Trace Element Characteristics of Pyrite and Arsenopyrite from the Golden Ridge Gold Deposit, New Brunswick, Canada: Implications for Ore Genesis

Abstract

The Golden Ridge gold deposit is located in southwestern New Brunswick, in the Canadian Appalachians. Gold mineralization is consistently associated with acicular arsenopyrite, and to a lesser degree with pyrite, disseminated in host rocks, sulphide veinlets, quartz-carbonate veins, and the breccia matrix. According to petrographic-based textural differences, four types of pyrite and two types of arsenopyrite are recognized with associated assemblages. Based on SEM-BSE imaging and LA-ICP-MS spot analyses of the different types of pyrites and arsenopyrites, “invisible gold” (solid solution in the crystal lattice of pyrite and arsenopyrite or <100 nm nanoparticles) and micrometer-size inclusions were identified as the main forms of Au. Four syn-gold mineralization pulses of fluid are suggested. The initial hydrothermal fluid, which generated low-grade pyrite (Py-I) enriched in Sb, Pb, Cu, Co, Ni, and Bi, was followed by a second pulse of fluid enriched in arsenic and gold, generating coprecipitated Py-II and Asp-I. The third and fourth pulses were enriched in both arsenic and gold and precipitated Py-III, then coprecipitated Py-IV and Asp-II, which constitute the most important Au depositional episodes. The repeated occurrence of growth zones with Au enrichment in the arsenian pyrites (Py-II, Py-III, and Py-IV) indicate surface growth during metal deposition and disequilibrium crystallization processes.

1. Introduction

The Golden Ridge gold deposit, formerly known as Poplar Mountain, was initially discovered in the 1990s when a series of Au-Ag-As-Hg-Sb-enriched till geochemistry samples collected by the New Brunswick Department of Natural Resources and Energy was further investigated given their anomalous contents [1]. Gold mineralization is principally associated with arsenopyrite and to a lesser degree with pyrite, stibnite, and sphalerite [2]. As no visible gold has been observed, it is thought that gold is contained in arsenopyrite as an ionic substitution or as submicron-sized inclusions [3]. This submicroscopic or “invisible” gold has been identified in a broad range of different geological environments [4,5,6,7,8,9,10,11], and its presence is regarded as economically negative as it impacts the recovery of Au from a mineral processing perspective [12,13,14,15,16,17,18,19]. The classification of the Golden Ridge mineralization is still debated. It has been regarded as a low sulphidation epithermal deposit, based on the presence of quartz-cemented breccia, locally cockscomb textures, with a Au-As-Sb association and the low contents of Cu [20]. Nonetheless, it was later argued that considering the proximity of the Early Devonian Pokiok Batholith and available geochronological data indicating mineralization was coeval with the age of the batholith, the ore-forming fluid had originated from a granitic intrusion. Moreover, due to the presence of a deep-seated granitic intrusion below the Middle Ordovician Poplar Mountain volcanic complex indicated by regional magnetic data and the CO₂-rich composition of the ore-forming fluid, mineralization was considered to be mesothermal granite intrusion-related [3,21].

At Golden Ridge, both arsenopyrite and pyrite are the main auriferous phases forming the quartz-sericite assemblage [1,2,3] and therefore can provide insights related to the ore-forming processes, including information about the evolution in gold deposition. However, mineral chemistry studies of these sulphides and their trace elements have not been previously documented at Golden Ridge, and the relationship between Au and other trace elements is still poorly understood. Therefore, in this contribution, geochemical vectors to gold mineralization are ascertained by evaluating the bulk rock chemistry of sericitized host rocks from low-grade to high-grade drill core samples. The paragenetic sequence of the mineralization, mineralogical texture of pyrite and arsenopyrite, and their in situ chemical composition via LA-ICP-MS spot analyses are comprehensively investigated in order to determine the mineralization process and their implications on the mechanism of gold saturation. This study complements the earlier studies of Chi and Watters [2] and Chi et al. [3] who focused on detailed petrographic work based on field and drill core observations followed by geochemical analysis of the different rock units, U-Pb and ⁴⁰Ar-³⁹Ar geochronology that attempted to constrain the age of the mineralization, and fluid inclusion and carbon and oxygen isotope analyses to characterize the mineralizing fluids. The significance of this study, which includes drill core material not studied previously, is the analysis of new bulk geochemical data and trace element compositions of arsenopyrite and pyrite that will enhance and reevaluate the paragenetic relationship between gold mineralization and sulphides to better understand this mineralizing system.

2. Geological Background

The Canadian Appalachians are a section of the Appalachian-Caledonian accretionary orogen, comprising an assemblage of microcontinental terranes sequentially accreted to Laurentia as a result of the closure of the Iapetus and Rheic oceans [22,23,24]. At first, the Canadian Appalachians were divided into ‘tectonostratigraphic zones’, i.e., Humber, Dunnage, Gander, Avalon, and Meguma, and subzones separated by major crustal-scale faults [25,26]; these initial subdivisions changed as a better understanding was obtained regarding the geological evolution of these zones that encompassed distinct tectonic elements and were separated by major faults [27,28,29]. Currently, the Humber zone refers to the remnants of Laurentia’s Appalachian margin during the Cambrian and Ordovician, and the Dunnage zone groups the remaining fragments of several continental and oceanic arc terranes in the Iapetus Ocean that were subsequently accreted to Laurentia [30]. The Gander, Avalon, and Meguma zones represent peri-Gondwanan (Ganderia, Avalonia, and Meguma) provenance microcontinents also accreted to Laurentia [31,32,33,34]. This sequential accretion of terranes and microcontinents caused multiple collisional events between the Late Cambrian and Late Carboniferous: The Taconic, Penobscot, Salinic, Acadian, Neoacadian, and Alleghanian orogenic cycles [24]. The microcontinent of Ganderia is composed of volcanic terranes that separated from, and were later re-attached to, Ganderia itself during the Penobscot and Salinic orogenic cycles [35]; in the New Brunswick segment of the Appalachians, Ganderia consists of seven terranes: Brookville, New River, Annidale, St. Croix, Miramichi, Elmtree, and Popelogan [36]. The Caledonia terrane represents a segment of Avalonia in southern New Brunswick [37,38,39].

The Poplar Mountain Volcanic Complex includes the Golden Ridge area (Figure 1), in New Brunswick, and is exposed within the southwestern end of the highly deformed sedimentary and volcanic rocks of the Miramichi terrane, which extends from Bathurst in the northeast to Woodstock in the southwest of New Brunswick and consists of the Woodstock Group (Late Cambrian to Early Ordovician) and the conformably overlying Meductic Group (Early Ordovician to Middle Ordovician) [2,3,40,41]. The Cambro-Ordovician Woodstock Group includes quartz-rich sedimentary strata, such as greenschist grade quartzose sandstone and shale, whereas the Ordovician Meductic Group contains felsic, intermediate, and mafic metavolcanic rocks in the lower section and clastic metasedimentary rocks in the upper section [36,40,41,42]. The Miramichi terrane is flanked to the northwest by the Matapedia Basin marked by the Woodstock Fault, and bounded by the Meductic Fault to the southeast, which separates it from the Canterbury Basin. The Matapedia Basin consists of carbonate rocks of the Matapedia Group and clastic sedimentary rocks of the Perham Group [43,44,45,46]. The Canterbury basin is composed of marine clastic and carbonate rocks of the Silurian Perham Group and the overlying Devonian Tobique Group, consisting of shallow marine clastic and volcanic rocks [47]. Several Middle Ordovician and Late Silurian to Middle Devonian age plutons intrude the Miramichi terrane, including the Middle Ordovician Benton Granite [48,49] and the Early to Middle Devonian Pokiok Batholith [50,51].

Both orogenic gold deposits and intrusion-related gold deposits have been identified in southern New Brunswick [21,52,53]. The closest major epithermal deposits to New Brunswick are the Hope Brook deposit and the Au high-sulphidation prospects in the Burin Peninsula both in Newfoundland [54]. At the Golden Ridge (Poplar Mountain) deposit, gold mineralization is hosted in the Poplar Mountain Volcanic Complex (PMVC), proximal to the Woodstock Fault, at the southwest end of the Miramichi Terrane. The PVMC is composed of three principal units: A porphyritic rhyodacite and similar rhyodacitic volcaniclastic rocks, and a mafic volcanic unit [55] (Figure 2). The porphyritic rhyodacite consists of 10% to 30% phenocrysts, primarily plagioclase and K-feldspar, and rarely quartz. The unit is characterized by porphyritic rocks, locally brecciated, containing fragments of porphyritic rhyodacite within a matrix of the same rock type. The rhyodacitic volcaniclastic unit contains heterolithic and monolithic volcanic breccias (Figure 3a), as well as ash tuffs and flows. Local dissemination of ultrafine black pyrobitumen occurs within the breccias (Figure 3c), and locally, a glassy matrix that has also been altered to an assemblage of chlorite and sericite also occurs. The mafic volcanic unit contains dark green to grey massive homogenous basalt and an intercalated subunit of plagioclase crystal tuff and chert. The mafic unit is divided into fine-grained basaltic rocks containing pyroxene, plagioclase, chlorite, titanite, and minor quartz, with occasional mm-size plagioclase phenocrysts, and a very fine-grained unit with amygdules filled with calcite [3]. The mafic unit stratigraphically overlies the felsic units, with the porphyritic rhyodacite interpreted to be a subvolcanic intrusion localized at the center of a volcanic dome and surrounded by rhyodacitic volcaniclastic rocks [2]. The volcanic and subvolcanic rocks exhibit brittle deformation in the form of fracturing and brecciation and generally lack any obvious foliation, present only locally as a preferential orientation of sericite in the deepest section of some drill holes. Extensive fracturing occurs in the porphyritic rocks, which are cut by minor brittle faults, locally filled with mineralized quartz veins and bordered by mineralized quartz-cemented breccia (Figure 3b). Chi et al. [3] performed U-Pb geochronology on zircons recovered from a sample of the porphyritic rhyodacite and obtained two populations clustering around 459 Ma and 390 Ma, defined by 16 grains and 2 grains, respectively.

3. Ore Deposit Geology

Gold mineralization occurs in several zones within the porphyritic rhyodacite and the mafic volcanic unit, predominantly associated with acicular arsenopyrite, occurring both in the phenocrysts and in the groundmass of the porphyritic rhyodacite, and less so with pyrite, which occurs disseminated in host rocks, sulphide veinlets, quartz-carbonate veins, and the breccia matrix [2,3]. There is no evident control of mineralization by individual faults nor structural boundaries between mineralized and non-mineralized areas, although high-density brittle fracturing constitutes a common feature in mineralized zones in combination with quartz-carbonate-sericite veins and quartz-cemented breccias [3]. The NW- and E–W-trending faults reported in the area were identified with the use of ground magnetic data. Subsequent detailed mapping identified subsidiary fractures and slickensides on the fault planes that indicated sinistral strike-slip movements along the NW- and NNW-trending faults and dextral strike-slip movements along the E–W-striking faults [1,3]. Chi and Watters [2] interpreted the orientation of these faults and the movement senses as an indication of being conjugate and related to NW–SE oriented shortening.

There are two main phases of alteration. The first consists of a chlorite–carbonate–quartz (chloritization) assemblage, with local illite and pyrite; the second consists of sericite–carbonate–quartz (sericitization), with minor illite and pyrite [55]. The chlorite–carbonate–quartz alteration occurs in the three principal bedrock units and is best preserved in non-mineralized zones, where chlorite preferentially replaces phenocrysts of hornblende and pyroxene, although it is locally present as cement in some hydrothermal breccias within the porphyritic rhyodacite surrounding the mineralized zones [2]. In mineralized zones, sericite-carbonate-quartz, i.e., sericitization, replaces chlorite-carbonate-quartz, with sericite replacing feldspar phenocrysts, and locally, the rock matrix groundmass with increasing alteration intensity [3].

The orientation of veins and veinlets is highly variable at both the outcrop and drill core scales. The first group consists of black-colored, millimeter- to centimeter-wide veinlets composed of fine-grained iron oxide, carbonate, albite, and quartz-cut chloritized areas. This first group is cut by dark-gray centimeter- to tens-of-centimeter-wide veins composed of microcrystalline quartz that are cut by later quartz-carbonate-sericite veins. Furthermore, many mutually crosscutting veins and veinlets of quartz–carbonate–sericite ± illite ± sulphides in the mineralization zones in the porphyritic rhyodacite overprint the chloritization and the earliest veining stages and, therefore, to a large extent are contemporaneous with sericitization and disseminated sulphide mineralization (Figure 3d) [2,3]. Chi et al. [3] performed ⁴⁰Ar/³⁹Ar geochronology on sericite, related to gold mineralization, and defined a plateau age of 411 ± 3.7 Ma, which falls within the age of the sequentially assembled and spatially related Pokiok Batholith (402–415 Ma) [50]. The different mineralizing events at the nearby Lake George mine, located along the eastern margin of the Pokiok Batholith, spanned a 4-million-year period, between 417 and 413 Ma. This period includes ⁴⁰Ar/³⁹Ar geochronology of hydrothermal muscovite from the sericitic alteration envelopes surrounding gold-bearing quartz–carbonate veins that yielded a ⁴⁰Ar/³⁹Ar plateau segment date of 414.1 ± 1.3 Ma [56]. As part of this contribution, in situ ⁴⁰Ar/³⁹Ar geochronology on illites from the sericitized porphyritic rocks was undertaken. However, considering the very fine grain sizes of the illite, it was extremely challenging to obtain a representative age, and most analyses produced imprecise data. One age based on a single spot analysis yielded an age of 390.1 ± 8.1 Ma.

The sulphides include arsenopyrite and the dominant sulphide phase, in addition to pyrite, and to a lesser extent, stibnite and sphalerite. No visible gold was found in hand samples, the drill core, or thin sections. The non-metallic minerals are primarily sericite, quartz, and ankerite, with minor illite. The stronger the sericitization, the more enriched the gold mineralization, and the abundance of both arsenopyrite and pyrite increases. The paragenetic relationships in the Golden Ridge deposit have been studied considering physico-chemical conditions, host rock type, and the minerals observed [2,3]. In these evaluations, it was indicated that mineralization is epigenetic. This is consistent with the observations made during this study, although a more detailed sulphide paragenesis is described here. Pre-gold mineralization pyrite (Figure 4a–d) is associated with chloritization of the country rock and occurs as fine-grained (~50–150 μm), anhedral to subhedral grains. Four types of pyrites in the syn-gold stage, Py-I, Py-II, Py-III, and Py-IV, and two types of arsenopyrites, Asp-I and Asp-II, are defined at the Golden Ridge gold deposit. The pyrite (Py-I) represents an early disseminated alteration mineral assemblage that includes pyrite (Py-I), sericite, carbonate, and quartz. It primarily occurs as highly fractured anhedral to subhedral grains (Figure 4e,h,i) with a diameter of several hundred microns to a few micrometers; it has a randomly oriented and inhomogeneous sieve texture infilled with silicates. Py-II occurs as subhedral grains (Figure 4e,g) with a minor sieve texture around the core, with diameters of several hundred microns to one micrometer, and as growth zones of Py-I with variable widths, in areas where small amounts of arsenopyrite occur. Both Py-I and Py-II are concentrated in areas with moderate sericite-ankerite-quartz alteration. Py-III is mid-size (300–800 μm) subhedral grains (Figure 4h–j) with increased microporosity, compared to Py-II, with minute inclusions of gangue minerals and arsenopyrite, also present as overgrowths and fillings in fractures. It is primarily related to quartz-carbonate-sericite veins, disseminated in both the fragments and matrix of breccias in intensely sericitized rocks and is commonly overgrown or rimmed by arsenopyrite. Py-IV (Figure 4k,l) is found in late mm-thick quartz-sericite-sulphide veinlets as aggregates with irregular boundaries and/or agglomerates, occasionally mantled by arsenopyrite or partially to completely replaced by it.

Arsenopyrite commonly occurs as small (200–400 μm) acicular or rhomb-shaped grains (Figure 4e–l), disseminated or dispersed in the host rocks, as inclusions within pyrite, and as intergrown aggregates. Arsenopyrite (Asp-I) is disseminated in host rocks (Figure 4f,g), quartz-carbonate-sericite veins, and breccias, typically acicular and rhomb-shaped. Arsenopyrite (Asp-II) is primarily restricted to late mm-thick sulfide quartz-sericite veinlets with acicular and granular shapes, or partially and locally replacing pyrite (Py-IV) (Figure 4k,l). Py-I is linked to an initial stage of arsenopyrite (Asp-I), sphalerite, and stibnite deposition, in addition to disseminated pyrite (Py-II) as individual grains or as growth/rims in Py-I. Py-III precipitated within veins and in hydrothermal breccias cemented by quartz and carbonates, along with the main stage of arsenopyrite (Asp-I) deposition. Py-IV is found in late, mm-thick quartz-sericite-sulphide (pyrite-arsenopyrite) veinlets that are commonly cut by quartz-carbonate veins that postdate ore deposition (Figure 5).

4. Materials and Methods

4.1. Sample Selection and Bulk Analytical Methods

Samples were collected from a halved NQ drill core kept by the New Brunswick Department of Natural Resources and Energy Development at its Sussex, New Brunswick facility. Rock samples were generally selected to better examine areas of interest based on rock alteration, texture, and mineralogy. The bulk compositions of 44 rock samples were evaluated after crushing, splitting to 100 g, then pulping with a soft iron swing mill. Twenty-six elements (Au, Ag, As, Cd, Cu, Mn, Mo, Ni, Pb, Zn, Ba, Bi, Ca, Cs, Fe, Ga, Ge, Hg, K, Na, Sb, S, Se, Te, Tl, and W) were determined at Actlabs by instrumental neutron activation analysis (INAA) [57] and with aqua regia digestion followed by ICP-MS analysis. Precision and accuracy were determined using several certified reference materials: Sulphide ore mill tailings (RTS-3a; CANMET-MMSL), two gold ores (DS-1 and MA-2c; CANMET-MMSL), diorite gneiss (SY-4; CANMET-MMSL), and zinc-tin-copper-lead ore (MP-1b; CANMET-MMSL).

4.2. Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry

Eleven polished thin sections were analyzed using a laser ablation–inductively coupled plasma–mass spectrometer (LA-ICP-MS) at the University of New Brunswick using a Resonetics S-155-LR 193 nm Excimer LA-ICP-MS with an Agilent 7700x quadrupole ICP-MS. Pyrite was analyzed with the beam size set to 33 μm with a laser repetition rate of 3 Hz. On-sample energy (fluence) was calibrated to 3 J/cm². A 30 s gas background was collected between every ablation. This ensured that backgrounds were calculated with enough time for signals to settle after sample “washout”. A total of 18 isotopes were monitored in pyrite (²⁹Si, ³⁴S, ⁵⁵Mn, ⁵⁶Fe, ⁵⁹Co, ⁶⁰Ni, ⁶³Cu, ⁶⁶Zn, ⁷⁵As, ⁷⁸Se, ¹⁰⁷Ag, ¹²¹Sb, ¹²⁵Te, ¹⁹⁷Au, ²⁰¹Hg, ²⁰⁵Tl, ²⁰⁸Pb, and ²⁰⁹Bi) from which element concentrations were calculated. Arsenopyrite was analyzed with a beam size of 24 μm with a laser repetition rate of 3 Hz. On-sample energy (fluence) was calibrated to 3 J/cm². A 30 s gas background was collected between every ablation. The isotopes ²⁷Al, ²⁹Si, ³⁴S, ⁵⁵Mn, ⁵⁶Fe, ⁵⁷Fe ⁵⁹Co, ⁶⁰Ni, ⁶³Cu, ⁶⁶Zn, ⁷⁵As, ⁹⁵Mo, ¹⁰⁵Pd, ¹⁰⁷Ag, ¹¹⁸Sn, ¹²¹Sb, ¹⁸²W, ¹²⁵Te, ¹⁹⁷Au, ²⁰¹Hg, ²⁰⁵Tl, ²⁰⁸Pb, and ²⁰⁹Bi were monitored in arsenopyrites from which element concentrations were calculated. Data deconvolution was undertaken using the Iolite 2.5 Trace Element Data Reduction Scheme [58], and spikes in the data were automatically filtered using the default 2σ outlier rejection in the Iolite internally standardized trace-element data reduction scheme. Although precautions were taken when selecting the places to be ablated, the presence of nano-inclusions at depth within the analyzed grains is inevitable and the LA-ICP-MS technique cannot always resolve chemically distinct exsolution phases in the subsurface, particularly those at the nanometer scale. Calibration and data reduction for the LA-ICP-MS spot analysis were performed using standards MASS-1 and NIST610.

Pyrite and arsenopyrite trace element mapping was performed by ablating sets of parallel lines across the samples [59]. A beam size between 13 and 24 μm was used depending on the grain size, trying to obtain the desired sensitivity of elements of interest and spatial resolution. The path separation between lines was set to be identical to the laser beam diameter. The laser was fired at a 10 Hz repetition rate and energy set to produce a fluence at the sample of ~1.5 J/cm². Raster lines of standards MASS-1 and NIST610 were analyzed in order to correct for instrument drift during the offline data reduction. Images were compiled and processed using Iolite 2.5, using the Fe CPS elemental map as a guide to masking the non-pyrite material [60,61]. To produce the quantitative 2D concentration maps from these intensity rasters, an Fe value of 46 wt.% was used as the internal standard. Concentration scales for each map are internally standardized and reported in absolute ppm.

4.3. Statistical Methods

Data with values below detection limits were imputed using the detection limit divided by the square root of two [62]. There were no above-detection limit values, so no sensitivity analysis was conducted. Correlation analysis, calculating the Pearson Product correlation coefficient (r), was conducted by applying a centered log-ratio (clr) transformation to the dataset in order to account for the sensitivity of the data set to peculiarities that induce incorrect correlations [63]. Principal Component Analysis (PCA) was used to simplify the interpretation of the geochemical data [64,65] and conducted on the clr-transformed data [66]. All analyses were conducted using both Minitab^® (version 20.2) and CoDaPack (v2.03.01).

4.4. Micro-X-ray Fluorescence (μ-XRF) Elemental Mapping and BSE Imaging

Selected polished thin sections were scanned by μ-XRF at the University of New Brunswick (UNB) using a Bruker M4 Tornado with a Rh X-ray tube, operating at 30 to 40 kV and 400 μA. The raster multi-element maps were produced by static X-ray spot sizes of 20 μm with 30 μm spacing [67].

Back-scattered electron (BSE) images were collected at the UNB Microscopy and Microanalysis Facility with a JEOL JSM-6400 Scanning Electron Microscope equipped with an EDAX Genesis 4000 Energy Dispersive X-ray (EDS) analyzer using an accelerating voltage of 15 kV and a beam current of 1.5 nA, with a working distance of 14 mm. Images were acquired using a Digiscan II controlled by Gatan Digital Micrograph software. Samples were coated with carbon for conductivity using an Edwards E306A evaporation coater.

5. Results

5.1. Bulk-Rock Geochemical Analyses

Table 1. Summary table for the bulk-rock compositions (in ppm).

Element *	Min.	Max.	Mean	Median	Std. Dev.	Q. 0.25	Q. 0.75	DL
Au	0.039	10.20	2.485	2.470	1.990	0.761	3.313	0.005
As	86	17,400	4471	4515	3646	1450	6040	2
Cu	2	1160	35	7	174	4	11	1
Mn	39	852	451	441	189	333.5	556	2
Ni	0.71	13	3.5	3.0	2.8	1.75	4.25	1
Pb	4	3500	99	14	525	8	26	2
Zn	16	1240	142	59	271	43.25	79	1
Bi	0.24	2.52	0.76	0.46	0.65	0.3275	1.06	0.1
Ca	400	57,100	19,077	18,650	9912	12,450	23,525	100
Cs	0.29	2.69	1.50	1.515	0.58	1.10	1.90	0.05
Fe	6100	57,200	30,070	31,350	11,267	22,600	34,750	200
K	800	4500	2923	3100	814	2575	3400	100
Na	200	49,000	7941	1850	10,568	975	11,150	100
Sb	6.4	1720	68.2	31.1	255.2	21.5	37.0	0.2
S	780	49,690	14,607	13,635	9631	9310	17,465	10
Se	0.07	2.90	0.39	0.30	0.46	0.10	0.50	0.10
W	2.8	28.0	10.9	9.0	7.3	5.8	15.0	4.0

* The raw data for some elements were reported as below detection limits. The data below the detection limit were imputed using the detection limit divided by the square root of two (Ni, Se, and W). Min: Minimum; Max: Maximum; Std: Standard; Dev.: Deviation; Q. 0.25: First quartile; Q. 0.75: Third quartile; DL: Detection limit.

contains a summary of the bulk geochemical results (Appendix A,

Table A1. Major and trace elements abundances of sericitized porphyritic rhyodacite from the Golden Ridge deposit.

	Au	Ag	As	Cd	Cu	Mn	Mo	Ni	Pb	Zn	Ba	Bi	Ca	Cs	Fe	Ga	Ge	Hg	K	Na	Sb	S	Se	Te	Tl	W
Sample	ppb	ppm	ppm	ppm	ppm	ppm	ppm	ppm	ppm	ppm	ppm	ppm	%	ppm	%	ppm	ppm	ppm	%	%	ppm	%	ppm	ppm	ppm	ppm
GR-MM-01	5380	<0.2	7430	<0.5	7	292	<2	4	26	115	<100	1.36	1.02	2.53	3.29	1	<0.1	<1	0.35	0.1	37	2.304	0.3	<0.1	<0.1	26
GR-MM-02	4740	0.4	10,000	<0.5	30	453	<2	6	77	161	<100	2.02	2.4	2.03	4.51	1	<0.1	<1	0.33	0.41	70.8	2.483	0.3	<0.1	0.1	6
GR-MM-03	482	<0.2	464	<0.5	7	337	<2	3	14	70	<100	1.11	2.16	2.19	3.26	1	<0.1	<1	0.24	2.61	10.7	1.421	0.2	<0.1	<0.1	9
GR-MM-04	3260	0.3	4540	<0.5	39	413	<2	13	39	86	<100	2.52	2.98	1	5.07	1	<0.1	<1	0.26	2.74	42.8	3.424	0.8	<0.1	<0.1	4
GR-MM-05	39	<0.2	86	<0.5	8	407	<2	1	11	40	<100	2.32	1.85	2.41	2.26	<1	<0.1	<1	0.41	0.21	15.6	0.103	<0.1	<0.1	0.1	5
GR-MM-06	345	<0.2	382	6	13	181	<2	<1	36	1210	<100	1.06	1.11	1.08	1.2	<1	<0.1	<1	0.17	0.04	27.1	0.762	<0.1	<0.1	<0.1	6
GR-MM-07	2030	<0.2	8490	<0.5	3	792	<2	4	4	25	<100	1.06	4	1.82	4.77	<1	<0.1	<1	0.32	0.14	24.1	2.892	0.4	<0.1	0.1	<4
GR-MM-08	4410	<0.2	5750	0.5	13	738	<2	6	26	75	<100	0.91	1.78	1.9	3.1	2	<0.1	<1	0.45	1.13	31.4	1.686	<0.1	<0.1	<0.1	20
GR-MM-09	2660	0.2	1960	7	4	294	<2	2	26	1240	<100	0.92	0.94	2.05	5.66	2	<0.1	<1	0.34	0.1	16.8	4.969	0.3	<0.1	<0.1	<4
GR-MM-10	774	<0.2	673	5.4	9	39	<2	2	65	837	<100	0.58	0.04	0.42	1.15	<1	<0.1	<1	0.08	0.04	49.9	0.947	0.3	<0.1	<0.1	6
GR-MM-11	2750	<0.2	4770	<0.5	7	420	<2	10	6	77	<100	1.78	2.52	1.03	3.45	2	<0.1	<1	0.34	2.55	20	1.381	0.5	<0.1	<0.1	<4
GR-MM-12	5390	<0.2	5860	<0.5	4	455	<2	4	11	52	<100	1.52	1.76	1.08	2.87	<1	<0.1	<1	0.31	0.1	23.6	1.521	0.3	<0.1	<0.1	10
GR-MM-13	4270	9.9	6320	1.5	1160	602	<2	2	3500	343	<100	2.16	2.02	0.96	2.66	<1	<0.1	<1	0.26	0.09	1720	1.014	2.9	<0.1	<0.1	8
GR-MM-14	5570	<0.2	7000	<0.5	9	579	<2	2	19	58	<100	2.12	3.1	2.24	5.72	1	<0.1	<1	0.33	1.73	36.5	3.187	1	<0.1	<0.1	13
GR-MM-15	89	<0.2	145	<0.5	11	155	<2	<1	21	44	<100	1.13	1.14	1.3	1.8	<1	<0.1	<1	0.16	2.07	12.8	0.167	0.1	<0.1	<0.1	<4
GR-MM-16	723	<0.2	1860	<0.5	2	467	<2	2	8	33	<100	0.53	2.06	0.56	3.89	1	<0.1	<1	0.16	0.07	8.8	0.456	<0.1	<0.1	<0.1	<4
GR-MM-17	57	<0.2	150	<0.5	9	341	<2	<1	8	74	<100	0.47	1.28	1.27	3.66	1	<0.1	<1	0.19	4.9	12.6	0.078	<0.1	<0.1	<0.1	<4
GR-MM-18	1940	<0.2	4650	<0.5	7	557	<2	8	5	49	<100	0.52	1.96	0.81	2.7	<1	<0.1	<1	0.22	1.92	18.5	1.433	0.1	<0.1	0.1	9
GR-MM-19	905	0.2	1700	<0.5	23	340	<2	2	32	35	<100	0.41	1.49	1.91	2.34	1	<0.1	<1	0.31	0.12	38.4	1.183	0.4	<0.1	0.1	16
GR-MM-20	680	<0.2	839	0.9	3	358	<2	<1	5	34	<100	0.3	0.22	1.63	1.46	1	<0.1	<1	0.33	0.06	13.5	0.434	<0.1	<0.1	<0.1	11
GR-MM-21	1550	<0.2	3890	<0.5	5	312	<2	3	13	56	<100	0.46	1.55	2.69	3.17	1	<0.1	<1	0.34	0.09	24.3	1.835	<0.1	<0.1	<0.1	22
GR-MM-22	435	<0.2	1090	<0.5	10	246	4	1	22	55	<100	0.5	0.93	1.48	1.98	<1	<0.1	<1	0.25	0.32	23.5	1.387	<0.1	<0.1	0.4	7
GR-MM-23	1660	<0.2	4260	<0.5	4	429	<2	3	9	41	<100	0.31	1.52	1.4	2.26	<1	<0.1	<1	0.27	0.08	29.8	1.173	0.5	<0.1	<0.1	15
GR-MM-24	2860	<0.2	8000	<0.5	11	388	<2	2	24	72	<100	0.42	1.82	1.74	3.45	<1	<0.1	<1	0.3	0.15	35.2	2.087	0.6	<0.1	<0.1	6
GR-MM-25	3260	<0.2	5640	<0.5	5	495	<2	5	14	69	<100	0.29	3.02	2.03	3.26	<1	<0.1	<1	0.3	0.35	36	0.992	0.7	<0.1	<0.1	13
GR-MM-26	3590	<0.2	5520	<0.5	9	556	<2	6	22	65	<100	0.46	1.72	1.18	3.25	<1	<0.1	<1	0.32	0.16	35.2	1.346	0.4	<0.1	<0.1	7
GR-MM-27	4800	<0.2	17400	<0.5	4	522	<2	<1	9	34	<100	0.4	1.5	1.73	4.47	<1	<0.1	<1	0.34	0.2	36.8	2.508	0.7	<0.1	<0.1	27
GR-MM-28	2620	<0.2	4770	<0.5	4	247	<2	<1	4	34	<100	0.42	0.73	1.41	2.04	<1	<0.1	<1	0.27	0.08	24	1.291	0.2	<0.1	<0.1	11
GR-MM-29	2580	<0.2	4100	<0.5	6	740	<2	3	14	113	<100	0.32	2.46	1.88	2.76	<1	<0.1	<1	0.33	0.84	27.9	1.064	0.3	<0.1	<0.1	14
GR-MM-30	3880	<0.2	5810	<0.5	3	533	<2	4	27	16	<100	0.33	2.22	2.15	3.55	2	<0.1	<1	0.43	0.74	34.1	1.496	0.7	<0.1	<0.1	12
GR-MM-31	2390	<0.2	4490	<0.5	3	506	<2	4	9	47	<100	0.37	2.07	1.9	3.75	1	<0.1	<1	0.37	0.81	22	2.193	0.6	<0.1	<0.1	6
GR-MM-32	1930	<0.2	3600	<0.5	15	852	<2	3	44	44	<100	0.36	5.71	2.01	2.19	<1	<0.1	<1	0.37	0.16	40.8	0.883	0.4	<0.1	<0.1	15
GR-MM-33	2200	<0.2	3150	0.7	5	556	<2	8	16	204	<100	0.34	3.19	1.61	2.2	<1	<0.1	<1	0.24	0.09	30.8	0.978	0.4	<0.1	<0.1	17
GR-MM-34	484	0.3	774	1.3	21	139	<2	<1	70	143	<100	0.39	0.8	0.29	0.61	<1	<0.1	<1	0.09	0.02	46.9	0.273	0.2	<0.1	<0.1	5
GR-MM-35	10,200	<0.2	13,900	<0.5	7	539	<2	2	13	57	<100	0.26	2.04	1.12	3.74	1	<0.1	1	0.32	1.53	37.3	2.301	0.7	<0.1	<0.1	25
GR-MM-36	55	<0.2	105	<0.5	3	382	<2	2	5	37	<100	0.37	2.47	1.55	2.72	1	<0.1	<1	0.2	2.55	6.4	0.145	<0.1	<0.1	<0.1	<4
GR-MM-37	1320	<0.2	1570	<0.5	2	264	<2	<1	5	31	<100	0.28	0.78	0.64	1.55	<1	<0.1	<1	0.29	0.06	62.5	0.809	0.4	<0.1	<0.1	9
GR-MM-38	2440	<0.2	3750	<0.5	9	512	<2	8	23	56	<100	0.52	2.35	1.22	3.38	1	<0.1	<1	0.37	0.17	32	1.627	0.6	<0.1	<0.1	11
GR-MM-39	2860	<0.2	5950	<0.5	5	323	<2	5	8	67	<100	0.32	1.07	1.9	2.51	1	<0.1	<1	0.37	0.16	29.9	1.451	0.1	<0.1	<0.1	28
GR-MM-40	199	<0.2	607	<0.5	3	357	<2	<1	7	62	<100	0.24	2.36	0.52	2.28	1	<0.1	<1	0.26	1.97	8.3	0.713	<0.1	<0.1	<0.1	<4
GR-MM-41	2960	<0.2	7920	<0.5	8	790	<2	6	7	56	<100	0.31	1.69	1.11	3.3	<1	<0.1	<1	0.34	1.1	35.1	1.699	0.1	<0.1	<0.1	15
GR-MM-42	3470	<0.2	6310	<0.5	14	614	<2	3	10	66	<100	0.48	1.88	1.41	3.31	<1	<0.1	<1	0.27	1.11	32.8	1.717	0.4	<0.1	<0.1	18
GR-MM-43	2620	<0.2	4700	<0.5	6	495	<2	4	10	85	<100	0.27	2.19	1.75	2.47	<1	<0.1	<1	0.28	0.94	37	1.122	0.2	<0.1	<0.1	20
GR-MM-44	2500	<0.2	6350	<0.5	12	837	<2	3	16	59	<100	0.26	2.04	1.23	3.29	<1	<0.1	<1	0.38	0.13	41.8	1.337	0.4	<0.1	<0.1	6

). The median Au concentration is 2.47 ppm, with 10.2 ppm being the maximum value. Arsenic concentrations are particularly high with the minimum value being 86 ppm. The potassium content is less than 0.5 wt.%, whereas Na ranges from below 0.1 wt.% to almost 5 wt.%. The median Ca content is less than 2 wt.%, though one sample contains almost 6 wt.%. Bismuth, Cs, and Se are below 3 ppm, and Sb, with a median of 31.1 ppm, yielded one outlier of 1720 ppm. The majority of the concentrations for Cu, Ni, Pb, and Zn are below 100 ppm. For over 95% of the Ag, Cd, Mo, Ba, Ga, Ge, Hg, Te, and Tl analyses, their concentrations are below the detection limit for each element, and therefore their statistical parameters are not reported. Gold is positively correlated with As and S, and significantly negatively correlated with Cu, Bi, Cs, and Na (α = 0.01); these correlations are an indication of the linear relationship between Au and each element based on the average behavior of the elements in the sample (Figure 6a).

A PCA was used to visualize the variability and facilitate the interpretation of the element concentrations within the samples (Figure 6b,c). The dataset develops three main components with eigenvalues larger than 1. Principal component 1 (PC1) accounts for 32.3% of the total variance while principal component 2 (PC2) accounts for 30% of the total variance, and principal component 3 (PC3) contains approximately 10% of the variance. Both Au and As contribute robustly to these three axes.

5.2. Trace Element Content of Pyrite

The concentrations for the isotopes ⁵⁵Mn, ⁶⁶Zn, ⁷⁸Se, ¹²⁵Te, ²⁰¹Hg, and ²⁰⁵Tl were below their detection limits and therefore were not computed. The spots analyses (

Table 2. LA-ICP-MS spot analyses data of pyrites. All in parts per million (ppm).

Type *	Item	Co	Ni	Cu	Ag	Sb	As	Au	Pb	Bi
Py-I (n = 12)	Minimum	0.33	0.83	65.86	0.12	7.43	51.23	0.16	6.96	0.32
	Maximum	841.8	247.2	946.4	55.3	583.7	7651.5	5.87	4117.3	34.2
	Mean	401.7	61.3	317.3	15.1	248.1	2195.3	2.32	1252.5	15.9
	Median	326.4	26.2	229.8	7.6	174.1	1621.9	1.86	1101.7	18.4
	Std. Dev.	308.2	73.5	268.2	18.1	216.4	2170.8	1.71	1187.8	10.7
Py-II (n = 16)	Minimum	0.01	0.09	8.82	0.01	0.16	689.6	3.18	0.35	0.01
	Maximum	124.5	26.0	214.1	9.2	162.4	23,382.7	55.7	970.1	8.58
	Mean	22.4	6.5	49.1	1.4	32.0	14,269.6	15.0	95.6	1.11
	Median	2.97	4.4	26.4	0.46	15.1	14,630.8	9.7	21.1	0.55
	Std. Dev.	38.0	7.9	53.5	2.5	43.7	6859.6	13.56	237.4	2.08
Py-III (n = 22)	Minimum	0.01	0.10	2.76	0.01	0.22	4083.2	1.12	0.16	0.01
	Maximum	186.0	39.6	218.2	5.6	61.9	45,753.7	91.0	229.5	12.8
	Mean	48.1	10.7	78.6	0.91	23.8	27,021.9	34.7	39.2	1.06
	Median	18.4	3.8	75.0	0.23	15.1	26,313.3	33.0	17.9	0.39
	Std. Dev.	60.8	12.3	52.1	1.41	20.1	11,574.1	28.2	55.1	2.66
Py-IV (n = 6)	Minimum	8.04	1.25	93.7	0.02	2.13	28,196.1	36.3	2.39	0.27
	Maximum	48.0	9.8	294.9	4.1	129.3	57,065.7	210.8	214.6	6.98
	Mean	28.3	4.7	184.1	1.8	44.9	43,384.8	105.1	68.9	2.43
	Median	30.3	3.4	181.2	1.1	20.4	43,891.2	82.0	32.9	1.35
	Std. Dev.	17.6	3.6	81.5	1.8	49.4	9,298.1	74.1	78.9	2.50

* The raw data for some elements were reported as below detection limits. The data below the detection limit were imputed using the detection limit divided by the square root of two (Co, Ni, and Pb).

) indicate that the gold content in Py-I varies from 0.16 ppm to 5.87 ppm with an average value of 2.32 ppm, much lower than in Py-II, with concentration from 3.18 ppm to 55.73 ppm, with an average value of 14.96 ppm. In contrast to the gold content of pyrite in rocks with weak to moderate sericitization (Py-I and Py-II), the dominant pyrite (Py-III), occupying quartz-carbonate-sericite veins and disseminated in intensely sericitized breccia fragments and matrix, is more enriched in gold with an average of 34.70 ppm. Py-IV grains are found in late sulphide veinlets and have the highest gold values ranging from 36.32 ppm to 210.8 ppm, with an average value of 105.1 ppm. This increase in gold content from Py-I to Py-IV is directly correlated with an increase in As with average values of 2195.3 ppm in Py-I, 14,269.6 ppm in Py-II, 27,021.9 in Py-III, and 43,384.8 ppm in Py-IV. Furthermore, Co, Ni, Cu, Ag, Pb, and Bi have different variability, and in general terms, have higher average values in Py-I; the variation of these elements in the different pyrite types is marked by high standard deviation values, a clear indication that the distribution of these elements in individual pyrite is heterogeneous. Minor amounts of stibnite occur in quartz-carbonate veins, where most pyrite corresponds to the Py-III type; however, the antimony concentration is higher in Py-I. Zn was only measurable in a few spots of Py-I and below the detection limit in all other pyrite types; sphalerite, nonetheless, appears in the matrix or as inclusions of Py-I, Py-II, and Py-III.

5.3. Trace Element Content of Arsenopyrite

The concentrations for the isotopes ⁵⁵Mn, ⁶⁶Zn, ⁹⁵Mo, ¹⁰⁵Pd, ¹¹⁸Sn, ¹²⁵Te, ²⁰¹Hg, and ²⁰⁵Tl were below their detection limits and therefore were not computed. The spots analyses (

Table 3. LA-ICP-MS spots analyses data of Asp-I and Asp-II. All in parts per million (ppm).

Type *	Item	Co	Ni	Cu	Ag	Sb	W	Au	Pb	Bi
Asp-I (n = 21)	Minimum	0.27	0.18	0.33	0.003	155.8	0.004	1.47	0.2	0.013
	Maximum	59.00	16.70	27.30	0.320	2880.0	8.000	166.0	9.15	3.66
	Mean	22.39	5.20	10.99	0.074	1015.2	0.972	43.50	3.294	1.06
	Median	17.90	3.81	9.80	0.040	540.5	0.036	28.50	2.31	0.652
	Std. Dev.	19.02	4.84	8.93	0.091	952.1	2.126	40.72	3.247	1.095
Asp-II (n = 15)	Minimum	0.10	0.02	7.40	0.002	0.06	0.003	4.20	0.05	0.008
	Maximum	320.0	90.60	155.0	1.94	1263	6.300	334.0	67.3	2.47
	Mean	28.20	7.33	46.90	0.67	394.6	1.570	131.3	23	0.94
	Median	2.10	0.60	35.60	0.22	152.0	0.041	88.90	12.7	0.86
	Std. Dev.	81.50	23.08	40.00	0.77	437.9	2.332	109.4	22.3	0.68

* The raw data for some elements were reported as below detection limits. The data below the detection limit were imputed using the detection limit divided by the square root of two (Ni, Cu, and Pb).

) indicate that arsenopyrite contains more gold than pyrite; the different types of arsenopyrite show contrasting gold concentrations, with mean gold values of 43.5 ppm and 131.3 ppm for Asp-I and Asp-II, respectively. The arsenopyrite (Asp-II) grains have approximately three times more gold than Asp-I. Furthermore, Co, Ni, Ag, and Bi display similar average values between Asp-I and Asp-II, although with important variability within the individual grains as evident from the minimum, maximum, and standard deviation values. Antimony concentrations show a higher average value of 1015.2 ppm for Asp-I and a lower average value of 394.6 ppm for Asp-II. Mn, Zn, Mo, Pd, Sn, Te, Hg, and Tl were not measurable in any arsenopyrites.

5.4. SEM-EDS Compositions of Pyrite and Arsenopyrite

SEM-EDS analyses showed that S and Fe concentrations of Py-I range from 65.02 to 66.16 at.% and from 33.45 to 34.79 at.%, respectively. In the case of Py-II, they range from 65.26 to 65.72 at.% for S and from 33.58 to 33.97 at.% for Fe. Analyses for Py-III and Py-IV indicated S and Fe concentrations in the ranges of 65.45–66.14 at.% and 32.95–33.57 at.%, and 65.05–65.34 at.% and 33.22–33.53 at.%, respectively. The average values of As for Py-I, Py-II, Py-III, and Py-IV are 0.15 at.%, 0.75 at.%, 0.92 at.%, and 1.46 at.%, respectively. The S/Fe ratios for Py-I to Py-IV are 1.87–1.96, 1.92–1.96, 1.95–2.01, and 1.94–1.96, respectively. SEM-EDS analyses reveal that arsenopyrite is systematically As deficient, ranging from 27.52 to 29.47 at.% and from 27.93 to 29.60 at.% for Asp-I and Asp-II, respectively. The Fe contents are almost uniform, and the As/S atomic ratio varies between 0.72 and 0.82, representing a characteristic compositional parameter (Appendix C,

Table A4. SEM-EDS compositions of pyrite and arsenopyrite from Golden Ridge.

	wt.%				at.%
Stage	As	S	Fe	Total	As	S	Fe	S/Fe
Py-I	0.36	51.60	47.39	99.35	0.20	65.35	34.46	1.90
Py-I	0.36	51.33	47.83	99.52	0.20	65.02	34.79	1.87
Py-I	0.24	51.68	47.29	99.21	0.13	65.47	34.40	1.90
Py-I	0.18	52.31	46.47	98.96	0.10	66.16	33.75	1.96
Py-II	1.39	51.69	46.62	99.70	0.75	65.39	33.86	1.93
Py-II	1.28	51.82	46.12	99.22	0.69	65.72	33.58	1.96
Py-II	1.47	51.65	46.63	99.75	0.80	65.34	33.87	1.93
Py-II	1.42	51.72	46.89	100.03	0.77	65.26	33.97	1.92
Py-III	1.81	51.45	45.96	99.22	0.99	65.45	33.57	1.95
Py-III	1.69	52.31	45.39	99.39	0.91	66.14	32.95	2.01
Py-III	1.58	52.02	45.80	99.40	0.86	65.85	33.29	1.98
Py-IV	3.16	50.83	45.21	99.20	1.73	65.05	33.22	1.96
Py-IV	2.51	51.22	45.44	99.17	1.37	65.34	33.29	1.96
Py-IV	2.30	50.63	45.34	98.27	1.27	65.20	33.53	1.94
	wt.%				at.%
	As	S	Fe	Total	As	S	Fe	As/S
Asp-I	41.96	22.40	37.27	101.63	29.07	36.27	34.65	0.80
Asp-I	41.20	22.78	37.12	101.10	28.56	36.91	34.52	0.77
Asp-I	40.15	23.74	37.46	101.34	27.52	38.02	34.45	0.72
Asp-I	41.00	23.04	37.08	101.12	28.36	37.23	34.41	0.76
Asp-I	42.05	22.65	37.06	101.75	29.06	36.57	34.37	0.79
Asp-I	41.42	23.07	37.06	101.55	28.56	37.16	34.28	0.77
Asp-I	41.78	22.56	37.09	101.44	28.96	36.54	34.50	0.79
Asp-I	42.19	22.32	36.38	100.89	29.47	36.43	34.10	0.81
Asp-I	42.24	22.11	37.30	101.64	29.35	35.89	34.77	0.82
Asp-II	42.25	22.06	36.49	100.79	29.60	36.11	34.30	0.82
Asp-II	41.13	22.78	36.70	100.61	28.64	37.08	34.28	0.77
Asp-II	40.52	23.51	36.99	101.03	27.93	37.87	34.21	0.74
Asp-II	41.71	22.42	36.97	101.10	29.03	36.45	34.52	0.80
Asp-II	41.64	22.41	36.65	100.69	29.08	36.57	34.35	0.80
Asp-II	40.81	23.17	37.14	101.11	28.19	37.40	34.41	0.75
Asp-II	42.09	22.16	36.72	100.98	29.41	36.17	34.42	0.81
Asp-II	40.86	22.48	36.68	100.01	28.66	36.83	34.51	0.78
Asp-II	41.13	22.85	36.87	100.85	28.57	37.07	34.36	0.77
Asp-II	42.14	22.46	36.25	100.86	29.42	36.64	33.95	0.80

).

6. Discussion

6.1. Element Association from Geochemical Data

The lithogeochemical results reported here are similar to those reported by Chi et al. [3] for sericitized porphyritic rhyodacite samples, where the intensity of alteration is directly correlated to the abundance of arsenopyrite and pyrite, and therefore to the Au and As concentrations. The significant negative correlation of Au with Na relates to the moderate depletion in Na₂O; the subsequent enrichment in K₂O with an increasing degree of sericitization is not reflected by the correlation analysis (Figure 6a). However, petrographic evidence and micro-X-ray fluorescence (μ-XRF) elemental mapping illustrate the strong relation between the presence of arsenopyrite and pyrite in sericitized rocks with K and Sb (Figure 7); the disseminated style constitutes the most important mineralization for both arsenopyrite and pyrite (Figure 7a), followed by quartz-sericite veins (Figure 7b,c). Stibnite is present as a minor constituent, although common in the quartz-sericite-arsenopyrite-pyrite mineral assemblage. A weak positive correlation between Au and As with Ni, Se, W, and Sb exists, whereas a negative correlation persists between Au and Zn, Cu, Pb, and Bi (Figure 6a). Silver and Hg are generally lower than their detection limit, regardless of the Au contents.

Petrological observations made in the field and in the drill core showed that Au in the Golden Ridge deposit is positively related to the presence of arsenopyrite, pyrite, stibnite, and sphalerite [2,3]. Despite this, it is not clear why Zn is not showing a statistical correlation with Au, as expected due to the presence of sphalerite. One reason might be that samples with low Au grades were part of the statistical analyses, and the presence of sphalerite was insufficient to display a positive correlation between Zn and Au. Another reason might lie in the Au distribution coefficient between solid and fluid phases for sphalerite, approximately two orders of magnitude lower than pyrite [68]. Tungsten has a weak positive correlation with Au, although no tungsten-bearing mineral has been identified at Golden Ridge. In the case of Te, the values below the detection limit suggest the absence of precious metal tellurides.

In accordance with the biplots illustrated in Figure 6b,c, both Gold (Au) and Arsenic (As) demonstrate prominent positive loadings in the first principal component (PC1). Furthermore, Antimony (Sb), Selenium (Se), Sulfur (S), Tungsten (W), and Lead (Pb) also present positive loadings, albeit to a lesser extent. It is critical to note the interpretative relevance of the length of links between two elements in clr-based biplots. Shorter links in this context denote a reduced log-ratio variance between the pair of elements, thereby suggesting a closer association between the mentioned elements. Consequently, such an observation provides an analytical basis for the assertion of a closer relationship between the two elements in question [66,69].

6.2. Correlation of Trace Metals in Sulphides

The concentrations of arsenic and gold in common sulphide minerals (e.g., pyrite) are positively correlated, as both gold and arsenic can replace iron and sulfur in the crystal lattice of pyrite [7]. Furthermore, it has been demonstrated that the maximum gold concentration is strongly correlated with the arsenic content, from which a solubility limit can be estimated [70]. In Figure 8a, Au and As display a positive correlation and there is also an increase in arsenic from Py-I to Py-IV. If pyrite with arsenic contents above 1 wt. % is considered arsenian pyrite [11,71], then Py-II to Py-IV can be classified as arsenian pyrites. All pyrite data points plot below the Au solubility limit (Figure 8a), an indication that gold occurs as structurally bound (invisible-refractory) gold in the arsenian pyrite structure [70,72,73,74] and exhibits a somewhat wedge-shaped-like distribution, a common feature of Au-bearing arsenian pyrite from hydrothermal ore deposits [70,75]. Py-I contains more Sb, Cu, Pb, and Ag than pyrites Py-II to Py-IV (Figure 8b–e); this suggests that Py-I originated from a different hydrothermal fluid than the one (or ones) from which pyrites Py-II to Py-IV originated. The hydrothermal fluid responsible for Py-II, Py-III, and Py-IV contains less Sb, Cu, Pb, and Ag and defines multiple pulses of hydrothermal fluids associated with the ore-forming process. Most pyrites, except for the Py-I type, have high Au/Ag ratios and plot above the Au-Ag = 1 ratio line. The Co and Ni content in pyrite shows a broad range of variability, with Py-I having the higher Co (401.7 ppm) and Ni (61.3 ppm) mean values, whereas the mean values for Py-II, Py-III, and Py-IV are lower and fairly similar. At Golden Ridge, most Co/Ni ratios in all generations of pyrite and arsenopyrite range between 1 and 10 (Figure 8f and Figure 9d), reflecting a consistent hydrothermal fluid composition for these metals [76,77,78,79].

The LA-ICP-MS data show higher values of Au in arsenopyrite (Figure 9a), only comparable to Py-III and Py-IV (Figure 8a, and indicating the incorporation of refractory Au (nanoparticles and/or lattice-bound) with a preferred enrichment of Au in arsenopyrite. Antimony is relatively enriched in both pyrite (Figure 8c) and arsenopyrite (Figure 9a), coinciding with the precipitation of stibnite and pointing towards relatively high Sb in the ore-bearing fluid. Arsenopyrites (Asp-II) have higher concentrations of Cu, Pb, and Ag than arsenopyrites (Asp-I) (Figure 9b,c), although in similar ranges to Py-II, Py-III, and Py-IV.

The textural characteristics and the mineral compositions suggest at least four pulses of fluid involved in the generation of pyrite during the gold mineralizing event at Golden Ridge. The first fluid was enriched in Sb, Pb, Cu, Co, Ni, and Bi. The second pulse of hydrothermal fluid was enriched in arsenic and was responsible for the formation of Py-II and arsenopyrite (Asp-I); the third and fourth pulses were enriched in both arsenic and gold and constituted the most important Au deposition stages. The LA-ICP-MS mapping results illustrate higher amounts of Co, Ni, Cu, Ag, Sb, Pb, and Bi in Py-I than Py-II. The maps also reveal oscillatory zoning of Au, As, and Cu in Py-II; Co and Ni tend to enrich in the rim (Figure 10). Although enriched in Au and As, Py-III and Py-IV are commonly low in most other trace elements: Ag, Sb, Pb, and Bi (Figure 11 and Figure 12). Both Py-III and Py-IV exhibit Au and Cu zonation. The patterns of the counts (Figure 10k, Figure 11k and Figure 12k) for the pyrite grains are relatively smooth for As and their parallel trend with Au reaffirms their described correlation. However, Au shows spiky patterns indicating the presence of a significant amount of micron-sized inclusions. Compared to pyrite, arsenopyrite (Figure 13 and Figure 14; see Appendix D for a rhomb arsenopyrite LA-ICP-MS map) is more enriched in gold, also showing gold zonation, with richer rims than cores, and both Asp-I and Asp-II also show local spikes in the gold patterns (Figure 13k and Figure 14k).

The As zoning in pyrite is observed in BSE imagery where the brighter areas contain more arsenic, and pyrites are increasingly brighter compared to Py-I to Py-IV (Figure 15a–c,e). SEM-EDS analyses of pyrite indicated S/Fe ratios in the range of 1.87–2.01 with an increase in the As concentration along with a decrease in Fe (Figure 15e), a well-documented incorporation mechanism of Au into a solid solution [70,80]. Both As and Au enrichment in arsenian pyrite at a low temperature is most likely the result of a refinement process within a prolongated or multi-episodic fluid flow also accompanied by a decrease in pyrite volume due to structural distortion and evident in a decrease in the size of the pyrite grains [72,79,81]. The addition of arsenic to pyrite will change both the incorporation mechanisms of gold into the structure and the surficial characteristics that will enhance adsorption [70,72,80]. In arsenopyrite, arsenic and sulphur show a strong negative correlation regardless of the arsenopyrite type (Figure 15f), with a variable As/S ratio ranging from 0.72 to 0.82. The high Au contents in this low arsenic arsenopyrite (Appendix C, Table A4) are related to its non-stoichiometry [82], i.e., As concentrations between 27.52 and 29.60 at.%, compared to 33 at.% as the stoichiometric value [83]. These variations, a common feature in arsenopyrite, lead to disorder in the host lattice and allow the incorporation of gold [84]. The low content of arsenic from the arsenopyrite suggests a low-temperature and non-equilibrium pyrite-association formation [85,86]. The average values of the total trace element concentrations for arsenopyrites obtained via LA-ICP-MS are 1235 ppm (range: 307–3818 ppm) and 634 ppm (range: 81–1500) for Asp-I and Asp-II, respectively. These data show that Au and Sb are the main trace elements in arsenopyrites. LA-ICP-MS analysis also indicated the presence of a sulphide mineral phase containing higher amounts of S-Sb-Cu-Pb and anomalous Te, Tl, and Bi. See Appendix C (

Table A5. LA-ICP-MS spot analysis data of unknown sulphide phase.

Sample	S	Cu	Zn	Pd	Ag	Sn	Sb	Te	W	Au	Tl	Pb	Bi
GR1164A-unknown-1	468,298	192,471	121	4.61	111	0.39	260,857	34.4	56.6	0.01	1.73	665,040	272
GR1164A-unknown-2	469,133	177,053	142	5.75	67	0.20	277,907	23.3	139.0	0.01	1.09	719,082	239
GR1164A-unknown-3	640,517	300,327	140	8.40	91	0.48	420,674	66.0	0.20	0.00	1.55	1,120,807	349
GR1164A-unknown-4	462,110	184,213	111	6.26	120	0.51	295,272	14.2	108.0	0.01	2.22	726,989	40
GR1164A-unknown-5	580,348	272,319	57	7.71	95	0.67	404,485	68.4	2.5	0.00	3.47	1,056,271	199
GR1164A-unknown-6	589,776	270,295	105	10.00	101	0.20	407,271	115.0	0.0	0.00	2.29	1,002,284	276
GR1164A-unknown-7	710,725	307,788	130	7.60	250	1.90	513,062	34.0	0.2	0.00	4.10	1,299,018	63
GR1164A-unknown-8	610,371	285,647	131	9.10	251	0.35	477,017	19.4	0.1	0.00	2.70	1,139,270	115
GR1164A-unknown-9	596,101	258,696	82	9.30	215	0.44	438,638	22.5	0.0	0.00	2.93	998,727	57
GR1164A-unknown-10	571,239	235,948	57	7.49	180	0.32	386,345	28.8	17.7	0.00	3.33	872,978	56
GR1164A-unknown-11	495,823	205,861	74	7.17	143	0.34	314,091	19.6	83.5	0.00	3.35	736,430	44

).

6.3. The Nature of “Invisible” Au

The average Au and As content in Py-I is 2.32 ppm and 2195.3 ppm, respectively. Py-I grains constitute the sieve textured core of low As content grains that are later mantled by As-richer zones (Py-II), narrow inclusion-rich (primarily arsenopyrite, Asp-I) corona-like texture (Py-III), and finally, pyrite grains with overgrowth zones partially or completed replaced by arsenopyrite (Asp-II). This concentration of trace elements in later growth zones, a common occurrence in pyrites from epithermal environments, accounts for surface interface-controlled growth during metal deposition [87,88] and indicates a disequilibrium crystallization process [89,90]. It is also observed that the amount of Au in the arsenian pyrites Py-II, Py-III, and Py-IV, in general, increases with decreasing sizes of the pyrite grains, a common feature in both Carlin-type and epithermal deposits [72,91,92].

The As content for pyrite grains Py-II, Py-III, and Py-IV ranges from approximately 700 ppm As to almost 6 wt.% of As. Py-I is interpreted to have formed before the introduction of high-As fluid to the system, whereas Py-II, Py-III, and Py-IV formed together with arsenopyrite. Invisible gold occurs in an elemental form as nanoparticles of Au(0) or as alloys with other trace elements, such as Bi, Te, and Ag, and as a chemically bound form. Nonetheless, the nature and redox state of the chemically bound form varies from Au³⁺ to Au^1- [84]. The different element correlations and the LA-ICP-MS data indicate “invisible gold” as the main form of Au present (solid solution in the crystal lattice of pyrite and arsenopyrite or submicron- to nanoparticles and micron-sized inclusions) as native gold and Au compounds (e.g., tellurides) have not yet been identified.

The distribution and size of gold particles in sulphide-sulphate solutions are ultimately controlled by the degree of supersaturation, fluid pH, sulphur concentration, and temperature (cooling favoring metal precipitation) [73]. The miscibility of fluid solutions is supported by high temperatures which leads to cooling as a destabilizing factor for Au dissolved in solid solution in the mineral lattice. Fluid-assisted recrystallization can also induce the breakdown of chemically bound Au that is later adsorbed on the surface of slow-growing crystals [93,94].

The vast majority of processes used to explain the high contents of gold in arsenian pyrite consider adsorption (surface interface) to be an important factor that derives from the inhomogeneous composition of high-Au and high-As pyrite [90,95]. The disequilibrium composition of the surface, in this case of Py-I, leads to deviations of the mineral composition from its equilibrium as nucleation and adsorption occurred during the input of an As-rich hydrothermal solution [71]. This process resulted in gold concentrations in places where gold could adsorb onto the surface of a growing crystal. Although it is observed that gold concentrations increase with increasing arsenic in different gold deposit types and it is believed that arsenic possesses a natural Au scavenging effect that results in the enrichment of invisible gold in both pyrite and arsenopyrite [7,8,72,80], the role of arsenic in gold incorporation in growing pyrite and arsenopyrite remains controversial. It is feasible that the preferential enrichment of gold in arsenopyrites (Asp-I and Asp-II) and pyrites (Py-II, Py-III, and Py-IV) is derived from primary coprecipitation and therefore the resulting absence of native gold as fracture fill or overgrowths in arsenopyrite.

Figure 15. SEM-Back-scattered electron (BSE) images of pyrite and arsenopyrite crystallites from the Golden Ridge gold deposit. (a) Pyrite grains with arsenic increasing from the center (Py-I) towards the rim (Py-II). Values are for As in ppm from LA-ICP-MS spot analysis; (b) pyrite grain enriched in arsenic with silicate and fine arsenopyrite inclusions concentrated around the center; note the arsenopyrite overgrowth on pyrite. Values are for As in ppm from LA-ICP-MS spot analysis; (c) arsenopyrite (Asp-II) and pyrite (Py-IV) in quartz-sericite mm-thick veinlet at the late stage of the mineralizing pulse; (d) disseminated rhomb and acicular arsenopyrite and pyrite in sericitized rock; (e) composition of pyrite from ore deposits on As-Fe-S ternary. Four different trends show substitutions of (i) As for S (As^1--pyrite; purple arrow); (ii) As²⁺ for Fe (As²⁺-pyrite; grey arrow); (iii) As³⁺ for Fe (As³⁺-pyrite; olive arrow); and (iv) divalent metals Me²⁺ for Fe (yellow arrow). After Deditius et al. [72] and Qian et al. [96]; (f) relationship between sulfur and arsenic concentrations in arsenopyrite (Asp-I and Asp-II). Qtz: Quartz, ank: Ankerite, ser: Sericite.

Figure 15. SEM-Back-scattered electron (BSE) images of pyrite and arsenopyrite crystallites from the Golden Ridge gold deposit. (a) Pyrite grains with arsenic increasing from the center (Py-I) towards the rim (Py-II). Values are for As in ppm from LA-ICP-MS spot analysis; (b) pyrite grain enriched in arsenic with silicate and fine arsenopyrite inclusions concentrated around the center; note the arsenopyrite overgrowth on pyrite. Values are for As in ppm from LA-ICP-MS spot analysis; (c) arsenopyrite (Asp-II) and pyrite (Py-IV) in quartz-sericite mm-thick veinlet at the late stage of the mineralizing pulse; (d) disseminated rhomb and acicular arsenopyrite and pyrite in sericitized rock; (e) composition of pyrite from ore deposits on As-Fe-S ternary. Four different trends show substitutions of (i) As for S (As^1--pyrite; purple arrow); (ii) As²⁺ for Fe (As²⁺-pyrite; grey arrow); (iii) As³⁺ for Fe (As³⁺-pyrite; olive arrow); and (iv) divalent metals Me²⁺ for Fe (yellow arrow). After Deditius et al. [72] and Qian et al. [96]; (f) relationship between sulfur and arsenic concentrations in arsenopyrite (Asp-I and Asp-II). Qtz: Quartz, ank: Ankerite, ser: Sericite.

6.4. The Golden Ridge Gold Deposit

Chi et al. [3] identified that sericitization, the mineralization-associated alteration, is accompanied by the addition of K, Rb, Cs, Ca, Sr, Ba, As, Sb, W, C, and S to the system. Most of these elements have the tendency to be concentrated in the top or contact zones of related intrusions and are known as granophile elements. However, many of these elements also form part of the general suite defined for epithermal deposits: Ag, As, Bi, Cu, Hg, Mo, Pb, Sb, Se, Sn, Te, Tl, W, and Zn [97]. Results from the INAA analysis on sericitized samples (Figure 6a) indicated that only As and S are significantly positively (α = 0.01 confidence level) correlated, whereas Cu, Bi, Cs, and Na are significantly negatively correlated (α = 0.01 confidence level). Furthermore, the PCA analyses (Figure 6b,c) showed that Au has a positive relationship with As and to a lesser degree with S, W, and Sb; a negative relationship is indicated for Bi, Zn, Cu, and Pb. These results do not show the common metal signature enriched in Bi-Te-W linked to intrusion-related gold systems, with the close association of Au with Bi, W, Mo, and Te [98,99,100]. Our data show a metal assemblage association of Au with As and Sb, and low concentrations of base metals, a feature common to both intrusion-related [101] and some epithermal systems [102,103,104]. A study carried out in gold and antimony deposits of southern New Brunswick using portable X-Ray Fluorescence Spectroscopy (pXRF) analysis [65] indicated that intrusion-related gold deposits occur in quartz-carbonate veins with sericitization being the main alteration type and S, As, and Sb as the gold pathfinder elements and arsenopyrite, pyrrhotite, and antimony-bearing minerals as the mineral phases associated with gold mineralization. Nonetheless, at Golden Ridge, a strong S-As-Sb association also occurs, a feature common as different deposit types can share a genetic link or be formed in a transitional environment [54]. Intrusion-related gold ores are enriched in Bi and Te and correlate with Au [52], a feature not seen at Golden Ridge.

The trace element content variations in the different pyrite and arsenopyrite grains suggest changes from an early fluid enriched in Sb, Ag, Cu, Pb, Co, and Ni (Figure 8a–f and Figure 9a–d) to later pulses of As- and Au-enriched fluids; there is also certain variability within the different pyrite generations that indicates that the fluid changed spatially at Golden Ridge. The strong correlation between Au and As and Au⁺¹ in solid solution as data plots below the gold solubility defined by Reicht et al. [70] and is consistent with studies performed on epithermal deposits [72,105,106]. The pyrites (Py-II) from the Golden Ridge gold deposit have As concentrations (mean = 14,269.55 ppm; Table 1) that are consistent with the global average (14,310 ppm) for low-sulphidation epithermal gold deposits [107]. LA-ICP-MS data indicate that neither pyrite nor arsenopyrite has either high Co and (or) high Ni, i.e., magmatic origin, but instead exhibit intermediate Co/Ni ratios and therefore suggest a hydrothermal origin [108]. The trace element changes between the different pyrite types suggest disequilibrium processes during rapid saturation, a common feature in many epithermal deposits [72,109,110,111]. Intrusion-related gold systems spread through a broad range of mineralization styles, always relative to intrusive centers, and therefore, can include common patterns described for other deposit types. In the case of intrusion-related distal deposits, these deposits include auriferous, mesothermal to epithermal quartz-sulphide veins, hydrothermal breccias and base-metal veins enriched in Ag ± Au, and Carlin-style deposits [101].

The solubility limit for a solid solution of Au as a function of As presented in Figure 8a is defined for a temperature range between 150° and 250 °C [70]. Fluid inclusion studies carried out on Golden Ridge [3,55] indicated the existence of two hydrothermal events, the first related to the chlorite–quartz–calcite alteration (T_h = 116 °C–159 °C) and the second related to the sericite–quartz–ankerite alteration (T_h = 220 °C–280 °C). This range of homogenization temperatures falls in the range of T_h from fluid inclusions defined for typical epithermal deposits, i.e., approximately 150° to 300 °C [97]. The results of C and O isotopic analysis [3] determined for calcite associated with chloritization and ankerite associated with sericitization showed that fluids in equilibrium with calcite (δ¹³C_PDB = −8.0 to −4.3‰; δ¹⁸O_SMOW = −2.1 to 3.3‰) exhibit values that account for meteoric water input, whereas fluids in equilibrium with ankerite (δ¹³_CPDB = −8.3 to −6.8‰; δ¹⁸O_SMOW = 6.4 to 8.3‰) display magmatic affinity values [112,113]. These characteristics are in line with epithermal deposits, known to be associated with magmatism that results in meteoric water circulation; therefore, the isotopic composition of the mineralizing solutions would show a magmatic water component, particularly in low-sulphidation deposits where the mixing of fluids of magmatic origin and meteoric water occurs [102,114,115,116,117].

The earlier fluid inclusion study [3,55] related to the mineralizing event indicated salinities of 2.3–10.6 wt.% NaCl equivalent, CO₂ as the dominant volatile species, and fluid pressures in the range from 770 to 1240 bars, equivalent to lithostatic depths of approximately 3–5 km. These results seem to be within the range of those from studies of fluid inclusions from epithermal deposits with generally < 3.5 wt.% NaCl equivalent [102], and others with a range between 0.1 and approximately 7.5 wt.% NaCl equivalent for precious metal deposits [97,118]. There is a relationship between precious metal ore with lower salinities in contrast to base-metal ores with relatively higher salinities (up to 23 wt.% equivalent values have been reported) in epithermal deposits [119,120] that also relates to a decrease in salinities and base-metal mineralization as the hydrothermal system evolves [121]. This feature is evident at Golden Ridge where the studied paragenetic stages indicated a decrease in base metals in Py-II, Py-III, and Py-IV compared to the initial stage Py-I (Figure 8c–e). Although uncommon, some hydrothermal systems in epithermal environments with near-neutral pH and deeply sourced fluids have been reported to be gas-rich, dominated by CO₂ [122,123,124,125]. The reason for these CO₂-dominated fluids at Golden Ridge is the interaction of oxidizing fluids with sediments and shales proximal to the Poplar Mountain Volcanic Complex. Most epithermal deposits form at shallow crustal levels [102,103,104], distinct from the > 3 km depth indicated by those fluid pressures from Golden Ridge. Regardless, epithermal activity can take place deeper than that generally described for this type of deposit in which the water–rock interaction or contributions from deep-seated magma chambers are the mechanisms promoting alteration and metal deposition [118,126]; the absence of boiling evidence, i.e., development of lattice textures, adularia, and crustiform-colloform banding [127] at Golden Ridge supports this hypothesis.

The fracturing and brecciation enhanced the penetration of meteoric waters that, along with the conversion of ferrous iron to pyrite in the chloritized areas, allowed the remobilization of gold as bisulphide complexes by these convecting meteoric waters percolating through this brittle environment [128,129]. The most important ligand at temperatures up to 350 °C is HS^- with Au(HS)₂^- being the dominant complex at low temperatures, with the stability constant reaching maxima at 250 °C, reducing conditions and pH near to neutral [130,131]. Additionally, further transportation can take place in response to a sudden decrease in temperature, when a saturated gold-bearing fluid may allow gold to remain in suspension, resulting in this negatively charged colloidal gold adsorbing onto minerals with a positive surface charge, such as arsenian pyrite and arsenopyrite [80,132,133,134,135].

7. Conclusions

At Golden Ridge, gold mineralization was focused within a brittle part of a Middle Ordovician rhyolitic dome, along the faulted margin of ductile graphitic metasedimentary rocks; the wall-rock alteration assemblage includes illite, chlorite, albite, pyrite, quartz, and calcite, reflecting an almost neutral-pH composition of the mineralizing fluid, but with low salt contents. The porphyritic rhyodacite unit exhibits the local development of brittle quartz-cemented breccia, dilatant siliceous veins and veinlets with cockscomb textures, and minor amounts of carbonate with stibnite and sphalerite. The gold mineralization is closely associated with elevated sulphide contents (>3–4%), primarily arsenopyrite and lesser pyrite. To the knowledge of the authors, no diagnostic features of low-sulphidation epithermal systems (i.e., adularia, bladed calcite, and colloform-crustiform textures) have been identified at Golden Ridge, indicating that fluid boiling was not the mechanism of precipitation. Both the bulk rock data and the LA-ICP-MS spot analyses of pyrite and arsenopyrite show a strong correlation between gold and arsenic, and indicated that Py-II, Py-III, and Py-IV are arsenian pyrites. SEM-BSE imaging also indicated “invisible gold” as the main form of Au present in the deposit, in solid solution in the crystal lattice of pyrite and arsenopyrite, <100 nm nanoparticles, and micron-size inclusions. The detailed textural characteristics studied and their compositions suggest the existence of at least four syn-gold mineralization episodes, as separate fluid pulses: The first fluid was enriched in Sb, Pb, Cu, Co, Ni, and Bi, followed by a second pulse of hydrothermal fluid enriched in arsenic and gold, with lesser contents of Sb, Pb, Cu, Co, Ni, and Bi. The third and fourth pulses are enriched in both arsenic and gold and constitute the most important Au deposition stages. The repeated occurrences of growth zones with Au enrichment in the pyrites at Golden Ridge reflect surface adsorption during growth with gold deposition, i.e., a disequilibrium crystallization process. All of these characteristics better describe Golden Ridge as an evolving distal epithermal deposit in which a deep-seated intrusion provided the ore-forming fluid (magmatic component) episodically mixing with meteoric waters and then dominated by meteoric fluids that percolated through brittle domains along the major Woodstock Fault.