Size-Activity Relationship of TiO₂-Supported Pt Nanoparticles in Hydrogenation Reactions

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Quantitative evaluation of the effect of shape on catalytic activity is extremely important for catalyst design and understanding of reaction mechanisms, but it is generally difficult to focus only on the size of catalyst particles.
The most important prerequisite is that all particles have the same surface properties, but this is something that is not known in this study.
In other words, there is no information that the electronic state is the same and the particles are not at all aggregated, so there is no guarantee that they have the same properties.
There is also a problem with the definition of TOF. Since it is defined simply by the molar amount of Pt, it seems to assume, needless to say, that the particle system is free of aggregation or coagulation.
Such a TOF cannot be used unless it is proven.
In other words, I think that the TOF should be defined after showing that the crystallite size and apparent particle size are the same, and then the discussion should proceed.
On the other hand, it is extremely dangerous to argue based on the fact that a slope value close to -1 was obtained. It is merely a leap of logic to conclude from this that the catalytic activity comes mainly from the surface sites of Pt nanoparticles, rather than from active sites at the interface, edges, or corners.
A more calm discussion is needed, but before that, more detailed analysis of the active site structure is essential, for example through CO and hydrogen adsorption experiments.

Author Response

Reviewer 1#: We sincerely thank Reviewer 1 for their thoughtful and constructive comments. Your insightful suggestions regarding the assumptions behind particle uniformity and the careful interpretation of turnover frequencies (TOF) have provided us an invaluable opportunity to re-evaluate critical methodological details in our study. Your professional and precise feedback has significantly strengthened the rigor and clarity of our manuscript, guiding us toward a more nuanced and accurate presentation of our results.

Comment 1: “The most important prerequisite is that all particles have the same shape, but this is something that is not known in this study.”

Response: We thank the reviewer for underscoring the importance of particle-shape uniformity to our size–activity analysis. High-resolution TEM images for every size fraction (Fig. 1 in the main text and Fig. S1 in the SI) show smooth, quasi-spherical Pt nanoparticles; analysis of more than 200 particles yields an average long-to-short-axis ratio of 1.07 ± 0.05, with no faceted, rod-like, or platelet-like morphologies observed. All five size populations were synthesized via the same polyol colloidal route under identical surfactant and reduction conditions, thereby minimizing shape variation. The absence of sintering necks or irregular aggregates in TEM, together with a single-phase XRD pattern, further corroborates morphological uniformity.

We now state explicitly on Page 2, line 79-82: “All five nanoparticle size fractions, each displaying sub-spherical Pt morphology, were prepared by the same polyol colloidal route under identical surfactant and reduction conditions, and TEM and XRD confirms the absence of sintering necks or irregular aggregates in every sample.”

Comment 2: “In other words, there is no information that the electronic states of the Pt nucleus are similar between particles, or that the particle system is free of aggregation or coagulation.”

Response: We thank the reviewer for bringing up the electronic state and aggregation issues. All Pt/TiO₂ samples were synthesized by the same polyol colloidal route and handled under identical surfactant, reduction, and deposition conditions, ensuring comparable surface chemistries. Newly added X-ray photoelectron spectroscopy data (Pt 4f region, see Fig. 6) show virtually identical binding energies for the metallic component (Pt⁰ ≈ 71.2 eV) and the slighty oxidized shoulder (Pt^δ+ ≈ 72.0 eV) across the entire size series, with shifts of ≤ 0.1 eV—well within instrumental error—indicating that the electronic states of the Pt cores are the same for all catalysts. This point has been incorporated into the revised manuscript (Fig. 6, Page 4, lines 171–174) to clarify the electronic uniformity of the catalysts. “XPS analyses were also carried out, and peak-deconvolution of the Pt 4f region (Fig. 6) shows virtually identical binding energies for both metallic Pt⁰ (4f7/₂ ≈ 71.2 eV) and the slightly oxidised Pt^δ+ component (4f7/₂ ≈ 72.0 eV) across all Pt/TiO₂ catalysts. This uniform valence-state distribution further substantiates our assumption of consistent particle morphology and comparable per-site activity.”

Regarding aggregation, we added the corresponding discussion to the revised manuscript; please see Response 1 for details.

Comment 3: Since it is defined simply by the molar amount of Pt, it seems to assume, needless to say, that the particle system is free of aggregation or coaqulation. Such a TOF cannot be used unless it is proven, In other words. l think that the TOF should be defined after showing that the crystalite size and apparent particle size are the same, and then the discussion should proceed. On the other hand. it is extremely dangerous to argue based on the fact that a slope value close to -1 was obtained. lt is merely a leap of logic to conclude from this that the catalytic activity comes mainly from the surface sites of Pt nanoparticles, rather than from active sites at the interace, edges.or corners. A more calm discussion is needed, but before that, more detailed analysis of the active site structure is essential, for example through CO and hydrogen adsorption experiments.

Response: Because not all Pt atoms are accessible to reactants—those buried beneath the nanoparticle surface, for instance—we report two turnover frequencies: TOF_overall, normalised to the total Pt loading, and TOF_real normalised to the number of surface-exposed Pt atoms. The slope of the ln(TOF)-ln(size of NP Pt) thus reflects how the reaction rate scales with the population of effective Pt sites. Relying solely on TOF_real (i.e., a TOF referenced only to crystallite-confirmed surface Pt) would obscure the purely geometric influence of particle size on the catalyst’s macroscopic activity. Although individual surface atoms may differ slightly in intrinsic activity, these variations are negligible compared with the much larger differences between surface and subsurface—or interface, edge, and corner—sites. We therefore consider the two assumptions underpinning ln(TOF)-ln(size of NP Pt) analysis, in Figure 7, to be well founded.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

 In this work “ Size-Activity Relationship of TiO2-Supported Pt Nanoparticles in  Hydrogenation Reactions “ a series of catalysts consisting of TiO₂-supported Pt nanoparticles with average sizes of approximately 2.32 nm, 4.38 nm, 9.63 nm, 14.64 nm, and 40.98 nm were synthesized using different preparation methods. The authors identified a correlation between the size of platinum particles and their catalytic activity in various hydrogenation reactions, as indicated by the analysis of turnover frequency (TOF) results. They concluded that the catalytically active sites are located on the surface of the platinum particles. Additionally, they suggested that the primary factor influencing catalyst performance is not the hydrogen spillover from platinum nanoparticles to the TiO₂ support.

I believe the article cannot be accepted for publication in this form.Here are my arguments:

1. The paper lacks a clear structure and is difficult to follow. I recommend reorganizing it for a more coherent and logical presentation.
2. To enhance understanding, many results presented in supplementary files should be included in the article (e.g. XRD characterization of Pt/TiO2, table of performances of hydrogenation by H₂ over Pt/TiO₂).
3. It is necessary to discuss the crystallinity of nanoparticles based on the size of platinum particles, as characterized by XRD.
4. The authors argue that the CO-DRIFTS presented in Figure 3 indicates that the platinum nanoparticles in this study exhibit relatively uniform adsorption behavior. This observation supports the hypothesis of uniform shape and activity. Linear CO adsorption signals were detected in the range of 2000–2100 cm⁻¹, with peaks observed at approximately 2088 cm⁻¹ and 2065 cm⁻¹ for CO adsorption on Ptδ⁺ and Pt⁰ sites, respectively. However, this argument is not strong enough to conclusively demonstrate the uniformity of the platinum particles' shape or the hydrogenation reaction mechanism. Additional characterizations, such as X-ray photoelectron spectroscopy (XPS) or H2-TPR, are necessary to determine the oxidation states of platinum and titanium.

5.      The turnover frequency (TOF) can be expressed in terms of the dispersion of the active component. The dispersion of platinum can be assessed through CO chemisorption. The authors calculate the TOF by relating the reaction conversion to the mole ratio of platinum to substrate. It is important to note that not all platinum centers are accessible for the reaction.6.       Therefore, additional selective H₂ chemisorption analyses are necessary for all catalysts to identify the active platinum centers.

Author Response

Reviewer 2#: We greatly appreciate Reviewer 2’s detailed and helpful critique. Your valuable suggestions concerning manuscript structure, data presentation, and additional characterization have been instrumental in highlighting key aspects needing clarification and improvement. Specifically, your recommendations to reorganize the manuscript and include supplementary data within the main text have prompted essential adjustments that will markedly improve the coherence and readability of our paper. We are genuinely grateful for your constructive feedback, which has substantially enhanced the overall quality and scientific rigor of our manuscript.

Comment 1. The paper lacks a clear structure and is difficult to follow. l recommend reorganizing it for a more coherent and logical presentation.

Response: We appreciate the reviewer’s observation that the manuscript’s structure could be clearer, and in response we have thoroughly reorganized the text. For example, (1) the literature survey on size-dependent Pt catalysis, including the work of Cargnello and Shekhar (now references 29–32), has been consolidated into Introduction part so that readers encounter the relevant context before the experimental details (see page 2, line 51-57). (2) The Results are now grouped into three clearly labeled subsections—(i) synthesis and structural characterization, (ii) electronic structure, and (iii) catalytic performance—followed by a dedicated Discussion that connects our findings to prior studies. We believe these changes render the manuscript far more coherent and easier to follow.

Comment 2. To enhance understanding, many results presented in supplementary files should be included n the article (e.g. XRD characterization of Pt/Ti02, table of performances of hydrogenation by H₂ over Pt/TiO₂).

Response: Thank you for the suggestion. In the revised manuscript we have moved the most pertinent supplementary data into the main text: the XRD patterns for all Pt/TiO₂ catalysts are now presented and discussed in Figure 2, and a concise table summarizing the H₂‑hydrogenation performance of representative substrates over 4‑Pt/TiO₂ appears as Table 2. We believe their inclusion greatly improves the readability and self‑containment of the article.

Comment 3. It is necessary to discuss the crystallinity of nanoparticles based on the size of platinumparticles.as characterized by XRD.

Response: Thank you for the suggestion. Because the Pt loading is only ≈ 1 wt % and the particles are nanometre-sized, the Pt reflections in the XRD patterns are extremely weak: only a faint, broadened Pt(111) shoulder at 2θ ≈ 39.8° is visible for the 15 nm and 41 nm samples, and it is too broad for reliable Scherrer analysis, while the smaller particles show no discernible Pt peaks at all. Consequently, we rely chiefly on HRTEM statistics to assess particle size and crystallinity.

Comment 4. The authors argue that the CO-DRlFTS presented in Figure 3 indicates that the platinum nanoparticles in this study exhibit relatively uniform adsorption behavior. This observation supports the hypothesis of uniform shape and activity. Linear CO adsorption signals were detected in the range of 2000-2100 cm-, with peaks observed at approximately 2088 cm-and 2065 cm- for CO adsorption on Pt* and Pt° sites, respectively. However, this argument is not strong enough to conclusively demonstrate the uniformity of the platinum particles shape or the hydrogenation reaction mechanism. Additional characterizations. such as X-ray photoelectron spectroscopy (XPS) or H2-TPR, are necessary to determine the oxidation states of platinum and titanium.

Response: Thank you for the suggestion. All Pt/TiO₂ samples were first reduced in 1 MPa H₂ at 200 °C for 1 h, after which H₂-TPR and XPS measurements (now included as Fig. S7) showed essentially identical reduction profiles and Pt 4f binding energies for every catalyst—a result that reflects the common pre-treatment rather than the particle morphology. CO-DRIFTS, however, is more site-sensitive: the linear-CO bands at 2088 and 2065 cm⁻¹ shift slightly but reproducibly with particle size, and the absence of bridged-CO features indicates that the nanoparticles are not agglomerated. Although CO-DRIFTS alone cannot prove perfect shape uniformity or elucidate the hydrogenation mechanism, when combined with TEM statistics it supports the conclusion that the Pt surfaces present comparable electronic environments across the series.

Comment 5: The turnover frequency (TOF) can be expressed in terms of the dispersion of the active component. The dispersion of platinum can be assessed through CO chemisorption. The authors calculate the TOF by relating the reaction conversion to the mole ratio of platinum to substrate. It is important to note that not all platinum centers are accessible for the reaction.

Response: Thank you for the suggestion. It is precisely because that not all platinum centers are accessible for the reaction, we correlated the TOF_overall and size of NP Pt through the effective Pt rate in total Pt atoms. If TOF was calculated based on numbers of effective Pt atoms, the TOFs of different sizes of NP Pt would present similar values. Furthermore, hypothetically speaking that all effective Pt atoms located interfacial sites, dispersion of platinum is obviously inappropriate to estimate the real TOF of effective Pt atom. If we talk about TOF based on dispersion of platinum assessed through CO chemisorption, TOF_surface might be a more suitable name. To avoid ambiguity and confusion, the TOF definitions in this work what we considered only involved TOF_overall and TOF_real, meanwhile avoided to regard TOF based on dispersion as TOF_real, although such cases are common but not universal.

Comment 6: Therefore, additional selective H, chemisorption analyses are necessary for all catalysts to identify the active platinum centers.

Response: Thank you for this suggestion. Because the Pt loading in every catalyst is deliberately kept at ≈ 1 wt % and three of the five samples contain sub-5 nm particles, the absolute H₂ uptake falls near the detection limit of our chemisorption set-up, making quantitative H₂-TPD or pulse chemisorption measurements unreliable. Instead, we relied on (i) CO-DRIFTS, whose linear-CO intensity scales with surface Pt atoms, and (ii) HRTEM/HAADF-STEM statistics, which provide a direct count of exposed surface sites, to normalise the intrinsic activity (TOF_surface) across the series. While these complementary methods have proved sufficient for the present correlation between particle size and rate, we recognise that selective H₂ chemisorption would offer an independent probe of the active ensemble. We are therefore upgrading our instrumentation to enable high-sensitivity H₂ adsorption microcalorimetry and plan to report those data in a follow-up study.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Please study more about active sites and TOF. This is based on the theory of surface science.

Turnover Frequency refers to the molar number of each catalytic site per unit time that converts reactants into the targeted product on a catalyst surface. It is used to evaluate the activity of catalysts by estimating the number of active sites involved in the reaction.

The active sites of a catalyst refer to the number of sites on the surface, not the amount of catalyst, molar of Pt. So the author's definition of TOF is completely wrong.

Author Response

Response: We thank the reviewer for highlighting the terminology issue, which is indeed a frequent source of confusion in heterogeneous catalysis. Two distinct normalisations coexist in the literature:

Symbol	Normalisation basis	Typical purpose	Representative references
TOF	Number of surface (or otherwise identified) active sites	Fundamental kinetic studies; structure-activity correlations	IUPAC Gold Book definition
TOFoverall	Total moles of metal in the catalyst	Rapid activity benchmarking when precise site densities are unavailable	Cargnello et al. (Science 2013); Macino & Hutchings et al. (Nat. Catal. 2019)

 

Both conventions are accepted, provided the calculation basis is stated unambiguously.

What we did in the manuscript

• We consistently used TOF_overall (moles product·mol_{Pt, total}⁻¹·h⁻¹) because (1) the total Pt loading is accurately known from ICP-OES, whereas the exact number of surface sites would require additional chemisorption or titration experiments that are beyond the scope of this study; (2) it enables direct comparison with earlier size-activity relationships that employed the same normalisation (e.g. refs 28–31 in our MS, Cargnello; Shekhar).
• To avoid any ambiguity we have now revised the text in the Abstract, Experimental Section and Figure legends to.
• In fact, the central conclusion (slope ≈ –1 implying surface-dominated catalysis) would remain unchanged because both normalisations scale identically with particle diameter under the assumptions laid out in Eq. S6.

Changes made to the manuscript

1. Abstract, line 17: “wherein overall turnover frequency (TOF) values (TOF_overall: calculated as moles of reactant converted per mole of Pt per hour)” (new wording).
2. Introduction, line 55: (unless otherwise specified, all TOF values refer to the overall TOF, calculated as moles of reactant converted per mole of Pt per hour). (new wording).
3. Table 1 and Table 3, Fig. 3 and Fig. 4 captions: “TOF values refer to the overall TOF, calculated as moles of reactant converted per mole of Pt per hour.”.

We hope this clarifies our usage of the term and satisfactorily addresses the reviewer’s concern.

Reviewer 2 Report

Comments and Suggestions for Authors The authors have improved the quality of the manuscript taking into account
my suggestions.
Therefore, I agree with the publication of the manuscript in its current form.

Author Response

Thank you for the positive evaluation and recommendation for publication. We appreciate your constructive feedback throughout the review process.

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

First of all, turn-over means that one molecule is transformed at a certain active site, so in terms of the concept of an active site, the original turn-over is limited to the surface. I know that there are other turn-overs, but as long as the author is aware of the difference between the original turn-over and the pseudo-turn-over and distinguishes between them, I think it is acceptable.