THE VIBRATIONAL SPECTROSCOPY OF MINERALS

WAYDE NEIL MARTENS

B. APPL. SCI. (APPL. CHEM.) M.SC. (APPL. SCI.)

Inorganic Materials Research Program, School of Physical and Chemical

Science, Queensland University of Technology

A THESIS SUBMITED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY OF THE QUEENSLAND UNIVERSITY OF TECHNOLOGY

2004

2

3

Toss another rock on the Raman…

4

KEYWORDS

Annabergite

Aragonite

Arupite

Baricite

Cerussite

Erythrite

Hörnesite

Infrared Spectroscopy

Köttigite

Minerals

Parasymplesite

Raman Spectroscopy

Solid Solutions

Strontianite

Vibrational Spectroscopy

Vivianite

Witherite

5

ABSTRACT

This thesis focuses on the vibrational spectroscopy of the aragonite and

vivianite arsenate minerals (erythrite, annabergite and hörnesite), specifically the

assignment of the spectra. The infrared and Raman spectra of cerussite have been

assigned according to the vibrational symmetry species. The assignment of satellite

bands to 18O isotopes has been discussed with respect to the use of these bands to the

quantification of the isotopes. Overtone and combination bands have been assigned

according to symmetry species and their corresponding fundamental vibrations. The

vibrational spectra of cerussite have been compared with other aragonite group

minerals and the differences explained on the basis of differing chemistry and crystal

structures of these minerals.

The single crystal spectra of natural erythrite has been reported and compared

with the synthetic equivalent. The symmetry species of the vibrations have been

assigned according to single crystal and factor group considerations. Deuteration

experiments have allowed the assignment of water vibrational frequencies to discrete

water molecules in the crystal structure. Differences in the spectra of other vivianite

arsenates, namely annabergite and hörnesite, have been explained by consideration of

their differing chemistry and crystal structures.

A novel approach to the assignment of site occupancy of ions in the erythrite –

annabergite solid solution has been reported. This approach has utilised vibrational

spectroscopy, in conjunction with careful consideration of the crystal structures of the

minerals. It has been shown that in the erythrite – annabergite solid solution Co

6

prefers metal site 2 contrasting nickel which prefers site 1. This study in conjunction

with other studies has yielded the trend that the more electronegative metal prefers to

occupy site 1, with the least electronegative metal preferring to occupy site 2.

Fundamentally this thesis has increased the knowledge base of the

spectroscopic properties of the aragonite and the vivianite minerals. The site

occupancy of metal ion substitutions in solid solution series of the vivianite group of

minerals has been further enhanced, with novel method of studying the site occupancy

of ions in solid solutions has been developed. A detailed knowledge and

understanding of factor group analysis applied to the study of minerals has been

achieved.

7

PUBLICATIONS WRITTEN IN THE COURSE OF STUDY

1. Martens, W., N., Rintoul, L., Kloprogge, J.T., and Frost, R., L. (2004) Raman single crystal study of cerussite. American Mineralogist, 89, 352-358.

2. Martens, W., N., Kloprogge, J.T., Frost, R., L., and Rintoul, L. (2004) Site

occupancy of Co and Ni in erythrite - annabergite solid solutions by vibrational spectroscopy. American Mineralogist, Submitted.

3. Martens, W., N., Kloprogge, J.T., Frost, R., L., and Rintoul, L.. (2004)

Single-crystal Raman study of erythrite,Co3(AsO4)2·8H2O. Journal of Raman Spectroscopy, 35, 208-216.

4. Martens, W.N., Frost, R.L., and Williams, P.A. (2003) The basic copper

phosphate minerals pseudomalachite, ludjibaite and reichenbachite: An infrared emission and Raman spectroscopic study. Neues Jahrbuch fuer Mineralogie, Monatshefte(8), 337-362.

5. Martens, W.N., Frost, R.L., Kloprogge, J.T., and Williams, P.A. (2003) The

basic copper arsenate minerals olivenite, cornubite, cornwallite, and clinoclase: An infrared emission and Raman spectroscopic study. American Mineralogist, 88(4), 501-508.

6. Martens, W., Frost, R.L., and Williams, P.A. (2003) Molecular structure of

the adelite group of minerals - a Raman spectroscopic study. Journal of Raman Spectroscopy, 34(2), 104-111.

7. Martens, W., Frost, R.L., and Williams, P.A.. (2003) Raman and infrared

spectroscopic study of the basic copper chloride minerals - implications for the study of the copper and brass corrosion and "bronze disease". Neues Jahrbuch fuer Mineralogie, Abhandlungen, 178(2), 197-215.

8. Martens, W., Frost, R.L., Kloprogge, J.T., and Williams, P.A. (2003) Raman

spectroscopic study of the basic copper sulphates- implications for copper corrosion and "bronze disease". Journal of Raman Spectroscopy, 34(2), 145-151.

9. Martens, W., Frost, R.L., and Kloprogge, J.T. (2003) Raman spectroscopy of

synthetic erythrite, partially dehydrated erythrite and hydrothermally synthesized dehydrated erythrite. Journal of Raman Spectroscopy, 34(1), 90-95.

10. Martens, W., and Frost, R.L. (2003) An infrared spectroscopic study of the

basic copper phosphate minerals: Cornetite, libethenite, and pseudomalachite. American Mineralogist, 88(1), 37-46.

11. Martens, W.N., Frost, R.L., Kristof, J., and Kloprogge, J.T. (2002) Raman

spectroscopy of dimethyl sulphoxide and deuterated dimethyl sulphoxide at 298 and 77 K. Journal of Raman Spectroscopy, 33(2), 84-91.

8

12. Martens, W.N., Frost, R.L., Kristof, J., and Horvath, E. (2002) Modification of Kaolinite Surfaces through Intercalation with Deuterated Dimethylsulfoxide. Journal of Physical Chemistry B, 106(16), 4162-4171.

13. Martens, W.N., Ding, Z., Frost, R.L., Kristof, J., and Kloprogge, J.T. (2002)

Raman spectroscopy of hydrazine- intercalated kaolinite at 77, 298, 323, 343 and 358 K. Journal of Raman Spectroscopy, 33(1), 31-36.

14. Martens, W.N., Kloprogge, J.T., Frost, R.L., and Bartlett, J.R. (2002) A

Crystallite Packing Model for Pseudoboehmite Formed during the Hydrolysis of Trisecbutoxyaluminium to Explain the Peptizability. Journal of Colloid and Interface Science, 247(1), 132-137.

15. Kristof, J., Frost, R.L., Martens, W.N., and Horvath, E. (2002) Separation of

Adsorbed and Intercalated Hydrazine in Hydrazine-Hydrate Intercalated Kaolinite by Controlled-Rate Thermal Analysis. Langmuir, 18(4), 1244-1249.

16. Kloprogge, J.T., Visser, D., Martens, W.N., Duong, L.V., and Frost, R.L.

(2003) Identification by RAMAN microscopy of magnesian vivianite formed from Fe2+, Mg2+, Mn2+ and PO4

3- in a Roman camp near fort Vechten, Utrecht, The Netherlands. Netherlands Journal of Geosciences, 82(2), 209-214.

17. Kloprogge, J.T., Martens, W.N., Nothdurft, L., Duong, L.V., and Webb, G.E.

(2003) Low temperature synthesis and characterization of nesquehonite. Journal of Materials Science Letters, 22(11), 825-829.

18. Kloprogge, J.T., Hickey, L., Doung, L.V., Martens, W., N., and Frost, R., L.

(2004) Synthesis and characterization of K2Ca5(SO4)6·H2O, the equivalent of görgeyite, a rare evaporite mineral. American Mineralogist, 89, 266-272.

19. Johnson, T.E., Martens, W., Frost, R.L., Ding, Z., and Kloprogge, J.T. (2002)

Structured water in hydrotalcites of formula MgxZn6-xAl2(OH)16(CO3).4H2O: a Raman microscopic study. Journal of Raman Spectroscopy, 33(8), 604-609.

20. Frost, R.L., Williams, P.A., Martens, W., Kloprogge, J.T., and Leverett, P.

(2002) Raman spectroscopy of the basic copper phosphate minerals cornetite, libethenite, pseudomalachite, reichenbachite and ludjibaite. Journal of Raman Spectroscopy, 33(4), 260-263.

21. Frost, R.L., Williams, P.A., Martens, W., and Kloprogge, J.T. (2002) Raman

spectroscopy of the polyanionic copper(II) minerals buttgenbachite and connellite: implications for studies of ancient copper objects and bronzes. Journal of Raman Spectroscopy, 33(9), 752-757.

22. Frost, R.L., Williams, P.A., and Martens, W. (2003) Raman spectroscopy of

the minerals boleite, cumengeite, diaboleite and phosgenite - implications for the analysis of cosmetics of antiquity. Mineralogical Magazine, 67(1), 103-111.

9

23. Frost, R.L., Williams, P.A., Kloprogge, J.T., and Martens, W. (2003) Raman spectroscopy of the copper chloride minerals nantokite, eriochalcite and claringbullite - implications for copper corrosion. Neues Jahrbuch fuer Mineralogie, Monatshefte(10), 433-445.

24. Frost, R.L., Weier, M.L., Martens, W., Kloprogge, J.T., and Ding, Z. (2003)

Dehydration of synthetic and natural vivianite. Thermochimica Acta, 401(2), 121-130.

25. Frost, R.L., Weier, M.L., Martens, W., Kloprogge, J.T., and Ding, Z.. (2003)

Thermal decomposition of the vivianite arsenates- implications for soil remediation. Thermochimica Acta, 403(2), 237-249.

26. Frost, R.L., Martens, W.N., and Williams, P.A. (2002) Raman spectroscopy

of the phase-related basic copper arsenate minerals olivenite, cornwallite, cornubite and clinoclase. Journal of Raman Spectroscopy, 33(5), 475-484.

27. Frost, R.L., Martens, W.N., Rintoul, L., Mahmutagic, E., and Kloprogge, J.T.

(2002) Raman spectroscopic study of azurite and malachite at 298 and 77 K. Journal of Raman Spectroscopy, 33(4), 252-259.

28. Frost, R.L., Martens, W.N., Kloprogge, T., and Williams, P.A. (2002)

Vibrational spectroscopy of the basic manganese and ferric phosphate minerals: Strunzite, ferrostrunzite and ferristrunzite. Neues Jahrbuch fuer Mineralogie, Monatshefte(11), 481-496.

29. Frost, R.L., Martens, W.N., and Kloprogge, J.T. (2002) Raman spectroscopic

study of cinnabar (HgS), realgar (As4S4), and orpiment (As2S3) at 298 and 77K. Neues Jahrbuch fuer Mineralogie, Monatshefte(10), 469-480.

30. Frost, R.L., Martens, W.N., Duong, L., and Kloprogge, J.T. (2002) Evidence

for molecular assembly in hydrotalcites. Journal of Materials Science Letters, 21(16), 1237-1239.

31. Frost, R.L., Martens, W., Williams, P.A., and Kloprogge, J.T. (2003) Raman

spectroscopic study of the vivianite arsenate minerals. Journal of Raman Spectroscopy, 34(10), 751-759.

32. Frost, R.L., Martens, W., Williams, P.A., and Kloprogge, J.T.. (2002) Raman

and infrared spectroscopic study of the vivianite-group phosphates vivianite, baricite and bobierrite. Mineralogical Magazine, 66(6), 1063-1073.

33. Frost, R.L., Martens, W., Kloprogge, J.T., and Williams, P.A. (2002) Raman

spectroscopy of the basic copper chloride minerals atacamite and paratacamite: implications for the study of copper, brass and bronze objects of archaeological significance. Journal of Raman Spectroscopy, 33(10), 801-806.

10

34. Frost, R.L., Martens, W., Kloprogge, J.T., and Ding, Z. (2003) Raman spectroscopy of selected lead minerals of environmental significance. Spectrochimica Acta, Part A: Molecular and Biomolecular Spectroscopy, 59A(12), 2705-2711.

35. Frost, R.L., Martens, W., Ding, Z., Kloprogge, J.T., and Johnson, T.E. (2003)

The role of water in synthesized hydrotalcites of formula MgxZn6-

xCr2(OH)16(CO3).4H2O and NixCo6-xCr2(OH)16(CO3).4H2O-an infrared spectroscopic study. Spectrochimica Acta, Part A: Molecular and Biomolecular Spectroscopy, 59A(2), 291-302.

36. Frost, R.L., Martens, W., Ding, Z., and Kloprogge, J.T. (2003) DSC and

high-resolution TG of synthesized hydrotalcites of Mg and Zn. Journal of Thermal Analysis and Calorimetry, 71(2), 429-438.

37. Frost, R.L., Kristof, J., Mako, E., and Martens, W.N. (2002) Modification of

the Hydroxyl Surface of Kaolinite through Mechanochemical Treatment Followed by Intercalation with Potassium Acetate. Langmuir, 18(17), 6491-6498.

38. Frost, R.L., Kristof, J., Horvath, E., Martens, W.N., and Kloprogge, J.T.

(2002) Complexity of Intercalation of Hydrazine into Kaolinite-A Controlled Rate Thermal Analysis and DRIFT Spectroscopic Study. Journal of Colloid and Interface Science, 251(2), 350-359.

39. Frost, R.L., Kloprogge, T., Williams, P.A., Martens, W., Johnson, T.E., and

Leverett, P. (2002) Vibrational spectroscopy of the basic copper phosphate minerals: pseudomalachite, ludjibaite and reichenbachite. Spectrochimica Acta, Part A: Molecular and Biomolecular Spectroscopy, 58A(13), 2861-2868.

40. Frost, R.L., Kloprogge, T., Weier, M.L., Martens, W.N., Ding, Z., and

Edwards, H.G.H. (2003) Raman spectroscopy of selected arsenates-implications for soil remediation. Spectrochimica Acta, Part A: Molecular and Biomolecular Spectroscopy, 59(10), 2241-2246.

41. Frost, R.L., Edwards, H.G.M., Duong, L., Kloprogge, J.T., and Martens,

W.N. (2002) Raman spectroscopic and SEM study of cinnabar from Herod's palace and its likely origin. Analyst (Cambridge, United Kingdom), 127(2), 293-296.

42. Frost, R.L., Duong, L., and Martens, W. (2003) Molecular assembly in

secondary minerals - Raman spectroscopy of the arthurite group species arthurite and whitmoreite. Neues Jahrbuch fuer Mineralogie, Monatshefte(5), 223-240.

43. Frost, R.L., Ding, Z., Martens, W.N., Johnson, T.E., and Kloprogge, J.T.

(2003) Molecular assembly in synthesised hydrotalcites of formula CuxZn6-

xAl2(OH)16(CO3).4H2O-a vibrational spectroscopic study. Spectrochimica Acta, Part A: Molecular and Biomolecular Spectroscopy, 59(2), 321-328.

11

44. Frost, R.L., Ding, Z., Martens, W.N., and Johnson, T.E. (2003) Thermal activation of copper oxide based upon the copper hydrotalcite CuxZn6-

xAl2(OH)16(CO3).4H2O. Thermochimica Acta, 398(1-2), 167-174. 45. Frost, R.L., Ding, Z., Kloprogge, J.T., and Martens, W.N. (2002) Thermal

stability of azurite and malachite in relation to the formation of mediaeval glass and glazes. Thermochimica Acta, 390(1-2), 133-144.

46. Frost, R., L., Martens, W., and Kloprogge, J.T. (2004) Synthetic deuterated

erythrite-a vibrational spectroscopic study. Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy, 60(1-2), 343-9.

47. Frost, R., L., Kloprogge, J.T., and Martens, W., N. (2004) Raman

spectroscopy of the arsenate and sulphates of the tsumcorite mineral group. Journal of Raman Spectroscopy, 35, 28-35.

48. Ding, Z., Martens, W., and Frost, R.L. (2002) Thermal activation of copper

nitrate. Journal of Materials Science Letters, 21(18), 1415-1417.

12

TABLE OF CONTENTS

KEYWORDS…………………………………………………………………………………….... 4

ABSTRACT……………………………………………………………………………………….. 5

PUBLICATIONS WRITTEN IN THE COURSE OF STUDY………………………………… 7

STATEMENT OF ORIGINALITY……………………………………………………………… 14

ACKNOWLEGMENTS…………………………………………………………………………... 15

1. INTRODUCTION……………………………………………………………………………… 16

1.1. Description Of Scientific Problem Investigated………………………………………… 16

1.2. Overall Objectives Of This Study………………………………………………………... 17

1.3. Specific Aims Of This Study……………………………………………………………… 17

1.4. Account Of Scientific Progress Linking The Scientific Papers………………………… 18

1.5. References…………………………………………………………………………………. 20

2. THEORETICAL CONSIDERATIONS OF MINERAL SPECTROSCOPY……………… 21

2.1. Introduction……………………………………………………………………………….. 21

2.2. Factor Group Analysis……………………………………………………………………. 24

2.3. Elements Of Symmetry…………………………………………………………………… 25

2.4. Prediction Of The Number Of Normal Modes Of Vibration…………………………... 27

2.5. Point Group Analysis Of Free Molecules………………………………………………... 28

2.6. Application Of The Internal Modes To The Crystal Structure………………………... 32

2.7. Lattice Modes……………………………………………………………………………... 35

2.8. Assignment Of Spectra, Selection Rules, Polarised

Orientated Vibrational Spectroscopy …………………………………………………….

36

2.9. Birefringence……………………………………………………………………………... 39

2.10. Porto Notation……………………………………………………………………………. 40

2.11. Olivenite Example………………………………………………………………………... 41

2.12. Vivianite Minerals And Solid Solutions………………………………………………… 44

2.13. References………………………………………………………………………………… 48

13

3. SINGLE CRYSTAL STUDY OF CERUSSITE……………………………………………… 52

4. SINGLE CRYSTAL STUDY OF ERYTHRITE…………………………………………….. 62

5. SYNTHETIC DEUTERATED ERYTHRITE-A VIBRATIONAL SPECTROSCOPIC

STUDY……………………………………………………………………………….………….

74

6. SITE OCCUPANCY OF CO AND NI IN ERYTHRITE - ANNABERGITE SOLID

SOLUTIONS BY VIBRATIONAL SPECTROSCOPY……………………………………...

84

7. GENERAL DISCUSSION……………………………………………………………………... 95

8. APPENDIX…………………………………………………………………………………….. 98

14

STATEMENT OF ORIGINALITY

The material presented in this thesis has not been previously submitted for a

degree at any other university or institution. To the best of my knowledge, this thesis

contains no material published or written by any other person except where due

acknowledgement is made.

Wayde Martens

15

ACKNOWLEDGMENTS

Thank you to all the people who made this project a possibility. By no means is this acknowledgment complete and by chance the author does not acknowledge you, your contribution however small it is gratefully appreciated. The effort, and financial support of my supervisor Associate Professor Ray Frost has made this thesis a reality. Thank you for always being encouraging, even though sometimes I did not appreciate the constant push you applied which has resulted in the completion of this thesis. You have imparted to me the knowledge of how to make the system work. Dr. Theo Kloprogge (unofficial co-supervisor), Thank you for all the help you have given me through out my journey of the thesis, in both life and academic life. With out your help this thesis would not contain the amount of mineralogy that it does. Although the emotional scars that you have imparted by you will not heal, thank you Dr. Llew (The instrument Nazi) Rintoul, for all your help and insults. Your help in the area of vibrational spectroscopy has left me with a very deep understanding of how the molecules of the universe vibrate. As such thanks for the simulation, mental that is. To all members of the inorganic materials group thank you for always distracting me. You have managed to have never allow me to get bogged down on any one thing. To the demolition expert (Simon Montgomery) please be safe in you future explosions. Sorry I mean exploits. Hopefully you are not required to operate a microwave in the future. Neil, you have seen the bane of my existence, nature always wins. I know your pain when you see SEM images of natural kaolinite growing on quartz. Matt Weier and James Brady may you suffer the rest of your existence for the pain you have caused me. Thanks for trying to keep me sane. However it is apparent that you have failed. Mr. Tony Raferty, I can not thank you enough for all the help you have given me with XRD, crystallography, and Siroquant. Loc Doung, Thor Bostrum and all at the AEMF thank you for allowing me access to your instruments.

16

1. INTRODUCTION:

1.1. DESCRIPTION OF SCIENTIFIC PROBLEM INVESTIGATED

Minerals have been important since ancient times, due to their wealth creation

potential and raw material production [1-3]. Minerals have been used as sources of

metals, building materials, additives, etc.[1-3] Some minerals were even used as

colorants for medieval manuscripts and cosmetics in ancient Egypt [4-8].

Identification of minerals is of the utmost importance as knowledge of secondary

mineralogy gives a direct indication of the primary mineralogy which is of particular

interest for mineral exploration [1]. Therefore, an identification procedure is required

for minerals that is non-destructive, fast, and effective. Raman spectroscopy has

proven a useful technique for the non-destructive identification of minerals [9-12].

Vibrational spectroscopy provides a tool for the identification of minerals with

mineral infrared (IR) spectra being routinely acquired [13]. Identification is easy

through a finger printing technique [12, 13], but the interpretation of spectra is

difficult. Interpretation of mineral vibrational spectra requires factor group analysis

and careful consideration of the crystal structures of the minerals. This study was

undertaken to acquire the spectra of selected groups of closely related minerals and

correlate their spectral differences to their crystal structures. Solid solutions are

known to hinder spectroscopic identification, due to changes in the spectra caused by

replacement of cations or anions in the minerals. This study has been undertaken to

understand the physical changes in the crystal structure due to solid solutions and

correlate this with the effect on the vibrational spectra.

17

1.2. OVERALL OBJECTIVES OF THIS STUDY

In this study the primary objective is firstly to gain an understanding of the

vibrational spectroscopy of minerals. Specifically cerussite in relation to the other

aragonite group minerals and the annabergite erythrite members of the vivianite

group of minerals and their solid solutions. The second objective is to use

vibrational spectroscopy to gain an understanding of the solid solution phenomenon

in the vivianite group of minerals. This study of the solid solutions of vivianites will

focus on the effect of cationic substitution on the vibrational spectra and crystal

structure in synthetic systems. The study utilises the analytical techniques of Raman

spectroscopy, infrared spectroscopy, SEM, and XRD.

1.3. SPECIFIC AIMS OF THE STUDY

The specific aims of the research presented in this thesis were:

• Acquire the vibrational spectra of the aragonite minerals and explain

the differences in the spectra due to changes in the crystal structure of

the minerals.

• Acquire and explain the single crystal Raman spectra of erythrite and

compare these with other vivianite arsenates. A detailed analysis of

bands is undertaken to pinpoint specific vibrating units.

• Determine and explain the solid solution phenomena of erythrite and

annabergite with respect to the vibrational spectra of the minerals.

• Determine the site occupancy of metal ions in the solid solution series

between erythrite and annabergite

18

1.4. ACCOUNT OF SCIENTIFIC PROGRESS LINKING THE SCIENTIFIC

PAPERS

To use vibrational spectra to explain differences related to crystal structures of

minerals it is necessary to gain a full understanding of factor group analysis. It is not

possible to assign any mineral spectra without such knowledge. To establish a basis

the vibrational spectroscopy of the vivianite minerals must be understood, before the

consideration of solid solutions. At this stage with detailed knowledge of factor group

analysis and knowledge of the spectroscopic behaviour of the vivianite arsenates it is

possible to undertake a study of the solid solution phenomenon in the erythrite –

annabergite series. To this end the following chapters are arranged to (Figure 1):

1. Show how factor group analysis is undertaken, how to assign spectra of minerals,

and particular assignment tools utilised in this thesis. (Chapter 2)

2. Assign the single crystal spectra of cerussite and compare the spectra of other

aragonite minerals, further illustrating the assignment of spectra of minerals.

(Chapter 3)

3. Determine and assign the single crystal spectra of erythrite, with comparison to

other vivianite arsenates to determine a basis for further work on

erythrite/annabergite. Comparison of the spectra to other vivianite minerals

establishes a crystallographic basis for variation of the spectra. (Chapter 4)

4. Investigate the spectra of deuterated erythrite to determine the assignment of water

vibrational modes, and track the independent water molecules. (Chapter 5)

5. Investigate the solid solution of erythrite annabergite using synthetic samples.

(Chapter 6)

19

Figure 1 Interaction and arrangement of chapters

Factor Group Analysis (FGA) of Minerals (Chapter 2)

Single Crystal Study of Cerussite (Chapter 3)

Application of FGA to assignment of spectra Spectral differences due to crystal structure

Single crystal study Of Erythrite (Chapter 4)

Basis study of spectroscopy of vivianites

Deuterated erythrite (Chapter 5)

Further assignment of spectra of vivianites particularly water modes

Solid solution of erythrite and annabergite (Chapter 6)

Determine site occupancy in solid solutions with vibrational spectroscopy

Appendix (Chapter 8)

Auxiliary mineral spectroscopy studies

20

References

1. Klein, C., Mineral Science. 22 ed. 2002, New York: Wiley. 2. Korbel, P. and M. Novak, The Complete Encylopedia Of Minerals. 2001,

U.K.: Grange Books. 3. Bishop, A.C., A.R. Woolley, and W.R. Hamilton, Philip's Minerals, Rocks

and Fossils. 3rd ed. 1999, London: Octopus Publishing Proup. 4. Frost, R.L., W. Martens, J.T. Kloprogge, and P.A. Williams, Journal of Raman

Spectroscopy, 2002. 33(10): p. 801-806. 5. Reiderer, J., Archaeometry, 1974. 16: p. 102. 6. Scott, D.A., Stud. Conserv., 200. 45: p. 39. 7. Kuchitsu, N., N. Kuroki, S. Inokuchi, and S. Mitsuishi, Hozon Kagaku, 1999.

38: p. 108. 8. Dei, L., A. Ahle, P. Baglioni, and D.F. Dini, E., Stud. Conserv., 1998. 43: p.

80. 9. Brooke, C., H. Edwards, and J. Tait, Journal of Raman Spectroscopy,

1999(30): p. 429. 10. Burgio, L., Ciomartan, and R. Clark, Journal of Raman Spectroscopy,

1997(29): p. 79. 11. Farmer, V.C., ed. The Infrared spectra of minerals. 1974, Mineralogical

Society: London. 12. Frost, R.L., J. Kloprogge, and J. Schmidt, Non-destructive identification of

minerals by Raman microscopy. The Internet Journal of Vibrational Spectroscopy. 3(4).

13. Hunt, G., Salisbry. Mod. Geol., 1971(2): p. 23.

21

2. THEORETICAL CONSIDERATIONS OF MINERAL SPECTROSCOPY

2.1. Introduction

Minerals are elements or combinations of elements in a crystalline form which

have been produced by a natural geological process [1-3], as such each different

mineral species must be different in composition or crystal structure [1-3].

Therefore, minor substitutions in a mineral structure are not classified as different

minerals unless this substitution occurs in constant specific ratios [1].

Identification of minor phases - present as associations - can also give an

indication of the geological history of the major mineral species [1]. Therefore, an

identification procedure that is non-destructive, fast, and effective is required to

identify these minerals. Current mineral identification techniques include:

(i) X-ray Diffraction (XRD):

X-ray diffraction is currently used as the main phase identification

technique. However, it is occasionally inappropriate due to the

necessity to grind the samples, which causes obvious damage to

valuable samples.

(ii) Optical Mineralogy:

Optical mineralogy is used as an identification procedure but 1 µm thin

sections must be prepared. The procedure is time consuming with user

experience playing a major role in the analysis.

(iii) SEM, AA/AE, ICP-MS, XRF:

These are the main suite of techniques used for compositional analysis.

Carbon coating (SEM) or dissolution (AA/AE, ICP-MS, XRF) is a

22

requisite for these analyses. A set of accurately known standards have

to be prepared for a reliable calibration of the method.

Vibrational spectroscopy provides an alternative tool for the identification of

minerals, with mineral IR spectra being routinely acquired [4]. Identification is easy

through a finger printing technique [4, 5], but the interpretation of spectra is difficult.

Vibrational spectroscopy is often the best method for the identification of minerals

when other techniques such as XRD can not be undertaken due to sample quantity

(minor phases), crystallinity (amorphous/semi-amorphous) and destructiveness

(grinding/dissolution)[5, 6]. Raman spectroscopy has proven a useful technique for

the non-destructive identification of minerals [5-8].

Copper arsenates, phosphates, chlorides, and carbonates have been extensively

used in many medieval manuscripts for pigments, but due to their nature, are easily

confused owing to their similarities in chemical composition, crystal structure, and

colour [9-16]. The dark green copper phosphate polymorphs - pseudomalachite,

reichenbachite, and ludjibaite - are difficult to distinguish. These copper minerals are

also confused with malachite Cu2(CO3)(OH)2 [17], and the copper arsenates

cornwallite, olivenite, and clinoclase [9-16]. Easy non-destructive methods of

analysis are required for these minerals, thus the utilisation of vibrational

spectroscopy is important for the study of these minerals and medieval manuscripts

[18, 19]. Raman spectroscopy can easily distinguish different anions such as

carbonate, arsenate, phosphate, nitrates, sulphides, chlorides, etc. [5, 6]. Minerals with

polymorphs can also be distinguished through the use of factor group analysis. Raman

spectroscopy has even been employed to identify the geographical origin of

23

minerals[19]. Raman microscopy allows the easy selective orientation of crystals and

has the ability to acquire the spectrum of a single crystal due to its spatial resolution

of 1µm [20]. Fluorescence from the sample matrix is also reduced with Raman

microscopy due to the reduction of sampling volume [20]. Birefringence can also

cause problems in Raman spectroscopy, because of the confocal nature of

microscopes; Raman microscopy evades this problem [20]. However, Raman

microscopy does suffer from some disadvantages, for instance a large amount of

leakage due to the microscope optics can occur. This is due to the analysed and

excitation radiation collected without a strictly parallel arrangement. The amount of

radiation that is not incident/ collected strictly parallel may be calculated [20].

Samples have an increased tendency to burn / decompose in Raman microscopy with

the laser power concentrated into a smaller volume. This is especially a problem with

blue samples when using He-Ne lasers as blue samples absorb red light, thus the laser

radiation is absorbed by the sample and dissipated as heat.

Solid solutions provide an added complication to vibrational spectroscopy due to

the broadening of spectra from random substitutions into the crystalline lattice causing

a loss of factor group. Standards are often not available for comparison and therefore,

identification is sometimes inconclusive, requiring the synthesis of minerals for

reliable identification standards.

Complex solid solutions also exist, the presence of which may be studied by

vibrational spectroscopy, as other techniques such as XRD may not detect them.

Many secondary mineral samples are also rare and found only in exclusive secondary

mineral collections. Techniques such as attenuated total reflectance (ATR), diffuse

24

reflectance (DRIFT) and Raman microanalysis are often used for the study of

minerals and other crystalline or semi-crystalline materials. Data analysis is often

fraught with danger and a number of critical questions aris e. What bands are

applicable to which molecular species? How many bands do we expect for each

molecule? And how do differences in crystal space group affect the vibrational

spectra of the mineral? For these questions to be answered, factor group analysis

must be employed. Factor group analysis of simple free molecules is often easy, but

the extension into crystalline materials is frequently troublesome. The following

sections offer a review of factor group analysis of simple molecules and then an

extension to crystalline lattices. These sections discuss symmetry elements, factor

group analysis of free molecules, site symmetry and factor group analysis of solids, as

well as the assignment of vibrations to molecular groups and symmetry species.

2.2. Factor Group Analysis

Factor group analysis is a method used for determining symmetry of vibrations

[6, 20-27]. It is useful for determining how many bands are expected of each type of

molecule in the vibrating unit and is also useful for ascertaining differences in spectra

of polymorphous materials. The variation in the spectra of polymorphs are due to the

difference in the symmetry of the polymorph [6, 20-27]. The symmetry elements and

the combination of symmetry elements is the crux of factor group analysis. An atom

in a vibrating unit, a molecule or crystal, may be thought of as having three degrees of

movement, namely x, y, and z directions. From the combination of the symmetry

elements of each vibrating unit the movement of the atoms, which results in an overall

25

change in the dipole moment (infrared spectroscopy) or polarisablity of the electron

cloud (Raman spectroscopy), may be inferred.

2.3. Elements of symmetry

To gain an understanding of factor group analysis it is necessary to first obtain

an understanding of elements of symmetry. It is these elements of symmetry and the

addition of them that predicts the normal modes of vibration of a molecule in a

crystalline lattice. Symmetry elements such as inversion centres, mirror planes,

rotation axis and the addition of these to get the overall symmetry species will be

looked at in this section. A symmetry element is an operation that once applied to a

shape leaves that shape unchanged after the operation [6, 21, 28]. One of the simplest

symmetry elements is possibly the mirror plane, an example of which is human hands.

If both hands are placed together with small fingers touching it is easy to see that if a

mirror is placed between the hands, one hand reflects onto the other hand. Each digit

on the hand has an identical image on both the left and right of the image. This is also

applicable to molecules such as the water molecule; a mirror plane may be drawn

down the centre of the molecule bisecting the oxygen and the angle of the two

hydrogens. Another mirror plane, which bisects all atoms, is also present, as shown in

Figure 1.

26

Figure 1 Symmetry elements in the water molecule

Another symmetry element is the rotation axis. An example of a rotation axis

is shown in Figure 1, where the water molecule may be rotated about the oxygen.

Other more complex images such as the tetrahedron also have rotation axis. Each

apex of the tetrahedron has a three fold rotation axis. If a shape has no symmetry

operation applicable to it the symmetry element is assigned to a C1 rotation axis and

represents the lowest symmetry possible. Inversion centres are centres of symmetry

where all elements are able to be transposed through the centre of symmetry to

achieve the same image as before. An example of an inversion centre is shown in

Figure 2.

Figure 2 Inversion centre

27

It can be seen that if all points are transposed through the origin to the other

side along the axis, the same shape is achieved. Elements of symmetry are able to be

combined to achieve a set of possible symmetry operations of a factor group. This

factor group will be discussed further in the following sections.

2.4.Prediction of the number of normal modes of vibration

The number of normal modes of vibration for a crystal or molecule may be

calculated from the number of degrees of freedom of it has to vibrate. This is related

to the number of atoms in the crystal or molecule. Since there are three degrees of

freedom for an atom to vibrate, there are 3N modes of motion possible where N is the

number of atoms in the molecule or crystal. As the whole crystal or molecule may

translate in three directions, then these must be subtracted from the total. Molecules

may also rotate about three directions for bent molecules, or two directions for linear

molecule, thus three or two modes must be deducted respectively. This allows the

following formula for the calculation of the normal modes of vibration of a molecule:

3N-5 Linear molecules

3N-6 Bent molecules.

Crystals have an added confusion, as the contents of the unit cell may not

reflect the content of the Bravais unit cell. Only the number of atoms in the Bravais

unit cell may be used to predict the number of normal modes of vibration of a crystal.

The number of atoms in the Bravais cell can be calculated by the formula:

LPZ

N =

Where N is the number of atoms in the Bravais unit cell, Z is the number of atoms in

the unit cell, and LP is the number of lattice point as determined from Table 1.

28

Table 1 Unit cell type and number of lattice points

Type A,B,C or I F P R LP 2 4 1 3 or 1*

* This lattice may have been reduced to primitive form

From the number of atoms in the Bravais cell the number of modes of

vibrations of a crystal may be determined by the formula:

3N-3 where N is the number of atoms in the Bravais unit cell. Only the three translational modes are subtracted in the case of a crystal as the

rotation of an atom or a molecule is not the same as the rotation of a neighbouring

group or atom.

2.5.Point group analysis of free molecules

The water molecule is one of the simplest cases which contains a combination

of symmetry elements. The water molecule has three atoms - one oxygen bound to

two hydrogens – at an angle of ~120° (Figure 1). A line can be drawn down the

middle of the oxygen, bisecting the angle between the two hydrogens. If the molecule

is rotated around this axis, there are two positions, which produce the same image as

the original starting position. This line represents a two fold rotation or commonly

referred to as the C2 axis and a rotation of 180° about the C2 axis is called a C2

symmetry operation. There are also two more symmetry elements in the water

molecule, the mirror planes. The first plane is the plane of the atoms; the second plane

is normal to the first and intersects it along the C2 axis. These mirror planes reflect the

atoms on one side of the molecule directly onto the other side. These are the Cs

planes, where s represents a mirror plane. Once the symmetry operations for the

29

molecule are known, the point group may be determined by following the flow chart

below (Figure 3). In this case, it may be deduced that the molecule has C2v symmetry.

The next step is to use the point group to assign the symmetry species of the

vibrations of the molecule. Consider the water molecule in terms of distinct elements;

the oxygen, which is sitting on a C2 axis and two mirror planes, and two identical

hydrogens sitting on a mirror plane. There are rules of convention governing the

assignment of the x, y, z axes [28].The z axis in vibrational spectroscopy is considered

the most important axis, which is commonly the axis that intersects the most atoms

with the most symmetry [28]. In the water molecule the z axis corresponds to the C2

axis. In this case the oxygen atom is sited on the highest point of symmetry [28]. The

x axis is now chosen to be perpendicular to the molecular plane and the y axis to be in

the molecular plane [28].

30

Figure 3 Flow chart for determination of point group [28]

Having assigned the axes, correlation tables such as those in Farmer may be

consulted to find the symmetry species of the vibrations of the molecules [11]. The

correlation table for C2v is shown below in Table 2. The first column shows the

highest symmetry group. In this particular case it is the C2v column that is the highest

symmetry species present. The letters in this column represent the modes, which are

available to the molecule to vibrate. Each distinct atom (with respect to symmetry and

atomic type) is considered for its particular symmetry elements. Consulting the

correlation for oxygen (C2v), it can be seen that there is displacement for the oxygen

in the A1, B1 and B2 symmetry operations. The hydrogens can be seen to be on

yzsC elements allowing for two displacements in the A1 mode, one in the A2 mode, one

for B1 and two for the B2 mode. The addition of these modes gives the reducible

31

representation given by in this case 3A1 + A2 + 2B1 + 3B2. However six of these

modes describe the modes of vibration which describe rotations and translations. In

Table 2 it can be seen that translations (T) and rotations (R), occur as the modes A1

(Tz), A2 (Rz), B1 (Tx,Ry), and B2 (Ty, Rx). The available modes of vibration of the

water molecules can then be obtained by the subtraction of the translation and

rotations possible. This is summarised in Table 3.

Table 2. Correlation table for the C2v point group

Table 3. Factor group analysis of the water molecule.

Available modes Elements of symmetry A1 A2 B1 B2

Displacements of the O atom C2v 1z 0 1x 1y Displacements of the H atoms Cs

yz 2p 1x 1x 2p Rotations and translations Tz Rz Tx,

Ry Ty, Rx

Available vibrations of the water molecules

2A1 - - 1B2

The irreducible representation of the water molecule is Γ = 2A1 + B2, which

predicts 3 modes for water. The 3N-6 rule discussed earlier allows for a check of the

number of modes possible, in this case confirming the prediction of three modes of

vibration for water. In Table 3 it is seen that translations apply to the normal modes

for the water molecule. This indicates that the dipole moment oscillates in the same

32

direction as the translation allowing for infrared activity. In addition it can be seen

that the polarisability tensor has modes applicable for the normal modes of water.

Thus, the modes will be Raman active also.

2.6. Application of the internal modes to the crystal structure

The discussion to date has dealt with the vibrational modes available to

isolated molecules. For the extension of factor group analysis to crystalline solids an

understanding of crystallography is required. Crystalline lattices exist as a specific

structure for which the atoms may only exist on only specific sites. There exist some

230 space groups which relate to 33 crystallographic space groups. Each specific

space group has a finite number of sites with specific symmetry. The “International

Table For Crystallography” has the relative co-ordinates for each symmetry site

contained in the unit cell which may be used in conjunction with single crystal XRD

to determine the identity of the atoms on each site. For the use of the XRD results in

vibrational spectroscopy the following table (Table 4) is needed to translate from

Hermann-Mauguin notation (XRD, crystallography) to Schoenflies notation

(spectroscopic) [11, 32].

33

Table 4 Relationship between Hermann-Mauguin and Schoenflies notation

Hermann-Mauguin

Schoenflies Hermann-Mauguin

Schoenflies

1 C1 422 D4

1_

Ci, S2 4mm C4v

2 C2 3_2m D2d

m Cs, S1

m

4mm

D4h

2/m C2h 6 C6 222 D2

6_

C3h

mm2 C2v m

6 C6h

mmm D2h 622 D6 3 C3 6mm C6v

3_

S6 6_

m2 D3h

32 D3 m

6mm

D6h

3m C3v 23 T

3_m D3d m3 Th

4 C4 432 O

4_

S4

4_

3m Td

m

4 C4h

m3m Oh

Once the site point group has been determined it is necessary to check that

there is no choice between the similar symmetry species. For instance, the cases of the

water molecule there is a choice for the Cs mirror plane in the yx or xz directions. In

this case the Wyckoff letter in combination with the Schoenflies superscript, obtained

from the “International Table For Crystallography”, will yield the answer. In this

instance xzsC are sites 1ef, 11d, 14c, 15c, 18d, 20c and the yz

sC are the 1gh, 2ab, 4c,

7a, 11e, 12a, 14de, 16b, 18c, 20d, and 22b (Table 2). [6]

34

Once the sites have been unequivocally determined, the factor group splitting

can be determined. Appropriate correlation tables can be consulted to determine how

the site symmetry reduces the internal modes to lower symmetry modes. If vibrational

modes are degenerate such as E and T modes, the site symmetry may split the modes

into lower symmetry species allowing multiple modes for the degenerate modes. A

crystal symmetry correlation table, such as Table 5, (Factor Group) can be used to

determine the splitting of the site mode which will result in splitting of these modes.

The occurrence of more than one of the identical molecules (with relation to

symmetry and atom type, could result in coupling). In the case of water molecules

(C2v) sitting on a C1 site in a C2h crystal, such as gypsum, the A1 and B2 mode form A

modes on C1 site and split to Ag, Bg, Au, and Bu modes in the C2h crystal as shown in

Table 6. This would suggest that in this case there is the occurrence of four of the

same molecules in the lattice, which will result in the splitting of A modes. It is also

noted that this results in a symmetric characteristic (Ag modes, vibrating in phase) and

an anti-symmetric mode (Bg, vibrating out of phase).

Table 5. Correlation table for the C2h Point group

35

Table 6 Factor group splitting for a water molecule on a C1 site in a C2h crystal

Free

C2v

Site

C1

Crystal

C2h

2A1

B2

3A

3Ag

3Bg

3Au

3Bu

The 3N-3 rule may then be applied to determine if all modes of vibration have

been described. In the case of crystals, the contents of the Bravis lattice must be

considered and not the total unit cell contents as discussed in the previous sections.

Taking this into account, three modes are available to the free water molecule, but in

the case above there are 4 water molecules in the Bravis cell and hence each mode

splits into 4, allowing 12 modes of vibration. All the modes of vibration determined in

this section have been concerned with the vibration of the molecule, the so-called

internal modes of vibration. There are also the vibrations of the molecule as a whole,

vibrating within the crystalline lattice.

2.7. Lattice modes

Lattice modes are dealt with by a similar mechanism to that of the vibrations

of molecules. The main difference between the factor group analysis of lattice

vibrations and internal modes, is that lattice modes are considered as the movements

of groups of atoms in the crystalline lattice [21, 25]. For instance, gypsum

CaSO4.2H2O, the water and sulphate molecules would be considered as free entities

then subjected to the symmetry constraints of the site, then the crystal symmetry. It is

then necessary to consider the motions of the sulphate group, the water molecules, as

36

well as the motion of the calcium atoms in the crystal. To consider the factor group

analysis of lattice modes it is necessary to know the site symmetry of each group [21,

25]. It is necessary to know the modes available for the molecule to translate and

rotate. While in the case of atoms, only translations need be considered due to their

rotational symmetry, it is however necessary to consider translations and rotations in

the case of molecules [25]. The appropriate modes may be deduced from correlation

tables. In the case of water molecules on C1 sites there are 6A modes available for the

water molecule to rotate and translate. This is then related to the C2h crystal group.

This predicts the formation of the reducible representation 6Ag + 6Bg + 6Au + 6Bu.

Each group of atoms is considered in turn to produce the active lattice vibrations

which are added to achieve the lattice modes of the crystal. It is again necessary to

consider the three modes that represent the translations of the whole crystal, which are

not active vibrations. These modes are subtracted to give the irreducible

representation 5Ag + 4Bg + 5Au + 4Bu.

2.8.Assignment of spectra, Selection Rules, Polarised Orientated Vibrational

spectroscopy

Assignment of spectral bands to various symmetry species can be achieved via

polarised single crystal Raman and polarised single crystal infrared investigations [20,

21, 25]. The symmetry species of the vibrations can be ascertained by orientating the

crystal of a mineral along different polarisations of the incident radiation. For Raman

measurements the polarisability tensor of the vibration is different along different axis

of the crystal [20, 21, 25]. The probability of a Raman transition is proportional to

[20, 21, 25]:

37

?? oa? i dt Where ?o is the vibrational wave function, a is the polarisability of the

molecule, ? i is the vibrational wave function, and dt implies the integration is carried

over all possible variables of the wave functions. The polarisability a is the ease with

which the electron cloud in the molecule may be distorted. This quantity is a tensor, a

3X3 array of components, i.e. axx, axy, axz, ayx, ayy, ayz, azx, azy, and azz, where akl = alk.

In the case where the integral is zero, then there is zero probability that the transition

will occur [20, 21, 25]. The integral will only be non zero in the case where the

product ? oa? is totally symmetric, that is +1. This can only be achieved when akl has

the same symmetry properties as ? i [21, 25]. The symmetry properties of akl are the

same as the symmetry properties of kl. Thus, if the normal mode has the same

symmetry as one of these binary combinations of x, y, and z, then a transition from the

ground state will be Raman active [21, 25]. A typical form of a polarisability tensor is

shown in Chapters 3 and 4. If a character is shown, then a transition along the axis

may be allowed. Modes of vibration that are not allowed will be of zero intensity or

very low intensity. Orientated single crystal Raman studies, with Raman spectra taken

of all possible orientations and polarisations possible, will yield the symmetry species

of the vibrations.

The polarisability tensors for a vibration are calculated along the spectroscopic

axis [20]. Thus it is necessary to know the relationship of the crystallographic axis in

relation to the spectroscopic axis. This can be achieved by the knowledge of the

translations of a spectroscopic group, available in correlation tables. In this way a

spectroscopic axis of translation can be seen to move in a direction associated with a

mode of vibration. From the correlation tables these modes can be related to the site

and space groups to determine the spectroscopic axis with relation to the crystal axis.

38

From an atomic coordinate model, the relationship between the spectroscopic and

crystallographic axis can be deduced. The spectroscopic axis can also be deduced

from the correlation tables, by looking at the symmetry species and Schoenflies /

Wyckoff letter if the space group has choices for the definition of the same symmetry

element.

Once the symmetry species of the bands are identified there is a possibility of

further assignment from deuteration, isotopic and analogous series studies (cationic

and anionic series) through the use of Hooke’s Law. Satellite bands are known to exist

in cerussite, with the intensity of these bands about 1% of the corresponding ν1 band

intensity[29-31]. This satellite is due to isotopic 18O substitution in the carbonate

group [29-31]. The calculated frequencies of this vibration are, from Equation 1, for

cerussite 1030, aragonite 1058, witherite 1034, and strontianite 1044 cm-1, which are

in good agreement with the experimental results [32].

V =µ

kcp

o

21

(Equation 1)

µ= reduced mass of participating atoms ko = force constant V = vibrational frequency in wavenumbers c= speed of light

Studies have shown a strong correlation between OH stretching frequencies

and both the O…O bond distances and the H…O hydrogen bond distances. [33-36] The

significant work of Libowitzky [37] showed that a regression function can be

employed, relating the above correlations with regression coefficients better than 0.96

[37]. The function is ?1 = 3592-304x109exp(-d(O-O)/0.1321) cm-1. Minerals where

39

two or more types of OH units are identified in the structure, the known hydrogen

bond distances may be used to calculate the hydroxyl stretching frequency. From this

predicted hydroxyl stretching frequency further assignment of the spectra may be

made.

Some vibrations that are treated as lattice vibrations such as hydroxyl

stretching vibrations, therefore may be hard to differentiate these vibrations from the

other lattice modes. Vector analysis may be employed to gain insight into these bands.

Vector analysis considers vibrations of a molecule as a vector. Interactions of this

vector with other vectors of an identical nature are deduced from the factor group. For

the below example of olivenite, it is unclear how many vibrations are expected from

the hydroxyl groups. As the hydroxyl groups are sitting on C1 sites there is only one

vector that the hydroxyl is allowed to vibrate on their site symmetry (A). When this is

allowed to split in the factor group there are 1Ag, 1Au, 1Bg, and 1Bu modes predicted,

of which, g modes are Raman active and u modes are infrared active. Olivenite does

indeed have two hydroxyl bands in the Raman [16]. This implies that in the structure

of olivenite there are four hydroxyls which couple to produce the four modes.

2.9. Birefringence

Birefringence is a well known hindrance to single crystal vibrational studies.

Birefringence is manifested as the differences in the refractive indices down each

independent optical axis. As such, there is a difference in the velocity of light down

these axes, hence a change in the resulting polarisation of the incident light. Erythrite

is known to be birefringent with refractive indices of a = 1.625, ß=1.661, and ? =

40

1.691 [38]. The optical axes or pseudo-optical axes, are orthogonal on all axes except

on the 010 face, in which the optical axis make a 30° angle with the c crystal axis

[39]. The amount of depolarisation of incident radiation on a birefringent crystal can

be calculated from the equation:

L = 100[cos2 ? – sin 2(G- ? ).sin 2G.sin2 (?/?).180] [39]

where L is the amount of radiation for which the plane of polarization is rotated by

90°, G is the angle the optical axis makes with the incident radiation, ? is the angle

between the analyser and polariser, ? is the difference between the refractive indices

multiplied by the thickness of the sample and ? is the wavelength of the incident

radiation [39]. This equation predicts that along the 010 face of a crystal of erythrite

incident polarised radiation is 28 % scrambled. This prevents the collection of reliable

oriented single crystal data from the 010-crystal face.

2.10. Porto notation

Porto notation is a way of expressing the orientation of the crystal with

reference to the polarisation of the laser in both the excitation and analysing directions

[25]. This notation consists of a four letter code expressing the direction of the

propagation of the incident radiation, the polarisation of the incident radiation, the

polarisation of the analysed radiation and the direction of propagation of the analysed

radiation. For example BACB, the first letter depicts the axis for the propagation of

the laser light, the second depicts the orientation of the polarisation direction i.e. the

direction that the laser is vibrating in the excitation, the third is the vibrational

41

direction of the Raman scattered photons that are being analysed and the last the

direction of propagation of the analysed laser beam [25]. This is all done with

reference to the crystal that is being analysed. An example is shown in Figure 4.

Figure 4 A example of Porto notation.

2.11. Olivenite Example

Olivenite is an olive-green copper arsenate mineral, and as with all secondary

copper arsenate minerals, it is formed from the oxidisation of native copper ore in

arsenate containing ground waters [16]. Crystals of olivenite may be short prismatic

to acicular; globular, reniform or nodular, fibrous structure; massive, granular to

earthy. Olivenite has the formula, Cu2(AsO4)OH, with four structural units per unit

cell. The crystal structure of olivenite is monoclinic, P21/n, with the Schoenflies

notation for this space group being C2h5 [16, 38]. From the atom coordinates, Figure 5,

it can be deduced that all atoms in the structure are on C1 sites. There are nine atoms

42

in a formula unit with four formula units per unit cell predicts 3N–3 (3*4*9 – 3 =

105) vibrations. The arsenate ion has the free symmetry of Td, with the arsenic atom

siting on all elements of symmetry, and all oxygen’s siting on a mirror plane and a

three fold rotation axis. From character tables, Table 7 may be constructed for the

arsenate ion.

Figure 5 Atom positions of each atom in the olivenite crystal structure.

Table 7. Factor group analysis of the free arsenate ion.

Available modes Elements of symmetry A1 A2 E F1 F2

Displacements of the As atom Td 0 0 0 0 1 Displacements of the O atoms C3v 1a 0 1p 1p 2

Rotations and translations R T Available vibrations of the arsenate ion A1 - E - 2T2

43

The arsenate vibrations can then be split to the site and space group in

accordance with Table 8. This shows that the A1 + E + 2T2 modes of the arsenate split

to A modes in the C1 site and split to 9Ag +9Bg + 9Au + 9Bu modes in C2h. The Ag and

Bg modes will be Raman active while the Au and Bu modes will be infrared active

[16].

Table 8. The factor group splitting of the arsenate ion in the olivenite structure.

Td C1 C2h A1 E

2T2

9A

9Ag 9Bg 9Au 9Bu

As all atoms in the crystal lattice have site symmetry of C1, only A modes are

available for the rotations and translations of the arsenate, the copper, and the

hydroxides. From this, construction of Table 9 can be undertaken to determine the

modes of vibration in the factor group.

Table 9. Factor group analysis of the lattice modes in olivenite

Site group Crystal C1

18A

52hC

18Ag 18Au

18Bg 18Bu

Γ = 18Ag + 18Bg + 17Au+ 16 Bg

69 Vibrations

It may be calculated that 105 normal modes of vibration are predicted for the

olivenite crystal from the 3N-3 rule. From the factor group analysis it has been

calculated that 105 vibrations are predicted, 36 internal modes of the arsenate ion and

44

69 lattice modes [16]. As both the 3N-3 rule and factor group analysis predicts the

same number of vibrations then all modes of vibrations have been calculated.

Once the symmetry species of the normal modes of vibration have been

determined, is necessary to determine the activity of these modes. Inspection of the

relevant correlation tables reveal that the Ag modes are active in Raman spectra of

XX, YY, ZZ and XY nature (Table 5). The Bg modes are active in the YZ, ZX spectra

(Table 5). These tensor elements are along the spectroscopic axes, and therefore great

caution must be practiced to ensure that the relationship between the spectroscopic

and crystallographic axes are known. In the case of the monoclinic cell the b

crystallographic axis is usually the unique axis which relates to the z spectroscopic

axis. Further to this, the relation of the optical axis to the spectroscopic axis must be

also checked to ensure that no birefringence occurs. In monoclinic cases,

misalignment of the optical and spectroscopic axis usually occurs in spectra excited

down the b axis. Other examples of factor group analysis are shown in Chapters 3 and

4.

2.12. Vivianite minerals and solid solutions

Vivianite group of minerals have the general formula A32+(XO4)2.8H2O, where

A2+ may be Co, Fe, Mg, Ni, Zn and X may be P or As [38]. Vivianite minerals are

monoclinic with point group 2/m [38] and two formula units per unit cell [38, 40].

The vivianite group consist of the minerals[38, 40, 41]:

45

Annabergite Ni3(AsO4)2.8H2O

Arupite Ni3(PO4)2.8H2O

Baricite (Mg,Fe)3(PO4)2.8H2O,

Erythrite Co3(AsO4)2.8H2O,

Hörnesite Mg3(AsO4)2.8H2O,

Köttigite Zn3(AsO4)2.8H2O,

Parasymplesite Fe3(AsO4)2.8H2O

Vivianite Fe3(PO4)2.8H2O.

These minerals are closely related to bobierrite Mg3(PO4)2.8H2O, and

manganese-hörnesite (Mn, Mg)3(AsO4)2.8H2O which have a b-axis twice of those in

the vivianite group [38, 40, 41].

Vivianites form solid solutions of mixed metal and mixed phosphates with

most metal ions producing complete solid solution mixing curves [42-55]. Erythrite

has been used for the preparation of pigments and glazes due to its brilliant pink

colour [56-61], with the remediation of arsenates from soils also receiving attention

[62-64]. The thermal transformations of vivianites has recently been studied with

Raman, infrared, and infrared emission spectroscopy [65-67].The infrared spectrum of

vivianite minerals has been previously determined [6, 68, 69]. Single crystal x-ray

diffraction studies have determined the crystal structure of erythrite [40]. The unit cell

contains 2 formula units in the space group 32hC with cell parameters of a = 10.251 Å,

b = 13.447 Å, c = 4.764 Å, and ß = 104.98° [38]. There are two sites occupied by two

independent cobalt atoms (C2h, C2 site symmetry), with one independent arsenate ion

(Cs site symmetry) and two independent water molecules (Ci site symmetry)[70].

46

Minerals of the vivianite group crystallize in the form of M (1)O2(H2O)4 octahedra

and M (2)2O6(H2O)4 double octahedral groups (where M is the metal), which are

linked via XO4 tetrahedra to complex sheets in (010), further interconnected by

hydrogen bonds only[70]. This structure is shown for erythrite in Figure 6.

Figure 6 The atomic structure of erythrite

Mixed metal vivianite samples, of chemical compositions

Co2.01Fe0.74Ni0.25(AsO4)2.8H2O for erythrite and Ni2.48Mg0.50Fe0.02(AsO4)2.8H2O for

annabergite, have been studied by energy dispersion x-ray spectroscopy and

microprobe analyses [70]. The M2+ cation distribution on the M(1) and M(2) sites in

erythrite and annabergite have been investigated by Moessbauer spectroscopy and site

occupancy refinements [70]. Magnesium substituted annabergite showed a strong

preference for Mg2+ to occupy the Ni(2) sites over the Ni(1) sites [70]. Moessbauer

spectroscopy of erythrite reveals an analogous preference of Fe2+ on the Co(2) site

[70]. Further evidence for the ordering of Mg and Ni site has also been demonstrated

in cabrerite, a currently disregarded mineral name[45]. Cabrerite has a monoclinic

47

space group C2/m, with a = 10.211, b = 13.335, c = 4.728 Å, and ß = 104.97° and Z =

2. Significant ordering of the Ni-Mg in the octahedral sites occurs. Single octahedra,

M(1), and double octahedral groups, M(2)::: M(2), are connected by AsO4 tetrahedra

to form complex sheets parallel to (010) [45]. Magnesian nickeloan erythrite has been

found with composition (Co0.54Mg0.29Ni0.15Zn0.02)(AsO4)2.8H2O giving an indication

of the complexity of solid solutions available to the vivianite minerals [62].

Furthermore, vivianite samples from the Big Chief pegmatite mine, Glendale,

Pennington Co., South Dakota, USA were found to contain oxide contents of FeO

38.9 %, MnO 4.2 %, NiO 0.13 %, CaO 0.5 %, P2O5 28.4 %, H2 O (28.7), with minor

As2O5, Na2O, ZnO and CuO [71]. This indicates that even metal ions that do not form

vivianite structures may substitute in the lattice of vivianite minerals. Due to this fact,

the vivianite group of minerals is an excellent choice for the study of solid solutions.

48

2.13. References

1. Klein, C., Mineral Science. 22 ed. 2002, New York: Wiley. 2. Korbel, P. and M. Novak, The Complete Encyclopaedia Of Minerals. 2001,

U.K.: Grange Books. 3. Bishop, A.C., A.R. Woolley, and W.R. Hamilton, Philip's Minerals, Rocks

and Fossils. 3rd ed. 1999, London: Octopus Publishing Proup. 4. Hunt, G., Salisbry. Mod. Geol., 1971(2): p. 23. 5. Frost, R.L., J. Kloprogge, and J. Schmidt, Non-destructive identification of

minerals by Raman microscopy. The Internet Journal of Vibrational Spectroscopy. 3(4).

6. Farmer, V.C., ed. The Infrared spectra of minerals. 1974, Mineralogical Society: London.

7. Brooke, C., H. Edwards, and J. Tait, Journal of Raman Spectroscopy, 1999(30): p. 429.

8. Burgio, L., Ciomartan, and R. Clark, Journal of Raman Spectroscopy, 1997(29): p. 79.

9. Martens, W. and R.L. Frost, An infrared spectroscopic study of the basic copper phosphate minerals: Cornetite, libethenite, and pseudomalachite. American Mineralogist, 2003. 88(1): p. 37-46.

10. Martens, W.N., et al., The Basic copper arsenate minerals olivenite, cornubite, cornwallite, and clinoclase: An infrared wmission and Raman spectroscopic study. American Mineralogist, 2003. 88.

11. Frost, R.L., et al., Vibrational spectroscopy of the basic copper phosphate minerals: pseudomalachite, ludjibaite and reichenbachite. Spectrochimica acta. Part a, molecular and biomolecular spectroscopy, 2002. 58(13): p. 2861-8. FIELD Reference Number: FIELD Journal Code:9602533 FIELD Call Number:.

12. Frost, R.L., et al., Vibrational spectroscopy of the basic copper phosphate minerals: pseudomalachite, ludjibaite and reichenbachite. Spectrochimica Acta, Part A: Molecular and Biomolecular Spectroscopy, 2002. 58(13): p. 2861-2868.

13. Frost, R., L., et al., Vibrational spectroscopy of the basic copper phosphate minerials: pseudomalachite, ludjibaite, and reichenbachite. Spectrochimica Acta. Part A, 2002. 58: p. 2861.

14. Frost, R.L., et al., Raman spectroscopy of the polyanionic copper(II) minerals buttgenbachite and connellite: implications for studies of ancient copper objects and bronzes. Journal of Raman Spectroscopy, 2002. 33(9): p. 752-757.

15. Frost, R.L., et al., Raman spectroscopy of the basic copper phosphate minerals cornetite, libethenite, pseudomalachite, reichenbachite and ludjibaite. Journal of Raman Spectroscopy, 2002. 33(4): p. 260-263.

16. Frost, R.L., W.N. Martens, and P.A. Williams, Raman spectroscopy of the phase-related basic copper arsenate minerals olivenite, cornwallite, cornubite and clinoclase. Journal of Raman Spectroscopy, 2002. 33(5): p. 475-484.

17. Frost, R.L., et al., Raman spectroscopic study of azurite and malachite at 298 and 77 K. Journal of Raman Spectroscopy, 2002. 33(4): p. 252-259.

18. Frost, R.L., et al., Raman spectroscopy of the basic copper chloride minerals atacamite and paratacamite: implications for the study of copper, brass and

49

bronze objects of archaeological significance. Journal of Raman Spectroscopy, 2002. 33(10): p. 801-806.

19. Frost, R.L., et al., Analyst, 2002: p. 127. 20. Turell, G. and J. Couset, Raman Microscopy: Development and Applications.

1996: Academic Press. 21. Davidson, G., Introductory Group theory for chemists. 1971, London: Elsevier

Publishing Co. 22. Coleyshaw, E.E., W.P. Griffith, and R.J. Bowell, Fourier-transform Raman

spectroscopy of minerals. Spectrochim. Acta, Part A, 1994. 50A(11): p. 1909-18.

23. Brown, A. and E.R. Clark, Infrared analysis of geological materials. J. Geol. Educ., 1980. 28(2): p. 92-5.

24. Ferraro, J.R., Factor group analysis for some common minerals. Appl. Spectrosc., 1975. 29(5): p. 418-21.

25. Nakamoto, K., Infrared and Raman spectra of inorganic and coordination compounds: Part A: theory and applications in inorganic chemistry. 5 ed. 1997, New York: Wiley-Interscience.

26. Plavsic, B., S. Kobe, and B. Orel, FTIR spectroscopy as a method for determination of the crystallization forms of calcium carbonate. Zb. Ref. Posvetovanja Slov. Kem. Dnevi, 1997: p. 487-492.

27. Plavsic, B., S. Kobe, and B. Orel, FTIR spectroscopy as a method for the study of the different crystallization forms of CaCO3. Key Eng. Mater., 1997. 132-136(Pt. 1, Euro Ceramics V): p. 268-270.

28. Salthouse, J.A. and M.J. Ware, Point group character tables and related data. 1972, London: Cambridge University Press.

29. Farmer, V.C., The Infrared Spectra of Minerals, ed. V.C. Farmer. Vol. 4. 1974, London.

30. Sterzel, W. and E. Chorinsky, Effect of heavy carbon isotopes on the infrared spectrum of carbonates. Spectrochim. Acta, Part A, 1968. 24(4): p. 353-60.

31. Cloots, R., Raman spectrum of carbonates MIICO3 in the 1100-1000 cm-1 region: observation of the .nu.1 mode of the isotopic carbonate (C16O218O)2- ion. Spectrochim. Acta, Part A, 1991. 47A(12): p. 1745-50.

32. Elderfield, H. and R. Chester, Effect of periodicity on the infrared absorption v2 frequency of anhydrous normal carbonate minerals. Amer. Mineral., 1971. 56(9-10): p. 1600-6.

33. Emsley, J., Very strong hydrogen bonding. Chemical Society Reviews, 1980. 9: p. 91-124.

34. Lutz, H., Hydroxide ions in condensed materials - correlation of spectroscopic and structural data. Structure and Bonding (Berlin, Germany), 1995. 82: p. 85-103.

35. Mikenda, W., Stretching frequency versus bond distance correlation of O-D(H)...Y (Y = N, O, S, Se, Cl, Br, I) hydrogen bonds in solid hydrates. Journal of Molecular Structure, 1986. 147: p. 1-15.

36. Novak, A., Hydrogen bonding in solids. Correlation of spectroscopic and crystallographic data. Structure and Bonding (Berlin), 1974. 18: p. 177-216.

37. Libowitsky, E., Correlation of the O-H stretching frequencies and the O-H...H hydrogen bond lengths in minerals. Monatschefte fÜr chemie, 1999. 130: p. 1047-1049.

38. Anthony, J.W., et al., Arsenates, Phosphates, Vanadates. 1 ed. Handbook of Mineralogy. Vol. 4. 2000, Tucson, Arizona: Mineral Data Publishing.

50

39. Bloss, F.D., Optical Crystallography. 1999, Washington, DC: The Mineralogical Society of America.

40. Wildner, M. and G. Giester, Structure and crystal chemistry of vivianite-type compounds: crystal structures of erythrite and annabergite with a mossbauer study of erythrite. European Journal of Mineralogy, 1996(8): p. 187-192.

41. Mandarino, J.A., Fleischer's Glossary of Mineral Species. 8 ed. 1999, Tuscon: The Mineralogical Record Inc.

42. Al-Borno, A. and M.B. Tomson, The temperature dependence of the solubility product constant of vivianite. Geochim. Cosmochim. Acta, 1994. 58(24): p. 5373-8.

43. Amthauer, G. and G.R. Rossman, Mixed valence of iron in minerals with cation clusters. Phys. Chem. Miner., 1984. 11(1): p. 37-51.

44. Ermolaev, M.I., et al., Preparation of water-soluble cobalt and nickel compounds from erythrin-annabergite minerals. Zh. Prikl. Khim. (Leningrad), 1977. 50(9): p. 1920-2.

45. Giuseppetti, G. and C. Tadini, The crystal structure of cabrerite, (Ni,Mg)3(AsO4)2.8H2O, a variety of annabergite. Bull. Mineral., 1982. 105(4): p. 333-7.

46. Jambor, J.L. and J.E. Dutrizac, Solid solutions in the annabergite-erythrite-hoernesite synthetic system. Can. Mineral., 1995. 33(5): p. 1063-71.

47. Kurnakov, N.S. and I.A. Andreevskii, Solid solutions of water and oxygen in the phosphates of the vivianite group. Ann. inst. anal. phys. chim., 1924. 2: p. 485-7.

48. Manceau, A., A.I. Gorshkov, and V.A. Drits, Structural chemistry of manganese, iron, cobalt, and nickel in manganese hydrous oxides: Part I. Information from XANES spectroscopy. Am. Mineral., 1992. 77(11-12): p. 1133-43.

49. Merlina, F.E. and T.V. Krylova, Mineral forms of cobalt and nickel in arsenic-cobalt ore and phase analysis methods. Obogashch. Rud, 1972. 17(6): p. 36-8.

50. Pizarro, J.L., et al., Synthetic pathways to obtain phosphates and arsenates of Co(II) and Ni(II) related to minerals: magnetic properties. Solid State Ionics, 1993. 63-65(1-4): p. 71-7.

51. Schmetzer, K., W. Horn, and W. Bartelke, Cobalt- and nickel-containing koritnigite from Jachymov (Joachimsthal), CSSR - a second occurrence. Neues Jahrb. Mineral., Monatsh., 1980(6): p. 237-40.

52. Schmetzer, K., W. Horn, and O. Medenbach, Cobaltkoritnigite, (Co,Zn)[H2O|AsO3OH], a new mineral and pitticite, Fe2O3.cntdot.As2O5.cntdot.9-10H2O, an amorphous iron-arsenate-hydrate. Neues Jahrb. Mineral., Monatsh., 1981(6): p. 257-66.

53. Wolfe, C.W., Classification of minerals of the type A3(XO4)2.nH2O. Am. Mineral., 1940. 25: p. 738-54,787-809.

54. Yakhontova, L.K., et al., X-ray diffraction study of erythrite and its varieties. Konst. Svoistva Miner., 1974. 8: p. 36-40.

55. Yakhontova, L.K., A.P. Grudev, and A.A. Petrova, Composition and nomenclature of minerals of erythrite-annabergite series. Zap. Vses. Mineral. O-va., 1979. 108(3): p. 287-93.

56. Hartwig, J., Production and use of Safre or Zaffera (cobalt) and Smalte by glassmakers from 16th to 18th century. Verre (Versailles, France), 2001. 7(4): p. 40-48.

51

57. Tourangeau, P., Preparation of colored mineral powders by thermal treatment and materials colored therewith, in U.S. 2000, (Can.). Us. p. 2020 pp.

58. Markl, G., The color of minerals and their advantage. Aufschluss, 1999. 50(6): p. 337-341.

59. Tourangeau, P., Pigments obtained by treatment of minerals, in Fr. Demande. 1999, (Fr.). Fr. p. 159 pp.

60. Colored powders useful as pigments, their preparation and use, in Fr. Demande. 1998, (Tourangeau, Paulette, Fr.). Fr. p. 165 pp.

61. Tourangeau, P., Pigments manufactured from minerals, in Can. Pat. Appl. 1998, (Can.). Ca. p. 2074 pp.

62. Beyer, H., Minerals from the dumps of the stibnite veins of Ahrbrueck (Eifel), especially a find of cabrerite. Aufschluss, 1978. 29(11): p. 401-8.

63. Mann, H., et al., Retardation of toxic heavy metal dispersion from nickel-copper mine tailings, Sudbury District, Ontario: role of acidophilic microorganisms. II. Structure and microanalysis of bioprecipitants. Biorecovery, 1989. 1(3): p. 173-87.

64. Wappler, G. and G. Tischendorf, Identification of some secondary minerals from the dumps of the Friedrichsglueck Mine, near Neustadt/Rennsteig [East Germany]. Z. Geol. Wiss., 1980. 8(11): p. 1397-402.

65. Frost, R.L., et al., Thermal decomposition of the vivianite arsenates-implications for soil remediation. Thermochimica Acta, 2003. 403(2): p. 237-249.

66. Martens, W., R.L. Frost, and J.T. Kloprogge, Raman spectroscopy of synthetic erythrite, partially dehydrated erythrite and hydrothermally synthesized dehydrated erythrite. Journal of Raman Spectroscopy, 2003. 34(1): p. 90-95.

67. Frost, R., L., et al., Dehydration of synthetic and natural vivianite. Thermochimica Acta, 2002. In. Press.

68. Gevork'yan, S.V. and A.S. Povarennykh, Characteristics of the IR spectra of water molecules incorporated into phosphate and arsenate structures. Mineral. Zh., 1980. 2(1): p. 29-36.

69. Frost, R.L., et al., Raman and infrared spectroscopic study of the vivianite-group phosphates vivianite, baricite and bobierrite. Mineralogical Magazine, 2002. 66(6): p. 1063-1073.

70. Wildner, M., et al., Structure and crystal chemistry of vivianite-type compounds: crystal structures of erythrite and annabergite with a Moessbauer study of erythrite. Eur. J. Mineral., 1996. 8(1): p. 187-92.

71. Ritz, C., E.J. Essene, and D.R. Peacor, Metavivianite, Fe3(PO4)2.8H2O, a new mineral. Am. Mineral., 1974. 59(9-10): p. 896-9.

52

Chapter 3 – SINGLE CRYSTAL RAMAN SPECTROSCOPY OF CERUSSITE

Wayde N. Martens, Llew Rintoul, J. Theo Kloprogge, and Ray L. Frost

Wayde N. Martens Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

Llew Rintoul Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

J. Theo Kloprogge Inorganic Materials Research Program, Queensland

University of Technology, Queensland 4001, Australia.

Ray L. Frost Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

American Mineralogist (2004), 89, 352-358

53

3.1. Statement of contribution

Wayde N. Martens : Wrote the manuscript, experimental design, conducted

experiments, and data analysis.

Llew Rintoul : Aided experimental design, major editing, aided data analysis,

aided writing of manuscript.

J. Theo Kloprogge: Major editing, aided data analysis

Ray L. Frost : Editing, data analysis, supervision

54

3.2. Abstract

Raman and infrared active modes of cerussite have been assigned according to

their symmetry species, and compared to other aragonite group minerals. Small

satellite bands at 823 and 1031 cm-1 to the low wavenumber side of the fundamental

vibrations ?2 and ?1, respectively, have been assigned to the isotopic substitutions of

13C and 18O. The Raman active ?1 and ?2 carbonate modes are observed at 1051 and

835 cm-1. The absence of the B2g component of the ?1 and ?2 vibrations has been

explained by the small coupling between the Ag and B2g modes. The Raman active ?3

carbonate anti-symmetric stretching mode is observed at 1361(Ag), 1376 (B1g), 1419

(B3g) and 1477 (B2g) cm-1, while the corresponding infrared active bands are observed

at 1396, 1432, and 1456 cm-1. The Raman active ?4 carbonate bending mode is

observed at 673 (Ag), 668 (B2g), 681 (B1g), and 694 (B2g) cm-1. The corresponding

infrared bands are observed at 670, 679, and 698 cm-1. In both ?3 and ?4, the factor

group splitting between the B1g and B3g modes is 1 to 3 times smaller than the

separation of the Ag and B2g modes. Raman active lattice vibrations are detected at

120 (B3g), 132 (Ag), 148 (B1g), 152(B2g), 174 (B2g), 179 (B1g), 213 (Ag), 226 (B3g),

and 243 cm-1 (B2g). Corresponding infrared active bands are detected at 573, 543,

573, 423, 375, 290, 205, 165, 146, and 134 cm-1. Raman bands at 949, 966, 989, 1000

and 1104 cm-1 and at 922, 946, 967, 988, 996, and 1007 cm-1 in the infrared spectra

are assigned to combination and overtone bands. Raman bands at 1676 (Ag), 1689

(Ag), 1730 (B3g), and 1740 (B1g) cm-1 are ascribed to combination modes of ?1 + ?4

with bands at 2052 and 2092 cm-1 assigned to 2?1. Corresponding infrared bands are

observed at 1729 and 1740 cm-1 (?1 + ?3). Bands at 2359, 2409, 2471, and 2521 cm-1

are ascribed to ?1 + ?3, with broad bands at 1246 and 1323 cm-1 assigned to 2?4

modes.

55

56

57

58

59

60

61

62

Chapter 4 - Single Crystal Raman Study of Erythrite Co3(AsO4)2.8H2O

Wayde N. Martens, J. Theo Kloprogge, Ray L. Frost and Llew Rintoul

Wayde N. Martens Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

J. Theo Kloprogge Inorganic Materials Research Program, Queensland

University of Technology, Queensland 4001, Australia.

Ray L. Frost Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

Llew Rintoul Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

Anonymous referee’s comments: “The paper is easy to read, clear, and scientifically

convincing. It reports a careful crystallographic and spectroscopic study of single

crystals of mineralogical interest, with potential applications in identification of

pigments……”

Journal of Raman Spectroscopy (2004), 35, 208-216

63

4.1 Statement of contribution

Wayde N. Martens : Wrote the manuscript, experimental design, conducted

experiments, and data analysis.

J. Theo Kloprogge: Major editing, aided data analysis

Ray L. Frost : Editing, data analysis

Llew Rintoul : Aided experimental design, editing, aided data analysis, aided

writing of manuscript.

64

4.2 Abstract

Single crystal Raman and infrared spectra of natural and synthetic erythrite

Co3(AsO4)2.8H2O are reported, and compared to the spectra polycrystalline, synthetic

annabergite (Ni3(AsO4)2.8H2O) and hörnesite (Mg3(AsO4)2.8H2O). Factor group

analysis and single crystal considerations have been used to interpret the experimental

data. The Raman spectra of erythrite reveal ?1 arsenate stretching vibration at 850

cm-1 (Ag) with the corresponding infrared band at 821 cm-1 (Bu). The ?3

antisymmetric vibration is split into three components, observed at 796 (Ag), 788 (Ag)

and 803 (Bg) cm-1. The ?2 symmetric bending modes are observed at 375 (Ag) and 385

(Bg) cm-1. The ?4 bending modes are predicted to split into three bands which are

observed at 441 (Ag), 446 (Bg) and 457 (Ag) cm-1. Lattice vibrations are found at 112

(Ag), 124 (Bg), 145(Ag), 157 (Bg), 165 (Ag), 179 (Ag), 189 (Ag), 191(Bg), 201 (Bg),

210 (Ag), 227 (Ag), 250 (Ag), 264(Ag), 264 (Ag), 280 (Bg), 302 (Bg), 321 (Bg), and 338

cm-1 (Ag). Hydroxyl stretching modes are observed at 3050, 3218, 3333, 3449 and

3479 cm-1, in the infrared spectrum. Raman active hydroxyl bands are detected at

3009 (Bg), 3052 (Ag), 3190 (Bg) 3203 (Bg), 3281 (Ag) and 3310 (Bg), 3436 (Bg) and

3443 (Ag) cm-1. Infrared hydroxyl bands at 3050, and 3218 cm-1 are from water type

II, short hydrogen bonding distances, and the bands at 3449 and 3479 cm-1 are due to

water I, long hydrogen bonding distances. Water bending modes are detected in the

infrared spectrum at 1571, 1621, 1641, and 1682 cm-1, but due to the inherent weak

Raman scattering cross section of water these could not be detected in the Raman

spectra.

65

66

67

68

69

70

71

72

73

74

Chapter 5 – Synthetic deuterated erythrite – a vibrational spectroscopic study

Ray L. Frost, Wayde N. Martens, and J. Theo Kloprogge

Ray L. Frost Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

Wayde N. Martens Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

J. Theo Kloprogge Inorganic Materials Research Program, Queensland

University of Technology, Queensland 4001, Australia.

Spectrochimica Acta Part A (2003), 343

75

5.1 Statement of contribution

Ray L. Frost: Wrote the manuscript, data analysis.

Wayde N. Martens : Aided in writing manuscript, experimental design, conducted

experiments, Editing, data analysis

J. Theo Kloprogge: Editing, aided data analysis

76

5.2 Abstract

A comparison of deuterated and non-deuterated erythrite has been made using

a combination of infrared and Raman spectroscopy. Infrared shows bands at 3442,

3358, 3194 and 3039 cm-1. The band at 3442 cm-1 is attributed to weakly hydrogen

bonded water and the band at 3039 cm-1 to strongly hydrogen bonded water.

Deuteration results in the observation of OD bands at 2563, 2407 and 2279 cm-1. The

ratio of these bands changes with deuteration. Deuteration shows that the strongly

hydrogen bonded water is replaced in preference to the weakly hydrogen bonded

water. Three HOH bending modes are observed at 1686, 1633, 1572 and DOD

bending modes at 1236, 1203 and 1176 cm-1. Deuteration causes the loss of intensity

of the bands at 841, 710 and 561 cm-1 and new bands are observed at 692, 648 and

617 cm-1. These three bands are attributed to the water librational modes. Deuteration

results in additional Raman band at 809 cm-1 with increasing intensity with extent of

deuteration. Deuteration results in the shift of Raman bands to lower wavenumbers.

77

78

79

80

81

82

83

84

Chapter 6 – Site occupancy of Co and Ni in erythrite - annabergite solid solutions by vibrational spectroscopy

Wayde N. Martens, J. Theo Kloprogge, Ray L. Frost and Llew Rintoul

Wayde N. Martens Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

J. Theo Kloprogge Inorganic Materials Research Program, Queensland

University of Technology, Queensland 4001, Australia.

Ray L. Frost Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

Llew Rintoul Inorganic Materials Research Program, Queensland

University of Technology, Brisbane, Queensland 4001,

Australia.

American Mineralogist, Submitted 02-04-04

85

6.1 Statement of contribution

Wayde N. Martens : Wrote the manuscript, experimental design, conducted

experiments, and data analysis.

J. Theo Kloprogge: Major editing, aided data analysis

Ray L. Frost : Editing, data analysis

Llew Rintoul : Aided experimental design, editing, aided data analysis, aided

writing of manuscript.

86

6.2 Abstract

The solid solutions series between erythrite (Co3(AsO4)2.8H2O) and

annabergite (Ni3(AsO4)2.8H2O) has been synthesised and studied by a combination of

X-ray diffraction, scanning electron microscopy, Raman and infrared spectroscopy.

The solid solution between erythrite and annabergite has been found to be complete,

with the monoclinic C2/m space group being retained throughout the solid solution

series. The unit cell parameters decrease in size along all crystallographic directions

as nickel is introduced into the unit cell. The ß angle in the unit cell also decreases

from 105.052° (Co3(AsO4)2.8H2O) to 104.886° (Ni3(AsO4)2.8H2O), as the unit cell

volume decreases. Crystals of annabergite and samples with high Ni content are

elongated along the a crystallographic direction, contrasting with crystals of erythrite

and high Co containing samples, which are elongated along the c crystallographic

direction. In the Raman and infrared spectra of the synthetic minerals, the band

positions were found to shift in accordance with the increase in bonding strength

associated with the decrease in the unit cell parameters. It has been shown that trends

in Raman band positions of the anti-symmetric arsenate stretching vibration are

sensitive to the site occupancy in the crystal structure. Changes in the crystal

morphology, unit cell parameters and vibrational spectra have been rationalised in

terms of the site occupancy of Co and Ni in the crystal structure. Substitution of

nickel has been shown to be directed to metal site 1 whereas cobalt is directed to

metal site 2.

87

88

89

90

91

92

93

94

95

7. GENERAL DISCUSSION

This thesis has increased the scientific knowledge of the crystallographic and

spectroscopic properties of the aragonite group and the vivianite group minerals. In

particular fundamental knowledge of the spectroscopy of the end members, erythrite

and annabergite, and their solid solutions has been obtained. A new method of

studying the site occupancy of ions in solid solutions has been developed which

utilises vibrational spectroscopy. A detailed knowledge and understanding of factor

group analysis applied to the study of minerals has been achieved. This has been

applied to a significant number of minerals as shown by the number of refereed

articles published by the author, and cited in the beginning of the thesis.

The inspection of the Raman spectra of aragonites reveals the presence of

isotopic bands. The assignment of the satellite band to 18O-12C allows the use of this

band to quantify the oxygen isotopic ratio. This can have important implications for

various areas in geology. Oxygen isotope fractionation between carbonates and

(sea)water for example is well known to depend on temperature. In principle this

allows one for example to determine the paleotemperatures of ancient oceans with a

resolution better than 0.5 °C based on fractionation between calcite and seawater. A

similar technique also exists for cerussite, and a variety of other minerals. The normal

analysis method for stable isotopes involves very careful preparation of your samples

under clean room conditions followed by measurement with ICP-MS (Inductively

Coupled Plasma-Mass Spectrometry). It may be suggested that, when properly

standardised, Raman spectroscopy can be a rapid and much cheaper alternative to get

an idea of the isotope fractionation in carbonates.

96

Consideration of the water stretching and bending vibrations and the factor

group splitting of the arsenate stretching vibrations has allowed the determination of

the site occupancy of metal cations in the structure of the minerals. This finding has

profound implications for the field of mineral science, particularly where there is an

inability to collect high quality X-ray diffraction data (i.e. synchrotron) for Rietveld

analysis. Rietveld analysis also requires high operator skills, and is difficult for even

the skilled users. The technique used in this thesis does require basic knowledge of

the crystal unit cell, which is only available from crystallographic data obtained by

XRD. The infrared and Raman based technique may aid in the structural

determination of many other minerals and solid solutions. Questions still remain

however about the wider validity of the technique to other mineral solid solution

series. Future work needs to be directed to other solid solution series with known site

occupancy preferences to fully confirm the technique. The investigation of the

erythrite - annabergite solid solution series continues. Rietveld analysis of high

quality X-ray diffraction data is expected to be available in due course. Potentially the

type of techniques developed in this thesis may be applied to quite a few solid

solution series. Of most interest are solid solution series which have temperature

dependencies such as Fe in sphalerite (ZnS), ilmenite, garnets pyroxenes and

plagioclase minerals. Such solid solutions are well known geothermometers and as

mineral solid solutions directly effect infrared and Raman spectroscopic data such

vibrational techniques may yield a new field in geothermometers. This would be of

particular use where the mineral samples may not be able to be subjected to chemical

analysis such as SEM microprobe analysis.

97

Much work has been done on the analysis of mineral vibrational spectra, but

prediction of mineral spectra, or the prediction of mineral crystal structures from

spectra is still not routinely possible. Quantum mechanical computer models are able

to predict mineral spectra from crystal structures, but these are cumbersome, usually

requiring many days of supercomputer time. This study and others have shown that

the vibrational spectroscopy of minerals are fundamentally controlled by unit cell

chemistry, symmetry and bonding distances. Much more work is required on the

vibrational spectroscopy of minerals, to allow Monte Carlo type calculations of the

mineral crystal structures. Statistical techniques such as chemometrics, principal

component analysis (PCA) and SIMCA may also be useful for discrimination of

mineral crystal structures. Few such studies have been conducted on minerals, with

most concerned with quantification of mineral mixtures. PCA and SIMCA techniques

applied to infrared and Raman spectra, may be able to discriminate between different

crystal structures as similar factor group splitting patterns occur for groups of

minerals as seen in this study.

Even though this study has been able to achieve the objectives proposed, it has

raised more questions than answers. More work is still required in the field of mineral

spectroscopy as the spectra of most minerals, especially Raman spectra, are still yet to

be reported. Mineral spectral databases are still in the early stages of development, as

such, even simple tasks such as mineral identification are at this stage not possible.

Fundamental work on the spectroscopy of minerals such as the work contained with

in this thesis needs to be conducted to understand mineral spectra and their

assignment to enable databases such as those already existing for X-ray diffraction

(powder diffraction file, PDF) to be developed.

98

8. Appendix The following papers are supporting studies which were conducted is parallel

with this thesis.

99

Frost, R.L., Martens, W., Williams, P.A. and Kloprogge, J.T., J. Raman Spectrosc.

2003, 34, 751-759.

Frost, R.L., Weier, M.L., Martens, W., Kloprogge, J.T. and Ding Z., Thermochimica

acta, 2003, 403, 237-249.

Frost, R.L., Weier, M.L., Martens, W., Kloprogge, J.T. and Ding Z., Thermochimica

acta, 2003, 401, 121-130.

Frost, R.L., Kloprogge, J.T., and Martens, W., J. Raman Spectrosc. 2004, 35, 28-35.

Martens, Frost, R.L. and W., Williams, P.A., J.T., J. Raman Spectrosc. 2003, 34, 104-

111.

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147