Isotopic composition analysis contributes significantly to the investigation of the biogeochemical cycle of zinc with its economical, environmental and health implications. Interpretation of isotopic measurements is however hindered by the lack of a set of equilibrium isotopic fractionation factors between Zn-bearing minerals. In this study, equilibrium mass-dependent Zn isotope fractionation factors in Zn-bearing minerals are determined from first-principles calculations within the density functional theory (DFT) scheme. A wide range of minerals belonging to sulfide, carbonate, oxide, silicate, sulfate and arsenate mineral groups are modelled to account for the natural diversity of Zn crystal-chemical environment. Calculated reduced partition function ratios (β-factors) span a range smaller than 2 at 22 • C. All studied secondary minerals (adamite, gahnite, gunningite, hemimorphite, hydrozincite, zincite) but zinc carbonate (smithsonite) are isotopically heavier than zinc sulfide minerals (sphalerite and wurtzite) from which they could form through supergene processes. Zinc-aluminium spinel and zinc silicate are the isotopically heaviest minerals. The investigation of the crystalchemical parameters at the origin of differences in isotopic properties shows an excellent linear correlation between ln β and Zn interatomic force constants. β-factors are also observed to increase when the Zn-first neighbour bond lengths decrease and charges on atoms involved in the bonding increase and vice versa. These findings are in line with the observation of heavy isotope enrichment in systems having the largest bond strength.

Theoretical mineral-solution equilibrium isotopic fractionation can contribute to the interpretation of Zn isotopic compositions. In this study, we investigate equilibrium isotopic fractionation properties of hexaaquo zinc complex, a major Zn aqueous species, using first-principles molecular dynamics (FPMD) based on density functional theory (DFT). Pentaaquo and tetraaquo zinc complexes, which can be relevant for specific mechanisms such as adsorption on mineral surfaces, are also considered. This approach takes into account both configurational and solvation effects that are less easily accounted for in molecular cluster models. The logarithmic value of the reduced partition function ratio of aqueous Zn (ln β 66 Zn/ 64 Zn) is 2.86 ± 0.04 at 295 K, while the same computational approach gave ln β 66 Zn/ 64 Zn in the range 2.2 − 4.1 for various Znbearing minerals. For example, calculations predict at 295 K, equilibrium fractionations between mineral and aqueous Zn of-0.46 for primary zinc sulfide sphalerite and of-0.20 and +0.34 for secondary phases gunningite and hydrozincite, respectively. Molecular clusters are also modelled, predicting isotopically heavier Zn with respect to related FPMD models (ln β increases by 0.09 to 0.23). These values are small but significant in the Zn isotopic system and supports the idea that a proper description of the dynamics of the system and the solvation effect are required for a reliable prediction of the isotopic properties of solvated ions.