Chapter 7
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
Approaches To Deal with Toxic Inhibitors during Fermentation of Lignocellulosic Substrates T. L. Richardson, N. K. Harner, P. K. Bajwa, J. T. Trevors, and H. Lee* School of Environmental Sciences, University of Guelph, Guelph, ON N1G 2W1, Canada *
[email protected] The recalcitrance of lignocellulose materials requires harsh pretreatment(s) to release sugar monomers for ethanol fermentation by microorganisms. Harsh pretreatment conditions result in the formation of compounds in the hydrolysate that are inhibitory to the fermenting microorganism(s). The main inhibitors include furan derivatives, organic acids, and phenolic compounds. Research has focused on physical, chemical and/or biological methods to deal with inhibitors. Many of these methods remove or convert inhibitors through a detoxification step prior to fermentation. Recently, biological methods have focused on increasing yeast inhibitor tolerance. This review examines research on the different methods used to detoxify hydrolysates or increase yeast tolerance to inhibitors.
Introduction Concerns over the effect of greenhouse gas (GHG) emissions and rising oil prices have led to efforts to develop clean and/or renewable fuel sources. Many industrialized countries, including Canada, have committed to the Kyoto Protocol which requires that, by 2012, emissions of carbon dioxide and other GHG be reduced to 6% below 1990 levels. In Canada, the transportation sector contributes 27% of the total GHG emissions (1). The Canadian government has mandated 5% renewable fuel content in gasoline and 2% content in diesel and home heating
© 2011 American Chemical Society In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
fuels by 2011. In total, 2.6 billion litres of renewable fuel will need to be produced with the goal of reducing GHG emissions by 4 million tons/year (2). Ethanol is an appealing alternative energy source that burns more completely than gasoline (3), making it both a substitute and additive for petroleum-derived fuels. Adding ethanol to gasoline oxygenates the fuel, allowing more complete combustion and reduced emission of pollutants. In Canada, gasoline mixtures containing up to 10% ethanol are typically used. In Canada, the storage carbohydrates in agricultural crops such as corn and wheat are used as substrates for bioethanol production (4). These crops are traditionally grown for food and animal feed, and may not be sustainable for large-scale ethanol production (5, 6). In contrast, vast quantities of sugars occur as structural polysaccharides in the form of cellulose and hemicellulose in lignocellulosic biomass residues. Lignocellulosic residues are more abundant than food crops, have fewer competing uses, and can be harvested with considerably less interference to the food economy and less impact on environmental resources. Using today’s technology, the bioconversion of lignocellulosic sugars to ethanol is comparatively difficult and expensive relative to starch conversion. Thus, there is interest to refine existing technology or develop new technologies to produce the second generation “lignocellulosic” ethanol to provide the large volume needed to satisfy the demand for the transportation sector. Lignocellulosic substrates envisioned for bioconversion include crop residues, municipal solid wastes, forestry residues and discarded material from paper mills. These low-value waste materials can be converted to produce ethanol. Lignocellulosic biomass is comprised of lignin (10-25%) as well as the structural carbohydrates hemicellulose (20-35%) and cellulose (35-50%) (7, 8). The composition of various lignocellulosic materials is summarized in Table I. Cellulose is a linear polymer of glucose molecules linked by β-1, 4 glycosidic bonds, making it highly crystalline, compact and resistant to biological attack (9). Hemicellulose is a branched heteropolymer composed of hexose sugars (glucose, mannose and galactose), pentose sugars (xylose and arabinose) and acids (glucuronic and acetic acids) (8). Xylose is the predominant sugar in hemicellulose and may comprise up to 25% of the dry weight in some biomass (8, 10). In the bioethanol industry, fermentation is typically conducted with the baker’s yeast Saccharomyces cerevisiae. This yeast utilizes hexose sugars in lignocellulosic substrates, but is unable to metabolize pentose sugars, despite the presence of a xylose transporter and all the enzymes needed for a full xylose-metabolic pathway (11). Both pentose and hexose sugars in potential lignocellulosic hydrolysates must be efficiently converted to ethanol to maximize the economic feasibility of a commercial bioconversion process. Native pentose-fermenting yeasts such as Candida shehatae, Candida tropicalis, Pachysolen tannophilus and Scheffersomyces (Pichia) stipitis can utilize the dominant pentose and hexose sugars and convert them to ethanol. However, when presented with lignocellulosic hydrolysates, this conversion becomes inefficient for several reasons. First, native pentose-fermenting yeasts suffer from hexose repression and inactivation (10, 12). When glucose and xylose are present, preferential utilization of hexose sugars occurs and pentose 172 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
metabolism is inhibited (10, 13). Second, they convert little substrate to ethanol in the presence of inhibitors arising from the pretreatment of lignocellulosic substrates, thereby limiting their fermentation performance (14). Third, they suffer from low ethanol tolerance which limits the amounts of ethanol that can be accumulated in the medium (15). To illustrate the problem, in a study of spent sulfite liquor (SSL) fermentation (16), S. cerevisiae was found to produce a greater ethanol yield (0.38 g/g total sugar) compared to Pachysolen tannophilus (0.12 g/g total sugar), Pichia stipitis (0.17 g/g total sugar) and Candida tropicalis (0.21g/g total sugar), even though the latter three ferment both pentose and hexose sugars. Thus, there is considerable scope for improving of the ability of the native pentose-fermenting yeasts to ferment potential lignocellulosic substrates.
Inhibitors Formed from Lignocellulosic Substrate Pretreatment To yield fermentable sugars, lignocellulosic substrates are subjected to a pretreatment process which beak apart the lignin, hemicellulose and cellulose matrix. Examples of pretreatment processes include acid hydrolysis, alkaline hydrolysis, steam explosion (autohydrolysis), hydrothermal treatments, organosolv, ammonia freeze explosion (AFEX), ozonolysis, wet oxidation (WO) and carbon dioxide explosion processes.
Table I. Composition of selected lignocellulosic biomass Lignocellulosic Material
Cellulose (%)
Hemicellulose (%)
Lignin (%)
Reference
41-50
11-33
19-30
(17)
Birch
40
39
21
Apsen
51
29
16
39-53
19-36
17-24
(17)
43
26
29
(18)
24-50
12-38
6-29
(17)
45
30
12
(19)
Corn stover
40
25
17
Wheat straw
30
50
20
Hardwood
Softwood Spruce Herbaceous plants Switchgrass
(18)
Agricultural residues (19)
173 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
A number of inhibitory compounds are formed in lignocellulosic hydrolysates from the harsh pretreatment processes. The main inhibitors include furan derivatives, organic acids and phenolics. These compounds adversely affect the viability, growth, and fermentative ability of yeasts (14, 17), thereby limiting the efficiency and economic feasibility of the bioconversion process. Generally, native pentose-fermenting yeasts are more sensitive to inhibitors than S. cerevisiae (18). The reason(s) for this is not known. Acetic acid is a weak acid released from the acetyl groups of acetylated xylan (20, 21). A concentration of 5 g/L can be inhibitory to S. stipitis and P. tannophilus (22) and hydrolysates may contain higher concentrations. For example, hydrolysates of sugarcane bagasse and corn stover may contain 10.4 g/L (23) and 13 g/L (24) of acetic acid, respectively. During dilute acidic pretreatment, furfural and hydroxymethylfurfural (HMF) are formed from the dehydration of pentose and hexose sugars, respectively. Furfural can be further degraded to formic acid, while HMF can be broken down to levulinic and formic acids (25). Various phenolic compounds such as vanillin, syringaldehyde and 4-hydroxybenzaldehyde are formed from lignin degradation (26). The phenolics are typically found at low concentrations in hydrolysates (27, 28). Nevertheless, when the inhibitors are present together, they may act synergistically, so even low concentrations of the inhibitors may contribute to the overall toxicity.
Mechanisms of Inhibition Organic Acids Acetic, formic and levulinic acid (Table II) are the most common weak acids found in lignocellulosic hydrolysates. The intracellular pH of yeast cells must be maintained within a physiological range for metabolic processes to proceed. The undissociated forms of these weak acids are lipid soluble and diffuse across the plasma membranes. Within yeast cells, they dissociate due to the higher intracellular pH (7.0 to 7.2), releasing protons and lowering the intracellular pH (14, 29). To neutralize the intracellular pH and maintain homeostasis, the cell requires ATP to transport excess protons out via a plasma membrane ATPase (29, 30). To generate the ATP required, sugars are utilized for energy and subsequent ethanol production and diverted from cell growth (31). Cytoplasmic acidification and cell death result from high levels of weak acids depleting ATP reserves, hampering cell growth and removal of protons (32). The concentration of undissociated acids in hydrolysates is highly pH dependent, and a pH of at least 5.5 improves fermentation by reducing the toxicity of weak acids (20, 33). Raising the pH above 5.5 in hardwood SSL can substantially reduce acetic acid toxicity by reducing the undissociated acetic acid concentration to less than 1.5 g/L (34).
174 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
Table II. Common organic acids found in selected lignocellulosic hydrolysates Organic acids
pKa (26)
Acetic acid
Chemical structure
Concentration in hydrolysates (g/L) Spruce (30)
Olive tree cuttings (35)
Yellow poplar (36)
Aspen Wood (37, 38)
4.75
2.4
6.8
7.1
6.1
Formic acid
3.75
1.6
2.5
ND
ND
Levulinic acid
4.66
2.6
ND
ND
ND
ND: No data available
Furan Derivatives Furfural and HMF (Table III) are chemically related compounds containing a furan ring and aldehyde group. Furan derivatives can be metabolized by some yeast such as S. cerevisiae during fermentation (39, 40). Furfural and HMF inhibit yeast growth and fermentation by interfering with glycolytic enzymes such as glyceraldehyde-3-phosphate dehydrogenase, alcohol dehydrogenase (ADH) and hexokinase. The greatest inhibitory effect of 50% was reported on purified triosephosphate dehydrogenase activity in S. cerevisiae by 1 g/L of furfural (41). Therefore, furfural adversely affects several key enzymes required for metabolism. Inhibition of ADH contributes to intracellular accumulation of acetaldehyde which at concentrations above 0.5 mM inhibit cellular functions such as DNA and protein synthesis, and can cause a lag in the growth of S. cerevisiae when furfural is present (26, 33, 42). In S. cerevisiae furfural causes reactive oxygen species to accumulate, vacuole and mitochondrial membranes damage, chromatin and actin damage (Almeida et al., 2007). Recently, Li and Juan (2010) also showed that furfural affects the biosynthesis of glycerol and metabolism of several important biochemicals at transcriptional level in S. cerevisiae.
175 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Table III. Furan derivatives found in selected lignocellulosic hydrolysates
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
Furan Derivatives
Chemical structure
Concentration in hydrolysates (g/L) Spruce Alder (46) (47)
Yellow poplar (36)
Corn stover (48)
Hardwood spent sulfite liquor (49)
Furfural
0.4
2.73.2
0.29
1.31.5
0.2
HMF
1.4
2.64.5
0.18
0.2
ND
ND: No data
Furfural decreases the specific growth rate of S. cerevisiae and inhibits growth at a concentration of 5 g/L (43). In the presence of 2.8 g/L furfural, the cell-mass yield of S. cerevisiae on glucose decreased from 0.06 to 0.05 g/g (42). Growth of Candida guilliermondii was reduced by 1 g/L of furfural and 1.5 g/L of HMF (44). The inhibitory effects of HMF are similar to but somewhat weaker than those of furfural (44). Both furfural and HMF inhibited CO2 production by S. cerevisiae in a synthetic medium containing 5% glucose (45) and presumably also ethanol production during fermentation. Furfural can act synergistically with other inhibitors, such as acetic acid on S. cerevisiae (50) and phenols on Escherichia coli (51) to inhibit fermentation. Though concentrations of a specific inhibitor may be too low to cause adverse effects, toxicity may be enhanced by the presence of other inhibitory compounds. The growth of S. cerevisiae was more inhibited by 2 g/L each of HMF plus furfural than by 4 g/L of either compound alone (52). In another study, furfural (0.3 g/L) and HMF (0.9 g/L), when combined, decreased cell growth, ethanol production, and slowed xylose utilization in P. tannophilus and S. stipitis. The inhibitory effects were more severe with both compounds present in the medium than each individually (14).
Phenolic Compounds Weakly acidic phenolic compounds listed in Table IV contain a hydroxyl group bonded to an aromatic hydrocarbon. The inhibitory mechanisms of these compounds are not well understood (26, 33, 38, 53). Many studies investigating the toxicity of phenols have been performed using concentrations significantly higher than those present in hydrolysates (54). It was suggested that low MW phenolic compounds partition into the lipid bilayer of biological membranes to adversely affect membrane structure and function, and hinder the ability of membranes to act as a selective permeability barrier (55).
176 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Table IV. Some common phenolic compounds found in lignocellulosic hydrolysates Concentration in hydrolysates (g/L)
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
Phenolic Compounds
Aspen wood (37)
Poplar (57)
Red oak (58)
Spruce (59)
Vanillin
0.23
0.003
0.09
0.096
Syringaldehyde
ND
0.002
0.213
ND
4-Hydroxybenzaldehyde
0.82
ND
ND
0.002
Chemical Structure
ND: No data available
Vanillin is formed from the degradation of the guaiacylpropane units of lignin and may be present in spruce, poplar, willow and pine hydrolysates (17). A low concentration of vanillin (1 g/L) was found to be toxic to S. stipitis, C. shehatae, Zymomonas mobilis and S. cerevisiae, resulting in almost complete inhibition of growth and ethanol production (54). Growth of C. guilliermondii was more severely inhibited by 1 g/L of vanillin compared to 1 g/L of syringaldehyde, and no growth occurred at 2 g/L of vanillin (55). The toxicity was correlated with the compound’s octanol/water partition coefficient which is a measure of hydrophobicity (56) and suggests a membrane site of action.
Strategies To Overcome Inhibition Physical, chemical and biological methods have been used to overcome inhibitors present in lignocellulosic hydrolysates (Table V). Many of these methods focus on removing or converting inhibitors through a detoxification step prior to fermentation. Recently, biological methods have focused on increasing yeast tolerance to inhibitors, thereby obviating the need for a detoxification 177 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
step. It would be advantageous to not have a detoxification step as it represents a significant portion of production cost. For example, the addition of calcium hydroxide and sodium sulphite to a willow hydrolysate has been reported to account for 22% of the production costs for ethanol (60).
Physical Methods
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
Vacuum Evaporation Vacuum evaporation is used to remove liquid and volatile compounds from samples. The procedure consists of reducing the pressure in a fluid filled container below that of the vapor pressure of the liquid, causing the liquid to evaporate along with volatile compounds. Vacuum evaporation has been used to remove inhibitors present in lignocellulosic hydrolysates (30, 37). In research comparing the efficiency of alkali treatment, sulfite treatment, vacuum evaporation, anion exchange, treatment with laccase, and treatment with Trichoderma reesei to detoxify a dilute-acid spruce hydrolysate, vacuum evaporation was observed to be one of the least efficient methods (30). The spruce hydrolysate contained 5.9 g/L HMF, 1.0 g/L furfural, 2.6 g/L levulinic acid, 2.4 g/L acetic acid, 1.6 g/L formic acid, and fourteen different phenolic compounds. In a study involving vacuum evaporation, either 10 or 90% of the initial volume was removed, and the concentration of the non-volatile compounds in the hydrolysate was restored by adding water to 100% of the initial volume. Evaporation of 90% of the initial hydrolysate volume was better at eliminating inhibitors than evaporation of 10% of the initial volume. When 90% of the hydrolysate was evaporated, 65% of acetic acid, 74% of formic acid, 100% of the furfural and 4% of the HMF were removed. When 10% of the spruce hydrolysate was evaporated, only 40% of the furfural was removed. Similar results were observed in a different study examining vacuum evaporation of dilute-acid aspenwood hemicellulose containing 6.1 g/L acetic acid and 0.28 g/L furfural (37). In this study evaporation of the hydrolysate to near dryness at 55˚C removed all of the furfural and 46% of the acetic acid. However, vacuum evaporation did not remove the phenolic compounds. Evaporation of 90% of the hydrolysate also resulted in better fermentation results with S. cerevisiae than when 10% was evaporated. The ethanol yield for 90% evaporation-treated spruce hydrolysate was 0.42 g/g, compared to 0.32 g/g, and 0.34 g/g for untreated spruce hydrolysate and 10% evaporation-treated spruce hydrolysate, respectively. Vacuum evaporation, unlike some other physical detoxification methods (ie., anion exchange), does not result in a decreased sugar content (30). However, these results along with other studies (61) suggest that volatile compounds in lignocellulosic hydrolysates are not the major inhibitors. Larsson et al. (1999) pointed out the possibility of combining laccase treatment and evaporation, as the laccase enzyme would remove phenols, and evaporation would remove furfural. However, it is desirable to eliminate or reduce detoxification steps for economical industrial production of ethanol.
178 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
Solvent Extraction Solvent extraction of lignocellulosic hydrolysates as a detoxification method has been examined (37, 62, 63). Ethyl acetate was used to extract low MW phenolics from an enzyme pretreated hydrolysate of aspen wood prior to fermentation by S. stipitis (37). The aspen wood hydrolysate contained 20.2 g/L xylose, 2.6 g/L glucose, 5.3 g/L acetic acid, 0.05 g/L furfural, 0.82 g/L hydroxybenzoic acid and 0.21 g/L vanillin. Ethyl acetate extraction at a 1:1 (v/v) ratio of ethyl acetate to aspen wood hydrolysate reduced the acetic acid to 2.7 g/L, and completely removed the furfural, hydroxybenzoic acid, and vanillin. No ethanol was produced from fermentation of untreated aspen hydrolysate. Fermentation of ethyl acetate extracted aspen hydrolysate gave an ethanol yield of 0.47 g/g. One disadvantage of liquid extraction is the requirement of a large volume of solvent, and distillation would be required to recycle the solvent, adding to production cost and time.
Table V. Common detoxification methods to increase the fermentability of lignocellulosic hydrolysates Method
Description
Reference
Vacuum evaporation
Evaporation of volatile inhibitors, such as furfural, acetic and formic acids.
(30, 37, 61)
Solvent extraction
Removal of phenolic compounds.
(37, 62, 63)
Ion exchange resins
Adsorb aliphatic acids, furans and phenolic compounds. Ion exchange resins are expensive.
(30, 46, 64, 65)
Activated charcoal
Adsorb furans, phenolic compounds, and acetic acid.
(64, 66, 67)
Lignin residue
Adsorb furans and phenolic compounds.
(68)
Addition of an alkali base to adjust pH to 9-10, followed by readjustment of pH to 5.5. Inhibitor removal is though chemical conversion. Loss of sugars has been reported.
(69–79)
Bioabatement
Treat hydrolysate with microorganisms that utilize or convert inhibitors to less toxic compounds.
(80–84)
Laccase enzyme treatment
Removal of phenolic compounds.
(64, 85–88)
Physical Methods
Chemical Method Alkali treatment
Biological Methods
Continued on next page.
179 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
Table V. (Continued). Common detoxification methods to increase the fermentability of lignocellulosic hydrolysates Method
Description
Reference
Fed-batch fermentation
Allows in-situ detoxification of inhibitors by fermenting yeast by keeping inhibitor concentration low through control of substrate feed rate.
(89–93)
Continuous fermentation with cell retention system
In-situ detoxification coupled with a cell retention system to maintain high cell density. Faster fermentation times than with fed-batch.
(94–101)
Yeast adaptation
Utilizes yeasts adaptation ability to increase inhibitor resistance.
(102–106)
Recombinant yeast
Molecular techniques are used to create recombinant yeast with increased resistance to inhibitors
(6, 107–114)
Evolutionary engineering of yeast
Development of inhibitor tolerant yeast through iterative genetic diversification and functional selection.
(115, 116)
Removal of Charged Inhibitors by Ion-Exchange Resins Ion-exchange resins have been used to detoxify lignocellulosic hydrolysates (30, 46, 65). Ion-exchange involves exchanging an ion from solution for a similarly charged ion attached to an immobile solid particle. Both cation and anion exchange resins have been used to remove inhibitors in hydrolysates. In one study, a dilute-acid spruce hydrolysate containing 5.9 g/L HMF, 1.0 g/L furfural, 2.4 g/L acetic acid, 2.6 g/L levulinic acid, 1.6 g/L formic acid, and 14 different phenolic compounds was treated with polystyrenedivinylbenzene-based anion-exchange resins at either pH 5.5 or 10 for 1 h in a batch procedure (30). The anion exchange resins were effective at removing the aliphatic acids, furan derivatives and phenolic compounds from the spruce hydrolysate. Anion exchange at pH 10 removed 70% of the HMF, 73% of the furfural, 97% of the formic acid, 96% of the acetic acid, and 93% of the levulinic acid from the spruce hydrolysate. The spruce hydrolysate contained 32.3 g/L of fermentable sugars. After treatment with anion exchange the hydrolysate was fermented with S. cerevisiae, resulting in an improved ethanol yield (0.49 g/g), volumetric productivity (1.42 g/L/h) and biomass yield (0.08 g/g) in comparison to the untreated hydrolysate’s ethanol yield (0.32 g/g), volumetric productivity (0.04 g/L/h) and biomass yield (0.01 g/g). Although, anion exchange at pH 10 was effective at removing inhibitors and improving hydrolysate fermentation, it resulted in a 26% decrease in sugar content. However, other studies using hydrolysates pretreated with anion exchange resins did not report any decrease in sugar content (46, 64, 65). 180 In Sustainable Production of Fuels, Chemicals, and Fibers from Forest Biomass; Zhu, J., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by MICHIGAN STATE UNIV on July 2, 2013 | http://pubs.acs.org Publication Date (Web): July 11, 2011 | doi: 10.1021/bk-2011-1067.ch007
The presence of counter ions, such as sulfate, was found to be important in preventing sugar loss when using ion exchange (65). Anion exchangers were found to have the ability to trap glucose monomers, but the addition of sulfate ions resulted in no significant loss of sugars (65). A comparison of a cation exchanger, anion exchanger, and a resin without charged groups at either pH 5.5 or 10 was done to determine which was most effective at removing inhibitors from a dilute-acid spruce hydrolysate (65) which contained 7.97 g/L aliphatic acids, 5.10 g/L HMF, 0.82 g/L furfural and 3.7 g/L phenolic compounds. Treatment with an anion exchanger at pH 10 was found to be the most effective at removing phenolic compounds, furan aldehydes and aliphatic acids. Anion exchange at pH 10 removed 96% of the aliphatic acids, 65% of the HMF, 68% of the furfural, and 81% of the phenolic compounds. For all three types of resins, greater improvement in fermentation was seen when treatment was conducted at pH 10 rather than pH 5.5. Fermentation of spruce hydrolysate treated by anion exchange resins at pH 10 by S. cerevisiae resulted in an ethanol yield of 0.46 g/g, compared to 0.06 g/g for the untreated hydrolysate (65). Horvath et al. (2004) reported that treatment with anion exchange resins made from styrene based matrices containing strongly basic functional groups improved fermentation of dilute-acid spruce hydrolysates and improved ethanol production by 7-fold compared to fermentation of untreated hydrolysates. The hydrolysate originally contained 15.7 g/L glucose, 12.8 g/L mannose, 1.4 g/L HMF, 0.4 g/L furfural, 1.4 g/L levulinic acid, 3.2 g/L acetic acid, 1.1 g/L formic acid, and 2.4 g/L phenolic compounds. Treatment with the anion exchange resins made from a styrene based matrix containing strongly basic functional groups removed 90% of the levulinic acid, 72% of the acetic acid, 73% of the formic acid, 38% of the furfural and 29% of the HMF. S. cerevisiae was used to ferment the hydrolysate. Treatment with resins containing weak basic functional groups led to the least improvement in ethanol productivity from fermentation of treated hydrolysate. Ion exchange resins represent an efficient method to detoxify lignocellulosic hydrolysates, and result in improved fermentation and ethanol yields. However, the use of ion exchange resins is costly and may be uneconomical and impractical for large-scale ethanol production.
Removal of Inhibitors by Activated Charcoal A few studies have examined the use of activated charcoal or carbon to remove inhibitors from lignocellulosic hydrolysates. Activated carbons are less expensive than ion exchange resins and can be regenerated using steam (66). Activated carbon has been successfully used to reduce acetic acid concentration from 10 to 4 g/L in a synthetic hydrolysate shaken at 250 rpm (67). However, activated carbon was found to be marginally successful at removing acetic acid from dilute acid corn stover hydrolysate (66). Five rounds of activated carbon adsorption were required to bring acetic acid levels from an initial concentration of 16.5 g/L to below 2 g/L. Negligible amount of glucose (