Pyrite Genesis and Its Implications for Uranium Mineralization at the Yahewan Deposit, Southern Ordos Basin, China

Abstract

The origin and metallogenic role of pyrite in sandstone-hosted uranium deposits remain difficult to constrain because pyrite may record both early diagenetic reduction and later ore-fluid overprinting. Here, pyrite and associated uranium minerals from mineralized and barren sandstones of the Middle Jurassic Zhiluo Formation in the Yahewan uranium deposit, southern Ordos Basin, were investigated using petrography, SEM-BSE imaging, EPMA, LA-ICP-MS trace-element analysis, and in situ sulfur isotope analysis. Pyrite occurs mainly as framboidal, colloidal, pore-filling, fracture-filling, and massive aggregates and is commonly associated with organic matter and coffinite or coffinite-like U-silicate minerals. Most pyrite domains show low Co/Ni ratios, suggesting a predominantly authigenic to sedimentary–diagenetic origin. High U, Mo, V, W, Se and As occur mainly in pyrite-rich microdomains, but high-U analyses are interpreted cautiously because some signals may include contributions from adjacent coffinite or coffinite-like U-silicate phases, fracture-hosted uranium minerals, or mixed ablation. Together, the petrographic, trace-element and sulfur isotope data support a two-stage model in which early organic matter and authigenic/biogenic pyrite created local reducing microenvironments that were later overprinted by U-bearing basinal fluids.

1. Introduction

Pyrite is a ubiquitous metallic sulfide mineral within the Earth’s crust. It frequently serves as a critical carrier for various ore-forming elements, such as uranium and gold, and is paragenetically associated with diverse rock types and mineral deposits [1,2,3]. Within sandstone-hosted uranium deposits, pyrite typically functions as a reducing agent or an indicator mineral for the redox conditions during the uranium mineralization process [4]. Its formation often spans the entire duration of the uranium mineralizing stages. Consequently, the geochemical composition of pyrite records vital information regarding the evolution of uranium-bearing fluids, serving as a reliable indicator for characterizing the properties of ore-forming solutions, the physicochemical environment, and the mechanisms of uranium mineralization [4,5,6]. In this context, resolving whether pyrite is of early diagenetic origin or was modified by syn-ore fluid overprinting is fundamental to understanding the development of reducing microenvironments and the mechanisms of uranium precipitation.

The Ordos Basin, the second-largest sedimentary basin in China, hosts abundant energy resources, including coal, oil, natural gas, and uranium [7,8]. The Yahewan uranium deposit is a recently confirmed medium-sized deposit located in the southern basin. Previous studies on the Yahewan deposit have addressed sand-body alteration, sedimentary characteristics, and their relationship with uranium mineralization. It is generally recognized that mineralization is primarily controlled by the lithofacies heterogeneity of the main braided channel sandstone bodies. The ore bodies typically occur in tabular or stratiform shapes at the contact zone between gray and green sandstones. The sedimentary materials are likely derived from the Qinling Orogenic Belt along the southern margin of the basin [9].

Previous work in the Yahewan area has documented sand-body architecture, alteration zoning and the spatial association among uranium mineralization, organic matter and pyrite. However, the timing and genetic role of pyrite remain unclear. In particular, it is uncertain whether pyrite mainly represents an early diagenetic reductant, a product of later ore-fluid influx, or a composite record of both processes. It is also unclear whether U and redox-sensitive elements in pyrite are lattice-bound, inclusion-related, or fracture-controlled, and how the large sulfur isotope variability reflects sulfur sources and pore-water evolution. Here, we combine petrographic observations, SEM-BSE imaging, EPMA, LA-ICP-MS trace-element analysis and in situ sulfur isotope measurements of pyrite from mineralized and barren Zhiluo Formation sandstones. The aim is to constrain pyrite genesis and evaluate how pyrite-organic matter assemblages contributed to uranium precipitation in the Yahewan deposit.

2. Geological Setting

2.1. Regional Geological Framework

The Ordos Basin is situated in the southwestern part of the North China Craton (Figure 1), covering an expansive area of approximately 370,000 km². The basin is exceptionally rich in energy resources and serves as a vital national production hub for natural gas, coal-bed methane, coal, and petroleum [10,11,12]. The basin also hosts numerous sandstone-hosted uranium deposits and has emerged as one of China’s most important uranium-producing regions [10,13,14,15,16,17,18]. The Yahewan uranium deposit is located in the western part of the Weibei Uplift in the southern Ordos Basin. Structurally, it occupies the transition zone between the Weibei Uplift and the Bin-Xun Sag, bordered by the Qinling Orogenic Belt to the south and the Yishan Slope to the north (Figure 1) [19,20]. The regional structure is characterized by a gentle monocline dipping toward the north, with relatively simple structural features and low stratigraphic dip angles. The primary structural framework is dominated by NE-trending and nearly E-W-trending faults [18]. Since the Late Cretaceous, tectonic uplift has exerted significant control over paleogeomorphology and groundwater circulation, thereby providing a favorable structural framework for the formation of sandstone-hosted uranium mineralization. At the basin scale, recent studies have further emphasized that large-scale sandstone-hosted uranium mineralization in northern China is jointly controlled by neotectonic activity, stratigraphic architecture, basin evolution, and fluid migration pathways [21,22]. These regional controls provide an important framework for understanding the localization of uranium mineralization in the southern Ordos Basin and help place the Yahewan deposit within the broader metallogenic system of the northern Chinese Meso-Cenozoic basins.

2.2. Geology of the Yahewan Uranium Deposit

The stratigraphic succession exposed in the Yahewan uranium deposit comprises, in ascending order, the Middle Jurassic Yan’an Formation (J₂y) characterized by gray coal-bearing sandstones; the Middle Jurassic Zhiluo Formation (J₂z) composed of variegated sandstones (with the basal light grayish-white and grayish-green pebbly coarse sandstones serving as the primary ore-hosting sand bodies); and the Lower Cretaceous Yijun Formation (K₁y) featuring purplish-red conglomerates, followed by the Luohe Formation (K₁l) comprising brownish-yellow coarse sandstones.

Uranium mineralization in the study area is predominantly hosted within the braided river channel sandstone bodies of the lower member of the Zhiluo Formation. These sandstone bodies exhibit significant thickness and high permeability. In cross-section, the ore bodies typically occur as stratiform or lenticular layers and are mainly localized along the oxidation–reduction transition zone. In the plan view, they display a sinuous morphology (Figure 2). The primary uranium mineral in the ore is coffinite, which mostly occurs as fine-grained disseminations. These minerals exhibit an intimate spatial and genetic association with reducing agents, particularly pyrite and organic matter.

3. Samples and Analytical Methods

3.1. Sample Collection and Preparation

A total of five representative, unweathered sandstone samples were selected from three drill cores in the Yahewan uranium deposit for petrographic and in situ geochemical analyses. These samples were collected from the lower member of the Middle Jurassic Zhiluo Formation (J₂z¹) and include four mineralized sandstone samples from ore intervals, namely 0-9-2, 0-9-3, 86-4-6, and 106-1-5, and one barren sandstone sample, 86-4-7. The sampling depths range from 318.0 m to 681.0 m. All samples are light grayish-white pebbly coarse sandstones, representing the main ore-hosting lithology in the study area. Detailed sample information, including drill-hole number, sampling depth, lithology, mineralization type, redox/color facies, uranium grade or radioactivity, pyrite morphology, and the number of EPMA, LA-ICP-MS, and sulfur isotope analytical spots, is provided in Table S1.

After initial processing, the samples were prepared into standard polished thin sections for microscopic and microbeam analyses. Petrographic observations were first conducted using transmitted- and reflected-light microscopy to identify detrital components, cement types, organic matter, pyrite textures, and mineral paragenetic relationships. SEM-BSE imaging was then used to further characterize the occurrence of pyrite, coffinite or coffinite-like U-silicate minerals, micro-fractures, and possible mineral inclusions. Based on these observations, representative pyrite domains were selected and marked for in situ micro-geochemical analyses, including Electron Probe Microanalysis (EPMA), LA-ICP-MS trace-element analysis, and in situ sulfur isotope measurement.

3.2. Analytical Methods

Petrographic observations, morphological characterization, and compositional analyses were performed at the Analytical Testing Center of the Beijing Research Institute of Uranium Geology (BRIUG). Basic petrographic identification was conducted using a Leica polarizing microscope (Leica Microsystems GmbH, Wetzlar, Germany) under both transmitted and reflected light. Micro-morphological observations and energy-dispersive spectroscopy (EDS) analyses were carried out using an FEI Nova Nano SEM 450 (FEI Company, Hillsboro, OR, USA) and a Zeiss Sigma 300 thermal field emission scanning electron microscope (SEM) (Carl Zeiss Microscopy GmbH, Jena, Germany). In situ quantitative analyses of major elements in uranium minerals were performed using a JEOL JXA-8100 electron probe microanalyzer (EPMA) (JEOL Ltd., Tokyo, Japan) equipped with an EDS system. The EPMA analyses were conducted at an accelerating voltage of 20 kV, a beam current of 10 nA, a take-off angle of 40° and a beam diameter of 2 μm. Albite was used as the standard for Si, Na, and Al; K-feldspar for K; Ca-bearing rhodonite for Ca; rhodonite for Mn; rutile for Ti; hematite for Fe; pyroxene for Mg; and metallic uranium for U. Before microbeam analysis, polished thin sections were examined by transmitted- and reflected-light microscopy and SEM-BSE imaging to identify uranium minerals, pyrite textures, micro-fractures, and possible mineral inclusions.

In situ trace element and sulfur isotope analyses of pyrite were conducted at Kehui Testing (Tianjin) Technology Co., Ltd. (Tianjin, China). Trace element analyses were performed using an Analytik Jena PQMS ICP-MS (Analytik Jena GmbH, Jena, Germany) coupled with an ASI RESOlution 193 nm excimer laser ablation system (Australian Scientific Instruments Pty Ltd., Canberra, Australia). The analyses were carried out in single-spot ablation mode, with a spot diameter of 38 μm, a repetition rate of 5 Hz, and a laser fluence of approximately 5 J/cm², using high-purity He as the carrier gas. Before analysis, the instrument was tuned to optimal performance using the NIST SRM 610 glass reference material. Each single-spot analysis lasted 85 s, including 20 s of background acquisition, 45 s of continuous sample ablation, and 20 s of washout [23]. During the analytical sequence, external reference materials, including NIST SRM 610, NIST SRM 612, BHVO-2G, BCR-2G, BIR-1G, and the sulfide reference material MASS-1, were inserted after approximately every ten unknown analyses for quantitative calibration and quality control. Offline data reduction, including background subtraction, instrumental drift correction, and concentration calculation, was performed using ICPMSDataCal 10.8. Reference values for the glass standards followed Pearce et al. [23], and the ICPMSDataCal-based data-reduction procedure followed Liu et al. [24], without applying an internal standard. During data reduction, time-resolved signals were checked to minimize the influence of obvious inclusions, mineral intergrowths, or mixed-ablation effects.

In situ sulfur isotope analyses were conducted using a Thermo Scientific Neoma Multi-Collector Inductively Coupled Plasma Mass Spectrometer (MC-ICP-MS; Thermo Fisher Scientific, Bremen, Germany) coupled with a RESOlution SE 193 nm solid-state laser (Applied Spectra Inc., West Sacramento, CA, USA). Single-point ablation was performed on the pyrite grains previously identified under the optical microscope, utilizing a spot diameter of 20–30 μm, a laser fluence of 3 J/cm², and a repetition rate of 6 Hz. The ablated aerosol was transported to the MC-ICP-MS using high-purity He as the carrier gas. The ³²S and ³⁴S signals were simultaneously detected in static mode using Faraday cups, with a single-point integration time of 0.131 s. Instrumental mass bias was corrected using a standard–sample–standard bracketing method with sulfide reference materials. The sulfur isotope compositions are reported as δ³⁴S_V-CDT values in per mille.

4. Results

4.1. Pyrite Petrography and Mineral Associations

Pyrite in the Yahewan uranium deposit predominantly occurs in medium- to coarse-grained feldspathic quartz sandstones of the Zhiluo Formation. These sandstones are well consolidated and commonly exhibit calcareous cementation. The core samples display low textural maturity, characterized by poor sorting and rounding of detrital grains, together with a mixed grain-size distribution. Compositionally, quartz and feldspar account for approximately 60% of the total detrital components, followed by igneous and metamorphic lithic fragments at approximately 30%, with minor sedimentary lithic fragments and mica flakes.

The uranium-bearing sandstones are mainly grayish-white, pebbly, medium- to coarse-grained sandstones (Figure 3a). Macroscopically, abundant pyrite (Figure 3b) and organic matter, such as carbonaceous debris (Figure 3c), are visible within the ore specimens. Microscopic observations show that pyrite commonly fills intergranular pores between detrital grains, such as quartz (Figure 3d,f), and also occurs along the cleavage planes of biotite (Figure 3e). Spatially, pyrite is closely associated with organic matter (Figure 3g,h) and locally shows a close paragenetic relationship with uranium minerals, mainly coffinite or coffinite-like U-silicate phases (Figure 3i).

Based on micro-petrographic observations and spatial distribution patterns, pyrite in the studied Zhiluo Formation sandstones can be grouped into three main types.

The first type is pore-filling pyrite, which fills intergranular pores and secondary dissolution pores and represents the most widely distributed pyrite type in the studied samples. This pyrite occurs mainly as disseminated, irregular massive, amorphous, colloidal, and locally euhedral cubic crystals, extensively filling intergranular spaces between lithic fragments and quartz grains (Figure 4a). Some pyrite is disseminated within secondary dissolution pores related to feldspar alteration (Figure 4b), or fills microfractures and interstitial spaces formed by mechanical fracturing of quartz grains (Figure 4c). Pyrite occupying the interstices of quartz and other detrital grains is commonly associated with fine-grained clay-mineral aggregates (Figure 4d).

The second type is organic-matter-associated pyrite, which shows a close spatial relationship with carbonaceous material at the microscopic scale. Pyrite is densely developed along the margins and within the micropores of carbonaceous debris and carbonaceous bands. Its morphologies include fine spherical pyrite, typical framboidal pyrite (Figure 4e), and spherical aggregates (Figure 4f). In localized microdomains where organic matter occurs in contact with carbonate cement, pyrite is also concentrated (Figure 4g).

The third type is cleavage- or fracture-hosted pyrite. Part of the pyrite occurs along mineral cleavage planes and micro-fractures, showing oriented filling or cross-cutting relationships. One subtype is distributed mainly along the cleavage planes of phyllosilicates, particularly biotite, where pyrite occurs as spherical or bead-like aggregates (Figure 4h). Another subtype fills micro-fractures and forms vein-like or dense massive aggregates that cross-cut rigid detrital grains such as quartz, with close contacts along the fracture margins (Figure 4i).

Uranium minerals are closely associated with pyrite in several textural settings. They typically occur as fine grains, veinlets, or irregular aggregates that adhere to the margins of colloidal pyrite (Figure 5a), locally forming partly enclosed or enclosed relationships with pyrite (Figure 5b). In addition, some uranium minerals fill micro-fractures within coarse-grained pyrite (Figure 5c) or occur within pyrite-rich domains (Figure 5d). These petrographic relationships indicate a close spatial association between pyrite and uranium minerals, although the exact host sites of uranium may vary at the microscale.

4.2. Geochemical Composition of Uranium Minerals Associated with Pyrite

To quantitatively determine the chemical composition of the uranium minerals intimately associated with pyrite in the Yahewan uranium deposit, in situ Electron Probe Microanalysis (EPMA) was conducted. A total of 14 valid analytical spots were obtained, with the representative results summarized in

Table 1. EPMA data of uranium minerals paragenetic with pyrite.

Sample	Spot	Analytical Results (wt.%)											Type
		UO2	SiO2	Y2O3	CaO	MgO	FeO	P2O5	Al2O3	TiO2	Others *	Total
0-9-2	1	63.84	14.36	2.36	1.25	0.02	0.19	0.63	0.87	0.3	0.33	84.15	Cof
	2	67.2	12.1	2.33	1.65	0.02	0.55	0.57	0.7	0.06	0.68	85.86
	3	52.86	12.89	2.5	1.18	0.1	7.34	0.65	1.01	7.49	0.22	86.24
	4	62.14	17.18	1.23	1.98	0.09	0.44	0.84	1.12	0.16	0.94	86.12
	5	59.11	19.81	1.48	2.21	0.31	1.36	0.82	2.21	0.24	2.9	90.45
	6	57.37	18.44	2.14	2.13	0.26	0.97	0.65	2.14	0.29	2.56	86.95
0-9-3	1	48.94	13.58	2.53	2.23	1	0.81	0.5	2.17	0.45	0.82	73.03
	2	52.03	18.44	2.47	2.18	0.64	0.94	0.72	2.97	0.49	1.16	82.04
	3	57.92	19.18	2.1	1.84	0.2	0.27	0.44	0.36	0.39	0.76	83.46
	4	43.58	17.47	1.77	2.38	0.47	0.34	0.92	3.11	0.84	2.83	73.71
	5	39.6	12.26	2.19	1.98	0.93	0.5	0.68	2.72	0.24	1.71	62.81
	6	47.51	14.08	2.75	1.27	0.1	13.41	0.91	0.18	-	0.11	80.32
	7	48.29	14.89	2.3	1.22	0.14	13.33	1.04	0.2	-	0.52	81.93
	8	47.62	14.94	2.29	1.25	0.14	13.17	1.13	0.18	-	0.67	81.39
Max		67.2	19.81	2.75	2.38	1	13.41	1.13	3.11	7.49	2.9	90.45
Min		39.6	12.1	1.23	1.18	0.02	0.19	0.44	0.18	0.06	0.11	62.81
Avg		53.43	15.69	2.17	1.77	0.32	3.83	0.75	1.42	0.99	1.20	80.73

* Others includes trace amounts of MnO, PbO, V₂O₃, Gd₂O₃, ReO₂, Sc₂O₃, CoO, NiO, Na₂O, K₂O, ThO₂, MoO₃, Cof = coffinite.

.

The analytical data reveal that the major chemical components of the studied uranium minerals are UO₂ and SiO₂. Specifically, the UO₂ content ranges from 39.60 wt.% to 67.20 wt.% (average: 53.43 wt.%), while the SiO₂ content varies between 12.10 wt.% and 19.81 wt.% (average: 15.69 wt.%). Notably, these minerals are significantly enriched in Y₂O₃ (average: 2.17 wt.%) and CaO (average: 1.77 wt.%), which may reflect the isomorphous substitution or admixture of elements such as yttrium and calcium during the formation of coffinite.

Furthermore, the FeO content exhibits considerable variation, ranging from 0.19 wt.% to 13.41 wt.% (average: 3.83 wt.%). The relatively high FeO values observed in certain analytical spots (up to 13.41 wt.%) are attributed to the excitation of micro-scale pyrite inclusions or intergrowths within the coffinite matrix during microbeam analysis. This geochemical observation corroborates the intimate spatial association between coffinite and pyrite observed under SEM-BSE imaging. Trace amounts of TiO₂, MgO, P₂O₅, and MoO₃ were also detected. Based on the major element chemistry and morphological characteristics, the primary uranium mineral phase in this deposit is confidently identified as yttrium-bearing coffinite.

It is noteworthy that the analytical totals for all measured spots are consistently low, ranging from 62.81 wt.% to 90.45 wt.% (average: 80.73 wt.%). This mass deficit is a well-documented phenomenon in the EPMA of coffinite. The analytical shortfall is primarily attributed to the presence of structural water or interlayer hydroxyl groups (OH⁻) incorporated within the crystal lattice, which are undetectable by electron microprobe analysis. Additionally, physical factors such as internal microporosity, commonly caused by radiation-induced metamictization, and surface roughness of the analyzed domains may also have lowered the analytical totals.

4.3. Elemental Characteristics of Pyrite

In situ LA-ICP-MS trace-element analyses were conducted on pyrite grains from light grayish-white pebbly coarse sandstones in the study area. The analytical results are summarized in Table S2.

4.3.1. Major Element Characteristics

The Fe content of the analyzed pyrite ranges from 44.90 to 50.26 wt.% (average: 47.51 wt.%), whereas the S content ranges from 46.42 to 53.01 wt.% (average: 51.19 wt.%). Minor Si and Mn signals were also detected in some analyses. The Si content reaches up to 1.99 wt.%, suggesting that some LA-ICP-MS spots likely sampled quartz, clay minerals, coffinite-like U-silicate phases, or mixed pyrite–mineral microdomains. These spots are therefore interpreted as pyrite-rich microdomain analyses rather than pure pyrite compositions. The Mn content ranges from 0.006 to 1.74 wt.%, indicating compositional heterogeneity among the analyzed pyrite-rich domains.

4.3.2. Trace Element Characteristics

The analytical results reveal broad variations in U, As, Mo, and other trace elements among the analyzed pyrite-rich domains. Uranium shows the most pronounced sample-scale heterogeneity. In the mineralized sample 0-9-2, all six LA-ICP-MS spots show elevated U contents, ranging from 212.04 to 37,915.74 ppm, with an average of 7841.75 ppm. By contrast, U contents in the remaining samples are much lower, ranging from 0.03 to 131.23 ppm, with an average of 11.33 ppm. This distribution indicates that U enrichment is spatially concentrated in pyrite-rich microdomains of sample 0-9-2. However, because several high-U spots are accompanied by lithophile-element signals, these analyses should not be interpreted as direct evidence for lattice-bound U in pyrite. Instead, they record close spatial coupling between pyrite-rich domains and coffinite or coffinite-like U-silicate mineralization. In addition, As contents range from 66.69 to 7183.15 ppm, with an average of 1209.29 ppm, whereas Mo contents vary from 9.58 to 720.78 ppm, with an average of 219.47 ppm.

Regarding iron-group transition metals, the overall concentration of nickel (Ni) is slightly higher than that of cobalt (Co). The Ni content ranges from 13.33 to 2565.57 ppm (average: 444.38 ppm), while the Co content ranges from 2.09 to 1384.17 ppm (average: 243.28 ppm). Among common chalcophile base metals, the average concentrations of lead (Pb) and copper (Cu) are 368.87 ppm and 230.60 ppm, whereas zinc (Zn) shows relatively lower concentrations (average: 28.98 ppm). In addition, several minor associated elements were detected in the analyzed pyrite-rich domains. Sb and Tl were detected in most analyses, with average concentrations of 22.39 ppm and 12.25 ppm, respectively. Se was detected only in part of the analytical spots, with an average concentration of 33.85 ppm for the available data. Minor Cd and Bi were also detected in some analyses.

Furthermore, the analyses detected several lithophile elements and large ion lithophile elements (LILEs) as trace impurities. These include barium (Ba; average: 272.89 ppm), strontium (Sr; average: 156.20 ppm), and zirconium (Zr; average: 48.51 ppm), along with trace amounts of vanadium (V), chromium (Cr), and thorium (Th). Notably, certain analytical spots exhibited distinct localized high-value anomalies for these specific elements.

4.3.3. Rare Earth Element Characteristics

The analytical results and characteristic parameters of rare earth elements (REEs) in pyrite from the Yahewan uranium deposit are summarized in

Table 2. Rare earth element compositions of pyrite.

Sam	0-9-2						0-9-3	86-4-6	86-4-7
	Ore sample						Ore sample	Ore sample	Barren sample
Dep	680 m						681 m	318 m	322 m
Stra	J2z1
Lith	Light grayish-white pebbly coarse sandstone
Element	1-1	1-2	1-3	1-4	1-5	1-6	2-3	3-1	4-1	4-2	4-3
La	2.31	22.40	0.56	6.02	306.89	0.57	0.38	3.57	22.37	5.56	1.24
Ce	5.90	114.84	2.68	35.78	992.67	6.67	2.14	7.91	53.92	9.81	3.14
Pr	1.47	17.07	0.57	5.92	248.59	1.30	0.51	1.06	6.64	0.99	0.31
Nd	4.30	70.33	1.69	22.66	1073.82	3.62	2.87	4.39	28.03	3.94	0.88
Sm	1.40	20.85	0.53	5.95	319.70	0.86	1.42	0.59	8.83	0.95	0.22
Eu	0.36	4.51	0.15	1.89	84.58	0.48	0.24	0.17	3.96	0.30	0.06
Gd	1.28	18.87	0.56	7.13	337.06	0.89	1.50	0.64	18.40	0.68	0.18
Tb	0.40	3.40	0.11	1.05	63.24	0.20	0.18	0.09	4.05	0.12	0.03
Dy	1.77	23.13	0.74	10.64	390.54	1.29	1.14	0.51	33.24	0.60	0.20
Ho	0.30	4.16	0.11	1.40	74.26	0.35	0.15	0.08	8.81	0.18	0.06
Er	1.02	11.95	0.32	4.18	204.71	0.89	0.87	0.24	30.45	0.44	0.13
Tm	0.17	1.98	0.05	0.55	30.80	0.16	0.07	0.03	4.67	0.06	0.02
Yb	0.74	10.57	0.18	4.15	177.98	0.64	0.92	0.17	33.47	0.31	0.11
Lu	0.12	1.55	0.03	0.51	24.94	0.28	0.06	0.01	5.31	0.06	0.02
Y	5.89	83.10	2.22	33.45	1394.78	5.96	6.59	1.93	321.28	3.63	1.33

Note: The REE contents of analyses 2-1 and 2-2 from sample 0-9-3, analyses 3-2 and 3-3 from sample 86-4-6, and all analyses from sample 106-1-5 are generally low; these data are therefore not listed or included in the statistical calculations.

, and the chondrite-normalized REE patterns are illustrated in Figure 6. REE values of zero or below the detection limit were omitted from the plotted lines and statistical calculations. The normalizing chondrite values are adopted from Sun and McDonough [25].

The results demonstrate that the total REE contents (ΣREE) of the pyrite grains exhibit profound variability, varying from 6.60 to 4329.79 ppm, which indicates highly heterogeneous enrichment at the microscale. Regarding the fractionation between light and heavy rare earth elements (LREEs and HREEs), the ΣLREE/ΣHREE ratios range from 0.89 to 9.99 (average: 4.17). After chondrite normalization, the (La/Yb)_N ratios vary from 0.28 to 14.27 (average: 3.93). The overall chondrite-normalized REE patterns display a right-dipping trend, characterized by relative LREE enrichment and HREE depletion.

Eu and Ce anomalies are commonly used as qualitative indicators of fluid–rock interaction, redox heterogeneity, and microdomain-scale geochemical variation [26]. These anomalies were calculated using the standard equations: Eu/Eu* = Eu_N/(Sm_N × Gd_N)^1/2 and Ce/Ce* = Ce_N/(La_N × Pr_N)^1/2. Analytically, the europium anomaly index (Eu/Eu*) of the studied pyrite ranges from 0.50 to 1.67 (average: 0.91). The REE patterns predominantly display a weak negative Eu anomaly, accompanied by localized positive anomalies at certain analytical spots. In contrast, the cerium anomaly index (Ce/Ce*) ranges from 0.77 to 1.87 (average: 1.18), indicating mostly weak to moderate positive Ce anomalies and locally pronounced positive anomalies on the normalized curves.

4.4. Sulfur Isotope Characteristics of Pyrite

The δ³⁴S_V-CDT values of the studied pyrite exhibit a broad range and a high degree of heterogeneity, varying from −27.49‰ to +36.06‰, yielding a maximum absolute variation of 63.55‰ (

Table 3. Sulfur isotope compositions of pyrite.

Sample	Spot	δ34SV-CDT (‰)	Sample	Spot	δ34SV-CDT (‰)
0-9-2	1	0.29	86-4-7	1	2.10
	2	−17.01		2	−9.61
86-4-6	1	−17.87		3	1.97
	2	−24.22		4	−20.97
	3	−19.63		5	−22.45
	4	13.69	106-1-5	1	34.87
	5	13.78		2	35.46
	6	13.34		3	36.06
	7	13.40		4	27.07
	8	11.96		5	30.65
	9	−27.49		6	31.88
	10	−26.20		7	20.75
	11	−12.14		8	28.84
	12	−16.54		9	32.48
	13	−18.74		10	35.10
	14	−5.64		11	17.53
	15	−14.64		12	6.87
	16	−19.05	max	36.06
	17	−22.74	min	−27.49
	18	−26.50

). The average measured abundances for ³²S, ³³S, and ³⁴S are 9.57, 0.0792, and 0.459, with their distribution ranges being 5.12–12.34, 0.0422–0.101, and 0.245–0.578, respectively. Furthermore, the average isotopic ratios for ³⁴S/³²S and ³³S/³²S are calculated to be 0.0480 and 0.0083, respectively, confirming the high reliability of the analytical data.

5. Discussion

5.1. Trace Element Constraints on Pyrite Genesis and Ore-Forming Fluid Properties

In modern sediments and Precambrian syngenetic sedimentary deposits, pyrite typically exhibits low cobalt (Co) concentrations (less than 100 ppm), with Co contents generally lower than nickel (Ni) levels [3]. Conversely, pyrite associated with intrusive and extrusive complex systems shows significantly higher Co and Ni concentrations, often exceeding 1000 ppm. Pyrite in volcano-sedimentary deposits typically has higher Co contents than those in normal sedimentary deposits, with Co/Ni ratios ranging between 5 and 100 [27,28,29]. Furthermore, hydrothermal pyrite also tends to have relatively high Co concentrations, mostly exceeding 100 ppm and commonly higher than Ni contents.

Co/Ni ratios in pyrite are commonly used to distinguish different genetic types, but their interpretation should be combined with petrographic and geochemical evidence. In the study area, most pyrite analyses show Co/Ni ratios below 1, consistent with a mainly sedimentary–diagenetic or authigenic origin under reducing microenvironments associated with organic matter and early pyrite formation [29,30]. A few coarse-grained pyrite domains show Co/Ni ratios near or slightly above 1, together with locally elevated V and U contents. These domains are not interpreted as unequivocal high-temperature hydrothermal pyrite. Instead, they are regarded as fluid-modified pyrite-rich domains, possibly related to later basinal fluid overprinting or local interaction with U-bearing microfractures. The similar trace-element signatures of these domains may indicate modification during a comparable later fluid-flow stage.

Pyrite is typically enriched in iron-group transition elements such as Co and Ni. Due to their similar atomic structures and physicochemical properties under natural conditions, Co²⁺ and Ni²⁺ commonly substitute for Fe²⁺ within the pyrite lattice via isomorphous substitution, forming CoS₂ and NiS₂ solid solutions [31]. The Co/Ni ratio in pyrite is a sensitive indicator of the specific geological setting and ore-forming fluid properties. In general, relatively high Co/Ni ratios may reflect non-sedimentary or fluid-modified pyrite, whereas low Co/Ni ratios are commonly associated with sedimentary–diagenetic or authigenic pyrite. However, Co/Ni ratios should be interpreted together with petrographic textures and other geochemical evidence.

Based on the genetic classification established by Bralia et al. [32], sedimentary-diagenetic pyrite typically yields Co/Ni ratios below 1.00. In contrast, magmatic-hydrothermal pyrite generally clusters between 1.00 and 5.00, while metamorphic-hydrothermal variants hover around 1.00. Volcanogenic massive sulfide (VMS) deposits exhibit the most extreme values, with ratios ranging from 5.00 to 50.00, characterized by high Co (avg. 480 ppm) and low Ni (<100 ppm).

In the study area, the pyrite displays significant variations in Co and Ni concentrations, with Co ranging from 2.09 to 1384.17 ppm (avg. 243.28 ppm) and Ni ranging from 13.33 to 2565.57 ppm (avg. 444.38 ppm). The Co/Ni ratios vary from 0.03 to 1.31, with an average of 0.55. Although the pyrite is characterized by high absolute abundances of both Co and Ni, the overall Co/Ni ratios remain relatively low. When plotted on the Co-Ni genetic discrimination diagram (Figure 7), the majority of the data points fall within the sedimentary-diagenetic field (Co/Ni < 1), yet exhibit a distinct distributional trend toward the fluid-modified field (Co/Ni ≈ 1). A small subset of analyses with Co/Ni ratios near or slightly above 1 may reflect later fluid modification of pyrite-rich domains [33,34,35].

Overall, the low average Co/Ni ratio and the common association of pyrite with organic matter support a mainly authigenic to sedimentary-diagenetic origin for most pyrite. However, a few pyrite-rich domains with Co/Ni ratios near or slightly above 1 and locally elevated U contents indicate modification during later fluid flow. These observations support a two-stage interpretation in which early pyrite formed a reduced diagenetic substrate that was subsequently overprinted by metal-bearing basinal fluids during uranium mineralization.

Trace-element covariation diagrams further illustrate the heterogeneous distribution of U and associated elements in the analyzed pyrite-rich domains (Figure 8). Some high-U analyses are accompanied by enrichment in redox-sensitive or lithophile elements, indicating localized geochemical heterogeneity within pyrite-rich microdomains. The repeated occurrence of high U in all analyzed spots from sample 0-9-2 suggests a real sample-scale U enrichment feature rather than a sporadic analytical artifact. However, the coexistence of high U with elevated Si and lithophile elements in some spots indicates that part of the U signal may be related to coffinite or coffinite-like U-silicate micro-inclusions, fracture-hosted uranium minerals, or mixed pyrite–uranium mineral ablation. Therefore, these data are best interpreted as evidence for later U-bearing fluid overprinting of pyrite-rich microdomains, rather than direct proof of lattice-bound U in pyrite.

The iron-group transition elements cobalt (Co) and nickel (Ni) represent the early sedimentary-diagenetic background. Co and Ni demonstrate an exceptionally strong internal correlation (ρ = 0.84), with data points highly clustered in scatter plots. This reflects their shared genetic origin and their broadly similar behavior during early pyrite crystallization.

In stark contrast, this background assemblage shows a negligible correlation with uranium (e.g., ρ = 0.09 for U vs. Ni). This geochemical decoupling suggests that the local U enrichment was not simply inherited from the early sedimentary–diagenetic pyrite-forming system, but was likely related to later U-bearing fluid overprinting with distinct chemical characteristics. The spatial convergence and overprinting of these two fluid end-members—each with distinct physicochemical properties—ultimately shaped the highly heterogeneous trace element patterns observed in the Yahewan pyrite.

Arsenic (As) exhibits significant enrichment and strong elemental coupling within the pyrite of the Yahewan uranium deposit. Quantitative correlation results between As, the major element Fe, and the key ore-forming element U were obtained through Spearman rank correlation analysis of in situ LA-ICP-MS data (Figure 9).

The scatter plots show positive covariation between As and U in some pyrite-rich domains (Figure 9a), suggesting that As enrichment may be associated with U-rich microdomains. As shows positive covariation with U, Mo, V, Se, and Pb in some pyrite-rich domains, suggesting localized enrichment of these elements during later fluid overprinting.

In the As-Fe correlation plot (Figure 9b), the relationship between As and Fe should be interpreted cautiously because Fe is the major structural component of pyrite and shows a relatively narrow concentration range. The enrichment of As in some U-rich pyrite domains may indicate coupled incorporation or micro-inclusion-related enrichment, but its lattice-bound versus inclusion-related occurrence requires further verification.

REEs can provide qualitative constraints on fluid-rock interaction and microdomain-scale geochemical heterogeneity during pyrite formation [36]. The REE signatures of the Yahewan pyrite provide qualitative information on fluid–rock interaction and microdomain-scale geochemical heterogeneity.

Unlike typical deep-seated magmatic-hydrothermal deposits, the studied pyrite exhibits a right-dipping REE pattern (ΣLREE/ΣHREE ≈ 4.17) with weak negative to localized positive Eu anomalies, aligning with inter-formational groundwaters that have undergone extensive water-rock interaction. Furthermore, the locally positive Ce anomalies may reflect redox heterogeneity during pyrite precipitation, although possible effects of micro-inclusions, mixed mineral phases, or fluid–rock interaction should also be considered.

The studied pyrite displays variable Eu and Ce anomalies, including localized positive Eu anomalies with Eu/Eu* values up to 1.67. These features suggest heterogeneous REE behavior within pyrite-rich microdomains. In low-temperature sandstone-hosted systems, localized positive Eu anomalies are tentatively attributed to fluid–rock interaction, inheritance from micro-inclusions or mineral intergrowths, and variable Eu complexation in basinal fluids. Therefore, Eu and Ce anomalies are used here as qualitative indicators of micro-domain heterogeneity and fluid–mineral interaction, rather than direct evidence for Eu²⁺ incorporation into the pyrite lattice or intensely reducing conditions.

The ore-hosting Zhiluo Formation contains abundant carbonaceous debris, organic matter, and early pyrite, which could have provided local reducing microenvironments favorable for uranium precipitation. The relatively lower Fe³⁺/Fe²⁺ ratios in mineralized intervals are consistent with more reducing conditions, although the redox state of the system likely varied at the microdomain scale. These features suggest that pyrite-rich and organic-rich domains acted as favorable sites for later fluid–mineral interaction during uranium mineralization.

5.2. Constraints of Sulfur Isotopes on Ore-Forming Material Sources and Mineralization Environments

The 37 in situ sulfur isotope (δ³⁴S_V-CDT) data obtained from pyrite in this study show a broad range and pronounced isotopic heterogeneity (Figure 10). The analytical results show that the δ³⁴S_V-CDT values vary from −27.49‰ to +36.06‰, yielding a substantial total range of 63.55‰ and a mean value of +2.34‰. Specifically, negative values account for 45.9% of the data (descending to −27.49‰), while positive values constitute 54.1% (peaking at +36.06‰). Such variability in sulfur isotopic composition differs from typical deep-seated magmatic-hydrothermal sulfur, which commonly falls within a narrow range around 0 ± 3‰ [29]. This contrast provides geochemical evidence for evaluating the sulfur sources and mineralization environment.

As shown in Figure 11, the sulfur isotopic (δ³⁴S) compositions of different natural reservoirs vary significantly. Mantle-derived or magmatic-hydrothermal sulfur is characterized by its stability within a narrow interval of 0 ± 3‰.

In contrast, modern sediments, coal, petroleum, and sedimentary sulfides influenced by bacterial sulfate reduction (BSR) often display a broad range of sulfur isotopes—frequently biased toward negative values—due to intense kinetic fractionation. Projecting the 37 measured pyrite data points from this study (ranging from −27.49‰ to +36.06‰) onto this diagram yields the following implications [37,38] (Figure 11).

The isotopic distribution of the studied pyrite overlaps with the ranges of sedimentary sulfides, coal, and biogenic modern sediments, while differing from the relatively narrow mantle-derived sulfur range near 0‰. This pattern suggests that sedimentary sulfate, organic sulfur reservoirs, and biogenic sulfur cycling within the host strata contributed to the sulfur budget of pyrite [39,40]. The prevalence of negative δ³⁴S values, particularly those between −27.49‰ and −17.00‰, is consistent with bacterial sulfate reduction under organic-rich pore-water conditions.

Consequently, the sulfur in the Yahewan pyrite was likely derived mainly from sedimentary sulfate and organic sulfur reservoirs within the host strata. These organic- and pyrite-rich sandstones likely provided reducing microenvironments that favored later interaction with U-bearing fluids [41,42,43].

The wide range of δ³⁴S_V-CDT values in pyrite indicates heterogeneous sulfur sources and multi-stage isotope fractionation under variable diagenetic to mineralizing conditions. The negative δ³⁴S_V-CDT values are consistent with sulfur isotope fractionation related to bacterial sulfate reduction, probably promoted by organic matter in relatively open to semi-open pore-water systems. In contrast, the positive δ³⁴S_V-CDT values, reaching up to +36.06‰, may reflect restricted pore-water evolution, residual sulfate enrichment, Rayleigh-type fractionation, or multi-stage pyrite growth. However, these processes cannot be fully distinguished using sulfur isotope data alone. Therefore, the sulfur isotope data are interpreted here as evidence for heterogeneous sulfur sources and evolving pore-water conditions, rather than as unique evidence for a strongly reducing mineralization environment. Together with the petrographic association of pyrite, organic matter, and uranium minerals, these isotopic features suggest that reducing microenvironments were locally developed in the sandstone and may have favored uranium precipitation during later fluid overprinting.

5.3. Implications of Pyrite for Uranium Mineralization

In sandstone-hosted uranium deposits, pyrite commonly acts as a reducing mineral and an indicator of redox conditions within the diagenetic-mineralizing system. The wide δ³⁴S_V-CDT range of pyrite, from −27.49‰ to +36.06‰, records heterogeneous sulfur sources and multi-stage sulfur isotope fractionation. Negative δ³⁴S_V-CDT values are consistent with bacterial sulfate reduction in organic-rich pore waters, whereas positive values may reflect restricted pore-water evolution, residual sulfate enrichment, Rayleigh-type fractionation, or multi-stage pyrite growth. Together with the close association between pyrite and organic matter, these sulfur isotope features suggest that reducing microenvironments were locally developed in the Zhiluo Formation sandstones and may have provided favorable sites for later interaction with U-bearing fluids.

Upon the influx of later U-bearing basinal fluids, the reducing microenvironments provided by early pyrite and organic matter may have favored localized uranium precipitation. The trace-element data suggest that some pyrite-rich domains were overprinted by later U-bearing fluids during uranium mineralization. The repeated occurrence of high U in sample 0-9-2 indicates localized U enrichment at the sample scale, whereas the association of U with Mo, V, W, Se, and As is consistent with interaction between pyrite-rich reducing domains and metal-bearing basinal fluids. However, because some high-U analyses may include contributions from coffinite or coffinite-like U-silicate micro-inclusions, fracture-hosted uranium minerals, or mixed pyrite–uranium mineral ablation, these data are best interpreted as records of U-bearing fluid overprinting of pyrite-rich microdomains rather than direct evidence for uranium incorporation into the pyrite lattice.

At the microscale, early pyrite and organic matter likely provided reducing microenvironments favorable for uranium precipitation when later U-bearing fluids entered the sandstone. The close spatial association between pyrite and coffinite or coffinite-like U-silicate minerals suggests that pyrite-rich domains acted as important sites for localized fluid–mineral interaction. In these microdomains, reduced sulfur species, ferrous iron, and organic matter may have contributed to the reduction of mobile U(VI) and the precipitation of U(IV) minerals. This process may have involved localized consumption of reduced components and limited material exchange within pore-scale reaction sites, which is broadly consistent with the heterogeneous and locally positive δ³⁴S_V-CDT values observed in some pyrite domains. Therefore, pyrite records, and may have partly mediated, the redox transition between early reducing pore waters and later U-bearing fluids. However, the exact timing of pyrite growth and uranium mineralization requires further geochronological and valence-state constraints.

6. Conclusions

• Pyrite in the Zhiluo Formation sandstones of the Yahewan deposit occurs mainly as framboidal, colloidal, pore-filling, fracture-filling, and massive aggregates. Its textures, close association with organic matter, and mostly low Co/Ni ratios indicate a dominant authigenic to sedimentary–diagenetic origin under reducing pore-water conditions. A few pyrite-rich domains with Co/Ni ratios near or slightly above 1 may record later fluid modification.
• Coffinite or coffinite-like U-silicate minerals occur along pyrite margins, within microfractures, and locally in pyrite-rich microdomains, indicating a close spatial association between uranium minerals and pyrite. Local enrichment of U, Mo, V, W, Se and As is consistent with U-bearing fluid overprinting of pyrite-rich domains, rather than direct proof of lattice-bound U in pyrite.
• The wide δ³⁴S_V-CDT range of pyrite (−27.49‰ to +36.06‰) records heterogeneous sulfur sources and multi-stage isotope fractionation, likely involving bacterial sulfate reduction and restricted pore-water evolution. These isotopic signatures indicate locally developed organic- and pyrite-rich reducing microenvironments that favored later interaction with U-bearing fluids.
• The main contribution of this study is to show that pyrite-rich microdomains record the transition from early diagenetic reduction to later uranium mineralization. Early organic matter and authigenic/biogenic pyrite provided local reducing sites, whereas later U-bearing basinal fluids overprinted these sites and promoted localized coffinite or coffinite-like U-silicate precipitation.