Bioethanol production from grass pea and wild oat hydrolysates using S.cerevisiae ETP53, K. marxianus ETP87, and P. fermentans ETP22

Conversion of lignocellulosic biomass into bioethanol is essential to reduce dependency on fossil fuels. After the grass pea and wild oat straws were characterized, they were hydrolyzed by live fungi (in situ), crude fungal enzymes, and chemical methods (sodium hydroxide and sulphuric acids at 0.5, 1, 2, and 3%). The fungi used for in situ degradation and crude lignocellulosic enzymes were Aspergillus niger JMC22344, Trichoderma reesei JMC22, Pleurotus ostreatus M2191, and Pleurotus sajor-caju M2145. Furfural content of hydrolysates was analyzed by GC–MS. The potential of activated charcoal, overliming, and sequential activated charcoal-overliming potential to reduce furfural in the acid hydrolysates were evaluated. The hydrolysates were fermented by S. cerevisiae ETP53, K. marxianus ETP87, and P. fermentans ETP22. Sulphuric acid hydrolyzed common wild oat (8.91 g/L) yielded higher sugars than NaOH (3.88 g/L) and NaOH treated wild oat (8.14 g/L) was superior in terms of sugars released than grass pea (4.21 g/L). P. ostreatus M2191 liberated the highest sugar (18.63 g/L) extracts than T. reesei JMC22676, A. niger ETP22344 and P. sajor-caju M2145. The 23–77% of the furfural in acid hydrolysates was reduced by overliming, activated charcoal, and the sequential activated charcoal-overliming. Yeasts produced optimal ethanol from straws hydrolyzed by crude enzymes from A. niger JMC22344. Generally, the ethanol titer was directly proportional to the amount of sugars released.


Introduction
The ever-increasing demand for energy and the problem of environmental pollution necessitate the search for renewable energy from various alternatives [1]. The biofuels obtained from agricultural residues, industrial wastes, and other lignocellulose sources are alternative cheap and sustainable sources of energy supply.
In 2016, 2.01 billion tons of solid waste was generated worldwide and it is expected to raise to 3.40 billion tons in 2050 [2]. Among the solid wastes, agricultural residues are important sources of biomass for ethanol production in Africa and the estimated total ethanol from agricultural residues and municipal solid waste (excluding paper) is 20.2 and 1.8 Billion liters per year respectively [3]. The highest available biomass was estimated to be 136 million tons per year in Western 1 3

Treatment of straws
Wild oat (3.13 g) and 3.11 g grass pea straws were transferred to screwed 250 mL bottles containing 100 mL of 0.5%, 1% and 3% sulfuric acid or sodium hydroxide [11]. The screw was closed very well but not tightly. The biomass was autoclaved at 121 °C for 90 min. After cooling, the pretreated biomass was filtered with filter paper to analyze the liquid portion for total reducing sugars and acid soluble lignin using a spectrophotometer whereas glucose, cellobiose, xylose, arabinose, and galactose were analyzed by high pressure liquid chromatography (HPLC). The furfural was quantified by GC-MS. The oven dried solid portion of previously treated straws was retreated with the same condition to evaluate for further possible saccharification.

Crude protein of grass pea and wild oat
Pre-dried grass pea (1 g) and wild oat (1 g) were put in 500 mL Kjeldahl digestion flask. A mixture of copper sulfate and potassium sulphate (10 g) and 25 mL of 0.1 N sulfuric acid was added. The flask was placed in an inclined position and heated gently. It was boiled for 30 min after the solution cleared. After cooling for 15 min, 250 mL of distilled water and then 90 mL of 40% NaOH solution were added. The distillation apparatus was connected with loops and 300 mL Erlenmeyer flask containing 30 mL boric acid indicator solution (methyl red plus methyl blue). Then it was distilled and titrated with 0.1 N HCl until colorless endpoint was reached.

Furfural reduction in acid hydrolysates
The detoxification of acid hydrolysate was run by employing the method of Kuhad et al. [24]. First, the acid hydrolysates (0.5 and 1% H 2 SO 4 ) of grass pea and wild oat were overlimed by mixing with calcium hydroxide until pH reached 10 [24]. Second, activated charcoal (1.5%, w/v) was added to acid hydrolysate according to Kuhad et al. [24]. The two mixtures were stirred with magnetic stirrer at higher speed for 30 min and 60 min to see also the effects of retention time on inhibitor reduction. Third, the overlimed hydrolysate was further mixed with activated charcoal (1.5%, w/v) and continuously stirred to investigate the synergetic effect of overliming and activated charcoal. Then, overlimed, activated charcoal and overlimed-activated charcoal treated acid hydrolysates were centrifuged at 8000 revolutions per minute (Eppendorf, Centrifuge 5418 R) to separate the precipitated calcium sulphate and activated charcoal from hydrolysates. The furfural was measured before and after treatment using GC-MS. The percentage of furfural reduction due to overliming and adsorbing with activated charcoal was calculated as where, FR = Percentage of furfural reduction, FAT = Furfural (g/L) after treatment with overliming and activated charcoal, FBT = Furfural (g/L) before treatment with overliming and activated charcoal.

Crude cellulase enzyme preparation and hydrolysis
One mL filtrate from above was centrifuged at 13,000g× for 10 min to remove cell debris and fungal spores and 3 mL cold acetone which was put at − 18 °C over-night was added to supernatant in acetone compatible tube. After vortexing, the acetone-supernatant mixture was incubated at − 18 °C for 2 h and centrifuged at 13,000g× (Eppendorf, Centrifuge 5418 R) for 10 min to separate crude enzymes from the mixture. The supernatant was carefully decanted. Finally, 0.5 mL acetate buffer at pH 5.0 was added to dissolve the enzyme pellet. Crude enzymes obtained from 10 mL of each culture filtrate were used to hydrolyze 1% (v/v) sulfuric acid pretreated straws (6% w/v on dry weight bases). As a positive control and comparison, carboxymethylcellulose (CMC) was hydrolyzed by crude enzymes from A. niger JMC22344. The mixture was incubated at 50 °C for 12 h. The sugar released was quantified by DNS [26]. The absorbance was measured at 540 nm (Jenway 6405 UV/Vis).

Yeast fermentation of acid or alkali hydrolysate
Bioethanol production from straw hydrolysate was carried out using three local yeast isolates; P. fermentans ETP22, S. cerevisiae ETP53, and K. marxianus ETP87 were isolated from compost, tella, and traditional yoghurt, respectively. The procedure was according to Jung et al. [11] and Martins et al. [27] with a little modification. The pH of acid and alkali treated hydrolysate were 0.9 and 13.2 respectively. The acid hydrolysate pH was adjusted near to 5 using first solid sodium hydroxide and then to 5.5 employing 3 N sodium hydroxide dropwise whereas the alkali hydrolysate pH was adjusted near to 5 using concentrated hydrochloric acid (45%) then to 5.5 by adding 3 N HCl dropwise. The solid NaOH and concentrated HCl were employed to avoid sugar dilution [27]. The hydrolysates were centrifuged at 10,000g× for 15 min and the filtrates were separated from solid particles by decanting. All these were done aseptically to prevent contaminations. Then 50 mL of the hydrolysates were supplemented with sterilized peptone (0.5 g/L) and yeast extract (0.1 g/L) and transferred into 250-mL flask, and inoculated with (approximately 7.5 × 10 6 ) P. fermentans ETP22, S. cerevisiae ETP53, and K. marxianus ETP87. They were incubated in shaker incubator at 30 °C and 130 rpm. Samples were taken at 8 and 24 h for yeast growth, sugar consumption, furfural content, butanol, isoamyl alcohol and ethanol analysis [11].

Ethanol fermentation of enzymatically degraded straws
The hydrolysates from enzymatic hydrolysis were boiled to reduce the volume by half and increase the sugar concentration two times. The hydrolysate (100 mL) was transferred to 250 Erlenmeyer flask and supplemented with yeast extract 1% and peptone 2% for ethanol production. Then it was inoculated with 7.5 × 10 6 cells of each yeast and incubated at 30 °C for 24 h.

High pressure liquid chromatography
The analysis of sugar contents of hydrolysates (such as cellobiose, glucose, galactose, xylose, and arabinose) was performed by high pressure liquid chromatography (HPLC) (Agilent 1100Series).The hydrolysates were centrifuged at 14,500g× (Eppendorf, Centrifuge 5418 R) two times for 10 min to remove any unsolublized materials or debris. The HPLC (Agilent 1100Series) was equipped with refractive index (RI) detector (G1362A) and quaternary pump (G1311A). MetaCarb Pb Plus (300 mm × 7.8 mm, Agilent A5241) column was operated at 80 °C and the flow rate was 0.1 mL/min. The injection volume was adjusted at 20 µL. The mobile phase consisted of 100% HPLC grade water.

Ethanol and furfural determination using gas chromatography
In order to determine the ethanol and furfural, the fermented products and hydrolysates were centrifuged at 14,000g× for 5 min. The supernatant was analyzed by Gas Chromatography (Agilent 6890N) coupled with a mass spectrometer (MS) with headspace auto-sampler (Agilent 7694E). The GC-MS was equipped with Mass Selective Detector (Agilent-5973Network) and a polar polyethylene glycol (PEG) phase DB-wax122-7032 capillary column. Helium was used as a carrier gas. The flow rate for the column was 1 mL/min. The column temperature was held at 40 °C for 4 min, and then raised to 240 °C at 40 °C/min; the inlet temperature was 220 °C. The GC was operated with 20:1 split injection of the headspace.

Yeast production
Yeast production was determined as a dry weight (g/L). One mL of yeast broth grown for 24 h was transferred into 1.5 mL eppendroff tube and centrifuged at 13,000g× (Eppendorf centrifuge 5418 R, Germany). The pellets were carefully transferred and dried in dry oven at 70 °C for 3 days and weighed using precision balance (Radwag WAS220/C/2, Poland).

Straw hydrolysis by dilute acid and alkali
The amount of reducing sugars released from acid-and alkali-treated lignocellulosic residues is shown in Table 2. The sugar yield of acid treated was between 5.3-8.9 g/L and 7.3-13.9 g/L from wild oat and grass pea, respectively compared with 3.4-4.2 g/L and 6.5-8 g/L obtained from the alkali treated wild oat and grass pea respectively, indicating that the sugar released from acid hydrolysis was 1.5 times higher than alkali hydrolysis. The sugar contents were drastically decreased when more than 2% concentrations of acid/alkali was used for the treatment. Jung et al. [11] attained optimal sugars (14.5 g per 100 g straw) from Thai Mission grass (Pennisetum polystachyon) which was hydrolyzed by 1% H 2 SO 4 (w/v). In the first hand, Juarez et al. [28] reported the liberating of higher reducing sugar from spent coffee at 5% (v/v) sulfuric acid and a hydrolysis time of 3 h. On the other hand, more reducing sugars were obtained when the algal (Chlorella vulgaris) biomass was hydrolyzed by 3% (v/v) sulfuric acid [29]. Similarly, 3% (v/v) sulfuric acid treatment of palm wood liberated higher reducing sugars [18].
The sugar yield of 4.26-8.19 g/L obtained from 2% alkali treated substrates in this study is much lower than 6.6 g/100 g straw released at 3% (w/v) NaOH hydrolyzed mission grass pretreated with microwave Tatijarern et al. [30], and the application of 4-7% NaOH resulted in the degradation of monomeric sugars released from switch grass [31]. Therefore, the lower sugar at higher alkali and acid treating chemical might be the reason.

Hydrolysate sugar composition
The composition of the different types of sugars in the acid and alkali treated hydrolysates is shown in Table 3. In all cases, the concentration of xylose was higher than glucose, except from 1% NaOH hydrolyzed grass pea and this suggests that the acid treatment might attack the hemicellulose portion of lignocellulose. This shows that most of the xylose (75-90%) present in hemicellulose could be extracted by dilute acid pretreatment technology [32] Similarly, de Cássia et al. [33] reported that D-xylose (19.19 g/L) was the major product in hydrolysates released from acid treated sugarcane bagasse hemicellulose whereas the concentrations of d-glucose (0.981 g/L) and l-arabinose (1.82 g/L) were very low. The authors showed that the mild hydrolysis condition was not enough to release d-glucose from the cellulose fraction. However, glucose was found to be dominant in dilute acid hydrolysis of algal biomass [29] and this might be higher starch present in the algae. Like de Farias Silva and Bertucco [29] studies and unlike this study, glucose concentration was higher than xylose in dilute acid hydrolysates of globe artichoke crop residues [34].

Ethanol production from acid and alkali hydrolysates
It is interesting to note that, although the acid-treated hydrolysates contained higher sugar contents than the alkalitreated ones, more ethanol was produced from the alkali treated ones (Table 4). Accordingly, the ethanol content from alkali-treated wild oat and grass pea hydrolysates was in the range of 0.8-2.6gL and 0.3-3.3 g/L, respectively. The higher ethanol was produced by S. cerevisiae ETP53 (2.6 g/L ethanol) (8 h) and P. fermentans ETP22 (4.2 g/L ethanol (24 h) and 3.3 g/L (8 h). The lowest ethanol (1.8 g/L) was detected from K. marxianus ETP87 (24 h). P. fermentans ETP22 was able to grow on xylose, the dominant sugar in hydrolysates and it might use xylose for growth and leave the glucose for more ethanol production. Huang et al. [35] suggested that xylose fermenting yeasts are highly needed to produce ethanol from acid hydrolysates that contains higher percentage of xylose than glucose. Candida tropicalis JH030 which was grown in non-detoxified rice straw acid hydrolysates containing 4.5 g/L glucose, 35.9 g/L xylose, and 4.3 g/L arabinose resulted 3.3 g/L ethanol [35]. The ethanol produced by the C. tropicalis JH030 was higher (3.3 g/L) than half of the glucose ( 1∕2 × 4.5 = 2.25 ) and this shows that C. tropicalis JH030 utilize xylose to produce ethanol.
In general, the yeast species showed variations in ethanol production upon different incubation periods. Thus S. cerevisiae ETP53 attained the highest concentration of ethanol (4.6 g/L) in 8 h; whereas P. fermentans ETP22 achieved optimum production of 4.2 g/L and 3.3 g/L in 24 h and 8 h on alkali treated hydrolysates. On the contrary, K. marxianus ETP87 did not show a significant pattern in ethanol production at different incubation times. Different studies have demonstrated variation on K. marxianus and S. cerevisiae potential to tolerate inhibitors and produce ethanol depending upon treatments, types of substrates and strains [36].
In this study, however, K. marxianus ETP87 was able to grow and ferment sugars both in acid and alkali hydrolysates but ethanol amount was lower than S. cerevisiae ETP53 (Table 4). K. marxianus is the efficient ethanol producer from non-Saccharomyces yeast and it is competent enough to produce ethanol from lignocellulosic hydrolysates [18]. Isoamyl alcohol was produced by only K. marxianus ETP87 that grew in acid treated wild oat and grass pea straws.

Detoxification of hydrolysates
The effect of treatment with liming and active carbon and their combination on the possible removal of the inhibitory by-products (furfurals) was tested on the acid treated hydrolysates ( Table 5). The data showed that the combination of overliming and activated charcoal removed 77% of the furfurals from wild oat and 70% from grass pea, followed by treatment with overliming with 53% removal efficiency of furfural (wild oat) and 40% (grass pea) in almost 30 min. However, treatment of the hydrolysates with activated carbon removed furfural by 35% and 43% from grass pea and 23% and 30% from wild oat upon 30 and 60 min, respectively, indicating that raising the adsorption time from 30 to 60 min slightly, but not significantly, decreased the furfural removed by activated charcoal and overliming but raising time had no effect on the synergetic application of overliming and activated charcoal (Table 5). Activated charcoal reduced not only furfural but also other inhibitors organic acid (acetic and formic) and 5-hydroxymethylfurfural and this enhanced ethanol production by S. cerevisiae [37].
Combined detoxification through overliming followed by activated charcoal strongly reduced furfural content by 77.45% and 68.62% for grass pea and wild oat hydrolysate respectively whereas overliming and activated charcoal separately detoxified the furfural to a less extent (Table 5). In the first hand, Keshav et al. [39] reported that activated charcoal treatment and ammonia neutralization of acid hydrolyzed oil palm empty fruit bunches removed 90% of the furfural [11].
The potential of the yeasts to produce ethanol from detoxified and non-detoxified hydrolysates is shown in Table 5. All the three yeasts produced higher ethanol in detoxified hydrolysates than non-detoxified counter-parts. The yeast S. cerevisiae ETP53 and K. marxianus ETP87 produced more ethanol from wild oat hydrolysate than the grass pea. On the contrary, P. fermentans ETP22 produced relatively high ethanol on grass pea hydrolysate than wild oat. Likewise, S. cerevisiae ETP53 produced the highest ethanol (2.86 g/L) in wild oat hydrolysate overlimed with Ca(OH) 2 whereas P. fermentans ETP22 produced the lowest (0.51 g/L) ethanol in wild oat hydrolysates treated with activated charcoal ( Table 5).
Fermentation of dilute acid hydrolysates of red sage by Pichia stipitis resulted in 5.16 g/L ethanol [24], which was higher than the present study. The maximum ethanol concentration produced by P. stipitis grown in over-limed acid hydrolysates of rice straw was 9.29 g/L ethanol which may be associated with xylose utilization [32]. In this study, however, even if P. fermentans ETP22 metabolized xylose, it did not produce enough ethanol. In general, the production of more ethanol in detoxified hydrolysates is due to increased cell growth and sugar consumption when inhibitors were reduced by different methods [12,36,37,40].

Ethanol production by employing sugars extracted from solid state grown fungi
The effect of enzymatically treated hydrolysates from T. reesei JMC22676, A. niger IMC22344, P. ostreatus M2191, and P. sajor-caju M2345 in wild oat and grass pea straw is indicated in Table 6.
The data indicated that the amount of sugar produced in the hydrolysates of the two treated and untreated substrates was in the range of 7.25-18.63 g/L. The highest reducing sugar was derived from oat straw ranging from 12.  twice as much as reducing sugars released from oat. The quantity of reducing sugar showed a pattern in that untreated substrates contained more sugars in their hydrolysates than the treated ones. The lower sugar in treated straw might be associated with loss of sugars by treatment before the fungus was inoculated. The hydrolysate derived from P. ostreatus on oat straw was the highest (18.63 g/L) followed by T. reesei and P. sajor-caju. The amount of ethanol produced from the hydrolysate also showed variations amongst the fermenting yeast (Table 6). Accordingly, K. marxianus ETP87 produced the highest amount of ethanol from oat hydrolysate (3.49-8.25 g/L) with average ethanol of 6.02 g/L, followed by S. cerevisiae ETP53 with 3.32-7.76 g/L with average ethanol production of 5.84; and P. fermentans ETP22, with a range of 2.69-6.95 g/L and mean of 5.11 g/L. However, Raja Sathendra et al. [18] were able to produce higher ethanol than this study by mixing, K. marxianus and T. reesei in acid pretreated palm wood.
The efficiency of ethanol production by all the yeasts corresponded with the amount of sugar released in acid treated/ untreated but enzymatically treated wild oat with P. ostreatus M2191 (Fig. 1); followed by the same substrate treated with by P. sajor-caju. The data showed that all the yeast strains produced the highest concentration of ethanol derived from the hydrolysates of P. ostreatus grown on wild oat (6.95-8.25 g/L) which was almost twice higher than the ones treated with A. niger although the difference in the amount of sugar released by the latter was only 21%.
Singhania et al. [1] reported that 10.5 g/L ethanol was generated by K. marxianus MTCC 4136 from 27 g/L glucose released by crude enzymatic activity from acid pre-treated and crude cellulases from Penicillium janthinellum.
In this study, the ethanol concentration in sugars released by crude enzymes obtained from treated substrates was lower than acid or alkali hydrolysates and fungi hydrolysate. This is due to 10% of enzymes extracted from 15 g of solid state fermented fungi were used to degrade 12 g of pretreated straws. Cunha et al. [41] suggested that due to complex structure and recalcitrant nature of lignocellulose, high crude enzyme loadings are needed in the production of ethanol from biomass.

Yeast production from acid and alkali hydrolysates
The yeast productivity in acid hydrolyzed straws was higher than alkali hydrolyzed straws (Fig. 2). The yeast dry weight obtained from P. fermentans ETP22 was higher on 1% acid treated grass pea with 7.0 g/L (biomass) followed by 6 g/L  from K. marxianus and 4-5 g/L with S. cerevisiae on the same substrate, respectively. The biomass recovered from acidhydrolysates was higher than alkali-hydrolysates (Fig. 2). Minimum yeast dry biomass was obtained in all yeasts grown on grass pea treated with 1% (w/v) sodium hydroxide. Generally, yeast growth and ethanol production are inversely related since most of the sugars are utilized for cell growth and functions during yeast growth.

Biomass production from in situ degraded straws
The effect of fungal treated hydrolysates on biomass production by the three test yeasts is shown in Fig. 3. Accordingly, K. marxianus ETP87 and P. fermentans ETP22 gave higher biomass than the S. cerevisiae ETP53; except on P. sajor-caju and T. reesei treated grass pea, the biomass production ranged from 5.0 to 8.0 g/L. The pattern showed that more biomass was recovered from wild oat than grass pea. This was much higher than 2 g/L biomass obtained from S. cerevisiae grown on the hydrolysates of Chrysanthemum waste by P. ostreatus degradation [42].

Conclusion and recommendations
The ethanol concentration produced from straws varies depending on the amount of sugars and inhibitors released during chemical hydrolysis. Alkali hydrolysis released less furfural and solubilized more lignin than acid hydrolysis. Both alkali and acid hydrolysis of grass pea resulted in higher sugar and ethanol production than wild oat. The pentose sugars, which were not fermented by S. cerevisiae ETP53, P. fermentans ETP22, K. marxianus ETP87, were dominant in chemical hydrolysates of wild oat and grass pea. This further reduces the ethanol concentration.
All the detoxifying methods such as overliming, activated carbon, and overliming followed by activated charcoal reduced the furfural content of acid hydrolysates. Prolonged time (greater than 30 min) for overliming and overlimingactivated charcoal did not further reduce the furfural content of acid hydrolysate but activated carbon increased the reduction when the incubation time was increased up to 60 min.