Anisotropic flexibility and rigidification in a TPE-based Zr-MOFs with scu topology

Tetraphenylethylene (TPE)-based ligands are appealing for constructing metal-organic frameworks (MOFs) with new functions and responsiveness. Here, we report a non-interpenetrated TPE-based scu Zr-MOF with anisotropic flexibility, that is, Zr-TCPE (H4TCPE = 1,1,2,2-tetra(4-carboxylphenyl)ethylene), remaining two anisotropic pockets. The framework flexibility is further anisotropically rigidified by installing linkers individually at specific pockets. By individually installing dicarboxylic acid L1 or L2 at pocket A or B, the framework flexibility along the b-axis or c-axis is rigidified, and the intermolecular or intramolecular motions of organic ligands are restricted, respectively. Synergistically, with dual linker installation, the flexibility is completely rigidified with the restriction of ligand motion, resulting in MOFs with enhanced stability and improved separation ability. Furthermore, in situ observation of the flipping of the phenyl ring and its rigidification process is made by 2H solid-state NMR. The anisotropic rigidification of flexibility in scu Zr-MOFs guides the directional control of ligand motion for designing stimuli-responsive emitting or efficient separation materials.

materials with L1, L2 or L1&L2 ligands show district behavior in terms of structural rigidification and optical properties. The results suggest an anisotropic flexible behavior upon installation of L1 and L2 ligands. In addition to the photoluminescence study, that author used these MOFs as stationary phases in gas chromatography for the separation of different isomers from various groups including hexane, octane, nonane, heptene, octene, and ethylbenzene.
Overall, the approach to tune anisotropically the flexible behavior of this particular MOF and in turn its optical and separation properties, is an interesting work. However, the novelty is limited because significant and similar work in terms of PL properties has been already published for this particular MOF. Furthermore, the manuscript is not well written. It is very hard to read, containing speculations, and difficult to follow the discussion of the experimental results. This work is not suitable for publication in Nature Communications. I recommend and more specialized Journal after major revisions. Specific comments are provided below.
1. How do the authors know that the obtained a non-interpenetrated scu MOF instead of the known double interpenetrated version? 2. Changes in PL properties are observed after heating the material at 250 oC. It is necessary to use heat in order to observe these changes? Is it possible by post synthetic modification, for example, by removing the coordinated terminal acetate anions to trigger these changes?
3. After heating, the PL properties are restored after treatment with DMF in the presence of acetic acid at 120 oC. Is the use of temperature and acetic acid crucial to restore the properties? Additional work is needed to clarify this behavior. 4. The pxrd patterns shown in SI Fig. 23 are not discussed in detail, along with the corresponding PL properties. For example, while for individual L1 and L2 containing MOFs the pxrd between the asmade and heated material is almost unchanged, this is not the case of the solid containing both L1 and L2 where the heated solid (Zr-TCPE-DLI-H) shows a larger unit cell. This is quite interesting because the installation of both L1 and L2 is expected to rigidify the scu framework. 5. Regarding, the isomer separation properties that results are confusing and not explained. For example, it is state "The Zr-TCPE-H showed seriously low separation ability compared with the original Zr-TCPE…". This behavior is attributed to "structural deformation" but no explanation is provided, supported by experimental data. Also the statement "Zr-TCPE-L1-H showed fair separation ability with obvious peak tailing compared with Zr-TCPE-L1 (Supplementary Fig. 40 and 41), which is due to the existence of flexibility along the c-axis." is confusing. 6. The statement "It was worth noting that there was almost no separation efficiency loss in Zr-TCPE-DLI-H ( Supplementary Fig. 43), indicating that not only its structural flexibility was completely rigidified" is in contrast to the pxrd data for this material (SI Fig. 23) where the Zr-TCPE-DLI-H phase shows a shift in the low angle peak towards lower 2theta values, suggesting an increase in the unit cell size and the presence of flexibility.
Reviewer #3: Remarks to the Author: This work describes the rigidification of an scu topology ZrMOF via linker insertion. The scu topology has two distinct pocket and here two different linkers were installed on each pocket stepwise and the resulting materials were compared. The resulting MOFs were characterized extensively and deutrated analogue of final some MOFs were prepared to establish structural rigidity via solid state NMR. While the topology, the MOF itself and linker insertion for rigidification is not new, it is worth publishing the demonstration of linker rotation and also the separation performance of the MOFs. However the following major points need to be addressed before its publication.
1-Firstly, while the authors cited the previously published TCPE MOFs with biphenyl analogs, the same ligand was also used for building scu topology Zr MOF. Authors need to cite the original work ( Zirconium-based metal-organic framework gels for selective luminescence sensing RSC Adv., 2020,10, 44912-44919) and mention this in both article and SI where the synthesis of this material is mentioned.
2-While authors cited other ligand insertion studies, they missed an improtant one for this research which shows that insertion of the same naphlene based ligand used here into an scu Zr MOF which increased its mechanical properties, more specifically bulk modulus (Robison et al. Chem. Mater. 2020, 32, 8, 3545-3552) 3-Were the Authors able to occupy all the possible sites to install these linkers? This should be mentioned in the discussion. 4-Authors reported the BET surface area of the materials studied here with decimal points. As it was mentioned in a recent Tutorial article (DOI: https://doi.org/10.1039/D1TA08021K), the surface areas obtained from the BET equation is subject to error depending on the pressure region selected so it is not accuarate to report surface areas with this precision as it is not that reproducible. So I recommend authors to either remove the decimal points or report the number with error bars.
5-Authors mentioned that they were not able to get the single crystal despite many trials. Their experimental section shows that they add water as well in addition to acetic acid. Water typically results in smaller particles. It could be because of the fact that they are getting other phases such as csq, in that case I recommend not using water but using a little more DMF. Not sure if this would result in diffraction quality crystals but I am nearly confident that this would give that larger crystals. 6-Authors also coated the inner surface of a column with the MOF to use it for hydrocarbon separation. However, their experimental section is not clear about exactly how the procedure was one. They mentioned it was was or x mL suspension was passed through column but they did not mention how this was done. This is an important step given that the capillary diameter is so small. I recommend authors to add more details to their experimental section. 7-Similar to above, for linker insertion, authors have cited a previous work without mentioning the amount of MOF used for each reaction. The reagnets amount needs to be added to ensure others can safely reproduce your results.
2 flipping of organic ligands exists, the organic ligands are firmly and identically immobilized into the framework along a, b, and c directions, and the framework is rigid." Please see Page 1, left column, second paragraph in the manuscript.
We have revised the incorrect illustration in Scheme. Please see Page 2, Scheme 1 in the manuscript.   The PXRD patterns, TEM images, and 1 H   NMR spectra confirmed the successful synthesis of Zr-TCPE, Zr-TCPE-L1, and Zr-TCPE-L2 constructed with H4TCPE-

7.
There are some small mistakes in the manuscript should be revised before publication.

Please double check.
Response: Thank you for pointing this out. We have carefully checked the manuscript.
We have revised some confusing expressions and revised mistakes. For example, we have revised "Meanwhile, with the dual linker installation (DLI), the Zr-TCPE-DLI was also obtained through the sequential synthetic method first with L2 then L1" to "Meanwhile, 10 with the dual linker installation (DLI), the Zr-TCPE-DLI was also obtained, in which the installation order is installing L2 first and then L1" in the manuscript. Please see Page 5, right column, the first paragraph in the manuscript. Overall, the approach to tune anisotropically the flexible behavior of this particular MOF and in turn its optical and separation properties, is an interesting work. However, the novelty is limited because significant and similar work in terms of PL properties has been already published for this particular MOF.
Response: Thank you for your comments. We have carefully read these comments and understand that the main consideration is the novelty. We apologize for not fully emphasizing our novelty which is different from theirs', leading to the misunderstanding.
Indeed, we have carefully read these two papers and fully evaluated the difference between these materials and our Zr-TCPE before the submission. We still believe our work investigates the anisotropic flexible TPE-based scu MOF from a special perspective with quite different phenomena. The differences between our work and others are listed below. Thus, we have confidently and objectively cited this paper in our previous submission. We understand your consideration, but we still would like to show the difference in major discoveries between our work and ref. #35.
The mechanism and modulation methods for PL changes are different. We agree with 13 the comment that PL changes in TPE-based MOFs are common phenomena. Thus, the special chemical modulation method and unique mechanism were crucial to the novelty. It We have revised "Meanwhile, with the dual linker installation (DLI), the Zr-TCPE-DLI was also obtained through the sequential synthetic method first with L2 then L1" to "Meanwhile, with the dual linker installation (DLI), the Zr-TCPE-DLI was also obtained, in which the installation order is installing L2 first and then L1" in the manuscript. Please see Page 5, right column, the first paragraph in the manuscript.
We have revised "Thus, we supposed that pocket A was fully occupied by L1 in Zr-TCPE-L1 as the possibility of L1 at pocket B was ruled out due to the size mismatching" to "Because of the size mismatching between the space of pocket B and the length of L1, the possibility of L1 coordinated on pocket B was ruled out. Thus, pocket A was fully occupied by L1 in Zr-TCPE-L1" in the manuscript. Please see Page 5, left column, second paragraph 18 in the manuscript.
We have revised "We supposed that the steric hindrance of pocket B in the original MOF prevented full coordination at pocket B" to "It resulted from the steric hindrance of pocket B in the original MOF that prevented full coordination at pocket B" in the manuscript. Please see Page 5, left column, second paragraph in the manuscript.
We have revised "We supposed that because L2 also possessed the ability to coordinate on the metal sites at pocket A" to "It was because L2 possessed the ability to coordinate on the metal sites at pocket A" in the manuscript. Please see Page 6, left column, the first paragraph in the manuscript.
We have revised "We supposed that the installation of L1 at pocket A rigidified the flexibility and restricted the intermolecular stacking between adjacent organic ligands along the b-axis" to "The installation of L1 at pocket A rigidified the flexibility and restricted the intermolecular stacking between adjacent organic ligands on the ab plane" in the manuscript. Please see Page 6, left column, second paragraph in the manuscript.
We have re-witted the discussion of GC separation experiments and added the discussion of the mechanism of separation difference. Please see Page 7, Isomer Separation, in the manuscript.

How do the authors know that the obtained a non-interpenetrated scu MOF instead of the known double interpenetrated version?
Response: Thank you for pointing this out. We have added the following discussion in Supplementary Information to support our statement. Please see Page S8, Supplementary We have added "On the one hand, due to the failure of single crystal synthesis, the structure of Zr-TCPE was analyzed by Pawley refinement. The low R-value indicated the refined profile matched the experimental PXRD pattern very well. The refinement results revealed that each TCPE ligand was connected to four Zr6 clusters and each Zr6 cluster was 19 coordinated with eight TCPE ligands. Then, the non-interpenetrated scu coordination structure was generated in Zr-TCPE. On the other hand, the distance of adjacent Zr6 clusters was 1.77 nm along the a-axis in the simulated non-interpenetrated structure, which was consistent with the HAADF image measurements (d=1.79 nm). However, the distance of adjacent Zr6 clusters in a double-interpenetrated structure would be 0.89 nm, which was not found in any of the HAADF image measurements. Meanwhile, the distance of adjacent Zr6 clusters was 3.02 nm along the b-axis in the simulated non-interpenetrated structure, which was also consistent with the HAADF image measurements (d=3.04 nm). Therefore, we assigned a non-interpenetrated scu structure to Zr-TCPE" Please see Page S11 in Supplementary Information.
We have revised "The measured distance of adjacent Zr6 clusters was around 1.77 nm, which was consistent with the simulated value along the a-axis. Along its perpendicular direction, the b axis, the distance between two adjacent Zr6 clusters was hard to observe directly" to "Due to the limitation of the resolution, the distance between two adjacent Zr6 clusters along the b-axis was hard to observe directly because of the interference from Zr6 clusters along the a-axis. Thus, the denoised images were acquired by an inverse-FFT process after applying a periodic mask to the FFT pattern for measuring the distance of Zr6 clusters. The measured distance of adjacent Zr6 clusters was around 1.79 nm and 3.04 nm ( Supplementary Fig. 4). These values were consistent with the simulated value along the a-axis and b-axis, respectively, in a non-interpenetrated structure, but differed from the distance of adjacent Zr6 clusters in a double-interpenetrated structure (0.89 nm and 1.51 nm, see Supplementary Fig. 3). The results further proved that the Zr-TCPE possessed a non-interpenetrated structure." in the manuscript. Please see Page 3, left column, the last paragraph in the manuscript.
We have added "Besides, the measured distance of adjacent Zr6 clusters along the baxis had also increased from 3.04 nm to 3.25 nm" in the manuscript. Please see Page 4, right column, the first paragraph in the manuscript.     Fig. 23  shows a larger unit cell. This is quite interesting because the installation of both L1 and L2 is expected to rigidify the scu framework.

The pxrd patterns shown in SI
Response: Thank you for pointing this out. We apologize for the mistake in the data processing. We have carefully rechecked the patterns and found that the PXRD pattern of

Regarding, the isomer separation properties that results are confusing and not explained.
For example, it is state "The Zr-TCPE-H showed seriously low separation ability compared with the original Zr-TCPE…". This behavior is attributed to "structural deformation" but no explanation is provided, supported by experimental data. Also the statement "Zr-TCPE-L1-H showed fair separation ability with obvious peak tailing compared with Zr-TCPE-L1 ( Supplementary Fig. 40 and 41), which is due to the existence of flexibility along the c-axis." is confusing.
Response: Thank you for your suggestion. We have added experiments and explanations in the discussion of the isomer separation properties.
We have revised "To investigate the influence of anisotropic rigidification of flexibility on MOF separation properties, high-resolution gas chromatography (GC) technique was implemented" to "It has been reported that the gas chromatographic Response: Thank you for pointing this out. We apologize for the mistake in the data processing. We have carefully rechecked the patterns and found that the PXRD pattern of hardly occupy the Zr-TCPE-DLI with all the possible sites to install these linkers.
We have further increased the amount of L1 and L2 in the synthesis procedure, however, the ratio of L1: L2: TCPE was 0.5: 0.2: 1, and the pocket B was not fully occupied.
Besides, we also tried to add two linkers simultaneously in the synthesis, however, the ratio  5. Authors mentioned that they were not able to get the single crystal despite many trials.
Their experimental section shows that they add water as well in addition to acetic acid.
Water typically results in smaller particles. It could be because of the fact that they are getting other phases such as csq, in that case I recommend not using water but using a little more DMF. Not sure if this would result in diffraction quality crystals but I am nearly confident that this would give that larger crystals.
Response: Thank you for your suggestion. We have tried the synthesis of Zr-TCPE without water but with more DMF, there were still no single crystals obtained. We have also tried other synthesis methods, and despite the increase in particle size, no diffraction-quality crystals were obtained.
We have carried out the trial synthesis of single crystals with more DMF: 10 mg ZrCl4 and 120 μL AA were dissolved in 2/3/4 mL DMF in a glass vial. The vial was heated at 100 o C for 1 h. After cooling down to room temperature, 160 μL AA, and 10 mg H4TCPE ligand 43 were added to the mixture. After sonication, the vial was heated at 120 °C for 24 h. The white product was collected by centrifugation and washed with DMF and EtOH three times, respectively, before drying at 60 o C under vacuum. The particle size of Zr-TCPE slightly increased with the increase of DMF content, but the size was non-uniform and these materials were not suitable for single crystal X-ray diffraction.  We have carried out the trial synthesis of single crystals with BA as a modulator: The 10 mg ZrCl4, 200/300/400/600/800 mg BA, and 10 mg H4TCPE were dissolved in 2 mL DMF in a glass vial. The vial was heated at 120 o C for 24 h or 72 h. After cooling down to room temperature, the product was collected by centrifugation and washed with DMF and EtOH three times, respectively, before drying at 60 o C under vacuum The particle size of Zr-TCPE-BA increased with the increase of BA content, but these materials were not suitable for single-crystal X-ray diffraction.  Figure R4. The PXRD patterns of Zr-TCPE-300 mg BA and Zr-TCPE-400 mg BA. 46 We have carried out the trial synthesis of single crystals with TFA as a modulator: The 10 mg ZrCl4, 50/100/150/200 L TFA, and 10 mg H4TCPE were dissolved in 2 mL DMF in a glass vial. The vial was heated at 120 o C for 24 h. After cooling down to room temperature, the product was collected by centrifugation and washed with DMF and EtOH three times, respectively, before drying at 60 o C under vacuum. The particle size of Zr-TCPE-TFA increased with the increase in TFA content, but these materials were not suitable for single-crystal X-ray diffraction.  Information.

Reviewers' Comments:
Reviewer #1: Remarks to the Author: It is always gratifying when an author chooses to take on board referees' comments, and uses them to improve the quality of the paper. The clarification of the experimental details makes this work more scientifically sound and significantly enhance the importance of this paper, and I am now happy to recommend acceptance.
Reviewer #2: Remarks to the Author: Dear Editor, The revised manuscript is improved. However, the quality of the experimental data, the discussion and clarity of the results, as well as the overall originality and novelty of the work, are not suitable for publication to Nature Communications. I recommend publication in a more specialized journal after taking into account the following comments. 1. The assignment to a non-interpenetrated scu structure is not convincing. This is mainly for the following reasons: a) Pawley (and not Powley as written in the legend of Fig. 1) refinement is a structureless methodology. This means that there are no atoms and atomic positions involved in the refinement, as in the case of Rietveld refinement. Therefore, the question is: how the authors obtained the structural images shown in the manuscript and in the SI file? The same question holds for the structure file used for the theoretical calculation. Did the authors apply a Rietveld refinement in their experimental pxrd data? If yes they should provide all the corresponding details, including where they obtained the initial structural model. b) It is evident from the HRTEM images that there is a superposition of single crystals. For accurate assignment of lattice distances HRTEM images from single crystals are necessary.
2. Perhaps a direct method for the assignment of a non-interpenetrated scu structure is to record accurate gas sorption isotherms such as N2 at 77K but preferable Ar at 87 K (accurate micropore analysis starting from very low relative pressures, eg 10-6 p/p0), from which accurate pore size distribution data can be extracted. A non-interpenetrated scu is expected to have a significant larger pore size and also a significantly higher total pore volume compared to a interpenetrated scu. This kind of data is not available and the authors are strongly advised to execute these experiments.

Comments:
The revised manuscript is improved. However, the quality of the experimental data, the discussion and clarity of the results, as well as the overall originality and novelty of the work, are not suitable for publication to Nature Communications. I recommend publication in a more specialized journal after taking into account the following comments.
Response: We highly appreciate your constructive comments. We totally understand your concern about the non-interpenetration of scu structure from the technical aspect.
We have added Rietveld refinement, ultrahigh-resolution low-dose HRTEM measurement, and Ar adsorption-desorption measurement at 87 K to confidently confirm the successful synthesis of non-interpenetration of scu structure.  We have revised "Powley" to "Pawley".
We have revised "Thus, the powder X-ray diffraction (PXRD) refinement was utilized to reveal the structure of this material" to "Thus, the powder X-ray diffraction (PXRD) Rietveld refinement was utilized to reveal the structure of this material ( Supplementary Fig. 1  spacing of (020), (001), and (021) was 1.53 nm, 1.26nm, and 0.95 nm, respectively, which was consistent with the d-spacing from PXRD refinement ( Fig. 1c and Supplementary Table 3). For better interpretation, the raw image was processed by correcting the effect of the contrast transfer function (CTF) of the objective lens ( Supplementary Fig. 5). Furthermore, the simulated electron diffraction (ED) pattern along the [100] direction was consistent with the selected area electron diffraction (SAED) pattern of the HRTEM image ( Supplementary Fig. 6). The average background subtraction filter (ABSF)-filtered CTF-corrected image in Fig. 1d matched well with the simulated structure of Zr-TCPE along the [100] direction zone axis ( Supplementary   Fig. 3). The distances between two adjacent Zr6 clusters were measured from the linear profiling along the b-and c-axis as 1.53 nm and 1.23 nm, respectively ( Supplementary   Fig. 7), which were consistent with the simulated structure of Zr-TCPE ( Supplementary   Fig. 3). Besides, the distance of Zr6 clusters along the c-axis was comparable to the caxis cell parameter in non-interpenetrated Zr-TCPE. Thus, not interpenetrated but noninterpenetrated structure was assigned to Zr-TCPE. We also obtained diffraction information of (020) and (130) planes from the FFT images ( Supplementary Fig. 8). All

Perhaps a direct method for the assignment of a non-interpenetrated scu structure is
to record accurate gas sorption isotherms such as N2 at 77K but preferable Ar at 87 K (accurate micropore analysis starting from very low relative pressures, eg 10 -6 p/p0), from which accurate pore size distribution data can be extracted. A non-interpenetrated scu is expected to have a significant larger pore size and also a significantly higher total pore volume compared to a interpenetrated scu. This kind of data is not available and the authors are strongly advised to execute these experiments.
Response 2: Thank you for pointing this out. To confirm the non-interpenetration structure, we have added the Ar adsorption isotherms of Zr-TCPE at 87K starting from very low relative pressure (3.17×10 -6 P/P0). The BET surface area, pore size distribution, and total pore volume obtained from the Ar adsorption isotherms are consistent with those from N2 adsorption isotherms. The main pore size distribution of Zr-TCPE calculated from the Ar adsorption isotherms was 8.7 Å, which was slightly smaller than the value (9.4 Å) obtained from N2 adsorption. This measured pore size was consistent with the simulated value (9.3 Å) of pocket A in Zr-TCPE. This result indicated the noninterpenetrated structure of Zr-TCPE. Please see Page S19, Supplementary Figure 13, and Supplementary Table 4 in Supplementary Information.
We have revised "To characterize the porosity of Zr-TCPE, N2 adsorptiondesorption isotherms were performed at 77 K. As shown in Supplementary Fig. 8, the isotherms presented fully reversible type-I behavior, indicating the microporous characteristic of Zr-TCPE. The Brunauer-Emmett-Teller (BET) surface area of Zr-TCPE was 934 m 2 /g and the pore size distribution of Zr-TCPE calculated by the DFT method was 9.4 Å, which was consistent with the simulated size of pocket A" to "To accurately characterize the porosity of Zr-TCPE, both the N2 and Ar adsorptiondesorption isotherms were performed at 77 K and 87 K, respectively. As shown in Supplementary Fig. 13, both the isotherms presented fully reversible type-I behavior, indicating the microporous characteristic of Zr-TCPE. The Brunauer-Emmett-Teller (BET) surface area, pore size distribution, and total pore volume obtained from the Ar adsorption isotherms are consistent with those from N2 adsorption isotherms (Supplementary Table 4). The main pore size of Zr-TCPE calculated by the DFT method from Ar and N2 sorption isotherms was 8.7 Å and 9.4 Å, respectively. The