Experimental constraints on stable potassium (K) isotope fractionation during phase separation in NaCl–KCl–H2O and KCl–H2O systems: implications for the K isotope composition of seafloor hydrothermal vent fluids

Abstract

Phase separation is a ubiquitous process in marine hydrothermal systems, significantly influencing the chemical and isotopic composition of vent fluids. Understanding its effects on elemental partitioning and isotopic fractionation is essential for using vent fluid chemistry as an accurate predictor of the chemical and physical conditions at depth in the oceanic crust. Here, we report the first laboratory experimental study investigating stable potassium isotope (⁴¹K/³⁹K) fractionation during phase separation of alkali (Na, K) chloride fluids under temperature and pressure conditions relevant to natural hydrothermal systems. Two distinct fluid compositions were tested in our experiments at 400 °C, including a mixed NaCl–KCl solution with a starting Na/K molar ratio of 10, and a pure KCl solution. This compositional difference allows for evaluating the role of Na on K isotope fractionation between coexisting vapor and liquid phases. For the NaCl–KCl–H₂O system, all experiments showed preferential enrichment of light K isotopes in the vapor phase during phase separation, yielding an average K isotope fractionation factor of −0.12 ‰ (±0.04 ‰, 1SD). In sharp contrast with the NaCl–KCl–H₂O experiments, the Na-free KCl–H₂O experiments revealed no measurable K isotope fractionation during phase separation. These results suggest that Na plays a critical role in modifying K bonding environments, likely through the formation of multi-cation polynuclear species in the concentrated liquid phase. The coexistence of the multi-cation polynuclear species in the liquid and compositionally simpler K free ions or KCl°(aq) ion pairs in the vapor may account for the observed K isotope fractionation in the NaCl–KCl–H₂O system. The absence of Na in the KCl–H₂O system prevents significant differences in K bonding environments between vapor and liquid phases, thereby precluding measurable K isotope fractionation. Using the K isotope fractionation factor quantified here, a Rayleigh fractionation model can successfully explain the previously unexplained low δ⁴¹K values (∼−0.8 ‰) reported for vapor-dominated fluids at Main Endeavour Field (NE Pacific Ocean). These results highlight the importance of considering phase separation alongside seawater interaction with the oceanic crust to fully understand K isotope variations in hydrothermal systems.