A practical preparation of bicyclic boronates via metal-free heteroatom-directed alkenyl sp2-C‒H borylation

Bicyclic boronates play critical roles in the discovery of functional materials and antibacterial agents, especially against deadly bacterial pathogens. Their practical and convenient preparation is in high demand but with great challenge. Herein, we report an efficient strategy for the preparation of bicyclic boronates through metal-free heteroatom-directed alkenyl sp2-C‒H borylation. This synthetic approach exhibits good functional group compatibility, and the corresponding boronates bearing halides, aryls, acyclic and cyclic frameworks are obtained with high yields (43 examples, up to 95% yield). Furthermore, a gram-scale experiment is conducted, and downstream transformations of the bicyclic boronates are pursued to afford natural products, drug scaffolds, and chiral hemiboronic acid catalysts.

positions ortho and para to the alkenyl group. The inductively electron-withdrawing substituent F does not sufficiently explore this question. In addition, for the N-directed and S-directed borylation of terminal alkenes (Fig. 3), results with directing groups NH2 and SH are notable absences. These examples are needed to fully test the generality of the method and should be included.
In addition, there are a number of other items that should be addressed, as follows: • In Figure 1d, it is recommended that structure B show a positive charge on O and a negative charge on B.
• The values in Table 1 do not appear to match the text. Specifically, the text says "When excessive A5 (2.0 equiv.) was used, the yield of 3aa dramatically decreased and significant amount of uncyclized side product o-isopropenylphenol 1bb was observed (entries 9-11)." However, entries 9-11 in Table 1 all show 100% conversion and 99% yield. Also, 2.0 equiv. of A5 is the condition used in Entry 6; the authors should make clear in describing the later examples that they are referring to excessive A5 relative to BBr3.
• The text describing Fig. 5 appears to be corrupted, i.e. immediately after reference 42, a sentence begins with the uncapitalized word "functionalized" and does not make sense.
• The use of this methodology to synthesize the beta-lactamase inhibitor QPX7728 provides a nice demonstration of its utility. However, the implication that this represents a 4-step synthesis from a commercially-available starting material, and the statement "compared with the previous synthesis, our concise and practical approach was cost-effective" ignores the fact that the starting material 3-bromo-4fluoro-2-methoxybenzaldehyde costs an exorbitant amount ($500,000/kg based on gram amounts). The wording should be revised to simply describe this as an efficient alternative method; while it is important to acknowledge the published synthesis, cost comparisons should be avoided. Also, as noted above, it would be useful to know if the boron insertion can be achieved with the carboxylate moiety already in place (presumably as an ester). • It is not at all clear how an MIC of 2.67 µg/mL against E. coli is obtained for compound 3bp. It is generally accepted that the MIC corresponds to "no growth." According to the plot on page S31 of the Supporting Info, this is achieved at 2.2 log10, which would be well over 100 µg/mL. If it is the case that all the MICs are ≥100 µg/mL, then the antibacterial evaluation should be removed entirely from the manuscript since many organic compounds have modest antibacterial activity, in most cases by nonspecific and uninteresting mechanisms of action. On the other hand, if the MIC vs. E. coli is in fact <3 µg/mL, it is indeed an interesting result that is worthy of reporting.
Reviewer #3 (Remarks to the Author): Tian and co-workers report a new, transition metal-free synthetic protocol for bicyclic boronate synthesis via heteroatom-directed C(sp2)-H borylation. These compounds are relatively rare in the literature but are found in several important β-lactamase inhibitors, as well as fluorescent sensors and functional materials. Hence, this report is likely to be of significant interest to the synthetic chemistry community, including medicinal chemists.
The scope of the method is fairly broad, and both electron-rich and electron-poor arene groups are compatible, as well as different alkene substitution patterns. However, no Lewis basic heteroaromatic substrates (e.g., methoxy vinyl pyridines) are demonstrated, and one would assume that they would complex the BBr3 non-productively? Notably, the reaction can be carried out cleanly on gram scale, which highlights its practicality.
The authors also provide extensive examples of further derivatisations of the bicyclic boronates, which is great to see. Moreover, they applied their method to the synthesis of the ultrabroad-spectrum βlactamase inhibitor (±)-QPX7728, and also showed that several of the boronates displayed antibacterial activity against gram-negative bacterium E. coli. This is a very well put together manuscript and was an enjoyable read. The Supporting Information is prepared to a good standard and the NMR spectra are all of high quality. Given the novelty of the approach to bicyclic boronates and the widespread interest in these unusual compounds, I have no hesitation in recommending this work for publication in Communications Chemistry.
We appreciate the time and effort that the editor and reviewers dedicated to providing feedback on our previous manuscript (Ms. No COMMSCHEM-22-0574A) and are grateful for the insightful comments on and valuable improvements to our paper. We have carefully reviewed the comments and have revised the manuscript extensively. Our response is given in a point-to-point manner as follows.

Reviewer #1:
The authors report a metal-free heteroatom-directed alkenyl sp2-C-H borylation. The alkene starting materials, in either (Z)-or (E)-form, could produce the desired bicyclic boronates, which significantly expanded the substrate scope of such borylative cyclization. In general, this manuscript is well-written and clear, the science is interesting and I consider it to be high-impact based on the ability to generate bicyclic boronates. Also noteworthy is the demonstration of the downstream transformations of products to afford natural products, drug scaffolds, and chiral hemiboronic acid catalysts. The work should be of broad appeal, however, significant revisions and additional experiments are required before the work is suitable for publication in COMMSCHEM.
[Q1] 1) What's the reactivity of BF3, BCl3 and BI3? The author should add these important information in the reaction optimization.
[Q3] 3) Regarding the mechanism, the authors proposed the ortho-quinone methide intermediate. However, I didn't see a solid evidence to prove this step. The authors should try some gem-substituted alkenes. Could they observe the rearrangement product. Furthermore, the quinone may be stabilized.
[Q4] 4) The authors proposed the dienol borate formed ahead of C-B bond formation. Why OMe in 3al is compatible. Is it possible that BBr3 coordinates with OMe and then actives the alkenyl group? They should give explanations and more experiments to support the proposed pathway.
[Q5] 5) In Supporting information, starting material and final product was synthesized, purified and thoroughly characterized by NMR, IR and HRMS. Although the quality of the data is good, the text of the SI should be proofread and carefully checked. There are multiple typos, editing issues, etc. Furthermore, the NMR spectra of some compounds, e.g. 1bf, 1ar, 1ba, 1bf, 3ba, 3bh and 3bj, show that these compounds are not pure enough. These samples must be re-purified.

Response to Reviewer 1
[Q1] 1) What's the reactivity of BF3, BCl3 and BI3? The author should add these important information in the reaction optimization.
Response: Thanks for the reviewer's advice. We tried other commercially available borylation reagents, such as BF3· OEt2 and BCl3, and the experimental results have been added in Table 1 (Optimization of Reaction Conditions). Because the reagent BI3 was expensive and need very long delivery period, so we did not test it.
[Q3] 3) Regarding the mechanism, the authors proposed the ortho-quinone methide intermediate. However, I didn't see a solid evidence to prove this step. The authors should try some gem-substituted alkenes. Could they observe the rearrangement product. Furthermore, the quinone may be stabilized. Response: Thanks for the reviewer's advice. At the beginning of preparing the manuscript, we proposed an intermediate Int-AB, which is a resonance of B. After discussing with colleagues and reading related literature, we thought intermediate B was more stable, so it was presented in the manuscript. Very recently, Michael Ingleson and coworkers reported "Haloboration of o-Alkynyl Phenols Generates Halogenated Bicyclic-Boronates" and discussed a similar intermediate I3A which is the structure of ortho-quinone methide.
Of course, according to reviewer's advice, a gem-substituted alkene 1ca was synthesized and tested in our reaction condition. From crude 13C-NMR and 11B-NMR, a peak of 185 ppm which indicated the formation of quinone methide was found. However, when we wanted to prepare the stable pinacol boronate (3ca), the signal of quinone disappeared and no desired product 3ca was obtained. Also, we tried to grow a crystal, but compound 3a cannot form a solid for long time under freezing condition.
[Q4] 4) The authors proposed the dienol borate formed ahead of C-B bond formation. Why OMe in 3al is compatible. Is it possible that BBr3 coordinates with OMe and then actives the alkenyl group? They should give explanations and more experiments to support the proposed pathway. Response: Thanks for the reviewer's advice. Because both lone pair electrons and π electrons can coordinate with Lewis acid BB3, when there is only 1.1 equivalent of BBr3 under our standard conditions, the complex of lone pair electrons and π electrons simultaneously coordinated with BB3 is the major product and participates in the following reaction. So, this is maybe the reason of -OMe compatible in 3al and 3am.
In addition, we proposed that both routes were possible and could lead to intermediate of quinone methide. Because the first stage of path b is a reversible reaction, the possibility of the irreversible path a is higher.
[Q5] 5) In Supporting information, starting material and final product was synthesized, purified and thoroughly characterized by NMR, IR and HRMS. Although the quality of the data is good, the text of the SI should be proofread and carefully checked. There are multiple typos, editing issues, etc. Furthermore, the NMR spectra of some compounds, e.g. 1bf, 1ar, 1ba, 1bf, 3ba, 3bh and 3bj, show that these compounds are not pure enough. These samples must be re-purified. Response: Thanks for your kind suggestions! The SI have been carefully rechecked. The typo errors and editing issues have been corrected. In addition, the impure compounds, such as 1bf, 1ar(1as), 1ba, 1bf, 3bh and 3bj, have been repurified, and their yield and NMR spectra have also been revised in the modified manuscript and SI. For the compound 3ba, we have repeatedly prepared it many times and the reported result is the best, also we have tried to re-purify it by preparative chromatography, but the impurity became more and more.

Reviewer #2:
This manuscript describes a novel and useful method for preparing bicyclic boronates using a [1,5]-sigmatropic rearrangement without metal catalysis and will be of high interest to readers in the field. It is clearly written and well-referenced; the characterization of compounds is thorough, and the spectral data show high purity. Overall, the subject and quality warrant publication; however, there are certain notable absences in the demonstration of scope that should be addressed prior to acceptance.
[Q1] Specifically, missing from the various substituents on the aromatic ring that have been explored are those that are strongly electronwithdrawing by resonance, such as -CO2R or -CN, particularly in the positions ortho and para to the alkenyl group. The inductively electron-withdrawing substituent F does not sufficiently explore this question.
[Q2] In addition, for the N-directed and S-directed borylation of terminal alkenes (Fig. 3), results with directing groups NH2 and SH are notable absences. These examples are needed to fully test the generality of the method and should be included.
In addition, there are a number of other items that should be addressed, as follows: [Q3] In Figure 1d, it is recommended that structure B show a positive charge on 0 and a negative charge on B.
[Q4] The values in Table 1 do not appear to match the text. Specifically, the text says "When excessive AS (2.0 equiv.) was used, the yield of 3aa dramatically decreased and significant amount of uncyclized side product o isopropenylphenol 1bb was observed (entries 9-11)." However, entries 9-11 in Table 1 all show 100% conversion and 99% yield. Also, 2.0 equiv. of AS is the condition used in Entry 6; the authors should make clear in describing the later examples that they are referring to excessive AS relative to BBr3.
[Q5] The text describing Fig. S appears to be corrupted, i.e. immediately after reference 42, a sentence begins with the uncapitalized word "functionalized" and does not make sense.
[Q6] • The use of this methodology to synthesize the beta-lactamase inhibitor QPX7728 provides a nice demonstration of its utility. However, the implication that this represents a 4-step synthesis from a commercially-available starting material, and the statement "compared with the previous synthesis, our concise and practical approach was cost-effective" ignores the fact that the starting material 3-bromo-4fluoro-2-methoxybenzaldehyde costs an exorbitant amount ($S00,000/kg based on gram amounts). The wording should be revised to simply describe this as an efficient alternative method; while it is important to acknowledge the published synthesis, cost comparisons should be avoided.
[Q7] Also, as noted above, it would be useful to know if the boron insertion can be achieved with the carboxylate moiety already in place (presumably as an ester).
[Q8] • It is not at all clear how an MIC of 2.67 µg/ml against E.coli is obtained for compound 3bp. It is generally accepted that the MIC corresponds to "no growth." According to the plot on page S31 of the Supporting Info, this is achieved at 2.2 log10, which would be well over 100 µg/ml. If it is the case that all the MICs are 100 µg/ml, then the antibacterial evaluation should be removed entirely from the manuscript since many organic compounds have modest antibacterial activity, in most cases by nonspecific and uninteresting mechanisms of action. On the other hand, if the MIC vs. E.coli is in fact <3 µg/ml, it is indeed an interesting result that is worthy of reporting.

Response to Reviewer 2
[Q1] Specifically, missing from the various substituents on the aromatic ring that have been explored are those that are strongly electron-withdrawing by resonance, such as -CO2R or -CN, particularly in the positions ortho and para to the alkenyl group. The inductively electron-withdrawing substituent F does not sufficiently explore this question. Response: Thanks for the reviewer's advice. During this research, we have planned to introduce an ester group in the positions ortho and para to the alkenyl group and tried to develope a concise route to synthesize QPX7728-Na, however, when the strongly electron-withdrawing substituents, such as -CO2R, -CHO,-CN, and -NO2 was introduced to the aromatic ring, the substrate become inert under the standard condition (presented in the following scheme). If the A5 was absent from the reaction system, only demethylated by-product was found without bicyclic boronate.
[Q2] In addition, for the N-directed and S-directed borylation of terminal alkenes (Fig. 3), results with directing groups NH2 and SH are notable absences. Response: Thanks for the reviewer's advice. The substrate 1ea with amine was explored under the same reaction condition as the substrate 1bh, although the reaction time was extended to 12 h at room temperature, there was also no reaction.
The substrate 1fe with the directing group SH was difficult to synthesize. We have tried some routes and finally obtained 50 mg through the following scheme E. We tried various reaction conditions, and changed the temperature, time, the amount of 2,6-di-tert-butylpyridine(A5), to our frustration, the results were that the 1fe disappeared, but no desired product was observed. The reviewer can also find the reaction scheme and NMR spectra of starting material as follows.

O-(2-(prop-1-en-2-yl)phenyl) dimethylcarbamothioate (1fc)
[Q3] • In Figure 1d, it is recommended that structure B show a positive charge on 0 and a negative charge on B. The reviewer can also find them Response: Thanks for the reviewer's advice. The charge has been added on the O and B in the revised manuscript, and the reviewer can also find it as follows.
[Q4] • The values in Table 1 do not appear to match the text. Specifically, the text says "When excessive AS (2.0 equiv.) was used, the yield of 3aa dramatically decreased and significant amount of uncyclized side product o isopropenylphenol 1bb was observed (entries 9-11)." However, entries 9-11 in Table 1 all show 100% conversion and 99% yield. Also, 2.0 equiv. of AS is the condition used in Entry 6; the authors should make clear in describing the later examples that they are referring to excessive AS relative to BBr3.
Response: Thanks for the reviewer's advice. The descriptions have been corrected in the revised manuscript, and the reviewer can also find it as follows.
[Q5] The text describing Fig. S appears to be corrupted, i.e. immediately after reference 42, a sentence begins with the uncapitalized word "functionalized" and does not make sense.

Response:
We very much appreciate the reviewer's advice, the typo error has been corrected in the revised manuscript, and the reviewer can also find it as follows.
[Q6] The use of this methodology to synthesize the beta-lactamase inhibitor QPX7728 provides a nice demonstration of its utility. However, the implication that this represents a 4-step synthesis from a commercially-available starting material, and the statement "compared with the previous synthesis, our concise and practical approach was cost-effective" ignores the fact that the starting material 3-bromo-4-fluoro-2methoxybenzaldehyde costs an exorbitant amount ($S00,000/kg based on gram amounts). The wording should be revised to simply describe this as an efficient