Research ArticleAPPLIED SCIENCES AND ENGINEERING

# Increasing the revenue from lignocellulosic biomass: Maximizing feedstock utilization

See allHide authors and affiliations

Science Advances  19 May 2017:
Vol. 3, no. 5, e1603301

## CONCLUSIONS

Production of biorenewable chemicals and fuels that are competitive with their petroleum-derived equivalents requires the generation of revenue from each of the fractions present in lignocellulosic biomass, that is, cellulose, hemicellulose, and lignin. Because of the different structures and chemistries of these components, an effective fractionation step is critical to separate each component without diminishing the value of any component. GVL has proven to be an excellent solvent to fractionate lignocellulosic biomass. The hemicellulose and lignin fractions can be solubilized in a GVL/water/acid mixture, leaving behind a high-purity cellulose stream that can be used as dissolving pulp to produce fibers. Because of the mild processing conditions that can be used to solubilize the lignin and the hemicellulose, the lignin retains its native structure and can be isolated at high purity. The hemicellulose can be converted into furfural at high yields without requiring the purification of the sugars, thus reducing processing cost. By converting lignocellulosic biomass into dissolving pulp, furfural, and lignin, we can obtain revenues of about \$500 per MT of dry biomass, which makes our technology economically competitive.

## MATERIALS AND METHODS

### Chemicals and materials

We acquired the GVL (>98% purity) from Shenzhen Nexconn Pharmatechs Ltd. All other reagents used were analytical grade, and we used them without further purification. We purchased sulfuric acid (>96%), glucose (>99%), xylose (>99%), mannose (>99%), galactose (>99%), fructose (>99%), arabinose (>99%), furfural (>98%), levulinic acid (>97%), formic acid (>99%), acetic acid (>99%), HMF (>98%), potassium iodide (ACS reagent), and sodium thiosulfate (1 N) from Sigma-Aldrich. We purchased potassium permanganate (0.100 N) and CED (1.0 M) from Ricca Chemical Company. We purchased Whatman filter paper (grade 42; ashless; diameter, 90 mm) from GE Healthcare. We acquired the white birch chips (approximately 2.54 cm × 2.54 cm × 0.65 cm) from Flambeau River Papers. We used white birch chips for the large-scale fractionation experiments. For the laboratory experiments, we used smaller chips created using a DR model CST1450-CHP knife mill fitted with a 1/8 screen. The chips were then screened with an 8-mm sieve screen to remove large pieces.

### Biomass fractionation and dissolving pulp production

We performed the biomass fractionation experiments using 60-ml and 1- and 20-liter reactors (see sections S1 to S3 for a detailed description of each reaction setup and experimental procedure). Typically, we added the biomass into the reactor and the GVL/water/acid solution in the proportions indicated in each experiment. We heated the reactors externally. At the end of the experiment, we separated the liquid from the solids by filtration and washed the solids, first with a GVL/water mixture and later with hot water.

We determined the pulp moisture content by placing the sample in an oven at 105°C following TAPPI-T258 om-02. We measured kappa numbers using TAPPI Standard Test Method TAPPI-UM246 (Micro-Kappa number). We measured the cellulose chain lengths in the pulps by measuring CED viscosity using TAPPI-T230 om-04 [viscosity of pulp (capillary viscometer method)]. We measured the alpha-cellulose content of the bleached pulp by TAPPI-T203 (alpha-, beta-, and gamma-cellulose in pulp). We determined the chemical composition of the solid residue following the National Renewable Energy Laboratory (NREL)/TP-510-42618, “Determination of structural carbohydrates and lignin in biomass.” We analyzed the liquid composition by high-performance liquid chromatography (HPLC) (Waters 2695 system with a Bio-Rad Aminex HPX-87H column) equipped with a refractive index (RI) 410 detector (for carbohydrates, organic acids, and GVL) and an ultraviolet (UV) detector (for furfural and HMF). We used a protocol similar to NREL/TP-510-42618 to convert soluble oligomeric carbohydrates into monomers. We diluted the liquid 10 times with 4.4 wt % sulfuric acid and heated at 120°C for 1 hour before analyzing the samples. We bleached the samples using a modified version of the standard OD(EP)D sequence (see section S3 for further details).

### Lignin production and characterization

We performed the lignin production, characterization, and upgrading studies using a GVL/liquid solution produced in a 20-liter twin digester reactor (see sections S3 and S4 for details). The reaction conditions to produce the lignin were 80:20 (w/w) GVL/water, 0.1 M sulfuric acid, 60 min at 130°C, and 0.21 to 0.24 MPa. We precipitated the lignin from the GVL/water solution by adding additional water. We recovered the lignin by centrifugation. We washed the lignin with water to remove the small amounts of GVL present and to minimize GVL interferences during the characterization.

We analyzed the chemical composition of the lignin following the NREL/TP-510-42618 protocol. We analyzed the molecular weight after an acetobromination derivatization procedure using a size exclusion chromatography system (Tosoh EcoSEC) equipped with a UV detector (265 nm) (see section S6A for complete details of the lignin characterization).

We determined the glass transition temperature (Tg) using a PerkinElmer Diamond differential scanning calorimeter. We conducted the experiments under nitrogen atmosphere using approximately 2 mg of lignin samples in triplicate. We first heated the lignin from 25° to 140°C at a heating rate of 100°C/min and held at that temperature until the change in thermal energy from the sample was zero to expel any moisture and to erase thermal history of the sample. We then cooled the sample and reheated it to 200°C at the same heating rate to calculate Tg. We measured the thermal decomposition of the lignin samples using a PerkinElmer Pyris 1 thermogravimetric analyzer. We conducted the experiments in duplicate using approximately 5 mg of specimens, heated from 20° to 105°C under a nitrogen atmosphere at a rate of 10°C/min, and held for 10 min to remove moisture. We then heated the sample to 950°C at the same heating rate.

We carried out nuclear magnetic resonance (NMR) analysis on a Varian 400-MR spectrometer. For 2D [heteronuclear single-quantum coherence (HSQC)] spectroscopy, we dissolved 100 mg of lignin in 0.75 ml of d6-DMSO; all samples were fully soluble. We recorded the NMR spectra at 25°C using the (HC)bsgHSQCAD pulse program. We used as internal reference the experiment, collecting 32 transients and 512 time increments in the 13C dimension. We performed the analysis of the data using Mnova version 10.0.1 (see section S6C for complete details of the NMR analysis and quantification).

### Carbon foam production

We produced the carbon foam and nanographitic carbon-carbon composites using lignin produced in the 20-liter reactor (see sections S3 to S5 for detailed lignin production and section S6 for detailed lignin characterization). We exposed the dry GVL-lignin powder through a series of heat and processing treatments to make a highly graphitic structure. The heating steps included a combined thermal stabilization/pyrolysis step followed by passivation. We first exposed the GVL-lignin powder to nitrogen (N2) with water vapor (H2O) in a tube furnace for thermal stabilization and pyrolysis. We passed the inert gas, N2, through a bubbler at room temperature and fed into the furnace at a volumetric flow rate of 3 liter min−1. During the thermal stabilization and pyrolysis step, we heated the furnace from 30° to 1000°C at a rate of 10°C min−1 and then held at 1000°C for 1 hour, producing a carbon foam (see section S7 for complete details on carbon foam production).

### Furfural production

We performed the initial experiments to study the production of furfural in the GVL/water solution in 10-ml high-pressure glass reactors (Alltech). We loaded the reactors with a solution of 80:20 (w/w) GVL/water, commercial xylose, and the catalyst. We mixed the reactors with a magnetic bar and heated to the reaction temperature in an oil bath. We analyzed the liquid by HPLC (Waters 2695 system with a Bio-Rad Aminex HPX-87H column), RI 410 detector (carbohydrates, organic acids, and GVL), and a UV detector (furfural and HMF). We studied the continuous production of furfural in an upflow reactor. We used the liquid solution recovered after the lignin precipitation to produce furfural. We concentrated the solution by evaporating the water at 100°C to a GVL/water ratio of 80:20 (w/w) before the experiment. We fed the liquid, which contained soluble C5 carbohydrates, through the flow reactor using an HPLC pump (LabAlliance Series I). The reaction conditions were 225°C for 1 min and 2.1 MPa. We used the residual sulfuric acid in the solution from the fractionation step as the catalyst for furfural production (see section S8 for a complete description of the reactor setup and experimental procedure).

### Techno-economic model

We developed a process model, based on the experimental data, to study the techno-economic feasibility using the Aspen Plus process simulator (V7.3, Aspen Technology). We scaled the storage, utilities, and wastewater treatment based on the NREL model (30). We performed the equipment sizing and cost analysis using the Aspen Process Economic Analyzer (V7.3, Aspen Technology) based on the simulation results (see section S10 for a complete description of the economic parameters and assumptions).

## SUPPLEMENTARY MATERIALS

Supplementary Materials and Methods

section S1. Biomass fractionation

section S2. Liquid/solid separation

section S3. High-purity cellulose production for dissolving pulp

section S4. Lignin recovery and precipitation experiments

section S5. Lignin recovery for carbon foam and battery anode production

section S6. Lignin characterization

section S7. Carbon foam production

section S8. Furfural production

section S9. Solvent stability study

section S10. Techno-economic model

fig. S1. Fractionation of white birch.

fig. S2. Lignin produced from white birch fractionation.

fig. S3. Lignin produced from white birch fractionation.

fig. S4. Thermal decomposition of lignin samples produced from white birch fractionation at 70:30 GVL/water, 125°C, 3 hours, and 0.1 M sulfuric acid and 80:20 GVL/water, 130°C, 1 hour, and 0.1 M sulfuric acid.

fig. S5. Aromatic (left) and side chain (right) region of the 13C-1H (HSQC) spectra of white birch, lignin samples produced from white birch fractionated at 70:30 GVL/water, 125°C, 3 hours, and 0.1 M sulfuric acid and 80:20 GVL/water, 130°C, 1 hour, and 0.1 M sulfuric acid.

fig. S6. X-ray diffraction patterns of carbonized and passivated GVL lignin compared to the original lignin (top).

fig. S7. Carbon foam produced from lignin (initial white birch fractionation conditions: 80:20 GVL/water, 130°C, 1 hour, and 0.1 M sulfuric acid).

fig. S8. Production of furfural in a batch reactor using 10 wt % xylose as feedstock.

fig. S9. Production of furfural in a continuous flow reactor using xylose as feedstock.

fig. S10. Gas chromatography–mass spectrometry (GC-MS) chromatogram of GVL (top), tetrahydrofuran (THF) (middle), and ethanol (bottom) after 12 hours at 130°C.

fig. S11. GC-MS chromatogram of GVL (top), THF (middle), and ethanol (bottom) after 12 hours at 130°C.

table S1. White birch fractionation optimization.

table S2. Properties of bleached high-purity cellulose sample from white birch wood chips to produce dissolving pulp.

table S3. Lignin solubility in GVL/water mixtures at room temperature.

table S4. Effect of water/GVL ratio on lignin precipitation.

table S5. Chemical composition and molecular weight of lignin samples.

table S6. Thermal properties of lignin samples.

table S7. Summary of NMR analysis of lignin samples.

table S8. Hydroxyl group content of lignin sample (70:30 GVL/water, 125°C, 3 hours, and 0.1 M sulfuric acid) obtained by quantitative 31P NMR spectroscopy (mmol/g).

table S9. Liquid composition after 12 hours at 130°C.

table S10. Liquid composition after 12 hours at 130°C.

table S11. List of economic parameters and assumptions.

table S12. Mass and energy balances (basis: 2000 tons of white birch per day).

table S13. Energy requirements before and after heat integration (basis: 2000 tons of white birch per day).

table S14. Capital and operating costs (basis: 2000 tons of white birch per day).

table S15. Process model development details.

References (3243)

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited.

## REFERENCES AND NOTES

Acknowledgments: Funding: This work has been funded in part by the NSF Small Business Innovation Research 1602713 and 1632394, and this material is based in part upon work supported by the U.S. Department of Energy (DOE) Great Lakes Bioenergy Research Center (DOE Office of Science Biological and Environmental Research; DE-FC02-07ER64494). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NSF or DOE Great Lakes Bioenergy Research Center. D.M.A. and S.H.H. thank Glucan Biorenewables LLC for its funding support. Author contributions: All the authors contributed to the design and optimization of the process, discussion of the experimental results, and write-up of the article. S.Z., C.J.H., and T.R. performed the biomass fractionation and bleaching experiments. W.W., K.H., and C.T.M. performed the techno-economic analysis. O.H., J.T., V.G.-N., N.L., and D.P.H. performed the lignin characterization and carbon foam production. A.H.M., M.A.M., S.H.H., D.M.A., and J.A.D. performed the lignin separation and furfural production experiments. S.H.H., D.M.A., and J.A.D. performed the biomass fractionation and separation experiments, lignin production and separation studies, furfural production experiments, and GVL recovery and stability studies and did the overall process design. Competing interests: D.M.A. and S.H.H. are employees of Glucan Biorenewables LLC, which is commercializing the technology. J.A.D. is a co-founder of Glucan Biorenewables LLC. J.A.D. and D.M.A. are authors on several patents related to this work held by the Wisconsin Alumni Research Foundation and Glucan Biorenewables LLC (US9359650 B2, US8772515 B2, US8962867 B2, and US8399688 B2). J.A.D. is an author on an additional patent held by the Wisconsin Alumni Research Foundation (US9045804 B2). D.M.A. and S.H.H. are authors on a patent application filed by Glucan Biorenewables LLC (15/216,363, filed 21 July 2016). D.P.H. and V.G.-N. are authors on a provisional U.S. patent filed by the University of Tennessee Research Foundation (62/359,754, filed 8 July 2016). D.P.H., O.H., and N.L. are authors on a patent application filed by the University of Tennessee Research Foundation (PCT/US2015047423, filed 29 August 2015). Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Additional data related to this paper may be requested from the authors.
View Abstract