Raman spectroscopy was assessed for its ability to rapidly identify asbestos phases by submitting the amphiboles anthophyllite, amosite, crocidolite and tremolite to spectroscopic analysis. All the phases were characterized by using XRD, SEM, TEM (CTEM, SAED), AEM and EDS techniques. This study demonstrates that accurate identification of the mineralogical phase can be attained by analysing the position in the Raman spectrum of the bands corresponding to the symmetric stretching modes (vs) and to the antisymmetric stretching modes (vas) of the different Si—O linkages. Raman spectroscopy thus proves to be an effective technique for rapidly distinguishing the different fibrous minerals examined.

The haüynophyre emitted from a parasitic vent of the Vulture stratovolcano is a S- and Cl-rich, leucitemelilite- bearing lava flow containing an unusually large amount of sodalite-group minerals (>23 vol.%). Mineralogical and chemical study of phenocrysts has led to the identification of black haüynes, blue lazurites and of Cl-rich white or black noseans. X-ray diffraction (XRD) study confirms the occurrence of nosean having a low symmetry (P23). Raman spectra and XRD data show that S is fully oxidized to SO4 in black haüynes and in white noseans, while it is partly reduced to form S3– groups in blue lazurites, which also contain H2O molecules. Structural and chemical data strongly question the validity of the Hogarth and Griffin (1976) method widely used to resolve the ratio S6+/S2– in sodalite-group phases from EMPA data. Among euhedral phenocrysts, large lazurites are only faintly zoned. All other phases show variable core-rim chemical zoning and many phenocrysts are partially resorbed and/or colour-zoned. Black haüynes have highly variable S/Cl and slightly lower SiO2/Al2O3 ratios, larger FeTOT contents and more compatible trace elements than lazurites. Thin opaque noseansodalite rims surrounding all crystals are interpreted as a result of rapid crystallization driven by exsolution of a S-scavenging fluid phase. We suggest that the extreme complexity of the mineralogical assemblage reflects variable aSiO2 and aH2O of the silicate melts.

Idiomorphic crystals (up to 3.5 mm) of ferroan platinum, cooperite and mertieite-II were found in a heavy-mineral concentrate from stream sediments of the Darya river in the Aldan Shield, Russia. Pt-Fe crystals display cubic and thin platy habits; occasionally they are twinned. The chemical composition ranges from Pt2.64Fe1.00 to Pt2.88Fe1.00 with Os, Ru, Ir, Rh and Pd below the analytical detection limit of the electron microprobe. X-ray diffractometry of Pt-Fe crystals suggests a F-centred cubic lattice, characteristic of ferroan platinum. Some of the ferroan platinum crystals have large (about 100 μm wide) cooperite overgrowth rims or are covered by a Au-Ag alloy. Cooperite also occurs as large euhedral crystals (up to 3 mm across, partly twinned). Crystals of mertieite-II are speckled with μm-sized (2–5 μm) inclusions of sperrylite and intergrown with minerals of cooperite-braggite solid solution, Pt-Pd-Hg alloy, keithconnite and Au-Ag alloy. Fractures along crystallographic planes of the mertieite-II crystals are filled with Pd-Pt-Fe-Sb-As-Hg-Te-Bi-bearing oxides. The coarse-grained PGM from the Darya have a geological setting similar to the Kondyor PGE placer 75 km to the northeast and are probably related to clinopyroxenite-hornblende-magnetite units of Alaskan-/Uralian-type intrusions.

The Soufrière Hills stratovolcano on the Caribbean island of Montserrat has been erupting since 18th July 1995. An enormous amount of respirable volcanic ash has been suspended into the atmosphere by the eruptions and wind re-suspension of deposited ash. The large amount of fine, airborne particulate matter, in particular the component 10 μm equivalent aerodynamic diameter (PM10), is a cause of medical concern. Airborne levels have frequently exceeded the UK environmental standard for PM10, (50 μg/m3), although it is noted that this standard was primarily set for urban PM10. The crystalline silica in the ash is mostly cristobalite, at reported levels up to 20%. The UK government’s Chief Medical Officer, referring to Montserrat, has suggested that long-term exposure to high levels of volcanic ash could lead to silicosis. These concerns have prompted government-funded investigations into the potential toxicity of well characterized volcanic ash samples from Montserrat. Given the well established toxicity of cristobalite, particular attention was paid to the amount of this mineral in the ash samples. Three ash samples were tested: (1) a vulcanian eruption ash, (2) ash released in a domecollapse pyroclastic flow, and (3) ash from a major vulcanian explosion that was wind-transported to, and deposited on, the neighbouring island of Antigua. Comparative toxicological studies were carried out on respirable preparations of these three samples together with appropriate control mineral dusts that matched the major components of the Montserrat samples: anorthite, labradorite, cristobalite/ obsidian and cristobalite. Alpha quartz (DQ12) was the positive control. All samples, including the controls, were characterized to establish particle-size distributions, particle morphologies, and to confirm the mineralogy. Rats were challenged with 1 mg via intratracheal instillation, and groups sacrificed at three time points (1, 3 and 9 weeks). Health assessment was made by examining endpoints of increasing lung damage such as inflammation, permeability (oedema), changes in epithelium, and increase in the size of broncho-thoracic lymph nodes. The data indicate that Montserrat respirable ash, derived from dome collapse pyroclastic flows or vulcanian explosions, has minimal acute bioreactivity in the lung. The feldspar standards showed low bioreactivity, in stark contrast to the cristobalite standard that showed progressive increases in lung damage. These results suggest that either the mass of cristobalite present in the Montserrat ash was insufficient to cause an effect in the lung, or the cristobalite in the ash was, for some as yet unknown reason, markedly less bioreactive than our pure cristobalite standard.

Lapilli and spatter of natrocarbonatite extruded from the T56B cone of Oldoinyo Lengai (Tanzania) in August 2003 are unusually rich in apatite, silicate, oxide and sulphide minerals. These minerals occur primarily within inclusions of quenched natrocarbonatite in gregoryite and fractures in nyerereite-gregoryite intergrowths. Silicates include members of the sodian cuspidine–niobian cuspidine–niocalite solid-solution series and manganoan monticellite. Oxides are represented by members of the magnesioferrite–jacobsite–magnetite solid-solution series. Sulphides occurring in decreasing order of abundance include: ferroan alabandite, manganoan ferroan sphalerite, galena and pyrrhotite. Petrographic and compositional data for these minerals are interpreted to indicate that all are high-temperature (<900–650°C), early-crystallizing phases from relatively-unevolved natrocarbonatite magma, with the sulphides forming prior to the silicates, and the latter before gregoryite. Sulphur fugacityand oxygen fugacity of natrocarbonatite magma are considered to decrease and increase, respectively, during the formation of the sulphide-oxide assemblage. Crystallization of cuspidine-niocalite, monticellite and Si-bearing apatite rapidly depletes natrocarbonatite magma of its initial low silica content, preventing crystallization of late-stage groundmass silicates.

Micro-inclusions (1 –10 μm) in 55 diamonds of cubic habit from the Udachnaya kimberlite pipe have been studied using vibrational spectroscopy. This has revealed a multiphase assemblage in cuboid diamonds from the Udachnaya kimberlite pipe. This assemblage includes carbonates, olivine, apatite, graphite, water and silicate glasses. The micro-inclusions preserve the high internal pressure and give confidence that the original materials were trapped during growth of the host diamond. The internal pressures, extrapolated to mantle temperatures, lie within the stability field of diamond and the relatively low temperatures are typical for the formation of cuboid diamonds. In contrast to previously reported data for African diamonds, the micro-inclusions in the cuboids from Udachnaya are extremely carbonatitic in composition (H2O/(H2O+CO2) ≈5 –20%) with the observed assemblage of microinclusions similar to some types of carbonatites. The low water and silica content testify that the material in the micro-inclusions of the Udachnaya diamonds was near-solidus carbonatitic melt. Vibrational spectroscopy has provided the evidence of carbonatitic melts in cuboid diamonds.

Skaergaardite, PdCu, is a new mineral discovered in the Skaergaard intrusion, Kangerdlugssuaq area, East Greenland. It occurs in a tholeitiic gabbro associated with plagioclase, clinopyroxene, orthopyroxene, ilmenite, titanian magnetite, fayalite and accessory chlorite-group minerals, ferrosaponite, a member of the annite–phlogopite series, hornblende, actinolite, epidote, calcite, ankerite, apatite and baddeleyite. The mineral is found in composite microglobules composed of bornite, chalcocite, digenite, chalcopyrite, with rare cobalt pentlandite, cobaltoan pentlandite, sphalerite, keithconnite, vasilite, zvyagintsevite, (Cu,Pd,Au) and Pt-Fe-Cu-Pd alloys, unnamed PdCu3, (Pd,Cu,Sn), Au3Cu and PdAuCu. Skaergaardite occurs as droplets, equant grains with rounded outlines, subhedral to euhedral crystals and as irregular grains that vary in size from 2 to 75 μm, averaging 22 μm. It is steel grey with a bronze tint, has a black streak, a metallic lustre and is sectile. Neither cleavage nor fracture was observed. The mineral has a micro-indentation hardness of VHN25 = 257. It is isotropic, non-pleochroic and exhibits neither discernible internal reflections nor evidence of twinning. Skaergaardite varies from bright creamy white (associated with bornite and chalcopyrite) to bright white (associated with digenite and chalcocite). Reflectance values in air (and in oil) are: 58.65 (47.4) at 470 nm, 62.6 (51.1) at 546 nm, 64.1 (52.8) at 589 nm and 65.25 (53.95) at 650 nm. The average of 311 electron-microprobe analyses gives: Pd 58.94, Pt 1.12, Au 2.23, Cu 29.84, Fe 3.85, Zn 1.46, Sn 1.08, Te 0.28 and Pb 0.39, total 99.19 wt.%, corresponding to (Pd0.967Au0.020Pt0.010)Σ0.997(Cu0.820Fe0.120 Zn0.039Sn0.016Te0.004Pb0.003)Σ1.002. The mineral is cubic, space group Pm3m, a = 3.0014(2) Å, V = 27.0378 Å3, Z = 1. Dcalc is 10.64 g/cm3. The six strongest lines in the X-ray powder-diffraction pattern [d in Å (I)(hkl)] are: 2.122(100)(110), 1.5000(20)(200), 1.2254(50)(211), 0.9491(20)(310), 0.8666(10)(222), 0.8021(70)(321). The mineral has the CsCl-type structure. It is believed to be isostructural with wairauite (CoFe), synthetic CuZn (β-brass) and is structurally related to hongshiite (PtCu). Skaergaardite developed from a disordered Pd-Cu-rich metal alloy melt that had exsolved from an earlier Cu-(Fe) sulphide melt. Ordering of Pd and Cu (beginning at T ≈ 600°C) results in development of the CsCl structure from a disordered face-centred cubic structure.

Serpentine minerals and brucite in ultramafic rocks from the South Chamorro Seamount were characterized chemically to investigate the serpentinization of the Mariana forearc mantle. Relict primary minerals of the serpentinites are olivine, enstatite and minor Cr-spinel and diopside. The secondary minerals are mostly serpentine and brucite with minor magnetite. The serpentine minerals, mostly lizardite and chrysotile, display large compositional variations. Al2O3 and Cr2O3 contents depend generally upon the nature of the primary mineral from which the serpentine was derived. Both serpentine minerals and brucite exhibit wide Mg, Fe and Mn substitution: the Mg# ranges are 95.1–77.2 and 88.9–60.8, respectively. These mineralogical and chemical features allowed us to estimate an upper temperature limit for serpentinization of ∼200–300°C, in agreement with recent thermal models which suggest that the serpentinized mantle wedge of the Izu-Bonin-Mariana subduction zone is cold. The high degree of serpentinization (40–100%, average >75%), and the serpentine + brucite paragenesis of these ultramafics imply that the Mariana forearc mantle has a significantly reduced density and strength down to ∼30 km, which provides a driving mechanism for serpentinite diapirism. Pervasive serpentinization of the forearc by fluids released from the décollement zone also explains the low seismicity of the Izu-Bonin-Mariana subduction zone.

Mesothermal gold ores in the Archaean Yilgarn Craton of Western Australia are dominated by a pyrite ± arsenopyrite ± pyrrhotite sulphide assemblage. Many of these ores are refractory to varying degrees and require treatment by roasting, bacterial oxidation or finer milling. The most common sulphide ore types can be sub-divided broadly into pyritic (pyrite±pyrrhotite) and arsenical types (pyrite+arsenopyrite± pyrrhotite). Arsenical ores vary from highly refractory to free-milling. Arsenopyrite in highly refractory ores is finer grained, As-deficient (27 –32.5 at.% As), contains high average concentrations of submicroscopic gold (60 –270 ppm), but does not contain inclusions of particulate gold. Arsenopyrite in free-milling ores is coarser grained, less As-deficient to slightly As-rich (30 –35 at.% As), contains low or negligible concentrations of submicroscopic gold, but contains inclusions and fracture fillings of particulate gold. In some refractory arsenical ores, pyrite also contains moderately high levels of submicroscopic gold (20 –40 ppm), the concentration of which is directly related to As content of the pyrite.

Pyritic ores are free-milling to mildly refractory, or rarely moderately refractory. Pyrite in pyritic ores contains negligible to low levels of submicroscopic gold (<5 ppm). Other reasons for refractory behaviour in pyritic ores include very fine-grained native gold inclusions in pyrite, or the presence of gold-bearing tellurides.

It is concluded that submicroscopic gold is incorporated into the crystal lattices of arsenopyite and arsenical pyrite at sub-greenschist to lower greenschist-facies temperatures, and is progressively expelled as inclusions and fracture fillings of native gold in sulphides, and ultimately into the gangue, as recrystallization proceeds through upper greenschist- into amphibolite-facies temperatures, during deformation and burial. Submicroscopic gold is expelled more rapidly from pyrite than arsenopyrite.

Pyrrhotite progressively replaces primary pyrite at higher temperatures, but rarely contains gold. Finally, a metallurgical classification scheme for refractory ores is presented which incorporates the above mineralogical conclusions.

Gold-silver vein deposits in the Taebaeksan district, Korea, coexist in time and space with a variety of other deposit types such as skarns, hydrothermal carbonate replacement bodies and Carlin-like deposits and reflect proximity to a magmatic source. The seven gold-silver deposits within the district show common features such as multiple complex veins, an abundance (up to 30% in ore) of base-metal sulphides, a wide range of Ag/Au ratios and the common occurrence of carbonate. Quartz-vein textures indicate open-space-filling at shallow crustal levels. On the basis of Ag/Au ratios of ore, mode of occurrence, and associated mineral assemblages, the seven deposits studied can be classified as follows; Au-dominant type (Group I), Au-Ag type (Group II), Ag-dominant type (Group IIIA) and base-metal and Ag-dominant type (Group IIIB). Group I is characterized by paragenetically early Pb-Zn basemetal sulphides with electrum and late, rare Ag-sulphides and Ag-sulphosalts, whereas Group III contains more Ag-sulphides and/or Ag-sulphosalts. Group II is gold-rich, but transitional to Group III. The Au contents and FeS contents of electrum and sphalerite, respectively, from all of the deposits decreased as mineralization proceeded. Temperature and log fS2 conditions of gold-silver mineralization tend to decrease from Group I and II to Group III deposits (i.e. 340–270°C, –9.3 to –11.8 bar, 320–240°C, –9.5 to –10.3 bar, 250–160°C, –12.5 to –16.9 bar, respectively) as well as from the main to late stages of mineralization in each deposit. The systematic mineralogy and variation of physicochemical conditions in Groups I, II and III are thought to be due to their relative positions with respect to a magma source that is genetically related to a low-to-intermediate-sulphidation porphyry system. Au-rich deposits are proximal to a magmatic source, whereas Ag-rich deposits are more distal.

Zircon- and reidite-type ZrSiO4 produced by shock recovery experiments at different pressures have been studied using infrared (IR) and Raman spectroscopy. The v3 vibration of the SiO4 group in shocked natural zircon shows a spectral change similar to that seen in radiation-damaged zircon: a decrease in frequency and increase in linewidth. The observation could imply a possible similar defective crystal structure between the damaged and shocked zircon. The shock-pressure-induced structural phase transition from zircon (I41/amd) to reidite (I41/a) is proven by the occurrence of additional IR and Raman bands. Although the SiO4 groups in both zircon- and reidite-ZrSiO4 are isolated, the more condensed scheelite gives rise to Si–O stretching bands at lower frequencies, suggesting a weakening of the bond strength. Low-temperature IR data of the reidite-type ZrSiO4 show an insignificant effect of cooling on the phonon modes, suggesting that the structural response of reidite to cooling-induced compression is weak and its thermal expansion is very small.

Tertiary ultrapotassic volcanic rocks from Serbia occasionally display low levels of K2O and K2O/ Na2O. In these rocks, analcime regularly appears as pseudomorphs after pre-existing leucite microphenocrysts. The process ofleucite transformation in Serbian ultrapotassic rocks is very thorough: fresh leucite survives only in ugandites from the Koritnik lava flows as well as in rare inclusions in Cpx. This paper focuses on the impact of ‘analcimization’ on the mineralogy and geochemistry ofthe Serbian ultrapotassic rocks, using the samples where leucite survived as a monitor for the process.

Analcimization has had a great impact on the geochemistry of the rocks, but affects only a restricted number of chemical parameters. These are the falsification of the original K2O/Na2O ratio, the decoupling oflarge-ion lithophile elements resulting in considerable depletion of Rb and K2O, but not ofBa, and sporadic, but extreme enrichment ofCs in some analcime-bearing samples (up to 900 ppm). Analcimization is also recognized by an increase in whole-rock δ18O values of ∼3% compared to fresh rocks, which correlates with the level of whole-rock hydration. Finally, the 87Sr/86Sr enrichment at nearly constant 143Nd/144Nd demonstrated by some rocks can also be explained by the analcimization ofleucite. For samples with variable 87Sr/86Sr from the same lava flow, 87Sr/86Sr values correlate with modal analcime abundance (ex-leucite), loss on ignition of whole-rock and whole-rock δ18O values. The extreme depletion in K and enrichment in Na, together with modification of other geochemical parameters, may have led to the misinterpretation of the origin and geodynamic affiliations of the Serbian ultrapotassic rocks, had the effects of analcimization not been taken into account.

Coherent intergrowths, at the lattice scale, between cuprobismutite (N = 2) and structurally related paděraite along both major axes (15 Å and 17 Å repeats) of the two minerals are reported within skarn from Ocna de Fier, Romania. The structural subunit, DTD, 3 layers of paděraite, is involved at interfaces of the two minerals along the 15 Å repeat, as well as in transposition of 1 paděraite unit to 2 cuprobismutite units along the 17 Å repeat in slip defects. Lattice images obtained by HRTEM across intervals of 200–400 nm show short- to long-range stacking sequences of cuprobismutite and padeÏ raite ribbons. Such nanoscale slabs mimic mm-scale intergrowths observed in back-scattered electron images at three orders of magnitude greater. These slabs are compositionally equivalent to intermediaries in the cuprobismutite-paděraite range encountered during microanalysis. Hodrushite (N = 1.5) is identified in the mm-scale intergrowths, but its absence in the lattice images indicates that, in this case, formation of polysomes between structurally related phases is favoured instead of stacking disorder among cuprobismutite homologues. The tendency for short-range ordering and semi-periodic occurrence of polysomes suggests they are the result of an oscillatory chemical signal with periodicity varying from one to three repeats of 15 Å, rather than simple ‘accidents’ or irregular structural defects. Lead distribution along the polysomes is modelled as an output signal modulated by the periodicity of stacking sequences, with Pb carried within the D units of paděraite. This type of modulator acts as a patterning operator activated by chemical waves with amplitudes that encompass the chemical difference between the minerals. Conversion of the paděraite structural subunit DTD to the C unit of cuprobismutite, conserving interval width, emphasizes that polysomatic modularity also assists interference of chemical signals with opposite amplitudes. Observed coarsening of lattice-scale intergrowths up to the mm-scale implies coupling between diffusion-controlled structural modulation, and rhythmic precipitation at the skarn front during crystallization.

A Th-rich mineral of the crandallite group has been investigated from the weathering profile of the Schugorsk bauxite deposit, Timan, Russia. It occurs within thin (up to 0.5 mm) organic-rich veinlets together with ‘leucoxene’ in the form of small shapeless grains which vary in size from 1—2 mm to 60—70 mm. Rare grains disseminated among boehmite crystals were also found. Microprobe analyses determined that the ThO2 content can be as high as 18 wt.%. The mineral composition is intermediate between crandallite CaAl3H(PO4)2(OH)6, goyazite SrAl3H(PO4)2(OH)6, Th-crandallite and svanbergite SrAl3PO4SO4(OH)6 in the beudantite group.

Comparatively high contents of Fe and Si and a very high positive Th and Fe content correlation (r = +0.98) suggest that the formula of the hypothetical Th-bearing end-member is ThFe3(PO4,SiO4)2(OH)6 with Th and Si substituting for REE and Prespectively (woodhouseite-type substitution). Another possible substitution is Th4+ + Ca2+ ⇋ 2REE3+ (florencite-type). A deficiency of cations in the X site can be explained by either the presence of carbon, undetectable by microprobe, in the crystal lattice or a lack of X-site cations due to radiation damage induced by Th. Some excess of cations in the B site (Al and Fe3+) can be explained by the presence of very small boehmite and hematite inclusions on the crandallite grain surfaces. Th-rich crandallite may be the result of alteration of an unidentified silicate mineral from the parent rock with a composition close to the simplified formula Fe2+ThSiO4(OH)2.

We have performed experiments to determine the extents to which Fe and H2O are lost from hydrous basaltic melts contained in Au80Pd20 and graphite-lined Pt capsules at 0.7 –1 GPa and 1300–1350°C. All experiments were performed in the piston-cylinder apparatus. In order to minimize the possibility of rupture of the AuPd capsule and to control H2O loss we used a double-capsule method. The inner welded 2 mm diameter Au80Pd20 capsule was placed inside a welded 3 mm diameter Pt capsule, the intervening space being packed with hydrous sample. Loss of FeO* from the sample was found to be ≤4% relative in both the Au80Pd20 and graphite-lined Pt capsules in experiments of up to 24 h duration. Loss of H2O is greater and it depends on the oxidation state of the starting materials and the nature of the capsule. For starting mixes fired at 1 log fO2 unit above the quartz-fayalite-magnetite (QFM) buffer at 1 atm, H2O loss from Au80Pd20 capsules averaged 9% relative. Starting mixes fired at 1 log fO2 unit below the QFM buffer at 1atm lost, on average, 32% of their H2O when run in Au80Pd20 capsules at high pressure. All samples run in graphite-lined Pt capsules experienced dramatic H2O loss, averaging 52% relative, irrespective of initial oxidation state. We conclude that Au80Pd20 capsules are suitable for high-pressure hydrous melting experiments and that the sample loses very little Fe. In order to minimize H2O-loss, however, it is important that the starting materials be relatively oxidized.

Pyrochlore has been considered as a waste form to immobilize high-level nuclear waste such as excess weapons-grade plutonium. In order to study the aqueous stability of pyrochlore, we have carried out hydrothermal experiments with a natural microlite (Nb + Ta > 2Ti; Ta ⩾ Nb) at 175°C in a neutral and acidic solution for 14 days. The starting material and the experimental products were studied by electron microprobe, backscattered electron (BSE) imaging, powder infrared (IR) and micro-Raman spectroscopy, and powder X-ray diffractrometry (XRD). The microlite has small U (∼200 ppm) and Th (∼800 ppm) contents and is crystalline. The hydrothermal treatment in a solution containing l mol/l HCl and l mol/l CaCl2 causes the partial replacement of the microlite (up to ∼10 μm) by a new pyrochlore phase. This new phase is characterized by a larger unit-cell volume and contains a large number of vacancies at the A site (A = Ca, Na) as well as anion vacancies, molecular water, and possibly OH groups. Analyses of the experimental fluid further revealed that U was also selectively lost to the solution. Treatment in pure water did not produce reaction zones detectable by BSE imaging or powder XRD. However, significant spectral changes in powder IR spectra of the reaction product and the detection of Na and Ca in the solution indicate that the microlite has also reacted in pure water. The experimental alteration features bear a remarkable resemblance to those seen in naturally altered microlite samples, suggesting that short-term experimental results can be applied to natural systems and vice versa in order to evaluate the long-term stability of a pyrochlore waste form.

The structural behaviour of a haüyne with a chemical composition of Na4.35Ca2.28K0.95[Al6Si6O24]- (SO4)2.03, at room pressure and from 33 to 1035°C on heating, was determined by using in situ synchrotron X-ray powder diffraction data (λ = 0.92249(5) Å). The satellite reflections in haüyne are lost at ∼400°C and a true substructure results because of this phase transition. There is a discontinuity in the a unit-cell parameter at ∼585°C. The α parameter increases rapidly and non-linearly to 585°C, but above 585°C, the expansion rate decreases. The percent volume change between 33 and 576°C is 2.0(3)%, and 0.6(3)% between 593 and 1035°C. Between 33 and 1035°C, the Al–O, Si–O and S–O distances are constant. Between 33 and 576°C, the angle of rotation of the AlO4 tetrahedron, jAl, changesfrom 11.5 to 5.8°, while the angle of rotation of the SiO4 tetrahedron, φSi, changesfrom 12.4 to 6.3°. The Al–O–Si bridging angle changesfrom 150.05(2) to 153.08(1)° from 33 to 576°C. Beyond 585°C, φAl and φSi angles remain nearly constant even though the maximum rotation of the tetrahedra is not achieved. Moreover, the Al–O–Si angle continues to increase at a slower rate from 585 to 1035°C by 1.05(2)°. From 33 to ∼585°C, the K atom position migrates at a slower rate than the Na and Ca atoms, and the structure expands at a high rate. Beyond 585°C, all the atomic positions of the interstitial cations (Na+, K+, Ca2+) remain nearly constant and the expansion of the structure is retarded.

The lamproite of Gaussberg is an ultrapotassic rock where leucite, olivine and clinopyroxene microphenocrysts occur in a glass-rich groundmass, containing microliths of leucite, clinopyroxene, apatite, phlogopite and rare K-richterite.

Abundant silicate melt inclusions occur in olivine, leucite and, rarely, in clinopyroxene microphenocrysts. Raman investigations on melt inclusions showed the presence of pure CO2 in the shrinkage bubbles. On the other hand, the glass of the groundmass is CO2-poor and contains up to 0.70 wt.% of dissolved H2O, as estimated by infrared spectra. It is inferred that CO2 was released at every stage of evolution of the lamproite magma (CO2-rich shrinkage bubbles), whereas H2O was retained for longer in the liquid. At Gaussberg, CO2 seems to have a major role at relatively high pressure where it favoured the crystallization of H2O-poor microphenocrysts; the uprise of the magma to the surface decreased the solubility of CO2 and caused a relative increase in water activity. As a consequence, phlogopite and K-richterite appeared in the groundmass.

The glass composition of both the groundmass and melt inclusions suggests different evolutions for the residual liquids of the investigated samples. Sample G886 shows the typical evolution of a lamproite magma, where the residual liquid evolves toward peralkaline and Na-rich composition and crystallizes K-richterite in the latest stage. Sample G895 derives from mixing/mingling of different batches of magma; effectively glasses from melt inclusions in leucite and clinopyroxene are more alkaline than those found in early crystallized olivine. Leucite and clinopyroxene crystallized early from a relatively more alkaline batch of lamproite magma and, successively, a less alkaline, olivinebearing magma batch assimilated them during its rise to the surface.

An investigation of the relative stabilities and equations of state of possible Fe3S polymorphs was conducted using first-principles pseudopotential calculations. These calculations were based on density functional theory and performed using ultrasoft Vanderbilt pseudopotentials within the generalized gradient approximation. In accord with experiment, we found that the tetragonal Fe3P-type polymorph is the only stable phase along the 0 K isotherm as a function of pressure. Fe3S exhibits permanent magnetism at ambient conditions (Fei et al., 2000), but magnetism is suppressed by pressure and temperature, and therefore non-magnetic data are appropriate ones to use for modelling planetary interiors. For this reason, and because the Fe3P-type polymorph of Fe3S contains 32 atoms per unit cell it was impractical to incorporate magnetic properties into the simulations of this phase, we studied the behaviour of the non-magnetic phase. We obtained values of 250 GPa for the bulk modulus, K0, and 4.61 for its first derivative withrespect to pressure, K0′, by fitting a 3rd order Birch-Murnaghan equation of state to the calculated internal energy as a function of volume for the non-magnetic Fe3P-type Fe3S. This suggests that a pressure far greater than that expected in the Martian interior would be needed to achieve a density comparable to that of the Martian core. We therefore conclude that it is unlikely that the core of Mars contains significant amounts of solid Fe3S.

La Colorada, in Zacatecas State, Mexico is an epithermal Ag-Zn-Pb system hosted in Mesozoic calcareous sedimentary rocks overlain by Tertiary volcanic rocks. The dominant vein is associated with a fault system that accommodates Tertiary normal and strike-slip faulting. The ore consists of fahlore [(Cu,Ag)10(Zn,Fe)2(Sb,As)4S13], polybasite [(Ag,Cu)16Sb2S11]–pearceite [(Ag,Cu)16As2S11] solid solution, pyrargyrite [Ag3SbS3]–proustite [Ag3AsS3] solid solution, acanthite-argentite [Ag2S], and native silver; associated sulphides include galena, sphalerite, chalcopyrite and pyrite. The Ag:Sb of the bulk concentrate from the mine is 1.076 and the Ag:Pb is 0.088. Compositions of the assemblages fahlore + pyrargyrite-proustite + sphalerite, and fahlore + polybasite-pearceite solid solution + (Ag,Cu)2S solid solution + sphalerite encapsulated in quartz and sphalerite indicate a primary depositional temperature of ∼325°C for a depth between 725 and 960 m below the inferred palaeosurface, which is in accord with fluid-inclusion data for higher elevations in the mine.