Mineralogy – the key to understanding and predicting the elution behavior of arsenic

Arsenic is a world-wide serious health problem occurring in various types of rocks, soil, and contaminated water resources. However, it is not always an acute health problem, as concentrations are often comparably low. But recently, it gained increasing attention with the introduction of new laws in Germany (and the EU) regarding the deposition of excavated sediments, e.g., in the context of the construction of tunnels or the mining of gravel and sand etc. The decision on whether excavated material can be used as backfill or needs landfilling depends on the analysis and classification relative to specified limit values. In the latter scenario, this can result in a significant increase in costs. It is mandatory to determine the heavy metals and arsenic (among other parameters) in the solid sample and to produce and analyze an eluate of the sample to identify soluble harmful elements that could potentially contaminate the groundwater.

At a first glance, arsenic contents in solid and in eluate often show erratic patterns, which could not be explained. In clastic sediments, an often observed (weak) correlation between arsenic and iron in solid analyses led to the assumption that arsenic is bound in iron-hydroxides. However, a comprehensive examination of the chemistry and mineralogy of a large number of samples did not reveal any clear correlations. Therefore, we studied drill-core samples and sediment profiles of poorly consolidated Miocene clastic sediments, including gravel, sands, clay and silt from different areas of the Northern Alpine foreland basin (mostly from the Munich and Ingolstadt areas).

Based on geochemical analyses, we found different “arsenic-types”: Some samples show the expected correlation pattern for arsenic in the solid and the eluate. Some samples, however, show high arsenic in the solid and low arsenic concentrations in the eluate. Also, remarkably, samples with rather low arsenic contents in the solid and high arsenic in the eluate were observed. No systematic correlation with other chemical elements or with macroscopic characteristics, e.g. grain size, could be identified. Our detailed mineralogical investigation of more than 30 samples showed that XRD analysis, which is usually used to identify the mineralogy in the finest fraction, is not sufficient to explain this behavior. Therefore, we separated all mineral phases (also in the clay fraction) and analyzed their mineral chemistry (in particular their arsenic content) with SEM, their texture with high-resolution Keyence microscopy (2000x), and combined the results with extensive leaching experiments.

Our results imply that mineralogy is the key to understanding the elution behavior of arsenic. Regarding the binding characteristics of arsenic, three different mineral types can be distinguished: 1) Fe-, Si- and Al-Hydroxides (minor tourmaline, apatite and zircon) can bind arsenic relatively well (no arsenic in the eluate), 2) chlorite and mica can adsorb high amounts of arsenate but with a weak bond (high As in eluate) and 3) smectite which releases its arsenate step-by-step with the increasing degree of swelling. The study showed that only with a detailed mineralogical and mineral-chemical analysis profound predictions on the elution behavior of arsenic can be made.