Zoned hibonites from a kalsilite, leucite, spinel, corundum, perovskite gneiss from the southern Indian granulite terrain near Punalur, Kerala, have rims that are the most Ti-rich yet recorded (0.83–0.87 Ti atoms per 19 O) and are essentially free of REE elements (ΣREE < 0.01 atoms per 19 O) while the cores are the most REE-rich compositions yet recorded (ΣREE = 0.55–0.65 atoms per 19 O). Within the limits of analytical uncertainty, the compositions of the hibonite can be related to the theoretical end-member CaAl12O19 by the substitutions REE R2+ ⇌ CaAl and R2+R4+ ⇌ Al2 with the REE-rich cores containing in excess of 50% of the REE R2+ Al11O19 end member. Minor substitution of Na for Ca occurs in the rims, while non-stoichiometry in both the cores and rims is indicated by partial 12-fold site occupancy. Ion-microprobe analysis of the REE-rich hibonites reveals strong enrichment in LREE with La/Lu c. 250 000.

The application of ion microprobe analysis to REE determination is discussed, with special reference to REE-bearing accessory minerals (e.g. allanite, monazite, apatite, xenotime, sphene). The main analytical problems are shown to be (a) interferences in the mass spectrum caused by molecular ions, and (b) matrix effects, i.e. variations in ion yield with sample composition. Interference suppression using either high mass resolution or secondary-ion energy discrimination is fairly effective, but entails a significant sacrifice in peak intensity. In the absence of a viable model for predicting matrix effects, it is proposed that empirical standardization should be used. Often better absolute accuracy can be obtained by combining ion microprobe and electron microprobe data for REE present in sufficiently high concentrations. Detection limits for REE are in the part per million region at present, with spatial resolution typically around 10 μm, but there is considerable scope for improvement. Better standards and improved understanding of matrix effects should also lead to greater accuracy in quantitative analysis.

A metallic mass brought to the Western Australian Museum from the Wongan Hills district N.W. of Perth has been identified as an iron meteorite of unique type. It has graphite inclusions about I mm across distributed throughout the metal giving a ‘raisin bread’ appearance. Its nickel content (6·65 %) is comparable with that of coarse octahedrites but the kamacite grain structure is anomalous. Its gallium, germanium, and nickel contents place it close to, but outside, Wasson's chemical group IIb. Taenite is absent and troilite is rare. Neumann bands in the kamacite are distorted and the kamacite has flowed around large schreibersite inclusions. The latter have an exceptionally low nickel content (7·0 %) and probably formed at an unusually high temperature. The kamacite contains more phosphorus than normal iron meteorites, and small schreibersite grains in the kamacite are relatively nickel-poor (22 %). The unusual structure of this iron is thought to be due to one or more of the factors high carbon, high phosphorus, and relatively rapid cooling.

Olivines from igneous rocks generally have low calcium contents. However, olivines crystallized from strongly silica-undersaturated magmas occasionally contain large amounts of kirschsteinite (Ki" CaFeSiO4) and monticellite (Mo: CaMgSiO4) components. The purpose of this note is to report the first natural occurrence of kirschsteinite crystals that have lamite (La: CazSiO4) in the mineralogical norm.

In ion microprobe analysis the specimen is bombarded with a focussed ion beam a few µm in diameter and the secondary ions produced are accelerated into the entrance slit of a mass spectrometer. An outline of the salient features of the instrument is given here, together with an account of the methods used for quantitative elemental and isotopic analysis.

The major part of this paper consists of a comprehensive account of the geological applications of ion microprobe analysis. These include elemental analysis, especially for trace elements (down to sub-ppm levels in many cases) and light elements (H-F) which are beyond the scope of the electron microprobe. The other main area of geological interest is isotopic analysis, where the ion microprobe has the advantage over conventional mass spectrometry of being capable of in situ analysis of selected points on polished sections, obviating the need for laborious specimen preparation, and enabling spatially-resolved data to be obtained, with a resolution of a few µm. The ion microprobe has been especially successful in U-Pb zircon dating and the study of isotope anomalies in meteorites. Other significant applications include diffusion and stable isotope studies.

Electron-probe microanalysis (EPMA) is applicable to rare-earth elements (REE) in minerals with relatively high REE concentrations (e.g. hundreds of parts per million). However, given that each of the 14 REE has at least 12 X-ray lines in the L spectrum, finding peak-free regions for background measurement can be problematical. Also, measured peak intensities are liable to require correction for interferences. Hitherto, little attention has been paid to the optimisation of background offsets and the implications of the wide variation in REE distribution patterns in different minerals. The ‘Virtual WDS’ program, which enables complex multi-element spectra to be synthesised, has been used to refine the conditions used for different REE distributions. Choices include whether to use the Lβ1 rather than the Lα1 line, background offsets, and counting times for comparable relative precision. Correction factors for interferences affecting peak and background measurements have also been derived.

Spinel lherzolite xenoliths from two localities in the Massif Central are undepleted in Al2O3, CaO, and Na2O. One suite from Tarreyres, is K2O depleted and amphibole-bearing whereas the other, from Monistrol d'Allier some 18 km away, is amphibole-free and has a higher mean K2O content of 0.035 wt.%. We present bulk major and minor element abundances in a harzburgite and a lherzolite from each locality and microprobe analyses of their constituent phases. Amphibole-bearing lherzolite and its pyroxenes are light-rare earth element (LREE) depleted, whereas amphibole-free lherzolite and its pyroxenes are LREE enriched. Both harzburgites and their pyroxenes are LREE enriched and one rock contains LREE enriched glass. The harzburgites are like harzburgite xenoliths from elsewhere but each lherzolite represents a previously unrecognized type of mantle in terms of the mineralogy and REE content. The implication for basalt genesis are briefly discussed.

Agates from a 430 Ma host at Stockdale Beck, Cumbria, England have been characterized. The crystallite size of the Stockdale Beck agates was found to be ~60% greater than any other agates from five regions aged 400–1100 Ma. Raman spectroscopy identified moganite in all agates tested except those from Stockdale Beck. Infrared spectroscopy showed that the silanol content of the Stockdale Beck agates was near zero. The properties of agates from Stockdale Beck and the 1.84–3.48 Ga metamorphosed hosts found in Western Australia were similar but different from agates found in other hosts aged 400–1100 Ma. Cathodoluminescence demonstrates further differences between agates from hosts aged 13–1100 Ma and those from Stockdale Beck and Western Australia. Agates from the latter areas have a lower proportion of defects causing a red emission band (~660 nm) but an increased proportion of defects causing blue (~470 nm) and orange (~640 nm) emission bands. Agates found in hosts aged 13–1100 Ma are also differentiated from the Stockdale Beck and Western Australian agates in a ternary plot of the relative intensities of violet to blue to orange emission bands. Single scans producing this combination of colours are only found in the Stockdale Beck and Western Australian agates. The properties shown by the Stockdale Beck and Western Australian agates demonstrate that an agate or chalcedony infill can be used to identify post-deposition palaeoheating within a host rock.

Chalcedony and agates from a variety of world-wide hosts have been examined using cathodoluminescence (CL). Gaussian fitting of the experimental data shows that there are two dominant spectral emissions at ∼400 and ∼660 nm. A third subordinate peak is also found at ∼470, ∼560 or ∼620 nm. An age-related link is shown between the respective decreasing and increasing relative intensities of the 660 and 620 nm emissions. It is proposed that this change is due to a condensation reaction between neighbouring Si–OH groups eliminating water and forming a strained Si-O-Si bond.

Agates from a variety of hosts and regions produced no clear demonstrable CL distinctions. However, a set of Western Australian agates was examined from host rocks that had been subjected to burial metamorphism. Cathodoluminescence produced different spectral emissions in the petrographic fibrous and granular regions of these agates. One agate shows a partial transformation of the petrographic fibrosity into granularity. This conversion is characterized by emission bands at 570 nm and 460 nm. Similar emission-band changes were produced by heating Brazilian agates for 35 days at 300°C. The identification of these changes in agate could serve as an indicator of palaeoheating within the parent rock.

Characterization of Brazilian agates containing a lower horizontally banded section and an upper chamber with bands parallel to the walls shows that these agates formed much later than the 135 Ma Paraná basalt host rock. Age differentiation between the two types of banding shows that the horizontal bands formed between 43 to 63 Ma ago with a final infill of wall-lining bands between 7 and 27 Ma later. The horizontal bands have a higher Al3+ concentration and a greater crystallite size than the wall-lining layers; they have a lower mogánite content and defect-site water content. The formation of these agates appears to be the result of a three-stage process. After the separate formation of horizontally banded and wall-lining agate, a silica infill seals the gap between the agate and the cavity wall. The detection of cristobalite in some specimens indicates that genesis of both the horizontally banded and wall-lining deposits in the Brazilian samples proceeds along an amorphous silica → opal-CT → opal-C → chalcedony pathway.

Extended abstract of a paper presented at Microscopy and Microanalysis 2012 in Phoenix, Arizona, USA, July 29 – August 2, 2012.