(239a) Cyclic Polyethylene Furanoate As a Monomer from Renewable Resources for Ring Opening Polymerization Conference: AIChE Annual MeetingYear: 2016Proceeding: 2016 AIChE Annual MeetingGroup: Process Development DivisionSession: Sustainable and Green Product Design Time: Monday, November 14, 2016 - 3:15pm-3:36pm Authors: Fleckenstein, P., ETH Zürich Storti, G., ETH Zurich Morbidelli, M., ETH Zürich Introduction The replacement of oil based chemicals by renewable resource based chemicals is one of the major issues chemical industry is facing right now. The alternative polymer poly (ethylene furanoate), PEF, is a possible substitute for poly (ethylene terephthalate), PET. Among different process alternatives, PEF can be produced at high polymerization rates and molecular weights via ring opening polymerization of cyclic oligo (ethylene furanoate), cyOEF. On the other hand, the effective production of these cyclic oligomers at high purity is still the major obstacle towards the process industrialization. In this work, different strategies aimed to produce these cyclic molecules from the two monomers Furan dicarboxylic acid (FDCA) and ethylene glycol (EG) are explored. And a model is presented to describe the equilibrium behavior of said cyclic oligomers. Three pathways towards cyOEF The three synthetic strategies shown in figure 1 have been investigated. In all cases, high dilutions in organic solvents are required to achieve high yields; the final product is a distribution of cyclic oligomers with different number of repeating units. Figure 1 Reaction pathways for the production of cyOEF. The first (scheme 1-2-6) is a fast cycle formation using the rapid esterification of acid chlorides with alcohols. Moderate selectivity towards cycles of around 50% are in contrast to a complex reaction setup and toxic chemicals involved for the chlorination. The product size distribution is kinetically controlled and consists mainly of C3, with C2 and C4 being the second most abundant species. The depolymerization (scheme 1-3-4-6) and reactive distillation (scheme 1-3-5-6) are the two more promising routes towards cycles. For depolymerization FDCA and EG are converted to oligomeric chains of around 12 repeating units. This prepolymer is subsequently diluted in a high boiling organic solvent and converted to a mixture of cyclics and linears at 180-200 °C. In case of reactive distillation the FDCA and EG (in excess) are converted to very short chains (1-2 repeating units). The chains are then diluted in a high boiling organic solvent and finally converted to cyOEF by distilling a portion of the solvent together with the excess EG. Both routes yield a similar distribution of cyclic species. Possible purification routes include precipitation and adsorption processes, which have both been applied to yield cycles with high purity starting from linear species (>99% by HPLC area). The yield of both reactions lies in the range of 60 to 80 % depending on the process conditions. One striking feature of this type of reaction is that it is possible to recycle the solvent and the linear side products in order to increase the yield of the process and furthermore make it more sustainable. Modeling of the ring chain equilibrium The equilibrium composition of cyclics and linears in 2-Methylnaphthalene and Dichlorobenzene (DCB) could be modeled using the Jacobson-Stockmayer model (Figure 2). Under high dilutions this equilibrium shifts towards the formation of cyclic oligomers. It is based on computing the probability of ring closure using the assumption of randomly coiled hydrocarbon chains for linear polymers. Two fitting parameters are involved: one is the equilibrium constant for the step growth polymerization, which is in equilibrium with the cyclisation reactions, and the second one is the effective length of the bonds in the polymer chains. As shown in the figure below, our own results from depolymerization and reactive distillation experiments were fitted using the model. These results suggest that it is indeed possible to use this approach to describe the ring chain equilibrium. The model is currently under refinement in order to describe system-specific effects such as the complete absence of cycles with six repeating units. Figure 2 Equilibrium distribution of cyclic oligomers in DCB. References  Gomes M, Gandini A, J. Polym. Sci., Part A: Polym. Chem., 49, 3759-3768 (2011).  Burch, RR, Lustig, SR, Spinu, M Macromolecules, 33, 5053-5064 (2000).  Stockmayer, WH, Jacobson H, J. Chem. Phys., 18, 1600-1606 (1950).