Abstract
The conversion of the most reactive hexose,
i.e. fructose, to 2,5-furandicarboxylic acid
(FDCA) proceeds through the formation of the intermediate 5-hydroxymethylfurfural (5-HMF)
by a dehydration reaction. Aerobic oxidation of this compound over a noble metal
catalyst leads to the desired final product. The two reaction steps were studied independently
and water was used as the solvent, despite causing extensive by-products formation.
In the first place, the effect of the acid catalyst choice on the aqueous fructose dehydration to
5-HMF was investigated in homogeneous solution under relatively mild reaction conditions (140°C). Salt addition experiments and the use of acids with different nature and strength
confirmed that, especially at higher pH values, the influence of anions present in solution on
the reaction rate and selectivity is not negligible.
Highly hydrophilic polyacrylates bearing propanesulfonic acid groups were synthesized and
tested for fructose dehydration under the same reaction conditions. The polymers with higher hydrophilicity (i.e. with higher swelling bility in water) and higher cross-linker contents were found to be more selective towards 5-HMF production and more prone to suppress by-products formation. The best catalytic performance was achieved with the material DAE25-40 (25 mol% of sulfonated monomer and 40 mol% of cross-linker), which gave 85% 5-HMF selectivity at 30% substrate conversion. One advantage of these materials with respect to PS-DVB-based polymers is that their higher thermal stability allows their use at higher reaction
temperatures, which can be adopted as a strategy to accumulate higher 5-HMF yields.
A one-step approach was developed for the production of mesoporous sulfonated carbon materials by means of an aerosol synthesis. The synthetic work on these materials was
performed by Nicolas Duyckaerts and Ioan-Teodor Trotuş.
Nebulizing a clear aqueous
solution of sucrose and sulfuric acid through a heated oven led to subsequent dehydration,
carbonization and sulfonation of the carbohydrate structure, in less than two seconds
residence time. Materials with ion-exchange capacities ranging from 0.1 to 0.6 mmol g-1
were obtained. Porosity could be easily introduced via salt templating and could be adjusted by varying the loading and type of salt used. The highest surface area was measured for the carbon material prepared by adding Li2SO4 to the precursor solution, which gave a BET surface area of 506 m2 g-1
and a mesopore size distribution between 2 and 8 nm.
Testing these materials for fructose dehydration and inulin hydrolysis showed that the carbons
synthesized by salt templating were more active than the bulk ones, especially for inulin hydrolysis, for which the initial activity was enhanced by a factor of 3.5, making these
materials competitive with the most active commercial resins.
In the second part of this study, the catalytic performance of Au, Pt, Pd and Ru monometallic
materials prepared with two different
synthetic procedures was evaluated for the 5-HMF
aerobic oxidation to FDCA (100°C, 8 bar O2, pH 9). The identity of the metal was found to
determine the reaction pathway and the accumulation of one oxidation intermediate over the
others, while the size of
the metal particles mainly influenced the reaction rate.
Interestingly,
for the reactions performed over Au and Pt, a structure sensitive behavior was observed.
The combination of metal couples on the same support to yield bimetallic materials was also
shown to have an effect on the catalytic performance. Electronic effects were found to
accelerate the oxidation of the hydroxyl-methyl moiety of 5-HMF in the case of Au-Pd alloys,
and could be identified as the sole cause for variations in activity and selectivity towards the
different oxidation intermediates in Au-Pt alloys.
Despite bearing relatively large metal particles (11-12 nm), the highest initial reaction rate
normalized to the corresponding total metal surface area was shown by Au-Ru catalysts,
suggesting that the combination of these two metals is favorable for 5-HMF conversion to
FDCA and that further materials optimization could improve their catalytic performance.