Chain-length distributions (CLDs) of polymers prepared by rotating-sector (RS) techniques under pseudostationary conditions were simulated for the case of chain-length-dependent termination and analyzed for their suitability of determining the rate coefficient of bimolecular chain termination ht. This was accomplished by first calculating (ht), the statistically correct but experimentally inaccessible "event-averaged" k(t), and checking whether a double-logarithmic plot of this quantity versus the mean chain length v' of the radicals in the moment of their termination was able to return the exponent b of the power law characterizing the chain-length dependence introduced in the simulations. After this test had been passed successfully, the same procedure was applied to experimentally accessible averages of k(t) such as <(k(t)(m))over bar>, an average calculated from the second moment of the CLD, and <(k(t)*)over bar>, the average resulting from the rate expression for RS polymerization. Both quantities proved to be excellently suitable to give con ect estimates of the average kr characterizing a specific experiment as well as to reproduce the exponent b although a slight tendency is to be noted for <(k(t)(m))over bar> to underestimate [k(t)]. The results are only slightly dependent on hero the chain-lengths of the two reacting radicals are averaged (i.e., which importance is assigned to the role played by the shorter one of the two chains), especially if the range of this averages is restricted to the more realistic cases. Taken in all, the determination of k(t) from RS polymerization data may at least keep up with methods based on the analysis of data originating from the recently more popular pulsed-laser polymerization (PLP), especially if some care is taken with respect to the choice of experimental conditions.
