An unusual highly hydrated and Na-depleted variety of elpidite was identified in a hydrothermally altered peralkaline pegmatite at Mt. Yukspor in the Khibiny alkaline complex, Kola Peninsula, Russia. It differs from ‘ordinary’ elpidite, ideally Na2ZrSi6O15⋅3H2O, in its crystal chemical features, infrared spectrum and optical characteristics. The chemical composition (wt.%, electron microprobe, H2O by TGA) is: Na2O 5.45, K2O 0.67, CaO 0.05, SiO2 60.32, TiO2 1.34, ZrO2 18.43, Nb2O5 0.65, H2O 12.80, total 99.71. The empirical formula calculated on the basis of 6 Si and 15 O atoms is [(Na1.05K0.08Ca0.01)Σ1.14(H3O)0.74]Σ1.88(Zr0.89Ti0.10Nb0.03)Σ1.02Si6O15⋅3.47H2O; the H2O:H3O ratio was calculated from the charge balance requirement, taking into account the results of crystal structure refinement. The highly hydrated variety of elpidite is orthorhombic, Pma2, a = 14.5916(6), b = 7.3294(3), c = 7.1387(2) Å, V = 763.47(5) Å3 and Z = 2. The crystal structure was solved from single-crystal X-ray diffraction data, R1 = 3.43%. The structure is based upon an elpidite-type heteropolyhedral Zr–Si–O framework with Na+ and H3O+ cations and H2O molecules in the zeolitic channels. Hydronium cations substitute for water molecules in one of the extra-framework sites. This variety of elpidite could be considered as an intermediate product of natural ion-exchange reaction between ‘ordinary’ elpidite and a low-temperature hydrothermal fluid.

A major debris flow, the Trig Point Hill flow, originating from Kerimasi volcano (Tanzania) contains numerous blocks of extrusive/pyroclastic carbonatites similar to those exposed at the rim of the currently inactive crater. The blocks of calcite carbonatite consist of: (1) large clasts of corroded and altered coarse grained calcite; (2) primary prismatic inclusion bearing phenocrystal calcite; and (3) a matrix consisting primarily of fine-grained prismatic calcite. The large clasts are inclusion free and exhibit a ‘corduroy-like’ texture resulting from solution along cleavage planes. The resulting voids are filled by brown Fe–Mn hydroxides/oxides and secondary calcite. The prismatic or lath-shaped phenocrystal calcite is not altered and contains melt inclusions consisting principally of primary Na–Ca carbonates which contain earlier-formed crystals of monticellite, periclase, apatite, Mn–Mg-magnetite, Mn–Fe-sphalerite and Nb-perovskite. Individual Na–Ca carbonate inclusions are of uniform composition, and the overall range of all inclusions analysed (wt.%) is from 28.7 to 35.9 CaO; 16.7–23.6 Na2O; 0.5–2.8 K2O, with minor SO3 (1.1–2.2) and SrO (0.34–1.0). The Na–Ca carbonate compositions are similar to that of shortite, although this phase is not present. The Na–Ca carbonates are considered to be primary deuteric phases and not secondary minerals formed after nyerereite. Monticellite shows limited compositional variation and contains 2–4 wt.% MnO and 12 wt.% FeO and is Mn-poor relative to monticellite in Oldoinyo Lengai natrocarbonatite. Periclase is Fe-bearing with up to 13 wt.% FeO. Spinels are Cr-free, Mn-poor and belong to the magnetite–magnesioferrite series in contrast to Mn-rich spinels of the magnetite–jacobsite series occurring in Oldoinyo Lengai natrocarbonatite. The matrix in which the ‘corduroy’ clasts and phenocrystal calcite are set consists of closely packed small prisms of calcite lacking melt inclusions, with interstitial fine-grained apatite, baryte, strontianite and minor fluorite. Pore spaces are filled with secondary Mn–Fe hydroxides/oxides, anhydrite and gypsum. The hypothesis that flow-aligned calcite in volcanic calciocarbonatites from Kerimasi, Tinderet, Homa and Catanda is altered nyerereite is discussed and it is considered that these calcite are either primary phases or altered melilite. The nyerereite alteration hypothesis is discussed with respect to the volumetric and compositional aspects of pseudomorphism by dissolution–precipitation replacement mechanisms. This study concludes that none of the volcanic calciocarbonatites containing flow-aligned calcite phenocrysts are altered natrocarbonatite.

The Neoproterozoic Sevattur complex is composed essentially of calcite and dolomite carbonatites together with pyroxenites and diverse syenites. This work reports the compositions and paragenesis of different pyrochlore generations hosted by albitite veins in this complex. The pyrochlore are distinctive, being exceptionally rich in uranium (26 to 36 wt.% UO2). Five types of pyrochlore (Pcl-I to Pcl-V) are recognised on the basis of composition and texture. With the exception of Pcl-V, the majority of the pyrochlore (Pcl-II to Pcl-IV) are surrounded by a thick orbicular mantle of Ba-rich potassium feldspar. This mantle around Pcl-V is partially-broken. Pcl-I is restricted to the cores of crystals, and associated with Pcl-II and -III and is relatively rich in Nb (0.53–0.62 apfu) together with more A-site vacancies (0.37–0.71 apfu) compared to Pcl-II to Pcl-IV. Other pyrochlore (Pcl-II to Pcl-IV) are characterised by elevated Ca and Ti compared to Pcl-I, which are related to the (3Nb5+ + Na+ → 3Ti4+ + U4+) and (2Nb5+ → 2Ti4+ + Ca2+) substitutions, respectively. These substitutions represent replacement of Pcl-II to Pcl-IV. Alteration and Ba-enrichment in all the pyrochlore are marked by interaction with an externally-derived Ba-rich hydrothermal fluid following the (2Nb5+ → 2Ti4+ + Ba2+) substitution. This substitution, coupled with extensive metamictisation leads to the formation of Ba-rich (15.9–16.3 wt.% BaO) patchy-zoned Pcl-V. The orbicular mantles around Pcl-I to Pcl-IV have prevented extensive metamictisation and extensive secondary alteration compared to Pcl-V, where mantling is partially disrupted. The compositional and textural variation suggests that Pcl-II to Pcl-IV form by nucleation on Pcl-I, and are transported subsequently as antecrysts in the host albitite.

The Quaternary carbonatite–nephelinite Kerimasi volcano is located within the Gregory rift in northern Tanzania. It is composed of nephelinitic and carbonatitic pyroclastic rocks, tuffs, tuff breccias and pyroclastic breccias, which contain blocks of different plutonic (predominantly ijolite) and volcanic (predominantly nephelinite) rocks including carbonatites. The plutonic and volcanic carbonatites both contain calcite as the major mineral with variable amounts of magnetite or magnesioferrite, apatite and forsterite. Carbonatites also contain accessory baddeleyite, kerimasite, pyrochlore and calzirtite. Zr and Nb minerals are rarely observed in rock samples, though they are abundant in eluvial deposits of carbonatite tuff/pyroclastic breccias in the Loluni and Kisete craters. Pyrochlore, ideally (CaNa)Nb2O6F, occurs as octahedral and cubo-octahedral crystals up to 300 μm in size. Compositionally, pyrochlore from Loluni and Kisete differs. The former is enriched in U (up to 19.4 wt.% UO2), light rare earth elements (up to 8.3 wt.% LREE2O3) and Zr (up to 14.4 wt.% ZrO2), and the latter contains elevated Ti (up to 7.3 wt.% TiO2). All the crystals investigated were crystalline, including those with high U content (a = 10.4152(1) Å for Loluni and a = 10.3763(1) Å for Kisete crystals). They have little or no subsolidus alteration nor low-temperature cation exchange (A-site vacancy up to 1.5% of the site), and are suitable for single-crystal X-ray diffraction analysis (R1 = 0.0206 and 0.0290; for all independent reflections for Loluni and Kisete crystals, respectively). Observed variations in the pyrochlore composition, particularly Zr content, from the Loluni and Kisete craters suggest crystallisation from compositionally different carbonatitic melts. The majority of pyrochlore crystals studied exhibit exceptionally well-preserved oscillatory- and sometimes sector-type zoning. The preferential incorporation of smaller and higher charged elements into more geometrically constrained sites on the growing surfaces explains the formation of the sector zoning. The oscillatory zoning can be rationalised by considering convectional instabilities of carbonatite magmas during their emplacement.