ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN ANN ARBOR Progress Report, January, 1956 THE EFFECT OF RADIATION ON CHEMICAL REACTIONS Joseph J. Martin Associate Professor, Department of Chemical and Metallurgical Engineering Leigh C. Anderson Professor and Chairman, Department of Chemistry Project Supervisors Bruce G. Bray E. I. Haiba Senior Research Ass.is~tant4 I. A. Baki L'. - D..Boddy Ben.G,. Bray ER. A. Carste4s W. %.;1heng H..E. C61en U'. Curtius H. M,: d':Emnus; R. M. Fitch D. Roper B. Schmall W. M. Sergy F. Sevcik J. Spencer D. E. Taylor L. D. Thomas Assistants in Research Project 1943-4 U.S. Atomic Energy Commission Contract No. AT (11-1)-162 Chicago 80, Illinois March 1956

Engineering Research Institute ~ University of Michigan PREFACE For the past five years a research group at The University of Michigan has been studying the effect of high-energy radiation on chemical reactions. The activities of that group during the last half of 1955 are the subject of this report. The primary support for this work has been from the United States Atomic Energy Commission through its sponsorship of a project in the Engineering Research Institute of the University. Support for research in the same field has also been received from Michigan Memorial-Phoenix Project No. 98, with funds contributed by the Chrysler Corporation. This report represents a collaborative effort of the personnel from both projects. These personnel have shared the same equipment and facilities in the Fission Products and Phoenix Project laboratories and have cooperated with each other in carrying out the overall objectives of the program. In addition to acknowledging the financial support of the Atomic Energy Commission and the Michigan Memorial-Phoenix Project, the opportunity is taken here to express appreciation to several other organizations which have been most helpful. The Engineering Research Institute itself, through a director's grant, supplied approximately five thousand dollars worth of new laboratory equipment that made it possible to conduct a number of experiments which could not have been performed otherwise. The gift of equipment included an Aminco high-pressure reactor bomb, a Magna-Dash magnetically stirred reactor, a Brown recorder-controller, a Flexopulse electronic timer, a gas booster pump, and miscellaneous high-pressure stainless-steel piping and fittings. The Dow Chemical Company supplied a large cylinder of oxygen-free ethylene, together with a chemical analysis of the cylinder's contents. The Continental Oil Company supplied samples of nonene and dodecene which were used in several reactions. The Koppers Company, Inc., supplied an interesting comparison of the physical properties of polyethylene produced by gamma radiation here at the University with those of polyethylenes produced by commercial processes. Special appreciation should also go to Dr. David E. Harmer, formerly with this project and now with the Dow Chemical Company, who gave generously of his own time in aiding some of the present laboratory assistants to master the techniques of conducting chemical reactions in gamma radiation fields. iii

Engineering Research Institute ~ University of Michigan Some of the material included in this report was presented at the International Conference on Peaceful Uses of Atomic Energy, Geneva, Switzerland, August 8-20, 1955. It is expected that most of the remainder of the subject material will be offered for publication in current scientific journals. iv

Engineering Research Institute ~ University of Michigan TABIE OF CONTENTS Page LIST OF FIGURES vi LIST OF TABLES Vii ABSTRACT viii INTRODUCTION 1 EXPERIMENTAL WORK 2 A. Polymerization of Ethylene 2 B. Copolymerization of Butadiene-Styrene Mixtures 12 C. Copolymerization of Sulfur Dioxide with Various Olefins 18 D. Oxo-Type Reaction 29 E. Chlorination of Some Aromatic Hydrocarbons 35 ECONOMICS OF RADIATION CHEMICAL PROCESSING 43 A. Design of a Chlorination Process Using Gamma Radiation 43 B. Cost Estimations 46 COOPERATIVE WORK WITH MICHIGAN MEMORIAL-PHOENIX PROJECT NO. 98 48 Preparation of Trichloroacetic Acid 48 CONCLUSIONS AND PROPOSALS FOR FUTURE WORK 49 BIBLIOGRAPHY 52

Engineering Research Institute ~ University of Michigan LIST OF FIGURES Page 1. Working drawing of the new reactor head for the Lewis reactor. 4 2. Disassembled view of the Aminco reactor. 5 3. Melting point and molecular weight of polyethylene. 10 4. Density and crystallinity of polyethylene. 11 5. Heavy wall pyrex vials used in small batch experiments. 14 6.Schematic diagram of vial-loading equipment. 15 7. Diagram of the stainless-steel reactor used in S02 copolymerizations. 19 8. High-pressure gas-loading rack. 21 9. Propylene-sulfur dioxide copolymerization at room temperature. 27 10. Disassembled view of the Magna-Dash reactor and agitator timer. 31 11. Large-volume reaction tank. 32 12. Schematic diagram of laboratory equipment used in aromatic chlorinations. 40 13. Redesigned gas rack used for measurement and control of reactant and tail gases in the chlorination experiments. 41 14. Flowsheet of proposed plant design for producing benzene hexachloride. 45 vi

Engineering Research Institute * University of Michigan LIST OF TABLES Page 1. Products from the Irradiation of Commercial Ethylene 6 2. Compositions of Ethylene Used 7 3. Products from the Irradiation of Oxygen-Free Ethylene 7 4. Physical Properties of Polyethylene Produced by Gamma Radiation of Commercial Ethylene 9 5. Comparison of Some Properties of Gamma-Initiated and Conventionally Produced Polyethylenes 12 6. Grade A 1,3-Butadiene —Styrene Copolymerization at 10~-15~ Centigrade 17 7. Grade B 1,3-Butadiene —Styrene Copolymerization at 10~-15~ Centigrade 17 8. Copolymerization of Various Alkenes with Sulfur Dioxide at 80 to 15~ Centigrade and 153,000 to 162:,000 Rep per Hour 23 9. Large-Scale Copolymerization of Various Alkenes with Sulfur Dioxide 25 10. Reaction Conditions and Product Yields for the Oxo Reaction Involving Various Olefin Reactants 34 11. Chlorination of Aromatic Compounds with Gamma Radiation 38 12. Cost Estimate for Producing 454 kg/Day of Gamma Isomer of Benzene Hexachloride with Four. Radiation Sources 47 vii

Engineering Research Institute ~ University of Michigan ABSTRACT During the period of this report investigations have been conducted on a half-dozen different types of reactions to determine the extent to which they are catalyzed or promoted by gamma radiation. The results of the experimental studies may be summarized briefly as follows: A. The study of the polymerization of ethylene was continued to include a wide range of temperature and pressure. The products of polymerization ranged from liquids to waxes to solids. The radiation yields were fairly high, indicating chain-reaction mechanisms and giving some promise for eventual commercial application. B. Various mixtures of 1,3-butadiene and styrene were copolymerized in different ways. Hiighest radiation yields were obtained in emulsion polymerization. The products were examined visually, but no physical properties were determined that would allow their potential use as plastics or elastomers to be clearly defined. C. Sulfur dioxide was reacted with a number of individual olefins. Polymerization reactions generally took place very rapidly with high radiation yields. The products appeared to be plastic-like compounds with rather high melting points. D. The Oxo reaction involving the condensation of hydrogen, carbon monoxide, and an olefin was the subject of considerable study. Solids, waxes, oils, and thin liquids were obtained as products; subsequent distillation and test of the fractions indicated the presence of aldehydes and alcohols in the product. For the runs made thus far, the radiation yields were not very high under the reaction conditions employed. E. The chlorination of aromatic hydrocarbons was extended to include benzene, toluene, xylene, mesitylene, naphthalene, and ethyl benzene. In general, these aromatics react vigorously with chlorine at room temperature in the presence of gamma radiation. Gamma radiation tends to promote the addition of chlorine to the benzene nucleus more than the substitution of chlorine for hydrogen either on the nucleus or on a side chain......... viii,

Engineering Research Institute ~ University of Michigan F. Trichloroacetic acid was produced by the chlorination of chloral hydrate in the presence of gamma radiation. The reaction was fairly rapid and is being subjected to further study. In addition to the experimental chemical reaction studies, a speculative plant design was developed to illustrate procedures which might be economically feasible for radiation chemical processing. The process chosen for study was the chlorination of benzene to produce benzene hexachloride, a wellknown insecticide. The present commercial process for the latter compound utilizes ultraviolet light, so that a comparison of the approximate costs of the gamma-radiation reaction and the ultraviolet-light reaction could be made. On the basis of somewhat limited information, which required estimation of radiation costs, it was concluded that the gamma-radiation process is not impracticable and could compete with the ultraviolet process. More precise costs of gamma radiation will lead to a more firm comparison of the two processes. ix

Engineering Research Institute ~ University of Michigan INTRODUCTION The promotion of chemical reactions by radiation is not a new field, as exemplified by the fact that ultraviolet light has long been employed in such studies. However, the use of high-energy radiation, or ionizing radiation as it is often referred to, is comparatively recent since it dates from the discovery of radium and the invention of the x-ray machine. Because of the somewhat limited amount of radiation available from radium or from x-ray machines, only a limited amount of work had been done on the influence of such radiation on chemical reactions. The advent of the atomic energy program, however, changed this situation. The vast quantities of radiation that are released either directly or indirectly as a result of nuclear fission have stimulated many research activities to discover uses for high-energy radiation, and the promotion of chemical reactions appears to be one of the more promising applications. Neutron, gamma, and beta radiation are released directly in a nuclear reactor. Some materials which may be circulated through the reactor may deliver gamma or beta radiation. Finally, the fission products which may be recovered from the spent fuel are very copious sources of gamma and beta radiation. At the present time these highly radioactive fission products are considered a waste material which requires rather expensive storage to prevent contamination of the surroundings. If the radiation from these fission products can serve as a catalyst in useful chemical reactions, a difficult wastedisposal problem may be alleviated to some extent and overall productivity increased at the same time. Ideally, a program to find chemical reactions promoted by fissionproduct radiation might well be carried out with fission-product sources; however, in the present stage of development of the atomic energy program, fissionproduct sources are not readily available. It has been necessary, therefore, to conduct investigations with other sources of radiation. Since beta radiation has a low penetrating power, it is not expected that this radiation from fission-product sources could be used effectively as catalysts because the fission products would have to be in almost direct contact with the chemical reactants. This could result in dispersal of the fission-product catalyst and, consequently, a very difficult contamination problem could arise. On the other hand, gamma radiation is extremely penetrating, and, therefore, chemical

Engineering Research Institute ~ University of Michigan reactions which are catalyzed by gamma radiation would constitute very favorable uses for the fission products. The fission products could be isolated in sealed stainless-steel containers which would transmit the radiation and the fission products then could be kept completely isolated from the chemical reactants. In the absence of fission-product sources and because of the expected use of gamma radiation, the experimental work at Michigan has been conducted with cobalt-60 sources. Although the gamma radiation from cobalt-60 has a somewhat shorter wavelength than that from fission products, it is believed on the basis of considerable scientific evidence that the chemical effects will be very similar. The Fission Products Laboratory is equipped with two powerful cobalt60 sources. One of these is nominally rated at ten kilocuries and has an actual strength of about three kilocuries; the other is nominally rated at one kilocurie and has an actual strength of about 300 curies. The Phoenix Project has acquired a new cobalt-60 source which is estimated to have a strength of about four kilocuries, though no accurate calibrations have yet been made. The reactions described in the subsequent sections were all carried out in the presence of one or more of these sources. The intensities of radiation were determined at the location of the chemical-reaction equipment by means of ferrous sulfate dosimetry and electronic ionization type instruments. These procedures have been described in previous reports along with a number of other chemical reactions which have been investigated. The reader is referred to the Bibliography for a complete description of the earlier work. EXPERIMENTAL WORK A. POLYMERIZATION OF ETHYLENE The study of the effect of gamma radiation on the polymerization of ethylene was initially reported by Lewis.17 This work has been extended with the primary emphasis upon the effect of temperature variation at pressures in the region of from 75 to 100 atmospheres. A portion of this work was reported 4 in Progress Report 7 of this laboratory and the work reported here is a continuation of these earlier experiments. Two grades of ethylene were used in this investigation. The first study involved commercial ethylene as available in cylinders from several industrial concerns; later, the ethylene was a special grade having a very low oxygen concentration. 1. Equipment Used.-The majority of the runs were conducted in the

Engineering Research Institute * University of Michigan high-pressure stainless-steel reactor designed and reported by Lewis.l7 A set of heaters built by Lewis was placed around the bomb, even when heating was not required, in order to insure that the radiation flux would be the same in all cases. Some difficulty was encountered at high temperatures with gas leakage through the thermocouple junction at the head of the bomb. This thermocouple was inserted into the pressure region through a port that was sealed with a Teflon gasket. After several hours irradiation at high temperatures and pressures, leaks developed in the gasket and vented the ethylene to the atmosphere. To correct this situation, a new reactor head was designed with an external entry thermowell. The thermowell extended into the center of the bomb to the same relative region as previously, but there were now no high<pressure seals to contend with. At steady-state conditions the temperature measurement was the same as previously. The design specifications for the new reactor head are found in Fig. 1. The head was turned from a solid piece of A.I.S.I. 304 stainless steel. The thermowell was constructed of 1/4-inch high-pressure A.I.S.I. 316 stainless-steel tubing welded to the head with A.I.S.I. 316 stainless-steel welding rod. The design conditions were 2000 psi at 2000C. All dimensions are to be found in Fig. 1. The new head has been used satisfactorily at the design conditions under radiation. The original head and reactor body have been completely described by Lewis. Some of the runs were made in a Standard Aminco reactor supplied by a grant from the Engineering Research Institute. An American Instrument Company 4-3/8-inch series reaction vessel constructed of 316 stainless steel (Cat. No. 406-25J3), 3-5/16-inch internal diameter by 10-inch internal height, was purchased for this work. This reactor was used in runs where pressures higher than 2000 psi were encountered. A pressure gage was attached to the head of the reactor for pressure readings. The bomb was equipped with a thermowell for thermocouple temperature measurements. The reactor was equipped with a heater for high-temperature work. A photograph of the disassembled reactor is to be found in Fig. 2. 2. Experimental Procedure.-The same loading and cleaning procedure was followed for each run. The reactor was cleaned with boiling xylene and rinsed with acetone and distilled water. The vessel was sealed mechanically and was pressure tested with nitrogen overnight at a pressure of 1000 psi or greater. On the day of loading, the nitrogen was vented to the atmosphere and the bomb was evacuated to a pressure of less than one millimeter of mercury. The ethylene was loaded to a desired pressure computed beforehand to give the approximate pressure condition desired at the reaction temperature. The reactor was placed in the source cave and brought to the desired temperature with the source in the well. When this predetermined temperature was reached, the source was raised into the irradiation position for

9/16-18 RH I HOLE 4.00 o.D. 1-3/4"-DRILL 4 HOLES 9/16 DIA. ({3 7CI /6 7/ GASKET 2.88 B.-C. ------------— ~ I II I THERMOCOUPLE WELL — 4.00' 0.0. 1 (4- /4" OD. STAINLESS STEEL AISI 316 TUBING - 1 —-/2' — I BREAK SHARP CORNERS WITH FILE I/8 "DIA. ROUND 0 5/8 WELD AND GRIND ROUND _ SECTION B-B 0 7/8" NOTE: MAT'L. FOR HEAD 304 S.S.TURNED FROM ONE PIECE WELDING ROD 316 SS. FOR ALL WELDS OPERATING CONDITIONS 2000 PSIA AT 2000C - 1. 245 —, 1.505 SECTION A-A Fig. 1. Working drawing of the new reactor head for the Lewis reactor.

0 C., 0ma, 0 0~ 0);' Cl00,0 0,0E. 0,,,0,.t,. l.t0t,00 g0,~gg0, 00SlEg.00.00000 0SLSE.iEE.Ri E.iEE. RiEEE, E.E R R E #REiE.iE.S.iSiEEi.. EEEEEE 0:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: C,,E,,..,.EE.......E,, t0)0,.::,w, C.iEEE.i.i E.iE.,EEiEiEi.ii E..EEEEEiEi E E.. ~ERER LEEE.iEEE iE..iR..,E..iEiE..... E EiEi E E iE E Ri.RE.

Engineering Research Institute * University of Michigan a specific time. During this irradiation period, temperature and pressure readings were made at various intervals. After the desired accumulated dose had been obtained, the excess ethylene was vented from the reactor. The reactor was then disassembled and the product was recovered and stored for further analysis. 3. Experimental Results. —A complete summary of the reaction conditions and polymer yields from the irradiation of commercial ethylene is given in Table 1. White, curdy, solid polymers were formed at reaction temperatures from 6~ to 130~C. A white wax formed at about 1600C, and above this temperature the product was an opaque liquid that separated into two layers after standing for several hours. The top layer was colorless and the bottom layer contained.a finely divided solid. The yield of polymer is presented as the A value,17 gm moles of ethylene reacted/ (metric ton of ethylene charged) (megarep of gamma radiation). The rep is based on the absorption of 93 ergs per gram. From the scatter of the data it is difficult to determine the effect of temperature on the yield since the largest yields were at 39~ and 217~C. A portion of Table 2 from Progress Report 7 is presented here in Table 1. The reason for this duplication in the presentation of the data is that the analyses of the products reported in Progress Report 7 were not complete at the time of writing. Complete analyses have now been made and are presented here. Runs 145839, 145833, 145845, 154712, and 145858,which are low-temperature and intermediate-temperature runs, have not been reported before. TABLE 1. PRODUCTS FROM THE IRRADIATION OF COMMERCIAL ETHYLENE Run Reaction Reaction Radiation Total Radi- Tensile on Temp. Pressure Dose Rate ation Dose Radiation Polymer Strength Page 0C Avg Psia Krep/Hr Megarep Yield A* State Psi 145839 6 755 61.0 4.20 1340 Solid 3050 145833 8 985 59.0 5.46 1630 Solid 3075 145800 13 1145 63.0 4.10 1530 Solid 3350 145801 13 875 63.0 5.28 2270 Solid 3100 145807 39 1380 63.0 4.47 3040 Solid 2520 145808 88 1325 26.5 1.90 490 Solid 145812 90 1300 62.5 5.85 433 Solid 145845 132 1700 43.0 4.05 955 Solid 154712 160 1490 31.0 2.76 945 Wax 145848 190 1125 25.0 2.88 1300 Liquid 145814 217 1100 33.0 3.77 4030 Liquid 145813 220 1175 62.5 4.86 1400 Liquid *A = gm moles reacted (metric ton charged) (megarep)

Engineering Research Institute ~ University of Michigan The commercial ethylene and, consequently, the gas compositions are the same as presented by Lewis.l7 This approximate analysis is found in Table 2 with the analysis of oxygen-free ethylene as supplied by the Dow Chemical Company for this work. TABLE 2. COMPOSITIONS OF ETHYLENE USED Percent Percent Percent Percent 02Material CO C02 N Combustibles Commercial.02.05.37 1.7.8% Pentane Oxygen Free <10 ppm Nil Nil Nil Z10 ppm Acetylene 1.5% Ethane A summary of reaction conditions and polymer yields from the oxygenfree ethylene is presented in Table 3. The product had the same appearance as that from the commercial ethylene. In general, the A values are higher than the A values for commercial ethylene at the same temperatures. Harmerl4 showed that small amounts of oxygen inhibited the reaction of chlorine with aromatic hydrocarbon. A comparison of Tables 1 and 3 indicates that oxygen may inhibit the polymerization reaction of ethylene in a similar manner. TABLE 3. PRODUCTS FROM THE IRRADIATION OF OXYGEN-FREE ETHYLENE Reaction Reaction Radiation Total Radi- Overall Polymer Temp. Pressure Dose Rate ation Dose Radiation Number ~C Avg Psia Krep/Hr Megarep Yield A* State 154039 8 1800 24.2 9.38 3340a Solid 154038 13 810 54.0 7.46 1990a Solid 154011 26 1000 24.4 3.98 3280 Solid 154034 63 1370 25.0 3.0o8 2450 Solid 154032 103 1310 27.5 2.90 970 Solid 154007 210 2400 10.5 2.40 13,000 Liquid A = gm moles reacted (metric ton charged) (megarep) a. Not believed true indication as excessive radiation was given. 7

Engineering Research Institute ~ University of Michigan 4. Evaluation of the Polymer Product. —Table 4 is a summary of the physical properties of the polyethylene produced from commercial ethylene under gamma radiation. Melting points of the solid and waxy products were taken on a melting-point bar as,described in previous reports. 177 Melting points of the liquids were observed in small tubes of the polymer which were allowed to warm up after being frozen in a dry-ice bath; two measurements were made on most samples and both values are listed in Table 4. It should be emphasized that only the lower limit of the melting-point range is presented. Because each product was a composite of polymers of varying molecular weights, a range of melting points was always observed. In general, the larger the dose of radiation applied, the larger the range of molecular weights because of the greater amount of cross linkage of polymer molecules in a high radiation field. The points at which melting could be detected first have been plotted in Fig.3as a function of reaction temperature. Solid products which were produced in appreciable quantity were pressed in a one-compartment mold at 1500 psi and 1500C for five minutes. Specimens were punched from the molded polymer sheet and placed in an Instron Tensile Test Machine. The tensile strengths were determined at a head and/or jaw speed of ten inches per minute. These tensile strengths are presented in Table 1. Specimens were also punched from the molded polymer sheet for the determination of melt viscosity. The melt viscosities of these solid polymers were obtained in a parallel-plate plastometer at about 1370C. Viscosities of the liquid products were determined in a modified Ostwald pipette. Molecular weights were estimated from the viscosities by the method of Dienes and Klemm9 and are shown in Fig. 3 as a function of reaction temperature. Because there is a range of molecular weights formed in the polymerization process, the molecular weights presented in Table 4 and Fig. 3 are only approximate, but a general trend is indicated. Densities of the solid polymers were determined by the Archimedes principle, while those of the liquids were measured by weighing in a calibrated pipette. The numerical values of the densities are found in Table 4. Figure 4 is a plot of the densities as a function of the reaction temperatures. The crystallinities or the degree to which the molecules are arranged in parallel positions were determined from the densities by the method of KirkOthmer.16 These crystallinities are found in Fig. 4 as a function of the reaction temperature. A portion of the polymer product from run no. 145833 was supplied to Koppers Company, Inc., for analysis and comparison with their own polyethylene products. The comparative analysis as run by the Koppers Company appears in Table 5. It is felt, however, that a range of melting points and molecular

Engineering Research Institute ~ University of Michigan TABLE 4. PHYSICAL PROPERTIES OF POLYETHYLENE PRODUCED BY GAMMA RADIATION OF COMMERCIAL ETHYLENE Lower Run Reaction Limit Density Crystallinity Approximate Temp. M it. ng w X Molecular Number Temp. Melting gm/cm Percent Molecular C Range R Weight 145839 6 724 0.953 77.5 25,200 734 145833 8 723 0.945 69 26,500 737 145800 13 718 0.950 77 24,150 145801 13 723 0.943 73.5 34,500 719 145807 39 717 0.940 72 37,100 and 720 38,600 145808 88 682 0.937 70 Samples 145812 90 677 0.930 66 Too small 145845 132 648 0.912 56 Too small 653 154712 16o 487 0.879 36 Too small 499 145848 190 393 0.849 Liquid 8,840 388 145814 217 371 0.725 Liquid 5,690 382 145813 220 395 0.860 Liquid 8,950 391 *... l... 9

800 700,~ 600 40PPW z~~~~~~~~~~~~~~~~ C~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~ rr~~~~~~~~~~~~~~~~~~~C 0 W 8O I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~, Z~~~~~~~ e Pa 1C 0 3600 ~,0 ~ a ~~ 200 Z,.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~U H I-~. C0 o o 0 z 0 20 40 60 80 100 120 140 160 180 200 220 REACTION TEMP OC Fig. 3. Melting point and molecular weigh~t of polyethylene. ~~~~40C ~ ~ ~ ~ ~ ~ -~~~~~~~~~~~~~-,~~~~~~~~~~~~~~~~~~~~~~~~~0.9 30C' l0,000 o (0 2 0 C ~ ~ ~ I I. _ _ _ _ _ _ _ _ _ _ 020 4060801I00 120 140 160 1802:002:20: REACTION TEMP 0C Fig. 5. Melting point and molecular \~eight of polyethylene.

Engineering Research Institute ~ University of Michigan LOO0 K0O 90 0,, 80.9001 - 80 E E I I I\ I I z Iz l.800 60 50 0 20 40 60 80 100 120 140 160 180 200 220 REACTION TEMP OC Fig. 4. Density and crystallinity of polyethylene. 11

Engineering Research Institute ~ University of Michigan weights were present in the gamma-initiated polymer instead of the single values that are indicated. This is shown in the case of the melting-point data for the same run taken in this laboratory and found in Table 4. TABLE 5. COMPARISON OF SOME PROPERTIES OF GAMMAINITIATED AND CONVENTIONALLY PRODUCED POLYETHYLENES Polyethylene Sp Gr MP, ~C(C) Molecular Weight(d) Gamma initiated(a) 0.9459 124.5 42,800 "Super Dylan" (b) (Ziegler) 0.95-0.98 135-139 40,000-120,000 "Dylan"(b) (conventional) 0.91-0.92 114-116 20,000-30,000 (a) Run 145833. (b) Koppers trade marks. (c) Crystalline melting point under crossed Nicol prisms. (d) Determined from viscosity of decalin solutions at 135~C. Intrinsic viscosity = 1.141 x 10-3 x (mol. wt)0 66 The physical properties of the polyethylene produced from oxygenfree ethylene have not been completely evaluated as yet. For this reason the presentation is delayed until complete analyses can be made. From the results presented in Tables 1, 3, and 4 it is clear that a wide variety of polymers of ethylene w'ere produced in the radiation reaction. From the results of oxygen-free ethylene polymerization it was seen that oxygen had a definite inhibiting effect upon the reaction. The yields were sufficiently high to consider the process feasible for commercial applications. At this stage, economic studies are in order for the determination of the possible competitive position of polyethylenes produced by radiation processing with respect to the position of those produced by conventional processes. B. COPOLYMERIZATION OF BUTADIENE-STYRENE MIXTURES Preliminary investigations have been made on the effect of gamma radiation upon the copolymerization of 1,3-butadiene and styrene and upon emulsions of 1,3-butadiene and styrene in various soap solutions. These 12

Engineering Research Institute ~ University of Michigan preliminary experiments involved small quantities of reacting materials in glass vials which were given various doses of gamma radiation. The amount of reactant used in the preliminary experiments was relatively small and no attempt, beyond a mere visual inspection, was made to evaluate the products obtained. 1. Equipment Used. —All these investigations were carried out in heavy-walled vials blown from thick pyrex tubing. Figure 5 is a diagram of the vials used and the average dimensions of the vials; when sealed, they will safely withstand pressures up to 400 psi. In this work the pressures encountered with 1,3-butadiene are of the order of 50 psia. A schematic diagram of the gas-loading rack, which was constructed of heavy pyrex tubing to facilitate loading the vials, is shown in Fig. 6. The rack was equipped with three loading lines, one of which was used as an emergency vent to a hood. A gas-scrubbing column and a gas-drying column, which were incorporated in one of the load lines, could be used to remove any impurity that was thought to be present in the gas, and a cold trap to condense vapors that might have harmed the vacuum pump was also included. The standard high-vacuum "Duo-Seal" pump capable of producing a vacuum of less than 100 microns of Hg was used in this work. Dry ice in a 60-40% mixture of chloroform and carbon tetrachloride provided a temperature of -75~C for the condensing baths in both the cold trap and the cold bath around the vial. 2. Chemicals Used in the Reactions. -Some of the chemicals used in these preliminary experiments were obtained from commercial sources and others were prepared in this laboratory. The analyses and methods of preparation were as follows: 1,3-butadiene Grade A: 0.4% butane and/or butene 0.5% acetylene 0.1% phenyl P-naphthylamine inhibitor 1,2-butadiene also present in unknown amount Grade B: 99.77% 1,3-butadiene Normal butane, butenes, and possibly a trace of acetylenes (0.23%). Styrene 500 ml distilled under vacuum from stabilized reagent-grade styrene, rejecting initial 50 ml and final 150 ml. 13

Engineering Research Institute ~ University of Michigan 1/16" A B 5" HEAVY WALL PYREX 8 v Fig. 5. Heavy wall pyrex vials used in small batch experiments.

Engineering Research Institute ~ University of Michigan SCRUBBING DRYING COLUMN COLUMN LOADING LINE #2 STOPCOCK STOPCOCK TO VACUUM E3 E2 E, PUMP N LOADING LINE #2 TO VIAL STOPCOCK E~~~~~4 ~LOADING OR VENT LINE #3 MANOMETER COLD COLD BATH TRAP Fig. 6. Schematic diagram of vial-loading equipment. Nacconal and Alconox Solutions Saturated solutions in distilled water were prepared in the laboratory and used throughout the experiments. Other Soap Solutions Solutions of sodium stearate, calcium stearate, zinc stearatej and lead oleate were prepared from reagent grades and used throughout the experiments. 3. Experimental Procedure. —A standard method for loading, irradiating, and handling the vials was adopted and followed in all cases. This procedure was as follows. Emulsifying solutions and liquid reactants at room temperature were introduced into the vial by pipette, weighed by difference, and frozen in the cold bath. The vial was connected to the system by means of high-pressure tubing and placed in the cold bath. The whole system was evacuated to less than l-mm-Hg pressure; a high vacuum was necessary to avoid possible inhibition of the reactions by the oxygen of the air. In the first eight runs the Grade A 1,3-butadiene was loaded directly to the vial through loading line 2 shown in Fig. 6. In subsequent runs an attempt was made to remove the amine inhibitor thought to be present in the 15

Engineering Research Institute ~ University of Michigan gas by using loading line 1. The gas was bubbled through a 6 Normal sulfuric acid solution and dried in a calcium chloride tower before it was condensed in the vial. The Grade B 1,3-butadiene and all other gases used in later experiments were loaded directly to the vial through loading lines 2 and 3. After introduction of the desired reactants, a check was made on the line pressure by connecting the vial to the vacuum-pump manometer. The temperature of the cold bath was low enough so that if sufficient time was allowed for complete condensation to take place, the vapor pressure of butadiene was below atmospheric pressure and the vial was ready for sealing. Stopcock E3 was then closed and the cold bath lowered to permit heating by the natural-gas -oxygen torch of the region AB on the vial, as shown in Fig. 5. After uniform heating of this region to a temperature close to the softening point, the tip of the flame was concentrated on the region CD, which rapidly became soft. Due to the weight of the vial, the region CD elongated and, since the external pressure was higher than the internal pressure, the wall collapsed upon itself, the end result giving a fine tip capable of withstanding high pressures. The vial was exposed to radiation at room temperature for various predetermined periods of time. After removal from the source of radiation, the vial was cooled to -75~C and broken at the tip. The vials were then placed in the hood and allowed to warm to room temperature to permit escape of unreacted gases. The product was then recovered and processed to eliminate the soap, water, and unreacted styrene. It was then weighed and stored. 4. Experimental Results.-The results of the copolymerization of 1,3-butadiene and styrene are presented in Tables 6 and 7. The two grades of butadiene used are presented separately. No evaluation of the polymer products was attempted beyond mere visual inspection. The observed state of the polymer is described on the tables in the last column under remarks. The yields are recorded with the total doses of radiation received by each vial. These values, together with the molecular weight, permit a standard A value to be calculated. The A value is the same as that described previously for the polymerization of polyethylene.l7 It is calculated from the expression: A gram moles reacted _. yield x 106 (metric ton charged) (megarep) (mol. wt) (dose, megarep) When two reactants were charged, the weight average molecular weight used in the above formula was determined by the relation: (mol. wtl1) (wt 1 charged) + (mol. wt2) (wt 2 charged) Mol. wtavg 1,2 Mol tavg 1,2 wt 1 charged + wt 2 charged...... 16

TABLE 6. GRADE A l,3-BUTADIEN-STY]RENE COPOLYMRIZATION AT 100 -l50 CENTIGRADE Run Total Weight Weight of Dose Components on Run Charged Product Pret Rate Meoae A Value Remarks Page Yield Rep/Hr g B 154852 9 5.5 0.2 5.7 166,000 57.4 28.2 sticky substance S ~~~~~154832 6 4.0 5.5 - 125,000 28.1 - viscous -creamy-polystyreeB + S ~~~~~~~~145851 2 2.5 trace O 0 120,500 30.44 0 yellow-cream flakes, coudbfirs B + 5 145851 5 2.2 ~~~~~ ~~~~0.15 68 95,000 207 59 yellow-crystalline0 B + S 145851 1 2.4 trace ~ 0 0 control 00 B + S + K2S208 + H20 145851 5 4.1 0.5 7.5 115,000 26.92 38.6 yellow-curdy-waxy B + S + K2S208 + H20 145851 4 4.4 0.9 20. 4 158,000 51.57 94.0 milky liquid (So) B+S+Nacconal soln. 1482 7 660915.6 1800 5.6.7 white-granular solid (So)C B + S + Nacconal soin. 145855 14 7.5 1.5 17.5 155,000 5.26 795.5 rubbery-white-spongy subsac B + S + Nacconal solm. 145858 34 4.9 0.5 10.2 148,000 6.99 198.1 white precipitate (So) B + S + Nacconal soln. 145835 15 8.5 0.9 10.8 0 control - yellow-waxy B + S + Alconox.soln. 145855 12 7.4 1.0 15.5 148,000 5.63 576.2 brown-curdy B + S + Alconox soln. 145835 15 5.4 <0.1 0 0 control - flaky white (so)0 TABLE 7. GRADE B 1,5-BUTADIENE —STYBENE COPOLYMERIZ.ATION AT 100 -l5o CEN'TIGRADE Run ~Total Weight Weight of Pret Dose Dos Components on Run Charged Product Yed Rate Doegrp A ValueReak Page Yield Rep/Hr M ae B + S 145840 41 2.7 0.05 1.85 140,000 8.88 51.1 cream-yellow-fibrous-somecytl B + 5 145840 40 1.6 trace ^10 0 control - B + S + Alconox soln.. 145840 42 7.7 1.2 15.6 127,000 8.04 299.2 spongy white polymer B + S + Alconox soln. 145840 43 8.5 0.5 5.5 0 control - yellow hard polymer B + S + Na Stearate soln. 145842 46 5.6 0.8 14.5 140,000o 5.7 576.8 milky liquid (so) B + S + Na Stearate soln. 145842 47 5.6 trace " 0 0 control - B + S + Na Stearate soin. 145845 52 4.7 0.5 10.61 155,000 4.5 330.6 curdy white polymer (So) B + S + Na Stearate soin. 145845 55 7.5 0.5 4.1l 0 control - cream-crystalline-flaky B + S + Zn Stearate soln. 145842 50 6.7 trace ~ 0 140,000 5.87 00 B + S + Zn Stearate soln. 145842 51 6.5 trace'I 0 0 control -- B + S + Zn Stearate soln. 145844 59 5.9 2.2 57.5 155,000 6.75 859.1 milky white liquid (so) B + S + Zn Stearate soln. 145844 58 7.0 trace 1110 0 control - B + S + Ca Stearate soln. 145842 48 6.0 1.0 16.7 155,000 4.29 604.6 milky white liquid (so) B + S + Ca Stearate soin. 145842 49 6.5 trace IVO 0 control - B + S + Ca Stearate soin. 145844 57 8.0 2.7 53.8 155,000 6.75 804.2 curdy white solid (so) C B + S + Ca Stearate soln. 145844 56 7.1 trace!=0 0 control -0 B + S + Pb Oleate soln. 145845 54 5.8 2.7 46.5 165,000 4.51 1,9489.1 yellow layer lighter thanwtr(o B + S + Pb Oleate soln. 145845 55 5.1 4 0.1 2 0 control - brown-flaky Abbreviations: B = 1,3-butadiene S = styrene So =possible soap and water left

Engineering Research Institute ~ University of Michigan A perusal of the first equation indicates that conclusions as to the rate or extent of a reaction, based solely on A values, could be misleading. When the yield approaches 100% in a finite reacting mixture, large quantities of radiation may be applied with small change in yield. Thus, an infinite dose conceivably could be given to a 100%-completed reaction, resulting in a zero A value. Dosimetry calibrations presented by Lewis et al.18 were used throughout this work in calculating radiation doses. Table 6 presents the results of a number of experiments performed with the Grade A butadiene and styrene. Except for the butadiene-styreneNacconal and the butadiene-styrene-Alconox products, which showed interesting elastomeric properties, the amount and quality of the products obtained were generally low. The results obtained with Grade B inhibitor-free 1,3-butadiene are presented in Table 7. In some runs the soap and water were not completely removed from the products and it is believed that the higher yields obtained are partly due to this fact. Comparing the results shown in Table 6 and in Table 7 for the butadiene-styrene and for the butadiene-styrene-Alconox copolymerizations, it appears that the effect of the amine inhibitor contained in the Grade A 1,3butadiene was slight. Although the yields obtained in the 1,3-butadiene — styrene copolymerization reactions were not particularly high, it may be seen that generally the percent yields of the controls were relatively insignificant when compared to those of the irradiated samples. C. COPOLYMERIZATION OF SULFUR DIOXIDE WITH VARIOUS OLEFINS The copolymerization of sulfur dioxide with olefins has been of interest for many years. The resulting polysulfone is a thermoplastic resin having many physical properties similar to those of some major commercial plastics. There are inherent deficiencies that require correction before they are substituted for conventional plastics, but the large quantities of S02 and olefins available at modest cost have led to considerable research on the polysulfone resins. The reactions between S02 and olefins proceed under additive catalysts such as peroxides, nitrates, and other oxidizing agents, 27 and it has also been found that actinic light will catalyze the reaction.27 Preliminary studies in this laboratory by Lewisl7 showed that gamma radiation catalyzed the copolymerization of S02 and ethylene. In the past year much work has been done on the copolymerization of 18

Engineering Research Institute ~ University of Michigan sulfur dioxide with other olefins. Preliminary experiments have been carried out in glass vials, and in larger-scale experiments a stainless-steel reaction vessel was used. The olefins that have been employed are propylene, butene-l, butene-2, isobutylene, 1,3-butadiene, nonene, dodecene, and styrene. The reactions have been carried out primarily at room temperature and under the vapor pressures of the mixtures. 1. Equipment Used. —In the preliminary experiments heavy-walled, pyrex-glass vials were charged with S02 and the olefin by condensation. These vials and the gas-loading rack are the same as those described for the butadiene-styrene experiments. The vials are shown in Fig. 5 and the gasloading rack is shown in Fig. 6. The larger-scale reactions were conducted in a stainless-steel reaction vessel constructed of A.I.S.I. 304 stainless steel shown in Fig. 7. O Hoke angle pattern valve type PY 275 _ _ Q 304 stainless-steel bar stock, threads: standard pipe ~ Groove for poured lead gasket ~ Body: 304 stainless steel, schedule-80 pipe Length: 8 in. ID: 1-9/16 in. Thickness: 5/32 in. All welds are fillet welds; 316 stainlesssteel welding rod used for all welds. Fig. 7. Diagram of the stainless-steel reactor used in S02 copolymerizations. 19

Engineering Research Institute * University of Michigan The body of the bomb was constructed of schedule-80 pipe. The heads, turned from bar stock, and body were threaded to provide a means of closing the bomb. The bottom head was placed on the body and welded with 316 stainlesssteel welding rod. The top head was machined for a lead gasket to provide a pressure seal. The inlet provided in the top head for loading consisted of high-pressure stainless-steel tubing welded to the head. A standard Hoke valve was attached to this inlet tube. The maximum working pressure of this bomb was 1500 psi. Although 200 psi was the maximum pressure used in this work, the bomb was hydrostatically tested before use at the maximum pressure. A high-pressure steel loading system was constructed for loading gases into the reaction vessel. A photograph of the system is shown in Fig. 8. All lines were made of 6000-psi 316 stainless-steel pressure tubing, except for the two ends which were fitted with flexible steel hoses rated at 2000 psi. The valves were standard Hoke valves rated at 2000 psi. The equipment was provided with four pressure gages of various ranges for pressure measurement. The high-vacuum pump and cooling mixture were the same as those used in the preliminary experiments. 2. Chemicals Used in the Reactions. —The chemicals used in both the preliminary and large-scale experiments were standard commercial grades, except as indicated. No individual analyses were made on gas cylinders and the following analyses were indicated by the commercial suppliers. Ethylene-0.02% 02, 0.05% C0, 0.37% C02, 1.7% N2, o.8% pentane, 97.06% C2H4. Propylene-C.P. grade: 99.0% C3H6. Principal impurity was propane. Butene-l-C.P. grade: 99.0%. Principal impurities were isobutane and small amounts of normal butane and butene-2. No isobutylene was known to be present. Butene-2-C.P. grade: 99.0%, both "cis" and "trans" forms. Principal impurity was butene-l; 0.1% acetylenes and lesser amounts of diolefins and other hydrocarbons were present. Isobutylene-C.P. grade: 99.0%, butene-l was the most probable impurity. 1,3-butadiene-99.77% 1,3-butadiene. Impurities were normal butane, butenes, and possibly a trace of acetylenes (0.23%). Sulfur dioxide-C.P. grade: 99.988% S02, 0.002% moisture, 0.010% noncondensable gases. 20

Ealgineoring Rosearch Institute University of Michigan....................................... ----------- ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ j~::: ol~~~~~~~~~~~~~~~:::~~:: ------------------------------—:iij:~::~' ~::~'ii: ~::::-:.............~~iiii l~iiiii...........i3'ii.......................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:~:::::;:::::%: ~::-:-:-:::::-:-:" r~~~~~~~~~~~~~~~~~................ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ (\...............~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~t.............................................................. Hi............... xx...,~~~~~~~~~~~~~~~~~~~b;-`::::;(~:~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.....,~i~i iiiiili~ii:,ili~i:- li;:: iiliilii i~i~i~i:$~S~ijr~ji'i~: Ijll~j: ~~i~i'ijiiiji~i~j~i~ii~i';.............;~~~~::::::::::::::::::::,:::::::::::::: fj~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... ~~~~~~~~::-: ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~........ ~~~::~~~~:~~~~i:~~~~:~~~~~~:~~~::~~~~~:Z~~~~il:~sliir:5;S': ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~....... i~~:~'~~'i:~~~~l~~~.~:i::~~~~iiiX'~~:~ii~~i~i ~ ~ -:::::-:i~~~:;lii~~~~l:~~.~iP.~~:i l~.......... s~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....................::::::::::: E X.-i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:: ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~................................................................. ~~~~~~~~~~~~~~~~~~~~~~~~~~~ E~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.....;~~~~~~liiiiiiilliiiilliiiiiliilliiiiiliiliii ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~................ Sriil~~~~~~~~~~~~~~~liii~~~~~~~~~~~~iiiirirriiliiiiilii~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i ~s —1~~~~~~~~............... ~~~~~ 8~~~~~~~~B~~~~~~~~~'ji~~~~~~~~~~~~~~~~~~~~i ~ ~ ~ ~ ~ ~ ~ ~ f~~L~~~~~~~~~~~~~~~~....... ~~~~~~~~~~~~~2:-a-:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~............ ~-f:: ~ ~ ~ ~ ~~~~~~~~~~~~i:~~~~~~~'iiiiiiiiiiiiiiii; — c~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.............:~~~:::::::::::;:::::::,-r4 iii8~~~~~~ ~:'"~~''''"':''''~....................................:::I:,iii~~~~~~~~~~~~~~~~~~~~~~~~~i~:iiiiliiiij-~~~~~~~~~~~~~~~~~~~~~~~....................:~ii~,_: Iliiiiiiiii ~ ti..........''''''''''' ~:::::: ~r~~l~~~~~~~~~~~~~~~~.....-.1 ~~~:: —i~~~~iii:~~~:_i::~~~~~~~~~~~iii~~~iiii..........................................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:::::::::::............~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i:ii:-:ii::Iiiil ggiiii:~~~~~~~~~~~~~~~~~~~~~~~~~:~~~~~~~:iiiiiiliiiilliilliliiiiiil~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~............... ~~i~~:~~~'~~'~~~~"~~:"'~~~~:~~~~~i~~~:i~~:~~~:~~l~~~~~~iiiii~~~~: i.............................. ~~~~~i~~~~~~~~i~~~~~~~~~i~~~~~~~~i~~~~~~~~i iiii~~~~~~~~~~~~~......................iii...............iiii~iii~i.................................................................~~~~~~~~~~:iii~i................................~~~iiii6..........................~~~~~~~~~~~~~~~~~~:~i ii:i:~~~~~~~~~~~~~~~~~~~~~~~........................ I::::::::~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..................................................................... s 9i~~~~~~~~~.............................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i ~~~~~~~~~~~~~~~~~~~~:~-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i:- ~ X~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~................ 6a ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~.....................................~~~~~~~~~~~~~~::::::_: j~~~~~~~~~~~~~~~~~~~~~~~............................................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i.......................................................~~~~~~~~~~~~~~::::: i-:::~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.................. iiiiiiiiiiiiiii~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~........................................................-:; ~~ ~ ~~~"~~~~~i-l~~~~~iiiiiiiii:;~~~~~~~~~~~~~~:liiiiilii~~~~~~~~~~~~~~iiiiii' ~~~~....... ~~iiiiiiiiiii~~~~~~~~~~iiiiiiiii~~~~~~~~i~........................ ~~~~,:iiiiiii~~~~~~~~~~~~~~~~~~~~~ii:-iii~~~~~~~~~~~~~~~~~i~~~~~~i-i~~~~~~~~~i~~~~~i~~~~~i~~~~~iiiiiiii: ~~~~~ ~ ~ ~ ~....... ~~i~iii~iiiiiiiiiiiiiii~~~iiiii~iiiii~~iii~i se~~i~i:::i~~i:::::i:::::i:::::i~~~i:::::i~i:-:::i:::i:::i:-.......................... F~~:::_:::-:_:-:::::-:::::-:::::::_:-::.......... ~~,::::i~~~~~~i:::i:::i:::::i:::::i:::i~~~~~~~~~~~~~~~~~~~~i......... Sii~~~~~iicii~~~~~~ailiiiiil~~~~~~~~~~iiiiiiiiiiiii..............................~~~~~~~~~~~~~~~~~r2 t

Engineering Research Institute ~ University of Michigan Nonene and dodecene-500 ml were distilled under vacuum from researchgrade reagents supplied by Continental Oil Company. Styrene-500 ml were distilled under vacuum from stabilized reagentgrade styrene, rejecting initial 50 ml and final 150 ml. 3. Experimental Procedure.-In the preliminary experiments the standard experimental loading procedure set up for the butadiene-styrene polymerization was followed, with one exception. When two gaseous components were charged to a vial, as was the case in the olefins of lower molecular weight, it was thought better to measure the amounts of condensed vapor by volume rather than by weight. The weighing procedure had two serious drawbacks: (1) the lines filled with air and, hence, oxygen when the vial was disconnected from the system, and (2) the vial warmed up slightly during weighing and released S02 and olefin vapors into the room. The volumes were computed by comparison with a calibrated graduated vial and the weights were obtained from densities found in standard reference handbooks. A standard procedure was also followed in loading and handling the stainless-steel reactor used in the larger-scale experiment. To provide the bomb with a tight seal, a strip of lead was placed in the gasket groove of the cap and melted in it by means of the natural-gas-oxygen torch, while care was taken to have the head perfectly horizontal to provide a uniform thickness of lead. When cool, the head was screwed onto the body of the bomb and the threads were coated with leaded pipe dope to give additional assurance of seal. Nitrogen was then introduced up to a pressure of 800 psig and the bomb placed in water and inspected to detect possible escape of gas. After the pressure-testing operation, the nitrogen was vented and the tare weight of the bomb recorded. The bomb was connected to the gasloading apparatus and the entire system evacuated to less than l-mm-Hg pressure. The feed gas cylinder was opened and the desired amount of gas allowed to condense in the bomb. The valve of the bomb was closed and the bomb weighed. When a second gas had to be introduced, the loading procedure outlined above was repeated and the final weighing made. After exposure to radiation, the bomb was placed in the hood and its gaseous contents allowed to escape. The bomb was opened and the product was removed mechanically. The product was weighed and stored until meltingpoint measurements and sulfur-content analysis were made. 4. Experimental Results.-Table 8 presents the results of the copolymerization of various alkenes with sulfur dioxide in the preliminary experiments. Generally, much higher yields were obtained with these polymers than in the case of the butadiene-styrene polymers. In the case of the

TABLE 8. COPOLYMERIZATION OF VARIOUS ALKENES WITH SULFUR DIOXIDE AT 80 to 150 CENTIGRADE AND 155,000 TO 162,000 REP PER HOUR Run Total Weight Weight of Percent Dose Components on Run Charged Product A Value Remarks Yield Megarep Page gm gm ae (aP SO2 + 1,3-Butadiene (B) 145846 62 6.3 5.3 84.0 3.85 3,184.7 white-creamy-hard polymer-melts above 310OC-color changes to brown SO2 + 1,3-Butadiene (B) 14+5846 63 6.6 3.3 50.0 control - white crystalline SO2 + Butene-l 145847 64 3.8 2.5 65.8 3.85 2,771.8 white and gray polymer-porous-melting range: 2340-2590C SO2 + Butene-l 145847 65 1.9 0.4 21.0 control - SO2 + Butene-l 145849 66 5.0 2.1 42.0 5.56 1,21L7.0 white-some colorless polymer-melting range: 2320-2620C S02 + Butene-l 145849 67 4.1 0.2 4.9 control yellowish white film polymer (A S02 + Butene-2 154700 68 2.6 2.4 82.5 5.56- 2,404.5 colorless hard polymer-melting range: 2310-2600C T~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~C S02 + Butene-2 154700 69 5.0 4.3 86.o control colorless hard polymer S02 + Nonene 154705 71 11.0 trace C!0 3.48 0c S02 + Nonene 154705 70. 4.4 0.0 1i 0 control - (A S02 + Dodecene 154706 81 3.7 0.3 8.1 3.68 2,065.3 golden brown liquid-black precipitate S02 + Dodecene 154706 80 4.4 trace " 0 control - S02 + Styrene 154707 91 5.9 2.0 34.0 3.48 1,258.2 fibrous white polymer-melting range: 2480-2840C + Styrene 154707 90 5.7 2.0 35.0 control hard white polymer (. S02 + i-butylene 154708 101 6.1 0.3 4.92 3.69 218.7 spongy white polymer S02 + i-butylene 154708 102 5.8 <0.1 <1.7 control - fine white polymer

Engineering Research Institute * University of Michigan 1,3-butadiene, styrene, butene-l and butene-2-,sulfur dioxide copolymerizations, the percent yield of the controls was also high, though consistently less than those of the irradiated samples. This relatively high yield in the control samples was believed due to the fact that the control vials were Stored in the light during the time that the other vials were being irradiated. The literature28 reports that sunlight or light in the region of 3000 to 3800 A catalyzes the reaction. The yields of polymer in the larger-scale control experiments conducted in the absence of light in stainless-steel vessels were negligible. Dodecene, nonene, and l-butylene —sulfur dioxide copolymerizations generally gave poor results. The products of the alkene-sulfur dioxide copolymerizations from the preliminary vial experiments are described below. A more accurate evaluation of the effects of gamma radiation upon the reaction is shown later in the discussion of the larger-volume experiments. Butene-2 —sulfur dioxide: The irradiated sample gave a transparent product which has the appearance of mica; the only visible difference between the control and the irradiated sample was that the control product was colorless. The inside wall of the vials was covered by a thin, brittle layer of product which was white in the case of the irradiated sample and colorless in the case of the control. Butene-l —sulfur dioxide: Both'the control and the irradiated samples gave a white-gray, crystalline, and porous product. 1,3-butadiene (Grade B) —sulfur dioxide: The irradiated sample gave a white product which turned brown on standing; the control was a pure-white crystalline substance. Styrene-sulfur dioxide: A yellowish product with a fibrous appearance was obtained. The control gave a product which had an initial elasticity but became hard on standing, due to further polymerization of the excess styrene. Table 9 presents the results of the large-scale experiments carried on with mixtures of various alkenes and sulfur dioxide. Generally, much smaller doses were used in these experiments than in the preliminary ones. The reactions with nonene and dodecene were not investigated further. The reaction with l-butylene and styrene gave comparatively mediocre results and their further study was relegated to a later time. Propylene-sulfur dioxide copolymerization was found to be quite successful. Most reactions were carried on at ambient room temperature. A few low-temperature runs were made in an attempt to determine the possible influence of temperature on the yield or physical aspect of the product. As yet, these reactions have not been studied at high temperatures.

TABL;E 9. LARGE-SCAL;E COPOLYMERIZATION OF VARIOUS ALKENES WITE SULFUR DIOXIDE Run Temp. of Total Weight Weight of MelIting -Point Percent Mole cular Alkene Used on R Moles Charged hre rdc au ng 0 i rnemt Run Charged ~~~~Percent Dose Rate Ds Orgni SO2 iel Kep/r reg"CProduct ose ~rA Valuen Page g Butene-2 154801 ~-74 0.36 0.41 4 6 76 38.3 39.4 489 12,900 Chars and boils at 244 54.1 1:1 (D Butene-2 154749 20 0.07 0.31 24 16.5 62.8 40.6 2,540 4,190 240-316 51.:l Butene-2 154815 28 0.55 0.31 51 1 3 2.6 0 0 Melts and decomposes at 268 49.5 1:1 $P Butene -2 154823 32 1.30 0.0 73 N0 0 68.8 4, 510 0 $O Butene-1 154802 20 0.38 0.29 40 0.3 0.75 32.8 541 232 = Buterie-1 154804 -75 0.13 0.30 26 0.6 2,~1 6.3 225 1,340 ( Butene-1 154808 -75 0.18 0.36 33 25.8 78.2 25.7 575 22,000 52.2.1:1. Butene-1 154814 26 0.21 O. 40 37 O. 1 O. 3 0 0 (O Bultene-1 154822 27 1.77 0.0 39 0.4 1.0 68.8 4,800 14.9 u 1,3-Butadiene 154810 21 0.54 1.08 98 73.0 74.5 29.0 498 24,600 292 54.0 1:1a 1,3-Butadiene 154811 -75 0 44 0.80 75 45.3 60 5 14 0 258 38,200 305, chars 53.'9 1:1' 1,3-Butadiene 154813 30 0.41 0.94 82 21.0 25.6 0 0 - Sublimes 5S2.8 1:1 r 1,3-Butadiene 154825 28 1.54 0.0 83 1.7 2.1 68.8 4,366 84.8 i -Butylene 154816 26 0.48 0. 53 61 2! O 0 2,850 ou i -Butylene 154824 31 1.94 0.0 109 "O 0 ~ 68.8 4,760 SO >312 ~ kD Syrn 154819 28 0.34 1.22 ll3 31.0 27.4 60.2 3,050 1,175 Chars at 150 None 154826 27 0.0 1.16 74 "0 ~0 68.8 2,945 ~0 >316 None 154827 29 0.0 1.72 ll0 O ~0 ~ 60.9 5,200 ~0 m316 Propylene 154817 26 0.88 0. 48 68 ~O ~ 0 0 0 - >316 Pro-pylene 154818 27 0.93 0.62 79 7i15 90.5 68.8 2,713 6,280 283 59.0 1:1:~ Propylene 154820 -75 0.76 0.56 68 59.0 86.8 24.0 2,073 7,830 273-316, chars 58.5 1:1-' Propylene 154821 27 1 5 0 0 65 0 6 0.9 51.2 3 095 71.4 Chars above 148 Propylene 154828 29 1.19 0.50 82 44.8 54.7 614 1,080 10,000 271-decomposes at 311 59 6 1:1 - Propylene 154829 25 1.07 0.95 106 91.1 85.9 61.8 6234,000 287 66.3 5:4 V Propylene 154830 26 1.54110 135 1.23646.5 63.4 69,700 284 58.2 1:1 - Propylene 154833 27 1.18 1.23 128 3.9 3.0 68.8 32.1 17,080 Chars below 167 63.2 1:1 2 Propylene 154834 27 2.28 1.97 223 90.0 40.4 68.8 180 41,220 285 5. 1:1 O Propylene 154837 28 1.07 0.84 99 87.3 88.1 68.8 1,021 15,990 58.4 1:1 - 9 Pro-pylene 154839 19 a. 53 0.69 108.4 ~r 0 ~ 68.8 399 0 Propylene 154840 25 1.8 1.12 116.8 54.8 47.0 68.8 375 22,580 59.6 1:1 ~ Propylene 154841 20 1.41 1.42 11.140.0 92.5 68.8 719 23,240 /~ Propylene 154842 17 1.13 1.23 126.5 106.9 84.0 68.8 542 27,800 ~8.2 1:1' Propylene 154115 14 1.40 1.6 62.6 118.9 73.1 68.8 0 632,100 Propylene 154843 18 2.63 0.0 110.5 <.4 c.35 68.8 5,754 22 Ethylene 145817 25 2.36 0.061 70 1 5121.6 37.8 3,720 2,170 Decomposed 24.7 2:19 Ethylene 145824 25.48 0.50 45 1 33.9 75.0 6.5 4,000 3,770 Decomposed 62.3 2:3 Ethylene 145826 100 0.97 0.97 89.2 57.0 64.0 37.0 3,630 3,680 316 65.1 1:1 Ethylene 145827 212.5 0.06 21.8 4.3 19.7 37.0 3,360 1,705

Engineering Research Institute ~ University of Michigan For each set of reactions, except for the styrene-sulfur dioxide copolymerization, a control was made for the alkene alone and for the alkene with sulfur dioxide. All but the 1,3-butadiene —sulfur dioxide control gave small yields. The melting point of the controls sometimes showed marked differences when compared to the irradiated samples. This was due to the higher range of molecular weights found in the polymers produced under gamma radiation. Two runs were made with sulfur dioxide alone, both under exposure to radiation. A brief description of the characteristics of the polymers formed in the larger-volume experiments and significant observations made during the runs are indicated below. Run numbers refer to specific runs presented in Table 9. 1-butylene-sulfur dioxide copolymerization gave no result in spite of the comparatively larger doses used. This was due to the fact that at room temperature we are above the "ceiling" temperature for the reaction. Styrene-sulfur dioxide copolymerization gave a low radiation yield and has not been investigated beyond the one run reported. Butene-2 —sulfur dioxide copolymerization data tend to show the characteristic exponential curve of the percent yield vs dose; it was felt, however, that the number of data obtained were not sufficient to plot a reasonably accurate curve. The compact, homogeneous, and transparent solids obtained with the butene-2 in the preliminary investigations were not obtained in the largerscale runs, except to some extent in Runs 154749 and 154801. Because the products were caked in the bottom of the bomb, they had to be broken up before they could be extracted. This handling, together with the fact that the products of the bomb polymerization were not allowed to stand for long periods of time, as was the case in the preliminary experiments, probably accounts for this lack of reproducibility. The melting-point range of the butene-2 —sulfur dioxide copolymers was between 2400 and 316~C. The average percent sulfur dioxide of the butene-2 product was found to be 51.6%, which indicated a one-to-one molecular arrangement. Butene-l-sulfur dioxide copolymerizations yielded white, hard powders; the percent sulfur dioxide calculated on the analysis of Run 154808 was 52.2% and indicated a molecular proportion identical to that of the butene-2 product. 1,3-butadiene (Grade B)-sulfur dioxide copolymerization yielded a

Engineering Research Institute ~ University of Michigan white product having the appearance of chalk. The product was very dry and difficult to grind by ordinary means. The lower limit of the melting-point range was in the vicinity of 292~ to 305~C. The product charred before the upper limit was attained. The sulfur dioxide content was found to average 53%, which also indicated a one-to-one molecular arrangement. Propylene-sulfur dioxide copolymerization was by far the most successful mixture. Two outstanding features were: (1) the importance of the relative amounts charged and (2) the sharp rise of the percent yield for a comparatively small increase in radiation. When the propylene to sulfur dioxide ratio of moles charged was close to 1, much higher A values were obtained than with propylene-rich mixtures. For instance, in Run 154818 6,280 kilorep were needed for a 90.5% yield with.927 mole of propylene to.624 mole of sulfur dioxide charged, while only 719 kilorep sufficed to obtain a 92.5% yield with the 1.41 moles of propylene and 1.42 moles of sulfur dioxide mixture in Run 154841. The other feature was best brought out by the graphical representation in Fig. 9, indicating that the percent yield rose from zero to 90% within a dose of 700 kilorep. 90 1 0 i 9~~~~~~0 80 70 60 w 50,, >- 0 40 30 0 20 0.2 4.6.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 RADIATION DOSE, MEGAREP Fig. 9. Propylene-sulfur dioxide copolymerization at room temperature. 27

Engineering Research Institute ~ University of Michigan One point (Run 154828) was considerably off the curve. This was believed to be due to the initial composition of the reaction mixture. The reaction probably reached completion on a sulfur dioxide basis, leaving unreacted propylene, and this would tend to indicate a low overall completion of reaction. In some cases very low' doses were used in attempts to determine the existence or nonexistence of an induction period; however, with a dose as low as 32 kilorep, some product was obtained. This indicated very little induction period to the reaction. The product obtained in all cases was a hard and dry white substance which could be ground to a fine powder. In all cases but one, the lower limit of the melting-point range was in the vicinity of 280'C and the upper limit at 310~C or higher. Charring usually occurred before any fluidity could be detected in the sample to be melted. The sulfur dioxide analysis showed that on the average the products contained 59.9% sulfur dioxide, from which the molecular arrangement was deduced to be one to one, as previously indicated for other alkene-sulfur dioxide copolymers. Ethylene-sulfur dioxide copolymerization: Analyses of the products were made on the two runs reported in Progress Report 74 and these are included in Table 8 with two runs made since that time. The product was a finely divided white powder. The yields may be somewhat obscured by the fact that polyethylene may have been formed in the runs. This was thought to be the case in Run 154817, where the ethylene-sulfur dioxide ratio is higher than ordinarily expected. Run 145827 was above the "ceiling" temperature for ethylene and sulfur dioxide, but polyethylene may still be formed. In general, the polymer decomposed before melting. This is a common phenomenon associated with all polysulfones of this type. 5. Evaluation of the Product.-Product yields are presented as the A value, as is customary for this work. A complete analysis of this quantity is to be found in the literature.l7 The A value and the molecular-weight calculations were the same as presented for the butadiene-styrene experiments. Melting points were taken on the laboratory melting-point bar described previously. Sulfur analyses were made in the laboratory by standard methods. The procedure followed was that outlined in the Parr Booklet 1048 for the analysis of sulfur in vulcanized rubber. For the products at hand, it was found more convenient to increase slightly the amount of potassium chlorate (0.6 gm in lieu of 0.5 gmin) and replace the 0.5 gm of sugar by 0.7 gm of benzoic acid. Approximately 0.2 gm of the finely ground product was weighed and thoroughly 28

Engineering Research Institute ~ University of Michigan mixed in the Parr bomb with the above-mentioned amounts of benzoic acid and potassium chlorate, and 10 gm of sodium peroxide. The mixture was then covered with an additional 4 gm of sodium peroxide and the bomb was sealed and fired. The contents were dissolved in warm distilled water and processed for further analysis as outlined in Willard and Furman. 32 The procedure consisted in making the solution acidic with hydrochloric acid, neutralizing with ammonium hydroxide, and boiling for 30 minutes after addition of 15 gm of ammonium carbonate. The precipitate was filtered and rinsed with a dilute sodium carbonate solution. The filtrate was then made acidic with hydrochloric acid and boiled free of carbon dioxide. After an addition of 5 cc of hydrochloric acid in excess, the solution was diluted to 400 cc and barium sulfate was precipitated by slowly adding 20 cc of a 10% solution of barium chloride. After checking for complete precipitation, the solution was kept warm for an hour. The precipitate was then filtered and washed with distilled water, ignited and weighed, and the last two steps were repeated until a constant weight was obtained. From the weight percents of organic constituents and sulfur dioxide in the products, a molecular composition was calculated. These values are presented in Table 9. In nearly all instances the molecular ratio was one to one. This corresponded to the molecular compositions found in the literature for the polysulfone products produced with conventional catalysts. 21 D. OXO-TYPE REACTION Since World War II the Oxo reaction has been of some interest in the United States. This hydroformulation reaction, as it is sometimes called, consists of the addition of carbon monoxide and hydrogen to an olefin, producing an aldehyde of one carbon atom greater than the original olefin. In general, a mixture of aldehydes are formed. The equations for the reactions involved are: 0 R-CH = CH2 + CO + H2 T, P catalyst) R-CH2-CH2-C// 0 H R-CH-C/ CH3 H The reaction has been investigated under a wide range of external conditions. 29

Engineering Research Institute ~ University of Michigan The present commercial industrial operation takes place at a temperature of from 1200 to 190'C at from 100 to 250 atmospheres of pressure in the presence of a cobalt carbonyl catalyst. 25 A complete review of the history, conditions, and operation of the reactions is presented by Orchin and Schroeder. 23 Various mechanisms have been suggested for the reaction. Evidence has been presented for cyclopropanone, carbonium-ion, hydrogenation, and freeradical type mechanisms. 2 It is this last mechanism that apparently is involved in the use of gamma radiation to promote chemical reactions, and for this reason it was considered of interest to study the promotion of the Oxo reactions under gamma radiation. Ethylene, butene-l, butene-2, propylene, and isobutylene were subjected to Oxo conditions under the influence of gamma radiation from the cobalt-60 source. No cobalt carbonyl catalyst was used in the reacting mixture. Different temperatures, pressures, and radiation doses were studied with varying success. The data that are reported are to be considered preliminary in nature,as no effective scheme was followed in varying the conditions. It was deemed advisable first to obtain a product and prove the reaction feasible before systematically controlling the variables present. 1. Equipment Used. —Several reactors were utilized in the study of the Oxo reaction in the laboratory. The reactor designed by Lewis was used in the case of the ethylene and butene-l Oxo reactions. Here, gas-phase reactions were studied at moderate pressures and the reactor was well suited to this work. In the high-temperature work involving higher pressures, the American Instrument Company reactor presented by the Engineering Research Institute was used. This reactor has been described in the section on polymerization of ethylene and is shown in Fig. 2 of this report. It has a larger volume and is more capable of withstanding high pressures at high temperature than is the Lewis reactor. In those systems where two phases were present in the reacting mixture, an Autoclave Engineer's "Magna-Dash" reactor with a variable magnetic stirrer was utilized. This reactor was purchased by the Engineering Research Institute for high-pressure two-phase work in this laboratory. The two-phase reacting system consisted of the liquid olefin and gaseous carbon monoxide and hydrogen at temperatures below the critical temperature of the olefin. The magnetic stirring mechanism provided violent agitation in the vessel, allowing greater surface contact between the two phases. Figure 10 is a photograph of the disassembled Magna-Dash reactor. The reactor was the standard Autoclave Engineer's Magna-Dash licensed by Standard Oil of Indiana. The reactor was constructed of 18-8 stainless steel with a capacity of 500-ml. The stirring mechanism consisted of an external solenoid acting upon an internal shaft to which the stirring plates were attached. There were no packing glands or 30

Engineerng Reseach Instiute Un vrsity ofM'chiga......................................................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r~......................................... -'i -S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~........................................ X.........................................~~~~~iiii~~~~~~Wii...... X...........................:. ~-I..,................................. ~ ~ ~ ~ ~ iiiiiii~::i-i i~-i::i::i:i:-.............................. X..............................~~ ~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~C~ i~~~~~~iiiiiiiiiii ~ ~ ~.........................................................~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~jj...............................................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~: 44....................:::: c ~:j~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...................................... ~"~~~ii'~~~ -D~~)~~....................................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r.................... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~Q......................... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ "~...............................................................................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~l~ X~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:-~::ii~~i~~_i~-: —: —:-::............~~~~~~~~~~~ X.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'."iiiii~iiiiiiii.............. F - _Fiijii~:iiiiiiillii'~;':iiii'~iijii t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.......... X,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~iilliliillii-iliii i...................i:-:::: —-:::i.......................................... ~ ~ ~ ~ 2.......... iii ii~l ~ i~ili P~~..................~ ~~ ~~ ~~ ~~ ~~ ~ ~~ ~~ ~ ~~ ~ ~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'iiiiiiiiiiiiiiiii:~-: ~~~~ —---— ~~~~~~~~iiiiiii~~~~~~iiiiiiiiiiiiiiiiiilii 3~~~~~~~...... I~~~~~~~~i~~~~~~~~iii ~~~~~~~~~~~~~~~~~''~~~...................i ~~~~~~i i~~~~~~~i~~~~i~~~~~~~~i i~~~~~~. ~~~~~~~~~~~ j~~~~~~~~~~,~~~~i~~~~~:~~~~~. i~~~~~~i~~~~~i:~~~~~~i:~~~~........ ~~~i~~~i~~~:j~:~~~~~~~i'.i~~~~~~~i: Ei~~~~~~~i~~~~iii;~~~~~~~i~~~i~~................. ~~ i~i~t i ~ l ~i ~I ~ _~:::: i~~I ~~: ~~: ~ I::................................................. ~ ~ ~ ~ ~ ~ ~ I~~................................................................................................ -j X lilxiiiiia~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... Ili'iiril~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~........... s~~~~~j:' Z~~~~~~~~~:j-:~~~~~~~~~I::~~~~~~~~:':::~~~~~~~~~~::~~~~~~i~........... cc:~:a:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....... ~~~~;~:::~~~~~~~~~~~~~::::::~~~~~~~~~~~~~~~~:::::jj::::j::::::~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:::~;:~~......... 31.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......

........Engineering Research Institute e University of Michign............ ext-e~rn-,.] ports,.: whli.ch. might a.]owl ras ]akagfe.'The reactor v.tas eq~uippe;(.d with amountb of agi.,tation. 9l.e operating, p~ressurce range w.as from 0 -to'jOO0 ps::.. T1he body, gasket head, stixrring.. m-echanismij,,and ad.jus'tab].e timfer,are sholm in Fig. ].O. T-he stri~p heaters and stband are not shown, as <armiy war surp].us. The vesselhaid a c a-acit y o[8):Liters and wspeur rated at'pOO psi. The reaction was run in the vapor phase at a ].ow pressuren at,), __ow prssur at room tmprtue The tank h)ad two "~/8 —inlchlp:[.pe o pnnsfor ].onad-n.-_, ng)and discharge. Only react~i.ons which went fromn w-por-phase reactants'Lo..:, qu-id-.. phase prod~~u ct e ecnutdin the tank 1, sice, lel-artngJ1! _Was othe~rwil-se, a uct rem-toval, and c]leankng of'the vessel,..[The tank:is show-n in F'iij ]..................:.:::::.:............. ~:,::::::::::::,:::,.............::.................,,,..................:........:...............................................................::::,:,:..................:,.:::;..............,::::::::::::.:::...............:.,,,,.,,::::::,,::...............:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....................I....... I............I...,......................... I.......................... and~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~......aiu...e.........".h na.ye o...ecro onx[ adhdrgna ea c.bon,, o ox.... 97 co, o co2, o o2, oyo']"~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....................:tyc: g.~~~~~~~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.......... ~~~~~~~~~~~~~~~~~.............. I' l.............. 5

Engineering Research Institute ~ University of Michigan The percent oxygen in the carbon monoxide is higher than desired for this work because oxygen may have an inhibiting effect upon the reaction; however, this was the best carbon monoxide obtainable in large quantities at the time. In some cases, the poor experimental results were thought to be due to the presence of this oxygen inhibitor. 3. Experimental Procedure. —The same experimental procedure was followed in all cases for all reactions in all reactors. The reactors were cleaned with various solvents, dried, and sealed for pressure testing. Nitrogen was used in the pressure-testing operation and was left in the reactors at a specific pressure for at least 12 hours. After venting the nitrogen, the vessels were evacuated to less than l-mm Hg before introduction of the reactants. The olefin was charged first by immersing a portion of the reactor in a -75~C cold bath of CC14 and CHC13. A specific amount of olefin was added and determined either by direct pressure measurement, weighing, or by heating to above the critical temperature and making a pressure measurement. The carbon monoxide and hydrogen were then charged separately to specific predetermined pressures to give approximately a one-to-one mole ratio of CO to H2. The exact proportions of all components were then calculated from standard tables of thermodynamic properties. The reactor was brought to the desired reaction temperature in the radiation-source room and the source was raised into irradiation position for a specific time. On completion of the run, the bomb was removed from the radiation and allowed to cool. Duplicate Orsat analyses were taken of the gases remaining in the reactor. An analysis was made for the olefin and carbon monoxide present and hydrogen was taken by difference. The rest of the gases were vented to the hood and the reaction vessel was opened for product recovery. The reactor was then cleaned for the next run while the product was stored for further analysis. 4. Experimental results. —In general, poor results were obtained from the Oxo reaction with most olefin reactants investigated. Table 10 is a summary of all reaction conditions and product yields obtained. The yields were so low and, in general, the doses were high enough that the calculation of A values was not warranted. The total dose and yield are presented, showing the relative values leading to this calculation. In the Orsat analyses made on the reactor off gas, no account was taken of the possible alkane formation by hydrogenation of the alkene. It is not known whether or not this can account for the low product yields. Another effect thought,but not proved, to be present is the oxygen inhibition of the reaction due to oxygen in the carbon monoxide feed stock. Harmer14 showed that oxygen inhibits other reactions in the presence of gamma radiation. Oxygen seems to inhibit the ethylene polymerization reaction, and the Oxo reaction is quite similar in many respects. 33

TABLE 10. REACTION CONDITIONS AND PRODUCT YIELDS FOR THE OXO REACTION INVOLVING VARIOUS OLEFIN REACTANTS Average Average Mole Radiation Total Product rn Page No. Temp. Press. Percent Charged Dose Rate Dose Yield Product Description Used ~C Psig Organic CO H2 Krep/Hr Megarep gm ( 154001 7 1120 ethylene 45 29 26 30.5 2.39 1.133 mp 1370C, light, fluffy, powder polymer 0 154003 143 1430 ethylene 40 31 29 29.3 1.89 o.o66 waxy, brown polymer 154005 17 1065 ethylene 56 21 23 37.0 3.43 2.071 mp 1140C, light, fluffy, powder polymer 0 154008 58 1375 ethylene 55 12 33 19.5 1.24 0.50 light, fluffy, powder polymer Q 154010 108 900 butene-l 11 44 45 33.5 2.85 sm. amt dark-brown, waxy polymer 0.1 154013 189 1235 butene-l 24 38 38 31.5 3.87 sm. amt tan, viscous, oily liquid C 154015 30 690 i-butylene 86 2 12 17.0 3.09 sm. amt brown, viscous, oily liquid %0.2 154018 181 2260 i-butylene 68 17 15 11.0 2.16 19.1 yellow liquid, pungent odor 154029 20 718 i-butylene 51 29 20 10.8 5.53 0.2 tan, oily liquid 0 154020 30 910 propylene 70 16 14 16.3 3.12 2.8 brown, viscous, oily liquid 154037 14 546 propylene 77 14 9 16.3 7.58 9.7 clear, tan, viscous wax-solid " 154027 190 1960 butene-2 49 27 24 11.1 4.74 14.0 clear, yellow liquid, pungent odor 154030 231 2020 butene-2 78 14 8 0 0 13.5 clear, yellow liquid, low viscosity, pungent odor Q 154036 14 180 butene-2 14 44 42 36.1 13.4 2.0 dark-brown liquid, low viscosity lI

Engineering Research Institute ~ University of Michigan Visual appearances of the products are included in Table 10. In general, the product yields were so low that this was the only measurement made. The small quantities of polymer and oil product were not sufficient for analyses of most physical properties. Melting points were taken where sufficient product was available. These are presented on Table 9 under product description. Melting points were taken on the laboratory melting-point bar as has been described previously. As is to be seen on Table 10, the majority of runs yielded polymer products. Solids, waxes, and heavy oily liquids were obtained in most cases rather than the expected liquid aldehydes. In general, these products were obtained in rather low yields. In three cases, 154018, 154027, and 154030, light, low-viscosity, pungent liquids were obtained in appreciable quantity. These products are being analyzed at the present time. Distillation curves have been run on Run 154018 and a variety of products are assumed to be present. Aldehydes, alcohols, and condensation products of the aldehydes are thought to be present from observation of the boiling-point curve. These data will be presented at a future time when complete analyses have been made. In the case of high product yields in runs with butene-2, it should be noticed that Run 145030 was a control run with no radiation. The product obtained was the same as that from Run 145027, where radiation was present. This leaves doubt of the effect of radiation upon the reaction. The reaction is to be investigated further to eliminate this question. E. CHLORINATION OF SOME AROMATIC HYDROCARBONS In general, chlorination of an aromatic hydrocarbon may result in three possible reactions: (1) substitution of chlorine for hydrogen in the benzene ring, (2) substitution of chlorine for hydrogen in an alkyl group attached to the ring, and (3) addition of chlorine to the ring. Friedel-Crafts catalysts promote substitution in the ring; high temperature in the absence of a catalyst favors substitution in the alkyl group, and radiation appears to favor addition to the ring. Ultraviolet,ll'12 roentgen rays,19 alpha radiation, and gamma radiationl'1 have been employed to promote the addition reaction. In this laboratory, Harmer14 initiated a study of the chlorination of benzene and toluene in the presence of gamma radiation. This has been extended to include xylene, mesitylene, ethyl benzene, and naphthalene. 1. Equipment Used. —In general, the glass reaction equipment designed and built by Harmerl4 was used for all runs. Some difficulty was encountered with breakage of the chlorine-flow tubes in the part between the second-floor analysis laboratory and the first-floor radiation cave where the reactor was located. The removal, repair, and replacement of this glass

Engineering Research Institute ~ University of Michigan tubing took a great deal of time, since the port to the radiation chamber contained a right-angle bend in the center to insure no radiation hazard to the workers in the second-floor laboratory from the radiation source below. It was found that Teflon tubing was a suitable replacement for this glass tubing through the port and the angle bend in the port. Teflon becomes brittle under gamma radiation, and so a Teflon-to-glass seal was made in the radiation cave. Glass tubing was still used where the lines came in contact with radiation. The Teflon-to-glass tubing seal is a standard shrink fit of plastic to glass covered by De Khotinsky cement at the junction to insure a full vacuum seal. During the runs it was noticed that the inlet chlorine rotameter float had a tendency to oscillate through a small range. The cause of this behavior was attributed to the bubbling of the chlorine gas through the reaction liquid in the reactor. Attempts to eliminate this oscillation were only partially successful. Dry traps were inserted in the flow lines before and after the rotameter to cushion the gas flow' and pressure differential due to bubbling. The cooling-system tanks were enlarged and provided a much more accurate temperature control. A mixture of methyl alcohol and water was used as the primary coolant around the reactor. An ether, chloroform, methyl alcohol, and dry-ice mixture was used as the secondary coolant in the low-temperature tank. The temperature was brought to any predetermined run value within a few minutes, allowing initial analyses samples to be taken sooner than was possible previously. A control valve was installed between the vacuum pump and the equipment proper to control the rate of evacuation. It was found that too rapid evacuation would collapse the rubber connections used in the lines and, consequently, reactor contents would be drawn into the lines and traps before start of the run. Previously it was necessary to have an observer in the source room at all times during the evacuation of the lines preceding a run. 2. Chemicals Used in the Reactions. —The analysis of the chemicals used in the reactions is presented here as an aid to the reader. Chlorine: 99.8% C12, trace of chloroform, hexachloroethane, carbon tetrachloride, air, and tetrachloroethylene Benzene: Reagent grade, distilled in the laboratory Toluene: Reagent grade, distilled in the laboratory Ethyl benzene: Reagent grade, distilled in the laboratory Mesitylene: Reagent grade, as obtained

Engineering Research Institute ~ University of Michigan Naphthalene: Reagent grade, as obtained recrystallized from ethyl alcohol Nitrogen: Prepurified, 99.9% N2, no H20,.001% 02,.001% H2, remaining inert. 3. Experimental Procedure.-A standard experimental procedure devised by Harmer14 was followed in all runs. Before the run, the reactor and all auxiliary glassware were washed, cleaned for several hours in hot chromic acid cleaning solution, steamed out in a specially constructed tower, and dried thoroughly at 1500C on glass in an oven,. The lead lines were washed with acetone between runs. The liquid reactants to be used were placed in the reactor and the head sealed on with De Khotinsky cement and the whole assembly placed in the steel cooling jacket in position in the source room. The lines to the second-floor analysis laboratory were connected and the run was ready to start. The whole system was successively evacuated and refilled with nitrogen three times to displace air and, hence, oxygen present. Then temperature control and chlorine flow were started and the cobalt-60 source was raised into the radiation chamber. The reaction time was measured from the moment the source was raised and to the midpoint of the period during which a sample of gas was collected. Radiation dosages were determined by the time and dosimetry calibrations mentioned previously. Rates of reaction between chlorine and the aromatic hydrocarbons were determined by chlorine balance. The amount of chlorine into the system was measured with a rotameter. Excess chlorine was passed through the reactor at all times to maintain saturated conditions. At certain intervals samples of the exit gas were absorbed in a solution of 0.2 Normal sodium arsenite and 3.0 Normal potassium hydroxide. The amount of free chlorine was determined by titrating the excess arsenite with ceric sulfate and total chloride ion was measured by the Volhard method. In this way the amount of substitution and addition could be determined. The complete analysis procedure is described by Harmer. 14 4. Experimental Results.-The results of the chlorination of xylene, mesitylene, ethyl benzene, and naphthalene are presented in Table 11, with the temperature and radiation conditions involved. In general, individual products were not isolated, but kinetic data were taken on an overall chlorine balance on the system. The data on the chlorination of benzene and toluene have appeared in previous progress reports14'17 and are presented in Table 11 for comparison purposes. Benzene, toluene, xylene, and mesitylene reacted vigorously with chlorine in the presence of gamma radiation. Addition of six chlorine atoms 37

Engineering Research Institute ~ University of Michigan TABLE 11. CHLORINATION OF AROMATIC COMPOUNDS WITH GAMMA RADIATION Run Aromatic Temp. Dose C12 Add. C12 Sub. Average Average on Chlorin- Reaction Rate moles/ moles/ G(b) G(b) SubPage ated 0C Krep/Hr liter liter Addition stitution 129767 30% Benzene 20 61.0 35.20 --- 234,000 --- 129770 20% Benzene 20 61.0 32.60 --- 252,000 --- 129771 10% Benzene 20 61.o 33.50 --- 90,000 129774 30% Benzene -10 61.0 13.70 --- 90,000 129773 10% Benzene -10 61.0 32.70 --- 57,000 150450 Toluene 20 14.0 17.42 7.54 523,000 226,000 150463 Toluene - 5 14.0 14.12 6.59 900,000 420,000 150456 Toluene 35 14.0 10.35 6.12 812,000 480,000 150479 Toluene 20 None 3.23 0.85 154561 Xylene 20 13.7 5.90 3.48 462,000 272,000 154570 Xylene - 5 13.7 7.45 2.94 448,000 177,000 154573 Xylene 20 None 3.20 2.63 154590 Xylene - 5 None 0.41 0.175 154270 Mesitylene - 5 12.5 2.86 12.0 380,000 945,000 154585 Mesitylene 20 13.7 4.20 15.10 210,000 827,000 154558 Ethyl Benzene 20 13.7 1.81 0.47 137,000 35,000 154555 Ethyl Benzene 20 None 1.61 0.13 5 — 154276 Ethyl Benzene - 5 12.5 2.22 0.233 360,000 42,600 154595 Ethyl Benzene - 5 None 0.61 1.71 --- 154250 Naphthalene 21 None --- 0.415(a) --- -- 154259 Naphthalene 20 12.5 1.27(a) 0.815(a) 893,000 647,000 154264 Naphthalene - 5 12.5 1.48(a) 0.615(a) 906,000 273,000 (a) Based on a 10% naphthalene, 90% carbon tetrachloride solution by weight. molecules reacted (b) G molecules reacted, based on chlorine. 100 e.v. absorbed to the ring and alkyl substitution appeared to be the primary reactions. The ratio of addition to substitution was different than that obtained in chlorination reactions without radiation. In general, this ratio appeared to be increased by radiation and decreased by increasing the temperature. The radiation yield is reported as G value (molecules chlorine reacted per 100 e.v. of radiation absorbed) in Table 11. For some reason radiation had little effect on the reaction with ethyl benzene. Whether this is due to inherent inactivity or to inhibiting impurities is not known. Harmerl4 demonstrated the inhibiting 38

Engineering Research Institute ~ University of Michigan properties of small amounts of benzyl chloride and oxygen in the chlorination of benzene and toluene. Any alkyl substitution product may have a similar inhibiting effect upon the reaction. The product formed in the chlorination of napthalene has not been determined as yet. It is not known whether the compound produced is wholly or partially saturated by the chlorination. Further work will be done on the reaction and product analysis in the future. The G values of addition for the naphthalene reaction based on chlorine are high; the G values based on naphthalene would be lower, depending on the amount of saturation to the rings. For example, if the product from Run 154264 was wholly saturated, the G value for addition based on the naphthalene would be: 906,000/5 = 181,200, where 5 is the mole ratio of chlorine to naphthalene involved in the reaction. Similarly, the G values for substitution will be lower if based on the aromatic compound rather than the chlorine. Radiation chlorination of benzene took place so quickly that temperature control was difficult. Also the production of the solid addition product tended to plug up the reactor unless the benzene was diluted with carbon tetrachloride. Toluene, xylene, and mesitylene did not react quite so fast and were charged to the reactor in a pure condition. Naphthalene was dissolved in carbon tetrachloride to give a liquid-phase reaction. Some solid product precipitated in the reactor during the run and the carbon tetrachloride provided a slurry medium to prevent plugging the reactor tubes. However, in all cases, reaction under gamma radiation takes place so rapidly as to make it very attractive for commercial application. This is particularly true where the addition product might be desired, for radiation favors the addition reaction. 5. Equipment Changes. —At the termination of the series of runs presented above, it was necessary to dismantle the rack for cleaning and repair. In the course of this dismantling, it was decided to redesign the rack and cooling system to give a wider versatility of operation and a greater ease of operation. In general, the same schematic system employed by Harmerl4 was followed. The changes mentioned previously were incorporated into the system more economically. Other aids in operation were employed. Figure 12 is the schematic diagram of the redesigned chlorination equipment. A photograph of the completed rack is to be found in Fig. 13. Essentially the changes involved were: (1) location of all cylinders on the rack proper, (2) central location of all stop cocks, (3) central location of measuring devices, (4) fewer rubber connections, (5) more systematized flow pattern on the rack, and (6) incorporation of new equipment. The N2 cylinder to purge the system and the 02 cylinder to cut off 39

THERMOCOUPLE TO 0 POTENTIOMETER LINEINLET H12 E2PLE LINE HEATER.. ONTRON i~~~~~~~~TEMCUL SAMLE LI 2 41 \-1CRULAN WEL / LTNE TRAP CONTINUOUS I COOLANT PROL VAC. MANOMETER RE~~~~~~~~~~~~ANOMT ER- POTENTIOMETER LIN PU~~~~~~~~~~~~~~~~~~~~SMP L ES'-C"YLINDANLTCALR- C BOTTLE ~ ~ ~ ~ ~ ~ ~ OUL COA KO WELL WELL~~~~~~~~~~~~~~~~~~~~~VETTOAM SPHR BP GASS BTT SRGAFLE NHLUTIONS CPUN. 2 AOTETER i I~~~~~~~~~~~~~~~~~~~~~~A L GA ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~PUMP L_.0.~l~~~~~~~~~~BTL o II ~~~~~~-ETTOAMSHEREO (0 Fig.N12. Schematic.dg oO UPLE uc Fig. 12. Schematic diagram of laboratory equipment- usedN1 / WELLn a.r....i —.chl.:in:_tTHRO

........ Engineerin n Research Institute o University of Michigan L~ CI ()...........~~C. )-) f.3 0) (.3 i_ -::::::~ia *-i:I-: ~i ~ ~I:i-iiiiii~i~.'~..:~.'.'.'i:B............................. - I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~U 0 0)) C.) e3 0) 0 (.').......i i-:-lll-.~ ii~ll,.~-~;li..;ii.)................... ~~~;;;;:..;;;:;.;;...~~~~~~~~~~~~~~~~~~~~t —.................. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~;................... ~ ~ ~ *`]I::.......................;~; ~ ~ ~iiiiii~~~~j~:~~i~i~~:i::i:::::iiil~ ~ ~.............. ~~~~~~~-:~: %-~-f~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.........:~:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....................... ~~L~~i:'i~~~~i'i:'-'ii:;':i::::::-:i:-iiiii~~~~~~~~~~~~~~~~~~i''''- t~~~~~~....................~i~ ~i ii~i r ~-............ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ I xxxx~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~f:~~~~~~~~~x: m ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ C~,~...........................~~~~~~~~~~~~~~~:............ ~ ~ ~ ~ ~ ~ 53 ~~ _ c~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i~~~~~~~~~~~~~~~~~~~ii:9 a F.I Q~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Tie ~: iii:: iiiiiiiiiiiiiiiiliiiiiii':~: i:::.: _:~:.............:i ~ R ~~~~li~~~~~~\:::-:::::::::::-,:-::::-:::i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~................................... ~ ~ ~~::;;::: —:::::-:1:1:'::':~1:1:1:1::: 1: r ~~~~~~~~........................................;............i-.................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i........................................................................................... I-' cii t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~............................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i........... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ *0..............~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~t~-t ~t~..........iii'i:ii'::~::`..............................'i....................................~.:...........~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:-......................... ~ ~ ~ ~ I~.........................................I

Engineering Research Institute ~ University of Michigan the reaction at completion of the run were previously located on individual carts external to the rack proper. They took up needed space in the operation area and were awkward to move around the laboratory. In the new rack facility they are located with the chlorine cylinder on the rack proper, as shown, in Fig. 13. This eliminated external tubing connections to the rack as well as providing more operation area in front of the rack. The more central location of stopcocks and measuring devices provided for greater ease of operation. The operator now may make all measurements and adjustments from one position. Any unexpected behavior taking place in the system may be compensated for immediately without major movement on the part of the operator. The number of rubber-to-glass connections has been reduced by twoor threefold. In general, this eliminated evacuation problems. It also cut down the possibility of chlorine leakage into the analysis laboratory during operation. Glass manifolds were used instead of individual glass stopcocks. In general, the flow pattern on the rack was more systematized so that the inexperienced operator would have no problem in tracing lines. The gas flow was indicated, stopcocks were numbered, and the number of lines was reduced to a minimum for ease in operation. As indicated in Fig. 12 and shown in Fig. 13, gases enter the system on the left side of the rack, pass through control and measuring devices, enter the reaction area from the center, pass back to the rack in the center, and are absorbed for analysis on the right. The cooling system and the temperature recorder-controller are not shown in Fig. 13. The cooling unit was located in the first-floor laboratory, near the door to the radiation cave. The temperature recorder-controller is located just to the left of the rack during an actual run. Some new equipment and aids to operation were installed on the rack at the time of rebuilding. A new Flexapulse electric timer was purchased by the Engineering Research Institute for this work. A new Duo-Seal Model 1400B vacuum pump was purchased and installed on the rack. As shown in Fig. 12, new' rotameters were installed in the tail-gas line from the reactor and the N2 purge line to the absorption bottles. These were aids in operation when running plots of the gas data were kept. They also provided accurate control of gases not immediately involved in the chlorine-balance measurements. A complete operation manual for the new redesigned gas rack has been written for the laboratory and for the aid of any new operator working on the chlorination equipment.

Engineering Research Institute ~ University of Michigan ECONOMICS OF RADIATION CHEMICAL PROCESSING The laboratory studies on the polymerization of ethylene and the chlorination of aromatic hydrocarbons have demonstrated the feasibility of promoting chemical reactions with gamma radiation. It is of interest to examine the costs of such radiation processes. The chlorination of benzene is well suited to cost estimation and comparison because the reaction is presently conducted on a commercial scale, using ultraviolet radiation. The addition product of benzene and chlorine is technically known as 1, 2, 3, 4, 5, 6 hexachlorocyclohexane, but it is usually referred to as benzene hexachloride. The gamma isomer of this compound is a powerful insecticide 15,26 which is sold under various trade names. In the usual radiation reaction this isomer constitutes about 12% of the total addition product. The present commercial process utilizes small quartz equipment necessary for transmission of the ultraviolet. The following calculations are made for the gammaradiation process using a single larger reaction vessel made of nickel alloy. Costs are estimated for four different gamma-radiation sources derived from fission products. A design for a similar facility was presented by this laboratory in Progress Report 74 The reactor designed there consisted of tubes passing close to a radiation source. The reactor design presented here involves a cylindrical vessel surrounding a radiation source located concentrically within it. The vessel and the reacting mixture act as a shield for the gamma radiation. The design here uses a gamma source more efficiently than the design presented previously. Many of the calculations are the same in both cases and for more completely detailed calculations reference should be made to Progress Report 7, pages 26 to 51. The complete data for the chlorination of benzene are reported in that reference and by Harmer et al.13 The plant design here has been made on the same overall yield basis as that previously reported. A. DESIGN OF A CHLORINATION PROCESS USING GAMMA RADIATION In a typical batch run at 20~C and 62 kilorep/hr of radiation, Harmer et al.13 obtained 63.6 gm of benzene hexachloride in 0.333 hr from a charge of 20 ml of benzene and 80 ml of carbon tetrachloride. The theoretical yield is 65.6 gm, based on the benzene density of 0.88 gm/ml. Neglecting the actual increase in volume of the reactor contents, the benzene concentration decreased from 20 to 0.6% by volume. Assuming the reaction is first order

Engineering Research Institute * University of Michigan with respect to benzene concentration C, the rate is dC/dG = - kC, with k being the reaction velocity constant and G the time. Integrating this from the beginning to the end of the batch reaction gives ln(20/0.6) = - k(0.333), or k = 10.6 hr-l As a basis it will be assumed that a plant is to be designed to produce 454 kg/day of the gamma isomer. This corresponds to a production of 541 gm moles/hr of mixed isomers. A cylindrical agitated-flow reactor is to be employed with a uniform benzene concentration of 80%. The gamma source is to be placed in a center hole, according to the scheme shown in Fig. 14. The source is to be of such strength that the mean reaction velocity constant is 1.0, so the reaction rate is (1.0)(0.8) ml benzene/hr-ml solution. This is equivalent to 9.02 gm mole benzene/hr-liter solution. Since a gm mole of benzene makes a gm mole of product, the volume of the reactor is 541/9.02 or 60 liters. This volume may be realized in a cylinder 50 cm high and 50 cm in diameter, with a hole 20 cm in diameter for the radiation source and cooling coils circling the inside of the walls. The reaction velocity constant is assumed to vary directly with radiation intensity I (Harmer14 showed the chlorination of toluene to vary as NIT), which is conservative for intensities less than that of the laboratory experiment. Thus, k = mI, or k = 0.171 I, from the fact that k = 10.6 at I = 62. For rough estimation, assume the cylindrical reactor is the equivalent of a 25-cm-radius sphere with a point source of radiation at the center. There will be a spherical hole of 10-cm radius at the center to account for the volume occupied by the actual source. If the radiation intensity on the surface of the hole at 10 cm from the center is 50 kilorep/hr, and if the intensity decreases according to the inverse-square law and the exponentialabsorption law, the average k throughout the reactor volume may be calculated as kavg r fi0.171(50)(lo)2e-0.o64(o.88)(r-lO)(4ir2)dr 1.0 v 1g 0 ~r2(4/3)t [E(25)3 - (10)3] Here 0.064 is the gamma absorption coefficient and o.88 is the density of the benzene. In order to obtain an intensity of 50 kilorep/hr at 10 cm from the center for a fission-product source, about 1250 curies would be required.20 This is multiplied by a factor of four to account for self-absorption, absorption by the wall and the agitator between the source and the benzene, and losses of radiation from the cylinder through the hole. Therefore, 5000 curies will be taken as a conservative estimate. The radiation chamber housing the reactor will be made 3 meters square to provide adequate working space. To insure that the intensity of radiation is less than 0.5 millirep/hr outside this chamber with a 5000-curie

RECYCLE C6 H6 C6 H6 I REFRIGERATION STORAGE SYSTEM C6 H6SATD. WITH Cl2 SOURCE IN MIXER REACTOR HS ) ~3~ ~ COLD BRINE COOLER' EXCESS S ITI C 12 REC YCL. H20 --- i~~ ~CONDENSOR REACTOR HOUSED IN SHIELDED C6 H6 CHAMBER PRODUCT C SEPARATION. C6 H6 C6 S M STEAM AND/OR HOT WATER I 0C6 H6 RADIATION H20+ C6 H 6 SOURCE IN WELL TOCHC TO SEPARATIONS - cl2 STORAGE a WELL OF H20 FOR SOURCE STORAGE Fig. 14. Flowsheet of proposed plant design for producing benzene hexachloride.

Engineering Research Institute ~ University of Michigan source, the concrete walls must be 85 cm thick.22 The other units in the system are calculated by conventional means. The process is straight-forward, with the benzene being saturated with chlorine before entering the reactor and excess chlorine bubbling through the reactor to keep the contents saturated The product from the reactor is flashed to separate out the solid benzene hexachloride. The benzene is condensed and returned to the chlorine saturator. Refrigerated coolant is required to maintain the reactor feed and contents at 200C. B. COST ESTIMATIONS The costs of fission-product sources are not yet available. Therefore, estimations are made for a range of values that seem reasonable. Because of varying half-life of different sources, it will be assumed that charges are made approximately according to the radiant energy delivered. Since the energy E of a radioactive source is given by fIoe-~0 dG = Io(1 - e- x)/k, at large values of time G the energy varies inversely as the disintegration constant X. But X is the reciprocal of the half-life, so the energy varies directly as the half-life. The costs will therefore be selected in proportion to the halflife. The cost of long half-life sources is to include shipping. Short halflife sources would involve excessive shipping costs and are assumed to be used in the vicinity where they are produced. If the source intensity is not to fall below', say, 80% of its initial strength, short half-life materials will require frequent replacement. These factors have been considered in Table 12, where the costs of producing the mixed isomers of benzene hexachloride are given as a function of the type of radiation source. In estimating the total radiation required, the source is to be composed of five sections. Whenever the intensity falls off 20%, one section is replaced with fresh radioactive material. In the case of cesium-137, it is apparent that no replacements are necessary. In fact, at the end of a five-year period, the cesium activity is still more than 90% of its initial value. In selecting costs for radiation sources, tenfold ranges were estimated for each case, but only the highest value is given in the table. It is noted that the variation in costs of radiation actually has only a small effect on the product cost. The costs of benzene and chlorine are taken from the latest trade journals. These also give the cost of commercial benzene hexachloride as l1.90 to $3.00 per kg of gamma isomer in 12% mixture with the other isomers. The gamma-radiation process evidently can compete with the ultraviolet process for making this compound. Of course, the true comparison between the two should be made on the basis of the comparative costs of the reactors and radiation sources, since all other equipment is the same. What this comparison does show, however, is that gamma-radiation processing is not out of line with other processing methods and that it should be given serious consideration by industry.

Engineering Research Institute ~ University of Michigan TABLE 12. COST ESTIMATE FOR PRODUCING 454 kg/DAY OF GAMMA ISOMER OF BENZENE HEXACHLORIDE WITH FOUR RADIATION SOURCES Radiation Source Gross Gross Gross CesiumFission Fission Fission 137 Product Product Product 0.5-yr. 1 - yr. 2 - yr. Half-life, years. 0.2 0.3 1.1 33 Total Curies Reqd.-5 years. 84,ooo 57,000 19,000 5,000 Selected High Cost/Curie, $. 0.05 0.10 0.25 5.00 Radiation Cost-5 yr, $. 4,200 5,700 4,800 25,000 Initial Investment in Radiation Source. 250 500 1,250 25,000 Radiation-Chamber Cost, $. 22,000 22,000 22,000 22,000 Process-Equipment Cost, $. 49,000 49,000 49,000 49,000 Total Investment, Including Radiation. 71,250 71,500 72,250 96,000 Annual Charges on Investment at 80o, Depreciation, Taxes, Interest, Maintenance, Etc. 57,000 57,200 57,800 76,800 Annual Radiation-Replacement Cost. 800 1,000 700 0 Salaries of Workers, $/yr. 21,000 21,000 21,000 21,000 Chlorine and Benzene, $/yr. 109,000 109,000 109,000 109,000 Utilities, $/yr. 3,000 3,000 3,000 3,000 Sales and Marketing, $/yr. 8,00 8,oo000 8,000 8,000 Total Annual Cost, $. 199,000 199,000 200,000 218,000 Annual Production, 290 Working Days, kg Gamma Isomer. 131,500 131,500 131,500 131,500 Cost/kg Gamma Isomer, in 12% Mixture, $. 1.51 1.51 1.52 1.66 47

Engineering Research Institute ~ University of Michigan COOPERATIVE WORK WITH MICHIGAN MEMORIAL-PHOENIX PROJECT NO. 98 PREPARATION OF TRICHLOROACETIC ACID Clerement5 reported that trichloroacetic acid was obtained in 631 yield by the oxidation of chloral hydrate with fuming nitric acid. The same author also reported the synthesis of the trichloro acid by oxidizing chloral hydrate with KMnO4.6 Other investigators were able to prepare the acid through the oxidation of an aqueous solution of chloral hydrate with KC10324 or 29 CaOC12. Ralph E. Plump30 stated in his patent that trichloro acid was obtained in over 60% yield by the oxidation of the corresponding trichloroaldehyde or the aldehyde hydrate with an aqueous solution of chlorates during agitation. The reaction is initiated at lower temperature in presence of a promotion catalyst such as NH4VO3 or Ce(S04)2. 33 ~ Stauff Schumache,33 in an attempted photochemical oxidation of chloral with 02 at 700 to 900, found that COC12 + CO + HC1 were the decomposition products obtained. Charles Strosacherl~ studied the preparation of trichloroacetic acid by the direct chlorination of acetic acid in the presence of a chlorinating catalyst such as S chloride; the monochloro acid was first formed, and when the chlorination was continued successively at high temperature up to 160~ the trichloro compound was formed. In our chemistry laboratory, repeated experiments at different temperatures gave no conversion of chloral hydrate to trichloroacetic acid by the action of chlorine alone. At room temperature there was no reaction of chlorine on anhydrous chloral, but at high temperature the chain reaction involving the formation of COC12, CO, and HC1 occurred at a very high reaction rate. When sealed tubes of chloral hydrate and liquid chlorine were subjected to gamma radiation, excellent yields of trichloroacetic acid were obtained. Attempts have been made to carry on the conversion as a continuous process; thus far there are not enough data available to indicate the necessary radiation doses, temperature, chlorine concentration, and other factors that affect the reaction. This work is being continued because the product of the reaction has real value, a process of this type merits considerable study, and its mechanism and kinetics should be investigated. 48

Engineering Research Institute ~ University of Michigan A typical experimental run is outlined below: Chloral hydrate, 5.5 gm (0.33 moles), and liquid chlorine, 16 ml (.3 mole), were irradiated for 15 hours in a sealed tube at a rate of 41,000 rep/hr at an average temperature of 8~ to 100. The tube was opened and the chlorine was allowed to evaporate at room temperature. The semi-solid residue was extractedfromthe tube with petroleum ether (60~ to 75~). The solvent was evaporated and the residue was fractionated in a small column and.9 gm of unreacted chloral hydrate distilled at 90~ to 96~. The residual viscous oil in the distilling flask was subjected to sublimation and 3.8 gm of trichloroacetic acid sublimed at 70~ to 740 at 2-mm pressure (80% yield). The colorless crystalline trichloroacetic acid melted at 550 to 570 and gave no depression when admixed with an authentic sample. The methylaniline salt came down quantitatively by mixing equimolecular quantities in a little benzene and leaving the warmed-up solution to cool. The salt crystallizes in colorless needles from CC14. The crystalline needles have a melting point of 970 31 The trichloroacetic acid could be isolated from the reaction mixture by extracting the reaction products with dry benzene and adding the proper amount of freshly distilled methylaniline. The methylaniline salt crystallizes when the mixture is allowed to stand overnight. Chlorine does not react with chloral hydrate in the absence of radiation. Chloral hydrate (5.5 gm) and liquid chlorine (16 ml) were mixed together in a sealed tube and allowed to stand in the dark for three weeks. When the reaction mixture was subjected to the treatment outlined above, chloral hydrate was recovered unchanged. CONCLUSIONS AND PROPOSALS FOR FUTURE WORK On the basis of the work completed and reported here, recommendations as to future work can be made. The work here, combined with the work by Lewis,l7 gives a generalized picture of the polymerization of ethylene influenced by gamma radiation. The data are sufficient to indicate general trends of temperature, pressure, radiation, and oxygen concentration on the reaction. The yield of polyethylene under gamma radiation is high enough to warrant investigation of the commercial feasibility of the radiation process. A preliminary plant design similar to the design presented for the chlorination of benzene will be made. The data on the copolymerization of butadiene and styrene show that, in general, the radiation reaction is not as favorable as the present emulsion 49

Engineering Research Institute ~ University of Michigan polymerization being employed commercially. However, because this compound has such a high commercial value, possibly a few runs will be made at other ranges of reaction conditions than those presented here. No major emphasis will be put upon this future work and the runs will be made at widely divergent reaction conditions. The number of runs will be kept at a minimum. The data taken on the copolymerization of sulfur dioxide with various olefins indicate that the reactions proceed in a fairly straightforward manner under radiation. The yield is a direct function of the radiation applied. There is little or no induction period as is present in many polymerization reactions under radiation conditions. The reaction proceeds fast enough to warrant commercial investigation. For these reasons the kinetics of reaction will be investigated further to determine the combined radiation and temperature effects upon the copolymerization of sulfur dioxide with a complete series of olefin hydrocarbons. Some olefins not considered in this report will be investigated to give a complete picture of the reaction under radiation conditions. The data on the Oxo-type reaction are, in general, poor. It is not clear what effect the radiation has upon the reaction. Several more runs will be made under different pressure and temperature conditions in the radiation field to try to answer this question. Olefins of higher molecular weight will be used in some of the future reactions. General trends are shown in the chlorination reactions for the effects of the addition of methyl groups to the benzene ring. No future runs are contemplated with toluene, xylene, or mesitylene. The effect of longer side chains on the benzene ring is not clear from the runs on the one compound (ethyl benzene) investigated. It is anticipated that runs with higher homologues of the same series will help clear up the question. Several runs will be made with propyl benzene and higher alkyl aromatics. The chlorination of higher ring structures,like naphthalene and possibly anthracene, is to be investigated further. These latter compounds will be run in a solution inert to chlorine such as carbon tetrachloride. Other gas-liquid reactions will receive literature study and some experimental work. Reactions involving ammonia and hydrogen sulfide as the gaseous reactants are of some interest and will receive attention. The glass reaction equipment is well suited to the measurement and control of these gases and therefore a minimum of equipment changes will have to be made. Analysis procedures must be investigated before any such reactions are carried out. Batch and flow bromination and iodination should receive some attention during the future period. These reactions should proceed in much the same manner as the chlorination reactions already investigated. 50

Engineering Research Institute ~ University of Michigan The preparation of trichloroacetic acid from the chlorination of chloral hydrate will be subjected to further study. The kinetics of reaction involving radiation is to be studied further by the group from the Michigan Memorial-Phoenix Project. In general, a variety of reactions will be studied in the immediate future. Both polymerization and gas-liquid reactions are well suited to the facilities and equipment available in the laboratory. It is in these general fieldswhere most of the work will be carried out. 51

Engineering Research Institute ~ University of Michigan BIBLIOGRAPHY 1. Alyea, H. N., "Chain Reactions Produced by Light and by Alpha Radiation," J. Am. Chem. Soc., 52:2743 (1930). 2. Anderson, L. C., Martin, J. J., Brownell, L. E., et al., "Utilization of the Gross Fission Products," Progress Report 2, Eng. Res. Inst. Project M943, Univ. of Mich., Ann Arbor, January, 1952. 3. Ibid., Progress Report 5, September, 1953. 4. Ibid., Progress Report 7, December, 1954. 5. Ann., 161:128 (1872); Comptes Rendus, 73:113 (1872). 6. Ann., 166:64 (1873); Comptes Rendus, 74:1492 (1872). 7. Dennis, L. M., and Shelton, R. S., "An Apparatus for the Determination of Melting Points," J. Am. Chem. Soc., 52:3128-32 (1930). 8. "Determination of Sulfur in Vulcanized Rubber Compounds," The Standard Calorimeter Co. Booklet 104, East Moline, Illinois. 9. Dienes, G. J., and Klemm, H. F., "Theory and Application of the Parallel Plate Plastometer," J. Appl. Phys., 17:458 (1946). 10. Dow Chemical Co., U.S. Patent No. 1,757,100. 11. Gonze, M., U.S. Patent No. 2,513,092. 12. Hardie, T., U.S. Patent No. 2,218,148. 13. Harmer, D. E., Anderson, L. C., and Martin, J. J., "Chlorination of Some Aromatic Compounds Under the Influence of Gamma Radiation, Chem. Eng. Prog. Symp. Series, 50:253-57 (1954). 14. Harmer, D. E., The Reaction of Chlorine with Aromatic Compounds Under Intense Gamma Radiation, Ph.D. Thesis, Univ. of Mich., 1955. (1943:4-41-T, Eng. Res. Inst., Ann Arbor, 1955.) 15. Hensill, G. S., "Lindane, A review of the background and development of this modern and versatile insecticide," Soap and San. Chem., 27:135 (1951).

Engineering Research Institute ~ University of Michigan 16. Kirk-Othmer, Encyclopedia of Chemical Technology, 10:938-57 (1953). 17. Lewis, J. G., and Martin, J. J., "Promotion of Some Chemical Reactions With Gamma Radiation, " Project 1943:4-1-T, Eng. Res. Inst., Univ. of Mich., Ann Arbor, 1954. 18. Lewis, J. G., Nehemias, J. V., Harmer, D. E., and Martin, J. J., "Analysis of Radiation Fields of Two Gamma-Radiation Sources," Nucleonics, 12:40-44 (1954). 19. Loiseau, L., French Patent No. 565,356, January 25, 1924. 20. Marinelli, L. D., Quimby, E. H., and Hine, G. J., "Radium Therapy," Am. J. of Roentgenology, 59:260 (1948). 21. Marvel, C. S., and Weil, E. D., "The Structure of Propylene Polysulfone," J. Am. Chem. Soc., 76:61 (1954). 22. National Bureau of Standards Bulletin 54, Protection Against Radiations From Radium Cobalt-60 and Cesium-137, Washington, D.C., 1954. 23. Orchin, M., and Schroeder, W. C., Hydroformulation (Oxo Reaction), Chapter IX, "Unit Processes in Organic Synthesis, edited by P. H. Groggins, 4th edition. New York: McGraw-Hill Book Co., 1952. 24. Seubert, Ber., 18:3336-69 (1885). 25. Sherwood, P. W., "Oxo Process in the United States for Petrochemicals From Olefins," Pet. Proc., 8:241-8 (1953). 26. Slade, R. E., "The y-Isomer of Hexachlorocyclohexane (Gammexane)," Chem. and Ind. Eng., 64:314-19 (1945). 27. Snow, R. D., and Frey, F. E., "The Reaction of Sulfur Dioxide with Olefins: The Ceiling Temperature Phenomenon," J.Am.Chem.Soc.,65:2417 (1943). 28. Snow, R. D., and Frey, F. E., "Reaction of Sulfur Dioxide With Olefins," Ind. and Eng. Chem., 30:176 (1938). 29. Suknewitsch, Z. phys. Chem., B48:880 (1929). 30. U.S. Patent No. 2,370,577. 31. Wheeler, S., and Jennings, F. D., J. Am. Chem. Soc., 49:1091 (1927).

3 9015 03695 1765 Engineering Research Instirur University of Michigan 32. Willard, H. H., and Furman, N. H., Elementary Quantitative Analysis. New York: D. Van Nostrand, Inc., 1952. 33. Z. phys. Chem.,,B48:17, (1941).