Züttel, AndreasZhao, Kun2020-06-162020-06-162020-06-16202010.5075/epfl-thesis-7701https://infoscience.epfl.ch/handle/20.500.14299/169320Renewable energy supply and energy storage in a closed materials cycle are the urgent global challenges of the 21st century. Carbon dioxide (CO2) hydrogenation over catalysts is a method to produce synthetic fuels from renewable energy in a CO2 neutral cycle. Numerous catalysts from pristine to novel materials have been investigated to achieve high CO2 conversion and selective hydrocarbon. Ahead of developing innovatively active catalyst with a high selectivity, it is crucial to understand the active sites and the related reaction mechanisms. Despite the fact that CO2 + H2 is a rather simple chemical equation, many reaction paths on a solid catalyst are possible. However, the proposed mechanisms from literatures have been controversial due to two main difficulties in 1) evidently identifying the reaction species, especially some vital transient species on the pathway of carbon hydrogenation and C-C coupling, and 2) proving the relation between surface structure of the catalysts and the reactivities. To address these difficulties, we built a new instrumental setup and developed an analysis program in order to investigate the catalyst surface and analyze the reaction mechanism in operando. We built up a diffuse reflectance infrared Fourier transform spectroscopy-mass spectroscopy-gas chromatography (DRIFTS-MS-GC) instrument for an operando study of the surface, gas, and liquid products using DRIFTS, MS, GC, and ex situ NMR. A bilevel evolutionary Gaussi-an fitting (BEGF) program was developed for dataset treatment including peak deconvolution and kinetic plot. Standard chemicals like carbonates, bicarbonates, formic acid, acetic acid, methanol, and ethanol, and isotopic spectra using 13CO2 and D2 in the reactions were em-ployed for supporting peak assignments. Based on these infrastructures, CO2 adsorption and hydrogenation reactions on pristine metals (Fe, Co, Ni, and Cu), metal hydride alloy (La-Ni4Cu), and metal-oxide (Co-CoO) were studied. We found out pure Fe, Co, Ni, Cu and La-Ni4Cu metals are not efficient for CO2 hydrogenation, and have no detectable adsorption spe-cies on the surface. High conversion (> 90%) of CO2 on Co-CoO surface emphasized the im-portance of oxide for CO2 chemisorption and for the observations of the surface species. The following investigation on oxide supported metal catalyst Ru/Al2O3 via an in situ control of the individual formation and hydrogenation of each adsorption species demonstrates that the oxide assists CO2 initial activation. CO2 methanation starts at the interface of the metal and the oxide and passes through the steps of CO2 ï ® HCO3-* (and/or HCOO-* in CO2+H2 co-adsorption condition) ï ® CO* ï ® CH4. We further explored the relationship between met-al/(metal+oxide) ratio and their reactivities. We synthesized Co/(Co+CoO) catalysts in the DRIFTS-MS-GC instrument, and investigated the CO2 hydrogenation reaction on site. The results reveal that high concentration of the oxide in the catalyst boosts CO2 methanation, be-cause the CO2 binding is moderate and the adsorption is enhanced when the oxide concentra-tion increased. In short, we established equipment for an operando study of CO2 hydrogenation on catalyst surfaces, unraveled the surface reaction mechanisms and the active sites, developed a highly active catalyst, as well as initiated an efficient data analysis program.enCO2 hydrogenationheterogeneous catalysismethanationreaction mechanismsurface ad-sorptionadsorbatescatalystmetal oxideIR spectradata analysisthesis::doctoral thesis