VALORISATION OF BIOMASS-DERIVED MOLECULES BY NOBLE METAL CATALYSTS

Valorisation of different biomass derived molecules was successfully approached and studied in this PhD project. The focus of the thesis was addressed to the catalysts preparation, passing through an accurate catalytic designed, to be then tested in academic and industrially appealing reactions. This approach led to the synthesis of different but equally interesting catalytic systems for the valorisation of substrates derived from the first and second generation of biomass feedstock. An extended study, at first, was conducted on the oxidation of glycerol (1st generation of biomass related), both in alkaline (needed for gold monometallic systems) and free pH (high industrial relevance) conditions. The target reaction was approached starting from the simplest Au/C catalytic systems, to finally move to more complicated and innovative materials: bimetallic once. Initially, the Au on carbon Vulcan (with the highest graphitisation degree) SOL derived catalysts showed a remarkable initial activity (IA= 1091 h-1) in comparison with the other carbonaceous supports (Norit and X40S) and the SMAD derived catalysts. This result pointed out the importance of the protecting agent (a polymer that surrounded the nanoparticles and is solely present for the SOL synthetic route) beside the importance of the support’s features. Similarly, electronic effects ascribed to the interaction with the support of the nanoparticles (i.e. the strong metal support interaction (SMSI) thermally induced on Au4Ag1/TiO2) showed to be the ruling factor to determine the oxidation state of the metals. This latter, subsequentially, influence the catalytic activity: an enhanced initial catalytic activity was detected for the Au4Ag1/TiO2 catalyst (IA= 1616 h-1), in comparison with the Au4Ag1/Al2O3 (IA= 963 h-1). The SMSI have influenced also the stability of the system, avoiding the enlargement of the nanoparticles during the thermal treatments. On the other hand, the SMSI induced the presence of Ag+ species onto the bimetallic nanoparticles titania supported, leading to a quite rapid deactivation of the catalytic system. The thermal treatments pointed out also the importance of the protecting agent (polyvinyl alcohol, PVA): on one side when it is present confers resistance to the system towards the nanoparticles aggregation, on the other when it is removed from the nanoparticles’ surface (by the same thermal treatment), the catalyst acquired an enhanced initial activity. AuPt/TiO2 catalytic systems were subsequentially exploited both in alkaline and free pH conditions. The gold content positively influenced the activity of the catalytic systems in both the conditions. In particular Au9Pt1/TiO2 was the most active catalyst in the alkaline condition (IA= 7389 h−1), and Au6Pt4/TiO2 showed the highest initial activity (IA= 301 h-1) in free pH condition. For all the bimetallic system mentioned and exploited in the valorisation of glycerol, furthermore, a synergistic effect was detected. The importance of gold as modifier to confer resistance to the catalytic system by stabilizing the oxidation state of the second metal was also established. Subsequentially, completely different designed and synthesised catalysts were prepared for the valorisation of substrates related with the 2nd generation of biomass. Bare carbon nanofibers (CNFs) and functionalised CNFs (CNFs-P, CNFs-O and CNFs-N), for instance, were employed as supports for Ru nanoparticles (introduced by incipient wet impregnation). All the catalysts prepared showed activity in the valorisation of cellulose derived molecules. In particular, it was observed how N-containing functionalisation of the support, promoted by a strong interaction with the Ru nanoparticles, led to the highest catalytic activity among the set of catalysts tested for the levulinic acid (LA) hydrogenation (88 % of conversion after 3 h) with a full selectivity to y-valerolactone (GVL). On the other side, exploring the 5-hydroxymethylfurfural (HMF) valorisation, Ru/CNF-N and Ru/CNF-P showed a lower activity but also a change in selectivity. In fact, these latter two catalysts enhanced the formation of ethers due to the reaction between 2,5-dihydroxymethylfuran and/or methylfurfuryl alcohol with the solvent (2-butanol). Similar support effects were also observed in the furfural hydrogenation over platinum nanoparticles (introduced by solvated metal atoms deposition, SMAD) supported on niobia and tailor-made modified niobia. Niobia was hydrothermally synthetized pure and doped with other two different metals (W and Ti, both 10 at.%) to tune the acidity of the system. In particular, we were able to enhance to 0.191 mmolPy/gCAT (W-Nb2O5) and decrees to 0.014 mmolPy/gCAT (Ti-Nb2O5) the acidity of the pure Niobia (0.078 mmolPy/gCAT). Platinum nanoparticles, showing a narrow particle size distribution (1.1-1.2 nm) for all the supports, have allowed a proper study of the acidity effect. The acidity, indeed, showed to be the ruling factor: the most acidic material showed the highest activity coupled with a selectivity addressed to the furan ether products (acid catalysed reaction’s step) at the expenses of furfuryl alcohol (highest selectivity of FA showed for the lowest acid catalyst). Unfortunately, the condition and the type of acidity (Lewis acidity) obtained were not sufficient to observe a high fraction of diols (target product, less than 10 % in selectivity), produced from the ring-opening of the substrate. Lastly, in the benzyl alcohol oxidation (model compound for the lignin) it was highlighted how gold-based materials characterised by comparable nanoparticles dimension (Au-Pd, Au-Pt, Au-Ru and Au-Cu, all supported on carbon) could change the catalytic behaviour and the bimetallic structure just by varying the second metal. For AuPd/C and AuPt/C, for example, alloyed structures were observed. On the other hand, for the case of Ru as second metal, a core-shell structure was found. When Cu was employed, bimetallic nanoparticles with Au:Cu molar ratio lower than the nominal one were detected suggesting the presence of segregated gold nanoparticles. All the catalysts were active and highly selective towards the desired and industrial appealing product (benzaldehyde, selectivity ≥ 99 %). Only in the case of AuPd/C and AuCu/C, however, a synergistic effect was observed. In particular, the AuPd/C bimetallic sample showed the highest activity (fully conversion of the substrate after 5 min). For the interesting Au-Cu system (the only catalysts that contain a not noble metal), furthermore, the role of the Cu was clarified and the composition effect was studied. The metals were deposited on a carbonaceous support by SMAD technique in order to avoid a protecting agent influence. More in details, it was speculated how Cu, promptly oxidised at CuO (if exposed to air), is responsible of the O2 activation, while the reaction took part at the Au-CuO interface. This reactivity is guided by a specific structure of the bimetallics particles finely characterized: Aucore-CuOshell structure. This last evidence highlighted once more the importance of having a good knowledge and control on the catalyst synthetic routes. Furthermore, synergistic effect was observed for all the active AuCu/C bimetallic systems, even when the amount of gold was very low (Au13Cu1/C, IA= 329 h-1). The highest initial activity, however, was reached with Au4Cu1/C catalysts (IA= 399 h-1). All the active AuCu bimetallic catalysts showed a high selectivity towards the desired product: benzaldehyde (≥ 95%). Good stability against deactivation was also observed. For the Cu-rich sample (Au1Cu17/C) case, distinguished by the negligible activity, it was assumed how the external copper oxide shells, by entirely covering the gold atoms, have repressed any catalytic activities.