High-purity lignin isolated from poplar wood meal through dissolving treatment with deep eutectic solvents

Deep eutectic solvents (DESs) have potential applications in biomass conversion and green chemicals due to their cost-effectiveness and environmentally friendly properties. This study reports on a feasible method of using DESs for lignin selective extraction from poplar wood meal. DESs obtained from various hydrogen-bond donors and acceptors were used to evaluate the dissolving capacity of lignin from poplar wood meal. Among the various DESs, lactic acid: choline chloride (9 : 1) exhibits the optimal extraction capacity, which is capable of selectively dissolving 95% of lignin from poplar wood meal at 120°C for 6 h. The purity of isolated lignin reaches 98% after regeneration in water. From Fourier Transform-IR, nitrobenzene oxidation and nuclear magnetic resonance analysis, the results demonstrate that the DESs can selectively cleave ether linkages and damage the non-condensation section of lignin, thereby facilitating lignin dissolution from wood meal. Thus, this study provides a promising route for the extraction of high-purity lignin from biomass materials.

ZW, 0000-0003-0201-7220 Deep eutectic solvents (DESs) have potential applications in biomass conversion and green chemicals due to their costeffectiveness and environmentally friendly properties. This study reports on a feasible method of using DESs for lignin selective extraction from poplar wood meal. DESs obtained from various hydrogen-bond donors and acceptors were used to evaluate the dissolving capacity of lignin from poplar wood meal. Among the various DESs, lactic acid: choline chloride (9 : 1) exhibits the optimal extraction capacity, which is capable of selectively dissolving 95% of lignin from poplar wood meal at 1208C for 6 h. The purity of isolated lignin reaches 98% after regeneration in water. From Fourier Transform-IR, nitrobenzene oxidation and nuclear magnetic resonance analysis, the results demonstrate that the DESs can selectively cleave ether linkages and damage the non-condensation section of lignin, thereby facilitating lignin dissolution from wood meal. Thus, this study provides a promising route for the extraction of high-purity lignin from biomass materials. by centrifugation at 5000 r min 21 for 10 min. The solid residue was washed using water and freeze-dried to evaluate the solubility of each DES. The solubility of DESs for poplar wood meal was called the dissolving removal rate (S) where m 1 is the mass of solid residue, and M is the mass of raw material.

Isolation of lignin with a lactic acid/choline chloride DES
The process of lignin isolation using a LA/ChCl DES is shown in figure 1. The wood meal (2 g, oven dry) and LA/ChCl (98 g) were added into a conical flask with continuous stirring. The dissolving temperatures were 1008C, 1108C, 1208C and 1308C, and the dissolving times were 3 h and 6 h, respectively. After dissolving, the solution and solid residue can be separated by centrifugation. The solid residue was washed using acetone three times followed by washing using water and freeze-dried to obtain separated cellulose (DES-C, figure 1). The acetone soluble fraction was mixed with the solution part, and it removed acetone by rotary evaporation. The mixtures (acetone free) were regenerated by deionized water and freeze-dried to obtain the isolated lignin (DES-L, figure 1), while hemicellulose still remains in the solution (DES-H, figure 1). The yield of solid residue (DES-C) and isolated lignin (DES-L) were calculated by the following equation: where m 2 is the mass of DES-C or DES-L, and M is the mass of raw material. The solubility (S) of components in raw materials is determined based on the content of each component (cellulose, lignin and hemicellulose) in DES-C, as shown in the following equations: The solubility of cellulose (SC): where OC, OL and OH are the respective masses of cellulose, lignin and hemicellulose in the original materials, and RC, RL and RH are the respective masses of cellulose, lignin and hemicellulose in solid residue (DES-C).

Lignin purity determination
Lignin is the total amount of acid soluble lignin (ASL) and acid insoluble lignin (AIL) in regenerated products (DES-L), according to standard methods [10]. The ASL was determined by UV-absorbance at   royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 181757 205 nm, which was performed on a UV-vis spectrophotometer (TU-1810, Puxi, Beijing, China). The content of sugar was measured by high-performance liquid chromatography (HPLC, Agilent 1200 Series, Santa Clara, CA). The lignin purity (PL) was calculated using the following: where RL is the mass of lignin in regenerated products (ASL þ AIL), and the R is the mass of regenerated products.

Characterization of isolated lignin (DES-L) and solid residue (DES-C)
The FT-IR spectra of DES-L and DES-C were obtained from the range of 4000-400 cm 21 on a VERTEX 80 V spectrometer (Bruker, Germany). The nitrobenzene oxidation (NBO) of lignin was determined based on the method of Chen [11]. The milled wood lignin (MWL) was prepared as the control sample, which was obtained through the method of Bjö rkman [12].
The structural characterization of purified DES-L and MWL by the 13 C NMR was conducted on a BBO 600 MHz NMR spectrometer following standard procedures for lignin analysis [13]. The acetylated DES-L was analysed using deuterated dimethyl sulfoxide (DMSO-d6) as a solvent (200 mg sample per 0.5 ml DMSO-d6) with a small amount of relaxation agent (Cr(acac) 3 ).
The 1 H-13 C two-dimensional NMRs (2D HSQC NMR) of the purified DES-L and MWL were performed on a Bruker 600 MHz instrument (AVANCE III, Switzerland) equipped with a cryogenically cooled 5 mm TCI z-gradient triple-resonance probe. The DES-L (50 mg) was dissolved into DMSO-d6 solvent (0.5 ml) according to the method of Kim et al. [14]. The spectral widths of 1 H and 13 C were from 0 to 16 ppm (9615 Hz) and from 0 to 165 ppm (24 900 Hz), respectively. The number of collected complex points was 2048 for the 1H-dimension with a recycle delay of 1.5 s. The number of transients was 64, and 256 time increments were recorded in the 13 C-dimension. The chemical peak shift of DMSO (delta C 39.5 ppm, delta H 2.5 ppm) was used to calibrate the data. In this work, the relative proportions of the various joint bonds, methoxy and the S/G ratio of lignin were calculated using the aromatic ring as the internal standard.
The crystal structure of cellulose in DES-C was subjected to X-ray diffraction (XRD) in reflection mode with a diffraction angle of 2u from 58 to 408 by a multifunction X-ray diffractometer (Ultima-IV) at 40 kV and 30 mA. The crystallinity (CrI) of the different samples was calculated using the following: where I 002 is the intensity of the 002 lattice plane at 2u ¼ 22.88, and I am is the intensity from the amorphous phase at approximately 2u ¼ 188 [15].

Solubility of poplar wood meal in different DESs
Firstly, a series of DESs, composed of various hydrogen-bond donors and ChCl, were prepared to find the optimal DESs for the dissolution of poplar wood meal. Figure 2a presents the solubility of poplar wood meal in different DESs under 1208C for 6 h. As shown, the solubility of poplar wood meal in LA/ChCl is higher than that for other DESs, which is more than 50%, thereby suggesting that LA/ ChCl is the optimal DES solvent for wood meal dissolution among these DESs. Furthermore, the effect of the LA/ChCl molar ratio on the solubility of poplar wood meal was investigated and is shown in figure 2b. The solubility of poplar wood meal increased with the LA/ChCl molar ratio (from 1 : 1 to 9 : 1), and LA/ChCl (9 : 1) exhibited the best dissolving capacity for poplar wood meal, which demonstrated LA plays a positive role in the solubility of polar wood meal. The chemical properties of DESs can be described by hydrogen-bond acidity, hydrogen-bond basicity and dipolarity/polarizability of DES solvent. The high hydrogen-bond acidity is conductive to breaking the lignin -carbohydrate complexes (LCCs) structure of wood meal, which is in agreement with Liu et al. [16]. The LA as a HBD mainly affects the hydrogen-bond acidity of DES solvent. The molar ratio of LA/ChCl (9 : 1) had the high hydrogen-bond acidity, enhancing the accessibility of wood meal and inducing disintegration. Furthermore, the utilization of DES can be described by Hole theory [17], which is related to the free volume and probability of finding holes for solvent molecules/ions to move into. Thus, a molar ratio of 9 : 1 was verified to be a priority in all ratios of LA/ChCl, but not 11 : 1. No waste was produced and the atom economy was nearly 100%. Thus, LA/ChCl (9 : 1) was chosen as the solvent system for wood meal and was further studied in the subsequent section. The effects of the dissolving temperature and time with LA/ChCl on the solubility of polar powder are shown in figure 3. The yields of the solid residue (cellulose) after the LA/ChCl dissolving treatment decreased gradually with the increase of the temperature (figure 3a). Under the same LA/ChCl molar ratio, increasing temperature can maximize the ionic characteristics and increase the molecular polarity of the DES, promoting the breakage of the intramolecular hydrogen-bond network and enhancing the solubility of lignin and hemicellulose [6]. Under the same temperature, the solubility of LA/ChCl for poplar wood meal also showed an upward trend with the increase of the dissolving time. The solubility of LA/ChCl for poplar wood meal reached more than 50% at 1308C for 6 h.
The dissolving efficiency of LA/ChCl for each component in the raw material is calculated based on the contents of cellulose, lignin and hemicellulose in the solid residue (DES-C). As shown in figure 3b,c, LA/ChCl possesses better dissolving efficiency for lignin and hemicellulose than cellulose. The dissolving removal rates of lignin and hemicellulose can be improved with the increase of the temperature and time, both of which can reach more than 95%. Furthermore, the LA/ChCl treatment hardly dissolve cellulose when the treatment temperature is less than 1308C. Only a small amount of cellulose is dissolved in LA/ChCl when the temperature is up to 1308C; the cellulose content of 95% could still be retained in solid materials. The results indicated that LA/ChCl has a high affinity to lignin and hemicellulose. In addition, lignin is easy regenerated from hemicellulose by the addition of deionized water. Therefore, lignin with high purity could be acquired from poplar wood meal by the pretreatment of LA/ChCl. The results are similar to those of a previous report [5].  figure 4 that the purity of lignin increases with increase of reaction temperature from 1008C to 1308C which could ascribe to the more serious degradation of hemicellulose under higher temperature. In addition, the purity of lignin in regenerated products can reach more than 95% under the given condition, which suggests that DESs treatment provides a feasible method for the separation and extraction of lignin.

Structural characterization of isolated lignin (DES-L)
The structures of DES-L were characterized by FT-IR, nitrobenzene oxidation, quantitative 13 C NMR, HSQC NMR and 1 H NMR analyses. Milled wood lignin (MWL) made by the same poplar was used as a control sample to compare with the DES-L. Figure 5 shows   The alkaline NBO is used to investigate the condensable connections of lignin [19]. The yields of the S and V units are less than those of the H unit in DES-C, thereby implying that the DES solvent is ready to dissolve the S and V units. Compared with the NBO products of poplar wood meal and DES-L, the NBO products yields of DES-L have a significant reduction, which is attributed to the cleavage of a large amount of noncondensation bonds of lignin, such as a-aryl, alkyl and b-aryl ether bonds. The V units of DES-L decreased, because most of the a-O-4 bond is occupied by the guaiacyl unit, and the phenolic a-O-4 bond can be rapidly degraded at the initial delignification stage [20]. The value of M(S)/m(V) is 1.85, which belongs to the typical guaiacin syringin-based lignin. 13 C NMR spectroscopy is an informative tool for the determination of lignin and its derivative. This method provides not only the information of the phenylpropane units and the side chain linkages but also the different types of hydroxyl groups, including the primary, secondary, C 5 -free and C 5 -substituted hydroxyl groups, if the lignin is acetylated. Figure 6 presents the 13 [4]. The three broad categories of quaternary, oxygenated (C Ar -O), non-oxygenated (C Ar -C) and methine (C Ar -H), in the aromatic region of the lignin 13 C NMR spectra (166-103 ppm), are important to quantify the different condensed moieties, such as the 5-5 0 and 4-O-5 0 linkages of phenolic and etherified types. Compared with MWL, the change of the DES-L signals in the aromatic region suggested that the side chains of lignin, especially the etherified structures, were damaged significantly after the DES treatment [21]. In addition, the absence of peak signals between 102 and 90 ppm, demonstrated that the carbohydrates were absent in DES-L. In other words, DES-L exhibited a high-purity property.
HSQC NMR is another effective tool to probe the structure of lignin. This method can determine the specific carbon-hydrogen functionalities that are unable to be identified in the 13      linkages. However, the signal intensity of the C g -H g correlations in the A' substructure of DES-L was considerably higher than that of MWL, which implied that esterification may occur between the g-OH in the A substructure and LA. Additionally, the signals of the B substructure were scarcely observed in the spectrum of DES-L, but those of the C substructure were almost unchanged compared with the HSQC NMR spectrum of the MWL, thereby indicating the b-b, a-O-g and g-O-a linkages were more stable than the b-5 and a-O-4 bonds in DES. It is intriguing that Hibbert's ketone (HK) linkages were detected in DES-L. In consideration of the analysis of the b-O-4 linkages, it is reasonable to deduce that the HK structure is mainly from the degradation of the b-O-4 structure. In addition, the absence of the carbohydrate signal in the side chain regions of the DES-L suggested that the DES-L possesses high purity, which is in good agreement with the analysis of 13 C NMR. The aromatic regions of MWL and DES-L in the HSQC NMR spectra were similar with the main signals corresponding to guaiacyl and sringyl units. This finding indicates that the fragmentization of lignin is mainly caused by the degradation of the side chains. The prominent signals assigned to p-coumarate were also observed, but the ferulate structure was only slightly detected in MWL and DES-L. During lignification, the p-coumarate and ferulate are frequently reduced to corresponding aldehydes, and these reactions are catalyzed by cinnamate: CoA ligase, which is distributed in various higher plants [22]. However, ferulate just provides the initiation sites from which the lignification event in the cell wall begins and p-coumarate is present via lignification using monolignol conjugation [23]. Therefore, the p-coumarate structure was easily detected, but the ferulate structure was hardly detected in the 2D HSQC spectra.
Besides, as shown in figure 3, almost 95% of cellulose remained in the residue under all the examined conditions, suggesting that DESs are good solvents for the separation of cellulose from poplar wood meal. The property of solid residue (DES-C) from poplar wood meal was further characterized by XRD (electronic supplementary material, figure S1) and FT-IR (electronic supplementary material, figure S2), to verify the effects of DES pretreatment on cellulose. The DES-C had the typical cellulose I structure with characteristic peaks at 2u ¼ 16.38 and 22.58 that were assigned to diffraction planes of 110 and 200 [24], respectively. The cellulose crystallinities of treated DES-C are 72.1, 68.0, 67.2, 71.2 and 64.7 for the treatment temperature/time of 1208C/3 h, 1308C/3 h, 1108C/6 h, 1208C/6 h and 1308C/6 h, respectively. The small difference in cellulose crystallinities suggested that the DES dissolving treatment hardly affected the crystal structure of cellulose under the studied dissolving temperature and time. From the FT-IR spectra, the presence of the adsorption peak of 1738 cm 21 is because the esterification of cellulose and LA at a high temperature created numerous C ¼ O bonds. The adsorption peak at 1633 cm 21 is ascribed to the hydrogen bond formed by the hydroxyl groups  royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 181757 in cellulose. In addition, the peak intensity at 1738 cm 21 increased, and the peak intensity at 1633 cm 21 decreased with the increase in the dissolving time, which further confirmed the esterification of cellulose and LA during treatment. The absence of adsorption peaks at 1595, 1512, 1462 and 1423 cm 21 indicated no benzene ring in DES-C, which suggested that a large amount of lignin can be removed from raw materials through the DES-dissolving treatment.

Conclusion
In this work, a feasible method of using DESs for lignin selective isolation from wood meal was provided. This method exhibited the following features: (i) the DES LA/ChCl has an excellent selective dissolving property for lignin. The optimal dissolving capacity can reach 95% when the molar ratio of LA/ChCl is 9 : 1 at 1208C for 6 h; (ii) the purity of regenerated lignin (DES-L) is up to 98.1%, and the FT-IR spectra demonstrated no cellulose in DES-L; (iii) the structural characterization of DES-L, such as nitrobenzene oxidation and NMR analysis, suggested that the DES can selectively cleave ether linkages and damage the non-condensation sections of lignin, thereby facilitating lignin dissolution from wood; and (iv) the content of cellulose in solid residue (DES-C) is more than 90%, and the crystal structure of cellulose changed only slightly as observed by XRD, thereby indicating that DES is a promising solvent for the separation of wood components. Competing interests. We declare we have no competing interests. Funding. This work is supported by the National Natural Science Foundation of China (grant no. 31870565) and the Natural Science Foundation of Jiangsu Province (BK20181397).