Mineralogy and Fluid Inclusion Constraints on the Genesis of the Recently Discovered Ag-(Ni-Co-Sb-As-Hg ± Bi) Vein Ore Shoot Mineralization in the Aouli Pb-Zn District (Upper Moulouya, Morocco)

Abstract

Unusual Ag-(Ni-Co-Sb-As-Hg ± Bi)-bearing fault-fill vein ore shoot mineralization set in a gangue of quartz, fluorite, and barite has been identified in Morocco’s Aouli deposit. The Paleozoic host rocks consist of a succession of Cambrian to Ordovician-aged folded and low- to medium-grade metasediments and metavolcaniclastic rocks with tuff interbeds and amphibolite sills, locally intruded by late Visean calc-alkaline to alkaline granitoid intrusions. Paragenetic relationships indicate that the sequence of ore precipitation comprises a succession of Ni-Co-Fe arsenides, followed by Pb-Sb-As-Ag-Hg sulfarsenides/sulfosalts and then Zn-Pb-Fe sulfides. Results indicate that the ore shoot mineralization formed from episodic stages of fracturing and subsequent fluid migration. Precipitation of ore phases is thought to have occurred as a result of isothermal mixing and subsequent fluid–rock interactions. The timing of mineralization is thought to have occurred between Late Triassic and Late Miocene, coinciding with major crustal extension and Middle Jurassic–Upper Cretaceous alkaline magmatism. Thermal convection and seismic pumping are proposed as the main driving force for the large-scale migration of the ore-forming brines. This research bears directly upon the potential for new exploration targets in Pb-Zn ± fluorite ± barite deposits hosted in Variscan inliers throughout North Africa.

1. Introduction

Unconformity-related Pb-Zn ± fluorite ± barite vein systems are widespread throughout the European Variscan belt [1,2] and its North African counterpart [3,4]. Sulfide mineralization occurs at different stratigraphic levels of the metamorphic basement and the unconformably overlying Mesozoic-to-Cenozoic sedimentary cover. In addition to Pb-Zn ± fluorite ± barite, recent mineralogic and geochemical studies revealed that part of these veins are host to unusual native As-sulfide ± native Ag ± Ni-Co-Fe arsenide ± antimonide ± Ag-Bi sulfosalt-bearing orebodies referred to as vein ore shoots [1,2,5,6,7,8].

The Aouli fault-fill deposits hosted by Paleozoic metasediments in the Upper Moulouya metallogenic province (Figure 1) rank together with those of Tighza [9,10,11], Roc Blanc [12], and Sidi Lahcen [13] among the four major producers of Pb-Zn ± fluorite ± barite-type argentiferous veins in Morocco, with significant Ag output (locally up to 4500 g/t Ag). Between 1926 and 1983, more than 9.6 Mt of run-of-mine ore were produced, with an average extraction grade of 5% Pb and 400 g/t Ag. Apart from the well-constrained Pb-Zn ± fluorite ± barite mineralization event [14,15,16], a new Ni–Co–Fe arsenide, Sb-As-Ag-Hg ± Bi-bearing sulfide and sulfosalt paragenesis, referred to thereafter as Ag-(Ni-Co-Sb-As-Hg ± Bi) vein ore-shoot mineralization, has recently been recognized in different mineral associations from the main exploited veins.

Here, we provide detailed investigations of the mineralogical, paragenetic, and fluid characteristics of these newly discovered vein ore shoots at the Upper Moulouya metallogenic province. As such, a refined paragenetic sequence that underpins a new genetic model is proposed based on an integrated approach that combines field observations and mineralogical, textural, and paragenetic constraints, along with fluid inclusion microthermometry. The combined use of these analyses provides a suitable framework for a better understanding of the significance and origin of the Ag-(Ni-Co-Sb-As-Hg ± Bi) ore shoot mineralization.

2. Regional and District-Scale Geological Settings

The ~15–50 km wide and 100 km long Upper Moulouya massif, which hosts the Aouli deposits (this study), is part of the eastern Meseta being sandwiched between the Middle Atlas to the northwest and High Atlas to the south (Figure 1). It is an intermontane Mesozoic basin that formed as a result of subsidence during the opening of the Jurassic Maghrebian Tethys and Atlantic Ocean [18,19], and subsequent inversion and uplift during the Late Cretaceous (Senonian) compressional–transpressional event [20,21]. There, the Paleozoic basement is mainly exposed in the Aouli-Mibladen and Bou Mia inliers surrounded by Mesozoic and Cenozoic strata (Figure 1).

Sedimentation rates increased through the Jurassic as rifting continued with the breakup of Pangea, and the opening of the neo-Tethys Ocean and the North Atlantic [22]. Two rifting episodes operating from the end of Early Jurassic to Late Jurassic were documented [23,24]. While the early Jurassic rifting 1 is related to the hyper-extension of the Atlantic Ocean, the late Jurassic rifting 2 is linked to the hyper-extension of the Maghrebian Tethys. During the Mesozoic, the Upper Moulouya massif formed an emergent structural high that separates two subsiding basins, i.e., the Middle Atlas to the northwest and the High to the south [25], whereas during the Cenozoic, the massif became an intramontane basin of uplift-related sediment input sourced from the surrounding Middle and High Atlas belts [26,27]

From a metallogenic perspective, the upper Moulouya massif is unique in containing a wide range of mineral deposits including the following (Figure 2): (1) vein-type (i.e., Aouli deposit described herein; 9.6 Mt at 5% Pb), (2) stratiform diagenetic-type (i.e., Zeida deposit; 16 Mt at 3% Pb), and (3) stratabound Mississippi Valley-type (MVT) (i.e., Mibladen deposit; 6.5 Mt at 5% Pb) [15].

In the studied district, the oldest rocks are of presumed Cambrian–Ordovician age [29,30] consisting of a succession of folded and low- to medium-grade metasediments and volcaniclastic rocks with tuff interbeds, and amphibolite sills (Figure 1). Locally near Engill vein (Figure 3), the exposed sedimentary succession shows the occurrence of a dark-colored muddy rock sequence, up to 120 m thick, consisting of organic-rich siliceous shale interbedded with thin fine-grained carbonate beds. The regional stratigraphic relationships for these black shales, which could be source rocks for petroleum and natural gas, remain poorly constrained, being assigned to the Carboniferous [31]. Several late Visean calc-alkaline to alkaline metaluminous to peraluminous granitoid intrusions (ca. 330–340 Ma [32]) intruded the Cambrian–Ordovician country (Figure 1).

Structurally, the metasedimentary host rocks were affected by two main phases of Variscan deformation, the earliest of which is thought to have occurred during the Tournaisian–Visean (i.e., Eovariscan D₁ phase at 360–330 Ma [30,33,34]), whereas the later Late Westphalian phase occurred under an extensional regime coinciding with the end of the D₁ phase and subsequent emplacement of the late Visean granitoid intrusions. A contact metamorphic aureole of ~1500-m width formed around the Aouli granitic batholith. The peak P-T conditions (400–640 °C, <3 kb), inferred from the main metamorphic assemblages along with Ti-in-biotite and amphiboles geothermometry, correspond to batholith emplacement depths ranging from 4 to 7 km [35,36,37].

Unconformably overlying the Neoproterozoic–Paleozoic basement, the Mesozoic to Cenozoic stratigraphy consists of syn-rift, ~400–500 m thick Triassic red bed conglomerates, salt-bearing argillites and arkoses with CAMP basalt interbeds, and Jurassic (Liassic, Toarcian, Aalenian–Bajocian) limestones, dolostones, and marls. The post-rift Cretaceous–early Tertiary formations comprise Cretaceous (Lower Cenomanian, Cenomanian–Turonian) conglomerates, shallow and marginal marine limestones, calcareous shales, dolostones, and interbedded anhydrite-rich shales, Cenozoic (Oligocene) continental fluvial to lacustrine deposits, and Plio–Quaternary alkali basalts (Figure 1). The major brittle structures comprise an array of NNE-SSW to NE-SW, and WNW-ESE-trending fault systems (Figure 1) which are interpreted as pre-Variscan in age but have been repeatedly reactivated during the later Alpine orogeny and currently represent significant features of extension (thinning) or trans-tension [38].

3. Sampling and Analytical Methods

Following extensive field mapping, representative fresh and hydrothermally altered host rocks along with barren and mineralized veins across each of the five ore zones (i.e., Aouli, Sidi Ayad, Poulet, Sidi Said, and Ansegmir; Figure 3) were collected for petrographic and mineralogical characterization and geochemical analysis based on texture, mineral assemblages, and paragenetic position. The sampling sites, preferably those of high-grade ore, include surface exposures, old mining pits, and underground mining galleries (Figure 3). Specimens of massive ores (n = 50) and associated gangue phases (n = 50) with visible arsenide (n = 20) were collected mainly from Sidi Said underground mining workings due to the inaccessibility of galleries from the other ore zones.

In the Sidi Said area, sampling targeted massive to stringer to disseminated arsenide (n = 20), sulfarsenide/sulfosalt (n = 20), and sulfide (n = 50) mineralization from both the veins and the adjacent wall rocks. The sample locations are shown in Figure 3. More than 120 polished thin sections were investigated under optical microscope and scanning electron microscopy (SEM) and scanning electron microscope-backscattered electron (SEM-BSE) imaging for mineral assemblage and textural relationship characterization prior to mineral, fluid inclusions.

3.1. Optical and Cathodoluminescence Microscopy

Mineral identification along with textural analysis and cold-cathodoluminescence (CL) imaging were carried out at the Isotope Studies Laboratory, Institute of Geosciences, UNAM, Juriquilla, Mexico, using an Olympus SZ-X microscope equipped with a Qimage Micropublisher 5 Mp digital camera, a Reliotron^® CL instrument, and a HITACHI TM3000 scanning electron microscope (SEM). The CL instrument was typically operated with a beam current of 500 μA and a corresponding voltage of ~15 kV. The CL imaging of the fluorite, quartz, apatite, and sphalerite was performed in complement to the classical optical petrography to better constrain textural relationships and relative age-constraints of the studied fluid inclusion assemblages (FIAs).

3.2. Electron Probe Microanalysis (EPMA)

Minerals in grain mounts and thin sections were imaged by backscattered electrons (BSE), and major and minor element data were collected at the microanalysis laboratory of Université Laval (Canada) using a CAMECA SX-100 electron microprobe. For arsenide, sulfosalt, and sulfide analyses, the instrument was operated with an accelerating voltage of 15 kV and a beam current of 20 nA with a spot diameter of 2 μm and counting times of 15–20 s for peak measurements. The element lines used for analysis included Kα lines for S, Cu, Ni, Co, Fe, Zn; Lα lines for Ba, Ag, As, Sb; Mα lines for Pb; and Mβ lines for Hg. The following in-house and published standards were used for calibration: sphalerite (for Zn), chalcopyrite (for Cu), pure Ni (for Ni), pure Co (for Co), hematite (for Fe), GaAs (for As), pure Ag (for Ag), Sb (for Sb), HgS (for Hg), galena (for Pb), and marcasite (for S).

3.3. Laser Ablation (LA)-ICP-MS Analysis

Trace element concentrations of apatite and rutile were measured at the Isotope Studies Laboratory, Institute of Geosciences, UNAM, Juriquilla, Mexico, using a Coherent Compex 102 excimer laser ablation system with an ATL excimer ArF laser source at 193 nm wavelength and ~5 ns pulse width coupled to a Thermo iCapQc inductively coupled plasma mass spectrometer instrument. Following a 15 to 20 s period of background analysis, apatite crystals were ablated with a 60 μm laser beam for 35 s at a 4 Hz repetition rate using a laser energy density of 4 J/cm² [39], and 15 s of washout. The ICP-MS was tuned following the experimental parameters reported in [40,41]. Spot ablation was carried out using a 24 µm spot size in the epoxy-grain-mount apatite samples and a 10 µm spot size in the thin-section apatite samples to avoid abundant inclusions. Data reduction and concentration calculations were performed using the Iolite 4 software [42,43]. The errors of trace element concentrations are typically better than 5% for most elements based on repeated analyses of secondary standards.

For rutile, a modified protocol from [44] was used. The analysis was performed using a laser beam energy maintained at 6 J/cm² with a laser repetition rate at a constant frequency of 5 Hz and a spot size of 60 µm. Analyses were conducted over 15 s of background collection, followed by 30 s of ablation and sample acquisition, with a final 15 s for cleaning. Ti measured as ⁴⁹Ti was used as a standard, normalized to a concentration of 59.94% in the NIST SRM 612 standard glass [45] and rutile R10 [46]. Rutile R19 (ID-TIMS: 489.5 ± 0.9 Ma; [46,47]) was used as an unknown to evaluate the data quality. Data were processed using the Iolite 4 software [42,43].

The LA-ICP-MS trace element compositions of sphalerite, pyrite, and chalcopyrite were determined at the EleMap laboratory, GFZ Helmholtz Centre for Geosciences, Potsdam (Germany) using an Analyte Excite 193 nm ArF excimer-based laser ablation (LA) system (Teledyne Photon Machines, Bozeman, MT, USA), coupled to a quadrupole-ICP-MS iCAP RQ from Thermo Fischer Scientific. The analytical and data processing methods and the full trace element dataset are presented in [48].

3.4. Fluid Inclusion Microthermometry

A total of 10 samples of ore-related fluorite and quartz were prepared as doubly polished wafers of approximately 120–200 μm thickness for fluid inclusion studies using optical and cathodoluminescence microscopy. Cathodoluminescence (CL) imaging was applied to reinforce fluid inclusion petrography by depicting different fluorite and quartz generations and growth zoning using a Reliotron^® cold cathode CL system under conditions of 15 kV accelerating voltage and a current density of ca. 10 µA [49]. Cathodoluminescence photomicrographs were taken with a very high sensitivity Qimage Micropublisher 5 Mp digital camera with a Peltier-cooled CCD.

Microthermometry measurements were carried out at the Crustal Fluids Laboratory of the Instituto de Geociencias at UNAM in Querétaro, Mexico, using an Olympus BX-51^® microscope equipped with a Linkam THMSG600 heating-cooling stage coupled to a QImaging Retiga 2000R CCD HD camera. For each fluid inclusion, freezing temperature (T_f), eutectic temperature (T_E), final hydrohalite melting temperature (T_m-HH), freezing final ice melting temperature (T_mice), and total homogenization temperature (T_h) were repeated during several runs for accuracy. Cryogenic experiments were carried out before heating to reduce the risk of decrepitating the inclusions. The volumetric fraction of phases was estimated at room temperature using the volumetric table of [50]. Total salinities expressed as equiv wt% NaCl-CaCl₂ were calculated using T_m-HH and T_m-ice temperatures according to the spreadsheet provided by [50] in the ternary NaCl-CaCl₂-H₂O system. In some cases, T_m-HH could not be observed. In this case, the corresponding salinities were calculated based on the HOKIEFLINCS_H₂O-NaCl spreadsheet of [50] in the H₂O-NaCl system and reported as total wt% NaCl eq. No hydrocarbon-bearing inclusions were detected with Raman spectroscopy.

4. Results

4.1. Vein Ore Shoot Distribution, Geometry, Ore Textures, Alteration, and Paragenesis

The Aouli Pb-Zn ± fluorite ± barite mining district extends across an exposed area of ~12 × 25 km at the approximate coordinates lat. 31°47′45″ N, long. 08°00′48″ W (Figure 1). The exploited fault-fill veins, which are mostly limited to and/or close to the rocks of the granite-induced contact metamorphic aureole, are grouped into five ore zones referred to as Aouli, Sidi Ayad, Poulet, Sidi Said, and Ansegmir (Figure 2). In the five ore zones, the ENE-WSW, WNW-ESE, and E-W trans-tensional steeply dipping veins extend laterally up to 10 km along strike and over 200 m down-dip, are spaced 50 to >100 m apart, dip between 35° and 65° to the east, and vary in thickness from several cm up to 10 m (averaging 4 m). The spatial distribution of the ore zones does not correlate with stratigraphy as the fault-fill veins are either entirely hosted in the granitic intrusion (i.e., Ansegmir group), in the Cambrian–Ordovician schistose pelite (i.e., Poulet, Aouli, and Sidi Said groups) or along the schistose pelite and the granite boundary, locally extending into the unconformably overlying Triassic rocks (i.e., Sidi Ayad group) (Figure 4A). Structural analysis indicates that all these fault-fill veins developed along mostly normal faults under an extension tectonic regime consistent with a NW-SE to NNW-SSE sub-horizontal σ₃ axis and a subvertical σ₁ axis [38].

Individual veins consist of a succession of strongly mineralized shear segments “ore shoots” separated by lower-grade or barren segments. The resulting Ag-(Ni-Co-Sb-As-Hg ± Bi)-bearing ore shoot mineralization occurs as irregularly distributed and lenticular orebodies typically on the same quartz-fluorite fault-fill veins, but is preferentially developed in dilatational jogs, changes in strike, splays, and tensile fracture arrays. The contact between veins and wall rocks is sharp, even though alteration halos with disseminated pyrite and sulfide-quartz veinlets may extend outward from the veins to the schistose pelitic host rocks. Wall-rock alteration consists predominantly of sulfidation and silicification with subordinate sericitization, carbonitization, argillization, and hematitization.

The majority of the Aouli ore shoot mineralization occurs as quartz-fluorite polymict clast-supported breccia veins (Figure 4B) and vug infill with ribbon, colloform, and cockade (Figure 4C) textures. The breccia bodies consist of angular to moderately rounded cm-sized wall-rock clasts set in a sand-sized hydrothermal blackish matrix of hydrothermally precipitated phases. Throughout the polymict breccia bodies, the Ag-(Ni-Co-Sb-As-Hg ± Bi)-rich ore is intergrown with quartz, sericite (Figure 4D), apatite (Figure 4E), rutile (Figure 4F), and adularia (Figure 4G), with the four latter hydrothermal phases being described hereinafter for the first time. The veins display comb, cockade, breccia, and crack-and-seal textures suggesting repeated stages of dilatation, fluid circulation and subsequent ore infilling. Reopening of the same fault-fill vein structures and their subsequent ore infilling during the repeated mineralization stages explain the texturally complex and compositionally variable vein generations. Although no consistent paragenetic sequence can be defined owing to repeated and episodic pulses of mineralization, the sequence of mineral deposition shows three successive sub-stages of mineralization (Figure 5). The early Ni-Co-Fe arsenide sub-stage comprises, from early to late, rammelsbergite (NiAs₂), skutterudite (CoAs₃), safflorite (CoAs₂), and loellingite (FeAs₂). The following sub-stage consists of a succession of Ag-As-Sb sulfarsenides and sulfosalts including fahlore group minerals (tennantite-tetrahedrite series), acanthite, proustite, polybasite, imiterite (Ag₂HgS₂), Ag-Hg amalgams, and native Ag. The late Zn-Pb sulfide sub-stage includes a succession of sequentially crystallized sphalerite, galena, pyrite, and chalcopyrite. The synchronous gangue phases for all three sub-stages comprise, in order of decreasing abundance, quartz, fluorite, and barite along with subordinate apatite, rutile, adularia, and sericite, and to a lesser extent carbonates (mostly dolomite and ankerite).

Strikingly comparable textural features and paragenetic sequence of ore minerals have been reported from several native metal-arsenide ore districts including the Ag-Bi-Co-Ni-As ± U veins of the Erzgebirge (Germany) [1,7,51,52], the Ni-Co-As mineralization in the Midwest polymetallic U deposit of the Athabasca Basin (Canada) [53], and those world-class deposits from the Bou Azzer Co-Ni-Fe-As(±Au ± Ag) and Imiter Ag-Hg-rich districts in the Moroccan Anti-Atlas [54,55,56]. Contrary to many of the so-called five-element vein systems, especially those in the Erzgebirge, which are commonly associated with substantial amounts of U mineralization, no evidence has been found so far either for the occurrence of U-bearing veins or U-disseminated mineralization, or for carbonate gangue phases in the Aouli granitic intrusions, or the schistose pelites.

In addition to the hydrothermal ores, a supergene stage is well developed in the uppermost sulfide-rich part of the exploited orebodies. The resulting secondary mineral assemblage consists of a variety of exotic phases such Pb- and Cu-carbonates (cerussite and malachite-azurite, respectively) and hydrated arsenate (erythrite: Co₃(AsO₄)₂•8(H₂O)).

4.2. Vein Ore Shoot Mineralogy and Mineral Chemistry

The Aouli deposit have a complex ore mineralogy which consists of (1) common metal sulfides (sphalerite, galena, pyrite ± chalcopyrite), (2) uncommon metal phases including tri- and diarseniades (skutterudite/Nickelskutterudite, safflorite/clinosafflorite, rammelsbergite, and loellingite), (3) Pb-Sb-As-Ag-sulfosalts (Ag-bearing tetrahedrite, polybasite, acanthite, proustite, and stannite), and (4) Ag-Hg amalgam and native silver (Figure 6). Volumetrically, common Zn-Pb sulfides are by far the dominant phases, whereas arsenides, Pb-Sb-As-Ag-sulfosalts, and Ag-Hg amalgam and native silver are less abundant (<5 vol.%), being spatially restricted to the upper parts of the exploited veins close to the unconformity between the Triassic basal red bed series and the underlying Paleozoic basement lithologies.

4.2.1. Arsenides

Among the Ni-Co arsenides, skutterudite/Nickelskutterudite (Figure 6A) is by far the dominant triarsenide phase. It typically occurs either as clusters of euhedral to subhedral crystals or as space fillings between quartz and fluorite ± ankerite-dolomite. Skutterudite is commonly intergrown with other members of the Ni-Co-Fe-bearing arsenide series including rammelsbergite and safflorite (Figure 6B). Locally, skutterudite is replaced by native silver, acanthite, proustite, stannite, and chalcopyrite along with supergene secondary phases such as arsenolite and erythrite (Figure 6B). The EDS analyses show large variations in As, Ni, and Co contents ranging from 70.02% to 74.40%, 6.01% to 19.43%, and 6.50% to 14.46%, respectively. Nonetheless, Fe and S are present at lower concentrations reaching up to 2.85% Fe and 3.06% S.

Diarsenide minerals commonly occur as thin continuous layers of safflorite/clinosafflorite, rammelsbergite, and loellingite displaying a complex corona-like overgrowth or star-shaped texture (Figure 6C). Texturally, diarsenide minerals occur either as (i) euhedral crusts or star-shaped aggregates, ranging from 10 μm to 500 μm in thickness, growing directly on skutterudite crystals, (ii) massive aggregates (up to 2 mm) that partially replace euhedral to subhedral skutterudite, and/or (iii) crystalline aggregates displaying an arborescent growth habit or a typical fortress texture materialized by concentric banding/zoning patterns with Ni/Fe ratios decreasing from the inner zones outward. Compositionally, the diarsenide phases exhibit overlapping chemical compositions with significant substitutions among Ni, Fe, and Co (Table S1; Figure 7A,B). In this respect, EPMA analyses show that rammelsbergite has As, Ni, Co, and Fe contents ranging from 65.06 to 69.19 wt%, 10.49 to 18.59 wt%, 4.29 to 10.69 wt%, and from 6.21 to 9.98 wt%, respectively. Safflorite has As ranging from 62.83 to 70.13 wt%, Co between 9.27 and 17.17 wt%, Ni from 0 to 9.65 wt%, and Fe from 5.01 to 13.85 wt%. Similarly, loellingite has As concentrations varying from 67.75 to 70.95 wt%, Ni from 0.19 to 8.78 wt%, Fe from 11.62 to 16.20 wt%, and Co between 5.76 and 13.30 wt%. Cu contents are up to 1.32 wt%. It is noteworthy that all analyzed diarsenides contain appreciable sulfur concentrations that range widely from 0.47 wt% up to 5.77 wt.% S.

In addition to the above-mentioned diarsenide phases, an unidentified (Ni-Co) AsS amalgam is also recognized. It occurs either as thin layers, up to 50 μm, overgrowing pyrite grains or as minute grains associated with, and/or replacing, rammelsbergite, safflorite/clinosafflorite, loellingite, and skutterudite/Nickelskutterudite. The EDS analyses indicate As, Ni, Co, Fe, and S concentrations of 46.74%, 8.86%, Co 24.99%, Fe 2.26%, and S 17.15%, respectively.

4.2.2. Pb-Sb-As-Ag-Sulfosalts

The Pb-Sb–As–Ag-sulfosalts are irregularly distributed in major veins and include Ag-bearing tetrahedrite, polybasite, proustite, imiterite, Ag-Hg amalgam, acanthite, and native silver.

Polybasite is by far the most abundant silver-bearing phase. It occurs as variably sized (up to 200 μm) anhedral crystals coexisting with Ag-bearing tetrahedrite, proustite, acanthite, sphalerite, and galena (Figure 6D). Polybasite also forms massive aggregates, corroding and cementing quartz grains (Figure 6E) or infilling small fractures. The EPMA analyses reveal that polybasite has Ag contents ranging from 58.24 to 73.96 wt%, Cu from 0.29 to 8.31 wt%, Sb from 1.98 to 12.54 wt%, and As concentrations in the range of 0.53 to 5.44 wt% (Table S1). The striking compositional feature of polybasite resides in its enrichment in Hg with contents ranging from 1.31 to 5.18 wt%. Ag-rich tetrahedrite occurs either as isolated anhedral grains, up to 200 µm in size, or as fracture-infill veinlets with coexisting chalcopyrite, polybasite, and acanthite; all set in a gangue of quartz and fluorite (Figure 6E). Geochemically, argentiferous tetrahedrite shows varying Ag (1.76–10.13 wt%), Sb (16.71–21.55 wt%), Cu (30.35–36.58 wt%), and As (4.82–8.53 wt%) contents; those of Fe and Zn range from 0.77 to 3.57 wt% and 4,00 to 6.59 wt%, respectively. Minor components include Hg and Pb whose concentrations range from 2.18 to 2.75 wt% Hg and 0.01 to 0.27 wt% Pb (Table S1).

Acanthite is closely intergrown with polybasite, imiterite, and Ag-Hg amalgam (Figure 6E). The EPMA analyses show highly variable Ag and S contents, ranging from 69.60 to 83.74 wt% and 8.62 to 15.26 wt%, respectively. The Hg concentrations range from 1.11 to 3.36 wt% (Table S1). Proustite forms anhedral to euhedral crystals ranging from 100 to 600 µm. It fills intergranular spaces between quartz and fluorite. Locally, proustite is associated with galena and sphalerite typically as coarse masses. The EPMA data reveal Ag contents between 63.23 to 65.73 wt%, As from 11.36 to 12.22 wt%, and Sb from 3.04 to 3.29 wt% (Table S1).

Ag-Hg amalgam is among the most abundant Ag-bearing phases (Figure 6E). It occurs as thin rims mantling and coexisting with Ag-bearing tetrahedrite, polybasite, acanthite, and native silver. The EPMA analyses show highly variable compositions with Ag and Hg contents ranging from 61.71 to 80.64 wt% and up to 33.23 wt%, respectively. Imiterite is a rare mineral, commonly associated with acanthite and Ag-Hg amalgam. It frequently occurs either as minute inclusions within polybasite or as intergrowths with other sulfosalts, replacing polybasite and acanthite. A single EPMA analysis reveals Ag, S, and Hg contents of 56.95 wt%, 8.59 wt%, and 32.37 wt%, respectively. Native silver occurs either as flake-shaped crystals up to several cm infilling the interstices between fluorite and quartz (Figure 6F), or as filamentous silver (i.e., wire silver) lining vugs (Figure 6G). The EPMA analyses indicate the composition of wire silver close to pure silver with Ag concentrations ranging from 96.3 to 97.6 wt% and sulfur content up to 1.1 wt%. Nonetheless, the Hg contents are low ranging from 0.22 to 0.34 wt% (Table S1).

Remarkably, the Aouli arsenide, Pb-Sb-As-Ag-sulfosalt, Ag-Hg amalgam, and native silver phases are characterized by the high percentages of Hg concentrations.

4.2.3. Sulfides

Sulfide phases consist predominantly of sphalerite and galena with subordinate pyrite, and to a lesser extent chalcopyrite. The full LA-ICP-MS trace element dataset on sulfides is available in [48].

Sphalerite typically occurs as µm to cm sized greenish- to brownish-colored scattered subhedral grains set in a gangue of quartz and fluorite ± dolomite-ankerite (Figure 6H). Texturally and compositionally, the sphalerite is intimately intergrown with Ag-bearing sulfides and sulfosalts. The EPMA analyses (Table S1) along with LA-ICP-MS (

Table 1. Summary LA-ICP-MS trace element compositions of sphalerite, pyrite, and chalcopyrite from the Aouli Ag-(Ni-Co-Sb-As-Hg ± Bi) vein ore shoot mineralization.

Sphalerite
Sample ID.	n		Mn	Fe	Co	Cu	Ga	Ge	As	Ag	Cd	In	Sn	Sb	Hg	Pb
S47	22	Min	0.13	7.46	<0.01	165.85	0.21	<0.06	24.89	66.86	987.34	0.00	0.02	90.62	9.98	32.80
		Max	4.39	3486.00	22.96	7934.87	13.62	90.24	976.46	8322.06	8749.17	16.05	2.09	8084.06	3683.78	2680.84
		Mean	1.57	1489.74	-	3048.48	2.67	-	393.26	2009.41	3227.92	1.59	0.30	2895.02	250.71	974.94
		Std. Dev.	1.55	1136.90	-	2503.15	3.46	-	326.74	1948.30	2329.35	3.61	0.43	2591.99	771.53	803.73
S38.1	57	Min	0.18	65.95	0.02	93.08	0.60	<0.03	19.28	183.32	91.54	<0.001	<0.01	75.36	23.88	584.60
		Max	3480.34	2452.99	7.94	2054.43	43.22	43.35	1791.00	43,790.03	1495.93	0.86	57.93	4078.82	1311.91	12,988.14
		Mean	126.72	1024.70	2.76	953.68	9.39	-	201.61	4129.06	718.03	-	-	1615.84	406.88	3239.50
		Std. Dev.	507.72	695.01	1.80	399.15	10.21	-	247.04	6709.34	287.01	-	-	975.99	332.22	2543.81
Pyrite
			Co	Ni	Cu	Zn	As	Se	Ag	Cd	Sn	Sb	Hg	Tl	Pb	Bi
S38.1	24	Min	110.77	88.39	390.92	1.79	2318.10	<0.16	5297.30	0.11	<0.01	65.56	75.85	8.96	932.11	<0.001
		Max	204.05	200.81	954.22	62.06	9769.39	0.55	19,954.92	0.83	0.18	184.29	171.48	91.97	6251.46	0.01
		Mean	144.23	136.70	703.78	13.63	6305.85	-	11,661.36	0.35	-	139.54	108.12	21.69	1558.19	-
		Std. Dev.	22.56	30.80	121.91	13.70	2461.34	-	4608.32	0.17	-	29.92	27.39	23.64	1149.65	-
Chalcopyrite
			Co	Ni	Zn	As	Se	Ag	Cd	Sn	Sb	Te	Hg	Tl	Pb	Bi
S9	24	Min	<0.02	<0.07	9.48	25.00	<0.52	16.95	0.24	2.94	40.81	<0.07	4.16	0.86	41.98	<0.01
		Max	0.31	0.31	34.30	158.92	1.45	92.92	1.44	31.39	190.44	0.19	117.95	6.79	246.28	0.14
		Mean	-	-	17.54	74.24	-	43.77	0.73	11.69	96.13	-	23.97	3.16	140.64	-
		Std. Dev.	-	-	5.87	37.82	-	15.96	0.29	6.80	37.04	-	22.61	1.41	56.17	-

Abbreviation: n = number of analyses; Min = minimum; Max = maximum; Std. Dev. = Standard Deviation.

) data reveal that sphalerite contains high concentrations of Ag, Hg, Sb, As, Cd, Cu, and Pb reaching up 4.3 wt% Ag, 0.36 wt% Hg, 0.8 wt% Sb, 0.17 wt% As, 0.8 wt% Cd, 0.79 wt% Cu, and 1.2 wt% Pb (Figure 8). Nonetheless, the concentrations of other elements, particularly Ga, Ge, In, and Fe, are commonly moderate to low with average value 43.22 ppm, 90.24 ppm, 16.05 ppm, and 3486 ppm, respectively.

Galena is by far the most abundant base metal sulfide in the studied deposits. Three textural varieties of galena are recognized, including (i) disseminations of varying grain size, (ii) crystalline, massive anhedral galena infills, and (iii) cm sized cubic crystals. Similar to sphalerite, galena from sub-stage II.3 is closely intergrown with sphalerite, pyrite, and chalcopyrite within a gangue of quartz and fluorite. Galena commonly hosts numerous minute inclusions of Ag-Hg amalgam and polybasite. The EPMA analyses reveal significant enrichment in Ag and Hg with contents of 4.11 wt% Ag and up to 1.53 wt% Hg (Table S1).

Pyrite is a minor sulfide. It occurs as an early fine-grained phase, forming subhedral to euhedral crystals up to 200 μm in size or larger subhedral aggregates up to 600 μm. Pyrite exhibits a distinct cataclastic texture, with the resulted brecciated crystals being cemented by tetrahedrite and Ag-bearing minerals, and locally cross-cut by thin veinlets of galena, sphalerite, and chalcopyrite. The LA-ICP-MS analyses show that pyrite has significant substitutions of Co (110.7–204.0 ppm), Ni (88.3–200.8 ppm), Cu (390.9–954.2 ppm), Sb (65.5–184.2 ppm), and Hg (75.8–171.4 ppm) whose concentrations contrast with those of Se, Te, Sn, Cd, and Bi which are low, commonly below detection limit. The Ni/Co ratio ranges from 0.74 to 1.52. Remarkably, As and Ag contents are strikingly high, reaching 0.63 wt% and 1.16 wt%, respectively (Table 1).

Chalcopyrite occurs either as massive mono- or polymineral aggregates commonly intergrown with quartz or as anhedral masses filling microfractures. The LA-ICP-MS analyses indicate a homogenous composition close to the ideal formula CuFeS₂ (Table 1).

4.2.4. Ore-Related Gangue Minerals

In addition to quartz and fluorite, which have been the focus of numerous studies [14,16], hydrothermal apatite, rutile, adularia, and sericite are described therein for the first time.

Apatite

Apatite occurs as homogeneously distributed fine-grained rounded, rarely elongate, to subhedral (10–500 μm across) crystals or aggregated grains closely intergrown with Ag-(Ni-Co-Sb-As-Hg ± Bi)-rich arsenides, sulfarsenides/sulfosalts, and sulfides along with sericite, adularia, and rutile. BSE images show that apatite grains have a homogenous appearance, contain locally quartz and galena inclusions, and fluoresce yellow light under CL (Figure 4E).

Geochemically, in situ LA-ICP-MS compositions of single apatite crystals show significant variations in trace elements, such as Mn (262–618 ppm), Sr (367–899 ppm), Pb (1.84–873 ppm), U (0.62–3.55 ppm), and Th (1.04–4.26 ppm). The Cl content is 215.98 ± 94.34 ppm (

Table 2. Summary of LA-ICP-MS trace element compositions of ore-related hydrothermal apatite from the Aouli Ag-(Ni-Co-Sb-As-Hg ± Bi) vein ore shoot mineralization.

Mineral	Apatite (n = 53)
	Min	Max	Avg.	Std. Dev.
Cl	39.96	656.58	215.98	94.34
Mn	262.96	618.09	417.60	94.46
Sr	367.91	899.79	525.56	122.93
Y	930.05	1605.61	1229.51	158.76
La	89.05	259.22	166.80	37.30
Ce	320.83	820.37	560.43	125.10
Pr	62.74	149.37	103.47	21.04
Nd	416.86	874.28	624.34	100.41
Sm	145.48	276.18	203.07	25.02
Eu	37.09	64.79	50.05	5.32
Gd	191.97	338.40	262.19	25.17
Tb	29.00	50.17	39.70	4.16
Dy	177.24	298.96	237.01	27.73
Ho	34.66	60.06	46.06	5.82
Er	88.75	152.69	115.42	15.52
Tm	9.96	17.28	13.24	1.82
Yb	55.39	94.81	74.19	10.17
Lu	7.25	12.39	9.52	1.30
∑REE	1718.49	3440.16	2505.49	339.80
∑REY	2648.54	4908.39	3735.00	404.05
∑LREE	928.04	2103.23	1455.04	279.24
∑MREE	615.43	1079.51	838.09	81.25
∑HREE	161.57	277.05	212.36	28.69
Pb	1.84	873.97	53.60	141.43
Th	1.04	4.26	1.73	0.50
U	0.62	3.55	1.08	0.54
Y/Ho	25.63	27.77	26.69	0.51
Sr/Y	0.28	0.94	0.44	0.14
Ce/Ce*Chondrite	0.92	1.05	1.00	0.03
Eu/Eu*Chondrite	0.58	0.71	0.66	0.03
Y/Y*Chondrite	0.83	0.91	0.87	0.02

Abbreviation: n = number of analyses; Min = minimum; Max = maximum; Avg. = Average; Std. Dev. = Standard Deviation.

). With respect to REE + Y contents, apatite grains are strongly enriched in REEs (ΣREE = 1718.5–3440.2 ppm), with average total light (La–Nd), middle (Sm–Ho), and heavy (Er–Lu) REE contents of 1455.1 ± 279.2 ppm, 838 ± 81.3 ppm, and 212.4 ± 28.7 ppm, respectively (Table 2). The chondrite-normalized REE + Y spectra show “hump”-shaped patterns (Figure 9A) with enrichment in MREE relative to LREE and heavy rare earth elements (HREE) in addition to exhibiting a prominent negative Eu anomaly with Eu/Eu* ratios of 0.6 to 0.7, and almost no Y and Ce anomalies (Ce/Ce* = 0.92–1.05). Key REE fractionation ratios are (La/Sm)_N = 0.3–0.5, (La/Yb)_N = 0.6–2.2, (Sm/Yb)_N = 2–3.7, and (Gd/Lu)_N = 2.6–4, where the subscript N indicates that the REE contents are normalized to chondrite [57]. The Y/Ho ratios are generally between 26 and 28. No consistent correlation between trace element compositions and CL colors has been noticed.

In summary, the Aouli apatite is characterized by high REE + Y (mean = 3735 ppm) and Pb contents (average = 54 ppm), but lower concentrations of U (average = 3.55 ppm). Regarding abundances of volatiles (i.e., Cl-F-OH), only Cl has been detected with concentrations showing large variations ranging from 39.96 ppm to 656 ppm (Table 2).

Rutile

As for apatite, rutile occurs as a subordinate but common phase closely intergrown with (Ni-Co)-As-S diarsenide, and tetrahedrite (Figure 4F). Under transmitted light microscopy, rutile is dark reddish-brown to opaque. It occurs as dispersed to clustered subhedral to euhedral µm-sized (10–400 μm) untwinned grains coexisting with quartz and fluorite. In BSE, the rutile grains are uniformly gray andhave no obvious chemical zoning (Figure 4F). Inclusions of galena were locally observed.

Geochemically, rutile grains are enriched in Fe (2029–75,592 ppm), V (181.32–1360.42 ppm), Zr (13–4600 ppm), Nb (207–514 ppm), W (821–10,247 ppm), and Sn (104–654 ppm), but depleted in Cr (<2–6.96 ppm), Ta (4–97 ppm), and Hf (3.79–93 ppm) (

Table 3. Summary of LA-ICP-MS trace elements compositions of ore-related hydrothermal rutile from the the Aouli Ag-(Ni-Co-Sb-As-Hg ± Bi) vein ore shoot mineralization.

Mineral	Rutile (n = 17)
	Min	Max	Avg	Std. Dev.
Sc	64.60	240.15	144.89	53.92
V	181.32	1360.42	552.31	316.47
Cr	<2	6.96	-	-
Fe	2029.75	75,592.55	9227.63	17,586.14
Sr	10.57	1236.32	167.73	301.95
Y	5.43	314.28	72.99	83.44
Zr	13.03	4600.56	680.22	1084.11
Nb	207.44	514.80	343.05	91.97
Mo	2.51	663.64	73.10	171.76
Sn	104.25	654.02	354.39	163.87
Sb	5.77	1091.10	169.62	315.94
Ba	21.65	586.37	169.85	171.71
La	0.72	154.80	29.26	43.97
Ce	2.44	426.56	79.43	120.64
Pr	0.36	54.79	10.07	15.41
Nd	1.72	211.69	39.98	59.91
Sm	0.60	44.44	9.21	12.71
Eu	0.21	22.76	4.42	6.21
Gd	0.94	43.92	9.72	12.29
Tb	0.19	9.56	2.06	2.61
Dy	1.20	83.63	17.36	22.54
Ho	0.23	22.32	4.46	6.03
Er	0.78	89.20	17.68	23.67
Tm	0.12	14.64	2.92	3.91
Yb	0.87	109.85	22.47	29.31
Lu	0.12	15.85	3.40	4.30
∑REE	10.90	1303.99	252.46	361.57
∑REY	16.33	1618.28	325.45	443.51
∑LREE	5.24	847.84	158.75	239.85
∑MREE	3.75	226.62	47.24	62.27
∑HREE	1.90	229.53	46.47	60.99
Hf	3.79	93.52	25.83	20.77
Ta	4.02	97.41	26.90	21.75
W	821.37	10,247.18	4664.37	2414.41
Pb	101.12	42,974.23	7694.61	14,082.39
Th	<0.001	3.27	-	-
U	0.30	199.18	31.80	53.66
Nb/Ta	4.41	90.36	20.77	19.58
Zr/Hf	0.88	49.19	19.95	12.40
Y/Nb	0.02	1.06	0.23	0.28
Fe/V	4.09	113.17	15.55	25.49
Y/Ho	14.08	30.37	19.86	5.13
Nb/V	0.20	1.70	0.83	0.50
Ce/Ce*Chondrite	1.06	1.16	1.12	0.03
Eu/Eu*Chondrite	0.84	2.01	1.32	0.26
Y/Y*Chondrite	0.53	1.32	0.72	0.22

Abbreviation: n = number of analyses; Min = minimum; Max = maximum; Avg. = Average; Std. Dev. = Standard Deviation.

; Figure 9B). There are large variations in the concentration of U (0.3–199.1 ppm) and consistently very low Th contents (0.0–3.27 ppm). As a result, the Th/U ratio is low (mean value = 0.01). Conversely, the Pb contents are remarkably high ranging from 101 to 42,974 ppm. Key ratios such as Nb/Ta, Zr/Hf, Y/Nb, Fe/V, and Y/Ho range from 4.4 to 90.4, 0.9 to 49.1, 0.0 to 1.1, 4.1 to 113.2, and 14 to 30.4, respectively.

Regarding REE + Y elements, the rutile samples are characterized by large variations in total REE + Y concentrations ranging from 16 to 1618 ppm, display consistently chondrite-normalized flat LREE-MREE patterns (Figure 9C), and lack distinct Ce, Eu, or Y anomalies with Ce/Ce*, Eu/Eu*, and Y/Y* values of 1.06–1.16, 0.84–2.01, and 0.50–1.32, respectively (Table 3). Comparatively, these patterns are typical of rutile of hydrothermal origin [58,59,60,61].

Adularia

Texturally, adularia occurs within breccia bodies as sparsely and homogeneously distributed very fine-grained (<20 µm across) euhedral rhombic and sub-rhombic crystals or grouped in clusters of aggregated tabular-shaped crystals intimately intergrown with sulfide mineralization and recrystallized fine-grained quartz (Figure 4G). Under cross-polarized light microscopy, adularia is pale to dark gray in color, commonly exhibiting complex twinning and optic variations. Locally, adularia consistently displays a mottled texture due to the presence of numerous minute (<15 µm across) inclusions of sericite. In BSE images, the adularia crystals are uniformly gray (Figure 4G) andhave no obvious chemical zoning or heterogeneous textures. SEM and BSE images show uniformly gray platy to polyhedral adularia morphologies with longest dimension lower than 20 µm across and have no obvious chemical zoning or heterogeneous textures. Chemically, the SEM-EDS analyses indicate the composition of adularia close to pure KAlSi₃O₈.

Sericite

Hydrothermal sericite occurs in relatively minor amounts as an aggregate of fine- to medium-grained flakes (up to 100 μm in length) closely intergrown with sulfides (sphalerite, pyrite, and galena) along with adularia, apatite, and quartz (Figure 4D).

4.3. Fluid Inclusion Analysis

4.3.1. Petrography

Fluid inclusions in ore-related quartz and fluorite were investigated using fluid inclusion petrography and microthermometry analysis. Moreover, the fine-grained nature of most quartz samples and/or the general lack of larger fluid inclusions limited our analyses to a restricted number of quartz-hosted fluid inclusions compared to those hosted in fluorite which are more abundant and larger. Nonetheless, both ore-related phases contain visible growth zones along which several fluid inclusion assemblages (FIAs; [62,63]) occur (Figure 10). The investigated fluid inclusion assemblages were classified as isolated, cluster, primary, pseudosecondary, and secondary according to [64,65].

Three types of fluid inclusions were identified in the ore-related fluorite and quartz based on their phase assemblages at room temperature and homogenization behaviors during heating, including the following: (i) liquid monophase (L), (ii) liquid-dominated bi-phase (vapor + liquid, L + V), and (iii) sporadically liquid-dominated tri-phase (L + V + S) inclusions in which the liquid and vapor homogenize to the liquid phase, but the solid phase(s) did not dissolve during heating (Figure 10D). Accordingly, these solid crystals were interpreted as entrapped phases and not daughter minerals, and as such were excluded from microthermometric measurements.

Overall, the two-phase L-V inclusions are by far the dominant type, with the degree of filling up to 10 vol.% of vapor phases. They are abundant in fluorite and to a lesser extent in quartz, occurring as irregular, subrounded to elongated-shaped inclusions mostly 5 to 40 μm in size (Figure 10). Larger inclusions up to 120 μm were observed in host fluorite. Examination under ultraviolet micro-spectroscopy did not identify hydrocarbon-bearing fluid inclusions as single-phase inclusions or as heterogeneously trapped aqueous-petroleum inclusions. Moreover, neither liquid CO₂ nor clathrates melting was observed, indicating that the CO₂ content in fluid inclusions is less than 1.5 mol% [66]. Remarkably, no evidence of boiling was observed as evidenced by the lack of coexisting liquid- and vapor-rich inclusions [64].

4.3.2. Microthermometry

Microthermometric measurements were performed on the two-phase L-V fluid inclusions considered to be of primary origin according to the criteria of [62]. A total of 198 two-phase L-V inclusions from 36 FIAs and clusters with consistent vapor-to-total ratios (<10%) were investigated for microthermometry. The results and related parameters are summarized in

Table 4. Summary of microthermometric data of fluid inclusions hosted in ore-related quartz and fluorite from the Aouli Ag-(Ni-Co-Sb-As-Hg ± Bi) vein ore shoot mineralization.

Mineral	Microthermometry															Calculated Salinity
	Th (°C)					Tice-melting (°C)					Tm(HH) (°C)					Salinity					NaCl/(NaCl + CaCl2)
	n	Min	Max	Avg.	Std. Dev.	n	Min	Max	Avg.	Std. Dev.	n	Min	Max	Avg.	Std. Dev.	n	Min	Max	Mean	Std. Dev.	n	Min	Max	Avg.	Std. Dev.
Ore-related quartz	27	48.7	168.1	128.5	29.8	32	−20.6	−10.5	−14.8	2.0	22	−26.2	−21.7	−23.8	1.3	32	14.5	22.8	18.4	1.6	22	0.6	1.0	0.8	0.1
Ore-related fluorite	174	90.9	141.1	122.1	8.4	174	−22.0	−8.2	−16.0	3.2	151	−32.1	−18.6	−23.6	1.8	174	12.0	23.7	19.3	2.6	146	0.4	1.0	0.8	0.1

Abbreviation: n = number of analyses; Min = minimum; Max = maximum; Avg. = Average; Std. Dev. = Standard Deviation.

 https://www.sciencedirect.com/science/article/pii/S0169136824005717 (accessed on 14 June 2025).

All the two-phase L + V inclusions homogenized into the liquid phase. Collectively, microthermometric data from fluid inclusions in ore-related fluorite and quartz are presented together, as both host minerals yield inclusions with roughly similar microthermometric properties. Excluding outliers, the majority of the FIAs and clusters exhibit consistent homogenization temperatures (T_h) that differ by less than 30 °C. Primary and pseudosecondary inclusions homogenized at temperatures ranging from 80 to 168 °C with 85% of the measurements between 110 and 150 °C. Fluid inclusions in quartz have on average slightly higher temperatures (140–170 °C) than those measured in fluorite (120–140 °C) (Figure 11A).

The first apparent melting temperature of ice (T_E) occurred mostly close to or lower than −50 °C, consistent with the eutectic of the H₂O-NaCl-CaCl₂ system [50]. The final melting temperature of ice (T_m(ice)) was in the range of −20.6 to −10.5 °C (mean = −14.8 °C), and that of hydrohalite between −26.2 to −21.7 °C (mean = −23.7 °C). During heating, hydrohalite exclusively melts before ice in all fluid inclusions. Based on the melting temperatures of ice and hydrohalite, the compositions of the fluid inclusions, including salinity (wt% NaCl + CaCl₂) and (NaCl/(NaCl + CaCl₂) mass ratio, were calculated according to [50].

Remarkably, the compositions of the fluorite- and quartz-hosted inclusions plot in the field of ice along two distinct horizontal arrays above the ice-hydrohalite and hydrohalite-halite cotectic boundaries, on both CaCl₂-dominated and NaCl-dominated sides (Figure 11B). The total salinities are mostly between 14 and 23 wt% NaCl + CaCl₂, though lower salinities as low as 12 wt% NaCl + CaCl₂ are observed in fluorite-hosted fluid inclusions from the late base metal sulfide sub-stage NaCl + CaCl₂ (Table 4).

5. Discussion

It is worth mentioning that the focus of this contribution is on the Ag-(Ni-Co-Sb-As-Hg ± Bi)-bearing vein ore shoot mineralization.

5.1. Physicochemical Evolution, P-T Conditions, and Sources of the Ore Fluids

Microthermometric measurements in ore-related fluorite- and quartz-hosted fluid inclusions indicate that ore-forming fluids correspond to evolved H₂O-NaCl-CaCl₂ basin-derived brines with temperatures of homogenization (T_h) ranging from 80 to 168 °C and salinities between 12 and 23 wt% NaCl + CaCl₂. These temperature and salinity ranges are similar to those documented for native metal-arsenide vein occurrences in Western Europe (e.g., Schwarzwald and Odenwald in Germany [1,67,68]), Morocco (e.g., Bou Azzer [54,55,69,70,71,72]), and elsewhere (e.g., Cobalt-Gowganda and Great Bear Lake/Canada [73,74,75]).

The isochore calculation and P-T estimation suggest a probable range of fluid temperature and pressure from 129 to 205 °C and 433 to 690 bars for a geothermal gradient of 30 °C/km, or from 117 to 193 °C and 308 to 480 bars for a geothermal gradient of 40 °C/km (Figure 12), assuming a surface temperature of 20 °C and a hydrostatic fluid pressure regime. A hydrostatic fluid pressure regime is considered most likely to be constrained by regional sedimentary overburden and the occurrence of breccia and ribbon and cockade textures (Figure 4C), which together indicate shallow crustal depths (i.e., 1–5 km; [16,76,77]). Assuming a depth of 3 km and a hydrostatic fluid pressure regime, the maximum fluid pressure is approximately 300 bars; therefore, a geothermal gradient around 40 °C/km is plausible, consistent with the modern thermal gradient of 30–45 °C/km that characterizes the northern part of Morocco [78]. Given these shallow depths of ore formation, the large fluctuations in calculated fluid pressures suggest that Aouli vein ore shoot mineralization formed from episodic stages of fracturing and subsequent fluid migration. Seismic pumping [79] is proposed as the likely mechanism that drove the released ore brines into the newly created and reopened veins, contributing to the vein ore shoot mineralization in the Aouli deposit. The regional high permeability faults such as those of Aouli, Georges, Ansegmir, and Amourou (Figure 3) would have served as the major pathways that focused fluid flow allowing basement-residing brines to ascend into the cover rocks.

Moreover, the T_h versus salinity relationships (Figure 13A) along with the ternary H₂O-NaCl-CaCl₂ plot (Figure 11B) show two main saline fluid inclusion populations with total salinities in the range of ~14 to ~20 wt% NaCl + CaCl₂, and between ~20 to ~24 wt% NaCl + CaCl₂, respectively. In this regard, the continuous linear correlations shown on temperature vs. NaCl/(NaCl + CaCl₂) (Figure 13B) and CaCl₂ vs. NaCl bi-plot diagrams (Figure 13C) are interpreted as binary isothermal mixing trends between two compositionally different endmembers including a CaCl₂-dominated and a NaCl-dominated brine. This compositional dichotomy indicates that the two ore-forming brines had distinct reservoirs, flow paths, and fluid–rock interactions. While the CaCl₂-dominated brine is interpreted to have been derived from the interaction between the NaCl-rich brine and Ca-rich protoliths in the basement, the Na-enriched brine more likely acquired its high salinity in a subsurface setting (i.e., basin) through extensive interaction with salt-bearing rocks. In the study area, the outcropping country rocks (Figure 1) show the occurrence of halite- to gypsum-bearing evaporites in the Triassic red bed succession [82], and as such, this latter aquifer could have acted as the sedimentary endmember. The Ca-enrichment of the CaCl₂-dominant ore-forming brine could have been acquired by Na/Ca and Na/K exchanges by albitization of plagioclase and K-feldspar, respectively, and Mg-loss by Mg-rich alteration of Ca-rich basaltic rocks.

In summary, the combined fluid inclusion microthermometry, isochore calculation, and P-T estimate indicate the involvement of two main fluid endmembers, including NaCl-rich and CaCl₂-rich brines.

5.2. Origin(s) and Fluid Evolution from the Trace Element Compositions of Syn-Ore Apatite and Rutile

In addition to being utilized for dating [47,83,84], the combined use of high-resolution in situ LA-ICP-MS trace element measurements on hydrothermal apatite and rutile have been shown to be reliable proxies for the physicochemical conditions and the oxidation state of the hydrothermal system during mineralizing events (i.e., temperature, pH, oxidation state, oxygen fugacity) [85,86,87,88,89]. The trace element variations shown by Aouli syn-ore apatite and rutile can be attributed to (i) the composition of the primary TiO₂-bearing mineral phases [60,90,91], (ii) origin and composition of ore-forming fluids (Cl- or F-rich), (iii) physicochemical conditions during mineral precipitation (i.e., temperature), and (iv) oxygen (f_O2) fugacity [87,88,89,92].

The middle rare earth elements’ (MREE) enrichment of apatite, except for Eu which shows a prominent negative Eu anomaly (Figure 9A), is interpreted to be related to dissolution/recrystallization processes driven by protracted fluid–rock interactions with the country rocks including igneous and metamorphic mafic and felsic protoliths along with organic-rich siliciclastic rocks. Moreover, the negative Eu anomaly shown by syn-ore apatite is interpreted as resulting from the breakdown of plagioclase, which more likely had occurred under oxidizing fluid conditions (i.e., high oxygen fugacity [93]), consistent with the rutile intragrain compositional variability of U, Th, Fe, V, and Nb (Figure 14). Altogether, the enrichment of syn-ore apatite in MREE and Cl-OH (Table 2) and rutile in Zr, Nb, and Sn-W (Figure 9B) both suggest the involvement of neutral to alkaline halogen-rich (F, Cl) hydrothermal fluids [59,94,95,96]. Additional evidence for the involvement of neutral F- and Cl-rich brines is provided by the widespread occurrence of fluorite and adularia along with the high salinity (14–23 wt.% NaCl + CaCl₂) of the ore-forming fluids as inferred from fluid inclusions microthermometry.

In summary, the large compositional variations shown by syn-ore apatite and rutile suggest that the hydrothermal pulses contained fluctuating levels of trace elements driven by mixing and subsequent fluid–rock interactions.

5.3. Fluid Mixing and Subsequent Fluid-Rock Interactions as Effective Depositional Processes

The T_h versus salinity relationships, and more importantly the linear correlations shown on Figure 12 and Figure 13, reveal variable degrees of isothermal mixing between two main mineralizing brines, including a deep-seated metal-rich CaCl₂-dominated brine and sulfide-bearing (H₂S and HS⁻) NaCl-dominated basinal brine. Most recent genetic models for the Ni-Co-As ore shoots invoke the involvement of redox reactions between upwelling oxidized ore fluids and reducing petroleum-bearing fluids (oil and gas) or organic-rich lithologies [1,6,98,99]. Some investigators have proposed that the metals were carried by hydrocarbons derived from black shales, and it was the oxidation of the metal-rich reducing fluids that triggered the precipitation of the metals [53,100]. Although the direct evidence for the potential involvement of hydrocarbon-bearing fluids in Aouli ore system is lacking (absence of petroleum- and gas-bearing fluid inclusions), the high Hg content of the ores (up to 14 wt%; Table S1) along with the occurrence of organic–organic-rich siliceous shale within the stratigraphic succession support the concept that organically-derived ore-forming fluids would have occurred at the site of ore deposition. By comparison, hydrocarbon-bearing brines [101] were invoked as responsible for deposition of the REE-rich fluorite mineralization in the adjacent El Hammam REE-rich fluorite deposit, whose geological setting, mineralogical attributes, and the age of mineralization are similar to those of Aouli [15,16].

Overall, the geochemical characteristics discussed above are consistent with mixing between three fluid endmembers, including (i) sulfide-bearing (H₂S and HS⁻) Na-rich brine, (ii) basement-derived metal-bearing CaCl₂-rich reducing brine, and (iii) oil/gas-bearing brine. Precipitation of arsenide, sulfosalt, and sulfide phases is thought to have occurred as a result of redox reactions induced by ternary mixing and subsequent fluid–rock interactions of the above-mentioned fluid endmembers.

5.4. Timing of Mineralization

Despite the discovery at Aouli deposit of hydrothermal ore-related phases reputed as representing suitable material for conventional radiometric measurements such as apatite, rutile, sericite, and adularia, our attempts to directly date hydrothermal mineralization failed to provide accurate age constraints. The U-Pb dating of apatite and rutile has been hampered by the low concentrations of U (typically <1 ppm) and moderate Th (mostly between 10 and 15 ppm), which depart from the consistently high common lead contents of both phases. Similarly, adularia and sericite were not amenable for K-Ar and/or Ar-Ar dating as their fine-grain size (<20 µm) is far lower than the ablation spot size (50 μm).

Accordingly, the relative timing of mineralization in the deposit is indirectly constrained by combining geological field observations and textural cross-cutting relationships. The Aouli vein ore shoots cut across the Triassic red-beds (Figure 4A), thereby constraining the age of mineralization to at least the Triassic. More importantly, radiometric ages along with thermochronological data [102,103,104,105,106] indicate that major development of supergene mineralization, which forms by oxidation at the expense of former sulfides, likely occurred as recently as the late Miocene. Based on these relationships, the inferred age of Aouli vein ore shoot mineralization is broadly bracketed between Late Triassic and Late Miocene time. Regionally, this timeframe coincides with the Late Jurassic–Early Cretaceous to Late Miocene time span, which represents a significant metallogenic epoch for the formation of major post-Variscan hydrothermal vein-type and MVT deposits across Europe and North Africa [1,107] and North Africa as well [108].

6. Metallogenic Model and Concluding Remarks

Previous studies at Aouli suggested that the deposit was formed by low-temperature hydrothermal fluids, but there has been no consensus as to the nature of the mineralizing fluids and the deposit classification [15,16,109]. This research reveals that the unusual Ag-(Ni-Co-Sb-As-Hg ± Bi) fluorite-barite ore shoot mineralization formed from episodic stages of fracturing and subsequent fluid migration. The mineralogy, fluid sources, and mechanisms of ore deposition at the origin of this newly discovered mineralization style are constrained therein based on an integrated approach that combines mineralogy, paragenesis, and fluid inclusion studies. The new data sets allowed us to refine preexisting genetic models [15,16], including those proposed based on structural constraints [38].

Paragenetic relationships indicate that the Aouli vein ore shoot mineralization formed as a result of three fluid pulses, including the early Ni-Co-Fe arsenides, the intermediate Pb-Sb-As-Ag sulfarsenide-sulfosalts, and the late Pb-Zn ± Cu sulfides with quartz, fluorite, and barite being precipitated in all three stages. Fluid inclusion petrography and microthermometry combined with trace element geochemistry of ore and related gangue phases along with isotopic constraints reveal that the ore shoot mineralization formed by variable degrees of isothermal mixing and subsequent fluid–rock interaction between two main ore brines, including a deep-seated metal-rich CaCl₂-dominated basement brine and sulfide-bearing (H₂S and HS⁻) NaCl-dominated basinal brine. The involvement of a third methane-dominated fluid–gas fluid which triggered ore deposition is inferred from the enrichment of the Aouli ore in Hg. The development of breccia structures in the ores and the fluid pressure fluctuation revealed by fluid inclusions suggest that the deposit formed from episodic stages of fracturing and subsequent fluid migration. Seismic pumping is proposed as the likely mechanism that drove the released mineralizing brines into the newly created and reopened veins contributing to the vein ore shoot mineralization. Precipitation of ore phases is thought to have occurred as a result of isothermal mixing. Fluid–rock interactions exerted strong control over the leaching process and transport of the released metals by the ore-forming brines from which Aouli ores precipitated.

In the absence of radiometric age constraints, the timing of Aouli mineralization is indirectly bracketed between Late Triassic and Late Miocene time, predating the climax of the Atlasic compressional event in the Late Miocene (late alpine phase [110,111]). From a geodynamic point of view, the time ascribed to the Aouli deposit coincides with major geodynamic events that developed as a causal far-field effect of the drifting of the Central Atlantic Ocean and formation of the Maghrebian Tethys passive margin [23,24,111,112,113,114,115]. These events include major crustal extension and widespread Middle Jurassic–Upper Cretaceous alkaline basaltic magmatism in the Atlas system [116,117,118]. Therefore, we propose a metallogenic model which involves a genetic link among northwestern African plate geodynamic, extensional tectonics, Middle Jurassic–Upper Cretaceous alkaline magmatism, and formation of Aouli deposit. In such a setting, we suggest that thermal convection induced by the generated high heat flow and subsequent increased geothermal gradient triggered development of small-scale convection cells that controlled the large-scale migration of ore-forming brines and formation of the Aouli unconformity-related polymetallic deposit in the Upper Moulouya metallogenic province.