Progress 12/15/14 to 12/14/18
Outputs
Target Audience:During the reporting period, results from our research were made available to researchers and academic, government, and industry professionals through publications, poster presentations, and oral presentations at national and international conferences including the Symposium on Biotechnology for Fuels and Chemicals and the American Institute of Chemical Engineers National Conference. Target audiences at the conferences included academics seeking to advance the knowledge of conversion of lignocellulose to fuels, government researchers seeking to understand the state of technology to make informed policy decisions, industrial practitioners seeking to stay up to date on the latest research on conversion of cellulosic biomass to fuels, and the general public wishing to understand the potential for production of low cost liquid fuels sustainably. Through publication in the journal ACS Catalysis, a top ranked journal in the catalysis field, results were provided to researchers working on fundamental understanding of the chemistry of heterogeneous, homogenous, and biocatalysts. Additionally, research was presented on campus engineering diversity organization primarily attended by underrepresented minorities in engineering, inspiring the next generation of researchers and engineers. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?A PhD candidate, Christian Alcaraz, was trained on aqueous biomass pretreatment, enzymatic hydrolysis, isomerization technology, development of anaerobic bacteria cultivation and propagation procedures, Aspen Modeling Software, Techno economic Evaluation, Industrial Fermentation, and Bioreactor Process Control. A second PhD candidate, Glen Svenningsen, was trained on chemical reactor process control, catalytic synthesis and characterization, high-throughput reactors, Aspen modeling software, and techno-economic evaluation. Five undergraduates (4 of whom are either underrepresented minorities and/or women STEM) were trained in aqueous phase biomass pretreatment enzymatic hydrolysis, isomerization technology, development of anaerobic bacteria cultivation and propagation procedures, data analysis, and bioreactor process control. How have the results been disseminated to communities of interest?The results of the project have been disseminated by either oral or poster presentation to conference members who attended the American Institute of Chemical Engineers National Conference between 2015- 2017 or the Symposium on Biotechnology for Fuel and Chemicals 2017-2019. Results were also published in the high ranking journal American Chemical Society: Catalysis. Throughout the entire project lifetime, the results were presented at the annual Chemical and Environmental Engineering Symposium. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Co-solvent Enhanced Lignocellulosic Fractionation (CELF) pretreatment was applied to milled corn stover at 130°C to 150°C for 10 to 40 min at a 5 wt% solids loading. The compositions of the pretreated solids and the liquids were analyzed and applied to mass balances. To gain insights into CELF, kinetic models were developed and fit to pretreatment data. From the models, fractions of slow and fast solubilizing xylan, bulk and residual lignin, and corresponding rate constants were determined. The fraction of fast and slow xylan in stover was 74% and 26%, respectively, and their conversion rate constants were 0.209 and 0.002 per minute at 130°C, 0.445 and 0.0207 per minute at 140°C, and 1.50 and 0.052 per minute at 150°C. The rate constants for xylan degradation at 130°C, 140°C, and 150°C were 0.00263, 0.00295, and 0.00673 per minute, respectively. The fractions of bulk and residual lignin were 79.3% and 20.7%, respectively. The rate constants for bulk and residual lignin conversion were 0.250 and 0.0015 per minute at 130°C, 0.290 and 0.005 per minute at 140°C, and .650 and 0.0372 at 150°C. These values revealed a strong correlation between fast xylan and bulk lignin and slow xylan and residual lignin, suggesting these fractions are strongly interlinked in corn stover. The rate constants more than doubled if operating temperature exceeded 140°C. This provided evidence of catalytic delignification and xylan removal, a phenomena ascribed to miscibility of THF in water above 145°C. In the miscible regime, lignin was better solvated by THF and removal of lignin solubilized more xylan. Additionally, xylose in solution experienced enhanced degradation at temperatures above 145°C. Models were also applied to gain kinetic insights into cellulose solubilization during pretreatment and subsequent enzymatic hydrolysis. Rate constants for cellulose solubilization to glucose at 130°C, 140°C, and 150°C were 0.000628, 0.00157, and 0.00372 per minute, respectively, after compensating for doubling acid concentration at 130°C and 140°C. The activation energy for cellulose hydrolysis was calculated to be 126 kJ/mol, a value generally attributed to hydrolyzing amorphous cellulose, suggesting glucose in pretreatment solution comes from amorphous cellulose. After pretreatment, pretreated solids were hydrolyzed with 65, 15, and 5 mg protein of enzyme/g glucan. The solids were 100% digestible in all cases at 65 mg enzyme/g glucan. At 15 mg enzyme/g glucan, only pretreatments at 140°C for 30 and 40 minutes and 150°C for 20, 30, and 40 minutes resulted in 100% digestibility. At 5 mg enzyme/ glucan, only 150°C at 30 and 40 minutes produced digestibility of at least 90%. To further understand factors influencing digestibility, a fractal kinetics model was applied to all enzymatic hydrolysis data for a 15 mg enzyme/g glucan enzyme loading. The h-factor that indicates impedance to fast and high yield enzymatic hydrolysis correlated strongly with the amount of lignin left in the solids (R-squared 0.74). Overall, these models suggest CELF should be at least at 150°C to realize the high lignin and xylan removal imperative to high enzymatic hydrolysis yields. Clostridium thermocellum was employed for cellulose saccharification to use glucose isomerase could in the same reactor as simultaneous saccharification and isomerization or SSI. However, despite supplementation with enzymes, maximum fructose yields were 10% possibly due to long cellulose hydrolysis times resulting in degradation and glucose lost to bacterial growth. Application of fungal cellulase and glucose isomerase to CELF pretreated corn stover at pH 5.5 and 50°C achieved fructose yields of 40% for a loading of 5 mg enzyme/g glucan. When enzymatic hydrolysis and glucose isomerase were separated, yields of the 50/50 equilibrium mixture of glucose and fructose reached 100% of theoretical maximum. Additionally, separating the isomerization and hydrolysis steps allowed borate addition to shift equilibrium to reach up to 80% fructose. Next, primary factors controlling selectivity of fructose dehydration to 5-hydroxymethylfurfural (HMF) were determined. A high-throughput 96-well Hastelloy plate reactor was developed to gather extensive data on fructose conversion to HMF with varying aqueous solvents, temperatures, times, and acid types and concentrations. In situ 13-C NMR analyzed the distribution of fructose tautomers as a function of solvent type and temperature. Of the aprotic solvents dimethylsulfoxide (DMSO), dimethylacetamide, dimethylformamide, and N-methyl-2-pyrrolidone, DMSO achieved the highest HMF yields. HMF selectivity was directly correlated with distribution of furanose and open-chain fructose tautomers as a function of reaction temperature, time, and solvent, supporting the hypothesis that selectivity in fructose conversion to HMF is primarily controlled by equilibrium between tautomeric forms of fructose. Difructose anhydrides were determined to protect intermediates to enhance HMF selectivity at longer times than predicted by fructose tautomeric distribution. This data provided insights into mechanisms controlling HMF selectivity from fructose dehydration by homogeneous acid catalysts and allowed development of conditions to achieve ~85% yield of HMF from fructose by maximizing stability of the furanose form of fructose. These findings were published in ACS Catalysis. These results confirmed selectivity control by increasing the furanose ratio. Although DMSO promoted furanose stabilization more than other solvents, its high boiling point made it impractical for economically separating HMF from DMSO. Thus, we hypothesized that porous silica particles functionalized with acid and methyl sulfoxide groups could generate microenvironments in the pores for lower boiling point solvent systems while promoting high selectivity and yields for fructose conversion. Accordingly, catalysts were functionalized by grafting 3-mercaptopropyl trimethylsiloxane onto silica supports with 60, 90, and 150 Angstrom pores, with 150 A proving most effective in dehydrating 50 g/L fructose in 1:1 THF:water solutions. The proportion of sulfoxide to sulfonic acid groups was controlled by addition of iodomethane to the mercaptopropyl group and oxidization with 25 wt% hydrogen peroxide to form sulfoxide and sulfonic acid. Fructose dehydration at 120oC for 8 hours demonstrated that adding 5000 mg iodomethane to 250 mg of functionalized silica increased selectivity to 55% compared to a selectivity of 35% with no iodomethane. Additionally, fructose conversion increased to 35% from 25%, likely due to increased rates in the microenvironment. Although, these yields were not as high as achieved for homogeneous catalytic systems, the systematic trends observed suggested that further optimization will allow high HMF yields from a heterogeneous catalytic system that employs 1:1 THF: water solvents and facilitate economical extraction of HMF from the mixture. Techno-economic analysis (TEA) was applied to fructose conversion to dimethylfuran using Aspen software. The newness of CELF technology invented at UCR necessitated simplifying process models for a conceptual biorefinery. HMF and DMF production were divided into fructose dehydration, hydrogenation, and purification sections with aqueous THF in continuously stirred tank reactors (CSTR). The process input was set at 500 metric ton/day of fructose, and the plant was amortized over 20 years. 25 parameters were varied to determine which factors had the greatest effect on the minimum selling price (MSP) for DMF. The MSP for DMF was estimated as $1.74/kg ($1.9/l). Sensitivity analyses demonstrated that the most significant parameters were feedstock cost, product yields, catalyst cost, and total purchased equipment costs.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2018
Citation:
Svenningsen G, Kumar R, Wyman CE, Christopher P. 2018. Unifying Mechanistic Analysis of Factors Controlling Selectivity in Fructose Dehydration to 5-Hydroxymethylfurfural by Homogeneous Acid Catalysts in Aprotic Solvents, ACS Catalysis, 8 (6), 5591-5600.
|
Progress 12/15/16 to 12/14/17
Outputs
Target Audience:The target audiences include academics seeking to advance the knowledge of biological conversion systems and train students, government researchers who seek to understand such systems and development policy, industrial practitioners who seek to apply advanced technologies for conversion of cellulosic biomass, and the general public who wish to better understand the potential for production of low cost liquid fuels sustainably. In this reporting period, the results have been disseminated in the form 1 journal paper and of 1presentationat the American Institute for Chemical Engineers 2017. In addition, the graduate students presented their data at the inter-departmental colloquium held at the Bourns College of Engineering, University of California, Riverside. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided? PhD graduate student, Christian Alcaraz, was trained on aqueous biomass pretreatment, enzymatic hydrolysis, and isomerization technology, development of anaerobic bacteria cultivation and propagation procedures, and data analysis. PhD graduate student, Glen Svenningsen, learned homogeneous and heterogeneous catalysis processing via a novel high-throughput methodology and was trained to use multiple laboratory characterization equipment for both catalyst characterization and data analysis. Two undergraduates were trained in aqueous phase biomass pretreatment enzymatic hydrolysis, and isomerization technology, development of anaerobic bacteria cultivation and propagation procedures, and data analysis. How have the results been disseminated to communities of interest? Svenningsen G*, Kumar R, Wyman CE, Christopher P. 2017. "Identifying the Mechanism of Fructose Dehydration to 5-Hydroxymethylfurfural in Aqueous Dipolar Aprotic Solutions by Homogeneous Acids," ACS Catalysis, 2017 (publication in press). Svenningsen G*, Kumar R, Wyman CE, Christopher P. 2017. "Selectivity Controlling Factors for Fructose Dehydration to 5-Hydroxymethylfurfural," AIChE 2017 Annual Meeting, Minneapolis, MN, November 17(oral presentation). Alcaraz C, Kumar R, Christopher PC, Wyman CE. 2017. "Low Severity CELF Pretreatment of Corn Stover for the Development of Novel Simultaneous Saccharification and Isomerization (SSI) technology for Catalytic Conversion into Drop-In Fuels" Chemical and Environmental Engineering Symposium 2017, UC Riverside, CA, September 18th (oral presentation). Svenningsen G*, Kumar R, Wyman CE, Christopher P. 2017. "Selectivity Controlling Factors for Fructose Dehydration to 5-Hydroxymethylfurfural," Chemical and Environmental Engineering Symposium 2017, UC Riverside, CA, September 18th (oral presentation). What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period, CELF pretreatments will be carried out with varying THF:water ratios and acid loadings to see their impact on sugar release/delignification. A mass balance will be carried out in each case and the solids will be used in objective 2 if sugars are retained and solids are highly digestible. Based on the results from last year, saccharification and isomerization will be attempted in parallel and in sequence in order to determine the maximum amount of fructose produced.Additionally, since saccharification and isomerization may occur in sequence, other isomerization catalysts will be explored that have shown high promise in terms of fructose yields (sodium borate and sodium aluminate). Afterwards, using Clostridium thermocellum will be revisited as a source of cellulase that can be coupled with glucose isomerase. To accomplish objectives 3 and 4, both the dehydration of fructose and the hydrogenation of HMF will be carried out in-situ in our high-pressure, continuously stirred, 1 L Parr reactor. High fructose concentrations of over 25 g/L will be reacted in the presence of 1:4 water/dioxane solutions. The technoeconmic model will be further refined.
Impacts
What was accomplished under these goals?
Co-solvent Enhanced Lignocellulosic Fractionation (CELF) pretreatment was applied to milled corn stover at 130°C-150°Cfor 10-40 minat 5 wt% solids loading.At 130°C, nearly 100% of xylan and glucan were recovered between the liquid and solids streams for all reaction times; however, lignin removal was still low. At 140°C, near 100% of xylan and glucan were recovered in the liquid and solids for reaction times of 10 and 20 min, but sugar degradation became apparent at 30 and 40 min. The amount of lignin removed was >35% for all reaction times.At 150°C, the amount of xylan lost to degradation products for reaction times 10, 20, 30, and 40 minutes was 7%, 9%, 15%, and 17%, respectively. The amount of delignification for the same time points was 50%, 78%, 78%, and 78%, respectively, and was higher than at 130°C and 140°C. The pretreated solids preparedat 130°C were not highly digestible even at high cellulase loading. Enzymatic hydrolysis was then performed on pretreated solids prepared at 140°C. The amount of glucan released reached 100% by day 1 at 65 mg cellulase loadingand >80% at 15 mg enzyme/g glucan for all pretreatment times. For pretreated solids prepared at 140°C, glucan yield was 100% at 15 mg enzyme/ g glucan for all pretreatment times and >60% at 5 mg enzyme/g glucan. For 5 mg cellulase loading, the yield was still on an upward slope and is expected to reach to theoretical yields given more time. Next the compatibility of cellulase with glucose isomerase (GI) for simultaneous saccharification and isomerization (SSI) process was evaluated.The results were as follows: 1) The addition of oxygen scavengers and nitrogen purging decreased fructose yields, 2) The amount of fructose produced increased from pH 5 to 5.5 but dropped dramatically at pH 6, and 3) The amount of fructose produced increased from temperature 50 to 55 but dropped dramatically at temperature 60°C. The SSI experiment was then performed on CELF pretreated corn stover. The total sugar released was 65% with 20% fructose and 45% glucose after 14 days of SSI reaction. Low sugar release is most likely due to different optimum temperatures and pH for GI and Accellerase. To address this, we are developing strategies to increase yields in SSI and developing separate hydrolysis and isomerization strategy to ensure maximum glucose yields from the pretreated solids before the isomerization step. Aqueous mixtures of DMSO containing water (ranging from 0 to 75 mol%) and fructose at 25 g/L in sulfuric acid (10 mM) were dehydrated at 70-150°C over range of times to determine the mechanism and factors that control selectivity in fructose dehydration to HMF by homogeneous acid catalysts. Increasing water content in DMSO mixtures demonstrated that the HMF selectivity for water contents below 50 mol% was relatively constant (between ~70-90%); however, at higher water contents, the selectivity was lower and equal to what is typically reported for pure aqueous solutions (~50-70%). In contrast, additional water was observed to decrease the rate of fructose conversion, demonstrating a selectivity to HMF that is independent from fructose conversion. This supports a mechanism in which the selectivity controlling step comes before the rate limiting steps. To further support this mechanism, the sulfuric acid catalyzed dehydration of fructose (25 g/L) was performed over a pH range of 1 to 5 at 150°C for 30 mins in DMSO mixture. With increase in pH (lower acid loading), the fructose conversion decreased; however, selectivity to HMF remained constant. It was seen that both the selectivity to HMF and the conversion increased with temperature. Finally, DMSO displayed the highest selectivity at 84% followed by NMP (77%), DMA (75%), and DMFA (73.1%). Previous nuclear magnetic resonance (NMR) studies on the tautomeric distribution of fructose propose that this distribution shifts from the frucotpyranose (six-membered ring) to the fructofuranose (five-membered ring) form of fructose with temperature. To further clarify this,13C-NMR was performed in a high pressure NMR tube under deaerated conditions to examine the change in the fructose tautomeric distribution in DMSO solutions at 70-120°C and NMP, DMA, and DMFA at 120°C for 30 mins and correlate it to HMF selectivity. The results indicate that the tautomeric distribution directly controls the selectivity to HMF in which the higher the ratio of the furanose to pyranose form the higher the selectivity to HMF. However, this relationship only occurred over short time frames due to the formation of difructose ahydrides. Our recent work showed that high HMF yields (~83%) were realized in aqueous mixtures of dimethyl sulfoxide (DMSO). To mimic the solvent effects of DMSO, the surfaces of heterogeneous acid catalysts were functionalized with 1, 3-methyl propyl sulfoxide. These sulfoxide groups were grafted near Bronsted acid sites on the catalyst surface to generate a surface environment similar to that of homogeneous Bronsted acid catalysts in the presence of DMSO. The resulting surface environment is expected to be able to match HMF yields produced in aqueous mixtures of DMSO, while aqueous mixtures of dioxane achieve high DMF yields. These systems will be further analyzed in the next performance period. Utilizing Aspen Plus V8, a process flow diagram was developed for the conversion of corn stover to both dimethylfuran (DMF) and methylfuran (MF). This techno-economic study compares CELF technology to several other studies which convert corn stover to furans. The goal of this study is to input experimental parameters and yields from CELF and catalytic experiments developed at CE-CERT. Additional data and parameters were adapted from the NREL economic models such as corn stover composition, waste water treatment, storage, etc. For this process flowsheet, corn stover was milled and pretreated in a 1:1 mixture of dioxane:H2O with a metal halide catalyst. This simulation utilized stoichiometric reactors and DSTWU distillers. This pretreatment would generate glucan rich solids and an aqueous mixture of soluble components, particularly furfural. The glucan rich solids are then filtered and washed with THF and water at 3:1 ratio of solvent to biomass. The aqueous phase was collected and then neutralized. The choice of dioxane as the pretreatment solvent was due to down-stream separation. Binary phase diagrams of MFand THF were calculated and seen to be almost identical, making it impractical to separate by distillation; therefore, a lower volatile aprotic cyclic ether, dioxane, was needed. In order to remove water, toluene was used as an extracting solvent in a 1:1 mixture to separate furfural from water. Excess water in the toluene:dioxane mixture was removed by a micro-sieve filtration. Then furfural and other soluble compounds underwent hydrogenation at 200°C to generate MF. MF was then easily separated by distillation from toluene and dioxane. MF was collected into storage tanks and then toluene and dioxane were further separated by distillation. Both solvents were then recycled to minimize the amount of organic solvent consumed. The glucan rich solids underwent simultaneous saccharification and isomerization to produce fructose rich streams. Enzyme loadings and cost were estimated in this stage as experimental data is still be processed. The fructose rich streams underwent dehydration at 120°C in a 1:1 THF: water mixture with an acidic heterogeneous catalyst to produce HMF. After neutralization and removal of water by microsieve filtration, the HMF was hydrogenated to DMF at 200°C. DMF was then separated from THF by distillation and the THF was recycled. Sensitivity analysis showed that this process is most economically sensitive to feedstock cost, solvent recyclability, and enzyme cost.
Publications
- Type:
Journal Articles
Status:
Under Review
Year Published:
2018
Citation:
Svenningsen G, Kumar R, Wyman CE, Christopher P. 2017. Identifying the Mechanism of Fructose Dehydration to 5-Hydroxymethylfurfural in Aqueous Dipolar Aprotic Solutions by Homogeneous Acids, in review.
|
Progress 12/15/15 to 12/14/16
Outputs
Target Audience:The target audiences include academics seeking to advance the knowledge of biological conversion systems and train students, government researchers who seek to understand such systems and development policy, industrial practitioners who seek to apply advanced technologies for conversion of cellulosic biomass, and the general public who wish to better understand the potential for production of low cost liquid fuels sustainably. In this reporting period,the results have been disseminated in the form 1 journal paper and of 3 presentations at the American Institute for Chemical Engineers 2016and the Symposium on Biotechnology for Fuels and Chemicals 2016. In addition,the graduate students presented their data at the inter-departmental colloquium held at the Bourns College of Engineering, University of California, Riverside. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided? PhD Candidate, Christian Alcaraz, was trained on aqueous phase biomass pretreatment and hydrolysis technology, development of anaerobic bacteria cultivation and propagation procedures, and data analysis using laboratory analytical equipment for mass balance calculations. PhD graduate student, Glen Svenningsen, learned homogeneous and heterogeneous catalysis processing via a novel high-throughput methodology and was trained to use multiple laboratory characterization equipment for both catalyst characterization and data analysis. Three undergraduates were trained in both the biomass and catalysis laboratories. Additionally, both graduate students were given the opportunity to gain a broader understand of their field in biomass special topic courses in which students investigated the state of biomass market penetration in various countries, the goals and strategies of the "50 hottest companies in Bioenergy" according to Biofuels Digest, and the effect of operating conditions being defined in our laboratory on the minimum ethanol selling price predicted by the NREL technoeconomic model. How have the results been disseminated to communities of interest?The results have been disseminated in the form of 1 paper and 3 presentations over the course of the last year as follows: Svenningsen G*, Kumar R, Wyman CE, Christopher P. 2017. "Identifying the Mechanism of Fructose Dehydration to 5-Hydroxymethylfurfural in Aqueous Dipolar Aprotic Solutions by Homogeneous Acids," ACS Catalysis, 2017 (publication in progress). Alcaraz C, Kumar R, Cai C, Wyman CE. 2016. "High Concentration Glucose Production from CELF Pretreated Corn Stover Using Wild TypeClostridium ThermocellumSupplemented with Beta-Glucosidase," AIChE 2016 Annual Meeting, San Francisco, CA, November 15(oral presentation). Svenningsen G*, Kumar R, Wyman CE, Christopher P. 2016. "The Role of Atmosphere, Solvents and Acids on the Production of 5-Hydroxymethylfurfural from Biomass Derived Fructose," AIChE 2016 Annual Meeting, San Francisco, CA, November 16 (oral presentation). Svenningsen G*, Kumar R, Wyman CE, Christopher P. 2016. "Development of Heterogeneous Catalysis for Conversion of Fructose to 5-Hydroxymethylfurfural in Aqueous Cosolvent Systems," 38th Symposium on Biotechnology for Fuels and Chemicals, Baltimore, MD, April 27 (oral presentation). What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period, CELF pretreatments will be carried out at lower solids loadings in the Parr reactor to improve mixing and obtain more uniform temperature in the biomass. Material balances will then be applied to validate the data and confirm high lignin removal and recovery of sugars in the solids and liquids from CELF. Large amounts of pretreated solids will be generated at CELF reaction conditions that give the highest yields for use in Objective 2. Based on the results from the last year, fermentation conditions will be further optimized to maximize glucose yields from cellulose saccharification by Clostridium thermocellum supplemented with beta-glucosidase. It is expected that a thermostable beta-glucosidase will be more compatible with Clostridium thermocellum and realize higher glucose yields. In this regard, beta-glucosidase from Thermoanaerobacter brockii is more thermostable than commercial Aspergillus niger beta-glucosidase, the latter having a maximum operating temperature that is 20°C lower than preferred for Clostridium thermocellum fermentations. In addition, the literature reports that beta-glucosidase from Thermoanaerobacter brockii combined well with Clostridium thermocellum. Thus, a strain of Thermoanaerobacter brockii from the American Type Culture Collection (ATCC) will be employed to produce beta-glucosidase for our upcoming experiments. Our results to date show that glucose production limits yields, and preliminary experiments indicate that media components used with Clostridium thermocellum do not inhibit the glucose isomerization enzyme. Thus, the primary concern is whether the addition of sodium tetraborate to shift the glucose/fructose equilibrium impacts Clostridium thermocellum growth. If glucose and fructose production prove to be too incompatible, the two steps will be applied separately. Although the high loadings of fungal enzymes required to achieve high yields of glucose from cellulosic biomass have proven too expensive in the past, the highly reactive solids produced by our CELF pretreatment can be saccharified to sugars with very low amounts of fungal enzymes that would be economically attractive. Thus, if Clostridium thermocellum proves to not realize high yields vital to economic viability, commercial fungal enzymes can be applied. To accomplish objectives 3 and 4, both the dehydration of fructose and the hydrogenation of HMF will be carried out in-situ in our high-pressure, continuously stirred, 1 L Parr reactor. High fructose concentrations of over 25 g/L will be reacted in the presence of 1:4 water/dioxane solutions. First, sulfoxide functionalized dehydration catalysts will be added to the reactor, followed by heating the reactor by a fluidized sand bath at temperatures below 150oC for several hours to fully convert fructose to HMF. Then the dehydration catalysts will be removed, and hydrogenation catalysts will be loaded into the reactor. After the reactor is sealed, pressurized hydrogen gas will be fed into the system at over 180oC for several hours to fully hydrogenate HMF to DMF at high yields.
Impacts
What was accomplished under these goals?
Our novel Co-solvent Enhanced Lignocellulosic Fractionation (CELF) pretreatment was applied to corn stover over a range of temperatures (140°C-150°C) and times (10-40 minutes) at 10 wt% solids loadings in a 1L stirred tank Hastelloy Parr reactor equipped with a double-stacked pitch blade impeller. The reactor was rapidly heated to reaction temperature by a sand bath. In all cases, 67% -100% of glucan and 1-17% of xylan were recovered in the solid residue left after CEFL pretreatment. Operation of CELF pretreatment at 150°C for 20 minutes removed 64% of the lignin and resulted in the highest combined sugar recovery in the solids: 100 % of the glucan and 14% of the xylan. The result was a solids composition of 68.8% glucan, 9.7% xylan, 0.68% arabinan, 13.6% lignin, 7.2% ash, and 0.02% other. The high lignin removal from corn stover realized at these conditions aids our second biological deconstruction objective by improving the accessibility of the polysaccharides in the solids to enzymes. These results were for a 10 wt% solids loading, but greater delignification is expected at lower solids loadings of 7.5wt% and 5 wt% that are more readily handled by our Parr Reactor. The solids yield was 42.6%, and most of the sugars not left in the solids were recovered in the liquid stream. The solids produced in Objective 1 were then subjected to simultaneous saccharification and isomerization (SSI) to convert the residual glucan into fructose. Initially, 10 g/L micro-crystalline cellulose Avicel PH101 was loaded into a 50 mL working volume pressure bottle under anaerobic conditions containing 15 mg enzyme per gram glucan of Beta-glucosidase from Aspergillus niger, 3 mg of immobilized glucose isomerase from Streptomyces murinus, and 2% Clostridium thermocellum inoculum (v/v). Because SSI for 24 hours at 60°C resulted in only 8% fructose yields, the process was divided into distinct steps of 1) glucan saccharification and 2) glucose isomerization to reveal opportunities to increase fructose yields. Because Clostridium thermocellum does not readily use glucose, supplementation of the fermentation with Beta-glucosidase was expected to increase yields of the latter. Thus, 100 g/L crystalline cellulose was loaded into a 50 mL working volume pressure bottle under anaerobic conditions with a 2% Clostridium thermocellum inoculum (v/v). After 6 days, glucose and cellobiose concentrations in the broth reached 5 g/L and 8 g/L, respectively. Next, the same experiment was repeated but with the addition of 15 mg Beta-glucosidase from Aspergillus niger per g glucan 3 days after the Clostridium thermocellum fermentation was started. The result was 100% conversion of cellobiose to glucose and a glucose concentration of 10 g/L in the broth after 5% glucan conversion. A novel high-throughput approach was developed to define reaction conditions (temperature, acid type and loading, and solvent type and loading) that result in the highest 5-hydroxymethylfurfural (HMF) yields. In particular, a 96 Hastelloy well plate reactor was employed to simultaneously define conditions that result in the highest fructose dehydration to HMF. The results were then used to develop extensive kinetic analysis with which to determine mechanisms and factors that control selectivity in fructose dehydration to HMF by homogeneous acid catalysts. By exploring a range of aqueous and aprotic mixtures over several temperatures and times, we were able to show constant selectivity to HMF over all temperatures and aqueous mixtures with DMSO and propose a mechanism not fully dependent on the tautomeric distribution. These results also demonstrated that DMSO had a unique stabilization effect that hindered unwanted side reactions. Additionally, we were able to show DMSO decomposition to acidic species in the presence of air, an aspect that had been continuously overlooked in the past. Based on these new findings, fructose dehydration in aqueous mixtures of DMSO without acid catalyst addition was compared under aerated and de-aerated conditions. The resulting 100% fructose conversion to HMF under aerated conditions at 150oC for 30 min demonstrated that acids formed by DMSO decomposition and not reaction catalysis by DMSO drove fructose dehydration in DMSO solvent systems. Several 10 mM homogeneous mineral acids were then employed for fructose dehydration at 120oC in de-aerated solutions. The results showed that acid type and loading did not influence HMF selectivity, thereby demonstrating that nucleophilic stabilization of carbocations did not influence rate or selectivity. Through analyzing fructose dehydration kinetics by homogeneous acids in aqueous-DMSO mixtures, we identified that the additive selectivity of fructose dehydration to DFA and HMF was roughly constant at 83% over increasing water contents and that DFA selectivity dropped with increasing temperature and conversion. The result was 83% HMF yields in aqueous solutions with >50 vol% DMSO co-solvent at optimal reaction conditions of 150oC for 30 mins with 25 g/L loading of fructose. These results clarified a number of mechanistic issues that were previously not well understood for this reaction. The flat selectivity vs. conversion profiles demonstrated that selectivity is thermodynamically controlled by reaction equilibration before the rate limiting carbocation dehydration steps. Furthermore, the poor selectivity dependence on reaction temperature suggested that selectivity is controlled by a combination of the fructose tautomer distribution and an internal equilibrium between the furanose and pyranose forms of the carbocation intermediate. These results also identified that DMSO only acts as a solvent that facilitates the fructose conversion reaction and that DMSO enhances HMF selectivity by stabilizing HMF from rehydration and increasing the tautomeric distribution of the fructose furanose form. Because combining 2 or more steps for a multistep process in a one-pot reactor can reduce total capital and operating costs, operating strategies that combine fructose dehydration to HMF and HMF hydrogenation to DMF in one reactor are under investigation. To merge these two steps, the solvent applied must be able to achieve high DMF yields from the combined operation. Our recent experimental work showed highly selective production of DMF from HMF by hydrogenation catalysts in dioxane mixtures, but acid dehydration of fructose to HMF suffered from poor yields in dioxane (<30%). In contrast, high HMF yields (~83%) were realized in aqueous mixtures of dimethyl sulfoxide (DMSO). To mimic the solvent effects of DMSO, the surfaces of heterogeneous acid catalysts were functionalized with 1, 3-methyl propyl sulfoxide. These sulfoxide groups were grafted near Bronsted acid sites on the catalyst surface to generate a surface environment similar to that of homogeneous Bronsted acid catalysts in the presence of DMSO. The resulting surface environment is expected to be able to match HMF yields produced in aqueous mixtures of DMSO, while aqueous mixtures of dioxane achieve high DMF yields.
Publications
- Type:
Journal Articles
Status:
Other
Year Published:
2017
Citation:
Svenningsen G*, Kumar R, Wyman CE, Christopher P. 2017. Identifying the Mechanism of Fructose Dehydration to 5-Hydroxymethylfurfural in Aqueous Dipolar Aprotic Solutions by Homogeneous Acids, in progress.
|
Progress 12/15/14 to 12/14/15
Outputs
Target Audience:The target audiences include academics seeking to advance the knowledge of biological conversion systems and train students, government researchers who seek to understand such systems and development policy, industrial practitioners who seek to apply advanced technologies for conversion of cellulosic biomass, and the general public who wish to better understand the potential for production of low cost liquid fuels sustainably. In this reporting period, thegraduate students presented their data at the inter-departmental colloquium held at the Bourns College of Engineering, University of California, Riverside.Their progress was alsopresented by the students at the 2015 Annual AIChE meeting in Salt Lake City, UT in Novmber 2015. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Two graduatestudents are being mentored by the PI,Co-PIs, and other senior graduate students, who have developed expertise in biomass conversion,in developing fundamental understanding and practical skills towards the overall objectives of the project. How have the results been disseminated to communities of interest?The graduate students presented their data at the inter-departmental colloquium held at the Bourns College of Engineering, University of California, Riverside.Their progress was alsopresented by the graduate students at the 2015 Annual AIChE meeting in Salt Lake City, UT. What do you plan to do during the next reporting period to accomplish the goals?High yield production of fructose from corn stover is the ultimate goal. To achieve this goal, pretreatment conditions to maximize glucan and xylan recovery will be defined by evaluating the effect of the following parameters on the performance of CELF pretreatment: solids loading, temperature, solvent to water ratio, and reaction time. Pretreatment performance over a range of these conditions will be screened using 10 mL Hastelloy tubular reactors followed by operation in a 1L Parr reactor to confirm results at the best conditions and produce more materials for evaluation. The glucan-enriched solids that are highly digestible will then be employed in the SSI environment. Clostridium thermocellum will be paired with a thermophilic anaerobe that produces β-glucosidase to define conditions that achieve high glucose yields. The two anaerobic bacteria of primary interest for the latter are Thermoanaerobacterium thermosaccharolyticum and Thermoanaerobacter brockii. Once the proper combination of biological systems and their operating conditions are defined, we plan to employ commercially available immobilized glucose isomerase in combination with the two organisms to develop the integrated SSI system. Definition of the best pH and temperature conditions will be determined using 200 mL pressure bottles. In addition, we plan to explore heterogeneous catalytic systems to shift the glucose to fructose equilibrium. Because HMF has been generated at highest yields utilizing homogenous DMSO acid catalysts but recovery of DMSO for recycle is expensive, novel heterogeneous catalysts will be synthesized to mimic high HMF yielding, homogeneous systems at the surface of catalytic supports that capture DMSO for repeated use. Thus, the objective is to maximize HMF yields from aqueous co-solvent systems while avoiding DMSO recovery limitations. This type of environment can be created on the surface of heterogeneous catalysts by grafting different functional groups that create cooperative effects that enable high yields in low-boiling point, aqueous co-solvent systems. Our future work will focus on optimizing catalytic effects in tetrahydrofuran/aqueous co-solvent systems to maximize HMF yields while reducing side reactions. In addition, we plan to publish two review papers in high ranked journal.
Impacts
What was accomplished under these goals?
This project proposes to couple thermochemical, biological, and catalytic transformational technologies to convert inexpensive, abundant cellulosic biomass into drop-in fuel additives compatible with the existing transportation infrastructure. The approach integrates 1) our recently invented tetrahydrofuran (THF) co-solvent enhanced lignocellulosic fractionation (CELF) technology for high yield production of furfural and glucan rich solids with 2) a novel combination of anaerobic enzyme production and glucose isomerization to convert co-solvent enriched glucan solids into sugar isomers 3) for dehydration to 5-hydroxymethylfurfural (5-HMF) 4) that along with furfural is catalytically converted into drop-in fuel components. The goal is to optimize this integrated process to achieve higher yields of fuels from biomass than by existing routes. This outcome is possible because our unique CELF technology removes virtually all the hemicellulose and about 90% of the lignin from cellulosic biomass while realizing higher furfural yields than possible by alternate routes. Anaerobically produced enzymes readily breakdown the glucan in the solids produced by CELF to achieve nearly theoretical yields of glucose in 24 h or less. Novel technology will be developed to simultaneously isomerize the glucose released into fructose with yields of about 90% of theoretical. Fructose in turn can be converted to HMF with much higher yields than possible from glucose. Heterogeneous catalysts react the volatile HMF and furfural to hydrocarbon fuels that are fungible with the existing fuel infrastructure. The resulting novel technology can be a major step forward for aqueous processing of biomass to fungible fuels. The primary focus to this point has been on developing a novel Simultaneous Saccharification and Isomerization (SSI) approach for fructose production from pure Avicel® PH101 cellulose as a model system. SSI was first applied using Novozyme®188 β-glucosidase (β-g), Clostridium thermocellum, and immobilized GI. Bottles were loaded with 5 g glucan/L, sealed, crimped, sterilized, and purged with nitrogen. Four mg of immobilized GI was then added to each bottle. Sterilized MTC media was added to the bottles followed by 1 mL of C. thermocellum cultured from stock solution and membrane filtered β-g at 30 mg protein/g glucan enzyme loading. All injections were over an open flame. The bottles were incubated at 60°C and shaken at 180 rpm. Based on samples taken from the broth at 16, 24, and 30 h, a maximum fructose yield of about 8% was realized at 24 h. The low yields to date were most likely due to incompatibility of all the enzymes and bacteria as β-g has a different optimum pH (4.8-5.0) than GI and C.thermocellum (6.0-7.0). Additionally, β-g appeared to inhibit C. thermocellum growth. For these reasons, different β-g production methods are being devised. The performance of GI with buffers at different pHs was also investigated. Because the enzyme performed better with MOPS than sodium citrate buffer, it appears that buffer and pH play key roles in GI's performance. Thus, the right buffer solution for the reaction is key to high fructose yields. Using immobilized GI results in a near 50-50 solution of glucose and fructose; however, our literature survey showed that sodium tetraborate in solution can shift equilibrium towards fructose. Thus, the impact of sodium tetraborate on C. thermocellum was investigated with a qualitative experiment in which C. thermocellum was grown on Avicel and then inoculated with sodium tetraborate. However, sodium tetraborate had an adverse effect on C. thermocellum in that Avicel, which is normally consumed by the bacteria within 48 h, did not disappear after a week. Thus, application of a heterogeneous catalyst is planned to shift the equilibrium to fructose. Two primary catalytic steps were proposed to transform sugars to 2,5-dimethylfuran (DMF), a fuel additive, with our current research focused on transforming fructose to 5-HMF. Numerous publications report HMF yields upwards of 100% utilizing a solvent known as dimethyl sulfoxide (DMSO). However, there has been much debate as to whether this solvent acts catalytically. Additionally, interactions between water, air, and DMSO are still not fully understood, leading to a lack of understanding in this field. On this basis, we tackled understanding this system from a chemical engineering perspective. Fructose (5 wt%) dehydration reactions were performed by loading Hastelloy tube reactors filled with varying ratios of DMSO (purity 99.8%) and distilled water into our unique steam chamber that is rapidly heated by a 75 kW electric steam boiler. HMF production did not improve until we increased the DMSO content to 75% or greater, and 95% DMSO drove the yields to a maximum that dropped at 99.8% DMSO. Although this outcome suggests that water is involved in fructose dehydration, many reports point out that water is the main contributor to lowering yields through generating levulinic acid, formic acid, and humins. Our goal is to determine the ratio of water to organic solvents in our systems that will balance these competing reactions. We desire aqueous systems as fructose is highly water soluble and water-based systems promote greener chemical processing. However, the high DMSO boiling point of 186°C leads to a greater cost for separating HMF and DMSO from more volatile water, and apparently this separation cost is the primary reason DMSO systems are not commercially used. On the other hand, heterogeneous catalysts are easily separated from liquid systems and can be utilized with lower-boiling point solvents to overcome separation and recycle issues. Thus, we are working on a heterogeneous catalysis approach based on the knowledge that DMSO has shown the highest HMF yields. To overcome the DMSO recovery challenge, we are developing a protocol for attaching DMSO-like functional groups onto the surface of heterogeneous metal oxide supports. This protocol began with silanization of metal oxides with organic silane compounds, in our particular case, mercaptopropylsilane (MPS). MPS was slowly added overnight under flowing nitrogen via a syringe pump to a stirred solution of toluene, pyridine, and silica gel at 115°C to allow the surface to be evenly distributed with 3-mercaptoproyl functional groups. We then reacted these groups with iodomethane in DMF to form 3-methylpropylsulfide groups. Finally, the sulfide group was oxidized to form a 3-methylpropylsulfoxide group that mimics DMSO. The characterization of these groups was performed by Fast-Fourier Transform Infrared spectroscopy (FTIR). From the 3-mercaptopropyl functionalized silica, the -CH2-S-H thiol stretching peak was observed at 2478 cm-1, and the -CH2-S- asymmetric and symmetric stretches to be 2950-2920 cm-1 and 2880-2845 cm-1, respectively. These results were confirmed by peaks at 2936 and 2847 cm-1. It was also expected to see -CH2- deformation peaks and strong wagging peaks at 1435-1410 cm-1 and 1270-1220 cm-1 that was confirmed by a peak at 1408 cm-1. As expected for 3-Methylpropylsulfide functionalized silica gel, CH3-S- asymmetric stretch and symmetric stretch peaks were seen at 2988 and 2917, respectively, confirming successful substitution of hydrogen with a methyl group. Lastly, because sulfide oxidization does not notably change CH3-S- or -CH2-S-, oxidization was primarily confirmed by a peak at 1030 cm-1 that is representative of the typical position for a sulfoxide peak. Promising initial results were obtained in initial trials of this catalysts in 1:1 THF/water systems at 150°C for one hour as a ten-fold increase in HMF yield compared to a system without catalyst. Optimizations are now underway to further increase HMF yields.
Publications
|
|