Optimization of biomass pretreatments using fractional factorial experimental design

Author(s)

Rezende, Camila A. ^[1] ; Atta, Beatriz W. ^[1] ; Breitkreitz, Marcia C. ^[1] ; Simister, Rachael ^[2] ; Gomez, Leonardo D. ^[2] ; McQueen-Mason, Simon J. ^[2]

Affiliation

^[1] Univ Estadual Campinas, UNICAMP, Inst Chem, POB 6154, BR-13083970 Campinas, SP - Brazil

^[2] Univ York, CNAP, York YO10 5YW, N Yorkshire - England

• Document type: Journal article
• Source: BIOTECHNOLOGY FOR BIOFUELS; v. 11, JUL 24 2018.
• Web of Science Citations: 6

Abstract

Background: Pretreatments are one of the main bottlenecks for the lignocellulose conversion process and the search for cheaper and effective pretreatment methodologies for each biomass is a complex but fundamental task. Here, we used a 2.5-1 fractional factorial design (FFD) to optimize five pretreatment variables: milling time, temperature, double treatment, chemical concentration, and pretreatment time in acid-alkali (EA) and acid-organosolv (EO) pretreatments, applied to elephant grass leaves. Results: FFD allowed optimization of the pretreatment conditions using a reduced number of experiments and allowed the identification of secondary interactions between the factors. FFD showed that the temperature can be kept at its lower level and that the first acid step can be eliminated in both pretreatments, without significant losses to enzymatic hydrolysis. EA resulted in the highest release of reducing sugars (maximum of 205 mg/g substrate in comparison to 152 mg/g in EO and 40 mg/g in the untreated sample), using the following conditions in the alkali step: {[}NaOH] = 4.5% w/v; 85 degrees C and 100 min after ball milling the sample. The factors statistically significant (P < 0.05) in EA pretreatment were NaOH concentration, which contributes to improved hydrolysis by lignin and silica removal, and the milling time, which has a mechanical effect. For EO samples, the statistically significant factors to improved hydrolysis were ethanol and catalyst concentrations, which are both correlated to higher cellulose amounts in the pretreated substrates. The catalyst is also correlated to lignin removal. The detailed characterization of the main hemicellulosic sugars in the solids after pretreatments revealed their distinct recalcitrance: glucose was typically more recalcitrant than xylose and arabinose, which could be almost completely removed under specific pretreatments. In EA samples, the removal of hemicellulose derivatives was very dependent on the acid step, especially arabinose removal. Conclusion: The results presented herewith contribute to the development of more efficient and viable pretreatments to produce cellulosic ethanol from grass biomasses, saving time, costs and energy. They also facilitate the design of enzymatic cocktails and a more appropriate use of the sugars contained in the pretreatment liquors, by establishing the key recalcitrant polymers in the solids resulting from each processing step. (AU)

• FAPESP's process: 16/10955-6 - Study on the production of second generation ethanol from elephant grass
• Grantee: Beatriz Widmer Atta
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• FAPESP's process: 16/05636-9 - Development of innovative analytical methods for analytes of pharmaceutical interest based on ultra-high supercritical fluid chromatography associated with chemometrics tools
• Grantee: Márcia Cristina Breitkreitz
• Support Opportunities: Regular Research Grants
• FAPESP's process: 16/13602-7 - Study on strategies to allow the integral use of lignocellulosic biomass
• Grantee: Camila Alves de Rezende
• Support Opportunities: Program for Research on Bioenergy (BIOEN) - Regular Program Grants