Electron probe microanalysis (JEOL JXA-8230) was carried out at Wuhan Sample Solution Analytical Technology Co., Ltd. (Hubei, China). Quantitative chemical analyses were obtained by wavelength-dispersive X-ray spectrometry (WDS) (acceleration voltage, 20 kV; beam current, 10 nA; beam diameter, 3–5 µm; number of point analyses, 11). Crystals used were PETJ (Sb and Te), TAP (As), LIFH (Fe, Cu, and Zn), and PETH (Ag, S, and Au). The reference materials were FeAsS for S and As, Ag for Ag, CuFeS₂ for Cu, and Sb₂Te₃ for Te and Sb. Because Ag-rich sulfosalts may be susceptible to electron-beam-induced Ag migration, the analytical stability of fengruiite was additionally evaluated using four analyses obtained with an 8 µm defocused beam at the same accelerating voltage and beam current. Each analysis was performed at a separate position on the grain to avoid repeated irradiation of the same area.

Because fengruiite occurs as very small intergrowths, a single domain of sufficient size and apparent homogeneity was selected from the polished section by reflected-light examination. The selected area showed no visible inclusions or fractures and was extracted as a crystal measuring 0.02 × 0.02 × 0.01 mm using a dual-beam focused ion beam platform (TESCAN GAIA 3) at the Analytical Laboratory of the Beijing Research Institute of Uranium Geology. X-ray powder and single-crystal diffraction analyses were conducted at the Science Research Institute, China University of Geosciences, Beijing. A Rigaku Oxford XtaLAB PRO-007HF single-crystal diffractometer, equipped with a rotating anode microfocus X-ray source (50 kV, 24 mA; MoKα, λ=0.71073 Å) and a hybrid pixel array detector, was used.

Raman spectroscopy of fengruiite was conducted on the same grain used for mineral structure analysis, using a Horiba LabRAM HR Evolution micro-Raman system equipped with a 50× objective at the Experiment Center of the School of Gemmology, China University of Geosciences (Beijing), China. The measurements were performed with an excitation wavelength of 523 nm and a laser output power of 25 mW. A grating with 600 grooves mm⁻¹ (centered at 500 nm) was employed, achieving a spectral resolution of 2 cm⁻¹ and a spot size of 5 µm. Spectra were collected over a range of 100–4000 cm⁻¹, with each spectrum acquired for 10 s and averaged over three accumulations.

The reflectance value of fengruiite and the coexisting cervelleite were carried out in air on the same polished section using an ultraviolet–visible–near-infrared (UV–VIS–NIR) microspectrophotometer of CRAIC Technologies Inc., at the Instrumentation and Service Center for Molecular Sciences, Westlake University, Hangzhou. The two minerals were measured during the same analytical session under identical instrumental conditions. Before measurement, the microspectrophotometer was calibrated using aluminum metal with an MgF₂ coating (λ=700 nm, R=90 %). Its calibration is traceable to NIST/NRC. Spectra were collected over the wavelength range of 400–700 nm with a spectral resolution of ≤ 2 nm. The mineral identities of the analyzed areas were independently confirmed by electron-probe microanalysis.

Fengruiite occurs as irregular strips less than 20 × 80 µm in size. The mineral is black gray with a steel-gray to iron-black streak and a metallic luster. It is brittle, shows an uneven fracture, and displays no cleavage or parting. The Mohs hardness is estimated to be 3–4 based on scratches produced by fine calcite and fluorite particles under reflected light. Micro-indentation hardness could not be determined due to the limited grain size. The calculated density is 6.62 g cm⁻³, derived from the empirical formula and unit-cell volume refined from single-crystal X-ray diffraction data. Fengruiite is nonfluorescent and nonmagnetic, with the latter property confirmed using a magnetic needle. In reflected light, the mineral appears gray with a bluish tint; is opaque; and exhibits no observable pleochroism, anisotropy, bireflectance, or internal reflections. The reflectance values of fengruiite and the coexisting cervelleite are compared in Supplement Table S1. Fengruiite shows Rmax values of 32.358 %–44.680 % and Rmin values of 30.657 %–36.184 % over the wavelength range of 400–700 nm, whereas cervelleite shows Rmax values of approximately 32.5 %–35.6 % and Rmin values of approximately 30.8 %–34.8 %. The reflectance ranges of the two minerals overlap substantially over most of the measured wavelength interval, although fengruiite exhibits a distinctly higher Rmax value at 400 nm. Therefore, reflectance data alone do not provide an unequivocal means of distinguishing fengruiite from cervelleite, and their identification in the studied assemblage is primarily based on chemical composition.

The Ag contents and analytical totals obtained with the 8 µm beam were comparable to those obtained with the 3–5 µm beam, and no systematic depletion of Ag was apparent between the two analytical conditions (Supplement Table S2). Time-resolved Ag count-rate profiles were not recorded during the original analyses; therefore, subtle short-term beam-induced Ag migration cannot be completely excluded. The compositional agreement between the two beam sizes, nevertheless, indicates that any such effect did not significantly influence the reported mean composition.

Electron probe microanalysis data show that fengruiite contains Ag (65.89 wt %–68.63 wt %), S (12.56 wt %–13.11 wt %), Sb (9.14 wt %–10.13 wt %), Te (6.34 wt %–7.56 wt %), Cu (3.26 wt %–3.76 wt %), and a minor amount of As (0.01 wt %–0.29 wt %) (Table 1). The empirical formula, normalized to 29 atoms per formula unit and partitioned between the A- and B-layer modules using the refined structural model, is [(Ag_5.87Cu_0.29)_Σ6.16(Sb_1.89As_0.03)_Σ1.92S7] [Ag₉CuS₂(Te_1.28S_0.63)S_1.91], and the simplified formula is [(Ag,Cu)₆Sb₂S₇][Ag₉CuS₂(Te,S)₂]. The ideal formula is [Ag₆Sb₂S₇][Ag₉CuS₂Te₂], corresponding to a calculated composition S 12.84 wt %, Ag 67.20 wt %, Cu 2.64 wt %, Te 7.21 wt %, and Sb 10.11 wt %, summing to 100 wt %.

The selected crystal was used to collect a hemisphere of diffraction data over the ranges -9≤h≤10, -8≤k≤9, and -16≤l≤16. The intensity data were corrected for Lorentz and polarization effects, and a multi-scan absorption correction was applied. The structure was solved by direct methods using SHELXT and refined against F2 by the full-matrix least squares method using SHELXL within OLEX2-1.3 (Dolomanov et al., 2009; Sheldrick, 2015a, b).

A total of 3288 reflections were measured, yielding 626 independent reflections (Rint=0.050), of which 548 had I>2σ(I). The refinement included 65 parameters and 1 restraint and converged to R1 = 0.0667 for observed reflections, wR2 = 0.109 for all data, and a goodness of fit of 1.29. Complete crystallographic and refinement details are provided in Table 2. The crystallographic information file is available as a supplement. The X-ray powder diffraction data of fengruiite (MoKα, λ=0.71073 Å) are presented in supplementary Table S3, with the seven strongest lines highlighted in bold. Unit-cell parameters obtained from the single crystal are a=7.6087(9) Å, c=11.970(2) Å, V=600.13(17) ų, and Z=1, while those obtained from powder data are a=7.6087(11) Å, c=11.9700(7) Å, and V=600.13(21) ų, with Z=1.

Fengruiite is trigonal and crystallizes in space group P3‾m1 (no. 164). Its unit-cell dimensions, symmetry, and stacking sequence identify the structure as the Tac polytype of the pearceite–polybasite group (Bindi et al., 2007). The structure consists of alternating A- and B-layer modules stacked along the c axis, consistent with the modular structural description established for this mineral group (Bindi et al., 2007; Bindi et al., 2013) (Fig. 4). The refined atomic coordinates, site occupancies, and equivalent isotropic displacement parameters are given in Table 3, and the anisotropic displacement parameters are reported in Table 4. The locally droplet-like electron density associated with Ag in the A-layer module was represented in the final refinement by two mutually exclusive split positions, Ag1 and Ag2, with refined occupancies of 0.78(4) and 0.21(4), respectively. Their occupancies sum to approximately 1; therefore, Ag1 and Ag2 do not represent two independent, simultaneously occupied crystallographic sites but alternative statistical positions of the same disordered Ag population. The extended Ag electron density in the B-layer module was approximated in the main refinement by the partially occupied Ag3, Ag4, and Ag5 positions using conventional harmonic displacement parameters. To further evaluate this diffuse electron density, a trial anharmonic refinement was performed in Jana2006 using third-order Gram–Charlier coefficients for the Ag sites. This refinement converged stably and improved the agreement factors (R1 = 0.0623 for 482 reflections with I>3σ(I); wR = 0.0576 for observed reflections; wR(all) = 0.0640 for all reflections). The anharmonic model confirms the strongly non-harmonic character of the Ag electron density, but it does not affect the species-defining S–Te distribution. Therefore, the conventional split-site model is retained here as the principal structural model, whereas the anharmonic refinement is provided as an auxiliary model in the Supplement, with the third-order Gram–Charlier coefficients listed in Supplement Table S5 and the corresponding auxiliary CIF provided as Supplement CIF 2.

The S–Te distribution was evaluated during the structure refinement for all crystallographically distinct anion sites. At the mixed Te1/S1 site in the B-layer module, Te and S were refined at the same position, with their occupancies constrained to sum to unity, yielding Te_0.680(12)S_0.320(12). Mixed S–Te occupancy was also tested at the S2 site in the A-layer module, but only a minor Te component was obtained, [S_0.95(2)Te_0.05(2)]; this contribution was not retained in the final model. The occupancies of the remaining S sites, S3 and S4, were also tested by free refinement. The refined occupancies were 0.996(15) for S3 and 1.01(2) for S4, both of which are statistically consistent with full occupancy by S. Therefore, the occupancies of the S3 and S4 sites were fixed at 1.0 in the final refinement. Thus, the Te1/S1 site in the B-layer module is Te dominant, whereas the anion sites in the A-layer module remain S dominant. This distribution provides the principal structural basis for distinguishing fengruiite from Te-rich polybasite-Tac, which lacks a Te-dominant crystallographic site, and from benleonardite, in which Te is distributed among anion sites of both structural modules. Accordingly, the identification of fengruiite is based primarily on the refined distribution of Te among crystallographic sites rather than on bulk Te content alone. Selected model-dependent interatomic distances are provided in Supplement Table S4.

The Raman spectrum of fengruiite displays bands at 144, 764, 937, 1474, and 1564 cm⁻¹ (Fig. S1). Because no lattice-dynamical calculations were performed and the available reference spectra were acquired under different analytical conditions, definitive assignments of these bands are not made here. The Raman spectrum is presented as supplementary characterization, whereas the species definition of fengruiite is based primarily on its chemical composition and single-crystal structure.

The crystal structure of fengruiite consists of two compositionally distinct modules stacked along the c axis. These are conventionally designated as an A module with composition [Ag₆Sb₂S₇] and a B module corresponding to [Ag₉CuS₂Te₂]. Each module forms a continuous layer parallel to (001), and their alternating repetition accounts for the trigonal metric and the one-layer stacking sequence.

In the A layer, Sb forms well-defined [SbS₃] pyramids. In contrast, the electron density assigned to Ag is locally droplet-like and was approximated in the refinement by the mutually exclusive Ag1 and Ag2 split positions, whose refined occupancies sum to approximately 1. Ag1 and Ag2 should therefore not be regarded as two independent, simultaneously occupied crystallographic sites. Rather, they represent a discrete model of an extended electron-density distribution. Consequently, the individual Ag–S distances and coordination geometries derived from these positions are model dependent and should be interpreted cautiously.

In the B layer, Cu is linearly coordinated by two S atoms, and the mixed Te1/S1 site, with an occupancy of Te_0.680(12)S_0.320(12), occurs near the center of the layer. The Ag electron density extends along pseudohexagonal two-dimensional pathways parallel to the ab plane and is represented by the partially occupied Ag3, Ag4, and Ag5 positions (Fig. 5). These positions correspond to statistical maxima along the extended electron-density distribution rather than fixed atomic sites with rigid coordination polyhedra. The observed density is consistent with pronounced positional disorder and possible Ag mobility, although the present room-temperature diffraction data do not distinguish unequivocally between static and dynamic disorder.

Comparable continuous or pseudohexagonal Ag electron-density distributions have been reported in other members of the pearceite–polybasite group and have been interpreted as potential pathways for Ag–ion migration, consistent with the ionic-conducting behavior documented for this mineral group (Withers et al., 2008; Bindi et al., 2013, 2015). Accordingly, the extended electron density shown in Fig. 5 may represent potential two-dimensional Ag-migration pathways within the B layer of fengruiite. However, no electrical-conductivity measurements or temperature-dependent diffraction experiments were performed in the present study. The electron-density distribution therefore provides structural evidence of possible Ag mobility but does not by itself demonstrate dynamic Ag diffusion or ionic conductivity in fengruiite. The auxiliary Gram–Charlier refinement is consistent with this interpretation but is not used as the principal model for the formal species description.

The overall architecture results from the ordered alternation of the negatively charged [Ag₆Sb₂S₇] layer and the positively charged [Ag₉CuS₂Te₂] layer. The interlayer compatibility of these modules stabilizes the trigonal symmetry and defines the structural identity of fengruiite.

Because members of the pearceite–polybasite group may show overlapping bulk chemical compositions, the following species-level comparison is based primarily on the refined occupancies and distributions of Te among crystallographically distinct anion sites. The crystal structure of fengruiite is topologically equivalent to the structures observed in all members of the pearceite–polybasite group, specifically the 111 pearceite-type structure (Tac polytype) (Bindi et al., 2007). A subsequent study, Bindi et al. (2013), identified a Te-rich variant of polybasite-Tac, characterized by significant concentrations of Te (occupancy of 0.49) at the S4 site within the B-layer module. Despite this Te-rich composition, the absence of Te-dominant sites in the polybasite crystal led to the conclusion that it did not qualify for classification as a new mineral.

In fengruiite, the Te1 site in the B-layer module is dominantly occupied by Te, with a refined occupancy of Te_0.680(12)S_0.320(12). This site occupancy distinguishes fengruiite from the Te-rich polybasite-Tac described by Bindi et al. (2013), in which Te is distributed among S sites but does not become the dominant constituent at any individual crystallographic site. Fengruiite also differs from benleonardite, in which Te is incorporated into anion sites of both the A- and the B-layer modules (Bindi et al., 2015). Accordingly, the distinction among these species is fundamentally based on the distribution and dominance of Te at crystallographically distinct anion sites, rather than on the total Te content alone.

The preferential incorporation of Te at the Te1/S1 site in fengruiite can be qualitatively related to the different local coordination environments of the anion sites. The Te1/S1 site is surrounded exclusively by the disordered Ag substructure in the B-layer module, providing a relatively weakly constrained and polarizable environment that can accommodate the larger Te anion. In contrast, S4 participates in the linear S–Cu–S coordination, S3 forms part of the [SbS₃] pyramids, and the anion sites in the A-layer module are incorporated into a more constrained Ag–Sb–S framework; these environments appear to favor S over Te. The broader distribution of Te in benleonardite may also reflect its substantially higher bulk Te content, which requires Te incorporation into several crystallographic environments. This interpretation remains qualitative because no energetic calculations were performed to quantify the relative stability of Te at the different anion sites.

Electron-probe microanalysis provides the bulk Te content but cannot determine its distribution between the A- and B-layer modules. Consequently, EPMA data alone may not permit an unambiguous distinction among Te-rich polybasite-Tac, fengruiite, and benleonardite. Total-Te ranges of Te_tot < 1 apfu, 1 ≤ Te_tot < 2.5 apfu, and Te_tot ≥ 2.5 apfu may be used only as provisional compositional guidelines for polybasite-Tac, fengruiite, and benleonardite, respectively, under the assumption that Te preferentially enters the B-layer module. These values should not be regarded as universal or formal species boundaries because Te may be distributed among several anion sites without becoming dominant at any one site, as demonstrated by Te-rich polybasite-Tac.

For compositions that overlap these proposed boundaries within analytical uncertainty or that fall within the compositional overlap among the three species, a definitive species assignment should not be made solely from EPMA data. In the absence of reliable site-occupancy information, such material is best reported as a Te-bearing member of the pearceite–polybasite group, together with its complete chemical composition, until suitable structural data become available. This limitation does not affect the type material of fengruiite, for which single-crystal X-ray diffraction directly demonstrates a Te-dominant Te1/S1 site in the B-layer module together with S-dominant anion sites in the A-layer module. These structural data establish fengruiite as a distinct mineral species of the pearceite–polybasite group. The principal chemical and structural differences among the related species are summarized in Table 5.