Kaili Zhang ,Ligang Wei ,Qingqin Sun ,Jian Sun,Kunlan Li ,Shangru Zhai ,Qingda An,*,Junwang Zhang

1 Faculty of Light Industry and Chemical Engineering,Dalian Polytechnic University,Dalian 116034,China

2 School of Life Sciences,Advanced Research Institute of Multidisciplinary Science,Beijing Institute of Technology,Beijing 100081,China

Keywords:Ionic liquid Formaldehyde Low cost pretreatment Lignocellulosic biomass Saccharification

ABSTRACT Low cost processing of lignocellulosic biomass is of great importance for sustainable chemistry and engineering.Herein,a low cost system composed of 1-butyl-3-methylimidazolium chloride([C4C1im]Cl),HCl and formaldehyde (FA) was developed for the pretreatment of corn stalk at 80 °C.The efficiency of this technology was compared with that in dioxane system or without FA addition.Due to FA stabilization,the extent of acid-hydrolysis of carbohydrate fraction can be significantly decreased while 70% above of lignin was removed with the pretreatment of [C4C1im]Cl/HCl/FA system at 80 °C for 2 h.A maximum reducing sugar yield of 93.7%and glucose concentration of 7.0 mg·ml-1 were subsequently obtained from enzymatic hydrolysis of the slurry.There were great differences in compositions of small molecule degraded products obtained with FA addition or not.The present[C4C1im]Cl based system exhibits great potential of substituting volatile organic solvents(i.e.dioxane)in developing low cost lignocellulosic biomass pretreatment at low temperature.Also,this work would gain insight into understanding on the roles of stabilization methods on the economic improvement of IL based biomass processing.

1.Introduction

With diminishing reserves of fossil fuels,lignocellulosic biomass could be utilized as an alternative renewable feedstock for producing fuels and chemicals,which is of great importance from the viewpoint of environmental sustainable development and energy economics change [1,2].As one of the three main components (30%-50%) of lignocellulosic biomass,cellulose is a linear carbohydrate polymers that are linked by β-1,4-glycosidic bonds[3],and,can be converted into fermentable sugars through enzymatic hydrolysis but is under challenging due to its recalcitrant nature.For this context,pretreatment of biomass has been considered as a prerequisite step for increasing the accessibility of enzymes towards cellulose in order to afford more fermentable sugars.Pretreatment technologies by using various chemicals(such as dilute acid,alkali,organic solvent) have been extensively studied,but they suffer from either high environmental pollution or low efficiency [4].

Recently,the use of ionic liquids (ILs) has gained significant attention [5,6].As one kind of environmental benign solvent [7],ILs have high potential for efficient biomass pretreatment due to their strong hydrogen bond capability,high solubility,and tunable physico-chemical properties [8-10].However,the high cost of some effective ILs (i.e.1-ethyl-3-methylimdazolium acetate([C2C1im]OAc)) limits this technology more widespread application in the industrial scale.As a potential cost-effective alternative,biomass pretreatment combining a relatively inexpensive IL(i.e.1-butyl-3-methylimidazolium chloride([C4C1im]Cl))and mineral acid has been developed to reduce the physical and biochemical barriers for the enzymatic hydrolysis of cellulose [11].For instance,Zhanget al.[12]reported a pretreatment protocol of sugarcane bagasse using hydrochloric acid (HCl) and [C4C1im]Cl solution (containing 20% water) at 130 °C for 30 min.Bagasse delignification and xylan removal were increased with decreasing solution pH of aqueous [C4C1im]Cl systems from 5.9 to 0.4.In the case of IL solutions with pH ＜0.9,the glucan digestibility of above 90%can be obtained through enzymatic hydrolysis after 72 h.Lianget al.[13] evaluated the HCl-assisted [C4C1im]Cl pretreatment of municipal solid waste and biomass blends at milliliter scale and 6 L (120-160 °C and 2 h),respectively.The results indicated that[C4C1im]Cl had better pretreatment performance than 1-ethyl-3-methylimidazolium chloride in terms of sugar yields,and the scale-up resulted in a lower glucose recovery and xylose yield at the same pretreatment conditions due to the inadequate blending.Despite the above advances,the use of acid leads to the excessive hydrolysis of cellulose and hemicellulose in the pretreatment step,which results in the loss of the carbohydrates for the downstream enzymatic saccharification.

To address the issue of acid induced decomposition of carbohydrate polymers,Shuaiet al.[14] found that high yields of glucose(＞80%) were obtained after enzymatic hydrolysis of the leftover cellulose-rich materials obtained from HCl-catalyzed pretreatment with 1,4-dioxane in the presence of formaldehyde(FA)at 80°C for 5 h.It was speculated that the acetal structure with cellulose hydroxyl groups could be formed by adding FA,as well as with lignin side-chain hydroxyl groups.This work suggested that adding FA could depress the acid-hydrolysis of cellulose in the pretreatment.Different organic solvents (i.e.1,4-dioxane,γ-valerolactone and tetrahydrofuran)were used in the biomass pretreatment with FA.However,to the best of our knowledge,less efforts on the effect of FA addition on the pretreatment of [C4C1im]Cl/acid have been made up to now.

In this study,the HCl based pretreatment of corn stalk in[C4C1im]Cl with FA addition were systematically investigated at the mild conditions (80 °C and 1-5 h).Low cost mineral acid,(e.g.HCl),was used for the acidolysis of corn stalk was selected as biomass raw material.The influence of FA on the pretreatment performances including chemical composition,structure features of the recovered cellulose-rich material,cellulose digestibility and fermentable sugar production were investigated.The in-depth knowledge of this process will be beneficial for raising the profitability of the IL based biorefinery of lignocellulosic biomass.

2.Experimental

2.1.Materials

The raw corn stalk was obtained from Lianyungang in Jiangsu province,China.The air-dried corn stalk was ground and those with the particle sizes of 180 μm were collected.The chemical components of the raw and treated materials were determined according to the laboratory analytical procedure provided by the National Renewable Energy Laboratory(NREL)[15].All these analyses were conducted in duplicate to measure the error limit.

The used [C4C1im]Cl was purchased from Lanzhou Institute of Chemical Physics,Chinese Academy of Sciences (CAS).The purity of [C4C1im]Cl could reach above 99%,and the water content was below 0.4%.NaOH (99% purity) and HCl (37% purity) were purchased from Tianjin komeio chemical reagent Co.Ltd.Ethanol(96% purity) and FA (37%) were supplied by Tianjin Damao chemical reagent factory.All chemical agents were used as received without further purification.Deionized(DI)water with conductivity of 0.056 μS/cm was used in this work.

2.2.Pretreatment of corn stalk

The pretreatment solvent systems were prepared by combining 15 g of [C4C1im]Cl with a specific amount of HCl and FA solutions.In a typical procedure,corn stalk (1 g) was incubated in the prepared treatment solvent in a 50 ml flask.The flask and the contents were heated and stirred(350 r·min-1)in an oil bath.All the experiments were conducted in triplicates.

After the pretreatment,a 200 ml of NaOH solution(3%(mass)of NaOH)was slowly added into the stirred slurry for regeneration of the dissolved cellulose from corn stalk.The obtained suspension was centrifuged at 10,000 r·min-1for 10 min.The precipitate was washed 5 times with DI water at 70 °C in order to remove the excess [C4C1im]Cl.The supernatant was rotationally evaporated to remove the excess water.The obtained liquor was adjusted to neutral with 4 mol·L-1HCl solution,and then was added 3 volumes of ethanol,standing for 12 h until the hemicellulose precipitation.After that,the obtained solution was once again centrifuged under the conditions mentioned above.The ethanol was rotationally evaporated under reduced pressure.The final volume was 200 ml.The solution was acidified to pH 2,with 4 mol·L-1HCl solution,to precipitate the lignin material.Centrifugation and washing of sediments were conducted as described above.The specific diagram of the recovery of cellulose-rich materials,hemicellulose and lignin can be found in Fig.S1 in Supplementary Material.All the solid products were lyophilized for further characterizations.The conversion of corn stalk/components,the recovery of components and lignin removal were calculated according to the equations listed in Part A in the Supplementary Material.

Compared with the [C4C1im]Cl-based system,corn stalk pretreated with 1,4-dioxane was also carried out.In a typical procedure,1 g of corn stalk was mixed with 9 ml of 1,4-dioxane,420 μl of a HCl solution (37%),and 1 ml of a FA solution (37%).The reaction was conducted in an oil bath at a specified temperature(e.g.80°C)and stirred at 350 r·min-1.The detailed experimental procedure can be found as reported [14].

2.3.Analytical methods

2.3.1.Degraded products

2.3.1.1.13C NMR analysis of obtained liquid mixtures.Liquid mixtures obtained from the transformation were analyzed by13C NMR using a Bruker AVANCE 500 MHz NMR spectrometer.The liquid samples were diluted with DMSO-d6(samples/DMSO-d6,85:15,by mass).A total of 20,000 scans were collected for13C NMR of the sample solutions at 115 MHz and 25 °C to get spectra with good resolution.

2.3.1.2.Gas chromatography-mass spectrometry (GC-MS) analysis of small molecular products.The liquid mixtures obtained from the transformation were firstly extracted by ethyl acetate (EA),and subsequently extracted by methyltetrahydrofuran (MTHF).The products of EA-soluble and MTHF-soluble were obtained by the evaporation of solvent.The small molecular chemicals producing in the transformation were analyzed with GC-MS (Agilent 7890B-5977A)equipped with a HP5-MS column.The inlet temperature was 280 °C with a shunt mode.High purity nitrogen as the carrier gas.The oven temperature program began at 50 °C for 3 min,followed by a heating rate of 10 °C·min-1to 150 °C,then raised to 220 °C at a rate of 5 °C·min-1.Last,the temperature was increased to 300 °C at a rate of 10 °C·min-1,held for 3 min.

2.3.2.Cellulose-rich materials

The untreated raw corn stalk,and cellulose-rich materials obtained from the pretreatment were respectively characterized by the following methods.

2.3.2.1.Fourier transform infrared(FT-IR)spectroscopic analysis.The functional-group compositions of the samples were analyzed with a Perkin Elmer 94416 FT-IR spectrometer.Samples were ground and mixed with potassium bromide (spectroscopic grade),then pressed in a pressure machine to produce tablets with standard diameter.FT-IR spectra were collected in the wavelength range of 4000-400 cm-1with the spectral resolution of 4 cm-1.

2.3.2.2.X-ray diffraction (XRD) analysis.XRD data were recorded using a Rigaku D/Max-2500 V/PC power diffractometer with Cu Kα radiation (λ=0.154 nm) at 60 kV and 300 mA.The recording 2θ range was from 10°to 50°with the step size of 0.02°.Expressed as Crystallinity index (CrI) was calculated by Eq.(1) according the XRD data [16]:

whereI002is the maximum intensity of the 002 peak(cellulose I)at about 2θ=22.5°,andIamis the scattered intensity of peak in the amorphous phase at about 2θ=18.5°.

2.3.2.3.Scanning electron microscopy (SEM) analysis.A scanning electron microscope (JSM-6460LV,JEOL,Japan) was performed to give information on the surface morphological features of corn stalk before and after pretreatment.

2.3.2.4.Enzymatic hydrolysis.The cellulase-catalyzed hydrolysis of untreated and pretreated corn stalk was carried out at a solid loading of 1% (mass/volume) in a 50 ml flask to which 0.1 mol·L-1sodium citrate buffer(pH 4.8)was added.Cellulase(enzyme blend)is a loading of 1000 unit/g regenerated solid.The hydrolysis was performed at 50 °C in a thermostatic shaker at a rotating speed of 150 r·min-1.The samples taken periodically were analyzed using a UV-vis spectrophotometer (Agilent Cary 60 spectrophotometer,Germany at 540 nm) with the 3,5-dinitrosalicyclic acid(DNS)method[17].The content of total reducing sugar was calculated with glucose as the standard according to the absorbance of saccharification solutions.All assays were performed in triplicate.Error bars show the standard deviation of triplicate measurements.The yield of reducing sugar and cellulose digestibility from the lignocellulosic biomass was calculated as follows:

2.3.2.5.High-performance liquid chromatography (HPLC) analysis.Glucose content in the hydrolysates was determined by an HPLC (Waters e2695 Separations Module).Biorada Aminex HPX-87H sugar column was employed and the mobile phase was consisted of 5 mmol H2SO4with a flow rate of 0.6 ml·min-1.The temperature for the column was 55 °C.

3.Results and Discussion

3.1.Compositional analysis

The effects of FA addition on corn stalk conversion under different pretreatment times and solvents were investigated.Unless otherwise noted,the pretreatments were conducted at 80 °C,1.1% HCl content,and FA (40% (mass),on the basis of biomass mass)(if FA was added).In order to simplify the discussion,a more convenient notation was given corresponding to the pretreatment solvent and conditions.For instance,IL-FA-5 represents a sample obtained from the pretreatment with 1.1%HCl content and FA/biomass mass ratio of 0.4 at 80°C for 5 h.The main experimental conditions and corresponding annotations are summarized in Table 1.

The compositions of corn stalk and cellulose-rich materials recovered from the pretreatment are listed in Table 1.Conversions of corn stalk and the carbohydrates at different pretreatment conditions including with or without FA were compared and the results were summarized in Fig.1.The untreated corn stalk used in this work contained 38.9% cellulose,28.7% hemicellulose,and 22.6% lignin (dry and benzene-alcohol extractives free basis)(Table 1).After pretreatment for 1-5 h without FA,the recovery of cellulose varied from 10.9%to 24.5%indicating significant degradation of a large amount of cellulose.Especially,in a case of IL-5 pretreatment,78.1% of the corn stalk was converted into soluble degraded products(Fig.1).The conversions of cellulose and hemicellulose were 88.4% and 93.6%,respectively.After IL-5 pretreatment,cellulose-rich solids with a high lignin content of 25.6%were recovered (Table 1).These findings suggested that after[C4C1im]Cl/HCl pretreatment,the carbohydrate polymer was degraded,but it was not converted into the desired sugar,which caused the loss and inefficiency of subsequent enzymatic conversion.Other researchers also confirmed this conclusion.Unlike in aqueous-phase reactions,lignocellulose dissolution in [C4C1im]Cl treatment facilitated carbohydrate hydrolysis and lignin fractionation.However,the acid-catalyzed conversion drastically destroyed the structure of native lignin and condensed the degraded fragmental molecules [18],thereby further valorizing the lignin extracted that would normally be inaccessible.

In IL-FA-5 pretreatment,59.2%of the corn stalk was conversion(Fig.1).Compared with IL-5 treatment,cellulose and hemicellulose conversion decreased to 41.9% and 77.2%,respectively,upon FA addition.The cellulose-rich materials recovered from the pretreatment contained 59.0% cellulose,16.1% hemicellulose,and 16.3%lignin (Table 1).In particular,54.4% of cellulose in the corn stalk was recovered,and this percentage was five times higher than that obtained from IL-5 treatment (10.9%) (Table 1),indicating that FA addition inhibited the acid-hydrolysis of cellulose.FA reacted with the cellulose hydroxyls to hemiacetal and acetal structure catalyzed by HCl,and the possible reaction diagram is indicated in Fig.S2.Compared with unstable hemiacetal structure,the crosslinked acetal structure preferentially depressed the H3O+activation of β-1,4-glycosidic linkages in cellulose [14,19].The hydroxyl groups involved in the acetal reactions can come from intermolecular/intermolecular hydroxyls of cellulose as well as hemicellulose.Similarly,hemicellulose was stabilized upon FA addition in IL-FA-5 pretreatment.

The effects of pretreatment time on corn stalk conversion are shown in Fig.1.As pretreatment time increased from 1 h to 5 h,cellulose (36.3%-46.2%) and corn stalk (42.7%-59.6%) conversion remained almost the same.The easily degradable cellulose parts(i.e.,amorphous or not FA-stabilized parts) were likely hydrolyzed into soluble products during the initial reaction stage,whereas the other cellulose parts were hardly destroyed under the experimental conditions.As an easily degradable component,hemicellulose conversion reached 60.0% after IL-FA-1 pretreatment,and then it increased to 77.0% after IL-FA-2 pretreatment as incubation time was extended to 2 h (Fig.1).However,hemicellulose conversion halted as incubation time extended to 5 h;this cessation in conversion may be attributed to the effects of FA stabilization.

To understand further the effects of FA addition on carbohydrate conversion in the [C4C1im]Cl system,we investigated the effects of different pretreatments on corn stalk conversion under different conditions.The results were summarized in Table S1.In case of complete dissolution of corn stalk before the pretreatment(run S1),the conversion of corn stalk was 66.1%,which was higher than that obtained without pre-dissolution (59.6%).However,the degradation of carbohydrate fractions was less than that of the IL-5 treatment,implying that the hydrolysis of cellulose was inhibited due to the stabilization of FA(Table S1,run S1,in Supplementary Material).Compared with IL-FA-5 pretreatment (80 °C),the carbohydrate fraction was degraded as pretreatment temperature rose to 100 °C (run S2).A similar result was obtained in the pretreatment with FA and a high HCl content of 3.2% at 80 °C for 5 h (run S3).In addition,the effects of this FA/biomass mass ratioon corn stalk conversion were greater than that in other treatments.Notably,89.7% of lignin from the corn stalk was removed at a 0.2 of FA/biomass mass ratio(run S4).An appropriate amount of water in [C4C1im]Cl/HCl system can promote the release of H+and decrease the viscosity,thereby improving the pretreatment efficiency.Excessive water content in the system results in a decrease of pH,and further weaken deconstruction of biomass recalcitrance by IL [12].In addition,excessive water content was unfavorable for the aldolization reaction.The water content was about 4.1%-9.5% with varying FA/biomass ratio and HCl content.According to the results,there wasn’t obvious influence of water content on the pretreatment.It could be attributed to varying water content within relatively narrow ranges in this work.Anyway,effect of water content should be further systematically investigated.These findings suggested that the pretreatment parameters should be further optimized.The use of functional ILs and implementation of FA stabilization increase the possibility of developing a highly efficient acidolysis of lignocellulosic biomass conversion.

Table 1Compositional analysis of corn stalk and cellulose-rich slurry recovered from the pretreatment (dry and benzene-alcohol extractives free basis)①

Fig.1.Conversion of corn stalk and the carbohydrates in the pretreatments.

In a dioxane-FA-5(DO-FA-5)pretreatment(Fig.1),only 33.1%of the corn stalk was converted into soluble products.Cellulose and hemicellulose conversion were 37.4% and 66.2%,respectively.The cellulose conversion obtained from the dioxane treatment was slightly lower than that in the [C4C1im]Cl system,suggesting that[C4C1im]Cl treatment with good cellulose solubility may aid the hydrolysis and increase the accessibility of glycosidic bonds in cellulose.The IL-FA-5 and DO-FA-5 pretreatments had similarity in terms of reaction characteristics.Results demonstrated that there might be a synergistic effect between [C4C1im]Cl and FA on the efficient pretreatment.

The cellulose-rich materials recovered from the pretreatment could be further used in enzymatic hydrolysis of fermentable sugars.In DO-FA-5 pretreatment,cellulose-rich materials containing 55.9%cellulose,21.0%hemicellulose,and 14.0%lignin with a cellulose recovery of 62.2%(Table 1).A similar result was achieved in ILFA-2 pretreatment.IL-FA-2 pretreatment showed potential applications in biomass pretreatment similar to the DO-FA-5 treatment described by Shuaiet al[14].Furthermore,hemicellulose content in the cellulose-rich materials was lower but lignin removal was evidently higher in [C4C1im]Cl/HCl/FA pretreatment than in DOFA-5 pretreatment.These conditions favoured the subsequent enzymatic hydrolysis of cellulose [20].The HCl based treatment of corn stalk in[C4C1im]Cl with FA can be potentially used in lignocellulose degradation or pretreatment by optimizing the operation parameters.

3.2.Characterizations on the products from polysaccharides conversion

The13C NMR spectra of the liquid mixtures obtained from the pretreatment are displayed in Fig.2.The resonance signals observed at δC92.4(C1),74.8(C2),73.1(C3),72.7(C4),76.5(C5),and 61.3(C6) were assigned to glucose,whereas those observed at δC97.1(C1),75.3(C2),73.3(C3),70.6(C4),and 66.1(C5) were attributed to xylose [21,22].Nevertheless,the resonance signals of the other monosaccharides (i.e.,arabinose and galactose released by hemicellulose hydrolysis) were too weak to be observed in the13C NMR spectra.This result confirmed that glucose and xylose were the predominant components in the hydrolysates.Some effective methods on the sugar recovery from IL mixtures have been proposed by the researchers.For example,alumina column chromatography (ACC) toward monosaccharides and ILs were extensively studied,indicating that ACC can effectively separate IL and monosaccharide from hydrolyzed mixture [23-25].

During the biomass treatment with aprotic solvents (i.e.,dioxane and THF),acetal formation in the monosaccharides (glucose or xylose) occurred upon FA addition [26].Acetal formation impeded the further conversion of the monosaccharides into furfural derivatives.However,the resonance signals assigned to the acetal structure were drowned in the level of signal noise and not as remarkable as those attributed to the monosaccharides in the13C NMR spectrum of the liquid mixtures obtained from ILFA-5 pretreatment (Fig.2b).Whether aldolization between the monosaccharides and FA occurred in the [C4C1im]Cl/HCl system could not be confirmed in this work.This question deserves further investigation.

Fig.2.13C NMR spectra (55-105) of the liquid mixtures obtained after IL-5 (a) and IL-FA-5 (b) pretreatments.

The liquid mixtures obtained from the pretreatment were subsequently extracted via EA and MTHF to identify clearly the compositions of the small molecular products.The GC-MS chromatograms of the small molecular products obtained from IL-5 and IL-FA-5 pretreatments are displayed in Fig.3.This treatment was complicated in terms of the number of products with dozens of chemicals generated.

Most species of detectable degraded products were furfural derivatives,esters,alcohols,aldehydes,and long-chain hydrocarbons (Table 2).These products normally originate from cellulose and hemicellulose [27].The relative abundance of these products with broad distributions was low,and even some of them were found in trace quantities.The effects of FA addition on the degraded products were reflected by the differences in compounds obtained from cellulose and hemicellulose depolymerization(Table 2).In IL-5 pretreatment,the detectable products contained ethyl levulinate (EA-soluble,8.56 min),4-penten-2-ol (MTHFsoluble,6.84 min),and ethyl hydrogen succinate (MTHF-soluble,10.93 min).When FA was added in the pretreatment,GC-MS could not detect these same compounds but recognized γ-valerolactone(MTHF-soluble,6.69 min),5-hydroxymethylfurfural (MTHFsoluble,11.39 min),and 5-[(tetrahydro-2H-pyran-2-yl)oxy]-penta nal (MTHF-soluble,12.66 min).This variation in the results of GC-MS analysis indicated that the acid degradation of biomass followed different mechanisms in the presence of FA.

Fig.3.GC-MS chromatograms of small molecular products obtained from IL-5 (a,c) and IL-FA-5 (b,d) pretreatments in the [C4C1im]Cl system.

Table 2Differences in the results of GC-MS analysis of EA-soluble and MTHF-soluble compounds

3.3.Characterizations of cellulose-rich materials

The changes in chemical compositions and structures of the corn stalk after the pretreatments were monitoredviaFT-IR analysis (Fig.4).The assignments of the main bands are listed in Table S2.Compared with the spectrum of the untreated corn stalk(Fig.4a),a number of bands in the FT-IR spectra of the celluloserich materials from IL-5 pretreatment (Fig.4b) disappeared or became weak,indicating that the cellulose-rich materials underwent considerable structural changes during the pretreatment.By contrast,upon FA addition,most functional groups in corn stalk biomass(i.e.,cellulose and hemicellulose)could be observed in the FT-IR spectra (Fig.4c-e) after IL-FA-5 or IL-FA-2 or DO-FA-5 pretreatment.

Fig.4.FT-IR spectra of untreated corn stalk (a) and cellulose-rich materials recovered after the pretreatments of IL-5 (b),IL-FA-5 (c),IL-FA-2 (d) and DO-FA-5(e).

The effects of different pretreatment conditions on cellulose structure were verified by the variations in the bands at about 3400 cm-1,which was ascribed to O-H stretching vibration(Fig.4).Compared with the FT-IR spectrum of the untreated sample,this O-H stretching band evidently became weak after IL-5 pretreatment,but no noticeable differences (the bands even became strong;Fig.4c) were observed upon FA addition (IL-FA-5 pretreatment).A similar tendency was also observed in the bands at about 2920 and 2850 cm-1,which were assigned to the stretching of-CH3and-CH2groups in cellulose,respectively.These findings indicated that the intermolecular hydrogen bond network in cellulose was prevented from being destroyed by [C4C1im]Cl/HCl pretreatment upon FA addition,which were consistent with those depicted in Fig.1.

The main differences in the FT-IR spectra of cellulose,hemicellulose,and lignin in corn stalk biomass generally existed at wavenumbers lower than 2000 cm-1.The bands at 1630,1605,and 1510 cm-1,which are considered as the characteristic absorbance of the aromatic skeleton in lignin,decreased in the spectra of the treated samples and the bands at 1510 cm-1even disappear.The decrease was due to the partial removal of lignin after pretreatment.The decrease in the band intensities was proportional to the rate of lignin removal(Fig.S3).The bands centered at about 1732 and 1250 cm-1were attributed to the stretching vibrations of C=O and C-O,respectively,in the ester linkages between lignin and hemicellulose [28,29].These bands almost disappeared in the spectra of the samples after pretreatments with and without FA (Fig.4b and c),suggesting that [C4C1im]Cl/HCl pretreatment was effective in breaking the ester linkages between lignin and hemicellulose despite FA addition.Furthermore,compared with the spectrum of the sample in IL-5 pretreatment,the intensities of peaks at 1056 cm-1(C-O stretching) and 1375 cm-1(C-H deformation),which were attributed respectively to cellulose and hemicellulose in the spectra of the samples after the pretreatment with FA addition (Fig.4c-e),increased.Specially,after DO-FA-5 pretreatment,the absorption peak is the strongest at the 1375 cm-1.The increase in peak intensities could be ascribed to the increase in cellulose and hemicellulose content in the samples as a result of lignin removal and FA stabilization.

Compared with the spectrum of the untreated sample(Fig.4a),the intensities of peaks at 1430 (-CH2bonding mainly derived from crystalline cellulose) and 898 cm-1(C-O-C stretching of β-1,4-glycosidic bonds in amorphous cellulose) after [C4C1im]Cl/HCl/FA pretreatment barely changed.The intensities of these bands were considerably stronger than those obtained after IL-5 pretreatment (Fig.4b),suggesting that FA stabilization inhibited cellulose degradation.These results agreed with those summarized in Table 1 and Fig.1.

Fig.5.XRD diffractograms of untreated corn stalk (a) and the cellulose-rich materials recovered after the pretreatments of IL-5 (b),IL-FA-5 (c),IL-FA-2 (d) and DO-FA-5 (e).

XRD analysis was performed to monitor the changes in cellulose crystalline in the corn stalk after pretreatment(Fig.5).The peaks in the XRD pattern of the untreated corn stalk located at 15.4°,16.6°,22.5°,and 34.7°corresponded to(1-1 0),(1 0),(2 0),and(0 4)lattice planes of crystalline cellulose I,respectively (Fig.5a).The flat peak at 18.5° was attributed to the amorphous regions related to disordered cellulose,hemicellulose,and lignin.After IL-5 pretreatment,peaks intensities corresponding to crystalline cellulose I dramatically decreased compared with those of the untreated sample(Fig.5a).Moreover,a diffraction peak at 21.5° representing the characteristic (0 2 0) lattice plane in cellulose II was observed.The crystallinity of the sample in IL-5 treatment (44.4 CrI) was lower than that of the untreated corn stalk (45.7 CrI).These findings indicated that the crystalline structure of cellulose was partially destroyed by [C4C1im]Cl/HCl pretreatment.

Upon FA addition,obvious shifts in typical peaks corresponding to cellulose I were not observed (Fig.5c-e).However,the intensities of these peaks were greatly affected.After DO or [C4C1im]Cl/HCl/FA pretreatment,CrI (47.1-58.1) evidently increased compared with the untreated sample.This result suggested that the amorphous components were considerably broken under the experimental conditions.This suggestion was confirmed by the decrease in intensity of the peak located at 18.5°.In particular,Huet al.[30] reported an increase in crystallinity of corn stalks after IL/water pretreatment.These results support that the structural destruction and removal of lignin/hemicellulose lead to the improvement of enzyme efficiency rather than the destruction of cellulose crystallinity.In addition,the increase in CrI indicated that this pretreatment was unable to break the inter-and intra-chain hydrogen bonds in cellulose fibrils.This phenomenon was similar to that observed in dilute acid pretreatment [31].

The XRD patterns of the samples treated without (Fig.5b) and with FA addition (Fig.5c-e) were vastly different.The water content in the[C4C1im]Cl system was about 6.1%because the addition of FA solution(containing 10%-15%methanol as the stabilizer)carried small amounts of water with it.Nevertheless,this event could not have resulted in the substantial influence of FA addition on the capability of dissolution or disruption of hydrogen bonding in crystalline cellulose[32].These findings indicated that FA stabilization protected the hydrogen bonds in cellulose from the cooperative actions of [C4C1im]Cl and H3O+.

SEM micrographs of the samples before and after the pretreatment are shown in Fig.S5.The surface topography of the untreated corn stalk (Fig.S5a) exhibited a clear and regular lignocellulosic fiber-shaped structure.[C4C1im]Cl/HCl pretreatment was sufficient to completely deconstruct the corn stalk into fine fibers or particles with a rough surface(Fig.S5b),suggesting that the corn stalk organization remarkably changed probably because of the extensive corn stalk degradation and lignin removal(Fig.1 and Table 1).After FA addition,lignocellulosic fibers remained (Fig.S5c),but these disorganized fibers were substantially shorter than those of the untreated corn stalk.Unlike[C4C1im]Cl/HCl pretreatment,lignocellulosic fibers appeared to be preserved to some content in the presence of FA,although their surfaces became irregular.These results arose probably because of the inhabitation of cellulose conversion due to FA stabilization.

To evaluate the efficiency of the pretreatments on sugar production,the corn stalk samples before and after pretreatments were subjected to a 96 h enzymatic hydrolysis.Fig.6 indicates the time profiles of reducing sugar yields from different samples.Fig.S4 shows cellulose digestibility for untreated and pretreated corn stalk,respectively.All the pretreated corn stalk had a higher reducing sugar yield in comparison with the untreated sample(37.9%) after 96 h.It can be attributed to the increase of cellulose accessibility because the biomass structure was destroyed and hemicellulose or/and lignin were partially removed in the pretreatments.As seen in Fig.S4,among samples for pretreatment,the highest cellulose digestibility was obtained via IL-FA-2 pretreatment,which was 97.7% at the 96 hour of enzymatic hydrolysis,being almost 3-fold more digestible than untreated corn stalk.This result is in good agreement with the SEM image displaying a remarkably deconstructed structure for pretreated corn stalk.[C4C1im]Cl pretreatment achieved a better cellulose digestibility than dioxane pretreatment.

Fig.6.Enzymatic hydrolysis of untreated corn stalk and cellulose-rich materials recovered in the pretreatments.

For the cellulose-rich materials obtained with [C4C1im]Cl/HCl pretreatment with FA addition,the reducing sugar yields from enzymatic hydrolysis after 96 h firstly increased and then decreased with increasing pretreatment time from 1 h to 5 h.The maximum reducing sugar yield (93.7%) was achieved with the samples pretreated for 2 h.A relative high content of lignin(16.6%) and hemicellulose (20.3%) remained in the sample pretreated at a short time (1 h) depressed the cellulose accessibility.When the pretreatment time was increased to 5 h,the reducing sugar yield dropped to 85.9%.In the pretreatment,condensation of lignin fragments was liable to occur.The condensed lignin could redeposit on the surface of the pretreated materials and preventing cellulose from enzymatic deconstruction.As depicted in Fig.6,the reducing sugar yield of the samples with [C4C1im]Cl/HCl pretreatment with FA addition the was similar to that with the DO-FA-5 pretreatment (90.1% after 96 h).It suggested that [C4C1im]Cl as the solvent,like 1,4-dioxane,also can achieve good pretreatment results.

Fig.7.Glucose concentrations obtained in enzymatic hydrolysis after 96 h.

As shown in Fig.7,the variation tendency of glucose concentrations of the hydrolysis liquors obtained in different conditions was similar to that of the reducing sugar yields.The highest glucose concentration of 7.0 mg·ml-1(hydrolysis for 96 h) was also obtained with the cellulose-rich material after IL-FA-2 pretreatment.

Qinget al.[33] demonstrated that the reducing sugar yield of about 82.9% was obtained in the case of HCl-catalyzed pretreatment of corn stover in 1,3-dimethylimidazolium dimethylphosphate at 110 °C for 2 h.Wanget al.[34] reported that pretreatment of bagasse pith in the [C4C1im]Cl solutions (1.0%HCl)at 120°C,and the maximum TRS yield reached to 89.9%after pretreatment and enzymatic hydrolysis of 72 h.Generally,like the other process reported in literatures,the [C4C1im]Cl/HCl pretreatment with FA addition demonstrated in this work was effective to destroy the biomass recalcitrance and enhance the efficiency of enzymatic hydrolysis,but was carried out at a relatively low temperature (80 °C).

4.Conclusions

Corn stalk pretreatmentsviaa low cost [C4C1im]Cl/HCl system with FA addition were investigated at a low temperature of 80 °C.[C4C1im]Cl was selected as the solvent because of the nonvolatile,high acid tolerance and good dissolution capacity of lignocellulosic biomass instead of volatile organic solvents(i.e.1,4-dioxane).With the aid of FA stabilization,acidolysis of the cellulose fraction of the corn stalk substantially decreased.In the meantime,above 70% lignin was removed and the structure recalcitrance of biomass was destroyed.For the sample after the IL-FA-2 pretreatment,the maximum reducing sugar yield (93.7%)was obtained as well as the highest glucose concentration of7.0 mg·ml-1(hydrolysis for 96 h).The efficient pretreatment was achieved with the combination of [C4C1im]Cl/HCl and FA at low temperature (80 °C) for a short time (2 h).On the basis of these preliminary results,we aim to develop an efficient the HCl based pretreatment process of lignocellulosic biomass with the cooperative action of structural stabilizers (i.e.,FA) and functional taskspecific ILs by optimizing operation parameters.

CRediT authorship contribution statement

Kaili Zhang:Conceptualization,Methodology,Software,Formal analysis,Writing -original draft.Ligang Wei:Writing -review &editing.Qingqin Sun:Writing-review&editing.Jian Sun:Supervision,Writing-review&editing.Kunlan Li:Supervision.Shangru Zhai:Writing -review &editing.Qingda An:Funding acquisition,Supervision.Junwang Zhang:Software,Validation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This study was supported by the National Natural Science Foundation of China(21776026,22078023)and Liaoning Revitalization Talents Program (XLYC1902037).

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2021.01.001.