Carbon In, Chaos Out

Electrochemical CO2 reduction (CO2R) could convert captured carbon dioxide into valuable fuels and chemical feedstocks (ethanol, ethylene, formate, methanol) using renewable electricity, closing the carbon cycle. However, CO2R on copper — the only known catalyst that produces multi-carbon products — generates an uncontrollable mixture of at least 16 different products simultaneously. At the current densities required for industrial viability (>200 mA/cm2), selectivity toward any single desired product rarely exceeds 40-50%, making downstream separation prohibitively expensive. No known catalyst material can selectively route CO2 reduction toward a single target product at commercially relevant rates.

Global CO2 emissions exceed 36 billion tons annually. Electrochemical conversion powered by renewable electricity could simultaneously reduce atmospheric CO2 and displace fossil-derived chemicals. Ethylene alone represents a $200+ billion market currently produced entirely from petroleum. If CO2R could achieve >80% selectivity for ethylene at scale, it would transform the chemical industry's carbon footprint. Without selectivity, CO2R produces dilute mixtures that cost more to separate than the products are worth, making the entire value proposition collapse.

Copper is the only monometallic catalyst that reduces CO2 beyond CO or formate to multi-carbon products, but its lack of selectivity appears fundamental to its mechanism: the same active sites that enable C-C coupling also facilitate competing pathways to methane, hydrogen, and oxygenates. Copper nanostructuring (nanocubes, nanowires, oxide-derived surfaces) shifts product distributions modestly but does not solve the selectivity problem. Bimetallic catalysts (Cu-Ag, Cu-Zn) can suppress some pathways but introduce new ones. Molecular catalysts achieve high selectivity for CO or formate but cannot produce the multi-carbon products that have the most economic value. Gas diffusion electrodes enable high current densities but exacerbate selectivity problems due to local pH gradients and CO2 mass transport limitations. Computational screening of catalyst surfaces has identified promising candidates that often fail to translate from DFT predictions to experimental reality because the models don't capture the complex solid-liquid interface under operating conditions.

Understanding and controlling the CO2R reaction mechanism at the atomic level under operando conditions — particularly the C-C coupling step that determines whether the pathway leads to ethylene versus ethanol versus propanol — would enable rational catalyst design. Single-atom or single-site catalysts that present a uniform active site geometry could enforce a single reaction pathway. Tandem catalysis strategies that spatially separate CO2-to-CO and CO-to-C2+ steps could decouple the competing selectivity requirements. Machine learning guided by high-throughput experimentation could efficiently navigate the vast composition-structure-condition parameter space.

A student team could build an electrochemical flow cell with online gas chromatography to systematically map how CO2R product selectivity on copper changes with electrolyte composition, pH, and flow rate — generating datasets that are currently sparse and poorly standardized across the literature. Alternatively, a team could use density functional theory to screen a family of bimetallic surface alloys for C-C coupling barriers and validate top candidates experimentally. Relevant disciplines include electrochemistry, materials science, chemical engineering, and computational chemistry.