Electrochemical reduction of carbon dioxide to liquid fuels : conversion of a thermal catalyst to an electrocatalyst

Thesis (PhD (Chemistry))--University of Pretoria, 2020.Having been a hot topic for some time, the interest in recycling carbon dioxide to renewable liquid fuels or other valuable chemicals has rocketed since the adoption of the Paris Agreement on Climate Change. This is due to the EU ruling that from 2020, a considerable fraction of renewable fuel of non-biological origin has to be added to gasoline and the commitment of large air carriers like UA to go 50% carbon neutral by 2050. The primary novelty of this thesis was the development and conversion of the thermal catalyst indium oxide to an electrocatalyst that could do the conversion of formic acid and CO2 at ambient conditions with water as the only hydrogen source and the second starting compound. Here, the synthesis of indium oxide (In2O3) and supported iridium oxide (IrO2) electrocatalysts were done in-house. The crystallinities and average particle size characterizations were examined via powder X-ray diffraction. Scanning electron microscopy was used to study the surface morphologies of both electrocatalysts. Three different anodic electrocatalysts including 60:40 wt% IrO2:TaC, 70:30 wt% IrO2:TaC and 100:00 wt% IrO2:TaC were fabricated and employed for water electrolysis, with 70:30 wt% IrO2:TaC demonstrated to be of superior electrochemical activity and further employed for subsequent studies. Currently, In2O3 is the best thermal catalyst for methanol formation from CO2 . Here, the as-synthesized thermal catalyst was converted to a cathodic electrocatalyst by firstly making electroconductive material in a nanosize form with very small crystallite grains, which contain numerous defects near the surface; thus, making it more conductive. This was used to prepare high-performance membrane electrode assemblies (MEAs). The reaction cell containing the MEA that was set up by spray-coating the respective catalyst inks onto either Nafion® or a carbon gas diffusion cloth. The PEM electrolysis cell configuration with Nafion® as the polymer electrolyte was used. It minimizes Ohmic losses, and a standard TaC-supported IrO2 water-splitting catalyst served as the anode, titanium mesh served as anodic gas diffusion layer, and the experiments were conducted at ambient temperature. The cathode consisted of In2O3 spray-coated on carbon paper which acts as a gas diffusion layer and titanium mesh current collector. The cathode electrocatalyst was enhanced by the addition of a small amount of polytetrafluoroethylene to the nanosized In2O3 to facilitate diffusion of FA and CO2. The electrochemical characteristics were examined via cyclic voltammetry, linear sweep voltammetry and chronoamperometric methods. The infrared spectroelectrochemical cell was also used because it permits in-situ analysis of the change of reactant concentrations and ideally the identification of intermediates Addition of PTFE to the In2O3 electrocatalyst layer for FARR has led to significant improvement in current density from 1.94 mA/cm2 (without PTFE) to 66.0 mA/cm2 (with 0.15 wt% PTFE) and 70.3 mA/cm2 (with 0.30 wt% PTFE) which is a factor of ca. 34 and ca. 36 respectively at 2.4 V cell voltage. This further reduces the onset potential of the electroreduction by 0.4 V and notably, the cell Ohmic resistance was reduced by a factor of 15, implying that the activation energy of the electrode and the transport resistance in the porous structure are significantly reduced. This is due to the increase in the hydrophobicity in the porous electrocatalyst layer. The Tafel slope was also used to investigate the electrochemical reaction of water splitting, co-electrolysis of 4.30 M formic acid and water on In2O3 and PTFE-In2O3 cathodes. Tafel values of all the samples over the respective number of LSV cycles were consistent with each other. Tafel analysis of the PTFE-In2O3 electrode improves significantly with the lower Tafel slope in comparison with the PTFE-free In2O3 electrode. The steady-state current density experiment in the absence of any flow showed excellent stability over the investigation period. A current density observed to be limited to ca. 26 mA/cm2 in the absence of any flow over 24 h from the initial current density of 70.3 mA/cm2, the limitation is a result of FA transport across the diffusion layers in the electrocatalyst surface. This behavior was further investigated using the Cottrell equation and this was observed to qualitatively reproduce the experimental behavior, thereby confirming a diffusion layer that builds up, resulting in a reactant depletion near the electrode surface. For a long time, it was thought that formic acid is a dead-end that does not lead to larger product molecules. For the first time, the co-electrolysis of water and aqueous 4.30 M formic acid, the first stable intermediate of CO2 electroreduction, results in a mixture of methanol, ethanol and isopropanol with a maximum combined Faraday efficiency of 82.6% at 3.5 V and a space-time-yield of 0.431 g(alcohol)/h/g(cat) that compares well with results from heterogeneous catalysis. It was further discovered here that high Faraday efficiency of the alcohols and current density can be achieved under a relatively low overpotential by tuning the amount of PTFE used. FTIR spectroelectrochemistry was used to monitor the disappearance of FA and the formation or disappearance of CO2 reaction intermediates as a function of time and potentials. The consumption of FA propelled significant decreasing of absorption of up to 6 vibrational modes in the observation window including bands: at 3670 cm-1 belonging to the O-H stretching vibration, 3037 cm-1 assigned to the C-H stretching mode, 2120 cm-1 attributed to the C=O stretching mode, a double band near 1667/1589 cm-1, assigned to the vibrational modes with major FA C-O stretching character, and finally one at 1225 cm-1 which are somewhat higher than corresponding literature values, suggesting interactions with the catalyst and the presence of the aqueous environment. In the experiment performed with CO2 catholyte on PTFE-In2O3 (in the absence of FA), the CO2 band disappears as expected with no FA build-up, suggesting that formic acid is bypassed as an intermediate. An additional convincing difference was that while R/R0 is >1 dominated by the FA disappearance and CO2 formation in FA catholyte, it is