Benzene- and pyridine-incorporated octaphyrins with different coordination modes toward two PdII centers

Expanded porphyrins have received considerable attention due to their unique optical, electrochemical and coordination properties. Here, we report benzene- and pyridine-incorporated octaphyrins(1.1.0.0.1.1.0.0), which are synthesized through Suzuki-Miyaura coupling of α,α′-diboryltripyrrane with m-dibromobenzene and 2,6-dibromopyridine, respectively, and subsequent oxidation with 2,3-dicyano-5,6-dichlorobenzoquinone. Both octaphyrins are nonaromatic and take on dumbbell structures. Upon treatment with Pd(OOCCH3)2, the benzene-incorporated one gives a Ci symmetric NNNC coordinated bis-PdII complex but the pyridine incorporated one gives Ci and Cs symmetric NNNC coordinated bis-PdII complexes along with an NNNN coordinated bis-PdII complex bearing a transannular C–C bond between the pyrrole α-positions. In addition, these two pyridine-containing NNNC PdII complexes undergo trifluoroacetic acid-induced clean interconversion.

The manuscript "Benzene-and Pyridine-Incorporated Octaphyrins (1.1.0.0.1.1.0.0) With Remarkably Different Coordination Modes Toward Two Pd(II) Metals" submitted by Jianxin Song and co-workers is a significant addition to expanded carbaporphyrinoid chemistry. Namely authors have elaborated the efficient methodology to construct octaphyrins (1.1.0.0.1.1.0.0) embedding two benzene (<b>12</b>) or two pyridine (<b>13</b>) moieties. These molecules acquire dumbbell geometry(D2h symmetry) with inverted six membered rings in the "trans" positions. Authors claim that the novel octaphyrins reveal the nonaromatic properties. Importantly the suitably prearranged geometries of these octaphyrins, which created two (NNNC) pockets, facilitated incorporation of two palladium(II) cations. Thus the insertion into benzene octaphyrins(1.1.0.0.1.1.0.0) <b>12</b>yielded the C2v symmetric molecule containing two identical Pd(II)-(NNNC) subunits (<b>14</b>). The insertion outcome was more complex for pyridine octaphyrins (1.1.0.0.1.1.0.0) (13). Here the analogous C2v Pd(II)-(NNNC) species was identified (<b>15</b>) to be accompanied, however, by the Cs Pd(II)-(NNNC) complex (<b>16</b>). Eventually the intriguing reversible <b>16</b> -<b>16</b> conversion in the presence of TFA was detected. The remarkable rearrangement of a pyridine octaphyrins (1.1.0.0.1.1.0.0) <b>13</b> skeleton was discovered as well. This process afforded the bis-Pd(II)-(NNNN) complex <b>17</b> bearing a transannular C-C bond between the pyrrole αpositions. The usual techniques (including MS, NMR, UV-Vis electronic spectroscopy and electrochemical methods) have been applied to characterize the novel macrocyclic compounds. The appropriate crystal structure have been also included. The DFT studies have been carried as well. Altogether the presented results are very significant. I highly appreciate the importance of this work, which provides a remarkable route to explore novel nontrivial coordination modes in expanded porphyrins and carbaporphyrinoids. Consequently, I can eventually recommend publication of this contribution in Nature Communication under condition that the remarks outlined below will be properly addressed. <b>Mechanism</b> The mechanism of the peculiar reversible <b>15</b> ⇄ <b>16</b> conversion is expected to be addressed in detail. Evidently the process requires the Pd-Cβ ⇄ Pd-(Cβ-H) transformation. Accordingly, the specific intramolecular activation of the Cβ-H bond, presumably facilitated by pyridine N-protonation can be of importance here. Interestingly, one can readily notice, that using TFA-d will result in selective β-deuteration of pyridine moieties, nicely providing the additional insight into the process. In addition, the nature of the intermediate containing presumably the Pd-(NCCC) and Pd-((CH)NNN) units can be considered using the DFT approach. Of course, one can wonder if a postulate above intermediate can be trapped by 1H NMR in the presence of TFA in very large excess.
<b>Peculiar chemical shift -issue</b> Consistently in the manuscript authors describe the reported molecules as nonaromatic.
Considering the nature of α,α'-incorporated benzene or pyridine units such an electronic structure seems to be rationally expected. Still, the analysis of the 1H NMR spectra at ESI, substantiated by examination of detailed chemical shifts, raises some essential scientific issue, which requires the sound addressing. In particular, the "enormous" upfield relocation of the inner-H resonances and remarkable difference between "inner" and "outer" chemical shifts have need of reasonable explanations (aromaticity?, the specific arrangement of six-membered units?). Below only the selected examples, which attracted my attention, are given (the assignments "outer" vs "inner" reflects the location with respect to the macrocyclic ring and has been given by this reviewer). a) Figure S4 12 m-phenylene (outer) 8.34 vs m-phenylene (inner) 5.63 b) Figure S6 13 m-pyridine (inner) 6.05 c) Figure S8 C6H3-H (inner) 5.32, C6H3-H (inner) 3.95, C6H3-H (outer) 8.67 d) Figure S10, Pyridine-H (inner)), 5.46, Pyridine-H (inner), 4.07 . e) Figure S12 pyridine-H (inner) 3.69 (coordinated) pyridine-H (inner, outer???) 7.67 , 7.03. I believe that the references 15, 18 (this contribution) and J. Am. Chem. Soc. 2014, 136, 4281-4286 will be useful in the further analysis. Evidently, the detailed assignments 1H resonances are absolutely necessary to provide the satisfactory explanation. <b>Other points</b> The acceptable HRMS data have been not been given in the description of new compounds.
Reviewer #2 (Remarks to the Author): This manuscript presents the synthesis of expanded porphyrins with either 1,3-phenylene or 2,6pyridinylene linkers, together with an investigation of the palladium complexes of these macrocycles. This study has uncovered several fascinating and unexpected features, such as the formation of the remarkable twisted bis-palladium complex 17. It is interesting that the 1,3phenylene-linked macrocycle 12 forms only one Pd2 complex, 14, whereas the 2,6-pyridinylenelinked macrocycle 13 forms three complexes 15, 16 and 17. All of these compounds are thoroughly characterized by NMR, mass spectrometry, electrochemistry, UV-visible spectroscopy and singlecrystal X-ray analysis. The work has been carried out to a high standard and the manuscript is well written. It is suitable for publication in Nature Communications after a few minor revisions.
(1) Chart 1: The structure of compound 7 appears to be incorrect; it is lacking a nitrogen atom.
(2) Page 3, line 68: Why does the equimolecular reaction of 9 and 10 yield only a trace amount of 12 whereas 1.2 equivalents 9 produced 12 in 8.0% yield? Is this a reproducible result? Can the authors offer a plausible rationalization? (3) Scheme 1. Replace "X = C" with "X = CH" (3 times).
(4) Page 11, lines 224-231: Were redox potentials measured on a CHI900 scanning electrochemical microscope or on an ALS660 electrochemical analyzer? (5) lines 240,260,277,288,327 and 338. The quantity of DDQ should be specified. It is not adequate just to write "excess"; e.g. does this mean 1.1 equiv., 2 equiv. or 20 equiv.?
Reviewer #3 (Remarks to the Author): The paper "Benzene-and Pyridine-Incorporated Octaphyrins(1.1.0.0.1.1.0.0) With Remarkably Different Coordination Modes Toward Two PdII Metals" by Liu et al. describes the synthesis of octaphyrins via Suzuki-Miyaura coupling. All compounds have been thoroughly characterized by NMR, SC-XRD, UV, CV and MALDI-TOF. Most SC-XRD data show however a rather bad agreement with the model. The disorders were just treated on a first level and these nice results demand and deserve a crystallographers love, e.g. for 13.cif the biggest residual densities belong to a 2nd site of the pyrole group next to the pyridine ring and for 14.cif the pheylene unit is bent due to a not treated disorder of the phenylene group to mention just the most obvious mistakes. As there a two position for the palladium metal there have to be two positions for the phenylene unit. Instead of properly treating the disorder FLAT and ISOR was used to mask the problem. This strategy was also used in the figures of the manuscript where thermal ellipsoids are shown with 30% probability. Even though this study is overall well done, I do not recommend the publication of the manuscript in its current state.

Reviewer #4 (Remarks to the Author):
The manuscript brings interesting results on the title compounds. In addition to their syntheses the authors used various methods of their characterization. I am not an expert in organic syntheses and NMR experiments but this study is worth of publishing. My comments may be summarized as follows: i) It must be put more exactly (line 95) that "Complex 14 also shows a nearly C2h symmetric structure …" due to its deviations from planarity.
ii) The Table 1 description is a little bit chaotic. I propose to move a substantial part of lines 163-165 into the line 162.
iii) The DFT method use is mentioned in the line 200 and Figs. S50 -S51 but authors did not mention the software used. The references on the method and basis sets used are missing as well.
Have the authors performed a geometry optimization of the structures under study (isolated or in solution?) or used X-ray geometries only (in such case -were the C-H and N-H bond lengths corrected and how the problem of disordered atoms has been solved)? iv) Frontier MOs might be briefly mentioned in the main text as well (Figs. S40 -S44). It is a pity that the authors have not presented some results of population analysis for Pd atoms and their coordination polyhedra (at least in Supplementary). The manuscript would be stronger. v) In Figs. S50-S51 the DFT energies of individual systems are compared. If the authors use optimized geometries, the Gibbs free energy data are more suitable for this purpose (temperature dependence). vi) Discussion is too brief and descriptive only. Much deeper insight is desirable. vii) In Method (lines 222-223) the X-ray experiment is described but the software used for the structure solution and related details are not mentioned at all. viii) The asterisks in Figs. S2-S4, S6, S8, S10, S12 and S14 are not explained (impurities?). ix) Some misprints and/or grammatical errors can be found by more careful reading. Finally it may be concluded that this manuscript demands major correction to be published. The appropriate crystal structures have been also included. The DFT studies have been carried as well. Altogether the presented results are very significant. I highly appreciate the importance of this work, which provides a remarkable route to explore novel nontrivial coordination modes in expanded porphyrins and carbaporphyrinoids. Consequently, I can eventually recommend publication of this contribution in Nature Communication under condition that the remarks outlined below will be properly addressed.

Response to Comments of Reviewers
(1) Reviewer 1 wrote: Mechanism. The mechanism of the peculiar reversible 15 ⇄ 16 conversion is expected to be addressed in detail. Evidently the process requires the Pd-C β ⇄ Pd-(C β -H) transformation. Accordingly, the specific intramolecular activation of the C β -H bond, presumably facilitated by pyridine N-protonation can be of importance here. Interestingly, one can readily notice, that using TFA-d will result in selective β-deuteration of pyridine moieties, nicely providing the additional insight into the process. Unfortunately, we could not observe the intermediate via in-situ 1 H NMR but confirmed that 16 was a sole product. We appreciated the nice advice from the reviewer 1. Figure S55 and Figure S56 show the relative energies of these intermediates indicated by DFT calculation.
(2) Reviewer 1 wrote: Peculiar chemical shift -issue. Consistently in the manuscript authors describe the reported molecules as nonaromatic. Considering the nature of α,α'-incorporated benzene or pyridine units such an electronic structure seems to be rationally expected. Still, the analysis of the 1 H NMR spectra at ESI, substantiated by examination of detailed chemical shifts, raises some essential scientific issue, which requires the sound addressing. In particular, the "enormous" upfield relocation of the inner-H resonances and remarkable difference between "inner" and "outer" chemical shifts have need of reasonable explanations (aromaticity?, the specific arrangement of six-membered units?). Below only the selected examples, which attracted my attention, are given (the assignments "outer" vs "inner" reflects the location with respect to the macrocyclic ring and has been given by this reviewer). a) Figure S4  For compound 16, the case is different from those in 12-15 due to their different symmetries. Apparently in 16 the chemical environment of C 5 NH unit differs from that of C 5 NH 3 unit. The protons of the C 5 NH unit lies above the center of the C 5 NH 3 unit, which means this proton locates in the shielding area. However, the 3 protons of the C 5 NH 3 unit are not so close to the center of the C 6 H unit, which indicates these 3 protons are less shielded.
As a result, the proton of the C 5 NH shows a chemical shift of 3.69 ppm, while the protons of C 5 NH 3 show chemical shifts of 7.67 and 7.03 ppm.
As suggested by reviewer 1 and to confirm our explanation above, DFT calculation was applied to elucidate these chemical shifts. These results were summarized in Table S8-S13, which indicated that the theoretical calculation met well with the experimental data. Sharp contrast between the centers of these molecules and the centers of hemiporphyrins in NICS value support our previous judgment that these octaphyrins are not global aromatic and that 'abnormal' chemical shifts are resulted from local aromaticity.
(3) Reviewer 1 wrote: The acceptable HRMS data have been not been given in the description of new compounds.

Response: We repeated all of the HRMS experiments, acceptable HRMS data were included
in the main text.

Response to Comments of Reviewer 2
Reviewer 2's general comments: This manuscript presents the synthesis of expanded porphyrins with either 1,3-phenylene or 2,6-pyridinylene linkers, together with an investigation of the palladium complexes of these macrocycles. This study has uncovered several fascinating and unexpected features, such as the formation of the remarkable twisted bis-palladium complex 17. It is interesting that the 1,3-phenylene-linked macrocycle 12 forms only one Pd2 complex, 14, whereas the 2,6-pyridinylene-linked macrocycle 13 forms three complexes 15, 16 and 17. All of these compounds are thoroughly characterized by NMR, mass spectrometry, electrochemistry, UV-visible spectroscopy and single-crystal X-ray analysis. The work has been carried out to a high standard and the manuscript is well written. It is suitable for publication in Nature Communications after a few minor revisions.
(1) Reviewer 2 wrote: Chart 1: The structure of compound 7 appears to be incorrect; it is lacking a nitrogen atom.

Response: Corrected.
(2) Reviewer 2 wrote: Page 3, line 68: Why does the equimolecular reaction of 9 and 10 yield only a trace amount of 12 whereas 1.2 equivalents 9 produced 12 in 8.0% yield? Is this a reproducible result? Can the authors offer a plausible rationalization?
Response: We repeated this experiment for several times. The yield of the equimolecular reaction of 9 and 10 was confirmed to be 3.8%. The relatively low yield of the equimolecular reaction may be due to some impurities in 9 or chemical reactivity of 9 under this cross coupling condition. Ref. (Angew. Chem. Int. Ed. 2016, 55, 6438; 2019, 58, 8124) (3) Reviewer 2 wrote: Scheme 1. Replace "X = C" with "X = CH" (3 times).
Response: We used ALS660 electrochemical analyzer to measure redox potentials, this was clarified in Materials and characterization part.
(5) Reviewer 2 wrote: lines 240,260,277,288,327 and 338. The quantity of DDQ should be specified. It is not adequate just to write "excess"; e.g. does this mean 1.1 equiv., 2 equiv. or 20 equiv.?
Response: A typical quantity of DDQ we applied is 2.4 equiv. The quantity was noted in the manuscript.
. (1) Reviewer 3 wrote: Most SC-XRD data show however a rather bad agreement with the model. The disorders were just treated on a first level and these nice results demand and deserve a crystallographers love, e.g. for 13.cif the biggest residual densities belong to a 2nd site of the pyrole group next to the pyridine ring and for 14.cif the pheylene unit is bent due to a not treated disorder of the phenylene group to mention just the most obvious mistakes. As there a two position for the palladium metal there have to be two positions for the phenylene unit. Instead of properly treating the disorder FLAT and ISOR was used to mask the problem. This strategy was also used in the figures of the manuscript where thermal ellipsoids are shown with 30% probability. Even though this study is overall well done, I do not recommend the publication of the manuscript in its current state.

Response to Comments of
Response: We thank the suggestion of Reviewer 3. As Reviewer 3 referred serious disorders were observed in 12, 13, 14 and 15. We split all of the atoms in 12 but did not treat 13, 14 and 15 in a similar way initially. Inspired by Reviewer 3 we re-refined crystallographic data of 13, 14 and 15 by trying to find the 2 nd parts in these compounds. For 13, the R1 value decreased significantly when the 2 nd positions were found and treated as the 2 nd part despite large amount of constrains and restrains involving AFIX, FLAT, SADI, ISOR and DELU were used. When this strategy was applied for 15, the 2 nd part could be isolate as well but the R1 value did not decrease. In 14 we could find only one 2 nd part of a half molecule in an asymmetric unit (one asymmetric unit contains 2 halves of molecule), however, the structure is not stable any more during refining when the 2 nd part of the other half molecule was isolated. This is mainly due to the low occupancy (~0.15) of the 2 nd part.

Response to Comments of Reviewer 4
Reviewer 4's general comments: The manuscript brings interesting results on the title compounds.
In addition to their syntheses the authors used various methods of their characterization. I am not an expert in organic syntheses and NMR experiments but this study is worth of publishing. My comments may be summarized as follows: Finally it may be concluded that this manuscript demands major correction to be published.
(1) Reviewer 4 wrote: It must be put more exactly (line 95) that "Complex 14 also shows a nearly C2h symmetric structure …" due to its deviations from planarity.
Response: We rephrased this sentence as "Complex 14 shows a nearly C i symmetric structure …" (2) Reviewer 4 wrote: The Table 1  Response: Gibbs free energy data were added in Figure S55 and Figure S56 and the corresponding data was discussed in main text.
(6) Reviewer 4 wrote: Discussion is too brief and descriptive only. Much deeper insight is desirable.
Response: We some analysis in 1 H NMR spectra was added in SI (Table S8 to S13).
(7) Reviewer 4 wrote: In Method (lines 222-223) the X-ray experiment is described but the software used for the structure solution and related details are not mentioned at all.
Response: Using Olex2, structures of compound 12-17 were solved with the ShelXS structure solution program using Direct Methods and refined with the ShelXL refinement package using Least Squares minimisation. Disordered solvent molecules were treated by SQUEEZE program of Platon. This description was added in Method part.

Response: The asterisks in
Response: Some misprints and/or grammatical errors were corrected, they are highlighted in the revised version.

REVIEWERS' COMMENTS
Reviewer #1 (Remarks to the Author): Authors have satisfactory addressed all previously raised problems, Consequently I can recommend this contribution for publication.
Reviewer #2 (Remarks to the Author): The revised version of this manuscript has been significantly improved and all the points raised by the referees have been addressed. This is a fascinating piece of work and it has been carried out to a high standard.
Two minor points: (1) In their response to Reviewer 1's first point, the authors have carried out an interesting study of the site of deuteration on treatment of compound 15 with TFA-d to form deuterated 16. These mechanistic results should be mentioned in the text and added to the SI.
(2) It is not very clear what is mean by "relative energy" in Figures S55 and S56 and why this is different from "delta G". Should "relative energy" be changed to "delta H"?
Reviewer #4 (Remarks to the Author): The manuscript brings interesting information on the title compounds but its presentation must be improved. My comments are related to its Supplementary information only and can be summarized as follows: i) The title and authors are missing. ii) How did you check the stability of the optimized structures (l. 408)? iii) Missing reference of 6-311G(D,p) basis sets (l. 411). iv) SDD denotes both pseudopotentials and basis sets (l. 412). v) Some details on NICS treatment are desirable at l. 412 (probe location -NICS(0) or NICS(1)?).
Reviewer #5 (Remarks to the Author): I have been requested to re-examine the updated crystallography. In general the authors have now appropriately dealt with the previously raised issues (disorder models) and the structures are now acceptable. The authors have also done a good job of dealing with the comments of the other referees and the manuscript is not largely acceptable.
In the future the authors should consider using the newer versions of shelxl for their refinements