Role of ZrO2 and CeO2 support on the In2O3 catalyst activity for CO2 hydrogenation

•ZrO2 supported In2O3 highly stable for CO2 hydrogenation.•CeO2 supported catalyst deactivates and only partly reversible.•Accumulation of OH groups deactivate CeO2 supported In2O3.•Deactivation due to structural changes is irreversible. Methanol synthesis from CO2 hydrogenation has drawn global attention as catalytic CO2 hydrogenation is an attractive choice to mitigate CO2 emissions and lessen dependency on fossil resources. In the present study, we have synthesized ZrO2 and CeO2-supported In2O3 catalysts for methanol synthesis from CO2 hydrogenation and the catalytic performances of Inx/ZrO2, and Inx/CeO2 (x = 1 % and 13 %) were compared. Specifically, the effect of the ZrO2 and CeO2 supports on In2O3 catalyst during CO2 hydrogenation was explored. This study reveals that ZrO2 support increased the catalytic activity while the CeO2 support decreased although both supports have almost the same indium loading and surface area. Various characterizations like XRD, DRIFT, CO2-TPD, H2-TPR and XPS analysis of catalysts provided insight into changes that arise after mixing the two oxides and during the reaction as well as after the reaction. The stabilities of In2O3, In13/ZrO2, and In13/CeO2 were tested for up to 50 h and we found In13/ZrO2 was stable during this time-on-stream, while In13/CeO2 lost activity after 2 h of reaction. XPS results of spent catalysts showed that In(OH)3 was observed significantly over the spent In13/CeO2. OH groups were also verified by DRIFT experiments, however at low levels due to low CO2 conversion at atmospheric pressure. XRD analysis confirmed the sintering of CeO2 support during the reaction. Thus, the hydrophilic nature of CeO2, redox properties of CeO2 and sintering of CeO2 support in the presence of water, were the main reasons for the early deactivation of In13/CeO2. A regeneration study was carried out to regenerate the catalyst and the results showed that partial regeneration of the In13/CeO2 catalyst is possible by Ar flushing. We, therefore, suggest that the build-up of OHgroups deactivate the In13/CeO2 catalyst and some of these OH groups could be removed during flushing with inert gas, causing a partial regeneration. However, the decreased surface area is not reversible, and this results in a continuous decrease in the activity of the catalyst after repeated experiments, even if the catalyst is flushed with Ar between the experiments.