Propagation Rate Coefficient of Poly(N-isopropylacrylamide) in Water below Its Lower Critical Solution Temperature Franc ¸ ois Ganachaud, † Robert Balic, ‡ M. J. Monteiro, § and R. G. Gilbert* Key Centre for Polymer Colloids, Chemistry School, Sydney University, NSW 2006, Australia Received April 7, 2000; Revised Manuscript Received August 28, 2000 ABSTRACT: Pulsed laser polymerization (PLP) of N-isopropylacrylamide (NIPAM) in water was performed over the range 2-20 °C (below its lower critical solution temperature), to obtain propagation rate coefficients (kp). While the value of kp deduced from these data obeyed some of the consistency criteria for PLP (e.g., that the multiple points of inflection give the same apparent value of kp), the apparent kp so obtained depended on monomer and initiator concentrations. For monomer concentration ∼0.4-0.8 M, the temperature dependence is approximated by kp(apparent)/dm 3 mol -1 s -1 ) 10 8.7 exp(-24.5 kJ mol -1 /RT). FTIR and osmometry measurements were used to infer the presence of significant amounts of dimer and to deduce the equilibrium constant for dimer formation. A model based on dimerization was derived to account for kp variations with monomer concentration but did not fit the experimental data, implying that a more complex treatment taking into account complexation with propagating chain ends or a bootstrap effect is required. Introduction Knowledge of rate parameters is essential for creating “designed” microstructures in free-radical polymer- ization. One important parameter is the propagation rate coefficient, k p . With the advent of pulsed laser polymerization (PLP; for recent reviews, see refs 1-5), accurate k p values can be determined for a wide range of vinyl monomers. The basic idea of PLP is as follows. Monomer (solution) and photoinitiator are irradiated by pulsed irradiation. Under favorable conditions a sig- nificant number of polymer chains which were initiated during one pulse will terminate with a short radical produced during the subsequent pulse, and others will be terminated by the second, third, ..., pulse. The degree of polymerization of the polymer chain terminated “instantaneously” by a radical generated by the nth pulse will be a value L n given by where [M] is the monomer concentration and t is the “dark” time between pulses (t -1 is the laser pulse frequency). Modeling molecular weight distributions under various conditions has shown 6-13 that L n corre- sponds closely (but not exactly) to a point of inflection on the GPC distribution w(log M). Although the technique is robust and simple, it is essential that the various criteria for reliability of the value of k p so obtained, identified by an IUPAC Working Party, 4,5,14 be obeyed (a point which is discussed in detail later). For example, data may be vitiated if the rates of chain transfer (especially to polymer, e.g., 15 a backbiting reaction which cannot be avoided even at low conversion but can be avoided at low temperatures 16 ) influence the molecular weight distribution (MWD) of the PLP sample: vinyl esters and acrylates exhibit a high rate of transfer to polymer, and a high pulse frequency and/ or low temperature is required to minimize these effects on the MWD. 16,17 Attempts have been made to apply PLP to polar monomers propagating in water, such as acrylic and methacrylic acids, 18 (meth)acrylamide, 19 and N-isopro- pylacrylamide (NIPAM). 20 Unfortunately, published PLP k p data for water-soluble monomers are not yet reliable and/or depend on quantities such as monomer concentration (and thus should be deemed apparent rate coefficients), for many different reasons: (i) Due to their polarity, these monomers are often insoluble in THF, and thus either water-phase GPC must be used, which is well-known to be subject to artifacts, or polymers must be derivatized, e.g., to poly- (methyl methacrylate) before GPC analysis, 21 which again is subject to artifacts (e.g., it is difficult to obtain totally quantitative conversion). (ii) Values of k p so obtained may depend on factors that should not play a role in the propagation reaction, such as monomer concentration (e.g., as reported for acrylamide 19 and methacrylic acid 18 ). (iii) A strong solvent effect has been observed in polar media, water exhibiting the greatest influence. 22 (iv) In addition to such technical issues, many of these data do not fulfill IUPAC recommendations 23 that k p determination should be independent of initiator con- centration. For example, acrylamide PLP data 19 were obtained prior to the development of reliable PLP criteria by the IUPAC Working Party, and tests such as the indepen- dence of apparent k p on laser pulse frequency were not carried out. Thus, the early acrylamide PLP k p values reported by Pascal et al. 19 cannot be deemed reliable until such checks are performed. The variations of apparent k p values determined for water-soluble monomers have been tentatively ex- plained either by the formation of dimers with different propagation rate coefficients or by the presence of † Present address: Laboratoire de Chimie Macromole ´culaire, UMR7610, Universite ´ Pierre et Marie Curie, Tour 44, 4 Place Jussieu, 75252 Paris cedex 05, France. ‡ Present address: Dulux Australia, PO Box 60, Clayton South, Vic 3169, Australia. § Present address: Department of Polymer Chemistry & Tech- nology, Eindhoven University of Technology, P.O. Box 513, 5600 MP Eindhoven, The Netherlands. * Author for correspondence and proofs. L n ) nk p [M]t (1) 8589 Macromolecules 2000, 33, 8589-8596 10.1021/ma000619l CCC: $19.00 © 2000 American Chemical Society Published on Web 10/25/2000