Mineralogy, Fluid Inclusions, and Oxygen Isotope Geochemistry Signature of Wolframite to Scheelite and Fe,Mn Chlorite Veins from the W, (Cu,Mo) Ore Deposit of Borralha, Portugal

Abstract

Scheelitization of Mn-bearing wolframite, scheelite, quartz, and Fe,Mn-chlorite veins was identified in the W, (Cu,Mo) ore deposits of Borralha, by optical microscopy, electron-microprobe analysis, and stable isotope geochemistry. Fluid inclusions derived scheelite crystallization temperature was compared with the oxygen isotope temperature estimated. Scheelite was formed mainly during stage I from a low salinity aqueous-carbonic fluid dominated by CO₂, where the homogenization temperature (T_h) decreased from 380 °C to 200 °C (average of 284 °C). As temperature decreased further, the aqueous-carbonic fluid became dominated by CH₄ (Stage II; (average T_h = 262 °C)). The final stage III corresponds to lower temperature mineralizing aqueous fluid (average T_h = 218 °C). In addition, salinity gradually decreased from 4.8 wt.% to 1.12 wt.%. The δ¹⁸O_Fluid values calculated for quartz-water and wolframite-water fractionation fall within the calculated magmatic water range. The ∆_{quartz-scheelite} fractionation occurred at about 350–400 °C. The ∆_{chlorite-water} fractionation factor calculated is about +0.05‰ for 330 °C, dropping to −0.68‰ and −1.26‰ at 380 °C and 450 °C, respectively. Estimated crystallizing temperatures based on semi-empirical chlorite geothermometers range from 373 °C to 458 °C and 435 °C to 519 °C. A narrower temperature range of 375 °C to 410 °C was estimated for Fe,Mn-chlorite crystallization.

1. Introduction

Tungsten mineralization in magmatic-hydrothermal deposits occurs predominantly as wolframite ([Fe,Mn]WO₄) and scheelite (CaWO₄). Wolframite is the main dominated tungsten mineral in granite-related systems, whereas in skarn deposits, it is scheelite [1,2,3,4]. Given the variability of the geological context in which scheelite occurs in nature- scheelite-quartz vein mineralizations and greisen in granite [5,6,7,8], quartz-scheelite veins in metabasite [9], scheelite in orogenic gold [10], and skarn deposits [11], published scheelite fluid inclusions (FIs) are somewhat limited.

Scheelite related to quartz veins and greisen (e.g., Yangjiashan, China) include vapor- and liquid-rich Fis types with detectable CH₄, CO₂, and N₂, where the T_h ranges (mean values) from 230 °C to 250 °C and salinity (mean values) from 9.3 wt.% to 6.6 wt.% NaCl equiv. [12]. In addition, ore-forming fluids show intermediate to low temperatures that belong to the H₂O-NaCl ± CO₂ system in the disseminated scheelite associated with potassic feldspathization and silicification of porphyritic biotite granite from the Wuyi metallogenic belt (China) [13].

In the quartz-scheelite veins hosted in metabasites (e.g., Metaggitsi area, Greece), the most abundant FIs consist of H₂O-CO₂ with highly variable CO₂ contents (20 ± 90 vol.%) and salinity between 0.2 wt.% and 8.3 wt.% NaCl equiv., where the homogenization temperature of aqueous and gas-rich type inclusions has been estimated to about 220 ± 25 °C (mean values) [9]. Furthermore, scheelite has a distinctive trace elements composition from various deposit types and geological environments, being considered an indicator mineral for orogenic gold [14]. Microthermometric and Raman spectroscopic studies of FIs have documented low salinity aqueous-gaseous (H₂O-CO₂-CH₄-NaCl) ore fluids, where gold and scheelite precipitated in orogenic gold deposits from the Dharwar craton (South India) [15].

The FIs of scheelite crystallized during the cooler period of skarn development (e.g., Kara, Tasmania) consist of primary H₂O-CO₂ with moderate salinity (12.0 wt.% and 17.8 wt.% NaCl equiv.) and homogenization temperatures of 350–580 °C [16]. A similar thermal history and scheelite formation were observed for other W skarn deposits (e.g., CanTung Mine, Northwest Territories, Canada) [17,18]. Recently, scheelite-bearing albitite (no FI data are known) was identified in the Zhuxi (China) skarn tungstate deposit as a product of a silicate-poor, H₂O-rich melt that formed by melt-melt-liquid immiscibility in an extremely fractionated residual magma [19].

The scheelite occurrences mentioned above showed a connection with a magmatic-hydrothermal, metamorphic, or hydrothermal-meteoric fluid of low to moderate salinity, where homogenization temperatures ranged from 220 °C to 580 °C. Concerning the source and evolution of ore-forming fluids, the mechanisms and thermodynamic conditions of scheelite deposition have yet to be addressed [20,21]. In addition, in most related cases, scheelite was found associated with quartz veins, greisen, albitite, or secondary K-feldspar. Nevertheless, there are no published data available in the literature regarding the association of scheelite with chlorite.

Scheelite from the W, (Cu,Mo) ore deposit of Borralha is associated with wolframite, Fe,Mn-chlorite, and quartz, where Mn-bearing wolframite has been replaced by scheelite. The relationships “chlorite vs. metal assemblages” have been revisited and explored, aimed towards using chlorite as a “proxy” for detecting porphyry ore deposits [22] because the large variety of chlorite minerals can serve as a tool to determine the direction towards a mineralized area, to estimate the temperatures under which the ore processes took place [23,24], or to estimate the heat flux away from a magmatic-hydrothermal centre [25].

The stability of the Fe²⁺ end-member wolframite (ferberite) is much more sensitive to the high variation of f_O2 than the Mn²⁺ end-member wolframite (hübnerite) and the composition of ferberite might also be strongly affected by the oxidation state of the system [26]. The physicochemical parameters (e.g., temperature, Mn²⁺/Fe²⁺ ratio in the mineralizing solution) during wolframite crystallization influence the hübnerite vs. ferberite proportions of wolframite [27].

Furthermore, redox state exchange between Fe- and Mn-distributions in chlorite associated with wolframite and scheelite from the Borralha deposit showed that its composition is directly influenced by thermal gradients of the evolution of different stages within the magmatic-hydrothermal system [28], where the diversity of chlorite minerals associated with W-mineralization serves as an excellent indicator to identify mineralization areas and to estimate crystallization temperatures. Geothermometers have been developed for the Al-, Fe-, or Mg-rich chlorites and little is known about the specific chlorite systems, like Mn-rich chlorite, where a new question arises on Mn-redox, analogously to Fe-redox processes [29].

The application of empirical geothermometers to ore deposit research depends on complementary data regarding the mineralogical reaction mechanisms, estimation of P-V-T-X parameters of the ore-forming fluids or ones circulated through rocks, and the stable isotope geochemistry of neoformed minerals, helping to understand mineral history. Understanding the principles governing isotopic variations in nature is still limited due to the magnitude of isotopic fractionation that accompanies various geological processes and the diversity of fluids circulation mechanisms in the oceanic or continental crust, where geochemical parameters and the kinetic isotope effects accompanying a conversion of one mineral to another under a variety of conditions.

The main aim of this study is to characterize the mineralogy of scheelite + Fe,Mn-chlorite ± wolframite scheelitized assemblages found in large crystalline Fe,Mn-chlorite masses in a W veins system, the chemistry of the fluids responsible for scheelite crystallization, where FIs data are coupled with Fe,Mn-chlorite geothermometry and the oxygen isotope geochemistry of wolframite, scheelite, quartz, and Fe,Mn-chlorite.

2. Geological Background

The Variscan orogeny is currently explained by an obduction-collision orogenic model [30,31,32], which distinguishes three main deformation phases (D1, D2, and D3) in northern Portugal [33,34]. The maximum crustal thickening was achieved by the end of the D1/D2 phases, whereas the D3 phase (intra-Westphalian age) corresponds to the final stages of the continental collision process. Most of the granitic intrusions and associated thermal metamorphic peaks are coeval with the D3 phase [30,35,36]. The syn- to late-orogenic magmatic activity generated voluminous granitoid batholiths often with contrasting compositions, where their emplacement was in most cases controlled by major Variscan structures.

The W, (Cu,Mo) deposit of Borralha is located in the Central Iberian Zone in northwestern part of the Iberian Peninsula of the (Figure 1a), where different syn-orogenic Variscan granites (e.g., two mica peraluminous, syn-D3, and biotite rich syn-D3) intruded Palaeozoic metasedimentary units. The ore deposits occur at the contact between metasedimentary formations (Silurian) and the syn-D3 porphyritic Borralha biotite granite (~315 Ma) and a syn-D3 two-mica granite (Figure 1b). The Borralha ore deposit was an important W deposit in Portugal explored for wolframite, scheelite, and argentiferous chalcopyrite during 1903–1985. W mineralization occurs in quartz veins and in the two breccia pipes known as Santa Helena and Venise [37]. The Santa Helena breccia (SHB) occurs in outcrops (Figure 1b), whereas the Venise breccia was found underground during mining works. Rock fragments of variable size, cemented with mineralized coarse-grained quartz aggregates, characterize both breccias. The underground works exploited vertical quartz veins (more than 45° dip, VV) and sub-horizontal veins or “flat veins” (<30° dip, FV). The main vertical veins mineralized with tungstates occur in the most central deposit part, located in the Santa Helena veins (e.g., Santo António, Santa Helena north, and Santa Helena south) and cross several lithologies including porphyroid biotite granite, tonalites, and schists. The dominant micaschist rocks to the east part are cross-cut by the main W veins subdivided into vertical veins sub-concordant with the schistosity, namely: The three north (3N) and three south (3S) veins. However, these lodes contain less mineralizations of tungstates and more sulphides (i.e., chalcopyrite). The ore mineral assemblages consist of tungstate (wolframite and scheelite), various sulphide minerals (chalcopyrite, pyrite, pyrrhotite, molybdenite, sphalerite, galena, bismuthinite, marcasite), native bismuth, and Pb-Bi-Ag sulphosalts. Fluid inclusion studies of quartz related to W mineralization have identified aqueous-carbonic fluids (300 °C < T < 400 °C and 50 MPa < P < 100 MPa) and aqueous fluids (250 °C < T < 300 °C and 30 MPa < P < 50 MPa) in sulphide mineralization [38,39].

3. Materials and Methods

Sample collection. Wolframite to scheelite alteration and scheelite samples (SHS and 3N “vertical veins”) with crystal sizes of a few centimeters associated with Fe,Mn-chlorite in large crystalline masses collected from the Santa Helena veins (SHS) and northern gallery (3N) from the exploitation level (−60 m underground) of the Borralha ore deposit were selected for mineralogy, FI, and oxygen isotope geochemistry analyses.

Optical microscopy. The petrography of gangue and ore minerals was conducted by optical microscopy (transmitted and reflected light) using a Nikon microscope ECLIPSE model E400POL (Minato, Tokyo, Japan), coupled with an AxioCam camera ZEISS model MRc, connected to a computer with an AxioVision imaging software (SE64 Rel.4.9).

Electron microprobe analysis. Mineral compositions were obtained with an electron probe micro-analyzer (EPMA) Jeol Hyperprobe JXA-8500F (Tokyo, Japan) operated at a 15 kV accelerating voltage and with a 10 nA beam current in the case of silicates; the operation mode for the tungstates involves 20 kV and 20 nA, with counting times of 10 s on the element peaks and 5 s on the background position. The major elements of the minerals were determined using detection limits that were acceptably (σ) above the mean background of 0.02 wt.% for most sulphides and silicates with mean counting times of 80 s. The standards used for the silicates were: Albite (NaKα), orthoclase (AlKα, SiKα, KKα), apatite (CaKα, PKα), MgO (MgKα), MnTiO₃ (MnKα), TiO₂ (TiKα), and FeO (FeKα); for W minerals: Cassiterite (Sn Kα), pure Ta (Lα), Mo (Lα), Ni (Kα), Co (Kα), Cu (Kα), Ag (Lα), and Au (Mα) metals, SrBaNbO₄O₁₂ (Nb Lα), scheelite (WLα), Sc₂O₃ (Sc Lα), UO₂ (U Mβ), ThO₂ (Th M_β), galena (Pb Mα), LaP₅O₁₄ (La Lα), CeP₅O₁₄ (Ce Lα), YAg (Y Lα), Cd (Lα), Sb₂S₃ (SbLα), AsGa (As Lα), and Bi₂Se₃ (BiMα). The calculation of the structural formula was based on four oxygens for tungstate minerals and 14 oxygens for chlorite.

Fluid inclusion microthermometry. Four bi-polished plate thin sections (two from SHS and the other two from 3N “vertical veins”) were prepared and studied. Fluid petrography was performed on four scheelite (SHS1, SHS2, 3N1, and 3N2) bi-polished slide samples, where about 50 FIs were observed per bi-polished sample plate. The petrographic study was carried out by transmission microscopy, using a Leica DM/LSP microscope (Wetzlar, Germany) equipped with objectives up to 50× and a computer-controlled ICC50HD imaging system using the LAS EZ4 W software. The classification of FIs was performed according to the genetic criteria of Roedder [41], where the petrographic concepts of generations of FIs and fluid phase petrography were exposed by Van den Kerkhof and Hein [42], using Shepperd’s et al. [43] abacuses. Microthermometric analyses were carried out using an Olympus microscope (Tokyo, Japan) with a maximum objective lens of 80× and a 12×, equipped with a Chaixmeca stage [44] for cryometry, and a Linkam TH600 stage, coupled to a NIKON POL-Optiphot microscope, for heating. The calibration of the equipment was done regularly using a set of SynFlinc standards. The estimated accuracy was −0.1 °C for the Chaixmeca stage and 1 °C in the Linkam stage. The salinity of the fluids was calculated with the help of the software developed by Bakker [45].

Raman spectroscopy. The study was conducted with a LabRam Dilor-JobinYvon-Spex (Lille, France) equipped with a He-Ne 632.8 nm laser source, coupled to an Olympus microscope with a 100× objective. The spectral decomposition of the bands and Raman parameters was carried out with Dilor LabSpec-JobinYvon-Spex Raman software (version 1.1). Raman spectrometry analysis of aqueous-carbonic inclusions allowed the estimation of the bulk composition and density of the fluids using the methods and equations of Touret [46] and Ramboz et al. [47]. The calibration and other technical details are described in Prieto et al. [48].

Oxygen isotopes. The scheelite, quartz, and chlorite samples were selected for oxygen isotope analysis carried out at the Laboratory for Stable Isotopes at the University of Salamanca (Salamanca, Spain). A conventional procedure was used for the extraction of oxygen from scheelite, chlorite, and quartz with BrF₅ and the quantitative conversion to CO₂ prior to the analysis by stable-gas mass spectrometry [49]. The δ¹⁸O values were reported in ‰ relative to the Vienna-Standard Mean Ocean Water (V-SMOW) and normalized to the NBS-30 reference (+5.17 per mil vs. V-SMOW) and NBS-28 (+9.34 per mil vs. V-SMOW). The accuracy of δ¹⁸O measurements was ±0.2‰ (1σ). The following isotope fractionation equations were used in this work between minerals and water in the system: Wolframite-H₂O [50], scheelite-H₂O [51,52], chlorite-H₂O [53], and quartz-H₂O [54].

4. Results

4.1. Mineralogy

The selected samples (3N and SHS) are characterized by the following mineral assemblages: (a) 3N: Wolframite, scheelite, Fe,Mn-chlorite, quartz, and (b) SHS: Fe,Mn-chlorite, scheelite. Fe,Mn-chlorite is represented by the following samples used for EPMA: FMN1, FMN2, FMN3, FMN8, FMN9, and FMN12.

Wolframite scheelitization. The wolframite crystals have been partially or totally replaced by scheelite in the samples analyzed. A net contact may be observed between wolframite (white field) replaced by scheelite (grey) (Figure 2a, NII). Small scheelite crystals occur along fissures or fractures and cleavages in wolframite, where the replacement of wolframite by scheelite developed a shredded texture (Figure 2b, NII) until complete replacement. Wolframite scheelitization implies a later circulation of a rich alkaline fluid (Ca²⁺) (Figure 2c, NII).

In addition, small inclusions of sulphide (pyrite or chalcopyrite) were identified along fractures in wolframite (Figure 2d, N+). EPMA data of wolframite show a mixed composition between ferberite (F) and hübnerite (H), where the Fe amount of 0.53 atoms per formula unit (apfu) is higher than Mn (0.43 apfu) (

Table 1. Electron microprobe data of wolframite and scheelite and calculus of structural formula based on four oxygens.

Oxides	10 Analyses Minimum	Fe,Mn-W Maximum	Fe,Mn-W Average	Sigma	14 Analyses Minimum	Scheelite Maximum	Scheelite Average	Sigma
WO3	75.12	76.35	75.94	1.23	79.29	80.85	80.34	2.14
FeO		12.88	12.51	0.7		0.07	0.02	0.0738
MnO		11.03	10.62	0.66		0.13	0.03	0.00994
CaO		0.01	0	0.0024		19.32	19.18	0.68
Sc2O3		0.04	0.01	0.0274		0.02	0.01	0.02
Nb2O5		0.26	0.18	0.13		0.2	0.04	0.164
SnO2		0.03	0.01	0.031		0.13	0.01	0.126
Ta2O5		0.14	0.04	0.012		0.3	0.06	0.2951
TiO2		0	0	0		0.04	0.01	0.0172
V2O3		0.04	0.02	0.02		0.08	0.02	0.0581
Bi2O3		0.03	0.01	0.018		0.06	0.02	0.0453
MoO3						0.241	0.09	0.0321
Sb2O5		0.05	0.02	0.02		0	0	0
Total		100.86	99.36			100.47	99.81
W		0.99	1			0.99	1
Fe		0.54	0.53			0	0
Mn		0.47	0.46			0	0
Ca		0	0			0.98	0.99
Sc		0	0			0	0
Nb		0.01	0			0	0
Sn		0	0			0	0
Ta		0	0			0	0
Ti		0	0			0	0
V		0	0			0	0
Bi		0	0			0	0
Mo						0.02	0
Sb		0	0			0	0
Total		2.007	1.99			1.994	1.994

).

The hübnerite and ferberite ratio (H/F) calculated shows values from 46.25% to 46.7%. Small amounts of Nb₂O₅ (0.18%), Ta₂O₅ (0.01%), and SnO₂ (0.01%) were also measured. EPMA data of scheelite show traces of Fe, Mn and also Nb, Ta, Mo, and Sn (Table 1). X-ray element maps of wolframite replaced by scheelite show the Fe, Mn, Ca, and W distributions in wolframite and scheelite crystals (Figure 3).

Scheelite. Scheelite (CaWO₄) aggregates adjacent to wolframite or Fe,Mn-chlorite were identified. The size of the scheelite aggregates reaches from 1 to 1.5 × 0.5 to 0.75 cm². The color is orange to yellow-brown and fluoresces with a blue-white color, indicating low MoO₃ content (average <0.09%). Traces of Sn, Pb, Bi, and Mo were measured in scheelite by EPMA (Table 1).

The average chemical composition of the scheelite associated with Fe,Mn-chlorite corresponds to W_1.00Ca_0.99O₄. The X-ray element maps show the W and Ca distributions in scheelite (Figure 4). Scheelite has a structure with tetrahedral [WO₄]²⁻ and irregular dodecahedral [CaO₈]¹⁴⁻ groups, where the hexavalent Mo can substitute for W, forming a complete solid solution series extending up to powellite (CaMoO₄) [55].

Fe,Mn-chlorite. Fe,Mn-chlorite aggregates observed in transmitted light under the microscope can be well identified by their green color pleochroism, varying from pale green for some massive aggregates to a dark pleochroism, a basal cleavage, and anomalous birefringence in cross-polarized light.

A strong pleochroism in green colors with an atypical dark orange interference color was observed (Figure 5), which is in accordance with the octahedral occupancy of heavy atoms (e.g., Mn²⁺) [56] and the Fe + Mn/Fe + Mn + Mg ratio [57]. The elongation sign is opposite to the optical one and the optical plane is almost parallel to (010). The Fe,Mn-chlorite aggregates are radial with a radius size ranging from 500 μm to 1000 μm.

The crystal chemistry of Fe,Mn-chlorite (

Table 2. Electron microprobe data of Fe,Mn-chlorite and calculus of structural formula based on 14 oxygens (Fe²⁺/Fe³⁺ and OH calculated assuming full site occupancy).

Oxides	FMN 1	FMN2	FMN 3	FMN 12	FMN 8	FMN 9	Averages 146 Analyses
	N = 31	N = 29	N = 19	N = 22	N = 34	N = 11
SiO2	23.18	22.7	23.2	23.11	23.73	22.46	22.82
TiO2	0.08	0	0.02	0.01	0.04	0.04	0.06
Al2O3	19.88	19.95	19.82	19.82	20.14	20.86	20.37
FeO	39.52	39.86	39.53	39.13	39.55	38.59	39.055
MnO	6.23	5.03	5.82	6.21	5.28	5.81	6.02
MgO	1.87	1.44	1.88	1.67	1.75	1.01	1.44
CaO	0.04	0.01	0.03	0.05	0.05	0.06	0.05
Na2O	0.03	0.03	0	0.01	0	0.01	0.02
K2O	0.01	0	0	0	0	0.01	0.01
Total corrected by Cl + F	90.84	89.02	90.3	90.02	90.54	88.85	89.845
Si	2.64	2.64	2.66	2.66	2.69	2.61	2.62
Al IV	1.36	1.36	1.34	1.34	1.31	1.39	1.37
Total T site	4	4	4	4	4	4	4
Al VI	1.31	1.37	1.33	1.34	1.39	1.46	1.385
Ti	0.01	0	0	0	0	0	0.005
Fe2+	3.77	3.87	3.78	3.76	3.75	3.75	3.76
Mn	0.6	0.5	0.56	0.6	0.51	0.57	0.585
Mg	0.32	0.25	0.32	0.29	0.3	0.17	0.245
Ca	0	0	0	0.01	0.01	0.01	0.005
Na	0.01	0.01	0	0	0	0	0.005
K	0	0	0	0	0	0	0
Total O site	6.02	6	6.01	6	5.96	5.97	5.995
Total Cations	10.02	10	10.01	10	9.96	9.97	9.995
OH−	8	8	8	8	8	8	8
Mg/(Fe + Mg)	0.08	0.06	0.08	0.07	0.07	0.04	0.06
Wt% H2O calculated	10.52	10.32	10.48	10.43	10.57	10.33	10.425
Total Wt% (plus H2O)	101.36	99.34	100.78	100.45	101.11	99.18	100.27
Vacancies
(AlVI-AlIV)/2 [58]	−0.02	0	−0.01	0	0.04	0.03	0.005
(AlVI Al IV-Na-K)/2 [59]	−0.02	0	−0.01	0	0.04	0.03	0.005
Calculation with Fe3+ estimated	0	0	0	0	0.08	0.06	0.02
Mg/(Fe2 + Mg)	0.08	0.06	0.08	0.07	0.07	0.04	0.06
Vacancies
6-(Fe2 + Fe3 + Mg + AVI) [60]	0.6	0.5	0.56	0.61	0.56	0.62	0.61
(AlVI-AlIV + Fe3-Na-K)/2 [59]	−0.02	0	−0.01	0	0.04	0.03	0.005

) studied by EPMA shows an average crystal chemical composition (140 points analysed) corresponding to: (Al₁.₃₈₅Fe₃.₇₆Mn₀.₅₉Mg₀.₂₄Ca_0.01)_5.99(Si₂.₆₂₅Al₁.₃₇₅)₄O₁₀(OH)₈. There is no tetrahedral substitution [Si⇔Al(IV)], confirmed also by the calculated vacancy site (0.005) [58,59]. The R²⁺ exchanges (Mn²⁺ for Fe²⁺) are evidenced in the octahedral sheet, also confirmed by the vacancy site calculated after Inoue et al. [59]. Small amounts of Zn²⁺, Ca²⁺, Na⁺, and K⁺ were identified in Fe,Mn-chlorite. The ratio of Fe/(Fe + Mg) shows values from 0.92 to 0.94, whereas the Mg/(Mg + Fe) ratio is about 0.06. We also calculated the vacancy sites (Al^VI-Al^IV-Na-K)/2 [58], (Al^VI-Al^IV)/2 [59], and 6-/Fe²⁺ + Fe³⁺-Mg + Al^VI [60] (see Table 2).

Compositionally, the Fe,Mn-chlorite falls in the middle field between daphnite to amesite (Figure 6) in the diagram of R²⁺ vs. Si⁴⁺ [61].

The oxidation state of Fe,Mn-chlorite has previously been estimated [62], where the Fe³⁺ exhibited values between 0.00 and 0.08, with an average value of 0.02 for 145 spots analyzed.

4.2. Fluid Phase Petrography and Microthermometry

Petrography. The scheelite crystals exhibited fractures, making difficult the distinction between primary and secondary and pseudo-secondary inclusions.

The FIs occur in three preferred modes, namely: (i) In trails; (ii) isolated inclusions, and (iii) zonate ones (Figure 7;

Table 3. Petrographic characteristics of fluid inclusions observed in scheelite. Degree of fill by liquid water is an approximate range. Size is generalized, the biggest fluid inclusions is 160 μm in the 3N1 sample (see photo Figure 8a). Total number of studied fluid inclusions was n = 190.

Samples	Type	Mode of Ocurrences	Shape	Number of Phases	Size (μm)	Φw (%)
3N1 and 3N2	Primary	Isolated	Irregular or rectangular	2	>20	50–80
3N1 and 3N2	Pseudosecondary	Intragranular trails and clusters	Negative crystal, rectangular or oval	2	<30	70–80
SHS1 and SHS2	Primary	Isolated	Negative crystal and irregular	2	>20	50–70
SHS1 and SHS2	Pseudosecondary	Intragranular trails and clusters	Rectangular and negative crystal	2	<30	70–95

). Isolated inclusions are generally larger than 20 µm, whereas trails and zoning plane inclusions are smaller (average ~10 μm). The biggest measured FIs has a size about 160 μm with a negative crystal or oval shape, making the crystallographic form of the tetragonal crystal easy to identify. The FIs appear mostly dark due to internal reflections and occur in random clusters (primary inclusions) or in intergranular arrangements (pseudo-secondary inclusions).

Generally, the distinguished FIs are bi-phasic. Isolated tri-phasic primary FIs with a size exceeding 100 μm (sample SHS1), primary isolated FI, and intra-granular trails of pseudo-secondary FIs are shown in Figure 8a. Isolated FIs with a negative-crystal shape (sample SHS1), primary ones with a negative-crystal shape, and pseudo-secondary oval-shaped FIs in intra-granular trails are shown in Figure 8b. Large variations in the degree of fill (Φ_w) were also observed, especially in the 3N1-3N2 primary FIs and in both primary and pseudosecondary FIs of SHS1 and SHS2 (Table 3).

Freezing temperature, microthermometry, and salinity. The freezing temperature data include the first melting point, freezing depression, and melting of clathrate. Three different types of fluids in the FIs hosted by scheelite were identified: (i) Aqueous-carbonic fluids with CO₂ as the dominant carbonic phase, sometimes containing CH₄ and N₂; (ii) aqueous-carbonic fluids with CH₄ as the dominant carbonic phase, containing small amounts of CO₂; and (iii) aqueous fluids generally with a low to medium salinity. Aqueous-carbonic FIs with a dominant CH₄ phases (sample 3N1) are shown in Figure 9a,b; those with dominantly CO₂ phases (samples 3N2.1c, SHS1.3a, and SHS2.5b) are shown in Figure 9c,d.

Microthermometric data of aqueous-carbonic fluids dominated by CO₂ are shown in

Table 4. Microthermometric data for aqueous-carbonic fluid. AqCO₂—aqueous CO₂ FI; AqCH₄—aqueous CH₄ FI; N—number of analysis; Φw—degree of fill liquid water; T_e—eutectic temperature; T_mice—ice melting temperature; T_clat—clathrate melting temperature; T_h—homogenization temperature; Sal.—salinity in weight percent equivalents of NaCl; Xgas—amount of CO₂ or CH₄ in mol%; D—density in g/cc; Min-Max Av.—minimum, maximum, average.

Type		Φw (%)	Te (°C)	Tmice (°C)	Tclat (°C)	Th (°C)	Sal (wt.eq. NaCl)	Xgas (mol%)	D (g/cc)
AqCO2 (n = 79)	Min-max	50 to 80	−75.7 to −50	−7.2 to 0	0.2 to 11.5	200 to 380	3.5 to 5.5	80 to 100	0.55 to 1
	Av.	70	−59	−2.3	5.9	284	4.8	94.2	0.77
AqCH4 (n = 11)	Min-max	60 to 80	−68.7 to −52	−6.3 to 0	7 to 10.4	220 to 323	4.3 to 5.1	100	0.6 to 0.8
	Av.	75	−59	−2.5	9.4	262	4.7	100	0.8
Aqueous (n = 96)	Min-max	50 to 95	−67 to −23.1	−5 to 0		130 to 265	0 to 6.2		1 to 1.0
	Av.	82	−54.8	−0.75		218.5	1.18		1.01

. The T_mice of aqueous-carbonic FIs consisting mainly of CO₂ ranges from −7.2 °C to 0 °C; T_clat from 0.2 °C to 11.5 °C; T_h from 380 °C to 200 °C; salinity from 3.5 wt.% to 5.5 wt.% NaCl equiv.; and 80 to 100 mol% XCO₂. N₂ was only identified in six of the CO₂-dominant aqueous-carbonic inclusions and always with XN₂ < 8 mol%. The T_mice of aqueous-carbonic FIs with dominant CH₄ ranges from −6.3 °C to 0 °C; T_clat from 7 °C to 10.4 °C; T_h from 323 °C to 220 °C; salinity from 4.3 to 5.1 wt.% NaCl equiv.; and 100 mol% CH₄ (Table 4). The data for aqueous FIs are: T_mIce from −5.7 °C to 0 °C; T_h from 265 °C to 130 °C; and salinity from 0 to 6.2 wt.% NaCl equiv. (Table 4).

Salinity (wt.% NaCl equiv.) measured for each type of fluid was plotted against T_h (°C) where three distinct evolution stages were distinguished (Figure 10). The first stage corresponds to the evolution of the aqueous-carbonic fluid with CO₂ dominating at high temperatures (<400 °C). The separation of carbonic phases shown in the T_h histogram occurred at about 330 °C through boiling.

At this stage, the fluids lose their volatile components and start to precipitate their metals at the boiling point. Evidence for the boiling is also shown by the variations of liquid/vapor ratios observed in the FI and homogenization in liquid and vapor phase. The CH₄ and water were separated during decompression at temperatures below 300 °C and then mixed with an aqueous fluid of lower salinity. The T_h, T_mice, and salinity (wt.% NaCl equiv.) histograms corresponding to the whole set of FIs are shown in Figure 11.

The T_mi and salinity histograms show that the hotter fluids were saltier (with about 5 wt.% NaCl equiv.) than the aqueous colder fluids which mixed at the end of the process. In addition, averages of the bulk compositions and densities of the FIs studied were calculated (see Table 4). No distinction between the primary and pseudosecondary inclusions could be established using the microthermometry of the fluid inclusion assemblages, but rather on the composition of the fluids. Raman spectrometry analyses were performed on representative aqueous-carbonic type (CO₂ or CH₄ dominant) FIs, where differences in the CO₂/CH₄ ratio were identified (Figure 12 and Table S1 in the Supplementary Materials).

4.3. Oxygen Isotope Analysis

Oxygen isotope measurements were obtained for quartz, wolframite, scheelite, and Fe,Mn-chlorite (

Table 5. Oxygen isotope data of quartz, wolframite, scheelite and Fe,Mn-chlorite minerals.

Samples	Location	δ18O (0) μεασυρεδ	Quartz-Water Pair			Wolframite-Water Pair			Scheelite-Water Pair			Chlorite-Water Pair
			300 °C	350 °C	400 °C	300 °C	350 °C	400 °C	300 °C	350 °C	400 °C	300 °C	350 °C	400 °C
Quartz (wf)	SHS 48 m	12.6	5.71	7.3	8.54	-	-	-	-	-	-	-	-	-
Quartz (sch)	3N	11.5	4.61	6.2	7.44	-	-	-	-	-	-	-	-	-
Wolframite	SHS- 48 m	3.8	-	-	-	6.11	-	6.31	-	-	-	-	-	-
Scheelite	3N	2.8	-	-	-	-	-	-	4.44	5.09	5.6	-	-	-
Scheelite	3N	3	-	-	-	-	-	-	4.64	5.29	5.8	-	-	-
Scheelite	SHS 48 m	3.8	-	-	-	-	-	-	5.44	6.09	6.6	-	-	-
Fe,Mn- chlorite	3N	3.1	-	-	-	-	-	-	-	-	-	3.09	3.17	4.36
Fe,Mn- chlorite	SHS	3.2	-	-	-	-	-	-	-	-	-	3.15	3.27	4.46

), and previously published data on Fe,Mn chlorite [28] were re-analysed. The δ¹⁸O isotope (V-SMOW) compositions of quartz (Q_wf) associated with wolframite shows a mean value of 12.6‰ (1σ) and a mean value of 11.5‰ (1σ) for the quartz join to scheelite.

The δ¹⁸O_Fluid of the quartz-water pair (10³lnα_{Quartz–water} = 3.38 × 10⁶T⁻²−3.40; [49]) yielded the following values for the precipitated quartz (Q_wf) associated with wolframite: +5.71‰ (1σ) at 300 °C, +7.30‰ (1σ) at 350 °C, and 8.54 (1σ) at 400 °C. For the precipitated quartz (Q_sch) associated with scheelite yielded: +4.61‰ (1σ) at 300 °C, +6.20‰ (1σ) at 350 °C, and +7.44‰ (1σ) at 400 °C (Table 5). The δ¹⁸O (V-SMOW) measured for wolframite yielded a value of 3.8‰ (1σ) (i.e., much lighter than quartz). The δ¹⁸O_Fluid of the wolframite-water pair (10³lnα_Wf–water = 1.03 × 10⁶T⁻²−4.96; and 10³lnα_Wf–water = 0.21 × 10⁶T⁻²−2.91 for temperature intervals between 250–370 °C and 370–420 °C [50]) yielded a value of 6.11‰ (1σ) for an estimated temperature of 300 °C and 6.31‰ (1σ) for 400 °C.

Combining the quartz-water fractionation equation [54] with the wolframite-water pair [50], the following oxygen isotope fractionation equations were proposed for the quartz-wolframite pair [63]: (i) 250–370 °C: 10³lnα_q-_Wf = 2.35 × 10⁶T⁻² + 1.56 and (ii) 370–420 °C: 10³lnα_q-_Wf = 3.17 × 10⁶T⁻²–0.49. The quartz-wolframite oxygen isotope fractionation estimated for temperatures ranging from 300 °C to 400 °C yielded 9 and 7.11‰ (1σ). The data plotted (not shown) in the 10³lnα vs. T °C [63] fall between 300 °C and 400 °C.

The δ¹⁸O isotope (V-SMOW) composition measured on scheelite yielded values of 2.8‰, 3‰, and 3.8‰ (1σ). The δ¹⁸O_Fluid of the scheelite-water pair (10³lnα_{scheelite-water} =1.39 × 10⁶T⁻²−5.87; [52]) yielded 4.44‰, 4.64‰, and 5.44‰ (1σ) for 300 °C, 5.09‰, 5.29‰, and 6.09‰ (1σ) for 350 °C, and 5.60‰, 5.80‰, and 6.60‰ (1σ) for 400 °C (Table 5. The oxygen isotope fractionation in the quartz-scheelite and quartz-wolframite pairs is large, offering a much more sensitive geothermometer [64].

Combining the quartz-water fractionation equation [54] with the scheelite-water equation [52], the following fractionation equation was obtained for quartz-scheelite fractionation: 10³lnα_q-scheelite = 1.99 × 10⁶T⁻² + 2.47. The ∆_{quartz-scheelite} calculated was 6.86‰ (1σ) for a T °C of 400 °C and 7.54‰ (1σ) for 350 °C. The ∆_{quartz-scheelite} values plotted in the diagram of 10³lnα vs. T °C (not shown) intersected the quartz-scheelite fractionation curve at about 350–400 °C.

The δ¹⁸O isotope (V-SMOW) signatures measured for Fe,Mn-chlorite yielded values of +3.1 and +3.2‰ ± 1.2‰, where the δ¹⁸O_Fluid calculated for the chlorite-water pair (10³lnα_{chlorite-water} = 1.56 × 10⁶T⁻² − 4.70; [53]) yielded +3.27‰ for a T °C of 350 °C, +4.46‰ for 400 °C (Table 5), and +4.92‰ for 450 °C. The ∆_{chlorite-water} fractionation factor calculated was about +0.05‰ for 330 °C and values of −0.68‰ and −1.26‰ corresponded to temperatures of 380 °C and 450 °C, respectively. After plotting the data in the 10⁶ × T⁻² (°K) vs. 10³lnα_{chlorite-water} diagram (Figure 13) [65], an average crystallization temperature of 400 °C was obtained.

4.4. Chlorite Geothermometry

Several chlorite thermometers based on empirical calibrations (linking T to Al^IV content) or semi-empirical models (linking T to the chlorite + quartz + water equilibrium constant) have been developed for Al-, Fe-, and Mg-rich chlorites over the last decade [58,59,60,67,68,69]. In a recent review on chlorite geothermometry, Bourdelle [29] noted that it could be worth applying these models for Mn-rich or Al-free chlorites to Mn-redox behavior, like Fe-redox.

In this study, 146 chemical analyses of Fe,Mn-chlorite (Table 2) were used for evaluating both empirical and semi-empirical models known in the literature. The average amount for FeO was about 39.05% and 6.05% for MnO, where the estimated Fe³⁺ was about 0.02 apfu. High contents of Fe [Fe/(Fe + Mg + Mn) = 0.4–0.95] and Mn [Mn/(Fe* + Mg) > 0.05] in chlorite minerals are also diagnostic of chemical environments related to ore-forming processes [70]. The Fe,Mn-chlorite temperature values obtained using semi-empirical geothermometers [58,59,60,71] are shown in

Table 6. Calculated temperatures using semi-empirical thermometry.

Samples Calculated Temperatures (°C)	FMN1	FMN2	FMN3	FMN12	FMN8	FMN9
Semi-Empirical Thermometry	T °C	T °C	T °C	T °C	T °C	T °C
Inoue et al. [60,71]. FeTotal = FeO + Fe2O3. Quadratic equation thermometer (valid for <350°, may be <400 °C)	229	246	230	220	212	228
Inoue et al. [71]. FeTotal = FeO (valid for <350 °C, may be <400 °C)	197	210	197	190	183	196
Bourdelle et al. [58] FeTotal = FeO. (Valid for <350 °C and <4 kb)	-	-	-	-	435	519
Lanari et al. [59]. Chl1: FeTotal = FeO + Fe2O3 (valid between 100–500 °C)	-	-	-	-	718	984
Lanari et al. [59] Chl2: FeTotal = FeO (valid between 100–500 °C and 1–20 kbar)	-	-	-	-	373	458

.

The average calculated temperature of Fe,Mn-chlorite [59,70] was estimated at about 228 °C (Fe_total = FeO + Fe₂O₃) and 196 °C (Fe_total = FeO). The temperatures estimated for samples FMN 8 and FMN 9 were respectively about 435 °C and 519 °C (Fe_total = FeO) after Bourdelle [58] and 373 °C and 458 °C (Fe_total = FeO) after Lanari [59]. Here, the geothermometers of Bourdelle and Lanari [58,59] are not applicable (

Table 7. Calculated temperatures using empirical thermometry *.

Samples Empirical Thermometry (FeTotal = FeO)	FMN1 T °C	FMN2 T °C	FMN3 T °C	FMN12 T °C	FMN8 T °C	FMN9 T °C
Cathelineau [66]. (valid for <350 °C)	375	376	371	371	359	387
Jowett [67]. For Si < 3.3 apfu and Ca < 0.07 apfu.	393	395	389	389	377	406
Fe/(Fe + Mg) < 0.6, and T° < 325 °C
Restrictions of each thermometer calibration (1 = Ok; 0 = not valid)
Bourdelle et al. [58].	NOT APPLICABLE	NOT APPLICABLE	NOT APPLICABLE	NOT APPLICABLE	NOT APPLICABLE	NOT APPLICABLE
Uncertainty of 30 °C for T < 300 °C and 50–60 °C for T > 300 °C
Na2O + K2O + CaO <1? (wt.%)	1	1	1	1	1	1
Si (apfu) between 2 to 4 apfu	1	1	1	1	1	1
Vacancies >0.05 apfu?	0	0	0	0	0	0
Lanari et al. [59].	NOT APPLICABLE	NOT APPLICABLE	NOT APPLICABLE	NOT APPLICABLE	NOT APPLICABLE	NOT APPLICABLE
Uncertainty of 50 °C.
Si < 3 apfu	1	1	1	1	1	1
Na2O + K2O + CaO <1? (wt.%)	1	1	1	1	1	1
Vacancies (Fe Total = FeOT) > 0.05 (for <0.05 = Thermometer Chl2 not applicable)	T° Chl2 should be not used	T° Chl2 should be not used	T° Chl2 should be not used	T° Chl2 should be not used	T° Chl2 should be not used	T° Chl2 should be not used
Inoue et al. [60,71].	Composition Valid	Composition Valid	Composition Valid	Composition Valid	Composition Valid	Composition Valid
Vacancies apfu < 1	1	1	1	1	1	1
Na2O + K2O + CaO < 0.5	1	1	1	1	1	1

* Thermometry calculations was carried out using the database of Verdecchia et al. [72].

). The average temperature calculated for Fe,Mn-chlorite crystallization was about 373 °C, after Cathelineau [67], and 392 °C after Jowett [68]. The chemical composition of Fe,Mn-chlorite plotted in the Si⁴⁺ vs. R²⁺ diagram [61] falls into a field with temperatures ranging from 400 °C to 550 °C (Figure 6).

5. Discussion

The successful application of chlorite minerals and stable isotope geothermometry to ore deposits requires several conditions to satisfy any pair of minerals: (i) Fractionation factors between them are well-calibrated and strongly temperature dependent; (ii) the mineral(s) species were in isotopic equilibrium at the time of deposition; and (iii) they have retained their isotopic composition since formation. Therefore, the sequence mineral pairs identified in the W, (Cu,Mo) ore deposit of Borralha satisfy all the requirements to evaluate the isotope geochemistry relationship between tungstates and neoformed silicates (quartz and Fe,Mn-chlorite) and empirical chlorite geothermometers, based also on FI microthermometry data.

5.1. Wolframite to Scheelite and Fe,Mn-Chlorite Crystallization

The scheelite samples collected are excellent samples due to their size, purity, and their genetic relationship with Mn-bearing wolframite and Fe,Mn-chlorite. Wolframite crystallized from fluids with slightly acidic to neutral pH, where WO₄²⁻ may be easily complexed with Fe²⁺, Mn²⁺, and Ca²⁺. It should be noted that the stability of the Fe²⁺ end-member (ferberite) is much more sensitive to the high variations of f_O2 than the Mn²⁺ end-member wolframite (hübnerite) since the composition of ferberite might be strongly affected by the oxidation state [26]. Only Mn-bearing wolframite is associated with the late evolving W mineralization, possibly triggered by an alkalinity rising in the fluid as a result of an increasing S activity and the availability of Ca²⁺ in the system, which is correlated with a late alkaline metasomatic reaction common in geochemical settings [73]. Several authors have suggested that Mn-rich wolframites precipitated at higher temperatures than Fe-rich wolframites [73,74,75,76]. Nevertheless, the H/F ratio used as a geothermometer [77] failed, because temperature is not the only factor controlling the composition of wolframite [27]. The H/F ratio can be used as an indicator of contrasting W deposition environments in perigranitic ore-forming systems, where hübnerite can be viewed as a relatively simple indicator of a strong magmatic control on W deposition [76]. The H/F ratio measured in wolframite samples from Borralha ranged from 46.25% to 46.70%, as a consequence of the high Mn/Fe ratio in the magmatic-hydrothemal fluid. By contrast, a H/F ratio < 40 and down to 0 indicates a W deposition environment where a significant non-magmatic contribution to the ore fluid and W deposition should be necessary [78].

Experimental studies on the stability of H₄[Si(W₃O₁₀)₄]2H₂O at high P-T conditions and variable pH in the presence of Ca²⁺ have confirmed the separation of silico-tungstic acid under the effect of solution neutralization [79], where [WO₄]²⁻ is preferentially combined with Ca²⁺ to form scheelite, favoring the incorporation of the available Fe²⁺ and Mn²⁺ from wolframite and SiO₂ from the fluid into the chlorite structure. The CO₂ loss from the hydrothermal fluid during boiling caused pH changes from acidic to alkaline conditions [80,81,82,83], accompanying the wolframite dissolution and scheelite deposition in the presence of Ca²⁺. Calcium for scheelite crystallization was supplied from the albitization and/or sericitization of plagioclase from granitic and pegmatite host rocks. Optical microscopy showed that the scheelitization of Mn-bearing wolframite occurred along fissures and fractures. This implies that high-pressure fluids are required to trigger fracture initiation and propagation, where the fluctuations of fluid pressure were recorded by FIs. Hence, Mn-bearing wolframite supported a dissolution process where [WO₄]²⁻ separation served for scheelite crystallization, whereas Mn and Fe for chlorite. It is important to note that the term “metamorphic fluid” is used here to describe a fluid that was partially equilibrated with metamorphic rocks and this does not necessarily imply that the fluid components were produced from prograde metamorphic reactions.

5.2. Characterization of the Mineralizing Fluid

Previous FIs studies on scheelite from Borralha revealed two generations of FIs (primary and pseudosecondary) characterized by the presence of an aqueous, low salinity fluid [38]. The fluids entrapped in scheelites showed a complex history with different fluids and boiling during W precipitation, where boiling (immiscibility), mixing, cooling, and fluid-rock interactions (e.g., Ca²⁺ needed for scheelite precipitation) played an important role in the scheelite precipitation. The primary fluid stage was characterized by a complex aqueous-carbonic fluids of low salinity, indicating a typical metamorphic fluid. The CO₂ ubiquitous recorded in FIs from a variety of mineralized systems underwent boiling/unmixing presumably leading to ore deposition. Scheelite was mainly precipitated during the early stage from an aqueous-carbonic fluid with a CO₂₋dominated composition, where the T_h of the ore-forming fluids decreased from 380 °C to 200 °C.

Fluid boiling/immiscibility has two fundamental effects on the fate of metals. Indeed, it is known to induce the precipitation of some metals from the liquid phase owing to enhanced metal concentrations and to change the redox conditions and ligand contents in the residual liquid as a result of the removal of volatile components (H₂S, H₂, HCl, and CO₂) into the vapor [81,82,83]. The partitioning of CO₂ and CH₄ into the vapor phase occurred through boiling, where both participated in redox reactions. The coexistence of vapor- and liquid-rich inclusions indicates fluid boiling/immiscibility. The boiling resulted in the separation of the aqueous-carbonic fluid with predominant CH₄ and mixing with an aqueous fluid of low salinity at about 330 °C, deduced from the histograms (Figure 8) and the diagram of salinity vs. T_h (Figure 9). The salinity (wt.% NaCl equiv.) vs. T_h (°C) diagram shows two distinct fields corresponding to a magmatic-hydrothermal vs. metamorphic fluid (Stages I and II) and a hydrothermal-meteoric fluid (Stage III) (Figure 9).

Stage I occurred around the maximum T_h measured, corresponding to the beginning of separation between CO₂-CH₄ fluids during boiling (immiscibility). The L-rich FIs appeared to be in equilibrium with the CO₂-CH₄ inclusions (see Figure 10 and Figure 11a), which is another indication of boiling. Boiling also increased oxygen fugacity, leading to the immiscibility of CH₄ where a disequilibrium isotopic fractionation occurred between CO₂ and CH₄ below 400 °C [83]. The process ended during the final of Stage II, when the aqueous saline FIs were generated. Changes in the aqueous-carbonic fluid composition were recognized as temperature decreased, where CH₄ became dominant (Stage II) and the T_h decreased (average of 262 °C). Stage III corresponded to the dilution of metamorphic fluids and cooling by the circulation of meteoric fluids. Exhaustion of deposition led to low salinity FIs. This stage corresponded to an aqueous fluid of lower temperature (average T_h = 218.5 °C). In addition, the salinity gradually decreased as temperature decreased (Figure 10 and Figure 11c). The large dispersal of T_h values in the histogram and the reduced salinity observed for the aqueous inclusions was indicative of mixing.

Experimental studies have demonstrated that the solubility of scheelite decreases substantially with decreasing temperatures [73], where tungstate destabilization was influenced by the pH changes of the fluid [84]. This process removed CO₂ from the hydrothermal fluid, a component that can not only be a buffer fluid pH [73,85], but also favors W transport [86]. The trace amount of N₂ addition to the mineralizing fluids derived from the devolatilization of metasediments [87].

5.3. Oxygen Isotopes

The δ¹⁸O_Fluid calculated for the quartz-water pair (sample Qwf) yielded values of + 5.7‰ at about 300 °C, + 7.30‰ at 350 °C, and + 8.54‰ at 400 °C, where the calculated values fall within the magmatic water field [88,89] from 5.5‰ to 10.0‰ [90,91,92]. In fact, from the isotopic point of view, quartz (Q_wf) and wolframite (+ 6.11 to + 6.31‰ at 300 and 400 °C, respectively) crystallized directly from a magmatic-hydrothermal fluid, which may represent primary magmatic water or water from any source whose isotopic composition is controlled by an exchange with a large volume of igneous rock. Nevertheless, the δ¹⁸O_Fluid calculated for quartz (Q_sch) and scheelite yielded values from 5.55‰ to 6.6‰ for sample SHS, for temperatures ranging from 300 °C to 400 °C; for sample 3N values of 4.44‰ to 5.8‰ were obtained in the same temperature range. The samples analyzed suggest several scheelite generations crystallized from a magmatic-hydrothermal vs. metamorphic fluid (sample SHS) at temperatures below 400 °C, which were later diluted by a hydrothermal meteoric fluid (sample 3N). In fact, the differentiation between magmatic and metamorphic fluids in terms of δ¹⁸O_Fluid is limited, creating difficult interpretations concerning the mixing of these two fluid phases. According with the FIs data, scheelite crystallization could be largely influenced by metamorphic fluids which equilibrated with the magmatic-hydrothermal fluid during the first wolframite crystallization. δ¹⁸O_Fluid data support an input of hydrothermal-meteoric waters, confirming an exchange and re-equilibration with the primary fluid at elevated temperatures and low water/rock ratios. This assumption is supported by FIs data obtained from the studied scheelite samples, where an input of aqueous FIs after boiling was observed (Figure 10). Based on δ¹⁸O data, it seems that Fe,Mn-chlorite crystallized in equilibrium with scheelite, but from a diluted hydrothermal-meteoric fluid contribution at a high temperature. The crystallization temperatures estimated from the 10⁶ × T⁻² (°K) vs. 10³lnα_{chlorite-water} diagram (Figure 13) [65,66] were between 350 °C to 450 °C, where the chlorite-water pair fractionation yielded values from +3.17‰ at 350 °C to +4.46‰ at 400 °C, confirming the presence of a hydrothermal-meteoric fluid in equilibrium with scheelite crystallization.

The mixing of magmatic and metamorphic fluids with different origin in granite systems has been inferred in the literature, where the external fluids (i.e., metamorphic, hydrothermal-meteoric) can mix and dilute the primary magmatic fluid, even when enriched by metals extracted from wall-rocks in the frame of the resulting convective circulation [93,94,95].

5.4. Chlorite Geothermometer

Chlorite consists of a 2:1 layer with a negative charge [(R^{2 +}, R^{2 +})₃(_xSi_4xR^{2 +}_y)O₁₀OH₂]⁻ balanced by a positively charged interlayer octahedral sheet [(R^{2 +}, R^{3 +})₃(OH)₆]⁺. R^2+,3+ is represented by Mg, Fe(II), and Fe(III) with Mg-rich chlorite (clinochlore) generally being metamorphic (high temperature), while the Fe-rich chlorite (chamosite) is typically diagenetic (low temperature). Additionally, Ni-rich chlorite (nimite) and Mn-rich chlorite (pennantite) are the other two less common varieties of chlorite. Most chlorites are trioctahedral in both sheets, i.e., the ferric iron content is low. Furthermore, it is well known that the ubiquitous and large thermodynamic stability of chlorite is in metamorphic, diagenetic, and hydrothermal systems, where its chemical composition is influenced by temperature and pressure. Thus, chlorite may crystallize over a large range of temperatures from 80 °C to more than 700 °C and pressures from 1 Kbar to 12 Kbar [96]. For example, cookeite, a Li-di,trioctaedral chlorite, is an example of a polymorph that has crystallized at low to high pressures [97,98].

The Fe,Mn-chlorite associated with W mineralizations from Borralha crystallized directly from the hydrothermal fluid and isotopically re-equilibrated with respect to scheelite. Hydrothermal fluids in the crust generally have a low oxidation potential and dissolved Fe predominantly occurs in its +2 oxidation state [99]. Under these reduced conditions, the amount of Fe³⁺ in hydrothermal fluids is negligible compared to that of Fe²⁺ [71]. Both Fe and Mn were available in the hydrothermal fluid from wolframite replaced by scheelite. Considering the textural relationships observed between wolframite to scheelite and Fe,Mn-chlorite, and the high amount of Fe and Mn in the hydrothermal fluid, chlorite crystallized after scheelite preserving the redox conditions during the fluid migration. The redox conditions are supported by the obtained FI data, where CO₂ and CH₄ subsequently reduced oxidation. By contrast, adequate oxidizing conditions were needed to buffer the oxygen activity in the case of a hydrothermal fluid with a Fe- or Mn-poor or -rich character and a low fluid/rock ratio. In this case, both Fe³⁺ and Mn³⁺ had a limited stability field in aqueous solution, being rapidly trapped during chlorite crystal growth.

Several authors have suggested that empirical or semi-empirical geothermometers have only been developed for the most common chlorites (i.e., Al-, Fe-, or Mg-rich chlorites) and it could worth extending to rarer chlorites, like Mn-rich chlorite [29], because the actual geothermometers do not include a Mn-rich end-member, such as pennantite. Nevertheless, the systematic presence of measurable amounts of Mn has been tested as a new pathfinder for chlorites developed in the course of ore-forming processes. A valid example is related to Fe,Mg,Mn-chlorite identified in Pb-Zn, Bi, and Ag ore veins from the Toyoha geothermal system, Japan [70]. In this case, the structural formula calculation, similar to ferric iron, will arise without the quantification of the total Mn³⁺/Mn ratio, but the ratio of Mn³⁺/Mn²⁺ would probably theoretically modify the chlorite’s structural formula [29]. Nevertheless, the presence of Mn³⁺ in chlorite, which would eventually be substituted with Al³⁺ in the octahedral site, is expected to lead to the structural distortion of chlorite due to the different ionic radii of Al³⁺ (0.535 Å) and Mn³⁺ (0.645 Å) [100].

The chemical compositions of Fe,Mn-chlorite are plotted in the field of daphnite (average Si = 5.30 apfu and Fe 7.51 apfu) in the Si vs. Fe_tot diagram (not shown) or in the field of daphnite-amesite according to the Si vs. R²⁺ diagram (Figure 6; [61]), which makes it valid for the four-thermodynamic-component solid-solution model (Mg-sudoite, Mg-amesite, clinochlore, and daphnite) used for semi-empirical geothermometers.

The estimated Fe,Mn-chlorite temperatures using the semi-empirical thermometer of Lanari [59] were about 373 °C and 458 °C (Fe_total = FeO) and 435 °C and 519 °C (Fe_total = FeO) (Bourdelle [58]). This represents average values of 415 and 477 °C. If we consider the uncertainly of 50–60 °C for T °C > 300 °C suggested by Bourdelle [58], we propose a temperature between 375 °C to 410 °C. The estimated temperatures calculated from semi-empirical thermometers seem to be supported by the δ¹⁸O_Fluid calculated using the fractionation equation for the chlorite-water pair [53] which confirmed a temperature of chlorite crystallization estimated at around 400 °C.

The magmatic-hydrothermal vs. metamorphic fluid circulation evidences was proposed in this work based on FI data and the oxygen stable isotope fingerprints obtained on tungstates, quartz, and Fe,Mn-chlorite (Figure 14). The presence of a metamorphic fluid with an aqueous-carbonic composition and low salinity was contributed to wolframite destabilization and scheelite precipitation in the presence of Ca²⁺ derived from plagioclase alteration. During boiling, the CO₂ was removed from the aqueous-carbonic fluid, favoring W transport [84], where Fe,Mn-chlorite crystallized after scheelite crystallization at a temperature close to the crystallization of scheelite.

6. Conclusions on Fe,Mn-Chlorite Thermometry

The diversity of chlorite minerals in fossils or active hydrothermal systems is connected in most cases with the alteration types of magmatic rocks, where chlorite may crystallize directly from a solution or replace the primary minerals. The sequence of mineral pairs identified in the W, (Cu,Mo) ore deposit of Borralha satisfies all conditions to evaluate the oxygen isotope signature between tungstates and silicates. Besides, the question arises on Mn-oxidation, the effect of fluid chemistry on the mineral-water fractionation factor in chlorite may represent an uncertainty however, fractionation factors between minerals and aqueous solutions are independent of solution chemistry.

Our multiproxy study based on FIs, stable isotopes, and semi-empirical chlorite thermometers showed good consistency between the results obtained. Combining FIs and oxygen isotope geochemistry data, it was revealed that a metamorphic fluid supported the crystallization of scheelite and quartz at temperatures below 400 °C. The geothermometers applied according to Lanari and Bourdelle for Fe,Mn-chlorite produced temperature estimates between 373 °C and 458 °C, and 435 °C and 519 °C, respectively. The ∆_{chlorite-water} calculated fractionation factor corresponded to temperatures of 380 °C and 450 °C, respectively. An average chlorite crystallization temperature of 400 °C was assumed considering an uncertainly of 50–60 °C for T °C > 300 °C, as suggested by the authors of the geothermometers.