Basalt glass from Kilauea, Hawaii (SBG) starts to crystallize upon heat treatment in air at ∽840°C. In addition, oxidation takes place. The kinetics of both processes may be readily quantified though the mechanism is difficult to access. This work investigates the relationship between these processes, crystallization and oxidation, and the escape of volatiles from the glass/liquid upon reheating. Evolved gas analyses and differential scanning calorimetry are the techniques used. In addition, isothermal heat treatment in argon is carried out in order to eliminate the external reason for oxidation, the oxygen gradient, and to focus on intrinsic oxidation. Products are examined with 57Fe Mössbauer spectroscopy. As expected, degassing temperatures of SBG, and of two MORB samples, considered for comparison, are above the glass transition temperature. We find no convincing evidence of intrinsic oxidation. Degassing is likely to induce interface-controlled crystallization.

Late Tertiary-Quaternary volcanism around Calatrava, within the Hercynian massif of central Spain, is alkaline mafic-ultramafic, with ∼250 centres, mainly monogenetic cones and vents, with melilitite the most abundant eruptive. Carbonatite may be expected in association with melilitite and a clear example of magmatic carbonate emerged from a brief field reconnaissance. It is a vent filled with a mixed eruption of glassy melilitite lapilli in a carbonate matrix. Levels and profiles of trace elements are inseparable from recognized carbonatite, and totally unlike those in local sedimentary limestone and caliche. C and O isotopes are the same as those reported for carbonatite ashes in other provinces. Carbonate is present as globules in the melilitite glass, and as inclusions within large clinopyroxene and olivine grains, which are largely xenocrystic. Euhedral spinels in melilitite and carbonate matrix are chromite mantled with titano-magnetite, reported previously only from high-temperature kimberlite. Wehrlite fragments indicate direct eruption from the mantle. Phlogopite and chromian spinel are found only as inclusions in the olivine, pointing to a phlogopite-carbonate mantle source region of the type favoured for carbonatite and melilitite melt generation. Calatrava represents the most southerly and westerly expression in mainland Europe of intra-continental carbonatite-melilitite volcanism. Follow-up field visits have shown that carbonate volcanism is extensive and voluminous throughout the province, requiring a long-term research programme.

The nature and mode of origin of quartz-cored tourmalines (QCT) are studied from hydrothermal quartz veins within massive quartz-tourmaline (MQT) rocks at Roche, SW England. The QCT are annular, have blue maximum absorption colour and occur together with tourmalines with brown cores rimmed by blue tourmaline. Where the quartz core is not continuous throughout the length of the QCT crystals, the tourmaline core has brown maximum absorption colour, similar to tourmalines without quartz cores. Both the blue and brown tourmalines are schorl, but are compositionally distinct showing different Fe/(Mg + Ti) ratios and Ca concentrations. Fluid inclusion data indicate quartz precipitation from a moderate salinity (∼20–25 wt.% NaCl eq.) brine which periodically boiled following pressure drops within the vein system. The QCT show rheomorphic and lobate textures on their inner margins indicating selective dissolution of their brown, relative Mg-, Ti- and Ca-rich tourmaline cores and replacement with quartz. This presents a problem in terms of the nature of the fluid responsible for such selective dissolution because tourmaline is generally highly resistant under the normal range of hydrothermal fluid conditions. It is proposed that the relatively high concentrations of Ti, Mg and Ca in the brown tourmaline caused significant lattice strain, which together with an increase in pH, and probably Al, in the boiling hydrothermal fluid caused the brown cores to become unstable compared with the Fe-rich blue tourmaline rims.

The [4]Al/Si and [6]Al/Mg order-disorder behaviour of minerals in the tremolite-tschermakite solid solution (namely, end-member tschermakite and the 50:50 composition, magnesiohornblende) has been investigated by Monte Carlo simulation, using a model Hamiltonian in which atomic interaction parameters Ji were derived from empirical lattice energy calculations, and chemical potential terms μj (to express the preferences of cations for particular crystallographic sites) were derived from ab initio methods. The simulations performed were increasingly complex. Firstly, ordering in one tetrahedral double chain with Al:Si = 1:3 (tschermakite) was simulated. Although the low-temperature cation distribution in this system was ordered, there was no phase transition (due to the quasi-one-dimensional nature of the system). Next, interactions between tetrahedral Al:Si = 1:3 double chains were included, and a phase transition was observed, with the cation distribution in one double chain lining up with respect to that in the next. Finally, interactions between tetrahedral and octahedral sites were incorporated, to model the whole unit cell, and compositions corresponding to tschermakite and magnesiohornblende were investigated. The whole-cell simulation results compare favourably with experimental conclusions for magnesiohornblende, in that Al at T1 is preferred over Al at T2, and Al at M2 is favoured over that at M1 and M3, but the significant amount of Al at M1 is at odds with experimental observation.

The crystal structure of allanite from granitic pegmatite, the Daibosatsu Pass, Yamanashi, Japan, has been refined under the constraint of chemical composition determined by electron microprobe analysis of rare earth elements. Back-scattered-electron images and X-ray element maps of the allanites show that each of their crystal grains has chemically homogeneous distribution of major elements. A typical formula for the chemistry is: (Ca0.920☐0.080)Σ1.000(La0.238Ce0.443Pr0.048Nd0.100Sm0.019Th0.042Mn0.008☐0.102)Σ1.000(Al0.607Fe0.3173+Ti0.076)Σ1.000(Al1.000)(Fe0.5432+Fe0.3653+Mn0.055Mg0.037)Σ1.000(SiO4)(Si2O7)O(OH).

The crystal structure of allanite, monoclinic, a 8.905 (1), b 5.7606 (5), c 10.123 (1) Å, β 114.78°(1), space group P21/m, Z = 2, has been refined to an unweighted R factor of 3.46% for 1459 observed reflections. Although the H atom position was not determined on the Difference-Fourier map, inspection of the bond valence sums demonstrates that the H atom is uniquely located at the O10 atom and involved in a hydrogen bond to O4. A systematic examination as to crystal chemistry of allanites suggests that the isolated SiO4 tetrahedron has the largest distortion of three kinds of the tetrahedron containing Si2O7 groups in the allanite structure. This observation is common to the epidote group minerals, while the larger distortion of A2 sites caused by occupancy by REE in allanites contrasts with the smaller one of A sites in other epidote group minerals. In the allanite groups the bond angles between the O10–H bond and hydrogen bond H···O4 are found to range from 170 to 180°.

Compilation of the chemical compositions of the title allanite and the others from granitic rocks, Japan, which reveals Th-incorporation as the coupled substitution of 3Th4+ + ☐ (vacancy) ⇌ 4REE3+, provides an explanation for the observation that higher Th concentrations characterize allanites from the island arcs. The ternary Al2O3-Fe2O3-ΣREE diagram illustrates that allanites are grouped, according to their origins, into three classes suggestive of tectonic backgrounds for the crystallization localities; (1) intracontinental, (2) island arc and (3) continental margin.

In classical toxicology, speciation of carbon is taken for granted and the carbon compounds responsible for toxicity are always described with the appropriate chemical nomenclature. By contrast, speciation of other elements is largely ignored and elements other than carbon are often condemned as toxic because of evidence relating toxicity to only a few of the chemical species in which they occur. In some cases, such as chromium, the distinction between oxidation states (chromium III and chromium VI) may be recognized as important. The fact that chromium VI may only be a toxicological problem in certain well defined forms, such as specific chromate salts that may cause nasal and lung cancer, and then only by the respiratory route of exposure, is largely ignored. Failure to consider properly chemical speciation of elements other than carbon can lead to poor use of our resources. Laws and regulations based on simple elemental analysis may wrongly condemn environmental media or products as toxic and prevent the use of important resources.

Water plays an important role in most processes within the Earth's mantle, e.g. transport phenomena, differentiation and seismic properties. This paper reviews the various aspects of water in the mantle, with a special emphasis on the water content in nominally anhydrous minerals, especially lower-mantle minerals. The saturation of the upper mantle with respect to water is calculated as a function of pressure, based on available water-solubility data obtained for upper-mantle minerals. The result indicates that the upper mantle is saturated at pressures between 2 and 4 GPa for bulk water contents of 250 to 700 ppm wt., as retrieved from measurements on mid-ocean ridge basalts (MORBs) and ocean island basalts (OIBs). Whereas up to 4000 ppm wt. of H2O could be dissolved in the upper mantle at pressures corresponding to 410 km depth, such a value is less than the 1.5–2.5 wt.% solubility stored in the hydrous phases of the transition zone. Water solubility in mantle perovskite is still controversial, because of the difficulty of synthesizing samples free of impurities. Reported data indicate that water solubility in perovskite decreases with increasing temperature and Al content. Water partitions preferentially into ferropericlase rather than into perovskite, and its water solubility increases with the incorporation of trivalent cations.

We have analysed a rare occurrence of orange-brown manganotantalite lamellae (visible in hand specimen), intergrown with microlite [(Ca,Na)2(Ta,Nb)2(O,OH,F)7], aggregates of ferrotapiolite, bismuth minerals and apatite to understand more about the mechanisms of crystal growth and secondary modification in Ta-rich minerals. The intergrowth occurs within amblygonite/montebrasite nodules near the quartz core of the highly fractionated rare-metal Li/Be/Ta pegmatite at Rubicon, Karibib, Namibia. Electron microprobe analyses show that manganotantalite lamellae are variable in composition. Primary microlite (Ta2O5 82%, 1.97 Ta a.p.f.u.) forms the matrix mineral between the lamellae. Textural relations suggest an exsolution origin for the lamellae. Manganotantalite is represented by three generations: (1) primary late magmatic; (2) disequilibrium exsolution lamellae; and (3) subsolidus replacement. Crystallization commenced with primary microlite and likely simultaneous intergrowth between ferrotapiolite and a first generation of late-magmatic primary manganotantalite with low Ta (1.1—1.5 a.p.f.u.). On cooling this was followed by exsolution of manganotantalite lamellae, generation (2) with low—medium Ta (1.27—1.7 a.p.f.u.). The replacement of microlite by a highly fractionated late-stage melt rich in Mn2+, Ca2+ with low Na+ finally produces a third generation (3) of manganotantalite with high Ta (1.72—1.99 a.p.f.u.) at the contact with microlite. Native bismuth and bismutite cut across microlite and pseudomorph lamellae as a final hydrothermal replacement. Apatite is ubiquitous at the contact with amblygonite. The stability field of microlite may be extended by incorporation of CaTa2O6-rynersonite and Ca2Ta2O7 — idealized, components in solid solution. However, rynersonite-CaTa2O6 with distorted octahedra has some structural templates which are similar to the structure of pyrochlore (microlite). Hence, via the perovskite/pyrochlore analogy, hypothetical exsolution of manganotantalite-type structures may occur from a microlite (pyrochlore) host by solid-state diffusion via metastable rynersonite-type intermediates. Such a mechanism has the potential to explain the crystallographically controlled intergrowth textures and the compositional heterogeneity.

Basalts associated with sapphire deposits are widespread in Europe, Africa, Asia and Australia. In Vietnam, blue, green and yellow sapphires are recovered from eluvial and alluvial placers hosted in basaltic fields of the Dak Lak Province. Two distinct basalt suites are recognized in this field: a tholeiitic suite without any xenocrysts and an alkaline suite with mantle and lower crustal xenocrysts. The sapphires are enriched in Fe (0.43 to 1.26 wt.%), with moderate contents of Cr (33–1582 ppm), Ti (35–1080 ppm), Ga (149–308 ppm) and V (28–438 ppm) and they are poor in Zn and Mg. Their O-isotope compositions range from 6.0 to 6.9% and are not in equilibrium with the alkali basalts which have δ18O values between 5.0 and 5.7%. The U-Pb dating of zircons recovered from the basaltic placers provides evidence of two eruptional events: one at ~6.5 Ma followed by another one at ~1 Ma. The petrography of the basalts and the oxygen isotopic composition of the sapphires suggest that the sapphires are xenocrysts and that they crystallized in a deep magma chamber, at the lower continental lithosphere and the upper mantle boundary, in evolved melts issued from the fractionation of alkali basaltic magmas contaminated with lower crustal fluids.

In the large serpentinite body at Zöblitz, which is part of the Gneiss-Eclogite Unit of the Saxonian Erzgebirge, ilmenite rods have been detected in olivine. Although these rods are ≤1 μm wide, they can be unequivocally identified as ilmenite using electron backscatter diffraction. We also applied this method to prove that ilmenite is topotactically intergrown with the olivine host ([100]ol ‖ [001]ilm). This relation was found previously, e.g. for ilmenite exsolved from olivine of garnet peridotite from Alpe Arami, Swiss Alps. The degree of this exsolution phenomenon in the Alpe Arami rocks was taken as an indication of the original formation of a Ti-bearing mineral at depths of ∽300 km. As in the olivines of the serpentinite from Zöblitz, the density of the ilmenite rods is significantly less than in the rock from Alpe Arami. We think that our observation is compatible with previous P-T estimates of close to 4 GPa and 1000–1100°C for the ultramafic bodies in the central Saxonian Erzgebirge.

We report a new type of högbomite in a spinel- and sapphirine-bearing silica-deficient Mg-Al rock from Manavadi at the northern margin of the Madurai Block along the Palghat-Cauvery shear zone system, which defines the Archaean-Proterozoic collisional boundary in southern India. Högbomite occurs here as fine- to medium-grained euhedral to subhedral crystals enclosed within or along grain boundaries of greenish spinel. It is compositionally very unique due to the lack of TiO2 (<0.2 wt.%) and enrichment in Al2O3 (66.4—70.8 wt.%) compared to other occurrences. The variety of högbomite described in this study, confirmed by laser Raman spectroscopy, has not been reported elsewhere previously. Högbomite in this study is subdivided into three groups based on the XMg (=Mg/ (FeTotal+Mg)) ratio; 0.54—0.57, 0.61—0.73 and 0.77—0.79. Compositional data suggest a linear element substitution of 3(Fe+Mg) = 2Al for the Ti-free högbomite, a trend that is distinct from previously reported Ti-bearing högbomites. Textures and mineral chemistry data suggest that the following reaction took place at retrograde stage (T <800°C): spinel + H2O + O2 → högbomite + corundum + magnetite, implying that introduction of hydrous and oxidized fluid as well as low bulk-Ti content played an important role in the formation of Ti-free högbomite.

Changes in the lability and local environment of Cd sorbed onto calcite, hydroxylapatite and humic acid have been investigated as a function of time (30 min–9 months) using a combination of X-ray adsorption spectroscopy (XAS) and isotopic dilution techniques with Cd concentrations between 1 and 1000 mmol kg –1. Enhanced X-ray absorption fine structure (EXAFS) shows that precipitation of Cd carbonate was initially observed on calcite. Ageing of a 10 mmol kg–1 sample showed replacement of Cd in the fourth shell with Ca, indicating diffusion between the two phases had occurred. This diffusion appeared to be ongoing for a 10 mmol kg–1 sample even after 9 months of reaction. Only 20% of the Cd in a sample with 10 mmol kg–1 Cd remained labile after 6 months of reaction, suggesting a high degree of incorporation into the solid phase, in agreement with the XAS results. The XAS results for hydroxylapatite indicated that Cd is not fixed in the mineral lattice and Cd lability in a sample loaded with 10 mmol kg–1 Cd was ∼55% after 6 months of reaction. This indicates that a relatively high proportion of the Cd sorbed to hydroxylapatite remained as kinetically active surface species. Little change was observed in the local environment of Cd adsorbed on flocculated Ca-humate samples as a function of time and ∼75% of the Cd remained radio-labile after 6 months of reaction regardless of initial Cd loading. High lability is expected because incorporation into a solid matrix or surface precipitation of a Cd phase is less likely than with the mineral phases.

To understand the rates of turnover of soil carbon, and hence interactions between soil carbon pools and atmospheric CO2 levels, it is essential to be able to quantify and characterize soil organic matter and mineral hosts for C. Thermal analysis is uniquely suited to this task, as different C compounds decompose during a heating cycle at different temperatures. In ‘air’ (80% He or N2, 20% O2), relatively labile cellulosic material decomposes between 300 and 350°C and more refractory lignin and related materials decompose between 400 and 650°C. Calcite and other common soil carbonate minerals decompose at 750–900°C. Using thermal analysis connected to a quadrupole mass spectrometer and to an isotope ratio mass spectrometer, it is possible to simultaneously determine mass loss during combustion, evolved gas molecular compositions, and carbon isotope ratios for evolved CO2. As an example of the potential of the technique, the evolution of a fungally-degraded wheat straw shows initial isotopic heterogeneity consistent with its plant origins (–23.8% v-PDB for cellulosic material; –26.1% v-PDB for ligninic material), which homogenizes at heavier δ13C values (–21.0% v-PDB) as lignin is preferentially degraded by fungal growth. Simultaneously, it is shown that the evolution of nitrogen compounds is initially dominated by decomposition of aliphatic N within the cellulosic component, but that with increasing fungal degradation it is the ligninic component that contributes N to evolved gases, derived presumably from pyrrolic and related N groups produced during soil degradation through condensation reactions. Overall, the use of thermal analysis coupled to quadrupole and stable isotope mass spectrometry appears to have considerable potential for the characterization of discrete carbon pools that are amenable to the modelling of carbon turnover within soil systems.

The atomic structure of synthetic, deuterated goslarite (ZnSO4·7D2O), a = 11.8176(6) Å, b = 12.0755(7) Å, c = 6.8270(4)Å, space group P212121, Z = 4, has been refined in a combined neutron powder diffraction and X-ray single-crystal data refinement to wRp 1.92%, Rp 1.45% and R(F2) 12.66% for the neutron powder data contribution and R(F2) 8.72% for the X-ray single-crystal data contribution. Both data sets were necessary to achieve the best overall fit agreement in the Rietveld refinement and reasonable geometry within structural units. The results of this study confirm that the H-bonding scheme for goslarite is the same as that of the other epsomite group minerals. Small but significant variations of the Zn–O bond lengths can be attributed to details of the H bonds to the O atoms of the Zn octahedra. This investigation of the atomic structure and hydrogen bonding of goslarite is groundwork for future studies into phase relationships and the mechanisms of hydration and dehydration in the ZnSO4–H2O system.

Coexisting Ba-rich (Cn56–59Or40–42Ab2An0–1) and Ba-poor (Cn7–15Or83–92Ab1–3An0–1) feldspars were found in a jadeitite collected from the Itoigawa-Ohmi district in the Renge high-P/T-type metamorphic belt, Japan. There is an apparent compositional gap in these barian feldspars at Cn15–56 in the ternary system BaAl2Si2O8–KAlSi3O8–NaAlSi3O8. Barian feldspars occur as interstitial phases between subhedral to euhedral jadeite and prehnite, and sometimes cut a jadeite crystal. The occurrence of barian feldspars combined with the texture and mineral assemblages of the host jadeitite suggest that barian feldspars were formed from residual fluids after the formation of jadeite and prehnite from primary aqueous fluids at low-T conditions (<350°C) and pressure of ~0.6 GPa. The Ba content in the residual fluids would increase locally during the formation of jadeite because Ba is incompatible with clinopyroxene. This study supports the suggestion that Ba-rich minerals are not uncommon in jadeitites and albitites/metasomatized rocks that occur as tectonic blocks in serpentinite-matrix mélanges. Barium is an important minor element in metasomatizing fluids related to the formation of jadeitites and might have been derived from subducted oceanic crust at the beginning of the subduction.

Naujakasite (Na6(Fe,Mn)Al4Si8O26) is a characteristic, rock-forming mineral of hyperagpaitic naujakasite lujavrite (arfvedsonite + naujakasite + steenstrupine + albite ± relict nepheline ± microcline ± ussingite) from the Ilímaussaq complex, South Greenland, and is only known from this complex. The qualitative stability relationships of naujakasite-bearing, low-variance liquidus mineral assemblages at near-solidus conditions in lujavritic magma can be modelled within the six-component system Si-Al-Fe-Na-O-H. The modelling suggests that an uncommon combination of elevated water activity and low oxygen fugacity is needed to stabilize arfvedsonite + naujakasite mineral assemblages in highly sodic liquids. These conditions are incompatible with the stable coexistence of aegirine and arfvedsonite. They are met in the arfvedsonite-bearing lujavrites of Ilímaussaq, but apparently not in rocks of similar composition in the Khibina and Lovozero complexes, Kola Peninsula, where aegirine- or aegirine-arfvedsonite-bearing mineral assemblages are characteristic. Liquid lines of descent leading to crystallization of naujakasite have to pass through the albite+nepheline+arfvedsonite liquidus field in the Si-Al-Fe-Na-O-H system, reaching this field at the high-Na side of the albite-nepheline-arfvedsonite plane.

A gonnardite-thomsonite-chabazite-calcite assemblage forms a cement in the Foveaux Formation, a fossiliferous gabbroic boulder bed that accumulated at the base of a sea cliff cut in a Permian igneous complex during late Oligocene±early Miocene time. Gonnardite was the earliest zeolite to form, locally following minor calcite. It was followed epitaxially by thomsonite, co-precipitating with chabazite. Crystal habits indicate a low-temperature origin. The maximum temperature to which the deposit may have been subjected is estimated as not more than ∽30°C. The chabazites are Ca-poor chabazite-K and chabazite-Na. Representative electron microprobe analyses are as follows, all + nH2O:

• thomsonite: Na3.77Ca7.73(Al19.39Si20.65)O80 and Na3.78K0.04Ca7.25Mg0.05(Al19.13Si21.05)O80;

• gonnardite: Na6.95K0.03Ca4.73(Al16.99Si23.15)O80 and Na8.56K0.03Ca4.05(Al17.32Si22.84)O80;

• chabazite-K: Na1.18K1.72Ca0.08Mg0.23(Al3.51Si8.49)O24 and Na1.67K1.92Ca0.18Mg0.17(Al4.11Si7.85)O24;

• chabazite-Na: Na2.51K1.13Ca0.17Mg0.02(Al4.08Si7.93)O24.

Such a Si-poor zeolite assemblage is unusual for marine sediments and is attributed to precipitation from marine water impoverished in silica in the gabbroic boulder bed and interacting with shell material and calcic plagioclase. In contrast, a dioritic clast in the boulder bed provides an example of less silica-poor zeolites originally formed in the parent igneous complex. Veinlets in the clast contain scolecite averaging Na1.19Ca7.36(Al15.84Si24.14)O80.nH2O, and mesolite averaging Na5.13K0.03Ca5.24 (Al15.93Si24.13)O80.nH2O, in part as sub-microscopic intergrowths. The composition of scolecite closely corresponds to the most Na-rich scolecite reported hitherto.

Cornwall (UK) has suffered extensive arsenic contamination due to the historic mining and processing of mineral ores. Standard procedures for contaminated land risk assessment (DEFRA and Environment Agency, 2002a) are probably unworkable in Cornwall, with a very large number of sites classified as contaminated by arsenic. Methods of measuring the speciation and mobility of arsenic are essential for effective and rapid risk assessments of arsenic contamination.

Three clusters of lysimeters were installed in three different areas of an arsenic-contaminated Cornish site. A novel phosphoric acid microwave extraction technique was applied to the soils removed from the lysimeter holes; HPLC-HG-AFS analysis showed the majority of solid-phase arsenic to be arsenate (AsV). Pore waters obtained from the lysimeters showed variable, relatively low levels of arsenite (AsIII) and arsenate (AsV) to be present (<1–129 μg l–1). Less toxic arsenate predominated in most pore waters, with the presence of minor amounts of arsenite suggesting heterogeneous redox conditions. Pore-water arsenic concentrations were strongly positively related to solid-phase arsenate concentrations.

The use of techniques that assess the speciation of arsenic both in the solid and aqueous phases of a soil provides important information about the mobility of arsenic. The methodology presented in this paper may offer a novel basis for risk assessments of other contaminated sites.