Etude de la promotion électrochimique de l'oxydation catalytique de l'éthylène sur des oxydes métalliques

The electrochemical promotion of heterogeneous catalytic reactions is a new development which draws the attention of the scientific community as much to fundamental research as to the possible direct applications of this phenomenon in automotive and in industrial catalysis. We have chosen to study the electrochemical promotion of the complete ethylene oxidation on metallic oxides catalysts. The principal oxides which we used has been the iridium oxide (IrO2) and iridium and titanium oxide mixtures (IrO2+TiO2), deposited on yttria-stabilized zirconia (YSZ). We have proved that the electrochemical promotion of ethylene combustion can be applied to these catalysts. Contrary to what was established before, we have demonstrated a certain persistence of activation of the reaction. mainly on the iridium oxide, after the end of polarization. The persistent reaction rate enhancement ratio is a function of the passed electric charge and levels off at a value of approximately 3. The other experimental significant fact for heterogeneous catalysis which has been discovered is the relation between chemical promotion and electrochemical promotion. Applying electrochemical promotion to a IrO2 + TiO2 composite catalyst, the two kinds of promotions have been confronted, since a IrO2 + TiO2 composite catalyst presents a chemical promotion, more precisely a synergy between the two components of the catalyst. The maximal amplitude of the electrochemical promotion of such catalysts has been demonstrated to be strongly reduced in comparison with the one of pure catalysts. As the maximal enhancement factor of the catalytic reaction rate was 12 for pure iridium oxide catalyst, the same factor dropped to less than 2 for a IrO2 + TiO2 composite catalyst. This result has enabled us to affirm that the mechanisms of chemical promotion and electrochemical promotion should involve a common type of surface activating species : a mobile surface spillover oxygen species. The catalyst work function on-line measurement with a Kelvin probe has enabled us to link the reaction rate with the work function change and to link the applied potential with the work function change, for al1 the phases of a transient polarisation experiment. The persistent activation of the catalyst after the end of a polarization has been linked to a persistent change in the work function. A major trend of this research consists in cyclic voltammetry. It is certainly one of the most powerful medium to extract experimental facts which have been useful to understand the electrochemical promotion principles. Due to the small amplitude of the potential which has been used in cyclic voltammetric experiments, the electrolyte-catalyst interface wasn't practically affected by these in situ measurements. By analogy to voltammetric charge formation mechanism in aqueous electrochemistry, we have taken the following chemical capacitance as responsible for the voltammetric charge in our system: IrO2 + δO- ⇄ IrO2+δ + δe- According to this mechanism, the reaction between O- and IrO2 can only occur at the interface between solid electrolyte and catalyst, where contact between the two species exists. In Our system, voltammetric charge has been demonstrated to be proportional to the thickness of the catalyst, proving by this way that the whole surface of the catalyst is in contact with the O- activating species coming from the solid electrolyte. So the implication of a spillover oxygen species during electrochemical promotion has been demonstrated by the way of cyclic voltammetry. The existence of this spillover oxygen species has been also directly proved by thermal desorption experiments under electrochemical promotion conditions. By considering the strong change in the decomposition temperature of the oxide induced by its polarisation, the thermal desorption technique has been used to display the influence of the polarization on the structure of the oxide catalyst itself.

Chemical and electrochemical promotion of supported rhodium catalyst

Olena Baranova

The chemical and electrochemical promotion of highly dispersed nanofilm Rh catalysts (dispersion: about 10 %, film thickness: 40 nm) has been investigated for the first time. To this end Rh metal was sputter-deposited, either on a purely ionic conductor (8 % Y2O3-stabilized ZrO2) or on a mixed ionic-electronic conductor (TiO2), the latter being a highly dispersed layer of TiO2 (4 µm) deposited on YSZ. These catalysts are designated as Rh/YSZ and Rh/TiO2/YSZ, respectively. It was established analytically that both in the Rh/YSZ and in the Rh/TiO2/YSZ system, the catalyst films have a nanoparticle-size grain structure. The Rh supported on titania is rather porous, exhibiting a higher dispersion and surface area than Rh on YSZ. Both after reduction (H2, T=400 °C) and after oxidation (O2, T=400 °C), Rh supported on TiO2 was found to be in a more highly reduced state than Rh on YSZ. After reducing treatment, the Rh/TiO2/YSZ samples contain a larger amount of weakly bonded oxygen, which can be attributed to oxygen "backspillover" from the TiO2. A major feature of this research was the electrochemical characterization of the oxygen/Rh/solid electrolyte three-phase boundary by steady-state polarization measurements and by impedance spectroscopy. These are powerful techniques for extracting experimental trends and details that are useful for an understanding of the electrochemical promotion principles. It was shown that the exchange current densities at the Rh/solid electrolyte interface are lower on account of the TiO2 layer. The exchange current densities are more than twice lower at Rh/TiO2(4 µm)/YSZ than at Rh/YSZ, demonstrating that the former interface is much more polarizable. The mechanism of oxygen exchange occurring close to equilibrium (O2/O2- couple) was investigated for the first time at Rh catalyst electrodes interfaced with solid electrolyte. The processes of cathodic oxygen reduction and anodic oxygen evolution are not symmetric, but they are similar in the two systems, Rh/YSZ and Rh/TiO2(4 µm)/YSZ. The cathodic process consists of three steps: dissociative adsorption of oxygen at the gas-exposed Rh surface, atomic oxygen diffusion to the electrochemical reaction sites (ERS), and a two-electron transfer to this oxygen on the ERS. The cathodic process is limited by interfacial diffusion of oxygen atoms from the gas-exposed metal surface to the ERS. The anodic process includes two steps: two-electron transfer reaction, which is the rate-determining step, and oxygen desorption to the gas phase. With data obtained from impedance spectroscopy at the equilibrium potential, it was possible to confirm the reaction scheme proposed. It was demonstrated that Rh nanofilm catalysts interfaced with YSZ or TiO2/YSZ can be electrochemically promoted for the reaction of ethylene oxidation. Small anodic currents cause periodic oscillations in catalytic rate and potential of the Rh/YSZ catalyst, while at Rh/TiO2/YSZ they give rise to a stable and reversible rate enhancement by up to a factor 80. The increase in ethylene oxidation rate is up to 2000 times larger than the electrochemical rate, I/2F, of O2- oxidation. The pronounced electrochemical promotion behavior that has been observed is due to the anodically controlled migration of O2- species from the electrolyte to the Rh/gas interface. At the Rh surface, these species destabilize the formation of rhodium surface oxide (Rh2O3). The existence of backspillover oxygen species has been confirmed by impedance measurements under positive applied potential. Another significant result for heterogeneous catalysis is the finding that thick as well as thin films of Rh/TiO2/YSZ catalyst are open to chemical promotion of ethylene oxidation and partial methane oxidation, ie, they offer a higher catalytic activity and stability than the Rh/YSZ catalysts. The modification of catalytic activity observed for Rh/TiO2/YSZ was attributed to either a "long-range" electronic-type SMSI mechanism or to a self-driven electrochemical promotion mechanism. In both cases, the ultimate cause of promotion are different work functions of catalyst and support. Equilibration of the work functions of two solids in contact induces surface charging, a migration of O2- ions to the catalyst/gas interface, and a weakening of the Rh-O chemisorptive bonds. It facilitates reduction of oxidized surface sites. Another important achievement of the present work was that of exploring and confirming the possibilities of current-assisted activation of Rh/TiO2/YSZ catalysts. In partial methane oxidation using close to stoichiometric gas compositions (CH4 : O2 = 2 : 1) at 550 °C, the inactive (oxidized) Rh/TiO2/YSZ catalyst was successfully activated by applied currents, either positive or negative. This phenomenon is an example of "permanent" electrochemical promotion furnishing a permanent rate enhancement ratio of γ = 11. The activation by negative currents is explained in terms of an electrochemical reduction of rhodium surface oxide, while the activation by positive currents can be explained by the mechanism of electrochemical promotion.

EPFL2005

Promotion électrochimique des catalyseurs à base de rhodium et d'iridium

Justyna Eaves

In this thesis, the electrochemical promotion of lr and Rh based catalysts was studied for the complete combustion of ethylene and propylene. The study was performed by using two types of electrocatalytical cells: a classical, pellet type cell and a tubular cell. This work has two major goals. The first one is the optimization of the employed electrocatalytical systems from electrochemical point of view. The second one concerns the study of the mechanism of the electrochemical promotion. The major part of this work concerns the electrochemical characterization of two cell types: a pellet type cell and a tubular cell. On the basis of the obtained results the modélisation of the potential and current distribution was performed, providing key parameters for the development of more efficient cell configurations. On one hand, these configurations ensure an uniform current distribution on the working electrode and enable correct catalyst potential measurement. On the other hand, it minimizes the current bypass in the bipolar tubular cell. The aim of the investigation in a two-compartment reactor was to evaluate and to compare the Au and Ir electrodes used in this work. Under 400°C, gold was found to be much worse material than lrÛ2 both as reference and counter electrode. Its potential was affected by oxygen and propylene adsorption and no equilibrium of electrochemical reaction was obtained at such low temperature. On the base of this knowledge, several materials better adapted to electrochemical promotion conditions have been proposed. Using the two-compartment reactor potential measurements under steady state conditions allowed to verify the mechanism of the propylene combustion on Rh and that of ethylene combustion on lr based catalysts proposed in the literature. Ameliorations of the existent kinetic models were proposed. The kinetic measurements were then analyzed with the additional help of ex situ XPS analysis. This study allowed to link the oxidation state of the catalyst surface with the extent of electrochemical promotion of the catalysts. It also allowed the ex situ observation of the appearance of new oxygen (oxide) species on the gas exposed surface of Ir catalyst. This new type of oxygen was different from the oxygen of Ir02 oxide, and it can represent the promoting species, according to backspiïïover theory. The same oxygen species was observed on well oxidized Ir02 catalyst surface, even before the polarization of this catalyst. Its presence was associated to the catalyst-support effect. It explains the high activity of Ir02 catalyst by comparison with the catalytic activity of lr metal. XPS analysis of Rh based catalysts indicated the possibly formation of an inferior Rh oxide (close to RhO) due to polarization. This may be a possible explanation of the observed permanent electrochemical promotion. Voltammetric charge measurements in very narrow potential window were performed while changing the catalyst thickness. The charge was demonstrated to be proportional to the catalyst thickness, suggesting the presence of the charged species on the gas exposed catalyst surface. The analogy between this situation and that of the emerged electrodes in liquid state electrochemistry was emphasized. The voltammograms obtained in this study were interpreted in the quite new, original way, which allowed to calculate the capacity of the different catalyst interfaces, the gas/catalyst interface, and the triple phase boundary, in good agreement with reference values from impedance spectroscopy measurements. Finally, the reaction rate transients during galvanostatic polarization of lr based catalyst have been moralized. The model fits well the experimental results.

EPFL2004

Effectiveness factor of thin-layer IrO2 electrocatalyst

Erika Herrera Calderon

Dimensionally Stable Anodes (DSA®) are electrodes composed of a metal support and an oxide coating (catalyst). Particularly in this thesis, the Ti/IrO2 electrode was used for the study of the fraction of the surface which participates effectively in the investigated reactions (effectiveness factor, Ef ). Four reactions with different kinetics have been investigated: (a) Fe3+/Fe2+ (fast reaction), (b) O2 and (c) Cl2 evolutions (slow reactions), and (d) isopropanol oxidation (complex reaction involving redox catalysis). This study has allowed us to apply for the first time in electrocatalysis the term, effectiveness factor, Ef , used before in other fields like in heavy metals recovery with 3D electrodes and in heterogeneous catalysis. On the other hand, in order to achieve the main objective of this thesis, an analytical and a qualitative approach was proposed. Furthermore, the Ti/IrO2 electrodes were characterized electrochemically using the cyclic voltammetry (CV) technique in which the voltammetric charge was measured at different potential scan rates, potential windows, and electrolyte temperatures using various Ti/IrO2 loadings. From the voltammetric charge, two contributions were found. The first contribution is related to the double layer capacitance and the second with the redox couples present on the electrode surface. The first reaction investigated in this work was the Fe3+/Fe2+ redox couple. In the first part of the investigation of this reaction, the CV technique was used; however, some issues related to the uncompensated electrolyte resistance were encountered. Therefore the effectiveness factor for this particular case was not possible to obtain. On the other hand, using rotating Ti/IrO2 disk electrodes (RDE) for the same reaction, the effectiveness factor, Ef , has been estimated, however, this value is much lower than the predicted one. The other investigated reactions in this work were O2 and Cl2 evolutions. The investigated kinetics of both reactions showed that the evolution of chlorine is much faster than that of oxygen. The effectiveness factor, Ef , for the oxygen evolution was close to the unity for all IrO2 loadings, contrary to the chlorine reaction that decreased with the loading. This behaviour was predicted according to what is given in the analytical approach. The last reaction investigated in this work was the oxidation of isopropanol. In this case a model was proposed based on three main reactions: Electrochemical IrO2 oxidation to IrO3, chemical oxidation of isopropanol via IrO3, and chemical decomposition of IrO3 to IrO2. The effectiveness factor, for the electrochemical reaction seemed not to correlate with the loading, whereas for the chemical one, it diminished with the loading.

EPFL2008