TNT TRANSPORT AND FATE IN CONTAMINATED SOIL

Past disposal practices at munitions production plants have contaminat ed terrestrial and aquatic ecosystems with 2,4,6-trinitrotoluene (TNT) . We determined TNT transport, degradation, and long-term sorption cha racteristics in soil. Transport experiments were conducted with repack ed, unsaturated soil columns containing uncontaminated soil or layers of contaminated and uncontaminated soil. Uncontaminated soil columns r eceived multiple pore volumes (22-50) of a TNT-(H2O)-H-3 pulse, contai ning 70 or 6.3 mg TNT L(-1) at a constant pore water velocity. TNT bre akthrough curves (BTCs) never reached initial solute pulse concentrati ons. Apex concentrations (C/C-o) were between 0.6 and 0.8 for an initi al pulse of 70 mg TNT L(-1) and 0.2 to 0.3 for the 6.3 mg TNT L(-1) pu lse. Earlier TNT breakthrough was observed at the higher pulse concent ration. This mobility difference was predicted from the nonlinear adso rption isotherm determined for TNT sorption. In all experiments, a sig nificant fraction of added TNT was recovered as amino degradates of TN T. Mass balance estimates indicated 81% of the added TNT was recovered (as TNT and amino degradates) from columns receiving the 70 mg TNT L( -1) pulse compared to 35% from columns receiving the 6.3 mg TNT L(-1) pulse. Most of the unaccountable TNT was hypothesized to be unextracta ble. This was supported by a 168-d sorption experiment, which found th at within 14 d, 80% of C-14 activity (added as C-14-TNT) was adsorbed and roughly 40% unextractable. Our observations illustrate that TNT so rption and degradation are concentration-dependent and the assumptions of linear adsorption and adsorption-desorption singularity commonly u sed in transport modeling, may not be valid for predicting TNT transpo rt in munitions-contaminated soils.