Environmental Protection Technology Series DEVELOPMENT OF MICROWAVE PLASMA DETOXIFICATION PROCESS FOR HAZARDOUS WASTES Phase I Municipal Environmental Research Laboratory Office of Research and Development U.S. Environmental Protection Agency Cincinnati, Ohio 45268 ------- RESEARCH REPORTING SERIES Research reports of the Office of Research and Development, U.S. Environmental Protection Agency, have been grouped into nine series. These nine broad cate- gories were established to facilitate further development and application of en- vironmental technology. Elimination of traditional grouping was consciously planned to foster technology transfer and a maximum interface in related fields. The nine series are: 1. Environmental Health Effects Research 2. Environmental Protection Technology 3. Ecological Research 4. Environmental Monitoring 5. Socioeconomic Environmental Studies 6. Scientific and Technical Assessment Reports (STAR) 7. Interagency Energy-Environment Research and Development 8. "Special" Reports 9. Miscellaneous Reports This report has been assigned to the ENVIRONMENTAL PROTECTION TECH- NOLOGY series. This series describes research performed to develop and dem- onstrate instrumentation, equipment, and methodology to repair or prevent en- vironmental degradation from point and non-point sources of pollution. This work provides the new or improved technology required for the control and treatment of pollution sources to meet environmental quality standards. This document is available to the public through the National Technical Informa- tion Service, Springfield, Virginia 22161. ------- EPA-600/2-77-030 April 1977 DEVELOPMENT OF MICROWAVE PLASMA DETOXIFICATION PROCESS FOR HAZARDOUS WASTES PHASE I by Lionel J. Bailin Barry L. Hertzler Lockheed Palo Alto Research Laboratory LOCKHEED MISSILES & SPACE COMPANY, INC. Palo Alto, California 94304 Contract No. 68-03-2190 Project Officer Donald A. Oberacker Solid and Hazardous Waste Research Division Municipal Environmental Research Laboratory Cincinnati, Ohio 45267 MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY CINCINNATI, OHIO 45268 ------- DISCLAIMER This report has been reviewed by the Municipal Environmental Research Laboratory, U.S. Environmental Protection Agency, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the U. S. Environmental Protection Agency, nor does mention of trade names or commercial products constitute endorse- ment or recommendation for use. 11 ------- FOREWORD The Environmental Protection Agency was created because of increasing public and government concern about the dangers of pollution to the health and welfare of the American people. Noxious air, foul water, and spoiled land are tragic testimony to the deterioration of our natural environment. The complexity of that environment and the interplay between its components require a concentrated and integrated attack on the problem. Research and development is that necessary first step in problem solution and it involves defining the problem, measuring its impact, and searching for solutions. The Municipal Environmental Research Laboratory develops new and improved technology and systems for the prevention, treatment, and management of wastewater and solid and hazardous waste pollutant discharges from munici- pal and community sources, for the preservation and treatment of public drinking water supplies, and to minimize the adverse economic, social, health, and aesthetic effects of pollution. This publication is one of the products of that research; a most vital communications link between the researcher and the user community. This report describes a current and successful research and development program whereby low-temperature microwave plasmas have been utilized to detoxify efficiently and safely various hazardous pesticides and organic wastes. As a result of the attainment of project objectives, further scale- up to larger size and capacity are under way at the pilot level, to be followed by field verification, which is in the planning stages. Francis T. Mayo Director Municipal Environmental Research Laboratory ill ------- PREFACE Toxic organic substances, such as chlorinated hydrocarbons, and many organophosphorus, organonitrogen, and organometallic compounds are com- ponents of pesticides which have been withdrawn from use, are obsolete, or are components of hazardous industrial wastes. These materials must be managed or disposed of safely and effectively. A primary responsibility of the U.S.Environmental Protection Agency's Solid and Hazardous Waste Research Division (SHWRD) has been to encourage and support research and development efforts in the area of hazardous waste disposal technology. For compounds of nominal toxicity, laudable achievements have been accom- plished in the technology of thermal destruction, chemical detoxification, long-term encapsulation, and special landfill methods. However, with the exception of high-cost incinerator processing, little or no new technology has been developed for the disposal of highly toxic, refractory, or extreme- ly persistent wastes. These materials are described in the 1974 EPA Re- port to Congress, Disposal of Hazardous Wastes, SW-115 The microwave plasma process described in this report is a relatively new application of what has been termed the "fourth state of matter", or the . "plasma state". It is the first practical application of a microwave dis- charge to the decomposition of chemical compounds in significant quanti- ties. IV ------- TECHNICAL ABSTRACT The process of microwave decomposition of organic materials has been applied to the detoxification/destruction of hazardous organic pesticides and wastes. The investigation began with the LPARL (Lockheed Palo Alto Re- search Laboratory) laboratory-size plasma system. Reaction efficiencies and reaction product identities were obtained. The detoxification process was then expanded to a larger-scale, continuous system employing custom- designed, commercially available hardware. A primary objective of the program was concerned with evaluation of the effectiveness of the expanded-scale system using hazardous organic com- pounds and wastes of current interest. The products were identified to de- termine whether the products were innocuous or toxic, and whether recovery of useful materials as by-products was feasible. Test Detoxifications in LPARL Unit The materials selected for study were passed through the LPARL lab- oratory-size reactor, which handles 1 to 5 g/hr quantities, to evaluate conversion efficiency and to determine product identities. Oxygen and argon as carrier gases in combination with the toxic materials were tested and evaluated. Conversion efficiency was calculated by means of mass-balance deter- minations; and product identity by means of instrumental analysis, e. g., infrared and mass spectroscopy. The mechanisms of the detoxification processes were postulated. Materials tested were pure Malathion liquid, liquid PCB's, phenylmercuric acetate solution, and methyl bromide gas. Expanded-Scale Plasma System An expanded-volume microwave plasma system was assembled from custom-designed commercially available microwave hardware. The plasma reactor tubes were constructed from transparent quartz. Materials feed systems for pure liquids, slurries, and solutions were designed and built to operate at reduced pressures, and installed with proper controls for monitoring the addition rates. Traps and separation units were installed as needed for the decomposition products. Test Detoxifications in Expanded-Scale System Liquid, slurry and solution materials were processed in the plasma reactor to evaluate conversion efficiency, by-product recovery, and product ------- identity. The principal goal was to maximize throughput at greater than 99. 9% conversion. Materials detoxified were pure Malathion liquid, liquid PCB's, phenylmercuric acetate solution, and Kepone in the form of a solid powder, an aqueous slurry, and a solvent solution. Metallic mercury was recovered as a by-product from the decomposition of phenylmercuric ace- tate pesticide. The Kepone and organomercurial pesticides were considered as "real-world" systems because of their commercial sources of manufac- ture, and the nature of their composition as mixed system pesticides. Throughput of the system was maximized at 450 to 3200 g/hr (1 to 7 Ib/hr), a multiplication factor of approximately 500 times that of the laboratory- scale unit. Specific engineering tasks have therefore been outlined for the design and construction of pilot-scale equipment and components for further expansion of the system to 20 Ib/hr, to be followed later by a 100 Ib/hr ver- sion. This report was submitted in fulfillment of Contract No. 68-03-2190 by the Lockheed Palo Alto Research Laboratory of Lockheed Missiles & Space Co., Inc., under the sponsorship of the U. S. Environmental Protection Agency. The report covers the period April 1975 through May 1976. vi ------- EXECUTIVE SUMMARY This report describes recent, successful, research and development efforts to utilize low-temperature microwave plasmas for the safe and effi- cient detoxification of various hazardous wastes which threaten man and his environment. Approximately 10 million tons of such wastes are generated yearly, and are made up in part of the following types of materials: • Pesticides which have been withdrawn from use • Obsolete or below-specification toxic materials which cannot otherwise be utilized • Industrial wastes from process streams, chemicals, explosives, etc. Perhaps 10 to 20 percent of these wastes will need special methods for disposal. Some may be incinerated, or disposed of by other means. How- ever, those containing highly toxic compounds will require more sophisti- cated techniques for disposal. The EPA Solid and Hazardous Waste Research Division was attracted to the microwave plasma process during its search for new and novel means for detoxification of hazardous and toxic waste streams. Previously, a Lockheed-built bench-scale device had won the attention and support of the U.S. Army Edgewood Arsenal for the detoxification of trace nerve gas simu- lants in contaminated air streams or interior building environments. Proof of Concept tests were conducted during 1970-1972, with considerable suc- cess. Subsequently, the EPA supported a laboratory study to test the effi- cacy of the technique, using liquids and solids instead of the diluted gases. This program included actual pesticides and toxic materials instead of simu- lants. During the first year's effort, 1975-1976, the process was remarkably successful in detoxifying all of the materials submitted for testing. These materials included a typical organophosphorus pesticide (Malathion), a ro- dent poison (methyl bromide), two polychlorinated biphenyl liquids (PCB's), a mercury-containing fungicide (phenylmercuric acetate), and Kepone, the highly chlorinated pesticide of some notoriety. Scale-up operations were also successful. Starting at from 1 to 5 grams per hour in the laboratory-size plasma unit, the first level of throughput ex- pansion was maximized at 450 to 3200 g/hr (1 to 7 Ib/hr) - a multiplication factor of approximately 500 times. vn ------- Further benefits of the process are competitive, reasonable costs of about $. 20/lb; the use of steam as well as oxygen as a carrier gas; and the potential, verified in one case, for resource recovery of valuable mate- rials, such as metallic mercury, from the organomercurial pesticide, phenylmercuric acetate. Prospects for additional applications include the detoxification of car- cinogenic compounds used by the U.S. Navy in their colored smoke formu- lations, the recovery of metal values from arsenic, cadmium, and zinc organometallic pesticides, and the decomposition of vinylchloride solid wastes and related products. The process warrants further R&D effort, namely, additional scale-up to larger sizes and capacities that will be com- patible with anticipated needs at the pilot- and field-verification levels. The disposal of a few hundred pounds per hour will be evaluated and developed as a safe, essentially simple, and economic process for materials of high toxicity. Field demonstration testing and construction of systems at na- tional or regional disposal sites are long-term projected goals. Vlll ------- CONTENTS Foreword ^ Preface iv Technical Abstract Y Executive Summary yi| Figures xi Tables xiii Acknowledgments x^v 1. Introduction 1 2. Microwave Plasma Characteristics and Basic Process Description 2 3. Equipment and Operating Procedures 4 3. 1 Laboratory-scale plasma unit 4 3. 2 Expanded-scale plasma system 9 3.3 Analytical methods 19 4. Selection of Pesticides and Hazardous Waste Materials 20 5. Results of Laboratory-Scale Oxygen Plasma Reactions 22 5. 1 Malathion 22 5.2 Poly chlorinated biphenyls (PCB's) 22 5. 3 Methyl bromide 24 5.4 Phenylmercuric acetate 24 6. Results of Expanded-Scale Oxygen Plasma Reactions 25 6. 1 Malathion 25 6.2 Poly chlorinated biphenyls (PCB's) 25 6. 3 Phenylmercuric acetate 27 6.4 Kepone - a "real world" mixture 27 6. 5 Steam plasma detoxification of PMA-30 28 6. 6 Process costs calculated for PMA-30 28 7 Discussion of Results 32 7. 1 Interpretation 32 7. 2 Chemistry of oxygen plasma reactions 33 7. 2. 1 Malathion - oxygen plasma reactions 33 7. 2. 2 PCB - oxygen plasma reactions 33 7. 2. 3 Methyl bromide - oxygen plasma reactions 34 7. 2. 4 Phenylmercuric acetate - oxygen plasma reactions 34 7. 2. 5 Kepone - oxygen plasma reactions 34 7. 3 Evaluation of ancillary equipment 35 7. 3. 1 Materials feed techniques 35 7. 3. 2 Packed bed technology 37 7. 3. 3 Reduced pressure systems 37 7. 3. 4 Analytical processes 37 IX ------- 8. Conclusions 40 9. Recommendations 42 References 44 Appendices A. Mass Spectra 45 A. 1 Equipment 45 A. 2 Pesticide and hazardous waste spectra 49 A. 2. 1 Malathion - oxygen plasma 49 A. 2. 2 Aroclor No. 1242 PCB - oxygen plasma 49 A. 2. 3 Aroclor No. 1254 PCB - oxygen plasma 49 A. 2. 4 Methyl bromide - oxygen plasma, laboratory- scale reactor, quantitative analysis 49 A. 2. 5 Troysan PMA-30, phenylmercurie acetate - oxygen plasma 54 A. 2. 6 Troysan PMA-30 - steam plasma 54 A. 2. 7 Kepone - oxygen plasma 54 B. Properties of Pesticides and Hazardous Wastes 5^ B. 1 Malathion 58 B. 2 Poly chlorinated biphenyls 59 B. 3 Methyl bromide gas 59 B. 4 Phenylmercuric acetate, PMA 60 B. 5 Kepone 60 C. Chemical Reactions in Argon and Hydrogen Plasmas 62 C. 1 Malathion-argon and malathion argon-oxygen plasmas 62 C. 2 PCB - argon plasma reactions 63 C. 3 PCB - argon-oxygen plasma reactions 63 C.4 PCB - hydrogen plasma reactions 63 C. 5 Methyl bromide - argon plasma reactions 64 C. 6 Methyl bromide - argon-oxygen plasma reactions 64 ------- FIGURES Number Page 1 Schematic diagram of laboratory microwave plasma system 5 2 Laboratory microwave plasma reactor system 6 3 Block diagram of microwave plasma system and related components 7 4 Reduced pressure liquid feed system — laboratory- scale 8 5 Quartz mesh basket within plasma reactor 10 6 Expanded-scale microwave plasma system — schematic 11 7 Expanded-scale microwave reactor — hardware and associated electronics 13 8 Series A microwave reactor system 14 9 Series B microwave reactor system 15 10 Series C microwave reactor system 16 11 Microwave power sources and control console 17 12 Hazardous liquid injector system, 2 liter 18 13 Infrared spectra of Kepone acetone extract (curve A) and its oxygen plasma reaction product (curve B) 36 A-1 Gas sampling system interface with plasma reactor and residual gas analyzer 46 A-2 Mass spectrometer residual gas analyzer — variable leak valve and quadrupole head 47 A-3 Mass spectrometer residual gas analyzer — power supplies and vacuum gages 48 XI ------- Number Page A-4 Mass spectra of gaseous effluent from reaction of Malathion in oxygen plasma 50 A-5 Mass spectra of gaseous effluent from reaction of AroclorNo. 1242 PCB in oxygen plasma 51 A-6 Mass spectra of gaseous effluent from reaction of AroclorNo. 1254 PCB in oxygen plasma 52 • A-7 Mass spectroscopic analysis of methyl bromide oxygen plasma reaction 53 A-8 Mass spectra of gaseous effluent from reaction of Troysan PMA-30 (phenylmercuric acetate methanol solution) in oxygen plasma 55 A-9 Mass spectra of gaseous effluent from reaction of Troysan PMA-30 (phenylmercuric acetate methanol solution) in steam plasma 56 A-10 Mass spectra of gaseous effluent from reaction of Kepone-methanol/acetone in oxygen plasma xu ------- TABLES Number Page Pesticides and Hazardous Waste Materials for Detoxification Tests 21 Representative Laboratory-Scale Oxygen Plasma Reactions 23 Summary of Expanded-Scale Oxygen Plasma Reactions 26 Electrical and Carrier Gas Costs for Plasma Reactions in Series C Reactor Expanded-Scale System 29 Influence of Packed Bed on PMA-30 (Phenylmercuric Acetate) Conversion in Series C Plasma System 38 xiu ------- ACKNOWLEDGMENTS This report was prepared by the Lockheed Palo Alto Research Labora- tory, Lockheed Missiles & Space Company, Inc., under U. S. Environmental Protection Agency Contract No. 68-03-2190. The work was performed in the Laboratory of Chemistry, Dr. Ernest L. Littauer, Manager. Dr. Lionel Bailin was Principal Investigator. Major contributions to the program were made by Dr. Barry L. Hertzler, plasma chemistry and mass spectroscopy; Dr. Gerald B. McCauley, infrared spectroscopy and related analysis; Dr. Sid J. Tetenbaum, plasma physics; and Prof. Alex T. Bell, University of California-Berkeley, consultant, plasma reactions and kinet- ics. The Project Officer for the Environmental Protection Agency was Mr. Donald A. Oberacker; Mr. Richard A. Carnes served as Technical Consultant. xiv ------- SECTION 1 INTRODUCTION Although the microwave plasma detoxification of pesticides and hazard- ous organic wastes is a relatively new process, investigations in plasma chemistry began as early as 1874 (1). In 1969, the Lockheed Palo Alto Research Laboratory (LPARL) was the first to investigate the application of a microwave discharge to the decomposition of toxic substances. A series of experiments was run in order to show feasibility for decom- posing toxic gases by an electrical discharge (2). Nerve gas simulants from the U.S. Army Edgewood Arsenal were tested in the process during 1970 - 1972. Decompositions were achieved at near 100 percent levels in the lab- oratory-size microwave cavity. Although the decomposition efficiencies were high, the throughput levels were low, about 1 to 5 g/hr (3, 4). This was the result of the small size of the reactor, which, however, utilized the largest cavity commercially available at the time. When it was determined that large-volume microwave applicators for expanded-scale plasma forma- tion could be obtained on a custom basis from microwave hardware sup- pliers (5), the EPA Solid and Hazardous Waste Research Division (SHWRD) approved a feasibility study to test the process on several pesticides and toxic wastes. In the following report are described the chemistry of the detoxification reactions, an initial scale-up of the microwave hardware, and an evaluation of the process in which up to 3200 g/hr (7 Ib/hr) were success- fully decomposed to harmless or easily disposable effluents. ------- SECTION 2 MICROWAVE PLASMA CHARACTERISTICS & BASIC PROCESS DESCRIPTION In the present context, the term "microwave plasma" denotes an ion- ized gas produced by microwave-induced electron reactions with neutral gas molecules. To a first approximation, the electrons, rather than thermal or radiative sources, are the major contributors of energy for promotion of the chemical reactions. The reactions are initiated by collisions between the reactant molecules and electrons in the plasma. Organic free radicals are generally considered as the primary intermediates in this process. The ionized gas, or plasma, is also derived from the carrier gas which serves to transport the molecules into the plasma zone. When microwave frequencies of 100 to 10, 000 MHZ are used to induce ionization, the field in the cavity oscillates so rapidly that the force on the electrons changes direction before the electrons can travel far; thus, the plasma is not swept out of the discharge region (6). Consequently, the electrons and ionized gas particles remain within the discharge, thereby contributing to the chemical reactions in the plasma. Microwave plasmas show many favorable characteristics (7) when com- pared with arc or electrode plasmas. These include the following: • Production of high ionization levels and molecular dissociation without excess heating of the contained gas • Construction of reaction vessels which are simple, free from contamination and less subject to damage because of the absence of internal electrodes • Production of little or no electrical interference • Absence of high voltages which can be easily contacted by operating personnel, i. e. absence of shock hazards. Microwave plasmas, rather than radio frequency (RF) plasmas (1 to 100 MHz), have been chosen because an RF discharge is difficult or impossible to develop in uniform cross-section (8). This difficulty is de- tailed for gas phase oxidation reactions in a recent publication by Brown and Bell (9). It appears that the ability of RF plasmas to interact fully with liquids and solids would therefore be severely compromised. ------- Since the plasma decomposition mechanism involves electronic rather than thermal energy, the microwave applicator power coupling equipment can be maintained at low temperatures, that is, barely hot to the touch. The materials of construction which are associated with furnace or inciner- ator devices will therefore be generally unnecessary, and maintenance and repair expenses should be low or nonexistent. The microwave plasmas described in this report were first produced in a laboratory-size resonant cavity, and then in three commercially-built trough waveguide applicators. Data on power levels utilized, flow rates, conversion efficiencies, and other plasma characteristics are enumerated in the sections associated with the respective toxic materials. Process equip- ment, hardware, and procedures are described in the next section. ------- SECTION 3 EQUIPMENT AND OPERATING PROCEDURES 3. 1 LABORATORY-SCALE PLASMA UNIT During the initial stages of the program, the plasma process was evalu- ated on pesticide materials in the same apparatus (4) which was utilized for decomposing the original organophosphonate simulants. In this way, a valid transition would be possible when the materials were run in the larger size microwave system, which was then under construction. The laboratory unit was effectively the same as used previously, with the exception that the feed techniques required modification for injection of liquids and solutions. A schematic line diagram is shown in Figure 1, and a photograph in Figure 2. A block diagram of the microwave plasma detoxi- fication system is given in Figure 3. This diagram applies to the different size systems irrespective of the type of microwave applicator or power source that is used. The microwave power source was a Varian PPS-2. 5A unit, which has an output frequency of 2450 MHz and develops up to 2. 5 kW. A Microlite 287 microwave leakage detector (Crystal Mfg. Co., Oklahoma City, OK) was used to monitor power leakage. Levels of leakage greater than 1 to 3 mW/cm2, which is the lower limit of sensitivity of the apparatus, were not detected in the immediate vicinity of the discharge tube. One materials-feed unit that was found to offer a practical solution for dropwise addition of liquids was a pressure-equalizing dropping funnel of approximately lOO-cm* capacity installed at the input to the plasma reaction tube, see Figure 4. The manipulative procedure involved, first, filling the dropping funnel with the liquid to be tested. The lowest Teflon needle-valve stopcock is in the closed position. The two remaining stopcocks are open. The entire sys- tem is evacuated to its minimum pressure, about 0. 1 torr (mm Hg). For pure or sol vent-free liquids, such as pure Malathion or polychlorinated bi- phenyls (PCB's), an ice water sample receiving trap is located directly below the plasma exhaust outlet, and liquid nitrogen-cooled traps are located down- stream from the gas-sampling tube to collect condensable products and to minimize contamination of vacuum pump oil. Next, pressure is adjusted to the desired level by regulation of the main vacuum valve. A Tesla coil is generally used to ignite the discharge. When the Teflon needle valve is ------- 02 GAS SUPPLY- Ar GAS VALVE FLOWMETER 1 SUPPLY—^X) VAC UUM BACK-FILL LINES QUADRUPOLE GAS ANALYZER DROPPING FUNNEL LN COLD TRAP U MANOMETER GAS SAMPLING TUBE PLASMA REACTOR DISCHARGE CAVITY MICROWAVE POWER SUPPLY ICE-H2O RECEIVING TRAP Figure 1. Schematic diagram of laboratory microwave plasma system ------- r. Figure 2. Laboratory microwave plasma reactor system ------- MATERIAL TO BE DETOXIFIED AND CARRIER GAS SOURCE MICROWAVE PLASMA CAVITY PRODUCT RECEPTOR TUNING CIRCUITRY MICROWAVE POWER SOURCE ANALYTICAL INSTRUMENTATION FOR CHEMICAL ANALYSIS Figure 3. Block diagram of microwave plasma system and related components ------- 30 CM 85 CM TEFLON NEEDLE- VALVE STOPCOCK TEFLON GASKET QUICK DISCONNECT TEFLON NEEDLE- VALVE STOPCOCKS TEFLON GASKET QUJCK DISCONNECT QUARTZ REACTOR TUBC 15 mm OD Figure 4. Reduced pressure liquid feed system — laboratory-scale 8 ------- opened, small drops fall under gravity through the quartz reaction tube. Reaction product sampling is performed at the gas sampling tube, and the ice water and liquid nitrogen traps. Calculations indicated that the rate of fall of a drop of pesticide at zero pressure through an average-size, 5-cm length plasma was 0. 1 sec, using for the approximation S = 1/2 g cr^ where S is the distance traveled, g is the gravitation constant (980 cm/sec2), and a is time in seconds. The actual residence time or will be shorter, since the liquid drops fall in the same direction as that of the carrier gas. However, since working pres- sures on the order of 10 to 100 torr are more the rule, the drag components associated with these pressures will lengthen the time over the estimated free-fall time. Therefore, the extent of change in a is related, in addition to the size of the liquid drop, its density, and the operating characteristics of the carrier gas. Nevertheless, it was apparent that a method would be needed to increase the time for passage of the drop through the discharge. The technique by which this was accomplished utilized a quartz mesh "basket" positioned at the center of the plasma zone. Quartz mesh fibers were loaded into the basket, and served as a contact area for the drops. The basket contained a sufficient number of holes to allow passage of the effluent reaction products, as shown in Figure 5. The process is continuous. The residence time has been estimated at 1/2 to 1 sec, the time for on-off occurrences of reaction flashes inside the plasma zone. For gas feeds, such as methyl bromide, the dropping funnel was by- passed, and the pesticide was fed directly into the plasma. With respect to additions of solids, this was deferred until the larger plasma system could be utilized. It was considered that development of a uniform method for feeding these materials into the narrow 15-mm o. d. reaction tubes would not be productive relative to the tasks prescribed for evaluation in the expanded-scale system. 3. 2 EXPANDED-SCALE PLASMA SYSTEM A primary objective of the-program was the evaluation of the effective- ness of an expanded-scale microwave plasma system in processing hazard- ous organic compounds, wastes, and pesticides of current interest. In this regard, the reaction products were identified to determine whether the products were innocuous or toxic, and whether there was the possibility for recovery of useful materials as by-products. The new microwave plasma applicator hardware and power supply- equipment were supplied by Gerling Moore, Inc., Palo Alto, California, according to LPARL specifications. Figure 6 shows a schematic diagram of the microwave electronics and the associated process equipment. The principal difference between this system and the laboratory model is the method of application of power. In the laboratory unit, the applicator uti- lized a resonant cavity, Varian Model EC2-DRS2, which was fed by a single 2. 5-kW power supply. In the expanded-scale unit, rectangular trough wave- ------- LIQUID PLUS CARRIER GAS TEFLON NEEDLE-VALVE STOPCOCK CARRIER GAS QUARTZ REACTOR TUBE QUARTZ — BASKET [TEFLON GASKET QUICK DISCONNECT I •*• PLASMA ZONE TRAPS AND VACUUM PUMP Figure 5. Quartz mesh basket within plasma reactor 10 ------- PESTICIDE DROPPING FUNNEL MICROWAVE POWER SOURCE MICROWAVE APPLICATOR MICROWAVE POWER SOURCE x-PLASMA REACTOR TUBE MASS SPECTROMETER Y Y I I FLOW METERS O SUPPLY ALTERNATE Z GAS SUPPLY 3-WAY STOPCOCK MANOMETER COLD TRAP —\ ri—0__*.VACUUM PUMP (LN) THROTTLE 1 VALVE RECEIVER^ COLD TRAP Figure 6. Expanded-scale microwave plasma system — schematic 11 ------- guides were used. Two 2. 5-kW power supplies fed the applicator, one sup- ply for each ridge. A diagram of the microwave hardware and the associated electronics is shown in Figure 7. In the design of the applicator, three lengths were chosen for the evalu- ations. These dimensions, exclusive of the two 21. 6 cm (8. 5 in. ) end- mounted radiation cutoff sections, are listed as follows: • Series A 172 cm (68 in. ) • Series B 92 cm (36 in. ) • Series C 41 cm (16 in. ) The interior dimensions (i. d. ) were the same for the three units: 21. 6 cm (8. 5 in. ) x 5. 7 cm (2. 24 in. ). The i. d. of the cutoff guides was 5. 7 cm (2. 24 in. ) x 5. 7 cm (2. 24 in. ). The ridges were 2. 5 cm (1 in. ) x 1. 25 cm (0. 5 in. ) attached at the narrower dimension axially to the 5. 7-cm section of the waveguide. In Figures 8 through 10, the three systems are shown in assembly. Figure 8 shows the applicator slanted approximately 30 degrees. The vari- able positioning was feasible through use of the flexible waveguides from the power supplies. The master control unit, power supplies, etc. , were in- stalled in a single movable rack (Figure 11). The reactor tubes were fabricated from transparent quartz with a 1. 5 mm wall thickness, and an o. d. of 49 to 50 mm. Raschig rings, fabri- cated from quartz were used to fill the reactor in order to increase the residence time of the toxic materials in the plasma zone. The ring dimen- sions ranged from 8 mm o. d. x 4 mm length to 10 mm o. d. x 10 mm length. See Subsection 7. 3. 2 for development of optimum size of the rings. The pesticide dropping funnel was a 2-liter version of the unit used for the laboratory-scale plasma tests (Figure 12). For relatively volatile solu- tions, a standard 250-cm3 volumetric dropping funnel was used for feeding directly into the reactor, using atmospheric pressure above the solution (Figure 10). The vacuum pump was a Welch DuoSeal Model 1397 oil-sealed two-stage mechanical pump with a free-air displacement of 425 liter/mm. Various traps were installed between the reactor output and the pump for product collection and to maintain cleanliness of the pump oil. The output from the pump was passed through a stack exhaust to the exit port on the roof of the building. The following procedure for operating the larger plasma systems was only slightly different from that of the laboratory unit. The system was eva- cuated initially to its minimum pressure, about 1 torr. With liquid-nitrogen or other low-temperature coolants in the trap condensers, the flow of oxygen was adjusted to maintain a pressure of about 10 torr. With all water and 12 ------- PART 5 PART 4 PART PART 3 2 PART 7 PART PART PART 5 PART 4 PART 6 1 w III 1 1 II 1 n LIIIIl L f l „ ,' Ijijiji' PART PART 2 - TRIPLE STUB TUNER PART 3 - DUAL POWER MONITOR PART 4 - FLEXIBLE WAVEGUIDE PART 5- 3-PORT CIRCULATOR AND WATER LOAD PART 6- VARIABLE POWER SOURCE AND CONTROL PART 7 - 208 VOLT WATT-METER Figure 7. Expanded-scale microwave reactor—hardware & associated electronics 13 ------- Flexible Wave Guide Dual Power Monitor Flexible Wave Guide Stack Exhaust Microwave Plasma Applicator Carrier Gas Flow Meters Pesticide Dropping Funnel LN Trap Bottom End of Reactor Tube Figure 8. Series A microwave reactor system 14 ------- Figure 9. Series B microwave reactor system 15 ------- Figure 10. Series C microwave reactor system 16 ------- Figure 11. Microwave power sources and control console 17 ------- ! Figure 12. Hazardous liquid injector system, 2 liter 18 ------- air cooling on, the microwave power sources were activated. As soon as the plasma ignited, about 1 sec, the flow of oxygen was increased, and the pump throttle valve adjusted to achieve the desired combination of oxygen flow and pressure. The microwave power was set to the desired level, and the tuning controls adjusted to give minimum reflected power. During the procedure, a Holaday Model HI-1500- 3 microwave radiation monitor was used to meter leakages in radiation,if any. Levels greater than 1/2 mW/cm2 were not detected in the immediate vicinity of the discharge tube. 3. 3 ANALYTICAL METHODS Two principal methods were applied for materials analysis. Of these, mass spectroscopy yielded the maximum information on the identification of the gaseous products produced in the plasma. Infrared analysis was used for determination of effluent materials, which included solids and liquids, deposited in the product receiver traps. Gas chromatography and standard quantitative techniques for mass balance and heavy metals were adopted as required; details of these analyses are given in the section pertaining to the individual pesticides. Mass spectrometric determinations were performed on a Varian Model 974-0002 residual gas analyzer, which includes an in-line continuous gas sampling system. Two ranges of atomic mass units (amu) were used as applicable to the specific pesticide or waste: 1 to 50 amu (for gases such as N2> °2, CO2, CO, H2O, CH4, C2H6, etc. ) and 1 to 250 amu maximum (for the detection of heavier gas, chlorinated hydrocarbons, etc.). Appendix A contains details of the instrumentation and sampling procedures. Infrared spectrograms were determined on a Perkin-Elmer 621 IR spectrophotometer with a range of 4000 to 400 cm"1 (2. 5 to 25n). Materials to be analyzed were ground with KBr and compressed to form KBr pellets which were scanned over the prescribed spectrum. 19 ------- SECTION 4 SELECTION OF PESTICIDES AND HAZARDOUS WASTE MATERIALS The pesticides and hazardous wastes which were detoxified or decom- posed by microwave plasmas are listed in Table 1, and classified as follows: • Organophosphorus compounds • Halogenated hydrocarbons • Heavy metal compounds The materials were tested in the form of a gas, as pure liquids, sol- vent solutions, aqueous slurries, and solid powder press cakes. Selections were made by EPA and LPARL personnel by mutual agreement on the»basis of diversity in form and chemical constitution, and on the extent of the\ en- vironmental problems which were associated with the materials to be de- toxified. The properties and specifications of these materials are detailed in Appendix B. In the following sections, a summary is presented of the conversion/ detoxification data obtained from the oxygen plasma reactions performed in the laboratory and expanded-scale systems. 20 ------- TABLE 1. PESTICIDES AND HAZARDOUS WASTE MATERIALS FOR DETOXIFICATION TESTS Classification Organophosphorus Pesticide Chlorinated Hydrocarbon Waste Brominated Hydrocarbon Rodenticide Heavy Metal Fungicide Chlorinated Hydrocarbon Pesticide Material Malathion PCB's (Poly chlorinated Biphenyls) Methyl Bromide Phenylmercuric Acetate Kepone Form Tested Pure Liquid Liquid Mixture Commercial Gas Commercial Methanol Solution 1. Commercial Powder 2. Aqueous Dispersion 3. Methanol Solution Manufacturer American Cyanamid Monsanto Matheson Gas Troy Chemical Allied Chemical Grade or Type ULV Aroclor 1242 Aroclor 1254 TroysanPMA-30 80% Powder Concentrate Technical Grade ------- SECTION 5 RESULTS OF LABORATORY-SCALE OXYGEN PLASMA REACTIONS The reactions in the laboratory plasma system, i. e., the system of 1-5 g/hr throughput, have illustrated the efficacy of the microwave plasma process on pesticides and related materials, using oxygen as carrier gas. Also evaluated were argon and hydrogen plasmas to determine their inter- actions, if any, with the toxic test materials. On the basis of the observed formation of extremely offensive mer- captans from Malathion-type compounds in argon plasmas, and the projec- ted formation of dimethylmercury from phenylmercuric acetate or other mercurials in argon and hydrogen plasmas, major efforts were applied to the development of an oxygen carrier gas technology. The reactions asso- ciated with the argon and hydrogen plasmas are given in Appendix C. A summary of typical results from the laboratory unit is listed in Table 2. 5. 1 MALATHION "Cythion" ULV Malathion, manufactured by the American Cyanamid Company, was passed through 200 to 250 Watt oxygen plasmas at 100 to 120 torr, using the quartz basket technique. (See subsection 3. 2 for a des- cription of the technique.) With the exception of a white etch zone and a high viscosity water-white liquid that formed below the plasma, all the products were gases. Mass spectroscopy and gas chromatography indicated CO,, CO, SOo, and water as effluent gases. Infrared spectroscopy showed the liquid product to be phosphoric acid. Materials balances indicated that metaphosphoric acid was the probable material from which conversion to orthophosphoric acid in moist air occurred in 1 to 2 days. Analysis for Malathion in the liquid reaction product was carried out colorimetrically (10). Percent conversion was 99. 98 + percent based on 0. 016 percent Malathion determined. 5. 2 POLYCHLORINATED BIPHENYLS (PCB'S) Monsanto Aroclor 1242 liquid was passed through a 250 Watt oxygen plasma at 100 torr pressure. Mass balance weighings showed no liquids attributable to the starting material. All the products of decomposition were gases. On the basis of control runs in the absence of the plasma 22 ------- TABLE 2. REPRESENTATIVE LABORATORY-SCALE OXYGEN PLASMA REACTIONS Pesticide/ Waste Malathion "Cythion" ULV PCB, Aroclor 1242 Methyl Bromide Phenylmercuric Acetate, "Troysan" PMA-30 Run No. 28-40 28-58 28-94 28-142 Microwave Plasma Power (W) 200-250 250 300-400 225-280 Pressure (torr) 100-120 100 50 120 Conversion 99. 98+ 99. 9+ 99* 99. 9 Reaction Products S09, C09, CO, H90, HPO,orH,,PO, * £i £t & «J iJ TT co2, co, H2o (ci2o, coci2)* CO2, CO, H2O, Br2, BrO2* Hg, C02, CO, H20 CO See text. ------- reaction, percent conversion was calculated at greater than 99. 9 percent. Gas products were identified as CO2, CO, H^O, C12O, and COC12. It will be of interest to note that the latter gases, chlorine oxide and phosgene, were not observed in the expanded-scale plasma reactions; rather, hydrogen chloride was the principal Cl-containing product (see section 6. 2). 5. 3 METHYL BROMIDE Gaseous methyl bromide was passed through 300 to 400 Watt oxygen plasmas at 50 torr pressure at 2 to 3 g/hr. The products of reaction were CO2> CO, H2O, and Br2> Oxides of bromine were found in the liquid nitro- gen traps, but were not otherwise produced at ambient temperatures. Quan- titative analysis of the plasma reactions was performed by mass spectro- meter, in which the ratios of CHgBr response heights before and during the plasma reactions were compared. (See Appendix A, Mass Spectra). Con- ventional analytical methods were precluded because of the toxicity of the methyl bromide gas. Decomposition was greater than 99 percent, which is the limit of precision for the method. 5. 4 PHENYLMERCURIC ACETATE Commercial Troysan PMA-30 solution (30 percent phenylmercuric ace- tate, approximately 70 percent methyl alcohol, plus ammonium acetate) was passed through a 225 to 280 Watt plasma at 120 torr pressure. Metallic mercury was observed as a metallic mirror on the glass tubing downstream from the plasma zone. Material balance indicated 99. 9 percent decompo- sition to mercury metal. Mass spectroscopy showed the products formed in addition to Hg were HoO, CO2, and CO. There was no evidence of dimethyl mercury or other organomercurials. 24 ------- SECTION 6 RESULTS OF EXPANDED-SCALE OXYGEN PLASMA REACTIONS The expanded-scale plasma system is defined as the equipment shown in Figures 8 through 11 which resulted in detoxification yields of 450 to 3200 g (1 to 7 Ib) per hr. Three microwave power applicators and three equivalent-size quartz reactor tubes were used in the evaluations. During the initial runs, the Series A 2. 7-liter applicator was used for oxygen plasma decomposition of PCB, Aroclor No. 1242. It was determined that the liquid had been decom- posed and that one of the reaction products — a black, soot-like deposit which coated the product receiver — contained no PCB, as determined by infrared spectroscopy. Several additional runs, in which feed, pressure, and absorbed power were varied, indicated, however, that the Series A sys- tem was too large in volume for the power available. As a result, the Series B and C applicators, 1. 46 and 0. 65 liters, respectively, were used for all detoxification reactions which are described in the following subsec- tions, and detailed in Table 3. It may be readily concluded, from the re- sults of these evaluations, that the Series C reactor yielded the highest throughput, the highest percentage of conversions, and the lowest costs, as calculated in Section 6. 6. 6. 1 MALATHION "Cythion" Malathion ULV liquid was drop-fed onto a porous, dielectri- cally heated quartz wool bundle positioned at the top applicator input to the plasma zone. By this means, in a similar mechanism to that which was used in the laboratory-scale system, large numbers of smaller droplets or vaporized particles were produced within the matrix of the heated wool, and propelled by the carrier gas through the plasma. Reaction products were gaseous sulfur, 803, H2O, CO, COo, and acetylene, plus a liquid phos- phoric acid. Deposits of a dark yellow-brown sulfur product mixed with a clear, water-white high-viscosity liquid dropped gradually down the sides of the reactor into the receiver. No carbon or any products resembling the starting material were observed. Analysis of the liquids from two reactions gave results of 12 ppm and <1 ppm Malathion, respectively, or conversions of 99. 9988 and 99. 9999 + percent. 6. 2 POLYCHLORINATED BIPHENYLS (PCB'S) Runs 31-8, 31-10, and 31-62 yielded HC1, CO, CO2, and H2O as deter- mined by mass spectrometer. No C12O or COC12 was observed. There 25 ------- TABLE 3. SUMMARY OF EXPANDED-SCALE OXYGEN PLASMA REACTIONS Pesticide/ Waste Malathion "Cythion" ULV Malathion "Cythion" ULV PCB Aroclor 1242 PCB Aroclor 1242 PCB Aroclor 1254 g PMA TroysanPMA-30 PMA TroysanPMA-30 PMA TroysanPMA-30 Kepone 80/20 20% Methanol Solution Kepone 80/20 10% Solids Aqueous Slurry Kepone 2- to 3-g Solid Discs Run No. 31-16 31-46 31-8 31-10 31-62 31-88 31-108 31-110 38-30 38-36 38-38 Reactor Series B B B B B C C C C C C Microwave Power (kW) 3. 7 4.7 4.6 4.2 4.5 4.6 4.0 4.3 4.6 4.2 4.6 Feed Rate g/hr (lb/hr) 504(1. 1) : 480(1. 1) 270(0. 6) Pressure (torr) 28-46 28-30 17-35 i 492(1. 1) 206(0.4) 1020(2. 25) 2380(5. 25) 2950(6. 5) (a) _ - 19-36 13-25 120-140 100-120 100-120 54 40 7 Oxygen Gas Flow (liters /hr) 361 480 323 395 360 960 792 792 720 None 90 Reactor Packing Wool. Plug (a> Wool Plug Wool Wool TDIll rt .Plug Solid Rings (a) Raschig Rings (a) Raschig Rings Raschig Rings Raschig Rings o Raschig Rings Raschig Rings Conversion 99. 9988 99. 9999 99 99 99 Complete estim, 99. 99^ Complete estim. 99. 99 Complete estim. 99. 99 99 99 99 (a) Quartz (b) see text ------- was formation of soot in the product receiver; infrared analysis gave no indication of PCB residues. It was determined, however, that at throughput levels of about 1 kg/hr in the B reactor, complete reactions had not occurred. This was determined by infrared analysis of the black tar-like liquid products in the receiver trap which indicated the presence of PCB starting material. 6. 3 PHENYLMERCURIC ACETATE Troysan PMA-30 was passed through the Series C reactor in several runs to determine the effect of the shortened length of the reactor, and the presence of quartz Raschig rings in the reactor tube. The reaction was con- sidered complete; i. e., no starting material passed through the plasma zone, if methanol peaks were not found by mass spectroscopy in the effluent gas. A series of reactions showed that at a throughput of 3600 g/hr (8 Ib/hr), small amounts of methanol solvent were detected in the gaseous effluent. This indicated that under the conditions of this series, between 7 and 8 Ib/hr can be considered for effectively complete detoxification/destruction of PMA-30. The principal gases of the reaction were CC>2, CO, and H2O. Metallic mercury was deposited in the traps downstream from the plasma. Additional information is given in subsection 7. 2. 4. It is also of importance to note that no organomercurials were detected by the mass spectrometer. 6.4 KEPONE - A "REAL WORLD" MIXTURE A commercial, clay-supported mixture, Allied Chemical 80-percent Kepone concentrate, Code 9406, was used for all tests. Approximately 250 g was available, which was converted as needed into water slurries, methanol, and methanol/acetone solutions, and solid press cakes. The solid press cakes were prepared by compressing 2-3 g batches of Kepone 80/20 in a die under 1000 psi pressure. The discs, which required strong finger pressure to fracture, were placed at the top of the Raschig ring area in the plasma reactor tube before the plasma was ignited. It was observed visual- ly that breakdown of the solids occurred in 10 to 30 seconds, depending on the flow of oxygen and the pressure within the reactor. An extrapolation of a 3-gram batch decomposition in 15 seconds yields 720 g/hr or 1. 6 Ib/hr detoxified. The gaseous reaction products were CO2, CO, HC1, and H2O. Hexachlorobenzene, phosgene, or chlorinated hydrocarbons were not detec- ted using a combination of techniques involving infrared and mass spectros- copy. Because of the limited quantity of the starting material, the reac- tions were not optimized with respect to throughput. Nevertheless, 3 kg/hr (7 Ib/hr) are considered to be well within the capabilities of the system. Safety and handling precautions were observed meticulously for the Kepone experiments. All solution and dispersion operations were carried out in a chemical hood which had a linear velocity of 100 feet per minute at the air inlet. Press cake operations performed in the open laboratory utilized a 4-in. -diameter suction tube which was connected to the chemical hood. The tube acted as a vacuum cleaner-type device to prevent dust from contaminat- ing the laboratory area; and decontamination of residual solids were carried out by dissolving in acetone which was subsequently stored in separate waste bottles. Paper wipes were also stored separately. 27 ------- 6. 5 STEAM PLASMA DETOXIFICATION OF PMA-30 Toward completion of the program, several successful exploratory experiments using steam instead of oxygen were performed in the expanded- scale Series C plasma unit. PMA-30 was evaluated in this reactor. The major products of the discharge were COo, CO, H2O and mercury metal. Very few, if any, signs of lower (Cj to C<£) hydrocarbons were ob- served. There was also no sign of the PMA methanol solvent. Thus, the water plasma decomposition of PMA appeared the same as that which oc- curred in the oxygen plasma. Because of the potential for significant cost savings and the fact that steam, when condensed, will develop its own vacuum, this additional tech- nology is considered to be of significant promise. In addition, when the water vapors condense below the plasma zone, dissolution or entrapment of acid gases can occur. The dissolved materials may then be treated with caustic by standard methods. It is therefore planned to include steam as carrier gas in future microwave plasma investigations at the pilot- and demonstration-levels. 6. 6 PROCESS COSTS CALCULATED FOR PMA-30 For the initial detoxifications in the expanded-scale plasma system, a cents-per-pound figure for electrical power and carrier gas costs was cal- culated. It was considered appropriate that the calculation be made exclu- sive of capital investment for later comparison with pyrolysis, high tempera- ture, and other forms of specialty incineration as data on those processes became available. This type of information was derived for the conversion of phenylmercuric acetate pesticide. It is presented in Table 4. For the determination of process costs for future 50 to 100 Ib/hr demon- stration-scale plasma systems, an accurate assessment is at this time replete with difficulties. Nevertheless, where the object is to show, in principle, the economic viability of the approach, it is possible to calculate values which can be derived from known process variables and estimated carrier gas, electrical, and related costs. The material chosen for the calculation was phenylmercuric acetate (PMA), first, because it was considered refractory or difficult to decompose completely and, second, because it has the significant advantage of recover- ability of its metallic mercury values. The following assumptions are listed below as part of the estimated costs: (1) Electrical costs are $0. 012/kWh, industrial usage. (2) Liquid oxygen costs are $0.005/SCF, large volume usage. Add $3000/yr storage fee. 28 ------- TABLE 4. ELECTRICAL AND CARRIER GAS COSTS FOR PLASMA REACTIONS IN SERIES C REACTOR EXPANDED-SCALE SYSTEM Run No. 31-88 31-108 31-110 Material PMA-30 PMA-30 PMA- 30 Carrier Gas(a> °2 °2 °2 Effective (Absorbed) Microwave Power (kW) 4.6 4.0 4. 3 Feed Rate [g/hr (lb/hr)] 1020(2.25) 2380(5. 25) 2950(6. 5) Carrier Gas Flow [std liter/hr (SCFH)] 960(34) 792(28) 792(28) Gas Cost ($/lb) 0. 18 0.064 0.054 Total Power Consumed (kWh/lb) 4.7 1.9 1.6 Power Cost ($/lb)(b> 0.06 0.023 0.019 Total Cost ($/lb) 0. 24 0.087 0.07 tO CO (a) 02, $0.012/ft3 (b) kWh, $0.012, Industrial Usage. ------- (3) Steam costs are $2. 00/1000 Ib. (4) Labor costs are one-half man per automated unit, i. e., one man operates two units, $12/hr. (5) In the case of PMA, the credit for metallic mercury is $1. 50/lb. This is based on a requirement for future purification. At present, the purity of the recovered metal has not been determined. A more pure product obviously has greater value, up to $4/lb. (6) The estimate for capital costs is $100, 000. It may appear at this time that 50 to 100 Ib/hr are low values when compared with municipal waste processes. However, the detoxification of even a small, 10-pound batch of a highly toxic or dangerous substance can give great concern to those who are responsible for its safe disposal. In the following calculation, a standard 330-day/year, three-shift operation is assumed. PMA-30 throughput is 50 Ib/hr, or 400, 000 Ib/yr. The process costs related to PMA-30 on a per-pound basis are as follows: • Oxygen 4. 3 SCF • Steam 1. 5 Ib • Electrical 1. 6 kWh Calculated total process costs are listed below: Oxygen Steam Variable Costs Operating Labor $ 47, 520 $ 47, 520 Maintenance (4% of Investment) 4, 000 4, 000 Oxygen or Steam 11, 600 1, 200 Electricity 7, 680 7, 680 Total Variable Costs $ 70, 800 $ 60,400 Fixed Costs Taxes and Insurance (2%/yr) $ 2,000 $ 2,000 Capital Recovery (10 yr - 10%) 16, 250 16, 250 Total Fixed Costs $ 18,250 $ 18,250 30 ------- Oxygen Steam Total Annual Costs $ 89, 050 $ 78, 650 Total Income From Recovered Mercury $108,000 $108,000 Total Net Profit $ 18,950 $ 29,350 Net Profit Per Pound Treated $ 0.047 $ 0.073 For detoxification of materials, such as Malathion or polychlorinated biphenyl wastes, where little or no useful by-products are recoverable, the costs may be higher. In oxygen and steam plasmas, for example, the costs per pound are calculated as $0. 22 and $0. 20, respectively. It should be noted that as microwave power technology advances, the electrical costs should decrease as the result of improvements in the coup- ling of microwave power to the plasma reactors. The potential for steam plasmas with their low cost, plus their capability to form their own vacuum, is readily apparent, and can therefore be considered for further reductions in electrical power requirements. The above data have been derived for a permanent-type disposal center, which may then be compared with other methods operating at similar levels of throughput. The costs associated with projected mobile, or back-of-truck type, detoxification units will vary, depending on the initial capital costs, which should be lower, and labor costs, which may be higher per unit. The latter will be dependent in part on the distance to the user's site as well as the toxicity of the materials to be treated. 31 ------- SECTION 7 DISCUSSION OF RESULTS An overview of the chemistry of microwave plasma detoxification and process variables as determined during the first year of the program is presented in the following subsections. 7. 1 INTERPRETATION • Microwave detoxification, using oxygen plasmas, has been success- ful and practical for all pesticides and hazardous wastes tested to date. These include Malathion, polychlorinated biphenyls, methyl- bromide gas, phenylmercuric acetate, and Kepone. • Optimization of plasma conditions was performed for the refractory organomercurial, phenylmercuric acetate. As a result of varying the several parameters affecting the detoxification efficiency, decom- position was 3. 2 kg/hr (7 Ib/hr), yielding principally gaseous CO2, CO, H^O, and metallic mercury. The mercury has a known cost recovery value, thereby resulting in low or nonexistent operating costs for that particular material. • Oxygen plasma reactions were uniform in their production of gaseous oxides of carbon, water, and acidic gases. • Pure liquids, solutions, aqueous slurries, and powders as press cakes were decomposed by oxygen plasmas. The only differences in response were associated with the methods of addition used for the respective materials. The chemical reactions were effectively iden- tical. • Further process cost reductions are potentially feasible through the use of steam plasmas, which, in addition to low costs relative to oxygen, can also develop their own reduced pressures in the plasma reactor. 32 ------- 7. 2 CHEMISTRY OF OXYGEN PLASMA REACTIONS 7. 2. 1 Malathion — Oxygen Plasma Reactions Complete oxidation of Malathion in the laboratory- scale system yielded colorless gases and a water-white liquid: SO2, CO2, CO, HoO, and phos- phoric acid products. Several equations may be written for the reactions. For simplicity, only CO2 is included, C10H19°6PS2 + 15O2~^2SO2 + 10CO2 + 8H2° + H3PO4 (°rth°PhosPhoric acid) C10H19°6PS2 + 15O2-*2SO2 + 10CO2 + 9H2° + HPO3 (metaPhosPhoric acid) 2C10H19°6PS2 + 30O2~~4SO2 + 20CO2 + 17H2° + H4P2°7 (pyr°PhosPhoric acid> 2C10HigOgPS2 + 30O2— 4SO2 + 20CO2 + 19H2O + P2°5 (phosphoric anhydride) Starting with 2. 7 g Malathion added, the stoichiometry indicates the following: H3P04 0. 80 g H4P20? 0. 73 g HPO3 0. 65 g P205 0. 58 g The experimentally determined value, twice run, was 0. 66 g. Although pyrophosphoric acid has been detected in air- oxidation incineration reac- tions of phosphorus -containing materials, metaphosphoric acid appears to have been produced in this system. The difference in phosphoric acid iden- tities may be related to the differences in the effective temperatures of the two reactions, and, consequently, their rates of formation in air (incinera- tion) versus pure oxygen (plasma process). On further exposure of the phosphoric acid liquid to laboratory air for several days, orthophosphoric acid was determined by infrared spectres copy to be the final product. In the expanded- scale Series B system, some sulfur and small amounts of acetylene were observed in addition to CO2, CO, H2O, and SO2. This is attributed to the incomplete oxidation of the starting material, since the sys- tem had not yet been optimized. It would not be considered difficult now to obviate acetylene or sulfur under currently optimized conditions in the Series C reactor. 7. 2. 2 PCB - Oxygen Plasma Reactions Reactions in the laboratory- scale system yielded CO2, CO, I^O, C^O, and COC12, whereas in the expanded- scale system, Series B, no C^O or COC12 was observed. Instead, HC1 was the Cl-bearing product detected, plus a soot-like deposit in the product receiver. It is probable that the 33 ------- following type of reaction occurred in the reactor configuration used at the time: C12H?C13 + 11/2O2 — 2H2O + 3CO2 + SCO + 6C(soot) + 3HC1 Optimization of the reaction in the Series C reactor should result in the dis- appearance of soot, and an additional formation of CO2, etc. 7. 2. 3 Methyl Bromide — Oxygen Plasma Reaction In the laboratory-scale system, oxidation yielded CC>2, CO, H2O, and Br2 as principal products, with some BrO2 and/or Br2O formed inside the liquid nitrogen traps. The principal reaction may be written as follows: 4CHgBr + 7O2 — 2Brg + 4CO2 + SHgO Side reactions involving Br2 and H2O may have produced HBr and HOBr. 7. 2. 4 Phenylmercuric Acetate — Oxygen Plasma Reactions In both the lab oratory-scale and expanded-scale plasma systems, the major products were CO2, CO, H2O, and mercury metal. Minor amounts of lower molecular weight hydrocarbons were observed in the expanded- scale system. The formation of metallic mercury may be explained by postulating HgO as an intermediate in the plasma reaction, followed by its decomposition. The dissociation pressure of HgO to give Hg metal and O2 is a few torr at 440°C (11). Thus, Hg metal can be produced as follows: 1) C8H802Hg + 19/202 -HgO 400-500°C 2) HgO " Hg+ 1/202 Addition of 1) and 2) yields: C8H802Hg + 902 - Hg + 8C02 Oxidation of methanol is considered to follow the usual mechanism to CO2 and H2O, depending on the quantity of oxygen available. 7. 2. 5 Kepone — Oxygen Plasma Reactions The plasma reaction products of 20 percent Kepone-methanol and Kepone-methanol/acetone solutions in the expanded-scale system were COo, CO, H2O, and HC1. Hexachlorobenzene, phosgene, or chlorinated hydro- carbons were not detected. Reactions in an oxygen plasma may therefore be written: C10C110° + 7°2 ~* 5C02 + 5CO + 5C12 34 ------- 2CH-OH + O0 — CO + CO0 + H0O (2) •j & & & Reaction of Cl2 with the H2O from the methanol reaction produces HC1 and HC1O; since the latter is unstable, it decomposes to HC1 and oxygen. It is of interest to note that the time periods for decomposition of small 3 g, 80/20 Kepone press cakes within the plasma zone were 10 to 30 sec; the actual values depended on the surface area of the solids, the flow of oxy- gen, and the pressure within the reactor. Percentage conversions are listed at >99 percent. The values are based on analysis of the effluent products. Absence of Kepone or hexachloro- benzene in the receiver, or any low molecular weight chlorinated hydrocar- bons in the gas effluent denotes a greater than 99 percent conversion. Infra- red spectrograms of Kepone extracted with acetone and its oxygen plasma reaction product are shown in Figure 13. 7. 3 EVALUATION OF ANCILLARY EQUIPMENT Four specific unit processes were developed and tested for use in the expanded-scale plasma detoxification system, as follows: • Materials feed techniques • Packed bed technology • Reduced pressure systems • Analytical processes 7. 3. 1 Materials Feed Techniques Several techniques were evaluated to optimize as far as practicable the desired materials introduction processes. These included one- and two- fluid spray nozzles, in which the feed stock was introduced upwards from bottom to top in order to maximize residence time and to minimize gravity effects. No fine-spray system was found completely satisfactory, since little or no liquid spray flare-out can be tolerated below the plasma zone in the waveguide cutoff. If Hare-out is permitted, thermal decomposition may occur in the cutoff, and the chemical identity of the products as well as the material balance becomes difficult or impossible to evaluate. Custom- fabricated two-fluid nozzles are now being designed to avoid these problems, which .can become more significant when larger quantities of materials are detoxified. The most successful technique was a vacuum-equalized liquid-feed sys- tem (Figure 12) into which several Teflon needle valves were constructed for improved drop-size control. This technique was used for gravity feed- ing PCBs and PMA-30 solutions. For feeding Kepone solutions and water slurries, the use of a volumetric buret was adequate; the feed was kept at 35 ------- 25 WAVELENGTH (M) 6 7 8 9 10 12 15 20 30 4050 100 60 ' .- u 40 20 3 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 WAVENUMBER (CM~') 200 Figure 13. Infrared spectra of Kepone acetone extract (curve A) and its oxygen plasma reaction product (curve B) ------- atmospheric pressure to avoid continual depletion of solvent during the addi- tion. Only the buret tip, inserted into the reactor above the plasma zone, was maintained at reduced pressure. Feed techniques for solids are also under investigation. These include a two-tier reduced pressure forechamber in which the solids are placed. The solids will be fed into a kiln-type, angularly-positioned reactor which is rotated by means of a suitable, motorized transport. 7. 3. 2 Packed Bed Technology Quartz Raschig rings were used in the reactor tube in order to increase the residence time of the feed materials in the plasma zone. Several varia- tions in ring size and bed length were evaluated to determine maximum throughput under different packed bed conditions. Maximum throughput was defined as that quantity passing through the reactor which showed no metha- nol solvent in the mass spectrometer scans. Data are listed in Table 5. Without rings, decreased detoxification levels were observed because of ex- cessively short, 0. 1 sec, residence times. On the basis of these runs, the application of packed bed technology can be considered for use where aqueous solution, pure liquids, and solvent sys- tems are to be treated. Ring materials may be quartz or high-alumina ce- ramics which are not "lossy", i. e. , do not show high dielectric loss factors, and thus do not "heat up" at 2450 MHz, which is the microwave frequency used in the process. In the batch detoxification of Kepone solids, the packed bed approach was used successfully. For the continuous feed of solids and powders into beds of the above geometry, further investigation will be re- quired for development of the technique. 7. 3. 3 Reduced Pressure Systems Sargent-Welch 1397B Duo Seal mechanical vacuum pumps (approximate- ly 400 liters/min pump capacity) were used for the development of the re- duced pressure required for plasma formation. Liquid nitrogen, ice, or mixed ice-acetone traps were used, depending on the materials being stud- ied, to trap out materials which would otherwise react with the oil or metal parts. This design has been adequate for the initial Phase I scale-up opera- tion. However, it is likely that at the next level of scale-up more practical pump systems will be required, such as liquid ring seal or steam ejector pumps. These will obviate in large part the need for traps, which require filling, cleaning, etc. , and can serve as a transfer medium for moving the acid gas products to a caustic scrubber for later disposal into municipal waste streams. 7. 3. 4 Analytical Processes The methods used in this study have been described above. (See Sec- tion 3. 3 and Appendix A. ) The principal objective has been to deter- mine the levels of detoxification/decomposition achieved in the expanded- scale plasma system, thereby evaluating the potential usefulness of 37 ------- TABLE 5. INFLUENCE OF PACKED BED ON PMA-30(PHENYLMERCURIC ACETATE) CONVERSION IN SERIES C PLASMA SYSTEM Run No. 31-88 31-108 31-110 38-6 38-8 , 38-14 Packing of 45-mm Reactor Ring Size o. d. x Length (mm) 8x8 8x8 8x8 10 x 10 8x4 8x8 Bed Length (cm) 16 45 45 34 31 31 Gas Flow (Standard liter/ min] 8 to 16 13. 2 1 3 , 2 13.2 13.2 to 18.4 13.2 to 16.0 Pressure (torr) Top 120 120 120 115 112 130 Bottom - 64 60 42 60 75 Absorbed Power (kW) 4.7 4.0 4. 3 4. 3 4.6 4. 7 Throughput (lb/hr) 2.25 5. 25 6. 5 4.25 6.0 8.0 5.4 Notes First high throughput run; reactor filled approximately one- half with rings Reactor completely filled with rings Repeat of 31-108 Largest rings of series Rings fused near bot- tom port of power input CH~3OH solvent ob- served at 8 lb/hr; disappeared when rate decreased to 5. 4 lb/hr oo ------- the process. The methods adequately served this purpose. It will be neces- sary in further process developments, however, to refine the analytical techniques whereby in situ determination of ppm reaction products can be obtained. This will be required for analysis of gaseous as well as liquid (water, steam) effluents, before these streams can be disposed of into muni- cipal sewage streams or directly to the air environment. In line with these requirements, high-sensitivity gas chromatograph units with appropriate detectors are required for in-line determination of impurities, if such exist. Improvements in mass spectroscopy, and, later, installation of a gas chromatograph-mass spectrometer plus computer scan and readout of pesticides and toxic materials, can be a significant aid in assessing the completeness of the oxygen/steam plasma reactions, and thereby its usefulness as a detoxification method. 39 ------- SECTION 8 CONCLUSIONS 1. The process of microwave plasma decomposition of organic mate- rials has been applied successfully to the detoxification/destruction of hazardous organic wastes. A primary objective of the program - evalua- tion of the effectiveness of an expanded-scale plasma system - has been completed using hazardous organic compounds and wastes of current inter- est. The reaction products were identified in order to estimate their toxi- city and to determine the potential for recycling/recovery of useful by- product materials. 2. This program has resulted in an expansion of the state of the art in microwave plasma detoxification from 1 to 5 g to 3 kg (7 Ib) per hour. One hundred fifteen (115) runs were performed in the series. This scale-up has changed the perspective from one of academic interest to that of a real- istic or practical level. 3. The commercial pesticides, Kepone, Malathion, phenylmercurie acetate, and methyl bromide; and the hazardous waste, polychlorinated bi- phenyl, can be detoxified beyond 99 percent in an oxygen plasma system. Malathion in pure liquid form was decomposed to > 99. 999 percent and liquid PCB to at least 99. 9 percent. 4. The potential for resource recovery has been clearly demonstrated for PMA-30, phenylmercuric acetate solution. The commercial form of this organomercurial was decomposed efficiently to yield metallic mercury, which is a salable item, plus oxides of carbon and water. The process may permit the additional recovery of valuable chemical feedstocks when applied to other organometallic pesticides or wastes which would otherwise be per- manently lost. 5. The process itself is eminently safe. Microwave leakage is not detected near the applicator or the power supplies. Because reduced pres- sures (low vacuum) are used, the equipment, by its nature, does not permit materials loss through buildup of excess pressures and the consequent pres- sure blowoffs. Further, dangerous leaks will not occur, because the plasma unit would first shut itself down, and, simultaneously, terminate by means of interlocks, the introduction of feed materials. Process controls have not been complex in the initial scale-up development, but high sensitivity toxi- city monitors at the ppm level will be required as the size of the equipment increases. 40 ------- 6. When compared with refractory incinerators, the microwave plasma systems have several advantages, such as potential for portability, or back- of-truck transport to a toxic materials' site, relatively low initial cost, leak- proof character, and, thus, the capacity for decomposing materials where no other techniques are satisfactory. For example, in an ordinary inciner- ator, the decomposition of phenylmercuric acetate, or any organomercurial waste, would result, unless all openings were totally closed, in an uncon- trolled release into the environment of mercury metal or oxide particles. In the plasma system, the mercury is separated from the effluent gas stream and collected. 7. Since the plasma decomposition mechanism utilizes electronic ener- gy, in which the applicator can be maintained barely hot to the touch, rather than thermal energy, materials-of-construction requirements are signifi- cantly less stringent. This results in lower materials costs, which reduces maintenance and repair expenses. 41 ------- SECTION 9 RECOMMENDATIONS This study, Phase I, Development of Microwave Plasma Detoxification Process, unquestionably points toward a phased, logical scale-up to a 20 Ib/hr unit, and then to a 100 Ib/hr version. A program, as outlined be- low, is recommended for the design, purchase, and construction of pilot- scale equipment and components. For development of the system, denoted as Phase II, specific engineering tasks are suggested. • Materials Feed Operations. Techniques need to be developed and evaluated for the feeding of materials to be detoxified into the plasma reactor. These materials may be gases, liquids, solids, or slurries. The problem should be defined in detail, and the necessary equipment designed, installed, and tested in accordance with data obtained from the program recently completed. • Microwave Power. Additional microwave power, above 5 kW at 2450, is suggested for the pilot process in order to expand capacity to the desired 20 Ib/hr or more throughput. Power-generating components and ancillary equipment are purchasable off the shelf from, microwave hardware suppliers. • Microwave Plasma Applicator/Reactor. A new high-power appli- cator is needed for transfer of the increased microwave energy to the plasma reactor. The applicator is available as a custom- fabricated unit. The reactor materials suggested for use are catalog items. • Product Receptor. For plasma reactions which involve in parti- cular the formation of metals, cooling traps/heat exchangers should be designed and evaluated for product separation and con- densation. • Vacuum System. High-capacity liquid ring pumps should be evalu- ated for obtaining reduced pressures, based, in part, on their capability for acting as a medium for dissolving many of the acid products emitted from the reactor. • Analytical Instrumentation. Process effluents should be analyzed for traces of potentially toxic impurities, if any of these materials are formed. Gas chromatographic units, infrared and ultraviolet spectrophotometers, and a quadrupole residual gas analyzer 42 ------- (mass spectrometer) should be installed in-line for obtaining the required data. It is also recommended that side-support studies be continued at the basic scientific level for evaluation of variables, such as microwave fre- quency and applicator/reactor technology to determine their potential ef- fect on maximum size and throughput. In conclusion, continued exploration of detoxification of "new" hazard- ous wastes should be maintained. In this regard, more than a hundred requests for information and data have been received during the last several months. These involved toxic waste and waste-related problems not pre- viously considered for this application, thus indicating clearly the wide interest this new technology has engendered. 43 ------- REFERENCES 1. McTaggart, F. K., Plasma Chemistry in Electrical Discharges, Else- vier, Amsterdam, 1967, p. 2. 2. Lockheed Missiles & Space Company, Inc., Dissociation of Toxic Gas Analogs in an Electrical Discharge, Lockheed Independent Research Program, Apr 1969. 3. Sibert, M. E., Vapor Decomposition by Microwave Discharge, Con- tract DAAA 15-70-C-04880, Final Report, LMSC-D243400, Sep 1971. 4. Bailin, L. J. , M. E. Sibert, L. A. Jonas, and A. T. Bell, "Micro- wave Detoxification of Toxic Vapor Simulants, " Environmental Science and Technology, 9 (3), 254-58, 1975. 5. Bailin, L. J., "Cold Plasma Detoxification of Toxic Wastes, " Water Quality Treatment, Final Report LMSC-D356581, Lockheed Independent Development Program, Dec 1973. 6. MacDonald, A. D. , Microwave Breakdown in Gases, John Wiley, New York, 1966, p. 71. 7. Fehsenfeld, F. C., K. M. Evenson, and H. P. Broida, "Microwave Discharge Cavities Operating at 2450 MHz, " The Review of Scientific Instruments, 36_(3), 294-8, 1965. 8. Bell, Alex. T., in "Techniques and Application of Plasma Chemistry, " John R. Hollahan and Alexis T. Bell, Ed., Chapter 10, Engineering and Economic Aspects of Plasma Chemistry, John Wiley, New York 1975, pp. 379-381. 9. Brown, Lloyd C., Alex. T. Bell, "Kinetics of the Oxidation of Carbon Monoxide in a Radiofrequency Electrical Discharge, " Industrial Engin- eering Chemistry, Fundamentals, 13, 203, 1974. 10. Norris, M. V., W. A. Vail and P. R. Averell, "Colorimetric Estima- tion of Malathion Residues, "Agricultural and Food Chemistry, 2 (11), 570-73, 1954. 11. Remy, H., Treatise of Inorganic Chemistry, Vol. II, Elsevier, Amsterdam, 1956, p. 464. 12. Brooks, G. T., Chlorinated Insecticides, Vol. 1, CRC Press, Cleve- land, 1974, pp. 97-104. 13. Safe, S. and O. Hutzinger, Mass Spectroscopy of Pesticides and Pollu- tants, CRC Press, Cleveland, 1973, pp. 28, 195. 44 ------- Appendix A MASS SPECTRA A. 1 EQUIPMENT A quadrupole mass spectrometer Varian Model 974-0002 residual gas analyzer was used to analyze the effluent gases from the reactor, as fol- lows. A small quantity of gas leaving the plasma reactor is pumped past a variable-leak sampling valve which is located approximately 6 meters from the plasma exhaust. Figure A-l is a schematic of the sampling valve inter- face with the plasma reactor and the residual gas analyzer. The sampling system consists of a quadrupole gas analyzer tube, a high-vacuum system for the analyzer tube, and a gas admittance system. A fraction of the plasma effluent gas is bled into the mass spectrometer, and is continually pumped through the sampling system by a second vacuum pump. Some of the gas is ionized by an electron beam ionizer in the analyzer section of the mass spectrometer. These gases not only form parent ions of the same mass as the neutral molecules, but also dissociate to form smaller ions. The ions are separated by mass by the quadrupole mass filter, and detected by an electron multiplier. The electron multiplier generates an electrical signal that is amplified and recorded. The filter is adjusted so that it passes ions of only one mass at a time. By controlling the filter, the distribution of ions is determined, yielding the mass spectrum of the gases present in the ionizer, up to a maximum of 250 mass units. Figures A-2 and A-3 are photographs of the residual gas analyzer hardware and electronics. It should be noted, in a mass spectrometer system that is pumped to 10-10 atm, that there remain approximately 109 molecules/cm^. A sensi- tive mass spectrometer will register these background gases as background ions. Vacuum pumps, particularly ion-getter or "Vac Ion" types, when used to maintain these low pressures, do not pump all gases with equal efficien- cy. Certain gases, e. g., argon, helium, etc., are pumped very slowly out of the mass spectrometer. This is the principal reason for the pres- ence of one of the main residual background gases, argon, which produces the Ar+ ion, mass/charge M/E = 40, and the Ar++ ion, M/E = 20. It should also be noted that, because of the dissimilar mass spectro- meter response characteristics of the different gases within a mixture, quantitative comparisons using their relative peak heights are not reliable, and should not therefore be attempted. 45 ------- ABSOLUTE MANOMETER SAMPLING LINE FROM PLASMA REACTOR SAMPLING VALVE ULTRA-HIGH VACUUM SYSTEM LEAK VALVE 'ROUGHING VALVE LIQUID NITROGEN TRAP ANALYZER TUBE VAC ION PUMP ISOLATION VALVE MECHANICAL PUMP Figure A-l. Gas sampling system interface with plasma reactor and residual gas analyzer 46 ------- -- VAC ION PUMP (TITANIUM GETTER) THERMOCOUPLE VACUUM GAGE X-Y RECORDER I GAS INLET VALVE FROM PLASMA SYSTEM QUADRUPOLE HEAD TOP OF OIL DIFFUSION PUMP •GAS SAMPLING VARIABLE LEAK VALVE Figure A-2. Mass spectrometer residual gas analyzer -- variable leak valve and quadrupole head ------- OSCILLOSCOPE RESIDUAL GAS ANALYZER CONTROL UNIT =1 \f: ^-DIGITAL VOLTMETER ION PUMP CONTROL UNIT ELECTRON MULTIPLER POWER SUPPLY IONIZATION GAGE FOR DIFFUSION PUMP VACUUM .A. POWER SUPPLY Figure A-3. Mass spectrometer residual gas analyzer power supplies and vacuum gages 48 ------- A. 2 PESTICIDE AND HAZARDOUS WASTE SPECTRA In the following spectra (Figures A-4 through A-10) are shown the masses of ions derived from the products of the respective oxygen and steam plasma reactions. These include scans from Malathion, PCB's, methyl bromide, phenylmercuric acetate, and Kepone. A brief commen- tary for each spectrum is included. A. 2. 1 Malathion — Oxygen Plasma (Figure A-4) The main gaseous products were carbon dioxide (M/E = 44, 28, 22, 16, 14, 12), carbon monoxide (M/E = 28, 16, 14, 12), water (M/E = 18, 17,16), sulfur dioxide (M/E - 64, 48, 32), and, possibly, a small amount of ace- tylene (M/E = 26, 25). Most of the ions appearing in the oxygen plasma back- ground were those formed from the residual background gases, such as argon (M/E = 40, 20), that are always present in the mass spectrometer. A. 2. 2 Aroclor No. 1242 PCB - Oxygen Plasma (Figure A-5) During the run, most of the oxygen was consumed. The main gases observed were carbon dioxide (M/E = 44, 28, 22, 16, 14, 12), carbon mon- oxide (M/E = 28, 16, 14, 12), and water (M/E = 18, 17, 16). Chlorine, Cl2, (M/E = 70, 72, 74) was not observed, but gaseous hydrochloric acid (M/E~= 35, 36, 37, 38) was detected. Most of the ions observed in the oxygen plasma background were those formed from residual background gases, such as argon (M/E = 40, 20), which are always present in the mass spectrometer. A. 2. 3 Aroclor No. 1254 PCB - Oxygen Plasma (Figure A-6) The main gases observed were carbon dioxide (M/E = 44, 28, 22, 16, 14, 12), carbon monoxide (M/E = 28, 16, 14, 12), and hydrochloric acid (M/E = 35, 36, 37, 38). Chlorine, Cl2, (M/E = 70, 72, 74), was not ob- served. No appreciable amounts of water (M/E =18, 17, 1 6) were observed; evidently, the hydrogen present in the Ci 2^05 molecules reacted to form hydrochloric acid HC1 preferentially. There appeared to be excess oxygen (M/E = 32) present for these reactions. Most of the ions formed in the oxy- gen plasma background were those formed from residual background gases, such as argon (M/E = 40, 20), that are always present in the mass spectro- meter. A. 2.4 Methyl Bromide — Oxygen Plasma, Laboratory-Scale Reactor, Quantitative Analysis (Figure A-7) ~~ Quantitative analysis of the plasma compositions was performed by mass spectroscopy. In this method, the ratio of CHgBr response (peak) heights were compared before and during the plasma reactions. Conven- tional analytical methods were precluded because of the toxicity of the methyl bromide gas. Figure A-7 illustrates the method. Decomposition was calculated as greater than 99 percent, which was the limit of sensiti- vity for the mass spectrometer unit. 49 ------- M/E' OXYGEN PLASMA BACKGROUND CO, i I i i i i I , , I i I I i i i I I i i . I 10 15 20 25 30 35 40 45 50 55 60 65 70 M/E' IN-LINE ANALYSIS MALATHION CO- SOT SO, i i I I i i i I i i i i I i i i i I i i i i I . 10 15 20 25 30 35 40 45 50 55 60 65 I I I i i I i i I 70 75 •RATIO OF MASS TO CHARGE; VALUES APPLY TO IONS FORMED FROM NEUTRAL GAS MOLECULES PASSING THROUGH MASS SPECTROMETER. Figure A-4. Mass spectra of gaseous effluent from reaction of Malathion in oxygen plasma (Run 31-16) 50 ------- OXYGEN PLASMA BACKGROUND CO! I I i i I I I i I i i I i I i >-».5 10 15 20 25 30 35 40 45 50 55 70 75 COT IN-LINE ANALYSIS AROCLOR NO. 1242 M/E*-*-5 10" 15 20 25 30 35 40 45 50 55 60 65 70 75 •RATIO OF MASS TO CHARGE; VALUES APPLY TO IONS FORMED FROM NEUTRAL GAS MOLECULES PASSING THROUGH MASS SPECTROMETER. Figure A-5. Mass spectra of gaseous effluent from reaction of Aroclor No. 1242 PCB in oxygen plasma (Run 31-10) 51 ------- CO, OXYGEN PLASMA BACKGROUND M/E-*- 5 20 25 30 35 50 55 60 65 I I _L CO I I IN-LINE ANALYSIS AROCLOR NO. 1254 COl Jj _h M/E' »• 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 •RATIO OF MASS TO CHARGE; VALUES APPLY TO IONS FORMED FROM NEUTRAL GAS MOLECULES PASSING THROUGH MASS SPECTROMETER. Figure A-6. Mass spectra of gaseous effluent from reaction of Aroclor No. 1254 PCB in oxygen plasma (Run 31-62) 52 ------- 100 90 80 70 1/1 50 -J 40 < Z o «/» 30 20 10 0 METHYL BROMIDE PEAK -CURVE 1, PLASMA OFF -CURVE 2, PLASMA ON 80 85 90 95 100 105 CH Br MASS SPECTROMETER IONS o Fraction of CI-LBr remaining when plasma on: Peak height plasma; On _ <0.5 < Peak height plasma: Off " 92 " Therefore/ CH«Br decomposed >99% in plasma reaction Figure A-7. Mass spectroscopic analysis of methyl bromide oxygen plasma reaction (Run 28-94) 53 ------- A. 2. 5 Troysan PMA-30, Phenylmercuric Acetate — Oxygen Plasma (Figure A^ In-line analysis indicated that almost all the oxygen carrier gas (M/E = 32) had reacted. The main gaseous products were carbon dioxide (M/E = 44, 28, 22, 16, 14, 12), carbon monoxide (M/E = 28, 16, 14, 12), and water (M/E = 18, 17, 16) with minor amounts of low-molecular weight hydrocarbons. Most of the ions observed in the background plasma were those which are formed from residual background gases, such as argon (M/E = 40, 20), which are always present in the mass spectrometer. A. 2. 6 Troysan PMA-30 - Steam Plasma (Figure A-9) The main gases observed were carbon dioxide (M/E = 44, 28, 22, 16, 14, 12), carbon monoxide (M/E = 28, 16, 14, 12), water (M/E - 18, 17, 16), and a small amount of benzene (M/E = 78, 77, 50, 51, 52). There was no sign of the methyl alcohol solvent, which has a unique ion at M/E = 31. No sign of the possible product, dimethylmercury, was observed in which the high-mass range (not shown) was searched at high sensitivity. Most of the ions observed in the plasma background were those formed from residual, background gases in the mass spectrometer, such as argon (M/E = 40, 20), or small amounts of aromatic ions (M/E = 78, 77, 76, 50, 51, 52). A. 2. 7 Kepone — Oxygen Plasma (Figure A-10) The main gaseous products observed were carbon dioxide (M/E = 44, 28, 22, 16, 14, 12), carbon monoxide (M/E = 28, 16, 14, 12), and water (M/E = 18, 17, 16). Further scanning in the high-mass spectrometer range, 1 to 250, using a high-sensitivity setting, showed no sign of phosgene or chlorinated hydrocarbons. Hydrogen chloride, extremely water-soluble, was found as an ice mass deposited in the liquid nitrogen trap. 54 ------- OXYGEN PLASMA BACKGROUND 100 x GREATER SENSITIVITY I I , I I I , I I I I , I I I I , M/E*-»5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 OH -| IN-LINE ANALYSIS TROYSAN PMA-30 M/E* 10 15 20 25 30 35 40 45 50 55 60 65 70 75 •RATIO OF MASS TO CHARGE; VALUES APPLY TO IONS FORMED FROM NEUTRAL GAS MOLECULES PASSING THROUGH MASS SPECTROMETER. Figure A-8. Mass spectra of gaseous effluent from reaction of Troysan PMA-30 (phenylmercuric acetate methanol solution) in oxygen plasma (Run 31-118) 55 ------- OH OXYGEN PLASMA BACKGROUND 100 x GREATER SENSITIVITY O2 ISOTOPE *-^5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 OH CO I i i IN-LINE ANALYSIS TROYSAN PMA-30 AND STEAM 100 x GREATER SENSITIVITY I < ' ' » I < ' ' ' t ' ' ' ' I M/E'*-5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 •RATIO OF MASS TO CHARGE; VALUES APPLY TO IONS FORMED FROM NEUTRAL GAS MOLECULES PASSING THROUGH MASS SPECTROMETER. Figure A-9. Mass spectra of gaseous effluent from reaction of Troysan PMA-30 (phenylmercurie acetate methanol solution) in steam plasma (Run 31-190) 56 ------- OXYGEN PLASMA BACKGROUND Ar A CO, L LJ_J I I I I I I I I I 01 -o M/E*—»• 5 10 15 20 IN-LINE ANALYSIS KEPONE SOLUTION 25 30 35 H, H i I I I . I I I I • i i i I I I I I I I i I I M/E* 10 15 20 25 30 35 40 45 50 •RATIO OF MASS DIVIDED BY CHARGE VALUES APPLY TO IONS FORMED FROM NEUTRAL GAS MOLECULES PASSING THROUGH THE MASS SPECTROMETER. Figure A-10. Mass spectra of gaseous effluent from reaction of Kepone-methanol/ acetone in oxygen plasma ------- Appendix B PROPERTIES OF PESTICIDES AND HAZARDOUS WASTES Chemical and physical properties and specifications of the materials studied in the program are listed in this section. B. 1 MALATHION American Cyanamid "Cythion" ULV (ultra low volume) grade was ob- tained from the manufacturer in commercial form, for which specifications are given as follows: Chemical name: Structural formula: Purity: Odor: Vapor pressure: Boiling point: Melting point: Solubility: Flash point, Tag Open Cup: Viscosity: Toxicity, , acute, oral: O, O- dimethyl phosphorodithioate of diethyl mercaptosuccinate S CH2C02C2H5 CH3O-P-S-CHCO2C2H5 OCH3 95%, minimum Slight, characteristic 4 x 10"5 torr at 30°C 156-57°C, 0. 7 torr, slight decomposition 2. 9°C 140 ppm in H2O, 25°C. Soluble in most alcohols, esters, aromatic solvents, and ketones. Poor solubility in aliphatic hydrocarbons Greater than 163°C (325°F) 40°C, 17 centipoises 25°C, 37 centipoises 1, 375 mg/kg male albino rats, anticholin- esterase behavior 58 ------- B. 2 POLYCHLORINATED BIPHENYLS Two different grades were evaluated. These were Aroclor Nos. 1242 and 1254, obtained from Monsanto, St. Louis, Mo. Specifications and properties are listed below: Aroclor No. 1242 Average molecular com- position: Chlorine (%) Appearance: Distillation range (°C): Flash point (°C), Cleveland Open Cup: Fire point (°C), Cleveland Open Cup: Evaporation loss (%) 163°C (325°F), 5 hr: B. 3 METHYL BROMIDE GAS C12H7C13 (98% between C12H8C12 and C12H5C15) 42 Aroclor No. 1254 CioHsClc (98% between Cl2H6Cl4 and C12H3C17) 54 Light-yellow viscous liquid 365-90 Clear, mobile oil 325-366 176-180 None to boiling point None to boiling point None to boiling point 3.0-3. 6 1.1-1.3 Methyl bromide was obtained as a liquefied gas from Matheson Gas Products Company. Its specifications are as follows: Chemical name: Formula: Purity: Appearance: Odor: Methyl bromide, or bromomethane CH3Br 99. 5% CH3Br (min. ) 0.015% H2O (max.) Colorless None, except at high concentrations where it has a chloroform-like odor Vapor pressure, 21°C&0°C): 13 psig (0. 9 kg/cm2 gage) Boiling point, 1 atm: 3. 5°C (38. 2°F) 59 ------- Freezing point, 1 atm: -93. 7°C (-136. 7°F) Flammability: Limited to narrow range of 13. 5-14. 5% in air Autoignition temperature: 538°C(998°F) Toxicity: Highly toxic; 20-ppm skin exposure, averaged, 8-hr day OS HA limits B.4 PHENYLMERCURIC ACETATE, PMA Commercial quantities of PMA solutions were obtained from Troy Chemical Company, and PMA 100 percent solids from Aldrich Chemical Company and American Drug and Chemical Company. Troysan PMA-30 was the principal material used for the detoxification studies. Specifica- tions and properties are listed as follows: Chemical name: Phenylmercurie acetate Structural formula: Melting point: 149°C PMA solids content in Troysan PMA-30: 30%, min. Metal content in PMA-30: 18%Hg Other components in PMA-30: Principally methyl alcohol, ammonium acetate Dilution in water PMA-30: Complete Toxicity, LD5Q, acute, oral: 22 mg mercury/kg, rat test animals B. 5 KEPONE Allied Chemical 80 percent Kepone concentrate, Code 9406, was used as the starting material. Approximately 0. 5 Ib was available for the tests. Properties are as follows: Chemical name: Decachloro-octahydro-1, 3, 4-metheno-2H- cyclobuta (c, d) pentalene-2-one, or chlor- decone. (Ref. 12) 60 ------- Structural formula: Cl Contents: Melting point: Solubility Toxicity, 80% Kepone, 20% inert clay-type support Technical product, more than 90% pure; sublimes at near 350°C, with some decom- position Soluble in acetone, methanol (small amount exclusive of clay-support not soluble in CE^OH). Isolatable as mono- or tri-hydrate; forms solvates with acids, alcohols, amines, and thiols, and soluble in strongly alkaline solutions. Recrystallizable from 90% aque- ous ethanol acute, oral: 114 to 140 mg/kg, rat test animals 61 ------- Appendix C CHEMICAL REACTIONS IN ARGON AND HYDROGEN PLASMAS This section includes the results of those plasma reactions which oc- curred in argon and hydrogen environments. The tests were continued during the program until it became apparent that these or related carrier gases would not be suitable for detoxification in a microwave plasma pro- cess. C. 1 MALATHION-ARGON AND MALATHION ARGON-OXYGEN PLASMAS Decomposition reactions of Malathion in 200-250 W, 100 torr pressure, pure argon, and mixed argon-oxygen plasmas were investigated in the lab- oratory-scale plasma system. In pure argon, the yellow-brown products were offensive and malodorous, similar to mercaptan and disulfide com- pounds. In mixed 50/50 vol.% argon-oxygen, the products were essentially identical to that of the 100% oxygen plasma reactions, whereas in 80/20 argon-oxygen plasmas, mercaptan (skunk-like) but not sulfide, odors were again evident. Although analyses of these materials were not pursued ac- tively, it was possible, nevertheless, to postulate the products of decom- position. This was based on the mass fragment formations which result from the ionization processes that occur during mass spectrescopic analy- sis of Malathion and related materials (13). They are presented as follows: S H I Ar plasma ] S S S CH.O-P -4-Sf C-CO9C,H,. 1) CH.O-P-SH 2) CH,O-P-S-S-P-OCH, O | ' I ^ ^ 3 O i «3 i i O OCH, C-CO.C-hL OCH_ OCH. OCH, O A ^ / 0 J O o H2 H H 3) HO-C-C-S-S-C-CO,H 4) CH-CO,C,H_ & | | i || * «m O H02C-CH2 I- 5) H-S and lower hydrocarbon sulfides, mercaptans 62 ------- C. 2 PCS - ARGON PLASMA REACTIONS Aroclor No. 1242 liquid was passed through a 250 W argon plasma at 100 torr. In contrast to oxygen plasma results, a sol vent-soluble, black tar-like oil formed in the reactor tube and dripped slowly down the sides during the reaction. The gases evolved were determined by mass spectres- copy to be the following: (1) Hydrogen chloride HC1 (2) Ethylene CH2 - CH2 (3) Acetylene HC^CH (4) Methyl chloride CH3C1 Although the tars were not analyzed extensively, later infrared deter- mination of similar processes in hydrogen indicated that the black materials contained substantial quantities of unreacted PCB's. C. 3 PCB - ARGON-OXYGEN PLASMA REACTIONS In a 50-50 volume percent oxygen/argon plasma, however, under lab- oratory-scale conditions similar to those described in subsection C. 2, tar- like liquids, were not formed in the reactor. There were not more than traces, if any, of unsaturated hydrocarbons, chlorine oxide, carbonyl chloride, or methyl chloride in the gas effluent. The products were as follows: (1) Hydrogen chloride HC1 (2) Carbon dioxide CO2 (3) Carbon monoxide CO (4) Water H2O C.4 PCB - HYDROGEN PLASMA REACTIONS PCB's 1242 and 1254 were passed through the expanded-scale system in which a hydrogen plasma at 50 torr had been activated. Powers ranged from 4. 0 to 4. 6 kW. The principal gaseous reaction products, as deter- mined by the mass spectrometer, were as follows: (1) Hydrogen chloride HC1 (2) Acetylene HC s CH (3) Methane CH4 63 ------- A black, oily liquid with an aromatic naphthalenic odor dripped into the receiving tube after several minutes of reaction time. Analysis for chlorine content and infrared spectroscopy indicated that the liquid was principally a form of the original starting material. C. 5 METHYL BROMIDE - ARGON PLASMA REACTIONS Gaseous methyl bromide (2 to 3 g/hr) was passed through 300 to 400 W argon plasmas at 50 torr in the laboratory- scale system. The products of reaction detected by the mass spectrometer were as follows: (1) Bromine Br2 (2) Hydrogen bromide HBr (3) Ethylene CH2=CH2 (4) Acetylene CH s CH (5) Methane CH4 Carbonaceous flake deposits were formed in the reactor tube. Quanti- tative analysis by the mass spectrometer showed that not less than 99 per- cent conversion had occurred. C. 6 METHYL BROMIDE - ARGON- OXYGEN PLASMA REACTIONS Reactions of methyl bromide in 50-50 vol. % argon- oxygen plasmas yielded the following products as estimated by the mass spectrometer: (1) Bromine (2) Hydrogen bromide HBr (3) Ethane C0HC Z D (4) Carbon dioxide COg (5) Carbon monoxide CO (6) Water HgO The unequivocal identification of several of these materials was diffi- cult because of similarity in their masses, e. g. , HBr+, M/E = 80, 82, which registers adjacent to Br+, M/E = 79, 81 on the spectrometer re- corder charts. There was also no indication of the carbonaceous deposits, i. e. , flakes, which were formed in the 100 percent argon plasmas. This is identical to the absence of carbon deposits reported above for 50-50 argon/ oxygen PCB plasmas. 64 ------- TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) REPORT NO. EPA-600/2-77-030 3. RECIPIENT'S ACCESSION-NO. 4. TITLE AND SUBTITLE 5. REPORT DATE DEVELOPMENT OF MICROWAVE PLASMA DETOXIFICATION PROCESS FOR HAZARDOUS WASTES - PHASE I April 1977 (Issuing Date) 6. PERFORMING ORGANIZATION CODE 7. AUTHOR(S) Lionel J. Bailin Barry L. Hertzler 8. PERFORMING ORGANIZATION REPORT NO. PERFORMING ORGANIZATION NAME AND ADDRESS Lockheed Palo Alto Research Laboratory Lockheed Missiles and Space Company, Inc. Palo Alto, California 94304 10. PROGRAM ELEMENT NO. IDC 618 11. CONTRACT/GRANT NO. Contract 68-03-2190 12. SPONSORING AGENCY NAME AND ADDRESS Municipal Environmental Research Laboratory—Cin.,OH. Office of Research and Development U.S. Environmental Protection Agency Cincinnati, Ohio 45268 13. TYPE OF REPORT AND PERIOD COVERED Final for Phase I, 1976 14. SPONSORING AGENCY CODE EPA/600/14 15. SUPPLEMENTARY NOTES This contract effort is continuing, and currently Phase II is underway. 16. ABSTRACT The microwave process described in this report is a relatively new application of what has been termed the "fourth state of matter", or the "plasma state". It is the first practical application of a microwave discharge to the decomposition of chemical compounds in significant quantities. This report describes a recent, successful, R&D effort in which a former "grams-per-hour" system was scaled up to a 5 to 7 pounds-per-hour system, and then its performance was verified with several typical hazardous materials. The materials tested and detoxified were Malathion, methyl-bromide, polychlorinated biphenyls, phenylmercuric acetate, and Kepone. Complete detoxification resulted. Further benefits of the process are the competitive, reasonable costs of about $0.20 per pound of material processed, including all costs. The process warrants further development, namely additional scale-up to pilot and field units. Presently, units up to 100 pounds per hour or so appear feasible to construct and be operable within two or three years. 17. KEY WORDS AND DOCUMENT ANALYSIS DESCRIPTORS b.lDENTlFIERS/OPEN ENDED TERMS COSATI Field/Group detoxification, degradation, hazardous materials, waste disposal, microwave equipment, pyrolysis, industrial waste treatment, waste treatment, organic wastes, electromagnetic radiation new microwave tech- nique, microwave- disposal process 13B 18. DISTRIBUTION STATEMENT Release to Public 19. SECURITY CLASS (ThisReport) unclassified 21. NO. OF PAGES 79 20. SECURITY CLASS (This page) unclassified 22. PRICE EPA Form 2220-1 (9-73) 65 # U. S. GOVERNMENT PRINTING OFFICE: '977-757-056/5621 Region No. 5-11 ------- |