Computer-aided solvent screening

In a helpful approach, Abildskov et al. described a computer-aided solvent screening methodology for biocatalytic systems that used group contribution methods for calculating the constrained properties related to chemical reaction equilibrium, substrate and product solubility, water solubility, boiling points, toxicity, and others. 

Abildskov J, Leeuwen MB, Boeriu CG, van den Broek LAM. Computer-aided solvent screening for biocatalysis.J Mol Catal B Enzym 2013 85—86 200—13. 

Phase-equilibrium calculations were discussed for vapor-liquid equilibria (VLB), liquid-liquid equilibria (LLE), and solid-solid equilibria (SSE). Results from VLE calculations often take the form of K-factors and relative volatilities, especially when thermodynamic calculations serve as intermediate steps in computer-aided process-design programs. In those situations, K-factors are routinely provided to subprograms that size distillation columns and gas-liquid absorbers. Similarly, the distribution coefficients computed for LLE serve as bases for sizing solvent-extraction columns moreover, liquid-liquid distribution coefficients may be helpful in screening candidate solvents for use in an extraction. 

In principle, aU these aspects related with the solvent selection will have an impact on the outcome of the LCA metrics of the various alternatives. Fig. 13.2 presents a generic flowsheet for the solvent-based postcombustion CO2 capture processes, illustrating also the main concerns from an LCA point of view. If the LCA scope is to screen solvents for postcombustion CO2 capture without performing process simulations (i.e., in a very early phase of process design, where perhaps the number of solvent molecules to be screened is immense, for instance, in computer-aided molecular design [CAMD] of solvent molecules [68]), the potential solvents should be characterized based on properties that would indicate their expected performance in the capture process (e.g., the standard flowsheet of Fig. 13.2). These properties can be thermodynamic in nature (e.g., solubility parameters between CO2 and the solvent, solvent heat of vaporization. 

Solvent/resin solubility matches can be made efficiently using a computer spreadsheet. A spreadsheet was used to screen six solvents and eight different resins by means of the radius of interaction equation (Equation 4.2). The results are listed in Table 4.2 where the resin is listed as soluble in the solvent if the calculated JR value is less then the resin spherical radius ( / ). If the resin radius is not available, the tentative method is to assume an average radius of 10. These compute- aided calculations can quickly suggest possible resin solvents or solvent blends. The spreadsheet can be designed to return a Y or N answer based on whether the calculated JR value is more or less than the resin sphere radius. 

---