Visible light-driven efficient palladium catalyst turnover in oxidative transformations within confined frameworks

Palladium catalyst turnover by reoxidation of a low-valent Pd species dominates the proceeding of an efficient oxidative transformation, but the state-of-the-art catalysis approaches still have great challenges from the perspectives of high efficiency, atom-economy and environmental-friendliness. Herein, we report a new strategy for addressing Pd reoxidation problem by the fabrication of spatially proximate IrIII photocatalyst and PdII catalyst into metal-organic framework (MOF), affording MOFs based Pd/photoredox catalysts UiO-67-Ir-PdX2 (X = OAc, TFA), which are systematically evaluated using three representative Pd-catalyzed oxidation reactions. Owing to the stabilization of single-site Pd and Ir catalysts by MOFs framework as well as the proximity of them favoring fast electron transfer, UiO-67-Ir-PdX2, under visible light, exhibits up to 25 times of Pd catalyst turnover number than the existing catalysis systems. Mechanism investigations theoretically corroborate the capability of MOFs based Pd/photoredox catalysis to regulate the competitive processes of Pd0 aggregation and reoxidation in Pd-catalyzed oxidation reactions.

the claims well and are coherent. The topic of Pd oxidations circumventing harsh conditions and stoichiometric additives is fascinating and is advanced by this work showing improvements over homogeneous conditions. Having said this, the general PS-catalyst electro-communication in UiO-67 approach described in this manuscript as well as the incorporation of molecular Pd catalysts in UiO-67 are not new and have been published several times over the last ~6 years. Comparable studies using the same conceptual approach with UiO-67 (molecular catalyst and photosensitizer) with similar conclusions that this reviewer is aware of: e.g., J. Phys. Chem. C 2018, 122, 3305, DOI: 10.1021J. Am. Chem. Soc. 2016, 138, 8698, DOI: 10.1021/jacs.6b04552); Catal. Lett. 2021, DOI: 10.1007/s10562-021-03719-0, J. Am. Chem. Soc. 2014, 136, 6566, DOI: 10.1021. This questions whether the strategy originality enables publication in Nature Communications. In addition, although interesting, the time-resolved spectroscopic experiments have been conducted in presence of Pd nanoparticles instead of the molecular complexes (in UiO-67-Ir-Pd) which limits the transferability of the conclusions to the actual case studied. If the authors can provide a more convincing argumentation and evidence of a conceptual novelty and a deeper fundamental understanding of mechanisms pointing towards an improved rational design of such systems, the work may well become suitable for Nature Communications.
Independently, this referee suggests considering the following points: 1. Scheme 1 is text-heavy and could be more suited as an "abstract/TOC" image. Particularly panel B is pure text and the Scheme may benefit from enlarging the benefits of this work (panel C) in direct contrast to homogeneous conditions (panel A). 2. Statement on line 46 regarding Pd catalyst turnovers often being limited by reoxidation of the Pd0 is lacking supporting references. 3. Related to the comment above on originality, it is warranted to include a short discussion and overview on comparable UiO-67 systems with dual-anchored molecular complexes for photocatalytic applications towards the end of the introduction (line 86). 4. The claim in lines 147-149 relating to the ideal diameter(s) for photocatalytic applications is interesting and is likely correct. However, several literature works should be cited to back up the individual claims of i) efficient sensitization, ii) exposure of more active sites, and iii) effective O2 adsorption. 5. The EXAFS fit in Figure 2b deviates to the experimental data in several places. A short explanation/clarification would be welcome. 6. The described homogeneous catalytic rationale in lines 212-229 is coherent but lacks citations. 7. In lines 292-294 the post-catalysis crystallinity is discussed and while Figure S14 shows clear retention of reflexes (in-line with the authors' analysis), a short comment/explanation on the distinct change in relative reflex intensities is warranted. 8. Table 2 is very interesting and shows a nice scope of tested substrates. For each reaction respectively, similar TONs/Yields are obtained. Can the main source of deactivation be distilled from this? Selectivity issues? 9. As stated above, the fs-TAS studies are a welcome addition to the mechanistic analysis. However, from the manuscript it only becomes clear at the end why this was performed on a "control" sample batch specifically including Pd nanoparticles instead of the sample UiO-67-Ir-Pd() actually used in catalysis. Including the data for the latter would be interesting to attain a full overview on photophysical behaviour instead of focusing on Ir-Pd0 interactions. 10. The same rationale is frequently repeated in a very similar fashion, first the introduction revolves around these points, afterwards in line 212-229 and 316-336 similar formulations for the same arguments are used. 11. Figure 1 (b,c,d) quality is poor (at least on this referee's computer) which makes its readability limited. 12. This reviewer believes that "superoxygen" l 431 and 433 should be corrected to "superoxide". 13. Sentence 433 should be rephrased since superoxide is unlikely to act as an oxidant. 14. There is likely to be a structure error in fig 3d Dear Editor and Reviewers: Thank you very much for your letter and the reviewer's comments concerning our manuscript entitled "Visible Light-Driven Efficient Palladium Catalyst Turnover in Oxidative Transformations within Confined Frameworks". These comments and suggestions are valuable and helpful for improving our paper. We have made careful corrections according to the reviewer's advices and the changes are marked with a yellow background in the revised manuscript. The responses to the reviewers are as follows: Reviewer #1 (Remarks to the Author): The authors incorporated Ir and Pd complexes into the matrix of UiO-67 MOF to obtain the UiO-67-Ir-PdX2 (X = OAc, TFA). The UiO-67-Ir-PdX2 (X = OAc, TFA) shows impressive catalytic properties towards a couple of oxidation reactions. The catalytic mechanism is discussed in detail. However, I have several serious concerns in the structural characterization of UiO-67-Ir-PdX2 (X = OAc, TFA) before and after characterization. I do not recommend its publication until the authors fully address the following comments.

Response:
We thank this reviewer for the positive comments.
1. The authors characterized the structure of obtained complexes by XRD, EDX maps and XAFS. However, from the presented analyses, none of them could be evident for the structure showing in Figure 1a. Please note that EDX maps only showed the elemental distribution rather than the crystal structure. Strong evidence on the crystal structure must be required. A combination of the fitting process of the XRD patterns, elemental analyses, IR and Raman spectra must be helpful. 2. The authors also showed the UiO-67-Pd in Table 1, what is the structure?
Response: UiO-67-Pd was synthesized according to previous reports (Green Chem. 4. For the XAFS results, a. Please make sure that the Ir spectra are tested in the K-edge or L3-edge. Ir-K has quite high energy.

Response:
The XAFS of Ir was tested in L3-edge, not K-edge. We are regretful for the typing error in the original manuscript and we appreciate the careful examination of this reviewer.
b. Please also provide the original XAFS data in E-and k-space in addition to the R-space.
Response: According to the reviewer's suggestion, the XAFS data in E-and k-space have been included in the revised manuscript (  Fig. 16). Relative description has been added in the revised manuscript. 5. In the catalytic reaction, the catalytic performance of the physical mixture of these three complexes is strongly recommended to be tested as a control group.
Response: According to the reviewer's suggestion, the catalytic performance of the physical mixture of three complexes have been tested and the corresponding catalytic results have been added in Table 1 (entry 12), Supplementary Table 6 (entry 12) and Table 7 (entry 12).
6. In Figure S14, the XRD pattern is required to a more extended range to show the feature of Pd NPs.

Response:
The PXRD patterns in Figure S14  UiO-67-Ir-PdX2 (X = OAc, TFA) after catalysis. This result proved again that no obvious Pd NPs were formed under the optimal reaction conditions. 7. In Figure S17, how the authors confirmed it is Pd NPs rather than Ir NPs or others?
Other characterizations are strongly required to confirm this point. Also, the morphology of UiO-67-Ir-Pd(TFA)2 showing in Figure S17a is different from the original morphology as well as the morphology showing in Figure S17b. This suggests some structure changes. Please identify and explain these differences.

Response:
We firstly appreciate the reviewer's suggestion for the confirmation of Pd NPs. The XPS measurements of the recovered catalysts operated in dark were conducted ( Supplementary Figs. 23a and 23b). The result showed that no characteristic peaks of Ir 0 was found and the oxidation state of Ir is still +3. While in Pd region, two peaks at 335.1 and at 340.5 eV, attributing to Pd 0 3d5/2 and Pd 0 3d3/2, respectively, were observed. These results undoubtedly show that the formed nanoparticles should be Pd NPs. Relative description and discussion have been added in the revised manuscript, as follows: When operated in dark, distinct Pd NPs with a size of approximate 5 nm were observed over the framework of the recovered MOF catalyst, which is termed as UiO-67-Ir-PdNPs ( Supplementary Fig. 23c). The Pd NPs nature was confirmed by XPS spectra that the appearance of peaks for Pd 3d5/2 at 335.1 and Pd 3d3/2 at 340.5 eV ( Supplementary Fig. 23a), respectively, prove the zero valent Pd, while the 4f7/2 and 4f5/2 Ir peaks do not almost change ( Supplementary Fig. 23b), indicating that Ir is still + 3 oxidation state.
Secondly, as for the morphological differences of the recovered catalyst and the fresh catalyst UiO-67-Ir-Pd(TFA)2, we think it is partly due to the slight structure distortion on the catalyst surface during the catalytic reaction. After all, this TEM picture ( Figure S17a which has become to Supplementary Fig. 22a in the revised SI) was taken after the catalyst has been used for five consecutive runs. Nevertheless, the overall morphology of the recovered catalyst still exists as octahedral particles, and their unchanged PXRD patterns ( Supplementary Fig. 19) suggested they are still long range ordered without obvious structure destruction. This slight morphology differences before and after catalysis in UiO series MOFs is not uncommon and can Response: We firstly appreciate the positive comments from this referee.
As the reviewer said, the merger of transition metal catalysts and photosensitizers into MOFs have been reported in recent years. Different groups have made important progress in this field and found successful applications of these MOF catalysts in organic transformations, water splitting, CO2 reduction, and etc. We have also stated such research process and cited related literatures in the original manuscript (line 76-80). However, the purpose of this manuscript is to use the designability of MOFs framework to solve the problems existing in Pd-catalyzed oxidation reactions, and use the easy characterization of crystalline MOFs framework to deeply explore the catalytic reaction mechanism, so as to provide a theoretical basis for the design of efficient and green Pd oxidation catalysts.
Lastly, we appreciate very much this referee' patience for our work again.
Independently, this referee suggests considering the following points: 1. Scheme 1 is text-heavy and could be more suited as an "abstract/TOC" image.
Particularly panel B is pure text and the Scheme may benefit from enlarging the benefits of this work (panel C) in direct contrast to homogeneous conditions (panel A).  Figure 2b deviates to the experimental data in several places. A short explanation/clarification would be welcome.

Response:
It is noteworthy that the slight deviation of the EXAFS fit of Ir to the experimental data is not uncommon in similar report (J. Am. Chem. Soc. 2020, 142, 8602;etc.). In addition, to further guarantee this EXAFS result, we added the EXAFS result of homogeneous pure Ir complex [Ir(bpy)(ppy)2](PF6) in the revised supplementary information (Supplementary Fig. 16a), and it also showed similar result. Meanwhile, the molecular model of the Ir III complex within UiO-67-Ir-PdX2 (X = OAc, TFA) has been added in the revised SI ( Supplementary Fig. 16c) to better interpret each path in Fig. 2b (in the original manuscript). 7. In lines 292-294 the post-catalysis crystallinity is discussed and while Figure S14 shows clear retention of reflexes (in-line with the authors' analysis), a short comment/explanation on the distinct change in relative reflex intensities is warranted.
Response: According to the reviewer's suggestion, a short explanation on the change in relative reflex intensities was added in the revised manuscript as follows.
It is worth noting that UiO-67-Ir-Pd(TFA)2 was stable under the photocatalytic reaction conditions as illustrated by the retention of PXRD pattern for the recovered UiO-67-Ir-Pd(TFA)2 ( Supplementary Fig. 19)  Indeed, for each reaction respectively, some substrates show similar TONs/Yields.
We would like to attribute this phenomenon more to inherent mechanism aspect of the reaction itself. Taking the decarboxylative coupling of allylic alcohols for example, it was evident that disubstituted benzoic acid facilitates carbon dioxide extrusion process in decarboxylative coupling (Chem. Eur.J., 2014, 20, 16680). In addition, the incipient anion which is generated after decarboxylation needs to be efficiently stabilized for further cross-couplings (Org. Lett., 2014, 16, 3934). These features make the dimethoxy substituted benzoic acid as the most commonly utilized substrates in Pd-catalyzed decarboxylative coupling (Org. Lett., 2009, 11, 2341Tetrahedron 2017Tetrahedron , 73, 2242J. Org. Chem., 2016, 81, 2521. In this work, the generation of the β-aryl ketones needs selective β-H elimination in the -OH side. The electron-donating dimethoxy benzoic acids could thus facilitate the selective β-H elimination and gave the β-aryl ketones product. Therefore, the dimethoxy benzoic acids are the dominating substrates to determine the yield, and different allylic alcohols with similar dimethoxy benzoic acids show similar TONs/Yields. 9. As stated above, the fs-TAS studies are a welcome addition to the mechanistic analysis. However, from the manuscript it only becomes clear at the end why this was performed on a "control" sample batch specifically including Pd nanoparticles instead of the sample UiO-67-Ir-Pd actually used in catalysis. Including the data for the latter would be interesting to attain a full overview on photophysical behaviour instead of focusing on Ir-Pd 0 interactions.

Response:
We totally understand the consideration of the reviewer. However, as we stated above, these three Pd-catalyzed oxidation reactions are initiated by the interaction between Pd II and substrate, accompanied by the generation of Pd 0 species.
No matter what interaction between Pd II and Ir III in UiO-67-Ir-Pd may happen, the Pd II are the only species to initiate the reaction. The excited [Ir III ]* moiety within the framework acts as an oxidant to reoxidize the in situ formed Pd 0 species to Pd II , that is, the electron injection from Pd 0 to [Ir III ]*, which can be monitored by the time-resolved fs-TA spectra and decay constants. While the fs-TA spectroscopic experiment of UiO-67-Ir-PdX2 cannot reflect the real electron transfer channel and Pd 0 reoxidation pathway. Thus, the fs-TA experiments of the complex containing Pd 0 and Ir III species seems reasonable to investigate the reoxidation process of Pd 0 species.
Anyway, we appreciate the reviewer for the kind suggestions.
10. The same rationale is frequently repeated in a very similar fashion, first the introduction revolves around these points, afterwards in line 212-229 and 316-336 similar formulations for the same arguments are used.
Response: According to the reviewer's advice, the relative statements have been adjusted to be more concisely in the revised manuscript.
11. Figure 1 (b,c,d) quality is poor (at least on this referee's computer) which makes its readability limited.

Response:
We readjusted the picture quality in Fig. 1 (b, c, d) ( Fig. 2 in the revised manuscript) to make it clear to read.

Response:
The term "superoxygen" has been corrected to "superoxide" in the revised manuscript.
13. Sentence 433 should be rephrased since superoxide is unlikely to act as an oxidant.
Response: According to the reviewer's advice, this sentence has been deleted, since