Mechanistic Studies On Reversible Addition-fragmentation-chain Transfer Free Radical (Co)Polymerization

Reversible addition-fragmentation chain transfer (RAFT) free radical polymerizationhas developed into one of the most important methods of controlled/living radicalpolymerization since its invention in 1998. The basic principle of the RAFT process isto protect the majority of propagating radicals from bimolecular termination reactions byreversible chain transfer to dormant chains with so call "RAFT agent". Up to now, thebasic reaction scheme involved in the RAFT process is still not very clear. There existsa debate regarding the magnitude of addition/fragmentation rate constants.In this thesis, the RAFT parameters, such as equilibrium constant K, chain transferconstant C_tr and group parameter (k_c/k_t)K were determined inhomo-polymerization systems. These parameters assisted in elucidation of the RAFTmechanism. A RAFT-mediated copolymerization model was then developed based onthe elucidated mechanism and elementary reactions. The model included the theoreticalequations of apparent equilibrium coefficient and apparent chain transfer coefficient.The model predicted polymerization kinetics and was verified by carefully designedexperiments. The type of RAFT agent in the copolymerization appeared to be veryimportant and its selection was discussed.In this work, three monomer types: styrene (St), butyl acrylate (BA) and methylmethacrylate (MMA), and three types of RAFT agents: benzyl dithioisobutyrate (BDIB.With Z=CH(CH₃)₂), 1-phenylethyl phenyldithioacetate (PEPDTA with Z=CH₂Ph) and2-cyanoprop-2-yl dithiobenzoate (CPDB with Z=Ph) were systematically investigated.The parametersK, C_tr and (k_c/k_t)K were determined for different monomer/RAFTsystems. Combined with some literature data, it was found that the electronic propertiesof Z group strongly affect the magnitude of K and C_tr. In the BA polymerizationsystem, K and C_tr decreased with the radical-stabilizing substituent from phenyl tobenzyl to isopropyl. This suggested a strong effect of the Z-group on fragmentation.With the same type of RAFT agent, the equilibrium constants for different monomer systems were in the order of K_MMA < K_St < K_BA, with BA suffered the most severe rateretardation effect. The rate constants of addition and fragmentation were related to thestability of propagating radical and Z group. With the same monomer type, the additionrate constant of isopropyl Z group was lower than that of benzyl Z group, which wasopposite to the fragmentation rate constant.Four assumptions were made in developing the RAFT copolymerization model.Firstly, a quasi-steady state of the total radical concentration (propagating andintermediate radicals) was satisfied. Secondly, the RAFT core equilibrium was fullyestablished. Thirdly, long chain assumption was valid (including cross propagation andcross chain transfer equilibrium). Finally, a terminal model was used for addition andfragmentation rate coefficients (penultimate effect influences propagation but not RAFTprocess). The theoretical equations of apparent equilibrium coefficient and apparentchain transfer coefficient for the RAFT copolymerization were derived as follows:The model revealed that the rate retardation in copolymerization had the samepattern as in homopolymerization. The phenomenon could be described by replacingthe group parameter (k_c/k_t)K in homopolymerization with an apparent groupparameter (c>)/t>). Because the independence of c>/t> on monomercomposition, it was the value that determined the rate retardation effect in theRAFT copolymerization (for a fixed type of RAFT agent). The apparent equilibriumcoefficient was a function of monomer composition as predicted by equation (2). In addition, the apparent chain transfer coefficient tr> could be predicted by equation(3) provided C_tr,ii and C_tr,ij (i≠j = 1, 2) were given.Some peculiar kinetic behaviors in the MMA/BA/CPDB copolymerization wereobserved and explained by the apparent equilibrium constant model. In the conventionalfree radical polymerization, the BA homopolymerization gave the highest rate and thecopolymerization rate decreased monotonously with increasing f_mma. However, withCPDB as the RAFT agent, the homopolymerization of MMA gave the highest rate.Increasing the BA fraction f_ba dramatically decreased the copolymerization rate and therate reached the lowest point at f_MMA=0.2. This was attributed to the RAFT retardationeffect in the copolymerization. In addition, it was found that the dependence of chaintransfer constant on monomer composition was nonlinear in both copolymerizationsystems of MMA/BA/PEPDTA and St/BA/BDIB. However, two systems followeddifferent patterns. In the copolymerization of MMA and BA, C_tr decreased monotonouslywith the increase of f_mma. Polymers with narrower molecular weight distribution wereprepared with lower f_MMA. But in the copolymerization of BA and St, C_tr decreased firstand reached a minimum value at f_St≈0.25, which was smaller than both C_tr1.1and C_tr2.2.In the RAFT copolymerization, the molecular weight distribution could be broader than intheir homo-polymerization counterparts in some composition ranges.The selection of RAFT agent is important in copolymerization. Some criteria weredeveloped in this work. The tr> value should be adequately large to ensurecontrolled molecular weight and narrow PDI. The should be relatively small toavoid severe rate retardation during polymerization. A large value increasespolymerization time, but also increases the amount of dead polymer resulting in broadmolecular weight distribution.