γ-valerolactone (GVL) is an attractive platform molecule that can be derived from biomass and can be converted to a variety of chemicals, including biofuel additives. It can be nicely derived from cellulose in few steps. The last step is the conversion of levulinic acid (LA) into γ-valerolactone (GVL), where ideally the hydrogen-source is the formic acid that is co-generated with LA. In mild conditions, this reaction is believed to start with the hydrogenation of the ketone function, followed by a cyclizing esterification. While Ru is a poor hydrogenation catalyst compared to Pt or Pd in gas phase, it is efficient in aqueous phase conditions to hydrogenate ketones such as the conversion of LA into GVL. Comparing experiments in THF and water, we showed that the activity of Ru/TiO2 catalyst is highly sensitive to its environment while Pt/TiO2 is not. Better insight is provided by periodic DFT calculations at the GGA level. The presence of a H-bonded water molecule can affect the adsorption of the reactant, intermediate or product. For instance, the obtained alcohol is not directly chemisorbed on the surface: it is strongly h-bonded to the chemisorbed water molecule. This water molecule dramatically reduces the energetic span of the reaction pathway on Ru, hence enhancing the catalytic activity. Conversely, the presence of such a H-bonded water molecule doesn’t not affect the energetic span of the reaction pathway on Pt. Finally, we predict also that this activation can be generalized to other oxophilic metals such as Co or Ni.[1] This study demonstrates that the knowledge accumulated over the years for gas phase reactions at metallic surfaces is not systematically transferable to aqueous phase conditions since water can also play an essential role in the catalytic activity. It also explains why ruthenium is so widely used to transform cellulosic feedstock while it is generally discarded for oil-based feedstock.[2] We further investigated the LA hydrogenation with FA as a hydrogen source using various Ru/C as efficient catalysts. We showed that the intermediate surface formate inhibits at least partially the LA hydrogenation. In addition, the FA decomposition is highly sensitive to the kind of the preparation method of the Ru/C catalyst: (i) the process looks structure sensitive favoured on larger particles and (ii) residual chlorine decreases significantly the FA decomposition rate.[3] REFERENCES [1] Michel, C.; Zaffran, J.; Ruppert, A. M.; Matras-Michalska, J.; Jedrzejczyk, M.; Grams, J. ; Sautet, P. Chem. Comm., 2014, 50, 12450-12453 [2] Michel, C. ; Gallezot, P. ACS Catalysis, 2015, 5, 4130-4132 [3] Ruppert, A. M.; Jedrzejczyk, M.; Sneka-Platek, O.; Keller, N.; Dumon, A. S.; Michel, C.; Sautet, P. & Grams, J. Green Chem. 2016, ASAP, 10.1039/C5GC02200B