Access Official Technical PDF

Share Summary:LinkedInEmail

In CO electroreduction, the oxygen found in a liquid product does not necessarily come from the carbon monoxide feed. Isotope-labeling experiments on copper electrodes show that solvent water can contribute oxygen atoms to products including ethanol and 1-propanol. For engineers and researchers assessing copper electrocatalysts, this distinction matters: product identity alone does not reveal how a reaction proceeds or which feedstock supplied each atom.

The work used water enriched in oxygen-18 (18O) and analyzed the products formed during CO reduction. By tracking the isotope in oxygen-containing products, the researchers could test whether solvent water participated in product formation. A series of control experiments addressed other possible sources of isotope incorporation, including exchange after a product had already formed.

What oxygen-18 labeling reveals about CO electroreduction

In an isotope-labeling experiment, a particular isotope acts as a tracer. Here, the electrolyte solvent was enriched with 18O, while the CO feed contained predominantly 16O. If a product contained 18O, that was evidence that its oxygen atom had come from the solvent rather than being retained from the CO molecule. The team identified oxygen-containing products using gas chromatography–mass spectrometry (GC–MS), examining mass fragments associated with the different isotopes.

CO reduction was conducted on copper surfaces with defined orientations, including Cu(100), Cu(111) and Cu(751), and under different electrolyte and potential conditions. The researchers also used nuclear magnetic resonance (NMR) to confirm methanol formation on Cu(111). This combination of product identification and isotope analysis provided a way to investigate not just what formed, but how oxygen could enter the products.

Which products incorporated oxygen from water?

The isotope results varied by product and reaction condition. Ethanol and 1-propanol showed substantial 18O incorporation in the reported experiments, indicating that solvent water contributed oxygen during their formation. In a phosphate-buffered experiment at pH 7.2, for example, the measured 18O fractions were 45.6% for ethanol and 51.5% for 1-propanol. These observations supported a reaction route involving water-derived oxygen, rather than an explanation based mainly on a secondary reaction between products in the electrolyte.

Not every oxygenate behaved the same way. The experiments did not detect 18O in methanol or allyl alcohol under the reported conditions. That difference is important: oxygen incorporation should not be treated as a universal feature of all products made during CO electroreduction. Product pathways can depend on the specific oxygenate and experimental conditions.

Acetate also requires careful interpretation. The researchers observed that acetic acid could exchange oxygen with water slowly, particularly after acidification for GC–MS analysis. Consequently, isotope measurements for acetate could be affected by sample handling. The team analyzed samples within a limited period after acidification to reduce this concern. This control illustrates why analytical preparation must be considered when interpreting isotope data.

Controls help separate reaction chemistry from isotope exchange

Several alternative explanations could have produced an apparent 18O signal without water contributing directly during electrochemical product formation. The experiments tested these possibilities:

  • CO–water exchange: CO passed through 18O-enriched water did not show evidence of exchanging its oxygen isotope with the solvent.
  • Exchange after product formation: Ethanol and 1-propanol showed no significant oxygen exchange with 18O water in the tested alkaline solutions, even after extended exposure. Acetic acid showed slow exchange, making its analysis more sensitive to handling time.
  • Electrode-mediated exchange: The tests did not support oxygen exchange between pre-existing ethanol, propanol or acetate and the labeled water simply because an electrical potential was applied.
  • Secondary aldehyde reactions: Experiments using phosphate buffer, which suppresses Cannizzaro-type disproportionation, still showed substantial 18O incorporation into ethanol and propanol. This indicated that such reactions were not the main explanation for the observed labeling.
  • Carbonate as an oxygen source: Tests with different electrolytes and labeled CO feeds did not support carbonate or bicarbonate as the reactant responsible for producing the labeled oxygenates.

The researchers also checked whether the mass spectrometer responded differently to 16O- and 18O-containing fragments. Mixtures with known product compositions gave agreement between predicted and measured isotope fractions, supporting the reliability of the GC–MS comparison.

Implications for copper catalyst pathways

The isotope findings, combined with computational analysis, point to more than one pathway for forming ethanol on Cu(100). In the proposed mechanism, a shared surface intermediate can proceed along branches that lead to different outcomes, including ethanol with oxygen retained from CO or ethanol with oxygen supplied by water. The calculations compared barriers for these pathways, and the predicted preference was consistent with the experimental product ratio reported for Cu(100).

For industrial automation and process development teams, the broader lesson is methodological rather than a ready-made operating recipe. Product selectivity data can describe what an electrolyzer produces, but isotope tracing and appropriate controls can reveal more about how those products form. That mechanistic understanding can help teams evaluate how electrode structure, electrolyte composition, pH and applied potential relate to product distributions—while recognizing that results observed on defined copper surfaces do not by themselves establish performance in a larger-scale system.

Practical takeaway

Oxygen-18 labeling shows that water can act as an oxygen source in the formation of selected oxygenates during CO electroreduction on copper. The product-specific results, together with controls for isotope exchange and analysis bias, make the case for treating reaction pathways as part of catalyst evaluation—not assuming that product composition directly identifies the origin of each atom. For engineers and technical decision-makers, that distinction is essential when interpreting selectivity data and assessing how copper electrocatalyst conditions shape oxygenate formation.

Access Official Technical PDF

About The Author