Boulangerite is a common mineral and a prominent member of the family of rod-based structures (Makovicky, 1993), present in many hydrothermal lead mineralisations. The single-crystal X-ray diffraction studies on natural ordered boulangerite of Mumme (1989) and Ventruti et al. (2012) show a monoclinic system; space group P21/a; and unit-cell parameters a=21.61, b=23.54, c=8.08 Å, β=100.71° and a=21.554(4), b=23.454(4), c=8.079 Å, and β=100.76(1)°. The ideal formula is Pb₁₀Sb₈S₂₂ (Z=4), reflecting the presence of 18 cation and 22 anion independent sites in the structure.

In contrast, the single-crystal X-ray diffraction studies on a synthetic phase with similar chemistry to boulangerite, described by Petrova et al. (1978) and Skowron and Brown (1990), yielded space group Pbnm and unit-cell parameters a=21.40(1), b=23.51(1), and c=4.036(1) Å and space group Pnam and unit-cell parameters a=23.490(5), b=21.245(1), and c=4.020(1) Å, respectively. The ideal formula is Pb₅Sb₄S₁₁ (Z=4), reflecting the presence of 9 cation and 11 anion independent sites in the structure. The synthetic phase is a disordered orthorhombic phase with c∼4.02 Å (i.e. half of c of natural boulangerite) and should not be named boulangerite but rather a “boulangerite-like” phase.

Arsenic is a known minor component in boulangerite, but most published analyses show limited substitution of Sb with As, with atomic As / (As + Sb) ratios that are typically below 0.125. For example, Jambor et al. (1982) quote a ratio of 0.125 (i.e. 2.02 As wt %) for As-bearing boulangerite from Madoc, Orlandi et al. (2008) describe boulangerite from Bottino with As / (As + Sb) equal to 0.106 (i.e. 1.7 As wt %), Ventruti et al. (2012) found boulangerite to have the highest As / (As + Sb) atomic ratio of only 0.0144 at this locality (i.e. 0.24 As wt %), while Birnie and Petersen (1977) describe ratios of As / (As + Sb) (in apfu) as high as 0.36 (i.e. 5.97 As wt %) in samples from the Lichito I vein of Huachocolpa, Peru.

Substantial arsenic concentrations in other sulfosalts have been observed in zinkenite (4.4 As wt % in Jas Roux, Johan and Mantienne (2000); from 0.0 wt % to 11.5 wt % in the Hemlo gold deposit, Harris (1989); and up to 7.6 wt % in the Vorontsovskoe gold deposit, Kasatkin et al., 2022), tsugaruite (∼12.5 As wt %, Shimizu et al., 1998, and Biagioni et al., 2021), vurroite (∼8.4 As wt %, Garavelli et al., 2005), tazieffite (∼9.7 As wt %, Zelenski et al., 2009), the sterryite group (from 2.8 As wt % to 8.3 As wt %, Moëlo et al., 2011; Bindi et al., 2016, and Biagioni et al., 2016), chovanite (3.37 wt %, Biagioni and Moëlo, 2017), and ginelfite (6.2 wt %, Biagioni et al., 2025). The single-crystal studies show that arsenic plays an important role in the crystal structure and the crystal chemistry (due to the As-dominant site or sites) of the aforementioned minerals, except for zinkenite. Low contents of As, as in andreadiniite (1.57 wt %) in the study of Biagioni et al. (2018) and launayite (2.7 wt %) in the study of Topa et al. (2025), were shown to play no structural role. The structural role of As in sorbyite (up to 6.2 wt %), playfairite (up to 3.3 wt %), and madocite (up to 3.1 wt %) as described by Jambor et al. (1982) remains unknown.

Here we describe in detail the new mineral lopatkaite, a high-arsenic-bearing analogue of boulangerite, from Madoc, Ontario, Canada. The new mineral is named after Eduard Lopatka (born 1961), who provided to us his sample collection of Madoc material, in which we found the new mineral.

The new mineral species and name were approved by the Commission on New Minerals Nomenclature and Classification of the International Mineralogical Association (IMA 2012-83; Topa et al., 2013). The holotype, consisting of a polished sample, was deposited in the mineralogical collection of the Natural History Museum, Vienna, Austria, under catalogue number O 3954 (note: the holotype was originally deposited in the collection of the Department of Materials Engineering and Physics, University of Salzburg, Salzburg, Austria, under catalogue number 15008 but was later transferred to Vienna).

The type locality of lopatkaite is the classical sulfosalt occurrence at the Taylor Pit in Madoc, Ontario, which was described previously by Jambor (1967a, b) and Jambor et al. (1982). Sulfosalts occur there in steeply dipping cipollino calcite–dolomite marbles near their contact with Precambrian granite. The bulk of sulfides is composed of dark-brown sphalerite, boulangerite, jamesonite, and fine-grained arsenopyrite and pyrite. A variety of rare sulfosalts (Jambor, 1967a, b) occur sparingly, intergrown with the just mentioned ones, and they are difficult to identify. They comprise semseyite, bournonite, playfairite, launayite, madocite, twinnite, veenite, guettardite, Sb-bearing baumhauerite, and tetrahedrite. They were identified in and described for this locality in the quoted publications.

Jambor et al. (1982) suggest that As-rich mineralisation came first at Madoc, followed by an Sb-rich one. The association galena–pyrite–sphalerite might be of synsedimentary origin, whereas hydrothermal activity connected with the intrusion of nearby granite plutons into the metamorphosed limestone triggered the formation of sulfosalts.

In the investigated samples, lopatkaite forms grains which attain a size of up to 300 µm and grain aggregates reaching up to more than 500 µm across. Grains of lopatkaite are embedded in calcite and are subhedral to anhedral. Neither distinct crystal forms nor twinning were observed. Enveloping by later boulangerite partly leaves traces of original crystal shape and partly leads to heavily corroded grains enclosed in boulangerite (Fig. 1). The colour is greyish black, with metallic lustre and dark-grey streak. Lopatkaite is opaque with a Mohs hardness (derived from the indentation measurements) of 3–3.5; VHN₅₀ data are as follows: a range of 189–208 and a mean of 200 kg mm⁻². Lopatkaite is brittle, without observed cleavage and parting. The density could not be measured because of the paucity of available pure material; the density (calc.) is equal to 6.168 Mg m⁻³ using the empirical formula and 6.096 Mg m⁻³ using the ideal formula. The a:b:c ratio calculated from the unit-cell parameters is 0.683:1:0.429.

In reflected light under one polariser, the colour is greyish white. No internal reflections were observed, and pleochroism or bireflectance is distinct and white to dark grey, especially in oil. Anisotropy is distinct, with rotation tints in shades of grey. Quantitative data on reflectance were obtained by means of a WTiC standard and Leitz MPV-SP microscope photometer and are given in Table 1.

Chemical analyses of lopatkaite were carried out at the Central Research Laboratories of the Natural History Museum, Vienna, using a JEOL “Hyperprobe” JXA 8530F field emission gun electron probe microanalyser (FE-EPMA). The analytical conditions were as follows: WDS measuring mode, accelerating voltage of 25 kV, beam current of 20 nA, beam diameter of 1.5 µm, and count times of 10 s in peak position and 5 s in background positions. JEOL software was used for the acquisition and processing of data. The following emission lines and standards were used: AsLα and TlLα (lorándite, TlAsS₂), PbMα (galena), AgLα (Ag metal), SbLα and SKα (stibnite), HgLα (cinnabar), and CuKα (chalcopyrite). Other elements such as Bi, Fe, and Hg were sought but not detected. Proper empirical correction was made for the interference of the third order of the SbLα line with the analytical AsLα line. Under the analytical conditions described, the detection limits for the measured elements in the lopatkaite matrix were as follows (expressed in wt %): S, Cu, and Fe ∼0.03; As, Ag, and Sb ∼0.06; and Hg, Tl, Pb, and Bi ∼0.09. Analytical data are given in Table 2.

The empirical formula (based on 40 apfu) is Pb_10.28Sb_6.06As_1.74S_21.92, leading to a simplified formula of Pb₁₀Sb₆As₂S₂₂, which requires the following (in wt %): 56.64 Pb, 19.97 Sb, 4.10 As, and 19.29 S, for a total 100.00. There are well-documented difficulties in obtaining the theoretical Pb / (Sb + As) ratio of 1.25 for a boulangerite-type structure, described in detail by Ventruti et al. (2012); explanations given for these problems vary between the correction problems of electron microanalyser programmes and the presence of structural defects in the rod-based sulfosalt structure, both without a valid and/or unambiguous proof.

Powder X-ray diffraction data could not be collected due to a lack of suitable pure material and were instead calculated based on single-crystal structure determination results. The seven strongest lines are shown in Table 3.

The structure of lopatkaite contains 18 distinct coordination polyhedra of cations and 22 anions (Table 5, Fig. 3). The asymmetric unit contains nine Pb sites, one mixed (Pb, Sb) site, two Sb sites, two mixed (As, Sb) sites with occupancy values of 0.67–0.58 As and 0.33–0.42 Sb, two mixed (Sb, As) sites, and two mixed (Sb, Pb) sites. Similarly to the closely related boulangerite, the lopatkaite structure (Fig. 4) is isotypic with the MDO1 polytype described by Ventruti et al. (2012) and contains large rod-like modules based on an SnS archetype described by Makovicky (1993) and Ferraris et al. (2008) and Moëlo et al. (2008).

The asymmetric unit content is presented in Fig. 2, and the crystal structure is shown in Fig. 3, in which the rods are outlined by shading in green. Figure 3 shows that the structure consists of (001) ribbons composed of [100] rods with a lozenge-like cross-section. Each rod consists of three broad double ribbons separated by interspaces which accommodate lone electron pairs of Sb³⁺ and As³⁺ cations.

The independent sites of the asymmetric unit (Fig. 4) build three ribbons, R₁, R₂, and R₃, each being three polyhedra long. The ribbons R₁ and R₂ are equivalent to the “ribbon C” and “ribbon M”, respectively, described by Ventruti et al. (2012).

Pure Pb sites (Pb1, 2, 3, 5, 6, 7) lie on ribbon R₃, which has a (100)_PbS configuration. The Pb atoms and the S atoms form a pseudotetragonal surface (Fig. 5). Polyhedra centred by Pb1, Pb5, Pb3, and Pb7 are bicapped trigonal prisms (CN = 8), while polyhedra centred by Pb2 and Pb6 are monocapped trigonal prisms (CN = 7). The ribbon R₂ hosts a central Sb2–(Sb,Pb)7 column of coordination polyhedra and two marginal, mixed Sb5–(As,Sb)8 and (Sb,As)4–Pb10 columns of polyhedra. The central ribbon R₁ of the rod hosts a central Sb1-Pb8 column of coordination polyhedra, and two marginal, mixed (As,Sb)3–Pb9 and (Sb,As)4–(Pb,Sb)4 columns of polyhedra.

The other two surfaces of the rod are composed of S atoms in a periodically sheared pseudohexagonal arrangement (Fig. 4). These surfaces face the pseudotetragonal surfaces of the adjacent rods, which are occupied by Pb atoms (as mentioned above), and the inter-rod space (left blank in Fig. 3) accommodates a complicated pattern of long Pb–S distances and weak lone-electron-pair stereoactivity of Pb. Its character is very different from the regular lone-electron-pair spaces inside a rod layer (Fig. 3).

Along the 8 Å a direction (Fig. 5), the (As,Sb)3 sites alternate with Pb9 sites; the (Sb,As)4 sites alternate with Pb10 sites; and the third of the mixed sites, (As,Sb)8, alternates with the Sb5 site. The bonding pattern is that which is typical of Sb sulfosalts: no crankshaft chains are present (Fig. 5), but an Sb2–Sb5 pair connected via a short S–S edge and via shorter Sb–S distances is present, whereas Sb7 and (As,Sb)8 share a common long edge of their coordination polyhedra. Only Sb1 is top-bonded; the rest of Sb and As atoms are side-bonded, as can be seen from the short Sb–S distances in Fig. 4.

The atoms Sb7 and Sb8 have typical short Sb–S bond lengths (2.33–2.44 Å) in relation to the vertex of the coordination pyramid and 2.47±0.01 Å at its base (Table 6), whereas those of Sb1 and Sb6 are longer than expected for the pyramidal base (2.55–2.61 and 2.61–2.66 Å, respectively). The (Sb,Pb)5 and (Sb,Pb)6 sites have a bonding scheme which deviates substantially from the usual situation for antimony, suggesting a flipping between opposing bond orientations. All the mixed sites are distributed regularly along the sheared pseudohexagonal boundaries of the lozenge-shaped rods; none of them are in the rod interior. The bonding schemes of dominant Sb sites show remarkable similarity to those in monoclinic boulangerite (Mumme, 1989; Ventruti et al., 2012), in which the Sb sites show typical Sb coordination and elongated short-bond versions and in which these sites, including the peculiar Sb6 site (it has the same name in Ventruti et al., 2012), occur at the same structural sites and with about the same bond lengths. The analogue of the Sb5 site in the structure description by Mumme (1989) has bond lengths of 2.73, 2.82, 2.87, and 2.88 Å; in the description by Ventruti et al. (2012) the corresponding values are 2.74, 2.84, 2.80, and 2.90 Å; both coordinations in the pyramidal base are slightly larger than those in lopatkaite, in agreement with the larger structure volume of boulangerite. The bond distances to the pyramidal S vertex are typical for Sb – 2.48 and 2.43 Å, respectively – in these two refinements.

The Sb1 site, with somewhat elongated short basal distances (above), is in a rather symmetrically positioned central site of the double layer. It is surrounded by Pb coordination polyhedra and compensates for these distances by means of shortened long Sb–S distances (Table 6). Similar compensation occurs for the marginal Sb6 site.

Arsenic is preferentially concentrated at peripheral ribbon positions, where steric constraints favour incorporation of the smaller As³⁺ cation relative to Sb³⁺. This distribution pattern is consistent with that observed in other Pb–Sb–As sulfosalts and reflects crystal–chemical control rather than random substitution (e.g. Moëlo et al., 2011; Biagioni and Moëlo, 2017).

The chemistry and the results of the crystal structure determination allow the calculation of a theoretical value of the arsenic content, which will define the border between As-bearing boulangerite and lopatkaite. Considering, theoretically, only two (As,Sb) sites with minimum As_0.505Sb_0.495 occupancy, the resulting formula is Pb₁₀As_1.1Sb_6.9S₂₂, and 2.23 wt % As is required. For two (As,Sb) sites with As_0.505Sb_0.495 occupancy and a third (Sb,As) site with Sb_0.6As_0.4 occupancy, the resulting formula is Pb₁₀As_1.5Sb_6.5S₂₂, which requires 3.05 wt % As. This suggests that a minimum of ∼3 wt % As content is necessary for the formation of lopatkaite.

Based on more than 250 spot analyses of lopatkaite from the Madoc deposit, we found a range of chemistries, from low-As lopatkaite to high-As lopatkaite, with empirical formulae varying from Pb_10.48As_1.57Sb_6.10S_21.84 to Pb_10.39As_2.60Sb_5.03S_21.98. The values of (Sb + As), Pb / (Sb + As), and As/Sb for low- and high-As varieties range from 7.67, 1.37, and 0.26 to 7.62, 1.36, and 0.52, respectively. Two clear trends are observed: (1) Sb + As is less than the ideal value of 8, and (2) the Pb / (Sb + As) ratio is higher than the ideal value of 1.25 for ideal boulangerite, Pb₁₀Sb₈S₂₂. The deviation from ideal values of (Sb+As) and Pb / (Sb + As) also appears for chemistries of boulangerite and As-bearing boulangerite measured on Madoc material. The binary plots of As vs. Sb, Pb vs. (Sb + As), and Pb / (Sb + As) vs. Sb / (Sb + As) are shown in Fig. 6. A detailed explanation and discussion of these observations will be given in a future publication, which will include the results of lopatkaite studies of samples from other localities.

The comparison of empirical and structural formulae and unit-cell parameters for boulangerite and lopatkaite is presented in Table 7. The comparison of independent atom sites and site occupancies for the boulangerite group members is given in Table 8. In Mumme's refinement the site A5, assigned as the pure Pb site, shows Me–S distances consistent with a (Pb,Sb) site ,and the site A5^′, assigned as the pure Sb site, shows Me–S distances consistent with a (Sb,Pb) site, in accordance with the results for A5 and A5^′ sites of Ventruti's solution and in accordance with Pb4 and Sb6 sites of our refinement.

Lopatkaite and monoclinic boulangerite share the same space group (P21/c), topology, coordination environments, and rod-based structural architecture. No subgroup–supergroup relationship, symmetry reduction, or topological modification is observed. According to the definitions of Lima-de-Faria et al. (1990), the two minerals are therefore isotypic. The substitution of As for Sb does not alter the structural type. In the studied material, arsenic is dominant over antimony at two crystallographically independent mixed (Sb,As) sites. Under the dominant-constituent rule adopted by the IMA–CNMNC, such site dominance is sufficient to define a new mineral species, even when the overall structure remains unchanged. The ideal end-member composition of lopatkaite is thus Pb₁₀As₂Sb₆S₂₂, representing the arsenic-dominant isotype of boulangerite. Intermediate compositions reflect continuous As–Sb substitution within the same structural framework; however, species definition is based on site occupancy dominance. Textural observations (Fig. 1) show that lopatkaite is overgrown and partially replaced by boulangerite, indicating changes in As/Sb ratios during mineralisation. These features are consistent with the two-stage sulfosalt deposition model proposed for the Madoc deposit (Jambor et al., 1982).

Lopatkaite is a pure Pb–Sb–As rod-based-structure sulfosalt mineral. It is an isotype of boulangerite with two structurally distinct cation sites, predominantly arsenic-occupied, and one more in a reversed ratio with Sb (Table 8). The above-described distribution of As in the rod volume, always under rod surfaces running diagonally to the crystal axes (Fig. 3), influences both the b and c parameters and, more strongly, the b parameter (Table 7). The unit-cell parameters are correspondingly reduced in relation to boulangerite (Table 7). The monoclinic boulangerite studied by Mumme (1989) has a 21.61 Å, b 23.54 Å, c 8.08 Å, β 100.71°, and space group P21/a compared with our values of a 8.0806 Å, b 23.360 Å, c 21.488 Å, β 100.709°, and space group P21/c. With respect to the discussion about the symmetry and space group of synthetic and natural boulangerite in the literature (Pbnm, Petrova et al., 1978; Pnam, Skowron and Brown, 1990; P21/c; Mumme, 1989; Ventruti et al., 2012), lopatkaite is unambiguously on the side of the ordered structures and monoclinic symmetry.

After the approval of the lopatkaite proposal (January 2013), we determined further occurrences of lopatkaite (based on EPMA data and crystal structure determinations) at two other localities: the Barika gold deposit, Azerbaijan Province, Iran, and the Balmat deposit, New York, USA. Kasatkin et al. (2022) described lopatkaite with 3.08 wt % As from the Vorontsovskoe deposit, Russia, and Biagioni et al. (2025) reported lopatkaite with 3.8 wt % As from the Jas Roux deposit, France, with both identifications being based on EPMA data. Although Birnie and Petersen (1977), who studied As-rich “boulangerite” samples from Huachocolpa, Peru, indicated an As / (As + Sb) ratio of only 0.36, without giving explicit EPMA data, one can assume based on the calculated 5.97 As wt % content of their samples that they represent the first analytical data of lopatkaite. A comparative study of our newly determined lopatkaite occurrences and these literature data will be the subject of a separate paper.