PDF - University of Arizona

124 downloads 190 Views 6MB Size Report
James Shigley, Ph.D.-Gemological Institute of America (GIA) ... The causes of color in various gem minerals and the electronic spectral features associated with .
CHARACTERIZATION OF VIBRATIONAL AND ELECTRONIC FEATURES IN THE RAMAN SPECTRA OF GEM MINERALS by Renata Jasinevicius

A Prepublication Manuscript Submitted to the Faculty of the DEPARTMENT OF GEOSCIENCES In Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE In the Graduate College THE UNIVERSITY OF ARIZONA 2009

ii

ACKNOWLEDGMENTS This work could not have been completed without the sponsorship of the Israel Diamond Institute (IDI). A special thanks to the following contributors for making this work possible: James Shigley, Ph.D.-Gemological Institute of America (GIA) Dr. M. Bonner Denton-University of Arizona, Dept. of Chemistry Bear Williams, Stone Group Labs (photos) Tom Tashey-Professional Gem Sciences, Inc. Charlene Estrada, photography Sue Robison, RRUFF Project, University of Arizona To my advisor, Bob Downs, you have challenged me in new ways and taught me how to network. You have helped me fine tune my writing skills and master the art of thinking scientifically. You have made me a better student and opened doors for my future. Thank you. I’d like to extend my gratitude to all the people who provided me with personal support and encouragement throughout my academic career: To my future husband, Jason Lafler, I never would have made it without you. I am the luckiest lady in the world. Madison Barkley, you are a dear friend, and having you in my life made all of my graduate school experiences more memorable. To my mentor and friend, Elizabeth Gordon, your guidance has helped me become the person I am today. You are an inspirational teacher and I appreciate everything you have done for me. To my best friend, Andrea Mikenas, you always believed I could do it and never failed to remind me. To my former roomie, Lindsay Draeger, your Sunday morning polka phone calls made me smile and made me feel a little less lonely. Last, but certainly not least, I’d like to thank my family. To my Mom and Dad who have always pushed me to achieve excellence, I love you and appreciate everything you have sacrificed to help me achieve my goals. Rich, you are a great big brother. To my future parents, Val and Eunice, since the first day we met you have taken me under your wings and treated me like your daughter. I love you both and feel incredibly blessed to have you in my life.

iii

CONTENTS INTRODUCTION Raman Spectroscopy IR Spectroscopy OH in Minerals Luminescence Spectroscopy Color Theory

1 2 14 14 15 16

RAMAN ANALYSIS 1. Beryl

19

2. Chrysoberyl

26

3. Corundum

30

4. Diamond

35

5. Diopside

43

6. Garnet

48

7. Olivine

74

8. Quartz

79

9. Spinel

82

10. Spodumene

87

11. Titanite

90

12. Topaz

97

13. Tourmaline

102

14. Zircon

114

15. Zoisite

125

APPENDICES Appendix A & B: Features in Raman Spectra

132

Appendix C: Unit Conversions

141

Appendix D: Cause of Color Chart

143

Appendix E: Point Symmetry Notations

147

iv

Introduction Raman spectroscopy, a non-destructive technique used to interpret atomic vibrations, has become a popular tool for the rapid identification of materials. Raman spectra produce unique vibrational fingerprints useful in identifying a multitude of materials. With the advent of numerous gem treatments and a variety of methods available for mineral synthesis, Raman spectroscopy is particularly useful in identifying and characterizing gemstones. Micro-inclusions in minerals can be analyzed using Raman spectroscopy providing evidence of mineral genesis or geologic origin. More recently, fluorescence features attributed to chromophoric ions and trace elements have been observed in Raman spectra, revealing important information about crystal chemistry. Analysis and interpretation of these features may help distinguish between natural, treated, and synthesized materials. Advances in optical technologies are bringing hand-held Raman spectrometers to the forefront of materials research. As new instruments are developed, with both increases in portability and decreases in production costs, hand-held Raman units will likely be fundamental to laboratory and field-based Geoscience and gemological research in the future. Therefore, the development of databases and interpretation of spectra in anticipation of the new instrumentation is required. Raman spectra and associated interpretation of spectral features for the important gemstones are presented in this study. In this study I will present a characterization of the vibrational and electronic features present in the Raman spectra of gem minerals including: ƒ ƒ ƒ ƒ ƒ

The effects of orientation Spectral features associated with the vibrational modes of OH and H2O The causes of color in various gem minerals and the electronic spectral features associated with particular color-inducing cations Spectral features associated with luminescence of REE The effects of metamictization

1

Basic Raman Theory Raman spectroscopy is a type of vibrational spectroscopy. Intense electromagnetic radiation (typically generated by a laser) interacts with a substance and is scattered into radiation of different wavelengths associated with nuclear motion and producing a unique spectral fingerprint of atomic vibrations (Smith and Dent, 2005). An intense light source is necessary in Raman spectroscopy because Raman scattering is a very weak process; only one out of every 106-108 photons will Raman scatter (Smith and Dent, 2005). In general, gem minerals at room temperature are in the ground vibrational state (lowest energy vibrational level). When light interacts with a crystal the incident radiation can be scattered in several ways that include: 1) Rayleigh scattering, in which the scattered photon retains the energy of the incident beam (no energy change, elastic scattering) (Fig. I1) and 2) Raman scattering, in which the scattered photon experiences a change in energy (inelastic). There are two types of Raman scattering: 1) Stokes and 2) anti-Stokes. Stokes scattering occurs when atoms in the crystal (at the ground vibrational state) absorb the energy from the incident photon and are ultimately promoted to a higher energy vibrational state; the incident photon loses energy relative to its original state and the wavelength of the scattered light is shifted towards the red end of the electromagnetic spectrum (Fig. I1). During antiStokes scattering, energy is transferred from the already excited atoms to the incident photon and subsequently, the atoms associated with this specific vibration are demoted to the ground vibrational state; the scattered light is higher in energy than the Rayleigh line and therefore, the wavelength of the scattered light is shifted towards the blue end of the spectrum (Fig. I1) (Smith and Dent, 2005). AntiStokes scattering occurs less frequently than Stokes scattering because it requires that the atoms already be in a higher energy vibrational state when the laser interacts with it (Smith and Dent, 2005). There are characteristic temperatures at which certain vibrations are activated in different sets of bonded atoms. For example, strongly bonded atoms like Si-O are not vibrationally active at room temperature; they are considered high energy vibrations. Fig. I1 Energy diagram showing transitions in various types of spectra (modified from Smith and Dent, 2005). IR

Rayleigh

Stokes Raman

Anti-Stokes Raman

Fluorescence

Electronic States

Vibrational Relaxation & Internal Conversion

(Emission)

Fluorescence

(Excitation)

Absorption

Virtual States

Vibrational States Ground State Whether or not a vibrational mode is Raman active depends on the polarization of the vibrating bonded atoms. When the incident beam interacts with the crystal, the atoms begin to oscillate at the same frequency of the incident radiation. As an atom oscillates, its electrons are pulled in various directions, depending on their distribution in the electron cloud, resulting in deformation of the cloud. As the electrons move, so does the atom’s nucleus producing a separation of charges in the atom called a dipole

2

(the atom becomes polarized). Changes in the polarization of bonded atoms produce Raman active vibrational modes (Ferraro et al., 2003). In minerals, chemistry and crystal structure dictate the types of vibrations that can occur. The way bonded atoms in a crystal can bend, stretch, or rotate, i.e. their degrees of freedom of movement, depends on the crystal symmetry (Smith and Dent, 2005). Complex correlation matrices involving specific site symmetries can predict Raman active modes in minerals (Ferraro et al., 2003). The details of this process are beyond the scope of this study. However, for a more detailed discussion of Raman selection rules, character tables, factor group analysis, and crystal field theory see the following sources: “Infrared and Raman Selection Rules for Molecular and Lattice Vibrations: The Correlation Method,” W.G. Fateley and F.R. Dollish, 1972 “Molecular Vibrations: The Theory of Infrared and Raman Vibrational Spectra,” E.B. Wilson, J.C. Decius, and P.C. Cross, 1955 “Chemical Applications of Group Theory,” F.A. Cotton, 1971 “Symmetry in Bonding and Spectra: An Introduction,” B.E. Douglas and C.A. Hollings Note: Raman mode analysis tables in this study were generated using the Raman mode prediction tool found on the Bilbao Crystallographic Server available online at http://www.cryst.ehu.es/.

References Bilbao Crystallographic Server II: Representations of crystallographic point groups and space groups”. Acta Cryst. (2006), A62, 115-128. Ferraro, J.R., Nakamoto, K. & Brown, C.W. (2003) Introductory Raman Spectroscopy, 2nd edn, Academic press, San Diego, CA. Smith, E. & Dent, G. (2005) Modern Raman spectroscopy: a practical approach, John Wiley & Sons, Chichester, West Sussex, England.

3

Experimental Procedures and the RRUFF Database All Raman spectra and X-ray diffraction data utilized in this study have been taken from the RRUFF database (found at www.rruff.info). This database, directed by Dr. Robert T. Downs at the University of Arizona, provides a large set of chemical and spectral data of minerals. All information is available to the public including the downloadable software, CrystalSleuth, used in this study to view, stack, and process Raman spectra (Laetsch and Downs, 2006). The development of the RRUFF database is beneficial to the advancement of Raman spectroscopic research, particularly to applications in geoscience, gemology, and material science. Two Raman spectrometers are used to collect spectral data in the RRUFF lab: 1) Thermo Nikolet Almega microRaman with 532 nm and 780 nm lasers (partially polarized, 1 μm spot size, 4 cm-1 resolution) and 2) a customized open beam system with a 514 nm tunable Ar laser and a Jobin Yvon Spex HR 460 spectrometer equipped with a liquid nitrogen cooled CCD and a 1200 grooves per mm grating; used specifically to collect Raman spectra of oriented samples.

Important Notes: Polarization: The 514 nm Ar laser in the open beam Raman system has vertical linear polarization, meaning the electric field of the laser oscillates vertically (up and down) with respect to the direction of propagation of the laser beam (Fig. I2). When the light leaves the laser enclosure it is polarized, however, interaction of the laser with various optical elements such as lenses, gratings, and mirrors can change the laser polarization due to geometrical changes along the laser’s path. Therefore, it is important to note that the optical components of a spectrometer must be very deliberately and precisely installed in order to maximize laser polarization over the entire optical bench. Background Correction: Raman spectra are available on the RRUFF database in two forms: 1) raw and 2) processed. Raw spectra are loaded into the CrystalSleuth program and any cosmic ray present in the spectrum, non-Raman modes, or features associated with the notch filter are removed (we call this process is called trimming) (Hill and Rogalla, 1992; Laetsch and Downs, 2006). Processed files are “background corrected,” meaning the spectra are forced to the baseline by an algorithm in the CrystalSleuth program (Laetsch and Downs, 2006). This allows for easy comparison of Raman spectra. Any Raman spectra used in this study that have been background corrected are marked as such. This process does NOT alter the Raman mode locations. An example is provided in Fig. I3. Crystal Structures: All crystal structure figures and images have been generated using XtalDraw software developed by Downs and Hall-Wallace (2003). Fig. I2 (below) Picture demonstrating vertical linear polarization of a laser; image modified from Paschotta (2009) Laser beam

4

Fig. I3 Raman spectra of spinel: 1) raw and 2) processed (background corrected); sample R050392, 780 nm laser

Intensity

1.

2. -1

Raman Shift (cm )

Notation of Spectral Features The positions of luminescence features in the Raman spectra of gem minerals are wavelength dependent. For convenience, the peak positions of luminescence centers observed in this study will be presented in both nm (as described in the literature) and Raman shifts (cm-1) using the following notation (see also Appendix C for unit conversions): laser wavelength (nm)

Raman shift cm-1

For example: “Luminescence bands located at 693 nm (5324366 cm-1) and 694 nm (5324396 cm-1) are attributed to octahedrally coordinated Cr3+ luminescence centers in corundum (Gaft et al., 2005).” Luminescence Features in this Study In this study, the luminescence features appear in the extended range (>1500 cm-1) of the Raman spectra of various minerals. In many cases the luminescence features are so intense that you can no longer see the Raman peaks associated with the atomic vibrations of the mineral. An example of this is provided below in the Raman spectrum of topaz (Fig. I4). Fig. I4 Extended range Raman spectrum of topaz; note how the luminescence features overwhelm the Raman peaks; sample R060024, Minas Gerais, Brazil, 532 nm laser, unoriented

Luminescence Features

Raman Peaks

5

Minimization of Luminescence In a gated Raman study, Gaft et al. (2009) reported that by pulsing the laser, luminescence features that would otherwise overwhelm the Raman spectrum of a mineral sample could be minimized (Gaft et al., 2009).

References Gaft, M. & Nagli, L. (2009) Gated Raman spectroscopy: potential for fundamental and applied mineralogy. European Journal of Mineralogy, Vol. 21, No. 1, pp. 33. Hill, W. & Rogalla, D. (1992) Spike-correction of weak signals from charge-coupled devices and its application to Raman spectroscopy. Analytical Chemistry, Vol. 64, No. 21, pp. 2575-2579. Downs, R.T. & Hall-Wallace, M. (2003) The American Mineralogist Crystal Structure Database. American Mineralogist 88, 247-250. Laetsch T. & Downs R. (2006) Software For Identification and Refinement of Cell Parameters From Powder Diffraction Data of Minerals Using the RRUFF Project and American Mineralogist Crystal Structure Databases. th Abstracts from the 19 General Meeting of the International Mineralogical Association, Kobe, Japan, 23-28 July 2006. Paschotta, R. (2009) Encyclopedia of Laser Physics and Technology. Online .

6

Relationship of Raman Peaks and Bond Length A study comparing bond lengths to Raman peak positions of asymmetrical stretching modes in anionic groups was conducted by Estrada et al. (2009) and a preliminary plot is provided and discussed here (Fig. I5). Overall, the data displays a negative correlation, in which increases in bond lengths are associated with smaller Raman shifts, and this data can be fit to a power law curve: ν = 2328.9 R-1.95 Within each anionic group there are also negative correlations, however, the slopes of these trends are much steeper. In general, the following statements have been supported by the correlations reported in this study: Shorter, stronger bonds produce Raman peaks located at higher Raman shifts. Longer, weaker bonds appear at lower Raman shifts and are much more difficult to correlate because the atoms are not as wellconstrained as in short, strong bonds, therefore oscillating in many different ways (Estrada, 2009). This information is very useful in approximately predicting Raman mode assignments in minerals containing the anionic groups presented in this study. . Fig. I5 Plot of bond length (Å) vs Raman peak positions of asymmetrical stretching modes of anionic groups; bonds specified in the key, reproduced with permission from (Estrada, 2009). N-O C-O III B-O IV B-O

-1

Raman Frequency, ν (cm )

1600

S-O P-O Si-O Be-O Cr-O

1400

As-O V-O Mo-O W-O

1200

1000

800

600 1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2

R(M-O) (Ǻ)

Reference Estrada, C. (2009) The Correlation of M-O Bond Length to Raman Stretching Frequency in Mineral Anionic Groups. unpublished.

7

Orientational Dependence of Raman Spectra The optical indicatrices of minerals dictate optical properties along the crystallographic axes including the orientational dependence of Raman spectra. Optical indicatrices are imaginary three-dimensional surfaces in which the lengths of the axes are proportional to the refractive index (Dyar et al., 2008). Examples of optical indicatries are presented below. Modified from (Dyar et al., 2008)

Crystal Systems

Optical Class

Isometric (cubic): a = b = c α = β = γ = 90o

Isotropic

Tetragonal: a = b ≠ c α = β = γ = 90o

Uniaxial

Hexagonal (trigonal): a = b ≠ c α = β = 90o γ = 120o

Uniaxial

Orthorhombic: a ≠ b ≠ c α = β = γ = 90o

Biaxial

Monoclinic: a ≠ b ≠ c α = γ = 90o β ≠ 90o

Biaxial

Triclinic: a ≠ b ≠ c α ≠ β ≠ γ≠ 90o

Biaxial

Isotropic Optical Indicatrix: The Raman spectra of minerals with an isotropic indicatrix do not exhibit changes in peak intensity with changes in the polarization of the incident beam. This is because no matter which may the crystal is oriented, the cross-section of the indicatrix is a sphere. An example is diamond.

8

Uniaxial Optical Indicatrix: Tetragonal and hexagonal minerals have uniaxial optical indicatrices. The Raman spectra of minerals with uniaxial indicatrices exhibit changes in Raman peak intensity with a change in the polarization of the laser. In the uniaxial optical indicatrix there is only one orientation in which the cross-section of the indicatrix is circular. An example is beryl. When a beryl crystal is oriented such that the polarization of the laser is parallel to the c-axis, there is no change in the intensity of the Raman peaks (the crosssection through the indicatrix is a circle). However, when the laser is polarized parallel to the a-axis, the peak intensities change (the cross-section of the indicatrix is an ellipse).

Uniaxial Positive Indicatrix

Circular Cross-section

Uniaxial Negative Indicatrix

Biaxial Optical Indicatrix: Orthorhombic, monoclinic, and triclinic minerals have biaxial optical indicatrices. A bixial indicatrix looks like an ellipsoid, however, unlike the uniaxial indicatrix, there are two angles that produce a circular crosssection.

Reference Dyar, M., Gunter, M. & Tasa, D. (2008) Mineralogy and Optical Mineralogy. Mineralogical Society of America. Chantilly, VA.

9

Factors Influencing the Raman Spectra of Minerals During collection, analysis, and processing of hundreds of Raman spectra of various minerals, some basic experimental observations have been made about factors that can affect the quality of Raman data. Heating When the incident laser beam interacts with a mineral surface, the surface is heated. Factors such as size, chemistry, and color of a mineral can alter the way it is capable of absorbing and dispersing the thermal energy produced by the laser. The effects of heating are illustrated by the following examples. Burning When a crystal is incapable of adequately absorbing and dispersing the energy of the incident beam, the sample can burn. Dark brown to black craters that are the size of the beam spot are indicative of burning. Sample size, chemistry, and color can influence the probability of burning. Size: Any very small or very thin crystals are at risk of burning, for example, single crystals prepared for X-ray diffraction analysis (50-100 microns in size). In addition, crystals that are poorly embedded in matrix can burn (the mineral may not be able to dissipate the heat of the laser). When dealing with small, thin, or poorly embedded crystals, the laser intensity is always reduced to the lowest laser intensity, 10%, to minimize the risk of burning. Samples that are embedded in matrix are often safe from burning. Chemistry: Gem minerals typically do not burn because they are hard, stable, strongly bonded, and usually transparent to translucent. However, there are many minerals that are at risk of burning due to their chemical compositions. Soft, weakly bonded minerals such as silver sulphides burn very easily. Acanthite (Ag2S) burns even under the lowest laser intensity setting. Color: The Raman system used in this study has two possible laser wavelengths, 532 nm (green) and 780/785 (red). Some crystals may burn under exposure to one of the lasers, but not the other due to the color of the mineral. An example is chalcanthite (CuSO4·5H2O). Chalcanthite does not burn under exposure to the higher energy 532 nm (green) laser, but does burn under the 780/785 nm (red) laser. This preferential burning occurs because chalcanthite is a blue mineral and when the red 780/785 nm laser hits the sample’s surface the sample absorbs more energy from the red light than from the green. The sample cannot disperse the thermal energy produced by the red laser because it is overwhelmed by absorption, and so it burns under only one laser. Not all blue minerals burn under the 780/785 nm laser. Color is typically combined with other factors such as size and chemistry in order to result in preferential burning. Dehydration Minerals containing water may dehydrate due to the heat produced by the lasers. The removal of water or hydroxyl from certain minerals can result in a phase change. Discoloration of mineral surface where the laser beam is focused on the sample is usually indicative of dehydration. In some cases, there is no visual evidence of dehydration. If the Raman spectrum of a sample unexpectedly changes during data during data collection, the sample has likely burned or dehydrated. Phase Changes (not related to water) Some minerals undergo a phase change when exposed to the heat of the lasers. An example is crocoite (PbCrO4). After less than 1 second of exposure to the 532 nm laser, the Raman spectrum of crocoite changes without any visual evidence on the sample’s surface to suggest alteration (there was no discoloration). Knight et al. (2000) discovered that crocoite experiences a phase change at 1068 K and transitions from the ambient temperature monazite structure to the barite structure. It is possible that the

10

laser heated the crocoite sample to 1068 K and ultimately caused the sample to change into its high temperature phase. Fluorescence Minerals containing rare-earth elements (REE) frequently produce fluorescence features that can overwhelm the Raman spectrum of a sample (Fig. I6 A & B). Minerals containing Ca, for example tremolite (†Ca2Mg5Si8O22(OH)2), frequently incorporate REE into their structures due to the size and coordination of the Ca-site and therefore, more commonly exhibit fluorescence in the Raman spectra. Rock-salt structure Minerals with the halite (or rock-salt) crystal structure produce Raman spectra with intensity equal to zero due to the presence of an inversion center (Fig. I7 & I8). An inversion center is a symmetry element describing a point within a crystal through which any straight line extends to points on opposite surfaces of the crystal at equal distances. When the incident beam causes the atoms in a mineral with the rocksalt structure to oscillate, the change in polarization of the atoms is negated because the distortion of the electron cloud in one direction is equal to the distortion of the cloud in another direction (a sort of destructive interference). The result is a Raman spectrum with intensity equal to zero. Metamictization Incorporation of radioactive elements like U and Th into the crystal structure of a mineral can result in metamictization (a loss of crystallinity). There are several gem minerals that are affected by metamictization, including titanite and zircon (see also sections on titanite and zircon). Metamictization can have dramatic effects on the Raman spectrum of a sample including loss of peak intensity, an increase in peak width, disappearance of peaks, and shifts in peak frequencies.

11

Fig. I6 A. Raman spectra of minerals exhibiting fluorescence.

B.

Fig. I7 A. Raman spectra of minerals with Raman intensities = zero due to symmetry

B.

12

Fig. I8 The halite structure; purple spheres are Cl atoms, teal spheres are Na atoms

References Knight, K.S. (2000) A high temperature structural phase transition in crocoite (PbCrO4) at 1068 K: crystal structure refinement at 1073 K and thermal expansion tensor determination at 1000 K. Mineralogical Magazine, Vol. 64, No. 2, pp. 291-300.

13

Infrared (IR) Spectroscopy Infrared spectroscopy, like Raman spectroscopy, is a tool used to measure the unique vibrational fingerprints of materials. In IR spectroscopy, radiation spanning a range of IR wavelengths is focused onto a sample (Smith and Dent, 2005). When this energy is absorbed by sets of bonded atoms, the atoms are promoted to a higher vibrational state. The energy of the incident light must match the energy of a specific atomic vibration in order for absorption to occur (Smith and Dent, 2005). Vibrations are IR active if a change in the dipole moment of the atoms (charge separation) is induced during absorption of the incident radiation (Ferraro et al., 2003). IR spectra are commonly plotted with the x-axis equal to either wavelength or wave number and the y-axis equal to either absorbance or transmission. It is possible for vibrations to be both IR and Raman active. An example of equivalent vibrational features in both the Raman spectra and the IR spectra of a gem mineral is a series of peaks attributed to molecular water occupying the structural channels of beryl (see also beryl section). References Ferraro, J.R., Nakamoto, K. & Brown, C.W. (2003) Introductory Raman Spectroscopy, 2nd edn, Academic press, San Diego, CA. Smith, E. & Dent, G. (2005) Modern Raman spectroscopy: a practical approach, John Wiley & Sons, Chichester, West Sussex, England.

Importance of OH- Incorporation in Minerals Investigation of the substitution of OH- and H2O molecules into the structure of minerals, principally those considered nominally anhydrous (those not requiring hydrous species to balance their stoichiometry), has been of great interest over the past decade. Incorporation of OH- can affect a mineral’s diffusion and dielectric properties (Rossman, 1996). Of even greater scientific interest is the incorporation of OH into nominally anhydrous minerals present in the mantle including garnets and olivine (Beran and Libowitsky, 2006). The incorporation of water into the structure of minerals can affect a wide variety of important mineral properties at depth including: melting conditions, bulk modulus (resistance to compression), and phase transition processes. Therefore, characterizing the influences of OH- on minerals in the mantle can help better constrain mantle dynamics (Beran and Libowitsky, 2006). Common crustal minerals such as quartz, feldspar, and pyroxenes can also contain trace OH- (Johnson, 2006). The primary method utilized in OH- studies is IR spectroscopy, however, Raman peaks associated with OH- and H2O in various orientations have also been observed in my study and are discussed in the following sections: beryl, garnet, titanite, topaz, tourmaline, zircon, and zoisite. References Beran, A. & Libowitzky, E. (2006) Water in natural mantle minerals II: olivine, garnet and accessory minerals. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 169-191. Johnson, E.A. (2006) Water in nominally anhydrous crustal minerals: speciation, concentration, and geologic significance. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 117-154. Rossman, G.R. (1996) Studies of OH in nominally anhydrous minerals. Physics and Chemistry of Minerals, Vol. 23, No. 4, pp. 299-304.

14

Luminescence Spectroscopy Additional spectral features are observed in the Raman spectra of mineral samples from the RRUFF database. These features are not Raman modes, they are luminescence features. The wavelengths of the lasers (514, 532, 780 and 785 nm) used to collect Raman spectra have excited multiple luminescence centers in various minerals. A basic introduction to the theory of luminescence spectroscopy (based on Gaft et al., 2005) is provided here. When electromagnetic radiation (light) interacts with a luminescent material, an electron is excited to a higher energy state. When the electron drops back down to the ground state it releases light of a specific wavelength or range of wavelengths (called radiative decay) (Fig. I1, p.2). Luminescence can be excited in minerals by radiation of many different wavelengths; luminescence induced by light sources in the UVvisible range is called photoluminescence (this is what we are observing in the Raman spectra of this study). Other means of exciting luminescence in minerals include excitation by a beam of electrons, called cathodoluminescence, and excitation by heat, called thermoluminescence. The ion responsible for a particular luminescence feature is called a luminescent center or activator. Correlating luminescence features with their associated centers is not a straightforward task. Simultaneous emission of multiple centers, luminescence decay, and symmetry constraints can affect interpretation and correlation of activators to their associated features. The majority of luminescence centers in minerals are transition elements and REE. Luminescence spectroscopy is used to measure the energy levels of luminescence centers. Energy levels are defined by characteristic states (ground state: lowest energy, excited states: higher energy). Electronic and vibrational transitions are of greatest interest in mineral luminescence. Luminescence emission occurs when an excited electron jumps from a higher energy level to a lower one, releasing a photon. Luminescent minerals emit radiation only when the excitation energy is absorbed. Symmetry and directional properties of orbitals determine the luminescent properties of a substance. Every atom has a scheme of energy levels which changes when the atoms combine to form the crystal structures of minerals. In my study, the majority of luminescence activators are transition elements and REE. The orbitals of interest in luminescence caused by transition elements and REE are the d and f orbitals. Crystal field and ligand theory discuss the manner in which bonded atoms are influenced by their nearest neighbors based on electon distributions and symmetry. For more information on crystal field or ligand theory, consult the following texts: “Chemical Applications of Group Theory,” F.A. Cotton, 1971 “Symmetry in Bonding and Spectra: An Introduction,” B.E. Douglas and C.A. Hollings

References Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany.

15

Causes of Color in Minerals The unique and variable combinations of hues, tones, and saturations of gem minerals are the most important characteristics dictating gemstone value and demand. Here is a brief discussion of the mechanisms producing color in gems (see also “Cause of Color Chart” in Appendix). For more detailed information see, “The Physics and Chemistry of Color,” by Kurt Nassau. There are five mechanisms responsible for color in gem minerals (Fritsch and Rossman, 1987): 1) 2) 3) 4) 5)

Metal ions Charge transfer Color centers Band theory Physical optics

1) Metal Ions The most common cause of color in minerals is the presence of color-inducing metal ions, sometimes called chromophores. Color is produced by chromophores when electrons undergo transitions between orbitals confined to a single metal ion (Fritsch and Rossman, 1988). When light interacts with a mineral, and consequently the ions dispersed throughout, the electrons become excited jumping from the ground state to an excited state as the ions absorb the incident energy. This excited state is unstable and the electrons can return back to the stable ground state in two ways: 1) by releasing the energy into the crystal lattice by atomic vibration (heat), or 2) by emitting the energy as a photon, a process called luminescence (see also section on “Luminescence Spectroscopy”). Colors produced by metal ions are the result of absorption of visible light. Common color-inducing ions present in gemstones include: Cu, Ni, Cr, V, Mn, Fe, Co, and Ti (Fritsch and Rossman, 1987). Copper, for example, produces the green color of malachite as well as the blue color of azurite. Color-inducing ions typically contain unpaired electrons in d or f orbitals (Nassau, 2001). This is significant because unpaired electrons can be easily excited by the energy of visible light, resulting in absorption. When ions are bonded together such that all of the electrons are paired off, having only completely filled or completely empty shells, the paired electrons become incredibly difficult to excite, requiring a large amount of energy to do so (UV). Accordingly, substances with all of their electrons paired off cannot selectively absorb visible light because visible light is not high enough in energy to excite the electron pairs and therefore, these substances remain colorless (Nassau, 2001). An example is halite (NaCl). The valence state or charge of an ion influences the color of a mineral due to the probability of occurrence of certain electronic transitions, as determined by quantum mechanic rules that are beyond 2+ the scope of this study (Fritsch and Rossman, 1987). An example is the presence of Mn in beryl; Mn 3+ produces a light pink hue (morganite), while Mn produces a bright red color (Fritsch and Rossman, 1987). Other factors affecting the color of gems containing chromophores include nearest neighbor atoms, coordination of the ion (the number of atoms the ion is bonded to), and the local symmetry (Fritsch and Rossman, 1987). Common coordinations referred to in this paper include tetrahedral (four neighboring atoms), octahedral (six neighboring atoms), and distorted cubic (eight neighboring atoms). Frequently these metal ions substitute (take the place of) atoms of similar size and charge in the crystal structure. 2) Charge Transfer While the mechanism producing color via dispersed metal ions involves electronic transitions within a single ion, the process of charge transfer occurs when electrons jump from one atom to another (Fritsch and Rossman, 1988). Transfer of electrons can occur between an ion and its nearest neighbors and even its next nearest neighbors (this process is called intervalence charge transfer). Yellow beryl 3+ (heliodor) is the result of the interaction of Fe with its nearest neighbor oxygen atoms. Intervalence 2+ 4+ charge transfer between Fe and Ti give sapphire its distinctive blue color (Fritsch and Rossman,

16

1988). Intervalence charge transfer between two different metal ions is called heteronuclear intervalence charge transfer (Nassau, 2001). Homonuclear intervalence charge transfer derives from interactions between two of the same metal ions, but with different valence state, for example between Fe2+ and Fe3+ (Nassau, 2001). 3) Color Centers A defect that causes light absorption is called a color center. Defects are commonly caused by natural or artificial radiation and include vacancies (missing atoms), extra atoms, changes in valence state of metal ions, or extracted electrons placed into an existing defect (Fritsch and Rossman, 1988). When a color center has one less electron than it would have had prior to irradiation, the center is called a hole center (Nassau, 2001). Vacancies in diamond, removal of carbon atoms from the structure, produced by natural irradiation can produce green diamonds (Fritsch and Rossman, 1988). Substitution of Al3+ for Si4+ in quartz combined with natural irradiation produces a hole center and creates the color of smoky quartz (Fritsch and Rossman, 1988; Nassau, 2001). In general, gemstones colored by centers tend to fade easily with exposure to heat or light because the electrons that have been displaced during the formation of the color center are weakly held by surrounding cations. Therefore, it does not take much energy to free the electron from its trap. The freed electron returns to its original position, and the mineral to its original color (Fritsch and Rossman, 1988). 4) Band Theory A less common cause of color in gem minerals is the result of delocalized electrons interacting with visible light (Fritsch and Rossman, 1988). The study of this interaction is called band theory and it is most commonly used to describe the properties of metals and semiconductors. In some minerals billions of atoms contribute to the possible energy levels of the substance producing an energy band. The lowenergy valence band is populated only by electrons and the high energy conduction band is typically empty. The energy separating the two bands is called a band gap. Instead of transitions between energy levels of atoms, transitions between bands produce colors in these minerals (Fritsch and Rossman, 1988). Transitions between bands occur when the electrons in the valence band absorb light that provides a sufficient amount of energy for the electrons to jump over the band gap into the conduction band. There are three possible transitions that can occur: 1) the incident visible light does not provide the electrons with enough energy to jump the band gap (band gap is greater than energy of visible light), all visible light is transmitted (none is absorbed) and the gemstone appears colorless (corundum, topaz, quartz, quartz, diamond, beryl); 2) the energy of the band gap is in the visible range (violet, blue, green light are absorbed) and the gemstone can appear a range of colors from deep yellow to deep red (depending on specific energy of the band gap) i.e. the red color of cinnabar; 3) the energy of the band gap is less than the lowest energy of the visible light resulting in total absorption (the sum of all colors is black) and in a metallic luster due to remission of the light by the electrons; i.e. gold, copper, pyrite, and platinum (Fritsch and Rossman, 1988). Trace amounts of impurities of atoms in these minerals can produce energy levels that are between the valence and conduction band resulting in different colors. Examples of this are incorporation of boron and nitrogen (Type Ib) in diamond producing blue and yellow colors, respectively (Fritsch and Rossman, 1988). 5) Physical Optics There are other physical mechanisms besides absorption of light that can produce various colors in gems. Scattering, dispersion, intereference and diffraction of light can produce various optical effects in gem materials including the play-of-colors in opal and labradorite (diffraction) (Fritsch and Rossman, 1988). These optical phenomena are not significant to the scope of this study as it relates to Raman spectra.

17

References Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: colors caused by bandgaps and physical phenomena. Gems & GemoA logy, Vol. 24, No. 2, pp. 81-102. Fritsch, E. & Rossman, G.R. (1988) An update on color in gems. Part 2: Colors involving multiple atoms and color centers. Gems & Gemology, Vol. 24, No. 1, pp. 3–15. Fritsch, E. & Rossman, G.R. (1987) An Update on Color in Gems. Part 1. Introduction and Colors caused by Dispersed Metal Ions. Gems and Gemology, Vol. 23, No. 3, pp. 126–139. Nassau, K. (ed) (2001) The Physics and Chemistry of Color: The Fifteen Causes of Color, 2nd edn, John Wiley and Sons, New York.

18

Beryl Gem names: emerald, aquamarine, heliodor, morganite, goshenite, green beryl

Ideal chemistry: Be3Al2Si6O18 Crystal system: hexagonal Point Group:

H-M: 6/mmm

S: D6h emerald from Habachtal, Salzburg, Austria

Space Group: P6/mcc

Table of Atomic Coordinates (Hazen and Finger, 1986): atom

x

y

z

Be

.5

0

.25

Al

1/3

2/3

.25

Si

.3876

.1159

0

O1

.3103

.2369

0

O2

.4985

.1456

.1453

Raman mode analysis:

Raman Active Modes Atom

Wyckoff Position Point Symmetry A1g E2g E1g

O2

24m

1

3

6

6

Si, O1

12l

m

2

4

2

Be

6f

222

-

1

2

Al

4c

32

-

1

1

(1 × 24m) + (2 × 12l) + (1 × 6f) + (1 × 4c) Raman mode analysis predicts the existence of 36 active Raman modes in beryl: 7A1g + 16E2g + 13E1g = 36

19

Orientational Dependence of Spectra When beryl is oriented with the a-axis parallel to the incident laser beam (the cross-section of the optical indicatrix is an ellipse), the intensities of the peaks change dramatically with rotation due to a change in the degree of freedom of atomic vibration (Fig. 1.1A). When the sample is oriented with the c-axis parallel to the laser (the cross-section of the optical indicatrix is a circle), there is no noticeable change in peak intensities due to an equal distribution of atomic vibration in all directions (FIG. 1.1B). In spite of the orientational dependence of its Raman peaks, beryl can be accurately identified by the spectrum of a randomly oriented sample. Note that in collecting the Raman spectra of beryl it is possible that the Raman peak at 1100 cm-1 may vary drastically in intensity due to orientation. FIG. 1.1

Raman spectra of oriented beryl crystals; sample R050368 (red beryl), processed

A. Raman spectra showing peak intensities as a function of orientation; 514.5 nm laser parallel to a* (through prism face); at 0° laser is polarized ll to c-axis; at 90° laser is polarized perpendicular to c-axis

B. Raman spectra showing peak intensities as a function of orientation. Notice that the intensities do not change as the direction of polarization changes; 514.5 nm laser parallel to c-axis; at 0° laser is polarized ll to b-axis; at 90° laser is polarized perpendicular to b-axis

FIG. 1.2 A.

B.

The crystal structure of beryl. Yellow spheres: channel-filling atoms; purple tetrahedra: BeO4 groups; green tetrahedra: SiO4 groups; blue octahedra: AlO6 groups (A) Looking down: c-axis; (B) Looking down a-axis.

20

Spectral Features Related to H2O As depicted in Fig. 1.2A, stacked six-membered rings of Si-tetrahedra linked by Be-tetrahedra and Aloctahedra form channels parallel to the c-axis in beryl. These channels are commonly occupied by cations (most commonly, but not limited to Na, K, Li, Cs), water, CO2 or some combination of the above (Lodzinksi et al., 2005). Channel-filling water molecules in beryl are designated either as type-I or type-II** (Wood & Nassau, 1968). These designations are orientationally dependent; type-I water molecules are oriented with the 2fold axis of the water molecule perpendicular to the 6-fold axis of beryl; type-II molecules are oriented with the 2-fold axis of the water molecule parallel to the 6-fold axis of beryl (Aurisicchio et al. 1994). Aurisicchio et al. (1994) suggest that in alkali-rich beryl, OH groups interact with alkali cations in the channels, resulting in a dominance of type-II water, over type-I. Lodzinski et al. (2005) report that peaks centered at 1386 and 1240 cm-1 in the Raman spectra of beryl are attributed to CO2. OH- Peaks (based on IR data reported by Lodzinski et al., 2005): Type-I H2O: 1598 cm-1, 3609-3606 cm-1 (stretching vibration), 3692/3696 cm-1, 3880 cm-1 (Figs. 1.3-1.4) Type-II H2O: 1628/1634 cm-1, 3594/3597 cm-1 (stretching vibration), 3651/3657 cm-1 (Figs. 1.3-1.4) In many experimental techniques, such as electron microprobe, OH- can be difficult to detect. Therefore, the OH- content of a mineral is often calculated, not measured. OH- can be detected by X-ray or neutron diffraction. However, a simpler, quicker, non-destructive method for detecting OH- in a sample is Raman spectroscopy. **R. I. Mashkovtsev and A. S. Lebedev (1993) proposed a third water type associated with channel-filling alkali cations (Type-III). Type-III water is oriented along the same axis as Type-I, however, the cations are situated at a greater distance from the water molecules producing IR active bands observed at 3705 and 1604 cm-1. Type-III water was not observed in this study.

FIG. 1.3 Raman spectra of light-blue beryl demonstrating a type-I H2O peak centered at 3600 cm-1 and a type-II -1 H2O peak centered at 3660 cm ; sample R050065, 532 nm laser, unoriented

TypeI H2O Type-II H2O Type-I H2O

FIG. 1.4 (right) Magnification of a type-I H2O peak centered at 3600 cm-1 and a type-II H2O peak centered at 3660 cm-1 ; sample R050065, 532 nm laser, unoriented, processed

Type-II H2O

21

Spectral Features Related to Chromophores and Other Ions Minute amounts of chromophoric (color-inducing) trace elements produce the many colors in beryl. Colorless beryl, goshenite, has low concentrations of color-inducing atoms as compared to colored beryls. Golden beryl, heliodor, is the result of Fe3+ substituting for either Be2+ or Al3+. A variety of color centers responsible for the hues of light blue beryl, also known as aquamarine, are attributed to ferrous and ferric iron, while red and pink beryl contain trace amounts of color-inducing manganese (Gaft et al., 2005). Distiguishing between an emerald and a green beryl is a controversial issue in gemology. The presence of Cr3+ substituting for Al3+ in a deep green-colored beryl was the original defining characteristic of an emerald. However, V3+ substituting for Al3+ can also produce deep green-colored beryl (Gaft et al., 2005). Whether or not a gemstone is marketed as an emerald or simply a green beryl can greatly affect the gemstone’s value. Luminescence studies reported by Gaft et al. (2005) attribute two distinct luminescence centers located at 680 and 685 nm in beryl to octahedrally coordinated Cr3+ (Fig. 1.5). FIG. 1.5 Peaks in the Raman spectra of beryl associated with Cr3+ luminescence centers located at 680 nm 532 -1 532 -1 ( 4095 cm ) and 685 nm ( 4170 cm ); note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060943, green beryl from Xinjiang, China, unoriented

A. Cr3+

Note: Microprobe data shows that this beryl sample contains a significant amount of vanadium, thus, some might argue that it is not an emerald, but rather a green beryl. Although vanadium is present in the crystal, there are no visible luminescence features associated with vanadium in this spectrum.

22

Spectra of Minerals with Inclusions Gemstones frequently contain mineral inclusions. These inclusions provide scientists and gemologists alike with valuable information about mineral genesis. Raman spectroscopy is a useful tool for identifying mineral inclusions in gemstones. In an emerald from Habachtal, Salzburg, Austria, electron microprobe analysis reveals the presence of actinolite micro-inclusions (Fig. 1.8). The resulting spectrum displays Raman peaks belonging to both beryl and actinolite; note the doublet around 670 cm-1 and the presence of actinolite peaks in the 150-200 cm-1 range that do not normally exist in a beryl spectrum (Fig. 1.7 & 1.9). Recognizing spectra containing Raman peaks of multiple minerals may help to distinguish between gemstones from various localities. FIG. 1.7 Spectrum of emerald from Habachtal, Austria with a split peak centered at 670 cm-1 demonstrating that Raman peaks from both beryl and actinolite are present in this spectrum; sample R060944, 532 nm laser, unoriented, processed

FIG. 1.8 (right) Polished surface of microprobe mount of an emerald crystal from Austria; the light colored fibers are micro-inclusions of actinolite in the emerald

Intensity

FIG. 1.9 (below) Comparison of the Raman spectra of 1) Austrian emerald, 2) aquamarine R040002, 3) actinolite R040063; demonstrates that the Raman spectrum of the Austrian emerald (1) contains peaks belonging to both beryl and actinolite; unoriented; 532 nm laser, processed

3. 2. 1. Raman Shift (cm-1)

23

Fluorescence of Natural Beryl Raman spectra of natural emerald collected with a 785 nm laser exhibit fluorescence that overwhelms the Raman signal (Fig. 1.10A). In natural beryl of any other color, fluorescence generated by the 785 nm incident beam is minimal (Fig. 1.10B), making this optical phenomenon diagnostic of green-colored beryls. The presence of Cr3+ is likely the cause of this fluorescence. The Raman spectra of green-colored beryls collected using a laser of a higher energy, such as 532 nm, minimizes this effect (Fig. 1.10A). FIG. 1.10 Fluorescence in the Raman spectrum of a green-colored beryl with 785 nm laser (black spectrum) and 532 nm laser (green spectrum); sample R060942, unoriented

Intensity

A.

785 nm

532 nm Raman Shift (cm-1)

Intensity

B. The presence of minimal fluorescence in the Raman spectra of a colorless beryl with 785 nm laser and 532 nm laser; sample R040002, unoriented

532 nm 785 nm

Raman Shift (cm-1)

24

References Aurisicchio, C., Grubessi, O. & Zecchini, P. (1994) Infrared spectroscopy and crystal chemistry of the beryl group. Canadian Mineralogist, Vol. 32, No. 1, pp. 55. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Hazen, R.M., Au, A.Y. & Finger, L.W. (1986) High-pressure crystal chemistry of beryl (Be3Al2Si6O18) and euclase (BeAlSiO4OH). American Mineralogist, Vol. 71, No. 7-8, pp. 977-984. Łodziński, M., Sitarz, M., Stec, K., Kozanecki, M., Fojud, Z. & Jurga, S. (2005) ICP, IR, Raman, NMR investigations of beryls from pegmatites of the Sudety Mts. Journal of Molecular Structure, Vol. 744, pp. 1005-1015. Mashkovtsev, R. & Lebedev, A. (1993) Infrared spectroscopy of water in beryl. Journal of Structural Chemistry, Vol. 33, No. 6, pp. 930-933. Wood, D. & Nassau, K. (1968) The characterization of beryl and emerald by visible and infrared absorption spectroscopy. American Mineralogist, Vol. 53, No. 5/6, pp. 777-800.

Additional Information Adams, D.M. & Gardner, I.R. (1974) Single-crystal vibrational spectra of beryl and dioptase. Journal of the Chemical Society, Dalton Transactions, Vol. 1974, No. 14, pp. 1502-1505. Charoy, B., de Donato, P., Barres, O. & Pinto-Coelho, C. (1996) Channel occupancy in an alkali-poor beryl from Serra Branca (Goias, Brazil): Spectroscopic characterization. American Mineralogist, Vol. 81, No. 3-4, pp. 395403. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Fritsch, E. & Rossman, G.R. (1988) An update on color in gems. Part 2: Colors involving multiple atoms and color centers. Gems & Gemology, Vol. 24, No. 1, pp. 3–15. Fritsch, E. & Rossman, G.R. (1987) An Update on Color in Gems. Part 1. Introduction and Colors caused by Dispersed Metal Ions. Gems & Gemology, Vol. 23, No. 3, pp. 126–139. Goldman, S., Rossman, G.R. & Parkin, K.M. (1978) Channel constituents in beryl. Physics and Chemistry of Minerals, Vol. 3, No. 3, pp. 225-235. Hagemann, H., Lucken, A., Bill, H., Gysler-Sanz, J. & Stalder, H.A. (1990) Polarized Raman spectra of beryl and bazzite. Physics and Chemistry of Minerals, Vol. 17, No. 5, pp. 395-401. Hofmeister, A., Hoering, T. & Virgo, D. (1987) Vibrational spectroscopy of beryllium aluminosilicates: Heat capacity calculations from band assignments. Physics and Chemistry of Minerals, Vol. 14, No. 3, pp. 205-224. Johnson, E.A. (2006) Water in nominally anhydrous crustal minerals: speciation, concentration, and geologic significance. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 117-154. Johnson, M.L., Elen, S. & Muhlmeister, S. (1999) On the identification of various emerald filling substances. Gems & Gemology, Vol. 35, No. 2, pp. 82–107. Kim, C.C., Bell, M.I. McKeown, D.A. (1995) Vibrational analysis of beryl (Be3Al2Si6O18) and its constituent ring (Si6O18). Physica B: Physics of Condensed Matter, Vol. 205, No. 2, pp. 193-208. Kloprogge, J.T. & Frost, R.L. (2000) Raman microscopic study at 300 and 77 K of some pegmatite minerals from the Iveland–Evje area, Aust-Agder, Southern Norway. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, Vol. 56, No. 3, pp. 501-513. Moroz, I., Roth, M., Boudeulle, M. & Panczer, G. (2000) Raman microspectroscopy and fluorescence of emeralds from various deposits. Journal of Raman Spectroscopy, Vol. 31, No. 6, pp. 485-490. Rossman, G.R. (2006) Analytical methods for measuring water in nominally anhydrous minerals. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 1-28. Rossman, G.R. (1996) Studies of OH in nominally anhydrous minerals. Physics and Chemistry of Minerals, Vol. 23, No. 4, pp. 299-304. 2+ 3+ Spinolo, G., Fontana, I. & Galli, A. (2007) Optical absorption spectra of Fe and Fe in beryl crystals. Physica Status Solidi B, Basic Research, Vol. 244, No. 12, pp. 4363. Taran, M.N. & Rossman, G.R. (2001) Optical spectroscopic study of tuhualite and a re-examination of the beryl, cordierite, and osumilite spectra. American Mineralogist, Vol. 86, No. 9, pp. 973-980.

25

Chrysoberyl Gem names: alexandrite, cat’s eye Ideal chemistry: BeAl2O4 Crystal system: orthorhombic Point Group:

H-M: mmm

S: D2h

Space Group: Pnma

Synthetic alexandrite gemstone

Table of Atomic Coordinates (Weber et al., 2007): atom

x

y

z

Al1

0

0

0

Al2

0.27282

0.25

-0.00503

Be

0.09289

0.25

0.43360

O1

0.09022

0.25

0.78822

O2

0.43316

0.25

0.24167

O3

0.16324

0.01554

0.25728

Raman mode analysis:

Raman Active Modes Atom

Wyckoff Position Point Symmetry Ag B1g B2g B3g

O3

8d

1

3

3

3

3

Al2, Be, O1, O2

4c

m

2

1

2

1

Al1

4a

1

-

-

-

-

(1 x 8d) + (4 x 4c) + (1 x 4a) Raman mode analysis predicts the existence of 36 active Raman modes in chrysoberyl: 11Ag + 7B1g + 11B2g + 7B3g = 36

26

Spectral Features Related to Chromophores and Other Ions Perhaps the most famous gem variety of chrysoberyl is alexandrite. Chrysoberyl gemstones that exhibit a color change from green or bluish-green in daylight to purplish-red under incandescent light are called alexandrite. This optical phenomenon also occurs in other gemstones such as sapphire, garnet, spinel, zoisite, and fluorite, though the colors may vary (Liu et al., 1994). In alexandrite, color-change is attributed to octahedrally coordinated Cr3+ substituting for Al3+ (Fritsch and Rossman, 1988). The substitution of Cr3+ for Al3+ is more complicated, however, because there are two different symmetries associated with the Al3+ sites, half of the Al3+ atoms are located at an inversion center (the Al1 site) and half are located on a mirror plane (the Al2 site) (Gubelin, 1976; Hassan and El-Rakhawy, 1974). Cr3+ is slightly larger than Al3+ and therefore, prefers to occupy the more spacious Al2 site (1.934Å). However, it has been reported that with an increase in pressure and temperature, Cr3+ will occupy the Al1 sites (1.890Å) (Gubelin, 1976; Hassan and El-Rakhawy, 1974). The ratio of Cr3+ occupying the Al1 site to Cr3+ occupying the Al2 site determines whether or not the alexandrite will exhibit colorchange. The greater the ratio of Cr3+ in Al1 to Cr3+ in Al2, the more pronouned the color-change becomes (Gubelin, 1976; Hassan and El-Rakhawy, 1974). The alexandrite effect in gemstones is not solely dependent on the Cr3+ content, but rather on the positions and intensities of the absorption and transmission regions resulting from the presence of Cr3+ in various structural sites (Gubelin and Schmetzer, 1982). Gaft et al. (2005) report the presence of multiple luminescence centers attributed to Cr3+ substituting for Al3+ in chrysoberyl located at 650, 655, 664, 679, 680, 693, 694, 700, 707, and 716 nm. The positions of Cr3+ peaks in the Raman spectra of chrysoberyl in this study are centered at: 650 nm, 655 nm, 664 nm 679 nm, 680 nm, 693 nm, and 700 nm (Fig. 2.1 A-D). Gaft et al. (2005) hypothesize that a luminescence center located at 703 nm may be attributed to V2+ (Fig. 2.1 A., B., & D.). Fig. 2.1 A. Peaks in the Raman spectra of chrysoberyl related to Cr3+ luminescence centers centered at 650, 655, 664, 679, 680, 693, and 700 nm; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R080110, Sri Lanka, λexcitation = 532 nm, unoriented

Cr3+

3+

3+

Cr

&

V?

Cr

27

Fig. 2.1 B. Magnification of Cr3+ peaks centered at 650 nm (5323370 cm-1), 655 nm (5323540 cm-1), 664 nm (5323745 cm-1), 679 nm (5324062 cm-1), 680 nm (5324100 cm-1), 693 nm (5324367 cm-1), and 700 nm (5324510 cm-1) and a peak 532 -1 centered at 703 nm ( 4573 cm ) possibly related to V in the Raman spectrum of chrysoberyl; sample R080110, Sri Lanka, unoriented

Intensity

Cr3+

Cr3+

3+

Cr

V?

3+

Cr

Raman Shift (cm-1)

Fig. 2.1 C. (left) Further magnification of weak Cr3+

Intensity

peaks centered at 650 nm (5323370 cm-1), 655 nm 532 -1 532 -1 ( 3540 cm ), and 664 nm ( 3745 cm ); sample R080110, unoriented Cr3+

Raman Shift (cm-1)

Fig. 2.1 D. (right) Further magnification of weak Cr3+

Intensity

Cr3+

Cr3+

Cr3+

V?

peaks centered at 693 nm (5324367 cm-1), and 700 nm (5324510 cm-1) and possibly a peak attributed to V2+ centered at 703 nm 532 -1 ( 4573 cm ); sample R080110, unoriented

Raman Shift (cm-1)

28

References Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Gubelin, E. (1976) Alexandrite from Lake Manyara, Tanzania. Gems & Gemology, pp. 203-209. Gubelin, E. & Schmetzer, K. (1982) Gemstones with alexandrite effect. Gems & Gemology, Vol. 18, pp. 197-203. Hassan, F. & El-Rakhawy, A. (1974) Chromium III centers in synthetic alexandrite. American Mineralogist, Vol. 59, pp. 159-165. Liu, Y., Shigley, J., Fritsch, E. & Hemphill, S. (1994) The "alexandrite effect” in gemstones. Color Research & Application, Vol. 19, No. 3, pp. 186-191. Weber, S.U., Grodzicki, M., Lottermoser, W., Redhammer, G.J., Tippelt, G., Ponahlo, J. & Amthauer, G. 57 Fe Mössbauer spectroscopy, X-ray single-crystal diffractometry, and electronic structure calculations on natural alexandrite. Physics and Chemistry of Minerals, pp. 1-9.

29

Corundum Gem names: ruby, sapphire, padparadascha Ideal chemistry: Al2O3 Crystal system: trigonal Point Group:



H-M: 3 m

S: D3d



Space Group: R 3 c Synthetic sapphire

Table of Atomic Coordinates (Lewis et al., 1982): atom

x

y

z

Al

0

0

0.35216

O

0.30624

0

0.25

Raman mode analysis:

Raman Active Modes Atom Wyckoff Position Point Symmetry A1g Eg O

18e

2

1

3

Al

12c

3

1

2

Raman mode analysis predicts the existence of 7 active Raman modes in corundum: 2A1g + 5E g = 7

30

Introduction to Raman Spectrum of Corundum The crystal structure of corundum consists of dense, closest packed layers of oxygen and octahedrally coordinated aluminum (Fig. 3.2). In general, we have found that minerals composed solely of octahedral polyhedra are poor Raman scatterers. A perfect octahedron has an inversion center at the cation site, therefore, no significant change in polarization of the atom can occur. An exception to this observation is corundum. In the structure of corundum, edge and face sharing of the Al-octahedra cause distortion, resulting in appreciable polarization of the atoms and intense Raman peaks. The Raman spectrum of corundum is provided below. Fig. 3.1 The Raman spectrum of corundum with mode assignments as reported by Porto and Krishnan (1967); sample X080003, synthetic yellow corundum, 780 nm laser, processed, unoriented

Intensity

418 A1g

378 Eg 432 Eg 578 Eg

645 A1g

751 Eg

451 Eg

Raman Shift (cm-1)

Fig. 3.2 The crystal structure of corundum; blue octahedra: AlO6 groups A. View down c-axis

B. View down a-axis

31

Orientational Dependence of Spectra Xu et al. (1995) and Porto and Krishnan (1967) report that when the incident beam is polarized parallel to the c-axis, the following three Raman peaks in corundum disappear: 576 (Eg), 643(A1g), and 749 (Eg) cm-1. When corundum is oriented with the a-axis parallel to the incident laser beam (the cross-section of the optical indicatrix is an ellipse), the intensities of the peaks change dramatically with rotation due to a change in the degree of freedom of atomic vibration (Fig. 3.3 A). When the sample is oriented with the caxis parallel to the laser (the cross-section of the optical indicatrix is a circle), there is no noticeable change in peak intensities due to an equal distribution of atomic vibration in all directions (FIG. 3.3 B). In spite of the orientational dependence of its Raman peaks, corundum can be accurately identified by the spectrum of a randomly oriented sample.

Fig. 3.3. Raman spectra of oriented corundum crystals; sample R040096, purple corundum, Sri Lanka, processed A. Raman spectra showing peak intensities as a function of orientation; 514.5 nm laser parallel to a*; at 0° laser is polarized ll to c-axis; at 90° laser is polarized perpendicular to c-axis, sample R040096, processed

B. Raman spectra showing peak intensities as a function of orientation. Notice that the intensities do not change as the direction of polarization changes; 514.5 nm laser parallel to c-axis; at 0° laser is polarized ll to a-axis; at 90° laser is polarized perpendicular to a-axis

32

Spectral Features Related to OHAn OH- band centered at 3310 cm-1 was recently reported by Beran and Rossman (2006) in natural corundum from worldwide localities during an IR study. This band is most often found in blue-colored corundum (sapphire) suggesting that the OH- is involved in a chemical reaction with Fe in the structure. Prior to this study, OH- bands in corundum were rarely observed in natural crystals, appearing instead in synthetic material (Beran and Rossman, 2006). There are no observable peaks associated with OHmodes in the Raman spectra of corundum in this study.

Spectral Features Related to Chromophores and Other Ions Gem quality corundum comes in a wide variety of colors, the mechanisms of which vary (Appendix D). In blue sapphires a charge transfer between Fe2+-O-Ti4+ in combination with charge transfer between ferric and ferrous iron create the diagnostic sapphire color. In rubies, octahedrally coordinated Cr3+ is the dominant color-inducing ion (Smith et al., 1997; Fritsch and Rossman, 1988). Trace amounts V3+ and Fe3+ may offer minor contributions to the red color in some rubies (Fritsch and Rossman, 1988). The positions of Cr3+ luminescence centers located at 693 nm and 694 nm, denoted R2 and R1 respectively, have been well established in the literature (Fig. 3.4 A. & B.) (Collins et al., 1960). The positions of the peaks in this doublet shift systematically with pressure (Collins et al., 1960; Mao et al., 1978). Therefore, the positions of Cr3+ peaks have been calibrated to determine pressures in diamond anvil cells (Mao et al., 1978). Fig. 3.4 A. Peaks in the Raman spectrum of corundum associated with Cr3+ luminescence centers located at 693

532 -1 532 -1 nm ( 4366 cm ) (R2) and 694 nm ( 4396 cm ) (R1); note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060020, colorless sapphire from Montana, unoriented

R1 Cr3+ R2

33

Fig. 3.4 B. Magnification of peaks in Raman spectrum of corundum associated with Cr3+ luminescence centers; sample R060020, colorless sapphire from Montana, λexcitation = 532 nm, unoriented

R1 Cr3+

Intensity

R2

Raman Shift (cm-1)

References Beran, A. & Rossman, G.R. (2006) OH in naturally occurring corundum. European Journal of Mineralogy, Vol. 18, No. 4, pp. 441. Collins, R.J., Nelson, D.F., Schawlow, A.L., Bond, W., Garrett, C.G.B. & Kaiser, W. (1960) Coherence, narrowing, directionality, and relaxation oscillations in the light emission from ruby. Physical Review Letters, Vol. 5, No. 7, pp. 303-305. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Lewis, J., Schwarzenbach, D. & Flack, H.D. (1982) Electric field gradients and charge density in corundum, Al2O3 . Crystal Physics, Diffraction, Theoretical and General Crystallography, Vol. 38, No. 5, pp. 7394. Mao, H.K., Bell, P.M., Shaner, J.W. & Steinberg, D.J. (1978) Specific volume measurements of Cu, Mo, Pd, and Ag and calibration of the ruby R fluorescence pressure gauge from 0.06 to 1 Mbar. Journal of Applied Physics, Vol. 49, pp. 3276. Smith, C.P., Gübelin, E.J., Bassett, A.M. & Manandhar, M.N. (1997) Rubies and fancy-color sapphires from Nepal. Gems & Gemology, Vol. 33, No. 1, pp. 24–41. Xu, J., Huang, E., Lin, J. & Xu, L.Y. (1995) Raman study at high pressure and the thermodynamic properties of corundum; application of Kieffer's model. American Mineralogist, Vol. 80, No. 11-12, pp. 1157-1165.

Additional Information Fritsch, E. & Rossman, G.R. (1988) An update on color in gems. Part 2: Colors involving multiple atoms and color centers. Gems & Gemology, Vol. 24, No. 1, pp. 3–15. Fritsch, E. & Rossman, G.R. (1987) An Update on Color in Gems. Part 1. Introduction and Colors caused by Dispersed Metal Ions. Gems and Gemology, Vol. 23, No. 3, pp. 126–139. Mao, H.K. & Bell, P.M. (1976) High-Pressure Physics: The 1-Megabar Mark on the Ruby R1 Static Pressure Scale. Science, Vol. 191, No. 4229, pp. 851-852. Porto, S.P.S. & Krishnan, R.S. (2004) Raman Effect of Corundum. The Journal of chemical physics, Vol. 47, pp. 1009. Szabo, A. (1970) Laser-induced fluorescence-line narrowing in ruby. Physical Review Letters, Vol. 25, No. 14, pp. 924-926.

34

Diamond Gem names: canary, cape Ideal chemistry: C Crystal system: isometric Point Group:

H-M: m3m S: Oh

Space Group: Fd3m

Table of Atomic Coordinates (Wyckoff, 1963): atom

x

y

z

C

0

0

0

Raman mode analysis:

Raman Active Modes Atom Wyckoff Position Point Symmetry T2g C

8a

4 3m

1

Raman mode analysis predicts the existence of 1 active Raman mode in diamond: 1T2g = 1

35

Raman Features of Diamond Diamonds are likely the most valuable of all the gemstones. The covalently bonded carbon atoms in the face-centered cubic crystal structure of diamond (Fig. 4.2) make it the hardest natural substance on Earth. For this reason, diamonds are commonly used as an industrial abrasive. The hardness of diamond, combined with its rarity and optical properties, also make it a popular gemstone. Spectra of diamond have one diagnostic active Raman mode, the carbon-carbon stretching mode, that produces a peak centered at 1332 cm-1 (Knight and White, 1989). Slight shifts in the frequency of the primary diamond peak occur depending on several factors including origin of the diamond and perfection of the diamond lattice (Fig. 4.1) (Huong, 1992). Frequency and width of this peak are also temperature dependent (Fig. 4.3) (Liu et al., 2000). In addition, isotopic effects can result in a frequency shift of the primary diamond peak. According to Chrenko (1988), natural diamonds typically contain 1 at.% of 13C. To investigate isotopic effects on the primary peak in the Raman spectra of diamond, several diamonds containing up to 91at.% 13C were grown by the temperature gradient method (Chrenko, 1988). In all cases, there was a single Raman peak in each spectrum, however, the location of this peak shifted from 1332.5 cm-1 at 1% 13C to 1288.7 cm-1 at 91% 13C. The absence of a second peak suggests that in this experiment the carbon isotopes were distributed homogenously throughout the diamond structure (Chrenko, 1988). In experimentally synthesized 12C:13C diamond films grown on natural 12C diamond substrates, two 1st order Raman peaks were observed (Behr et al., 1993). In the synthesized pure 13C film, in addition to the peak at 1333 cm-1 associated with the 12C substrate, a peak attributed to 13C appeared at 1283 cm-1 (Behr et al., 1993). In the 50%12C/50%13C film, the 13C peak shifted to 1313 cm-1 (Behr et al., 1993). 12C and 13C spectral differences have not been observed in this study. Additional weaker peaks at 1817, 1864, 2025, 2177, 2254, 2333, 2458, 2519, and 2667 cm-1 represent second-order Raman scattering (Fig. 4.4 A. & B.) (Solin and Ramdas, 1970). Analysis of these peaks is complex however, increasing the 13C content of diamond results in a frequency shift of these peaks suggesting that these peaks are related to carbon and not to trace impurities (Chrenko, 1988). Very weak peaks centered at 3300 and 3825 cm-1 represent third-order Raman scattering (Fig. 4.4 A. & 4.5) (Bormett et al., 1995). In this study, no visible differences were observed in the Raman spectra of natural and synthetic diamonds (Fig. 4.6). The shift in frequency of a Raman peak due to isotopic effects can be predicted using a simple equation. Gillet et al. (1996) used the following calculation to predict the positions of Raman peaks in calcite containing various amounts of 16O and 18O: v1/v1* = √ m16O/ m18O v1= theoretical peak postion; v1*= actual peak position; m = mass of the isotopes

This equation can also be used to predict the peak position of the primary Raman peak in diamonds composed of 13C. For example, the mass of 12C is 12 and the mass of 13C is 13. When you plug these values into the equation above you get: v1/1332 cm-1 = 0.96 and so the theoretical peak location in diamond composed of 100% 13C is v1 = 1278 cm-1. This calculated position is very close to the peak position of diamond composed of 91% 13C centered at 1288.7 cm-1 reported by Chrenko (1988). 5. 5. 4.

Intensity

Fig. 4.1 (right) Slight shift in the frequency of the primary Raman peak in diamond; 1) natural type IIa diamond, sample R050204, 2) natural type IIa diamond, sample R050206, 3) synthetic yellow diamond, sample X080011; 4) natural type I diamond, sample R050207; 5) natural type IIa diamond, sample R050205; λexcitation = 532 nm, background corrected, unoriented

3. 2. 1.

Raman Shift (cm-1)

36

Intensity

Fig. 4.2 The face-centered cubic crystal structure of diamond; blue spheres: carbon atoms

2.

1.

1334 cm

-1

Fig. 4.3 (left) Shift in frequency and change in width of primary Raman peak in diamond when cooled to liquid nitrogen temperature; yellow synthetic diamond, sample X080011, λexcitation = 532 nm, unoriented; 1) room temperature, 2) liquid nitrogen temperature

1336 cm-1

-1

Raman Shift (cm )

Fig. 4.4 A. Second order Raman scattering and third order Raman scattering (not visible in this figure); natural type IIa diamond; sample R050205; λexcitation = 532 nm, unoriented

rd

2

nd

Order Raman Scattering

3 Order Raman Scattering

37

Fig. 4.4 B. Magnification of second order Raman scattering; natural type IIa diamond; sample R050205; λexcitation =

Intensity

532 nm, unoriented

-1

2465 cm

-1

2667 cm

-1

2030 cm

Raman Shift (cm-1)

Intensity

Fig. 4.5 Magnification of third order Raman scattering; natural type IIa diamond; sample R050206; λexcitation = 532 nm, unoriented

-1

-1

3300 cm

3825 cm

Raman Shift (cm-1)

Fig. 4.6 Spectra of natural and synthetic diamond showing negligible spectral differences; 1) synthetic diamond,

Intensity

sample X080011; 2) natural diamond, sample R050207; λexcitation = 532 nm, processed, unoriented

2. 1.

Raman Shift (cm-1)

38

Causes of Color in Diamond Diamonds are divided into two types depending on the dominant chromophores present: Type I or Type II (Deljanin and Simic, 2007). Type I diamondsType I diamonds are subdivided based on the way nitrogen is dispersed throughout the crystal structure. Type Ia diamonds contain aggregates of nitrogen: IaA contain a pair of nitrogen atoms (A-center); IaB contain four nitrogen atoms surrounding a vacancy (B-center); IaAB contain a mixture of both B-centers and A-centers. 97% of all diamonds are Type Ia. They can appear colorless, near-colorless, yellow, or brown. Type Ib diamonds account for less than 1% of all diamonds. Diamonds of this type contain isolated atoms of nitrogen that replace carbon atoms (C-centers). Type Ib diamonds are yellow, orange, or brown (Deljanin and Simic, 2007). Type II diamondsType II diamonds are generally absent of nitrogen (containing less than 1ppm). Type II diamonds make up approximately 2% of all diamonds. Diamonds containing no nitrogen and no boron are called Type IIa. They can be colorless, near-colorless, brown, and pink. Type IIb diamonds contain boron. When boron replaces carbon in the diamond lattice the nonequivalent charge creates a hole in the diamond lattice allowing for Type IIb diamonds to conduct a positive electrical charge. Type IIb diamonds are blue, gray, light brown, or near-colorless (Deljanin and Simic, 2007). Non-nitrogen Related Colors: Pink/red: Plastic deformation of the diamond lattice can result in displacement of the carbon atoms along glide planes. This is the cause of both pink and red colors in diamond (Fritsch et al., 2007b; Shigley, 1993). Purple: Like pink diamonds, purple diamonds are also believed to be the result of plastic deformation. The details of color-inducing defects in purple diamonds are still being explored (Titkov et al., 2008). Brown: Brown diamonds typically contain nitrogen impurities and graining associated with deformation of the diamond lattice (Massi, 2005). Black: Black to gray diamonds contain micro-inclusions of dark minerals such as graphite, magnetite, hematite, and native iron (Titkov et al., 2003). Green: Green colored diamonds are produced by irradiation. Surface “staining” of natural diamonds is commonly caused by the interaction of radioactive alpha and beta particles with the diamond surface, producing a green-colored “skin” on the diamond (Kane et al., 1990). The historic Dresden green diamond has green color throughout the body of the stone and not just at the surface therefore the color of this famous gemstone is likely the result of deeper penetrating ionizing radiation. Exposure to this type of radiation forced some of the carbon atoms out of their sites producing vacancies in the diamond structure. These vacancies produce what is known as the GR1 color center. This center absorbs light from the red portion of the spectrum, resulting in a diamond that is green (Kane et al., 1990). Chameleon Diamonds: As the name suggests, chameleon diamonds change color, from grey-green to yellow, with a change in temperature (thermochromic behavior) or with change in light exposure (photochromic). These diamonds contain high concentrations of hydrogen along with some nitrogen and nickel. Chameleon diamonds are type IaA/B and the suggested model for the color-change involves an electron trap caused by the interaction of hydrogen atoms with A-aggregate nitrogen (Fritsch et al., 2007a).

39

Spectral Artifacts Related to Nitrogen in Synthetic Diamonds The 638 nm center, or NV- center, is the result of an isolated nitrogen atom trapping a vacancy in the diamond structure (Zaitsev, 2001). An isolated nitrogen atom substituting for a carbon atom is bound to the nearest vacancy filled with a single electron resulting in the nitrogen atom relaxing away from the vacancy at a distance equal to approximately 8% of the normal C-C bond (Zaitsev, 2001; Loubser and van Wyk, 1978; Mainwood, 1994). This center occurs naturally in any nitrogen-containing diamonds that have been irradiated with high-energy ions, however, it is especially noticeable in type Ib diamonds (Zaitsev, 2001; Nishida et al., 1989). In this study, it has only been observed in synthetic diamonds (Fig. 4.7). At liquid nitrogen temperature the 638 nm peak becomes more resolved (Fig. 4.8). The weaker peaks accompanying the 638 nm peak are also related to the NV- center (Zaitsev, 2001). A note on HPHT and GE POL treatments: Fisher and Spits (2000) report that diamonds exposed to HPHT (high-pressure and high-temperature) treatment commonly display the 638 nm peak and its neutrally charged counterpart at 575 nm. They add that the presence of isolated nitrogen in type IIa diamonds is indicative of HPHT treatment, a technique used to change the color of undesirable diamonds. Isolated nitrogen atoms present in GE POL diamonds (natural diamonds enhanced by General Electric using undisclosed methods of HPHT and annealing techniques) were likely generated from the disaggregation of A- and B-centers at temperatures above 1960oC and 2240oC respectively (Fisher and Spits, 2000). The presence of the luminescence features described above provides valuable information about synthetic diamonds and further investigation of these features may lead to the development of a method to distinguish between synthesized, treated, and natural diamonds.

Intensity

Fig. 4.7 A comparison of the spectral features related to the 638 nm (5323100 cm-1) NV- center in the Raman spectra of a natural diamond and two synthetic diamonds; 1) natural type IIa diamond, sample R050205; 2) synthetic yellow diamond, sample X080011; 3) DeBeers Element6 type IIa synthetic diamond, sample X07001; λexcitation = 532 nm, room temperature, unoriented

3. NV2.

1. Raman Shift (cm-1)

40

Intensity

Fig. 4.8 The NV- center (638 nm) at 1) room temperature and at 2) liquid nitrogen temperature; synthetic yellow diamond, sample X080011; λexcitation = 532 nm, unoriented

NV-

Raman Shift (cm-1)

Diamond Simulants Common diamond simulants such as colorless zircon, cubic zirconia, yttrium aluminum garnet (YAG), gadolinium gallium garnet (GGG), and strontium titanate are easily identified by Raman spectroscopy (Fig. 4.9). Fig. 4.9 Comparison of the Raman spectra of natural diamond with the spectra of common diamond simulants. 1) Diamond, sample R050204; 2) Zircon, sample R050488; 3) YAG, sample X090003; 4) GGG, sample X090005; 5) GGG, sample X090007; 6) Cubic zirconia, sample R040142; 7) Strontium titanate (synthetic tausonite); sample X090004; λexcitation = 532 nm, processed, unoriented

7.

Intensity

6. 5. 4. 3. 2. 1. -1 Raman Shift (cm )

41

References 13 Behr, D., Wagner, J., Wild, C. & Koidl, P. (1993) Homoepitaxial C diamond films studied by micro-Raman and photoluminescence spectroscopy. Applied Physics Letters, Vol. 63, No. 22, pp. 3005-3007. Bormett, R.W., Asher, S.A., Witowski, R.E., Partlow, W.D., Lizewski, R. & Pettit, F. (1995) Ultraviolet Raman spectroscopy characterizes chemical vapor deposition diamond film growth and oxidation. Journal of Applied Physics, Vol. 77, pp. 5916. Chrenko, R.M. (1988) 13C-doped diamond: Raman spectra. Journal of Applied Physics, Vol. 63, No. 12, pp. 58735875. Deljanin, B. & Simic, D. (2007) Laboratory-Grown Diamonds: Information Guide to HPHT-grown and CVD-grown Diamonds, 2nd ed, Gemology Headquarters International, India. Fisher, D. & Spits, R.A. (2000) Spectroscopic evidence of GE POL HPHT-treated natural type IIA diamonds. Gems & Gemology, Vol. 36, No. 1, pp. 42-49. Fritsch, E., Massi, L., Rossman, G.R., Hainschwang, T., Jobic, S. & Dessapt, R. (2007a) Thermochromic and photochromic behaviour of “chameleon” diamonds. Diamond & Related Materials, Vol. 16, No. 2, pp. 401-408. Fritsch, E., Rondeau, B., Hainschwang, T. & Quellier, M.H. (2007b) A contribution to the understanding of pink color in diamond: The unique, historical Grand Condé. Diamond & Related Materials, Vol. 16, No. 8, pp. 1471-1474. Gillet, P., McMillan, P., Schott, J., Badro, J. & Grzechnik, A. (1996) Thermodynamic properties and isotopic 18 fractionation of calcite from vibrational spectroscopy of O-substituted calcite. Geochimica et Cosmochimica Acta, Vol. 60, No. 18, pp. 3471-3485. Huong, P.V. (1992) Diamond and diamond simulants as studied by micro-Raman spectroscopy. Materials science & engineering.B, Solid-state materials for advanced technology, Vol. 11, No. 1-4, pp. 235-242. Kane, R.E., McClure, S.F. & Menzhausen, J. (1990) The legendary Dresden green diamond. Gems & Gemology, Vol. 26, No. 4, pp. 248–266. Knight, D.S. & White, W.B. (1989) Characterization of diamond films by Raman spectroscopy. Journal of Materials Research, Vol. 4, No. 2, pp. 385-393. Liu, M.S., Bursill, L.A., Prawer, S. & Beserman, R. (2000) Temperature dependence of the first-order Raman phonon line of diamond. Physical Review B, Vol. 61, No. 5, pp. 3391-3395. Loubser, J.H.N. & van Wyk, J.A. (1978) Electron spin resonance in the study of diamond. Rep.Prog.Phys, Vol. 41, No. 8, pp. 1201. Mainwood, A. (1994) Nitrogen and nitrogen-vacancy complexes and their formation in diamond. Physical Review B, Vol. 49, No. 12, pp. 7934-7940. Massi, L., Fritsch, E., Collins, A.T., Hainschwang, T. & Notari, F. (2005) The “amber centres” and their relation to the brown colour in diamond. Diamond & Related Materials, Vol. 14, No. 10, pp. 1623-1629. Nishida, Y., Mita, Y., Okuda, S., Mihara, T., Kato, R., Ashida, M., Sato, S. & Yazu, S. (1990) Color centers in synthetic Ib diamonds and their application to opto-electronics in Science and Technology of New Diamond, eds. S. Saito, O. Fukunaga & M. Yoshikawa, KTK Scientific, , pp. 363-367. Shigley, J.E. & Fritsch, E. (1993) A notable red-brown diamond. Journal of Gemmology, Vol. 23, No. 5, pp. 259–266. Solin, S.A. & Ramdas, A.K. (1970) Raman Spectrum of Diamond. Physical Review B, Vol. 1, No. 4, pp. 1687-1698. Titkov, S.V., Shigley, J.E., Breeding, C.M., Mineeva, R.M., Zudin, N.G. & Sergeev, A.M. (2008) Natural-Color Purple Diamonds from Siberia. Gems & Gemology, Vol. 44, No. 1, pp. 56. Titkov, S.V., Zudin, N.G., Gorshkov, A.I., Sivtsov, A.V. & Magazina, L.O. (2003) An investigation into the cause of color in natural black diamonds from Siberia. Gems and Gemology, Vol. 39, No. 3, pp. 200–209. Wyckoff, R.W.G. (1963) Crystal Structures. Vol. 1, Interscience Publishers. Zaitsev, A.M. (2001) Optical properties of diamond, Springer New York.

Additional Information Collins, A.T. (1980) Vacancy enhanced aggregation of nitrogen in diamond. J.Phys.C: Solid St.Phys, Vol. 13, pp. 2641-2650. Collins, A. (1982) Colour centres in diamond. Journal of Gemmology, Vol. 18, pp. 37-75. King, J.M., Shigley, J.E., Gelb, T.H., Guhin, S.S., Hall, M. & Wang, W. (2005) Characterization and Grading of Natural-Color Yellow Diamonds. Gems & Gemology, Vol. 41, No. 2, pp. 88. McNamara, K.M., Gleason, K.K., Vestyck, D.J. & Butler, J.E. (1992) Evaluation of Diamond Films by Nuclear Magnetic Resonance and Raman Spectroscopy. Diamond & Related Materials, Vol. 1. Moss, T.M., King, J.M., Wang, W. & Shigley, J.E. (2002) A highly unusual, 7.34 ct, Fancy vivid purple diamond. Journal of Gemmology, Vol. 28, No. 1, pp. 7-12.

42

Diopside Gem names: chrome diopside Ideal chemistry: CaMgSi2O6 Crystal system: monoclinic Point Group:

H-M: 2/m

S: C2h

Space Group: C2/c

Chrome diopside, Photo courtesy of Stone Group Labs

Table of Atomic Coordinates (Redhammer, 1998): atom

x

y

z

MgM1

0

0.9078

0.25

CaM2

0

0.3014

0.25

SiT

0.2860

0.0916

0.2302

O1

0.1184

0.0871

0.1432

O2

0.3623

0.2498

0.3175

O3

0.3487

0.0201

0.9983

Raman mode analysis:

Raman Active Modes Atom

Wyckoff Position Point Symmetry Ag Bg

Si, O1, O2, O3

8f

1

3

3

Mg, Ca

4e

2

1

2

(4 x 8f) + (2 x 4e) Raman mode analysis predicts the existence of 30 active Raman modes in diopside: 14Ag + 16B g = 30

43

Introduction to the Raman Spectrum of Diopside Diopside is an inosilicate consisting of single chains of SiO4 groups (Fig. 5.2). Diopside is a pyroxene and is commonly found in ultramafic igneous and metamorphic rocks. Phase equilibrium studies indicate that diopside is stable at temperature and pressure conditions similar to those of the upper mantle (Swamy et al., 1997). The Raman spectrum of diopside is presented below (Fig. 5.1 A-C). The most intense peak in the spectrum, centered at 1014 cm-1, is associated with Si-O stretching vibrations with non-bridging oxygen atoms (Richet et al., 1997). Swamy et al. (1997) report that the frequencies of Raman peaks in diopside decrease with an increase in temperature. Richet et al. (1997) report that increases in temperature also affect the peak width. The peaks centered at ~600 cm-1, associated with Si-O-Si bending modes, widen with increases in temperature (Richet et al., 1997). Fig. 5.1 A. The Raman spectrum of diopside with mode assignments reported by Swamy et al. (1997); sample R060171, 532 nm laser, processed, unoriented

1014 Ag

Intensity

667 Ag 139 Ag

323 Ag

390 Ag 367 Bg

See Fig. 5.1 C

See Fig. 5.1 B

&

357 Ag

508 530 Ag Ag

710 Bg

856 Ag

917 Bg

1048 ? Bg

Raman Shift (cm-1)

Fig. 5.1 B. Magnification of Raman peaks in 460-580 cm-1 range; mode assignments reported by Swamy et al. (1997) sample R060171, 532 nm laser, processed, unoriented 515 Bg

530 Ag

Intensity

508 Ag

465 Bg

560 Bg

Raman Shift (cm-1)

44

Fig. 5.1 C. Magnification of Raman peaks in 160-300 cm-1 range; mode assignments reported by Swamy et al. (1997) sample R060171, 532 nm laser, processed, unoriented

Intensity

181 Ag

163 Bg

254 Ag 194 Bg

301 Bg

229 Bg

Raman Shift (cm-1)

Fig. 5.2 The crystal structure of diopside; green octahedra: MgO6 groups, blue tetrahedra: SiO4 groups, orange ellipsoids: Ca

45

Spectral Features Related to Chromophores and Other Ions Gem quality diopside is most commonly green or yellowish green though the mechanisms responsible for these colors can vary. Octahedrally coordinated Cr3+, V3+, a combination of Cr and V, or a charge transfer between Fe2+-Fe3+ can produce green color in diopside (Fritsch and Rossman, 1988; Andrut et al., 2003). A rare purple color in diopside is generated by charge transfer between Fe2+ and Ti4+ (Herd et al., 2000). Comparison of the peak positions of spectral features in diopside, centered at 671 nm, 679 nm, 684 nm, 700 nm and 722 nm (Fig. 5.3 A. & B.), with the positions of luminescence centers in the spectra of other minerals studied by Gaft et al. (2005) suggests that these peaks are likely related to Cr and V. This has yet to be confirmed by experimental luminescence studies. Fig. 5.3 A. Peaks centered at 671 nm (5323900 cm-1), 679 nm (5324082 cm-1), 684 nm (5324196 cm-1), 700 nm

(5324526 cm-1), and 722 nm (5324958 cm-1) in the Raman spectrum of diopside likely related to Cr and V luminescence centers; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060085, green diopside, Tyrol, Austria, unoriented

Cr?

V? Cr?

Fig. 5.3 B. Magnification of features likely related to Cr and V luminescence centers in the Raman spectrum of diopside; sample R060085, green diopside, Tyrol, Austria, λexcitation = 532 nm, unoriented

Intensity

Cr?

V?

Cr?

Raman Shift (cm-1)

46

References Andrut, M., Brandstätter, F. & Beran, A. (2003) Trace hydrogen zoning in diopside. Mineralogy and Petrology, Vol. 78, No. 3, pp. 231-241. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Fritz, E.A., Laurs, B.M., Downs, R.T. & Costin, G. (2007) Yellowish green diopside and tremolite from Merelani, Tanzania. GEMS AND GEMOLOGY, Vol. 43, No. 2, pp. 146. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Herd, C.D.K., Peterson, R.C. & Rossman, G.R. (2000) Violet-colored diopside from southern Baffin Island, Nunavut, Canada. Canadian Mineralogist, Vol. 38, No. 5, pp. 1193. Redhammer, G.J. (1998) Mossbauer spectroscopy and Rietveld refinement on synthetic ferri-Tschermak's molecule CaFe3+(Fe3+Si)O6 substituted diopside. European journal of mineralogy, Vol. 10, No. 3, pp. 439-452. Richet, P., Mysen, B.O. & Ingrin, J. (1998) High-temperature X-ray diffraction and Raman spectroscopy of diopside and pseudowollastonite. Physics and Chemistry of Minerals, Vol. 25, No. 6, pp. 401-414. Swamy, V., Dubrovinsky, L.S. & Matsui, M. (1997) High-temperature Raman spectroscopy and quasi-harmonic lattice dynamic simulation of diopside. Physics and Chemistry of Minerals, Vol. 24, No. 6, pp. 440-446.

47

Garnet General Formula: X3Z2(SiO4)3 Crystal system: isometric Gem names: pyrope, almandine, grossularite, spessartite, andradite (demantoid), uvarovite, rhodolite, tsavorite, hessonite

Point Group:



H-M: m 3 m

S: Oh Spessartine, Fujian Province,China

Space Group: Ia3d

Table of Garnet Chemistries Garnet Variety

Ideal Chemical Formula

pyrope

Mg3Al2(SiO4)3

almandine

Fe2+3Al2(SiO4)3

grossular

Ca3Al2(SiO4)3

spessartine

Mn2+3Al2(SiO4)3

andradite

Ca3Fe3+2(SiO4)3

uvarovite

Ca3Cr2(SiO4)3

Note: The members of the garnet group are isostructural (they have the same crystal structure), therefore, atomic coordinates and Raman mode analysis only for pyrope (the most common gem garnet) have been provided because its structure is representative of the other garnet species.

Table of Atomic Coordinates (Merli et al., 2000): atom

x

y

z

Mg

0.125

0

0.25

Al

0

0

0

Si

0.375

0

0.25

O

0.033

0.0503

0.6533

48

Raman mode analysis:

Raman Active Modes Atom Wyckoff Position

A1g

Eg

T2g

O

96h

3

6

9

Si

24d

-

1

3

Mg

24c

-

1

2

Al

16a

-

-

-

Raman mode analysis predicts the existence of 25 active Raman modes in all garnet species: 3A1g + 8E g + 14T2g = 25

49

Garnet Chemistry The garnet group consists of multiple isostructural species that are chemically distinguishable from one another based on a variety of possible atomic substitutions in the X (divalent dodecahedral) and Y (trivalent octahedral) sites of the crystal structure (see “Table of Garnet Chemistries” on previous page) (Novak and Gibbs, 1971). Pyrope, almandine, and spessartine are three end-member species of a garnet solid-solution series and are denoted, chemically separate from the other garnet species, as ‘pyralspite’ (Winchell, 1933). The calcium-rich garnet species uvarovite, grossular, and andradite are classified as ‘ugrandite’ (Winchell, 1931). Rarely are crystals of pure end-member chemistry found in nature, therefore, garnets are frequently described as having multiple chemical components (Novak and Gibbs, 1971). A recent study conducted by Henderson (2009) at the University of Arizona has demonstrated that the Raman spectra of garnets can be used to calculate their crystal chemistry. Raman spectroscopy of garnets can be utilized as a quick, non-destructive alternative technique to microprobe analysis for constraining the chemical composition of a sample. Using a correlation matrix to compare shifts in the frequency of six peaks associated with stretching modes in the Raman spectra of garnet to the change in chemical composition based on microprobe data, the chemistry of multiple garnet varieties can be predicted to within 5% (overall error of bulk composition) of the microprobe values (Fig. 6). Henderson (2009) also noted that reported a correlation between chemistry and peak intensity: in Raman spectra of calcic garnets (those containing >50% Ca in the X-site, such as the ugrandites) the most intense peak is -1 peak 6 (centered at ~350 cm ), while in garnets containing less than 50% Ca (pyralspites), the most intense peak is peak 2 (centered at ~850 cm-1) (Fig. 6). Note: the octahedral (Y) site sits on a center of inversion and therefore, there are no Raman peaks associated with the vibrations of atoms occupying this site (Hofmeister and Chopelas, 1991). Fig. 6 Raman spectra of end-member garnets; the six labeled peaks are the ones used in the chemical composition calculation (reproduced with permission from Henderson, 2009)

50

Cause of Color Garnets come in a wide variety of colors and a discussion of the various causes of color is provided below (see also Table of Color Causes, Appendix). Pyrope (Mg3Al2(SiO4)3): Brown-red pyropes contain Fe2+ while pure red pyropes typically contain a combination of Fe2+ and octahedrally coordinated Cr3+ (Fritsch and Rossman, 1988; Manning, 1967). Almandine (Fe2+3Al2(SiO4)3): The red color-inducing cation in almandine is Fe2+ and a garnet containing components of both pyrope and almandine, well-known to gemologists as ‘rhodolite’, also owes its distinctive reddish purple color to Fe2+ (Fritsch and Rossman, 1988; Manning, 1967). Spessartine (Mn2+3Al2(SiO4)3): The orange color of spessartine is produced by the presence of Mn2+ in distorted cubic coordination (Fritsch and Rossman, 1988; Gubelin, 1982; Manning, 1967). Uvarovite (Ca3Cr2(SiO4)3): The intense green color of uvarovite is due to the presence of the common chromophore, octahedrally coordinated Cr3+ (Fritsch and Rossman, 1988; Manning, 1969). Grossular (Ca3Al2(SiO4)3): The green variety of grossular, commonly called tsavorite, contains octahedrally coordinated V3+, while orange grossular (also known as hessonite) typically contains Mn2+ or Fe2+ (Fritsch and Rossman, 1988; Manning, 1970). Andradite (Ca3Fe3+2(SiO4)3): Yellow-green andradite contains Fe3+ in the octahedral site and the green variety called demantoid contains octahedrally coordinated Cr3+ (Fritsch and Rossman, 1988). Color-change garnets: Several garnet species exhibit a color-change, or alexandrite effect (Carstens, 1973; Gubelin, 1982; Schmetzer and Bernhardt, 1999). According to Schmetzer and Bernhardt (1999), there are two divisible groups of color-change garnets, 1) chromium-rich (> 3 wt%) pyropes with a green to blue-green color change, and 2) pyralspites containing octahedrally coordinated V and/or Cr with a red to purple-red color change. Gubelin (1982) describes chromium-rich pyropes that are blue-green in daylight and wine red under incandescent light due to the presence of Cr and/or V. Less Cr and V are necessary to cause color-change in garnets with a major spessartine component (Gubelin, 1982).

51

Spectral Features Related to Chromophores There are multiple luminescence features in the Raman spectra of garnet associated with both Cr and V. Gaft et al. (2005) attribute luminescence features located at 690, 695, 698, 703, and 717 nm to the presence of vanadium in grossular and a band at 694 nm to Cr3+ in rhodolite (pyrope-almandine mixture). Polarized UV-VIS absorption spectra of uvarovites conducted by Andrut and Wildner (2001) attribute features at 686, 701, and 695 nm to octahedrally coordinated chromium. Absorption studies of chromium-doped gallium garnet laser crystals report the presence of peaks associated with Cr3+ centers located at 692 and 696 nm (Struve and Huber, 1985). Based on the location of Cr3+ and V luminescence features in garnet described by the aforementioned authors and the location of features in other chromium- and vanadium-bearing mineral species such as topaz, beryl, spinel, corundum, chrysoberyl, and zoisite (see Appendices A & B), the following table has been provided to address the luminescence features present in each garnet specie and assign them to the corresponding chromophores (Table 6.1). In this study, there are more observable luminescence features present in the Raman spectra of colorless pyrope, than there are in deep red-colored pyrope (Fig. 6.3 C.). Typically, colorless minerals are colorless due to a lack of chromophores or color centers. The fact that there are actually more luminescence features in the Raman spectra of colorless pyrope samples is, therefore, unexpected. Colorless pyrope samples exhibit eight well-resolved peaks attributed to luminescence centers. Peaks centered at 683 nm, 686 nm, 690 nm, 695 nm, 693 nm, 699 nm, and 701 nm are likely related to Cr, while a single peak centered at 729 nm may be related to V (Gaft et al., 2005; Andrut and Wildner, 2001). There are only five peaks associated with Cr luminescence centers in red pyrope located at 683 nm, 686 nm, 689 nm, 692 nm, and 695 nm (Gaft et al., 2005; Andrut and Wildner, 2001). These peaks appear broader and less resolved than the corresponding peaks present in colorless pyrope.

52

Table 6.1 Peaks related to Cr3+ and V luminescence centers in garnets, Raman spectra: λexcitation = 532 nm, unoriented

Sample

Description/Origin

Garnet

Spectral Features nm

R040159

R060448

R070637

R060099

R060279

R080053

R060382

R060477

red, Meronitz, Bohemia

red, India colorless, Piedmont, Italy

red, Alaska

red-brown, Shigar Valley, Pakistan orange, East Africa

transparent tan, Tanzania

green, Ural Mts., Russia

Pyrope

Pyrope

Pyrope

Almandine

Spessartine

Spessartine

Grossular

Uvarovite

Ion

Figure

Reference Gaft et al., 2005 (topaz & beryl) Andrut and Wildner, 2001; Gaft et al., 2005 (spinel)

cm-1

683

532

4166

Cr3+

6.1 & 6.3 B

686

532

4225

Cr3+

6.1 & 6.3 B

689

532

4290

Cr3+?

693

532

4360

Cr3+

696

532

4431

Cr3+ or V?

6.1 & 6.3 B 6.1 & 6.3 B 6.1 & 6.3 B

693

532

4366

Cr3+

6.2

694

532

4393

Cr

3+

6.2

683

532

4172

Cr

3+

6.3

686

532

4236

Cr3+

6.3 C

690

532

4316

Cr3+?

6.3 C

693

532

4380

695

532

4427

Cr3+ or V?

699

532

4492

Cr3+?

6.3 C

701

532

4550

Cr3+

6.3 C

729

532

5095

V?

6.3 C

688

532

4277

Cr3+?

6.4

693

532

694

Cr

3+

6.3 C 6.3 C

4372

Cr

3+

6.4

532

4401

Cr3+

6.4

706

532

4642

V?

6.4

693

532

4370

Cr

710

532

4705

V?

693

532

694

3+

6.5

Gaft et al., 2005 (chrysoberyl & corundum) Struve and Huber, 1985; Gaft et al., 2005 Gaft et al., 2005 (chrysoberyl & corundum) Gaft et al., 2005 Gaft et al., 2005 (topaz & beryl) Andrut and Wildner, 2001; Gaft et al., 2005 (spinel) Gaft et al., 2005 (chrysoberyl & corundum) Struve and Huber, 1985; Gaft et al., 2005 Andrut and Wildner, 2001

Gaft et al., 2005 (chrysoberyl & corundum) Gaft et al., 2005 Gaft et al., 2005 (chrysoberyl & corundum)

6.5

4370

Cr

3+

6.6

532

4395

Cr3+

6.6

697

532

4450

Cr3+

6.7

701

532

4544

Cr3+

6.7

717 721

532

4840 532 4930

V V?

605

532

2262

Mn

6.7 6.7 6.7 A. & C.

696

532

4440

Cr3+or V?

6.8

701

532

4542

Cr3+

6.8

721

532

4930

V?

6.8

Gaft et al., 2005 (chrysoberyl & corundum) Gaft et al., 2005 Andrut and Wildner, 2001; Struve and Huber, 1985 Andrut and Wildner, 2001 Gaft et al., 2005 Gaft et al., 2005 Struve and Huber, 1985; Gaft et al., 2005 Andrut and Wildner, 2001

53

Pyrope Fig. 6.1 A. Peaks likely related to Cr3+ and possibly V luminescence centers; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R040159, red pyrope, Meronitz, Bohemia; λexcitation = 532 nm, unoriented

V?

Cr3+

Fig. 6.2 A. Peaks likely related to Cr3+ centers; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060448, red pyrope, India; λexcitation = 532 nm, unoriented

Cr3+

Fig. 6.1 B. (right) Magnification of peaks located at

V?

Intensity

683 nm (4166 cm-1), 686 nm (5324225 cm-1), 689 nm (5324290 cm-1), and 693 nm (5324360 cm-1), related to 3+ 532 Cr centers and a peak located at 696 nm ( 4431 -1 cm ) possibly related to a V center; sample R040159, red pyrope, Meronitz, Bohemia; unoriented

Cr3+

Raman Shift (cm-1)

54

Intensity

Fig. 6.2 B. (left) Magnification of peaks located at 693 (5324366 cm-1) and 694 nm (5324396 cm-1) associated with Cr3+ luminescence centers; sample R060448, red pyrope, India, unoriented

Cr3+

-1 Raman Shift (cm )

Fig. 6.3 A. Peaks in Raman spectrum of colorless pyrope located at 683 nm (5324172 cm-1), 690 nm (5324316 cm-1),

532 -1 532 -1 532 -1 532 -1 695 nm ( 4427 cm ), 693 nm ( 4380 cm ), 699 ( 4492 cm ), and 701 nm ( 4550 cm ), are likely related to Cr, 532 -1 while a single peak centered at 729 nm ( 5095 cm ) may be related to V sample; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R080060, synthetic pyrope, unoriented

Cr3+ V

Cr3+

Fig. 6.3 B. Peaks in Raman spectrum of red pyrope located at at 683 nm (5324171 cm-1), 686 nm (5324225 cm-1),

689 nm (5324289 cm-1), 692 (5324358 cm-1), and 695 nm (5324429 cm-1) related to Cr3+ centers; sample R040159, Sunset Crater, AZ, unoriented

V?

Cr3+

55

Fig. 6.3 C. Comparison of magnified luminescence features associated with Cr and V centers in red pyrope (3-4) and colorless pyrope (1-2); 1) sample R070637, colorless pyrope, Piedmont, Italy, 2) sample R080060, synthetic colorless pyrope, 3) sample R040159, red pyrope, Bohemia, 4) sample R050446, red pyrope, Sunset Crater, AZ; 532 nm laser, unoriented V?

Intensity

Cr3+ 4.

Cr3+ 3.

Cr3+

2.

V

3+

Cr

1.

Raman Shift (cm-1)

Almandine Fig. 6.4 A. Peaks in Raman spectrum of almandine associated with Cr3+ and V luminescence centers; sample R060099, red almandine, Alaska, λexcitation = 532 nm, unoriented

Intensity

Fig. 6.4 B. (right) Magnification of peaks associated with Cr3+ located at 688 nm (5324277 cm-1), 693 nm 532 -1 532 -1 ( 4372 cm ), and 694 nm ( 4401 cm ) and a peak located at 706 nm (5324642 cm-1) related to a V center; sample R060099, Alaska, unoriented Cr3+

V?

Raman Shift (cm-1)

56

Spessartine

Fig. 6.5 A. Peaks in the Raman spectrum of spessartine associated with Cr3+ and possibly V luminescence centers; sample R0600279, spessartine, Pakistan, λexcitation = 532 nm, unoriented

Cr3+

V?

Fig. 6.5 B. Magnification of a peak associated with Cr3+ located at 693 nm (5324370 cm-1) and a peak located at 710 532

4705 cm-1) possibly related to V in spessartine; sample R060279, Pakistan, unoriented

Intensity

nm (

Cr3+ V?

Raman Shift (cm-1)

Fig. 6.6 A. Peaks in the Raman spectrum of spessartine associated with Cr3+ luminescence centers; sample R080053, East Africa, λexcitation = 532 nm, unoriented

Cr3+

57

Fig. 6.6 B. Magnification of peaks located at 693

nm (5324370 cm-1) and 694 nm (5324395 cm-1) 3+ associated with Cr centers in spessartine; sample R080053, East Africa, unoriented Intensity

Cr3+

Raman Shift (cm-1)

Grossular Fig. 6.7 A. Peaks associated with Cr3+, Mn, and possibly V centers in the Raman spectrum of grossular; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060382, Tanzania, λexcitation = 532 nm, unoriented

V? Cr3+3+ Cr Mn

Fig. 6.7 B. (right) Magnification of peaks located at 697 nm (5324450 cm-1) and 701 nm (5324544 cm-1) related to Cr3+ 532 -1 centers and peaks located at 717 nm ( 4840 cm ), 721 nm (5324930 cm-1) possibly related to V centers; sample R060382, grossular from Tanzania, unoriented Intensity

V?

Cr3+

Raman Shift (cm-1)

58

Fig. 6.7 C. (left) Magnification of a broad feature located at 605 nm (5322262 cm-1) likely related to Mn; sample R060382, grossular from Tanzania, unoriented

Intensity

Mn

-1 Raman Shift (cm )

Uvarovite Fig. 6.9 A. Peaks in the Raman spectrum of uvarovite associated with Cr3+ and possibly V centers as well as centers associated with unidentified REE; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060477, Ural Mts., Russia, λexcitation = 532 nm, unoriented

V? Cr3+ ?

Fig. 6.9 B. Magnification of peaks located at 696 nm (5324440 cm-1) and 701 nm (5324542 cm-1) attributed to Cr3+

centers and a peak possibly related to V located at 721 nm (5324930 cm-1) in the Raman spectrum of uvarovite; sample R060477, Ural Mts., Russia, unoriented

?

Intensity

V Cr3+

-1 Raman Shift (cm )

59

Spectral Features Associated with Rare-earth Elements Incorporation of REE’s into the structure of garnet is of great importance to geoscientists because of its applications to geochemical and thermodynamic modeling. Trace-element partitioning and diffusion between garnets and silicate melts are vital to understanding the dynamics of the mantle (Quatieri et al., 2002). Very few studies have been conducted to determine the site preferences and coordination of trace and rare-earth elements in garnets (Quatieri et al., 2002). Although garnets incorporate relatively low concentrations of rare-earth elements into their structures compared to other silicate minerals such as titanite and zircon, isotopic studies (Nd-Sm, U-Pb, Rb-Sr, Lu-Hf) of metamorphic garnets still provide valuable data necessary to constrain temperature and pressure conditions as well as dates of metamorphic and igneous events (Prince et al., 2000). There are several luminescence features present in the Raman spectra of garnet that are likely related to trace amounts of REE. Although many of these features have not been described in the literature, comparison of spectral features in the various garnet species with luminescence features of other REEbearing minerals (like titanite and zircon) studied by Gaft et al. (2005), suggests that several of these 3+ peaks are likely related to Nd . According to luminescence studies of titanite conducted by Gaft et al. 3+ (2003), nearly all Nd luminescence emission bands appear in the IR region. Nd3+ has been assigned to features located at 860, 870, 878, 880, 888, 906/907, 930, 940/942, 1047, 1060, 1070/1071, 1080, 1089/1090, 1100, 1115, and 1131 nm in titanite (Gaft et al., 2003). Additional luminescence studies done on zircon also attribute bands at 817 and 885 nm to Nd3+ (Gaft et al., 2005). Denisov et al. (1986) provide images of spectral features associated with Nd3+ in their study of gallium garnet crystals synthesized for use in lasers. Although no table of assigned centers was provided, the band locations appear to be consistent with the observations made by Gaft et al. (2003, 2005) for titanite and zircon (Denisov et al., 1986). Based on the location of Nd3+ spectral features provided by the aforementioned authors, the following table and corresponding figures address the various Nd3+ peaks present in the Raman spectra of each garnet variety as well as describe additional peaks likely related to unidentified REE luminescence centers (Table 6.3). Laser ablation ICP-MS data for sample R060382, grossular from Tanzania, is provided in Table 6.2 (Breeding, 2007). Table 6.2 Laser ablation ICP-MS data for grossular, sample R060382, Lalatema, Tanzania; numerical values represent concentrations in parts per million (ppm) of each element taken at three different spots on the sample (Breeding, 2007).

7Li

24Mg 2966 2974 2872

29Si 18.5 18.2 17.7

68.2 62.5 67.6

43Ca 28.4 27.9 27.5

44Ca 27.4 26.7 26.6

45Sc 16.0 15.4 16.4

48Ti 2612.0 2650.0 2672.0

51V

3.7 3.4 3.7

55Mn 2802.0 2721.0 2620.0

56Fe 1110.0 992.7 949.2

57Fe 1306.0 1159.0 1121.0

69Ga 24.5 23.2 22.5

72Ge

88Sr 2.6 2.5 2.4

1.2 1.2 1.2

89Y 95.6 93.7 99.7

90Zr

13.2 24.5 35.5

141Pr

146Nd 5.4 5.4 5.6

147Sm 5.5 5.1 5.4

153Eu 2.9 2.8 2.8

157Gd 8.0 7.9 7.7

159Tb 1.9 1.9 1.9

163Dy 14.6 14.8 15.4

165Ho

166Er 10.5 10.4 11.2

169Tm 1.6 1.6 1.6

172Yb 13.5 13.4 14.2

175Lu 1.7 1.8 1.9

178Hf

52Cr

140Ce 0.9 0.9 0.9

0.4 0.4 0.4

31P

212.6 230.6 231.7

24.7 25.4 26.3

3.4 3.4 3.4

1.1 1.0 1.3

60

Grossular Fig. 6.10 A. Peaks associated with Nd3+ luminescence centers in the Raman spectrum of grossular; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060382, Lalatema, Tanzania, λexcitation = 785 nm, unoriented

Nd3+

? ?

Nd3+ 3+ Nd Nd3+ Nd3+

Fig. 6.10 B. Magnification of peaks centered at 860 nm (7851123 cm-1), 875 nm (7851314 cm-1), 879 nm (7851367

cm-1), 888 nm (7851490 cm-1), 933 nm (7852023 cm-1) attributed to Nd3+ centers and peaks located at 895 nm (7851568 -1 785 -1 785 -1 cm ), 911 nm ( 1763 cm ), and 950 nm ( 2216 cm ) likely related to luminescence centers of unidentified REE; sample R060382, Lalatema, Tanzania, unoriented

Intensity

Nd3+

Nd3+

?

?

?

Nd3+ Nd3+

Raman Shift (cm-1)

Fig. 6.10 C. Magnification of peaks centered at 1041 nm (7853135 cm-1), 1061 nm (7853315 cm-1), 1070 nm (7853396

Intensity

cm-1), and 1082 nm (7853502 cm-1) attributed to Nd3+ centers; sample R060382, Lalatema, Tanzania, unoriented

Nd3+ Nd3+ Nd3+ Raman Shift (cm-1)

61

Uvarovite Fig. 6.11 A. Peaks associated with Nd3+ and other unidentified REE luminescence centers in the Raman spectrum of uvarovite; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060477, Ural Mts., Russia, λexcitation = 780 nm, unoriented

?

Nd3+

Nd3+ Nd3+ Nd3+

Fig. 6.11 B. Magnification of peaks centered at 861 nm (7801206 cm-1), 874 nm (7801387 cm-1), 878 nm (7801441 cm-1), 888 nm (7801559 cm-1), and 932 nm (7802091 cm-1) related to Nd3+ and peaks located at 894 nm (7801632 cm-1), 780 -1 780 -1 910 nm ( 1830 cm ), and 948 nm ( 2278 cm ) likely related to luminescence centers of unidentified REE in the Raman spectrum of uvarovite; sample R060477, Ural Mts., Russia; unoriented

Intensity

? Nd

3+

? Nd

3+

? Nd

Nd3+

3+

-1 Raman Shift (cm )

Pyrope

Fig. 6.12 A. Peaks associated with Nd3+ and other unidentified REE luminescence centers in the Raman spectrum of pyrope; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R080060 λexcitation = 780 nm, unoriented

Nd3+ ? ? ?

Nd3+ Nd3+

62

Fig. 6.12 B. Magnification of peaks centered at 863 nm (7801239 cm-1), 871 nm (7801348 cm-1), 877 nm (7801426

cm-1), 879 nm (7801446 cm-1), 885 nm (7801526 cm-1), and 942 nm (7802207 cm-1) related to Nd3+ centers and peaks 780 -1 780 -1 780 -1 located at 894 nm ( 1633 cm ), 903 nm ( 1750 cm ), and 951 nm ( 2312 cm ) related to luminescence centers of unidentified REE in the Raman spectrum of pyrope; sample R080060, unoriented

Intensity

Nd3+ ? ? ? Nd3+

Raman Shift (cm-1)

Spessartine Fig. 6.13 A. Peaks associated with Nd3+ and other unidentified REE luminescence centers in the Raman spectrum of spessartine; sample R060279, red-brown spessartine from Pakistan, λexcitation = 785 nm, unoriented

Nd3+

Fig. 6.13 B. Magnification of peaks centered at 877 nm (7851335 cm-1), 886 nm (7851457 cm-1), 930 nm (7851991

cm-1), and 942 (7852126 cm-1) related to Nd3+ centers and peaks located at 904 nm (7851675 cm-1) and 912 nm 785 -1 ( 1777 cm ) likely related to luminescence centers of unidentified REE in the Raman spectrum of spessartine; sample R060279, red-brown spessartine from Pakistan, unoriented

Intensity

Nd3+

Nd3+

?

Raman Shift (cm-1)

63

Feature Related to Unidentified REE Luminescence Centers: Fig. 6.14 A. Peaks likely related to unidentified REE luminescence centers in the Raman spectrum of grossular; sample R060443, unknown locality, λexcitation = 532 nm, unoriented

Fig. 6.14 B. Magnification of peaks centered at 790 nm (5326142 cm-1), 801 nm (5326321 cm-1), and 807 nm

Intensity

(5326408 cm-1) likely related to luminescence centers of unidentified REE in the Raman spectrum of grossular; sample R060443, unknown locality, unoriented

? ? ?

Raman Shift (cm-1)

Fig. 6.15 A. Peak likely related to an unidentified REE luminescence center in the Raman spectrum of spessartine; R050063, China, λexcitation = 785 nm, unoriented

?

64

Fig. 6.15 B. Magnification of peak centered at

968 nm (7852414 cm-1) likely related to an unidentified REE luminescence center; sample R050063, spessartine from China, unoriented

Intensity

?

Raman Shift (cm-1)

Fig. 6.16 A. Peaks likely related to unidentified REE luminescence centers in the Raman spectrum of uvarovite; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R06104, Finland, λexcitation = 532 nm, unoriented

Fig. 6.16 B. Magnification of peaks centered at 1580 nm (5326411 cm-1), 1598 nm (5326484 cm-1), and 1621 nm

Intensity

(5326571 cm-1) likely related to unidentified REE luminescence centers; sample R061041, uvarovite from Finland, unoriented

?

? ? Raman Shift (cm-1)

65

Table 6.3 Table of spectral features in garnet related to the presence of Nd3+ and possibly other unidentified REE Sample No.

Description/Origin

R060382

transparent tan, Lalatema Tanzania

Garnet Specie

Spectral Feature nm

R060443

R060477

R061041

R080060

yellowish tan, unknown locality

green, Ural Mts., Russia

green, Outokumpu, Finland

colorless, synthetic, P = 23.5 kb, T = 1000 deg C for 23 hours

Grossular

Grossular

Uvarovite

Uvarovite

Pyrope

Ion

Figure

Reference

1123

Nd3+

6.10 A & B

Gaft et al., 2003 (titanite)

1314 1367 785 1490 785 1568 785 1763 785 2023 785 2216 785 3135 785 3315 785 3396 785 3502

Nd? Nd3+ Nd3+ ? ? Nd3+ ? Nd? Nd3+ Nd3+ Nd3+

6.10 A & B 6.10 A & B 6.10 A & B 6.10 A & B 6.10 A & B 6.10 A & B 6.10 A & B 6.10 A & C 6.10 A & C 6.10 A & C 6.10 A & C

?

6.14

-1

cm

860

785

875 879 888 895 911 933 950 1041 1061 1070 1082

785

790

532

801 807 818

532

6321 6408 532 6584

? ? Nd?

6.14 6.14 6.14

Gaft et al., 2005 (zircon)

861

780

1206

Nd3+

6.11

Gaft et al., 2003 (titanite)

874

780

1387

Nd?

6.11

878

780

1441

3+

6.11

Gaft et al., 2003 (titanite)

888

780

1559

3+

Nd

6.11

Gaft et al., 2003 (titanite)

894

780

1632

?

6.11

910

780

1830

?

6.11

932

780

932

780

2091

948

780

2278

1580

532

1598

785

6142

532

Nd

Gaft et al., 2003 (titanite) Gaft et al., 2003 (titanite)

Gaft et al., 2003 (titanite)

Gaft et al., 2003 (titanite) Gaft et al., 2003 (titanite) Gaft et al., 2003 (titanite)

3+

6.11

Gaft et al., 2003 (titanite)

3+

Nd

6.11

Gaft et al., 2003 (titanite)

?

6.11

6411

?

6.16

532

6484

?

6.16

1621

532

6571

?

6.16

863

780

1239

Nd3+

6.12

Gaft et al., 2003 (titanite)

871

780

1348

Nd3+

6.12

Gaft et al., 2003 (titanite)

877

780

1426

3+

Nd

6.12

Gaft et al., 2003 (titanite)

879

780

1446

Nd3+

6.12

Gaft et al., 2003 (titanite)

885

780

1526

3+

Nd

6.12

Gaft et al., 2005 (zircon)

894

780

1644

?

6.12

903

780

1750

?

6.12

942

780

2207

Nd

6.12

951

780

2312

?

6.12

2091

Nd

3+

Gaft et al., 2003 (titanite)

66

Sample No.

Description/Origin

Garnet Specie

Spectral Feature nm

R060279

R050063

red-brown, Pakistan

red-brown, China

Spessartine

Spessartine

Ion

Figure

Reference

cm-1

877

785

1335

Nd3+

6.13

Gaft et al., 2003 (titanite)

886

785

1457

Nd3+

6.13

Gaft et al., 2005 (zircon)

904

785

1675

?

6.13

912

785

1777

?

6.13

930

785

942

785

2126

968

785

2414

1991

3+

6.13

Gaft et al., 2003 (titanite)

3+

Nd

6.13

Gaft et al., 2003 (titanite)

?

6.15

Nd

67

Spectral Features Related to OHThe study of structurally incorporated water in garnets is of particular interest to geologists studying the properties of the mantle. These nominally anhydrous minerals represent storage sites for hydrogen (Andrut et al., 2002). The presence of hydrogen can strongly affect the physical properties of minerals, and therefore, investigation of hydrogen substitution in minerals, such as garnet, may provide vital seismic and thermodynamic insight into the mechanics of the mantle (Andrut et al., 2002). Garnets can contain anywhere from 5 wt%), such as members of the grossular-hydrogrossular series, IR modes associated with hydroxyl are strong, broad, overlapping peaks centered at approximately 3660 and 3600 cm-1 (Beran and Libowitsky, 2003; Rossman and Aines, 1991). Substitution of an (OH)4 for a SiO4 group, known as the hydrogrossular substitution, produces these modes (Beran and Libowitsky, 2003). Lager et al. (1989) describes a slightly different set of IR peak locations due to hydrogrossular substitution than the aforementioned authors at 3598 and 3677 cm-1 (Lager et al., 1989). The hydrogrossular substitution has been observed not only in members of the grossular-hydrogrossular series, but also in Ti-rich andradites (Beran and Libowitsky, 2003). The details of hydrogen substitution in garnets with low OH- concentrations remain ambiguous (Beran and Libowitsky, 2003; Johnson, 2006). In general, garnets with low concentrations of OH- display multiple, fine-structured IR bands which may be affected by cation substitutions in proximity of the OH- (Johnson, 2006). IR studies of synthesized pyrope conducted by Withers et al. (1998) reveal an IR mode centered at 3630 cm-1, likely representing hydrogrossular substitution. In the same study, bands located at 3622 (shoulder at 3612), 3561, 3602, 3656, and 3665 cm-1 were attributed to hydroxyl incorporation in synthesized grossulars (Withers et al., 1998). Rossman et al. (1989) conducted IR experiments on pyrope from Dora Maira Massif in the Western Alps and reported peaks located at 3661, 3651, 3641 and 3602 cm-1 in the spectra, modes similar to those found in low-OH- grossular. Multiple spectroscopic studies of uvarovite reveal fourteen hydroxyl modes in the IR spectra (Andrut et al., 2002). Modes located at 3559/3540, 3572/3565, 3595/3588, and 3618 cm-1 are likely associated with hydrogrossular substitution, while modes located at 3652/3602 and 3640 cm-1 may represent substitution of OH- into cation vacancies resulting in SiO3(OH) tetrahedral groups (Andrut et al., 2002). Vibrational modes associated with hydroxyl were observed in the Raman spectra of grossular (Fig. 6.17), spessartine (Fig. 6.18), and andradite (Fig. 6.19 & 6.20). A doublet in the spectrum of spessartine with peaks at 3583 and 3632 cm-1 (Fig. 6.18) and a single mode located at 3578 cm-1 in andradite (Fig. 6.19) may represent hydrogrossular substitution (Lager et al., 1989; and Andrut et al., 2002). The more complex features in grossular (3534, 3570, 3610, 3645, and 3686 cm-1, Fig. 6.17) and andradite (3538, 3564, 3596, 3611, and 3634 cm-1, Fig. 6.20) are more difficult to interpret. Due to the complexity of the features it is likely that the sample contains low concentrations of OH- (Johnson, 2006). Arrendondo and Rossman (2002) conducted IR and Raman studies on two suites of garnets containing OH to determine whether or not water content could be calculated using Raman spectra. They concluded that Raman spectra are not well suited for the quantitative determination of water in garnet (Arrendondo and Rossman, 2002). Studies conducted by Thomas et al. (2008) contradict this conclusion, claiming that OH content can be quantitatively analyzed in garnets by confocal Raman spectroscopy and comparison with glass standards of known chemical compositions, also known as the “Comparator Technique”. Di Muro et al. (2006) confirm the accuracy and reliability of Raman spectroscopy to quantify water content in glass chips.

68

Fig. 6.17 A. Peaks in the Raman spectrum of grossular associated with OH-; sample R050312, Eden Mills, Vermont, 532 nm laser, unoriented

OH-

Intensity

Intensity

Fig. 6.17 B. (below) Magnification of peaks centered at 3534, -1 3570, 3610, 3645, and 3686 cm attributed to OH in grossular; sample R050312, Eden Mills, Vermont, 532 nm laser, unoriented

OH-

OHRaman Shift (cm-1)

Raman Shift (cm-1)

Fig. 6.18 B. (above) Magnification of doublet centered at 3583 and 3632 cm-1 associated with OH in sppessartine, sample R050063, Fujian Province, China, 532 nm laser, unoriented

Fig. 6.18 A. Peaks in the Raman spectrum of spessartine associated with OH-; sample R050063, Fujian Province, China, 532 nm laser, unoriented

OH-

69

Fig. 16.19 A. Peaks in the Raman spectrum of andradite associated with OH-; sample R060350, San Benito Cty., CA, 532 nm laser, unoriented

OH-

Fig. 6.19 B. (right) Magnification of peak centered at 3578 cm-1

Intensity

attributed to OH- in andradite, sample R060350, San Benito Cty., CA, 532 nm laser, unoriented

OHRaman Shift (cm-1)

Fig. 16.20 A. Peaks in the Raman spectrum of andradite associated with OH-; sample R040001, Stanley Butte, Graham Cty., AZ, 532 nm laser, unoriented

OH-

70

Fig. 6.20 B. Magnification of peaks centered at 3538, 3564, 3596, 3611, and 3634 cm-1 attributed to OH- in

Intensity

andradite, sample R040001, Stanley Butte, Graham Cty., AZ, 532 nm laser, unoriented

OH-

Raman Shift (cm-1)

71

References Andrut, M., Wildner, M. & Beran, A. (2002) The crystal chemistry of birefringent natural uvarovites. Part IV. OH defect incorporation mechanisms in non-cubic garnets derived from polarized IR spectroscopy. European Journal of Mineralogy, Vol. 14, No. 6, pp. 1019. Andrut, M. & Wildner, M. (2001) The crystal chemistry of birefringent natural uvarovites: Part I. Optical investigations and UV-VIS-IR absorption spectroscopy. American Mineralogist, Vol. 86, No. 10, pp. 1219-1230. Arredondo, E.H. & Rossman, G.R. (2002) Feasibility of determining the quantitative OH content of garnets with Raman spectroscopy. American Mineralogist, Vol. 87, No. 2-3, pp. 307-311. Bell, D.R. & Rossman, G.R. (1992) Water in Earth's mantle: The role of nominally anhydrous minerals. Science, Vol. 255, No. 5050, pp. 1391-1397. Beran, A. & Libowitzky, E. (2003) IR spectroscopic characterization of OH defects in mineral phases. Phase transitions, Vol. 76, No. 1-2, pp. 1-15. Breeding, M. (2007) Personal communication. Carstens, H. (1973) The red-green change in chromium-bearing garnets. Contributions to Mineralogy and Petrology, Vol. 41, No. 3, pp. 273-276. Denisov, A.L., Ostroumov, V.G., Saidov, Z.S., Smirnov, V.A. & Shcherbakov, I.A. (1986) Spectral and luminescence 3+ 3+ properties of Cr and Nd ions in gallium garnet crystals. Journal of the Optical Society of America B, Vol. 3, No. 1, pp. 95-101. Di Muro, A., Villemant, B., Montagnac, G., Scaillet, B. & Reynard, B. (2006) Quantification of water content and speciation in natural silicic glasses (phonolite, dacite, rhyolite) by confocal microRaman spectrometry. Geochimica et Cosmochimica Acta, Vol. 70, No. 11, pp. 2868-2884. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Gaft, M., Nagli, L., Reisfeld, R. & Panczer, G. (2003) Laser-induced time-resolved luminescence of natural titanite CaTiOSiO4 . Optical Materials, Vol. 24, No. 1-2, pp. 231-241. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Gubelin, E. & Schmetzer, K. (1982) Gemstones with alexandrite effect. Gems & Gemology, Vol. 18, pp. 197-203. Henderson, R.R. (2009) Determining chemical composition of the silicate garnets using Raman spectroscopy. Prepublication Manuscript, University of Arizona. Hofmeister A.M. & Chopelas A., (1991) Vibrational spectroscopy of end member silicate garnets, Physics and Chemistry of Minerals 17, 503-526. Johnson, E.A. (2006) Water in nominally anhydrous crustal minerals: speciation, concentration, and geologic significance. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 117-154. Lager, G.A., Armbruster, T., Rotella, F. & Rossman, G.R. (1989) OH substitution in garnets: X-ray and neutron diffraction, infrared, and geometric-modeling studies. American Mineralogist, Vol. 74, pp. 840-851. Manning, P.G. (1969) Optical absorption studies of grossularite, andradite (var. colophonite) and uvarovite. Canadian Mineralogist, Vol. 9, No. 5, pp. 723. Manning, P.G. (1967) The optical absorption spectra of the garnets almandine-pyrope, pyrope, and spessartine and some structural interpretations of mineralogical significance. Canadian Mineralogist, Vol. 9, No. 2, pp. 237. Manning, P.G. & Harris, D.C. (1970) Optical-absorption and electron-microprobe studies of some high-Ti andradites. Canadian Mineralogist, Vol. 10, No. 2, pp. 260. Merli, M., Callegari, A., Cannillo, E., Caucia, F., Leona, M., Oberti, R. & Ungaretti, L. (1995) Crystal-chemical complexity in natural garnets; structural constraints on chemical variability. European Journal of Mineralogy, Vol. 7, No. 6, pp. 1239. Novak, G.A. & Gibbs, G.V. (1971) The crystal chemistry of the silicate garnets. American Mineralogist, Vol. 56, pp. 791-825. Prince, C.I., Kosler, J., Vance, D. & Günther, D. (2000) Comparison of laser ablation ICP-MS and isotope dilution REE analyses—implications for Sm–Nd garnet geochronology. Chemical Geology, Vol. 168, No. 3-4, pp. 255274. Quartieri, S., Boscherini, F., Chaboy, J., Dalconi, M.C., Oberti, R. & Zanetti, A. (2002) Characterization of trace Nd and Ce site preference and coordination in natural melanites: a combined X-ray diffraction and high-energy XAFS study. Physics and Chemistry of Minerals, Vol. 29, No. 7, pp. 495-502. Rossman, G.R. & Aines, R.D. (1991) The hydrous components in garnets; grossular-hydrogrossular. American Mineralogist, Vol. 76, No. 7-8, pp. 1153-1164. Rossman, G.R., Beran, A. & Langer, K. (1989) The hydrous component of pyrope from the Dora Maira Massif, Western Alps. European Journal of Mineralogy, Vol. 1, No. 1, pp. 151. Schmetzer, K. & Bernhardt, H.J. (1999) Garnets from Madagascar with a color change of blue-green to purple. Gems & Gemology, Vol. 35, No. 4, pp. 196–201. 3+ Struve, B. & Huber, G. (1985) The effect of the crystal field strength on the optical spectra of Cr in gallium garnet laser crystals. Applied Physics B: Lasers and Optics, Vol. 36, No. 4, pp. 195-201.

72

Thomas, S.M., Thomas, R., Davidson, P., Reichart, P., Koch-Muller, M. & Dollinger, G. (2008) Application of Raman spectroscopy to quantify trace water concentrations in glasses and garnets. American Mineralogist, Vol. 93, No. 10, pp. 1550. Winchell, A.N. (1933) Optical Mineralogy II, Wiley and Sons, New York. Withers, A.C., Wood, B.J. & Carroll, M.R. (1998) The OH content of pyrope at high pressure. Chemical Geology, Vol. 147, No. 1-2, pp. 161-171.

Additional Information Aines, R.D. & Rossman, G.R. (1984) The hydrous component in garnets; pyralspites. American Mineralogist, Vol. 69, No. 11-12, pp. 1116-1126. Beran, A., Langer, K. & Andrut, M. (1993) Single crystal infrared spectra in the range of OH fundamentals of paragenetic garnet, omphacite and kyanite in an eklogitic mantle xenolith. Mineralogy and Petrology, Vol. 48, No. 2, pp. 257-268. Beran, A. & Libowitzky, E. (2006) Water in natural mantle minerals II: olivine, garnet and accessory minerals. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 169-191. Gillet, P., Fiquet, G., Malezieux, J.M. & Geiger, C.A. (1992) High-pressure and high-temperature Raman spectroscopy of end-member garnets; pyrope, grossular and andradite. European Journal of Mineralogy, Vol. 4, No. 4, pp. 651. Kolesov, B.A. & Geiger, C.A. (1998) Raman spectra of silicate garnets. Physics and Chemistry of Minerals, Vol. 25, No. 2, pp. 142-151. Manning, P.G. (1972) Optical absorption spectra of Fe3+ in octahedral and tetrahedral sites in natural garnets. Canadian Mineralogist, Vol. 11, No. 4, pp. 826. 3+ 3+ Mazurak, Z. & Czaja, M. (1995) Optical properties of tsavorite Ca3Al2 (SiO4)3: Cr , V from Kenya. Journal of Luminescence, Vol. 65, No. 6, pp. 335-340. O'Donnell, K.P., Marshall, A., Yamaga, M., Henderson, B. & Cockayne, B. (1989) Vibronic structure in the photoluminescence spectrum of Cr3+ ions in garnets. Journal of Luminescence, Vol. 42, No. 6, pp. 365-373. Rossman, G.R. (1996) Studies of OH in nominally anhydrous minerals. Physics and Chemistry of Minerals, Vol. 23, No. 4, pp. 299-304. van Westrenen, W., Blundy, J. & Wood, B. (1999) Crystal-chemical controls on trace element partitioning between garnet and anhydrous silicate melt. American Mineralogist, Vol. 84, No. 5-6, pp. 838-847.

73

Olivine Gem names: peridot Ideal chemistry: (Mg,Fe2+)2SiO4 Crystal system: orthorhombic Point Group:

H-M: mmm

S: D2h

Space Group: Pbnm

Burmese peridot; Photo courtesy of Stone Group Labs

Table of Atomic Coordinates - forsterite (Kirfel et al., 2005): atom

x

y

z

Mg1

0

0

0

Mg2

0.50846

0.77742

0.25

Si

0.07353

0.59403

0.25

O1

0.73408

0.59155

0.25

O2

0.22160

0.44704

0.25

O3

0.22253

0.66316

0.46697

Raman mode analysis:

Raman Active Modes Atom

Wyckoff Position Point Symmetry Ag B1g B2g B3g

O3

8d

1

3

3

3

3

Mg2, Si, O1, O2

4c

m

2

1

2

1

Mg1

4a

1

-

-

-

-

(1 x 8d) + (4 x 4c) + (1 x 4a) Raman mode analysis predicts the existence of 36 active Raman modes in forsterite: 11Ag + 7B1g + 11B2g +7B3g = 36

74

Introduction to Raman Spectrum of Forsterite Forsterite is an isosilicate and consists of hexagonal closest packed SiO4 groups and octahedrally coordinated Fe2+ and Mg (Fig. 7.2). It is the most important mineral in the upper 400 km of the Earth because it reflects the Fe component in the mantle. Forsterite crystals containing 8-10% of Fe are a desirable yellow-green color, known to gemologists as peridot. Pure, end-member forsterite is colorless. Innumerable studies have been conducted involving the structural and chemical properties of olivine due to its geologic importance. The Raman spectrum of forsterite is provided below. The two most intense peaks, centered at 824 and 856 cm-1, are associated with SiO4 stretching modes (Chopelas, 1991). Fig. 7.1 A. The Raman spectrum of forsterite with mode assignments reported by Chopelas (1991); sample R040052, synthetic forsterite; 514.5 nm laser, processed

856 Ag

Intensity

824 Ag

920 B3g

See Fig. 7.1 B

226 Ag

374 B3g

965 Ag

592 545 B3g 608 Ag Ag Raman Shift (cm-1)

Fig. 7.1 B. Magnification of Raman peaks in 200-440 cm-1 range; mode assignments reported by Chopelas (1991);

Intensity

sample R040052, synthetic forsterite; 514.5 nm laser, processed

226 Ag

329 Ag

339 Ag

374 B3g

304 Ag

315 B3g

410 B3g

422 Ag

Raman Shift (cm-1)

75

7.2 The crystal structure of forsterite; green octahedra: MgO6 groups, blue tetrahedra: SiO4 groups A. Viewed down the c-axis

B. Viewed down the b-axis

C. Viewed down the a-axis

76

Spectral Features Related to Chromophores The yellowish-green gem variety of forsterite is well-known as peridot. Peridot contains octahedrally coordinated Fe2+, with occasional trace amounts of Cr3+ producing its characteristic color (Fritsch and Rossman, 1987; Fritsch and Rossman, 1988). Forsterite and chrysoberyl are isostructural and therefore, it is likely that the weak peaks centered at 693 nm (5324373 cm-1) and 694 nm (5324400 cm-1) in the Raman spectra of forsterite (Fig. 7.3 A. & B.) are associated with octahedrally coordinated Cr3+ luminescence centers also present in chrysoberyl (Gaft et al., 2005). Fig. 7.3 A. Peaks centered at 693 and 694 nm that are likely associated with Cr3+ luminescence centers; sample R060539, brown forsterite gem from Sri Lanka, λexcitation = 532 nm, unoriented

Cr3+

Fig. 7.3 B. Magnification of peaks associated with Cr3+ luminescence centers; sample R060539, brown forsterite gem from Sri Lanka, λexcitation = 532 nm, unoriented

Intensity

Cr3+

Raman Shift (cm-1)

77

References Chopelas, A. (1991) Single crystal Raman spectra of forsterite, fayalite, and monticellite. American Mineralogist, Vol. 76, No. 7-8, pp. 1101-1109. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Fritsch, E. & Rossman, G.R. (1987) An Update on Color in Gems. Part 1. Introduction and Colors caused by Dispersed Metal Ions. Gems and Gemology, Vol. 23, No. 3, pp. 126–139. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Kirfel, A., Lippmann, T., Blaha, P., Schwarz, K., Cox, D.F., Rosso, K.M. & Gibbs, G.V. (2005) Electron density distribution and bond critical point properties for forsterite, Mg2SiO4, determined with synchrotron single crystal X-ray diffraction data. Physics and Chemistry of Minerals, Vol. 32, No. 4, pp. 301-313. Kolesov, B.A. & Geiger, C.A. (2004) A Raman spectroscopic study of Fe–Mg olivines. Physics and Chemistry of Minerals, Vol. 31, No. 3, pp. 142-154.

Additional Information Kolesov, B.A. & Tanskaya, J.V. (1996) Raman spectra and cation distribution in the lattice of olivines. Materials Research Bulletin, Vol. 31, No. 8, pp. 1035-1044.

78

Quartz Gem names: amethyst, citrine, ametrine, praseolite, smoky, rose, milky, rutilated, tourmalinated Ideal chemistry: SiO2 Crystal system: trigonal Point Group:

H-M: 32

S: D3 Synthetic citrine gem

Space Group: P3221 Table of Atomic Coordinates (Ikuta et al., 2007) atom

x

y

z

Si

0.4696

0

0

O

0.4132

0.2679

0.1191

Raman mode analysis:

Raman Active Modes Atom

Wyckoff Position

Point A1 E Symmetry

O

6c

1

3

6

Si

3a

2

1

3

Raman mode analysis predicts the existence of 13 active Raman modes in quartz at room conditions: 4A1 + 9E = 13

79

Introduction to the Raman Spectrum of Quartz Quartz is one of the most common minerals on the Earth’s surface, second only to feldspar. It is a framework silicate consisting entirely of SiO4 groups (Fig. 8.2 A. & B.). Quartz is one of the few minerals in this study that does not have an inversion center, therefore, all of the predicted vibrational modes can be represented by peaks in the Raman spectra (Fig. 8.1). The most intense peak, centered at 463 cm-1, is associated with Si-O-Si bending (Sato and McMillan, 1987). The second most intense peak, centered at 205 cm-1, is associated with a soft mode (Jayaraman et al., 1967). This vibrational mode is much more flexible than the mode associated with the 463 cm-1 peak, and therefore, it exhibits both temperature and pressure dependence. Quartz undergoes a transition from the low temperature α phase (P3221) to the high temperature β phase (P3121) at 573oC (Shapiro et al., 1967). With an increase in temperature, the 205 cm-1 peak exhibits a decrease in frequency as quartz approaches the α-β transition (Shapiro et al., 1967). The peak is nonexistent in the Raman spectrum of β-quartz (Shapiro et al., 1967). Jayaraman et al. (1967) report that the 205 cm-1 peak exhibits a large initial increase in frequency and a decrease in peak width with pressure. Fig. 8.1 The Raman spectrum of quartz with mode assignments as reported by Sato and McMillan (1987), a peak -1 related to an E mode centered at 1066 cm is not visible in this spectrum; sample R040031, 514 nm laser, processed, unoriented

Intensity

463 A1

205 A1 128 E

263 E

401 & 354 393 A1 E

697 E

808 E

1083 1160 1231 A1 E E

Raman Shift (cm-1)

Fig. 8.2 The crystal structure of quartz; blue tetrahedra: SiO4 groups A. Viewed down the a-axis

B. Viewed down the c-axis

80

Causes of Color in Quartz The mechanisms responsible for the various colors of quartz are still controversial. Hole centers (missing electrons in the crystal structure) created by the presence of various ions produce a variety of colors in quartz. The distinctive purple color of amethyst is the result of a hole center created by incorporation of Fe2+ and Fe3+ into the crystal structure combined with irradiation (Paradise, 1982). Exposure to heat destabilizes this color center, and in the presence of Fe3+, amethyst changes color becoming yellow citrine. If Fe2+ is present in amethyst, heat will alter the quartz to green-colored prasiolite (Paradise, 1982; Fritsch and Rossman, 1988). A hole center created by substitutional AlO4 groups produces the diagnostic black color of smoky quartz, though the details of this process are still debated (Maschmeyer et al., 1980; Fritsch and Rossman, 1988). The color of rose quartz is attributed to pink fibrous microinclusions, likely a mineral species closely related to dumortierite (Goreva et al., 2001). The pink color of these inclusions is the result of charge transfer between Fe2+ and Ti4+ in the M1 site of the inclusions (Ma et al., 2002). There are no observable peaks in the Raman spectra of quartz associated with the aforementioned color centers in this study.

References Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Goreva, J.S., Ma, C. & Rossman, G.R. (2001) Fibrous nanoinclusions in massive rose quartz: The origin of rose coloration. American Mineralogist, Vol. 86, No. 4, pp. 466-472. Ikuta, D., Kawame, N., Banno, S., Hirajima, T., Ito, K., Rakovan, J.F., Downs, R.T. & Tamada, O. (2007) First in situ X-ray identification of coesite and retrograde quartz on a glass thin section of an ultrahigh-pressure metamorphic rock and their crystal structure details. American Mineralogist, Vol. 92, No. 1, pp. 57-63. Jayaraman, A., Wood, D.L. & Maines, R.G. (1987) High-pressure Raman study of the vibrational modes in AlPO4 and SiO2(α-quartz). Physical Review B, Vol. 35, No. 16, pp. 8316-8321. Ma, C., Goreva, J.S. & Rossman, G.R. (2002) Fibrous nanoinclusions in massive rose quartz: HRTEM and AEM investigations. American Mineralogist, Vol. 87, No. 2-3, pp. 269-276. Maschmeyer, D., Niemann, K., Hake, H., Lehmann, G. & Räuber, A. (1980) Two modified smoky quartz centers in natural citrine. Physics and Chemistry of Minerals, Vol. 6, No. 2, pp. 145-156. Paradise, T.R. (1982) The natural formation and occurrence of green quartz. Gems & Gemology, Vol. 18, No. 1, pp. 39. Sato, R.K. & McMillan, P.F. (1987) An infrared and Raman study of the isotopic species of alpha-quartz. Journal of Physical Chemistry, Vol. 91, No. 13, pp. 3494-3498. Shapiro, S.M., O'Shea, D.C. & Cummins, H.Z. (1967) Raman scattering study of the alpha-beta phase transition in quartz. Physical Review Letters, Vol. 19, No. 7, pp. 361-364.

81

Spinel Ideal chemistry: MgAl2O4 Crystal system: isometric Point Group:

H-M: m3m

S: Oh

Space Group: Fd3m Synthetic spinel

Table of Atomic Coordinates (Martignago et al., 2003): atom

x

y

z

Mg

0.125

0.125

0.125

Al

0.5

0.5

0.5

O

0.26338

0.26338

0.26338

Raman mode analysis:

Raman Active Modes Atom Wyckoff Position Point Symmetry A1g Eg T2g O

32e

3m

1

1

2

Al

16d

3m

-

-

-

Mg

8a

4 3m

-

-

1

Raman mode analysis predicts the existence of 5 active Raman modes in spinel: 1Ag + 1Eg + 3T2g = 5

82

Spectral Features Related to Chromophores and Other Ions Natural Spinel The metal ions that produce the various colors of natural spinel include Cr3+, Fe3+, Fe2+, and Co2+ (Fritsch and Rossman, 1988) (See Appendix D). Octahedrally coordinated Cr3+ produces pink and red hues in spinel (Fritsch and Rossman, 1988). The addition of tetrahedrally coordinated Fe2+ to octahedrally coordinated Cr3+ in the spinel structure creates a purple hue (Fritsch and Rossman, 1988). In addition to these ions, synthetically grown spinels can contain trace amounts of Cu, V, and Mn that may contribute to color (Cain, 1988). Multiple prominent peaks centered at 676 nm, 686 nm, 698 nm, 708 nm, and 718 nm (Fig. 9.1 A. & B.) in the Raman spectra are related to Cr3+ luminescence centers in spinel (Gaft et al., 2005). These features are present in the Raman spectra of spinel samples of several colors, but not in the spectra of black spinel (Fig. 9.2). The peaks in the Raman spectra of black spinel are also poorly resolved when compared to the Raman peaks in the spectra of the other colored samples (Fig. 9.3).

Fig. 9.1 A. Peaks centered at 676 nm (5324000 cm-1), 686 nm (5324211 cm-1), 698 nm (5324466 cm-1), 708 nm

(5324657 cm-1), and 718 nm (5324859 cm-1) in the Raman spectrum of natural spinel are related to the Cr3+ luminescence center; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R050392, pink Burmese spinel, unoriented

Cr3+ Cr3+

Fig. 9.1 B. Magnification of peaks attributed to Cr3+ luminescence centers; sample R050392, natural pink Burmese spinel, λexcitation = 532 nm, unoriented

Cr3+ 3+

Intensity

Cr

Raman Shift (cm-1)

83

Fig. 9.2 Peaks associated with Cr3+ luminescence centers in natural spinel of a variety of colors; note the absence

of Cr3+ peaks in black spinel; 1) sample R060799, blue spinel from Pakistan; 2) sample R050392, pink Burmese spinel; 3) sample R050411, very pale pink Burmese spinel; 4) sample R050259, purple Burmese spinel; 5) sample R070013, black Thai spinel; λexcitation = 532 nm, unoriented

Intensity

5.

4. 3. 2. 1. Raman Shift (cm-1)

Fig. 9.3 Comparison of the Raman spectra of spinel of various colors (both synthetic and natural) with the spectra of natural black spinel; 1) natural pink spinel, Mogok, Burma, sample R050411; 2) natural blue spinel, Pakistan, sample R060799; 3) natural black spinel, Cuba, sample R060798; 4) natural black spinel, Thailand, sample R070013; 5) synthetic blue spinel, sample X080014; 6) synthetic red spinel, sample X080013; λexcitation = 532 nm, processed, unoriented

6.

Intensity

5.

4.

3. 2. 1. Raman Shift (cm-1)

84

Synthetic Spinel The peaks attributed to Cr3+ luminescence centers in natural spinel visibly differ from the peaks present in the spectra of synthetic spinel samples (Fig. 9.4 A. & B.). The peaks in synthetic spinel appear broader and poorly resolved when compared to the spectra of natural spinel. The highest intensity Cr3+ peak in the spectra of synthetic spinel is also shifted to a higher frequency, by 83 cm-1, than the peak with the highest intensity in natural spinel. Based on luminescence studies conducted by Tijero and Ibarra (1993), when chromium-containing spinel are heated and annealed (700-950oC), the spectral features associated with the luminescence centers can change dramatically. The peaks associated with Cr3+ centers in spinel that have been heated and annealed are broader and less resolved. The changes in the spectra are related to the disordering of the Mg and Al in the crystal structure. As the proportion of four and six coordinated Mg and Al changes, there is a decrease in the distance between Cr3+ and O2-. As the spinel are heated, the population of ordered centers associated with Cr3+ decreases, and therefore, the peaks associated associated with the Cr3+ disappear (Tijero and Ibarra, 1993).

Fig. 9.4 A. Peaks related to Cr3+ centers in spinel; 1) synthetic red spinel gem, sample X080013; 2) synthetic blue

Intensity

spinel gem, sample X080014; 3) natural pink Burmese spinel, sample R050392; notice the differences in shape and width of the highest intensity peak; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; λexcitation = 532 nm, unoriented

3.

2. 1.

Raman Shift (cm-1)

Fig. 9.4 B. Magnification of the peaks related to Cr3+ centers in spinel; 1) synthetic red spinel gem, sample X080013; 2) synthetic blue spinel gem, sample X080014; 3) natural pink Burmese spinel, sample R050392; notice the differences in shape and width of the highest intensity peak; λexcitation = 532 nm, unoriented

Intensity

3.

2.

1. Raman Shift (cm-1)

85

Structural Effects on Raman Spectra The ordered structure of spinel, called the normal structure, has the following cation distribution: X[Y2]O4, where X are divalent tetrahedrally coordinated cations (Mg), and Y are octahedrally coordinated trivalent cations (Al) (Uchida et al., 2005). Most natural spinels, and all synthetic spinels, are disordered, meaning Al occupies the tetrahedral site and both Al and Mg occupy the octahedral site (Uchida et al., 2005). This disordering produces structural defects, electron traps, and vacancies, in the crystal structure (Cain et al., 1988). These defects can complicate the interpretation of the spectra of spinel. In nature, spinel can cool slowly enough to create completely ordered crystals. An example are spinel of Burma. There are five predicted Raman-active modes in spinel, however, only four peaks have been observed in the Raman spectra of natural spinel. Based on analysis of synthetic spinel, the missing fifth peak should appear at 492 cm-1. Unlike natural spinel, synthetic spinel frequently produce six to seven Raman peaks including an additional peak centered at 727 cm-1 (Cynn, 1992). In natural, ordered spinel, at ambient conditions, this peak is not visible, however, once a sample is heated and then quenched, the peak appears in the spectrum. The 727 cm-1 peak is attributed to the stretching vibration of AlO4 groups created by the rearrangement of some Al ions from octahedral to tetrahedral sites (Cynn, 1992). This additional peak has not been observed in this study. At this time, no definitive study relating the ordering of Mg and Al to the Raman spectra of spinel has been conducted. References Cain, L.S., Pogatshnik, G.J. & Chen, Y. (1988) Optical transitions in neutron-irradiated MgAl2O4 spinel crystals. Physical Review B, Condensed matter, Vol. 37, No. 5, pp. 2645-2652. Chopelas, A. & Hofmeister, A.M. (1991) Vibrational spectroscopy of aluminate spinels at 1 atm and of MgAl2O4 to over 200 kbar. Physics and Chemistry of Minerals, Vol. 18, No. 5, pp. 279-293. Cynn, H., Sharma, S.K., Cooney, T.F. & Nicol, M. (1992) High-temperature Raman investigation of order-disorder behavior in the MgAl2O4 spinel. Physical Review B, Condensed matter, Vol. 45, No. 1, pp. 500-502. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. 3+ 3+ Martignago, F., Negro, A.D. & Carbonin, S. (2003) How Cr and Fe affect Mg–Al order–disorder transformation at high temperature in natural spinels. Physics and Chemistry of Minerals, Vol. 30, No. 7, pp. 401-408. Tijero, J.M.G. & Ibarra, A. (1993) Use of luminescence of Mn2+ and Cr3+ in probing the disordering process in MgAl2O4 spinels. Journal of Physical Chemistry, Solids, Vol. 54, No. 2, pp. 203-207. Uchida, H., Lavina, B., Downs, R.T. & Chesley, J. (2005) Single-crystal X-ray diffraction of spinels from the San Carlos Volcanic Field, Arizona: Spinel as a geothermometer. American Mineralogist, Vol. 90, No. 11-12, pp. 1900-1908.

86

Spodumene Gem names: kunzite, hiddenite Ideal chemistry: LiAlSi2O6 Crystal system: monoclinic Point Group:

H-M: 2/m

S: C2h

Space Group: C2/c

Kunzite, Photo courtesy of Stone Group Labs

Table of Atomic Coordinates (Cameron et al., 1973): atom

x

y

z

Si

0.2941

0.0935

0.2560

Al1

0

0.9066

0.25

Li2

0

0.2752

0.25

O1

0.1099

0.0823

0.1402

O2

0.3646

0.2673

0.3009

O3

0.3565

0.9871

0.0578

Raman mode analysis:

Raman Active Modes Atom

Wyckoff Position Point Symmetry Ag Bg

Si, O1, O2, O3

8f

1

3

3

Al, Li

4e

2

1

2

(4 x 8f) + (2 x 4e) Raman mode analysis predicts the existence of 30 active Raman modes in spodumene: 14Ag + 16B g = 30

87

Spectral Features Related to Chromophores and Other Ions Spodumene is isostructural with diopside. Spodumene contains two nonequivalent octahedral sites occupied by Al and Li (Souza, 2004). The Li site is distorted. Al and Li are frequently replaced by colorinducing ions such as Mn, Cr, and Fe (Souza, 2004). The pink color of kunzite is associated with trace amounts of manganese, while the diagnostic green color of hiddenite is associated with octahedrally coordinated Cr and V (Fritsch and Rossman, 1988). Luminescence studies of spodumene conducted by Walker et al. (1997), Souza et al. (2003), and Gaft et al. (2005) report the presence of a broad luminescence feature located between 600-630 nm. The interpretations of this feature vary. Walker et al. (1997) and Gaft et al. (2005) theorize that this peak is associated with a Mn2+ center. However, Souza et al. (2003) report that this feature is not related to the presence of Mn. Instead, Souza et al. (2003) theorize that the broad feature is likely related to a hole captured by Al3+ atoms in the crystal structure. During experimentation with doped, artificial spodumene crystals, Souza et al. (2003) observed that the presence of Mn intensifies this hole-center luminescence band, while Fe quenches it. A broad peak located between centered at 620 nm is present in the Raman spectra of every spodumene sample in this study and is likely related to the theorized centers described above (Fig. 10.1). Fig. 10.1 A broad peak located at 620 nm (5322670 cm-1) in the Raman spectrum of spodumene may be related to a

Al3+-hole center or Mn2+ center; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R040050, variety kunzite, unoriented

Intensity

Al3+ hole center 2+ or Mn ?

Raman Shift (cm-1)

References Cameron, M., Sueno, S., Prewitt, C.T. & Papike, J.J. (1973) High-temperature crystal chemistry of acmite, diopside, hedenbergite, jadeite, spodumene, and ureyite. American Mineralogist, Vol. 58, pp. 594-618. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Souza, S.O., Ferraz, G.M. & Watanabe, S. (2004) Effects of Mn and Fe impurities on the TL and EPR properties of artificial spodumene polycrystals under irradiation. Nuclear Inst.and Methods in Physics Research, B, Vol. 218, pp. 259-263. Souza, S.O., Watanabe, S., Lima, A.F. & Lalic, M.V. (2007) Thermoluminescent mechanism in lilac spodumene. ACTA PHYSICA POLONICA SERIES A, Vol. 112, No. 5, pp. 1001. 3+ Walker, G., El Jaer, A., Sherlock, R., Glynn, T.J., Czaja, M. & Mazurak, Z. (1997) Luminescence spectroscopy of Cr 2+ and Mn in spodumene (LiAlSi2O6). Journal of Luminescence, Vol. 72, pp. 278-280.

88

Additional Information Chandrasekhar, B.K. & White, W.B. (1992) Polarized luminescence spectra of kunzite. Physics and Chemistry of Minerals, Vol. 18, No. 7, pp. 433-440. Pommier, C.J.S., Denton, M.B. & Downs, R.T. (2003) Raman spectroscopic study of spodumene (LiAlSi2O6) through the pressure-induced phase change from C2/c to P2/c. Journal of Raman Spectroscopy, Vol. 34, pp. 769-775.

89

Titanite Gem names: sphene (old name) Ideal chemistry: CaTiSiO5 Crystal system: monoclinic Point Group:

H-M: 2/m

S: C2h

Space Group: P21/a or C2/c Table of Atomic Coordinates (Ghose et al., 1991) (P21/a) atom

x

y

z

Ca

0.2421

0.4185

0.2511

Ti

0.5137

0.2540

0.7496

Si

0.7483

0.4327

0.2491

O1

0.7494

0.3219

0.7491

O2

0.9097

0.3161

0.4332

O3

0.0881

0.1849

0.0644

O4

0.3829

0.4609

0.6456

O5

0.6191

0.0399

0.8532

Raman mode analysis:

Raman Active Modes Atom Ca, Ti, Si, O1, O2, O3, O4, O5

Wyckoff Position Point Symmetry Ag Bg 4e

1

3

3

(8 x 4e) Raman mode analysis predicts the existence of 48 active Raman modes in titanite: 24Ag + 24B g = 48

90

Metamictization Due to Radiation Damage The crystal chemical properties of titanite not only make this stone valuable to collectors, but also to geologists. The reaction of titanite crystals to radioactive elements can provide valuable information about the geologic conditions under which they form (see also “Spectral Features Related to REE,” p. 69). Incorporation of U and Th into the crystal structure of titanite results in damage of the crystal, called metamictization, due to radioactive decay (Vance and Netson, 1985). Zircon also experiences metamictization, however, X-ray diffraction studies conducted by Vance and Netson (1985) reveal that titanite is two to three times more sensitive to damage by α-decay. The reactivity of titanite to certain radioactive elements may play a role in the disposal of nuclear fuel waste in the future. Vance and Netson propose that titanite-containing glass ceramics could act as the immobilizing host of nuclear waste. An understanding of the metamictization of titanite by radioactive compounds is not only of mineralogical interest, but also of importance to this proposed waste disposal process (Vance and Netson, 1985). The effects of metamictization in titanite are apparent in various laboratory techniques and include: broadening of X-ray diffraction (XRD) peaks, decrease in XRD peak intensity, loss of anisotropy or orientational dependence of spectral features, and broadening and loss of resolution in IR bands (Zhang et al., 2002). In addition, metamict titanite generally contains more OH than its crystalline counterpart (see below, “Spectral Features Related to OH “) (Hammer et al., 1996; Zhang and Salje, 2003). Although an in-depth Raman spectroscopic study of metamict titanite has yet to be conducted, Raman scattering of partially amorphous titanite from the RRUFF database produce poorly resolved, broader Raman peaks and concomitant widened diffraction peaks, as is the case in metamict zircons (Fig. 11.1) (Nasdala et al., 1995). Note: A structural phase transition occurs in titanite with increase in temperature. Raman spectroscopic studies conducted by Salje et al. show that at temperatures above 860K, the structure of titanite changes from P21/a symmetry to A2/a, resulting in peak differences in the Raman spectra (Salje et al., 1992).

Intensity

.193 4. .079

Intensity

Fig. 11.1 (Right) Portion of XRD pattern with associated band width values and (left) corresponding portion of Raman spectra of titanites showing increase in metamictization from bottom to top; 1) sample R040033, 2) sample R050124, 3) sample R050114, 4) R050039 (this sample has undergone the most damage); 514 nm laser, Laser parallel to -b* (0 -1 0). Fiducial mark perpendicular to laser is parallel to -c [0 0 -1], background corrected

3. 2.

.078

1. .059 -1

Raman Shift (cm )

2 Theta

91

Spectral Features Related to OHWhen exposed to radiation, titanite becomes metamict (slightly amorphous). Radioactive decay produces particles that smash through the crystal structure, breaking bonds and destroying crystallinity. Metamictization alters the titanite structure allowing OH- to replace the oxygen atoms (O1) shared by the bent chains of TiO6 groups in the crystal structure (Fig. 11.2) (Hammer et al., 1996). The O1 site is underbonded making the substitution of O1 for OH- and F common (Frost et al., 2001). Degree of metamictization may determine how much OH- a titanite sample contains; damaged titanites contain more OH- than those without exposure to radiation (Hammer et al., 1996). A peak centered at 3485 cm-1 in the Raman spectra of titanite corresponds to an IR band associated with OH- reported by Isetti and Penco (1968) at the same location (Fig. 11.3 A. & B.). Fig. 11.2 (Right) Crystal structure of titanite; arrows

pointing to pink O1 atoms where OH substitutes; green octahedra: TiO6 groups, blue tetrahedra: SiO4 groups, red spheres: oxygen atoms, teal polyhedra: 7-coordinated Ca atoms

Fig. 11.3 A. Weak peak located at 3485 cm-1 attributed to OH-;

Intensity

titanite from Brazil, sample R040033, 532 nm laser, unoriented

OH -

Raman Shift (cm-1)

Fig. 11.3 B. (right) Magnification of weak OH- peak located

Intensity

at 3485 cm-1; titanite from Brazil, sample R040033, 532 nm laser, unoriented

OH-

Raman Shift (cm-1)

92

Cause of Color in Titanite Gem-quality titanite (sphene) is typically green, yellow-green, yellow, brown and less frequently, pink, in color. Iron is usually responsible for yellow, green, and brown colors in titanite. However, octahedrally coordinated Cr3+ can also produce a green color in titanite (Fritsch and Rossman, 1988). The presence of pink carbonate inclusions gives some crystals a pink color. In addition, Mn-rich titanite is pink due to octahedrally coordinated Mn2+ (Fritsch and Rossman, 1988). No spectral features attributed to the aforementioned chromophores have been observed in this study.

Spectral Features Related to Rare-earth Elements The large, 7-fold Ca site in the crystal structure of titanite (Fig. 11.2) can incorporate many different large atoms such as U, Th, Pb and Mn, as well as a variety of rare Earth elements (REE) including Sm, Eu, and Nd (Frost et al., 2001; Gaft et al., 2005). In pegmatite-derived titanites, REE’s can account for over 4 wt % of a crystal, although a typical titanite crystal usually contains less than .02 wt% of REE (Frost et al., 2001). In addition, titanite can incorporate anywhere from 10 to over 100 ppm’s of U into its structure. The geochemistry of titanite, in combination with its high closure temperature (max. 700oC), make it an excellent geochronometer providing important U-Pb isotopic data used to date geologic events (Frost et al., 2001). There are multiple features in the Raman spectra of titanite that are associated with a variety of REE luminescence centers, particularly Sm3+, Eu3+, and Nd3+. Gaft et al. (2005) attribute peaks centered at 574, 578, 589, 600, 613, 617, and 620 nm to Sm3+ centers (Fig. 11.4 A. & B.). Peaks centered at 689 and 703 nm are attributed to Eu3+ luminescence features with corresponding features located at 4263, 4545, and 4651 cm-1 in the Raman spectra (Fig. 11.4 A. & 11.4 C.) (Gaft et al., 2005). Luminescence features located at 867, 883, and 894 nm are attributed to Nd3+ centers (Fig. 11.5 A. & B.) (Gaft et al., 2005; Gaft et al., 2003). There are multiple peaks present in the Raman spectra of titanite that have not been presented in the literature, but are likely related to the luminescence centers of unidentified REE. The locations of the peaks are as follows: 639 nm, 647 nm, 658 nm, 664 nm, 718 nm, and 729 nm (Fig. 11.4 A. & C.); 854 nm, 905 nm, 923 nm, 934 nm, and 972 nm (Fig. 11.5 A. & B.); 794 nm, 799 nm, and 806 nm (Fig. 11.4 A. & 11.6). Fig. 11.4 A. Peaks in the Raman spectra of titanite associated with Sm3+ and Eu3+ luminescence centers as well as peaks likely related to unidentified REE centers; sample R050114, Pakistan, λexcitation = 532 nm, unoriented

?

Sm3+ ?

Eu3+

?

93

Intensity

Fig. 11.4 B. Magnification of peaks attributed to Sm3+ luminescence centers in titanite located at 574 nm (5321394 cm-1), 578 nm (5321526 cm-1), 589 nm (5321720 cm-1), 600 nm (5322151 cm-1), 613 nm (5322491 cm-1), 617 nm (5322597 -1 532 -1 cm ), and 620 nm ( 2682 cm ); sample R050114, unoriented

Sm3+ Sm3+

Sm3+

-1 Raman Shift (cm )

Fig. 11.4 C. Magnification of peaks attributed to Eu3+ luminescence centers in titanite located at 689nm (5324263

Intensity

-1 532 -1 532 -1 cm ) and 703 nm ( 4545 cm and 4651 cm ) as well as peaks likely related to unidentified REE centers located 532 -1 532 -1 at 639 nm ( 3162 cm ), 647 nm ( 3360 cm ), 658 nm (5323619 cm-1), 664 nm (5323737 cm-1), 718 nm (5324870 -1 532 -1 cm ), and 729 nm ( 5086 cm ); sample R050114, Pakistan, unoriented

Eu3+ ?

? ?

Raman Shift (cm-1)

Fig. 11.5 A. Peaks related to Nd3+ as well as peaks likely related to unidentified REE centers in titanite from Kingman, AZ; sample R050039, λexcitation = 785 nm, unoriented

Nd3+

? ?

94

Fig. 11.5 B. Magnification of peaks related to Nd3+ luminescence centers located at 867 nm (7851269 cm-1), 883 nm (7851374 cm-1), and 894 nm (7851588 cm-1) and multiple peaks likely related to unidentified REE centers located at 854 785 -1 785 -1 785 -1 785 -1 785 -1 nm ( 1038 cm ), 905 nm ( 1696 cm ), 923 nm ( 1904 cm ), 934 nm ( 2039 cm ), and 972 nm ( 2455 cm ) in the Raman spectrum of titanite from Kingman, AZ; sample R050039, unoriented

?

Intensity

Nd3+

? ?

?

-1 Raman Shift (cm )

Fig. 11.6 Magnification of peaks located at 794 nm (5326217 cm-1), 799 nm

Intensity

532 -1 532 -1 ( 6290 cm ), and 806 nm ( 6393 cm ) likely related to unidentified REE centers in titanite from Pakistan; sample R050114, unoriented

? ?

?

Raman Shift (cm-1)

95

References Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102 Frost, B.R., Chamberlain, K.R. & Schumacher, J.C. (2001) Sphene (titanite): phase relations and role as a geochronometer. Chemical Geology, Vol. 172, No. 1-2, pp. 131-148. Gaft, M., Nagli, L., Reisfeld, R. & Panczer, G. (2003) Laser-induced time-resolved luminescence of natural titanite CaTiOSiO4 . Optical Materials, Vol. 24, No. 1-2, pp. 231-241. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Ghose, S., Ito, Y. & Hatch, D.M. (1991) Paraelectric-antiferroelectric phase transition in titanite, CaTiSiO5 . Physics and Chemistry of Minerals, Vol. 17, No. 7, pp. 591-603. Hammer, V.M.F., Beran, A., Endisch, D. & Rauch, F. (1996) OH concentration in natural titanites determined by FTIR spectroscopy and nuclear reaction analysis. European Journal of Mineralogy, Vol. 8, No. 2, pp. 281. Isetti, G. & Penco, A.M. (1968) La posizione dell’Idrogeno ossidrilico nella titanite. Mineralogica et Petrographica Acta, Vol. 14, pp. 115–122. Nasdala, L., Irmer, G. & Wolf, D. (1995) The degree of metamictization in zircon; a Raman spectroscopic study. European Journal of Mineralogy, Vol. 7, No. 3, pp. 471. Salje, E., Schmidt, C. & Bismayer, U. (1993) Structural phase transition in titanite, CaTiSiO5: A Raman spectroscopic study. Physics and Chemistry of Minerals, Vol. 19, No. 7, pp. 502-506. Vance, E.R. & Metson, J.B. (1985) Radiation damage in natural titanites. Physics and Chemistry of Minerals, Vol. 12, No. 5, pp. 255-260. Zhang, M. & Salje, E.K.H. (2003) Spectroscopic characterization of metamictization and recrystallization in zircon and titanite. Phase Transitions: A Multinational Journal, 76, Vol. 1, No. 2, pp. 117-136. Zhang, M., Salje, E.K.H., Bismayer, U., Groat, L.A. & Malcherek, T. (2002) Metamictization and recrystallization of titanite: An infrared spectroscopic study. American Mineralogist, Vol. 87, No. 7, pp. 882-890.

Additional Information Beran, A. & Libowitzky, E. (2006) Water in natural mantle minerals II: olivine, garnet and accessory minerals. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 169-191. Beran, A. & Libowitzky, E. (2003) IR spectroscopic characterization of OH defects in mineral phases. Phase transitions, Vol. 76, No. 1-2, pp. 1-15. Gaft, M., Nagli, L., Reisfeld, R. & Panczer, G. (2003) Laser-induced time-resolved luminescence of natural titanite CaTiOSiO4. Optical Materials, Vol. 24, No. 1-2, pp. 231-241. Johnson, E.A. (2006) Water in nominally anhydrous crustal minerals: speciation, concentration, and geologic significance. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 117-154. Rossman, G.R. (2006) Analytical methods for measuring water in nominally anhydrous minerals. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 1-28. Rossman, G.R. (1996) Studies of OH in nominally anhydrous minerals. Physics and Chemistry of Minerals, Vol. 23, No. 4, pp. 299-304.

96

Topaz Gem names: sherry topaz, Imperial topaz Ideal chemistry: Al2SiO4F2 Crystal system: orthorhombic Point Group:

H-M: mmm

S: D2h

Space Group: Pbnm

topaz from Antonio Peireira Topaz mine, Minas Gerais, Brazil

Table of Atomic Coordinates (Gatta et al., 2006): atom

x

y

z

Al

0.9053

0.13131

0.0816

Si

0.39976

0.94088

0.25

O1

0.7064

0.0311

0.25

O2

0.4516

0.7564

0.25

O3

0.2114

0.98979

0.09284

F

0.5982

0.25263

0.05988

Raman mode analysis: (3 x 8d) + (3 x 4c)

Raman Active Modes Atom

Wyckoff Position Point Symmetry Ag A1g B2g B3g

Al, O3, F

8d

1

3

3

3

3

Si, O1, O2

4c

m

2

1

2

1

Raman mode analysis predicts the existence of 54 active Raman modes in topaz: 15Ag + 12B1g + 15B2g +12B3g = 54

97

Spectral Features Related to OHIn the crystal structure of topaz, two F atoms bond with an AlO4 group to form distorted octahedra. Hydroxl (OH-) commonly substitutes for fluorine in varying concentrations, depending on the origin of the topaz sample (Beny and Piriou, 1987). Multiple peaks associated with OH- are can be observed in Raman spectra of topaz. Pinheiro et al. (2002) report the presence of a single peak centered at 3647 cm-1 in the Raman spectra of topaz containing less than 10 wt% OH- and associate this peak with the substitution of F for OH- (Fig. 12.5). In samples containing greater than 10 wt% OH-, this peak becomes asymetrical. The asymmetry is related to the presence of a second OH- Raman peak centered at 3639 cm-1 (Pinheiro et al., 2002). Pinheiro et al. (2002) suggest that the appearance of the second OH- peak at 3639 cm-1 is related to a change in the local symmetry from D162h to C92v producing two nonequivalent F sites that can be occupied by OH-. A Raman peak centered at 1165 cm-1 has been assigned to the inplane stretching mode of hydroxl (Fig. 12.1) (Beny and Piriou, 1987). Pinheiro et al. (2002) report the presence of two IR bands associated with hydroxyl in topaz; 1) a band centered at 1150 cm-1, related to Al-OH- bending and 2) a band centered at 3600 cm-1 associated with an OH- stretching mode. OHcontent is not related to the various colors of topaz (see next section on chromophores in topaz). FIG. 12.1 Raman peaks associated with hydroxyl in topaz; a peak centered at 3650 cm-1 is related to the

substitution of F for OH- and a peak centered at 1165 cm-1 is assigned to a stretching mode of OH- ; sample R050176, colorless topaz from Nigeria, λexcitation = 532 nm, unoriented

OH-

-

OH

98

Spectral Features Related to Chromophores and Other Ions Unlike beryl, the color varieties of topaz are dominantly the result of radiation-induced color centers, rather than the result of electronic transitions produced by chromophoric ions. The pink color in topaz is undoubtebly produced by substitution of octahedrally coordinated Al3+ for Cr3+ however, the details surrounding irradiation-induced centers in other color varieties of topaz remain ambiguous (Taran et al., 2003; Fritsch and Rossman, 1988). Gaft et al. (2005) attribute yellow color in topaz to Cr3+ in combination with a radiation induced O- hole. Cr3+ combined with a F-center (vacancy with a trapped electron) produce orange-red topaz (Gaft et al., 2005). Red-brown Cr-deficient topaz is colored by O- and F center combinations (Gaft et al., 2005). Blue topaz is produced by R-centers: two F-centers with two trapped electrons (Gaft et al., 2005). In a study of topaz from various Brazilian localities, Taran et al. (2003) discovered that Cr3+ is also responsible for pink to violet colors, while Cr4+ produces a red-orange Imperial topaz color, and a combination of Cr3+ and Cr4+ results in pink-orange colors. Luminescence studies reported by Gaft et al. attribute peaks located at 680, 696, 712 and 730 nm to multiple Cr3+ luminescence centers (Fig. 12.2) (Gaft et al., 2005). FIG. 12.2 Two distinct peaks centered at 680 nm (5324075 cm-1) and 683 nm (5324160 cm-1) related to Cr3+ centers; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R040121, pink topaz from Mexico, unoriented

Cr3+ Cr3+

The Raman spectra of two brown topaz crystals from different mines in Minas Gerais, Brazil have a unique set of peaks distinguishing them from the other color varieties of topaz in this study (Fig. 12.3 A. & B.). Peaks centered at 678 nm, 683 nm, 712 nm, and 733 nm are likely related to the Cr3+ luminescence centers observed by Gaft et al. (2005). FIG. 12.3 A. Peaks centered at 678 nm (5324055cm-1), 683 nm (5324175 cm-1), 712 nm (5324762cm-1), and 733 nm

532 -1 3+ ( 5155 cm ) in the Raman spectrum of topaz attributed to Cr luminescence centers; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060024, brown topaz from Minas Gerais, Brazil, unoriented

Cr3+

99

FIG. 12.3 B. Magnification of peaks in Raman spectra attributed to Cr3+ luminescence centers; brown topaz from Minas Gerais, Brazil, sample R060026 (blue spectrum), sample R060024 (black spectrum), λexcitation = 532 nm, unoriented

Intensity

Cr3+

Cr3+

Raman Shift (cm-1)

References Beny, J.M. & Piriou, B. (1987) Vibrational spectra of single-crystal topaz. Physics and Chemistry of Minerals, Vol. 15, No. 2, pp. 148-159. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Gatta, G.D., Nestola, F., Bromiley, G.D. & Loose, A. (2006) New insight into crystal chemistry of topaz: A multimethodological study. American Mineralogist, Vol. 91, No. 11-12, pp. 1839-1846. Pinheiro, M.V.B., Fantini, C., Krambrock, K., Persiano, A.I.C., Dantas, M.S.S. & Pimenta, M.A. (2002) OH/F substitution in topaz studied by Raman spectroscopy. Physical Review B, Vol. 65, No. 10, pp. 104301. Ribbe, P.H. & Gibbs, G.V. (1971) The crystal structure of topaz and its relation to physical properties. American Mineralogist, Vol. 56, pp. 24-30. Taran, M.N., Tarashchan, A.N., Rager, H., Schott, S., Schürmann, K. & Iwanuch, W. (2003) Optical spectroscopy study of variously colored gem-quality topazes from Ouro Preto, Minas Gerais, Brazil. Physics and Chemistry of Minerals, Vol. 30, No. 9, pp. 546-555.

Additional Information Dickinson, A.C. & Moore, W.J. (1967) Paramagnetic resonance of metal ions and defect centers in topaz. Journal of Physical Chemistry, Vol. 71, No. 2, pp. 231-240. Fritsch, E. & Rossman, G.R. (1988) An update on color in gems. Part 2: Colors involving multiple atoms and color centers. Gems & Gemology, Vol. 24, No. 1, pp. 3–15. Fritsch, E. & Rossman, G.R. (1987) An Update on Color in Gems. Part 1. Introduction and Colors caused by Dispersed Metal Ions. Gems & Gemology, Vol. 23, No. 3, pp. 126–139. 3+ Gaft, M., Nagli, L., Reisfeld, R., Panczer, G. & Brestel, M. (2003) Time-resolved luminescence of Cr in topaz Al2SiO4 (OH, F) 2. Journal of Luminescence, Vol. 102, pp. 349-356. Kloprogge, J.T. & Frost, R.L. (2000) Raman microscopic study at 300 and 77 K of some pegmatite minerals from the Iveland–Evje area, Aust-Agder, Southern Norway. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, Vol. 56, No. 3, pp. 501-513. Komatsu, K., Kagi, H., Okada, T., Kuribayashi, T., Parise, J.B. & Kudoh, Y. (2005) Pressure dependence of the OHstretching mode in F-rich natural topaz and topaz-OH. American Mineralogist, Vol. 90, No. 1, pp. 266-270. Krambrock, K., Ribeiro, L.G.M., Pinheiro, M.V.B., Leal, A.S., Menezes, M. B.C. & Spaeth, J.M. (2007) Color centers in topaz: comparison between neutron and gamma irradiation. Physics and Chemistry of Minerals, Vol. 34, No. 7, pp. 437-444. Londos, C., Vassilkou-Dova, A., Georgiou, G. & Fytros, L. (1992) Infrared studies of natural topaz. Physica status solidi A, Applied research, Vol. 133, No. 2, pp. 473-479.

100

Rossman, G.R. (2006) Analytical methods for measuring water in nominally anhydrous minerals. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 1-28. Rossman, G.R. (1996) Studies of OH in nominally anhydrous minerals. Physics and Chemistry of Minerals, Vol. 23, No. 4, pp. 299-304. Schott, S., Rager, H., Schurmann, K. & Taran, M. (2003) Spectroscopic study of natural gem quality “Imperial" topazes from Ouro Preto, Brazil. European Journal of Mineralogy, Vol. 15, No. 4, pp. 701. Shinoda, K. & Aikawa, N. (1997) IR active orientation of OH bending mode in topaz. Physics and Chemistry of Minerals, Vol. 24, No. 8, pp. 551-554. Souza, D.N., Fernandes de Lima, J., Valerio, M.E.G., Fantini, C., Pimenta, M.A., Moreira, R.L. & Caldas, L.V.E. (2002) Influence of thermal treatment on the Raman, infrared and TL responses of natural topaz. Nuclear Inst. and Methods in Physics Research, B, Vol. 191, No. 1-4, pp. 230-235. 3+ Tarashchan, A.N., Taran, M.N., Rager, H. & Iwanuch, W. (2006) Luminescence spectroscopic study of Cr in Brazilian topazes from Ouro Preto. Physics and Chemistry of Minerals, Vol. 32, No. 10, pp. 679-690.

101

Tourmaline General Formula: X Y3 Z6 Si6O18 (BO3)3 (OH,O, F)4

Crystal system: Trigonal Gem names: rubellite, indicolite, Paraiba, watermelon

Point Group: H-M: 3m S: C3v Zoned Elbaite (“watermelon” tourmaline)

Space Group: R3m List of Ideal Chemistries of Major Gem Tourmalines: Elbaite:

Na(Al1.5Li1.5)Al6(BO3)3Si6O18(OH)4

Liddicoatite: Ca(Li2Al)Al6(BO3)3Si6O18(OH)3F Dravite:

NaMg3Al6(BO3)3Si6O18(OH)4

Uvite:

CaMg3(Al5Mg)(BO3)3Si6O18(OH)3F

Note: Raman mode analysis and atomic coordinates for elbaite are provided below

Table of Atomic Coordinates (Bosi et al., 2005): atom

x

y

z

NaX

0

0

0.23267

AlY

0.12318

0.06159

0.62744

AlZ

0.29758

0.26086

0.61151

B

0.10969

0.21938

0.45457

SiT

0.19192

0.18996

0

O1

0

0

0.78262

O2

0.06091

0.12182

0.48293

O3

0.26857

0.13429

0.50990

O4

0.09335

0.18670

0.07149

O5

0.18670

0.09335

0.09304

O6

0.19695

0.18684

0.77565

O7

0.28554

0.28578

0.08043

O8

0.20989

0.27071

0.44137

H3

0.2664

0.1332

0.3818

102

Raman mode analysis:

Raman Active Modes Atoms AlZ, SiT, O6, O7, O8 AlY, B, O2, O3, O4, O5, H3 NaX, O1

Wyckoff Position

A1

E

18c

3

6

9b

2

3

3a

1

1

(5 × 18c) + (7 × 9b) + (2 × 3a) Raman mode analysis predicts the existence of active Raman modes in : 31A1 + 53E = 84

103

Spectral Features Related to OHMultiple studies have been conducted to interpret IR bands associated with hydroxyl in various tourmaline species. OH- can substitute into two positions in the crystal structure of tourmaline (Gonzales-Carreño, 1988; Castañeda et al., 2000). The first position, denoted OH1, is located at the center of the hexagonal rings with the oxygen coordinated by three octahedral Y-site cations (Fig. 13.1) (Gonzales-Carreño, 1988; Castañeda et al., 2000). In the second position, denoted OH3, the hydroxyl occupies a position along the edge of the hexagonal columns where the oxygen is coordinated by one Y and two Z cations (Fig. 13.1) (Gonzales-Carreño, 1988; Castañeda et al., 2000). The type of cations occupying the Y- and Z-sites in the crystal structure affect the frequency and bandwidth of the OH- bands (Gonzales-Carreño, 1988). In general, the highest frequency bands (3600+ cm-1) are associated with OH1 while those appearing at 3600-3400 cm-1 are assigned to OH3 vibrations (Castañeda et al., 2000). Factor group analysis conducted by Castaneda et al. (2000) predicts three IR active modes for OH- in tourmaline, one for OH1 and two for OH3; only two of these modes are Raman active. Two tables comparing the locations of IR bands attributed to OH- observed in the literature with Raman peaks positions observed in this study are provided below (Table 13.1 & 13.2). Table 13.1 Locations of IR bands attributed OH- and their associated coordinated cations in tourmaline Tourmaline Variety

Dravite Schorl Schorl Elbaite Ca-Elbaite

OH1 Location (cm-1) 3738 3738 3633 3650

Coordinated Cations MgYMgYMgY MgYMgYMgY FeYFeYFeY LiYAlYAlY

3650/3680

LiYAlYAlY Y

Y

Y

Tourmaline (various)

Coordinated Cations R2+YAlZAlZ

3553 3583 3463 3586/3604 3474/3507 3594 3568 3492

R2+YAlZAlZ LiYAlZAlZ AlYAlZAlZ LiYAlZAlZ AlYAlZAlZ LiYAlZAlZ R2+YAlZAlZ AlYAlZAlZ

Li Mn Al

3692

LiYAlYFeY

3594 3558 3492

LiYAlZAlZ R2+YAlZAlZ AlYAlZAlZ

3641, 3646-47

(Al,Li) YAlYAlY

3457-60, 3580-82

(Al,Li)YAlZAlZ

Fe-Elbaite

Fe-Elbaite

Location (cm-1) 3568

3670 Mn-Elbaite

Elbaite

H2O

OH3

3585, 3592

3645, 3648-51 3628-31

Y

Y

Li Al Fe

Y

(Al,Li) YAlYAlY Y

Y

Fe Al Al

Y

3478-79, 3484, 3557, 3560 3468-70, 3478, 3490-92 3559, 3564 3585-3597

Y

Z

Z

(Fe,Li,Al) Al Al

AlYAlZAlZ Y

Z

Z

Fe Al Al LiYAlZAlZ

Reference

GonzalesCarreño, 1988

3170, 3340 3165, 3175, 3360, 3380

Castañeda et al., 2000

Zhang et al., 2008

104

Table 13.2 Raman peaks in tourmaline associated with hydroxyl and the possible cation coordination assignments based on IR data; 532 nm laser, unoriented

Tourmaline Variety

Elbaite

Elbaite

Liddicoatite

Liddicoatite

Dravite

Uvite

Sample #

R050487, R060652, R050260

R060560, R050119, R060003, R060566

R060635

R060969

R050077

R050172

OH peaks cm-1

Possible Assignment

3496

OH3 AlYAlZAlZ

3566

OH3 FeYAlZAlZ, OH3(R2+YAlZAlZ)

3598

OH3 LiYAlZAlZ

3476

OH3 AlYAlZAlZ

3592

OH1 LiYAlYFeY

3656

?

3511 3609

? ?

3489

OH3 AlYAlZAlZ Y

Z

Z

3595

OH3 Li Al Al ?

3571

?

3740

OH1MgYMgYMgY

3743

13.2 B. & C.

13.2 A. & C.

Reference Gonzales-Carreño, 1988 Zhang et al., 2008; Gonzales-Carreño, 1988 Zhang et al., 2008 Gonzales-Carreño, 1988 Castañeda et al., 2000

13.3 A. & C.

3657

3584

Figure

OH1 LiYAlYFeY, OH3 LiYAlZAlZ ?

13.3 B. & C. 13.4 A. & B. 13.5 A. & B.

Zhang et al., 2008 Zhang et al., 2008

Gonzales-Carreño, 1988 Castañeda et al., 2000; Zhang et al., 2008

Fig. 13.1 The crystal structure of elbaite viewed down the c-axis showing sites of hydroxyl substitution; red spheres: oxygen; blue spheres: Al; green spheres: substitutional site for OH3; pink sphere substitutional site for OH1; teal cation below OH1 substutional site is Na

OH3

OH1

105

Fig. 13.2 A. Peaks associated with OH in Raman spectra of elbaite; 1) sample R050119, 2) R060560, 3) R060003, 4) R060566; 532 nm laser, unoriented

-

OH

4.

3. 2. 1.

Fig. 13.2 B. Peaks associated with OH in Raman spectra of elbaite; 5) sample R060562, 6) R050260, 7) R050487; 532 nm laser, unoriented OH7.

6.

5.

106

Fig. 13.2 C. Magnification of peaks associated with OH in elbaite, note the variation in frequency and peak shape; 1) sample R050119, 2) R060560, 3) R060003, 4) R060566, 5) R060562, 6) R050260, and 7) R050487; 532 nm laser, unoriented

Intensity

7.

6.

5.

4.

3. 2. 1.

Raman Shift (cm-1)

Fig. 13.3 A. Peaks associated with OH in the Raman spectrum of liddicoatite; 1) sample R060635, brown gem of unknown locality; 532 nm laser, unoriented

OH-

107

Fig. 13.3 B. Peaks associated with OH in the Raman spectrum of liddicoatite; sample R060969, red crystal from Madagascar; 532 nm laser, unoriented

OH-

Fig. 13.3 C. Magnification of peaks associated with

Intensity

OH in liddicoatite, note the variation in frequency and peak shape; 1) sample R060635; centered at 3511 -1 and 3609 cm (no literature match), brown gem of unknown locality; 2) sample R060969, centered at 3489 and 3595 cm-1, red crystal from Madagascar, 532 nm laser, unoriented

2. 1.

Raman Shift (cm-1)

Fig. 13.4 A. Peaks associated with OH in dravite; sample R050077, dark brown crystal from Western Australia, 532 nm laser, unoriented

OH-

108

Fig. 13.4 B. (left) Magnification of peaks centered at 3571 and 3740 cm-1

Intensity

associated with OH in dravite; sample R050077, dark brown crystal from Western Australia, 532 nm laser, unoriented

OHRaman Shift (cm-1)

Fig. 13.5 A. Peaks associated with OH in uvite; sample R050172, 532 nm laser, unoriented

OH-

Fig. 13.5 B. (right) Magnification of peaks centered 3584 and

Intensity

3743 cm-1 associated with OH in uvite; sample R050172, 532 nm laser, unoriented

OH-

Raman Shift (cm-1)

109

Spectral Features Related to Chromophores Tourmaline comes in a wide variety of colors due to its chemical complexity (see “Table of Color Causes”, Appendix). Several members of the tourmaline group are often multi-colored and zoned due to compositional changes during crystal growth (Dunn et al., 1977). Liddicoatite (the structural analog of elbaite) from Madagascar is famous for its striking color patterns (Dunn et al., 1977). Elbaite, uvite, and dravite commonly incorporate iron into their crystal structures, Fe2+ into the Z-site and Fe3+ into the Y- and Z-sites, and this greatly affects the color (Mattson and Rossman, 1987). There are many colors of tourmaline for which the dominant color-inducing mechanism is charge transfer involving Fe including blue, green, and brown elbaite and liddicoatite and yellow to brown dravite (Fritsch and Rossman, 1988). Green dravite and uvite are produced by octahedrally coordinated Cr3+ and V3+ (Fritsch and Rossman, 1988). The presence of Mn is likely responsible for yellow-green, pink, and red tourmalines (Fritsch and Rossman, 1988). Recent studies of the newly discovered cuprian elbaite from Paraiba, Brazil attribute its distinctive, nearly neon blue and green hues to a high concentration (greater than 0.1 wt%) of Cu2+ (Rossman et al., 1991). Although extensive studies have been conducted to understand the processes that produce color in tourmaline, further investigation is necessary in order to fully comprehend the details. Several spectral features related to chromophores have been observed in the Raman spectra of tourmaline. Gaft et al. (2005) attribute luminescence emission centers located at 697, 707, 691/692 nm to the presence of Cr3+ in tourmaline. Peaks likely related to the presence of Cr3+ centers were observed in the Raman spectra of uvite centered at 680 nm and 683 nm (Fig. 13.6 A. & B.), and dravite centered at 4185 cm-1 (684 nm) (Fig. 13.7 A. & B.). Although the positions of these peaks do not match those provided by Gaft et al. (2005) for Cr3+ luminescence centers in tourmaline, they do correspond to the locations of trivalent chromium centers in several other minerals including beryl, chrysoberyl, and topaz. Fig. 13.6 A. Doublet associated with Cr3+ luminescence centers in uvite; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R050301, Bahia, Brazil, λexcitation = 532 nm, unoriented

Cr3+

110

Fig. 13.7 A. Peak centered at 684 nm (5324185 cm-1) associated with a Cr3+ luminescence center in dravite; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R040088, brown crystal from Western Australia; unoriented

Cr3+

Fig. 13.6 B. (right) Magnification of doublet centered at 680

3+

(5324110 cm-1) and 683 nm (5324160 cm-1) associated with Cr uvite; sample R050301 from Bahia Brazil, unoriented

in

Intensity

Cr3+

Raman Shift (cm-1)

Intensity

Cr3+

Fig. 13.7 B. (left) Magnification of peak centered at 684 nm 532

-1

3+

( 4185 cm ) associated with a Cr luminescence center in dravite; sample R040088, brown crystal from Western Australia; unoriented

Raman Shift (cm-1)

111

Spectral Features Related to Rare-earth Elements The geochemistry of REE distribution in tourmalines is a valuable tool for modeling the development of various hydrothermal ore deposits, particularly gold deposits (Jiang et al., 2004). Sm-Nd isotopic data can help constrain pressure, temperature, and depth conditions as well as preserve geochemical information regarding ore-forming fluids (Jiang et al., 2004). Although tourmaline typically contains only very small concentrations of REE (ppm’s), its resistance to weathering and ability to precipitate from hydrothermal systems over a broad range of conditions make it a viable marker mineral for economic geologists (Jiang et al., 2004). Gaft et al. (2003) report the presence of multiple luminescence centers attributed to Nd3+ in titanite located at 860 nm, 869 nm, 878 nm, 896 nm, and 1053 nm. The same centers are observed in the Raman spectra of elbaite and liddicoatite in this study (Fig. 13.8 & 13.9). Fig. 13.8 Peaks located at 860 nm (7801196 cm-1), 869 nm (7801314 cm-1), 878 nm (7801433 cm-1), 896 nm (7801668 3+ -1 780 -1 cm ), and 1053 nm ( 3323 cm ) associated with Nd centers in elbaite; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060631, unknown locality, unoriented

Nd3+

Nd3+

Fig. 13.9 Peaks located at 860 nm (7801196 cm-1), 869 nm (7801314 cm-1), 878 nm (7801433 cm-1), 896 nm (7801668 3+ cm-1), and 1053 nm (7803323 cm-1) associated with Nd centers in liddicoatite; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060635, unknown locality, unoriented

Nd3+

Nd3+

112

References Bosi, F., Agrosi, G., Lucchesi, S., Melchiorre, G. & Scandale, E. (2005) Mn-tourmaline from island of Elba (Italy): Crystal chemistry. American Mineralogist, Vol. 90, No. 10, pp. 1661-1668. Castañeda, C., Oliveira, E.F., Gomes, N. & Soares, A.C.P. (2000) Infrared study of OH sites in tourmaline from the elbaite-schorl series. American Mineralogist, Vol. 85, No. 10, pp. 1503-1507. Dunn, P.J., Appleman, D.E. & Nelen, J.E. (1977) Liddicoatite, a new calcium end-member of the tourmaline group. American Mineralogist, Vol. 62, No. 11-12, pp. 1121-1124. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: colors caused by bandgaps and physical phenomena. Gems & GemoA logy, Vol. 24, No. 2, pp. 81-102. Gaft, M., Nagli, L., Reisfeld, R. & Panczer, G. (2003) Laser-induced time-resolved luminescence of natural titanite CaTiOSiO4 . Optical Materials, Vol. 24, No. 1-2, pp. 231-241. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer. Gonzalez-Carreño, T., Fernandez, M. & Sanz, J. (1988) Infrared and electron microprobe analysis of tourmalines. Physics and Chemistry of Minerals, Vol. 15, No. 5, pp. 452-460. Jiang, S.Y., Yu, J.M. & Lu, J.J. (2004) Trace and rare-earth element geochemistry in tourmaline and cassiterite from the Yunlong tin deposit, Yunnan, China: implication for migmatitic–hydrothermal fluid evolution and ore genesis. Chemical Geology, Vol. 209, No. 3-4, pp. 193-213. 2+ 3+ Mattson, S.M. & Rossman, G.R. (1987) Fe -Fe interactions in tourmaline. Physics and Chemistry of Minerals, Vol. 14, No. 2, pp. 163-171. Rossman, G.R., Fritsch, E. & Shigley, J.E. (1991) Origin of color in cuprian elbaite from São José de Batalha, Paraíba, Brazil. American Mineralogist, Vol. 76, No. 9-10, pp. 1479-1484. Zhang, A., Wang, R., Li, Y., Hu, H., Lu, X., Ji, J. & Zhang, H. (2008) Tourmalines from the Koktokay No. 3 pegmatite, Altai, NW China: spectroscopic characterization and relationships with the pegmatite evolution. European Journal of Mineralogy, Vol. 20, No. 1, pp. 143.

Additional Information De Oliveira, E.F., Castañeda, C., Eeckhout, S.G., Gilmar, M.M., Kwitko, R.R., De Grave, E. & Botelho, N.F. (2002) Infrared and Mossbauer study of Brazilian tourmalines from different geological environments. American Mineralogist, Vol. 87, No. 8-9, pp. 1154-1163. Gasharova, B., Mihailova, B. & Konstantinov, L. (1997) Raman spectra of various types of tourmaline. European Journal of Mineralogy, Vol. 9, No. 5, pp. 935. Manning, P.G. (1973) Effect of second-nearest-neighbour interaction on Mn3+ absorption in pink and black tourmalines. Canadian Mineralogist, Vol. 11, No. 5, pp. 971. Peng, M., Mao, H.K., Chen, L.G. & Chan, E.E.T. (1988) The Polarized Raman Spectra of Tourmaline. Annual Report of the Director of the Geophysical Laboratory, Carnegie Inst.Washington, Vol. 1989.

113

Zircon Ideal chemistry: ZrSiO4 Crystal system: tetragonal Point Group:

H-M: 4/mmm

S: D4h

Space Group: I41/amd Photo courtesy of Stone Group Labs

Table of Atomic Coordinates (Finch et al., 2001): atom

x

y

z

Si

0

0.75

5/8

Zr

0

0.75

1/8

O

0

0.0657

0.1961

Raman mode analysis:

Raman Active Modes Atom Wyckoff Position Point Symmetry A1g B1g B2g Eg O

16h

m

2

2

1

3

Si

4b

4 2m

-

1

-

1

Zr

4a

4 2m

-

1

-

1

Raman mode analysis predicts the existence of 12 active Raman modes in zircon: 2A1g + 4B1g + 1B2g + 5E g = 12

114

Introduction to Raman Spectrum of Zircon Zircon is an isosilicate consisting of isolated SiO4 groups and dodecahedrally coordinated Zr atoms creating open channels parallel to the c-axis in the crystal structure (Fig. 14.2). The Raman spectrum of zircon is presented here (Fig. 14.1). The most intense peak in the spectrum, centered at 357 cm-1, is associated with restricted rotation of SiO4 groups (Kolesov et al., 2001). The next most intense peak, centered at 1008 cm-1, is associated with Si-O stretching modes, while the peak centered at 438 cm-1, is associated with O-Si-O bending (Kolesov et al., 2001).

Fig. 14.1 The Raman spectrum of zircon; mode assignments reported by Kolesov et al. (2001); sample R050203, 532 nm laser, processed, unoriented

Intensity

357 Eg

215 B1g

202 Eg

439 A1g

1008 B1g

225 Eg

974 A1g

393 B1g

Raman Shift (cm-1)

Fig. 14.2 The crystal structure of zircon; blue tetrahedra: SiO4 groups, pink polyhedra: ZrO8 groups A. View down the a-axis

B. View down the c-axis

115

Orientational Dependence of Spectra When zircon is oriented with the a-axis parallel to the incident laser beam (the cross-section of the uniaxial optical indicatrix is an ellipse), the intensities of the peaks change dramatically with rotation due to a change in the degree of freedom of atomic vibration (Fig. 14.3 A). Notice that when the laser is polarized parallel to the c-axis, the peaks centered at 968 and 1001 cm-1, associated with Si-O stretching modes nearly disappear. When the sample is oriented with the c-axis parallel to the laser (the crosssection of the uniaxial optical indicatrix is a circle), there is no noticeable change in peak intensities due to an equal distribution of atomic vibration in all directions (FIG. 14.3 B). In spite of the orientational dependence of its Raman peaks, zircon can be accurately identified by the spectrum of a randomly oriented sample. Fig. 14.3 Raman spectra of oriented corundum crystals; sample R050034, reddish brown zircon, Oaxaca, Mexico, processed

A. Raman spectra showing peak intensities as a function of orientation; 514.5 nm laser parallel to a*; at 0° laser is polarized ll to c-axis; at 90° laser is polarized perpendicular to c-axis; sample R050034, processed

B. Raman spectra showing peak intensities as a function of orientation. Notice that the intensities do not change as the direction of polarization changes; 514.5 nm laser parallel to c-axis; at 0° laser is polarized ll to b-axis; at 90° laser is polarized perpendicular to b-axis

116

Cause of Color in Zircon The color centers of zircon are complex and the details surrounding the color-inducing mechanisms are still debated. Zircon comes in a variety of colors and most of these colors are generated by natural irradiation. Zircon, like titanite, frequently incorporates radioactive atoms such as U and Th into its crystal structure. Red zircon have radiation-induced color centers in which Nb4+ substitutes for Zr4+ (Fielding, 1970; Fritsch and Rossman, 1988). Blue zircon is attributed to the presence of U4+ (Mackey et al., 1975; Fritsch and Rossman, 1988). No spectral features attributed to these color centers have been observed in this study.

Metamictization Due to Radiation Damage Zircon typically incorporates radioactive atoms like U and Th into its crystal structure. As these elements radioactively decay over time, the crystal structure is compromised resulting in metamictization (breakdown of the crystal structure) (Johnson, 2006). Metamictization can make a crystal slightly or completely amorphous depending on the amount of radiation exposure; this can affect the X-ray diffraction patterns, unit cell parameters, and IR and Raman spectra of the crystal (Zhang et al., 2000; Nasdala et al., 1995). Raman spectroscopic studies of radiation-damaged zircons conducted by Zhang et al. (2000) and Nasdala et al. (1995) show widening of Raman peaks as well as frequency shifts due to an increase in the irregularity of bond lengths and angles in the crystal structure. In particular, the peaks associated with the stretching modes of Si-O, located at 900-1000 cm-1, broadened and although the zircon structure was damaged, Zhang et al. (2000) conclude that in comparison with nearly amorphous silica, the SiO4 groups in metamict zircon are less polymerized. The Raman spectra of metamict zircon display peaks associated with both a damaged crystalline zircon, and an amorphous zircon (observed between 850-1100 cm-1) (Zhang et al., 2000). Comparisons of diffraction peaks and Raman spectra of zircon samples that may have undergone very slight radiation damage are shown in Fig. 14.4. The unit cell parameters of the zircon samples included in this study are provided in Table 14.1.

Fig. 14.4 (Right) Portion of the XRD pattern of zircon with associated band width values and (left) portion of the Raman spectra of zircon showing possible slight radiation damage; 1) sample R050034, 2) sample R050286, 3) sample R050203, 4) R050488 (this sample has likely been exposed to the most radiation); 514 nm laser, processed

4.

Intensity

Intensity

.126

.120 .099

3. 2.

.095 1. -1 Raman Shift (cm )

2 Theta

117

Table 14.1 Unit cell parameters of zircon samples from RRUFF database Zircon color, Origin, & Sample number

Unit cell parameters

Structural State

Reference

a (Å)

c (Å)

V (Å3)

red-brown; Oaxaca, Mexico (R050034)

6.6077(1)

5.9957(2)

261.78(1)

likely well crystallized

RRUFF database

brownish red; Renfrew Cty., Ontario, Canada (R050286)

6.6132(3)

6.0038(3)

262.57(2)

likely well crystallized

RRUFF database

dark red; Eastern Thailand (R050203)

6.6049(1)

5.9801(2)

260.88(1)

likely well crystallized

RRUFF database

brown; Sigulani Village, Tambani Area, Nyassaland (R050488)

6.6058(5)

5.9818(8)

261.03(4)

likely well crystallized

RRUFF database

Spectral Features Related to OHThe details behind the incorporation of OH- and H2O into various structural sites of zircon remain controversial. As with titanite, an increase in metamictization results in an increase in OH- concentration. Well-crystallized zircon exhibit sharp, anisotropic IR peaks associated with OH-, whereas the IR spectra of damaged crystals usually display an additional peak associated with the presence of H2O molecules (Beran and Libowitsky, 2003). IR and Raman studies performed by Nasdala et al. (2001) confirm the presence of at least three peaks centered at 3180, 3385, and 3420 cm-1 associated with OH- defects in crystalline zircon (Fig. 14.5 A & B) (Dawson et al., 1971). Fig. 14.5 A. Raman peak centered at 3424 cm-1 attributed to OH-; sample R050286, natural red zircon, Ontario, Canada, λexcitation = 532 nm, unoriented

OH-

118

Fig. 14.5 B. Magnification of Raman peak centered at 3424

Intensity

cm-1 attributed to OH-; sample R050286, natural red zircon, Ontario, Canada, λexcitation = 532 nm, unoriented

OH-

Raman Shift (cm-1)

119

Spectral Features Related to Rare-earth Elements Like titanite, zircon crystals commonly contain a wide variety of trace elements including, but not limited to, REE, Y, Hf, P, and radioactive atoms, such as U and Th (Hanchar et al., 2001). U-Pb dating constrained by zircon crystals has provided the timing of nearly all geologic events. The high closure temperature of zircon along with its ability to isotopically record geologic processes, make zircon invaluable to geoscientists (Hanchar et al., 2001). There are numerous luminescence features associated with REE present in the Raman spectra of zircon, specifically peaks associated with Eu3+, Sm3+, and Nd3+ centers. Luminescence studies conducted by Gaft et al. (2005) report the presence of Eu3+ centers located at 591, 596, 604, and 614 nm (Fig. 14.6 A. & B.). Additional luminescence features located at 702 and 707 nm in the Raman spectra may also be related to Eu3+ (Fig. 14.6 A. & C.) (Gaft et al., 2005). Peaks related to Sm3+ luminescence centers in zircon are centered at 660 and 666 nm (Fig. 14.6 A. & D.) (Gaft et al., 2005). Gaft et al. (2005) report the presence of peaks located at 870 nm, 879 nm, and 891 nm that are likely related to luminescence centers produced by Nd3+ (Fig. 14.6 A. & B.) (Gaft et al., 2005). There are multiple peaks in the Raman spectra of zircon that have not been presented in the literature, but are likely related to the presence of unidentified REE luminescence centers. The locations of the peaks are as follows: 719 nm, 722 nm, 728 nm (Fig. 14.6 A. & 14.5 A.); 797 nm, 803 nm, and 810 nm (Fig. 14.6 A & 14.8 B.); 909 nm, 917 nm, 928 nm, and 976 nm (Fig. 14.7 A & 14.9). Fig. 14.6 A. Peaks in the Raman spectra of zircon associated with: Eu3+ and Sm3+ luminescence centers, a peak associated with OH-, and multiple peaks of unknown origins; sample R050203, natural red zircon, Thailand, λexcitation = 532 nm, unoriented

Eu3+ ? ?

3+

OH-

Sm

Eu3+

Fig. 14.6 B. Magnification of peaks in the Raman spectrum of zircon attributed to Eu3+ centers located at 591 nm

532 -1 532 -1 532 -1 532 -1 ( 1910 cm ), 596 nm ( 2012 cm ), 604 nm ( 2260 cm ), and 614 nm ( 2560 cm ); sample R050203, natural red zircon, Thailand, unoriented

Intensity

Eu3+

Eu3+

-1 Raman Shift (cm )

120

Fig. 14.6 C. Magnification of peaks in the Raman spectrum of zircon related to Eu3+ centers located at 702 nm (5324572 cm-1) and 707 nm (5324728 cm-1); sample R050203, natural red zircon, Thailand, unoriented

Intensity

Eu3+

-1 Raman Shift (cm )

Fig. 14.6 D. Magnification of a peak attributed to OH- centered at 3424 cm-1 and peaks related to Sm3+

luminescence centers located at 660 nm (5323654 cm-1) and 666 nm (5323772 cm-1) in the Raman spectrum of zircon; sample R050203, natural red zircon, Thailand, unoriented

Intensity

Sm3+

OH-

-1 Raman Shift (cm )

Fig. 14.7 A. Peaks attributed to Nd3+ luminescence centers and unidentified REE; sample R050286, natural red zircon, Ontario, Canada, λexcitation = 780 nm, unoriented

Nd3+

?

121

Fig. 14.7 B. Magnification of peaks attributed to Nd3+ luminescence centers located at 870 nm (7801327 cm-1), 879 nm (7801450 cm-1), and 891 nm (7801600 cm-1); sample R050286, natural red zircon, Ontario, Canada, unoriented

Intensity

Nd3+ Nd3+

Raman Shift (cm-1)

Fig. 14.8 A. (right) Magnification of peaks centered at 719 nm

Intensity

(5324882 cm-1), 722 nm (5324950 cm-1), and 728 nm (5325073 cm-1) likely associated with the luminescence centers of unidentified REE; sample R050203, natural red zircon, Thailand, unoriented

Intensity

Raman Shift (cm-1)

Fig. 14.8 B. (left) Magnification of peaks centered at

532 -1 532 -1 797 nm ( 6254 cm ), 803 nm ( 6348 cm ), and 810 532 -1 nm ( 6450 cm ) likely associated with the luminescence centers of unidentified REE; sample R050203, natural red zircon, Thailand, unoriented

Raman Shift (cm-1)

122

Intensity

Fig. 14.9 Magnification of peaks centered at 909 nm (7801819 cm-1), 917 nm (7801922 cm-1), 928 nm (7802047 cm-1), and 976 nm (7802580 cm-1) likely associated with the luminescence centers of unidentified REE; sample R050286, natural red zircon, Ontario, Canada, unoriented

Raman Shift (cm-1)

References Beran, A. & Libowitzky, E. (2003) IR spectroscopic characterization of OH defects in mineral phases. Phase transitions, Vol. 76, No. 1-2, pp. 1-15. Dawson, P., Hargreave, M.M. & Wilkinson, G.R. (1971) The vibrational spectrum of zircon (ZrSiO4). Journal of Physics C: Solid State Physics, Vol. 4, pp. 240-256. Finch, R.J., Hanchar, J.M., Hoskin, P.W.O. & Burns, P.C. (2001) Rare-earth elements in synthetic zircon: Part 2. A single-crystal X-ray study of xenotime substitution. American Mineralogist, Vol. 86, No. 5-6, pp. 681-689. Gaft, M., Panczer, G., Reisfeld, R. & Uspensky, E. (2001) Laser-induced time-resolved luminescence as a tool for rare-earth element identification in minerals. Physics and Chemistry of Minerals, Vol. 28, No. 5, pp. 347-363. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Hanchar, J.M., Finch, R.J., Hoskin, P.W.O., Watson, E.B., Cherniak, D.J. & Mariano, A.N. (2001) Rare earth elements in synthetic zircon: Part 1. Synthesis, and rare earth element and phosphorus doping. American Mineralogist, Vol. 86, No. 5-6, pp. 667-680. Johnson, E.A. (2006) Water in nominally anhydrous crustal minerals: speciation, concentration, and geologic significance. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 117-154. Kolesov, B.A., Geiger, C.A. & Armbruster, T. (2001) The dynamic properties of zircon studied by single-crystal X-ray diffraction and Raman spectroscopy. European Journal of Mineralogy, Vol. 13, No. 5, pp. 939. 4+ Mackey, D.J., Runciman, W.A. & Vance, E.R. (1975) Crystal-field calculations for energy levels of U in ZrSiO4. Physical Review B, Vol. 11, No. 1, pp. 211-218. Nasdala, L., Beran, A., Libowitzky, E. & Wolf, D. (2001) The incorporation of hydroxyl groups and molecular water in natural zircon (ZrSiO4). American Journal of Science, Vol. 301, No. 10, pp. 831. Nasdala, L., Irmer, G. & Wolf, D. (1995) The degree of metamictization in zircon; a Raman spectroscopic study. European Journal of Mineralogy, Vol. 7, No. 3, pp. 471. Zhang, M., Salje, E.K.H., Farnan, I., Graeme-Barber, A., Daniel, P., Ewing, R.C., Clark, A.M. & Leroux, H. (2000) Metamictization of zircon: Raman spectroscopic study. Journal of Physics-Condensed Matter, Vol. 12, No. 8, pp. 1915-1926.

123

Additional Information Ashbaugh, C.E. (1989) Gemstone irradiation and radioactivity. Gems and Gemology, Vol. 25, No. 4, pp. 196-213. Beran, A. & Libowitzky, E. (2006) Water in natural mantle minerals II: olivine, garnet and accessory minerals. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 169-191. Götze, J., Kempe, U., Habermann, D., Nasdala, L., Neuser, R.D. & Richter, D.K. (1999) High resolution cathodoluminescence combined with SHRIMP ion probe measurements of detrital zircons. Mineralogical Magazine, Vol. 63, No. 2, pp. 179-187. Nasdala, L., Zhang, M., Kempe, U., Panczer, G., Gaft, M., Andrut, M. & Plotze, M. (2003) Spectroscopic methods applied to zircon. Reviews in Mineralogy and Geochemistry, Vol. 53, No. 1, pp. 427-467. Rossman, G.R. (2006) Analytical methods for measuring water in nominally anhydrous minerals. Reviews in Mineralogy and Geochemistry, Vol. 62, No. 1, pp. 1-28. Rossman, G.R. (1996) Studies of OH in nominally anhydrous minerals. Physics and Chemistry of Minerals, Vol. 23, No. 4, pp. 299-304. Tomašić, N., Bermanec, V., Gajović, A. & Linarić, M.R. (2008) Metamict Minerals: an Insight into a Relic Crystal Structure Using XRD, Raman Spectroscopy, SAED and HRTEM. Croatica Chemica Acta, Vol. 81, No. 2. Zhang, M. & Salje, E.K.H. (2001) Infrared spectroscopic analysis of zircon: radiation damage and the metamict state. Journal of Physics Condensed Matter, Vol. 13, No. 13, pp. 3057-3072. Zhang, M., Salje, E.K.H. & Ewing, R.C. (2002) Infrared spectra of Si-O overtones, hydrous species, and U ions in metamict zircon: radiation damage and recrystallization. Journal of Physics Condensed Matter, Vol. 14, No. 12, pp. 3333-3352. Zhang, M., Salje, E.K.H., Ewing, R.C., Farnan, I., Ríos, S., Schluter, J. & Leggo, P. (2000) Alpha-decay damage and recrystallization in zircon: evidence for an intermediate state from infrared spectroscopy. Journal of Physics Condensed Matter, Vol. 12, No. 24, pp. 5189-5200.

124

Zoisite Gem names: tanzanite Ideal chemistry: Ca2Al3(Si2O7)(SiO4)(OH)

Crystal system: orthorhombic Point Group:

H-M: mmm

S: D2h Tanzanite gem

Space Group: Pnma Table of Atomic Coordinates (Comodi and Zanazzi, 1997):

atom

x

y

z

Ca1

0.3668

0.25

0.4373

Ca2

0.4518

0.25

0.1150

Si1

0.0813

0.25

0.1150

Si2

0.4105

0.75

0.2824

Si3

0.1600

0.25

0.4357

Al1

0.2497

0.9970

0.1897

Al2

0.1055

0.75

0.3004

O1

0.1307

-0.0006

0.1453

O2

0.1011

0.0137

0.4309

O3

0.3587

0.9897

0.2450

O4

0.2193

0.75

0.3004

O5

0.2276

0.25

0.3119

O6

0.2718

0.75

0.0600

O7

0.9916

0.25

0.1639

O8

0.9960

0.75

0.2952

O9

0.4211

0.75

0.4431

O10

0.2682

0.25

0.0754

H

0.263

0.25

0.976

125

Raman mode analysis:

Raman Active Modes Atom Al1, O1, O2, O3 Ca1, Ca2, Si1, Si2, Si3, Al2, O4, O5, O6, O7, O8, O9, O10, H

Wyckoff Position

Point Symmetry

8d

1

3

3

3

3

4c

m

2

1

2

1

Ag B1g B2g B3g

(4 x 8d) + (14 x 4c) Raman mode analysis predicts the existence of 132 active Raman modes in zoisite: 40Ag + 26B1g + 40B2g + 26B3g = 132

126

Spectral Features Related to OHThe crystal structure of zoisite, a member of the epidote group, is composed of chains of edgesharing AlO6 groups bound by SiO4 groups. Hydroxl frequently creates a weak bridging bond oriented parallel to the c-axis between the AlO6 octahedra (Fig. 15.1). Changes in the energy of this OH stretching mode can reveal valuable information about the crystal structure (Winkler et al., 1989). This bridging OH produces a peak in both IR and Raman spectra centered at 31603170 cm-1 (Fig. 15.2) (Winkler et al., 1989; Langer and Lattard, 1980; Winkler et al., 2008). With an increase in pressure, this peak shifts to lower frequencies at a rate of approximately 34 cm-1 per GPa due to shortening of the hydrogen bond (Winkler et al., 2008). The frequency shift per GPa is much larger in zoisite than in the structurally similar mineral, clinozoisite, because the bridging hydrogen bond in zoisite is considerably straighter and shorter than the hydrogen bond in clinozoisite. The OH bond in clinozoisite is bent at an angle of greater than 30o, thus, reducing its compressibility with pressure (Winkler et al., 2008). Winkler et al. also observed two IR bands at 2330 and 2900 cm-1 and suggest that these bands are associated with CO2 and C-H stretching of organic material, respectively (Winkler et al., 1989). Neither of these peaks were observed in the Raman spectra of zoisite in this study.

Fig. 15.1 Crystal structure of zoisite viewed down b-axis; note location of hydrogen (red spheres) bonds bridging AlO6 groups (green octahedra); orange tetrehedra: SiO4 groups, purple spheres: Ca cations

Intensity

Fig. 15.2 Peak centered at 3166 cm-1 in the Raman spectra of zoisite associated with bridging OH- oriented parallel to the c-axis; sample R050038, λexcitation = 532 nm, unoriented

OH-

Raman Shift (cm-1)

127

Spectral Features Related to Chromophores and Other Ions Although it has been established that the ion responsible for the distinct blue-violet color of tanzanite is vandium, the valence and the specific nature of the color center remain controversial (Franz and Liebscher, 2004). Fritsch and Rossman (1988) attribute the color to a combination of octahedrally coordinated V4+ and V3+. Green color in zoisite is produced by the presence of Cr3+ and pink color is due to Mn3+ (Fritsch and Rossman, 1988). Luminescence data collected by Gaft et al. (2005) and Koziarska et al. (1994) both attribute peaks located at 692 and 710 nm to vanadium centers, however, the valence is still in dispute (Fig. 15.3 A. & B.). Koziarska et al. (1994) attribute the features to V3+, while Gaft et al. (2005) attribute them to V2+. Zoisite can incorporate a variety of trace and rare-earth elements into its structure including, but not limited to, Cr, Sr, Eu, Tb, Dy, Nd, Y and Hf (Frei et al., 2004). Luminescence features in the Raman spectra of a tanzanite from Umba Valley, Tanzania, located at 872 nm, 880 nm, 890 nm, and 897 nm are likely related to the presence of Nd3+ (Fig. 15.4 A. & B.) (Gaft et al., 2005). Laser ablation ICP-MS data for this sample is provided below (Table 15.1, Breeding, 2007). Peaks located centered at 586, 587, 589, 591, 611, and 617 nm are likely produced by REE, however, no luminescence or infrared data have been published to confirm this (Fig. 15.5). Gaft et al. (2005) report the presence of peaks centered at 575, 544, 440 nm related to Dy, Tb, and Eu luminescence centers respectively, however, no equivalent peaks in the Raman spectra were observed in this study.

Fig. 15.3 A. Peaks centered at 692 nm (5324340cm-1) and 710 nm (5324695 cm-1) associated with vanadium luminescence centers in zoisite; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample R060567, variety tanzanite, Umba Valley, Tanzania, unoriented

V

128

Fig. 15.3 B. Magnification of peaks centered associated with vanadium luminescence centers in zoisite; sample R060567, variety tanzanite, Umba Valley, Tanzania, λexcitation = 532 nm, unoriented

Intensity

V

Raman Shift (cm-1)

Fig. 15.4 A. Peaks in the Raman spectrum of zoisite likely related to the presence of REE luminescence

785 -1 785 -1 785 -1 centers; peaks centered at 872 nm ( 1270cm ), 880 nm ( 1378 cm ), 890 nm ( 1504 cm ), and 897 785 -1 3+ nm ( 1602 cm ) are related to Nd ; sample R060567, tanzanite from Umba Valley, Tanzania, unoriented

Nd3+

Fig. 15.4 B. Magnification of peaks in the Raman spectrum of zoisite likely related to Nd3+ luminescence

Intensity

centers; sample R060567, tanzanite from Umba Valley, Tanzania, λexcitation = 785 nm, unoriented

Nd3+ Nd3+ Raman Shift (cm-1)

129

Table 15.1 Laser ablation ICP-MS data from zoisite, sample R060567, Umba Valley, Tanzania; numerical values represent concentrations in parts per million (ppm) of each element taken at three different spots on the sample; (Breeding, 2007) 24Mg 240.3 220.2 225.7

27Al 19.5 20.7 20.1

29Si 18.0 18.4 18.2

31P 77.0 87.5 58.2

43Ca 18.1 18.0 17.9

45Sc 10.5 12.7 11.4

48Ti 424.5 430.5 423.7

51V 2209.0 2675.0 2499.0

52Cr

55Mn 17.5 18.9 18.4

69Ga 163.8 185.0 177.9

72Ge 16.5 14.7 16.5

88Sr 1355.0 1340.0 1375.0

89Y

90Zr 61.7 94.5 79.0

0.9 2.0 1.5

137Ba 2.8 2.9 3.2

139La 8.7 15.5 11.7

140Ce 18.3 33.3 25.1

141Pr 2.5 4.6 3.5

146Nd 13.0 21.8 16.8

147Sm 4.7 8.0 6.3

153Eu 1.6 2.7 2.3

157Gd 5.5 9.2 7.3

159Tb 1.2 1.8 1.5

163Dy 8.4 13.4 10.8

165Ho 1.8 2.8 2.3

166Er

169Tm 0.5 0.8 0.7

172Yb 3.6 5.3 4.7

175Lu 0.4 0.7 0.6

208Pb 0.6 0.2 0.2

232Th 1.6 2.4 2.0

238U

137.1 166.3 158.0

4.7 7.4 6.2

6.8 13.1 10.9

Fig. 15.5 A. Peaks of unknown origins centered at 586 nm (5321748 cm-1), 587 nm (5321788 cm-1), 589 nm

(5321830 cm-1), 591 nm (5321887 cm-1), 611 nm (5322440 cm-1), and 617 nm (5322593 cm-1); peaks may be related to REE luminescence centers; note that the luminescence peaks overwhelm the Raman peaks in this spectrum; sample X060001, variety tanzanite, unoriented

V

?

130

Fig. 15.5 B. Magnification of peaks of unknown origins; sample X060001, variety tanzanite, λexcitation = 532

Intensity

nm, unoriented

Raman Shift (cm-1)

References Breeding, M. (2007) Personal communication. Comodi, P. & Zanazzi, P.F. (1997) The pressure behavior of clinozoisite and zoisite: An X-ray diffraction study. American Mineralogist, Vol. 82, No. 1-2, pp. 61-68. Franz, G. & Liebscher, A. (2004) Physical and chemical properties of the epidote minerals-an introduction. Reviews in Mineralogy and Geochemistry, Vol. 56, No. 1, pp. 1-81. Frei, D., Liebscher, A., Franz, G. & Dulski, P. (2004) Trace element geochemistry of epidote minerals. Reviews in Mineralogy and Geochemistry, Vol. 56, No. 1, pp. 553-605. Fritsch, E. & Rossman, G. (1988) An update on color in gems. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, Vol. 24, No. 2, pp. 81-102. Gaft, M., Reisfeld, R. & Panczer, G. (2005) Modern Luminescence Spectroscopy of Minerals and Materials, Springer, Heidelberg, Germany. Koziarska, B., Godlewski, M., Suchocki, A., Czaja, M. & Mazurak, Z. (1994) Optical properties of zoisite. Physical Review-Section B-Condensed Matter, Vol. 50, No. 17, pp. 12297-12300. Langer, K. & Lattard, D. (1980) Identification of a low-energy OH-valence vibration in zoisite. American Mineralogist, Vol. 65, No. 7-8, pp. 779-783. Winkler, B., Gale, J.D., Refson, K., Wilson, D.J. & Milman, V. (2008) The influence of pressure on the structure and dynamics of hydrogen bonds in zoisite and clinozoisite. Physics and Chemistry of Minerals, Vol. 35, No. 1, pp. 25-35. Winkler, B., Langer, K. & Johannsen, P.G. (1989) The influence of pressure on the OH valence vibration of zoisite. Physics and Chemistry of Minerals, Vol. 16, No. 7, pp. 668-671.

131

Appendix A Features in Raman Spectra of Gem Minerals* Sorted by increasing Raman shift * garnets are excluded see charts in garnet section for mode locations and assignments laser (nm)

Raman shift (cm-1)

nm

ion/molecule

mineral

reference

780

860

Nd3+

elbaite & liddicoatite

Gaft et al., 2003

785

872

Nd3+

titanite

Gaft et al., 2005

785

872

Nd3+

zoisite

Gaft et al., 2005

780

869

Nd3+

elbaite & liddicoatite

Gaft et al., 2003

780

870

Nd3+

zircon

Gaft et al., 2005

785

880

Nd3+

titanite

Gaft et al., 2005

785

880

Nd3+

zoisite

Gaft et al., 2005

780

878

Nd3+

elbaite & liddicoatite

Gaft et al., 2003

780

879

Nd3+

zircon

Gaft et al., 2005

785

890

Nd3+

zoisite

Gaft et al., 2005

785

897

Nd3+

titanite

Gaft et al., 2005

780

891

Nd3+

zircon

Gaft et al., 2005

785

897

Nd3+

zoisite

Gaft et al., 2005

780

896

Nd3+

elbaite & liddicoatite

Gaft et al., 2003

topaz

Beny & Piriou, 1987

diamond

McNamara et al., 1992

1196 1269 1270 1314 1327 1374 1378 1433 1450 1504 1588 1600 1602 1668

-

1165

N/A

1280-1289

N/A

OH (stretching) 13

C

532

574

Sm3+

titanite

Gaft et al., 2005

532

579

Sm3+

titanite

Gaft et al., 2005

1598

N/A

Type-I H2O

beryl

Lodzinski et al., 2005

1628/1634

N/A

Type-II H2O

beryl

Lodzinski et al., 2005

585

Sm3+

titanite

Gaft et al., 2005

diamond

Solin & Ramdas, 1970

diamond

Solin & Ramdas, 1970

1394 1526

532

1720

nd

1817

N/A

1864

N/A

2 order Raman 2nd order Raman

132

laser (nm)

Raman shift (cm-1)

nm

ion/molecule

mineral

reference

532

591

Eu3+

zircon

Gaft et al., 2005

532

596

Eu3+

zircon

Gaft et al., 2005

1910 2013

nd

2025 532

2151

N/A

2 order Raman

diamond

Solin & Ramdas, 1970

600

Sm3+

titanite

Gaft et al., 2005

diamond

Solin & Ramdas, 1970

diamond

Solin & Ramdas, 1970

zircon

Gaft et al., 2005

diamond

Solin & Ramdas, 1970

diamond

Solin & Ramdas, 1970

nd

2177

N/A

2254

N/A

532

2260

604

2 order Raman 2nd order Raman Eu3+ nd

2333

N/A

2458

N/A

532

2 order Raman 2nd order Raman

614

Eu3+

titanite

Gaft et al., 2005

N/A

2nd order Raman

diamond

Solin & Ramdas, 1970

532

614

Sm3+

zircon

Gaft et al., 2005

532

617

Sm3+

titanite

Gaft et al., 2005

diamond

Solin & Ramdas, 1970

spodumene

Gaft et al., 2005; Souza et al., 2003

2491

2519 2560 2597

nd

2667

N/A

2 order Raman Al3+-hole or Mn2+?

532

620

532

620

Sm3+

titanite

Gaft et al., 2005

532

638

NV- center

diamond

Zaitsev, 2001

N/A

bridging OH-

zoisite

Winkler et al., 1989

2670 2682 3100

3160-3170

rd

N/A

3 order Raman

diamond

Bormett et al., 1995

650

Cr3+

chrysoberyl

Gaft et al., 2005

3424

N/A

OH-

zircon

Dawson et al., 1971

3476

N/A

OH3

elbaite

Gonzales-Carreño, 1988

3584

N/A

OH1 or OH3?

elbaite

Castañeda et al., 2000; Zhang et al., 2008

3485

N/A

OH(stretching)

titanite

Hammer et al., 1996

3489

N/A

OH3

liddicoatite

Zhang et al., 2008

3496

N/A

OH3

elbaite

Gonzles-Carreño, 1988

655

Cr3+

chrysoberyl

Gaft et al., 2005

3300 532

3370

532

3540

133

laser (nm)

Raman shift (cm-1)

nm

ion/molecule

mineral

reference

3566

N/A

OH3

elbaite

Zhang et al., 2008; GonzalesCarreño, 1988

3592

N/A

OH1

elbaite

Castañeda et al., 2000

3594/3597

N/A

Type-II H2O (stretching)

beryl

Lodzinski et al., 2005

3595

N/A

OH3

liddicoatite

Zhang et al., 2008

3598

N/A

OH3

elbaite

Zhang et al., 2008

3606-3609

N/A

beryl

Lodzinski et al., 2005

3650

N/A

topaz

Pinheiro et al., 2002

3651/3657

N/A

Type-II H2O

beryl

Lodzinski et al., 2005

660

Sm3+

zircon

Gaft et al., 2005

3673

N/A

OH-

andalusite

Rossman, 1996

3692/3696

N/A

Type-I H2O

beryl

Lodzinski et al., 2005

3740

N/A

OH1

dravite

Gonzales-Carreño, 1988

532

664

Cr3+

chrysoberyl

Gaft et al., 2005

532

666

Sm3+

zircon

Gaft et al., 2005

532

3654

3745 3772

Type-I H2O (stretching) OH- substituting for F

rd

3825

N/A

3 order Raman

diamond

Bormett et al., 1995

3880

N/A

Type-I H2O

beryl

Lodzinski et al., 2005

532

671

Cr3+

diopside

Gaft et al., 2005

532

676

Cr3+

spinel

Gaft et al., 2005

532

678

Cr3+

topaz

Gaft et al., 2005

532

679

Cr3+

chrysoberyl

Gaft et al., 2005

532

679

Cr3+

topaz

Gaft et al., 2005

532

679

Cr3+

diopside

Gaft et al., 2005

532

680

Cr3+

beryl

Gaft et al., 2005

532

680

Cr3+

chrysoberyl

Gaft et al., 2005

532

680

Cr3+

uvite

Gaft et al., 2005

532

683

Cr3+

uvite

Gaft et al., 2005

532

683

Cr3+

topaz

Gaft et al., 2005

3900 4000 4055 4062 4075 4082 4095 4100 4110 4160 4160

134

laser (nm)

Raman shift (cm-1)

nm

ion/molecule

mineral

reference

532

683

Cr3+

beryl

Gaft et al., 2005

532

683

Cr3+

topaz

Gaft et al., 2005

532

684

Cr3+

dravite

Gaft et al., 2005

532

684

Cr3+

diopside

Gaft et al., 2005

532

686

Cr3+

spinel

Gaft et al., 2005

532

689

Eu3+

titanite

Gaft et al., 2005

532

692

V

zoisite

Gaft et al., 2005

532

693

Cr3+

corundum

Gaft et al., 2005

532

693

Cr3+

chrysoberyl

Gaft et al., 2005

532

693

Cr3+

forsterite

Gaft et al., 2005

532

694

Cr3+

corundum

Gaft et al., 2005

532

694

Cr3+

forsterite

Gaft et al., 2005

532

698

Cr3+

spinel

Gaft et al., 2005

532

700

Cr3+

chrysoberyl

Gaft et al., 2005

532

700

Cr3+

diopside

Gaft et al., 2005

532

701

Eu3+

titanite

Gaft et al., 2005

532

702

Eu3+

zircon

Gaft et al., 2005

532

703

V2+

chrysoberyl

Gaft et al., 2005

532

706

Eu3+

titanite

Gaft et al., 2005

532

708

Cr3+

spinel

Gaft et al., 2005

532

709

V

zoisite

Gaft et al., 2005

532

710

Eu3+

zircon

Gaft et al., 2005

532

712

Cr3+

topaz

Gaft et al., 2005

532

718

Cr3+

spinel

Gaft et al., 2005

532

722

V3+

diopside

Gaft et al., 2005

532

733

Cr3+

topaz

Gaft et al., 2005

4170 4175 4185 4196 4211 4263 4340 4366 4367 4373 4396 4400 4466 4510 4526 4545 4572 4573 4651 4657 4695 4728 4762 4859 4958 5155

135

Appendix B Features in Raman Spectra of Gem Minerals* Sorted alphabetically by mineral name *garnets are excluded see charts in garnet section for mode locations and assignments

laser (nm)

mineral

Raman shift (cm-1)

nm

ion/molecule

reference

beryl

1598

N/A

Type-I H2O

Lodzinski et al., 2005

beryl

1628/1634

N/A

Type-II H2O

Lodzinski et al., 2005

beryl

3594/3597

N/A

Type-II H2O (stretching)

Lodzinski et al., 2005

beryl

3606-3609

N/A

Type-I H2O (stretching)

Lodzinski et al., 2005

beryl

3651/3657

N/A

Type-II H2O

Lodzinski et al., 2005

beryl

3692/3696

N/A

Type-I H2O

Lodzinski et al., 2005

beryl

3880

N/A

Type-I H2O

Lodzinski et al., 2005

beryl

532

680

Cr3+

Gaft et al., 2005

beryl

532

683

Cr3+

Gaft et al., 2005

chrysoberyl

532

650

Cr3+

Gaft et al., 2005

chrysoberyl

532

655

Cr3+

Gaft et al., 2005

chrysoberyl

532

664

Cr3+

Gaft et al., 2005

chrysoberyl

532

679

Cr3+

Gaft et al., 2005

chrysoberyl

532

680

Cr3+

Gaft et al., 2005

chrysoberyl

532

693

Cr3+

Gaft et al., 2005

chrysoberyl

532

700

Cr3+

Gaft et al., 2005

chrysoberyl

532

703

V2+

Gaft et al., 2005

corundum

532

693

Cr3+

Gaft et al., 2005

corundum

532

694

Cr3+

Gaft et al., 2005

4095 4170 3370 3540 3745 4062 4100 4367 4510 4573 4366 4396

136

laser (nm)

mineral

Raman shift (cm-1)

nm

diamond

1280-1289

N/A

diamond

1817

N/A

2nd order Raman

Solin & Ramdas, 1970

diamond

1864

N/A

2nd order Raman

Solin & Ramdas, 1970

diamond

2025

N/A

2nd order Raman

Solin & Ramdas, 1970

diamond

2177

N/A

2nd order Raman

Solin & Ramdas, 1970

diamond

2254

N/A

2nd order Raman

Solin & Ramdas, 1970

diamond

2333

N/A

2nd order Raman

Solin & Ramdas, 1970

diamond

2458

N/A

2nd order Raman

Solin & Ramdas, 1970

diamond

2519

N/A

2nd order Raman

Solin & Ramdas, 1970

diamond

2667

N/A

2nd order Raman

Solin & Ramdas, 1970

638

NV- center

Zaitsev, 2001

diamond

532

3100

ion/molecule 13

C

reference McNamara et al., 1992

rd

diamond

3300

N/A

3 order Raman

Bormett et al., 1995

diamond

3825

N/A

3rd order Raman

Bormett et al., 1995

671

Cr3+

Gaft et al., 2005

679

Cr

3+

Gaft et al., 2005

Cr

3+

Gaft et al., 2005

Cr

3+

Gaft et al., 2005

diopside

532

diopside

532

diopside

532

diopside

532

diopside

532

722

V

Gaft et al., 2005

forsterite

532

693

Cr3+

Gaft et al., 2005

forsterite

532

694

Cr3+

Gaft et al., 2005

spinel

532

676

Cr3+

Gaft et al., 2005

spinel

532

686

Cr3+

Gaft et al., 2005

3900 4082 4196 4526 4958 4373 4400 4000 4211

684 700

3+

137

laser (nm)

mineral

Raman shift (cm-1)

nm

ion/molecule

reference

spinel

532

698

Cr3+

Gaft et al., 2005

spinel

532

708

Cr3+

Gaft et al., 2005

spinel

532

718

Cr3+

Gaft et al., 2005

spodumene

532

620

Al3+-hole or Mn2+?

Gaft et al., 2005; Souza et al., 2003

titanite

785

872

Nd3+

Gaft et al., 2005

titanite

785

880

Nd3+

Gaft et al., 2005

titanite

785

897

Nd3+

Gaft et al., 2005

titanite

532

574

Sm3+

Gaft et al., 2005

titanite

532

578

Sm3+

Gaft et al., 2005

titanite

532

589

Sm3+

Gaft et al., 2005

titanite

532

600

Sm3+

Gaft et al., 2005

titanite

532

613

Sm3+

Gaft et al., 2005

titanite

532

617

Sm3+

Gaft et al., 2005

titanite

532

620

Sm3+

Gaft et al., 2005

N/A

OH- (stretching)

Hammer et al., 1996

titanite

4466 4657 4859

2670

1269 1374 1588 1394 1526 1720 2151 2491 2597 2682

3485

titanite

532

689

Eu3+

Gaft et al., 2005

titanite

532

703

Eu3+

Gaft et al., 2005

titanite

532

706

Eu3+

Gaft et al., 2005

N/A

OH(stretching)

Beny & Piriou, 1987

733

Cr3+

Gaft et al., 2005

N/A

OH- substituting for F

Pinheiro et al., 2002

topaz topaz topaz

4263 4545 4651

1165 532

5155

3650

138

laser (nm)

mineral

Raman shift (cm-1)

nm

ion/molecule

reference

topaz

532

678

Cr3+

Gaft et al., 2005

topaz

532

679

Cr3+

Gaft et al., 2005

topaz

532

683

Cr3+

Gaft et al., 2005

topaz

532

683

Cr3+

Gaft et al., 2005

topaz

532

712

Cr3+

Gaft et al., 2005

tourmaline

780

860

Nd3+

Gaft et al., 2003

tourmaline

780

869

Nd3+

Gaft et al., 2003

tourmaline

780

878

Nd3+

Gaft et al., 2003

tourmaline

780

896

Nd3+

Gaft et al., 2003

4055 4075 4160 4175 4762 1196 1314 1433 1668

tourmaline

3476

N/A

OH3

Gonzales-Carreño, 1988

tourmaline

3489

N/A

OH3

Zhang et al., 2008

tourmaline

3496

N/A

OH3

tourmaline

3566

N/A

OH3

tourmaline

3592

N/A

OH1

tourmaline

3595

N/A

OH3

Zhang et al., 2008

tourmaline

3598

N/A

OH3

Zhang et al., 2008

Gonzales-Carreño, 1988 Zhang et al., 2008; Gonzales-Carreño, 1988 Castañeda et al., 2000

tourmaline

532

680

Cr3+

Gaft et al., 2005

tourmaline

532

683

Cr3+

Gaft et al., 2005

tourmaline

532

684

Cr3+

Gaft et al., 2005

zircon

780

870

Nd3+

Gaft et al., 2005

zircon

780

879

Nd3+

Gaft et al., 2005

zircon

780

891

Nd3+

Gaft et al., 2005

zircon

532

591

Eu3+

Gaft et al., 2005

4110 4160 4185 1327 1450 1600 1910

139

laser (nm)

mineral

Raman shift (cm-1)

nm

ion/molecule

reference

zircon

532

596

Eu3+

Gaft et al., 2005

zircon

532

604

Eu3+

Gaft et al., 2005

zircon

532

614

Eu3+

Gaft et al., 2005

N/A

OH-

Dawson et al., 1971

zircon

2013 2260 2560

3424

zircon

532

660

Sm3+

Gaft et al., 2005

zircon

532

666

Sm3+

Gaft et al., 2005

zircon

532

702

Eu3+

Gaft et al., 2005

zircon

532

710

Eu3+

Gaft et al., 2005

zoisite

785

872

Nd3+

Gaft et al., 2005

zoisite

785

880

Nd3+

Gaft et al., 2005

zoisite

785

890

Nd3+

Gaft et al., 2005

zoisite

785

897

Nd3+

Gaft et al., 2005

N/A

bridging OH-

Winkler et al., 1989

zoisite

3654 3772 4572 4728 1270 1378 1504 1602

3160-3170

zoisite

532

692

V

Gaft et al., 2005

zoisite

532

709

V

Gaft et al., 2005

4340 4695

140

Appendix C Unit Conversions Raman spectroscopy utilizes Raman shift (cm-1) to describe the location of Raman modes. The abbreviation cm-1 can be misleading because Raman shift represents a change in wave numbers, a shift from the frequency of the laser. The abbreviation Δcm-1 would be more appropriate, however it is convention to describe Raman shift with cm-1. When comparing Raman and IR modes to luminescence features, a conversion from cm-1 to nm is required. The simple conversion is as follows: Note: these conversions are laser wavelength dependent. From nm to cm-1: 1) Convert the wavelength of the laser used for Raman spectroscopy from nm to wave numbers. Do this by converting nm into cm and then to wave numbers. 1 nm = 1.0 x 10-7 cm ex. 532 nm laser = 532 x 1.0 x 10-7 = .0000532 cm From cm to cm-1: 1/.0000532 cm = 18796.99248 cm-1

The 532 nm laser is equal to 18796.99248 cm-1 2) Convert the luminescence feature (nm) to wave numbers by same process described in step 1). ex. Luminescence feature at 693 nm = 14430.01443 cm-1

3) Since Raman shift represents the shift in energy from the initial energy of the laser beam, the final step requires that you subtract the converted luminescence feature in wave numbers from the converted laser wavelength in wave numbers. This will give you the location in the Raman spectrum where you would expect to see the luminescence feature you have observed. ex. (laser) 18796.99248 cm-1 - (feature) 14430.01443 cm-1 = 4366.978051 cm-1 (Raman shift)

To switch between Raman spectra from different laser wavelengths, simply replace the laser wavelength with the wavelength being utilized; ie. 514 nm, 780 nm, 785 nm, etc. From cm-1 to nm (reverse the process): 1) Convert your value in Raman shift to wave numbers by subtracting the Raman shift value from the wavelength of the laser. ex. Raman shift value: 1378 cm-1 Laser: 532 nm = 18796.99248 cm-1 -1

-1

18796.99248 cm – 1378 cm

-1

= 17418.99 cm (wave numbers)

2) Convert wave numbers into cm and then nm. -1 -5 1/17418.99 cm = 5.74 x 10 cm

5.74 x 10-5 cm x 107 = 574 nm

141

Wave lengths of radiation are frequently described in angstroms (Å). The simple conversion from angstroms to nm is provided here: Angstroms to nm: 1 Å = .1 nm

To convert from Å to nm simply take the wavelength in angstroms and multiply by 0.1

Example: 5448Å = 544.8 nm Wave lengths of radiation can also be described in electron volts (eV). The simple conversion from eV to nm is provided here: eV to nm: 1239.8424121/ x eV = nm Example: 1239.8424121/2.16 eV = 574 nm

142

Appendix D Cause of Color Chart (modified from Fritsch and Rossman, 1988)

Mineral

Color

beryl

emerald green green beryl green with yellow or blue undertones heliodor (golden) dark blue (Maxixe) light blue (aquamarine) dark aquamarine pink (morganite) red

chrysoberyl

alexandrite yellow

corundum

ruby red

blue sapphire purple

orange-pink (padparadscha)

orange to brown pink

Cause

Reference

octahedrally cooridnated Cr3+ octahedrally cooridnated V3+ O2- Fe3+ charge transfer, Fe2+ filling channels O2- Fe3+ charge transfer color centers due to irradiation

Fritsch and Rossman, 1988; Gaft et al., 2005 Fritsch and Rossman, 1988; Gaft et al., 2005

channel-filling Fe2+ Fe2+O Fe3+ intervalence charge transfer octahedrally coordinated Mn2+ octahedrally coordinated Mn3+ octahedrally coordinated Cr3+ octahedrally coordinated Fe3+ octahedrally coordinated Cr3+ with some V3+ and Fe3+ in octahedral coordination Fe2+ O Ti4+charge transfer Fe2+ O Ti4+ charge transfer; with octahedrally coordinated Cr3+ still debated; presence of Cr possible with other color centers, valence is debated octahedrally coordinated Cr3+with some Fe3+ octahedrally coordinated Cr3+

Fritsch and Rossman, 1988; Gaft et al., 2005 Fritsch and Rossman, 1988; Gaft et al., 2005 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988; Gaft et al., 2005 Fritsch and Rossman, 1988; Gaft et al., 2005 Fritsch and Rossman, 1988; Gaft et al., 2005 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988; Gubelin, 1976; Hassan and El-Rakhawy, 1974 Fritsch and Rossman, 1988

Fritsch and Rossman, 1988

Fritsch and Rossman, 1988 Fritsch and Rossman, 1988

Fritsch and Rossman, 1988

Fritsch and Rossman, 1988 Fritsch and Rossman, 1988

143

Mineral corundum

Color yellow

green diamond

yellow, orange, brown, colorless, near colorless brown, pink, colorless, near colorless blue, gray, light brown, near colorless

multiple possibilities relating to Fe (possible charge transfer or Fe-pairs) several possibilities involving either Fe and/or Cr3+ Type I diamonds; contain nitrogen aggregates

Reference Fritsch and Rossman, 1988

Fritsch and Rossman, 1988 Deljanin and Simic, 2007

Type IIa diamonds; nitrogen absent

Deljanin and Simic, 2007

Type IIb diamonds; nitrogen absent, contain boron)

Deljanin and Simic, 2007

pink/red

deformation of lattice

Fritsch et al., 2007; Shigley, 1993

purple

deformation of lattice

Titkov et al., 2008

brown black chameleon (color change green to yellow) diopside

Cause

green (chrome diopside) yellowish green purple (rare)

nitrogen impurities and lattice deformation dark mineral microinclusions electron trap due to H and N interactions octahedrally coordinated Cr3+, V3+, or a combination charge transfer Fe2+ Fe3+ charge transfer Fe2+ and Ti4+

Massi, 2005 Titkov et al., 2003 Fritsch et al., 2007 Fritsch and Rossman, 1988; Andrut et al., 2003). Fritsch and Rossman, 1988; Andrut et al., 2003). Herd et al., 2000

garnet almandine

red (rhodolite)

andradite

green (demantoid) yellow green

grossular

pyrope

green (tsavorite)

Fe2+ in distorted cubic coordination octahedrally coordinated Cr3+ octahedrally coordinated Fe3+ octahedrally coordinated V3+

Fritsch and Rossman, 1988; Manning, 1967 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988; Manning, 1970 Fritsch and Rossman, 1988; Manning, 1970

orange (hessonite)

Mn2+ or Fe2+

red

octahedrally coordinated Cr3+and some Fe2+

Fritsch and Rossman, 1988; Manning, 1967

brownish-red

Fe2+

Fritsch and Rossman, 1988; Manning, 1967

144

Mineral

Color

Cause

Reference

spesartine

orange

Mn2+

uvarovite

green

colorchange

green to blue, red to purple, green to red

Gubelin, 1982; Manning, 1967 Fritsch and Rossman, 1988; Manning, 1970 Gubelin, 1982; Schmetzer and Bernhardt, 1999; Carstens, 1973

olivine

green, yellowgreen (peridot)

quartz

purple (amethyst) yellow (citrine)

spinel

green (praseolite)

Fe2+

black (smoky)

hole center

pink (rosy)

pink microinclusions

pink/red purple

cobalt blue blue green spodumene

pink to purple (kunzite) green (hiddenite) pale green

titanite

octahedrally coordinated Cr3+ octahedrally coordinated Cr3+and V3+ octahedrally coordinated Fe2+with minor Cr3+ hole center created by presence of Fe and radiation O2- Fe3+ charge transfer

octahedrally coordinated Cr3+ octahedrally coordinated Cr3+with tetrahedrally coordinated Fe2+ Co and Fe in tetrahedral coordination tetrahedrally coordinated Fe2+and Fe3+ Mn3+ octahedrally coordinated Cr3+ and V3+, with some Mn possible Fe charge transfer and some Mn

Fritsch and Rossman, 1988 Paradise, 1982; Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Paradise, 1982; Fritsch and Rossman, 1988). Maschmeyer et al., 1980; Fritsch and Rossman, 1988 Goreva et al., 2001 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988

Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988

green

Fe

Fritsch and Rossman, 1988

chrome sphene (green)

octahedrally coordinated Cr3+

Fritsch and Rossman, 1988

pink

Mn

Fritsch and Rossman, 1988

yellow

Fe

Fritsch and Rossman, 1988

brown

Fe

Fritsch and Rossman, 1988

pink

octahedrally coordinated Cr3+

Taran et al., 2003; Fritsch and Rossman, 1988)

145

Mineral

Color

Cause

Reference

3+

topaz

yellow orange-red red-brown

Cr and a radiation induced O- hole Cr3+combined with a F-center O- and F center combinations

Gaft et al., 2005 Gaft et al., 2005 Gaft et al., 2005

blue

R-centers

Gaft et al., 2005

purple

Cr3+

Taran et al., 2003

red-orange (imperial)

Cr4+

Taran et al., 2003

pink-orange

Cr3+ and Cr4+

Taran et al., 2003

tourmaline dravite

green yellow to brown red

elbaite and liddicoatite

blue (indicolite) green brown yellow-green pink to red (rubellite) neon blue/green (Paraiba)

uvite

green

zircon

red

zoisite

octahedrally coordinated V3+ and Cr3+ Fe and Ti charge transfer Fe Fe and charge transfer Fe and Ti charge transfer Fe and Ti charge transfer Mn and Ti charge transfer Mn and possible irradiation center Cu2+ octahedrally coordinated Cr3+ and V3+ 4+ Nb and irradiation center

Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Fritsch and Rossman, 1988 Rossman et al., 1991 Fritsch and Rossman, 1988 Fielding, 1970; Fritsch and Rossman, 1988 Mackey et al., 1975; Fritsch and Rossman, 1988

blue

presence of U4+

blue-violet (tanzanite)

V4+ and V3+

Fritsch and Rossman, 1988

green

Cr3+

Fritsch and Rossman, 1988

pink

Mn3+

Fritsch and Rossman, 1988

146

Appendix E Point Group Symmetry Notations There are two notations used to describe point symmetry in crystals and molecules: 1) Hermann-Manguin notation (preferred by crystallographers) and 2) Schönflies notation (preferred by spectroscopists and chemists). A table correlating crystallographic point groups with the corresponding Schönflies notation is provided below (Boisen and Gibbs, 1990). HM 1

S C1

2

C2

3 4 6 222

HM 1 2/m

S Ci

HM m

S Cs

C2h

4

S4

C3 C4 C6 D2

3 4/m 6/m mmm

C3i C4h C6h D2h

6 mm2 3m

C3h C2v C3v D2d

32

D3

3m

D3d

422 622 23

D4 D6 T

4/mmm 6/mmm m3

D4h D6h Th

432

O

m3m

0h

235

I

m3 5

Ih

4 2m 4mm

6 2m 6mm 4 3m

C4v D3h C6v Td

A brief description of Schönflies notation is provided below (Gaft et al., 2005; Ferraro et al., 2003). Schönflies notation: C: (cyclic) one rotational axis; C1: identity, Cn: has an n-fold axis D: (dihedral) orthogonal axes (perpendicular); Dn: n-fold axis of symmetry perpendicular to n 2-fold axes Cubic: T (tetrahedral) 4 axes, O (octahedral) 8 axes, I (icosahedra, describes molecules, not crystals) 20 axes i: inversion center E/I: identity Planar Symmetry: -one subscript letter follows the rotation symmetry; if letter is h (horizontal): mirror plane parallel to the rotation axis; v (vertical): mirror plane perpendicular to rotation axis; d (diagonal): diagonal mirror ex. C3h: has one 3-fold rotational axis with a mirror parallel to the 3-fold axis -Sn: S is mirror symmetry and invariance (remains unchanged) by a n-fold rotation followed by reflection in plane perpendicular to axis

147