The basalt-hosted Semenov-2 hydrothermal field on the Mid-Atlantic Ridge is host to a rather unique Cu-Zn–rich massive sulfide deposit, which is characterized by high Au (up to 188 ppm, average 61 ppm, median 45 ppm) and Ag (up to 1,878 ppm, average 490 ppm, median 250 ppm) contents. The largest proportion of visible gold is associated with abundant opal-A, which precipitated after a first generation of Cu, Fe, and Zn sulfides and before a second generation of Fe and Cu sulfides. Only rare native gold grains were found in earlier sulfides. Fluid inclusions in opal-A associated with native gold indicate precipitation at 300° ± 40°C from fluids of salinity higher than that of seawater (3.5–6.8 wt % NaCl equiv). According to laser ablation-inductively coupled plasma-mass spectrometry analyses, invisible gold is concentrated in secondary covellite (23–227 ppm) rather than in the primary sulfides (<1 ppm). Silver minerals (native silver, stutzite, and naumannite) rarely occur in the sulfides and in aragonite associated with opal-A; invisible silver was detected in all sulfides, but, again, covellite contains more Ag (>1,000 ppm) than all other sulfides (<250 ppm). Covellite replacing Zn sulfides (covellite-A) is enriched in all analyzed trace elements relative to covellite replacing Cu-Fe sulfides (covellite-B). The enrichment of covellite-A in trace elements may be related to the dissolution of inclusions of various minerals hosted in former sphalerite, which were the source for Au and Ag (native gold), Pb and Tl (galena), Se (chalcopyrite, Se-bearing galena, naumannite), Te and Bi (Bi tellurides), As (tennantite, chalcopyrite), and Sb (tennantite). The formation of covellite-A was favored by hydrothermal fluid/seawater mixing or direct oxidation of sulfides by seawater, as suggested by the relatively high contents of typical “seawater” elements (U and V). The degree of seawater involvement was apparently lower for covellite-B.
Although the Semenov-2 field is basalt hosted, several geochemical features of the massive sulfides studied are similar to those of the Mid-Atlantic Ridge ultramafic-hosted Cu-Zn–rich massive sulfides, such as Fe:Cu:Zn ratios close to 1:1:1, high Sn, Se, Au, and Ag contents, and high Au/Ag ratios. However, the strong enrichment in SiO2, the moderate Mn and Co contents, very low Ni contents, and the Co/Ni ratio >1 are more consistent with a mafic signature. Thermodynamic modeling of hydrothermal fluids produced by reactions between various proportions of seawater and basalt or peridotite at 350°C shows that mineral assemblages broadly similar to those of the Semenov-2 deposit can precipitate from fluids produced in a mafic environment, but that Au and Ag minerals are not predicted to precipitate from such fluids over a wide temperature range. These results suggest that an additional contribution to the hydrothermal system is required in order to achieve saturation in precious metals. A magmatic input is suggested by the occurrence of plagiogranites and tonalites dredged on sea floor in the Semenov area, which could be a potential source of Au-rich magmatic fluids, and by mineralogical and geochemical similarities with magma-related, low-to intermediate-sulfidation epithermal systems, namely high Au and Ag grades, high Au/(Cu + Zn + Pb) and Au/Ag ratios, and presence of Ag, Bi, and Te minerals. The likely crucial role of silicic melts in producing high Au and Ag grades suggests that exploration for precious metal-rich, volcanic-hosted massive sulfide deposits should be primarily directed to sites in which evolved igneous rocks occur on sea floor. Both in modern and ancient mafic-hosted deposits, zones characterized by abundant deposition of silica could be good clues to the presence of significant gold.