Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX, I) and octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) hydrolyze at pH > 10 to form end products including NO2-, HCHO, HCOOH, NH3, and N2O, but little information is available on intermediates, apart from the tentatively identified pentahydro-3,5-dinitro-1,3,5-triazacyclohex-1-ene (II). Despite suggestions that RDX and HMX contaminated groundwater could be economically treated via alkaline hydrolysis, the optimization of such a process requires more detailed knowledge of intermediates and degradation pathways. In this study, we hydrolyzed the monocyclic nitramines RDX, MNX (hexahydro-1-nitroso-3,5-dinitro-1,3,5-triazine), and HMX in aqueous solution (pH 10-12.3) and found that nitramine removal was accompanied by formation of 1 molar equiv of nitrite and the accumulation of the key ring cleavage product 4-nitro-2,4-diazabutanal (4-NDAB, O2NNHCH2NHCHO). Most of the remaining C and N content of RDX, MNX, and HMX was found in HCHO, N2O, HCOOH, and NH3. Consequently, we selected RDX as a model compound and hydrolyzed it in aqueous acetonitrile solutions (pH 12.3) in the presence and absence of hydroxypropyl-beta-cyclodextrin (HP-beta-CD) to explore other early intermediates in more detail. We observed a transient LC-MS peak with a [M-H] at 192 Da that was tentatively identified as 4,6-dinitro-2,4,6-triaza-hexanal (O2NNHCH2NNO2CH2NHCHO, III) considered as the hydrolyzed product of II. In addition, we detected another novel intermediate with a [M-H] at 148 Da that was tentatively identified as a hydrolyzed product of III, namely, 5-hydroxy-4-nitro-2,4-diaza-pentanal (HOCH2NNO2CH2NHCHO, IV). Both III and IV can act as precursors to 4-NDAB. In the case of the polycyclic nitramine 2,4,6,8,10,12-hexanitro-2,4,6,8,10,12-hexaazaisowurtzitane (CL-20), denitration (two NO2-) also led to the formation of HCOOH, NH3, and N2O, but neither HCHO nor 4-NDAB were detected. The results provide strong evidence that initial denitration of cyclic nitramines in water is sufficient to cause ring cleavage followed by spontaneous decomposition to form the final products.

The sorption−desorption characteristics of 2,4,6-trinitrotoluene (TNT), 4-amino-2,6-dinitrotoluene (4-ADNT), and 2,4-diamino-6-nitrotoluene (2,4-DANT) within a natural topsoil, an illite shale, and a sandy aquifer material (Borden sand) were studied. The sorption capacity constant ( ) of the three nitroaromatic compounds (NACs) increased with the number of amino groups (i.e., 2,4-DANT > 4-ADNT > TNT) for topsoil, and there was significant sorption−desorption hysteresis. Traces of 4-N-acetylamino-2-amino-6-nitrotoluene (4-N-AcANT) formed during sorption of 2,4-DANT by nonsterile topsoil (22 h), but this did not account for the hysteresis. For longer contact times (66 h), 4-N-AcANT accounted for 26% of the biotic disappearance of 2,4-DANT, and traces of 2-N-acetylamino-2-amino-6-nitrotoluene (2-N-AcANT) were detected. For illite, the increased with the number of nitro groups (i.e., TNT > 4-ADNT > 2,4-DANT), and there was also sorption−desorption hysteresis. Most of the 2,4-DANT was neither desorbed nor extractable by acetonitrile from illite or topsoil. Sorption of the NACs by Borden sand was slight or nonexistent. This study illustrates that soil and NAC type will have a significant effect on the as well as the formation of acetylated metabolites.

A separation method using cyclodextrin-modified capillary electrophoresis has been developed for analysis of the 16 U.S. EPA priority pollutant PAHs in contaminated soils. In this procedure, a mixture of negatively charged sulfobutyl ether-β-cyclodextrin (SBβCD) and neutral methylβ-cyclodextrin (MβCD) was added to the running buffer, and separation was effected on the basis of differential distribution (partitioning) of the PAH components between the two CD types. Satisfactory separation of all 16 PAHs was achieved in under 20 min with 35 mM SBβCD and 15 mM MβCD, with efficiencies for all components greater than 10 5 theoretical plates, as expected for CE methods. Injection and electrophoresis methods were identified such that samples provided in dichloromethane were diluted 10-50-fold in methanol and injected directly. Laser-induced fluorescence detection provided sensitive detection of 11 of the 16 components, with detection limits measured typically in the low microgram per liter (ppb) range. Contaminated soil was extracted using CO 2 supercritical fluid, and the extract was provided in dichloromethane, methanol, and a 50/50 mixture. The diluted extract was analyzed using fluorescence detection, and 10 of the 11 compounds were detected above the detection limit. HPLC was used to verify the concentration of six of these components, giving fairly good agreement, though with a possible systematic underestimation of the concentration by the CE method.