Mobility and Retention of TNT AND RDX IN SOILS

I.K. Iskandar and H.M. Selim

Transport experiments were conducted to describe the potential mobility of TNT (2,4,6-trinitrotoluene) and RDX (2,3,5-trinitro-1,3,5-triazine) in a pure system (montmorillonite clay) and two selected soils (Norwood and Kolin) which differ in their clay content, pH, and organic carbon content. The kinetics of release of TNT, RDX and other explosive contaminants from the Louisiana Army Ammunition Plant (AAP) contaminated surface (Kolin) soil were also investigated. Transport experiments were based miscible displacement techniques in uniformly packed soil columns under steady flow conditions. Adsorption and desorption of TNT and RDX was also measured using batch methods for a range of input concentrations. Adsorption isotherms for RDX appeared to follow a linear trend for Norwood and Kolin soils with much lower retention than for TNT. Isotherms for TNT exhibited distinct nonlinearity for Norwood and Kolin soils and to lesser extent for SWy-1 montmorillonite. The Freundlich model provided best fit of TNT isotherms. Results of retention kinetics indicated that there was continued decrease with time in TNT concentration for Norwood and Kolin soils. This was perhaps due to the formation and subsequent adsorption of transformation products. In contrast, for SWy-1 montmorillonite, TNT retention was rapid initially and appeared to reach equilibrium within 1 d. Unlike Kolin and Norwood soils, for SWy-1 montmorillonite there was a lack of nonsingularity or hysteretic behavior of TNT adsorption-desorption isotherms and a mass balance calculation indicated fully reversible reaction mechanisms.

Transport results from bentonite/sand miscible-displacement columns indicated that TNT was a highly mobile contaminant and was considered fully conservative in the presence of methanol as the background solution. In contrast, TNT was strongly retained with as much as 50% of amount applied within the Kolin, Norwood, bentonite/sand columns for a prolonged period when the applied TNT pulse was in 0.005 M Ca(NO3)2 as the background solution. Transformations of TNT to 4-D-Am-DNT and other analytes which were present and identified in the effluent solution was postulated. In contrast, for all miscible displacement experiments there was only a limited retention of RDX during transport in the soil columns. Transport results provided the data sets necessary to examine the applicability of a mechanistic nonlinear multireaction approach which incorporates kinetic and equilibrium retention processes into the convection-dispersion transport equation to describe TNT and RDX transport in soils.

I.K. (Alex) Iskandar
Chief, Geochemical Sciences Division
U.S.Army Cold Regions Research and Engineering Laboratory (CRREL)
72 Lyme Road
Hanover, NH 03755-1290
phone: 603-646-4198
fax: 603-646-4561
e-mail: iskandar@crrel.usace.army.mil

H.M. Selim
Louisiana State University
Baton Rouge, LA 70803
e-mail: Selim@LANMAIL.OCS.LSU.EDU