Porphyry copper-gold deposits host more than 70% of the world's total copper reserves (Sillitoe, 2010). They are exclusively related to oxidized magmas, with abundant coexisting sulfate and sulfide (Ballard et al., 2002; Liang et al., 2006; Sillitoe, 2010; Sun et al., 2015, 2017; Zhang et al., 2017). This was either attributed to sulfate reduction (Sun et al., 2004, 2013), or sulfite disproportionation (Richards, 2015). Recent experiments using diamond-anvil cell equipped with in-situ Raman spectrometer find that sulfate and sulfide are dominant species, with considerable amount of S₃^-; and S₂^-; radical ions under conditions before porphyry mineralization started. No sulfite is detected (Fig. 1). This strongly supports that sulfate reduction, rather than sulfite disproportionation, is the process that controls porphyry mineralization (Sun et al., 2013).

The experiment was carried out through in-situ Raman spectroscopic analyses of samples in a highpressure optical cell. Four different silica glasses with different alkalinity ([(K+Na)/Al]=0.6, 1.1, 1.3, and 3.0) and aqueous thiosulfate solutions (either K₂S₂O₃ or Na₂S₂O₃) were used (Colin et al., 2020). Sulfur redox balance was imposed through stoichiometric breakdown of thiosulfate and enable oxygen fugacity and acidity buffering via the dominant reactions:

The experimental conditions are dramatically different from that of porphyry and other hydrothermal deposits. These experiments were carried out at 700 ℃, 0.3–1.5 GPa, and oxygen fugacity in the vicinity of the sulfide-sulfate transition (near the nickel-nickel oxide (NNO) buffer) (Colin et al., 2020). All these conditions are different from those of porphyry Cu deposits.

The temperature of the experiments (700 ℃) is higher than that of porphyry mineralization (mostly<500 ℃, mostly less than 400 ℃), but is similar to that of primitive porphyry magmas. The pressures of the experiments (0.3–1.5 GPa) correspond to depths of 9–45 km. Note, under such high pressures, no hydrothermal deposit forms because aqueous fluids are usually dissolved in magmas (Sun et al., 2007). Porphyry Cu deposits usually form at depths of 2–4 km (Sillitoe, 2010). Therefore, the experimental conditions are dramatically different from that of porphyry and other hydrothermal deposits. Nevertheless, most porphyry deposits are connected to underlying composite plutons at depths of 5–15 km (Sillitoe, 2010), with pressures overlap with that of the experiments (Colin et al., 2020), supporting that sulfate and sulfide are the dominant sulfur species in the porphyry magmas.

The oxygen fugacity of the porphyry magmas (FMQ+(1.5–2.0)) (Sun et al., 2017; Zhang et al., 2017) is marginally higher than that of the experiment (~ FMQ+1, near the NNO buffer). Given that the higher the oxygen fugacity, the more stable the sulfate is, whereas sulfide is far more soluble in hydrous fluids than sulfate (Huang and Keppler, 2015), porphyry magmas should have higher sulfate/sulfide ratios than the experiments.

Previous authors once proposed that the dominant sulfur species in porphyry magmas is sulfite, which forms the coexistence of sulfate and sulfide through disproportionation during mineralization (Richards, 2015). This is not supported by the experiment results of Colin et al. (2020), which clearly show that sulfate and sulfide, rather than sulfite, are the stable sulfur species in porphyry magmas. Considering that the detection limits of laser Raman is quite low (Zhang et al., 2020), the absence of sulfite in the experiments indicates that sulfite is not responsible to porphyry mineralization nor other hydrothermal deposits.

Oxygen fugacity is a key factor that controls sulfur species, and consequently, porphyry mineralization (Sun et al., 2015; Liu et al., 2020), although high oxygen fugacity alone cannot form porphyry deposits (Lee et al., 2012). Previous studies suggest that sulfate is the dominant sulfur species in porphyry magmas before the onset of mineralization (Zhang et al., 2017). It is subsequently reduced to sulfide during magnetite crystallization at the late stage of magmas evolution (Sun et al., 2004; Liang et al., 2009). This process is called the "magnetite crisis" (Jenner et al., 2012). Given that sulfide is highly soluble in hydrous fluids (Huang and Keppler, 2015), the newly formed sulfide is scavenged into the ore forming hydrothermal fluids that prop up porphyry mineralization.

In addition to sulfide (H₂S and HS^-) and sulfate (HSO₄^- and SO₄^2-), S₃^-, and S₂^- radical ions were also discovered in the experiments (Colin et al., 2020), which may favor the mobility of metals (Au, Cu, and Mo) in geological fluids (Sun et al., 2013; Colin et al., 2020). Nevertheless, the experiments used aqueous thiosulfate solutions as the starting materials, which may artificially increase the proportion of S₃^- and S₂^- radical ions in the run products.

In-situ Raman spectroscopic analyses makes hydrothermal diamond-anvil cell and other highpressure optical cells, e.g., fused silica capillary, more practical in hydrothermal experiments. More attentions should be paid on systems close to natural processes, e.g., porphyry copper deposits are usually associated with intermediate adakitic rocks. Future experiments should focus on sulfate reduction, rather than decomposition of aqueous thiosulfate. In addition, hydrothermal mineralization mostly occurs at pressures lower than 0.2 GPa and temperatures lower than 500 C. Such experiments may better be carried out using fused silica capillary.