No abstract

The pulsed-laser polymerization (PLP) with subsequent GPC analysis of the resulting polymers was applied to the bulk polymerization of styrene and methyl methacrylate in the temperature range 25−70 °C. The chain length dependence of the rate coefficient of chain propagation k p previously found in 25 °C experiments was confirmed; its extent increases even slightly with temperature. The effect is interpreted to be caused by a progressive displacement of monomer from the proximity of the radical chain end by the rest of the chain, thus decreasing the local monomer concentration which results in a formal decrease of k p if the average monomer concentration is inserted into its calculation. Aided by computer simulations based on this concept this decrease could be modeled and parametrized; the difference between the extrapolated values of k p for zero and infinite chain length k p(0) and k p(∞) amounts to about 40−60% of k p(0), somewhat depending on temperature, monomer, and modeling function. The variation of the k p data actually observed is markedly less, of course (about 25−35%). The chain length for which the decrease from k p(0) to k p(∞) is 50% complete was calculated to be on the order of 100, somewhat higher for the (stiffer) poly(MMA) chain.

On the basis of recently developed numeric procedures, the results obtained for the propagation constants kp by means of the PLP-SEC method (pulsed laser polymerization followed by size exclusion chromatographical analysis of the chain length distribution) were subjected to a critical reevaluation and reanalysis particularly with respect to the chain length dependence of k p. In this context the errors arising from a broadening caused by axial dispersion in the SEC were largely eliminated together with other uncertainties in assessing the characteristic chain length L 0 on which the evaluation of kp is based. The main result that refers to the bulk polymerization of methyl methacrylate and styrene in the temperature range 25-70 °C is that the previously found negative long-range chain length dependence of k p survived this procedure to an extent that widely exceeds the known variation of kp for extremely short chains and extends over a range of several hundreds of propagation steps.

A survey is given on a selection of recently developed methods for the evaluation of the rate coefficient kt of termination and its chain‐length dependence. In particular these are the time‐resolved single‐pulse pulsed laser polymerization (TR‐SP‐PLP), the single pulse pulsed laser polymerization in combination with the analysis of the molecular weight distribution produced (SP‐PLP‐MWD), the methods yielding an average kt either from the second moment of the chain‐length distribution (CLD) or from the rate of polymerization, and a method focusing on the chain‐length dependence of kt consisting in an analysis of the CLD resulting from PLP experiments carried out at low pulse frequencies (LF‐PLP). The results obtained by these methods are compared and discussed. The role of the shielding of the two radical chains by their appendant coils is emphasized.

The chain length dependence of bimolecular termination in the solution polymerization of styrene was examined at 25 °C with the solvents (50% per volume) toluene, cyclohexane, bis(3,5,5trimethylhexyl)phthalate, and ethyl acetate. Irrespective of which method was applied, the "good" solvent systems with toluene and ethyl acetate exhibited a chain length dependence comparable to or even greater than that observed for the bulk polymerization of styrene, while a markedly weaker chain length dependence was observed with the "bad" solvents cyclohexane and bis(3,5,5-trimethylhexyl)phthalate. This is documented by the exponent b in the power law k h t ) A‚ν′ -b connecting the average termination coefficient k ht and the chain length ν′ of the radical chains at the moment of undergoing bimolecular termination: b ≈ 0.2 for styrene (in bulk) and in toluene solution, b ≈ 0.3 in ethyl acetate solution and 0.10 to 0.14 in cyclohexane and bis(3,5,5-trimethylhexyl)phthalate solution, respectively. Only in the case of the bad solvents the bulk viscosity and k ht or the prefactor A are inversely correlated. In the case of the good solvents, the bulk viscosities were too close to detect such a relationship. The observed values of the parameter b (with the exception of the ethyl acetate system) are consistent with the idea of a chain length dependent shielding exercised by the rest of the chains on the two radical chain ends (some sort of "kinetic" excluded volume effect) which is much weaker in bad solvents.