Kishonite, VH₂, and Oreillyite, Cr₂N, Two New Minerals from the Corundum Xenocrysts of Mt Carmel, Northern Israel

1. Introduction

The study of the xenoliths occurring in pyroclastic ejecta of small Cretaceous basaltic volcanoes exposed on Mt Carmel [1,2,3,4,5,6,7,8] has revealed a very peculiar, exotic and ultra-reduced mineral assemblage reflecting the interaction of deep-seated basaltic magmas with mantle-derived CH₄ + H₂ at high fluid/melt ratios. During these ongoing studies, two new minerals, kishonite (VH₂) and oreillyite (Cr₂N) have been identified and are described here.

The sample containing kishonite was recovered from bulk alluvial samples in the Kishon Mid Reach Zone 2. These are part of a xenolith assemblage that includes coarse-grained aggregates of hibonite + grossite + spinel assemblages that carry inclusions of V⁰ and V-Al alloys, indicating fO₂ down to ≤ΔIW − 9 [4].

The sample containing oreillyite was recovered from the Kishon Mid Reach Zone 1 and consists of a 1.7 mm crystal of corundum with a hollow centre and raised rims, indicating hopper growth, and it has an intense purple-red colour [5].

The name kishonite is for the Kishon river, which drains Mt Carmel and enters the sea near Haifa in Northern Israel (and where the placer gemstone deposits with the studied xenoliths are located), whereas oreillyite honours Suzanne O’Reilly (born 1946), professor in Geology at Macquarie University, Sydney, Australia, Director of the ARC National Key Centre for Geochemical Evolution and Metallogeny of Continents (GEMOC) and Director of the ARC Centre of Excellence for Core to Crust Fluid Systems (CCFS).

Both new minerals and mineral names have been approved by the Commission on New Minerals, Nomenclature and Classification of the International Mineralogical Association (No. 2020-023 and 2020-030a for kishonite and oreillyite, respectively). Holotypes are deposited in the collections of the Museo di Storia Naturale, Università degli Studi di Firenze, Firenze, Italy, catalogue number 3364/I. Oreillyite is also deposited at the Centre for Microscopy, Characterization and Analysis, The University of Western Australia, Australia, catalogue number 1174-C_FF.

2. Description and Physical Properties

Kishonite occurs as one subhedral single crystal, 17 × 25 × 38 μm (Figure 1), which is the holotype material. It is opaque. Colour, luster, streak, hardness, tenacity, cleavage, fracture, density, and optical properties could not be determined because of the small grain size.

Calculated density is 4.523 g·cm⁻³, using the empirical formula and X-ray single-crystal data (see below). Kishonite is associated with hibonite, grossite, spinel, vanadium, and V-Al alloys. A more detailed description of the kishonite occurrence is given by Bindi et al. [9].

Oreillyite occurs as one subhedral volume sandwiched between corundum and native chromium (Figure 2), which is the holotype material.

Colour, lustre, streak, hardness, tenacity, cleavage, fracture, density, and optical properties could not be determined because of the small grain size. Calculated density is 6.64 g·cm⁻³ using the ideal formula and the unit-cell volume of the triple ordered cell (Z = 3; see below).

3. Chemical Composition

Quantitative elemental microanalyses of the type grain of kishonite were carried out using a CAMECA 100× electron microprobe (WDS mode, 15 kV and 10 nA, 1 μm beam size, counting times 20 s for peak and 10 s for background; Gennevilliers, France). Analyses were processed with the CITZAF correction procedure and gave average values of (in wt.% of elements): V 93.71 and Al 2.61. Metallic vanadium and aluminium were used as standard [9]. Hydrogen was not measured (given the rarity of the phase), and the stoichiometry is given on the basis of the high-quality single-crystal X-ray refinement (see below). The VH₂ ideal formula requires V 96.19 and H 3.81.

A preliminary chemical analysis of oreillyite using EDS revealed the presence of Cr only. TEM-EDS spectra extracted from the maps collected on oreillyite in the FIB foil gave (in wt.% of elements) Cr 86.9 and N 13.1. Electron energy loss spectroscopy (EELS) data (Figure 3) indicate that the valence of Cr is less than 3 in oreillyite and that the mineral is actually oxygen-free. Chromium in low-Cr and high-Cr parts of the corundum occurs as Cr³⁺, whereas in oreillyite, Cr seems to be +1 or +2. The Cr₂N ideal formula requires Cr 88.13 and N 11.87.

4. X-ray Crystallography

The small kishonite fragment in Figure 1 was handpicked from the polished section under a reflected light microscope and mounted on a 5 μm diameter carbon fibre, which was, in turn, attached to a glass rod. Then, the fragment was tested by single-crystal X-ray diffraction. Single-crystal X-ray studies were carried out using a Bruker D8 Venture diffractometer equipped with a Photon II CCD detector (Billerica, MA, USA), with graphite-monochromatized MoKα radiation (λ = 0.71073 Å), and with 100 s exposure time per frame; the detector-to-sample distance was 7 cm. Kishonite is cubic, space group Fm$\overline{3}$m, with a = 4.2680(10) Å, V = 77.75(3) ų and Z = 4 [9].

The kishonite structure was solved by direct methods and then refined using the program Shelxl-97 [10]. The site occupation factor (s.o.f.) at the V site was allowed to vary (V vs. Al) using scattering curves for neutral atoms taken from the International Tables for Crystallography [11]. At the last refinement stage, with anisotropic atomic displacement parameters for V, the residual value settled at R1 = 0.0134 for 25 unique reflections and four parameters. Atom coordinates and isotropic displacement parameters are given in

Table 1. Atomic coordinates and equivalent isotropic displacement parameters (Ų × 10³) for kishonite. U_eq is defined as one third of the trace of the orthogonalized U^ij tensor.

Site	x	Y	z	Ueq	s.o.f.
V	0	0	0	10(2)	V0.96(7)Al0.04 V0.96(7)Al0.04
H	1/4	1/4	1/4	38(17)	1

s.o.f. = site occupancy factor.

.

The crystal structure of kishonite is shown in Figure 4. It is a CaF₂ structure type with V in cubic close packing and cubic eight-fold coordination and H occupying the T_d interstitial sites. V-H bond distance of 1.8481(3) Å. The solution of the crystal structure gave the VH₂ stoichiometry, with minor Al substituting for V (i.e., V_0.96Al_0.04), in excellent agreement with the electron microprobe data. The results obtained are also in excellent agreement with what has been reported for synthetic VH₂ [12].

As concerns oreillyite, we extracted a grain (30 × 38 × 66 μm in size) containing both native chromium and oreillyite (Figure 5) and tried conventional single-crystal X-ray diffraction experiments using a Rigaku XtaLAB Synergy (Tokyo, Japan) with a HyPix detector using MoKα radiation. We collected images with 480 s of exposure in the θ range 8.5–27.3°. Oreillyite was present as single crystals, whereas native chromium was present only as a homogeneous polycrystalline material producing rings that were indexed with the Im$\overline{3}$m structure of native Cr (a = 2.889 Å; [13]). Oreillyite was indexed with a = 2.7628(3), c = 4.4630(6) Å, and V = 29.503(6) ų, representing the hexagonal disordered form of CrN₂ (space group P6₃/mmc) [14,15,16,17,18] with lattice parameters a = 2.748 and c = 4.438 Å.

The oreillyite structure was solved by direct methods in the P6₃/mmc space group. It is a hexagonal close packing of N with Cr in octahedral coordination. Cr occupies a Wyckoff position 2c and N partially (nearly 50%) occupies a 2a position with Z = 1. The ordering of N in alternating 2a positions leads to the trigonal P$\overline{3}$1m structure first reported by Vallas and Calvert [19], which appears as a consequence of aging of disordered h.c.p. Cr₂N [20]. Electron densities and geometric details are in keeping with a Cr₂N stoichiometry. Atom coordinates and atomic displacement parameters are given in

Table 2. Atomic coordinates and equivalent isotropic displacement parameters (Ų × 10³) for oreillyite. U_eq is defined as one third of the trace of the orthogonalized U^ij tensor.

Site	x	Y	z	Ueq	s.o.f.
Cr	1/3	2/3	1/4	12(2)	1
N	0	0	0	15(11)	N0.47(9)

s.o.f. = site occupancy factor.

. Bond distances are as follows: Cr–N = 1.9466(2) (×6) and Cr–Cr = 2.7430(3) Å, which compares with values of 1.952 Å and 2.779 Å observed by Cabana et al. [21] for the ordered P$\overline{3}$1m structure.

Crystallographic information files for both kishonite and oreillyite are available as Supplementary Materials.

Calculated X-ray powder diffraction data (

Table 3. Calculated X-ray powder diffraction data for kishonite (left) and oreillyite (right). Only reflections with I_calc > 2 are reported. The strongest reflections are given in bold.

h	k	l	dcalc	Icalc	h	k	l	dcalc	Icalc
1	1	1	24.641	100	1	0	1	30.501	3
2	0	0	21.340	42	1	1	0	24.055	14
2	2	0	15.090	27	0	0	2	22.387	21
3	1	1	12.869	26	2	−1	1	21.190	45
2	2	2	12.321	7	1	1	1	21.190	55
4	0	0	10.670	3	1	1	2	16.388	9
3	3	1	0.9791	10	2	−1	2	16.388	10
4	2	0	0.9544	10	3	0	0	13.888	16
4	2	2	0.8712	10	1	1	3	12.682	6
					2	−1	3	12.682	7
					3	0	2	11.802	11
					2	2	1	11.616	4
					4	−2	1	11.616	5
					0	0	4	11.193	2
					4	−2	3	0.9365	3
					2	2	3	0.9365	3
					4	1	1	0.8910	2
					3	0	4	0.8715	10

, in Å for CuKα) for both kishonite and oreillyite are given in Table 3.

5. Transmission Electron Microscopy (TEM) for Oreillyite

FIB foils from the region of interest (bottom part of Figure 2) were prepared for TEM using a dual-beam FIB system (FEI Helios G3 CX). High-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) imaging and element mapping were carried out using an FEI Titan G2 80–200 TEM/STEM with ChemiSTEM Technology operating at 200 kV at the Centre for Microscopy, Characterization and Analysis (CMCA), The University of Western Australia, Perth, Australia. The element maps were obtained by energy dispersive X-ray spectroscopy using the Super-X detector on the Titan with a probe size ~1 nm and a probe current of ~0.25 nA. Total acquisition times of 20–30 minutes were used to obtain good signal-to-noise ratios. Electron diffraction was carried out using a field limiting aperture that selected an area approximately 600 nm in diameter.

Synthetic Cr₂N can exhibit an ordered (a = 4.81, c = 4.48 Å; S.G. P$\overline{3}$1m) or a disordered (a = 2.75, c = 4.48 Å; S.G. P6₃/mmc) structure [22]. The X-ray study (see above) did not reveal the presence of superstructure reflections leading to the trigonal cell (P$\overline{3}$1m model), which were masked by the extremely small size of the crystal and the weak nature of these reflections. The TEM study was, therefore, especially focused to try to elucidate the degree of disorder.

Two selected area electron diffraction (SAED) patterns are shown in Figure 6.

The first (Figure 6a) was matched to the [100] zone axis of the P$\overline{3}$1m Cr₂N structure [21,22,23]. The spacings for the two nearest spots making up the rectangle are 4.4 Å (corresponding to the spacing of (001) planes) and 4.1 Å (corresponding to the spacing of (010) planes), in agreement with what observed for the P$\overline{3}$1m structure. The other SAED pattern (Figure 6c) was matched to the [210] zone axis of the P$\overline{3}$1m Cr₂N structure. The spacings for the two nearest spots making up the rectangle are 4.4 Å (corresponding to the spacing of (001) planes) and 2.4 Å nm (corresponding to the spacing of (−120) planes), again in agreement with that observed for the P$\overline{3}$1m structure. The calculated diffraction patterns for the P$\overline{3}$1m structure down the [100] and [210] zone axes (Figure 6b,d) match the experimental SAED patterns in Figure 6a,c. As can be seen, the ~4.1 Å spacing we observe (Figure 6b) is only possible with the geometry observed for the trigonal modification (corresponding to the 010 reflection). Even if this reflection should be very weak (I_calc < 2) in the P$\overline{3}$1m structure (see Table 1), its intensity is enhanced because of dynamic effects in the thick foil.

As additional confirmation, if we take into consideration the α and β sample tilts used to obtain the SAED patterns in Figure 6b,c (α = 15.88, β = −4.07 and α = −14.43, β = −1.76, respectively) it is evident that we tilted approximately 30° to get from one zone axis to the other. The calculated angle between these two axes in the P$\overline{3}$1m space group is actually 30.0°.

Thus, we conclude that oreillyite exhibits the ε-type P$\overline{3}$1m ordered structure (Figure 7), although the single-crystal X-ray experiment was able to detect only the reflections belonging to the hexagonal substructure. It consists of a hexagonal close-packed (h.c.p.) sublattice of Cr with interstitial N atoms occupying some of the octahedral interstices with N-Cr distances of 1.959 Å ([22]; to be compared to 1.9466(2) Å obtained in the disordered model obtained by single-crystal X-ray diffraction). The trigonal cell of the ordered structure can be obtained from the disordered hexagonal one by applying the transformation matrix |2 1 0/−1 1 0/0 0 1|. The cell so obtained for oreillyite is: a = 4.7853(5) Å, c = 4.4630(6) Å, V = 88.51 ų, and Z = 3. Oreillyite is very hard and stable at high-pressure and temperature according to first principles calculations by Ma et al. [23], who calculated a bulk modulus of 289 GPa and a shear modulus of 145 GPa by density functional theory.

6. Remarks on the Origin of Kishonite and Oreillyite

The coarse (cm-size crystals) V-bearing hibonite + grossite + spinel aggregates containing kishonite appear to represent a late stage of the evolution of this magma [4]. The grain sizes, magmatic microstructures, and evidence for an extended sequence of magmatic crystallization [5,7] clearly rule out more speculative origins such as lightning strikes [24], meteorite impact, or shallow hydrothermal circulation. Petrographic evidence for the reaction liquid + corundum → anorthite, and the stability of grossite suggest that this process operated at or near to the base of the crust (25–30 km depth; [3]), and at temperatures ranging from 1450–1500 °C to a minimum of >1150 °C.

Oreillyite was found inside a 1.7 mm crystal of corundum with a hollow centre and raised rims, indicating hopper growth, and with an intense purple-red colour (Figure 2; [5]). The surface of the crystal is decorated with balls of native chromium up to 60 microns across, and very small grains of an Fe-Ni alloy. Corundum is strongly zoned in Cr from central parts with 1–4% Cr₂O₃, rising to mean values of ca 23 wt.% Cr₂O₃ toward both the outer rim and the edge of the central cavity. Single electron microprobe spots in these zones contain up to 33% Cr₂O₃, and corundum adjacent to the chromium balls has mean values >31 wt.% Cr₂O₃. The most Cr-rich portions of the crystal are finely granular in BSE imagery, suggesting an intergrowth of higher- and lower-Cr corundum on the scale of a few microns.

The two ultra-reduced minerals described here imply coexistence with a strongly hydrogen-dominated fluid [3,4], probably in localized systems isolated from the oxidized lithospheric mantle. Such fluids can be derived from a metal-saturated sublithospheric mantle, where fO₂ is buffered near IW, and any COH fluid consists almost entirely of CH₄ + H₂; similar fluids are observed in sublithospheric diamonds [25,26] and predicted theoretically [27,28,29]. The existence of such fluids and their movement through the lithosphere could have major implications for the transport of carbon, hydrogen, and other volatile species from the deep mantle to the surface.