Kinetic study of an in situ inhibited polymerization using robust M-statistics

The presence of multicollinearity, leverage and strong outliers in kinetic data of polymerization systems interfere with correct estimation and interpretation of the kinetic parameters. In this work, a biased statistical technique based on M-estimation is developed to cope with such data sets. To validate the model, the in situ inhibited free radical solution copolymerization of isobutyl methacrylate (i-BMA) with lauryl methacrylate (LMA) is studied by differential scanning calorimetery (DSC). The “Pseudo-kinetic rate constant” (PKRC) method is employed to reduce the kinetic treatment of the copolymerization system to homopolymerization counterparts. A couple of inhibited polymerization experiments with four various levels of initial inhibitor/initiator molar ratio is designed. To monitor the kinetics of the inhibited polymerization system, the isothermal DSC experiments are carried out. Moreover, to realize the effect of the polymerization media on the rate of initiation, decomposition kinetics of initiator is studied by non-isothermal DSC and the results compared with in situ polymerization counterparts. It is indicated that robust M-estimation outperforms existing methods of estimation in terms of statistical precision and computational speed, while keeping good robustness properties.