Atomic-Scale Insights into Cu-Modified ZrO₂ Catalysts: The Crucial Role of Surface Clusters in Phenol Carboxylation with CO₂

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper provides insights regarding improving the catalytic activity of Phenol Carboxylation reaction using Cu dopants and small Cu clusters on ZrO₂ catalysts. It brings interesting perspectives in catalytic activity tuning with changes in reaction mechanism. However, there are some issues that needs to be addressed prior to publication.

Comments:

• Experimentally, what is the Cu⁰ size? My concern is that weather Cu₂₂ is big enough to represent the system.
• Authors should perform the stability tests for Cu²⁺ and Cu₂₂ cluster over the ZrO₂ If the stability is low, potential particle growth, such as sintering could possibly happen. More detailed discussion is recommended.
• In section 2.1, when describing the surface models, I recommend showing the surface models side by side, which can provide a direct view for the readers to clearly see the structures.
• When simulating ZrO₂ system, is there a reason that Hubbard U parameters are not considered to correct d-electrons delocalization, as suggested by Gebauer (https://doi.org/10.3390/cryst13040574)? I recommend including Hubbard U parameters to accurately capture the electronic structures of the systems, hence reaction energetics and kinetics. 
• I recommend generating potential energy surface (PES) or free energy diagram (FED) to map out the reaction mechanisms with both reaction thermodynamics and kinetic barriers, which provides direct view and comparison of the energetics of proposed reaction mechanisms.
• Is there electronic communication between the ZrO₂ and Cu single site or clusters, which enable the interfacial chemistry between Cu/ZrO₂ to improve catalytic activity? Density of states (DOS) analysis should be included.
• Based on the data presented, I agree that small clusters would be beneficial to Phenol Carboxylation; however, these small particles would feature high surface energy, it would be great to elaborate on how to stabilize these particles to ensure improved catalytic activity.

Author Response

Please see attached.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper uses DFT to compare Cu as a dopant in ZrO₂ versus Cu clusters on ZrO₂ for phenol→p-hydroxybenzoic acid carboxylation with CO₂.

Major revision. The core idea is interesting and potentially impactful. With the methodological clarifications (dispersion/free energies, slab thickness, spin/+U, convergence), better bookkeeping of H/protonation, and a small size-sensitivity or microkinetic check, the work will be much stronger and easier to trust.

Major comments, questions and suggestions:

1. There is clearly not enough information about the main material ZrO₂, the relevance is not disclosed, and it is especially unclear whether the authors are familiar with the latest trend of this material. See some recent publications:

Salikhodzha, Z.M. et al. Density Functional Theory Study of Pressure-Dependent Structural and Electronic Properties of Cubic Zirconium Dioxide. Ceramics 2025, 8, 41. https://doi.org/10.3390/ceramics8020041

Kozlovskiy, A. L., Konuhova, M., Borgekov, D. B., & Anatoli, P. (2024). Study of irradiation temperature effect on radiation-induced polymorphic transformation mechanisms in ZrO2 ceramics. Optical Materials, 156, 115994. https://doi.org/10.1016/j.optmat.2024.115994

2. You omit dispersion corrections and ZPE/entropy “as only adsorbed phases were considered” (2.2, lines ~114–116). For weakly bound phenol and interfacial CO₂, dispersion can materially change adsorption strengths, geometries, and thus barriers; entropic effects matter for comparing ER vs. LH. Please (i) add PBE-D3 (or similar) checks for key states and (ii) report ΔG(…T) for the rate-limiting steps, or at least provide a sensitivity analysis.
3. Pure PBE on ZrO₂/Cu can misplace levels; you don’t mention DFT+U or spin polarization. Please justify not using +U on Zr (and possibly Cu), and state whether spin-polarized calculations were performed; if so, provide magnetic moments and confirm that the reported TS/IS are the ground-spin states. (2.2, lines ~103–113.)
4. You use a 400 eV cutoff and Γ-only (1×1×1) for the enlarged Cu@ZrO₂ cell (2.2, lines ~106–109). For Zr PAW and metallic Cu, 400 eV can be borderline. Please show convergence tests (cutoff, k-mesh) for representative barriers (e.g., LH step 1 and 2 on Cu@ZrO₂).
5. Both surfaces include an O vacancy from earlier work (2.1, lines ~86–98), but the vacancy already creates highly active sites and will interact with Cu. Please justify the vacancy density and show its location relative to dopants/clusters. Did you test pristine surfaces or different vacancy positions?
6. The pure and doped slabs use 4 layers; the cluster model uses 2 layers “to save computational time” (2.1, lines ~98–101). Two layers can allow spurious polarization and substrate softness, especially at an interface driving reactivity. Please repeat key barriers with ≥4 layers and report dipole corrections/vacuum thickness. Also clarify whether bottom layers were fixed; you state all atoms relaxed “to ensure symmetry,” which reads unusual for asymmetric slabs.
7. You conclude ER is disfavored on Cu-enriched ZrO₂ and LH dominates, with C–H cleavage as RDS (3.3; Table 1). However, barrier differences are within a ~0.5–0.7 eV window and could shift with dispersion/entropy. Please provide a minimal microkinetic analysis (coverages, TOF trends) at a plausible T and partial pressures to substantiate the RDS and the ER/LH crossover. (3.1–3.3; Table 1.)
8. You replace a surface Zr with Cu at very low nominal doping (1:15, lines ~90–92). With an O vacancy present, Cu may favor different coordinations/oxidation states (Cu⁺/Cu⁰). Please report Bader (or similar) charges/spins for Cu and nearby O/Zr, and clarify whether Cu binds to vacancy sites or segregates to the interface in the cluster model. (2.1–2.2.)
9. Only Cu₂₂ is studied (lines ~93–101). Please add at least one other cluster (e.g., Cu₁₃ or Cu₅₅) to show the robustness of the “interfacial C–C coupling barrier drop” with cluster size/shape. Alternatively, provide an argument from adhesion/strain and show that the identified interfacial motif persists across sizes.
10. You attribute trends to d-band shifts and “geometric distortion,” but do not show DOS/PDOS or d-band centers for your models (3.3, lines ~231–240, 246–258; refs. 36–38). Please add PDOS/d-band analyses (Cu sites, interfacial Cu/ZrO₂, aryl radical bonding) to support the narrative.
11. Table 1 includes numbers for “pure ZrO₂” from earlier work (ref. 20). Please confirm that the same slab, vacancy density, functionals, and numerical settings were used here; if not identical, highlight differences and provide recalculated key barriers for apples-to-apples comparison. (3.3; Table 1; ref. 20.)

Author Response

Please see attached.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I have gone through the updated manuscript, and I'm satisfied with the revision, which has addressed the previous comments, so I recommend for publication.

 

Reviewer 2 Report

Comments and Suggestions for Authors

after revision this paper can be accepted