Experimental insights into plasma-catalytic CO2 hydrogenation

Abstract

Electrifying the chemical industry with renewable energy is key to reducing anthropogenic CO₂ emissions, and non-thermal plasma is a promising technology in this context, though energy efficiency is hampered by poor product selectivity. Catalyst integration rarely improves selectivity in practice, leaving plasma catalysis far from competitive with conventional processes, largely because the mechanisms governing plasma–catalyst interactions remain poorly understood. This thesis addresses that gap experimentally, focusing on CO₂ hydrogenation to CO, CH₄, and methanol (CH₃OH) using dielectric barrier discharges (DBDs), the most common plasma type combined with catalysts (Chapter 1). Chapter 2 examines how surface oxygen vacancies (OVs) on CeO₂ affect plasma-catalytic CO₂ hydrogenation. CeO₂ samples with progressively higher OV concentrations are evaluated in a temperature-controlled packed-bed DBD reactor with detailed discharge characterization. Contrary to expectation, CO₂ conversion decreases with increasing OV concentration, paralleled by a decrease in microdischarge number and intensity. DFT calculations show OVs lower the CeO₂ work function, promoting fewer and/or less intense microdischarges, the primary driver of CO₂ activation in DBD plasma. Discharge characteristics thus depend sensitively on catalyst material and oxidation state, and non-selective gas-phase chemistry dominates over surface catalysis in conventional packed-bed DBDs. Chapter 3 presents a novel in-situ DBD transmission FTIR cell, fabricated mainly from PEEK, compatible with high-voltage operation, and improved over previous designs. The cell probes OV formation on CeO₂, Cu/CeO₂, and thermally prereduced Cu/CeO₂ (Cu/CeO₂-TR) under H₂/Ar plasma. On Cu/CeO₂-TR, a clear gradual increase in Ce³⁺ concentration indicates plasma-driven OV formation; switching to H₂/CO₂ plasma establishes a dynamic equilibrium between OV replenishment by H₂ and healing by CO₂. OV concentration thus emerges as a dynamic, condition-dependent property, underscoring the value of in-situ characterization. Chapter 4 critically reassesses the plasma-catalytic CO₂-to-methanol literature. Using direct online GC analysis, CH₃OH selectivity is negligible without a catalyst (<0.05%) and remains below 0.5% with Cu-based catalysts (Cu/Al₂O₃, Cu/MCM-41), two to four orders of magnitude below published values, cross-validated at the University of Manchester (~0.4% for the same catalyst). The indirect offline quantification method widely used in the literature consistently overestimates CH₃OH selectivity. The chapter proposes good practices for rigorous online quantification and discusses implications for economic viability. Chapter 5 offers a broader critical perspective, identifying four key challenges facing plasma catalysis. In-situ product removal is argued essential for suppressing gas-phase decomposition and improving energy efficiency, and the chapter closes with a discussion of which gas conversions plasma catalysis is most likely to succeed at.