Aiden E Burgess, Fariborz Shaahmadi, Andries J. Burger, Jamie T. Cripwell
The monomer fraction, X , or the molecular fraction of a component’s molecules not participating in hydrogen bonding, is an explicitly calculable quantity in the SAFT thermodynamic framework. Researchers have long proposed the measurement of X using FTIR spectroscopy and incorporating these data into parameterisation algorithms to overcome parameter degeneracy issues, particularly for polar, solvating species. This work looks to address the lack of monomer fraction data, and indeed the lack of an unbiased approach to its generation, by drawing on spectral processing techniques well-established in the spectroscopic literature. Multivariate Curve Resolution with Alternating Least Squares (MCR-ALS) is investigated as an approach to generate monomer fraction data in solvating mixtures. MCR-ALS resolves spectra for different compositions into statistically significant spectral and concentration profiles, which can be physically interpreted to justify a full range of hydrogen bond interactions, while approximating monomer fraction under appropriate assumptions. This overcomes the inherent bias and spectral ambiguity associated with curve fitting approaches used in previous thermodynamic studies of monomer fractions. The method was benchmarked against published ethanol/n-hexane data before application to ethanol/acetone. An apparent acetone monomer/free-carbonyl fraction was estimated across the composition range, while the expected low fraction of ethanol monomers could not be elucidated from the instrument noise and signal processing. An initial assessment of polar-SAFT models to simultaneously predict phase equilibrium and monomer fraction with previously published parameter sets highlighted valid approaches for incorporating these data in model parameterisation and development in future work.