High yield electrosynthesis of oxygenates from CO using a relay Cu-Ag co-catalyst system

As a sustainable alternative to fossil fuel-based manufacture of bulk oxygenates, electrochemical synthesis using CO and H2O as raw materials at ambient conditions offers immense appeal. However, the upscaling of the electrosynthesis of oxygenates encounters kinetic bottlenecks arising from the competing hydrogen evolution reaction with the selective production of ethylene. Herein, a catalytic relay system that can perform in tandem CO capture, activation, intermediate transfer and enrichment on a Cu-Ag composite catalyst is used for attaining high yield CO-to-oxygenates electrosynthesis at high current densities. The composite catalyst Cu/30Ag (molar ratio of Cu to Ag is 7:3) enables high efficiency CO-to-oxygenates conversion, attaining a maximum partial current density for oxygenates of 800 mA cm−2 at an applied current density of 1200 mA cm−2, and with 67 % selectivity. The ability to finely control the production of ethylene and oxygenates highlights the principle of efficient catalyst design based on the relay mechanism.

The authors designed Cu-Ag bimetal catalyst with Ag-Cu interfaces, which are crucial for CO-tooxygenates electrosynthesis, achieving a partial current density of 800 mA cm-2 of high faradaic efficiency of 67%.Also by in-situ characterization techniques and theoretical methods to confirm the mechanistic pathway, the authors showed impressive results.However, I would like to recommend this manuscript for publication, but only after the authors clarify satisfactorily on two major key points.
The authors have described the association between the *COH intermediate (formaldehyde) and the main product, acetic acid, using online DEMS and in-situ Raman analysis.However, in the Cu/30Ag catalyst configuration, ethylene also occupies a significant portion as a main product.The authors should additionally explain the high selectivity of *COH intermediate towards ethylene.
The authors assumed an Ag(100)/Cu(100) interface for the DFT and AIMD calculations and described their findings based on this assumption.However, when examining the XRD data in Figure 2a, it is evident that the Cu/Ag catalyst does not exhibit the crystallographic orientation of Ag(100) or Cu(100), but rather prominently displays Cu(111) and Ag(111) orientations.Therefore, in order to establish a connection between the DFT and AIMD calculation data and the experimental data, it is necessary to provide calculations for the Ag(111)/Cu(111) interface.
Reviewer #2 (Remarks to the Author): In this article, the authors investigate Cu-Ag nanocomposites for CO electroreduction to oxygenate products.Their premise is that CO adsorbs more strongly in the liquid phase on Ag than on Cu, contrary to trends in the gas phase.This has been demonstrated using TPD curves in the gas phase, LSV, and stripping experiments in the liquid phase.Therefore, if a composite Cu-Ag catalyst is used, CO adsorbs on Ag, which they claim can convert to COH* and is then transported to Cu, where it undergoes C-C coupling reactions.They used sequential deposition techniques to form the desired nanocomposites with different Cu-Ag ratios.Experimental characterization using XRD, STEM, HR-TEM shows no alloy formation is observed but the Cu nanoclusters coated by Ag.They also observe lattice fringes corresponding to the 111 facets of both metals in TEM.They then performed flow cell experiments (MEA) at high current densities to show that the FE for oxygenates is indeed high at 800 mA/cm2, with Cu/30Ag performing the best.The stability of the catalyst for 28 hrs shows no degradation of performance.To understand the mechanism behind the improvement, they performed DEMS and in-situ flow-cell experiments where they observed peaks corresponding to the formation of formaldehyde on Ag while the same was not observed on Cu or Cu/30 Ag.Through this, they conclude that Ag helps in the formation of formaldehyde-like species.However, they claim that on the interface even though formaldehyde like species is produced, they are consumed very fast leading to non-detection of these species.They then performed a few DFT/AIMD simulations to check the stability of COH on interfaces and the movement across it.The article makes interesting observations regarding the role of Ag, and the characterization and catalytic experiments look promising.However, the paper raises a lot of questions, especially in terms of mechanistic aspects and the DFT simulations which do not merit the direct publication of the work.
1.The authors, based on previous articles (ref. 26, 27 in the paper), make an assumption that CO to COH can be rate-determining for C2 formation instead of C-C coupling steps.Both the references that the authors cite do not definitively conclude that only CO to COH is RDS; they claim that CO to COH/CHO can be RDS for this reaction.However, the authors only consider the COH* formation as RDS, but following their logic, how do they explain the formaldehyde product detection from their DEMS and Raman experiments on Ag? Formaldehyde will be produced from the CHO pathway, not the COH (Energ Environ Sci 2010, 3 (9), 1311-1315).COH converts to carbon or CHOH and then leads to methane formation.I feel this assumption of COH formation as RDS cannot help in understanding or explaining the experimental results.
2. Based on the mechanism proposed, how do authors explain the increased selectivity to oxygenate products?If all the C-C coupling does occur on the Cu surfaces, why is there an increase in oxygenate product formation?Since Cu is inherently known to produce both ethylene and ethanol.
3. The authors should also consider going through the following paper and possibly cite it: Acs Catal 2020, 10 (7), 4059-4069.Even in this work, CuAg composite catalysts were used albeit for CO2 reduction with more detailed DFT simulations.4. Furthermore, the current set of DFT simulations in the paper adds minimal value to the study while a lot more can be done.

4(a)
. Firstly, the basic assumption of using Cu(100)|Ag(100) interface is not at all justified.The HRTEM in Figure 2(d) shows lattice fringes corresponding to Cu(111)|Ag(111).Why wasn't this interface simulated?The current Cu(100)|Ag(100) interface looks very sub-optimal to me; what is the lattice strain of the interfaces when they are placed side-to-side.

4(b)
. What is the difference in thermodynamic or kinetic barriers for COH/CHO formation from CO on Ag vs. Cu surfaces.Since CO hydrogenation step is claimed to be the RDS in this study and how does it compare to C-C coupling reaction energetics.4(c).The authors make an interesting claim through experiments in Figure 1 that CO adsorption can actually be stronger on Ag as compared to Cu in a liquid medium.Can this be proved using DFT simulations?4(d).In the AIMD simulations in Figure 5c, how can authors justify a change in energy of 6 eV?Furthermore, I cannot fathom COH* displacing so much and crossing the interface in just 2 ps of AIMD simulations.I would have expected the authors to use some accelerated AIMD techniques like meta dynamics or slow-growth simulations to simulate this process.
Reviewer #3 (Remarks to the Author): Herein, Meng et al successfully address a long-standing scholarly debate concerning CO capture and activation over Ag.Furthermore, leveraging this property, they propose a rationally designed strategy to employ Ag as a catalytic site for capturing CO and converting it into a COH* oxygenates precursor.This process creates a COH*-rich environment around Cu nanoclusters, serving as a highly efficient relay electrocatalyst for synthesizing oxygenates products.The effectiveness of Ag and the transfer of COH* from Ag to Cu are solidly demonstrated through CO stripping, in situ DEMS/Raman/XAFS experiments, robustly supported by Ab initio molecular dynamics simulations.By briefly controlling the areal density of Ag-Cu interfaces to match or balance the generation rate of the oxygenates precursor with its consumption rate, the record partial current density of 800 mA cm-2 with decent 67% selectivity in 1200 mA cm-2 CO electrolysis is reached.Significantly, through optimization of reaction parameters, this results in an approximate profit of 234 USD per tonne of acetic acid.As a result, this well-organized manuscript offers a novel understanding of the Cu-Ag catalytic system and contributes to enhancing overall electrolysis efficiency for oxygenates electrolysis.Given the significance of these findings and the solid experimental/mechanism analysis, I strongly recommend the publication of this manuscript in Nature Communications.Minor revisions are suggested to address the following issues.
1.For further reflecting the CO capture capacity of Ag in the composite, the CO stripping curve of the champion Cu/30Ag catalyst should be added.
2. The V-t curves over the champion catalyst under the different applied current densities (50 mA cm-2 to 800 mA cm-2) should be provided for further describing its stability.

Table of Contents (TOC) of this work can be drawn and provided.
4. To enhance the reproducibility of this study, the additional information regarding the electrode preparation process should be added.5. Page 10, line 22. '234 USD per tonne of the products' should be revised to '234 USD per tonne of the product' or '234 USD per tonne of the acetate acid product'.