Electrochemical C–H phosphorylation of arenes in continuous flow suitable for late-stage functionalization

The development of efficient and sustainable methods for carbon-phosphorus bond formation is of great importance due to the wide application of organophosphorus compounds in chemistry, material sciences and biology. Previous C–H phosphorylation reactions under nonelectrochemical or electrochemical conditions require directing groups, transition metal catalysts, or chemical oxidants and suffer from limited scope. Herein we disclose a catalyst- and external oxidant-free, electrochemical C–H phosphorylation reaction of arenes in continuous flow for the synthesis of aryl phosphorus compounds. The C–P bond is formed through the reaction of arenes with anodically generated P-radical cations, a class of reactive intermediates remained unexplored for synthesis despite intensive studies of P-radicals. The high reactivity of the P-radical cations coupled with the mild conditions of the electrosynthesis ensures not only efficient reactions of arenes of diverse electronic properties but also selective late-stage functionalization of complex natural products and bioactive compounds. The synthetic utility of the electrochemical method is further demonstrated by the continuous production of 55.0 grams of one of the phosphonate products.


REVIEWER COMMENTS
Reviewer #1 (Remarks to the Author): In this manuscript Hao Long et al report the electrochemical C-H phosphorylation in flow using phosphite precursors. With this protocol a broad variety of substrates were able to be transformed into the desired aryl phosphonate derivates. Hao Long and co-workers were capable to address electron-poor substrates with high functional group tolerance such as esters, cyano, amides, halides, etc, and even late-stage functionalization was feasible. The approach on electron-poor substrates was a big challenge and has been demonstrated successfully. The electrochemical conversion in flow has been laid out properly and decent considerations have been made regarding the reaction mechanism and the reactivity of the phosphite precursor. Still some issues need to be stressed a bit. Introduction: 1) First sentence: I would like to see more common organic chemical applications of aryl phosphonate compounds (2-3 sentences). You just have one small sentence. Which feels like you did not pay that much attention on the application. So, a bit more detail is highly desired. 2) You are saying in the introduction [..] arene or the phosphorus nucleophile […] The term nucleophile is misleading. Because its resulting electron deficient intermediate will be most likely electrophilic. Just rephrase that 3) Later you are referring to reaction published where they used electron deficient arenes and they needed to use three electrode arrangement. In one of your Ref. they were using Co-cat in both compartments which means it is stable under oxidative and reductive conditions and no deposition can be seen. Please rethink that. Results and discussion: 4) Can you explain this counterintuitive behaviour of the equivalents? 7eq or 3 eq does not work anymore. Have you ever tried to switch the excess components? Try to use the arene as excess since you proposed the phosphite is the one which is oxidized. An excess of trapping agent might be reasonable. 5) Can you explain the role of HBF4 except for being a proton source for the hydrogen evolution at the platinum cathode? Did you use any additional supporting electrolyte such as Tetraalkylammonium salts? Tetrafluoroboric acid tends to be easily solvated which mean under these conditions it might also formes acid-base-salts which helps in terms of conductivity. Can you give any information on the cell potential of the reaction with TFA, AcOH, Sc(OTf)3 and HBF4? 6) Table 1: Have you ever tried TfOH. Which might be interfering with the phosphite and function as unproductive side reaction. But still interesting Mechanistic studies: 7) Have you ever measured CV with the whole electrolyte system? HBF4 and with or without water? 8) Conditions I: Have you seen the formation of the alkyl phosphonate instead? Which might occur under non-aqueous conditions with an acid present 9) […] Since trialkyl phosphites underwent rapid acid-promoted hydrolysis to dialkyl phosphites, […]. Isn't it rather a H-phosphonate than a dialkyl phosphite? 10) Regarding the hydrolysis of the phosphite. The oxygen will not be incorporated at all. Hence, this result was expected. 11) Have you quantify production of H-phosphonate with standard conditions? Might be interesting if its related to the amount of water? Because with conditions II you have been using the H-phosphonate in the same quantities like the amount of water with the standard conditions. 12) My suggestions: a. Change the excess component 1 eq phosphite and 5 eq of arene b. also try to use TfOH. Just as control reaction. c. The role of the H-phosphonate is intriguing. Have you ever tried using catalytical amount of it? Would be very interesting to understand the role of this intermediate. Have you tried to monitor the process on-line by any means of NMR, IR. 13) Later you discuss the loss of the alkyl group by any nucleophilic species. In the Ref. 66 they mentioned that phosphonium salt is attacked either from water to produce MeOH or by the acid salt to produce the methyl ester. Might be an option here as well. Please discuss this in more detail In this manuscript, Xu and coworkers have developed a method for electrochemical C-H phosphonylation of arenes via the generation of a phosphorous radical cation. In this reaction, a trialkyl phosphite is electrochemically oxidized to generate the phosphorous radical cation which can react with the arene. Upon loss of an electron and a proton, dealkylation of the phosphonium occurs affording the final phosphonate product. This platform offers several advantages over previous C-H phosphonylation strategies: 1. Transition-metal catalyzed approaches require pre-installation of a directing group, limiting the arene scope of the transformation. 2. Undirected approaches usually proceed via oxidation of the arene substrate or the phosphorus nucleophile. However, the scopes of such processes are limited to electron-rich arenes due to the high oxidation potentials of the arene substrates or the reduced nucleophilicity of neutral phosphorus radicals. Xu and coworkers circumvented these challenges through the direct oxidation of trialkyl phosphites, reporting the seminal example of engaging phosphorus radical cations in C-H phosphonylation. The authors demonstrate that this method is synthetically useful through a broad scope, late-stage functionalization, product derivatization, and large scale synthesis.
Given the utility of these products and the clear advantages of this new protocol relative to state-of-the-art approaches, this manuscript will likely be of interest to the readership of Nature Communications and I recommend publication with minor revisions (vide infra).
In order to further improve this already strong manuscript, I recommend the authors consider the following suggestions that could improve this already strong manuscript: 1) The authors state that since there was no 18O incorporation in the product that HPO(OR)2 is critical for the reaction and seem to indicate that water's role is forming this species. However, based on the NMR spectrum provided in FIgure 4b, HPO(OR)2 is formed in the absence of water. The authors should clarify this point and the role of water in the formation of HPO(OR)2 as much as possible. These are interesting observations but currently almost completely unexplained in the manuscript. Have the authors examined whether trialkylphosphates could similarly promote the reaction?
2) The authors state that this reaction proceeds without the need for an oxidant. However, electrons must always be accepted in some way, in this case likely through cathodic proton reduction. The wording should be altered to state the transformation does not necessitate the use of a conventional chemical oxidant.
3) The authors screen several acid additives in the reaction. Is there a rationalization for why HBF4 is the best acid for this transformation?
4) The clarity of Figure 4e would be improved by inclusion of the % 18O incorporation results.
5) The authors suggest that the increased yield observed in the scaleup reaction is probably due to the in situ mixing of HBF4 and P(OEt)3. Have the authors tested this in their standard conditions?
6) The manuscript states that the reaction performed is phosphorylation (C-O bond formation). However, since a C-P bond is formed, phosphonylation would be a more precise term.
Reviewer #3 (Remarks to the Author): In this contribution, Xu and coworkers present a flow electrochemical procedure for the C-H phosphorylation of arenes. The authors claim that this is a novel transformation, applicable to electron-deficient arenes.
Unfortunately, it appears that the authors have not carried out a literature background check. There are many precedents on this type of electrochemical phosphorylation on electron poor aromatics. For example: 10.108010. /10426507.201810. .1540488, 10.101610. /j.cattod.201610. .06.001, 10.100710. /BF00953100, 10.108010. /10426507.201810. .1541897, 10.108010. /10426507.201610. .1212051, 10.108010. /10426507.2018. The chemistry presented is therefore not as novel as the authors suggest and, in opinion of this referee, it should not be accepted for publication in a top journal such as Nature Communications. It should be submitted to a more specialized journal. There are some important additional issues listed below: -The authors must cite all the literature mentioned above. -Since the chemistry is known, the only novelty of this work is the adaptation of the electrolysis to continuous flow conditions. However, the development of the required conditions are not shared and are simply stated. It would be beneficial to the reader to understand the optimization process of these electrochemical parameters.
We thank the reviewers for taking their valuable time to check the manuscript and for giving constructive suggestions to improve it. We have worked tirelessly to address every comment of the reviewers. A point-to-point response to the comments is as following. The original reviewer comments are in blue.

Reviewer #1 (Remarks to the Author):
In this manuscript Hao Long et al report the electrochemical C-H phosphorylation in flow using phosphite precursors. With this protocol a broad variety of substrates were able to be transformed into the desired aryl phosphonate derivates. Hao Long and co-workers were capable to address electron-poor substrates with high functional group tolerance such as esters, cyano, amides, halides, etc, and even late-stage functionalization was feasible. The approach on electron-poor substrates was a big challenge and has been demonstrated successfully. The electrochemical conversion in flow has been laid out properly and decent considerations have been made regarding the reaction mechanism and the reactivity of the phosphite precursor. Still some issues need to be stressed a bit.

Response:
We thank the reviewer for the positive comments.

Introduction:
1) First sentence: I would like to see more common organic chemical applications of aryl phosphonate compounds (2-3 sentences). You just have one small sentence. Which feels like you did not pay that much attention on the application. So, a bit more detail is highly desired.

Response:
We have added more details and the first sentences now reads as the following. We would be happy to revise further if the review has additional suggestions on the wordings.
Aryl phosphorus compounds have wide applications in medicinal chemistry, 1 material science, 2 and catalysis as ligands and Lewis acid catalysts. 3,4 In addition, brigatinib, which contains a phenylphosphine oxide motif and serves as an anaplastic lymphoma kinase (ALK) inhibitor, has been achieved commercial success for treating metastatic non-small-cell lung cancer (NSCLC). 5 2) You are saying in the introduction [..] arene or the phosphorus nucleophile […] The term nucleophile is misleading. Because its resulting electron deficient intermediate will be most likely electrophilic. Just rephrase that Response: The sentence has been revised to the following: In these reactions, either the arene or the phosphorus reagent is oxidized under… 3) Later you are referring to reaction published where they used electron deficient arenes and they needed to use three electrode arrangement. In one of your Ref. they were using Co-cat in both compartments which means it is stable under oxidative and reductive conditions and no deposition can be seen. Please rethink that.

Response:
We have revised the sentence (see below). It is interesting that the dehydrogenative cross coupling can be achieved in both anodic and cathodic compartment. It is unclear how that works and why a divided cell is necessary if both electrodes can be used to promote product formation. More discussions on the refs of these authors are provided in responding the comments of reviewer 3.
Reactions of electron deficient arenes requires metal catalysts and impractical three electrode configuration in divided cells. 42,43 Results and discussion: 4) Can you explain this counterintuitive behaviour of the equivalents? 7eq or 3 eq does not work anymore. Have you ever tried to switch the excess components? Try to use the arene as excess since you proposed the phosphite is the one which is oxidized. An excess of trapping agent might be reasonable.
Response: For these reactions, an optimal concentration of P(OEt)3 is important for optimal results. With 3 equiv of P(OEt)3, a large portion of P(OEt)3 is hydrolyzed to HPO(OEt)2 because 2 equiv of H2O is added to the system. In addition, some P(OEt)3 is protonated by the strong acid HBF4 [pKa in H2O: HP(OMe)3 + , 2.6; HBF4, −0.3]. As a result, the amount of P(OEt)3 available is small leading to reaction failure. At higher concentration of P(OEt)3 such as 7 equiv, our hypothesis is that P(OEt)3 can now compete with the arene to react with the radical cation [(EtO)3P •+ ], leading to decomposition of P(OEt)3 and recovery of most arene starting material. With 7 equiv of P(OEt)3, we have observed the formation of EtPO(OEt)2, which likely arises from the reaction of P(OEt)3 with [(EtO)3P •+ ].
EtPO(OEt)2 is not observed under the standard conditions. We agree that a higher concentration (excess) of arene is helpful for the arene to trap the P-radical cation [(EtO)3P •+ ]. But the arene is usually the more valuable component and it is better to use an excess of P-reagent for practical applications. We have conducted experiments with 5 equiv of arene and 1 equiv of P(OEt)3 with or withour H2O (2 equiv).
The reaction failed in the presence of H2O because P(OEt)3 is hydrolyzed to HPO(OEt)2.
The reaction without H2O provided 2 in 10% yield. These results under anhydrous conditions are consistent with early reports by A. N. Pudovik et al. (Russ. Chem. Bull. 1983, 32, 566;ref 44). Pudovik and coworkers studied the electrochemical reactions of PhH, PhEt, and PhMe with P(OEt)3 under anhydrous conditions in a divided cell. They started that "as large an access as possible ArH" was needed to ensure good results (no details on the exact amount). Obviously, 5 equiv of compound 1 is not enough to obtain good yield for 2. Hence, the use of excess of arene is not a practical solution. 5) Can you explain the role of HBF4 except for being a proton source for the hydrogen evolution at the platinum cathode? Did you use any additional supporting electrolyte such as Tetraalkylammonium salts? Tetrafluoroboric acid tends to be easily solvated which mean under these conditions it might also formes acid-base-salts which helps in terms of conductivity. Can you give any information on the cell potential of the reaction with TFA, AcOH, Sc(OTf)3 and HBF4?
Response: Role of HBF4. In addition to helping hydrogen evolution, HBF4 facilitates the hydrolysis of P(OR)3 to produce HPO(OEt)2, which is critical for the success of the electrochemical reaction. Without acid, the electron deficient cationic intermediate [ArP(OR)3 + ] or the product ArPO(OEt)2 can undergo reductive side reactions. The addition of 2 equiv of HBF4 indeed reduces the cell potential from > 50 V (without HBF4) to 2.9 V (standard conditions). So HBF4 is also helpful to increase conductivity. For the reaction of compound 1, replacing HBF4 with nBu4NBF4 (1 equiv) (cell potential = 5.4 V) leads to no formation of 2. Hence, the role of HBF4 is not just to increase conductivity. In the manuscript, we have stated the following.
In addition to promote hydrolysis of P(OEt)3, the acidic additive HBF4 serves as the supporting electrolyte and is also helpful for H2 evolution and avoiding unwanted cathodic reduction of electron-deficient species such as 60 and 61.
The cell potentials of the reaction with 2 equiv of TFA, AcOH, TfOH, Sc(OTf)3, HBF4 are 5.8 V, (increase over time), 2.5 V, 23 V, 2.9 V, respectively. Mechanistic studies: 7) Have you ever measured CV with the whole electrolyte system? HBF4 and with or without water?

Response:
We have added to the Supplementary Information the CV of the whole system consisting of compound 1 (1 equiv), P(OEt)3 (5 equiv), HBF4•Et2O (2 equiv), and H2O (2 equiv). A copy is displaced below (the blue trace in the figure below). The two oxidation peaks can be assigned to P(OEt)3 and compound 1 and HBF4•Et2O. The CV of HBF4•Et2O with or without water (1 equiv) has also been included in the Supplementary Information (also see below). We have change dialkyl phosphites to H-phosphonates to avoid any confusion.
10) Regarding the hydrolysis of the phosphite. The oxygen will not be incorporated at all.
Hence, this result was expected.
Response: We agree. The 18 O experiment also confirms that the PO(OR)2 moiety of the Ar-PO(OR)2 product did not originate from HPO(OR)2.

11) Have you quantify production of H-phosphonate with standard conditions? Might be
interesting if its related to the amount of water? Because with conditions II you have been using the H-phosphonate in the same quantities like the amount of water with the standard conditions.

Response:
We have used an internal standard (added after the electrolysis) to quantify the amount of H-phosphonate under the stand conditions. P(OPh)3 was added to the eluent of the cell and 31 P NMR was taken directly without workup. The results suggest that the yield for HPO(OEt)2 is about 62% from P(OEt)3. By following your suggestions, we have studied the reaction in the presence of different amount of HPO(OEt)2 (vide infra).

12) My suggestions:
a. Change the excess component 1 eq phosphite and 5 eq of arene

Response:
We have done this experiment. Please see response to number 4) above.
b. also try to use TfOH. Just as control reaction.

Response:
We have done this experiment. Please see response to number 6) above.  Determined by 1 H-NMR analysis using 1,3,5-trimethoxybenzene as the internal standard.
Unreacted 1 was shown in brackets. c Isolated yield.
The results on the catalytic amount of HPO(OEt)2 prompted us to apply these conditions to the difficult synthesis of 4 and to some of the substrates that are failed under the standard conditions (see below). But these conditions with catalytic HPO(OEt)2 did not afford better results than our standard conditions for the difficult substrates. It is also unclear if these conditions with catalytic HPO(OEt)2 are general and robust. But the results with compound 1 do show that it is possible to use less phosphorus reagents, e.g. 3 equiv of P(OEt)3 and 0.2 equiv of HPO(OEt)2. These conditions will be applied to our ongoing investigation on the synthetic applications of P-radical cations.
We did not monitor the process on-line. 13) Later you discuss the loss of the alkyl group by any nucleophilic species. In the Ref.
66 they mentioned that phosphonium salt is attacked either from water to produce MeOH or by the acid salt to produce the methyl ester. Might be an option here as well. Please discuss this in more detail

Response:
We agree that nucleophilic species such as H2O or alcohol should be the responsible for accepting the alkyl group. We have conducted additional experiments to address this point by detection of Bu2O formation during the synthesis of 34. The use of HBF4•Et2O as an additive prevented us to study with P(OEt)3. Hence, we studied the reaction of P(OBu)3. The nBu group may be picked up by H2O or nBuOH (generated from P(OBu)3 and H2O) to produce nBuOH and nBu2O respectively. We have observed formation of nBu2O by GC analysis of the reaction mixture, suggesting nBuOH is at least one of the species to accept the alkyl group. This experiment has been included in the Supplementary Information. We added the following discussion to the manuscript: …which loses an alkyl group to a nucleophilic species in the reaction mixture such as H2O or alcohol produced from hydrolysis of P(OR)3 or during the workup to afford the final phosphonate product 61. Given the utility of these products and the clear advantages of this new protocol relative to state-of-the-art approaches, this manuscript will likely be of interest to the readership of Nature Communications and I recommend publication with minor revisions (vide infra).

Response:
We thank the reviewer for the positive recommendation.
In order to further improve this already strong manuscript, I recommend the authors consider the following suggestions that could improve this already strong manuscript: 1) The authors state that since there was no 18O incorporation in the product that HPO(OR)2 is critical for the reaction and seem to indicate that water's role is forming this species. However, based on the NMR spectrum provided in 2) The authors state that this reaction proceeds without the need for an oxidant. However, electrons must always be accepted in some way, in this case likely through cathodic proton reduction. The wording should be altered to state the transformation does not necessitate the use of a conventional chemical oxidant.

Response:
We have changed "chemical oxidants" to "conventional chemical oxidants" and "oxidant-free" to "external oxidant-free".
3) The authors screen several acid additives in the reaction. Is there a rationalization for why HBF4 is the best acid for this transformation?
Response: HBF4, which contains a non-nucleophilic anion and is a strong acid, serves as electrolyte and promotes H2 evolution and the hydrolysis of some of P(OEt)3 to HPO(OEt)2.
Acids such as TFA and AcOH with nucleophilic anions can compete with the arene to react with the radical cation derived from P(OEt)3. As a result, no product 2 was formed with TFA and AcOH. TfOH is known to cause the decomposition of P ( 6) The manuscript states that the reaction performed is phosphorylation (C-O bond formation). However, since a C-P bond is formed, phosphonylation would be a more precise term.

Response:
We have changed phosphorylation to phosphonylation. We agree that phosphonylation clearly describes the reaction since the products are phosphonates.
Phosphorylation refers to the attachment of a phosphoryl group (PO3 2-) to a molecular (e.g.

Reviewer #3 (Remarks to the Author):
In this contribution, Xu and coworkers present a flow electrochemical procedure for the C- The chemistry presented is therefore not as novel as the authors suggest and, in opinion of this referee, it should not be accepted for publication in a top journal such as Nature Communications. It should be submitted to a more specialized journal.

Response:
We thank the reviewers for providing these references, which are summarized in Table R1 as refs I to VI (see below). We believe without doubt that these previous

works do not reduce the novelty of our work. On the contrary, they highlight the advantage of our method and increase the novelty of our work.
Two of these references (refs I and II) have already been cited in our original submission.
We 1. All the 28 references cited in the paper are self-citations. It is rather unusual for a paper published in 2019 to cite only their own work.
2. There is a lack of details on the reactions. a) No information on the scale of the reactions is given.
b) The reactions are conducted using a three-electrode system, but no information is given on the potentials applied.
c) The yields are referred to those of the main product as suggested by   Budnikova et al., Phosphorus, Sulfur, and Silicon and the Related Elements 2019, 194, 415, DOI:10.1080/10426507.2018 Review on the authors' own work.
There are some important additional issues listed below: -The authors must cite all the literature mentioned above.

Response:
The references have been cited as 42-47.
-Since the chemistry is known, the only novelty of this work is the adaptation of the electrolysis to continuous flow conditions. However, the development of the required conditions are not shared and are simply stated. It would be beneficial to the reader to understand the optimization process of these electrochemical parameters.
Response: Our work is not just simple adaption of the electrolysis in continuous flow as discussed above and represents significant advances in arene C-H phosphorylation.
Divided cells, constant potential conditions, and metal catalysts are used because simple conditions involving undivided cells, constant current and metal-free do not work, not because people prefer the complicated setup and inconvenient conditions. We make it work through innovation on technology, mechanism, and reaction conditions. Simply moving the reactions in the references listed by the reviewer to flow will not work. The reactions of Budnikova and coworkers proceed through addition of P-radicals to the arenes. To achieve efficient reactions in undivided cells without metal catalysts, we resort to a different mechanism involving reactions of P-radical cations with arenes. Ref III follows a similar mechanism but required a divided cell and "as large an access as possible Mechanistic Studies: "The HPO(OR)2 formed in situ through the hydrolysis of P(OR)3 likely forms reversibly with radical cation 59 an adduct 62, which reduces the decomposition of 59 and buys more time for its reaction with the arene." -Isn't it rather 57 than 59? -It is reasonable to state that this intermediate 62 will act as reservoir.
Reviewer #2 (Remarks to the Author): I am completely satisfied by the revisions to this manuscript. I think the authors did an excellent job responding to the comments of all 3 reviewers. I found the arguments, text clarifications, and new data compelling and recommend publication without further change.
A point-to-point response to the comments of reviewers is as following. The original reviewer comments are in black, and our responses are in blue. Maybe it is related to the double layer polarity which is different if you compare the nbutylammonium salt with HBF4 close to the positive polarized cathode.

Response:
We thank the reviewer for this question. At the Pt cathode, the most easily reduced species accepts electrons from the electrode. With an acid present, the most easily reduced species is proton. Hence protons undergo reduction to H2 when HBF4 is included as an additive. However, without an added acid, the proton concentration in the reaction mixture is low, and other electron-deficient species such as [ArP(OR)3] + or the product ArP(OEt)2 become the most easily reduced species and may undergo to reduction at the cathode leading to their decomposition. We have modified the discussion on the role of acid in the text to the following.
Under these acidic conditions, protons, which are the mostly easily reduced species in the reaction mixture, accept electrons at the Pt cathode to generate H2. In addition to promote hydrolysis of P(OR)3, the acidic additive HBF4 serves as the supporting electrolyte and a proton source for H2 evolution, avoiding unwanted cathodic reduction of electron-deficient species such as 60 and 61.
Mechanistic Studies: "The HPO(OR)2 formed in situ through the hydrolysis of P(OR)3 likely forms reversibly with radical cation 59 an adduct 62, which reduces the decomposition of 59 and buys more time for its reaction with the arene." -Isn't it rather 57 than 59?
Response: We thank the reviewer for pointing out the typo. It is indeed 57. We have corrected it to show as 57 (see below).
The HPO(OR)2 formed in situ through the hydrolysis of P(OR)3 likely forms reversibly with radical cation 57 an adduct 62, which reduces the decomposition of 57 and buys more time for its reaction with the arene.
-It is reasonable to state that this intermediate 62 will act as reservoir.

Response:
We thank the reviewer for agreeing on our proposal.

Reviewer #2 (Remarks to the Author):
I am completely satisfied by the revisions to this manuscript. I think the authors did an excellent job responding to the comments of all 3 reviewers. I found the arguments, text clarifications, and new data compelling and recommend publication without further change.
Response: We thank the reviewer for taking time to review the manuscript again and for the positive recommendation.