Oct., 1950
COPOLYMERIZATION OF ACETYLENE DERIVATIVES WITH OLEFINS
I t has been suggested that these complexes are closely related structurally to those of the complexes formed by silver ion and aromatic substances. At the present time a structure is favored
[CONTRIBUTION N O .
101 FROM
THE
4681
for the product of interaction of bromine and the aromatic ring in which the centers of the ring and the two halogen atoms lie on a straight line, DAVIS,CALIFORNIA RECEIVED MARCH 6, 1960
GENERAL LABORATORIES OF THE ~JNITED STATES RUBBER COMPANY]
Copolymerization. XV. Copolymerization of Acetylene Derivatives with O l e h s . Retardation by Radicals from Acetylenes BY KENNETH W. DOAK
Much work has been done to determine the relative reactivities, in free radical copolymerization, of various olefinic monomers with different radicals. Mayo, Lewis and Walling2 have established a general reactivity series based on a consideration of monomer reactivity ratios. A polarity series, in which monomers are placed according to their ability to donate or accept electrons, has been established by a consideration of the products of the reactivity ratios. No quantitative data have been determined for the radical copolymerization of acetylene derivatives which would enable this class of monomers to be placed in the reactivity and polarity series for olefins. It was the objective of this work to determine the reactivity ratios for the copolymerization of the representative acetylenes, hexyne-1, phenylacetylene, and diphenylacetylene, with the olefins acrylonitrile and methyl acrylate. The copolymerization of styrene and phenylacetylene was also studied. In order to compare the reactivity of the double and triple bond, hexene-1 was copolymerized with acrylonitrile and methyl acrylate. Since the acetylene derivatives caused marked retardation in most copolymerizations, some experiments were carried out in order to gain some information upon the mechanism of chain termination. The rate of polymerization of three systems was determined as a function of the catalyst concentration, in order to test for the formation of radicals which do not propagate the kinetic chain. Thus, it was shown by Bartlett and Altschula that the rate of polymerization of allyl acetate is proportional to the catalyst concentration, instead of being a squarC root function, as occurs for most polymerizations. This presumably is due to the formation of allyl radicals which do not propagate the chain. The molecular weight of the polymers is independent of the catalyst concentration. Experimental Monomers.-The acetylene derivatives were obtained commercially. The diphenyhcetylene (m. p. 59-61') was used without further purification. Hexyne-1 and (1) This paper was presented at the Atlantic City Meeting of the American Chemical Society, September, 1949. For paper XIV in this series, see Wdling, fumdnps, Briggs and M a y o , Txrs JOITRNAL, T8, 48 (1960). (a) ~ r y ok, r r i ~and Wdling, #bid., 70, 1629 (1948h ( 8 ) Bvtlctt rad Albchul. W d . , W, 816 (1944).
phenylacetylene were carefully fractionated before being used. The physical constants were: hexyne-1, b. p. 71.8-72.0 (760 mm.), @D 1.3993; phenylacetylene, b. p. 75.2 (90 mm.), lZmD 1.5485. Styrene, acrylonitrile and methyl acrylate were commercial samples, redistilled before use. Physical constants closely checked literature values. Copolymerization for Reactivity Ratios.-Copolymerizations were carried out in evacuated tubes, with benzoyl peroxide as catalyst, as previously discussed by Mayo and Lewis.4 The amount of peroxide used was 0.1 mole %, TABLE I COPOLYMERIZATION OF ACETYLENES Time, Polymer ~ hr. analysis Expt. [Ml]on [MI]" [ M Z I O[MzI0 Methyl Acrylate (MI)-Phenylacetylene (MI)
1 2 ' 3 qb
85.22 85.23 40.42 40.62
71.53 71.73 37.68 37.95
14.20 14.32 57.27 55.35
11.25 70 11.38 70 54.72 145 52.72 145
63.66,63.49%C 63.61,63.71 75.97,75.89 76.41,76.44
Srrylonitrile (MI)-Phenylacetylene (MI)
8'
83.46 84.63 39.13 39.58
9 10 11
80.44 50.62 51.02
12 13 14
79.67 50.95 50.91
5 6 7b
76.42 77.54 36.87 37.32
14.09 14.09 64.72 54.99
11.40 97 11.23 97 51.96 145 52.22 145
15.40,15.l9%N 14.96,14.73 7.90 7.88
Acrylonitrile (M~)-Diphenylacetylene(MI)
72.78 20.17 20.03 15 46.48 50.04 49.76 160 47.96 49.85 49.63 160
25.0.24.707,N 21.56 21.31
Methyl Acrylate (MI)-Diphenylacetylene (MI) 65.58 19.980 19.913 15 56.11,56.26%C
42.48 49.92 49.74 42.07 50.01 49.84
48 48
57.57,57.33 57.28.57.25
Acrylonitrile (MI)-Hexyne-l (Md
15 16 17
121.13 114.68 73.74 73.05 120.14 114.27 73.91 73.26 49.94 48.42 45.82 45.61
54 R4
32
22.6507, i Y 22.51 21.80
Acrylonitrile (MI)-Hexene-1 (Md
18 19
59.29 59.75
20 21 22
50.54 35.59 35.23
23 24 25
50.02 49.66 49 49
44.11 38.02 37.19 47.12 37.97 37.20
3.5 24.30%X 3.5 24.08
Methyl Acrylate (MI)-Hexyne-I (MI)
33.39 46.72 45.35 75 16.51 31.64 29.49 115 14.83 34.72 32.31 115
58.10% C 58.90 39.03
Methyl Acrylate (MI)-Hexene-1 (M?)
18.56 43.41 38.72 17.71 43.27 39.22 18.39 43.32 39.36
4.6 59.59% C 4.5 59.08 4.6 59.09
Phenylacetylene (Homopolymerization)
26b 0
Millimolea.
46.81
44.28 160
0.2 mole % benzoyl peroxide.
except for a few experiments, indicated in Talde I , in which 0.2mole % was used. The copolymers of methyl acrylate with hexyne-1, hexene-1, diphenylacetylene, and phenylacetylene (low proportion), and of styrene with phenylacetylene, were purified by the frozen benzene technique.' Those from methyl acrylate and a higher proportion of phenylacetylene had a lower molecular weight, and the monomer mixture was removed under vacuum, at room temperature, from the polymer which formed thin layers in flasks. Drying was finished at 140' for a few minutes. A homopolymer of phenylacetylene was dried in the same manner. The copolymers of acrylonitrile with diphenylacetylene, hexyne1, and hexene-1 precipitated from the reaction mixture as insoluble powders, which were extracted with alcohol aiid hexane and dried a t 60' and 1 mni. The copolymers of acrylonitrile and phenylacetylene (low proportion) wcre dissolved in acetone and precipitated (three times) from hexane as finely divided powders. Those from acrylonitrile and phenylacetylene (higher proportion) were freed from most of the monomer mixture by vacuum distillation, then were precipitated as finely divided powders and dried. The data for these experiments appear in Table I. The graphical determination of reactivity ratios for phenylacetylene with methyl acrylate and acrylonitrile is s h o w in Fig. I .
0,s 1 /'
0.6
-
/
--+/----
+ I
/
4'
0.4
-
0.2
-
/ /
/
0.2
0.4 r2.
0.6
0.s
I
centration. The catalyst chosvii was l-azo-bis-l-cyclohexanecarbonitrile, which has a half-life at 60' of about 1300 hoursee The experiments were carried out by pipetting aliquot parts of a mixture of the two monomers into tubes containing different amounts of the catalyst. The monomers were degassed three times, sealed a t lo-' mm., and polymerized a t 60'. The styrene copolymers were precipitated from hexane (three times for the phenylacetylene system, four times for the diphenylacetylene system) and dried by the frozen benzene technique. The methyl acrylate copolymers were freed from the monomers by vacuum distillation from very thin layers of polymer, the drying being completed a t 65". This drying procedure appears to be adequate, since the methyl acrylate-phenylacetylene system, and phenylacetylene alone, have ail extremely low rate of polymerization ((j. Tables I wid I11 ) . The rate data are rcarded in Table I I I . TABLE 111 RATESOF COPOLYMERTZATIOS Moles/l x 100
0.000 0.193 37" .731 1 205 1 553 2 38 32
-188 ti
09
0.00" ,410 811 1.66 2 86 4 n8 0 u)uo
Fig. 1.-Reactivity ratios for copolymerization of phenylacetylene with methyl acrylate (- - -) and acrylonitrile (-); 12 is phenylacetylene radical.
0 385 0 855
Rates of Copolymerization.-Some experiments were carried out to determine the effect of small amounts of phenylacetylene upon the rate of polymerization of styrene. These experiments are summarized in Table 11.
3 29
1.668 488
Time, hr.
Yield, g.
%/hr.a
x 100
Mol. w t .
x
Phen ylacet ylene-St yreneb 100 0.043 100 ,223 2.04 100 ,296 2.89 100 ,513 5.46 ti 4 .510 8.74 lj4 ,661 11.58 0-1 ,821 14.69 29.6 ,561 21.46 29.6 .702 27.7 ,837 83.0 29.6 123.3 108 123.2 ,568 4.32 123.3 718 5.82 48 , :-; , .510 11.21 46 3 ,729 16.70 48 :; 048 32.29 I'hcnyl:ict~tyle~~e-Methyl Acrylate' 10& 0,0187 0.00 108 1731 1.53 108 ,312 2.94 85.5 ,441 5.43 85.5 I745 9.45 ,961 12.37 85.5 Dip henylacet y lene-Styrened 24.5 0.458 19.6 34.5 ,591 25.9 24.5 ,927 42.4 18.3 1.043 65.3
10-3
53
... 46 41 3(i 37 . . I
14 :33 63 31 ...
42 39 .,,.I ,
9.37 1.007 1.38 63,000 9.78 0.585 1.68 96,000 9.99 408 I .90 112,000 10.17 199 2.26 141,000 10.37 IO00 2.60 177,000 Seventeen hours at go', with 0.025 g. of benzoyl peroxide. Calculated from intrinsic viscosity in benzene, by the method of Gregg and Mayo, THISJOURNAL, 70, 2373 (1948).
0.434 , . . ... 0.794 167 1.604 3.19 142 1 77 18.3 1.283 d2.2 126 6 35 18.3 1.491 97.6 113 0.0 67.5 0,076 a Corrected for blank, and for small decrease in amount of monomers. 10.0-c~.samples (9.00 g.) of mixture of 10.0100 g. of styrene and 24.0 g. of phenylacetylene. cc. samples (9.40 g.) of mixture of 51.6 g. of methyl acrylate and1 5.3 g. of phenylacetylene. * 1 0 . 0 ~ samples ~. (9.31 g.) of mixture of 50.0 g. of styrene and 21.0 g. of diphenylacetylene. e Two sets of experiments were carried out a t different times to check reproducibility.
More carefully controlled rate measurements were made t o determine the rate of polymerization of a mixture of r t p n e and phcnylacetylene 89 a function of catalpst con-
Discussion Relative Reactivities of Acetylenes and Ole5s.-The monomer reactivity ratios for the
EFFECTOF Styrene, g.
TABLE I1 PHENYLACETYLENE UPON POLYMERIZATION STYRENE' Phenyl-
acetylene, g.
Yield, g.
OF
Molecular
wdghtb
(6) Lawla and Matheeon, T m JOVaNaL, T& 747
(lWO),
Oct,, 1950
COPOLYMERIZATION OF ACETYLENEDERIVATIVES WITH OLEFINS
4683
TABLEI V
MONOMER REACTIVITYRATIOSAND RELATIVEREACTIVITIES ri Ma 7a MI 0.26 * 0.03 Phenylacetylene Acrylonitrile 0.33 * 0.06 Phenylacetylene 0.62 * 0.02 Methyl acrylate .27 * 0.04 13.6 * 1.0 Acrylonitrile Diphenylacetylene ( .o)o 55 * 5 Methyl acrylate Diphenylacet ylene ( .o)" Acrylonitrile Hexyne-1 5.4 * 0 . 3 ( .O)" 11.2 * 2 Methyl acrylate Hexyne- 1 ( .OY 12.2 * 2 . 4 Hexene-1 Acrylonitrile ( .O)" Hexene-1 8.5 * 2 Methyl acrylate ( .OY 0 . 0 4 * 0.04 Acrylonitrileb .40 * 0 . 0 5 Styrene 0.18 * 0.02 Methyl acrylate' Styrene .75 * 0.03 a These r2 values mere assumed to be zero, and rl was calculated from the intersection with Calculated assuming maximum 11 of 0.08. ref. 8b. Ref. 8b. Ref. Sa.
Relative. radical reactivitres (M,) AcryloMethyl nitrile acrylate
1.00 1.oo
0.019 0.011 0.048
0.055 0.021
0.073 3.2d 3.5 the rl axis accordingly.
')
systems studied are recorded in Table IV. Relative reactivities of the acetylenes with the acrylonitrile and methyl acrylate radicals were determined from the reciprocals of YI, phenylacetylene being taken as unity. Data for hexene-1 and styrene (data for the latter taken from the literature) make it possible to compare the reactivities of the acetylenic and olefinic bonds toward attack by free radicals. The acetylenes have the following order of Feactivity toward both the acrylonitrile and methyl acrylate radicals : phenylacetylene > hexyne-1 > diphenylacetylene. Thus, substituting a phenyl group for an alkyl group greatly increases the reactivity of the acetylenic bond, possibly because of the contribution of structures such as H C = C= ~ .to the transition state. The secR ond phenyl group causes a large decrease in the reactivity. This decrease presumably is due to steric hindrance. Nozaki' postulated that symmetrical substitution always caused deactivation due to steric hindrance, although Mayo, Lewis. and Walling2 found that diethyl fumarate is more reactive than methyl acrylate toward the styrene radical, which is a good donor, but less reactive toward a poorer donor radical such as vinylidene chloride. Hexyne-1 and hexene-1 have comparable reactivities (within experimental error) toward the methyl acrylate radical, but hexyne-1 is about twice as reactive as hexene-1 toward the acrylonitrile radical, which is a stronger acceptor than methyl acrytate. It thus appears that hexyne-1 is a better donor than hexene-1, presumably because the aliphatic triple bond is more easily polarized than the double bond. Styrene is over three times as reactive as phenylacetylene toward both the acrylonitrile and methyl acrylate radical, showing that in a conjugated system the olefinic bond is more reactive. Thus, it appears that resonance structures such as R - c H ~ CH ~ in styrene make a greater contribution to the ac(7) Nozaki, J. Polymer Sci., 1, 455 (1946).
6.
tivated com lex than do structures such as R-CH=C in phenylacetylene. A comparison of the products of rl and ra indicates that styrene and phenylacetylene have about the same alternating tendency with methyl acrylate. With acrylonitrile, a better acceptor, styrene appears to alternate better than phenylacetylene. A conjugated double bond thus appears to be more reactive and a better donor than a conjugated triple bond. Mechanism of Chain Termination-The data in Table I show qualitatively that the rates of copolymerization of phenylacetylene with methyl acrylate and acrylonitrile are much slower than the corresponding systems containing styrene.* The data in Table I1 show that adding small amounts of phenylacetylene (up to about 10%) reduce the over-all rate of polymerization of styrene by a factor of about two, accompanied by a decrease in the molecular weight of the polymer. This relation suggested that phenylacetylene formed radicals which do not readily propagate the chain, but have a strong tendency to terminate chain growth. Experiments (Table 111) were carried out to determine the rate of polymerization of a mixture of styrene and phenylacetylene as a function of the catalyst concentration. 1Azo-bis-1-cyclohexanecarbonitrilewas used as a catalyst; because of the long half-life, about 1300 hoursI6a constant rate of radical formation will result over fairly long periods of time. In a monomer mixture containing 19.7 mole % phenylacetylene, the rate of copolymerization, when plotted against the catalyst concentration (Fig. 2), gives a curve which obeys the empirical relationship Rate = K1[CIL/y &[C] Ki = 0.334,Kz 4.06 in which rate is expressed in %/hour by weight, and [C], the concentration of the catalyst, in .moles/liter. This expression can be derived by
+
(8) (a) Lewis, Walling, Cummings, Briggs and Mayo, THIS 70, 1519 (1948); (b) Lewis, M a y o and Hulse. ibid e?, 1701 (1945).
JOURNAL,
KENNETHw.D0'4.K
48x.Q
Val. 72
Dividing (31 by i4) and eliminating S . arid P., one obtains
3
x
Since the rate constants, and S and P are constant (ti'
I C1.l
l
1
1.7
j
0.02 0.03 0.04 o.ori mi i C I Tmolesiliter Pig. 8.- -Kate of polyirierizdtion of styrene atid phenylxetylene: - -, calcd. for rate = 0.334 z / C 4.ObC. 0,one set of simultaneoui evperimerits, 8 , second srt of , 0 , moleculai weight5 experiments, Table H I , 1101
+
making the following assumptions : (a) constant rate of radical formation; (b) chain propagation occurs by attack of radicals upon both styrene aiid phenylacetylene; ( c ) the radicals formed from a phenylacetylene unit are relatively unreactive, some propagating the chain by attack principally upon styrene, others coupling t o terminate the chain. (d) I n the steady state, the concentration of any species of radicals is constant. The reactions are (
h f S ~,
I,
P Jrh
k, --+ s
--
Itlltlatloll
k:
+s. k
p. (l'lop'tg~lllorl
ki
-3-s k
P. t- lJ
--f
x
1 trtrllllcLtlDl1
C is catalyst concentrahon, S and S. are styrene and styrene radical concentrations, P and P are phenylacetylene and phenylacetylene radicals. The reactions of catalyst with P, and P. with P are neglected, since they are probably small in n system containing a four-fold excess of styrene. Tf the rate of formation o f radicals eqiials t h e r x t e of destruction, then kiC= k P -
11
Since the rate of formation of P. IS equal to tht rate of destructioii knSY kd't~.ht k,P $2 The rate of polymer formation, neglecting the addition of phenylaeetylene to the phenylacetylene radicall,is represented by d Pol/dt
5
kzSS
+ kaSP + k4P.S
(3 1
'The rate of formation of radicals 15 t l C cit = k l c
Equation (6) is the differential form of the enipiriral relationship observed experimentally. If k h is assumed to be Lero (in the case that the phenylacetylenc: radical does not add to styrttriei, (6) heconies dPol/dC = k-", equivalent t(J the polymerizatiori of allyl acetate. If one assunies that P. is a radical formed from phenylacetylene hy chain transfer instead of by copolymerization, the kinetics are unchanged. They merely indicate that phenylacetylene forms radicals which do not readily propagate the kiiietic chain. 'The molecular weights of the polymers decreast with increasing catalyst concentration, then approach a constant value. Thus, a t high catalyst concentration, the chain termination appears t u be malogous to t hac in the polymerization o f allyl
'
.i~'ekdte.
'The extent oi copoly iiierizatioii in this system is unknown, due to lack of a satisfactory analytical inethod. However, if phenylacetylene is nearly one-third as reactive as styrene toward the styrene radical (Table IV), the polymers should contain uearly one-twelfth phenylacetylene (assuming no alternating tendency). This much cannot bt. incorporated as end-groups, since the molecular weights are over 30,000. Price and Greene") have found the reactivity ratios for the system 2vinylpyridine and phenylacetylene to be 4.0 and 0.2, respectively, for the two radicals, showing copolynierization as well as retardation. Walling, et al., *Od have shown 2-vinylpyridine and styrene to have the reactivity ratios 0.55 and 1.14. respectively, for the two radicals. The system styrene-diphenylacetyleneshows a rate of polymerization which is proportional to the square root of the catalyst concentration (Fig. 3), indicating normal radical termination. l 1 'The rates of polymerization are higher than for the system styrene-phenylacetylene by a iartor of 7-13. Since the reactivity data with methyl acrylate and acrylonitrile indicate that dipheriylacetylene is much less reactive than phenylacetylene, it i s likely that diphenylA ralculation of kr/k, gives au unreasonable value, if ks is ab the rate constant for addition of the styrene type radical to phenylacetylene. If one asaumes that phenylacetylene forms radicals both by copolymerization and transfer, and that the latter type radicals do not readily propagate the kinetic chain, a relationship Of the form shown by (6) can be derived, in which the constants have a more complex form. (10) Price and Greene, J . Polymer Sci., in press (loa) Walling, Bripg5 and W'olfstirn. THISJOWRNM, TO, la&.( fVj
,umed t o b e
1948) 111
(4)
l o r till.
IIic
c u r b , doer !lot c x r r a p o l t L < to thc o n g l u i l l h l prezrni t , m e
15 LIIIIIIIM-I1
Itu
ied.01
COPOLYMERIZATION OF ACETYLENEDERXVATIVES WITH OLEFINS
Oct., 1950
acetylene copolymerizes with styrene to a negligible extent. The number of diphenylacetylene radicals formed, even if they are relatively unreactive, is thus probably too small to affect the normal square root termination. The system phenylacetylene-methyl acrylate, with 20 mole yo of phenylacetylene, also shows a rate of copolymerization which is proportional to the square root of the catalyst concentration (Fig. 3).11 These two monomers were shown by reactivity ratios to copolymerize readily with a considerable tendency to alternate. Thus, chains ending in phenylacetylene radicals are definitely formed, but the square root relationship indicates that there is no specific coupling of phenylacetylene radicals. However, if chain termination is by cross-termination, the square root relationship will result, as shown in the following derivation
k:, M.+P + P
+P
P.
Me
ks ---f
P.
ks + Pa + X
C, ,M, Ma, P, and Pa represent concentrations of catalyst, methyl acrylate monomer and radical and phenylacetylene monomer and radical. The steady state relationships are : k5M.P = k7P.M k,C = k8P.M.
(7) (8)
4685
0.10
1.0
0
,
0 0.8 f
2 0.08
e 0.06 E
E
E \
0.6
W
2U 0.04 U
0.4 % 3:
0.02
0.2 I
I
I
0.06 0.10 0.15
I
0.20
I
]
0.25
70 Adenine, % 10.0 Thymine, 8.4 The molecular weight of thymus DNA has been variously reported, but a value of 35,000 was used for calculations, based on the measurements of Jungner, Jungner and Allgen' and Hammarsten.* The acid- and alkali-treated samples of DXA mere prepared according to Gulland, Jordan and Taylor.$ Analysis for alkali-treated sample : S , 14.1; P, 9.0; for acid-treated sample N, 13.3; P, 8.0. The rosaniline was a commercial sample obtained from the Allied Chemical and Dye Corporation. After rec.rystallization from water, the nitrogen and chlorine analybes indicated 98 t o 100% purity. In 0.05 M potassium phosphate huffer the extinction coefficient was 79,600 a t pH 6.6 and 66,300 a t PH 6.7 (5400 A.1. It was shown to obey Beer's law under the conditions of the study. Anal. Calcd. G H m N ~ C I : N, 12.48; C1, 10.51. Found: iu, 12.63; C1, 10.31. Method.-The binding of dye by DNA was determined by the method of equilibrium dialysis. Experiments were carried out at PH 5.6 and 6.7 in 0.05 M potassium phosphate buffer. DNA solutions varied from 0.05 t o 0.2V0. Five milliliters of DNA solution in 0.06 M buffer contained in a Visking cellophane bag were immersed in 5 nil. of dye in 0.05 ~ buffer. li A group (ca. 24) of testtubes was placed in a shaking device overnight which was sufficient time for equilibrium t o be attained. The o tical density of the solution of free dye (outside the bag? was determined in a Beckman spectrophotometer, Model DU, a t w v e length of 540 mfi and the concentration calculated. Results were reproducible t o within about 3%. Experiments were carried out at 3 * 0 . 5 O , 27 * 1' and 32 lo. Concentrations of DNA and dye were chosen such that a large proportion of dye was bound with respect t o free dye concentration. The amount of dye adsorbed by the cellophane casing at each equilibrium concentration wab determined from separate runs and found t o be about 20y0 of the free dye concentration a t PH 6.7. At PH 5.6 the cellophane adsorption ranged from about 30% at lon ilvr- concentrations to 15% at high dye concentrations.
