Effects of Metal Chloride Salt Pretreatment and Additives on Enzymatic Hydrolysis of Poplar

: Metal chloride salt pretreatment was performed to isolate and convert cellulose to glucose from poplar. A glucose yield of 82.0% ± 0.7 was achieved after 0.05 mol/L AlCl 3 pretreatment conducted at 180 °C for 20 min, ascribing to the removal of hemicellulose, the alteration of crystallinity, surface morphology, and the retention of the majority of cellulose. Then, the influence of different additives on glucose yield was assessed, generating the highest glucose yield of 88.5 ± 0.06 with the addition of PEG 8000. Meanwhile, a similar glucose yield of 82.8% ± 0.3 could be obtained with PEG 8000 when hydrolysis time was reduced by a quarter and enzyme dosage by three - quarters. It can be seen that AlCl 3 pretreatment is a viable and efficient pretreatment method for poplar, while the addition of PEG 8000 can enhance the enzymatic efficiency and reduce cellulase loading, ascribing to the reservation of free enzyme and enzyme activity in the supernatant and the reduction in surface tension, which provide an idea to improve the economics of the enzymatic conversion of poplar.


Introduction
The problems of excessive energy consumption and climate change prompted countries all over the world to seek green and sustainable energy to replace fossil fuels [1,2].Bioethanol produced from lignocellulosic biomass is seen as a potential alternative energy [3].The conversion of lignocellulosic biomass into bioethanol includes the hydrolysis of cellulose to monosaccharide and the fermentation of monosaccharide to ethanol, in which enzymatic hydrolysis is the rate-limiting step [4,5].However, the stable and complex cross-link between hemicellulose and lignin and the natural lattice structure of cellulose result in an anti-degradation barrier for enzymatic hydrolysis [6].Therefore, it is essential to eliminate this obstacle through appropriate pretreatment to improve the cellulose accessibility to cellulase and make it easier to convert into fermentable sugars [7].
Various pretreatment methods have been investigated to fractionate biomass to improve enzymatic hydrolysis [8,9].Recently, acid, alkali, and hot water have been introduced to steam explosion, roasting, and organic solvent pretreatment, which could eliminate the carbohydrate complex of lignin by removing hemicellulose or lignin [10,11].Due to the production of "black liquor" during delignification pretreatment, pretreatment by removing hemicellulose was proposed to enhance enzymatic hydrolysis [12].Compared with inorganic acid pretreatment, metal chlorides are less corrosive and can be recycled by forming alkaline precipitation and reused for subsequent pretreatment, which reduces the chemical dosage and pretreatment cost.The previous literature reported that low-temperature (140 °C) pretreatment showed good selectivity for hemicelluloses in the presence of AlCl3, but there was no significant decomposition of cellulose and lignin [13].The presence of AlCl3 reduces the effect of decomposition temperature on biomass and significantly changes the solid structure of the acid due to its high dehydrating activity [14].Though the removal of hemicellulose could improve enzymatic hydrolysis, the retained lignin after AlCl3 pretreatment still impeded the efficiency due to being non-productive with cellulase and steric hindrance [11,15].To decrease the negative effect of lignin, additives (such as surfactants and proteins) were added to improve enzymatic efficiency.
Surfactants contain hydrophilic groups and lipophilic groups; the lipophilic groups are often hydrocarbons and water molecules, and the hydrophilic groups are charged or amphoteric groups that interact strongly with water molecules [16].Surfactants can significantly reduce the interfacial tension of solvents (usually water) and liquid-liquid interfacial tension [17].According to this property, different types of surfactants were used to bind with lignin, which reduced the affinity between cellulases and lignin by forming hydrophobic interactions with the residual lignin [15].The previous literature reported that the addition of additive (Tween 80 or BSA) could lessen enzyme loading by 33-50% [18].And the economic analysis indicated that adding PEG 3000 in enzymatic hydrolysis showed a 56% reduction in total costs [19].The presence of PEG 4600 during enzymatic hydrolysis also increased the sugar conversion yield [20].In addition to enzyme loading, the hydrolysis time also has an important role in affecting enzymatic cost.The enzymatic hydrolysis rate could be enhanced in a short time with the addition of additives, thus shortening the hydrolysis time while producing a similar sugar yield.However, there is less research about the reduction in cellulase loading and hydrolysis time simultaneously with the addition of various additives.
In this study, different metal chloride salt pretreatments were compared to remove hemicellulose and retain cellulose.Then, the optimal reaction temperature was also determined to obtain higher glucose yield.Meanwhile, the characterization of raw and pretreated poplar by XRD, FT-IR, and SEM was implemented to determine the changes in surface morphology and internal structure and their effects on subsequent enzymatic hydrolysis.Furthermore, additives were added to investigate the effects of types, surfactant loading, and cellulase loading on the glucose yield and increased yield.In addition, the alteration of the free enzyme, enzyme activity, and surface tension with the addition of additive was investigated.

Pretreatment and Enzymatic Hydrolysis
Poplar (15 g) was mixed with 150 mL of water containing 0.05 mol/L of metal chloride.The samples were heated at target temperature for 20 min at 300 rpm in a stainless steel reaction vessel and then cooled with icy water until they reached room temperature.The pretreated solids were filtered with a Brynner funnel and washed to neutral pH with deionized water.Subsequently, the pretreated solids were kept in the 4 °C refrigerator for further study [13].
Enzymatic hydrolysis of the raw and pretreated poplar (2 g) was implemented using commercial cellulase (20 FPU/g dry substrate) in 100 mL of sodium acetate buffer solution (pH 4.8).All samples were reacted at 50 °C for 72 h with a shaking speed of 150 rpm.Furthermore, the effect of additives (PEG 8000, WP, Tween 80, CL, and Triton X-100) on enzymatic hydrolysis was evaluated.During the hydrolysis, 1 mL of supernatant was sampled regularly for glucose analysis.All analyses were conducted in duplicate and results were expressed as mean ± standard deviation.

Analytical Methods
The chemical composition of raw and pretreated poplar was estimated using the analytical method of the National Renewable Energy Laboratory (NREL) [21].Briefly, the sample was hydrolyzed at 30 °C in 72% sulfuric acid for 60 min and then autoclaved at 121 °C for another 60 min after diluting to 4% sulfuric acid with deionized water.Then, the mixture was separated by filtration, and the supernatant was transferred to the HPLC system to determine the sugars concentration.The solid was dried and podzolized to determine the content of acid insoluble lignin (AIL).The HPLC was equipped with a KS-801 cation-exchange column and a refractive index detector (RID) using ultrapure water (0.4 mL/min) as the mobile phase [22].The solid recovery, glucose yields, and the increased yields of glucose with the additive were calculated according to the following formula: where Wbefore pretreatment and Wafter pretreatment were the weight of dry poplar.

Characterization of Raw and Pretreated Poplar
The X-ray diffraction (XRD) patterns of samples were recorded by Bruker D8-ADVANCE (Karlsruhe, Germany) with a Cu Kα radiation source (λ = 1.54 Å), and it was operated between 2θ angles of 5° to 60°.The crystallinity index (CrI) was determined using the following equation: in which I crystalline is the diffracted intensity of crystalline regions at about 2θ = 22.5°, and I amorphous is the diffracted intensity of the amorphous portion at about 2θ = 18.2°.The surface morphology features of different poplar solids were investigated with scanning electron microscopy (EVO18, ZEISS, Jena, Germany).Functional group changes in untreated and pretreated poplar were analyzed by Fourier transform infrared spectroscopy (FTIR) (Vertex70, Bruker, Berlin, Germany).The surface tensions of the supernatant samples were determined with a surface tension instrument (DCAT21, Dataphysics, Filderstadt, Germany) at 25 °C.

Determination of Activity and Content of Cellulase
The activity of cellulase was determined according to the 3,5-dinitrosalicylic acid (DNS) method [23].Specifically, filter paper (50 ± 5 mg) was added into 0.05 M citrate buffer solution (1.5 mL), then 0.5 mL of diluted enzymolysis supernatant was added into all the test tubes, except the blank test tube.After 60 min in a water bath at 50 °C, 3 mL of DNS was added into all tubes.Meanwhile, 0.5 mL of dilution was added into the blank tube.All tubes were incubated immediately for 10 min in a boiling water bath, then all tubes were fixed to 25 mL, and the absorbance was determined at 540 nm.Using bovine serum albumin as the protein standard, the content of cellulase protein in supernatant was determined according to Bradford method [24].

Statistical Analysis
The significance of the results of the glucose yield and increased yield was analyzed and set when the p value < 0.05.

Effect of Different Metal Chloride Pretreatment on Chemical Composition and Glucose Yield
Metal chlorides (MgCl2, ZnCl2, AlCl3, MnCl2, CuCl2, FeCl3) with low pKa were selected as acid catalysts by forming hydronium, which would cleave the glycosidic linkages in poplar during pretreatment [25].All pretreatments were carried out with 0.05 mol/L metal chloride with a solid/liquid ratio of 1/10 at 160 °C for 20 min.The effect of metal chloride pretreatment on chemical composition, solid recovery, and glucose yields were evaluated and demonstrated, as shown in Figure 1.For raw poplar, the contents of glucan, xylan, acid in-soluble (AIL), and ash were 45.6% ± 0.7, 15.0% ± 0.8, 22.9% ± 0.4, and 2.7% ± 0.2, respectively.As shown in Figure 1A, after the metal chloride pretreatment, the solid recovery ranged from 69.1% ± 0.6 to 91.1% ± 0.2.When MgCl2, ZnCl2, CuCl2, MnCl2, and FeCl3 were introduced to the pretreatment, the contents of cellulose, xylan, and AIL did not present obvious differences with raw poplar, and the removal of xylan (less than 42.3%) and slight degradation of AIL (less than 18.3%) contributed to the nearly 20% reduction in the solid [26].However, the weight loss for AlCl3-pretreated poplar was about 30%, ascribing to the large removal of xylan (92.3%), resulting in a low xylan content (1.7% ± 0.1) in the pretreated solid [27].This led to the increment of glucan and AIL to 60.7% ± 0.1 and 28.6% ± 1.5, respectively.It was reported that metal chlorides with lower pKa and chemical hardness values are better suited for biomass pretreatment [25].Though FeCl3 (2.46) had a lower pKa value than AlCl3 (4.85), the higher chemical hardness of AlCl3 (45.8 eV) led to the faster degradation of hemicellulose, suggesting that AlCl3 pretreatment had better ability to remove hemicellulose than FeCl3 pretreatment.
In order to compare the effect of metal chloride pretreatment on the following enzymatic hydrolysis, the glucose yield of pretreated solids after enzymatic hydrolysis was tested, as shown in Figure 1B.For raw poplar, the highest glucose yield of 11.3% ± 1.6 was obtained after hydrolysis for 24 h.Afte metal chloride pretreatment, the glucose yield increased steadily as the hydrolysis was progressed, which was higher than that obtained from untreated poplar [28].For AlCl3-pretreated solid, the glucose yield increased to 25.1% ± 0.5 after 6 h.As the hydrolysis proceeded to 72 h, the glucose yield increased sharply to 59.4% ± 0.6, which was 2.13 times higher than that obtained from other metal chloride pretreatments.This could be because AlCl3 pretreatment removed most of hemicellulose, as shown in Figure 1A, destroyed the intact structure of poplar, provided more accessible sites for enzymes, and promoted enzymolysis.Shen et al. found that the AlCl3-catalyzed hydrothermal pretreatment (170 °C, 1 h) removed almost all of the hemicelluloses and cleaved most of the β-O-4 linkages in lignin of Eucalyptus camaldulensis, yielding 77.8% cellulose conversion [13].Obviously, the glucose yield of AlCl3 pretreatment was much higher than that of other metal chlorides, and there were significant differences between AlCl3 and other metal chlorides, so AlCl3 was selected for the follow-up study.

Effect of Temperature on Chemical Composition and Glucose Yield
The previous literature reported that the temperature of pretreatment was a key factor in affecting enzymatic hydrolysis.Excessively low temperature was unable to disrupt the intact structure, while exorbitant temperature resulted in excessive energy expenditure and cellulose loss [29].Hence, the optimal temperature of AlCl3 pretreatment on poplar was investigated and the results were described in Table 1 and Figure 2. As shown in Table 1, the chemical composition of poplar pretreated under different temperature was presented.With the increase in pretreatment temperature from 150 °C to 190 °C, the solid recovery decreased gradually from 76.3% ± 0.2 to 51.8% ± 1.2, ascribing to a vast removal of xylan (77.5-100%), which resulted in the decreasing contents of xylan (0-4.4%).Meanwhile, though there was no obvious variation in the degradation of AIL (10.2-16.6%), the contents of AIL increased from 25.0% to 39.6% with the increment of pretreatment temperature.This phenomenon suggested that AlCl3 pretreatment had a strong ability to remove hemicellulose but had little effect on the lignin degradation [30].At the temperature of 180 °C, the glucan content reached 63.7% ± 0.3 with a glucan recovery rate of 88.7% [31].However, when the pretreatment temperature went up to 190 °C, the glucan content dropped sharply to 42.2% ± 0.8, proving that unduly high temperature could lead to excessive glucan degradation.Meanwhile, the influence of pretreatment temperature on enzymatic efficiency was implemented and the glucose yield was also detected and depicted in Figure 2. As shown, after AlCl3 pretreatment at 150 °C, the glucose yield was only 36.1% ± 1.9, which was a little higher than that obtained from raw poplar (11.3% ± 1.6).This phenomenon suggested that low pretreatment temperature was not sufficient to decompose the tight structure of poplar, leading to low enzymatic hydrolysis efficiency [32,33].With the increment of pretreatment temperature from 150 to 170 °C, the glucose yield increased gradually to 79.8% ± 2.2 after 72 h.When the pretreatment was conducted at 180 °C, the glucose yield of 79.8% ± 0.1 after 48 h was obtained, suggesting that a similar glucose yield can be achieved by severer pretreatment conditions and shorter hydrolysis time.Further extending the hydrolysis time to 72 h, the highest glucose yield of 82.0% ± 0.7 was obtained, accounting for about 92.5% of the glucose in the pretreated solids.However, as the pretreatment temperature continuously increased to 190 °C, the glucose yield after hydrolysis for 72 h did not raise, but decreased to 57.2% ± 1.3.This decline in glucose yields may be related to the loss of glucan in the pretreatment stage (shown in Table 1).Meanwhile, the nonproductive combination between lignin and enzyme decreased the accessibility of enzyme to cellulose, thus inactivating the cellulase activity and then reducing the efficiency of enzymatic hydrolysis [19,34].Thus, it can be seen that AlCl3 pretreatment can effectively improve enzymatic efficiency of poplar.Although results for glucose yield between 170 °C and 180 °C at 72 h were considered insignificant (Sig.= 0.386 > 0.05), there was significant difference in the yield between 170 °C and 180 °C at 48 h (Sig.= 0.041 < 0.05).The yield of glucose at 170 °C and 72 h (79.8% ± 2.2) was similar to that at 180 °C and 48 h (79.8% ± 1.0), indicating that higher temperatures can shorten the reaction time.Therefore, the optimal temperature of 180 °C was selected for subsequent investigation.

Physical and Chemical Structure Characterization of the Raw and Pretreated Poplar
The majority of xylan and some amount of lignin were removed during AlCl3 pretreatment, as illustrated in Table 1, which would change the crystallinity of the lignocellulosic biomass, thus affecting the following enzymolysis [35].Hence, the XRD patterns and corresponding crystallinity index (Crl) of raw poplar and AlCl3-pretreated solid (180 °C) were analyzed and shown in Figure 3A.Three peaks appear at about 2θ = 15.8°,22.5°, and 34.5°, corresponding to the crystal planes 101, 002, and 034 of crystalline cellulose I, respectively [36].There was no obvious alteration in these diffraction peaks except for the intensity, suggesting that AlCl3 pretreatment did not destroy the crystallization zone [37].For raw poplar, the CrI value reached 63.8%.After AlCl3 pretreatment, the CrI value was increased to 75.3%, attributing to the removal of amorphous hemicellulose and lignin, which destroyed the complete structure of poplar, thus providing more reactive sites for the accessibility of enzymes and improving the enzymolysis.
The changes in the infrared spectroscopy of raw and AlCl3-pretreated poplar were compared in Figure 3B.As shown, the absorption peak at about 1740 cm −1 related to the acetyl in hemicellulose almost disappeared in AlCl3-pretreated solid, indicating that hemicellulose was basically removed after AlCl3 pretreatment.The absorption peak of the aromatic ring in lignin at 1510 cm −1 appeared in both raw and pretreated poplar, suggesting that lignin was not significantly removed after AlCl3 pretreatment, which was in accordance with the chemical analysis (shown in Table 1) [38].Both the raw material and the AlCl3-pretreated solid showed strong absorption peaks of β-glycosidic bonds at 898 cm −1 , which are considered as the characteristic peak of the cellulose in lignocellulose [39].It can be seen that in contrast to the raw material, the intensity of characteristic peaks after AlCl3 pretreatment had little change, suggesting that the AlCl3 pretreatment on poplar would not lead to a large amount of cellulose degradation.These results further proved that AlCl3 pretreatment could achieve a large amount of hemicellulose removal, but had no significant effect on lignin degradation, and retained most of the cellulose, which was also consistent with the previous composition analysis.Figure 3C showed the surface morphology and ultrastructure of raw and pretreated poplar (180 °C) at 5000× magnifications.As shown, the surface structure of raw poplar was smooth with a compact morphology.The ordered arrangement of cellulose, hemicellulose, and lignin made it difficult for cellulase to access the internal cellulose bundle, so the glucose yield was extremely low [40].After pretreatment, the solid surface was obviously damaged, the structure became looser, and a large number of cracks appeared, corresponding to the large removal of hemicellulose, as in previous analysis.These cracks exposed cellulose to enzymes, leading to the improvement of enzymatic hydrolysis [41].

Effect of Additives on Enzymatic Hydrolysis
The previous literature reported that the remaining lignin after pretreatment would affect the subsequent enzymatic hydrolysis by non-productive adsorption and steric hindrance [15,32].Hence, different types of additives (non-catalytic proteins: WP; nonionic surfactants: PEG 8000, Tween 80, Triton X-100; ionic surfactants: CL) were added to investigate their effect on enzymolysis (Figure 4).As shown, obvious improvement with different additives on enzymatic hydrolysis had been observed [42].For the control sample without additive, the glucose yield of 82.0% ± 0.7 was obtained after 72 h.With the addition of additives, the glucose yield could reach this comparable level after hydrolysis for 24 h, suggesting that adding additives could accelerate the enzymolysis, shorten the enzymolysis time, as well as improve the enzymatic efficiency [15].Among these additives, PEG 8000 presented the optimal performance in promoting enzymolysis and produced the highest glucose yield of 88.5% ± 0.06, followed by WP (86.6% ± 0.2), CL (86.4% ± 0.06), and Triton X-100 (86.0% ± 0.01).The addition of Tween 80 had a relatively small increase on glucose yield (84.5% ± 1.2).The previous literature reported that the addition of Tween 80 to 0.5% HCl-catalyzed ethylene glycol pretreated sugarcane bagasse maintained a 91.4% glucose yield after 72 h [18].This phenomenon was not consistent with our results, ascribing to the different materials and pretreatment method.Lu et al. found that the addition of PEG 3000 to hydrothermal-alkaline/oxygen two-step pretreated reed enhanced the glucan conversion rate by 30.7% compared to that without surfactant [19].PEG 8000 combines its hydrophobic part with the hydrogen part of lignin in lignocellulose, which effectively prevents the non-productive binding of enzyme and lignin and promotes the combination of cellulase and cellulose, thus improving the efficiency of enzymolysis [43,44].Hence, the addition of PEG 8000 was a viable approach for weakening the adverse effect of lignin on cellulase during enzymatic hydrolysis.
To further investigate the enhancement of glucose yield by additives, the increased yield of glucose was detected at 6 h, 24 h, 48 h, and 72 h, and presented in Figure 4B.It can be seen that the addition of additives had different effects on enzymatic hydrolysis under extended hydrolysis time.After hydrolysis for 6 h, the range of increased yield was between 42.4% and 58.1%.As the hydrolysis time was extended to 24 h, the increased yield dropped sharply to 15.3-23.5%,and there was only a 3.1-7.9%increase in yield observed after hydrolysis for 72 h.These results showed that the additives improved the enzymatic hydrolysis mainly in the initial stage, which greatly reduced the unproductive adsorption of cellulase [45,46].As mentioned above, the addition of additives shortened the enzymatic hydrolysis time, and most enzymatic hydrolysis can be completed within 24 h.This was attributed to the fact that large amounts of cellulose had been transformed to glucose as the reaction went on, leading to less accessibility to the cellulose [47].According to the higher glucose yield (88.5% ± 0.1), increased yield (7.9-58.1%),and significance analysis (Sig.< 0.05), PEG 8000 was selected as the optimal additive for AlCl3-pretreated poplar for improving the efficiency of enzymatic hydrolysis and shortening the enzymatic hydrolysis time by preventing the contact between lignin and enzyme.Meanwhile, the effect of PEG 8000 loading on the enzymatic hydrolysis of AlCl3-pretreated poplar was explored.The glucose yield and increased yield under different hydrolysis time were depicted in Figure 5.As presented, with 5 mg/g substrate PEG 8000 loading, the glucose yield after 6 h reached 59.9% ± 0.1 while increased yield was 29.6%.Further extending the hydrolysis time to 72 h, the glucose yield rose gradually to 87.5% ± 0.9 with an increased yield of 6.8%, suggesting that the increased yield decreased as time went on.As the PEG 8000 loading increased to 25 mg/g substrate, the glucose yield increased to 71.9% ± 0.6 after 6 h while increased yield was 55.7%.As the hydrolysis time was prolonged to 12 h, a similar glucose yield was obtained as that without additive after 72 h, suggesting that the additive could shorten hydrolysis time by 5/6.Further increasing the PEG 8000 loading did not improve the glucose yield significantly because the maximum cellulose conversion was almost achieved [48].This phenomenon indicated that a small amount of surfactant can fully promote the contact between the cellulose and enzyme.Hence, the PEG 8000 loading of 25 mg/g solid was chosen for the subsequent study.During enzymatic hydrolysis, increasing the loading of cellulase could improve the conversion of lignocellulose to glucose; however, its high price also increases the product cost [49].Hence, the glucose yield and increased yield with additives under lower loading of enzyme were studied and presented in Figure 6.As shown, without PEG 8000, the glucose yield was only 54.6% ± 0.6 after 72 h under 2.5 FPU/g substrate enzyme loading.However, it increased sharply to 79.1% ± 0.6 by adding PEG 8000.Meanwhile, the increased yield presented an elevation from 15.7% at 6 h to 45.0% at 72 h, which was distinctive with that operated under high enzyme loading, indicating that lower enzyme loading worked better for the gradual enhancement of the increased yield [50].When the cellulase loading increased to 5 FPU/g substrate, the glucose yield without additive after 72 h was only 58.2% ± 0.4, and it was evaluated to 84.3% ± 0.1 with the addition of PEG 8000, which was higher than that obtained without the addition of surfactant under 20 FPU/g cellulase loading (82.0%), indicating that a preferable glucose yield could be achieved with the 3/4 reduction in cellulase loading by adding PEG 8000 [51,52].Meanwhile, the highest increased yield of enzymatic hydrolysis was achieved at 5 FPU/g cellulase loading.Compared with the control group, the increased yield reached 73.4% after 6 h and 72.0% after 24 h.Further extending the hydrolysis time to 72 h, the increased yield decreased to 44.9%.When the loading of cellulase was increased from 5 FPU/g to 20 FPU/g, a small increment in glucose yield from 84.3% ± 0.1 to 88.0% ± 0.4 was observed; however, a four-fold increase in cellulase loading was required [53].For increased yield, the higher cellulase loading at 72 h resulted in a lower increased yield, probably because more cellulase could be available for enzymatic hydrolysis under higher cellulase loading, which weakened the ability of the surfactant to combine with lignin to promote cellulose and cellulase binding; therefore, the enhancement effect of the surfactant was limited [54].Thus, the cellulase loading in this study was reduced by approximately 3/4.The previous literature indicated that the addition of PEG 3000 saves 85% of cellulose loading, this difference might be attributed to the different lignin content in pretreated substrate, which is larger in poplar in this work than in reed [19].Hence, the addition of PEG 8000 lowered the cellulase loading, boosted up the enzymatic hydrolysis, and enhanced the economic feasibility of the enzymatic hydrolysis.

The Mechanism of PEG 8000 on Enzymatic Hydrolysis
To explore the effect of PEG 8000 on the alteration of cellulase, the enzyme activity and the free enzyme in the enzymatic supernatant were analyzed and shown in Figure 7.All enzymatic hydrolysis was carried out under enzyme loading of 5 FPU/g substrate.As shown, the change trend of enzyme activity and free enzyme with or without PEG 8000 was similar.With the extension of enzymatic hydrolysis, the enzyme activity and the free enzyme in the hydrolysate decreased rapidly, and then gradually slowed down after 12 h, corresponding to the above-mentioned enzymatic hydrolysis that rapidly reacted at first and then gradually slowed down.For the control sample, the enzyme activity decreased to about 60.0% ± 2.2 after hydrolysis for 12 h.Meanwhile, about 36.1% ± 0.9 of cellulase was adsorbed on the substrate.This phenomenon suggested that the enzyme is adsorbed onto lignin and gradually inactivated due to irreversible adsorption [55].When PEG 8000 was added, the enzyme activity and the free enzyme were superior to those of the control group.After hydrolysis for 72 h, about 63.6% ± 0.6 of cellulase activity was retained, and 72.2% ± 1.7 of free enzyme was reserved in the enzymatic supernatant.This phenomenon was ascribed to the interaction between PEG and lignin on the surface of the substrate, preventing unproductive adsorption of enzyme, promoting the release of ineffective absorption enzymes, and providing more available cellulose, thus enhancing the enzymatic hydrolysis [56].The surface tension of the supernatant after enzymatic hydrolysis with or without surfactant was also determined in this study.As depicted in Figure 8, the X-axis represented the time required to stabilize the sample, the Y-axis was the surface tension, and the final surface tension value was analyzed by the software.The surface tension of the supernatant containing PEG 8000 (57.1 mN/m) was significantly lower than that without the addition of surfactant (65.1 mN/m).These results showed that the addition of PEG 8000 reduced the surface tension, reducing the contact of enzyme with air-liquid interface, resulting in enhancing the enzymatic hydrolysis.It was reported that the surface tension decreased by about 12.3% after adding 0.2 g/L rhamnolipid in the fermentation system, which decreased the enzyme loading from 15 to 10 FPU/g cellulose while achieving the same cellulose conversion rate [57].Similarly, the addition of PEG8000 effectively reduced the enzyme loading and improved the enzymatic hydrolysis efficiency, which further proved that PEG 8000 was an ideal additive.

Conclusions
Metal chloride salt pretreatment could remove a large amount of hemicellulose and retain most of the cellulose, where AlCl3 pretreatment at 180 °C achieved 100% hemicellulose removal and produced an 82.0% ± 0.7 glucose yield after hydrolysis for 72 h.The characterization studies further demonstrated that the AlCl3 pretreatment was able to improve the accessibility of enzyme to cellulose, thereby increasing the glucose yield.Furthermore, the addition of additives can effectively promote the enzymatic hydrolysis, and the highest glucose yield of 88.5% ± 0.6 was achieved with 150 mg PEG 8000 based on per g substrate by weakening the unproductive adsorption of lignin to cellulase.However, the parallel glucose yield (82.8% ± 0.3) could be achieved after 48 h with PEG 8000 loading of 25 mg/g substrate and cellulase loading of 5 FPU/g substrate, which reduced hydrolysis time by 1/4 and enzyme dosage by 3/4.The enhanced enzymatic efficiency was ascribed to the higher enzyme activity (63.6% ± 0.6), a greater percentage of free enzyme in the supernatant (72.2% ± 1.7), and the reduction in surface tension with the addition of PEG 8000.These findings will improve the production of fermentable sugars from lignocellulosic biomass using metal salt pretreatment and additives at low enzyme loading, thus making the production of biofuels more affordable.

3 BFigure 1 .
Figure 1.Chemical compositions of solid after pretreatment with metal chlorides (A) and glucose yield after metal chloride pretreatment (B).abcdefg : values listed in the same column with the same letter are not statistically different at 95% confidence intervals.

Figure 2 .
Figure 2. Effect of temperature on glucose yield of AlCl3 pretreatment.abcd : values listed in the same column with the same letter are not statistically different at 95% confidence intervals.

Figure 4 .
Figure 4. Glucose yield (A) and increased yield (B) under different surfactants (cellulase loading: 20 FPU/g, surfactant loading: 150 mg/g).abcd : values listed in the same column with the same letter are not statistically different at 95% confidence intervals.

Figure 5 .
Figure 5. Glucose yield (A) and increased yield (B) under different PEG 8000 loading (cellulase loading: 20 FPU/g).abcd : values listed in the same column with the same letter are not statistically different at 95% confidence intervals.

Figure 6 .
Figure 6.Glucose yield (A) and increased yield (B) under different cellulase loading (PEG 8000 loading: 25 mg/g).abc : values listed in the same column with the same letter are not statistically different at 95% confidence intervals.

BFigure 7 .
Figure 7. Enzyme activity (A) and the percentage of free enzyme (B) in the enzymatic suspension.

Figure 8 .
Figure 8. Surface tension of the enzymatic supernatant.

Table 1 .
The chemical composition of AlCl3-pretreated poplar under different pretreatment temperature.