Trisulfur radical anion S₃^•−—A major carrier for platinum in hydrothermal fluids

Platinum (Pt) and other platinum group elements (PGE), including palladium (Pd), iridium (Ir), rhodium (Rh), ruthenium (Ru), and osmium (Os), have many advantageous characteristics, including catalytic properties, corrosion resistance, and high melting points (Table 1), which lend themselves to various applications ranging from automotive catalysts to pharmaceutics to use in jewelry (1). In particular, the chemical behavior of PGE is quite unique: On the one hand, they are inert and highly resistant to corrosion by most chemicals; on the other hand, they can serve as catalysts, accelerating the rates of many petrochemical reactions. However, PGE have very low elemental abundances in Earth's crust, all at part per billion (ppb) levels (Table 1) (2, 3). Hence, PGE-bearing ore deposits are a critical resource, and their extracted PGE metals are highly sought after by modern industries. To effectively utilize sparse PGE resources, it is essential to understand the formation of PGE ore deposits for their targeted exploration and to develop efficient chemical approaches for PGE extraction and separation.

Table 1.

Some characteristics of PGE and associated Au

PGE or Au	Z	Melting point (°C)	Abundance in Earth's crust (ppb)	Metallic radius (Å)	Valence electron configuration
Pt	78	1,768	5	1.39	6s25d84f14
Ir	77	2,446	1	1.36	6s25d74f14
Os	76	3,033	1.5	1.35	6s25d64f14
Pd	46	1,555	15	1.37	4d10
Rh	45	1,964	1	1.34	5s14d8
Ru	44	2,334	1	1.34	5s14d7
Au	79	1,064	4	1.44	6s15d104f14

From refs. 2 and 3.

Although the natural abundances of PGE are low, fortunately, they are not evenly distributed within Earth and other planets. Because of their aversion to bonding with oxygen and their high affinity for reduced sulfur, PGE tend to either occur in the metallic state (siderophile, metal loving) or bond with sulfur or other group VA and VIA ligands (chalcophile, sulfur loving) to form mineral components (4). Consequently, PGE are concentrated in core-forming Fe-Ni alloys and in sulfide phases of PGE-bearing ore deposits (5). This distinct behavior of PGE, as well as their radioisotope systematics, can serve as indicators of many geological processes such as early solar system evolution, and planetary differentiation and formation (6).

Of critical importance to PGE-bearing ore deposits is the enrichment of PGE in Fe-Ni-Cu sulfide minerals, originating from the high extraction of PGE by sulfide liquids with respect to coexisting silicate magmas during ore deposit formation (5). Although several basic types of PGE deposits have been recognized, many aspects of PGE ore formation, especially the role of hydrothermal fluids, are not well known. Specifically, the three interrelated key properties of a hydrothermal system—solute speciation, solid phase solubility, and aqueous complex stability—and their variations with temperature, pressure, and fluid chemistry need to be thoroughly addressed for PGE-containing systems. The majority of previous studies on aqueous solutions of PGE, especially Pt^II, have focused on their complexation with common ligands such as Cl⁻, OH⁻, and SO₄²⁻ (7–9). However, the obtained concentrations of aqueous Pt species in equilibrium with Pt-bearing minerals are too low to explain the Pt mobility observed in many hydrothermal−magmatic fluids. Although hydrogen sulfide (HS^–) complexes have been investigated as plausible agents to transport Pt [as well as associated Au, also a chalcophile element (10)], there is a large range in the reported solubility data for Pt sulfide solid phases, spanning almost four orders of magnitude under different redox conditions (11–14). Therefore, it is conceivable that other Pt aqueous species may be largely responsible for Pt transport at hydrothermal conditions.

In a PNAS contribution, Pokrovski et al. (11) identify a ligand with which Pt complexes in hydrothermal fluids—trisulfur radical ion S₃^•−—using in situ X-ray absorption spectroscopy (XAS) and solubility experiments combined with atomistic and thermodynamic simulations. More specifically, Pt L₃-edge X-ray absorption near-edge structure (XANES) spectroscopy coupled with quantum chemical simulations reveal that, in a sulfur-bearing aqueous solution at 275 °C and 700 bar with f_O2 = HM–0.2 (expressed in log10 [bars] scale relative to the hematite−magnetite [HM] buffer) and pH = 5.9, Pt^II complexes with the previously recognized HS⁻ and the newly identified S₃^•− to form four-coordinated, square-planar Pt^II(HS)₂(S₃)₂^2–. By contrast, in a more acidic and oxidized solution (f_O2 = HM+0.8; pH = 4.9) at the same pressure−temperature conditions, Pt^IV complexes with HS⁻, S₃^•−, and H₂O to form six-coordinated, octahedral Pt^IV(HS)₃(H₂O)(S₃)₂^–. Further extended X-ray absorption fine structure (EXAFS) analyses show systematic increases in the first-shell mean Pt-S distance and coordination number with increasing f_O2 and decreasing pH, indicating coexistence of the Pt^II and Pt^IV species in different proportions in solutions of varying f_O2 and pH. EXAFS data also reveal the occurrence of second-shell S atoms, which is in agreement with the presence of the S₃^•− ligand in both types of Pt species. These results are further corroborated by PtS(s) solubility measurements combined with thermodynamic modeling. Compared with HS^– complexes, Pt(HS)_n²⁻ⁿ, commonly assumed to form in the literature (12, 13), the newly identified Pt^II(HS)₂(S₃)₂^2– and Pt^IV(HS)₃(H₂O)(S₃)₂^– species can enhance PtS(s) solubility by four to five orders of magnitude under similar conditions (11). This finding helps explain observed PGE enrichment in hydrothermal sulfide deposits of common metals (Cu, Mo, Zn, Au, and Ag), whereas earlier models based only on conventional sulfide and chloride ligands cannot (11). In fact, the S₃^•− and HS⁻ ligands have also been found, by Pokrovski et al., to coordinate with Au to form soluble and stable Au(HS)S₃^– in S-rich hydrothermal fluids (10). Therefore, S₃^•−, which is ubiquitous in a range of systems (15), likely plays a key role in the aqueous speciation of PGE and related metals at hydrothermal conditions and in the eventual formation of their ore deposits, despite the presence of more common ligands such as Cl⁻.

Although both the HS⁻ and S₃^•− ligands participate in Pt speciation in hydrothermal fluids, the affinity of Pt^II for S₃^•− is much higher than for HS⁻ (11). Calculation of the equilibrium constant of the ligand exchange reaction Pt(HS)₄^2– + 2 S₃^•− = Pt(HS)₂(S₃)₂^2– + 2 HS⁻ at 300 °C and 500 bar yields a value of 10^8.7, implying that Pt^II is at least ∼10⁴ times more likely to bond to S₃^•− than to HS⁻ in these conditions (11). Such remarkable affinity of Pt^II for S₃^•− can be explained by the electronic/molecular structures of Pt^II and S₃^•−, resulting in the energetically favorable bonding configuration of Pt(HS)₂(S₃)₂^2–. In contrast, the chemical affinities of Au^I for S₃^•− and HS⁻ in Au(HS)S₃^– are very similar, although Au is next to Pt in the periodic table and both are siderophile elements (10). Thus, subtle differences in the electronic structures of PGE and other coexisting metal atoms can result in significant disparities in their bonding behavior with S₃^•−. Given the systematic variations of PGE in their valence electron configurations (Table 1), it is possible that PGE may form different bonding configurations with S₃^•−. In particular, since Pt, Pd, and Ir are more chalcophile than Ru, Rh, and Os (11), the prior subgroup may have higher tendencies to complex with S₃^•− than the latter, rendering S₃^•− a potentially effective ligand for fractionation of these two PGE subgroups in natural S-rich hydrothermal fluids and for their separation and purification for technological applications.

The discovery of the [Pt-S₃^–]- and [Au-S₃^–]-type complexes by Pokrovski et al. (10, 11) relies heavily on in situ synchrotron XAS using a hydrothermal apparatus. With advances in synchrotron sources/instrumentation and the development of various sample environment cells, synchrotron X-ray scattering has increasingly allowed for unprecedented, in situ, real-time structural studies. Future research directions include 1) measurements of additional elements, including PGE other than Pt, which will determine whether there are systematics in [M-S₃^–]-type (M, metal) configurations as a function of M; and 2) optimization of hydrothermal cells to access the higher pressure−temperature regime of hydrothermal−magmatic systems, where both S₃^•− and its higher-temperature (>500 °C) counterpart, S₂^•−, may complex with Pt and other metals (11). Hydrothermal diamond anvil cell (HDAC), which can achieve temperatures/pressures up to 1,200 °C/25 kbar, has previously been combined with XAS to investigate transition metal-, rare-earth− and actinide-containing fluids (16–18), but XAS/HDAC has not yet been used to study aqueous PGE species.

In addition to XAS, other spectroscopic techniques, in particular, Raman and ultraviolet-visible spectroscopy, have been employed to measure aqueous species in various temperature, pressure, and chemical conditions. For example, Raman spectroscopy combined with HDAC was used to probe sulfur speciation in aqueous fluid-silicate melt systems (19) and to characterize uranyl chloride complexes in hydrothermal solutions (20). Together with in situ XAS studies, these techniques have allowed us to identify species that are stable only in extreme environments. In addition to novel complexes involving nontraditional ligands such as S₃^•−, as described above, a variety of species in which a metal cation is bonded with one or more common ligands, such as Cl⁻, OH⁻, and SO₄²⁻, have been identified at hydrothermal conditions. In some cases, such species (e.g., UO₂Cl₂⁰) tend to be lowly or neutrally charged, presumably due to decreases in water's dielectric constant with increasing temperature and thus weakening in the shielding of charged aqueous species in hot water (18, 20). Hence, many metal elements that are traditionally considered “immobile” based on room/low-temperature solubility data may actually be quite mobile at hydrothermal conditions, thereby explaining the wide occurrence of hydrothermal deposits with relatively high metal concentrations.