Effectively Converting Cane Molasses into 2,3-Butanediol Using Clostridiumljungdahlii by an Integrated Fermentation and Membrane Separation Process

Abstract

Firstly, 2,3-butanediol (2,3-BDO) is a chemical platform used in several applications. However, the pathogenic nature of its producers and the expensive feedstocks used limit its scale production. In this study, cane molasses was used for 2,3-BDO production by a nonpathogenic Clostridium ljungdahlii. It was found that cane molasses alone, without the addition of other ingredients, was favorable for use as the culture medium for 2,3-BDO production. Compared with the control (i.e., the modified DSMZ 879 medium), the differential genes are mainly involved in the pathways of carbohydrate metabolism, membrane transport, and amino acid metabolism in the case of the cane molasses alone. However, when cane molasses alone was used, cell growth was significantly inhibited by KCl in cane molasses. Similarly, a high concentration of sugars (i.e., above 35 g/L) can inhibit cell growth and 2,3-BDO production. More seriously, 2,3-BDO production was inhibited by itself. As a result, cane molasses alone with an initial 35 g/L total sugars was suitable for 2,3-BDO production in batch culture. Finally, an integrated fermentation and membrane separation process was developed to maintain high 2,3-BDO productivity of 0.46 g·L^-1·h^-1. Meanwhile, the varied fouling mechanism indicated that the fermentation properties changed significantly, especially for the cell properties. Therefore, the integrated fermentation and membrane separation process was favorable for 2,3-BDO production by C. ljungdahlii using cane molasses.

Keywords: 2,3-butanediol; Clostridium ljungdahlii; cane molasses; membrane separation.

Figure 1

Effect of sugars (a), the ratio of fructose:glucose (b), and total sugars in molasses (c) on 2,3-BDO production. The basal medium was the same as the modified DSMZ 879 medium, while the sugar (i.e., fructose) was substituted by a different substrate in Figure 1. The fermentation was carried out in 250 mL screw-cap bottles. Data are given as the mean ± SD, n = 3.

Figure 2

Effect of fermentation medium on 2,3-butanediol production. The components of M1, M2, M3, M4, M5, and M6 were shown in Table 1. The fermentations were carried out in 250 mL screw-cap bottles. Data are given as the mean ± SD, n = 3.

Figure 3

Effect of batch culture on 2,3-BDO production. The cultures were carried out in a 2.7 L bioreaction. Data are given as the mean ± SD, n = 2.

Figure 4

KEGG functional classification (a) and COG functional classification (b) of the differential genes. The differential genes between fructose were used as the carbon source in the modified DSMZ 879 medium (control) and cane molasses alone was used as the fermentation medium.

Figure 4

KEGG functional classification (a) and COG functional classification (b) of the differential genes. The differential genes between fructose were used as the carbon source in the modified DSMZ 879 medium (control) and cane molasses alone was used as the fermentation medium.

Figure 5

Effect of fed-batch culture on 2,3-BDO production. The time courses refer to (a) an initial 35 g/L total sugars with a final concentration of 35 g/L total sugars in the broth from each fed-solution, and (b) an initial 55 g/L total sugars with a final concentration of 55 g/L total sugars in the broth from each fed-solution. The cultures were carried out in a 2.7 L bioreaction. Data are given as the mean ± SD, n = 2.

Figure 6

Effect of exogenous KCl (a), sugars (b), 2,3-BDO (c), and acetic acid (d) on 2,3-BDO production. The basal medium was the same as that in Figure 1c with 35 g/L sugars in molasses. In (b), the sugars in the horizontal axis mean the simulated sugars in molasses, which were further added to the initial 35 g/L real molasses solution in Figure 1c. In (c,d), the concentration of 2,3-BDO or acetic acid in the vertical axis means the final concentration of 2,3-BDO or acetic acid in the broth subtracted its initial concentration in the medium. The fermentations were carried out in 250 mL screw-cap bottles. Data are given as the mean ± SD, n = 3.

Figure 7

Kinetic curves of 2,3-BDO production in a membrane bioreactor with repeated-batch culture. Data are given as the mean ± SD, n = 2.

Figure 8

Determination of fouling mechanisms from model fitting to experimental data in Figure 7: (a) cycle 1, (b) cycle 2, and (c) cycle 3.