9-2 Characterization of corn stover via high throughput thermochemical pretreatment and enzymatic hydrolysis
Tuesday, April 26, 2016: 1:25 PM
Key Ballroom 3-4, 2nd fl (Hilton Baltimore)
M. van den Berg*, L. Coulier, A. Carvalho de Souza and M. Appeldoorn, DSM, Delft, Netherlands; S.R. Decker, G.B. Turner, E. Gjersing, R. Sykes, M.P. Tucker, C. Doeppke and L.A. Schuster, National Renewable Energy Laboratory, Golden, CO, USA
Corn stover was pretreated in the NREL High-Throughput Biomass Recalcitrance Screening Pipeline.  A range of pretreatment conditions (varied H2SO4 concentration, time, temperature) was conducted and mass loss, sugar release, acid-soluble lignin content and lignin degradation products were determined.  The hydrolysate liquors for each pretreatment condition were analyzed for xylo- and gluco-oligomers and inhibitor formation (hydroxymethylfurfural and furfural). The solids were analyzed for structural carbohydrates, acid-insoluble lignin and cellulose crystallinity. 

High severity (180oC, 17.5min, 2% acid) resulted in major losses of xylose through degradation, while low severity (120oC, 9min, 0% acid) resulted in minimal sugar release.  The middle range resulted in high xylose yields, upto ~80% of theoretical yield at optimal severity, where acid soluble lignin levels were ~20% of total lignin. Both xylo- and gluco-oligomers were released in a small window of cSF, with the xylo-oligomers being released at lower cSF compared to the gluco-oligomers.  Both correlated with the release of their corresponding monomer.  As expected, xylose release increased with severity and then decreases dramatically due to degradation to furfural.  Glucose release, while typically low, also increases with severity, however HMF formation is much slower than furfural and overall glucose levels continue to rise even during HMF formation.

The enzyme hydrolysis rate determining composition (monomers vs oligo/polymers distribution, lignin decomposition and degradation products) varied with the actual settings of the pretreatment (H2SO4 concentration, time, temperature) and was only partly correlating with cSF. This illustrates that the cSF number is not fully predictive for the impact of pretreatment on enzyme hydrolysis.