Volume 110, Issue 4

The high-pressure behavior of iron nitrides has garnered significant attention due to the potential of deep nitrogen reservoirs within the Earth’s interior. Here, we investigate the magnetic, structural, electrical, and thermal properties of Fe 3 N up to 62 GPa and 2100 K, using multiple probes (including synchrotron X-ray diffraction, synchrotron Mössbauer spectroscopy, and electrical measurements) coupled with the diamond-anvil cell technique. Fe 3 N undergoes a magnetic phase transformation from the ferromagnetic to paramagnetic state at ∼17–20 GPa, 300 K. The equation of state was determined as V 0 / Z = 42.8(1) Å 3 , and K 0 = 151.8(1) GPa, with K ' fixed at 4. Additionally, Fe 3 N exhibits unexpectedly low electrical and thermal conductivity under high-pressure and high-temperature conditions. This result suggests that deep nitrogen cycling may contribute to the thermal evolution of the deep interiors of Earth and other terrestrial bodies.

RamanCrystalHunter (RCH) is a new software program designed to pre-process, analyze, and identify Raman spectra by comparison with spectra in the RamanCrystalHunter Database (RCHDB). The software is free and can be downloaded from the website https://www.fabrizionestola.com/rch. RCH is characterized by a simple graphical user interface, making it suitable for both specialist and non-specialist users, and it has been developed mainly for applications in Earth Sciences (processing the spectra of minerals) but can be used to process the Raman spectra of any synthetic or natural inorganic or organic material. RCH allows users to visualize, pre-process (e.g., using smoothing, noise reduction, and baseline correction operations), and analyze (e.g., using fitting or various calculation tools) Raman spectra. Moreover, it is equipped with the RCHDB, a new database of high-quality mineral spectra that can be downloaded for free, along with the RCH program. The RCHDB contains the Raman spectra of minerals (including single- and multi-phase inclusions within mineral hosts, for example, diamonds) and related synthetic compounds, allowing for rapid and accurate identification of unknown spectra. The RCH software includes highly customizable yet efficient and user-friendly methods for processing and analysis of Raman spectra and represents a valuable contribution to the field of Raman spectroscopy, whose applications have expanded greatly in recent years, especially in Earth Sciences. Two practical examples of novel ways in which this software can be used for geoscience applications are presented.

Quantitative phase analysis (QPA) using the Rietveld method and X-ray diffraction (XRD) patterns is useful for predicting the reactivity of a rock to carbon dioxide (CO 2 ) and for quantifying mineral carbonation. Lizardite and smectites in kimberlite are reactive to CO 2 , but they are structurally disordered and cannot be quantified using the standard Rietveld approach. In this study, the Partial Or No Known Crystal Structure (PONKCS) method was used to model the peak profiles of smectite and lizardite to account for turbostratic stacking disorder in synthetic samples of processed kimberlite. Lizardite and montmorillonite PONKCS models were made using XRD patterns collected with three X-ray diffractometers: two XRDs from the same manufacturer and of similar model (XRDs B1 and B2) and another XRD from a different manufacturer (XRD A1). Five synthetic samples of processed kimberlite of known compositions were prepared and used to test the results of these PONKCS models for data collected using all three instruments. The results provide a total bias ranging from 4.8–14.1 wt% using correctly calibrated, instrument-specific PONKCS models. We also tested the sensitivity of the PONKCS method to changes in instrument geometry: PONKCS models calibrated for one instrument (XRD B1) were used in refinements with XRD data collected on an instrument made by a different manufacturer (XRD A1), or on a similar instrument made by the same company but having a slightly different geometry (XRD B2). Results were highly inaccurate when PONKCS models calibrated to XRD B1 were used with patterns collected on XRD A1 (32.1–71.6 wt% total bias for our weighed mixtures). Our results show that the smaller differences in instrument parameters between XRD B1 and XRD B2 can also lead to inconsistent and less accurate QPA results using PONKCS (9.8–32.7 wt% total bias). Therefore, correct calibration of PONKCS models to a specific XRD instrument is required for accurate QPA and quantification of CO 2 mineralization in clay-rich rocks.

Amorphous iron-arsenate precipitates are significant As sinks in natural and industrial settings and often serve as precursors to the crystallization of ferric arsenate minerals such as scorodite (FeAsO 4 ·2H 2 O) and kankite (FeAsO 4 ·3.5H 2 O). These amorphous phases have varied structures and compositions depending on the geochemical conditions under which they form but have not been well characterized except for the frequently encountered amorphous ferric arsenate (represented as AFe III in the present study). Here, we intend to characterize another phase, amorphous ferrous arsenate (AFe II ), which is formed in partially oxidized Fe(II)-As(V) systems, and carry out a comparative study on the properties and related scorodite mineralization processes relative to those of AFe III . We synthesized both AFe II and AFe III , determined their compositions and structures, and finally examined the amorphous-crystalline phase transition processes leading to scorodite mineralization. Computed chemical formulas showed AFe III and AFe II can be represented by Fe(III) 0.99 AsO 4 ·4.1H 2 O and Fe(III) 0.48 Fe(II) 0.43 (HAsO 4 ) 0.7 (AsO 4 ) 0.3 ·4.0H 2 O, respectively, with AFe II containing 47.3% Fe(II), as deduced from the pre-edge data of Fe K -edge X-ray absorption spectroscopy. Structurally, the Fe-O bond at 2.02 ± 0.01 Å in AFe II was significantly longer than that of 1.99 ± 0.01 Å in AFe III and scorodite. In addition, AFe II appeared to have a higher stability (slower kinetics of scorodite crystallization) under either atmospheric or aqueous conditions. Together with solution chemistry measurements, these findings suggest AFe II controls scorodite formation by limiting the occurrence of aqueous Fe 3+ and retards the amorphous-scorodite phase transition compared to the case of AFe III . These findings improve our understanding of iron arsenate system and may find industrial applications to arsenic sequestration via scorodite mineralization.

In anorthite-diopside-tridymite paralava of the Hatrurim Complex, Daba-Siwaqa, Jordan, phosphides from two facies environments were found. Large xenomorphic phosphide aggregates, presented by minerals of the barringerite-transjordanite series, murashkoite and zuktamrurite, were noted at the intimate contact of the paralava with the host rocks. Moreover, the contact facies of the paralava, composed of large diopside crystals up to 1 cm in size and cemented by hydrosilicates, zeolites, and calcite, are enriched in phosphides. In the contact facies, it was possible to observe pseudomorphs of barringerite and murashkoite after fish bones and inclusions of minute needle-like barringerite crystals in diopside with morphological manifestations of the simultaneous crystallization of these minerals. In the central part of the weakly altered paralava with abundant amygdules, two nodules containing ore minerals were detected. In the first nodule, barringerite and schreibersite were found; the second contained nickelphosphide. In this paper we discuss the hypothesis that phosphides at the contact facies of the paralava formed as a result of reducing carbothermal reactions with the participation of thermally altered and graphitized fish-bone remains as a source of carbon and phosphorus and oxidized pyrite framboids as an iron source: Fe 2 O 3 +3C = 2Fe(lq)+3CO(g), 2Ca 5 (PO 4 ) 3 F+ 14 C= 3P 2 (g)+ 14 CO(g)+10Cao+F 2 (g) and nFe(lq)+½P 2 (g) = Fe n P, where n = ½, 1, 2, 3. The process of phosphide formation during hot paralava injection, the temperature of which exceeded 1400 °C, into previously altered host rocks with contamination of material containing bone remains, took place in a small volume in a kinetic mode, which led to the preservation of the primary form of bone remains when they were replaced by phosphides. The formation of phosphides in the central part of the paralava occurred with the participation of gases that transported phosphorus reduced as a result of carbothermal reactions. The phosphorous was absorbed by drops of metal or sulfide melt, which led to the formation of phosphides.

This study investigates silicate liquid immiscibility (SLI) microstructures in the Chang’E-5 (CE-5) lunar ferrobasalt sample, the youngest recovered m are basalt (ca. ∼2.0 Ga). Employing advanced high-resolution imaging techniques and chemical analysis, we examined a subophitic fragment, revealing two distinct types of microstructures indicative of multi-stage SLI events. The first type is observed in the mesostasis pockets and exhibits both “sieve” and “maze” textures, where the Si-K-rich glassy phases are interconnected with Fe-rich minerals, e.g., fayalite. This type of microstructure, similar to previous observations in Apollo and Luna samples, is the product of a stable SLI event. The second type is characterized by K-free but high-Si melt inclusions occurring as emulsions in the rims of plagioclase. The entrapment of these emulsions followed a metastable SLI event, with the Fe-rich liquids serving as precursors to subsequent stable SLI processes. Additionally, the Fe-rich droplets within the emulsions underwent coarsening via Ostwald ripening, a phenomenon in which smaller particles in solution dissolve and deposit on larger particles. Our simulation of this coarsening process suggests a duration of at least 15–32 days for the SLI processes, alongside a slow cooling rate (<0.3 °C/h) of the late-stage CE-5 lava. We propose that metastable SLI may have influenced the effusive signature of the CE-5 lava flow during its late-stage evolution. The metastable SLI process can potentially lead to the formation of various phases during the late-stage evolution of lunar ferrobasaltic magmas, thereby contributing to the diversity of lunar rock types.

Results from X-ray remote sensing aboard NASA’s MErcury Surface Space ENvironment GEochemistry and Ranging (MESSENGER) spacecraft have demonstrated that Mercury has a low, but measurable, concentration of Fe on its surface. However, ultraviolet to near-infrared spectroscopic measurements of the mercurian surface do not show the 1 μm absorption band characteristic of ferromagnesian silicates. This observation is consistent across multiple Fe-bearing terranes with a range of ages, suggesting the Fe present on Mercury’s surface may not be stored within silicate phases. To further constrain the possible mineralogy and composition of Fe-bearing phases on Mercury, we used various spectroscopic techniques to characterize synthetic olivine with minor amounts of Fe (i.e., Fo 99.62 –Fo 99.99 ) and more Fe-rich natural olivines. Our results indicate that the distinctive 1 μm absorption band of olivine is detectable in reflectance spectra of olivine at a concentration as low as 0.03 wt% FeO and 0.01 wt% in continuum removed data. Additionally, MESSENGER’s lack of a 1 μm absorption, taking into account Mercury Dual Imaging System (MDIS)’s limited spectral resolution and Mercury Atmospheric and Surface Composition Spectrometer (MASCS)’s high signal-to-noise ratio, suggests there is <0.38 wt%, and likely <0.01 wt%, FeO on the surface of Mercury. Because the 1 μm band is not observed in surface spectra, these results indicate that the Fe observed on the surface of Mercury is not bound in an olivine structure. Rather, we posit that Fe is present as nano-phase and macroscopic Fe-rich metal or Fe-sulfide that formed as a result of space weathering and igneous smelting processes. Looking forward to ESA/JAXA’s BepiColombo mission that has a planned Mercury orbit arrival time in December 2025, Mercury Radiometer and Thermal Infrared Imaging Spectrometer (MERTIS) mid-infrared spectra should provide a mineralogical detection or absence of olivine where MIR spectral features are still present even in synthetic olivines with minor amounts of Fe (Fo 99.99 ).

Lead-bismuth (Pb-Bi) minerals of the galena (PbS)-matildite (AgBiS 2 ) series and lillianite homologs (Pb 3–2xAgxBi2+xS 6) are intergrown with electrum (Au-Ag alloy) and chalcopyrite (CuFeS 2 ) in specific bands within a colloform-banded vein at the Switchback epithermal deposit in southern Mexico. A macro-to nanoscale study revealed that these minerals fill small (<200 μm) cavities in the gangue minerals, showing curvilinear boundaries, bleb-like morphologies, and rounded nanoparticles (∼100 nm). These observations are consistent with growth in a molten system from a precursor Pb-Bi melt containing Au, Ag, and Cu. Minerals in the galena-matildite series typically display Widmanstätten textures (i.e., octahedral-like or basket-weave matildite ss lamellae in the galena ss matrix), which have been traditionally linked to the decomposition of a high-temperature solid solution. However, galena ss and matildite ss show nanoscale sinuous reaction fronts and replacement relicts (“islands”) while maintaining the [011] Galena ∥ [100] Matildite crystallographic orientation relationship. This suggests a topotaxial growth of matildite ss mediated by coupled dissolution-reprecipitation reactions between galena ss and the metallic melt upon cooling. A similar scenario is proposed for Pb-Bi sulfosalt intergrowths, which replace galenamatildite and electrum and grow topotaxially along (200) Galena . Collectively, these results suggest that Pb-Bi melts can exist in epithermal fluids, acting as precursors for the crystallization of ore minerals and being able to sequester precious metals. This model explains abnormally high-Au-Ag enrichments observed in some deposits that contain Pb-Bi ores.

A study of a skarn sample from the Malmkärra iron mines, Norberg, Västmanland (Sweden), revealed the occurrence of a peculiar epidote-supergroup mineral. It was examined using electron microprobe, single-crystal X-ray diffraction, Mössbauer, and Fourier-transform infrared spectroscopy (FTIR) techniques. Structure refinements combined with electron microprobe data indicate the following cation populations: A 1 = C a 0.96 R E E 0.03 3 + M n 0.01 ; A 2 = R E E 0.99 3 + C a 0.01 ; M 1 = M g 0.40 A l 0.32 F e 0.26 3 + F e 0.02 2 + ; $A 1=\mathrm{Ca}_{0.96} \mathrm{REE}_{0.03}^{3+} \mathrm{Mn}_{0.01} ; A 2=\mathrm{REE}_{0.99}^{3+} \mathrm{Ca}_{0.01} ; M 1=\mathrm{Mg}_{0.40} \mathrm{Al}_{0.32} \mathrm{Fe}_{0.26}^{3+} \mathrm{Fe}_{0.02}^{2+} ;$ M 2 = A l 0.98 F e 0.02 3 + ; M 3 = M g 0.72 F e 0.17 2 + F e 0.11 3 + ; T 1 , 2 , 3 = S i 2.93 A l 0.07 , $M 2=\mathrm{Al}_{0.98} \mathrm{Fe}_{0.02}^{3+} ; M 3=\mathrm{Mg}_{0.72} \mathrm{Fe}_{0.17}^{2+} \mathrm{Fe}_{0.11}^{3+} ; T 1,2,3=\mathrm{Si}_{2.93} \mathrm{Al}_{0.07},$accounting for a total positive charge of 24.64. The presence of Fe 2+ is confirmed by Mössbauer data. The remarkable number of divalent cations at both M 1 and M 3 (>1 pfu, per formula unit) demands more than one monovalent anion pfu in the structure. As the mineral lacks fluorine, charge neutrality must be achieved through additional H + (about 0.4 apfu). Only one independent hydrogen atom is located within the structure, with O10 as donor and O4 as acceptor, as in other epidote-supergroup minerals. Nonetheless, another O-O distance is suitable for a hydrogen bond, namely O10-O2. Although the existence of the additional OH group was not directly proved by vibrational spectroscopy, FTIR data provided information related to this potential O10-O2 bridge. In the IR spectrum acquired, several bands are observed in the OH-stretching region, and a secondary peak at 2140 cm −1 could be assigned to the bending mode of the O10-H···O2 group. To shed light on this puzzling observation, one single crystal was subjected to annealing experiments at temperatures from 500 to 700 °C, in 50 °C steps, while a second one underwent a heat treatment at 700 °C. After the heat treatment, the IR spectrum showed a decreased intensity of all observed bands, in agreement with a dehydrogenation occurring at high temperatures. Although the structural position of the second hydrogen is still uncertain, it is reasonable to describe the composition of the epidote-supergroup mineral from Malmkärra as a solid solution between dissakisite-(Ce) (32%), ferriallanite-(Ce) (28%), and a yet undescribed end-member (40%), the (OH)-analog of dollaseite-(Ce), ideally CaREE 3+ Mg 2 AlSi 3 O 11 (OH) 2 .

The high field strength elements (HFSE) Ti, Nb, and rare earth elements (REE) are commonly regarded as immobile during hydrothermal activity and metamorphism, making them important tracers of geological processes. Here, we report a Ti-REE-Nb-As mineralization recently discovered in quartz, feldspar, muscovite ± biotite, fluorapatite, hematite, epidote, and dravite-schorl veins hosted in quartz ± feldspar ± muscovite ± biotite gneisses from the Monte Leone nappe (Switzerland/Italy). The veins formed during prograde metamorphism and were boudinaged and/or folded during peak metamorphism under lower amphibolite facies. The mineralization consists of megacrysts (≫2 cm) of allanite-(Ce) and Nb-REE-rich titanite-(I). Titanite-(I) displays prominent primary oscillatory- and sector-zonings in Y+REE and Nb. Allanite-(Ce) and titanite are also present as metamorphic minerals disseminated in the host-rock. The vein-hosted megacrysts and their host rocks have identical Nd isotope systematics, indicating that the HFSE mineralization results from small-scale remobilization of host-rock components. Localized, fluid-assisted dissolution of vein-hosted allanite-(Ce), epidote, and dravite-schorl during retrograde alpine deuteric alteration resulted in cavities lined with chlorite, muscovite, hematite, and diverse REE minerals. The same fluids caused titanite-(I) to break down into a porous assemblage of acicular niobian rutile with lamellae of crichtonite-group minerals and/or hematite and a suite of REE-Nb-Ti micro-minerals. A few titanite (titanite-II) crystals preserve an intermediate stage of the dissolution-reprecipitation process. Unlike titanite-(I), they display a patchwork-like micro-texture (100 μm sized subgrains with inhomogeneous Nb concentrations); they host lamellae of crichtonite-group minerals within cleavage planes of the parent titanite, as well as secondary Y+Nb+REE oxides and calcite along subgrain boundaries. The occurrence of calcite indicates that CO 2 -enriched fluids promoted the destabilization of titanite-(I). Highly localized fluid flow accounts for the common occurrence of fresh and altered allanite-(Ce) and titanite in close proximity. The HFSE-enriched veins reveal a complex history of mobility of minor elements (Ti, Nb, REE, As ± B, Be) together with major components (Si, Al, K, Na, Fe) from the host rock, resulting in their early (prograde) concentration within the veins, and their remobilization upon the action of oxidized CO 2 -bearing fluids during retrograde metamorphism. In general, crystallization of enriched phases during prograde metamorphism may be an important step in determining the fertility of a source rock for hydrothermal HFSE deposits.

Sodic volcano-plutonic terranes in the Archean can be well preserved, but why oxidized S-rich sodic magmas and porphyry-type Cu-Au deposits are so rare remains poorly understood. Here we addressed this issue by measuring the S concentration and S 6+ /ΣS ratio of primary apatite grains in >2.7 Ga felsic volcanic rocks from the well-characterized Neoarchean Abitibi Greenstone Belt of the Superior Province, Canada. Whereas apatite grains inmost samples contain low-S concentrations (<0.01wt%, n = 24), a few apatite samples are S-rich (0.14 ± 0.03 wt%, 1σ) and have low-S 6+ /ΣS ratios (0.56 ± 0.17; 1σ, n=4). Samples with S-poor apatite have variable whole-rock La/Yb ratios (generally <30) and zircon 10 000*(Eu/Eu*)/Yb ratios of 11 ± 8 (1σ), which may be products of plume-driven or over-thickened crustal melting. In contrast, the samples with S-rich apatite have elevated La/Yb ratios of 49 ± 15 (1σ), zircon 10 000*(Eu/Eu N *)/Yb ratios of 26 ± 7 (1σ), and zircon δ 18 O values of 5.8± 0.1 ‰ (1σ), consistent with a deep, hydrous and homogeneous mantle-like source for the melts dominated by amphibole ± garnet fractionation that is reminiscent of subduction-like process. These are the first reported results documenting the predominant accommodation of relatively reduced S in S-rich apatite grains crystallized from terrestrial silicate melts, possibly reflecting slight oxidation associated with the hydration of Neoarchean mantle and crystal fractionation over the magma evolution. The more common S-poor apatite data suggest that suppressed oxidation of the parental sodic magmas led to weak S emission from Earth’s interior to its evolving surface, explaining the rarity of porphyry-type Cu deposits in >2.7 Ga Archean sodic volcano-plutonic terranes.

The new mineral jianmuite, Z r T i 4 + T i 5 3 + A l 3 O 16 $\mathbf{Z r T i}^{4+} \mathbf{T i}_{5}^{3+} \mathbf{A l}_{3} \mathbf{O}_{16}$ IMA No. 2023-057) was found: (1) as an inclusion in a corundum grain from the Cr-11 orebody near Kangjinla (29°11 ʹ N, 92°18 ʹ E), Tibet, China, with wenjiite, Ti 10 (Si,P,□) 7 , Al-rich spinel, osbornite, TiN, a Ca-amphibole particularly enriched in Ti, a Sr-rich dmisteinbergite, CaAl 2 Si 2 O 8 , and amorphous phases (probably residual melt) (holotype and first cotype), and (2) as several grains with corundum, mullite, Al 6 Si 2 O 13 , tistarite, Ti 2 O 3 , and kaitianite, T i 2 3 + T i 4 + O 5 , $\mathrm{Ti}_{2}^{3+} \mathrm{Ti}^{4+} \mathrm{O}_{5},$ in the matrix of the CV3 carbonaceous chondrite Allende (second cotype). Eight energy-dispersive spectroscopic analyses of the holotype normalized to 100% gave SiO 2 1.83, ZrO 2 23.17, Al 2 O 3 18.41, Ti 2 O 3 54.32, Sc 2 O 3 1.25, MgO 1.03, total 100.00 wt%. The empirical formula calculated on the basis of cations and 16 O atoms is V I I I Z r V I I T i 3.034 3 + T i 0.602 4 + Z r 0.364 Σ 4.000 V I T i 1.845 3 + A l 1.838 M g 0.186 S c 0.131 Σ 4.000 I V A l 0.780 S i 0.220 Σ 1.000 ${ }^{\mathrm{VIII}} \mathrm{Zr}^{\mathrm{VII}}\left(\mathrm{Ti}_{3.034}^{3+} \mathrm{Ti}_{0.602}^{4+} \mathrm{Zr}_{0.364}\right)_{\Sigma 4.000}{ }^{\mathrm{VI}}\left(\mathrm{Ti}_{1.845}^{3+} \mathrm{Al}_{1.838} \mathrm{Mg}_{0.186} \mathrm{Sc}_{0.131}\right)_{\Sigma 4.000}{ }^{\mathrm{IV}}\left(\mathrm{Al}_{0.780} \mathrm{Si}_{0.220}\right)_{\Sigma 1.000}$ O 16 . The simplified formula is Z r T i 3 + , T i 4 + , Z r 4 T i 3 + , A l 4 ( A l , S i ) O 16 $\mathrm{Zr}\left(\mathrm{Ti}^{3+}, \mathrm{Ti}^{4+}, \mathrm{Zr}\right)_{4}\left(\mathrm{Ti}^{3+}, \mathrm{Al}\right)_{4}(\mathrm{Al}, \mathrm{Si}) \mathrm{O}_{16}$ and the ideal chemical formula is Z r T i 4 + T i 5 3 + A l 3 O 16 , $\mathrm{ZrTi}^{4+} \mathrm{Ti}_{5}^{3+} \mathrm{Al}_{3} \mathrm{O}_{16},$ which requires calculated contents of ZrO 2 17.23 wt%, TiO 2 11.16 wt%, Ti 2 O 3 50.23 wt%, and Al 2 O 3 21.38 wt%, total 100 wt%. Six energy-dispersive spectroscopic analyses of the first cotype normalized to 100% gave SiO 2 2.86, ZrO 2 23.06, Al 2 O 3 18.10, Ti 2 O 3 53.78, Sc 2 O 3 1.17,MgO1.04, total 100.00 wt%. Three-dimensional electron diffraction of the holotype determined jianmuite as I -centered tetragonal, space group: ( I 4̄) (#82), with a = 10.3675(10) Å, b = 10.3675(10) Å, c = 9.8125(10) Å, V = 1054.70(18) Å 3 , and Z = 4. Wavelength-dispersive spectroscopic electron probe microanalyses of Allende jianmuite (second cotype) gave for the two grains (5 analyses each) SiO 2 2.32, 3.02, ZrO 2 13.95, 10.79, Al 2 O 3 26.82, 21.83, Ti 2 O 3 44.68, 49.47, TiO 2 12.62, 14.75, Sc 2 O 3 0.12, 0.04, MgO 0.19, 0.80, FeO, 0.52, 0.55, CaO 0.17, 0.08, total 101.40, 101.34 wt%. TiO 2 (wt%) and Ti 2 O 3 (wt%) were calculated from formula units Ti 4+ and Ti 3+ , which were determined assuming a stoichiometry of 10 cations and 16 O atoms. Electron backscatter diffraction (EBSD) patterns on the Allende jianmuite (second cotype) can be indexed by the tetragonal I 4̄ structure obtained on the holotype, with a mean angular deviation of ∼0.7°, thereby confirming identification of the Allende mineral as jianmuite. The structure of jianmuite consists of cations packed in a distorted I -centered lattice. There are eight independent cation sites: two inside distorted cubes with tetragonal ridged faces (CN = 8), two inside capped trigonal prisms (CN = 7), two inside distorted octahedra (CN = 6), and two inside tetrahedra (CN = 4). All coordination polyhedra are connected by edges, except for the tetrahedra, which are connected by vertexes with other polyhedra. Despite the similarity in composition with carmeltazite, Z r T i 4 3 + A l 2 O 11 , $\mathrm{ZrTi}_{4}^{3+} \mathrm{Al}_{2} \mathrm{O}_{11},$ the structures of the two minerals bear little resemblance to one another, most notably in the number and type of cation-oxygen polyhedra. Instead, the jianmuite structure has several features in common with structures of pyrochlore-supergroup minerals, including similarity in unit-cell parameters, cubic a ≈ 10 Å and tetragonal a ≈ c ≈ 10 Å in pyrochlore and jianmuite, respectively, and cation packing being cubic I -centered in both structures.

Cabrerite (IMA2023-123), NiMg 2 (AsO 4 ) 2 ·8H 2 O, is a newly approved mineral species from the Nickel mine, Cottonwood Canyon, Table Mountain district, Churchill County, Nevada, U.S.A., that was originally described in 1863 from Sierra Cabrera, Almería, Andalusia, Spain. At the Nickel mine, cabrerite occurs in divergent groups of green blades up to 1 mm long. Blades are elongated and striated parallel to [001], flattened on {010}, and exhibit the forms {010}, {110}, and {201̄}. The mineral is transparent with vitreous luster and white streak. The Mohs hardness is ∼2½. The mineral has moderately sectile tenacity, irregular and stepped fracture, and three cleavages: perfect on {010}, fair on {100}, and poor on {1̄02}. The measured density is 2.93(2) g·cm −3 . The mineral dissolves slowly in RT dilute HCl. The mineral is optically biaxial (+), α = 1.609(2), β = 1.633(2), γ = 1.667(2) (white light); 2 V meas = 82(2)°; slight r < v dispersion; orientation X = b , Z ^ c = 37° in obtuse β; nonpleochroic. Electron probe microanalysis provided the empirical formula (Mg 1.46 Ni 1.55 )Σ 3.01 (As 1.00 O 4 ) 2 ·8H 2 O. Cabrerite is monoclinic, C 2/ m , a = 10.2054(11), b = 13.3772(13), c = 4.7382(4) Å, β = 105.057(7)°, V = 624.66(11) Å 3 , and Z = 2. The mineral has a vivianite-type structure ( R 1 = 0.0353 for 668 I > 2σ I reflections) in which M lO 2(H 2 O) 4 octahedra and M 22 O 6(H 2 O) 4 edge-sharing octahedral dimers are linked together via T O4 tetrahedra (where T = P or As), and hydrogen bonds to form layers parallel to {010}; successive layers are linked by hydrogen bonds only. Cabrerite is the ordered intermediate between hörnesite and annabergite, with Ni dominant at M 1, Mg dominant at M 2, and T =As.

Natural quasicrystals, compounds with characteristics intermediate between crystalline material and glass, have been so far discovered in extraterrestrial materials only. Furthermore, their occurrence in nature is limited to metallic, Al-bearing alloys. The presence of metallic aluminum in these minerals raised doubts about their potential occurrence in terrestrial rocks. The geochemical conditions needed to form metallic Al are so reducing that they are considered very unlikely in a terrestrial environment. Given the report of dodecagonal symmetry in the synthetic Ta 1.6 Te quasicrystal, the search for terrestrial quasicrystals was focused on natural tellurides. Here, we report the discovery of the first terrestrial approximant of a dodecagonal quasicrystal, a Pd-Ni-telluride with formula Pd 3 Ni 4 Te 8 and tetragonal symmetry, which was found as small inclusions in a rock sample from Kalgoorlie, Western Australia. Periodic approximants are crystalline materials that share a similar chemical composition with quasicrystals but have a slightly altered atomic structure, aligning their symmetry with the traditional principles of three-dimensional crystallography. These crystalline approximants provide insights into the local atomic structure of their corresponding quasicrystals. Natural Pd 3 Ni 4 Te 8 has been approved as a new mineral by the International Mineralogical Association with the name proxitwelvefoldite (IMA 2024-034). The Pd 3 Ni 4 Te 8 composition has never been reported to form quasicrystal approximants among synthetic products and could indicate the possible existence of a 12-fold quasicrystal in the Pd-Ni-Te system. The discovery points to the possibility that the quasicrystalline structure may be much more common than previously thought, even in non-alloy systems.