Sugar beet leaves (Beta vulgaris) from mature ECN115 variety beets (seeds provided by KWS Betaseed, Inc.) were collected from the research farm of the University of California, Davis. Some of the fresh leaves were randomly picked and processed for compositional analysis. The remaining leaves were stored in a sealed container at 4˚C for enzymatic hydrolysis immediately after determining compositional data. For compositional analysis, fresh leaves were chopped, ground using a food processor, and dried using a freeze drier (FreeZone Freeze Dry System, LABCONCO, MO). The dried sample was milled with a 40-mesh Wiley mill (Thomas Scientific, NJ). For enzymatic hydrolysis, the leaves were chopped and ground using a food processor. Fresh and dried samples were analysed for moisture (MC), total solids (TS), and volatile solids (VS) contents.

Extractive, structural carbohydrate, and lignin contents for sugar beet leaf samples were determined following NREL Chemical Analysis and Testing Standard Procedures [24] [25] . The extractives were determined by extracting the biomass with water and ethanol for eight and sixteen hours, respectively, using a Soxhlet apparatus. The supernatants were vacuum-filtered and collected for analyses of sugars and sugar acids. The filtered solids were freeze-dried, followed by acid hydrolysis with strong sulfuric acid (72% H₂SO₄, w/w) at 30˚C for one hour, and diluted sulfuric acid (4% H₂SO₄, w/w) at 121˚C for one hour in an autoclave. The autoclaved hydrolysis solution was vacuum filtered and analysed for lignin and structural carbohydrate contents.

The concentrations of sugars in supernatants were determined using a high performance liquid chromatography (HPLC) system (Shimadzu, Columbia, MD) equipped with an ion exchange Aminex HPX-87P column (Bio-Rad Laboratories, Hercules, CA), micro-guard column, and a refractive index (RI) detector (RID-10A, Shimadzu, Columbia, MD). The separation was run at 85˚C with a sample injection volume of 25 µL and 18 mQ NANOpure water as a mobile phase at a flow rate of 0.6 ml∙min⁻¹. The aqueous solutions, containing glucose, xylose, galactose, arabinose, mannose, fructose and sucrose at four concentration levels, were used to create standard curves to quantify the sugar content in the sample. Organic acid and alcohol concentrations in supernatants were determined using an Aminex HPX-87H column (Bio-Rad Laboratories; Hercules, CA) with the HPLC system (Shimadzu, Columbia, MD) equipped with micro- guard column, a refractive index (RI) and photodiode array (PDA) detectors (SPD-M20A, Shimadzu, Columbia, MD). The HPLC analysis was run at 60˚C using 5 mM H₂SO₄ in 18 mQ NANOpure water as a mobile phase at a flow rate of 0.6 ml∙min⁻¹ to elute the sample injected at a volume of 25 µL. The aqueous solutions, containing galacturonic acid, lactic acid, formic acid, acetic acid and ethanol at four concentration levels, were used to create standard curves to quantify the acid contents in the sample.

Cellic CTec2 (cellulase) and Cellic HTec2 (xylanase) enzymes were provided by Novozymes North America, Inc (Franklinton, NC). Pectinex Ultra SPL (P2611) (pectinase) was purchased from Sigma Aldrich, Inc. Representative cellulase, xylanase, and pectinase activities of each enzyme product (Cellic CTec2, Cellic HTec2 and Pectinex Ultra SPL) were determined using a method described by Ghose and Bisaria [26] [27] . The corresponding activity units were reported as a filter paper unit (FPU), a xylanase unit (XU) and a polygalacturonase unit (PGU), respectively.

Enzymatic hydrolysis of sugar beet leaves was performed 100-ml Erlenmeyer flasks at 10% w/v total solids loading by mixing ground biomass with sodium citrate buffer (pH 4.8) to obtain a final concentration of 0.05 M in total working mass of 50 g. Penicillin (20 mg∙l⁻¹) and sodium azide (0.2% w/v) were added to the mixture to prevent microbial contamination. Preliminary studies on the effect of antibiotic addition showed that penicillin and sodium azide are needed for the inhibition of microbial contamination and growth. Enzymes (CTec2, HTec2 and Pectinex Ultra SPL) were added to the mixture at concentrations according to the experimental design (Table 1). Hydrolyses were performed at 50˚C and the flasks shaken at 150 rpm for 72 h. Samples (2 ml) were taken every 24 h from each flask, diluted (5 times) with deionized water, and immediately centrifuged at 8000 rpm for 15 min at 4˚C (HeraeusMultifuge X1R centrifuge,

^aPolygalacturonic acid (PGA) is the major component of pectin.

VWR International, Brisbane, CA). After centrifugation, part of the supernatant was stored at −20˚C until analysed for nitrogen and protein. The remaining supernatant was immediately filtered through a 0.2-μm membrane filter and stored at −20˚C for later sugar and acid analyses using HPLC. As shown in Equation (1), sugar conversion was calculated as a percent of the theoretical yield, which was estimated from the total potential monomeric sugar present in the leaves. Total amount of sugar released was analysed after enzymatic hydrolysis, includes galacturonic acid, which is a sugar acid produced from the hydrolysis of pectin. Total sugar content in the biomass was estimated from the compositional results.

Total nitrogen content for the biomass and hydrolyzed samples, before and after three days of hydrolysis, were determined using the Persulfate Digestion Method [28] and a Hach DR 2700 spectrophotometer (Hach Company, Loveland, CO). The protein content was estimated from the multiplication of the nitrogen content by 6.25. Solubilized protein content was determined using the standard Bradford protein assay [29] . Bovine serum albumin (BSA) at four concentration levels was used as a standard for calibration. To exclude the protein from the enzyme addition, the protein content of enzyme controls was used to subtract the protein content from enzymatic hydrolysates.

Response surface methodology (RSM) was used to optimize the enzymatic hydrolysis of beet leaves at 10% TS loading using an enzyme mix of CTec2, HTec2 and Pectinex Ultra SPL. A three-factor rotatable central composite design (CCD) was used for determination of the effects and interactions of cellulase, xylanase and pectinase on the enzymatic hydrolysis. Enzyme loadings were the variable parameters in a range of 10 - 30 FPU g⁻¹ glucan, 500 - 1500 XU g⁻¹ xylan, and 50 - 150 PGU g⁻¹ polygalacturonic acid for cellulase, xylanase, and pectinase respectively. Sugar conversion after 24 h, 48 h and 72 h of hydrolysis were used as response variables. Each of the enzyme loadings was studied at five levels with six replicates at the center point. A total of 20 experiments were performed, as shown in Table 1. A quadratic polynomial equation was fitted for evaluation of the mathematical relationship between the variables and response. Analysis of variance (ANOVA) was used for the estimation of the statistical parameters, using JMP Pro 10 statistical software (SAS Institute Inc.; Cary, NC).

Solids analysis results for sugar beet leaves show that the percentages of TS, VS, and VS/TS are 14.20% ± 0.05% wb, 11.35% ± 0.06% wb, and 79.92% ± 0.06% db, respectively. The chemical compositional data for the beet leaves are shown in Table 2. On a dry basis, sugar beet leaves are mainly composed of structural carbohydrates (32.4%), protein (26.9%), and soluble sugars (10.0%). The structural carbohydrates are composed of polymeric sugars in nearly equal amounts of glucan (11.0%), hemicellulosic-derived sugars (xylan, galactan, and arabinan) (10.4%), and polygalacturonic acid (11.0%).

To use sugar beet leaves as a feedstock for biofuel production, it is important to access and utilize structural carbohydrates, as they are vital carbon sources. Enzymatic hydrolysis is crucial for depolymerizing structural components to soluble sugars, and for liquefying feedstock at high solids loadings to improve mixing and pumping.

As present in a considerable amount, galacturonic acid a pectin-derived sugar-acid is also a major carbon source. The fermentation of galacturonic acid can

increase ethanol yield significantly. Although wild type Saccaromycescerevisea cannot ferment these substrates to ethanol, many novel microorganisms have been researched and metabolically engineered for the ethanolic fermentation of pectin [30] - [35] . A novel metabolic pathway from Escherichia coli KO11 was developed to allow the fermentation of galacturonic acid to equal molar amount of ethanol and acetate [32] [33] . In addition, Clostridium species (C. phytofermentans [36] [37] and C. cellulyticum [38] ) and ethanologenic constructed bacteria in genus Erwinia (E. chrysanthemi EC16 and E. carotovora SR38 [34] ) showed potential for the direct hydrolysis and fermentation of pectin to ethanol. The use of these microorganisms would make the overall ethanol production from pectin-rich biomass, such as sugar beet leaves, more economically feasible.

Results from the enzyme activity assays conducted are shown in Table 3. Filter paper, xylanase, and polygalacturonase assays were conducted for the representative measurement of these enzyme activities, acknowledging that these are not a full description of all enzyme activities that may be present. The primary specific activities of Cellic CTec2 (cellulase), Cellic HTec2 (xylanase) and Pectinex Ultra SPL (pectinase) are 125 FPU ml⁻¹, 9685 XU ml⁻¹, and 9848 PGU ml⁻¹, respectively. Most commercial enzyme products are a mixture of various enzyme types, and thus they can catalyze several reactions. CTec2 was found to exhibit relatively high xylanase activity, with 4261 XU ml⁻¹. However, for the experimental design, enzyme loading is determined based on the primary specific activity of each enzyme.

The enzymes used in this experiment have been studied and reported on in other research works [39] [40] [41] [42] . The enzyme activities determined in this experiment are in the same range as those reported previously, as shown in Table 4. The xylanase activities for Cellic HTec that can be found in previous studies range from 1090 to 10,596 XU ml⁻¹ [40] [41] [42] . Such a range might be attributable to the sensitivity of the activity measurement (due to the lack of linearity between enzyme concentration and product liberation) and the accuracy of the measurement since the measurement of xylanase activity is complex [26] .

Soluble proteins released from the biomass after 72 h of enzymatic hydrolysis ranged from 59% to 68% (an average of 62.8% ± 3.0%) of total protein, comparing to 52% of total protein released from hydrolysis without enzyme addition. To exclude the protein from the enzyme addition, the protein content in the hydrolysate was subtracted by those in the enzyme controls. From the selected range of enzyme loadings, the statistical analysis (significance of regression models and ANOVA) shows that the type of enzymes, enzyme loadings, and the interaction and polynomial terms are not significant (p < 0.05) in relation to solubilized protein. The protein remaining in the biomass after 72 h of hydrolysis is approximately 37.2% ± 3.0% of total protein. The insoluble protein could be potentially separated from the hydrolysate and used as nitrogen supplement or in protein meals.

The high protein content of sugar beet leaves would provide sufficient nitrogen for microbial requirement during a sequential fermentation process [46] . In addition to carbon sources for biofuel conversion, nutrients are essential for microbial growth and maintenance during the fermentation. Nitrogen is a major cellular component of microorganisms. The amount of nitrogen required for biomass synthesis can be estimated from the empirical formula for a bacterial cell as C₅H₇O₂N [47] . Previous studies on the effect of nitrogen on ethanol fermentation have shown the adverse effects of nitrogen deficiency [48] [49] [50] . The lack of nitrogen diminishes yeast's metabolic activity, sugar uptake rate, and the cell biomass during fermentation.

Excess nitrogen from sugar beet leaves after fermentation can be recovered as amino acid supplements, nitrogen source, or fertilizer. Nitrogen fertilizer can be recovered from biomass residues after biofuel production, potentially through anaerobic digestion. Recovering nitrogen as fertilizer helps reduce large nitrogen consumption during plant biomass cultivation, and thus improves soil quality and the cost for fertilizer usage. The co-production of fertilizer products also helps improve the overall economic of the biofuel production system.