Introduction

Corn fiber is an abundant and low-value byproduct of corn processing (wet milling, dry grind, dry milling, and modified dry grind), with an annual production of 4–5 million tons for wet milling alone at a price of about $60 per US ton (Raiagopalan et al. 2005). The utilization of corn fibers is crucial to the profitability and sustainability of a commercial corn processing operation.

Destarched corn fiber contains about 70% arabinoxylan (Doner and Hicks 1997) and is hence a good source for the production of xylooligosaccharides and other soluble dietary fibers. Soluble dietary fibers help to suppress blood cholesterol levels, especially low density lipoprotein-cholesterol, by binding to bile acids, which contain cholesterol, in the gastrointestinal tract and carrying them out of the body as waste (Woo and Kim 2005).

Strong alkalines are used to extract hemicellulose, mainly composed of arabinoxylan, from corn fibers for possible uses as corn fiber gum, non-toxic adhesives, thickeners, emulsifiers, stabilizers, film formers and paper additives (Doner and Hicks 1997; Hespell 1998). Conversion of carbohydrates in corn fibers to fermentable monomeric sugars, such as glucose and xylose, has also been carried out (Saha et al. 1998; Saha et al. 1999; Schell et al. 2004; Mosier et al. 2005). A few recent researches have explored means to produce xylooligosaccharides from biomass, including corn stover and corn cobs (Yang et al. 2006; Zhu et al. 2006). There has been no research effort dedicated to the production of xylooligosaccharides or soluble dietary fibers from corn fibers into forms for human consumption, such as nutraceuticals, functional foods, or functional food ingredients.

The objective of this research was to develop a corn fiber defractionation method for enriching arabinoxylan for the production of xylooligosaccharides or other soluble dietary fibers from the pentose portion in corn fibers.

Materials and methods

Destarching of corn fibers

Corn fibers containing up to 60% moisture (wet basis) were provided by Tate & Lyle (Decatur, IL). For improving storage stability and better handling (weighing and grinding), the fibers were dried at 105°C to a constant weight. They were then ground and sieved through a #20 mesh. Dry corn fibers, 20 g, were mixed with 80 ml sodium acetate buffer (50 mM, pH 4.5) and autoclaved at 121°C for 1 h. After cooling to 55°C, 2 ml G-ZYME 480 ethanol solution (Genencor International, Rochester, NY) was added and the mixture was shaken at 250 rpm for 16 h. After centrifugation at 6,900g for 15 min, the supernatant was discarded. The precipitate was washed with 100 ml ethanol and then centrifuged. After removing the supernatant, the precipitate was washed with ∼100 ml deionized H2O and centrifuged again at 6,900g for 15 min. Starch content in corn fiber was measured.

Pretreatment and enzymatic hydrolysis of destarched corn fiber

Destarched corn fibers were subjected to a second pretreatment under 9 different conditions (Table 1). The pretreatment was conducted with stainless steel tubular reactors (2.5 cm OD × 17.5 cm length) with a working volume of 50 ml. A SBL-2D fluidized sand bath with a TC-8D temperature controller was used to heat the reactors. Corn fiber slurry was loaded into the reactors that were submerged in the fluidized sand bath and heated to a target temperature (Table 1). Xylanase digestion of destarched corn fiber was performed in an Aquatherm water bath shaker at 40°C for 48 h. Enzeco Xylanase AN900 (900,000 BXU/gram) was provided by Enzyme Development Corporation (New York, NY), which also has glucoamylase and cellulase activity and can breakdown various complex carbohydrates. The enzyme was used at a ratio of 2 ml enzyme/100 g destarched corn fibers (dry wt).

Table 1 Experimental conditions for pretreatment and enzymatic hydrolysis of destarched corn fibers
Full size table

Scanning electron microscopy

The corn fiber samples before and after a pretreatment were examined with scanning electron micrographs to observe morphology changes.

Sugar analysis

The supernatant from samples taken during the enzymatic hydrolysis was analyzed to determine the concentration of monomeric sugars and oligomers. Glucose and xylose concentrations were analyzed by HPLC using a Bio-Rad HPX-87P column (30 × 4.6 mm) at 85°C and a refractive index detector. The mobile phase was ultrapure water (18.1 MΩ cm) at 0.6 ml/min. For oligomer analysis, the supernatant (0.4 ml) was mixed with 0.4 ml of 4 M trifluoroacetic acid (TFA) and held at 100°C for 1 h. The samples were cooled, neutralized with NaOH (0.4 ml of 4 M) and the xylose and glucose concentrations analyzed with HLPC. The concentration of the polymeric sugars was estimated from the concentration difference of the monomeric sugars in the supernatant before and after the TFA-treatment. An anhydro correction was used (0.88 for xylose, 0.9 for glucose) in sugar quantification (Theander and Westerlund 1986).

Results and discussion

Destarch of corn fibers

The major carbohydrate components in corn fibers are starch 18 (% w/w), arabinoxylan 32 (% w/w) (xylose, 20; arabinose, 10), and cellulose 24 (% w/w) (Leathers and Gupta 1996; Leathers et al. 2006). Large number of starch granules can be observed on the surfaces of the wet milled fibers (Fig. 1A). After the destarching process, the starch content was decreased to 0.5 ± 0.1% with almost no adherent starch granules on the fiber surfaces. During the destarch process, the recalcitrant fiber structure remained intact (Fig. 1B).

Fig. 1
figure 1

Scanning electron microscopy (SEM) images of corn fibers. (A) Wet milling corn fiber without pretreatment; (B) destarched corn fiber; and (C) destarched corn fiber after pretreatment at 170°C and enzymatic hydrolysis. Corn fiber samples were dehydrated in a graded series of ethanol solutions (25, 50, 70, 95 and 100%) and dried in a CO2 critical-point drier. The dried samples were then mounted on stubs and sputter-coated with gold/palladium for 80 s by a sputter coater before taking the SEM images

Full size image

Pretreatment and enzymatic hydrolysis of destarched corn fiber

The surface condition of fibers after the second pretreatment and enzymatic hydrolysis is shown in Fig. 1C. Compared to destarched samples (Fig. 1B), the recalcitrant fiber structure was totally destroyed by Enzeco Xylanase AN900. Typical yields of glucose, xylose, cellooligosaccharide (COS) and xylooligosaccharides are given in Fig. 2. The concentrations of glucose were higher than that of xylose and xylooligosaccharides in all cases. As destarched fibers were used as the substrate, all glucose in the hydrolysate must have arisen from cellulose. This indicates that the mild hot water pretreatment at 140–170°C followed by an enzymatic hydrolysis is effective in removing the hexose component in the fiber with the remaining solid residues rich in arabinoxylan.

Fig. 2
figure 2

Sugar yield during the enzymatic hydrolysis of hot water pretreated corn fibers. (A) Treatment at 160°C for 30 min; (B) treatment at 140°C for 30 min; and (C) treatment at 170°C for 30 min, all with a solid loading of 15% (w/v) in the enzymatic hydrolysis. Symbols: ■ Glucose; ▲ Xylose; ● Cellooligosaccharides; and ▼ Xylooligosaccharides

Full size image

Saha and Bothast (1999) evaluated several pretreatments (hot water, alkali, and dilute acid) and enzymatic saccharification procedures for conversion of starch, cellulose, and hemicellulose in corn fibers to fermentable sugars. They also found that a hot water pretreatment (121°C, 1 h) facilitated the enzymatic saccharification of starch and cellulose but not hemicellulose (Saha et al. 1999; Saha 2003). Our results confirmed their observations. Hot water pretreatments at <160°C made cellulose easier to be hydrolyzed than corn fiber hemicellulose, even though the enzyme (AN900) used has strong xylanse activity. The corn fiber hemicelluloses, mostly branched arabinoxylans, are highly cross-linked by diferulic bridges, which may be responsible for their high resistance to enzymatic hydrolysis (Saulnier et al. 1995; Saha 2003). The use of enzyme preparation AN900 may not be a best choice since it has a high xylanase activity. A further improvement of the two-step process tested in this study will be to combine the two hexose separation steps into one with a hot water pretreatment followed by a glucoamylase and cellulase simultaneous enzymatic hydrolysis.

In all the 9 treatments, the glucose concentration increased during the enzymatic hydrolysis process, while the concentrations of xylose and xylooligosaccharides remained almost unchanged. Generally, a decrease in cellooligosaccharides was accompanied by an increase in glucose. The increase in glucose, however, was more pronounced than the decrease in cellooligosaccharides (Fig. 2B and C). Under those conditions, the enzyme might have solubilized cellooligosaccharides into glucose faster than did for conversion of cellulose into cellooligosaccharide. High solid/liquid ratios during enzymatic hydrolysis produced more glucose. At the lowest solid/liquid ratio (10%) (treatment E), the overall sugar yield was the lowest among all treatments.

Based on the estimated composition of corn fiber (Leathers and Gupta 1996; Leathers et al. 2006), the remaining xylose and glucose after the TFA hydrolysis were calculated (Fig. 3A). Among all the experiments, treatment G (140°C, 30 min, solid loading 15%) had the highest pentose retention (>90%) in the solid residue, with 45% remaining hexose in the solid. Since the purpose of the two-step treatment is to separate the pentose from the corn fibers for value-added products and utilize the hexose for bioethanol production, C5/C6 separation factor, a ratio of pentose to hexose in the solids, was used as another criterion in the selection of best treatment, as shown in Fig. 3B. The treatment D (160°C, 30 min, solid loading 15%) had the highest C5/C6 separation factor and a high remaining pentose of about 87%, and was hence chosen as the best treatment under the conditions tested in this study.

Fig. 3
figure 3

Xylose and glucose remaining in the solid after a 48 h enzymatic hydrolysis. (A) Remaining percentage of xylose and glucose after pretreatment and enzymatic hydrolysis; (B) C5/C6 separation factor which was defined as the ratio of pentose to hexose in the solids. The xylose percentage was estimated using 20 g xylose/100 g dry corn fiber as 100% while the base for glucose was 42 g glucose/100 g dry corn fiber (Leathers and Gupta 1996)

Full size image

Conclusion

After destarching and a mild pretreatment by hot water, the cellulose in wet milling corn fibers is more susceptible to enzymatic digestion than arabinoxylan. With this method, most of the hexose components (the starch and cellulose portion) in wet milling fibers can be removed, while arabinoxylan is enriched. This finding suggests that a two-step fractionation process can be used to separate hexose components from corn fibers for biofuel production, with the remaining arabinoxylan-rich solid for the production of value-added products.