Alcools dehydrogenation by metallic catalysts: Probing the co-adsorbates influence by DFT calculations
Résumé
1 Introduction
The necessary swap from petroleum feedstock to biomass feedstock triggers the development of new catalysts, in particular in the conversion of oxygenated products.[1] In this context, the dehydrogenation of alcohols is one of the central issues. In alkanes, the C-H bond scission is mainly catalyzed by supported metallic particles, the metal and the support being tailored depending on the desired products. For the C-H and O-H bond dissociation in alco- hols, supported metallic particles are also used, but they have to be adapted to the presence of the aqueous phase, the alcohols being generally solubilized in water.[2-3] Numerous co-adsorbates may affect the elementary steps of the alcohols dehydrogenation : water, hydroxyl groups, etc. Experimental studies have already shown that the addi- tion of water can promote the conversion of aliphatic alcohols into the corresponding ketone or aldehyde.[4] Howe- ver, a better understanding of the influence of water and other co-adsorbates is still required. Using periodic calcu- lations in the DFT framework, we will focus on the role of H-bonded neighbors on the C-H and O-H bond scissions in water, ethanol and glycerol.
2 Computational Details
The metallic surfaces have been modeled by a slab made of four layers, separated by 5 layers-equivalent of vacuum. The two bottom layers are frozen in the ab initio bulk position. The supercell is a 3x3 cell. The calcula- tions have been performed at the DFT level using the VASP code. Various XC energy functionals have been used. A tight convergence of the plane-wave expansion was obtained with a cutoff of 400 eV. The electron-ion interac- tions were described by the projector augmented wave method (PAW). A Monkhorst-Pack mesh of 3x3x1 K-points was used for the 2D Brillouin zone integration. Geometries have been converged until forces were less than 0.01eV. Transitions states have been characterized by the presence of a unique imaginary frequency.
3 Results and discussion
Water is a simple molecule of great importance in the catalytic valorisation of biomass but also in a much wider range of context. We will first deeply analyse the water adsorbtion at transition metal close packed surfaces and the subsequent OH sicssion, questioning the influence of a co-adsorb water on those processes. Then, we will shift to alcohols, using ethanol as a model to better understand the specificity of polyols such as glycerol. The role of coad- sorbates such as water but also hydroxyl groups will be probed on two different metals: the widely used platinum and the more oxophilic rhodium.
OH bond scission in water --
We have considered close-packed surfaces of Ru, Co, Rh, Ir, Ni, Pd, Pt.[5] At a 2/9 ML coverage, water preferen- tially adsorbs as a dimer rather than as two monomers: a first water molecule is adsorbed at atop site, the second one is H-bonded as the H-bond acceptor, farther from the surface. This process is strongly synergetic: both the H-bond and the metal-O bond are shortened, and the energy gain is greater than the formation of a H-bond plus the weak interaction between the surface and the second water molecule. The synergy can be quantified around 0.25 eV per dimer. In this water dimer, the OH bond scission is facilitated compared with the isolated water for all the considered metals. Those calculations have been performed using a GGA functional (PW91). For sake of comparison, the same analysis has been also performed using PBE but also PBE+D to take better into account the dispersion [5-7] and optB86, a functional recently proposed by Klimes et al.[8] based on the non-local Van der Walls density functional initially proposed by Dion et al.[9] The conclusions do not change with the choice of the functional: the OH scission is activated in the water dimer compared with the monomer. This ef- fect is much stronger on oxophilic metals such as Ru or Rh (up to 30% decrease of the activation barrier) than for later metals such as platinum (only 7% decrease of the activation barrier).

OH and CH bond scission in alcohols --
Then, we have focused on the effect of an adsorbed water molecule on O-H and C-H scission in ethanol on Rh(111) and Pt(111).[10,11] The situation is analogous to the water dimer, the ethanol replacing the physisorbed H-bonded water molecule. Here again, the OH scission is strongly assisted on Rh and weakly on Pt. In addition, the CH scission is inhibited (up to 30% increase of the activation barrier). Thus, the presence of the adsorbed water modifies strongly the reactivity of ethanol at a metallic surface: the OH scission becomes easier than the CH scission and rhodium becomes more active than platinum. At a lesser extent, the same effect can be pinpointed in polyols such as glycerol.[10,12]
However, those results are contradictory with some experimental results: the oxidation of aliphatic alcohols using supported platinum catalysts is promoted by the addition of water.[4] Under oxidative conditions, water is expected to generate hydroxyl groups on metallic surfaces. The co-adsorbtion of such species strongly modifies the reactivity of the metallic catalyst, even more than the co-adsorbtion of water. In presence of hydroxyl groups, breaking the OH bond of ethanol is a low barrier process on Rh(111) and even a barrierless process on Pt(111). Thus, water in- creases the activity of the platinum catalyst as a source of hydroxyl groups, not as a co-adsorbate.
4 Conclusions
According to DFT calculations using various functionals, H-bonded neighbors are key players in the reactivity of oxygenated compounds at metallic surfaces. Their influence depends on the metal in consideration. Water can strongly activates the OH bond breaking, especially on oxophilic metals, while hydroxyl groups modify drastically platinum based catalysts toward alcohols oxidation.
Acknowledgements
Calculations were performed using the local HPC resources of PSMN and of GENCI (CINES/IDRIS), project x2010075609. The French ANR supports this project (GALAC).
References
[1] A. Corma, S. Iborra, A. Velty, Chemical Reviews, 107 (2007) 2411
[2] J. Chaminand, L. Djakovitch, P. Gallezot, P. Marion, C. Pinel, C. Rosier, Green Chemistry, 6 (2004) 359 [3] J.N. Chheda, G.W. Huber, J.A. Dumesic, Angewandte Chemie International Edition, 46 (2007) 7164 [4] A. Frassoldati, C. Pinel and M. Besson, Catalysis Today, 173 (2011) 81
[5] C. Michel, F. Göltl, Ph. Sautet Physical Chemistry Chemical Physics, 14 (2012) 15286
[6] S. Grimme, Journal of Computational Chemistry, 27 (2006) 1787
[7] W. Hujo, S. Grimme, Physical Chemistry Chemical Physics, 13 (2011) 13942
[8] J. Klimes, D.R. Bowler, A. Michaelides, Journal of Physics-Condensed Matter, 22 (2010) 022201
[9] M. Dion, H. Rydberg, E. Schröder, D.C. Langreth, B.I. Lundqvist, Physical Review Letters, 92, (2004) 246401
[10] C. Michel, F. Auneau, F. Delbecq, Ph. Sautet, ACS Catalysis, 1, (2011) 1430
[11] S. Chibani, C. Michel, F. Delbecq, C. Pinel, M. Besson, Catalysis Science and Technology, in press (2013) (DOI: 10.1039/c2cy20363d)
[12] F. Auneau, C. Michel, F. Delbecq, C. Pinel, Ph. Sautet, Chemistry: A European Journal, 17 (2011) 14288