Solubilization and Fractionation of Japanese Beech Wood with LiCl and DMSO

The solubility of beech wood cell wall in 8% lithium chloride/dimethyl sulfoxide (LiCl/DMSO) was investigated with an ethylenediamine (EDA) pretreatment without ball milling. EDA pretreatment of the wood cell wall was found to be an efficient method for the solubilization of the majority of the wood cell wall while avoiding the destructive effects of milling on the structure of the wood cell wall components. The yield of the final insoluble fraction was only approximately 31.6% based on the original Wiley wood sample after two EDA pretreatments and the dissolving processes. The solubility of lignocellulosic material in 8% LiCl/DMSO increased with decreasing lignin content after an EDA pretreatment. The yield of the insoluble fraction decreased from 62.9% to 9.2% with a decrease of lignin content from 28.0% to 14.1%. When the lignin content of lignocellulosic material was relatively low (approximately 10.5%), it could be dissolved in 8% LiCl/DMSO after an EDA pretreatment. The EDA pretreatment of wood meal had a much greater effect on polysaccharides than on the lignin in wood cell wall.


INTRODUCTION
Wood is an important natural composite consisting of different polymers, mainly cellulose, hemicellulose and lignin.Evaluation of the interactions between these polymers has been one of the key subjects in the field of wood and pulping chemistry.As monolignols polymerize within the cell wall and are embedded in previously accumulated polysaccharide gels, it is possible that molecular associations and covalent bonds are produced between lignin and carbohydrates. 1 Phenylpropane units are linked by various types of ether and C-C bonds in a lignin macromolecule, and the frequency of each bonding type varies between cell walls.Such diversity of polymers in cell walls and the chemical and physical associations with carbohydrates make it impossible to isolate the entire lignin fraction without serious degradation. 2 , 3 , 4It is important to clarify whether chemical bonds exist between these polymers in the native wood and corresponding pulp because the difficulties encountered in the latter stages of delignification are partially attributed to such chemical bonds between lignin and carbohydrates.
One promising method to examine these interactions is wood dissolution.It is nearly impossible to dissolve wood in its native state in conventional solvents.Although dissolving wood is challenging, recent discoveries have demonstrated that suitable media exist that may allow the dissolution of lignocellulosic materials.
In recent decades, many solvent systems have been reported to dissolve cellulose, which is the major component of chemical pulp. 5 , 6 , 7 , 8However, the direct application of a cellulose solvent system to the dissolution of underivatized lignocellulosic material with relatively high lignin content has not been successful. 9 , 10The development of a solvent system for pulps with high lignin content is required not only to examine the chemical nature of pulp components and interactions between them but also to utilize such pulps as a new source of natural fibrous materials.Kilpelä inen et al. demonstrated that both hardwoods and softwoods are readily dissolved in various imidazolium-based ionic liquids under gentle conditions.However, the complete wood dissolution can be achieved only at high temperature. 11Lu and Ralph described that finely ground plant cell wall can be dissolved in dimethyl sulfoxide and tetrabutylammonium fluoride (DMSO/TBAF) or dimethyl sulfoxide and N-methylimidazole (DMSO/NMI) binary solvent systems.The ability to dissolve wood meal without serious degradation of lignin would significantly improve methods for the analysis of the entire lignin fraction. 12However, the DMSO/NMI solvent system requires a rather long milling time prior to dissolution of the wood meal.It should be emphasized that milling of wood meal always causes a structural change of lignin, depending on the degree of milling. 13,14t is of interest to develop a novel solvent system that can completely dissolve wood meal with a minimal milling time.In a previous study, a novel solvent system consisting of lithium chloride/dimethyl sulfoxide (LiCl/DMSO) was developed.This solvent system can dissolve wood meal that has been finely ground by a planetary ball-mill for 2 h. 15he fractionation of wood cell wall components was achieved by the application of the LiCl/DMSO solvent system with a ball-milling pretreatment. 16However, the cellulose crystallinity of the ground wood meal is dramatically decreased even by 1 h of milling.Thus, a pretreatment method other than milling is a prerequisite to dissolution of highly crystalline cellulose or pulp for the analysis of the entirety of plant cell wall components.An ethylenediamine (EDA) treatment has been developed as a pretreatment for the dissolution of pulp with high lignin content in LiCl/DMSO. 17This method does not require a milling pretreatment, so structural modification of lignin and serious degradation of cellulose can be avoided.Various hardwood and softwood chemical pulps, including those with relatively high lignin content (up to approximately 10.5%), were completely dissolved without milling in LiCl/DMSO after a pretreatment with EDA.
In this study, the solubility of beech wood cell wall in 8% LiCl/DMSO was investigated with an ethylenediamine (EDA) pretreatment without ball milling.A new fractionation method was applied to wood for structural analyses of lignin, LCC, and polysaccharide by the use of this dissolution method.

Materials
Japanese beech (Fagus crenata Blume) wood was ground in a Wiley mill (40-80 mesh) and extracted with ethanol/benzene (1:2, v/v) for 8 h in a Soxhlet apparatus.The extracted Wiley wood was dried in air and subsequently dried under vacuum.
Delignification was conducted by treating 2.5 g of Wiley wood with 1 g of NaClO2 and 0.2 ml of acetic acid in 150 ml water at 75 °C for 1 h.This treatment was conducted one to three times to give DW1 (delignified wood 1), DW2 (delignified wood 2) and DW3 (delignified wood 3), respectively, with different lignin contents.
All chemicals were purchased from Wako Chemicals (Takasaki, Japan) and used as received without further purification.

Pretreatment of lignocellulosic material with EDA
The Wiley wood or partially delignified wood was treated with ethylenediamine (EDA) as a pretreatment for dissolution without milling.The EDA pretreatment method was the same as that in a previous study. 17In this study, 10 g samples were soaked in 100 ml EDA and stirred for 1 day at room temperature.The EDA treated samples were then freeze-dried.The Wiley wood and delignified woods treated with EDA are referred to as "wood-EDA complex" and "DW-EDA complex," as shown in Figures 1 and 2, respectively.

Fractionation of wood cell components
The fractions were separated by 8% LiCl/DMSO extraction from EDA pretreated Wiley wood and delignified wood.The wood-EDA complex and DW-EDA complexes were suspended in 8% LiCl/DMSO.The mixture was stirred at room temperature for 24 h and then heated to 75 °C and stirred at this temperature for several hours.The soluble and insoluble fractions were separated by centrifugation.As shown in Fig. 1, the wood-EDA complex was extracted by 8% LiCl/DMSO to obtain the soluble fraction WE-DL8S (F1.1) and the insoluble fraction WE-DL8R (F1.2).The residue WE-DL8R (F1.2) was freeze-dried after regeneration by dialysis with water.The dried WE-DL8R (F1.2) fraction was again treated with EDA.The EDA treated WE-DL8R residue (WRE) was again extracted by 8% LiCl/DMSO.The soluble fraction WRE-DL8S (F1.3) and the insoluble fraction WRE-DL8R (F1.4) were obtained.

WRE-DL8R (F1.4) Soluble Insoluble
Fig. 2 Scheme II: Procedure for isolation of various fractions from delignified wood purified by dialysis (molecular weight cutoff of 14000) with water for seven days.The samples were then freeze-dried and weighed.

Analyses of neutral sugars and lignin
The amount of neutral sugars was determined by the alditol-acetate method to characterize polysaccharides in the fractions. 18The lignin content (the total amount of Klason lignin and acid soluble lignin) in the fractions was determined by the Klason method. 19Alkaline nitrobenzene oxidation was performed according to Chen. 20

Solubilization of wood cell wall components with an EDA pretreatment
As previously reported, various lignocellulosic materials such as cotton, cellulose, holocellulose, hardwood and softwood chemical pulps, including those with relatively high lignin content, can be completely dissolved in LiCl/DMSO after a pretreatment with EDA without milling. 17When a chemical pulp contained approximately 14.6% lignin, it could not be completely dissolved in LiCl/DMSO after EDA pretreatment.Therefore, it is quite important to understand why an insoluble fraction is left when the lignin content is above a certain level.
Wiley wood, without milling, was subjected to the dissolution process consisting of EDA pretreatment and dissolution in 8% LiCl/DMSO.The EDA treated freeze-dried wood is called "wood-EDA complex".Because the yield of the soluble fraction WE-DL8S (F1.1) from this process was only approximately 24.8% (Fig. 3), the residue was again subjected to the same dissolution process.
High lignin content was considered to cause the incomplete solubility of EDA pretreated wood meal.Therefore, wood meals were delignified by NaClO2 under acid conditions.By repeating the delignification treatment, delignified woods with different lignin content were prepared.The yield of delignified wood and its lignin content are shown in Figure 4.Both the yield and lignin content of the delignified wood decreased with additional delignification steps.The lignin content of the DW3 (delignified wood 3) is still 14.1%, even though the delignification treatment was repeated 3 times.Delignified woods were soaked in EDA with stirring for 1 day at room temperature followed by freeze-drying.
EDA pretreatment of wood cell wall is an efficient method for the solubilization of the majority of wood cell wall while avoiding the destructive effects of milling on the structure of wood cell wall components.The yield of the final insoluble fraction WRE-DL8R (F1.4) prepared by Scheme I was 50.3% based on WE-DL8R (F1.2, the residue from the first dissolution process) and was only approximately 31.6% based on the original Wiley wood.On the other hand, the solubility of lignocellulosic material in 8% LiCl/DMSO increased with decreasing lignin content after an EDA pretreatment.The 62.9% yield of the insoluble fraction WE-DL8R (F1.2) decreased to 9.2% for DWE3-DL8R (F2.6).When the lignin content of lignocellulosic material is relatively low (approximately 10.5%), it can be dissolved in 8% LiCl/DMSO after an EDA pretreatment. 17

Fractionation of wood cell wall components
Different cell wall fractions were obtained by varying the dissolving capacity of the solvent system, which was controlled by the LiCl concentration in the LiCl/DMSO solvent system in our previous study. 16On the contrary, in this study, the LiCl concentration was maintained at 8% in DMSO.Namely, the dissolving capacity of solvent system was fixed at the highest level.Instead, the different soluble and insoluble fractions of wood cell wall components were obtained by varying the solubility of each sample in the 8% LiCl/DMSO system.One method employed to change the solubility of the sample was to conduct the second EDA pretreatment on the insoluble residue from the first dissolution (Scheme I, Figure 1), and another was to partially delignify the wood before subjecting it to EDA pretreatment (Scheme II, Figure 2).In Scheme I (Figure 1), first the EDA pretreated Wiley wood was extracted with 8% LiCl/DMSO, then the soluble fraction WE-DL8S (F1.1) and insoluble fraction WE-DL8R (F1.2) were separated.Finally, the insoluble fraction WE-DL8R (F1.2) was treated with EDA and extracted by 8% LiCl/DMSO again.The second extraction process produced soluble fraction WRE-DL8S (F1.3) and insoluble fraction WRE-DL8R (F1.4).In Scheme II (Fig. 2), the delignified wood samples with different lignin content were treated by EDA and extracted by 8% LiCl/DMSO.Different soluble and insoluble fractions were obtained from delignified wood samples.
The yield of each soluble and insoluble fraction and their lignin contents obtained using Scheme I and Scheme II are shown in Figures 3 and 4, respectively.As shown in Figure 3, the Wiley wood still could not be completely dissolved in 8% LiCl/DMSO by the second EDA pretreatment.In Scheme I, the yield of each soluble fraction was lower than that of the corresponding insoluble fraction.As shown in Figure 3, the first EDA treatment of Wiley wood yielded 24.8% of soluble fraction WE-DL8S (F1.1) in 8% LiCl/DMSO.After the second EDA treatment of the remaining residue WE-DL8R (F1.2), the yield of soluble fraction WRE-DL8S (F1.3) reached 41.0% based on the weight of WE-DL8R (F1.2), leaving 31.7% of the original Wiley wood as a residue.
The yield of the soluble and insoluble fractions separated from delignified wood by Scheme II is shown in Figure 4.The yield of each soluble fraction obtained from Scheme II was higher than that of the corresponding insoluble fraction, due to the low lignin content in delignified wood.The yield of the insoluble fraction decreased with an increase in the degree of delignification, i.e., the solubility of delignified wood in LiCl/DMSO increased with a decrease in the lignin content of each delignified wood sample.From Figure 4, the yield of insoluble fraction DWE3-DL8R (F2.6) was only 9.2% when the lignin content of delignified wood (DW3) was a relatively low 14.1% (Figure 4).The results fully agree with the results reported in our previous study, wherein the chemical pulp could be dissolved completely after EDA pretreatment when the lignin content of the pulp was up to approximately 10.5%, but when the lignin content of the pulp reached 14.6%, only a suspension was obtained. 17It is speculated that if a lignocellulosic material contained a relatively low lignin content, the cellulosic components would be more accessible during the EDA pretreatment and resulted in a better dissolution in 8% LiCl/DMSO.The sums of the yield of each soluble and corresponding insoluble fraction separated by Scheme II DWE1-DL8S (F2.1) + DWE1-DL8R (F2.2),DWE2-DL8S (F2.3) + DWE2-DL8R (F2.4), DWE3-DL8S (F2.5) + DWE3-DL8R (F2.6), were 78.6%, 69.3% and 70.2%, respectively (Figure 4).The yield was lower compared with the result from Scheme I (87.7%).In the case of a milling pretreatment, which was conducted for milled wood, the sum of each fraction was approximately 92%.For Scheme I, the lignin content in the soluble fraction was lower than that of the corresponding insoluble fraction, as shown in Figure 3.These results are quite different from that obtained from milled wood.The results suggest that the EDA pretreatment of wood meal had a greater effect on the polysaccharide fractions than on the lignin of the wood cell wall.For Scheme II, it is interesting that the lignin contents in the soluble fractions of DWE2-DL8S (F2.3) and DWE3-DL8S (F2.5) were higher than those in the corresponding insoluble fractions DWE2-DL8R (F2.4) and DWE3-DL8R (F2.6), as shown in Figure 4.This result indicates that after a delignification process, the fractions rich in lignin were easily dissolved in 8% LiCl/DMSO after an EDA pretreatment, leading to the higher lignin contents in the soluble fractions.

Structural characterization of lignin in each fraction
To investigate the structural features of the aromatic portion of lignin, alkaline nitrobenzene oxidation was conducted on the fractions separated using Scheme I and Scheme II.Figures 5 and 6 show the yields of the nitrobenzene oxidation products syringaldehyde + syringic acid (S), vanillin + vanillic acid (V), S+V, and the syringyl ratio S/(S+V) of each fraction.From Figure 5, the S/(S+V) ratio of the soluble fraction was slightly higher than those of the corresponding insoluble fractions.And the yields of S+V of each fraction were almost the same.Thus, the lignins in the soluble or insoluble fractions separated by Scheme I were found to have similar structural characteristics.These results differ from the results obtained from milled wood. 16.For the milled wood, it was proposed that the secondary cell wall lignin with a higher S/(S+V) ratio was more difficult to extract than the primary cell wall lignin with a lower S/(S+V) ratio.With an increase in the dissolving capacity of the solvent, more of the fractions with higher S/(S+V) ratios became extractable, suggesting that the contribution of secondary cell wall lignin to the soluble fraction increased with the heightened solubility of the solvent.
Conversely, Figure 6 shows that both the yield of S+V and the S/(S+V) ratio in the soluble fractions were always lower than those of the corresponding insoluble fractions, except for the S/(S+V) ratios in DWE1-DL8S (F2.1) and DWE1-DL8R (F2.2).The high ratio S/(S+V) in DWE1-DL8S (F2.1) was due to the high lignin content remaining in DW1 after only one delignification treatment, i.e., less degradation occurred in DW1, from which lignin with a higher ratio S/(S+V) was extracted by 8% LiCl/DMSO after EDA treatment.After the second and third delignification procedures followed by EDA treatment, the yields S+V and the ratios S/(S+V) of both the soluble and insoluble fractions decreased.Such decrease agreed well with the decrease in the yield S+V and the ratio S/(S+V) in the delignified wood after repeating the NaClO2 treatment, as shown in Table 1.

Composition of neutral sugars
The weight ratios of neutral sugars (based on total sugars) in each fraction prepared by Scheme I and Scheme II are shown in Figures 7 and 8, respectively.As shown in Figure 7, the weight ratio of glucose in WE-DL8S (F1.1) was quite low, and the weight ratio of xylose was very high.As the lignin content was only 19.7%, which is extremely low compared with the first solubilized fraction, the main component of this fraction was xylan. 16The question is why only the solubilization of hemicellulose (xylan) was facilitated by the first EDA pretreatment.One possible explanation is that the high lignin content restricted the formation of a cellulose-EDA complex in the wood cell wall during the first EDA pretreatment.This notion can explain why cellulose was not solubilized.However, the solubilization of xylan cannot be fully understood.Another explanation is that the formation of a cellulose-EDA complex was restricted by a lignin-xylan complex, solubilization of which proceeded under high LiCl concentration.By the removal of this fraction during the first dissolution process, the formation of a cellulose-EDA complex could be achieved during the second EDA treatment.By the second EDA treatment, the soluble fraction WRE-DL8S (F1.3) obtained from WE-DL8R (F1.2) showed a weight ratio of xylose only slightly higher than or similar to that of the corresponding insoluble fraction WRE-DL8R (F1.4).This result is reasonable as the amount of xylose in fraction WE-DL8R (F1.2) was already quite low after the first extraction.These results agree well with the results obtained by Scheme II.As shown in Figure 8, the weight ratios of glucose in the soluble fractions were always lower and the weight ratios of xylose in soluble fractions were always higher than those in corresponding insoluble fractions.

Correlation between carbohydrate composition and lignin content
The dependence of the weight ratio of sugars on the lignin content of each fraction prepared by Scheme I and II are shown in Figures 9 and 10.Except for the fraction WE-DL8S (F1.1, 19.7% lignin content), the weight ratio of glucose and the weight ratio of sugars from hemicelluloses were similar for all other fractions, even though their lignin contents were different.It appeared that the weight ratios of sugars of each fraction prepared by Scheme I and II were not affected by the lignin content.These results were different from those reported in our previous study, wherein a significantly high correlation was observed between the weight ratio of sugars and the lignin content. 16This difference was due to the EDA and/or delignification pretreatment in this study that changed the solubility of lignocellulosic materials (Scheme I and Scheme II).In the previous study, the dissolving capacity of the solvent system was varied by controlling the LiCl concentration in DMSO. 16In this study, the dissolving capacity of the solvent system was fixed by fixing the LiCl concentration at 8% in DMSO, and the chemical composition of the soluble or insoluble fractions were mainly determined by the properties of the sample after the pretreatment.
It is interesting to note that for each case in the present study, the weight ratio of glucose in the soluble fraction was always lower than that in the insoluble fraction, while the weight ratio of xylose in the soluble fraction was always higher than that in the insoluble fraction, regardless of the lignin content.Apparently, the EDA pretreatment facilitated the dissolution of xylan, suggesting that an EDA-complex was probably formed between xylan or lignin-xylan with EDA, which in turn increased the solubility of xylan/xylan-lignin in the LiCl/DMSO solvent system.

CONCLUSIONS
With EDA pretreatment, the lignocellulosic materials of Japanese beech wood can be partially or completely dissolved in LiCl/DMSO, depending on the lignin content.Two methods to fractionate wood cell wall components were proposed.The difference of the two methods was with or without a delignification treatment, which had a significant effect on the solubility of the wood and the properties of the soluble and insoluble wood fractions in the LiCl/DMSO solvent system.Without delignification treatment, the yield of the final insoluble fraction was 50.3%, based on the residue from the first dissolution process, and 31.6% based on the starting wood material.With delignification treatment, the solubility of the lignocellulosic material in 8% LiCl/DMSO increased markedly.The yield of the insoluble fraction decreased from 62.9% to 9.2% when the lignin content of the wood material was reduced from 28.0% to 14.1%.The results also showed that the EDA pretreatment was efficient in promoting the solubilization of the wood material in the LiCl/DMSO solvent system, without the need of intensive mechanical treatment (e.g.milling).

Fig. 1
Fig. 1 Scheme I: Procedure for isolation of various fractions from EDA treated Wiley wood

Fig. 3
Fig. 3 Yield and lignin content of each fraction prepared by Scheme I.

Fig. 4
Fig. 4 Yield and lignin content of the delignified wood samples and each fraction prepared by Scheme II (Lignin content of the original wood was 28.0%).

Fig. 5
Fig. 5 Yield of S+V and the S/(S+V) ratio of each fraction prepared by Scheme I.

Fig. 6
Fig. 6 Yield of S+V and the S/(S+V) ratio of each fraction prepared by Scheme II.

Fig. 7 Fig. 8
Fig. 7 Weight ratio of xylan and glucose in each fraction prepared by Scheme I. (S: soluble fraction, R: insoluble fraction)

Fig. 9
Fig. 9 Dependence of the weight ratio of xylan and glucose on the lignin content of each fraction prepared by Scheme I (S: soluble fraction, R: insoluble fraction).

Fig. 10
Fig. 10 Dependence of the weight ratio of xylan and glucose on the lignin content of each fraction prepared by Scheme II (S: soluble fraction, R: insoluble fraction).

Table 1
Yield of S+V and ratio of S/(S+V) of the delignified woods.