Geochronology and Genesis of the Carboniferous Shikebutai Iron Deposit in Western Tianshan, Northwestern China

Abstract

Submarine volcanic-hosted iron oxide deposits are critical archives for reconstructing the interplay between hydrothermal activities and marine redox conditions, yet the genesis of these deposits remains controversial. Here, we present a comprehensive geochronological and geochemical study on the Shikebutai iron deposit in the Western Tianshan, northwestern China, to constrain the mineralization age, the source of iron, and deposit genesis. The stratiform-to-lenticular orebodies are hosted within the Late Carboniferous marine volcanic–sedimentary sequence of the Yishijilike Formation. The iron ores consist primarily of hematite and quartz, with minor siderite and barite, exhibiting massive to locally banded textures. SHRIMP zircon U-Pb dating of the overlying andesite yields an age of 315.8 ± 1.5 Ma, consistent with the Sm–Nd isochron age of the iron ore samples (319 ± 26 Ma), precisely constraining the mineralization age to the Late Carboniferous (ca. 315–320 Ma). The geochemical compositions of the iron ore samples indicate negligible syn-depositional detrital contamination, as evidenced by low Al₂O₃ (<1.00 wt%) and TiO₂ (<0.20 wt%) contents. Low abundances of trace elements, including Sr (0.33–31.18 ppm), Hf (0.05–1.77 ppm) and Rb (1.49–39.02 ppm), further support the minimal detrital influence. Geochemical signatures, such as pronounced positive Eu anomalies (Eu/Eu = 1.62–7.12, mean 4.14), LREE enrichment ((La/Yb) _(PAAS) = 0.58–4.78), and near-chondritic Y/Ho ratios (mean 28.5), suggest a significant high-temperature (>250 °C) hydrothermal contribution. Moreover, the ε_Nd(t) values of iron ore samples (+1.99 to +2.93) are comparable to those of coeval andesites (+2.75 to +3.44) but exceed those of associated metasiltstones (+0.41 to +0.95), suggesting that ore-forming materials were derived from hydrothermal fluids leaching juvenile crust. The Shikebutai iron deposit exhibits geochemical and mineralogical similarities to modern Red Sea and East Pacific Rise metalliferous sediments, establishing the deposit as a product of active vent-proximal hydrothermal systems rather than marine chemical sediments such as banded iron formations.

1. Introduction

Marine Fe(III) oxyhydroxide sediments related to hydrothermal activity are widespread in modern seafloor settings [1,2,3,4]. Some ancient stratiform submarine volcanic–hydrothermal sedimentary Fe-oxide deposits, composed predominantly of hematite, are considered to be chemical sedimentary rocks that exhibit similarities to these modern Fe(III) oxyhydroxide precipitates [5,6,7,8]. Marine Fe-oxide deposits represent critical geochemical archives that can be utilized to reconstruct iron sources [9,10,11,12]. Representative examples include the Middle to Late Devonian Lahn-Dill-type iron deposit in the Rhenish Massif of Germany [13,14], the Ordovician iron ores hosted in the Løkken ophiolite of central Norway [9], and the Mesoproterozoic Jingtieshan iron deposit in the North Qilian region of northwestern China [15].

In the Western Tianshan Orogenic Belt of northwestern China, the Awulale Fe-oxide metallogenic belt hosts several submarine volcanic-hosted Fe-oxide deposits [16,17,18], most notably the Shikebutai hematite deposit and the Motuosala iron–manganese deposit, both of which formed within the marine volcanic sequences [19]. However, the genesis of submarine volcanic-hosted hematite deposits associated with sedimentary processes, such as the Shikebutai deposit, remains highly controversial and poorly understood. Currently, two primary genetic models have been advanced for the Shikebutai deposit: (1) formation through direct iron-rich hydrothermal precipitation induced by submarine volcanic activity [20,21,22,23]; and (2) precipitation from mixed seawater–hydrothermal fluids within anoxic to suboxic depositional basins, with formation processes analogous to those of Precambrian banded iron formations (BIFs) [24,25].

The Shikebutai iron deposit, located at the westernmost end of the Awulale Fe-oxide metallogenic belt, is hosted in Carboniferous volcano-sedimentary sequences and comprises stratiform-to-lenticular orebodies [13,20,26]. The ore exhibits predominantly massive textures and is dominated by hematite, quartz, siderite, and barite [16,17], with some samples displaying alternating iron- and silica-rich layers that closely resemble those in Precambrian BIFs [19]. While previous studies have characterized the deposit using Fe-O-C-Si isotopes and mineral chemistry [22,24,25], several critical questions remain unresolved. First, the precise timing of mineralization is still poorly constrained, as previous age estimates rely on the ages of host volcanics rather than direct dating of the ore itself. Second, the relative contributions of hydrothermal fluids versus seawater to the iron budget have not been quantified, and the provenance of iron-bearing fluids has not been directly constrained using Nd isotopic systematics. Third, the genetic affinity of the Shikebutai deposit with Precambrian BIFs versus modern vent-proximal hydrothermal systems has not been evaluated through direct comparison of geochemical signatures. To address these gaps, this study provides the first Sm–Nd isochron age obtained directly from the Shikebutai iron ores, which, combined with high-precision SHRIMP zircon U-Pb dating of the overlying andesite, establishes a more stringent age bracket for the mineralization. By integrating these new geochronological constraints with a systematic Nd isotope and trace element study, we aim to: (1) precisely define the mineralization age; (2) quantitatively constrain the relative contributions of hydrothermal fluids and seawater to iron mineralization; and (3) evaluate whether the Shikebutai deposit represents a Phanerozoic analogue of active vent-proximal hydrothermal systems or a chemical sedimentary sequence genetically comparable to that of Precambrian BIFs.

2. Geologic Background

The Chinese Western Tianshan Orogenic Belt is located on the southwestern part of the Central Asian Orogenic belt, which is sandwiched between the Tarim Craton and the Junggar Terrane (Figure 1a). From south to north, this orogenic belt is composed of the South and North Margins of the Tarim Craton, the Central Tianshan Terrane, and the Yili Block (Figure 1b) [27,28]. The Western Tianshan Orogenic belt has undergone complex tectonic evolution (Figure 1b), beginning with the accretion and breakup of Paleo-to-Neoproterozoic supercontinents. This was followed by the collision of multi-continental microplates and island arcs within the ancient Asian Ocean during the Early Paleozoic [28,29]. Subsequently, the Paleozoic ocean between the Junggar terrane and the Tarim craton closed in the Late Paleozoic, leading to post-collisional extension during the Mesozoic [18,27,28].

Carboniferous volcano-sedimentary strata, including the lower Dahalajunshan, middle Akeshake, and upper Yishijilike formations, are widely distributed along the southeastern margin of the Yili Block and Central Tianshan Terrane. These sequences represent the primary host strata for the regional Fe deposits (Figure 1b) [24,30]. The Lower Carboniferous Dahalajunshan Formation is composed of marine intermediate-acid volcanic rocks and pyroclastic rocks with sedimentary rocks. The Dahalajunshan Formation consists mainly of andesite, rhyolite and andesitic tuff, with a small amount of limestone and sandstone. In addition, many magmatic–hydrothermal magnetite deposits, such as Beizhan, Dunde, Zhibo, Chagangnuoer, Nixintage and Songhu iron deposits, are found in this stratum (Figure 1b) [18]. The Lower Carboniferous Akeshake Formation is composed of clastic rocks, limestone, siliceous rocks and a small amount of volcanic rocks, which produced exhalative sedimentary iron–manganese deposits (e.g., the Motuosala deposit) and sedimentary manganese deposits (e.g., the Zhaosu manganese deposit) [31]. The Upper Yishijilike Formation is composed of submarine–subaerial intermediate-acid volcanic rocks and volcaniclastic sedimentary rocks mixed with sedimentary rocks, with the major lithologies being andesite, rhyolite, volcanic breccia and andesitic tuff, partially intercalated with metasiltstones [24]. The Shikebutai hematite deposit studied occurs in this stratum. In addition to the iron and manganese deposits hosted in the Carboniferous strata, a series of copper, zinc and gold deposits are also distributed in the surrounding area [17]. Additionally, it is widely suggested that the Carboniferous volcanic–sedimentary strata in the Western Tianshan Orogenic belt were deposited in a subduction-related arc tectonic setting, with sedimentary iron and manganese deposits likely formed within extensional back-arc basins [17,19].

3. Ore Deposit Geology

The Shikebutai iron deposit (43°33′ N, 83°37′ E) is located in the western section of the Carboniferous volcanic–sedimentary sequences in the Western Tianshan Orogenic belt (Figure 1b), with reserves of ~28.39 Mt and an average grade of 56% total Fe (TFe), and occurs within the upper Yishijilike Formation [20,32]. The Yishijilike Formation is subdivided into four major lithological members (Members 1–4; mapped as C₂Y¹–C₂Y⁴) from the bottom to the top, with the second member being the ore-bearing horizon (Figure 2). Member 1 (C₂Y¹) consists of rhyolite, andesite, andesitic tuff, tuff, with the top transitioning to tuffaceous metasandstone and tuffaceous phyllite. Member 2 (C₂Y²) is characterized by chlorite–sericite phyllite, ferruginous sericite phyllite, iron orebody, metasiltstone, sericite–quartz schist, tuffaceous phyllite and andesitic tuff from bottom to top, which is the main ore-bearing horizon and mining locality. Member 3 (C₂Y³) is composed of andesite, andesitic tuff, andesitic volcanic breccia and volcanic agglomerate. Member 4 (C₂Y⁴) is dominated by andesite, trachyandesite, trachyte and sedimentary volcanic breccia, and the bottom is composed of complex volcanic breccia, which is absent in the study area [26,32].

The Shikebutai iron deposit is situated on the southern limb of an east–west-trending syncline structure, with strata dipping southward at angles ranging from 60° to 80°. The iron orebodies are structurally controlled by east–west-trending faults, which are locally crosscut by later northeast- and northwest-trending faults. The orebodies extend approximately 4.6 km along strike and 1.3 km in width (Figure 2). Three major orebody zones have been identified along the east–west strata, classified from north to south as follows: (1) the Main and Eastern, (2) Western and Southwestern, and (3) South and Southeastern zones. The main orebody is characterized by layers and lenses of hematite ore hosted within ore-bearing horizons near the contact between metasiltstone and andesite (the hanging wall of the orebodies) and ferruginous sericite phyllite (the footwall of the orebodies), in association with chlorite–sericite phyllite, sericite–quartz schist, and tuffaceous phyllite (Figure 3). The orebody is subdivided into 14 ore layers, extending 1000 m east–west and 100 m north–south. Among these, three layers are notably thick, extending 400 to 600 m along strike, with a maximum thickness of ~24 m and an average thickness of ~8 m. These orebodies generally pinch out along the east–west strike and thin gradually at depth. In contrast, other ore zones contain fewer ore layers with relatively smaller thicknesses (maximum ~10 m; average ~2 m). The continuity of these ore layers along the strike and dip is relatively variable [19,20,26].

4. Sampling and Analytical Techniques

4.1. Sample Preparation

Field observations were conducted primarily on the open pit of the Shikebutai iron deposit (Figure 3). Forty samples were collected from iron ores and associated rocks (andesite and metasiltstone). During sampling, we specifically targeted fresh specimens and avoided obvious fault zones or highly altered features to ensure data reliability. A detailed petrographic examination, using both transmitted and reflected light, was carried out to determine mineral assemblages and the paragenesis of all samples. Polished thin sections were examined using scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (SEM-EDS) at the Institute of Geology and Geophysics, Chinese Academy of Sciences (IGGCAS).

4.2. SHRIMP Zircon U-Pb Geochronology

Zircon grains were separated from one andesite sample (19SW-11) located in the hanging wall of the iron orebody, using conventional magnetic and density techniques. These grains were handpicked under a binocular microscope to ensure purity and cast in an epoxy mount along with Plešovice (PLE) zircon standards. They were then polished to expose the center of the grains. Following this step, photographs were taken in both reflected and transmitted light, and to further reveal their internal structures, cathodoluminescence (CL) images were gained using a TESCAN GAIA3 scanning electron microscope (SEM) (TESCAN ORSAY HOLDING, Brno, Czech Republic) at the Beijing Research Institute of Uranium Geology. Thereafter, mounts were cleaned with deionized water and alcohol, and vacuum-coated with gold prior to secondary ion mass spectrometry (SIMS) analyses.

Measurements of zircon U, Th, and Pb isotopes were conducted using a Sensitive High-Resolution Ion Micro Probe (SHRIMP II) (Australian Scientific Instruments, Canberra, Australia) at the Beijing SHRIMP Center, Chinese Academy of Geological Sciences. Procedures and conditions were referred to those described by Williams [33]. The intensity of the primary O²⁻ ion beam was ~3.5 nA. The spot size was ~25 μm, with 120 s raster prior to each isotopic analysis, and the total measurement time for one spot was about 12 min. The reference zircon Plešovice (²⁰⁶Pb/²³⁸U age = 337.1 Ma, [34]) was analyzed for every third analysis to calibrate the Pb/U ratios. During the session, ten analyses of the PLE standard yielded a weighted mean ²⁰⁶Pb/²³⁸U age = 337.1 ± 2.1 Ma (MSWD = 1.6), shown in Supplementary Table S1, which is in excellent agreement with its recommended value. A common Pb correction was applied using the measured ²⁰⁴Pb abundances and model Pb compositions of Cumming and Richards [35]. Data processing and assessment were carried out using the SQUID 2.50 and ISOPLOT 4.15 programs [36]. The uncertainties for individual analyses on the concordia diagrams are presented at the 1σ level, whereas uncertainties for the final weighted mean ages are quoted with a 95% confidence interval.

4.3. Major and Trace Element Analysis

Major and trace element analyses were measured at the Wuhan Sample Solution Analytical Technology Co., Ltd., Wuhan, China. Major oxides were analyzed with a Zsx Primus II wavelength dispersive X-ray fluorescence spectrometer (XRF) on fused glass beads. LOI was analyzed by the gravimetric method. The analytical uncertainty for reference materials is better than 2% for elements with content > 1 wt%. Ferrous iron content (FeO) was completed by titration with potassium dichromate. The limit of detection was greater than 0.5% with the relative standard deviations (RSD) below 0.8%. Whole-rock trace elements were analyzed by Agilent 7700e inductively Coupled Plasma Mass Spectrometer (ICP-MS) after HNO₃ + HF digestion of about 50 mg sample powder in high-pressure Teflon vessel. These analyses were performed with an ELEMENT spectrometer using international standards samples BHVO-2, BCR-2 and RGM-2. Accuracies based on repeated analyses of internal standards are ± 5% for REE (rare earth element) and ±5%–10% for trace elements.

Shale-normalized REY (REE + Y) patterns (subscript “SN”, normalized to the post-Archean Australian Shale (PAAS), after McLennan [37]) are presented for all iron ore samples. Considering that the chemical behavior of Y is similar to that of other REEs, Y is inserted between Dy and Ho based on its ionic radius [38]. Thus, Y and the REEs are considered together. The La anomaly was calculated using the procedure of Bolhar et al. [39]: (La/La*) _SN = La _SN/(3Pr _SN − 2Nd _SN); the Eu anomaly was calculated as (Eu/Eu*) _SN = Eu _SN/(0.67Sm _SN + 0.33Tb _SN); and the Ce anomaly was calculated as (Ce/Ce*) _SN = (2Ce _SN/(La _SN + Pr _SN)) [38].

4.4. Sm-Nd Isotopic Analysis

The procedure of Sm-Nd isotopic analysis is similar to that described by Li et al. [40,41]. Whole-rock powders (~150–300 mg) for Sm-Nd isotopic analyses were dissolved in Savillex Teflon screw-top capsule after being spiked with the mixed ¹⁴⁹Sm-¹⁵⁰Nd tracers prior to HF + HNO₃ + HClO₄ dissolution. Sm and Nd were separated using the classical two-step ion exchange chromatographic method and measured using a Triton Plus (Thermo Scientific, Bremen, Germany) multi-collector thermal ionization mass spectrometer at the IGGCAS in Beijing. The whole procedure blank was lower than 100 pg for Sm-Nd. Measured ¹⁴³Nd/¹⁴⁴Nd ratios were corrected for mass-fractionation using ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219. The JNdi-1 international standard sample was employed to evaluate instrument stability during the period of data collection, yielding values of ¹⁴³Nd/¹⁴⁴Nd = 0.512110 ± 0.000008 (2 SD, n = 3). The USGS reference material BCR-2 was measured for Sm/Nd isotopic composition to monitor the accuracy of the analytical procedures, with the following results: ¹⁴³Nd/¹⁴⁴Nd = 0.512630 ± 0.000008. The ¹⁴³Nd/¹⁴⁴Nd data of BCR-2 showed good agreement with previously published data by TIMS techniques Li et al. [40,41]. Depleted mantle model ages (T_DM(Nd)) were calculated according to DePaolo [42].

5. Results

5.1. Petrography

The Shikebutai iron ores display mainly massive and locally banded/lamellar structures (Figure 4). Massive hematite ore is steel-gray (Figure 4a) and dominated by hematite (>85%) with minor quartz, siderite, barite, and organic matter (Figure 4b). The petrographic study reveals that hematite exhibits a platy texture with weak orientation and grain sizes up to 50 μm, while associated quartz grains frequently host inclusions of extremely fine-grained hematite (Figure 4c). Banded jasper–hematite ore shows discontinuous red jasper bands (0.5–2 cm wide) alternating with steel-gray hematite (Figure 4d). Its mineralogy comprises hematite (20%–80%), quartz (10%–70%), ~5% siderite, ~5% barite, minor sericite and chlorite, trace pyrite, and locally preserved organic matter. Hematite occurs mainly as flaky to bladed crystals ranging from 50 to 200 μm in size, associated with irregular patches of organic matter that was identified by its low backscatter intensity in BSE images and confirmed by SEM-EDS spot analysis revealing high carbon concentrations (Figure 4e). Some sheet-like hematite preserves deformation textures without obvious preferred orientation. Particulate hematite (1–10 μm) within jasper matrix likely represents primary precipitation (Figure 4c,f).

The andesite sample is greyish green to dark grey and exhibits a porphyritic texture characterized by approximately 15%–20% phenocrysts and 80%–85% cryptocrystalline to fine-grained matrix (Figure 5a). The phenocrysts are primarily composed of plagioclase, which forms subhedral-to-euhedral tabular crystals ranging in size from 0.5 to 2 mm, randomly distributed within the dark matrix (Figure 5b). The plagioclase phenocrysts exhibit clear crystal outlines, displaying bright interference colors, predominantly in light gray, with prominent twinning and some crystals showing concentric zoning (Figure 5c). The matrix exhibits a pilotaxitic texture under plane-polarized light, and is mainly composed of microcrystalline plagioclases, mafic crystallites, opaque minerals, and interstitial glass, appearing deep brown to black in color (Figure 5c).

The metasiltstone is dark gray to black and exhibits a massive structure with indistinct lamination. The rock exhibits a uniform fine-grained matrix, with mineral grains being extremely fine (grain size < 0.01mm), presenting a cryptocrystalline-to-microcrystalline texture. The matrix is dominated by primarily of clay minerals (60%–70%), fine-grained quartz (20%–25%), and opaque minerals (5%–10%) (Figure 5d). A small amount of coarser clastic minerals is visible, distributed in a scattered pattern, primarily consisting of silt-sized quartz grains (grain size approximately 0.01–0.05mm), which are sub-rounded to sub-angular and display first-order-gray-to-pale-yellow interference colors (Figure 5d). The overall structural characteristics suggest that the metasiltstone was formed in a low-energy depositional environment, experiencing compaction with very low metamorphic alteration.

The chlorite–sericite phyllite is gray green to light gray in color, displaying a typical phyllitic texture. The rock shows distinct foliation, with smooth and silky lustrous surfaces. Under the microscope, the rock displays a strong preferred orientation, with mineral grains aligned in a streamline-like pattern parallel to the foliation (Figure 5e). The mineral assemblage is composed of fine-grained phyllosilicates (50%–60%, predominantly sericite and chlorite), granular quartz (20%–30%), and metallic minerals (5%–10%, mainly hematite) (Figure 5e). The platy minerals are extremely fine-grained (<0.05 mm) and are aligned to form continuous foliation planes. The sericite and chlorite microcrystals display interference colors (bluish green to greenish), aligned along the foliation direction, forming a distinct lepidoblastic texture with band widths ranging from approximately 0.1 to 0.5 mm (Figure 5f). Quartz appears in lens-shaped or banded forms between the foliation planes, exhibiting typical residual clastic features (Figure 5f). In some areas, dark metallic minerals (hematite) are observed as bands or disseminated spots in certain areas.

5.2. Zircon SHIRMP U–Pb Dating

Twenty-nine spots were analyzed on zircons collected from the andesite (sample 19SW-11). The zircons are predominantly colorless to pale brown, euhedral to subhedral, and prismatic or fragmented, with lengths of 60–80 µm and aspect ratios of 1:1.2 to 1:2. Although zircon is often less abundant in intermediate lavas, the host andesite (19SW-11) is relatively silica-rich (SiO₂: ~60%; Table S1), which facilitated zircon saturation during magma differentiation. Most of the zircon grains exhibit well-developed magmatic oscillatory or sector zone in CL images (Figure 6a; Supplementary Figure S1); the slightly sub-rounded morphology of a few grains is attributed to magmatic resorption or polishing effects rather than sedimentary abrasion, as their internal oscillatory zones remain sharp and untruncated. A few zircons have dark cores with homogenous or sectored structures (Figure 6a). The oscillatory-zoned cores and rims indicate a magmatic origin. ²³⁸U and ²³²Th abundances are 99–723 ppm and 63–950 ppm, respectively, yielding ²³²Th/²³⁸U ratios of 0.39–1.57 (Table S1). These high ratios, with the majority exceeding 0.4, indicate a typical magmatic origin for the analyzed zircons [43]. Among the twenty-nine analytical spots, twenty-one analyses plot in a small region overlapping the concordia curve, with a concordia ²⁰⁶Pb/²³⁸U age of 315.8 ± 1.5 Ma (MSWD = 1.7), and a weighted mean ²⁰⁶Pb/²³⁸U age of 315.8 ± 3.4 Ma (MSWD = 0.31) (Figure 6b). The remaining eight data points were excluded from the age calculation due to their significant deviation from the main cluster. Specifically, spots 4, 16, 19, 26, and 31 yielded scattered ²⁰⁶Pb/²³⁸U ages ranging from 350.5 to 449.3 Ma; their occurrence in dark, structured cores suggests they are inherited grains captured during magma ascent (Figure 6a). Conversely, spots 8, 10, and 17 produced younger, slightly discordant ²⁰⁶Pb/²³⁸U ages (297.4–305.0 Ma), which are attributed to post-crystallization lead loss or the presence of microscopic inclusions. Consequently, these eight points were omitted to ensure the geological accuracy of the andesite’s crystallization age.

5.3. Whole-Rock Major and Trace Elements

Whole-rock major and trace elements of the iron ore samples are provided in Table S2. The chemical composition of the ten iron ore samples is predominantly characterized by high contents of TFe₂O₃ (54.54%–98.26%) and SiO₂ (0.9%–19.05%). The Al₂O₃ contents of these samples between 0.27 wt% and 4.43 wt% is likely attributed to the addition of minor terrigenous material during their deposition. Specifically, two banded iron ore samples (e.g., 19ST-4-2 and 19ST-4-8) exhibit relatively high concentrations of FeO + MnO + MgO + CaO (8.2% and 13.75%), accompanied by correspondingly high loss on ignition (LOI) values (6.62% and 9.24%). Other typical oxides, including TiO₂ (0.01%–0.18%), Na₂O (0.00%–0.88%), K₂O (0.05%–0.94%) and P₂O₅ (0.01%–0.23%), were present at low concentrations (i.e., <1 wt%).

The trace elements abundances in the iron ore samples exhibit significant variations, such as Ba (7.22–1873.6 ppm), V (68.1–617.1 ppm), Mo (5.95–172.4 ppm), Cu (0.52–96.6 ppm), and Zn (0.63–37 ppm). The relatively high concentrations of these elements are likely attributable to the presence of minor barite and other sparse sulfide minerals. Additionally, the samples exhibit significantly low concentrations of high-field strength elements (HFSE), such as Zr (2.52–74 ppm), Ta (0.004–0.18 ppm),Nb (0.1–1.82 ppm), Th (0.09–2 ppm), and Hf (0.05–1.77 ppm), relative to Post-Archean Australian Shale (PAAS; Zr: 210 ppm; Ta: 1.28; Nb: 33.9 ppm; Th: 14.6 ppm; Hf: 5 ppm; [37]). Since these elements are typically immobile and concentrated in detrital minerals, their depletion suggests minimal terrigenous clastic input during the deposition of the iron ores [38]. Furthermore, the distinct depletion of large ion lithophile elements (LILEs), including K (0.04%–0.78%), Rb (1.49–39.02 ppm), Cs (0.04–0.77 ppm), and Sr (0.33–31.18 ppm), compared to crustal abundances (K: 3.1%; Rb: 82 ppm; Cs: 4.9 ppm; Sr: 320 ppm; [44]) indicates that the samples have preserved their primary chemical signatures and have not undergone significant post-depositional alteration or metasomatism.

PAAS-normalized REY patterns of iron ore samples are characterized by relative enrichment of light REEs (LREEs) with respect to heavy REEs (HREEs) (Figure 7a), with (La/Yb) _SN and (Sm/Yb) _SN ratios ranging between 0.58 and 4.78 and 1.84 and 6.93, respectively (Table S2). All REY patterns are also characterized by positive La anomalies ((La/La*) _SN = 1.05–2.56) and near-chondrite Y/Ho ratios (22.96–32.61). All analyzed iron ore samples exhibit pronounced positive Eu anomalies ranging from 1.62 to 7.12 (average 4.17) and lack true negative Ce anomalies (Figure 7a).

5.4. Sm-Nd Isotopes

Sm-Nd isotopes and concentrations, as well as the calculated ε_Nd(t) for andesites, metasiltstones and iron ore samples, are reported in Table S3. The andesites contain 1.46–6.03 ppm Sm and 5.11–24.50 ppm Nd, with ¹⁴⁷Sm/¹⁴⁴Nd and ¹⁴³Nd/¹⁴⁴Nd ratios of 0.1281–0.1733 and 0.512641–0.512746, respectively, and ε_Nd(t = 315.8 Ma) values of +2.75 to +3.44 (average of +3.11). The metasiltstones show Sm contents of 0.86 to 5.44 ppm, Nd contents of 2.95 to 28.06 ppm, ¹⁴⁷Sm/¹⁴⁴Nd ratios of 0.1175 to 0.1761, ¹⁴³Nd/¹⁴⁴Nd ratios of 0.512499 to 0.512644, and ε_Nd(t = 315.8 Ma) values of +0.41 to +0.95 (average of +0.7). By comparison, the iron ores (banded iron ore and massive iron ore) display Sm and Nd abundances of 0.27 to 5.07 ppm and 0.73 to 20.26 ppm, respectively, corresponding to ¹⁴⁷Sm/¹⁴⁴Nd ratios of 0.1429 to 0.2811, ¹⁴³Nd/¹⁴⁴Nd ratios of 0.512671 to 0.512964, and ε_Nd(t = 319 Ma) values of +1.99 to +2.93 (average of +2.59).

In addition, Sm–Nd isotope data of all iron ores (eight samples) are plotted along a correlation line in a conventional isochron diagram (Figure 8) with an apparent age of 319 ± 26 Ma (MSWD = 2.1). The initial ¹⁴³Nd/¹⁴⁴Nd value is 0.50930 ± 0.00018. The analytical data are presented on isochron diagram with 2σ errors, and the age error is quoted at the 95% confidence level.

6. Discussion

6.1. The Mineralization Age

The Late Carboniferous Yishijilike Formation, exposed in the Awulale belt of the Yili Block, comprises a marine volcanic sequence dominated by basaltic porphyry, andesitic porphyry, andesite, dacite, rhyolite, volcanic clastic rocks, and tuffaceous clastic rocks [49]. Geochronological investigations have established multiple age constraints on the Yishijilike Formation. Whole-rock Rb-Sr isochron ages of 317 ± 16 Ma and 319 ± 8 Ma were reported for andesites in the Aketasi area, Xinyuan County [50]. Similarly, Zhu et al. [51] obtained a SHRIMP zircon U-Pb age of 313 ± 4 Ma for trachy-andesites on the northern slope of Nalati Mountain south of Xinyuan County. In the Awulale Mountains, the formation age is further constrained by a U-Pb zircon date of 313 ± 2.6 Ma from a rhyolite at the top of the sequence [52]. The Shikebutai iron deposit is hosted within the marine volcanic rocks of the Yishijilike Formation. Regarding the Shikebutai iron deposit, age constraints have primarily relied on the footwall volcanic rocks. Reported ages include a U-Pb zircon date of 313 ± 2 Ma from a tuff at the base of the iron body [53] and 320 ± 3 Ma from a dacite at a similar stratigraphic position [54]. However, the lack of geochronological data for the overlying strata has prevented a precise determination of the upper age limit for mineralization. To address this, we conducted SHRIMP zircon dating on andesites from the Yishijilike Formation in the hanging wall of the Shikebutai ore body.

Zircons from an andesite sample (19SW-11) are morphologically short-prismatic and exhibit clear magmatic oscillatory zoning indicative of a magmatic origin (Figure 6a). Twenty-one analyses yielded apparent ²⁰⁶Pb/²³⁸U ages ranging from 307.9 to 323.2 Ma (Table S1). These analyses plot on or near concordia and yield a concordia age of 315.8 ± 1.5 Ma (MSWD = 1.7) and a weighted mean ²⁰⁶Pb/²³⁸U age of 315.8 ± 3.4 Ma (MSWD = 0.31) (Figure 6b), which is interpreted as the crystallization age of the andesite. Notably, although the analyzed samples were collected from the hanging-wall andesite overlying the orebody, its crystallization age (ca. 315.8 Ma) is slightly older than the previously reported age of the underlying tuff (313 ± 2 Ma; [53]), suggesting post-depositional faulting likely has disrupted the local sequence. This apparent stratigraphic inversion, although subtle within analytical uncertainties, may indicate localized structural disruption (e.g., faulting or folding) that has affected the original sequence—a common occurrence given the complex tectonic evolution of the Western Tianshan Orogen [27,28]. Alternatively, the slight age discrepancy could stem from inter-laboratory calibration differences or the overlapping nature of these volcanic events within a narrow time window. Accordingly, combined with the 320 ± 3 Ma age from the underlying dacite [54], the depositional age of the Shikebutai iron deposit can be more precisely constrained to the time interval of 315–320 Ma in the Late Carboniferous.

This temporal framework is independently corroborated by the direct dating of the iron ores. As illustrated in Figure 8, the Sm–Nd isotopic data for the iron ore samples define a linear array, yielding an isochron age of 319 ± 26 Ma (MSWD = 2.1). Although the analytical uncertainty of the Sm–Nd method is larger than that of zircon U–Pb dating, this age overlaps with the 315–320 Ma interval defined by the volcanic host rocks. The concordance between the crystallization ages of the wall rocks and the formation age of the ore minerals provides compelling evidence that the iron mineralization was coeval with the Late Carboniferous submarine volcanism.

6.2. Effects of Syn- and Post-Depositional Processes

Primary geochemical compositions in ancient iron deposits can be obscured by terrigenous input and metamorphic overprints. Consequently, prior to interpreting REY systematics or other tracers in terms of depositional processes, the potential influence of syn-depositional clastic contamination and post-depositional alteration must be evaluated to ensure that the data reflect original signals [39,55].

Generally, pure chemical metalliferous sediments are enriched in Fe and Mn, whereas the influx of terrigenous clastic or volcaniclastic materials is characterized by elevated concentrations of immobile elements such as Al and Ti [56,57]. Most of the iron ore samples are chemically pure, characterized by high SiO₂ + TFe₂O₃ (average 85.6 wt%), low concentrations of Al₂O₃ (generally <1.00 wt%) and TiO₂ (<0.10 wt%) and HFSEs (e.g., Nb (0.1–1.82 ppm), Th (0.09–2 ppm), and Hf (0.05–1.77 ppm)), indicating a negligible detrital contribution. As shown in Figure 9a, the majority of the Shikebutai ore samples plot within the ‘Hydrothermal’ field, confirming that the observed sedimentary textures, such as locally banded/lamellar structures, resulted from the direct seafloor precipitation of hydrothermal fluids rather than normal sedimentary processes [58,59]. Exceptions are observed in the banded jasper–hematite samples (19ST-4-2 and 19ST-4-8), which show slightly elevated Al₂O₃ (4.25–4.43 wt%) and TiO₂ (0.15–0.18 wt%) levels and plot near the boundary of detrital input (Figure 9a). These enrichments are consistent with the presence of minor Al-bearing phyllosilicates (e.g., sericite and chlorite). A strong positive correlation between Al₂O₃ and TiO₂ (Figure 9b) is observed, suggesting that the banded ores were subjected to a certain degree of syn-depositional mixing with fine-grained clastic or volcanic material rather than random post-depositional alteration. Nevertheless, no significant correlations were observed between Al₂O₃ and key paleo-environmental tracers (e.g., Y/Ho, Eu/Eu*; Figure 9c,d), suggesting that the primary signals are essentially preserved.

The REY system, particularly the redox-sensitive elements Ce and Eu, is susceptible to mobilization during diagenesis and metamorphism, which may obscure primary depositional signatures [56,60]. Specifically, extensive interaction with reducing, high-temperature metamorphic fluids typically results in the reduction of Eu³⁺ to Eu²⁺ and its subsequent desorption, leading to negative Eu anomalies in iron ore samples formed by chemical sedimentary processes [56]. However, this is not observed in iron ore samples here. As shown in Figure 7a, all samples—ranging from pure massive ores to contaminated banded ores—exhibit pronounced positive Eu anomalies (Eu/Eu* > 1) rather than Eu depletion. The absence of negative anomalies indicates that the REY system was not affected by metamorphic overprinting and fluid leaching. Moreover, the conservative behavior of the REY suite is further evidenced by the covariance in redox-sensitive elements with their immobile neighbors. In the iron ore samples, Ce defines a perfect linear trend with Pr (R² = 1.00), and Eu correlates well with Sm (R² = 0.76). According to Bolhar et al. [61], these correlations indicate that the REY system remained closed, resisting decoupling during post-depositional events. Thus, the REY signatures from all iron ore samples of the Shikebutai deposit are interpreted as pristine depositional signals.

Sm-Nd isochron ages are widely used to assess the integrity of isotopic systems in ancient precipitates [62]. The analyzed iron ore samples form a correlation trend (Figure 8) that corresponds to an age of 319 ± 26 Ma (MSWD = 2.1). Significantly, this age overlaps with the accepted formation age of the deposit, constrained by the ~315.8 Ma zircon U-Pb age of the underlying andesite. If the Sm-Nd system had been altered by later geological events, geochronological resetting would be expected, resulting in a discrepancy between the isochron and zircon ages. The observed consistency, therefore, argues against significant element mobilization, implying that the pristine Nd isotopic fingerprint (initial ¹⁴³Nd/¹⁴⁴Nd = 0.512360 ± 0.000034) has been preserved.

6.3. Inferences for Iron Sources

Major element compositions of the Shikebutai iron ore samples suggest a predominant hydrothermal origin, with high SiO₂ + TFe₂O₃ (average 85.6 wt%) and low Al₂O₃ (≤1.02 wt%) and TiO₂ (<0.18 wt%), consistent with pure chemical precipitates minimally affected by detrital influx [63,64]. The high Fe/Ti and Fe/Al ratios of the analyzed iron ore samples are comparable to those of typical hydrothermal sediments [65,66]. To assess the hydrothermal contribution, we employed the Fe/Ti vs. Al/(Al + Fe + Mn) discrimination diagram (Figure 10) [67]. In Figure 10, the iron ore samples plot exclusively along the hydrothermal end-member curve at low Al/(Al + Fe + Mn) values (<0.1) and high Fe/Ti ratios (>10), overlapping with the reference fields of modern metalliferous sediments from the Red Sea and East Pacific Rise (EPR), and are clearly separated from modern terrigenous sediments and pelagic sediments that plot toward higher Al/(Al + Fe + Mn) values. This distribution supports a significant hydrothermal input potentially exceeding 80%, although it should be noted that this diagram is semi-quantitative and the exact proportion may vary depending on the degree of mixing with terrigenous or volcanogenic components (Figure 10) [63]. Furthermore, the samples exhibit low transition metal ratios (e.g., Co/Zn and Ni/Zn), which are diagnostic of a hydrothermal origin, distinct from hydrogenous ferromanganese crusts (Table S2) [66].

REE patterns offer particularly robust discrimination between hydrothermal and seawater end-members. High-temperature (>250 °C) submarine hydrothermal fluids exhibit pronounced positive Eu anomalies (Figure 7b) resulting from preferential mobilization of Eu²⁺ during plagioclase breakdown in the oceanic crust reaction zone, coupled with LREE-enriched patterns reflecting fluid–rock equilibration at elevated temperatures [45,48]. The Shikebutai iron ore samples display significant positive Eu anomalies ranging from 1.62 to 7.12 (mean 4.14) (Figure 7a), substantially exceeding typical terrigenous sediments (Eu/Eu* ~0.65; PAAS) or low-temperature (<200 °C) hydrothermal fluids (Eu/Eu* < 1.1; [45]). The positive correlation between Fe/Al ratios and Eu/Eu* values (R² = 0.4) further supports that samples with the strongest hydrothermal signatures (highest Fe/Al) preserve the most pronounced Eu enrichment, consistent with coupled delivery between Fe and REY from high-temperature fluids [68]. However, it should be noted that positive Eu anomalies can also be influenced by factors such as diagenetic processes, extensive fluid mixing, or the inheritance of signatures from plagioclase-rich source rocks. In this study, the lack of significant correlation between Eu anomalies and detrital indicators (Figure 9d) suggests that these signatures likely reflect primary hydrothermal control. A comparison with modern submarine hydrothermal systems reveals clear compositional similarities. Figure 7 compares the Shikebutai iron ore samples REY patterns with modern high-temperature black smoker fluids (>250 °C), low-temperature diffuse flow fluids (~30 °C), and Pacific deep seawater compositions (data from [45,46,47,48]). The Shikebutai samples show LREE-enriched patterns ((La/Yb) _(PAAS) = 0.58–4.78) that closely follow the trends of high-temperature hydrothermal fluids, contrasting with the LREE-depleted character of modern seawater ((La/Yb) _(PAAS) < 0.5). Additionally, positive La anomalies (La/La* > 1.0) found in several samples indicate either signatures derived from seawater or partial oxidation of rising hydrothermal plumes [38], consistent with precipitation at the fluid–seawater interface where mixing occurs. Notably, the absence of a negative Ce anomaly in the iron ores samples (Ce/Ce* < 1.05) is best explained by rapid, near-vent precipitation following hydrothermal discharge, rather than by prolonged mixing with oxygenated ambient seawater that would impart a strong seawater-like negative Ce signature. This rapid kinetic scavenging potentially “freezes” the REY signature of the vent fluids (or the immediate mixing zone) into the precipitating particles before the distinct negative Ce anomaly of the bulk seawater can be imprinted [66,69]. As a result, the primary reducing signature of the hydrothermal fluids was likely preserved.

Y/Ho ratios provide kinetic constraints on precipitation rates. Modern oxygenated seawater displays super-chondritic Y/Ho ratios (44–74) due to preferential Ho scavenging onto particles under equilibrium conditions, whereas hydrothermal fluids and rapidly precipitated metalliferous sediments exhibit near-chondritic to moderately super-chondritic values (26–32) reflecting non-equilibrium fractionation [38,70]. The Shikebutai iron ore samples yield Y/Ho ratios averaging 28.5 (range 26.49–32.61), intermediate between chondritic and seawater end-members, which may suggest rapid precipitation that precluded full Y-Ho fractionation equilibration. However, this interpretation should be considered cautiously, as Y/Ho ratios in ancient chemical sediments can also be influenced by the degree of basin restriction and water column stratification. In a restricted or semi-restricted back-arc basin setting, as inferred for the Yishijilike Formation depositional environment [17,19], limited exchange with open ocean seawater could independently produce near-chondritic Y/Ho values even without exceptionally rapid precipitation kinetics. Nevertheless, the convergence of near-chondritic Y/Ho ratios with pronounced positive Eu anomalies and LREE-enriched REY patterns collectively lends support to a vent-proximal hydrothermal precipitation model. This pattern mirrors modern East Pacific Rise hydrothermal precipitates (Y/Ho ~28–35; [71]), consistent with rapid Fe(II) oxidation and deposition upon hydrothermal fluid discharge into oxygenated shallow seawater.

Neodymium isotope compositions of chemical sediments serve as robust tracers for constraining the provenance of iron and evaluating the relative mass fractions derived from competing sources [55,72,73]. It is broadly accepted that the Nd isotopic composition of chemical precipitates reflects that of the ore-forming fluids, which, in turn, represents a balance between two isotopically distinct end-members: a continental end-member characterized by negative ε_Nd(t) values derived from weathering and erosion of older crustal material, and a hydrothermal end-member with positive ε_Nd(t) values inherited from leaching of juvenile oceanic or arc crust during sub-seafloor circulation [74,75,76]. The Shikebutai iron ore samples display consistently positive, depleted mantle-like initial ε_Nd(t) values (ranging from +1.99 to +2.93; average +2.59; Figure 11a), coupled with pronounced positive Eu anomalies. These features are consistent with the interpretation that the dissolved Nd (and by inference, Fe) was significantly, if not dominantly, derived from sub-seafloor hydrothermal systems circulating through juvenile oceanic crust. Notably, comparison with global iron formations reveals that the Shikebutai ores display markedly more positive ε_Nd(t) values than typical Precambrian BIFs (e.g., Brockman BIF, Pietersburg BIF, Temaqami BIF, and BIF of the Qingyuan greenstone belt; Figure 11a), which generally record ambient seawater-dominated signatures, further underscoring the disproportionately large contribution of juvenile arc-derived hydrothermal fluids at Shikebutai [75,77,78,79]. Typically, seafloor-vented hydrothermal fluids acquire their isotopic signature via high-temperature alteration in the volcanic basement [73,75].

In this regard, the mean ε_Nd(t) value of the andesites is +3.11 (ranging from +2.75 to +3.44; Figure 11b), which is consistent with the depleted mantle-like values reported for the underlying volcaniclastic rocks [25]. Therefore, this value serves as a robust constraint for the hydrothermal end-member, representing fluids equilibrated with the juvenile arc crust of the Yishijilike Formation. However, the Nd isotopic composition of a depositional basin is rarely controlled by a single source; rather, it reflects a balance between hydrothermal flux and ambient seawater, the latter being dominated by dissolved riverine input and weathering of continental margins [80,81]. We acknowledge that the two-component mixing framework employed here represents a simplification of what was likely a more complex system. In particular, the Nd budget of the depositional basin could have been influenced by additional factors, including: (1) the residence time of Nd in the water column, which in restricted or semi-restricted basins is typically short (on the order of 50–200 years) relative to open ocean settings (~400–1000 years), meaning that local hydrothermal and terrigenous inputs exert a disproportionately strong control on basin water Nd isotopic compositions [22,80]; and (2) the degree of basin restriction, which governs the extent of exchange between hydrothermal vent fluids, local pore waters, and open seawater. The inferred back-arc basin setting of the Yishijilike Formation [17,19] implies that seawater circulation was likely restricted, amplifying the isotopic imprint of local hydrothermal sources relative to distal continental inputs. These factors collectively support the dominance of hydrothermal Nd in the ore-forming system, although they also introduce uncertainty into any precise quantitative partitioning between end-members. Notably, the Shikebutai iron ores display ε_Nd(t) values (average +2.59) that are distinctly lower than those of the andesitic hydrothermal end-member (average +3.11). Studies on global marine Nd isotopic compositions show that the Nd budget of modern oceans is dominated by dissolved riverine input and material mobilized from continental erosion products, such as eolian dust and continental margin sediments [80,82,83,84]. For instance, weathering of young mantle-derived rocks around the Pacific Ocean produces more radiogenic ε_Nd values than the Atlantic Ocean, whose margins are dominated by old continental crust [81]. It can thus be inferred that the Nd isotope composition of Late Carboniferous ambient seawater was also closely linked to local Nd sources. Consequently, the geochemistry and Nd isotopic data of the associated metasiltstones provide critical constraints on this continental end-member [74,75].

To constrain this ambient seawater end-member, we examine the associated metasiltstones, which yield consistently lower ε_Nd(t) values (average +0.70; +0.41 to +0.95; Figure 11b), reflecting the isotopic signature of terrigenous detritus derived from the weathering of the hinterland. Metasiltstones record the isotopic signature of their detrital source material rather than directly measuring dissolved seawater Nd. However, in restricted basin settings where the residence time of Nd is short, the dissolved Nd composition of ambient seawater tends to converge toward that of the dominant particulate input. In this case, terrigenous detritus derived from the surrounding continental margin [75,76,80]. The fine-grained, low-energy depositional characteristics of the metasiltstones are consistent with slow settling of continent-derived material in a restricted basin, suggesting that their ε_Nd(t) values approximate the isotopic composition of dissolved Nd in the background seawater with which the iron-precipitating fluids interacted. Since the dissolved Nd load of ambient seawater is generally coupled to the provenance of detrital inputs in restricted basins [75,76], these metasiltstones provide the best available approximation for the background seawater signal. Using the measured ε_Nd(t) values of the andesites (+3.11) and metasiltstones (+0.70) as hydrothermal and seawater end-members, respectively, a simple binary mixing calculation suggests that the iron ore samples (average ε_Nd(t) = +2.59) received approximately 75%–80% of their Nd budget from the hydrothermal end-member, with the remaining 20%–25% attributable to ambient seawater. While this estimate carries inherent uncertainties associated with the two-component approximation discussed above, it demonstrates that hydrothermal input overwhelmingly dominated the dissolved REE and Fe budget of the depositional system. The observed range of ε_Nd(t) values in the iron ores (+1.99 to +2.93; Figure 11a,b) likely reflects a dynamic depositional system in which the proportions of hydrothermal versus seawater-derived Nd varied spatially or temporally, possibly driven by fluctuations in vent activity or changes in the degree of basin restriction. Contributions from varying proportions of unradiogenic older continental material and radiogenic juvenile mafic volcanic rocks within the ambient seawater cannot be fully excluded, and may account for part of the isotopic variability observed. Yet, the Shikebutai ores exhibit a strong isotopic affinity toward the hydrothermal (andesitic) end-member rather than the ambient seawater (metasiltstone) signal. This pattern suggests that the hydrothermal input did not merely contribute to the basin but overwhelmingly dominated the dissolved REE and Fe budget. The high ε_Nd(t) values indicate that the rate of hydrothermally sourced Fe and Nd supply far exceeded the input from background seawater circulation. Consequently, the Nd isotopic systematics strongly support the interpretation that the Shikebutai deposit is the product of focused, intense submarine volcanic exhalative activity, where the rapid discharge of high-temperature fluids leaching the juvenile arc crust served as the primary driver for mineralization, with minimal dilution by ambient Carboniferous seawater.

As summarized in

Table 1. Comparative characteristics of Precambrian BIFs, modern hydrothermal sediments, and the Shikebutai iron deposit.

Criterion	Precambrian BIFs			Modern Hydrothermal Sediments		This Study
	Hamersley BIF ~2.47 Ga	Isua BIF ~3.77 Ga	Pongola BIF ~2.9 Ga	Red Sea	East Pacific Rise	Shikebutai Iron Deposit ~315–320 Ma
	Australia	Greenland	Africa	Atlantis II Deep	9–21° N	Tianshan, China
Metallic minerals	Magnetite, hematite, siderite, minor pyrite	Magnetite, hematite, minor pyrite	Magnetite, hematite, siderite	Goethite, Fe–Mn oxyhydroxides, sphalerite, chalcopyrite, galena	Amorphous Fe–Si oxyhydroxides (ferrihydrite), goethite, minor pyrite	Hematite (>85%), minor siderite, barite, trace pyrite
Non-metallic minerals	Chert (microcrystalline quartz), stilpnomelane, greenalite	Chert, carbonate (siderite, ankerite), hornblende (metamorphic)	Chert, stilpnomelane, minor carbonate	Fe-montmorillonite (nontronite), anhydrite, amorphous silica	Amorphous silica, nontronite, barite	Quartz (jasper), sericite, chlorite, barite
Ore texture/structure	Well-developed mesobanding (cm) and microbanding (mm), laterally continuous over 100 s km	Banded, often complexly deformed (amphibolite facies metamorphism)	Well-banded, laterally continuous, more consistent than Isua	Massive, unconsolidated, very fine-grained; poor crystallinity	Loose, poorly crystalline, rapidly deposited mounds	Massive (dominant) and banded jasper–hematite; banding less continuous and laterally discontinuous
TFe2O3 (wt%)	~50–60	~40–65	~40–60	Variable; up to ~60	~30–60	54.54–98.26 (avg. ~80)
SiO2 (wt%)	~40–50	~30–50	~40–55	Low to moderate	Variable	0.90–19.05
Eu/Eu* (PAAS)	~1.0–2.0 (slight positive)	1.5–4.0 (pronounced positive; anoxic ocean + high-T hydrothermal)	1.0–2.5	1.5–5.0 (positive)	2.0–8.0 (strongly positive near vents)	1.62–7.12 (avg. 4.14) (strongly positive)
Ce/Ce* (PAAS)	~0.8–1.0 (slight negative, oxygenated water)	~1.0 (no anomaly; pre-GOE anoxic ocean)	~0.9–1.0	~0.9–1.1 (near-neutral)	~0.9–1.1 (near-neutral near vent)	<1.05 (no negative anomaly; rapid near-vent precipitation)
(La/Yb) (PAAS)	0.3–0.8 (LREE depleted)	0.5–1.5 (variable)	0.3–0.9	0.5–3.0 (LREE enriched)	1.0–5.0 (strongly LREE enriched)	0.58–4.78 (LREE enriched)
Y/Ho ratio	40–65 (super-chondritic; seawater-like)	28–45 (intermediate)	40–60	28–38	26–35 (near-chondritic)	26.49–32.61 (avg. 28.5) (near-chondritic)
La anomaly (La/La*) (PAAS)	~1.0–1.5 (slight to moderate)	~1.0–1.8	~1.0–1.5	Variable	Variable	1.05–2.56 (positive)
εNd(t)	−2 to +2 (seawater-dominated, continental influence)	+1 to +4 (stronger hydrothermal influence)	+1 to +3	Positive (juvenile oceanic source)	Positive (MORB-like source)	+1.99 to +2.93 (avg. +2.59; strongly positive, arc-derived)
References	[11,38,62,64,68]	[39,62,74]	[55,61]	[63,65,67,85]	[45,48,70,71,86]	[24,25]; This study

, the Shikebutai iron ores are consistently distinguished from Precambrian BIFs by their markedly more positive ε_Nd(t) values (+1.99 to +2.93 vs. −2 to +2 for seawater-dominated BIFs), lower Y/Ho ratios (avg. 28.5 vs. 40–65), and stronger positive Eu anomalies (average. 4.14) while closely resembling modern Red Sea and East Pacific Rise metalliferous sediments across all key geochemical and isotopic proxies. These systematic differences indicate that the Shikebutai deposit formed through focused, high-temperature hydrothermal exhalation rather than by chemical precipitation from a seawater-buffered ocean, establishing it as a Phanerozoic analogue of vent-proximal hydrothermal mineralization rather than a BIF.

7. Conclusions

The Shikebutai iron deposit is hosted within the Late Carboniferous Yishijilike Formation in Western Tianshan. It features two principal ore types: massive hematite ores and banded jasper–hematite ores, primarily composed of hematite and quartz, with minor siderite and accessory minerals. SHRIMP zircon U-Pb dating of the overlying andesite (315.8 ± 1.5 Ma) and the Sm–Nd isochron age of the iron ore samples (319 ± 26 Ma), combined with the previous age of the underlying dacite (320 ± 3 Ma), refine the timing of mineralization to the tight interval of 315–320 Ma.

Geochemical signatures, including low Al₂O₃, TiO₂, and HFSEs concentrations, indicate negligible detrital contamination. High Fe/Ti and Fe/Al ratios, coupled with positive Eu anomalies, LREE enrichment, positive La anomalies, and near-chondritic Y/Ho ratios, affirm that high-temperature submarine hydrothermal fluids provided the dominant contribution to iron mineralization, analogous to modern vent systems.

Sm-Nd systematics of the Shikebutai iron ores samples indicate mixing between hydrothermal fluids and ambient seawater. Crucially, the strong isotopic affinity to host andesites confirms that mineralization was overwhelmingly driven by intense submarine exhalation leaching juvenile arc crust, with minimal dilution by continental-derived seawater inputs.

Therefore, the Shikebutai deposit represents a Phanerozoic analogue of proximal hydrothermal exhalative mineralization formed via direct hydrothermal fluid discharge.