Higher-generation type III-B rotaxane dendrimers with controlling particle size in three-dimensional molecular switching

Type III-B rotaxane dendrimers (T3B-RDs) are hyperbranched macromolecules with mechanical bonds on every branching unit. Here we demonstrate the design, synthesis, and characterization of first to third (G1–G3), and up to the fourth (G4) generation (MW > 22,000 Da) of pure organic T3B-RDs and dendrons through the copper-catalyzed alkyne–azide cycloaddition (CuAAC) reaction. By utilizing multiple molecular shuttling of the mechanical bonds within the sphere-like macromolecule, a collective three-dimensional contract-extend molecular motion is demonstrated by diffusion ordered spectroscopy (DOSY) and atomic force microscopy (AFM). The discrete T3B-RDs are further observed and characterized by AFM, dynamic light scattering (DLS), and mass spectrometry (MS). The binding of chlorambucil and pH-triggered switching of the T3B-RDs are also characterized by 1H-NMR spectroscopy.


1.
Have the authors tried the synthesis of the functionalized G4 dendrons with azide and acetylene moieties and the targeted G4 rotaxane dendrimer? If so, the authors should provide the primary results about the synthetic results. I understand that it is very difficult to get G4 rotaxane dendrimer. But it will be very helpful for readers to understand better about this chemistry if the authors provide some primary results;

2.
In the case of the purity of the ionic and neutral G1-G3 rotaxane dendrimers, only DOSY experiments were performed in the manuscript. However, HPLC or GPC analysis might be very helpful to check the purity; 3.
In the Figure 8, the AFM images showed the morphology and size information of the rotaxane dendrimers G1-G3. However, the morphology of the G1 is inhomogeneity, which might be due to the slight aggregation as claimed by the authors. In order to give a better description of the increase of size, a uniform morphology is preferred. Furthermore, TEM measurements should also be performed to visual the morphology and size information;

4.
In the supporting information, the peak of G3 in the DLS analysis ( Figure S23) was too broad and not a single peak, which was not accurate enough to give any size information;

5.
The sizes of the resultant rotaxane dendrimers were measured by three different analysis means, i.e, DOSY, AFM and DLS. However, the values of the same generation strongly varied in different analysis methods. For example, by AFM analysis, the average height of G3 is ~10.96 nm, whereas DLS measurement showed ~4.57 nm, and DOSY displayed ~1.35 nm. A reasonable explanation of the difference should be provided;

6.
In order to display a more intuitive descriptions of the size change after deprotonation, AFM and DLS analysis of the corresponding neutral rotaxane dendrimers should be performed;

7.
Simply based on the NMR titration, how to confirm the exact numbers of guest molecules binding with the rotaxane dendrimers?

8.
As a critical parameter on host-guest chemistry, the binding constants of the rotaxane dendrimers G1-G3 towards the guests are very necessary.
Reviewer #2 (Remarks to the Author): In the previously published paper (W. Wang, L.-J. Chen, X.-Q. Wang, B. Sun, X. Li, Y. Zhang, J. Shi, Y. Yu, L. Zhang, M. Liu and H.-B. Yang, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 5597), Yang and coworkers has well established the synthesis, characterization, and functionalization of highergeneration (up to fourth-generation) organometallic rotaxane branched dendrimers using pillar [5]arene and sequential coupling-deprotection-coupling processes. Although this manuscript use the another host-guest interaction and coupling reactions, the point (fourth-generation) is still in absence of innovation. Therefore, this manuscript should not publish in Nature Communications.
Reviewer #3 (Remarks to the Author): I would like to commend on this paper according to the guideline for reviewers.
1. What are the major claims of the paper? -The author successfully synthesized and fully characterized a series of type III-B rotaxane dendrimers (dendrons) up to 4th generation. They also observed the pH controlled switching behaviour and their preliminary application for controlled drug release.
2. Are they novel and will they be of interest to others in the community and the wider field? -Synthesis of mechanically locked rotaxane dendrimers is a very challenging task, especially the type III-B dendrimers. This work is of particularly interest to the general readers in supramolecular chemistry. It is also interesting for readers in structural organic chemistry and organic materials science.
3. Is the work convincing? -I have carefully checked all the characterization data (1D and 2D NMR, HR MS, LS, AFM , etc.), I think all the compounds are well characterized and it is convincing.
4. Do you feel that the paper will influence thinking in the field? -Yes, the synthetic strategy is new, and the switching behaviour and the preliminary test give some new insight into the potential application of using rotaxane dendrimers or hyperbranched polymers for controlled drug release applications. 5. Appropriateness and validity of any statistical analysis, as well the ability of a researcher to reproduce the work.
-The data are well analysed and is reproducible.
I only have minor issue. The authors talked about the after deprotonation, the crown ether will stay at the triazole unit based on the theoretical calculations. It would be nice if they can further elaborate this, for example, what is the energy difference between the two possible states?
The paper was well written and the supporting information is clear.

Reviewer 1
We greatly thank and appreciate the reviewer acknowledged the novelty of our work, also giving us some valuable feedback and useful suggestions to modify and enhance the quality of the manuscript. We have now taken into account carefully of all suggestions and concerns given by the reviewer, and have modified the manuscript accordingly.
1. Have the authors tried the synthesis of the functionalized G4 dendrons with azide and acetylene moieties and the targeted G4 rotaxane dendrimer? If so, the authors should provide the primary results about the synthetic results. I understand that it is very difficult to get G4 rotaxane dendrimer. But it will be very helpful for readers to understand better about this chemistry if the authors provide some primary results.

Response:
We have tried the synthesis of the functionalized G4 dendrons with azide and acetylene moieties and the targeted G4 rotaxane dendrimer. However, the resulted data are too preliminary, wherein we only got the 1 H NMR and the 13 C NMR spectroscopic results. The preliminary data are shown below: 1 In order to fully study the G4 [31]rotaxane dendrimer, it requires a much longer or unknown time to finish up all repeated synthetic steps and characterization, including 1-D NMR, 2-D NMR, MS, AFM, DLS, GPC analysis, etc. Therefore, at this stage, we tend to not disclose the preliminary results on the G4 [31]rotaxane dendrimer. In page 14 of the revised manuscript, we have added a sentence, stating that we are continuing the further investigation on G4 and higher generation rotaxane dendrimers.

In the case of the purity of the ionic and neutral G1-G3 rotaxane dendrimers,
only DOSY experiments were performed in the manuscript. However, HPLC or GPC analysis might be very helpful to check the purity.

Response:
We agreed with the reviewer concerning about the purity of the rotaxane dendrimers with only DOSY experiments. We then performed an additional GPC analysis for the neutral G1 rotaxane dendrimer. From the GPC chromatogram, we can observe a single peak, further confirming the purity of the compounds.
For other higher generation neutral rotaxane dendrimers, unexpectedly due to the strong adsorption with the GPC columns, we cannot obtain satisfactory results.
For the ionic G1-G3 rotaxane dendrimers, we have not performed the GPC/HPLC analysis. Since the neutral rotaxane dendrimers have already strongly adhered on the columns and that the ionic rotaxane dendrimers carry high charges, they may not be able to elute out from the column and would cause column damage. Therefore, for the cases of ionic G1-G3 rotaxane dendrimers, we think that the combinations of clear 1-D NMR ( 1 H, 13 C, 31 P), 2-D NMR (NOESY and DOSY) spectroscopy and ESI-MS analysis can confirm the purity of the targeted compounds. Figure 8, the AFM images showed the morphology and size information of the rotaxane dendrimers G1-G3. However, the morphology of the G1 is inhomogeneity, which might be due to the slight aggregation as claimed by the authors. In order to give a better description of the increase of size, a uniform morphology is preferred. Furthermore, TEM measurements should also be performed to visual the morphology and size information.

In the
Response: We agreed with the reviewer about the inhomogeneity and aggregation of AMF image (Figure 8) of G1 rotaxane dendrimer. We then carefully prepared and retried the G1 AFM sample analysis. We clearly observed the monodispersed uniform G1 rotaxane dendrimer on mica surface, and the Figure 8 and its discussion (page 11) were modified accordingly. 5 We have also tried the TEM analysis for rotaxane dendrimers, however, we cannot observe the useful morphology and size information, possibly due to (1) relatively low resolution to observe objects of 1-10 nm and (2)  Since AFM has a better resolution towards the pure organic small molecules of 1-10 nm, we therefore believed that the images obtained from AFM are relatively reliable to obtain the morphology and size information of all ionic and neutral rotaxane dendrimers.
4. In the supporting information, the peak of G3 in the DLS analysis ( Figure S23) was too broad and not a single peak, which was not accurate enough to give any size information. While in AFM, it is measured the rotaxane dendrimer in solid state, after the spin coating on the mica surface. In the case of G3, due to the rigidity bring by the mechanical bonds, as well as the high generation, the adsorption forces between rotaxane dendrimer and mica surface will decrease with higher generation, thus G3

Response
would not be flatten (nor spherical) on mica surface, giving a higher height than DOSY and DLS. (Langmuir 2000, 16, 5613-5616.) From these three techniques, we can observe a general increase trend of size increase from G1-G3 gradually. This point has been discussed in pages 11-12.
6. In order to display a more intuitive descriptions of the size change after deprotonation, AFM and DLS analysis of the corresponding neutral rotaxane dendrimers should be performed.

Response:
We agreed with the reviewer, concerning the characterization of size change of rotaxane dendrimers after the deprotonation using AFM and DLS. We study the AFM of neutral G1-G3 rotaxane dendrimers, the results showed that the heights in neutral G2 and G3 rotaxane dendrimers were higher than the ionic ones, while neutral G1 showed a similar height. The below paragraph and new figures were added to the revised manuscript (pages 11-12) and SI (S27-29). "The morphologies of neutral G1-G3 T3B-RDs were also be visualized. (Supplementary Fig. S27-S29) In neutral G1, similar height (1.23 ± 0.02) was observed, as there were only four shuttling triazole stations for the ring after deprotonation and having a relatively small molecular weight. The height difference between the ionic and neutral G1 rotaxane dendrimer is 7%. In contrast, the heights of neutral G2 (Δh = 0.87 nm, 22%) and neutral G3 (Δh = 5.89 nm, 54%) showed higher heights than the original ionic states. Since G2 rotaxane dendrimer contains 12 triazoles while G3 contains 28 triazoles, the ring shuttling between the triazoles are significant, and that the molecules would tend to form the most stable conformation, avoiding the steric hindrance from dendrons after the deprotonation (relaxed state). As the generation increases, a larger height difference was observed due to the number of switching state was also increased. These T3B-RDs generally possessed some rigidity, whereas they would not be easily flatten on surface." Neutral G1 For DLS analysis, we have tried several conditions in the analysis of neutral G1-G3 rotaxane dendrimers, however, no satisfactory results were obtained. Since the (n)triazole in the rotaxane dendrimers were equivalent to each other, the macrocycle will keep shuttling between (n)triazole, together with flexibility of the molecules itself in solution state. As DLS measurement depends on the scattered light radiation intensity decay as a function of time, when the particles dispersed in solution, the rotaxane dendrimer (particle) size will keep changing, thus resulting the instability in DLS analysis. More intuitive descriptions of the size change after deprotonation are now realized by the new AFM analysis in addition with the DOSY NMR data.

Simply based on the NMR titration, how to confirm the exact numbers of guest molecules binding with the rotaxane dendrimers?
Response: We thank the reviewer to pointing out the unclear explanation of molecules binding with the rotaxane dendrimers based on NMR titrations. In terms of determination of the number of guest molecules binding with rotaxane dendrimers, we compared their NMR chemical shifts of the original (ionic) dendrimer and the guest-bound dendrimers.

Neutral G3
Neutral G2 In G1 dendrimer's NMR titration ( Figure S42), when titrating up to 2.0 equivalents of guest molecules, the unique hydrogen-bonded DBA proton signals are still not restored (green color), meaning that the macrocycle is still located at the triazole, while the DBA interacts with the carboxylate. Once excess acid (guest) was added, the carboxylate reprotonated to carboxylic acid and the macrocycle moved from triazoles back to the original site. Therefore, G1 could bind with two guest molecules by observing from NMR titration. In the case of G2 dendrimer's NMR titration ( Figure S44), we could observe the unique DB24C8-DBA binding signals after the addition of 4.0 equivalents of guest molecules, therefore one G2 dendrimer was able to bind with 4 guest molecules. In G3 dendrimer's NMR titration ( Figure S46), similar peak shifts were used for the determination of guest molecules binding to the rotaxane dendrimers, and 8 guest molecules were bound. In conclusion, guest substrates can be bound to the amine sites at the outer layer of the dendrimers. This point has been discussed in the revised manuscript (page 13 and figures S43-S47).   The general structure of the two rotaxane dendrimers:

Type III-A rotaxane dendrimer Type III-B rotaxane dendrimer
The work presented by Yang et al on type III-A rotaxane dendrimers is different from our type III-B rotaxane dendrimers, in terms of structures, nature (organometallic and organic) and synthetic approaches (divergent and convergent). Also, in our work on higher generation type III-B rotaxane dendrimers, we have demonstrated the pH responsive switching process and the size difference from the discrete ionic and neural G1-G3 rotaxane dendrimers, where the dendrimers synthesized by Yang et al cannot be achieved. We also summarized the differences of our work in comparison to Yang's work in the following

Surface modification
Ferrocene -Therefore, our work on higher generation type III-B rotaxane dendrimers is different from Yang's published type III-A rotaxane dendrimers in various aspects.

Reviewer 3
I only have minor issue. The authors talked about the after deprotonation, the crown ether will stay at the triazole unit based on the theoretical calculations. It would be nice if they can further elaborate this, for example, what is the energy difference between the two possible states?