A study of the distribution of REE in epidote-bearing metaluminous granitoids from Sierra de Chepes, Sierras Pampeanas, Argentina, reveals that a large proportion of the REE reside in the accessory minerals (allanite, epidote, titanite, apatite and zircon), and therefore these minerals control the behaviour of REE in granitic magmas. Well-developed chemical zonation in titanite indicates that the REE content decreases in the melt during crystallization of this mineral. The textural and chemical characteristics of euhedral epidote suggest a magmatic origin, and in that case it may have played an important role in the fractionation of the REE. The amount of silica and any other geochemical parameter indicative of fractionation progress in the dominant granodioritic-tonalitic facies (gtf) do not correlate with observed variations in the REE patterns. When many accessory minerals are involved, as in the gtf, the differentiated melts (e.g. aplites) are REE poor. Thus, the presence/absence of accessory minerals in granitoids can be indicative of the generation of differentiated melt enriched or poor in REE and other trace elements. This may have an economic significance, as it may allow us to predict the probable geochemistry of the differentiated melts (i.e. those that tend to develop mineralization) from the textural analysis of the ‘regional’ granitic rock.

Finally, the type and abundance of accessory minerals in the granitic suite can also help us to define the geotectonic environment where magmas were generated.

The Eonyang amethyst deposit is thought to be spatially and temporally associated with the biotite granite of the Cretaceous Kyongsang Basin, South Korea. The euhedral quartz crystals in cavities in the aplite which intrudes biotite granite are colour-zoned from white at the base to amethystine at the top. Fluid inclusions from rock-forming quartz in granitic rocks and euhedral quartz crystals in cavities were examined. Three types of primary inclusions were observed and three isochores for inclusions representing each type are constructed to constrain the trapping conditions and fluid evolution involved during the formation of the amethyst. The intersection of the isochore representing the early fluid inclusions with solidus temperature of the host granite indicates initial quartz formation at ~600°C and 1.0–1.5 kbar. Intermediate quartz formation, associated with the high-salinity inclusions, occurred at somewhat lower temperatures (400°C) and pressures of ~1 kbar. The amethystine quartz formed from H2O–CO2–NaCl fluids at temperatures between 280 and 400°C, and pressures of ~1 kbar. Based on the texture and mineralogy of host minerals and on the fluid inclusion characteristics, the euhedral quartz began growing at near solidus conditions of the granite and the pressure did not vary significantly until the end of crystallization of amethystine quartz crystals in cavities. Early quartz in cavities formed from moderately saline fluids that either exsolved from or were in equilibrium with the granite, whereas the amethystine quartz apparently grew from fluids of at least partial sedimentary origin. The granite crystallized at considerable depth under relatively low water pressures probably in the root zones of porphyry-type systems. Hydrothermal activities, fluid compositions and erosion factors combined to provide favourable conditions for the formation of the Eonyang amethyst deposit and its presence near the Earth's surface today.

Total neutron scattering measurements, analysed using a modification of the reverse Monte Carlo modelling method to account for long-range crystallographic order, have been used to describe the temperature-dependent behaviour of the structure of quartz. Two key observations are reported. First, the symmetry change associated with the displacive α–β phase transition is observed in both the long-range and short-range structural correlations. Secondly, some aspects of the structure, such as the Si–O bond length and the thermally-induced dynamic disorder, the latter of which sets in significantly below the transition, are relatively insensitive to the phase transition. These results are used to show that the α-domain model of the β-phase disorder is inappropriate and that the classical soft-mode picture of the phase transition is too simplistic. Instead, it is argued that the structural behaviour is best described in terms of its ability to respond to low-frequency, high-amplitude vibrational modes. This view is supported by additional single-crystal diffuse neutron scattering measurements.

Chemical composition and cell parameters were ascertained on new data from Ti-Nb-Ta complex oxides from the miarolitic pegmatites of the Baveno pink granite (Southern Alps, NW Italy). The crystals are tiny, single or aggregated in small sprays, prismatic or tabular, from yellow-orange to brownish in colour. Typical associated minerals include fluorite, zinnwaldite, gadolinite-group minerals and Sc-minerals. Cavity paragenesis is typical of NYF pegmatites, and shows two stages of crystallization developing in magmatic-pneumatolitic and hydrothermal conditions. X-ray data show that some oxides belong to the aeschynite mineral group; others are polycrase and fersmite. Aeschynite and polycrase are chemically heterogeneous and structurally disordered because of their metamict state. This disorder does not always seem to be related to radionuclide contents. Two main trends are indicated, considering the behaviour of Y. The high Y contents fit with very low Ca and LREE contents in the A site; the HREE contents follow the Y trend. In the B site, Ti is the dominant cation, followed by some Nb and very little Ta. Small quantities of Y fit with increasing Ca, U, Th and REE contents. In the B site, Nb cations often exceed those of Ti. The Th contents are often greater than those of U. Besides the already known aeschynite-(Y) and vigezzite, new varieties, ‘titano-vigezzite’ and ‘niobo-aeschynite-(Y)’, are identified here in the Baveno miarolitic cavities. Samples 14 (analysis a) and 2502 (analysis b) have Ca as the main occupant of the A site, followed by Y, Th and REE; in the B site, Ti prevails over Nb. These compositions cannot be considered as pure vigezzite, but as a new variety called ‘titano-vigezzite’. In the same way, analysis a of sample 3 may be considered a new variety of aeschynite-(Y), with Nb prevailing over Ti in the B site, and here called ‘niobo-aeschynite-(Y)’. Neither variety has ever been mentioned in the literature. Epitaxial growth of aeschynite on polycrase (sample 3194) allows some inferences on the crystallization sequence.

The crystal structure of hydrothermally synthesized Rb-feldspar (monoclinic, space group C2/m, a= 8.839(2)Å, b= 13.035(2)Å, c= 7.175(2)Å, β = 116.11(1)8, V= 742.3(3)Å3, Z= 4) has been refined to a final R of 0.0574 for 692 independent X-ray reflections. Microprobe analyses of the Rb-feldspar suggest deviation from stoichiometry, with excess Si and Al, resulting in a unit formula of Rb0.811□0.127Al1.059Si3.003O8. Infrared (IR) spectra indicate the structural occupancy of large H2O content, which implies that the □Si4O8 substitution favours the structural incorporation of the H2O molecule at the M-site. The mean T–O distances are 1.632 Å for T1 and 1.645 Å for T2, revealing highly disordered (Al,Si) distribution with Al/Si = 0.245/0.755 (T1 site) and 0.255/0.745 (T2 site).

There are two geochemical implications from this refinement: (1) identification of both rubicline triclinic with (Al,Si) ordered distribution and synthetic monoclinic RbAlSi3O8 with (Al,Si) disordered distribution implies that Rb cannot be one of factors disrupting the (Al,Si) ordered and disordered distributions in feldspars; and (2) natural and synthetic feldspars capable of accommodating the large cations tend to incorporate □Si4O8, excess Al and H2O components in their crystal structures.

Crystal growth experiments in the system BaO–Al2O3 yielded single crystals of a new Ba oxo-aluminate with composition Ba4Al2O7. The title compound is orthorhombic, with cell parameters a = 11.3126(5), b = 11.7045(9), c = 27.1850(14) Å, space group Cmca. The structure was solved by direct methods followed by difference Fourier synthesis from X-ray single crystal diffraction data (R1 = 0.048 for 1213 independent observed reflections and 135 parameters). The main building units of tetrabarium aluminate are Al3O10 trimers and isolated AlO4 tetrahedra. Seven crystallographically different Ba sites crosslink between the tetrahedral groups and are co-ordinated by six to nine oxygen ligands. The structure is closely related to the mineral kilchoanite (Ca6[SiO4][Si3O10]) and the first representative of an unbranched oligo-aluminate with triple tetrahedra.

The alteration mineralogy of two outcrops of basaltic lavas in northeast Ireland is studied with respect to the hydrothermal alteration history of basalts. Evidence for contact between basaltic flows and palaeorivers or palaeolakes is reported in both outcrops, which belong to two different formations: the Lower Basaltic Formation (LBF) and the Inter Basaltic Formation (Causeway Tholeiite Member, CTM). The secondary minerals consist of layer silicates (chlorite-smectite, C-S, di- and trioctahedral smectites), zeolites, calcite and silica minerals and there is a repetition of the sequence of alteration mineralogy in each formation. The mineralogy of phyllosilicates from the LBF (C-S) corresponds to hotter conditions of alteration than those of CTM (di/trioctahedral smectites), while the zeolite mineralogy is controlled by the fluid composition. In consequence, the mineralogy and the distribution of alteration minerals suggest that the crystallization of phyllosilicates and zeolites did not develop during burial of lavas a long time after the eruption, but reflects thermal gradients driven by lava heat flow and marked by local perturbation caused by regional water flow through the aquifer zone in the lava during the last cooling of each lava sequence.