| dc.description.abstract | Cost-effective synthesis of fine chemicals/biorenewable fuels via fermentation invariably hinges upon efficient separations from the dilute broth. Biphasic/in situ extraction can positively impact both upstream fermentation and the downstream product separations. For developing environmentally sustainable processes, ionic liquids are touted as greener alternative to organic solvents not only because of their relatively low volatility but also due to the ability to tune their properties and design new ionic liquids for task specific needs. Solvent selection for in situ fermentation is depended on high solute partitioning and their biocompatibility with the microorganisms. Such information for these new set of solvents, ionic liquids that comprise of ions is very limited compared to organic solvents. Here, a new methodology to enable successful use of ionic liquids in in situ extraction is developed by focusing on two model systems: 1. Fine chemical - (1R,2S)-1,2-napthalene dihydrodiol (NDHD), produced from the biotransformation of naphthalene by Escherichia coli strain JM109(DE3) pDTG141 and 2. Biorenewable fuel - Biobutanol (along with Acetone and Ethanol), produced from the Fermentation of sugars by Clostridia. The partitioning of each of the solutes, NDHD, Acetone, 1-butanol and ethanol between water and ionic liquid has been measured in the range of concentrations typical of actual fermentations. Different cationic classes of ionic liquids including 1-alkyl-3-methylimidazolium ([RMIm]), trialkylmethylammonium ([TRMAm]) and trihexyltetradecylphosphonium ([P6,6,6,14]) were investigated along with anions such as halides and hydrophobic bis(trifluoromethylsulfonyl)imide [Tf2N] and (trifluorotris(perfluoroethyl)phospate [FAP]. The model ionic liquid, [HMIm][Tf2N] (an IUPAC/IACT standard), demonstrated the highest affinity for NDHD (KC = 2.8) while phosphonium and ammonium ionic liquids with bulkier alkyl side chains had the lowest extractability (KC 2) illustrating the role of molecular design for improvement. Ternary diagrams for the extraction of ABE solutes from water using model ionic liquid, [HMIm][Tf2N] were developed. Acetone and 1-butanol solutes were favorably extracted from water with high selectivities while its affinity for ethanol was low. NRTL activity coefficient model was used to model the ternary data and a regression program was written to obtain the binary interaction parameters for the ternary system. Simulation of 1-butanol extraction using [HMIm][Tf2N] was conducted using process simulator, Aspen Plus. Energy analysis was performed on an optimized flowsheet and these results along with equipment costs were compared with traditional organic solvent extraction and distillation. The toxicity of twenty different ionic liquids to the mutant strain of E. coli was tested with results varying from biocompatible to antimicrobial evidenced from EC50 values of growth rates. Here, the molecular toxicity was measured and EC50 refers to concentration of the ionic liquid that reduces cell growth rate by 50% at molecular level; the EC50 was compared with water solubilities, to determine if the ionic liquid was toxic at molecular level. As widely known, the increase of alkyl-chain length increased the toxicity. However, even highly alkylated cations may become biocompatible by the choice of anion, for e,g, trioctylmethylammonium bromide was antimicrobial while with [Tf2N] anion, it was biocompatible. The mechanism of growth inhibition in presence of ionic liquids was studied through imaging and initial explanations of possible inhibition mechanisms includes the effect of dissociation of these ionic liquids on how the cations/anions interact with the cell membrane. Fermentation broths can be viscous and exhibit non-Newtonian behavior and efficient liquid-liquid contacting is required for higher mass transfer rates and hence faster extractions. Intensification of contacting of non-Newtonian rheology fluids was studied by the application of electric fields. Continuous phase was Mineral oil containing a rheological modifier while aqueous carboxymethylcellulose (CMC) solutions were used as dispersed phase with the apparent viscosities varying between 1 cP to 1000 cP. Significant reduction in drop size was observed when the applied voltages were varied between 0 to 15 kV; viscosity of the continuous phase resulted in lower terminal velocities while dispersed phase viscosity affected the droplet formation times. Empirical correlations for droplet diameter as a function of physical properties, nozzle dimensions and electric field strength were developed and discussed here. | |
