Date of Award
Doctor of Philosophy
Agricultural and Biosystems Engineering
Kenneth J. Moore
D R. Raman
The purpose of this research was twofold: (i) to develop a system for screening lignocellulosic biomass feedstocks for biochemical conversion to biofuels and (ii) to evaluate brown midrib corn stover as feedstock for ethanol production.
In the first study (Chapter 2), we investigated the potential of corn stover from bm1-4 hybrids for increased ethanol production and reduced pretreatment intensity compared to corn stover from the isogenic normal hybrid. Corn stover from hybrid W64A X A619 and respective isogenic bm hybrids was pretreated by aqueous ammonia steeping using ammonium hydroxide concentrations from 0 to 30%, by weight, and the resulting residues underwent simultaneous saccharification and co-fermentation (SSCF) to ethanol. Dry matter (DM) digested by SSCF increased with increasing ammonium hydroxide concentration across all genotypes (P>0.0001) from 277 g kg-1 DM in the control to 439 g kg-1 DM in the 30% ammonium hydroxide pretreatment. The bm corn stover materials averaged 373 g kg-1 DM of DM digested by SSCF compared with 335 g kg-1 DM for the normal corn stover (P<0.0001). Of the bm mutations, bm3 had (i) the greatest effect on cell-wall carbohydrate hydrolysis of corn stover, (ii) the lowest initial cell-wall carbohydrate concentration, (iii) the lowest dry matter remaining after pretreatment, and (iv) the highest amount of monosaccharides released during enzymatic hydrolysis. However, bm corn stover did not reduce the severity of aqueous ammonia steeping pretreatment needed to maximize DM hydrolysis during SSCF compared with normal corn stover.
In the remaining studies (Chapters 3 thru 5), a system for analyzing the quality of lignocellulosic biomass feedstocks for biochemical conversion to biofuels (i.e., pretreatment, enzymatic hydrolysis, and fermentation) was developed. To accomplish this, a carbohydrate availability model was developed to characterize feedstock quality. The model partitions carbohydrates within a feedstock material into fractions based on their availability to be converted to fermentable sugars, including non-structural carbohydrates (CN) (monosaccharides, starches, oligosaccharides), biochemically available carbohydrates (CB) (structural carbohydrates susceptible to enzymatic hydrolysis) with an associated 1st-order availability rate constant (kB) and unavailable carbohydrates (CU) (hemicellulose and cellulose in close association with lignin). The model partitions the non-carbohydrate dry matter into extractives, lignin, and ash. Quality parameters were determined using a biomass quality assay that combined established wet-chemistry analyses techniques, including total non-structural carbohydrates (TNC), alcohol insoluble residue (AIR), simultaneous saccharification and fermentation (SSCF), and Klason lignin. The assay was used to analyze four compositionally diverse biomass feedstocks: corn cobs (Zea mays L.), hybrid poplar (Populus x canadensis Moench), kenaf (Hibiscus cannabinus L.) and switchgrass (Panicum virgatum L.). In these feedstocks, CN ranged from 27 to 127 g kg-1 DM, CB ranged from 34 to 344 g kg-1 DM, kB ranged from 0.071 to 0.415 h-1, total available carbohydrates (CA) ranged from 61 to 517 g kg-1 DM, and Klason lignin ranged from 139 to 244 g kg-1 DM.
The next study evaluated multiple high-throughput (HTP) modifications to the original assay methods, including (i) using filter bags with batch sample processing, (ii) replacement of AIR with neutral detergent fiber (NDF) as a cell-wall isolation procedure, and (iii) elimination of the fermentation organism in the SSCF procedures used to determine biochemically available carbohydrates. The original and the HTP assay methods were compared using corn cobs, hybrid poplar, kenaf, and switchgrass. Biochemically available carbohydrates increased with the HTP methods in the corn cobs, hybrid poplar, and switchgrass, but remained the same in the kenaf. Total available carbohydrates increased and unavailable carbohydrates decreased with the HTP methods in the corn cobs and switchgrass and remained the same in the hybrid poplar and kenaf. There were no differences in total carbohydrates (CT) between the two methods. The HTP methods consistently assayed less lignin than did the original method. Despite the slight differences parameter values, the HTP assay methods essentially gave a similar summary of the feedstock quality as did the original assay methods while significantly reducing the time and cost for feedstock quality analysis. The HTP assay methods was used to analyze 19 additional biomass feedstocks, including cool-season grasses, warm-season grasses, corn residues, and woody materials.
The final study evaluated the variability of biomass quality parameters in a set of corn stover samples, and developed calibration equations for determining parameter values using near infrared reflectance spectroscopy (NIRS). Fifty-two corn stover samples harvested in Iowa and Wisconsin in 2005 and 2006 were analyzed using the HTP assay for determining feedstock quality for biochemical conversion. Non-structural carbohydrates ranged from 84 to 155 g kg-1 DM, CB ranged from 354 to 557 g kg-1 DM, kB ranged from 0.20 to 0.33 h-1, CA ranged from 463 to 699 g kg-1 DM, and neutral detergent lignin (NDL) ranged from 32 to 74 g kg-1 DM. Significant differences (P<0.0001) among samples were observed for all parameters, except for the availability rate constant of CB. Near infrared reflectance spectroscopy calibration equations were developed for CN, CB, CA, CU, CT, and NDL. It was not possible to generate a meaningful calibration equation for kB.
Patrick Thomas Murphy
Murphy, Patrick Thomas, "Development of a system for characterizing biomass quality of lignocellulosic feedstocks for biochemical conversion" (2009). Graduate Theses and Dissertations. 11002.