Approaching 100% Selectivity at Low Potential on Ag for Electrochemical CO2 Reduction to CO Using a Surface Additive

Summary¶

The work combines electrochemical measurements on polycrystalline Ag foil modified by drop-cast quaternary ammonium bromide salts with reactive molecular dynamics (RMD) using ReaxFF to relate CO2 reduction selectivity to interfacial structure and local CO2/H2O availability at −0.8 V versus RHE.

Methods¶

MD application (reactive ReaxFF)¶

Engine / code: ReaxFF reactive molecular dynamics (RMD) as described in the article, following a prior Cu–electrolyte interfacial workflow with parameters transferred to Ag (per ACS Catal.). Specific package name (N/A — not quoted in this note; ACS Catal. / SI for LAMMPS-style implementation details if reported).
System size & composition: Aqueous electrochemical Ag interface models including H₂O, CO₂, and Br⁻; charge-neutral variants with about 26 Br⁻ (and 24–25 Br⁻ variants reported as similar) plus associated quaternary ammonium cation arrangements from the modifier series.
Boundaries / periodicity: Slab/interfacial periodic cells implied by the standard RMD setup (full boundary prescription in pdf_path/SI).
Ensemble / temperature: NVT at 298 K; 2 ns equilibration (reported as sufficient in the paper).
Timestep: N/A — not restated in this note (see pdf_path/SI).
Duration / stages: 2 ns equilibration; production/analysis segments as in the full Methods/SI; additional staging N/A — not itemized in this short summary.
Thermostat / barostat: Thermostat as implemented for NVT in the RMD setup (N/A — type string not restated in this note). N/A — NPT barostat: constant-volume interfacial sampling, not a bulk hydrostatic pressure-controlled segment.
Pressure: Interfacial effective pressure / stress in the classical RMD is not a direct stand-in for electrochemical potential; treat electrochemical bias in experiments separately.
Electric field / bias in MD: RMD is used with an implicit electrochemical framing in the text; the classical trajectory does not embed the full −0.8 V RHE bias as an explicit efield (N/A — see ## Limitations and pdf_path for the stated modeling approximation).
Widom / post-processing: CO₂ local distributions near the interface are analyzed (including Widom test-particle style sampling after equilibration as described in the paper). H-bond coordination used a 0.4 nm cutoff; interfacial tension from RMD is compared to measured contact angles (SI).

Experiments and characterization¶

Chronoamperometry / product analysis: 35 min chronoamperometry in a previously reported cell, extracting Faradaic efficiency and partial currents for CO and H₂ at controlled potentials.
GIWAXS / wettability: GIWAXS on drop-cast quaternary ammonium modifiers; NMR of rinsed electrodes (SI); contact angle goniometry (SI) linked to the MD surface-tension discussion.

Force-field training and standalone QM in this work¶

N/A as a new ReaxFF reparameterization article—the study reuses/extends an established interfacial ReaxFF setup for Ag from a prior Cu workflow, with DFT/continuum details for training referenced in the main text rather than re-derived here.

Findings¶

With 2-C16 (dihexadecyldimethylammonium bromide), FE_CO reaches about ~97% at ~−0.75 to −0.8 V versus RHE while the geometric CO partial current rises roughly ~9× (e.g. 0.14 → 1.21 mA·cm⁻² in a quoted comparison) and H₂ partial current drops to ~0.001 mA·cm⁻² from ~0.44 mA·cm⁻² (values as tabulated/quoted in the article). Longer C16 chain modifiers improve CO selectivity relative to C10 in ways the authors argue exceed contact-angle hydrophilicity alone.

Simulation/experiment link: RMD with 2-C16 is associated with higher interfacial CO₂ population (more CO₂ in the first interfacial monolayer and out to about 4 nm versus other modifiers) while maintaining enough H₂O to service HOCO*-mediated CO formation without excess water that would drive the HER branch. 2-C16 adopts a bilayer with inward-facing headgroups that can concentrate CO₂; a 1-C16-class arrangement forms a monolayer with chains toward the metal. GIWAXS is used to connect larger d-spacing and more gauche content for 2-C16 to looser packing and higher interfacial CO₂ solubilization.

Caveat / limitations: RMD omits a full first-principles treatment of the applied bias; numerical settings for NVT MD and 2 ns staging should be verified in pdf_path before quantitative reuse (see ## Limitations).

Limitations¶

The atomistic models treat a simplified electrolyte interface and finite simulation times; electrochemical potentials are not fully replicated in the classical reactive MD framework.

Relevance to group¶

Co-authors include W. A. Goddard III; ReaxFF is used to interpret electrochemical interface structure and selectivity.