6
Low-Temperature, Liquid-Phase Processes

This chapter reviews several chemical, biological, and other processes designed to destroy chemical agents under low-temperature conditions, those from about 20°C to less than 100°C. The lower value is room temperature. The upper value allows the use of aqueous systems at atmospheric pressure to minimize the risks of leakage of high concentrations of agent from pressurized autoclaves.

Three main types of low-temperature, liquid-phase processes are reviewed below: processes for detoxification, processes for oxidation of organic residue, and biological processes.

Detoxification processes have been the most intensively studied. They offer promising approaches for all three major agents in the U.S. stockpile (GB, VX, and H). This set of processes includes chemical processes and the use of ionizing radiation for detoxification.

In addition to detoxification by conversion of agent to other compounds, demilitarization requires the process to be essentially irreversible (see Chapter 1). This requirement will, in some cases, require a two-step sequence of initial detoxification followed by further chemical processing, which could be accomplished with additional chemical reaction but would not require complete oxidation to carbon dioxide. Detoxification reactions that produce irreversible products are therefore of most interest.

Relatively little effort has been directed to achieving complete oxidation at low pressure and temperature. However, there are several leads in this direction, arising from studies of chlorinated hydrocarbon destruction and agent decontamination. In view of recent advances in synthetic and catalytic chemistry, the discovery of improved oxidation processes seems possible. The time pressure to develop these processes is less than for detoxification processes, because material detoxified to meet the treaty demilitarization requirements can be stored safely and in compliance with the international treaty, until an improved oxidation process is developed and tested.

Biochemical processes to destroy chemical agents have received relatively little study, and if they are to be used, both exploratory and basic



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions 6 Low-Temperature, Liquid-Phase Processes This chapter reviews several chemical, biological, and other processes designed to destroy chemical agents under low-temperature conditions, those from about 20°C to less than 100°C. The lower value is room temperature. The upper value allows the use of aqueous systems at atmospheric pressure to minimize the risks of leakage of high concentrations of agent from pressurized autoclaves. Three main types of low-temperature, liquid-phase processes are reviewed below: processes for detoxification, processes for oxidation of organic residue, and biological processes. Detoxification processes have been the most intensively studied. They offer promising approaches for all three major agents in the U.S. stockpile (GB, VX, and H). This set of processes includes chemical processes and the use of ionizing radiation for detoxification. In addition to detoxification by conversion of agent to other compounds, demilitarization requires the process to be essentially irreversible (see Chapter 1). This requirement will, in some cases, require a two-step sequence of initial detoxification followed by further chemical processing, which could be accomplished with additional chemical reaction but would not require complete oxidation to carbon dioxide. Detoxification reactions that produce irreversible products are therefore of most interest. Relatively little effort has been directed to achieving complete oxidation at low pressure and temperature. However, there are several leads in this direction, arising from studies of chlorinated hydrocarbon destruction and agent decontamination. In view of recent advances in synthetic and catalytic chemistry, the discovery of improved oxidation processes seems possible. The time pressure to develop these processes is less than for detoxification processes, because material detoxified to meet the treaty demilitarization requirements can be stored safely and in compliance with the international treaty, until an improved oxidation process is developed and tested. Biochemical processes to destroy chemical agents have received relatively little study, and if they are to be used, both exploratory and basic

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions research will be required. The discussions of the three types of processes below reflect their developmental differences. An advantage of these technological alternatives is that they generally permit highly controlled, dosed environments. Most of these approaches will involve minimal gas emissions, and are suitable for in situ analysis of the progress of the course of agent destruction. A variety of toxicological tests has indicated that the chemical hydrolysis products have very low toxicity, making it possible to effectively manage the waste streams resulting from most of these processes. There has been extensive study of the alkaline hydrolysis of nerve agent GB, and the chemical products have been identified. Although the chemical and biological systems have a certain, inherent technological simplicity due to their mild environmental requirements, there are several important considerations common to all of them: There are various chemical specificities in the nature of the chemical or biological reactions that may limit their applicability to some of the contaminating ingredients of many stockpile elements. The gelatinous or insoluble nature of some of the stockpile material configurations may hamper their conversion. Each approach will result in a variety of process-specific reaction products with various levels of hazard that must be handled as separate waste streams in many cases. It may be necessary to sequence several processes, in some cases integrating chemical and biological technologies to adequately manage all of their reaction products. These low-temperature processes are generally not applicable to dunnage or the 5X decontamination of metal parts. Therefore, other processes would still be needed to handle these streams in the overall demilitarization system. The implication is that alternative demilitarization chemical processes must be carefully tailored to meet the requirements for each agent. A readily available low-temperature, liquid-phase technology is unlikely to be appropriate for all of them. CHEMICAL DETOXIFICATION PROCESSES A number of chemical processes that are known conceptually, experimentally, or in practice to destroy agent will (or can) run at moderate temperatures (between room temperature and 100°C) and at atmospheric pressure. Most of these processes are relatively simple to carry out and can be used in reactors commonly used in the chemical industry. Although no

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions single chemical treatment appears to destroy all three agents (GB, VX, and H), reactions to destroy all three could probably use the same basic equipment. However, parallel sets of some components may be desirable for each process because of differences in reaction rates, mixing conditions, and heating and cooling requirements. A distinct advantage of these chemical processes is that they can be operated in batch mode, with complete containment during the process and opportunity to verify satisfactory agent destruction. The destruction level should satisfy treaty requirements and hazardous material requirements for subsequent on-site storage and transportation to another site. Alternatively, further on-site treatment could be used to complete oxidation. A small gas evolution is expected for acid chlorinolysis and other processes that use oxidizing agents. A further advantage of these low-temperature, atmospheric-pressure, liquid-phase processes is that leaks of lethal agents are less likely to occur than from pressurized equipment. Although all these method are generally effective for destroying the chemical agents of chief interest, they generate different waste streams. Some might yield chemicals now used in the civilian economy, but economically viable recovery is not expected. By using some of the processes, relatively simple operations at each Army storage site could convert lethal chemical agents into material that meets demilitarization requirements. Such material could be managed according to standard practices for chemical waste. This approach would permit local storage or transport of the demilitarized material, of hazard no greater than that for industrial chemicals regularly shipped, to a central location where it could be processed by conventional waste treatment technology. In the United States, research on the chemical detoxification of bulk agent for demilitarization was mostly discontinued in 1982, when the decision was made to use incineration. Fortunately, a research and testing program on decontamination techniques for battlefield use has continued and produced many advances in understanding and applying chemistry relevant to detoxification. An informative review was recently published (Yang et al., 1992). Battlefield decontamination systems face a number of constraints not directly relevant to demilitarization of bulk agents. They are restricted to ambient temperatures and the need to minimize damage to surfaces. In addition, speed and ease of operations are essential. For bulk agent demilitarization, minimizing disposal problems is more important than for battlefield decontamination. Also, some chemical reactions that have not been found useful for battlefield decontamination might, under conditions of intense mixing and higher temperature, be useful for bulk agent detoxification.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions The following sections address the applicability of specific chemical processes to the agents of greatest interest. GB (Sarin) Reaction with sodium hydroxide in water. The hydrolysis process in which GB (isopropyl methylphosphonofluoridate, or Satin) is reacted with sodium hydroxide (NaOH) in water, a process often called neutralization, was studied intensively by the Army a decade ago (Eq. 1). Trial runs, including some at full scale, were conducted. On the basis of that experience, the process was abandoned in 1982 in favor of the current incineration technology. Major considerations were the large residue of hydrolysis products and problems encountered in scaling up the mixing of GB with aqueous NaOH solution (Flamm et al., 1987; Coale and DePew, 1992). Large quantities of residual salts were formed because of the use of excess NaOH, which was used in an attempt to destroy GB completely. Although it is now evident that apparently incomplete reactions were analytical artifacts related to a diester impurity in GB, acceleration might be achieved by use of small mounts of catalysts, especially ortho-iodosobenzoate salts. The latter accelerate reactions of structural analogs of GB with sodium hydroxide (Moss et al., 1984). As to the problem of residual salts, use of just slightly more than enough NaOH to satisfy the requirements of the balanced equation (Eq. 1 above) should be sufficient.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions In view of the high reactivity of GB with hydroxide ion in water (rate constant, k=23.7 liter/mole/second), it is probable that solutions of ammonia or low-molecular-weight amines in water would effectively detoxify it (Gustafson and Martell, 1962). This would need to be verified by research. Note that use of NH4OH instead of NaOH would change the nature of the waste stream from the process. The products from reaction of GB with aqueous NaOH are only slightly hazardous and are suitable for shipment, storage, or further processing. They will contain some of the mentioned diester, a substance not very hazardous itself and a contaminant in stockpiled GB, but convertible to GB on reaction with hydrofluoric acid (HF). Maintaining an excess of NaOH is therefore indicated. Although discontinued by the Army in favor of incineration (Flamm et al., 1987), this process has been used in the United Kingdom and other countries to destroy relatively small stocks of GB (see Chapter 3; Manley, 1992a, b). The hydrolysis of VX and GB by NaOH and Ca(OH)2 is further discussed below. Reaction with alkali in an alcohol The process of reacting agent with alkali in an alcohol may also be called neutralization. In this process, agent is combined with a solution of NaOH or potassium hydroxide (KOH) in an alcohol solvent. The alcohol may be methanol (McAndless, 1992a), 2-methoxyethanol (EPA, 1991), polyethylene glycol (Picardi et al., 1991), or (in principle) ethylene glycol. In Canadian experience, the solutions remaining after destruction of agents with KOH in methanol were then incinerated (McAndless, 1992a). Equation 2 shows the reactions that occur in this process. For H, or mustard, its reactions with KOH in methanol are shown; in other alcohols, the alkyl group in the ether (instead of CH3) would be that contributed by the structure of the alcohol.1 The ratio of the ether product to divinyl sulfide will depend on conditions. In the case of GB, probably some replacement of fluorine by the alkoxy group of the alcohol (e.g., minus-OCH3 with methanol) occurs, forming an intermediate that reacts further to form products of the type shown. This type of process is suitable for destruction of GB and H. There are doubts about its suitability for destruction of VX. Although reactions with VX are fast (Durst, 1992; Yang, 1992a-d), it appears that a by-product is a highly toxic compound also obtained from reaction of VX with NaOH in water 1   Most alcohols are composed of a hydroxyl group (OH) attached to a group composed of carbon and hydrogen atoms, such as C2H5 in the case of ethyl or grain alcohol. Thus, ethyl alcohol is C2H5OH.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions (Epstein, 1992a,b). This compound is described below in the discussion of VX hydrolysis. Energetics such as TNT are expected to react with methanolic NaOH or KOH, to yield azoxy compounds (Hickinbottom, 1957). Similar reactions are expected with solutions of these alkalis in the other alcohols mentioned. Other energetics may also react, depending on their chemical structures. The U.S. Army decontaminating agent DS2 is a solution of NaOH (2 percent) in 2-methoxyethanol (28 percent) and diethylenetriamine (70 percent). Its action is based on hydrolysis chemistry like that described here. KOH in 2-methoxyethanol is the active reagent in the proprietary DeChlor/KGME process (EPA, 1991), which, although intended for destruction of polychlorinated biphenyls (PCBs), should also be effective with GB and H. Solutions of KOH in polyethylene glycols are used in other proprietary processes to destroy PCBs and dioxins, such as the Galson and General Electric KPEG processes (Picardi et al., 1991). The reagents were developed to manage toxic materials (aryl chlorides) far less reactive than the agents H, GB, and VX. The general class of processes described here appears to be broadly applicable to agent destruction. A disadvantage common to these processes is the relatively large quantity of organic waste produced. Acid-catalyzed hydrolysis. The rate of acid-catalyzed hydrolysis depends on the concentration of hydrochloric acid (HCl) or sulfuric acid and on the temperature. At a hydrogen-ion concentration of 1 mol/L (about 4 percent HCl in water), the half-life of agent is 138 minutes, which implies 99.9999

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions percent destruction in 27.6 minutes at 25°C (77°F) (Epstein, 1992b). This rate pertains to the first stage of reaction shown in Equation 3; the second stage is thought to be slower. (Its rate constant needs to be determined.) The products of the first step lack the enormous toxicity of GB. In principle, hydrolysis of diisopropyl methylphosphonate (the diester of structure shown above) should also be catalyzed by acids, but the actual reaction rate appears not to have been measured. GB undergoes some spontaneous hydrolysis in water. The acid product, HF, would be expected to accelerate the hydrolysis. It is therefore conceivable that dissolving GB in water and letting it stand for considerable time would allow the spontaneous reaction and ensuing autocatalysis to completely destroy the GB. A preferred alternative would be to recycle the HF-containing hydrolysis products to eliminate the need for adding other acids. A second reactor would be needed to serve as a source of acid and to attain the desired level of agent conversion. A possible disadvantage of acid-catalyzed hydrolysis is that corrosion of process equipment is likely to be more severe than for reaction of GB with aqueous NaOH. The chemical reaction described in Equation 3 is well known from laboratory studies (Epstein, 1992a,b). Additional laboratory work should be conducted to determine rate constants for diester hydrolysis and the second stage of GB hydrolysis, to allow comparison with the rate constant for reaction with NaOH and to form a basis for pilot plant work if this alternative is to be pursued. Reaction with ethanolamine. When dissolved and heated with ethanolamine (a commercial product), GB reacts to form products of lower toxicity that are suitable to store or ship for further processing (Eq. 4). Because the products contain a diester, further processing to destroy the

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions diester or to remove HF is necessary to prevent formation of GB from reaction of HF with the diester. Advantages of this process are that the volatility of GB is much reduced when GB is dissolved in ethanolamine, that the reaction occurs cleanly as depicted (Greenhalgh and Weinberger, 1967), and that the corrosion of process equipment is minimal. Its principal disadvantage is the total mount of organic material produced, which will require further treatment. The nitrogen content is expected to increase the production of nitrogen oxide compounds on final oxidation. This methodology was used by Soviet authorities to destroy about 200 tons of nerve agents in faulty chemical munitions that contained no explosives (Leonov, 1991). The committee is unaware of any American pilot plant studies. It has also been found useful for HD detoxification (see later section on mustard). VX Reaction with NaOH with and without hydrogen peroxide, in water. VX reacts with aqueous NaOH (Eq. 5), but the resulting product A-1 is very toxic—nearly as toxic as VX—when administered intravenously to rabbits (Yang et al., 1990b). Further hydrolysis of product A-1 should be achievable, possibly at more severe conditions or through use of catalysts or improved hydrolysis systems. Addition of hydrogen peroxide (H2O2) to aqueous NaOH was recently found to avoid formation of product A-1 and to give the same products shown for reaction of VX with the commercial product OXONE® (Eq. 6) followed by neutralization by the NaOH present (Yang, 1992c). The combination of NaOH and H2O2 reagent might be effective for the detoxification of GB, but the committee does not know of any research specifically on this subject. H2O2 is relatively inexpensive and its residue after reaction is water.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions Reaction with oxidizing agents in acid solutiorn. VX is rapidly detoxified by several oxidizing agents, of which chlorine in aqueous acid (acid chlorinolysis) and OXONE are of interest for demilitarization purposes. Equation 6 shows the reaction of VX with OXONE, which is a mixture of KHSO5, KHSO4, and K2SO4 in 2:1:1 molar proportions (Yang et al., 1992a). The reactions in acid chlorinolysis and in oxidation with H2O2 in aqueous NaOH share an important characteristic with the OXONE reaction, namely, that the P—S bond is broken and products of reduced toxicity are formed. Although these oxidative methods can all be effective for demilitarization, they differ in some attributes. Acid chlorinolysis and treatment with OXONE are conducted in a strongly acidic solution, which may encourage equipment corrosion. Chlorination involves use of chlorine gas, the first poison gas used in World War I but a common substance in most American communities, where it is used for chlorination of swimming pools and drinking water supplies. OXONE is a mixture of salts that, when combined with the salts of the acidic products of reaction with VX, would increase the waste stream for ultimate disposal.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions H (Mustard) For H and other mustard agents, insolubility, inclusion of thickeners, and the formation of gels and solid deposits during long storage present problems in carrying out chemical detoxification. Possibilities for overcoming these problems will be discussed after the chemistry is considered. Hydrolysis of mustard. Mustard agent in water solution hydrolyzes rapidly, but for demilitarization purposes the reaction is impaired by both physical and chemical factors. The physical problem is the extraordinarily low solubility of mustard in water. Investigators who have determined rate constants for mustard hydrolysis carried out their studies in mixed solvents, such as aqueous acetone (Bartlett and Swain, 1949; Yang et al., 1990c) or aqueous ethanol (Yang et al., 1987). The chemical problem is that the intermediate products are cyclic or oligomeric sulfonium salts, which are relatively unreactive and which moreover have the potential for slowly reforming mustard (Yang et al., 1990c). An estimate for the mustard hydrolysis rate constant at 90ºC can be made, based on the enthalpy of activation (ΔH*) of 18.5 kcal/mol reported by Yang et al. (1987) and on the rate constant 0.261 rain-1 at 25°C tabulated by Ward and Seiders (1987). (They pointed out reasons to consider this value as being of uncertain quality.) The estimate is 18.1 rain-1 at 90°C, or half-life 2.3 seconds. The actual hydrolysis reaction is thus very fast; if the solvent is just water, the rate-limiting step is the dissolution of mustard in water rather than the hydrolysis. Formation of the troublesome oligomeric sulfonium salt can be avoided if a strong nucleophile is present in the water; hydroxide ion fulfills this role. In principle, it can be supplied either as NaOH or calcium hydroxide Ca(OH)2.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions In NaOH solutions, mustard reacts with an initial (rate-limiting) internal displacement to form a cyclic sulfonium salt (Bartlett and Swain, 1949; Ward and Seiders, 1984). This reaction is then followed by further reaction with NaOH. However, the surface reaction of water on liquid mustard agent is reported to form a complex set of ionic products, which then diffuse into the bulk phase. Overall performance of this system has been rated as unsatisfactory for demilitarization purposes (Durst et al., 1988). In contrast, reaction of mustard with Ca(OH)2 in water at 90° to 100°C was reported (Reichert, 1975), although this reaction seems not to have been used for large-scale demilitarization. It was used in England to convert 125-gallon batches of HD (mustard) to thiodiethanol in an apparatus not much more complicated than a tub (Eq. 7). The committee considers this method to be very promising but advises further study before any use is made of it. Reaction with oxidizing agents. HD can be oxidized in the liquid phase by several strong oxidizing agents. All of these reactions are limited by HD's low solubility in water. Approaches to increasing the contacting between reactants include improving the physical dispersion and forming microemulsions of HD. As discussed above, the rate of reaction at the HD surface appears to be fast enough to prevent significant diffusion of HD into the aqueous phase. Even though HD solubility may increase with temperature, corresponding increases in reaction rates may still prevent penetration into the aqueous phase of dissolved HD. Regardless of the specific mechanism, overall reaction rate will be controlled by the HD-liquid surface area. Mechanical emulsification (as in milk homogenization) can reduce droplet size to a few micrometers, forming a very large surface area (about 5 m2/gram, for a droplet of 1 μm in diameter). Work on physical dispersion should be an important componem of a program on liquid-phase detoxification of HD. The need for dispersion with gelled or solidified agent should also be included in such a program. A surfactant and a cosurfactant (such as butyl alcohol) in proper proportions, along with a hydrocarbon, readily form an emulsion with HD so freely dispersed that it appears dear. A compound less toxic but very similar to mustard (CH3CH2SCH2CH2Cl instead of ClCH2CH2SCH 2CH2Cl) was found

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions TABLE 6-1 Oxidation Potential of Different Chemical Species Oxidation Species Potentiala Fluorine 3.06 Hydroxyl radical 2.80 Atomic oxygen 2.42 Ozone 2.07 Peroxydisulfate 2.06 Chlorine dioxide 1.96 Ag2+ 1.98 Peroxymonosulfate 1.98 Hydrogen peroxide 1.77 Perhydroxyl radicals 1.70 Hypochlorous acid 1.49 Chlorine 1.36 Ferric ion 0.77 a At 1M hydrogen ion concentration. The potentials change with pH. million or less. Electrical energy has been used to generate the ozone and UV light. The committee found no information on the treatment of concentrated organic wastes or chemical warfare agents. The best application of this technology would appear to be for final treatment of dilute solutions after bulk destruction and oxidation have been accomplished by other means. BIOLOGICAL PROCESSES Introduction and Overview The use of biological processes to destroy chemical warfare agents is at an early stage of development (Ward, 1991; Harvey and DeFrank, 1992; Landis and DeFrank, 1991). Biological processing may be useful in detoxifying neat organophosphorus nerve agents and in destroying the reaction products from initial chemical detoxification of agents. In general, biological systems

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions are most appropriate for processing dilute aqueous solutions.2 One of the most important issues about the applicability of available systems to chemical demilitarization is whether biological processes or biochemical reactions can be developed into functional engineering processes. The most promising potential applications of biological processes to chemical demilitarization appear to be the following: direct detoxification of stockpiled organophosphate nerve agents using cellular or enzyme-based reactions (potentially applicable to GB and VX but not to sulfur mustard agents); biodegradation and mineralization of reaction products from chemical destruction of the nerve agents GB and VX; biodegradation of thiodiglycol or other products from hydrolysis or chemical oxidation of mustard (H); and biodegradation used as a final polishing process for aqueous effluents from other detoxifying processes, such as chemical or thermal oxidation. The first two applications use biological processes in the primary detoxification stream or in secondary processing streams. They would entail the modification and integration of large-scale fermentation and waste treatment technologies developed for other applications, but not for agent demilitarization (Irvine and Ketchurn, 1989). If biological processes are used for initial detoxification of agents, the control and management of agent toxicity during fermentations or enzyme-catalyzed reactions would be of critical concern. For example, residual toxicity from the partitioning or sorption of agents onto microbial cell mass or immobilized enzyme support matrices must be considered. In addition, the capability of enzyme or cellular-based processes to completely degrade the agents (e.g., to greater than 99.99 percent destruction efficiency) has not been demonstrated in a practical reactor system. A final consideration is the characterization of gaseous, soluble, and solid by-products of the biological processes. These include waste cell mass (sludge), products of incomplete biological mineralization, and agent or reaction products potentially volafdized during process aeration. 2   Dilution to 5 to 10 percent aqueous solutions is an initial estimate of the maximum concentration of dissolved organic substrates (reaction products from initial detoxification of chemical agents) in solution that would be biodegradable. This is based on previous work in biological treatment of high strength industrial waste waters (Enzminger et al., 1987; Lepore et al., 1989; 1990a,b; 1991). Initial biological treatment studies would need to define this upper concentration limit for the particular reaction products to be treated. This would be accomplished through toxicity inhibition by using a climated cultures.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions Biologically based technologies used as alternatives to inchaeration in destroying the current agent stockpiles are assessed below. The discussion is limited to the potential for destroying purified chemical agent stored in bulk containers or collected through the disassembly of chemical weapons. Direct Destruction of GB and VX Technology description. Enzyme-based systems that can directly degrade GB have been identified from numerous systems (Table 6-2). The initial, enzyme-catalyzed hydrolysis of GB would result in the production of hydrogen fluoride (HF) and mono-isopropyl methylphosphonate. Several microbial strains capable of hydrolyzing VX have reportedly been isolated, but none has been characterized to date (Harvey and DeFrank, 1992). Several enzyme and cellular systems have been identified that are capable of cleaving the P-C bond and degrading the methylphosphonate products of agent hydrolysis. Microbial cultures capable of degrading the relatively nontoxic phosphonic acids are described in the section below on the biodegradation of the reaction products from chemical processing of GB and VX. The main issues for enzyme-or microbial-based hydrolysis of the chemical nerve agents are not the intrinsic capability of biological systems to detoxify the agents, but rather system integration questions: What is the greatest degree of destruction that can practically be achieved, and what is the associated system efficiency? What are the impacts of impurities and stabilizing agents in the chemical warfare agent stockpiles? How much aqueous dilution will be necessary if biological processing requires solubilization with a solvent or another biological product? What volume of biomass and neutralization salts will be produced by biological destruction, and what are the characteristics of and management considerations for these components? What, if any, metabolic products result from biological treatment of materials containing significant amounts of trace compounds? (See the section on chemical processes for trace by-products.) Do the various physical states and chemical impurities of the stockpile materials interfere with the biological processes? Are the kinetics of the biological processes adequate for practical scaleup of degradation? What types of analytical and process control systems must be developed to monitor the processing streams.?

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions TABLE 6-2 Enzymes Capable of Degrading Organophosphorus Neurotoxins   Substrate Source (protein) P-O P-F P-X Genetics Reference Human serum A P,C GD>GB pac cloned Hasset et al. (1991); Furlong et al. (1991); Gan et al. (1991); Smolen et al. (1991) Human serum B P,C GB>GD pac cloned Hasset et al. (1991); Furlong et al. (1991); Gan et al. (1991); Smolen et al. (1991) Rabbit serum P GB>GD pac cloned Furlong et al (1991); Zimmerman and Brown (1986) Rat serum - GD>GB>DFP GA - de Jong et al. (1989) Rat liver P GB>GD>DFP GA - Little et al. (1989) Hog kidney - GD>DFP>GB GA - Mazur (1946); Hoskin (1990) Sheep serum P DFP - - Main (1960); Mackness and Walker (1981) Pig liver P GB>GD>DFP GA - Whitehouse and Ecobichon (1975) Squid nerve - DFP>GD>GB - cloned Hoskin (1990); Ward (1991) Squid muscle - GD>DFP - - Hoskin (1990) Clam foot - DFP>GD - - Landis (1991); Anderson et al. (1988) T. thermophila - DFP>GD - - Landis and DeFrank (1991); Landis et al. (1987) B. sterothermophilus NPEPP GD>GB - - DeFrank and Cheng (1991); DeFrank (1991) Ps. diminuta P,C+ GB>DFP>GD VX cloned Dumas et al. (1989a,b); McDaniel et al. (1988); Lewis et al. (1988); Serdar et al. (1989) E. coli - GD>GB>DFP - - Zech and Wigand (1975) Thermophilic bacteria - GD - - Chettur et al. (1988) Halophilic bacteria - GD>GB GA cloned DeFrank and Cheng (1991) Substrates: P—O bond: P, paraoxon; C, coumaphos; NPEPP, 7-nitrophenyl ethyl (phenyl) phosphinate; +, others. P—F bond: GB, Sarin; GD, Soman; DFP, diisopropyl fluorophosphate; M, mipafox. P—X bonds: GA, Tabun (P-CN bond); pac, phenyhcetate (P-C bond); VX (P-S bond). The > sign indicates which agent is used more effectively by the substrates. Source: Based on Dave et al. (1993).

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions The following sections address the scientific principles of primary detoxification, availability of appropriate biological systems, and systems issues. Biological systems. Numerous biological systems have defined enzyme-based capabilities for detoxification of nerve agents (Table 6-2). Several degradative gene systems have been dolled and subjected to genetic manipulation (Pseudomonas, halophilic bacteria, mammalian serum enzymes, and squid nerve enzyme). Several degradative enzymes (from squid nerve, soil bacteria, halophilic bacteria, and mammalian serum paroxonases) have been purified and to various degrees their interactions with G agents (GA, GB, and GD) characterized. The diversity and substrate preferences of agents treatable by enzyme-based systems are indicated in Table 6-2. In addition, numerous other biological systems have been shown to possess degradative capabilities, although the genes and enzymes have not been extensively detailed with multiple substrates and their specificity for GB and VX has not been determined (Mounter et al., 1955; Attaway et al., 1987; Mulbry and Karns, 1989a,b). Bulk liquid GB has been directly hydrolyzed at laboratory scale by defined enzyme-based systems. Several biological systems that have been used to degrade GB are also applicable to other G agents. Only a few of the enzymatic reactions identified in these various biological systems (Pseudomonas, the halophilic bacterium Altermonas , and squid nerve ganglion) have been well characterized with G agents; most of the others represent preliminary whole cell identification of biochemical capabilities. The direct use of biochemical degradation for G agents was first suggested for the squid enzyme in 1982 (Hoskin and Roush, 1982). The enzymes from Pseudomonas and squid have been purified and immobilized in active form on various matrices (Hoskin and Roush, 1982; Caldwell and Raushel, 1991), suggesting a potential for their use in bioreactor development. The organophosphate hydrolyzing systems of soil bacteria (the opd gene) have been successfully used in field studies to detoxify the insecticide coumaphos, a neurotoxic surrogate for the G agents (Kearney et al., 1988). Although the enzyme involved is also capable of GB hydrolysis (Dumas et al., 1990), the coumaphos studies address neither the actual substrate concentrations that could feasibly be used in direct treatment of agents nor the effect of stockpile contaminants on the activity and stability of these enzymes. These biological systems need further R&D before being used in biodegradation technologies (see Engineering Prospects).

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions Biodegradation of Reaction Products from the Chemical Processing of GB and VX An alternative to the direct biological or biochemical treatment of GB or VX would be to couple initial chemical processing of the agent with biological degradation of the reaction product, especially for the complete destruction of the products of well-defined chemical reactions (see the earlier sections in this chapter on chemical processes). Biological systems offer much promise to manage bulk materials of lower toxicity. The potential hazards of the longer reaction times and storage of these systems, and associated fermentation safety concerns would all be greatly reduced because of the lower toxicity of the chemical process streams. For this application, there are several major issues that require consideration: What are the reaction products of chemical processing of the primary agent and of the contaminants present with the agent? Are there efficient biological systems that have been adequately characterized for each specific purpose? What is the impact of carrier materials in the streams? For example, how much dilution of the products from chemical processing of GB and VX is necessary to provide concentrations of organic compounds appropriate for biological processing? What are the management considerations for the volume and characteristics of the resulting biomass and neutralization salts? Each chemical process discussed earlier in this chapter results in reaction products that could be further treated by biodegradation (see earlier section on chemical detoxification processes). As an example, the biodegradation of the reaction products of H will be discussed in the next section. When organic solvents are used for the chemical degradation of GB and VX, subsequent biodegradation must contend with the reaction solvent, reaction products, and impurities or their products. For example, reaction with potassium hydroxide in methanol would require degradation of the methanol carrier solvent as well as the reaction products; the aqueous reaction of GB and VX with ethanolamine would require degradation of ethanolamine as well as the reaction products. Biodegradation of the solvents is readily achievable; however, biodegradation of many of the reaction products has not been directly investigated. The most likely biodegradation pathways include the following: cleavage of the phosphonate ester linkage followed by oxidation of the resulting alcohol;

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions cleavage of the amide linkage followed by reduction of the resulting cleavage products; and cleavage of the methylphosphorus bond of the phosphonate to produce methane (Harkhess, 1986; Schowanek and Verstraete, 1990a,b) or biodegradation of the alcohols (the hazards associated with the methane, the main constituent of natural gas, would need to be addressed). Chemical Hydrolysis and Bioremediation of Mustard The direct biodegradation of mustard agents containing sulfur is not promising because there are no corresponding microbial or enzyme-based systems. Unlike the G agents and VX, mustard compounds are toxic to most biological systems. For this reason, initial chemical processing, possibly to form thiodiglycol, might be used followed by biological degradation to eliminate possible regeneration of agent by reaction with HCl. Mustard has reportedly been hydrolyzed by Ca(OH)2, yielding thiodiglycol as the primary product. The thiodiglycol can be degraded by two different strains of recently isolated bacteria (Pseudomonas sp. and Alcaligenes xylosoxidans) that are able to use thiodiglycol as their sole source of carbon and sulfur. Mustard from 1-ton containers at Aberdeen Proving Ground was used to demonstrate that chemical hydrolysis by NaOH with NH3, followed by biodegradation, was directly applicable to chemical agent stockpiles. The resulting culture medium was determined by bacterial toxicity studies to be nontoxic (Harvey and DeFrank, 1992). Bioremediation of Explosives and Energetics Bioremediation of explosive and energetic materials has been demonstrated for dilute, purified materials (Kaplan, 1993). However, applications to the explosive and energetic materials in the stockpile do not seem expeditious because of the burnable characteristics of these materials and the ease by which they can be destroyed through combustion. Engineering Prospects Numerous steps are required to assess the progression from scaling up of biological concepts to the practical engineering systems for chemical demilitarization.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions Direct destruction of GB and VX. Further evaluations of the efficacy of enzymes and cellular-based systems for direct destruction of GB and VX should focus on several areas: identifying appropriate enzymes and cellular systems capable of detoxifying VX; defining the maximum extent of reaction (percent agent destruction) achievable for selected representatives of each potential system; defining the maximum aqueous concentration of chemical agent that each selected system can treat; and determining the usable life of each enzyme or cellular system and the quantities required for practical application. If the results of these investigations encourage further studies, subsequent research should also focus on the following areas: defining a suitable reactor configuration, including reaction and reactor kinetics, to determine reactor size and processing time requirements for bioprocessing; determining the production requirements of the enzymes or cellular systems in quantities sufficient for scaleup; and defining the quantities and characteristics of process effluents, including exhausted enzymes or cells, and nonbiodegradable organic species and salts. Biochemical processing would most likely involve either enzymes or whole cells dispersed in dilute chemical agent or immobilized as a catalyst bed through which the chemical agent would flow. The whole cells method would most likely require production of smaller quantities of enzymes or cells. In either case, considerable effort may be required to produce large quantities of purified enzymes. Application of enzyme deactivation for the treatment of nerve agent using whole cells may be a viable option. Specific concerns over use of whole cells in this application would be (1) the sorption of non-deactivated agent to cellular material (biomass) resulting in residual toxicity problems; (2) potentially more rapid deactivation of enzymes; and (3) greater limitations on operating conditions (temperature, pH, agent concentration, etc.) for whole cell systems rather than isolated enzymes. Whether to use purified enzymes or whole cell systems should be based on the ease of enzyme purification and the relative activities of the two systems. Use of the agent (non-deactivated) as the carbon and phosphorus source for microbial growth would most likely be unattractive because of the relatively slow growth rates (compared to deactivation rates of either purified enzymes or those present in whole cells)

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions and the attendant requirement that the cell growth reactor be configured to contain active agent. Incineration of the resulting whole suspension is unattractive because of the large quantifies of water that would have to be incinerated. Biodegradation of chemical degradation products. Future investigations for biodegradation of chemical detoxification processes should initially focus on identifying organisms or mixed microbial populations with broad degradation capabilities for the defined categories of reaction products of selected processes. Success of this secondary degradation will be determined by the development of suitable microbial consortia capable of carrying out the sequence of biodegradation steps necessary for treatment of the mixed-reaction products. Scaleup of potential biodegradation processes could be accomplished by using well-established fermentation and biodegradation technology. The most likely approach would use a series of sequencing batch reactors, as is common in the biodegradation waste treatment industry, or batch fermentation, as practiced in the biotechnology industry (Irvine and Ketchum, 1989). In either case, a series of batch reactors would need to be operated in parallel to permit the greatest process efficiency (this allows a reactor to start operation at a high substrate concentration and continue until the lowest residual concentrations are achieved). These engineering goals must be balanced with complete process control; all reactor contents should be tested before transport or final disposal. The design of a biodegradation reactor will depend primarily on whether the initial chemical reaction is carried out in aqueous or organic solution. Preliminary process design estimates for aqueous chemical reactions suggest that a typical batch bioreactor size would be 10,000 gallons and that two reactors operating in tandem would be required for the biodegradation of residual products from the chemical processing of 1 ton of agent per day. The biodegradation of reaction products from chemical reactions carried out in organic solvents (methanol or ethanolamine) would require approximately a 10-fold increase in reactor size to accommodate biodegradation of the carrier solvent. (To facilitate biodegradation the carrier solvent most likely would have to be diluted with water to less than 10 percent by weight.) Waste streams. Biodegradation of chemical reaction products would result in the following process waste streams: process waste water, including nondegradable process reaction or biodegradation products and neutralization salts; sludge from produced microbial cell mass; and gaseous bioreactor effluents.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions Process waste water could most likely be treated through reverse osmosis or evaporation to remove the salts accumulated from pH neutralization and halogen ion released through agent degradation. Recovered water could be recycled into the chemical reaction process or biodegradation process steps. The resulting dry salt stream could be disposed of by conventional waste-disposal practices after the absence of residual toxicity is confirmed. Sludge from microbial cell mass could be disposed of by using conventional commercial facilities for disposal of waste water treatment sludge disposal. Approximately 600 pounds (dry weight) of sludge could be expected for each ton of organic solute degraded. Gaseous bioreactor emissions (CO2, O2, N2, and methane [CH4]) would result from aeration of the bioreactors and the CO2 produced by organic solute mineralization. Emissions can be minimized by using either oxygen enrichment or pure oxygen instead of air to supply oxygen to the reactor, depending on microbial sensitivity to oxygen concentration and reactor design. Initial estimates are that between 20,000 and 40,000 cubic feet of oxygen (at standard temperature and pressure) would be required for each ton of organic solute biodegraded. This quantity is small enough to permit complete capture and testing of gaseous emissions before release. Effluent CO2 could be captured in alkaline solution to eliminate gaseous discharge. Effluent CH4 would need to be managed by the industrial processes commonly used for potentially flammable gaseous emissions. Developmental status. Limited investigation has been carried out to date on the potential for biodegradation of chemical reaction products. The following steps would be required for process development and scaleup: identifying microorganisms capable of biodegrading specific chemical reaction products; determining the maximum initial concentration of reaction products that can be degraded without adverse effects on the microbial community; determining the maximum extent of biodegradation achievable; determining biodegradation stoichiometry and rates; and developing process-control strategies. Some DOD agencies have established University Research Initiative programs involving several research centers for biodegradation to address the basic science and engineering of biodegradation of environmental contaminants. This joint work, which is being performed at the Army Research Office/Texas A&M University, Office of Naval Research/University of Washington (Seattle) and Advanced Research Projects Agency/Rutgers University, may help address some of the questions above.

OCR for page 109
Alternative Technologies for the Destruction of Chemical Agents and Munitions Summary of the Potential Application of Biological Processes Some of the major observations in the preceding discussion lead to the following generalizations about potential applications of biological processes to demilitarization of the U.S. Stockpile: Processes where microbial consortia oxidize products from detoxification and partial oxidation can probably be developed. The nerve agent GB, and probably VX, can be directly detoxified by enzyme reactions; mustard probably cannot be directly detoxified. The application of biological processes to propellant and explosives does not appear useful at this time. Biological processes are not applicable to dunnage.