Hydrothermal Scheelite Associated with Upper Cretaceous Intrusions in Romania: A Mineralogical Insight to the W Metallogeny

Abstract

Hydrothermal scheelite from three Romanian occurrences was analyzed in order to ascertain its structural, physical, vibrational, paragenetic, and crystal-chemical peculiarities as an important tool for characterizing the metallogenetic behavior and facilitating the ore-processing. All three occurrences, i.e., Ciclova and Oravița in Banat and Băița Bihor in the Bihor Mountains, are related to skarn deposits developed at the contact of Upper Cretaceous granodioritic bodies with Mesozoic calcareous deposits. Typical crystals show {001}, {111}, and {101} forms and are up to 15 mm across. The structure was successfully refined as tetragonal, space group I4₁/a, with R₁ = 0.0165 (Ciclova), 0.0204 (Oravița), and 0.0237 (Băița Bihor), respectively. The cell parameters refined for the same samples are a = 5.2459(10) Å and c = 11.3777(5) Å at Ciclova, a = 5.2380(2) Å and c = 11.3679(8) Å at Oravița, and a = 5.2409(2) Å and c = 11.3705(6) Å at Băița Bihor. The multiplicity of bands in both infrared absorption and Raman spectra is consistent with the S₄ punctual symmetry of the tungstate anion, agreeing with the structural data. In all cases, the analyzed scheelite is close to the CaWO₄ end-member. Cathodoluminescence peculiarities at the level of single crystals suggest that they crystallized in a slightly oxidizing to reducing environment from late hydrothermal solutions. Textural and paragenetic peculiarities suggest that scheelite from the three occurrences crystallized from epithermal, low-temperature, fluoride- and boron-bearing aqueous solutions.

1. Introduction

Scheelite, ideally CaWO₄, accounts for a significant source of tungsten. Present in about 65% of the tungsten deposits throughout the world, the mineral is of major economic importance (e.g., [1,2,3] and referred works). Due to its very significant economic importance and high supply risk, tungsten was listed as one of the 32 critical materials in Europe (e.g., [4]) and 50 in the USA (e.g., [5]) and is also considered a hi-tech metal [6]. As such, tungsten is primarily used to make wear-resistant metals but also in manufacturing electrodes, heating elements, field emitters, filaments in light bulbs, and cathode ray tubes. The detailed mineralogical study of scheelite as a major source for tungsten is then of great importance, offering an important tool to facilitate ore processing. Scheelite itself is widely used in scintillators for X-ray and gamma-ray detection, whereas compounds with scheelite structure were found to have extra low thermal conductivity [7] and were also investigated as photoanodes [8], photocatalysts [9], or even for nuclear waste immobilization [10]. Last but not least, scheelite can be used, in special conditions, as an appreciated gemstone [11,12].

Reported for the first time at the beginning of the 20th century [13], scheelite from Romania characteristically occurs in skarn deposits, the best-known occurrences being those at Ciclova (e.g., [13,14]), Oravița [14,15], and Băița Bihor: [14,16]. In all three of these occurrences, complex characterization mineralogy is lacking. The goal of this study is to fill the gaps existing in the mineralogical knowledge of the Romanian scheelite associated with Upper Cretaceous intrusions, insisting on its structural, physical, vibrational, paragenetic, and crystal-chemical peculiarities, of real use not only for mineralogists but also for prospectors, ore processors, and metallurgists. Last but not least, we appreciate that scheelite could represent, at least at Băița Bihor, an important component of a complex tungsten ore.

2. Geological Setting

The Ciclova, Oravița, and Băița Bihor occurrences are related to proximal or distal skarns developed at/near the contact of granodiorite bodies ascribed to the so-called “Banatitic Magmatic and Metallogenetic Belt”—BMMB [17]. The BMMB consists of a series of discontinuous magmatic and metallogenetic districts that are discordant over a Middle Cretaceous-aged nappe structure ([18,19,20,21,22] and references therein). The intrusive bodies are ascribed to calc-alkaline hydrated magmas and belong to I-type/magnetite series granitoids, derived from a primitive monzodiorite parent with a deep crust or upper mantle source region. These magmatic complexes (high-K calc-alkaline to calc-alkaline) are all productive in terms of ore formation [17,23].

The formation and evolution of the BMMB was early ascribed to subduction models involving either slab-rollback or slab-tear mechanisms affecting the subduction front of the Vardar Ocean ([20] and references therein). Other authors regarded magmatites in BMMB as post-collisional type I magmatites and ascribed the abundance of ore deposits in the eastern part of the Alpine–Balkan–Carpathian–Dinarides (ABCD) realm to a slab break-off of subducted lithosphere fragments during the Vardar Ocean closure, which imposed the rolling back and steepening of slabs. This process leads to upper crust extension and favors the access of melts to high crustal levels. The Cenozoic collision between the Tisia and Dacia blocks might be responsible for overall crustal thickening, for the formation of extensional faults that facilitated magma upwelling, and for a clockwise rotation of 60–80° with respect to the initial east-west trend at the time of magmatic emplacement ([19,21,22], and references therein).

The locations of the three occurrences in the general context of BMMB are given in Figure 1.

In all three occurrences, the granitoids were emplaced at shallow depths. Plutons show little evidence for differentiation, which is observed in other skarn systems (e.g., Fe, Au, and Zn skarns: [25]). Specific information on the analyzed bodies and on the skarn protolith is reported in

Table 1. Geological features of the analyzed scheelite-bearing skarns.

Occurrence	Intrusion	Age of Intrusion	Protolith	Age of Protolith	Structural Unit
Ciclova Oraviţa	Ciclova-Oraviţa pluton	79 ± 3 to 74 ± 3 Ma [26] (1) 73.9 ± 3.2 to 71.2 ± 4.1 Ma [27] (2) 86.77 ± 0.5 to 87.70 ± 0.5 [20] (3)	carbonated veins in endoskarn	Mesozoic	Locva unit
Băiţa Bihor	Bihor batholith	77 ± 3 to 67 ± 3 Ma [28] (4) 70 ± 5 Ma [29] (5) 80.63 ± 0.3 to 78.69 ± 0.4 Ma [20] (3) 80.3 ± 1.6 Ma [27] (2)	dolostones	Anisian-Carnian; Carnian-Norian	Vălani unit; Vetre unit

⁽¹⁾ K-Ar on biotite; ⁽²⁾ U-Pb on zircon; ⁽³⁾ Re-Os data on molybdenite; ⁽⁴⁾ K-Ar ages on whole rock or on “femic” minerals; ⁽⁵⁾ Rb-Sr isochron.

.

Skarn deposits are one of the most representative types of mineralization from BMMB [17]. Copper metallogeny is predominant and distinguishes the BMMB in the context of the larger Alpine-Balkan–Carpathian–Dinaride belt [19]. A Cu-Mo zone can be distinguished in most of the Romanian deposits, including the Ciclova, Oravița, and Băița Bihor occurrences. Towards the external zones, Mo-rich ores grade into Mo-W-Bi-Te (in calcic skarns) or Cu-W-Bi (in magnesian skarns) [17,30,31].

Scheelite forms significant concentrations in the Cu-Mo (W) mineralization of Băița Bihor [30] and Oravița–Ciclova [15]. At Băiţa Bihor, scheelite occurs in distal magnesian skarns forming typical metasomatic columns whose outer zones contain magnesian borates (i.e., ludwigite, kotoite, suanite, fluoborite, and szaibélyite), whereas the inner zones typically contain magnesian silicates (e.g., diopside, forsterite, humite-group minerals, and phlogopite) [32]. In this deposit, scheelite typically associates with chalcopyrite, galena, sphalerite, molybdenite, bornite, Bi-Cu sulphosalts, and Bi tellurides (e.g., [33,34]).

The scheelite occurrences from Ciclova and Oraviţa subscribe to vein systems developed at the level of endoskarn. The typical mineral association comprises, beside calcite, chalcopyrite, arsenopyrite, pyrite, gersdorffite, glaucodot, fluorite, and, according to [15], adularia, cobaltite, cubanite, molybdenite, and Bi-Cu sulphosalts.

3. Materials and Methods

The samples described in this paper were taken from three waste dumps pertaining to the Lobkowitz Gallery, located on Anton Brook in Ciclova (45°03′09″ N–21°44′25″ E); the Emil Gallery, located on Chinisea Valley in Oravița (45°04′42″ N–21°43′16″ E); and a collector located inside the Baia Roşie basal (transportation) Gallery, emergent on Crişul Negru Valley, in Băița Bihor (emergence: 46°29′14″ N–22°36′53″ E). Galleries from Ciclova and Oravița were mined for copper during the second half of the 19th century and reopened in the 50th year of the 20th century. The gallery at Băița Bihor is still active and is used for the Cu ore evacuation. The samples from Băița Bihor probably originate from Antoniu or Baia Roşie skarn bodies (metasomatic columns), where scheelite is particularly abundant [16].

Electron-microprobe analyses (EMPA) were performed using a Jeol JXA 8530F Hyperprobe (JEOL Ltd., Tokyo, Japan), equipped with a field emission-assisted thermo-ionic (Schottky) emitter, five wavelength-dispersive spectrometers (WDS), and one SD EDS detector. The apparatus is hosted by Rice University, Department of Earth, Environment, and Planetary Science (Houston, TX, USA). The apparatus was set at an accelerating voltage of 15 kV and a beam current of 20 nA (measured at the Faraday cup), for beam diameters of 5–10 µm. Natural chalcopyrite (Cu Kα), natural periclase (Mg Kα), natural hematite (Fe Kα), natural rhodonite (Mn Kα), natural scheelite (W Kα), natural molybdenite (Mo and S Kα), natural galena (Pb Kα), natural diopside (Ca Kα), metallic Bi (Bi Kα), and metallic Sb (Sb Kα) served as standards. Counting time was 20 s per element. Data were reduced using the Phi-Rho-Z matrix correction [35].

The fast identification of mineral phases in all samples was performed by Energy-Dispersive Spectrometry (EDS) analysis, using a JEOL Silicon Drift (SD) X-ray Detector with a 10 mm² active area and 133 eV resolution. The detector is attached and integrated into the JEOL JXA 8530F Hyperprobe described before. The analytical conditions used for EDS analysis were as follows: 15 kV accelerating voltage, 20 nA beam current, live time 20 s. Dead Time (DT) during the analysis was 35%–40% with count rates ranging from ~45,000 to ~100,000. The beam size used was “spot” size (~300 nm). Cathodoluminescence (CL) imaging was performed using a Gatan CL system (Gatan Inc., Pleasanton, CA, USA). Images were acquired at a voltage of 10 kV and contrast-enhanced for better observation.

X-ray powder diffraction (XRD) analysis of selected samples was performed on two different Brucker (AXS) D8 Advance diffractometers equipped with Bruker AXS Hi-star Area Detectors (Bruker, Karlsruhe, Germany) hosted by the Geological Institute of Romania (Bucharest) and by the University in Liège, Belgium, respectively. Both diffractometers used Ni-filtered CuKα radiation (Kα₁ = 1.54056 Å) with Cu anode Kristalloflex K780 X-ray tubes, a step size of 0.02° 2θ, and a counting time of 6 s per step. An operating voltage of 40 kV for a current of 30 mA, a slit system of 1/0.1/1 with a receiving slit of 0.6 mm, and a scanning range of 4 to 100° 2θ were used for all measurements. Unit-cell parameters were calculated by least-squares refinement of the XRD data, using the computer program of [36] modified for PC use by [37]. Synthetic silicon (NBS 640b) was used as an external standard in order to verify the accuracy of the measurements. As the sample grinding seems to lower the crystallinity of the mineral, as well as to produce its partial carbonation, as in the case of Ca silicates [38], separate determinations of the unit-cell values were carried out based on single-crystal X-ray diffraction.

Structure refinements of scheelite crystals were performed by single-crystal X-ray diffraction on a Rigaku Agilent Xcalibur EOS diffractometer equipped with a CCD Detector (both manufactured in Osaka, Japan), housed at the Laboratory of Mineralogy, University of Liège. Data were collected at room temperature (293 K) with monochromatized MoKα radiation (λ = 0.71703 Å) at 40 kV and 40 mA. The instrument has Kappa geometry (ϕ/ω scan). Data collection, subsequent data reduction, and face-based absorption corrections were carried out using the CrysAlis Pro 41.123a software [39]. The initial solution of the structure in space group I4₁/a was determined by the charge flipping method using the Superflip algorithm [40], and the structural model was subsequently refined on the basis of F² with the Jana2006 software [41].

The infrared-absorption spectrum of a sample from Oravița was obtained at the Geological Institute of Romania with a SPECORD M-80 IR spectrophotometer (Carl Zeiss, Jena, Germany), using a conventional pressed disk technique. Carefully dried mineral powders were embedded in KBr; a dilution of 2.5 wt.% was used. This spectrum was recorded between 250 and 4000 cm⁻¹. FT-IR spectra of representative samples from Ciclova and Băița Bihor, as well as details of the stretching region of (WO₄)²⁻ groups in the scheelite structure, were obtained using a BRUKER FTIR S 12 spectrometer (Ettlingen, Germany) also hosted by the Geological Institute of Romania. In this case, the records were made in the frequency range between 400 and 4000 cm⁻¹, using the standard pressed-disk technique, after embedding 2 mg of mechanically ground mineral powder in 148 mg of dry KBr and compacting at 2500 N/cm² pressure. The spectral resolution was 0.1 cm⁻¹. All spectra were recorded at 25 °C.

Raman spectra were recorded using a Renishaw SEM-Raman system (Artisan Scientific, Champaign, IL, USA) hosted by the Geological Institute of Romania, at 25 °C, using both structural and chemical analysis (SCA) and inVia interfaces. The spectrometer was equipped with a 10 mW, 532 nm diode-pumped solid-state laser as an excitation source. The spectral resolution was 1 cm⁻¹ for a 1 μm spatial resolution. Analyses were collected using a 50× objective, a confocal aperture of 400 µm, a 150 µm slit width, and 1800 lines·mm⁻¹ grating. The spectra were collected in the range of 10–1500 cm⁻¹ (10-s accumulation time, 10 scans). The instrument was calibrated with synthetic silicon.

The mean density of various crystals of scheelite was measured using a pycnometer (Rainhard Co., Austin, TX, USA) at 25 °C, using a mixture of methylene iodide and toluene as the immersion liquid.

Indices of refraction were measured at room temperature (25 °C) using a conventional JENAPOL-U petrographic microscope (Carl Zeiss, Jena, Germany) with a spindle stage and calibrated immersion liquids (Cargille or temperature-calibrated oils), with a 589-nm interference filter. Thin-section photographs were obtained using a Zeiss Axio Imager A2M microscope (Carl Zeiss, Jena, Germany) with 2.5×/10×/20× objectives.

UV-luminescence tests were performed using a portable Vetter ultraviolet lamp (Vetter, Lottstetten, Germany) with 254 and 366 nm filters, respectively.

4. Mode of Occurrence and Morphology

At the macroscopic level, scheelite crystals occur as roughly tetragonal crystals up to 30 mm across, easily identified in UV light (Figure 2). At Oravița, the crystals are generally anhedral, due to secondary overgrowths (see below). Previous studies have described scheelite nodules up to 1 cm in diameter at Ciclova [15], as well as well-formed crystals having {001} (basal pinacoid), {111} (pyramid), and {101} (prism) forms [13]. At Băița Bihor, euhedral to subhedral crystals vary in size from microns to 3 cm [16]. Twinning on {110} was rarely recognized, as were contact twins with a composition plane of (110).

The microscopic study reveals that the mineral can occur as clusters or streaks of crystals, but also as individual crystals, as already observed by [16]. The crystals are generally subhedral or even anhedral at Oravița [15]. In most of the cases, partly preserved tetragonal contours can be observed (Figure 2). The distinct cleavage on {101} can be observed, as it is sometimes lined by secondary hydrotungstite.

The CL imaging at the electron microscope is relevant for the tetragonal symmetry of the mineral. Transversal sections of typical crystals of scheelite from Ciclova, Oravița, and Băița Bihor are given in Figure 3 and generally reveal oscillatory CL zoning suggesting the tetragonal symmetry. Overgrowths of secondary scheelite on previous crystals (Figure 3B,C) could explain the anhedral contours observed in some cases [15].

5. Chemical Data

Electron-microprobe analyses of 30 representative scheelite crystals from Ciclova, Oravița, and Băița Bihor are given in

Table 2. Representative electron-microprobe analyses of scheelite from Ciclova *.

Sample	2506a	2506b	2506c	2509a	2509b	2509c	2534a	2534b	2575a	2575b	Mean
N (1)	12	15	16	11	13	13	14	11	12	11	128
WO3	79.81	79.41	78.31	78.91	79.24	79.14	79.40	79.16	79.24	78.49	79.10
MoO3	0.63	0.63	1.21	1.21	1.14	1.05	0.77	1.12	1.12	0.87	0.97
Bi2O3	0.03	0.02	0.03	0.01	0.04	0.02	0.01	0.03	0.03	0.04	0.03
CaO	19.44	19.40	19.45	19.49	19.38	19.37	19.31	19.31	19.32	19.25	19.37
MgO	0.00	0.02	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
MnO	0.02	0.03	0.04	0.03	0.02	0.03	0.06	0.04	0.03	0.02	0.03
FeO (2)	0.03	0.06	0.03	0.02	0.04	0.05	0.02	0.02	0.04	0.04	0.04
PbO	0.03	0.05	0.03	0.03	0.05	0.07	0.07	0.02	0.05	0.00	0.04
CuO	0.01	0.01	0.03	0.03	0.02	0.02	0.05	0.01	0.03	0.02	0.02
Total	100.00	99.63	99.13	99.73	99.93	99.75	99.67	99.71	99.86	98.73	99.60
Number of Cations on the Basis of 4(O)
W	0.988	0.986	0.974	0.976	0.979	0.980	0.986	0.980	0.980	0.982	0.981
Mo	0.013	0.013	0.024	0.024	0.023	0.021	0.015	0.022	0.022	0.018	0.019
Bi	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Ca	0.995	0.996	1.000	0.996	0.990	0.992	0.991	0.989	0.988	0.996	0.994
Mg	0.000	0.001	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Mn	0.001	0.001	0.002	0.001	0.001	0.001	0.002	0.002	0.001	0.001	0.001
Fe2+	0.001	0.002	0.001	0.001	0.002	0.002	0.001	0.001	0.002	0.002	0.002
Pb	0.000	0.001	0.000	0.000	0.001	0.001	0.001	0.000	0.001	0.000	0.001
Cu	0.000	0.000	0.001	0.001	0.001	0.001	0.002	0.000	0.001	0.001	0.001
Composition in End Members (mol.%)
scheelite	98.70	98.70	97.60	97.60	97.70	97.90	98.50	97.80	97.80	98.20	98.10
powellite	1.30	1.30	2.40	2.40	2.30	2.10	1.50	2.20	2.20	1.80	1.90

* EMP analyses, results expressed in wt.%; ⁽¹⁾ number of point analyses on the same crystal; ⁽²⁾ total iron taken as divalent.

,

Table 3. Representative electron-microprobe analyses of scheelite from Oraviţa *.

Sample	2537a	2537b	2537c	2538a	2538b	2538c	2539	2568	2569a	2569b	Mean
N (1)	5	6	7	9	7	6	11	16	12	11	90
WO3	79.86	79.30	79.87	80.38	80.36	80.48	79.42	77.79	77.95	77.82	79.03
MoO3	0.14	0.33	0.25	0.37	0.32	0.27	1.32	2.85	1.92	1.97	1.29
Bi2O3	0.01	0.01	0.00	0.01	0.01	0.00	0.02	0.02	0.00	0.01	0.01
CaO	19.29	19.53	19.55	19.42	19.36	19.44	19.56	19.90	19.46	19.42	19.53
MgO	0.00	0.01	0.01	0.00	0.01	0.01	0.01	0.01	0.00	0.00	0.01
MnO	0.00	0.00	0.01	0.00	0.00	0.00	0.00	0.00	0.01	0.01	0.00
FeO (2)	0.05	0.04	0.03	0.03	0.05	0.17	0.02	0.02	0.02	0.02	0.04
PbO	0.04	0.00	0.03	0.01	0.05	0.02	0.02	0.02	0.04	0.01	0.02
CuO	0.02	0.00	0.01	0.00	0.01	0.00	0.01	0.00	0.00	0.00	0.01
Total	99.41	99.22	99.76	100.22	100.17	100.39	100.38	100.61	99.40	99.26	99.94
Number of Cations on the Basis of 4(O)
W	0.997	0.990	0.993	0.994	0.995	0.994	0.975	0.944	0.963	0.963	0.975
Mo	0.003	0.007	0.005	0.007	0.006	0.005	0.026	0.056	0.038	0.039	0.025
Bi	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Ca	0.996	1.008	1.004	0.993	0.991	0.993	0.993	0.998	0.994	0.993	0.996
Mg	0.000	0.001	0.001	0.000	0.001	0.001	0.001	0.001	0.000	0.000	0.001
Mn	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Fe2+	0.002	0.002	0.001	0.001	0.002	0.007	0.001	0.001	0.001	0.001	0.002
Pb	0.001	0.000	0.000	0.000	0.001	0.000	0.000	0.000	0.001	0.000	0.001
Cu	0.001	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Composition in End Members (mol.%)
scheelite	99.70	99.30	99.50	99.30	99.40	99.50	97.40	94.40	96.20	96.11	97.50
powellite	0.30	0.70	0.50	0.70	0.60	0.50	2.60	5.60	3.80	3.89	2.50

* EMP analyses, results expressed in wt.%; ⁽¹⁾ number of point analyses on the same crystal; ⁽²⁾ total iron taken as divalent.

and

Table 4. Representative electron-microprobe analyses of scheelite from Băița Bihor *.

Sample	2536	2544	2549	2552	2576	2577	2590	2591	2592	2593	Mean
N (1)	11	10	8	9	12	10	10	13	11	11	105
WO3	79.70	79.48	79.56	78.41	78.81	78.93	78.60	79.28	79.49	78.97	79.13
MoO3	0.63	0.73	0.61	1.13	1.29	0.88	1.16	1.17	0.84	0.88	0.95
Bi2O3	0.03	0.03	0.00	0.04	0.04	0.01	0.01	0.00	0.02	0.05	0.02
CaO	19.48	19.4	19.37	19.42	19.36	19.26	19.43	19.5	19.26	19.23	19.37
MgO	0.00	0.00	0.05	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
MnO	0.02	0.07	0.01	0.04	0.03	0.02	0.03	0.02	0.01	0.03	0.03
FeO (2)	0.04	0.03	0.09	0.03	0.07	0.03	0.01	0.03	0.04	0.03	0.04
PbO	0.04	0.02	0.08	0.03	0.04	0.01	0.03	0.05	0.04	0.00	0.03
CuO	0.01	0.00	0.01	0.02	0.03	0.03	0.03	0.03	0.03	0.01	0.02
Total	99.95	99.76	99.78	99.12	99.67	99.17	99.30	100.08	99.73	99.20	99.59
Number of Cations on the Basis of 4(O)
W	0.987	0.986	0.987	0.976	0.975	0.984	0.976	0.978	0.986	0.984	0.982
Mo	0.013	0.015	0.012	0.023	0.026	0.018	0.023	0.023	0.017	0.018	0.019
Bi	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.001	0.000
Ca	0.997	0.995	0.993	0.999	0.991	0.992	0.998	0.994	0.988	0.991	0.994
Mg	0.000	0.000	0.004	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Mn	0.001	0.003	0.000	0.002	0.001	0.001	0.001	0.001	0.000	0.001	0.001
Fe2+	0.002	0.001	0.004	0.001	0.003	0.001	0.000	0.001	0.002	0.001	0.002
Pb	0.001	0.000	0.001	0.000	0.001	0.000	0.000	0.001	0.001	0.000	0.001
Cu	0.000	0.000	0.000	0.001	0.001	0.001	0.001	0.001	0.001	0.000	0.001
Composition in End Members (mol.%)
scheelite	98.70	98.50	98.80	97.70	97.40	98.20	97.70	97.70	98.31	98.20	98.10
powellite	1.30	1.50	1.20	2.30	2.60	1.80	2.30	2.30	1.69	1.80	1.90

* EMP analyses, results expressed in wt.%; ⁽¹⁾ number of point analyses on the same crystal; ⁽²⁾ total iron taken as divalent.

. The samples are indicated by numbers, whereas letters, when used, indicate different crystals in the same sample. Each analysis represents the average of N points of analysis across the same crystal from a thin section. Formulas were normalized on the basis of 4(O) atoms per formula unit (apfu), the resulting crystal-chemical formulas being, in all cases, definitory for terms close to the scheelite end-member of the scheelite-powellite solid solution series.

A few remarks must be drawn based on the results in Table 2, Table 3 and Table 4, as follows:

• No obvious chemical zoning was observed at the level of individual crystals, in spite of the CL zoning described below. Qualitative elemental X-ray maps exemplified in Figure 4a,b don’t show any zoning.
• A slight overcompensation in the four-fold coordinated cations (i.e., W + Mo > 1) could be observed in some samples, particularly at Oraviţa (Table 3). This phenomenon was already observed in some other cases (e.g., [12,42,43]) and is probably due to the EMPA interference effect of tungstite (ideally WO₃·H₂O) or rather hydrotungstite (ideally H₂WO₄·H₂O), whose presence in all three occurrences was already mentioned [44].
• The isomorphism in the scheelite–powellite solid solution series is very restricted, reaching up to 2.4 mol.% powellite at Ciclova, up to 5.6 mol.% at Oravița, and up to 2.6 mol.% at Băița Bihor. The X-ray elemental maps failed to identify a chemical zoning at the crystal level, in spite of the general opinion that the dark CL oscillatory bands described above are enriched in Mo (i.e., [45,46]).
• Ba was sought but not detected, so the isomorphism toward ronpetersonite can’t be considered.
• The contents of stolzite (ideally PbWO₄) in the solid solution are insignificant, reaching up to 0.1 mol.%.
• The small contents of Cu recorded in part of the samples could be rather due to submicrometer-sized cuprotungstite—ideally Cu₂(WO₄)(OH)₂—inclusions, probably intruded on the cleavage directions; these inclusions are too small to be seized by the X-ray imaging.
• The collection of qualitative elemental X-ray maps (Figure 4a,b) together with SEM-EDS analyses confirmed that Sb and Bi are present in minor amounts in some scheelite grains. Apparently, both trivalent Sb and Bi are substitutes for Ca²⁺ through a mechanism similar to those reported by [47] and referred to in the case of REE and Y, i.e., 2Bi³⁺(Sb³⁺) + □ Ca = 3Ca²⁺, where □ Ca represents a vacancy site.

6. X-Ray Powder Diffraction Data

Cell parameters of 5 representative samples of scheelite were successfully refined by least squares, based on a tetragonal I4₁/a cell (e.g., [48]). The patterns were indexed on the basis of Miller indices obtained by structural refinement that closely correspond to PDF 41-1431 [49]. Three XRD powder patterns given for scheelite samples from Băița Bihor by [16] and three others given for scheelite samples from Oravița–Ciclova and Băița Bihor by [44] were also used for calculation of cell parameters. The results are given in

Table 5. Cell parameters of selected samples of scheelite from the analyzed skarn occurrences.

Sample	a (Å)	c (Å)	V (Å3)
Ciclova
1189 (1)	5.230(2)	11.356(7)	310.67(8)
2492 (2)	5.2431(1)	11.3743(5)	312.68(2)
2506	5.243(4)	11.366(14)	312.46(53)
2509	5.234(2)	11.393 (15)	312.06(37)
Oraviţa
535 (1)	5.231(1)	11.343(3)	310.41(14)
2491	5.239(4)	11.402 (9)	313.02(53)
2537	5.229(4)	11.372(14)	310.92(57)
2569 (2)	5.2425(2)	11.3703(7)	312.499(2)
Băiţa Bihor
Sch 01 (3)	5.240(2)	11.373(5)	312.25(20)
Sch 02 (3)	5.2384(9)	11.370(4)	312.00(12)
Sch 03 (3)	5.2377(8)	11.366(2)	311.82(10)
1050 (1)	5.232(2)	11.375(8)	311.34(25)
2536	5.243(1)	11.380(4)	312.85(16)
2576 (2)	5.2409(2)	11.3705(6)	312.314(2)
2593 (2)	5.2388(2)	11.3667(9)	311.959(1)

⁽¹⁾—calculated after the XRD powder patterns given by [44]. ⁽²⁾—data obtained by single-crystal refinement. ⁽³⁾—calculated after the XRD powder patterns given by [16].

, together with 4 sets of cell parameters obtained by single-crystal X-ray refinement.

The cell parameters in Table 5 are generally close to the cell parameters reported by various authors for stoichiometric scheelite {i.e., a = 5.24294(6) Å and c = 11.373(7) Å according to [49]; a = 5.2419(5) Å and c = 11.376(1) Å according to [50]; and a = 5.23643(5) Å and c = 11.33815(33) Å according to [51]} and also to similar values reported for scheelite samples close to stoichiometry: e.g., [43,52]. The low content of powellite in the solid solution series of the analyzed scheelite samples is quite evident since powellite has lower a and higher c cell parameters {e.g., a = 5.2256(5) Å and c = 11.434(1) Å according to [50], a = 5.222(1) Å and c = 11.425(3) Å according to [52], or a = 5.220(2) Å and c = 11.407(10) Å according to [53]}.

7. Physical Properties

Macroscopically, scheelite from all three occurrences occurs as translucent, turbid-white to yellow crystals. A glassy luster can be observed on the distinct cleavage on {101}, whereas the normal luster is waxy vitreous. The mineral is nonfluorescent in long-wave UV light (366 nm) and shows generally strong blue fluorescence in short-wave UV light (254 nm). Exceptions occur at some samples from Oravița, where a yellow tint of fluorescence also occurs. The usual fluorescence tints are better depicted in Figure 2 and Figure 5 and indicate low contents of powellite in the solid-solution series [54].

Compared with the standard color card proposed by [54], the scheelite samples from Ciclova, Băița Bihor, and the large majority of those from Oravița may contain up to 3 mol.% powellite in solid solutions, whereas the yellow-tinted scheelite from Oravița contains up to 6 mol.% powellite, which perfectly agrees with data in Table 2, Table 3 and Table 4. Scheelite samples from Băița Bihor seem to correspond to the scheelite III type described by [16] based on fluorescence colors.

Under the electron beam, scheelite samples show a strong cathodoluminescence (CL) response of light blue color, which can be due to the self-activated emission band (also referred to as a self-luminescence band, SB), which produces a broad spectral peak that is intrinsic to the mineral, this being attributed to the (WO₄)^2– tetrahedral complex ([45] and referred works). A cathodoluminescence zoning of the individual crystals, described in many worldwide occurrences (e.g., [42,45,46,47,55,56]), can be observed in all the analyzed samples and is illustrated in Figure 3. The subhedral scheelite grains display in many cases oscillatory CL zones characterized by alternating dark and bright CL bands with sharp boundaries, similar to Type 1 scheelite described by [46] or S-type described by [55]. Crystals of scheelite could also display bright peripheral CL concentric zones with central homogenous dark CL zones, suggesting the optical lamellar zoning of garnet and resembling Type 2 scheelite described by [46]: Figure 3.

Optically, the mineral is uniaxial positive. The indices of refraction of three representative samples (one for each occurrence), taken as the average of five measured values on 5 different crystals from the same sample, are given in

Table 6. Physical constants of hydrothermal scheelite from the analyzed skarns.

Sample	2509b	2539	2544
Occurrence	Ciclova	Oraviţa	Băiţa Bihor
ω	1.928	1.929	1.931
ε	1.945	1.946	1.948
n	1.934	1.935	1.937
Mmol (1)	286.298	285.614	286.801
V (Å3)	313.11	311.90	312.31
Dm (g/cm3)	6.08(1)	6.08(1)	6.10(1)
Dx (g/cm3)	6.071	6.080	6.098
KP (2)	0.1536	0.1538	0.1536
KC	0.1492	0.1494	0.1488
1 − KP/KC	−0.0298	−0.0294	−0.0323
KP′ (3)	0.1538	0.1538	0.1537
1 − KP′/KC	−0.0313	−0.0293	−0.0327

⁽¹⁾ Molecular mass in atomic mass units; ⁽²⁾ based on measured density (D_m); ⁽³⁾ based on calculated density (D_x).

. The indices of refraction in Table 6 are in the range of the higher values reported by [57] for scheelite, i.e., ω = 1.911 − 1.934 and ε = 1.926 − 1.953. As the refraction indices presumably decrease with increasing content of Mo [57], the values in Table 6 are indicative of terms with low contents of powellite in the solid solution series. The mean refraction index, calculated as n = (2ω + ε)/3 [58], is also given in Table 6.

The D_x calculated densities of three representative samples (one for each occurrence) were obtained on the basis of chemical data in Table 2, Table 3 and Table 4 and of unit cell volumes in Table 6, for Z = 4 formula units per cell [59]. The values obtained are given in Table 6. The mean measured density D_m, taken as the average of measurements on 10 different grains, is also included in Table 6 and compares well with the calculated densities. Both values, as well as the mean refraction indices, were used to calculate the physical refractive energy as K_P = (n − 1)/D.

The chemical refractive energy (K_C) values in Table 6 are based on the oxide content issued from EMP analyses of the three samples and on the constants for oxides given by [58,60], excepting that for WO₃, taken from [61], i.e., k (WO₃) = 0.133.

The calculation of the Gladstone-Dale compatibility index (1 − Kp/Kc) yielded the values given in Table 6, indicative of excellent agreement between physical and chemical data [60]. The use of a Gladstone-Dale constant for WO₃ as indicated by [58,60], i.e., k (WO₃) = 0.152 or k (WO₃) = 0.171, respectively, results in poor compatibilities.

8. Structure

The atomic structure of scheelite was refined by both neutron diffraction [48] and single-crystal X-ray diffraction ([52,59]). The results are coincident in describing a structure consisting of two intercalated polyhedral lattices, one composed of slightly deformed WO₄ tetrahedra and another by CaO₈ dodecahedra (bidisphenoids) described by eight-fold coordinated Ca cations. The structure features a tetragonal unit cell with space group I4₁/a (Z = 4), which was confirmed during this study. As the powellite contents in the analysed samples are low, the eventual Mo atoms were not included in the refinement of the X-ray data.

Details of the scheelite structure obtained during this work are shown in Figure 6.

Table 7. Data collection and structure refinement details for Romanian scheelite.

Sample	2509	2538	2544
Occurrence	Ciclova	Oravița	Băița Bihor
a (Å)	5.2459(10)	5.2380(2)	5.2409(2)
c (Å)	11.3777(5)	11.3679(8)	11.3705(6)
V (Å3)	313.108(18)	311.90(3)	312.31(3)
Z	4	4	4
Dx (g/cm3)	6.108	6.132	6.124
Crystal size (mm3)	0.177 × 0.124 × 0.091	0.35 × 0.22 × 0.1	0.283 × 0.21 × 0.195
Absorption coefficient (mm−1)	38.351	38.502	38.449
F(000)	504	504	504
Max. 2ө (°)	56.994	57.766	57.74
Range of indices	−6 ≤ h ≤ 6	−5 ≤ h ≤ 6	−7 ≤ h ≤ 6
	−5 ≤ k ≤ 6	−4 ≤ k ≤ 6	−7 ≤ k ≤ 5
	−13 ≤ l ≤ 15	−14 ≤ l ≤ 14	−12 ≤ l ≤ 15
Number of measured reflections	1031	1011	1051
Number of unique reflections	188	191	194
Independent non-zero reflections	176	180	180
Criterion for observed reflections	I > 2σ(I)	I > 2σ(I)	I > 2σ(I)
Number of refined parameters	16	16	16
R int	0.0212	0.0261	0.0327
R sigma	0.0158	0.0181	0.0216
R1 (F) with F0 > 4 σ(F0) *	0.0144	0.0189	0.0216
R1 (F) for all the unique reflections *	0.0165	0.0204	0.0237
wR2 (F2) *	0.0348	0.0477	0.0539
S (“goodness of fit”)	1.227	1.267	1.206
Min./max. residual e density, (eÅ−3)	−0.88/0.61	−2.46/0.77	−1.71/1.00
Weighing scheme	1/(σ2(I)2+ 0.0025(I)2

* Notes: R₁ = Σ(|Fobs| − |Fcalc|)/Σ|Fobs|; wR₂ = {Σ[w(F²obs − F²calc)²]/Σ[w(F²obs)²]}^1/2. W = 1/[s²(F₀²) + (aP)² + bP], where P = [2F_c² + Max(F₀²,0)]/3, where a, b are shown in the refinement process.

depicts the data collection and structure refinement details for the analyzed samples, whereas positional and displacement parameters are given in

Table 8. Anisotropic displacement parameters (Ų) for scheelite.

Atom	U11	U22	U33	U23	U13	U12
Sample 2509 Ciclova
Ca	3.4(9)	3.4(9)	1.8(11)	0	0	0
W	3.96(17)	3.96(17)	4.9(2)	0	0	0
O	14.0(19)	14.7(18)	12.0(15)	−2.0(14)	−1.1(13)	−2.7(14)
Sample 2538 Oravița
Ca	2.2(12)	2.2(12)	2.8(15)	0	0	0
W	2.6(2)	2.6(2)	6.2(3)	0	0	0
O	11(2)	11(2)	10(2)	1.9(17)	−2.0(17)	0.3(15)
Sample 2544 Băița Bihor
Ca	7.7(13)	7.7(13)	4.2(16)	0	0	0
W	7.6(3)	7.6(3)	8.4(3)	0	0	0
O	16(2)	18(2)	15(2)	−1(2)	−3(2)	1.9(18)

,

Table 9. Atom coordinates and isotropic displacement parameters (Ų) for scheelite.

Atom	x	y	z	U (eq)
Sample 2509 Ciclova
Ca	0	7500	8750	2.9(8)
W	0	2500	6250	4.27(15)
O	1515(6)	4909(6)	7108(3)	13.6(10)
Sample 2538 Oravița
Ca	0	7500	3750	2.4(10)
W	5000	7500	6250	3.8(2)
O	2594(7)	5980(7)	5392(4)	10.7(12)
Sample 2544 Băița Bihor
Ca	0	7500	3750	6.5(11)
W	5000	7500	6250	7.8(2)
O	2592(9)	6000(8)	5393(4)	16.4(14)

,

Table 10. Bond distances (Å) and occupancies observed in the crystal structure of scheelite.

Sample	2509	2538	2544
Occurrence	Ciclova	Oravița	Băița Bihor
Bond distances (Å)
Ca-O x 4	2.443(4)	2.442(5)	2.440(5)
Ca-O x 4	2.478(4)	2.474(4)	2.482(5)
W-O x 4	1.784(4)	1.781(4)	1.778(5)
Occupancy
Ca	0.83(15)	0.84(2)	0.83(2)
W	0.91(10)	0.91(14)	0.87(15)

and

Table 11. Detailed bond-valence table (v_u) for the crystal structure of scheelite *.

	Sample 2509 Ciclova			Sample 2538 Oravița			Sample 2544 Băița Bihor
	Ca	W	Σ	Ca	W	Σ	Ca	W	Σ
O	0.276 × 4 ↓	1.436 × 4 ↓	1.963	0.277 × 4 ↓	1.444 × 4 ↓	1.98	0.278 × 4 ↓	1.456 × 4 ↓	1.98
	0.251 × 4 ↓			0.254 × 4 ↓			0.249 × 4 ↓
Σ	2.11	5.73		2.12	5.77		2.11	5.82

* bond-valence parameters adjusted after [62].

. The corresponding *.cif files are given as Supplementary Materials.

The structure can be visualized as an assembly of CaO₈ dodecahedra and WO₄ tetrahedra. Each CaO₈ dodecahedron is connected to four other dodecahedra through edge-sharing, resulting in zigzag chains that develop along the c axis. WO₄ tetrahedra are isolated from each other and connected to CaO₈ dodecahedra via bridging oxygen atoms; each oxygen atom is connected to two Ca atoms and one W atom. CaO₈ dodecahedra are described by two distinct Ca−O distances (e.g., Table 11) and eight distinct O−Ca−O angles, while WO₄ tetrahedra could be fairly described by a single W−O distance and two distinct O−W−O angles. In fact, the isolated WO₄ tetrahedron is compressed along the c axis by 7% over a regular tetrahedron [51], which is better depicted by data in Table 10. Note that the site occupancies observed in Table 10 suggest mixed occupancies in both Ca and W sites and the non-stoichiometry of the analyzed samples with site populations involving substitutions in both structural sites, as given in Chapter 5.

9. Infrared and Raman Behavior

The infrared and Raman spectra recorded for representative scheelite samples from Ciclova, Oravița, and Băița Bihor are shown in Figure 7 and Figure 8, respectively. An attempt to assign the vibrational bands in the Raman and infrared spectra to specific vibrational modes, together with the character and relative intensities of the bands in the infrared spectra, is included in Table 12.

Assigning vibrational bands in the Raman and infrared spectra to specific vibrational modes was considerably facilitated by the structure determinations, as well as by previous works of [11,12,63,64,65,66,67,68,69,70,71] for infrared spectra and of [11,65,70,72,73,74,75,76,77,78] for the Raman behavior.

IR spectra in Figure 7 show, at first view, the deviation of the main vibrational bands assignable to the WO₄ structural groups from the values given by [79] for the WO₄ tetrahedra with ideal T_d symmetry (i.e., ν₁ = 931 cm⁻¹, ν₃ = 833 cm⁻¹, ν₂ = 373 cm⁻¹ and ν₄ = 320 cm⁻¹), as well as a superior band multiplicity. Note that, in scheelite, the ν₂ out-of-plane bending occurs at higher wavenumbers than the ν₄ in-plane bending, representing an opposite relation to the usual order. This relation, which implies that ν₂ > ν₄, was proven, based on force field studies, by [68]. Band assumptions in Table 12 were made consequently. Excepting the ν₁ symmetric stretching, vibrational modes of the [WO₄]²⁻ group (i.e., the ν₃ antisymmetric stretching and the two bending modes) are double degenerate. The band assignable to the ν₁ symmetric stretching, predicted by [68] to occur at 903 cm⁻¹, is materialized by a shoulder on the complex of bands at 750–900 cm⁻¹, which materializes the ν₃ antisymmetric stretching vibration of the WO₄ group (simply degenerate according to [67], and double degenerate according to [11,65,68,70]. The four bands found by [12] in the same region are certainly due to admixed secondary tungstates or, more probably, to the presence of two generations of scheelite, with different powellite contents. The ν₂ out-of-plane bending vibration of the [WO₄]²⁻ group was found in the range of 428–441 cm⁻¹: [11,12,65,66,67,71]. Accepting that the two shoulders of the broad complex of bands centered at ~ 810 cm⁻¹ materializes the ν₁ symmetric stretching (shoulder at higher frequency) and the ν’₃ antisymmetric stretching (shoulder at lower frequency) of the [WO₄]²⁻ group, respectively, and that the band recorded at 286 cm⁻¹ in the infrared spectrum in Figure 7 materializes in reality the ν’₄ in-plane bending [67], the multiplicity of bands in the infrared absorption spectrum (1ν₁ + 2ν₃ + 2ν₂ + 2ν₄) is consistent with the S₄ punctual symmetry of the tungstate anion, agreeing with the structural data. This point group symmetry was also deduced by [73,76,80] on the basis of the multiplicity of the Raman bands. Comparative studies on the IR behavior of CaWO₄ vs. CaMoO₄ (e.g., [65,67]) failed to prove any dependence of the positions of the main bands in the IR spectra on the Mo-for-W substitution. The slight differences in the position of the IR bands in our spectra do not consequently depend on the different powellite contents in the solid solution series of the analyzed samples.

Table 12. Positions and assignments of the FTIR and Raman bands recorded for scheelite from Romanian occurrences ⁽¹⁾.

Structural Group	Vibrational Mode	Wavenumber (cm−1)								Character, Intensity (4)
		Calculated		Ciclova		Oravița		Băița Bihor
		IR (2)	Raman (3)	FTIR	Raman	IR	Raman	FTIR	Raman
WO4	ν1 (Ag) symmetric stretching	903	912	908	911	910	911	908	913	shd, s
WO4 (?)	ν’1 (Bg) (?)	-	893	-	-	-	-	-	883	-
WO4	ν3 (Bg) antisymmetric stretching	808	838	810	838	811	839	810	841	sh, vs
WO4	ν’3 (Eg) antisymmetric stretching	795	780	797	797	795	797	797	797	shd, s
WO4	ν2 (Ag) out-of-plane bending	418	439	440	434	441	434	440	437	sh, m
WO4	ν’2 (Eg + Bg) out-of-plane bending	-	403	-	401	-	400	-	402	-
WO4	ν4 (Bg) in-plane bending	359	-	-	369	358	369	-	370	sh, w
WO4	ν’4 (Eg) in-plane bending	348	330	-	333	326	333	-	332	sh, m
WO4	rotation R‖ (Ag)	286	280	-	275	285	-	-	274	sh, w
WO4	rotation R⊥ (Eg)	-	207	-	211	-	211	-	210
CaO8, WO4	translation (Eg) Ca/Ca	-	196	-	193	-	193	-	193
CaO8, WO4	translation (Bg) W/W	-	115	-	116	-	116	-	115
CaO8, WO4	translation (Eg + Bg) W/W	-	83	-	84	-	84	-	85

⁽¹⁾ Assumptions according to the authors referred to in the text; ⁽²⁾ after [68]; ⁽³⁾ after [70,80]; ⁽⁴⁾ character of the bands on IR (FTIR) spectra: s = strong; m = medium; w = weak; vs = very strong; sh = sharp; shd = shoulder.

Table 12. Positions and assignments of the FTIR and Raman bands recorded for scheelite from Romanian occurrences ⁽¹⁾.

Structural Group	Vibrational Mode	Wavenumber (cm−1)								Character, Intensity (4)
		Calculated		Ciclova		Oravița		Băița Bihor
		IR (2)	Raman (3)	FTIR	Raman	IR	Raman	FTIR	Raman
WO4	ν1 (Ag) symmetric stretching	903	912	908	911	910	911	908	913	shd, s
WO4 (?)	ν’1 (Bg) (?)	-	893	-	-	-	-	-	883	-
WO4	ν3 (Bg) antisymmetric stretching	808	838	810	838	811	839	810	841	sh, vs
WO4	ν’3 (Eg) antisymmetric stretching	795	780	797	797	795	797	797	797	shd, s
WO4	ν2 (Ag) out-of-plane bending	418	439	440	434	441	434	440	437	sh, m
WO4	ν’2 (Eg + Bg) out-of-plane bending	-	403	-	401	-	400	-	402	-
WO4	ν4 (Bg) in-plane bending	359	-	-	369	358	369	-	370	sh, w
WO4	ν’4 (Eg) in-plane bending	348	330	-	333	326	333	-	332	sh, m
WO4	rotation R‖ (Ag)	286	280	-	275	285	-	-	274	sh, w
WO4	rotation R⊥ (Eg)	-	207	-	211	-	211	-	210
CaO8, WO4	translation (Eg) Ca/Ca	-	196	-	193	-	193	-	193
CaO8, WO4	translation (Bg) W/W	-	115	-	116	-	116	-	115
CaO8, WO4	translation (Eg + Bg) W/W	-	83	-	84	-	84	-	85

⁽¹⁾ Assumptions according to the authors referred to in the text; ⁽²⁾ after [68]; ⁽³⁾ after [70,80]; ⁽⁴⁾ character of the bands on IR (FTIR) spectra: s = strong; m = medium; w = weak; vs = very strong; sh = sharp; shd = shoulder.

All the 13 Raman-active vibrations recognized by [63] or [80] can be observed in our spectra (Table 12, Figure 8). In fact, the factor-group analysis indicates, for scheelite-group structures, a band multiplicity describing a representation of Γ_i = 3A_g + 5B_g + 5E_g (13 vibrational modes), from which 7 are internal [stretching ν₁ (A_g); ν₃ (B_g) and ν’₃ (E_g); bending ν₂(A_g), ν’₂(B_g), ν₄(B_g) and ν’₄(E_g)] and 6 are external (2 rotations, i.e., A_g and E_g, and 4 translations, i.e., 2B_g and 2 E_g): [70,73,75,81]. A fourteen-band, detected by us at 883 or 893 cm⁻¹ (Table 12), is normally Raman inactive but was also observed by [78] and assumed to be strain-induced ν’₁ (B_g) symmetric stretching.

As a characteristic feature, the Raman spectrum displays an intense band at ~911 cm⁻¹, attributed to the ν₁ symmetric stretching vibration of the WO₄ tetrahedra (Table 12) and identified at 909 cm⁻¹ by [77] and [78], at 910 cm⁻¹ by [74], at 911 cm⁻¹ by [76], and at 916 cm⁻¹ by [70]. The frequency of this band decreases with the progression of the Mo-for-W or Pb-for Ca substitutions [75,76].

The presence of two components of one ν₂ mode cannot be clearly observed because they probably overlap. A closer inspection of the mode located near 400 cm⁻¹ allows us to conclude that it is indeed a double mode, as very weak additional bands or shoulders occur in its case (Figure 8).

The low-frequency translational E_g and B_g modes occur at 85 cm⁻¹ and overlap [80].

A band located at ~695 cm⁻¹ can be observed in all the Raman spectra in Figure 8. This band, which doesn’t pertain to scheelite, can be assumed to be the O-W-O antisymmetric stretching in hydrotungstite [82]. Another band, not reproduced in Figure 8 and Table 12 and located at ~1092 cm⁻¹, was recorded in all but five of our Raman spectra and materializes the ν₁ symmetric stretching of the (CO₃)²⁻ groups in the admixed calcite [83].

10. Genetic Considerations

Most tungsten deposits, including tungsten skarns, are typically associated with evolved A- or S-type granites. The three analyzed Romanian deposits, not included among the 203 worldwide W-skarn deposits identified by [25], can be defined and classified as Cu-W skarn deposits. As well as the “true” W-skarns described by [25], they associate with calc-alkaline equigranular plutons in an orogenic belt (i.e., BMMB).

In our specific case, tungsten mineralization is typically associated with fractionated felsic magmas that form under slightly oxidized to reduced fO₂ conditions [84]. In all three cases, the analyzed scheelite samples have low powellite (Mo) contents and are locally associated with molybdenite, which implies that they formed in a reduced environment [85,86]. Associated granodiorites pertain to the ilmenite series of I-type granites as defined by [87], which solidify under reducing conditions, differing from magnetite series magmas [87]. Sulfur in a reduced magma is kept in the silicate melt, whereas sulfur in an oxidized magma escapes to the aqueous phase [88]. The associated scheelite and sulfide-bearing deposits described in this paper seem to be genetically connected with the reducing nature of the ilmenite series granitic magma, as in part of similar Japanese deposits described by [89]. The ore-forming fluid is initially related to the magmatic fluid, particularly to the exolution and segregation of a hydrothermal fluid phase, which can afterwards mix with metamorphic or meteoric waters ([56] and references therein). The low Mo contents in the analyzed samples are compatible with a reducing environment; however, fluctuating Eh conditions may be evident. Mo migrates as Mo⁶⁺ in oxidizing conditions; thus, Mo substitution is likely to occur in oxidizing environments [90]. This is consistent with thermodynamic calculations, which indicate that scheelite and molybdenite exhibit an overlapping field of stability in terms of oxygen and sulfur fugacities under normal skarn system conditions [45,85].

The hosting skarns are formed at relatively low temperature (up to 600 °C) and pressure (up to 3 kbar according to [32]) in a reducing environment. Scheelite, however, appears to crystallize from late hydrothermal solutions during a retrograde stage of skarn evolution. At Ciclova and Oravița, the mineralizing solutions may have been supplied through original fissures, which were later filled with late quartz and calcite veins in tourmaline-rich endoskarns [15]. At Băița Bihor, scheelite-rich bands occur in the mass of diopside- and chondrodite + norbergite-rich magnesian skarn at the limit of the outer, borate-bearing zone of the metasomatic columns. In all three occurrences, fluorite occurs as micrometer- to millimeter-sized individual anhedral crystals (Figure 9), sometimes associated with scheelite, which is not unusual as it was also described in other scheelite occurrences throughout the world (e.g., [11,45,47,55,56,89]). Conversely, boron-bearing minerals (i.e., schorlite at Ciclova and Oravița and magnesian borates at Băița Bihor) are particularly abundant in the scheelite-bearing skarns.

These observations suggest that W was introduced in the skarn system by fluoride-bearing aqueous fluids at relatively low temperatures and vapor-saturated water pressure (e.g., as per experimental conditions used by [92]). Accepting that these experimental conditions were fulfilled during the hydrothermal evolution of the skarn systems described in this paper, we must accept that tungsten was dissolved predominantly as H₃WO₄F²⁻ in solutions containing 0.05–0.30 m HF at pH (T) values between 1.5 and 3.4 and temperatures between 100 and 250 °C. The occurrence of abundant fluorite in scheelite-bearing samples from quartz-vein deposits from Ciclova and Oravița supports the hypothesis of scheelite crystallization from F-bearing fluids. The ideal reaction to form both scheelite and fluorite may be written, in our case, as: 2 H₃WO₄F²⁻ + 3 CaCO₃ = 2 CaWO₄ + CaF₂ + 3 H₂CO₃.

Alternatively, a number of authors (e.g., [93] and referred works) showed that a large amount of boron can be dissolved in F-bearing hydrothermal solutions, which implies that speciation of B-bearing minerals is possible in similar conditions as for scheelite and explains the presence of boron minerals in the W-bearing skarns at Ciclova, Oravița, and Băița Bihor.

11. Conclusions

Physical and crystallographic properties and vibrational behavior, as well as crystal chemistry, show that scheelite from Ciclova, Oravița, and Băița Bihor is close to stoichiometry.

In all cases, the analyzed samples show, at least locally, oscillatory zonation in cathodoluminescence, characteristic of scheelite from proximal intrusion-related systems such as skarns. CL peculiarities suggest that, in the analyzed occurrences, scheelite formation was controlled by multiple factors, e.g., the nature of ore-forming fluids and crystal diffusion or growth rate. In all cases, the CL images of the crystals vary from homogenous to zoned, behaving like hydrothermal scheelite and suggesting slight oscillations in the Eh of the crystallizing solutions, from slightly oxidizing to reducing. Textural relations and petrographic evidence indicate that the analyzed scheelite crystallizes late in the skarn paragenesis, apparently from epigenetic hydrothermal fluids.

Scheelite described in this paper is typically associated with hydrous minerals and retrograde alteration but also with F- and B-bearing minerals, suggesting that the mineral crystallized from epithermal, low-temperature fluoride- and boron-bearing aqueous solutions.

The Ciclova, Oravița, and Băița Bihor deposits in Romania can be defined as Cu-W skarns but also contain significant contents of other metals (i.e., Pb, Zn, Bi, Mo, Ag, and Au). At least the deposit from Băița Bihor, which is a quite important source of Cu, Mo, and base metals, can produce a significant amount of byproduct W.