ENERGY TRANSFER IN POLYETHYLENE AND POLYETHYLENE-POLYBUTADIENE MlXTURES DURING GAMMA IRRADIATION * BY MALCOLM DOLE AND T. F. WILLIAMS Dept. of Chemistry, Northwestern University, Evanston, Illinois Chemistry Division, A.E.R.E., Harwell Received 14th January, 1959 Disappearance of initial vinyl or vinylene unsaturation on gamma-ray irradiation of highly crystalline linear polyethylenes follows a first-order decay law in the liquid state, but only over a limited dose range in the solid. Evidence is given that this decay must be the result of transfer of energy of excitation within localized spurs. Breakdown of the first-order law in the solid state when a certain concentration of radiolytically and randomly produced vinylene groups has been attained is explained on the basis of a pro- tective action of the latter.Concentrations of cis- or trans-1 : 4-polybutadiene up to 5 % have little effect on the radiolysis of crystalline polyethylene in the polyethylene + polybutadiene mixture, but in the liquid state the polybutadiene must interact both with ions and excited states of the polyethylene because G(H2) and k l , the vinyl group decay constant, are both lowered markedly. cis-Polybutadiene exerts less of a protective action than trans-polybutadiene. The difference between the behaviour in the liquid and solid states is explained on the basis of a more uniform distribution of the polybutadiene in the liquid and a greater frequency of collisions between the polybutadiene segments and ions or excited groups of the polyethylene. In 1953 Dole and Keeling1 noted that vinylidene-type double bonds -CH2CCH2-, about one per molecule, in a low density polyethylene disappeared II CHZ on irradiation in a heavy-water pile much more rapidly than would have been expected on a purely statistical basis.Thus, after an irradiation dose sufficient to liberate from -CH2- units 3.6 molecules of hydrogen per number average polyethylene molecule of molecular weight 32,000, essentially a11 of the vinylidene groups had disappeared. In a later more accurate paper, Dole, Milner and Williams 2 found that during gamma-ray irradiation, the initial rate of hydrogen evolution in molecules evolved per 100 eV of energy absorbed, G(H2), was com- parable to the initial G-value of vinylidene group elimination, despite the fact that the concentration of vinylidene groups, 0.414 x 10-4 mole/g was almost 2000-fold smaller than the concentration of CH2 groups, 7-1 x 10-2mole/g.It was also demonstrated 2 ~ 3 that a similar effect was observed for the disappear- ance of the vinyl double bonds in a high density polyethylene, Marlex-50. The decay of the latter initially followed the first-order law, - dPi]/dD = kiwi], (1) where P i ] represents the vinyl group concentration, kl the first-order constant and D the dose. The first-order constant was the same for two different poly- ethylenes whose initial vinyl group concentration differed five-fold. From the data of Lawton, Balwit and Powell 3 it appears that kl is also independent of dose rate. * The research described here will be submitted by one of us (T.F. W.) to the Uni- versity of London in partial fulfilment of the requirements leading to the Ph.D. degree. 74M. DOLE AND T. F . WILLIAMS 75 The selective reactivity of the vinyl and vinylidene groups has been inter- preted 2 in terms of transfer of excitation energy. The present paper deals with a further study of this phenomenon, not only with respect to pure Marlex-50, but also to Marlex-50 containing up to 5 "/o of either trans- or cis-1 : 4-polybutadiene. In addition a polyethylene containing initially only vinylene unsaturation, -CH= CH-, was studied. EXPERIMENTAL The trans-vinylene group concentrations were determined in this laboratory by the infra-red method while the percentage compositions were given by the supplier. Table 1 contains a list of materials studied and their composition.TABLE 1 .-COMPOSITION OF MATERIALS STUDIED Y.3 yo trans-1 : 4- Yo cis-1 : 4- [ trans-Vl] Marlex-50 polybutadiene polybutadiene moles/p x 104 supplier Phillips Petroleum Co. 100 0 0 0.06 98-5 1-35 0.15 2.1 95 4.5 0.5 7.0 95 0.2 4.8 0.3 Standard Oil Co. of Indiana *75 The vinyl group concentration in the pure Marlex-50 varied slightly from sample to sample, but was of the order 1 x 10-4 moleslg. The Standard Oil Co. of Indiana (SOI) material and the Marlex-50 were linear high- density polyethylenes. The Marlex-50 component of the Phillips Petroleum Co. materials was antioxidant free, but the polybutadiene contained about 1.5 % of phenyl-/3-naphthyl- amine antioxidant. As far as the experiments of this paper are concerned no difference could be detected in the behaviour of pure Marlex-50 with or without antioxidant.The radiation source, Co-60, associated equipment and dosimetry methods were the same as previously described.*, 4 The technique of measuring the concentration of the different unsaturated groups by the infra-red method was also the same as previously described 2 except for the cis-vinylene concentration determinations. Relative changes in the concentration of the latter were estimated by measuring changes in the optical density of the infra-red spectrum at 6.05 p. Inasmuch as the vinyl group absorbs at 6.1 p, the cis-vinylene group absorption had to be detected and measured from the height of the '' step " in the curve of the recording of the infra-red absorption in this spectral region.Measurements of changes in the height of this step enabled estimates to be made of relative changes in the cis-vinylene con- centration. Other techniques such as that used for the evaluation of G(H2) were the same as described previously.29 4 RESULTS The observed changes in concentration of the cis- and trans-vinylene groups at 142" as brought about by the gamma irradiation in all of the different systems studied are illustrated in fig. 1 and 2. If the concentrations of the vinylene groups in pure Marlex-50 at the same total dose at 142" are subtracted from the measured vinylene concentrations of the Standard Oil Co. of Indiana (SOI) polyethylene, the residual vinylene concentra- tions given in fig. 3 can be seen to decline with dose according to the first-order law.The method of plotting the data in fig. 3 is mathematically equivalent to assuming the validity of the zero-order growth and first-order decay law given in a previous publication.2 At room temperature (r.t.) the first-order decay law no longer holds for the disappearance of the vinylene groups that were initially present in the SO1 polyethylene. Instead, the semi-empirical expression used 2 to describe vinyl decay at room temperature in Marlex-50 was again found to be valid, see fig. 4. In eqn. 2, [Vl] represents the vinylene group concentration in units of moles/g, k2, the first-order decay constant for the disappearance76 ENERGY TRANSFER IN POLYETHYLENE FIG. 2.-Decay o f vinylene unsaturation in 5 % polybutadiene + Marlex-50 mixtures at 142".Open circles, trans-po 1 y b u t a d i e n e mixture ; solid circles, cis-polybutadiene mix- ture, The cis-curve is uncorrected for isomer- ization which accounts for only 11 % of the total decay. The truns- curve has been correct- ed for normal vinylene growth in the Marlex- 50 f r a c t i o n of the mixture . FIG. 1 .-Growth and decay of trans-vinylene unsatura- tion at 142°C : solid circles, Mar lex - 5 0 ; ha 1 f -filled circles, SO1 polyethylene ; open circles, Marlex-50 + adiene mixture. 1.5 % trans-1 : 4-polybut- dose, eV g-1 x 10-20 I I 8 I2 I 4 dose, eV g-1 x 10-20M. DOLE AND T. F. WILLIAMS 77 I I 1.8 FIG. 3.-Vinyleneunsaturation 0 decay in SO1 polyethylene at t~ 142" after correction for nor- 3 ma1 vinylene growth. + \o 1.7 dose, eV g-1 x 10-20 I I I I I I 0.2 05 1.0 1 - exp (- k2D) FIG.4.-Decay of unsaturation at room temperature plotted according to eqn. (2) and (3). Open circles, vinyl groups in pure Marlex-50 ; solid circles, trans-vinylene groups in SO1 polyethylene.78 ENERGY TRANSFER I N POLYETHYLENE of vinylene groups randomly produced by the irradiation, and the subscripts zero, infinity and M refer to zero and infinite dose and to the pure Marlex-50 respectively. For SO1 polyethylene, eqn. (2) takes the form where k2 represents the first-order decay constant for the vinylene groups initially present. For polyethylene, the value of kZ given in table 2 for the decay of the vinylene groups TABLE z.-FIRST-ORDER DECAY CONSTANTS FOR DISAPPEARANCE OF UNSATURATION g/eV x 1021 material Marlex-SO so1 5 % polybutadiene trans cis group r.t . 142 O r. t. 142' 142 O 142 O vinyl 1.61 2 . 0 9 1 *42 1 -92 time time In Marlex-50 trans-vin ylene 0 . 5 2 0.64 0.46 0-65 1.2 initially present was calculated to be 0.46 X 10-21 g/eV. In this calculation, [Vllo, was necessarily taken to be zero. At low doses where the exponential term of eqn. (2) can be replaced by (1 - kzD), eqn. (2) reduces to the first-order decay law, eqn. (1). I I 5 1 0 dose, eV g-1 x 10-20 FIG. 5.-Decay of trans-vinylene unsaturation in polybutadiene + Marlex-50 mixtures at 142". Open circles, 5 % trans-1 : 4-polybut- adiene ; half-filled circles, 1.5 % trans-1 : 4-polybutadiene. Both curves corrected for normal trans- vinylene growth in the Marlex-50 fraction of mixture. If the trans-vinylene growth curve in the Marlex-50 fraction of the mixture as estimated from the data on pure Marlex-50 was subtracted from the measured trans-vinylene con- centrations of the 1.5 and 5 % trans-1 : 4-polybutadiene + Marlex-50 mixture, the residual vinylene group concentration at 142" decreased with dose according to the first-order law, fig.5. Fig. 6 illustrates the influence at 142" of 1.5 and 5 % trans- and 5 % cis- polybutadiene on the decay of vinyl groups in the Marlex-50 + polybutadiene mixtures. Finally the effect of the cis- and trans-polybutadiene on the vinyl group decay constant kl and on G(H2) is illustrated in fig. 7.M. DOLE AND T. F. WILLIAMS I I I I I I 5 10 15 dose, eV g-1 x 10-20 79 FIG. 6.-Vinyl unsaturation decay in Marlex-50 at 142" as affected by polybutadiene.Solid circles, pure Marlex-50 ; horizontally half-filled circles, 5 % cis-1 : 4-polybutadiene ; vertically half-filled circles, 1 -5 % trans-1 : 4-polybutadiene ; open circles, 5 % trans- 1 : 4-polybutadiene. 1 I I 4 5 yo polybutadiene FIG. 7.-Effect of added polybutadiene on G(H2) and kl for vinyl decay in Marlex-50 at 142". Open circles, trans-1 : 4-polybutadiene ; half-filled circles, cis-1 : 4-polybutadiene.80 ENERGY TRANSFER I N POLYETHYLENE The various decay constants are collected in table 2. Data for the decay of the cis- vinylene groups are not given for reasons to be discussed below. Actually, the observed rate of decay of the cis-vinylene groups in the 5 % cis-mixture was the same within the limits of experimental uncertainty as the rate of decay of trans-vinylene groups in the 5 % trans-mixture, see fig.2. Data for the decay constants of the trans- or cis-vinylene groups at room temperature in the 5 % polybutadiene + polyethylene mixtures are also not given because the cis- and trans-vinylene concentrations changed so little with dose that such changes could not be accurately determined. DISCUSSION For pure Marlex-50 at room temperature the high G-values for vinyl decay were probably the result of activation of the vinyl groups by excitation energy released in localized regions corresponding to the spurs of fast electron tracks followed by reaction to form end-links, as previously postulated.2 Processes resulting mainly from ionization were probably evolution of hydrogen, vinylene double-bond formation and cross-linking.It was also postulated that as vinylene groups were produced by the irradiation in the Marlex-50, mostly in the crystalline regions since Marlex-50 film is about 85 % crystalline, these randomly created vinylene groups served to protect partially the vinyl groups from further decay. It was estimated 2 that each sphere of excitation corresponding to the dissipation of 100 eV of energy was large enough to contain approximately 10 vinyl groups at the initial concentration of the latter. The concentration of vinylene groups at the dose when the vinyl decay began to deviate from the first-order law was about two per sphere of excitation. As far as could be told from the data, the decay of the randomly created vinylene groups followed the first-order law over the whole dose range.A Ziegler-type polyethylene was studied in which the initial vinyl concentration was about 0.2 of that in Marlex-50 ; nevertheless despite this low vinyl concentration the initial decay followed the first-order law until the same dose was reached at which vinyl decay deviated from the first-order law in the Marlex-50 case. These doses in the two cases corresponded to approxim- ately equal vinylene concentrations. Similar effects were observed in the studies on the SO1 polyethylene in which the initial unsaturation was all vinylene. In this case the decay of the initially present vinylene groups was first-order until the concentration of the randomly created vinylene groups had again reached a value of about two per excitation sphere.The similarity in behaviour of the vinyl groups of Marlex-50 and the initially present vinylene groups of SO1 polyethylene suggests that the latter must have existed either near the ends of the molecular chains or in the amorphous regions. If they had been randomly located throughout the crystalline mass of the polyethylene, one would have expected no difference in resistance to the gamma radiation between the vinylene groups initially present and those randomly created by the radiation. For the Marlex-50 + polybutadiene mixtures at room temperature, the very little influence of the trans- or cis-vinylene groups of the polybutadiene on the vinyl decay, hydrogen evolution, and cross-linking in the Marlex-50 fraction, indicates that transfer either of electric charges or of excitation energy from crystalline Marlex-50 to the amorphous regions containing the polybutadiene has a low efficiency.This is probably partly due to the segregation of the polybuta- diene in the amorphous regions ; on a sub-microscopic scale its distribution must have been far from random. The behaviour of the pure polyethylenes and the polyethylene + polybuta- diene mixtures in the molten phase at 142" was considerably different from the room-temperature behaviour. In the first place, the vinyl group decay followed the first-order law over practically the whole dose range. Although the initial G-value and the kl constant rose somewhat with temperature, it is possible that energy transfer was less efficient, but that the probability of end-linking followingM .DOLE AND T . F . WILLIAMS 81 activation was greater so as to give an overall greater decay constant. G(H2) and C (cross-linking), G(X), both rose with temperature.4 It is unlikely that such an increase was due to less energy transfer to the vinyl groups, because the rate of hydrogen evolution, at room temperature at least, was independent of vinyl group concentration. We attribute the greater G(H2) and greater G(X) at 142" com- pared to the room temperature values as due to greater segmental mobility of the CH2 chains at 142" which enabled neighbouring -CH2- groups or -CH*- free radicals to approach closely enough for C-C bond cross-links to form, and to a smaller caging effect of liberated hydrogen, either atomic or molecular.In a research only partly published,s Arvia and Dole demonstrated that increase of dissolved hydrogen in crystalline polyethylene lowered the hydrogen yield, pre- sumably because of a back reaction involving molecular hydrogen. This idea finds confirmation in the work of Soviet workers 6 who have shown that deuterium is radiologically built into the polyethylene when deuterium gas and polyethylene are irradiated together. At 142" the polybutadienes, either cis or trans, but trans- greater than cis-, had a marked effect in lowering G(H2) and the vinyl-group decay. In the liquid phase there is greater mobility of the polybutadiene segments as well as of the methylene groups. This greater mobility must increase the frequency of random collisions and make possible charge transfer or transfer of excitation energy.The ionization potential of olefins is of the order 9.2 V as compared to about 10.2 V for long chain normal paraffinic hydrocarbons. The reduction in G(H2) we at- tribute to charge transfer while the reduction in kl for vinyl decay probably resulted from partial transfer of excitation energy from the bulk of the Marlex-50 to the polybutadiene molecules rather than to the vinyl groups of the Marlex-50. In other words the vinylene groups of the polybutadiene competed with the saturated paraffinic chains for positive ions and with the vinyl groups for the energy of excitation. It should also be pointed out that in the liquid phase a more uniform dis- tribution of the polybutadiene must exist than in the solid mixture where both crystalline and amorphous regions can be recognized.The concept of localization of chemical effects has been applied with con- siderable success to a study of the direct effect of ionizing radiation on macro- molecules of biological importance.8.9 Although the actual mechanism of inactivation is in doubt, estimates of molecular size have been made with fair precision based upon the empirical assumption that each primary ionization and excitation zone occurring at random within a single macromolecule is sufficient to cause its inactivation. Recent work 10 on the protective effect of glutathione on the enzyme catalase when irradiated in the dry state bears a strong similarity to the results on the protective action of polybutadiene on the vinyl decay of Marlex-50 in the liquid state given here.Calculation of the decay constant for pure catalase yielded a value of 5 x 10-21 g/eV which is of the same order as that observed for the decay constants of unsaturated groups in polyethylene. Although the presence of the polybutadiene depressed the vinyl decay constant of polyethylene in the liquid state, the polybutadiene itself showed considerable damage in terms of the disappearance of its vinylene groups. This type of pro- tective action in which the protector exhibits sacrificial damage is similar to the effect noticed by Alexander and Charlesby 11 on mixtures of polymethyl meth- acrylate and aniline. They also showed that aromatic ring systems built into polymers or hydrocarbons lowered the radiation damage in terms of either main chain breakage or cross-linking between chains.All the evidence would suggest that migration of chemical activity in large molecules ensuant upon irradiation in the solid state is largely the result of excitation or ionization transfer processes. The effect of added cis- and trans-1 : 4-polybutadienes in lowering G(H2) for polyethylene in the liquid state is qualitatively similar to the results obtained by82 ENERGY TRANSFER I N POLYETHYLENE Manion and Burton 12 for mixtures of cyclohexane + cyclohexene and cyclo- hexane + benzene. Their data were explained in terms of ionization and ex- citation transfer from the component of higher ionization or excitation potential to that of the lower. Benzene has lower values than cyclohexane and exhibits a form of self-protection in which the excitation energy is degraded to heat by internal conversion without accompanying chemical effects. It is clear that although polybutadiene participated in significant energy transfer from poly- ethylene, the subsequent activation of the polybutadiene resulted in extensive chemical changes within the molecule.Thus, although the net G(H2) was lowered there was a much increased disappearance of total unsaturation for the mixture. Energy transfer processes are therefore not always beneficial in terms of a decrease in total decomposition values, but depend on the inherent molecular properties and the variety of processes leading to the depopulation of excited states. Other possible mechanisms for vinyl and vinylene decay such as isomerization, scavenging by free radicals or atomic hydrogen, activation by ionic migration, etc., have been considered, but space does not permit a detailed explanation of why these mechanisms are not considered to be significant in the radiation chemistry of polyethylene and polyethylene + polybutadiene mixtures. Support of this research by the U.S. Atomic Energy Commission and the gift of materials by the Phillips Petroleum Company and the Standard Oil Com- pany (Indiana) are gratefully acknowledged. 1 Dole and Keeling, J. Amer. Chem. SOC., 1953, 75, 6082. 2 Dole, Milner and Williams, J. Amer. Chem. Soc., 1958, 80, 1580. 3 Lawton, Balwit and Powell, J. Polymer Sci., 1958, 32, 257. 4 Williams and Dole, submitted for publication. 5 Dole, Williams and Arvia, Proc. Int. Conference Peaceful Uses of Atomic Energy 4 Varshavskii, Vasilev, Karpov, Lazurkin and Petrov, Doklady Akad. Nauk. S.S.S. R., 7 Field and Franklin, Electron Impact Phenomena and the Properties of Gaseous Ions 8 Lea, Actions of Radiations on Living Cells (Cambridge Univ. Press, 2nd ed., 1955), 9 Pollard, Adv. Biol. Med. Phys., 1953, 39, 53. 10 Norman and Ginoza, Radiation Res., 1958, 9, 77. 11 Alexander and Charlesby, Nature, 1954, 173, 578. 12 Manion and Burton, J. Physic. Chem., 1952, 56, 560. 13 Miller, Lawton and Balwit, J. Physic. Chem., 1956, 60, 599. (Geneva, 1958), paper 818. 1958, 118, 315. (Academic Press, N.Y., 1957), p. 122. p. 69.