Optimisation of the processing conditions of hydrolytic hydrogenation of cellulose using carbon nanofiber supported Ni catalysts

14 figures, 6 tables.
The synthesis of sorbitol from cellulose over Ni catalysts is a promising valorisation route in the biorefinery scenario, applying relatively simple preparation methods and earth-abundant metals. The overall selectivity, however, depends on the kinetic control of a complex reaction network, involving the hydrolysis of cellulose to monosaccharides via cello-oligomers, glucose hydrogenation into sorbitol and hydrogenolysis side-reactions of sugars and sorbitol to low molecular weight polyols. Therein, subtle changes in the catalyst composition and process conditions might have a strong impact on the final product distribution. Driven by these challenges, this work first-time provides novel insights into the hydrothermal hydrogenation of cellulose over a carbon nanofibre supported Ni catalyst (Ni/CNF). Firstly, the impact of the duration of a ball-milling pre-treatment step and the influences of the hydrothermal time and temperature were thoroughly analysed. The experimental results obtained highlighted the importance of process control to promote the first transformation of cellulose to glucose and its subsequent hydrogenation to sorbitol to minimise the extension of side reactions. Finally, an additional study was conducted to palliate the recalcitrant nature of cellulose by decreasing mass transfer limitations to a minimum extent. This was achieved by including an additional mix-milling of the amorphous cellulose produced in the first pre-treatment with the catalyst and increasing the H2 pressure of the hydrothermal hydrogenation process. This allowed attaining a sorbitol yield as high as 62% at 190 ºC using an initial H2 pressure of 8 MPa for 26 h, which is one of the best results reported in the literature.
The authors are grateful for the financial support from the Spanish Ministry of Economy and Competitiveness (MINECO, Project ENE2017-83854-R) and the I+D+i project PID2020-115053RB-I00, funded by MCIN/ AEI/10.13039/501100011033. J.R. and D.T. are grateful to the Spanish Ministry of Science, Innovation and Universities for the Juan de la Cierva Incorporación (JdC-I) fellowships (Grant Numbers: IJC2018-037110-I and IJC2020-045553-I, respectively) awarded.