Kumada catalyst transfer polycondensation for controlled synthesis of polyfluorenes using 1,3-bis(diarylphosphino)propanes as ligands

Three new bis(diphenylphosphino) propane (dppp) derivatives, i.e., 1,3-bis(di(4-methylphenyl)phosphino)propane (L1), 1,3-bis(di(3-methylphenyl)phosphino)propane (L2) and 1,3-bis(di(2-methylphenyl)phosphino)propane (L3), were synthesized for studying the effect of the ligand on the Kumada catalyst transfer polycondensation (KCTP) of 7-bromo-9,9-dialkylfluorenylmagnesium chloride (M1) using Ni(acac)(2)/L (L = dppp, L1, L2 or L3) as the catalyst. Ni(acac)(2)/L1 exhibited a polymerization performance comparable to Ni(acac)(2)/dppp, and the polymerization could be well-controlled with >= 1 mol% catalyst. When the most hindered ligand L3 was used, the polymerization was out-of-control and only poly(9,9-dioctylfluorene)s (PF8s) with relatively low number-average molecular weights (M-n) and large polydispersity indices (PDIs) could be obtained. In contrast, the moderately hindered catalyst Ni(acac)(2)/L2 outperformed Ni(acac)(2)/dppp. The polymerization was well-controlled with a catalyst of >= 0.5 mol%, as confirmed by the linear correlation of M-n versus [monomer]/[Ni], and PF8s with M-n up to 91.1 kDa and fluorene-thiophene block copolymers (PF8-b-P3HTs) with M-n up to 78.4 kDa could be prepared in a controlled manner. Density functional theory (DFT) calculations, P-31 NMR spectroscopy studies and block copolymerizations revealed that the great performance of L2 can be attributed to the stronger affinity of the L2Ni(0)-polymer p-complex and the higher stability of active chain ends. The active chain ends of the high molecular weight polymers were unstable, probably owing to unknown side reactions assisted by chain aggregation or entanglement, which may be responsible for the out-of-control nature of the polymerization at low catalyst loadings.