Carbonates Mineralogy and Geochemistry
 0939950154

Table of contents :
Page 1
Titles
REVIEWS in
Volume 11
CARBONATES:
MINERALOGY and CHEMISTRY
RICHARD J. REEDER, Editor
The Authors:
Series Editor:
MINERALOGICAL SOCIETY OF AMERICA
Page 1
Titles
REVIEWS in MINERALOGY
Mineralogical Society of America
ii
Tables
Table 1
Page 1
Titles
CARBONATES:
MINERALOGY and CHEMISTRY
TABLE of CONTENTS
1
Chapter
CRYSTAL CHEMISTRY of the RHOMBOHEDRAL CARBONATES
Page 2
Titles
2
3
Page 3
Titles
4
5
Page 4
Titles
Fossil shells
. . . .
6
Page 5
Page 6
Titles
8
Page 7
Titles
9
Page 8
Titles
Stacking faults
Stylolites - "Pressure solution"
. ... .
R
REFERENCES
Page 1
Titles
CARBONA TES:
MINERALOGY and CHEMISTRY
FOREWORD
PREFACE and ACKNOWLEDGMENTS
iii
Page 2
Page 1
Titles
1
CR YST AL CHEMISTRY of the RHOMBOHEDRAL CARBONATES
Page 2
Page 3
Titles
l
o
3
Tables
Table 1
Page 4
Titles
h + k
h
k:
+ £
h
z
~
Tables
Table 1
Page 5
Page 6
Page 7
Titles
........... %
Page 8
Page 9
Titles
9
Page 10
Titles
Atomic thermaZ vibrations.
Page 11
Titles
11
Tables
Table 1
Page 12
Titles
12
Tables
Table 1
Page 13
Titles
StructuraZ variation.
Page 14
Titles
i;
'"
::.
N' A Zn
Mn
UNIT CELL VOLUME (1.')
'.0
·Cd
.Fe
.Zn
• Mg • Mn
2.'
~
o
14
Tables
Table 1
Page 15
Tables
Table 1
Page 16
Page 17
Titles
The magnesian calcite solid solution.
Tables
Table 1
Page 18
Page 19
Titles
.;;:r
="
.,.
~
Page 20
Titles
::bd
The CaC03-CdC03 solid solution.
The Ni-Mg carbonate soZid solution. Based on the similarity of ionic
Page 21
Page 22
Tables
Table 1
Page 23
Page 24
Titles
.,

~
~
o
-o
'"
~
.::
"
-o
-
o
"
~
~
o
"
'"
Il
....
u
9
o
9
0-
o
:i:
ur
~
e
.!::
:
o
.,_
'"
~
-
.~
.!::
'"
III
III
-o
OJ
OJ
24
Page 25
Titles
25
Page 26
Titles
OctahedraZ distortion.
Page 27
Titles
re
...
Table 9
Tables
Table 1
Table 2
Page 28
Titles
Thermal parameters.
Ankerite and ferroan dolomite.
Page 29
Page 30
Page 31
Titles
Kutnahorite.
Other transition metaZ doZomites.
Tables
Table 1
Page 32
Titles
Thermal disordering.
Page 33
Page 34
Page 35
Titles
Structural changes.
Disorder in low-temperature dolomites.
Page 36
Page 37
Page 38
Titles
..
~
"
.:
,.0
..
.
°0,
~o
/0
.: ° f::,
38
Page 39
Titles
::il_§:=~:
Page 40
Page 41
Titles
U H HUH H
Page 42
Titles
-:
c
Page 43
Page 44
Titles
44
Tables
Table 1
Table 2
Table 3
Page 45
Page 46
Titles
(a) Colcite- type (b) CoCO;, II
~ \ ~ \
---®----@----@----@------------~-----@---~------0-
=~
~ \ \ \
Page 47
Page 48
Page 1
Titles
2
PHASE RELATIONS of RHOMBOHEDRAL CARBONATES
Page 2
Page 3
Page 4
Tables
Table 1
Page 5
Titles
~
u
~
! + /""
j :::t~ j:"
Page 6
Titles
-
-
-
-
I
15
I
10
I
5
Solution
/.> '"
Calcite /Z~/-
Solid J/
A/
if
1'1
"
> I Calcite + Dolomite
> I
o
9001-
5001-
54
Tables
Table 1
Page 7
Titles
Relations at higher pressures and temperatures.
Page 8
Page 9
Titles
Cd+Cm
60 80 100
MgC03
Ar+Cd
20 40
00 20 40 60 80 1000
MgCO~
\
\
+
T
1500
500
1000
u
57
Tables
Table 1
Page 10
Titles
AdditionaZ considerations in subsoZidus relations.
Page 11
Page 12
Tables
Table 1
Page 13
Titles
::k XXXX
XXXX
X
x
X
Tables
Table 1
Page 14
Tables
Table 1
Page 15
Titles






,
,
,
Tables
Table 1
Page 16
Page 17
Titles
Zn-Dolomite I Zn- Dolomite
+ +
Calcite Smithsonite
Zn- Calcite
Ca - Smithsonite
T
Page 18
Page 19
Titles
I
Page 20
Titles
-----
Page 21
Titles
-----
Page 22
Page 23
Titles
._. __ _.-..CoMnco,',
"
, 'j
, ,
f .. :
,/'
71
Page 24
Titles
\
eeo-c
Page 25
Page 26
Page 27
Page 28
Page 1
Titles
3
Page 2
Page 3
Page 4
Page 5
Tables
Table 1
Page 6
Titles
..
;./:\
:",
.. '
.. \
..
... 7.\
•.. ~
..... \
. ' .. '\
. . .,: .i:-
~ .' \
. . .
_':: .. ::::: .. ::\~\~ .. ...!....OL: ., '. :., : :':': ... ".: .. : .:.:/:: .
Page 7
Titles
Co Mn (C03)2
.\
/: .
. ...
'._. .
. .
...... .
.....
... .
:
:
.. '
..
'.'
._::_.~:.!' • .:t:u_: ;M" •• " \~.:: •• ;.:: :.~(./)::; .:. :':~" ~,~ ~ :: •• .. __ --"'- __ ->
83
Page 8
Page 9
Page 10
Page 11
Titles
87
----
Page 12
Page 13
Titles
/~2~
Page 14
Page 15
Page 16
Page 17
Page 18
Page 19
Page 20
Page 1
Titles
4
MAGNESIAN CALCITES: LOW-TEMPERATURE 'OCCURRENCE,
Fred T. Mackenzie, William D. Bischoff, Finley C. Bishop,
Page 2
Page 3
Tables
Table 1
Page 4
Page 5
Titles
- TOC
o 28
24
20
16
12
8
o
WATER TEMPERATURE 'C
~
..
...
g
N
'"
101
Tables
Table 1
Table 2
Table 3
Page 6
Tables
Table 1
Page 7
Tables
Table 1
Page 8
Page 9
Page 10
Titles
.~
NBS Oolomit. __ • :/~~se~~;n~Jitl! 0
Tables
Table 1
Page 11
Page 12
Page 13
Titles
...
J
Tables
Table 1
Page 14
Tables
Table 1
Page 15
Page 16
Page 17
Titles
7.2

7.4

.,
] 76
o
o
8.2
8.4
8.6
o
11
o
10 20
MOLE %MgC03
11
30
113
Page 18
Page 19
Page 20
Titles
r ... --------1
t I - ...... _
i I ~ - · .___:».
[disordered calcite]
IT at
low P
[NaCI]
149
Tables
Table 1
Page 6
Page 7
Page 8
Tables
Table 1
Page 9
Titles
o
~
,.
Page 10
Page 11
Titles
350
300
Temperature -c
155
Tables
Table 1
Page 12
Titles
° 0
°
Page 13
Page 14
Page 15
Titles
Sr
Pb
Page 16
Page 17
Titles
Table 4. Anhydrous rhombohedral and orthorhombic carbonates
-----------------_.-------------------_ .... --_._-----------------
-_ ... ---_ ...... ".----== ..... ===-""'.,..= .... _-_ ....... =_."" ... = .. = ... "",.=--------------
--------_ .. _-----------_._--------------------_ ... _-----------------
-----------------_.-----------------------_.-----_._------.------
Recent shells.
Tables
Table 1
Page 18
Titles
Table 5. Trace element content of aragonitic skeletons.
...... _-_ ....... _-,.""---------------------- .. _---------------------------
}
u
--_ ... _------------------_._---------_._------ ... _-----._---------
:d
162
Page 19
Page 20
Titles
ionic radius (Al
Page 21
Tables
Table 1
Page 22
Titles
40
Tables
Table 1
Table 2
Page 23
Titles
Fossil ehel.l.e ,
Page 24
Titles
i
\ t
;--l-+/ "''+'"
Page 25
Page 26
Tables
Table 1
Page 27
Page 28
Titles
r : ~
.(
.. i·
Page 29
Titles
temperature (OC)
Tables
Table 1
Page 30
Page 31
Titles
b 1000 I,...,-.,---,-r-,...---,-.----,-~
CaBa(C03)2
.._
E
mole
percent
Tables
Table 1
Table 2
Page 32
Titles
Pb
Page 33
Page 34
Titles
t~ ~!
~ ~. ~I U)
'"
~~----------------------~a
Page 35
Page 36
Page 37
Titles
\
a
Page 38
Page 39
Tables
Table 1
Table 2
Page 40
Titles
3 4.0
~
cation atomic moss
'"
E
E
6.0
5.0
100
140
180
220
100
equivalent mol % SrC03
witherites
cation atomic mass
E
E
Tables
Table 1
Page 41
Page 42
Titles
aile
Tables
Table 1
Table 2
Table 3
Page 43
Page 44
Titles
-7
Page 45
Page 46
Tables
Table 1
Page 1
Titles
6
The POL YMORPHS of CaC03
and the ARAGONITE- CALCITE TRANSFORMATION
William D. Carlson
Page 2
Titles
oco, PHASE RELATIONS
Page 3
Page 4
Titles
0' 'j»ti:)J-'¥" lor 'V' 'Ol' '°1°' I!!)r
350 400 450 500 550
16 [ ContOU'S on Sr-rich limit of
:012
W
~8
~6
194
Page 5
Page 6
Page 7
Titles
lii
rJ
1", T "1
--- - -X-X'-_1
. \
,
t
Tables
Table 1
Table 2
Table 3
Page 8
Titles
Evidence for disordering.
Page 9
Titles
- ~ .. ,\.
ON.
199
Tables
Table 1
Page 10
Titles
Possible disordering schemes.
Page 11
Page 12
Titles
The transformation under conditions of aragonite stability.
Page 13
Titles
The transformation outside the aragonite stabiZity fieZd. Experiments
Page 14
Titles
Petrographic observations.
Experiments measuring buZk transformation rates.
Page 15
Titles
"
.~
_.> e
o 0
..,
"' ...
e.!l
o
~.&
~.s
.. "
...
" ..
..
~ §
...
.. '"
.. ~
.... '"
Page 11
Titles
Dislocations.
Page 12
Page 13
Titles
361
Page 14
Titles
"
j
~ ....
~o
e
"
-e
e
o
.....
"''''
~::!
"'"
.!1

Citation preview

REVIEWS in MINERALOGY Volume 11

CARBONATES: MINERALOGY

and CHEMISTRY

J.

RICHARD

REEDER, Editor

The Authors: DAVID J. BARBER

FRED T. MACKENZIE

Dept. of Physics University of Essex Colchester, England

JANE SCHOONMAKER C04

3SQ

WILLIAM D. BISCHOFF

Dept. of Oceanography University of Hawaii Honolulu, Hawaii 96822

FINLEY C. BISHOP

JOHN

Dept. of Geological Sciences Northwestern University Evanston, Illinois 60201

Dept. of Oceanography Texas A & M University College Station, Texas 77843

W. MORSE

WILLIAM D. CARLSON

RICHARD

Dept. of Geological Sciences University of Texas at Austin Austin, Texas 78712

Dept. of Earth & Space Sciences State University of New York Stony Brook, New York 11794

J. REEDER

ERIC J. ESSENE

J. ALEXANDER

Dept. of Geological Sciences University of Michigan Ann Arbor, Michigan 48109

Dept. of Geological Sciences Virginia Polytechnic Institute State University Blacksburg, Virginia 24061

JULIAN

SPEER &

R. GOLDSMITH

Dept. of Geophysical Sciences University of Chicago Chicago, Illinois 60637 MICHELE

LOIJENS

ROLAND

WOLLAST

HANS-RUDOLF

Laboratoire d'Oceanographie Universite Libre de Bruxelles 1050 Brussels, Belgium

Series Editor: Virginia

JAN VEIZER Dept. of Geology University of Ottawa Ottawa, Ontario, Canada

KIN 6N5

WENK

Dept. of Geology & Geophysics University of California Berkeley, California 94720

PAUL H. RIBBE Dept. of Geological Sciences Polytechnic Institute & State University Blacksburg, Virginia 24061

MINERALOGICAL

SOCIETY OF AMERICA

COPYRIGHT MINERALOGICAL

1983

SOCIETY

PRINTED

of AMERICA

BY

BookCrafters, Inc. Chelsea, Michigan 4-8118

REVIEWS in MINERALOGY (Formerly:

SHORT

COURSE

NOTES)

ISSN 0275-0279 Volume

11:

CARBONATES:

MINERALOGY

and CHEMISTRY

ISBN 0-939950-15-4-

those

ADDITIONAL listed below

COPIES of this volume as well as may be obtained at moderate cost from:

Mineralogical

Society of America

2000 FLORIDA AVENUE, N.w.

WASHINGTON, D. C. 20009

Volume 1

SULFIDE MINERALOGY, P.H. Ribbe, editor

(1974) (1975; revised 1983)

284 p.

2

FELDSPAR MINERALOGY, P.H. Ribbe, editor

3

OXIDE MINERALS, Douglas Rumble III, editor

4

MINERALOGY and GEOLOGY of NATURAL ZEOLITES, F.A. Mumpton, editor (1977)

232 p.

5

ORTHOSILICATES, P.H. Ribbe, editor

450 p.

6

MARINE MINERALS, R.G. Burns, editor

7

PYROXENES, C.T. Prewitt, editor

8

KINETICS of GEOCHEMICAL PROCESSES, A. C. Lasaga and R.J. Kirkpatrick, editors (1981)

(1976)

(1980; revised 1982) (1979)

(1980)

9A AMPHIBOLES and OTHER HYDROUS PYRIBOLES - MINERALOGY, D.R. Veblen, editor (1981)

362 p. 502 p.

380 p. 525 p. 391 p. 372 p.

9B AMPHIBOLES: PETROLOGY and EXPERIMENTAL PHASE RELATIONS, D.R. Veblen and P.H. Ribbe, editors (1982) 390 p. 10 11

CHARACTERIZATION of META}IORPHISMthrough MINERAL EQUILIBRIA, J.M. Ferry, editor (1982) CARBONATES,

R.J. Reeder, editor

ii

(1983)

397 p.

CARBONATES: MINERALOGY

and CHEMISTRY

TABLE of CONTENTS COPYRIGHT;

LIST OF PUBLICATIONS

FOREWORD PREFACE;

Page ii

......•.

iii

ACKNOWLEDGMENTS

iii Chapter

1

CRYSTAL

of the RHOMBOHEDRAL

CHEMISTRY

INTRODUCTION

1

RHOMBOHEDRAL

VERSUS HEXAGONAL

THE C03 GROUP AS A STRUCTURAL

AXIAL SYSTEMS

2

UNIT

4

THE R3c CARBONATES

6

The calcite structure Calcite isotypes

6 8

Atomic thermaZ vibrations Structural variation Solid solutions of R3c carbonates The magnesian calcite soZid solution The CaC03-CdC03 solid solution The Ni-Mg carbonate solid solution Lattice

parameters

THE R3 CARBONATES The dolomite

in other solid solutions

Dolomite

22 23 26 28 28 28 31 31 32 32 32 35 35 37

isotypes

carbonates

Thermal disordering Structural changes Disorder in low-temperature dolomites Lattice

parameters

of dolomite-type

OTHER DOUBLE

CARBONATES



carbonates

40

AT HIGH TEMPERATURES

AND PRESSURES

Thermal expansion High-temperature transformations CaC03(II) structure

of calcite

FOR FUTURE WORK

ACKNOWLEDGMENTS

17 17 20 20 21 22

Calcian dolomites and ankerites Cation order in dolomite-structure

SUGGESTIONS

14

structure

Ankerite and ferroan dolomite Kutnahorite Other transition metal dolomites

CHEMISTRY

10

. . • . . • •

Interatomic distances Octahedral distortion Thermal parameters

CRYSTAL

CARBONATES

J. Reeder

Richard

43 43 45 45 46

. • . . • .

47 v

Chapter

2

PHASE RELATIONS

2

of RHOMBOHEDRAL Julian

CARBONATES

R. Goldsmith Page

INTRODUCTION

AND EXPERIMENTAL

THE END-MEMBER DOLOMITE-TYPE BINARY

RESULTS

49

CARBONATES

51

COMPOUNDS

PHASE

51

RELATIONS

51

CaC03-MgC03

Relations at moderate pressures Relations at higher pressures and temperatures Additional considerations in subsolidus relations CdC03-MgC03 CaC03-MnC03 CaC03-FeC03 Additional binary

joins with

a single

solvus

CaC03-NiC03; CaC03-CoC03; MgC03-NiC03 Binary joins with extensive solid solubility The join CaC03-ZnC03 and Zn-dolomite A note on the asymmetry of the solvi TERNARY

PHASE

RELATIONS

66

CaC03-MgC03-FeC03 CaC03-MgC03-MnC03 Order-disorder relations in Fe- and Mn-containing dolomites The systems CaC03-MgC03-CoC03 and CaC03-MgC03-NiC03 SUGGESTIONS

FOR FUTURE

ACKNOWLEDGMENTS

51 51 55 58 60 60 62 63 63 64 64 65

WORK

66 70 70 72

75

. . . . . .

76 Chapter

3

SOLID SOLUTIONS and SOLVI among METAMORPHIC CARBONATES with APPLICATIONS to GEOLOGIC THERMOBAROMETRY

3

Eric J. Essene

INTRODUCTION

77

EXPERIMENTAL

DATA BASE

77

DETERMINATION

OF CARBONATE

COMPOSITIONS

78

COMPOSITIONS

OF METAMORPHIC

CARBONATES

80

Rhombohedral Orthorhombic SOLVUS

LIMITS

Solvi Solvi APPLICATIONS

carbonates carbonates

IN METAMORPHIC

in the system in the system

81

84 CARBONATES

CaC03-MgC03-FeC03 CaC03-MgC03-MnC03

OF CALCITE-DOLOMITE

Regional metamorphic Contact metamorphic THE ARAGONITE-CALCITE

. •

THERMOMETRY

rocks rocks

TRANSITION

85 85 86 88 89 91

AS A THERMOBAROMETER vi

93

CHAPTER 3, continued COEXISTING

ORTHORHOMBIC

SUGGESTIONS SUMMARY

Page AND RHOMBOHEDRAL

CARBONATES

94

FOR FURTHER WORK

95

..

95

ACKNOWLEDGMENTS

96 Chapter

4

MAGNESIAN

CALCITES:

SOLUBILITY Fred T. Mackenzie, Michele Loijens,

4

LOW-TEMPERATURE

and SOLID-SOLUTION

OCCURRENCE, BEHAVIOR

William D. Bischoff, Finley C. Bishop, Jane Schoonmaker & Roland Wollast

INTRODUCTION

97

LOW-TEMPERATURE

OCCURRENCE

98

Skeletal magnesian calcites Magnesian calcite cement UNIT CELL PARAMETERS: High-temperature

98 98

BIOGENIC AND SYNTHETIC INORGANIC PHASES.

. • ..

synthetic materials

104 104 105 107 108

Synthesis techniques Unit cell parameters Low-temperature synthetic materials Biogenic materials SOLUBILITIES

AND SOLID SOLUTION BEHAVIOR

Solubilities

of magnesian

ll2

calcites

114 114 118 118 119

Dissolution experiments Precipitation experiments Interpretation of experimental data Theoretical

considerations

ThermodYnamic equilibrium Stoichiometric saturation Experimental tests of stoichiometric saturation Expression of solubility

120 124 126 127

Evaluation of thermodynamic properties from dissolution experiments Other approaches to estimation of magnesian calcite properties

Solid solution properties Estimation of free energy of mixing from high-temperature data Estimation of excess lattice energy Evaluation of the heat of mixing by calorimetric measurements Hypothesis

of an hydrated magnesian

SOME CONCLUDING

REMARKS

ACKNOWLEDGMENTS

. . . . CRYSTAL

129 132

132 135 136 138

calcite

140 142 144

Chapter

5

104

5

CHEMISTRY and PHASE RELATIONS ORTHORHOMBIC CARBONATES J. Alexander

of

Speer

INTRODUCTION

145

CRYSTALLOGRAPHY

145

vii

CHAPTER 5, continued

Page

Crystal structures Charge distribution AC03 polymorphism Twinning Morphology The CaBa(C03)2 phases Carbocernaite CHEMISTRY

145 150 151 152 152 154 157

. . . . .

158

Inorganic orthorhombic carbonates Biogenic orthorhombic carbonates Recent she l.le Fossil shells ISOTOPIC

158 161 161 167

COMPOSITION ..

169

Dependence Dependence Dependence

on crystal chemistry on physical conditions on source

PHASE RELATIONS

.

170 170 171

.

.

Unary systems Binary systems

177 minerals

OCCURRENCE OF ORTHORHOMBIC CARBONATES AND DISCUSSION COMPOSITIONS AND MINERAL ASSEMBLAGES PHYSICAL PROPERTIES

178 OF THEIR

. .

182 184 185 188 188 188

FOR FUTURE WORK . .

189 Chapter

6

178 182

Density Lattice parameters Optical properties Luminescence Magnetic properties Infrared and Raman spectra SUGGESTIONS

171 171 173 173 174 174 174 175 175 176

CaC03-SrC03 CaCOrBaC03 CaCOrPbC03 SrC03-BaC03 SrC03-PbC03 BaCO 3-PbCO 3

Ternary systems Phase relations in aqueous systems Phase relations involving noncarbonate

.

6

The POL YMORPHS of CaC03 and the ARAGONITE--CALCITE

TRANSFORMATION

William D. Carlson INTRODUCTION THE POLYMORPHS

191 OF CALCIUM CARBONATE

191

The calcite-aragonite equilibrium CaC03(II) and CaC03(III) CaC03(IV) and CaC03(V) The role of rotational disorder of C03 groups Evidence for disordering viii

193 195 196 198 198

CHAPTER 6, continued

Page

Possible disordering schemes Speculations on disorder in CaC03 as a function of pressure and temperature ARAGONITE-CALCITE

TRANSFORMATIONS

The ca1cite-to-aragonite

IN THE SOLID STATE

The transformation under conditions of aragonite stability The transformation outside the aragonite stability field transformation

Petrographic observations Experiments measuring bulk transformation rates Independent determination of growth rates Extraction of nucleation rates from bulk transformation rates Summary THE ARAGONITE-CALCITE Observational

TRANSFOR}~TION

IN THE PRESENCE OF WATER . . . ..

evidence for the nature of the aqueous transformation

Marine environments Fresh water environments Mechanisms

of the transformation

Dissolution at surface of parent crystal Nucleation of product crystal Transport of complexes in solution Precipitation at surface of product crystal Interpretation mechanisms

of natural occurrences

211 212 212 212 213 214 214 214 214 215

attempts to quantify reaction kinetics

The significance of transport processes The time dependence of the overall transformation Inconsistency of a time-squared volume dependence with other observations A possible explanation for the time-squared volume dependence CONCLUSIONS

202 202 203 203 204 204 207 207 209

in terms of reaction

Reasons for extremely limited extent of transformation in sea water Reasons for apparent differences in transformation type Reasons for oriented overgrowths and topotaxial replacement Reasons for selective transformation of skeletal materials Reasons for long-term preservation in ancient limestone Experimental

201 202

transformation

The aragonite-to-ca1cite

200

AND OPPORTUNITIES

FOR FURTHER RESEARCH

. . . . . . . . ..

ACKNOWLEDGMENTS

215 215 217 218 218 219 219 219 220 222 224 225

7

Chapter

The KINETICS of CALCIUM CARBONATE

7 DISSOLUTION and PRECIPITATION

John W. Morse INTRODUCTION

227

BASIC PRINCIPLES

228

REACTION KINETICS

231

IN SIMPLE SOLUTIONS

231 231 234

Dissolution

General observations Models and mechanisms ix

CHAPTER 7, continued

Page

Precipi tation

238

General observations and models Secondary nucleation

238 239

DISSOLUTION AND PRECIPITATION REACTIONS OF NONBIOGENIC CARBONATES IN COMPLEX SOLUTIONS . . . . . . . . . . . . . . . . . . . . . . .. General considerations

Major influences Models for reacting surfaces and inhibitors Individual seawater-component

241 242

influences

244

Magnesium Sulphate

244 246

Reaction kinetics

in seawater and related solutions

246

Dissolution Precipitation

246 247 249

Other specific influences

Phosphates Heavy metals Organics

249 252 253

SPECIFIC TOPICS . . . .

254

Biogenic carbonate dissolution kinetics

254 254 256 258 261 261 262

General considerations Influence of grain size Prediction of solubility from kinetics Kinetic influence on coprecipitation

reactions

Experimental methods Cation coprecipitationwith calcite ACKNOWLEDGMENTS

264 Chapter

8

TRACE

241 241

ELEMENTS

and ISOTOPES

8

in SEDIMENTARY

CARBONATES

Jan Veizer

INTRODUCTION INCORPORATION

265 OF TRACE ELEMENTS INTO CARBONATE MINERALS

INCORPORATION OF STABLE ISOTOPES OF OXYGEN AND CARBON INTO CARBONATE mNERALS Theoretically predicted composition of carbonate minerals Isotopic variations in natural waters Isotopic composition of natural carbonates INCORPORATION

OF RADIOGENIC

Radiocarbon U-series disequilibrium Isotopes of strontium DIAGENETIC REPARTITIONING RECORD OF TERRESTRIAL

ISOTOPES INTO CARBONATE MINERALS.

267 271 272

276 278 279 279 282 284

nuclei

OF TRACERS

285

EVOLUTION IN ANCIENT CARBONATES

Oxygen isotope paleothermometry Post-Triassic paleoceanography Carbon isotopes of ancient oceans: the story of life Strontium isotopes and buffering of the oceans x

288 289 289 292 294

CHAPTER 8, continued

Page

Oxygen isotopic composition of sedimentary Secular variations in chemical composition CONCLUDING

carbonates of carbonates

299

REMARKS

ACKNOWLEDGMENTS

300 Chapter

9

296 298

MICROSTRUCTURES Hans-Rudolf

9 in CARBONATES

Wenk, David J. Barber

& Richard

J. Reeder

INTRODUCTION

301

ORIGINS OF MICROSTRUCTURES

303

Deformation microstructures Transformation microstructures Growth microstructures

303 306 3ll 3ll 3ll 313 313 313

Dislocations Growth bands Faults Stacking disorder Twins METHODS OF ANALYSIS

. . . .

313

Scanning electron microscopy (SEM) Transmission electron microscopy (TEM) Scanning transmission electron microscopy (STEM) High resolution electron microscopy (HREM) POLYMORPHIC

TRANSFORMATIONS

. . . . . .

Cation ordering in dolomite Aragonite ~ calcite transformation

Observations in sedimentary rocks Observations of the solid state transformation in the TEM VARIATIONS

IN STACKING ORDER

• . . . . . . . . . . . . • .

Periodic basal superstructure in dolomite Stacking disorder and polytypes in rare earth carbonates MODULATED

STRUCTURES

PARALLEL TO r

=

{1014} . . . . . . . . .

322 326 328 328 329 332

352 352 354

. . . . . .

Deformation

319 321

352

CARBONATES

Recent dolomites Magnesian calcites DEFORMATION

319

332 338 343 344 348

Calcian dolomites Saddle dolomites Calcite Carbonatites Interpretation OTHER SEDIMENTARY

314 315 317 318

354 355 356 357 359

mechanisms

Calcite Dolomite The effects of twinning

Biaxiality

xi

CHAPTER

9, continued

Other defects

in deformed

Page carbonates

Dislocations Stacking faults Stylolites - "Pressure solution" ACKNOWLEDGMENTS

R

.

. . . . . .

REFERENCES

xii

...

.

359 359 360 364 367

369

CARBONA TES: and CHEMISTRY

MINERALOGY

FOREWORD This is Volume 11 of REVIEWS in MINERALOGY, Course Notes" for the Mineralogical of the mineralogy the M.S.A.

and chemistry

of carbonates

in 1974.

This review

appears just four years after

Short Course on Marine Minerals which covered marine manganese

oxides and iron oxides, barite,

a series begun as "Short

Society of America

evaporite,

silica polymorphs,

and placer minerals.

Volume 6, Marine Minerals,

because,

coverage

mineral

of this important

and the zeolite, clay, phosphorite, Carbonates were not included in

as editor Roger Burns put it, " ... group warrants

a separate monograph."

Several years ago, Rich Reeder, who was and still is very excited about new vistas

of research

tron microscope bonates

centered around application

to carbonates,

volunteered

for the Mineralogical

Society of America.

by this volume with its fifteen authors, 1983, just prior to the annual meetings of America

at Indianapolis,

of the transmission

to organize

elec-

a short course on car-

His success is commemorated

nine of whom lectured October of M.S.A. and the Geological

28-30,

Society

Indiana. Paul H. Ribbe Blacksburg, VA September 1, 1983

PREFACE and ACKNOWLEDGMENTS This volume of Reviews in Mineralogy understanding

attempts

to synthesize

our present

of certain aspects of the mineralogy

and chemistry

of the rock-

forming carbonates. of research. mentary)

Hopefully,

it reflects

the presently

more active areas

This review follows, by ten years, a major assessment

carbonate

minerals

of subject material,

by Lippmann

(1973).

of (sedi-

There is only minor overlap

and I hope that this difference

reflects

fairly how this

field has developed. In some respects mineral chemical

carbonates

groups providing processes

lative importance

an abundant

throughout

Moreover

record of biological,

much of geologic

in sedimentary

more emphasis here.

are unique, for they are one of the few

time.

rocks, low-temperature

the obvious iii

correlation

physical,

Because

and

of their re-

examples

are given

with energy resources

has been a significant terest in this area. reflection

factor contributing However,

of their widespread

vironments,

including

to the current resurgence

the broader

interest

occurrence

metamorphic

in carbonates

in vastly different

and igneous settings,

ciation of their role in both atmospheric

of in-

is also a

geologic

en-

as well as an appre-

and oceanic chemistry,

both past

and present. In this volume, crystal chemistry

some of the papers are general

and phase relations),

damental

nature and are of interest

coverage

and generally

geochemistry. a relatively

reflect

and they provide overviews

to many.

new and powerful in carbonate

technique

approaches

The various efforts.

A hurried

all gratefully

schedule,

these should be brought

for mineralogical

ways; in particular

however,

without

I thank my friends and colleagues

errors

in their to persist;

Sarah Vaughan,

the cooperation

Polytechnic

and

in various

at the State University

Support in various but essential

the final product.

torial assistance

were expeditious

Many others contributed

thanks go to Paul H. Ribbe for advice,

Ada Simmons,

of manuscripts

in this volume is still "fresh."

allows for unnoticed

all of whom I thank.

vided both by SUNY and by Virginia

for producing

that has

to my attention.

of New York at Stony Brook.

Special

research

the completion

acknowledged,

This volume would not have been possible help of the authors,

electron microscopy,

research.

much of the newer material

reviewers,

Institute reviews,

forms was pro-

and State University. copy editing,

Typing of the final manuscript

Theresa Beddoe and Margie Strickler,

and

was done by with edi-

by Julie Anne Ribbe. Richard J. Reeder Stony Brook, NY September 1, 1983

SPECIAL ACKNOWLEDGMENT The Mineralogical Company,

Society

of America

ARCO Oil and Gas Company,

for providing

in

used in carbonate

transmission

Owing to the short time interval between and publication,

of a fun-

Others are more specialized

the different

The final chapter introduces

great potential

(i.e., those addressing

scholarships

for graduate

is grateful

and SHELL Development students

carbonates.

iv

to AMOCO Production

to attend

Corporation

the Short Course on

1

CR YST AL CHEMISTRY

of the RHOMBOHEDRAL

CARBONATES

Richard J. Reeder INTRODUCTION Crystal minerals, whether

chemistry plays an important role in understanding

both in terms of reactivity

and stability.

in the solid state or in aqueous

Stability

preting

the potential

structure

differences

Crystal growth processes,

imposed by coordination

among natural minerals,

for chemical

and bonding,

properties.

A multitude

is available

to study these, but in all cases the essential

understanding

of the crystal structures

The history carbonates

of crystal structure

of calcite was firmly established that of dolomite Merwin

(1924).

sophisticated precise

Accurate

determinations

element is an

of the common rhombohedral

sporadic in nature.

The general structuI

early in the century by Bragg

just ten years later by Wasastjerna structural

data collection

refinements

methods,

of positional

in

of techniques

involved.

investigations

has been long and somewhat

and

so important for inter-

change, reflect detailed differences

as do physical

of

systems, are influenced by the arrangE

ment of atoms in the crystal and the constraints bonding.

the behavior

(1914), and

(1924) and Wyckoff and

awaited the development

and by 1965, studies had yielded

and thermal parameters

for calcite

of more fairly (Sass

et al., 1957; Inkinen and Lahti, 1964; Chessin et al., 1965) and dolomite (Steinfink and Sans, 1959).

Lippmann

at that time, giving a more complete tinued interest

(1973) thoroughly historical

in carbonate minerals

recent studies; The preferred although

carbonates.

have been undertaken

transmission

have been used (see Chapter

based on the more

to much of the older work, see Lippmann

technique for structural

Raman) of carbonates

for most of the com-

The present review is primarily

for references

recently

Since then, con-

has fostered many new crystal structure

studies, and today modern refinements mon rhombohedral

reviewed data available

development.

studies has been x-ray diffraction,

electron microscopy

9).

(1973).

and electron diffraction

Studies of the vibrational

have been reviewed by White

spectra

(IR and

(1974) and Scheetz and White

(1977) • Of the rock-forming accounting

carbonates,

calcite and dolomite are the most abundant

for more than 90% of natural

carbonates.

The structure

of calcite

is descriptively

fairly simple, and it is the same as that taken by several

other anhydrous,

single carbonates

rhodochrosite

(MnC03), otavite

(CoC03), and gaspeite Single carbonates

(NiC03).

including magnesite

(CdC03), smithsonite

(MgC0 ), siderite (FeC0 3 3 (ZnC0 ), sphaerocobaltite 3

with cations larger than Ca2+ form with the orthorhombic

aragonite

structure. The explanation given for this preference usually relates 2 to the size of Ca +, which is near the limit for 6-fold coordination. In the

aragonite however,

structure,

the cat-ion is 9-coordinated.

that at high temperatures

invert to a rhombohedral,

the orthorhombic

calcite-like

RHOMBOHEDRAL

structure

VERSUS HEXAGONAL

When dealing with the rhombohedral either a rhombohedral

or hexagonal

It is interesting BaC0

and SrC0

3

3

(Chang, 1965). AXIAL SYSTEMS

carbonates,

axial system.

smallest unit cells for both calcite and dolomite hedrons

carbonates

to note,

one has the choice of using It should be noted that the

are acute primitive

rhombo-

(Fig. 1) with cell contents of 2 CaC0

and 1 CaMg(C0 )2' respectively. 3 3 Rhombohedral cells are described entirely by two parameters -- a , the edge r length, and ~r' the angle between any two of the three equal length edges forming the apex of the rhomb.

Since the line connecting

the apices of the rhombo-

hedron is a three-fold

axis, all lattice points of the rhombohedron

equally well described

using a hexagonal

have two possible

orientations

axial system.

with respect

can be

The rhombohedron

to the hexagonal

may

axes called the

obverse and the reverse settings (International Tables for X-ray Crystallography, Vol. I).

The obverse setting shown in Figure 2 is now standard.

both calcite and dolomite,

the heights of the hexagonal

for the acute rhombohedral

cells, although

respectively,

In

cells are the same as

they contain 6 CaC0

3

and 3 CaMg(C0 ); 3

and are triply primitive.

The primary advantage tion of the axes relative

of the hexagonal

with respect to one another, and all lie perpen-

0

dicular to the c axis.

The three a axes, of equal length,

to one another.

intersect at angles of 120

This arrangement

makes it relatively

geometric aspects of the crystal structure, tinct levels along the c axis. hedral axes, particularly

especially

easy to visualize

planar features at dis-

This becomes rather awkward using the rhombo-

so when comparing

angle between axes changes from one mineral Modern descriptions

axial system lies in the disposi-

of rhombohedral

different minerals because the to the next.

carbonates

are given almost exclu-

sively in terms of the hexagonal

cell.

uses a rhombohedral

which may be based on either the true unit

description,

cell or the morphologic

Much of the older literature,

cell (also called a cleavage cell).

however,

The latter (shown

in Fig. 3) should be avoided since it is not a correct unit cell.

A proper

morphologic

unit cell does exist, but its a cell edge is twice that of the

morphologic

cell shown in Figure 3, and its cell contents are 16 times that of

the acute rhombohedral

cell.

Owing to the presence

of the fourth axis

2

(as) used in the hexagonal

axial

Figure 1 (to the left). Schematic illustration of the true rhombohedral unit cell for calcite containing 2 CaCO). The apical angle a is 'V 46° for this acute rhombohedron. Closed circles represent Ca positions; two C03 groups are shown. After Lippmann (1973)"

l 5

Figure

4

obverse lation view, X-ray

3 2

2 (below). Illustrations showing the setting of the rhombohedron in reto the hexagonal unit cell: (a) side (b) plan view. From Int'l Tables for Crystallography, Vol. I.

o

b.,

a

.~

.j '0

'0

'0

.~ .j '0

.: .j ,0

1

0

·i

+x,

-yo .j

aR

.~

.~ •j

= 46' 07"

«« = 101' 55"

Figure 3. Schematic illustration ahowf.ng the relationships between the true rhombohedral unit cell (the acute rhombohedron) and the morphologic (or cleavage) cell in calcite. Notice that the true cell is doubled in height with respect to the morphologic cell. The hexagonal unit cell is also shown. From Hurlbut and Klein (1977)"

3

.j

system, four-symbol h + k + i writing

=

0,

indices should be used -- h k i £, where

Miller-Bravais

Although

some authors prefer to drop the superfluous

indices, its presence

leaves no doubt that hexagonal

i when

indices are in

fact being used, The following and rhombohedral h k h

k:

r

- £

r

expressions

h

r

between hexagonal

Transformations

r

for fractional

2/3 hh+ 1/3 kh + 1/3 £h

r

k

r

+ kr + £

r

are useful when converting

indices:

r

-1/3 hh+ 1/3 kh + 1/3 £h

zr

-1/3 hh- 2/3 kh + 1/3 £h

atomic coordinates

between

the two systems

are as follows: xh

=

2/3 xr - 1/3 Yr - 1/3 zr

x

r

=

Yh

=

1/3 xr + 1/3 Yr - 2/3 zr

Yr

=

-xh + Yh + zh

zh

=

1/3 xr + 1/3 Yr + 1/3 zr

Z

=

- Yh + zh

Cell parameters

r

xh

+ zh

can be related by the following

conversions:

~

2a

r

sin

2r

a

r

(3a 2 h

= 1/3

+ c2)

2 h "

~ sin....!. = a /2a 2 h r All descriptions

in the present

chapter will refer to the hexagonal

unit

cell. THE C03 GROUP AS A STRUCTURAL

UNIT

The C03 group is the fundamental chemical unit from which the carbonate minerals derive their identity, This anion is also present in some minerals that might not be best described

as carbonates,

scapolite being a good example,

There in fact exists a large number of minerals where the C0 several anions

(or polyanions)

in the structure.

common "other" anion in these mixed-anion Despite

the wide variety of carbonate

Hydroxyl

group is one of 3 is perhaps the most

carbonates, mineral

structures,

the basic con-

figuration cases,

of the C03 group is found to be remarkably uniform in all but a few In its general form the C03 group resembles an equilateral triangle with

oxygen atoms at the corners and a carbon atom in the center, figuration

is attained

in calcite and its isotypes, where the local point sym-

metry of the group is 32, (6m2) characterizes

This exact con-

Zemann

(1981) notes that a higher point symmetry

one of the two distinct C03 groups in fairchildite 4

(K Ca 2

(C0 )2)' although the structure of the group is essentially identical with that 3 in calcite. Slight deviations from this form reduce the point symmetry below that in calcite, and reflect

differences

in coordination

or electronic

environ-

ment. Zemann

(1981) has compiled recent structural

data for C0 groups in 30 dif 3 He found the mean value for the O-C-O angle to be

ferent carbonate minerals. 120

in agreement with the ideal value.

0

rather small, and the great majority

o-c-o

angle approximates

large deviations

Deviations

of carbonates

from this are generally have C0

groups where the 3 In a few minerals,

the ideal value rather closely.

may be found

(up to 11

in one case), but seem to occur only

0

where the edge of a C03 group is shared with an edge of a coordination polyhedron of an adjacent cation. In acid carbonates, such as NaHC0 and KHC0 , 3 3 the o-c-o angles (now for an HC03 radical) are found to deviate from 1200 by several degrees. Zemann found the mean

between

son is provided

bond length to be 1.284

by the x-ray structural

calcite, magnesite, structural

siderite,

A

on the

c-o

result of the short

Tabulated

c-o

but quite small variation.

values range

A very useful compariet al. (1981) for

and smithsonite.

of the

Within this iso-

A,

bond length is only 0.005

Thus the effect of different

bond length within this structure

c-o

with a standard devia-

data of Effenberger

rhodochrosite

series the maximum variation

a clearly measurable cations

c-o

A

for average values in a C03 group. 1.25 and 1.31 A in all but a few extreme cases.

tion of only 0.004

type is minimal.

bond length is a rather close approach

1he

for oxygen atoms

the C03 group -- 2.22 A. Slightly shorter nonbonding 0 ..•0 distances are known for NaN03 and stishovite (cf. Table 5 of Shannon and Prewitt, 1969).

within

to the C03 group, the c-o bond lengths in the acid carbonates within individual HC03 ions. Sass and Scheurman (1962) rep or bond lengths of 1.346, 1.264, and 1.263 for NaHC03. In contrast

vary considerably

A

In those carbonates

where layers of C03 groups separate layers of dLf f eren: cation types, the carbon atom is displaced from the plane formed by the 3 oxygel atoms.

Dolomite,

with its two distinct cation layers,

Recent x-ray structural show displacements Althoff,

refinements

A

1977) appears

is reported

of ordered dolomites

of 0.018, 0.020, 0.022 and 0.027 to be somewhat

more typical for dolomite. by Knobloch

serves as a good example

A.

(see Table 8 below) The latter value (from

large, and the other values are perhaps

In buetschliite et al. (1980).

K2Ca(C03)2' The direction

a displacement

of the displacement

always toward the layer of cations with the smaller radius. position

of unlike cation layers forms an ordered

displacement

is sensitive

is

Since the juxta-

superstructure,

to the degree of cation ordering. 5

of 0.038

the average

For dolomite,

Reeder and Wenk

the deviation

(1983) have shown that the magnitude as cation disorder

ment decreases

the C0

reinforced

3 by consideration

in calcite

of the

c-o

(Peterson et al., 1979) and dolomite data) show a buildup

0, supporting

this view.

In general,

than any other M-O bond typically bond strength calculated

When we consider

the

c-o

density dis-

(Effenberger

et al.,

In calcite,

Brown-Shannon

strongE

for examplE method

(Bro.

four times greater than for the Ca-O bond. in the calcite structure

section, it will become clear that the excellent to this substantial

is

regarded as

bond will be considerably

using the empirical

atomic positions

to

of charge density between C and

found in carbonates.

and Shannon, 1973) is approximately

attributed

where

Such a viewpoint

Recent studies of electron

unpublished

c-o

unit.

bond, which is generally

in character.

1983; Reeder,

the

displace-

in most carbonates

of bond lengths and angles, it is reasonable

group as a fairly rigid structural

being strongly covalent tribution

of the average

However

occurs it is quite small.

Based on the uniformity consider

increases.

difference

in the following

cleavage can, in part, be

in bond strengths.

The cleavage

plane, {10I4}, is one that breaks the least number of Ca-O bonds and no

c-o

bonds. Several authors have considered (Effenberger

the charge distribution

et al., 1981; Yuen et al., 1978; and references

there is much disagreement

in earlier

studies, Yuen et al. and Effenberger

al. find point charges of approximately for single rhombohedral

in the C03 group therein). While et

+1 and -1 for C and 0, respectively,

carbonates. The R3c CARBONATES

The calcite structure Traditionally structure

the calcite structure has been described

as a starting point.

The primitive

cubic NaCl structure is a rhombohedron

using the NaCl

unit cell of the face-centered

whose 3-fold axis is coincident

the 3-fold axis of the cubic unit cell (i.e., the body diagonal), volume is one quarter of the latter.

with

but whose

If Na atoms are replaced by Ca atoms

and Cl atoms by C03 groups, a structure somewhat resembling calcite results. By necessity, the rhombohedron must be compressed in a relative sense along the 3-fold axis to accommodate

the "flat" C0

3

groups now lying perpendicular

to this axis. At once we see the limitations spherical

symmetry, being centers of symmetry

C03 groups are triangular. successive

of this description.

carbonate

The different

layers are nowhere 6

The Cl atoms have a

in the NaCl structure, while the

orientations

foreshadowed

of the C03 groups in in the NaCl structure.

z:

Figure 4 (to the left). Schematic illustration of the hexagonal unit cell of calcite, a ""4.99 a = 17.06 Ca atoms are open circles; CO) groups are triangles; z. coordinates for Ca layers are shown on the right. From Lippmann (1973) .

1.

........... %

A.

..........%=213

.............. 3/6= 1/2

Figure 5 (above). The CO) layer in calcite as seen down the c axis. The CO) groups have identical orientations throughout the layer. Dashed circles show the positions of Ca atoms projected from the cation layer above the CO) plane. Relative sizes of the Ca and 0 circles correspond to crystal radii as assigned by Shannon and Prewitt (1969).

In fact, we will see that the rhombohedral

cell derived from the NaCl structure

is not a proper unit cell for calcite. Figure 4 shows a schematic model for the calcite structure.

It can be seen

layers along the c axis.

that layers of Ca atoms alternate with carbonate

The

C0

groups have like orientations within each layer, but reversed orientations 3 in successive layers. Thus the translation along c to similarly oriented C03 layers is doubled

ture.

in comparison

This requires Perhaps

a doubling

to the analogous

a better description

of the oxygen atoms are teferred Megaw, 1970, 1973).

translation

in the NaCl struc-

of the unit cell height as shown in Figure 3. of the structure

is provided

to a pattern of hexagonal

While only approximate,

if the positions

close packing

(cf.

this analogy does show the orien-

tational relationships

among C03 groups in successive layers. Within a given layer of hexagonally-packed oxygens, carbon atoms are introduced such that each

oxygen coordinates

with only one carbon reSUlting

of C in that layer.

in a hexagonal

Calcium atoms then fill those octahedral

(between oxygen layers) that avoid coordination

distribution

intersticies

with two oxygens of the same

C0

group, since that 0-0 distance would be very short (Fig. 5). C atoms in 3 successive oxygen (or C03) layers are distributed so as to avoid superposition over either Ca or C atoms in adjacent cation and C03 layers. This ensures that Ca is then coordinated to six oxygens, each belonging to different C03 groups. The unit translation

along the direction 7

of stacking

(i.e., along the c axis)

Table I.

Fornal

description

of calcite

is equivalent

structure

to the height of six C0

3

layers.

Ce11 parameters':

Atom

pos i t Ions

In reality,

R3c (No. 167) Hexagonal. rhombohedrally-centered 6 CaC03

Space group: Unit cell: Cell contents:

s ":

the oxygens are not

closest packed, and in fact, the ideal hexagonal pattern

a = 4.9896 j\ C =17.0610 j\

is not attained owing

to the foreshortening

Ca in 6 (b): 0, 0, 0 C in 6 (a): 0,0, Ii o in 18(e): x , 0, Ii x = 0.2568

*Cell parameters for other calcite carbonates are given in Table 2.

different C03 groups are 3.19 K (across C03 planes) and 3.26 A (in-plane), both

structure

considerably

**Oxygen positional parameters for other carbonates are given in Table 2.

A

of the 0-0 distancE

within the C03 groups (2.22 K). In calcite, the shortest 0-0 distances between

and 2.93

A,

(2 x 1.36

0-0 distance~

in magnesite

are 2.85

which in many ways resembles

the other hand, it faithfully

=

tive ionic radius of oxygen

2.72 K). description,

The corresponding

respectively.

it does not strongly enough emphasize

larger than twice the effec-

Another

the description

disadvantage of corundum,

the rigid nature of the C0

represents

3

the proper orientations

of thi~ is that

group.

On

of the C0

3

groups, and leads to a correct choice of the unit cell. The formal description

of the calcite structure (and that of its isotypes) 2 The M + cation position is taken as the origin and is a

is given in Table 1. special position.

In addition

to being a center of symmetry, this site also As noted previously, the M2+ cation is 6-coordinated

lies on the triad axis. in the calcite structure, The dispositions 7.

forming a slightly distorted

of the octahedra

relative

Within a given layer, the octahedra

nor corners.

Adjacent

corners of octahedra

octahedra

octahedron

(Fig. 6).

to one another are shown in Figure

are independent,

sharing neither edges

are linked both by C0

groups and through the 3 (i.e., along the c axis).

from layers above and below

Each octahedron shares its corners with other octahedra,

three from layers above

and three from below. It is also useful to consider C03 group.

the coordination

Figure 8 shows such an arrangement.

sphere around an individual

Each oxygen is bonded to one

carbon and two cations, each of the latter from adjacent Both C and 0 lie on special positions,

as well.

symmetry 32 and the oxygen position point symmetry parameter which is not constrained

c-o

bonds are aligned parallel

2.

cation layers.

The C position has point The only positional

by symmetry is the x parameter

to the three a axes.

for O.

The

The overall point symmetry

for calcite is 321m. Calcite

isotypes

Charge balance

requires

that carbonate 8

isotypes of calcite contain only

Figure 6. Stereoscopic projection of a portion of the calcite structure showing the central Ca coordinated to six oxygens from distinct CO) groups, prepared using the program ORTEP (Johnson, 1965). Vibrational ellipsoids represent the 90% probability e Let ron distribution surfaces. Note the orientations of the oxygen ellipsoids.

Figure 7. Schematic projection down the a-axis showing one layer of Ca06 octahedra in calcite. Large closed circles are Ca atoms (not to scale) at the center of the octahedra. Open circles at corners of the octahedra are oxygen atoms; dashed lines of octahedra depict lower edges. Cj.osed and open circles between octahedra are C atoms in the upper and lower CO) layers, respectively. CO) groups link corners of neighboring Ca06 octahedra, as shown at left. Octahedra also share corners with other octahedra from layers above and below. The dotted triangle at top shows the basal edges of a Ca06 octahedron from the next layer above.

Figure 8. Stereoscopic projection of the coordination environment of a CO) group in calcite, prepared using the program ORTEP (Johnson, 1965). Vibrational ellipsoids represent 90% probability electron distribution surfaces. Note that the Ca and C ellipsoids are nearly isotropic, while that for oxygen is anisotropic.

9

divalent cations as major elements. to form calcite-type

carbonates

For at least two multivalent

(Mn and Fe), various

shown them to exist in the divalent Wildeman,

spectroscopic

state (e.g., Takashima

1970; Urch and Wood, 1978).

divalent cations:

Pistorius

the synthesis at high pressure

studies have

and Ohashi,

The calcite structure

bonates of the following (1960) reports

cations known

1968;

is found for car-

Ca, Mg, Fe, Cd, Mn, Zn, Co, and Ni. of a phase that might

be CuC0 , although confirmation is lacking. 3 Recent x-ray structure refinements have been done for calcite, magnesite, siderite,

otavite, rhodochrosite,

magnesian

calcite and a Cd-rich calcite.

most recent or complete

smithsonite,

and two solid solutions -- a

Selected structural

studies are presented

in Table 2.

some insight into the changes produced by substituting

data from the

These data provide

different

cations in the

structure. In order to understand

Atomic thermaZ vibrations. atom in its local environment, of their vibrational this description,

factors reported by Effenberger

3.

equivaZent

Ellipsoids

axis.

For

of the principal

axes

et al. (1981) for calcite, magAlong with these amplitudes,

isotropic temperature factors

of revolution,

the shapes and sizes

from the refined anisotropic

and smithsonite.

for the M2+ and C positions

to be ellipsoids

of an

as deduced from the thermal parameters.

ellipsoids were calculated

nesite, siderite, rhodochrosite, calculated

it is useful to consider

the root mean square (RMS) amplitudes

of the vibrational temperature

ellipsoids

the behavior

(B) are also given in Table

are constrained

the revolution

by point symmetry

axis coinciding with the 3-fold

For 0, which lies on a diad axis, the ellipsoid

is only constrained

to

axis coincident with the diad (i.e., the c-o bond direction) The amplitudes for M2+ and C in the R3c carbonates listed in Table 3 indi-

have one principal

cate that these ellipsoids hand, the ellipsoids coincides with the the

c-o

are nearly spherical

for oxygen are clearly anisotropic.

c-o

bond direction.

bond is by far the strongest

inclined with respect

shapes of the ellipsoids

The shortest axis'

This result is to be expected

in the structure.

in Fig. 8).

The ratio of amplitudes

Not surprisingly,

since

The longest axis is

in other respects as well.

This is especially

Calcite

isotropic

is somewhat anomalous

temperature

factors

larger than for the other pure carbonates,

pronounced

of this

for the longest to shortest axes is roughly

The equivalent

for calcite are considerably

the direction

to the "plane" of the 3-fold coordination

1.4 to 1.5, except in calcite where it is 1.9.

atoms.

On the other

to the c axis at angles varying from 22° to 480 (see the

longest axis is roughly perpendicular of oxygen.

in all cases.

(B) for all

for 0 where the B value for calcite is

twice that of the other pure carbonates. 10

These values reported

(for calcite)

,...., 00

a>

,....,

..... '-'

....;

....;

I-t

bI) .....

.... 4)

.....

,....;

4)

I-t .....

I-t

4) '-' bI)

4)

I-t '"

I-t

'"a>

....OS I-t

t.:>

a>

a> "I

,....,

-e- ,-,

'-'

00 0

"I "I

'-' 00 '-'

-o

N '-' 0 00 "I

"I'"

.....

,....,

'"

-o

'-'

00 r-,

oJ)

,-,,-, '-' ..... N 00 '-' '-' a>a>N

,-,

-e

00 '-'

a>

'" '"

'-' 0

... ...'" 00 0

a>

'"

"I

,-,

~

,-, N '-'

'" '"'" '" '" '" .....

oJ)

'-' r-, 0

'" '" '"

..... .....

...... .....

'-' 0 0

.....

.....

,-, "I '-' 00 00 N

N N

N

"I

N

"I

,...., ..... ,-,

,-,

8 .....

,....,

,-,

"I

1.1")

-o

NN

.....

oJ)

00

oJ)

a>

.....

.....

"I

'-' N N "I

,-, "I '-'

'-'

"I

00

oJ)

.....

N '-'

a>

'-'

oJ)

....

..... .....

'-'

'-'

-o

,-,

,....,

,-,

-o

r-, '-'

"I '-'

'-'

a>

00

.....

.....

'"

""

"" r-, "I

-o

-o

'"

"I

'"'" r-,

oJ)

N "I

N

"I

N

I-t

....

000

00

0

0

0

0

0

0

...

,....,

,-,

~ .....

;;;-

'-'

'-'

>


r-,

,....,

'" '-'

N '" ...."I 00 U":

oJ)

.......... .....

..........

,...., ,-,

,-,

00

a>

OJ

::0

4)

N

oJ)

r-,

N N '-' r-,

c2 '-' "'I

ell

N a> ,-, '-' ._, N

'" '" a> a>

'-' 00 "" 00 00

,-,

oJ)

..... 00 N

N 00

-o

""

00

"I 0

00

a>

.....

-o

..... '"

.....

oJ)

-o

,-, r-, '-'

,-,

a> a> a> """""

""""

.....

'"

-e-

z

oJ)

'-'

-o

1.1")

'-'

z

,-,

.....

N ,-, '-' 00 '-' N r-,

.....

I-t

0

'-'

.....

oJ)

4)

I-t

0

,....,

'" "" '" -o'" a>

N

.....

oJ)

-o

,a

,-, N '-'

.....

Vl

ell N

I" Calcite + Dolomite

5001-

-

>I

o

I

I

I

5

10

15

Figure 4. A portion of the calcite limb of the calcite-dolomite so Lvus . The reversal brackets and the solid curve are from the (unpublished) data of Fanelli and Wyllie (indicated as "This St udy" in the figure), using electron microprobe analyses of crystals grown 1n the presence of a liquid phase (see text). The dashed curve is the po!ybaric so!vus of Goldsmith and Newton (1969, see Fig. 3), and the dot-dashed curve from Goldsmith and Heard (1961).

Weight % MgC03 the dolomite

limb of the solvus is further complicated

order on lattice parameters tle under any circumstances.

by the effect of dis-

as well as by the fact that it changes rather litHarker and Tuttle

(1955) list a value of the

dolomite limb at 900°C as 43 wt % MgC03, or 47 mol %, the same as the Ca53Mg47 by Goldsmith and Heard (1961). At 500°C, Harker and Tuttle

determination

found 45 wt % MgC0

in the dolomite (their Sample 15), or 49.6 Mol %, and 3 and Heard report Mg . . Little has been done on the dolomite limb 49 5 since these studies. The Mg-rich dolomite limb and the magnesite limb in

Goldsmith

equilibrium

with it are steep and little affected by temperature

Consideration

of the magnesite-rich

at crustal temperatures

up to 900°C.

portion of the system, of little interest

and pressures,

will be considered

shortly with rela-

tion to studies under upper mantle conditions. Fanelli and Wyllie completed

(in preparation;

a study at 2 kbar of subsolidus

see also Fanelli

et al., 1983) have

and melting relations

in the system

CaO-MgO-C0 -H 0, and have re-examined the calcite limb of the solvus using 2 2 electron microprobe analyses. At temperatures of 650°C and above, the calcite and dolomite are in equilibrium crystals

grow to 25-30 µm.

for microprobe

analysis.

enough to analyze, reversed

with a hydrous carbonate

The dolomite and magnesite

melt, and the calcite

crystals are too small

Below 650°C the calcite crystals

although the system is subsolidus.

at 600°, 650°, 700°, and 750°C.

The equiiibrium

was

The Fanelli and Wyllie data are

shown in Figure 4, with the curves of Goldsmith 54

are still large

and Heard

(1961) and Goldsmith

and Newton Metz

(1969).

Earlier determinations,

(1976) investigated

discussed

above, are omitted.

tremolite + dolomite ~ foster site

the reaction

+ calcite + CO2 + HZO, and found the MgC03 content of the magnesian calcites produced

to be within 0.5 mol % of the solvus data of Goldsmith

(1961) and Goldsmith Puhan

and Newton

and Heard

(1969); see also Gordon and Greenwood

(1970).

(1978) determined

the MgC03 content of calcites produced by the reaction dolomite + K-feldspar + H20 ~ phlogopite + calcite + CO in the temperature 2 range 4l0o-640°C and found them in good agreement with the Goldsmith and Newton (1969) solvus. It would appear from the conformity techniques

that the calcite limb of the solvus is reasonably

the metamorphic temperature

temperature

range from greenschist

data would be desirable,

sitions is healthy,

but considering

and ease of exsolution in the carbonates

facies.

Lower

and further testing of equilibrium

compo-

the relatively

responsive

kinetic behavior

amount of back-reaction

rocks is to be expected

and is observed

Unless one can be sure that dolomite ex solved during retro-

gression can be distinguished

from that present at the "equilibrium

ture", and that it can also be

analyzed

tempera-

still 3 in solid solution in the calcite, little more than a "minimum" meta-

morphic temperature greater

well known in

to granulite

in this system, a significant

of metamorphic

(Goldsmith, 1960).

present

of these data obtained by different

can be determined.

(Gittins, 1979).

In addition,

along with the residual MgC0

The problem

in carbonatites

is even

the effect of other components

such

as FeC03 and MnC03 is not adequately known. Thus, for the purposes of geothermometry, further refinement of the binary CaC0 -MgC0 solvus appears 3 3 superfluous.

Relations at higher pressures and temperatures. melting and decarbonation

reactions

complicate

above studies moderate pressure was adequate In considering upper mantle, cifically

the Ca-Mg carbonates decarbonation

treated.

At higher tempeatures

the phase relations;

to maintain

under conditions

carbonate

in the solids.

more appropriate

to the

behavior will be indicated, but will not be spe-

Little will be said about melting, beyond indicating

on the CaC0 -MgC0 join. 3 3 Byrnes and Wyllie (1981) investigated

the subsolidus

it

and melting rela-

tions for the join CaC0 -MgC0 at 10 kbar, and Irving and Wyllie (1975) at 3 3 30 kbar, later corrected by Byrnes and Wyllie (1981) to 27 kbar. In addition, Irving and Wyllie deduced the relations on their pressure schematic

correction,

isobaric

calcite-dolomite

at 5, 6, 12, 15 and 25 kbar (or based

at 10% less than these values),

diagrams at these pressures.

and presented

The representation

of the

solvus and its behavior with pressure was taken from 55

Goldsmith

and Heard

(1961) and Goldsmith

and Newton

tion of fields in the diagrams not directly deduced from P-T data on decarbonation and Wyllie,

1975; Goldsmith

(1969), and the configura-

determined

experimentally

were

and melting of MgC03 and CaC03 (Irving decarbonation

and Heard, 1961), of dolomite

(Goldsmith and Heard, 1961), and P-T-X relations with respect to the calcite-aragonite

(Goldsmith and Newton,

1969)

transition

in the system CaC03-MgC03. Figure 5 shows the system aL 27 kbar, at which pressure no decarbonation takes

place, and the join is binary.

The aragonite-magnesian

tions and the lower temperature

portion

extrapolated

from Goldsmith

and Newton

calcite-dolomite

of the dolomite-magnesite

rela-

solvus,

(1969) are also illustrated.

Figure 6 shows the system at 10 kbar, only partially binary because of the presence

of peric1ase

(MgO) and vapor

of periclase

+ liquid + vapor, dolomite-like is included

for comparison.

The schematic

6, 12, 15 and 25 kbar from Irving and Wyllie The derivation

involving reactions

tailed here, but left for the aficionado 1975, especially

Fig. 6A).

+ vapor,

+ vapor.

isobaric

The 27

sections at

(1975) are shown as Figure 7 A-D.

of these diagrams from P-T data, extrapolations,

mated considerations

Wyllie,

in ternary fields

carbonate + periclase

carbonate + periclase

and a small field of magnesite-like kbar diagram

(C02) represented

and some esti-

in the join CaC03-MgC03 is not deto pursue on his own (see Irving and

Extrapolation

of the "dolomite"-magnesite

solvus at 27 kbar into the melting interval of Figure 6 gives a temperature approximately

l410°C for the solvus crest, still metastable

At somewhat higher pressures

the melting

of

at this pressure.

interval consists of a simple loop

with a minimum on the liquidus. In Figure 7 B-D the encroachment the calcite-dolomite of the Mg-free between

of aragonite

solvus is apparent.

aragonite

(plus Mg-calcite)

it and the single-phase

At still higher pressures continues

field of Mg-calcite

reaches the crest of the solvus, as indicated variant point where aragonite, It should be emphasized

on the Mg-calcite

Mg-calcite,

that the diagrams

limb of the field

to grow until the boundar: at higher temperatures

in Figure 8, producing

an in-

and dolomite are in equilibrium. of Figures

7 and 8 are largely

deduced and not known in detail. Nothing reactions

substantive

of the various

MgC03 at relatively Byrnes and Wyllie, fining pressures

has been said about the very important decarbonation carbonates.

high pressures

1981), all other investigations

that exceeded P

tions would significantly ena, and equilibria

With the exception

and temperatures

'

of the system CaC0 3 1975;

(Irving and Wyllie,

were carried out at con-

To discuss the known decarbonation

relaC02 lengthen this review, and along with melting phenom-

in hydrous systems, is beyond the scope of this paper, and 56

1500

u

0

~ ~

1000

C1>

a. E

~

500,

Ar+Cd

Cd+Cm

5 kb Ar+Cd

00

20 40 60 80 1000 20 40 60 80 100

CaC03 Figure

6 (to

the

join at 10 khar.

Mole % MgC03 right).

Figure 5 (to the left). On the. left is the isobaric temperature-composition section of the CaCOrMgC03 join at 30 kbar (later corrected to 27 khar) , in part extrapolated from data of Goldsmith and Newton (1969). Cc, Cd and em refer to carbonate solid solutions related to calcite, dolomite, and magnesite, respectively, at lower pressures and temperatures.. On the right is the section at 5 kbar. From Irving and Wyllie (1975, Figs. 6B and 6C).

MgC03

The CaCO MgC0

r

3

Salvus curves from Gold-

smith and Heard (1961). Also shown are results of Irving and Wyllie (1975) at 30 kbar, subsequently recalibrated to 27 kbar. Cc:: c a Lcd t e solid solution. em = magnesite

solid solution, Do - dolomite solid solution, Cd - carbonate solid Bolution near dolomite composition, Pe - periclase, L_ liquid, Wyllie

V - CO2 vapor. (1981).

From Byrnes

and

Figure 7 (below). Schematic isobaric sections of the join CaC03-MgC03 (see Irving and Wyllie, 1975, for details of construction). The isobars show successive changes as pressure increases from 5 kba r (Fig. 7D), to 30 kbar (corrected to 27 kbar , Fig. 5, left-hand diagram).

CoC°3 V

Ferroon Calcite Solid Solution




Zn

Fe

>

>

Co

>

Ni »

Cu.

doubt as to the stability of Zn-dolomite,

it falls

higher in their series than does Fe, and may therefore be considered being a contender Garavelli CaZn(C03)2

et al. (1982) described

from Tsumeb, Namibia,

minrecordite.

close to

for dolomitic honors. the first occurrence

associated with dioptase,

It shows evidence of cation ordering

of a nearly pure that they named

consistent with the

space group, and is also associated with an impure, magnesian well as a garden variety

of zincian dolomite

R3

variety, as

in several dioptase druses.

The

rarity of CaZn(C03)2 argues against a large stability field, but the existence of an ordered compound also argues against a crystallization as a metastable compound;

metastable

or configurationally Minrecordite "second oxidation

crystallizations

tions is presented

is questionable

it was formed in the

evidence of instability,

in Figure 15.

representation

The temperature

phases might be rather low.

amount of the natural material

A note on the asymmetry

but it is

is a compound with a maximum temperature

and highly conjectural

tion to two disordered obtain an adequate

at a low temperature;

The fact that it was not formed by reaction of the

that Zn-dolomite

A schematic

disordered

zone" (Garavelli et a L, , 1982), along with dioptase, a hy-

end members at 600-750°C

ence.

of high-entropy,

simple forms.

crystallized

drous copper silicate.

possible

are generally

of exist-

of its phase rela-

of the suggested decomposiIt would be desirable for experimental

to

study.

of the solvi

All of the solvi under consideration er amount of solid solution

is observed 65

are asymmetric,

such that the great-

in the limb representing

the larger

cation.

This phenomenon

is not restricted

tional solid solution it is generally

to carbonates,

the larger one with greater ease than vice versa. observed

in the reaction

mechanism

smith and Graf, 1960).

for in a substitu-

true that the smaller ion can replace This behavior

can also be

of, for example, MnC0

At low temperatures

with MgC0 (Gold3 3 when reaction is incomplete, it

was observed

that the MnC03 has taken some MgC0 into solid solution, leaving 3 pure MgC03 unreacted. This disequilibrium example of Mg-rich compositions reacting less completely than the Mn-rich ones is analogous to

essentially

the strain-energy

control of the equilibrium

solvi with asymmetrical

limbs.

TERNARY PHASE RELATIONS The combinations of CaC03 and MgG03 with FeC0 , MnC0 , CoC0 , and NiC0 3 3 3 3 are the only ternary systems to be experimentally investigated to date. It will become apparent that additional systems

(CaC03-MgC03-FeC03)

work is needed, and only one of the four

has even been considered by a second laboratory.

CaC03-MgC0 -FeC03 3 This system has been examined by Rosenberg et al. (1962).

Rosenberg

sels with samples in gold-foil carbonate

solution to promote

largely between

2-3 kbar.

(1960, 1967), and by Goldsmith

worked from 350° to 550°C in cold-seal pressure vesenvelopes

(containing a drop of O.OlM alkali

+ ~ 2% CO atmosphere 2 et al. (1962) produced isothermal sec-

reaction)

Goldsmith

open to a CO

tions at 600°, 650°, 700°, 750°, and 800°C, for the most part at 15 kbar, in an opposed-anvil

device.

Figures

16 and 17 contain the data of Rosenberg

(1967) and Figures 18 and 19 those of Goldsmith The two sets of data have no overlapping general configuration

of the diagrams

aries are significantly

different.

phase triangle containing

et al. (1962) at 600° and 650°C.

temperature

range, and although

is the same, the locations

the

of the bound-

At 550°C the ankerite corner of the three-

a calcite-rich,

a siderite-rich,

and an Fe-rich dolo-

determination is (mol %) FeC0 = 36, MgC0 3 3 9, CaC03 = 55. Goldsmith et al. at 600°C show FeC0 = 21, MgC0 = 27, CaC0 3 3 3 = 52. Rosenberg's data show rather small changes with temperature. At 450°, mite (ankerite) in Rosenberg's

the temperature where most of the work was done, the ankerite

corner is at

= 33, MgC03 = 13, and CaC03 = 55. Goldsmith et al. show a larger shift, in only 50° to 650° to FeC03 = 27, MgC03 = 21, CaC0 = 52. Both data sets 3 indicate that CaFe(C03)2 is not stable under these conditions, but Rosenberg

FeC03

shows significantly

more Fe-substitution

three-phase

The single~phase

field.

field of dolomite-ferroan

ite in both studies lies on the Ca-rich and natural ankerites

in ankerite along with a narrower

side of the CaMg(C03)2-CaFe(C03)2

also show excess CaC0 . 3 66

dolomite-ankerjoin,

Figure 16 (left). Subsolidus r lations in the system CaCOrMgCOrFeC03 at 450 C. D dolomite solid solution, C = calcite solid solution, S :::; siderite-magnesite solid solution. Open circles, one-phase; filled circles, two-phase; and triangles, three-phase assemblages. From Rosenberg (1967. Fig. 1).

5

=

Figure 17 (below). Three-phase triangles in the system CaCO)-MgC03-FeC03 for temperatures of 4000C, 4500, (shaded) and SOOoC shown superimposed. From Rosenberg (1967, Fig. 4).

/

,I

CaC03

It is unlikely attributed

that the sizeable differences

to the difference

calcite and magnesite

in pressure.

in the two studies can be

The sum of the molar volumes of

is very little larger than that of dolomite,

sure has very little effect on the stabilization molar volumes of ankerites

of dolomite.

to reconcile

and pres-

Although

the

are not known, it is also likely that pressure

would have rather little influence on solid-solubility difficult

I

the differences

relations.

It is also

on the basis of the temperatures

of

the studies. I see no basis for selecting as being closer to the truth. mental data base is clearly boundaries

Figures

Either or both can be in error.

inadequate,

in modern apparatus

temperatures,

one over the other of these investigations

is essential

to assure equilibrium.

20 and 21, the single-phase

composition

of dolomite,

composition

CaFe(C03)2'

CaMg(C03)2'

The experi-

and, at least some reversals

field originating

of phase At higher at the

extends as a thin sliver toward the

curving toward and enlarging 67

into the CaC03 apex

CoCO. Figure

The system CaCOrMgC03-FeC03 at 600oC,

18.

under a confining pressure of 15 khar. The solid solubilities shown in Figures 20-24 for the binary join CaCO)-MgC03 are taken from Goldsmith

and Heard (1961). from cles

Those in the join

CaCOrFeC03

the smoothed curve of Figure 15. Open cirindicate single phases, filled circles, two-

phase assemblages. Fig. 5).

From Goldsmith et al.

(1962,

CoMg(CO.),

MgCO.

FeCO.

CoCO. Figure under

19.

The system CaCOrMgCOrFeC03 at 650°C,

confining

in Figure

pressure

18.

of

15 kbar.

From Goldsmith e t al.

Symbols

as,

(1962, Fig. 4).

CoMg(CO.),

----MgCO.

after traversing Goldsmith temperature

approximately

that have Fe-contents

(1978) states

of the distance.

and suggested

The tendency

diagenetic

equivalent

of natural

ankerites

Matsumoto

metastability

(1978) and Matsumoto

from coal fields in Japan

to those of the ordered compounds

et al. (1962) at temperatures (p. 355), "It is noteworthy

68

low-

that some of them may

for compositional

has already been noted.

(1981) describe

by Goldsmith

ankerites,

compositions.

in several carbonates and Iijima

two-thirds

et al. (1962) noted the high Fe-contents

(sedimentary)

have metastable

FeCO.

of 600°C and above. that the carbonates

produced

Matsumoto contained

in

CoCO. Figure 20. The system CaC03-MgCOrFeC03at 700oC, under confining pressure of 15 kbar. Open circles indicate single phases, filled circles, twophase assemblages. From Goldsmith et a L, (1962 Fig. 3). I

----Mgco.

FeCO. CoCO,

Figure under

21.

The system CaCOrMgCOrFeC03 at 800°C,

a confining

as in Figure Fig. 1).

pressure

20.

of

15 kba r .

Symbols

From Goldsmith et a1.

(1962,

MgCO,

clastic

sediments

of chemical

FeCO,

of the coalfield

composition

as shown in Figure 11, compared with synthetic car-

bonates at higher temperature Rosenberg

(1967)."

Although

an4 pressure

given that the equilibrium

thermometer.

et al. (1962) and

in the investigations

of Gold-

(1967) make for real uncertainties,

Fe-content

is completely

makes the use of the Fe-content as a geologic

by Goldsmith

the differences

smith et a1. (1962) and Rosenberg

low temperatures,

regions studied provide much wider ranges

of ankerite,

unknown,

in ankerites

At higher

at rather

of metastability

and ferroan dolomites

(metamorphic) 69

particularly

the possibility

and

temperatures,

suspect where

equilibrium

might be more likely, it would be desirable

compositions

against an independent

to check ankerite

geothermometer.

CaC03-MgC03-MnC03 This system was investigated

at 500°, 600°, 700°, and 800°C by Goldsmith

and Graf (1960) in an opposed-anvil

device at 10 kbar.

trate the system at 500°, 600°, and 800°.

Figures 22-24 illus-

At 500° the presence of a solvus

between CaMn(C03)2 and MnC03 is shown in'Figure 22; at the higher temperatures solid-solubility between CaC03 and MnC03 is complete. The diagrams, particularly the size and shape of the three-phase

area, illustrate

greater solid

solubility

in general than in the system CaC03-MgC03-FeC03• The join CaMg (C03)2-CaMn(C03)2 is binary only at temperatures above approximately 700°C,

probably a result of the fact that Mn enters the Ca as well as the Mg position of the dolomite

structure,

producing

the case of Fe substitution Order-disorder

relations

a Ca-deficient

where ankerites

ordered compound,

unlike

show excess Ca.

in Fe- and Mn-containing

dolomites

In the system CaC03-MgC03-FeC03, at temperatures below approximately 675-700°C the ankerite corner of the three-phase triangle (Figs. 16-19) marks the limit of Fe solubility, At the higher temperatures

represented of Figures

by an ordered dolomite-type 20 and 21 the equilibrium

structure.

degree of

order of the ferroan dolomites within the single phase field originating the composition

of CaMg(C03)2 Figure 25 is a representation

is a function of temperature

of the limits of order as observed by the pres-

ence of several of the dolomite ordering reflections matter is discussed probable

in some detail by Goldsmith

at 700° and 800°C.

This

et al. (1962), and it is

that some degree of cation order extends beyond the temperature

compositional

region in which ordering reflections

powder diffraction

at 500°C and 800°C. in x-ray scattering

the compositional

factor produced by the substitution

Furthermore,

corner of the three-phase

of Mn for Mg makes

in x-ray powder diffraction

photographs

limits of detectable

more

order are even

Mn enters the Ca as well as the Mg-site, adding a

degree of freedom, which enhances disorder.

a better indication

region in which cation order

in the system CaC0 -MgC03-MnC0 can be detected 3 3 As discussed by Goldsmith and Graf (1960), the increase

so in this system the observed

more suspect.

are detectable with x-ray

structure)

detection of order reflections difficult,

and

techniques.

Figure 26 illustrates (the dolomite-type

at

and composition.

It may well be that the upper

triangle at the several temperatures

of the limits of cation order.

70

studied may be

Figure 22. Subsolidus relations in the system CaC03-MgC03-MnC03 at 500·C and 10 kbar total pressure. Filled circles indicate reaction products that are single phases; open circles, twophase; diamonds, three-phase assemblages. Tie lines are shown in the two-phase areas. The join CaMg(C0 ) 2-CaMn(CO ) 2 is indicated by a dotted 3 line. An additional inferred ,two-phase region is indicated by the dashed line. From Goldsmith and Graf (1960, Fig. 1).

Figure 23. Subsolidus relations in the system CaCOrMgCOrMnC03 at 600·C and 10 kbar total pressure. Symbols and tie lines as in Figure 22. The join CaMg(C03)2-CaMn(C03)2 is indicated by the dashed line. From Goldsmith and Graf (1960, Fig. 2).

Figure 24. Subsolidus relations in the system CaCOrMgCOrMnC03 at 600·C and 10 kbar total pressure. Symbols and tie lines as in Figure 22. The join CaMg(C03)2-CaMn(C03)2 is indicated by a dotted line. The position of the threephase triangle at 800°C was determined from only two runs, and the dashed line indicates somewhat greater uncertainty in its position than at lower temperatures. The inferred two-phase region along the base of the three-phase triangle is also indicated. From Goldsmith and Graf (1960, Fig. 4)

._. __

_.-..CoMnco,',

," ,

,'j

'0'

f ..:

,/'

71

700'C {IO·I}

visible

{OI'5}

and

{OO'3}

visible

from P 10 Q

{02"}

visible

from

from

P to 5

R to T

800'C

Figure

{IO.'}

visible

{OI'5}

ond{02.t}

from P'io

{OO'3}

visible

Q'

visible

from

R'

from

p'to

S'

T'

to

Compositional regions in the system CaC03-MgC03-FeC03 at (a) 70Qoe, and the several ferroan dolomite ordering reflections are visible in powder X-ray diffraction films. These two diagrams are enlarged central portions of Figures 22 and 24 to which letters have been added. In determining the positions of the letters, the presence or absence of the ordering reflections was noted in X-ray film of products from the runs shown in the figures. Inferred tie lines through the experimental points were used to locate the ferroan dolomite compositions on the upper and lower boundary curves of the single-phase region. From Goldsmith et a1. (1962, Fig. 8).

(b)

25.

BOOoe in which

\

eeo-c Figure 26. The compositional regions in the sys.tem CaC03-MgC03-MnC03 the dolomite-type structure can be observed in powder X-ray diffraction The grid of dashed lines gives compositions at intervals of 5 mol %. (1960, Fig. 7).

The systems CaC0 -MgC0 -CoC03 3 3

and petrology.

Their investigation

shed on the nature of divalent bility of the dolomite forming cations.

at 500° and 800°C, in which diagrams, are stippled. From Goldsmith and Graf

and CaC0 -MgC0 -NiC0 3 3 3

Neither of these systems plays any significant mineralogy

MnC03

was undertaken

cation substitutions

structure

Opposed-anvil

in the presence apparatus

deal of difficulty with decomposition

role in the real world of for the light

in dolomite and the sta-

of the various carbonate-

was used for both systems.

of NiC0 - and CoC0 -rich 3 3

A great

mixtures was

encountered. This system was investigated and 750°C at a confining pressure 0

tem at 600

0

and 750

of 15 kbar.

and are not reproduced

0

here.

The resemblance 72

650

0

,

Figure 27 illustrates

the 650 and 700°C diagrams are intermediate 0

;

at 600

,

700

0

,

the sys-

in appearance

to the system CaC0 -MgC0 -FeC0 3 3 3

CoCO,

CoMg(CO')2

MgCO, CoCO, Figure 27. The system CaC03-MgC03-CoC03 (a) at 600°C and (b) at 750°C under confining pressure of 15 kbar. Open circles indicate reaction prod':" ucts that are single phases; filled circles, two phases; triangles, three phases. The join CaMg(C03)2-CaCo(C03)2 is indicated by the light broken line. Fields that are inferred or not well delineated by experimental runs are indicated by broken lines. From Goldsmith and Northrop (1965, Figs. 1 and 4). CaMg(CO,l,

MgCO,

at temperatures

CoCO,

below 6750 is apparent.

At 600°C the field of cobaltoan

dolo-

mite extends approximately 22% toward the hypothetical CaCo(C0 )2; approxi3 mately 11 mol % Mg2+ is replaced by Co2+. At 750°C the upper limit of substitution of Co2+ in dolomite is approximately 26%. The ternary relations

in this system are more com-

plex than in those of CaC03 and MgC03 with MnC03, FeC03, and CoC0 . The mis3 cibility gap between magnesite and NiC0 produces an additional three-phase 3 triangle. The system was studied at 6000, 7000, and 7500, and Figures 28a and 28b are the results

at 600° and 750°C.

In Figure

28a the fields labelled

I-XI are as follows: Single-phase I

Calcite

fields: containing

Ni2+ and Mg2+ in solid solution 73

CoCO, I

a

600 C D

Figure 28. The system CaC03-MgC03-NiC03 (a) at 600°C and (b) at 750°C under confining pressure of 15 kbar. Open circles indicate reaction products that are single phases; filled circles, two phases; triangles, three phases. See text for discussion of the two three-phase triangles and for identification of fields. The join CaMg(C03)2 -CaNi (C03) 2 is indicated by the light broken line. Fields that are inferred or not well delineated by experimental runs are indicated by broken lines. From Goldsmith and Northrop (1965, Figs. 6 and 8).

b

7S0 C D

CoMg(CO,),

II III

Nickeloan dolomite 2+ 2+ Magnesite containing Ni and some Ca in solid solution IV Nickel carbonate containing Mg2+ and Ca2+ in solid solution Two-phase fields: V

VI VII VIII IX

Coexisting Coexisting Coexisting Coexisting Coexisting

Three-phase X

XI

calcitic solid solutions and nickeloan dolomites nickeloan dolomites and magnesite solid solutions magnesite solid solutions and NiC0 solid solutions calcitic solid solutions and NiC0 3 solid solutions 3 solutions nickeloan dolomites and NiC0 SOlld 3 triangles:

The compositions of the three coexisting phases are determined by the apices of the triangle: nickeloan dolomite, calcitic solid solution, and NiC03 solid solution The compositions of the three coexisting phases are determined by the apices of the triangle: nickeloan dolomite of a composition slightly different than that in X, magnesite solid solution, and NiC03 solution slightly different from that in X 74

of Ni2+ for Mg2+ in dolomite

The substitution between

600

0

dolomites

and 750

increases very little

in this range it is from 13 to 15 mol %.

0 ;

synthesized

in this and in the Co-containing

All of the

system appear to have

a high degree of cation order. SUGGESTIONS The disconcerting

differences

FOR FUTURE WORK

in the results of Rosenberg

smith et al. (1962) in the system CaC03-MgC03-FeC~3

(1967) and Gold-

require resolution;

equi-

librium reversals in modern apparatus are necessary to establish the extent of Fe2+ substitution in dolomite as a function of temperature, as well as the compositions three-phase

of the coexisting

triangle of Figures 16-19.

Of the ternary carbonate Mg has been independently

be desirable,

systems that have been investigated,

examined by more than one laboratory.

tion of the systems containing

verified

calcitic and sideritic phases making up the

on any carbonate

features be

system other than CaC0 -MgC0 . 3 3 with dolomite, because of its mineralogical

logical importance

would also

There has been rather little modern experimental

The preoccupation

sideration

Re-investiga-

Mn, Co, and Ni in modern apparatus

at least to the extent that their most important

or disproved.

only Ca-Fe-

is understandable.

of compositional

variants

work

and petro-

but lessons can be learned from the conof the dolomite-structure

type.

Kutna-

horite, CaMn(C03)2 is known as an ordered (R3) compound from several localities although it is a rather rare mineral. Although solid solutions along the join CaC03-MnC03 are readily synthesized (Goldsmith and Graf, 1957; de and Peters, 1981, and others), the ordered compound CaMn(C0 )2 has 3 not been produced, although the problem of detection of ordering reflections Capitani

in x-ray diffraction synthesis

of ordered structures,

low temperatures carbonates

powder patterns

particularly

has been considered

in Goldsmith,

has been mentioned.

of

those that are stable at rather

elsewhere

(see discussion

1959), and the onset of disorder

horite was observed at a temperature

The difficulty

as low as 450°C.

with respect to

in a natural kutna-

Thus, the crystalliza-

tion of ordered CaMn(C03)Z may at least in part be complicated by difficulties inherent in syntheses at low temperatures. Little consideration has been given to the possibility compounds;

of significant

a high degree of short-range

order would go undetected Ca-Mn carbonates

order in dolomite-like

order in the absence of long-range

by x-ray methods.

No structural

refinements

have been carried out, and other determinative

might be worthwhile. dolomites

short-range

Schindler and Ghose

with electron

paramagnetic

amounts of Mn, and although

(1970) have examined

resonance;

a preference 75

both contained

of

techniques two natural but trace

of Mn for the Mg site was apparent,

no work has been done on Mn-rich material.

Nuclear magnetic resonance tech-

niques on Ca-Mn compounds might also prove fruitful. The recent discovery of minrecordite (Zn-do1omite is a much more descriptive and palatable name) whets the appetite of the friends of the dolomite minerals.

Determination of the stability relations of CaZn(C03)2 would make at least several of these friends very happy, and might even be of interest to a larger segment of the mineralogical and geochemical population.

Synthe-

sis of Zn-dolomite has not been successful, but it has not been attempted at temperatures below 600°C, and lower temperature syntheses should be tried. If natural crystals are available, combined thermal (heating under elevated CO2 pressures) and single-crystal x-ray work could produce information on stability and/or order-disorder relations. It has been mentioned that additional thermodynamic information on carbonates is needed.

Data on solid solutions, as well as on ordered versus

disordered phases, is lacking.

For example, although it is known that mag-

nesian calcites and aragonite are metastable in sea-water with respect to essentially Mg-free calcite, the relations between the common Mg-calcites and aragonite are unknown.

At what magnesium concentration does calcite

become unstable with respect to aragonite?

The paucity of data is at least

in part due to calorimetric problems produced by the release of CO2 during Capobianco and Navrotsky (1983) have reported what

dissolution of carbonates.

appears to be the first successful work on carbonates.

The solvent was an

aqueous mixture of RCI and LiCl, and although it is not known whether or not the CO2 remains in solution (A. Navrotsky, pers. comm.), data are apparently reproducable. It is hoped that future work with this technique or perhaps with the use of a pressurized calorimeter will be carried out to produce the data badly needed to fully understand carbonate stability relations. Little attention has been given to thermal dissociation of carbonates. Good high temperature-high pressure data are not available for most of the carbonates, including many of the common minerals.

For example, calculation

of CO2 properties at high temperatures and pressures above 10 kbar relies on experimentally determined decarbonation reactions. These reactions involve magnesite and the accuracy of the calculation is strongly influenced by knowledge of the thermodynamic properties of magnesite. magnesite has not been accurately determined.

The Pc0 -T curve for 2 This and similar data on other

carbonates would provide better information for equation of state projections, adding to our knowledge of mantle geochemistry.

The inferred carbonates in

peridotite compositions are dolomite at lower pressures and magnesite at higher pressures.

ACKNOWLEDGMENTS

The author is indebted to R.C. Newton for the concept of Figure 12 and for helpful discussions, to David M. Jenkins for a variety of helpful acts, and to the National Science Foundation for Grant EAR 78-13675. 76

3

SOLID among

SOLUTIONS

METAMORPHIC

APPLICATIONS

and

SOL VI

CARBONATES

to GEOLOGIC

with

THERMOBAROMETRY

Eric J. Essene

INTRODUCTION There has been great interest ing carbonates experimental

in using the compositions

as geothermometers,

work of Goldsmith

due in large part to the extensive

and coworkers

(see Chapter Z for a discussion

of coexist-

on various

of the available

rhombohedral

carbonates).

petrologists

have directed most attention

carbonate

experimental

Of all the available

equilibria,

systems

data on

metamorphic

to the system CaC0 -CaMg(C0 )Z 3 3 in many marbles.

because of the common occurrence of calcite and dolomite The first applications powder diffraction

of calcite-dolomite

measurements

to estimate

despite the problem of retrograde and Heard, 1961).

exsolution

relied on X-ray

the MgC0

content of calcite

3 (Goldsmith,

1960; Goldsmith

With the advent of electron microprobe

followed a period of general disenchantment because of the recognition cite compositions applications

thermometry

of retrograde

to low temperatures.

of carbonate

thermometry

studies, there

with carbonate

resetting However,

thermometry

of many natural

the number of successful

has steadily mounted

ten years, and it is now clear that calcite-dolomite useful in contact and regional metamorphic

cal-

in the last

thermometry

rocks subjected

can be

to tempera-

tures between 400° and 650°C. In contrast

there have been few successful

able phase equilibria ates.

Despite

the availability

CaC03-MgC03-FeC03, estimates

in other carbonate

of metamorphic

with these systems

equilibria

temperatures

invariably

have been obtained petrologist themselves

or are these carbonate

and pressures?

These

below.

EXPERIMENTAL Of the many experimental

carbon-

in the systems

are the experiments

of equilibrium,

reset at low temperatures

questions will be discussed

of the avail-

few accurate

Has the metamorphic

these experiments,

representation

experiments

and CaC0 -srCo , 3 3 and/or pressures

(pace Barron, 1974).

been remiss in ignoring still an inadequate

of numerous

CaC03-MgC03-MnC03,

applications

systems to metamorphic

DATA BASE

determinations

of carbonate

only a few in the systems CaC03, CaC03-CaMg(C03)Z' 77

equilibria,

and CaC0 -FeC0 3 3

have

been tightly

reversed

(see Essene,

1982, for a discussion

of solvus

reversals).

The solvil in the systems CaC03-CaMg(C0 )2 and CaC0 -FeC0 3 3 3 as well as the calcite-aragonite transition in the system Caco have been 3 well reversed and appear to be accurately located equilibria for crustal metamorphites.

Many of the early experimenters

ment of both compositional

and structural

0

rapid at T > 500 C, although and Wyllie

(1982).

assumed

equilibrium

this has recently

that the attain-

was relatively

been questioned

Some of the early data were obtained

and related nonhydrostatic

experimental

reversed

data subsequently

obtained

cylinder

or rod bomb devices)

revisions

in the locations

crepancies

among different

devices,

although more carefully

in more hydrostatic

have generally

experiments

systems

necessitated

of the equilibria.

by Byrnes

in piston-anvil

However,

(piston-

only minor remaining

(e.g., see Anovitz

dis-

and Essene,

1983, on the system CaC03-MgC03-FeC03) suggest that the assumption of equilibrium in all experiments should be tested, ensuring careful reversals of appropriate

structural

by characterization

of run products with microprobe

electron

in addition

microscopy

Goldsmith provide

the experimental

Compositions

data relevant

by X-ray diffraction

1970), by wet-chemical

analysis

rocks have been obtained 1955; Goldsmith

1960; Sheppard

(Engel and Engel,

1970; Gittins,

1973; Puhan,

Michigan).

1976; Essene and coworkers

Powder x-ray by measuring

tional change

is relatively

in solid solution

(sub)microscopic intergrowths

allows

exsolution

1958; Carpenter,

analysis

(e.g., Hutcheon

at the University

estimation

or cell parameter

(Graf, 1961; Goldsmith

this technique

components

diffraction

a d-value

of carbonate

sensitive

et al., 1961; Bischoff

in calcite.

are formed in magnesian

of com-

to composiet al., 1983).

rapid, results are affected

(Goldsmith

et al.,

and Schwarcz,

1979), by atomic absorption

(Kretz, 1980) and by electron microprobe

and Moore,

positions

analysis

6

COMPOSITIONS

in metamorphic

1955; Goldsmith,

and Schwarcz,

techniques.

in Chapter

to this chapter.

(e.g., Graf and Goldsmith,

1955; Harker and Tuttle,

1967; Sheppard

5, and Carlson

OF CARBONATE

of carbonates

chemistry

and analytical

to standard x-ray diffraction

in Chapter 2, Speer in Chapter

DETERMINATION

While

states, as well as determining

by other

et al., 1961) and are plagued by Indeed if submicroscopic

calcites

then coherency

coherent

strain may

1 The term solvi will be used loosely in this paper for the miscibility gaps in carbonates of broadly similar struotures (e.g., rhombohedral) whether or not complete solid solution has been found at high enough temperatures. 78

shift the d-values

sufficiently

to yield incorrect

for even the calcite

lamellae.

absorption

analysis,

X-ray diffraction

analytical

technique

which disregards

possibility

of compositional

powder diffraction obtaining

analyst.

20 µm whenever

amperes

is followed.

Unfortu-

under the same conditions

and obtain poor results.

are best analyzed

Because

long wavelength

to 10-

is desired,

it is

less than 0.007 micro-

conditions.

due to differential poorly understood

heating factors.

The correction

by fixed stoichiometry

and other

the carbon

the

in the

to avoid erroneous

With due care and adequate analytical

standards,

totals

it is

(including

CO2) of 100 ± 2 wt %. analysis

by staining

the textural

examined

before preparing

and redistributes

Thallium

of coexisting

in thin section.

the polished

thin section.

Acid Phthalate,

RAP 79

Red S - dilute

Staining

the

flatness

in the calcite.

are common in magnesian

(1) zoning from magnesian

car-

This is often fa-

it alters the required

Mg and other minor elements

Two kinds of heterogeneities rocks:

relations

the bottom of the chip with an Alizarin

side should be avoided because

metamorphic

calculate

or fixed weight

to obtain acceptable

should be carefully

HCl mixture

by grain-size

program used for reducing

data should explicitly

Before beginning

cilitated

such as TAP or RAp2,

for Mg even under these

checks should be run on both

which is affected

microprobe

corrections.

of time

in the development

beam damage may vary from sample to sample

electron

stoichimometric

crystals

analyses

Volatilization

because

unknowns

possible

of recent advances

excellent

for the

at 10 or 12 kV instead of

(typically

spectrometer

to obtain

and unknown

atomic number

used

Volatilization

beam should be opened

If a 1-2 µm spot analysis

to use low currents

10-20 seconds).

restrictive

polished

is now a quantitative

practice

and the electron

possible.

it is now possible

bonates

tool for systems.

sample current on brass) and to count for short periods

of sensitive

usually

and ignores the

and some synthetic

of carbonates

carbonates

silicates

voltage,

important

(typically

standard

relations

is a bulk

under the electron beam can be a serious problem Carbonates

15 kV operating

especially

in natural

analysis

analyze

for the more refractory

unwary

textural

and atomic

practiced

For these and other reasons, x-pay

tool if proper analytical

some operators

of the carbonate

like wet chemical

as routinely

zoning.

compositions

microprobe

and accurate nately

compositions

should not be used as the primary analytical

carbonate

Electron

In addition,

inferred

calcites

cores to calcian

Rubidium

from

rims, and

Acid Phthalate.

(2) microscopic is presumably

exsolution regulated

of dolomite.

by diffusion

other cases the distribution to interpret.

analyze

closest

calcites.

cores of the larger to peak metamorphic

1982; Wada and Suzuki,

1983).

calcites

ter (1967), and Nesbitt

calcite

are present

the rims deficient Perkins

dolomite

stepping

annealed

calcite-dolomite excess dolomite 1967; Valley calcites

granules

into a granular

avoiding

1973; Puhan,

Exsolution

1976;

lamellae

assemblage

and Essene,

in dolomite-rich

The released

of interstitial

dolomite

texture resembling

1980).

buffered

by

1960; Carpenter,

It is safer to apply thermometry

marbles

throughout

(Goldsmith,

which

a high-temperature

but which was not necessarily

COMPOSITIONS The compositions

low temperatures

et al., 1979).

at the peak of metamorphism

with excess dolomite

within

exsolution

by step analysis

or else anomalously

(Puhan, 1976; Perkins

often forms external

may become

et al. (1982) all

If microscopic

1983).

in

(1960), Carpen-

have often been lost to the host and these areas

during

will be obtained

(1979) and Perkins

and Essene,

and Essene,

of dolomite

Goldsmith

(Hutcheon and Moore,

et al., 1982; Treiman

near grain boundaries must be avoided

exsolution

they must be reintegrated

in dolomite

the

(Sobol and Essene,

1982; Nesbitt

of this feature.

and

one should

grains to obtain

conditions

Microscopic

and Essene

but in

profiles

For zoned calcites

has long been known;

have shown photomicrographs lamellae

during exsolution,

and

is more erratic and difficult

1973; Brown et al., 1978; Bowman and Essene,

former magnesian

is often regular

(1983) show some excellent

in metamorphic

the magnesian

temperatures

of Mg/Ca

of magnesium

Wada and Suzuki

maps of magnesium

The zoning

which must have remained

to

saturated

metamorphism.

OF METAMORPHIC

of most metamorphic

CARBONATES

carbonates

are well represented

the tetrahedron

are usually coexisting

CaC03-MgC03-FeC03-MnC0 . Carbonates from marbles 3 close to the CaC03-CaMg(C03)2 binary, and one often finds calcite and dolomite. Carbonates in carbonated ultramafics

and some iron-formations no miscibility

lie along the MgCo -FeC0 binary, and because 3 3 gap is found along this join, only one carbonate phase is

found in this portion of carbonate many metamorphosed

manganese

1981) while others

locus of a solvus between

space.

The carbonates

deposits

some showing no compositional Peters,

composition

of

lie close to the CaC0 -MnC0 join, 3 3 gaps (Winter et al., 1981; de Capitani and

show a gap consistent rhodochrosite

with the experimental

and kutnahorite

(Goldsmith

and

DC

__

.....

.~- . ..

600l

-

500f

B

. - .........

"at

I

Am M

,... . ... .

I

I

I

• .....

.

••••

lA

4 ....

.•"--.'15

I

50

CoCO:l Figure MnC03' tures. tirama 1980).

....

,M ••

(Mn.Mg.r.lCo,

Mole %

1.

Carbonate compositions from metamorphosed manganese deposits on the binary CaCO) Microprobe analyses of natural (Ca,Mn)CO]-rlch carbonates with their estimated temperaAm = Serra do Navio (Scarpelli, 1970), M = Muretto Pass (Peters et a1. 1980), B = Buri(Peters et a1., 1977), A = Alagna (Peters et a l , , 1980), and S"" Scerscen (Peters et a l , , From de Capitani and Peters (1981). I

Graf, 1957; de Capitani carbonates

and Peters,

in the systems CaC03-MgC03-FeC03 or CaC03Each of these subsystems has been investigated experimentaland other workers

0

than 400 C, and applications Rhombohedral

group

(see Reeder,

dolomite

particularly

to metamorphic

at temperatures

greater

rocks await the petrologist.

carbonates

These may be separated

magnesite

Other metamorphic

are well-described

MgC0 -MnC03• 3 ly by Goldsmith

calcite

1981) (Fig. 1).

into two major

structural

(space group R3c) and the dolomite Chapter

1).

Within

subdivisions,

the

group (space group R3)

the calcite group are calcite

(CaC0 ), 3

(MgC0 ), siderite (FeC0 ) and rhodochrosite (MnC0 ), while 3 3 3 (CaMg(C03)2)' ankerite (CaFe(C03)2)3, and kutnohorite

(CaMn(C0 )2) are contained in the dolomite group4. Two parallel planes 3 through the carbonate tetrahedron, MgC03-FeC03-MnC03, and CaMg(C03)2CaFe(C03) 2-CaMn(C03) 2 , each traverse isostructural solutions and illustrate much of the solid solutions

shown by natural

samples in the carbon-

ate tetrahedron. While this may be taken as the idealized end-member formula for ankerite, the end-member itself is unstable with respect to calcite and siderite. 4 X-ray reflections indicating long-range order are not always observed in kutnohorite (Winter et al., 1981), but the name will be applied in this chapter to compositions near CaMn(C03)2 without requiring this information. The spelling kutnohorite is now preferred to kutnahorite by the I.M.A. and will be used as such in this chapter. 81

./:\

;

:", . .. ....'

\

.. ..

... 7.\ •.. ~ .....

..

'

.

.. _':: .. ::::: .. ::\~\~

.. ...!....OL:

.,

'.

:.,

,_._'.,_,

\

.. . .,: ...i:-'\ ~ .'

I_

..

V .....

:

\

: \f :':':I ... ".: .. : .:.:/::,j:.\ . II'

~::

••

Ie

:,:~.

Figure 2. Carbonate compositions (mol %) in the system MgC03-FeC03MnC03' Carbonates with more than 5 mol % CaC0 are excluded. 3 Data from Doelter (1912,1914,1917), Yosimura (1939) Pal ache et al. (1951), Muta (1957), Deer et al. (1962), Trdlicka (1964), Pisa (1966), Klein (1966,1974,1978), Machamer (1968), Butler (1969). Wenk & Maurizio (1970). Pearson (1974a, b), Beran (1975), Curtis e t; al. (1975), Floran & Papike (1975, 1978), Pinsent & Smith (1975). Klein & Fink (1976), Trdlicka & Hoffman (1976), Kashima & Motamura. (1977), Shimada (1977), Beran & Thalmann (1978), Pekkala & Puustinen (1978). Peters et al. (1977 ,1978), Sanford (1978), Ta1antsev & Sazonov (1979), Haase (1979), Hall (1980), Koba1ski (1980), Sivaprakash (1980), Klein & Gole (1981), Fukuoka (1981), Winter et al. (1981), Frost (1981,1982), Frost et al. (1982), Anovitz & Essene (1983b). I

Carbonates from literature

plotting

close to the plane MgC03-FeC0 -MnC03 are taken 3 The complete solution between MgC0 3 FeC03 and MnC03 is shown by these data. The lack

sources

(Fig. 2)5.

and FeC0 and between 3 of known carbonate compositions been interpreted

by Paulishin

between magnesite and Slivko

and rhodochrosite

(1967) to be a solvus gap.

has This

5 The carbonate compositions in Figures 2 and 3 are taken from hydrothermal and sedimentary as well as metamorphic occurrences because it is often difficult to distinguish these, especially in the older literature. Unlike magnesian calcites and calcian dolomites from sediments, these carbonates show no obvious tendency towards metastable compositions. Plots of the data from undoubted metamorphic occurrences show the same degrees of solid solution but with fewer points. 82

.\

Co Mn (C03)2 ,\

/:. .. ... . ...... . '

..

'._.

..... ... .

:

..

..

:

'

._::_.~:.!'•.:t:u_: ;M"

•• " \~.:: •• ;.:: :.~(./)::;

'.'

.:. :':~" ~,~ ~ ::••..

__

--"'- __

->

Figure 3_ Carbonate compositions (mol %) in the system CaMg(C03)2CaFe(C03)z-CaMn(C03)2. Carbonates with more than 60%or less than 40% CaC03 in

solid

solution

are

excluded

from this

diagram.

Data from Doe1ter (1912,1924,1917), Yosimura (1939), Palache et aL, (1951), Lee (1955), Fronde1 & Bauer (1955), Howie & Broadhurst (1958), Deer et aL, (1962), Trdlicka (1963), L1ambias (1964), Tsusue (1966), Klein (1966,1974,1978), Trdlicka & Sevgu (1968), Butler (1969), Wenk & Maurizio (1970), Pearson (1974a,b), Miyahisa e t e.l. (1975), Pf.nsent; & Smith (1975). Floran & Papike (1975,1978), Filho (1976), Klein s Fink (1976), Shibuya & Harada (1976), Trdlicka & Hoffman (1974,1976), Shimada (1977), Lesher (1978). Pekkala & Paustinen (1978), Sanford (1978), Ta1antsoy & Sazonov (1979). Haase (1979), Van Lamoen (1979), Hall (1980), Koba1ski (1980), Zak & Povondra (1981), Frost (1981,1982), Winter e t; aL, (1981), Irwin (1981), Dill (1983), Anovitz & Essene (1983).

is only permissive rhodochrosites metamorphic geochemical manganese

evidence

with mutual

rocks before

and one would like to find magnesites exsolutions

accepting

or petrological carbonate

deposits

as well as two-phase

this interpretation.

reasons why intermediates contain

calcium which

and

materials

in

There may be are unknown.

is strongly

Most

partitioned

into carbonates

over other phases

(Winter et al., 1981), while few mag-

nesite deposits

carry significant

manganese,

and so there may be few pro-

toliths which would yield magnesite-rhodochrosite 83

compositions.

R3

The

ternary composition

space CaMg(C03)2-CaFe(C03)2-CaMn(C03)2

shown in Figure 3 along with available

carbonate

this plane.

As has long been known,

and ankerite

does not extend completely

analyses

the solid-solution

exceed 2/3 CaFe(C03)2' phase equilibria of Rosenberg

et al. (1962) who found a three-phase content of ankerite

triangle

dolomite

limit appears

and manganoan

in preliminary

the CaFe(C0 )2 3 The

to extend well towards by tielines

kutnohorite. between manganoan

siderite which have been found by Rosenberg

unreversed

two-phase

limiting

of calcite + siderite.

by the coexistence

miscibility

Natural

between

(1960, 1963b) and of Goldsmith

If this is a real gap, it must be penetrated calcite

is

lying close to

to CaFe(C0 )2' and only a few 3 This is generally consistent with the

experimental

ankerite

analyses

(1960)

experiments

materials

on the join CaFe(C0 )2 - CaMn(C0 )2' 3 3 should be sought to confirm this interpreta-

tion. Extensive horite

solid solution occurs among dolomite,

(Fig. 3).

Kutnohorites,

manganoan

mites are more common than generally miscibility

gap between

1976) although ity gap here Stefanova

previous

data.

A more complete

group.

Orthorhombic

carbonates

cerussite

tianites")

of data

set

(Fig. 3) largely

The remaining

complexity

of potential

versus

history

in this long-range

of minerals

petrologic

in

significance6

(CaC03), strontianite

(PbC03), but few instances

in metamorphic

of a miscibil-

1980) based on a more limited collection

of a

et al., 1967; Minceva-

states of short-range

These may have solid solutions aragonite

dolo-

(Shibuya and Harada,

the presence

on the time(t) - temperature(T)

the dolomite

and kutno-

field and at present there is no reason

to suspect a solvus on this join. plane may be due to variable

amongst

suggested

1967; Kobalski,

ankerite

and manganoan

There is no evidence

1960; Minceva-Stefanova

fills the dolomite-kutnohorite

order depending

thought.

and kutnohorite

evaluations

(Rosenberg,

and Gorova,

of carbonate

dolomite

ankerites

rocks.

However,

(SrC0 ), witherite (BaC0 ), and 3 3 of such solid solutions are known

calcian

strontianites

with ~20% CaC03 in solid solution

("calciostron-

are found in sediments

often

6 Alstonite, BaCa(C03)2' is an example. With a structure related to the orthorhombic carbonates, it is presumably the high-pressure form of barytocalcite, and distributions of these two rare polymorphs should have barometric implications.

coexisting Harder,

with calcite

(e.g., Dietrich,

1960; Mitchell

and Pharr,

1964; Salter and West, 1965), and similar compositions

sought in metamorphosed

limestones.

Ore deposits

(Deer et al., 1962) and at L&ngban, carbonates.

Most attention

finer-grained (Sundius,

gangue may contain

1965) which

Sweden

studied with microprobe

1965) contain barium

to vein materials,

two or even three coexisting

should constrain techniques.

should be

in the north of England

(Sundius,

has been directed

1961;

equilibrium

carbonates

relations

In the presence

but the

if carefully

of magnesium,

ben-

stonite (Ba6Ca6Mg(C03)13) or norsethite (BaMg(C0 )2) may occur and these 3 phases may have solid solutions of Pb and Sr for Ba, Mn for Ca, and Fe for Mg (Sundius,

1943, 1965; Sundius

and Blix, 1964; Steyn and Watson,

1967). While

the solid-solutions

carbonates

may indirectly

gaps that petrologists conditions presence

because

are most likely

to gain insight

into metamorphic

is bounded

by the

of the second phase.

The carbonate to have a vertex

IN METAMORPHIC

CARBONATES

tetrahedron

of calcite

CaC03-MgC0 -FeC0 -MnC0 can be considered 3 3 3 (R3d), a base of magnesite-siderite-rhodo-

(R3c), and a plane half-way between vertex and base composed of

the R3 phases dolomite-ankerite-kutnohorite. there are complex from the

and orthorhombic

it is with miscibility

the extent of the solid solution

SOLVUS LIMITS

chrosite

found in rhombohedral

give some information,

R3

solvi in the tetrahedron

plane and both these from the

At metamorphic separating

R3c

base.

temperatures

the calcite vertex Experiments

on two

faces of the tetrahedron, 2) reveal

CaC03-MgC0 -FeC0 and CaC0 -MgC0 -MnC0 (see 3 3 3 3 3 some of these solvi7. We will examine information on

natural

two-phase

assemblages

learned

from them.

Chapter

on these two faces to see what might be

Solvi in the system CaC0 -MgC0 -FeC0 3 3 3 These are reflected two-phase

assemblages

dolomite,

calcite

in natural

carbonate

in the subsystem

- siderite,

dolomite

minerals

forming

equilibrium

CaCO-MgC0 -FeC0 , e.~., calcite 3 3 - magnesite, ferroan calcite -

7 So far there have been no adequate experiments or observations of natural two-phase assemblages constraining connections of these solvi which must occur inside the tetrahedron.

85

ankerite, natural

and ankerite carbonate

function

- magnesian

assemblages

of metamorphic

siderite.

Analyzing

should allow delineation Data from the biotite

zone of the green-

schist facies

(T

=

450°C) (Fig. 4), and the staurolite

zone of the amphi-

bolite

(T

=

5500C)(Fig.

facies

grade.

the coexisting of the solvi as a

5) are taken from Anovitz

(1983a).

While three-phase

reported,

few have been analyzed

Smith,

1975).

Rosenberg detailed

carbonate

assemblages

to the writer's

These have been interpreted

(1960, 1967) and of Goldsmith discrepancies

of the three-phase

(3) the partitioning ankerite - magnesian These inconsistencies

assemblages tioning

1983a).

offer potential

tative results

that of there are

ferroan

calcite

solid

- ankerite

the experimental

and

data

et al., and no simple explanation

or temperature

While

calcite-siderite-

in the magnesite-siderite

are also found in comparing

in terms of pressure

(Barron,

triangle

of Mg/Fe between siderite.

with those of Goldsmith

(Anovitz and Essene,

(Pinsent and

similar

et al. (1962) although

(2) the amount of CaC03 dissolved solutions;

can be made

knowledge

in a manner

been

in:

(1) the location ankerite;

of Rosenberg

and Essene

have occasionally

differences

the natural

as thermometers

1974), the discrepancies

two-phase

in the data ankerite-bearing

in terms of Ca/Mg/Fe must be resolved

before

partiquanti-

can be expected.

Solvi in the system CaC03-MgC03-MnC03 These are not reflected Despite

the experiments

in natural

revealing

two-phase

miscibility

carbonate

assemblages.

gaps in the subsystem

CaC0 -MgC0 -MnC0 (Goldsmith et al., 1962), the writer knows of no equili3 3 3 brium natural two-phase carbonates in this system, although Goldsmith (Chapter

2) and A. M. Gaines

lamellae

and local domains

Franklin-Sterling binary

between

of kutnohorite

an

tetrahedron

R3e

and an

R3

Ironically

kutnohorite difficult

to attain. - magnesian

phase,

to delineate

system to metamorphic

Until equilibrium rhodochrosite

presumably «400°C?)

from the

manganoan

because

the crest

that equilibrium calcite

Moreover,

experimental

is

- magnesian

pairs are found in nature,

or apply the available minerals.

calcite

this is the one

where no solvus has been experimental-

of the solvus is at such low temperatures difficult

1983) have found exsolution in manganoan

Hill area, New Jersey.

in the carbonate

ly located

(pers. comm.,

it is

solvi in this

until the problems

in the

Figure 4. Analyses of carbonates in the system CaCOrMgCOrFeC03from the biotite zone of greenschist facies. Carbonates with >5 mol % MDCD3 are excluded. Crossing tielines presumably mark disequilibrium

pairs

low temperatures.

Figure

5.

Analyses

or

carbonates

reset

at

very

From Anovitz & Essene (1983).

of

carbonates

in

the

system

CaC03-MgC03-FeC03 from the staurolite zone of the amphibolite facies. carbonates with >5 mol % MnC03excluded. From Anovitz & Eaaene (1983).

---87

CaC0 -MgC0 -FeC0 and CaC0 -MgC0 -MnC03 systems are resolvea, only the 3 3 3 3 3 solvus between calcite and dolomite seems presently suited for geothermometry. APPLICATIONS Calcite-dolomite

OF CALCITE-DOLOMITE

thermometry

ditions of metamorphism

THERMOMETRY

is a useful tool for evaluating

con-

because:

(I) calcite and dolomite in marbles binary in metamorphic terranes;

are close to the CaC0 -CaMg(C0 )2 3 3

(2) the thermometer is nearly independent of pressure requiring gross pressure estimates in order to apply a correction;

only

(3) unlike most silicate isograds, it is independent of fluid composition. Its disadvantages are largely related to rapid reequilibration which limits it to minimum estimates of temperature in many rocks. Calcite-dolomite once dissolved metamorphic

thermometry in calcite

temperatures

equilibrated which,

nesian cores of calcites analysis.

is obtained

with dolomite

in turn, is obtained

or reintegrating

While Hutcheon

and Moore

Rice (1977a) have all obtained

the mol % MgC0

by determining

heavily

3

at or near peak by analyzing

exsolved

areas by step

(1973), Bickle and Powell

algebraic

expressions

mag-

(1977) and

for the inferred

temperatures writer

in terms of % MgC03 in calcite (and other parameters), this prefers graphical use of the actual solvus (see Fig. 4, Chapter 2)

to obtain

temperatures,

Goldsmith

and Newton

expressions

combined

(1969), if necessary.

include erroneous

actual P-T-X range for which new experimental

with a small pressure

data.

they were fit and difficulty

because

and is still imperfectly

represent

the dolomite

applications of greater

limb is considerably

of calcite-dolomite than 6500C

calcite grains

in most geological

the

in accomodating is presently

less sensitive

thermometry

(but see Gittins, even exsolved

systems.

to

to temperature

The limits of 400-6500C

are the practical

meter should be applied

in most rocks.

seldom yield

1979) which may dolomite

within

On the other hand, below

solvus limit is not well constrained

and is steep and insensitive

beyond

located.

an upper limit of preserving

4000C the calcite

of dolomite

from

of the algebraic

of the functions

Use of the composition

temperature Careful

Disadvantages

extrapolation

impractical

temperatures

correction

experimentally

(Anovitz and Essene,

1983).

range over which this thermo-

Successful

applications

have been

.500 .530

/~2~

• Ceo Dol o

Ce

':>~':>

o

o Apsley

5Km

L____J

Bannockburn 0

Figure 6a Temperatures obtained from calcite-dolomite thermometry and contoured maximum.temperatures across a segment of Grenville terrane near Bancroft, Ontario. From Sobol (1973).

made to both regional discussed Regional

and contact metamorphic

metamorphic

rocks

Calcite-dolomite

thermometry

rocks by several workers, al. (1973), Hutcheon (1977), Nesbitt

temperatures

temperature isotherms

and Moore

to regional

Sobol and Essene

(1973), Puhan (1980), Nesbitt

metamorphic

(1973), Hatcher

et

(1976), Bickle and Powell and Essene

(1982), and Per-

Sobol and Essene were able to contour approximate

across a region

(Fig. 6).

temperatures

has been applied

including

(1979), Kretz

kins et al. (1982).

Ontario

rocks which will be briefly

below.

While

in the Grenville

it is difficult

terrane near Bancroft,

to determine

whether

or not these

have been reset by perhaps as much as 500C in the higher rocks,

the results

are parallel

(1977) evaluated

agree well with Al Si0 isograds and the 2 5 isograds. Bickle and Powell

with other mapped

the experiments

the influence

of small amounts

thermometry.

They applied

of Goldsmith of FeC0

et al. (1962) to correct

solid solution

3 their model to calcite

89

for

on calcite-dolomite

- ferroan dolomite

pairs

from the Tauern window, Austria, to 490°C for greenschist and Nesbitt

obtaining

reasonable

facies marbles.

and Essene

(1982) in marbles

isograd;

from 425° at the biotite

methods

with dolomite

exsolution

including

(1973), Puhan

al. (1979), and Valley

and Essene

erroneous

in applications

temperatures

al metamorphites.

(1960), Sheppard

erratic results thermometry

of ~5000C

in the N. W. Adirondacks,

where subsequently ranging

Sheppard

and Schwarcz

(1977), Brown et and/or

to region-

relied on X-ray diffraction

temperatures

morphic

Earlier

Graf and Goldsmith

of carbonate

low temperatures

occasionally

isograds.

(1976), Garde

(1980), obtained

The early workers

and as a result obtained

on pelitic

in their carbonates.

(1955), Engel and Engel (1958), Goldsmith and Moore

gave temper-

gave much lower temperatures

On the other hand, various workers,

(1970), Hutcheon

to amphi-

Tennessee,

et al. (1973) using X-ray diffraction

to obtain calcite compositions,

consistent

of 410° (1979)

zone to 540°C at the staurolite

these agreed well with information

work in the same area by Hatcher

of Nesbitt

from the greenschist

bolite facies Murphy Marble belt east of Ducktown, atures ranging

temperatures

Thermometry

methods,

for Grenville

marbles

Brown et al. (1979) obtained

up to 625°C with microprobe

studies.

and Schwarcz calcites

chemically calcites

(1970) showed that X-ray determinations of metasignificantly lower T (by 70-180oC) than bulk 0 calcites. They obtained 645-705 C for bulk analyzed

yielded

analyzed

from the Haliburton-Bancroft-Denbigh-Renfrew

Canada in upper amphibolite isograds.

facies rocks,

area of Ontario,

in good agreement

with silicate

They proceeded

with X-ray determinations of calcites from and obtained scattered temperatures of 175-2800C zone, 230-4350C for the biotite zone, and 290-5750C for

Vermont metamorphics for the chlorite

the staurolite-kyanite reasonable

approximation

values are probably determinations

zone.

The upper temperatures

of peak metamorphic

explained

occasionally

by dolomite

appear

it would be useful

to reexamine

Hutcheon

(1973) applied

and Moore

greenschist-amphibolite Ontario.

ranging

to be successful

this thermometer

exsolved

astride

This isograd

this region and the scattered

in the Vermont

to marbles

carbonate

and obtained

rocks,

methods. from the

terrane near Marble

calcites by microprobe,

content for geothermometry,

isograd.

and the lower While X-ray

these samples by microprobe

3 from 400° to 600°C for marbles

tremolite-calcite

conditions,

exsolution.

facies in the Grenville

They reintegrated

the highest MgC0

of each zone are a

Lake,

selected temperatures

the dolornite-quartz-

should be found at 500-550oC temperatures

are therefore

in

questionable.

Hutcheon and Moore

(1973) also report carbonate

tempera-

tures of ~8000C for marbles

10 km east near the kyanite-sillimanite isograd, too high by perhaps 150-2000C. It would appear that their reintegration

technique

included too many dolomite-rich

primary dolomite

grains.

metry to marbles

in a cordierite-granulite

Damara,

S. W. Africa.

rocks.

Puhan (1976) applied calcite-dolomite

He carefully

calcites and obtained

reintegrated

cores of exsolved

620°C which appears to be some 100°C low for these

Garde (1977) criticized

the general application

after finding calcite compositions

in apparent

dissonance

with silicate

varying

erratically

isograds in greenschist

which yield T of 300-5000C,

3

of carbonate and

facies

His data show calcites with 2-5 mol % MgC0

from W. Greenland.

and 0.2-1.8 mol % FeC0 effect of Fe.

thermo-

facies terrane of Central

thermometry

marbles

areas and/or

3 for the

uncorrected

However by selecting

the maximum MgC0 contents in the 3 the Bickle and Powell correction for FeC0 , his 3 data can be interpreted to show a gradient across the section from 475° calcites and applying

to 5250C in reasonable

agreement with silicate isograds and the regional

grade of metamorphism. amphibolite obtained

Valley and Essene (1980) investigated

to granulite

facies marbles

T < 500°C, suggesting

subsequently

that these high-grade

reset during cooling.

is best suited to greenschist

in the Adirondacks

rocks have been

It appears that carbonate

to middle amphibolite

rocks, and that upper amphibolite

carbonates

do not record peak metamorphic

Contact metamorphic

(1977a, b), Bowman

thermometry

temperatures.

including

Carpenter

(1967), Suzuki (1977), Rice

(1978), Burton and Jackson (1979), Brown

(1980), Bowman

(1982), and Wada and Suzuki (1983) have applied carbonate to contact metamorphic

intruded by high-level X-ray observations

plutons.

aureoles Carpenter

showed extensive

in which carbonate

to infer minimum

temperatures

units are

(1967) noted that optical and

resetting of calcite compositions

and relied on textural evidence of original one-carbonate of 760°C at Crestmore,

is in good agreement with the occurrence

assemblages

California.

This

of other high-temperature

minerals found at Crestmore. Rice (1977a, b) obtained temperatures 0 450-475 C at the tremolite-calcite-quartz-diopside isograd in the Marysville

for

to granulite

rocks

A number of workers,

and Essene

thermometry

facies terranes

regional metamorphic generally

upper

and uniformly

aureole, Montana.

Bowman

(1978) and Bowman and Essene

of

(1980)

Figure tained

500

o

o

100m '-----'

Undifferentiated .ediments

contoured Montana

the temperatures

obtaining

silicate

these appear

Burton and Jackson

syenite

intrusion

the authors accept

the Wind Mountain Wada and Suzuki

carbonates

carbonates

from a contact metamorphic temperatures

gradually

increasing

(1967) data show

by microprobe thermometry

and methods.

to carbonates

in the Kasuga area, Japan.

to 570-6800C at le% MgC03 a < 0.1

Specimen

0

~(~p)

0

~(A)

_£/~

_£(A)

1.2

367.4(1)

4.979(1)

17.111(6)

3.437(2)

6.2

361.67(6)

4.9646(3)

16.944(3)

3.4130(8)

6.5

361.6(1)

4.9635(6)

16.950(4)

3.415(1)

9.1

358.6(1)

4.949(1)

16.904(6)

3.416(2)

Trlpneustes esculentis (test)

10.5

357.4(1)

4.9460(8)

16.869(4)

3.411(1)

Lytechinus varlegatus (test, specimen 2)

11.2

356. 6( 9)

4.9425(9)

16.854(6)

3.410(2)

Lytechinus variegatus (teeth, specimen 1)

11.3

357.4(1)

4.9461(6)

16.871(4)

3.411(1)

Diadema

11.8

355.74 (8)

4.9398(6)

16.834(4)

3.408(1)

12.0

356.2(1)

4.9421(8)

16.840(5)

3.408(2)

12.2

355.71(9)

4.9392(6)

16.837(3)

3.409(1)

Barnacle Diadema (spine,

antillarum specimen

1)

Lytechinus variegatus (spine, Diadema

(spine

(test,

1)

specimen ant illarum

tip,

specimen 1)

antil1arum

specimen 1)

Lytechinus (test,

varlegatus

specimen

1)

Diadema antl11arum (test,

specimen

2)

Homotrema rubrim

12.4

355.99(6)

4.9397(4)

16. 846( 2)

3.4101(7)

Echinometra

lucanter

13.3

353.94(8)

4. 9323( 5)

16.800(4)

3.406(1)

lucanter

13.6

353.9(1)

4.9312(8)

16.803(5)

3.408(2)

(teeth) Echinometra

(test) Lithothamnium ~.

15.5

354.1 (1)

4.9317(8)

16.809(4)

3.408(1)

Amp hiroa ~~.

19.5

349.3(1)

4.9139(8)

16.704(5)

3.399(2)

samples revealed surrounding asymmetric

domains

material.

(~10 µm) up to 10 mol % higher in MgC0

These domains apparently

peaks in the Lithothamnium

contributed

sp. x-ray pattern

than the 3 to the strongly

(cf. Milliman

et al.,

1971) . Previous

microprobe

that small domains within an otherwise al., 1969).

studies of coralline

homogeneous

skeleton

either as high-Mg domains

genic specimen

in this study exhibited

high-Mg

The existence

ever, cannot be excluded.

O'Neill

(1981) suggested

spinulosis,

geneity in the 200

A

sp., no other bio-

domains on a scale that could

analysis.

Echinaster

in the

(Schmalz, 1965; Weber and Kauf-

sp. and Lithothamnium

be detected by microprobe

phase varies on the crystallite

Mg not apparent

(Goldsmith et al., 1955; Milliman

et al., 1971), or as a separate brucite phase Aside from the Amphiroa

have shown may exist

(Moberly, 1968, 1970; Schroeder et

Many algal skeletons contain significant

x-ray patterns,

man,1965).

algae and echinoids

(20-40 µm) of lower or higher Mg concentration

of smaller domains, howthat Mg in the carbonate

size scale (1300-3600 K) in the echinoid

although Blake and Peacor

(1981) demonstrated

size range in the crinoid Neocrinous 110

blakei.

homo-

Moberly

(1970) has

suggested

that brucite domains in coralline

algae are submicron

in

size. Inhomogeneities

on a larger scale were noted during atomic absorption

analyses of the biogenic

Spines of Lytechinus

skeletons.

variegatus have

6.5 mol % MgC03, whereas teeth have 11.3 mol % MgC03• Comparable variations were found in Diadema antiZZarum: tests are 11.8 mol % MgC03, spines are 6.2 mol % MgC03, and spine tips are 9.1 mol % MgC03. Variatioµs in Mg content between tests and spines and within spines were noted for many sea urchins analyzed variations

by Chave

(1954b) and Weber

in the compositions

(1969).

(Table 3).

Chave

variations

(1954b) found compositional

test of one specimen of Lytechinus Substitution could distort

two specimens of Lytechinus

of tests between

variegatus and Diadema antiZZarum

We also observed minor

From 12 diffraction

patterns,

of 3.4 mol % MgC03 within the

variegatus.

of trace amounts of large ions into the biogenic

or increase the size of unit cells.

Sr were all analyzed by electron microprobe, found in quantities

exceeding

volume from substituting made by comparing

but only Sr, Na, and S were

a few hundred ppm.

An estimate of the excess

Sr for Ca in the biogenic

strontianite

50% of the volume discrepancy

and aragonite between

skeletons

K, Ba, Mn, Fe, Na, Sand

magnesia~ calcites was

unit cell volumes.

the biogenic

Only 5 to

phases and the calculated

volume from the results on the synthetic phases could be due to Sr substitution.

This calculated

volume discrepancy

ments caused only local distortion ously in silicates

not be excluded

of the crystal structure as found previ-

above 200 ppm were noticed, but the possibility

by organics or seawater, even after extensive (cf. Blake and Peacor, 1981).

within the crystal volume because

if the trace ele-

(Iiyama, 1974).

Na and S concentrations of contamination

would decrease

structure,

no correlation

they apparently

cleaning,

can-

If these elements are present have little effect on the excess

exists between Na

+

S

+

Sr and the observed

volume deviations. Infrared of O-H bonding

spectroscopic in biogenic

studies discussed magnesian

calcites.

later demonstrate

the existence

As yet, it has not been deter-

mined if the O-H bonding arises from hydroxyl ions or H20 molecules. The presence of hydrated Mg2+ ions in the lattice could increase unit cell volumes. On the other hand, substitution

of OH- for CO~- would probably

cell volume.

The intensity of O-H absorption

the magnesium

content of biogenic

deviations

for biogenic

phases.

phases, however,

nesium concentration.

111

decrease

unit

in the infrared increases with

No trend in unit cell volume is apparent as a function of mag-

The crystal

chemical

and structurally

studies show that biogenic

inhomogeneous.

in skeletal microarchitecture and solubility

phases ~re chemically

These phases also exhibit (Walter, 1983).

large differences

These factors can affect rate

studies and should be kept in mind when interpreting

dissolu-

1

tion data on skeletal magnesian

calcites and solid solution models derived

from these data. SOLUBILITIES Knowledge

of the solubilities

and aragonite, calcites,

AND SOLID SOLUTION BEHAVIOR

and members

is of utmost

these minerals

of the calcium carbonate

in natural

of the CaC03-MgC03 solid solution, the magnesian to an understanding of the behavior of

systems.

The precipitation

Mackenzie,

in numerous

solubility studies

1974; Berner, 1975b;MBller

for magnesian

and Parekh, 1975; Garrels, Wollast,

investigations

1979a; Wollast and

in aqueous

involved dissolution

solutions

' synthetic

including

Ca-Mg-HC0

C02 Besides dissolution, other experimental flow-through

apparatus,

A compilation

of solubility

solubility

determinations

product,

l't g-ca c~ e for solubility

below).

future reference.

ous carbonate

including

and analytical

system.

differences

techniques.

authors

studies.

investigators

is expressed as the I-x x aCa2+ aMg2+ aC032- , has been debated (see

in materials,

Another

possible,

and Busenberg

to problems

solutions,

and

source of divergency

between

is the use of 'different models for the aque-

In Figure II and the remainder

whenever

aqueous model of Plummer In addition

have included use of a

Curves have been drawn through three sets of data for

the results of different

have recalculated,

and seawater.

The large amount of scatter in Figure II undoubtedly

reflects many factors experimental

mag-

water in equili-

by numerous

solubility

IAPM

the use of this expression

discussion

distilled

studies and precipitation

In this diagram,

The major-

of biogenic

solutions,

3 approaches

Ca-Mg exchange

is given in Figure II. stoichiometric

cal-

(e.g., Chave et al., 1962;

data, 1977; Walter and Hanor,

brium with a fixed P

although

product

1980; Walter and Morse, 1983b; Mucci and Morse, 1983a).

ity of experimental nesian calcites

7,

1966; Land, 1967; Schmalz, 1967; Weyl, 1967; Plummer and

and Konig, unpublished Marijns,

behav-

(see Morse, Chapter

at Earth surface T and P are reasonably

Values of the equilibrium

cites have been reported

and dissolution

is fairly well documented

this volume) and their solubilities

Chave and Schmalz,

calcite

importance

ior of calcite and aragonite

well known.

dimorphs,

resulting

112

the experimental

of this section, we data using the recent

(1982). from differences

in experimental

7.2



7.4



., ] 76 o v

o

° -' I

8.2

11

8.4 8.6

o

11

o

10 20 MOLE %MgC03

30

Figure 11. Summarydiagram of experimental determination of magnesian calcite solubilities. Solubilities are all expressed in terms of -log IAPMg-calcite' The three curves are drawn through the data of Plummer and Mackenzie (1974), the general trend of the dissolution experiment results, and the precipitation experiment results of Mucci and Morse (1983a).

o

!_!..,

Chave.!.!. distilled

~

Chave and

and

A variety

(20

of

.. 1 atm,

2 1966.

Schmalz,

AzIIphiroa

biogenic

lII.te~1als;

+

25·C.

Lytech1nus

eoj e % MgCO);

(11

mole

% MgC03)

sea ....a r e r , Peo

..

Berner,

1975.

Calculation

eqUivalent

in

cent .. in1ng

about

seeded

ahowing

solubility 8.5

of

that

is

c.. lcite

% )'!geo) , using

mole

precipitation

arasonite

to a magnelian

magnesian

data

frolll

calcite.

2

10-3.5

IiiiiJ

1962.

ve t er . PC0

atm,

Sctmalz:,

25·C.

..

1967.

Pinna

(2.5

spines

(10 mole

Hgeo);

distilled

water,

1967.

As above;

% Hgeo»,

% HgC03),

mole and

Amphiroa

(15.4

• 10-0.01

PC0

Lytech1nus ioole

%



25"C.

.ttll,

HOller

and Parekh,

ca-Hg

ion

Garrels,

exchange Wollast,

Sphaerech1nus

1975.

From exper1.lllents

behavior

on pure

and Konig,

unpub.

granular

on the

calcite

surfaces

data,

(11 mole % HgCO);

is

.

1977. distilled

2 ~

Schmalz,

distilled

water,

Peo

-

water,

pure

and

spiked

with

ea++

Hg++,

or

P

f)

10-)·5

atm,

Weyl,

1967.

25·C.

()

solution,

1967.

seawater,

IJ

Land,

of

biogenic

of

10-0.01

A variety

biogenic atll'l,

of

distilled

1977).

water,

o

biogenic

A variety Pe0

2

ea++-

Walter

water,

and Hanor,

1979a.

\l

materu1s;

ea r e r La Ls ;

of

6

25"C. by Thorstenson

biogenic

pure

and

Tripneustes

(18 mole spiked

with

% Mf;CO)

(12 mole

% MeCO)

i

distilled

phosphate,

P

C02

- 1 atm,

25·c.

)O·C.

25·C.

water,

and Plul1llller,

seawater,

materials; apparatus,

PCO • 10-0 .. 01 atlll, 2 PlutQlller and Mackenzie, 19710 (revised

distilled •

c02

1967.

-

materials;

and Neogoniol1thon

!low-through

A variety

P

biogenic

)O"e.

A variety

Weyl,1967.

Land,

of

apparatus,

1-1& ++-Heo;

CI

2S·C.

A variety

flov-through

- 1 atm, e02

2

Wollast

spiked

w1th

Walter

and Horae,

water

1 a till , 25"c.

1980.

(5 mole % l1gCO);

tigCO)

materials;

and Harijns.

spines

Sphaerechinus

distilled

CaC12, MeCl2 and NaMeo), 198)b,

and Goniol1thon spiked

with

elypeaster

and

Pe0

solutions

.. 1 atm, %

distilled

HgC1 , P .. 10-2.5 2 e02

25·C.

'Y

l'lIc:ci and Morse, from free with

113

drift

variable

198)(1. experiments;

Ca:l1g ratio,

Incr gan Ic precipitates artificial 25·C.

25·C.

2 (12 mole

(18 ec Le % Hgeo);

Cae12 and

granularis

water

seawater

e tm ,

procedures,

theoretical

matter of debate. difficulties

interpretation

Dissolution

because

of the experimental

experiments,

the incongruent

data has been a

in particular,

are fraught with

nature of the reaction prevents

attain-

ment of equilibrium. In an attempt to clarify the interpretation Thorstenson

and Plummer

(1977) reviewed

of the experimental

the theoretical

criteria

brium between a binary solid solution and an aqueous phase. relationships

derived by Thorstenson

application

to the experimental

results

1978).

Recent experimental

of the equilibrium

1980; Walter,

theoretical

considerations

nesian calcites. thermodynamic

In addition,

properties

from calorimetric

1983; Walter and Morse, 1983b).

Solubilities

of magnesian

several approaches

of experimen-

used to obtain fundamental

calcites are discussed

and new data

calcites

have attempted to determine

As shown in Figure II, many investigators

the solubilities

of magnesian

studies in terms of compositional

(Chave et al., 1962; Land, 1967; Plummer and Mackenzie, dissolution

deter-

are presented.

Dissolution experiments. three most comprehensive

work on solubility

of various solution models to the mag-

of the magnesian

measurements

criteria derived by

involved in interpretation

tal data, and the possible application

1978; work, de-

calcites, has led to contradictory

In this section, we review the experimental mination,

(1974) has been

(Berner, 1978b; Garrels and Wollast,

and Plummer to the magnesian

(Wollast and Marijns,

the

and Plummer appear to be valid, their

and Plummer,

signed to test the applicability Thorstenson

for equili-

Although

data of Plummer and Mackenzie

a source of much argumentation Lafon, 1978; Thorstenson

data,

of biogenic magnesian

calcites

calcites.

The

range studied

1974) all involved

in distilled water saturated with

0.97 atm CO2• Thus, the results of these three studies may be compared directly. We will not consider here the numerous other studies in which values of the equilibrium

solubility product

cites were obtained generally under different

for some single magnesian experimental

conditions

cal(e.g.,

in seawater). Chave et al , (1962) limited their measurements the dissolution

reaction and estimated

the pH at infinite time by extrapola-

tion of plots of pH versus the reciprocal plots were chosen empirically nite time is minerals

approached

calcites, however,

the dissolution

of the square root of time.

reactions

These

they usually yield linear plots as infi-

when calcite, aragonite

are used (Garrels et al., 1960).

case of magnesian because

because

to changes of pH during

or other simple carbonate

Chave et al. claimed that in the

only relative solubilities

are irreversible 114

and therefore

are obtained the material

can have no true equilibrium Solubilities

solubility.

of annealed

skeletal carbonates were estimated

by Land

(1967) by allowing the system to reach a steady state pH and by determining dissolved Ca2+ and Mg2+ concentrations. Dissolution of the magnesian calcites in Land's experiments

appeared

to be congruent

in the sense that the steady

state solution had a molar Ca/Mg ratio similar to that of the solid. At the suggestion the behavior

of Owen Bricker,

of magnesian

calcites during dissolution

of Mg2+ and Ca2+ were monitored

pH and concentrations ments.

They were able to distinguish

reaction

(Fig. 12).

Plummer and Mackenzie

(1974) studied

in greater detail. throughout

up to three stages in the dissolution

The first stage was congruent dissolution,

of the most reactive phase present when the biogenic magnesian were not initially homogeneous

phases.

presumably calcite samples

The congruent portion of dissolution

was followed by a stage in which dissolution

continued but a low magnesium

calcite was precipitated

Solution chemistry

on grain surfaces.

more than a twofold supersaturation 2.

Finally

The

the experi-

indicated

with respect to pure calcite during stage

these authors attributed

stage 3 to the precipitation

of a Ca-rich

phase from the bulk solution. It is interesting dissolution dissolved

to note that the congruency

of the initial step of

is still maintained when the aqueous solution initially contains

Ca2+ or Mg2+ (Fig. 13; Wollast

The distinction

between

tant in the extrapolation

congruent

and Marijns,

and incongruent

of experimental

data.

1980). dissolution

is impor-

Plummer and Mackenzie

extrap-

olated their pH data to infinite time on plots of pH versus the reciprocal

of

the square root of time, as did Chave et al., but used only the congruent portions of the reactions.

They assumed

that the extrapolated

pH values repre-

sented equilibrium

with the initial magnesian calcite composition. These pHs 2 2 were then used, with P and the Ca + and Mg + concentrations, to calculate C02 solubilities from the following expression: (I-x) Ca

2+

+ xMg

2+

23

+ C0

lAPMg-calcite The final pH values obtained during the three studies discussed here are presented ates.

in Figure 14 as a function of the magnesium

Except for the low-Mg calcites,

The lowest pH values, corresponding

to the lowest solubilities,

tically obtained by Land and the highest Plummer

and Mackenzie

assumed

content of the carbon-

there is a large scatter of the data.

values by Plummer

that the lower values reported 115

were systema-

and Mackenzie. earlier provide

,. g

Stag. 2

'1

it I

12

~ _ 10 ~ I ~ ~ I

I-......_ ~ - · .___

E

Q)

0..6

0.7

400

0..01 20.0.

0..8 0.55 ( 10.3/ T , OK)

temperature

19 (above). Pressure-temperature of melting of (a) BaCOr and (b) Data from Baker (1963, 1964).

Figure 20 (right).

en en

... a.

,, 0..5

0.65

diagram in the SrC03-C02 sys-

Pressure-temperature

200

40.0.

600

800

temperature diagram. for Peo,'" lotm PbCo,~

the PbC03-C02 system. Data for phase relations at elevated pressures from Grisafe and White (1964); data for low-pressure sequences from Ball and Casson (1975) and Yamaguchi e t al. (1980).

(OC)

t

t

t

PbCO,PbO3*

PbCO, ~

PbCO,2PbO ~

PbCO,'2PbO

200·

0

of cerussite pressure.

over the range of 200-900 C

Their work is presented

of oxycarbonates

Binary

carbonates

has been designed hedral

and 15 to 1400 bars carbon dioxide

in Figure

20.

The reaction

sequences

(1975) and Yamaguchi

by the

et al. (1980).

systems

Phase equilibria rhombic

co,\ 310.

at 1 atm and less than 1 atm have been supplemented

studies of Ball and Casson

liquid

:::J

0.1

Q)

-2 I-

Figure region tems.

o

SrO •CO2

e

5. _I

~

N

-

o

;; :;!

p..

""o z

II

u

205

All studies concur that the transformation vated nucleation

and growth.

one is immediately

Looking

beyond

struck by the disparity

also in the form and degree of inferred

proceeds

by thermally

acti-

this point of agreement,

not only in absolute

time, temperature,

however,

rates, but

and pressure

depen-

dences. Because activation

of the exponential

energy

(G*), differences

lead to immense differences from laboratory

of absolute

0

(say 400 C)

Thus, even if two investigations

to geologic

measured

identical

energies measured

measured

is nearly

magnitude,

it is understatement

data to natural

occurrences

form derived

dependence

formation

and because

(1965) is excepted1, and Mehl

to describe

energy) and absolute

study of Kunzler

that the major determinants

rates at

absolute

rates

but vary over more than two orders of of these rate

and Goodell

the observed

nucleation.

In all cases, samples with the highest nucleation

transformed

the rate of the transformation

time dependence

variability

most rapidly.

of

in temperature in the careful-

That work demonstrated energy for the bulk trans-

to the opportunity

preferential

of

(1939, 1940,

rates were treated (1970).

of rate and activation

were factors which contributed

rate equations

(1939) and Avrami

The reasons for the remaining

(activation

ly controlled

(say 2000C).

the total range of acti-

to assert that the application

by Johnson

1941) appear to be well-suited the transformation.

temperatures

is made

is problematical.

If the study of Davis and Adams the general

small in Table 1

extrapolation

extrapolated

Because

300 kJ/mole,

near 400°C are not identical

and

rates at 400°C, a differ-

energy of only 30 kJ/mole yields

200°C that differ by an order of magnitude. vation

rate on temperature

in G* which may appear

in rates when the required

temperatures

ence in activation

dependence

for heterogeneous

densities

of sites for

The pressure

is not well characterized,

dependence

but appears

of

to be

small except very near equilibrium.

1 The work of Davis and Adams (1965) contrasts sharply with other studies; the rate functions they employ are purely empirical, so the parameters involvec have no direct physical significance. Their study, however, merits special attention because it is also the only one in which pressures in excess of 3000 bars were utilized. The most unusual feature of their data, and the one which causes their measurements to depart from an Avrami model, is that the rate of transformation levels off and approaches zero before complete conversion is achieved; when these rates are then extrapolated to 99% conversion, extraordinarily long transformation times are required. It appears that the transformation of some appreciable fraction of the reactants is somehow inhibited.

Independent

determination

suggest

that growth kinetics

morphic

aragonite,

extraction

and because

Carlson

and Rosenfeld

spheric pressure

inversion

of nucleation

of calcite 0

the capriciousness

of the geologic

rates independent

375-455 C

are rate-determining

Because natural

textures

in the inversion

of meta-

of nucleation

of growth rate data from bulk transformation

understanding

vations

of growth rates.

measurements

an

of growth

rates.

(1981) measured

crystals

growing

on a petrographic

[Fig. 11].

necessitates

prevents

measurements,

growth rates directly

topotactically heating

Grain diameters

from aragonite

from obserat atmo-

stage over the temperature

range

were found to be linear functions

of

time, and growth rates were found to vary over a factor of about two with crystallographic dation ed.

orientation

of growth due to buildup

The temperature

yielding

migration

s

V

G* ~G T

energy of l63±4 kJ/mole.

(1956) for grain boundary

0 • V • exp(-G*/RT) . [1 - exp(-~G/RT)]

the experimental

to calculate

results

known as functions

determined

Carlson and Rosenfeld,

1981, p. 632-635,

aragonite,

interest,

nucleation

especially

(1983) reanalyzed

as described

The general

Avrami

of nat-

are of considerable transforma-

With this in

(1970) bulk transfor-

theory, using the values

of growth

above.

form of the Avrami

1940, 1941 and Christian,

Although

kinetics

as the aragonite-calcite

and Goodell's

(see

3, this volume).

rates.

of the inversion

for more complex reactions. Kunzler

data in terms of classical

rates measured

Chapter

rates and processes

insofar

tion finds use as a simple model mind, Carlson

and Essene,

determinants

assemblages,

paths for these rocks

rates from bulk transformation

growth rates are the critical ural metamorphic

of aragonite-bearing

pressure-temperature-time

of nucleation

G*, and all other parameters

of P and T, the above equation was used

growth rates during uplift

which in turn constrain

mation

The rate data showed excel-

theory of Turnbull

rate of grain boundary migration; interplanar spacing perpendicular to interface; characteristic atomic vibrational frequency; molar activation free energy; molar free energy difference between aragonite and calcite; absolute temperature; and R = universal gas constant.

Because

theoretical

stress and strain were detect-

of growth rate was found to be exponential,

with the kinetic

are independently

Extraction

drag or retar-

[Fig. 13]:

x =

x

of interfacial

dependence

an activation

lent agreement

where

[Fig. 12]; no effects of impurity

transformation

1975, section 207

4) is:

function

(see Avrami,

1939,

Calcite crystal in monacrys talline aragonite sectioned parallel to (001) produced experimentally by heating on petrographic stage. Note stress-induced lamellar twins and radial fractures; incipient circumferential fractures are visible at lower left. Maximum diameter of crystal, which measures 38 um, is parallel to [100] in aragonite; minimum diameter is parallel to [010] in aragonite. From Carlson and Rosenfeld (1981) .

-13

-13 rGROWTH RATES MEASURED ON

(001) SECTION OF ARAGONITE -14

-14

,..,

r1

U $-15

'0 ~ -15

]



o

-8

o

4

Figure 2. Log rate of calcite dissolution as a function of pH in acids at 25°C. All data are far from equilibrium and independent of P (C02)' The slope is -0.95 and rate may be assumed, within the uncertainties of the data, to be first order in all. Open circle, Tominaga e t a1. (1939): filled circle, Plummer e t a1. (1978); fiJJ.ed square Sjoberg (1978); open star, Wentzler (1971); open square, Berner and Morse (1974); filled star, King and Liu (1933); vertical bar, Wey1 (1958). After Plummer et al. (1979).

6

pH only as an approximation

where activity coefficients

and thermal and electrical Equation The dissolution unit area. matter,

neutrality

9, while relatively rate is dependent

are assumed constant,

effects are minor.

simple, contains

elements for controversy.

on surface area, since the flux is per

For smooth, spherical, non-porous

but for rough, irregularly-shaped

particles

this is a simple

porous materials

carbonates,

it can be very complex

coefficient

is not the same for all components

such as biogenic

(Walter and Morse, 1983a).

The diffusion

involved in the reaction

(e.g., DH+ is several times greater than DCa2+, Robinson and Stokes, ]959). The thickness of the stagnant boundary layer is strongly dependent on hydrodynamic conditions.

It also can depend on particle

size.

A commonly used general equation for dissolution of crystals, similar to Equation

9, is: (10) r

where R is the dissolution

rate, A is the total surface area of the dissolv-

ing solid, and r is particle 1961; Nielsen,

radius

(e.g., Frank, 1950; Terjesen et al.,

1964; Berner, 1971; Berner and Morse, 1974).

demands an essentially tic Le use for particles

infinitely
z s

where A and B are constants.

'

A plot of sin X (called the contrast transfer

function) shows that for a particular Cs there is a value of 6F, defining the Scherzer focus for which the focal and aberration effects tend to annul one another to give a reasonably constant phase shift of electrons, over a range of scattering

angles.

Figure 14 shows calculated contrast transfer

functions for a 500 kV microscope with parameters

very similar to those of

the instrument used to produce Figures 2ge, 33, 42b and 44b.

Images of very

thin crystals formed under these conditions have the correct phase information and are structure images, i.e., they show the distribution potentials projected orientations

denote atomic positions.

Phase effects in images not formed

strictly under these conditions may still be interpretable computer simulations

of atomic

in the beam direction, which for zone axial crystal

(e.g., see Kihlborg,

1979).

In this review we rely on defect identification review of the SEM literature

on sedimentary

318

with the aid of

with the TEM.

carbonates,

For a

the reader is

1.0,-------------

10

-----,

b

a 10

--_._._

-2.0

Figure

14.

Example of

500 kV; Cs• 2.0 incident

referred

contrast

axial

111m,

semi-angle

=

0.4

to Bathurst

transfer

functions

illumination mred ,

focal

=

Reeder

a high

100 A (courtesy

resolution

electron

(b) effect

of

D.J.

microscope

TRANSFORMATIONS

in dolomite

(Chapter 1) has already discussed cation ordering relations

dolomite.

In this section we briefly examine experimentally-induced

structures

that form during the

a change from the calcite-

R3c ~ R3

transformation

to dolomite-type

(1982) used TEM to examine microstructures quenched from above the transformation

structure).

stacking defects and domain boundaries.

absent in samples quenched

from lower temperatures.

image from Reeder and Nakajima in a transformed thicknesses

dolomite.

spanning

tural relationship

only a few cation layers.

transformation.

which occur pervasively

Morphologically,

Figure l5a is a lattice

Although

platelets with their actual struc-

they apparently

reflect

sequences in very localized areas which formed

tified by contrast analysis variant.

in the starting mate-

localized basal stacking defects

is not yet clearly established,

imperfect cation stacking

These dolo-

of the cations,

Both were found to be

These defects are effectively

during the cation ordering boundaries

revealing

dolomites

(~1125°C).

were found to contain two types of defects not present

to

Reeder and Nakajima

of stoichiometric

temperature

in

micro-

(corresponding

mites, having gone from a random to an ordered arrangement

rial--basal

at

of damping functions:

Smith).

(1975). POLYMORPHIC

Cation ordering

for

(a) coheren~ illumination, spread

Figure l5b shows the domain

throughout

as twin boundaries

the specimen.

these curved boundaries 319

They were iden-

(TB's) caused by a rotation resemble APB's

(e.g.,

Figure 15. TEMimages showing ordering defects from a dolomite specimen annealed at 12000C and quenched rapidly. (a) Lattice image of 0003 fringes showing a coherent basal stacking defect terIRinating at a twin boundary (TB) denoted by arrows. Fringes are offset across the TB. (b) Darkfield aicrograph showing twin d01lain microstructure. FrOllReeder and Nakajima (1982).

Fig. 8a) but geometrically

are TB's.

tions, the contrast in adjacent APB it would be the same.

domains across the TB is different;

While APB's correspond

and offset the whole structure

tion around a. but superposes

non-equivalent

(in the case of dolomite,

layers are affected,

(Fig. l6b).

In reciprocal

for an

to a lattice translation

across the boundary

both cation layers and carbonate offset the cation lattice

For at least some operating reflec-

Fig. l6a), TB's only

The twin operation

is a 180° rota-

space this produces no new reflections, reflections.

The presence

twins is confirmed by single crystal x-ray intensity

of transformation

data.

The identifica-

tion of TB's and not APB's suggests that at l125°C only the cations were disordered,

i.e., the C03 groups maintained their ordered arrangement. Notice, however, that in NaN03 (which at lower temperatures is isostructural with calcite) when heated above 275°C, the N03 groups show rotational disorder (Paul and Pryor, 1971). Subsequent ordering of the N03 groups to give the low-temperature, calcite-type structure corresponds to a change in space group (R3m

+

R3c) with a doubling of the c cell dimension

and should result in the formation

of APB's.

As reviewed

(8.41

A

+

16.82

in previous

A),

chap-

ters (1 and 6), there is evidence become rotationally 800° to 1000°C

disordered

(e.g., see Cohen and Klement, 1973; Mirwald,

Salje and Viswanathan, are ordered,

suggesting that the C03 groups in calcite upon heating within the temperature range

1976).

1976, 1979a;

In dolomite, where both cations and anions

several possibilities

exist (e.g., see the discussions

tias et al .• 1979 and Reeder and Nakajima,

by Gra-

1982).

Defects of the kinds found in these annealed and quenched dolomites are 320

TB

Figure 16. Schematic diagrams showing (a) an APBand (b) a TB in a section through a (1120) plane in the dolomite structure. Ca and Hg atoms, open and closed circles respectively, are segregated alternating planes normal to the a-axis. C03 groups are represented by triangles which have been tipped up here to illustrate their different orientations in successive layers. FromReeder and Nakajima (1982). onto

not generally possibility reasonable

found in natural dolomites.

that such microstructures to conclude

the cation ordering conclusion

On this basis, and excluding

may have been destroyed,

that (most) natural dolomites have not passed through

transformation.

Although

this might appear a trivial

in view of the rather high critical ordering temperature,

significant

the

it seems

with regard to the possibility

of metastably-disordered

it is dolo-

mites that subsequently

order.

In this context, the lack of these specific

ordering

defects

in natural dolomites

transformation

ing reactions

occurring

cipitation.

by different mechanisms,

Furthermore,

strictly to the R3c

+

does not preclude order-

such as dissolution-repre-

our review of the experimental

R3 transformation

results pertains

and does not consider changes in

the degree of ordering which may occur. Aragonite

+

calcite transformation

In contrast transformation changes

to the ordering

from orthorhombic

in primary coordination

transitions aragonite

and waa considered

typical example of a reconstructive

show many common features

Gillet and Madon transformed

(1982) have proposed

genetic environments cipitation,

rearrangement. substituting

tion relations

energies.

On the other hand, the

of partial dislocations. mechanism.

aragonite

So far

seems to occur by dissolution-repre-

often contains

is richer in Mg and Fe.

have been described

can be

In most dia-

by chemical changes which accompany

Specifically, calcite

(1961) as a

which requires breaking

(Fig. l7c-f; see also Bragg, 1924).

for such a martensitic

the transformation

as documented

calcite involves

a scheme by which aragonite

into calcite by the movement

there has been no evidence

above, the solid state

by Buerger

transformation,

of primary bonds and has large activation two structures

discussed

to rhombohedral

the structural

traces of Sr while

Nevertheless,

definite

which document that the aragonite

ture exerts some control on the calcite. 321

orientastruc-

CaQ-lime

R 3c

calcite

Fm3m

aragonite

01234A

»

I,

Pmcn ® r:

,---+ [100J=0

.;"., Co

I

e

,1,

J

'I

-

0-

I

I I

(;OlJ

,,

a

[1;-OJ

I.""-c [2110J=0

iI " , [iii I

/

t [111J

e

1

e ~ [OlOJ = b

o

o

0-0

o

[OOlJ

t e

la

0

e

a_}

12

Turner

(1972)

Barber et a1.

(1981) ,

&

UJ f-

s 0 ...l

0

c

slip ·c (0001)

25-700

5

1/3

-v

1/3 of- (i012)

Barber et

25-700

5

1/6

r-(1014}

50-130

300-500

5

> 500

S

"'250-600

170-100

al. (1983) (1981)

Barber et al.

twinning

I

Of (1012 }

they found a homogeneous boundaries

associated

of microstructures

distribution

of Mg throughout

with mosaic structure.

in biogenic materials

tricacies of their subsequent demonstrated

90-100

the usefulness

and noted some tilt

Clearly the characterization

is vital to understanding

stabilization,

of the TEM/STEM

the in-

and these early efforts have microanalysis

approach.

DEFORMATION Deformation

mechanisms

Deformation

is primarily

i.e., calcite, dolomite

of interest

and aragonite.

in the rockforming

The two main rhombohedral

calcite and dolomite -- form large parts of the sedimentary and are therefore pressure

geologically

environments

important;

environments

(Vance, 1968).

information

about deformation

on the rhombohedral knowledge

aragonite

such as sUbduction.zones

biogenic phase in modern sediments

crust

(Brown et al., 1962), as a 5), and unusual geochemical

for aragonite,

calcite and dolomite.

by Table 1, taken .from Wenk 354

varieties--

continental

is found mainly in high

For these reasons and because

carbonates

is summarized

(see Chapter

mechanisms

carbonates,

there is little

we shall concentrate

The present

state of

(1984), which represents

Figure 39. Temperature dependence of c rd t Lca L resolved shear stress for principal tion mechanisms in (a) calcite and (b) dolomite. FromBarber et a1. (1981).

results of laboratory pressure.

experiments

The temperature

on single crystals tested under confining

dependence

of the principal

is illustrated

in Figure 39.

and twinning.

The former, as explained previously,

and movement

Both materials

of dislocations,

and it is difficult

ferent twinning mechanisms. grain boundary

sliding,

can occur, additional modated

In contrast,

need not involve dislocations)

concentrations

their characteristics

by slip

involves the creation stress for the ini-

deformation

to define critical

and the diffusion

mechanisms

twinning

(which

tends to be promoted at stress

At high temperatures,

mechanisms

(see Paterson,

deformation

can deform plastically

so that there is a critical

tiation of slip on each slip system. in principle

deforma-

stresses for the dif-

where dislocation

climb,

of point defects to grain surfaces

occur whereby changes of shape can be accom-

1979; Langdon,

1984, for reviews).

Mechanisms

in calcite and dolomite will now be considered

and in more

detail. Twinning

Cal.o i.t:e,

and especially

on e

=

shear stress) for r-slip is high. also occur

{01I8} occurs readily up to ~500°C

at low temperatures,

for which the crss (critical resolved Minor twinning on rand

(Paterson and Turner, 1970).

(l859), various workers

Following

have studied e-twinning

(for a review see Barber and Wenk, 1979a).

in rhombohedral

carbonates

have a preferred

The sense of a deformation

=

{OlI2} can

in calcite, a classic mineral

in this context

for shear.

f

the early work of Pfaff

Slip and twinning

sense, i.e., a definite direction

is positive,

as defined by Turner et

al. (l954a), if rotations

during glide tend to move the pole of the glide

plane towards the c-axis,

and negative 355

if the pole moves away.

e-twinning

is positive.

Elastic

(i.e., reversible)

twins can also be produced

cite (Garber, 1938, 1947), but these are a curiosity,

irrelevant

in cal-

to plastic

deformation. Rhombohedral

=

slip on r

{l014} , in the negative

over a wide range of temperatures, et al., 1954a) and polycrystals f-slip is also an important somewhat enigmatic;

sense, prevails

as shown by work on single crystals

mechanism.

Despite much work, r-slip remains

in many instances r-slip is not well defined at the TEM

level, the dislocations

commonly being very curved, not lying in well-defined

planes and interacting

strongly with each other, even at low temperatures.

The basic lattice translation

vector in the {l014} planes is 7.7

large for the Burgers vector of a primary

slip system.

cessive glide of two partial

with Burgers vectors

+

i c],

(Turner

(e.g., Turner et al., 1956; Wenk et al., 1973).

i.e., half of 7.7

A

dislocations

K

long,

Slip via the suc[jal + ja2

in length, is possible, which would give stacking

faults in the C03 sublattice. produced no concrete evidence

But TEM and HREM investigations for the mechanism.

have so far

There is little evidence

for the operation of other slip systems, although Turner and Orozco

(l976)

have reported a case of basal slip. Flow laws have been determined marbles

(e.g., Heard and Raleigh,

1977; 1980).

for fine-grained

1972; Rutter,

limestones

For high strain rates at low temperatures, .

exponentially-dependent

on stress, G (E

=

and coarse

1972; 1974; Schmid et al., strain rate ~ is

exp CG) with C ~ 2.5 kbar

-1

for

Solnhofen limestone at

... ... ...... QI'"

...u ~ .. ..." U

.. 0

...." ".. ".., 0>

.... QI

~ §

.... ~~ ..... ~

I

0

....

,"... -e ...

e-,

...::r

....~'" QI

::r

....... ~'"

358

Biaxiality metamorphic

rocks and also in carbonatites

attributed

moderate-to-strong

rally-deformed lamellae

in calcites2 from all grades of

is known to be widespread

calcites

biaxiality

(Turner, 1975a).

to overlap between optically-visible

and the host crystal.

Hauser and Wenk

the optical properties

of submicroscopically-twinned

suggested

that strong biaxiality

was largely attributable

(1975b) concluded

attributed

that slight biaxiality

He also concluded

a model that calcite and

to this effect.

(5° < 2V < 10°) could not be had a submicro-

that in natural rocks, biaxiality

late and minor strain effect of no general geological cance.

thin e-twin

to (optically) visible twinning and, by inference,

scopic origin.

The findings of Barber and Wenk

(l975b)

and natu-

(1976) proposed

predicted

Turner

Turner

in both experimentally-

or petrogenic

was a signifi-

(1979a) may somewhat question

these

views, however,

since they found that a strongly biaxial calcite

(2V ~ 45°)

showed evidence

of repeated

(Fig. 40b).

submicroscopic

twinning and recovery

The calcite also contained high dislocation boundaries,

which were not incorporated

(l976), nor anticipated

e-twinning

tion on naturally-deformed lent.

and many sub-grain

by it.

In calcite mylonites, axial symmetry,

densities

into the model of Hauser and Wenk

which show very strong textures approximating is often pervasive,

but the more sparse informa-

dolomites

that f-twinning

indicates

Both findings are in general accord with experimental

temperature

relationships.

Weiss

is not preva~

data on twinning-

(l954) showed that crystallographic

orienta-

tions of e-twins and the a-axis for each grain in a deformed aggregate be used to determine and the resultant

the net compressive

strain directions.

tally with Yule marble

to

could

and tensile stresses which had acted

The method has been tested experimen-

(Borg and Turner, 1953) and can be applied

to geologi-

cal formations

(e.g., Weiss, 1954; Groshong,

since twinning

occurs at low stresses in calcite and may therefore postdate

a main deformational Other defects

event.

in deformed carbonates

A comprehensive be premature

description

of dislocation

in the light of current knowledge

more space than is available.

Dislocations.

behavior

however,

359

occupy

and comparisons

of

several problems.

There are marked differences

is uniaxial.

would

and this section includes recent

to help in resolving

perfect calcite

in carbonates

and would undoubtedly

Some general observations

calcite and dolomite are possible, HREM results which promise

2 Structurally

1972), although with caution

in the general appearance

of dislocation temperatures,

microstructures as mentioned

in calcites and dolomites

earlier, and exemplified

(Barber and Wenk, 1976; 1979b; Barber, Kohlstedt, products

in TEM observations

1977; Barber et al., 1981; Goetze and

1977; White and White, 1980).

of r-slip,

deformed at low

Dislocations

are far less regimented

in calcite, mostly

and crystallographically

trolled than those in dolomite deformed under comparable is especially acteristic

conditions.

This

true for dolomite yielding by c-slip, which has a quite char-

appearance,

with well-defined

slip planes (commonly in slip bands)

and long straight

mainly-screw

r-slip in calcite

and c- and f-slip in dolomite,

segments of dislocation.

dolomite is most common at higher temperatures are not so straight deformed natural

Heavily-strained

with dislocation

Figures 4la,b,c show

respectively.

f-slip

in

(>300°C) and the dislocations

as basal plane dislocations

calcites of low metamorphic

are very high, so high that individual guish.

con-

at low temperatures.

grade the dislocation

dislocations

are difficult

In many

densities to distin-

laboratory

densities

specimens have the same characteristics, 2 ~lOlO cm- . Such high densities are extremely

uncommon in dolomites. TEM evidence more difficult

and flow stress data (Table l) suggest that r-slip may be

to initiate in calcite at low temperatures

at 300°C) than c-slip is in dolomite is the reason, in calcite

dislocations

generated

(the crss is 64 MPa

Whether or not this

by r-slip at low temperatures

«200°C)

spread laterally out of the slip planes, causing much interaction

with the generation type of behavior

f

deformed

dislocation

of many small dislocation

are easy to identify up to 700°C. specimens of both carbonates

microstructure.

at high temperatures in dolomites

loops and other debris.

is not found in dolomite until ~400°C, although

slip planes, cand prevalent,

(50 MPa at 25°C).

deformed

Twinning

below 300°C.

When climb becomes

are more closely similar in

is more prominent

(>600°C) than in calcites,

This

the active

in dolomites

deformed

and it does not occur at all

Work on experimentally-deformed

samples

(Barber and Wenk, 1979a; Barber et al., 1981) shows that dislocations more likely to be created by twinning in calcite than in dolomite. tions involved

in both c- and f-slip in dolomite

as described

in the next section.

dissociation

in calcite.

Stacking faults.

In dolomite

and those on

analyzed by the methods

f

crystals

deformed between

Figures

contrast

for

300 and 500°C,

those on .a always terminating

much less frequently.

of diffraction

(Barber et al., 1981; 1983).

Disloca-

into partials,

So far there is no certain evidence

long faults are visible on the a- and f-planes, at dislocations

can. dissociate

are

The a-faults are easily

(Barber et al., 1978) and HREM

42a and b show low and high resolution

Figure 41. (a) Dislocations created by r-slip in calcite deformed at 25°C. (b) Dislocations created by a-slip in dolomite deformed at 500°C, (c) Dislocations created by r-slip in dolomite deformed at 800°C (minor r-slip also present), (d) Climbing dislocations associated with a-slip in dolomite deformed at 700°C. (AlII MeV micrographs). 361

."...o :