Anti-Markovnikov hydro(amino)alkylation of vinylarenes via photoredox catalysis

Photoredox catalysis is a powerful means to generate odd-electron species under mild reaction conditions from a wide array of radical precursors. Herein, we present the application of this powerful catalytic manifold to address the hydroalkylation and hydroaminoalkylation of electronically diverse vinylarenes. This reaction allows for generalized alkene hydroalkylation leveraging common alkyl radical precursors, such as organotrifluoroborate salts and carboxylic acids. Furthermore, utilizing easily accessible α-silyl amine reagents or tertiary amines directly, secondary and tertiary amine moieties can be installed onto monoaryl and diaryl alkenes to access valuable products, including γ,γ-diarylamines pharmacophores. Thus, under a unified system, both hydroalkylation and hydroaminoalkylation of alkenes are achieved. The substrate scope is evaluated through 57 examples, the synthetic utility of the method is demonstrated, and preliminary mechanistic insights are presented.

In this manuscript, Hull and coworkers present a photoredox-based approach to accomplish the anti-Markovnikov hydroalkylation of vinylarenes. This is clearly a valuable reaction-the authors do an excellent job contextualizing their advance relative to prior work. While there are other ways to make these products, this new method is clearly complementary. I particularly appreciate that a range of radical precursors (BF3K salts, carboxylic acids, alpha-amino C(sp2)-H bonds) each participate readily in this transformation. Additionally, a range of styrene derivatives (EWG, EDG, alpha-and beta-substituted, etc) readily engage in this reaction. However, I do note that many of the examples where the radical component is varied employ 1,1-diphenylethylene as a radical trap. This is clearly a much better vinylarene radical trap than most vinylarenes and, accordingly, they are much easier functionalize using this new method (for example, compare entry 26 to 41 in both reaction time and yield). If any of the alpha-amino radical precursors were tested with styrene or a simpler vinyl arenes but not presented (due to low yield), I'd appreciate to seeing these entries included in the SI. Generally, however, the SI was thorough and contained sufficient experimental detail to reproduce these experiments and verify the identity and purity of the compounds produced.
I found the mechanistic experiments reasonably convincing that a electron transfer/proton transfer (ET/PT) mechanism is involved-specifically the d7-DMF and MeOD experiments. The interpretation of the acid/base additive experiment was not clear to me as the yield change upon addition of AcOH was relatively minor and it was not clear that the carbonate base should be meaningfully deprotonating MeOH under these conditions and it seems that it could interfere with the reaction through a variety of alternative mechanisms. The deuteration experiments more clearly rule out a simple HAT mechanism under these conditions. While the authors show these experiments lead to similar conclusions for the carboxylic acid precursors, I was hoping they could shed some light on the RBF3K salts given that under those conditions they optimized to include a phenol additive that was rationalized as an HAT catalyst. Can this simply be replaced with a carboxylic acid? Furthermore, the authors note at the outset that reduction of the benzylic radical is thermodynamically demanding. Can the origin of the electron be more clearly identified for the proposed ET/PT mechanism? Finally, I was left curious whether the authors saw any deuteration of the amine fragment in the MeOD experiment (although I do not feel this is a crucial point for the authors to address).
I have one minor concern: the redox auxiliary approach employed to render the hydroaminoalkylation more general was well-described from a practitioner perspective, however, this TMS redox-auxiliary approach was extensively developed by the electrochemistry community and is summarized well in this review: Chem. Rev. 2008, 108, 2265-2299. I'd like to see this review, or the seminal papers in this area, cited in addition to the more recent work exploiting this strategy in photoredox catalysis (such as Ref 57).
Overall, I find this manuscript likely to be interesting to the broad readership of Nature Communications and recommend this manuscript for publication with minor revisions according to the concerns outlined above.
Reviewer #2 (Remarks to the Author): The authors describe a Photoredox-catalyzed Hydro(amino)alkylation of styrene derivatives from either carboxylic acids, BF3K salts or aminosilanes. The general premise behind this work is that a carbon-based radical is generated from one of the three aforementioned radical precursors using a photoredox catalyst that undergoes a Giese addition in anti-maokovnikov fashion followed by a hydrogen atom transfer to give the final adducts. A range of hydro(amino)alkylated styrenes is synthesized via this reaction and the mechanism is mostly plausible (though H-atom transfer is the most likely final step of the catalytic cycle rather than reduction/protonation).
As for the innovation, there really is none to speak of in this submission. Carboxylic acids, BF3K salts and aminosilanes have been utilized as radical precursors by multiple groups including MacMillan, Yoon, Nicewicz, Molander and others in many types of radical reactions including Giese additions to Michael acceptors and in some cases, with absolute stereocontrol. The omission of the large body of this work in the discussion and the references section is both egregious and disingenuous. The only new work here is the use of styrenes, which is viewed as incremental advance at best. The products of this transformation are mildly interesting at best and there are certainly far better disconnections to make these adducts from alternative starting materials. This work is better suited for a more specialized organic chemistry journal such as J. Org. Chem. or Eur. J. Org. Chem.
Reviewer #3 (Remarks to the Author): Visible light-driven photocatalytic alkene functionalization is an area of intense research in organic chemistry. In this manuscript, Hull and co-workers take advantage of potassium trifluoroborate salts, carboxylic acids and tertiary amines as alkyl radical precursors to accomplish anti-Markovnikov hydro(amino)alkylation reaction through visible light-driven photoredox catalysis. Under the optimized conditions, a wide range of electronically diverse vinylarenes are well tolerated. This protocol could also be extended to alpha-TMS tertiary amines. Successful application of this methodology to several natural product derivatives also highlights its synthetic utility. The corresponding anti-markovnikov hydro(amino)alkylation products are obtained with moderate to good yields. Key to the success of this protocol is the photocatalytic regioselective addition of carbon radical to vinylarenes and identification of efficient HAT reagents. A series of control experiments also support the mechanistic hypothesis. Considering readily available radical precursors, practical conditions, and significance of the products, I think this protocol should be of great interest to the synthetic community.
Moreover, the manuscript is well written, and the supporting information is thorough and provides all the expected data. In conclusion, this is a nice paper and can be accepted in Nature Communication. I do not have any technical issues as the work is scientifically sound.
Minor points: 1) The R group of product 5 should be indicted in Table 2 We appreciate the kind comment of referee 1 on our work: "This is clearly a valuable reaction-the authors do an excellent job contextualizing their advance relative to prior work. While there are other ways to make these products, this new method is clearly complementary." and "Overall, I find this manuscript likely to be interesting to the broad readership of Nature Communications and recommend this manuscript for publication with minor revisions according to the concerns outlined above." (a) Original comments: "However, I do note that many of the examples where the radical component is varied employ 1,1-diphenylethylene as a radical trap. This is clearly a much better vinylarene radical trap than most vinylarenes and, accordingly, they are much easier functionalize using this new method (for example, compare entry 26 to 41 in both reaction time and yield). If any of the alpha-amino radical precursors were tested with styrene or a simpler vinyl arenes but not presented (due to low yield), I'd appreciate to seeing these entries included in the SI." We agree with the reviewer that 1,1-diaryl ethylene is a better carbon-centered radical trapper than simple styrenes. Indeed, as the reviewer mentioned, by comparing the yield and reaction time, the reactivity trend appears to be: 1,1diaryl ethylene ~ electron-deficient styrene > electron-neutral styrene > electronrich styrene. This is consistent with the polarity-matched radical addition principle. Moreover, electron-withdrawing groups further increase the reduction potential of the resulting benzylic radical, which facilitates the single electron transfer from the reduced ground state photocatalyst (Ir II in Scheme 3) to form the benzylic anion.
We have also included two unsuccessful substrates in Supplementary Table 19. We agree with reviewer I on this point. There are a range of mechanisms and effects by which these additives could change reaction performance, and they would be difficult to precisely pinpoint. Additionally, photoredox reactions are more susceptible to perturbations in photon flux, which is a further confounding variable when considering the reaction with Cs2CO3. Given the lack of clarity in these studies, we think this result is best removed from the manuscript. We note that this does not change the overall mechanistic framework of the work.

(c) Original comments: "While the authors show these experiments lead to similar conclusions for the carboxylic acid precursors, I was hoping they could shed some light on the RBF3K salts given that under those conditions they optimized to include a phenol additive that was rationalized as an HAT catalyst. Can this simply be replaced with a carboxylic acid? "
We thank reviewer I for this kind suggestion. Indeed, we did try acetic acid as the H-source, and the desired product 1 was obtained in 15 % yield. A new entry now is added to the Table 1 to provide this insight to the prospective reader.
Based on our mechanistic studies, we believe the ET/PT mechanism is more likely when carboxylic acids and alpha-TMS amines are used with diaryl alkenes or electron deficient monoaryl alkenes. As for RBF3K salts, as electron rich phenols were the optimal H-source, a HAT mechanism might well be occurring in this case.
(d) Original comments: "Furthermore, the authors note at the outset that reduction of the benzylic radical is thermodynamically demanding. Can the origin of the electron be more clearly identified for the proposed ET/PT mechanism? " This is an excellent point, and one we have pondered over for quite some time. Indeed, reviewer I is correct that reduction of the benzylic radical is thermodynamically challenging. To put this in perspective, the reduction potential for the styrene-derived radical (E1/2 = -1.60 V vs. SCE) is well out of range of the reduced photocatalyst (E1/2 = -1.37 V vs. SCE). This rules out the sequential ET/PT mechanism for the electron-neutral and -rich monoaryl alkene substrates presented in the manuscript. On the other hand, a direct H-atom abstraction from the MeOH donor is also thermodynamically unfeasible based on bond-dissociation energy considerations. The BDE for a benzylic C-H bond is around 85 kcal/mol and the BDE for MeOH is around 107 kcal/mol. To reconcile this, we favor a PCET mechanism in these cases where the thermodynamics are mismatched. The combination of MeOH (pKa = 29) and the Ir photocatalyst (E1/2 -1.58 vs. Fc) furnishes a BDFE of 58 kcal/mol, which is much lower than that of the benzylic C-H (85 kcal/mol), thereby rendering the step thermodynamically feasible.
However, this mechanism is not a catch-all for every substrate we have presented. The mechanism for the analogous diarylalkene substrates presented could very well be different. The reduction potentials here are E ≈ -1.34 V vs. SCE. This is now within the range of the photocatalyst. Therefore, a sequential ET/PT mechanism could be operative for this class of substrates.
To address this reviewer comment, we have added a brief discussion near the end of the mechanism section of the manuscript.

(e) Original comments: "Finally, I was left curious whether the authors saw any deuteration of the amine fragment in the MeOD experiment (although I do not feel this is a crucial point for the authors to address)"
We thank reviewer I for this question. Indeed, we didn't observe deuterium incorporation into the amine fragment in the MeOD experiment, as indicated by both 1H and 2H NMR of the product. These studies and spectrums were included in the "Mechanistic Studies" section of the SI.
1 H NMR of product 2 H NMR of product (f) Original comments: "I have one minor concern: the redox auxiliary approach employed to render the hydroaminoalkylation more general was well-described from a practitioner perspective, however, this TMS redox-auxiliary approach was extensively developed by the electrochemistry community and is summarized well in this review: Chem. Rev. 2008, 108, 2265-2299 We thank reviewer I for this kind suggestion and we completely agree that these early studies should be mentioned in the manuscript. We've added following sentence in the TMS-amine section: "This redox auxiliary approach has been extensively studied in modern electroorganic synthesis. 61 " where this Chem Rev paper was cited as ref 61.

II.
Re: referee 2 (a) Original comments: "A range of hydro(amino)alkylated styrenes is synthesized via this reaction and the mechanism is mostly plausible (though H-atom transfer is the most likely final step of the catalytic cycle rather than reduction/protonation)" For hydroalkylation with carboxylic acid and hydroaminoalkylation with alpha-TMS amines, we believe that the reduction/protonation mechanism is more likely based on the deuterium studies in Scheme 5 and SI (page S36). Moreover, hydrogen atom abstraction of either carboxylic acid or MeOH by the resulting benzylic radical from is thermodynamically unfavored.
As for RBF3K salt, a HAT mechanism might occur in this case as phenols are good hydrogen atom donor.
(b) Original comments: "Carboxylic acids, BF3K salts and aminosilanes have been utilized as radical precursors by multiple groups including MacMillan, Yoon, Nicewicz, Molander and others in many types of radical reactions including Giese additions to Michael acceptors and in some cases, with absolute stereocontrol. The omission of the large body of this work in the discussion and the references section is both egregious and disingenuous. The only new work here is the use of styrenes, which is viewed as incremental advance at best."