(450d) Sugar Concentration for Continuous Enzymatic Saccharification Using Nanofiltration and Reverse Osmosis Membranes

Cost effective conversion of lignocellulosic biomass into bioethanol requires maximizing ethanol yields.  Enzymatic hydrolysis of polysaccharides, e.g. cellulose, starch, is usually conducted in a batch reactor.  Disadvantages of the classical batch reactor are: product variation from batch to batch, higher overall investment costs due to larger reactor volumes, higher running costs due to frequent startup/shut down, one time use of enzymes as well as catalyst/enzyme separation costs.  Development of a continuous saccharification process overcomes the limitations of batch operation.  Additional potential advantages include: recovery and reuse of enzymes, improvement of product yield and kinetics, and reduction in inhibition of enzymes.  During a continuous saccharification process it is desirable to concentrate the product sugar stream prior to fermentation. 

Here we focus on sugar concentration using nanofiltration and reverse osmosis membranes.  Nanofiltration, which originated in the 1970s, is one of the newest pressure driven membrane filtration processes.  Low pressure reverse osmosis membranes came to be known as nanofiltration membranes with some of the earliest applications being described in the 1980s.  Characteristics of nanofiltration membranes include greater than 99% rejection of multivalent ions, 0-70% rejection of monovalent ions and greater than 90% rejection of small organic compounds with molecular weights in the range 150-300. 

We have investigated concentration of aqueous sugar streams using a number of different commercially available nanofiltration and low pressure reverse osmosis membranes.  Experiments have been conducted using model feed streams as well as corn stover derived biomass hydrolysates.  Model feed streams consisted of DI water containing D-Xylose and D-glucose at a ratio of 3:1.  In addition compounds that negatively impact ethanol yields during fermentation such as acetic acid, furfural and 5-hydroxymethylfurfural were also added to the feed stream in the ratio 6:2:1.  The effect of feed pH was determined.  Membrane performance was characterized by determining the level of sugar (xylose and glucose concentration) as well as removal (passage through the membrane) of acetic acid, furfural and 5-hydroxymethylfurfural.  The percentage decrease in permeate flux was also determined.  Corn stover hydrolysates obtained from the National Renewable Energy Laboratory, Golden Colorado, were also tested. 

Membrane surfaces have been characterized using X-ray photoelectron spectroscopy.  Differences in membrane performance have been related to interactions between the various species present in the feed and chemical groups present on the membrane surface. The result presented here may be used to identify suitable membranes and operating conditions for sugar concentration and removal of species such as acetic acid, furfural and 5-hydroxymethylfurfural.