The Catalysis of the Selective Electrochemical Oxidation of Hydrocarbons

Decreasing prices for renewable electricity are making electrification of the chemical industry more interesting, both as a means of energy storage, and to save on nonrenewable energy carriers. Such an electrochemical approach is interesting for the partial oxidation of hydrocarbons, where the catalytic challenge is to avoid formation of the thermodynamically most stable product CO2, while providing sufficiently high reaction rates.

In electrochemical processes, the driving force for the reaction is provided by applying an electrical potential, which allows for better control over reaction products and operation at room temperature. In this thesis, I studied the electrochemical partial oxidation of propene which is highly relevant for two reasons: First, propene has two different sites for oxidation; the double bond, and the allylic carbon.

This makes it ideal for mechanistic studies as different products will result from different adsorption geometries on the surface when varying reaction conditions. Second, several propene oxidation products, such as allyl alcohol, acrolein, acrylic acid, and propylene glycol are important commodity chemicals with annual production volumes of several megatons.

The catalysts used for propene oxidation experiments were high surface area Pd, Au and AuPd alloys, prepared by the hydrogen bubble template method and thoroughly characterized using XRD, XPS, SEM and LEIS. The choice of materials was based on previous literature and DFT calculations by my collaborators.

Initially, electrochemical propene oxidation with ex-situ product analysis was carried out on Pd electrodes in acidic electrolyte. Both activity and product distribution depend on the potential, in alignment with reports from the literature. The highest activity was observed at ca. 0.9 Vvs. RHE; at the same potentials, selectivity for allylic oxidation to acrolein and acrylic acid is the highest.

Thorough product analysis led to the identification of two products not previously reported under the same conditions, to the best of my knowledge: Allyl alcohol (primarily at potentials below 0.9 V vs. RHE), and propylene glycol (at potentials above 0.9 V vs. RHE). DFT modeling and allyl oxidation experiments suggest that allyl alcohol is an intermediate in the formation of acrolein, and that oxygen insertion is the rate limiting step.

Based on these results we used a combination of DFT and advanced stripping experiments with chip based electrochemical mass spectrometry (EC-MS) to show that surface coverage dictates the reaction mechanism: Propene degradation under reaction conditions leads to the formation of a layer of adsorbates which facilitates propene coordination to the catalyst solely through the allylic carbon, resulting in the observed selectivity for acrolein and acrylic acid.

This surface layer can be influenced by alloying Pd with Au: Already at 10% Au content, the activity is improved, while still providing high selectivity for allylic coordination. To further understand the nature of the adsorbed species, we performed experiments on Pd using surface-enhanced infrared absorption spectroscopy (SEIRAS).

However, due to the adsorption geometry of the adsorbed species, signal intensities are very weak. Nonetheless, our results show that the degradation mechanism for propene oxidation is fundamentally different from allyl alcohol oxidation. During allyl alcohol oxidation, the surface is covered with CO.

In the presence of propene, however, no CO was observed, which supports the hypothesis that oxygen insertion is the rate limiting step. SEIRAS is particularly sensitive for determination of adsorbed CO. Therefore we used CO as a probe for determining propene adsorbates. We observed that some of the adsorbed propene can be displaced by CO, while the distribution of CO between different adsorption sites is significantly altered in the presence of propene in comparison with only CO saturated electrolyte.