5.1. The chemical pathway

Oxidation of xylose has been numerously reported in the literature although focus interest, both on the biological as well as chemical point of view has been focused toward a simple oxidation of xylose to xylonic acid (see Figure 12).

Oxidation of xylose has been reported for a variety of different metallic catalyst including gold for high conversion rates [171]. Using a process performed a little higher than room temperature in a basic pH for 1 hour, they were able to reach a 78 % conversion of xylose to xylonic acid. Using comparable catalyst, Pruesse et al. [172] were able to reach 99 % selectivi­ty with a conversion rate of 21 mmol/min/g (Au) in a continuous reactor. Nevertheless, con­trarily to Bonrath, Pruesse and co-worker used a mixture of gold and palladium to perform this oxidation and temperature slightly higher (60 °C as compared to 40 °C).

Copper has also been indirectly investigated for the conversion of xylose to xylonic acid in that sense that Van der Weijden et al. [173] used C5 sugars (including xylose) for the reduc­tion of copper sulfate in wastewater with very promising results. Although emphasis was not put on the carbohydrate itself, results showed that the reduction of copper from (II) to elemental was possible yet economical at larger scale. Xylonic acid was also observed as by­product of xylose oxidation using chlorine, as a side reaction of lignin oxidation. In this work [174], the concentration of xylonic acid increased by a factor of 40 after the chlorination process. Interesting enough, the xylitol concentration also increased, which might lead to the conclusion that oxidation, was probably not the sole factor here and that side reactions as the Cannizarro reaction between two xylose molecules could have been occurring. Jokic et al. [175] showed that it was possible up to an efficiency of 80 % to convert xylose simultane­ously to xylonic acid and xylitol using electrotechnologies. Such process could be to a cer­tain extent compared to the Cannizarro reaction where the original aldehyde is acting as redox reagent.

Further oxidation of xylose leads to a trihydroxydiacid, more specifically xylaric acid as de­picted in Figure 13 below.

Conversion of C5 sugars and to a smaller extent of xylose into aldaric acids has been descri­bed in literature in a few reports. Kiely et al. [176] reported that a conversion up to 83 % xy­lose into 2,3,4-trihydroxyglutaric acid was achievable in a reaction mixture composed of nitric acid and NaNO2. The side product of this reaction was reported to be disodium tetra — hydroxysuccinate. Conversion of xylose to xylaric adic was also reported [177] using oxygen under a platinum catalyst all of this in an alkali promoted medium. Comparable conversion process [178] was obtained without any alkali, though still performed the reaction in water at 90 °C under 75 psi of oxygen. The conversion for this process was 29 %. Fleche et al. [179] reported a maximum conversion of 58% once again using platinum supported on alumina.

Severer oxidizing conditions leads to a breakage of the carbon-carbon bonds in the carbohy­drate molecule leading to the production, mostly, of small organic acids as formic and acetic acid on glucose [180]. A simplified scheme of such a reaction is presented in Figure 14 below:

An example of sever oxidation of xylose in a mixture of hydrogen peroxide and ammonium hy­droxide have been recently reported [181] with a conversion of 96 % at room temperature for 1 h. Similar conversion of xylose was reported [182] for a process using oxygen and a molybde­num and vanadium catalyst. The reaction was done for 26 h at 353 K and 30 bar for a conversion of up to 54 % into formic acid with carbon dioxide as by-product.

5.2. The biological pathway

Xylonic acid synthesis from xylose has been reported for Acetobacter sp. [183], Enterobacter cloacea [184], Erwinia sp. [185, 186], Fusarium lini [187], Micrococcus sp. [188], Penidllium cory- lophilum, Pichiaquer cuum [185], Pseudomonas sp. [189, 190], Pullularia pullulans [191], Glucono — bacter and Caulobacter [192, 193].

In metabolic pathways, xylose is converted to xylonate via 2 key enzymes. First, a xylose de­hydrogenase (XD) oxidizes xylose to D-xylono-1,4-lactone (xylonolactone) using either NAD + or NADP+ as cofactor. This reaction is followed by the hydrolysis of xylonolactone to xylo­nate either spontaneously or by an enzyme with lactonase activity [194, 195]. It is hypothe­sized that Pseudomonas and Gluconobacter sp. both carry a membrane-bound pyrroloquinoline quinine (PQQ)-dependent XD and a cytoplasmic one [195, 196]. Stephens et al. [193] recently proposed a full xylose catabolic pathway for C. crescentus. Note that a similar pathway was proposed for arabinose yielding L-arabonate [197]. As shown in Figure 15, the proposed metabolic pathway for C. crescentus shows that xylonate is an intermediate in catabolic reactions that is quite different from the XI or XR/XDH previously discussed which were more intensively studied.

Researches on highly efficient microbial xylonic acid production are scarce compared to bio­fuels or xylitol. Even if the identification of xylonate producing species began as early as 1938 [187], the first attempt to isolate a possible industrial biocatalyst was done by Buchert et al. [185], who identified P. fragi ATCC4973 as a potentially high efficiency xylonate producer (92 % of initial sugar converted to xylonic acid with initial xylose concentration of 100 g/L). In further work, P. fragi and G. oxydans showed yields of over 95 % but the low tolerance of those native strains to inhibitors tends to be problematic for industrial uses [192]. As dis­cussed above, the metabolic pathways implied by xylonate have been investigated in the re­cent years [193,196]. The first recombinant microorganism engineered for the industrial production of xylonate was done by Toivari et al. [198]. By introducing the heterologous Tri — choderma reesei xydl gene (coding for the NADP+ dependant XD) in S. cerevisiae, they were able to obtain up to 3.8 g/L xylonate with 0.036 g/L/h productivity and 40 % yield. Nygard et al. [195] engineered K. lactis by introducing T. reesei xydl and deleting the putative xyll gene coding for the XR. Up to 19 g/L xylonate where produced when grown on a xylose (40 g/L) and galactose (10.5 g/L) medium. The native ability of fast xylose uptake was an advantage, but high intracellular xylonate concentration was observed, which may indicate difficulties with product export. Liu et al. [199] used similar approach engineering E. coli by disrupting the native xylose metabolic pathways of XI and XK (as shown in Figure 16). The native path­way of xylonate was also blocked by disrupting xylonic acid dehydratase genes. The XD from C. crescentus was introduced and 39.2 g/L of xylonate from 40 g/L of xylose in minimal medium was obtained at high productivity 1.09 g/L/h. From these results it is clear that re­search is at its genesis and significant efforts will be required for the creation of a highly pro­ductive and effective xylonate production biocatalyst.

At this point it is rather hard to verify the potential or the economic value of oxidation prod­ucts from xylose. Complete oxidation to formic acid could be the most suitable approach at this point since the market for xylonic and xylaric acid is not as well defined as for the sim­ple methanoic acid with its actual market value between 750-950 USD/tonne [200] and an annual world demand suspected to reach 573 Ktonnes in 2012 [201]. Conversion of xylaric acid into glutaric acid (pentanedioic acid) would lead to a very interesting market as a plas­ticizer but dehydration or reduction of the three central hydroxyl groups may be a challenge that could be winning at lab scale although a multiple synthesis pathway would be very dif­ficult to reach economic at an industrial level.

D-xylonic acid

Figure 16. D-xylose and D-xylonic acid metabolic pathways in E. coli. The symbol X denotes that the gene is disrupted.

6. Conclusion

Second-generation ethanol or "cellulosic ethanol" relies on the utilisation of lignocellulosic biomass as a source of carbohydrates via the "bio" conversion route (keeping in mind that other pathway, as thermocatalytic pathways, may also lead to cellulosic ethanol). Produc­tion of ethanol thus requires isolation of cellulose from lignocellulosic matrix, then hydroly­sis of cellulose to glucose prior to fermentation. Both of the previously mentioned steps represent challenges for industry, but the whole economic of the process is perhaps the most challenging part of cellulosic ethanol production. Cellulose is usually available in lignocellu­losic biomass in the 45-60 % range which, assuming a perfect conversion implies production of 300-400 L/tonne of lignocellulosic biomass processed. At an actual price of 0.48 USD/L, each ton of biomass has a potential value of about 150-200 USD/tonne of biomass processed.

The conversion of lignocellulosic biomass is rather more complex and to a certain extent more expensive than starch-based feedstock as corn and therefore, one can assume that the
conversion price is going to be higher than classical or first generation ethanol production. Keeping that fact in mind, the conversion of cellulose to glucose itself is a major technologi­cal challenge since it either requires enzymes, ionic liquids or strong acids that are rather ex­pensive to buy or expensive to recycle and since it is of outmost importance for the production of the ethanol, technology is to a certain extent limited by this reality.

The remaining carbon content of lignocellulosic biomass is also an important factor to be considered. Since the maximum production of ethanol from the total feedstock could vary around 300-400 L per tonne, there is at this point a necessity to generate co-products from the biomass in order to make this whole process economic at the end thus coping for techno­logical problem as conversion of cellulose to glucose. Lignin is one of the most abundant macromolecule on earth bested only by cellulose. The aromatic nature of lignin is a chal­lenge for ethanol production but not for added value compounds as aromatic monomers that could displace actual monomers used in the polymer industry that are usually obtained from non-renewable materials.

Hemicelluloses are also an important part of the lignocellulosic biomass. Hemicelluloses, contrarily to cellulose that is characterized by an amorphous and a crystalline part, are high­ly ramified and easy to hydrolyse. Usually, a simple diluted alkali solution, acidic solution or even hot water can allow conversion of hemicellulose to simple sugars. The major prob­lem with hemicellulose is the heterogeneous composition including but not limited to small acids and a variety of C6 and C5 sugars. Whilst the C6 sugars could be easily fermented to ethanol, pending reduction of the organic acids and other inhibitors, the C5 sugars require speciality yeasts for fermentation.

Other than the classical fermentative pathway, C5 sugars can as well be converted, biologi­cally as well as chemically into a wide variety of added value products and "green" com­pounds. In this paper, we have identified 4 pathways for the conversion of C5 sugars but more specifically xylose, a common carbohydrate in biomass hemicelluloses.

Reaction of xylose under an acid catalyst is probably one of the most investigated fields in this domain. The target for this conversion being furfural, a well-known chemical as well as precursor for other compound as furan, Me-THF, THF and furfuryl alcohol, a reactant used in the polymer industry. The best approach for the conversion of xylose furfural, to the best of our knowledge, is chemical as no microorganism allowing conversion of C5 sugars to fur­fural has been identified so far. The conversion of xylose to furfural was reported to reach more than 95 % for both heterogeneous and homogeneous catalyst. On the other hand, the selectivity toward furfural is not always as efficient since the latter undergoes polymerisa­tion in acidic medium, which often also leads to deactivation of the catalyst.

A basic catalyst leads to a conversion of C5 sugars to lactic acid although this pathway as not been deeply investigated in the literature. Lactic acid is a compound well in demand on the market but the limitations for the chemical transformation is the lack of stereospecificity of the products. Conversion of xylose under a base catalyst leads to the production of a race­mic mixture of D — and L-lactic acid and thus reducing the market value of the product, par­ticularly if the polymer industry is targeted. On the other hand, the biological conversion of xylose to lactic acid is a well-known and extensively reported process for which the produc­tion was reported to reach 6.7 g/L/d for genetically modified organisms as, in this specific case, Lactobacillus sp. RKY2. According to the reports, the production of lactic acid would be more efficient by the biological approach since it can lead to a stereospecific and a higher market value.

Reduction of xylose can lead to many different products including xylitol for lower severi­ty up to diols as ethylene glycol and propylene glycol at higher severity. It is ambiguous to determine at this point if either the chemical or the biological pathway is more efficient for the production of xylitol since reports on both pathways have shown promising re­sults. The main problem with the xylitol market is that although it is increasing, it is fair­ly small and therefore it is harder to fit in a new production of xylitol. On the other hand, a more severe reduction of xylose, leading to diols, could be a very interesting opportuni­ty for the production of ethylene glycol and propylene glycol, two very important prod­ucts in the chemical industry. The downside of this approach would be the production of glycerol as a side-product.

Finally, oxidation of xylose is, at this point, the approach with the lower potential for a rapid commercialisation since the market for xylonic acid and xylaric acid is hard to size at present. The conversion process, both chemical and biological seems to have significant po­tential in terms of scalability but the end usage is not well defined at this point. The best option would be to produce glucaric acid from xylaric acid, which could be used as a plasti­cizer. On the other hand, such a process, overall rather complicated, would add a significant cost for a product that would land in the commodity range.

Acknowledgement

We would like to acknowledge Enerkem, Greenfield Ethanol, CRB Innovations and the Min­istry of Natural Resources of Quebec for financial support of the Industrial Chair in Cellulo — sic Ethanol.

Author details

Ariadna Fuente-Hernandez, Pierre-Olivier Corcos, Romain Beauchet and Jean-Michel Lavoie*

*Address all correspondence to: jean-michel. lavoie2@usherbrooke. ca

Industrial Research Chair on Cellulosic Ethanol (CRIEC), Departement de Genie Chimique et de Genie Biotechnologique, Universite de Sherbrooke, Sherbrooke, Quebec, Canada