# Ni-SAC-2.5 CO2-to-CO performance: first research briefing

## Executive summary

Yang et al. report that Ni-SAC-2.5, a carbon-supported nickel single-atom catalyst containing 2.5 wt.% Ni, reached a peak CO Faradaic efficiency (FE_CO) of 98.9% at −1.2 V versus the reversible hydrogen electrode (RHE) in CO2-saturated 0.1 M KHCO3. This is an electron-selectivity result: 98.9% of the measured charge was assigned to CO formation under that test condition. It is not a reported energy efficiency and does not show that 98.9% of the supplied CO2 was converted. The peak-selectivity claim is supported by the potential-dependent FE_CO measurement in main-text Figure 6b and the loading comparison in Figure 6e (paper pp. 6–7).

Durability was tested separately at −0.8 V versus RHE. Figure 6c shows only a slight decrease in current density over 20 h, while Figure 6d compares Ni K-edge XANES before and after CO2 reduction and reports nearly unchanged spectra (paper pp. 6–7). These observations support short-duration retention of electrochemical current and average Ni electronic/coordination-state similarity under the stated test, but they do not establish 20 h retention of the 98.9% FE_CO measured at −1.2 V. The source does not report energy efficiency, CO2 single-pass conversion, carbon utilization, or industrially relevant full-cell performance.

**Source:** Yang et al., “A universal ligand mediated method for large scale synthesis of transition metal single atom catalysts,” *Nature Communications* 10, 4585 (2019), [DOI 10.1038/s41467-019-12510-0](https://doi.org/10.1038/s41467-019-12510-0).

## Evidence table

| Question | Source-supported observation | Interpretation | Locator |
|---|---|---|---|
| What is the peak CO selectivity? | Ni-SAC-2.5 reached FE_CO = 98.9% at −1.2 V vs RHE. The authors also state FE_CO remained above 90% from −0.7 to −1.5 V vs RHE. | Strong selectivity toward CO within the reported half-cell conditions and potential range. FE_CO allocates electrical charge among detected products; it is not energy efficiency or CO2 conversion. | Paper p. 6, Fig. 6b; paper p. 7, Fig. 6e; DOI above |
| Were competing products assessed? | H2 FE was reported as low across potentials; no liquid products were detected by NMR after CO2RR. | The reported product analysis supports CO as the dominant detected reduction product. It does not by itself establish a complete carbon balance or quantify unreacted CO2. | Paper p. 6; Supplement pp. 47–48, Supplementary Figs. 51–52 |
| How was durability tested? | At −0.8 V vs RHE, current density decreased only slightly over 20 h. | This supports current retention at −0.8 V for 20 h. It must not be combined with the 98.9% FE_CO measured at −1.2 V. | Paper p. 6, Fig. 6c |
| Was catalyst state checked after testing? | Ni K-edge XANES spectra before and after CO2RR were described as almost identical. | This supports similarity of the ensemble-averaged Ni XANES signal before and after testing, but is not a complete demonstration that every active site or the carbon support remained unchanged. | Paper pp. 6–7, Fig. 6d |
| What was the electrochemical configuration? | A sealed three-electrode H-type cell used a catalyst-coated glassy-carbon working electrode, Ag/AgCl reference, Pt-wire counter electrode, and 0.1 M KHCO3. FE tests held constant potential for 30 min; gaseous products were analyzed by gas chromatography. | The evidence is a laboratory half-cell assessment. It does not directly establish device-level voltage efficiency, power efficiency, or practical reactor conversion. | Paper p. 8, Electrochemical measurements |
| How does Ni loading affect selectivity at −1.2 V? | FE_CO values were 98.9%, 98.5%, 98.0%, and 95.3% for Ni loadings of 2.5, 3.4, 4.5, and 5.3 wt.%, respectively. | Within this series, the 2.5 wt.% sample gave the highest reported FE_CO; the authors associate the lower value at 5.3 wt.% with small Ni clusters. | Paper p. 7, Fig. 6e; structural context on p. 5, Fig. 5 |

## Metric distinctions

**Faradaic efficiency to CO (FE_CO)** is the fraction of total transferred charge accounted for by detected CO production. It answers, “Where did the electrons go?” A high FE_CO indicates suppression of competing electrochemical products such as H2 under the stated conditions.

**Energy efficiency** would additionally account for the electrical energy input, including operating voltage or overpotential and relevant thermodynamic potentials. The paper does not report an energy-efficiency value, so FE_CO = 98.9% must not be restated as 98.9% energy efficiency.

**CO2 conversion** is the fraction of CO2 feed consumed, often specified as single-pass conversion in a flowing reactor. Neither FE_CO nor the absence of detected liquid products determines CO2 conversion. The paper does not report feed flow, outlet CO2 balance, single-pass conversion, or carbon-utilization efficiency.

## Methodological limits

The performance evidence comes from a three-electrode H-type half-cell rather than a membrane-electrode assembly or industrial flow cell. Reported FE measurements used 30 min constant-potential holds, while the durability trace monitored current for 20 h at a different potential. The text and displayed evidence do not establish that FE_CO remained constant throughout the 20 h run.

The paper does not report energy efficiency, cell voltage for a complete electrolyzer, CO2 single-pass conversion, carbonate losses, full carbon balance, or long-duration operation beyond 20 h. It also does not present, in the cited performance section, replicate counts, statistical uncertainty, or error bars sufficient to quantify run-to-run reproducibility. Current retention and similar before/after XANES are useful but limited durability indicators: they do not exclude local restructuring, minority-site changes, support corrosion, catalyst detachment, or performance decay on longer timescales.

The authors report gas chromatography for gaseous products and NMR showing no liquid products, which supports the product-selectivity claim. However, without a complete mass balance and explicit detection limits, “no liquid products detected” should not be read as proof that every possible carbon product was absent. Likewise, the large-scale synthesis demonstration and similar catalytic properties of a kilogram-scale batch concern manufacturability and batch performance; they do not substitute for reactor-scale CO2 conversion or process-energy data.

## Bottom line

The defensible teaching claim is narrow: under the reported H-cell conditions, Ni-SAC-2.5 achieved a peak FE_CO of 98.9% at −1.2 V vs RHE, and a separate test at −0.8 V vs RHE showed only slight current-density loss over 20 h with nearly unchanged before/after Ni K-edge XANES. The study supports high electrochemical selectivity and short-term current retention, while leaving energy efficiency, CO2 conversion, carbon utilization, full-cell performance, statistical reproducibility, and longer-term durability untested or unreported.