Mechanistic Aspe~.ts of -Free-Radical Catalytic Chain Transfer Polymerization
Darren Forster
School of Chemical Engineering and Industrial Chemistry
University of New South Wales
Thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy
November 1999 ii
Summary
This thesis contains the results of an investigation into several aspects relating to the kinetics of free-radical catalytic chain transfer polymerization. This study is an endeavour to obtain a greater mechanistic understanding of the transfer process in bulk and solution polymerizations. The specific areas of investigation include the rate of transfer to methacrylates, the kinetics of catalytic chain transfer polymerization in a supercritical medium, the role of monomer on the transfer process, the effect of solvents on the kinetics of the transfer reaction and the kinetics of catalytic chain transfer polymerization in styrene.
The catalytic chain transfer polymerizations of methyl, ethyl, and butyl methacrylates are studied over the temperature range 40-70°C with cobaloxime boron fluoride (COBF) and its tetraphenyl derivative (COPhBF). It is found that for both catalytic chain transfer agents, the chain transfer constant decreases in going from methyl to butyl methacrylate, and that there is no significant temperature effect on the observed chain transfer constants. The re~ults are consistent with a diffusion-controlled rate coefficient for the chain transfer reaction.
The mechanism of catalytic chain transfer polymerization with methyl methacrylate is studied in a range of media. The chain transfer reaction in supercritical C02 is found to be significantly enhanced compared with similar experiments in toluene or in bulk methyl methacrylate. The results provide further evidence for a diffusion-controlled rate-determining step in the transfer process with cobaloxime catalysts. The gas like viscosities in the supercritical medium result in an approximate chain transfer coefficient (i The role of monomer in catalytic chain transfer polymerization is studied by determination of the chain transfer constant of COPhBF in methyl methacrylate at 60°C, varying the monomer concentration instead of the COPhBF concentration, as is common practice. Toluene and tert-butyl acetate were used as diluents in these studies iii and it was found that the chain transfer constants obtained in the present studies were not significantly different from those observed in conventional experiments. These results suggest the absence of a direct participation of monomer molecules in the hydrogen abstraction step in catalytic chain transfer. The catalytic chain transfer constants of COBF and COPhBF, of methyl methacrylate is studied in bulk and the presents of toluene and t-butyl acetate at 60°C, and are found to be similar in all cases. The results indicate that there is no solvent effect on catalytic chain transfer polymerization from weakly co-ordinating solvents. The effect of a strong base ligand, pyridine, was tested with COBF and COPhBF in MMA and styrene. Pyridine was found to a detrimental effect, with both catalysts, on the transfer reaction and rate of polymerization in MMA. In styrene, however, the transfer reaction was enhanced at low concentrations of pyridine, but on further additions of pyridine the transfer constant dropped again. This behaviour is thought to be caused by, at first, a single pyridine as the axial ligand catalyst, changing the catalytic process. Secondly it is thought that an inactive species is formed on further additions of pyridine, which maybe a hi-ligand species. The effect of pyridine was shown to have a significant effect on the kp of MMA, but had no effect on the kp of styrene. In the transfer polymerization of styrene, in the presence of COBF and COPhBF, the transfer constant is found to decrease rapidly in the initial stages of the polymerization, due to the formation of cobalt-carbon bonds between the styryl radicals and the catalyst. The transfer constant is found to increase dramatically after 80°C, due to the cobalt carbon bonds formed between the styryl radicals and the cobalt species becoming more labile changing the equilibrium between the active cobalt species and the inactive Co(III). There is little effect of the catalyst in p-methoxy styrene possible due to the catalyst binding irreversibly to the radical. Overall this thesis provides insight into many aspects of the mechanism of catalytic chain transfer polymerization. iv Declaration I hereby declare that this submission is my own work and that, to the best of my knowledge and belief, it contains no material previously published or written by another peFson nor material which to a substantial extent has been accepted for the award of any other degree or diploma of the university or other institute of higher learning, except where due acknowledgement is made in the text. I also declare that the intellectual content of this thesis 1s the product of my own work, even though I may have received assistance from others on the style, presentation and language expression. Darren Forster v Acknowledgments I would like to sincerely thank everybody who in some way is connected with the work in this thesis, especially: My supervisor Tom Davis, for giving me the opportunity to do a Ph.D., and being there with good advise, ideas, and inspiration, for the project, and creating a friendly, relaxed work environment. I would like give special mention of Dr. Johan P. A. Heuts who was an inspiration throughout the project, and who's insight led to some very fruitful discussions. I would like to acknowledge the financial support by ICI and ORICA. I would also like to thank all my colleagues, from the University of New South Wales for not only help when it was needed but also for all the good times that we shared, especially Lucy Baker, Michelle Coote, Heidi Kapfenstein, Dax Kukulj, David Morrison, Leesa Morris, Lisa Muratore, Evan Rodrigues, Lachlan Yee and Michal Zammit. Finally, I would like to thank my wife Chattragom and my parents for there encouragement, love and support. vi Contents 1 'INTRODUCTION ...... 1 1.1 GENERAL INTRODUCTION ...... 1 1.2 AIMS OF INVESTIGATION ...... 1 1.3 OUTLINE OF THE THESIS ...... 2 1.4 PUBLICATIONS ...... : ...... 4 2 BACKGROUND ...... 5 2.1 FREERADICALPOLYMERIZATION ...... 5 2. I. I Introduction ...... 5 2.I.2 Mechanism ...... 5 2. I .3 Diffusion Controlled Reactions ...... I I 2.I.4 Molecular Weight Distributions ...... I I 2. I. 5 Measuring Rate Coefficients ...... I 3 2.2 CATALYTIC CHAIN TRANSFER POLYMERIZATION ...... 19 2.2.I Introduction ...... I9 2.2.2 Mechanism ...... I9 2.2.3 Cobalt-Carbon Bond Formation ...... 22 2.2.4 Cobalt catalysts ...... 24 2.2.5 Temperature effects ...... 28 2.2.6 Solvent effects ...... 28 2.3 EXPERIMENTAL TECHNIQUES ...... 32 2.3. I Molecular weight analysis ...... 32 2.3.2 Rate Analysis...... 36 2.4 REFERENCES ...... 37 3 THE EFFECTS OF ESTER CHAIN LENGTH AND TEMPERATURE ON THE CATALYTIC CHAIN TRANSFER POLYMERIZATION OF METHACRYLATES ...... 41 3.1 INTRODUCTION ...... 41 3.2 CATALYTIC CHAIN TRANSFER POLYMERIZATION IN THE BULK POLYMERIZATION OF THE HOMOLOGOUS SERIES OF METHACRYLATES.- ...... 43 vii 3.2.1 Experimental ...... 43 3.2.2 Results and Discussion ...... 46 3.3 CONVENTIONAL AND CATALYTIC CHAIN TRANSFER IN THE FREE-RADICAL POLYMERIZATION OF 2- PHENOXYETHYL METHACRYLATE ...... 59 3.3.1 Experimental ...... 60 3.3.2 Results and Discussion ...... 62 3.4 CONCLUSIONS ...... 68 3.5 REFERENCES ...... 69 4 CATALYTIC CHAIN TRANSFER POLYMERIZATION OF METHYL METHACRYLATE IN SUPERCRITICAL CARBON DIOXIDE: FURTHER EVIDENCE FOR A DIFFUSION CONTROLLED TRANSFER PROCESS ...... 73 4.1 INTRODUCTION ...... 73 4.2 EXPERIMENTAL ...... 75 4.2.1 Materials ...... 75 4.2.2 General polymerization procedure ...... 75 4.2.3 Phase behaviour...... 76 4.2.4 Molecular weight analysis ...... 77 4.2.5 Matrix-assisted laser desorption ionisation (MALD1) mass spectrometry...... 78 4.3 RESULTS AND DISCUSSION ...... 78 4.3. 1 Catalytic chain transfer of COPhBF in super critical MMA/carbon dioxide ...... 78 4.3.2 Mechanistic 1nterpretation...... 82 4.3.3 The Effect of Non-Homogeneity ...... 86 4.4 CONCLUSION ...... 87 4.5 REFERENCES ...... 88 5 SOLVENT EFFECTS ON CATALYTIC CHAIN TRANSFER ...... 89 5.1 THE ROLE OF MONOMER IN THE CHAIN TRANSFER REACTION IN COBALOXIME-MEDIATED FREE- RADICAL POLYMERIZATION ...... 90 5.1. 1 Experimental procedures ...... 92 5.1.2 Results and Discussion ...... 94 5.1.3 Summary of the role ofmonomer on catalytic chain transfer ...... 98 5.2 THE EFFECT OF WEAK CO-ORDINATING SOLVENT ON THE CHAIN TRANSFER REACTION IN COBALOXIME-MEDIATED FREE-RADICAL POLYMERIZATION ...... 99 5.2.1 Experimental ...... 100 5.2.2 Results and Discussion ...... 100 5.2.3 Summary of the effect of weak co-ordinating solvent on the chain transfer reaction in cobaloxime-mediated free-radical polymerization ...... 104 viii 5.3 THE EFFECT OF PYRIDINE ON THE CHAIN TRANSFER REACTION IN COBALOXIME-MEDIATED FREE- RADICAL POLYMERIZATION ...... 104 5.3.1 Experimental ...... 105 5.3.2 Results and Discussion ...... 106 5.3.3 Summary of the Effect of Pyridine on the Kinetics of Catalytic Chain Transfer Polymerization ...... 119 '5 .4 THE EFFECT OF PYRIDINE ON THE PROPAGATION RATE COEFFICIENT ...... 119 5.4.1 Experimental ...... 120 5.4.2 Results and Discussion ...... 120 5.4.3 Summary of the Effect of Pyridine on the Propagation Rate coefficient of MMA and styrene ...... 124 5.6 REFERENCES ...... 124 6 REVERSIBLE COBALT-CARBON BOND FORMATION IN THE CATALYTIC CHAIN TRANSFER POLYMERIZATION OF STYRENE ...... 127 6.1 TRANSFER CONSTANT AS A FuNCTION OF CONVERSION ...... 129 6.1.1 Experimental ...... 129 6.1.2 Results and Discussion ...... 130 6.1.3 Summary of the transfer constant as a function of conversion ...... 142 6.2 TEMPERATURE DEPENDENCE OF THE TRANSFER CONSTANT IN STYRENE ...... 142 6.2.1 Experimental ...... 143 6.2.2 Results and Discussion ...... 144 6.2.3 Summary of the transfer constant as a function of temperature ...... 148 6.3 TRANSFER CONSTANT OF COPHBF IN P-METHOXYSTYRENE ...... 148 6.3.1 Experimental ...... 149 6.3.2 Results and Discussion ...... 150 6.3.3 Summary of catalytic chain transfer polymerization of4-metho:xystyrene ...... 152 6.4 REFERENCES ...... 152 7 CONCLUSIONS AND RECOMMENDATIONS ...... 155 APPENDIX 1 ...... 159 APPENDIX 2 ...... 167 APPENDIX 3 ...... 175 Introduction 1 Chapter 1 Introduction 1.1 General Introduction Free-radical polymerization has been widely used for the production of polymers for over 50 years. The reason for the wide use of free-radical polymerization is in the versatility and robustness of the process, its high tolerance to additives, functional groups, water and impurities. Free-radical polymerizations can occur over a wide range of operating conditions, such as, bulk, solution, emulsion and suspension processes. Though the technology has well been established, and the dominant reaction mechanisms understood, the free-radical process is highly non-selective thus controlling the molecular weight still poses great challenges to industry and academics. In recent years, however, there have been several novel techniques for the control of free-radical polymerizations that have emerged. These novel techniques, in particular, catalytic chain transfer, and "pseudo-living" radical polymerization, have allowed the control of polymer properties, molecular weight, polydispersity, end functionality, and also allows the formation of different architectures such as block copolymers, graft, and others. 1.2 Aims of Investigation The overall objective of this thesis is to investigate, and understand, the mechanisms of catalytic chain transfer polymerization. This process is of great importance both industrially and academically. Chapter 1 2 This thesis specifically investigates the use of catalytic chain transfer agents for control of molecular weight in free-radical polymerization. This relatively new technology is beginning to find applications within industry, though, the mechanisms behind the transfer process is still little understood. The aim of this thesis is to investigate the mechanism of catalytic transfer using a range of monomers and a range of polymerization conditions. 1.3 Outline of the Thesis Chapter 2: Background This chapter provides an introduction to free-radical polymerization mechanisms and kinetics, and the analytical techniques used in this thesis are given. The most important features pertaining to kinetics of free-radical catalytic chain transfer polymerization and catalytic chain transfer agents are also discussed. Chapter 3: The Effects of Ester Chain Length and Temperature on the Catalytic Chain Transfer Polymerization of Methacrylates The catalytic chain transfer polymerizations of methyl, ethyl and butyl methacrylates are studied over the temperature range 40°C to 70°C with cobaloxime boron fluoride (COBF) and its tetra phenyl derivative (COPhBF). The results are compared to the monomer viscosities at the different temperatures. The transfer constants of COPhBF and Dodecyl mercaptan are determined in Phenoxyethyl methacrylate at 60°C, and the results compared to the known transfer constants in MMA Chapter 4: Catalytic Chain Transfer Polymerization of Methyl Methacrylate in Supercritical Carbon Dioxide: Further evidence for a Diffusion Controlled Transfer Process This chapter describes initial experiments performed to test the potential of CCT in supercritical carbon dioxide, for the synthesis of acrylic oligomers. The effect of the low viscosity medium on CCTP will be evaluated by determining the transfer constant for COPhBF in supercritical MMA/C02. Introduction 3 Chapter 5: Solvent Effects In this chapter the role that monomer plays in the chain transfer reaction of cobalt mediated free-radical polymerization is investigated. The kinetics of catalytic chain transfer are measured in the presents of weak coordinating solvents and coordinating solvents. Chapter 6: Reversible Cobalt-Carbon Bond Formation in the Catalytic Chain Transfer polymerization of Styrene This chapter will further probe several aspects of the effect of cobalt-carbon bond formation on catalytic chain transfer polymerization and its possible role in the actual mechanism of the actual chain transfer reaction. The chain transfer constant, Cs, of both COBF and COPhBF was measured as a function of percent conversion during the initial stages of the free-radical polymerization of styrene. This study will also investigate the effect temperature on the chain transfer constant of the cobalt complexes in styrene. The chain transfer constant of COPhBF was measured in p-methoxystyrene to determine the effect of the radical strength on the transfer mechanism. Chapter 7: Conclusions and Recommendations The major findings and results of this thesis are summarised, and recommendations for further work in each of the areas investigated are given. Chapter 1 4 1.4 Publications The following publications, arising from the work in this thesis, have appeared to date. Further publications from this thesis, will be submitted to international journals. ' "The role of monomer in the chain transfer reaction in cobaloxime-mediated free-radical polymerization." Heuts J.P.A. Forster D.J. Davis T.P. Macromolecular Rapid Communications. 20(6):.299-302, (1999) "The effects of ester chain length and temperature on the catalytic chain transfer polymerization of methacrylates." Heuts J.P.A. Forster D.J. Davis T.P. Macrof!Zolecules. 32(12):3907-3912, (1999) "Catalytic chain transfer polymerization of methyl methacrylate in supercritical carbon dioxide: Evidence for a diffusion-controlled transfer process." Forster D.J. Heuts J.P.A. Lucien F.P. Davis T.P. Macromolecules. 32(17):5514-5518, (1999) "Conventional and catalytic chain transfer in the free-radical polymerization of 2- phenoxyethyl methacrylate." Forster D.J. Heuts J.P.A. Davis T.P. Polymer. 41:1385, (2000) Background 5 Chapter 2 Background 2.1Free Radical Polymerization 2.1.1 Introduction Free-radical polymerization has been widely adopted in the plastics and coatings industry for over 50 years. The reason for this includes the versatility of the reaction, owing to its high tolerance to impurities, water, functional groups, and additives. Free radicals, however, tend to be highly reactive and non-selective intermediates and it is difficult to isolate a single product from synthesis involving radicals, but a range of molecular weights and compositions, in the case of copolymers. 2.1.2 Mechanism Free radical polymerization is characterised by a chain reaction mechanism consisting of the following steps: (i) initiation, where new radicals are being formed, (ii) propagation, where monomeric units are added to the radical giving chain growth, (iii) termination, which involves the reaction of two radicals to give dead polymer chains, (iv) chain transfer, which is a process by which a radical passes from the growing polymer chain to a smaller molecule, thus terminating the polymer chain but not the radical. These processes will be described in more detail below. Chapter 2 6 2.1.2.1 Initiation The initiation process involves two steps, (1) decomposition of an initiator molecule, usually thermally or by electromagnetic radiation, to give free radicals, t, with dissociation rate coefficient kd. (2) The addition of the formed primary radicals to the monomer, with the rate coefficient kpi· Initiator (2.1) (2.2) Equation (2.1) represents of the decomposition of initiator molecules to give two identical radicals. The radicals may then add to monomer, in which case initiation of the polymer chain has taken place as in equation (2.2). The rate-determining step for initiation is usually decomposition of the initiator and so the rate of initiation can be expressed as (in equation (2.3)): (2.3) where f, is the efficiency of the initiator, which is the ratio of the number of initiator molecules that add to a monomer unit compared to the number of radicals formed during decomposition of the initiator. Typically the value off is in the range of 0.3-0.8 due to recombination of the radicals after initiator decomposition. 2.1.2.2 Propagation Propagation is the process by which the polymer chains are formed; it provides the mechanism for polymer growth and it involves the subsequent addition of monomer units to the active radicals: Background 7 kn R"n + M -----!p-~) R"n+l (2.4) where R •n is a growing polymer chain containing n monomer units, and kp n is the propagation rate coefficient for the n-th monomer addition. The rate of propagation can be expressed as: (2.5) Recent studies show that the propagation rate coefficient is dependant on the chain length, and that the value of kp at very short chains can be 4 - 10 times higher than for the long chain limit.l-5 2.1.2.3 Termination Termination is a chain stopping process that involves the destruction of active radicals in the system. This involves a chemical reaction between two propagating radical chains resulting in two dead polymer chains. This process may occur in two different ways, combination where the radicals simply couple with each other to produce a single polymer chain (as in equation (2.6)), or disproportionation, where an atom (usually a hydrogen) is transferred from one propagating radical chain to another to terminate both growing chains: R"n + R" m (2.6) km,n td ) --»:---4- Pn + Pm R"n + R" m (2.7) Chapter 2 8 Both of these termination modes are possible and can occur simultaneously during a single polymerization process. The ratio of the termination modes depends on the structure of the terminal monomer unit in the growing radical chain. Termination of polymer chains from vinyl monomers, eg styrene, terminates predominantly via combination. Polymer chains from a-methyl vinyl monomers, eg methyl methacrylate, involves a significant amount of disproportionation.6 Unless there is a need to distinguish between the modes of termination it is usual to use a single termination rate coefficient, kt: (2.8) Termination rate coefficients are diffusion controlled and so the actual values of the termination rate coefficient are system dependent 7. Diffusion is a function of molecular weight and the media viscosity, which in turn is a function of conversion and molecular weight, hence it is common to see a decrease in the termination rate coefficient of several orders of magnitude. To avoid these problems an average termination rate ). coefficient is taken, (k1 The termination rate coefficient can then be related to the number average degree of polymerization by the equation: (2.9) the rate of termination can then be expressed as: (2.10) Background 9 2.1.1.4 Chain Transfer In addition to termination, the propagating polymer chains can also be stopped by chain transfer. The transfer reaction causes the death of one chain, and the reinitiation of a new polymer chain. The chain transfer reaction can be expressed as: (2.11) (2.12) The transfer reaction can occur to any species, S, that may be present in the reacting system ie monomer, solvent, initiator, polymer or a chain transfer agent added deliberately, such as a mercaptan. The transfer process, equation (2.11), usually proceeds via a transfer of a hydrogen or a halogen atom from the transfer agent to the n meric radical producing a dead polymer of chain length n, Pn, and a small radical s• which, usually, can reinitiate propagation, equation (2.12). The net effect of chain transfer is a reduction in molecular weight, but no loss in radical activity. In classical kinetics (no chain length dependence on the termination rate coefficient), there is no effect of chain transfer on the rate of the reaction. However chain transfer will usually reduce the overall rate of polymerization due to there being more short radicals which have a higher termination rate coefficient, ie non-classical kinetics. 2.1.1.5 Kinetic expressions In deriving simple kinetic equations for the preceding reactions, it is necessary to make the following assumptions: (a) The long chain assumption, where the growing polymer chains are assumed to be sufficiently large so that the total rate of monomer consumption is equal to the monomer Chapter 2 10 consumption in the propagating reaction alone. Thus the polymerization rate is simply equal to the rate of propagation: _ d[M] =R (2.13) dt p b) Kinetic parameter chain length independence, where the rate coefficients for the propagation reaction are independent of chain length and conversion, thus the rate coefficient for the first propagating steps are assumed to be equal to the long chain propagating rate coefficient. (c) The steady state assumption, where the steady state concentration of the radicals is thought to be constant (d[R"]/dt = 0), thus the rate of formation of radicals equals the rate of consumption of radicals. (2.14) By combining equations (2.3), (2.10) and (2.14) the following expression for the steady state free radical concentration: Yz [M.] =( jk;,[l]) ' (2.15) Substitution into equation (2.5) yields for the rate of polymerization: (2.16) The kinetic chain length, which is the ratio of rate of propagation to the rate of initiation, by combining equations (2.3), (2.5) and (2.15), becomes: Background 11 (2.17) 2.1.3 Diffusion Controlled Reactions During the polymerization reaction the reaction mixture becomes increasingly viscous, and the diffusion coefficients of all species present decrease. If the reaction mixture became glassy at high conversions the diffusion coefficients can drop by many orders of magnitude. From the encounter pair model, 8,9 the overall rate coefficient, k, of a bimolecular process with contributions from both chemical, kchern and diffusion, kdiff' components can be written as: 1 1 1 -=--+- k k chem k diff (2.18) when a reaction is chemically controlled, kdiff >> kchern and so k = kchern. Though when the mobility of the reactants are slowed sufficiently so that kchern >> kdiff, then the reaction is said to be diffusion controlled. 2.1.4 Molecular Weight Distributions The molecular weight of a polymer plays a very important role in its application. Generally when a polymer is made a range of molecular weights are synthesized, so it is usual to describe the molecular weight distribution of a polymer sample. There are Chapter 2 12 various methods that have been developed for calculating the molecular weight of a polymer, based on endgroup analysis, colligative properties, dilute viscosity, light scattering and viscosity.lO These techniques, however, only give the molecular weight averages and do not show the overall molecular weight distribution. In a kinetic analysis of the processes of polymer synthesis, it is often useful to look at the entire molecular weight distribution. The use of gel permeation chromatography, or size exclusion chromatography, enables the entire molecular weight distribution to be analysed. Often it is convenient to describe the molecular weight distribution of a polymer in terms of its averages. Thus polymers can be easily compared. The common averages used are the number average molecular weight, M0 , weight average molecular weight, Mw, and polydispersity index, PDI, and these are defined as follows: fMP(M)dM M =~---- (2.19) n JP(M)dm JM2P(M)dM M = -=-=----- (2.20) w fMP(M)dM (2.21) Where P(M) is the number of chains of molecular weight M. The weight of chains W(M) is defined as W(M) =M P(M). The two forms of presenting the molecular weight distribution, that are used throughout this work, are: (a) the log weight distrybution, w(logM) vs logM, and (b) the natural log Background 13 of the number distribution ln P(M) vs M. The w(logM) vs logM plot is useful for giving a pictorial representation of the molecular weight distribution, and the ln P(M) distribution is useful for determining the chain stopping mechanisms, and can be useful in obtaining kinetic parameters. The relationship between these distributions is presented in 2.3.1.1. 03 -10-.------, (a) (b) Mw -12 -14 (12 Mn -16 i' g ~ -18 .9 "' Ul -3l -22 lf1 QO 1.0<1<1' 1.5xld' 20dd' logM M Figure 2.1 Two forms of presenting molecular weight distributions of a typical polymer sample. (a) w(logM) vs log M, and In P(M) vs M. 2.1.5 Measuring Rate Coefficients 2.1.5.1 Propagation Rate Coefficient Several methods for determining the propagation rate coefficient, kp, have been used in literature, however many of these are model dependant and rely on assumptions about the termination rate coefficient, distribution of polymers produced by pulsed-laser polymerization (PLP). A unique feature of PLP, schematically shown in Figure 2.2, is that PLP, which relies of non steady state kinetics, allows kp to be determined independently of any of any other kinetic parameters. Chapter 2 14 In PLP the monomer and photo initiator are exposed to short pulses of laser light, in the order of 10-7 s, delivered at a constant repetition rate, with dark times, tc, greater than 10-3 s. Since the flash time is instantaneous compared to the dark time a population of radicals are produced together at regular intervals (Figure 2.2). In the dark time between pulses the radicals propagate with propagation rate coefficient, kp. Some bimolecular tetmination will occur during this time, but if the appropriate conditions are selected, then most of the polymer chains will still be growing at the next pulse. At the subsequent laser pulse, a large population of radicals is produced. These react with the growing polymer causing premature and instantaneous termination. The surviving radicals propagate until the next laser pulse and the process is repeated. As a result of this process a significant number of polymer chains are initiated and terminated by consecutive pulses, resulting in a characteristic peak in the molecular weight distribution corresponding to the polymer that grew between two consecutive flashes. In addition a secondary peak and even tertiary peaks sometimes appear, as a result of the polymer chains that survive two or three pulses respectively. flash x flash x+-1 flash x+-2 flashx+-3 Lo Figure 2.2: Schematic of a pulsed-laser polymerization experiment. "Instantaneous" initiation occurs at 'flash x', which starts polymer chains propagating during dark time LO. At the next flash (flash x+ 1), a second genaration of radicals is produced increasing the probability of bimolecular termination. Some macroradicals survive the subsequent flash and are terminated after lor 2 more time periods. Background 15 By measuring the degree of polymerization that develops during dark period, L0, kp can be determined from the following equation: k=~ (2.22) P [M]tr Determination of Lo has been the topic of much discussion in recent years. In general the low molecular weight inflection point gives the best estimate of L0, see Figure 2.3. 1.5"T"""------, inflection 1.0 point GPCtrace QO 1st Deriva.ti\te -Q5 1cP logM Figure 2.3 Molecular weight distribution, w log(M) vs log M, and its first derivative, obtained by a pulsed laser polymerization methyl methacrylate at 60°C. Note the large primary peak, and inflection point, due to radicals being terminated at the next pulse, and the overtone, due to radicals surviving for two pulses. Chapter 2 16 2.1.5.2 Chain Transfer Rate Coefficient A measure of the reactivity of a chain transfer agent is given by the chain transfer constant, Cs, which is defined as the ratio of chain transfer and the propagation rate coefficients, ku./kp. Mayo Method The most widely used method for determining the chain transfer constant of any transfer agent, solvent, monomer, is the use of the Mayo methodl0,12. This method has seen widespread use, and an extensive collection of chain transfer constants is presented in the Polymer Handbook.l3 The Mayo equation defines the number average degree of polymerization, DPn, as a function of the rate of chain growth, divided by the sum of the rate of all chain stopping events, and is expressed as: (2.23) Inverting both sides of equation (2.22), we obtain: (2.24) In this expression, A is the fraction of termination by disproportionation, and eM (=ktr,Mikp) and Cs (=ktr,slkp), are the chain transfer constants to monomer and chain transfer agent, S, respectively. Since the expression is based on overall chain growth and chain stopping rates, no assumptions regarding any mechanisms other than the overall kinetic scheme and the long chain assumption are introduced, and hence the expression is exact. Background 17 The usual procedure for measuring the chain transfer constant, involves determining the average degree of polymerization for a series of polymerizations with varying [S]/[M] values, and plotting the data as DPn-1 vs [S]/[M] (ie a Mayo plot, Figure 2.4(b)). The chain transfer constant, Cs, is obtained by taking the slope of the Mayo plot. (16 (a) Q014 (b) Increasing [S]/[M] (1012 Q4 QOtO o.tXJI i ll-." ,g §1 ;1: nro; Q2 Q rum o.tlll to' to' to' t Figure 2.4 (a) A series of molecular weight distributions of Phenoxyethyl methacrylate showing a decrease in molecular weight with increasing [S]/[M], The transfer agent, S, is dodecyl mercaptan (DDM), (b) shows a Mayo plot with the slope of the line being Cs. The number average degree of polymerization is calculated from the number average molecular weight, Mn, divided by the monomer mass. This method suffers difficulties in assigning values for the number average molecular weight, due to sensitivity in the base line selection. An alternative method of determining the degree of polymerization is by dividing the weight average molecular weight, Mw, which is less susceptible to baseline errors, by monomer mass multiplied by the polydispersity index. Using Mw/(2·monomer mass) as an estimate for DPn, is justified in systems which are fully controlled by chain transfer (i.e., the polydispersity index, Mw!Mn, is equal to 2), but it has been found that this estimate gives more reliable results than the estimate based on Mn.l4-19 Chapter 2 18 Chain Length Distribution Method A more recent procedure, for determining the chain transfer constant, developed by Gilbert and co-workers20-23 is based upon taking the high molecular weight slope of the number distribution, P(M), plotted as ln(P(M)) vs M. The high molecular weight slope, denoted as A, is defined as follows: ll·m ktr,s[S]+ktr,M[M]+(k 1 )[R•] ,_M A =M~~ In P(M) = (2.25) where m0 is the mass of the monomer. The similarity of equation (2.25) and the Mayo equation (equation (2.24)) is immediately obvious, the only difference between equations (2.25) and (2.24) is the fact that the termination term in the latter does not depend on the fraction of termination by disproportionation or combination. For a polymerization which is dominated by chain transfer, equation (2.25) Can be reduced to: A =lim In P(M) oc -C [S] . M (2.26) M~~ s [M] mo Measurement of Cs in this case involves the measurement of A as a function of [S]/[M] and a plot of -Amo vs [S]/[M] then yields a straight line with slope Cs. Recent work by several groups14,15,17,24,25 showed that in systems that are dominated by chain transfer, more reliable results can be obtained if A is taken from the peak molecular weight region of the distribution. Background 19 2.2 Catalytic Chain Transfer Polymerization 2.2.1 Introduction Catalytic chain transfer (CCT) has emerged in recent years as a highly effective method of controlling molecular weight in free radical polymerization. The first studies of the use of low spin cobalt (II) complexes, in free radical polymerization, were carried out by Enikolopyan and coworkers26-28, where porphyrin complexes were found to have chain transfer constants 3-4 orders of magnitude higher than the mercaptan based transfer agents, in the free radical polymerization of methyl methacrylate. But the cobalt porphyrins are highly coloured, expensive and sparingly soluble in polar media. The concentration of the cobalt-porphyrin was measured spectroscopicaUy26, and was shown not to change throughout the polymerization process, and all the catalyst could be recovered, thus the transfer process was shown to be genuinely catalytic. Since then many other cobalt species have been identified as catalytic chain transfer agents. 2.2.2 Mechanism In catalytic chain transfer polymerization, initiation, propagation and termination steps are thought to proceed via a free radical mechanism. NMR studies of the polymer reveal that the products of the process are reflective of a system controlled by transfer to monomer, ie a hydrogen atom is transferred from the propagating radical to a monomer unit. Shown in scheme 2.1 Chapter 2 20 H CH3 H CH3 H CH H CH3 I I \ I Co( II) I II 2 I I R-C-C• + c=c R-C-C + H-C-C• I \ I \ • I \ I \ H C=O H C=O H C=O H C=O I I I I 0 0 0 0 \ \ \ \ CH 3 CH3 CH3 CH3 Scheme 2.1: Mechanism for the catalytic chain transfer polymerization of methyl methacrylate via a Co (III)-H intermediate. Although it has since long been established that a hydrogen atom is abstracted from a ~ position (preferably from an a-methyl group) in the growing radical and transferred to a monomer molecule to give a monomeric free radical, no conclusive evidence regarding the actual chain transfer reaction exists. The most widely accepted view seems to be that the mechanism is a two-step process26,27,29-33. R: +Co(II) Pn +Co(ID)-H (2.27) Co(ID)- H + M ~ Co(II) + R: (2.28) Scheme 2.2: Mechanism for the catalytic chain transfer reaction via a Co(III)-H intermediate. Firstly, the Co(II) complex abstracts a hydrogen atom from a growing radical chain, R0 •, yielding a dead polymer chain, P0 , and a Co(ID)-H intermediate. The Co(ID)-H intermediate subsequently reacts with a monomer molecule, M, to yield a monomeric free radical and the original Co(ll) complex: The rate of transfer for the mechanism shown in scheme 2.1, Rtr, is determined by equation (2.27) of scheme (2.2), and ~an be written as: Background 21 (2.29) In addition to scheme 2.2, two other possible mechanisms have been mentioned26,34: (i) a~ reaction of a coordinated radical with a monomer molecule (equations (2.30) and (2.31)), or (ii) a reaction of a coordinated monomer molecule with a radical (equations (2.32) and (2.33)). It is not important for the kinetic scheme whether the second substrate molecule also needs to be coordinated or reacts directly with the coordinated substrate molecule. Kt Co(ll) + R~ f-- [Co- -Rnl (2.30) ku-,1 [Co--Rn] + M Co(II) + Pn + R~ (2.31) K2 ) Co(II) + M f-- [Co--M] (2.32) (2.33) Scheme 2.3: Mechanism of catalytic chain transfer via the cobalt catalyst coordinating with the radical (equation 2.30)or monomer (equation 2.32) It is then easily seen that for both mechanisms the following expression is derived for the rate of chain transfer (Rtr,i): (2.34) Chapter 2 22 where i = 1 or 2 and Ki is the equilibrium constant of the association equilibria of equations (2.30) and (2.32). 2.2.3 Cobalt-Carbon Bond Formation It is well known that organo-cobalt complexes can be formed by the radical-radical recombination of a Co(II) complex and an organic radical, and many such organo-cobalt compounds have been observed.35,36 The formation of the Co-C bonds in these compounds is reversible and in· some cases, such as in the case of the free-radical polymerization of acrylates in the presence of Co(II) complexes, this process leads to "living" polymerization characteristics (see Scheme 2.4).37-40 • Rn + Co(II) Co(III)-Rn u Monomer addition Scheme 2.4: Living free-radical polymerization involving the formation of cobalt carbon bonds Electron paramagnetic resonance (EPR) studies reported by Gridnev et al41, and later reproduced by Heuts et al19(See Figure 2.5), showed that there is significant Co-C bond formation in the cobaloxime mediated catalytic chain transfer polymerization of styrene, whereas this seems to be absent in the case of methyl methacrylate. Considering the fact that the observed chain transfer constant is a function of active transfer agent present (equation (2.24)), the Co-C bond formation leads to a decrease in Background 23 the concentration of free catalyst. This implies that the chain transfer constants for styrene determined so far may be only apparent values. It was also found by Heuts19 that the formation of R-Co(III) in the early stages of the polymerization causes an induction period, which was ascribed to a reaction of the Co(II) complex with the initiator-derived primary radicals in both methyl methacrylate and styrene polymerizations and a further reaction with propagating styryl radicals in styrene polymerization. For the case of styrene polymerization it was found that for the first few percent conversion there was an increase in the number average molecular weight. This suggests that the observed chain transfer constant for styrene is conversion dependent. b100 '§:100 :;::. (a) (c) so I"· ~t~ so 0 v~~ -...... - 0 ·SO LJ ·SO ~ -100 -100 2800 3000 3200 3400 3600 3800 2000 3000 3200 3400 3600 3600 IGl IGJ -250 'b eoo r ::::.200 (b) -g: (d) lSO 400 100 (\ ~if l 200 50 0 0 1-'------v-..-.--...~ -SO v~ n -200 -tOO If\ v -150 -400 ·200 ·250 -600 I I I 3200 3400 3600 3800 2800 3000 3200 3400 3600 38CO IGl IGI Figure 2.5 Electron paramagnetic resonance spectrum of COPhBF showing no loss in intensity of spectrum after the addition of radicals for MMA (a) in bulk MMA, (b) in bulk MMA after heating at 60°C in the presents of an initiator, but spectrum lost after the addition of radicals to styrene (c) in bulk styrene, (d) in bulk styrene after heating at 60°C in the presents of an initiator. (Figures taken from Heuts et al19) Chapter 2 24 2.2.4 Cobalt catalysts The cobalt(II) catalyst has its origins in biochemistry, where vitamin Biz is used to conduct many free radical reactions. Vitamin Biz consists of a corrin ring, and is the only metal-carbon bonds biochemistry. The use of vitamin Biz and related Co ~ complexes as catalysts in organic synthesis was reviewed by Scheffold.42 The most effective Co(II) catalyst are invariable low spin Co(II) coordinated compounds.35,43 Co(II) has a d7 electron configuration and can exist in either the high spin state (with three unpaired electrons) or low spin state (with only one unpaired electron). The energy levels of these states are close together, and it is the nature of the ligand which governs whether the Co(II) will exits in either a high or a low spin state. For example, if the macrocycle consists of 2 oxygen atoms and 2 nitrogen atoms or 4 oxygen atoms coordinated with the cobalt atom (Figure 2.6), then the cobalt (II) species will exist in the high spin state, causing the catalyst to be ineffective. Structure 1 Structure 2 Figure 2.6 Ineffective catalytic chain transfer agents. Most cobalt(II) complexes that contain non macrocyclic tetradentate ligands with either oxygen or nitrogen coordinated to the cobalt atom are in a high spin state. However there are a number of tetradentate macrocycles with sufficient high field ligand to allow the Co(II) species to exist in a low spin state, such as structures 3,4,5 and 6. Background 25 Structure 3 Structure 4 Figure 2. 7 Effective catalytic chain transfer agents The most effective method of determining if a catalyst is in the high or low spin state, is by determining the magnetic moments of the catalyst. Low spin Co(II) has a magnetic moment of about 1.73 J.ls and the high spin value is 3.46 J.18 . This is illustrated in Table 2.1, which lists the magnetic moments of four different cobalt catalysts, two of, which are effective and two of, which are ineffective. Table 2.1. Magnetic moments for selected Co(II) macrocycles Catalyst Magnetic Spin state Catalytic chain moment(BM) transfer agent Structure 5 1.86 Low Yes Structure 6 1.58 Low Yes Structure 1 4.48 High No Structure 2 4.28 High No data taken from Davis et ai44 An alternative method to screen potential catalysts is by means of NMR and ESR. It is difficult to obtain good NMR spectra of an efficient catalyst due to the paramagnetic effects of the low spin Co(II) species, but high spin Co(II) complexes give acceptable Chapter 2 26 spectra44. Conversely low spin Co(ll) complexes give excellent ESR spectra in contrast to high spin catalysts. Gridnev45 pointed out the use of peroxide initiators whilst using cobalt complexes as catalytic chain transfer agents should be avoided. The peroxide initiators were found to r poison the catalyst irreversibly, due to the formation of oxygen as the peroxide decomposes. Acidic solvents and monomers were also found to degrade the cobalt catalyst. A range of alternative macrocycles was then explored.46,47 O'Driscoll et ai.30 successfully replaced the porphyrin ring with dimethyl glyoxime (Co-DMG) (structure 4). These structures were coordinated with base ligands such as pyridine or methanol. The cobaloxime was found to have chain transfer constants an order of magnitude higher than the porphyrins (i.e. five orders of magnitude higher than the mercaptan based transfer agents). They were also less coloured, less expensive, and more soluble in organic media. The cobaloxime is however, susceptible to hydrolysis and oxidation, although less than the porphyrin catalyst. Gridnev31 reported the use of cobaloxime, and reported the effectiveness of the catalyst as a chain transfer agent in the presence of methacrylates and for styrene. Catalysts similar to this (stabilized with pyridine and phosphines as ligands), which are easily made (even in situ), were patented by Glidden46,48. The problem of oxidation and hydrolysis were overcome by the addition of BF2 to convert cobaloxime to cobaloxime boron fluoride (COBF) (structure 5). This particular catalyst has received the most attention in recent literature. The DuPont series of patents feature examples of COBF in use, covering terpolymerization49, suspension polymerization 50, and emulsion polymerizationS I. Background 27 COBF COPhBF Structure 5 Structure 6 Figure 2.8 Structure of catalytic chain transfer agents cobaloxime boron fluoride (COBF) and tetra phenyl cobaloxime boron fluoride (COPhBF) The partitioning of the catalyst between the aqueous phase, and the organic phase is of 66 great importance, when designing a catalyst for emulsion polymerization. Kukulj Showed that the partitioning of the cobaloxime boron fluoride catalyst, could be controlled by the selection of glyoxime used in the preparation of the catalyst. COBF was found to partitioned reasonably well between the aqueous phase and organic phase for methyl methacrylate, ie. substantial quantities of the catalyst are present in both phases. It was found that a reasonable large amount of chain transfer occurred using this catalyst. Tetra-phenyl cobaloxime boron fluoride (COPhBF) (structure 6) was found to partition solely in the organic phase. Thus the mass transfer of the catalyst from the monomer phase to the growing latex particle is limited, and the chain transfer was found to be reduced by a large amount compared to COBF. The two different cobaloxime boron fluoride catalysts were also found to have different chain transfer constants52. The value for COBF was found to be approximately 40000 in MMA at 60°C. And for COPhBF the value was reported to be approximately 20000 for the same conditions. A possible reason for the difference in catalytic activity, proposed by Haddleton et al53, is the reduced the mobility of COPhBF due to an increase in cross sectional area. Chapter 2 28 2.2.5 Temperature effects The effect of temperature on the catalytic chain transfer process was reported by both Sanayei54 and Kukulj34. In the first study54 a very strong temperature dependence of the catalytic chain transfer reaction of COBF in methyl methacrylate was reported. The - 1 activation energy for the chain transfer constant was reported to be -10.1 kJ·mor with a 3 pre-exponential factor in the order of 10 , which corresponds to a pre-exponential factor 10 for ktr in the order of 10 . This result was contradicted in a subsequent study34, based on a limited set of data, which found only a weak temperature dependence. The activation energy for the chain transfer constant, in this study was reported to be -4.63 1 3 kJ·mor with a pre-exponential factor in the order of 10 • Which also corresponds to 10 pre-exponential factor in the order of 10 . Kukulj34 also reported a strong temperature dependence of the catalytic chain transfer reaction of COBF in styrene. The activation energy reported for the chain transfer constant of COBF in styrene was determined to be -41.8 kJ·mor1 with a pre-exponential 4 factor in the order of 10- , which corresponds to a pre-exponential factor ktr in the order 3 of 10 . Though the final conversions of these experiments varied, thus if there is a strong dependence of conversion on the chain transfer constant in the catalytic chain transfer polymerization of styrene, then this result may not be representative of a temperature effect, but a conversion effect. 2.2. 6 Solvent effects 2.2.6.1 Weakly-coordinating solvents There seems to be some contention as to the actual effect of non-coordinating solvents on CCTP. In one study by Gridnev45, it was stated that additives such as acetone, ethyl acetate, benzene and toluene (up to 40% v/v) did not cause changes in CCTP rate and kinetics of polymerization, no data on this was shown in the paper. In another study by Suddaby et al15, based on a limited set of data, he showed there was a decrease in the Background 29 chain transfer constant, of COBF from 36 x 103 to 25 x 103 on addition of toluene to the catalytic chain transfer polymerization of methyl methacrylate at a ratio of 2:1 respectively. 2.2.6.2 Coordinating solvents The square-planar cobalt complexes feature a base ligand that is situated in the axial, position coordinated to the central metal atom. (Figure 2.9) The nature of this ligand depends on the synthetic route taken in synthesising the catalyst, for example if the catalyst were synthesised in methanol, then the base catalyst would be methanol. Alternatively the catalyst could be synthesised in pyridine making pyridine the base ligand. Figure 2.9: Positioning of the base ligand, situated in the axial position coordinated to the central metal atom. The base ligand has been shown to have an effect on the activity of the catalyst47. Janowicz47 showed the influence of a number of base ligands, at a range of base ligand:catalyst ratios, on the molecular weight of the polymer formed. Figure 2.10 shows the effect of pyridine on the activity of the CCT agent at a fixed concentration of the cobaloxime complex (6.0 x 10-6 M). Janowicz also tested the effect of a number of other Lewis bases to try and enhance the transfer process, ranging from triethylamine (Et3N) to trimethoxyphosphine (Me0)3P, which appeared to have the strongest effect. Chapter 2 30 'I CslfsN 40000 * - * • (C6Ifs)3P 4 30000 r- • - s= • ::;E 20000 1- * * 10000 1- • 0 • .. * "" 10 100 1000 10000 [Base]/[Catalyst] Figure 2.10 Effect of pyridine and triphenylphosphine on the efficiency of CCT at a fixed concentration of CCT agent ( data from Ref. 47) Janowicz47 also tested the commercially available octaethylporphyrin cobalt complex in admixture with triphenyl phosphine, and noted a decrease in the molecular weight, with an increase in base concentration, as with the cobaloxime. When the same experiment was repeated with pyridine instead of triphenylphosphine, it failed to demonstrate the anticipated improved molecular weight control. Thus different Lewis bases can have different effects on the activity of the catalyst depending on the axial ligands. A study of more strongly co-ordinating solvents, by Haddleton et al53, showed a marked decrease in the Cs of COBF in methanol and butanone. The interpretation of these results, was a competition between the monomer, solvent and the polymer for reaction at the cobalt centre (equation (2.35)). [Co]-solvent H [Co]+ solvent (2.35) Background 31 As was discussed earlier it is thought, in the mechanism of cobalt mediated catalytic chain transfer, that the cobalt(II) species binds with the radical in some way to form a [ Co(ill)-C ] as an intermediate or transition state. A number of authors31,55,56 have stated that the nature of the axial ligand affects the strength of this bond. For the series of Lewis base ligands examined the same conclusion was formed by all that is, the Co-C bond dissociation energy increases systematically with basicity of the ligand. One of the main points that can be discerned from Halpems work is that an increase in steric crowding from the axial ligand decreases the Co(ill)-C bond strength. The increase in basicity of the ligand will put more non-binding electron density on the cobalt centre, this should stabilise the Co(ill)-C bond formed. However it is very difficult to draw any firm conclusions of the real effect of the base ligand on the CCT process since there is very little data of this effect in literature. It should also be noticed that any change in the bond strength should result in a change in rate of polymerization, but there is no data supporting this in the literature. Gridnev et al35,41 pointed out that the base ligand could also take part in side reactions in CCTP. For example [Pyridine-Co(ill)-H] could under go deprotonation to give [Co(I)r and Pyridine-W. Active Co(II) can be regenerated by the reaction of [Co(l)r and [Co(ill)-H] to give 2x[Co(II)], and H2 being evolved. It was noted by Davis44 that aprotic solvents may also promote this reaction. These reactions are thought to be fast, compared to the competing radical reaction. Thus aprotic solvents and bases should be avoided in CCTP as they can act as a radical sink, although some bases can enhance the activity of the catalyst as shown above. 2.2.6.3 Catalytic inhibition The effect of some solvents can even have the effect of catalytically inhibiting the polymerization process. O'Driscoll and co-workers57 found that dimethyl formamide (DMF), which is known as a strong coordinating solvent, catalytically inhibited the Chapter 2 32 polymerization of methyl methacrylate. The catalytic inhibition was thought to follow the following mechanism: R • + Co(ll)-DMF ~ P + H-Co(III)-DMF (2.36) R• + H-Co(III)-DMF ~ P + Co(ll)-DMF (2.37) The main point of catalytic inhibition is that the cobalt hidride formed is stable towards monomer. Thus reinitiation does not occur, but the Co(III)-H can react with a second polymeric radical terminating the polymer. The net effect is two chains are terminate by disproportionation. 2.3 Experimental Techniques 2.3.1 Molecular weight analysis 2.3.1.1 Gel Permeation Chromatography Gel Permeation Chromatography (GPC), or Size Exclusion Chromatography (SEC) as it is sometimes referred is the most common method for determining the molecular weight distribution of a polymer sample. GPC is a column fractionation method in which molecules are separated according to their hydrodynamic volume in solution. Separation of the polymer molecules is accomplished when the polymer solute is passed through a column with a porous polymer packing, typically made with cross-linked poly(styrene/divinyl benzene). As the polymer passes through the column the molecules will venture into the pores of the packing increasing its retention time. As the Background 33 hydrodynamic volume of the molecule increases, only the larger pores will trap the molecules, thus the retention time shortens. This results in the separation of the polymer molecules as the smaller molecules are retained longer and are eluted at later times. To obtain a molecular weight distribution it is first necessary to transform the raw signal, which relates detector response as a function of elution volume. This is achieved by the use of a calibration curve, which relates the elution volume, Vel. to molecular weight, M. This is usually generated using a series of low polydispersity polymer standards. A third order polynomial is fitted to the calibration curve which is plotted as log M vs Vel· If the standards are the same polymer as the polymer of interest then an absolute molecular weight distribution can be obtained. If, however, the standards are not available then a universal calibration can be used. This relates the hydrodynamic volume, HDV, to molecular weight and intrinsic viscosity, ['fl], by the expression:58 HDV=M[rJ] (2.38) The intrinsic viscosity can be related to the molecular weight by the use of the Mark Houwink-Sakurada equation:9,59 [17] = KMa (2.39) Where Kanda are parameters which describe a specific polymer/solvent/temperature combination. Combining these two equations gives the following relationship for universal calibration: log(HDV) =logK +(a+ l)logM (2.40) Chapter 2 34 For a quantitative analysis of the GPC chromatogram, it is necessary to transform the raw-data raw-signal (detector response vs elution volume) into a molecular weight distribution.60 This is performed using the calibration curve, and the signal response from the polymer. The chromatogram is first normalized to give the weight of polymer per unit volume increment: (2.41) By dividing the normalized chromatogram by the derivative of the calibration curve, d(log M)/dVeh the distribution in terms of molecular weight is obtained: W(M) = dw = dw . dVei d(log M) dVe1 d(log M) = __dw__ . loge d(logM) M loge =W(logM)· (2.42) M From the linear distribution the, W(M), the number distribution, P(M), can be calculated: P(M)= W(M) (2.43) M Background 35 2.3.1.2 Matrix-Assisted Laser Desorption Ionization (MALDI) Time-Of -Flight (TOF) Mass Spectrometry Recent years has seen the emergence of mass spectrometry techniques, such as Matrix Assisted Laser Desorption Ionization Time Of Flight Mass Spectrometry (MALDI-TOF MS), for obtaining the absolute molecular weight distribution of synthetic polymers.61- 64 MALDI-TOF-MS uses a soft ionization technique, which vaporizes the polymer molecules without fragmentation of the chains. In principle, the sample and the matrix solution are mixed in such a way that matrix separation of the polymer is achieved. A laser pulse excites the matrix to photo excite the matrix material. The excitation explodes the matrix causing both matrix and polymer to be transferred to the gas phase. Once the analyte is ionized (using cations added to the matrix as a salt), it is accelerated and analysed in a time of flight mass spectrometer. As a result, the analyte is separated according to the mass/charge ratio, and an oligomer distribution is obtained. The distributions of kinetic and potential energies are negated by the use of delayed extraction and a reflectron flight chamber. MALDI-TOF-MS is able to resolve the individual peaks, each being separated by the mass of the monomeric unit, of a molecular weight distribution and even isotopic resolution can be achieved. This allows the identification of endgroups present in the polymer sample.65 Chapter 2 36 2.3.2 Rate Analysis 2.3.2.1 Gravimetry In gravimetry, the sample is collected after a period of time and weighed. The polymer < is isolated by evaporating off the monomer, and the polymer is weighed. The ratio of the mass of polymer to the mass of the monomer and polymer gives the conversion attained at the time the sample was taken. By taking samples periodically will give an indication of the rate profile, or the final conversion can be obtained 2.3.2.2 Dilatometry In theory dilatometry is a very simple technique, which requires only the containing of a known volume of liquid in a reaction tube with a capillary attached. The dilatometer follows the movement of a meniscus in the capillary. In reality the technique is not quit so simple. Dilatometry utilises the volume change to follow conversion verses time. As the monomer is converted into polymer the overall density increases, leading to a decrease on voume of the reaction mixture. A dilatometer generally consists of a reaction flask fitted with a capillary. Changes in the reaction volume are noted in the change of height in the meniscus. The fractional conversion, x(t), can then be related to change in height, h(t), of the meniscus the following equation. a x(t) = h(t)-- (2.44) . mo·f where a is the cross sectional area of the capillary tube, m0 is the initial mass of monomer, and f is the contraction factor. Assuming ideal mixing f is given by the equation. Background 37 1 1 J=--- (2.45) Pmon Ppoly where Pmon is the density of the monomer, and Ppoiy is the density of the polymer. A ~problem that can occur is that there is considerable heat given of during polymerization, (about 50kJ mor1 for MMA); thus the heat must be removed efficiently. As the dilatometer acts as a thermometer and any fluctuation in reaction temperature will be noted in the height of the meniscus. This problem can be overcome with the correct design of ampoule. 2.4 References 1. Heuts, J.P. A.; Gilbert, R. G.; Radom, L. Macromolecules, 28, 8771 (1995). 2. Deady, M.; Mau, A. W. H.; Moad, G.; Spurling, T. H. Makromol. Chern., 194, 1691 (1993). 3. Krstina, J.; Moad, C. L.; Moad, G.; Rizzardo, E.; Berge, C. T.; Fryd, M. Macromol. Symp., 111, 13 (1996). 4. Moad, G.; Rizzardo, E.; Solomon, D. H.; Beckwith, A. L. J. Polym. Bull., 29, 647 (1992). 5. Gridnev, A. A.; Ittel, S.D. Macromolecules, 29, 5864 (1996). 6. Moad, G. Soloman., D.H., in Comprehensive Polymer Science, Eastmond, G.C.; Ledwith, A., Russo, S., Sigwalt, P., Eds., 1989, Permagon: London. 7. Benson, S. W.; North, A.M. J. Am. Chern. Soc. 1962,84,935. 8. See for example: Atkins, P. W. Physical Chemistry; 3rd ed.; Oxford University Press: Oxford, 1987. 9. See for example: Gilbert, R. G. Emulsion Polymerisation, A Mechanistic Approach; Academic Press: London, 1995. 10. See for example: Odian, G. Principles of Polymerization; 2nd ed.; Wiley: New York, 1981. 11. Olaj, 0. F.; Bitai, 1.; Hinkelmann, F. Makromol. Chern., 188, 1689 (1987). Chapter 2 38 12. Mayo, F. R. J. Am. Chern. Soc., 65,2324 (1943). 13. See for example: Polymer Handbook; Wiley Interscience: New York, 1989. 14. Heuts, J. P. A.; Kukulj, D.; Forster, D. J.; Davis, T. P. Macromolecules, 31, 2894 (1998). 15. Suddaby, K. G.; Maloney, D. R.; Haddleton, D. M. Macromolecules, 30, 702 '(1997). 16. Stickler, M.; Meyerhoff, G. Makromol. Chern., 179, 2729 (1978). 17. Moad, G.; Moad, C. L. Macromolecules, 29, 7727 (1996). 18. Kapfenstein, H. M.; Heuts, J. P. A.; Davis, T. P. Macromol. Chern. Phys., In preparation. (1999) 19. Heuts, J.P. A.; Forster, D. J.; Davis, T. P.; Yamada, B.; Yamazoe, H.; Azukizawa, M. Macromolecules, 32, 2511-2519 (1998). 20. Whang, B. Y. C.; Ballard, M. J.; Napper, D. H.; Gilbert, R. G. Aust. J. Chern., 44, 1133 (1991). 21. Clay, P. A.; Gilbert, R. G. Macromolecules, 28, 552 (1995). 22. Christie, D. I.; Gilbert, R. G. Macromol. Chern. Phys., 197,403 (1996). 23. Christie, D. I.; Gilbert, R. G. Macromol. Chern. Phys., 198, 663 (1997). 24. Kapfenstein, H. M.; Davis, T. P. Macromol. Chern. Phys., 199,2403 (1998). 25. Heuts, J.P. A. Davis, T.P.; and Russell, G. T. submitted 1999. 26. Enikolopyan, N. S.; Smirnov, B. R.; Ponomarev, G. V.; Belgovskii, I. M. J. Polym. Sci., Polym. Chern. Edn., 19, 879 (1981). 27. Smirnov, B. R.; Morozova, I. S.; Pushchaeva, L. M.; Marchenko, A. P.; Enikolopyan, N. S. Dokl. Akad. Nauk USSR (Engl. Transl.), 255, 609 (1980). 28. Smirnov, B. R.; Plotnikov, V. D.; Ozerkovskii, B. V.; Roshchupkin, V. P.; Yenikolopyan, N. S. Polym. Sci. USSR, 23, 2807 (1981). 29. Smirnov, B. R.; Marchenko, A. P.; Plotnikov, V. D.; Kuzayev, A. 1.; Yenikolopyan, N. S. Polym. Sci. USSR, 23, 1169 (1981). 30. Burczyk, A. F.; O'Driscoll, K. F.; Rempel, G. L. J. Polym. Sci., Polym. Chern. Edn., 22, 3255 (1984). 31. Gridnev, A. A. Polym. Sci. USSR, 31,2369 (1989). 32. Gridnev, A. A.; lttel, S. D.; Wayland, B. B.; Fryd, M. Organometallics, 15, 5116 (1996). Background 39 33. Gridnev, A. A.; lttel, S. D.; Fryd, M.; Wayland, B. B. Organometallics, 15, 222 (1996). 34. Kukulj, D.; Davis, T. P. Macromol. Chern. Phys., 199, 1697 (1998). 35. Gridnev, A. A.; lttel, S. D.; Fryd, M.; Wayland, B. B. Organometallics, 12, 4871 (1993). 36;Gridnev, A. A.; lttel, S. D.; Fryd, M.; Wayland, B. B. J. Chern. Soc., Chern. Commun., 1010 (1993). 37. Wayland, B. B.; Poszmik, G.; Mukerjee, S. L.; Fryd, M. J. Am. Chern. Soc., 116, 7943 (1994). 38. Wayland, B. B.; Basickes, L.; Mukerjee, S.; Wei, M.; Fryd, M. Macromolecules, 30, 8109 (1997). 39. Wayland, B. B.; Mukerjee, S.; Poszmik, G.; Woska, D. C.; Basickes, L.; Gridnev, A. A.; Fryd, M.; Ittel, S. D. Control of Radical polymerizations by Metalloradicals; Wayland, B. B.; Mukerjee, S.; Poszmik, G.; Woska, D. C.; Basickes, L.; Gridnev, A. A.; Fryd, M.; lttel, S.D., Ed.; American Chemical Society: Washington, DC,; 685, 305 (1998). 40. Avranitopoulos, L. D.; Greuel, M. P.; King, B. M.; Shim, A. K.; Harwood, H. J. Photochemical Polymerizations Initiated and Mediated by Soluble Organocobalt Compounds; Avranitopoulos, L. D.; Greuel, M. P.; King, B. M.; Shim, A. K.; Harwood, H. J., Ed.; American Chemical Society: Washington, DC,;. 685, 316 (1998). 41. Gridnev, A. A.; Belgovskii, I. M.; Enikolopyan, N. S. Dokl. Akad. Nauk USSR (Engl. Transl.), 289, 1408 (1986). 42. Scheffold, R., Rytz, G. and Walder, L. Modern synthetic methods, 3, 355. 43. Wayland, B. B.; Gridnev, A. A.; Ittel, S. D.; Fryd, M. Inorg. Chern., 33, 3830 (1994). 44. Davis, T. P.; Kukulj, D.; Haddleton, D. M.; Maloney, D. R. Trends Polym. Sci., 3, 365 (1995). 45. Gridnev, A. A. Polym. J., 24,613 (1992). 46. Carlson, G. M. a. A., K.J. US Patent 4 526 945; Carlson, G. M. a. A., K.J., Ed.; Carlson, G.M. and Abbey, K.J.: US, (1985). 47. Janowicz, A. H. US Patent 4 694 054; Janowicz, A. H., Ed.; Janowicz: US, (1987). Chapter 2 40 48. Lin, J. C., Abbey, K.J.,US Patent 4 680 354; Lin, J. C., Abbey, K.J., Ed.; Glidden: US, (1987). 49. Janowicz, A. H. a.M., R.L. US Patent 4 680 352; Janowicz, A. H. a.M., R.L., Ed.; Janowicz: US, (1987). 50. Melby, L. R., Janowicz, A.H. and lttel S.D. European Patent 0 119 436 Al; Melby, 'L. R., Janowicz, A.H. and lttel S.D., Ed.; Dupont: US, (1986). 51. Janowicz, A. H. US Patent 5 028 677; Janowicz, A. H., Ed.; Dupont: US, (1991). 52. Suddaby, K. G.; Maloney, D. R.; Haddleton, D. M. Macromolecules, 30, 702-713 (1997). 53. Haddleton, D. M.; Maloney, D. R.; Suddaby, K. G.; Muir, A. V. G.; Richards, S. N. Macromol. Symp., 111, 37 (1996). 54. Sanayei, R. A. a. 0. Driscoll, K.F. J. Macromol. Sci.-Chem., A26(8), 1137-1149 (1989). 55. Halpern, J. Science, 227, 869 (1985). 56. Ng, F. T. T.; Rempel, G. L.; Halpern, J. J. Am. Chem. Soc., 104, 621 (1982). 57. Suddaby, K. G.; ODriscoll, K. F.; Rudin, A. J. Polym. Sci., A: Polym. Chem., 30, 643 (1992). 58. Grubisic, Z., Benoit, H.R.j. polym. sci., polym letters edn., 5, 753 (1997). 59. See for example: Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, New York, (1953). 60. Shortt, D. W. J. Liquid Chromat., 16, 3371 (1993). 61. Bahr, U.; Deppe, A.; Karas, M.; Hillenkamp, F.; Giessman, U. Anal. Chem., 64, 2866 (1992). 62. Castro, J. A.; Koester, C.; Wilkins, C. Rapid Commun. Mass Spectom., 6, 239 (1992). 63. Dey, M.; Castoro, J. A.; Wilkins, C. L. Anal. Chem., 67, 1575 (1995). 64. Danis, P. 0.; Karr, D. E.; Simpsonsick, W. J. Macromolecules, 28, 1229 (1995). 65. Lloyd, P.; Scrivener, E.; Maloney, D. R.; Haddleton, D. M.; Derrick, P. J. ACS Polym. Preprints, 37, 847 (1996). 66. Kukulj, D., PhD Thesis "Mechanistic Studies of Free-Radical Polymerisation", University of New South Wales (1997) The effect of ester chain length and temperature in CCTP of Methacrylates 41 Chapter 3 The Effects of Ester Chain Length and Temperature on the Catalytic Chain Transfer Polymerization of Methacrylates 3.1 Introduction Catalytic chain transfer polymerization is a relatively recent technique for the production of vinyl end-functionalized low-molecular weight polymers via free-radical polymerization.1-5 This technique is based upon the property of certain low-spin Co(ll) complexes to catalyze the chain transfer to monomer reaction (shown in Scheme 1 for the polymerization of methyl methacrylate).2,3,6-11 It has been well-established that virtually all chains are initiated by a hydrogen atom and terminated by a vinyl endgroup,6-8 but despite many efforts,6-19 most notably those by Gridnev and co workers20-27 a fully consistent and unambiguous mechanism has not yet appeared. In mechanistic studies the Arrhenius parameters of the overall reaction rate coefficient provide very valuable information about the kinetics and character of the rate determining step and the nature and importance of substituent effects (for example in a homologous series of monomers). To our knowledge, only two studies of the Arrhenius parameters18,19 and one study of the catalytic chain transfer polymerization of a homologous series of methacrylatesl3 have been published to date. In one study18 a very strong temperature dependence of the catalytic chain transfer reaction in methyl methacrylate was reported. This result was contradicted in a subsequent study,19 based Chapter 3 42 on a limited set of data, which found only a weak temperature dependence. The effect of the size of the ester group in a series of methacrylates was studied by Mironychev et al., 13 and it was found that the chain transfer constant (Cs) of a cobalt(ll) porphyrin in the catalytic chain transfer polymerization of alkyl methacrylates decreases with increasing size of the ester group. However, the published interpretation of this ob~ervation,l3 i.e., a specific complexation between the catalyst and monomer, appears unlikely when considered in the light of the findings of chapter 5 in this thesis. Since no temperature dependence was determined in the earlier work 13, it has never been established whether the reported trend in Cs is constant with temperature (which is unlikely if complex formation plays an essential part in the mechanism). This chapter investigates mechanistic aspects controlling the rate of transfer in catalytic chain transfer polymerization of methacrylates. The first section of this chapter (section 3.2) investigates the catalytic chain transfer polymerization, in the bulk polymerization of the homologous series of methacrylates, and investigates the temperature dependence of CCT in methacrylates. The second part of this chapter (section 3.3) deals with the free-radical polymerization of the viscous monomer 2-phenoxyethyl methacrylate (POEMA) at 60°C. POEMA was studied both in the presence of n-dodecanethiol and a catalytic chain transfer agent (COPhBF). cobaloxime boron fluoride tetra-phenyl cobaloxime boron fluoride (COBF) (COPhBF) Figure 3.1: Structures of the catalytic chain transfer agents used in this chapter. The effect of ester chain length and temperature in CCTP of Methacrylates 43 3.2 Catalytic chain transfer polymerization in the bulk polymerization of the homologous series of methacrylates. The aim of the work reported in this section was to investigate the effect of a homologous series of methacrylates on the kinetics of catalytic chain transfer. The effect ~ of the temperature dependence of the chain transfer constants for both cobaloxime boron fluoride (COBF) and its tetra phenyl derivative (COPhBF) in the catalytic chain transfer polymerization's of methyl, ethyl and butyl methacrylate was also investigated. This work will also determine the effect of the viscosities of methyl, ethyl and butyl methacrylate, on the catalytic chain transfer constants of COBF and COPhBF. 3.2.1 Experimental 3.2.1.1 Synthesis of COBF The bis(methanol) complex of cobaloxime boron fluoride (COBF), more correctly , [bis[J.!-[(2,3butanedione dioximato )(2-)-0:0"]] tetrafluorodiborato(2-)-N,N' ,N" ,N"'] cobalt, was synthesised according to a modification of the method described by Bakac and Espenson28. Cobaloxime boron fluoride (COBF) preparation consists of a two step reaction. Synthesis and isolation of cobaloxime (equation (3.1)), and the substitution of the oxime protons by the BF2 groups (equation (3.2)). The intermediates involved in the synthesis of COBF are extremely air sensitive, the reactions should be carried out in an oxygen free atmosphere. The steps of the procedure were performed using standard schlenk apparatus. Chapter] 44 Co(CH3COO)zAHzO + 2dmgHz ~ Co(dmgH)z.2HzO + 2HCH3COO + 2Hz0 (3.1) Co(dmgH)z.2HzO + 2BF3Etz0 ~ Co(dmgBFz)z.2HzO + 2HF + 2Etz0 (3.2) Co(dmgBF2)z.2H20 + 2MeOH ~ Co(dmgBF2)z.2MeOH + 2Hz0 (3.3) Firstly, 1.9g of dimethyl glyoxime,dmgH2 (Aldrich, 99%), was dissolved in 75ml of analytical grade methanol (the methanol was deoxy_genated by bubbling nitrogen through for at least 1 hour prior to addition.). To this, cobalt (II) acetate tetrahydrate, Co(CH3COO)z.4H20 (2g), was added and allowed to stir at ambient temperature for 4 hrs. The methanol was filtered of using a cannula, and the precipitate was washed once with 20ml of deoxygenated methanol. The dark red solid was dried under vacuum To the dried cobaloxime, 50 ml of deoxygenated diethyl ether was added and the slurry stirred in an ice bath until the mixture is at approx 0°C. lOml of boron trifluoride diethyl etherate, BF3Et20, was added to the reaction flask, and was allowed to stir overnight, warming to room temperature. The solid product is collected and washed with 10ml of diethyl ether and dried. The product was then twice washed in 20ml of deoxygenated water then twice washed in deoxygenated methanol then dried and collected. Yields were typically around 50% due to some loss of the product in the wash solvents. Yield in this synthesis is sacrificed in the final washes to increase catalyst purity. It is very difficult to characterize the catalyst by conventional methods, so the purity of the catalyst is determined by measuring the transfer constant of the catalyst in the bulk polymerization of MMA at 60°C. Ideally the catalyst should have a transfer constant 4 above 3 x 10 . The tetraphenyl version of COBF (COPhBF) was synthesised using the same route replacing dimethyl glyoxime with equimolar amounts of diphenyl glyoxime. The effect of ester chain length and temperature in CCTP of Methacrylates 45 3.2.1.1 General Polymerization Procedure Methyl (MMA; Aldrich, 99% ), ethyl (EMA; Aldrich, 99%) and butyl methacrylate (BMA; Aldrich, 99%) were purified by passing through a column of activated basic alumina (ACROS, 50-200 micron) and purged with high purity nitrogen (BOC) for 1.5 hours prior to use. AIBN (DuPont) was recrystallized twice from methanol and used as initiatior. Polymerizations were carried out as follows: Two stock solutions were prepared: (i) an initiator stock solution, and (ii) a catalyst stock solution. (i) The initiator solution was prepared by dissolution of approximately 400 mg of AIBN in 150 ml of monomer. (ii) The catalyst stock solution was prepared by dissolution of approximately 3 mg of catalyst into 10 ml of solution (i) and a subsequent 10-fold dillution with solution (i). Five reaction mixtures were then prepared, each containing 4.0ml of initiator solution (i) and 0.00, 0.10, 0.20, 0.30 and 0.40 ml of catalyst solution (ii), respectively. The reaction ampoules, specially modified for use with standard Schlenk equipment, were deoxygenated by two freeze-pump-thaw cycles and subsequently placed in a thermostatted waterbath. Final conversions were maintained below 10%. 3.2.1.3 Molecular Weight Analysis Molecular weight distributions were determined by size exclusion chromatography using a Shimadzu LC-10 AT VP pump, a Shimadzu SIL-10AD VP Autoinjector, a column set consisting of a Polymer Laboratories 5.0 Jlm bead-size guard column (50 x 5 4 7.5 mm) followed by three linear PL columns (10 , 10 and 10\ and a Shimadzu RID lOA differential refractive index detector. Tetrahydrofuran (BDH, HPLC grade) was used as eluent at 1 ml/min. Calibration of the SEC equipment was performed with narrow poly(methyl methacrylate) standards (Polymer Laboratories, molecular weight 6 range: 200 - 1.6·10 ). Mark-Houwink-Sakurada parameters used for the universal 5 1 5 calibration are as follows: {KMMA = 12.8·10- ml·g- , aMMA = 0.697}, {KEMA = 15.5·10- 1 5 1 ml·g- , aEMA = 0.679}, and {KBMA = 14.8·10- ml·g- , aBMA = 0.664}.29 Chapter 3 46 3.2.1.4 Viscosity Measurements Absolute monomer viscosities were measured using an Ostwald viscometer (size A) immersed in a temperature-controlled water bath. 30 The accuracy of the calibration was verified by cross-checking the viscosity obtained for MMA with previously published values by Stickler et at31 3.2.2 Results and Discussion 3.2.2.1 Effect of Alkyl Chain Length in Ester Group One objective of this chapter is to investigate whether there is a systematic effect on the observed chain transfer constant with increasing size of the ester group in the methacrylate monomer. The chain transfer constants were measured for COBF and COPhBF for the bulk polymerizations of methyl, ethyl and butyl methacrylate at 60°C. Tables 3.1 and 3.2, show the experimental results for the transfer constant measurements for COBF and COPhBF respectively. The corresponding Mayo-plots are shown in Figures 3.2 and 3.3. Table 3.1: Experimental Results for the Determination of the Chain Transfer Constant of COBF at 60°C for Meth;yl~ Eth;yl and But;yl Methacr;ylate Methyl Methacrylate Ethyl Methacrylate Butyl Methacrylate [Co]/[M] Mw/10~ [Co]/[M] Mw/10~ [Co]/[M] Mw/10~ 0.00 469 0.00 514 0.00 998 9.84·10-8 54.7 1.42·10-7 52.6 1.82·10-7 91.0 1.92·10-7 28.1 2.77·10-7 27.8 3.55·10-7 45.1 2.81·10-7 17.6 4.06·10-7 19.7 5.19·10-7 32.4 3.67·10-7 16.8 5.30·10-7 15.5 6.77·10-7 26.7 Cs= 34·10~ Cs= 27·10~ Cs= 16·10~ The effect of ester chain length and temperature in CCTP of Methacrylates 47 Table 3.2: Experimental Results for the Determination of the Chain Transfer Constant of COPhBF at 60°C for Meth;rl2 Eth;rl and But;rl Methacr;rlate Methyl Methacrylate Ethyl Methacrylate Butyl Methacrylate [Co]/[M] Mw/10~ [Co]/[M] Mw/10~ [Co]/[M] Mw/10~ 0.00 420 0.00 684 0.00 921 1.01·10"7 83.0 1.31·10"7 110 1.73·10"7 183 1.96·107 63.3 2.55·10"7 49.2 3.37·10"7 76.4 2.88·10"7 39.6 3.74·10"7 36.3 4.94·10"7 54.1 3.75·10"7 25.4 4.87·10"7 25.7 6.43·10"7 43.4 Cs= 19·10~ Cs= 17·10~ Cs= 10·10~ 0.012 ~= 0.008 ...... 0.004 O.(XX) +---"T""---,r---r----.--..-..,.--,..----.-..,.---,..----.-..,.----.----r--1 0 Ixl0-7 2xl Figure 3.2: Mayo plots for the determination of the chain transfer constants of COBF in the free-radical polymerizations of (e) MMA, (6.) EMA and (T) BMA. Data based upon DPn-1 =2mo/Mw. Chapter 3 48 0.010-r------, 0.008 • 0.006 ~c - 0.004 0.002 0.000 +----,--,----,r---,----,r---,--.---,--.---,--..---r-...----t 0 [Co]/[M] Figure 3.3: Mayo plots for the determination of the chain transfer constants of COPhBF in the free-radical polymerizations of (e) MMA, (.6.) EMA and (T) BMA. Data based upon DPn-1 =2mo/Mw. It is clear from these results that the chain transfer constants for both COBF and COPhBF decrease with increasing size of the ester group. This result is in accordance with the data previously reported by Mironychev et al.l3 for the chain transfer constants of a cobalt(ll) porphyrin in the free-radical polymerization of a wide range of methacrylates at 60°C. These workers ascribed the reduction in chain transfer constant mainly to an increasing steric hindrance with bulky ester side chains and an enhanced stability of the complex formed upon axial ligation of the monomer with the Co(ll) center.l3 Although the possibility that these factors play a role cannot be discounted, it is doubtful that they dominate. The work reported in chapter 5, showed that direct participation of monomer molecules in the hydrogen-abstraction step is unlikely and that solvent effects through axial ligation are small unless strong ligands, such as pyridine, are included in the reaction mixture. An alternative interpretation of the data is that the results are consistent with a diffusion-controlled hydrogen-abstraction rate coefficient. The effect of ester chain length and temperature in CCTP of Methacrylates 49 3.2.2.2 Temperature Effects Tables 3.3 and 3.4 presents the results determined for chain transfer constants COBF and COPhBF, respectively, with the three monomers at 40, 50, 60 and 70°C. It is clear from these results (the transfer constants vs temperature plots are shown in Figures 3.4 and 3.5), that to within experimental error, no significant temperature effects are observed for either catalyst in any of the studied monomers. This conclusion accords with results obtained in a previous study by Kukulj19 (ie only a small temperature dependence was reported for COBF in methyl methacrylate). However, this finding contradicts conclusions made in a different study where an activation energy for the chain transfer constant of COBF in methyl methacrylate of -10.1 kJ·mor1 was reported.l8 This apparent anomaly was investigated by re-analyzing the earlier data,18 3 3 3 3 yielding Cs-values of 33.5·10 , 32.8·10 , 35.2·10 , and 32.5-10 at 60, 70, 80, and 90°C, respectively, indicating the absence of any significant temperature dependence. Table 3.3: Chain Transfer Constants of COBF for Methyl, Ethyl and Butyl Methacr;rlate in the Tem~erature Range of 40 - 70°C Temp a Methyl methacrylate Ethyl methacrylate Butyl methacrylate Cs (Mn)jj Cs (Mwt Cs (Mn)jj Cs (Mw)c Cs (Mn)b Cs (Mw)c 40 38·103 33·103 30·103 27·103 14·103 15·103 50 44·103 39·103 30·103 30·103 15·103 16·103 60 40·103 34·103 26·103 24·103 16·103 16·103 27·103 27·103 70 31·103 28·103 26·103 24·103 16·103 15·103 28·103 27·103 24·103 22·103 a) Temperature in oc; b) Chain transfer constant of COBF determined using DPn = Mnf(monomer mass); c) Chain transfer constant of COBF determined using DPn =Mw/(2·monomer mass) Chapter 3 50 Table 3.4. Chain Transfer Constants of COPhBF for Methyl, Ethyl and Butyl Methacrilate in the Tem~erature Range of 40 - 70°C Tempa Methyl methacrylate Ethyl methacrylate Butyl methacrylate Cs (Mn)6 Cs (Mw)c Cs (Mn)b Cs (Mwl Cs (Mn)6 Cs (Mwl 40 18·103 19·103 15·103 17·103 6.5-103 7.8·103 19·103 20·103 7.7·103 8.1·103 9.4·103 8.8·103 50 19·103 18·103 15·103 17·103 9.2·103 10·103 16·103 16·103 7.5·103 10·103 9.0·103 10·103 60 21·103 19·103 17·103 17·103 9.7·103 10·103 17·103 18·103 8.6·103 9.5·103 18·103 17·103 9.9·103 11·103 70 21·103 20·103 15·103 17·103 11·103 8.4·103 16·103 17·103 9.3·103 9.5·103 17·103 17·103 8.2·103 8.9·103 a) Temperature in °C; b) Chain transfer constant of COPhBF determined using DPn = Mn/(monomer mass); c) Chain transfer constant of COPhBF determined using DPn = Mw/(2·monomer mass). The effect of ester chain length and temperature in CCTP of Methacrylates 51 5.0xHf 4.0xHf- • • 3.0xHf • A A I cY A t 2.0xHf ...... l.Oxlcf 0.0 _.__---.----.------r-----r----r-----r-----r----1 40 50 60 70 Temperature CCC) Figure 3.4: Temperature dependence of the chain transfer constants of COBF in the free-radical polymerizations of (e) MMA, (.._) EMA and (T) BMA. In the temperature range of 40 - 70°C 25xlcf 2.0x104 A l • A• l A A A 15x104 cY l.Oxlcf- ... 5.0xlo3- ' ' 0.0 -'---.----.------r-----r-----r----r-----.-----1 40 50 60 70 Temperature CCC) Figure 3.5: Temperature dependence of the chain transfer constants of COPhBF in the free-radical polymerizations of (e) MMA, (.._) EMA and (T) BMA. In the temperature range of 40 - 70°C Chapter 3 52 To facilitate discussion of the Arrhenius parameters it is useful to introduce the important parameters by means of equation (3.4): B s -B P) = Aexp ( -~B ) (3.4) RT RT where As, Ap and A are the pre-exponential factors of the chain transfer rate coefficient, propagation rate coefficient and chain transfer constant, respectively, and Bs, Ep, and Bact are the activation energies of the chain transfer rate coefficient, propagation rate coefficient and chain transfer constant, respectively. The Arrhenius parameters calculated for Cs and those calculated for ktr are listed in Table 3.5. Table 3.5: Arrhenius parameters for Cs and ktr for the polymerization of Methyl, Ethil and But!l Methacr!late, with COBF and COPhBF Cs ktr Catalyst Monomer Baa) A b) Baa) Ac) 3 10 COBF MMA -5.5 4.3 X 10 15.3 0.6 X 10 3 10 BMA -5.5 3.5 X 10 16.8 1.0 X 10 3 10 BMA -1.2 9.8 X 10 19.3 1.8 X 10 4 10 COPhBF MMA 1.4 3.2 X 10 23.3 6.4 X 10 4 10 BMA -1.7 0.9 X 10 21.3 3.6 X 10 4 10 BMA 2.3 2.1 X 10 22.8 3.9 X 10 a) kJ·mor1 b) dimensionless From the chain transfer constants listed in Tables 3.3 and 3.4 it is clear that Bact= 0 in all cases and therefore Bs = Ep for all three monomers (Table 3.5). This in turn implies The effect of ester chain length and temperature in CCTP of Methacrylates 53 that the pre-exponential factor of the chain transfer rate coefficient is approximately given by: (3.5) Using this expression yields values for As of the order of -1010 dm3 mor1 s- 1 for all three monomer systems, which is in agreement with previous work.18,19 The value of the activation energy, Es = 17 kJ·mor1 (for COBF) and Es = 23 kJ·mor1 (for COPhBF), for the chain transfer reaction, simply indicates the presence of a reaction barrier; unfortunately it is not possible to draw significant mechanistic conclusions on the basis of this result alone. More information can be obtained from the 10 3 1 1 pre-exponential factor As which is of the order of -10 dm mor s- • This value is atypical for a chemically-controlled bimolecular (elementary) reaction between a polymeric radical and a small molecule, as experimental data,32-36 backed up by theoretical transition state theory arguments,37-41 suggest that typical pre-exponential 6 8 3 1 1 factors are several orders of magnitude lower (ie, -10 - 10 dm mor s- ). The much higher value for As reported here is therefore indicative of an unusual chain transfer mechanism. High As values are more likely to signify a unimolecular rate-determining 1 step37 (in which case the units of As should be s- ). A unimolecular rate-determining step is feasible if the mechanism of catalytic chain transfer involves a ~-elimination of a hydrogen atom from a radical coordinated to the cobalt center. 27,42 Although this possibility cannot be discarded at present, it seems unlikely due the findings in chapter 5. An alternative explanation, already alluded to in previous studies,2,11,19 is that the chain transfer reaction involving methacrylate monomers is diffusion-controlled. The 7 3 1 1 chain transfer rate coefficients, ks ( -10 dm mor s- ) are similar to the rate coefficients obtained for bimolecular termination reactions in free-radical polymerization- these are known to be diffusion-controlled.43 Chapter3 54 For a diffusion-controlled reaction, the rate coefficient is often expressed in terms of the Smoluchowski equation,43-45 which states that the rate coefficient is proportional to the mutual diffusion coefficients of the two reactants, which can be approximated by the sum of the two self-diffusion coefficients: 43 (3.6) Here Dco is the diffusion coefficient of the cobalt catalyst and Di the diffusion coefficient of an i-meric radical. This implies that the temperature dependence of the diffusion coefficients will dictate the temperature behaviour of the rate coefficient.46 An extensive study of the self-diffusion of toluene in a polystyrene solution has shown that the activation energy of the diffusion coefficient is about 11 kJ·mort at zero polymer concentrations and increases with increasing polymer content.47 Furthermore, diffusion coefficients at 25°C and 40°C reported by Griffiths et al. 48 for the diffusion of methyl methacrylate and butyl methacrylate oligomers in rubbery polymer matrices suggest that the activation energy for the diffusion of these species has a value between about 10 and 40 kJ·mort. Hence, an activation energy of -17 - 23 kJ·mort is not inconsistent with a diffusion-controlled reaction. An estimated value for the pre exponential factor in a diffusion-controlled reaction can be obtained from the reaction rate coefficient for the recombination reaction of two primary radicals, which is of the order of 109 dm3 mort s-t.43 In conjunction with an activation energy of about 11 kJ·mort (taken from the toluene study, vide supra), this corresponds to a pre exponential factor of about 10to- 1011 dm3 mort s-t. The estimate from the current study, i.e., As - 10t0 dm3 mort s-t is consistent with this value. The effect of ester chain length and temperature in CCTP of Methacrylates 55 3.2.2.3 The effect of monomer viscosity Further indications of a diffusion-controlled chain transfer reaction can be found by comparing the chain transfer constants obtained for the three different monomers. It is evident from Tables 3.3 - 3.4 that the larger the ester group in this homologous series, the smaller the chain transfer constant. This decrease in chain transfer constant can only be partially explained by an increasing propagation rate coefficient upon increasing size of the ester group.49 Since the chain transfer constants contain the rate coefficients of both chain transfer and propagation, the discrepancy can only be ascribed to the chain transfer rate coefficients, as indicated in equation (3.7), where the chain transfer constants for methyl and butyl methacrylate are compared. CS.MMA = kS,MMA • kp,BMA (3.7) CS,BMA kS,BMA kp,MMA At 40°C, Cs,MMA/Cs,BMA ::::: 2.0 - 2.2 (see Tables 3.3 and 3.4) and kp,BMAikp,MMA ::::: 1.3,32,49 which leaves a factor of about 1.7 unaccounted for. This factor should then correspond to the ratio ks,MMAfks.sMA· If these chain transfer rate coefficients are diffusion-limited then they should be proportional to the diffusion coefficients in these systems (see equation (3.6)), which in tum are generally found to be roughly proportional to the inverse of the solvent viscosity f):50 1 Doc- (3.8) 1'/a where a lies generally between 0.5 and 1.0. 50 Since the chain transfer rate coefficient is proportional to the sum of the self-diffusion coefficients of the cobalt complex and the i meric radicals (see equation (3.6)), it should also be roughly proportional to f)-a. (i.e., ks Chapter 3 56 oc 11-a). Hence, if the chain transfer reaction is diffusion controlled, equation (3.9) should apply: (3.9) Equation (3.9) in tum implies the following: (3.10) In Table 3.6, the individual values for Cs, kp, and measured monomer viscosities are listed. The equality specified in equation (3.10) is tested for MMA, EMA and BMA at 40, 50, 60 and 70°C using three different values for a, viz. 1, 0.7 and 0.5.50; the results are shown in Table 3.7. The data in Table 3.7 clearly suggest the validity of equation (3.10) within experimental uncertainty, which provides further evidence for a diffusion-controlled chain transfer reaction. The effect of ester chain length and temperature in CCTP of Methacrylates 57 Table 3.6: Summary of relevant kinetic and physical data of the studied methacrylates at 40, 50, 60 and 70°C. Monomer Temperaturea kcp 3 Methyl 40 33 X 10 497 0.45d Methacrylate 3 50 39 X 10 649 0.40d 3 60 34 X 10 833 0.37d 3 70 27 X 10 1054 0.33d 3 Ethyl 40 27 X 10 555 0.48e Methacrylate 3 50 30 X 10 728 0.44e 3 60 26 X 10 946 0.39e 3 70 23 X 10 1196 0.35e 3 Butyl 40 15 X 10 676 0.70e Methacrylate 3 50 16 X 10 863 0.62e 3 60 16 X 10 1085 0.55e 3 70 16 X 10 1347 0.49e a) Temperature in °C; b) Chain transfer constants of COBF taken from Table 3.3 (Mw results); 3 1 1 c) Propagation rate coefficient (dm ·mor -s. ) taken from ref 29; d) Absolute viscosity (centipoise) of MMA taken from ref 30; e) Absolute viscosity of the monomer (centipoise) determined in the present study. Chapter3 58 Table 3.7. Comparison of Values for Cskp'Yia for the COBF-mediated Free-Radical .... Polymerizations of Methyl, Ethyl and Butyl Methacrylate Temp a MMA EMA BMA 40 7.8·106 7.2·106 7.3·106 7 6 6 ~ Cs kp 11 50 1.0·10 9.5·10 8.5·10 60 1.0·107 1.0·107 9.3·106 70 9.7·106 9.7·106 1.0·107 40 9.9·106 9.0·106 8.1·106 7 7 6 Cs kp ,o.7 50 1.3·10 1.2·10 9.9·10 60 1.4·107 1.3·107 1.1·107 70 1.3·107 1.3·107 1.3·107 40 1.2·107 1.0·107 8.7·106 7 7 7 Cs kp ,o.s 50 1.6·10 1.4·10 1.1·10 60 1.7·107 1.6·107 1.3·107 70 1.7·107 1.7·107 1.5·107 a) Temperature in °C. 3.2.2.4 The effect of catalyst structure: The effect of catalyst structure on the chain transfer constant can be easily seen in Tables 3.3 - 3.4, where COPhBF consistently displays a chain transfer constant about 50% less than those observed with COBF. A possible explanation for this observation was previously published by Haddleton et al., 11 who suggested that the lower chain transfer constants for COPhBF may be ascribed to a larger cross-sectional area of COPhBF compared to COBF, slowing down the diffusion process. The effect of ester chain length and temperature in CCTP of Methacrylates 59 3.3 Conventional and catalytic chain transfer in the free-radical polymerization of 2-phenoxyethyl methacrylate In the previous section strong evidence was shown that diffusion controls the rate of tra:gsfer in the catalytic chain transfer polymerization of methacrylates. In this section we further probe the idea of a diffusion-controlled chain transfer step in catalytic chain transfer polymerization. In this section of work, the chain transfer behaviour of 2- phenoxyethyl methacrylate (POEMA, Figure 3.6) with n-dodecanethiol and COPhBF at 60°C, will be investigated. If one compares the chain transfer reaction of n-dodecanethiol, which is chemically controlled, within a homologous series, to catalytic chain transfer, the picture is very different. Based upon transition state theory arguments it has been suggested39 that the chain transfer constant of a thiol in a homologous series should be nearly constant; since the transition states of the chain transfer and the propagation reactions are similar enough for the changes in ktr,DDM and kp caused by a changing monomer to cancel, one does not expect great changes in CnnM (=ktr,DnM/kp) within a homologous series.39 This argument is substantiated by results from Hutchinson et al. 51 who measured a constant CnnM (- 0. 7) for MMA, EMA and BMA. <0 Figure 3.6: Structure of 2-Phenoxyethyl methacrylate, POEMA Chapter 3 60 3.3.1 Experimental 3.3.1.1 Materials The tetra-phenyl derivative of cobaloxime boron fluoride was prepared as described in section 3.2.1.1. The monomer 2-phenoxyethyl methacrylate (POEMA; Sartomer) was r purified by vacuum distillation prior to use, n-dodecanethiol (DDM; Aldrich, 98%) was used without further purification, and 2,2' -azobisisobutyronitrile (AIBN; DuPont) was recrystallized twice from methanol and used as initiator. 3.3.1.2 Measurement of chain transfer constant to DDM A stock solution of approximately 100 mg AIBN in 130 ml POEMA was prepared from which four samples of -5 ml were transferred into glass ampoules (all quantities were accurately weighed). To each of these solutions a varying amount of n-dodecanethiol (ranging roughly from 20 to 100 mg) was added and the solutions were subsequently purged with high purity nitrogen gas (BOC gases) for 10 minutes. The ampoules were subsequently sealed with rubber septa and placed in a thermostatted water bath. Conversions were maintained below 5%. 3.3.1.3 Measurement of chain transfer constant to COPhBF Measurements of the chain transfer constant of COPhBF in POEMA were carried out as described previously. Two stock solutions were prepared: (a) an initiator stock solution I, and (b) a catalyst stock solution II. (a) The initiator solution was prepared by dissolution of approximately 100 mg of AIBN in 130 ml of monomer (-5·10-3 M) - solution I. (b) The catalyst stock solution was prepared by dissolution of approximately 3 mg of catalyst into 10 ml of solution I and a subsequent 10-fold dilution with solution I- solution II. Five reaction mixtures were then prepared, each containing 4.0 ml of initiator solution I and 0.10, 0.20, 0.30, 0.40 and 0.50 ml of catalyst solution II, respectively. The reaction ampoules, specially modified for use with standard Schlenk equipment, were deoxygenated by two freeze-pump-thaw cycles and subsequently placed in a thermostatted waterbath. Final conversions were maintained below 10%. Each replicate was performed independently. The effect of ester chain length and temperature in CCTP of Methacrylates 61 3.3.1.4 Molecular weight analysis Molecular weight distributions were determined by size exclusion chromatography using a Shimadzu LC-10 AT VP pump, a Shimadzu SIL-lOAD VP Autoinjector, a column set consisting of a Polymer Laboratories 5.0 J..tm bead-size guard column (50 x 5 4 7.5 mm) followed by three linear PL columns (10 , 10 and 10\ and a Shimadzu RID lOA differential refractive index detector. Tetrahydrofuran (BDH, HPLC grade) was used as eluent at 1 ml/min. Calibration of the SEC equipment was performed with narrow poly( methyl methacrylate) standards (Polymer Laboratories, molecular weight 6 range: 200 - 1.6·10 • No special consideration was given to the Mark-Houwink Sakurada parameters of poly(POEMA), but it is expected that the introduced errors arising from this approximation are small. 3.3.1.5 Viscosity measurements Absolute monomer viscosities were measured using an Ostwald viscometer (size A) immersed in a temperature-controlled water bath at 60°C. 30 The calibration of the capillary was further refined by cross-checking the viscosity of MMA with those obtained by Stickler et al. 31 3.3.1.6 Pulsed laser polymerization The propagation rate coefficient of POEMA at 60°C was measured using the technique of pulsed laser polymerization.52 Purified monomer and the photoinitiator benzoin were weighed into pyrex sample tubes (10 mm diameter by 60mm height), which were then purged with high purity nitrogen for 5 minutes and sealed with rubber septa. The reaction mixtures were equilibrated at the reaction temperature prior to laser exposure. The polymerizations were initiated by a pulsed Nd:Yag laser (Continuum Surelite 1-20) with a harmonic generator (a Surelite SLD-1 and SLT in series), which was used to produce the 355 nm UV laser radiation, and a wavelength separator (Surelite SSP-2), which was used to isolate the 355 nm beam. Constant pulsing frequencies of 10 and 6.67 Hz were used. Polymerization activity was terminated by removing the sample Chapter 3 62 from the laser, and precipitating the polymer into methanol. The molecular weight distributions of obtained polymers were determined subsequently by size exclusion chromatography. 3.~.2 Results and Discussion 3.3.2.1 Propagation rate coefficient In order to determine the transfer rate coefficient, we need to know the value of the propagation rate coefficient of POEMA at 60°C. This value was determined using pulsed laser polymerization (PLP) using the inflection point molecular weight (Minf) as a measure for the characteristic chain length Lo (i.e. Lo =Minr/monomer mass) in equation (3.11): (3.11) where v is the pulsing frequency of the laser. Typical molecular weight distributions obtained against a polyMMA calibration curve are shown in Figure 3.7 and it is clear that PLP characteristics are observed. It is of interest to note that the number of overtones is relatively large (i.e at least the 2nd and 3rd overtone can be easily distinguished) compared to the typical MMA PLP derived molecular weight distribution, which under the same conditions shows only a small secondary overtone as shown in Figure 5.18. It is well known that bimolecular radical termination is diffusion controlled, thus the increased viscosity of POEMA decreases the termination rate coefficient allowing more radicals to survive each laser flash. In Table 3.8 the experimental results are listed and from these results a propagation rate coefficient of 953 ± 30 dm3 mor1 s-1 (against a polyMMA calibration curve!) is obtained. The effect of ester chain length and temperature in CCTP of Methacrylates 63 6.67Hz 0.4 10Hz l l \~... /-·-·---...... ~ /"o...______•••••••...... \ i g 0.2 ~ ' ' ' ' ' ' '• •, '• ''"-·,...... _, M Figure 3. 7 Molecular weight distribution from the pulsed laser polymerization of POEMA at 60°C, for 10Hz and 6.67Hz Table 3.8: Summarl: of ex[!erimental[!ulsed laser [!Oll:merization results at 60°C d ke f kg Exp va pl3 [Mt Minf,l p Minf,2 p 1 10 1.046 5.07 98·103 938 194·103 926 2 10 1.046 5.07 104·103 998 209·103 1000 3 6.67 1.046 5.07 149·103 948 290·103 923 4 6.67 1.046 5.07 150·103 955 292·103 931 a) Pulsing frequency in Hz; b) Monomer density in g·ml-1; 3 c) Monomer concentration in mol·dm- ; d) First inflection point; e) Propagation rate coefficient derived from first inflection point (dm3 mol~ 1 s-1); f) Inflection point of first overtone; g) Propagation rate coefficient derived 3 1 1 from inflection point of first overtone (dm mor s- ) Chapter 3 64 3.3.2.2 Chain transfer constant of COPhBF in 2-Phenoxyethyl methacrylate Chain transfer constants were obtained using the Mayo procedure and the results obtained for COPhBF in POEMA are listed in Table 3.9 and shown in Figure 3.8. 3 There is a good reproducibility of the data and Ceo,POEMA is found to be 2·10 , which is 3 much smaller than Ceo,MMA. previously determined to be around 20·10 • Table 3.9: Summary of experimental results for the determination of the chain transfer constant to COPhBF at 60°C Exp. 1 soln. 1/ml soln. IJ/ml [Co]/[M] · Mn Mw a 0.1 4.0 2.12·10-7 198·103 610·103 b 0.2 4.0 4.15·10-7 168·103 382·103 c 0.3 4.0 6.07·10-7 127·103 282·103 d 0.4 4.0 7.91·10-7 113·103 229·103 e 0.5 4.0 9.67·10-7 94.8·103 192·103 3 3 Ceo= 1.5·10 1.9·10 Exp.2 soln. 1/ml soln. IJ/ml [Co]/[M] Mn Mw a 0.1 4.0 2.12·10-7 264·103 595·103 b 0.2 4.0 4.15·10-7 174·103 375·103 c 0.3 4.0 6.07·10-7 130·103 269·103 d 0.4 4.0 7.91·10-7 102·103 210·103 e 0.5 4.0 9.67·10-7 85.9·103 178·103 3 3 Ceo= 2.2·10 2.2·10 The effect of ester chain length and temperature in CCTP of Methacrylates 65 0.0025 0.0020 0.0015 t:l-.1') ...... ~ 0.0010 0.0005 0.0000 0.0 l.Oxl0-6 [COPhBF]/[POEMA] Figure 3.8 Mayo plots for COPhBF in POEMA at 60°C This result is consistent with the idea of a diffusion-controlled chain transfer reaction for the methacrylate series of monomers, as the measured viscosity of POEMA at 60°C, TJroEMA = 2.51 centipoise, is much larger than the viscosity of MMA at 60°C, TJMMA = 0.37 centipoise. If the chain transfer reaction is indeed diffusion controlled, the relationship given by equation (3.10), shown to be applicable to MMA, EMA and BMA, should also apply in the present case. Or, in other words, equation (3.12) should apply: fcco fcco (3.12) \: kP TJ\froEMA ::::: \: kP TJ)MMA Using the value of kp for POEMA at 60°C, we can calculate the values of Cca·kp·TJ for MMA and POEMA: 3 6 MMA: Ceo X "" X TJ = 18.5 ·10 X 833 X 0.37 = 5.7 ·10 (3.13) Chapter 3 66 3 6 POEMA: Ceo X kp X ll = 2·10 X 953 X 2.5 = 4.8·10 (3.14) It is clear that these two values are very similar, which strongly suggests that viscosity effects are very important in the chain transfer reaction, and that it is very likely that the chain transfer reaction in the methacrylate series of monomers is diffusion controlled. A~ observant reader may object that the molecular weight distributions were not analyzed using appropriate Mark-Houwink constants for polyPOEMA. However, since the molecular weight ranges obtained in our CCT and PLP are very similar (see Tables 3.8 and 3.9), the effect of a different set of Mark-Houwink constants will be approximately the same for the molecular weights obtained in the two sets of 1 experiments. Since Ceo oc molecular weighf and kp oc molecular weight, the product of Cs and kp is unlikely to be affected significantly by a different set of Mark-Houwink constants 3.3.2.3 Chain transfer constant of n-dodecanethiol in 2-Phenoxyethyl methacrylate In order to confirm that this large decrease in chain transfer constant is in fact not due to an unusual solvent effect, we measured the chain transfer constant of n-dodecanethiol in POEMA. The obtained results are listed in Table 3.10 and shown in Figure 3.9, and good reproducibility of the data is seen. It can be seen that a chain transfer constant CooM,POEMA of about 0.7 is obtained, which is close to the ones previously obtained in MMA (0.751 and 0.836). This is indeed what is expected from theory, as mentioned in the introduction. If any unusual steric or electronic effects were important in this system, a larger difference would have been observed. The effect of ester chain length and temperature in CCTP of Methacrylates 67 Table 3.10. Summary of experimental results for the determination of the chain transfer constant to DDM at 60°C Exp.1 mooM I g mpoEMA/ g [DDM]/[M] Mn Mw a 0.0210 5.7670 3.71·10·3 68.2·103 139·103 b 0.0479 5.7223 8.52·10-3 36.5·103 65.6·103 c 0.0647 5.8307 11.3·10-3 28.5·103 49.8·103 d 0.0814 5.7104 14.5·10-3 23.6·103 39.9·103 e 0.0984 5.7366 17.5·10-3 19.4·103 33.5·103 CooM= 0.55 0.68 Exp.2 mooM/ g mpoEMA/ g [DDM]/[M] Mn Mw a 0.0198 5.7534 3.50·10-3 64.8·103 133·103 b 0.0445 5.7166 7.93·10-3 37.9·103 68.5·103 c 0.0648 5.6949 11.6·10-3 27.1·103 48.4·103 d 0.0802 5.6840 14.4·10-3 23.5·103 40.6·103 e 0.0979 5.7146 17.4·10-3 20.2·103 34.2·103 CooM= 0.51 0.64 Chapter3 68 0.014 0.012 0.010 0.008 P-.r:: Q- 0.006 0.004 0.002 ~ 0.000 0.000 0.004 0.008 0.012 0.016 0.020 [DDM]/[POEMA] Figure 3.9: Mayo plots for DDM in POEMA at 60°C 3.4 Conclusions The data presented clearly show a dependence of the catalytic chain transfer constant on the size of the ester group in the methacrylate series, i.e., Cs,MMA > Cs,EMA > Cs,BMA for both catalyst tested. Furthermore, a virtually constant chain transfer constant was observed with changing temperature, yielding a pre-exponential factor A of about 1010 and an activation energy of about 17 - 23 kJ mor1 for the rate determining step in the chain transfer reaction. These observations are consistent with a diffusion-controlled rate-determining step in the catalytic chain transfer reaction of methacrylates. The conclusion of a diffusion-controlled rate-determining step is reinforced by an inverse correlation of the Cs values with monomer viscosity, as 'flMMA < 'flEMA < 'flBMA· Furthermore, it was shown that for a diffusion-controlled chain transfer reaction, the The effect of ester chain length and temperature in CCTP of Methacrylates 69 product of Cs, kp and 'lla should be constant - which was confirmed for MMA, EMA and BMA at 40, 50, 60 and 70°C. The conclusion of a diffusion-controlled rate-determining step, in the catalytic chain transfer reaction of methacrylates, was found to be consistent when tested with the highly viscous POEMA. The product of Cs, kp and 'lla, for POEMA at 60°C, was found to be consistent with that of MMA, EMA and BMA at the same temperature. Since the chain transfer constant of n-dodecanethiol in POEMA is found to be very close to the one found in MMA, it can be concluded that the observed effect in the catalytic chain transfer polymerization cannot be attributed to unusual solvent effects. 3.5 References 1. Karmilova, LV.; Ponomarev, G. V.; Smimov, B. R.; Belgovskii, I. M., Russ. Chem. Rev., 53, 132 (1984) 2. Davis, T. P.; Haddleton, D. M.; Richards, S. N., J. Macromol. Sci., Rev. Macromol. Chem. Phys., C34, 234 (1994) 3. Davis, T. P.; Kukulj, D.; Haddleton, D. M.; Maloney, D. R., Trends Polym. Sci., 3, 365 (1995) 4. Krstina, J.; Moad, C. L; Moad, G.; Rizzardo, E.; Berge, C. T.; Fryd, M., Macromol. Symp., 111, 13 (1996) 5. Davis, T. P.; Zammit, M. D.; Heuts, J. P. A.; Moody, K., Chem. Commun. , 2383 (1998) 6. Smimov, B. R.; Morozova, I. S.; Pushchaeva, L M.; Marchenko, A. P.; Enikolopyan, N. S., Dokl. Akad. Nauk USSR (Engl. Trans!.), 255,609 (1980) 7. Smimov, B. R.; Morozova, I. S.; Marchenko, A. P.; Markevich, M.A.; Pushchaeva, L. M.; Enikolopyan, N. S., Dokl. Akad. Nauk USSR (Engl. Trans!.) , 253, 891 (1980) 8. Smimov, B. R.; Plotnikov, V. D.; Ozerkovskii, B. V.; Roshchupkin, V. P.; Yenikolopyan, N. S., Polym. Sci. USSR, 23,2807 (1981) Chapter 3 70 9. Enikolopyan, N. S.; Smimov, B. R.; Ponomarev, G. V.; Belgovskii, I. M., J. Polym. Sci., Polym. Chem. Edn., 19, 879 (1981) 10. Burczyk, A. F.; O'Driscoll, K. F.; Rempel, G. L., J. Polym. Sci., Polym. Chem. Edn., 22, 3255 (1984) 11. Haddleton, D. M.; Maloney, D. R.; Suddaby, K. G.; Muir, A. V. G.; Richards, S. N., "Macromol. Symp., 111, 37 (1996) 12. Heuts, J. P. A.; Kukulj, D.; Forster, D. J.; Davis, T. P. , Macromolecules , 31, 2894 (1998) 13. Mironychev, V. Y.; Mogilevich, M. M.; Smimov, B. R.; Shapiro, Y. Y.; Golikov, I. V., Polym. Sci. USSR, 28,2103 (1986) 14. Smimov, B. R.; Pushchayeva, L. M.; Plotnikov, V. D., Polym. Sci. USSR, 31, 2607 (1989) 15. Smimov, B. R.; Marchenko, A. P.; Plotnikov, V. D.; Kuzayev, A. I.; Yenikolopyan, N. S., Polym. Sci. USSR, 23, 1169 (1981) 16. Suddaby, K. G.; O'Driscoll, K. F.; Rudin, A., J. Polym. Sci., A: Polym. Chem., 30, 643 (1992) 17. Suddaby, K. G.; Maloney, D. R.; Haddleton, D. M., Macromolecules,30,702 (1997) 18. Sanayei, R. A., O'Driscoll, K.F., J. Macromol. Sci.-Chem., A26(8), 1137-1149 (1989) 19. Kukulj, D.; Davis, T. P., Macromol. Chem. Phys., 199, 1697 (1998) 20. Gridnev, A. A.; Belgovskii, I. M.; Enikolopyan, N. S., Dokl. Akad. Nauk USSR (Engl. Trans!.), 289, 1408 (1986) 21. Gridnev, A. A., Polym. Sci. USSR, 31, 2369 (1989) 22. Gridnev, A. A., Polym. J., 24, 613 (1992) 23. Gridnev, A. A.; Ittel, S. D.; Fryd, M.; Wayland, B. B., Organometallics,12,4871 (1993) 24. Gridnev, A. A.; Ittel, S. D.; Fryd, M.; Wayland, B. B., J. Chem. Soc., Chem. Commun.,1010 (1993) 25. Gridnev, A. A.; Ittel, S. D.; Wayland, B. B.; Fryd, M., Organometallics,15, 5116 (1996) 26. Gridnev, A. A.; Ittel, S. D.; Fryd, M.; Wayland, B. B., Organometallics,15, 222 (1996) The effect of ester chain length and temperature in CCTP of Methacrylates 71 27. Gridnev, A. A.; lttel, S.D.; Wayland, B. B.; Fryd, M., Organometallics, 15, 5116- Supplementary material (1996) 28. Bakac, A.; Brynildson, M. E.; Espenson, J. H., Inorg. Chem.,25, 4108 (1986) 29. Hutchinson, R. A.; Paquet, D. A.; McMinn, J. H.; Heuermann, S.; Fuller, R. E.; Jackson, C., DECHEMA Monographs, 131, 467 (1995) 30~ Rabek, J. F. Experimental Methods in Polymer Chemistry; John Wiley & Sons: Chichester, (1983) 31. Stickler, M.; Panke, D.; Wunderlich, W., Makromol. Chem.,lSS, 2651 (1987) 32. Heuermann, S.; Buback, M.; Davis, T. P.; Gilbert, R. G.; Hutchinson, R. A.; Olaj, 0. F.; Russell, G. T.; Schweer, J.; Van Herk, A.M., Macromol. Chem. Phys.,19S, 1545 (1997) 33. Maeder, S.; Gilbert, R. G., Macromolecules ,31, 4410 (1998) 34. Tobolsky, A. V.; Offenbach, J., J. Polym. Sci. , 16, 311 (1955) 35. Buback, M.; Gilbert, R. G.; Hutchinson, R. A.; Klumperman, B.; Kuchta, F.-D.; Manders, B. G.; ODriscoll, K. F.; Russell, G. T.; Schweer, J. , Macromol. Chem. Phys., 196, 3267 (1995) 36. Kapfenstein, H. M.; Heuts, J. P. A.; Davis, T. P., Macromol. Chem. Phys., In preparation (1999) 37. Gilbert, R. G.; Smith, S.C. Theory of Unimolecular and Recombination Reactions; Blackwell Scientific: Oxford, (1990) 38. Heuts, J.P. A.; Gilbert, R. G.; Radom, L., Macromolecules ,28,8771 (1995) 39. Heuts, J.P. A.; Sudarko; Gilbert, R. G., Macromol. Symp.,111,147 (1996) 40. Heuts, J. P. A.; Gilbert, R. G.; Radom, L., J. Phys. Chem., 100,18997 (1996) 41. Heuts, J.P. A.; Pross, A.; Radom, L., J. Phys. Chem.,100,17087 (1996) 42. Brookhart, M.; Green, M. L. H.; Wong, L.-L., Progr. Inorg. Chem.,36,1 (1988) 43. Russell, G. T., Macromol. Theory Simul.,4,497 (1995) 44. Pilling, M. J.; Seakins, P. W. Reaction Kinetics; Oxford University Press: Oxford, (1996) 45. See for example: Atkins, P. W. Physical Chemistry; 3rd ed.; Oxford University Press: Oxford, (1987) 46. Russell, G. T., Macromol. Theory Simul. ,4,549 (1995) 47. Pickup, S.; Blum, F. D., Macromolecules,22,3961 (1989) Chapter 3 72 48. Griffiths, M. C.; Strauch, J.; Monteiro, M. J.; Gilbert, R. G., Macromolecules, 31, 7835 (1998) 49. Zammit, M. D.; Coote, M. L.; Davis, T. P.; Willett, G. D., Macromolecules, 31, 955 (1998) 50. Reid, R. C.; Prausnitz, J. M.; Sherwood, T. K. The Properties of Gases and Liquids; ' 3rd ed.; McGraw-Hill: New York, (1977) 51. Hutchinson, R. A.; Paquet, D. A.; McMinn, J. H., Macromolecules, 28,5655 (1995) 52. Olaj, 0. F.; Bitai, 1.; Hinkelmann, F., Makromol. Chem.,188, 1689 (1987) CCTP of methyl methacrylate in supercritical C02 73 Chapter 4 .Catalytic Chain Transfer Polymerization of Methyl Methacrylate in Supercritical Carbon Dioxide: Further evidence for a Diffusion Controlled Transfer Process 4.1 Introduction The economic controlled synthesis of low molecular weight polymers is assuming greater importance as applications in the coatings, detergents and water treatment industries continue to develop. There is a driving force to decrease molecular weight in oligomeric additives to detergents, dispersants and scale inhibitors as biodegradability is facilitated as molecular weight decreases. One standard approach to molecular weight reduction involves the use of chain transfer agents such as mercaptans. This strategy has two major disadvantages, viz, the mercaptan is incorporated in the small chains adversely affecting the oligomer properties and also the malodorous nature of the mercaptans is often a problem. Another approach is to use high initiator concentrations, which is often uneconomic and therefore impractical. Molecular weight reduction can also be effected in semi continuous solution polymerization, which can result in high organic solvent emissions and problems in isolation of the oligomeric product. In many practical instances a combination of these approaches is taken. Chapter4 74 Two technologies have emerged in recent years that may lead to a step-change in green oligomer synthesis. Firstly, catalytic chain transfer (CCT) polymerization is now well known as a highly efficient method for producing oligomersJ,2 CCT has many favourable characteristics; the catalysts are used in low concentrations and are generally non-toxic, and the free-radical nature of the reaction means that many conventional mO'nomers can be polymerized without special plant requirements. The other new technological area, i.e., polymerization in supercritical fluids (SCF) such as carbon dioxide, has two primary advantages for oligomer synthesis - the elimination of organic solvent emissions and facile oligomer isolation. The. limited solubility of common polymers in supercritical C02 is also overcome to some extent by targeting a low molecular weight product. Dada et al. (Rohm & Haas )3 have patented a high temperature process for oligomer synthesis in supercritical C02, which demonstrates the potential commercial utility of a successful CCT/SCF approach. This chapter describes initial experiments performed to test the potential of CCT in supercritical carbon dioxide, for the synthesis of acrylic oligomers. The effect of the low viscosity medium on CCTP will be evaluated by determining the transfer constant for COPhBF in supercritical MMNC02• COPhBF Figure 4.1: Structure of the catalytic chain transfer agent used in supercritical fluid studies CCTP of methyl methacrylate in supercritical COz 75 4.2 Experimental 4.2.1 Materials The bis(methanol) complex of COPhBF were prepared according to the method <' described in section 3.2.1.1, replacing dimethyl glyoxime in the described procedure for diphenyl glyoxime. Methyl methacrylate monomer was passed through a column of activated basic alumina (ACROS, 50-200 micron) and was purged with high purity nitrogen (BOC) for 1.5 hours prior to use. AIBN (Dupont) was recrystallized twice from methanol and used as initiator. toluene, used without further purification, was purged with high purity nitrogen (BOC) for 1.5 hours prior to use. The COz (99.5%) was supplied by CIG. 4.2.2 General polymerization procedure. Polymerizations were carried out as described in the previous chapter. Two stock solutions were prepared: (i) and initiator stock solution and a (ii) catalyst stock solution. (i) The initiator stock solution was prepared by dissolution of approximately 200 mg of AIBN into 50 ml of monomer. (ii) The catalyst stock solution was prepared by dissolution of approximately 2 - 3 mg of catalyst into 10 ml of solution (i). Four reaction mixtures were prepared by the addition of between 2 and 3ml of solution (ii) to 24ml of solution (i). For the supercritical reactions a Jerguson sight gauge (50ml), equipped with a recirculating pump (Figure 4.2), was purged with C02 before reaction mixture was introduced. The reaction mixture (7ml) was introduced via a glass syringe to the reactor. Carbon dioxide was allowed to flow into the reactor slowly::::: lOml·min-1 to ensure a homogeneous reaction mixture is attained at the super critical pressure. The recirculating pump was turn off as the pressure got close to the super critical point to stop the monomer rich liquid phase spraying into the dense gas phase. After the super critical point is passed the recirculating pump is turned back on for continuous mixing. The experiments were performed at 150 bar and the temperature was maintained at 50°C as the entire sight gauge was immersed in an isothermal water bath. The control Chapter4 76 reaction mixture (7ml) was diluted with toluene (43ml) and charged to a Schlenck vessel which was then immersed in a 50°C water bath. Conversions were maintained below 5%. IS CO Pump l :reactor i ca.lbon dioxide ~···································'f······························· supply water bath --v<~--,..~.., sample collection Figure 4.2: Apparatus for polymerization of MMA in super critical carbon dioxide 4.2.3 Phase behaviour. Reactions in supercritical media are conducted at high pressure and under these conditions the phase behaviour of the reaction mixture can be complex. Subramaniam and McHugh4 have noted that many studies fail to identify the phase behaviour of the reaction mixture. The number and nature of extant phases has a direct impact on the progress of the reaction and so knowledge of the phase behaviour is essential for the proper interpretation of experimental data. Prior to running the reaction experiments, the phase behaviour of the C02/methyl methacrylate binary mixture was examined. This involved the determination of the transition pressure of the mixture at 50°C. At pressures in excess of the transition pressure the mixture exists as a single, homogeneous phase which exhibits gas-like transport properties due to the presence of supercritical C02.5,6 Various quantities of CCTP of methyl methacrylate in supercritical C02 77 methyl methacrylate (5- 10 ml) were added to the Jerguson sight gauge. After purging the system with C02, the pressure was increased slowly with COz so as to observe the behaviour of the meniscus, visible along the length of the sight gauge. The addition of COz was accompanied by a significant expansion of the liquid phase, especially once the critical pressure of C02 had been exceeded. At pressures in excess of 95 bar, the meniscus was no longer visible thus indicating the formation of a homogeneous mixture. It is noteworthy, that when approximately 7 ml of methyl methacrylate was used, the position of the meniscus became stable towards the middle of the sight gauge. As the pressure was increased to around 95 bar the meniscus became cloudy and eventually disappeared as the pressure was increased further. Such behaviour typically indicates that the mixture critical point has been reached for a given mixture composition.? ,8 It was therefore concluded that operating at a pressure of 150 bar at 50°C was sufficient to achieve a homogeneous reaction mixture. 4.2.4 Molecular weight analysis. Molecular weight distributions were determined by size exclusion chromatography using a GBC Instruments LC1120 HPLC pump, a Shimadzu SIL-10A Autoinjector, a column set consisting of a Polymer Laboratories (PL) 3.0 J.!m bead-size guard column 6 5 4 3 (50 x 7.5 mm) followed by four linear PL columns(10 , 10 , 10 and 10 ) (300 x 7.5 mm) and a VISKOTEK Duel Detector Model 250 differential refractive index detector. Tetrahydrofuran (BDH, HPLC grade) was used as the eluent at lrnl/min. Calibration of the SEC equipment was carried out with narrow poly(methyl methacrylate) standards 6 (Polymer Laboratories, molecular weight range: 200- 1.6xl0 ). Chapter4 78 4.2.5 Matrix-assisted laser desorption ionisation (MALDI) mass spectrometry. Mass spectrometry analysis was carried out on a Biosystems, Voyager DE™RP instrument operated in reflectron mode. The mass scale was calibrated using substance F. The matrix used was 2,5-dihydroxybenzoic acid (DHB) with NaCl as the cation source, and was dissolved in a 50:50 mixture of water and methanol. The polymer analysed was dissolved in THF at a concentration of 5mg/ml. The matrix solution was deposited onto a gold plated target and allowed to dry. The polymer was then deposited onto the matrix. 4.3 Results and Discussion 4.3.1 Catalytic chain transfer of COPhBF in super critical MMA/carbon dioxide The preliminary experiments indicated that low molecular weight polymer could be synthesised and the experiment was subsequently optimised to decrease the time taken for pressurisation and thermal equilibration. The time taken for depressurisation was also found to be important. In cases where a significant time interval passed in these initial and terminal stages then high molecular weight material could be detected in the subsequent SEC analysis (see Figure 4.3). CCTP of methyl methacrylate in supercritical C02 79 0.6-r------, Supercritical C0 I 2 0.4 Toluene M Figure 4.3: Molecular weight distributions obtained in supercritical C02 and in toluene solution. In order to quantify the efficacy of the catalytic chain transfer reaction, we endeavoured to measure the chain transfer constant of COPhBF in the MMA/C02 solution. The results were compared with control reactions performed either at the same concentration of MMA and [Co]/[MMA] ratio in toluene or in bulk MMA at ambient pressure, at 50°C. The molecular weight distributions obtained in the different media are significantly different, as shown in Figures (4.3) and (4.4), where it is clear that extremely low molecular weights are generated in the supercritical C02• It is evident from Figures (4.3)- (4.4) that the CCT polymerization performed under supercritical conditions has a substantially higher apparent chain transfer constant. Figure 4.5 shows the Mayo plots, which were generated using Mw/(2·m0), for the reactions conducted in supercritical C02 and those performed by bulk and solution polymerization at 50°C. The actual Cs values obtained are given in Table 4.1. Chapter4 80 1.5-.------., 1.0 M Figure 4.4 Molecular weight distributions obtained in supercritical C02 (solid line) and the control experiments run contemporaneously in bulk MMA (dotted line). 0.20....------., • 0.15 0 0 0.10 ~ 0 • 0.05 • ¢ • 0.00111~~~---.-----.------.----.----~-----1 o.o 5.0x10-7 l.Oxlo-6 1.5:xlo-6 [COPhBF]/[MMA] Figure 4.5 Mayo (Mw/2)(•) and CLD (0) plots for MMA in supercritical C02 and Mayo (A) and CLD (T) plots for bulk MMA. Also shown is the Mayo plot (*) for MMA solution (66% toluene) CCTP of methyl methacrylate in supercritical C02 81 The Mayo plots were generated assuming DPn = Mw/(2·m0). This assumption is generally applicable for a chain transfer dominated system, in which case Mw is equal to 2 x Mn (except for very low molecular weights)_9,10 In this work, using Mw resulted in improved linearity of the Mayo plots, however, as such low molecular weights were g~nerated, and hence Mw :5 2 x Mn, some error is inevitable and the Cs values should only be treated as indicative measures of the increased efficiency of transfer in the supercritical medium. Table 4.1 Cs values obtained for CCT with MMA in three different media, using the CLD and Mayo methods. (The variation in the Cs value for MMA in supercritical C02 reflects the difficulty in accurately characterising the molecular weight distribution at very low chain lengths- as discussed in the main text). Cs Medium Mayo CLD BulkMMA 20K 21K Toluene 23K 26K Supercritical C02 378K 113K As monomer can be generated as part of the CCT reaction, this can lead to an under estimate of Cs as this is excluded from the evaluation of the molecular weight moments, particularly when very low molecular weights are generated. In this regard the chain length distribution (CLD) method may be preferable for evaluating the true Cs value. The lnP(M) plots for polymerizations run in supercritical C02 and bulk MMA are shown in Figure (4.6)- exemplifying the enormous difference in chain transfer activity in the two media. The Cs values derived from the CLD analyses are also reported in Table 4.1. Chapter4 82 l.Oxlcf 20xlcf 3.0xlcf 4.0xlcf M Figure 4.6: In P(M) distributions obtained in supercritical C02 (solid line) and the control experiments run contemporaneously in bulk MMA (dotted line). 4.3.2 Mechanistic Interpretation. The mechanism of CCT has never been unambiguously proven. The generally accepted mechanism is shown in Scheme 4.1, where a transient cobalt hydride intermediate is generated. Characteristically the final products formed during CCTP are the same as the final products of the transfer to monomer reaction. The chain transfer reaction is remarkably fast (compared with conventional chain transfer reactions), and strong evidence was given in the previous chapter that catalytic chain transfer is diffusion controlled in the case of methacrylates. CCTP of methyl methacrylate in supercritical C02 83 R: +Co(II) Pn +Co(ID)-H (4.1) Co(ID)- H + M ~ Co(ll) + R ~ (4.2) Scheme 4.1: Mechanism for the catalytic chain transfer reaction via a Co(III)-H intermediate. A MALDI spectrum showing the molecular weight distribution of PMMA generated in a CCT reaction in supercritical C02 is shown in Figure (4.7). Each peak mass (m) is consistent with the expression m = (100.1 x n) + 23.0; where n is the degree of polymerization, 100.1 is the mass of the repeat unit and 23.0 is the mass of the sodium counterion. There are no chains evident with an AIBN initiator fragment (mass = 68), which would have a mass equal to m = (100.1 x n) + 68 + 23.0. This result is concordant with a polymerization process dominated by catalytic chain transfer. 1123 1223 2000 1000 1500 til § 8 1000 1100 1200 1300 1000 500 1000 1500 2000 mlz Figure 4.7: MALDI-TOF-MS spectrum of PMMA synthesised using CCT in supercritical C02• Chapter4 84 Thus it is clear that CCT polymerization of MMA with cobaloximes in supercritical C02 is more effective than the equivalent reaction in toluene at ambient pressure. The raises the question as to why this should be? The first possibility to consider is that there is a significant solvent effect on the propagation reaction, as Cs is the ratio of the elemental transfer and propagation rate coefficients. Van Herk and co-workers, 11,12 have shown that the propagation rate coefficient is virtually the same in supercritical C02 as in bulk (there is a minor solvent effect - but this cannot account for the large change in Cs observed in the current work). This leaves the second possibility, a significant solvent effect on the chain transfer reaction,. as the basis of the reduced Cs. The solvent effect can be exerted conceivably via three different mechanisms: 1. Complexation or ligation 2. Partitioning of the catalyst 3. Viscosity of the reaction medium In what follows, each of these possibilities will be discussed in tum. 4.1.2.1 Complexation or Ligation There has been some conjecture in the literature that CCT reactions can be strongly influenced by solvent. In chapter 5, it is shown that the solvent effect on CCT (with MMA and styrene) is not significant for solvents that are weak ligands, such as toluene and butyl acetate. A solvent effect of this nature only becomes important for strong ligands such as pyridine - this has also been observed in previous studies. In the current case neither toluene or C02 are strong ligands and therefore it seems unlikely that the strong effects we observe can be attributed to direct chemical interactions. 4.1.2.2 Partitioning of the Catalyst. An enhanced chain transfer process may be possible if the local cobaloxime concentration around the radical is increased. In this work, we were careful to select COPhBF as the CCT catalyst as we assumed that the relatively hydrophobic phenyl CCTP of methyl methacrylate in supercritical C02 85 rings (together with the BF2 groups) would increase the solubility of the catalyst in supercritical C02• Klingler and Rathke13 studied the hydrogenations of dicobalt octacarbonyl and dimanganese decacarbonyl in supercritical C02 over a wide temperature range and found no significant solvent effects in the supercritical medium in comparison to nonpolar organic solvents. Therefore it seems unlikely that the ~ increased transfer efficiency in supercritical C02 can be attributed to partitioning. 4.1.2.3 Viscosity of the Reaction Medium As mentioned earlier in this chapter, and shown in the previous chapter, the chain transfer process may be diffusion controlled, as indicated by the high chain transfer rate coefficient in the CCT polymerization of MMA in organic solvents. In addition the individual Cs values for the methacrylates could be correlated with monomer viscosity, i.e., the higher the viscosity, the lower Cs. These findings are consistent with theories for a diffusion-controlled reaction, in which the rate coefficient is generally found to be proportional to the self-diffusion coefficients of the two reactants: (4.3) Here Dco is the diffusion coefficient of the cobalt catalyst and Di the diffusion coefficient of an i-meric radical. Since diffusion coefficients are generally found to be roughly proportional to the inverse of the solvent viscosity TJ (D oc TJ-a, where a lies generally between 0.5 and 1.0), the following relationship should approximately be valid for the chain transfer constant: 1 a (4.4) TJ It is well known that supercritical fluids are characterized by gas-like viscosities, which in general have a viscosity an order of magnitude smaller than that for liquids. Therefore we would predict an increase in the -rate of transfer relative to propagation (which is Chapter4 86 chemically controlled) in the supercritical medium if the chain transfer reaction is indeed diffusion controlled. This is in qualitative agreement with our experimental observations. 4.3.3 The Effect of Non-Homogeneity Figure 4.8 shows the w(log M) vs M plot of a PMMA sample formed when the supercritical fluid reaction mixture was not homogenous during the entire reaction. During the initial charging of the reactor, for example, if the reactor is allowed to pressurise rapidly, the MMA will not swell with C02 enough before the super critical pressure was reached. As a result of this the reaction mixture formed a suspension, ie the reaction mixture goes cloudy, with monomer rich phase within the super critical COz. It is doubtful that the reaction mixture within the monomer rich droplets "go supercritical", thus higher molecular weight polymer is formed during this time, than would form in a homogeneous supercritical fluid (Figure 4.8). As was mentioned earlier in this chapter, PMMA has a very low solubility in supercritical C02, thus at a critical mixture concentration ratio the high molecular weight polymer, formed during charging of the reactor, begins to precipitate forming a suspension once again. This time the monomer, which has a higher affinity for the polymer than for the supercritical C02, enters the particles formed by the polymer precipitate. Very high molecular weight polymer is formed within the particles (Figure 4.8), formed by the polymer precipitate, due the solubility of the cobalt catalyst used (COPhBF), being very high in supercritical COz, thus there is no mass transport of COPhBF from the supercritical C02 to the particles. CCTP of methyl methacrylate in supercritical C02 87 0.5~------. Polyrrer fonred in the particles forrred by the precipitate 0.4 0.3 Polymer formed during the charging of the reactor l 0.1 0.0 M Figure 4.8: w(log M) vs M plot of a PMMA sample formed when the supercritical fluid reaction mixture was not homogenous during the entire reaction 4.4 Conclusion The value of the transfer constant of COPhBF in supercritical C02 was found to be enhanced by an order of magnitude compared to that expected for bulk or solution polymerization, and it is unlikely that this factor can be attributed to partitioning or direct chemical interactions. A result of, which provides further evidence that the rate determining step of catalytic chain transfer polymerization of MMA with COPhBF, is diffusion controlled. Catalytic chain transfer in supercritical C02 thus represents an extremely efficient method for synthesising oligomers using free-radical polymerization. Chapter4 88 4.5 References 1. Davis, T. P.; Kukulj, D.; Haddleton, D. M.; Maloney, D. R. , Trends Polym. Sci. , 3, 365.(1995) 2. Karmilova, L. V.; Ponomarev, G. V.; Smimov, B. R.; Belgovskii, I. M. , Russ. Chern. Rev. , 53, 132.(1984) 3. Dada, E. A.; Lau, W.; Merritt, R. F.; Paik, Y. H.; Swift, G. SCF oligomer patent; Dada, E. A.; Lau, W.; Merritt, R. F.; Paik, Y. H.; Swift, G., Ed.; Rohm & Haas, (1994.) 4. Subramaniam, B.; McHugh, M. A. , Ind. Eng. Chern. Process Des. Dev. , 25, 1.(1986) 5. Dehghani, F.; Wells, T.; Coton, N. J.; Foster, N. R. , J. Supercrit. Fluids , 9, 263.(1996) 6. Page, S. H.; Sumpter, S. R.; Coates, S. R.; Lee, M. L.; Dixon, D. J.; Johnston, K. P. , J. Supercrit. Fluids, 6, 95.(1993) 7. Hicks, C. P.; Young, C. L., Chern. Rev. , 75, 119.(1975) 8. Li, L.; Kiran, E. , J. Chern. Eng. Data , 33, 342.(1988) 9. Stickler, M.; Meyerhoff, G., Makromol. Chern. , 179, 2729.(1978) 10. Kapfenstein, H. M.; Heuts, J. P. A.; Davis, T. P. , Macromol. Chern. Phys. , In preparation (1998) 11. Quadir, M.A.; DeSimone, J. M.; Van Herk, A.M.; German, A. L., Macromolecules '31, 6481.(1998) 12. Herk, A. M. v.; Manders, B. G.; Canelas, D. A.; Quadir, M. A.; DeSimone, J. M. , Macromolecules, 30, 4780.(1997) 13. Klingler, R. J.; Rathke, J. W. , J. Am. Chern. Soc. , 116, 4772.(1994) Solvent effects on CCTP 89 Chapter 5 Solvent Effects on Catalytic Chain Transfer Solvents are used extensively in commercial polymerization processes. Solution polymerization can be used to produce a wide range of molecular weights, but more applicable is the production of low molecular weight oligomers in the paints and coatings industry. Catalytic chain transfer, as described in previous chapters, is a highly effective free-radical method of producing low molecular weight oligomers, as only ppm quantities of catalyst are required for a significant molecular weight reduction. This chapter describes the effect of solvents on the catalytic chain transfer reaction. The first section of this chapter, studies the role that monomer plays in the chain transfer reaction of cobalt mediated free-radical polymerization, in an attempt to distinguish between the different mechanisms of catalytic chain transfer described in section 2.2.2. The effects of two types of solvents on CCTP are studied: Weak co-ordinating solvents and coordinating solvents. The effects of the two classes of solvents are expected to be different. Coordinating solvents, as described in section 2.2.6.2, could influence the chain transfer process by complexing with catalyst in the base ligand position, by changing the nature of the cobalt centerl. Weak co-ordinating solvents, on the other hand, should have the same effect on the catalyst, as the monomer, hence no significant change in the chemical nature of the catalyst is expected. These mechanisms will be described in more detail in the appropriate sections. The effects of solvents on the catalytic chain transfer process will be examined in the presence of two cobaloximes with different oxime ligands. The catalysts used, shown in Figure 5.1, are cobaloxime boron fluoride (COBF), and tetra-phenyl cobaloxime boron fluoride (COPhBF). ChapterS 90 cobaloxime boron fluoride tetra-phenyl cobaloxime boron fluoride (COBF) (COPhBF) Figure 5.1: Structures of the catalytic chain transfer agents used in this chapter. 5.1 The role of monomer in the chain transfer reaction in cobaloxime-mediated free-radical polymerization Although it has long been established that a hydrogen atom is abstracted from a 13 position (preferably from an a-methyl group) in the growing radical and transferred to a monomer molecule to give a monomeric free-radical, no conclusive evidence regarding the actual chain transfer reaction exists. The most widely accepted view seems to be that the mechanism is a two-step process2-9 as shown in scheme 2.2. Incorporation of this mechanism into the overall kinetic scheme of free-radical polymerization leads to the following form of the Mayo equationlO, which expresses the 1 reciprocal number average degree of polymerization, DPn- , as a function of the individual kinetic parameters: 1 r [Co(II)] + eM + '-'S (5.1) D~ [M] Solvent effects on CCTP 91 In this expression, 'A is the fraction of termination by disproportionation, (kt) is the average termination rate coefficient, [R •1 the overall radical concentration, kp the propagation rate coefficient, CM the chain transfer constant to monomer and Cs the chain transfer constant to the Co(II) complex, which is defined is kt.rfkp. The efficiency of the transfer agent is expressed in Cs. Although the predictive value of this mechanism, and hence the Mayo equation expressed as equation (5.1), is very high, it has one major shortcoming. According to equation (5.1), where 1/DPn is a function of the ratio of the catalyst concentration to the monomer concentration, the molecular weight should continuously decrease with increasing conversion, but experiments clearly indicate that the molecular weight distribution remains virtually unchanged during the course of polymerization8,11. A possible explanation is poisoning of the catalyst and although it is known that this is indeed operative during the polymerization12, the possibility that this would exactly compensate for the decrease in monomer concentration, seems unlikely. An alternative explanation would be the direct participation of monomer in the hydrogen abstraction process (i.e., the equivalent of the normal chain transfer to monomer process), in which the chain transfer rate would be first order in monomer concentration. The monomer concentration would then also appear in the numerator of the last term in equation (5.1) and therefore cancel from this expression, leaving 1/DPn a function of the catalyst concentration only. This section discusses the kinetics that would be observed in case of a mechanism, which involves the direct participation of monomer, and this was subsequently tested experimentally. ChapterS 92 5.1.1 Experimental procedures 5.1.1.1 Materials The bis(methanol) complex of COPhBF were prepared according to the method ~ described by Bakac et al13. and presented in section 3.2.1.1, replacing dimethyl glyoxime in the described procedure by diphenyl glyoxime. Methyl methacrylate (Aldrich, 99%) was passed through a column of activated basic alumina (ACROS, 50- 200 micron) and purged with high purity nitrogen (BOC) for 1.5 hours prior to use. AIBN (DuPont) was recrystallized twice from methanol and used as initiator. Tert-butyl acetate (Merck), distilled under reduced pressure, and toluene (Ajax Chemicals, Analytical Reagent), used without further purification, were used as solvents and purged with high purity nitrogen (BOC) for 1.5 hours prior to use. 5.1.1.2 General polymerization procedure Sample preparations and polymerizations were carried out similar to the chain transfer constant measurements described previously using standard Schlenk and syringe techniques to keep oxygen out of the solutionslO). Firstly, 250 mg of AIBN was dissolved in 70 ml of monomer (solution 1), and 107 mg of AIDN in 30 ml of toluene/t butyl acetate (solution II). Subsequently, 3-4 mg of COPhBF was dissolved in 10 ml of solution I. From this solution, 3 ml were further diluted with 27 ml of solution I (solution ill). Reaction ampoules, specially adapted for use with Schlenk equipment were then charged with the solutions according to Table 5.1, and subsequently degassed by two freeze-pump-thaw cycles. Polymerizations were carried out in a water bath thermostated at 60 ± 0.5°C ensuring conversions to remain below 10%. Solvent effects on CCTP 93 Table 5.1: Composition of the runs performed. Run Solution (I) Solution (II) Solution (Ill) /ml /ml /ml A1 4 0 1 A2 3 1 1 A3 2 2 1 A4 1 3 1 A5 0 4 1 B1 4 0 3 B2 3 1 3 B3 2 2 3 B4 1 3 3 B5 0 4 3 5.1.1.3 Molecular weight analysis Molecular weight distributions were determined by size exclusion chromatography using a GBC Instruments LC1120 HPLC pump, a Shimadzu SIL-10A Autoinjector, a column set consisting of a Polymer Laboratories 3.0 J.lm bead-size guard column (50 x 6 5 4 3 7.5 mm) followed by four linear PL columns (10 , 10 , 10 and 10 ) and a VISCOTEK Dual Detector Model 250 differential refractive index detector. Tetrahydrofuran (BDH, HPLC grade) was used as eluent at 1 ml/min. Calibration of the SEC equipment was performed with narrow poly(methyl methacrylate) standards (Polymer Laboratories, 6 molecular weight range: 200- 1.6x10 ). ChapterS 94 5.1.2 Results and Discussion 5.1.2.1 Derivation of kinetic expressions As stated in the introduction, the expression given by equation (5.1) cannot explain an unchanging molecular weight distribution with changing conversion. It was mentioned that catalyst poisoning at a rate similar to that of monomer conversion is not very likely, and that a direct participation of monomer in the chain transfer reaction is a possible alternative explanation. In the latter case, the last term in equation (5.1) should be replaced by a single chain transfer constant ( CM,catalyzed) for the catalyzed chain transfer to monomer process: 1 + eM + CM,catalyzed (5.2) D~ Let us now consider the possibilities that would give rise to such form of the Mayo equation. Firstly yve must realize that the Mayo equation involves solely the rate of dead polymer formation by the chain transfer reaction, and not on any subsequent re-initiation step. Thus the monomer needs to be directly involved in a reaction with the polymeric radical. Two possible mechanisms, where the monomer is directly involved in the chain stopping event, are given in scheme 2.3. The following expression, for both mechanisms, is then easily derived for the rate of chain transfer (Rtr,i): Rtr,i = ~.iKJCo(II)][R"][M] (5.3) where i = 1 or 2 and Ki is the equilibrium constant of the association equilibria of equations (2.29) and (2.31). Substitution of equation (5.3) into the Mayo equation given by equation (5.1) would then yield the following expression for CM,catalyzed: Solvent effects on CCTP 95 C = ~ i Ki [Co(II)] (5.4) M, catalyzed ~ Comparison of equations (5.1) and (5.2) yields an expression for the thus far measured chain transfer constant C8: (5.5) This expression suggests that the value of Cs is proportional to the monomer concentration, which is consistent with the observation reported by Suddaby et al.14 that the chain transfer constant (Cs) of cobaloxime boron fluoride in methyl methacrylate polymerization was reduced by approximately 50% when diluted with 50% toluene. Assuming that monomer indeed directly participates in the chain transfer reaction, it is now interesting to examine what we expect to observe in a conventional catalytic chain transfer constant measurement. Commonly, the average degrees of polymerization are measured for several different [Co(II)]/[M] ratios in which [Co(II)] is varied but [M] kept constant. From equation (5.4) it is clear that CM,cataiyzed increases with increasing [Co(II)], and hence DPn decreases. Since [M] is constant, a plot of DPn-l vs [Co(II)]/[M], will then indeed give a straight line with a slope equal to Cs which is given by equation (5.5). Repeating the same experiment with a lower, but constant, monomer concentration will also lead to a straight line, but with a smaller slope, again given by equation (5.5). To summarize this argument, our expected results are completely consistent with conventional wisdom. However, if we were to carry out the experiment by changing the [Co(II)]/[M] ratio through a varying [M] and constant [Co(II)], the situation becomes different. According to equation (5.4), CM,catalyzed will be constant in the range of experiments, because this parameter only depends on [Co(II)], which is held constant, and hence a constant DPn is expected. Since [M] is not constant, the slope of this plot(= 0) is not equal to Cs (being a function of [M]). ChapterS 96 In order to test this prediction, we need to measure the average degree of polymerization for different ratios of catalyst and monomer concentrations, keeping the former constant and varying the latter. This can, in principle, be accomplished by diluting the monomer solution to varying extents. However, significant solvent effects can occur in the current system, as a solvent molecule may act as the axial ligand for the Co(ll) complex and ma"y significantly alter the electronic properties of the catalystS. The latter effect may lead to a different reactivity of the catalyst and hence the experiment may not be suitable for testing our predictions. In the current study, toluene and tert-butyl acetate as diluents were chosen as they are known to be weak ligands, and we do not expect them to significantly alter the properties of the catalyst. Figures 5.2 and 5.3 show the Mayo plots obtained for the runs conducted in the presence of toluene and t-butyl acetate respectively. All plots clearly have a positive slope similar to what is observed in a conventional experiment. This result, which is in contrast with what is expected from equation (5.2), could be caused by one of two possible cases: (i) there are unusual solvent effects or (ii) the rate of chain transfer is not first order in monomer concentration. Since the results obtained in both solvents are very similar, the latter explanation is the most conceivable one. The results are consistent with what is predicted from equation (5.1), and therefore we may assume the slopes of these lines are indeed equal to C8. In Table 5.2, a summary of the determined Cs values is given. It can be seen that within experimental error the Cs values do not depend on the used COPhBF concentration nor the solvent, as expected from equation (5.2), and are of similar magnitude to those obtained for a methyl methacrylate/toluene (1:1 v/v) system 3 (Cs = 24 x 10 ) as seen in the next section of this chapter, using the conventional measurement (i.e., varying [Co(II)] and constant [M]). Solvent effects on CCTP 97 0~~------. 0.15- ~ 0.10- Q ...... 0.05 Figure 5.2: Mayo plots obtained for the experiments conducted in toluene: (•) Run A; (D) Run B; (e) Run A (repeat); (0) Run B (repeat) 0~ D 0 0.15- 0 • A-.= 0.10- D 0 Q ...... • 0 0 • D 0.05- • e 'iii • • 0.00 0.0 2.0x10-6 4.0x10-6 6.0x10-6 s.Oxio-6 [COPhBF]/[MMA] Figure 5.3: Mayo plots obtained for the experiments conducted in t-butyl acetate: (•) Run A; (D) Run B; (e) Run A (repeat); (0) Run B (repeat) ChapterS 98 Table 5.2 Transfer constant (Cs) of COPhBF, in the solution polymerization of MMA in toluene and tert-butyl acetate, determined by varying the monomer concentration with a constant catalyst concentration. Toluene Tert-Butyl Acetate Run [COPhBF]/M Mna Mwb Mna Mwb A 1.1·10-5 25.4·103 21.4·103 22.4·103 19.2·103 A (repeat) 1.1·10-5 24.0·103 23.7·103 23.9·103 30.5·103 B 2.4·10-5 25.4·103 29.8·103 22.6·103 24.1·103 B (repeat) 2.4·10-5 25.9·103 24.0·103 27.4·103 28.6·103 a) Cs measured using DPn =Mn/100.12; b) Cs measured using DPn = Mw/(2·100.12) 5.1.3 Summary ofthe role of monomer on catalytic chain transfer In this work the possibility of a direct participation of the monomer in the hydrogen abstraction step in catalytic chain transfer polymerization was tested. The kinetic expressions corresponding to a mechanism in which there is such a direct participation indicate a constant degree of polymerization in situations with a constant COPhBF concentration but a varying monomer concentration. However, this constant degree of polymerization was not found in the current studies. The obtained results are consistent with the commonly accepted two-step mechanism involving a Co(ill)H intermediate, which still leaves us without a satisfactory explanation for the observation that a constant molecular weight distribution is produced throughout the entire course of polymerization. Solvent effects on CCTP 99 5.2 The effect of weak co-ordinating solvent on the chain transfer reaction in cobaloxime-mediated free-radical polymerization Weak co-ordinating solvents such as toluene and acetates are used quite often in commercial applications. Also, in studying the mechanism of CCT, dilution of the ' reaction solution with solvent is often needed; any effect of the solvent could hinder this work. Thus the effect that these solvents have on the kinetics of free-radical catalytic chain transfer polymerization is of great importance. There have been three studies on the effect of weak co-ordinating solvents to date, and there seems to be some contention as to the effect of these solvents on CCTP. In the previous section of this chapter, on the effect of monomer concentration on CCTP, it was observed that there seemed to be no change in the Cs of the tetra-phenyl derivative of COBF, over a varying toluene concentration. In another study by Gridnev 7, it was stated that additives such as acetone, ethyl acetate, benzene and toluene (up to 40% v/v) did not cause changes in CCTP rate and kinetics. This was contradicted in a study by Suddaby et al14, based on a limited set of data, he showed there was a 3 3 decrease in the transfer constant, of cobaloxime boron fluoride, from 36x10 to 25x10 , on addition of toluene to the catalytic chain transfer polymerization of methyl methacrylate at a ratio of 2: 1. This section investigates the effect of toluene and tert-butyl acetate on the catalytic chain transfer reaction involving COBF and COPhBF in methyl methacrylate. Two mass ranges of catalyst were used in this study to provide a broad mass range for analysis. ChapterS 100 5.2.1 Experimental 5.2.1.2 General polymerization procedure Polymerizations were carried out as described in previous chapters. Two stock solutions were prepared: (i) and initiator stock solution and a (ii) catalyst stock solution. (i) The initiator stock solution was prepared by dissolution of approximately 200 mg of AIBN into 70 ml of monomer. (ii) The catalyst stock solution was prepared by dissolution of approximately 2.5 mg of catalyst into. 10 ml of solution (i) and a subsequent 10-fold dilution with solution (i). For the high catalyst concentration experiments no dilution of solution (ii) is needed. Five reaction mixtures were then prepared containing 0.1, 0.2, 0.3, 0.4 and 0.5 ml of solution (ii) and made up to 4.5 ml total volume with solution (i). The reaction mixtures were then diluted to the required dilution ratio with solvent. The reaction ampoules were then further degassed by two freeze pump thaw cycles and placed in a thermostated water bath at 60°C. Final conversions were maintained below 5%. Polymer was collected by evaporating the unreacted monomer to avoid fractionation on precipitation. 5.2.2 Results and Discussion Figure 5.4 shows the Mayo plots for COBF at the high and low catalyst concentration ranges, in the solution polymerizations of MMA in toluene and t-butyl acetate, compared to the Mayo plot of COBF in the bulk polymerization of MMA. The Cs values for these experiments are tabulated in Table 5.3. The molecular weights of the polymer produced for the low catalyst concentration set ranged from 6K to 30K, where as the high catalyst concentration range produced polymer of molecular weight, 800 to 4.5K. This provided a broad range of analysis, to determine if there is any catalyst concentration effect or molecular weight dependency, on the transfer constant, in the presence of the weak co-ordinating solvents. Solvent effects on CCTP 101 It can be seen from the Mayo plots of COBF that, to within error, there is no effect of the added solvent. These results are in contrast to the results obtained by Suddaby et al14, but are in accordance with the observations made by Gridnev7. These results are also consistent with the findings in the previous section of this chapter where it was shown that the monomer concentration does not play a role in the transfer step. 0.20 ~ [J <)0 [J • 0.15 [J • •oo (:J..r::. -.:>0 e o.w 2.0x10"7 4.0x10·7 - 0.0 2.0xlo-6 4.0xlo-6 6.0xlo-6 [COBF]/[MMA] Figure 5.4: Mayo plots obtained for COBF in MMA conducted in: (•)Bulk(high); (D)Bulk(low); (O)Toluene(high); (e)Toluene(low); ( ¢ )t-butyl acetate. Insert shows a blow up of low concentration experiments. Table 5.3 Transfer constant (Cs) of COBF, in the solution polymerization of MMA in toluene and tert-butyl acetate, Cs measured usingDPn =Mw/(2·100.12) COBF Run 1 Run2 3 3 Bulk (high) 35 X 10 36 X 10 3 3 Bulk (low) 33 X 10 34 X 10 3 3 toluene (high) 33 X 10 29 X 10 3 3 toluene (low) 33 X 10 34 X 10 3 3 Tert-butyl acetate (high) 34 X 10 33 X 10 Chapter5 102 Figure 5.5 shows the Mayo plots for COPhBF at the high and low catalyst concentration ranges, in the solution polymerizations of MMA in toluene and t-butyl acetate, compared to the Mayo plot of COPhBF in the bulk polymerization of MMA. The Cs values for these experiments are given in Table 5.4. It can be seen, in Figure 5.5, that the Mayo plots for the experiments conducted in bulk and the experiments conducted in solution also overlay each other, thus there is no effect of the added weak co-ordinating solvents on the transfer reaction of COPhBF with MMA. There is good agreement of the Cs values, given in Table 5.4, for the experiments conducted in solution, with the Cs values for COPhBF obtained in the previous section in this chapter, Table 5.2. These values are slightly higher than the ones obtained in the bulk polymerization of MMA. It is not within the scope of this work to explain these results, though one possible explanation for this is that the solvent may allow increased mobility of COPhBF in solution, than in the experiments conducted in bulk. It can also be seen that the transfer constant of both catalysts is independent of catalyst concentration in bulk, which was expected and pointed out by Haddleton5. The transfer constant is also ~ndependent of catalyst concentration in the solution polymerizations of MMA in toluene and t-butylactetate. This result would not be expected if the solvent had an effect on the kinetics of CCT. Thus providing further evidence that non coordinating solvents do not play a significant role in the kinetics of CCT. Solvent effects on CCTP 103 0.16 • 0 • D oo 0 0.12 ~0 0 • ~ 0 (:J.;= • 0.08 7 7 • e 2.0x10' 4.0x10' - 0 • • 0.04 • .~ • 0.00 0 1xl0-6 2x10-6 3x1o-6 4x10-6 5x10-6 [COPhBF]/[MMA] Figure 5.5: Mayo plots obtained for COBF in MMA conducted in: (•)Bulk(high); (D)Bulk(low); ( 0 )Toluene(high); ( • )Toluene(low); ( + )t-butyl acetate; ( ¢ )t-butyl acetate • Insert shows a blow up of low concentration experiments. Table 5.4 Transfer constant (Cs) of COPhBF, in the solution polymerization of MMA in toluene and tert-butyl acetate, Cs measured usingDPn =Mw/(2·100.12). COPhBF Run 1 Run2 3 3 Bulk (high) 20 X 10 19 X 10 3 3 Bulk (low) 18 X 10 19 X 10 3 3 toluene (high) 27 X 10 24 X 10 3 3 toluene (low) 24 X 10 23 X 10 3 Tert-butyl acetate (high) 24 X 10 3 Tert-butyl acetate (low) 19 X 10 ChapterS 104 5.2.3 Summary of the effect of weak co-ordinating solvent on the chain transfer reaction in cobaloxime-mediated free-radical polymerization In this work the effect of the weak co-ordinating solvents toluene and t-butyl acetate on ' catalytic chain transfer polymerization was tested. The data presented clearly indicate that the transfer constant for both COBF and COPhBF are not effect by the addition of toluene or t-butyl acetate as solvents, over all mass ranges. Using a higher catalyst concentration the transfer constants are marginally higher probably due to increased transfer at low molecular weights. 5.3 The effect of pyridine on the chain transfer reaction in cobaloxime-mediated free-radical polymerization In the previous section it was shown that weak coordinating solvents had no effect on the molecular weight distributions formed by CCTP. In the presence of strongly coordinating solvents, however, the solvent will complex with the catalyst as the base ligand, and alter the electronic nature of the cobalt species. The effect of strong coordinating ligands on CCTP has only been previously been reported twice that we are aware of. Janowiczl5, who showed an increase in catalytic activity after the addition of pyridine and tri-alkyl phosphines; and Haddleton5 who reported a decrease in activity if the polymerizations were conducted in methanol solvent. There has been no systematic study on the effects strong complexing ligands, such as pyridine, on the kinetics of catalytic chain transfer polymerization. In this section, several aspects of the effect of pyridine on the chain transfer constants of COBF and COPhBF in methyl methacrylate and styrene, and rates of polymerization Solvent effects on CCTP 105 were studied. A study on the effect of pyridine on the UV NIS spectra of the catalysts was performed in an attempt to identify the changing nature of the cobalt catalyst. 5.~.1 Experimental 5.1.2.1 General polymerization procedure Polymerizations were carried out as described in previous chapters. Two stock solutions were prepared: (i) and initiator stock solution and a (ii) catalyst stock solution. (i) The initiator stock solution was prepared by dissolution of approximately 200 mg of AIBN into 70 ml of monomer. (ii) The catalyst stock solution was prepared by dissolution of approximately 2.5 mg of catalyst into 10 ml of solution (i) and a subsequent 10-fold dilution with solution (i). Five reaction mixtures were then prepared containing 0, 0.1, 0.2, 0.3, and 0.4 ml of solution (ii) and made up to 4.4 ml total volume with solution (i). The reaction mixtures were then diluted to the required dilution ratio with pyridine. The reaction ampoules were then further degassed by two freeze pump thaw cycles and placed in a thermostated water bath at 60°C. Final conversions were maintained below 5%. Polymer was collected by evaporating the unreacted monomer, avoiding fractionation on precipitation. 5.1.2.2 UVNIS spectrophotometry The UVNIS spectra were obtained using a CarylE UV-Vis Spectrophotometer. Solutions were prepared by dissolving 3mg of catalyst into 10ml of oxygenated monomer. 0.5ml of pyridine was added to the sample in a quarts sample vial after each successive scan. Chapter 5 106 5.3.2 Results and Discussion 5.3.2.1 Dependence of the transfer constant of COBF and COPhBF on the addition of pyridine. Figure 5.6a shows the transfer constants of COBF, in MMA, as a function of pyridine concentration, and the values are given in Table 5.5. It can be seen in this plot, to within error, that the transfer constant doesn't seem to be affected by the addition of pyridine to the reaction mixture. Though, looking at the Mayo plots for these reactions it is seen that the Mayo plots become curved on the addition of pyridine, as shown in Figure 5.6b. Thus the apparent transfer constant may seem to increase if a linear regression is used on this data. Though the molecular weights produced are not lower, as a higher Cs value would indicate. 4.5xHf 0.025 • • (a) (b) 0.020 / • 4.0xl ~ / • / § 0.010 / I'"'"" / ~ .tv" / 3.0xl<>" / a / / 0.005 / / ...... 25xl Figure 5.6: The effect of a constant concentration of pyridine on the transfer reaction of COBF in MMA, at 60°C. (a) Dependence of the transfer constant versus the pyridine concentration. (b) dependance of the Mayo plot on the addition of pyridine; (e) Mayo plot in the absence of pyridine (.A.) Mayo plot after the addition of 4.4% pyridine Solvent effects on CCTP 107 40000 35000- • • 30000- en u • 25000 20000- • 15000~~.-----~-----~~-----r-,-r-----~,~-----~.-----~~.-----~ 0.0 l.Ox104 2.0x104 3.0x104 4.0xl04 5.0x104 [pyridine]/[COBF] Figure 5. 7: The effect of the ratio of pyridine to catalyst, on the transfer constant of COBF in MMA, at 60°C. 20000.------, 0 15000 en 10000 u • 5000 • • • 0+------~-----r-~------.------r-----.------~--~-----~~• 0.0 3.0x104 6.0x104 9.0x104 1.2x10-3 1.5x10-3 [pyridine] Figure 5.8 The effect of the pyridine on the transfer constant of COPhBF in MMA, at 60°C. Chapter 5 108 Table 5.5: The transfer constants of COBF and COPhBF in MMA and styrene at 60°C, in the presents of pyridine. Cs values determined using DPn = Mw/(2·m0) Monomer Catalyst [pyridine] [pyr]/[mon] Conversion Cs /mol% % MMA COBF 0 36604 1.1 43068 2.2 37454 4.4 42686 0 36403 1.38E+04 32908 2.92E+04· 28023 4.68E+04 18698 COPhBF 0 17494 1.1 6192 1.1 5265 4.4 4773 4.4 2014 8.8 892 Styrene COBF 0 2.8 1761 1.1 2.5 4553 2.2 2.2 3739 8.8 2.2 2194 COPhBF 0 3.2 470 1.1 3.0 720 2.2 3.0 628 8.8 2.9 249 It is possible for the Mayo plot to become curved if the nature of the catalyst (thus the observed transfer constant), or the radical concentration, is changing throughout the range of analysis. It is unlikely that the radical concentration is changing, as dilometric studies, which is reported later in this chapter, of this system showed straight conversion verses time plots. Thus, its seems that by changing the ratio of the catalyst and pyridine, the nature of the catalyst also changes. It can be seen in the Mayo plots in Figure 5.6b that at low COBF concentrations, ie high pyridine to COBF ratio, the molecular weight of the polymer formed is considerably higher than the polymer formed at the same Solvent effects on CCTP 109 COBF concentration, in the absence of pyridine. At the higher COBF concentrations (lower pyridine to COBF ratio) there is no real effect of the pyridine seen on the molecular weight of the polymer formed. Figure 5.7 gives the effect of pyridine on the transfer constant of COBF in MMA when the ratio of pyridine to COBF is kept constant. The transfer constant of COBF, in this case, decreases on an increase in the ratio of pyridine to COBF. Figure 5.8 shows the transfer constant of COPhBF in MMA, as a function of a constant pyridine concentration. It can be seen in this plot that the transfer constant for COPhBF in MMA dropped dramatically at the lowest concentration of pyridine and leveled out on addition of more pyridine. This result is not in accordance with the results obtained by Janowiczl5. Where, Janowicz showed a reduction in the molecular weight of PMMA as the pyridine to catalyst ratio increased. Though, in Janowicz's experiments, the polymerization of MMA was conducted in methanol as the solvent. It has been shown that there is a solvent effect of methanol on the kinetics of CCT5. It was found that the transfer constant of COBF in MMA decreased from 40K to 10K on the addition of methanol. The transfer constants of the cobaloxime catalyst, in Janowicz's experiments in MMA, ranged from 8 to 1000, with the addition of pyridine. Thus the catalyst showed very low activity in methanol, but on the addition of pyridine the activity increased. No conclusions can be drawn, from Janowicz's results, on the current work of this thesis. Figure 5.9a and Figure 5.10 shows the transfer constants of COBF and COPhBF respectively, in styrene, as a function of pyridine concentration, and the values are also given in Table 5.5. The transfer constants for both COBF and COPhBF increased significantly in styrene at low pyridine concentrations then started to drop over the rest of the range. The Mayo plot of COBF in styrene on the addition of pyridine, Figure 5.9b, are still straight over the range of analysis, when pyridine is added, as opposed to those of the obtained in the presence of a constant concentration of pyridine in MMA. It can be ChapterS 110 from these Mayo plots, that there is a genuine increase in the transfer constant. And that over the entire range of catalyst concentrations that the polymer obtained, the molecular weight is lower, at 1.1% pyridine, than the experiment conducted in the absence of pyridine. 5.3:2.2 Mechanistic interpretation The fact that the transfer constant goes up at the lowest concentration of pyridine, and then starts to drop, indicates that there are two different effects of pyridine on the catalyst behaviour. The first effect is obviously a single pyridine, complexing with the catalyst, in the base ligand position. The reason for the increase in the transfer constant is not so clear. It has been shown11,16, and in the following chapter, that the formation of cobalt carbon bonds between the propagating radical and the cobalt species dominates the kinetics of CCT in styrene. The effect of the cobalt-carbon bond formation is that the observed transfer constant is effectively lowered due to the reduction of the active cobalt species. It is possible that the pyridine, in the base position of the catalyst, changes the ratio of bound catalyst to free catalyst. The reason of this change is not clear as it has been shown that the stronger the base ligand, the stronger the cobalt carbon bond 1. The second effect of the pyridine, on the catalyst, causes the observed transfer constant to drop. It is seen that further additions of pyridine to the mixture causes further decrease in the transfer constant. A possible explanation for this decrease is the possibility that the pyridine takes the second position on the catalyst in the trans-position to the first ligand, as seen in scheme 5.1, thus occupying the reaction site. An observant reader will notice that in the following chapter the transfer constant of cobaloximes in styrene is very conversion dependant and that the transfer constant can vary dramatically depending on the conversion it was taken at. The conversions in all the experiments for these experiments are very close together, thus we can rule out the effect of conversion being a contributing factor for the trends observed in this set of experiments. Solvent effects on CCTP 111 500) • (a) 0.0032 (b) 4000 • 0.0028 3IXXJ 0.0024 !:I. Cl) 0.0020 • u 200) • § • 0.0016 1000 0.0012 0 0.0 5.0x1o4 l.Oxi0-3 1.5xl Figure 5.9: The effect of a constant concentration of pyridine on the transfer reaction of COBF in styrene, at 60°C. (a) Dependence of the transfer constant versus the pyridine concentration. (b) dependance of the Mayo plot on the addition pyridine; (e) Mayo plot in the absence of pyridine (.6.) Mayo plot after the addition of 4.4% pyridine. 800 • (ffi- • u 400 u<.ll 200 • 0~----~----~~----~-----r-.----~----~ 0.0 5.0x.HT4 l.Ox.HT3 [Pyridine] Figure 5.10: The effect of a constant concentration of pyridine on the transfer reaction of COPhBF in styrene, at 60°C. ChapterS 112 L L L ..... I ...... •····· ···eo··· ...... ••···· ...... ······eo····· ,. ' - ;::~o~ ,..-I"' L L Scheme 5.1. Tentative mechanism for the ligation of cobalt(II) oximes with a strpng axial ligand, L. The speculative mechanism depicted in Scheme 5.1 shows on addition of a small amount of a strong complexing ligand, such as pyridine, the equilibrium moves quickly to form a single base ligand complex. On further additions of the strong complexing ligand the equilibrium then moves to a bi-base ligand complex, where the base ligands are trans to each other, which is probably not very active as the reactive site on the catalyst is occupied, though there is only circumstantial evidence for this formation. 5.3.2.3 Dependence of the rate of polymerization, on the addition of pyridine, on cobalt mediated free-radical polymerization. Figure 5.1la and Figure 5.1lb shows the final conversions of the polymerization of MMA in the presence of COBF and COPhBF respectively, with varying amounts of pyridine, as a function of the catalyst concentration. It can be seen that in both figures, that on the addition of pyridine the final conversion drops significantly. In the case of COBF it can be seen that the final conversion of MMA drops from 4% to below 1% at the same concentration of catalyst. The drop in conversion of MMA in the presence of COPhBF is less dramatic than in the case of COBF. The final conversion, with COPhBF as the catalyst, drops to approximately half in the presence of pyridine compared to the series conducted in the absents of pyridine. The drop in final conversion in MMA was also observed by Haddleton5 using polar solvents, but did not provide an explanation. Figure 5.12a and Figure 5.12b shows the final conversions of styrene in the presence of COBF and COPhBF respectively, with varying amounts of pyridine, as a function of the catalyst concentration. In both plots it can be seen that, to within error, there is no effect of either the transfer agent or the transfer agent with pyridine on the final conversions obtained. Solvent effects on CCTP 113 It was not certain whether the fall in the final conversion, of the experiments in MMA, could be attributed to a fall in the rate of polymerization, or an extended inhibition period due to the added pyridine. Figure 5.13 shows the results of the rate of polymerization obtained by a dilometric method compared to the results obtained by gravimetry, for the polymerization of MMA in the presence of COPhBF, in bulk and with 1% pyridine. It can be seen in this plot that there is good agreement between the two sets of data. Thus origin of the loss of conversion is indeed a fall in the rate of polymerization, and not an extended inhibition period. There are two possible explanations for the fall in the polymerization rate. The first explanation is the pyridine competing with the monomer for the hydride. Gridnev has pointed out that pyridine can take part in side reactions during CCTP, for example [Pyridine-Co(III)-H] could under go deprotonation to give [Co(I)r and Pyridine-It". Active Co(ll) can be regenerated by the reaction of [Co(I)r and [Co(III)-H] to give 2x[Co(ll)], and 1/2H2 being evolved. The result of this reaction would be a reduction in the radical concentration and therefore a reduction in the rate of polymerization. Though this does not account for the reduction in final conversion as seen by Haddleton5 using methanol as the solvent. The second possibility is an increase in the stability on the hyrididocobaloxime. Schrauzer et al17, found that the stability of the hydridocobaloximes, which is believed to be the intermediate for the catalytic chain transfer process (Scheme 2.2), is highly dependent on the nature of the axial base component. The cobalt hydride in the CCT process is thought to be very short lived, but relatively stable hydrides are formed if the axial ligands are n-bonding as in the case of tri-alkylphosphines. It is then possible, that the cause of the reduction in rate, is due to the reinitiation step becoming slower, or in fact the hydridocobaloxime may live long enough to become significant in termination process, by disproportionating with the radical (equation (5.6)). It is seen that the loss of rate of polymerization is not as significant with COPhBF, as with COBF. A possible reason for this could in part be due to, at the same ChapterS 114 concentration of catalyst; the change in molecular weight reduction is only half as great for COPhBF than for COBF. Though, the effect of pyridine on the transfer constant of COPhBF in MMA is different, as it is seen that the fall in activity of the catalyst is faster in COPhBF than for COBF, and thus the two systems can not really be compared. Co(III) - H + R • ~ Co(ll) + Dead Polymer (5.6) If it is indeed the case, that the hydride can become stable enough that it slows the polymerization rate. Then it should not be discounted that the formation of the hydridocobaloxime species is, in part, a cause of the reduction in the rate of polymerization in normal bulk polymerization of MMA in the presence of catalytic chain transfer agents. 6,,------. (a) (b) 5 • 6 4 • • • •... • • .~ .. • • • ~ • §4 J .. • ~ .. • • • ...• • •Jl 0+---~--~--~----~--~~ 2 7 7 7 7 0.0 20xl Figure 5.11 : The effect of pyridine on the final conversions of the experiments conducted in MMA and (a) COBF (•) 0% pyridine (e) 1.1 pyridine (•) 2.2 pyridine (T) 4.4 pyridine and (b) COPhBF (•) 0% pyridine (e) 1.1% pyridine (•) 4.4% pyridine (T)8.8% pyridine Solvent effects on CCTP 115 4 5 (a) (b) 4 3 • • • • I • • A• • § t .,.A• .,.•A • 8 3"' I A -~ 2 • -~ • • ;> • ~2 ~ ~ 1 0+-~-.--~.-~-.--~.-~~ 0+---~--.---~---.--~--~ 0 0.0 [COBF]/[STY] [COPhBF]/[STY] Figure 5.12 : The effect of pyridine on the final conversions of the experiments conducted in styrene, Measured by gravimetry. (a) COBF (•) 0% pyridine (e) 1.1 pyridine (.6.) 2.2 pyridine and (b) COPhBF (•) 0% pyridine (e) 1.1% pyridine (.6.) 2.2% pyridine (T)8.8% pyridine Figure 5.13 : A comparison of the rate of polymerization of MMA; (•) by dilotometry and (D)gravimetry: in the presence of COPhBF over the rate of polymerization and in the absents of catalyst (e)by dilotometry and (O)gravimetry, versus the catalyst over the monomer concentration. Chapter 5 116 5.3.2.4 The effect of pyridine on the UVNIS absorption of COBF and COPhBF. Figure 5.14 and 5.15 shows the UVIVIS spectra for COBF in MMA and styrene respectively, and the effect of small editions of pyridine to the solution. The first difference that can be seen between the two spectra, before the addition of pyridine, is that the cobaloxime in MMA and styrene have different peak maxima. Confirming that in normal bulk polymerization the monomer acts as the axial ligand for the catalyst. On addition of a small amount of pyridine, in both cases, the spectral peak shifted down field. This shift in the UV NIS maxima is due to the stronger ligand taking up the axial position of the catalyst in place of the monomer. On further additions of pyridine a new absorption peak is starting to form at 493nm. Even at the lowest concentration of pyridine tested the peak at 493nm is starting to form. Figure 5.16 and 5.17 shows the effect of small additions of pyridine to the UVIVIS spectra for COPhBF in MMA and styrene solvent respectively. The same trend can be seen with COPhBF as was seen with COBF ie. the absorbance peak initially shifts down field on the addition of pyridine and then forms a new broader peak back up field, at 445nm for both .solvents, at further additions of pyridine for both solvents. It is interesting to note that the position of the absorbance peak at high excess of pyridine is catalyst dependent but not solvent dependent. We cannot conclude much about the structure of the catalyst, when there is a large excess of pyridine, from this result. The only thing we can deduce is that there are two species formed one at low pyridine concentrations and one at higher pyridine concentrations. These results concur with the results obtained when pyridine is added to the polymerization of styrene in the presence of CCT agents, where it was shown that the observed transfer constant increased significantly on the first addition of 1% pyridine, then decreased on further additions of pyridine. We could then say that the first species formed, which is a cobaloxime with a single pyridine ligand occupying an axial position, has a positive effect on the transfer constant of COBF and COPhBF in styrene (which is not seen in the MMA system, probably due to the rate of transfer for cobaloximes in methacrylates being at diffusion control limits already, See Chapters 3 and 4). The second species that is formed has probably a very low transfer constant rate of transfer, if at all, thus as this species forms the transfer constant is seen to fa.ll rapidly. Solvent effects on CCTP 117 Q~,------, Ul5 UIO ~ Q(l5 0.00 400 :ro run Figure 5.14: The effect of pyridine on the UVIVIS spectra of COBF in MMA solvent U3 ...... run Figure 5.15: The effect of pyridine on the UVIVIS spectra of COPhBF in MMA solvent ChapterS 118 0.20 ~-'-· t:·_, 0.15 (/) e3"'-<, 0.10 0.05 0.00 ' 375 400 425 450 475 500 525 550 run Figure 5.16 : The effect of pyridine on the UV NIS spectra of COBF in Styrene solvent U4~--~----~--~----~--~----~----r---~----, 493 ~ U2 400 450 500 run Figure 5.17 : The effect of pyridine on the UVNIS spectra of COPhBF in styrene solvent Solvent effects on CCTP 119 5.3.3 Summary of the Effect of Pyridine on the Kinetics of Catalytic Chain Transfer Polymerization In this work the effect of pyridine on the kinetics of catalytic chain transfer polymerization was tested. It was shown that pyridine decreases the transfer constant of both COBF and COPhBF in MMA. The transfer constant of these two catalysts in styrene however, showed at first an increase at the first addition of pyridine, then a decrease on further additions of pyridine. This behaviour is thought to be caused by, at first, a single pyridine as the axial ligand catalyst, changing the way in which the catalyst behaves. Secondly it is thought that an inactive species is formed on further additions of pyridine, which maybe a bi-ligand species. The UVNIS spectra confirmed that there is more than one catalyst species formed on the addition of pyridine. The rate of polymerization was seen to drop dramatically in the presence of both pyridine and catalyst in the MMA experiments, but not in the pyridine experiments. 5.4 The effect of Pyridine on the Propagation rate coefficient In the previous section the effect of pyridine on the kinetics of catalytic chain transfer was explored. It is known that pyridine has the largest solvent effect on the propagation rate coefficient, but it is not known as to what extent this solvent effect is. Bramford18 stated that the effect of pyridine on the propagation rate coefficient is as high as 45%, but this is assuming that there is no effect of pyridine on the termination rate coefficient. In order to be able to completely interpret the effect of pyridine on the kinetics of CCTP, it is necessary to know the propagation rate coefficient as a function of pyridine concentration. This was measured using the technique of pulsed laser polymerization (see section 2.1.5.1). ChapterS 120 5.4.1 Experimental 5.4.1.1 Pulsed laser polymerization The propagation rate coefficient of MMA and Styrene in pyridine at 57 .5°C was r measured using the technique of pulsed laser polymerization.l9 Purified monomer, pyridine, and the photoinitiator benzoin were weighed into pyrex sample tubes (10 mm diameter by 60mm height), which were then purged with high purity nitrogen for 5 minutes and sealed with rubber septa. The reaction mixtures were equilibrated at the reaction temperature prior to laser exposure. The polymerizations were initiated by a pulsed Nd:Yag laser (Continuum Surelite 1-20) with a harmonic generator (a Surelite SI..D-1 and SLT in series), which was used to produce the 355 nm UV laser radiation, and a wavelength separator (Surelite SSP-2), which was used to isolate the 355 nm beam. Constant pulsing frequencies of 10Hz were used. Polymerization activity was terminated by removing the sample from the laser, and precipitating the polymer into methanol. The molecular weight distributions of obtained polymers were determined subsequently by size exclusion chromatography. 5.4.2 Results and Discussion Figure 5.18 shows the typical molecular weight distribution of MMA generated by the PLP experiments, indicating the low molecular weight inflection point (used to calculate kp) and an overtone (at twice the molecular weight of the inflection point). It can be seen that the PLP characteristics are well defined, as there is a sharp primary peak and a small overtone, thus there was no difficulty in obtaining kp from these plots. Table 5.6 and Table 5.7 list the experimental conditions and the results obtained from the PLP experiments of MMA with pyridine, and styrene with pyridine respectively, at 57.5°C. Solvent effects on CCTP 121 1.4 1.2 (\ 1.0 0.8 i- 0.6 ~ 0.4 0.2 0.0 } -0.2 - v 1cf M Figure 5.18: Molecular weight distribution, w(logM), and the first derivative from the pulsed laser experiment of MA at 57.5°C, tr = 0.1. Note the point of inflection and the overtone at twice the molecular weight. It can be seen in Figure 5.19 that there is an increase of the propagation rate coefficient, kp, of MMA in the presence of pyridine. It can be seen that there is an increase of about 17%, from approximately 770 dm3.mor1.s-1 to 905 dm3.mor1.s-1 at 55mol% of pyridine in the feed. This increase, although quite large, is not as large as predicted by Bramford,l8 in where he assumed that the increase in rate due to kp only, and assumed kt to be unaffected by the presence of pyridine. These results indicate that at the concentrations of pyridine used for the previous set of work, the only effect on Cs is due to the transfer rate coefficient. As there is no noticeable change in kp. Figure 5.20 shows the effect of pyridine on the propagation rate coefficient, kp, of styrene. It can be seen from this plot that, to within error, there is no· effect of the pyridine on the kp of styrene. Chapter 5 122 Table 5.6: Results ofPLP exeeriments ofMMA in eiridine at 57.5°C fpyr [AIBN] tr LogMinf 100> 950- 900- • .~ • ---- 850------• ------800- ----·• • ~ ,__ ----- ·---- 750 • 700 650 600 I 0.0 0.1 02 03 0.4 05 0.6 ~ Figure 5.19: Dependence of Table 5.7: Results of PLP experiments of styrene in pyridine at 57.5°C fpyr [AIBN] tf LogMinf ~ 0.30 2.4 0.25 4.68 283 0.26 2.4 0.25 4.68 265 0.37 2.3 0.25 4.65 281 0.48 2.4 0.25 4.55 261 0.59 2.3 0.25 4.47 283 0.00 4.5 0.25 4.81 286 0.07 4.4 0.25 4.76 267 0.14 4.4 0.25 4.69 241 0.26 4.5 0.25 4.65 246 0.36 4.3 0.25 4.59 244 0.47 4.0 0.25 4.51 245 0.59 4.4 0.25 4.46 278 400 350 300 • • I - .-- .- • - 250 -. ¥" • • • • 200 150- 100 I 0.0 0.1 02 03 0.4 05 0.6 fpyr Figure 5.20: Dependence of 5.4.3 Summary of the Effect of Pyridine on the Propagation Rate coefficient of MMA and styrene In this work the solvent effect of pyridine on kp was tested in MMA and Styrene at 57 :-5°C. The data presented indicate that the propagation rate increases by as much as 17% at composition of 50 mol% pyridine, but there is no visable effect on styrene. 5.6 References 1. Halpern, J.; Ng, F. T. T.; Rempel, G. L. J. Am. Chern. Soc., 101, 7124 (1979). 2. Enikolopyan, N. S.; Smimov, B. R.; Ponomarev, G. V.; Belgovskii, I. M. J. Polym. Sci., Polym. Chern. Edn., 19, 879(1981). 3. Suddaby, K. G.; Maloney, D. R.; Haddleton, D. M. Macromolecules, 30, 702-713 (1997). 4. Sanayei, R. A.; ODriscoll., K.F. J. Macromol. Sci.-Chem., A26(8), 1137-1149 (1989). 5. Haddleton, D. M.; Maloney, D. R.; Suddaby, K. G.; Muir, A. V. G.; Richards, S. N. Macromol. Symp. 1996, 111,37 (1996). 6. Gridnev, A. A.; Ittel, S. D.; Fryd, M.; Wayland, B. B. Organometallics, 15, 222 (1996). 7. Gridnev, A. A. Polym. J., 24,613 (1992). 8. Heuts, J. P. A.; Kukulj, D.; Forster, D. J.; Davis, T. P. Macromolecules, 31, 2894 (1998). 9. Kukulj, D.; Davis, T. P. Macromol. Chern. Phys., 199, 1697 (1998). 10. Mayo, F. R. J. Am. Chern. Soc., 65, 2324 (1943). 11. Heuts, J.P. A.; Forster, D. J.; Davis, T. P.; Yamada, B.; Yamazoe, H.; Azukizawa, M. Macromolecules, 32, 2511-2519 (1998). 12. Gridnev, A. A. Polym. Sci. USSR, 31, 2369 (1989). 13. Bakac, A.; Brynildson, M. E.; Espenson, J. H. Inorg. Chern., 25,4108 (1986). 14. Suddaby, K. G.; Maloney, D. R.; Haddleton, D. M. Macromolecules, 30, 702 (1997). 15. Janowicz, A. H. US Patent4 694 054; Janowicz, A. H., Ed.; Janowicz: US, 1987. Solvent effects on CCTP 125 16. Gridnev, A. A.; Belgovskii, I. M.; Enikolopyan, N. S. Dokl. Akad. Nauk USSR (Engl. Trans!.), 289, 1408 (1986). 17. Shrauzer, G. N.; Holland, R. J. J. Am. Chern. Soc., 93, 1505 (1971). 18. Aliwi, S. M.; Bamford, C. M. J. Chern. Soc. Faraday I, 70, 2092 (1974). 19. Olaj, 0. F.; Bitai, 1.; Hinkelmann, F. Makromol. Chern., 188, 1689 (1987). Chapter 5 126 Reversible cobalt-carbon bond formation on CCTP in styrene 127 Chapter 6 , Reversible Cobalt-Carbon Bond Formation in the Catalytic Chain Transfer polymerization of Styrene Catalytic chain transfer, described in the previous chapters, has been shown to be a very efficient method for controlling molecular weight in free radical polymerization of methacrylates, where an a-methyl group is present. It is generally assumedl-11 that the hydrogen abstraction and the subsequent hydrogen addition to monomer proceed through a Co(III)-H intermediate (see scheme 2.2), but despite many efforts, most notably those by Gridnev and coworkers12-14, no conclusive evidence for this mechanism exists at present. Furthermore, assuming that this two step process is indeed involved in catalytic chain transfer polymerization, the mechanism of the actual hydrogen abstraction step leading to a dead polymer chain and the Co(III)-H is still not known. This mechanism could, for example, be a disproportionation reaction between the propagating radical and the Co(II) complex or it could involve a (3-elimination step, which requires the formation of a bond between the propagating radical and the Co(ll) complex. It has been well established that cobalt-carbon bond formation occurs between Co(ll) complexes and organic radicals, and many such compounds have been observed6,7. In the case of the acrylates this process even leads to living behavior.l5-17 In chapters 3 and 4 of this thesis it was shown that there is good evidence supporting a diffusion controlled transfer reaction. Although the discussion regarding diffusion controlled reaction rates in the catalytic chain transfer polymerization of the methacrylates does not seem to apply to the catalytic chain transfer polymerization of styrene. Firstly, reported chain transfer constants for COBF and COPhBF in styrene are about one to two orders of magnitude lower than those reported in MMA, leading to 5 3 1 1 values of ktr,Co of the order of 10 dm mor s- , which are clearly too low for a Chapter6 128 diffusion-controlled reaction under ordinary free-radical polymerization conditions. Secondly, the only studylO on the temperature dependence of the catalytic chain transfer polymerization of styrene that we are aware of that has appeared in the literature to date, indicates very different trends when compared to the methacrylates. Although there is a large scatter in the data, a negative activation energy and a frequency factor of -103 dm3 1 mor S-l were found for ktr,Co in styrene. It is clear that a negative activation energy is not consistent with an ordinary chemically-controlled reaction rate, and that the low frequency factor is not consistent with a diffusion-controlled reaction rate. At this point possible explanations for these observations will not .be discussed, but will simply conclude there are obviously some significant differences between the catalytic chain transfer polymerizations of the methacrylates and styrene. A recent publication by Heuts et al18, studied the reversible cobalt-carbon bond formation in styrene by electron paramagnetic resonance and also followed conversion and molecular weights in time. This study clearly indicated that cobalt-carbon bond formation is significant in the free-radical polymerization of styrene and needs to be invoked to explain the kinetics and the molecular weight evolution in the early stages of polymerization. Further more Heuts18 briefly discussed the possible effect of this cobalt-carbon bond formation on the measured chain transfer constant (i.e., a macroscopic property), which may be significantly lower than the real chain transfer constant. This chapter will further probe several aspects of the effect of cobalt-carbon bond formation on catalytic chain transfer polymerization and its possible role in the actual mechanism of the chain transfer reaction. The chain transfer constant, Cs, of both COBF and COPhBF are measured as a function of percent conversion during the initial stages of the free-radical polymerization of styrene. This study also investigates the effect temperature on the chain transfer constant of the cobalt complexes in styrene. The chain transfer constant of COPhBF is measured in p-methoxystyrene to determine the effect of the radical strength on the transfer mechanism. Reversible cobalt-carbon bond formation on CCTP in styrene 129 6.1 Transfer Constant as a Function of Conversion The aim of the work presented in this section was to investigate the effect of the formation of cobalt-carbon bonds on the initial stages of the catalytic transfer polymerization of styrene. The catalysts used in this section are COBF and COPhBF, ~ which showed transfer constants in the order of 35,000 and 20,000 in methyl methacrylate respectively. 6.1.1 Experimental 6.1.1.1 Materials The COBF was prepared according to the method described in section 3.2.1.1. The bis(methanol) complex of COPhBF was prepared using the same procedure, replacing dimethyl glyoxime in the described procedure for diphenyl glyoxime. Styrene (Aldrich, 99%) monomer was passed through a column of activated basic alumina (ACROS, 50- 200 micron) and was purged with high purity nitrogen (BOC) for 1.5 hours prior to use. 6.1.1.2 General polymerization procedure Polymerizations were carried out as described in the previous chapters. Two stock solutions were prepared: (i) and initiator stock solution and a (ii) catalyst stock solution. (i) The initiator stock solution was prepared by dissolution of 250 mg of AIBN into 150 ml of monomer. (ii) The catalyst stock solution was prepared by dissolution of 2 mg of COBF, or 3.3 mg of COPhBF into 10 ml of solution (i). Five sets of four reaction mixtures were prepared by the addition of 0.2, 0.3, 0.4 and 0.5 ml of solution (ii) to 4ml of solution (i). The polymerizations were carried out in a thermostatted water bath at 60°C. Sample sets were removed periodically for conversion measurement (by gravimetry) and molecular weight analysis. Molecular weight analysis was carried out as described in previous chapters. Chapter6 130 6.1.2 Results and Discussion Figures 6.1 and 6.2 show the evolution of the cumulative molecular weight distributions of polystyrene, formed in the presence of a constant concentration of catalytic chain ~ transfer agent, at 60°C, as a function of conversion, with the molecular weight distributions scaled for conversion. It is obvious from these plots, that as conversion increases the molecular weight distribution is moving toward higher molecular weight (for the initial stages of polymerization). This is consistent with the pseudo living polymerization shown by Heuts et al18, for the polymerization of COPhBF in styrene (i.e. during the initial stages, the system resembles that of a living system where the observed molecular weight grows with conversion). Though it is clearly observed in Figures 6.1 and 6.2, that the molecular weight evolution and the broadening of the molecular weight distribution with increasing conversion, is more likely to be attributed to the molecular weight distribution that is formed at any one time, moving toward higher molecular weight. Thus the low molecular weight polymer formed at the beginning of the reaction, becomes part of the low molecular weight tail that is seen at later conversions. The cobalt-carbon bond formation at the initial stages of polymerization will have a two-fold effect. The first effect is that it will reduce the amount of active radicals, during the initial stages. This effect can be attributed to by two factors; (i) the primary factor is the reduction of active radicals due to the formation of the cobalt-carbon complex, (ii) and to a lesser extent the increase in the termination of propagating chains due to the formation of lower molecular weights. Previously inhibition periods were measured for styrene systems with a very high catalyst concentration18. At the high concentrations of catalyst used in the previous study, the transfer rate should be high enough at the initial stages that an induction period should exist. At the lower catalyst concentrations used in this work, the chains have a higher probability for chain growth prior to encounters with the cobalt complex. In this case an induction period may not be observed. Reversible cobalt-carbon bond formation on CCTP in styrene 131 The second effect, which is more important, is the reduction of the effective concentration of Co(ll) in the system, implying that the transfer constants measured from a typical Mayo plot based on the initial Co(ll) concentration could be too low. This effect has already been mentioned in previous works.10,18 The- chain transfer constant, Cs, of COBF and COPhBF in the bulk polymerization of styrene, was measured as a function of conversion using both the Mayo, obtaining Mn directly from the GPC analysis, and the approximation of Mn = Mw/2, and the CLD method taking slopes the high molecular weight side of the peak molecular weight of the ln(P) vs M plot. The results from the three methods are given in Table 6.1. Figures 6.3 and 6.4 show the transfer constant of COBF and COPhBF in styrene at a constant initiator concentration of 1x10-2 mol·r\ at 60°C. It is obvious from these results that there is a sharp decrease in the transfer constant during the first percent conversion, due to the formation of cobalt-carbon bonds, which is explained in greater detail further on in this chapter. It can also be seen, that if the curves of the plots were extrapolated to the first instance polymer is made, the chain transfer constant would be of a similar order of magnitude, as that obtained for the transfer constants of COBF and COPhBF in monomers with a a-methyl grouping. Thus, it is conceivable that the transfer reaction of the cobalt (IT) complexes in styrene could also be diffusion controlled. It is interesting to note the difference in the Cs values obtained from the different methods of analyse. It is clear that as the equilibrium is reached the value for the transfer constant obtain using the Mayo (Mn) method is much higher than the values obtained for the Mayo (Mw) or the CLD method. The reason for this is simply that, at the beginning of the reaction there is low molecular weight polymer formed, as can be seen in Figures 6.1 and 6.2. Since the Mn of a polymer is very sensitive to any low molecular weight tail it is the method most sensitive to the formation of any low molecular weight tail to the distribution. Chapter6 132 O.Dl5 0.010 0.005 ···· ... ·. 0.~~~~~~~~~~~~--~~---~----~-~···~·-.. ~~~ 1~ l Figure 6.1 Evolution of the cumulative distributions observed in the catalytic chain transfer polymerization of styrene in the presence of COBF, at 60°C. The broadening of the distribution is caused by the instantaneous molecular weight distribution moving to higher molecular weight as conversion increases. O.OXl 0.015 } ~ 0.010 0.005 l Figure 6.2 Evolution of the cumulative distributions observed in the catalytic chain transfer polymerization of styrene in the presence of COPhBF, at 60°C. The broadening of the distribution is caused by the instantaneous molecular weight distribution moving to higher molecular weight as conversion increases. Reversible cobalt-carbon bond formation on CCTP in styrene 133 ~.------, 400) 0 • • U:XX> A 0 A 0 0~--~--~;---~~--.---~-.--~--.-~--~-~ .6 0 2 4 6 8 10 12 %conversion Figure 6.3 Dependance of the observed transfer constant, Cs, of COBF in styrene at 60°C, on conversion. Cs determined using the (T) Mayo method (•) Mayo (Mw/2} and (0) the CLD method, taking the slopes on the high molecular weight side of the peak molecular weight. U:XX>- *0 !D) • A tro • cY 0 400 • A • 200 ~ o ~;I 0 ~ n 0 I ~ 0 2 4 6 8 10 12 %conversion Figure 6.4 Dependance of the observed transfer constant, Cs, of COPhBF on conversion. Cs determined using the (T) Mayo method (•) Mayo (Mw/2) and (0) the CLD method, taking the slopes on the high molecular weight side of the peak molecular weight. Chapter6 134 Table 6.1 Chain Transfer Constants of COBF at 40 and 60°C and COPhBF at 60°C, in styrene calculated at different conversions. Transfer constant (Cs) Catalyst Temperature a conversionb Mnc Mwf2d ClDe COBF 60 0.5 4700 3980 3960 1.6 2200 885 485 2.8 1600 345 120 4.6 1130 160 70 10.9 930 50 20 COBF 40 0.12 11400 12100 9950 0.45 10000 8100 6600 1.6 6600 5300 4300 2.4 5400 4700 3580 8.5 1060 80 25 COPhBF 60 0.3 1015 1015 960 0.7 850 650 500 1.5 675 285 200 1.8 370 200 150 2.4 280 130 95 2.8 175 120 90 3.5 150 105 65 5.7 45 58 32 7.4 85 40 18 11 49 25 31 a) Temperature in °C; b) Percent conversion calculated by gravimetry; c) Chain transfer constant of COBF determined using DPn =MJ(monomer mass); d) Chain transfer constant of COBF determined using DPn = Mw/(2·monomer mass); e) Slopes of the ln(P) vs M plots taken at the high molecular weight side of the peak molecular weight. Reversible cobalt-carbon bond formation on CCTP in styrene 135 6.1.2.1 Mechanistic interpretation As briefly stated in the introduction and discussed in detail by Heuts et al18, the catalytic chain transfer polymerization of styrene is characterized by a significant occurrence of the reversible formation of bonds between the propagating radicals and the' Co(ll) complex (The fact that this reaction seems to be significant in styrene polymerization and not methyl methacrylate is simply that in the former case tertiary radicals are present. Firstly tertiary radicals are more stable than secondary radicals, and secondly the formation of a bond between Co(ll) and a tertiary radical has greater steric requirements 11 ). initiation I~R· (6.1) propagation R~I + M (6.2) Reversible capping R~ + Co(ll) ~ R- Co(ill) (6.3) chain transfer klr R~I + Co(ll) pi + Co(lll) -H (6.4) Co(ll) + M klr R; + Co(ll) (6.5) termination k~c R~ R~ (6.7) I + J pi+j k~d R~ R~ ) pi (6.8) I + J + pj Scheme 6.1 Mechanism for the catalytic chain transfer reaction via a Co(III)-H intermediate with the reversible formation of cobalt-carbon bonds. Scheme 6.1 shows all the possible reactions that can take place during the CCTP of styrene. Equation (6.3) shows the reversible formation of cobalt-carbon bonds between the catalyst and the propagating radical, effectively reducing the concentration of Co(ll) in the reaction medium. A closer look at scheme 6.1 shows, that the reaction between Chapter6 136 the cobalt radical and the propagating radical (reactions 6.3 and 6.4) resembles the termination reactions of two propagating radicals (reactions 6.6 and 6. 7). This statement is reasonable since · there is strong evidence that the transfer reaction is basically a radical-radical reaction and the rates are approaching diffusion control, as are the termination reactions. The ratio of the termination mode for methyl methacrylate is known to be predominantly disproportionation 19. And it has been shown that, for the reaction of the cobalt (IT) species in methyl methacrylate, the concentration of the active cobalt (IT) during the polymerization of methyl methacrylate does not vary by much. The termination mode for styrene is thought to be predominantly recombination19. Hence for the reaction of cobalt (IT) with styryl radicals, it has been shown that most of the active cobalt (IT) species is consumed, due to the formation of cobalt-carbon bonds (i.e. recombination), during the free radical polymerization of styrene in the presence of cobalt {IT) species. It is also thought that, in the polymerization of acrylates, the termination mode is almost exclusively combination20. The polymerization of acrylates in the presence of low spin cobalt (IT) complexes results in a pseudo living polymerization system. The active acrylic radicals are reversibly capped by the cobalt (IT) species, and there is negligible chain transfer to cobalt, with the right conditions the molecular weight can be seen to grow proportional to conversion and low polydispersities are obtained.l5-17 The fact that there seems to be a correlation between the termination reaction and the transfer reaction indicates that there may not be much difference between the mechanisms of the transfer reaction to catalyst and a radical radical bimolecular termination reaction. 6.1.2.2 Kinetic Model In the previous section the molecular weight evolution of styrene in the presence of CCT agents as a function of conversion was studied, and it is the aim of this section to arrive at a comprehensive model to explain the features observed. The reactions shown in scheme 6.1 represents the set of chemical reactions that is modelled in this section. Reversible cobalt-carbon bond formation on CCTP in styrene 137 To obtain an equation or set of equations that describe the reactions in scheme 6.1. Firstly, the appropriate differential equations need to be defined and then solved. Starting with the expressions for initiator decomposition and the radical concentration we get: d[I] (6.9) d[R•] = (6.10) dt where k1 is the rate coefficient of coupling, and k2 is the rate coefficient of dissociation of the cobalt-carbon bond. At this point the model will deviate from previous models of similar systems. The main deviation is that at the onset of polymerization the equilibrium of equation (6.3) has not been established. Thus at the onset of polymerization the rate of polymerization is slowed by the k1[R.][Co(II)] term in the above equation, and hence it is possible that the retardation at the start of the reaction could be mistaken as an inhibition period. To model the loss of active cobalt in the system given in scheme 6.1, firstly we must assume that consumption of cobalt(m only occurs by the reaction given by equation (6.3) and this is the rate determining step of the transfer process. Then the rate of loss of cobalt is given by: d[Co(II)] =k ·[Co(II)]·[R*]-k ·[Co(ill)-R] (6.11) dt I 2 The description of the monomer consumption (which allows one to determine the time dependence on conversion and the conversion dependence on the observed transfer constant) is given by the differential equation expressed in equation (6.12), though as both the radical concentration and the cobalt (II) concentration are changing during the time of interest, and the fact that these values are unknown then it is pointless to go further with this equation. Chapter6 138 d[M] = - (k,[M] + k.[MJX R" l = -( k,[M] + kc. [~~~ll) R ·] dt (6.12) = - (kP[M]+ kcJco(II)])[R"] If we now define the concentration of Co(ID)-R as: [Co(III)- R] = [Co(II)]0 - [Co(ll)] (6.13) and by substituting equation (6.13) into equation (6.11) and integrating, gives the expression for the concentration of cobalt(II) with respect to time. (6.14) where [Co(II)lt is the concentration of free cobalt(II) at time t. [Co(II)]o is the concentration of active cobalt at time zero. By making the following assumptions: • The transfer reaction only occurs by free cobalt(II), and the transfer constant of the free cobalt(II) (assuming no bound cobalt) is 15000, which brings it closer to the transfer constant of the methacrylates. • The transfer reaction only occurs via equation (6.4). • The final value of [Co(II)]/ [Co(II)]O is very low, i.e. around 0.01. • The concentration of the radicals is low. This assumption is valid at the beginning of the reaction where we are most concerned, as the formed radicals are being capped by the cobalt species. But once the equilibrium is being reached the concentration of the radicals could go up by a couple of orders of magnitude. Reversible cobalt-carbon bond formation on CCTP in styrene 139 Since DPn-1 is directly related to the product of the chain transfer constant and the concentration of chain transfer agent, as given by the Mayo equation (equation (2.23)), the relationship between the observed and the true chain transfer constants is simply given by: (6.15) A graphical illustration of equation (6.14) can then be given in the form of the observed transfer constant as a function of time, and is shown in Figure 6.5. It can be easily seen that the shape of the plot, shown in Figure 6.5, that equation (6.14) has a similar trend as that shown by the observed transfer constant of COBF in styrene, shown in Figure 6.3. 15(XX)- 12500 10000 ~ ~ 7500 a~ 5(XX)- 2500- 0 0 40X) (ill) 10000 T:hre (sec) Figure 6.5: The time dependence of the observed transfer constant in catalytic chain transfer polymerization of styrene, according to equation 6.14. Chapter6 140 In the light of these results, it may be concluded that incredibly good control of conversion is required for catalytic chain transfer constant measurements in styrene. This may explain some of the unusual observations that have been reported in the literature. This result also indicates that there is not much difference between the rate of the hydrogen being extracted from the a-methyl position or being extracted from the b~ckbone of the polymer, which was mentioned as a possible explanation for the difference between the transfer constants in methyl methacrylate and styrene.lO Figures 6.6 shows the evolution of the cumulative molecular weight distribution of polystyrene, formed in the presence of COBF, at 40°C, as a function of conversion, with the molecular weight distributions scaled for conversion. It is also seen from this plot that as conversion increases the molecular weight of the polymer formed also increases as is to be expected. In this experiment though it is seen that there is a bimodal distribution once the reaction reached a higher conversion. At first look this distribution would seem to be an experiment that went wrong apart from the fact that this sort of distribution is reproducible at 40°C. The bimodal distribution may be caused by the cobalt-carbon bond formed between the catalyst and the styryl radical being reasonably stable, and that once the reaction has reached equilibrium the concentration of free cobalt is effectively zero allowing the polymer to go to high molecular weights. Figure 6. 7 shows the dependence of the transfer constant of COBF in styrene at 40°C. From this figure it can be seen that the transfer constant takes longer to reach equilibrium, than is seen when the reactions took place at 60°C, as shown in Figure 6.8. This can simply be explained by the fact that at 40°C the rate of radical production is two orders of magnitude lower than the rate of radical production at 60°C. Thus it then seems reasonable that it will take longer to obtain the equilibrium given by equation (6.3). Reversible cobalt-carbon bond formation on CCTP in styrene 141 UOOOr------, M Figure 6.6 Evolution of the cumulative distributions observed in the catalytic chain transfer polymerization of styrene in the presence of COBF at 40°C. 1400) 1200) ~ I tml- .& rY 0 (ill) .. • I 400)- :ml .. 0+---~-.---r--~--r--.--~--~~~~-.--~--;I I - 0 2 4 6 8 10 12 %conversion Figure 6.7 Dependance of the observed transfer constant, Cs, of COBF on conversion. Cs determined using the (T) Mayo method (.A.) Mayo (Mw/2) and (0) the CLD method, taking the slopes on the high molecular weight side of the peak molecular weight. Chapter6 142 1 liD) 0 (ill) 0 cY 0 400)- • :;ro)- 0 • • - - 0 2 4 - 6 8 10 - 12 %conversion Figure 6.8 Comparison of the obseved Cs of COBF in styrene at (e) 60°C and (0) 40°C. 6.1.3 Summary of the transfer constant as a function of conversion In this work the transfer constant of COBF and COPhBF in styrene were determined as a function of conversion. The transfer constant for both catalysts tested was found to decrease rapidly in the first few percent conversion, due to the formation of cobalt carbon bonds formed between the styryl radicals and the cobalt species. This explained some of the inconsistent data that has been published on CCTP in styrene, as it is obvious now that good control over the conversion to obtain any meaningful data. The rate of change of the observed Cs was also found to be dependent on the rate of radical production. 6.2 Temperature dependence of the transfer constant in styrene In the previous section it was shown that the kinetics of CCT in styrene is dominated by the formation of cobalt-carbon bonds. The significant formation of cobalt-carbon bonds Reversible cobalt-carbon bond formation on CCTP in styrene 143 and its effect on the observed chain transfer constant is also likely to cause a very different temperature dependence of the chain transfer constant, as the equilibrium of equation (6.3) should shift towards an increasing [Co(ll)] at higher temperatures. However, the only experimental study published on styrene thus far does not appear to be consistent with this argument~ for reasons unknown at the time of that publication and which were outlined above, the presented data may not be reliable enough to draw any mechanistic conclusions. It is well known that heat can affect the equilibrium of equation (6.3).15 This section will investigate the effect that temperature has on the observed transfer constant of COBF and COPhBF in styrene. 6.2.1 Experimental Polymerizations were carried out as described in the previous chapters. Two stock solutions were prepared: (i) and initiator stock solution and a (ii) catalyst stock solution. (i) The initiator stock solution was prepared by dissolution of 250 mg of AIBN or 250 mg of 2'2-azo-bis-cyclohexylpropionitrile (for the reactions 100 and 120C) into 150 ml of monomer. (ii) The catalyst stock solution was prepared by dissolution of 2 mg of COBF, or 3 mg of COPhBF into 10 ml of solution (i). Five sets of four reaction mixtures were prepared by the addition of 0.2, 0.3, 0.4 and 0.5 ml of solution (ii) to 4ml of solution (i). The polymerizations were carried out in a thermostated water bath at 40, 60 and 80°C, and in a thermostated water oil bath at 100 and 120°C. Sample sets were removed and conversion was measured (by gravimetry) and molecular weight analysis was performed. Only the samples at 5% conversion were used for this analysis. Molecular weight analysis was carried out as described in previous chapters. Chapter6 144 6.2.2 Results and Discussion The chain transfer constants for COBF and COPhBF were determined at varying temperatures, and the results are given in Table 6.2 and Figure 6.9. The transfer constant for each was only taken when the reactions had reached approximately 5%. This was to ensure that the transfer constant had reached a steady level, as seen in Figures 6.3 and 6.4. The chain length distribution method was used as the method of determining the chain transfer constant, as it is possible to choose the portion of the molecular weight distribution in which to analyze21,22 and thus, is the method least influenced by the formation of low molecular weight material in the initial stages of polymerization. The chain transfer constant for both catalysts is seen to increase after 80°C (Figure 6.9). The transfer constant for COBF is seen to increase from approximately 50, at temperatures less than 80°C, to approximately 1300 at l20°C. And for COPhBF the effect of temperature was less dramatic as there was only an increase from 20, below 80°C, to 135 at l20°C. This effect is due to the equilibrium in equation (6.3) being driven to the side of free cobalt(II), as the cobalt carbon bond becomes more labile at higher temperatures. The change in the transfer constant above 80°C is seen to be more dramatic in the case of COBF than for COPhBF. There does not seem to be any explanation for this other than the possibility that the cobalt-carbon bond formed between COPhBF and the styryl radical is more stable than the bond formed between the styryl radical and COBF. By modeling the effect of a varying equilibrium constant, for the reaction in equation (6.3), such that there is an increase in the active free radical concentration after the equilibrium is reached, and using equation (6.14). Figure 6.10 shows the model prediction when there is an increase in the active cobalt(II) after the equilibrium has been reached, we can see from this plot that the final value of the transfer constant obtained with an increase of active Co(II) also increases, which is what we would expect. Reversible cobalt-carbon bond formation on CCTP in styrene 145 11ID 1400 I 1:a:D • 1CXD • en 8Xl (.) liD 400 I :m 0 0 140 Temperature ( "C ) Figure 6.9: Dependence of the observed transfer constant, Cs, of (• )COBF and (O)COPhBF on temperature. 15000 12500 10000 10 7500 Cll u 5000 ,·, ,. lrae:sirg(~II})/[~IOJo ~ -+--'.....,<·...... ·~_~·-.....::· ·_::· ~-;::~ ·_==· ~==·~-=~ ~;;:::.~::::!:: t·~=·~=~:::;::~-~=~-=~·-=· ~-=;~ ·=~ ~=· ~ ·=~ ~=;- ~;=-~=~· ~=-~=~· _""--~=~·=~ ~=· ~:::;:-~ ~=· ~=-~=~· ~=;-~=~-=~·-=· ~-~~. 0 10000 titre (sec) Figure 6.10: The dependence of the observed transfer constant in catalytic chain transfer polymerization of styrene on the ratio of free cobalt(II) to initial cobalt(II), according to equation 6.14. Chapter6 146 Table 6.2: Chain Transfer Constants of COBF and COPhBF at 40, 60, 80, 100, and 120°C% in st;yrene calculated using the CLD method. Temperature COBF COPhBF Conversion Cs (ClD) Conversion Cs (CLD) 40 5.2 16 5.6 18 40 5.1 21 5.5 20 40 5.7 76 5.3 17 40 5.7 80 5.1 32 60 4.6 31 5.5 20 60 4.4 27 5.3 22 60 5.6 48 5.1 18 60 5.5 46 4.9 20 80 5.4 62 5.3 27 80 4.7 54 5.2 30 80 5.3 46 6.5 36 80 5.1 45 5.7 22 100 5.7 375 5.5 98 100 5.4 448 5.4 71 100 5.6 395 5.3 72 100 5.7 501 5.5 96 120 5.5 1207 8.1 132 120 5.9 1063 8.4 130 120 6.0 1395 8.4 140 120 5.6 1442 8.7 130 Reversible cobalt-carbon bond formation in CCTP of styrene 147 Figure 6.11 shows the effect of temperature on the rate of polymerization of styrene, in the presence of COBF and COPhBF. It can be seen in this figure that at 40 and 60°C there is no noticeable dependence, of the rate of polymerization of styrene, on the catalyst type. When the reactions are conducted at 80°C there is a slight decrease in the rate of polymerization of styrene in the presence of COBF compared to COPhBF. As tne temperature of the reaction increases, the difference between the two catalyst becomes larger. In light of the effect of the transfer constant with respect to temperature, this effect of the slowing down of the rate of polymerization of styrene in the presences of COBF above 80°C, can be explained simply by the fact that, as seen in Figure 6.9, the transfer reaction becomes more dominant, for COBF, at temperatures above 80°C. Thus lower molecular weights are being produced compared to that of the reactions with COPhBF, and so lower rates are being observed. 0.6-r--~------. 0.4 I• ~~ 0.2 g 0 0 • ("'\ 0.0 -.. I 40 60 80 100 120 Ten:perature (°C) Figure 6.11: The effect of temperature on the rate of polymerization in the presence of ( 0) COBF and ( •) COPhBF Chapter6 148 6.2.3 Summary of the transfer constant as a function of temperature In this work the transfer constant of COBF and COPhBF in styrene was determined as a function of temperature. The transfer constant at 5% conversion for both catalysts tested was found to increase dramatically after 80°C, due to the cobalt carbon bonds formed between the styryl radicals and the cobalt species becoming more labile changing the equilibrium between the active cobalt species and the inactive Co(III). The rate of polymerization of styrene in the presence of the two catalyst was found to diverge after 80°C due to transfer becoming more dominant in the CCTP of styrene with COBF than with COPhBF slowing the rate of the former in comparison to the later. 6.3 Transfer constant of COPhBF in p-methoxystyrene In the previous section of this chapter it was shown that the polymerization of styrene in the presence of a CCT agent is dominated by the formation of cobalt carbon bonds, affecting the efficiency of the transfer process. It was also shown that the equilibrium between the Co(ll) and the Co(Ill) can be affected by temperature. It is known that substituents (electron withdrawing or electron donating), on the phenyl ring of the styrene, affects the amount in which the unpaired electron is delocalized within the phenyl ring. p-methoxystyrene is known to terminate predominately but not solely by combination of the radicals23. The stability of the secondary radical formed by p-methoxystyrene is affected by an additional contributing structure (scheme 6.2)24. Thus it is of interest to determine the effect of the stability of the styryl radical on the CCTP kinetics. Reversible cobalt-carbon bond formation on CCTP in styrene 149 Scheme 6.2: Radical stabilisation of the secondary radical formed in the free radical polymerization of p-methoxystyrene. This section investigates the polymerization of p-methoxystyrene in the presence of a catalytic chain transfer agent (COPhBF) to determine the effect of radical stability on the mechanism of catalytic chain transfer in styrenes. 6.3.1 Experimental 6.3.1.1 Materials The COPhBF was prepared according to the method described in section 3.2.1.1, replacing dimethyl glyoxime in the described procedure for diphenyl glyoxime. P methoxystyrene (Aldrich, 99%) monomer used without further purification, and was purged with high purity nitrogen (BOC) for 1.5 hours prior to use. 6.3.1.2 General polymerization procedure Polymerizations were carried out in toluene solvent. Two stock solutions were prepared: (i) and initiator stock solution and a (ii) catalyst stock solution. (i) The initiator stock solution was prepared by dissolution of 50 mg of AIBN into 50 ml of toluene. (ii) The catalyst stock solution was prepared by dissolution of 3.3 mg of COPhBF into 10 ml of solution (i). Two sets of five reaction mixtures were prepared by the addition of 0.2, 0.3, 0.4 and 0.5 ml of solution (ii) and made up to 4 ml with solution (i), and 4 ml of degassed monomer was added to each mixture. Chapter6 150 The polymerizations were carried out in a thermostatted water bath at 60°C. Sample polymer was collected, after the conversion reached 5%, by precipitation in cold methanol. Molecular weight analysis was carried out as described in previous chapters. 6.3.2 Results and Discussion Figure 6.12 gives the molecular weight distribution obtained for the determination of the transfer constant of COPhBF in 4-methoxystyrene. It can be seen from the distributions that the presence of the catalytic chain transfer agent has little effect on the final molecular weight of the polymer. And it can be seen that in the plot of In P(M) vs M (Figure 6.13) there is no difference in the slope of the line above the peak molecular weight indicating that there is no transfer occurring. It is clear from this result that COPhBF is much less effective in 4-methoxystyrene, than the equivalent reaction in styrene. This raises the question as to why this should occur. The first possibility is that the stability of the radical formed in the case of p-methoxystyrene is such that there is no interaction between the cobalt species and the radical. Though the low molecular weight portion of the molecular weight distributions is clearly different for the reactions conducted in the presence of CCT agent compared to the distribution formed in the absence of CCT agent. Thus there is an obvious effect of the CCT agent indicating a reaction with the radical at some stage. The second possibility is there are some solvent effects of toluene on the mechanism of transfer. Though in the previous chapter it was shown that weak coordinating solvents such as toluene do not affect the kinetics of CCTP. This leaves the second possibility, catalyst poisoning during the initial stages of the reaction. The results obtained by both replicates indicating the same trend of a zero transfer constant, so the catalyst is not poisoned by anything other than the propagating radical. It was shown in the previous section of this chapter that the observed transfer constant drops rapidly in the first percent of conversion due to the loss of active cobalt (IT) species by combining reversibly with the radical. It is possible that the bond formed, between the cobalt species and the propagating p-methoxystyrene radical, is stable enough that it reduces the concentration of the active cobalt complex to zero. The Reversible cobalt-carbon bond formation on CCTP in styrene 151 difference in the low molecular weight side of the molecular weight distributions is caused at the beginning of the reaction before all the cobalt (II) species become bound. 0.2 0.1 0.04=~----~--~~--~~~~------~--~~~~~~ l<:f lcf logM Figure 6.12 Molecular weight distributions observed in the catalytic chain transfer polymerization of p-methoxystyrene in the presence of COPhBF at 60°C -14 -16 ~ ..s -18 -, l.Oxlcf 20xlcf 3.0xlcf 4.0xlcf ' M Figure 6.13 In P(M) vs M plots observed in the catalytic chain transfer polymerization of p-methoxystyrene in the presence of COBF at 60°C Chapter6 152 6.3.3 Summary of catalytic chain transfer polymerization of 4- methoxystyrene In this work the transfer constant of COPhBF in styrene was determined in 4- methoxystyrene. The CCT agent seemed not to have an effect on the molecular weight ' distribution after the initial stages of polymerization. Though a low molecular weight tail indicated that the catalyst may have had an effect in the initial stages of polymerization. The catalyst is thought to then form a stable bond with the propagating radical effectively reducing the concentration of the catalyst to zero. 6.4 References 1). Karmilova, L. V.; Ponomarev, G. V.; Smirnov, B. R.; Belgovskii, I. M. Russ.Chem.Rev., 53, 132 (1984). 2. Davis, T. P.; Haddleton, D. M.; Richards, S. N.J. Macromol. Sci., Rev. Macromol. Chem. Phys., C34, 234(1994). 3. Davis, T. P.; Kukulj, D.; Haddleton, D. M.; Maloney, D. R. Trends Polym. Sci., 3, 365 (1995). 4. Burczyk, A. F.; ODriscoll, K. F.; Rempel, G. L. J. Polym. Sci., Polym. Chem. Edn., 22, 3255 (1984). 5. Gridnev, A. A. Polym. Sci. USSR, 31,2369 (1989). 6. Gridnev, A. A.; Ittel, S. D.; Fryd, M.; Wayland, B. B. Organometallics, 12, 4871 (1993). 7. Gridnev, A. A.; Ittel, S. D.; Fryd, M.; Wayland, B. B. J. Chem. Soc., Chem. Commun., 1010 (1993). 8. Haddleton, D. M.; Maloney, D. R.; Suddaby, K. G.; Muir, A. V. G.; Richards, S. N. Macromol. Symp., 111, 37 (1996). 9. Suddaby, K. G.; Maloney, D. R.; Haddleton, D. M. Macromolecules, 30, 702 (1997). 10. Kukulj, D.; Davis, T. P. Macromol. Chem. Phys., 199, 1697 (1998). Reversible cobalt-carbon bond formation on CCTP in styrene 153 11. Heuts, J. P. A.; Kukulj, D.; Forster, D. J.; Davis, T. P. Macromolecules, 31, 2894 (1998). 12. Gridnev, A. A.; lttel, S. D.; Fryd, M.; Wayland, B. B. Organometallics, 15, 222 (1996). 13. Gridnev, A. A.; lttel, S. D.; Wayland, B. B.; Fryd, M. Organometallics, 15, 5116 r (1996). 14. Gridnev, A. A.; lttel, S.D.; Wayland, B. B.; Fryd, M. Organometallics, 15, 5116- Supplementary material (1996). 15. Wayland, B. B.; Poszmik, G.; Mukerjee, S. L.; Fr~d, M. J. Am. Chern. Soc., 116, 7943 (1994). 16. Wayland, B. B.; Basickes, L.; Mukerjee, S.; Wei, M.; Fryd, M. Macromolecules, 30, 8109 (1997). 17. Wayland, B. B.; Mukerjee, S.; Poszmik, G.; Woska, D. C.; Basickes, L.; Gridnev, A. A.; Fryd, M.; lttel, S. D. Control of Radical polymerizations by Metalloradicals; Wayland, B. B.; Mukerjee, S.; Poszmik, G.; Woska, D. C.; Basickes, L.; Gridnev, A. A.; Fryd, M.; lttel, S.D., Ed.; American Chemical Society: Washington, DC,; 685, 305 (1998). 18. Heuts, J.P. A.; Forster, D. J.; Davis, T. P.; Yamada, B.; Yamazoe, H.; Azukizawa, M. Macromolecules, 32, 2511-2519 (1998). 19. Zammit, M. D.; Davis, T. P.; Haddleton, D. M.; Suddaby, K. G. Macromolecules, 30, 1915 (1997). 20. Moad, G.; Solomon, D. H. The Chemistry of Free Radical Polymerization; Pergamon: Oxford, 1995. 21. Heuts, J.P. A. Davis, T.P.; and Russell, G. T. submitted 1999. 22. Moad, G.; Moad, C. L. Macromolecules, 29,7727 (1996). 23. Berry, R. W. H.; Ludlow, A. J.; Mazza, R. J. Macromol. Chern., Phys. 198, (1997). 24. Berry, R. W. H.; Ludlow, A. J.; Mazza, R. J. Macromol. Chern. Phys., 198, 1579- 1595 (1997). Chapter6 154 Conclusions and Recommendations 155 Chapter 7 Conclusions and Recommendations This thesis contains the results of an investigation into several aspects relating to the kinetics of free-radical catalytic chain transfer polymerization. This was an endeavour to obtain a greater mechanistic understanding of the transfer process in bulk and solution polymerizations. The specific areas investigated include the rate of transfer to methacrylates, the kinetics of CCTP in a supercritical medium, the role of monomer on the transfer process, the effect of solvents on the kinetics of the transfer reaction and the kinetics of CCTP in styrene. In chapter 3 The chain transfer constant for COBF and COPhBF was measured in MMA, EMA, BMA and POEMA. The data presented clearly show a dependence of the catalytic chain transfer constant on the size of the ester group in the methacrylate series, ie., Cs,MMA > Cs,EMA > Cs,BMA > Cs,POEMA for both catalysts tested. Furthermore, a virtually constant chain transfer constant was observed with changing temperature, yielding a pre-exponential factor A of about 1010 and an activation energy of about 17 - 23 kJ mor1 for the rate determining step in the chain transfer reaction. These observations are consistent with a diffusion-controlled rate-determining step in the catalytic chain transfer reaction of methacrylates. The conclusion of a diffusion controlled rate-determining step is reinforced by an inverse correlation of the Cs values with monomer viscosity, as f]MMA < f1EMA < f1BMA < f1POEMA. Furthermore, it was shown that for a diffusion-controlled chain transfer reaction, the product of Cs, kp and f]a should be constant - which was confirmed for MMA, EMA and BMA ·at 40, 50 and 60°C. Recommendations for further work: The transfer constants for these sets of experiments were conducted in bulk-and at low conversions. It may be of interest to Chapter 7 156 extend this to higher conversions, to see if there is still a correlation between the viscosity of the reaction medium to the rate of transfer. In chapter 4 The chain transfer constant of COPhBF was determined in supercritical CQ2• The value of the transfer constant of COPhBF in supercritical COz was found to be enhanced by an order of magnitude compared to that expected for bulk or solution polymerization, providing further evidence for a diffusion controlled transfer reaction. Recommendations for further work: This work only involved CCTP of MMA in supercritical fluid. There is an opportunity to extend this work to the production of low molecular weight fluorinated polymers, as they should be more soluble in the supercritical medium. In chapter 5 The 'role of the monomer on the transfer process was investigated, also the transfer constant of COBF and COPhBF was determined in the presences of coordinating and non-coordination solvents. It was found that the transfer process was zero order with respect to monomer. Non-coordinating solvents were found to have no effect on the kinetics of catalytic chain transfer, while pyridine was found to a detrimental effect on the transfer reaction and rate of polymerization in MMA. In styrene, however, the transfer reaction was enhanced at low concentrations of pyridine, but on further additions of pyridine the transfer constant dropped again. This behaviour is thought to be caused by, at first, a single pyridine as the axial ligand catalyst, changing the catalytic process. Secondly it is thought that an inactive species is formed on further additions of pyridine, which maybe a bi-ligand species. The effect of pyridine Conclusions and Recommendations 157 was shown to have a significant effect on the kp of MMA, but had no effect on the kp of styrene. Recommendations for further work: The first two sections of this work is clear, but the work with pyridine could be extended to include the effect of pyridine on the molecular weight distribution as the reaction reaches higher conversion. Also some ESR studies of the effect of pyridine on CCT in styrene may show if the cobalt is still predominantly bound to the radical end. In chapter 6 the transfer constant of COBF and COPhBF was determined in styrene as a function of conversion and also as a function of temperature. The transfer constant of COPhBF was also found in p-methoxy styrene. The transfer constant for both catalysts tested was found to decrease rapidly in the first few percent conversion, due to the formation of cobalt carbon bonds formed between the styryl radicals and the cobalt species. The transfer constant was found to increase dramatically after 80°C, due to the cobalt carbon bonds formed between the styryl radicals and the cobalt species becoming more labile changing the equilibrium between the active cobalt species and the inactive Co(lll). There was little effect of the catalyst in p-methoxy styrene possible due to the catalyst binding irreversibly to the radical. Recommendations for further work: The reactions in this chapter were conducted using thermal initiation. Further work on this could involve the determination of the effect of strong UV light on the Co(II)/Co(III) equilibrium in polymerization reactions. Chapter 7 158 Appendix 159 Appendix 1 Al.l Data for C8 of COBF in Methyl methacrylate Table Al.l: Experimental conditions and results for the determination of C8 for COBF in methyl methacrylate. T a) Cone b) [AIBN] c) t d) Mne) Mwf) PDi 40 0 16 120 770 1440 1.9 40 9.84 x 10-8 16 120 34 58 1.7 40 1.92 x 10-7 16 120 10 27 2.6 40 2.s1 x 10-7 16 120 10 24 2.3 40 3.67 x 10-7 16 120 7.0 16 2.2 50 0 16 40 436 823 1.9 50 9.84 x 10-8 16 40 30 52 1.7 50 1.92 x 10-7 16 40 12 26 2.1 50 2.s1 x 10-7 16 40 8.0 18 2.3 60 0 16 20 269 469 1.8 60 9.84 x 10-8 16 20 32 55 1.7 60 1.92 x 10-7 16 20 16 28 1.8 60 2.81 x 10-7 16 20 7.2 18 2.5 60 3.67 x 10-7 16 20 7.6 17 2.2 70 0 16 10 131 245 1.9 70 9.84 x 10-8 16 10 31 55 1.8 70 1.92 x 10-7 16 10 12 29 2.4 70 2.81 x 10-7 16 10 9.9 21 2.1 70 3.67 x 10-7 16 10 8.6 18 2.2 70 0 16 10 110 195 1.8 70 1.22 x 10-7 16 10 27 49 1.8 70 2.37 x 10-7 16 10 13 26 2.0 70 3.48 x 10-7 16 10 10 21 2.1 70 4.53 x 10-7 16 10 7.2 15 2.0 a) Temperature /°C b) [COBF]/[MMA] c) Initiator /mmoi.L-1 d) Reaction time /min e) Mn X w-3 f) Mw x w-3 160 A1.2 Data for C8 of COBF in Ethyl methacrylate Table A1.2: Experimental conditions and results for the determination of C8 for COBF in ethyl methacrylate. Ta) Conch) [AIBN] c) t d) Mne) Mwt) PDi ~ 40 0 16 120 720 1410 2.0 40 1.42 x 10-7 16 120 24 65 2.7 40 2.11 x 10-7 16 120 13 35 2.7 40 4.06 x 10-7 16 120 9.4 22 2.4 40 5.30 x 10-7 16 120 7.5 18 2.5 50 0 16 40 380 6260 1.7 50 1.42 x 10-7 16 40 28 53 1.9 50 2.11 x 10-7 16 40 11 27 2.4 50 4.06 x 10-7 16 40 10 19 1.8 50 5.30 x 10-7 16 40 6.8 14 2.1 60 0 16 20 325 515 1.6 60 1.42 x 10-7 16 20 28 53 1.9 60 2.11 x 10-7 16 20 13 28 2.1 60 4.06 x 10-7 16 20 11 20 1.8 60 5.30 x 10-7 16 20 7.5 16 2.1 70 0 16 10 164 313 1.9 70 1.1s x 10-7 16 10 35 67 1.9 70 2.25 x 10-7 16 10 19 37 1.9 70 3.29 x 10-7 16 10 12 28 2.2 70 4.29 x 10-7 16 10 10 21 2.1 70 0 16 10 138 288 2.1 70 1.29 x 10-7 16 10 27 56 2.0 70 2.51 x 10-7 16 10 16 32 2.0 70 3.68 x 10-7 16 10 12 27 2.3 70 4.79 x 10-7 16 10 9.0 19 2.1 a) Temperature /°C b) [COBF]/[EMA] c) Initiator /mmoi.L-1 d) Reaction time /min 3 e) Mn X 10-3 f) Mw X 10- Appendix 161 A1.3 Data for C8 of COBF in Butyl methacrylate Table A1.3: Experimental conditions and results for the determination of C8 for COBF in butyl methacrylate. T a) Cone b) [AIBN] c) t d) Mne) <' MwO PDi 40 0 16 120 1288 2340 1.8 40 1.82 x 10-7 16 120 56 107 1.9 40 3.55 x 10-7 16 120 32 60 1.9 40 5.19 x 10-7 16 120 18 30 1.6 40 6.77 x 10-7 16 120 16 30 1.9 50 0 16 40 733 1370 1.9 50 1.82 x 10-7 16 40 46 87 1.9 50 3.55 x 10-7 16 40 22 40 1.8 50 5.19 x 10-7 16 40 18 31 1.7 50 6.77 x 10-7 16 40 14 26 1.8 60 0 16 20 523 1000 1.9 60 1.82 x 10-7 16 20 543 91 1.7 60 3.55 x 10-7 16 20 263 45 1.7 60 5.19 x 10-7 16 20 18 32 1.8 60 6.77 x 10-7 16 20 13 27 2.1 70 0 16 10 268 519 1.9 70 1.82 x 10-7 16 10 49 84 1.7 70 3.55 x 10-7 16 10 20 47 2.3 70 5.19 x 10-7 16 10 17 35 2.1 70 6.77 x 10-7 16 10 12 26 2.1 a) Temperature /°C b) [COBF]/[BMA] c) Initiator /mmol.L-1 d) Reaction time /min 3 e) Mn X 10-3 f) Mw x 10- 162 A1.4 Data for C8 of COPhBF in Methyl methacrylate Table A1.4: Experimental conditions and results for the determination of C8 for COPhBF in methyl methacrylate. T a) Cone b) [AIBN] c) t d) Mne) MwO PDi 40 0 16 120 644 1120 1.8 40 1.00 x 10-7 16 120 61 100 1.7 40 1.96 x 10-7 16 120 34 55 1.6 40 3.75 x 10-7 16 120 15 28 1.9 50 0 16 40 312 617 2.0 50 1.00 x 10-7 16 40 50 88 1.8 50 1.96 x 10-7 16 40 47 97 2.1 50 2.89 x 10-7 16 40 15 36 2.3 50 3.75 x 10-7 16 40 14 27 1.9 60 0 16 20 243 420 1.7 60 1.00 x 10-7 16 20 47 83 1.8 60 1.96 x 10-7 16 20 35 63 1.8 60 2.89 x 10-7 16 20 22 40 1.8 60 3.75 x 10-7 16 20 11 25 2.3 70 0 16 10 130 259 2.0 70 1.00 x 10-7 16 10 41 72 1.8 70 1.96 x 10-7 16 10 34 58 1.7 70 2.89 x 10-7 16 10 17 31 1.8 70 3.75 x 10-7 16 10 11 24 2.2 a) Temperature /°C b) [COBF]/[MMA] c) Initiator /mmoi.L-1 d) Reaction time /min 3 e) Mn X 10-3 f) Mwx 10- Appendix 163 A1.5 Data for C8 of COPhBF in Ethyl methacrylate Table Al.S: Experimental conditions and results for the determination of C8 for COPhBF in ethyl methacrylate. Ta) Cone b) [AIBN] c) t d) Mne) Mwf) PDi ' 40 0 16 120 1084 1670 1.7 40 1.31 X IQ-7 16 120 63 122 1.9 40 2.55 X IQ-7 16 120 29 54 1.8 40 3.74 X IQ-7 16 120 20 36 1.8 40 4.87 X IQ-7 16 120 15 27 1.7 40 0 16 120 1199 1920 1.6 40 1.31 X IQ-7 16 120 63 127 2.0 40 2.55 X IQ-7 16 120 31 56 1.8 40 3.74 X IQ-7 16 120 20 34 1.7 40 4.87 X IQ-7 16 120 12 24 2.0 50 0 16 40 637 1190 1.9 50 1.35 X IQ-7 16 40 77 155 2.0 50 2.64 X IQ-7 16 40 34 66 2.0 50 3.86 X IQ-7 16 40 22 40 1.8 50 5.03 x w-7 16 40 14 27 1.9 50 0 16 40 550 899 1.6 50 1.35 X IQ-7 16 40 89 175 2.0 50 2.64 X IQ-7 16 40 35 70 2.0 50 3.86 X IQ-7 16 40 19 40 2.1 50 5.03 X IQ-7 16 40 15 28 1.9 60 0 16 20 397 684 1.7 60 1.31 X IQ-7 16 20 55 110 2.0 60 2.55 x w-7 16 20 25 49 2.0 60 3.74 X IQ-7 16 20 19 36 1.9 60 4.87 X IQ-7 16 20 13 28 2.0 60 0 16 20 407 698 1.7 60 1.31 x w-7 16 20 66 129 2.0 60 2.55 X IQ-7 16 20 27 53 1.9 60 3.74 x w-7 16 20 18 34 1.9 60 4.87 X IQ-7 16 20 14 25 1.9 164 60 0 16 20 313 5.5 1.6 60 1.13 X 10-7 16 20 53 97 1.8 60 2.21 X 10-7 16 20 31 57 1.8 60 3.23 X 10-7 16 20 20 38 1.9 60 4.22 X 10-7 16 20 14 31 2.1 70 0 16 10 194 404 2.1 70 1.35 x 10-7 16 10 54 125 2.3 70 2.64 X 10-7 16 10 32 61 1.9 70 3.86 X 10-7 16 10 19 36 1.9 70 5.03 X 10-7 16 10 14 26 1.8 70 0 16 10 215 407 1.9 70 1.35 X 10-7 16 10 64 130 2.0 70 2.64 x 10-7 16 10 29 56 1.9 70 3.86 X 10-7 16 10 18 35 1.9 70 5.03 X 10-7 16 10 13 25 1.9 70 0 16 10 136 270 2.0 70 1.13 X 10-7 16 10 44 82 1.9 70 2.21 X 10-7 16 10 26 48 1.9 70 3.23 X 10-7 16 10 20 37 1.9 70 4.22 X 10-7 16 10 14 28 2.0 a) Temperature fDC b) [COBF]/[EMA] c) Initiator /mmol.L-1 d) Reaction time /min e) Mn x 10-3 f) Mw x 10-3 Appendix 165 A1.6 Data for C8 of COPhBF in Butyl methacrylate Table A1.6: Experimental conditions and results for the determination of C8 for COPhBF in butyl methacrylate. T a) Cone b) [AIBN] c) t d) Mne) Mwt) PDi ' 40 0 16 120 1450 2920 2.0 40 1.50 x w-7 16 120 131 262 2.0 40 2.93 x w-7 16 120 71 128 1.8 40 5.60 x w-7 16 120 38 64 1.7 40 0 16 120 1297 2290 1.8 40 1.73 x w-7 16 120 110 211 1.9 40 3.37 x w-7 16 120 50 93 1.9 40 4.93 x w-7 16 120 37 67 1.8 40 6.43 x w-7 16 120 28 55 1.9 40 0 16 120 1531 2450 1.6 40 1.73 x w-7 16 120 93 195 2.1 40 3.37 x w-7 16 120 40 100 2.5 40 4.93 x w-7 16 120 29 80 2.7 40 6.43 x w-7 16 120 24 46 1.9 50 0 16 40 909 1570 1.7 50 1.73 x w-7 16 40 81 177 2.1 50 3.37 x w-7 16 40 41 79 1.9 50 5.19 x w-7 16 40 31 56 1.8 50 6.77 x w-7 16 40 23 43 1.9 50 0 16 40 930 1700 1.8 50 1.50 x w-7 16 40 123 217 1.8 50 2.93 x w-7 16 40 63 105 1.7 50 4.30 x w-7 16 40 41 71 1.7 50 5.60 x w-7 16 40 34 47 1.5 50 0 16 40 838 1450 1.7 50 1.73 x w-7 16 40 92 179 2.0 50 3.37 x w-7 16 40 45 95 2.1 50 5.19 x w-7 16 40 31 58 1.9 50 6.77 x w-7 16 40 24 43 1.8 166 60 0 16 20 499 921 1.9 60 1.73 x 10-7 16 20 78 183 2.3 60 3.37 x 10-7 16 20 39 76 2.0 60 4.93 x 10-7 16 20 27 54 2.0 60 6.43 x 10-7 16 20 22 43 1.9 60 0 16 20 527 960 1.8 ~ 60 1.50 x 10-7 16 20 99 169 1.7 60 2.93 x 10-7 16 20 52 95 1.8 60 4.30 x 10-7 16 20 37 65 1.8 60 5.60 x 10-7 16 20 28 50 1.8 60 0 16 20 560 938 1.7 60 1.73 x 10-7 16 20 92 183 2.0 60 3.37 x 10-7 16 20 41 80 1.9 60 4.93 x 10-7 16 20 29 55 1.9 60 6.43 x 10-7 16 20 22 40 1.9 70 0 16 10 334 601 1.8 70 1.73 x 10-7 16 10 60 125 2.0 70 3.37 x 10-7 16 10 35 92 2.6 70 4.93 x 10-7 16 10 31 67 2.1 70 6.43 x 10-7 16 10 17 46 2.6 70 0 16 10 276 548 2.0 70 1.73 x 10-7 16 10 68 141 2.1 70 3.37 x 10-7 16 10 35 75 2.1 70 4.93 x 10-7 16 10 30 57 2.0 70 6.43 x 10-7 16 10 22 43 2.0 70 0 16 10 275 534 1.9 70 1.50 x 10-7 16 10 80 147 1.9 70 2.93 x 10-7 16 10 46 84 1.8 70 4.30 x 10-7 16 10 37 67 1.8 70 5.60 x 10-7 16 10 27 51 1.9 a) Temperature fCC b) [COBF]/[BMA] c) Initiator /mmol.L-1 d) Reaction time /min 3 e) Mn X 10-3 t) Mw X 10- Appendix 167 Appendix 2 A2.1 Data for the role of monomer in the chain transfer reaction in cobaloxime-mediated free-radical polymerization Table A2.1: Experimental conditions and results for role of monomer in the chain transfer reaction in cobaloxime-mediated free-radicalJ!Ol~merization Solvent Cone a) Mn Mw PDi A1 t-butyl acetate 6.29 x 10-6 741 1615 2.2 3.15 x 10-6 2207 4722 2.1 2.10 x 10-6 2635 4656 1.8 1.57 x 10-6 3696 6474 1.8 1.26 x 10-6 4382 7616 1.7 A1 t-butyl acetate 6.14 x 10-6 678 1122 1.7 repeat 3.07 x 10-6 1027 2095 2.0 2.05 x 10-6 1625 3367 2.1 1.53 x 10-6 2492 5756 2.3 1.23 x 10-6 3701 6601 1.8 B1 t-butyl acetate 6.29 x 10-6 872 2134 2.4 4.72 x 10-6 1003 2027 2.0 3.77 X 10-6 1233 2535 2.1 3.15 x 10-6 1798 3540 2.0 2.10 x 10-6 1696 3799 2.2 B1 t-butyl acetate 6.14 x 10-6 641 1232 1.9 repeat 4.60 x 10-6 902 1722 1.9 3.68 x 10-6 940 2019 2.1 3.07 x 10-6 1376 2863 2.1 2.63 x 10-6 1859 3173 1.7 A1 Toluene 5.98 x 10-6 685 1563 2.3 2.99 x 10-6 1258 2754 2.2 1.99 x 10-6 2016 4018 2.0 1.50 x 10-6 3380 6790 2.0 1.20 x 10-6 3920 7678 2.0 A1 Toluene 6.14 x 10-6 702 1416 2.0 repeat 3.07 x 10-6 1297 2652 2.0 168 2.05 x w-6 2156 4285 2.0 1.53 x w-6 3218 6523 2.0 1.23 x w-6 3829 7724 2.0 B1 Toluene 5.98 x w-6 682 1248 1.8 4.49 x w-6 893 1732 1.9 3.59 x w-6 1131 2239 2.0 2.99 x w-6 1404 2842 2.0 2.56 x w-6 1664 3393 2.0 B1 Toluene 6.14 x w-6 666 1387 2.1 repeat 4.6o x w-6 921 1889 2.1 3.68 x w-6 1216 2398 2.0 3.07 x w-6 1385 2786 2.0 2.63 x w-6 1702 3365 2.0 a) [COBF]/[MMA] Appendix 169 A2.2 Data for the determination of C8 of COBF, in the solution polymerization of MMA in toluene and tert-butyl acetate Table A2.2: Experimental conditions and results for Transfer constant (Cs) of COBF, in the solution polymerization ofMMA in toluene and tert-butyl acetate Solvent Cone a) solvent/M Mn Mw PDi Run 1 Toluene 1.21 x 10-6 50150 2590 5340 2.1 2.37 x 10-6 1570 2740 1.7 3.47 x 10-6 1130 1760 1.6 4.53 x 10-6 920 1400 1.5 5.53 x 10-6 750 1100 1.5 Run 1 Toluene 1.15 x 10-7 50150 36100 70800 2.0 3.31 x 10-7 12100 23700 2.0 4.31 x 10-7 7900 15100 1.9 5.27 x 10-7 6430 12200 1.9 Run2 Toluene 1.15 x 10-6 50150 4680 8710 1.9 2.26 x 10-6 2360 4300 1.8 3.31 x 10-6 1360 2520 1.9 4.31 x 10-6 1040 1820 1.7 5.27 x 10-6 860 1410 1.6 Run2 Toluene 1.21 x 10-7 50/50 35500 65700 1.9 3.47 x 10-7 10900 20800 1.9 4.53 x 10-7 7100 14000 2.0 5.53 x 10-7 5680 11300 2.0 Run 1 t-b-acetate 1.15 x 10-6 50150 2620 5690 2.2 2.26 x 10-6 1620 2700 1.7 3.31 x 10-6 1130 1870 1.7 4.31 x 10-6 900 1450 1.6 5.27 x 10-6 720 1110 1.5 Run2 t-b-acetate 1.15 x 10-6 50150 2690 6010 2.2 2.26 x 10-6 1630 2650 1.6 3.31 x 10-6 1060 1980 1.9 4.31 x 10-6 790 1400 1.8 5.27 x 10-6 650 1120 1.7 a) [COBF]/[MMA] 170 A2.3 Data for the determination of C8 of COPhBF, in the solution polymerization of MMA in toluene and tert-butyl acetate Table A2.3: Experimental conditions and results for Transfer constant (Cs) of COPhBF, in the solution polymerization of MMA in toluene and tert-butyl acetate ~ Solvent Conca) solvent/M Mn Mw PDi Run 1 Toluene 0 50/50 128000 243000 1.9 1.21 x 10-6 3530 6510 1.9 2.43 x 10-6 1790 3060 1.7 3.65 x 10-6 1060 1950 1.8 4.86 x 10-6 870 1530 1.8 Run 1 Toluene 0 50/50 126000 242000 1.9 1.21 x 10-7 38000 68600 1.8 2.43 x 10-7 16600 34400 2.1 3.65 x 10-7 11100 20500 1.8 4.86 x 10-7 8930 16700 1.9 Run2 Toluene 0 50/50 140000 271000 1.9 2.03 x 10-6 3460 6260 1.8 4.07 x 10-6 1390 2340 1.7 6.11 x 10-6 880 1500 1.7 8.14 x 10-6 620 1030 1.7 Run2 Toluene 0 50150 109000 190000 1.7 6.71 x 10-7 6420 13500 2.1 1.31 x 10-6 3370 6140 1.8 1.88 x 10-6 2650 4620 1.7 2.59 x 10-6 1970 3190 1.6 Run 1 t-b-acetate 1.01 x 10-6 50/50 4689 9126 1.9 2.03 x 10-6 3469 5864 1.7 3.05 x 10-6 1864 3428 1.8 4.07 x 10-6 1249 2119 1.7 5.09 x 10-6 952 1721 1.8 Run 1 t-b-acetate 1.01 x 10-7 50150 42600 78500 1.8 2.03 x 10-7 34600 64200 1.9 3.05 x 10-7 14900 30600 2.1 4.07 x 10-7 12800 25900 2.0 5.09 x 10-7 9410 19700 2.1 a) [COPhBF]/[MMA] Appendix 171 A2.4 Data for the determination of C8 of COBF, in the solution polymerization of MMA in pyridine Table A2.4: Experimental conditions and results for Transfer constant (Cs) of COBF, in the solution polymerization of MMA and styrene in pyridine. Monomer Conca) [pyridine] conv Mn Mw PDi MMA 0 0 5.3 195000 392000 2.0 1.33 x 10-7 0 4.7 15400 35100 2.3 2.60 x 10-7 0 4.2 7980 16300 2.1 3.81 x 10-7 0 4.0 6220 14500 2.3 4.96 x 10-7 0 4.0 5780 11100 1.9 MMA 0 1.4 x 10-4 5.0 256000 482000 1.9 1.33 x 10-7 1.4 x 10-4 3.2 19800 47400 2.4 2.6o x 10-7 1.4 x 10-4 1.8 9510 20500 2.2 3.81 x 10-7 1.4 x 10-4 1.2 6100 12000 2.0 4.96 x 10-7 1.4 x 10-4 0.8 4820 9550 2.0 MMA 0 2.7 x 10-4 4.8 241000 438000 1.8 1.33 x 10-7 2.7 x 10-4 2.9 26600 55600 2.1 2.60 x 10-7 2.7 x 10-4 1.5 13200 37100 2.8 3.81 x w-7 2.7 x w-4 0.6 5910 15400 2.6 4.96 x w-7 2.7 x 10-4 0.4 5170 10500 2.0 MMA 0 5.4 x 10-4 4.6 2123000 445000 2.1 1.33 x w-7 5.4 x 10-4 2.9 57400 124000 2.2 2.60 x w-7 5.4 x w-4 1.3 13500 24600 1.8 3.81 x w-7 5.4 x 10-4 0.7 8110 14800 1.8 4.96 x w-7 5.4 x w-4 0.4 5080 9470 1.9 MMA 0 o* 155000 297000 1.9 1.15 x w-7 o* 23100 39700 1.7 2.26 x 10-7 o* 10800 23000 2.1 3.31 x w-7 o* 7530 15000 2.0 4.31 x 10-7 o* 6010 12400 2.1 MMA 0 1.4 X 104 * 195000 362000 1.9 1.15 x w-7 1.4x 104 * 56700 117000 2.1 2.26 x w-7 1.4x104 * 17200 32100 1.9 3.31 x 10-7 1.4x 104 * 10080 21900 2.2 4.31 x 10-7 1.4x 104 * 6870 13700 2.0 172 MMA 0 2.9 X 104 * 171000 312000 1.8 1.15 x 10·7 2.9 X 104 * 68100 125000 1.8 2.26 x 10·7 2.9 X 104 * 24200 40200 1.7 3.31 x 10·7 2.9 X 104 * 12300 26800 2.1 4.31 x 10·7 2.9 X 104 * 6890 15600 2.3 MMA 0 4.7 X 104 * 178000 314000 1.8 1.15 x 10·7 4.7 X 104 * 61600 117000 1.9 2.26 x 10·7 4.7 X 104 * 26300 50400 1.9 3.31 x 10·7 4.7 X 104 * 13400 34600 2.6 4.31 x 10·7 4.7 X 104 * 9030 23100 2.6 Styrene 0 0 2.8 94900 167000 1.8 1.18 x 10·7 0 2.8 81500 146000 1.8 2.31 x 10·7 0 2.7 69200 128000 1.9 3.38 x 10·7 0 2.8 54800 114000 2.1 4.41 x 10·7 0 2.7 48500 97000 2.0 Styrene 0 1.4 x 10·4 2.6 87600 161000 1.8 1.18 x 10·7 1.4 x 10·4 2.5 67000 125000 1.9 2.31 x 10·7 1.4 x 10·4 2.5 48700 92100 1.9 3.38 x 10·7 1.4 x 10·4 2.5 39700 75100 1.9 4.41 x 10·7 1.4 x 10·4 2.4 35400 66100 1.9 Styrene 0 2.1 x 10·4 2.3 81300 148000 1.8 1.18 x 10·7 2.1 x 10·4 2.2 78500 142000 1.8 2.31 x 10·7 2.1 x 10·4 2.3 52400 100000 1.9 3.38 x 10·7 2.1 x 10·4 2.2 43400 80700 1.9 4.41 x 10·7 2.1 x 10·4 2.1 37200 70200 1.9 Styrene 0 10.3 x 10·4 2.2 89100 159000 1.8 1.18 x 10·7 10.3 x 10·4 2.2 86500 142000 1.6 2.31 x 10·7 10.3 x 10·4 2.1 58500 110000 1.9 3.38 x 10·7 10.3 x 10·4 2.1 55400 106000 1.9 4.41 x 10·7 10.3 X 104 2.0 38300 73900 1.9 a) [COBF]/[Mon] *) [Pyridine]/[ COBF] Appendix 173 A2.5 Data for the determination of C8 of COPhBF, in the solution polymerization of MMA in pyridine Table A2.5: Experimental conditions and results for Transfer constant (Cs) of COPhBF, in the solution polymerization of MMA and styrene in pyridine. Monomer Cone a) [pyridine] conv Mn Mw PDi MMA 0 0 5.7 256000 492000 1.9 1.08 x 10·7 0 5.5 194000 348000 1.8 2.15 x 10·7 0 5.1 56600 108000 1.9 3.22 x 10·7 0 4.8 26300 44700 1.7 4.30 x 10·7 0 4.6 13300 25400 1.9 MMA 0 1.4 x 10·4 6.3 270000 480000 1.8 1.08 x 10·7 1.4 x 10·4 5.6 215000 390000 1.8 2.15 x 10·7 1.4 x 10·4 4.5 102000 204000 2.0 3.22 x 10·7 1.4 x 10·4 3.8 61500 117000 1.9 4.3o x 10·7 1.4 x 10·4 3.2 47000 83800 1.8 MMA 0 5.4 x 10·4 5.2 261000 506000 1.9 1.08 x 10·7 5.4 x 10·4 4.9 262000 424000 1.6 2.15 x 10·7 5.4 x 10·4 3.8 152000 296000 2.0 3.22 x 10·7 5.4 x 10·4 3.1 109000 211000 1.9 4.30 x 10-7 5.4 x 10-4 2.5 87200 161000 1.9 MMA 0 10.3 x 10·4 5.0 242000 460000 1.9 1.08 x 10·7 10.3 x 10·4 4.5 233000 409000 1.8 2.15 x 10-7 10.3 x 10-4 3.7 209000 363000 1.7 3.22 x 10-7 10.3 x 10-4 3.0 155000 288000 1.9 4.30 x 10-7 10.3 x 10-4 2.8 130000 247000 1.9 Styrene 0 0 3.3 94800 165000 1.7 1.31 x 10-7 0 3.2 84700 149000 1.8 2.61 x 10-7 0 3.2 81300 143000 1.8 3.92 x 10·7 0 3.2 80200 144000 1.8 5.23 x 10·7 0 3.2 79100 141000 1.8 Styrene 0 1.4 x 10·4 2.9 92700 161000 1.7 1.31 x 10-7 1.4 x 10-4 2.8 81400 152000 1.9 2.61 x 10-7 1.4 x 10-4 2.8 81800 145000 1.7 3.92 x 10-1 1.4 x 10-4 3.0 76800 138000 1.7 5.23 x 10-7 1.4 x 10-~ 2.8 74400 133000 1.8 174 Styrene 0 2.1 x 10-4 2.9 96100 168000 1.7 1.31 x 10-7 2.1 x 10-4 2.8 88500 154000 1.7 2.61 x 10-7 2.1 x 10-4 2.8 84700 150000 1.7 3.92 x 10-7 2.1 x 10-4 3.0 80700 146000 1.8 Styrene 0 10.3 x 10-4 2.9 91000 159000 1.7 1.31 X 10-7 10.3 x 10-4 2.8 86400 153000 1.8 2.61 x 10-7 10.3 x 10-4 2.8 85300 149000 1.8 3.92 x 10-7 10.3 x 10-4 3.0 83400 150000 1.8 5.23 x 10-7 10.3 x 10-4 2.8 83100 148000 1.8 a) [COPhBF]/[Mon] Appendix 175 Appendix 3 A3.1 Data for C8 of COBF in styrene at 60°C Table A3.1: Experimental conditions and results for the determination of C8 for COBF in styrene Vs conversion at 60°C. t a) Cone b) [AIBN] c) Convd) Mne) Mwt) slope g) PDi 15 2.43 10 0.50 10 23 -9.42E-05 2.2 15 3.56 10 0.49 6.5 15 -1.32E-04 2.3 15 4.64 10 0.45 5.0 12 -1.81E-04 2.3 15 5.67 10 0.43 4.1 9.5 -2.15E-04 2.3 30 2.43 10 1.75 23 65 -2.88E-05 2.8 30 3.56 10 1.66 12 46 -3.61E-05 3.8 30 4.64 10 1.60 11 41 -3.91E-05 3.7 30 5.67 10 1.53 8.7 34 -4.45E-05 3.9 60 2.43 10 2.84 56 144 -1.40E-05 2.6 60 3.56 10 2.87 27 141 -1.40E-05 5.3 60 4.64 10 2.77 24 139 -1.39E-05 5.7 60 5.67 10 2.72 21 129 -1.47E-05 6.3 90 2.43 10 4.70 44 131 -1.47E-05 3.0 90 3.56 10 4.64 32 122 -1.49E-05 3.8 90 4.64 10 4.52 22 102 -1.69E-05 4.6 90 5.67 10 4.50 18 102 -1.64E-05 5.8 200 2.43 10 11.04 56 144 -1.40E-05 2.6 200 3.56 10 11.02 27 141 -1.40E-05 5.3 200 4.64 10 10.74 24 139 -1.39E-05 5.7 200 5.67 10 10.72 21 129 -1.47E-05 6.3 a) Time (min) b) [COBF]/[styrene] x 106 c) Initiator /mmol.L-1 d) Conversion measured by gravimetry 3 e) Mn X 10-3 f) Mw x 10- g) slope for In P(M) vs M plot 176 A3.2 Data for C8 of COPhBF in styrene Table A3.2: Experimental conditions and results for the determination of C8 for COPhBF in styrene Vs conversion at 60°C. t a) Cone b) [AIBN] c) Convd) Mne) Mwt) slope g) PDi 10 2.74 10 0.37 39 71 -2.97 1.9 10 4.01 10 0.34 28 51 -4.04 1.8 10 5.23 10 0.27 20 40 -5.02 2.0 10 6.39 10 0.23 16 31 -6.39 1.9 20 2.74 10 0.82 50 95 -2.39 1.9 20 4.01 10 0.73 40 72 -3.00 1.8 20 5.23 10 0.66 26 55 -3.57 2.1 20 6.39 10 0.65 21 46 -4.15 2.2 30 2.58 10 1.64 61 110 -1.90 1.8 30 3.78 10 1.56 47 92 -2.15 1.9 30 4.93 10 1.50 45 87 -2.32 1.9 30 6.03 10 1.44 24 70 -2.57 2.9 40 2.74 10 1.91 67 127 -1.77 1.9 40 4.01 10 1.84 52 110 -2.03 2.1 40 5.23 10 1.78 43 97 -2.17 2.2 40 6.39 10 1.71 35 88 -2.31 2.5 53 2.74 10 2.54 71 139 -1.52 2.0 53 4.01 10 2.38 52 112 -1.79 2.2 53 5.23 10 2.32 46 110 -1.83 2.4 53 6.39 10 2.29 40 101 -1.90 2.5 63 2.74 10 2.94 60 139 -1.46 2.3 63 4.01 10 2.88 61 128 -1.58 2.1 63 5.23 10 2.76 48 114 -1.73 2.4 63 6.39 10 2.72 44 107 -1.79 2.4 70 2.58 10 3.56 75 151 -1.45 2.0 70 3.78 10 3.52 74 142 -1.52 1.9 70 4.93 10 3.46 58 129 -1.57 2.2 70 6.03 10 3.32 55 118 -1.67 2.1 95 2.58 10 5.71 88 180 -1.36 2.0 95 3.78 10 5.69 75 169 -1.39 2.2 Appendix 177 95 4.93 10 5.60 72 163 -1.41 2.3 95 6.03 10 5.55 77 152 -1.47 2.0 130 2.58 10 7.57 93 199 -1.31 2.1 130 3.78 10 7.46 91 183 -1.34 2.0 130 4.93 10 7.30 85 183 -1.35 2.2 130 6.03 10 7.10 73 172 -1.37 2.4 ~ 200 2.58 10 11.04 97 196 -1.24 2.0 200 3.78 10 11.02 95 185 -1.30 2.0 200 4.93 10 10.74 89 182 -1.32 2.1 200 6.03 10 10.72 84 180 -1.35 2.2 a) Time (min) b) [COPhBF]/[styrene] x 106 c) Initiator /mmoi.L-1 d) Conversion measured by gravimetry 3 e) Mn X 10-3 f) Mw X 10- 5 g) slope for In P(M) vs M plot (x 10 ) 178 A3.3 Data for C8 of COBF in styrene at 40°C Table A3.3: Experimental conditions and results for the determination of C8 for COBF in styrene Vs conversion at 40°C. t a) Cone b) [AIBN] c) Convd) Mne) MwD slope g) PDi 63 2.67 10 0.56 7.5 16 -11.5 2.2 63 3.92 10 0.49 4.5 10 -17.1 2.3 63 5.11 10 0.40 3.3 7.0 -25.4 2.1 63 6.24 10 0.31 2.0. 5.0 -34.0 2.4 95 2.67 10 1.93 8.7 22 -9.32 2.5 95 3.92 10 1.69 4.7 13 -14.3 2.7 95 5.11 10 1.53 3.7 9.5 -18.7 2.6 95 6.24 10 1.41 2.9 7.2 -24.2 2.5 130 2.67 10 2.77 8.4 23 -9.09 2.8 130 3.92 10 2.48 6.1 15 -12.7 2.4 130 5.11 10 2.22 4.3 10 -16.8 2.4 130 6.24 10 1.92 3.3 8.1 -21.4 2.5 1000 2.67 10 8.55 3.3 188 -0.55 5.6 1000 3.92 10 8.44 2.4 175 -0.56 7.3 1000 5.11 10 8.50 1.7 154 -0.61 8.8 1000 6.24 10 8.37 1.5 151 -0.62 9.9 a) Time (min) b) [COBF]/[styrene] x 106 c) Initiator /mmoi.L-1 d) Conversion measured by gravimetry 3 e) Mn X 10-3 f) Mw x 10- 5 g) slope for In P(M) vs M plot (x 10 ) Appendix 179 A3.4 Data for C8 of COBF in styrene with varying temperature Table A3.4: Experimental conditions and results for the determination of C8 for COBF in st~rene versus tem~erature Ta) t b) Cone c) Convd) Moe) MwO slope g) PDi 39 1483 1.25 5.60 28 251 -0.425 9.0 39 1483 2.43 5.29 48 224 -0.500 4.7 39 1483 3.56 5.14 25 190 -0.514 7.7 39 1483 4.64 5.02 19 190 -0.497 9.8 39 1483 5.67 4.88 22 '193 -0.508 8.9 39 1482 1.25 5.50 32 271 -0.393 8.5 39 1482 2.43 5.28 38 211 -0.497 5.5 39 1482 3.56 5.02 26 208 -0.495 8.1 39 1482 4.64 4.92 25 214 -0.480 8.6 39 1482 5.67 4.56 22 185 -0.509 8.5 39.2 1500 1.25 6.05 61 247 -0.532 4.1 39.2 1500 2.43 5.86 39 148 -0.712 3.8 39.2 1500 3.56 5.79 26 124 -0.759 4.8 39.2 1500 4.64 5.62 24 117. -0.798 4.8 39.2 1500 5.67 5.44 20 94 -0.892 4.6 39.2 1500 1.25 5.89 61 3.26 -0.488 5.3 39.2 1500 2.43 5.68 42 1.65 -0.632 3.9 39.2 1500 3.56 5.75 30 1.40 -0.703 4.7 39.2 1500 4.64 5.61 23 1.17 -0.759 5.1 39.2 1500 5.67 5.73 31 1.10 -0.848 3.5 60 134 1.18 4.61 54 203 -0.812 3.8 60 134 2.31 4.56 43 182 -0.857 4.3 60 134 3.38 4.49 47 175 -0.899 3.7 60 134 4.41 4.52 36 165 -0.918 4.6 60 134 5.39 4.49 35 159 -0.936 4.5 60 134 1.18 4.47 72 203 -0.809 2.8 60 134 2.31 4.40 70 187 -0.845 2.7 60 134 3.38 4.30 45 171 -0.878 3.7 60 134 4.41 4.36 37 161 -0.891 4.4 60 150 1.25 5.69 57 202 -0.848 3.6 60 150 2.43 5.61 45 180 -0.915 4.0 180 60 150 3.56 5.54 47 165 -0.972 3.5 60 150 4.64 5.47 33 152 -1.01 4.6 60 150 5.67 5.45 36 146 -1.05 4.1 60 150 1.25 5.58 59 207 -0.825 3.5 60 150 2.43 5.50 53 180 -0.904 3.4 60 150 3.56 5.49 42 166 -0.942 4.0 60 150 4.64 5.52 41 162 -0.966 3.9 60 150 5.67 5.46 32 144 -1.04 4.5 80.4 22 1.31 5.48 41 134 -2.03 3.3 80.4 22 2.55 5.43 37 105 -2.11 2.9 80.4 22 3.74 5.45 31 101 -2.18 3.3 80.4 22 4.87 5.37 28 93 -2.24 3.3 80.4 22 5.96 5.38 28 98 3.5 80.4 22 1.31 4.77 35 107 -2.07 3.1 80.4 22 2.55 4.73 35 103 -2.15 3.0 80.4 22 3.74 4.71 28 90 -2.23 3.2 80.4 22 4.87 4.70 27 92 -2.25 3.4 80.4 22 5.96 4.71 22 84 -2.32 3.8 80.9 22 1.25 5.34 39 98 -2.09 2.5 80.9 22 2.43 5.33 36 93 -2.13 2.6 80.9 22 3.56 5.33 30 89 -2.18 2.9 80.9 22 4.64 5.32 32 86 -2.24 2.7 80.9 22 5.67 5.34 28 84 3.0 81 22 1.25 5.32 40 97 -2.11 2.4 81 22 2.43 5.00 42 95 -2.18 2.3 81 22 3.56 5.01 32 86 -2.23 2.7 81 22 4.64 4.98 30 84 -2.26 2.8 81 22 5.67 4.94 29 83 -2.31 2.9 100 70 1.31 5.95 45 193 -1.24 4.3 100 70 2.55 5.55 30 111 -1.87 3.8 100 70 3.74 5.57 34 106 -2.22 3.1 100 70 4.87 5.64 29 98 -2.48 3.4 100 70 5.96 5.47 21 62 -3.03 3.0 100 70 1.31 5.60 46 187 -1.41 4.0 100 70 2.55 5.25 30 111 -1.88 3.7 100 70 3.74 5.34 27 86 -2.40 3.1 Appendix 181 100 70 4.87 5.36 27 81 -2.97 3.0 100 70 5.96 5.31 21 70 -3.36 3.4 100 70 1.25 5.87 56 177 -1.35 3.2 100 70 2.43 5.62 39 120 -1.86 3.1 100 70 3.56 5.65 31 93 -2.10 3.0 100 70 4.64 5.63 28 80 -2.53 2.8 100 70 5.67 5.40 22 64 -3.12 2.9 100 70 1.25 5.99 43 192 -1.33 4.5 100 70 2.43 5.75 36 117 -1.74 3.2 100 70 3.56 5.61 35 104 -2.36 3.0 100 70 4.64 5.53 29 85 -2.89 2.9 100 70 5.67 5.38 26 71 -3.43 2.8 120 16 1.31 6.13 39 99 -2.66 2.5 120 16 2.55 4.95 19 49 -4.24 2.5 120 16 3.74 5.52 16 35 -5.54 2.2 120 16 4.87 5.30 14 51 -6.99 3.6 120 16 5.96 5.37 13 65 -8.01 5.0 120 16 1.31 6.26 32 97 -2.26 3.0 120 16 2.55 6.25 21 53 -3.85 2.5 120 16 3.74 6.11 20 40 -5.02 2.0 120 16 4.87 6.15 16 50 -6.21 3.1 120 16 5.96 5.50 14 28 -6.99 2.0 120 16 1.31 6.13 41 108 -2.72 2.6 120 16 2.55 6.11 24 67 -4.57 2.7 120 16 3.74 6.18 19 75 -6.14 3.9 120 16 4.87 5.38 14 48 -7.73 3.4 120 16 5.96 5.01 12 48 -8.91 3.9 120 16 1.31 5.79 39 92 -3.00 2.3 120 16 2.55 5.71 20 54 -5.00 2.6 120 16 3.74 5.81 16 59 -6.80 3.6 120 16 4.87 5.19 13 41 -8.56 3.2 120 16 5.96 5.33 12 46 -9.25 4.0 a) Temperature /°C b) Time (Min) c) [COBF]/[styrene] x 106 d) Conversion measured by gravimetry 3 e) Mn X 10-3 f) Mw X 10- 5 g) slope for In P(M) vs M plot (x 10 ) 182 A3.5 Data for C8 of COPbBF in styrene with varying temperature Table A3.5: Experimental conditions and results for the determination of C8 for COPhBF in st~rene versus teml!erature. ; Ta) t b) Cone c) Convd) Mne) Mwf) slope g) PDi 39 1498 1.16 5.90 216 466 2.2 39 1498 2.27 5.57 144 398 -0.328 2.8 39 1498 3.33 5.54 53 354 -0.345 6.7 39 1498 4.33 5.61 89 339 -0.364 3.9 39 1498 5.30 5.34 83 329 3.9 39 1498 1.16 6.01 179 444 -0.308 2.5 39 1498 2.27 5.85 145 387 -0.326 2.7 39 1498 3.33 5.73 133 385 -0.323 2.9 39 1498 4.33 5.50 80 302 -0.357 3.8 39 1498 5.30 5.07 42 299 -0.393 7.0 39.1 1500 1.20 5.87 50 370 -0.341 7.4 39.1 1500 2.35 5.29 61 328 -0.362 5.4 39.1 1500 3.44 5.64 41 319 -0.350 7.8 39.1 1500 4.48 5.21 49 303 -0.380 6.1 39.1 1500 5.48 4.66 57 276 -0.420 4.9 39.1 1500 1.20 6.05 82 390 -0.338 4.7 39.1 1500 2.35 5.48 47 345 -0.348 7.4 39.1 1500 3.44 5.05 79 325 -0.381 4.1 39.1 1500 4.48 4.30 48 186 -0.44 3.9 39.1 1500 5.48 4.38 42 260 6.2 60 150 1.16 5.46 53 237 -0.761 4.5 60 150 2.27 5.68 67 242 -0.793 3.6 60 150 3.33 5.52 66 225 -0.814 3.4 60 150 4.33 5.42 70 217 -0.830 3.1 60 150 5.30 5.57 67 217 -0.841 3.2 60 150 1.16 5.25 84 248 -0.768 2.9 60 150 2.27 5.40 66 228 -0.793 3.5 60 150 3.33 5.39 59 215 -0.825 3.7 60 150 4.33 5.50 63 207 -0.849 3.3 Appendix 183 60 150 5.30 5.16 63 209 -0.850 3.3 60 150 1.20 5.23 79 265 -0.705 3.3 60 150 2.35 5.18 84 249 -0.746 3.0 60 150 3.44 4.87 77 253 -0.746 3.3 60 150 4.48 5.04 78 238 -0.756 3.0 60 150 5.48 4.90 61 222 -0.795 3.6 60 150 1.20 5.08 95 284 3.0 60 150 2.35 4.10 70 252 -0.725 3.6 60 150 3.44 5.35 88 253 -0.738 2.9 60 150 4.48 4.98 85 234 -0.772 2.8 60 150 5.48 4.79 65 226 -0.781 3.5 80 22 1.20 5.31 45 109 -1.79 2.4 80 22 2.35 5.50 48 113 -1.85 2.4 80 22 3.44 5.26 42 103 -1.87 2.5 80 22 4.48 5.19 46 101 -1.91 2.2 80 22 5.48 5.23 39 100 -1.90 2.5 80 22 1.20 5.25 48 108 -1.83 2.2 80 22 2.35 5.21 44 103 -1.88 2.3 80 22 3.44 5.13 42 101 -1.91 2.4 80 22 4.48 5.13 41 99 -1.94 2.4 80 22 5.48 5.16 41 97 -1.96 2.4 80 22 1.16 6.52 50 118 -1.93 2.4 80 22 2.27 6.46 49 111 -2.00 2.2 80 22 3.33 6.53 45 109 -2.01 2.4 80 22 4.33 6.58 36 102 -2.04 2.8 80 22 5.30 6.50 41 100 -2.09 2.4 80 22 1.16 5.76 43 118 -1.92 2.7 80 22 2.27 5.68 47 115 -1.95 2.4 80 22 3.33 5.66 46 114 -1.98 2.5 80 22 4.33 5.69 47 119 -1.99 2.5 80 22 5.30 5.65 48 115 -2.01 2.4 100 39 1.24 4.78 50 355 -0.509 7.1 100 39 2.43 5.27 61 284 -0.654 4.6 100 39 3.55 5.86 41 235 -0.747 5.7 100 39 4.63 6.04 44 208 -0.843 4.7 100 39 5.66 6.12 55 199 -0.911 3.6 184 100 39 1.24 4.87 47 307 -0.534 6.5 100 39 2.43 5.30 56 298 -0.593 5.3 100 39 3.55 5.39 45 248 -0.687 5.5 100 39 4.63 5.95 37 206 -0.764 5.6 100 39 5.66 6.04 50 212 -0.827 4.2 100 39 1.20 4.74 33 317 -0.494 9.7 100 39 2.35 5.28 43 295 -0.572 6.8 100 39 3.44 5.28 63 270 -0.655 4.3 100 39 4.48 5.78 41 229 -0.729 5.6 100 39 5.48 5.89 26 199 -0.787 7.5 100 39 1.20 4.96 73 360 -0.481 4.9 100 39 2.35 5.48 45 290 -0.575 6.4 100 39 3.44 5.80 51 251 -0.686 5.0 100 39 4.48 5.80 41 219 -0.780 5.3 100 39 5.48 6.14 42 137 3.3 120 16 1.20 6.99 52 221 -0.812 4.3 120 16 2.35 7.97 66 208 -0.970 3.2 120 16 3.44 7.20 60 189 -1.11 3.2 120 16 4.48 8.95 67 173 -1.27 2.6 120 16 5.48 9.28 42 156 -1.34 3.7 120 16 1.20 7.69 73 232 -0.814 3.2 120 16 2.35 8.43 76 205 -0.975 2.7 120 16 3.44 8.58 60 185 -1.11 3.1 120 16 4.48 9.14 61 157 -1.27 2.6 120 16 5.48 9.21 56 152 -1.33 2.7 120 16 1.20 7.30 66 209 -0.850 3.2 120 16 2.35 6.60 57 203 -0.917 3.5 120 16 3.44 8.11 70 178 -1.15 2.5 120 16 4.48 9.49 63 172 -1.29 2.7 120 16 5.48 9.31 56 153 -1.39 2.7 120 16 1.20 7.47 80 223 -0.857 2.8 120 16 2.35 8.01 69 198 -1.00 2.9 120 16 3.44 8.93 73 185 -1.16 2.5 120 16 4.48 9.30 61 152 -1.29 2.5 120 16 5.48 9.90 66 163 -1.38 2.5 a) Temperature /°C b) Time (Min) c) [COBF]/[styrene] x 106 d) Conversion measured by gravimetry Appendix 185 3 e) Mn X w-3 f) Mw x 10- 5 g) slope for ln P(M) vs M plot (x 10 ) A3.6 Deravation of the model of Cobalt(II) concentration used in Chapter6 <' In this appendix, an expression for the loss of active cobalt(ll) from the system due the formation of cobalt carbon bonds which dominates in the CCTP of styrene. The derivation is based on the reaction scheme given in chapter 6 (Scheme 6.1) - d[Co(II)] =k ·[Co(II)]-[R*]-k ·[Co(ll)-R] (A.1) dt I 2 - [Co(ll)- R] = [Co(II)]0 [Co(II)] (A.2) By substituting equation A.2 into A.1 we obtain: d[Co(II)] • - dt = k 1 ·[Co(II)}R ]-k2 ·[Co(II)]0 +k2 ·[Co(II)] rearranging gives: 186 intergrating for [Co(II)] between timet and 0 and rearranging gives: [Co(ll)]·(k,[R"]+k2 )-k2 ·[Co(ll)]} ( [ •] ) ln [ ] ( [ ] ) [ ] = -t k 1 R + k 2 { • • Co(ll) 0 k 1 R" +k2 -k2 Co(ll)