Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Appendix D Results from Technical Risk Reduction Program Activity 2a, Phase II, for GB and Activity 11 for VX TRRP 2a, Phase II, For GB AnaLysis lize the GB. In all cases, the aqueous phase constituted >98 percent of the total hydrolysate volume, and the percentage The performance of the new analytical method was constituted by the organic phase was very small. Methylene adequate to ensure that measured GB concentrations in the chloride extraction of the hydrolysates followed by gas hydrolysate were low enough for secondary treatment using chromatography and mass spectrometry (GC/MS) showed supercritical water oxidation (SCWO). Method detection tributylamine, diisopropylmethyl phosphonate, dibutyl limit (MDL) values in hydrolysate were 2.2 µg/L (2.2 ppb) acetamide, andin the case of the diisopropyl carbodiimide with 68 percent recovery and were calculated in accordance stabilized materialdiisopropylurea and diisopropylnitros with standard U.S. Environmental Protection Agency (EPA) amine. Residual GB was not mentioned. methods (40 CFR Part 136, Appendix B). This value is well Method application tests of hydrolysate produced on below the release criteria of 75 ppb for GB in the hydroly- the bench showed that in 9 of 10 batches, GB destruction sate. A more significant performance parameter is the target exceeded 99.9999 percent with greater than 95 percent action limit (TAL), which is the concentration for which 95 confidence. In one batch, residual GB was detected at con- percent of the measurements will be below the release cri- centrations ranging from 61 to 74 ppb. This showed that the teria (Malloy et al., 2007). In an analysis of GB performed method was effective for analyzing the hydrolysate. on hydrolysate generated from two batch reactor studies An issue of significant concern was whether GB could conducted at Battelle, the TAL values were calculated at 57 be analyzed in the solids that are found in the munitions. ppb and 52 ppb. Crystals that tend to form in the munitions have been shown The study also showed that the destruction method was to be diisopropylurea, which is derived from diisopropyl car- effective for destruction of GB in a number of other matrices, bodiimide (Rosso et al., 2005). The TRRP study showed that including the diisopropylurea crystals that form from the the analytical method worked well for detecting GB in the diisopropyl carbodiimide during storage. diisopropylurea crystals, and that GB associated with the di- A series of batch reactor studies were conducted to isopropylurea crystals was hydrolyzed (Malloy et al., 2007). validate the Blue Grass Chemical Agent Destruction Pilot For diisopropyl urea crystals that had been washed with Plant (BGCAPP) GB neutralization process (Malloy et al., caustic, residual GB was detected at high concentrations, 2007). Three different GB batches were used that had con- on the order of 70 to 450 ppm. When the diisopropylurea centrations ranging from 75 percent to 90 percent, and di- crystals were mixed with GB/diisopropyl carbodiimide and isopropylmethyl phosphonate and the stabilizer compounds then added to the reactor, no GB was detected. made up the remainder of the material. The batch tests used The analytical method also worked well for measuring 6 percent sodium hydroxide; the pH at the end of the batch GB in the energetics neutralization hydrolysate. BGCAPP- runs was ~12. 104B had an MDL of 13 ppb, based on an analysis of spiked At the conclusion of the hydrolysis experiments, an or- enegetics hydrolysates from the neutralization reactor. ganic layer was present. The composition of this layer (pres- The TRRP study also showed that GB could be re- ent in some hydrolysates) was analyzed not by Battelle but formed in the SCWO feed, but only when pH was lowered by the Armyâs Edgewood Chemical and Biological Center, to acidic values. At longer reaction times, however, the GB which reported that the layer was largely tributylamine and concentration begins decreasing again. This is explained in some additional compounds that are manufacturing impuri- terms of GB re-formation upon acidification, followed by ties in the tributylamine that was used by the Army to stabi- acidic hydrolysis. GB re-formation without pH adjustment 61
62 Review of secondary waste disposal Planning was evaluated by analysis of hydrolysate from one of the 1 g/L (1,000 ppm), which must be demonstrated for transfer method application tests 1 month after the hydrolysis was of the hydrolysate from the ANR to the SCWO. However, performed. One of the reanalyses measured GB at 5.7 ppb, the method is temperamental in that the baseline detector which suggested that slow re-formation might be taking response is substantial and varying, retention times vary, place. peak shapes are variable, and the detector is nonspecific. The TRRP also evaluated the performance of the GB Nevertheless, analysis of simulated ANR hydrolysate spiked analytical method for measuring agent in SCWO and reverse with 400 ppb EA2192 resulted in detection with good P&A osmosis (RO) effluents. Analysis of GB in simulated SCWO in analysis of hydrolysate derived from two VX batches. effluent using method BGCAPP-104b reproducibly mea- Re-formation of VX in the hydrolysate was a concern sured GB at concentrations below 5 ppb, with a calculated based on reactions of hydrolysis products catalyzed by sta- MDL of 0.63 ppb. Similar results were achieved for the RO bilizers (Brickhouse et al., 1998). This is considered more rejectate. These values are well below the target value of 20 likely to occur as the pH decreases from the high values (>12) ppb. When blended hydrolysate was evaluated, an MDL of found in the unmodified hydrolysate. As in previous cases, 1.7 ppb was measured. analysis was complicated by analytical difficulties stemming from the tendency of VX to protonate at pH < 12, which re- sults in inefficient extraction. Therefore, BGCAPP-204 was Results from TRRP 11 For VX AnaLysis modified by raising the pH of the extract above 7 to drive the A TRRP was conducted to evaluate the analytical VX into the organic phase. Using this analytical approach, methods for measuring VX in the matrices anticipated at no VX was detected in the hydrolysate for up to 60 days. BGCAPP and the effectiveness of agent destruction. A However, because the recoveries were low, the possibility of precision and accuracy (P&A) study of the refined method re-formation in an acidified hydrolysate could not be com- (referred to as BGCAPP-204) was performed by conducting pletely discounted. Reanalysis using a liquid chromatography triplicate analyses on a hydrolysate spiked at four different method also failed to detect VX in the hydrolysate. However, levels. The P&A study produced a detection limit of 4 µg/L the TRRP report (Dejarme and Lecakes, 2008) stated that (ppb) and a limit of quantification (LOQ) of 10 ppb for VX. the study did not adequately represent process conditions The TAL, which accounts for the method imprecision, was planned for BGCAPP, and the authors recommended that a measured at 107 ppb, a value that is well below the 160 ppb more detailed re-formation study be performed. acceptance criteria of the SCWO. When the method was ap- Residual VX and EA2192 were also evaluated in the plied to hydrolysates from the three different VX batches, de- SCWO effluent and in the RO rejectate. To make the mea- tection limits were similarly low, ranging from 9 to 28 µg/L. surements, the analytical method for VX needed to be modi- A modified 12-sample P&A study of hydrolysates from four fied, because the effluents were slightly acidic, which im- different sources of VX showed an average detection limit of pedes extraction of VX. These modifications were effective, 4 ppb, an LOQ of 8 ppb, and a TAL of 125 ppb. When the resulting in VX MDL values of 14 and 12 ppb for the SCWO P&A study was extended to 48 samples over 4 days, the mean and RO effluents, respectively. For EA2192, MDL values of MDL was 16 ppb, the LOQ was 35 ppb, and the TAL was 290 and 470 ppb were achieved for the SCWO and RO ef- 107 ppb. A final test of analytical efficacy involved analysis fluents. The SCWO VX method is called BGCAPP-604, and of simulated agent neutralization reactor (ANR) hydrolysate the method modified for EA2192 is BGCAPP-704B. spiked with 80 ppb VX. In this study, VX was detected with Members of the public expressed concern about the good P&A in hydrolysate samples derived from two separate different chemistries of the two layers that emerge in VX VX batches. These studies indicated that the instrumental hydrolysate. Specifically, there is uncertainty over whether method used for clearing the VX hydrolysate for further VX could be present in the organic layer. In examining the SCWO treatment is adequate. hydrolysates from the four batches of VX, one of the four For instrumental analysis of EA2192, a liquid chro- had a layer that appeared to account for several percent matography method was developed that utilized either a of the total volume, while each of the other three had an diode array detector or an ultraviolet detector. Initial studies organic layer that accounted for a much lower fraction. In were conducted using liquid chromatography/electrospray TRRP activity 11, GC/MS analysis of the upper organic ionization/MS/MS, with the objective of confirming peak layer showed that it consists mainly of the disulfide, bis[2- identification, which cannot be done with confidence using (diisopropylamino)ethyl] disulfide, with lesser amounts of an ultraviolet detector alone. Once the peaks were identi- the thiolamine and related thiols, sulfides, and conjugates of fied in the liquid chromatogram, studies were conducted to those molecules with the stabilizer diisopropyl carbodiimide. evaluate the performance of the diode array detector method VX was not detected in this analysis; however, the MDL for (BGCAPP-304B), which produced detection limits averag- this approach is not known. ing 61 ppm. When the ultraviolet light detector was used, The TRRP activity 11 report (Dejarme and Lecakes, detection limits were calculated that ranged from 18 to 106 2008) concluded with remarks that while the BGCAPP VX ppm. These levels are much lower than the clearance level of clearing method was working, it should be tested to evaluate
APPENDIX D 63 its robustness in an actual plant neutralization environment. Dejarme, L., and G.D. Lecakes. 2008. Bench-Scale Evaluation of VX Furthermore, extensive testing to characterize the potential for Hydrolysis, TRRP #11, Test Report, Rev. 0. Aberdeen Proving Ground, Md.: Program Manager for Assembled Chemical Weapons Alterna- VX re-formation was beyond the scope of TRRP activity 11. tives. Malloy IV, T.A., L. Dejarme, C. Fricker, J. Guinan, G.D. Lecakes, and A. Shaffer. 2007. Bench-Scale Evaluation of GB Hydrolysis, TRRP #02a References Phase II, Test Report, Rev. 0. Aberdeen Proving Ground, Md.: Program Brickhouse, M.D., B. Williams, D. McGarvey, H.D. Durst, and R.J. Manager for Assembled Chemical Weapons Alternatives. OâConnor. 1998. Coupling reactions between dialkylcarbodiimides, Rosso, T.E., P.L. Abercrombie, A.B. Butrow, G. Hondrogiannis, J.M. phosphonates, and thiols. Pp. 617-623 in Proceedings of the ERDEC Lochner, J.J. Loss, R.J. Malecki, J.M. Meuser, D.K. Rohrbaugh, and Scientific Conference on Chemical and Biological Defense Research. Y.-C. Yang. 2005. Characterization of Solid and Liquid GB Samples Report number ECBC-SP-004. Aberdeen Proving Ground, Md.: Edge- Collected from M55 Rockets Processed at Anniston Chemical Agent wood Chemical and Biological Center, U.S. Army Soldier and Biologi- Disposal Facility (ANCDF). Aberdeen Proving Ground, Md.: Edgewood cal Chemical Command. Chemical and Biological Center.
