Chemical Recovery of γ-Valerolactone/Water Biorefinery

We introduce the optimization of the pulping conditions and propose different chemical recovery options for a proven biorefinery concept based on γ-valerolactone (GVL)/water fractionation. The pulping process has been optimized whereby the liquor-to-wood (L:W) ratio could be reduced to 3 L/kg without compromising the pulp properties as raw material for textile fibers production. The recovery of the pulping solvent was performed through combinations of lignin precipitation by water addition, distillation at reduced pressure, and liquid CO2 extraction. With a two-step lignin precipitation coupled with vacuum distillation, more than 90% of lignin and GVL could be recovered from the spent liquor. However, a significant part of GVL remained unrecoverable in the residue, which was a highly viscous liquid with complicated phase behavior. The recovery by lignin precipitation combined with liquid CO2 extraction could recover more than 85% GVL and 90% lignin without forming any problematic residue as in the distillation process. The remaining GVL remained in the raffinate containing a low amount of lignin and other compounds, which can be further processed to isolate the GVL and improve the recovery rate.


Carbohydrate, lignin, furanic compounds and organic acids analyses of solid and liquid samples
The carbohydrate and lignin contents in the pulp and lignin samples were analyzed in accordance to the 2-step hydrolysis method described in the NREL/TP-510-42618 standard. The pulp was firstly hydrolyzed in 72% H 2 SO 4 , with an acid-to-pulp ratio of 10 mL/g, at 30°C, for 60 minutes. The hydrolyzed suspension was subjected to the second hydrolysis in 4% H 2 SO 4 , with an acid-to-material ratio of 300 mL/g, at 121°C, for 60 minutes. The hydrolyzed suspension was then filtered through a Robu® glass crucible (porosity 4). The monosaccharide content in the filtrate was determined by high performance anion-exchange chromatography (HPAEC) in a Dionex™ ICS-3000 device. The content of furanic compounds (furfural and HMF) and organic acids (formic acid, acetic and levulinic acid) in the liquid samples was determined by high performance liquid chromatography (HPLC) in a Dionex UltiMate 3000 device. The HPLC system was equipped with a UV diode array detector and a Rezex™ ROA-Organic Acid H+ (8%) LC column (7.8 mm × 300 mm). The UV detection wavelength was 210 nm and 280 nm for organic acids and furanic compounds, respectively. The column and detectors were at 55ºC. The eluent was 0.0025 mol/L sulfuric acid with the flow of 0.5 mL/min. The samples were filtered through a 0.45 μm syringe filter before the analysis.

Molecular mass determination for lignin samples
The molecular mass distributions, the number and weight average molecular masses (M n , M w , respectively) of the lignin samples were determined by gel permeation chromatography (GPC) in an Agilent 1100 HPLC/VWD device. The GPC system was equipped with a UV detector, a Phenogel™ pre-column (7.8 mm × 50 mm, particle size 5 μm) and two Phenogel™ size exclusion columns of styrenedivinylbenzene with pore sizes 50 and 1000 Å (7.8 mm × 300 mm, particle size 5 μm). The UV detection wavelength was 260 nm and 280 nm for the standards and samples, respectively. The column and detector were at 35ºC. The eluent, also the lignin solvent, was LiChrosol®-grade tetrahydrofuran (THF) with the flow of 1 mL/min. Lignin was not acetylated before the GPC analysis. The samples were prepared in THF with a concentration of about 2 mg/mL and filtered through a 0.45 μm syringe filter before the analysis. Calibration was performed with two standard solutions, one containing toluene, syringol, 2,2'dihydroxybiphenyl, PS474, PS3470 and PS76600, and the other one containing toluene, polystyrene dimer PS208, PS1270, PS7000 and PS18200. The standard PS474 was divided by the columns to several oligomer peaks, of which polystyrene trimer PS312, tetramer PS417 and pentamer PS521 were included in the actual calibration. A molecular weight cut-off at 201 g/mol (one assumed phenylpropane unit in the eucalyptus GVL lignin) 3 was used in processing the results.

GVL/water ratio determination for liquid samples
The GVL/water mass ratio in the liquid samples was determined by gas chromatography (GC) in an Agilent 6890N device coupled with 7683 Series liquid injector. The GC system was equipped with a thermal conductivity detector (TCD) and a polar capillary Agilent DB-WAXetr column (0.32 mm × 30 000 mm, film thickness of 1 μm). Solid particles (mostly lignin) were separated from the analyzed mixture by the glass wool liner at the GC inlet. The GC inlet temperature and the split ratio was 250°C and 10:1, respectively. Helium was the carrier gas with an initial average velocity of 29 cm/s at constant flow mode. The GC oven temperature started at 80°C for 5 minutes, then it was raised with a rate of 60°C/min to 140°C and held for 2 minutes; after that, the temperature was raised with a rate of 60°C/min to 200°C and held for 6 minutes. The detector temperature was at 250°C. The method can quantify water, GVL, acetic acid, formic acid and furfural in the liquid samples. However, due to the closeness of their retention time on the chromatogram, furanic compounds and organic acids were analyzed by HPLC as described earlier, while the GVL/water mass ratio was determined by the GC method.
The gravimetric samples were prepared to calibrate the response factors with acetone as the internal standard. The response factors of TCD were calculated as in equation (1).
(1)  The energy consumption for the recovery of GVL by distillation at reduced pressure and liquid CO 2 extraction was estimated by simulation models constructed in the ASPEN PLUS v.10 environment. For simplification, only the main components (GVL-water-CO2 for extraction and GVL-water for distillation) were included. Dissolved compounds (lignin, carbohydrates and furanic compounds) were not considered. For comparability purpose, similar separation capacity was targeted in both models, which are >98 wt% purity of GVL in the organic phase and >99 wt% purity of water in the aqueous phase. Spent liquor containing 50 wt% GVL and 50 wt% water was continuously feed at a rate of 1 kg/s. UNIQUAC and SRK thermodynamic models were used for the liquid and vapor phase, respectively, in the distillation column, flash tanks and compressor. Fixed K-factors obtained from the experimental data were used in the extraction unit.

Distillation at reduced pressure
The spent liquor was firstly evaporated at reduced pressure. In a real process, this step prevents the lignin from entering the distillation column.

Liquid CO 2 extraction
The spent liquor was extracted with CO 2 (solvent-to-feed ratio of 0.45 : 1, i.e. twice the minimum ratio which is obtained by the modified McCabe-Thiele graphical method) in a three-staged extraction unit.
The extractor was modelled as three decanters with ideal phase equilibrium stages, operating isothermally and connected in counter-current mode. The distribution coefficients of the components (K i = x i, extract / x i, raffinate ) were fixed for each component within each ideal phase equilibrium. The K-values were calculated from the measured phase equilibrium 1 . The simulation model took into account the mole balance, phase equilibrium and energy balance.