The untreated rice straw consisted of cellulose

The untreated rice straw consisted of cellulose (45.15%), hemicellulose (22.87%), lignin (18.61%), and ash (10.33%). Results of chemical composition changes of pretreated rice straw are presented in Table 1. According to results with increasing pressure and retention time, the percentage of cellulose content increased in pretreated rice straw compared with that of the unpretreated. However, other components such as lignin, hemicellulose, and ash have been reduced significantly. The analysis demonstrated that in the cycle 10 bar and 10 min with blank moisture, followed by 15 bar and 1 min, 15 bar and 5 min, resulted in more lignin and ash removal. The content of hemicellulose was generally reduced in the pretreated rice straw. Pretreatment in the cycle 5 bar and low retention time approximately led to the same amount of the lignin, ash, and carbohydrate as the unpretreated sample (Table 1). These results indicate that the pretreatment technique is capable of removing lignin and hemicellulose under high pressure condition.

Pressure is the most important parameter in steam explosion pretreatment compared to the other parameters such as reaction time and moisture ratio. During the steam explosion pretreatment, the hemicellulose is first degraded followed by the lignin when the temperature >150 °C (Ma et al., 2014; Xiao et al., 2014). In addition to the removal of the hemicellulosic, the steam explosion induces important modifications in the structure of lignins. Lignin degrades in the temperature range of 120-200 °C and it divides into smaller particles and is separated from celluloses (Fan et al., 2016). It damaged the cell wall of rice straw by disrupting the lignin structure.

The analysis show that pretreated rice straw contained 91.8% total solids and 60.9% volatile solids on a wet weight basis that is increased by different cycles of pretreatment. Table 2 shows the percentage of VS and TS of the pretreated rice straws. These results also show that increase pressure of the pretreatment led to the increase of VS content. This is an indication of higher carbohydrate contents in the pretreated rice straw (Dehghani et al., 2015).

3.3 Effects of Pretreatment on Rice Straw Structure
Results of changes in the structure and the functional groups of rice straw are shown in Table 3. Bands have been found in all of the samples in the range of 450-3,600 cm-1. The band between 3,600 and 3,100 cm-1 caused by the presence of alcoholic and phenolic hydroxyl groups involved in hydrogen bonds that is related to OH stretching vibrations present in the cellulose, hemicellulose, and lignin (Rahnama et al., 2013). The intensity of band 3415 cm-1 was decreased, as a result after the steam explosion pretreatment in the high-pressure cycles 10 bar for 5-10 min and 15 bar for 1-5 min with blank moisture, 15 bar for 10-15 min with moisture 35% and 10 bar, 10 min and 15 bar for 5-10 min with moisture 70%. Therefore it indicates that the partial hydrogen bond in cellulose was destroyed. This is a key step toward increasing the accessibility of enzymes and microorganisms to cellulose (He et al., 2008). The intensity of peaks at 2920 and 2861 cm-1 was related to C-H stretching vibrations. The intensity of both peaks indicates the distinguished features of cellulose (methyl and methylene) and hemicellulose. The intensity of bands was reduced after the pretreatment pressure increased (Kazeem et al., 2017). Structural changes in lignin and loss of aromatic units were shown by changes of intensity in the 1,649 cm-1 band, and the intensity of bands at 1516 cm-1 is related to C=C stretching of the aromatic ring of lignin (Isroi et al., 2012). The intensity of both peaks decreased with increase in pretreatment pressure in the cycle 10 bar for 5-10 min and 15 bar for 1-5 min with blank moisture, 10 bar and 10 min and 15 bar for 10-15 min with moisture 35% and 10 bar for 5-10 min and 15 bar for 5-10 min with moisture 70%. This was an indication of lignin structure changes.

The intensity of band obtained at 1064 cm-1 is usually attributed to the structural specifications of cellulose and hemicelluloses (Rahnama et al., 2013). It was clear that the absorption band at 1064 cm-1 was lower compared to the unpretreated rice straw. The drop observed in this band is related to decreased hemicellulose content after steam explosion pretreatment in the cycles 10 bar for 5-10 min and 15 bar for 1-5 min with blank moisture, 10 bar, 10-15 min and 15 bar for 10-15 min with moisture 35% and 10 bar for 10-15 min and 15 bar for 1-5-10 min with moisture 70%. The increase in the intensity of this band might be dependent to the dissolution of non-cellulose components that causes the increase of cellulose content in the rice straw (Ang et al., 2012). The absorption band 1430 cm-1 is related to amorphous cellulose (Kazeem et al., 2017). The absorption band by 1430 cm-1 has reduced after the steam explosion pretreatment and it means the degraded of the amorphous part structure, especially the lignin structure changes has occurred. Lignin structure was degraded in the cycles 10 bar for 5-10 min and 15 bar for 1-5 min with blank moisture and 15 bar for 10-15 min with moisture 35% and 15 bar for 5-10 min with moisture 70%.

Crystallinity index of the lignocelluloses is determined by two parts amorphous and crystalline. The amorphous part consists of hemicelluloses and lignin, while the crystalline constituent part includes cellulose (Sakdaronnarong & Jonglertjunya, 2012). The crystallinity index CrI (%) unpretreated and pretreated rice straws are shown in Table 4. Two peaks were observed at 2? of between 18°-19° and between 22°-23°, that relating amorphous and crystalline regions of the rice straw. According to results, the crystallinity index increased in the pretreated rice straw than to the unpretreated rice straw. The highest crystallinities of pretreated rice straw were 28.8%, 28.58% and 29.3% in the cycles 10 bar and 10 min with blank moisture, 15 bar and 15 min with moisture 35% and 15 bar and 10 min with moisture 70%, respectively, than to the unpretreated rice straw 22.9%. In all pretreated samples, the region of crystalline cellulose was higher than to the unpretreated rice straw (Table 4). During the steam explosion pretreatment, Hemicellulose and lignin are hydrolyzed with the increase of pressure, which mainly constitutes amorphous regions of rice straw. This led to the Increase of crystallinity index after pretreatment.

Previous research has also suggested that the crystallinity index of rice straw could enhance by other thermal pretreatments, and this increase not have negative effect on fermentation (Kshirsagar et al., 2015). Increase in crystallinity index due to the effect of pretreatment was more in the amorphous region than to the crystalline region. Rahnama et al. (2013) reported an enhancement in the crystallinity index from 50.81% to 61.41% in the alkali pretreated rice straw. Crystallinity index increased in the pretreated rice straw with microwave and acid to 61.36% compared with pretreated rice straw 52.2% (Renu Singh, 2014).

3.3 Effect of Steam Explosion Pretreatment on Biogas Production
The cumulative biogas production from the pretreated rice straw after 60 days of incubation is presented in Table 5. Retention time, pressure and moisture level were considered as pretreatment variable parameters, and the effects of the pretreatments were evaluated using the accumulated methane production as the response variable. The initial production rates were measured as the average of the produced methane per day during the first ten days of anaerobic digestion process, and are shown in Table 5.

The yield of methane produced by inoculum was deducted from the total of produced methane from each sample. The highest initial production rate was after pretreatment of rice straw in the cycle 10 bar-10 min with an increase of 59% from 92 to 146 N ml/g VS and followed by 144 N ml/g VS, compared with that of the unpretreated (Fig 1c). The unpretreated rice straw resulted in almost the same initial production rate in methane production as compared to some the pretreated rice straw in the blank moisture with low pressure and time (Fig. 1a). The highest initial production rate was obtained in pretreated rice straw in the cycle 15 bar-10 min with moisture 35%, with an increase of 166% from 92 to 245 N ml/g VS followed by 244 N ml/g VS (Fig. 2c). The maximum initial production rate was after the pretreatment of rice straw in the cycle 10 bar-15 min with moisture 70% with an increase of 141% from 92 to 222 N ml/g VS followed by 221 N ml/g VS (Fig. 3b).

The highest methane production was obtained for pretreated rice straw in the cycles 10 bar-10 min and 15 bar-5 min with blank moisture 418 and 370 N ml/g VS compared with that of the unpretreated rice straw 196 N ml/g VS (Table 5) and this means increased by 114% of the methane yield. The pretreated rice straw in the cycles 15 bar for 15-10 min with moisture 35% was obtained the highest methane production 399 and 387 N ml/g VS, respectively. Methane yield increased by 196 to 399 N ml/g VS, and this means increased by 104% compared to the unpretreated rice straw. The highest methane production was reached for the pretreated rice straw after pretreatment in the cycles 15 bar for 10-5 min with moisture 70% to 496 and 450 N ml/g VS, respectively (Fig. 3c). The methane yield of rice straw increased by 196 to 496 N ml/g VS, and this means increased by 154% compared to the unpretreated sample.

According to the improvement of methane production yield, structural modification of rice straw was performed by steam explosion pretreatment, and methane production in primary days of the anaerobic digestion process was increased for the pretreated rice straw. According to the initial production rate of pretreated rice straw in the cycles 10 bar-10 min, 15 bar-15 min with moisture 35% and 15 bar-10 min with moisture 70%, were about 35%, 57% and 37% of methane production yield occurred in 10 days of incubation, respectively.

The lowest initial production rate was observed, in the cycles 5 bar-15 min with blank moisture, leading to a methane level of 91 N ml/g VS than to the unpretreated rice straw 92 ml/g VS (Fig. 1b). The lowest methane production was in the cycle 10 bar-15 min with blank moisture, leading to a methane production of 178 N ml/g VS that the less was than to the unpretreated rice straw 196 ml/g VS (Fig. 1d). The lowest initial production rate and methane production was observed, for pretreated rice straw in the cycle 5 bar-1 min with moisture 35%, leading to a methane production of 166 and 267 N ml/g VS compared with that of the unpretreated 92 and 196 ml/g VS, respectively (Fig. 2a and 2b). Pretreated rice straw in the cycle 15 bar-10 min with moisture 70% resulted in the lowest initial production rate of 180 N ml/g VS (Fig. 3c). The lowest methane production was in cycle 5 bar-5 min with moisture 70% leading to a methane production of 272 N ml/g VS (Fig. 3a).

Parameters effects, confidence intervals of steam explosion pretreatment were calculated on the anaerobic digestion experiments. The response surface after steam explosion pretreatment was with two levels of each parameter. The parameters were used as covariates to be able to measure the size of the effects of the pretreatments. The parameters were considered significant when the probability (p-value) was less than 0.05. The ANOVA analysis of the response surface design model (Table 6) showed pressure and moisture as significant factors in increasing methane yield, while the time was not statistically significant (Fig. 4).

According to the results of fits and diagnostics for unusual observations, two pretreatments in the cycles 10 bar-10 min and 15 bar-10 min with the moisture level of the blank were the best and worst pretreatment conditions considering standard residual 2.50 and -2.91 and fit 413.6 and 447.5 respectively.

4 Conclusions
Steam explosion is suitable pretreatment method for improving biogas production efficiency from rice straw. The initial production rates were increased by 59%, 166%, and 141% compared with that to the unpretreated rice straw. The methane gas was obtained after 60 days 535, 516 and 613 N ml/g VS that each has increased 114%, 104%, and 154% than to the rice straw unpretreated, respectively. Steam explosion pretreatment was effective in remove lignin and hemicellulose and also opening up the crystalline structure of the cellulose of rice straw prior to anaerobic digestion. This demonstrates the ability of the steam explosion pretreatment to modify the structure of rice straw.