Enhanced delignification of cornstalk by employing superbase TBD in ionic liquids

Abstract

The delignification of cornstalk was efficiently accomplished by using 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD) as an additive in 1-allyl-3-methylimidazolium acetate ([Amim][OAc]). When 1.0 wt% TBD was added to [Amim][OAc], the cellulose and lignin contents of cellulose rich material (CRM) were achieved to be 39.12% and 6.74%, respectively. With the addition of 0.1 wt% TBD to [Amim][OAc], the lignin content of CRM could even be reduced to 2.06% without considering the cellulose content. There could be two possible reasons for the enhanced delignification of cornstalk by adding TBD in [Amim][OAc]. One is the alkalinity and exposed nitrogen atoms of TBD, which make it an efficient dibasic nucleophile and helpful for the lignin β-O-4 ether bond cleavage reaction. The other one is the decreased interaction energy of [Amim]⁺ and [OAc]⁻ from 99.1 kcal mol⁻¹ to 89.2 kcal mol⁻¹ with the addition of TBD, which makes the [Amim]⁺ and [OAc]⁻ easier to interact with the cornstalk components. Simultaneously, the CRM regenerated from the system of [Amim][OAc] + TBD was effectively hydrolyzed by cellulase with 98% enzymatic hydrolysis yield, which proved that the cellulose structures were highly disrupted and lignin was significantly removed in the CRM.

---

Introduction

Biomass utilization has been greatly developed and more environmentally friendly methods are being researched due to the current energy crisis. The lignin seal and crystalline structure of cellulose are major drawbacks suppressing the energy production and chemical processes of biomass utilization. The lignin hinders carbohydrate molecules being exposed to the pretreatment solvents, so delignification is a necessary procedure to improve the pretreatment efficiency.^1,2 A successful biomass pretreatment should result in high cellulose and hemicellulose recovery with lower lignin content, meanwhile, the accessibility of cellulose to the cellulase should also be increased in order to improve the enzymatic hydrolysis rate. Many traditional pretreatment methods, like steam explosion,³ acidolysis and alkaline pretreatment,^4,5 have been explored to overcome the drawbacks and increase the enzymatic accessibility of biomass.^6,7 These methods contribute a lot to the development of biomass utilization but with serious environmental consequences, strong corrosion and are regarded as highly energy intensive processes.

Ionic liquids (ILs) are composed of anions and cations that are liquids around or below 100 °C, many kinds of ILs have been used as solvents for cellulose and pretreatment medium for lignocellulosic materials.⁸ 1-Butyl-3-methylimidazolium chloride ([Bmim]Cl) was first found effective to dissolve cellulose.⁹ Hereafter many kinds of ILs were investigated to dissolve cellulose. And the most effective ILs are composed of various anions ([Cl]⁻, [Br]⁻, [SCN]⁻, [OAc]⁻, [HCOO]⁻, [(C₆H₅) COO]⁻, [(NH₂)CH₂COO]⁻, [(CH₃CH₂)₂PO₂]⁻, [(CH₃O)₂PO₂]⁻, [HSCH₂COO]⁻) and cations ([C_nmim] (n = 2, 3, 4, 5, 6, 7, 8), [Amim], [Bu₄P]).^10–16 And it is believed that ILs can be used to disrupt the hydrogen bonds among different polysaccharide chains in the biomass, so ILs are widely applied in biomass applications.¹⁷ For examples, about 35% reduction of lignin content in the reconstituted CRM from pine was achieved with 1-ethyl-3-methylimidazolium acetate ([Emim][OAc]),¹⁸ 62% separation of cellulose was realized from wood in 1-allyl-3-methylimidazolium chloride ([Amim]Cl),¹⁹ and approximately 38% hemicellulose separation from spruce was accomplished in switchable ILs.²⁰ It has been demonstrated that the effective anions and cations are [Cl]⁻, [(CH₃CH₂)₂PO₂]⁻, [(CH₃O)₂PO₂]⁻, [OAc]⁻ and the imidazolium cations with short carbon chains.²¹ Especially, the imidazolium-based acetate ILs attract more attention due to the low viscosity, weak corrosivity, and high biomass dissolving power as compared to other kinds of ILs.²² However, it is still very hard to realize the industrialization of biomass utilization by using pure IL systems for biomass pretreatment. Therefore, many combined ILs systems with catalysts or additives were also used to improve the biomass separating efficiency. Lithium salts were added to increase cellulose solubility,²³ acidic catalyst was used to enhance the extraction of the constituents, and aqueous ammonia was added to increase the delignification.^24,25 Such catalysts and/or additives have improved the biomass components separation to a certain degree, resulting in an effective method for enhanced separation efficiency by developing better additives for biomass ILs pretreatment process.

The objective of this work was to obtain the enhanced delignification of cornstalk by adding organic N-bases in ILs. Jia and co-workers reported that a phenolic lignin model compound could be effectively decomposed while adding 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD) in 1-butyl-2,3-dimethylimidazolium chloride ([BDMIm]Cl), which had high activity to assist the β-O-4 bond cleavage reaction.²⁶ So two kinds of N-bases TBD and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) were introduced in this work owing to their excellent lignin decomposition properties. Several kinds of biodegradable ILs containing Ch⁺ were widely used to remove lignin.²⁷ The ChCl is an inexpensive organic quaternary ammonium salt and also contains Ch⁺, therefore, this kind of Ch⁺ source was added in [Amim][OAc] to detect whether it was helpful for delignification of cornstalk. The solid acid (SO₄²⁻/ZnO–TiO₂) had a large number of acid sites and showed excellent catalytic performance in glycolysis of polyethylene terephthalate.²⁸ So ChCl, together with solid acid (SO₄²⁻/ZnO–TiO₂) as an organic quaternary ammonium salt and a kind of solid acid, were also used as additives to compare with above two kinds of superbases. The cornstalk pretreatment was investigated by changing additive type, ILs kind, dissolution time and TBD amount. Moreover, the cornstalk sample, microcrystalline cellulose (MCC), and CRM were characterized to understand the dissolution process of cornstalk by differential thermal analysis (TG/DTA), Fourier transfer infrared (FT-IR), X-ray diffraction (XRD) and solid ¹³C-nuclear magnetic resonance (¹³C-NMR). At last, the CRM regenerated from the system of [Amim][OAc] + TBD + cornstalk was hydrolyzed by cellulase to evaluate the pretreatment effect.

Simultaneously, the anions and cations of ILs are not all dissociative but some in ion-pairs, even in the clusters form, which greatly influence the properties of ILs.²⁹ The selected additives were perhaps able to break or weaken the interactions between cations and anions of ILs, which makes it easier for the anions and cations of ILs to interact with the biomass components, so the accelerated dissolution and enhanced delignification of cornstalk could be achieved. Therefore, in the best delignification system combining IL and additive, the density functional theory calculations were introduced to investigate the interaction energy changes between the anion and cation of ILs when adding additive to prove our hypothesis.

Materials and methods

Materials

Cornstalk provided by Henan Province in China was chopped, air-dried, milled and sized to pass through 0.125 mm screens. Most of the cellular contents including fat, sugar, starch and protein were removed from raw cornstalk for preparing the cornstalk sample by neutral detergent solvent (containing lauryl sodium sulfate). After being washed, the cellulose, hemicellulose, lignin and ash contents of the cornstalk sample were 36.82%, 44.19%, 18.49% and 0.5%, respectively. The cornstalk sample was kept in the desiccator before being used.

2 M HCl and 72% H₂SO₄ were prepared according to the method of Van Soest in local experiment laboratory. Alkali lignin was purchased from TCI (Shanghai) Chemical Industries Development. Microcrystalline cellulose (MCC) with a degree of polymerization (DP) of 210–240 was purchased from National Pharmaceutical Group Chemical Reagent. Commercially available regents purchased from Sigma-Aldrich and local chemical companies were all analytical grade and used as received without any further purification. The ILs (>98%) used in this work were purchased from Henan Lihua pharmaceutical and Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences. The ionic liquids used in this work were also kept in the desiccator. Before being used, the ionic liquids were dried for 24 h in order to control the water content at a very low level. Karl Fisher titration of the dried ionic liquids showed that the water content were all less than 50 ppm. The cellulase (140 FPU g⁻¹) was purchased from Ningxia Xiasheng Industrial (Group) Co., Ltd.

Procedure of cornstalk delignification in IL + additive

10 g IL, 1 wt% additive (w additive/w IL) and 5 wt% (w cornstalk/w IL) cornstalk sample (particle size < 0.125 mm) were added to a 100 mL round-bottomed flask with a vigorous stirring at 130 °C for 3 h. The viscosity of the mixed solution sharply increased after cornstalk dissolution, so that it is difficult to separate the non-dissolved residue from the mixed solution. After the mixed solution was cooled to room temperature, DMSO was then added as diluents aiming to separate the non-dissolved residue from the mixed solution. DMSO is not environmentally benign, but it has excellent diluents performance and almost has no effect on the dissolution ability of three main components of cornstalk in the course of dilution. So it is commonly used in biomass IL pretreatment in order to obtain accurate dissolution data.¹⁸ The residue was filtered through an organic filter membrane (aperture < 0.8 μm) and washed with deionic water for three times, finally dried to a constant weight at 70 °C overnight. Dissolution rate of cornstalk was calculated according to eqn (1):where dissolution% is the mass percent of dissolved cornstalk, M_cs represents the mass of cornstalk sample (g), which is added to the initial solution. And M_ur is the mass of non-dissolved residue left in the solution (g).

In the dissolved solution, 100 mL acetone–water (v/v = 1 : 1) mixed solvent was used as anti-solvent to regenerate CRM with vigorously stirring at room temperature according to the method reported by Sun and co-workers, because acetone can restrain the regeneration of lignin from IL solution.¹⁸ Centrifuged at 9800 rpm for 6 min, the CRM was continually washed with deionic water for three times to remove the traces of [Amim][OAc]. After being isolated by filtering, the CRM was dried in a convection oven at 65 °C overnight prior to the analysis. And the regeneration% was measured according to eqn (2):

where regeneration% is the mass percent of CRM compared with cornstalk sample, M_CRM is the mass of CRM (g), M_cs is the mass of cornstalk sample added to each solution (g).

The scope of this work mainly focused on the cellulose rich material regenerated from the cornstalk solution, because the amount of non-dissolved residue is so less that it is not enough to be analysed by the Van Soest method under the most condition of experiments. So we just discussed the contents of the cellulose, hemicellulose and lignin in the regenerated CRM in this paper.

The analysis of sample contents was conducted according to the Van Soest process.³⁰ 1 g raw cornstalk was first dried for at least 4 h in a 105 °C oven, and then boiled in 50 mL neutral detergent for 1 h with refluxing. The cooked sample was diluted to neutral pH with deionic water, and then filtered by organic membrane. Washed by acetone for two times, the residue was dried at 105 °C for 4 h to determine the neutral detergent fiber (NDF). The cornstalk sample used in this study was prepared according to the procedure above mentioned. The remaining sample was boiled for 50 min in 50 mL of 2 M HCl to measure the acid detergent fiber (ADF), the residue left was digested in 72% H₂SO₄ subsequently for obtaining the acid detergent lignin (ADL). The hemicellulose content was in the NDF–ADF difference, and the ADF–ADL was regarded as cellulose content, the weight left from the ADL process was considered to be lignin and ash. ADL was then placed in the muffle furnace at 550 °C for 4 h and dried to constant, resulting in ash content. The ash is mainly composed of the silica, which couldn't be regenerated in the CRM by adding anti-solvent (acetone–water) into the cornstalk solution. Therefore, there was almost no ash retained in the CRM.

Our research was mainly focused on the CRM, so the ash of the CRM was not considered in this manuscript. So the hemicellulose, cellulose and lignin contents of the CRM are calculated according to eqn (3)–(5).

Characterizations methods

Thermogravimetry (DTG-60H, SHIMADZU, Japan) was used to detect the TGA of the samples during a temperature range from room temperature to 600 °C at a heating rate of 10 °C min⁻¹ in nitrogen flow (30 mL min⁻¹). The samples were also characterized by solid ¹³C-NMR spectra, which were measured by BRUKER AVANCE-III 400 M solid state spectrometer (Bruker, Germany). The XRD curve of cornstalk sample, MCC, and CRM were examined by X'Pert PRO MPD diffract meter at an accelerating voltage of 40 kV and an emission current of 40 mA with Cu Kα radiation. The diffracted intensity of Cu Ka radiation (λ = 0.1542 nm) was measured in a 2θ range over 10–50° with a step size of 0.01° and step time of 1 s. The FT-IR spectra of cornstalk sample, MCC, and CRM, were recorded in the range 4000–400 cm⁻¹, and accumulated for 32 scans at a resolution of 4 cm⁻¹ on a Nicolet 380 spectrometer (Thermo Fisher Scientific, America).

Enzymatic hydrolysis of CRM from the system of [Amim][OAc] + TBD + cornstalk

1.0 g sample and cellulase (20 FPU g⁻¹) were added to 20 mL citrate buffer (pH 4.9, 50 mM) in a 50 mL flask. Enzymatic hydrolysis was conducted in a thermo shaking incubator at 50 °C for 115 rpm. The sugar content was determined by Aminex HPX-87H column at 65 °C with 5 mM H₂SO₄ as eluent, the flow rate was 0.7 mL min⁻¹. The conversion rate was calculated according to eqn (6):where volume is hydrolysis volume, and M₁ is the mass of sample added to enzymatic hydrolysis, cellulose% is the cellulose content of the sample.

Results and discussion

Effect of different additives in [Amim][OAc] on the cornstalk delignification

In the Fig. 1a and b, the effect of different additives on the cornstalk pretreatment was investigated. 1.0 wt% TBD, DBU, ChCl and solid acid (SO₄²⁻/ZnO–TiO₂) were added to [Amim][OAc], respectively, for screening the most appropriate additive for cornstalk delignification. The components contents of cornstalk sample and the CRM obtained from using [Amim][OAc] only were also showed in Fig. 1b. In the cornstalk sample, the cellulose, hemicellulose and lignin contents were 36.82%, 44.19% and 18.49%, respectively. In the CRM obtained when using [Amim][OAc] only, the cellulose, hemicellulose and lignin contents were 30.19%, 61.47% and 8.34%, respectively. When different additives were added to [Amim][OAc], the lignin% of the CRM decreased in the order of 7.17% (DBU) > 6.74% (TBD) > 4.72% (ChCl) > 2.47% (solid acid). The N-bases DBU and TBD would assist cleaving the β-O-4 bonds of lignin in the dissolution process.²⁶ From the result in the Fig. 1, the delignification of cornstalk was obvious when adding ChCl, which could be attributed to the Ch⁺ cation.²⁷ The effective delignification by solid acid (SO₄²⁻/ZnO–TiO₂) may due to the acidity and catalytic activity.²⁸ So these additives are helpful for the delignification of cornstalk. However, the cellulose% of CRM increased only when adding TBD to [Amim][OAc], the order of cellulose% of the CRM is as followed: 17.93% (solid acid) < 18.63% (DBU) < 23.07% (ChCl) < 39.12% (TBD). In these four combined pretreatment systems, only TBD reduces the lignin% and increases the cellulose% of CRM under the applied conditions. Simultaneously, the dissolution% and regeneration% of cornstalk in the system of [Amim][OAc] + TBD were both the highest among the four combined pretreatment systems. So TBD is chosen as desired additive in the following optimized experiments.

Fig. 1 Effect of different additives in [Amim][OAc] on the cornstalk delignification: (1) TBD, (2) DBU, (3) ChCl, (4) solid acid, (5) no additive, (6) cornstalk sample, (a) the dissolution% and regeneration%, (b) the components of the CRM.

Effect of different ILs with TBD on the cornstalk delignification

To compare with the TBD's role in different IL systems, three kinds of imidazolium-based ILs, 1-ethyl-3-methylimidazolium diethylphosphate ([Emim][DEP]), [Bmim]Cl and [Emim][OAc], were also used to dissolve cornstalk. From the results shown in Fig. 2a and b, the cellulose% of the CRM increased in the order as followed: 21.08% ([Bmim]Cl) < 25.95% ([Emim][DEP]) < 30.76% ([Emim][OAc]) < 39.12% ([Amim][OAc]). The high hydrogen-bonding ability (β) value of ILs expresses the ability to accept hydrogen bonds, and higher β value of results in better cellulose solubility.³¹ The β values of those ILs was followed in an order of [BMIM]Cl (0.84) < [Emim][DEP] < [Emim][OAc] (1.074) < [Amim][OAc] (1.175).³² So the high cellulose% of CRM from [Amim][OAc] + TBD system was mainly due to the high β value of [Amim][OAc]. The lignin% of the CRM decreased in the order of 14.72% ([Bmim]Cl) > 10.32% ([Emim][DEP]) > 6.74% ([Amim][OAc]) > 6.04% ([Emim][OAc]). The cornstalk dissolution% and regeneration% both reached the highest value in the system of [Bmim]Cl + TBD, however, the cellulose% of CRM from the system of [Bmim]Cl + TBD was not very high. The highest cellulose% and lower lignin% of CRM were obtained from the system of [Amim][OAc] + TBD. This could be due to the high β value of [Amim][OAc] and the π-interactions of the cation with the aromatic compounds of lignin,³³ which exhibited not only within its imidazolium ring, but also on its side-chain of [Amim][OAc]. Based on the above results, the system of [Amim][OAc] + TBD is better for cornstalk delignification.

Fig. 2 Effect of different ILs with TBD on the cornstalk delignification: (1) [Emim][DEP], (2) [Bmim]Cl, (3) [Emim][OAc], (4) [Amim][OAc], (a) the dissolution% and regeneration%, (b) the components of the CRM.

Effect of dissolution time on the cornstalk delignification

To further understand the role of TBD, cornstalk dissolution in the systems of [Amim][OAc] or [Amim][OAc] + TBD were conducted for different dissolution time, as shown in the Fig. ESI 1a† (in ESI) and 3a, respectively. With the dissolution time increasing from 1 h to 9 h, the cornstalk dissolution% in the systems of [Amim][OAc] alone and [Amim][OAc] + TBD were ranged from 34.08% to 90.75%, from 42.63% to 92.28%, respectively. The highest regeneration% in the systems of [Amim][OAc] and [Amim][OAc] + TBD were 77.47% at 7 h and 75.78% at 5 h, respectively. So the dissolution time of highest regeneration% of the system of [Amim][OAc] + TBD was shortened 2 hours. In order to understand why the dissolution was accelerated, the density functional theory (DFT) calculations were carried out to analyse the interaction of reactants as shown in Fig. 9, ESI 2 and 3.† All calculations were carried out using the Gaussian 03 program. With the influence of TBD, the interaction energy of cation and anion decreased from 99.1 kcal mol⁻¹ to 89.2 kcal mol⁻¹. That result may indicate that it becomes easier for [Amim]⁺ or [OAc]⁻ to contact with cornstalk components and disrupt the hydrogen-bonds among the cellulose, hemicelluloses and lignin,³⁴ therefore, the dissolution speed increased.

Fig. 3 Effect of the dissolution time on the cornstalk delignification with [Amim][OAc] + TBD, (a) the dissolution% and regeneration%, (b) the components of the CRM.

The TBD addition caused the lignin% of CRM decreasing to 10.39% at 1 h and 6.74% at 3 h, the cellulose% of CRM increasing to 41.41% at 1 h, 39.12% at 3 h, respectively, as shown in Fig. 3b. The cellulose% of CRM from the system of [Amim][OAc] + TBD was higher 17% and 9% than those from the system of [Amim][OAc] alone (in the Fig. ESI 1b†). Those results indicate that TBD has an effect on accelerating the dissolution and improving the CRM. Taking into account both the cellulose% and lignin% of CRM, 3 h would be better dissolution time for cornstalk delignification in the system of [Amim][OAc] + TBD.

Effect of TBD amount in [Amim][OAc] on the cornstalk delignification

As shown in Fig. 4a and b, the effect of different TBD amount (0.1, 0.3, 0.5, 0.7, 1.0, 1.2, and 1.5 wt%) was conducted in [Amim][OAc]. When 0.1 wt% TBD was added, the lignin% decreased to 2.06% of CRM from 18.49% of cornstalk sample, the cellulose% of CRM decreased to 19.59%. As the TBD amount increased, the dissolution% increased. The highest value of dissolution reached 66.06% at 0.5 wt% TBD, then decreased and kept at a stable level. When the TBD amount increasing from 0.1 wt% to 1.5 wt%, the lignin% of the CRM increased to the highest point then decreased. The change trend of the cellulose% was similar to that of the lignin% in the CRM. In order to get CRM with higher cellulose% and lower lignin%, 1.0 wt% TBD is a better amount for cornstalk pretreatment in [Amim][OAc]. Meanwhile, 0.1 wt% TBD is also a better choice if only considering delignification of cornstalk. Simultaneously, the price of TBD is 549 CNY/5g, the low dosage of TBD in [Amim][OAc] and the reuse of the [Amim][OAc] + TBD solution make this delignification method economical and convenient.

Fig. 4 Effect of TBD amount in [Amim][OAc] on the cornstalk delignification, (a) The dissolution% and regeneration%, (b) the components of the CRM.

In this delignification process, the appropriate concentration of TBD would increase the dissolution of cornstalk. Overmuch or insufficient TBD is adverse for the dissolution and delignification of cornstalk. On one hand, the additive TBD decreases the interaction energy between the cation and anion of [Amim][OAc], which is helpful for the cellulose and lignin dissolution in ionic liquid. On the other hand, the additive TBD would react with lignin by cleaving the β-O-4 bond in the lignin due to the catalytic property of TBD.²⁶ And the relationship between those two effects of TBD is competitive, so the dissolution and delignification of cornstalk were diverse when adding different amount of TBD in [Amim][OAc].

When the TBD dose is very little, the main effect of TBD is reacting with lignin, so the delignification is obvious. And little amount of TBD is used in decreasing the interaction energy between the cation and anion of [Amim][OAc], so the dissolution of cornstalk does not increase greatly at 0.1 wt% TBD in [Amim][OAc]. As increasing TBD loading from 0.1 wt% to 0.7 wt% in [Amim][OAc], the amount of TBD used in decreasing the interaction energy between the cation and anion of [Amim][OAc] increases, so the dissolution of cornstalk, the cellulose and lignin contents of the CRM all increase. Therefore, the effect of TBD is mainly in decreasing the interaction energy between the cation and anion of [Amim][OAc] when the TBD amount during 0.1 wt% to 0.7 wt%. However, when the TBD amount exceeds 1.0 wt%, the competitive relationship between those two effects of TBD results in decreased dissolution of cornstalk and low lignin content in the CRM, which indicates that the TBD mainly reacts with lignin. This would be verified by both experiment and DFT calculations in our further research.

Analysis and characterization

Fig. 5–8 present the TGA and ¹³C-NMR, XRD and FT-IR spectra of cornstalk sample, MCC, and CRM from the system of [Amim][OAc] + TBD, respectively, which indicate that the crystallinity of the cellulose in the CRM reduces and the amorphous regions increases, so the accessibility to cellulase can also be improved. In the TGA spectra, the thermal decomposition of cornstalk sample and MCC are about 338 °C and 346 °C, respectively, and the thermal decomposition of CRM is 297 °C. The lower thermal decomposition temperature of CRM from the system of [Amim][OAc] + TBD may be caused by the structural reconstitution during the dissolution and regeneration process, which is in accordance with the ¹³C-NMR and XRD results.

Fig. 5 TGA spectra for (a) MCC, (b) cornstalk sample, (c) CRM from the system of [Amim][OAc] + TBD.

Fig. 6 ¹³C-NMR spectra for (a) MCC, (b) cornstalk sample, (c) CRM from the system of [Amim][OAc] + TBD.

Fig. 7 XRD spectra for (a) MCC, (b) cornstalk sample, (c) CRM from the system of [Amim][OAc] + TBD.

Fig. 8 FT-IR spectra for (a) MCC, (b) cornstalk sample, (c) CRM from the system of [Amim][OAc] + TBD.

The cellulose structures of cornstalk sample, MCC, and CRM from the system of [Amim][OAc] + TBD were characterized by the solid-state ¹³C-NMR spectroscopy shown in Fig. 6. In the solid-state ¹³C-NMR spectra of MCC, the C4cr (C4 chemical shift at about 89 ppm) and C6cr (C6 chemical shift at about 65 ppm) are the crystalline cellulose characteristic peaks, respectively. The C4am (C4 chemical shift at about 84 ppm) and C6am (C6 chemical shift at about 63 ppm) are the amorphous cellulose characteristic peaks, respectively. The C2, 3, 5 and C1 are assignment of the chemical shifts for the C2, C3, C5 and C1 positions in cellulose.³⁵ In the cornstalk sample, the cellulose is also in microcrystalline state. The peaks of C4cr and C6cr are contributed to the crystalline region. The peaks of C4am and C6am are contributed to the amorphous region. However, in the CRM, the peaks of C4cr and C6cr were almost disappeared, and the peaks strength of C4am and C6am were enhanced, which indicated that the cellulose in the CRM was almost entirely amorphous cellulose.

The XRD spectra of cornstalk sample, MCC, and CRM from the system of [Amim][OAc] + TBD were showed in Fig. 7. The XRD data of cornstalk was used to calculate crystallinity index (CrI) as the eqn (7):

where CrI is the crystalline index, I₀₀₂ is the maximum intensity of the 002 lattice diffraction, near 2θ = 22.5°, and I_am is the intensity diffraction at suitable locations for the amorphous background at 2θ = 18.0°.³⁶

The major cellulose I peaks around 2θ of 14.9° and 16.5° are correspond to the 1–10 and 110 planes, while 22.4° and 34.6° are correspond to the 200 and 004 planes.³⁷ As shown in Fig. 7, the main peaks of MCC and cornstalk are around 2θ of 14.9°, 16.5°, 22.4° and 34.6°, which indicate that the cellulose of MCC and cornstalk sample are mainly in cellulose I form. However, the peaks from around 2θ of about 20.5° and 21.6° are correspond to the 110 and 200 planes of cellulose II, and the CRM shown in Fig. 7 are mainly in cellulose II.³⁸ According to the equal (7), the CrI were in an order of MCC (64.89) > cornstalk sample (37.87) > CRM from the system of [Amim][OAc] + TBD (21.93). The cellulose crystal form of cornstalk sample is similar to that of MCC, however, the CrI of cornstalk is lower than that of MCC, the regenerated CRM from the system of [Amim][OAc] + TBD is amorphous and in different crystal form. The results indicate that the cellulose crystalline form of cornstalk changes from cellulose I to cellulose II with an amorphous form during the dissolution and regeneration process. The more amorphous cellulose caused by adding TBD is conducive to the biomass utilization, which is evidenced by the high enzymatic hydrolysis rate.

The cornstalk sample, MCC, and CRM from the system of [Amim][OAc] + TBD were characterized by FT-IR and showed in Fig. 8. The increase intensity of peaks at 1370 cm⁻¹, and 898 cm⁻¹ (C–O stretching vibration in lignin, C–H bending vibration in cellulose and hemicelluloses and C–H deformation vibration in cellulose) in the CRM spectrum demonstrated that the cellulose or hemicelluloses were richer than that of cornstalk sample.³⁹ Meanwhile, the peak at 898 cm⁻¹ in CRM is enhanced compared with the cornstalk sample, demonstrating that the addition of TBD caused higher cellulose content in the CRM, which was verified in Fig. 8. The decreased intensity of peaks at 1740 cm⁻¹ (carbonyl group in hemicelluloses), 1603 cm⁻¹, 1513 cm⁻¹ and 1252 cm⁻¹ (aromatic rings and typical stretching vibrational bands of the C–C, C=C conjugate system of lignin) in the CRM indicated that large quantities of lignin had been removed by the pretreatment. And in the CRM, the peak at 1603 cm⁻¹ nearly disappeared, the intensity of peaks at 1740 cm⁻¹ and 1513 cm⁻¹ are also weaker compared with that of the cornstalk, which showed that the addition of TBD enhanced delignification of cornstalk sample.

Possible mechanism of cornstalk delignification in the system [Amim][OAc] + TBD

Combining the experiment results with the simulation of interaction energy of [Amim]⁺ and [OAc]⁻ when adding TBD, there could be two possible reasons to explain why this system [Amim][OAc] + TBD is helpful for delignification of cornstalk.

One is that the TBD itself alkalinity and exposed nitrogen atoms, which enable TBD to work more effectively as a dibasic nucleophile, assisting the cleavage of β-O-4 bonds in lignin. Fig. 9 illustrates the structures of the TBD and lignin portion, the exposed nitrogen atoms in TBD are active to cleave the β-O-4 bonds of lignin.²⁶ The alkalinity of TBD could be helpful for dissolving more lignin in [Amim][OAc], then the dissolved lignin would be catalysed into small molecules by cleaving the β-O-4 bonds in lignin with TBD. There may be an equilibrium between lignin dissolving and β-O-4 bonds cleavage in lignin, so adding different amount of TBD results in disparate delignification of cornstalk.

Fig. 9 The structure of TBD, and lignin portion is taken from ref. 40.

Secondly, the addition of TBD caused the interaction energy of [Amim]⁺ and [OAc]⁻ decreasing. So the [Amim]⁺ and [OAc]⁻ were more easily to be free from the ion-pairs or clusters form.²⁷ The structures of cation [Amim]⁺, the anion [OAc]⁻ and TBD were shown in Fig. 10, the interaction of TBD and [OAc]⁻, [Amim]⁺ were investigated, respectively, as shown in Fig. ESI 2 and 3.† There are possible positions in which anion locates around cation, and different initial configurations were optimized at B3LYP/6-311+G(d,p) level. Five conformers of [Amim][OAc] with different energies were finally obtained and shown in Fig. 10 (from 10A to 10E). Based on the structure with lowest energy, the interaction energy of [Amim]⁺ and [OAc]⁻ was calculated and obtained as 99.1 kcal mol⁻¹. Based on the structure TBD interacting with [Amim][OAc], the interaction energy of [Amim]⁺ and [OAc]⁻ was calculated and obtained as 89.2 kcal mol⁻¹. The optimized structures of TBD, [Amim]⁺ and [OAc]⁻ by B3LYP/6-311+G(d,p) was shown in Fig. 10F. With the influence of TBD, the interaction energy of cation and anion decreased from 99.1 kcal mol⁻¹ to 89.2 kcal mol⁻¹, which meant [Amim]⁺ and [OAc]⁻ were easier to release to interact with the components of biomass. So the cornstalk dissolution was accelerated and the delignification of cornstalk was enhanced by adding TBD in [Amim][OAc]. The interaction energy changes may be a useful parameter in choosing appropriate additive in ILs for biomass pretreatment.

Fig. 10 The structures of cation [Amim]⁺, the anion [OAc]⁻ and TBD, optimized structures of [Amim][OAc] ion pairs and optimized structures of TBD and [Amim][OAc] ion pair by B3LYP/6-311+G(d,p).

According to the decreased interaction energy of [Amim]⁺ and [OAc]⁻, a possible interaction pathway was proposed in Fig. 11. The anions and cations of ILs are not all dissociative but some in ion-pairs, even in the clusters form.²⁷ So in the [Amim][OAc], the [Amim]⁺ and [OAc]⁻ may interact with each other to produce acetic acid according to the pH of ionic liquid that Ober and co-workers reported.⁴¹ When adding TBD into the [Amim][OAc], the acetic acid would interact with TBD and produce acetate anion to produce TBD⋯H⁺⋯[Amim-H]. To investigate the TBD effect on reducing the interaction energy of [Amim]⁺ and [OAc]⁻, the mixed solution of [Amim][OAc] and TBD (with different mass ratios) was characterized by ESI-MS (shown in Fig. ESI 4†). The 140 amu peak was attributable to TBDH⁺, and the 262 amu peak was attributable to TBD⋯H⁺⋯[Amim-H]. So the TBD⋯H⁺⋯[Amim-H] could be existed in the mixed solution. When adding a little amount of TBD, the amount of TBD⋯H⁺⋯[Amim-H] was to very little compared with large amount of [Amim][OAc], so the 262 amu peak was weak. As the TBD amount was increased, the TBD⋯H⁺⋯[Amim-H] and TBDH⁺ peaks were enhanced, the [Amim]⁺ peak was almost disappeared. Therefore, the additive TBD could change the interaction energy of [Amim]⁺ and [OAc]⁻. The TBD enlarged the distance between [Amim]⁺ and [OAc]⁻ by interacting with [Amim]⁺ and [OAc]⁻, respectively. Therefore, the interaction energy of [Amim]⁺ and [OAc]⁻ was changed from 99.1 kcal mol⁻¹ to 89.2 kcal mol⁻¹ when adding TBD, and the decreased interaction energy may be helpful for the [Amim]⁺ and [OAc]⁻ to interact with the cornstalk components.

Fig. 11 Proposed mechanism for TBD decreasing the interaction energy of [Amim]⁺ and [OAc]⁻.

Those two factors together result in an effective system combining [Amim][OAc] and TBD for cornstalk pretreatment, which is helpful for accelerating the dissolution and enhancing the delignification of cornstalk.

Enzymatic hydrolysis

In order to further investigate the effect of TBD on cornstalk pretreatment, the CRM from the system of [Amim][OAc] + TBD at optimized condition together with cornstalk sample were practiced in the enzymatic hydrolysis. The enzymatic hydrolysis rates of CRM was about 98% at the 20 h, which was almost three times higher than that of cornstalk sample as shown in Fig. 12. The high cellulose conversion rate was contributed to the cellulose morphological structure changes and more amorphous cellulose obtained in the CRM, which is more accessible to the cellulase.⁴² This result is well consistent with the lowest decomposition temperature of CRM in the TGA spectra and the amorphous structures of CRM in the characterizations.

Fig. 12 Enzymatic hydrolysis of the cellulose in CRM from the system of [Amim][OAc] + TBD and cornstalk sample.

Conclusions

TBD is helpful for improving delignification and accelerating dissolution of cornstalk in [Amim][OAc] by decomposing the cleavage of β-O-4 bonds among lignin and changing the interaction energy between [Amim]⁺ and [OAc]⁻, which makes the [Amim][OAc] easier to interact with cornstalk components. Interaction energy changes between anion and cation may be a useful parameter in choosing appropriate additive in ILs for biomass delignification, and further investigations with diverse additives by means of both experimental and DFT calculations will be carried out to understand and prove the proposed hypothesis in the future. The obtained CRM is also proved to possess a high enzymatic hydrolysis rate. This facile approach is also considered to be effective in delignification of other lignocellulosic biomass.

Acknowledgements

This research was supported financially by the Projects of International Cooperation and Exchanges NSFC (no. 21210006, no. 21336002), National High Technology Research and Development Program of China (863 Program) (no. 2012AA063001), Key Project of Natural Science Foundation of Beijing of China (no.2131005).

Notes and references

1. B. Yang and C. E. Wyman, Biotechnol. Bioeng., 2004, 86, 88–95.
2. N. Mosier, C. Wyman, B. Dale, R. Elander, Y. Y. Lee, M. Holtzapple and M. Ladisch, Bioresour. Technol., 2005, 96, 673–686.
3. C. Cara, E. Ruiz, M. Ballesteros, P. Manzanares, M. J. Negro and E. Castro, Fuel, 2008, 87, 692–700.
4. T. A. Lloyd and C. E. Wyman, Bioresour. Technol., 2005, 96, 1967–1977.
5. A. H. Gao, M. V. Bule, D. D. Laskar and S. L. Chen, J. Agric. Food Chem., 2012, 60, 8632–8639.
6. P. Kumar, D. M. Barrett, M. J. Delwiche and P. Stroeve, Ind. Eng. Chem. Res., 2009, 48, 3713–3729.
7. B. Li, I. Filpponen and D. S. Argyropoulos, Ind. Eng. Chem. Res., 2010, 49, 3126–3136.
8. R. D. Rogers and K. R. Seddon, Science, 2003, 302, 792–793.
9. R. P. Swatioski, S. K. Spear, J. D. Holbery and R. D. Rogers, J. Am. Chem. Soc., 2002, 124, 4974–4975.
10. S. Barthel and T. Heinze, Green Chem., 2006, 8, 301–306.
11. T. Erdmenger, C. Haensch, R. Hoogenboom and U. S. Schubert, Macromol. Biosci., 2007, 7, 440–445.
12. Y. Fukaya, K. Hayashi, M. Wada and H. Ohno, Green Chem., 2008, 10, 44–46.
13. B. Kosan, C. Michels and F. Meister, Cellulose, 2007, 15, 59–66.
14. J. P. Mikkola, A. Kirilin, J. C. Tuuf, A. Pranovich, B. Holmbom, L. M. Kustov, D. Y. Murzin and T. Salmi, Green Chem., 2007, 9, 1229–1237.
15. J. Vitz, T. Erdmenger, C. Haensch and U. S. Schubert, Green Chem., 2009, 11, 417–424.
16. H. Zhao, Z. Song, O. Olubajo and J. V. Cowins, Appl. Biochem. Biotechnol., 2010, 162, 13–23.
17. P. Mäki-Arvela, I. Anugwom, P. Virtanen, R. Sjöholm and J. P. Mikkola, Ind. Crops Prod., 2010, 32, 175–201.
18. N. Sun, M. Rahman, Y. Qin, M. L. Maxim, H. Rodríguez and R. D. Rogers, Green Chem., 2009, 11, 646–655.
19. X. Wang, H. Li, Y. Cao and Q. Tang, Bioresour. Technol., 2011, 102, 7959–7965.
20. I. Anugwom, P. Mäki-Arvela, P. Virtanen, S. Willför, R. Sjöholm and J. P. Mikkola, Carbohydr. Polym., 2012, 87, 2005–2011.
21. H. Tadesse and R. Luque, Energy Environ. Sci., 2011, 4, 3913–3929.
22. H. Zhao, C. L. Jones, G. A. Baker, S. Xia, O. Olubajo and V. N. Person, J. Biotechnol., 2009, 139, 47–54.
23. A. R. Xu, J. J. Wang and H. Y. Wang, Green Chem., 2010, 12, 268–275.
24. J. van Spronsen, M. A. Tavares Cardoso, G. J. Witkamp, W. de Jong and M. C. Kroon, Chem. Eng. Process., 2011, 50, 196–199.
25. T. A. Nguyen, K. R. Kim, S. J. Han, H. Y. Cho, J. W. Kim, S. M. Park, J. C. Park and S. J. Sim, Bioresour. Technol., 2010, 101, 7432–7438.
26. S. Y. Jia, B. J. Cox, X. W. Guo, Z. C. Zhang and J. G. Ekerd, Holzforschung, 2010, 60, 577–580.
27. X. D. Hou, T. J. Smith, N. Li and M. H. Zong, Biotechnol. Bioeng., 2012, 109, 2484–2493.
28. M. L. Zhu, S. Li, Z. X. Li, X. M. Lu and S. J. Zhang, Chem. Eng. J., 2012, 168–177.
29. R. Ludwig, Phys. Chem. Chem. Phys., 2008, 10, 4333–4339.
30. P. J. Van Soest and J. B. Robertson, in Standardization of Analytical Methodology in Feeds, ed. W. J. Pigden, C. C. Balch, and M. Graham, International Research Development Center, Ottawa, Canada, 1980, pp. 49–60.
31. A. Brandt, J. P. Hallett, D. J. Leak, R. J. Murphy and T. Welton, Green Chem., 2010, 12, 672–679.
32. A. Brandt, J. Gräsvik, J. P. Hallett and T. Welton, Green Chem., 2013, 15, 550–583.
33. I. Kilpelainen, H. Xie, A. King, M. Granstrom, S. Heikkinen and D. S. Argyropoulos, J. Agric. Food Chem., 2007, 55, 9142–9148.
34. B. G. Janesko, Phys. Chem. Chem. Phys., 2011, 13, 11393–11401.
35. T. Mori, E. Chikayamac, Y. Tsuboi, N. Ishida, N. Shisa, Y. Noritake, S. Moriya and J. Kikuchi, Carbohydr. Polym., 2012, 90, 1197–1203.
36. C. Li, B. Knierim, C. Manisseri, R. Arora, H. V. Scheller, M. Auer, K. P. Vogel, B. A. Simmons and S. Singh, Bioresour. Technol., 2010, 101, 4900–4906.
37. N. Reddy and Y. Q. Yang, Bioresour. Technol., 2009, 100, 3563–3569.
38. P. Mansikkamäki, M. Lahtinen and K. Rissanen, Cellulose, 2005, 12, 233–242.
39. N. Labbe, T. G. Rials, S. S. Kelley, Z. M. Cheng, J. Y. Kim and Y. Li, Wood Sci. Technol., 2005, 39, 61–77.
40. J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius and B. M. Weckhuysen, Chem. Rev., 2010, 110, 3552.
41. A. Ober and R. B. Gupta, Ind. Eng. Chem. Res., 2012, 51, 524–2530.
42. N. I. Haykir, E. Bahcegul, N. Bicak and U. Bakir, Ind. Crops Prod., 2013, 41, 430–436.

---

Footnote

† Electronic supplementary information (ESI) available: Effect of the dissolution time on the cornstalk pretreatment with [Amim][OAc] alone, the density functional theory (DFT) calculations of TBD and [Amim][OAc] ion pairs and MS of [Amim][OAc] + TBD with different mass ratio. See DOI: 10.1039/c4ra02510e

---