The wet by-product from the ethanol industry has been used for a long time for the purpose of animal feeding in Vietnam. However, the nutrient composition of this potential by-product has not been elucidated, except for little data recently published [10, 11]. In contrast, there is a huge bank of data for wheat and corn DDGS which have been published several years [2, 15–17]. In fact, rice and cassava have been known as more popular crops in Asian and African countries than maize and wheat. This study was the first systematic work focusing on the nutrient compositions of by-products from different ethanol factories in Vietnam using rice, and cassava as raw materials.
4.1 The composition of DDG depend on bioethanol technology and raw materials
In this study, the variation of composition including crude protein, starch, fat, crude fiber and ash of 7 DDG samples collected from 5 ethanol factories was observed. The variation in composition was also proven for DDGS from corn and wheat [2, 16]. The authors showed that the types of raw material, the production technology and WDG treatment influence on the composition of DDGS, and this was an explanation for the variation in proximates observed with rice and cassava DDGs in this study. Proximate of these DDGs were different due to not only the raw material but also ethanol processing and WDG treatment (Table 1, Fig. 1). In general, for all materials, bioethanol manufacture starts with a liquefaction step in which milled raw material is treated at high temperature of above 100oC and supported by a heat-stable α-amylase. Among these, a high-pressure steam treatment (temperature of 105 -110oC) was applied in Factory 3 and Factory 5 to achieve the highly extensive liquefaction of starch. A saccharification step was performed either separately (as in Factory 1, Factory 2 and Factory 5) or simultaneously with fermentation (as in Factory 4 and Factory 5). Factory 5 carried out a continuous fermentation meanwhile others used batch fermentation. The SLSF sample was different from others because the liquefaction, saccharification and fermentation were simultaneously performed at 30oC with yeast in the supporting of liquefying and saccharifying enzymes.
Moreover, the WDG was differently harvested and treated from the factory to others. The WDG was separated by a plate-frame filter before the alcohol distillation to avoid blocking the distilling tower as at Factory 1 and Factory 2. Meanwhile, in other factories, the WDGs were harvested from the whole stillage after the alcohol distillation by a high-speed decanter (Factory 3, Factory 4) or by frame-plate filter (Factory 5). Moreover, in Factory 3, the thin stillage after the decantation was partly reused to reduce the pH of the rice mash. Interestingly, in Factory 4, spent from the cassava starch harvesting process, containing fiber up to 50% dry matter, was combined with WDG from the bioethanol process.
The composition of raw materials such as white rice, cassava chips and corn has been extensively elucidated. Based on dry matter, rice (milled rice or white rice) is composed of less crude fiber (0.1–0.8%) than corn (3%) or cassava (3.7- 4.0%) [18–21]. On the other hand, white rice contains 7–8% of crude protein, which is less than that in corn (9.42%) and higher than in cassava (1–3%) [18–21]. Three types of raw materials were composed of a comparable level of carbohydrates with mainly starch, i.e., 80% for rice and cassava and 72% for corn. During the bioethanol processing, the un-converted components such as crude fiber in the raw material, which are mostly left in DDG, could lead the difference in DDG composition. This explains the much higher level of protein in the rice DDG than that in the corn DDG and the cassava DDGs (Table 2). Unlike wheat/corn DDGS, even though the soluble in thin stillage was not taken into account (Fig. 1), the rice DDGs in this study contained more protein than the corn DDGS which ranged from 26–32% [2, 17, 22–24]. Also, in this work the protein content in the corn DDG (Factory 5-CO) was slightly higher than the range (35.6%) (Table 2). Yeast biomass has been known as a rich source of protein which was reported in a range of 38.8–70.7% dry matter [25]. The amount of protein in DDGS is affected by yeast as shown in a review by Liu et al [2]. In an uncertain estimation, the authors indicated that yeast contributes approximately 5.3% of the protein in DDGS [2]. This number is much lower than the amount of protein observed in rice DDG in this study (50–80% of dry matter). This suggests most of the protein in rice DDG was from rice material. This statement is also supported by the low protein value in cassava DDGs (13.38–16.40%) (Table 2). If protein in DDG were mainly contributed by yeast biomass, the protein content in cassava-DGG would be much higher than the observed values.
The higher fiber in cassava resulted in the higher of this parameter in cassava DDGs (up to 36.5% in Factory 5-CS) in comparison to the rice DDGs (Table 2). The high content of fiber in the cassava DDG Factory 4-CS was explained by the addition of the cassava spent by-product of starch processing which contains up to 50% of fiber. The fiber content in the corn DDG (Factory 5-CO) in this study was in agreement with the reported range for DDGS (7.22–10.2%) [1, 17, 24].
The content of residual starch in DDG samples indicated the effectiveness of heat treatment, amylolytic conversion by amylases and fermentation by yeast [17]. Starch content in corn DDGS ranges from 3–6% [2, 17]. The starch content in DDGs in this study might be overestimated due to the analysis method using HCl hydrolysis. The reducing sugar from non-starch carbohydrates can be generated after acid hydrolysis.
Despite the lower ash content in cassava [18, 20], the ash content of cassava DDG was higher than that in rice DDG (1.44–2.61%) and corn DDG (3%) (Table 2).
There was no correlation between fat content in raw material and DDG samples. For example, the fat content in rice (1–2%) and cassava (0.5-1.0%) is lower than in maize (4.72%) [19, 20], and the fat in the rice-DDG varied from 0.69% − 10.41% (Table 2). We also found in this work that the fat content in the corn DDG Factory 5-CO was 4.56%, which is half to one-third of the reported range from 10.2–14.5% of DDGS [2, 17, 23, 24].
The difference in compositions of the rice DDG samples from Factory 2-R, Factory 3-R, Factory 4-R and the SLSF-R suggests that the ethanol processing technology leads to a significant effect. Especially, the Factory 2-R and the SLSF-R used the same rice material but different process technology showing varied data in protein, starch, fiber and other components. Interestingly, the SLSF-R differs remarkably from that of Factory 2-R on protein (51.49%), fiber (18.42%) and fat (7.47%). It is likely because of the uncooked and simultaneous saccharification and fermentation in the SLSF process (Fig. 1).
4.2 The potential of rice in food application and cassava DDG in animal feeding
DDGS was considered not only for animal feeding but also for food application recently. This study showed that the rice DDG which contained protein up to 50–80% of dry matter was highly potential for food application, especially as the source of plant protein. In comparison to soybean seed which is well-known as a main source of plant protein for food application currently, the protein in rice DDG was much higher than the protein in soybean seed (ca 40% of dry matter [26]). Moreover, the profile of amino acids in the rice DDG is balancing of essential and non-essential amino acids (Table 3). Similarly in rice as reported previously [27], in the rice DDG the essential amino such as isoleucine, leucine, phenylalanine, valine, threonine are also present at higher levels than others. This also means the ethanol process did not change the amino acid profile of raw material to DDG as mentioned previously [2]. Lysine, an important amino acid, is considered the first limiting in cereals such as maize, cassava and even in some varieties of rice [28]. In some varieties of Vietnamese rice, lysine was in the range of 3–5% of total protein [27], which is in agreement with the lysine content in rice DDGs found in this study (approximately 2.5–4% of rice DDG) (Table 3). Lysine presents in rice DDG at a level of 2.5-4 times higher than in cassava or corn DDG (Table 3).
Among rice DDGs, SLSF-R was not treated with heating, therefore protein (~ 51% of dry matter) in the SLSF-R was expectedly unmodified as in other rice DDGs due to heat treatment. Protein from SLSF-R was preliminarily harvested with alkaline extraction and precipitation by pH 4.5 adjustment. The main protein fraction in rice, glutenin (amounted to 75–90% of total rice protein) [29], was reported as well-dissolved in alkaline condition [30], therefore NaOH 0.5M was used for extraction. The protein content was increased to 70–72% in resulted precipitate with a recovery yield of 54–55% (Table 4). The enriched protein precipitate from SLSF-R was applied to replace wheat flour in the bread recipe. With such high protein, the replacement of enriched-protein precipitate from SLSF-R to wheat flour can improve the protein content in bread, however, it led the reduction of gluten which is needed for the textural structure of bread, therefore the rheological properties of bread were negatively affected, especially at the high ratio of replacement [3]. Among these, the ratio of 5% replacement was acceptable in flavor and taste. Even though, with a such high protein content, it is highly potentially applied in many non-structural food products. Protein from rice DDG, furthermore, could be used for enzymatic hydrolysis to improve the digestibility [13] or to obtain the bioactive peptides [31].
Cassava DDG contained a low level of protein but a high level of fiber. Fiber, even though does not generate energy, has to be included in the diet to maintain normal physiological functions in the digestive tract [32]. Using molds for fermentation to enrich protein in cassava DDG has been reported by Mai-Dinh Vuong et al. with protein content increased from 11.96 to 15.27% using Trichoderma harzianum strain [11], similar results were observed in this study with the protein content increased from the initial value of 16.42% up to 18.08% − 20.7% of dry matter, depending on strains (Table 6). Among used fungal strains, Rhizomucor miehei FCH 116.3 and Thermothelomyces thermophilus FCH 112.2 showed better protein improvement than others. Moreover, the evaluation of carbohydrate-active enzymes showed that the strains had the ability to produce the three enzymes required in animal feed on cassava-based WDG substrate (Fig. 3). The on-site production of enzymes through fermentation for animal feed production and high protein biomass generation is a promising strategy.