Methods for Enhancing the Desorption of High Explosives from Activated Carbon

Adsorption to granular activated carbon (GAC) is an effective method for removing the high explosives RDX and HMX from contaminated water. However, the process is relatively expensive because of replacement and disposal costs for spent GAC. One potential method for reducing operating costs is off-line bioregeneration, in which contaminants are desorbed from GAC and biodegraded in a separate reactor. Because slow desorption kinetics may impede the regeneration of spent GAC, several methods for increasing the rate and extent of desorption of RDX and HMX were explored. The specific objectives of this research were to determine the effectiveness of several enhanced desorption methods and to determine if existing adsorption models could accurately predict enhanced desorption of contaminants in small-scale columns.Batch tests were used to test the effectiveness of cosolvents, surfactants, and cyclodextrins in desorbing RDX from GAC. At an initial RDX loading of 71.1 mg/g, only 3% of the adsorbed RDX was desorbed by water over 11 days, whereas 92.6% of the RDX was desorbed by 100% ethanol over the same period. Lower concentrations of ethanol in solutions with water were less effective in desorbing the RDX. At all concentrations in water, methanol was somewhat less effective than ethanol. One anionic and three nonionic surfactants were tested at three concentrations each. Sodium dodecyl sulfate (SDS) was the most effective of the tested surfactants. At a concentration of 500 mg/L (about 20% of its critical micelle concentration, or CMC), SDS desorbed 56% of the RDX. At a concentration of about twice its CMC, SDS desorbed slightly more than 70% of the RDX. β-cyclodextrin provided minimal improvement over water, desorbing 5.3% to 8.1% of the RDX at concentrations of 1 g/L and 10 g/L, respectively.Two ethanol solutions (5% and 10%) and two of the surfactants (SDS and Tween 80) were used to desorb RDX from loaded GAC in small scale, continuous flow column tests. Of the enhanced desorption solutions that were tested, the 10% ethanol solution was the most effective in desorbing RDX with the least amount of fluid. To desorb 50% of the adsorbed RDX required 22,500 bed volumes of buffered water, whereas only 4,300 bed volumes of 5% ethanol and 2,100 bed volumes of 10% ethanol were required. SDS at 500 mg/L was very effective, desorbing RDX nearly as fast as the 5% ethanol solution; however, SDS precipitated in the GAC column, impeding flow. Initial column modeling efforts successfully predicted the breakthrough curve for adsorption of RDX from water and desorption of RDX from GAC with water. Two different approaches were attempted to model enhanced desorption with ethanol solutions. The approach that assumed competition between ethanol and adsorbed solutes was the more accurate of the two approaches.