i n t m o d e l s w i t h t h e r a t i o o f t h e r e l a t i v e i n t e n s i t i e s o f t h e i o n s a t m/e 106 and 107 o b t a i n e d i n mass spectroscopic fragmentation. I n t e r a c t i o n of the r i n g nitrogen w i t h the substituent i n the 2-poeition i s responsible f o r the f r a g m e n t a t i o n pathway as w e l l as odor q u a l i t y . The Q - n.m.r. chemical s h i f t s were a l s o h i g h l y c o r r e l a t e d w i t h odor i n the v e c t o r m o d e l w i t h t h e f i r s t and s e c o n d d i m e n s i o n s o f t h e o d o r q u a l i t y space. I n f r a r e d s t r e t c h i n g f r e q u e n c i e s , l o g k' ( a measure o f column r e t e n t i o n used i n h i g h p r e s s u r e l i q u i d c h r o m a t o g r a p h y ) , and l o g Ρ ( o c t a n o l - w a t e r p a r t i t i o n c o e f f i c i e n t ) were o n l y weakly c o r r e l a t e d w i t h q u a l i t y . The h i g h c o r r e l a t i o n s f o r n e g a t i v e i d e a l p o i n t s ( u n l i k e p o s i t i v e i d e a l p o i n t s ) do n o t provide strong support f o r a physicochemical parameter.

Table I I D i r e c t i o n Cosines o f F i t t e d Physicochemical

Vectors

1

Dimension 2

3

1

-0.9966

-0.0101

-0.0815

l o g k'

2

-0.9294

-0.3662

0.0468

carbonyl s h i f t

3

-0.6541

0.7561

0.0216

carbonyl

4

-0.7974

0.5873

-0.1384

concentration

5

0.6900

-0.7227

0.0399

mass spectrum r a t i o

6

0.9133

-0.4054

-0.0387

l o

*

P

oct

frequency

S t u d y 3: P y r a z i n e s . A n o t h e r m e t h o d o f r e l a t i n g p h y s i c o c h e m i c a l p a r a m e t e r s t o a n MDS s p a c e , c a n o n i c a l c o r r e l a t i o n , was e m p l o y e d by S c h i f f m a n and L e f f i n g w e l l ( 7 ) . I n t h i s c a s e p y r a z i n e s w e r e arranged i n a 3-dimensional space on the b a s i s o f odor q u a l i t y ( F i g u r e 4 a and 4b). Then a s e t o f d e s c r i p t o r s g i v e n i n T a b l e V w e r e g e n e r a t e d by ADAPT, a c o m p u t e r s y s t e m f o r a u t o m a t e d d a t a a n a l y s e s by p a t t e r n r e c o g n i t i o n t e c h n i q u e s ( S t u p e r and J u r s , 8 ) . The s u b s t r u c t u r e s t h a t w e r e u s e d t o g e n e r a t e t h e e n v i r o n m e n t d e s c r i p t o r s a r e g i v e n i n F i g u r e 5. C a n o n i c a l r e g r e s s i o n ( s e e 1) was u s e d t o f i n d l i n e a r r e l a t i o n s h i p s b e t w e e n t h e two s e t s o f v a r i a b l e s , t h a t i s , t h e s t i m u l u s c o o r d i n a t e s o f t h e MDS s p a c e and t h e p h y s i c o c h e m i c a l attributes. The e q u a t i o n s f o r c a n o n i c a l c o r r e l a t i o n a r e : h i

'

a

k o

+

a

k i

+

· · ·

+

a

kr

^iP

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

44

COMPUTERS IN FLAVOR AND FRAGRANCE RESEARCH

Table Π Ι

Coordinates of Ideal Points

1

3

^ o c t

1

negative i d e a l point

-0.11853

0.05568 -0.02633

log k'

2

negative i d e a l point

-0.11396

0.06849 -0.03036

carbonyl s h i f t

3

negative i d e a l point

-O.05810

-0.04761 -0.03163

carbonyl frequency

4

positive i d e a l point

-0.32895

0.18924 -0.06022

concentration

5

positive i d e a l point

-0.08281

-0.01206 -0.02496

mass spectrum r a t i o

6

positive ideal point

0.12824

-0.06688 -0.04036

P

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch004

Dimension 2

Table I ? C o r r e l a t i o n s

(a)

(b)

(c)

I d e a l P o i n t Model

V e c t o r Model

1

0.9584

0.3933

2

0.9489

0.3995

3

0.9725

0.9058

c a r b o n y l frequency

4

0.4861

0.4730

concentration

5

0.9499

0.8250

masβ spectrum r a t i o

6

0.9715

0.9555

l Q

g

P

oct

l o g k' carbonyl

shift

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

SCHIFFMAN

Design of Odorants and Tastants

Η Ο^ΝΛνΗ 3

H

3

C

V

N CH V

3

7

3

C H

3

II

Η MH (H20) . + mH 0 +

n

n

m

(1)

2

where M i s the trace species to be analyzed. The resulting product ion i s further dehydrated by exposure to the ultra-pure N curtain gas and by low energy c o l l i s i o n a l activation (CID on Figure 1) brought on by the application of slight e l e c t r i c a l f i e l d s during the free-jet expansion into the analyser portion. The result i s an APCI mass spectrum dominated by molecular or pseudo-molecular ions (usually MH ) with minimal associative or dissociative product ions. The negative mode reactions are dominated by proton abstraction (2) from Bronsted acids and electron capture (3) by Lewis acids again providing pseudo-molecular and molecular ions respectively. 2

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch007

+

R- + M

» (M-H)- + RH

(2)

R- + M

> M- + R

(3)

Because of the chemical complexity of gaseous phase perfume mixtures, i t i s sometimes necessary to exploit the s e l e c t i v i t y of the APCI process. In the positive mode, only those compounds which are more basic than the reagent ion H (H 0) species are ionized. Through the addition of a reagent with higher proton a f f i n i t i e s than water the s p e c i f i c i t y of the ion source i s increased. This also applies in the negative mode as w e l l , exploiting relative gas phase a c i d i t i e s and electron a f f i n i t i e s of the reagent to the analyte in order to alter ion source specif i c i t y . Alternatively, the APCI chemistry can be altered to provide charge transfer as opposed to proton transfer ion reactions. This i s accomplished through the addition of a charge transfer reagent (such as C5H6) to generate a reagent ion "plasma" which only ionizes those compounds which have an ionization potential (IP) lower than that of the reagent's. Thus, for example, C5H5 reagent ions would only ionize compounds with an IP - Abs

a

F i g u r e 7.

Simplest

calibration

(single

wavelength).

t Abs

i

I —

F i g u r e 8.

V j

k

Wavelength — *

I n t e r f e r e n c e b e t w e e n two

spectra.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

BURNS

Near IR Reflectance Analysis

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch008

Cone

Figure 9·

Two-wavelength c a l i b r a t i o n plane.

Cone

AbsX Figure 10. space.

1

Least squares best f i t o f plane through p o i n t s

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

102

COMPUTERS IN FLAVOR A N D F R A G R A N C E RESEARCH

where η represents the number of wavelengths r e q u i r e d to account f o r a l l i n t e r f e r e n c e s ( t y p i c a l l y 3-6) and F i s a s o r t of "composite o f f s e t " ( e q u i v a l e n t t o the "b" i n the equation y = mx + b). Q

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch008

Operation In o p e r a t i o n , samples may be s o l i d s or l i q u i d s . Most s o l i d s a r e placed i n sample cups which have a quartz window on top so the r e l a t i v e p o s i t i o n under the i n t e g r a t i n g sphere i s r e p r o d u c i b l e . Semi-solids a r e sometimes handled i n open cups to s i m p l i f y cleanup. L i q u i d s a r e u s u a l l y i n j e c t e d v i a s y r i n g e i n t o a s p e c i a l c e l l f o l l o w i n g passage through a heat exchanger ( s i n c e temperature plays a bigger r o l e i n l i q u i d measurements than i t does with s o l i d s ) . A s p e c i a l l i q u i d drawer i s interchangeable with the standard s o l i d drawer. A more s o p h i s t i c a t e d instrument with a scanning monochromator i s f r e q u e n t l y used with a l e a r n i n g s e t to c r e a t e the equation which w i l l be used f o r subsequent determinations of c o n s t i t u e n t concentrations i n unknowns. The scanning instrument i s q u i t e s i m i l a r to the f i l t e r instrument except that i t can make up to 700 measurements on each sample r a t h e r than being l i m i t e d to 19 as on the simpler v e r s i o n w i t h a f i l t e r wheel. The computer which accompanies the Technicon I n f r a A l y z e r 500 system, f o r example, i s a Hewlett-Packard 1000 with a 12" screen, thermal p r i n t e r , and floppy d i s c d r i v e s . I t s program i s u s e r - f r i e n d l y and menu-driven, meaning that most of the time a s i n g l e keystroke with one f i n g e r i s s u f f i c i e n t to move l o g i c a l l y through the steps r e q u i r e d to go from sample scanning to report generation. Correlation A set of standards which covers the a n a l y t i c a l range of i n t e r e s t w i l l produce a f a m i l y of s p e c t r a i n which ( i t i s hoped) the change i n absorbance at one or more wavelength w i l l c o r r e l a t e with the changing concentrations of t h e a n a l y t e . In the simplest case, the computer program m e t h o d i c a l l y examines these changes at each wavelength where measurements o f absorbance were made ( t y p i c a l l y 350 data p o i n t s over the range 1100-2500 nm at 4 nm increments) and i d e n t i f i e s that wavelength f o r which the c o r r e l a ­ t i o n i s h i g h e s t . Holding t h i s wavelength constant, i t then searches f o r the next best wavelength ( i . e . , the next highest c o r r e l a t i o n ) , then the 3rd best, e t c . , u n t i l no s i g n i f i c a n t im­ provement i s seen. There a r e other search s t r a t e g i e s (such as those i n v o l v i n g a t e s t of a l l combinations of two or more wave­ lengths) , but most a r e based upon the improvement o f the c o r r e l a ­ t i o n c o e f f i c i e n t of repeated r e g r e s s i o n analyses (an expression with which the reader should be f a m i l i a r , s i n c e i t i s a most b a s i c c a l c u l a t i o n i n s t a t i s t i c s and i s fundamental t o an under­ standing of c o r r e l a t i o n transform spectroscopy).

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch008

8.

BURNS

103

Near IR Reflectance Analysis

C a l i b r a t i o n s which are generated on the t o p - o f - t h e - l i n e instrument can be t r a n s f e r e d to a l e s s expensive instrument f o r r o u t i n e use by l e s s s k i l l e d personnel. Computer programs handle a l l the d e t a i l s , such as a f i l t e r transform (to make data c o l l e c t e d v i a monochromator look l i k e i t was c o l l e c t e d v i a i n t e r f e r e n c e f i l t e r ) , a data r e d u c t i o n (to l i m i t a l a r g e data base of as many as 700 wavelengths to only those 19 wavelengths represented by t h e f i l t e r s i n a smaller instrument), and a c a l i b r a t i o n (the c r e a t i o n o f an equation by examining known standards and maximizing the c o r r e l a t i o n as described above). The computer p r i n t o u t s are lengthy and f i l l e d with enough s t a t i s t i c s to s a t i s f y the most s o p h i s t i c a t e d user, but f o r purposes of i l l u s t r a t i o n , w e ' l l l i m i t our example t o the h i g h l i g h t s of the data from a 3-wavelength search on wheat samples f o r the c o n s t i t u e n t p r o t e i n . I t i s noteworthy that such i l l - d e f i n e d molecules as p r o t e i n s can be determined so r a p i d l y , with so l i t t l e sample p r e p a r a t i o n (a few seconds of g r i n d i n g ) , with such low e r r o r s ( t y p i c a l l y with standard e r r o r s of p r e d i c t i o n o f about 0.2 over the range 9-18 percent p r o t e i n ) .

H i g h l i g h t s of P r i n t o u t

STANDARD ERROR CORRELATION COEFFICIENT COEFFICIENTS:

WAVELENGTHS

B(0) B(l) B(2) B(3)

NUMBER OF WAVELENGTHS 3 2 1 0.55 1.47 7.11 .997 .999 .923 -105 260

120 1389 -1391

1440

1664 1824

90 775 -962 545 1652 1820 1292

As more wavelengths are added, the standard e r r o r diminishes and approaches some minimum v a l u e , and the c o r r e l a t i o n c o e f f i c i e n t increases and approaches u n i t y . The equation becomes more comp l i c a t e d with the a d d i t i o n of more c o e f f i c i e n t s so that i t may b e t t e r d e f i n e the analyte's c o n c e n t r a t i o n . F e a s i b i l i t y Studies with F l a v o r Constituents In order t o e s t a b l i s h the p o t e n t i a l of NIRA i n any given f i e l d , three methodical steps are r e q u i r e d : a f e a s i b i l i t y study (with low, medium, and high l e v e l s of the sought a n a l y t e ) , a f u l l c a l i b r a t i o n (with 25-30 samples covering the range of a n a l y t i c a l i n t e r e s t ) , and c o n f i r m a t i o n that the generated equation works i n

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch008

104

COMPUTERS IN FLAVOR AND F R A G R A N C E RESEARCH

the r e a l world (by c o r r e c t l y determining a group of about 20 r e f e r e n c e samples which were not used i n the development of the equation). The reader must understand the p r o p r i e t a r y nature of i n f o r mation obtained from many companies, and the requirement that names and a c t u a l c o n s t i t u e n t l e v e l s be a r b i t r a r i l y scaled to conc e a l the i d e n t i t i e s . A c c o r d i n g l y , a l l values i n the f o l l o w i n g examples have been r e s c a l e d to set the lowest l e v e l s to ten. One manufacturer of microencapsulated f l a v o r s wished to analyze h i s product f o r moisture and o i l essence by NIRA and compare r e s u l t s with h i s r e f e r e n c e method. P r e l i m i n a r y r e s u l t s were as shown i n Table I and Table I I . For the o i l essence, the lower r e l a t i v e standard d e v i a t i o n (RSD) f o r the I n f r a A l y z e r method i n d i c a t e s a more p r e c i s e measurement than was obtained with the manual method upon which i t was based, although accuracy would depend upon that of the r e f e r e n c e method. The slopes are near 1.0, showing good c o r r e l a t i o n between the two methods. The NIRA method gives s l i g h t l y higher r e s u l t s w i t h a range ( f o r "A") of 10.8-11.5 compared with 10.0-11.0 f o r the r e f e r e n c e . RSD's are o f t e n 1% times higher than the r e f e r e n c e method, so the moisture determinations are e q u a l l y good. For a f e a s i b i l i t y study with such a small data s e t , these f i n d i n g s are w i t h i n experimental e r r o r and suggest that f u r t h e r work should be done to e s t a b l i s h a robust c a l i b r a t i o n . In another study, f l a v o r beads were analyzed f o r carbohydrate and water. In t h i s case, however, the product was made i n a v a r i e t y of c o l o r s (white, red, yellow, green, blue) which must be regarded as i n t e r f e r e n c e s , s i n c e the l e v e l of a n a l y t e was u n r e l a t e d to the v i s i b l e c o l o r . I t r e q u i r e d 25 samples and f i v e wavelengths to account f o r the i r r e l e v a n t c o l o r s and produce these r e s u l t s : Level Range C o r r e l a t i o n Coef Normal Low

10-10.8 10-10.3

0.95 0.90

I t would normally r e q u i r e more than 25 samples i n a l e a r n i n g set to produce a r e a l l y good equation. Despite t h i s l e s s than optimum number, the search a l g o r i t h m was able to i d e n t i f y a comb i n a t i o n of f i v e wavelengths (from the 19 d i s c r e t e f i l t e r s i n the I n f r a A l y z e r 400 system) and generate an equation which could e x t r a c t carbohydrate and water concentrations from s p e c t r a which came from h i g h l y c o l o r e d m a t e r i a l . C l e a r l y , the NIRA technique shows promise f o r t h i s a n a l y s i s . F i n a l l y , a food processor presented 11 samples f o r a f e a s i b i l i t y study on cheese powder to i n d i c a t e i t s usefulness i n q u a n t i f y i n g c a l o r i c content as w e l l as such c o n s t i t u e n t s as moisture, p r o t e i n , f a t , ash, and carbohydrate. The composite absorbance t r a c i n g s of F i g u r e 11 produced a 3-wavelength equation, the r e s u l t s of which are shown i n Table I I I . The

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Near IR Reflectance Analysis

8. BURNS

105

Table I . O i l Essence I n f r a A l y z e r Method RSD Level (%) (%)

Slope

1 2 3 4

10.0 11.0 10.6 10.9

4.2

10.9 11.5 10.9 11.0

10.8 11.3 10.9 11.1

2.2

1.04

Β

1 2 3

11.0 10.3 10.0

4.8

10.5 9.9 9.6

10.6 10.0 10.3

3.8

0.97

C

1

10.0

9.6

9.6

A

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch008

Reference Method Level RSD (%) (%)

ι

Product

0.96

Moisture

Table I I .

Reference Method RSD Level (%) (%)

Product

I n f r a A l y z e r Method RSD Level (%) (%)

Slope

A

1 2 3 4

15.0 10.0 12.5 12.5

16.3

15.0 11.6 12.8 12.5

15.6 10.6 13.1 12.8

12.6

1.04

Β

1 2 3

10.0 13.3 13.9

17.0

10.8 12.8 13.0

11.4 13.6 13.0

8.7

1.00

C

1

10.0

11.5

11.5

Table I I I .

1.15

Cheese Powder

Constituent

Range

Calories Moisture Protein Fat Ash Carbohydrate

10-10.3 10-12.9 10-10.3 10-11.3 10-10.5 10-10.4

Corr Coef

0.81 0.93 0.83 0.70 0.89 0.70

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

RSD (%) 0.6 3.6 0.5 3.0 4.5 0.9

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch008

Figure 11. Spectra o f cheese powder (composite o f 25 samples learning s e t ) .

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

in

8.

BURNS

Near IR Reflectance Analysis

107

Instrument used here was the I n f r a A l y z e r 500 system with the scanning monochromator. Except f o r the carbohydrate a n a l y s i s , a l l analyses (even the r a t h e r non-descript c a l o r i c values) can l i k e l y be handled q u i t e w e l l i f the standard d e v i a t i o n of the r e f e r e n c e method could be improved.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch008

Conclusions Near I n f r a r e d Reflectance A n a l y s i s (NIRA) i s f i n d i n g ever i n c r e a s i n g use i n the food i n d u s t r y as a r a p i d and n o n - d e s t r u c t i v e t e s t i n g technique. Sample p r e p a r a t i o n i s u s u a l l y simple, and sometimes i s n ' t r e q u i r e d at a l l . With a good r e f e r e n c e method upon which t o base the NIRA method, one can expect good c o r r e l a t i o n , accuracy, and p r e c i s i o n . Methods developed on a s o p h i s t i cated scanning instrument can be used by l e s s s k i l l e d personnel on lower cost f i l t e r instruments a t the most u s e f u l s i t e s (e.g., manufacturing p l a n t , QC l a b ) . The c a l i b r a t i o n s are u s e r t r a n s f e r a b l e among instruments w i t h i n the I n f r a A l y z e r l i n e . There i s so much i n f o r m a t i o n present i n the NIR region that NIRA i s probably destined to become a p r e f e r r e d a n a l y t i c a l method. RECEIVED A p r i l 6, 1984

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

9 Multivariate and Gas Chromatographic Techniques in Flavor Research M. A. JELTEMA, B. W. GOOD, F. S. HSU, and M. E. PARRISH

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

Philip Morris U.S.A., Research Center, P.O. Box 26583, Richmond, VA 23261

A s u b s t a n t i a l e f f o r t has been made t o develop sound o b j e c t i v e measurements f o r q u a l i t y evaluat i o n o f a consumer product's f l a v o r , aroma and taste. I t is apparent that progress i n f l a v o r research was dependent on advances i n a n a l y t i c a l s e p a r a t i o n techniques as w e l l as the a p p l i c a t i o n of computer technology to provide not j u s t data management, but more importantly, d i r e c t i o n f o r f u r t h e r research. The growth and success o f t h i s research a c t i v i t y has been t i e d c l o s e l y t o the development o f gas chromatography and more recentl y t o the a p p l i c a t i o n o f p a t t e r n r e c o g n i t i o n techniques f o r data a n a l y s i s and s i m p l i f i c a t i o n . Input from the a n a l y t i c a l , sensory and statistical d i s c i p l i n e s have r e s u l t e d i n the development o f techniques which are p l a y i n g an i n c r e a s i n g l y important r o l e i n s o l v i n g q u a l i t y assurance and product development problems. The impact o f the d i r e c t i o n o f t h i s research on our work, as w e l l as two s p e c i f i c examples o f the use o f m u l t i v a r i a t e analyses techniques f o r tobacco research will be discussed. A n a l y t i c a l Data A n a l y s i s . The development and commercialization of the gas chromatograph i n the mid 1950 s had a dramatic e f f e c t on f l a v o r research because the technique made i t p o s s i b l e t o o b t a i n o b j e c t i v e measurements o f the numerous compounds which made up the f l a v o r o f the product under i n v e s t i g a t i o n . Data a n a l y s i s was reasonably simple and s t r a i g h t f o r w a r d , as the number o f r e s o l v e d peaks was s m a l l . However, as chromatographic techniques were r e f i n e d and high r e s o l u t i o n c a p i l l a r y columns and microprocessor c o n t r o l l e d GC's were introduced, the use o f computers and m u l t i v a r i a t e a n a l y s i s techniques have become e s s e n t i a l f o r data a n a l y s i s and r e d u c t i o n . 1

0097-6156/ 84/ 0261 -0109506.50/ 0 © 1984 American Chemical Society

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

110

COMPUTERS IN FLAVOR AND FRAGRANCE

RESEARCH

These most recent developments have played an important r o l e i n our research i n v e s t i g a t i o n s on c i g a r e t t e smoke. As a consumer product, c i g a r e t t e smoke can be considered a very complex and dynamic f l a v o r system. One f a c e t of our research e f f o r t s has been to develop a n a l y t i c a l methods f o r c h a r a c t e r i z ing the c i g a r e t t e smoke generated from d i f f e r e n t tobaccos used i n the product. For addressing the task of d i f f e r e n t i a t i n g c i g a r e t t e types, s t a t i s t i c a l analyses are necessary f o r data r e d u c t i o n , p a t t e r n e x t r a c t i o n and ranking the importance of the gas chromatographic peaks. This approach has been used by researchers i n many d i f f e r e n t areas (1-5). These techniques have been a p p l i e d s u c c e s s f u l l y i n our l a b o r a t o r y to the low b o i l i n g v o l a t i l e s comprising the gas phase (6) and the higher b o i l i n g v o l a t i l e s e x t r a c t e d from a f i l t e r pad with Freon-11, followed by t r i m e t h y l s i l y l a t i o n and automatic sampling onto a fused s i l i c a c a p i l l a r y column (7). Our emphasis continues to be the development of a n a l y t i c a l techniques f o r o b t a i n i n g r e p r o d u c i b l e chromatographic c a p i l l a r y p r o f i l e s which r e f l e c t to the g r e a t e s t extent the p o s s i b l e f l a v o r a t t r i b u t e s of c i g a r e t t e smoke and tobacco, and to become more experienced i n the proper use of p a t t e r n r e c o g n i t i o n t e c h niques f o r data a n a l y s i s . Recently, however, we a l s o have recognized the need to c o r r e l a t e a n a l y t i c a l data to sensory data. J u s t as p a t t e r n r e c o g n i t i o n has been demonstrated to be v a l u a b l e i n the d i s t i n c t i o n of samples based on t h e i r gas chromatographic p r o f i l e , p a t t e r n r e c o g n i t i o n a l s o i s u s e f u l i n the a n a l y s i s of sensory data. Sensory. Although the b a s i s f o r m u l t i v a r i a t e analysis was developed i n the e a r l y 1900 s, i t s use i n sensory a n a l y s i s i s r e l a t i v e l y recent. These types of s t a t i s t i c s , however, have been v a l u a b l e i n d e a l i n g with two fundamental problems which occur i n sensory t e s t i n g . F i r s t there are d i f f i c u l t i e s encountered when one attempts to breakdown complex sensory parameters i n t o s i n g l e semantic terms which can be r a t e d , and second i t i s d i f f i c u l t to achieve the goal of every p a n e l i s t having the same i n t e r n a l understanding of each term. Approaches to minimize these d i f f i c u l t i e s i n c l u d e d : 1) e v a l u a t i o n of semantic terms used by the panel to determine i f the v a r i a b l e s are unique or can be condensed to a new s e t of unique v a r i a b l e s ; 2) evaluat i o n of the p a n e l i s t s use of semantic terms to determine incons i s t e n c i e s as w e l l as the r e l a t i v e importance of the terms to food q u a l i t y or d i s c r i m i n a t i o n . ( 8 ) While some researchers f e e l t h a t i n d i v i d u a l components making up the sensory response need to be c h a r a c t e r i z e d (9), others f e e l that methods d e a l i n g with the composite s e n s a t i o n are more a p p r o p r i a t e . For t h i s reason methods which do not r e l y on i n t e r n a l r e p r e s e n t a t i o n of terms have been a p p l i e d . One such method i s multidimensional s c a l i n g (MDS), which t r e a t s data based on a persons' t o t a l p e r c e p t i o n of the d i s s i m i l a r i t y between objects.(10) f

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

9.

JELTEMA ET AL.

111

Multivariate and GC Techniques

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

One o f the major uses o f m u l t i v a r i a t e techniques has been the d i s c r i m i n a t i o n o f samples based on sensory scores, which a l s o has been found t o provide i n f o r m a t i o n concerning the r e l a t i v e importance o f sensory a t t r i b u t e s . Techniques used f o r sensory d i s c r i m i n a t i o n i n c l u d e f a c t o r a n a l y s i s , d i s c r i m i n a n t a n a l y s i s , r e g r e s s i o n a n a l y s i s , and multidimensional s c a l i n g (8, 10-15). S p e c i f i c t o our research, the multidimensional techniques such as MDS, f a c t o r a n a l y s i s , c a n o n i c a l c o r r e l a t i o n and regress i o n a n a l y s i s have been used not only to analyze sensory and a n a l y t i c a l data, but a l s o to perform c o r r e l a t i o n between the two sets o f data. O b j e c t i v e and Sensory C o r r e l a t i o n s . The wish to correlate p h y s i c a l parameters t o sensory parameters i s not new. I t was f i r s t used to c o r r e l a t e s u b j e c t i v e and o b j e c t i v e measurements o f c o l o r and t e x t u r e . I t was not u n t i l the I960 s that researchers began t o c o r r e l a t e f l a v o r with gas chromatographic data, basi c a l l y due t o the l a c k o f good a n a l y t i c a l techniques. Since that time, research i n the area has a c c e l e r a t e d . Although some researchers have been s u c c e s s f u l i n r e l a t i n g the c o n c e n t r a t i o n of s i n g l e components t o sensory response (16,17), i t has been recognized that f o r most a p p l i c a t i o n s , a m u l t i v a r i a t e approach i s necessary s i n c e sensory response i s due t o i n t e r a c t i o n s o f a multitude o f components. F o r i n s t a n c e , i n a mixture o f 12 compounds there i s a p o s s i b i l i t y o f 4095 sources o f combinations which could generate a sensory response (15). T h i s problem i s compounded by a l a c k o f knowledge o f how we respond t o sensory mixtures and models t o p r e d i c t such a response. The d i f f i c u l t i e s o f r e l a t i n g o b j e c t i v e measurements t o sensory scores are compounded f u r t h e r by the f a c t that the presence o f one compound may suppress the p e r c e p t i o n of another or the sensory response may be due t o compounds which are transparent t o the a n a l y s i s , thereby making c o r r e l a t i o n s more d i f f i c u l t t o p e r c e i v e . A l s o , while a n a l y t i c a l response i s l i n e a r , t h i s i s not t r u e o f sensory response; t h e r e f o r e l i n e a r f u n c t i o n s may not always be approp r i a t e (18). While the problem o f r e l a t i n g sensory response t o a simple mixture i s d i f f i c u l t , t h i s i s compounded when e f f o r t s are made to r e l a t e sensory response to the thousands o f components cont a i n e d i n c i g a r e t t e smoke. As with many food systems, the d i f f e r e n c e s are e s s e n t i a l l y q u a n t i t a t i v e r a t h e r than q u a l i t a tive. The use o f m u l t i v a r i a t e techniques are e s s e n t i a l s i n c e they are designed t o deal with a l l the peaks o f a chromatographic p r o f i l e . F o r t u n a t e l y , many o f these components are h i g h l y c o r r e l a t e d with others, and t h e r e f o r e simpler v a r i a b l e s can be e x t r a c t e d through techniques such as f a c t o r a n a l y s i s . Our e f f o r t s t o c o r r e l a t e instrumental and sensory data have i n i t i a l l y centered around two areas, both o f which are somewhat simpler than the p r e d i c t i o n of sensory response t o c i g a r e t t e smoke. The f i r s t was aimed a t developing a model f o r a q u a l i t y 1

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

112

COMPUTERS IN FLAVOR A N D FRAGRANCE

RESEARCH

assurance (Q.A.) procedure where i t i s e s s e n t i a l to know i f a misforraulation would f a l l out o f a p r e s e t sensory s p e c i f i c a t i o n . The second d e a l t with tobacco aroma where we were i n t e r e s t e d i n d i s c o v e r i n g i f d i f f e r e n c e s which could be seen a n a l y t i c a l l y between types of tobacco could be r e l a t e d to d i f f e r e n c e s found by a sensory panel. Each of these problems i s d i f f e r e n t , t h e r e f o r e the use of d i f f e r e n t sensory as w e l l as a n a l y t i c a l techniques was necessary.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

A p p l i c a t i o n to Q u a l i t y Assurance

(Q.A.)

Implementation o f a sensory q u a l i t y c o n t r o l s p e c i f i c a t i o n can be extremely d i f f i c u l t because p r o d u c t i o n w i l l o f t e n exceed panel capabilities. A model which p r e d i c t s panel t e s t r e s u l t s with even moderate success would be extremely v a l u a b l e i n a v o i d i n g the acceptance o f poor p r o d u c t i o n l o t s . For q u a l i t y c o n t r o l problems, i t i s o f t e n s u f f i c i e n t to e s t a b l i s h that a d i f f e r e n c e has been detected which i s great enough to exceed present specifications. I t i s g e n e r a l l y assumed that a r e l a t i o n s h i p (not n e c e s s a r i l y l i n e a r ) e x i s t s between percent c o r r e c t response i n d i f f e r e n c e t e s t s and the a c t u a l degree of d i f f e r e n c e between the stimuli (19). Therefore, i n t h i s study r e p l i c a t e d t r i a n g l e t e s t s were ured as the sensory i n p u t . The samples used f o r the study i n c l u d e d blends o f 5 multicomponents where the l e v e l o f each multicomponent i n the blend was i n d i v i d u a l l y a l t e r e d . M o d i f i c a t i o n s ranged from -100% (missing) to + 100% (2 times reference l e v e l ) o f the c o n c e n t r a t i o n of each multicomponent i n the reference mixture. Sensory. T r i a n g l e t e s t s were performed by a panel of t e n experienced s u b j e c t s . Each subject performed t e n t r i a n g l e t e s t s a t 60 second i n t e r v a l s so that a l l subjects evaluated a l l experimental t r i a n g l e t e s t comparisons. A triangle test typically c o n s i s t e d o f t e n randomly p l a c e d combinations of AAB and ABB. P a n e l i s t s were asked to choose the odd sample based on odor. I t was found that the i n i t i a l three r e p l i c a t i o n s were p a r t o f a l e a r n i n g process by the p a n e l i s t , t h e r e f o r e the f i r s t three t e s t r e s u l t s were not used. As shown i n F i g u r e 1, while i n d i v i d u a l r e p l i c a t i o n s ( i e . 10 subjects performing a t r i a n g l e t e s t ) v a r i e d s u b s t a n t i a l l y about the s i g n i f i c a n c e l e v e l , percent cumulative response showed a c o n s i s t e n t mean c o r r e c t panel response a f t e r the f i r s t few t r i a l s , with l i t t l e evidence of f a t i g u e . Analytical. Samples were chromatographed on a Hewlett-Packard 5880A gas chromatograph which was f i t t e d with a 30M fused s i l i c a c a p i l l a r y column (DB5) and an automatic sampler. The GC was interfaced t o a Hewlett-Packard 3354 Laboratory Automation System (LAS). Raw data was a u t o m a t i c a l l y t r a n s f e r r e d t o the LAS where peaks were s e l e c t e d by r e t e n t i o n time, i n t e g r a t e d and stored i n a processed f i l e . Processed data was then t r a n s f e r r e d

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

JELTEMA ET AL.

Multivariate and GC Techniques

Figure 1.

Panel response on t r i a n g l e

tests.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

114

COMPUTERS IN FLAVOR A N D FRAGRANCE RESEARCH

over t e r m i n a l l i n e s to a DECSYSTEM 2060 f o r s t a t i s t i c a l a n a l y s i s using BMDP programs (20). Since the sensory data c o l l e c t e d i n v o l v e d degree o f sample d i f f e r e n c e from a reference, i t was f e l t that the a n a l y t i c a l data should be analyzed i n a s i m i l a r manner. In cases where some peaks making up a multicomponent mixture are known t o be s p e c i f i c to t h a t mixture, t h i s i s a r e l a t i v e l y simple matter. In such cases, the peak areas of the known components can be compared to a reference and average percent d i f f e r e n c e c a l culated. However, i f i t i s not p o s s i b l e to p i c k out peaks that are c l e a r l y s p e c i f i c to a s i n g l e multicomponent mixture, a more s o p h i s t i c a t e d technique such as f a c t o r a n a l y s i s i s r e q u i r e d . There are circumstances where a l l peaks are common t o each multicomponent mixture, i . e . q u a l i t a t i v e l y s i m i l a r but q u a n t i t a tively different. Also there are cases where peaks are found only i n one o f the multicomponent mixtures, but i t i s not c l e a r to which mixture they belong. In these cases f a c t o r a n a l y s i s i s r e q u i r e d to e x t r a c t patterns that are c h a r a c t e r i s t i c o f the s p e c i f i c multicomponent mixtures. A n a l y t i c a l concentrations o f each o f the multicomponent mixtures are then c a l c u l a t e d as a s e t of f a c t o r scores where each score i s d i r e c t l y p r o p o r t i o n a l to the a c t u a l c o n c e n t r a t i o n o f each multicomponent mixture. Data Reduction. Data r e d u c t i o n was performed i n two separate ways: f a c t o r a n a l y s i s and peak area r a t i o s . A t y p i c a l chromatogram o f the component blend used i s shown i n F i g u r e 2. F a c t o r a n a l y s i s (20) was performed to reduce and s i m p l i f y the data. The chromatographic data was thereby reduced from the 31 s e l e c t ed peaks to 5 f a c t o r s . Each f a c t o r was p r i n c i p a l l y comprised o f components which o r i g i n a t e d from one multicomponent mixture but the a n a l y s i s a l s o i n d i c a t e d which components were found i n more than one o f the multicomponent mixtures. The f a c t o r scores, which were c a l c u l a t e d f o r each sample, gave a composite score f o r the c o n c e n t r a t i o n o f each i n d i v i d u a l multicomponent. Samp l e s were t h e r e f o r e separated by p l o t t i n g t h e i r f a c t o r scores (Figure 3 ) . F i g u r e 3 demonstrates the s e p a r a t i o n t h a t was achieved by p l o t t i n g samples on f a c t o r 1 (x a x i s ) versus f a c t o r 2 (y a x i s ) . The samples which contained +100% o f multicomponent 1 r e c e i v e d a high p o s i t i v e score on f a c t o r 1 (~3.0), while samples c o n t a i n i n g -100% o f that multicomponent received a high negative score. Likewise, samples which contained +100% or -100% o f component 2 r e c e i v e d high p o s i t i v e and negative scores on f a c t o r 2. Smaller a l t e r a t i o n s i n c o n c e n t r a t i o n (±20%) produced lower f a c t o r scores (±0.6). Samples i n the center which were not separated from each other i n c l u d e d the reference ( f a c t o r score ~0.0) and samples which contained the same conc e n t r a t i o n o f multicomponents 1 and 2 as i n the reference. These samples had been a l t e r e d i n the c o n c e n t r a t i o n o f 3, 4 or 5; however, t h i s d i d not a f f e c t t h e i r f a c t o r scores f o r f a c t o r 1 and 2.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

F i g u r e 2.

Chromatographic p r o f i l e o f a r e f e r e n c e .

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

L/1

s

•δ* ε

δ­ α

Ci

S'

Ï

r

>

H

m q m > m

(_



COMPUTERS IN FLAVOR A N D FRAGRANCE RESEARCH

+100%

0

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

0

0 + 20%

o à-ioo%

_

2 0 %

CP

• UNALTERED IN 1 OR 2 AMULTICOMPONENT 1 0 MULTICOMPONENT 2 I 4

-2

1+20%

/Ρ^ΏΏ

00

M+100%

-20%

-100%

I

I

0

2

4

FACTOR 1 Figure 3.

Factor score p l o t s o f a l t e r e d

samples.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

9.

JELTEMA ET A L .

117

Multivariate and GC Techniques

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

In the case o f t h i s multicomponent mixture, i t was known which peaks belonged t o which multicomponent. Therefore peak areas o f known components could be e a s i l y compared t o a r e f erence and the average percent d i f f e r e n c e c a l c u l a t e d . This was done f o r each sample with a simple program. This program was a u t o m a t i c a l l y i n i t i a t e d a t the time a sample was chromatographed by the HP3354 LAS. An example o f the output i s shown i n Table I. This sample was formulated t o c o n t a i n -100% (or 0%) o f a multicomponent C and -50% (or 1/2) the normal c o n c e n t r a t i o n o f a multicomponent A. From the l i s t o f missing peaks and the f r a c t i o n o f i n d i v i d u a l peaks, t h i s was found to be t r u e . The extra peaks which are l i s t e d can be used to p i n p o i n t p o s s i b l e sources of contamination.

Table I. Computer Output from C a l c u l a t i o n o f Percent D e v i a t i o n of Each Component Based on a Reference Mixture

** FINGERPRINT SPECIFICATION PROGRAM ** SAMPLE NAME: 50A W/0 C ** MISSING PEAKS ** EXPECTED RT 11.59 13.38

NAME CI C2

** EXTRA PEAKS ** ACTUAL RT 3.3 15.77 PEAK 1 2 3 4 5 6 7

TIME 2.55 3.79 11.37 16.19 19.83 20.64 29.81

FRACTION 0.52 0.52 0.52 1.05 1.04 0.21 0.54

NAME Al A2 A3 Bl B2 BC1 A4

P r e d i c t i o n o f Panel Response. I t was b e l i e v e d that the best t h e o r e t i c a l model t o p r e d i c t sensory response from the a n a l y t i c a l data would be a quadratic f u n c t i o n (Figure 4 ) , s i n c e as the c o n c e n t r a t i o n o f a multicomponent was e i t h e r increased or decreased from the reference, more of the panel should be able to d i f f e r e n t i a t e the odd sample from the reference.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

118

COMPUTERS IN FLAVOR A N D F R A G R A N C E RESEARCH

ANALYTICAL X

Figure k. T h e o r e t i c a l panel response t o a l t e r a t i o n s i n the conc e n t r a t i o n o f a multicomponent mixture.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

9.

JELTEMA ET AL.

Multivariate and GC Techniques

119

The t h e o r e t i c a l minimum of t h i s curve i s a 33.3% c o r r e c t response which represents a chance response of s e l e c t i n g the odd sample when a r e f e r e n c e from one l o t i s compared to a reference from another l o t . This p r e d i c t i o n u s i n g e i t h e r f a c t o r scores or a n a l y t i c a l d e v i a t i o n from a r e f e r e n c e i s shown i n F i g u r e 5 f o r multicompo­ nent 1. The types of curves obtained f o r the other multicomponents were s i m i l a r , although the steepness v a r i e d according to the s e n s i t i v i t y of the p a n e l i s t s f o r changes i n each multicompo­ nent. In order to s e t up an a n a l y t i c a l method based on f a c t o r scores or a n a l y t i c a l d e v i a t i o n i n a Q.A. environment, s e v e r a l precautions would need to be followed. For i n s t a n c e , changes i n raw m a t e r i a l s would need to be monitored and e i t h e r new ref­ erence m a t e r i a l s used or f a c t o r analyses repeated. Changes i n chromatographic c o n d i t i o n s would r e q u i r e a system which allowed constant updating to avoid s h i f t i n f a c t o r s c o r e s . While the peak r a t i o technique i s the simpler of the two, f a c t o r scores provide a more s o p h i s t i c a t e d method which can be used where overlap of components does not lend i t s e l f to the simpler meth­ ods. These two techniques f o r d e r i v i n g the c o n c e n t r a t i o n s of the multicomponents i n a mixture were shown to p r e d i c t sensory response when one multicomponent was a l t e r e d . However i t a l s o could be used i n a m u l t i d i m e n s i o n a l formula to p r e d i c t response when more than one multicomponent has been a l t e r e d . A p p l i c a t i o n to Product Development The second study focused on the use of a n a l y t i c a l / s e n s o r y cor­ r e l a t i o n s f o r a product development a p p l i c a t i o n . Since the sensory c h a r a c t e r of a c i g a r e t t e i s due i n p a r t to v o l a t i l i z a ­ t i o n products which are d i s t i l l e d o f f of the tobacco i n the rod by the heat of the c o a l , d i f f e r e n c e s i n the odor of tobaccos may c o r r e l a t e to sensory d i f f e r e n c e s upon smoking. Analytical. F i v e tobaccos were a v a i l a b l e i n a cased ( f l a v o r e d ) and uncased (unflavored) mode: 100% b r i g h t , (A) 100% b u r l e y , (B) 100% o r i e n t a l , (C) 33.3% b r i g h t / 3 3 . 3 % burley/33.3% o r i e n t a l and 60% b r i g h t / 30% b u r l e y / 1 0 % o r i e n t a l . A 10 gram sample of c o n d i t i o n e d tobacco (23°C, 60% RH, 48 hrs) was p l a c e d i n a 70°C water-jacketted desorbing chamber. Two l i t e r s of helium were passed through the tobacco over a 20 minute p e r i o d , t r a p p i n g the v o l a t i l e s on an a i r sampling probe c o n t a i n i n g 500 mg of Tenax 35/60 mesh. The Tenax c o n t a i n i n g the tobacco v o l a t i l e s was p l a c e d on a CDS 320 Concentrator equipped with a thermal desorber to accept the 6 χ 1/2 probe. A l i q u i d n i t r o g e n cryogenic t r a p was c o n t r o l l e d by the LAS f o r f u r t h e r f o c u s i n g . Temperature pro­ gramming and data a c q u i s i t i o n s were performed by the P e r k i n Elmer Sigma 1 and the Hewlett-Packard 3354 LAS r e s p e c t i v e l y . M

M

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

120

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

COMPUTERS IN FLAVOR A N D FRAGRANCE RESEARCH

ι -100

1 -75

1 -50

1 1 Γ " -25 0 25 ANALYTICAL %

50

75

100

100

-I 0 1 FACTOR 1 SCORE Figure 5. Panel response t o a l t e r a t i o n s i n the concentration o f multicomponent mixture one. Top, based on a n a l y t i c a l d e v i a t i o n from a reference: R = 0 . 8 6 ; and bottom, based on f a c t o r scores: R = 0.91. 2

2

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

9.

JELTEMA ET AL.

Multivariate and GC Techniques

121

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

Twenty-two peaks, 5 o f which were designated as r e t e n t i o n reference peaks, were s e l e c t e d a r b i t r a r i l y without i d e n t i f i c a t i o n based on r e p r o d u c i b i l i t y o f i n t e g r a t i o n over the e n t i r e data s e t . Data Reduction. A gas chromatographic p r o f i l e o f the Tenaxtrapped v o l a t i l e s from a 33.3% bright/33.3% burley/33.3% o r i e n t a l tobacco sample i s shown i n F i g u r e 6. The 22 s e l e c t e d peaks are designated. A s t a t i s t i c a l program, BMDP4M (20) was a p p l i e d t o the 22 peaks from the t e n tobacco types and four b r i g h t / b u r l e y blends individually. The purpose o f f a c t o r a n a l y s i s was to f i n d which peaks c o r r e l a t e with each other and what i n f o r m a t i o n a peak (or group o f peaks) might impart as to tobacco type, blend or casing. In the a n a l y s i s o f the t e n tobacco types, f i v e f a c t o r s were c a l c u l a t e d that explained 81% o f the variance o f the data set. P l o t s o f f a c t o r scores i n F i g u r e 7 make i t easy t o see the u t i l i t y o f c e r t a i n peaks f o r d i s t i n g u i s h i n g the tobacco types and casings. F a c t o r 1 (derived p r i n c i p a l l y from peaks 11, 13, 15 and 18) explained the g r e a t e s t v a r i a n c e , but was not u s e f u l f o r c l a s s i f y i n g the samples i n any recognizable order. F a c t o r 2 was c l e a r l y i n d i c a t i v e o f the presence o f c a s i n g (peaks 6, 7, 8, 10, 19 and 22). F a c t o r 3 (peaks 1, 3, 4, 16 and 20) vs. f a c t o r 5 (peaks 2, 17 and 21) o f f e r e d the g r e a t e s t c l a s s i f i c a t i o n . F a c t o r 3 appeared to represent the b u r l e y c h a r a c t e r , and f a c t o r 5 the b r i g h t character (negative a x i s ) and o r i e n t a l character (positive axis). F a c t o r 4 (not shown-contains peaks 5, 9, 12 and 14) v i s u a l l y o f f e r e d no s e p a r a t i o n o f the samples, but the f a c t o r scores from f a c t o r s 4 and 5 had a c o r r e l a t i o n c o e f f i c i e n t of 0.74. Another u s e f u l technique i s d i s c r i m i n a n t a n a l y s i s (BMDP7M). This technique i s used to f i n d the f u n c t i o n which i s able to best separate samples i n t o p r e - s e t groups by maximizing i n t e r group distances while minimizing intra-group d i s t a n c e s . Disc r i m i n a t i o n o f the 10 groups of tobaccos (Figure 8) was achieved with 100% c o r r e c t c l a s s i f i c a t i o n . J a c k k n i f i n g , a procedure which uses one sample a t a time as a t e s t case and r e c l a s s i f i e s using a l l other cases, showed 87% c o r r e c t c l a s s i f i c a t i o n . The r e s u l t i n g d i s c r i m i n a n t p l o t showed t h a t s e p a r a t i o n along funct i o n one was based p r i m a r i l y on tobacco type. B r i g h t r e c e i v e d a high p o s i t i v e score, b u r l e y was centered near zero and o r i e n t a l received a negative score. Function 2 appeared t o separate by casing. Sensory. Multidimensional s c a l i n g was used i n t h i s study to evaluate the s u b j e c t i v e d i f f e r e n c e s between tobacco types. MDS t r e a t s data based on a person's t o t a l p e r c e p t i o n of the degree of d i s s i m i l a r i t y between objects presented i n a p a i r w i s e f a s h i o n (12). A map i s produced which groups the o b j e c t s by degree o f s i m i l a r i t y on s e v e r a l dimensions.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

122

COMPUTERS IN FLAVOR AND FRAGRANCE RESEARCH

Φ

-ρ «Η

Ο

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

φ · ώ α! ft -Ρ ω β τ* (ϋ σΐ · Η 0) U Λ Ο Ο ^ ϋ on ο · co on , Ω on ο _ «M

S

Ο

φ s Η •Η

Φ Η

«Η



Ο

2

ϋ

on

•Η

·

Λ on ft on CO



*

b0 -ρ 0

θ*

•Ρ bO



8·° on on • on

vo Φ

-

2 α

to Φ •H H

P«4 RÛ

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

9. JELTEMA ET AL.

123

Multivariate and GC Techniques

d b

e θ β

β

c c

D r

d dE Ε

d

c

b

B b ë

C

β

A

C

B Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

b

D

e

b

d

Ε

A C

C

c

M

A

c

FACTOR 1 20%

h

c

d

-B%

S-

Ρ Ε d

A A

FACT0R3 16%

F i g u r e 7. F a c t o r score p l o t s o f 100$ b r i g h t (A,a), 100$ b u r l e y (B,b), 100$ o r i e n t a l (C,c), 33.3#/33.3#/33.3$ CD,d), and 60%/ 30%/l.0% (E,e). Uppercase were cased. Lowercase were uncased samples. Top, f a c t o r 1 vs. f a c t o r 2; and bottom, f a c t o r 3 v s . f a c t o r 5·

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

124

COMPUTERS IN FLAVOR AND FRAGRANCE

RESEARCH

The d i m e n s i o n a l i t y to be used i n an MDS s o l u t i o n can be based on s e v e r a l f a c t o r s , two of which are s t r e s s value and ease of i n t e r p r e t a t i o n . G u i d e l i n e s f o r the number of dimensions chosen i s that s t i m u l i should range from 4-6 times the number of dimensions (12). For t h i s study, where 12 types of tobacco were compared, d i m e n s i o n a l i t y should l i e between 2-3. Stress i s o f t e n used by f i n d i n g an elbow (point at which the curve f l a t ­ tens) i n the s t r e s s value as i t approaches zero. For the 12 types of tobaccos, a s o l u t i o n i n v o l v i n g 2 dimensions was chosen, based on both s t r e s s values and ease of i n t e r p r e t a t i o n . Twelve p a n e l i s t s (5 females, 7 males) were chosen who per­ formed w e l l on p r e l i m i n a r y t r i a n g l e t e s t s . P a n e l i s t s were f a m i l i a r i z e d with samples before t e s t i n g began by sniffing r e p r e s e n t a t i v e p a i r s of tobacco samples. Each sample was e v a l ­ uated a g a i n s t every other sample i n a p a i r w i s e f a s h i o n . Panel­ i s t s rated samples on degree of d i f f e r e n c e using a. 10 cm l i n e . Two r e p l i c a t i o n s were performed. Data from the score cards were entered i n t o the DECSYSTEM 2060 computer v i a a graphics t a b l e t . D i s s i m i l a r i t y scores were evaluated by MDS using the s t a t i s t i c a l program ALSCAL 4 (21). Data from each p a n e l i s t was s e p a r a t e l y evaluated f o r s t a b i l i t y of response and s i m i l a r i t y between panelists. From t h i s a n a l y s i s i t i s c l e a r that dimension 1, which accounted f o r most of the v a r i a t i o n i n the data, separated the types of tobacco based on c a s i n g (Figure 9). The p a n e l i s t s were more s e n s i t i v e to presence of casing than to the p a r t i c u l a r type of tobacco. Uncased b u r l e y was found to be very d i f f e r e n t from cased b u r l e y , while cased b r i g h t or o r i e n t a l were s i m i l a r to t h e i r uncased counterpart. Dimension 2 d i f f e r e n t i a t e d samples based on tobacco type. B r i g h t and o r i e n t a l were found to be most d i f f e r e n t from each other. Blend 2 (60(A), 30(B), 10(C)) was most s i m i l a r to b r i g h t , while blend 1 (33.3(A), 33.3(B), 33.3(C)) l o s t b r i g h t character and was more s i m i l a r to b u r l e y and o r i e n t a l . There were more d i f f e r e n c e s i n tobacco type between uncased samples than those which had been cased. That i s , most of the d i f ­ ferences seen between uncased b r i g h t and b u r l e y were l o s t when casing was added. Tobacco types from d i f f e r e n t crop years were very s i m i l a r i n d i c a t i n g more d i f f e r e n c e s due to tobacco type than crop year. C o r r e l a t i o n of A n a l y t i c a l / S e n s o r y R e s u l t s . Sensory data was c o r r e l a t e d with headspace data of tobacco v o l a t i l e s by f a c t o r a n a l y s i s (BMDP4M) and c a n o n i c a l c o r r e l a t i o n BMDP6M. A n a l y t i c a l data i n c l u d e d f a c t o r scores and d i s c r i m i n a n t analyses scores; sensory data i n c l u d e d scores from the two MDS dimensions. Sorted rotated f a c t o r loadings of combined s e n s o r y / a n a l y t i c a l data u s i n g f a c t o r a n a l y s i s are shown i n Table I I . F a c t o r one contained those v a r i a b l e s from the a n a l y t i c a l and sensory data which r e l a t e d to d i f f e r e n c e s between b r i g h t (A), b u r l e y (Β), and o r i e n t a l (C) (Figure 10). These i n c l u d e d dimension 1 i n the

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

9.

JELTEMA ET AL.

125

Multivariate and GC Techniques

EE CM

d

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

cc

FUNCTION 1 48% Figure 8. Discriminant f u n c t i o n score p l o t s . i n Figure 7 ·

Sample codes are as

CM

Δ ORIENTAL •ORIENTAL ABLEND 1 • BLEND 1

A.BURLEY BURLEY '81 Δ

nBLEND 2

•BURLEY

ΔBLEND 2 • BRIGHT

A BRIGHT ' 8 1 ABRIGHT

ΔUNCASED • CASED

(M

-2

-1

β

1

2

DIMENSION 1 Figure 9· Multidimensional s c a l i n g score p l o t s o f tobacco smoke. 8 l designates 1 9 8 l crop year; a l l others are 1979 crop year. f

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

126

COMPUTERS IN FLAVOR AND F R A G R A N C E RESEARCH

Table I I . F a c t o r A n a l y s i s o f A n a l y t i c a l / S e n s o r y C o r r e l a t i o n of Headspace Data. Factor 1

Factor 3

Factor 4

Factor 5

ANFAC5

0.960

--

--

--

MDSC0M2

0.937

--





--

ANFAC4

0.789

0. 398



--



-0.712

0.,527

-•

--

DISCRIM1

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

Factor 2

-0. 329

ANFAC2



0.,883







MDSC0M1

--

0..836



--

--

ANFAC3

--

0..845





DESCFAC4



-•



DESCFAC5

--

-"

0..418

DESCFAC1

--

DESCFAC3



DESCFAC2



-

-0 .456

-

0,.372

-

DISCRIM2



0 .495

0 .252

-•



2 .493

1 .368

1 .356

1 .097

-0..637

--

0..628

-•

0..573

-•

-•

0..913 -0,.316

Variance Explained

3.071

d i s c r i m i n a n t a n a l y s i s , a n a l y t i c a l f a c t o r s 4 & 5 and MDS dimens i o n 2. In a l l o f these f i g u r e s , o r i e n t a l was most d i f f e r e n t , while b r i g h t and b u r l e y were more s i m i l a r . Blend 1 (33.3/33.3/ 33.3) was centered c l o s e to b u r l e y while blend 2 (60/30/10) remained c l o s e to b r i g h t . F a c t o r 2 contained those v a r i a b l e s which r e l a t e d to casing. These i n c l u d e d MDS dimension 1 and a n a l y t i c a l f a c t o r 2. P l o t s of f a c t o r s 1 and 2 ( F i g . 10) showed that these a n a l y t i c a l / s e n sory f a c t o r s contain both tobacco type and casing information. The second data treatment was based on c a n o n i c a l c o r r e l a t i o n (BMDP6M) which f i n d s combinations o f two sets o f data which show high c o r r e l a t i o n , i . e . what combination o f a n a l y t i c a l v a r i a b l e s (A1-A2) shows a high c o r r e l a t i o n with another combinat i o n o f sensory v a r i a b l e s (S1-S2) (Table I I I ) . Canonical f a c t o r 1 ( A l and S I , R =99%) contained information r e l a t i n g b r i g h t vs b u r l e y vs o r i e n t a l as can be seen by high loadings on those 2

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

9.

JELTEMA ET AL.

Multivariate and GC Techniques

127

FACTOR1 24%

F i g u r e 10. F a c t o r score p l o t s o f a n a l y t i c a l / s e n s o r y data. are as i n Figure 7.

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Labels

128

COMPUTERS IN FLAVOR A N D F R A G R A N C E RESEARCH

a n a l y t i c a l and sensory dimensions which d i s c r i m i n a t e d between b r i g h t , b u r l e y , and o r i e n t a l ( a n a l y t i c a l f a c t o r s 4 & 5, d i s c r i m inant dimension 1 and MDS dimension 2). Canonical f a c t o r 2 (A2

Table I I I . Canonical Factors from A n a l y t i c a l / S e n s o r y from Headspace Data. Canonical Factors S-2 S-l 0.886 0.451 -0.406 0.913

MDSCOMl MDSC0M2 Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

Correlation

ANFAC1

A-l -0.109

A-2 0.041

ANFAC2

0.379

0.752

ANFAC3

0.256

-0.300

ANFAC4

0.802

-0.017

ANFAC5

0.845

-0.288

DISCRIM1

-0.467

0.796

DISCRIM2

0.336

0.094

2

and S2) c o r r e l a t e d with an R of 70%. V a r i a b l e s which showed high loadings i n these f a c t o r s c o r r e l a t e d to casing information ( a n a l y t i c a l f a c t o r 2, MDS dimension 1,). Both f a c t o r a n a l y s i s and canonical c o r r e l a t i o n techniques were s u c c e s s f u l i n demonstrating that d i f f e r e n c e s between tobacco type and casing could be detected from both a n a l y t i c a l and sensory data, and that those d i f f e r e n c e s found a n a l y t i c a l l y were h i g h l y c o r r e l a t e d t o sensory d i f f e r e n c e s . From t h i s type o f data c o r r e l a t i o n , components can be pinpointed which may be responsible f o r sensory d i f f e r e n c e s between tobacco types. Summary The two research i n v e s t i g a t i o n s reported here - the sensory quality control s p e c i f i c a t i o n model and the a p p l i c a t i o n of sensory and a n a l y t i c a l data f o r d e f i n i n g d i f f e r e n c e s i n tobacco aroma - both demonstrate the usefulness of m u l t i v a r i a t e a n a l y s i s techniques f o r analyzing a n a l y t i c a l and sensory data as w e l l as c o r r e l a t i n g these data. Although these tasks do not compare i n complexity to that of the p r e d i c t i o n o f sensory response t o a n a l y t i c a l data c o l l e c t e d on c i g a r e t t e smoke, our research to date has revealed no element which i n d i c a t e s that t h i s i s an impossible task. In f a c t , the r e s u l t s of these and s i m i l a r

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

9.

JELTEMA ET A L .

Multivariate and GC Techniques

129

i n v e s t i g a t i o n s (6,7,22) support the p r o p o s i t i o n that i n f o r m a t i o n does e x i s t and i s d e t e c t a b l e i n the a n a l y t i c a l data s e t , which i s necessary f o r the s u c c e s s f u l c o r r e l a t i o n and hence p r e d i c t i o n of the sensory response to c i g a r e t t e smoke. Acknowledgments

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

The authors wish to thank those people who c o n t r i b u t e d to the c o l l e c t i o n of the data used f o r the a n a l y t i c a l / s e n s o r y c o r r e l a ­ tions. These i n c l u d e V i v i a n W i l l i s , C h r i s K r o u s t a l i s , Danny E n n i s , Tim Crews, Mary Buckner, Janet Shelton and Duane Watson. Anne Donathan i s a l s o acknowledged f o r her work on the prepara­ t i o n of t h i s manuscript. Literature Cited 1. 2. 3. 4. 5. 6. 7.

8. 9.

10.

11.

12. 13. 14.

15.

16.

C l a r k , Η. Α.; J u r s , P. C. Anal. Chem. 1975, 47, 374. Hoff, J . T.; Herwig, W. C. ASBC J o u r n a l , 1975, 34, 1. Saxberg, B. E. H.; Duewer, D. L.; Booker, J . L.; Kowalski, B. R. A n a l . Chim. Acta. 1978, 103, 201. Blomquist, G.; Johansson, E.; Söderstrom, B.; Wold, S. J . Chromatogr. 1979, 173, 19. McConnell, M. L.; Rhodes, G.; Watson U.; Novotný, M. J. Chromatogr. 1979, 162, 495. P a r r i s h , M. E.; Good, B. W.; Jeltema, Μ. Α.; Hsu, F. S. Anal. Chem. Acta. 1983, 150, 163. Hsu, F. S.; Good, B. W.; P a r r i s h , M. E.; Clabo, D. A. Jr.; Crews, T. D. J . High Resoln. Chromatogr. Chromatogr. Commun. 1982, 5, 658. Ennis, D. M.; Boelens, H.; Haring H.; Bowman, P. Food Tech. 1982, 36(11), 83. Powers, J . J . i n Food F l a v o u r s , P a r t Α.; Morton, I.D.; MacLead, A. J . , Eds.; Elsevier S c i e n t i f i c Publishing Company: Amsterdam, 1982; pp 121-168. Schiffman, S. S.; Reynolds, M. L.; Young, F. W. "Introduc­ t i o n to M u l t i d i m e n s i o n a l S c a l i n g " ; Academic Press, Inc.: New York, 1981. W i l l i a m s , A. A. i n "Progress i n F l a v o u r Research"; Land, D. G.; Nursten, Η. E., Eds.; A p p l i e d Science P u b l i s h e r s LTD: London, 1979; pp 287-305. Pangborn, R. M. i n "Flavour '81"; S c h r e i e r , P., Ed.; Walter de Gruyter & Co.: B e r l i n , 1981; pp 3-32. Kwan W.; Kowalski, B. R. J . Food S c i . 1980, 45, 213. Clapperton, J . F.; i n "Progress i n F l a v o u r Research"; Land, D. G.; Nursten, H. E., Eds.; A p p l i e d Science P u b l i s h e r s LTD: London, 1979; pp 1-14. Powers, J . J . "Sensory E v a l u t i o n Methods f o r the P r a c t i c i n g Food T e c h n o l o g i s t " ; Johnston, M. R., Ed.; I n s t i t u t e of Food T e c h n o l o g i s t s : Chicago, 1979; ρ (9) 1-17. Warner, K.; Evans, C. D.; L i s t , G. R.; Dupuy, H. P.; Wads-

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

130

17. 18. 19.

20. 21.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch009

22.

C O M P U T E R S IN F L A V O R A N D F R A G R A N C E

RESEARCH

worth, J . I.; Goheen, G. E. J . Am. O i l Chem. Soc. 1978, 55, 252. Schmitt D. J . ; Hoff, J . T. J . Food S c i . 1979, 44, 901. Trant, A. S.; Pangborn, R. M.; Little, A. C. J . Food S c i . 1981, 46, 583. Amerine, Μ. Α.; Pangborn, R. M.; R o e s s l e r , Ε. B. " P r i n ­ c i p l e s of Sensory E v a l u a t i o n of Food"; Acad. Press N.Y.: 1965; pp 3-32. "BMDP Statistical Software", W. J . Dixon, Ed., U n i v e r s i t y of C a l i f o r n i a Press, 1981. Young, F. "ALSCAL 4"; U n i v e r s i t y of North C a r o l i n a : Chapel Hill, N.C., 1977. Jeltema, Μ. Α.; Good, B. W.; K r o u s t a l i s , C. S.; Ennis, D. M.; Watson, D. C.; Willis, V. E. Presented at 43rd Annual Meeting of the I n s t i t u t e of Food T e c h n o l o g i s t s , A b s t r a c t #134, New Orleans, June 20, 1983.

R E C E I V E D April 6, 1984

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

10 Analytical Flavor Data Enhancement with Computer Techniques RICHARD A. DEPALMA

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch010

The Procter & Gamble Company, Winton Hill Technical Center, 6071 Center Hill Road, Cincinnati, O H 45224

The j o i n i n g of f l a v o r a n a l y s i s by c a p i l l a r y gas chromatography with computerized data handling techniques i s a n a t u r a l combination of information generators with information a n a l y z e r s . The chro­ matographic analyses of f l a v o r e x t r a c t s from food matrices produces hundreds of d i s c r e t e peaks. A dedicated data system i s needed to process t h i s data, and a programmable system i s needed to develop software to e x t r a c t the maximum i n f o r ­ mation from the f l a v o r data. The experience of our l a b o r a t o r y , with the Hewlett-Packard 3357 Labor­ atory Automation System, will be discussed. The chromatographic/software a p p l i c a t i o n s are d i v i d e d i n t o four c a t e g o r i e s : standard chromatographic software, d i s p l a y c a p a b i l i t i e s , file searching, and m u l t i v a r i a t e a n a l y s i s . Computers are pervading a l l areas of our l i f e and, i n t h i s same way, they are pervading a l l areas of chemical research. In f l a v o r a n a l y s i s , computers with t h e i r information a n a l y s i s c a p a b i l i t i e s are a n a t u r a l a l l y with separation science which i s an information generator. Our l a b o r a t o r y has extensive experience with the Hewlett-Packard 3357 Laboratory Automation System. The s p e c i f i c hardware which we have i s given i n Table I. Table I.

Hewlett-Packard

3357 Laboratory Automation System

Ε S e r i e s C e n t r a l Processor 85 Mbyte F i x e d Disc 320 Kwords Memory Magnetic Tape Unit 4 Graphics Terminals 3 Non-Graphics Terminals Plotter 0097-6156/ 84/ 0261 -0131 $06.00/ 0 © 1984 American Chemical Society

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

132

COMPUTERS IN FLAVOR AND FRAGRANCE RESEARCH

The s y s t e m i s a c t u a l l y composed o f two s e t s o f s o f t w a r e : t h e RTE-6/VM o p e r a t i n g system s o f t w a r e which p r o v i d e s a l l the f e a t u r e s w h i c h y o u e x p e c t from a r e a l t i m e m u l t i - t a s k i n g o p e r a t i n g s y s t e m ( h i g h l e v e l l a n g u a g e p r o g r a m m i n g , d i s c f i l e a c c e s s and m a i n t e n a n c e , u t i l i t i e s , e t c . ) , and the L a b o r a t o r y A u t o m a t i o n S o f t w a r e w h i c h c o l l e c t s d a t a from c h r o m a t o g r a p h i c - t y p e i n s t r u m e n t s , integrates t h e d a t a , s t o r e s t h e d a t a on t h e d i s c , and d i s p l a y s t h e r e s u l t s i n standard r e p o r t form. Before g o i n g o n , the reasons f o r the a c q u i s i t i o n o f such a l a r g e ( e x p e n s i v e ) d a t a s y s t e m needs t o be a d d r e s s e d . Experience i n o u r l a b o r a t o r y w i t h c h r o m a t o g r a p h i c d a t a has shown t h a t a d a t a s y s t e m needs t h e a t t r i b u t e s i n T a b l e I I .

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch010

Table I I . 1. 2. 3. 4. 5. 6.

Data System Requirements

Turnkey chromatography c a p a b i l i t i e s . A b i l i t y t o t a k e d a t a from s e v e r a l t y p e s o f i n s t r u m e n t a t i o n ( c a p i l l a r y G C , p a c k e d c o l u m n GC, L C , A u t o A n a l y z e r ) . A b i l i t y t o t a k e d a t a from i n s t r u m e n t s f o r s e v e r a l v e n d o r s . A b i l i t y t o t a k e d a t a from 15 i n s t r u m e n t s and have c a p a c i t y t o grow. (Instruments l o c a t e d hundreds o f feet a p a r t . ) O n - l i n e s t o r a g e o f c h r o m a t o g r a p h i c d a t a , raw d a t a and i n t e grated data. A b i l i t y to program i n h i g h l e v e l languages to a) d i s p l a y the c h r o m a t o g r a m s , and b) p e r f o r m s t a t i s t i c a l a n a l y s i s on c h r o matographic data.

A t t h e t i m e o f p u r c h a s e , t h e HP d a t a s y s t e m was the o n l y one w h i c h met a l l c r i t e r i a and o u r e x p e r i e n c e w i t h i t has n o t been disappointing. The m a j o r a d v a n t a g e o f s u c h a s y s t e m i s t h a t e a c h u s e r has t h e f u l l power o f t h e s y s t e m a v a i l a b l e t o them a t a l l t i m e s , when t h e y a r e a t any t e r m i n a l , r e g a r d l e s s o f t h e number o f o t h e r u s e r s and demands on t h e s y s t e m ( a l t h o u g h t h e s y s t e m does s l o w down when t h e number o f u s e r s i n c r e a s e s ) . This a v a i l a b i l i t y of c o m p u t a t i o n a l power i s t h e most s i g n i f i c a n t l e a r n i n g from o u r experience. The u s e r s o f t w a r e i s d i v i d e d i n t o two s e t s . The t u r n k e y c h r o m a t o g r a p h i c s o f t w a r e and i n - h o u s e s o f t w a r e w h i c h p l o t s chromatograms and p e r f o r m s s t a t i s t i c a l a n a l y s i s . The t u r n k e y s o f t w a r e p r o v i d e s d a t a a c q u i s i t i o n and i n t e g r a t i o n w h i c h l e a d s t o a r e p o r t as shown i n F i g u r e 1. The r e p o r t i n c l u d e s a l l t h e s t a n d a r d i t e m s o f r e t e n t i o n t i m e , peak a r e a , i d e n t i f i c a t i o n , e t c . I n a d d i t i o n , a BASIC p r o g r a m r e f o r m a t s t h e r e p o r t i n t o a form which i s d i r e c t l y sent to the s u b m i t t e r ( F i g u r e 2 ) . FORTRAN programs c a n a l s o be u s e d t o r e f o r m a t t h e d a t a o r p e r f o r m o t h e r p o s t - r u n a n a l y s i s o f the d a t a , s u c h as ( a ) c o l u m n r e s o l u t i o n c h e c k s , and ( b ) s e c o n d a r y s t a n d a r d a n a l y s i s .

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

10.

DePALMA

Analytical Flavor Data

133

ζ

οt-» ht-t

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch010

ω οα . οu

PJ PJ

y— ^ Z> A A



~ i 3 00 00 £s. ΓΟ «r 00 PJ *- Ch

σ

00

C

a:

οf-

03

Η-I

CJ

Ζ ο i—! 31 ο

\

Ι>

-

ο Κ ο

:> . Ω

ο

Y-

Ld



a: Ο ζ

ο Κ ο Ο ΙΠ 3 Χ

i —i ω

:ε »Η-t



ο

α

ο

Analytical Flavor Data

ο

^

*

·

Γ

θ

Γ

*

>

~

*

·

ο

σ

~

£

147

^

Γ

θ

ο

θ

Ν

θ

Γ

θ

>

θ

2

ο

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch010

~

- + ^

^

u"î

eu

ro

>o

LO

U"Î

ω

ro

^-

CD

CD

»-

^"

»-

>o

American Chemical Society Library

1155 16th St., N.W. Washington, D.C. 20036

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

^

148

COMPUTERS IN FLAVOR A N D FRAGRANCE RESEARCH

Literature Cited 1. 2. 3. 4. 5.

Heckers, H.; Dittmar, K; Melcher, F. W.; Kalinowski, H. O. J . Chromatogr., 1977, 135, 93. Gibson, T. W.; Strassburger, P. J . Organic Chem., 1976, 41, 791. Arctander, S. "Perfume and F l a v o r Chemicals"; S t e f f e n Arctander, P u b l i s h e r , 1969; V o l . I and I I . Davis, J . C. " S t a t i s t i c s and Data A n a l y s i s i n Geology"; John Wiley & Sons, New York, 1973; p. 467. M a y f i e l d , H. T.; Bertsch W. Computer Appl. Lab., 1983, 1, 130. 1984

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch010

RECEIVED M a y 30,

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

11 Experiences in Use of Robotics in an Analytical Research Laboratory W. J. HURST, K. P. SNYDER, and R. A. MARTIN, JR.

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch011

Hershey Foods Corporation Technical Center, 1025 Reese Avenue, P.O. Box 805, Hershey, PA 17033-0805

The use of the Zymate Laboratory Automation System allows the s t a n d a r d i z a t i o n and automation of many r o u t i n e operations i n an a n a l y t i c a l chemistry labor a t o r y . I t a d d i t i o n a l l y allows f o r a c l o s i n g of the a n a l y t i c a l automation loop of sample p r e p a r a t i o n and a n a l y s i s t h e r e f o r e p o t e n t i a l l y decreasing the need f o r personnel with a r e s u l t a n t increase in productivity. These operations i n c l u d e , but a r e not l i m i t e d to, weighing, p i p e t t i n g , d i l u t i n g , b l e n d i n g , heating, liquid-solid e x t r a c t i o n , and filtration. Since our l a b o r a t o r y f r e q u e n t l y uses HPLC f o r the final determination step, those assays were first chosen f o r automation. Each procedure was subdivided i n t o d i s c r e t e l a b o r a t o r y u n i t operations f o r final i n c l u s i o n i n t o the Zymate program. Each of these operations was a l s o assigned to a module such as hand, master l a b s t a t i o n , or blender. The sequence of operations and modules was then merged to a r r i v e at a final procedure. T h i s final procedure was then "taught" to the robot using a s e r i e s o f user-defined terms which could then be coupled i n t o a program f o r that sample p r e p a r a t i o n . Since many of the l a b o r a t o r y operations are the same f o r many assays, an a n a l y s t needs to d e f i n e only a l i m i t e d number of terms to be intermixed i n t o a v a r i e t y o f programs. Examples of system layout and method uses will be given. A d d i t i o n a l l y , data is presented as to the p r e c i s i o n a v a i l a b l e in sample p r e p a r a t i o n using laboratory robotics. In the a n a l y t i c a l chemistry f i e l d , there i s i n c r e a s i n g emphasis on automation of the l a b o r a t o r y . In many cases implementation of automation i s l i m i t e d to the i n s t a l l a t i o n o f a l a b o r a t o r y computer which a s s i s t s i n l a b o r a t o r y management, sample flow and 0097-6156/ 84/0261-0149S06.00/0 © 1984 American Chemical Society

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

150

COMPUTERS IN FLAVOR AND F R A G R A N C E RESEARCH

use of any of a v a r i e t y of autosamplers f o r a n a l y t i c a l i n s t r u ments. These autosamplers are t i e d to the l a b o r a t o r y computer f o r f i n a l data a n a l y s i s and r e p o r t generation. While these are steps i n automation, there i s u s u a l l y human i n t e r v e n t i o n of some kind i n v o l v e d i n sample p r e p a r a t i o n , t h e r e f o r e l e a v i n g a v o i d i n the automation loop. The human i n t e r v e n t i o n steps are labor i n t e n s i v e i n sample p r e p a r a t i o n f o r both s e r v i c e and a n a l y t i c a l methods development a p p l i c a t i o n s . Sample receipt

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch011

F i g u r e 1.

Human Intervention

Analysis

A n a l y t i c a l Sample Flow.

The i n t r o d u c t i o n of l a b o r a t o r y r o b o t i c s f o r sample p r e p a r a t i o n and u l t i m a t e l y methods development serves to c l o s e t h i s gap. These l a b o r a t o r y robots bear no resemblance to C3P0 and R2D2 of Star Wars fame, but r a t h e r they are complex computer c o n t r o l l e d u n i t s s p e c i f i c a l l y manufactured f o r use i n a n a l y t i c a l chemistry and are capable of a l a r g e number of tasks. They can be obtained commercially or can be l a b o r a t o r y manufactured (1,2). The i n i t i a l a p p l i c a t i o n i n our l a b o r a t o r y was to automate the p r e p a r a t i o n of samples f o r a f i n a l HPLC determination of sorbate i n chocolate syrup. The b a s i c robot u n i t used i n t h i s study (Zymate, manufactured by Zymark Corporation, Hopkinton, MA) c o n s i s t s of a c o n t r o l l e r which c o n t r o l s motion through a Cathode Ray Tube and keyboard, and the robot arm which has three dimensional motion. The u n i t has many a u x i l i a r y items of equipment i n c l u d i n g a master lab s t a t i o n f o r p i p e t t i n g and d i l u t i n g , a Vortex u n i t f o r sample mixing, a d i g i t a l balance f o r automated sample weighing, a heating b l o c k and a power/event c o n t r o l l e r which serves to monitor switch c l o s u r e s and actuate s p e c i f i c equipment. Various h o l d i n g racks are d i s p e r s e d i n s t r a t e g i c l o c a t i o n s around the robot. These racks are f o r t e s t tubes, p i p e t t e t i p s and HPLC autosampler v i a l s . The u n i t a l s o i s equipped with an RS-423 i n t e r f a c e f o r data sharing between the robot and the l a b o r a t o r y computer. A p r i n t e r serves as a means of program documentation and data review. A c o n f i g u r a t i o n of the system i s shown i n Figure 2. The robot u n i t has a l a r g e number of a v a i l a b l e operations. These i n c l u d e automatic weighing, d i l u t i n g and other l i q u i d handling, s e p a r a t i o n ( p a r t i t i o n i n g ) , mixing, f i l t r a t i o n , dispensing, manipulation with hands, data r e d u c t i o n and documentation. The robot operates by moving to a p o s i t i o n i n three dimens i o n a l space. For one f a m i l i a r with d e a l i n g i n the complex methodology, the use of the robot i s a humbling experience s i n c e a l l i t s operations must be described i n t h e i r most b a s i c movements. For example, i f a person i s moving a p i e c e of glassware from one point to another and encounters an o b s t a c l e such as a s h e l f or bracket, that person knows to avoid t h i s o b s t a c l e . Unless t o l d

In Computers in Flavor and Fragrance Research; Warren, C., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Robotics in a Research Laboratory

HURST E T A L .

Robot Controller and Printer

Publication Date: August 22, 1984 | doi: 10.1021/bk-1984-0261.ch011

Power and Event [Controller

HPLC Injector

Ο

ompletejl Sample IVials

Master Lab Station

Hands

'^7 Liquid ispenserj

Turntable

Ο

Digital Balance

Block Heater