Epsilon Archive for Student Projects

Abstract

Arsenic (As) is a toxic element to all life and exists in nature both naturally and anthropogenically. Naturally As contaminated soils are formed due to high geogenic background As concentrations which is released when bedrock is eroded. Asia is a continent with a high amount of naturally As contaminated soils and groundwater. Anthropogenically As contaminated soils can originate from several sources such as industry, agriculture, burning of As containing fossil fuels etc. The main issue with As contaminated soils is that it ends up in the groundwater contaminating it and possibly in turn local drinking water. In nature, As exists predominantly as either As(III) in anoxic conditions or as As(V) in oxic conditions. As(III) is more soluble than As(V) leading to As being more soluble in anoxic conditions. The use of ZVI to immobilize As in the ground is a remediation technique that has become a more common the past decades. It is a technique that is favoured because of several factors, some being its efficiency, economical cost, and simplicity to use. ZVI also has the ability of effectively immobilizing As in oxic and anoxic conditions. In this report the As solubility in untreated and ZVI (mZVI and nZVI) -treated samples was analysed. The samples were collected in an anoxic aquifer within a CCA (Chromated copper arsenate) -contaminated area in Hjältevad. First, experiments were carried out to determine if phosphate and oxalate extractions are a good method in determining how As is bound in untreated and ZVI-treated sediment samples. Second, batch experiments were carried out to first determine the reaction kinetics of the experiment. Then, a pH-dependent As solubility experiment with different amounts of added As was carried out. Results from the extraction experiment and batch experiment were then used as input parameters in the chemical equilibrium software Visual MINTEQ. The purpose of this was to see if it was possible to find a model that fit the measured results by adjusting the available As for reaction and the amount of active ferrihydrite for the As to bind to. The results from the phosphate extraction were unexpected and showed an increase of As for ZVI-treated samples over time. This increase was also observed for Fe. This released Fe was most likely amorphous Fe that also had As bound to it, which lead to an increase of As over time. The reason for the amorphous Fe to be released into solution is not fully understood but the high concentration of PO43- (0.5 M) in the solution could over time perhaps interrupt the structure of the amorphous Fe. The geochemical modelling proved to be a challenge and no model was able to be made that fit the measured values with a great degree. The results did indicate that all the oxalate-extractable As was taking part in the equilibrium reactions. However, since the models did not fit perfectly it is difficult to determine if this is the actual case. The best fit model indicated that about 54% of the oxalate-extractable ferrihydrite is active and takes part in the reactions.
Keywords: Hjältevad, zerovalent iron, phosphate extraction, Oxalate extraction, Visual MINTEQ, geochemical modelling