Factors Controlling the Gallium Preference in High-Al Chromitites

Abstract

Gallium (Ga) belongs to the group of critical metals and is of noticeable research interest. Although Ga³⁺ is highly compatible in high-Al spinels a convincing explanation of the positive Ga³⁺–Al³⁺ correlation has not yet been proposed. In the present study, spinel-chemistry and geochemical data of high-Al and high-Cr chromitites from Greece, Bulgaria and the Kempirsai Massif (Urals) reveals a strong negative correlation (R ranges from −0.95 to −0.98) between Cr/(Cr + Al) ratio and Ga in large chromite deposits, suggesting that Ga hasn’t been affected by re-equilibration processes. In contrast, chromite occurrences of Pindos and Rhodope massifs show depletion in Ga and Al and elevated Mn, Co, Zn and Fe contents, resulting in changes (sub-solidus reactions), during the evolution of ophiolites. Application of literature experimental data shows an abrupt increase of the inversion parameter (x) of spinels at high temperature, in which the highest values correspond to low-Cr³⁺ samples. Therefore, key factors controlling the preference of Ga³⁺ in high-Al chromitites may be the composition of the parent magma, temperature, redox conditions, the disorder degree of spinels and the ability of Al³⁺ to occupy both octahedral and tetrahedral sites. In contrast, the competing Cr³⁺ can occupy only octahedral sites (due to its electronic configuration) and the Ga³⁺ shows a strong preference on tetrahedral sites.

1. Introduction

Gallium (Ga) is a vital metal for the economy, due to its use in high-technology applications, such as electronics industry, electric cars, solar panels. Although bauxite deposits are traditionally mined for their Al content and are important sources of Ga as a byproduct commodity [1], the distribution of Ga in chromite ores may be of particular research interest, due to its relationship with the major element composition of chromite.

Chromite belongs to the subgroup of spinels, which accommodate a wide variety of cations in their structure with the general formula AB₂O₄. Many authors emphasized that despite their simple structure, many spinels exhibit complex disordering phenomena involving the two cation sites, which play an important role both in their thermochemical and their physical properties [2,3,4,5,6]. The movements of cations between tetrahedral and octahedral sites, as a result of cation substitution, have been discussed under the aspect of structural parameters, such as tetrahedral and octahedral bond lengths, cation-cation and cation-anion distances, bond angles and hopping lengths, which were calculated by experimental lattice constants and oxygen parameters [2,3,4,7,8,9,10,11,12]. The ability of Ga³⁺ (r = 0.62 Å) to replace Al³⁺ (r = 0.54 Å) in aluminum minerals is related to their geochemistry (Group III of the periodic table), while Ga might be expected to behave in a similar way in chromite and magnetite as they share similar ionic radii (Cr³⁺, r = 0.62 Å; Fe³⁺ = 0.64 Å ) for octahedral and tetrahedral coordination (Al³⁺, r = 0.39; Ga³⁺ = 0.47; Fe³⁺ = 0.49) [3]. Gallium levels reported in Cr-spinel grains from ophiolites, varying from 10 to 50 ppm [13,14] are consistent with experimental mineral-melt data on the partition coefficient (D_Ga = 0.9–11.2) [15]. Although mineral–melt partition coefficients are not constants, depending on a number of factors (pressure, temperature, oxygen fugacity or mineral and melt composition) on the basis of experimental data it has been suggested that Ga is volatile and there is no significant effect of temperature, magma composition and at very low oxygen fugacity conditions [15].

A tectonic discrimination of peridotites, using the oxygen fugacity (fO₂)–Cr#[Cr/(Cr + Al)] diagram, and the Ga–Ti–Fe³⁺# [Fe³⁺/(Fe³⁺ + Cr + Al)] systematics in chrome-spinels, has been proposed [16]. A negative correlation between Ga and Cr in chromitites has been established, that may be related to the composition of parental magmas [17,18,19,20,21,22,23,24] or to the outer electronic structure of Ga that is similar to that of Al [25]. Also, the investigation of spinels in lithospheric mantle xenoliths from distinct tectonic settings has demonstrated that trace elements contribute in discriminating between spinels hosted in peridotites and those crystallized from magmas [11]. However, a convincing explanation of the positive correlation between Ga³⁺ and Al³⁺ has not yet been offered.

In the present study we characterize the spinel chemistry, bulk ore composition, including Ga, from chromitite samples of selected ophiolite complexes in Greece (Pindos, Central Vourinos and Skyros), the Rhodope–Serbo–Macedonian zone (SMZ) massifs (all of Mesozoic age) and the Kempirsai Massif (Kazakhstan) in the Urals (Palaeozoic age), all hosting both high-Cr and high-Al chromitites. The investigated samples are representative of large chromite deposits and small occurrences, in order to define potential relationships between major, minor or trace elements and Ga and the effect of re-equilibration processes, during a long evolutionary time of the ophiolites. We apply available platinum-group element (PGE) data to define potential correlations between the Ga content and fractional crystallization, and experimental literature data for the structure of spinels, aiming to investigate the role of intra-crystalline cation exchange, and contribute to still uncertain factors controlling the positive Al-Ga correlation of chromitites.

2. Materials and Methods

2.1. Mineral Analysis

Polished sections of all chromitite samples were examined using a reflected light microscope and a scanning electron microscope (SEM), equipped with energy-dispersive spectroscopy (EDS). The SEM-EDS back-scattered electron images (Figure 1) and analyses of chromite ores (

Table 1. Electron scanning electron microscope (SEM)/energy-dispersive spectroscopy (EDS) analyses of chromite from chromitites of Greece, Bulgaria and Kempirsai (Urals).

Vourinos						Pindos							Skyros
wt%	Vour. 1	Vour. 2	Vour. 3	Vour. 4	Vour. 5	Pi.1.	Pi.2	Pi. 3	Pi. 4	Pi. 5	Pi.6	Pi.7	Sky. 1	Sky. 2	Sky. 3
TiO2	0.3	0.2	0.2	0.1	0.2	0.2	0.3	0.2	0.2	0.1	0.2	0.1	0.2	0.2	0.2
Al2O3	11.4	9.4	9.8	23.9	11.3	26.2	27.2	32.9	34.5	16.5	15.5	8.2	20.11	24.5	22.1
Cr2O3	59.7	63.2	61.3	44.8	60.4	41.5	39.5	35.9	34.8	52.8	52.1	61.4	48.8	45.3	47.1
MgO	13.6	12.1	13.6	16.2	12.4	14.5	13.6	16.4	16.7	12.4	10.1	11.7	13.9	15.7	10.2
FeO	13.1	14.1	12.7	11.1	14.6	13.5	15.7	12.3	12.5	15.6	18.7	14.9	13.6	11.8	19.7
Fe2O3	1.7	0.1	2.1	4.2	.5	3.4	3.1	2.3	2.6	2.9	2.7	2.8	2.2	2.3	0.3
MnO	0.3	0.2	0.2	0.2	0.2	0.3	n.d.	0.1	n.d.	0.2	0.3	0.3	0.1	0.1	0.2
NiO	0.2	0.1	n.d.	n.d.	0.2	0.3	n.d.	0.2	n.d.	0.2	0.1	0.2	0.1	0.2	0.2
Total	100.3	99.5	99.9	100.5	99.8	99.9	99.4	100.1	101.2	100.7	99.7	99.6	99.01	100.1	100
Cr/(Cr+Al)	0.77	0.81	0.81	0.56	0.78	0.52	0.49	0.42	0.4	0.70	0.69	0.81	0.62	0.55	0.58
Mg/(Mg+Fe2+)	0.65	0.62	0.66	0.72	0.60	0.65	0.61	0.71	0.72	0.59	0.49	0.58	0.64	0.69	0.51
Fe3+/(Cr+Al+Fe3+)	0.031	0.000	0.025	0.036	0.0055	0.038	0.036	0.025	0.027	0.0413	0.028	0.036	0.027	0.026	0.0033
Numbers of Cations on the Basis of 32 Oxygens
Ti	0.019	0.039	0.040	0.018	0.039	0.036	0.054	0.035	0.014	0.019	0.039	0.020	0.037	0.036	0.038
Al	3.426	2.889	2.997	6.736	3.451	7.415	7.758	8.966	9.241	4.650	4.728	2.555	5.896	6.922	6.532
Cr	12.037	13.031	12.523	8.471	12.383	7.891	7.558	6.563	6.254	10.645	10.662	12.835	9.602	8.586	9.340
Mg	5.170	4.703	5.250	5.774	4.790	5.210	4.906	5.652	5.658	4.712	3.896	4.611	5.156	5.610	3.813
Fe2+	2.744	3.271	2.744	2.203	3.164	2.706	3.149	2.326	2.377	3.222	4.056	3.300	2.840	2.367	4.142
Fe3+	0.498	0.001	0.400	0.757	0.087	0.620	0.575	0.401	0.436	0.659	0.533	0.570	0.427	0.420	0.053
Mn	0.065	0.044	0.045	0.040	0.044	0.061	0.000	0.020	0.000	0.043	0.065	0.067	0.021	0.020	0.042
Ni	0.157	0.021	0.000	0.000	0.042	0.058	0.000	0.037	0.000	0.041	0.020	0.042	0.020	0.039	0.040
Rhodope Massif											Urals Kempirsai
	Skyros		Greece			Bulgaria
									Goliamo		Northern Part		Southern Part
wt%	Sky. 4	Sky. 5	Soufli 1	Soufli 2	Gomati	Broucevci	Jacovitsa	Pletena	Kamenyane 1	Kamenyane 2	Batamshinsk		Main Ore Field
TiO2	0.1	0.2	n.d.	0.2	0.2	0.4	0.2	0.5	0.4	0.3	0.3	0.3	0.1	0.2
Al2O3	11.2	12.6	15.8	19.6	30.8	27.9	5.5	10.4	4.5	1.2	24.4	23.3	9.3	9.8
Cr2O3	59.1	58.1	53.9	47.8	35.5	37.3	58.8	49.8	33.2	24.9	46.6	47.1	61.3	60.2
MgO	13.8	14.4	13.4	14.1	15.3	16.6	10.2	8.2	11.9	7.2	14.3	14.2	15.5	15.2
FeO	12.1	11.5	13.4	13.5	13.9	10.4	16.2	20.7	13.6	20.2	14.1	14.1	9.9	10.3
Fe2O3	2.4	2.2	3.1	5.0	5.1	6.0	7.2	9.5	35.1	45.6	0.3	0.8	4.5	4.3
MnO	0.2	0.2	0.2	0.2	n.d.	0.2	0.6	0.5	n.d.	0.2	n.d.	n.d.	n.d.	n.d.
NiO	0.2	0.2	n.d.	0.2	n.d.	0.2	0.2	0.2	0.3	0.4	0.2	0.2	0.1	0.3
Total	99.1	99.7	100.2	99.8	100.3	99.1	99.1	100	99.2	99.9	100.4	99.9	100.8	100.3
Cr/(Cr+Al)	0.75	0.79	0.70	0.63	0.44	0.47	0.88	0.74	0.83	0.94	0.56	0.58	0.82	0.80
Mg/(Mg+Fe2+)	0.68	0.69	0.62	0.65	0.66	0.72	0.52	0.37	0.52	0.38	0.64	0.65	0.74	0.73
Fe3+/(Cr+Al+Fe3+)	0.03	0.026	0.036	0.059	0.057	0.068	0.093	0.123	0.455	0.62	0.0015	0.009	0.053	0.052
Numbers of cations on the basis of 32 oxygens
Ti	0.019	0.038	0.019	0.036	0.035	0.035	0.041	0.102	0.082	0.062	0.054	0.054	0.019	0.038
Al	3.414	3.813	4.636	5.591	8.475	7.828	1.767	3.277	1.452	0.386	6.987	6.,675	2.787	2.947
Cr	12.066	11.695	10.745	9.396	6.553	7.021	12.670	10.582	7.168	5.648	8.880	9.070	12.323	12.143
Mg	5.321	5.512	5.036	5.231	5.324	5.890	4.143	3.291	4.858	3.088	5.137	5.151	5.874	5.780
Fe2+	2.613	2.442	2.940	2.805	2.711	2.067	3.715	4.653	3.114	4.839	2.879	2.864	2.125	2.197
Fe3+	0.480	0.414	0.580	0.940	0.901	1.079	1.481	1.936	7.215	9,844	0.024	0.146	0.852	0.833
Mn	0.044	0.044	0.043	0.000	0.000	0.040	0.138	0.115	0.000	0.046	0.000	0.000	0.000	0.000
Ni	0.041	0.041	0.000	0.000	0.000	0.038	0.044	0.044	0.114	0.088	0.039	0.038	0.020	0.062

) were carried out at the Faculty of Geology and Geoenvironment, National and Kapodistrian University of Athens (NKUA), using a JEOL JSM 5600 (Tokyo, Japan), scanning electron microscope, equipped with ISIS 300 OXFORD (Oxford shire, UK), automated energy dispersive analysis system. Analytical conditions were 20 kV accelerating voltage, 0.5 nA beam current, <2 μm beam diameter and 50 s count times. The following X-ray lines were used: FeKα, NiKα, CoKα, CuKα, CrKα, AlKα, TiKα, CaKα, SiKα, MnKα and MgKα. Cr, Fe, Mn, Ni, Co, Ti and Si, MgO for Mg and Al₂O₃ for Al. Contents of Fe₂O₃ and FeO were calculated on the basis of the spinel stoichiometry.

2.2. Whole Rock Analysis

Major and trace elements in massive chromitite samples (more than 95 vol %) were determined by ICP-MS analysis, at the ACME Analytical Laboratories Ltd., Vancouver, BC, Canada (currently Bureau Veritas Commodities Canada Ltd.). The samples were dissolved using an acid mix (HNO₃–HClO₄–HF) digestion and then the residues were dissolved in concentrated HCl. The rare earth elements La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb and Lu as well as Li, K, Ge, Sr, Y, Zr, Mo, Sb, Cs, W, Pb, Th, and U were lower than the detection limits of the analytical methods. The detection limits of the method for the presented elements are 1 ppm for Ga and V, 0.2 ppm for Co and Zn, 0.1 for Ni and 0.01 wt % for Fe. On the basis of the quality control report provided by the Analytical Labs, the results of analyses of the reference material in comparison to expected values, and the results from multistage analysis of certain samples, showed accuracy and precision in good agreement with accepted values for international standards. The analytical error, for Ga, for example, was <5%. Although the PGE data reported in

Table 2. Trace element contents of high-Cr and high-Al chromitites.

Location	SEM/EDS		Trace Element (ppm)					wt%	ppb
	Cr/(Cr+Al)	Mg/(Mg+Fe2+)	Ni	Co	V	Zn	Ga	Fe	ΣPGE *	Pd/Ir *
Vourinos 1	0.77	0.65	2000	240	500	260	14	8.1	140	0.06
Vourinos 2	0.81	0.62	1580	210	560	550	15	7.9	135	0.35
Vourinos 3	0.81	0.66	1900	200	400	300	13	8.72	92	0.55
Vourinos 4	0.56	0.72	1800	170	780	360	27	9.2	30	0.54
Vourinos 5	0.78	0.61	1800	180	620	280	14	7.76	109	0.63
Pindos 1	0.52	0.63	1300	140	600	400	23	7.84	51	1.0
Pindos 2	0.48	0.6	1500	260	710	410	32	9.77	143	1.0
Pindos 3	0.42	0.71	1500	260	710	520	34	7.8	117	6.33
Pindos 4	0.4	0.72	1630	240	580	460	32	7.3	6123	34.2
Pindos5	0.67	0.57	750	270	760	520	16	9.7	3875	12.2
Pindos 6	0.69	0.49	720	240	760	620	23	10.1	2098	7.2
Pindos 7	0.81	0.58	1450	290	560	490	17	10.2	181	0.22
Skyros 1	0.61	0.64	1300	250	1200	540	34	10.9	2300	0.08
Skyros 2	0.55	0.69	1600	200	1000	400	36	9.45	464	0.7
Skyros 3	0.58	0.51	1500	240	870	450	40	9.6	251	0.67
Skyros 4	0.75	0.69	1250	220	640	420	14	10.7	145	0.33
Skyros 5	0.79	0.69	1200	200	620	400	11	10.1	145	0.1
Othrys (n = 4)	0.54	0.69	1400	210	960	370	33	9.8	91	0.36
Rhodope Massif
Greece
Soufli1	0.70	0.62	1700	230	380	580	16	13.2	150	0.2
Soufli2	0.63	0.65	1150	220	460	280	12	10.6	82	1.0
Gomati	0.49	0.67	1030	130	730	280	24	9.6	104	0.25
Bulgaria
Broucevci	0.47	0.72	1300	230	790	420	45	11	60	0.92
Jacovitsa	0.88	0.52	2250	310	240	760	9	13.9	197	0.46
Pletena	0.74	0.37	890	290	330	1030	13	17.9	563	0.07
Goliamo Kamenyane 1	0.83	0.52	1550	80	1000	450	12	11.6	87	0.14
Goliamo Kamenyane 2	0.94	0.38	2260	970	370	4030	6	64.3	40	1.93
Kemprsai (Urals)
Northern	0.80	0.73	1600	210	160	160	14	9.1
Batamshinsk	0.82	0.74	1600	230	200	190	16	9.5
Southern	0.56	0.58	1500	240	680	480	48	12.6
XL Let Kazakhstan	0.64	0.65	1700	230	730	340	49	10.1

Symbol * = Data on PGE from literature [26,27,28,29,30,31,32].

have been published previously (Table 2) a portion from the same samples was used for the presented trace element analyses.

3. A Brief Outline of Characteristics for the Studied Chromitites

All chromitite samples selected for the present study come from deposits and occurrences, which have been the subject of detailed geological, mineralogical and geochemical investigation [26,27,28,29,30,31,32,33,34,35,36] and references therein. The main ophiolite complexes of Greece (Vourinos, Othrys and Pindos) belong to the Upper Jurassic to Lower Cretaceous Tethyan ophiolite belt, and are characterized by heterogeneous deformation and rotation, during their original displacement and subsequent tectonic incorporation into continental margins [33]. The studied samples of chromitites are massive (Figure 1) and exhibit variations in the chromitite tonnage, the composition of chromite (Table 1 and Table 2), the degree of transformation of ores and the associated ophiolites [27,28,29,30,31,32,33,34,35,36,37].

Chromite ores in the Vourinos complex occur in the mantle and cumulate sequences, with a tonnage estimated to approximately 10 Mt of high-Cr type, but at the central part of the complex there are high-Cr and high-Al ores in a spatial association, with low PGE contents [26]. The Othrys complex has a relatively high tonnage (approximately 3 Mt) of high-Al massive chromite ores and low PGE content [27].

The chromitite occurrences in the Pindos ophiolite complex are small (a few tens of m (x) a few tens of cm) and are hosted within completely serpentinized and weathered, intensively deformed dunite-harzburgite blocks, due to a strong plastic and brittle deformation that was superimposed on primary magmatic textures [17,18,34]. Chromitites throughout the Pindos complex are high-Cr and high-Al, often in a spatial association. The most salient feature of the Pindos chromitites is the enrichment in Pt and Pd at the area of Korydallos, at a level of 7 ppm PGE_total [28,29] and up to 29 ppm [35]. In the Achladones area on the Skyros island small massive chromitite bodies are of high-Al type and have elevated PGE contents, up to 3 ppm ΣPGE, although both high-Cr and high-Al types having low PGE content are found on the entire island [30].

Ophiolites associated with the Serbomacedonian massif (Gomati) and Rhodope massif including the ophiolites of Soufli (Greece), Dobromirci, Jacovitsa, Broucevci and Goliamo-Kamenjane (Bulgaria) host small (a few thousand tons) high-Cr and high-Al ores in a spatial are association, which occasionally contain elevated PGE concentrations. They are completely serpentinized, locally sheared and metamorphosed to antigorite-tremolite and/or talc schists. Detailed description of the characteristic mineralogy and texture of those chromitites have been published in previous studies [17,18,31,32,36].

The Kempirsai massif, covering an area of 2000 km², is divided by a shear zone into two parts: the southeastern part that is called Main Ore Field (MOF), hosting large high-Cr chromite deposits, and the northwestern area, the so-called Batamshinsk Ore Field (BOF), hosting much smaller high-Al chromite deposits [37]. An excellent description of the petrography and mineral chemistry, including mineral inclusions in the chromite of the giant chromite deposit of Kempirsai has been provided and discussed by Melcher et al [37]. These authors have interpreted their formation by a multistage process: High-Al chromitites may be derived from MORB-type tholeiitic melts, and high-Cr ones from boninitic magmas, during a second stage by interaction of hydrous high-Mg melts and fluids with depleted mantle in a supra-subduction zone setting.

4. Results

4.1. Compositional Variations in Chromite

The chromite samples from the central part of the Vourinos complex, the Pindos, Skyros island, Serbomacedonian, Rhodope and Kempirsai massifs show a wide variation in major elements from high-Cr, with the Cr/(Cr + Al) atomic ratio ranging from 0.81 to 0.69, to high-Al with the Cr/Cr + Al) ratio ranging from 0.63 to 0.4 (Table 1), falling in the range of metallurgical and refractorytype, respectively [21]. In addition, in the Bulgarian Rhodope massif (Jacovitsa, Pletena and Goliamo Kamenyane areas) altered chromite grains are dominant, having relatively high FeO and low Al₂O₃ and MgO contents (Table 1). As a consequence, the Cr/(Cr + Al) atomic ratios of those chromitites are significantly higher than those of high-Cr chromitites from the Vourinos complex (Table 1).

4.2. Distribution of Trace Elements in Chromitites

The geochemical data from whole rock analyses show a wide variation in major and trace element contents (Table 2). Gallium contents are lower in high-Cr chromitites (11 to 23 ppm) compared to high-Al ones (27–49 ppm), that seems to be independent on the degree of fractionation of parent magma, as exemplified by the Pd/Ir ratio [38]. The highest Co, Mn, Zn, Fe and lowest Ga were mainly recorded in strongly altered small chromite occurrences from the Rhodope massif in Bulgaria. They are in a good agreement with other chromitites [21,22,23,24,25] and are independent of the age of the associated ophiolites. Platinum-group elements (PGE) show total contents ranging from 30 to 6120 ppb and Pd/Ir ratios from 0.06 to 34, which are independent of the major element composition of chromitites (Table 2).

The results show a strong negative correlation (R ranges from −0.98 to −0.95) between the Cr/(Cr + Al) atomic ratio and Ga for the relatively large chromite deposits of Vourinos, Kempirsai massif (Urals) and the Skyros island. In addition, there is a less strong negative correlation for small chromite occurrences from the Pindos and Rhodope massifs (R ≥ −0.76 and R ≥ −0.83, respectively). Apart from Ga, the best correlation is found between Cr/(Cr + Al) and V for the Vourinos (R = −0.84), Skyros and Kempirsai (R ≥ −0.93), whereas no significant relationship for chromitites from the Pindos and Rhodope massifs (Figure 2b) or between Cr/(Cr + Al) and other minor and trace elements is observed.

5. Discussion

5.1. Factors Controlling the Spinel Chemistry

5.1.1. Magmatic Versus Post-Magmatic Processes

The wide variation of the Cr/(Cr + Al) atomic ratio for the chromitite samples from Greece, Bulgaria and Kempirsai massifs (Table 1 and Table 2) fall in the range of metallurgical and refractory types. Differences in the trace element content (Table 2) may suggest trace element partitioning, depending on a number of factors, such as temperature, pressure, oxygen fugacity and the chemical composition of parent magmas [17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47].

As the partition coefficient of elements is defined as the ratio of the element content in a mineral and the melt [15], the Pd/Ir ratio can be used as an indicator of the degree of fractionation of parent magma for chromitites [38] and the presence of very low Pd/Ir values (low degree of fractionation) for both high-Cr and high-Al chromitites, suggest their origin from different magma sources. It has been argued that high-Cr chromitites, which have higher Sc, Mn, Co and Ni, and lower Ti, V, Zn and Ga contents may be derived from boninitic magmas, while high-Al ores may be derived from MORB-type tholeiitic magmas [17,18,19,20,21,22,23,24]. Experimental data at high temperature have shown that Ga is compatible in spinel with D values ranging between 0.9 and 11.2, and slightly lower D values in the most reducing experiments, while experimental data at temperatures >1300 °C and low oxygen fugacity have shown that there is no significant effect of temperature, composition and redox conditions [15]. However, the negative correlation between the Cr/(Cr + Al) atomic ratio and Ga content in natural chromitites points to the potential effect of the composition of the parent magma, while a positive trend between the Fe³⁺/(Al + Cr + Fe³⁺) atomic ratio and Ga content for large chromite deposits (Figure 2c,d) may suggest the effect of the redox conditions on the Ga distribution in chromitites. Specifically, high-Cr chromitites formed earlier from a primary magma (under relatively reducing conditions), compared to high-Al ones formed later from an evolved magma (and more oxidized conditions) magma [47].

In addition, differences in the negative correlations between the Cr/(Cr + Al) atomic ratio and Ga content and the slope of correlation lines for the different occurrences (Figure 2) may suggest that in addition to the composition of parent magmas, which is a major factor for large deposits (like Vourinos and Kempirsai massif, Urals) other factors such as temperature, pressure or redox conditions may be responsible for the observed deviation from linearity for small metamorphosed occurrences of chromitites, such as those from the Pindos and Rhodope massifs (Figure 2b). The lack of significant relationships between major and trace elements, in small chromitite occurrences from the Pindos and Rhodope massifs (Figure 2) may be related with post-magmatic processes. The elevated Mn, Co, Zn and Fe contents and depletion in Ga (Table 2) is consistent with the spinel chemistry in the Rhodope massif of Bulgaria, showing a trend of depletion in Ga, in the metamorphic Fe-chromite rims surrounding the cores of chromite grains, implying that most tetrahedral sites are still occupied by Fe²⁺ [17]. In addition, it has been suggested that the Mg cations can be replaced by Mn, Zn or Co, whereas Al and Fe³⁺ compete for the octahedral sites, hampering the entry of Ga [17,18,24].

Despite the recorded modification in trace elements by re-distribution during post-magmatic processes, as exemplified by bulk analysis (Table 1 and Table 2) and spinel chemistry [17,18,24] limited only to relatively small chromitite occurrences, the well-established relationship (R = −0.95 to −0.98) between Cr/(Cr + Al) ratio and Ga (Table 2; Figure 1a) that is comparable to literature data for chromitites hosted in other ophiolite complexes [21,22,23,24,40,41,42,43,44,45,46,47] seems to be a salient feature.

5.1.2. Spinel Structure

The structure of spinel is a cubic close-packed array of 32 oxygen ions, with 64 tetrahedral vacancies and 32 octahedral vacancies in one unit cell each, containing 8 formula units, with the general formula: A²⁺B³⁺₂O₄, where A = Fe²⁺, Mn²⁺, Mg²⁺, Co²⁺, Zn²⁺, Ni²⁺ and B = Fe³⁺, Cr³⁺, Al³⁺, Ga³⁺, V³⁺ [36]. Spinels are traditionally denoted as either “normal”, where the A cation occupies T sites, the B cation occupies M sites, whereas in the “inverse” type cation B occupies the T site and the M site is occupied by both cations A and B [4,8,9,10,11,12]. The degree of inversion x characterizing the cation distribution can show values between x = 0 (normal spinel) and x = 1 (inverse spinel). The spinel structure is able to accommodate many cations (at least 36) by enlarging and decreasing its tetrahedral and octahedral bond distances, while the oxygen positional parameter (u) should be regarded as a measure of distortion of the spinel structure from cubic close packing or as the angular distortion of the octahedron [10,12]. The movements of cations between tetrahedral T and octahedral M sites, as a result of Mg²⁺ substitution, can be discussed based on structural parameters, such as bond lengths, cation-cation and cation-anion distances, bond angles and hopping lengths, which were calculated using experimental lattice constants and oxygen parameters [4,8,9,10,11,12].

5.1.3. Applications to Natural Spinels

Despite post-magmatic compositional changes in Fe-chromite within the chromitite ores, resulting in the remobilization of cations during metamorphism (700 °C to 450 °C) of chromitites [17,18,21,24,32], the structural incorporation of Ga into the chromite lattice is evidenced by the progressive and linear increase of Al or decreasing Cr/(Cr + Al) atomic ratio (Figure 2a). Experimental data on cation distribution versus temperature may provide valuable information related to the preferences of cations in the spinel lattice. MgAl₂O₄ is the most prominent example of a normal spinel because Mg²⁺ is much larger than Al³⁺ [2,6,48]. Gallium (Ga³⁺) is smaller than Mg²⁺, but significantly larger compared to Al³⁺, leading to an ordering which is called mainly inverse, at least for the end-member composition MgGa₂O₄ [4].

The intra-crystalline exchange reaction in spinels has been modeled [49,50] and the order-disorder process has been described by the following exchange reaction: ^TAl + ^MMg = ^TMg + ^MAl (where T = tetrahedral and M = octahedral site) in which the forward reaction implies an exchange of Mg with Al at the M site (ordering process), and backwards (disordering process). The cation distributions for both disordering and ordering experiments were obtained by measuring the oxygen positional parameter (u), the inversion parameter (x) (Al in T) site, using samples with varying composition. The Mueller kinetic model was satisfactory applied to the experimental data and allowed the calculation of the kinetic ordering constants K, linearly related to temperature by means of Arrhenius equations [48,49,50,51].

Martignago et al. [52] performed crystal structure refinements on three natural spinels, on low-Cr spinel containing small Fe³⁺ quantities and two other samples with high Cr (8.4 wt % Cr₂O₃) and low Cr (3.3 wt % Cr₂O₃) contents [52]. These experiments have shown that both parameters (u) and (x) remained constant for the three different samples up to 600 °C, independently of Cr³⁺ contents. The distortion of spinels, started at higher temperature, near to 650 °C (Figure 3). The degree of distortion at the highest temperatures is inversely correlated with Cr³⁺ contents. Since Cr³⁺ has a tendency to be completely partitioned on the octahedral site, due to its electronic configuration and size [2,52,53,54], Al³⁺ cation is unable to substitute Cr³⁺. The most salient feature derived from the above experimental data [52] is the abrupt increase of the inversion parameter (x) having the highest values for the sample with the lowest Cr content (L-Cr sample), and the lowest values for the sample with the highest Cr content (H-Cr) [52].

Such structural changes should cause modifications of the structural, physical and thermal properties of the spinels. Although the spinel structure is complicated and determination of several parameters may be required, the above results [52], which show that the Cr content in spinels affects the occupancy of Al in the tetrahedral site, may suggest that the co-existence of Ga³⁺ in high-Al chromitites (Table 2) [2,4,5,6,7,8,9,10,11,12] is related to the degree of disorder that is inversely correlated with Cr contents.

Therefore, potential controlling factors on the Ga preference in high-Al chromitites are (a) the composition of the parent magma (geotectonic setting), temperature and redox conditions, (b) the electronic configuration of Cr³⁺ resulting in occupation of M sites only, the ability of Al³⁺ to occupy T and M sites and the strong preference of Ga³⁺ to T sites, and (c) elevated values of the inversion parameter (x) that is inversely correlated with Cr³⁺ content and favors the co-existence of Ga³⁺ in high-Al chromitites.

6. Conclusions

The presented geochemical and mineral chemistry data on chromitites associated with ophiolite complexes, in conjunction with experimental literature data allowed us to draw the following conclusions:

(1)

The lower Ga contents in high-Cr chromitites (11 to 23 ppm) compared to high-Al ones (27–49 ppm) suggest that the composition of the parent magma may be a major factor controlling the preference of Ga in high-Al chromitites.

(2)

The positive trend between the Fe³⁺/(Al + Cr + Fe³⁺) atomic ratio and Ga content for large chromite deposits may suggest the effect of the redox conditions on the Ga distribution in chromitites.

(3)

Plot of the Cr/(Cr + Al) atomic ratios versus Ga content exhibits differences in terms of the slope of correlation lines for the different occurrences, suggesting that, in addition to the composition of parent magmas, other factors such as temperature, pressure or redox conditions may affect the observed deviation from linearity for small metamorphosed chromitite bodies.

(4)

The depletion of Ga and Al, and elevated Mn, Co, Zn and Fe contents in certain small chromitite occurrences, transformed during post-magmatic metamorphism, suggest potential change of the Ga content in Cr-spinel during sub-solidus reactions.

(5)

Assuming that low-Cr spinel is characterized by the highest value of the inversion parameter (x) at higher than 650 °C, then the high Al content in spinels may be a driving force for the degree of inversion in the structure that facilitate the substitution of Al³⁺ for Ga³⁺ at magmatic conditions.