[] TL: ALTERNATIVE TECHNOLOGIES FOR THE DESTRUCTION OF CHEMICAL WEAPONS: AN INFORMATION PAPER SO: Greenpeace International (GP) DT: October 1990 Keywords: chemical weapons disposal alternatives gp reports toxics military / [part 1 of 4] Prepared by: Alfred Picardi Environmental Science and Assessment Services 1155 Connecticut Avenue, N.W., Suite 400 Washington, DC 20036 with contributions and review from: Ruth Stringer Jim Puckett Paul Johnston Sebia Hawkins Lisa Bunin Wayne Landis Pat Costner Eugine Euglis (?) Dieter Miessaer (?) Prepared for: Greenpeace International 1436 U Street, N.W. Washington, DC 20009 October 12, 1990 --------------------------------------------------- TABLE OF CONTENTS PG I. INTRODUCTION.......................... 1 II. CHEMICAL REPROCESSING.................. 18 III. CHEMICAL DESTRUCTION PROCESSES........... 26 IV. NEUTRALIZATION/DECONTAMINATION.......... 61 V. BIOLOGICAL DESTRUCTION.................. 71 APPENDICES A. Chemical Weapons Explosives and Propellants B. Chemical and Toxicological Properties of Chemical Agents C. Hydrolysis Reaction Equations D. Summary of Neutralization Data E. Hydrolysis, Photodegradation and Microbial Breakdown Products of Chemical Warfare Agents F. Toxicology of Primary Degradation Products of Chemical Warfare Agents LIST OF FIGURES ------------------------------------------------------- 1. INTRODUCTION The following information paper reviews the history of the U.S. Army CW Demilitarization Program technology; and alternative technologies to high temperature incineration, the demil choice to date, are reviewed. Greenpeace has reviewed the Environmental Impact Statements and associated documentation concerning the Army plans to incinerate chemical weapons at Johnston Atoll and in eight locations in the contiguous United States (Greenpeace EIS Reviews, February 1990, July 1990).(Note:Greenpeace EIS Reviews hereinafter referred to as GP Reviews - spelled out Greenpeace) The Greenpeace reviews detailed inherent problems in the high temperature incineration (HTI) destruction technology chosen by the Department of the Army. Environmental problems associated with high temperature incineration include the production of dioxins, furans, and other products of incomplete combustion in the incineration process, in particular, the condensation products such as dioxins and furans formed in the cooler parts of the incinerators downstream of the combustion chambers in the air scrubber units and the stacks.(? outside the stacks) These toxics generated in the high temperature incineration process (? will partition) between air emissions and the scrubber brines. They will pose environmental problems in (?either) case. High temperature incineration was once viewed as the panacea for destruction of hazardous waste. It was thought that this was a simple and inexpensive destruction technology for a broad spectrum of waste materials. In recent years, high temperature incineration has proven to be both an expensive means of destruction and one that has unacceptable environmental consequences. The air pollution I abatement devices associated with the incineration units has multiplied the cost of these units as shown by the cost overruns of the JACADS facility. High temperature incineration can no longer be thought of as a simple, inexpensive way for completely destroying a broad spectrum of hazardous waste materials. (?Inject the rotary kiln/living corrosion issue and other tech probs?) This information paper reviews an extensive technology base available for the destruction of chemical weapons. In the years since the decision to use high temperature incineration for the destruction of chemical weapons was made, a diverse range of technologies which could be applied to the destruction of chemical weapons has emerged. These technologies are a result of the quest for technologies which are environmentally benign for the disposal of hazardous waste materials. These technologies combine a variety of chemical and biological as well as physical techniques. While many will not be applicable to all of the chemical weapons waste materials, a combination of techniques could be chosen in such a manner as to draw on the power of each method. Certain methods could be applied in series in order to degrade the chemical agents in step-wise fashion to relatively benign byproducts. CHARACTERIZATION OF THE CHEMICAL WARFARE AGENT STOCKPILE Chemical formulas of the stockpiled weapons, including nerve agents and mustards, are shown in Figure One. Explosives and propellants are listed in Appendix A. The chemical agents fall into two basic categories: the nerve agents are in a class 2 of chemicals known as organophosphates, the "mustard" agents, which act as vesicants. The organophosphate nerve agents include GA (tabun), GB (sarin), GD (soman), and VX. The chemical properties, including solubility and volatility of these agents are given in Appendix B. GB is fairly soluble in water, whereas VX is only moderately soluble in water. GB is highly volatile, whereas agent VX is less volatile. The "mustard" agents are vesicants. The vesicant agents in the unitary stockpile scheduled for destruction are: H (Levinstein mustard); BIS (2-chloroethyl) sulfide; HD (distilled mustard); HT (a plant run mixture containing about 60% HD and less than 40 agent T, which adds stability and persistence to the blend by lowering the freezing point; agent T also has considerable toxic properties of its own and is highly mutagenic in fruit flies); and Lewisite (agent L, an organic arsenical).(1) HISTORY OF U.S. ARMY DEMILITARIZATION OPERATIONS In the years prior to 1969, disposal methods included open bit burning, atmospheric dilution, burial, and ocean dumping. In 1969, the National Academy of Sciences performed a scientific review of chemical agent munitions disposal methods. The NAS recommended abandoning ocean dumping as a method of disposal and (1) Chemical Stockpile Disposal Program, Summary of U.S. Army Experience, 1987. 3 suggested chemical neutralization of nerve agent GB and incineration of the blister agents H and HD. In 1972, a Senior Advisory Panel report confirmed the original NAS recommended dual method approach to disposal and added that the Army should continue to test incineration for disposal of agents GB and VX. Incineration of hazardous wastes was an emerging technology at that time and was thought to be an environmentally benign method of completely destroying waste materials.(1) The Army first attempted neutralization on an industrial scale because of the Army's familiarity with neutralization in field disposal and decontamination operations. Neutralization in this sense is meant to describe a chemical reaction where the toxic agent is rendered nontoxic through a chemical reaction. In some cases, this is in fact, an acid base neutralization reaction, forming a salt.(2) According to U.S. Army documentation(3) , mustard agent can be neutralized by hydrolysis or by reacting with an excess of monoethanolamine. The Army pointed out, however, that the liquid organic waste by-product created disposal problems. No disposal method other than incineration was investigated for this organic waste. (1) Ibid. (2) Chemical Stockpile Disposal Program, Summary of U.S. Army Experience, 1987. (3) Chemical Stockpile Disposal Program, Final Programmatic Environmental Impact Statement, January 1988. ***Figure #1 page 4 5 Since incineration would have been used to dispose of the neutralization byproducts, no industrial scale neutralization was carried out for the mustard agents. Instead, an incineration process for the mustard agent was developed. The nerve agent VX was neutralized by the Army through the use of acid chlorinolysis. However, neutralization of VX was never demonstrated on the pilot plant scale.(1) The Army did carry out industrial scale neutralization of the nerve agent GB. Caustic sodium hydroxide was used in this neutralization process. The process is exothermic, meaning that a great deal of heat is generated by the chemical reactions. It was found that the neutralization reaction was sensitive to reactant concentrations, pH and temperature. Some recombination of reactants was found to occur in the drying of the salts formed in the reaction. Choice of salt dryer heating fuel as well as a change in the mechanism used for drying the salts proved to eliminate most of these problems. Rotary drum dryers were found to be far superior to spray dryers in relation to mitigating the problem of GB reformation. Fuels without acid gas combustion products also helped solve the problem. Approximately 3.7 million kg (8.4 million pounds (?tonnes)) of GB were chemically (1) Chemical Stockpile Disposal Program, Summary of U.S. Army Experience, 1987. 6 neutralized on an industrial scale at the Rocky Mountain Arsenal in Denver, Colorado and at the Chemical Agent Munitions Disposal System (AMDS) in Tooele, Utah.(1) Operations at the Rocky Mountain Arsenal included (?incineration) of mustard from ton containers between July 1972 and March 1974, (FN) as well as (demilitarization (?how)) of GB containing cluster bombs, (2422 ton containers (?out of place)), 106 warheads, each containing 368 bomblets, 1,222 fused bomblets, and 39,632 unfused bomblets. Nerve agent GB stored in underground tanks was also disposed of in the industrial scale neutralization process.(2) The neutralization process proceeded as follows: the nerve agent GB was pumped to a holding tank. From the holding tank it was pumped to a Venturi type mixing T, where it was mixed with a caustic solution consisting of 18 percent sodium hydroxide. From the mixing T the GB and caustic solution flowed into a reacting vessel. The reacting vessel also contained some amount of caustic solution. The reaction mixture was continuously agitated and recirculated. Heat generated from the exothermic reaction was removed by a heat exchanger.' (1) Chemical Stockpile Disposal Program, Summary of U.S. Army Experience, 1987. (2) Ibid. (3) Ibid. (FN) Add in quantity of mustard & nerve agent incinerated at RMI during 1970's 7 When the neutralization was determined to be complete by sampling and testing, the brine was transferred to a spray dryer, where it was reduced to salt. The water vapor from the spray dryers was scrubbed before being discharged to the atmosphere. The salt was packed in drums for disposal. No further processing of the salts was attempted. Wastewater from the scrubbing process and from periodic wash-down of the reacting vessels was transferred to an industrial sump or lagoon. Solid waste from this neutralization process included metal parts which were decontaminated, furnace ash and spray-dried salt. The decontaminated metal parts were sold as scrap. The furnace ash was disposed of 1986 in a hazardous waste landfill.(2) The furnace ash contained heavy metals which led to its classification as a hazardous waste. The furnace ash resulted from the incineration process used in the decontamination of metal parts. 21.5 million pounds of spray-dried salt was placed in an approved hazardous waste landfill in 1986.(3) (?Add in a characterization of the salts - i.e., toxicity) The industrial process of neutralization was improved and operations carried out at the Tooele Army Depot. Two major operations occurred between 1979 and 1981, (2) Ibid. (3) Ibid. 8 when 13,951 M-55 rockets were demilitarized. The process consisted of draining the liquid GB, cutting the rocket into pieces, incinerating the explosive/propellant, thermally decontaminating the metal parts and storing the GB for neutralization.(1) Between July 1981 and July 1982, 12,673 projectiles were demilitarized. The process consisted of extracting the nose closure, puffing the burster well, draining the liquid GB, thermally decontaminating the empty projectiles, and storing the GB for neutralization.(2) The GB was neutralized, using what was called the Agent Destruction System (ADS). The ADS incorporated improvements over the Rocky Mountain Arsenal system. These improvements consisted of mixing the caustic solution and GB in the reactor itself, rather than in a mixing T. This alleviated line clogging problems. The heat of reaction was removed from the reactor by a recirculation system and cooling jacket, rather than a heat exchanger. These modifications were incorporated to eliminate the foaming and fine plugging problems experienced at Rocky Mountain Arsenal and to improve process operations in general. These changes were also necessary to accommodate caustic neutralization of the VX acid brine in the second (1) Ibid. (2) Ibid. 9 step of the VX neutralization process. VX, however, was never neutralized on this industrial scale.(1) Other significant changes and improvements over the Rocky Mountain Arsenal industrial process included use of drum dryers rather than spray dryers for evaporating the water and reducing the brines from neutralization to salts. The main reason for the changeover was to avoid the conditions which led to minuscule GB reformation. to be high operating temperature and exposure to acidic combustion gases. Furthermore, the drum dryers involved much These conditions were found less air volume, since the air was used not for heat transfer, but only as effluent to carry away water vapor. A smaller volume of air had to be scrubbed in the drum dryer operation. The operation of the drum dryers was found to be more cost effective than the spray dryers.(2) THE RATIONALE FOR CHOOSING INCINERATION The Army was dissatisfied with the process undertaken at the Tooele Army installation for two main reasons. Minute quantities of nerve agent were found in the brines and the process took significantly longer than expected. Only (caustic (?what does this mean) was (1) Ibid. (2) Ibid. 10 used in this neutralization process. Other types of caustics or other types of reactants were not tried. Since the process took longer than expected, large excess amounts of caustics were added to the reactors. This produced a large amount of salt from the neutralization process.(? how much 6x's) No attempt was ever made to further process the salts, however.(1) The rationale for abandoning neutralization was based on a number of factors: 1) The sheer complexity of the process as compared to incineration, which was the emerging industrial technology for disposal of organic substances. At that time, the incineration process was thought to be simple. 2) The sensitivity of the process to numerous parameters that would slow the reaction or even promote hydrolysis reversal, reforming GB. 3) The quantity and nature of the waste that was produced (however, this waste was never processed further (?toxicity) 4) The high capital costs (at this time, incineration was regarded as a simple and cheap process.) For these reasons, in March 1982, the configuration policy board of the Army officially decided to abandon neutralization and adopt HTI incineration only for disposal ofchemical agent munitions.(2) (1) Ibid. (2) Ibid., Chemical Stockpile Disposal Program, Final Programmatic Environmental Impact Statement, 1988; Final Environmental Impact Statement for Project Eagle, Phase 1, 1971. (?Also, use National Academy of Sciences rpeat reference) 11 In retrospect, these problems with the neutralization must seem trivial compared to the problems encountered in safe incineration. The solutions to these problems in the neutralization process itself must seem quite cheap in comparison to the incineration process, which has run into continuous delays because of engineerinng inadequacies and huge cost overruns.(1) (?Add in technical requests from the Army) The National Research Council of the NAS report stated that were the Army to choose chemical neutralization as the principal disposal method, separate facilities would be needed for each of the agents to be neutralized, or alternately one agent could be processed in the facilities, then modified to accommodate other agents.(2) Apparently, at this time, the understanding of high temperature incineration was such that, it was thought that the same incineration technologies and methods could be used for different types of incinerator feeds. The NRC report goes on to point out that should chemical neutralization be used, it is important to know what agent will be found when an artillery projectile is opened. The NRC report states that the records on these agent weapons are not entirely complete, implying that the contents of some of the weapons are not reliably known. (?New) As far as cost of incineration compared with neutralization technology, the NRC report states that the design for a planned disposal facility on Johnston Atoll substituted a $2.5 minion liquid (1) GAO Report,(?which one) . p. 15-19; Tony Capaccio, "Army Hit by Added Delays on Chemical Weapons Cleanup," Defense Week, Volume 11, No. 28, p.1-2; David Evans, "Chemical Arms Disposal Plans Called Unsafe," Chicago Tribune n n , June 17, 1990; David Evans, "Chemical Assaults on the Credibility of Our Government," Chicago Tribune June 29, 1990. (2) Ibid.(?what's the reference. the NAS report?) 12 incinerator for the $29 million chemical neutralization system first tried at Tooele Army Depot (Scott, 1984a). The NRC report states. ... . . according to Little (?as in Arthur D.) (1981), the total cost savings when burning agents in two JACADS (furnaces) rather than neutralizing them is $29,521,000."(1)(?are comparative costs available) In retrospect, it can be seen that the cost analyses were totally and utterly inaccurate by huge factors. The analyses of the technologies available then should be re-evaluated in light of the many technologies and further development of these technologies now available circa 1990. The National Research Council report, in view of these emerging technologies stated that "despite these problems, the Army's technology development program, and its concept development phase, included one contract to explore chemical procedures such as hydrolysis, anhydrous chlorinolysis, and aqueous chlorinolysis (Jody et al, 1983). None of these chemical processes showed particular promise (Mitre Corporation, 1983d). Further, each required incineration of the by-products." At the time these evaluations were made, apparently complete degradation of the waste materials was not technically feasible through alternate technologies, although nothing more, than a one-step process was apparently ever considered.Further processing of, for for instance, neutralization by-products by processes other than incineration was not considered. [copy edited here] Ibid. 13 [The report stated that, in spite of an extensive search, no other chemical by-products of commercial value could be identified (Jody et al, 1983). The National Research Council report concludes that: . . . when compared to disposal by incineration, chemical neutralization processes are slow, complicated, produce excessive quantities of waste that cannot be certified to be free of agent, and would require higher capital and operating cost. The panel agrees with the Army's decision to abandon chemical neutralization processes in favor of incineration." OTHER OPTIONS EXPLORED BY THE DEPT. OF THE ARMY The National Academy of Sciences suggested in a report in 1984 that alternatives to the destruction of chemical weapons were the use of chemical processes such as hydrolysis, caustic neutralization, anhydrous chlorinolysis, and aqueous chlorinolysis. The report commented that such methods were difficult and slow for the agents and munitions in the current stockpile, although they could be used to decontaminate the disposal plant and equipment. The report further commented that chemical methods produced large quantities of hazardous waste materials. At this time, the waste disposal and toxic emission problems associated with incineration was apparently not fully understood, nor was the complexity or cost of high temperature incineration fully understood. The National Academies of Science study comments on chemical processes:(1) (1) Disposal of Chemical Munitions and Agents, Committee on Demilitarizing Chemical Munitions and Agents, National Research Council, 1984. 14 "An intuitively attractive approach to destroying chemical agents is to mix them with other inexpensive chemicals and produce harmless (perhaps even useful) end products. As noted in Chapter 2 when CAMDS began operation in 1979, it tested a process of caustic neutralization to destroy GB. These tests, however, were not encouraging. Based upon laboratory measurements of chemical reaction rates, GB has a half-life of less than one second in a 5% aqueous solution of sodium hydroxide (Yurow and Davis, 1982). Thus, caustic neutralization was expected to progress rapidly. In practice, however, it was difficult to achieve the necessary mixing of components. To speed up the process, excess quantities of sodium hydroxide were added. Still, more than two weeks were often required before the GB was adequately neutralized; i.e., could no longer be detected (Scott, 1984a; Paulic, 1984). Not only did the use of excess caustic increase the operating cost, but it also produced larger quantities of waste than had been anticipated. Calculations, based on simple chemistry, indicated that for every pound of GB destroyed, about 1.5 pounds of salt waste would be produced (Little, 1982). In actuality, about five pounds of waste were produced (Jody et al, 1983). These unexpected extra wastes came from (1) additional quantities of sodium hydroxide in an effort to speed up the neutralization process, and (2) the solutions used to wash down and decontaminate the equipment. Although the salts contain no detectable amounts of agent, they could not be disposed of in an environmentally acceptable way because they contained sodium chloride, phosphonates, and heavy metals (Jody, et al, 1983). . . . The neutralization process for VX is even more uncertain than that for GB (Jody, et al, 1983; Scott, 1984; Yurow and Davis, 1982). It seems that (1) VX contains a contaminant, about 10% that resists hydrolysis; (2) the reaction for VX is highly exothermic might "run away," leading to an explosion; and (3) chemical neutralization of VX has never been demonstrated at the pilot plant scale. The prospects for chemical neutralization of mustard appear even less attractive because of its low solubility and the imperfect characterization of products of the reaction process. Further, high temperature and pressure would be needed to practical reaction rates (Jody et al, 1983). 15 The production of excessive waste from the neutralization process, must be viewed in light of the fact that 1) the excessive quantities of waste were generated from the neutralization of GB due to an excess of reagents used; and 2) the excess reagents were used to speed the reaction due to the inadequacy of the mixing technique used in the process design. There are various ways of mixing reagents which could have proven much more efficient. In fact, the neutralized salts were certified free of agent before disposal. ALTERNATIVE TECHNOLOGIES HAVE BEEN DEVELOPED The following sections of this report will show a full range of available technologies for processing the chemical weapons into environmentally benign by-products, as well as processing the chemical weapons to generate useful chemical by-products. The neutralization processes described as being "slow" in the National Research Council (NRC) report, were slow only by virtue of inadequate process engineering of the neutralization process itself. The laboratory analyses indicated a very fast reaction rate.(1) The difficulty lies in the mass transfer of the reagent materials. This same difficulty is inherent in the incineration process. The NRC report stated that the chemical processing was complicated. In retrospect, the complexity of the JACADS five- furnace high temperature incineration system should be reviewed with respect to the various chemical and biological processes described below. (1) Disposal of Chemical Munitions and Agents, NRC, 1984. 16 A recent delisting petition developed by the Army for the neutralization by-products of various chemical agents asserts that "the decontamination protocol and procedures are safe, scientific, and result in the total destruction of chemical agents.(1) OTHER TECHNOLOGIES PROPOSED TO THE DEPARTMENT OF THE ARMY The Department of the Army received various unsolicited proposals before the implementation of "Project Eagle," the disposal of chemical weapons by incineration and neutralization at the Rocky Mountain Arsenal. The unsolicited proposals covered electrochemical, chemical and biodegradation processes. While the Army chose not to pursue any of these unsolicited technological proposals, the report, concluded that biodegradation and electrochemical processes should be investigated for further use in mustard and other disposal problems.(2) The Department of the Army and the Department of Energy have had ongoing programs for alternate hazardous waste and chemical weapons disposal since technologies since 1983. Methods such as biodegradation, molten salt degradation, and other emerging technologies have been supported by both of these agencies. These processes and others will be reviewed in the following sections. (1) Support for the Debating of Contaminated Liquid Chemical Surety Materials as Listed Hazardous Waste, CRDEC Aberdeen Proving Ground, 1988. (2) Final Environmental Impact Statement for Project Eagle, Phase I, Disposal of Chemical Agent Mustard at Rocky Mountain Arsenal, Denver, Colorado, 1971. 17 11. CHEMICAL REPROCESSING The most compelling options for chemical weapons destruction in these days of budget crisis and deficit would be recycling or reprocessing of the original materials to other chemical forms having economic value. None of the documentation from the Department of the Armny, Department of Energy, or the National Research Council has addressed the issue of recycling propellants and explosives by reconfiguring these into other munitions or stockpiling for further use. A current cost analysis should be performed in order to review the relative total cost of conversion versus the various available disposal methods, including more realistic cost projections for incineration. The pesticide industry occasionally reprocesses materials during normal manufacturing and formulating operations (e.g., reworking to meet specifications), but this has seldom been applied to finished formulations already released to the market and distribution systems (Pesticide Disposal and Detoxification Processes and Techniques, p. 176). The chemical weapons in question fall into two basic categories: organophosphate, similar to commercial pesticides, and the chlorinated mustard agents. Two conversion methods for transforming the unwanted chemical agents into useful chemicals are chlorolysis, and catalytic hydrodechlorination. 18 Chlorolysis Exhaustive chlorination as a method of disposing of pesticides and other chemical waste has been developed at least since 1974. There are two U.S. patents describing basic improvements in chlorination (Krekeler et al, 1972a; 1972b). The EPA has studied the usefulness and economics of chlorolysis for the destruction of pesticides and other chlorohydrocarbon wastes and a followup report has been issued concerning process details, engineering cost estimates, and marketing information (Shiver 1976; DesRosiers 1978, NITS). Chlorolysis in this instance refers to the exhaustive chlorination of aliphatic or aromatic feedstocks over a range of pressures and temperatures depending on the feedstock. A process has been developed by Farbwerke Hoechst Ag, of Frankfurt/Maine, Germany, in which hydrocarbons and their oxygenated or chlorinated derivatives are completely converted to carbontetrachloride, phosgene, and hydrogen chloride at pressures up to 240 atmospheres and temperatures up to 620 degrees C (1140 degrees F). A schematic of the chlorolysis process is shown in Figure Two (Krekeler et al, 1975). As of 1980, there were chlorolysis units operating successfully in Frankfurt/Maine, Germany and two in Soviet Russia. Control of potential gas and liquid effluents from a large scale chlorolysis unit is easily achievable through low temperature distillation units, condensors, traps, scrubbers, absorbers, and other means. Phosgene waste generally occurs as a contaminant in gaseous vent streams. Proper waste management options include hydrolysis with steam followed by sodium 19 carbonate solution scrubbing, sodium hydroxide or ammonia solution scrubbing in a packed tower; or reaction with alcohols to form chloroformates and subsequently polycarbonates (Ottinger et al, 1973). The chlorolysis unit in Frankfurt/Maine, Germany is a 50,000 metric ton per year reactor, which is used to process still bottom residues from aliphatic chemical manufacturing operations. Cost estimates for a 25,000 metric ton per year plant are shown below in Table One. These cost estimates are dated (Samfield 1978) and will need revision both for values of products as well as capital costs. The capital cost estimate for the incineration facility made at about the same time. The capital cost estimate for the 25,000 metric ton per year plant would be on the order of $27 million (1978 dollars) for primary and auxiliary facilities. The plant would yield approximately 15% discounted cash flow rate of return on investment at a disposal cost of approximately $133 (1978 dollars) per metric ton. The cost to destroy chlorohydrocarbons may be 13 cents per kilogram, assuming a market price of $300 per metric ton for carbontetrachloride (Samfield, 1978). The downside to the chlorolysis facility is the fact that 90-95% of carbontetrachloride produced is utilized in the manufacture of chlorofluorocarbons. An analysis of the future market as well as products which would be less potentially environmentally destructive would have to be made as part of an environmental impact analysis for the chlorolysis technology. Other drawbacks may include the ***Figure 2, page 20 21 fact that waste materials containing sulfur, nitrogen or phosphorous, all of which are contained in the chemical munitions, may have adverse effects on the chlorolysis process. The presence of sulfur bearing pesticides in excess of 25 parts per million sulfur in the hydrocarbon feedstream may cause severe corrosion of the nickel tube catalytic reactor (Shiver 1976). There's also some question as to whether nitrogen trichloride and phosphorous trichloride or phosphorous pentachloride would be formed in applying the chlorolysis process to nitrogen and phosphorous containing organophosphate, and of the hazards if these products were formed. Catalytic Hydrodechlorination The rationale behind catalytic hydrodechlorination is to achieve a dechlorinated or partially dechlorinated product which will be more readily biodegradable and may have possible use a chemical intermediate, solvent, or fuel. Researchers at the Worcester Polytechnic Institute have researched the basis for catalytic hydrodechlorination. The dechlorination is carried out by reaction with high pressure hydrogen gas in the presence of a catalyst. Figure Three is a representation of a batch operation. A large scale reactor had not been designed and the figure represents technology based on laboratory scale experiments (Kranich et al, 1977). The process as conceived, proceeds as follows: The chlorinated compounds are charged as a batch into a rotary extractor. Extraction takes place in hot ethanol and the extracted material is pumped into the reactor vessel. Alcoholic caustic (sodium hydroxide) is then added. The function of the caustic is ***Table 1, Page 22 23 to react with liberated hydrogen chloride gas which can deactivate the catalyst and also lead to corrosion problems. Next, the reaction vessel is pressurized to 32 to 50 atmosphere with hydrogen gas. Catalysts used are nickel 61%, or 10% palladium on charcoal. The catalyst is pre-reduced in hydrogen at 370 degrees C (698 degrees F). The catalyst particles of .15 to .30 centimeters in diameter are retained on a stainless steel screen during the reactor vessel discharging (698 degrees F). The quantity of catalyst required is 0.20/o of the weight of the compound to be dechlorinated. After the reaction is complete, the reaction mixture is pumped to a solvent recovery still. The catalyst particles are recovered on the stainless steel screen in the reactor vessel. Ethanol is distilled through a refractionating head and subsequently returned to the solvent holding tank. The hydrocarbons and possibly some incompletely dechlorinated materials, together with salt and small amounts of water introduced from the feed material are washed with water in a liquid extraction device. The salt water is discarded and the hydrocarbon materials are stored for possible alternate uses. Hydrogen gas at elevated pressures is used in this process, presenting an explosion hazard. However, the use of hydrogen gas under these conditions is a standardized industrial procedure. A waste stream from this procedure will contain hydrocarbons, perhaps partially chlorinated species and salt. It may be possible to reflux partially chlorinated species, or th cycle them through one of the chlorinated waste processes discussed below. 24 111. CHEMICAL DESTRUCTION PROCESSES Chemical destruction processes for the agents stockpiled chemical warfare are discussed below. Some of these processes are applicable to the nerve agents and the mustard agents, as well as the explosives and propellants. Some processes are more appropriate for the chlorinated mustard agents and some are more appropriate for the organophosphate nerve agents. Some processes would have to be coupled in order to achieve complete degradation into environmentally benign by-products. Several processes such as super-critical water oxidation (SCWO), wet oxidation and molten sodium salt destruction are nonspecific as to waste feed and can be used for all of the various classes of weapons materials as well as biological wastes. Several processes are reported to yield complete destruction to innocuous compounds. Caution should be exercised in screening these processes in the pilot plant and industrial scales to verify these envirnmentally benign characteristics. The incineration process was once thought to be relatively environmentally benign, yielding only the simple oxidation products of water, carbon dioxide and carbon monoxide. It is now known that the high temperature incineration process in practice yields highly toxic products of incomplete combustion as well as condensation products, such as dioxins and furans formed in the cooler parts of the incinerator. These condensation products and PICs have been documented in all incineration operations studied whenever organic materials are burned with a halogen source. ***Figure 3, Page 25 26 Super Critical Water Oxidation Process The Super Critical Water Oxidation (SCWO) process for hazardous waste destruction, utilizes temperatures and pressures of water above the critical point of water (374 degrees C and 218 atmosphere). The process is applicable to a broad range of organic compounds as well as oxidizable inorganics such as ammonia and cyanide. Bench scale tests have shown no creation of dioxins, furans or other refractory organics and pilot scale tests have shown net decrease in dioxin content from contaminated waste materials to effluent. The unique properties of super critical water are the key to the operation of this process. Gases including oxygen and organic substances are completely soluble in super critical water, whereas inorganic salts exhibit greatly reduced solubilities under process conditions. Organic substances dissolve in the super critical water, and oxygen and the organic substances are brought into intimate single phase contact at temperatures and molecular densities that allow the conventional oxidation reactions to proceed rapidly to completion. Bench and pilot plant tests indicate that residence times of less than one minute are usually sufficient to reduce the concentrations of organic compounds to levels below analytical detection Emits. The effluents from the process are contained and consist of clean water, gas consisting of carbon dioxide, oxygen and nitrogen, and solids if the waste contains inorganic salts or organics with halogens, sulfur or phosphorous. The water is near potable quality and is clean enough to meet National Pollution Discharge Elimination System (NPDES) discharge requirements. The effluent gases contain no 27 oxides of nitrogen or acid gases such as hydrogen chloride or sulfur oxide. The process generates no particulates and less than 10 parts per minion carbon monoxide has been measured (Thomason et al, 1990). BACKGROUND Above the critical temperature and pressure of water, organic substances are completely miscible in all proportions, while salts are insoluble above certain super critical temperatures. The solubility characteristics are strongly dependent upon density. Near the critical point, between temperatures of 300 degrees C and 450 degrees C, and densities from 0.2 to 0.7 gm/cm3, the density varies rapidly with relatively small changes in temperature or pressure. The dialectric constant of water, normally 80, drops to between 1 and 10 under super critical conditions. A dialectric constant of between 1 and 10 is similar to that of nonpolar organic solvents. When the organiclwater mixture becomes supercritical, by definition there is one single phase; thus, the components are miscible in aB proportions. At densities of less than 0.7 gm/cm3, the solubility of inorganic salts in water is as unusual as that of organics. Many salts that have high solubilities in liquid water have extremely low solubilities in super critical water. Water goes through a complete reversal in solvating behavior towards typical organic and inorganic substances through the temperature of 350 to 450 degrees C. When organic compounds and oxygen are dissolved in water above the critical point, they are immediately brought into intimate molecular contact in a single 28 phase at high temperature. With few mass transport constraints, the oxidation reaction proceeds rapidly to completion. The products of hydrocarbon oxidation are carbon dioxide and water. Organic heteroatoms are converted into inorganic compounds; usually acids, salts or oxides in high oxidation states. Phosphorous is converted to phosphate and sulfur to sulfate . Nitrogen containing compounds are oxidized to molecular nitrogen and N20. Because of the relatively low reactor temperature, neither oxides of nitrogen or sulfur dioxide is formed. The process can successfully treat wastes in which organic concentrations are as low as 0 percent or as high as 100 percent. Initial development has been targeted for aqueous wastes with 1 to 20 percent organic concentrations. This is perceived to be the niche in which this process can compete successfully with other processes for hazardous wastes. Initial projections show that costs are very competitive with alternative technologies such as incineration. PROCESS DESCRIPTION A schematic of the SCWO process is shown in Figure 4. The process as described by Thomason et al (1990) is as follows. Feed 1. Organic waste materials in an aqueous medium are pumped from atmospheric pressure to the pressure in the reaction vessel. 29 ****Flow Chart #1, goes between Pages 29 & 30 2. Oxygen, stored as a liquid, is pumped to the pressure of the reaction vessel and then vaporized. Alternately air can be compressed to system pressure and metered into the reaction vessel. 3. The heating value of the waste stream is adjusted by dilution or addition of fuels. 4. When organic waste contains heteroatoms which produce mineral acids and it is desired to neutralize these acids and form appropriate salts, caustic is injected as part of the feed stream. 5. A recycle stream from a portion of the super critical reactor effluent is mixed with the feed streams to raise the combined fluids to a high enough temperature to ensure that oxidation reaction goes rapidly to completion. Reaction and Salt Separation 1. Organics are oxidized in a controlled but rapid reaction. Since the oxidizer operates adiabatically, the heat released by the readily oxidized components is sufficient to raise the fluid phase to temperatures at which all organics are oxidized rapidly. 2. Since the salts have very low solubility in super critical water, they separate from the other homogeneous fluids and fan to the bottom of the separation vessel where they are removed. 3. The gaseous products of reaction, along with the super critical water, leave the reactor at its top. A portion of the super critical fluid is 30 recycled to the oxidizer by a high temperature, high pressure pump. This operation provides for sufficient heating of the feed to bring the oxidizer influent to optimum reactor conditions. 4. The remaining reactor effluent (other than that recycled) consisting of supercritical water and carbon dioxide is cooled to discharge carbon dioxide and water at atmospheric conditions. Cooling and Heat Recovery 1. Most of the heat contained in the effluent can be used to generate steam for use outside the SCWO process. Pressure Letdown 1. The cooled effluent from the process separates into a liquid water phase and a gaseous phase, the latter containing primarily carbon dioxide along with oxygen, which is in excess of the stoichiometric requirements (and nitrogen when air is the oxidant). 2. The separation is carried out in multiple stages in order to minimize erosion of valves as well as to optimize equilibrium. 3. Salts are removed from the separator as a cool brine through multiple letdown stages, and are either dried (and water recovered) or discharged as brine depending on operating requirements. 31 The authors state that for larger scale systems, energy recovery could take the form of power generation by direct expansion of reactor products through a super critical steam turbine. The system would be capable of generating significant power in excess of that required for air compression, oxygen pumping or waste feed pumping. The largest capital cost of the system is the air compressor. Table 3 shows the compounds thus far oxidized in the super critical water oxidation system on the bench and pilot plant scale. Results of pilot plant scale demonstrations have shown that the process has the ability to destroy toxic or persistent organic contaminants in liquid waste streams without producing hazardous by-products (Staszak, Malinowski and Killilea, 1987). The super critical water oxidation is a closed loop. Any process upset shuts the system down and the effluents are contained, not released to the environment. This feature is consistent with the U.S. EPA definition of a "totally enclosed treatment facility." The process can be adapted to a wide range of waste feed mixtures and scales of operation. Systems can be designed as skid-mounted transportable units or as larger stationary units. A pilot scale test was also conducted on themophilic bacteria spores. The results indicated that there is a potential application for waste of genetic and bioengineering processes as well as infectious wastes. 32 ****Table #3, Page 33 Bench scale tests were run in 1985 on a liquid feed comprising of methyl ethyl ketone contaminated with dioxins. At lower reaction temperatures, some chlorinated compounds were detected in the effluent, with concentrations decreasing as the reaction temperature was increased. Although the analyses specifically searched for chlorinated dibenzo-p-dioxins, none were detected above the analytical detection limit of one microgram per kilogram in any of the effluents from these tests. Wet Oxidation The principle behind the wet air oxidation process for organic waste is that if sufficient heat and pressure is applied, organic compounds can be oxidized by air or oxygen. Figure 5 shows a schematic of the wet air oxidation process. The temperatures are between 150 and 340 degrees C and pressures applied are between 450 and 2500 psig. At these temperatures and pressures, sewage sludges are oxidized to alcohols, aldehydes, acids and finally at the higher temperatures and pressures, to carbon dioxide, and water. Sulfur, nitrogen and phosphorous may remain in solution as salts. Heavy metals may be precipitated as sulfates, phosphates, oxides or hydroxides, or may remain in solution (Astro, 1977). Thirty to sixty minutes are required for oxidation to these states. In wet air oxidation, destruction of organic chemicals can be as high as 99.9% but many materials are refractory to this type of treatment, e.g., chlorobenzenes and PCBS. Total COD (Chemical Oxygen Demand) removal is usually only 75% to 95%. Because the oxidation is not complete, the off-gas formation from the process can contain appreciable concentrations of volatile organics and may require additional treatment 34 before released to the atmosphere (Resource Number 9 from Thomason et al, 1990, Innovative Hazardous Waste Treatment Technologies). Wet air oxidation is exceptionally economical when dilute solutions of materials are concerned (Flynn, 1977). Wet air oxidation has been combined with biological and activated carbon digestion of dilute waste streams. The wet air oxidation system is used to regenerate the activated carbon. The toxic of organic volatile compounds can be treated up to levels where they interfere with biomass growth, about 1,000 parts per million, or 1% in an aqueous solution (EPA 1989). Molten Salt Processes The Atomic International Division of Rockwell International Corporation spent over two decades developing expertise in molten salt technology. The salts used are caustic, usually sodium carbonate or a mixture of sodium and potassium carbonate. The molten salt bath as inherent advantages over conventional combustion techniques in that excellent thermocontact is maintained between the heat source and the carbonaceous waste material and the resulting off-gases generally do not contain acidic component, that is, oxides of sulfur or nitrogen. (Mass transport for complete combustion is a problem in high temperature incineration.) The salt bath is stable, nonvolatile, relatively inexpensive, nontoxic, and may be recycled for further use. This process is applicable to a variety of hazardous wastes including pesticides (similar tosuch as the organophosphate nerve agents, and explosives. The process has also been used on hospital wastes, low level radioactive wastes, and 36 ***Figure #5 Page 35 other hazardous chemicals, including carcinogens (Pesticide Disposal and Detoxification Techniques). [] TL: ALTERNATIVE TECHNOLOGIES FOR THE DESTRUCTION OF CHEMICAL WEAPONS: AN INFORMATION PAPER SO: Greenpeace International (GP) DT: October 1990 Keywords: chemical weapons disposal alternatives gp reports toxics military / [part 2 of 4] PROCESS DESCRIPTION Figure 6 shows a schematic of a molten salt pilot plant. The combustible waste material is shredded in a hammer mill and conveyed into a hopper to await entry into the bath. The bath is contained in a large vessel which contains the molten salt, maintained at between 800 and 1,000 degrees C. The waste material is blown into the reactor by an air stream which also furnishes oxygen for the combustion process. The bath is preheated on start-up; however, once the reaction is in operation, the heat of combustion makes the process self-supporting. A pilot scale model was in operation which was able to process feed rates of 50 to 200 pounds per hour, depending on the application (Pesticide Disposal and Detoxification Techniques). The carbonaceous material containing carbon, hydrogen, oxygen, nitrogen, sulfur, chlorine, and phosphate, plus oxygen in the molten salt bath with an iron catalyst is converted with the addition of heat to off gases including carbon dioxide, water, molecular nitrogen and molecular oxygen waste is converted to sulfates, phosphates, and chlorides. The off gases are sent through a baffle assembly to trap and remove entrained salt particles, and to a high energy Venturi scrubber unit to remove any fine particulate matter before emission to the atmosphere. Acidic 37 scrubber units are not needed since these components, that is, acid gases, are absorbed by the alkaline carbonate molten salt. In the pilot plant sited from this reference (Pesticide Disposal and Detoxification Techniques), the gaseous emissions were monitored and essentially only carbon dioxide, carbon monoxide, water, molecular oxygen and molecular nitrogen were found in the exhaust gas. However, small amounts of nitric oxides, less than 20 parts per million, organic chlorine, less than 0.4% of the original pesticide, and traces of hydrocarbons, less than 10 parts per million, were found. The pesticides tested represented several classes of chemicals, including the chlorinated hydrocarbons, DDT, 2,4-D, and chlordine, and organophosphorus compounds. This indicates that the process is applicable to organophosphate as well as chlorinated compounds, such as the mustard agents. The molten carbonate is recycled once it accumulates a dissolved impurities content of 20% by weight. These impurities consist of sodium chloride, phosphate, sulfate, and other noncombustible material. These affect the fluidity of the melt. The recycling operation consists of dissolving the carbonate by transferring to a tank where it is treated with water or aqueous sodium bicarbonate solution. The dissolved carbonate solution is filtered to remove insoluble materials and treated with carbon dioxide gas to precipitate sodium bicarbonate. Filtration is used to recover the bicarbonate salt. The bicarbonate salt is then returned to the molten 39 ***Figure #6 page 38 salt furnace, where it is reconverted to carbonate. The remaining bicarbonate solution is normally recycled, but eventually it accumulates too much chloride, phosphate and sulfate and must be discarded. Figure 7 shows a schematic for recycling the carbonate melt as described above (Birk, 1973). CURRENT DEVELOPMENT Currently, the Sandia National Laboratories in Albuquerque, New Mexico have been working on chemical oxidation of organic and mixed waste using molten nitrate salts. The Sandia objective is to develop a chemical oxidation process using molten nitrate salt in a simple reactor system to destroy a variety of hazardous waste streams including wastes in liquid, gaseous, or perhaps solid form. The targeted waste streams include chemical warfare agents explicitly. The process would reduce waste volume and concentrate waste metals and radionucleides in the salt, that is the molten salt bath. In the proposed Sandia system, the waste streams are heated and mixed with the molten nitrate salts in a direct contact reactor vessel at about 500 degrees C and atmospheric pressure. The melt is a mixture of potassium and sodium nitrate salts and is an extremely strong oxidizer which reacts chemically with organics including halogenated compounds such as chlorinated mustards. The organics form carbon dioxide, water vapor and simple salts. The inorganics including heavy and reactive metals and radionucleides are trapped in the melt and ultimately can be separated 40 ***Figure #7 Page 41 for disposal as simple salts such as metal oxides or halides. The reduced salts after the reaction can be regenerated by reaction with injected oxygen. The system has advantages over high temperature incineration in that it can concentrate heavy metal or mixed waste contaminants. The materials are stainless steel and operate at lower temperatures than high temperature incineration. The reactors are relatively inexpensive and there are no auxiliary fuel requirements, eliminating both fuel combustion products such as carbon dioxide and oxides of nitrogen and dilution of the product gases, thus eliminating or reducing stack gas cleanup. The inexpensive chemically stable nitrate salts are easily handled and can be recycled. They pose no hazardous properties of their own. The capital and operating costs are expected to be less than that of incineration. The Sandia system can be designed as an easily transportable and compact system for on-site processing (Tyner, et al, 1990). Sandia has built a reactor which will be used in Fiscal Year 90 for continuous flow of gas and liquid waste experiments as well as larger scale experiments with solids. During the remainder of Fiscal Year 90 and 91, Sandia's activities will include assessment of nitrate melt characteristics including reaction kinetics and catalytic effects, effects of mixtures and impurities on physical and chemical properties, and materials and engineering issues. Sandia will continue to evaluate the potential of the process for application to concentrated organic waste streams (Tyner, et al, 1990). 42 Ultraviolet Light (UV) and Titanium Dioxide (TiO2) Catalyst Sandia National Laboratories has been developing a system in which a titanium dioxide catalyst is used in conjunction with ultraviolet light concentrated by parabolic sun reflectors to degrade toxic organic compounds including chlorinated organics such as PCBS. Researchers at the University of Wisconsin have been experimenting with titanium dioxide as a catalyst in photochemical degradation of toxic organic compounds (Tunesi, et al, 1987). In the photocatalytic degradation of toxic organic compounds, titanium dioxide particles act as both semiconducting photoelectrochemical cells (PECS) and adsorbants. In the Sandia experiments, using solar power for the ultraviolet light source, titanium dioxide has been mixed into the waste and water and run through a long glass tube which sits at the focus of a 720-foot- long parabolic trough. Recent experiments have also been run on fixed catalyst supports within the glass tube, onto which the titanium dioxide is fixed (Pacheo et al, 1990; Pool, Science Volume 245, Hazardous Waste Consultant, January/February 1989). The experimental results of the Sandia solar system indicate a change in reaction kinetics from 0 order to first order. Tests were run on trichlorylethylene (TCE) in water at concentrations of 120 to 5,000 ppb. The tests were run until the effluent concentration of TCE was 5 parts per billion (ppb) which was the detection limit of the gas chromatograph, as well as the Environmental Protection Agency's regulated 43 effluent Emit. Sandia has also evaluated the effect of hydrogen peroxide on the destruction of TCE. The addition of hydrogen peroxide tended to increase the decomposition rate of the TCE. This system, thus far, appears to be applicable to more dilute aqueous solutions and may have application after dilution of the chemical agents or as in conjunction with another process (Pacheco et al, 1990). Ozone/Ultraviolet Irradiation Houston Research has developed a method of destroying or detoxifying hazardous chemicals in solution, including heavy metal cyanides or pesticides using a combination of ozonization and LJV irradiation. The technique involves rather simple apparatus including a reactor vessel, an ozone generator, a gas diffuser, a mixer, and a high pressure mercury vapor lamp. Pesticides have been reduced from about the 50 ppm concentration to levels less than 0.5 ppm (Pesticide Disposal and Detoxification Procedures and Techniques, Mauk et al, 1976). The material available indicated only bench scale investigation of this technique. The overall chemical reactions that occurred depend on the specific waste material being detoxified. In general, the process involves activation of the organic molecule to a highly energetic state by LTV illumination, followed by vigorous attack by ozone. A schematic for the ozone/LJV irradiation process is shown in Figure 8. 44 ***Figure #8 Page 45 Similar work using the combination of an ozone/LJV irradiation is being performed by Westgate Research Corporation (Zeff, 1977). Zeff and coworkers studied the treatment of wastewater that contained residues of explosives manufacture (Pinkwater) and the treatment of pesticide intermediates in groundwater near Rocky Mountain Arsenal, Denver, Colorado. The complete results of these tests have not been published. Cost estimates and other economic data are being developed for the treatment of explosives residues in process water (Roth, 1977). A study funded by the EPA Site Program for technology demonstration, used ultraviolet radiation with ozone and hydrogen peroxide to destroy (oxidize) toxic organic compounds, especially chlorinated hydrocarbons in water. The process oxidizes compounds that are toxic or refractory (resistant to biological oxidation) in concentrations of ppm or ppb. The system consists of a reactor module and air compressor/ozone generator module, and a hydrogen peroxide feed system. It is skid mounted and portable and permits on-site treatment of a wide variety of liquid wastes. A field scale demonstration was completed in March 1989 of a hazardous waste site in San Jose, California. Very low total organic carbon removal was found implying that partial oxidation of organics occurred without complete conversion to carbon dioxide and water. This technique may have use in conjunction with a biological system but is applicable to waste streams in the parts per million concentration range (EPA 1989). 46 Laser Stimulated Photochemical Oxidation This technology was designed to photochemically oxidize chlorinated organic wastes. The process was envisioned as a polishing step in treating organic contamination in groundwater. The most efficient destruction of clorobenzene occurred with concentrations of 12.5 to 50 milligrams per liter. Concentrations either too low, 3 milligrams per liter, or too high, 100 grams per liter, were less efficient (EPA 1989). Photolysis/Ultraviolet Irradiation (UV) Photodecomposition is an important pathway for environmental detoxification of pesticides such as organophosphate. The organophosphorus nerve agents decompose through photodecomposition in the environment. Decomposition products of the chemical warfare agents are listed in Appendix E. Photolysis is used in several processes for hazardous waste treatment in conjunction with chemical catalysts. Photolysis involves the absorption of light energy above 290 nm. The molecule can also receive energy through a transfer process from another molecule called photosensitization. The environment surrounding the molecule during photodecomposition is important in the subsequent steps of photodecomposition. The initial step of the photolytic reaction usually involves fission of the parent molecule to form free radicals. These unstable intermediates react further with solvent, other organic molecules, inorganic species, radicals, etc. The resulting degradation products may be a complex mixture in which isomerization, substitution, oxidation, or reduction processes have occurred. Photosensitization is the process by 47 which certain molecules absorb fight energy and transfer it to the species of interest, that is, the toxic molecule which would not otherwise absorb light at that wavelength. An important example of photosensitization is the photolysis of chlorodioxins by interaction with a hydrogen donor solvent molecule (Crosby, 1977; Homberger et al, 1976). Pesticides undergo several different types of photolytic reactions including: ring fission (Plimmer et al, 1967); condensation (Rosen et al, 1970; Primmer and Keamey, 1969); bond rearrangement (Benson, 1971); reductive loss of chlorine (Plimmer and Hummer, 1969); replacement of chlorine by hydroxyl group (Crosby and Tutass, 1966); and replacement of halogen by phenyl group (Ugochukwu and Wain, 1965). Catalytic Ozone Oxidation with UV Chen and Smith at Southern Illinois University researched treating solutions of wastewater by a combination of ultrasonic energy ozone and inactivated raineynickel catalyst at room temperature. Called sonocatalytic ozonization, the process reduced a phenol solution from 500 to 25 ppm. The process produced a series of oxidized species including catechol hydroquinone, pyrogallol and ultimately small amounts of carbon dioxide and water. Chen et al (1975) investigated an alternate system for treating wastewater containing phenol and ethyl acetoacetate using ozonation and a special FE203 catalyst and found that it was very effective in reducing chemical oxygen demand and total organic carbon in industrial waste water. More recent experimental work by Chen and Smith has focused on chemically catalyzed 48 ozonation rather than sonocatalysis. These workers are now of the opinion that essentially the same results for treatment of waste water can be accomplished by catalytic ozonation without generation and expenditure of ultrasom'c energy. Sonocatalysis in combination of ozone/UV irradiation are feasible processes for waste water treatment but are inherently more expensive than chemically catalyzed ozonation (Smith, 1977). As part of the EPA technology demonstration program, catalytic ozone oxidation utilizing ozone ultraviolet fight and ultrasound was used in a soil washing technique. This technique is applicable to soils, solids, sludges, leachates, and groundwater containing organics such as *PCB, PCP, pesticides, herbicides, dioxins and inorganics such as cyanides. The total contaminant concentrations could range from I part per million to 20,000 parts per million for the technology to be effective. The technology includes a multi-chamber reactor where ozone gas is applied to the contaminated water along with ultraviolet light and ultrasound. Ultraviolet light and ultrasound catalyze the oxidation of the contaminants by ozone. The treated water flows out of the reactor to a storage tank and is reused to wash another batch of soil (EPA, 1989). Plasma Reactors Various plasma reactors have been developed for the thermal destruction of hazardous waste. Retech Inc. has developed a plasma reactor in which a plasma 49 torch is used to create a molten bath to detoxify contaminated soils. Organic contaminants are vaporized and react at very high temperatures to form innocuous products. Solids melt and are incorporated into the molten bath. Metals are retained in this phase and when cooled, this phase is a non-leachable matrix. Figure 9 shows a diagram of the plasma reactor. The contaminated waste material (soils) enter through the bulk feeder. The interior of the reactor rotates during waste processing. The centrifugal force created by this rotation retains the waste and molten material from flowing out of the reactor through the bottom. Heat and electrical energy are also evenly distributed throughout the molten phase. The reactor is periodically emptied. The molten solids fall into a collection chamber where they solidify. Gases travel through the secondary combustion chamber and then through a series of air pollution control devices designed to remove particulates and acid gases. In the event of a process upset, a surge tank has been installed to allow for reprocessing of off-gases produced. Liquid and solid organic compounds can be treated by this technology and it is appropriate for treatment of hard-todestroy organic compounds and wastes contaminated with metals. A demonstration is planned for early 1990 at the Department of Energy research facility in Beaute, Montana. During the demonstration, the reactor will process approximately 4,000 pounds of waste at a feed rate of 100 pounds per hour. Effluent streams will be sampled to assess the performance of this technology. 50 ***Figure #9 Page 51 While the initial plasma phase of this technology differs from other types of high temperature incineration, the process may have the same types of problems as high temperature incineration (EPA, 1989). Microwave plasma destruction has been studied by Lockheed Missiles and Space Company, Palo Alto Research Laboratory, as a detoxification method for hazardous materials including pesticides (Oberacker and Lees, 1977). A demonstration unit which can process 1 to 7 pounds per hour of hazardous organic compounds, has been built. The EPA has provided additional support for the fabrication and demonstration of a 10-30 pounds per hour apparatus. Figure 10 shows the microwave plasma reactor schematic. Energy in the 2,450 MHZ was supplied to a quartz reaction tube. The organic hazardous waste material is added to the apparatus as a pure liquid, as a slurry or solution in water or methanol, or as a compressed cake or pellet. The apparatus operates at a reduced pressure (10 to 100 torr pressure) with a carrier gas and devices to admit and control the addition rate of the sample to the plasma reactor section. The hazardous material moves under the combined forces of gravity and a carrier gas (oxygen, oxygen-argon, or steam) through a quartz reaction tube held with quartz Raschig rings or alternately the reaction tube contains a quartz basket filled with quartz fibers. The purpose of the rings or fibers is to increase the resonance time of the hazardous material in the reaction tube. The reaction products may be contained in traps cooled with ice water or liquid nitrogen which is placed between the reaction tube and the vacuum pump. 52 ***Figure 10 Page 53 Destruction of the hazardous material takes place in the reaction tube by application of microwave radiation, which generates an ionized carrier gas. The microwave induced electrons, then react with neutral organic molecules to form free radicals which ultimately dissociate and/or react with oxygen to form simple reaction products. A tesla coil is generally used to ignite the discharge. After destruction of the hazardous material, the reaction products as gases, vapors, and readily cjondensible material, pass through the condensing phase before the carrier gas re-enters the vacuum pump. In a test on the destruction of U.S. Navy red dye (Bailin, 1977a), the final products included polyaromatic hydrocarbons (less than 2 ppm), with essentially no dye starting materials remaining (less than 5 ppm). Several vacuum feed techniques were evaluated for the dye, including different physical forms such as powders, a solution, and a water slurry. As part of the entire program, a survey of thighly toxic materials stored within the continental United States was performed (Bailin, 1977b). Included in this list of toxic materials were nerve poisons, acutely toxic organometallic compounds and heavy metal complexes (Pest Disposal and Detoxification Techniques). Electrochemical Processes Electrochemical techniques can be classified into four basic groups: 1) electrodialysis in which charged species (ions) can be removed from an effluent stream and concentrated into a smaller stream, 2) the electro flotation process is used to separate dilute suspensions into slurries and clear liquid. The technique uses 54 electrolytically generated gas bubbles in the classic flotation type of separation (Elect. of Cleaner Env., 1972). The above two techniques are physical separations achieved through electrochemical means. In addition, metal ions can be reduced at the negative electrode to the elemental state of the metal. Electrochemical techniques are strong techniques for separation of metals from waste streams. The chemical warfare agents and in particular the neutralized products contain heavy metal contaminants, the removal of which could be accomplished by electrochemical techniques. Furthermore, at the positive electrode, the anode, oxidation occurs, either directly or by oxidizing agents generated at the anode. Organic waste can be processed by chemical oxidation using electrochemical techniques (Elect. of Clean Env., 1972). [] TL: ALTERNATIVE TECHNOLOGIES FOR THE DESTRUCTION OF CHEMICAL WEAPONS: AN INFORMATION PAPER SO: Greenpeace International (GP) DT: October 1990 Keywords: chemical weapons disposal alternatives gp reports toxics military / [part 3 of 4] ANODIC DESTRUCTION PROCESSES The toxic species can be directly destroyed by electrochemical reaction at the anode or alternatively the electrolytic solution can be designed to generate oxygen or sodium chloride can be added to the solution or hypochloride. These species then serve as oxidants to destroy the organic species. Indirect oxidation using oxidants generated at the anode, is not restricted to two- dimensional contact of the effluent with the electrode itself The oxidizing species can diffuse throughout the solution. Where sodium chloride or seawater is added to the solution, chlorine or hypochloride is formed. Where the anode and cathode products are not separated, the 55 product of the reaction is sodium hypochloride, NACIO (Electrochemistry of Cleaner Environments, 1972). Direct oxidation at the anode of such species as ethane and other organic species has been extensively studied. Cyanide destruction can be achieved through electrochemical processes. Cyanide is one of the more toxic degradation products of mustard gas. The cyanide is electro-oxidized at the anode, followed by dimerization, then alkaline hydrolysis. The cyanate then rapidly reacts with further alkalyd in solution to give NH,HCO31 Na2CO31 and NH40H (Neumann, 1910, Lister, 1955). According to most authors, not only simple cyanides, but also complex cyanides can be decomposed at the anode. Practical electricochemical cyanide destruction plants are common in industry. Furthermore, the addition of salt into an electrochemical cell can reduce cyanide through reaction with hypochloride. This process is probably commercially competitive with destruction processes using chlorine gas or calcium hypochloride (Electrochemistry for Cleaner Environments, New York Plenum, 1972). Reductive Degradation Utilizing Metallic Couples A catalytic reductive degradation process has been patented in the U.S. by Sweeney and Fisher (1973). The system uses various metallic couples such as zinc/copper, iron/copper, aluminum/copper, etc. in the presence of dilute acids (pH 1.5 to 4.0). At room temperature, decomposition of dilute DDT emulsifiable concentrate is claimed. Typically, the concentration of an effluent containing 400 to 500 ppm 56 DDT (a chlorinated organic) could be lowered to 1 ppm. While the process is useful in detoxification, complete destruction is not achieved. A schematic flow diagram for the reductive degradation process is shown in Figure 11 (Pesticide Disposal and Detoxification Techniques). Chemical Decontamination Chemical techniques can effect the detoxification of the agent or may result in the complete or nearly complete decomposition of the agent. The agents involved are in two basic classes: the organophosphate nerve agents and the chlorinated mustards. Some of the chemical decontamination methods will not be applicable or appropriate for both classes of compounds. In particular, dechlorination methods, or hyperchlorination, chlorolysis methods would be applicable to the chlorinated mustard agents. Dechlorination Dechlorination can effect the detoxification or neutralization of the agent as well as rendering the chemical more amenable to further processing through biological or chemical means. The environmental toxicity of a compound is greatly reduced once it is dechlorinated. Dechlorination technologies can render a commercially valuable product are reviewed above. The following dechlorination technologies may not necessarily yield a commercially valuable product. Dechlorination technologies which do not affect nearly complete or complete dechlorination have not been included in this review. 57 ***Figure 11 Page 58 Metallic sodium is used to treat PCB and PCB contaminated materials by dechlorination. The technology is considered a proven technology; however, metallic sodium is a hazardous material and can lead to spontaneous combustion when it is allowed to contact water. The material is pyrophoric. A process developed by Galston Remediation Corporation has been extensively tested on soils contaminated with PCBS, dioxins and furans. The technology involves reacting potassium hydroxide (KOH) and polyethylene glycol (PEG) to form an alkoxide and water. The alkoxide reacts with one of the chlorine atoms in the biphenyl ring, producing an ether and potassium chloride. This renders both of the reaction products water soluble and renders them easily separable from the contaminated media. This is not a complete dechlorination or destruction technique dimethyl sulfoxide (DMSO) is used as a solvent and catalyst in the dechlorination process (Hazardous Waste Consultant, November/December, 1989). Dimethyl Sulfoxide The contaminated waste material (soil) is combined with an equal mass of the solvents KPEG/DMSO reagents. The resulting slurry is mixed and heated to between 200 and 350 degrees C and allowed to react for one to five hours. Excess reagent is removed by decantation and can be recycled. After the soil has been washed thoroughly with water, it is discharged. The entire process can be carried out in a single reaction vessel (Hazardous Waste Consultant, November/December, 1989). 59 General Electric's KPEG process is similar to the Galston Remediation process, but GE does not use DMSO as a catalyst. Heated transformer oil in the GE process is mixed with small amounts of potassium hydroxide (KOH), and PEG in a stirred vessel. According to GE, complete reaction of the PCBs occurs quickly, producing decontaminated transformer oil and a non-PCB by-product. The by-product is insoluble in transformer oil and is easily removed from decanting (Hazardous Waste Consultant, November/December, 1989). Such dechlorination processes as these could be used to detoxify the mustard agents prior to processing in another manner such as biological degradation. 60 IV. NEUTRALIZATION/DECONTAMINATION The Stockholm International Peace Research Institute (SIPRI) and the U.S. Army Armament Munitions Chemical Command at Aberdeen Proving Ground, Maryland, have assembled extensive reviews of neutralization/decontamination methods (Trapp, 1985; CRDEC, 1988). The two major chemical procedures effective in the destruction of toxic chemical agents (both mustards and nerve agents) are hydrolysis and oxidation. There are numerous reagents available for hydrolysis and oxidation processes. There is a chemical equivalence between these various available reagents. Moreover, the reactive species responsible for the decontamination are the same, while the reagents and admixtures or solvents may differ. The reactive behavior of the various reagents, hydrolytic or oxidants, can be considered chemically equivalent decontamination procedures on a mole per mole basis (CRDEC, 1988). For example, there are no major differences when sodium carbonate is substituted for sodium hydroxide insofar as mechanism and products are concerned. The various hydrolysis reactions for different agents and different reagents are shown in Appendix C of this report. A summary of the neutralization data for the various agents is shown in Appendix D. Differences in the various reagents may indicate solutions to problems identified in industrial scale neutralization plants such as mass transfer in the reactor vessels, line clogging, corrosion problems and formation of toxic by-products. 61 HYDROLYSIS Hydrolysis is the chemical reaction where water is added to a reactive molecule. Subsequently, some fragment of that molecule is eliminated, or cleaved, after the addition of water into the aqueous solution. For example, water will react with the nerve agent GB to produce one mole of hydrogen fluoride and one mole of GB acid. If sodium hydroxide (NAOH) is added to a water/GB solution, a rapid bimolecular reaction is observed whereby hydroxide attacks the phosphorous center of the GB and subsequently releases a fluoride ion. Hydroxide is known as a strong base as well as an excellent nucleophile. OXIDATION Just as many reagents which produce the hydroxide ion in water are useful in hydrolysis, many reagents make use of the hypochlorite anion for use in oxidation reactions. Hypochlorite is the preferred reagent for decontamination of mustards and the nerve agent VX. The reaction is highly exothermic and care must be taken because of the great amount of heat generated as the reaction proceeds. The reaction is usually carried out therefore in aqueous media. Moreover, the reagents preferred for oxidation decontamination release in a controlled manner the hypochlorite anion. For instance, sodium hypochlorite, calcium hypochlorite, or the organic N-chloramine compounds (CRDEC, 1988). 62 Hydrolysis and oxidation reactions can be catalyzed through the use of metallic ions or biologic enzymes. Various biologic enzymes have been discovered which hydrolyze the organophosphate nerve agents at rates well over a million times the normal hydrolysis reaction rate. A class of chemical compounds known as iodobenzoates mimics the enzymatic activity of the biologic enzyme. Furthermore, a soil bacteria has been isolated (Wild, 1990) which will catalyze the decontamination of organophosphate. The enzyme is active after the bacteria have been grown, processed and dried, and the viability of the bacteria is not necessary for the activity of the enzyme when reacted with the organophosphate. This research will be discussed below in the section on Immobilized Enzymes. The categories of reagents used for decontamination include water, strong bases, complexing agents and nucleophiles with a subcategory of oxidants. Table Two shows a listing of detoxicants, their reactions and applications. Water While the organophosphate GB and GA are relatively miscible in water, their reaction rates at neutral (pH 7) conditions is extremely slow, 75 hours at 25 degrees C (Crabtree and Sarver, 1977). VX is only slightly miscible in water and takes 40 days at pH 7 to hydrolyze (Crabtree and Sarver, 1977). The mustard agents are even less soluble in water. Furthermore, they tend to form agglomerations and can persist in the environment for many, many years (Landis, 1990). 63 ***Table 2 Page 64 Strong Bases The agent GB hydrolyzes readily in strong base. Because of its relatively low solubility in water, VX takes a considerably longer time in strong base. Because of the relative insolubility of Lewisite or its oxide in aqueous solution, use of a cosolvent such as alcohol, is recommended (CRDEC 1988). Reichert (1975) demonstrated that HD could be hydrolyzed with calcium oxide in excess. Aqueous sodium hydroxide and ammonium hydroxide were also used for hydrolysis of HD. The nerve agent VX is more resistant to cleavage by bases than GA, GB, or GD. The sulfur and nitrogen degradation products including di-isopropyl amino ethanol are not commercially reusable (Hedley, et al, 1977). One of the hydrolysis products of VX exhibits a mild toxicity upon injection and thus acid chlorinolysis is the decontamination method of choice for large scale destruction of VX in demil procedures (Crdec, 1988). Sodium hydroxide is the decontaminant of choice for the mustard agent L (CRDEC 1988). Ashcroft and Parker (1978) have suggested that substitutes for sodium hydroxide or sodium carbonate could include strongly basic sodium salt such as tri- sodium phosphate or sodium silicate. These systems have not been studied in any detail (CRDEC 1988). Partly Aqueous or Nonaqueous Solutions Partly aqueous or nonaqueous solutions would facilitate solution of agents which are only slightly miscible in water. However, partially or completely nonaqueous solutions have lower dielectric constants than water and may actually slow the 65 hydrolysis reaction. A number of multi-component strongly basic mixtures have been studied for the decontamination of HD, GA, GB, GD, L, and VX. A proprietary mixture known as DS-2 (Jackson 1960) was found to be effective for agents GB and the mustard agent HD but somewhat less effective for the nerve agent VX. Monoethanolamine (MEA), an organic solvent, has been shown to be useful for the decontamination of the mustard agent HD (Brankowitz 1978). MEA is a relatively strong base, as well as an organic solvent. MEA has a relatively high flashpoint, is relatively nontoxic (TLV of 3 ppm), is noncorrosive to metals, inexpensive, relatively stable, and generates a moderate heat of reaction with HD. A volume ratio of 5:1 MEA to HD is required (Pistritto and Eng, 1974; Davis and Sass, 1974). Monoethanolamine (MEA) has also been used for the detoxification of organophosphorus pesticides (Wolverton 1973). Wolverton's work was part of the technical support for the Wharf Chemical Warfare Program for the Department of the Air Force in the late 1960s and early 1970s. The decontamination procedure involved a dilution of 1:10 organophosphorus to MEA. Wolverton claimed that the resulting reaction mixtures of the decontamination were less toxic to fish than the decontamination solution alone (Wolverton, 1973). 66 Complexing Agents and Nucleophiles Metallic salts such as copper, uranium, zirconium, thorium and molybdenum are employed in solutions closer to neutrality than the bases mentioned above. Only a few of these systems have been translated into useful decontamination procedures. Several promising compounds have been investigated (Ward and Hovanec 1987). Alpha Nucleophiles Even though the alpha nucleophiles are less basic, the reaction rates are more rapid than hydroxide anion for detoxification of such agents as HD, GA, GB, GD and VX. This group of reagents includes hydroperoxides, hypochlorites, oxiines, and hydroxamics (Crdec 1988; Larsson 1958; Hackley et al 1955; Demek et al 1962; Epstein et al 1956; Swidler et al, 1959; Stolberg et al, 1959; Vogel 1941). Of these reagents, only sodium hypochlorite has been developed for large scale decontamination. The class of oximes are used for antidotes to organophosphate exposure (DOA 1988). The hypochlorites fall more properly under the heading of oxidants discussed below. The enhanced reactivity of the alpha nucleophiles is related to the presence of an unshared electron pair on the atom next to the one bearing the negative charge, which decreases charge repulsion during interaction. In addition to the alpha nucleophiles, bidentate nucleophiles, such as pyrocatechol and pyrogallol anions, have been studied in the laboratory and found to hydrolyze organophosphate rapidly. The laboratory results have not been developed into 67 practical full scale systems (CRDEC, 1988, Jandorf, 1951; O'Neil et al, 1953; Higuchi, 1963; Epstein et al, 1964; Epstein et al, 1956). Lastly, sodium thiosulfate reacts rapidly to hydrolyze HD (Benes and Weidenthaler, 1957). agents. GB can This reaction has not been applied to bulk quantities of mustard be hydrolyzed at pH 7.6 in the presence of pyridinium bases. This reaction has not been developed outside of the laboratory (Grochowski et al, 1966). Oxidants HD mustard and VX organophosphate nerve agent are readily subject to oxidation due to their sulfur inclusions. The destruction of HD can be accomplished with bleach, that is, aqueous sodium hypochlorite solution, chlorinated lime, and calcium hypochlorite. Calcium hypochlorite yields the highest percentage of available chlorine and is currently most often used currently for decontamination. The reaction is violently exothermic and is more easily controlled in an aqueous slurry (CRDEC 1988). The oxidation products of HD with bleach in excess show conversion into numerous products, none of which are mustard. In fact, toxicological tests on mustard decontaminated with bleach show no toxic effects (CRDEC 1988). Organic chlorinating agents are even more efficient than bleach. Chloramine B and chloramine T are more expensive and not considered useful for large scale operations (CRDEC 1988). 68 Chlorine Chlorine has been used on a large scale as a decontaminant for agent Vx at Tooele Army Depot (Bauer et al, 1958; Benson et al, 1974). At Tooele, hundred pound batches of VX from munitions were dissolved in 1.5 N hydrochloric acid and chlorine added to a green color. Reaction was rapid and strongly exothermic yielding a destruction efficiency of 99.999999%. After reaction (acid chlorinolysis), the products were converted to drum dried salts (Valis and Vigus, 1977). Chlorine Dioxide Chlorine dioxide reacts with agent VX to give carbon dioxide carbonyl, sulfide, sulfate ion, phosphonic acid, and di-isopropylamine. The kinetics of this reaction are quite rapid. However, explosive gas is generated, and thus far large scale work has not been pursued (Epstein et al, 1969). Potassium Permanganate Potassium permanganate and acetone has been used for the oxidative destruction of HD (CWSR 1918). VX has been reacted with pen-nanganate in neutral solutions and formed reaction products which were "disposal problems" (CRDEC). VX, when reacted with permanganate in highly basic solutions, yielded mainly hydrolysis products (Davis et al, 1987). 69 Nitric Acid Oxidation of mustard with concentrated nitric acid produces a yield of mustard sulfoxide of greater than 99% (CRDEC, 1988). lodosobenzoates Perhaps the catalytic activity of the iodosobenzoates bridges the gap between pure chemical methods and biological methods of toxic waste destruction methods. The iodosobenzoates catalyze the hydrolysis of toxic organophosphorus esters (Katritzky et al, 1988; Durst et al, 1988; Longo, 1988; Landis, 1990). A catalyst can increase the rate of a chemical reaction many hundreds, thousands or minions of times over the uncatalyzed rate. Enzymes that hydrolyze organophosphate increase the rate of hydrolysis factors of 1 to 2 million or more. In this fashion, chemical reactions which are not favored thermodynamically can become quite rapid. The Army has done work on the class of catalysts known as the iodosobenzoates at the Chemical Research Development and Engineering Center in Aberdeen, Maryland. Furthermore, the Army has conducted research that shows that nerve agent can be decomposed even more efficiently by the use of iodosobenzoates with surfactants. Surfactants, like soaps, act to increase the solubility of materials in water (Landis 1990). 70 V. BIOLOGICAL DESTRUCTION TECHNIQUES Biological techniques have been used to treat municipal and industrial waste for many years and are considered proven technology. Biological wastewater treatment plants have been constructed for the last 80 years and the engineering parameters are well known. The use of biological techniques in toxics disposal has gained a good deal of attention in recent years. Great strides have been made in isolating the types of organisms which are useful in degradation of toxics, and enhancing their activity. Biotechnology has enabled the manipulation of genetic material, further enhancing the capabilities of the organisms used in biological treatment. The enzyme systems used by these organisms have in many instances been studied, characterized and separated from the organisms (Landis, 1990, Wild, 1990). Immobilized enzymes have been used in waste detoxification (Wild 1990, Munnecke, 1977). In fact, enzymes which hydrolyzes organophosphate has been separated and used in the treatment of organophosphate pesticide without the necessity of keeping the microbes themselves alive and growing (Wild, 1990). This enables the placement of these enzymes in various reactor conformations such as lining the inside of tubes through which the toxic waste material can be cycled (Wild, 1990). Biological reactors for batch and flow through have been developed beyond the traditional activated sludge and trickling filter bioreactors. These biological waste treatment systems have been used on a variety of industrial toxic wastes, including pesticides. Capabilities of the biological methods include the dechlorination of refractory compounds in anaerobic conditions; the hydrolysis of organophosphate including nerve agents, as well as the catabolic breakage of the carbon phosphorous bond. 71 The carbon phosphorous bond is almost nonexistent in nature, as is the organophosphate ester moiety. The evolution of biological enzyme systems is a matter of current speculation (Landis, 1990, Wild, 1990). An aqueous medium is used in activated sludge (biomass), anaerobic digestion, trickling filter, oxygenated lagoons, and practically all systems common to industrial biological treatment (Atkins, 1972; Gruber, 1975; Grunham, 1965; and Todd, 1970). Recent industrial processes have been developed in which a porous media is used in a flow through system. The porous media is used to immobilize the microbial cells which then metabolize the waste materials as they flow past in the aqueous media (Landis, 1990). If the waste mixes easily with water, contact with the active microbes will be accelerated. Often surfactants are used to facilitate the solubility of the waste materials. Biodegradation in soil is also possible with soil incorporation methods (Sanbom et al, 1977). The microbes used in biodegradation processes form complex communities consisting of saprophytic bacteria, fungi, and protozoa, with a complement of autotrophic organisms. The primary organisms are species that are indigenous to soil and sewage or mutant strains thereof (Pesticide Disposal and Detoxification Techniques). Recent research has identified a number of enzymes called organophosphorus acid (OPA) anhydrases that are able to rapidly break down organophosphate nerve agents such as VX and GB. Successful research in breaking down chemical agents 72 with enzymes was demonstrated in 1983 by the research group of Dr. F. C. G. Hoslcin at the Illinois Institute of Technology. Dr. James R. Wild and Dr. Wayne Landis have conducted research on a broad range of organophosphates including some nerve agents as well as some pesticides that are broken down by these enzymes (Landis, 1990; Wild, 1990). Chemical agents can be broken down into less toxic forms by bacteria and fungi that actually grow on the toxic material. For example, a chlorinated riot control material called CR was broken down into relatively nontoxic materials by microbial action. There are commercially available techniques in which microbial degradation is used to decontaminate groundwater, contaminated soils and stored hazardous wastes (EPA 1989). Freeze-dried bacterial cultures which are specifically suited to the degradation of certain types of waste are also commercially available (Pesticide Degradation and Detoxification). Bacterial cultures or consortia of mixed strains can be optimized for degradation of a specific type of waste material by exposing the culture to greater and greater concentration of the waste material (Haley et al, Aquatic Toxicology Vol. 13). The sections that follow will discuss the specific capabilities of microbial cultures for each of the different chemical bond moieties of chemical warfare agents and their neutralized by- products. Bioreactor design and specific experimentation on different classes of chemical compounds will then be discussed, after which the possibilities of biotechnology and bioengineering will be briefly examined. 73 Economics of Biological Treatment Biological treatment is favored by industry for dilute aqueous solutions because it is cheaper than incineration. Incineration costs are from 25 cents to $1 a pound for hazardous waste materials. The cost for an aerobic wastewater treatment system depends on the equipment used to aerate the system. A 2 million gallon per day (MGD) plant capital cost would be in the neighborhood of $8 to $10 million. Yearly operational costs would be in the neighborhood of $500,000 to $1 million. The decision to use incineration is based only on the disposal costs and not on environmental considerations. If a waste has heating value, incineration is the cheaper disposal method. Therefore, wastes with an organic content of greater than 10 percent are favored for incineration because of the fuel value. Because hazardous wastes disposal costs are based on volume or weight, dilution of hazardous waste prior to treatment is strictly avoided by industry (Anonymous, 1990). Industrial Use of Biological Methods Several of the largest U.S. chemical manufacturers are working on improved biological treatment technologies and have active and large biotechnology programs (Landis, 1990). Some of these wastewater treatment systems have dealt with chemicals that are not very water soluble. These systems include a wide variety of activated sludge type wastewater treatment systems. The research and development program that would typically be pursued in the industries using biological treatment would be to identify the microbes that can degrade the various waste compounds 74 and then apply them on a biological treatment system. Municipal wastewater typically has an organic carbon content of 300-400 mg/liter. Industrial wastewater systems of the anaerobic variety have a capability of degrading the wastewaters with 3,000-4,000 mg/liter organic carbon content. Aerobic systems have a capability of 30,000-40,000 mg/liter organic carbon content (anonymous 1990). Of the chemical agents in question, mustard is a difficult compound for aqueous treatment systems because it is marginally miscible in water. Furthermore, it skins over when exposed to water and forms globules. These globules are extremely resistant to envirnmental weathering. In fact, there is some in the ground left over from World War I in areas of France (Landis, 1990). A surfactant may be required for processing of mustard agents in aqueous treatment systems (Landis 1990). Surfactants can bring compounds that are typically soluble in the 5-10 ppm range into the 50 ppm range (Anonymous 1990). Biological treatment of nerve agents would be fairly straightforward. The microbes used for hydrolysis of the agents would yield phosphonic acid and hydrogen fluoride. The *phosphonic acid would then be cycled through a wastewater treatment system. Of the hydrolysis products of the chemical agents discussed in this paper, hydrogen cyanide (HCN) from nerve agent GA and arsenite from the mustard agent Lewisite represent the greatest health hazards. Hydrogen cyanide could be processed in the 75 oxidation processes discussed above, including electrolysis and chlorine techniques. Hydrolysis, biodegradation and photolyss products of chemical agents are shown in Appendix E. A review of the toxicology of these primary, degradation products is shown in Appendix F (DOA 1988). The Chlorine Carbon Chemical Bond Chlorinated solvents have been degraded by microorganisms (Boyer et al, 1988; Oldenhuis et al, 1989b; Galli and McCarthy, 1989; Mayer et al, 1988; Vogel and McCarthy, 1985). Other chemicals and classes of chemicals for which degrading microorganisms have been identified include chloroguaiacols, chloroveratroles, chlorocatechols (Nielson et al, 1987), chlorobenzenes (Bosma et al, 1988) substituted benzenes (Goulding et al, 1988), pentachloraphenol (Topp and Hanson, 1990), 2,4-D (Kelly et al, 1989) and dioxin (Clecka and Gibson, 1980). Some groups of species, notably the pseudomonas species and fungal species such as the white rot fungus have been found to have potential for degradation of a variety of groups of chemicals and are the subject of considerable research. Pseudomonas species have been found to break down chlorobenzenes (Haigler et al, 1988; Spain and Nishino, 1987), chlorotoluene (Haigler and Spain, 1989), lindane (Imai et al, 1989) and PCBs (Parsons and Sijm, 1988). The fungus Aspergillus niger is reported to be able to degrade the lower chlorinated technical mixtures of PCBs (Dmochewitz and Ballschmiter, 1988). The white rot fungus (Phanerochaete chrysosporum) can degrade DDT (Bumpus and Aust, 1987), 76 crystal violet (Bumpus and Brock, 1988), and pentachlorophenol (Mileski et al, 1988). The enzyme employed by the white rot fungus is a ligniase and this has been extracted from the fungus and used to oxidize a wide variety of refractory toxic compounds including PAHs and dioxins (Hammel et al, 1986; Thilly, 1990). The use of immobilized enzymes directly on the toxic materials is an alternate method of biological degradation technique. There is a great deal of interest in the capabilities of the ligniase associated with the white rot fungus in its broad capability to degrade toxic refractory organic materials. The chlorinated herbicide diuron was degraded by a pond water culture of microorganisms (Ellis and Camper, 1982). The microorganisms were isolated from pond water and sediments and examined in laboratory tests. Cultures capable of degrading the herbicide were isolated by enrichment techniques. Several mixed fungal/bacterial cultures and some mixed bacteria cultures had this capability. The mixed cultures degraded from 67-990/*, of the added herbicide, forming from 6 to 7 degradation products. Biodegradation and reduction in aquatic toxicity of the persistent riot control material 1,4-Dibenz-Oxazepine has revealed that this compound was degraded in a series of steps by naturally occurring microorganisms, specifically Alcaligenes denitrificans (Halley et al, Aquatic Toxicology, Vol. 13). Members of this genus of bacterium have been isolated from both terrestrial and aquatic environments (Kerster and DeLay, 1984). There have also been cases of Alcahgenes being isolated from 77 various human body fluids, such as blood, urine, and the spine (Fedorak and Westlake, 1983). This genus of gram negative bacteria is apparently widespread and quite versatile in xenobiotic metabolism. Halley and coworkers pointed out that if degradation by-products could be identified through radio tagged carbon, a consortium of bacteria could be arranged to degrade the by- products and completely eliminate CR toxicity. Further, Hailey and researchers point out that it may be possible to use microorganisms incorporated into a bioreactor matrix for toxic and hazardous waste cleanup. The organisms in Hailey's study markedly reduce the toxicity of the compound in question through degradation of the compound over a period of 20 days. The naturally occurring microorganism culture was obtained after 90 days of incubation and isolation, using successively higher concentrations of the toxic CR. Finally, an isolate was obtained that was resistant to CR in concen-trations up to 200 mg/Liter. Degradation of the parent compound CR was below detectable limits after 22 days (Haley et al, Aquatic Tox., Vol. 13). Anaerobic degradation of halogenated organics has been reviewed by Palmer and coworkers (Pahner et al, 1988). Anaerobic degradation is studied in this case because halogenated compounds are often resistant to aerobic degradation. The report reviews isolation of microorganisms capable of dehalogenation. Nutritional requirements of the organisms and the genetics of dehalogenation are reviewed by Palmer (1988). The ulti*mate goal of this research is to use the information collected to develop genetically engineered microorganisms with unproved anaerobic dehalogenation capabilities (Pahner 1988). 78 Organisms with Enzymes that Catalyze Organophosphate Hydrolysis Landis et al (1987a, 1987b) identified organophosphate acid anhydrase activities in the common brackish water clam Ringia cuneata. Tissue extracts from this clam were found to degrade potent neurotoxins, DFP and Mipafox. Enzymes capable of hydrolyzing DFP and related anti-acetylchohnesterase inhibitors, including soman nerve agent, have been reported in the tissues of many animals, including hog kidney, escherichia choli, mammalian tissues, the protozoan Tetrahymana thermophila and the clam Rangia cuneata (Hoskin, Kirkish and Steinman, 1984; Landis et al, 1987a, 1987b). The brackish water clam R. cuneata contained enzymatic activity that hydrolyzed soman (nerve agent) and was slightly enhanced by magnesium Mn 2** (Anderson, Durst and Landis, 1988). Five OPA anhydrases have been recognized in T. thermophila ranging in molecular weight from 67,000 to 96,000 D (Landis et al, 1987a). These enzymes differ from the squid type enzymes found in cephalapods that hydrolyze DFP and soman at a somewhat slower rate that have molecular weight of approximately 26,000 D (Hoskin, Kirkish and Steinman, 1984). Inactivation of organophosphorus nerve agents by the phosphotriesterase from Psuedomonase diminuta has demonstrated (Dumas et al 1990) this enzyme to be the most effective of those known for the specific activity against Saran, the major organophosphate nerve agent stockpiled by the United States and thought to be the major component of the Soviet arsenal (Robinson, 1980). The purified enzyme from P. diminuta has been shown hydrolyze the phosphorous-fluorine bond of soman and saran nerve agents. The authors state that the substantial rate enhancement exhibited by this enzyme for the hydrolysis of a wide variety of 79 organophosphorus nerve agents make this enzyme the priiine candidate for the biological detoxification of insecticide of mammalian acetylchohnesterase inhibitors (Dumas et al, 1990). The kinetics show that the catalytic rate constant for this enzyme represents a rate enhancement of 22 million times over saran hydrolysis in water at pH 7. This enzyme catalyzes the hydrolysis of soman 1.5 million times faster than the uncatalyzed reaction at pH 7 in water. The Carbon-Phosphorous Bond The carbon-phosphorous bond is an exceedingly nonreactive constituent of unsubstituted alkyl- and aryl-phosphonate nerve agents, insecticides, herbicides, fungicides, and flame retardants and several other economically important categories of chemicals. The C-P bond has been considered to resist cleavage by higher organisms (Menn, 1971; Menn and McBain, 1974). Daughton, Cook and Alexander (1979) have documented the bacterial conversion of alkyl-phosphonates to natural products via carbon- phosphorous bond cleavage. The phosphorous containing breakdown products of O-alkyl alkylphosphonate toxicants which are particularly resistant to cleavage at the C-P bond were fully degraded to natural products by Pseudomonas testosteroni. In an aerobic culture, where the phosphonate was the sole and limiting phosphorous source, the compound was degraded by release of the alkoxy as the alcohol followed by cleavage of the alkyl-phosphorous bond (methyl, ethyl, or propyl) to produce the respective alkane and an inorganic phosphorous compound that was detected as inorganic orthophosphate. P. testoteroni was not able to cleave the bonds of other carbon-heteroatoms such as arsonates, sulfonates 80 and mercurials. This research does show that an organophosphorus toxicant can be degraded biologically to yield natural products such as alcohols, alkanes, and phosphate (Daughton, Cook, Alexander, 1979). These same researchers have documented the cleavage of the carbon-phosphorous bond of phenylphosphonates to the molar yield of benzene from the phosphonate of 89 %. The organism was identified as a strain of Klebsiella pneumoniae (Cook et al, 1979). While this capability of carbon-phosphorous bond cleavage in the phenylphosphonate is interesting, it is not directly relevant to the chemical warfare agents in question. Biodegradation of Alkanes Soil microbes have a unique ability to utilize compounds that are biochemically inert to other biota. Aerobic bacteria can catalyze early steps in degradation cannot accomplish to form metabolites that can enter common pathways of metabolism such as the Krebs cycle or the fatty acid cycle. An alkane remains biochemically inert until a terminal carbon has been oxidized. The fatty acid then formed can enter the metabolic pathway known as beta oxidation. If a hydroxyl group is inserted at the appropriate point in a saturated ring structure, the molecule can be further oxidized to break the ring providing fragments that can be degraded. The aerobic microbes have the ability to use molecular oxygen in a controlled manner at the commencement of a metabolic sequence sufficient to initiate catabolism (Stanley Dagley, Degradation of Synthetic Organic Molecules in the Bio- 81 sphere). N-alkyl compounds can be degraded by beta oxidation following an initial attack by alpha or beta oxidation. However, extensive branching of an alkyl of moiety or terminal substitution rendering the alpha and beta positions inaccessible to the initial attack may i*mpede biological degradation (McKenna and Kaflio, 1964) in degradation of synthetic organic molecules in the biosphere. Moreover, if a quaternary carbon atom (or neopentyl group) occurs at the end of an alkane chain, the result is a molecule quite resistant to microbial attack (McKenna and Kallio, 1964). [] TL: ALTERNATIVE TECHNOLOGIES FOR THE DESTRUCTION OF CHEMICAL WEAPONS: AN INFORMATION PAPER SO: Greenpeace International (GP) DT: October 1990 Keywords: chemical weapons disposal alternatives gp reports toxics military / [part 4 of 4] Biological Treatment Systems Biological treatment systems are usually aqueous systems; however, biodegradation in soil is also possible with soil incorporation methods (Sanbom et al, 1977). Aqueous systems include those in which the microbes are fixed on a biofilm such as trickling filters, submerged filters, downflow or upflow filters, and fluidized beds. Rotating disks are also used as bioreactors. Other biological treatment systems use the microorganisms suspended in the waste material such as activated sludge, anaerobic digestion, and aeration tanks. Recent innovations include the immobilizing of selected strains onto porous media in a flow-through system in which the waste is cycled past the microbes. Traditional biodegradation processes use complex communities of microbes consisting of saprophytic bacteria, fungi, and protozoa, as well as autotrophic organisms. Extremes in pH or temperature may have a deleterious effect on the biological treatment system. Oxygen levels are maintained 82 as high as possible in aerobic systems, while anaerobic systems require the absence of oxygen (Pesticide Disposal and Detoxification Techniques). Often the different types of biological treatment units described above are combined in series. Different types of biological treat*ment reactors are shown in Figure 12. The activated sludge process consists of a tank or trough in which the waste stream is aerated in the presence of recycled activated sludge or biomass. This sludge is the source of acclimated microorganisms which accelerate the biodegradation of the unwanted substance (Pesticide Disposal and Detoxification Techniques). In the activated sludge process, the microorganisms are suspended in the aqueous media. A trickling filter, on the other hand, uses a bed of gravel or synthetic substrate on which the microorganisms attach in thin biofilms. The waste stream is sprayed onto the filter bed and allowed to trickle through. Good aeration is achieved due to the thin film of waste sprayed onto the filter bed. Trickling filters are not as effective as activated sludge, but are less sensitive to shock loads of changing waste composition. They are frequently used as the preliminary stage in wastewater treatment (Pesticide Disposal and Detoxification Techniques). An activated sludge system for the disposal of agent orange (2,4- D and 2,4,5-T) was conducted by the Utah State College of Engineering Water Resource Laboratory. T'he system was designed under contract with the U.S. Air Force to dispose of 2.3 83 ***Figure 12 Page 84 million gallons of surplus herbicide orange (Watchiski et al, 1974). The laboratory scale experiments in aqueous systems containing 230, 1,380, and 3,450 mg/Liter of the agent orange mixture. The mixtures were seeded with sewage soil and a proprietary preparation of freeze-dried bacterial*inoculum, which degrades phenolic materials (Phenobac). After a 16-day test period, the proprietary preparation of bacterial inoculum demonstrated a significant reduction of herbicide orange in all samples (64-730/o). The samples were incubated at 18 degrees C and presu*mably would have been more efficient at the optimum growth temperature of 30 degrees C. A tapered fluidized bed bioreactor was developed at Oak Ridge National Laboratory for microbial degradation of coal conversion, aqueous waste, hydrogen sulfide, ammonia, phenols, thiocyanates and other hydrocarbons (Lee and Scott, 1977; Hancher 1977)*. The fluidized bed bioreactor is shown in Figure 13 and consists of a column with a cone-shaped bottom. The waste material is introduced at the bottom of the column through the fluidized bed of coal particles and out the top of the column. An aeration gas stream agitates the coal particles within the column. The coal is thereby suspended in the waste liquid and serves as a substrate for growth of biofilm of the desired organisms for the immobilization of an enzyme (Scott and Hancher, 1976). 85 ***Figure 13 Page 86 Trickling filters and activated sludge processes are used widely by the pesticide manufacturing industry (Atkins, 1972; von Rumker et al, 1974). There have been several innovations which duplicate or enhance the functions of trickle filters and activated sludge reactors and require much less space. These are the rotating disk and the fluidized bed reactors. The rotating disk or biodisk has been common in Europe for about 18 years (Autotrol, 1971). The process uses thin film biodegradation on large plastic disks that are half submerged in the waste water. The microorganisms colonize the disk surfaces and receive maximum oxygenation as the disks are slowly rotated. Many disks are aligned on a single shaft and many of these units are placed in series providing a maximized well aerated surface requiring relatively little space. The rotating disk bioreactor is shown in Figure 14. Fluidized bed reactors are commercially available at full scale and have been used for processing municipal sewage as well as several industrial waste streams including pharmaceutical wastes and petroleum tank truck wastes (Jeris et al, 1977; Owens 1977). A variation of the fluidized bed bioreactor incorporates a highly porous packing material to which the microbes adhere in the reactor column. Air is supplied by fine bubble membrane diffusers mounted at the bottom of each column. One such system can be run under anaerobic conditions also. A system demonstration under the EPA site program incorporates a conditioning step in which the contaminated 87 ***Figure 14 Page 88 waste stream enters a mix tank where the pH is adjusted and inorganic nutrients are added. If necessary the waste stream is heated to the opti*mum temperature (EPA 1989). One of the EPA site program demonstrations includes a liquid/solid contact digestion system in which sludges or soils containing from 2,000 to 800,000 ppm total organic carbon are degraded biologically in several stages. T*he process makes use of acclimated seed bacteria and aerobic biological oxidation. Emulsifiers are added to increase solubility of the organic materials. The technology is thought to be suitable for treating halogenated as well as nonhalogenated organic compounds and some pesticides and herbicides. The system has been demonstrated on liquids, sludges and soils with high organic concentrations. A schematic of the system is shown in Figure 15. A demonstration project is proposed for testing this technology on contaminated soils from a wood preserving facility in April of 1990 (EPA 1989). Mutant Bacterial Seed Several freeze-dried biochemical preparations containing mutant bacteria and substances which enhance their growth are commercially available. These products may be added to new or existing biological aerobic wastewater treat*ment systems. Some may withstand a wide variety of waste stream conditions. One such preparation was demonstrated at the Mobay Chemical Company in West Virginia 89 ***Figure 15 Page 90 after a natural gas shortage had shut down Mobay's activated sludge process. The activated sludge process contained only 5% of the biomass expected during normal operations. Attempts to revive the process using sludge from a nearby treatment and an easily degradable waste feed failed. Considerable success was obtained by using the commercial mutant bacterial seed and the activated sludge system was fully effective after four weeks (Pesticide Degradation and Toxification). Immobilized Enzymes Enzymes capable of degrading organophosphate (hydrolysis) have been isolated from the parent cells and used to degrade organophosphate after immobilization on a substrate material (Munnecke, 1977). J. R. Wilde at Texas A&M isolated a soil bacterial enzyme that degraded soman and saran and immobilized it on the inside of plastic tubing. When dilute aqueous solutions of pesticides were passed through the tubing, 100% degradation was achieved (Wild, 1990). Bioengineering The gene for the organophosphate hydrolysis enzyme was isolated in 1984 and has been successfully introduced in a variety of organisms (Wilde, 1990). While the research and capabilities are at hand, the process has not been funded for development on an industrial scale thus far. 91 END