Enhancement of Liquid Hot Water Pretreatment on Corn Stover with Ball Milling to Improve Total Sugar Yields

: Conversion of the lignocellulosic biomass to bioethanol contributes to the reduction of greenhouse gas emissions, enhancement of energy security, utilization of waste materials, and the promotion of sustainable agricultural practices. In this study, we report the e ﬀ ect of combining ball milling followed by liquid hot water (LHW) pretreatment of corn stover to lower the amount of enzyme required while also greatly increasing the recovery of xylose in fermentable form compared to either pretreatment alone. Short-duration ball milling for 60 min reduces the particle size of corn stover to 37.3 µm; however, the glucose only increased to 47% compared to 32% for unpretreated corn stover. In contrast, liquid hot water pretreatment alone can achieve increasing enzyme hydrolysis yields of cellulose from 49% to 93% as the pretreatment severity factor is increased from 3.24 to 4.41. However, the xylose yield decreased to 36% due to the fact that a considerable part of the xylose was degraded into furfural and humins. Surprisingly, the combination of mild ball milling (30 min) followed by mild liquid hot water pretreatment (190 °C, 15 min) could achieve both high glucose (83%) and xylose (72%) yields for a total sugar yield of 79%, theoretically. Thus, combining ball milling with liquid hot water pretreatment allows for milder conditions for both processes that lead to enhanced cellulose conversion without sacri ﬁ cing xylose to degradation, which hinders enzymatic hydrolysis.


Introduction
The lignocellulosic biomass has gained considerable attention as a promising and sustainable resource for fuel production.This interest has been driven by growing concerns about climate change resulting from the excessive use of fossil energy.[1][2][3].Corn is grown on approximately 160 million farms and is currently left on the fields to decompose, making it a promising feedstock for the production of biofuels and chemicals.[4].Bioethanol is produced from renewable lignocellulosic biomass sources, such as corn, making it a sustainable energy source.The conversion process of the lignocellulosic biomass to bioethanol primarily involves four steps: pretreatment, enzymatic hydrolysis of cellulose and hemicellulose into monosaccharides, fermentation of monosaccharides into ethanol, and the subsequent separation and purification of bioethanol.The lignocellulosic biomass primarily comprises cellulose, hemicellulose, and lignin, which are interconnected.The complexity of these components contributes to biomass recalcitrance, leading to low efficiency and high costs in the conversion of bioethanol [5,6].Therefore, pretreatment of the lignocellulosic biomass is crucial to modify its structure and enhance the accessibility of cellulose and hemicellulose.Various pretreatments have been utilized to enhance sugar yields of the lignocellulosic biomass, such as mechanical grinding, liquid hot water, acid, and alkaline pretreatments [7,8].
Mechanical grinding is usually considered as the first step to pretreat the lignocellulosic biomass and has been extensively utilized in the bioprocess industry [9,10].Mechanical grinding reduces the particle size and increases the porosity of this compound, which could effectively enhance the contact reaction between cellulose/hemicellulose and the enzyme.Moreover, at the more extreme conditions, ball milling destroys the plant cell wall matrix (cellulose), hemicellulose and lignin, thus increasing the enzymatic hydrolysis yields of the cellulose.It is worth noting that the destruction of crystalline cellulose has been shown to be more important than particle size reduction in the enhancement of sugar yields through ball milling [11].Several studies have shown that mechanical grinding at a cellular scale, such as ball milling, could significantly destroy cell wall structure and crystalline cellulose, resulting in a notable improvement of enzymatic hydrolysis efficiency [12,13].However, the high cost of energy input during long ball milling process times makes the process unattractive for commercial implementation; thus, it is essential to reduce the energy cost during the mechanical grinding process.
Liquid hot water pretreatment (LHWP) is an environmentally friendly process without chemicals and entails low operation and capital costs compared to traditional acid-or alkaline-based processes [14].Hydronium ions serve as catalysts to hydrolyze and solubilize hemicellulose at high temperatures, and lignin is partially melted and redistributes within the matrix structure, which increases the accessibility of cellulose to enzyme catalysts in the subsequent hydrolysis process.However, lignin exposed on the surface of the biomass, after LHWP, inhibits the enzyme's access to cellulose by non-productively adsorption of enzyme [15].Another challenge is to achieve a high recovery of both pentose and hexose with one-step LHWP because hemicellulose is more susceptible to hydrolysis and the xylose is more easily degraded than cellulose and glucose.For example, LHWP at high temperatures could achieve high glucose yields but sacrifices xylose yields, which are degraded to by-products, such as furfural.Under these pretreatment conditions, furfural can further react to form pseudo-lignin that redeposits on the cellulose surface, hindering enzymatic hydrolysis [16].Glucose yields and xylose yields are both important for the bioethanol production process since both are potentially fermentable to bioethanol [17].
Recently, there has been a notable increase in attention towards the integration of physicochemical or chemical pretreatments with physical pretreatment methods.These combined pretreatment approaches are capable of effectively overcoming the limitations of individual pretreatment methods while amplifying their respective advantages.To overcome the shortcoming of ball milling pretreatment and liquid hot water pretreatment, combined pretreatment was also studied.Kim et al. proposed hot water pretreatment at 160-200 °C for 4-8 min followed by disk milling, resulting in an improvement of glucose and xylose yields by 89% and 134%, respectively [18].However, the sequential application of LHWP followed by milling has a significant drawback in terms of its high energy demand for the water drying process.
Therefore, in this study, we combined short-duration ball milling followed by liquid hot water pretreatment under mild conditions on a corn stover to achieve both high glucose and xylose yields compared to ball milling and liquid hot water pretreatment alone.The effects of combining pretreatments on the chemical composition of the biomass throughout the process and particle size distribution were investigated.The total sugar yield (including glucose and xylose yield) at low and high enzyme loading as well as energy cost were also discussed in this study.The enhancement of bio-conversion efficiency with the combined pretreatment of ball milling and liquid hot water pretreatment could effectively contribute to mitigating climate changes and achieving sustainable development goals.

Raw Materials
Corn stover, collected from a commercial corn field, was milled using a hammer mill (Thomas Ltd., Oak Park, IL, USA) with the final sample passing through a 1.00 mm screen.The moisture of corn stover was 9.26%.The extractives, carbohydrates, lignin, and ash contents were determined using the NREL procedure.

Ball Milling Method
A Planetary Micro Mill Pulverisette 6 (Fritsch, Idar-Oberstein, Germany) was used to ball mill corn stover.Sample (10 mL) was milled in an 80 mL zirconium oxide grinding ball with 20 mL zirconium oxide balls (φ = 10 mm) at a speed of 300 rpm for 20, 30, and 60 min.To prevent high temperatures from affecting the milling process, milling cycles were alternated for 10 min of active milling followed by 10 min of rest to cool.

Liquid Hot Water Pretreatment
LHWP of raw and ball-milled corn stover was conducted in stainless steel tubes according to Kim (2014b) [15].Each tube was 45 mL in total volume and 11.4 cm in length.Corn stover (5.57g) and water (28.13 mL) were mixed in the stainless steel tubes to keep the total volume at 33.7 mL, providing enough space for liquid expansion at high temperature.The tubes were preheated at 140 °C for 10 min and then heated to the target temperature (Table 1) for 15 min.After each pretreatment, the tubes were immediately placed in cool water for 10 min for quenching.The slurry was washed with 100 mL 80-90 °C hot water and then vacuum filtered to recover solid and liquid fractions.
All pretreated corn stover samples are shown in Table 1.Severity factor (log R0) of LHWP was determined according to the following equation [19]: where t is the reaction time (min); T is the pretreatment temperature (°C).
The liquid fraction was analyzed for monosaccharides and by-products (e.g., furfural and 5-hydroxymethyl furfural) according to NREL procedure [20].To determine the oligosaccharides in liquor, the samples were hydrolyzed with 4% sulfuric acid at 121 °C for 1 h.Total phenolics in liquid fraction were determined with Folin-Ciocalteau method [21].

Particle Size Distribution
Particle size distribution was determined with laser diffraction particle analysis using a Mastersizer3000 (Malvern Instrument Ltd., Worcestershire, UK).Regarding liquid hot water pretreated samples, the washed solids were kept at −80 °C for 48 h and then freeze dried (Labconco Ltd., Fort Scott, KS, USA).Corn stover powder was dispersed in deionized (DI) water to form a liquid suspension before being added into the instrument.D10, D50 (mean particle size), and D90 were determined according to particle size distribution curves, which separately represent the 10th, 50th, and 90th percentiles of the total volume.Measurements were performed in duplicate.

Enzymatic Hydrolysis
Untreated and treated solids were hydrolyzed with CellicCtec2 enzyme (Novozymes, Lyngby, Denmark), which had a protein content of 150 mg/mL and 90 FPU/mL as measured using the NREL procedure [22].Enzymatic hydrolyses were conducted as previously described [23] in 50 mM sodium citrate buffer (pH 4.8) with 0.3 mM soldium azide to prohibit microbial growth.Each enzymatic hydrolysis used an enzyme loading of 20 FPU/g dry corn stover (33.33 mg protein/g) in 50 mL centrifuge tubes with 5% (w/v) biomass solids concentration at 50 °C with a 200 rpm rotation speed.Samples were taken for analysis at 2, 4, 8, 24, 48, and 72 h.The experiments were performed in duplicate.
Based on economic considerations and comprehensive comparisons of combining pretreatments to a single pretreatment, low-enzyme-loading enzyme hydrolysis experiments were also performed on all samples with a CellicCtec2 loading of 3 FPU/g (5 mg protein/g) for 72 h.

Sugar Yields
Sugar concentrations in acid and enzymatic hydrolysates were analyzed via HPLC (Waters, Milford, MA, USA) equipped with an HPX-87H column (Bio-Rad, Hercules, CA, USA).The mobile phase was 5 mM H2SO4 at a flow rate of 0.6 mL/min.The column temperature was kept at 60 °C.The sugar yields were determined as follows:

Statistical Analysis
Reported results are shown as the replicate mean ± standard deviation with Excel 2019 (Microsoft, Redmond, WA, USA).Pearson's methods for correlation analysis were performed using SPSS 20.0 software.

Chemical Compositions of All Corn Stover Samples
The composition of the raw corn stover (on a dry weight basis) was 11.8% water extractives, 2.3% ethanol extractives, 33.2% glucan, 19.1% xylan, 0.8% arabinan, 11.8% lignin, 18.5% ash, and 3.12 others, which is comparable to other studies [24].The glucan, xylan, and total lignin are summarized in Table 2.The high ash content might be caused by soil and dust bonding to the surface of corn stover which was collected from the ground after the grain was mechanically harvested [25].Chemical composition and remaining solid fractions of liquid hot water of pretreated and combined pretreated samples are shown in Table 2.The remaining solid fraction decreased as the severity factor increased for liquid hot water pretreatment, corresponding to an increased dissolution of the biomass into a solution as the severity of the pretreatment increased.At the lowest severity factor tested (log R0 = 3.24, T = 170 °C), the glucan, xylan, and lignin content have all significantly increased due to the removal of extractive components.While at the highest severity factor (log R0 = 4.41, T = 210 °C), the xylan has been completely removed from the solids.The analysis of the liquid fraction after the LHW pretreatment (Figure 1) indicates that pretreatment severity strongly affects hemicellulose removal and subsequent degradation of xylose.At the lowest severity (log R0 = 3.24, LHW170), almost 100% of the xylan remains in the solids; only 6.5% of the xylan was hydrolyzed into water-soluble components.For LHW210 (log R0 = 4.41), no detectable xylan remained in the solids while the soluble components, including furfural, accounted for less than 36% of the initial xylan, suggesting that most of the xylan completely degraded into humins [26].Therefore, in this study, 190 °C was selected for the combining pretreatment with ball milling based on the fact that more than 85% of the xylan was preserved as soluble xylo-oligosaccharides and xylose.Xylan and lignin content of the solids after both liquid hot water treatment and ball milling was similar to LHW190 (Table 2).However, the glucan content of dry solid decreased with the increase of ball milling time, suggesting that part of the cellulose degraded into soluble sugar in the liquor.This was confirmed by a measured increase in cello-oligosaccharides in the hydrolysate (Table S1).This is consistent with the previous literature, which also found that increased ball milling time would enhance the soluble sugar in the water extractives [27].The xylose recovery (86-91%) of combined pretreatment samples was also consistent with the single-liquid hot water pretreatment at 190 °C.

Particle Size Distribution
Particle size distribution of single ball milling, liquid hot water pretreated, and combined pretreated samples are shown in Figure 2; D10, D50, and D90 values are shown Table 3. Ball milling significantly reduced the particle size of corn stover.The peak of particle size distribution curve shifted to the left with an increase in ball milling time, resulting in a mean particle size from 568 µm to 37 µm.The curves of single ball milling samples showed bimodal distributions, likely due to the different mechanical properties for various parts of corn stover.Most of the larger fractions were likely derived from the stem, while finer fractions were more likely to be derived from the leaf [28,29].Liquid hot water pretreatment alone also slightly reduced the particle size.The particle size reduced with an increase in the severity factor, and the smallest mean particle size of 383 µm was found in LHW210.This may be because solubilization of the hemicellulose and other cell wall components during the liquid hot water pretreatment resulted in tissue separation, causing particle size reduction [30].One narrow peak also indicated that the particle size was evenly reduced for all tissue types during the liquid hot water pretreatment.

Enzymatic Hydrolysis
Ball milling and LHW pretreatment both individually enhanced the enzymatic hydrolysis efficiency (Figure 3).Although ball milling could significantly reduce the particle size of corn stover, the increase in glucose yields was limited.Ball milling for 60 min only improved glucose yield to 47% compared to 32% for the untreated corn stover.Several studies have shown that particle size reduction is not the main factor in increasing digestibility [11,31].Our previous work also demonstrated that cellulose crystallinity and destruction of the cell wall structure were the dominant factors for improving glucose yield [23].It may be inferred that 60 min ball milling might not be long enough to totally deconstruct the cell wall structure for corn stover, and that glucose yield would continue to increase if the ball milling time was extended.However, the energy cost will also increase with the extension of milling time [32].
The enhancement of cellulose conversion to glucose (46%) was limited for LHW pretreatment at a low severity factor of 3.24 (T = 170 °C) due to low xylan solubilization (Figure 1).When the severity factor increased to 3.83 (T = 190 °C), the glucose yield (62%) showed a significant improvement and almost all the cellulose was hydrolyzed to glucose at a severity factor of 4.41 (T = 210 °C).Many studies that focused on LHW pretreatment reported similar results [15,33].
All three combined pretreatments showed higher glucose yields (72~90%) than the liquid hot water pretreatment alone.It is likely that the tissues and physical structures were destroyed with ball milling, making corn stover more accessible to the liquid hot water at a high temperature.In other words, the ball milling process enhanced the effects of the following liquid hot water pretreatment, resulting in cellulose being more accessible to enzymes [34,35].It was interesting to note that ball milling could significantly improve the initial enzymatic hydrolysis rate (0-4 h).For example, BM60LHW190 and BM30LHW190 showed an enzymatic hydrolysis rate of 97 mg/h and 50 mg/h, respectively, during the initial enzymatic hydrolysis stage.In contrast, although LHW210 had a higher glucose yield than the combined pretreatment samples, the initial hydrolysis rate was only 34 mg/h.Wang et al. [35] reported glucose yields of 60.26% at an enzyme loading of 35 FPU/g of substrate by using a combination of a biological pretreatment with a liquid hot water pretreatment.Kim et al. [18] reported the highest observed glucose yield of 87.3% by combining 1% dilute acid and disk milling for a corn stover.The effect of lowering the enzyme loading to 72 h cellulose hydrolysis times is shown in Figure 4.The trend of the glucose yield for different samples at a low enzyme loading was consistent with the results at a high enzyme loading, but with lower overall yields.The highest glucose yield was only 59% for LHW210.The fact that the glucose yield for BM60LHW190 (52%) at an enzyme loading of 3 FPU/g was higher than BM60 (47%) at an enzyme loading of 20 FPU/g indicated that the combined pretreatments could save the cost for producing enzymes compared to the short-duration ball milling pretreatment.This may be due to less lignin redistribution and deposition occurring on cellulose surfaces, which occur at higher pretreatment severities under acidic conditions, including liquid hot water pretreatments [15,36,37].The 72 h xylose hydrolysis yield showed a similar result in Figure 5.During the single ball milling treatment, it was observed that as the milling time increased from 20 min to 60 min, the xylose yield increased to 35% when a high enzyme loading was used.However, when the enzyme loading decreased to 3 FPU/g dry solid, the xylose yield decreased to 21%.Additionally, the liquid hot water pretreatment at a low severity also enhanced the xylose yield.However, at a high severity, the liquid hot water pretreatment resulted in the destruction of xylan components in corn stover, leading to a zero xylose yield.The highest xylose yield was achieved in the combined pretreatment samples.

Total Sugar Yield
The efficiency of the pretreatment for ethanol production was determined by the total yield of sugars recovered from both glucan and xylan.Therefore, in this study, total sugar yield including total sugars (monosaccharides plus oligosaccharides) in the pretreated liquor and enzymatic hydrolysis of pretreated solids was calculated in relation to the raw corn stover (Table 4).Ball milling for 60 min without liquid hot water pretreatment only increased the total sugar yield from 28.6% to 42.7% at a high enzyme loading and from 18.8% to 30.3% at a low enzyme loading.Although LHW210 had the highest glucose yield, most of the xylan degraded into by-products, leading to a lower total sugar yield than the combined mid-severity LHW pretreatment with ball milling.In addition, by-products, such as furfural, that are known to be strong fermentation inhibitors were eliminated [38].BM30LHW190 had the highest total glucose yield of 79%, achieving both a high glucose yield (83%) and xylose yield (72%) at an enzyme loading of 20 FPU/g glucan.The combined pretreatment samples also showed both a higher glucose and xylose yield than the single liquid hot water pretreatment in 190 °C.
Ball milling could significantly reduce the particle size of corn stover; however, the total sugar yield showed a low correlation (r = −0.04)with the mean particle size.Therefore, it could be inferred that the first stage of ball milling loosened the microstructure of the cell wall and made hemicellulose more accessible for the removal of the liquid hot water pretreatment in the second stage, resulting in an increase in both glucose and xylose yield.
The total sugar yield (42.9%) of BM20LHW190 at 3 FPU/g was higher than the one observed for BM60 with 20 FPU/g (42.7%), indicating that combining ball milling for 20 min with the liquid hot water pretreatment at 190 °C could save 67% of the ball milling time and 75% of the energy cost based on the comparable total sugar yield.As the energy cost was linear with regard to the ball milling time, it could be concluded that combining the ball milling pretreatment and the liquid hot water pretreatment could save both energy and the cost of producing enzymes [32,33].This order of pretreatment (ball milling followed by LHW) also saves energy for drying if the order is reversed, e.g., hot compressed water followed by a ball milling pretreatment, as reported previously [32].
The total sugar yield (42.9%) of BM20LHW190 at 3 FPU/g was higher than the one observed for BM60 with 20 FPU/g (42.7%), indicating that combining ball milling for 20 min with the liquid hot water pretreatment at 190 °C could save 67% of the ball milling time and 85% of the enzyme cost based on the comparable total sugar yield.Our previous study demonstrated that, as the ball milling time increased from 30 min to 60 min, the energy requirement during the mechanical fragmentation process doubled [39].It can be inferred that the combined pretreatment of ball milling and liquid hot water has the potential to reduce energy costs compared to the ball milling pretreatment alone.However, detailed energy calculations need to be carried out in the future to confirm this hypothesis.

Conclusions
Liquid hot water and ball milling pretreatments are both environmentally friendly and do not contain chemicals.Ball milling for 30 min followed by liquid hot water at 190 °C for 15 min had the highest total glucose yield of 79%, achieving both a high glucose yield (83%) and xylose yield (72%) without the production of any detectible xylose degradation by-products (furfural).This combined pretreatment also could save ball milling energy and enzyme loading based on the comparable total sugar yield.It could be concluded that the destruction of tissues and cell wall structures with ball milling in the first stage intensified the accessibility of cellulose to enzymes following the LHW pretreatment, resulting in the improvement in both the glucose and xylose yield by allowing for the lower severity pretreatment to achieve high sugar yields.This approach offers a promising solution for improving the overall efficiency of lignocellulosic biomass conversion processes and supports the sustainable utilization of biomass resources.

Figure 1 .
Figure 1.Xylan recovery for combining and single-liquid hot water pretreatment samples.These data were calculated based on chemical composition of dry solid and pretreated liquor.Xylan recovery included xylan in solid particles, humins (determined by difference subtraction), xylo-oligosaccharides in the pretreated liquor, and xylose in the pretreated liquor and furfural.The numbers shown on the graph are the percentage of total water-soluble components on the basis of the initial biomass.

Figure 2 .
Figure 2. Particle size distribution for all samples.The data were determined with a laser diffraction particle size analyzer.(a) Particle size distribution for ball-milled samples.(b) Particle size distribution for liquid hot water pretreated samples.(c) Particle size distribution for combined pretreated samples.Ball milling pretreatment could significantly reduce the particle size of corn stover.Liquid hot water pretreatment slightly reduces the particle size of corn stover.

Figure 3 .
Figure 3. Glucose yield of enzymatic hydrolysis for all samples.Untreated and pretreated corn stovers (dry basis) were hydrolyzed with a CellicCtec2 loading of 20 FPU/g dry corn stover (33.33 mg protein/g) for 2, 4, 8, 24, 48, and 72 h at a 5% (w/v) concentration.Error bars represent the standard deviation of glucose yield.

Figure 4 .
Figure 4. Glucose yield (72 h) of enzymatic hydrolysis for all samples.The data show the cellulose conversion to glucose at 72 h at both high enzyme loading (20 FPU/g dry solid) and low enzyme loading (3 FPU/g dry solid).

Figure 5 .
Figure 5. Xlycose yield (72 h) of enzymatic hydrolysis for all samples.The data show the xylan conversion to xylose at 72 h at both high enzyme loading (20 FPU/g dry solid) and low enzyme loading (3 FPU/g dry solid).

Table 1 .
Pretreatment conditions of corn stover.
N/A means Not applicable.

Table 2 .
Partial chemical composition corn stover before and after pretreatment.
Chemical compositions were calculated based on pretreated corn stover; data are shown as the replicate mean ± standard deviation.N/A means Not applicable.

Table 3 .
Particle size distribution of all samples.

Table 4 .
Total sugar yield for all samples.