Sulfur electronic environments in α-NiS and β-NiS: examination of the relationship between coordination number and core electron binding energies

X-ray photoelectron spectra have been obtained under the same experimental conditions for synthetic α-NiS and natural β-NiS in order to establish any difference in S electronic environment, and to test the proposition that S core electron binding energies increase measurably with coordination number when the same metal is in different sulfide structures or lattice sites. The Ni and S electronic environments in the two NiS structures have been further probed by near-edge X-ray absorption fine structure (NEXAFS) spectroscopy, and the NEXAFS spectra interpreted by reference to spectra simulated by ab initio calculations. The photoelectron and NEXAFS spectra for freshly prepared surfaces of α-NiS and β-NiS were found to be similar, with only subtle differences in electronic environment evident in the experimental and simulated NEXAFS spectra. The measured and calculated core electron binding energies did not support the previously postulated relationship between S coordination number and electron binding energies.     

The oxidation states of copper and iron in mineral sulfides, and the oxides formed on initial exposure of chalcopyrite and bornite to air

Metal L_2,3, sulfur K and oxygen K near-edge X-ray absorption fine structure (NEXAFS) spectra for chalcopyrite, bornite, chalcocite, covellite, pyrrhotite and pyrite have been determined from single-piece natural mineral specimens in order to assess claims that chalcopyrite should be regarded as Cu^IIFe^IIS₂ rather than Cu^IFe^IIIS₂, and that copper oxide species are the principal initial oxidation products on chalcopyrite and bornite exposed to air. Spectra were obtained using both fluorescence and electron yields to obtain information representative of the bulk as well as the surface. Where appropriate, NEXAFS spectra have been interpreted by comparison with the densities of unfilled states and simulated spectra derived from ab initio calculations using primarily the FEFF8 code and to a lesser extent WIEN2k. Metal 2p and S 2p photoelectron spectra excited by monochromatised Al K_α X-rays were determined for each of the surfaces characterised by NEXAFS spectroscopy. The X-ray excited Cu LMM Auger spectrum was also determined for each copper-containing sulfide. FEFF8 calculations were able to simulate the experimental NEXAFS spectra quite well in most cases. For covellite and chalcocite, it was found that FEFF8 did not provide a good simulation of the Cu L₃-edge spectra, but WIEN2k simulations were in close agreement with the experimental spectra. Largely on the basis of these simulations, it was concluded that there was no convincing evidence for chalcopyrite to be represented as Cu^IIFe^IIS₂, and no strong argument for some of the Cu in either bornite or covellite to be regarded as Cu(II). The ab initio calculations for chalcopyrite and bornite indicated that the density of Cu d-states immediately above the Fermi level was sufficient to account for the Cu L₃-edge absorption spectrum, however these incompletely filled Cu d-states should not be interpreted as indicating some Cu(II) in the sulfide structure. It was also concluded that the X-ray absorption spectra were quite consistent with the initial oxidation products on chalcopyrite and bornite surfaces being iron oxide species, and inconsistent with the concomitant formation of copper-oxygen species.     

Interpretation of Ni2p XPS spectra of Ni conductors and Ni insulators

Ni2p_3/2 X-ray photoelectron spectral peak binding energies of Ni metal, NiS, and NiAs (all conductors) span a range of about 0.5 eV and are, consequently, insensitive to formal Ni oxidation state and to the nature of the ligand to which Ni is bonded, relative to other metals (e.g., Fe). Ni2p_3/2 peak structures and binding energies reflect two energetic contributions. The major contribution is that associated with the electrostatic field produced by ejection of the Ni(2p) photoelectron, the minor contribution is the relaxation energy associated with filling unoccupied, conduction band 3d⁹ and 4s Ni metal orbitals. These conduction band orbitals become localized on the Ni photoion (and sometimes filled) in response to the field created by the photoemission event. Because only the core Ni2p electron and nonbonding orbitals of predominantly metallic character are affected, the main peak of all three conductors are affected similarly, leading to similar Ni2p_3/2 main peak binding energies. NiO, Ni(OH)₂, and NiSO₄ are insulators in which Ni is divalent and is bonded to oxygen. Although Ni is bonded to oxide in these phases, Ni2p binding energies differ substantially, and reflect primarily the nature of the ligand (O²⁻, OH⁻, SO₄ ²⁻) to which Ni is bonded. The influence of the ligand is the result of charge (electron) transfer from valence band bonding orbitals of dominantly ligand character, to unoccupied conduction band orbitals localized on Ni photoions. Relaxation energy resulting from charge transfer is acquired by the emitted photoelectron, thus Ni2p_3/2 photopeak binding energies of these insulators reflect the nature of the ligand to which Ni is bonded. The Ni2p main peak binding energy of these conductors and insulators is a poor guide to Ni oxidation states. The Ni2p_3/2 binding energies of insulators reflect, however, the nature of the ligand in the first coordination sphere of Ni. The intensity of the Doniach–Sunjic contribution to Ni2p XPS spectra of NiS and NiAs is dependent on the nature of the ligand. The Doniach–Sunjic contribution to ligand XPS core-level photopeaks (e.g., S2p of NiS and As3d of NiAs) has not been explained and is poorly understood. Received 24 May 1999 / Revised, accepted: 30 June 1999     

Crystal orbital contributions to the pyrrhotite valence band with XPS evidence for weak Fe–Fe π bond formation

 Synchrotron excited X-ray photoelectron spectra (SXPS) of hexagonal pyrrhotite reveal three distinct Fe 3d-derived photopeaks within its outer valence band. The t _2gα band (majority spin) is centered at about 2.5 eV, the e _g α band at about 1.0 eV and the t _2gβ (minority spin) contribution at about 0.25 eV. From these data the ligand field splitting energy is 1.5 (±0.2) eV and the majority spin pairing energy is 2.25 (±0.2) eV. These are the first such XPS measurements for this mineral. S 3p-derived bonding and non-bonding bands are identified, with the former centred at about 6.5 eV and the latter near 4.5 eV. The XPS results are remarkably consistent with SCF-Xα scattered wave molecular orbital calculations. Although the calculations and the collected spectra are consistent, they differ from a recent interpretation of the pyrrhotite valence band. An explanation for the discrepant results is provided. Auger resonant enhancement of Fe 3d photopeaks at 60 eV photon energy results in the t _2gα emission (at 2.5 eV) being strongly enhanced and broader than the t _2gβ emission (0.25 eV). The explanation of these observations requires the presence of weak Fe–Fe π and π^* crystal (molecular) orbitals located near 2.5 eV, and separated by no more than about 0.5 eV. The π-bonded crystal orbitals are derived from weak mixing of adjacent Fe t _2g atomic orbitals along the c crystallographic axis. Received: 15 June 2000 / Accepted: 11 June 2001     

An X-ray photoelectron and absorption spectroscopic investigation of the electronic structure of cubanite, CuFe2S3

X-ray photoelectron and absorption spectra have been obtained for natural specimens of cubanite and compared with the corresponding spectra for chalcopyrite. Synchrotron X-ray photoelectron spectra of surfaces prepared by fracture under ultra-high vacuum revealed some clear differences for the two minerals, most notably those reflecting their different structures. In particular, the concentration of the low binding energy S species formed at cubanite fracture surfaces was approximately double that produced at chalcopyrite surfaces. However, the core electron binding energies for the two S environments in cubanite were not significantly different, and were similar to the corresponding values for the single environment in chalcopyrite. High binding energy features in the S 2p and Cu 2p spectra were not related to surface species produced either by the fracture or by oxidation, and most probably arose from energy loss due to inter-band excitation. Differences relating to the Fe electronic environments were detectable, but were smaller than expected from some of the observed physical properties and Mössbauer spectroscopic parameters for the two minerals. X-ray absorption and photoelectron spectra together with the calculated densities of states for cubanite confirmed an oxidation state of Cu^I in the mineral. It was concluded that the best formal oxidation state representation for cubanite is Cu^I(Fe₂)^VS ₃ ^?II .     

The bonding of ferrous iron to sulfur and oxygen in tetrahedral coordination: A comparative study using SCF Xα scattered wave molecular orbital calculations

Molecular quantum mechanical calculations have been performed on high-spin ferrous iron tetrahedrally coordinated to sulfur and oxygen, respectively. The molecular orbital energies obtained from the calculations are compared with experimental optical and X-ray emission spectra. Good agreement was found between calculated and experimental spectral transition energies for the optical absorption spectra of Fe²⁺ in sphalerite, of Fe²⁺ in FeAl₂O₄, staurolite and (Zn, Fe)O, and for the FeKβ X-ray emission spectra of FeCr₂O₄. This both clarified interpretation of the spectra and established the validity of the calculations. Distinct differences occur in the molecular orbital structures of the sulfide and oxide clusters. In the sulfide, the crystal field type (mainly Fe 3d) molecular orbitals lie within the nonbonding (mainly S 3p) orbitais in energy, whereas in the oxide, they lie well above the 02p nonbonding orbitals. This also results in a wider valence band in the oxide than in the sulfide. The crystal field type (Fe 3d) molecular orbitais have more ligand character in the sulfide than the oxide and the chalcophilic properties of iron are partly attributed to this observation.     

Electronic and optical properties of Fe, Zn and Pb sulfides

Ab initio quantum-chemical calculations of the spatial and electronic structures of sphalerite (ZnS), pyrite (FeS₂) and galena (PbS), using the density functional theory (DFT) local density approximation (LDA) and generalized gradient approximation (GGA), the Hartree–Fock (HF) method and the hybrid functional B3LYP, have been carried out. For galena, the DFT LDA and GGA functionals provided the best estimate of the band gap, from within –0.1 eV to +0.4 eV of the measured value. B3LYP and RHF gave rise to errors of +1.3 and +5.4 eV, respectively. The unit cell parameter error varied from between –1.1% and +2.3% for all the functionals examined. For sphalerite the B3LYP functional provided the best estimate of the band gap (error +0.3 eV). The unit cell parameter error varied between –2.1% and +2.0% for the various DFT functionals and B3LYP. RHF gave rise to an error of +3.8%. For FeS₂, the DFT-GGA approach provides the best results for both the unit cell and the band gap. This may be due to mutual cancellation of the crystal field splitting and band separation force, which are of equal but opposite magnitudes. The calculated density of states (DOS) for the conduction band is used to interpret the experimental features of the S 1s XANES (X-ray absorption near-edge structure) spectra obtained using synchrotron radiation. Because of the l = ±1 selection rule for electron excitation, the S K-edge XANES spectra represent a transition of the S 1s electron to conduction band S p-like orbitals. The near-edge region, up to 15 eV past the edge is approximated well by the DOS. Individual peaks in the DOS correlate with peaks in the XANES spectra. In addition, the imaginary part of the dielectric function, which reflects the transitions from occupied to unoccupied levels, is used to model the near-edge region of the XANES, using the DFT-GGA formalism. Individual peaks in the XANES spectrum are moderately well resolved using the dielectric function, especially for ZnS and FeS₂, while the DOS for the conduction band is more successful in predicting the shape of the XANES spectra for all three minerals.     

C 1s K-edge near edge X-ray absorption fine structure (NEXAFS) spectroscopy for characterizing functional group chemistry of black carbon

Black carbon (BC) is considered ubiquitous in soil organic matter (OM) and therefore plays an important role in soil biogeochemistry. Its complexity, particularly within environmental matrices, presents a challenge for research, primarily as a result of techniques which may favor detection of certain functional group types rather than capturing total sample C. The objective of this study was to utilize carbon (C) 1s near edge X-ray absorption fine edge structure (NEXAFS) spectroscopy to characterize the C chemistry of a broad range of BC materials. Characteristic resonances in the NEXAFS spectra allowed direct molecular speciation of the total C chemistry of the reference materials, environmental matrices and potentially interfering materials, obtained from an earlier BC ring trial. Spectral deconvolution was used to further identify the functional group distribution of the materials. BC reference materials and soils were characterized by a large aromatic C region comprising around 40% of total absorption intensity. We were able to distinguish shale and melanoidin from BC reference materials on the basis of their unique spectral characteristics. However, bituminous coal shared chemical characteristics with BC reference materials, namely high aromaticity of more than 40% identified by way of a broad peak. Lignite also shared similar spectra and functional group distributions to BC reference materials and bituminous coal. We compared the results of spectral deconvolution with the functional group distributions obtained by way of direct polarization magic angle spinning (DPMAS) ¹³C nuclear magnetic resonance (NMR) spectroscopy. Correlations between aromatic type C values for DPMAS ¹³C NMR and NEXAFS gave r² = 0.633 (p < 0.05) and the values for NEXAFS were around 30–40% lower than for ¹³C NMR. Correlations were also drawn between the aromatic C/O-alkyl C ratio values for the two methods (r² = 0.49, p < 0.05). Overall, NEXAFS was applicable for a wide range of environmental materials, such as those measured, although some limitations for the technique were addressed.     

Polarized X-ray absorption spectroscopy of metal ions in minerals

We have made use of the nearly complete linear polarization of synchrotron radiation to study the polarization dependence of X-ray absorption near-edge structure (XANES) and extended fine structure (EXAFS) in oriented single crystals of gillespite (BaFe²⁺ Si₄O₁₀; Fe^{2 +} in square-planar coordination, point symmetry C ₄), anatase (TiO₂; Ti⁴⁺ in octahedral coordination, point symmetry D _2d), and epidote (Ca₂(Al, Fe³⁺)₃SiO₄)₃(OH); Fe³⁺ in distorted octahedral coordination, point symmetry (C _s). For gillespite, the Fe K-XANES spectrum varies strongly with E-vector orientation of the incident X-ray beam. When the E-vector lies in the plane of the FeO₄ group (i.e., perpendicular to the c-axis), multiple-scattering features at 7127 and 7131 eV intensify, whereas when the E-vector is perpendicular to the plane of the FeO₄ group (i.e., parallel to the c-axis), a strongly-polarized 1s to 4p bound state transition occurs at 7116 eV and a localized continuum resonance occurs at 7122 eV. The Fe-K-EXAFS spectrum of gillespite is also highly polarization dependent. When the E-vector is perpendicular to c, all four nearest-neighbor oxygens around Fe²⁺ contribute to the EXAFS signal; when E is parallel to c, the EXAFS signal from nearest-neighbors is reduced by at least 86%. The unpolarized Ti K-XANES spectrum of anatase has three relatively strong pre-edge features at 4967.1, 4969.9, and 4972.7 eV which have resisted definitive interpretation in past studies. The lowest energy feature has a strong xy polarization dependence, suggesting a large amount of 4p _x,y character, and it is also very sharp, indicating a well-defined transition energy. Both of these observations are consistent with an excitonic state with a binding energy of 2.8 eV. The two higher energy features, which are characteristic of octahedrally-coordinated Ti⁴⁺, show little polarization dependence and are probably due to 1s to 3d bound-state transitions, with a small degree of np character in the final state wavefunction. Interpretation of the polarization dependence of Fe K-XANES spectra for epidote is not as straightforward due to the lower space group symmetry (P2₁/m) relative to gillespite (P4/ncc) and anatase (I4₁/amd) and the lower point group symmetry (C _s) of the M(3) site which contains most of the Fe³⁺ in the epidote structure. However, the presence of a shoulder at 7121 eV in the E parallel to b spectrum and its absence in the E normal to bc spectrum are consistent with it being a 1s to 4p _z bound-state transition. Strong, weakly x, y polarized features near 7126 eV in both spectra are most likely due to localized continuum transitions. Also, the 1s to 3d pre-edge intensity varies in intensity with E-vector orientation which is consistent with displacement of Fe³⁺ from the center of the M(3) octahedral site. Analysis of EXAFS spectra of epidote in these two polarizations yields bond distances which are within 0.04 Å of previous single-crystal X-ray diffraction analysis. This study demonstrates the utility of polarized X-ray absorption spectroscopy in quantifying the energies and orbital compositions of final state wavefunctions associated with various X-ray induced transitions in transition-metal containing minerals. It also shows that reasonably accurate M-O distances can be obtained for individual bonds oriented in crystallographically non-equivalent directions.     

Ab initio calculations on (OH)4 defects in α-quartz

Ab initio total energy calculations based on a new optimized oxygen psuedopotential have been used to study the structures and relative energies of α-quartz, a partly (OH)₄ substituted version of the α-quartz structure, and interstitial water molecules in α-quartz. Hydrogen bonds formed from two hydroxyl groups of the (OH)₄ defects in the substituted α-quartz structure promote a stable structure for the defect α-quartz at low temperature. Comparable ab initio calculation of the energy of the interstitial water molecule in the quartz structure indicates that, energetically, the (OH)₄ defect is likely to be strongly favoured as a mode for the incorporation of water. Ab initio stress calculations confirm that the (OH)₄ defect in quartz has a large associated stress field which is likely to lead to segregation of these defects on supersaturation in wet quartz. The calculations indicate that segregation should occur in the plane (10 0) of the α-quartz structure.     

Calculation of the UV Absorption Spectra of As(III) Oxo- and Thio- Acids and Anions in Aqueous Solution and of PF3 in the Gas-phase

Changes in the UV spectra of As(OH)₃ solutions with variations in pH and temperature have recently been used to determine the temperature dependence of the pK_a of the acid. In previous studies I used quantum mechanical techniques to study changes in structure and vibrational spectra as a function of pH for arsenites and thioarsenites. I previously calculated UV spectra for ``molecular' minerals, like realgar As₄S₄. Here I use a number of different quantum mechanical methods, both Hartree-Fock and density functional theory based, to calculate the UV spectra for both a related simple well-characterized gas-phase molecule PF₃ and for As(OH)₃ and As(SH)₃ and their conjugate anions and some neutral and anionic oligomers in aqueous solution. For the monomeric species small numbers of water molecules have been explicitly included, in a supermolecule or microsolvation approach. I find that UV absorption energies accurate to a few tenths of an eV can be obtained both for gas- phase PF₃ and for neutral arsenious acid in aqueous solution, for which the UV absorption maximum is calculated to occur around 6.5 eV, consistent with experiment. Accurate calculation of the UV energies for arsenite anions in aqueous solution is much more difficult, since basis set size and solvation effects are considerably larger than for the neutral molecules, but fairly reliable results can still be obtained. Deprotonation is found to reduce the lowest calculated UV transition energy by about half an eV. Oligomerization also reduces the lowest calculated UV energy by at least half an eV. Replacement of one or all the –OH groups by –SH groups reduces the lowest calculated UV energies by about 2 eV. UV excitation energies have been calculated for oligomeric species as large as As₃E₃(EH)₃ and As₄E₆, where E = O, S, and may be useful for identifying such species in solution.     

Synchrotron-excited luminescence of natural zircon

The luminescence properties of two single zircon crystals from kimberlite of Yakutia have been studied, excited by the DORIS HASYLAB synchrotron, Germany, within energy range from the visible to the soft X-ray region (5–25, 50–200, and 500–620 eV) at temperatures of 300 and 10 K. The luminescence spectra in the range of 2.5 to 6.0 eV and excitation spectra of the main bands have been examined, the physical nature of the luminescence centers has been discussed, and the luminescence properties of a crystal containing growth (radiation) structural defects and a crystal with the same impurities but annealed in air at 1200°C are compared. The zoned structure of the mineral has been considered and the value of the energy gap (E _g) in the mineral has been estimated at 7.1 eV. Two groups of luminescence bands caused by impurities of intrinsic (growth, radiation) nature (E _max = 2.1, 2.7–2.8, and 3.2–3.3 eV) and matrix luminescence (E _max = 4.4−4.7 and 5.4 eV) probably with the participation of excitons were distinguished on the basis of selective excitation of zircon with different synchrotron energies relative to the gap value (E _excit < E _g, E _excit ∼ E _g, and E _excit ≫ E _g). The short-lived component with a response time of 4 ns has been revealed in the afterglow of zircon in the region of 5.4 eV.     

Calculation of the visible-UV absorption spectra of hydrogen sulfide,bisulfide, polysulfides,and As and Sb sulfides,in aqueous solution

Recently we showed that visible-UV spectra in aqueous solution can be accurately calculated for arsenic (III) bisulfides, such as As(SH)₃, As(SH)₂S^- and their oligomers. The calculated lowest energy transitions for these species were diagnostic of their protonation and oligomerization state. We here extend these studies to As and Sb oxidation state III and v sulfides and to polysulfides S _n ^2- , n = 2–6, the bisulfide anion, SH^-, hydrogen sulfide, H₂S and the sulfanes, S _n H₂, n = 2–5. Many of these calculations are more difficult than those performed for the As(iii) bisulfides, since the As and Sb(v) species are more acidic and therefore exist as highly charged anions in neutral and basic solutions. In general, small and/or highly charged anions are more difficult to describe computationally than larger, monovalent anions or neutral molecules. We have used both Hartree-Fock based (CI Singles and Time-Dependent HF) and density functional based (TD B3LYP) techniques for the calculations of absorption energy and intensity and have used both explicit water molecules and a polarizable continuum to describe the effects of hydration. We correctly reproduce the general trends observed experimentally, with absorption energies increasing from polysulfides to As, Sb sulfides to SH^- to H₂S. As and Sb(v) species, both monomers and dimers, also absorb at characteristically higher energies than do the analogous As and Sb(III)species. There is also a small reduction in absorption energy from monomeric to dimeric species, for both As and Sb III and v. The polysufides, on the other hand, show no simple systematic changes in UV spectra with chain length, n, or with protonation state. Our results indicate that for the As and Sb sulfides, the oxidation state, degree of protonation and degree of oligomerization can all be determined from the visible-UV absorption spectrum. We have also calculated the aqueous phase energetics for the reaction of S₈ with SH^- to produce the polysulfides, S _n H^-, n = 2–6. Our results are in excellent agreement with available experimental data, and support the existence of a S₆ species.     

Theoretical studies of the electronic structure of copper in tetrahedral and triangular coordination with sulfur

Molecular orbital calculations are presented for the copper-sulfur polyhedral clusters CuS ₄ ^7? , CuS ₄ ^6? , CuS ₃ ^5? and CuS ₃ ^4? , which occur in many minerals. Calculated and experimental optical and X-ray energies are found to be in good agreement. The crystal field orbitals of Cu⁺ in tetrahedrally coordinated sulfides are found to be less tightly bound than the S3p nonbonding orbitals by about 2–3 eV whereas the e and t ₂ crystal field orbitals are split by about 1 eV. The crystal field splitting of Cu²⁺ in tetrahedral coordination is about 0.7–0.8 eV while the separation of the S3p nonbonding orbitals and the partially filled t ₂ crystal field orbital is about 2 eV. In triangular coordination both the Cu⁺ and Cu²⁺ crystal field orbitals are more stable than in tetrahedral coordination, more widely split and more strongly mixed with the S3p orbitals. CuS is shown to be unstable as the mixed oxidation state compound Cu^2+III (Cu^+IV)₂S^2?(S ₂ ^2? ); rather each Cu is predicted to have a fractional oxidation state and partially-empty crystal field orbitals.     

The local structure of Ca-Na pyroxenes. I. XANES study at the Na K-edge

 X-ray absorption Na K-edge spectra have been recorded on synthetic endmember jadeite and on a series of natural Ca-Na pyroxenes compositionally straddling the Jd-Di join. The C2/c members of the series are systematically different from the P2/n members. Differences can be interpreted and explained by comparing the experimental spectra with theoretical spectra. These have been calculated by the multiple-scattering formalism from the atomic positional parameters determined by single-crystal X-ray diffraction structure refinement on the same samples. In the full multiple scattering region of the spectra (1075 to 1090 eV) C-pyroxenes exhibit three features which reflect the 6-2 configuration of the O back-scattering atoms around the Na absorber located at the center of the cluster (site M2 of the jadeite structure). P-pyroxenes show more complicated spectra in which at least four (possibly five) features can be recognized; they reflect the two types of configuration (6-2 and 4-2-2) of O around Na in the two independent M2 and M21 eight-fold coordinated sites of the omphacite structure. A weak, sometimes poorly resolved peak at 1079 eV is diagnostic and discriminates C- from P-pyroxenes. The Garnet Ridge C2/c impure jadeite exhibits a spectrum which is intermediate between those of jadeite and omphacite. The Hedin-Lundqvist potential proves best for these insulating materials and allows multiple-scattering calculations agreeing well with experiments. Received: July 11, 1996/Revised, accepted: October 21, 1996     

A theoretical study of the absorption spectra of Pb+ and Pb3+ in site K+ of microcline: application to the color of amazonite

The blue-green color of amazonite has been assigned by various authors to ions Pb⁺ (6 s)² (6 p) and/or Pb³⁺ (6 s) in site of K⁺ of microcline. Owing to the complex which forms between the ion Pb³⁺ and the lone pairs of the oxygen atoms surrounding it, the peripheral electron of Pb³⁺ passes on the levels (6 p) of the latter, which results in a great similarity of the spectra of Pb⁺ and Pb³⁺ in amazonite (the transition energies are multiplied by a factor greater than 1), whereas, in the isolated state, these spectra are completely different from one another. An analytical development of the crystal field around a site K⁺ is established. Under the effect of the crystal field, the transition ² P _1/2→² P _3/2 (6 p) is split into two double transitions. The lower transition only falls in the visible domain (1.6–1.8 eV for Pb⁺), the second in U−V. The green color would arise from the ion Pb⁺, whereas the blue one would be attributed to the ion Pb³⁺. Received: 23 January 1997 / Revised, accepted: 10 September 1997     

Sulfur K- and L-edge X-ray absorption spectroscopy of sphalerite,chalcopyrite and stannite

Sulfur K-edge x-ray absorption spectra (XANES and EXAFS) and L-edge XANES of sphalerite (ZnS), chalcopyrite (CuFeS₂) and stannite (Cu₂FeSnS₄) have been recorded using synchrotron radiation. The K- and L-edge XANES features are interpreted using a qualitative MO/energy band structure model. The densities of unoccupied states at the conduction bands of sphalerite, chalcopyrite and stannite are determined using S K- and L-edge XANES features (up to 15 eV above the edge), combined with published metal K-edge XANES. The SK- and L-edge XANES also indicate that, for sphalerite, the Fe²⁺ 3d band at the fundamental gap has little or no bonding hybridization with S 3p and S 3s orbitals; for chalcopyrite, the Cu⁺ 3d and Fe³⁺ 3d bands have strong mixing with S 3p and S 3s states, while for stannite the Cu⁺ 3d band strongly hybridizes with S 3p and S 3s orbitals, but the Fe²⁺ 3d band does not. The post-edge XANES features (15–50 eV above the edge) of sphalerite, chalcopyrite and stannite are similar. These features are related to the tetrahedral coordination of sulfur in all these structures, and interpreted by a multiple scattering model. The resonance energies from both the K-edge and L-edge XANES for these minerals are well correlated with reciprocal interatomic distances and lattice spaces. Sulfur K-edge EXAFS analyses using Fourier transform and curve fitting procedures are presented. Comparison of the structural parameters from EXAFS with x-ray structure data shows that the first shell bond distances (BD) from EXAFS are usually accurate to ±0.02 Å, and that coordination numbers (CN) are generally accurate to ±20 percent. For sphalerite, EXAFS analysis yields the structure parameters for the first three neighbour shells around a sulfur atom; the BD and CN even for the third shell are in close agreement with the x-ray structure, and the Debye-Waller term decreases from the first shell to the third shell. It is shown that sphalerite (ZnS) is a good model compound for EXAFS analysis of sulfur in chalcogenide glasses and metalloproteins.     

Temperature dependence of the polarized electronic absorption spectra of olivines. Part I – fayalite

 Polarized electronic absorption spectra of orthorhombic fayalite, Fe₂SiO₄, [E || a(|| Z),E || b(|| X), E || c(|| Y)], space group Pbnm, have been studied in the temperature range 293 ≤T/K ≤ 1273. The spectra were analysed into component bands originating from spin-allowed dd transitions of iron(II) at the different sites, M1 and M2, in the structure. The assignments of bands, made on the basis of the polarization dependence of the spectra and considerations of transition energies, were confirmed by the analysis of the temperature-dependent spectra. The temperature dependencies of integral intensities, half band widths and energy positions of absorptions bands caused by Fe²⁺ on the different octahedral sites, M1 and M2, were evaluated for the individual transitions. Independent of the site symmetry, absorption bands shift to lower energies and half band widths increase on rising temperature. The temperature dependence of band intensities depends on site symmetry. The integral intensities are found to increase with temperature for the transition metal ion on a centrosymmetric site, or remain constant when the site is missing an inversion centre. This is consistent with the general conclusion of Taran et al. (1994). Received: 11 October 2001 / Accepted: 17 January 2002     

Atomistic calculations of structural and elastic properties of serpentine minerals: the case of lizardite

The physical properties of the hydrous phyllosilicate lizardite have been investigated by atomistic simulation using the GULP code based on transferable semi-empirical interatomic potentials. Lizardite behavior was first investigated during structure relaxation at room temperature. The Helmholtz free energy is minimum for an equilibrium structure that is in agreement with experiment. The bulk, shear, and Young modulii for lizardite were calculated along with the Poisson ratio. From the shear and bulk modulii, we also calculated translational and longitudinal acoustic wave velocities that are important quantities for tectonophysics models. As expected, lizardite is stiffer in the a direction parallel to the layers than in the c perpendicular direction; the variation of the unit cell parameters with pressure is in good agreement with experiment. The cohesive energy between two successive layers along c direction was calculated at 0.33 eV (i.e., 0.11 eV per OH bond) in good agreement with recent ab initio calculations. Upon pressure and temperature variations, we evidenced that structural changes are mainly pressure induced; pressure being accommodated by a decrease of the c parameter up to 10 GPa. We also found that the change of slope in the derivative of the c cell parameter with respect to pressure occurring around 2 GPa originates from the bending of the interlayer hydroxyl groups with respect to the layer normal direction.     

Electronic environments in carrollite, CuCo2S4, determined by soft X-ray photoelectron and absorption spectroscopy

Fracture surfaces of a natural carrollite specimen have been characterised by synchrotron and conventional X-ray photoelectron spectroscopy and near-edge X-ray absorption spectroscopy. For the synchrotron X-ray measurements, the mineral surfaces were prepared under clean ultra high vacuum and were unoxidised. The characterisation was undertaken primarily to establish unequivocally the oxidation state of the Cu in the mineral, but also to obtain information on the electronic environments of the Co and S, and on the surface species. Experimental and simulated Cu L_2,3-edge absorption spectra confirmed an oxidation state of Cu^I, while Co 2p photoelectron and Co L_2,3 absorption spectra were largely consistent with the Co^III established previously by nuclear magnetic resonance spectroscopy. S 2p photoelectron spectra provided no evidence for S to be present in the bulk in more than one state, and were consistent with an oxidation state slightly less negative than S^-II. Therefore it was concluded that carrollite can be best represented by Cu^ICo^III₂(S₄)^-VII. The Cu^I oxidation state is in agreement with that expected for Cu tetrahedrally coordinated by S, but is in disagreement with the Cu^II deduced previously from some magnetic, magnetic resonance and Cu L-edge X-ray absorption spectroscopic measurements. A significant concentration of S species with core electron binding energies both lower and higher than the bulk value were formed at fracture surfaces, and these entities were assigned to monomeric and oligomeric surface S species. The density of Cu d states calculated for carrollite differed from that previously reported but was consistent with the observed Cu L₃ X-ray absorption spectrum. The initial oxidation of carrollite in air under ambient conditions was confirmed to be congruent, unlike the incongruent reaction undergone by a number of non-thiospinel sulfide minerals.