Quantifying the bioaccessibility of chromium contaminated soil

There are numerous sites on Department of Defense (DOD) and Department of Energy(DOE) lands along with other industrial facilities that are contaminated with toxic metals such as chromium and are awaiting possible clean up and closure. Ingestion of contaminated soil by children is the usual risk driver that motivates remediation for sites contaminated with toxic metals. The purpose of this study was to investigate the effect of soil properties on the bioaccessibility of Cr(III) and Cr(VI) as a function of contaminant concentration and aging, and to develop a simple statistical model based on common soil properties to estimate the bioaccessibility of Cr(in) contaminated soil upon ingestion.The A- and upper B- horizons of two well characterized soils were treated with varying concentrations of Cr(IlI) and Cr(VI) and allowed to age. The bioaccessibility of the contaminated soils was measured using a physiologically based extraction test (PBET)designed to simulate the digestive process of the stomach. Sorption of Cr(III) and Cr(VI)varied significantly as a function of soil type and horizon as well as the oxidation state of the contaminant. Cr(lII) showed significantly greater solid phase concentrations thanCr(VI) for any given initial Cr concentration. These results are consistent with one of the expected mechanisms of sequestration of Cr(III) vs. Cr(VI) by the soils: the precipitation of Cr(III) - hydroxides, which can result in the accumulation of large mass fractions of contaminant on mineral surfaces. Overall, bioaccessibility of Cr decreased as the duration of exposure to the soil increased. This occurred for all soils and at all solid phase concentrations, with Cr(III) exhibiting the most pronounced aging effects. The Decrease in Cr bioaccessibility was rapid for the first 50 d and then slowed dramatically from 50 to 200 d.The statistical model for bioaccessibility of Cr(III) was developed using data from thirty-six uncontaminated soils from seven major soil orders which were spiked with Cr(III),aged and then tested using the PBET method. The bioaccessibility of Cr(III) varied widely as a function of soil type with most soils limiting bioaccessibility to < 45% and <30% after 1 and 100 days aging, respectively. Statistical analysis revealed that Cr(III)sequestration by the soils was strongly correlated with the clay, total inorganic carbon(TIC), pH, and Mn-oxide content of the soils. Soils that contained higher quantities of clay, inorganic carbon (i.e. carbonates), higher pH and low Mn-oxide contents generally sequestered more Cr(III). Further statistical analysis showed that the bioaccessibility ofCr(III) on soil was also correlated with the clay content, TIC content, and pH of the soil with bioaccessibility decreasing as the soil TIC content and pH increased and as the clay content decreased. The model was statistically rigorous, passing both a normality test and a constant variance test, with no indication of parameter collinearity. The model yielded an equation based on common soil properties that could be used to predict theCr(III) bioaccessibility in soils with a reasonable level of confidence. Thus, the model should be useful for assessing Cr(III) bioaccessibility at contaminated site so that the appropriate remediation criteria can be determined.