Abstract: Abstract:To determine the bioaccessible fraction of PAHs concentration, a desorption experiment assisted by XAD-2 resin was employed to characterize the desorption process and bioaccessibility of high molecular weight PAHs (HMW PAHs) in the soils of a coke and chemical plant. In order to evaluate the sorption of PAHs with different functional groups in the soils, FTIR technique was used to characterize the infrared spectrogram of the soil samples before and after the desorption experiments. The desorption leveled off after 60 days, and the rates in the first 30 days (fast desorption phase) were two to three orders of magnitude higher than the ones in the last 60 days (slow desorption phase), which could be regressed by dual-phase model. Bioaccessibilities of the corresponding HMW PAHs were 0.18-0.47 in the soil samples contaminated through leakage of PAHs-containing liquid, while they were 1.00 in the surface soil samples contaminated via precipitation. This suggests that it may be conservative to simply assume bioaccessibility in all soil is 100%. The difference between infrared spectrograms of the samples before and after desorption was negligible, indicating the main function groups in soil organic matters was not changed during the experiment. However, the peaks of hydroxyl and carboxyl in the spectrograms shrunk, implying that PAHs adsorbed on these function groups might be more bioaccessible. The results suggest that bioaccessibility should be considered in a specific site to reduce the conservativeness of health risk assessment.