1.1. The chemical pathway

Either in cyclic or aliphatic form, xylose then tends to dehydrate thus leading to the produc­tion of furfural whilst losing three molecules of water. Although this approach could explain the formation of furfural, it is not the sole options and many detailed reports have shown, by correlating the intermediaries with the actual structure, could be formed by many ap-

proaches depending on the reactant as reported by Marcotullio et al. [20] using halogen ions and proceeding only via the aliphatic form or as reported by Nimlos et al. [21] either via an aliphatic or a cyclic pathway (D-xylopyranose). Many different types of acid catalyst, either Br0nsted or Lewis have been tested for the production of furfural. Although most of the acids reported in literature have been efficient so far for the production of the targeted mole­cule, one of the major side-reaction of furfural is polymerisation which influences the con­version rates and the selectivity of most of the processes reported in literature. An example of the abundance of research on this specific conversion is shown in Table 2 for different de­hydration reactions under acid catalyst..

For these reactions, the temperature is generally between 140-240 °C under proportional pressure allowing the mixture to remain liquid. Many researches also use a co-solvent, often toluene in order to isolate furfural from the aqueous mixture. The reason why toluene is so popular to this purpose is mostly related to the fact that toluene has affinity for fufural thus inhibiting its polymerization.

Heterogeneous catalyst has been proven to be very efficient for the process [22,23] although polymerisation tend to reduce the surface activity thus leading to a short-term deactivation of the catalyst. On the other hand, homogeneous catalyst was also shown to be efficient but at this point the whole technique relies on how the organic solvent is dispersed in the aque­ous mixture. Reducing the size of the organic solvent particles in water (or vice-versa) to the maximum should allow the best transfer between the aqueous phase to the organic phase,

assuming of course that furfural has suitable affinity for the solvent and that the partition coefficient favours the solvent.

Production of furfural itself is of course of significant interest because, amongst many fac­tors, this chemical is commonly used in the industry as a solvent (mostly in oil chemistry). The average world production for furfural is 250 000 t/y and the actual market price evolves around 1000 USD/t [36] with recent market value reported to be closer to 1600 USD/tonne [37]. Furfural can also be a gateway to other products that could be used either as biofuels or as biomolecules. Example of such would be furfuryl alcohol via partial reduction of furfural (see Figure 2 below).

Furfuryl alcohol is also of interest since it is used as resins, adhesives and wetting agent, it has been mentioned that most of the 250 Kt/y of the furfural production is oriented toward production of furfuryl alcohol. The market value of this compound has been reported to be around 1800-2000 USD/tonne [38] and many reports in open literature mentions high selec­tivity for the conversion of furfural with iridium and ruthenium catalyst [39], rhodium [40], iron [41] and with zirconium oxide [42].

Another possible target for the transformation of furfural is for the production of 2-meth — yltetrahydrofuran (Me-THF) (see Figure 3). The latter is actually accredited as an additive for fuel and therefore, the possible market is virtually very important. It is also used in the petroleum industry to replace tetrahydrofuran (THF) that usually comes from non-re­newables.

Reduction of furfural to Me-THF seems to represent an important challenge since there is fewer reports mentioned in literature on the subject, as compared, as an example, to the re­duction of furfural to furfuryl alcohol. Wabnitz et al. [43, 44] patented a one and two step process allowing conversion of furfural to Me-THF under a palladium-based catalyst and a mixture of palladium and copper oxide and chromium oxide as for the two step process.

Lange [45] patented a process using palladium and titanium oxide whilst Zheng et al. [46] worked with a copper alloy. Value for Me-THF could be estimated from the price of THF which is around 3000 USD/tonne [47] and the gap between the value of furfural and Me — THF could justify the process although hydrogen value can be estimated to be around 4.5 USD/Kg (estimated with the actual price of natural assuming reforming of the latter).

Another potentially interesting approach for a transformation of furfural would be decar­boxylation to furan. The general process is depicted in Figure 4 below.

Many researches have focused on decarboxylation including work by Zhang et al. [48] who mentioned decarboxylation with potassium-doped palladium, and Stevens et al. [49] who re­ported conversion with copper chromite in supercritical CO2.

Results reported in literature show that xylose, under an acid catalyst, tend invariably to de­hydrate to furfural thus limiting the possibilities for side-products in such specific condi­tions. The acids could be Brrnsted or Lewis type, all lead to the production of furfural furthermore when temperature are raised above 150 °C.

1.2. The biological pathway

Although furfural is a very common route for the conversion of xylose under an acid cata­lyst, furfural itself is rarely related to microorganisms in that sense that it is often considered as an inhibitor instead of a metabolite. Nevertheless, to the best of our knowledge, no report mentioned a biological conversion of xylose to furfural.