The distribution of Mg and Ni over the octahedrally-coordinated sites M(l,2,3) in Ni-substituted potassic richterite solid solution, AKB(NaCa)c(Mg2.5Ni2.5)Si8O22(OH)2, synthesized at 0.1 GPa/750°C (25 days), has been studied in situ to 700°C by neutron powder diffraction with Rietveld structure refinement. Using a 2.7 g sample it was possible to make short data collections (3 h) at each temperature at intervals of 50° from 50 to700°C and locate the onset of cation exchange. Above 700°C the amphibole decomposes rapidly and so only an on-heating dataset was collected. Unit-cell parameters increase smoothly upon heating, with no discontinuities evident. Site occupancies of M(l), M(2) and M(3) sites were refined from site-scattering values. The initial XNi values of M(l), M(2) and M(3) sites, corresponding to the synthesis temperature of 750°C, are 0.57(1), 0.34(1) and 0.68(2), respectively. Above 400°C, there is initial ordering whereby thermal annealing allows the disordered Mg-Ni distribution of the quenched synthesis product to adjust to be closer to those appropriate for lower temperatures of the heating sequence. This ordering involves the exchange of Mg and Ni between M(l) and M(2), with no significant change in M(3) occupancy. At 700°C, XNi values of the M(l), M(2) and M(3) sites are 0.64(1), 0.27(1) and 0.69(2), respectively. No disordering was observed, due to the short duration of the high-temperature data collections and decomposition at 750°C. The results of this study indicate that divalent cations exchange between octahedral sites in amphiboles in a matter of hours.

The crystal structure of the mineral uytenbogaardtite, a rare silver-gold sulfide, was solved using intensity data collected for a crystal from the type locality, the Comstock lode, Storey County, Nevada (USA). The study revealed that the structure is trigonal, space group R3̄c, with cell parameters a = 13.6952(5), c = 17.0912(8) Å and V = 2776.1(2) Å3. The refinement of an anisotropic model led to an R index of 0.0140 for 1099 independent reflections. The structure consists of a sub-lattice of sulfur atoms forming a distorted body-centred cubic arrangement. The structure contains distinct tri-atomic linear groups (S–Au–S) and Ag atoms bonded to four S atoms (from four different linear groups) in a distorted tetrahedral arrangement. On the basis of information gained from this characterization, uytenbogaardtite is here definitively proved to be structurally different from petzite, Ag3AuTe2 and fischesserite, Ag3AuSe2. By use of high-quality single-crystal diffraction data, the symmetry of the mineral was found to be trigonal, and not tetragonal as erroneously supposed. A revaluation of the powder diffraction data listed in the scientific literature for uytenbogaardtite according to the structural results obtained here leads to an excellent agreement. Crystal-chemical features of uytenbogaardtite, Au2S, petrovskaite AgAuS, uytenbogaardtite–fischesserite series Ag3Au(S2–xSex) and acanthite–naummanite series Ag2(S1–xSex) are compared.

The migration and accumulation of immiscible silicate liquids may play a significant role in the differentiation of crustal magma bodies and the formation of some economic mineral deposits. However, our understanding of the processes that control the segregation of these liquids is currently limited by the short timescales of petrological experiments. Detailed microstructural investigations of Palaeogene basaltic dykes from Northeast England, coupled with simple 1D thermal models, constrain the effects of cooling rate on the microstructure of unmixed immiscible silicate liquids under natural conditions. The size of unmixed Fe-rich droplets within a continuous silicic phase is related to the cooling rate by a power law, with droplet diameter increasing with decreasing cooling rate, accompanied by an increase in the number of droplets. Fe-rich droplet coarsening is a result of diffusion-controlled growth. The average apparent aspect ratio and grain size of matrix plagioclase crystals indicate that nucleation and growth of these grains probably occurred in a static (or only weakly convecting) fluid dynamical regime.

Dargaite, ideally BaCa12(SiO4)4(SO4)2O3, is an additional member of the arctite group belonging to minerals with a modular intercalated antiperovskite structure derived from hatrurite. The holotype specimen was found at a small outcrop of larnite pseudoconglomerates in the Judean Mts, West Bank, Palestinian Autonomy. Larnite, fluorellestadite–fluorapatite, brownmillerite, fluormayenite–fluorkyuygenite and ye'elimite are the main minerals of the holotype specimen; ternesite, shulamitite and periclase are noted rarely. Dargaite, nabimusaite and gazeevite occur in linear zones with higher porosity within larnite rocks. Pores are filled with ettringite and Ca-hydrosilicates, less commonly with gibbsite, brucite, baryte, katoite and calciolangbeinite. Dargaite is colourless, transparent with a white streak and has a vitreous lustre. It exhibits pronounced parting and imperfect cleavage along (001). Mohs’ hardness is ~4.5–5.5. The empirical formula is (Ba0.72K0.24Na0.04)Σ1(Ca11.95Mg0.04Na0.01)Σ12([SiO4]0.91 [PO4]0.05[AlO4]0.03[Ti4+O4]0.01)Σ4([SO4]0.84[PO4]0.14[CO3]0.02)Σ2(O2.54F0.46)Σ3. Dargaite is trigonal R$\overline 3 $m, the unit-cell parameters are: a = 7.1874(4) Å, c = 41.292(3) Å, V = 1847.32(19) Å3 and Z = 3. The crystal structure of dargaite was refined from X-ray single-crystal data to R1 = 3.79%. The calculated density is 3.235 g cm–3. The following main Raman bands are distinguished on the holotype dargaite (cm–1): 122, 263, 323, 464, 523, 563, 641 and 644, 829 and 869, 947, 991 and 1116. The formation conditions of dargaite are linked to the local occurrence of pyrometamorphic by-products (gases, fluids and melts) transforming earlier mineral associations at ~900°C.

The crystal structure of nikmelnikovite, Ca12Fe2+Fe3+3Al3(SiO4)6(OH)20, a new member of the garnet supergroup from Kovdor massif, Kola Peninsula, Russia (R$\bar{3}$, a = 17.2072(6), c = 10.5689(4) Å, V = 2710.1(2) Å3 and Z = 3) has been refined to R1 = 0.046 on the basis of 1184 unique observed reflections. Nikmelnikovite is the first mineral species in the garnet supergroup that has a trigonal (rhombohedral) symmetry. The relationship between its unit cell and the pseudocubic (ideal garnet) unit cell can be described by the transformation matrix [1$\bar{1}$0 | 01$\bar{1}$ | ½½½]. The crystal-chemical relations between the ideal Ia$\bar{3}$d garnet and the nikmelnikovite structure type can be described by the following series of imaginary modifications: (1) the symmetry is lowered according to the Ia$\bar{3}$d → R$\bar{3}$ group–subgroup relationship; (2) the cation sites are split according to the following sequences: X → {X1, X2}; Y → {Y1, Y2, Y3, Y4}; Z → {Z1, Z2}; (3) the X sites remain fully occupied by Ca; (4) each Y site is occupied predominantly by a distinct chemical species: Y1 → Al (Al site), Y2 → Fe2+ (Fe1 site), Y3 → Fe3+ (Fe2 site), Y4 → vacancy (Mn site); (5) one of the Z sites (Z1) is occupied by Si, whereas the other site (Z2) is predominantly vacant. The crystal-chemical formula that takes into account the transition between the archetype and the nikmelnikovite structure type can be described as X{Ca12}Y[Fe2+Al4Fe3+2□]Z(Si6□6)O24(OH)20□4. The structural complexity of nikmelnikovite (4.529 bit/atom and 434.431 bit/cell, after H-correction) is higher than those for andradite, grossular and katoite, which is typical for low-temperature minerals formed after primary minerals with simpler structures.

Hectorite was synthesized from a Eu(III)-bearing brucite precursor in a multistep procedure. In a separate experiment, Eu(III) ions were adsorbed onto hectorite in suspension. Colloids were extracted from both samples. The size distributions in the colloidal fractions were characterizedby application of the asymmetrical flow field-flow fractionation (AsFlFFF) method and the corresponding elemental compositions were obtained by ICP-MS. Extended X-ray absorption fine structure (EXAFS) spectroscopy was used to characterize the local chemical environment surrounding Eu in thebulk samples and in the colloidal fractions.

The EXAFS results show that Eu is associated with hectorite upon co-precipitation or adsorption. Results from AsFlFFF suggest that Eu is structurally associated with the colloidal fraction extracted from bulk Eu-bearing co-precipitated hectorite.The AsFlFFF data are different for the colloidal fraction containing Eu(III) adsorbed on hectorite; in this sample they are consistent with a surface retention mechanism. These small but significant differences enable surface sorbed Eu to be distinguished from co-precipitated Eu. Eu is verylikely located in a clay-like environment in the co-precipitation experiment, and it forms inner-sphere surface complexes in the adsorption experiment. The results obtained using the different experimental techniques agree, and show the benefits of using multiple methods of analysis.

Trivalent europium was used as non-radioactive chemical homologue for trivalent actinides. Similar retention mechanisms are expected for the trivalent actinides if they are co-precipitating with or adsorbing onto sheet silicates. The present study provides information which can be usefullyadded to the safety assessments required for deeply buried nuclear waste disposal sites.

Compreignacite, a naturally occurring potassium uranyl oxide hydrate, is an expected alteration product of spent nuclear fuel subjected to oxidative corrosion in the presence of water. Ion-exchange experiments were performed using natural crystals of compreignacite in a 2 M CsCl solution at 180°C for 24 h, and in a 100 ppm CsCl solution at 90°C for 14 days. Exchange of Cs into crystals of compreignacite was demonstrated by crystal-structure analysis for a crystal from 2 M solution and chemical analysis for crystals from both exchange experiments. The structure of Cs-exchanged compreignacite is hexagonal, space group P2m, a= 14.1014 (18) Å, c = 15.127(3) Å, V = 2605.1(7) Å2. It was solved by direct methods and refined on the basis of F2 for all unique reflections to a final R1 = 6.37%. The structure determination demonstrated almost complete exchange of Cs for K in the interlayer of the structure, as confirmed by compositional analysis of the crystals, and provided the formula Cs6[(UO2)12(OH)10O10](H2O)3.5, Z= 2. The structure of Cs-exchanged compreignacite contains α-U3O8-type sheets of uranyl pentagonal bipyramids that are topologically identical, although compositionally distinct, to those in compreignacite. Compreignacite crystals placed in 100 ppm CsCl solution also incorporated substantial Cs by ion exchange. Ion exchange experiments, using Cs-exchanged compreignacite from earlier experiments, in 2 M KCl solution at 90°C for 14 days showed that compreignacite will retain significant Cs in the presence of a solution rich in K. These experiments indicate that formation of compreignacite structure-type phases in a geological repository due to alteration of nuclear waste may significantly impact upon the mobility of Cs, either by direct incorporation of Cs into growing crystals, or by exchange of Cs into earlier-formed crystals of compreignacite when they contact Cs-bearing solutions.

The crystal structure of the sheet oxychloride mineral parkinsonite, nominally Pb7MoO9Cl2, has been determined for synthetic and natural crystals of analysed compositions, (Pb7.28Mo0.72) O8.96Cl1.96 and (Pb7.23Mo0.40V0.37)O8.90Cl1.82, respectively. Parkinsonite is tetragonal, space group I4/mmm. Unit-cell parameters for synthetic and natural crystals are: asynthetic = 3.9773(4) Å, csynthetic = 22.718(4) Å, Vsynthetic = 359.38(5) Å3, and anatural = 3.9570(3) Å, cnatural = 22.634(5) Å, Vnatural = 354.40(5) Å3. Final agreement indices (R1, wR2) for refinements of the two crystals are 0.024, 0.067 (synthetic) and 0.036, 0.078 (natural). Although a superlattice has been identified by electron diffraction for crystals of both samples (Welch et al., 1996), only the substructure could be determined by X-ray diffraction. This X-ray invisibility of the superstructure has also been observed for the closely related sheet oxychlorides asisite and schwartzembergite, for both of which superstructure motifs have been identified by electron diffraction. The Pb(1) site of both parkinsonite crystals is fully occupied by Pb. Refinement of the Pb content of the Pb(2) site for the synthetic and natural crystals gives occupancies of 0.85(1) and 0.70(1) respectively, corresponding to 3.40 and 2.80 Pb(2) a.p.f.u. respectively. The substituent cation Mo (synthetic crystal) and [Mo+V] (natural crystal) was located at a distance of 0.5 Å from Pb(2), being displaced along the fourfold axis. The reduced occupancy of Pb(2) is due to substitution by Mo or [Mo+V]. No evidence for separate Mo and V sites in the substructure of natural parkinsonite was found. Refined occupancies of the Cl site are 0.84(4) and 0.91(5) for the synthetic and natural crystals, respectively, and are consistent with the 9:1 superstructure component identified by electron diffraction.

The phase equilibrium diagrams of CaMgSi2O6-CaTiAl2O6 with 10, 20, and 30 wt % Ca2MgSi2O7 in the join CaMgSi2O6-Ca2MgSi2O7-CaTiAl2O6 were determined. Six three-phase assemblages, six four-phase assemblages, and two five-phase assemblages (isobaric trivariant, divariant, and univariant) were confirmed. The univariant assemblages are diopsidess+forsterite+åkermanitess + perovskite + liquid and diopsidess+åkermanitess+perovskite+spinel + liquid. Because of the complex solid solutions of diopside and åkermanite the crystallization ceases before invariant assemblages are reached and the final phase assemblage in the diopside-rich region is diopsidess+åkermanitess+perovskite, in the diopside-poor region diopsidess+åkermanitess+perovskite + spinel. Two invariant points, diopsidess+forsterite+åkermanitess+perovskite+spinel+liquid and diopsidess+åkermanitess+anorthite+perovskite+spinel+liquid in the silica-poor region of the system CaO-MgO-Al2O3-TiO2-SiO2 are estimated. The diopside solid solution in this join is discussed with its bearing on the natural titanpyroxenes from alkalic rocks including melilite.

Textures and relict mineralogy in the Dalradian gneisses from the southern contact aureole of the Glen Doll diorite trace the development of three stages of metamorphism through regional and contact metamorphic phases. Regional metamorphic stage I is characterized by the stability of sillimanite+muscovite; recognition of a subsequent regional metamorphic stage II of lower grade is based on textural criteria, the stability of kyanite and staurolite + quartz, and geothermometry/geobarometry. The breakdown of zincian staurolite occurred under the conditions of contact metamorphism (stage III). Textural evidence from the outer part of the contact aureole suggests that zincian staurolite broke down by the following oxidation reaction:

zincian staurolite + muscovite + quartz + O2 → andalusite + Zn-rich spinel + magnetite + biotite + H2O.

Various stages of completion of this reaction have been observed in different parts of a sample. Predominance of magnetite over Zn-rich spinel (⩽ 14 wt. % ZnO) as a breakdown product can be explained by the initial breakdown of Fe-staurolite component + muscovite + quartz+O2 to form andalusite+biotite+magnetite. These product phases were joined by the Zn-rich spinel when sufficient Zn-staurolite component had concentrated in the unreacted staurolite. Rare local examples where Zn-rich spinel is dominant over magnetite may reflect lower O2 fugacity and/or higher initial Zn contents of the staurolite.

Petrographic, microscopic, and X-ray investigations of a uranium occurrence near Tiraun, Graubünden, Switzerland, are described. The uranium mineralization occurs in greenschist facies metamorphic rocks of the Tavetsch massif and consists essentially of uraninite of various habits with minor chalcopyrite, linnaeite, millerite, galena, marcasite, pyrite, hematite, magnetite, and skutterudite. The ore structure, characterized by concentrations along foliation planes, boudinage, and rotated (helicitic) textures of uraninite porphyroblasts, is indicative of a synmetamorphic (Alpine) remobilization, recrystallization and enrichment of the ore, which was probably derived from an original ‘protore’ in psephites and psammites (? sandstone-type mineralization) of at least Variscan age. The synmetamorphic emplacement of the ore took place under medium- to high-temperature conditions (approx. 350–400°C), as demonstrated by the lattice constants of the uraninite phases and by the ore paragenesis. Late to postmetamorphic tectonism resulted in a partial remobilization (over cm distances) and redistribution of the ore along veinlets that cut the foliation planes. The characteristics of the ore from Tiraun are similar to those from the synmetamorphic ore from Preit, northern Italy, as described by Cevales (1961).

The distribution of uranium in a suite of variably deformed and metamorphosed rocks from the leucocratic member of the Glendessarry syenite has been determined using the fission track method. The uranium content of the magma increased during crystallization and uranium was concentrated in accessory minerals such as monazite, zircon, sphene, allanite, apatite, and microinclusions of a Zr- and Ti-rich phase. Contamination of the magma by pelitic metasediment enhanced the uranium content and monazite and zircon formed instead of sphene, allanite, and apatite.

Evidence of subsolidus uranium mobility in late stage magmatic or metamorphic fluids is presented here and shows: (a) Intracrystalline redistribution of uranium, especially in grains of sphene. (b) Intergranular mobility in a fluid phase, which affected the uraniferous accessory minerals in several ways.

The angular relationships among sets of olivine plates in barred chondrules were investigated on the universal stage to test a suggestion that such chondrules represent exsolution and inversion from a pre-existing jadeitic pyroxene. In one of sixty-seven cases studied, adjacent plates represent olivine twinning on (100); the others vary randomly and correspond to neither olivine nor pyroxene twinning. On this basis and from other evidence, we conclude that the mineralogy and texture of barred chondrules are primary and due to rapid crystallization of ultrabasic magma in a low-pressure environment.

We have studied the structures of Al/Si disordered leucite and bicchulite with a quantum mechanical version of the virtual crystal approximation, VCA. In leucite, the average tetrahedron has a composition of (AlSi2)O4, while bicchulite represents an extreme case with (Al2Si)O4 tetrahedra. Both structures are well described with the VCA. In conjunction with an earlier study, where we have shown that the (AlSi)O4 tetrahedra in gehlenite and Al,Si-disordered octahedra are also well reproduced, we have now established that the VCA gives a reliable description of the averaged structure of disordered aluminosilicates over the whole compositional range. The current calculations confirm that Al/Si ordering is not driving the cubic to tetragonal phase transformation in leucite. In bicchulite, the model calculations are consistent with hydrogen on Wyckoff position 8c, in agreement with the result of a single crystal X-ray diffraction study, but in variance with results based on a neutron powder diffraction study.

Armalcolite and pseudobrookite have been crystallized from synthetic melts of basaltic composition with variable amounts of added TiO2. It was found that the composition of the armalcolite and pseudobrookite are dependent on the TiO2 content of the melt and the conditions (particularly oxygen fugacity, but also temperature and quenching rate) of formation. The maximum Al2O3 content of pseudobrookite appears to be ∼14–15 wt.% Al2O3, a spinel-like phase being exsolved at higher Al2O3 concentrations. Generally armalcolite is formed at low pO2 and pseudobrookite at higher oxygen fugacities.

Hydroxylian pseudorutile in an adamellite from Cameroon has been identified using EPMA, FT-IR and single-crystal XRD analysis. TiO2 and Fe2O3 contents range between 60.6 and 67.9 wt.% and 22.2 and 36.5 wt.%, respectively, and totals vary from 90.1 to 98.1 wt.%. The SiO2 (0.03 —1.40 wt.%), Al2O3 (0.02—0.68 wt.%) and CaO (0.03—0.44 wt.%) contents increase systematically and MnO (0.04—0.69 wt.%) decreases slightly with decreasing Fe3+/Ti values. The MgO content is <0.06 wt.%. H2O contents (1.3—7.5 wt.%) are greater in parts of crystals with lower Fe3+/Ti, suggesting solid solution between the end-members Fe3+2Ti3O9 and Fe3+Ti3O6(OH)3. The lattice parameters are a = 2.85 Å and c = 4.57 Å with hexagonal symmetry and point group 6/mmm.

Danalite is the Fe2+ end-member of the minerals of the helvite group, which have the general formula M8(BeSiO4)6S2, with M = (Mn,Fe2+,Zn). These minerals are relatively uncommon, although limited amounts are known at many localities round the world (e.g. Ragu, 1994; Larsen, 1988; Kwak and Jackson, 1986, and references therein). Their typical host rocks are skams, but occurrences in mineralized veins pegmatites, and altered alkali granites have also been reported. Danalite is rarer than helvite (Mn endmember) and genthelvite (Zn end-member) and, unlike the other two end-members, has never been found nor synthesized as a pure mineral (Mel'nikov et al., 1968).