Evaluation of Waxy Grain Sorghum for Ethanol Production

Cereal Chern. 88(6):589-595 The objective of this research was to investigate the fermentation performance of waxy grain sorghum for ethanol production. Twenty-five waxy grain sorghum varieties were evaluated with a laboratory dry-grind procedure. Total starch and amylose contents were measured following colorimetric procedures. Total starch and amylose contents ranged from 65.4 to 76.3% and from 5.5 to 7.3%, respectively. Fermentation efficiencies were in the range of 86.0-92.2%, corresponding to ethanol yields of 2.61-3.03 gallonslbushel. The advantages of using waxy sorghums for Unlike wheat, com, and rice, grain sorghum is a starch-rich cereal that can be grown economically in the semiarid regions of the world. In the United States, sorghum is the second-ranking feed grain and is cultivated primarily in the Great Plains, including the Midwest and the Southwest. Although it is primarily used as feed in the United States, grain sorghum has been reported in wide uses such as wallboard, fermented beverages, traditional foods (porridges and flat breads), and conventional pan bread for gluten-free markets (Owuama 1997; Rooney and SernaSaldivar 2000; Schober et al 2005; Taylor et al 2006). Sorghum utilization by the ethanol industry has been growing in the United States in recent years (RFA 2007; Sarath et al 2008). Currently, about 95% of U.S. fuel ethanol is produced from com and .,,4% is from sorghum grain, which uses 30-35% of the total sorghum production in the United States (Kubecka 2011; USDA-NASS 2011). Sorghum could make a larger contribution to the nation's fuel ethanol requirements (Farrell et al 2006; Rooney et al 2007; Wu et al 2007). Overall, sorghum composition is similar to com. Starch is the major grain component, followed by protein. Most sorghum starches contain 20-30% amylose and 70-80% amylopectin, but waxy and heterowaxy sorghums contain 0-15% amylose and 85100% amylopectin (Rooney and Serna-Saldivar 2000). Starch content in grains is a good predictor of ethanol yield (Lacerenza et al 2008; Zhao et al 2009). The presence or absence of amylose may influence ethanol yield and conversion efficiency. Wu et al (2007) reported that low amylose content in sorghum grain may be associated with increased ethanol conversion efficiency. One of *The e-Xtra logo stands for "electronic extra" and indicates that Figure I appears in color online. I Contribution number 11-327-J from the Kansas Agricultural Experiment Station, Manhattan, KS 66506. 2 Department of Biological and Agricultural Engineering, Kansas State University, Manhattan, KS 66506. 3 Present address: C. W. Brabender Instrument, Inc., 50 E. Wesley Street, S. Hackensack, NJ 07606. 4U.S. Department of Agriculture, Agricultural Research Service (USDA-ARS), Center for Grain and Animal Health, Manhattan, KS 66502. 5 USDA-ARS, Grain, Forage, and Bioenergy Research Unit, Lincoln, NE 68583. Names are necessary to report factually on available data; however, USDA neither guarantees nor warrants the standard of the product, and the use of the name by the USDA implies no approval of the product to the exclusion of others that may also be suitable. 6 Department of Agronomy, Kansas State University, Manhattan, KS 66506. 7 Corresponding author. Phone: (785) 532-2919. Fax: (785) 532-5825. E-mail: [email protected] http://dx.doi.org/10.1094 / CCHEM-04-11-0056 © 2011 AACC International, Inc. ethanol production include easier gelatinization and low viscosity during liquefaction, higher starch and protein digestibility, higher free amino nitrogen (FAN) content, and shorter fermentation times. The results showed a strong linear relationship between FAN content and fermentation rate. Fermentation rate increased as FAN content increased, especially during the first 30 hr of fermentation (R = 0.90). Total starch content in distillers dried grains with solubles (DDGS) was less than 1 % for all waxy varieties. the aims for this study, which was conducted on 25 varieties of waxy grain sorghum, was to investigate further whether ethanol yield and fermentation efficiency were influenced by the contents of amylose and amylopectin in waxy grain sorghums. Both ethanol yield and fermentation efficiency have been studied to evaluate the performance of grain sorghum in ethanol production (Wu et al 2007). Recent research has shown that key factors affecting the ethanol yield from grain sorghum include grain hardness and particle size, starch content and digestibility, amount and types of phenolic compounds present in sorghum, amount of amylose, formation of amylose-lipid complexes during mashing (Wu et al 2007; Wang et al 2008; Yan et al 2009), level of extractable proteins, protein-starch interaction, and mash viscosity (Zhao et al 2008). Sorghum as a raw material can be converted to ethanol with a wide range of efficiency (Wu et al 2007). Currently, almost 100% of industrial ethanol is produced by yeast from starch-rich or sugar-rich biomass. The availability of yeast food is vital to yeast growth and its efficiency in converting fermentable sugars into ethanol during fermentation. As such, most yeast fermentation systems need nutrient supplementation. Yeast uptakes not only fermentable sugars for ethanol production but also nutrients (amino acids, minerals, and vitamins) for its own growth and functional maintenance (e.g., levels of invertase and permeases, which are responsible for sugar transportation and conversion). Free amino nitrogen (FAN) is an essential nutrient for yeast growth during fermentation (Pickerell 1986; Taylor and Boyd 1986). Protein is the second major component in grain sorghum. Protein degradation could provide nitrogen for yeast growth during fermentation. Recent research has found that ethanol yield and conversion efficiency significantly increased as FAN increased in laboratory-germinated and field-sprouted grain sorghum (Yan et al 2009, 2010). Yeast can only utilize FAN and short peptides, not large intact proteins. Much research has been conducted on the effects of protein and protein digestibility on ethanol fermentation of various cereal grains such as wheat and barley (Lacerenza et al 2008), sorghum and maize (Perez-Carrillo and Serna-Saldivar 2007; Perez-Carrillo et al 2008; Zhao et al 2008), and different varieties of maize (Wu 1989; Wang et al 2005), but little research has been conducted on the effect of FAN on the conversion efficiency of sorghum varieties in ethanol fermentation. Sorghum is a large, variable genus with many cultivars. Many varieties of sorghum exist, and more are being developed through plant breeding to select and concentrate desired characteristics in new varieties for food and feed applications (Rooney and SemaSaldivar 2000; Mace and Jordan 2010). We believe genetically improving the quality of grain sorghum for ethanol production could increase the utilization of sorghum for ethanol production Vol. 88, No.6, 2011 589 in the near future. The main objective of this research was to investigate the fermentation performance of waxy grain sorghum for ethanol production. MATERIALS AND METHODS Grain Sorghum Twenty-five waxy grain sorghum varieties were obtained from the U.S. Department of Agriculture, Agricultural Research Service (USDA-ARS), Grain, Forage, and Bioenergy Research Unit (Lincoln, NE). The origin of these waxy varieties was from 10 different countries around the world, and the seeds of these accessions were increased by the USDA-ARS in Nebraska (Pedersen et al 2007). Detailed sample information is listed in Table I. The samples were manually cleaned by removing plant debris and foreign materials and then were ground to flour with a UDY cyclone sample mill (Fort Collins, CO) with a 1.0 mm screen. Preparation of Mashes and Inoculation of Yeast Liquozyme SC DC, a heat-stable a-amylase from Bacillus licheniformis was used for liquefaction (Novozyme, Franklinton, NC). The listed enzyme activity was 240 KNU/g (one kilo Novo unit, or KNU, is the amount of enzyme that breaks down 5.26 g of starch per hour at Novozyme's standard method for determination of a-amylase). Spirizyme Fuel (Novozyme), an amyloglucosidase from Aspergillus niger, was used for saccharification. Its listed enzyme activity was 750 AGU/g (one AGU is the amount of enzyme that hydrolyzes 1 J.lIllol of maltose per minute under specified conditions). Ethanol red active dry yeast (Saccharomyces cerevisiae) from Fermentis (Milwaukee, WI) was used for simultaneous saccharification and fermentation (SSF). Before inoculation, dry yeast (:»2 x 1010 live cells per gram) was activated by adding 1.0 g of dry yeast cells into 19 mL of preculture broth (containing 20 g of glucose, 5.0 g of peptone, 3.0 g of yeast extracts, l.0 g of KH2P04, and 0.5 g of MgS04'H20 per liter) and shaking at 200 rpm in a 38°C incubator for 30 min. The activated yeast culture had a cell concentration of roughly 1 x 109 cells/mL. Thirty grams (db) of sorghum flour for each sample was dispersed in 100 mL of water (containing 0.1 g of KH2P04 and preheated to about 60°C) in a 250 mL Erlenmeyer flask. Twenty microliters of high-temperature a-amylase (Liquozyme, 240 KNU/g) was added into the sorghum flour slurry. The flasks were transferred to a 70°C water-bath shaker operating at 170 rpm. The water-bath temperature was gradually increased from 70°C to 85°C over a 30 min period. The liquefaction process continued at 85°C for another 60 min. The flasks were then removed from the water-bath shaker and cooled to room temperature. Materials sticking to the inner surface of each flask were scraped back into the mash with a spatula, and then the inner surface was rinsed with 2-3 mL of distilled water with a fine-tipped polyethylene transfer pipette. The pH of the mashes was adjusted to 4.2-4.3 with 2N HCI. After pH adjustment, 100 ilL of amyloglucosidase (Spirizyme Fuel), 1 mL of activated yeast broth, and 0.3 g of yeast extract (1 mL of freshly prepared 30% yeast extract solution) were added to each flask. The inoculated flasks were then sealed with S-shaped airlocks and transferred to an incubator shaker for SSE All samples were run in duplicate. Fermentation and Distillation Ethanol fermentation was conducted at 30°C in an incubator shaker (12400, New Brunswick Scientific, Edison, NJ) operating at 150 rpm for 72 hr. The fermentation process was monitored by measuring the weight loss from evolution of carbon dioxide (C02) during fermentation. The weight loss was related to ethanol yield during fermentation (C6H120 6 ~ 2C2H60 + 2C02!). The ratio of ethanol to carbon dioxide is theoretically 46:44. After 72 hr of fermentation, finished mash in each 250 mL flask was entirely transferred to a 500 mL distillation flask, and the Erlenmeyer flask was washed four times with 100 mL (25 mL x 4) of distilled water. Two drops of antifoam agent 204 were added into the distillation mash to prevent foaming during distillation. The contents were distilled in a distillation unit, and the distillates were collected into a 100 mL volumetric flask that was immersed in ice water. When the distillates in the volumetric flask approaching the 100 mL mark «0.5 mL to the mark), the volumetric flask was removed from the distillation unit and the distillation process was stopped. The distillates in the volumetric flask were equilibrated for a few hours in a 25°C water bath and TABLE I Sample Information, Chemical Composition (%, db), and Fermentation Efficiency (%) of Waxy Sorghums' Accession No. Local Name Origin GBSS Allele Amylose Starch Protein Fat Fiber Ash Efficiency Tannin PI220636 Nai-Shaker Afghanistan No wx" 6.6 ± 0.57 66.80 14.46 6.56 1.56 2.64 88.4 + PI2323 I Brown Kaoliang China Yes wxb 6.8 ± 0.38 67.46 13.75 5.02 2.22 2.01 89.4 + PI548008 Huang Ke Jiao China No wx" 5.5 ± 0.70 68.74 15.09 4.98 1.52 2.54 87.7 + PI563576 LV 129 China No wx" 6.9 ± 0.87 68.38 15.80 5.70 1.59 2.38 89.8 PI563670 L 1999B-17 China Yes wxb 6.2 ± 0.20 76.34 11.22 3.78 1.75 1.85 88.3 PI56367I L 1999B-18 China Yes wxb 5.8 ± 0.90 72.41 12.40 3.60 1.98 1.94 90.3 PI586524 IS 27929 China No wx" 6.2 ± 0.35 69.88 13.53 5.21 1.86 2.16 89.6 + PI586526 IS 27931 China No wx" 7.0 ± 0.15 69.48 12.40 5.37 1.75 1.89 90.6 + PI586529 IS 27935 China No wx" 6.7 ± 0.45 66.79 14.29 5.69 1.55 1.72 92.2 + PI455543 ETS 3634 Ethiopia No wx" 6.6 ± 0.15 70.76 13.60 5.35 1.53 2.01 89.4 PI586448 Cody Hungary No wx" 5.8 ± 0.64 75.19 12.02 4.81 1.77 1.96 89.9 PI586454 Leoti Hungary No wx" 6.8± 0.72 67.71 14.17 5.11 1.80 2.12 89.1 + PI217897 305 Indonesia Yes wxb 5.9 ± 0.20 68.94 12.02 5.18 1.61 1.71 89.9 + PI234456 Unknown Japan No wx" 6.4± 1.12 71.30 12.15 4.94 1.58 1.75 90.2 + PI82340 Kaoliang-WX Korea No wx" 6.5 ±0.30 72.30 12.34 4.23 1.89 2.09 88.3 + PI87355 Bomususu Korea No wx" 6.1 ±0.40 69.71 14.40 5.17 1.88 2.14 89.9 + PI88004 Susu zairai shu Korea No wx" 6.1 ± 0.59 69.23 13.68 5.04 1.63 2.16 88.8 + PI563015 Kaura Mai Faran Kona Nigeria No wx" 6.6 ±0.51 65.36 15.46 5.64 1.75 1.93 88.4 + PI567803 Yungju South Korea No wx" 7.0±0.35 67.19 13.51 5.46 1.57 2.06 91.7 + PI567809 Unknown South Korea No wx" 6.6 ± 0.38 67.55 14.29 5.27 1.51 1.82 90.9 + PI5678 I I Unknown South Korea No wx" 7.3 ± 0.20 66.80 13.43 5.09 1.70 1.97 91.3 + PI562758 Basuto Red Q2-1-29 USA No wx" 6.2 ± 1.02 72.18 16.75 4.28 2.03 2.28 86.0 PI563068 IS 8303 USA No wx" 6.2 ± 0.83 71.53 14.30 3.00 2.25 2.06 88.5 PI563402 IS 10497 USA No wx" 6.1 ±O.IO 69.94 15.04 3.78 1.94 2.49 89.6 Ellis USA, wild No Wx 6.3 ± 0.10 72.86 13.28 4.46 1.91 1.71 90.5 a Protein contents were calculated by 6.25 x N contents from the Leco method (AOAC method 990.03). GBSS = granule-bound starch synthase; wx" and wxb are different forms of the allele. 590 CEREAL CHEMISTRY then brought to the 100 mL mark with distilled water. Ethanol concentrations in the distillates were analyzed by HPLC with a Rezex RCM column (Phenomenex, Torrance, CA) and refractive index detector (Wu et al 2006). Morphological Structure of Waxy Grain Sorghum The microstructures of waxy sorghum kernels were examined with a scanning electron microscope (SEM) with an accelerating voltage of 5.0 kV (S-3500N, Hitachi Science Systems, Tokyo). Samples were vacuum coated with a mixture of 60% gold and 40% palladium particles with a Desk II combined sputter coater and etch unit (Denton Vacuum, Moorestown, NJ). Single Kernel Characterization and Particle-Size Analysis Kernel hardness, weight, and size of waxy sorghum samples were analyzed through the single kernel characterization system (SKCS) 4100 (Perten Instruments, Springfield, IL) controlled by Microsoft Windows software SK4100. The reported data were the means of 300 kernels. The particle size of ground sorghum flour was measured by an LS 13 320 single wavelength laser-diffraction particle-size analyzer (PSA) with Tornado dry powder system (Beckman Coulter, Miami, FL). Samples were run in duplicate. Pasting Properties Pasting properties of the sorghum flour samples were measured with a Rapid Visco Analyzer (RVA) (RVA-3D, Newport Scientific, Warriewood, Australia). For sample preparation, 4 g of sorghum flour (14% moisture basis) and distilled water (25 mL) were added to an aluminum canister at room temperature. A plastic paddle was inserted into the canister and then jogged and rotated manually for about 30 sec to break up any lumps. The paddle (with the sample canister) then was attached to the electric motor in the head of the RVA. The sample was premixed by initially running the motor at 960 rpm for 10 sec, and then the motor was slowed to 160 rpm for the rest of the test. The standard 23 min profile of AACC International Approved Method 76-21.01 (2010) was followed for sample testing. Each sample was analyzed in duplicate. Thermal Properties Differential scanning calorimetry (DSC) (Pyris I, Perkin-Elmer, Norwalk, CT) measurement was conducted and calibrated with indium. Sorghum samples were weighed accurately (~10 mg) into stainless steel pans with a microbalance. Deionized distilled water was added carefully with a micropipette into the sample pan. The weight ratio of water to dry flour was 3: I. The pans were sealed and allowed to rest for about 1 hr. Samples were analyzed at heating and cooling rates of 10°C/min. The temperature regime consisted of heating from 25 to 150°C with an initial I min hold. Data from the DSC scans were analyzed with Pyris software 7.0 for Windows (Perkin-Elmer). Enthalpies are reported on a dry flour weight basis. Each sample was analyzed at least in duplicate. Protein Digestibility Protein digestibility was determined following the method of Mertz et al (1984) with modification: 200 mg sorghum samples were suspended in 35 mL of pepsin solution (1.5 g of enzymelL of O.IM potassium phosphate buffer, pH 2.0) and incubated with vigorous shaking at 37°C. Pepsin (P-7000, Sigma-Aldrich, St. Louis, MO; activity 924 units/mg of protein) digestion was stopped by addition of 2 mL of 2M NaOH at the end of the 2 hr digestion course. After centrifugation at 4,000 x g for 15 min, the supernatant was discarded, and the residue was washed in 10 mL of O.IM phosphate buffer (pH = 2.0) and centrifuged as before. After the second washing and centrifugation, the residue was frozen and then lyophilized. The freeze-dried residue was then weighed and analyzed for nitrogen content. Analytical Methods AOAC official methods (2000) were used to analyze sorghum flour samples for dry matter and moisture (925.10), crude protein (990.03), ash (942.05), crude fiber (962.09), and crude fat (920.39). Total starch and amylose contents were measured through colorimetric procedures (Megazyme total starch and amylose/amylopectin kits; procedures are available at http:// secure.megazyme.comldownloads/enidatalK-TSTA.pdf and http:// secure.megazyme.comldownloads/en/datalK-AMYL.pdf). The presence of amylose in the waxy sorghum kernels was also qualitatively examined through iodine-staining techniques (Pedersen et al 2004). FAN was analyzed through the European Brewery Convention method (EBC 1987) with modification. Grain sorghum flour (150 mg) was mixed with 1.5 mL of deionized distilled water in a 2.5 mL microcentrifuge tube and vortexed five times in 10 min and then centrifuged at 12,000 rpm for 20 min. The supernatant was then ready for FAN analysis. A tannin bleach test followed the Xiang method (2009). Glucose, glycerol, and ethanol in samples were determined by HPLC (Shimadzu Scientific Instruments, Columbia, MD) according to the method described by McGinley and Mott (2008). The column used was a Rezex ROA column (Phenomenex), and the detector was a refractive index detector (RID-lOA, Shimadzu) maintained at 40°C. The mobile phase was 5mM sulfuric acid at a flow rate of 0.6 mL/min, and the oven temperature was 65°C. HPLC data were analyzed with Shimadzu EZStart 7.4 software. Fermentation efficiency was calculated as the ratio of the actual ethanol yield (grams of ethanol determined by HPLC) to the theoretical ethanol yield (= starch grams x 1.11 x 0.511) (Yan et aI2009). Statistical Analyses All experiments were performed at least in duplicate. The tabular results presented were the mean values of repeated experimental data. Regression analyses were conducted in Microsoft Excel with the linear regression function. RESULTS AND DISCUSSION As clearly indicated by the major components of the samples from proximate analysis, the waxy sorghum samples used in this project had diverse genetic backgrounds and physical and chemical properties. Normal cultivars on the market have starch content of 72-76% (db) and protein content of around 12% (db). The starch content of these samples ranged from 65 to 76% (db), and their protein content was 12-15.8% (db). Details of the proximate analysis results are listed in Table I. Effect of Starch on Ethanol Production Figure I shows correlation between total starch content and ethanol yield from fermentation of 25 waxy grain sorghum samples. Ethanol yield (gallons/bushel) was linearly correlated with total starch content (R2 = 0.7946). This result is in agreement with those reported by Wu et al (2007) and Lacerenza et al (2008). Sorghum cultivars with high starch and low protein contents are cultivars of choice for fuel ethanol production. Wu et al (2008) reported that higher starch content means higher ethanol yield, better processing efficiency, and lower amounts of residues after fermentation; therefore, total starch content of waxy grain sorghum can be a predicator of ethanol yield. Average ethanol yield from waxy grain sorghum is similar to com (Lemuz et al 2009). Although the sorghum samples tested in this study have diverse genetic backgrounds, which translate into different starch and protein contents, the ethanol yields ranged from 2.6 to 3.0 gallons per bushel, with an average of 2.8 gallons per bushel. Endosperm of waxy grain sorghum showed little or no sign of amylose when tested by rapid iodine-staining techniques (Pedersen et al 2004). If enough amylose is present in the grain, iodine will bind with amylose in the endosperm of a grain kernel and Vol. 88, No.6, 2011 591 ~--==--------~-------------------------,~---------=----~----------~--~-tum its color dark blue; waxy grains contain no or little amylose and will tum reddish brown (Pedersen et al 2004). The iodinestaining test showed that amylose contents were low in most waxy sorghum samples and slightly higher in a few other samples. The Megazyme amylose assay and DSC analysis results further confirmed the iodine-staining test results. All 25 tested cultivars of waxy grain sorghum had small amounts of amylose, ranging from 5.5 to 7.3%. Fortunately, amylose content in waxy grain sorghum had no significant effect on ethanol yield (R = 0.1341, Figure 1). This was probably because amylose contents in the tested samples were all very low «7.3%) and within a narrow range (5.5-7.3%, Table I). The chances for such small amounts of amylose to complex with lipids in waxy grain sorghum were lower than in normal grain sorghums, which have amylose contents of 23-30% (Wu et al2006, 2010). DSC results confirmed that only four (PI220636, PI217897, PI548008, and PI562758) out of the 25 tested waxy cultivars showed an amylose-lipid complex enthalpy peak at temperatures around 100°C (Table II). Fermentation efficiencies of those four cultivars (with amylose-lipid complex peaks) were lower than those of cultivars without amylose-lipid complex. Actually, two of these waxy sorghum samples (PI562758 and PI548008) had the lowest fermentation efficiencies among all 25 tested samples, 86.0 and 87.7%, respectively; the average efficiency of all 25 samples was 89.6%. Previous research conducted on different ratios of commercial amylose and amylopectin for ethanol production showed that high amylose content led to low ethanol yield (Wu et al 2006). In normal wheat, com, and sorghum, amylose is located in the amorphous region of starch granules and ________________ __ .I> ___________________ L~'_ :~"-o.~~~~_:+-_ ~~!~~ __ 3.00 6. R' = 0.1341 A j ::::. 2.90 ~ :s! .~ 2.80