INSIGHTS INTO LOW-TEMPERATURE CHEMICAL ALTERATION OF BASALT USING MASS ANALYSIS: APPLICATION TO LAVAS FROM THE COLUMBIA RIVER AND DECCAN FLOOD BASALT PROVINCES

The “mass analysis” methodology developed from evaluation of Columbia River Basalt (CRB) lava chemistries (Sawlan, 2018, Geosphere) is based on mass balance relationships that allow the determination of material compositions removed by mineral ± glass dissolution, the proportions of reactive phases, and the parent element abundances for samples mass-depleted by water-rock interaction. A key part of this method, called “mass normalization”, eliminates the inherent normalization of analyses caused by the analysis of mass-depleted samples as a unit mass, and gives “mass-normalized abundances” (MNA). MNA are the product of analyzed abundances (normalized to 100% volatile-free) and the remaining sample mass fraction (M) as calculated from a parent:sample immobile element ratio. Prior expressions assessing chemical alteration incorporate mass normalization (e.g., tau (Anderson et al., 2002, GSAB), but also include additional terms emphasizing relative changes. Direct evaluation of MNA (in wt%, ppm) as a function of M for mobile elements shows linear correlations over as much as 40 wt% sample mass loss. For each element the slope gives the weight proportion or “transport abundance” of the entire dissolved material, termed the “transport composition”. These correlations indicate that steady state dissolution is the dominant process modifying sample chemistry during low-T water-rock interaction.

Linear MNA-M correlations are observed for CRB altered under either anoxic or oxic conditions. Transport compositions, however, differ significantly between these two environments and indicate differences in phase reactivity as well as redox-sensitive element solubility (e.g., Fe). Notably, glass and plagioclase under anoxic conditions are nearly unreactive, but both phases under oxic conditions are strongly reactive. Samples from paleosols in Deccan basalt (Babechuk et al., 2014, Chem. Geol.) give transport compositions similar to those for CRB paleosols, although for samples with >30 wt% mass loss some elements’ MNA deviate from linear trends. This analysis also reveals low-level erratic enrichments (ppm to tenths wt%) in some minor and trace elements in CRB and Deccan lavas. Their influence on the overall sample chemistry, however, is minimal compared to that of steady state dissolution.