ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN ANN ARBOR Final Report EFFECTS OF NUCLEAR RADIATION ON POLYETHYLENE AND CERTAIN CHEMICAL TRANSFORMATIONS Lo M. Hobbs J. A, Manson Project 243o SPENCER CHEMICAL COMPANY KANSAS CITY, MISSOURI July 1957

The University of Michigan * Engineering Research Institute ACKNOWLEDGMENTS We wish to thank Dr. John Ro Brown, Dr. Go E. Hulse, Dr. A. Deutschmann, and Mr. Ko Ao Kaufman-all of the Spencer Chemical Company-for their advice and continuing interest. We are also indebted to Mr. W. C. Bull, Mr. Go Chenoweth, Mro Lo Landrum,and Mr. P. Strickler for helpful discussions. Progress of the work was helped considerably by Dro R. Go Craig and Mr. So Co Kothari, both of whom worked full-time on the project for several monthso We wish to acknowledge the part-time assistance with experimental work given at various times by the following: Messrso Fo Baumgartner, R. To Bigda, Wo Graessley, Do Fuerst, Jo Gallini, and L. Po Sharmao The cooperation of Mr. Jo Vo Nehemias, of the Fission Products Laboratory, and Mr. Ao Ho Emmons and Mr. Wo Ro Dunbar, of the Phoenix Laboratory, in making arrangements for the irradiation of our samples is also gratefully acknowledged. Finally, we appreciated assistance with the molecular weight distribution problem given by the following members of the faculty of The University of Michigan~ Dr. C. Co Craig, Director of the Statistical Research Center, and Dr. No D. Kazarinoff, Department of Mathematics. ii

The University of Michigan * Engineering Research Institute TABLE OF CONTENTS Page LIST OF TABLES v LIST OF FIGURES vii ABSTRACT viii OBJECTIVE ix GENERAL INTRODUCTION 1 Io THE POLYMERIZATION OF ETHYLENE BY GAMMA RADIATION 2 Introduction 2 Experimental 3 Results and Discussion 4 Conclusions 13 IIo EFFECTS OF GAMMA RADIATION ON POLYETHYLENE 14 Ao Crosslinking of Polyethylene 14 Introduction 14 Experimental 14 Results and Discussion 15 Conclusions 18 B. Modification of Crosslinking in Polyethylene 19 Introduction 19 Experimental 20 Results and Discussion 21 Conclusions 39 C. Graft Polymerization of Monomers to Polyethylene 40 Introduction 40 Experimental 40 Results and Discussion 41 Conclusions 44 Do Determination of Molecular Weight Distribution in Polyethylene by an Irradiation Technique 44 Introduction 44 Experimental 45 Results and Discussion 47 Conclusions 53 iii

The University of Michigan * Engineering Research Institute TABLE OF CONTENTS (Concluded) Page IIIo Miscellaneous Studies 53 A. Mooney Viscosity Measurements on Polyethylene 53 Introduction 53 Experimental 54 Results and Discussion 54 Conclusions 57 B. Synthesis of Cyclohexanone Oxime Using Gamma Radiation 58 Introduction 58 Experimental 58 Results and Discussion 60 Conclusions 61 C. Fixation of Nitrogen by Gamma Radiation 61 Introduction 61 Experimental 61 Results and Discussion 62 Conclusions 63 APPENDIX - CALCULATION OF MOLECULAR WEIGHTS FROM SOLUBILITY DATA OF IRRADIATED POLYETYLENE 64 General 64 Correction for Branching 64 Correction for Chain Scission 65 Determination of Molecular Weight and Distribution65 Sample Calculation for A-22 66 REFERENCES 68 iv

The University of Michigan * Engineering Research Institute LIST OF TABLES Table Page I EFFECT OF GAMMA RADIATIO ON THE POLYMERIZATION OF STYRENE IN SOLUTION 5 II EFFECT OF REACTION VARIABLES ON THE SENSITIZED POLYMERIZATION OF STYRENE USING GAMMA RADIATION 6 III THE EFFECT OF GAMMA RADIATION ON THE POLYMERIZATION OF ETHYLENE IN THE PRESENCE OF DILAUROYL PEROXIDE 8 IV THE EFFECT OF GAMMA RADIATION ON THE POLYMERIZATION OF ETHYLENE IN THE PRESENCE OF AZO COMPOUNDS 10 V THE EFFECT OF GAMMA RADIATI- ON ONTHE. POLYMERIZATION OF ETHYLENE IN THE PRESENCE OF PEROXIDES 12 VI THE EFFECT OF ADDITIVES ON THE PHYSICAL PROPERTIES OF IRRADIATED POLY-ETH oo0085 22 VII THE EFFECT-OF TYPE OF CARBON BLACK ON THE PHYSICAL PROPERTIES OF CARBON-FILLED IRRADIATED POLY-ETH o1008 5 23 VIII THE EFFECT OF FLAKE GLASS ON PHYSICAL PROPERTIES OF IRRADIATED POLY-ETH 1008o 5 25 IX THE EFFECT OF ABSORBED VINYL CHLORIDE ON PHYSICAL PROPERTIES OF IRRADIATED POLYETHYLENE SHEETS (POLY-ETH. 1008.5, 20-MIL) 25 X THE EFFECT OF ABSORBED MONOMERS ON PHYSICAL PROPERTIES OF IRRADIATED POLYETHYLENE SHEETS (POLY-ETH 1008.5, 20-MIL) 27 XI THE EFFECT OF ABSORPTION OF MONOMERS -ON THE HEAT RESISTANCE OF IRRADIATED POLYETHYLENE SHEETS (POLY-ETH 1008.5, 20-MIL) 28 XII THE EFFECT OF TOTAL DOSE ON HEAT RESISTANCE OF IRRADIATED TREATED POLYETHYLENE SHEETS (POLY-ETH 1008.5, 20-MIL) 30 XIII THE EFFECT OF TIME OF CONTACT WITH VINYL CHLORIDE PRIOR TO IRRADIATION ON PHYSICAL PROPERTIES OF POLYETHYLENE SHEETS (POLY-ETH 008.5, 20-MIL) 3355 XIV THE EFFECT OF TIME OF DEGASSING ON PHYSICAL PROPERTIES OF IRRADIATED TREATED POLYETHYLENE SHEETS (POLY-ETH 1008. 5, 20-MIL) 355 v

The University of Michigan * Engineering Research Institute LIST OF TABLES (Concluded) Table Page XV PHYSICAL PROPERTIES OF IRRADIATED TREATED POLYETHYLENE FILM (POLY-ETH 1008.5, 2-MIL) 36 XVI PHYSICAL PROPERTIES OF SOME IRRADIATED TREATED POLYETHYLENE SHEETS (POLY-ETH 100805, 20-MIL) 37 XVII PHYSICAL PROPERTIES OF POLYETHYLENE FILMS IRRADIATED IN THE PRESENCE OF MONOMERS 42 XVIII PHYSICAL PROPERTIES OF POLYETHYLENE IRRADIATED IN THE PRESENCE OF MONOMERS IN SOLUTION 43 XIX MOLECULAR WEIGHT DISTRIBUTION DATA DERIVED FROM SOLUBILITY CURVES OF IRRADIATED POLYETHYLENE 50 XX SOLUBILITY DATA FOR IRRADIATED POLYETHYLENE 51 XXI MOONEY READINGS FOR POLYETHYLENE AT 280~F 55 XXII MOONEY READINGS FOR POLYETHYLENE AT 240~F 56 XXIII PREPARATION OF CYCLOHEXANONE OXIME 59 I________________________________ vi

The University of Michigan * Engineering Research Institute: LIST OF FIGURES Figure Page 1 Effect of gamma radiation on the polymerization of ethylene in the presence of dilauroyl peroxide. 9 2 Effect of gamma radiation on the tensile strength of Poly-Eth 1005 (irradiated at -70~C). 16 3 Effect of absorption of monomers on the heat resistance of irradiated polyethylene sheets (Poly-Eth 100o85, 20mil). Test I. 29 4 Effect of vinyl chloride on the heat resistance of irradiated polyethylene sheets (Poly-Eth 1008.5, 20-mil). Test VII. 31 5 Effect of vinyl chloride on the heat resistance of irradiated polyethylene sheets (Poly-Eth 1008.5, 20-mil). Test IX. 32 6 A typical solubility curve for irradiated polyethylene (A-22). 48 7 A typical solubility curve for irradiated polyethylene (A-10). 49 8 Dependence of Mooney scale readings for polyethylene on time. 57 vii

The University of Michigan * Engineering Research Institute - ABSTRACT The polymerization of gaseous ethylene by gamma irradiation in the presence of sensitizers such as peroxides and azo compounds was investigated at several temperatures (60, 80, and 100~C) and at a pressure of 2000 psi. Results show that the initial rate of polymerization at 60~C in the presence of dilauroyl peroxide was increased l.4-fold; this difference in rates would correspond to a temperature difference of about 5~C if equal rates were maintained. On the other hand, no increases in the rates of conversion of ethylene to solid polymer at 80 and 100~C were observed with the following initiators: di-t-butyl peroxide, dicumyl peroxide, azobisisobutyronitrile, and diisopropyl azisobutyrate. In fact, in many cases, a decrease in rate resulted from irradiation, apparently because of the formation of nonsolid products under the influence of radiation. Effects of gamma radiation on polyethylene were also studied; samples of Poly-Eth 1005 (melt index 1.8) and Poly-Eth 1008o5 (melt index 20) were irradiated at temperatures ranging between -70 and +120~C, and physical properties determined. In general, the tensile and yield strengths passed through a maximum as the radiation dose was increased, while the melt index and percent elongation decreased steadily. However, doses required to improve the strength properties resulted in fabrication difficulties, especially with Poly-Eth 1005o In another attempt to improve the properties of polyethylene, the effect of incorporating modifiers and fillers in polyethylene prior to irradiation was examined. The addition of such substances as carbon black, styrenep.oraniline, resulted in a deterioration of properties; on the other hand, the irradiation of glass-filled polyethylene resulted in a product much tougher than filled unirradiated polyethyleneo The use of vinyl monomers to modify the nature of irradiated polyethylene was also attempted. Best results were found with the grafting of -gaseous vinyl chloride to 2-mil film and to 20-mil sheets of Poly-Eth 1008.5. Provided that the polyethylene had been freed from occluded oxygen, irradiation of a film or sheet with a dose of from 0.8 x 106 to 1.2 x 106 rep in the presence of vinyl chloride yielded a product having improved toughness and remarkable resistance to deformation (that is, shrinking and buckling) on heating. Because of the close relationship between irradiation-induced crosslinking in polyethylene and the molecular weight distribution, the effect of irradiation on the gel fractions of several polyethylene samples was studied. It was found that a careful examination of curves relating gel fraction to radiation dose could give useful information about the breadth of the molecular weight distribution, and, in some cases, the molecular weight. viii

The University of Michigan * Engineering Research Institute Preliminary experiments with the use of gamma radiation in the synthesis of cyclohexanone oxime and the fixation of nitrogen were conducted; results were not promising, as yields were low and difficult to reproduce. In the one nonradiation study made, it was found that samples of polyethylene having the same melt index could be differentiated by measurements of Mooney viscosity. OBJECTIVE The primary objective of this investigation was to study the application of nuclear radiation (especially, gamma radiation) to the synthesis and modification of polyethylene. The secondary objectivewasto study the use of nuclear radiation to promote certain chemical transformations such as the synthesis of cyclohexanone oxime and the fixation of nitrogen. ix

The University of Michigan * Engineering Research Institute GENERAL INTRODUCTION The interaction between various types of radiation (a, P, y, light, etc.) and matter has been of at least academic interest for many years. When radiation is absorbed by a material, energy is transferred to the electrons of the molecules. A complex set of processes then occurs culminating in the formation of excited molecules, ions, and free radicals; the free radicals formed this way then either combine with each other or react with other molecular species in the medium. Thus radiation (a, P, y, light, etc.) may be used to initiate many free radical reactions. For many years tne most common type of radiation used for the initiation of reactions were ultraviolet light and x-rays; sources of C, i, and 7 radiation were generally less common. However, during the last few years, sources of nuclear radiation such as Co-60 or spent fuel elements. from nuclear reactors have become fairly readily available; at the same time, the cost of these radiation sources has decreased steadily. Indeed it seems likely that nuclear radiation will soon become a competitively priced source of energy; considerable attention has been given, both in industrial and academic laboratories, to the possibility of using this radiation for the initiation of chemical reactions on a commercial scale. This possibility has been especially interesting to the Spencer Chemical Company because of its concern both with commercial power and with the production of a wide range of chemicals and polymers. One of the most obvious applications of nuclear radiation is the initiation of chemical reactions known to be initiated by ultraviolet lights In spite of some basic differencesbetween the effects of light and nuclear radiation, it is known, for example, that some reactions, such as the chlorination of hydrocarbonsp which may be initiated photochemically, are also initiated by nuclear radiation.1 Perhaps the most extensive work with the applications of nuclear radiation to chemical reactions has been in the field of polymer chemistry. For example, a good deal of interest has developed in the use of nuclear radiation to initiate polymerization; the polymerization of vinyl monomers such as styrene and ethylene has been studied in some detail. 2' Another type of reaction, especially interesting becauseasmall amount of chemical reaction leads to large effects on physical properties, is the modification of polymers already formed. This type of reaction has led to much experimental work, and, indeed, to commercial application. Thus several laboratories both in this country and in Europe have studied the effects of radiation on polyethylene;,'5 already some. irradiated polyethylene products are available commercially. 6,7 For these reasons, it was decided to undertake a study of the use of 1

The University of Michigan * Engineering Research Institute - nuclear radiation to promote chemical reactions of commercial importance. More specifically, it was decided to study the synthesis and modification of polyethylene by gamma radiation, and also the synthesis of some compounds of interest to the sponsors-namely, cyclohexanone oxime and nitric oxide or ammonia. I. THE POLYMERIZATION OF ETHYLENE BY GAMMA RADIATION INTRODUCTION The initiation of polymerization in ethylene by gamma radiation has been studied by several research groups. Bretton and others8 obtained solid polymers by the polymerization.of ethylene at one atmosphere pressure and at temperatures between room temperature and 167~C.o Lewis, Martin, and Anderson3 reported preparation of solid polyethylene having high crystallinityo According to another study, by Hayward,9 the use of pressures up to 21 atmospheres is advantageous in this polymerization to avoid the formation of liquid polymers. On the other hand, little attention has been paid to the incorporation in the ethylene of radiation-sensitive compounds (that is, compounds that form free radicals on dissociation by radiation) to enhance the effect of radiation. A British patentlO claims that chlorohydrocarbons, aliphatic alcohols, and other oxygen-containing compounds are suitable sensitizers, but it gives no exampleso The most obvious sensitizers would be, of course, ordinary initiators such as peroxides. Except for some work on the photochemical decomposition of peroxidesll-l4 however, no studies of the effects of radiation.on peroxides appear to have been published. It was decided, therefore, to examine the effects of gamma radiation on the polymerization of.ethylene in the presence of some typical chemical initiators such as peroxides and azo compounds. It was hoped that it might be possible to conduct a conventional polymerization at a given rate in the presence of radiation at a temperature lower than that required for the same rate of polymerization in the absence of radiation. Originally the study was intended to deal with the continuous polymerization of ethylene at high pressures (20,000 psi). When this was found to be impractical, the batch polymerization of ethylene at lower pressures (2000 psi) was investigate-d it was hoped that the results obtained could be. extrapolated to continuous operation at higher pressures. To obtain preliminary information about the effect of radiation in initiators, the polymeri zation of styrene by irradiation in the presence of several initiators was examinedo Results are presented and discussed below. A preliminary investigation was also made to test the feasibility of ---- 2 ---

The University of Michigan * Engineering Research Institute irradiating an initiator prior to polymerization. Results of this study:, which consisted of the irradiation of a frozen suspension of dilauroyl peroxide, are also given in this sectionO EXPERIMENTAL Styrene (99%) was obtained from the Dow Chemical Company; solvents were of technical grade. Two cylinders of ethylene were used-one supplied by the Matheson Chemical Company, and one by the Spencer Chemical Company, data used for the comparison of a radiation experiment with its blank were always obtained using ethylene from one cylindero Except for the diisopropyl azoisobutyrate (Spencer Chemical Company) and the dicumyl peroxide (Hercules Powder Company), liquid and solid initiators were supplied by the Lucidol Division of Wallace and Tiernan, Incorporatedo Dispersions-of initiators in mineral oil were supplied by the Spencer Chemical Company. The diphenylpicrylhydrazyl used in some experiments was prepared by the oxidation of diphenylpicrylhydrazine hydrochloride with lead dioxide;15 the concentration of hydrazyl in solutions was determined spectrophotometrically. Polymerizations of styrene were conducted in sealed glass tubes. The tubes were filled with 15 ml of toluene, 5 ml of freshly distilled styrene, and, unless specified otherwise, 0.20 g of initiator; before polymerization the contents were degassed. After polymerization the product was precipitated by pouring the contents of the tubes into methanol. For the ethylene polymerizations, a 250-ml Magne-Dash autoclave was used. The procedure was as followso After addition of a solution of the initiator (e.go, l0O-g initiator/50-ml benzene), the autoclave was sealed and then flushed-first, three times with nitrogen (charged to 2000 psi), and finally with ethylene (charged to 900 psi)o Next, the autoclave was charged with ethylene and stirred until a constant pressure of between 900 and 950 psi was observed. After charging, the autoclave was placed in an oil bath regulated to ~ 1~C at the temperature desired, equilibrium was usually reached within one hour o Ethylene was bled off as required to maintain an initial operating pressure of 2000 (+~ 50) psi. For example, at 2000 psi and 60~C, the autoclave was found to contain approximately 80 g of ethylene. After the warm-up period, the autoclave (in the oil bath) was exposed to gamma. radiation at nominal dose rates in air* ranging from 27,000 to 33,000 rep/hro It should be noted that the time of irradiation was necessarily less than the time in the bath (Tables III, IV, and V) because of interruptions of the irradiation to remove other samples. When desired, the autoclave was removed from the bath and quenched with cold water; unreacted ethylene was released and the polymer precipitated by the addition of methanolo Blank runs were made in' a similar way in the absence of radiationo The extent of polymerization of ethylene by the direct action of *Because of absorption of radiation by the bath and autoclave, the actual dose rates would be lower (possibly 1/3 lower), but would, of course, still be relative. 3

The University of Michigan * Engineering Research Institute radiation was found to be negligible. The yields of polymers did not always correspond to the decrease in pressure during reaction. The probable explanation is that low molecular weight products, the formation of which would be reflected in a pressure decrease, were not recovered by the precipitation technique used (see also results and discussion)o RESULTS AND DISCUSSION 1o Polymerization of Styrene To obtain general information about the effects of radiation on different initiators, tIe polymerization of styrene in the presence of the following initiators was studied in the presence and absence of gamma radiation: di-t-butyl peroxide, dilauroyl peroxide, benzoyly peroxide, dicumyl peroxide, azobisisobutyronitrile, diisopropyl azoisobutyrate,.cyclohexanone peroxide, isopropylbenzene hydroperoxide, and cumene:ydroperoxideo The degree of conversion was plotted against time for both radiation and blank experiments, and initial rates of..conversion determined. As may be seen from Table I, the rate of polymerization was invariably increased by the rate of' radiation, the magnitude of the effect varied over a wide range (for di-t-butyl peroxide, the initial rate with.radiation was about 3 times the initial blank rate, while for cumene hydroperoxide, the ratio of the two rates was only.ol)o The effects of reaction variables, such as dose rate, initiator concentration, and temperature, were also examined. Results are given in Table IIo If the rate of initiation is increased by the absorption of radiation, the rate of polymerization should be proportional to the square root of the dose rate. This expectation is confirmed by the results (Table II-a). This proportionality means in turn that the higher the dose rate the lower the efficiency of polymerization per repo It may also be seen (Table II-b) that the percent conversion increases regularly with the peroxide concentrations it may be shown,..too, that the rate of polymerization varies approximately with the square root of the peroxide concentrat on.. Table II-c shows that the lower the temperature, the more pronounced is the effect of radiation. This result is consistent with the assumption that the effect of radiation should be independent of temperature. As the temperature is increased, this constant effect is masked more and more by the increasing contribution of thermal polymerizationo Thus the effects of gamma radiation in these experiments are quite straightforward o p4.....

The University of Michigan * Engineering Research Institute TABLE I EFFECT OF GAMMA RADIATION ON THE POLYMERIZATION OF STYRENE IN SOLUTION Initial Rate Temp, Ratio of InitiatorInitial Slopes a of Conversion, C Initial Slopes %/hr Di-t-butyl peroxide 77 2.87b 3 2.50C Dicumyl peroxide 77 2.00 2 Dilauroyl peroxide 77 1.97b 23 1.84C 113 1.40 Diis.opropylbenzene hydroperoxide 77 1.55 3 Diisopropyl azoisobutyrate 77 1.35 20 Benzoyl peroxide 77 1.29 9 Azobisisobutyronitrile 77 1.28 20 Cyclohexanone peroxide 77 1.10 8 Cumene hydroperoxide 77 1.08 4 aThat is, the ratio of the initial slope of the conversion curve to the initial slope for a blank run. Unless specified otherwise, dose rate was 7.0 x 10 rep/hr. Dose rate: 5.8 x 104 rep/hr. Dose rate: 2.4 x 104 rep/hr.

The University of Michigan * Engineering Research Institute TABLE II EFFECT OF REACTION VARIABLES ON THE SENSITIZED POLYMERIZATION OF STYRENE USING GAMMA RADIATIONa (a) Effect of Dose Rate at 25~C ae1 /2 Conversion Rate of Dose Rates (Dose Rate 2 ) 4 se Rte, (De R /2 of Styrene, Conversion, rep/hr (rep/hr) _/ _% /hr..... "1.03 0.25 7.0 x 103 84 1.82 0.46 5.0 x 104 172 2.39 0.60 4.2 x 10 205 2.38 0.60 1.2 x 105 347 3.48 0.87 (b) Effect of Initiator Concentration at 77~C Concentration C onvers ion Dose Rate Initiator, Doe Re, of Styrene g/25-ml soln rep/hr 0.050 4.2 x 104 23.9 0. 050 20.3 0.100 4.2 x 10 30.1 0.100 --- 27.0 4 0.200 4.2 x 10 42.5 0.200 --- 36.6 0.400 4.2 x 104 52.6 0.400 — 45.5 (c) Effect of Temperature'emp, Conversion Temp, Dose Rate, ConversionRatio of |~C ~ rep/hr of Styrene, Initial Slopes 25 4.2 x 10 2.38 2.31 - 1.5035 60 4.2 xo10 20.5 1.64 12.5 77 4.2 x 10 42.5 1.16 36.6 113 4.2 x 10 54.0 1.20 ______ —^___ 44.8 aInitiator: Dilauroyl peroxide; time: 4 hr. -------- 6

The University of Michigan * Engineering Research Institute 2, Polymerization of Ethylene It was originally intended to study the effect of gamma radiation on the continuous polymerization of ethylene at high pressures (say, 20,000 psi) using peroxides or azo compounds as initiators or sensitizers. However, both the cost and the time required for installation and trial operation would have been unduly great in view of the limited time available. Since on the other hand, an autoclave was already available, and since experimental problems are fewer and less difficult at lower pressures, it was decided that more basic information could be obtained in a short time with the autoclave than with the high-pressure unit. Accordingly, plans were changed to comprise batch experiments at relatively low pressures (say, 2000 psi). Studies of the following initiators, both in the presence and absence of radiation, were completed: dilauroyl peroxide (at 60~C); dicumyl peroxide, di-t-butyl peroxide, and diisopropyl azoisobutyrate (each at.80 and 100~C). In addition, blank experiments were made with dilauryl peroxide at 80 and 100~C and for dicumyl peroxide, di-ttbutyl peroxide, and diisopropyl azoisobutyrate at 60~Co Generally the products obtained from the solution in the autoclave were whitish waxy granuleso (However, with dilauroylperoxide at 100~C, a solid mass was obtained rather than a liquid benzene solution. ) It is believed that these products are usually similar; molecular weights calculated (Refo 16, p. 310) from the melting points (118-127~C) would be of the order of 103 (if the product is essentially an n-paraffin) or 104-105 (if it is a high molecular weight polyethylene)o a. Dilauroyl Peroxide —The data summarized in Table III show that, in general, the yields of solid polymer obtained at 60~C in the presence of radiation were greater than the yields for the unirradiated blanks. Melting points of the products were similarO The data also show that the precision of the results is fairly good for experiments of this nature (see Fig. 1); the deviation of a point from the "best" line is usually within + 20 percent. If an induction period of 30,minutes is assumed (that is half the time required to raise the temperature of the autoclave from 20 to 600C);a ratio for slopes of initial yield curves (irradiated/unirradiated) of 1.45 may be calculated. Since the dose rates in these experiments were similar, effects of dose rate were disregarded in this calculation. From the data for blanks at 60, 80, and 100~C, an approximate activation energy of 18,000 kcal/mole was calculated. From these data it was found that an initial slope ratio of 1045 correspondsto a temperature differenc if rates of polymerization are equal, of about 3~C, In other words, the initial rate of polymerization in the presence of radiation at 60~C corresponds to the initial rate of polymerization in the absence of radiation at about 63~Co If the contribution of the blank is excluded, the G value would be 2O3 moles of ethylene reacted/1000 rep/1000 g of ethylene charged. ---------------------- 7 ---------— 7

The University of Michigan * Engineering Research Institute TABLE III THE EFFECT OF GAMMA RADIATION ON THE POLYMERIZATION OF ETHYLENE IN THE PRESENCE OF DILAUROYL PEROXIDE Initial pressure~ 2000 psi Solvent: 50-ml benzene Reactor volume: 275 ml Initiator concentration: 0.10 M Run Temp, Time Dose Rate, Dose, Yield, Melting Final No. ~C in Bath, rep/hr x 10-3 rep x 10 Point Pressure, hr_ _ pC P___:-1-...psi 36 60 1.0 - 0.18 - 2060 38 60 2.0 0 0.47 - 2060 45 60 2.5 0 0.56 - 1980 59 60 2.5 - 0 0.44 - 2020 35 60 3.5 -- 0.67 - 2050 40 60 4.5 0 1.24 125-127 1620 46 6 5.8 -- 0 1.35 - 1850 44 60 7.0 -- 0 1.68 - 1820 33 60 3.0 32 54 1.04 -2250 43 60 4.2 27 81 1.44 - 1850 68 60 6.0 33 106 1.76 125-126 184o 56 81 1.0 0 0.05 - 1990 58 81 1.5 0.08 - 2005 57 81 2.0 0 1.35 1960 55 81 3.0 -- 0 2.90 - 1940 59 81 4.5 - 0 5.25 122-124 1850 67 102 0,75 0 3.0 118-120 1990 66 102 1.5 0 5.38 1970 ----------------------- 8 -------------------

The University of Michigan * Engineering Research Institute 2.0 1.5 |~ B i.^/ Initiator: Dilauroyl peroxide C y / 0 T:60~C; P:2000psi u r - With radiation?;~~~~ t.~|^n~~~~~~~~ >0- O-Without radiation 1.0 0.5 0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 TIME, HR Figo 1 Effect of gamma radiation on the polymerization of ethylene in the presence of dilauroyl peroxide, This effect does not seem large, especially in view of the assumption that radiation effects would be most noticeable at low temperatures. This assumption was based in turn on the assumption that the effects of radiation are independent of temperature. However, the possibility that radiation effects are not always independent of temperature has been suggested recentlyo16,17 Thus the possibility that a better (higher) temperature could be found should not be excluded without further studyo b. Azo Initiators o-Data for the azo initiators are given in Table IV. Products were white granular solids having similar melting points. It is clear that the rate of conversion of ethylene to solids at 80~C in the presence of azobisisobutyronitrile is not increased by exposure to radiation. In fact, the rates are invariably lower in the presence of radiation, and the higher the dose rate, the yloer theyareey Now the rate of decrease of pressure was essentially the same whether or not radiation was used. Thus this curious behavior must reflect an effect of radiation resulting in a lower degree of polymerization, and hence in a greater solubility; this lowering effect should be greater the higher the dose rate. At 100~C, an enhancement of the rate of conversion may exist,.9

The University of Michigan * Engineering Research Institute cii ^S*H 000000000000 * CD P 0 0 0 0 0 0 0 0 0 0 0 0LL\ L\O — OO h c o ( o o \ co c - O O - p d - -- r-l r-l r- r-lr-r- r-l- l r-l ( I b. - - -}-'. -- Lr 0 0O ~ -PN L ( C LI CI C0 CJ COJ ~: r~ ~ C)r —i r-i r4 r-4 r — r-4 r-4 r- 0~ 0J OJoJ J c-th c o oP J C C\J0 C\J C\i r-4 h r- P1 1 1 1 r 1 r-1 r1 rr- r-4 LU t LD C~ U CC CC L\ OOOO CO CO >a~ ~~~ O ~ -a-t ( I 0cr-4 ( LPC) \ \,Or- - \,O O Lr\K IMl -P (D... H0 0 0 C Fzl 0 * ^Or- OJK O^9LP\ 0 90.. o H H U2Q L. r'CO l > prq P*H O X aCO KOL\ 9 L5) s >.H ~ H a p 0 4 Hn lH \ o r- - rO - r- r —I r — H r-I - C\J v \ S \O r z SH H I P9O O:q 0 X 0 4D 4 H H' C <; -1 *rlr-i.. POrM~r ^ -P c c 0H ~- CO L O p= *~ I: V I~~~~~~~OOr- O O I I I I i O oo co co c c~ r~ co c 0 0U cH H HN OJ r- H C\j Vx. 0 # N.0 5 -H ~H ~ Cd P4 - -P (D ~ r -— I I, O - >- - RI- 1 CO 0i r | H0 4*H 0Q, 0 0r 00000000000 (D N tr ~ Q o cE CO CO \CO C0 0 0 CO oO 0 Sri h~ —.d E rd r — r — r-! r-4 -r Z -4 r- -Pm H ciO *H *H J....CO-.. HCD~ H B( ) CDj 1 Epl 0 Dq rq 1 O r)LT (D o OOC hC O CCOO h9 OCO- h -c CO Oh

The University of Michigan * Engineering Research Institute although a definite conclusion is impossible owing to the scarcity of data~* If we accept the few data at face value., the ratio of conversion rates (irradiated/nonirradiated) would be only 1.o2 On the other hand, the.rate of decrease in pressure was not consistently greater for the radiation experiments. In this case, the effect of dose rate noted at 80~C does not exist. Results of the few experiments made with diisopropyl azoisobutyrate, as with the azobisisobutyronitrile at 80~C, indicate that the use of radiation at 80~C results in a lower rate of conversion of ethylene to solid polymers than is observed for a blank runo -An explanation similar to that suggested for the azobisisobutyronitrile experiments may also be applicable in this caseo c. Other Peroxide Initiators — Data'for these initiators are present,in Table V. Again, products were white granular solids having similar'melting points o With di-t.-butyl peroxide solid products were not obtained at 8o0~C although the pressure decreased measurably during the reaction. Small yields were, however, obtained at 100~C. Just as was obseryed in the previous section, the rate of conversion to solid polymer was lowered by irradiation; on the other hand, irradiation may have resulted in- an increase in the rate of decrease of pressure A similar situation exists with dicumyl peroxide. At 80~C, a decrease in pressure was observed but no solid polymer was obtained. At' 00~C, solid polymers were obtained, but with a lower rate of conversion in the radiation run. d.o General Comments on the Sensitized Polymerizations of Ethylene.So far, the only clear-cut case in which irradiation, improves' the rate of conversion of ethylene to solid polymers under the experimental conditions used here is that'of dilauroyl peroxide. In the other cases, irradiation apparently usually results in the formation of lower molecular weight, and hence more soluble, polymers. Further and more accurate pressure measurements would be required to determine unequivocally whether or not irradiation in these latter cases results in an increased rate of consumption of ethylene. The fact that some of the products are soluble in methanol suggests that the products are probably high molecular weight waxes rather than regular polyethylenes o It had been hoped that results of these low-pressure studies could be extrapolated to the conventional high-pressure polymerization of ethylene. The conclusion that the higher the dose rate, the lower the degree of polymerization -should certainly be applicable to a conventional polymerization. However, a final answer regarding the effect of radiation on a conventional polymeriza*As with dilauroyl peroxide at 100~C, the polymerization resulted in a solid mass rather than a solution. -- 11j -— 1

The University of Michigan * Engineering Research Institute ON CT5?H0 0 0 0 0 0 0 0 0 CH Od I.'r O O O O O O O O O ~r4 cn ~ co Lr~ 0% O cO o 0C o\ F D C I - r-l r-clr-l - r-l r-l r r P-0 0 l - Q) O4O O OI PEM v0 O" I cO O O O 0 r- 0 o t O I I IH (U 0.0 o o oI Z. 0 Hi H H~~~~0~~~~~~~~~~~~~~' H 4, CO.4 H iiH t4 ) LCN ~ ~ I.~O A' o O00 C1 *I I I i O * I HI 0 O O I I 0, n O I OC I Ei i,-4 4C o- 0 A ^c\ 0 0 0r8 0 pq o oj (l ir4 H ~ E ~ r-I ( -L) ( Ii! 1 i (L i i. H dT Q)! CD P E4' * Q. P_ 0 0 P0 a ) H U r)0 I I 01 I( r O H O O -O HQ Er4Q r' i O' 01 0 0 0 0lN 0 0 0 C1 He vo.. Q O O O 0 O O O o0 *H rc$ (L 4 r4 0 f [x3N in -" )o 00 CO 00 0 00000 0 0 0 03 H HE-r ri cS rH H H Hl 4 o~ o ~ O C ~ Pl P. P4 Pco 4 t - P P P4 Poo H r0 4' H H Hq rC-( 4lO ^ 4' rl *HI' o a,?o t >- i- KO r>- co cO KO i- 0o cd

The University of Michigan * Engineering Research Institute tion-system may properly require more experiments at higher pressures which would result in a polymer more typical of polyethyleneo e. Pre-irradiation of Initiators — Some experiments with the irradiation of frozen solutions of dilauroyl peroxide indicated the presence of free radical activity (as measured by the consumption of the stable free radical, diphenyl picrylhydrazyl) after thawingo To see if it would be possible to irradiate an initiator for polyethylene rather than the reactor itself, 3 gallons of a suspension of dilauroyl peroxide were irradiated (dose, 1.6 x 106 rep) while frozen, and 3 gallons of a similar suspension were frozen as a blank These samples of initiator were sent to the pilot plant of the Spencer Chemical Company for evaluation. However, the efficiency of the irradiated initiator when thawed proved to be no higher than the efficiency of the unirradiated initiator suspension~ The basic problem seems to lie in getting the frozen free radicals in contact with the monomer before recombination.with one another can occur. So far, then, it seems unlikely that the irradiation of initiators prior to polymerization is practicalo CONCLUSIONS.o The rate of polymerization of styrene (in solution) in the presence of peroxide or azo compounds is increased by gamma radiation. Ratios of initial conversion slopes (irradiated/unirradiated) range from 2.9 with di.-t~butyl peroxide to lol with cumene hydroperoxide (both at 77~C). Effects of initiator concentration, dose rate, and temperature were as expected, at least with dilauroyl peroxide. 2. The rate of polymerization of ethylene (to form solid products) at 2000 psi and 60~C in the presence of dilauroyl peroxide may be increased 1.4-fold by gamma irradiation. Products obtained, white waxy solids, are similar to products obtained in the absence of radiation. 3o The rate of conversion of ethylene to solid products at 2000 psi and 80~C in the presence of azobisisobutyronitrile is decreased by gamma irradiation, probably because of a decrease in the molecular weight, and hence an increase in solubility of the product in the precipitant used, caused by irradiationo At 100~C there may be a slight increase in the rate of conversion due to irradiationo 4, The rate of conversion of ethylene to solid products at 2000 psi in the presence of diisopropyl azoisobutyrate or dicumyl peroxide at 800C, or di-t-butyl peroxide at 100~C is decreased by gamma irradiation, presumably because of a decrease in the molecular weight of the product. 5o The irradiation of an initiator solution prior to polymerization.appears to be impractical, unless some way can be found to prevent recombination of any free radicals formedo 15

The University of Michigan * Engineering Research Institute II.. EFFECTS OF GAMMA RADIATION ON POLYETHYLENE A.o Crosslinking of Polyethylene INTRODUCTION It is well established that the irradiation of polyethylene results in several reactions, even though the mechanisms of these reactions are not yet clear (Ref. 16, Chapter 40o7 and Refs., 17-20). Considerable evidence shows that the irradiation of polymers, for example, polyethylene, results An crosslinking, chain scission, and the development of unsaturation; 5'212 with a crystalline polymer like polyethylene, a loss of crystallinity also occurs. The effects of these reactions on physical properties have been examined in some detail in several laboratorieso As the radiation dose in increased, the density decreases at first owing to a decrease in crystallinity; at higher doses, when much of the crystallinity has been destroyed, the density increases due tO tighter packing (Refo 16, po 151)o A similar decrease followed by an increase has been noted for Youngts modulus (Ref 16, po 151).o Lawton, Bueche, and Balwit2 found a maximum in the curve relating the tensile strength of irradiated polyethylene to dose; they also reported a maximum in the curve relating the ultimate elongation to dose. On the other hand, Dole., Keeling, and Rose25 found little change in tensile strength on irradiation as long as oxygen was absent; Ballantine and others28 also reported almost negligible initial increases in tensile strength on irradiation (and also a steady, consistent decrease in ultimate elongation as the dose was increased). Other properties are also affecteds for example, it is claimed that several commercial products prepared by irradiating polyethylene have improved resistance to heat, chemicals, and stress-cracklingg 7' Charlesby also claims a similar improvement in resistance to heat and solventso29 Although fabrication of these irradiated products is difficult, an improvement in this respect has recently been announcedo30 It seemed logical, therefore, to investigate the effects of irradiation-on samples of typical kinds of polyethylene produced by the Spencer Chemical Company, to see if the physical properties could be improved by irradiation without a corresponding decrease in ease of processinga Results of this study are presented below. EXPERIMENTAL The following samples of polyethylene were supplied by the Spencer Chemical Company~ Poly-Eth 1005 (PE 1005), Nat B, lot 42, melt index 1.8; PolyEth 1008.5 (PE 1008.5), Nat B, melt index 20. 14

The University of Michigan * Engineering Research Institute Irradiations were conducted using the 1-kilocurie Co-60 source at the Fission Products Laboratory. Polymer samples were irradiated, under an atmosphere of nitrogen, in sealed glass tubes. Samples for tensile measurements were prepared by compression molding at 300~F and 2500 psi. Tensile measurements were made using an Instron tester. Cross-head speeds of either 5 or 10 in./min were used, tests revealed no measurable difference between tensile properties determined at either speed. RESULTS AND DISCUSSION In the first experiment, a sample of Poly-Eth 1005 (melt index 1.8) was exposed at room temperature to doses ranging between 1 x 105 and 2 x 107 repo It was found30 that a plot of the tensile strength as a function of dose passed through a maximum at a dose of 1o5 x 106 rep; however, any.dose as high or higher than this value resulted in values for melt index sufficiently low to preclude satisfactory extrusion. The softening point was improved by irradiation, but during unmolding tests the molded irradiated samples tended to revert to the original pellets, Extrusion of the material was difficult. To see if a better balance between crosslinking and processability could be achieved, the effect of temperature of irradiation was investigated. Samples of Poly-Eth 1005 were irradiated at -70~C (that is, below the glass temperature) and at 120~C (that is, above the crystalline melting point). Experiments were also conducted with Poly-Eth 1008.5 (melt index 20) with the hope that this polymer might show a lesser tendency to become crosslinked to a degree that could reduce the ease of fabrication. Physical properties, including resistance to heat distortion but excluding tests of stress-cracking, were determined. Similar tests were also made with a sample of Poly-Eth 1005 that had been heated at 120~C but not irradiatedo Samples of the following materials were sent to the Plastics Research. Laboratory for evaluation.o31 Poly-Eth 1005 (l.Ox 105 < dose in rep < 1.5 x 106) and Poly-Eth 1008.5 (1o0 x 105 < dose in rep < 5,0 x 106), Results are summarized below. (a) Poly-Eth 1005 (Irradiated at -70~C) Dose: O - 2 x 107 rep Tensile strengths Increased from 1860 psi at zero dose to a maximum of 24350 psi at a dose of 1.o5 x 106 rep (see Fig. 2)o Yield strength: Increased from 1180 psi at zero dose to a maximum of 1330 psi at a dose of lo5 x 106 rep (see Fig. 2). Percent elongation: Decreased from 570% at zero dose to 540%. at a dose of 1o5 x 106 rep, and rapidly thereafter (see Figo 2)o 15

The University of Michigan * Engineering Research Institute 600 3000C \ 0- Tensile strength 500 \ O.- Elongation 2600 \ A -Yield point / 4- \ U 2200 \ O s' b"\ \ z 5 \ -- -3000 l 1800 \ \ C3 (/\ \ \ z J V. — 200" z 1400 10-iS~ Q^00s~' -2-00 A,0~k —~~ A'"'_ -0..-. -1.... 10000 I! I I I I I I I I 0 2 4 6 8 10 12 14 16 18 20 RADIATION DOSE, REP(x10-6) Figo 2. Effect of gamma radiation on the tensile strength of PolyEth 1005 (irradiated at -70~C)o Melt index: Decreasedg 1o4 at a dose of 1.0 x 105 rep, and 0 at a dose of lo5 x 106 rep. Heat distortions* Samples that had been given doses greater than lo5 x 105 rep tended to curl and break apart. Remarks: Irradiated samples more transparent than control sheet, but not uniformly so. (b) Poly-Eth 1005 (Irradiated at 120~o) Dose.0 - 8.0 x 106 rep Tensile strength. Increased to 1900 psi at a dose of o.5 x 106, and decreased thereafter. Yield strength: Increased slightly to a maximum of 1200 psi at a dose of 350 x 106 rep, and decreased thereaftero *~In tests reported in this section, samples were held at 220~F for 30 minutes. 16

The University of Michigan * Engineering Research Institute Percent elongation: Decreased steadily; 400% at a dose of 1.5 x 106 rep. Melt index: Decreased to 0o8 at a dose of 1o0 x 105 rep, and to 0 at a dose of 1.5 x 1060 Heat distortion: Samples which had been given doses greater than 1 x 106 rep tended to curl and break apart. Remarks. Irradiated samples adhered to walls of glass vessel more than did the unirradiated one. Thus exposure of Poly-Eth 1005 to a low dose resulted in an initial increase in tensile strength and yield strength, especially when the irradiation is conducted at a low temperatureo The overall resistance to heat distortion was somewhat improved in comparison with unirradiated samples. However, the amount of radiation required to reach the maximum in tensile strength is large enough to lower the melt index to a level that makes the material difficult to fabricate. (c) Poly-Eth 1008.5 (irradiated at 10~C) Dose: 0 - 5o0 x 106 rep Tensile strength: Increased from 1430 psi at zero dose to 1550 psi at a dose of 350 x 106 repo Yield strengths Increased from 920 psi at zero dose to 1040 psi at 3.0 x 106 repo Percent elongation: Decreased from 450% at zero dose to 320% at a dose of 5,0 x 106 repo Melt index: DecreasedO 0.14 at a dose of 1.5 x 106 rep, and 0 at a dose of 5.0 x 106 repo Heat distortion: Samples that had been given doses greater than 1.5 x 106 rep tended to curlo Also, a large sample of Poly-Eth 1008o5 was irradiated at -10~C (dose is 7.5 x 105 rep) and tested for processability at the Plastics Research Laboratory?2 The melt index after irradiation was 2.3o Although extrusion on film was unsatisfactory, injection molding behavior was satisfactory. It was also observed that the irradiated polymer was more transparent that the control usedo Other physical properties were similar, though slightly inferior, to the sample of Poly-Eth 1005 used as controlo (d) Poly-Eth 1008o5 (irradiated at -70~C) Dose: 0 - 1o5 x 106 repo ------ 17

The University of Michigan * Engineering Research Institute Tensile strengths Essentially unchangedo Yield strengtho Decreased slightly from 1090 psi at zero dose to 1050 psi at a dose of lo5 x 106 -repo Percent elongations Essentially unchanged~ Melt index: Decreased from 20 at zero dose to 16.o7 at a dose of 350 x 105 rep to 0 at a dose of 1o5 x l06 repo Heat distortiono Samples that had been given doses greater than lo0 x 106 rep tended to curlo The results show that the melt index of Poly-Eth 1008.5 may be reduced at will by irradiation. However, as with Poly-Eth 1005, fabrication of the final product is difficulto General Remarks,.-So far, at least with Poly-Eth 1005 and Poly-Eth 1008o5, it has not been possible to introduce enough crosslinking to improve strength properties without affecting at least some fabrication properties adverselyo Nevertheless, tensile properties of the large sample prepared by the irradiation of Poly-Eth 1008o5 were almost as good as those of the control sample of PE 1005o While the mere duplication of Poly-Eth 1005 is not in itself of interest, it is possible that some properties not determined, such as stress-cracking or resistance to reagents might be better than the corresponding properties for Poly-Eth 1005 It is also possible that other samples of polyethylene might behave differently; for example, control of the molecular weight distribution is probably important. It. seems likely from tests reported in a later section that Poly-Eth 1005 contains a small amount of a very high molecular weight tail, which, of course, forms gel at a very low dose; Poly-Eth 1008o5 may also have a broad distribution. Small amounts of gel formed from a high molecular weight tail could very easily cause difficulties in fabricationo The degree of branching may also be an important variable. Although it is claimed that one product m.y be extruded satisfactorily,7 the commercial irradiated polymers mentioned above must, in general, be irradiated after fabricationo Hence the problem of achieving a suitable balance between crosslinking and ease of falbrication has still not been solved CONCLUSIONS o1 The tensile and yield strengths'of Poly-Eth 1005 may be improved by gamma irradiation (in the absence of air) with a dose of about 1o5 x 106 rep, especially at low temperature:s the resistance to heat distortion may also be improved in this way. However, the dose required to attain the maximum strength lowers the melt index to a level that results in fabrication difficultieso 18

The University of Michigan * Engineering Research Institute 2o The tensile and yield strengths of Poly-Eth 1008.5 may also be improved by gamma irradiation (in the absence of air) at 10~C, but to a lesser extent than is the case with Poly-Eth 1005; the resistance to heat distortion is also slightly improvedo The melt index may be reduced at will by varying the dose received. Again, however, fabrication becomesdifficult; for example, the physical properties of a sample of irradiated Poly-Eth 1008.5 with a melt index of 2o5 were only slightly inferior to the properties of Poly-Eth 1005, but extrusion as a film was not satisfactory. 35 It is difficult to obtain a satisfactory balance of properties by irradiation of the two samples studied. Possibly better results might be achieved if a polymer having a narrow molecular weight distribution could be used, irradiation of such a polymer should result in less difficulty in fabrication caused by small amounts of galo Bo Modification of Crosslinking in Polyethylene INTRODUCTION When the difficulty of achieving a suitable balance between a desirable degree of crosslinking in polyethylene and ease of fabrication became apparent attention was turned to the possibility of modifying or controlling the crosslinking due to irradiationo If, for example, the use of an additive in polyethylene altered the ratio of crosslinking to degradation, the properties of irradiated polyethylene might be different. Some information is already available about the effects of additives to polymers on the behavior under irradiation. Charlesby, Alexander, and others33-35 have done considerable work on the protection of polymers against radiation damage by adding free radical acceptors such as aniline to the polymer before irradiation. Meikle and'Graham36 claim that the use of unspecified additives to polyethylene improves the characteristics (especially at high temperatures) of the irradiated products the process concerned is now being licensed for commercial use ("Hyrad" process)o The addition of carbon black to polyethylene before irradiation has been claimed to be advantageous937538 a commercial application of the use of carbon has also been madeo59 Although the mechanism of the action is not firmly established, it seems possible that the carbon particles act as "traps" for polymeric free radicals and bind together several polymer molecules, the- new bonds formed.may thus increase mechanical strength.38 40 Another effect of additives may be to combine with residual oxygen and thus inhibit degradative processes otherwise catalyzed by oxygen.534 Because of this interest in the modification of reactions induced in high polymers by irradiation, a preliminary investigation was begun. Several experiments were therefore made in which varying amounts of aniline or carbon 19.

The University of Michigan * Engineering Research Institute black were milled into Poly-Eth 1008.5 before irradiation. Since styrene, like aniline and carbon, should be a good free radical acceptor, the irradiation of polystyrene-containing polyethylene [in the form of a polyethylene-styrene graft copolymer, (Section C, below)]was also completed. In addition, since it had been observed previously (Section A) that the adhesion of polyethylene to glass was increased by irradiation, several experiments were made with the incorporation of finely divided glass, or "flake" glass, in Poly-Eth 1008.5 before irradiation. Results of these experiments are presented and discussed in this section. Now the incorporation of all these additives or fillers requires milling, which,of course, decreases the strength properties of polyethylene (see Tables VI-VIII)o The incorporation of a modifier in such a manner as to retain the inherent structure of the polyethylene seemed especially desirable. One way of accomplishing this purpose would be to saturate (to some degree) the polyethylene with a gaseous modifier before irradiation. Experiments were accordingly conducted with the irradiation of Poly-Eth 1008.5 that had been previously more or less saturated with vinyl chloride, which, as well as its polymer, should act as an efficient chain transfer agent for reactions of polymeric radicals. At.the same time, the formation of polyvinyl chloride should be limited, for polyvinyl chloride is degraded by radiation. Because some information was already' available about the reactions of monomers with polyethylene film (see Section C), the first experiments were made with films. or sheets; a few experiments were made with pelletso The results of this work are reported below. Similar experiments were made using vapors of the following two groups of monomers: vinyl acetate, and isobutylene —the polymers of which are degraded by irradiation; butadiene and propylene-the polymers of which.are, or might be expected to be, crosslinked by irradiation. EXPERIIENTAL Pellets, 2-mil film, and 20-mil sheets of Poly-Eth 1008.5 were supplied by the Spencer Chemical Company. Styrene (99%) was obtained from the Dow Chemical Company, aniline (reagent grade) from Merck and Company, and vinyl chloride (99.8%) from the Matheson Company. The carbon black used —HCR, SRF, and Thermax-was obtained from the Binney and Smith Company. Flake glass was supplied! by the Owens-Corning Fiberglas Corporation. The styrene-containing polymer was sample 7, Table XVIIIo *Although the reaction of the vinyl chloride and other monomers with polyethylene does, in one sense, constitute grafting, the purpose of these experiments was not to make a graft polymer er se (polymer A on polymer B). but merely to modify the structure of polyethylene.by use of a chain transfer agento For this reason, these experiments are discussed here, rather than in Section C, which deals with graft copolymerso 20

The University of Michigan * Engineering Research Institute Additives were incorporated in the polyethylene by milling at 240~F for 5 minutes; a sample of polyethylene was milled in a similar manner as a blanks The milled samples were sealed into evacuated tubes for.irradiation at room temperature by the 3-kilocurie Co-60 source at the Phoenix Laboratory (dose rate: about 3 x 105 rep/hr)0 Tensile properties were then determined as in Section B. Adhesive ratings were found by determining the force in grams required to strip a length of Scotch tape from a sheet of polyethylene (the tape having been subjected to a constant weight for a constant time). In the experiments using monomers, the polyethylene samples were first placed in glass tubes and then degassed using a high-vacuum line for varying periods of time before charging with monomer vapor (see tables). Gaseous monomers were charged to approximately atmospheric pressure directly from cylinders after evacuation of the connecting lines; the. other monomers were charged from degassed ampoules of the liquid monomers to the maximum pressure possible. Blank experiments were conducted in a similar way except for the omission of monomers. After irradiations (at room temperature, unless specified otherwise) were complete, the tubes were opened to the atmosphere at once,.and physical properties determined [both parallel to (+), and transverse to (-), the machine direction, in the case of films or sheets]. Tests showed that the tensile properties of samples cut from different places on the sheet roll did not vary significantly. It was also found that the tensile properties of monomertreated film were unchanged by degassing residual monomer. Settings of the Instron are given in relevant sections below. Tensile measurements were usually made with 8 or 10 samples, and the values averaged. Tests of heat distortion -were macLe by exposing samples cut from the sheets or films and placed on asbestos board to the desired temperature in a circulating ovenr tests showed the heating of the asbestos board to be uniform. Infrared measurements were made using a Baird recording spectrophotometer, model B. RESULTS AND DISCUSSION 1, Solid and Liquid Additives In Table VI, results are presented for the experiments with the incorporation of. styrene,aniline, and carbon black in polyethylene before irradil ation; more detailed results for several types of carbon black are given in Table VII. The results show that the.presence of aniline hindered the initial increase in tensile strength expected on irradiation; in fact, the irradiation led to a slight decrease in tensile strength., especially at the higher concentration of aniline. Since all properties listed are considerably decreased from the original properties, no advantage is given by the addition of aniline, at least under conditions used in this study. Similarly, the incorporation of.21

The University of Michigan * Engineering Research Institute TABLE VI THE EFFECT OF ADDITIVESa ON THE PHYSICAL PROPERTIES OF IRRADIATED POLY-ETH 1008.5 Dose, Tensile Adhesive lon Dose. Elongation, Sample Additive rep x -6 Strength, Rating, rep x 10- -.. -......_..... -.*.. Psi.......g.._.. _. Styrene graftb None -- 160 255 540 Styrene graft None 0.22 1180 146 265 PE-1008.5 2% aniline 0.51 1230 190 240 PE-1008.5 5% aniline 0.51 1200 182 240 PE-1008 5 2% aniline -- 1230 195 230 PE-1008.5 5% aniline -- 1230 182 220 PE-1008.5 2% C black 0.51 1210 261 PE-1008.5 2o2 C black -- 1180 264 PE-1008.5 5% C black 0.51 1190 286 PE-1008.5 5% C black -- 1290 254 PE -1008.5 PE-1008.5 None 0.5 1400 230 535 (Milled) aAdditives milled in with polyethylene, unless specified otherwise. These samples were not milled. 22

The University of Michigan * Engineering Research Institute TABLE VII THE EFFECT OF TYPE OF CARBON BLACKa ON THE PHYSICAL PROPERTIES OF CARBON-FILLED IRRADIATED POLY-ETH 1008.5 Tensile Adhesive Type of Dose, Str h Elongation, Cabo rp 16 _ Strength, Rating, Carbon rep x 10-6 %.......... psi g.. Blankb -- 1230 222 535 Blank 1.0 1500 240 535 Blank 1.75 1575 209 (375) Blank 3.75 1540 213 535 HCR-8 -- 1190 268 65 HCR-8 1.0 1230 268 65 HCR-8 1.75 1350 277 65 HCR-8 3.75 1180 227 65 SRF -- 1110 225 65 SRF 1.0 -- 250 -- SRF 1.75 1250 227 65 SRF 3.75 1200 222 65 Thermax -- 1060 262 65 Thermax 1.0 1120 182 65 Thermax 1.75 1360 240 65 Thermax 3.75 1220 226 65 aAll samples contained 10 carbon black. All blank samples were milled before testing. 23

The University of Michigan * Engineering Research Institute styrene in the form of a polyethylene-styrene graft copolymer resulted in a marked lowering of all properties measured after irradiation.. Although the use of small amounts of carbon black resulted in a decrease in tensile strength in both the blank and irradiated samples, the adhesive rating of the sample containing 5% carbon was increased by irradiation (Table VI). As might be expected, the effect of carbon black depends on the type used (Table VII)o Although the addition of 10% carbon black lowered the overall tensile strength and percent elongation, tensile strengths at a dose of 1.75 x 106 rep were all higher than the tensile strength of unfilled, unirradiated polyethylene'the relative increase of tensile strength was greatest for the samples containing Thermax carbon black. Adhesive ratings were in most cases higher than for the unfilled unirradiated polyethylene. Thus the irradiation of carbon-filled polyethylene may to some extent offset the loss in tensile strength incurred by the addition of the filler. Possibly the reason for the failure to observe dramatic effects with the use of. additives lies in the fact that irradiations were conducted in the absence of oxygen; also, tests. showed no evidence for post-irradiation degradation that could result from the presence of residual'oxygen. This explanation is consistent with the observations of Alexander and Toms reported above.34 Results for the glass-filled polyethylene are given in Table VIII. The presence of the glass resulted in a considerable lowering of the percent elongation, but little change in the tensile strength. Irradiation with a fairly low dose (4.8 x 105 rep) resulted in substantial increases in percent elongation. Indeed, the perdent elongation for a sample containing 5% glass was increased 15-fold to the level characteristic of unfilled polyethylene. In other words, irradiation of a glass-filled polyethylene results in a substantial increase in toughnesso 2. Vinyl Monomers as Modifiers ao Polyethylene Sheets -Results from preliminary experiments with the irradiation of 20-mil polyethylene sheets are given in Table IXo A striking increase in both tensile strength and percent elongation may be noted for the treated irradiated sheets as compared with the sheets that-had been merely irradiated under vacuum. In other words, treatment with vinyl chloride results in increased.toughness. Both tensile strength and percent elongation reached a maximum at a dose of 8.5 x 105 rep; at this dose, the gain in weight of the film was only 2.ol percent. The incorporation of this small amount of vinyl chloride also affected the permeability slightly. After immersion in benzene for 48 hours, a typical sheet (Noo 5) gained 10.6% in weight, as compared to 11o3% for irradiated sheet (having the same dose),.and 9.1% for a plain sheet. Infrared measurements revealed no major differences between tie treated and untreated filmo 24

The University of Michigan * Engineering Research Institute TABLE VIII THE EFFECT OF FLAKE GLASS ON PHYSICAL PROPERTIES OF IRRADIATED POLY-ETH 1008.5 Flake Tensile Dose, Elongation, Glass, Strength rep x 10-5 Stength,'% L psi 1.0 4.8 1145 280 5.0 - 1220 20 5.0 4.8 1100 330 10o0 - 1230 70 10.0 4.8 1130 165 15.0 - 1200 65 15.0 4.8 1100 185 0,0 5.5 1190 330 TABLE IX THE EFFECT OF ABSORBED VINYL CHLORIDE ON PHYSICAL PROPERTIES OF IRRADIATED POLYETHYLENE SHEETS (POLY-ETH 1008.5, 20-MIL)a Tensile Adhesive Change in Dose, Elongation, Sample Ds Strength, Rating, Weight, rep x 10-5.b % psi g - -1670 713 2' 3.0 1715 735 - - 7' 4.8 1570 655 5' 8.3 1850 798 175 3' 13.0 1820 720 180 - 4185 1960 760 180 - 2 3.0 1760 710 115 - 7 4.8 2150 825 145 5 8.3 2370 800 165 2.1 3 12.4 2250 820 165 4 18.5 1870 7535 202 0.5 aBefore irradiation the sheet was kept in contact with vinyl chloride for 3 days. bCross-head speed 10 in./min; all samples transverse to the machine direction. 25

The University of Michigan * Engineering Research Institute On the other hand, as may be seen from Table X, treatment with the other monomers* had either little effect (as with vinyl acetate) or resulted in deterioration of the properties measured. Again, changes in weight were very small-of the order of parts per thousand. Thus vinyl chloride appears to be unique. Further experiments with this technique were therefore restricted to the use of vinyl chlorideBefore proceeding with further experiments, tests of heat distortion were made on rectangular samples cut from all sheets studied up to that time. Results are given in Table XI; the appearance of the samples after testing (originally the same size and shape) is shown in Fig. 3. Again, the effect of vinyl chloride is unique;j although use of the other monomers resulted in sheets having a poorer resistance to heat (especially with.respect to shrinkage) than did plain irradiated sheets, the use of vinyl chloride and a dose of (0.8.- 12) x 106 rep resulted in a sheet having remarkable stability with respect to both buckling and shrinkageo This optimum dose corresponds to the dose yielding maximum tensile strength and percent.elongation, and hence maximum toughness, It was then decided, first, to try to determine the experimental conditions leading to heat-resistant sheets,** and second, to study the relationships between the unusual stability and' tensile properties in both the transverse and machine directionso- The experiments with sheets were accordingly repeated; effects of time in contact with the vinyl chloride, time of degassing, and dose were examined in some detail at room temperature. Sinc'e particular attention to elastic modulus was desired, the cross-head speed of the Instron tester was set at a very slow speed, 2 in./min. As before, tests of heat resistance were made' since the conditions of test varied somewhat from case to case, some variations in results may be noted below. Effect of total doses Results of heat tests made on some typical irradiated treated polyethylene sheets are given in.Table XII' see also Figs. 4 and 5. As was the case with the preliminary experiments, the irradiation -of a vinyl chloride-treated sheet with a dose of about 8 x 105 rep (sample 47) resulted in very good resistance both to shrinkage along the machine direction and to buckling' In contrast, irradiation of untreated film with the same dose (sample 47") resulted in rather poor resistance to heat; the sheet flowed and shrank considerably-to about the same extent as an unirradiated sheet. The *In several of these experiments the sheets were irradiated immediately after charging, instead of after a soaking period of 3 days, as with samples described in Table IXo A check was therefore made with a vinyl chloride experiment in which the sheet was irradiated immediately after charging (sample 27). Although the tensile strength was, indeed, lower than the value predicted on the basis of Table IX, the strength was still high. 41 **The grafting of vinylidene chloride to polyethylene film has been describ~ed, but no physical properties were reported; the use of vinyl chloride was not, however, mentioned. 26

The University of Michigan * Engineering Research Institute TABLE X THE EFFECT OF ABSORBED MONOMERS ON PHYSICAL PROPERTIES OF IRRADIATED POLYETHYLENE SHEETS (POLY-ETH 1008.5, 20-MIL) Tensile Change in Dose, Elongation, Sample Monomer o, Strength We ight _____~rep x 10 6 p ~_ 13 Butadieneb 0.27 1425 540 -0.2 13' Butadienea'b 0.44 1650 580 +0.2 14 Butadienea, 0.72 1600 525 +0.2 14' Butadieneab 1.16 1510 490 +0.3 15 Butadieneab 1.61 1420 505 +0.2 Blank None None 1735 510 16 Isobutylene 0.67 1610 440 +0.1 17 Isobutylenec 0.99 1610 435 +0.1 18 Isobutylene 1.49 1710 500 +0.1 19 Vinyl acetatec 0.67 1710 510 0.0 20 Vinyl acetatec 0.99 1730 550 +0.1 21 Vinyl acetatec 1.49 1800 500 0.0 25 Vinyl acetatea 0.96 1490 475 +3.0 29 Propylenea 0.96 1580 510 +0.1 27 Vinyl chloridea 0.96 2100 600 aSheet degassed for 24 hours. bontact time of sheet with monomers before irradiation: 2-1/2 days. CSheet degassed for 4 hours. 27

The University of Michigan * Engineering Research Institute TABLE XI THE EFFECT OF ABSORPTION OF MONOMERS ON THE HEAT RESISTANCE OF IRRADIATED POLYETHYLENE SHEETS (POLY-ETH 1008.5, 20-MIL) Test: 15 min at 130~C......Approximate....Dose........... Ge.ra... G e n... Sample Monomer De6 General Shrinkagea Gradeb Jrep x 10- u Appearance 7 None 0.5 No buckling 25 A51 None 0.8 Buckling 25 A3' None 1.3 Buckling 25 A4' None 1.8 Buckling 25 A2 Vinyl chloride 0.3 A little distortion 10 A 5 Vinyl chloride 0.8c No apparent change 0 A+ 3 Vinyl chloride 1.2 Very little change 5-10 A 4 Vinyl chloride 1l8 Buckling 30 B+ 13 Butadiene 0.3 Soft and distorted 60d B 13' Butadiene 0.4 Soft and buckled 60d B 14' Butadiene 1.2 More buckled 60 B15 Butadiene 1.6 Some buckling 40d B+ 16 Isobutylene 0.4 Buckling 60 A17 Isobutylene 1.0 Buckling 60 B+ 18 Isobutylene 1.5 Buckling 60 A19 Vinyl acetate 0.4 Buckling 60 A20 Vinyl acetate 1.0 Buckling 60 B+ 21 Vinyl acetate 1.5 Buckling 60 Aaln machine direction of sheet. bSee Ref. 31. CCorresponds to maximum ultimate tensile strength. dAlso, shrinkage transverse to machine direction. 28

The Udniversity of M/ichigan Engrine~ering Research Insrtitute — ~~~~~~~~~....................................~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~c —.....................................................I..........~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~*s~~...................................................~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~.......................................11.................................................................................................................................................................................''l'.....................-................................................1................................~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~ j~j~~~i~ii ~ ~ X~ilX~~i:iX............................................................................................................ iiiil................................................................................................................................................................................................................................... 1....................................................................................................................................................................................................:::::::::::::::i:::j:::.....................................................................................................................:\::::::::::::.......................................................................................................................................................................................................................................................................................................................~~~~~~~I ~:- ~~.......................................................................................................................... 11................................................................................................................................................................................................................................................................~~ xxx~~x~~::xx~~:~xxx-................. l~i ~...........................................................................................................................::::~:j::i:::a::::::~: x~.r)............................................................................f C:I:I i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ilx~~~~~~~~~~~~~~~.........................................~ ~ ~ ~ ~ ~ ~~~~~~~~t~ t.......................................................................~............................................ - X x............................................................................~............................................. -......................................... -..................................... i~~~~~~~~.:i~~~~~~~~~~~~i~~~~~i~~~~~iiil~~~~~~~~~~~i))~~~~~~~~~i~~~~~:~~~~...................................................... ~ ~ ~ ~-5~~......................................... —...........~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4... ~-......................~ ~ ~ ~ ~~~~~~~~g ~.............................................. -~~~-...............~~~~~~~~~~tJ............................~~~~~~~~~~~~~~~~~~~~~~~~~F-~~ O..................................:~~:~~:~~: ~~ii~i j~~ i...............~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I:'~:ji:::jjj:~:j~l13........................~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~i~~ii~i~i~ji~Xii -~...................::~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~X.X.:~~~li~ ~Xll~~: l~ l.............~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~lllii~.................................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r............................ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~E ~~:........................ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ i~ l~ i.................................~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~~~~~~~~~~~Illi~i8:~ii~l"~Q -.......................:':.........::..............................................................................

The University of Michigan * Engineering Research Institute hD H a a r - 4 O - C O ) = H H 4-D 4-) rd rd rd r rd rd r rd rdrd rd rd rO O 4 ) C) C) 0 -P o o o o o o o o o 0 o o r o -H ) C H rd ctS cc Fcd -P ed Od ed ed ed edCd ct Cd et 5C c t J d Ic. ( (D (D (U (D <D (U (D? J (D (D 31 \CJ rH r- 1 r- rN-H o COCO co c co C CO Cc coC CO co COCO u2 CJ 40 co O ~ Hri HHH-H HH- HH r- H H r L.n n Lrn Ln n Ln HI w4X a) tECH O < ori oC) [ S ~ W 1t -P- g | n 0C I C'\ > U o C\SO I I C\lr — I Hi (\ J!C.JCM I I I I 0CU 01 I. CO dO Od) H 0 E Ei PE ) OI C H t t -I- - I! UI 1 U) 0- W r~ H. I P cr 00O O 0 0n 0 O Lf0 0 0 0 0 1 n 0 n 0 CI l Cu j I - * P4H o c IX I t 2 [f I b CO Px] COJ (U 0 P, I - I c 30 EH ~ ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ FH'^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,

The Ulniversity of Michigan Engineering ReSearchf I~nstitute ~i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ f illi~~~~~~~~~~~~~~~l si C3~:~~s -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~3~~~~~~I t O03~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~e a: 14~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I (. nI (N)~I~3 c )_... ~ i~ii~i~i'i~iiiiji'iiii'i~i~iiii~ii'iji C) ii3~iiicxixsliisiassiii 0iiiiii~iiiiiiiiiiii iiiiU:~~~~ rn~;;~ - iiiiii i~i~iiiiii~iiiii ~ ~'~~''~~~0'~~'~''~" ~ ~:l:lilP3~:::*:.~:;: iil~~~~~~~il~~~~~~~~~ll( ()lll:::::::~PF:::;.. iiiil~: Illlil~~~~~~~~~ll~~~~il~~~~i' -3 1:. "::::::i::4' M O)iii~~illii ~~~:::::::a:::::::: j:i: ~ ~ ~ iiliiillc~~iiciliiiiilillillllirlo C iiicjc iii~:il iiiiliil iiH'S.)l r~~~~~~~~::t):::::a::'4 bI I:::::::::::~::::::::::::::::::::::'C l 0iiii~iii~iii~i jiiiiii Uiiiiiiiil iiiiiiliiiil iillH ~~i ii iii~~'ijii i ii i ii i ii ~a~iiiiiir~i lllii~jilH3~

. The University of Michigan ~ IEngineering Research Institute.............................. w....................................................................................................................................................................................................................................................... 11 i D'4 00}i~~~:)... 0~i i.'" ~-' | I Ef:fect of vinyl chloridc~e o~n the h t resi t nce o'f irradiated polIyethylene sheets (Poly-Fth 1008.', 2id e, 1). Te.t..[., efec t of higher doses fs not, however, clear. Results:or sheets 49 and >2 suprt the erlir rvaton (Fig. 3 tat doe highe than I x b0' rep rc uit; in det";eroration of heat resist.ance, Neverthel.ess, snples 50 and 5 be haved well when exposed to heat, Regardless of the effects of high doses, it seesnv reasonable to conelude that expoae of a polyethylene het that hs been tate wth vinyl chlori~de to a radija~t'on dosfe of ~bout; B x l0~ rep resubt~fs in a ma.ter lfab htt.vinrg remarkable resistance'to heat. This resistance is much greater than that conr Effect of time of contact wth vii lI chloride: In the pr, Ii inactry expherlriment0 detsc bed earler, resul t's'ugesed thats w the tonsile strength of the fintal product was hsiigher the greater the length of time the she etwas w n contact with the vinyl chloride. The importanc e of a long contact time is demonstrated again in Tatble XII. Although the heat re'istane of -S~feet o-_____,"n,~ cotat stllch1-2 5

The University of Michigan * Engineering Research Institute Hi M ~r-l (D H 0 0 a-4H I H O o,> r o (D - oi O \) 0 O r- t lA U V l r- r ~1 CJ O. 0i! r C6-0 4EH O U D 5 M O 0 o ( r-4 d 4 0+ + I H E- o ei: M. r Q o 0.H: O X p! O'OL - OJO Ord -4 H *HD O0 0- C OO O _ - P:I T V o L('. n~ LCL C L —: C C 0O P;>-i CQd rd P \CDO ro j H E- r1 0 + +I + I H o H l ~ rd H.i.. --' H. H j-H>- O. EH Fpq l +? + + H * c~ I. H -. H PE~~-) ~4) HO E- O O hD (Fz^r H.-p e 00 00 00 00 0I r-l4 D r4rt 1- I HCJ c-0\ [i: 0 0 OO i Or- LC O O PE' ( P4 ~'. O JOJ r H r-.H VU V U4 + + + O -P E-l — P c FzC *rl C) O L 0 LO | l n H EO ~ % C O-p ^i 1., -1 O 0 X! CO CO CC?4 P4 CU2 q a 4 c U) 4-D HO 0. P -P- d H, 0 ~ (U J E' ",- Q=! *H ^ HLo 0 &-' ci 0U ) H^ Ei I' cC C) ci) r- -55

The University of Michigan * Engineering Research Institute a sheet (sample 26) irradiated immediately after charging with monomer was poor, the heat resistance was markedly improved by an increase in the time of contact with vinyl chloride to 3 hours, or better, to 24 hours. Apparently, then,the diffusion of monomer into the structure of polyethylene sheet must be an important factor in the development of heat resistance in the sheet. Effect of time of degassing: Because of the notorious behavior of oxygen in free radical reactions, one might expect that the extent to which oxygen is removed from the polyethylene sheets before irradiation affects physical properties considerably. This expectation is confirmed by the data in Table XIV, which gives properties as a function of the length of time the sheets were degassed before irradiation. As the time of degassing was increased, the heat resistance increased also. Thus the removal of residual oxygen from the polyethylene film before irradiation appears to be desirable if good heat resistance is to be obtained. b. Polyethylene Film.-Because of the considerable interest in thin films, some experiments with the irradiation of thin (2-mil) films saturated with vinyl chloride were begun. Results given in Table XV show that, as with the 20-mil polyethylene, exposure of the film to a dose of 8 x 105 rep resulted in superior heat resistance. However, the heat tests with this film were less extensive than with 20-mil sheets, and also less reproducible, possibly because of a high sensitivity to the cutting procedure. c. Polyethylene Pellets.-A tube filled with pellets of Poly-Eth 1008.5 (degassed for 22 hours) was charged with vinyl chloride and irradiated with a dose of 9o6 x 105 rep. After irradiation, the pellets were compressionmolded, and the tensile properties determined. It was difficult to obtain homogeneous sheets; values of tensile strength ranged between 1370 and 1890 psi-the average, 1650 psi, being considerably higher than the value of 1450 psi interpolated from previous data (Section A). Possibly the dose was high enough to induce enough crosslinking to cause molding difficulties. d. Relationship Between Heat Resistance and Other Properties. —So far the best correlation between the unusual stability with respect to heat and physical properties is with ultimate tensile strength and elongation, that is, with toughness (see Tables IX, XII, and XVI). Although it is difficult to visualize the precise meaning of toughness, it seems reasonable that the greater the toughness (and hence the greater the total energy required to deform a sample to the breaking point), the greater should be the resistance to deformation due to heating. Certainly one would expect the stability to be related to the elastic ---------------------- 4 —----------

The University of Michigan * Engineering Research Institute bD *H r-4 ^ ^Trrd Cd Cd Cd Cd rd rCd m H 00 0) 0 0 0 0 0 C c) wr J Hrr-A r- r- r-H H r-O r- 0 co1X 1 4 Z ^ o Q -p ) O ) C) C) C) C) C) C) i * C) 1 0U ^Q3 ^,Q \ Q Q \ Q Q Q0 \0 QQ 4 S Eq z C) E-0 4 o o *.^ *'.t *'\."..'.?',. *.t. (DO 00 C0 C0 o L r0- LC \L C J o 4ci r CU C d Cu CD CD CD CD CD CD CD COCO co CO CO CO Co CO CO CO CO. H 0 0 CO 4 2 LI\OOOOOOOLCOOLC\00000 H ^ CD ^.D4 r...Q r..D K o C.D o G U:2 C- Q n j O0 0 t -O O O K\H 0 C\' 0 0 0 0 0 H I O 4 H l bO n) \0 L t K\ K\ - _J- n nL \ Lr\-j Ln _L: _:J I;;(X) 1& 1 1 1 + + + + + +.+ *+ I O 1 H *- 0 CLI1, pr-JHH I I 0C\^Ot K r\V\ COO CDOOPO I O -0O1>-V- O -^0O -t —-l K- 0 t-ti - OHC) C 0 d~-l ^Q'0 W\LF\OOOOLC\OOLC0POOOOOO H O H * oooot oooo ooooooo ><1 HCO 0 Q \_ j HHHH-HH OHO\ HOJi\JHHOr-iaji\ r-4 OJH H Co n' n OJ-P - I- L CH o k\ \ C'Q U2 0'-CO b- co \-0 b- b- ~- L'- \) \ D \ -- b- I. l- Cr-O O + + + + + + + H' H-,.' Cd ~4 > 05 P I0 [D 0 H H H H\ L OC 0 O H T (D r- l Or- r r — r-i r — - C r-1. 0 r-' rA r-4-' r *..rS].7 P | ~r -! r-! r — r-4 r-4 r-I r-l- -l r-l rI r —! r -l r —/ J r I r-i r-4 0 - 2 H1 U?- +, + + +, + 1 + + + +,0 1'I; zCO OHE< 0 Cd H ~ o EE-i. 4 O r- - 0 0 O O O OOOOOOO O O O rq 0c —.- CC) - 0 C\ L N CO CO h-c 00 ( -— C n + —l I' " HI~u () PQ CJ - -. r —I r- r- r-i r-l r- oo r-! r- CQ C- i r- O U!E -- p4+ + + + + + I +,+ - 0o o0 c E~~~H Q (U ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 0 CH.H ^ _0 - - WP^4 -P L, 1-) *H i CO LA U) k9 ~ s F CX *H 0 wHcd H 0) QI (\J CC) CCJ:) co CdH] W ~~~~~~~.CH 0 H 0 H (DI 0 (D 0 CH'-e So- - - ^ 0 (D5

The University of Michigan * Engineering Research Institute I I I 4 ) H,H U 0. - Pt? Pq $-4 4-3? 4 -P' ( i ^ * r5" rdl, rd *H H * H. rr r JT}d hDD l d J a *H 0) r'H *r-1 rdC) +~ r -4 H I o- I rl ^ r - -PH ~' +- n'- - Q 1 rd rd C. 0) C) O ^i1 1) 4O @ E- - r 0 *r-H H r-4.- r-4 I 0 *H O o t rd?q'q*H ) Cn - CH r-4 r ^ 0 H. oj O CT cd O o. O o^ 0 g oj Cd Cd Cd C. C (0 u - r-4 r- LCN mt o *, T s o B C; *h rO e rd (, e CO KC\ O H 0 0 O gO O r jo.-I rid o d O O r-1 c H'- C j.H c H *H r4 Cd.r4-D - 4-+? -P 0? -P 0,_c l rd,-,, rd C p: q CC ca' cr p C +..or-" H H -P 0"' 0.0 0o0 0Lo 0 LC\ O.L'\ 0 00 ID H-'oj \ K'.- r-Hi HV HC -\- H C \OI\, r l'cK itlr-4 K'~ r — l\' r-I r,' r-.,c r-l-d- r-IK'::I +1 +' + + + + 0::,.-4 E- O O O O O O O O O O O O -.r-l - L - cO Coo O o o Lo oo -o o-O ^ ^ — Q ds\ p -.\I \ \' \ r\ \\ 0 - -\MtHCdO 4 Ci C\J r-l i- I r O H HH OH H 0 H H 0 0 HE H r- O rd- r - r- - r-l r- r- r rl - r i r~1~ + + + + + + + + + + > LCfX -P O' 0 00 00 00 0 00 00 00 3 * H brl hDO S CO 0cC ), J 0 O' N l _ 0 — 0 0'\ GCO 0La) g.t o'J \ -J -- - -- L- C- JK \ O C o a) * r- A N r-4 CQ r-4 Jr-4 - Jr C\J rO + I + + + I + + ff E-1" > r -P 00 0 0 00 00 COLn 00 00 Oi.rJ. OHO'0OH - O' O -0 O O - CO LO\ 0 o m to O - Lrn - bL - - 0\ 0\ - - a- \ \K O0 UCO P-, C) a H OIj 0H CJ r- C i O r- OJC H CoJ CJC C\ Hv | ^ + EIM + + +1 - + + I E- C. 0 o)H 0 0 0 0 0 0 O o.CO. O O O O O O P4 H 0) I C) C O C *H 00) toQHc I CMJ COJ COJ OJ1 c1 U b0 a i) -.- L~ kO r-I K OJ d t.DI I I I I I - H H H EH H H H-. 36

The University of Michigan * Engineering Research Institute IP OOOOOOOOUooOO OOOOOOOOO Ln\ oooo 0- 00000 04 r C O O000 00 0 0 00L00 00 C" Itp 0 t — 0 O \0 00 N o o --!a: r- 0" Ln O",CO -' hI -- tn oh00 Ln n 0oJ r- d + + + + + + + + +1+1+1+ + 0 r-I rq E-. -. o0, o o 0o 00-,o o. O o' o o o O O c o\ 00o o o o o m p C~ O b- \ rl ~ -- — 4 - O-\ K, 0n Q -) L 0 — 0- o \9 O 0- -- - - N o Cd rd 04 \D N \O \N \D L- \,OO \ L LCCOO t-O O U-Ck o rdO) 00OOOOO0W\OOOOOOOOOOL\OOOj\00 P0 () PA H —H H — 4 — 4 r — r' —! r'-l -H - — r r -H r -. r'-H r -H r --' r - r r -4 r 4rr-1 r-4 ~ + + + + + + + +,+ 1 + + + + |+ + + C'l H o 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 H. br-l rJ OJ 0 \ \ L* -- OOd CH n K\ 0 0 -- RCO CO r- 0 r- ~- f - > O O a C 1. — - H \- - -— r-t rd- \.-H \,D )- b- - t H L'n 1D b-Cr 0 0 L — b — O\ n < E! Pq rdr, -4.-4 r~-4,r-r! -I CJ C\J r^ ~r - rd r -r-!0 r r —.! r-t r-~ r-d rd J COJ -,r-I r- i-rPq *H ao A od +, +, +, +, +. +, +, +, +, +. +, +, g,4 +, +, +, +, + + +, +, +, + + +'+ 1 0 CO ~ O! a r- r-'4O oo -0 ) E3 p i O CO C) () O $Z Xi H CG C' CG I l 1 1 P-l - H H s-p.U *rl 3-1 u vz. ~,.- Lr''.-. --.l). 0 C l ^3 p, OJ OJ:1, I ~ I I P- 0P P /2 PP ^ - - -. ^ a

The University of Michigan * Engineering Research Institute modulus.. The data, however, are not conclusive (Table XVI)o In any case, one might conclude that the toughness and strength due to crosslinking result in a considerable resistance to buckling. Although it is not clear how vinyl chloride acts so specifically, several possibilities exist. For example, the vinyl chloride may disrupt the crystallinity of the polyethylene slightly, especially at the surfaces. Possibly the resulting situation in which a crystalline material is faced with a more or less amorphous layer is analagous to the situation existing in a polished metal, in which the surface consists of a disorganized layer of metal. It is believed that the strength of a polished metal is associated with the "skin" just mentioned. Alternatively, the vinyl chloride may have a "healing" effect on surface imperfections; it may relieve stresses in the polymer sheet and thus increase the resistance to bucklingo It is more difficult to consider the diminished shrinkage in the machine direction. Apparently in some way treatment with vinyl chloride must more or less equalize "something." One might postulate that the use of vinyl chloride induces crosslinking between regions oriented by fabrication rather than between molecules in the same oriented region; this might be reasonable, as the vapor could diffuse more readily into the amorphous areas between the oriented regions than into the oriented regions themselveso Indeed, it has already been observed that the diffusion of vinyl chloride is important for the development of heat resistance (Table XIII), However, one would expect to find this equalization reflected in-physical properties as a function of sheet direction. But strangely enough, neither the tensile and yield strengths nor percent elongation for transverse (-) and par'allel (+)' samples show a consistent trend that could be correlated with either the stability with respect to heat or toughness. Similarly, no consistent trends may be noted in the elastic modulus. e. Vinyl Chloride and Crosslinking —Although the value of treating films and sheets with vinyl chloride before irradiation seems clear, the question of fabricating treated polymers arises. For this reason the milling of some irradiated treated sheet (47), plain sheet, and irradiated sheet (47') was tried at 220~Fo While the plain sheet flowed smoothly, it was necessary to increase the temperature to handle the other sheets. Eventually, the irradiated sheet formed a crinkly sheet; the treated sheet seemed to flow a little at first, and then formed a sheet that was much more crinkly than the irradiated sheet. When square samples of each of the three milled sheets were exposed to a temperature of 125~C for 20 minutes, the plain sheet melted, the irradiated sheet flowed into an elongated rectangle, but the treated irradiated sheet retained both its square form and its original appearance, though with a little (10-20%) shrinkage in each direction. In other words, the treated irradiated polymer behaved as though it were highly crosslinked, These results suggest two alternatives: that the use of vinyl chloride results in a greatly enhanced degree of cross linking at the time of irradiation, or that a crosslinking reaction occurs daring heating as well as during irradiation. If 38

The University of Michigan * Engineering Research Institute the former alternative is true, then fabrication of treated polymers (as in the initial work with pellets) may be no more easy than the fabrication of polymers that have been merely irradiated. If, on the other hand, the latter alternative is true, the use of an inhibitor should help prevent the second-stage crosslinking during molding or milling.* Further experiments would be required to settle this matter: for example, (a) a study of the effect of inhibitors on fabrication; (b) a study of the effect of very high doses on heat stability of films and sheets; and (c) determination of gel content for plain, irradiated, and irradiated treated'sheets and filmso CONCLUSIONS 1 By the irradiation of thin polyethylene sheets or films (20-mil or 2-mil) that have been, to some degree, saturated with vinyl chloride, the resistance to shrinking and buckling on exposure to heat may be increased markedly. 2o The optimum dose appears to be about (8-9) x 105 repO Best results are obtained by allowing the monomer to remain in contact with degassed polyethylene for a long period of time, say, 24 hours; degassing of the polyethylene beforehand is essentials 35 The resistance to heat may be correlated with toughness. Yield strengths and elastic moduli do not, on the other hand,'bear much relationship to the thermal behavior, 4.. Even if the vinyl chloride acts as a chain transfer agent, the mechanism leading to the observed effects is not clearo So far, several possibilities, not necessarily mutually exclusive, exist: a "healing" effect on surface imperfections, a "skin" effect resulting in greater toughness, the crosslinking of regions of high crystallinity formed by orienting, and the promotion of crosslinking in general. More evidence will be required to understand the mechanism or mechanisms more thoroughly. *Some support of this alternative is given by the observation that the shrinkage observed in heating the irradiated treated sheet was more or less equal in both dimensions. This fact would be consistent with the occurrence of crosslinking at a high temperature, for little distortion of the resulting structure should occur in further heating; on the other hand, a sample already crosslinked at a low temperature before fabrication should tend to shrink on heatingo 39

The University of Michigan * Engineering Research Institute Co Graft Polymerization of Monomers to Polyethylene INTRODUCTION In the search for new high polymers or desirable modifications of old ones, the technique of graft polymerization has received much attention.42 By this technique, branches composed of one polymer may be added to a backbone chain of another polymer. Functional groups of interest may be added to a backbone chain having desirable characteristics of its own, for example, strength; the properties of the resultant product are generally determined by the sum of the properties of the individual polymers rather than some function less than the sum. It has been reported that the grafting of monomers to a backbone chain of, say, polyethylene preceeds very favorably under the influence of gamma radiation.431,44' Ballantine and others43 have described the grafting of monomers such as styrene and acrylonitrile' to polyethylene film by means of irradiation. Similar experiments have been described by Chen and others,44 who reported the following: by the grafting of styrene to Teflon, a substantia increase in adhesive power; by the grafting of acrylonitrile to polydimethyl siloxane, an increase in solvent resistance; and by the grafting of vinyl carbazole to polyethylene, an improvement in temperature performance as a lowloss dielectric. Because of these claims of versatility fbr grafting, it was decided to investigate the grafting of several monomers to polyethylene to see if any useful products would be formed. Results of this study are described below. EXPERIMENTAL Pellets of Poly-Eth 1008.5 were obtained from the Spencer Chemical Company. When desired, crumbs were prepared by milling pellets for 5 minutes at 220~F. Polyethylene' film was a common commercial wrapping type. The monomers were commercial grade, and were distilled shortly before use. Graft polymerizations were conducted in tubes sealed under vacuum; the contents of the tubes were degassed before sealing. A.few experiments were also made in which samples of Poly-Eth 1008.5 in sheet form that had been irradiated previously with a dose of 9.6 x 105 rep were immersed in acrylonitrile. After reaction, polyethylene film was removed, washed with an appropriate solvent, and dried at 75~C- Polyethylene solutions were precipitated in methanol, filtered off, washed, and dried in the same manner as the film. -------------------— 4 o ----------

The University of Michigan * Engineering Research Institute Tensile measurements and tests of adhesive rating were made as described in Sections B and C. RESULTS AMD DISCUSSION Best results were obtained with grafting to films and grafting in solution. This section is, therefore, concerned with these two topics. Polyethylene Film, 4.5-mil: Results are given in Table XVII for experiments with acrylonitrile, styrene, and vinyl acetate. It may be seen that the use of acrylonitrile resulted in a marked increase in tensile strength, little change in percent elongation, and, surprisingly, a marked decrease in the adhesive rating. The irradiation of a film.in contact with the monomer gave a slightly higher tensile strength than the irradiation in vacuum of a film swollen with monomer. The amounts of acrylonitrile concerned in the grafting reaction must have been small, for no change in the weight of the film could be detected after graftingo In general, somewhat higher tensile strengths were observed for the styrene copolymers. The adhesive rating was also higher, at least for samples 7,. 8, and 8'-and, indeed, higher than for the blankso On the other hand, the percent elongation was lower for samples 7, 8, and 8". In this case, the irradiation of a film swollen with monomer, rather than the irradiation of an immersed film, gave the higher tensile strength. Increases in the weight of the films were also observed-14.8% and 15.o2 for samples 8 and 8', respectivelyo In contrast to the other monomers, the use-of vinyl acetate resulted in no consistent difference in properties-certainly in no improvement.o Thus, the grafting of acrylonitrile and styrene to polyethylene is promoted by relatively small doses of gamma radiation0 Polyethylene in Solutiono These experiments were made in the hope that grafting in solution would yield a more homogeneous product. Typical results of grafting experiments using acrylonitrile and styrene as monomers are given in Table XVI oII. With acrylonitrile, an improvement in tensile strength over plain film was observed;* the tensile strength was independent of the concentration of acrylonitrile in the solution. The most interesting observation, however, was that transparency of the polyethylene was considerably improved by grafting, especially with the sample for which the smallest proportion -of acryloni*The blanks were not milledo However, milling reduces the tensile strength, and hence any differences observed between values for graft polymers and values for blanks are probably less than the true differences. -41

The University of Michigan * Engineering Research Institute 4-)> 4UbD rb a) p I r-H H Cd C a) a) 0 *H *H C0 -P0 C Q C d d p () *H *H U) c U> L ~ Cd Cd'c o0 ) () a -H -P k4 Kt 040 -z1 LCN Cd'r* rd 0 C]O Pi o,JH 0 H i r 0r- n 0 0 0 0 0 0 0 0 0 0- 0 r- P CdJ K-N Ln i \ O CUj KN Ln 00 C O \ - C- LC\ K^ *H Z; 0 * ^0 [> O> 0 VO ^O C O z o O CHd o H 1 Q a - U) 0 > bD - CH H c - O D O OO O O O 0 0 0 O a) UO )H b 0 O cO 0 O LCn \1 0 Ln ON \) ~O H - I)P - A C A r- r- r- CMU 0 C I r I (> HCQ'hDb H O O 0 0 0 0 0 00 0 0 0 iO O U U 0 _o -t K\ _ CM CO ON t ct- C \ LC n 0 r-A daP -z.- -.- -0 LN, CO r,, LC\ -c- 0N C — M' LN ) a) CM CM CM CMD CM C CMj CM CM CM CMN C\J -P! Cd E *H 0 54 *H - *H *H ~, ct cd X~~~~~~~~-a) a) 01 (D (D r CY I o O A 0l \0 0 i o P r \ Cd Cd 0C H'CHC C w#O wP4 p- I I - l Cd Cd C0 C) C) C) -p - - U0 0 I I.HH Hr-I.r'R 1 O O CO o o o o o d 0 C D ac 0 0 0 0 4 p p.r —.r- 0 0 0U r —i r — H r-^ O C) P) PC~ -P ~ ~rlO(a ) uz~ m4 ~4 m m 42co ~ ~ ~ ~ ~ ~ c C a ------------------------------ 4 2 ------------------------------

The University of Michigan * Engineering Research Institute -P ~'. -p -P~ Cd Cd Cd'd C4 F1! -c~~~~~~~-?-?- i P4 P4 P4 CQ 0L CQ CQ C) Cd Cd Cd Cd rP 4 4 P C a ri~~~~~~~~O cci c(5 c^I o3 ~A 8 b CQb -p -P -P o Cd ~a 3 CD I H H H 4d 4d Cd Cd -P -P~ -P -P a40 0 0 0 HI ~I d H- Hl H H 0 S 0 E0) 0l ir r-l nJ T 4-D-p - +3 l (] ( ( 0 ^ bo b O -P -P -P -P U) ^-'5'5 *r5 -p 4' 4-P -P 0 H lr-4 H 0 0 0 H o Cd. 0 0 0 0 0 0) 0 LFI\ rr~ 0 bO o H r 3 r0 0 O 8 O rO 4 CON CQ CO CU U c! 1-I rd C-) z P-i 0 U)~~~~~~~~~~~~~~~~~ 0 c[0-^ 0 0 0 0 0 0 0 0 0 ~ d 01 0 01 01 CM0 014 C\ 01 01 tf SI ^o o- co - o t> Lor!Z! (1) n L^n tn \Ln Ln - LJ I H^~~~ ^~O0 C) C)~~~~~~C H P H~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C 0 Ed (D t -^ t - 3 r-4u r-4 rA sI r_4 r-I 4 Q) O r-4 r- -l r _ izI- r _-, r-I I Ird 0 H oE 0t 01 0 01 01 01 01 01 01 0 0 1 O P-i U)- bi ^ K _l ^- i }IOa K < C) aU o 0 H Q r-l~~~~ ~~~~~~~-P O Cd i> H * H 0 0 0 0 000 X EP 03 I-IO OH HO O r 5 H o 0<c &^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~D PX FC ('.. 0 0 0 0 r I l U) ~ (Do H H H H 0 0 0r 0r r-l I n pg ~ ~~~~~~~~~~~~~~~~~ E-i r-iI (-1 a) a, a H I C ) C ~Cd P-0 LC (]- O - 0.0H C -P- ~LIPHeCd H H, P-U h-U r C O C) p^ CD X! * * * * * - i - 3^`t I H 15;~~~?i O ) O - - ~I ~ e CO (Ue d ed m m c mc i3 Fr;~r r-l r- rH r-. p^u *H T- *H *r-l P-1 1 3^ ^ 1 vl O O O O ^ CJ ^ CJ I' (-3 r ~ i- i-l i l (D ( (D (U' * (

The University of Michigan * Engineering Research Institute trile had been used. Thus the branching reaction apparently disrupted the crystallinity enough to increase the transparency; at the same time, the tensile strength was not adversely affected. Higher tensile strengths were observed when styrene, especiallya small amount: of styrene, was used. However, the samples had a mottled appearance. All these results show that it may be possible to select particular monomers, such as acrylonitrile, to modify the polyethylene structure and at the same time contribute some virtues of their own. If special combinations of properties are required, the technique of grafting may:be useful. CONCLUSIONS 1. The tensile strengths of polyethylene film may be increased by irradiation in the presence of acrylonitrile or styrene. 2o The adhesive rating of polyethylene film may be increased by irradiation in the presence of styrene. 3. The structure of polyethylene may be modified without a corresponding reduction in tensile strength by irradiation in a solution containing small amounts (say, 5%) of acrylonitrile. 4. The tensile strength of polyethylene may be increased considerably by irradiation in a solution containing small amounts (say, 7%) of styrene. 5- Since only relatively small doses are required, the technique of grafting provides a convenient method for adding desired groups into polyethylene either for their own sake or to modify the polyethylene structure. D. Determination of Molecular Weight Distribution in Polyethylene by an Irradiation Technique INTRODUCTION Although the molecular weight distribution in polyethylene is a very important property, experimental determination of the breadth of distribution, as indicated by the ratio Mw/Mn, is difficult. Recently a new method for determining this ratio was suggested by Baskett. 4 This method is based upon a mathematical analysis of curves relating the soluble fraction, (l-x), of a polymer crosslinked by irradiation to the fraction of carbon atoms crosslinked, a (:x assumed to be proportional to the dose received). The method is essentially an extension of work done by Charlesby.4647 44

The University of Michigan * Engineering Research Institute Baskett suggests fitting experimental values of (l-x) to a linear combination of a series of functions of p [p being equal to -log(l- x)]whose inverse Laplace transforms are known. If, for example, 1 - x o= bao + bla W(p), (P~a0).2 (p+a )2 (p+a) 2 the transform of the series becomes boaome-am + blal2me-am.... = W(m), where W(m) is the distribution function, that is, the weight fraction of molecules containing m carbon atoms in the original unirradiated material, and where ao, al, bo, and bl are arbitrary constants. From the exponential series given, Mw and Mn can be readily found: DPw = 2 + L and DPn = l/(aobo + alb..) Two complications must, however, be considered: chain scission, which reduces the gel fraction corresponding to a given degree of crosslinking, and the fact that the extent of scission depends on the degree of branching. Baskett has derived a correction for these effects by assuming that a real sample of polyethylene may be treated as a polymer of simple exponential distribution that has been subjected to a certain degree (ao) of preliminary crosslinking. Since this method avoids the problems associated with light-scattering and osmotic pressure measurements at elevated temperatures, it was decided that a study of this method would be very useful. In this section results obtained for the following polymers will be presented and discussed. A5L, A10, A22, A34, DYNH, TD385, Blend 7, and Poly-Eth samples 1005, 1405, 1503, and 2205, Irradiations were completed for TD304, TD386, and Poly-Eth 1205, but time did not permit solubility determinations. EXPERIMENTAL Samples (0.2-0.5 g) in pellet form were degassed by evacuation for from 2 to 4 hours using a high-vacuum line (mercury diffusion pump backed by an oil pump), sealed in glass tubes, and irradiated using the 1-kilocurie or 3kilocurie Co-60 source. Doses ranged from 5~0 x 104 to 3.0 x 107 rep, and corresponded approximately to those used by Baskett. All samples were supplied by the Spencer Chemical Company. When the irradiation was complete for a given sample, the product was extracted by means of a hot Soxhlet extraction for approximately 48 hours. It was found that the use of pellets gave results identical with results obtained 45

The University of Michigan * Engineering Research Institute for the thin discs recommended by Baskett. Although the first extractions were conducted using reagent-grade benzene at 80~C, in later work it was necessary to use reagent grade toluene at 100~C. After drying for 24 hours at 80~C under vacuum, the weight of the insoluble fraction, and hence of the soluble fraction, was determined. Some difficulty was experienced in obtaining reproducible weights for some of the extraction thimbles. Best results for the weights of the dried residues were obtained in the following way. First the thimble containing the dried residue was allowed to come to equilibrium, without desiccation, in the balance room, which was kept at constant temperature (25~C) and humidity (50%, relative), usually a period of 6 hours was sufficient. Next the thimble plus contents was weighed, ignited at 1000~C for 12 hours, and then reweighed. Finally, after being allowed to reach equilibrium again in the balance room, the ignited thimble was weighed, and the weight of residue burned off was determined by difference. Values of the soluble fraction were then plotted against the total dose received, and corrected for chain scission and the presence of branching by the procedure suggested by Baskett. The resulting corrected curve corresponds approximately to a hypothetical linear polymer of simple exponential distribution,*which has been slightly crosslinked to form the actual polymer subjected to irradiation. The data were then treated in two ways. In the early experiments, the plotting of the data as a linear combination of Laplace transforms was attempted. Later, the general procedure described below was used; details of the calculations may be found in the appendix. First, values of Mw were calculated from the intercept of the upper portion of the uncorrected soluble curve at infinite solubility, Mw being equal to the reciprocal of the intercept and being independent of distribution. The values of Mw for the hypothetical parent polymer (see above) were calculated in a similar way from the intercept of the corrected soluble fraction curve. Since this hypothetical parent polymer has a distribution approximating a normal one, the value of Mn was taken to be one-half of the value found for Mw. Next this value of Mn was corrected to allow for the presence of branching, this new value would correspond approximately to the value for the real polymer. In other words, the effect of a certain number of crosslinks on Mn was determined, the number of crosslinks was taken to be the same number assumed in Baskett's method of correcting the observed soluble fraction curves for branching. *That is, W(m) = a2me-am 46

The University of Michigan * Engineering Research Institute RESULTS AND DISCUSSION Most consistent results were obtained using the method described in the appendix. Even with assistance from the Statistical Research Center, attempts to plot the experimental data as a linear combination of Laplace transforms were unsuccessful. Results presented in this section were, therefore, obtained as described in the appendix. Typical curves of the soluble fraction, (1 - x), as a function of radiation dose are given in Figs. 6 and 7. Two general features of the curves are worth noting: 1. The general shape of the curve is related to the breadth of the molecular weight distribution. The broader the distribution, the lower the slope and the greater the curvature of the major portion of the curveo Thus, the curve for A-22 (Mw/Mn'' 6) is relatively steep and linear, while the curve for A-10 (MW/Mwnr 20) slopes off more gradually. 2. At low doses, the curves tend to become asymptotic. It is true that measurements of the soluble fraction are difficult as infinite solubility is approached. Nevertheless, repeated determinations have confirmed that this behavior is, indeed, characteristic of most samples of polyethylene, whether or not the unirradiated sample is completely soluble under the extraction conditions. Apparently, then, most samples of polyethylene must contain very small amounts of very high molecular weight material that either is insoluble to start with or becomes insoluble after exposure to an extremely small amount of radiation, this conclusion is consistent with general experience. Clearly a precise extrapolation to determine cc and hence Mw is difficult, if not impossible, if such a "tail" effect exists. Until it is possible to account quantitatively* for the effect of this small amount of material, one must ignore its presence and try to extrapolate the major portion of the curve. Of course, any comparison of Mw with Mw from lightscattering, it should be realized that some or all of this high molecular weight component is centrifuged out before light-scattering measurements are made. Results obtained so ifar are given in Table XIX; incomplete solubility data are given in Table XX. Samples I, II, III, IV: Use of data for these samples taken from Baskett's paper, yielded ratios of Mw/Mn that agree well with ratios cal*Mr, Graessley is studying the effects of small proportions of high molecular weight material on the solubility curve. So far, it appears that only 2-3 percent of material with Mw 500 times the Mw for the major portion of the polymer could result in the tail effect observed.. —--- 47

The University of Michigan * Engineering Research Institute 8 - 8 o Pu,~' P~ — ~- O O r —4 /O o X ~0.^-I~0 r-4 e_ 0 ) r4 0 o5 00^ uT~0 rd 3 C (x - I) NOI138S 3(I 48 -t0 0- 0 ~ o) 000 o rOx-l —.~- 0 oo~~~~ I6- 0 W 0 6 (- 0 x,-O - 6 d 6 o ~ g (x-I)'NOliDvSJ 3'18'oS 48

The University of Michigan ~ Engineering Research Institute --------- 001 y r__g - lox / ~ X x 0 ~ $3 ~~/^~~~~~~ ~ 0,-! a)'L o E I 0 //'^ < g Co \bQ Cc.2 o.m_.- Cl) o r 0(x-i) N01iiova4 3)n-mos

The University of Michigan * Engineering Research Institute TABLE XIX MOLECULAR WEIGHT DISTRIBUTION DATA DERIVED FROM SOLUBILITY CURVES OF IRRADIATED POLYETHYLENE Sample Mwn M, w/Mn M w/Mn Comments x 10-5 x 10- observed comparison _ _ _ _ a b I - - 8 6b II a - 10 8b Solubility fraction data IIIa 10 12b from reference 45~ IVa -- - 4 5b A-5L - -- -- -- Prelimination measurements indicated broad distribution. A-10d 9.0 0.44 20 13c Presence of high-M tail dis20.2 0.47 43 counted in calculations. A-22 8.4 1.4 6 7 Presence of high-M tail discounted in calculations. A-34 0.5 0.36 14 (Narrow)c Pronounced high-M tail discounted in calculations. PE-1005 -6.4 1.6 o40 -0 A "best" value for Mw/Mn might be 70. PE-1405 -- -- -- -- High-M tail. PE-1503 -- -- -- - High-M tail. PE-2205 1.26 1.01 1.25 High-M tail. TD-385 2.1 1.2 17 DYNH 1.58 1.66 9.5 Relatively narrow distribution except for a high-M tail (discounted). aData obtained from Ref. 45; raw data were only approximate. Values from Ref. 44; Mw from gel curves, Mn from osmotic measurements quotedo Values given in duPont information bulletin (Ref. 48). The two sets of.data arise from different extrapolations; see text. 50

The University of Michigan * Engineering Research Institute TABLE XX SOLUBILITY DATA FOR IRRADIATED POLYETHYIZNE PE-1503 Blend 7 PE-1405 Dose, Soluble Dose, Soluble Dose, Solible rep x 10-6 Fraction rep x 10-6 Fraction rep x 10-6 Fraction 0.085 0.973 10.0 0. 687 0.085 0. 990 0.5 0.963 15.0 0.595 0.50 0.990 0.972 0.932 20.0 0.361 1.0 0.981 1.0 0.991 30.0 0.354 1.8 0.99 0.955 o. 964 2.1 0.99 1.8 0.9538 4.o 0.994 0.946 6.0 o 0978 2.1 0.978 0.967 6.5 0.95 4.0 0.979 10.2 0.78 0.932 15.4 O 0.57 6.o 0.951 0.961 7.8 1.00 0.937 10.2 0.843 51

The University of Michigan * Engineering Research Institute culated using osmotic data given in the paper. In one case, sample II, the value of Mn calculated by our method was closer to the osmotic value than:-was Baskettts value* on the other hand, with sample III, the value found for Mn did not agree with the osmotic value. ASL: Because of the relatively high content of carbon black, few measurements were made on this. sample. However, a broad distribution was indicated by the preliminary measurementso A-22 (99-100% soluble): The extrapolation is not as difficult as in some other cases because of the steep slope of the right-hand portion of the curve. If the effect of the high molecular weight tail is ignored, a "best' value of 6 may be estimated for Mw/uno Although this ratio is Close to the approximate value supplied by the du Pont Company4 7 the absolute values of Mw and Mn are in disagreement with the du Pont values* —300,000 and 45,000, respectively. No reason for this discrepancy is apparent. A-10 (98% soluble): With a very broad distribution as exists with this sample, extrapolation is very difficult. Two possible extrapolations are evident in Fig. 7, and results of these alternatives are presented in Table XIX. Again, the values of Mw/Mn, at least for one extrapolation, is fairly consistent with the value supplied by the du Pont Company. A-34 (99% soluble): This sample appears to have a broad distribution, contrary to the data supplied by the du Pont Company. In particular, a pronounced high molecular weight tail was observed. Poly-Eth 1005 (100% soluble): The solubility curve for this sample resembles the curve for A-10. A very broad distribution of molecular weights must exist for the "best" value, for the ratio Mw/Mn would be approximately 70. Poly-Eth 1405 (99 soluble): Although more solubility data at high doses would be required to complete the study of this sample, acceptance of all the existing data would suggest a fairly narrow distribution. Poly-Eth 1503: For some.reason the- data obtained were difficult to reproduce. Since this polymer was fairly soluble even after a dose of 1 x 107 rep, both Mw and Mn must be relatively low. Poly-Eth 2205 (99.8% soluble): Again, the presence of a high molecular weight tail is obvious. If, however, this tail is ignored, the distribution of most of the polymer appears to be fairly narrow, at least for polyethylene. *These values were supplied through the courtesy of Dr. Billmeyer, Polychemicals Department, E. I. du Pont de Nemours and Company..... —-- - 52

The University of Michigan * Engineering Research Institute TD 385 (99% soluble): The solubility data imply a moderate breadth for the molecular weight distribution. DYNH (99% soluble): If the presence of a high molecular weight tail is discounted in calculations, the distribution should be fairly narrow. These results illustrate the advantages and disadvantages of this technique. In a general way, the method reveals whether or not a polymer sample has a broad or narrow distribution of molecular weights. This is true whether or not one wishes to include a high molecular weight tail in considering the breadth of distribution. Once some experience has ben gained, the technique is relatively easy to use, although very precise determinations of gel fractions are necessary when the solubility is close to infinite. The determination of absolute values of Mw and Mn, however, still is uncertain. The reason why good absolute values of Mn are obtained in some cases and not in others is not clear'. Undoubtedly there are many uncertainties in the method used, such as the conversion of dose to degree of crosslinking, the choice of P, the accurate determination of gel fraction, etc. But these errors might be expected to affect all values of Mn; thus the source of error does not appear to be a consistent one. CONCLUSIONS 1. Examination of the curve relating the soluble fraction of irradiated polyethylene to dose can reveal useful information about the breadth of the molecular weight distributiono Although precise determination of the ratio M/M-n may be difficult if the distribution is broad or if a high molecular weight tail is present, at least a qualitative indication of the breadth may be obtained. 2. It appears possible to obtain' reasonable estimates of the absolute value of Mn in some cases, but not in others. However, even if the estimate of Mn is in error, the ratio Mw/Mn seems to agree reasonably well with the ratio determined from measurements of light.-scattering and osmotic pressure, at least in the few cases for which such data were available. III. MISCELLANEOUS STUDIES A. Mooney Viscosity Measurements on Polyethylene INTRODUCTION Complex relationships exist between such physical properties of polyethylene as the following: melt index, intrinsic viscosity, molecular --- 553

The University of Michigan * Engineering Research Institute weight, molecular weight distribution, and degree of branching. At present, measurements of melt index and intrinsic viscosity are routine; measurements of the other properties listed are, however, tedious and difficult. Unfortunately, because of the influence of these latter properties, two samples of polyethylene having the same melt index (and perhaps the same intrinsic viscosity as well) may have quite different characteristics. To see if such samples could be differentiated by means of simple methods, it was decided to study the viscosity, measured with a Mooney viscometer, of several samples of polyethylene having the same melt index. EXPERIMENTAL The following samples were studied: Poly-Eth 1005 (Nat B, Lot 3643), Poly-Eth 1205 (Nat B,, Lot 3756), Poly-Eth 2205 (Nat B, Lot 3019), Poly-Eth 2405 (Nat B, Lot 2466), and Poly-Eth 1008.5 (Nat B, Lot 3161). Preliminary measurements were made using pellets; later molded sheets were used. The Mooney viscometer was obtained from Scott Testers, Inc.; the large rotor was used in these experiments. Cellophane was used between the polymer and metal surfaces. For the pellet experiments, the viscometer was operated at 280~Fthis temperature being high enough for molding the most viscous polymer. After the sample had remained in the closed mold for 10 or 15 minutes, the motor was. started; after 1.5 minutes, gauge readings (which did not change much thereafter) were recorded. Samples were examined for scoring after each run to be sure the slippage had not occurred along the mold faces. For the sheet experiments, pellets were molded into 1/8-inch -sheets at 300~F and 30>000 pounds for 10 minutes, released while hot, and then quenched; these sheets were cut into pieces approximately 2 x 2 inches, and introduced into the viscometer mold chamber, which was, in this case, heated to 240~F. After 5 minutes the mold was closed; 25 minutes later the motor was started, and gauge readings were recorded periodically over a period of 3 hours. As before, samples were examined for scoring after each-run. RESULTS AND DISCUSSION Results of the preliminary pellet experiments at 280oF are given in Table XXI. Although the scale readings are inconveniently low, differences between the samples are evident. In an attempt to obtain larger scale readings and hence larger differences in scale readings for different samples, the temperature of operation was lowered to 240~F. At this temperature it was necessary to use layers of molded sheets for the sample. Results are given in Table XXII; a typical curve 54

The University of Michigan * Engineering Research Institute TABLE XXI MOONEY READINGS FOR POLYETHYLENE AT 280~F Sample: pellets Time for reading: 1.5 min Temperature: 280~F Molding time: 10-15 min Large rotor Sample Triala Scale Units Sampl e Trial Mooney Reading_ Average Reading PE-1005 1 7.8 7.8 7.6+ 0.2 2 7.5 PE-1205 1 2.6 2.6 3.0 + 0.5 2 3.5 3.5 PE-2205 1 3.8 3.8 3.3 + 0.8 2 2.5 3.0 PE-2405 1 7.9 7.3 + 1.3 2 6.0 8.6 PE-1008.5 1 1.5 1.5 2.0 + 0.5 2 2.5 aEach trial refers to a separate charging of sample. 55

The University of Michigan * Engineering Research Institute TABLE XXII MOONEY REASDINGS FOR POLYETHYLENE AT 240~ F Sample: 6 1/8-in. sheets Time for reading: variable Temperature: 240 (+ 1)~F Molding time: 30 min Large rotor Mooney Reading, Scale Units Sample Trial. 2 min 10 main 30 min 2 hr 3 hr PE-1205 1 - - - 24.0 2 32 28 25.5 22.5 21.8 24.6 24.2 22.6 22.0 22.0 4 42 33 29.0 24.5 5 _41.5 30 5 27.1 22.53 21.5 Avg 5 + 10 28.9 26 + 3 23 1. 21.8 0.5 PE-1005 1 14.9 14.8 14.3 15.5 13.2 2 16.4 15.7 i4.5 13.0 12.8 Avg 16 + 1 5.2 ~ 0.5 14.4 + 0.1 15.2 ~ 0.3 13.0 + 0.2 PE-2205 1 15.1 14.0 12.8 12.0 11.9 2 15.1 14.7 13.8 13.0 12.8 Avg 15.1 14.3 + o.4 13.3 + 0.5 12.5 + 0.5 12.4 0o.5 PE-2405 1 15.3 14.2 12.8 11.3 11.0 2 15.0 14.1 12.7 11.2 10.9 Avg 15.1 + 0.2 14.1 ~ 0.1 12.7 + 0.1 11.2 + 0.1 11.0 + 0.1 PE-1008.5 1 3.5 3-2 3.3 3.2 3.2 2 3.2 3.1 3.3 3.1 3.1 Avg 3.3 ~ 0.2 3.2 + 0.1 3.3 3.2 + 0.1 3.1 + 0.1 _ —--------- ~~~~56

The University of Michigan * Engineering Research Institute is given in Fig. 8. Except for trials 2 and 3 (Poly-Eth 1205), which were made while the technique was being mastered, the experimental precision of the average values of scale reading was generally + 0.5 scale units. Although at 2 minutes large differences between the rather imprecise data for Poly-Eth 1205 and Poly-Eth 1005 exist, differences between data for Poly-Eth 1005, Poly-Eth 2205, and Poly-Eth 2405 are not apparent. However, at longer times, the precision of data for PE 1205 and PE 1005 improves, and slight but significant differences between data for PE 1005, PE 2205, and PE 2405 appear. The cause of this spreading out at long times is not clear. 20 O 0 D- Trial I |> |0- Trial 2 1 5 o 0 l E'l -Trial l 0 30 60 90 120 150 180 TIME, MI.N Fig. 8. Dependence of Mooney scale readings for polyethylene on time. Results of the two sets of experiments do not agree. Possibly there is a difference in temperature dependence of the shear behavior of the samples. However, the data for the sheet experiments are probably the more reliable. In any case, the Mooney viscosity values of polyethylene samples having the same melt index are different. These differences might be related to a property such as the degree of branching, and could be useful in the routine characterization of polyethylene. CONCLUSIONS The following conclusions may be drawn: ---------------- 5 7

The University of Michigan * Engineering Research Institute 1. The Mooney viscometer can be used to differentiate between samples of polyethylene having the same melt index. 2. So far, best precision (deviation from the average about ~ 0.5 scale units) and best differentiation may be obtained by using layers of molded sheets as samples,.a molding time of 30 minutes, a long reading time (say, 30 minutes, or longer), and a temperature of 240~F. B. Synthesis of Cyclohexanone Oxi'me Using Gamma Radiation INTRODUCTION For some years, it has been known that aliphatic hydrocarbons such as n-heptane or n-hexane may be nitrosylated by nitrosyl chloride under the influence of light (that is, ultraviolet radiation) to form products that may be isomerized to the corresponding ketoximes.4952 The nitrosylation of cycloalkanes has also been accomplished using ultraviolet radiation; good yields of cyclohexanone oxine have been claimed for the reaction between cyclohexane and nitrosyl chloride.53-57 Because the cost of ultraviolet radiation is, and probablywill remain, high, it was decided to study the synthesis of cyclohexanone oxime using gamma radiation, the cost of which will probably decrease eventually. Results of this investigation are presented below. EXPERIMENTAL Gamma radiation was obtained from either the Co-60 source (3-kilocurie) or the MTR fuel element source. Unless specified otherwise, reagents were used as received. Nitrosyl chloride was obtained from the Matheson Chemical Company, cyclohexane from the Baker and Adamson Company, and formic acid'analyzed" grade) from the Baker Chemical Company. In experiments using carefully purified reagents, the cyclohexane was treated with a sulfuric-nitric acid mixture,54 washed, redistilled (B.P. 78.5-79.0~C), and dried over Drierite; the infrared spectrum was characteristic of pure (99%) cyclohexane. For these latter experiments, nitrosyl chloride was refluxed gently for 50 minutes, and then fractionated by a procedure outlined by Coleman and others.57 Hydrogen chloride was prepared by the reaction between concentrated sulfuric acid and hydrochloric acid, and dried by means of concentrated sulfuric acid. Before irradiation, contents of several tubes were degassed (see Table XXIII). Before charging, cleaned reaction tubes were heated to redness to remove residual organic materialo -------------------------- ~~58 -------------

The University of Michigan * Engineering Research Institute TABLE XXIII PREPARATION OF CYCLOHEXANONE OXIME Weight Weight Weigt,Temp, Total Dose Weght Yield Run NOC1/100 ml p Tol o6 Product, Y Comments Nol/en 0C rep x 10 -6 % _ 2Solventalline gprodu 1 2.0 0.0 0.66 - -- Crystalline product 2 2.0 0.0 0.66 -- -_ Crystalline product b 3 2.0 0.0 0.23 -- -- Crystalline product 4 2.9 0.0 0.32 0.07 3.6 Crystals + clear oil 5 2.9 0.0 0.82 0.04 1.9 Crystals + clear oil 6 3.2 -60 0.70 0.48 18 Clear oil 20 1.0c 0.0 2.0 - -- Little reaction 21 1.0d 0.0 1.8 -- -- Clear oil 22 1.0 0.0 2.0 - - Little reaction 23 1.0d 0.0 1.8 - -- Clear oil 24 1.0 0.0 2.0 -- - Little reaction 25 1.0 0.0 1.8 - - Little reaction 26 1.0 0..0 2.0 -- - Little reaction 27 1.0e 0.0 0.79 Clear oil 28 2.0' 0.0 0.79 -- -- Clear oil 29 1.0c,e 0.0 0.79 -- -- Clear oil 30 88 -- 0.86 - - Clear oil 31 8 -- 0.86 - -- Clear oil ~Based on amount of nitrosyl chloride charged. This run consists of supernatant liquid from run 2. dCyclohexane presaturated with dry hydrogen chloride. Cyclohexane presaturated with formic acid. Degassed before irradiation. 59

The University of Michigan * Engineering Research Institute Infrared spectra were obtained using a Baird recording spectrophotometero RESULTS AND DISCUSSION In the preliminary experiments with gamma radiation, tubes containing cyclohexane and small amounts of nitrosyl chloride (lo5 - 35o0) were irradiated at room temperature, the total dose received by each tube was 3 x 106 repo An oily product was obtained in the tubes, containing 3.0% nitrosyl chloride, However, satisfactory characterization of the oily product was not achieved, although treatment of hydrolyzed product with 2,4-dinitrophenylhydrazine yielded a hydrazone derivative that appeared similar to the hydrazone derivative of cyclohexanone. Because high yields of oxime are claimed to be favored by irradiation at low temperatures53 the next set of experiments was conducted at lower temperatures-0 and -60~C. Results of these experiments are given in Table XXIII. In runs 1, 2, and 3, the product was crystalline in nature,while in runs 4, 5, and 6, the product consisted of both crystalline and oily, or just oily, material. In all cases, the product was deposited on the walls and bottoms of the reaction tubes. The reaction was continuous; when the product was removed from run 2, further irradiation (run 3) led to the formation of more product. Some of the crystals from run 2 were recovered and recrystallized from petroleum ether. Although it had been expected that this product would be the oxime hydrochloride (M.P. 148-1500C), the melting point, 87-89~C, was very close to the melting point, 90-910C, observed for a sample of pure oxime prepared in the conventional manner from cyclohexanone. With runs 4, 5, and 6, yields were determined; these yields, even in run 6, are rather low considering the dose required. Also, yields are not consistent; for example, the yield for run 4 is greater than for run 5, although the dose received was lower. Attempts were then made to determine, first, the conditions governing the yielding of a crystalline rather than an oily product, and, second, the conditions required for optimum yields. It has been expected that a combination of low temperature, moderately low concentration of nitrosyl chloride, and moderate doses should lead to the best yields of crystalline material. Temperatures of irradiation were accordingly varied between 0 and -60~C, the dose between lo. x 106 and 2.0 x 106 rep, and the concentration of nitrosyl chloride between 0.7 and 20.0%.o To make sure that any product formed would exist as a hydrochloride rather than as a mixture of oxime and oxime hydrochloride, one sample was saturated with dry hydrogen chloride before irradiationo Results of these experiments were, however, unsatisfactory; no consistent relationships between the variables were observed, and the results were not reproducibleo No reason for this variability was found. In a final attempt to reproduce the successful experiments described 60

The University of Michigan * Engineering Research Institute above, carefully purified reagents were used. Reaction conditions are given in Table XXIIIo Although an oily product was obtained in each case, the amount was not as large as the largest amounts found previouslyo Presaturation of the cyclohexane with dry hydrogen chloride53 had no apparent effect on the reaction; the concentration of nitrosyl chloride also was unimportant No evidence of discoloration was observed. Again, satisfactory characterization of the product was not achieved. To avoid the mechanical losses associated with working up the product, infrared measurements were tried as a means of characterization. Unfortunately, the product was soluble only in solvents, such as benzyl alcohol, that absorb too strongly in the infrared regions of interest for identification of appropriate bands. Because of limited time, the low conversion of this reaction, and the difficulty of identifying the products, work on this problem was suspended. CONCLUSIONS 1. Gamma irradiation, at low temperatures, of a nitrosyl chloridecyclohexane solution leads to the formation of cyclohexanone oxime (or its hydrochloride). 2. The yields of this reaction are undesirably low; conditions for optimum yields are as yet unknown. C, Fixation of Nitrogen by Gamma Radiation INTRODUCTION Since the energy of gamma radiation from a Co-60 source is high enough (1.1 and 1.3 Mev) to activate or ionize both nitrogen and oxygen, the possibility of fixing nitrogen by irradiating mixtures of nitrogen and oxygen was considered. It was hoped that nitric oxide could be formed by the initiation of a chain oxidation of nitrogen by gamma radiation either at a lower temperature than is normally required, or under completely different conditions, for example, in the liquid state. Results of experiments designed to investigate this possibility are reported below. EXPERIMENTAL Both nitrogen and oxygen were of commercial quality. Other reagents were of ordinary laboratory grades, 61

The University of Michigan * Engineering Research Institute Analyses for nitric oxide were made using a Consolidated-Nier mass spectrometer, model 21-201. RESULTS AND DISCUSSION In the first experiment, a mixture of liquid nitrogen and oxygen (N2/02 = 2/1) was irradiated at a dose rate of 1.0 x 105 rep/hr; the total dose received was 1,8 x 106 rep. After irradiation, some of the nitrogen and oxygen was evaporated, and the residue examined in the mass spectrometer. Since only 100 ppm of nitric oxide were found,.it seemed clear that no chain reaction, without which the reaction could not be economically feasible, had taken place. Another sample (95% liquid nitrogen and 5% oxygen) was given a dose of 3 x 106 rep; again, yields were in the range of parts per million. Next, in an attempt to increase the efficiency of absorption of gamma radiation, a sample consisting of 98% nitrogen and 2% oxygen was irradiated with 1.5 x 106 rep in the presence of silica gel containing a cobalt salt as indicator. Once again, the yield of nitric oxide was almost negligible. Since the formation of nitrate ions or ammonia by the interaction between pile radiation and dissolved nitrogen and oxygen or hydrogen in the cooling water of the pile is known to occur, the irradiation of (1) gaseous nitrogen plus oxygen (95% N2) and (2) gaseous nitrogen plus hydrogen (95% N2)both in the presence of water-was tried. In addition, a sample of gaseous nitrogen plus hydrogen (95% N2) was irradiated in the presence of aqueous hydrogen bromide, (the bromide being used as a chain carrier). After a dose of 4.5 x 106 rep had been received by each tube, the gaseous hydrogen bromide was swept out from its tube and the pH of each liquid was tested; since the liquids were still neutral, it was assumed that the formation of nitrate ions, ammonia, or ammonium ions was negligible. These results are not surprising, for Harteck and Dondes58 obtained nitrogen tetroxide from the irradiation of nitrogen-ozygen mixtures in a nuclear reactor only at a high temperature (200~C), pressure (20-25.atm) and dose rate (3 x 108 rep/hr). Also Primak and Fuchs59 claim the radiation yield (analogous to quantum yield) for the production of nitric acid by the irradiation of humid air is unity; in other words, these authors found no evidence for a chain reaction, which would, of course, lead to a radiation yield greater than unityo Since the radiation-induced reaction between nitrogen and oxygen wou be of interest only if yields of nitric oxide would be high (that is, only if a chain reaction existed), no further work on this problem was done. 62

The University of Michigan * Engineering Research Institute CONCLUSIONS It is concluded that the irradiation of a mixture of liquid nitrogen and oxygen produces small quantities of nitric oxide. However, yields are undesirably low; no chain reaction seems to exist~ The addition to the liquid mixture of silica gel containing cobalt salts to absorb the radiation or the use of a chain carrier such as hydrogen bromide with gaseous nitrogen and oxygen does not appear to be helpful in obtaining a more efficient reactiono 65

The University of Michigan * Engineering Research Institute APPENDIX CALCULATION OF MOLECULAR WEIGHTS FROM SOLUBILITY DATA OF IRRADIATED POLYETHYLENE In this appendix details of the calculations used for the determination of molecular weights and distribution in polyethylene are given. The treatment* is based mainly upon assumptions and considerations used by Baskett, but leads to a numerical calculation of Mn instead of to a calculation by means of the graphical analysis suggested by Baskett (see II, section D, above; GENERAL Two assumptions are made initially: that the polyethylene is branched, and that the polyethylene is subject to chain scission, as well as crosslinking, during irradiationo Curve 1 of Figo 6 represents a typical experimental solubility curve for such a polymer. If, on the other hand, no branching is present and no chain scission occurs, a curve like curve 2 would be found; this curve corresponds to a linear polymer, having a normal molecular weight distribution (that is, M/Mn = 2), that does not undergo scission during irradiation. The basis of this treatment is the correction of the experimenta solubility curve for a real polymer for the effects of both branching and chain scission to find the curve for a corresponding hypothetical linear polymer that does not undergo chain scission. From this hypothetical curve, both Mw and Mn for the real polymer are estimated as described belowo CORRECTION FOR BRANCHING Let us assume that the experimental branched polymer can be obtained from a linear polymer by adding a certain amount of crosslinking, ao In othe words, let us assume that our experimental polymer behaves like a linear polymer that has been subjected to an amount of radiation equivalent to aoo To obtain the curve for this linear polymer, we add a constant number to the values for a in the experimental curve 1 so that the upper portion of the curve becomes a straight line on the log-log plot. It should be noted that with this plot the effect of this correction on the shape of the curve is greatest for the higher soluble fraction portion. *This treatment was developed by Mr, W. Wo Graessleyo -------- 64

The University of Michigan * Engineering Research Institute CORRECTION FOR CHAIN SCISSION Because of the fracture of main-chain bonds by radiation, the lower part of the experimental curve will eventually level off at high doses instead of continuing down as does curve 2 of Fig. 6. Although doses given in Fig. 6 are not high enough for this levelling-off to become evident, the data must still be corrected for the effect of scission. The pertinent formula to use is given by Charlesby,47 and may be expressed as (1 - x)observed 1 + a l (1 - X)corrected 2 Vc + In this relationship, 3 is the chain scission parameter (number of bonds broken per crosslink formedj the value of P for polyethylene is given as 0o2)047 The shape of the curve is most affected at the lower end. The final curve obtained after correction should approximate a straight line (curve 2) whose slope is the same as would be predicted for a polymer having a normal distribution and not subject to scission, Thus the corrected points are fitted to a line having the slope equal to the theoretical slope for a polymer of normal distribution and not subject to scissiono DETERMINATION OF MOLECULAR WEIGHT AND DISTRIBUTION According to the accepted theories of gelation,2 the weight-average degree of polymerization, DPw, for the real polymer is found from the intercept of the experimental curve, cc: DPw Oc The more complicated procedure for obtaining the number-average degree of polymerization will now be considered. First, the weight-average degree of polymerization, DPw, for the hypothetical polymer may be determined from the intercept at infinite solubility of the hypothetical curve (curve 2, Fig. 6): DPw wc + c0 21 44 This relationship is valid for all types of distributiono. Assuming a normal distribution, the number-average degree of polymerization, DPn, for the hypothetical polymer is then given by DP~ = n 2(%c + %o) We now have a curve for a hypothetical linear polyethylene sample, which upon irradiation is not degraded, but rather crosslinked by an amount %Q 65

The University of Michigan * Engineering Research Institute To determine the number-average molecular weight of the experimental polymer from the hypothetical polymer, we reverse the procedure used for correcting for the degree of branching. In other words, we add to the hypothetical polymer an amount of crosslinking, ao and determine the resulting number-average degree of polymerization. If an amount of crosslinking represented by ao is added to the hypothetical polymer, the number of molecules, N, is changed according to the following expression: N = No - number of crosslinks, where No is the number of hypothetical polymer molecules before crosslinkingo Then DPn (for the real polymer) is given by:l Dn = no. of monomer units present DPn n N 1/Dn - 1/2 ol 2ac + 3/2 ao The ratio of molecular weights (Mw/Mn = DPw/Dn) for the real polymer may now be determined: DPw 2ac + 3/2 aO DPn C SAMPLE CALCULATION FOR A-22 First to express our solubility curves in terms of the degree of crosslinking, a, we must convert the radiation doses to corresponding values of a. In the absence of other evidence, the conversion must be based on data given by Charlesby. According to Charlesby, exposure of polyethylene to a radiation dose of 45 x 106 rep should result in the crosslinking of 0.5% of the carbon atoms present (that is, a = 0o005).* Thus a = radiation dose (in rep) x 1.11 x 10o10o Next values of ao and ac are determined from the solubility graph. For sample A-22, *It should be noted that this value is uncertain because of the lack of preci molecular weight data 16 66

The University of Michigan * Engineering Research Institute aO = 4 x 106 nc = 1.5 x 106. = 14- x 10 = 14,000, 1.11 x (9 x 106) and - 14 x 10o10 1.11 x 1.5 x 106 67

The University of Michigan * Engineering Research Institute REFERENCES 1. Harmer, D. E., Anderson, L. C., and Martin, J. J., The Reaction of Chlorine with Aromatic Compounds Under Intense Gamma Irradiation, The University of Michigan, Engineering Research Institute Report 1943:4-41-T, Ann Arbor, May, 1955. 2. Manowitz, B., Chem. Eng. Progr., 50, 201 (1954). 3. Lewis, J. G., Martin, J. J., and Anderson, L. C., Chem. Eng. Progr., 50, 249 (1954). 4. Charlesby A., Proc. Roy. Soc., A, 215, 187 (1952); A, 222, 60 (1954). 5. Miller, A. A., Lawton, E. J., and Balwit, J. S., J. Phys. Chem., 60, 599 (1956). 6. General Electric Company, G. E. Irrathene, 101, April 27, 1954. 7. American Agile Corporation, Properties and Application of Agilene - HT, 1954. 8. Bretton, R. H., Progress Report IV (NYO-3311), Yale University, October, 1952. 9. Hayward, J. C., Ph.D. Dissertation, Yale University, June, 1955. 10. Buckley, G. D., and Seed, L. (Imperial Chemical Industries), Br. Patent 714,843, Sept., 1954. 11. Bell, E. R., Rust, F. F., and VaughanW. E., J. Am. Chem. Soc., 72, 337 (1950). 12. Dorfman, L., and Salzburg, Z., J. Am. Chem. Soc., 73, 255 (1951). 13. Volman, D. H., and Graven, W., J. Am. Chem. Soc., 75, 311 (19553). 14. Tobolsky, A. V., and Mesrobian, R. B., Organic Peroxides, Interscience Publishers, Inc., New York, 1954. 15. Goldschmidt, S., and Renn, K., Ber., 55, 628 (1922). 16. Raff, R.A.V., and Allison, J. B.,'"Polyethylene," High Polymers, Vol. XI, Interscience Publishers, Inc., New York, 1956. 68

The University of Michigan * Engineering Research Institute REFERENCES (continued) 17. Charlesby, A., Lecture at The University of Michigan, April, 1957. 18. Willard, J. E., Paper presented at the 131st meeting, American Chemical Society- Miami, April, 1957. 19. Chapiro, A., Comptes Rend., 237, 247 (1953). 20. Dole, M., Lecture at The University of Michigan, March, 1957. 21. Okamoto, H., and Isihara, A., J. Polymer Sci., 20, 115 (1956). 22. Charlesby, A., J. Polymer Sci., 10, 201 (1953); Nature, 171, 167 (1953); Nucleonics, 12, 19 (1954); J. Polymer Sci., 15, 547 (1954). 23. Charlesby, A., and Hancock, N. H., Proc. Roy. Soc., A, 218, 245 (1953). 24. Charlesby, A., and Ross, M., Proc. Roy. Soc., A, 217, 122 (1953). 25. Dole, M., Keeling, C. D., and Rose, D. G., J. Am. Chem. Soc., 76, 4304 (1954). 26. Lawton, E. J., Bueche, A. M., and Balwit, J. S., Nature, 172, 76 (1953); Ind. Eng. Chem., 46, 1703 (1954). 27. Simha, R., and Wall., L. A., J. Phys. Chem., 61, 425 (1957). 28. Ballantine, D. S., Dienes, G. J., Manowitz, B., Ander, P., and Mesrobian, R. B., J. Polymer Sci., 13, 410 (1954). 29. Charlesby, A., Br. Patent 732,047, June 15, 1955; Br. Patent 747,478, April 4, 1956. 30. Materials and Methods, 45, 167 (1957). 31. Hughes, R. L., Interim Report No. 55-20, Spencer Chemical Company, January 4, 1956. 32. Hughes, R. L., Interim Report No. 56-46, Spencer Chemical Company, August 1, 1956. 33. Alexander, P., and Charlesby, A., Nature, 173, 578 (1954). 34. Alexander, P., and Toms, D., J. Polymer Sci., 22, 343 (1956). 35. Alexander, P., Charlesby, A., and Ross, M., Proc. Roy. Soc., A, 223, 392 (1954). 69 ----------------------— 69 --------------

The University of Michigan * Engineering Research Institute REFERENCES (continued) 36. Meikle, J. B., and Graham, B., Rubber and Plastics Age, 37, 678 (1956). 37. Charlesby, A., Br. Patent 741,826, June 15, 1956. 38. Szwarc, M., J. Polymer Sci., 19, 589 (1956). 39. General Electric Company, Chem. Eng. News, 33, 5091 (1955). 40. Smith, P., J. Polymer Sci., 21, 143 (1956). 41. Griffin, D., and Rubens, L. C., Paper presented at the 130th meeting of the American Chemical Society, Atlantic City, September, 1956. 42. Immergut, E. H., and Mark, H., Makromol. Chem., 18, 322 (1956). 43. Ballantine, D. S., Glines, A., Metz, D. J., Mesrobian, R. B., Behr, J., and Restaino, A. J., J. Polymer Sci, 19, 219 (1956). 44. Chen, W. K. W., Mesrobian, R. B., Ballantine, D. S., Metz, D. J., and Glines, A., J. Polymer Sci., 23, 903 (1957). -45. Baskett, A. C., Paper presented at the International Symposium of Macromolecular Chemistry, Turin, 1955. 46. Charlesby, A., Proc. Roy, Soc., A, 222, 542 (1954). 47. Charlesby,A., J. Polymer Sci., 11, 513 (1953). 48. E. I. duPont de Nemours and Company, Information Bulletin, Alathon Polyethylene Resins, No. X-79a, July 10, 1956. 49. Lynn, E. V., J. Am. Chem. Soc., 41, 368 (1919). 50. Lynn, E. V., and Hilton, 0., J. Am. Chem. Soc., 44, 645 (1922). 51. Lynn, E. V., and Arkley, H., J. Am. Chemo Soc., 45, 1045 (1923). 52. Mitchell, S., and Carson, S. C., J. Chem. Soc., 1005 (1936). 53. Naylor, M. A., and Anderson, A. W., J. Org. Chem., 18, 115 (1953). 54. Ito, Y., Bull. Chem. Soc. Japan, 29, 227 (1956). 55. Austrian Patent 172,618, September 25, 1952. 70

The University of Michigan * Engineering Research Institute REFERENCES (Concluded) 56. U. S. Patent 2,719,116, September 27, 1955. 57. Coleman, G. H., Lillis, G. A., and Goheen, G. E., Inorganic Syntheses, 1, 55 (1939). 58. Harteck, P., and Dondes, S., J. Chem. Phys., 24, 619 (1956). 59. Primak, W., and Fuchs, L. J., Nucleonics, 13, 39 (1955). 71