Ligand-promoted cobalt-catalyzed radical hydroamination of alkenes

Highly regio- and enantioselective intermolecular hydroamination of alkenes is a challenging process potentially leading to valuable chiral amines. Hydroamination of alkenes via metal-catalyzed hydrogen atom transfer (HAT) with good regioselectivity and functional group tolerance has been reported, however, high enantioselectivity has not been achieved due to the lack of suitable ligands. Here we report a ligand-promoted cobalt-catalyzed Markovnikov-type selective radical hydroamination of alkenes with diazo compounds. This operationally simple protocol uses unsymmetric NNN-tridentate (UNT) ligand, readily available alkenes and hydrosilanes to construct hydrazones with good functional group tolerance. The hydrazones can undergo nitrogen–nitrogen bond cleavage smoothly to deliver valuable amine derivatives. Additionally, asymmetric intermolecular hydroamination of unactivated aliphatic terminal alkenes using chiral N-imidazolinylphenyl 8-aminoquinoline (IPAQ) ligands has also been achieved to afford chiral amine derivatives with good enantioselectivities.

Lu and co-workers report Markovnikov-type selective radical hydroamination of alkenes w ith diazo compounds by means of hydrogen atom transfer. The tridentate ligand effect and broad scope is quite intriguing. The authors synthesized the scalemic products using chiral ligand. The mechanism of this transformation is also quite intriguing. Unfortunately, the scope is too narrow (only two examples). It should be noted that Pronin recently reported the enantioselective hydroalkoxylation of tertiary allylic alcohol by hydrogen atom transfer -initiated process (JACS, 17532, 2019). A similar transformation using diazo compound has been reported Cui and co-worker, as the authors cited in ref. 30. The authors should correct the reference 19 in Scheme 2 (maybe 49). Overall, this work does not meet requirements of novelty to accept publication in Nature communication.
Reviewer #2 (Remarks to the Author): In this submission Lu and coworkers have developed a highly efficient hydroamination reaction of terminal olefins and indenes w ith diazo compounds, w ith the product being hydrazone derivatives, which are shown to be readily converted to useful structures as shown in Table 2c. The authors have demonstrated that the choice of the ligand (L2-4, which were developed by the same team) is crucial for the desired transformation. They have thoroughly investigated the substrate scope for this new reaction and proposed that reaction likely proceeds through a radical (HAT) mechanism. Most importantly, for the first time such alkyl radical interception event (by diazo) is achieved in high stereoselectivity (up to 85% ee). I am enthusiastic about this paper and especially excited about the enantioselectivity demonstrated.
Catalyst controlled stereoselectivity has long been elusive in first-row transition metal catalyzed HAT radical hydrofunctionalizations and this paper represent perhaps one of the first intermolecular examples w ith Co in high ee. This particular ligand class has rarely been considered in the context of HAT catalysis and the original work show n here w ill significantly broaden the horizon of catalytic HAT reactions. It would be very interesting to know how this catalyst system is so efficient to induce an inner-sphere functionalization process of an alkyl radical species, and I am looking forward to see more work on this from this team. Overall this is a very nice piece of work w ith high originality and likely w ill be impactful to the synthetic organic community. Therefore I strongly support publication as it is. A few minor points: 1) Page 1, right column bottom, "it should be noted that…… classic alkene insertion", references are needed. 2) Scheme 3, the charges in C ("+" should be at the inner "N") and D need to be corrected.

Reviewer
#3 (Remarks to the Author): The nitrogen-containing compounds are very important and useful in many research areas or industry. Although metal-catalyzed hydroamination of alkene has been reported via HAT, the lack of efficient ligand did limit this type of transformation. Lu and co-workers reported an interesting cobalt-catalyzed radicaltype hydroamination reaction of alkenes to access the nitrogen-containing compounds w ith excellent regioselectivity and functional group tolerance. In this manuscript, the novel unsymmetric NNN tridentate ligand was designed to efficiently promote the radical hydroamination of alkenes. The scope of substrates is quite broad. The gram-scale reaction could be conducted smoothly. Due to the ease of further derivatization, various amides and bioactive molecules could be obtained efficiently. Additionally, the first intrermolecular asymmetric radical hydroamination of unactivated simple alkenes has been illustrated by using new ly developed IPAQ ligand. The enantioselectivity is up to 92.5 :7.5 er which is the best result in intrermolecular asymmetric hydroamination of aliphatic simple alkenes. All these results demonstrated the potential utility of this protocol. It may also open a new avenue for the highly enantioselective hydroamination of unactivated simple alkenes. The SI was well prepared. It should be noted that compared to the author's previous works on hydroboration and hydrosilylation of alkenes, this work may be a new direction. So this reviewer suggested that this manuscript was suitable for Nature Communications. presented. 3. In the scheme 3, the Co(III) hydride species might form to undergo HAT. The chemical valence on cobalt should be mentioned in the catalytic cycle. If the ligand w ere non-innocent, the chemical valence "n" or "n+1" could be used for cobalt. 4. On left column in page 3, the word "complex" in "be suitable for late-stage functionalization of complex molecules" should be instead by "complicated". 5. On left column in page 3, the word "alcohols" in "Alkenes, bearing phenyl, ether, tosyl, free alcohols, amine," should be "alcohol".

Answer to comment 2:
Thanks for your advice. Recent reviews and examples of metal-catalyzed relay asymmetric radical reactions have been cited in the revised manuscript, as reference 57 and 58.

Comment 3-2
To prove the utility of asymmetric transformation, more examples on asymmetric hydroamination could be presented.

Answer to the comments:
Thanks for your advice. Substrate scope on asymmetric hydroamination of alkenes has been studied and shown in Table 3 in the revised manuscript. 19 examples have been added. "The scope of substrate was quite broad as similar as that of racemic reactions.
Various functional groups, such as halide, ether, free alcohol and indole, could be tolerated."

Comment 3-3
In the scheme 3, the Co(III) hydride species might form to undergo HAT. The chemical valence on cobalt should be mentioned in the catalytic cycle. If the ligand were non-innocent, the chemical valence "n" or "n+1" could be used for cobalt.

Answer to the comments:
Thanks for your advice. Due to the possible redox non-innocent property of the ligand, the exact chemical valence has not been confirmed yet. For example, the formal chemical valence of cobalt was changed from "n" in complex A to "n-1" in complex B, however, the exact chemical valence of cobalt in complex B may be not "n-1". So the chemical valence was not described.

Comment 3-4
On left column in page 3, the word "complex" in "be suitable for late-stage functionalization of complex molecules" should be instead by "complicated".

Answer to the comments:
Thanks for your advice. The word "complex" has been changed to "complicated".

Answer to the comments:
Thanks for your advice. The word "alcohols" has been changed to "alcohol".
The references on amine and its derivatives have been modified. The author nicely explained the enantioselectivity of mechanism, and presented 19 asymmetric examples in the revised MS. Unfortunately, the ee% are not unsatisfactory. In my opinion, the enantioselectivity to be accepted in a high-impact journal should be more than 90%ee (one or two examples is enough). Otherwise, the author should disclose the story of optimization. Also, I would recommend the authors reconsider inclusion of the statement "the scope of substrate was quite broad as similar as that of racemic reactions". For examples, Table 3 does not cover many products in Table 2. I understand that any methods contain each deficiency in scope. At least, the author should clarify the limitation of this method in the Table  3  The authors have nicely addressed the review comments. I recommend publication as it is. Unfortunately, the ee% are not unsatisfactory. In my opinion, the enantioselectivity to be accepted in a high-impact journal should be more than 90%ee (one or two examples is enough).

Reviewer
Answer to Comment 1-1: Thanks for your comment. So far, we do our best to access chiral aliphatic amine derivatives with up to 92.6 : 7.4 er.
To the best of our knowledge, it is the best er achieved in metal-catalyzed hydroamination of unactivated aliphatic terminal alkenes to date.
Additionally, high enantioselective chiral amine derivatives with up to 98.6 : 1.4 er could be obtained via simple recrystallization.
Comment 1-2: Otherwise, the author should disclose the story of optimization. Comment 1-3: Also, I would recommend the authors reconsider inclusion of the statement "the scope of substrate was quite broad as similar as that of racemic reactions".
Answer to Comment 1-3: Thanks for your advices. After consideration, the text "the scope of substrate was quite broad as similar as that of racemic reactions" was changed into "the scope of substrate was quite broad". Table 3 does not cover many products in Table 2. I understand that any methods contain each deficiency in scope.

Comment 1-4: For examples,
At least, the author should clarify the limitation of this method in the Table 3 and sentence (include negative results).
Answer to Comment 1-4: Thanks for your advices. Due to time limitation, we do not cover all products in Table 2. Alkene bearing an indole moiety could undergo this transformation to afford the corresponding amide with a slight lower er. Alkenes bearing free alcohol gives corresponding product in a lower yield. These results have been added in the table 2.