Feb., 1963
RADIATION CHEMISTRY OF ISOTACTIC AND ATACTIC POLYPROPYLENE
linear segment of the 2000-atm. curve to be correct we find i = 0.93 and P = 5.0. Rlrasurements at 5000 atm. are only scmiquantitative. We cannot predict how P will vary with pressure, since both thc rates of diffusion and of decomposition of acctoxy radicals are involved. The variation of 1 with pressure is qualitatively in agreement with cq. 1 since it must decrease with pressure. Measurements at high pressure were limited by the
295
solubility of iodine, which we estimate to be mole fraction at 25’ and GOO0 atm., and this was qualitatively confirmed by separation of crystals from solutions at somewhat higher initial conccntrations. T o demonstrate the pressurc dcpendrnw of Wt, results of sevcraI iodinc mole fraction, appcar in runs, all at 2.5 x J’ig. 2 . f4inc.e P varics in a cwmplicatcd way with pressure and 2 is known at only two prrssurcs, no detailed analysis of the results in Vig. 2 is possible.
RADIATION CHEMISTRY OF ISOTACTIC AXD ATACTIC POI,YPl~Ol’YLESE. I. GAS EVOLUTION AND GEL STUDIES BY W. SCHNABEL ANI) M. DOLE Chenkical Laboratory of Northwestern University, Evunston, Illinois Received July 13, 1962 A study has been made of gas evolution and gel formation when both atactic and isotactic polypropylene arc irradiated by co-60 ?-rays in vacuo a t room temperature. The G-value for hydrogen evolution, 2.7, is somewhat less than for polyethylene, while G(CH4) is 0.08. The gas from the irradiation of atactic polypropylene contained more methane and less hydrogen than that from the isotactic sample. The gel data can be interpreted on the basis of the Charlcsby-Pinner gel equation, but very low values for cross-linking are obtained. We conclude that the cross-linking G-value, G(X), lies between 0.088 and about 0.3, that G(X) for isotactic polypropylenc is about 0.6 that of atactic polypropylene, that most of the hydrogen evolution results in double bond formation or possibly intramolecular cross-linking, and that G(S) is practically identical for atactic and isotactic polypropylene. These conclusions are for G-values beyond the gel point and are estimated assuming constancy of G-values with dose.
I. Introduction Continuing our study of the radiation chemistry of synthetic semicrystalline polymers’ polypropylene has been chosen as thc subject of this invest’igation. I’olypropylene is especially intercstirig for two reasons; first, it is a polymer that undergoes simult’aneous and approximately equal cross-linking and degradation during the irradiation; and second, it ca.n be obtained in a st,ereospecific isotactic form as well as in the atactic form with random sequences of d and I asymmetric centers along the chain. l’revious studics of the radiation chemistry of polypropylene have been carried out primarily by Black and Lyons2v3 and by W a d d i n g t ~ n . ~Black and Lyons found that tho intrinsic viscosity decreased initially on irradiation of isotact’ic polypropylene, but t’hat’subsequently gelat.ion occurred at higher radiation doses. They concluded that one vinylidene double bond was produccd for each main chain fracture. The validity of this and other conclusions of Black and Lyons is discussed below. Waddington4 measured swelling ratios, elastic moduli, and dose to the gel point for two different samples of polypropylene, nature unspecified, but] presumably isotactic polypropylene. GuptsL6~6 irradiat,ed isotactic polypropylene in a nuclear reactor to different’ neutron dosages and concluded from nuclear magnetic resonance (n.m.r.) measurements t’hai: “our previous inference (from experiments on irradiat’cd polyethylene) t,hat cross-links are produccd predominantly in the amorphous region is (1) liar the most recent riaper from this Laboratory on linear polyethylene see &I. Dole and 1’. Cr:icco. J. I’htjs. Chcm., 66, 193 (1RG2). (2) R. hI. I3luck and I < . -1. Lyons, A‘alure. 180, 1346 (1957). (3) R. M. I h o k and I < . .J. Lyons, I’voe. Roil. S o c . (London), 8 3 6 3 , 322 ( 1 HA!)).
(4) 1’. 13. Wnddiiiatun J. I’ulumer ( 5 ) 11. 1’. Qiii)ta. Kolloid Z.,174, 74 (1961). (ti) See also It. 1’. G u ~ i t n J. , I’hys. Ckcm., 66, 8-10
(1962).
confirmed by this experiment on polypropylene.” This is jn contrast to the generally accept,ed findings of Charlesby, Von Arnim, and Callaghan’ t’hat the Qvalue for cross-linking, G(X), wa,s t’he same for four diff went polyet’hylencs of diff erent degrees of crystallinity. A preliminary note on the electron spin rcsonance (em-.) spectra of polypropylene has been published by Libby, Ormerod, and Charlesby.8 F’ischer and Hellwegeg carried out a more extensive e.s.r. investigat,ion of irradiated stretched isotactic polypropylene and concluded that’the dominant, free radical foy irradiation at liquid nitrogen temperat’ure was the -C(CHJC112- radical, while on warming to room temperature or on irradiation at room temperature the dominant and persistent free radical was the resonance stabilized frce radical, -CH-C(CHs)CII-. This conclusion agrees with our observation that, D-I1 atom exchangelo between irradiated polypropylene end dcuterium gas occurs less readily than in the case of polyet’hylene. Presumably D-H exchange does not involve allyl type free radicals. Tsvctkov, Molin, and Voevodskiill also made e.8.r. studies. l‘hc oxidation of polypropylene during irradiat,ion was investigated by IOLYPROPYLET\rE
I
2.0
lo/
297
POLY PROPYLENE I POLYPROPYLENE I. ISOTACTIC FILM 2. ATACTIC 3. ISOTACTIC FLAKE 3.
,://y
0’
1.2 1.2
0.8
Fig. 1.-Solubility of polypropylene as a function of dose plotted according to the relationship of Charlesby and Pinner, eq. 4.
large, eq. 4 is not valid over any dose range except,, possibly, at enormously high doses. If eq. 4 is applied as is to the estimation of G(X), the values given in Table V are obtained TABLE V MAXIMUMAND MIXIMUMESTIMATES OF G(S) AND G(X)
--
From eq. 4---
Polypropylene type
G(9) 2G(X)
Atactic Sample 1 0.448 Isotactic Sample2(flake) .755 Sample3(film) .734
r r o m eq. 8 and 9 G(X) G(S)
G(X)
G@)
0.272
0.244
0.115
0.10:3
,181 .141
.273 .207
.068 .069
.lo2 .lo1
Charlesby and Pinner] 0 also have derived equations relating G(S) and G(X) for the initial condition that li of eq. 6 is equal to infinity. They call such a molecular weight distribution “pseudo-random.” Assuming an initial pseudo-random molecular weight distribution foir the atactic and isotactic polypropylenes, these relationti may be derived from the equations of Charlesby and Pinner
I
=
2[1 - [In (1
+ Y)l/Yl
(7)
where I is defined above, and l0OyN~ G(X) =-2r,IM,,o
G(S) = 2IG(X)
(9)
Knowing I , y may be calculated from eq. 7 and then inserted into (8) to calculate G(X) and G(S). Because the initial value of b of eq. 6 is neither 2 nor infinity but in the neighborhood of 5 , it might be thought that the G-values of Table V bracket the true values. However, there are several indications that G(S) initially must be quite high. If this is true, then the use of the initial weight average molecular weight to calculate G(X) is inadmissible and the G-values of Table V are without quantitative significance. IV. Discussion of the Results First of all it should be noted that the G-values for hydrogen evolution, Table 11, are somewhat smaller than those observed in the case of polyethyleneJZ0 2.8 as compared to 3.8 at room temperature. Furthermore, hydrogen evolution from atactic polypropylene was slightly less than that from the isotactic samples. Dew( 2 0 ) T. F. Williams and M. Dole, J. Am. Chsm. Soc., 81, 2919 (1959).
298
W, SCHNABELAND M. DOL&
hurstZ1 studied hydrogen and other yields from branched alkanes and concluded that the hexanes showed a regular decrease in hydrogen yield with increase in methyl substitution. However, 2-methylheptane and 2,4-dimethylpentaiie, the latter compound closest in structure to polypropylene, exhibited hydrogen yields equal to those for n-hexane, n-heptane, and n-octane. Two methyl groups substituted on the same or adjacent carbon atoms lowered the hydrogen yield significantly. Perhaps in the atactic polypropylene there are some ‘(head to head” structures which would reduce the hydrogen yield and raise slightly the methane yield, by analogy with Dewhurst’s results. There are several possible explanations for the smaller over-dl yield of hydrogen during the radiolysis of polypropylene as compared to polyethylene. There may be more scavenging of hydrogen atoms by double bonds or free radicals in the polypropylene case. Chain scission was definitely much greater during the irradiation of polypropylene with the production of more free radicals or double bonds due to this effect than in the case of polyethylene. The relatively lorn cross-linking G-value may partly be the result of the scavenging of free radicals by hydrogen atoms. As shown by Dole and Craccol molecular hydrogen probably reacts with free radicals to produce momentarily hydrogen atoms whether or not the material is being irradiated. Considering the energetics of various reactians as estimated from bond strengths and ionization potentialsZ2 of low molecular weight hydrocarbons in the gaseous phase as listed in Table VI, it appears that reactions of the parent ion to liberate molecular hydrogen are more endothermic in the case of polypropylene than in the case of polyethylene. This may explain, in part, the lower hydrogen yield on radiolysis.
+ + -CHZCHZ-
1. -CHzC&2.
+ -CH=CH-+ + --+ HZ + -CH=CH--+ Hz
CHs +
I
3. -CHZCHCH2CH3
I
4. -CH&HCHZ+-
+Hz + -CH2
where p expresses the breadth of the initial molecular weight distribution through the equation
p=-
(trans)
-0.15
(cis)
-
iHZ++ CH2CH?
I +
+Hz + -CH&=CH-
1 bo - 1
MnG(S)
2=-
AHo, e.v./moleoule
-
.I1
1
dx
.04
(21) H. A . Dewhurst, J. Am. Chem. Soc., 80, 5607 (1958). ( 2 2 ) D a t a chiefly from F. H. Field and J. L. Franklin, “Electron Impact PhenomenP,” Academic Press, New York, N. Y., 1957, a n d G. C. Fettis a n d A. F. Trotman-Dickenson, J. Chem. Soc., 3937 (1961). (23) D. A . Boyle, W. Simpson, and J. D. Waldron, Polumer, 2,338 (1961). (24) J. L. Franklin and F. W. Lampe, Trans. Faraday Soc., 67, 1449
~OONA
when r is expressed in units of e.v. g.-’. (10) with respect to s, we find
.38
The liberation of methane in the case of polypropylene is in agreement with previous work showing the tendency of side chains to be degraded from the main chain by radiation. For example, Boyle, Simpson, and Waldron23 demonstrated that the yield of butane-& on irradiation was a linear function of the number of butyl-dl: side chains per 1000 main chain carbon atoms in the solid polymer. Although the C-H bond strength usually is considerably greater than the C-C bond strength, the difference between the two may become much smaller or even reverse its sign in the case of the parent ion. Thus, Franklin and LampeZ4have quite successfully calculated the relative strength of one-
(1961).
electron C-C and C-H bonds for a number of hydrocarbons. I n many cases the C-H+ bond strength is less than that of C-C+ bonds. A complete understanding of the monomolecular decomposition rates requires also a knowledge of the frequency factors for the decompositions involved. I n addition, in the solid state the cage effect undoubtedly partly influences the yields; this effect possibly explains the reduced yield of methane in isotactic as compared to that in the atactic polypropylene. Although G-values for cross-linking and scission are given in Table V, there is considerable doubt concerning the quantitative validity of these data. The reasons follow. (A) Change of b with Dose.-In order for the Charlesby-Pinner equation (4) to be valid the initial molecular weight distribution should be random, b equal to 2, or the distribution as a result of degradation alone should be nearly random by the time that the gel point is reached. To lower b from its initial value of 5 to nearly 2 requires a much higher G ( S ) than given in Table V. Making use of eq. 29 of S ~ l i t . 0Inokuti28 ,~~ has derived the general expression for the change in b with dose due only to random degradation, namely
and
TABLE VI ESTIMATED REACTION HE~TS Reaction
Vol. 67
x3
+ s(1 +$)(I (I
+;)I
Differentiating
(13)
+;)l+b
or in the limit of zero dose when x
