Raw material

Straw obtained from sorghum [Sorghum bicolor (L.) Moench.] of the RB-Paloma variety was donated by the National Farming Research Institute, Campo Experimental Río Bravo, Tamaulipas, Mexico. The straw was cut into pieces, 5 cm in length, dried in a stone with ventilation at 60 °C for 24 h, grounded and sieved to obtain particles with an approximate diameter of 500 mm.

Preliminary studies

Given that the most widely used pretreatment for acid hydrolysis is grinding, in order to obtain a larger surface area, we determined the feasibility of using fine flour (particle size ≤0.05 µm) and another, thicker one (≥500 mm, product of sieving), or using a mixture of both types of flour in a 1:1 proportion to use all the residue. Hydrolysis was carried out under the following conditions: sulfuric acid at 4 % for 80 min at 120 °C, with a solid-liquid ratio of 1:6 with all three treatments; these conditions were selected according to studies by ^{Aguilar et al. (2002)}, ^{Téllez-Luis et al. (2002)} and ^{Sepúlveda-Huerta et al. (2006)}.

Hydrolysis and detoxification

Hydrolysis was carried out with sulfuric acid under the following conditions: acid 2, 4 and 6 %; solid-liquid ratio of 1:6, 1:8 and 1:10. Reaction time and temperature were constant (80 min and 120 °C, respectively) and nine hydrolysis treatments with sorghum straw flour (Table 1).

Afterwards, the best conditions for hydrolysis were chosen, the syrups obtained were divided into two equal parts, and one was treated with active carbon to eliminate inhibiting compounds, furfural and acetic acid. The detoxification conditions were as follows: pH 5 with a time of 45 min and a charge of active carbon at 1.5. The other part of the syrups was not detoxified, in order to evaluate the growth of yeast in the presence of furfural and acetic acid. Next, the pH of the hydrolysates and of the untreated ones was adjusted with active carbon; calcium carbonate (CaCO₃) was added, followed by a vacuum filtering and measured using a potentiometer (Ultra Basic UB-10 Denver Instrument). A 50 g sample of the syrupactive carbon was placed in 250 mL beakers, stirred at 150 rpm and 45 °C in an incubator. The samples were then centrifuged at 10 000 rpm for 10 min, and the concentrations of xylose, acetic acid, and furfural were analyzed before and after adding active carbon. The concentration of xylose and acetic acid were determined using High Performance Liquid Chromatography (HPLC HP-1100) and a Transgenomic ION-300 column, with a stone temperature of 45 °C, and an isocratic elution of 0.4 mL min^-1 of flow of mobile phase (H₂SO₄ 0,0025 M). The concentration of furural was measured using UV-vis (UV-1800 Shimadzu) spectrophotometry at a wavelength of 270 nm.

Microorganism and culture medium

The liophilized strain of D. hansenii NRRL Y-7426 was reactivated in Erlenmeyer beakers with100 mL of medium with xylose (30 g L^-1), peptone (5 g L^-1) and yeast extract (3 g L^-1), the pH was adjusted to 5.5, incubated at 28 °C, stirring at 150 rpm for 48 h. Afterwards, the production of xylitol in the media based on xylose, product of the hydrolyzed acid from the sorghum straw.

Production of xylitol

The production of xylitol by D. hansenii was evaluated in media with detoxified and non-detoxified hydrolysates with 30, 40 and 50 g L^-1 of xylose, at 30 and 35 °C and a stirring speed of 120 and 200 rpm; the concentration of xylitol was measured every 12 h up to 96 h of fermentation. The variables determined were yield or conversion factor (Y_P/S, g xylitol produced times g xylose consumed) and volumetric productivity (Qp, g xylitol produced per L of medium in 1 h). The concentration of xylitol was measured by HPLC using the same conditions as those mentioned for xylose.

With the data, a multi-factorial ANDEVA was carried out to determine the effect of the independent variables on the composition of the hydrolysate and production of xylitol. The differences between treatments were evaluated using a Fisher analysis for means comparison (p≤0.05).

Acid hydrolysis of RB-Paloma sorghum straw and detoxification

Treatment TH3 (H₂SO₄ at 6 %, 120 °C for 80 min and solid-liquid ratio 1:6) produced the greatest concentration of xylose (63.82 ±0.78 g L^-1), while the lowest concentration was obtained in treatment TH7 with 23.09 ±4.56 g L^-1, with 2 % of H₂SO₄ at 120 °C for 80 min in a proportion of 1:10 (Table 3). This amount of xylose is similar to that reported by Téllez-Luis et al. (2001) with a concentration of 16 g L^-1 of xylose with the acid hydrolysis of red sorghum hay with H₂SO₄ at 6 %, for 60 min and 120 °C. ^{Herazo et al. (2009)} hydrolyzed rice straw with H₂SO₄ at 2 % for 30 min and obtained 32.5 g L^-1 of reducing sugars and 9.9 g L^-1 of xylose.

For the growth of D. hansennii and production of xylitol, we used the hydrolysis conditions of treatment TH9 because it produced less furfural and acetic acid. These compounds derive from the degradation of lignin and act as inhibitors of microbial growth by reducing the metabolism (^{Mussatto and Roberto, 2004}; ^{Oliva et al., 2006}; ^{Pereira et al., 2011}); consequently, it is important to have low values of these toxic compounds. Fermentations were carried out in detoxified and non-detoxified media in order to evaluate the growth of yeast in the presence of inhibiting compounds, and therefore reduce production costs and time.

Detoxification

With the process of detoxification with active carbon, acetic acid and furfural decreased by 72 and 65.5 %, respectively (Table 4). Kamal et al. (2011) performed acid hydrolysis on palm trunk, then detoxified the syrups obtained with 1 and 2.5 % of active carbon, and the maximum reduction found was 58 % for furfural with 2.5% of active carbon. ^{Villarreal et al. (2006)} performed an acid hydrolysis on eucalyptus residues, and reported a reduction of 100 and 10.6 % of furfural and acetic acid, respectively, using 5 % active carbon at a pH of 5.5.

In order to evaluate which of the two factors (concentration of acid, solid-liquid ratio, or both) has a greater effect on the concentration of xylose obtained, a two-way ANOVA was carried out with a confidence level of 95 %. The result showed that the percentage of acid was the main factor in the hydrolyses performed (p≤0.05), and in this study, the solid-liquid ratio shows no significant difference (p>0.05) between treatments (Table 5).

Production of xylitol in detoxified and nondetoxified media

The highest concentration of xylitol in the media detoxified with active carbon was obtained in TH8 (40 g xylose, 35 °C, 200 rpm) with 28.8 g L^-1 after 48 h of fermentation; whereas in non-detoxified media, TH11 (30 g xylose, 35 °C, 150 rpm) produced a maximum concentration of 29.23 g L^-1 after 48 h. The concentration of xylitol was similar in the media with detoxified and non-detoxified hydrolysate. Although ^{Carvalheiro et al. (2005)} mentioned that xylitol production is favored by high concentrations of xylose, in our study, the lowest concentration of xylitol was in the medium with 50 g L^-1 of the carbon source. ^{Carvalheiro et al. (2005)} carried out detoxification treatments in hydrolyzed acids from rice straw, and found that the use of ionic exchange resin eliminated the highest concentration of furfural, although the media designed with these detoxified hydrolysates had a low xylitol yield (0.51 g g^-1). This may be due to the fact that these processes eliminate inhibiting compounds as well as other compounds, such as minerals that can serve as growth factors for yeast.

The highest concentration of xylitol was 29.23 g L^-1 (Table 6) and was obtained using a nondetoxified hydrolysate, although the detoxification process promotes the ability of fermentation of microorganisms in media based on lignocellulosic hydrolysates (^{Alriksson et al., 2011}; ^{Cavka and Jönsson, 2013}). Treatments 1 to 8 are fermentations with detoxified hydrolysates, while treatments 9 to 19 are non-detoxified. After reaching the maximum production of xylitol, the concentration of xylose decreases, due to its conversion into xylitol, since the latter is a primary metabolite, meaning it is related to the growth of yeast; that is, D. hansenii uses xylose as a source of carbon to grow and therefore the relation xylose-growth is inversely proportional. Therefore, it is important to establish the optimum time of fermentation, because when the microorganism enters a stationary phase, the concentration of the metabolite decreases.

Some of the most common detoxification methods include adjusting the pH with calcium carbonate, followed by adsorption with active carbon (^{Carvalheiro et al., 2005}). ^{Greetham et al. (2016)} mention that a low concentration of acetic acid exerts an antagonistic effect on furfural, which helps yeast grow under these stress conditions.

^{Villarreal et al. (2006)} performed a hydrolysis with hemicellulosic eucaliptus (Eucalyptus grandis) and they used a strain of Candida guilliermondii responsible for the conversion of xylose; the highest concentration of xylitol reached was 32.7 g L^-1 after 48 h of fermentation, with a yield of 0.57 g g^-1 and a Qp of 0.68 g L^-1 h^-1. Mussatto and Roberto (2001) hydrolyzed rice hay with 0.1 M H₂SO₄ at 121 °C for 20 min and a solid-liquid ratio of 1:10, and for the conversion of xylose by C. guilliermondii FTI 20037, they used 20 g L^-1 of xylose in the culture media and they obtained 0.72 g g-1 and a volumetric productivity of 0.61 g L^-1 h^-1.

Carvalheiro et al. (2006) evaluated the production of xylitol by D. hansenii in media based on acid hydrolysates from beer grains, they processed detoxified and non-detoxified hydrolysates with active carbon, and the values obtained for Y_P/S were 0.51 and 0.5 g g^-1 , and for Qp, 0.29 and 0.33 g L^-1 h^-1 in detoxified and non-detoxified media, respectively. These amounts are lower than those obtained in our study: top values of 1.16 g g^-1 and 1.13 g L^-1 h^-1 for Y_P/S and Qp respectively, after 24 h of fermentation in a non-detoxified medium; in a detoxified hydrolysate, values for Y_P/S and maximum Qp were 1.18 g g^-1 and 0.79 g L^-1 h^-1. The tendency in both studies was similar regarding higher volumetric productivity of xylitol from non-detoxified media. ^{Huang et al. (2011)} isolated the strain of C. tropicalis JH030 from residual sludge from a bioethanol-producing plant, they evaluated its growth and production of xylitol in non-detoxified rice hay media and reported a Y_P/S= 0.71 g g^-1. In our study, the media based on RB-Paloma white sorghum straw, still not detoxified, are a considerable amount of carbon for the D. hansenii growth process and its conversion from xylose to xylitol due to a good yield and volumetric productivity, which is higher to reports from other studies with lignocellulosic materials.

^{Wang et al. (2011)} used hydrolysates from maize ears and carried out fermentations at 150 and 200 rpm with 140 g L^-1 of xylose and a temperature of 30 °C; the highest amount of xylitol was obtained with 200 rpm, and a Qp of 2.12 g L^-1 h^-1 after 24 h of fermentation. Although the volumetric productivity is double than the one in our study, the concentration of xylose used is almost five times higher, which increases production costs.

Finally, with the highest concentrations of each fermentation treatment, a multifactorial ANOVA was carried out, with a 95 % confidence and no significant differences were found (Table 7). Therefore, these results help us deduce that there is no difference between using detoxified or nondetoxified hydrolysates, which means a potential increase in the production of xylitol with a reduction in time, cost, and materials, by eliminating the detoxification process in the hydrolysates.