3.2.2. Composition of Fibers and Extracts

The carbohydrate and lignin contents in the fiber after steam refining are important process characteristics and figured out in Figure 3. Therefore, the main carbohydrates glucose and xylose, the further hemicellulose monomers (arabinose, galactose, mannose, rhamnose) and as well the acid-insoluble hydrolysis residues were analyzed. The composition was calculated based on the fiber fraction (Figure 3a) and based on the original raw material (Figure 3b).

Regarding the fiber fraction (Figure 3a), it can be stated that with increasing severity the proportion of glucose in the fibers increases. In contrast to that, a strong decreasing tendency is visible for xylose, which is degraded or can be found in the extract fraction due to the preferential solubilization of hemicelluloses. This impact can also be seen for the hydrolysis residue, which is increasing in the fiber fraction due to the loss of xylose and the accumulation of lignin.

**Figure 3.** Comparison of carbohydrate and residue content of the steam-refined corn stover fiber fraction. The composition of fibers is referred to as fiber fraction (**a**) and raw material input (**b**).

If the obtained fiber yield is considered, the components can be calculated based on the dry material input. With this view, it can be seen that in contrast to the fiber-based view, the amounts of detected glucose and of the hydrolysis residue are nearly stable. Regarding the xylan content, the xylan is heavily degraded or dissolved (Figure 3b) and is thus decreasing.

The described tendencies for the carbohydrates in the fiber fraction, with nearly stable glucose contents and decreasing xylose amounts with increasing severity, are also described in the literature for steam refining of poplar at similar severity factors by Schütt [39]. For the mild steam explosion of corn stover, similar values of lignin, cellulose and hemicellulose are reported recently [38].

Referring to this, Bura et al. [40] described similar findings for steam treatment of corn stover with addition of SO<sup>2</sup> at low (log *R*<sup>0</sup> = 3), medium (log *R*<sup>0</sup> = 3.4) and high (log *R*<sup>0</sup> = 4.2) severities. They report increasing glucose yields, decreasing xylose yields and increasing lignin residue for the fiber fraction. Nevertheless, based on raw material the amounts of glucose are in comparison lower, whereas the amounts of xylose in the fiber fraction are much higher for corn stover experiments in contrast to poplar wood results [39].

The extract fraction was analyzed as well regarding the carbohydrate and residue content and is illustrated in Figure 4.

The composition of the extract is referred to as raw material input. When the yield data of the extract are referred to the raw material input (Figure 4), the maximum of xylose obtained between severities between 3.5 and 4 is, of course, less pronounced. This kind of presentation illustrated clearly that rather small quantities of the raw material components can be retrieved in the extract fraction, especially at severities higher than four. However, it is interesting to see that there is a gap between the raw material-based xylose yields from the fibers (Figure 3b) and the extract xylose yields in Figure 4, especially at severities higher than 3.5. This loss of hemicellulose can be explained with the formation of degradation products, like 5-HMF and furfural in the next chapter. Schütt [39] described the extract fraction of steam-refined poplar wood increasing xylose contents up to severities around 4. Subsequently, the xylose degradation occurred due to more intense pretreatment conditions. These findings are quite similar to the findings for the xylose content of the corn stover extracts after steam refining in the present study. Bura et al. [40] report slightly increasing glucose yields and increasing xylose yields in the extract fraction after steam pretreatment of corn stover at low (log *R*<sup>0</sup> = 3), medium (log *R*<sup>0</sup> = 3.4) and high (log *R*<sup>0</sup> = 4.2) pretreatment conditions. They did also report a slight increase of the xylose in the extract fraction from a severity of 3.4 to 4.2. It must be remarked that the authors used SO<sup>2</sup> as acid catalysts for the dataset. Additionally, the medium pretreatment condition was used without an acid catalyst. Higher yields of xylose in the extract fraction at medium conditions

are clearly visible for the experiments with SO2, but by calculating the total xylose mass balance there is a higher xylose loss visible compared with the results without a catalyst. However, the results without an acid catalyst are comparable with the results presented in this study.

Σ Other carbohydrates **Figure 4.** Comparison of carbohydrate and residue content of the steam-refined corn stover extract fraction.
