Although pyrolysis of carbohydrate-rich biomass should theoretically yield large amounts of sugar, the presence of alkali and alkaline earth metals (AAEMs) in most biomass prevents this from happening.

Industry statistics indicate that technology‐learning rates can dramatically reduce both feedstock and biofuel production costs. Both the Brazilian sugarcane ethanol and the United States corn ethanol industries exhibit drastic historical cost reductions that can be attributed to learning factors. Thus, the purpose of this paper is to estimate the potential impact of industry learning rates on the emerging advanced biofuel industry in the United States. Results from this study indicate that increasing biorefinery capital and feedstock learning rates could significantly reduce the optimal size and production costs of biorefineries. This analysis compares predictions of learning‐based economies of scale, S‐Curve, and Stanford‐B models. The Stanford‐B model predicts biofuel cost reductions of 55 to 73% compared to base case estimates. For example, optimal costs for Fischer‐Tropsch diesel decrease from $4.42/gallon to $2.00/gallon. The optimal capacities range from small‐scale (grain ethanol and fast pyrolysis) producing 16 million gallons per year to large‐scale gasification facilities with 210 million gallons per year capacity. Sensitivity analysis shows that improving capital and feedstock delivery learning rates has a stronger impact on reducing costs than increasing industry experience suggesting that there is an economic incentive to invest in strategies that increase the learning rate for advanced biofuel production. © 2014 Society of Chemical Industry and John Wiley & Sons, Ltd

Thermochemical processing is a promising method for the rapid depolymerization of biomass. This study investigated switchgrass, corn stover, red oak, hybrid poplar, and loblolly pine in terms of heteropolymer and elemental composition, and the distribution and composition of the fast pyrolysis products. Corn stover differed from other biomass types in that less of the biomass was recovered as sugar or phenolic oil (PO) and more of the biomass was recovered as bio-char and bio-gas. The sugar-rich aqueous stream recovered from the bio-oil heavy fraction was characterized in terms of sugar content and distribution, inhibitor content, and ability to support production of ethanol by Escherichia coli KO11+lgk as a model biorenewable product. Levoglucosan was the most abundant sugar from each type of biomass, followed by either xylose or cellobiosan. For hybrid poplar, cellobiosan accounted for 30 wt% of the total sugar pool. Each of the sugar streams also contained a variety of inhibitors, particularly 5hydroxymethylfurfural (5-HMF) and methylcyclopentenolone. Methylcyclpentenolone, maple lactone, was found to decrease the specific growth rate of E. coli by 50% when present at 0.72 wt%, indicating that it is less toxic than furfural, acetic acid and guaiacol. Sugars produced from switchgrass contained 4fold less contaminants on a per-sugar basis than those from poplar and pine. All of the sugar streams contained too many inhibitors to be used at an industrially feasible concentration without additional detoxification. The poplar-derived pyrolytic sugar syrup was particularly inhibitory, possibly due to the high abundance of aromatic hydrocarbons, such as xylenes, and anisoles.