ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN ANN ARBOR THE EFFECT OF GAMMA RADIATION ON CHEMICAL REACTIONS A collection of material originally presented as part of the Progress Reports of the Fission Products Laboratory from 1951 to 1955 Joseph J. Martin Associate Professor, Department of Che~{cal and Metallurgical Engineering Leigh C. Anderson Professor and Chairman, Department of Chemistry Project Supervisors David E. Harmer John G. Lewis Bruce G. Bray Senior Research Assistants F. Bashore R. L. Klemm R. A. Carstens R. R. Nissle C. E. Eckfield E. M. Rosen D. J. Goldsmith W. M. Sergy J. R. Hallman S. A. Stolton J. P. Holmes H. H. Yang R. L. Kinney Assistants in Research Project M943-4 U. S. ATOMIC ENERGY COMMISSION CONTRACT NO. AT(11-1)-162 CHICAGO 80, ILLINOIS June, 1956

INTRODUCTION A research team in the Engineering Research Institute of The University of Michigan has been investigating various uses of radiation for industrial applications. This investigation was conducted under the general title "Utilization of the Gross Fission Products." The effects of radiation on such things as combustion-engine performance, promotion of chemical reactions, and preservation of foods, were studied. Various cooperative programs with other departments of the University and with industrial concerns were instigated. The results of these investigations were collected and published semiannually as progress reports of the whole laboratory facility. The demand for reprints for specific sections of these reports warranted this reproduction in a collected form. This volume is the collected work on "The Effect of Gamma Radiation on Chemical Reactions." This study began in 1951 and has continued unbroken until the present time. Presented in this volume is all the work that was reported from August, 1951, until December, 1954, in Progress Reports 1 through 7 of the Fission Products Laboratory. A portion of this work may be found in the doctoral dissertations of D. E. Harmer and J. G. Lewis of The University of Michigan. A number of papers based on these reports were written and published in the technical literature from time to time. The page numbers, figure numbers, and table numbers are the same here as in the original reports. The contributors to the work are listed on the title page and each appears at the head of the section in which he was involved. A table of contents, list of figures, and list of tables appears before each individual section.

TABLE OF CONTENTS PROGRESS REPORT NO. 1 (C00-86), "Utilization of the Gross Fission Products," by L. C. Anderson, J. J. Martin, L. E. Brownell, et al., Eng. Res. Inst. Project M943, Univ. of Mich., Ann Arbor, August 31, 1951, pp. 28-60. A complete literature survey was made on the effects of radiation upon chemical reactions. Ninety-two references involving both experimental and theoretical work are listed. PROGRESS REPORT NO. 2 (AECU-1889), "Utilization of the Gross Fission Products," by L. C. Anderson, J. J. Martin, L. E. Brownell, et al., Eng. Res. Inst. Project M943, Univ. of Mich., Ann Arbor, January 31, 1952, pp. 30-42. A high-pressure stainless-steel reactor was designed for using highpressure reactions under gamma radiation. Some preliminary experiments were conducted with the 1000-curie Co-60 source and with a palladium-109 source. PROGRESS REPORT NO. 3 (C00-91), "Utilization of the Gross Fission Products," by L. C. Anderson, J. J. Martin, L. E. Brownell, et al., Eng. Res. Inst. Project M943, Univ. of Mich., Ann Arbor, June, 1952, pp. 31-39. Experiments have been conducted on the polymerization of styrene and natural oils and additional information has been obtained on the ammonia synthesis reaction. PROGRESS REPORT NO. 4 (C00-124), "Utilization of the Gross Fission Products," by L. C. Anderson, J. J. Martin, L. E. Brownell, et al., Eng. Res. Inst. Project M943, Univ. of Mich., Ann Arbor, March, 1953, pp. 24-36. It has been found that gamma radiation readily promotes the chlorination of benzene by substitution to produce benzene hexachloride. Gamma radiation also promotes the chlorination of toluene. Ethylene can be polymerized under the influence of gamma radiation to produce a polymer having properties which differ from the conventional polymer produced by existing methods. Sulfurous acid can be converted to sulfuric acid by passing oxygen through sulfurous acid in the presence of gamma radiation. PROGRESS REPORT NO. 5 (C00-196), "Utilization of the Gross Fission Products," by L. C. Anderson, J. J. Martin, L. E. Brownell, et al., Eng. Res. Inst., Project M943, Univ. of Mich., Ann Arbor, September, 1953, pp. 21-68.

TABLE OF CONTENTS (concluded) The study of the effect of radiation on the promotion of chemical reactions has been continued with studies on the polymerization of ethylene and the chlorination of benzene and toluene. Initial studies indicate that the chlorination of toluene gives the hexachloro addition product. A discussion of the thermodynamics and kinetics of reactions is included. Dosimetry studies on both the l-kc and 10-kc sources are reported, and the nominal activities and those determined from dosimetry measurements are compared. The actual activity of both the l-kc and 10-kc sources is found to be considerably less than the reported activity. Personnel of Brookhaven National Laboratory have reached the same conclusion regarding the true activity of the l-kc source, based on reactor computations. PROGRESS REPORT NO. 6 (CO0-198), "Utilization of the Gross Fission Products," by L. C. Anderson, J. J. Martin, L. E. Brownell, et al., Eng. Res. Inst. Project M943, Univ. of Mich., Ann Arbor, March, 1954, pp. 12-48. Ethylene was polymerized under gamma radiation at pressures of about 1000 psi and room temperature. Radiation dosages of 5 to 7 megarep produced a hard, tough polymer having a tensile strength up to 2300 psi with 79 percent elongation upon rupture, and a molecular weight of 37,300. Lesser dosages of radiation produced a soft, brittle, waxy polymer. Toluene was chlorinated under gamma radiation and produced.the addition compound. This reaction is considered to be unique in that it is believed to be promoted only by gamma radiation. The product toluene "hexachloride" is being evaluated as an insecticide. PROGRESS REPORT NO. 7 (AECU-2981), "Utilization of the Gross Fission Products," by L. C. Anderson, J. J. Martin, L. E. Brownell, et al., Eng. Res. Inst. Project M943, Univ. of Mich., Ann Arbor, December, 1954, pp. 9-53. Further studies on the polymerization of ethylene under gamma radiation were made using higher temperatures and slightly higher pressures. At 2200C an oily liquid was obtained whereas tough or waxy solids had been obtained at lower temperatures. Toluene was chlorinated under gamma radiation. The products obtained by dehydrohalogenation and subsequent oxidation of the chlorinated toluene confirm the hypothesis that chlorine adds to the ring of toluene. New glass reaction equipment has been constructed to study the kinetics of the chlorination reaction and analytical procedures have been developed. A design was developed for a chemical plant to chlorinate benzene under gamma radiation for the production of insecticides. Sources of cesium-137, 6-month-old mixed fission products, and cooling reactor-fuel elements were compared. Estimated prices of producing the addition product using gamma radiation compared favorably with existing prices of the product prepared under ultraviolet radiat ion. vr

Progress Report No. 1 UTILIZATION OF THE GROSS FISSION PRODUCTS The Effect of Gamma Radiation on Chemical Reactions L. C. Anderson D. E. Harmer J. J. Martin J. G. Lewis

TABLE OF CONTENTS Progress Report No. 1 Page PROJECT M943C - THE EFFECT OF RADIATION ON CHEMICAL REACTIONS 28 1. Introduction 28 2. Literature Review for Project M943C 29 a. Inorganic Reactions 29 b. Organic Reactions 33 c. Literature Review - Theoretical 44 3. Proposed Research on Selected Reactions 51 a. Inorganic Reactions 51 b. Organic Reactions 52 4. Bibliography for Project M943C 54

B. PROJECT M943C - TE EFFECT OF RADIATION ON CHEMICAL REACTIONS 1. Introduction It is the purpose of this project to study the promotion of chemical reactions by the use of radiation from waste fission products. Interest has been stimulated in this and other applications of waste radioactive materials by the availability of vast quantities of these materials as by-products from the operations of the nuclear reactors of the Atomic Energy Commission. Both from technical and economic considerations it is of interest to attempt to employ fission products for industrial purposes. If fission products should prove capable of promoting some chemical reactions to an extent which is attractive from a technical viewpoint, the processes which would be suggested by such reactions would still have to be examined to determine their economic feasibility. This project is primarily concerned with radiation chemistry, which deals with the effects of high-energy photons and charged particles resulting from radioactivity. A study of the influence of radiation, consisting of these high-energy photons and particles, upon those chemical reactions which are thermodynamically feasible under the existing operating conditions seems to be a logical and appropriate starting point. Such reactions would be those which would yield a favorable ratio of desired product to reactant if brought to equilibrium under the conditions of the reaction. Judging from a study of the previous work done in the field of radiation chemistry, it appears to be debatable whether or not radiation may cause a displacement of the position of equilibrium in a chemical reaction. At any rate, the first studies of this project will be to determine the effect of radiation upon the reaction rates for selected systems. "Radiocatalysis" is the term which will be used to designate promotion of reactions in the nmanner just described. Thus "radiocatalysis" as used here and "catalysis" as generally used, have much the same basic significance, i.e., both terms are used to describe the changing of the rates of a chemical reaction without changing the composition of the reacting masses at equilibrium. Radiocatalysis simply indicates that some form of radiation is employed to change the reaction rates. The radiation from fission products consists of beta and gamma rays. The gamma radiation, because of its great penetrating power, could be useful in commercial processes in which the chemical reactants are separated from the fission products by solid walls. On the other hand, the beta radiation, because of its small penetrating power, may be utilized if reactants can be brought into intimate contact with the fission products. In the course of this study gamma-ray and beta-ray sources and x-ray machines will be employed. Comparison of results using different sources of radiation may serve to indicate the relative efficiency of different kinds and different energies of radiation in promoting reactions. 28

Tests may be made of the relative effects upon chemical reactions of radiation alone and of radiation in combination with solid catalysts. It is thought that perhaps radiation may cause a catalyst ordinarily used for a given reaction to promote the reaction under less severe physical conditions (such as lower temperature and pressure) than are ordinarily required, or that the addition of radiation to the usual installations may increase the rates of reaction. There is the possibility that because of the low absorption of the gamma rays, little reaction might occur, especially in those cases where the reactants are not very dense. In order to utilize the energy of the gamma rays in the reacting masses, a chemically inert material might be added to the reactants in the form of a liquid or a gas, or else a solid packing might be placed in a reaction vessel. Such added materials might influence the rate of reaction by one or more of several mechanisms. Secondary emissions might be produced, such as softer gamma rays or electrons. Some gases or vapors might be added which would be more opaque to gamma radiation than the reactant molecules or which would be capable of absorbing and transferring energy to the reactant molecules without being permanently altered themselves. Laboratory experiments determining the effect of radiation on chemical reactions are not new. However, the industrial application of these reactions has not been feasible until this time because of the lack of cheap sources of radiation. With the vast quantities of fission products available, it is worthwhile to review these experiments with the obvious intention of trying to find some reactions which might have great industrial promise. Although there are many papers on the general subject, the following review covers only those which are closely related to the objectives of this work. 2. Literature Review for Project M943C a. Inorganic Reactions: There is some evidence in the literature that ammonia may be formed from the elements under the influence of charged particles. Lind and Bardwell (49) gave data for the flow method of producing ammonia from nitrogen and hydrogen activated by alpha radiation from radon. The resulting gas mixture was analyzed and showed the ratio of the molecules of ammonia formed to the ion pairs of reactants produced to be 0.2 to 0.3. According to these investigators the ratio of moles of ammonia decomposed to those ionized is one, and the equilibrium is reached when five times as much radiation falls on the hydrogen and nitrogen molecules as on the ammonia molecules. This is found to occur when there are 10 volumes of reactant (3H2 + N2) gases to one volume of ammonia, or a mixture containing 9.09 per cent ammonia by volume. Results in which decomposition of ammonia proceeded to 73 per cent showed that the reactionvelocity constant falls steadily as the decomposition reaction proceeds, as if a reverse reaction were occurring. 29

Lind (44) discusses radiochemical equilibrium in the synthesis of ammonia. He designates "M" as the number of moles of ammonia that are formed and "N" as the number of ion pairs formed. Using M/N = 0.2 afn -M/N = 1.0 (49), for the formation and decomposition of ammonia, respectively, Lid and Bardwell calculated equilibrium at 250 to be 83.3 per cent decomposition, or 9.09 per cent ammonia by volume. Ponsaert (70) uvsd 0.32 for M/N, ahd 1.08 for -N/ll to calculate equilibrium at 13.5 per cent ammonia by volume. The aetual equillbrift found by D'Alieslager and Jungers (21) was only 4,7 per cent ammonia by volume. In calculating it was assumed that the mechanisms of the two reaction were independent of each other in intermediate steps. Only 0.11 for M/N would be required to give an equilibrium concentration of 4,7 per cent ammonia by volume. Accordingly, Lind (44) maintains that the intermediate steps are not independent and that there wmust be an exchange of activation energy in the direction to produce decomposition. The shift of equilibrium is in the correct direction to be accounted for by an exchange of intization from H2+ (16 volts) or N2+ (17 volts) to give HE + (11 volts).* This behatior favors decomposition at the expense of synthesis. The type and directton of thift Just described may be general in other similar reactions because a large molecule usually has a lower Ionization potential than that of its components. Boulle tested the effeet Of cathode rays from warioun metallic cathodes upon the ammonia synthesis. Methods and apparatus for the catalysis of N2 + 3 H2 = 21H3 at about 3 mm pressure by the radiations from various metallic cathodes were described. Aluminum, antimony, silver, tin, platinum, lead, and silicon in various chemical and physical forms were tested as catalysts. Optimum currents, voltages, and pressures for the best yields per unit of power input wete determined. A platinum coil was found best as a catalyst. Ammonia yields were comparable to those of a high-pressure reaction. The thermodynamic equilibrium of the reaction and the temperature of the cathode diseharge were determined. No catalysis was found at the anode. Williams and Essex (90) studied the ion yield in nitrous oxide bombarded by alpha rays at 10-20 cm absolute pressure. The ion yield increased with electric field stresgth above half saturation values because of electron acceleration. * Later measurements show these values to be 15.4 volts for H2+, 15.5 volts for 12+, and 11.2 volts for N3+. (Friedlander, G., and Kennedy, -., Introduction to Radiochemistry. New York: John Wiley and Sons, Inc., 1949.)

Lind (42) studied the kinetic behavior of the combination of oxygen and hydrogen under the influence of alpha rays. The velocity of the reaction was found to depend only upon the quantity of radiation and the pressure. A kinetic equation was given, Hydrogen and oxygen were both found to be activated. Varying proportions of reactants caused changes in the rate of reaction, due to the different specific ionizations of the reactants. The temperature coefficient of the reaction was found to be zero between 0 and 25 ~C. Lind (43) investigated the rate of reaction of hydrogen with oxygen under the influence of alpha rays at small volumes and low pressures. Under the conditions mentioned the rate of reaction was observed to be abnormally high. This effect was attributed to the action of "recoil ions" resulting from the recoil of an atom from which an alpha particle had been ejected. The approximate statistical agreement between ionization and chemical action was cited for cathode rays, beta rays, alpha particles, and recoil atoms, with respective masses from 1/1700 to 220 times the mass of the hydrogen atom. The catalytic influences of the ions of the inert gases during the bombardment of certain gases by alpha particles was studied by Lind and Bardwell (47). The effects were noted of alpha particles on acetylene, cyanogen, hydrogen cyanide, the oxidation of CO and H2, the decomposition of CO and the decomposition of lIH3, as catalyzed by the inert gases nitrogen, helium, neon, argon, krypton, xenon, carbon dioxide, and hydrogen. The ions of the inert gases acting as catalysts were quantitatively equivalent to those of the reactants in producing chemical reaction. Nitrogen and carbon dioxide failed to autocatalyze the reactions in which they were generated. This behavior was exceptional. The catalyst may have had an ionization potential either higher or lower than that of the reactants. This possibility precluded a primary step consisting of an exchange of charges between the ionized catalyst and the neutral reactants. By observation, it was determined that for any fraction of ionization of the gaseous catalysts up to 0.50, the catalyst efficiency was 100 per cent. Lind and Bardwell (46) have investigated the reactions of carbon dioxide and carbon monoxide, Carbon monoxide alone under alpha rays gave carbon dioxide, carbon, and an unknown suboxide of carbon, A mixture of carbon monoxide and oxygen was oxidized to carbon dioxide by alpha radiation. This reaction proceeded at the temperature of liquid air at about one-half the rate observed at room temperature. Carbon monoxide and hydrogen gave a white solid which was neither aldehyde nor sugar, was insoluble in water, and had the approximate composition of a polymer of formaldehyde. Carbon 31

dioxide and hydrogen gave a polymer of formaldehyde different from that mentioned above, water, and a small amount of carbon monoxide, but no methane. It was stated that there was no chain effect. Carbon dioxide alone was unaffected by alpha radiation. Watson, Vanpee, and Lind (86) mixed carbon monoxide in an 8-cmdiameter glass flask with radon having an initial activity of 100 mc and allowed the mixture to stand for "more than a month". Carbon dioxide, graphite, and one additional solid were obtained as products. The graphite and the other solid were examined by x-ray powder patterns, by means of which the graphite was identified. Lines appeared which could not be identified.and were attributed to a suboxide of carbon, C302, which was presumed to be the solid other than graphite. Examination of the solids by electron microscopy indicated the presence of hexagonal particles, supporting the identification of graphite by powder diffraction. The overall reaction was given as 6co = 2C02 + C + C302. This reaction, it was suggested, proceeds by two or more reactions of lower order, as a result of the ionization of the CO by the alpha particles. Ammonium persulfate dissolved in anhydrous glycerine was irradiated with 0.8 A x.-rays by Broda (7). The decomposition of the persulfate was measured iodometrically. The decomposition was proportional to the dosage of radiation and was of the first order with regard to persulfate concentration. Glycerine was used because it was unaffected by the x-rays. Solutions of potassium dichromate were subjected to electron bombardment by Treiman (84). Acidic solutions were steadily reduced, while neutral solutions became alkaline and the rate of reduction became slower. Aerated solutions gave the same yields as deoxygenated ones. Mund (65), assuming Geiger's law, gave formulas and numerical tables for the calculation of the number of ions which are produced in a spherical vessel by the alpha rays of a given amount of radon in equilibrium with RaA and RaC wholly deposited on the walls. An evaluation was given by Snyder and Powell (77) of the usefulness of vatiousformulas in calculating the efficiency of different materials in absorbing gamma radiation. Experimental values were given for the absorption coefficients of aluminum, iron, copper, silver, tin, tantalum, lead, and vanadium as functions of the mev of the radiation. Absorption data were alao given for nitrogen, oxygen, carbon, water, air, and tissue. 32

b. Literature Review - Organic Reactionso It is indicated in the literature that the types of reactions which are most frequently promoted by radioactive discharge are those which involve polymerization and/or dehydrogenation. Saturated hydrocarbons have been observed to undergo both reactions, yielding an unsaturated product of higher molecular weight. Unsaturated hydrocarbons were found to polymerize. A 20 cc ampule of monomeric styrene, pure or in solution, was placed 4.5 cm from 100-400 mc of radium wrapped in 1 mm lead foil (3). At 12'C the material polymerized at the rate of 0.015 per cent polymer per curie-hour exposure. For pure styrene less than 5 per cent polymerized, and for styrene in methanol less than 15 per cent polymerized the percentage converted was linear with exposure time aid was proportional to the square root of the radiation intensity. The reaction was assumed to involve free radicals, since polymerization was blocked by 1 per cent benzoquinone. Styrene in solutions 20 mole per cent with primary or secondary amines, acetone, propionitrile, benzene, or cyclohexane was polymerized faster than when pure. When in solution in hydrocarbons, the rate of polymerization was reduced, Coolidge (14) used hot-cathode, high-vacuum tubes, allowing cathode rays to pass out through a window of aluminum foil 0.00265 mm thick, and 1.7 mm in diameter. There was produced from acetylene a yellow compound resembling the product both from corona discharge in acetylene and from the use of radium emanation, Under the influence of the rays described, castor oil changed rapidly to a solid. Crystals of cane sugar turned white and evolved a gas upon subsequent heating. Aqueous solutions of cane sugar became acid to litmus. The theory and experimental methods and procedure were discussed for the polymerization of acrylonitrille and methacrylonitrile by means of gamma rays and x-rays (15). The results were discussed, and many references were given. Aqueous solutions of acrylonitrile of several concentrations were given various dosages of gamma radiation from radium (16). Polymerization occurred as a second-order reaction with respect to the concentration of monomer. The reaction was independent of the 3trength of 'the source. The induction period was dependent upon the strength of the source and upon the concentration of monomer. The induction period was thought to be caused by the presence of an inhibitor which was destroyed by the products of the primary process, as well as by reaction with growing polymer chains. 33

The action of radon upon various hydrocarbons was reported by Heisig (32-). Compounds acted upon by radon were allene, methylacetylene, dimethylacetylene, 2-butene, and isoprene. Details of the experiments were given, with a short discussion pertaining to the structure of the compounds. The polymerization of allene and methylacetylene were similar, in that in the presence of radiation a fog formed, which in turn condensed to a light-colored liquid. The liquid became more viscous, and a lightcolored solid formed. A fog formed immediately in dimethylacetylene and condensed to a moderately viscous liquid, resembling a medium lubricating oil in color and viscosity; no solid formed. A fog formed in butene-2, and after several hours the droplets settled. The liquid became darker and more viscous as the action progressed, but no solid formed. In isoprene, a fog formed immediately, droplets settled, and a heavy, colorless, viscous, rubbery mass collected. Styrene (13), and acrylonitrile (15,16), were polymerized during radiation by gamma rays. The polymerization of methyl methacrylate at room temperature under the action of a radium preparation was determined (71). Polymerization occurred not only during but after irradiation. In both periods the reaction was autocatalytic, though much less rapid in the latter. Hopwood and Phillips (37) used a volume dilatometer to study the rate of polymerization of methyl methacrylate with neutrons, beta rays, and gamma rays. The rate of polymerization was somewhat higher using both gamma rays and neutrons than when using gamma rays alone. The sources of radiation were: for gamma rays, 78 mg Ra(SO4) in platinum needles; for gamma rays plus neutrons, 78 mg Ra(S04) plus beryllium in Monel tubes. Data and discussion were presented by Burr and Garrison (8) for some investigations of the changes of physical properties of 25 different plastics and synthetic rubbers. Samples were irradiated with beta and gamma rays, and then the specific electrical resistance the hardness, and the tensile strength were determined. Beta and gamma rays gave about the same chane in tensile strength for the total energy absorbed, evidently because of heating within the material. The presence of polar impurities and polar groups alike within a plastic appeared to cause temporary reduction of the electrical resistance under irradiation. Tests were made by Davidson and Geib (22) to determine the possibility of vulcanizing uncured natural rubber and a butyl stock by means of pile radiations. It was also desired to test the possibility 34

of introducing unsaturation into materials such as polyisobutylene, polyvinylchloride, acrylates, and others not mentioned, in order to make these materials vulcanizable. The effect of introducing a boron salt into the materials was also checked to determine the results of the reaction B10 (n,c) Li7. No procedure was reported to yield a cure of natural rubber at all comparable to sulfur vulcanization. No unsaturation was introduced into polyisobutylene. The butyl stock was permanently degraded by pile radiation. Natural rubber showed some radioactivity for days after irradiation, possibly because of its mineral content. Heisig (33) reported experiments in which propylene and cyclopropane were radiated by alpha particles from radon. In the propylene, a fog appeared shortly after mixing, and collected as a mobile, colorless oil. It was not identified. In the cyclopropane, a fog appeared after five minutes, and after six hours, a small pool of a mobile, colorless liquid condensate had formed. A preliminary report was made of a study of the effect of radon on methane, ethane, propane, butane, ethylene, acetylene, cyanogen, hydrogen cyanide, and ammonia (45). The reactions observed were as follows: oxidation of the foregoing except ethylene and hydrogen cyanide; hydrogenation of acetylene, ethylene, and cyanogen; polymerization of acetylene, of cyanogen, and of hydrogen cyanide. The effect on methane, propane, and butane was the production of liquid and, on further radiation, some light-yellow solid. Methane and hydrogen were produced in these reactions. Cyanogen yielded a brownish-black solid with the elimination of 5 per cent of nitrogen. Ethylene yielded a liquid and much free hydrogen. Acetylene yielded a yellow solid similar to cuprene and 2 per cent hydrogen. The oxidation of methane and ethane (under radiation) proceeded completely: with propane and butane, less completely. The oxidation of acetylene gave a clear, colorless liquid and no solid. The product combined with oxygen in the ratio 1:1, with slight formation of carbon dioxide. The oxidation of cyanogen gave a white powder thought to be (CNO)x, with some nitrogen and carbon dioxide. A 1:1 mixture of nitrogen and acetylene gave a yellow product, as with acetylene alone, but proceeded at a faster rate, which increased during the reaction. The nitrogen did not combine and was referred to as an ionic catalyst, i.e., the N2+ was believed to furnish additional clustering and polymerizing centers. Carban dioxide, hydrogen, and methane did not affect the rate. Twenty acetylene molecules polymerized for each N2+ ion, the same as for the C2H2+ ion. The N2+ had the same influence on the polymerization of cyanogen and hydrogen cyanide.

Gibbs and Lockenvitz (27) studied the relative molecular stopping power of n-butane, isobutane, and butene-l, butene-2, and isobutene. The specific ionization method, with the apparatus of Colby and Hatfield, was used to determine the extrapolated ionization range. Honig and Sheppard (36) compared the effects of deuter:ns and alpha particles on methane and n-butane. The products from the two types of bombardments were found to be quite similar. The liquid obtained from the butane under deuterons, showed a wide range of molecular weights with evidence of the presence of both olefin and ring structures. According to Viallard and Magat (85), the impact of electrons with energies of tens to hundreds of ev on polyatomic molecules produced ionized fragments and free neutral radicals in addition to ions of those molecules. In a homologous series, the percentage of ionized fragments produced with simple C-H bond rupture diminished with increasing chain length. The presence of multiple bonds increases this proportion. In fluorinated C chains, the formation of ions of free radicals is more probable than formation of ions of molecules. Simultaneous rupture of two C-C bonds is practically nonexistent, The percentage of CH3+ and CH:2+ does not increase with an increase in chain length. The ratios of CH2+/CH+ and C2H4+/C2H5+ approach limiting values for long chains. The rupture of the ends of saturated chains was infrequent. A double bond in the 1-2 position lessens the relative number of fragments obtained by cutting this bond and neighboring bonds. A double bond in the middle of the chain augments the number of fragments. The relative ionization probabilities of ten hydrocarbons and carbon dioxide, carbon monoxide, nitrogen, oxygen, nitrous oxide, helium, neon, and argon have been determined by Otvos (68) for beta particles from C140O and C 402 For a hydrocarbon series, the ionization probability increases linearly with the number of valence electrons: for periodic neighbors carbon, nitrogen, and oxygen, the ionization probability showed simple additive relationships based on valence electrons: methane and neon are isoelectronic, but the ionization probability for methane is higher, due to the lack of centralization of the nuclear charge. Ion ization probabilities at high energies of C14 beta particles bear no relation to ionization potentials or chemical properties, but seem to be governed by quasi-geometrical factors, such as molecular volume. The interpolation of ionization probabilities for high electron energies should be possible on the basis of valence electrons, distribution of nuclear charge, and the position in the periodic table, with values for He, Ne, A relating the first three rows. 36

Lind and Bardwell (48) reported work done on the radiation of saturated hydrocarbons by alpha particles. Hydrogen was liberated with the production of a liquid, which then continued to react under the irradiation. The liquid phase was unsaturated. For all saturated hydrocarbons except methane, one-fifth as much methane was liberated as hydrogen. The gaseous products contained only saturated compounds, which were either higher or lower homologs. A mixture of methane and carbon dioxide produced a wax-like solid under alpha rays. Lind, Bardwell, and Perry (50) reported considerable information ncm the chemical action of gaseous ions produced by alpha particles on unsaturated carbon compounds. Acetylene was polymerized by the radiation to give a light-yellow powder, with 20 molecules reacting for each ion pair which was calculated to be formed. The molecular weight of the resulting powder could not be found because of its insolubility. Further radiation of the powder liberated more hydrogen, probably as a result of further condensation of the solid. No methane was found. Radiation of cyanogen produced a black powder which gave off increasing amounts of nitrogen gas. It was concluded that the nitrogen formed had a catalytic effect on the reaction. Hydrogen cyanide was found to polymerize similarly to cyanogen. A dark solid with a reddish cast not apparent in the cyanogen polymer was formed, with the evolution of nitrogen and hydrogen. Ethylene condensed with the liberation of hydrogen and methane. The liquid which first formed became a solid under further radiation with increasing evolution of hydrogen and methane. Attempts at the hydrogenation of ethylene gave no evidence of such a reaction. In fact, hydrogen appeared to act as a center for reaction without actually reacting. In the attempted hydrogenation of acetylene, a solid formed, as it did with only acetylene present. The hydrogen was shown to act as an ionic catylyst and also to combine in some way. Hydrogen and cyanogen combined under alpha radiation in the ratio of 3:2 to form a dark reddish solid. Oxygen combined with cyanogen to give a yellow powder of formula (CNO), with no trace of the black powder which was formed by the radiation of the cyanogen alone. Carbon dioxide, nitrogen, and some carbon monoxide were also liberated during the radiation. The oxidation of acetylene gave a colorless liquid and no solid. The products were (C2H3)xand carbon dioxide, a fact which contradicted the former statements that a (CHO)x polymer resulted from this reaction. Lind and Schiflett (56) have reported on the oxidation of cuprene produced by alpha rays. Acetylene was polymerized by alpha particles from radon and the resulting polymer was analyzed. In an acetylene atmosphere, this polymer was insoluble in water, ethanol, ether, acetone, carbon disulfide, carbon tetrachloride, and benzene. The gases resulting from

the rapid oxidation of the polymer (the oxidation was followed manometrically) indicated 7.13 per cent CO and 92.85 per cent 02, with traces of CO2 and H20. The carbon-to-hydrogen ratio of the remaining solid was 11.65 and 12.08. The amount of 02 which reacted was 5-1/2 times the CO formed. This ratio suggested an oxidation solid of C 4O or an original polymer of (C2H2)20. The heat of polymerization was calculated as 47 kcal per gram-mole of C2H2 reacted. Mund and Koch (66) have investigated the rate of polymerization of acetylene under alpha radiation, and have also investigated the influence upon the rate of oxygen, pressure, and temperature. They concluded that these influences had no effect on the number of molecules of acetylene that were polymerized per ion pair produced. It was also found that 20 molecules of acetylene reacted per ion pair formed. Rosenblum (72) has given a short review of the identification of benzene in acetylene radiated by alpha rays from radon. Approximately one fifth of the reacting acetylene was utilized for benzene formation. A small amount of benzene was believed to react further. The probable reason that benzene was missed by earlier workers was that 50 to 100 fold higher radiation was used formerly, and this accelerated the further reaction of benzene. The theory was presented that the reaction proceeded by successive bimolecular reactions between normal acetylene molecules and excited molecules or polymers. It was suggested that benzene forms from a cyclization of the activated trimer. It was also concluded that benzene is formed in the polymerization of acetylene by beta and gamma rays. Fricke, Hart, and Smith (25) irradiated gas-free aqueous solutions of CO, alcohols, aldehydes, ketones, and acids in the concentration range 10 micromolar to 1 M, between pH of 1 to 13, and studied the reactions principally by gas analysis and potentiometric acid analysis. Oxidation and condensation reactions with the evolvement of gaseous hydrogen were observed. CO2 was produced from certain acids, especially from those having an oxygen-containing group in the alpha position. No liberation of CO, hydrocarbons, or 02 was found. The pH of the solutions affected both the rate and the nature of the reactions. Newton (67) subjected a number of alcohols, in the liquid state, to high-energy alpha particles. Various oxidized and reduced products resulted, as well as methane and hydrogen. As the alcohols became more branched, the methane yield increased, while the hydrogen yield decreased, the latter seeming to indicate that the hydrogens on the carbon to which the hydroxyl is attached were especially subject to attack. No evidence of polymerization was noticed. The mechanism involved was thought to be excitation and ionization with the formation of free radicals. 38

The effects of radioactivity on fatty acids has been reported by Shepard and V. Burton (75). They tested the hypothesis that radioactivity might be a factor in the production of petroleum from gaseous paraffins. Several fatty acids which included acetic, caprylic, lauric, and palmitic acids, were bombarded with alpha particles from radon. The production of a gas which consisted of H2, CO, C02, H20, CHi), and higher hydrocarbons was proportional in the initial stages to the fraction of radon decayed. Under radiation, lauric acid and palmitic acid yielded n-undecane and n-pentadecane, respectively, as products, indicating decarboxylation to be predominant. The experimental procedure and results of the radiation were discussed. The processes resulting from the bombardment can be summarized as follows: dehydrogenation, decarboxylation, the formation of low-molecular-weight water-soluble acids, the formation of methane and other hydrocarbons, and the formation of CO and H20. The effects of the bombardment of oleic acid with deuterons was discussed by V. Burton (12). Of the original acid 31 per cent was unreacted. Of the converted material, 10 per cent was non-saponifiable material, 52.5 per cent was a polymerized acid, and about 1.7 per cent was stearic acid. The presence of the stearic acid in the products indicated that hydrogen produced under the influence of radioactivity could be removed from the gas phase by reaction with the unsaturated components produced during the bombardment. Hart (31) has recently studied the mechanism of formic acid oxidation by gamma rays in air-free aqueous solutions. It was concluded that the oxidation occurs as a result of a reaction between formic acid and the H and OH free radicals produced by the action of the gamma radiations on the solvent water. The effect could be varied by the addition of hydrogen peroxide, and it was suggested that the reactions proceed by a chain mechanism. Work done by Penneman (69) on the effects of radiation on aqueous carboxylic acid solutions shows that for various amounts of x-, electron, and deuteron radiation, both the reducing and acid equivalents of oxalic and formic acids are decreased. Quantitative data are given. Some types of electrical discharges have been found to exert influences upon saturated hydrocarbons similar to the effects produced by alpha particles. Cathode rays also were capable of polymerizing acetylene. Gaseous and liquid reaction products were obtained by Lind and Glockler (51) from the action of a 12,000-volt silent discharge upon gaseous ethane. The composition of the gas and liquid produced corresponded to CnH1l8n and agreed rather closely with the composition of the products obtained by irradiating ethane with alpha rays. About 10 kwh of electricity was usedtp produce 5 grams of oil. 39

The effect of electric discharge on ethane and the subsequent control of the liquid hydrocarbons produced have been reported by Lind and Glockler (52). Silent, corona, and high-frequency discharges caused condensation of the ethane to liquid with the liberation of various gases. In semicorona and corona discharges, cracking was apparent, resulting in the formation of free carbon. In corona discharge a solid film deposited on the wall. Variation in the molecular weight of the liquid products appeared to bedependent upon the time the earlier products had remained in the discharge subject to further ionization. The average molecular weight was regulated between 467 and 105 by controlling the time of reaction. Evidence of a delayed condensation was attributed to "open bonds, " which react slowly to form liquid without repeated ionization. Methane, ethane, propane, butane, and ethylene were condensed to liquid and solid hydrocarbons in a semicorona discharge (53). Hydrogen and methane were eliminated, as with alpha rays. The liquid products from different hydorcarbons or from the same hydrocarbon in different tubes, were similar in physical properties and were complex. The solid products were gummy, resinous, and inert toward solvents and reagents except strong oxidizing agents. The extent of the reaction was found to be dependent upeon time. Paraffins, cycloparaffins, olefins, cyclo-olefins, and aromatics were irradiated by Schoepfle and Fellows (73) with cathode rays at 170,000 volts and 0.3 ma. The total quantity of gas released from the hydrocarbons was largest in the case of paraffins and decreased in the order in which the compounds are named. In general, as the molecular weight of a given series increased, the percentage of hydrogen in the gas given off increased, and the percentage of methane decreased. The branched-chain compounds gave higher percentages of methane and of gaseous saturated hydrocarbons than the straight-chain compounds. Ozonizers were used to pass an electrical discharge through butane (54). The experimental procedure was given for the preparation of about one liter of liquid. The liquid was fractionated into three fractions. The light fraction I was fractionated into eleven subfractions. Light fraction I-6 (the largest) was refractionated. Light fraction I-6-2 was examined. The properties resembled those of 2,4-dimethylhexane and 2-methyl-3-ethylpentane. The density, C-H ratio, molecular weight, halogenation number, and freezing point -- all indicated unsaturation. Presumably the products were octylenes. It was not yet possible to identify the isomers.

The condensation of hydrocarbons in ozonizers was compared with the condensation by alpha particles (55). The similarity of these processes was confirmed by the following: the analogy of pressure changes during reaction; the amount of free hydrogen produced is of the same order for both; the similar percentage of hydrogen in the hydrogen-rethane gas phase for both types; the similar percentages of liquid conversion; the similar composition of the liquids approximating CnH2n, as calculated from the analysis of the gas phases; and the similarity of the amounts of total hydrocarbon reacted in both types of processes. Unsaturated hydrocarbons were suspected in the gas phase produced by stopping the condensation of butane at the point of maximum pressure. Loiseleur, LatarJet, and Crevisier (60) have carried out work on oxygen containing organic compounds. Hydrogen peroxide and organic peroxides were formed when 0.01 - 0.00001 M solutions of crotonic, succinic, fumaric, acetic, and benzoic acids, formaldehyde, methanol, and ethanol were irradiated with x-rays. Stein and Weiss (74) investigated the effects of ionizing radiations upon aromatic organic compounds. Benzene suspended in oxygen-free water was radiated with 106 roentgens of x-rays. Analysis showed formation of a trace of phenol. Diphenyl was also isolated. Benzoic acid solution treated similarly formed 0.1 millimole of hydroxybenzoic acids, and salicylic acid was isolated. Similar experiments using a neutronalpha ray source (radium and beryllium powder) resulted in the products mentioned above and catechol, together with products from opening of the ring. The total yield was stated to depend upon pH. The formation of diphenyl suggested free phenyl radicals in the reaction. It might be noted here that the free-energy change for the reaction resulting in phenol from benzene is a positive quantity, indicating that additional energy must have been supplied from the x-rays in order to cause the reaction. The polymerizations which were reported served to substantiate the idea that gamma rays may be used to produce the same reactions which are found as a result of alpha and beta radiation. Breger and Burton (6) have studied the effect of alpha particles and deuterons on e naphthenic acid. Cyclohexanecarboxylic acid was used to determine whether anticipated decarboxylation would lead to forration of ring compounds or whether ring cleavage would lead to formation of straight-chain hydrocarbons. Analysis of material subjected to alpha particles showed decarboxylation with some dehydrogenation. The results showed little or no difference between the chemical effects of alpha particles and deuterons. The experimental data given dealt with the bombardment and with the analysis of the resulting mixtures. It is presumed that cyolohe.ane and cyclopentane rings were not opened by the bombardment.

Radiation of nitrogenous compounds by x-rays has been found to produce deamination (20), The ion yield for the reaction increased approximately exponentially with increasing concentration of the aqueous solution. Aqueous solutions of L-serine of varying concentrations were given x-radiation (18). Because of the greater solubility of this compound, higher concentrations were possible than had been previously obtainable. It was found that ionic yield increased with increasing concentration and then, at highest concentrations, "leveled-off."' Solutions of glycine were deaminated by alpha radiation (19). However, the ionic yields for this reaction using alpha radiation, were only 15 to 19 per cent of that produced by x-radiation. This lower ion yield for alpha rays than for x-rays is an unusual circumstance. The rates of reaction were measured by Alyea (4) in a solution of chlorine in benzene and in a solution of oxygen in sodium sulfite solution, both with and without radon present. Earlier work was cited regarding similar treatment of mixtures of H2 and C12 and of CO and C12. The ratio M/N varied from 700 to 200,000, depending upon the purity of the materials tend the intensity of the radiation. The data could be explained more readily in terms of a chain mechanism rather than in terms of the "ion cluster" theory. The rate of decomposition of chloroform by radon was found by Harker (30) to be greatly influenced by the presence of the products of decomposition (C12 and HC1). The presence of iodine in a potassium bisulphite solution increased the rate of oxidation of the latter under gamma radiation. A series of articles on chemical actions of ionizing radiations on aqueous solutions is currently being published in the Journal of the Chemical Society (British). The purpose of this series is to study the action of the radiations, to study the reactions of the free radicals formed by the radiation in the absence of interfering reagents, and to study the products of those reactions which are similar to those in biological systems (24). X-rays, neutrons, and alpha rays are used in these studies, Solutions of benzene and benzoic acid in water were irradiated by x-rays. Phenol, diphenyl, and terphenyl were produced from the benzene (79) (cf. Stein and Weis (74)). Salicylic acid and p-hydroxybenzoic acid 42

were the main products isolated from the benzoic acid solution. Results of varying the atmosphere above the solution and comparison of energies involved in alternative reaction paths led to the conclusion of a free radical mechanism for these reactions. When benzene was subjected to bombardment of alpha particles and neutrons, it was found that in addition to phenol and diphenyl, which had also been produced by gamma and x-radiation, polyphenols and a straightchain dialdehyde were formed (80). This additional reaction could be attributed to further reaction of the phenol molecules due to the intensity of ionization, hence free radicals, along the tracks of the bombarding particles. Aqueous solutions of glycine, alanine, and serine were irradiated by x-rays, under various conditions (81). Deamination occurred, giving ammonia, molecular hydrogen, and aldehydeso It was concluded that both the atomic hydrogen and hydroxyl radicals produced by the x-radiation attack the amino acid, and an oxidative and reductive mechanism are operative in the deamination. Saturated solutions of nitrobenzene in air-saturated water were irradiated with 3o5 x 104 energa units of X-ray. (58).;Each 200 ml of this solution yielded about 10- moles of the mixed phenols of nitrobenzsee. Solutions of cholesterol and 3-B-hydroxypregn-5-en-20-one, which are both naturally occurring steroids, were irradiated with 10 r of x-rays (38)o The isolated double bond in the sterol ring was attacked by hydroxyl groups, resulting in adjacent OH groups in the ring. The labile hydrogens adjacent to the double bond in the cholesterol were also attacked, leaving a ketone group in this position. The mechanism for this change may also involve free radicals. Solutions of the sodium salt of cholic acid in water were subJected to x-radiation (39). The product which was isolated was 3a:12Gdihydroxy-7-keto cholanic acid. This reaction represents the change of the hydroxy group in the 7-position to a keto group. The 7-position in cholesterol was similarly attacked (39). Both these reactions can be explained by attack of OH radicals which are produced by the radiations and subsequent elimination of water to leave the keto group in that position. Radiation of 0.2 per cent aqueous oxygen-containing benzoic acid solutions produced a yield of the mixed isomers of salicyic acid (59). Up to a dose of 5 x 104 energy units, the formation of salicylic acid, in the presence of oxyge$, is a linear fAnction of the dosage. The yield for 200 ml of solution given 5 x 10 energy units was approximately 1.2 Ag.

Both alkaline and acid solutions of (+) - oestrone-b were irradiated by x-rays (40). The resulting compounds from both were identical. This compound was apparently a lactone similar to that produced by hydrogen peroxide or other oxidation methods. In effect, the carbon-to-carbon bond at the 17-position in the five-membered steroid ring is broken, and a six-membered lactone ring is formed. If attack by OH radicals on a double bond formed by enolization is assumed, it is most probable that the ring is broken between the carbons of 16- and 17-positions. c. Literature Review - Theoretical: In the earlier experimental work done in the field of radiation chemistry, and especially in that done by Lind and his associates, large yields per ion pair and polymerization were often explained by a "cluster theory", which assumed the groupIng of molecules around the central ion aFs being responsible for these effects. Later work has, in general, disproved this original theory. Eyring, Hirschfelder, and Taylor (23) regarded clustering as playing a very minor role. They proposed a mechanism of ionization involving formation of excited molecules, ions, and radicals. This mechanism may be summarized by equations for three steps (as reviewed by Burton (10)): ALA+ + e 1. Ionization A + C +e SB+ C + e 2. Discharge A+ + e — A* (excited) stable molecules 3. Decomposition A* m c free radicals Burton (9) has given a unified picture of the theory involved in radiation chemistry. He differentiates radiation chemistry from photochemistry in terms of the energy involved. In the primary acts (in radiation chemistry), electrons are released and trapped at some remote point. The succeeding processes depend upon the nature of the ions involved and their stability in their environment. The significant reaction, where solvation does not occur is given as, AB+ + e —A + B. or, AB+ + M —PA + B + M. (ci. Eyring, Hirschfelder, and Taylor (23)). According to the FranckCondon principle, the electron moves to the positive ion in such a short

time that the constituent atoms are left in positions whose energy states are above those necessary for dissociation of the bond concerned. This state leads to the reaction given above. As the size of the molecule increases, the ionic configuration is more nearly like that of the unchanged molecule, so the energy of neutralization may not be localized enough for bond rupture to occur within one vibration period. As a result, the process of relocation of potential energy of the molecule may lead to decomposition to ultimate molecules before rupture can occur. This process may be important in many cases where free-radical decomposition had been assumed. In the liquid state, the energy could leak from the excited molecule before decomposition can occur, thus lowering the decomposition yield. In photochemistry, it is often possible to choose wave lengths which promote a reaction in one direction, and therefore do not aid the reverse reaction. However, in radiation chemistry, this does not appear to be the case; hence a steady state may be attained in which the forward and reverse reactions take place at equal rates. Alpha particles, deuterons, and protons rarely make direct nuclear impact, but rather, cause a large degree of ionization along their paths. An energetic electron, or beta particle produces a much smaller degree of ionization, which is also more diffuse and homogeneous. Gamma and x-rays interact with molecules to produce ions and energetic electrons, which in turn, are responsible for much of the observed effects. Fast neutrons scatter the nuclei with which they collide, and for sufficiently high velocities, any ejected nucleus may leave one or more electrons behind it. Fast neutrons were found to cause displacement of atoms of solids from their lattice positions. This was called the "discomposition" effect. Coloration of ionic crystals was explained on the basis of electrons trapped in negative ion vacancies. Effects of radiation on water may be represented by the reactions given previously. Tables of data for the effect of 170-kv cathode rays on various hydrocarbon compounds show that the methane yield increases with the number of methyl groups. Unsaturation tends to decrease hydrogen yields, and increase polymerization. The hydrogen yield decreases with increasing complexity of structure in accordance with the principle of increased probability of internal conversion with increased molecular complexity. Burton has also discussed the effects of radiation on organic compounds in a paper presented before the Symposium on Radiation Chemistry at the 110th Meeting of the American Chemical Society at Chicago, Illinois, in September, 1946 (11). He states his views in the summary of the paper: "All the processes which occur in photochemical reactions of organic compounds occur also in radiation-chemical processes. In addition, there are

reactions resultant from the peculiar sequence characteristic of radiation chemistry: i.e., ionization, discharge, and decomposition. In general, any electron in the molecule is equally susceptible to ionization in the initial act; this fact must be constantly recalled in any interpretation of radiation-chemical mechanisms. "Since, in general, the excitation energy lies in any part of the molecule, the yield of a particular product is closely related to the number of parent groups in the molecule. Gas production, particularly in unsaturated compounds, is an inadequate criterion of the resistance of a compound to high-energy radiation. In the liquid state, the excessive excitation energy tends to minimize the Franck-Rabinowitch effect (i.e., decrease in yield due to collisional deactivation and cage effect). Factors which increase resistance of organic compounds to radiation (and ratio of ultimate molecules to free-radical processes) are molecular complexity, resonance in the molecule, and all properties of the molecule which tend to increase the correspondence between ionic and molecular configurations. Among the latter are molecular symmetry (cf. benzene) and molecular size (cf. palmitic acid). Apparently, increase of molecular size tends to channel the decomposition along a particular path rather than to diversify the products." Dainton, in 1948, gave a report on radiation chemistry in the British Annual Reports on the Progress of Chemistry (17). He defines '"radiation chemistry" as chemical effects produced by the absorption of all types of rays whose energy is above 50 ev which result from radioactive processes or by the absorption of electrons or positive ions of similar energy. The sources of various radiations were discussed, as well as the dosimetry. Positively charged particles lose their energy by elastic impacts with particles in their path. The average energy dissipated per ion pair is about 30 ev. Electrons, being of low mass, are easily deflected, giving badly defined tracks.. They lose energy by elastic impact and by the production of bremsstrahlung. Photons must be absorbed in a single elementary act. Those of high energy have three modes of aosorption: ejection of a photoelectron usually from the K-orbit, Compton scattering, and positron-electron pair production when the energy is high enough. In the primary act, the charged particle leaves a path of positive ions, surrounded by a more distant field of the electrons which have been knocked out by the particles. The positive ion may or may not dissociate. The free electrons may be captured by neutral atoms to produce negative ions. Production of new, uncharged species may occur by a charge-neutralization process, or directly, when the molecules can be excited to nonionic repulsive levels. The "cluster" theory, which was formerly proposed, was discarded in favor of the "atom-radical" theory 46

(cf. lyring, Hirschtf;lder, and Taylor (23)). Several reaction 3ystems are discussed. In many of the single inorganic substances, the yield is so large as to suggest that more than one radical is formed per ion pair. A discussion of the theory and kinetics of the effect of radiation on water is given in this report by Dainton. Burton, in one section of his report for the Notre Dame Symposium (10) has summarized the effects of the types of particles. Energetic heavy particles produce one ion for every 5-10 molecules of path, while electrons with the same velocity produce one ion in every 500 molecules of path. In the theoretical discussion of their work, Sheppard and Honig (76) show why the alpha particles and deuterons should produce similar chemical effects. They also point out that the amount of change is proportional to the number of ion pairs formed, because for every ion pair formed, a given amount of energy is absorbed and a given amount made available for each electronic process. This rule implies that the total amount of reaction should be proportional to the amount of radiation, which in turn produces the ion pairs. The number of molecules which condense per ion pair formed (-M/N ratio) is stated by Heisig (34) to be highest for the substances having negative heats of formation from the elements in their standard states. The condensation process for saturated hydrocarbons is nearly isothermal, having small -M/N values. For unsaturated hydrocarbons, -M/N values vary with their negative heats of formation. Their condensation is exothermic. Experimental results and theoretical interpretations were given by Toulis (83) for the decomposition of water by radiation. He concurs with the hypothesis that the primary process in water is the creation of H-atoms and OH free radicals. The decomposition of water was found to depend upon the ratFr of energy loss of the radiation. X-rays, gamma rays, electrons, and extreme ultraviolet light and particles losing energy at a rate less than 70 mev/gm/cm2 of water showed little effect. The outstanding property of a given reaction was thought to be the probability of the capture of a free radical while in a solvent cage. Emphasis was placed upon this concept rather than upon the usual rate constant for a chemical reaction. The probability of capture is independent, to a first approximation, of a solute and depends only upon the type of radlcl. with which the radical in question reacts. The types of reactions possible are: radical-radical, H-radical-molecule, and OH radical-molecule. Studies were made of the rate of decomposition of both pure (conductivity) water and of aqueous solutions of H2, 02, and H202 under the influence of 47

x-rays. Most data were evidently deleted from the text. X-rays were used in preference to particles because the x-rays gave a more nearly uniformly distributed effect throughout the solution, and simplified calculations resulted thereby. The dissolved impurities were found to influence the equilibrium compositions and rates of reaction, apparently by capturing free radicals. A further comment worthy of note was that in the case of ionizing radiation (whether photon or pa3r!icle, was nat stated) the free radicals formed were segregated into two coaxial cylindrical regions about the track of the radiation. The H were in the outer 0 s cylinder, of 150 A diameter, and the OH, in the inner cylinder, of 8 A diameter. Thus, back-reaction to H20 was hindered in preference to other reactions. Allen (1) in a review of existing data indicates that covalent compounds are decomposed by ionizing radiations. The change of rate of reaction with progressive conversion is discussed. A mechanism involving free radicals is proposed for the decomposition of water. This proposed presence of free radicals is used to explain the great effect of dissolved solutes upon the behavior of irradiated aqueous solutions. Another article by the same author (2) has a discussion of the effects of inizig radiations upon chemical compounds in various physical states of aggregation. Several possible theories were discussed for the mechanism by which a chemilcal reaction proceeds as a result of irradiation of the reactants. The effects of excited and ionized molecules, positive ions, energy absorption by inert gases present, decomposition of ions, reactionLs of ions with molecules, reaction of an electron with a molecule having electron affinity, breaking of bonds before neutralization of an ion, formation of "clusters", significance of ion yield, and production of chain reactions were all discussed in their relation to reaction mechanism. The reversibility of reactions induced by radiation was discussed, as was the effect upon equilibrium of certain "promoters" or "inhibitors". Some basic differences were pointed out in the behavior of covalent, ionic, and metallic solids under irradiation. A general theoretical discussion was tendered by Steacie (78), showing similarities between photochemistry and radiation chemistry and posing questions resulting frcm their differences. The discussion included the primary process, the secondary process in respeft to ions and in respect to excited molecules, and the application of the knowledge of thermal reactions to the investigation of the secondary processes. Many references to experimental data were cited. 48

It was pointed out by Garrison (26) that several alternative hypotheses are available to explain the polymerization of acetylene by ionizing radiations. The ion cluster theory of Lind was mentioned. Alternative to this theory were suggested the following mechanisms: the action of C-2H2+ ions as acid catalysts by combining with the negative carbon atom of the ionic-resonance-form of acetylene; and the polyrmerization via a free radical mechanism because of the unshared electron in i2n 2+ A discussion of radiochemistry, a comparison with photochemistry, and the ion-excitation theory (Eyring, Hirschfelder,and Taylor (23)), are given in an article by Wildschut (89). He also gives a mechanism for polymerization of hydrocarbons starting with the ion produced in the primary act, which is similar to that given by Garrison above. The mechanism of the radiochemical reactions in aqueous solutions was discussed by Weiss (87). An attempt was made to interpret the facts of the radiochemistry of solutions on the basis of known photochemical and chemical reactions in solutions. It was stated that the products to be obtained from the irradiation of solutions depend upon the nature of the solute and the pH of the solution. The pH and nature of the solute both determine the oxidizing properties of the solute. A reducing solute would react with OH radicals, and free H2 would be produced. An oxidizing solute would permit oxygen to be freed by combining with N radicals. The general principles involved in the chemical and biological action of radiation are examined by Weiss (88). An important difference between photochemistry and radiation chemistry is that in the latter, the absorption of radiation energy is not specific and is approximately proportional to the mass but almost independent of the chemical linkage. Therefore, in dilute solutions, most of the energy is absorbed by the solvent, so that most of the primary changes must take place in that medium. The direct or indirect action of the radiation may, however, lead to the same qualitative result. It was found that the recombination process following primary formation of radicals is of considerable importance. If recombination can be neglected, then the effects are approximately independent of the nature and wave length of the radiation and depend only on the total dosage. Lind and Vanpee (57) studied the effect of xenon ions in the chemical action of alpha particles. Xenon has a higher ionization potential than acetylene. Hence ionization pasges from Xe to acetylene by collision. Therefore these two gases can not be used to prove that the nature of an ion is indifferent in causing the polymerization of

acetylene to cuprene. Both H2 and 02 have ionization potentials higher than Xe, so that the reaction 2H2 + 02 = 2H20 can be used to test the effect of added Xe+ ions. The Xe+ ions constituting 70-95 per cent of the total ionization caused the reaction to proceed 12 times as fast as in their absence. Xe+ ions added to 2C0 + 02 = 2C02 gave a slight positive effect, The presence of Xe+ ions in the reaction of CO to a carbon suboxide had no effect. Zimmer (91) has stated that there is no indication of a transfer of energy from solvent to solute. His experimental data seem to indicate the questionability of the hypothesis of activated solvent molecules and point to a possibility that the energy transfer takes place by diffusing molecules. However, a paper which was submitted by Manion and Burton (64) at the Symposium on Radiation Chemistry at the 119th Meeting of the Americal Chemical Society in April, 1951, reemphasizes the significance of ionization transfer in radiation chemistry, especially in the liquid state. Studies of hydrocarbon mixtures radiated with 1.5 mev electrons show results which are explained in terms of ionization and excitation transfer, as well as removal of free radicals through attack on unsaturated bonds. In a recent article (61), Magee has outlined a model of a system being radiated with particles, from which an equation is set up which allows mathematical treatment of the effects arising from the variation of ionization density due to the lack of homogeneity caused by the tracks of the charged particles. Most former considerations had assumed that all intermediates are created homogeneously in space. In a further discussion (63), Magee and Burton have considered the negative ion formation by electron capture. When its energy is sufficiently low, an electron may be captured by a neutral molecule to form a negative ion. If a thermal electron is to be captured in a dissociative process, the electron affinity of the ion produced must exceed the strength of the bond which is ruptured. Magee and Burton have published a theoretical discussion (62) of the mechanism by which the electron is captured in the process of discharging the positive ion produced by an ionizing radiation. Capture of the electron most prQbably leads to formation by dissociation of two particles, one of which is excited. Dissociation into radicals is favored over dissociation into molecules. However, in the liquid state, the production of ultimate molecules increases in importance. 50

In the precedinx.review only those papers were discussed which are closely related to the objectives of this *ork. For a comprehensive list of references related to radiocatalysis, one may turn to the A.C.S. Monograph by Lind (41) whiih lests 508 references, Dainton's review (17) which lists 149, the symposium in J. of Phys. and Colloid. Chem. (82) with 339 references, and the sectioon on amm"onia in Gielins Handbuch (28), with 131 references. 3. Proposed Research on Selected Reactions a. Proposed Research - Inorganic Reactions: It is proposed to study the formation of ammonia from the elements under the influence of x-, beta, and gamma radiation. It appears that if ammonia can be synthesized in commercially attractive yields by this process, an important supplement can thereby be provided to facilities for the production of ammonia. The reason for this statement is that evidently the reaction can be made to proceed at ordinary temperatures. Considerable savings in ammonia-plant maintenance and construction should result if the conversion step can be carried out at considerably decreased temperatures. In addition, of course, the use of lower temperatures would favor the presence of ammonia in the equilibrium mixture. There are indications from thermodynamic data, as is well known, that nitric acid can be formed directly from its elements. It is possible that the reaction forming water may predominate; however, it is proposed to test the possibility of producing HO03 from H2, N2, and 02. It is proposed to conduct preliminary studies on the oxidation of sulfur dioxide by means of radiation. If this reaction were to take place at ordinary temperatures or above, it should prove interesting in sulfuric-acid manufacture. A single stage of adiabatic reaction might prove feasible in the conversion step, with the final temperature still low enough to produce near-quantitative yields. This might prove attractive., especially if it were necessary to use recovered sulfur dioxide from smelter operations, etc., for raw material. A careful economic study of plant performance, freight rates, etc., would be necessary to determine the commercial importance of this reaction, should the reaction prove possible. Various means have been proposed for the recovery of elemental sulfur from hydrogen sulfide. It is proposed to study the direct oxidation of hydrogen sulfide under the influenc of radiation in order to determine the extent to which oxidation would proceed under the influence of radiation. 51

Ilemental sulfur may also be recovered from stack gases. A irtur. of sulfur dioxide, carbon monoxide, and water has free-energy reJationships such that it should be completely converted to carbon dioxide Id hydrogen sulfide if at equilibrium at 250C and one atmosphere pressure. If equilibrium could be partially attained under irradiation, the resulting hydrogen sulfide and unreacted sulfur dioxide would react in the presence of moisture to yield elemental sulfur. This reaction would facilitate tremendously the processing of power-plant, smelter, and other wIste gases to avoid atmospheric pollution. At the same time the recovery o elemental sulfur should permit attractive pay-off times for the installations, while adding to the national supply of elemental sulfur. Carbon monoxide might be oxidized to carbon dioxide by the use of radiation as a catalyst. An important purpose of this reaction could be to supplement existing methods for purifying the exhaust gases from internal-c ombust ion engines. b. Proposed Research - Organic Reactions: Since polymerization -ad/or dehydrogenation have previously been reported under the influence of radiatioAl, these types of reactions will be attempted first. Success in simple reactions would lead to the trial of more complex reactions. Acetylene can be polymerized to benzene within a wide range of temperatures if the reaction can be activated. Radiocatalysis may bring about the desired activation. More interesting is the fact that at certain temperatures the free-energy change for the conversion of ethylene or ethane to acetylene or benzene becomes negative, indicating a possible reaction, The formation of acetylene from methane by dehydrogenation is thermodynamically possible at elevated temperatures. An alternative method for attempting to produce acetylene from methane might be the partial oxidation of the methane, with an accompanyig polymerization to the higher hydrocarbon. Since radiation may produce polymerization reactions, it is conceivable that the desired oxidation and polymerization might be promoted by a radiocatalyst. According to free-energy data, the reaction involving the partial oxidation of methane should yield a favorable proportion of acetylene at room temperature. At sufficiently high temperatures the conversion of methane or natural gas to benzene has been accomplished (29); however, the yields reported were small and considerable coking occurred, It is hoped that under the influence of radiation, suitable yields of benzene might be obtained from methane. The commercial importance of this conversion can hardly be overestiated. A process which could convert natural gas to bmoa or its. deritives would be of great iorane to the chemical her.~~~~~~~~5

It may be worthwhile to attempt the hydrogenation of hydrocarbons, although the literature seems to indicate that the reverse reaction is predominant under radiation. Three reactions which resemble hydrogenation are the FischerTropsch, methyl alcohol, and "Oxo" syntheses. All three of these involve the hydrogenation or partial reduction of carbon monoxide with hydrogen. Special catalysts and conditions are needed for all three syntheses. The conditions and thermodynamics of the methyl alcohol synthesis are very similar to those for the production of ammonia. Since the literature has reported ammonia syntheses effected by means of radiation, the synthesis of methyl alcohol may be worth studying. In the Fischer-Tropsch process carbon monoxide and hydrogen are used to produce hydrocarbons containing varying numbers of carbon atoms. Therefore, some sort of polymerization is involved. Since radiation has often been shown to produce polymerization, this reaction appears to merit some study. The "Oxo" process, producing, as a final product, aldehydes and primary alcohols, also involves polymerization. A type of reaction which possesses interesting industrial possibilities is the hydration of unsaturated hydrocarbons to alcohols. This reaction normally does not proceed easily, but under the influence of radiation it might be caused to proceed fairly rapidly. The production of small amounts of phenol from benzene and water was reported in the literature (74). Apparently, under the conditions reported the free-energy change was positive. If conditions could be found to give a negative free-energy change, the promotion of this reaction by radiation might be industrially feasible. It might also be possible to conduct this reaction as a partial oxidation, similar to the production of acetylene from methane. Many other organic reactions such as cracking, condensation, isomerization, cyclization, esterification, and nitration, may be studied at a later date. The reactions which have been chosen for the initial work are those which seem to have the greatest promise, Judging from the experiments reported above in the literature survey. 53

Bibliography for Project M943C 1. Allen, A. 0., Chemical Effects of Ionizing Radiation on Simple Inorganic Compounds and Aqueous Solutions. U. S. Atomic Energy Commiss'ion, MD)DC-363, 1946. 2. Allen, A. 0., Effects of Radiation on Materials. U. S. Atomic Energy Commission, MDDC-962,, 1947. 3. Allen, A. 0., 'Radiation Chemistry of Aqueous Solutions," J. Phys. and Colloid. Chem. 2_ 479-90 (1948). 4. Alyea, Hubert N., "Chain Reactions Produced by Light and by Alpha Radiation," J. Am. Chem. Soc. 52, 2743 (1930). 5. Boull, Andre, "Catalysis by Cathodic Projection," Bull. Soc. chim. 10, 361-71 (1943). 6. Breger, Irving A., and Burton, Virginia L., "The Effects of Radioactivity on a Naphthenic Acid," J. Am. Chem. Soc. 68, 1639-42 (1946). 7. Broda, E., "Mode of Chemical Action of X-rays on a Non-aqueous Solution," Nature 151 448 (1943). 8. Burr, J. G., and Garrison, W. M., The Effect of Radiation on the Physical Properties of Plastics. U. S. Atomic Energy Commission, AECD-2078, 1948. 9. Burtonp Milton, "Radiation Chemistry," J. Phys. and Colloid. Chem. 51, 611-625 (1947). 10. Burton, Milton, "Radiation Chemistry IV. An Interpretation of the Effect of State on the Behavior of Some Organic Compounds and Solutions," J. Ph s. and Colloid. Chem. 52, 564-577 (1948). 11. Burtonp Milton, "Effects of High-Energy Radiation on Organic Compounds," J. Phy and Colloid. Chem. 51, 786-797 (1947). 12. Burtonp Virginia L., "The Effects of Radioactivity on Oleic Acid," J. Am. Chem. Soc. 71, 4117 (1949). 13. Chapiro, Adolphe, "Polymerization by Gamma Rays," Compt. rend. 228, 1490-2 (1949). 14. Coolidge, W. 0., "High Voltage Cathode Rays Outside the Generating Tube," Science 62_ 441-2 (1925). 15. Daintoni F. S., "On the Existence of Free Atoms and Radicals in Water and Aqueous Solutions Subjected to Ionizing Radiation," J. Phys. and Colloid. Chem. 52, 490-517 (1948). 16. Dainton, F. S. '"Effect of Gamma and X-rays on Dilute Aqueous Solutions of Acrylonitrile," Nature 16_, 268 (1947).

Bibliography (cont'd) 17. Daintonm F. S., "Chemical Reactions IndUced by Ionising Radiations," Annual Reports on the Progress of Chemistry, vol. 45, 1948. Landon, Chemical Society, Richard Clay and Co. Ltd., 1949 (149 references). 18. Dale, W. M4, and Davies, J. V., "Deamination of Aqueous Solutions of L-Serine by X-Radiation," Nature 166 1121 C1950). 19. Dale, W. M., Davies, J. V., and Gilbert C. W., "The Deamination of Glycine by Alpha-Radiation from the Disintegration of Boron in a Nuclear Reactor," Biochemical. J4i5 543-.6 (1949). 20. Dale, W. M., Davies, J. V., and Gilbert, C. W.,. '"The Kinetics and Specificities of Deamination of Nitrogenous Compounds by X-Radiation," Biochemical J. 45, 93-99 (1949). 21. D'Alieslager and Jungers, Bull. Soc. chim. 31g. B 1e 75 (1931). 22. Davidsoan W. L., and Geib, G., "The Effects of Pile Bombardent Uncured Elastomers," U. S. Atomic Energy Commission, MDDC-1449 (l1947). 23. Eyring, H., Hirschfelder, J. 0., and Taylor, H. S., 'The Radiocheuical Synthesis and Decomposition of Hydrogen Braide," J. Chem. s. 570-5 (1936). 24, Farmer, F. T., Stein, G., and Weiss, J., "Chemical Actions of Iontsing Radiations on Aqueous Solutions. Part I. Introductory Remarks amd Description of Irradiation Arrangements," J. Chem. Soc. 3241-5 (1949). 25. Fricke, Hugo, Hart, Edwin J., and Smith, Homer P., "Chemical Reactions of Organic Compounds with X-ray Activated Water," J. Chem. Phys. 6 229-40 (1938). 26. Garrison, Warren M., "On the Polymerization of Unsaturated Hydrocarbons by Inizing Radiations," J. Chem. s 15 78-9 (1947) 27. Gibbs, Thorns E., and Lockenvitz, Arthur E., "A Comparison of the Relative Molecular Stopping Power of Some Eydrocarbon Isomers for Alpha-Particles from Polonium," 'ys. Rev. 73, 652 (1948). 28. Gmelins Handbuch der anorgranischen Chemie, 8. Aufl. Syst. Nr. h4, Lef. 2, Berlin 1936. Deut. Chem. Ges. (131 references on ammonia formtion and decomposition by means of alpha rays, canal rays, cathode rays and ultraviolet light). 29. Goldstein, Richard Frank, The Petroleum Chemicals Industry New York, WIl ey, 1950. 50. Harker, George, "Influence of Sensitisers on Chemical Reactions Produced by Gamma Radiation," Nature _33 378 (19314). 55

Bibliography (Cont' d) 31. Hart, E. J., "Mechanism of the Gamma-ray Induced Oxidation of Formic Acid in Aqueous Solution," J. Am. Chem. Soc. 73, 68-73. 32. Heisig, G. B., "The Action of Radon on Some Unsaturated Hydrocarbons," J. Am. Chem. Soc. 5 3245 (1931). 33. Heisig, G. B., "Action of Radon on Some Unsaturated Hydrocarbons, II. Propylene and Cyclopropane," J. Am. Chem. Soc. 5 2328-32 (1932). 34. Heisig, G. B., "Heats of Formation and -M/N ratios," J. Phys. Chem. 36, 1000-5 (1932). 35. Hirschfelder, J. 0., "Chemical Reactions Produced by Ionizing Processes,' J. Phys. and Colloid. Chem. 52, 447-50 (1948). 36. Honig, R. E., and Sheppard, C. W., "An Experimental Comparison of the Chemical Effects of Deuterons and of Alpha Particles on Methane and n-Butane," J. Phys. Chem. 50 119-143 (1946). 37. Hopwood, F. L., and Phillips, J. T., "Polymerization of Liquids by Irradiation with Neutrons and Other Rays," Nature 143, 640 (1939). 38. Keller, M., and Weiss, J.., "Chemical Actions of Tonising Radiations in Solution. Part VI. Radiation Chemistry of Sterols. The Action of X-rays on Cholesterol and 3-B-Hydroxypregn-5-en-20-one," J. Chem. Soc. 1950, 2709. 39. Keller, M., and Weiss, J., "Chemical Actions of Ionising Radiations in Solution. Part VII. Radiation Chemistry of Sterols. The Action of X-rays on Cholic Acid in Aqueous Solution," J. Chem. Soc. 1951, 25-6. 40. Keller, M., and Weiss, J., "Chemical Actions of Ionising Radiations in Solution. Part IX. Radiation Chemistry of Sterols. The Action of X-rays on (+) QEstrone-b in Aqueous Solution," J. Chem. Soc. 1951, 1247-9. 41. Lid,. C., The Chemical Effects of Alpha Particles and Electrons, 2nd ed., ACS Mnaograph Series No. 2, New York, The Chemical Catalogue CcKpn~y, 1928 (508 references). 42. Lind, S. C., "Chemical Action Produced by Radium Emanation, I. The Combination of Hydrogen and Oxygen," J. Am. Chem. Soc. 41, 531-51 (1919). 43. Lind, S. C., "Chemical Action Produced by Radium Emanation, II. The Chemical Effect of Recoil Atoms,t' J. Am. Chem. Soc. 41, 551-59 (1919). 44. Lind, S. C., "'adiochemical Equilibrium in Ammonia Synthesis,' J. Am. Chem. Soc. _L 2423-4 (1931). 45. Lind, S. C., and Bardwell, D. C., "The Chemical Effects in Ionized Organic Gases,"t' Science 62, 422-24 (1925).

Bibliography (cont 'd) 46. Lind, S. C., and Bardwell, D. C., "Chemical Action of Gaseous Ions Produced by Alpha Particles, VI. Reactions of the Oxides of Carbon," J. Am. Chem. Soc. 47, 2675 (1925). 47. Lind, S. C., and Bardwell, D. C., "The Chemical Action of Gaseous Ions Produced by Alpha Particles, VIII. The Catalytic Influence of Ions of Inert Gases," J. Am. Chem. Soc. _ 1575-84 (1926). 48. Lind. S. C., and Bardwell, D. C., "Chemical Action of Gaseous Ions Produced by Alpha Particles, IX. Saturated Hydrocarbons, " J. Am. Chem. Soc. 48, 2335 (3926). 49. Lind, S. C., and Bardwell, D. C., "The Synthesis of Ammonia by Alpha Rays," J. Am. Chem. Soc. 50, 745-8 (1928). 50. Lind, S. C., Bardwell, D. C., and Perry, J. H., "The Chemical Action of Gaseous Ions Produced by Alpha Particles, VII. Unsaturated Carbon Compounds," J. Am. Chem. Soc. 48, 1556-75 (1926). 51. Lind, S. C., and Glockler, George, "The Chemical Effect of Electrical Discharge in Ethane," Trans. Am. Elect. Soc. 52, 37 (1927). 52. Lind, S. C., and Glockler, George, "Control of the Molecular Weight of Liquid Hydrocarbons Produced by Electrical Discharge in Ethane," J. Am. Chem. Soc. 50, 1767-72 (1928). 53. Lind, S. C., and Glockler, George, "III, The Chemical Effects of SemiCorona Discharge in Gaseous Hydrocarbons," J. Am. Chem. Soc. 5X 2811-21 (1929). 54. Lind, S. C., and Glockler, George, "IV. The Chemical Effects of Electrical Discharge in Butane. Fractionation of the Liquid Product," J. Am. Chem. Soc. 5 3655-60 (1929). 55. Lind, S. C., and Glockler, George, "V. The Condensation of Hydrocarbons by Electrical Discharge. Comparison with Condensation by Alpha Rays," J. Am. Cem. SoCem. Soc 52,, 4450-61 (1930). 56. Lind, S. C., and Schiflett, C. H., "Studies of the Oxidation of Alpha Ray Cuprene," J. Am. Chem. Soc. 5_L 411-13 (1937). 57. Lind, S. C., and Vanpee, M., "Tihe Effect of Xenon Ions in Chemical Action by Alpha Particles," J. Phys. and Colloid. Chem. L 898 (1949). 58. Laebl, H., Stein, G., and Weiss, J., "Chemical Actions of Ionising Radiations on Aqueous Solutions. Part V. Hydroxylation of Nitrobenzene by Free Radicals Produced by X-rays, " J. Chem. Soc. 1950, 2704. 57

Bibliography (cont' d) 59. Laebl, H., Stein, G., and Weiss, J., "Chemical Actions of Ionising Radiation on Aqueous Solutions. Part VIII. Eydroxylation of Benzoic Acid by Free Radicals Produced by X-rays, " J. Chem. Soc. 1951, 405-7. 60. Loiseleur, J. Latarjet, R., and Crovisier, C., "Formation of Peroxides in Aqueous Solutions of Organic Compounds Irradiated With X-rays," Corpt. Rend. Soc. Biol. 136 57-60 (1942). 61. Magee, J. L., "Theory of Radiation Chemistry. I. Some Effects of Variation in Ionization Density," J. Am. Chem. Soc. 73, 3270-5 (1951). 62. Magee, J. L., and Burton, M., "Elementary Processes in Radiation Chemistry I. Saam Considerations of Mechanism of Electron Capture," J. Am. Chem. Soc_. 1965-74 (1950). - 65. Magee, J. L., and Burton, M., "Elementary Processes in Radiation Chemistry II. Negative Ion Formation by Electron Capture in Neutral Molecules," J. Am. Chem. Soc. 7y 523-32 (1951). 64. Manion, J. P., and Burton, M., "Radiolysis of Hydrocarbon Mixtures," Abstracts of Papers, 11th Meeting Am. Chem. Soc. 41p, 1951. 65. Mmud, W., "Ianization by Radon in Spherical Vessels," J. Phys. Chem. 30_ 890-894 (1926). 66. Mund, W., and Koch, W., "The Chemical Action of Alpha-Particles on Acetylene," J. ys. Chem.30, 289-93 (1926). 67. Newton, A., "Radiation Chemistry of Alcohols," from Wakerling, R. K.. Summary of the Research Progress Meeting, Dec. Un 1950, Uiversity of California, Radiation Laboratory, UCRL-1175. 68. Otvos, John W., "Ionization by C14 Radiation in the Ionization Chamber," Phys. Rev_. 73 537 (1948). 69. Penneman, R. A.,, Effects of Radiation on Aqueous Solutions of Carboxylic Acids. U. S. Atomic Energy Comm;ission, AECD-21367Y19487 70. Ponsaert, Bull. Soc. Chim. Belg. 38, 110 (1929). 71. Rexer, Ernst, "Accelerated Polymerization by Radiation with Gamma and Rontgen Quanta," Reichsber. Physik (Beihefte, Physik. Z.) 1, 111-19 (1944). 72. Rosenblum, Charles, "Benzene Formation in the Radiochemical Polymerization of Acetylene,' J. s. and Colloid. Chem. 52, 474-8 (1948). 73. Schoepfle, C. S., and Fellows, C. H., "Gaseous Products from Action of Cathode Rays on Hydrocarbons," nd. Eng. Chem. L 1396 (1931). 58

Bibliography (cont'd) 74. Stein, Gabriel, and Weiss, J., "Chemical Effects of Ionizing Radiations, Nature 161, 650 (1948). 75.. 2:hep-pird, Charles W., and Burton; Virginia L., "The Effects of Radioactivity on Fatty Acids," J. Am. Chem. Soc. 68, 1636-39 (1946). 76. Sheppard, C. W., and Honig, R. E., "A Theoretical Analysis of the Relative Chemical Effects of Alpha Particles and Deuterons," J. 0Phys. Chem. 50, 144-51 (1946). 77. Snyder, W. S., and Powell, J. L., Absorption of Gamma Rays, U. S. Atomic Energy Commission, AECD-2739, 1949. 78. Steacie, E. W. R., "The Relation of Radiation Chemistry to Photochemistry," J. Phys. and Colloid. Chem. 52, 441-6 (1948). 79. Stein, G., and Weiss, J., "Chemical Actions of Ionising Radiations on Aqueous Solutions. Part II. The Formation of Free Radicals. The Action of X-rays on Benzene and Benzoic Acid," J. Chem. Soc. 1949, 3245-54. 80. Stein, G., and Weiss, J., "Chemical Actions of Ionising Radiations on Aqueous Solutions. Part III. The Actions of Neutrons and of Alpha Particles on Benzene, J. Chem. Soc. 1949, 3254-6. 81. Stein, G., and Weiss, J., "Chemical Actions of Ionising Radiations on Aqueous Solutions. Part IV. The Action of X-rays on Some Amino-acids," J. Chem. Soc. 1949, 3256-63. 82. "Symposium on Radiation Chemistry and Photochemistry." J. Phys. and Colloid. Chem. 52, 437-611 (1948) (339 references). 83. Toulis, William J., "The Decomposition of Water by Radiation," U.S. Atomic Energy Commission, UCRL-583 (1950). 84. Treiman, L. H., "Effect of Electron Beams on Aqueous Dichromate Solutions," U. S. Atomic Energy Commission, MDDC-1732, 1948. 85. Viallard, R., and Maget, M., "The Fragmentation of Linear C Chains by Electron Impact," Compt. rend. 228, 1118-20 (1949). 86. Watson, John H. L., Vanpee, Marcel, Lind, S. C., "The Solids Condensed from Carbon Monoxide by Alpha Particles," J. Phys. and Colloid. Chem. 54, 391-400 (1950). 87. Weiss, J., "Radiochemistry of Aqueous Solutions," Nature 153, ~]h-l50 (19?4). 88. Weiss, J., "Some Aspects of the Chemical and Biological Action of Radiations," Trans. Faraday Soc. 43, 314-24 (1947). 59

Bibliography (cont ' d) 89. Wildschut, A. J., "Stralingschemie en Polymerisatie," Chemisch Weekblad,5, 2-8 (1949). 90. Williams, Nelson T., and Essex, Harry, 'Effect of Electric Fields on the Decomposition of Nitrous Oxide by Alpha-rays, " J. Chem. Phys. 16, 12, 1153 (1948). 91. Zimmer, K. G., Naturwissenchaften 32, 375-6 (1944). 60

Progress Report No. 2 UTILIZATION OF THE GROSS FISSION PRODUCTS The Effect of Gamma Radiation on Chemical Reactions J. J. Martin L. C. Anderson D. E. Harmer J. G. Lewis

TABLE OF CONTENTS Progress Report No. 2 Page THE EFFECT OF RADIATION ON CHEMICAL REACTIONS (SUBPROJECT M943-C) 30 A. Introduction 30 B. Experimental Results 31 C. Design and Construction of New Equipment 35 D. Future Program 42 LIST OF ILLUSTRATIONS Figure Page 26. Pyrex Glass Tubes for Irradiation of Gases at Low Pressure 33 27. Low-Pressure System for Gases in the Presence of Radioactive Catalysts 34 28. Batch Reaction System Utilizing High-Pressure Reactor 34 29. Stainless-Steel Reactor and Auxiliary Fittings for Use with Gases at High Pressure 36 30. Working Drawing for the Construction of the Reactor 37 31. Flow Diagram for a Continuous System Employing the High-Pressure Stainless-Steel Reactor 38 32. Catalyst Holder for Reactor 40 33. Washing Packing for High-Pressure Reactor 41 Table IV. Theoretical Percentages of Ammonia in a System Initially Consisting of a Stoichiometric Mixture of Nitrogen and Hydrogen 35

PART I TEE EFFECT OF RADIATION ON CHEMICAL REACTIONS (SUBPROJECT M943-C) Personnel Subproject Supervisors: J. J. Martin, Associate Professor of Chemical Engineering; L. C. Anderson, Professor of Chemistry and Chairman of Department. D. E. Harmer, Research Assistant; J. G. Lewis, Research Assistant. A. Introduction The basic objective of this work was stated in Progress Report I; namely, to study the promotion of chemical reactions by the use of radioactive fission products. In the course of the study answers to two questions are being sought: When does radiation accelerate the rate of approach to thermodynamic equilibrium in a chemical reaction or cause a displacement of the equilibriumpoint, and how does the extent of the reaction vary with the quantity of radiation? When the answers to these questions have been found, it is believed it will be merely an application or developmental procedure to determine whether radiation from fission products can be effectively and economically utilized in the promotion of reactions of industrial importance. Among the reactions which have been shown by previous investigators to be affected by radioactivity are the synthesis of ammonia, polymerization of some liquids, and dehydrogerlation of some hydrocarbon gases. These reactions have, for the most part, been carried out under the influence of alpha and beta radiation. Comparatively little has been published about the effect of gamma radiation. Since fission products yield considerable amounts of gamma radiation and since the project has had available for the past several months a 1000-curie, cobalt-60 gamma source prepared at Brookhaven National Laboratory, a number of experiments have been conducted using this source of radiation. Also as this report is being written, a small piece of palladium foil containing 3 curies of Pd-109 has been received after irradiation at Chalk River, and some reactions have been studied under the influence of the 1 Mev beta radiation from this source. 3o

B. Experimental Results The first reactions to be studied under the influence of gamma radiation were conducted in small 9 mm by 20 cm pyrex tubes. The procedure was simply to seal the chemical reactants in the tubes and place them in the cobalt-60 vault for approximately 24 hours. At the end of this period the tubes were removed for observation and tested for any changes that might have taken place. In the case of a number of hydrocarbon gases, including isobutane, butene-2, ethylene, butane, isobutylene and acetylene, the analysis involved breaking the sealed tubes in a closed system and observing the pressure change which resulted. It was found that none of these hydrocarbon gases originally sealed at atmospheric pressure and room temperature underwent any appreciable change in pressure, and this fact has been interpreted to mean that no reaction occurred. In the case of a tube of liquid butane which was examined after 24 hours of gamma radiation, an appreciable amount of unsaturated gas was found by absorption in sulfuric acid and the presence of hydrogen was confirmed by combustion with copper oxide. This result indicated that significant decomposition had occurred. Since gaseous butane gave no reaction, it appears that the ability of the material to absorb and utilize radiation depends upon the density of the material, as would be expected. Stoichiometric mixtures of nitrogen and hydrogen charged to the small tubes showed no measurable change after being exposed to the cobalt-60 source for 24 hours. However, in this case, as well as in the case of the hydrocarbon gases, the amount of reactant charged was very small and detection of any reaction was not considered very accurate. Therefore, large pyrex tubes of approximately 100 ml capacity were fitted with stopcocks and charged with nitrogen and hydrogen for further studies. Some of these samples were subjected to 24 hours of gamma radiation from the cobalt-60, and some of them were placed 6 inches from the target of a 200 KVP X-ray machine operating at 20 ma and 155 KVP for 3 hours. After radiation the gases in these large tubes were drawn through Nessler's test reagent for ammonia, which is capable of detecting a small fraction of one per cent of ammonia in the sample. In no case was any appreciable amount of ammonia formed. These experiments included tests in which nitrogen and hydrogen were dried before charging and also tests in which these gases were saturated with water vapor before charging. In the next set of experiments sample tubes were constructed which had glass stopcocks at one end and ground-glass-joint caps at the other end. The caps permitted charging solid catalyst, or any other solid, into the tubes (see Fig. 26, page 33). A doubly promoted ammonia catalyst was obtained from the Pennsylvania Salt Mfg. Co. and prepared by reduction in hydrogen at 450~ to 31

5000C. This reduced catalyst was transferred under nitrogen to the reaction tubes which were then evacuated and filled with a stoichiometric mixture of nitrogen and hydrogen. Using both dry and wet reactant gases, no appreciable amount of ammonia was detected by Nessler's test after these samples had the usual 24 hours in the cobalt-60 vault. It was noted in these experiments that the catalyst tended to adsorb gases very readily since it was quite porous. It is conceivable that very small amounts of ammonia might have remained on the catalyst, though the consistently negative results seem to indicate absence of ammonia. Because of the necessity of using stopcock grease on the ground-glass joints, the reaction tube with the catalyst could not be heated to very high temperatures after radiation to assure the liberation of minute amounts of ammonia. In an attempt to determine whether gamma radiation might be favoring the decomposition of ammonia rather than its synthesis, a sample tube was charged with pure ammonia. Analysis was made on an Orsat-type gas analyzer using 10 per cent sulfuric acid as the absorber for ammonia gas. After 24 hours of gamma radiation no appreciable amount of unabsorbed gas remained, indicating no decomposition of ammonia. Experiments are now in progress to determine whether the presence of a catalyst might have any effect on the decomposition of ammonia when irradiated with gamma rays at room conditions. It might be pointed out that the results of thermodynamic calculations for the system nitrogen, hydrogen, and ammonia at room temperature and atmospheric pressure, as given later in this report, show that equilibrium in the absence of radiation greatly favors ammonia. Therefore, if any appreciable decomposition is to be found, the equilibrium point must change. Furthermore, it seems probable that if decomposition does take place, it will vary directly with the quantity of radiation supplied, because the radiation must satisfy the positive freeenergy-change requirement of the reaction. On the other hand, the extent of the synthesis reaction to make ammonia might well be independent of the quantity of radiation once the reaction is started, for the negative free-energy change forms an ideal setting for the initiation of a chain reaction. In another experiment on the radiation of a liquid with gamma rays, it was found that a sample of acrylonitrile (practical grade) in a sealed tube formed many small nuclei of solid polymer during the first part of the irradiation; and at the end of 24 hours of irradiation, a hard white solid was obtained. Monomeric styrene showed evidence of polymerization by a gradual increase in viscosity during irradiation. The viscosity continued to increase after the irradiated styrene was removed for the cobalt-60 vault while unirradiated check samples remained unchanged. Results obtained up to the time of this writing have not been consistent enough to warrant any general conclusions, although the viscosity increase after radiation seems to indicate that the long induction period ususlly found in thermal polymerization of plastics may be shortened in the presence of gamma radiation. Further experiments have been planned.

Since it is knowen from the literature that beta particle energy is more readily absorbed than high energy photon radiation, an investigation wtas started to find out whether the electrons produced by the Comton effect might be used as a means of imparting energy to chemical reactants. An increase in count was observed with Geiger counters when thin metal sheets were placed in front of the counter windows and gamma radiation passed parallel to the surface of the window. As an application of these observations, nitrogen and hydrogen were charged to reaction tubes containing copper gauze or brass sheets or aluminum dust. Although no ammonia was detected after the usual 24-hour gamma radiation from cobalt-60, it is felt that the possibility of using these secondary emissions should be investigated further, especially at higher pressures where absorption of' the emitted electrons would be nearly complete. It is also planned to investigate other types of secondary radiation, particularly that which emanates from materials which fluoresce or phosphoresce when activated by beta or gamma rays. More recent experiments involving beta radiation directly have not given any positive results. In one experiment nitrogen and hydrogen were charged to a two-liter flask (Fig. 27) along with a piece of palladium foil containing Fig. 26. Pyrex Glass Tubes for Irradiation of Gases at Low Pressure. Tube on the left is charged with a solid ammonia catalyst. 33

........ Fig. 27. Low-Pressure System for Gases in the Presence of Radioactive Catalysts. Flask behind lucite shield. contains palladium foil with a small percent of Pd-109. some palladium-109, whose total activity was about 5 curies. At the end of about 12 hours no ammonia could be detected by Nessler's test. In another experiment a stainless steel reactor tube (Fig. 28) was charged with the palladium foil Fig. 28. Batch Reaction System Utilizing High-Pressure Reactor. whose activity was now approximately 1 curie, and a stoichiometric mixture of nitrogen and hydrogen was added at a pressure of about 55 atmospheres. At the end of about 11 hours the Nessler test indicated that no appreciable amount of ammonia had been produced at room temperature by the beta radiation. This is the first experiment to be conducted at a pressure higher than atmospheric, and considerable confirmation and extension of this work remains to be done. 23,

C. Design and Construction of New Equipment The preliminary experiments to date indicate that high density of the chemical reactants is most favorable to the promotion of a reaction by radiation. This means that, in the case of gases, high pressures appear to give the greatest promise, both because of greater absorption of radiation and because of the increased conversion at equilibrium in many systems. As an example of a reaction in which elevated pressure increases the normal equilibrium percentage of desired product, the ammonia synthesis is considered in detail. The equilibrium molar percentages of ammonia present in a reaction mass originally consisting of nitrogen and hydrogen in the stoichiometric ratio are given in Table IV. TABLE IV THEORETICAL PERCENTAGES OF AMMONIA IN A SYSTEM INITIALLY CONSISTING OF A STOICHIOIMETRIC MIXTURE OF NITROGEN AND HYDROGEN (Results of Thermodynamic Calculations) Temperature Pressure Percentage of (~F) (psia) Ammonia at Equilibrium 1 70 147 ~ 100 -2 70 1500 100 -3 450 14.7 35.9 4 450 1500 72.1 5 932 4400 26.4 (Data from Curtis,l Dodge, and Hougen and Watson3. Lines 1 through 4 represent conditions of temperature and pressure attainable in the apparatus which will soon be available in this laboratory. Line 5 represents conditions approximating those in industrial ammonia reactors. It is evident from Line 2 that it is desirable to operate an ammonia reactor at as high a pressure and low a temperature as possible in order to obtain the maximum conversion to desired product at equilibrium. However, if the usual catalysts are employed for the reaction, the rate of reaction decreases with 1 Curtis, H. A., Fixed Nitrogen. New York, Reinhold (1932). 2 Dodge, B. F., Chemical Engineering Thermodynamics. New York, McGraw-Hill (19144) 3 Hougen, O. A., and Watson, K. M., Chemical Process Principles. New York, Wiley (1947).

temperature decrease in such a way that at room temperature there is no conversion whatsoever. Now it seems possible that radiation may speed up the reaction and permit lower operating temperatures either with or without catalyst. Similar remarks could be made for the methanol synthesis and for some other reactions of commercial importance. By constructing a reaction system in which both temperature and pressure can be varied over considerable ranges, it is thought that much more information can be secured concerning the kinetics and equilibria of the systems to be investigated. In addition, it is possible that some reactions which would escape detection if carried out under milder conditions may be detected by operation under severe conditions. The experimental reaction system being constructed is centered around a high temperature, high pressure reaction vessel. Fig. 29 is a photograph of the completed vessel, and Fig. 30 is a working drawing for the construction of the Fig. 29. Stainless-Steel Reactor and Auxiliary Fittings for Use with Gases at High Pressure. vessel. Accessory equipment will be provided to permit the reactor to be operated either batch-wise or as a part of a contirluous-flow reaction system, as shown on the flow sheet in Fig. 31. The reactor itself was designed according to the ASME code to operate at maximum conditions of 2000 psi and 6500F and was given a hydrostatic test at 2800 psi and room temperature. The reactor is constructed of stainless steel, and fits snugly into the 1-1/2-inch diameter access opening in the cobalt-60 vault. It will be necessary to limit the temperature of the outside of the reactor to approximately 350'F in order to avoid damage to the vault. It is thought that higher temperatures might cause warping of the aluminum case enclosing the cobalt or warping of the stainless-steel inner shell of the vault, or might even cause softening of the lead shielding. Provisions must be made for cooling the chemicals coming from the reactor since the accessory supply and control tubing will be of alumina.un. The accessory equipment may be operated at 2000

TOP OFVALT TOP OF 4-ME DIA. HEAD LRANGE BODANFLANGE BOTTOM HEAD-SEE DETAIL DETAIL2CA-N193 BOLT STUDS -1/2 IN. DI. DETAIL.0 BAR 1/21NOID~~~~~ NUTRSSASBERPIE NT-ASIAEFRMSANLS EN E S N P DETAIL L ADAPER PIECE -ASEE DETAIL IN SECTION H-R E t~~~~~~~~~~~~~~~ ASSEMBLY -3 DRIll 4/6 HOLES 9/16DtA._ FULR EPENETRATION \L 1.~~~~~~~~70 N7 I I I I N - N i t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~EDN WELDOTFAE H/S~~~~~~ A. 4 ORCOEFLEESTIAL BUTT WELDS 020 ORES0 -CORNESWT20- C TOC WELDS-R2DWO ENO TO ENOTHTAF - A 22FR AGEUT 27.341 C.A = --- —- =750 WBDLLIBG ORFIN -U I E NET FINDSH CRO L FEB0QDOFPRPE PO C PE RFTER WELDING A.tE 9/16~~-C RI2/ORE N DEPRESSIONS BELOW PIPE T H51E34 FLEHL WALL PERMITTED S N151 E04CFFSETMETAL NCI —EE DL.220 DRLL4 HOLES 9/16 DIA FULL PENETRATION BEING 19 T R.H FOlo T IA LUIELD 5/16 ______ WE BOLTS/OUTSERABA TCSEA-ST /L 1/~~~~~~~~~~~~~~~~~~~~~I16 0125 lIC 1020 ALLFWELDSIALESBAUTTAREL I L mDEAL1 1050 ~ J _- IC2S IMO 1.150 \~~ t/~ OMLO TORECEIVE TABNG CAT EALT STUDSDORN TO /IE1N. LENGTH 3/16-I ADAPTER RIR SCARF FOR WELDING 3m I -1.s~~~C AS SWHO ITWN BREAK SHARP -I —4 OIN —EN — C 0.355E~~~~~~~~~~~~~~~~~~~~~~~~~~ CORNERS WITH FPLE CORERFINISH SMOOTH AFTER M RG LA CR1TI/O SAC HANALING AFAE" 0250 ODRILL FOR FLANGE NUTS NGSEAT FILE ON I/RO EBA 28 ORIL(O.1405O)- IGR.H 1/9 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~FULL PENETRATION NTG.H IL - REQUI E A O S I I 2.88 SC. - — 4 IIH RW OL23, 5 _ GASKET 400 O.QAFE WELDING e~ooR THNEAR......E.BE YAP E E -9 SPOT FACEI STION ACA 2 HOLED ANE N O.O81'Ob4?047 ~ ~ ~ ~ ~ ~ ~ ETI CEETTO FITN-ND T B4 O.D — C LEON B Fig.~ 30. Working Drawing forS thE Construction of the.BeaIC TOP HEAD -1.750 FILLET WELDS ATSIDE SECTION P-F ~ ~ ~ ~ ~ ~ ~~~I 0250 DRILL- 2 nOLES -+- DRILL 4 HOLES 9/16 DIP, r~~~~~~~~~~~~~~~~ESIN ONDTIOS 200 SI NC-F R.O. SECTIONFSF0.5 DET0IL 2 CLASS 2 FIT N-Q5 RL-2HLSDIL4HLS91 I./6 GRINID THREADS - L FULL THROS MING 7-a0 30" 50 0125 HOLLOW OUT THIS HOLER TEST L500 1 125 1,410 1,750 ONLY TO RECEIVE TUBING E ADAPTER PIECE ~NC-1-yq 5 THDS/IN.-R.H. j-vZRUT DET CAIL OF FITTIGRING - ~N51DE TOPEADS DI ) FABRICATE NUTSTAFRDTHED Flg. 30. AT~~~~~~~~~~~r~~~t~:Dr~~~~~~~t~ for th~~~1/8 eMIFN CRoMsr to o~ thAe:ec',o BREAK SHARP CORNERS 30 18 30, 1. 50 ITH FILE F1( GASKET GROOVEII REFERENCE HOLLOW OUT OE HOLE ERI-'TUSE SOURCE HANDLING SA.FE' 0.125 CLEAR ~~~~~~~~~~~~~~~~ONLY AS SHOWN - 2.50 ~~~~~~~~~KEEP FITf- NUA C)E C.TIN - AS POSSIBLE TO THIS LINE 4.00 O.D.-SCIO - ADAPTER WITHOUT G0143 OUTSIDE DESIGN CONDITIONS 200 PSIG ----- ~~~~~~~~~~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~OF THE L NE BODY FLANGL 6501F FABRICATE FROM 1.750 O.D. BAR STOCK SILVERSLER OPERATING CONDITIONS 1500 PSIG 3/8 O.D.q ALUMINUM TUBING 70-200'F:1:-::F. DO NO SUPPLY ON THE JOB ~~~~~~~~~~~TEST CONDITIONS 4000 PSIG - HYDROTS CUT PIPE THREADS OFF AR qnWN 70-FPI -HME SECTION H-n ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ALL WELDS WITH 1-1 0. 1-3/4 LB DETAIL OF FITTING - INSIDE TOP HEAD HM~~FLLLNT ALWLE Fig. 30. Working Drawing for the Construction of the Reactor

POWER PRESSURE GAGES T.C. WIRES POTENTIOMETER POWER WIRES G H L DERR VENT AC PRESSURE EGAGS OIIESS TO SAMPLING TRAIN SATURATOR SUPPLY GA5 CYLINDER5 GAMMA RAYS | VACUUM PUMP GAS HOLDER 'VVVL/L'Lt I r I II I EXHAUST REACTOR DESIGN CONDITIONS TRAP FLOW SHEET PILOT PLANT FOR REACTOR 2000PSI 6500F. RADIATION CHEMISTRY STUDIES Fig. 31. Flow Diagram for a Continuous System Employing the High-Pressure Stainless-Steel Reactor

psi at room temperature, with the exception of the sampling, analytical and lowpressure metering apparatus. These three types of equipment will be operated at atmospheric pressure. Fig. 32 shows the internal fittings for the reactor, which are patterned after some types of commercial reactors. The apparatus is designed in such a way that it may readily be removed from the reaction vessel shown in Fig. 30. Internal heat exchangers are provided in the form of baffles, so that incoming gases may cool the walls of the pressure vessel. A thermal radiation shield may be added if necessary in order to maintain the desired temperature in the interior and in order to prevent thermal radiation from the electric heater element from overheating the wall of the reactor. The heating element will be wrapped around the constriction just above the catalyst holder. Spiral countercurrent heat exchangers are provided to exchange the heat between the incoming and outgoing streams. It may be discovered that the systems being investigated are so transparent to gamma radiation that little use is made of the energy of the radiation even at higher pressures. Since electrons in passing through a given thickness of material will lose a much higher percentage of their energy than will gamma rays, it may prove desirable to cause the gamma rays to eject electrons from a suitable material placed in the reacting system. If the rate of chemical reaction depends upon the rate of ionization, the rate of reaction should be increased by this secondary emission. Fig. 33 is a drawing of a special packing which may be readily installed in or removed from the reactor vessel. This special packing consists of a series of washers placed parallel to the maximum component of the gamma radiation from the surrounding cylindrical source of cobalt-60. This packing should cause electrons produced in the washers by Compton-scattering of gamma rays to be dispersed in the reacting medium. The arrangement of the packing in washers was governed by the thought that a perferred orientation of the scattered electrons may exist and that this orientation may be different from the direction of the incident gamma radiation. Although the reaction system described above was designed to fit into the arrangement of the existing cobalt-60 vault for studies using gamma radiation, it will be equally possible to use the entire system for studies in which a source of radioactivity is placed inside the reaction vessel itself. The latter arrangement would be particularly desirable when beta sources are used. The reaction vessel has now been completed and batch tests at room temperature will probably be conducted in this apparatus while construction on its internal fittings and the necessary equipment needed to study flow reactions is proceeding. The greatest share of the materials needed to complete the construction of the experimental unit is now on hand, while most of the remainder of the materials should be secured within three months. 39

STAINLESS SPIRAL 28 GA.~ 21-13/16 A - 1-9/16 5- 1/2- 2. 12 3/4\ l~ | ~ ~ g1/4" 1/4 1I 0'"O.D x20 A. 13/Y16O. D 20 GA. V/8ROD x 20 GA 14 A. WIRE CROMEL A 3/A + _ t,1.5 e:6~~~~~~~~~~~~~~~~~~~~~~~~~~SILVER SOLDER IN PLACE (SUPPORT _/ " -20ODA OLT FITO S L~UERSLE IETNSIDE LL2 1/4 ''L l ' XlX C 0.40 DOO SECTION D-D SUPPORTS 6 REQ'D II GA. WIRE- TURN DOWN / \,.,32".230,-O.D AT COIL TO FIT INSIDE /o ~%"l-I..~ ---~.. 5/B"00,0.56 I.D 20 GA.TUBE SILVER 0.09 S.269 FITLVER SOLDER END TURNS I OF*WIRE --------— "r..k....II ~J.IJ~......GAS PORTS 0.065 — TO CLEAR 14 A. SCALE- DOURBLE lx0AUTTY7PISCAL SLOT FOR 14. 0-j~j~i oss 14 GAI WIRE-TURN DOWN TO....... FIT ISIDE 13/16 O.D...74 DIMENSIONS OF SLOT ARE THE 0 L......75. SAME ON THE 13/16" a 5/8" TUNE 0.38' 5/8" x 20 GA. 13/16", 20 GA, 0.93t FIT IN. I x 20. GA. DIAMETERS OF PLUG- DOUBLE SCALE Fig. 2. Catalyst Holder for Reactor

STAGGER -1/8 INCH CUT-OUTS INSERT SPACE FOR HEATER 3/8 CARBON STEEL TUBING 5/I6_ 12-3/4 7-1/16 22-1/8 ASSEMBLY / — STD. NUT TO FIT INSIDE AM. STD.SILVER SOLDER / /8 192 REQUIRED 192 REQUIRED NOTE: FLATTEN WASHERS BEFORE ASSEMBLY ELECTRON SCATTERING DEVICE ONE REQ'D. AS SHOWN DWG. No. M943C-S3 Fig. 33. Washer Packing for High-Pressure Reactor

D. Future Program It is planned to continue the studies on the ammonia synthesis reaction, varying the temperature and pressure over wide ranges and using the new stainlesssteel reactor. Since many commercial reactors operate at 4400 psia and 5000C, it is quite reasonable that no positive results were found at room temperature and atmospheric pressure. It is expected that studies at higher temperatures and pressures will prove more enlightening and will indicate the range of usefulness of beta and gamma radiation in promoting the ammonia synthesis, as well as other reactions, such as the methanol synthesis and Fischer-Tropsch reactions. While the higher-pressure gas reactions are being studied, further work will be carried on with liquids at lower pressures. It is planned to study some systems in which chain reactions have been postulated or might be possible. The polymerization of liquids such as styrene and acrylonitrile are worthy of further study. Also some reactions between different liquids and solids, such as the Diels-Alder reaction will be investigated in this program. Gamma radiation will continue to be tried for all reactions studied, and an increasing amount of beta radiation will be utilized. It is planned to try some P-32 as the beta source, as well as Pd-109. Both of these will be charged inside the reacting system, as their penetrating power is quite low. Secondary particle and photon radiation will also be studied thoroughly. 42

Progress Report No. 3 UTILIZATION OF THE GROSS FISSION PRODUCTS The Effect of Gamma Radiation on Chemical Reactions J. J. Martin J. G. Lewis L. C. Anderson J. R. Hallman D. E. Harmer S. A. Stolton

TAIBLE OF CONTENTS Progress Report No. 3 Page SUBPROJECT M943-C, THE EFFECT OF RADIATION ON CHEMICAL REACTIONS 31 A. Introduction 31 1. Styrene 31 2. Exploratory Investigation of Reactions Influenced by Radiation 34 B. Future Work 38 1. Experimental Studies 38 2. Need for Modified Gamma Source 39 Bibliography 39 LIST OF ILLUSTRATIONS Figure Page 17. Glass Apparatus for Vacuum Distillation of Styrene, and Glass-stoppered Tube for Irradiation and Polymerization of Styrene 33 18. Graph of Percent of Styrene Polymerized vs. Hours of Polymerization at 1250C for Control Samples and Samples Given 24 Hours of Radiation 35 I. Irradiation of Natural Oils with Palladium-109 Beta Rays 36 II. Changes in Viscosity of Soya Oil after Irradiation and Subsequent Heating 37

PART II. SUBPROJECT M943-C, TE0 EFFECT OF RADIATION ON CHEMICAL REACTIONS Personnel: Subproject Supervisors: J. J. Martin, Associate Professor of Chemical Engineering; L. C. Anderson, Chairman of the Department of Chemistry. Research Assistants: D. E. Harmer, J. G. Lewis, J. R. Hallman, S. A. Stolton. A. INTRODUCTION Tle objective of this work is indicated by the title of this subproject and has been discussed in earlier reports. Recently, the work has been concentrated on more nearly quantitative determinations of some effects of radiation, but studies are also being continued with the objective of finding which reactions are affected by radiation. The effects of gamma rays from the cobalt-60 source have been receiving the most attention. More work will probably be done on beta radiation in the future. 1. Styrene In the last progress report made for this projectl, it was reported that the viscosity of samples of monomeric styrene after irradiation in the cobalt-60 gamma source continued to increase after the samples were withdrawn from the source, while unirradiated check samples remained unchanged. This effect has been investigated more fully. Others working in this field have produced a high degree of polymerization of styrene at room temperature with radiation2. The concern of this project, however, has been to study the effects of radiation on the progress of the conventional thermal polymerization of styrene. It is thought that industrial applications for radioactive materials may be more quickly realized by these methods. In original experimental runs, relative viscosity was used as a qualitative measure of the degree of polymerization of unstabilized styrene which had been heated at 125~C. Samples of styrene were given 24 hours of 31

radiation in the kilocurie cobalt-60 gamma source. They were then subjected to thermal polymerization at 1250C. Viscosities of these samples were measured by observing the time required for stainless-steel ball bearings to fall through a given layer of the material. The viscosities of irradiated samples were found to be two to three times as great as the viscosities of corresponding unirradiated control samples. This relationship appeared to hold over the entire range of viscosities which could be measured by the method described. The diameters of the glass tubes containing the styrene and falling ball were somewhat variable, since ordinary Pyrex tubing was employed, so a large variation in results was obtained. However, the data were regarded as being good evidence that radiation had produced a noticeable effect on the polymerization rate. When 12 or 48 hours of radiation were given, the viscosity increase of the irradiated sample was correspondingly less or more than that of the 24-hour runs, as would be expected. Runs made on styrene stabilized with tertiary butylcatechol also showed an increase in reaction of the irradiated sample over the unirradiated control. The samples used for this particular series of experiments were irradiated and polymerized under their own vapor pressure. Since the experiments just described indicated that the irradiation of the monomeric styrene was giving a noticeable radiocatalysis to the subsequent thermal polymerization, data were sought to determine more quantitatively the effect of the prepolymerization irradiation period on the percantage conversion at various times of thermal polymerization. A standard method for the purification of the styrene monomer has been developed. All glassware which would be used to handle the dried monomer was first heated to remove moisture and volatile impurities. The monomer (E. K. Co. No. 1465 stabilized with tertiary butylcatechol) was then run through a 10 x 400-mm tube packed with fresh Ascarite to act as a drying agent and to remove as much as possible of the phenolic inhibitor by salt formation. The dried styrene was placed in an all-glass distillation apparatus (see Fig. 17), after which the inlet tube was sealed with a torch. A stopcock was provided for evacuation of the apparatus, and was placed so that the liquid styrene did not come in contact with it or with the stopcock grease. The system was evacuated and the stopcock was closed. The distillation was carried out by a process of evaporation from the surface of the styrene, rather than by ebullition with the accompanying danger of spray being carried over through the Vigreaux column into the receiver. It was found that such a distillation could be carried out by immersing the boiling flask in hot water, while cooling the receiver in a dry-ice, chloroform, carbon tetrachloride bath. Under these conditions the vapor came over at a temperature of 600-750C, and was frozen immediately in the receiver. When 32

Fig. 17. Glass apparatus for vacuum distillation of styrene, and glass-stoppered tube for irradiation and polymerization of styrene. about three-fourths of the material had been distilled, air was admitted to the system, and the frozen distillate was brought to room temperature. The sealed ends of the glass apparatus were broken open in order to pour the purified monomer, under air, into cleaned and dried glass-stoppered sample tubes, (See Fig. 17). The design of the distillation apparatus allowed the monomer to be poured out through the exit tube without coming into contact with the vacuum stopcock, while the undistilled material was retained by the shoulder of the boiling flask. When the tubes of samples could not be used immediately, they were stored either in ice or in a dry-ice bath to prevent appreciable polymerization. In runs made up to the time of this report, the irradiated samples were given 24 hours in the cobalt-60 gamma source, corresponding to about 1,920,000 rep in air. During the irradiation, the samples were stoppered in an air atmosphere, and were at room temperature. Polymerization of control and irradiated samples was carried out in an oil bath maintained at 125~ + 2~C. 5

At the end of the polymerization period, the sample tubes were cooled to room temperature in a water bath, wiped clean, and weighed. The polymer solution was poured into absolute ethyl alcohol, and the tube, with the small amount of solution that remained in the tube, was weighed again. The known weight of polymer mixture was determined by the difference in weight of the original and emptied tube. This known weight of solution was allowed to remain in the alcohol for a few minutes. The resulting precipitate was then placed in a weighed Gooch crucible and extracted in a Soxhlet extractor with absolute alcohol for about 24 hours. The crucible and polymer were then placed in a vacuum dessicator at 68-75~C for 24 hours. At the end of this procedure, the cooled crucibles were weighed again with the remaining polymer. It is felt that by this process, the significant part of the unpolymerized styrene and the absorbed alcohol was removed, so that the percentage which had polymerized could be ascertained. Some samples were polymerized to such an extent that they could no longer be poured into the alcohol. A slight modification of the above procedure was employed for these samples. The glass sample tube was broken, so that the solidified polymer mixture could be removed. The rod of polymer mixture was then crushed, weighed, and dissolved in benzene. The benzene solution could then be poured into alcohol, and the polymer precipitated as before. From data thus obtained for a series of different times, a graph of per cent conversion versus time of polymerization at 1250C for irradiated (24 hours in the cobalt-60 gamma source) and the unirradiated samples was drawn (Fig. 18). Examination of this plot shows that some amount of polymerization has taken place during the irradiation process. The amount which was polymerized during irradiation is somewhat greater than that obtained by other investigators2 This difference may be attributed to the fact that the styrene used in the data for this report was polymerized in an air atmosphere. The upper line of the graph, representing the irradiated styrene, shows that such samples are always polymerized to a greater extent than the corresponding unirradiated samples. This fact is especially important in the highest percentages polymerized, since it is here that the reaction rate is the slowest. This graph does not correspond exactly to data published by the Dow Chemical Company3, but the slightly slower rate of polymerization of the control samples may be attributed to traces of inhibiting impurities, or to a different amount of dissolved air in the samples. 2. Exploratory Investigation of Reactions Influenced by Radiation A considerable proportion of the work of this laboratory has been devoted to the search for chemical reactions which may be promoted to an

110 - I00 lA FIG. 18. GRAPH OF PERCENT OF STYRENE 70 _/__ POLYMERIZED VS. HOURS OF POLYMERIZATION AT 1250 C. w I N FOR CONTROL SAMPLES AND SAMPLES GIVEN 24 HOURS so60 OF IRRADIATION -J rl~~ I ILEGEND ~-CONTROL o o —I RRADIATION. 50 I. I rI 40 a. 30 _ 20 10 0 0 8 16 24 32 40 48 56 64 72 80 88 96 104 112 HOURS OF POLYMERIZATION industrially important extent by radiation. Several different reactions have been investigated, using macro methods of chemical analysis in attempts to detect changes produced by radiation. From the data included in this report it can be seen that some reactions have given much more promising results than have others. Below are described some experiments which were conducted in order to find easily detectable effects of radiation on chemical systems. Acetylene was irradiated with 1.9 x 106 rep in air of cobalt-60 gamma radiation. The acetylene was dissolved in acetone which had first been absorbed in a dried portland cement, asbestos mixture. The cakes of cement were placed in a stainless-steel pressure vessel designed to hold 2000 psig and fitted with an aluminum rupture disc designed to burst at 150 psig. This assembly was known as the "bomb". (The pressure vessel is that described on p. 37 of Progress Report 21 of this laboratory.) The bomb was evacuated with the cement in place, flushed with nitrogen, and evacuated again. Then 105 grams of acetone was added and acetylene introduced until an equilibrium pressure of 5 psig was reached. The bomb thus charged was irradiated for 24 hours, after which the volatile contents of the bomb were recovered by immersing the bomb in hot water and heating the discharge pipes with infrared lamps. Subsequently, the bomb was evacuated while being heated in the manner just described. During the heating period all effluent material was passed through dry-ice traps.

The condensed liquid was distilled in a Podbielniak column, where it was observed that the overhead temperature was not significantly different from that of acetone during the distillation. About 0.2 gram of a brownish, waxy solid was recovered from the pot of the column. This material was insoluble in acetone. Some natural oils, donated by the Wyandotte, Michigan, plant of the Archer-Daniels-Midland Company, of Minneapolis, Minnesota, were subjected to radiation by beta rays from palladium-109. Watch glasses were arranged on a wire frame. One drop of each oil to be observed was weighed on a separate watch glass. A thin polyethylene film was placed over the watch glasses, and the palladium-109 foil, encased in manila paper, was placed over the film of polyethylene. The oils were allowed to stand at ambient room temperature for one week in the apparatus described above. Control samples were run on the same oils. The control samples received no radiation, but were treated the same as the test samples in other respects. A summary of the results is given in Table I. TABLE I IRRADIATION OF NATURAL OILS WITH PALLADIUM-109 BETA RAYS Material Percentage Gain Remarks in Weight Test Control Test Control raw +7.8 +0.53 tough, rough film no change linseed boiled +5.7 +7.0 tough, rough film tough, rough film linseed degummed +5.8 +0.39 tough, smooth film no change soya castor +1.3 -o.58 no change no change refined +9.3 +4.5 tough, smooth film no change menhaden cotton- -0.36 -0.19 no change no change seed pitch

Some of the gains in weight observed may have been caused by oxidation in the air. It is interesting to note that on the irradiated samples of raw linseed, degummed soya, and refined menhaden, hard, tough films appeared while on the unirradiated samples of these same oils, no films appeared. A preliminary investigation was conducted on the effect of radiation on the increase in viscosity of dry, refined, degummed soya oil. Samples of the oil were placed in sealed containers and irradiated for 24 hours by gamma rays from cobalt-60. Immediately after a sample was removed from the cobalt-60 vault, it was placed in an apparatus designed to maintain the temperature and pressure of the sample at some fixed values. Nearly all runs reported were conducted at 1 mm absolute pressure. One set of runs was conducted at 575~F, and one set was conducted at 600~F. The results are tabulated in Table II, as follows: TABLE II CHANGES IN VISCOSITY OF SOYA OIL AFTER IRRADIATION AND SUBSEQUENT HEATING The starting material used in each run was dry, refined, degummed soya oil. Pressure, Run Irradiation Tempera- Time of Viscosity mm mercury, No. ture, 0C Heating, Hr G-H Poises absolute 11 24 hr, Co60 300 6 N+1/3 3.5 1 12 -- 300 6 J 2.5 1 13 24 hr, Co60 300 6 L 3.0 1 14 24 hr, Co6O 300 6 L 3.0 1 15 -- 300 NG-air leak -- --- 12 16 24, hr, Co60 300 6 L 3.0 1 17 -- 300 6.3 L 3.0 1 18 -- 300 6.2 J 2.5 1 19 -- 300 6 J 2.5 1 26 24 hr, Co60 316 6 Z4 63.4 1 27 -- 316 NG-air leak -- > --- 10 28 24 hr, Co60 316 6 Z4+1/4Z5 69.4 1 29 -- 316 NG-boiled -- ---- 1 over 30 -- 316 6 Zl+2/3Z2 33.0 3 31 -- 316 6 Z3+1/2Z4 54.8 3-4 37

The samples heated at 3000C after irradiation had viscosities of 3.0 to 3.5 poises, whereas those not irradiated but otherwise treated in the same way had viscosities of 2.5 to 3.0 poises. This difference in viscosities was approximately 20 per cent of the lower value. The samples heated at 3160C after irradiation were generally more viscous than those not irradiated but otherwise treated in the same nlanner. Some relatively mobile liquid distilled off the samples tested at 3160C. Because of the difficulties experienced in controlling the material balances at this temperature, further work under these conditions is indicated and planned. Equipment has been constructed and is now being tested for the study of the kinetics of the ammonia synthesis at approximately 9300F. These tests will be conducted both in the presence and in the absence of gamma radiation from cobalt-60. Construction is also proceeding on equipment to test gaseous reactions in the presence of cobalt-60 gamma radiation at 2000 psig and temperatures up to 9300F. It is expected that operation at conditions of temperature and pressure approaching those used in the industrial processes for the manufacture of ammonia and methanol may provide -useful additional information regarding the influence of radiation on these reactions. B. FUTURE WORK 1. Experimental Studies Future work being planned with styrene includes investigations of the variables of polymerization temperature, irradiation time, and length of time elapsed between irradiation and thermal polymerization. Apparatus has now been constructed to permit thermal polymerization and irradiation of the styrene simultaneously in the gamma source. With this new arrangement, it may be possible to ascertain whether the effects of temperature and of radiation on this reaction may be combined to give a reaction rate greater than the combined individual effects, or whether the two effects will operate independently. The use of impure samples of styrene, in which an induction period for the reaction is to be expected, may yield information about the usefulness of radiation to shorten this induction period. When sufficient data have been obtained in the case of styrene, those effects which appear to increase the rate of reaction will be investigated in other vinyl systems, and for other olefins. It is planned to conduct additional tests on the oxidation and/or polymerization of natural oils under the influence of gamma radiation. More information regarding the changes taking place could be secured by subjecting

the treated samples to some of the tests ordinarily used in testing natural oils, such as iodine number, unsaponifiable material, free fatty acids, and melting point. The work on the kinetics of gaseous reactions at high temperatures will be continued. The influence of radiation will be studied in the presence of catalysts ordinarily used for the reactions in question. 2. Need for Modified Gamma Source The available space within the shielding of the present souree of gamma radiation is so limited that the introduction of chemical equipment into the radiation field is severely handicapped. The demand for use of the present source is great, and only one small sample at a time can be accommodated. As a consequence, the work on chemical reactions is retarded, particularly by the necessity of designing and constructing highly specialized equipment needed to conduct the usual operations of chemistry within a space 1-1/2 inches in diameter by 13 inches long. If the working space within the shielding of the present source were to be increased, the flux density in the resulting volume would be reduced materially below the level now available. The productive effort of this laboratory would be measurably increased if more working space were available. However, in order to maintain the intensity of gamma radiation in the working space at some value approximating that now available, a source of greater intensity would be needed. Such an arrangement would permit more frequent scheduling of experiments, each of which would require less time to prepare. In addition, experiments needed to explore the maximum effects of radiation upon given reactions could be scheduled for the long, continuous exposures required without interfering with other experiments. Bibliography: 1. Martin, J. J., Anderson, L. C., and associates, "Utilization of the Gross Fission Products. Progress Report 2 (COO-90). University of Michigan, Engineering Research Institute. Ann Arbor, Michigan. Jan. 31, 1952. 2. Manowitz, B., Horrigan, R. V., and Bretton, R. H., "Preliminary Studies on Industrial Applications of Intense Gamma Radiation." Progress Report on Fission Products Utilization I. Brookhaven National Laboratory. Upton, N. Y., Dec. 1, 1951. BNL 141 (T-27). 3. Polymerization of Styrene. Plastics Technical Service Bulletin, The Dow Chemical Company, Midland, Michigan. 39

Progress Report No. 4 UTILIZATION OF TEE GROSS FISSION PRODUCTS The Effect of Gamma Radiation on Chemical Reactions J. J. Martin D. J. Goldsmith L. C. Anderson J. R. Hallman D. E. Harmer R. L. Kinney J. G. Lewis E. M. Rosen

TABLE OF CONTENTS Progress Report No. 4 Page SUBPROJECT M943-4 (FORMERLY M943-C), THE EFFECT OF RADIATION ON CHEMICAL REACTIONS 24 A. Introduction 24 B. Experimental Work 24 1. Halogenation 24 a. Chlorination of Benzene 26 b. Chlorination of Toluene 32 c. Bromination and Iodination Reactions 32 2. Oxidation 32 a. Using Elementary Oxygen 32 b. Using Other Oxidizing Agents 33 3. Polymerization 33 4. Other Reactions 34 C. Future Work 35 D. Bibliography 35 LIST OF ILLUSTRATIONS Figure Page 13. Flow sheet for additive chlorination of benzene in cobalt-60 gamma ray source. 25 14. Control equipment for chlorinations (Second floor, Fission Products Laboratory). 26 15. Inner part of glass reactor tube, showing inlet and exit tubes and thermocouple well. 27 16. Assembled glass reactor tube, containing liquid reactants. 27 17. Glass reactor tube in its steel jacket, connected to cooling system. Cobalt-60 vault is in the background. 28 18. Chlorine feed cylinder and tail-gas absorber bottles. 29 19. When in place in the cobalt-60 gamma source, the glass reactor can be connected to feed lines by means of tools which allow the workers to stand in a field of low radiation intensity. 29 20. Second-floor view of reactor and its glass feed lines and cooling system. 30 Table IV. Analysis of Some Samples of Benzene Hexachloride 31

PART II. SUBPROJECT M943-4 (FORMERLY M943-C), THE EFFECT OF RADIATION ON CHEMICAL REACTIONS Personnel: Subproject Supervisors: J. J. Martin, Associate Professor of Chemical Engineering, L. C. Anderson, Chairman of the Department of Chemistry. Senior Research Assistants: D. E. Harmer, J. G. Lewis. Assistants in Research: D. J. Goldsmith, J. R. Hallman, R. L. Kinney, E. M. Rosen. A. INTRODUCTION In the investigation of the effect of radiation on chemical reactions, two types of reactions are being studied: (1) chemical reactions of commercial, or possible commercial, importance that may be influenced by radiation, and (2) chemical reactions of a general kind which aid in explaining the mechanism of radiation effects. B. EXPERIMENTAL WORK In choosing chemical reactions for study of the effect of radiation, two different avenues of approach may be followed: A single reaction which has been found to be affected by radiation may be studied in detail, determining quantitatively the effect of all variables involved: or a large number of reactions may be tried in a preliminary sort of investigation to determine which reactions are affected appreciably by radiation. In general it is the latter approach which has been more closely followed so far. Of the reactions considered below, only the chlorination of benzene has received special attention. 1. Halogenation As used in this work, the term halogenation (chlorination, bromination, or iodination) refers to the reaction of an organic compound with a halogen. The products may be the result of addition or substitution processes or combination thereof. 24

LNE CHROMEL - ALUMEL MICROMAX THERMOCOUPLE LEADS TAIL GAL RUBBER STOPPER SEALED WITH DE KHOTINSKY CEMENT FEED GAS MANOMETER CONTROLLER GLASS GROUND JOINT CHLORINE RUI.BER STOPPER LINE GLASS REACTOR METHANOL RETURN DRY BOTTLE SPLASH GUARDS METHANOL FEED ER IC IC COBALT - 60 GAMMA [~V ~ RAY SOURCE \J1 REACTANT (BENZENE: r.._FOAM GUARD CARBON TETRACHLORIDTHERMOMETER THERMOMETER CARBON TETRACHLORIDECHLOROFORM -DRY ICE INSULATION LEAD SHIELD METHANOL STEEL JACKET NITROGEN F.F w e f a t cPUMP -MOTOR Fig. 13. Flow sheet for additive chlorination of benzene in cobalt-60 gamma ray source.

a. Chlorination of Benzene. It is well known that ultraviolet radiation promotes the addition of chlorine to the benzene ring, Slator12 having demonstrated this fifty years ago. Alyeal showed that alpha radiation produces a similar effect. Therefore, an experiment was undertaken to determine whether this reaction could be promoted by gamma radiation. When chlorine is bubbled through benzene, an addition reaction takes place at a very rapid rate. In view of the importance of the product, benzene hexachloride, it was decided to make a more intensive study of this reaction. After preliminary runs had been made, apparatus was constructed and assembled in which the reaction could be carried out under controlled conditions. A schematic diagram of this system is presented as Fig. 13. For convenience of operation in the Fission Products Laboratory, most of the controls for the experimental work were located on the second floor, adjacent to the opening over the kilocurie cobalt-60 source. (See laboratory plan on page 14 of Progress Report 14.) Thus the gas cylinders, manometers, tail-gas absorption train, and temperature recorder-controller were located on the second floor (Fig. 14), while the kilocurie cobalt-60 source, glass reactor with its steel jacket, and the cooling system were located on the first floor. Glass tubing was used Fig. 14. Control equipment for chlorinations (Second floor, Fission Products Laboratory). 26 - 117 7 7. '

to carry gases between the reactor and the supply cylinders and absorption train. The glass reactor consisted of a 25-mm glass tube long enough to extend completely out of the cobalt-60 source. The lower end of the glass reactor was closed, while the upper end was fitted with a standard ground joint. Four tubes, a chlorine inlet, a nitrogen inlet above the liquid level, a tail-gas exit, and a thermocouple well, were held in place in the inner ground joint by means of a rubber stopper sealed with resin cement (see Fig. 15). The assembled reactor (Fig. 16) was inserted in a steel jacket through a rubber stopper which also contained feed lines for the coolant liquid (Fig. 17). Splash guards were affixed to the steel jacket and compressed air was circulated through the vent hole at the bottom of the source container in order to prevent any possible damage to the cobalt-60 source in event of a spill or a chlorine leak. The tail gases from the reaction were absorbed in sodium hydroxide solution for analysis at the end of the experimental run (Fig. 18). Fig. 15. Inner part of glass Fig. 16. Assembled glass reactreactor tube, show- or tube, containing ing inlet and exit liquid reactants. tubes and thermocouple well. 27

Fig. 17. Glass reactor tube in its steel jacket, connected to cooling system. Cobalt-60 vault is in the background. The assembled glass reactor tube in its steel jacket was placed in the open kilocurie cobalt-60 gamma source and all glass connecting tubes were attached to the feed lines by extension tools fabricated for that purpose (Fig. 19). The design of the lead shielding of the kilocurie cobalt-60 gamma source at the University of Michigan is such that a high-intensity beam is located directly over the open source, but the radiation field at the sides of the container is low enough in intensity to allow loading operations in this area (see calibration of field around the open source, page 14 of Progress Report 14). Cooling of the reactor was accomplished by circulating methanol chilled by dry ice. When dry ice was added directly to the methanol, the centrifugal circulating pump became filled with carbon dioxide gas, causing the pump to lose its prime. Consequently, it was found necessary to chill the methanol indirectly by passing it around a can containing a mixture of carbon tetrachloride, chloroform, and dry ice. This part of the cooling system was 28

Fig. 18. Chlorine feed cylinder and tail-gas absorber bottles. Fig. 19. When in place in the cobalt-60 gamma source, the glass reactor can be connected to feed lines by means of tools which allow the workers to stand in a field of low radiation intensity. 29

insulated with sections of polystyrene foam. The thermocouple of a Micromax temperature recorder-controller was placed inside the glass reactor tube. This controller alternately turned the pump on and off. The cold methanol was pumped into the steel jacket surrounding the reactor. The temperature control was greatly improved in later runs by circulating methanol continuously through the reactor cooling jacket. The temperature of the methanol was controlled by by-passing some of the methanol through the dry-ice chamber by means of a second pump. This pump was controlled by the Micromax regulator. The assembled apparatus was located on the first floor; its appearance from the second floor may be seen in Fig. 20. Before each run, the entire system was purged with nitrogen in order to exclude all oxygen from the system. In early experiments, pure benzene was placed in the reactor. It was found, however, that the reaction then proceeded so rapidly that temperature control was virtually impossible, and the inlet lines soon became plugged with solid product. To prevent this complete solidification of the product, the reactant benzene was diluted with carbon tetrachloride. Mixtures containing as much as 30% benzene by volume produced Fig. 20. Second-floor view of reactor and its glass feed lines and cooling system. an isgls fe lns n

a semifluid slurry of solid product in the carbon tetrachloride. The reaction has been run to near completion in as little as 15 minutes when using benzene solution kept saturated with chlorine. The solid product has been analyzed and was found to be nearly pure 1,2,3,4,5,6-hexachlorocyclohexane, or "benzene hexachloride".* This compound is formed as a mixture of five stereoisomers, to which other investigators have assigned the first five letters of the Greek alphabet. The gamma isomer is known commercially as "Lindane" and possesses marked insecticidal properties which make it of considerable commercial value. The annual production of the mixed isomers in 1951 was about 125 million pounds7 Although the material may be used either as the unseparated mixture of isomers or in the purified gamma form, the insect-killing power is almost entirely dependent on the gamma isomer content. Table IV shows the gamma TABLE IV ** ANALYSIS OF SOME SAMPLES OF BENZENE HEXACHLORIDE Approximate Temperature l Gamma Isomer in of Reaction, L Benzene $ CC14 (solvent) Benzene Hexachlor~C (by volume) (by volume) ide (from infra(by volume) (by volume) red analysis) 11.3 (over 20 10 90 chlorinated) 20 20 80 12.5 -10 10 90 12.3 -10 30 70 12.8 * The authors are indebted to the Process Engineering Department of Wyandotte Chemicals Corporation and to Dr. L. E. Liggett of that company for an analytical method for total organic chlorides by the use of sodium diphenyl reagent. The authors wish to express their gratitude to E. I. du Pont de Nemours and Company, Engineering Service Division, for furnishing the analyses for gamma isomer listed here. Primary standard samples of the alpha, beta, gamma, and delta isomers were kindly supplied by the Hooker Electrochemical Company, Physical Chemical Laboratory. These samples were for the purpose of setting up analytical facilities in this laboratory. 31

isomer content of a few samples which were analyzed by infrared. It should be noted that the temperatures listed are only approximate. In each case the temperature went somewhat higher (as much as 50C) at the height of the reaction. The continuous cooling system that was installed later brought this maximum temperature down several degrees. The percentages of gamma isomer are not significantly different from those obtained in processes from the addition of chlorine to benzene under the activation of ultraviolet light. b. Chlorination of Toluene. Recently, work has been undertaken to chlorinate toluene under conditions similar to those used for benzene. Although both ultraviolet light and gamma radiation activated the addition of chlorine to the benzene nucleus, it is known that in toluene ultraviolet radiation activates substitution of chlorine for hydrogen in the methyl group. Therefore, it is of great interest to determine whether gamma radiation will activate the side-chain substitution or the nuclear-addition reaction. If it is the latter reaction which is predominantly promoted when toluene is chlorinated in the presence of gamma radiation, it is possible that gamma radiation may serve as a new tool to promote reactions that cannot be accomplished by other means. Preliminary results at the time of this writing indicate that both reactions, the addition and the substitution, are taking place under gamma irradiation; however, separation and analysis of reaction products is still in progress. The reaction between toluene and chlorine takes place at a rate comparable to that between benzene and chlorine under similar conditions. c. Bromination and Iodination Reactions. Because of the success in the promotion of chlorine addition to benzene, experiments have been tried using bromine and iodine. Evaporation of a solution of bromine in benzene which had been given about 16 hours of gamma irradiation in the kilocurie cobalt-60 source left a residue of a small quantity of crystals in a liquid having noticeable lachrymatory properties. At the time of writing, this small yield of reaction product had not been identified. A solution of iodine in benzene which was given about 25 hours in the kilocurie cobalt-60 gamma source gave no noticeable residue on evaporation. No change in color of either the bromine or iodine solution could be observed during the irradiation. A solution of linseed oil, benzene, and iodine was found to become lighter in color when given about 21 hours of irradiation in the kilocurie cobalt-60 gamma source. 2. Oxidation a. Using Elementary Oxygen. Sulfur dioxide and oxygen in a gaseous mixture of stoichiometric proportions were irradiated while under pressures of about 50 psig and temperatures of about 300~F. No sulfur trioxide was detected by iodometric-acidimetric titration procedures.3 32

Oxygen was bubbled through liquid sulfur dioxide held in the cobalt60 vault at temperatures of about 250'F until the total pressure reached about 800 psig. No visible drop in total pressure was observed when introduction of oxygen was stopped. However, when the reaction products Mere absorbed in water, considerable amounts of sulfate were precipitated by the addition of barium chloride. When sulfur dioxide and oxygen were agitated in aqueous solution at atmospheric pressure in the presence of gamma radiation, considerable amounts of sulfate were detected in the resulting solution. These results indicate the formation of sulfuric acid by the oxidation of sulfur dioxide dissolved in water. This can be compared with the results of Alyeal and Backstrom2 on the oxidation of sodium sulfite in aqueous solution in the presence of ultraviolet radiation. b. Using Other Oxidizing Agents. From time to time, when space has been available in the cobalt-60 source, single experiments were made on mixtures in which there was reason to believe, on the basis of other experience, that reaction might occur. In addition to investigating the possibility of brominating or iodinating benzene, as mentioned above, attempts have been made to react sulfur with benzene, pyridine, and naphthalene. Changes in odor after irradiation indicated that some reactions had apparently taken place, but these reactions have not been investigated further up to this time. While other workers have also used elementary oxygen in reactions carried out in radiation fields, the use of other oxidizing agents has not been reported. Solutions of potassium permanganate in acetone and in an acetone-benzene mixture have been irradiated in the kilocurie cobalt-60 gamma source. The solutions containing benzene were decolorized and yielded a dark precipitate which appeared to be manganese dioxide. No change was apparent in irradiated solutions of permanganate in acetone. Potassium iodide solutions were also irradiated in the cobalt-60 gamma source, using neutral solutions, some slightly acidic with acetic acid, and some 1 N with hydrochloric acid. Visible amounts of iodine were present in all samples irradiated for 30-40 hours. Titratable quantities of iodine (with 0.1 N thiosulfate) were present in the irradiated acidic samples. Irradiated aqueous acidic (acetic acid) potassium iodide solution displayed a measurable potential with respect to an unirradiated control solution, indicating a difference in concentration. Also, a measurable difference was detectable when each solution was measured separately with respect to a calomel half-cell. 3. Polymerization Acetone has been reported to decompose into free methyl and acetyl radicals.13 Acetone was therefore tried as a reaction initiator under gamma 33

irradiation. Irradiation appeared to change the odor of mixtures of acetone with ethanol and acetone with ethyl ether to some extent. It was consequently thought that some amount of ester may have been produced. However, distillation analysis on the Podbielniak column failed to yield any larger amount of any material other than the reactants. Mixtures of acetone and ethylene have been irradiated while under pressures of about 1000 psig. Yields of 1 to 5% by weight of white powder have resulted from these runs, and also with ethylene alone held at 900 psig and irradiated for about 50 hours with cobalt-60 gamma radiation. Solutions of 1% glacial acetic acid in styrene have been irradiated, and viscosity measurements indicate that the acetic acid activates the polymerization of styrene under gamma irradiation. Liquid isobutylene was irradiated for some time under its own vapor pressure. A small amount of liquid residue was isolated. This material had a terpene-like odor. When a mixture of 50% (by volume) of styrene in linseed oil was irradiated for about 20 hours in the kilocurie cobalt-60 gamma source it showed slightly increased viscosity and a slightly bleached color. A similar mixture of linseed oil and acrylonitrile was given 39 hours in irradiation time, producing a thick, thixotropic paste. Odor indicated that some monomer was still present in this mixture. 4. Other Reactions Certain similarities in the electron configurations of the C12 molecule and the H202 molecule led to the trial of hydrogen peroxide as a reactant with benzene. In some runs acetone was used as a mutual solvent. In one such run a crystalline product with a sharp melting point was obtained, but subsequent runs produced explosive products, which were probably organic peroxides. This work has been temporarily discontinued. The reaction of benzene and chlorine is promoted by both ultraviolet and gamma radiation. It is therefore of some interest to compare the behavior of systems which have been subjected to known amounts of radiation from each of these two ranges of frequencies. For purposes of comparison, the benzophenoneisopropanol reaction was selected, since under certain conditions of concentration, sunlight, and geometrical arrangement of apparatus, this solution will show a visible precipitation of the bimolecular reduction product, benzopinacol, in about 5 hours.8 A sample of this solution failed to show any such reaction after about 50 hours of gamma irradiation in the kilocurie cobalt-60 source. Apparently this reaction is not appreciably promoted by gamma radiation. The energy absorbed by this system in the kilocurie cobalt-60 gamma source was estimated to be 0.15 cal/ml-hr. This value is based in part on the calibration of the cobalt-60 gamma source as given in Progress Report 3.6 In addition, the rate of energy absorption was calculated for this system when exposed to sunlight. The fraction of the solar constant in the range 3300A to 3400A was

estimated. The sun was assumed to be a black-body radiator, and the Planck radiation function was employed, as given by Jahnke and Emde.10 It is further assumed that no absorption of incident solar radiation occurs in this frequency range, either by the atmosphere or by the glass container. When all radiation in the range of 3300A to 3400A was assumed to be absorbed by the reaction mixture, a value of 0.39 cal/ml-hr was found. It will be noted that the rates of energy absorption for this system in sunlight and in the source are of the same order of magnitude. C. FUTURE WORK The investigations of this laboratory indicate that the reactions which are most likely to be accelerated by ionizing radiation are those which proceed by a free-radical chain mechanism. The evidence to date also indicates that dense phases (liquids or gases under very high pressures) are necessary to absorb appreciable amounts of gamma radiation to produce sufficient effects for the products to be measured by the usual macro-methods of chemical analysis. Future work will, therefore, follow lines consistent with these observations. It is expected that reactions of liquids at atmospheric pressure and varying temperature, such as the benzene chlorination, will receive continued emphasis, as will reactions of gases under pressure such as the polnmerization of olefins or the hydrogen-carbon monoxide reactions. Every effort is being made to control the temperature of the reactions more closely, since this is so important in comparing reactions with and without radiation. Reactions which normally take place at high temperature are especially important in this study; it is desirable to determine whether gamma radiation can cause them to proceed at a rapid rate at a lower temperature. If other reactions are found to proceed as rapidly under gamma radiation as the chlorination of benzene, these reactions will probably be given special consideration and studied in some detail. However, the broad approach of trying a number of reactions is still very much in effect in the program at Michigan. D. BIBLIOGRAPHY 1. Alyea, H. N., "Chain Reactions Produced by Light and by Alpha Radiation", Jour. Amer. Chem. Soc., 52, 2743 (1930). 2. Backstrom, H. L. J., Jour. Amer. Chem. Soc., 49, 1460 (1927). 3. Bodenstein, M., and Pohl, W., Z. Fur Elek., 11, 373 (1905).

4. Anderson, L. C., Martin, J. J., et al., Utilization of the Gross Fission Products, Progress Report 1, (C00-86) Eng. Res. Inst., Univ. of Mich., Ann Arbor, Mich. (Aug. 31, 1951). 5. Ibid., Progress Report 2 (C00-90), Jan. 31, 1952. 6. Ibid, Progress Report 3 (COO-91), June 30, 1952, p. 84. 7. Chem. and Eng. News, 30, 3078 (1952). 8. Fieser, Laboratory Manual, 2nd Ed., p. 202-4 9. Handbook of Chemistry and Physics, 33rd Ed., Chemical Rubber Publishing Co., Cleveland, Ohio, 1951-1952. 10. Jahnke, E., and Emde, F., Tables of Functions, 4th Ed. Dover, New Nork, 1945, addenda, p. 44. 11. Selke, W. A., Engel, S., Kardys, C., Jazel, R. C., and Sherry, E. V., Utilization of Waste Fission Products in Chemical Reaction, Columbia University, May 5, 1952. 12. Slator, A., Z. Physik. Chem., 45, 540 (1903). 13. Steacie, E. W. R., Atomic and Free Radical Reactions, Reinhold, New York, 1946, p. 200.

Progress Report No. 5 UTILIZATION OF TEE GROSS FISSION PRODUCTS The Effect of Gamma Radiation on Chemical Reactions J. J. Martin C. E. Eckfield L. C. Anderson J. P. Holmes D. E. Harmer R. L. Kinney J. G. Lewis E. M. Rosen

TABLE OF CONTENTS Progress Report No. 5 Page SUBPROJECT M943-4, THE EFFECT OF RADIATION ON CHEMICAL REACTIONS 21 A. Introduction 21 B. Chemical Reactions 21 1. Polymerization 21 2. Chlorination 26 3. Oxidations 31 4. A Dosimetry Reaction 33 C. Thermodynamics and Kinetics of Reactions 36 D. Dosimetry Problems 40 1. Experimental Procedure 40 2. Calculation Procedure 43 3. Consideration of Neutron Activation Equations 55 E. Dose Rate Within a Cylindrical Pressure Reactor 57 F. Equipment Changes 62 G. Future Work 68 H. References 68 LIST OF ILLUSTRATIONS Figure Page 28. Effect of Oxygen on Rate of Ethylene Polymerization 25 29. Ethylene Yields for Successive Runs 25 30. Rate of Ethylene Production for Successive Runs 25 31. Energy Distribution Among Molecules 37 32. Log Ktr Versus Reciprocal Temperature Plot 39 33. Fraction of Activated Molecules-as a Function of Dose Rate 39 34. Location of Samples —lO-kc Source Dosimetry 41 35. Location of Samples —10-kc Source Dosimetry 41 36. Source with Negligible Wall Thickness 43

LIST OF ILLUSTRATIONS (Continued) Figure Page 37. Source with Finite Wall Thickness 46 38. Differential Equation and Its Integration for Source with Finite Wall Thickness (Dosimetry Calculations) 46 39. Dose Rate on Axis of 10-kc Cobalt-60 Source 48 40. Dose Rate on Midplane of 10-kc Cobalt-60 Source 48 41. Dose Rates Parallel to Midplane Interpolated from Measurements —10-kc Source 48 42. Dose Rates Parallel to Axis Interpolated from Measurements —10-kc Source 48 43. Isodose Surfaces Interpolated from Measurements —10-kc Source 49 44. Dose Rate on Axis of l-kc Cobalt-60 Source 553 45. Calculated Dose R. tes Parallel to Midplane —l-kc Source 53 46. Calculated Dose Rates Parallel to Axis —l-kc Source 53 47. Calculated Isodose Surfaces —l-kc Source 53 48. Increase in Activity During Neutron Irradiation 56 49. Gamma Source and Pressure Reactor Shown Diagrammatically 57 50. Attenuation of Gamma Radiation by Absorption and Distance 58 51. Heaters for Pressure Reactor 62 52. Wiring Diagram for Heaters for Pressure Reactor 63 53. Details of Heaters for Pressure Reactor 63 54. Rack for Pressure Reactors 63 55. Pressure Reactor in Rack and Connected to High Pressure Gas Cylinder 64 56. Rack for Pressure Reactor 64 57. Sling for Pressure Reactor 64

LIST OF ILLUSTRATIONS (Continued) Figure Page 58. Cone Joint to Iron Pipe Thread Adapter 64 59. Location of Tubing between Fission Products Laboratory and 10-kc Source 66 60. Details of Tubing Shown in Figure 59 66 61. Socket Wrench Brazed to End of Electrical Conduit Used for Installing Aluminum Tubing for 10-kc Source 67 62. Glass Apparatus Used in Chlorination Runs 67 63. Glass Apparatus Used in Chlorination Runs 67 Table II. Irradiation of Ethylene with Cobalt-60 Gamma Rays 22 III. Irradiation of Propylene in l-kc Cobalt-60 Gamma Source 24 IV. Distillation at 4.55 cm Hg Pressure of the Products of a Toluene Chlorination Run with Gamma Irradiation 27 V. Distillation at 4.50 cm Hg Pressure of the Products of a Toluene Chlorination Run Without Irradiation 28 VI. Effect of the Presence of Benzyl Chloride on the Reaction Between Benzene and Chlorine 29 VII. Irradiation of Oxygen-Benzene Mixtures in l-kc Cobalt-60 Gamma Source at 1 atm abs and Room Temperature 30 JIII. Irradiation of Oxygen-Toluene Mixtures in l-kc Cobalt-60 Gamma Source at 1 atm abs and 68~F 31 IX. Irradiation of Potassium Permanganate - Toluene Mixtures in l-kc Cobalt-60 Gamma Source at 1 atm abs and Room Temperature 32 X. Irradiation of Iodine - Toluene Mixture in l-kc Cobalt60 Gamma Source at 1 atm abs and Room Temperature 33 XI. Irradiation of Solutions of Potassium Iodide in l-kc Cobalt-60 Gamma Source at 1 atm abs and Room Temperature: Possible Dosimetric Method. 34

LIST OF ILLUSTRATIONS (Concluded) Table Page XII. Irradiation of Ferrous Sulfate Solutions in Cobalt-60 Source - Dosimetry by Method of Weiss (Data from 10 -Kilocurie Source Unless Noted Otherwise) 42 XIII. Dose Rates on Axis of 10-kc Source 47 XIV. Dose Rates on Midplane of 10-kc Source 50 XV. Estimates of Activities from Measurements of Dose Rates 51 XVI. Dose Rates on Axis of l-kc Source 52

LIST OF ILLUSTRATIONS (Concluded) Table Page XII. Irradiation of Ferrous Sulfate Solutions in Cobalt-60 Source - Dosimetry by Method of Weiss (Data from 10 -Kilocurie Source Unless Noted Otherwise) 42 XIII. Dose Rates on Axis of 10-kc Source 47 XIV. Dose Rates on Midplane of 10-kc Source 50 XV. Estimates of Activities from Measurements of Dose Rates 51 XVI. Dose Rates on Axis of l-kc Source 52

PART II. SUBPROJECT M943-4, THE EFFECT OF RADIATION ON CHEMICAL REACTIONS Personnel: Subproject Supervisors: Joseph J. Martin, Associate Professor of Chemical and Metallurgical Engineering; and Leigh C. Anderson, Professor of Chemistry and Chairman of Department. Research Assistants: David E. Harmer and John G. Lewis Assistants in Research: C. Eckfield; J. Holmes; R. Kinney; and E. Rosen. A. INTRODUCTION Reference to previous progress reports will reveal the general nature of the studies being carried on at this laboratory. During the period covered by this report, the activities of the personnel of this subproject have not been confined strictly to determining experimentally, the effect of ionizing radiation on the promotion of chemical reactions. Consequently, the work is being reported under four major topic headings as follows: Chemical Reactions, Thermodynamics and Kinetics of Reactions, Dosimetry Problems, and Equipment Changes. These topics will now be considered in that order. B. CHEMICAL REACTIONS 1. Polymerization The polymerization of ethylene at pressures as high as 1600 psi and temperatures up to 400~F has received considerable attention. As shown in Table II, the early runs showed fairly high yields of a fluffy white powder. This was true for the runs in which the ethylene and a small amount of acetone were irradiated and for the runs in which ethylene alone was irradiated. The acetone was added with the idea that it would release free radicals on radiolysis and initiate chain polymerization of the ethylene. Later runs have 21

TABLE II IRRADIATION OF ETHYLENE WITH COBALT-6o GAMMA RAYS Ave. Ave. Grams Position Starting Pressure Grams Hours Other Page Pres., Temp. Polymer Source of Base Date psig psia OF Polymer Irrad. Reactants 1521153 19 Dec 32 785-830 822 69 1.6+ 69.7 0.02+ 1 ml acetone 1-kc - 132118 12 Jan 55 795-820 822 70 0.666 61 0.011 - 1-kc 152147 3 Mar 55 1330-1450 1395 250 0.292 17.2 0.017 - 1O-kc 8" NE 9" UP 152149- 7 Mar 53 f700-820 f775 50 0.004 r 0.9 0.0005 1O-kc 8" NE 9" UP 1424-435 445 132159 26 Nov 52 780-790 800 68 0.594 42 0.009 1 ml acetone 1-kc 132250 10 Mar 55 715-810 777 50 0.167 16.2 0.010 - 10-kc Qon Qat Base 132252 11 Mar 55 1330-1450 1395 200 1.04 16.1 o.o6 - 10-kc Do. 135225 12 Mar 55 692-698 710 45 1.48 87.7 0.017 - 10-kc 8" NE 9" UP 132254 17 Mar 53 655 670 45 - 23.3 - [650 psi NA 10-kc 8" NE 9" UP 15 psi Air LVent, Chg. C2H4 132255 18 Mar 55 710-800 770 45 0.088 21.5 0.004 - 10-kc 8" NE 9" UP 132256 19 Mar 55 760-840 8i5 420 0.581 15.5 0.025 - 10-kc 8" NE 9" UP

132258 23 Mar 53 70-810 455 45 0.010 39. 0.0003 10 ml acetone 10-kc 132259 26 Mar 53 183-788 485 61 - 15.3 - - 10-kc ~ on~ at Base 132262- 4 May 53 1425-1600 1528 72 - 21.9 - - 1-kc I 132263- 5 May 53 595-620 623 72 0.022 23. 0.001 - l-kc II 132264 7 May 53 600-615 622 69 0.0001 20.7 0.000,005 100 ml 02 l-kc 132265 8 May 53 608-620 631 67 0.0001 21.4 0.000,005 25 ml 02 l-kc 132266 9 May 53 577-594 601 72 0.034 25.6 0.0013 5.0 ml 02 l-kc 132267 10 May 53 595-600 603 75 0.001 17.0 0.000,05 2.0 ml 02 1-kc 132268 11 May 53 597-617 622 70 0.017 22.3 o.oo08 10.0 ml 02 1-kc 132269 12 May 53 590-605 612 71 0.030 21.1 0.0014 10.0 ml 02 1-kc 132270 13 May 53 629-637 638 66 0.016 14.8 0.0011 5.0 ml 02 1-kc 132275 4 Jun 53 525-550 553 70 0.031 16.8 0.002 - 10-kc 8" sw 13-7/8" UP TABLE II (concluded)

produced very erratic results, and as noted in Table II, the yields have fallen considerably. At first it was thought this might be caused by the presence of oxygen in the system, and therefore a number of runs were made in which a known amount of oxygen was introduced. Although the results of these runs showed that the extent of reaction is dependent on the amount of oxygen, as shown in Fig. 28, the effect was far smaller than the decrease in reactivity from the early runs to the later runs. See Figs. 29 and 30. Therefore, it seemed unlikely that oxygen alone could be responsible for inhibiting the reaction. It should be noted that some uncertainty exists as to the slope of the curve of Fig. 28 for oxygen content near zero. Other ideas which have been suggested are that the cylinder of ethylene contained impurities which acted either as promoters for the early runs or as inhibitors for the later runs. Alternative possibilities are that there might have been some polyethylene powder in the original cylinder of ethylene, that impurities might have been introduced by leakage during evacuation, that impurities might have been adsorbed on the packing or sealing washers, that acetone might have been present in the early runs even though it was not intentionally added except for three runs, and that some kind of separation reaction had taken place in the feed cylinder which resulted in different products over a period of time. These various ideas are being investigated and it is hoped to have a more definite information soon on just what causes ethylene to polymerize and what inhibits that polymerization in the system under investigation. It should also be mentioned that the analysis for polymer after irradiation has been very crude. The procedure involves bleeding off all the gas, opening the reactor, and scraping out the fluffy powder. Analyses of the bleedoff gas are contemplated for future runs, as well as molecular-weight determinations on the polymer that is formed. Liquid propylene was irradiated for 67.6 hours at a dose rate of about 25,000 rep/hr; the results are reported in Table III. A small quantity of a TABLE III IRRADIATION OF PROPYLENE IN 1-KC COBALT-60 GAMMA SOURCE Starting Hours Dose Rate, Temp., Pres., Date Date Irrad. Rep/hr Dose, Rep psia Remarks 132140 6 Feb 53 1735 0. 65,ooo + 10% 0. 71. 147 2.8-in.-depth liq. propylene in bomb; 3000-lb gauge. 132140 9 Feb 53 1313 67.6 6s, oo000 + 10% 4,400,000. 655 145 Out of vault. 132140 18 Feb 53 1000 -- -- -- Room 25-30 0.2-0.3-in.-depth liquid in bottom, quickly evaporated.

WEIGHT POLYMER PRODUCED, GRAMS p 0 0 0 - - - N o n,:~ m iD;o io b 0~ b b o O'q i'~~ Ci) (1) ~ ~ WEIGHT POLYMER FORMED / HOUR,GRAMS xIO4 0 -- b) b 0 ODo I-~ O) O/ _Ft~, O O ~ ct o D, CD mo C+l CC CD Z cc o (D K r I I 0I 1 h) ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ R O ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ O 0 CD~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C (D rn ~ ~ ~ ~ ~ ~ ~ ~ HD o -J o o~ (D hC;aa Pi I 1 I I i Ic Cf~ ~ ~~~~~~~~~~~~~~ r~~~~~~~~C ro E (- -~ H co mm WEIGHT OF POLYMER, GRAMS / HOUR N~ Z ru~... t~ I I-' I: o ~r 3 0 ~~~ ~~~ P P P ~~~~~~ 00 ~~~~ O~0FO ON_ __~~~ CD 0 N ~~ 0) CD 0 N ~~~ 0) oc~( 00 H-'0 H H~~~~~~~~..- -.I o (Dr c0 m o r~ t- G) WEIGHT OF POLYMER, GRAMS — HOUR N f'o o,.. __-. __ _. __ __. _ __ c~f Pa l h u~~e o bCb b b b b CD CD p 0 0 ODP, a f 0 o'Q m4 1(D 0 10 CO - (D 0~~~~~~~~~~~~~~~~~~~~~~~~~~( co ";0 (D O Z 0 c 0m CD (D Z;1 rC o rlo

volatile liquid was found in the reactor when it was vented and opened as in the ethylene runs. Undoubtedly some polymerization took place, although the quantitative amount has not yet been determined. 2. Chlorination The addition of chlorine to benzene under gamma radiation to form the addition compound, benzenehexachloride, has been reported by this laboratory previously. The speed and extent of this reaction have led to investigations of the chlorination of other hydrocarbons such as toluene. Also, further work is anticipated on the kinetics of the benzene chlorination under gamma radiation to compare it with the reaction induced by ultraviolet radiation. The reaction of chlorine with toluene has been carried out at 200C. The experimental runs consisted of a 10-minute nitrogen purge through the toluene and the entire system, followed by passing an excess of chlorine gas through the toluene for 30 minutes during irradiation. After this, nitrogen was again passed into the system for 10 minutes, in order to allow the apparatus to be disconnected without escape of chlorine gas into the room. Equipment for the runs was identical to that used for benzene runs described in Progress Report 4 An attempt was made to separate the products of this reaction by distillation at atmospheric pressure. Fractions were obtained up to 225~C, but a vigorous decomposition began at 170'C with evolution of hydrogen chloride, About 30 per cent of the total chloride content of the sample was given off in this decomposition. When the original sample was extracted with water, it was found that less than 1 per cent of the total chloride passed into the aqueous layer, indicating that the chloride was not present in an active or ionic form. A distillation of the products of another toluene chlorination run was carried out in a column packed with glass helices and operated at a pressure of 4.55 cm Hg. Table IV summarizes the fractions and some of their properties. The organic chloride analyses were carried out by the sodium diphenyl reagent method. Table IV shows that the largest single fraction (excepting the recovered starting material) is that boiling between 206 and 216~C (fraction X). This fraction has a chlorine content which is remarkably close to the theoretical percentage for a hexachloro addition product of toluene. Because of the possibilities which might be realized from the production of a new type of toluene derivative, further investigations are being made. A control run was made on the reaction of chlorine and toluene in the absence of gamma radiation. All conditions were the same as for the runs previously described. Table V summarizes the fractions obtained from the distillation of the reaction product mixture at 4.50 cm Hg pressure. This control run 26

TABLE IV DISTILLATION AT 4.55 CM HG PRESSURE OF THE PRODUCTS OF A TOLUENE CHLORINATION RUN WITH GAMMA IRRADIATION Volume, Temperature Per Cent Fraction Comment ml Range, ~C Chlorine I 28.5 35-360 Unreacted toluene II and III 8.9 62-104~ IV 5.6 95-96~ 26.5 Mostly benzyl chloride (pure benzyl chloride is 28% Cl) V to VII 6.2 96-127~ VIII 2.8 142-157~ 56.2 Mostly benzo trichloride (pure is 54.4% C1) IX 4.6 157-180~ X 15.0 206-216~ 68.8 Probably the hexachloro addition product. Theoretical chlorine content is 69.8%. XI 1.0 216-222~ Slight decomposition Residue 12.0 70 shows that very little reaction takes place in the absence of gamma radiation. Experiments which will allow comparison of the effectiveness and nature of products obtained under gamma irradiation and ultraviolet irradiation are contemplated. 27

TABLE V DISTILLATION AT 4.50 CM HG PRESSURE OF THE PRODUCTS OF A TOLUENE CHLORINATION RUN WITHOUT IRRADIATION Temperature Per Cent Fraction Volume Comment Range, ~C Chlorine I 85.5 340 Toluene (identified by odor) II 1.9 80-87~ 25 Mostly benzyl chloride or chlorotoluene (theoretically 28% C1) Residue 4.1 68 High chlorine content, but dissimilar appearance to fraction X of Table IV. Since fraction X of Table IV appeared to be somewhat unique, it was investigated further. If the structure is assumed to be the hexachloro addition compound of toluene, it should be possible to remove three molecules of hydrogen chloride and obtain 2,4,6-trichlorotoluene, or some corresponding isomer, by procedures similar to those which yield trichlorobenzene from hexachlorocyclohexane. C1 H 3 3 HC1 H C1 The thermal decompositions observed during distillation might lead to this type of compound in addition to other products. A sample of fraction X from Table IV was refluxed with pyridine for 63 hours. Aliquot portions of the pyridine mixture were placed in a mixture of equal volumes of benzene and ether and extracted with 0.1 N sodium hydroxide. The alkaline extracts were analyzed for ionic chloride. Of the total chlorine content of the original fraction, 52.8 per cent was found in this aqueous layer. This fact indicates that the hexachloro addition product of toluene was probably present in the original 28

material and suffered elimination of three molecules of hydrogen chloride while being teated with pyridine. The proof of structure is being given further consideration. In an effort to ascertain whether the addition reaction could be applied to other substituted aromatic systems, benzyl chloride, benzal chloride, and benzotrichloride were each treated with chlorine under gamma irradiation at 20~C for 30 minutes. In each case, practically no increase in chlorine content was found and the starting material remained virtually unchanged. Because of the nearly complete nonreactivity of the side-chainsubstituted chlorine compounds of toluene, it seemed valuable to ascertain the effect their presence would have on the rate of reaction of chlorine with benzene under gamma irradiation. A series of runs was made in which mixtures of benzene and benzyl chloride were treated with chlorine under gamma irradiation for a 15-minute interval, at the end of which time the solid benzene hexachloride was separated and weighed. Table VI gives the results of these runs. TABLE VI EFFECT OF THE PRESENCE OF BENZYL CHLORIDE ON THE REACTION BETWEEN BENZENE AND CHLORINE Mole Per Cent Per Cent Yield of Benzene Temperature Rise of Benzyl Hexachloride in over Control Point Chloride 15 Minutes 5 0.72 None observed 4 0.01 None observed 2 3.6 None observed 1 5-7 10~F after 13 minutes 0.5 5-6 22~F after 9 minutes 0.5 20~F after 10 minutes The above table shows that the presence of benzyl chloride has a pronounced effect on the rate of addition of chlorine to benzene. The exact mathematical character of the function is not satisfactorily indicated by the data because of the experimental difficulties which arose at the minimum and maximum concentrations. The time for running the experiments was set at 15-minutes because previous work in this laboratory has shown that solutions of benzene 29

TABLE VII IRRADIATION OF OXYGEN-BENZENE MIXTURES IN 1-KC COBALT-60 GAMMA SOURCE AT 1 ATM ABS AND ROOM TEMPERATURE Starting Hours Dose Dose, Page Time Id Rate Procedure - Observations Date Irrad. Re/hr Rep Rep/hr 132121 17 Jan 53 1615 0 65,000+10% 0 100 ml Benzene into vault 132121 17 Jan 53 1626 0.18 12,000. Oxygen on 132121 17 Jan 53 1705 0.83 54,000. Oxygen off 132121 17 Jan 53 1707 o.87 57,000. Out of vault No changes in temperature or tail gas rate. 132122 20 Jan 53 - - - - 5 ml of 132121 No visible Ppt. in liquid. Product with Cloudiness in benzene sat'd. Bromine layer. water, Evaporate on watch glass 132122 21 Jan 53 - - - Small oily smear on water glass, smelled of halogenated hydro-carbons.

diluted with carbon tetrachloride reacted almost completely in this length of time. Although great care was taken to control the temperature of the reaction by adequate cooling, some of the reactions displayed a tendency to "run away" after an initial period of good temperature control; this effect was more marked at lower concentrations of benzyl chloride. In the last run reported in Table VI, the mixture heated up so rapidly that chlorine escaped and no quantitative yield could be ascertained. 3. Oxidations Several attempts have been made to oxidize partially some aromatic compounds by procedures analagous to those used in chlorinations runs. Pure oxygen was bubbled through benzene for 30 minutes while the mixture was being subjected to gamma irradiation at the rate of 50,000 rep/hr. The results are reported in Table VII. No changes in temperature or tail-gas flow rate were noted during irradiation, no changes were noted in the color or odor of the product, and no visible reaction occurred between a sample of the product and saturated bromine water. An attempt to oxidize toluene as reported in Table VIII, was carried out in a manner similar to that just described for benzene. The same equipment and dose rate were used and oxygen was bubbled through the toluene at 680F for 1/2 hour. No changes were noted in temperature, tail-gas flow rate, color, or odor of the reaction product. TABLE VIII IRRADIATION OF OXYGEN-TOLUENE MIXTURES IN 1-KC COBALT-60 GAMMA SOURCE AT 1 ATM ABS and 68~F Dose Page Starting Hours Dose Page Starting Time Hours Rate Dose, Procedure Observations Date Irrad. Rep/hr Rep 132126 31 Jan 53 1708 0. 65,000+10% 0. 100 ml Toluene into source. 132126 31 Jan 53 1712 0.07 65,000+10% 4,500. Oxygen on 132126 31 Jan 53 1732 0.40 65,000+10% 25,000. Oxygen off 132126 31 Jan 53 1735 O.45 65,000+10% 30,000. Toluene out of No noticeable vault change in toluene 31

TABLE IX IRRADIATION OF POTASSIUM PERMANGANATE -TOLUENE MIXTURES IN l-kc COBALT-60 GAMMA SOURCE AT 1 ATM ABS and ROOM TEMPERATURE Dose Page Starting Tim ours Rate, Dose, Procedure Observations Date Irrad. Rep/hr Rep 132125 29 Jan 53 1056 0 65,000+10% 0. 0.5g permanganate 25 ml acetone, 25 ml toluene 132125 29 Jan 53 1400 3.1 65,000+o10% 200,000. inspect sample Supernatant liquid clear pink when shaken. 132125 29 Jan 53 1409 3.1 65,000+10 - resume irradiation 132125 29 Jan 53 1530 4.5 65,000+10% 300,000. inspect sample still faint pink O 132125 29 Jan 53 1531 4.5 65,000+10% - resume irradiation 132125 29 Jan 53 1616 5.1 65,000+10% 330,000. inspect sample Supernatant liquid clear. Faint pink when shaken. Pungent odor. 132125 29 Jan 53 1627 5.1 65,000+10% - resume irradiation 132125 30 Jan 53 1519 28.0 65,000+10% 1,800,000. inspect sample Supernatant liquid clear and remained clear on shaking; decolorized bromine water; slowly decolorized a solution of permanganate in acetone; did not react visibly with aqueous bromidebromate. The central sample had also decolorized.

In a different type of oxidation a sample consisting of 25 ml of acetone, 0.5 gram of potassium permanganate, and 25 ml of toluene was irradiated 28 hours at a dose rate of about 50,000 rep/hr, as reported in Table IX. Observations were made at intervals, and it was noted that there was a gradual bleaching of the permanganate color. The unirradiated control sample also bleached, but not as rapidly as the irradiated sample. A solution of 0.25 gram of iodine in 50 ml of toluene was irradiated 44.5 hours at a dose rate of about 50,000 rep/hr, as reported in Table X. No difference was discernible between the irradiated sample and the unirradiated control. TABLE X IRRADIATION OF IODINE -TOLUENE MIXTURE IN 1-KC COBALT-60 GAMMA SOURCE AT 1 ATM ABS AND ROOM TEMPERATURE Page Starting Time Hours Dose Rate, Dose, Rep Procedure Observations Date Irrad. Rep/hr 132130 4 Feb 53 2108 44.5 65,000+10% 2,900,000. 0.25 g. iodine No discernible dissolved in difference be50 ml toluene tween run and control kept in dark. 4. A Dosimetry Reaction Preliminary tests were made of the use of starch-iodide solutions for dose measurements. A solution of 1 gram of potassium iodide and 1 drop of glacial acetic acid in 100 ml of water was prepared together with a 1-per cent water solution of soluble starch. Samples consisting of 1 part of the starch solution to 19 parts of the iodide solution were subjected to doses of 5,000 to 300,000 rep, and the results are reported in Table XI. Visual observations indicated correspondence between dose and depth of color. Since the above results indicate that an iodide system might be used for dose measurements, application of the method is considered: The possibility of measuring the concentrations of iodine produced in the oxidation of an iodide by means of a concentration potential for the solution of the mixture was discussed in Progress Report 4. Consider the usual voltage equation for an electrode dipping into a solution of iodine and iodide: 2.303 RT (Y7 BCR)b E = E~ loglo N F (-YACA) 0.05915 log0 (YBCB)b = ~ 10 ( /ACA) 5533

TABLE XI IRRADIATION OF SOLUTIONS OF POTASSIUM IODIDE IN 1-KO COBALT-60 GAMMA SOURCE AT 1 ATM ABS AND ROOM TEMPERATURE: POSSIBLE DOSIMETRIC NETHOD Stock Solution was: 100 ml water, 1 g potassium iodine, 1 drop glacial acetic acid. Starch Solution was: 100 ml water, 1 g soluble starch. Statig Hur Dos Rate, Page Startng Time urD Rose/hRat Dose, Rep Observation 132127 3 Feb 55 1430 0 65,000+10* 0. Add 1 ml starch soln. to 100 ml stock soln. 5 samples, 25 drops each, then put into vault Remove one at stated time. 152127 5 Feb 55 1435 o.o83 65,ooo+10* 5,500. Faint blue tint 152127 3 Feb 55 1500 0.50 65,000+10* 33,000. Moderate blue w, 152127 5 Feb 55 1650 2.0 65,000+10* 150,000. Purplish blue 132127 3 Feb 55 2050 6.0 65,000+10% 390,000. 2 out-both blue-black; put one sample into clean tube, added starch, color darkened. 132128 4 Feb 55 0950 0. 65,000+10% 0. Add 75 drops starch soln. to 90 ml stock soln. and starch from 152127. Made 5 samples, 25 drops each, put all into vault. Remove one at stated time. 132128 4 Feb 55 0935 0.083 65,000+10* 5,500. No change - clear. 132128 4 Feb 5 09)40 0.17 65,000+10* 11,000. No change - clear. 132128 4 Feb 55 1010 0.7 65,000 1 45,000. Moderate blue. 132128 4 Feb 55 1550 6.o 65,000o+10* 390,000. Deep purple; control clear. 132129 4 Feb 55 1050 0. 65,ooo+10* 0. 10 g potassium iodide, 100 ml water, 1 drop glacial acetic acid, 5 ml starch soln. Procedure as above. 132129 4 Feb 55 1035 0.083 65,000+10* 5,500. Faint blue. 132129 4 Feb 55 1040 0.17 65,000+10* 11,000. Light blue. 132129 4 Feb 55 1110 0.7 65,000+10* 45,000. Deep blue; control clear.

where E = the single-electrode potential at 25~C of the electrode, E~ = the single-electrode potential of the electrode when the concentration of all reactants and products involved is one molar (unit activity), F = coulombs per Faraday of charge, N = number of electrons involved in the electrode reaction, CA = concentration of the reactant, iodide, for which the activity coefficient is Y A CB = concentration of the product, iodine, for which the activity coefficient is /'B, and a and b are the stoichiometric coefficients of the reactants and products respectively, in the electrode reaction. Now let two identical aqueous solutions of acidic iodide be prepared, such that a reaction will occur in either of them when it is subjected to gamma irradiation. Then concentration of iodine in the irradiated solution will be increased by a small amount. A concentration cell employing platinum electrodes and these two solutions, one irradiated and one not, will then have a potential as follows: E = E1- E2 = (E 2.303 RT (CB)b control Y Bb E0 - -log,0 NF 0 (CA)a control 7Aa - _ 2.303 RT (CB)b test r7Bb -NF log(C )a test a N Aa where CA = iodide concentration, and CB = iodine concentration. Since the iodide concentration of the control solution is approximately equal to the iodide concentration of the test solution for small degrees 35

of reaction, the above equation simplifies to: b CB test E = 0.05915 log10 CB control (7) Equation (7) shows that a small change in the amount of product may be detected by measuring accurately the oxidation potential corresponding to the ratio of the test value to the control value. C. THERMODYNAMICS AND KINETICS OF REACTIONS In the study of chemical equilibria, the standard free energy of reaction is a criterion of the degree of reaction to be anticipated at equilibrium in a given system. As usually defined, a negative standard free energy for a given reaction indicates the possibility of obtaining a preponderance of the product of the reaction in the equilibrium mixture. In addition, a reaction with a negative standard free energy of reaction may proceed spontaneously, but a positive free-energy change precludes the possibility of a spontaneous reaction. Only reactions with negative standard free energies of reaction will be considered in the following discussion. It is often assumed that the standard free energy of reaction is merely the net change in the energy of the system, and that the rate of the reaction may be explained on the assumption that a certain energy of activation must be possessed by a molecule or group of molecules before a reaction can occur. (See Fig. 31). In a given system the fraction of molecules possessing a given energy is a statistical function of the energy possessed. Energies greater than or equal to the energy of activation needed for chemical reaction will be possessed by a certain fraction of molecules at a temperature above absolute zero. The rate of a chemical reaction proceeding as a result of thermal activation will vary exponentially with temperature if other conditions are fixed, since the number of molecules possessing energies equal to or greater than the energy of activation will vary exponentially with temperature. As an alternative to thermal activation, energy might be introduced into a system in the form of electromagnetic radiation. For photon energies from 500 kev to 5 mev the predominant process would be ionization as a result of Compton scatter. The resulting ions or excited molecules, radicals or other excited species would have large energies compared with those ordinarily derived from statistical distribution of thermal energy. Each excited species activated by the encounter with radiation would probably possess sufficient

TEMR I X TEMP. 2 c I) O M ~w ~Fi. z 0 ow tr 0 z W=: W w ow-'EI E2 E3 E4 ENERGY OF MOLECULE ~ --- Fig. 31. Energy Distribution Among Molecules. energy to react. Therefore, radiation might be used to induce chemical reactions in a system at a temperature too low for these reactions to occur measurably by thermal activation. come the activation barrier were left to the Maxwell-Boltzmann distributiion of energies among the molecules, the fraction reacting in unit time would be a function of the temperature only. In addition, for an exothermic reaction the net energy of the reaction is released and increases the random energy of the system as a whole. Thais in a sense the passage of each molecule over the barrier of activation helps to raise other molecules over this barrier. However, if ions are produced by the absorption of highly energetic electromagnetic radiation and these ions react directly or indirectly to produce chemical products, then such an occurrence would not be a function of the temperature of the system. Rather the probability of such an occurrence would be governed by the intensity and energy of incident radiation. The energy of reaction and probably also the energy of ionization (barring fluorescence) would be released to the system and cause an increase in the random energy of the system as a whole, similar to that noted by thermal activation. However, there are cases in which the temperature is so low that this additional energy is not elctomgntc adato ad heeios eatdieclyo idiecl t podc

sufficient to increase thermal activation by a measurable amount. Therefore, each molecule reacting as a result of radiation activation does not "lift" another molecule over the barrier of activation. Under these conditions, molecular interactions induced by radiation would be random events when considered as a function of the reacting system only, but the number of such interactions would be proportional to the dose rate of radiation. Furthermore, since the rate of reaction due only to thermal activation increases exponentially with the temperature, it should be possible to attain a temperature above which the contribution of radiation to the rate of reaction would become negligible compared with that of thermal activation. In a chain reaction the energy of activation is transmitted directly from reacted molecules to unreacted molecules. It then becomes unnecessary for the energy of activation to be distributed at random among all molecules and then concentrated by random motion to activate the unreacted molecule. In effect the rate of a single-hit reaction is maintained in the chain reaction, but multiplied by the chain length. The influences of radiation and temperature on the rate of a given reaction may be described quantitatively as follows: The Arrhenius equation may be assumed to hold for the specific reaction velocity constant, kt. kt = Ae-Et/RT (8) where R is the gas constant, T is the absolute temperature, Et is the energy of activation, and A is a frequency factor for the reaction. It can be shown5 that the factor e-Et/RT represents the fraction of molecules having energies equal to or greater than Et, the energy of activation. By postulate, these molecules are the only ones with sufficient energy to initiate reaction. The specific reaction velocity constant, ktr, of a reaction occurring in a field of radiation may be examined and equation (8) may be modified to: ktr = A(e-Et/RT + g), (9) where g is the fraction of molecules activated by irradiation. Equations (8) and (9) may be rearranged and combined as follows: in (ktr - Ag) = In Ae-Et/RT (10) = in kt 38

W ktr- REACTION VELOCITY CONST. J z DUE TO COMBINED TEMP. a 0 RADIATION; T- ABSOLUTE TEMPERATURE I- DOSE RATE OF RADIATION @ I: i 5 < = AT 1/,| J | > AT 12 | go 9(12 o g L3) AT 13 l/T, I,DOSE RATE - Fig. 32. Log Ktr Versus Reciprocal Fig. 33. Fraction of Activated MolTemperature Plot. ecules as a Function of Dose Rate. The validity of equation (10) may be tested as described below. Note that a plot of in (ktr - Ag) versus 1/T should be linear. Consequently a plot of in ktr versus 1/T would be expected to be curved (see Fig. 32). If such a plot were obtained experimentally (as by Bretton et al.2), it could be straightened by the subtraction of an empirically determined parameter from ktr* The parameter determined by this method would equal Ag, the part of ktr due to radiation alone. The values of Et and A could be closely approximated from equation (8) at temperatures sufficiently high that ktr would be nearly indistinguishable from kt. Then the value of g could be computed from equation (9). Then g could be plotted as a function of the dose rate, as shown in Fig. 33. Thus the separate effects of temperature and gamma radiation on the kinetics of a reaction with no temperature-dependent chain could be studied. Figure 51 indicates that the fraction of molecules with energy sufficient to react is almost solely a function of radiation effects at temperature 1. However, at temperature 2 the area under the curve due to thermal excitation 39

in the region lying above E3 is a considerable proportion of the total area lying above E3. These observations illustrate the conclusion drawn from equation (10) above. That is, the rate of a reaction induced by gamma radiation is dependent on the dose rate and may be nearly independent of temperature over a wide range of temperatures less than temperature 2. However, at temperatures above the neighborhood of temperature 2 the rate of reaction becomes a function of both the temperature and dose rate due to radiation. At still higher temperatures the reaction due to thermal activation is sufficiently great that the rate due to radiation becomes small by comparison. D. DOSIMETRY PROBLEMS For any quantitative study of the effect of radiation on chemical reactions, it is necessary to know the dose rate of the radiation. Considerable work has been done since the last progress report on measurement of dose rates. The following discussion is taken from an article on that subject which was written by four of the authors of this report (Harmer, Lewis, Martin and Nehemias) for the 1953 Gordon Research Conference on Radiation Chemistry. 1. Experimental Procedure Dose rates were measured chemically by the method employing the oxidation of ferrous sulfate solutions. Dilute solutions of ferrous sulfate (5 x 10-4 M) in aerated 0.8 N sulfuric acid solution were exposed to gamma radiation for doses of between 5 and 20 kilorep. The ferric ion produced by the gamma radiation depends on the presence of a small amount of oxygen, which is readily furnished by first passing air through the solution. For quantitative determinations of the ferric ion produced by irradiation, the spectrophotometric method described by J. Weiss9 of Brookhaven National Laboratory was employed. This method makes use of a spectral absorption peak of ferric ion at about 304 millimicrons in the ultraviolet region. Optical densities of the irradiated solutions are measured at 305 millimicrons and compared with those of known ferric solutions made up by dilution of standardized ferric stock solution. In converting the chemical yield to radiation dosage, a value of 15.4 micromoles/liter-kilorep was used. This value is based on the absorption of 93 ergs/gm of water for each equivalent roentgen of radiation. The solutions were irradiated in glass bottles about 3 cm in inside diameter, which were filled to a depth of about 4 cm. The bottles of solution were placed inside and outside the l0-kilocurie source, as shown in Figs. 34 and 35, and inside the l-kilocurie source. (See 40

PLAN '0 SOURCE 7i 1 1.3 " F DIAg. 4. Location of2.4"T Samp HOULDERslO-kc Source Dosimetry6 IiI -3.82- 0 ALL BOTTLES VERTICAL DURING _ ~....-. gI IRRADIATION EXCEPT 2,4 a 5, WHICH 16-7-390 - T 6 WERE HORIZONTAL __.._V.:?, 1 -1- 17 55.13 1-1 Fig. 5. Location of Saples ---kc Source Dosimetry. 5.13IOI~T1ON WEPT 2.485.WH41

Table XII). Proper exposure times were calculated to fall within the range of the method of ferric ion determination. Measurement of dose rate in the l-kilocurie source was carried out at times separated by an interval of 1 year, and values were found to be consistent after corrections for radioactive decay were applied. Measurements using a ceric sulfate system were also made and found to agree within experimental error with the ferrous sulfate results. TABLE XII IRRADIATION OF FERROUS SULFATE SOLUTIONS IN COBALT-60 SOURCE - DOSIMETRY BY METHOD OF WEISS9 DATA FROM 10-KILOCURIE SOURCE UNLESS NOTED OTHERWISE Page 132308 Page 132306 Date 16 Mar 53 Date 13 Mar 53 See Fig. 34 For location of Samples See Fig. 35 For Location of Samples Sample Dose Rate, Sample Dose Rate, Number Kilorep/Hr Number Kilorep/Hr 1 280 1 144 2 250 2 194 3 242 3 249 4 244 4 261 5 292 5 266 6 342 6 281 7 38 7 243 8 102 8 234 9 80 9 248 10 8 10 274 1 kilocurie 57 11 168 1 kilocurie 60 12 86 13 115 14 52 15 24 16 13 17 22. 18 17 19 74. 20 42. 21 4.4 22 2.6 23 2.0 1 kilocurie 55. 42

Physical determinations of dose rate have also been carried out on both sources. Two instruments have been employed in these determinations: a Victoreen roentgen rate meter, which measures the current flow between electrodes in an ionization chamber which is placed in the radiation field; and a Victoreen r-meter, which measures the drop in potential of a charged condenser due to ionization current caused by the radiation. The rate meter was calibrated against radium standards by the manufacturer, while the r-meter was calibrated against a cobalt standard at the University of Michigan. Within 50 cm of the center of the 10-kilocurie source the rate-meter readings were 15 to 20 per cent lower than the ferrous sulfate determinations. The r-meter readings were 15 to 20 per cent higher than the ferrous sulfate measurements in the 10-kilocurie source, and were 25 to 30 per cent higher in the 1-kilocurie source. The detailed significance of these differences is not clear. 2. Calculation Procedure Since gamma radiation from a point source may be assumed to follow the usual inverse-square relation, it is possible to calculate the dose rate at any position in the neighborhood of a source of known shape and total actiz vity by an integration technique similar to that employed in radiant heat transfer. If the geometry of the source shape is complicated, the resulting integration may be difficult. As a simple shape somewhat similar to the two P(R,O,Z, cobalt-60 sources, consider first a hollow cylinder of negligible wall thickness and assign to it the power or curie rating of the actual source. The dimensions of the cylinder are taken to dA correspond as nearly as possible to those of the actual source, and the assigned curies are assumed to be distributed uniformly over the surface of the cylinder. Then the contribution to the radiation intensity at any given point due to an element of source area, dA, iYY 7Kat a distance p is given by equation (11); see Fig. 36. SHEET SOURCE dA dI =. (11) Fig. 36. Source With Negligible Wall Thickne s s. 4.

The total intensity at the given point is obtained by summing the contributions from all elemental areas as Z=L 0=: Z=L rdGdZ (12) Z=O I=0 R2+r2-2Rr cosG + (Z Z)2 (12) Integrating equation (12) gives 2cr R+r 1 R+r(13) I 2R+r F(tan ZL, k)-F(tan-1 R+rk (13) for Z> L >O. R> O. r> 0 0 <tan-1 R+r tan1 Rr < 2 k7l 2 and, -2rr { 2K(k)- F(tan-l R,k)+F(tan1 R+r k) (14) R+r L-Z1 Z for L> Z> 0, R> 0, r> O, Rfr, O 0tan- LRr, tan- 1 < L- Z 2 1 An alternative form may be obtained as shown by equation (15). Dewes and Goodale3 have indicated the preliminary steps in this development. 2cr -l Z1 1 Z1-L I = R+r F(tan 1 r-R k)-F(tan-1 r-R k) (15) for >0 R> O, r> 0, R/r, '- 1ZL 2 <ta n- 1 1 <2 ~for - <~,02 Ir-RI 2' Ir-RI 2 A relation given by Hancock4 permits the transformation of equations (13) and (14) into equation (15), and vice versa. Equation (15) is considered more convenient in most computations, except for R=r, Z1 >L, where equation (13) may be used to advantage. The symbols used above are defined as follows: I = dose rate, equivalent roentgens/hour; A = area of source; 44

p = distance from elemental area dA to the point at which I is taken; = (total activity, curiesy10millicuriesyequivalent roentgens at 1 cm area of source, cm2 A curie Ahour x millicurie point source r = radius of source or constant radius vector of cylinder; R = radial distance of point at which I is taken from axis of source; 9 = central angle from R to r; Z = distance parallel to axis of source from base of source to element d Z1 = Z coordinate of point at which I is taken; k = 2/R; R+r F(O,k) = elliptic integral of first kind of modulus k and amplitude 0; and K(k) = complete elliptic integral of first kind of modulus k. Self-absorption of a hollow cylindrical source of finite thickness may be approximated along the axis of the source by the following procedure (see Fig. 37). It will be assumed (1) that the source is of uniform unit-volume activity and density, (2) that absorption occurs only within the source, (3) that scattered radiation due to the absorber will not affect the dose rate, (4) that radiation intensity and dose rate vary inversely with the square of the distance from a point source and inversely with an exponential function of absorber thickness, and (5) that the part of the source lying outside the cone, 0 = tan-1 [r/(Z1-L)], also fulfills the foregoing assumptions. The resulting differential equation and its approximate integration are as shown in Fig. 38, where P = distance between point and element of volume; p = density, grams/cm3 = mass absorption coefficient, cm2/gram; V=total activity, curies millicuriesquivalent roentgens at 1 cm V = 00== volume of source, cm3 curie hour x millicurie point source dv = element of volume of source; and all other terms are defined as above or in Fig. 37. 45

(Z,) d l= V e-L P EQ(1'6) I L VL rd9dZdr -/-LP(r-r)csc EQ.(17) Jz =0 ~ r-2+ ( Z, -Z)2 if a three-term approximation to the exponential is employed: I \ +27V 2r2 r2 'r sin r = | V |2+ (2V/l ZpL(arrL-rsin,,-sinl'. - ) r2 + - Z,-L In ()2I 2 2 In O'Vp t(-2, a r2' r2 Fr. (sinite Wall Th sin h metry Calculations ele + (Z,- L) r, In + Z,-r, In Z_ 2 2 1(+1 '2

Equations (15) and (18) were applied to both the 1000- and 10,000-curie sources. In the case of the 1000-curie source, it was simple to assume a cylinder with dimensions corresponding to the actual cobalt cylinder. In the case of the 10,000-curie source, the nest or bundle of 100 rods was assumed equivalent to a cylinder whose inside and outside diameters were the shortest and longest diametrical distances across the rod bundle. The 10,000 curies was assumed to be uniformly divided throughout this volume and the density of the assumed cylinder was taken so that its mass equalled that of the rods themselves. Calculated and observed values of dose rate for the 10-kilocurie source of cobalt-60 are compared in Tables XIII and XIV and are plotted in Figs. 39 and 40. The calculated values were based on an assumed activity of 10,000 curies. The observed values are considerably less than the calculated values. TABLE XIII DOSE RATES ON AXIS OF 10-KC SOURCE L Z1- 2 = Distance above midplane, cm; R = O Calculated Rep/hr Measured Rep/hr for 10,000 curies Sheet Ferrous Victoreen L Annular Source Source L Oxidation* Rate Z1- 2. Z1- 2 Meter No With No Absorption Absorption Absorption 0 1,020,000 830,000 1,010,000 0 242,000 (16 Mar 53) 6.35 910,000 747,000 928,000 2.5 249,000 12.7 665,000 527,000 662,000 2.5 234,000 25.4 222,000 154,000 218,000 8.9 194,000 38.1 97,000 61,000 96,000 15.2 144,000 63.5 - - 32,000 20.3 ------- 61,000 21.6 74,000 22.8 ------- 48,ooo 25.4 ------- 38,ooo 26.7 42,000 38.1 16,000 50.8 8,100 60.8 5,000 76.2 3,300 * 13 Mar 53 unless noted. 47

O 10,000l CURIE POINT SOURCE -CALCULATED, NO ABSORPTION,, 0 10,000 CURIE ANNULAR SOURCE -CALCULATED, NO ABSORPTION X 10,000 CURIE SHEET SOURCE -CALCULATED, NO ABSORPTION,h A 10,000 CURIE ANNULAR SOURCE -CALCULATED, WITH ABSORPTION. 10,000 CURIE POINT SOURCE 5 \ \ \ V FER9~~~ROUS OX[DATION MEASURE- < / -CALCULATED, NO ABSORPTION 5V FERROUS OXrDATION MEASURE- X 10,000 CURIE SHEET SOURCE 10 MENTS, 13 MARCH '3 1000 CURIE SHEET SOURCEPTION.- + VICTOREEN RATE METER MEASU- -CALCULATED, NO ABSORPTION I X y ~~\ REMENTS, MARCH 53 V FERROUS OXIDATION MEASURE-&J I t\ I \< RENI 'II InENTS, 13 MARCH '53.O FERROUS OXIDATION MEASUREzcr x~ - ~ MENTS, 16 MARCH'53 U; W ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~O h __ A FERROUS OXIDATION MEASURE\ z MENTS, 9 JUNE '53 z:w y-\ z z S T /- + VICTOREEN RATE METER I^Jw \ MEASURE, _ MEASUREMENTS, MARCH '53 OU. Y 1. /.- I1VICTOREEN R METER D 10 v MEASUREMENTS, MARCH '53 > \ 0 I CI 1.0 1O 100 DISTANCE ABOVE MID-PLANE, CM. LO 10 100 DISTANCE FROM AXIS, CM. Fig. 39. Dose Rate on Axis of 10-kc Fig. 40. Dose Rate on Mid-Plane of 10-kc Cobalt-60 Source. R 10.67 I-1 25.4 * i I I I I I 111 1 1 _ I II I I I I 111 a4 33.1 r I 1 [! I |lb - I- I I 1 1111. I R10. 0.1 1. R10 I 4 25.~~~~ ~48 0 \\ R:77.6 I I 0.1 T. 0 100 DISTANCE FROM MID-PLANE,: CM. 48

40 E 30 20 Lu i0oooo~ I HR o 10 2.0 30 4( 50 60 70 80 90 DISTANCE FROM AXIS, cm Fig. 43. Isodose Surfaces Interpolated From Measurement —10-KC Source. For any given method of measurement the observed values are a nearly constant fraction of the calculated values. In Figs. 41, 42, and 43 appear cross plots of equation (15) when the latter is made to agree with data from the oxidation of ferrous ion. The data were taken on the midplane and on the axis in March, 1953. The source was irradiated at the Chalk River NRX reactor and was rated at 9250 curies on shipment from the Chalk River site in January, 1953. The activity computed from each means of measurement appears in Table XV. In the extreme right column of Table XV the ratio of the curies estimated from observed values of dose rate to the 9250-curie nominal value after correction of the latter value for decay appears. If self-absorption is not considered, the activity is estimated to be from 17 to 33 per cent of the nominal value. If selfabsorption is considered, the activity is estimated to be from 28 to 33 per cent of the nominal value. These figures are computed from data taken both on the midplane and on the axis. No estimate of self-absorption was made on the midplane however. Calculated and observed values of dose rate and curies for the l-kilocurie source are compared in Table XVI and in Fig. 44. The 1-kilocurie source was irradiated at Brookhaven National Laboratory and was assumed to have a nominal activity of 1000 curies in July, 1951. The activity computed from each means of 49

TABLE XIV DOSE RATES ON MIDPLANE OF 10-KC SOURCE Z1 = = 12.7 cm; R = Distance from Axis, cm. Rep/hr for Rep/hr for Rep/hr for Sheet Source; Ferrous Oxidation Victoreen Meters R No Absorption R 13 16 R Rate RCalculation for Mar Mar Meter Meter 10,000 curies 53 5 0 1,000,000. 0 249,000. 242,000. 21.3 61,000. 0 234,000. - 4.85 1,120,000. 3.30 243,000. 250,000. 23.1 52,000. 8.70 1,800,000. 3.30 248,000. 244,000. 26.2 43,000. 9.70 oo 6.30 281,000. 280,000. 30.8 32,000. 10.7 1,800,000. 6.30 274,000. 292,000. 31.8 ------ 40,000. 12.0 1,150,000. 14.7 168,000. - 38.4 18,500. ------ 19.4 379,000. 18.1 86,000. 102,000. 51.0 15,500. 29.1 160,000. 20.6 115,000. 79,500. 64.0 7,500. ------ 38.8 92,000. 25.7 52,000. 73.9 ------ 7,500. 100. 14,000. 38.4 24,000. 140. 1,800. 51.0 13,000. 165 1,200. 89.1 4,400. 114.8 2,600. 140.0 2,000. 20.6 82,000 9 June 53

TABLE XV ESTIMATES OF ACTIVITIES FROM MEASUREMENTS OF DOSE RATES Mean Value Measurements Estimate of Activity, Curies* Dided Divided Source Where Method Date Maximum Minimum Arithmetic By Decayed Taken Mean Nominal Value 10-KC Axis Ferrous Mar 53 2500 2100 2300 0.26 Oxidation 3100A 2800A 2950A 0.33A Victoreen Mar 53 1700 1400 1550 0.17 Rate 2600A 2500A 2550A 0.28A Meter Mid- Ferrous Mar 53 2600 2200 2400 0.27 plane Oxidation Victoreen Mar 53 2500 2000 2250 0.25 Rate Meter Victoreen Mar 53 3000 2900 2950 0.33 R Meter 1-KC Axis Ferrous May 52 140 0.16 Oxidation 180A 0.20A May 53 130 0.16 170A 0.21A Victoreen Feb 53 150 110 130 0.16 Rate 190A 150A 170A 0.20A Meter Victoreen May 52 170 0.19 R Meter 230A 0.26A May 53 160 0.20 210A 0.26A * A after value indicates self-absorption was considered. 51

TABLE XVI DOSE RATES ON AXIS OF 1-KC SOURCE L Z1- 2 = Distance above midplane, cm R = O r = 2.493 cm Calculated Rep/hr Measured Rep/hr for 1,000 curies L Annular Source Sheet L Ferrous Victoreen Victoreen 1- 2 Source 1 2 Oxidation Rate Meter R-Meter No With No Absorption Absorption Absorption Feb 55 O 442,000. 341,000. 460,000. 0 62,300. 79,000.1 May, 52 | May, 52 8.75 429,000. 336,ooo. 450,000. 0 57,200.1 72,000. May, 53 J May, 53f 17.5 232,000. 184,000. 240,000. 1.3 51,600. 35.0 ------- 15,200. 3. 8 52,800. 87.5 -------- ------- 1,900. 6.3 54,000. 8.9 55,800. 11.4 57,500. 14.0 46,800. 16.5 19,200. 19.0 9,300. 21.6 5,150. 24.2 3,240. 26.7 1,860. 29.2 1,320. 34.2 490.

~~~~~~~~~~~~~~~~106 6 1,000 CURIE POINT SOURC 6 X -CALCULATED, NO ABSORPTI N 0 1,000 CURIE ANNULAR SOURCE -CALCULATED, NO ABSORPTION 17.5 -CALCULATED, NO ABSORPTION / --- 1,000 CURIE ANNULAR SOURCE -CALCULATED, WITH ABSORPTION 0' \ O VICTOREEN R METER MEASURE- 106 * 1O@ ' \ 1 MENT, MAY '52 zr!O \\uVICTOREEN R METER MEASURE- _ \ MENT, MAY '53 wn t + I++ \ V FERROUS OXIDATION MYEASURE- -_' + - [ ' MENT, MAY '52 ' \ \ o ~ous FERROUS OsIDATION MEASUREwlb 1 | ~~~ \ \ \ MENT, MAY 53 4 - -x- + VICTOREEN RATE METER MEASURE- 4 - 1 M I W MENT, FEB. '53 5 w I0.1 Cobalt-60 Source allel to Mid-Plane —1-kc Source. i06 1 _I0 u -~~~~~~~~~~~~~~~~~0 I 'l l] I I I 1 I I l! ] t ": t 1 l 8 7.5 -I. i / ~ ~___...._, [.:', I,, I1 I + ' 1 1 1 1 1 1 1 L 1 1 1 I I I I I I I I I 101. I., 10 00 DISTANCE FROM AXIS R, CM. 103,._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ I _ _ \ - 1.0 10 100 DISTANCE ABOVE MID-PLANE, CM. Fig. C44. Dose Bate on Axis of 1-kc Fig. 45. Calculated Dose Bates ParCobalt-60 Source. allel to Mid-Plane —l-kc Source. 106 ~~~~R '~~~~~~~~~~306 I 50,o ' i i \ o I I N I 4 2 0 030 1031 40 V-1" — ' 10 " 20 501 1 50 0 70 0.1 I. 10 i00 DISTANCE FROM AXIS,cm DISTANCE FROM MID-PLANE, 2I,-,CM Fig. 46. Calculated. Dose Rates Par- Fig. 47. Calculated Isodose Surallel to Axis —l-kc Source. faces —l-kc Source.

measurement is given in Table XV. The ratios of observed to decayed nominal curies appear in the right column. If self-absorption is not considered, the activity is estimated to be from 16 to 20 per cent of the nominal value. If self-absorption is considered, the activity is estimated to be from 20 to 26 per cent of the nominal value. These figures are computed from data taken on the axis only, since it was not possible to make measurements external to the source. However, it was desired to compare the dose rates predicted by equation (15) for the l-kilocurie source with those predicted for the 10-kilocurie source in order to observe differences caused by the different geometrical proportions of the two sources. Consequently Figs. 45, 46, and 47 are presented to portray the dependence of dose rate on position in the neighborhood of the 1-kilocurie source. The data for these figures were computed on the assumption that the source actually contained 1000 curies. Judging from the above results, there appears to be about a threefold discrepancy between the curies in the 10-kilocurie source as estimated from ionization measurements and as calculated from neutron absorption. The comparisons for the 1-kilocurie source are not so meaningful, since no firm estimate of the activity of the source was supplied by Brookhaven National Laboratory. The errors in the methods of calculation summarized in equations (13), (15), and (18) probably arise chiefly from the simplifying assumptions made. The assumption that the source has no thickness is evidently justified by the agreement of values calculated on this assumption with those in which the thickness of the source is considered (see Figs. 39 and 44). The results in Tables XII and XVI show that absorption is not negligible. However, in Figs. 39 and 40 it can be seen that the plots from data and from equations (13), (15), and (18) differ by an approximately constant factor between any pair of curves. This result is interpreted to mean that equations (15) and (15) may be used within limits to predict the distribution of dose rates without consideration of selfabsorption, but that accurate prediction of dose rates requires consideration of self-absorption. The l-kilocurie source is evidently not of uniform activity throughout its whole volume. This conclusion was reached from a study of Fig. 44. Note that the measured dose rates in the 1-kilocurie source do not vary with distance along the axis in the manner predicted by the calculated curves. The depression near the midplane is probably caused by lower unit activity inside the source in this region, which in turn is probably caused by failure of neutrons to penetrate to the interior of the source near the midplane as abundantly as near the ends.A The other assumptions introduced are thought to be reasonably acceptable, although equation (18) converges much more slowly as Z1 is increased. The value of 13.5 equivalent roentgens/hour at 1 cm per millicurie point source of cobalt60 was taken from the work of Marinelli, Quimby, and Hine7, and was assumed to be correct within our experimental error.

Although there were some differences between the chemical and physical dosimetry measurements, they were not sufficiently large to account for the factor of three or four between the nominal activities of the sources and those which result from the dosimnetry measurements themselves. In an attempt to understand this discrepancy, attention was turned to the calculations of activity of a cobalt sample which is subjected to neutron radiation. Neutron absorption calculations are generally used to predict activities of irradiated samples. 3. Consideration of Neutron Activation Equations Exact computation of the absolute amount of an isotope produced during neutron bombardment is dependent on precise knowledge of several factors. The relationship commonly used for such computations is: C =0.60cr Mr (e_-o.693t/T) = 5.7xl0 10A ~l-e /), (19) 3.7xlOlOA where C = total activity of isotope produced, curies; = thermal neutron flux, particles/cm2-sec; C = activation cross section, barns; A = atomic weight of irradiated isotope, grams; t = length of irradiation, arbitrary time units; T = half-life of isotope produced, same time units; M = mass of element present, grams; and r = fraction of irradiated isotope present. The factors which might contribute to a serious error in the computed value of C will be considered. Of these factors, A is known accurately in most cases and t, the length of irradiation, may be measured as accurately as desired, the limiting consideration being the length of time required for insertion and removal of the samples. Half-life may in general be measured as accurately as desired. Exceptions to this statement must be made for extremely long- and extremely shortlived isotopes. The buildup of activity under bombardment is exponential (Fig. 48). The plateau or saturation value of activity is simply the coefficient of the

SATURATION VALUE 0 15/16 7/8,,, 3/4 -nz to 1/2 01 O T 2T 3T 4T ST DURATION OF IRRADIATION T- HALF LIFE OF ISOTOPE RESULTING FROM IRRADIATION Fig. 48. Increase in Activity During Neutron Irradiation. bracketed term in equation (19). It can readily be seen that in the case of short-lived isotopes, i.e., for TP( t, the exponential term is of negligible importance, and that errors in the determination of T have little effect on the computed value of C at the time of removal. For long-lived isotopes, however, the exponential term becomes: l_e-0.693t/T -a 1- (1- +0.693t/T) = 0.693t/T (20) and errors in the determination of T have a direct effect on the computed value of C. The case of cobalt-60, with a half-life of the order of 5 years and irradiations of the order of months, is still on the nearly linear portion of the activation curve. However, recent determinations of the half-life of cobalt-60 are within 5 per cent of one another. Differences of this magnitude, although serious in many cases, cannot be considered as significant contributions to the discrepancy under discussion. The accurate determination of the remaining factors (neutron flux density, 0, and activation cross section, r ) is not as straightforward as those previously discussed. There are several reasons for this difficulty. First, the determinations as commonly made are interdependent. To measure 0 using an isotope of known a, or to measure a using a neutron source of known 0 is relatively simple. However, the calibration of the quantity to be used as a reference is more complex. A closer examination of these measurements might uncover a factor partially responsible for the observed discrepancy, but it seems unlikely that such a factor could explain a threefold error.

Second, in any practical case the observed neutron flux is not completely thermalized but contains a finite range of energies higher than thermal. The activation cross section is a function of energy. Therefore it is necessary in principle to know the neutron spectrum and the cross section as a function of energy. In general, however, the cross section at nonthermal energies is vanishingly small, with the exception of occasional resonance energies. Such activations could contribute a source of error to activity computations, but the computed value would then be too low rather than the observed high value. Third, local neutron depression caused by the insertion of irradiation samples of high cross section must be considered. It is this consideration which seems the most probable source of the observed discrepancy. Recent work of Levin and Hughes6 indicates that this correction is of the order of magnitude of the observed discrepancy. Therefore, at this writing it is believed that the activity of a source calculated from neutron absorption, such as the 9250 curies mentioned above for our source, may be several times too high because of the depression of the neutron flux in the vicinity of the cobalt. E. DOSE RATE WITHIN A CYLINDRICAL PRESSURE REACTOR All the foregoing calculations and dosimetry measurements are for points in air lying at different distances from the sources. In the experiments reported on the polymerization of ethylene, however, the reaction took place inside a stainless-steel pressure vessel. A series of calculations was therefore made on the intensity of radiation inside the pressure vessel, taking into account the absorption of gamma radiation by the walls of the vessel (see Fig. 49). Z..-L R REACTOR - \S - SOURCE Fig. 49. Gamma Source and Pressure Reactor Shown Diagrammatically. 57

It is assumed (1) that the source of gamma radiation is a cylindrical sheet of no thickness and of uniform activity per unit area, (2) that the source is transparent to its own radiation, (3) that the dose rate varies inversely with the square of the distance and inversely with an exponential function of absorber thickness, and (4) that no absorption occurs inside the reactor. Let the terminology be defined as in Fig. 49 and as follows: I = dose rate at P(R,Z), rep/hr; factivity of source, curies\ 0millicuriesquivalent roentgens at 1 c = (activitr of source, curies 00 area of source, cm2 curie x millicurie point sourc */ = absorption coefficient, cm2/gram; A = area of source, cm2; P = distance from dA at P(r,Z,G) to P(R,Z); P' = distance through bomb wall, cm; K(k) = complete elliptic integral of first kind of modulus k; kl = 2 JRr; R+r p = density of bomb wall, grams/cm3; and X = distance in Fig. 50 from source to point at which I is measured. See Fig. 50. POINT (P) AT WHICH DOSE RATE IS TO BE MEASURED SOURCE ABSORBER I 11 1 13 14 I —X X2 X3 - X4 X, I x2z xS x4 Fig. 50. Attenuation of Gamma Radiation by Absorption and Distance.

From assumption (3) above the following equation may be written: dl = -AIdX X- IdX (21) Integration and substitution of limits yields the expressions: 2 I'2 ()'l I(22) X 3 2 I3 x2 2 X 2 X2) (24) 2 3 Combining equations (22), (23), and (24) results in the expression: 4 =(K-2) I e/- X) - 5) (25) From equation (25) we may deduce that the location of an absorber is immaterial as long as it is between the source and point P. Only the thickness of the absorber need be considered. Now the equations may be written for dose rate on the axis of the bomb (see Fig. 49): = b csc 0, (26) R2+r2_- 2Rr cos G + (Z1-Z)2 dI = and eLP (28) dI = p-' -P (28) Equation (28) must be integrated over the entire source, as shown by equation (29). A three-term approximation to the exponential is employed.

I = J 2 czrdGdZ Z=0 =0 R2+r2-2Rr cos + (Z1-Z)2 -&bp /R2+r2-2Rrcoss + (Z!-Z)2 VR2+r2-2RrCos O +.2b2p2 (29) 2'. (R2+r2-2Rr cos g) J The first term within braces has been integrated above, and presented as equation (15). Integration of equation (29) yields equation (30). 2czar 1 Z1 1 Z1-L I = R+r LF(tan 1 k -F (tan Ir-Rl' k) ' ---i~~ I Z1-Z=Z1 K(k)dk CztrL 2b2 2 p fr-RI' I -r Y (30) 1' Z=Z12L -2c/b c/A R Z1-Z=Zl-L 1-(k/k 1)2 R2-r2 where k = k (Z1-Z) 1 V(R-r)'2 + (z1-Z)2 Equation (30) holds for Z3?. L> O, O <tan-l, tan -----, R-r Ir-RI Ir-RI 2 If, however, L Z1> o, Rzrf0 then I = 2, F(tan k) -F(tan Z1-L k) R+r Ir-RI lr-RI z _-z=~ K(k)dk Z__ K(k)dk - 2c.b z Pr z z L- i,-(k/k1) 2 z -zo 1-(k/k )2 Z1-Z=Z1-L 1 0(3) +czrL2b2p2 R26r2

Since k, defined above, is the modulus of an elliptic integral of the first kind, l1> k O. Consequently, for equation (31) the following definitions are employed. (Zi-Z) k = k(z-z) 1 /(R-r)2 + (Z1-Z)2 for Zl>_Z k = k1 (Z-z1) /(R-r)2 + (z1-_)2 for Z >Z1 If Z > L > O, R=O then I = 2 (tan 1 1 -tan ( L)) - 2bpln( Z1 + /r2 + Z2) (Z-L) + r+(ZL)2 + cxLp2b2p2 (32) r where o <tanl( 1 ) tan1 -L < <g/2 If, however, L> Z1> O, R=O, then I = 2act [tanl( r ) + tan-l (L-Z (L-Zl)+ /r2+(L-Zl)2] CZl+ /r2+Zl ] - 2a,/bpdln r2 a+L 2b2p2 (_) r 61

where where <tan-l( Z1 ) tan-1 ( rZ1) <i/2 In equations (30) and (31) terms of the form: K(k) dk may be integrated graphically. F. EQUIPMENT CHANGES A number of additions to and modifications of the equipment in the laboratory have been made during the past several months. These are as follows: Some special heaters were designed and built to maintain the stainlesssteel reactor at elevated temperatures. The heaters were designed to be added to or removed from the reactor readily and to absorb a minimum of gamma radiation. Chromel-A resistance wire previously wound around a 1/8-inch arbor was wound around porcelain tubes. The turns were insulated from each other by winding asbestos rope soaked in water glass between turns. An insulating jacket of asbestos paper and a protective jacket of steel shim stock were applied over each element. A photograph of the heaters appears in Fig. 51, and details of construction appear in Figs. 52 and 53. Fig. 51. Heaters for Pressure Reactor. A steel rack was designed and built for the purpose of positioning the stainless-steel reactor approximately symmetrically with respect to the 10-kilocurie source; see Figs. 54 and 55 for details. Extension legs (Fig. 56) permit use of the rack to hold the reactor on the axis of the source. The new steel 62

O-IOAMP, 0-135VOLT, BOMB HEATER # 14-2 COND. "ADJUSTAVOLT" R.C. AUTO TRANSFORMER \ AMP 0 A - 110 V. Fig. 52. Wiring Diagram for Heaters for Pressure Reactor. SCALE I2'1I'-0" A / \ A TWO " DRILL 16 3/32 DIAM. I MACNINE SCREWS WOVEN ASBESTOS INSULATION DRILL........18GA..... C L A ON' ~' 18 GA. CHROMEL A ONE16..I/....A.. CROME A I - 5: /SILVER SOLDERE6 TO ~I/B'-ARBOR -2 —'1 25. *126A. COPPER LEADTO CORD CAP I/8'ASBESTOS CORD - - - - _9-I * 1 I _ -_ _ N SOAKED IN WATERGLASS -IN I PORCELAIN TUBE 30 'AE "11- - WEL 0006"STEEL PLAN - SHIM STOCK PLAN O.OO6 STEEL ' i I/S6THICK ASBESS SECTION A-A ELEVATI ON Fig. 53. Details of Heaters for Fig. 54. Rack for Pressure Pressure Reactor. Reactors.

Fig. 55. Pressure Reactor in Rack Fig. 56. Rack for Pressure Reactor. and Connected to High Pressure Gas Cylinder. I SCALE: I"=1'-0" 8IXIX1L~~~ SCALE:I-I-O ^ t PRESSURE GAUGE-I /2" IPS BOLTS WITH WING NUTS MACHINE SEAT RETAINING SLEEVE -__________ CARBON STEEL SCALE FULLSIlEi BR w ADAPTER CONE AISI 304 2 DRILL 4 HOLES IN FIEL X T / X = 11 1 I 11B|RFOR /4 TUBE l l l r ( ~~~~~~~~E~ t~OLsE >~'9~ 41- - 3/8. ~EZJ ZZI~590 ASSEMBLY I/B- - 2 4 4 4L /12X XPDRILL TWO HOLES ON " O SCALE 4", I'-0 -- I - 2 2r~~~~~~~! lBDBRILLLL2 HOLES.A2AEILL~D~tiL '2 Ho ' I I Fig. 57. Sling for Pressure Reactor. Fig. 58. Cone Joint to Iron Pipe ' ----~6 Thread Adapter.

rack was designed to accommodate the sling (Figs. 56 and 57) previously constructed for the purpose of supporting the reactor with its attached tubing in the 1-kilocurie source. A special connection was constructed to permit attaching a compression tubing fitting to a pressure gauge having an iron-pipe-size thread. The details are given in Fig. 58. The pressure gauge fitting was machined out to provide a cone seat to match the standard cone on the end of the tubing. This kind of joint was used to provide a better seal than was thought to be possible by the use of tapered pipe threads. An access opening was provided during the construction of the shielding for the 10-kilocurie source; see Figs. 59 and 60. Electrical leads and aluminum tubing lines were placed in this opening to permit the operation of the pressure reactor remotely from the second floor of the laboratory. Two copper leads were also carried through this opening to connect a thermocouple cold junction in the vault with a potentiometer on the second floor. Assembly of the screwed compression joints for the aluminum tubing was accomplished by means of specially constructed socket wrenches (Fig. 61). Two wrenches were required, one to be inserted from each leg of the right-angled access opening and over the length of tubing extending out from the tubing ell. The wrenches were constructed by brazing spark-plug wrenches to the ends of pieces of electrical conduit of appropriate diameter. The socket was then turned by means of a pipe wrench. A special wooden holder was devised in order to protect and support glass tubing introduced into the same opening with the electrical leads and aluminum tubing. A modified form of the original glass chlorination reactor has been designed and constructed for use with the 1-kilocurie cobalt-60 source. This equipment is illustrated in Figs. 62 and 63. The use of concentric tubes joined at the top by means of ground joints facilitates cleaning of the reactor and at the same time assures a tighter seal for the gaseous contents than the earlier design afforded. In runs where a large heat of reaction is not anticipated, cooling may be done with water. The water is in turn cooled by recycling through a chamber containing ice frozen around a can containing a mixture of chloroform, carbon tetrachloride, and dry ice. If a large heat of reaction is expected, the water may be replaced with methyl alcohol cooled well below the freezing point of water. A knife-type heater has also been installed, actuated by means of a double-throw switch on the temperature controller, in order to provide better temperature control.

t.'-w-.;-F.p5 - ERMETO ELBOW PASSAGE LAB. ERMETO CHUTE FITTING TUBING At -g t - | SOURCE ROOM 3T,~ II 4 ERMETO FITTINGS;, ",' ' -|\REACTOR REACTOR RACK F18-4" \1 CONCRETE SHIELDING PLAN SCALE- 3/6" I' TUBING AS SEEN IN POSITION A-A SCALE- 3/8' I FISSION PRODUCTS < LAB. I CHUTE K ///////Z CEILING LEVEL SECOND FLOOR 7 I TUBING I kt~:~~~~9.~~~5e~~-~11O S REERMETO FITTINGS ~I *~ XAPC OU RCE k I -REACTOR RACK REACTOREILING LEVEL ELEVATION AS SEEN IN 0-8 Fig. 59. Location of Tubing Between Fig. 60. Details of Tubing Shown Fission Products Laboratory and in Figure 59. 10-kc Source. 10-kc Source.

v. Em a 5/8"SPARK PLUG WRENCH BRAZE SHIM ELECTRICAL CONDUIT, 1.0 2 1.17"Fig. 61. Socket Wrench Brazed to End of Electrical Conduit Used for Installing Aluminum Tubing for 10-kc Source. Fig. 62. Glass Apparatus Used in Fig. 63. Glass Apparatus Used in Chlorination Runs. Chlorination Runs. 67

G. FUTURE WORK It is expected that the work in the near future will concentrate on the following objectives: (1) Determine the conditions that cause ethylene to polymerize (and also not to polymerize) under the influence of gamma radiation. (2) Analyze more completely the products produced by irradiating ethylene. (3) Continue studies on the analysis of products from the chlorination of toluene. (4) Chlorinate other hydrocarbons. (5) Study the kinetics in the chlorination of benzene under gamma radiation. (6) Perhaps try some partial oxidations of olefins and aromatics and also some nitrations. H. REFERENCES (1) Anderson, L. C., Martin, J. J., et al., Utilization of the Gross Fission Products, Progress Report 4 (C00-124), Eng. Res. Inst., Univ. of Mich., Ann Arbor, Michigan, March, 1953. (2) Bretton, R. H., et al., Effect of Gamma Radiation on Hydrocarbon Gases, Progress Report IV (NYO-3311) Dept. of Chemical Engineering, Yale University, New Haven, Conn., October 30, 1952. (3) Dewes, R. A., and Goodale, E. E., Utilization of the Gross Fission Products, SO-1100, General Electric Laboratory, December, 1951. (4) Hancock, Harris, Elliptic Integrals, Wiley, New York, 1917. (5) Kennard, Earle H., Kinetic Theory of Gases, 1st ed., McGraw-Hill, New York, 1938. (6) Levin, J. S., and Hughes, D. J., "Flux Depression and Self Protection in the Production of Radio-Cobalt',' Nucleonics, 11, No. 7, 8 (July, 1953). (7) Marinelli, L. D., Quimby, E. H., and Hine, G. J., American Journal of Roentgenology and Radium Therapy, 59, 260 (1948). (8) Rosenzweig, W., A Dosimetric Study of High-intensity Gamma Ray Sources, BNL-1254, Brookhaven National Laboratory, November, 1952. (9) Weiss, Jerome, "Chemical Dosimetry Using Ferrous and Ceric Sulfates", Nucleonics, 10, No. 10 28-31 (1952). 68

Progress Report No. 6 UTILIZATION OF THE GROSS FISSION PRODUCTS The Effect of Gamma Radiation on Chemical Reactions J. J. Martin R. A. Carstens L. C. Anderson C. E. Eckfield D. E. Harmer J. P. Holmes J. G. Lewis R. L. Kinney F. Bashorm R. L. Klemm

TABLE OF CONTENTS Progress Report No. 6 Page SUBPROJECT M943-4, THE EFFECT OF RADIATION ON CHEMICAL REACTIONS 12 A. Introduction 12 B. Polymerization of Ethylene by Means of Gamma Radiation 12 1. Experimental Procedure 17 2. Results of Polymerization of Ethylene 17 3. Discussion of Polymerization of Ethylene 19 4. Evaluation of the Polyethylene Product 21 C. Chlorination of Aromatic Compounds 31 1. Description of Apparatus 31 2. Experimental Procedure and Results 38 D. Suggestions for Future Work 45 E. Summary 47 1. Polymerization of Ethylene 47 2. Chlorination of Aromatic Compounds 47 F. References 47 LIST OF ILLUSTRATIONS Figure Page 4. Radiation Yield as Function of Dose of Radiation in Polymerization of Ethylene 14 5. Rack and Sling for Pressure Reactor 18 6. Pressure Reactor: Tubing Assembly and Gas Cylinder 18 7. Molecular Weight as Function of Radiation Yield of Polyethylene 26 8. Melting-point Bar 27 9. Molecular Weight and Crystallinity as Functions of Radiation Dose for Polymerization 28 10. Crystallinity and Tensile Strength as Functions of Radiation Yield for Polyethylene 29 11. Stress-strain Plots for Test Specimens of Polyethylene 30

LIST OF ILLUSTRATIONS (Continued) Figure Page 12. Melting Points as Functions of Radiation Yield of Polyethylene 32 13. Apparatus for Reaction of Liquids with Chlorine Gas in the 10-kc Cobalt-60 Gamma Source 33 14. Individual Parts of the Apparatus for Reaction of Liquids with Chlorine Gas 33 15. Drawing of Apparatus for Reaction of Liquids with Chlorine Gas 34 16. Photograph Showing Detail of the Jet Injector for Gases Used on the Apparatus for Liquid-gas Injections in the 10-kc Source 355 17. Component Parts of the Glass Jet Injector, Disassembled 35 18. Schematic Drawing of Apparatus for Temperature Control of Reactions in the 1 and 10-kc Cobalt-60 Gamma Source 37 19. Apparatus for Reaction of Liquids with Chlorine Gas, Assembled to Go into the Portable Hood Next to the 10-kc Gamma Source 37 20. Portable Hood and Glass Apparatus in Use for Reactions in the 10-kc Source 37 21. Wooden Support Board for Glass Tubing to be Put through the Chute Providing Access to the 10-kc Source Room 39 22. Special Tool for Attaching and Removing Clips from Joints Located at Right Angle of the Access Chute to 10-kc Gamma Source 39 23. Diagram of Complete Set-up for Reactions of Liquids with Chlorine Gas in the 10-kc Source 40 24. Chart and Graph Showing Radiation, Temperature, and Material Balance at Various Times during a Reaction of Chlorine and Toluene in an Average Radiation Flux of 22 k rep/hour 42

LIST OF ILLUSTRATIONS (Concluded) Table Page 4. Analyses of Ethylene from Storage Cylinders 15 5. Analyses of Ethylene from Reactor 16 6. Irradiation of Ethylene 20 7. Properties of Polyethylene Produced 25 8. High Boiling Fractions Obtained from Redistillation of Products of Reaction between Toluene and Chlorine 41 9. Summary of Three Reaction Runs of Toluene and Chlorine in the 10-kc Cobalt-60 Gamma Source 44 10. Reaction of Chlorobenzene with Chlorine under Gamma Irradiation 46

PART II. SUBPROJECT M943-4, THE EFFECT OF RADIATION ON CHEMICAL REACTIONS Personnel: Subproject Supervisors: Joseph J. Martin, Associate Professor of Chemical and Metallurgical Engineering; and Leigh C. Anderson, Professor and Chairman of the Department of Chemistry. Senior Research Assistants: David E. Harmer and John G. Lewis. Assistants in Research: F. Bashore, R. Carstens, C. Eckfield, J. Holmes, R. Kinney, R. Klemm. A. INTRODUCTION Since the last report the efforts of this group have been concentrated on studying the effects of gamma radiation on two types of chemical reactions: (1) the polymerization of ethylene and other gases at high pressures; and (2) the chlorination of aromatic hydrocarbons. In the case of the ethylene polymerization, considerable amounts of ethylene polymer were produced as a powder and then formed, by heating under pressure, into sheets that could be subjected to physical testing. In the case of chlorination, major interest centered on producing and analyzing the addition product of chlorine and toluene. A large amount of this material was made and some of it was sent to the Department of Agriculture for testing as a possible insecticide. The following sections of the report discuss the details of these reactions, both of which are of sufficient importance to be considered seriously for patenting. B. POLYMERIZATION OF ETHYLENE BY MEANS OF GAMMA RADIATION The erratic nature of the results observed when attempting to polymerize ethylene by exposure to gamma radiation has been mentioned previously (see Progress Report 5.2 In the earlier work some attempt was made to correlate the observed rates of polymerization with oxygen content in the monomeric ethylene 12

and with the order of the run made. Neither of these approaches yielded satisfactory results. Other ideas were therefore advanced in order to account for the erratic polymerization rates observed. It was suggested that some polyethylene might have been present in the storage cylinders and might have been introduced during charging of the reactant to the pressure reactor; however, the conditions usually required for the polymerization of ethylene were unlikely to have prevailed in the storage cylinder. Another possibility was that some unknown inhibitor or some unknown promotor was present sporadically. The substances most likely to fall into these categories are impurities in the ethylene, gases from the air, materials used in cleaning the reaction equipment, and the reaction equipment itself. The last possibility was tested tentatively by allowing polymer to accumulate on the walls of the reactor and then checking the rate of reaction in a subsequent run; no influence on the rate of reaction was noted. The influence of various solvents and other materials thought possibly to have been present accidentally in the successful runs was checked by adding the following materials successively to separate batches of the reactant ethylene: acetone, acetaldehyde, air and acetone, air and water, carbon dioxide, sulfur dioxide, and aluminum chloride. Sulfur dioxide and aluminum chloride were the only additives producing detectable effects, and the latter material produced a tar instead of the white powder sought. The addition of sulfur dioxide resulted in the production of a white powder at relatively high rates of reaction (see Fig. 4); however, this powder proved to have a sulfur content rather close to that of the equimolar addition product of sulfur dioxide and ethylene. Matthew and Elder10 and Snow and Freyl2 have reported similar reactions between sulfur dioxide and olefins under ultraviolet light. Next, the composition of the reactant gases was examined in some detail. The ethylene was analyzed (see Tables 4 and 5) immediately on removal from the storage cylinders, after charging to the reactor but before irradiation, and on removal from the reactor after irradiation. Components determined were "soluble in bromine," carbon dioxide, oxygen, carbon monoxide, paraffin hydrocarbons, and nitrogen. Higher olefins and acetylenic compounds were not detected separately by the methods used. From the results of the above experiments it was concluded that an inhibitor could have been present and responsible for the erratic yields observed, but the inhibitor could not be identified. However, the answer was finally found when the reactor was then left in the radiation field for periods longer than those previously used. Larger yields of polyethylene were obtained as a result, and it was shown that after a certain minimum induction period the yield increased to a nearly constant and reproducible value. 13

10000 I I YIELD OF ADDITION PRODUCT OF SULFUR DIOXIDE AND ETHYLENE YIELD OF POLYETHYLENE z 0 0 w 0 10 x 2o Q. w 0 w POLYMERIZATION OF ETHYLENE V BY IRRADIATION WITH CEoT a- GAMMA RAYS 0 o CONDITIONS N MBIENT ROOM TEMPERATURE 0 e250 TO 1600 PSIA RESSURNE II O 104 N- CHARGED WITH ETHYLENE )0 w V OXYGEN ADDED TO ETHYLENE I 0 X NON-CONDENSIBLE GASES VENTED HFROM LIQUID ETHYLENE 1 SULFUR DIOXIDE ADDED TO ETHYLENE I C~ 1.0- - ' OTHER REACTIONS w A ADDITION PRODUCT OF SULFUR DIOXIDE AND ETHYLENE 0.1 arD0 2 3 4 5 6 7 DOSE, MEGARER Fig. 4. Radiation Yield as Function of Dose of Radiation in Polymerization of Ethylene. 14

TABLE 4 ANALYSES OF ETHYLENE FROM STORAGE CYLINDERS Material MgC2 q~z2 %coN2. Combustible Number of No. as marked Determinations 1 Math o.o6 0.02 0.02 0.15 0.1 propane? duplicate FF737 Math 2 Math 0.06 0.02+ 0.00 0.07 0.29 propane duplicate 5772 3 OC 0.10+ 0.02+ 0.00+ 0.47 total ethane? single sample H ~G28087 C and C 4 C and C 0.37 0.02 0.05 1.7 0.8 pentane triplicate JK370330 USI ~~5 U ~IC 0.08 0.05 0.005(?) 0.21 0.45 methane duplicate Ic-lo65

TABLE 5 ANALYSES OF ETHYLENE FROM REACTOR Material Dose G Page CO* 2* CO* N2 Combustible Number of Determinations No. No. as marked 4 3.09 2380 132363 0.37 0.02 0.05 0.00 3.8 ethane duplicate; first sample 0.62 0.07+ 0.07 of 1.1 or more 2 2.40 750 132366 0.06 0.02 0.00 0.12 0.14 ethane(?) duplicate 0.05 0.008 0.002 0.02 0.05 0.01 4 0.57 790 132370 0.34 0.01 0.01+ 2.49 total -?- duplicate 0.54 0.01 0.01+ 2 0.91 298 132372 0 0.02 0.00 0.23 total -?- duplicate; acetaldehyde added 2 6.97 2200 132373 o0o6 0.02 0.00 0.09 0.18 methane single 0.05 20.01 0.00 3 6.45 1915 1275 0.10 0.02 0.00 0.16 0.32 ethane(?) single 0.11 0.02 0.009 5 4.29 745 132376 0.08 0.05.5(?) 10 0.63 methane single 0.07 0.009 0.00 * Top, before irradiation (no removal by distillation and no addition to ethylene unless noted) Bottom, after irradiation

1. Experimental Procedure. In all of this work ethylene was irradiated with cobalt-60 gamma radiation at room temperature and at pressures of 250 to 1600 psi. Some tests were made in which ethylene was reacted alone and some in which the ethylene was used with other reactants. A stainless-steel bomb (Figs. 5 and 6) was used as the reaction vessel. The bomb was evacuated to pressures of less than 1 nm of mercury absolute and ethylene added from a cylinder. Pressures of about 1100 psi and room temperature were employed in most of the tests. The bomb was then placed in either the 1-kilocurie source or the 10-kilocurie source until the proper dose had been accumulated. After irradiation, the bomb was removed from the source, the unreacted ethylene was vented and analyzed by an Orsat analyzer, and the accumulated polymer was removed mechanically. In order to remove possible oxygen or other volatile gases from the ethylene, the bomb was evacuated, ethylene was charged under cylinder pressure, and then the ethylene was condensed by immersing the bomb in a flask containing dry ice. The bomb was then vented until the pressure had dropped to a predetermined value or until a given volume of gas had been released. The ethylene was then vaporized and the bomb and contents irradiated as before. 2. Results of Polymerization of Ethylene. In order to conduct a quantitative study of the effect of gamma radiation on the polymerization of ethylene, it was necessary to adopt some criterion of gamma radiation effectiveness. For this purpose it seemed desirable to calculate the radiation yield in terms of the G value, i.e., the number of molecules of ethylene undergoing polymerization per 100 electron-volts of energy absorbed from the radiation. The roentgen equivalent physical, or rep, was adopted as the unit of measure of the absorption of gamma radiation by ethylene. The rep was assumed to correspond to the absorption of 93 ergs per gram of absorber. The dose rate in rep per hour in the absorber was determined by a combination of methods. Chemical dosimetry was conducted, using the method of Weiss14 and the results were correlated and extended by the methods of Lewis, Nehemias, Harmer and Martin.8 The dose rates determined in this way were applied to the calculation of the radiation yields. The degree of polymerization of the ethylene was determined by terminating the reaction at a stage such that the conditions during reaction had been reasonably constant, venting the unreacted ethylene, and then removing the polymer mechanically and weighing it. The conditions prevailing during reaction were averaged and the average values were assumed to have prevailed throughout the course of the reaction. Instead of G. the quantity A was sometimes calculated in this work. A was defined as the gram moles reacted per metric ton subjected to one megarep. By a comparison of units the following relation can be established: 17

:~,:::~;::;:ic-I:8; ij:::::: iZ::::::ii::j: _-::::91:::Sil -a:::;::-_i:~:,;-:::.:: _: -,;,:,;i::i:: -:::,::::i,:i.,~:::;:;:dii iii:,:,:'':::. —;-: Q:d-a i:~:::::Juaii3'j:cX"'::'-:;;:':::' '":l!,pTil;ii ai;:i;:-::i::::::i: eriu-::i::::::najlia__::ii:: i,::;-: :::::::,: 1 w " "c; B 1 ; '"'" ii: S8 '::: --; —; --- —-; —;-~ ---;~;;~;;-; ---; —; —;-~ 'BSB" -%BBB$BBi8Si95-9 j6gi " enB- asw -I a8ggg I agpag t-l Bg CO SWX::,:: j '-P',-:::::r. y " 7 Fig. 5. Rack and Sling for Pressure Fig.6. PressureReactor: Tubing Reactor. Assembly and Gas Cylinder.

A gram moles reacted G molecules reacted (metric ton)(megarep) 1.04 " 100 electron volts Consequently, A is almost equal numerically to G. With some simplification the following relation was developed: A gram moles reacted _ weight fraction ethylene reacted x 106 ' (metric ton)(megarep) (molecular weight of ethylene)(dose, megarep) This simplified relation was used in calculating A values in the following work. A white, solid polyethylene resulted from the irradiation of ethylene with cobalt-60 gamma rays. See Table 6 for the experimental results. The yield of polymer was found to be quite small until the system had received a dose of about 1/2 megarep. The yield increased rapidly to a value of about 2500 gram moles reacted/(metric ton)(megarep) at about 3 megarep, and remained nearly constant up to doses of 7 megarep, the highest dose studied (see Fig. ), About one-third of the monomer was polymerized in three days in the center of the l0-kilocurie source. The amount of ethylene charged to the bomb was calculated from the observed temperature, pressure, and volume and the thermodynamic properties of ethylene taken from the work of York and White.5 3. Discussion of Polymerization of Ethylene. From Fig. 4 it can be seen that the yield of polymer per unit of energy absorbed from the radiation is a function of the total dose of radiation. This relation is evidently due to the presence of an induction period for the reaction. No correlation could be observed between contents of the following gases in the monomer and the yield as a function of dose: carbon dioxide, oxygen, carbon monoxide, hydrogen, paraffin hydrocarbons, nitrogen, and sulfur dioxide. It appeared, however, that the venting of noncondensable gases from the liquid ethylene did increase the initial rate of reaction a little. The data for the analyses of gases before and after irradiation are given in Tables 4 and 5. In view of the evidence just mentioned, it cannot yet be stated what effect chemical additives have on the rate of polymerization of ethylene by gamma radiation. It is possible, however, that the induction period is caused by the presence of chemical compounds other than ethylene. Average values of dose rates used in these studies varied from about 30 kilorep/hour to about 90 kilorep/hour. It should be noted, however, that errors exist in the method of calculating the dose rates used in estimating the G or A values. A Victoreen ratemeter was used to measure the dose rates on the axis of the bomb. This instrument would detect the secondary photons produced by scatter from the wall of the bomb, but probably would not detect 19

T A BLE 6 I RR A D IA TION O F E TH YL E NE Order Page Starting Ps ig Avg. Avg. Grams Bos Grass 0-Moles Reacted Averaged over Reactor Other Psto fClne Temp., ~ ~ ~ ~ ~ ~ ___________ Dose Rate7 Total DoseSore PstnofCldrRears No. No. Date PaisTa. Polyme~r 'I'rrd Polymer A = Metric Ton)(Meagr~ep) 1lNr MgsP Reactants Sore Reactor Mfr. No Initial Final -F per RourVlr H meapNo 2 132113 19 Dec 92 830 785 822 69 1.6* 69.7 O. o2+ 608. 29. 2.02 1 Ml acetone 1 kcc M- atheson FF 737 - 5 132118 12 Jan 53 820 795 822 70 0. 666 61. 0.011 291. 28.9 1.76 - 1 kcc4 132147 3 Mar 53 1430 1350 1395 230 0.292 17.2 0.017 485. 29.5 0.507 — 10 ho 8" NE 9"' up 5 13149II Ma 53 820 70 7 0 oo..00031274. o660kc 81NE9 u 435 424 445 15. 1 132159 26 Nov 92 790 780 8oo 68 0.394 42. 0.009 258 29.5 1.24 1 Ml acetone 1 kc - 6 132290 10 Mar 53 810 715 777 50 0.167 16.2 0.010 94.5 96. 1.55 — 10 kcc Bas at Ban 7 132252 11 Mar 53 il3o 1350 1395 200 1.04 16.1 0.6 513. 82.5 1.33 — 10 icc Do. 8 132253 12 Mar 953 698 692 710 45 1.48 87.7 0.017 423. 42.0 3.69 — 10 icc 8", NE 9", up 650 Psi N2 and 9 132254 17 Mar 53 635 655 660 45 -- 23.3 -- -- 42.0 0.979 l5 psi ais 1 lkc 8" NE 9"1up 10 132255 18 Mar 53 800 ~~~~~~~~~~~~~~~~~~~~~~~~~vent, chg. C2B 10 132255 18 Mar 53 8oo 710 770 45 o.o88 21.5 o.oo4 865. 42.0 0.904 — 10 icc 8" NE 9" up 11 132256 19 Mar 53 84o 760 815 420 0.381 15.3 0.025 1800. 29.3 o.448 — 10 kc 8", NE 9", up 12 132258 23 Mar 53 81o 70 455 45 0.010 39. 0.0003 12.5 41.9 1.63 10 Ml 10 kc- " La 13 132259 26 Mar 53 788 183 485 61 -- 15.3 -- — 95.5 1.46 — 10 icc Bas at Ban 14 132263-i 4 may 53 1600 1425 1528 72 -- 21.9-Ba7.7e at Ba1sce15 132263-II 5 May 53 620 595 623 72 0.022 23. 0.001 45.8 27.7 o.638 — 1 kc - 16 132264 7 May 53 615 600 622 69 <0.0001 20.7 <O-000005 <0.25 27.6 0.572 100 Ml 02 1 icc17 132265 8 May 53 620 6o8 631 67 <0.oooi 21.4 <0.000005 <0.24 27.6 0.591 25 Ml 02 1 icc18 132266 9 May 53 594 577 601 72 0.034 25.6 0.0013 67.5 27.5 0.703 5.0 Ml 02 1 kc - 19 132267 10 May 53 600 575 603 75 -.wO.01 17.0.0.00005 3. 1 27.5 o.468 2.6 Ml 02 1 icc20 132268 11 May 53 617 597 622 70 0.017 22.3 o.ooo8 36.6 27.5 o.613 10.0 Ml 02 1 icc21 132269 12 May 53 605 590 612 71 0.030 21.1 0.00o14 71. 27.5 0.581 10. 0 Ml 02 1 kc - 22 132270 13 May 53 637 629 638 66 m~16 14.8 0.0011 50. 27.5 0.407 5. 0 Ml 02 1 kc - 23 132275 4 Jun 53 530 525 553 70 0.031 16.8 o. oo2 94. 31.3 0.326 — 10 icc 8" SW 13-7/8", up 24 132276 7 Jul 53 490 490 505 76 0.021 22.1 0.001 63. 27.0 0.597 13 iccir -- 95 Ml airon 25 132277 8Jul 53 460 460 475 78 -- 20.1 -- 0.0 27.0 0.543 9Mlar 1kccro 26 132278 9 Jul 53 460 455 473 73 0.011 18.3 0.006 43.1 27.0 0.494 215 mmn 1 icc27 132279 10 Jul 53 445 440 458 -75 C 136 69.8 o.o02 147. 27.0 1.88 33 ma S02 1 icc28 132280 13 Jul 53 426 395 425 87 Lost, seves al 16.5 -- -- 27.0 0.446 4.598 gm 1 iccgramn is:- AiC3 29 132281 14 Jul 53 43o 420 44o 81 0.013 20.0 0.001 32.2 27.0 0.540 3 — 1 ic - 30 132282 15 Jul 53 120 410 130 79 0.657 23.2 0.028 184. 27.0 0.627 1 atm SO2 1 icc -"" A0: 31 13241 u s 420 385 417 83 2.692 16.1 0.167 241. 27.0 0.435 S0 1 icc AS02 = 3380 32 132285 17 Jul 53 393 70 246 83 n0.9. 64.0o cv43 2o 27.0 1.73 239gm1kAS =40 33 132287 20 Jul 53 400 390 410 -- -- -- -- -- 0 0 27.9 ga ----- S02 54 132293 3 Aug 53 810 790 815 60 0.208 17.0 0.012 357. 26.8 o.456 — 1 icc- Ohio Chea. 028087 - 35 132294 4 Aug 53 930 910 935 76 o.472 21.8 0.022 443. 26.8 o.584 — 1 ic - 36 132295 5 Aug 53 615 610 628 76 0.010 16.0 0.006 3.26.8 o.428 Alk. pyrogallol 1 icc- Matheson FF 737 - 37 132297 6 Aug 53 955 950 968 75 4.45 71.5 0.o623 730. 26.8 1.91 — 1 icc-"" Powdery, soft lup, ht 38 132350 10 dug 53 320 320 335 75 0.005 15.7 0.0003 35.6 26.8 0.421 — 1 icc39 132351 11 Aug 53 985 965 990 80 0.84 16.3 0.005 67. 26.8 0.437 — 1 icc40 132353 12 Aug 53 945 930 950 75 0.085 14.6 0.06 112. 26.8 0.392 — 1 icc41 132354 13 Aug 53 860 840 865 80 0.150 16.3 0.009 234. 26.8 0.437 — 1 icc42 132355 14 Aug 53 870 815 857 74 1.70 63.6 o.0268 638. 26.8 1.70 — 1 icc43 132356 17 Aug 53 935 900 932 70 0.309 11.8 o.0262 509. 26.7 0.315 — 1 icc- Ohio Ciem. 028087 - 44 132357 18 Aug 53 875 870 88 72 0.428 18.5 0.0231 551. 26.7 0.494 -1 icc -- 1 45 132359 20 Aug 53 1015 970 1002 75 '.0.005 17.4 0a.0002 7.3 26.7 o.465 — 1 icc- Matheson 5772 - 46 1323460 21 Aug 53 920 865 907 74 2.6 66.8 0.40 802. 26.7 1.78 — 1 icc47 132361 24 Aug 53 825 785 820 79 -- 15.7 -- 0.0 26.5 0.416 I- icc48 132362 25 Aug 53 975 946 975 79 0.029 14.o 0.0021 28.0 26.5 0.372 — 1 iccdarhide In small to,mdialms 49 132363 26 Aug 53 1200 1000 1115 90 21.5 116.5 0.185 2380. 26.5 3.09 — 1 icc- and JK 370331 slightly off whitadrtl Carbon coherent 50 132366 5 Sept 53 1010 324 980 77 4.9 90.5 0.0.41 730. 2.5 2.4 1c Mathesons 77 Powdery, in asall. otlms 51 132369 9 Sept 53 1005 325 980 75 -- 17.0 -- 0.0 26.5 0.451 -1 icc whit Carbide 52 132370 10 Sept 53 1000 970 1000 70 1.5 21.3 0.0705 790. 26.5 0.565 — 1 icc- and 2K 370331 - Carbon 53 132372 11 Sept 53 513 475 510 60 0.16 34.5 0.0464 298. 26.5 0.912 Acetaldehyde 1 ic - Matheson 5772 - 54 132373 14 Sept 53 1500 960 1245...70 51.3 262.7 0.195 2200. 26.5 6.97 -1 icc Hafies toall cutoheet ht 55 132375 25 Sept 53 1570 1080 1315 as65 43.6 71.8 0.607 1915. avg 89.6 6.43 — 10 icc on - fines allb cutoheeti~ t 17.0 29.5 0.30 - 1 icc.. nd 56 132576 30 Sept 53 1780 1110 1450 79 12.5 23.7 o.18 avg 745. avg 89.5 2.12 — 10 icc 0n U.. d 1065 White, soft powdevr ml up 28.8 58.0 1.67 -- 10 icc 6- out Chem.

the scattered electrons. These scattered electrons would be quite effective in producing chemical reaction because nearly all their energy would be imparted to the chemical system. Consequently it can be seen that more ionization probably occurred than was taken into account by the calculations, in which the effect of the bomb wall was neglected. The effect of this error is that the G values given are too high. On the other hand, as a calculation device the primary beam was assumed to undergo no appreciable absorption within the ethylene in the bomb. Rather the beam was assumed to maintain within the bomb a value which would be attained on the axis if the bomb were full of air. It was recognized of course that absorption within the ethylene was assumed to be causing the reaction. If account were taken of absorption of primaries within the ethylene, then somewhat greater credit for initiating reaction would have to be given to each primary photon, and this would increase the G values given. Thus, neglect of nonequilibrium secondaries and neglect of the absorption gradient of primary gamma intensity within the ethylene compensate each other to some extent. The importance of accounting for the above errors in dosimetry is recognized. However, the complexity of the measurement problems would seem to indicate the desirability of pursuing this work further in future studies. Therefore, the values given for G in Fig. 4 should be regarded as relative rather than absolute, since all determinations were made in the same equipment and using similar procedures. No consistent effect of pressure on the G value was noted. Elevated temperatures were investigated only briefly., but preliminary results indicated that increased rates of polymerization would-result in irradiated systems at temperatures of 200-14000F as compared with those obtained at room temperature. 4i. Evaluation of the Polyethylene Product. a. General: The polyethylene obtained as a result of gamma irradiation of ethylene was subjected to a brief program of evaluation. The properties considered most basic to an understanding of the material were investigated. Most experimental work was concerned with determinations of solution viscosity, melt viscosity, density, and tensile strength. Melting points of some samples were also determined. Molecular weights were estimated from the determinations of viscosities of solutions and of melts. Crystallinity was estimated from determinations of density. The other measurements were made by conventional means. These measurements and derived quantities probably need no further ex

molecules of polymer are arranged parallel to each other. An arrangement of parallel molecules results in a repetitive structural pattern such as that found among the molecules of a crystal. A random orientation of molecules similar to a pile of jackstraws might be expected to be less dense than a parallel arrangement such as that just described, and it has been found that percentage crystallinity may be correlated with the density of polyethylene (see KirkOthmer).7 From the data obtained in experiments on the polymer, molecular weight and crystallinity are presented as functions of dose and all the properties of the polyethylene are presented as functions of the radiation yield of the polymerization reaction because of the following considerations. The radiation yield of the polymerization of ethylene may be expressed as the G value. Lind9 has shown that in many gaseous systems, approximately one molecule reacts per ion pair formed in the system. In the irradiation of ethylene a variable number of molecules, usually much greater than one, react for each ion pair formed. The polymerization of ethylene is therefore evidently a chain reaction. For this calculation it is assumed that one chain is initiated for every ion pair formed, that all chains are of equal length, and further that the formation of each ion pair requires 32.5 electron-volts of energy, a value approximately correct for gases at one atmosphere. The densities of ethylene under the conditions of reaction were greater than at one atmosphere, however, and therefore the energy required per ion pair may be quite different from the value given. The G value may therefore be divided by three to give the approximate number of molecules reacted for each ion pair formed, and this result may then be multiplied by the molecular weight of the monomer in order to arrive at the molecular weight of the polymer. Consequently, the G value is directly proportional to the molecular weight which would be expected of the polymer if the above assumptions held. Furthermore., the properties of a polymer are frequently found to be functions of its molecular weight. It therefore seems advantageous to consider the properties of the polyethylene as functions of the G value. The results of most determinations could be correlated against the G value., or radiation yield, somewhat better than they could against dose, although the G value has been shown to be a function of dose. In Fig. 4, for example., the G value was about 0.1 to 1.0 until about 0.5 megarep had been received. The G value then increased rapidly with increasing dose until it reached a nearly constant value of about 2000 molecules per 100 electron-volts for doses of about 5 to 7 megarep. b. Experimental: All the samples of polyethylene were white. Some

Portions of each of the samples of polyethylene which were obtained in yields of 4 grams or more were molded into sheets as an operation preliminary to further examinations. A two-compartment mold was used, one compartment at a time. Samples were placed between aluminum foil in the mold, preheated to 3000F, pressed at 1000 psi, and cooled to about 1250F under pressure. The resulting sheets were 2.5 by 4 by 0.025 inch. All such sheets proved to have the characteristic milky, translucent appearance of polyethylene. The sheets molded from the powders were brittle, while those from the tough reaction products were also tough. Molecular weights were estimated from viscosities of solutions, measured as follows: Solutions of some samples were prepared in concentrations of 0.01 percent and of 0.125 percent by weight in tetralin. Viscosities of these solutions and of the tetralin were measured in modified Ostwald pipettes at 212F. Specific viscosities were calculated and divided by the respective concentrations. The resulting ratios were plotted as a function of the concentration of polymer, and the plots were extrapolated to zero concentration to give intrinsic viscosity. Intrinsic viscosity was assumed to be directly proportional to molecular weight. The concentration was computed in units of gram moles of monomeric ethylene per liter of solution. The constant of proportionality was computed by the author to cause the observed value for the molecular weight of Bakelite DYNH to agree with the value of 20,000 for the weight-average molecular weight given by Dienes and Klemm.4 The value of the constant was computed in this way to be 0.42 x 104 liter per gram. (See also the work of Tani'3 on intrinsic viscosities of polyethylene in tetralin.) The method of Dienes and Klemm4 was used to estimate molecular weights from melt viscosities. Viscosities were measured in a parallel-plate plastometer with an attached dial gauge reading to 0.01 millimeter. The entire assembly was placed in an oven. Temperatures of 2480F and 2660F were used. The samples were placed between sheets of aluminum foil about 1-1/2 mils thick. The thickness of the sheets of foil was measured in the plastometer before each determination'. Crystallinity was estimated by correlation with density (see KirkOthmer).7 Densities were determined by the use of Archimedes' principle. Weighings were made directly in water, after the sample had first been degassed by use of reduced pressure while it was 'immersed in water. Tensile properties of the polyethylenes were examined by the following procedure. Specimens for testing were cut from the molded sheets by means of a die. The resulting specimens were 0.079 by 0.025 inch in the smallest cross section. The narrowest section was 1-1/2 inches long. Tension was applied

Melting points were determined on a melting-point bar of the type used by Dennis3 (see Fig. 8). c. Results. Results of evaluation of the properties of the radiationpolymerized polyethylene are summarized in Table 7. The molecular weight is plotted as a function of radiation yield in Fig. 7, and the molecular weights and crystallinities are plotted as functions of dose in Fig. 9. The significance of the determination of molecular weights by means of solution viscosity is not clear. The values obtained were assumed to be weight-average molecular weights, based on the weight-average molecular weight of Bakelite DYNH of 20,000 (see Table 7). However, differences in crystallinity and cross-linking, mentioned above, may invalidate the comparison of the thermally polymerized sample with the radiation-polymerized samples. Determinations of molecular weight by melt viscosity may be subject to similar criticism. As shown in Figs. 7 and 9, the molecular weights determined by solution viscosity do not agree well with those determined by melt viscosity. Neither do the molecular weights from solution viscosity appear to display any regular variations with dose or with G value, in contrast to the regular behavior of molecular weights from melt viscosities. The reasons for these discrepancies are not clear. Values of crystallinity are plotted as a function of radiation yield in Fig. 10. The crystallinities varied from about 77 percent for samples of low radiation yield to ab out 71 percent for samples of high radiation yield. All these samples were of considerably higher crystallinity than was the Bakelite DYNH, which had a crystallinity of about 61 percent. It is possible that the radiation-polymerized samples were of higher crystallinity-than the'thermally polymerized sample of DYNH because the temperature of polymerization was lower for the radiation-polymerized samples. The samples of low radiation yield would be expected to be more highly crystalline than those of high radiation yield. Since radiation yield has been shown to increase with dose (Fig. 4.), crosslinking and branching would probably also increase with dose, and increases in ei'ther cross-linking or branching would cause decreased crystallinity. Tensile properties are reported in terms of stress as a function of strain in Fig. 11. The irradiated samples all have properties similar to those of a brittle material., and the samples subjected to higher doses have higher tensile strengths and are more ductile than those subjected to lower doses of radiation. Such behavior would be likely if the irradiation increased crosslinking and branching. The Bakelite DYNH shows the characteristic elongation of several hundred percent before rupture (see Kirk-Othmer7 page 942). Ultimate tens -_-Ile- stes asa untonofrditinyildi plotted i Fig. 10. The ult- I4

TABLE 7 PROPERTIES OF POLYETHYLENE PRODUCED Page Dose, Radiation Melting Density Ultimate Elongation, Crystallinity Molecular Weight Number Megarep Yield, A Point, F g/Cm3 Tensile, Percent Percent by Melt by Solution Lower/Upper psi at rupture Viscosity Viscosity 132250 1.55 95 219/226 132268 0.61 37 216/225 132269 o.58 71 205/217 152276 o.60 63 207/214 132281 o.54 52 196/205 132297 1.91 750 241/244 0.951 450 4 77 26,300 insol. ro 132562 0.37 28 234/255 132363 3.09 2400 248/689 0.941 2200 42 71 34,400 insol. 132366 2.40 730 210/248 0.951 770 2 77 28,100 4200 132369 0.45 0.1 132570 0.57 790 203/252 8800 132372 0.91 298 199/207 insol. 132373 6.97 2200 234/720 0.943 2100 29 72 40,500 insol. 132375 6.43 1900 241/610 0.941 2300 79 71 37,300 insol. 132376 4.29 745 241/244 0.951 630 5 77 11,900 3700 Bakelite 20,000 DYNkelit 0.921 1500 550 61 21,80 DYNH.

0 w z w FROM MELT VISCOSITY 4 0~~~~~ 0. IL. ro 0 i;:i 0 FROM SOLUTION VISCOSITY 0 500 1000 1500 ~~~2000 2000 30 RADATON IEDGRAM MOLES REACTED RADIATIONYIELDC TON JMEGAREP Fig. 7. Molecular weight as Function of Radiation Yield of Polyethylene.

or-I b-O co rr, ~....2...7

0~ ~~~~~~~~~~6 CRYSTALLINITY BY DENSITY --- SO 50~ 50,9000 5 5z _1 49040 IMOLECULAR WEIGHT0w C,J BY MELT VISCOSITY z 30,OOC ____30 ca 4~~~~~~~~~~~~~~~ C)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C ma 20_00_ 20 0~~~~~~~~~~~~ 10,000__ _ __ _ _ __ _ _ __ __ _ _ 10 MOLECULAR WEIGHT J-BL OLUTION VISCOSITY - 0 ~~~~~~~~~~~~~~~0 0 2 45 6 7 DOSE., MEGA REP

PERCENTAGE CRYSTALLI NITY 71 72 73 74 75 76 2400 _,~ 0 2200 2000 RCENTAGE _/ tYSTALLI NITY\.TIMATE TENSILE/ a.RENGTH w - '. w U)X 1600 o _~PERCENTAGE A-CRYSTALLINITY w _ULTIMATE TENSILE 0STRENGTH 41400 w >' 1200 z 0 1000 800 6000 400 800 1200 1600 2000 2400 2800 3200 ULTIMATE TENSILE STRENGTH, PSI

2500 1 { x 132 375 132 363 llao -132 373 2000,15001 I I 0. en |BAKELITE DYNH 1000 F I i-. $ 132 366-ULTIMATE Io 132 376-ULTIMATE 5oo0 132 297-ULTIMATE ____ 00L 0 I2 3 4 56 STRAIN

with dose. A set of structural properties such as those just described for the radiation-polymerized polyethylene might be desirable for certain applications, but the properties differ from those of most polyethylene currently marketed. Melting points are plotted as a function of radiation yield in Fig. 12. Curves are given for both the upper and the lower ends of the melting-point range. The results show that there is a small increase in the temperature of initial softening with increase in radiation yield, and that there is a large increase in the temperature of complete melting. The higher melting points indicate higher degrees of cross-linking as a result of the higher doses of radiation, and are thus in conformity with the-results of the other determinations. C. CHLORINATION OF AROMATIC COMPOUNDS The reactions of chlorine with various aromatic hydrocarbons have been reported by this laboratory previously. During this period covered by this report, the study of toluene reactions has been continued and monochlorobenzene has also been reacted with chlorine. 1. Description of Apparatus. In order to utilize the -kc cobalt60 gamma source for reactions employing large volumes of toluene, new equipment was designed and constructed. In this equipment, up to 1. liters of toluene may be reacted with chlorine gas. The chlorine was fed from a cylinder located in the second floor chemical laboratory of the Fission Products Laboratory. Gases returning from the reacting system passed through bubbler bottles' or solutions for absorption and were vented to the outdoors. This part of the system was the same as that illustrated in Progress Report 4i (1, pages 26 and 29) in the system used with the l-kc source. The apparatus located in the source room consisted of a water-cooled reactor section,. a storage section, and inlet and outlet tubes for gases leading to the chemistry laboratory. The apparatus is shown in Fig. 15, and details of individual parts of this apparatus appear in Fig. 14. Referring to Fig. 15, the gases entered the apparatus through tube (1)., and passed through the jet injector assembly (2). The gas then bubbled through the outer jacket of the reactor section (3), passed through the connecting tube (4) and bulb (5), and emerged from tube (6). In passing through the jet injector (see Figs. 16 and 17 for detail), and in rising through the vertical reactor section, the gases caused the reacting liquids to circulate through the tubes of the system. After being carried across the upper connecting tube (4), the liquid returned to the

80,000 6,000 4,000____ MELTING BEGAN __ _ _ _ _ _ _ _ _ __ _ _ _ _ wo 2,OO ___ __ * N C c. wo uid 0 7 1_6000 x Soo 4 400 W. ~a: 400 I w.~ 200 z 0 ti 4 4 100__ _ _ __ _ _ 60 40 20 100 200 300 400 500 600 700 S00 MELTING POINT, OF

Fig. 13. Apparatus for Reaction Fig. 14. Individual Parts of the of Liquids with Chlorine Gas Apparatus for Reaction of Liquids in the 10 kc Cobalt-60 Gamma with Chlorine Gas. Source.

GAS APPARATUS TO CHLORINE CYLINDER6 3~~~~ LES AND MERCURY SEALED '-F TRAPS 5 as/~~~ THERMOCOUPLE

Fig. 16. Photograph Showing Detail of the Jet Injector for Gases use on the Apparatus for Liquid-Gas Injections in the 10 kc-Source. FG ~)~~5

reacting mixture, since the apparatus was initially filled to the level of the upper connecting tube (4). It also served to separate the exit gases from spray and foam at the surface of the liquid. Vent lines were included at the inlet (9), and exit (10) of the apparatus. These lines were connected to dry bottles and thence to mercurysealed traps so that safety was assured in case of rapid reaction or buildup of gas pressure due to plugged lines. For the recording and controlling of temperature, thermocouples were fastened to the outside of the glass at points (11) and (12). The temperatures of the thermocouples were recorded on chart recorders located in the chemical laboratory. The thermocouple on the upper connecting tube (4) also activated a temperature controller, so that the temperature of the coolant circulating through the inner jacket of the reactor section (3) was controlled by the temperature of the liquid as it left the reacting section. The coolant was recirculated through a modification of the cooling system described in Progress Report 4 (1, pages 28 and 30). Figure 18 is a schematic drawing of this apparatus. By means of a pump (1) the coolant was recirculated through the cooling jacket of the reaction apparatus, and returned to the mixing can (3). A knife heater (6) in this mixing can was activated by the recorder-controller to which the thermocouple was attached and was wired to a recirculating pump (2) through a double-throw switch. When this pump was activated by the controller, water was pumped into an upper cooling can (4), from which it overflowed into the mixing can (15) again. In the cooling can (4) was a smaller can (5) containing a mixture of dry ice, carbon tetrachloride., and chloroform which cooled the recirculation liquid as it was pumped around it. When only moderate cooling was required, water was used as the recirculating coolant, in which case a layer of ice was kept frozeii around the inner part of the cooling can arrangement. For low-temperature runs, methanol can be used as the recirculating coolant. When ready for use, the glass reactor equipment was assembled on a portable support frame, and placed in a portable exhaust hood adjacent to the l0-kc gamma source (Figs. 19 and 20). In use., the reaction section could be centered as close as 2 inches from the gamma source. A plastic shield was placed over the open-side of the hood to prevent the corrosive liquids from splashing out into the source room in case the glass reaction equipment should break. After the apparatus had been set in place in the source room, it was connected to the gas control rack in the chemical laboratory by means of glass

LEADS AND NORTHRUP CONTROLLER r~~ --- —y___ii r-okI II fI I THERMOCOUPLE i X sLi WELL INTERMITTENT PUMP,COOLINGT CONTINUOUS PUMP Fig. 18. Schematic Drawing of Apparatus for Teniperature Control of Reactions in the 1 and 10 kc Cobalt-60 Gamma Source. g. 19. Apparatus for Reaction of Liquids Fig. 20. Portable Hood and Glass~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... Gamm Sore II: t

(page 66) access to the source room from the chemical laboratory was gained through a small chute with a right angle through the shielding. In this chute were the leads for thermocouples and electrical cable. The problem of making a gas-tight connection of glass tubing at the right angle of the chute was solved in the following manner. A wooden support board was constructed which was of the same width and somewhat longer than the horizontal portion of the access chute (see Fig. 21). The glass tubes, which were fitted with ball-andsocket joints, were affixed to this board by means of small spring clamps made from spring clothespin clips. One end of each tube had a short right-angle bend in it. These were rotated to lie flat against the supporting board so that the entire assembly could be pushed into the access tube. Wooden strips along the sides and one one end of the board acted as runners and allowed the glass tubing to hang from the bottom of the board, while cables and aluminum tubing lay on top of the support board in the chute. Once in place, the glass tubing was rotated so that the right-angle bends lead downward placing the ground glass ball in position to receive the next (vertical) section of tubing with its ground-glass socket. Since clips are necessary to assure a gas-tight seal at the ball-and-socket ground joint, a special tool was devised to affix a clip to the joint located at the right angle of the chute. This tool (Fig. 22) was inserted through the vertical section of the chute from below in the source room, and could be manipulated to attach or remove the ball-joint clips. A diagram of the entire setup is shown in Fig. 23. 2. Experimental Procedure and Results. A number of runs using toluene and chlorine were made in the apparatus described above. For these runs., about 1.5 liters of toluene were placed in the reactor. Nitrogen was run through the system for 10 to 20 minutes before each run to displace all air and the solutions were saturated with chlorine before the source was raised at the start of the reaction, During the entire period of chlorination to minimize photochlorination., only red light-was used whenever it was necessary to observe the equipment. It should be noted that when the source is raised in the dark., a faint blue glow can be seen in the toluene solution, probably caused by Cerenkov radiation. At periodic intervals, the source was lowered to allow personnel to add more dry ice to the temperature control system. When desirable, however, it is possible to place part of the cooling system at a remote position so that uninterrupted runs can be carried out. In most runs, the storage-bulb secti~on of the glass apparatus (7, Fig. 15) was shielded by lead bricks from the direct radiation of the source. In one run, however, no shielding was employed and the reaction became so violent after about 15 seconds of irradiation that the chlorine and hydrogen chloride escaping from the hot toluene solution blew open a glass joint. In another run, the shielding bricks were removed after the run had continued for

Fig. 21. Wooden Support Board for Glass Tubing F ig. 22. Special Tool f or Attaching and(eovn to be put through the Chute Providing Access Clips from Joints Located at Right Angl f h to the 10 kc Source Room. In Use, the Side Access Chute to 10 kc Gamma Source. Shown is Placed Down, when the Assembly is Put into the Chute.

GAS CONTROL RACK TO EXHAUST,,ABOVE ROOF TUBING SUPPORT BOARD IN ACCESS CGHUTE SEODFLOOR " '-..'~ /~~~~~~~~~~~~~~~~I DUC TOBOE.... "' "~ 'SPPOR —' —'i" i '"- ':.,, GLSSTUBING,~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~10 A Al-, IOK'C.-, E TUBING.:, '.',~,.~~~~~~~~~~~ 4 ~a" ~.. '~.;,'.. ~ ~.'~~~~ Fig~~~~~~~~~. 23..~ Diara of Coplt Se-pfr ecin o iud with~~~~, ChoieGs nte1 ore

The reaction mixture from one of these runs was distilled under vacuum, and high-boiling fractions were vacuum-distilled again. Table 8 summarizes the information on these fractions. Overlapping of temperature ranges was probably due to variations in pressure as receivers were changed. The theoretical chlorine content for a hexachloro addition product of toluene is 69.78% C1. TABLE 8 HIGH BOILING FRACTIONS OBTAINED FROM REDISTILLATION OF PRODUCTS OF REACTION BETWEEN TOLUENE AND CHLORINE Fraction Volume, Temperature Pressure, Number ml Range, 0C Microns Hg C1 VII-A 6.6 85-102 100-200 VII-B 29.4 98-111 100-200 68.6 VII-VIII 4.0 110-115 100-200 VIII-A 18.4 111-115 100-200 69.6 VIII-B 18.0 112-120 100-200 70.5 VIII-IX 3.1 120-125 100-200 IX-A 10.0 125-155 100-200 70.7 The individual fractions were syrupy, crystal-clear liquids. After several weeks of standing, part of which was under refrigeration, some of the liquids became cloudy and crystals appeared to be precipitating out slowly. This production of liquids is perhaps not unusual, for if the product is 1,2,,4,5,6-hexachloro-l-methylcyclohexane, a great number of stereoisomers may be expected from total addition of chlorine to toluene. It is possible to write twenty such different stereoisomers, and it seems probable that at least half of these should be important constituents of the mixture of reaction products. Thus it is not surprising that the combined effects of many mixed isomers and high viscosity prevented appreciable crystallization of the products. In an effort to study the behavior of the reaction during its course, a run was made in which the chlorine-gas cylinder was weighed continually, and in which samples of the exit gas were taken for analysis. Figure 24 summarizes the resulting data. The analytical methods were not precise 'in this run,. so data on the material balance of the chlorine can be regarded as only roughly

REACTION OF TOLUENE AND CHLORINE -10 KC SOURCE 141540 TIME SOURCE REACTION CHLORINE HCL CHLORINE CHLORINE UP TEMPERATURE INPUT OUT EXCESS IN BY (HATCHED OUT ADDITION AERA) MILLIMOLES/MIN. MILLIMOLES/MIN. MILLIMOLES/MIN. MILLIMOLES/MIN. I0~ 203 0~ 1115 1125 1156 1158 90 1203 75 9 1205 1216 90 1219 9016 4 70 1222 1238 103 43 1242 1330 69 2 5 1339 76 1402 67 1420 80 1431 102 36 0.5 36 '433 21 1542 73 502 1545 1603 140 1619 31 1657 1707 -1729 31 -1732 1738 1742 49 1744 9 1750 9

The behavior of the temperature curve is a striking example of the dependence of the reaction on the radiation field. Because of the nature of the temperature control system a short time for readjustment was always reuired after a sudden change in heat production of the system. Thus, it can be seen that each time the source was raised into position for irradiation the temperature displayed a sharp increase, while each time the source was lowered the temperature dropped off. Furthermore, this effect is more pronounced at the start of the reaction than at a later time, when the rate had apparently decreased. A summary of information on the runs made with toluene in the 10-kc source is presented in Table 9. The estimated average gamma flux rate was based on flux measurements in air at the centers of each of the two components (reaction and storage) of the glass apparatus. The overall yield was based on moles of supposed hexachloro addition products (high-boiling fractions with chlorine content in the 69-71% range) obtained from a given number of moles of toluene originally placed in the apparatus. Calculation of radiation yields in moles per liter per kilorep and in terms of "G" (molecules reacted per 100 electron-volts of gamma energy) was made using the density which was determined from a reactor solution used in one of the runs saturated with chlorine gas. Inspection of the G values (Table 9) for the addition reaction in these compounds reveals that they are all in the neighborhood of 17,000. The high values for G found in all these cases are somewhat unusual in radiation chemistry at the present time, and indicate a long reaction chain. Assuming about three primary ionization events per 100 el ectron-volts,11 the chain lengths must be of the order of 5 x 103 or greater. This figure also assumes that-each priLmary ionization event produces a chain of molecules which react completely and solely to give the hexachloro addition product. The fact that appreciable substitution takes place indicates that the chain length for the addition process itself might be much larger. It is very interesting that the products obtained by gamma activation differ markedly from those usually obtained through actinic radiation activation (mostly sidechain substitution in this case). In view of the long reaction chain lengths the product isolated contained negligible amounts of those molecules originally activated in the pri~mary process. Thus the reaction chains in the gamma-activated process would seem to be propagated by a different mechanism than those of the ultravioletactivated process. This difference is being investigated further. From one of the reaction runs of toluene and chlorine., 120 grains of those fractions which distilled over at 90-1560C (90-240 microns Hg pressure) were sent to be tested for entymological poison activity, since the supposed

TABLE 9 SUMMARY OF THREE REACTION RUNS OF TOLUENE AND CHLORINE IN THE 10 KC 60 GAMMA SOURCE Estimated Weight of Volume ~TotalAergAvae Volume Control Total Average Total Hexachloro Average Average Run of Temp.., up Gamma Flux Dose Addition Moles of Overall Moles per- Value of Run of ~~~~Gamma Flux Dose Addition Prdc Yil (it) Gfo Number Sample Temp., Time, Product ml C min ae, Rece ivred, rou Produced (kilorep) Addition kilorep/ kioepi hr air grams -ra 141500 1440 20 71 24 28 185 0.61 4.5 0.014 14,000 141534 1450 20 121 22 44 371 1.2 8.8 0.017 18,000 141540 1500 20 226 22 82 778 2.5 18 0.018 19,000

After consideration of the rapid addition of chlorine to benzene and to toluene, and the nonreactivity of benyl, benzal, and benzo trichlorides under the same conditions2 (page 29), a run was carried out in which chlorobenzene was treated with chlorine gas in the l-kc cobalt-60 gamma source. In this case, the reaction proceeded at a more rapid rate than that for toluene. Table 10l summarizes the information obtained in this run. The yield in Table 10 is based on the weight of crude product obtained by steam-distilling off the unreacted chlorobenzene and drying the solid residue; this procedure may yield values that are high. A sample of recrystallized material gave a chloride analysis of 76.0%, as compared to the calculated value of 76.35 for heptachlorocyclohexane, which would result from addition of six chlorine atoms to chlorobenzene. A run has been made in darkness with other conditions the same as those used during the gamma-induced reaction. Analysis of reaction products gave a value for chlorine content which was so close to the calculated value for unreacted chlorobenzene that virtually no reaction was indicated.. SUGGESTIONS FOR FUTURE WORK It would be of interest to investigate further the polymerization of ethylene under gamma radiation with the objective of determining the cause for the induction period observed for the polymerization. It seems possible that small concentrations of impurities are responsible for the induction period. However., it is also possible that such an induction period might be characteristic of the gamma-induced polymerization and independent of chemical parameters. If the ethylene could be removed shortly after polymerizing, then a continuous process for the polymerization might be developed. Such a procedure would permit closer control of those properties of the polymer which are dependent on total dose of radiation, As mentioned earlier, under "Discussion of the Polymerization of Ethylene, elevated temperatures in conjunction with irradiation caused greater rates of polymerization of ethylene than did irradiation alone. Investigation of the influence of both elevated temperatures and irradiation of the polymerization of ethylene is certainly in order, Some suggestions for such a program were advanced in Progress Report 3, 2 based on some of the foregoing studies of radiation chemistry and on the work of Kennard.6 The electrical properties of the polyethylene made by gamma irradiation might well be studied.

TABLE 10 REACTION OF CHLOROBENZENE WITH CHLORINE UNDER GAMMA IRRADIATION Weight of Total Gamma Total g Sample Control Reaction Fl~ Hexachloro Moles of Reaction Flux Rate dose, Addition Product Ov er)ll Rate dose Overall ~~~Moles per G for Volume, Temp., Time, in air, kilorep dt Producted Yield, it ) Addition ml ~C min kilorep/hr in air Grams (kilorep) in 152004Grams 100 '20 30 52 26 38 0.12 12 0.042 43,000

information concerning the behavior of this reaction. The physical properties of this polymer molded into massive form should be investigated. The work on chlorination of aromatic compounds should be pursued with the object of attempting to find and correlate those reactions which differ from reactions promoted by chemical catalysis or ultra-violet radiation. This work would involve further study of reaction rates for those reactions already partially investigated, and the study of the reaction of other aromatic compounds with chlorine. Such studies might provide insight into the manner in which gamma irradiation promotes reactions which have large G values, and establish unique uses for the gamma radiation of fission products. E. SUMMARY 1. Polymerization of Ethylene. (a) Ethylene has been polymerized by exposure to gamma irradiation from cobalt-60. The rates of reaction were sufficiently large that further work on this reaction appears to be promising. (b) Polyethylene formed by gamma irradiation has been subjected to a preliminary evaluation. The polymer was found to be denser, less ductile, and of a higher ultimate strength than Bakelite DYNH polyethylene. Molecular weights of the radiation-polymerized materials increased with radiation dose to a value of about 40,000 when estimated from melt viscosities. Most samples were insoluble in tetralin, but estimates of molecular weights from solution viscosities were not conclusive. Crystallinities estimated from densities varied from 71 to 77 percent. 2. Chlorination of Aromatic Compounds. (a) Toluene and chlorine were reacted in the presence of gamma irradiation using up to 1.5 lof toluene in one run. Fractional vacuum distillation of the reaction products yielded high boiling materials whose chlorine content is close to that for a hexachloro addition product of toluene and chlorine in addition to products of substitution reactions. For some runs of varying total gamma dosage., the G value for addition was in the neighborhood of 17.,000. Preliminary rate data for one run have been taken. (b) Chlorobenzene and chlorine were reacted under gamma irradiation. It was found that the addition reaction proceeded with a G value of about 45,,000. F. *REFER~ENCES 1. Anderson, L. C., Martin., J. J.., et al.., "Utilization of the Gross Fission

2. Anderson, L. C., Martin, J. J., et al., "Utilization of the Gross Fission ProdctsProgress Report 5 (C00oo-196), Univ. of Mich., Eng. Res. Inst. Proj. M94, September, 1953.. Dennis, L. M, and Shelton, R. S., J. Am. Chem. Soc. 52, 3128-32 (1930).. Dienes, G. J., and Klemm, H. F., "Theory and Application of the Parallel,Plate Plastometer" J. Appl. Phys. 17, 458 (1946). 5. Gardlner, H. A., and Sward, G. G., Physical and Chemical Examination of Paints., Varnishes, Lacquers, and Colors, Gardner, New York. 10th ed., 1946. 6. Kennardl, E. H., Kinetic Theory of Gases, McGraw-Hill, New York, 1st ed., 1958. 7. Kirk-Othmer, Encyclopedia of Chemical Technology, 10, 938-57 (1953). 8. Lewis, J. G., Nehemias, J. V., Harmer, D. E., and Martin, J. J., Nucleonics 12, No. 1, 4-4 (1954). 9. Lind, S. C. The Chemical Effects of Alpha Particles and Electrons, The Chemical Catalog Co., New York, 2nd ed., 1928. 10. Matthew,, F. E., and Elder., H. M., Brit. Pat. 11,655, 1914. 11. Pollard, E. C.,9 and Davidson, W. L., Applied Nuclear Physics, John Wiley and Sons, Inc., New York 2nd ed., 1951, p. 25. 12. Snow, BR. D..,and Frey, F.* E., Ind. Eng. Chem. 50, 176 (1958). 15. Tani., H., Chem. High Polymers (Japan) 4., 151-7 (1947). 14. Weiss, Jerome, Nucleonics 10., 28-51 (1952). 15. York, R., and White., E. F.., Trans,. Am. Inst. Chem. Eng. 40., 227 (1944).

Progress Report No. 7 UTILIZATION OF THE GROSS FISSION PRODUCTS The Effect of Gamma Radiation on Chemical Reactions J. J. Mart in C. E. Eckfield L. C. Anderson J. P. Holmes B. G. Bray R. R. Nissle D. E. Harmuer W.,M. Sergy R. A. Carstens H. H. Yang

TABLE OF CONTENTS Progress Report No. 7 Page SBPROJECT m943-4, THE EFFECT OF RADIATION ON CHEMICAL REACTIONS 9 A. Introduction 9 B. Polymerization of Ethylene 10 1. Equipment Used 10 2. Visual Inspection of the Products 10. Physical Properties of the Solid Polyethylene Product 11 4. Other Work on Polymerization of Ethylene 12 C. Polymerization of Ethylene and Sulfur Dioxide 1 D. Chlorination of Aromatic Compounds 14 1. Investigation of the Structure of the Primary Product Obtained from the Gamma Ray Activated Reaction of Toluene with Chlorine 14 (a) Dehydrohalogenation 14 (b) Oxidation of the Dehydrohalogenated Material15 2. Separation of Products of Chlorination of Toluene 20 (a) Attempted Crystallizations 20 (b) Analysis of Solids from the Reaction Mixtures 22 5. Investigation of Reaction Rates between Toluene and Chlorine 22 (a) Apparatus 22 (b) AnalyticalMethods and Preliminary Results 25 (c) Equipment Changes 26 E.. Design of a Plant for the Chlorination of Benzene under Gamma Radiation 26 1. Laboratory Chlorination of Benzene in the Presence of Gamma Radiation 28 2. Rate of Reaction for the Addition Product of Chlorine and Benzene 29 5. Process Design Calculations for a Continuous Flow Reactor 52 4. Description of the Proposed Plant Process 3 5. Calculation of Gamma Radiation Source Strength and Shielding 37 6. Estimation of Cost for the Proposed Equipment and Radiation Chamber 59 ConsantCouc Stre ngtha- 42L! - --

TABLES OF CONTENTS (Concluded) Page F. Work Contemplated for the Near Future 51 G. References 51 LIST OF ILLUSTRATIONS Figure Page 8. Regions of Investigation of Polymerization of Ethylene under the Influence of Gamma Radiation 1 9. Melting Point Ranges for Fractions of Isomeric Tri chlorobenzoic Acids Separated by Ion-exchange Chromatography 18 10. Flowsheet for Equipment Used to Determine Reaction Rates between Toluene and Chlorine 24 11. Gas-c-ontrol Equipment Rack Used for Re~actions between Toluene and Chlorine 25 12. Drawing of Glass Reactor Used in Determining Reaction Rates between Toluene and Chlorine 27 15. Glass Reactor Used for Determining Reaction Rates between Toluene and Chlorine 28 14. Glass Reactor, Disassembled 28 15. Reaction Velocity Constant vs Reciprocal Temperature for the Chlorination of Benzene 51 16. Flowsheet of Proposed Plant Process for the Addition of Chlorine to Benzene under the Influence of Gamma Radiation 355 17. Plan View of Proposed Reaction Cave Showing Reactors in

TABLES OF CONTENTS (Concluded) Page F. Work Contemplated for the Near Future 51 G. References 51 LIST OF ILLUSTRATIONS Figure Page 8. Regions of Investigation of Polymerization of Ethylene under the Influence of Gamma Radiation 1 9. Melting Point Ranges for Fractions of Isomeric Tri chlorobenzoic Acids Separated by Ion-exchange Chromatography 18 10. Flowsheet for Equipment Used to Determine Reaction Rates between Toluene and Chlorine 2 11. Gas-c-ontrol Equipment Rack Used for Re~actions between Toluene and Chlorine 25 12. Drawing of Glass Reactor Used in Determining Reaction Rates between Toluene and Chlorine 27 15. Glass Reactor Used for Determining Reaction Rates between Toluene and Chlorine 28 14. Glass Reactor, Disassembled 28 15. Reaction Velocity Constant vs Reciprocal Temperature for the Chlorination of Benzene 31 16. Flowsheet of Proposed Plant Process for the Addition of Chlorine to Benzene under the Influence of Gamma Radiation 355 17. Plan View of Proposed Reaction Cave Showing Reactors in

PART II. SUBFROJECT M943-4, THE EFFECT OF RADIATION ON CHEMICAL REACTIONS Personnel: Subproject Supervisors: Joseph J. Martin, Associate Professor of Chemical and Metallurgical Engineering; and Leigh C. Anderson Professor and Chairman of the Department of Chemistry Senior Research Assistants: B. G. Bray, David E. Harmer Assistants in Research: R. A. Carstens, C. E. Eckfield, J. P. Holmes, R. R. Nissle, W. M. Sergy, H. H. Yang A. INTRODUCTION Since the preparation of Progress Report 6, work has been carried out on three major fronts: (1) The polymerization of ethylene under gamma radiation has been studied over a wide range of temperature. At the same time the copolymerization of ethylene and sulfur dioxide has received a small amount of additional attention.. (2) The rate of reaction of chlorine with toluene under gamma radiation has been studied in more detail than reported by this laboratory earlier, and fractions of the product have been subjected to further tests to prove the presence of large amounts of the addition product of chlorine and toluene. (3) On the basis of reactionrate data for the chlorination of benzene under gamma radiation., a preliminary design has been made of a plant to produce benzene hexachloride. A comparison is given between the estimated cost of the product from this gamma-activated process and product from the conventional ultraviolet process which is employed commercially. The following sections of this report describe these studies in greater detail.

B. POLYMERIZATION OF ETHYLENE The study of the effect of amma radiation on the polymerization of ethylene as inaugerated by Lewis has been considerably extended. Lewis carried out all polymerization runs at room temperature and at varying pressures and radiation dosages. In the work reported here, primary interest has been in varying the temperature of the polymerization reaction to determine its effect on the rate of the reaction and the nature of the products. 1. Equipment Used The high-pressure stainless-steel bomb built by Lewis was used for all polymerization runs. A set of heaters built by Lewis was placed around the bomb in all runs, regardless of whether heating was required, so that the radiation flux within the bomb would be the same in all cases. The bomb was loaded and emptied in the same manner described by Lewis.21 The temperature was raised from 13 to 2200C, while the reacting pressure of pure ethylene was varied from 1075 to 1950 lb/sq in. 2. Visual Inspection of the Products In the case of the three runs at the lowest temperatures (runs 145800, 145801, and 145807 described in Table 2) three distinct layers of solid polyethylene product were found in the bomb at the conclusion of the runs. The top layer was a spongy, white mass that adhered to the sides of the bomb with a physical consistency much like that of a fine porous sponge. The middle layer was a white curdy material, much resembling the curds in cottage cheese, and was directly opposite the center of the gamma radiation source; theref ore., receiving the highest dose rate. The bottom layer was a hard solid layer that rested on the bottom of the bomb. This layer was from three quarters to two inches thick and had to be cut out of the bomb with a chisel. Runs 145808 and 145812 yielded a f inely divided white powder which adhered to the sides of the bomb. There was no particular variation in the physical appearance of this powder within the bomb. The powder did not always fall to the bottom of the bomb, but usually remained attached on the sides of the stainless-steel vessel.

liquid of high viscosity with an odor similar to that of used motor oil. After standing several days, a finely divided white powder precipitated our from the product and left a clear cream-colored liquid layer. TABLE 2 IRRADIATION OF ETHYLENE Averaged over Radiation Yield Run Initial Reactor Polymer, A, Number Pres- Reaction Hours Dose Total gm gm-moles reacted sure, Temp, Irradi- Rate, Dose, metric megapsig ~C ated kilorep/hr megarep ton / \ rep 145800 1450 13 65.1 54.2 3.53 21.8 2370 145801 1260 13 91.75 53.8 4.94 39.7 3570 145807 1950 39 70.9 53.5 3.79 45.6 5860 145808 1500 88 71.33 26.6 1.90 1.4 438 145812 1400 90 91.93 47.0 4.32 3.8 542 145813 1425 220 77.8 53.0 4.12 6.3 1695 145814 1075 217 114.5 35.4 4.o6 17.9 3860 145817* 940 25 98.25 37.8 3.72 15.1 2130 145824** 410 25 64.o 62.5 4.oo 33.9 7030 *42 psig gaseous SO2 added to reaction. **20 gm liquid S02 added to reaction; vapor pressure was 42 psig. 3. Physical Properties of the Solid Polyethylene Product The radiation yields (A values) given in Table 2 are in general higher than those reported by Lewis21 but the dose rate to the reactor also was usually higher. The melting points and tensile strengths of the polyethylene prepared by polymnerizing ethylene in a field of gamma radiation are listed in Table 5. The tensile strengths are somewhat higher than those reported by Lewis. These tensile strengths were made on an Instron tensile tester with a head and/or jaw speed of ten inches per minute. This speed was selected to eliminate the cold flow phenomeon observed in plastics in general and polyethylene in particular. The tensile strengths determined by Lewis were made on a Gardiner-Parks tensile testing machine where the coldflow phenomeon was present.

TABLE 3 PROPERTIES OF IRRADIATED ETHYLENE Run Sample Tensile Elongation Number Section Melting Point, OF Strength, at Rupture, Lower Upper psi in./in. 145800 Middle 258.4 >700 335554.24 145801 Middle 263.0 >700 3211.23 145801 Top 259.0 >700 2947.30 145807 Top 257.2 >700 2455.24 145807 Bottom 260.4 >700 2600.25 145808 Total 222.0 230.41 Too small 145812 Total 217.2 221.7J [ sample J 145813 Total Liquid Liquid 145814 Total Liquid Liquid 145817 *Total >465 145824 *Total >450 * Decomposes before melting at the temperature indicated. the film with a standard sample die. The sample was somewhat larger than those used by Lewis. Melting points of the polyethylene solids were determined on the laboratory mel-ting-point bar5 and in general are higher than those obtained by Lewis. They are, however, consistent with those of Lewis when account is taken of the difference in dose rates. 4. Other Work on Polymerization of Ethylene Hayward and Bretton17 at Yale University, working with a 400-curie gamma ray colbalt-60 source, have investigated the polymerization of ethylene under conditions ranging from 8o to 46o0F and from 0.5 to 21 atmospheres. Both liquid and solid products were obtained. Lewis21 observed the effect of pressure on the polymerization of ethylene in both 1-kilocurie and 1O-kilocurie sources at this laboratory.

A comparison of the ranges of temperature and pressure covered by Hayward and Bretton and by Lewis is shown in Fig. 8 of this report. 200 LEWIS (21) 150 ~~~~~~150 _ _ E THIS REPORT [] HAYWOOD AND BRETTON (17) 1a 1r00 50 LIQUID 0.5 ~~~SPECIFIC VOLUME 1.0 5 Pressure-Volume Diagramn for EthlIene (drawn from the data of Perry's Chemical Engieer Handbook) Fig. 8. Regions of Investigation of Polymerization of Ethylene under the Influence of Gamma Radiation. C. POLYMERIZATION OF ETHYLENE AND SULFUR DIOXIDE As reported by Lewis, the addition of sulfur dioxide to ethylene in the reaction bomb under gamma radiation produces a copolymer of the two compounds. The results of the two runs conducted in this study are reported in Tables 2 and 5. A fluffy white granular product was obtained which decomposed before melting at greater than 4-i50OF; leaving a brown and black residue. All attempts to -press this polymer into a film in order to obtain tensile test specimens resulted in decomposition of the product. The film

These two runs are reported here in the interest of coupling with the results of prior work. The physical properties of these products are being investigated more fully and future work is palnned in the copolymerization of the two compounds. D. CHLORINATION OF AROMATIC COMPOUNDS 1. Investigation of the Structure of the Primary Product Obtained from the Gamma Ray Activated Reaction of Toluene with Chlorine (a) Dehydrohalogenation. The supposed hexachloro addition product which has been obtained by reaction of toluene with chlorine under gamma radiation4 has been investigated more closely in order to establish whether it is indeed the postulated product. The first step in the investigation was the degradation to a supposed trichloro toluene by means of dehydrohalogenation in boiling pyridine. C ~ 5Cl < CH5 H~~~~~~~~~~~~~~~~~~~~~ H C1 Boiling pyridine 1 H - HC1 Cl C1 H C1 and other isomers Previously it was demonstrated that nearly one-half of the total chlorine content was eliminated during this reaction.3 As starting materials for the reaction., mixtures of high-boiling products obtained from the chlorination runs in the lO-kilocurie-source equipment4 were used and several dehydrohalogenations were carried out. The initial materials used for one dehydrohalogenation were 176 gm of a material boiling at 125 to 1540C (30 to 60 microns of Hg) mixed with 89 gm of a material boiling at 123 to 1580C (65 to 75 microns Hg). The theoretical chlorine content for a hexachloro addition product of toluene and chlorine is 69.8 percent; the materials used contained 69.8 to 70.1 percent chlorine, respectively, indicating the presence of some higher chlorinated material in the latter of the two. This mixture of materials was added to about 1.5 1 of redistilled pyridine and refluxed for 47 hours. On cooling, crystals, presumably pyridine hydrochloride, formed in the dark green liquid. The material in the flask was added to a large excess of

remained on the filter paper. The filtrate was a dark brown oil density 1.4 gm/cc, weight 87.7 gm. Assuming the starting materials to be C7H8C16 and the oil to be C7H5C13, a yield of 53.3 percent is indicated for the dehydrohalogenation reaction. The oily product from the dehydrohalogenation was fractionally distilled in a Podbielniak column at atmospheric pressure; a summary of these data is presented in Table 4. Each of the first nine fractions obtained by distillation was recrystallized from methanol, while the remainder, being nearly insoluble in methanol, was recrystallized from ethanol. Fractions three through eight gave two phase mixtures when the methanol was partially evaporated and the resulting residues cooled, so these phases were separated and crystals taken from each. Melting points were taken on a melting-point bar. Some fractions were analyzed for chlorine content by the use of sodium diphenyl reagent followed by a Volhard titration, but the values thus obtained are only approximate because of small sample size. Carbon-hydrogen analyses were also taken on certain samples. It is apparent from the chemical analyses of these materials that both tri- and tetrachloro compounds are present, although the majority of material appears to be the expected trichloro compound. The presence of the tetrachloro material may arise from either or both of two causes; the dehydrohalogenation may have been incomplete, or the higher chlorine content of part of the starting material may have been carried through as an extra chlorine atom in some of the mol ecules. The melting-point ranges found in many of these compounds indicate that separation was not complete. However, they indicate that one of the products is undoubtedly the known compound,, 2,.74,5-tri'chlorotoluene., whose melting point is given in the literature as 8o to 82ocllY~24,28 and whose boiling point has been given as 229 to 2300C 3 (b) Oxidation of the Dehyd-rohalogenated Material. A quantity of supposed hexachloro addition product of toluene and chlorine was dehydrohalogenated as before and then oxidized to obtain an acid. For this treatment, 60 gin of material boiling at 90 to 1560C (90 to 240 microns of Hg) were placed in 100 ml of pyridine and heated at just below the reflux temperature for 48 hours. Water was added to extract the pyridin'e hydrochloride and excess pyridine. The organic layer was neutralized with nitric acid, then 750 ml of 2o% nitric acid were added. About one-half of this twophase mixture was sealed into glass tubes and heated for 66 hours at 155 to 1530C. On cooling, large floculant crystals and a solid phase separated out from the solution.

TABLE 4 FRACTIONS OBTAINED BY THE DISTILLATION AND RECRYSTALLIZATION OF THE DEHYDROHALOGENATED PRODUCT FROM REACTIONS OF CHLORINE WITH TOLUENE UNDER GAMMA IRRADIATION Boiling Range, Upper or Lower Fraction Weight, ~C Layer from Melting Point, % by Analysis Number gm (atm pressure)Methanol Mixture ~C C1 C H 1 0.9 210.5-231.0 2 1.3 228.0-238.5 3 8.o 238.5-241.5 Lower 77.5 54.0 43.0 2.68 4 1.5 242.0-242.5 Lower 77.3-78.6 Upper 79.8 5 9.3 241.5-242.0 Lower 77.5-78.0 54.2 Upper 77.8-78.6 6 4.4 242.0-243.0 Lower 78.0-78.6 56.8 43.0 2.50 Upper 76.4-77.8 7 2.7 243.0-245.5 Lower 76.4-78.0 Upper 77.0 8 3.0 245.5-257.0 9 o.6 258.0-272.0 66.7-70.2 10 1.3 273.0-277.0 80.2-88.8 11 3.3 277.0-279.5 88.2-89.8 61.4 36.7 1.86 12 1.1 279.5 80.2-86.6 13 1.1 277.0-278.5 87.6-88.8 61.1 14 1.6 278.5 87.5-90.0 61.1 15 5.1 Hold-up drained Not 94.0-96.8 61.4 from column recrystallized Theoretical values for percent of C, H, and Cl: Trichlorotoluene: C = 43-0o '%, H= 2.58 %, Cl = 54.4* Tetrachlorotoluene: C = 36.6 %, H =1.75 '%,, Cl = 61.7* Tetrachloromethyl cyclohexadiene (incomplete dehydrohalogenat ion product): C = 56.2 %v, H = 2.61 %, Cl = 61.2 *

The solids were filtered off from the nitric acid solution and dissolved in ammonium hydroxide solution. This solution was extracted with toluene to remove nonionic materials. The organic acid was then precipitated by addition of nitric acid, filtered, washed, and dried. The yield was 9.5 gm, or 42 percent overall, based on the original starting materials and assuming trichlorobenzoic acid as the product. Identification of the final product required separation of the crude product into its various isomers. Since each isomer should differ in its acidstrength, the first attempted separation was based on this property. A series of extractions was carried out on 4 gm of acid so that during each step one-half of the acid would dissolve in an aqueous ammonium hydroxide layer, while the other half would be retained in the organic phase. These extractions were carried out until 32 divisions were obtained. However, when the melting points of these 52 fractions were taken, long ranges inicated that practically no separation had been effected. Since the differences in acid strength of these materials thus appeared to be quite small, it seemed advisable to turn to a chromatographic type of separation. Still making use of the acid strength characteristics of the material, ion-exchange chromatography was employed. A highly basic anionic resin, Dowex-l, 8 percent cross linked., 200 to 4oo mesh, was used. Two columns were used in series, the first 400 x 10 mm ID and the second 20 x 6 mm ID, each being packed two thirds full of resin. The first column was jacketed and water at 550C was circulated around it. Before use, the resin was backwashed. and. put through a chlorid~e-hydroxi'de cycle twice. After several experiments, it was found. that the following cond~itions were satisfactory for operation of the column. The resin was prepared. for use in the chlorid~e form. The mixed. isomers of the acid. (about 1 gin) were introd~uced. at the top of the column as the ammonium salt solution in 75% ethanol (10 ml volume). The column was washed. with d~istilled. water and. elution was carried. out with 0.01N and. 0.lN hydrochloric acid. in 75% ethanol (25% water). Twenty 5-ml samples were collected. at a slow rate, 1 to 16 with 0.01N acid. and. 17 to 20 with 0.lN acid.. Initially, d~ark band~s were observed. to move d~own the column, but these became ind~istinguishable as they approached. the base of the upper section It was found. on evaporation of the solvent from the fractions that the first two contained. nearly no solid~s, while numbers 5 through 9 contained. about equal amounts of solid~s. Numbers 10 through 12 contained. small amounts of impure solid~s) and. the remaining fractions contained. only traces of solid~s. The solid. materials were recrystallized. from hot water.

200 190 180 70? 160 0' 150 - 140 I 1300 Z,.J w120 'IC 100 90 80 702 I 4 5 2 3 4 5 6 7 8 9 10 FRACTION NUMBER F ig. 9. Melting-point Ranges for Fractions of Isomeric Trichlorobenzoic Acids Separated by Ion-exchange Chromat ography. The melting-point maxima are accompanied by a decrease in temperature range showing an increase in purity of the material in that particular fraction. The much lowered melting points and long ranges of fractions 2 and 8 indicate possibly that these were collected at a point of transition from one chromatographic band to another. Fractions 7 and 10 exhibit the sharpest melting points and lie in a range close to that reported for isomers of trichlorobenzoic acid. The increased range but higher first melting point of fraction 10 could perhaps be attributed to the presence of amounts of a much higher melting isomer in this fraction. All melting points of this series of fractions are recorded in Table 5, together with chloride analyses and data for comparison with reported values for trichlorobenzoic acids. Fractions 4 and 5 merit special attention. During the determina

TABLE 5 FRACTIONS OBTAINED BY CEROMATOGRAPHIC SEPARATION OF ACIDS FROM THE DEHYDROHALOGENATED PRODUCT FROM REACTION OF TOLUENE AND CHLORINE Melting Point % Calculated from Fraction Melting Point, * by Analysis Comparison Compound of Comparison Comparison Compound Number ~C C1 C H Compound C1 C H 2 118.7-155.0 3 147.0-154.4 4 156.0-160.6 5 155.6-162.0 \1 6 145.6-153.2 7 159.4-160.8 49 37.4 1.69 2,4,5 trichlorobenzoic acid 163 47.2 37.3 1.34 2,3,6 trichlorobenzoic acid 163-4 2,4,6 trichlorobenzoic acid 164 8 110.5-120.7 9 170.2-175.8 10 182.6-183.6 46 2,5,4 trichlorobenzoic acid 186-7 47.2 11 180.8-187.2

was found to sublime, leaving a material of a higher and sharper melting point. In a further investigation, quantities of fractions 4 and 5 were placed in a sealed melting-point tube. This tube was partially immersed in an oil bath at 180'C until a film of material had condensed on the cooler walls of the upper end of the tube. The tube was then sealed off at the center and melting points of the more volatile materials were taken. Table 6 is a summary of information on the products of this separation procedure and data for comparison with values reported for isomers of trichlorobenzoic acid. Work on degradation of the product of the reaction of toluene with chlorine under gamma irradiation can be summarized by the following equations: H C1 C1 C CH CHE3 Cl Pyridine ~ + 3HC1 Cl iC1 Reflux 1lCl H ClH C and other isomers (1) Cl Cl CH5 ~COOH I ~~~~~2C$ HNO5 Cl Heat 1 or other or other 1 Cl isomers isomers The prod~ucts obtained. by d~ehyd~rohalogenation and. subsequent oxid~ation confirm the hypothesis that chlorine ad~ds to the ring of toluene and. further ind~icate that substitution o f more halogen atoms for hydrogen atoms occurs in some of the molecules of the hexachloro ad~dition compound.. 2. Separation of Prod~ucts of Chlorination of Toluene (a) Attempted. Crystallizations. Several unsuccessful attempts were mad~e to obtain a crystalline solid. from the syrupy form in which the hexa

TABLE 6 MELTING POINTS OF SEPARATED COMPONENTS OF FRACTIONS 4 AND 5 OF TABLE 5 Melting Melting Point % Calculated Fraction Point, Comparison Compound of Comparison for Comparison * by Analysis ~C Compound Compound C H C H 4-sublimate 166.0-166.5 2,3,6 trichlorobenzoic acid 163-164 37.3 1.34 537. 4 1.22 2,4,6 trichlorobenzoic acid 164 in mixture in mixture before before 4-residue 163.2 2,4,5 trichlorobenxoic acid 163 37.3 1.34 sublimation sublimation 5-sublimate 166.0-166.5 2,3,6 trichlorobenzoic acid 163-164 37.3 1.34: 37.4 1.69 2,4,6 trichlorobenzoic acid 164 in mixture in mixture before before 5-residue 162.0 2,4,5 trichlorobenzoic acid 163 37.3 1.34 ~sublimation sublimation

mixed with the syrup, but repeated extraction procedures gave no evidence of separation of components of the chlorinated material or of production of crystalline solids. (b) Analysis of Solids from the Reaction Mixtures. The liquid mixtures consist chiefly of unreacted toluene, benzyl chloride, and the chlorine addition products of toluene. The liquid product from several runs became cloudy after a period of standing and a small amount of very fine crystals slowly separated. The solids which formed were investigated further to determine whether they were crystalline 1,2,3,4,5,6-hexachloro-methylcyclohexane. They were separated from the mother liquors by centrifugation, placed on porous plate for oil removal, and recrystallized from methanol. Analysis for chlorine content showed variable composition (ranging from 72.1 to 72.8* chlorine) for materials from various runs and can probably be regarded as impure heptachloro products (theoretically 73.1 chlorine) arising from substitution of additional chlorine in the hexachloro product. 3. Investigation of Reaction Rates between Toluene and Chlorine (a) Apparatus. A series of reactions between toluene and chlorine was carried out to obtain data from which the kinetics of this reaction could be determined in order to gain better insight into the nature of the activation produced by the gamma irradiation. The equipment used was a combination of certain parts d....escribed. in previous reports. Gas hand.ling and. control equipment were similar to that outlined. in Fig. 23 of Progress Report 6.~ The glass reaction vessel shown in Figs. 62 and. 63 of Progress Report 55 was used. at various d~istances from the source and. was connected. to the glass lines lead~ing to the gas-control apparatus located. on the second. floor. In early experimental runs the amount of chlorine was ascertained. by weighing the gas cylind~er d~uring the run, but it soon became apparent that this proced~ure was not accurate enough because of the inherent insensitivity of the type of balance usable und~er.the large weight of the steel cylind~er. An orifice-type meter was next used. in which the pressure drop of the gas was measured. as it passed. through a calibrated. glass orifice; results in this case also lacked. reprod~ucibility. Satisfactory measurements of the inlet gas flow were finally obtained. with the use of a pair of 'tRotometer" type flowmeters. These instruments were mad~e of Pyrex with sapphire floats and. were connected. in series so that the two d~ifferent ranges might be used. without the need. of stopcocks. They were calibrated. against nitrogen, ethylene, butane, carbon d~ioxid~e, and. ethane, accord~ing 22

A four-way stopcock was used for taking samples of exit gas from the reaction; this made it possible to purge a gas absorption bottle with nitrogen, pass all the exit gas through it for a specified time, then purge out the connecting tubes with nitrogen again. The absorption bottles were provided with fritted-glass dispersion tubes to facilitate efficient absorption in the solution. The fritted tubes were calibrated frequently to assure that the pressure drop of the gas through each was not significantly different from the rest. During experimental runs, except for the periods during which samples were taken, the exit gas was vented to the atmosphere or passed through large absorption bottles containing 50% sodium hydroxide solution. Figures 10 and 11 illustrate the gas-handling equipment. (b) Analytical Methods and Preliminary Results. The method used for analysis of exit gases, and hence, the choice of absorption solution, received considerable attention and several difficulties were encountered. Methods making use of liberated oxygen from hydrogen peroxide or of oxidation of iodine compounds were found to lack precision under the experimental conditions of this work. Since it can be shown theoretically that ferrous ion should be oxidized quantitatively to ferric ion by chlorine, an analytical procedure was devised making use of this reaction. In order to gain efficient solution of the gas, two liquids were used in each absorption bottle. A lower layer of reagent-grade carbon tetrachloride provided rapid absorption of excess chlorine in the exit gas, while an upper layer of aqueous acidic ferrous sulfate dissolved all hydrogen chloride in the gas stream. Examination of two absorption bottles in series showed that gas absorption was complete at the flow rates ordinarily used. After absorption of the gas sample,, the absorption bottle was shaken vigorously. This caused the dissolved chlorine in the lower layer to react completely with the ferrous sulfate of the upper layer. An aliquot portion of this layer was then titrated with ceric sulfate for ferrous ion content, which when subtracted from the original content gave the amount of chlorine absorbed. A second aliquot was oxidized with hydrogen peroxide to convert all ferrous ion to ferric ion and analyzed for chlorides by the Volhard method. Thus., free chlorine and total chloride could be determined from a single absorbed gas sample. Preliminary investigations indicated the method should be capable of giving precise results on free chlorine determinations. A series of experiments using the equipment previously described and using the ferrous absorption method were made. When the gas flowmeter had been properly calibrated and the reaction rates calculated., it was found that results were very scattered. Therefore, a very thorough investigation of analytical methods was imperative. Using samples of 99.5 to 99.84 pure chlorine gas, runs were carried out in which the gas passed through dry glass tubes

9 || CREO TERACLOR~~~~~~~I DE-l l| g| || VNoT +: | | I N L ~E T MGAS K 4LINE L T5. * eFLOWMETERS |&N H \COOLANT l l l l NIROENANALYTICAL LSNM ICROCUP~ |LN R 1 WRECATABSORPTION IMCOCLNEBOTTLE I~~ H CONTROLLER H- ITU STIRRm ONTROLLER ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 GAM I A I~ lOOT" DOUBLE THROWC YINR RELAY O I I ~~~~~~~~ ~ ~~~~~~~~~TAIL GAS UBOTLES VACUUM MANOMETER _____________________PUMP___THERMOSTAT BATH CARBON TETRACHLORIDE- INE ANMTR +1 I ],I CHLOROFORM-DRY ICE VENT INLET MANOMETERS I I ~~~~~~~~~~~~~TO 4. ( it ~~~~~~~~~~~~~ ATMOSPHERE HHEATER ',~.~~.~ THERMOCOU PLE. POTENTIOMETE I -- COOLANT - SOLUTIONS SURGE BOTT.E BP:,~~~~~~~~~~~~~~~~~~~~~ LEADS1 I INTERMITTENT:tZ CONTINUOUS PUMP I THERMOCOUPLE' - ----- REACTANT SOLUTION WELLS STEEL JACKET REACT~~~~~~~~~~~~l I I Hi iIII liiii 4.~~~~~~~~~~~~~~~~~~~T OUPLE ~ ~ ~ ~ ~ ~ OUC FIc Fig. 10.FlwheforEupetUetoDtrieRatoRaebeweTounanChrn.

Fig. 11. Gas-Control Equipment Rack Used for Reactions between Toluene and Chlorine. samples were absorbed in potassium iodide and titrated with sodium thiosulfate, a significantly different deviation from the flowmeter values was obtained. When potassium hydroxide was employed as abosrbent, followed by hydrogen peroxide reduction of the hypochlorite formed., the chloride ion titrate in this solution showed still a different deviation from the flowmeter value. A satisfactory explanation for the failure of the ferrous sulfate method of analysis has not been found, but it was abandoned in favor of a more precise Absorption of chlorine in alkaline sodium arsenite has been published as an analytical method. 12,34,37 By use of 0.2N sodium arsenite in N potasacidified and analyzed for chlorid.e ion by the Volhard titration method.. Thus.,

(c) Equipment Changes. Further refinements of technique for kinetic runs have involved certain modifications in equipment. A new all glass reactor, illustrated in Figs. 12, 13, and 14, has been constructed. The stream of chlorine gas is caused to bubble back and forth across a series of five baffle plates. Above this is a space for foam separation. Further up the tube is a series of baffles, placed so that the exiting gases follow a spiral path throwing entrained spray against the outer walls of the tube. The chief advantages of this design are larger contact time between the gas andliquid phases, more efficient separation of entrained liquid, less open space for gas above the liquid with resulting decreased lag in the time required for gases to return for analysis, complete separation of parts for cleaning, and location of the thermocouple well in a position for most accurate indication of temperature. A steel jacket has been constructed to fit the new glass reactor in order to provide for a very rapid linear flow of coolant along the walls of the reactor tube. New larger diameter tubes have been installed in the lO-kilocurie-source room to permit the coolant circulation equipment to be placed outside the source room without causing undue pressure drops in the leadlines. This combination of improved cooling equipment should make temperature control more precise and flexible. In order to insure that the glass reaction vessel is always free from contaminants which might catalyze or inhibit the reaction, special cleaning procedoures have been ad.opted. and equipment constructed to facilitate these proced~ures. After cleaning with solvents and. drying, the reactor is placed. in a mixture of concentrated. sulfuric acid. and. chromic acid. for a few hours. It is then rinsed. in d~istilled. water and. placed. in a steam cleaning tower. About one and. one-fourth liters of water is vaporized. and. passed. through this tower so that all surfaces of the dissassembled. reactor are thoroughly steamed.. The parts of the reactor are then d-ried. at 1500C in a special oven wherein the glass parts rest on a plate glass shelf and. d~o not come in contact with any metal. When blown out with filtered. air and. cooled. the reactor is then assembled. for use and. filled. with reactant liquid.. This new equipment is now being used. to stud~y the kinetics of the chlorination reactions. E. DESIGN OF A PLANT FOR THE CHLORINATION OF BENZEN~E UNDER GAMAMA RADIATION Benzene and. chlorine react und~er a variety of cond~itions to prod~uce either ad~dition or substitution of chlorine to the benzene ring. In 1903, Slator32 first showed. that ultraviolet light selectively promotes the hexa

12 /5 ~~ 12/30 T 1 2/30 ' 7mm 10/30 10/30 4-5 cm 3.7 1 cm.2 16.6 cm 25 cm mm 25 mm 6m-a 3cm 7mm 5mm 12/5 ~{5~I~m 8 cm 34/45 6mm 1 5mm 4mm 34/45 39 cm I.6crm FROM j. TO THIS LINE 1.8cmm 28 mm25m 6mm- - E E 12.3 cm 0 ~~~~~~~~~~~~E:lz ~~~~OVERALL H H ~~~~LENGTH H (9 ( z z 45cm LU LUL 6mm 4 mm 3.7mm 4cm u ~~~~~~~~~~~4cm c THERMOCOUPLE 5 mm WELL TO TERMIN- ~6mm ATE 8 cm FROM BOTTOM OF OUTER TUBE INLET TUBE TO c TERMINATE I cm4c FROM BOTTOM OF OUTER TUBE 2cm ~6mm

Fig. 15. Glass Reactor Used for Fig. 14. Glass Reactor, Determining Reaction Rates between Dissassembled. Toluene and Chlorine. l~~~l~~~4~,22,2 1am aito 2,162 ultraviolet light 10,14'22,25 alpha radiation, and gamma radiation. 621 Studies using gamma radiation were made in this laboratory with cobalt-60. The commercial importance of this addition product lies in the insecticidal properties of the gamma isomer. These properties have been described by Slade51 and others.9,10l18 1. Laboratory Chlorination of Benzene in the Presence of Gamma Radiation The procedure for chlorinating benzene under the influence of gamma radiation has been described previously-2,l6)21 In brief this procedure is as follows: A glass reactor containing benzene and carbon tetrachloride was lowered into the cobalt-60 gamma source and purged of air with nitrogen. A cooling medium was passed through an outer jacket of the reactor. Chlorine was bubbled into the reaction mixture until the first traces of HC1 or C12 appeared in the off gas from the reactor and then was regulated at a rate so that there was no HC1 or C12 in the off gas. The reaction was usually conducted until all the benzene was consumed. 28

It was repcorted2 that the temperature could not be controlled during the reaction if pure benzene was used; therefore, carbon tetrachloride, an inert in the reaction, was added to give greater reaction volume for closer temperature control and also to act as a solvent for the solid reaction prod~uct. Part of the data taken has been reported22,16,21 and this is further supplemented by Table 7 and Fig. 15. Due to equipment breakage and recovery problems, the solid product in some cases was not obtained. Rather than discard these runs as useless, a weight percent completion of reaction has been assumed. In those runs where an obvious excess of chlorine was added, the average completion obtained was assumed; this was 95-percent completion based on the benzene. In those runs where the reaction was stopped before completion, the amount of chlorine added was recorded and 95 percent of the benzene that would react with the chlorine stoichiometrically was calculated. This amounted to 90 percent completion based on the total benzene concentration. The reaction velocity constant for the complete reaction was calculated. This constant ktr is due to the combined temperature and radiation effects. A dscussion of this combined constant is to be found in the theoretical 5 work by this laboratory in Progress Report 53, pages 36-40. Figure 15 is a plot of this constant versus the reciprocal of the absolute temperature at a constant dose rate of 105 kilorep per hour. [G] values have also been calculated and are to be found in Table 7. 2. Rate of Reaction for the Ad~dition Prod~uct of Chlorine and. Benzene The original investigation of this laboratory was concerned. with the effect of gamma rad~iation on the yield. of gamma isomer in the mixed. isomer reaction prod~ucts and. it was d~etermined. that there was no apparent affect of gamma rad~iation on yield. of gamma isomer. The rate of ad~dition of chlorine to the reaction mixture was not constant nor uniform., and. no d~ata on the kinetics of the reaction based. on chlorine concentration are available. This type of d~ata is necessary to d~esign a flow reactor for the system. In the literature, Noyes and. Leighton in Photochemistry of Gases2 report the rate as approximately r =kI 1/ [l [616 In Chemical Action of Ultraviolet Raysl0 the rate is given as

TABLE 7 IRRADIATION OF BENZENE AND CHLORINE BY COBALT-60 GAMMA RADIATION AT A DOSE RATE OF 105 KILOREP PER HOUR Average Weight Reaction Run Volume Yield Reaction Reaction Percent Velocity Constant, G Number C6H6, C6H6C16, Temp, Time, Completion gm-moles gm ~R min on C6H6 ktr (hr)(ft3) 129762 100 129765 100 49.6 498 35 15.2 82.6 47,100 129766 30 *93 520 40 **95 136.0 51,60 129767 30 *93 525 35 **95 155.5 59,000 129768 30 102.8 526 25 105 241.0 91,40 129774 30 40.0 474 33 40.8 70.4 26,800 129861 25 *52.6 528 15 ***90 205.0 76,0O0 129855 20 *62 528 14 **95 258.5 91,500 129769 20 *48 520 24 ***90 116.5 42,500 129770 20 63.6 528 20 97.3 186.0 68,0oo 129771 10 32.0 528 25 98.0 74.7 26,100 129772 10 31.7 483 53 97.2 34.8 12,200 129773 10 40.8 477 46 125 51.8 18,100 *Calculated values using assumed percent completion. **Weight percent completion on C6H6 assumed by average of other runs where excess chlorine was introduced. Weight percent completion on C6H6 assumed. on runs where stoichiometric amounts of chlorine were ad~ded.. r ot1/2PC12 PC~6H r ca I(1.6) Luther and. Guld~berg report 22 r a~1C2

1000 900 9 800 \ Reaction Velocity Constant ktg Moles Rx 800 tr (hr) (ftt3) i 700 for the Reaction: 600 CeHe+ 3C1 C HCl 500 400 30C 200 U) ~ ~ Q t-100 N 0 go ~ ~ 8O0 0 870 80 ' 70 \0 --- —-- 60 \ 50 \ 40 30 20 10 1.8 1.9 1 OR X 102 2.0 2.i Fig. 15. Reaction Velocity Constant vs Reciprocal Temperature for the Chlorination of Benzene. and. 10 molecules of reaction product per quantum of light, where P = total pressure p = partial pressure I= ultravoilet intensity T = temperature r = rate t = time k = reaction velocity constant L]= concentration.

. Process Design Calculations for a Continuous Flow Reactor For the purpose of this design calculation, it is necessary to make several assumptions regarding the data available. Enumerated with easons for their selection, these assumptions are: (1) The reaction is first order with respect to the chlorine concentration. An excess of benzene is to be used and recycle through the reactor. This excess will reduce the effect of small benzene concentration on the reaction rate and also provide a medium to slurry out the solid product. This also assumes a similar mechanism of reaction under gamma radiation as under ultraviolet rays. Luther and Goldberg, quoted above, give reaction as first order with respect to chlorine. (2) The data obtained in this laboratory for different reaction temperatures give the reaction velocity constant. The plant is designed using the same radiation rate as was present in the laboratory. Only temperature is to be extrapolated. (3) Gunther14 and others9 report a yield of 42% gamma isomer in the reaction of chlorine and benzene under ultraviolet light rays on the addition of NaOH to the reaction mixture. An addition of 2 NaOH in the mixture is assumed to give 32% gamma isomer by weight. (4) The same reaction mechanism holds for the case of a mixture of liquid chlorine in benzene as for the case of bubbling gaseous chlorine through benzene. This permits the use of the reaction velocity constant of Table 7 and Fig. 13. The same percentage completion on the chlorine will be held. This allows use of these overall constants. With these assumptions., the equation and material balance for the reaction capable of producing 1000 lb of gamma isomer per day on 95 percent completion of the chlorine are as follows: C6116 + 3C12 + gamma radiation =C6H6C16 Basis: 1000 lb gamma isomer per day Wt mixed isomers per day -.32 3120 lb Wt benzene per dy = (5120)(78) 8

Wt chlorine per day = (3120) (71)(3) = 2410 lb (.95) (291) Wt NaOH per day = (3290) (.02) = 65 lb. If 100 percent excess benzene is used for the reasons in assumption 1, then the weight entering the reactor is (82)(2) + (2410) + (65) = 4239 lb day or 177 lb/hr This corresponds to a feed rate of 2.38 ft3/hr at 850F. The equations for the rate of the reaction in the flow reactor (with the assumptions noted) of volume VR are r =ktr [012] d NC 12 I dNC12 J/'dVR = f-~l = __ r k C12 = dNC12 kN C12 where F ft3/hr Feed Rate Cl ~ (lb-moles Cl2) Chlorine Rate (hr) 1 =R ft3 By selecting 85oF as the operating temperature, a liquid phase reaction may be obtained at not too high a pressure. With the range of data available, long range extrapolation of the data is not advisable due to the possibility of error. Assuming that 95 percent of the chlorine is reacted in a single pass through the reactor, the limits of integration are

Outlet C2 Cl = (1.41)(.o5)=.07 (lb-oles C12) (hr) The reaction velocity constant from Fig. 15 at 850F is 445 (gm-moles) (hr)(ft3) Converted to units of /(hr) this is 445 (gm-moles) (ft3) (lb-mole) 78 lb 1.4 (hr)(ft) (.88)(62.4)lb (454)(gm-moles) (lb-mole) (hr The volume of the reactor under these conditions is then by integration of the above expression VR (2.38) Loge 1.41 = 5.1 cu ft. (1.4).07 Using a nominal three-inch OD high alloy stainless-steel or nickel pipe for a reactor, approximately one hundred feet are necessary. This would correspond to ten passes ten feet long in the field of gamma radiation. The heat of reaction for this system calculated. by an approximate method.51519 is 271,000 Btu/(lb mole) of benzene hexachlorid~e. Thus, (271,000) ((3120)/(24)(291)) = 121,000 Btu/hr must be removed. from the reaction mixture. A five-inch CD steel pipe located. concentric to the three-inch reactor pipe would. make an effective heat-transfer jacket. The cooling med~ium would. pass through the annuli between these pipes. 4. Description of the Proposed. Plant Process The flowsheet of the proposed. process is to be found. in Fig. 16. The liquid. benzene and. chlorine pass from the storage to a mixer., and. then in the liquid. phase at 15 atmospheres pressure through the nickel reactor in the rad~iation chamber. A plan view of this chamber is to be found. in Fig. 17. Ten passes are necessary at the d~ose rate of 105 kilorep per hour.

- --—!~~~I ___I ~ SOURCE| MAKE AGEST UPAVE I~~~~~~ Ij I )I REFRIGERATIONREACOR UNIT IC___ 'qk~~~~~~~~~~~~~~~~~~~~~~~"l HOT WATER I WiATE TO SERDA S T IOAN ' Fig. i6. Flowsheet of Proposed. Plant Process for the Ad~dition of Chlorine to Benzene und~er the Inf luence of Gamma Rad~iation. reduced. to 5 atmospheres. This will vaporize any remaining chlorine which may then be compressed. and. recycled. back into the system. The benzene hexachlorid~e stream from this chlorine vaporizer is pumped. through a benzene vaporizer where it is sprayed. on hot water or into steam. The pressure is red~uced. from 5 to 1 atmosphere. The excess benzene is vaporized. and. passed. through a cond~enser and. then recycled. at 15 atmospheres into the reactor. The vaporizing benzene leaves behtind. the benzene hexachlorid~e in water which can be either dried. or purif ied.. The gamma isomer may be separated. from the mixed. isomers and. sold. as pure Lind~ane or as d~usting powd~er in the unpurified. form. This separation into the pure isomers is not of interest in this plant d~esign. Method~s of analysis and. separation are reported. in th'kenI + liteature I1,c

24' K 3- 4" 5- 5 IP 4",o~~4,...o. ____,* co. A A OD 'I.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I REACTORS OTANK SOURC "''~~~~~~~~~~~T. " '.'D"".::"-:L-"IP" "OO-I"G:JA'KE-..'...- P I STD 3' O.D. '.PPERACO Fig. 17. Plan View of Proposed Reaction Cav Showing Reactors in Positn.. 56

Since the operating temperature was selected as 850F for several reasons listed above, it is necessary to use a system of refrigeration to effect heat removal from the process. The cooling in the reactor, chlorine condenser, and benzene cooler must be accomplished by this refrigeration. This amounts to a removal of Reactor 121,000 Btu/hr Chlorine condensor (.07)(lb moles) (8800) Btu 620 Btu/hr hr lb mole Benzene cooler (1.05) (882) lb (.46) Btu (70) OF = 1,250 Btu/hr 24 hr lb OF 122,870 Btu/hr or approximately 123,000 Btu/hr from the three sources. The benzene condenser may have water as the cooling medium as it involves heat transfer at higher temperatures. At a higher operating temperature this refrigeration unit could be replaced by water as the cooling medium. Calculation of Gamma Radiation Source Strength and Shielding Two possible arrangements of the reactor tubes in relation to the source are considered. The gamma radiation source may be located inside the reactor or concentric to an annular flow reactor, or the source may be a finite rectangular plaque with the reactor located on either side. This latter arrangement has been assumed. for these calculations. The reactor size has been calculated. to be ten passes each ten feet long. In ord~er to utilize a rectangular rad~iation source most efficiently, the arrangement of reactor tubes is as pictured. in cutaway AA of the plan view of Fig. 17. Five reactor tubes are placed. on either sid~e of the source with the first row six inches from the source. The second. row is placed. on an equilateral triangle pitch behind. this first row. Assuming a d~ose rate in the first row of tubes equal to that used. in the laboratory work and. correcting for any absorbing med~ia between the source and. this point the rate of gamma emmission on the surface of the source 3Y13 ~~~~4 may be calculated.. Gomberg, et al. 51 and. Brownell, et al. have d~one this. Using the equation d~eveloped. by Lewis21 for the intensity of rad~iation through an absorption med~ium which in turn utilizes the ad~sorption coef ficients 7 -7 ~ 3

I= Io e-ZLPX X i2. 08)CM (1.0) (2.25)in.+(-075)cm (8 c m* =m g cm gm om —7 g' gm where 2 H20 and C6H6 = 08 cm Fe and Ni 075 gm Fe and N~~g 105 = Io e-1.215 Io= (3-37)(105) krep = 354 kre I = radiation intensity, krep/hr I = absorption coefficients, (cm2/gm) x = thickness absorber, (cm) p = density of absorber, (gm/cm3). The specific activity Of the gamma radiation source is assumed to be 3.8xl03 (rep)/(hr)(curie)i'313 A rectangular source plaque two feet by eight feet is assumed for the commercial design as compared to 16 cm2 used in the laboratory experiment. Then the approximate total activity of the source plaque is 5354(krep)x10 (rep) 2(ft) 8(ft) 144(in 2) 6.45(cm2) (hr) (krep ft~27 (17n. ~) =86,, oo curies. 38x 1000 (krep) (16)(cm2) (hr) (curie) The source-room size is selected and the concrete shielding necessary to protect the human operators can be calculated. Assuming that the dose rate in air varies immensely with the square of the distance from the source, the dose rate at a distance of five feet from the source would be = 12 I (1) I54oo

I = 985 rep hr Assuming a.05 mr/hr to be a safe dose rate for the human operators, I = Io ex0 (.05) mr 985 Mr -.2x (.05mr = 985 rep x 1000 ehr hr rep where = 0.2 cm1 for concrete x = thickness of concrete in cm e=2x = 985,000 = 1.97x107.05.2x = in 1.97x107 x - (7.294)(2.303) = 84 cm (.2) x = 331 in. Assuming.05 mr/hr to be a safe dosage for these calculations, based on narrowbeam coefficients., a wail thickness of forty inches or three feet four inches is seiected. (Based on broad-beam coefficients and forty inches of concrete, the radiation ievei is computed to be increased to about one mr/hr.) A weii of tweive feet of water wiii permit iowering the source to a sufficient depth to provide adequate shieiding. A pian view of the radiation chamber is shown in Fig. 17. A cutaway perspective view of this chamber showing the reactor in place is shown in Fig. 18. The reactor-end supports and the source elevator have been omitted from the figure to clarify the picture. 6. Estimation of Cost for the Proposed Equipment and Radiation Chamber An itemized cost summary of the proposed radiation chamber is given in Table 8. The cost estimation for all equipment used in the process

J4 C) Fig. CLjtaT ng Of Reac tor Cave Sho.. lng 1012 SOLIrce al2d React or

TABLE 8 COST ESTIMATES FOR RADIATION CHAMBER Concrete for walls, roof, floor, and well 130 yd at $20.00/yd $ 26 Reinforcing 4 Ecavation for footings and well 8 Forms for concrete 8oo Labor 18 Stainless-steel well 1500 Elevator for source 1500 Ion-exchange equipment 2000 Monitoring equipment 18 Painting, wiring, plumbing, and ventilation for cave including labor 3200 $16,oo Miscellaneous contingencies (20%) 33 Engineering (5) 8o Contractors fee (140) 16 Total $22,100 TABLE 9 COST ESTIMATES FOR EQUIPMENT EXCLUDING ISOMER SEPARATION Nickel or stainless-steel reactor pipe and fittings into cave $ 900 Heat-exchange jacket 4oo Pumps and pipe from pumps 1900 Two-flash vaporizers 3ooo Chlorine compressor and connections 8ooo Heat exchanger for benzene 300o Refrigeration unit 18000 Benzene and chlorine storage 4o00 Total $39,200 Installation (50% of cost) 11800 Building for housing pumps, control equipment, etc. 9000 Total $60,000

7. Comparison of Gamma Radiation Sources to Give Constant Source Strength The source plaque as described in calculating the total gamma-radiation strength could be one solid sheet of radioactive material or it could e composed of a number of individual sections. This would be essentially a ack which held the source made of several sections. These sections could be in the form of strips or cannisters and could be removed and/or replaced as t became necessary. This replacement time would depend on the permissable flux variation of the source and the half life of the material used for the ource. For the following calculations, it is assumed that a flux variation of 20 percent will not affect the rate of reaction of benzene and chlorine ppreciably. The time (t) necessary for a radioactive material of original intensity (Io) to decay to an intensity (I) is I = Io e-kt I t - 1 In o k I here k is a constant of the material. Letting the material decay to half its original value, then I0 2 t - ln 2 =.69 k k or k=.695; where t.5 is the defined half life of the material. Then for any other time of decay I I o e(69) Letting I =.8o 10, then

or (.693)t.80 1 = n- i n 1.25.22 t.80.5.223.8 = 693 t.5 This equation is the relation of the time for any material of half-life to decay to 80 percent of its original value of intensity. Both t8 and t 5 must be in the same units. Assuming a finite number of sections of the source, then at the end of one.80-life, a quantity of gamma radiation equal to 20 percent of the rated source strength must be added to the source. If this means replacement of any section of the source, then the strength of the replaced section must also be present in the replacement section to keep the rated radiation value constant. For example, assume that the rectangular plaque consists of twenty individual sections. At the beginning of any.80-life period the plaque will be at the necessary rated strength calculated previously. The ratio of the section strength to the rated section strength will be (.8)n, where n is the number of.8-lives since the section was replaced. Assuming one replacement per.8-life, the source will consist of twenty sections with strengths Section 1 2 5 4.... 19 20 1 (.8)1 (.8)2 (.8)3.... (.8)18 (.8) 19. At the end of any.8-life each section will have decayed by a.8 amount and the source will consist of twenty sections of strengths Section 1 2 5 4. 19 20 (.8)1 (.8)2 (.8) 3 (..8)4.. (.8) 19 (.8) 2 This shows that the addition of one section of rated section strength must secton srenth Smus then beeqalt

[.000 - (.8)20] s = (86,5oo)(.20) curies, where 86,500 curies is the total rated source strength; then [.000 -.0115] S = 17,300 curies 17,9300 ~curies S= 17,00 = 17,500.988 gamma section Table 10 is a comparison of four d~ifferent gamma rad~iation sources of known half life. Three gross fission products, six-months old, one-year old, and two-years old, are compared with cesium-137. The approximate half lives of the fission products were taken from the literature.) TABLE 10 COMPARISON OF FOUR GAMMA RADIATION SOURCES TO GIVE CONSTANT SOURCE STRENGTH FOR FIVE YEARS Fission- W Section Total I Total Radiation Product Half life t.8 yr ISections Replacements Sections Necessary for Material t5 yr to Start in 5 yr in 5 yr 5 yr, curies Gross 6-mo old.20*.o64 5 78.1 83.1 1,455,000 Gross 1-yr old..30*.097 5 51.5 56.5 990,000 Gross 2-yr old 1.10*.354 5 14.1 19.1 334,000 Pure Cs-137 33. 10.6 5 0 5 86,500 Approximate half life for mixtures of elements of varying half life. The calculations for the six-month-old. fission-prod'uct material are illustrated. here. The others follow similarly.

The number of gamma radiation sections to start is five. The number of gamma replacements in five years is Replacements = 5.00 5.00 t. 80. ( = 78.1 Then the total gamma sections necessary for five-years operation are Starting sections + Replacement sections = Total sections Total sections = 5 + 78.1 = 85.1 Then the total rated radiation in curies necessary for five years operation is curies (Total sections) ( ci Total radiation =Total radiation = (85.1)(17,500) 1,4n55,000 curies In the case of cesium-137 it is seen that the half life is long enough that no replacement sections are needed. The original five sections have not decayed to the permissable.8o level of radiation. 8. Cost Comparison of Benzene Hexachloride from Four Gamma Radiation Sources and from Present-Day Process Little is known about the cost of either gross or concentrated fission products. There are some figures available for pure radioactive isotopes forspecific purposes. A range of values expressed in dollars per curie is assumed in Table 11 for the four gamma radiation sources considered. The range is wide enough to take into consideration use of the fission product at the nuclear reactor site where no shipping is necessary and large quantities of hot fission products are presently in storage. These costs have been assumed on the basis of approximate equal cost per unit of energy

TABLE 11 COMPARISON OF TOTAL PLANT COST FOR FOUR GAMMA RADIATION SOURCES Fission- Range of Estimated Product Cost for Radiation, Total Radiation Total Plant Material Dollars per Curie Cost, Dollars Cost, Dollars ~Low High Low High Low High Gross 6-mo old.000.05 728 72,8 82900 154,900 Gross 1-yr old.001.05 990 49,5100 11,6 Gross 2-yr old.02.25 6,680 83,500 8 Pure ~Cs-117.50 5. 00 43,250 432,500 125,4 5146 t E = Io e = Io (1-et )/\ 0o it follows that the energy varies inversely as \ for large values of t. Since %~ is the reciprocal of the half life, it follows that the energy delivered varies directly as the half life. Thus, a source with three times the half life of another can deliver three times as much radiation energy and is assumed to cost about three times as much. The cost of long half-life sources is estimated to include paqkaging, shipping, and installation in a plant at some distance from where the source is prepared. The cost of very short half-life sources would be excessively high because of the large shipping cost. Therefore, it is assumed that the short half-life sources are to be used in the vicinity where they are produced. In this way it is believed that the cost per curie of the short half-life material (which might even be spent fuel elements) although low,, is not unrealistic. Table 11 is a comparison of the total plant cost including the radiation source for the radiation sources considered. The values and the

(No. curies)(cost/curie) = Total cost for radiation (1,455,000ooo curies)(.0005 $/curie) $728 The total radiation cost plus the cost for equipment and for the radiation cave found in Tables 8 and 9 is the total plant cost. Total cost for radiation = $ 728 Cost for radiation chamber = 22,100 Cost for equipment = 60,ooo $82,828 or $82,900 Table 12 is self explanatory. Carrying through the calculations for the low range value of six-month-old fission products and the annual cost of the proposed plant, amortizing the plant investment over a period of five years, and assuming 10 percent interest on the total plant, the investment amounts to Annual plant cost, 82,900/5 $16,56 Interest, (.10)(82,900) 8290 Working Capital 20,000 Operation and maintanence 0,0 Total Annual cost of plant operation $74,850 per year Assuming operation and. prod~uction d~uring 290 d~ays of the year., the d~aily cost of the operating plant would. be $74,850o $258 per d~ay. 290 The cost of raw material i7),2 Benzene at $.42 per gal =$50.50 Chlorine at $5.50 per 100 lb =84.5o or a total operating cost of $393 per d~ay. The cost of prod~ucing 5120 lb of mixed. isomer is then 51209l per d~ay - $.126~ per lb mnixed.qisome-rs.

TABLE 12 COMPARISON OF TOTAL PRODUCTION PRICE OF THE GAMMA ISOMER OF BENZENE HEXACHLORIDE PRODUCED FROM FOUR GAMMA RADIATION SOURCES Fission-Product Gross Gross Gross Pure Material 6-mo old 1-yr old 2-yr old Cs-137 Dollars Low High Low High Low High Low High Annual plant cost 16,560 31,000 16,610 26,300 17,720 33,100 25,100 103,000 on 5-yr amortization Annual interest to 8,290 15,490 8,310 13,6oo00 8,880 16,560 12,500 51,460 stockholders, lo% on total plant cost Working capital 20 20,000 2 0,000 20, 000 2 0,000 20,000 20,000 2 0,000 2000 Operation and 350 30,000 30,000 30,000 30,000 30,000 30,00 000 30,000 maintenance of plant Total annual cost, 74,850 96,490 74,920 89,900 76,600 99,660 87,600 204,460 excluding raw materials Daily cost assuming 258 333 258 310 264 344 302 705 290 operating days per year Total daily operat- 393 468 595 445 599 479 457 84o ing cost, including raw materials $/( 'gunit)(lb.00595.oo468.00595.oo445.00599.00479.00457.oo84o isomers) on32 isomer yield $/(I unit)(lb.oo63.0075.oo63.00714.oo64.00768.0070.01545 isomers), assuming 20* Y isomer yield

Defining the gamma unit as percent of gamma isomer per pound of mixed isomers, the cost per gamma unit assuming 32* gamma isomer produced is.126 3126 = $.00395 per gamma unit per pound. 32 If the recovery was only 20 percent gamma isomer in the reaction of mixed isomers, as is true in many commercial ultraviolet units today, the production price for the same production of mixed isomers would be.126 $.00oo63 per gamma unit per pound. 20 Benzene hexachloride is being sold at present market conditions8,26 at $.009 to $.015 per gamma unit per pound in concentrations of 12 to 14* gamma isomer per pound. Using this lower figure for comparison, a savings of 50 to 225 percent can be realized by producing benzene hexachloride by gamma radiation activation as against conventional ultraviolet light activation. This is true for the whole range of costs of all the gross mixed fission products. Only the highest value estimated for cesium-137 falls above $.009 per gamma unit per pound. This is, however, in the range of the price being asked today. The separation cost of the isomers would be comparable to that method. used. to0day and. has not been calculated.. These calculations, however, reveal a more efficient means of obtaining the crude isomer mixture. 9. Conclusions and. Summary Several important improvements in existing plant operations have been overlooked. in presenting this economic stud~y. Some of the more importaRt improvements present in this proposed. plant d~esign are (1) Elimination of necessarily small costly equipment d~ue to the necessity for ultraviolet rad~iation penetration. Gamma rad~iation will penetrate to the very center of the reaction mixture d~ue to its very nature and. wavelength. (2) Elimination of glass or quartz equipment necessary to transmit ultraviolet rad~iation. Steel equipment may be used. with gamma rad~iation.

(4) The full capabilities in the use of gamma radiation have not been fully realized in this study due to the nature of the data used. There are several important regions where the use of gamma radiation could be investigated more fully. Enumerated they are: (a) Higher rates of radiation should be investigated. The data presented are for a constant rate of radiation. The possibility of utilizing larger radiation sources to give larger radiation fluxes and increased rate of reaction should be studied. (b) Running a larger chlorine recycle stream and/or running to a less complete reaction based on the chlorine would undoubtedly increase the overall rate of reaction. The rate of reaction is faster at higher concentrations of chlorine and running to 95 percent completion on the chlorine decreases the overall rate of reaction. (c) Running at even higher pressure than presented here would allow running a higher temperature with the chlorine still in the liquid phase. The data available, although dependable and consistent, cover too small a temperature range to allow long extrapolation. At a higher temperature the reaction rate would be greater. (d) For simplicity of design only one effective absorbing bank of tubes has been used (see Fig. 3, Sec. AA). Other rows of tubes could be placed behind the tubes shown to absorb radiation not used in this design. Brownell, et al)'-4 show an increase in efficiency of 500 percent in using a multipass package plant as against a single pass dead carcass plant for irradiation of meat. By increasing the capacity of the plant and putting more reactor bundles in the radiation chamber, the cost of crude benzene hexachloride would be reduced. Only equipment costs would increase in the cost estimate. The major expense of the radiation source would be the same. (e) An annular reactor could be used with the source of gamma radiation in the center. This would presumably utilize a higher percentage of the radiation emitted and decrease the rated strength necessary in the source. (f) The plant cost was written off in five years, but it must be realized that in all cases there is gamma activity left at the end of five years. Less than 10 percent of the radiation of Cs-1357 has been used at the end of this five-year write-off period.

undoubtedly show a greaterdifference in the cost of benzene hexachloride as it is produced today by ultraviolet radiation and benzene hexachloride produced utilizing gamma radiation as presented here. F. WORK CONTEMPLATED FOR THE NEAR FUTURE Present plans are to make a few more runs on the polymerization of ethylene with gamma radiation to supplement the data already taken. Following that it is probable that no more experimental work will be undertaken on this reaction. However, a preliminary plant design, such as has been given for the chlorination of benzene, will probably be made to obtain an estimate of the cost of polyethylene produced by gamma radiation for purposes of comparison with the cost of polyethylene produced by the conventional high pressure process employed industrially today. oday. A few additionalrs will be made on the rate of reaction of chlorine with toluene under gamma radiation utilizing the best technique for analysis of the off-gases of chlorine and hydrogen chloride. The proof that the addition compound is the principal product is now complete. Chlorination of other compounds mentioned in previous reports will probably receive very high priority in the experimental work. The aromatics closely related to benzene and toluene, such as ethyl benzene, the xylenes, and naphthalene, will be chlorinated to determine whether gamma radiation favors the addition of chlorine to the double bonds. It is also expected to chlorinate some simple aliphatic alcohols., aldehydes, and acids. Some preliminary work along this line has shown that under gamma radiation chlorine probably converts a hydrated aldehydic group to an acidic group., a reaction which apparently does not occur when ultraviolet light is employed. Some other high pressure polymerizations are scheduled to be tried soon. Also it is hoped to carry out a copolymerization under gamma radiation, with the combination of butadiene and styrene at low temperatures; this reaction will probably be tried first. G. REFERENCES 1. Alyea, H. N., J Am Chem Soc, 52, 2743~ (1950). 2. Anderson, L. C., Martin, J. J., Brownell, L. E.., et al., "Utilization of the Gross Fission Products??, Progress Report 4 (COO-124) Eng. Res. Inst.

4. Ibid., Progress Report 6 (C00oo-198) April 1954. 5. Anderson, W. T., Jr., Indus and Eng Chem, 9, 844-6 (July, 1947). 6. Chamlin, G. R., J Chem Educ, 23, 283-4 (June, 1946). 7. Chem and Eng News, 32, 1227-50 (March 29, 1954). 8. Ibid., 32, 2597-2620 (June 28, 1954). 9. Chem Ind, Staff Report, 60, 418-20, March, 1947. 10. Ellis, C., Wells, A. A., and Heyroth, F. F., Chemical Action of Ultraviolet Rays, American Chemical Society, Washington, D. C. 1941 527542-44. 11. Feldman, I. K. and Kopelisvich, E. L., Arch Pharm, 27, 488-96 (195). 12. Gleu, K., Z Anal Chem, 95, 385 (1933). 13. Gomberg, H. J., et al., "Design of a Pork Irradiation Facility Using Gamma Rays to Break the Trichinosis Cycle" Eng. Res. Inst., Report No. M943:4-2-T, Univ. of Mich. and Mich. Mem. Phoenix Proj. 54, Ann Arbor, Michigan, February, 1954. 14. Gunther, Chem and Ind, 65, No. 44, 599 (1946). 15. Handbook of Chemistry and Physics, Chemical Rubber Pub. Co., 28th Ed, 1944, 1445. 16. Harmer, D. E., Anderson, L. C., and Martin, J. J.., "Chlorination of Some Aromatic Compounds under the Influence of Gamma Radiation", Paper presented at First International Nuclear Engineering Congress sponsored by AIChE, Ann Arbor, Michigan, June, 1954. 17. Hayward, J. C. and Bretton., R. H.., "Kinetics of the Ethylene Reaction Initiated by Gamma Radiation"l, Paper presented ar First International Nuclear Engineering Congress sponsored by AIChE, Ann Arbor., Michigan, June, 1954. 18.. Hensill, G. S., Soap and San Chem, 27, 155 (September, 1951). 19. Hougen, 0. A. and Watson, K. M., Industrial Chemical Calculations, Chapter VI, 2nd Ed, Wiley and Sons., New York, 1947.

21. Lewis, J. G., "Promotion of Some Chemical Reaction with Gamma Radiation Thesis, Eng. Res. Inst. Report No. M94:4 —T, Univ. of Mich., Ann Arbor, Michigan, January, 1954. 22. Luther and Goldberg, Z Physik Chem, 6, 4 (1906). 23. Martin, J. J., Chem Eng Prog, 45, 338-42 (1949). 24. Musante, C. and Fusco, R., Gasy Chim ital, 66, 69-48 (196). 25. Noyes and Leighton, Photochemistry of Gases, Reinhold Publishing Company, New York, 1941, 294-296. 26. Oil, Paint, and Drug Reporter, 166, No. 5, 12, 19 (August 2, 1954). 27. "Photochlorination", Chem Soc J. 5578-89 (December 1950). 28. Qvist, W., Acta Acad Aboenisis Math et Phys, 15, No. 5 (1946). 29. Saur, R. L., "Preparation and Properties of the Gamma Isomer of,2 4 5,6-Hexachlorocyclohexane", Thesis, MSC, 1949. 30. Seelig, E., Annalen der Chemie, 237, 1 (1887). 51. Slade, Chem and Ind. 64, 514-519 (October 15, 1945). 52. Slator, A., Z Physik Chem, 45, 450 (1905). 55. Snyder and Powell, "Absorption of Gamma Rays", AECD-2759, Oak Ridge National Laboratory, Oak Ridge, Tennessee, August, 1950. 54. Solray Process Division, Allied Chemical and Dye Corporation, The Analysis of Liquid Chlorine and Bleach, 3rd Ed, 1954, 15-17. 55. Stanford Research Institute, "Industrial Uses of Radioactive Fission Products", Report to U.S. AEC, Stanford, California, September, 1951. 36. "Symposium on Photochemistry", J Phys Coll Chem, 52, 457-611 (March,1948). 57. Treadwell, F. P. and Christie, W. A. K., Z f'ur angew Chemn 18,9 1950 (1905). 58. Whittingham, D. 3. and Garmaise, D. L., Can 3 Research, 27B, 415-20 (1949).

UNIVERSITY OF MICHIGAN 3 901 5 03095 030011111111 3 9015 03095 0300