Coupling in vitro assays with sequential extraction to investigate cadmium bioaccessibility in contaminated soils

To understand how Cd in different fractions contributes to Cd bioaccessibility by in vitro assays, Cd bioaccessibility in 12 contaminated soils was determined by four assays (UBM, SBRC, IVG, and PBET) and correlated with different Cd fractions based on a sequential extraction scheme. The Cd bioaccessibility in the gastric phase (GP) was high (35-107%, averaging at 77%), implicating high risk to human health, while it decreased to 19-88% averaging at 47% in the intestinal phased (IP). From the GP to IP, the reduction of extractable Cd (0.45-48 mg kg^-1) and Fe (118-3884 mg kg^-1) showed significant correlation (R = 0.54-0.74) via UBM, SBRC, and IVG, suggesting co-precipitation with Fe and/or sorption onto Fe oxides maybe responsible for decrease in Cd bioaccessibility. Although Cd bioaccessibility varied among assays, their results show some consistency based on their correlation in the GP (R = 0.56-0.90) and IP (0.34-0.73, excluding UBM-IP and PBET-IP). Sequential extraction data show that Cd was primarily associated with the exchangeable fraction (E1; 7.05-72.9%, averaging 39.4%). The carbonate (C2; 6.86-44.8%, 21.9%) and Fe/Mn oxides fraction (F3; 12.5-53.6%, 28.2%) were similar, while organic (O4; 0.62-25.0%, 7.91%) and residual fraction (R5; 0.22-8.54%, 2.62%) were the lowest. Significant correlation (R = 0.59-0.88) between the first two fractions (E1+C2) and bioaccessible Cd suggest they were the main sources of bioaccessible Cd in those contaminated soils.