1.3. Catalytic Chain Transfer Polymerisation
1.3.3. CCTP – Catalysts
Effective catalytic chain transfer agents are invariably low-spin Co(II) complexes with octahedral geometry, derived from a square planar ligand and two available axial sites.155 Co(II) is d7with the potential to exist in either a high spin, with three unpaired electrons, or low spin conformation, with one unpaired electron (Figure 1.16). Interestingly, this is an important consideration in designing a catalytic chain transfer agent, as the nature of the ligands surrounding the central cobalt can easily result in spin crossover from low spin to high spin, as the relative energy levels are close together; to date no reason has been found for this empirical observation.
Figure 1.16: Electronic configuration of Co(II) complexes in both low (left) and high spin (right) configurations
On first discovery of CCTP the Co(II) complexes were based on porphyrin complexes e.g. Co(II) hematoporphyrin IX tetramethyl ether (1). This first generation of CTA’s produced Csvalues of 2.4 E3,156 but due to the expensive isolation and purification of
porphyrins, the cost was considered too high for application of CCTP in industry. Other disadvantages included their high colour, and insolubility in a range of solvents, hence alternative catalysts were required. Based on the criteria for active CCTP catalysts outlined above, cobaloximes were investigated.
The second generation of Co(II) complexes were cobaloximes, (2), which are an order of magnitude more active than porphyrins (Cs 2.0E4 - 2.3E3), with the added benefit
that the synthesis is less costly, they are low in colour compared to porphyrins, have good solubility and their properties can be tuned through the axial and equatorial ligands, in turn affecting the chain transfer constant, which has also been shown to increase the stability of catalyst solutions to oxidation in air.122, 157-159A disadvantage to this generation of cobaloximes was they are susceptible to hydrolysis and oxidation; hence the third generation of cobaloximes incorporates a BF2 bridge, which stabilizes
the cobaloxime towards these reactions (Cs4 E4).158 These complexes were given the
general name CoBF, which is commonly used to describe CTA’s of general structure3.
1.3.3.1. Measuring Catalyst Activity
The most commonly used measure of catalytic activity, and a commonly used indicator of catalyst purity, is measurement of the Cs, defined as the ratio of the rate constant for
chain transfer to the rate constant for propagation (ktr,s/kp). Csvalues for conventional
chain transfer agents, such as mercaptans, are an order of magnitude of 1 - 10 for methacrylates, in contrast, Cs values for cobaloximes can be four orders of magnitude
greater, and are not consumed during the reaction. This makes them highly effective catalysts, hence only ppm amounts are often required to achieve large reductions in molecular weight.127
Typically Cs values are determined using variations of the Mayo equation as described
in Equation 1.5, from which a linear Mayo plot of the 1/DPn vs. [S]/[M] can be plotted,
the gradient of which is a measure of the Cs value for that particular chain transfer
agent under the conditions used in the polymerisation. The intercept gives the 1/DPn0,
defined as the DPnin the absence of chain transfer agent.
0
1
1
[ ]
[
]
s n nS
C
DP
DP
M
Equation 1.5: General form of the Mayo equation, whereby DPn is the number average degree of
polymerisation of the polymer, DPn0is the number average degree of polymerization in the absence
of chain transfer agent, and [S] and [M] are the concentrations of chain transfer agent and monomer respectively.
Polymerisations carried out in order to obtain Mayo plots should ideally be conducted under the same conditions, preferably using more than three chain transfer agent concentrations, and should be terminated at low conversions in order to ensure that [S]/[M] remains constant throughout the polymerisation. DPn can be calculated in one
of two ways, either by division of the Mn, usually obtained by size exclusion
chromatography (SEC), by the monomer mass or division of the Mw by two times the
monomer mass. Use of Mw as a measure of DPn is only justified when the
polymerisation is fully controlled by chain transfer, hence a PDi value of approximately two is obtained, however, in general practice this method has proved more accurate than measurements based on Mn as it is less susceptible to baseline deviations.127, 160, 161
The Cscan also be measured by chain length distribution (CLD), also described as
molecular weight distribution, which can be obtained directly via SEC.127, 158, 160, 162 The CLD of a polymer contains a history of the kinetic events which have occurred throughout the polymerisation. Information on chain transfer kinetics can be readily extracted from the CLD of a polymer. Equation 1.6 was derived by Gilbert et al. differentiation of which gives the limiting slope of Pivs.i (Equation 1.7)162
lim
→ஶܲ∝݁ݔቆ
݇௧,ெ[ܯ] +݇௧,௦[ܵ]
݇[ܯ] ݅ቇ
Equation 1.6: Instantaneous CLD, Pi, obtained by SEC whereby ktr,M is the transfer rate to
monomer, ktr,Sis the transfer rate to CTA and kpis the rate of propagation ݀(݈݊ܲ) ݀݅ = −ቆ ݇௧,ெ[ܯ] +݇௧,ௌ[ܵ] ݇[ܯ] ቇ= −൬ܥெ +ܥௌ [ܵ] [ܯ]൰ Equation 1.7: Limiting slope of Pivs. i, where CMis chain transfer to monomer.
The slope of this can be represented by Equation 1.8 as Pi= (1-T) Ti-1:
݀(݈݊ܲ)
݀݅ =݈݊ܶ
Where,
ܶ=ܴ ܴ
+ܴ௧+ܴ௧
Equation 1.9: Rp, Rtrand Rtare the rates of propagation, transfer and termination respectively
Hence, the slope of a plot of lnPivs. i gives lnT. From this the transfer constant can be
obtained from the slope of the plot lnT vs. [S]/[M], preferably for more than three concentrations of chain transfer catalyst, where the intercept will be 1/DPn0.163
An advantage to using this method for estimation of the Csis that essential information
can be obtained by analysis of a small portion of the CLD, providing high molecular weight data is used. As the whole CLD is not analysed this method is less sensitive to noise, poor baseline selection and the presence of artefacts in the SEC spectra. The catalytic activity of Co(II) complexes can vary dramatically depending on a number of factors such as sterics and electronics of ligands, solvent, viscosity, monomer and impurities.155, 156, 164-167 As it is difficult to measure purity of catalytic chain transfer agents, often measurement of Csis the most reliable measure.