Photo-initiated polymerization of N-vinylcarbazole in dichloromethane solution containing azobisisobutyronitrile

European Polymer Journal Vol. 18, pp. 285 to 288, 1982 Printed in Great Britain. All rights reserved

0014-3057/82/040285-04503,00/0 Copyright © 1982 Pergamon Press Ltd

PHOTO-INITIATED POLYMERIZATION OF N-VINYLCARBAZOLE IN D I C H L O R O M E T H A N E SOLUTION C O N T A I N I N G AZOBISISOBUTYRONITRILE R. G. JONES and N. KHALID University Chemical Laboratory, University of Kent at Canterbury, Canterbury CT2 7NH. England (Received 16 June 1981)

Abstract--Azobisisobutyronitrile, as a photosensitizer of the polymerization of N-vinylcarbazole in dichloromethane solution at 2 = 366nm, has also been shown to quench the polymerization that normally occurs in its absence. As the initiator concentration is raised, the polymer yields at first decrease and pass through a minimum before increasing to about twice the initial value, at the solubility limit of the initiator. The product polymers are predominantly formed through a cationic mechanism though a free radical fraction is also evident, particularly at high initiator concentrations. Kinetic mechanisms compatible with the experimental evidence are considered and indicate that the higher molecular weight cationic fraction is formed through the intermediacy of the first excited triplet states of both the monomer, and a complex of the monomer and solvent. Quenching of the former species by azobisisobutyronitrile results in initiation of the polymerization, but quenching of the latter species quenches polymerization. The low molecular weight radical fraction, on the other hand. is considered to be formed through the quenching of the first excited singlet state of the monomer by azobisisobutyronitrile.

INTRODUCTION In a recent paper [1] we presented a detailed kinetic analysis of the photo-initiated polymerization of N-vinylcarbazole (VCZ) in the presence of azobisisobutyronitrile (AZBN) in benzene solution. The major conclusions to be drawn from that study were that the polymerization propagated through non-interacting radical and cationic mechanisms and that these were respectively initiated by the quenching of the lowest excited singlet and triplet states of VCZ by AZBN. Though it has previously been recognized [2, 3] that, even in the absence of any added initiator, the photosensitization of the charge transfer polymerization of VCZ in dichloromethane solution can be realised, and that this would be a complicating feature, it was nonetheless felt to be worthwhile to study the V C Z - A Z B N system in that solvent. The higher dielectric constant of dichloromethane, exerting a greater stabilizing influence on any ion-pair initiating and propagating species that might be produced, should lead to polymers whose variations of molecular weight distribution with initial A Z B N concentration correlate with the proposed mechanism in a manner that is more pronounced, and thus more compelling, than those of polymers that are produced in benzene solution.

EXPERIMENTAL

Reactions were carried out at 303 K in a 3.5 cm dia., 1 mm path length silica cell at an irradiation wavelength of 366 nm. The path length and wavelength of illumination were chosen so as to avoid the build up of large concentration gradients during the course of reaction. In order that product polymers should be sensibly representative of initial conditions, conversions were held below 15~o. A reaction duration of 30 min was suitable for this purpose.

Initial concentrations were such that light absorption by AZB'N was responsible for <0.05~ of the total absorption. All materials, other apparatus and procedures have been described previously [2]. RESULTS AND DISCUSSION Figure 1 shows the gel permeation chromatograms of polymers prepared from 1 tool d m - 3 solutions of VCZ in dichloromethane containing A Z B N at concentrations up to a value (5 x 1 0 - 4 m o l d m -3) approaching the solubility limit. The corresponding molecular weight data and yields are listed in Table 1. The molecular weight distributions of all the polymers are broad and clearly bimodal. As the initial A Z B N concentration is raised to 3 x 10-s tool dm-3, the yield and the ratio of the low to high molecular weight fractions both decrease. As the A Z B N concentration is further increased, this pattern of behaviour is reversed. On the assumption that, as in benzene solution, propagation is through non-interacting radical and cationic mechanisms, polymers were also prepared in the presence of dimethylformamide (about 0.01 tool dm -3) to scavenge the cationic centres and thus characterize the two fractions. These products had a narrow monodal distribution with a peak molecular weight corresponding to that of the lower molecular weight fraction of the other polymers. Data for a representative polymer so prepared are included in Fig. 1 and Table 1. The higher molecular weight fraction, being suppressed by the inclusion of D M F in the reaction mixture, is presumably formed through a cationic centre, and a dispersity of about 1.7 for the remaining low molecular weight fraction is quite consistent with propagation through a radical mechanism. The order of molecular weight for these two fractions is the reverse of that which obtains in benzene solution.

285

286

R.G. JONES and N. KHALIO

(al

{el

/2 A

\

B

A I

B I

A

B I

{f}

Fig. 1. Gel permeation chromatograms of samples of poly (VCZ) prepared in dichloromethane solution at 303 K at representative AZBN concentrations (in moldm-3): (a) 0; (b) 2 x 10-5; (c) 3 x 10-5; (d) 10-4; (e) 5 × 10-4; (f) 10-4 ( + DMF). [VCZ] = I mol dm-3. The vertical deflections are refractometer readings in arbitrary units and the horizontal displacement is elution volume. (A) and (B) correspond to molecular weights of 40,000 and 2900 respectively.

Clearly, the cationic centres are stabilized by the higher dielectric m e d i u m whereas the p r o p a g a t i n g radical lifetime is reduced by chain transfer to dic h l o r o m e t h a n e (cf. C~ for styryl radicals to benzene and d i c h l o r o m e t h a n e at 323 K; 4 × 10 -6 a n d ~ 10 - 4 respectively 1"3]). The previously proposed m e c h a n i s m is summarized below, where V C Z is represented by M, the cyanoisopropyl g r o u p by R a n d excited states are distinguished by asterisks.

(iii) energy transfer (quenching of $1 by A Z B N ) 1M*+AZBN

k 2 , M + IAZBN*

(iv) deactivation of ~AZBN* IAZBN.

k,) A Z B N

(v) initiation

Radical scheme 1AZBN* + M

(i) excitation

M + hv

) 1M*

R'+M

(ii) excited state decay (both radiative, non-radiative and including that t h r o u g h intersystem crossing to T0 IM*

k. , R M . + R - + N2 ,RM"

(vi) propagation

k, ) M

R M i" + M

kp ) R M i + t .

Table 1. [AZBN] x

IO s

(mol d m - 3)

"o Yield ' overall

10 -4 x ~1n

Mw/M.

0 2 3 5 10 50 10*

7.50 2.00 0.82 4.58 13.43 13.59 3.68

2.79 5.49 5.59 3.27 2.86 2.11 0.45

4.3 3.3 2.6 3.5 5.3 5.7 1.7

* Including DMF in the reaction mixture.

Cationic fraction 10 -4 x ?,~, Yield M. Mw/M. 6.97 1.96 0.81 4.36 12.63 11.85 --

5.4 7.5 6.7 5.4 6.3 4.9 --

2.4 2.4 2.2 2.2 2.4 2.8 --

Radical fraction 10- 3 x % Yield M, M,/Mw 0.63 0.04 0.01 0.22 0.83 1.72 3.68

4.4 4.3 4.7 4.2 3.1 4.3 4.5

1.7 1.6 1.7 1.6 1.8 1.8 1.7

Radical to cationic yield ratio 0.091 0.023 0.016 0.050 0.067 0.143 --

Photo-initiated polymerization of N-vinylcarbazole R~,~. Thus in dichloromethane solution,

(vii) termination RM~" + RMj"

Rp = Rpr + Rp, + R~r + R'pc.

k' , polymer.

(viii) intersystem crossing *M*

~, ) a M *

(ix) triplet state deactivation 3M* X", M (x) charge transfer (quenching of Ta by AZBN) 3M* + AZBN k, ) M+ ' AZBN ~ (xi) donor-acceptor complex deactivation M'+,AZBN ~ k , , M + A Z B N (xii) initiation M +" AZBN;

(2)

At low concentrations of AZBN the ratio R),r/Rp, is approximately given by Krk}k6/Kc[AZBN] ½ and it is thus predicted that the radical to cationic yield ratio must initially decrease with increasing AZBN concentration. At high concentrations of AZBN, however. Rpr/R,,c is approximated by Krk~k.I[AZBN]/Kc and the reverse trend would then obtain. These predictions are in accordance with experimental observation. Tada and co-workers [3] have proposed that the photopolymerization of VCZ in dichloromethane solution without any additive proceeds via a cationic mechanism through the intermediacy of the VCZ triplet state. Our own results indicate that there is also a radical fraction formed; however, of greater significance is the fact that yields initially fall as AZBN is added to the system. Were the polymerization in the absence of AZBN to be initiated through quenching of the VCZ excited states by dichloromethane, then it can be readily shown that R'pr and R'p~ will take the forms defined by Eqns (3) and (4):

Cationic scheme

k, , RM + ... R N ;

R'pr = K'r(k, + k2 [AZBN])- i (xiii) propagation RM + .

287

(3)

R'r~ = K'~[(kl + k2[AZBN])(k6 + kT[AZBN])] -1.

RN2 . . + .M G

(4)

RM++I... RN~-

(xiv) termination RM + ...RN~-

~; , RMiNzR.

Instantaneous stationary state conditions of all transient species were assumed and the rate expression (II was derived Rp = Rp, + Rp¢

(1)

where

Rp~=K~(-kl +-k2~N]j[AZBN]"]1/2 and Rp~ = K~ (kl + k2[AZBN])(k6 + kv[AZBN]) " Rp is the overall rate of polymerization and Rp~ and Rp~.the radical and cationic contributions respectively,

(2k2k2k4Ia"] 1/2 [M]3/z k'pkskTk9I~ Kc -- k;(ks + k9) [M], and I, is the intensity of the light absorbed by VCZ. In order to simplify the interpretation of the present results in terms of the above expressions, we will temporize about the possible mechanism of initiation of the polymerization in the absence of AZBN, and merely acknowledge its occurrence by modifying Eqn (1) through inclusion of terms R~,~and

Substitution into Eqn 2 followed by the usual processes of calculus shows that Rp cannot initially fall and pass through a minimum as the AZBN concentration is raised. In order to reconcile this experimental observation, it is necessary to propose that the polymerization in the absence of AZBN is initiated through the intermediacy of at least one other excited state, and that this is more readily quenched by AZBN than either the first excited singlet or triplet states of VCZ itself. The products of the quenching reaction must also be stable to the initiation of polymerization. These requirements are met if a weak ground state charge transfer complex of VCZ and dichloromethane is postulated, an excited state of which is the precursor of the initiating species of the polymerization in the absence of AZBN. Such an excited state donoracceptor complex would have lower energy than the corresponding excited state of either component so it is further proposed that it is readily quenched by AZBN in a simple process in which the energy that is dissipated is insufficient to result in either bond fission within, or electron transfer to, AZBN. Assuming that this quenching reaction takes place at a diffusion controlled rate and that polymer formation mirrors the quenching kinetics, then the quantum yields for polymerization in the absence (4)o) and presence (4)(9 of a quencher Q will be related through the familiar Stern-Volmer expression, 4)0/4)o = 1 + 101° r [ Q ] where r is the excited state's natural lifetime. The initial yield of polymer is halved at an AZBN concentration of about 10 -5 m o l d m -3 so it follows that z ~ 10 -Ssec, a value that is consistent with Tada's proposal that the polymerization is initiated through the intermediacy of a triplet state. We assert, however, that it is in fact the first excited triplet state

288

R.G. JONESand N. KHALID

of a ground state complex of VCZ and dichloromethane that is responsible, and not the dynamic quenching by the solvent of the triplet state of VCZ itself, thus p~

(xv) excitation and intersystem crossing VCZ-CH2CIz

h,. , I(VCZ.CH2C12), , 3(VCZ. CH2C12)* Rpr

(xvi) initiation of polymerization 3(VCZ'CH2CI2)*

k,o, predominantly cationic polymerization¢

O

(xvii) quenching by AZBN 3(VCZ'CH2CI2)* + AZBN k,, ~ inactive products. The variations of R~,~ and R~,~ with AZBN concentration will then take the forms of Eqns (5) and (6): Rp~ = K'r(k I o + k l 1 [AZBN])- ~

(5)

R'p~ = K'c(klo + k l l [AZBN])-1.

(6)

The plots of Fig. 2, based on the variations predicted from the expressions for Rpc, Rpr, R~,c and R~,r show how it is possible for the polymer yields to fall to a minimum as the AZBN concentration is raised; they are not exact but represent the trends. Necessarily k ~ / k ~ o must be very much greater than k2/k i. The latter ratio we have previously reported [1] to be 53 dm 3 mol-~; so accepting that the reactive excited state of the complex has the calculated natural lifetime and that it is quenched by AZBN at a diffusion controlled rate, then k x l / k a o is given by 10a°z = l05 dm a mol- ~ and the requirement is indeed met.

CONCLUSION Over the years there have been many studies of photosensitized polymerizations of VCZ, particularly in the presence of well characterized electron acceptors. Products formed through both cationic and radical mechanisms are a well established common feature of these reactions, and the supreme importance of donor-acceptor interactions in both the ground and excited states is also realized. Because of the complexity of the systems, few detailed kinetic studies have been undertaken and a correlation of mechanism of polymerization with nature of excited state (singlet or triplet, simple or complexed) has t The small free radical fraction is not visualized as being formed via the singlet state as the formations of both fractions are apparently quenched by AZBN with about equal efficiency.

EAZBN]

Fig, 2. Sketch plots indicating the predicted variations of yield with azobisisobutyronitrile concentration at constant monomer concentration and light intensity according to Eqn 2 and the derived expressions for Rp¢, Rp,, R~,, and R'.. barely b e e n considered. From our studies of the VCZ/AZBN and VCZ/benzoyl peroxide systems [2, 5] in both benzene and dichloromethane solutions, however, there is an emerging pattern that might be extendable to the other systems. It can be summarized as follows: 1. If the first excited triplet state of the monomer is dynamically quenched by the initiator, the latter behaves as an electron acceptor and cationic propagation results. 2. If the first excited singlet state of the monomer is dynamically quenched by the initiator then energy transfer followed by homolytic bond fission occurs and free radical propagation ensues. 3. If the monomer and initiator form a ground state donor-acceptor complex, excitation of this species will always lead to cationic polymerization. A possible corollary to 3 that is suggested by 1 and 2 would be that the triplet state of the excited donoracceptor complex is responsible for cationic initiation, and that in those systems which form identifiable ground state complexes and on excitation give some radical polymerization, it is the singlet state that is responsible. The usually high efficiency of the cationic process would be attributable to a generally efficient intersystem crossing process. REFERENCES 1. R. G. Jones and R. Karimian, Polymer 21,832 (1980). 2. K. M. Z. AI-Abidin and R. G. Jones, J. chem. Soc. Faraday 1 75, 774 (1979). 3. K. Tada, Y. Shirota and H. Mikawa, J. Polym. Sci. I 1, 2961 (1973). 4. S. Tazuke, Pure appl. Chem. 34, 329 (1973). 5. K. M. Z. AI-Abidin and R. G. Jones, J. chem. Soc. Faraday I In press.