Electrochemical Nickel Catalysis for Sp2-Sp3 ... - ACS Publications

---

Letter pubs.acs.org/OrgLett

Electrochemical Nickel Catalysis for Sp2‑Sp3 Cross-Electrophile Coupling Reactions of Unactivated Alkyl Halides Robert J. Perkins, Dylan J. Pedro, and Eric C. Hansen* Chemical Research and Development, Pfizer, Inc., Eastern Point Road, Groton, Connecticut 06340, United States S Supporting Information *

ABSTRACT: A constant-current electrochemical method for reducing catalytic nickel complexes in sp2-sp3 cross-electrophile coupling reactions has been developed. The electrochemical reduction provides reliable nickel catalyst activation and turnover and offers a tunable parameter for reaction optimization, in contrast to more standard activated metal powder reductants. The electrochemical reactions give yields (i.e., 51−86%) and selectivities as high or superior to those using metal powder reductants and provide access to a wider substrate scope.

N

reaction additive1a and as the electrolyte (Table 1). A reticulated vitreous carbon (RVC) cathode was paired with a sacrificial metal rod anode9 in an undivided cell. RVC was chosen as the cathode material both for its high surface area for reduction and for its chemical inertness. Screening several readily available metal rods for the sacrificial anode showed a significant effect on the reaction efficiency (Table 1, Entries 1−4). Both magnesium and aluminum anodes led to a large amount of metal halide precipitate in the reaction, with the magnesium anode yielding no productive reaction (Table 1, Entry 1) and the aluminum anode giving a low yield of 8%, poor selectivity favoring hydrodehalogenation, and incomplete conversion even after an excess of charge (4.0 F/mol) had been passed (Table 1, Entry 2). Zinc and iron anodes, however, produced significant amounts of cross-coupled product with low amounts of hydrodehalogenation and aryl−aryl homocoupling side reactions, giving cross-coupled yields of 91% (Table 1, Entry 3) and 80% (Table 1, Entry 4), respectively. Noticeably, while the cross-coupling reaction fails to initiate with iron powder as the reductant, it proceeds well in the presence of the iron anode, indicating that it is indeed a cathodic reduction of the nickel catalyst rather than a reduction by the anode metal surface. Furthermore, no reaction occurs with either anode in solution in the absence of current. These observations are consistent with previous reports that a metal rod anode makes a minimal contribution to catalyst reduction.6a Though good yields were obtained at room temperature, the current efficiency of the reaction was not ideal, as it required twice the theoretical amount of charge (4.0 F/mol) to be passed to achieve full conversion. Heating the reaction to 65 °C to increase the rate of cross-coupling relative to the currentcontrolled Ni(II) reduction led to much better current efficiencies, with the reaction using a Zn anode completing in

ickel-catalyzed cross-electrophile couplings have undergone substantial development in recent years, having been shown to be versatile reactions for the formation of sp2sp3 bonds without the need to preform an organometallic reagent.1 Despite this development, several limitations still remain. Selectivity for the cross coupled product over homocoupled and hydrodehalogenated byproducts is sometimes poor and not well understood. These reactions also require multiple equivalents of a metal powder reductant, which often requires surface activation, to turn over the nickel catalyst.1,2 Conversions and yields can vary dramatically based on the lot, age, and storage conditions of the metal powders.1g Such inconsistency is of particular concern with regards to scaling these reactions in an industrial setting. Furthermore, use of metal reductants can lead to poor yields when using more reactive substrates where direct reduction of one of the crosscoupling partners can compete. While the exclusion of metal reductants has been achieved using organic3 and inorganic4 reductants as well as photoredox catalysis,5 we sought an alternative, robust catalyst reduction method without the need for specialized reagents via electrochemistry.6 With some precedent for the electrochemical reduction of nickel catalysts for aryl−aryl couplings7 and couplings with activated α-chloro ketones and esters,8 we hoped to generalize electrochemical reduction techniques to unactivated aryl−alkyl cross-couplings while circumventing the above-mentioned shortcomings of such reactions with metal powders. Moreover, we sought to thoroughly examine the adaptation of conditions with a metal powder reductant to electrochemical conditions and the details of reaction optimization thereafter. Herein, we report the development of an electrochemical method for reducing catalytic nickel for sp2sp3 cross-electrophile coupling reactions. Initial studies were carried out on the cross-coupling of ethyl4-bromobenzoate 1 (1.3 mmol) and 1-bromo-3-phenylpropane 2 (1.3 equiv) at room temperature in N,N-dimethylacetamide (DMA) with NiCl2(dme) (10 mol%), 4-4′-dimethoxy-2-2′bipyridine (dmbpy, 10 mol%), and NaI serving as both a © 2017 American Chemical Society

Received: May 26, 2017 Published: July 13, 2017 3755

DOI: 10.1021/acs.orglett.7b01598 Org. Lett. 2017, 19, 3755−3758

Letter

Organic Letters Table 1. Optimization of Electrochemical Nickel-Catalyzed Cross-Coupling of 1 and 2

entry 1 2 3 4 5 6 7 8 9 10e

anode Mg Al Zn Fe Zn Fe Zn Zn Zn

temp(°C) 22 22 22 22 65 65 65 65 65 65

Current (mA)

charge passed (F/mol)

10 10 10 10 10 10 50 16 5

yielda 3 (%)

Ar−Hb 4 (%)

Ar−Ara 5 (%)

8 (11)d 91 80 85 68 22 50 65 61

20 3 6 11 24 21 26 30 39

1 6 9 4 5 3 3 3 0

b

N/A 4.0c 4.0 4.0 2.0 2.6 6.0 3.5 2.0

a

NMR yield using internal standard. bNo reaction. N/A = not applicable. cIncomplete conversion. Nonelectrochemical reaction with 2 equiv Zn powder activated with 0.1 equiv of trifluoroacetic acid.

e

d

Recovered starting material.

well as substrate control2,11 provide handles for controlling selectivity, presumably by altering the relative rates of steps in the catalytic cycle. Electrochemistry provides direct control specifically of the reduction step via the current, giving access to an additional handle for reaction optimization, which can be controlled in a continuous fashion. A proposed mechanism for the reaction, based on that proposed by Weix and co-workers for metal powder reductive coupling1c is given in Figure 1, where Ni(0) enters the catalytic cycle via reduction of Ni(II) at the cathode while the metal rod anode is oxidized.12

the theoretical 2.0 F/mol (Table 1, Entry 5) and that with the Fe anode completing with a slight excess of charge of 2.6 F/mol (Table 1, Entry 6). Only a slight drop in selectivity compared with the room temperature reaction was observed with the Zn anode, still giving an 85% yield of cross-coupled product. The reaction with the Fe anode saw a larger drop in yield and selectivity at higher temperature, giving a 68% yield of crosscoupled product and larger amounts of hydrodehalogenation. The use of a Zn anode gave higher yields and better selectivity in general and was used exclusively in subsequent optimization. Notably, the electrochemical cross-coupling reaction with the Zn anode (Table 1, Entry 5) performed better than the reaction under nearly identical conditions but with Zn powder as the reductant (Table 1, Entry 10) which gave a 61% yield with 39% hydrodehalogenation. This case shows a clear difference in selectivity using Zn powder versus optimized electrochemical conditions with a Zn anode. Though a minor drop in selectivity was observed with the Zn anode between room temperature and at 65 °C, final reaction optimization was continued with heating for two reasons. First, many other aryl bromide substrates either perform better with or require heating to proceed. Second, the increase in reaction rate with heating allows for shorter reaction times and higher current efficiencies. Final reaction optimization involved an evaluation of reaction currents. At a current of 50 mA, the reaction suffered from low yield, poor selectivity, and low mass balance (Table 1, Entry 7). Furthermore, the reaction required 6.0 F/mol of charge to be passed to achieve full conversion. Reducing the current to 16 mA (Table 1, Entry 8), still required 3.5 F/mol of charge with yield, selectivity, and mass balance intermediate to that between the reactions at 10 mA and 50 mA. Finally, at a lower current of 5 mA (Table 1, Entry 9) the reaction recovered the high current efficiency and mass balance obtained with 10 mA, however selectivity for cross-coupled product over hydrodehalogenation was negatively affected, giving only 65% of product with 30% hydrodehalogenation. Overall, the correlation between current and selectivity suggests that minimizing side reactions can be achieved by tuning the rate of reduction to maintain a balance of the catalyst in the various oxidation states. It has previously been observed that additives10 and ligands as

Figure 1. Proposed mechanism for electrochemical cross-electrophile coupling.

After observing that the reaction is sensitive to the reduction rate, we sought to identify how the current needed to be scaled with the size of the reaction, as well as catalyst loading. We hypothesized that if the absolute amount of nickel catalyst were increased, then the current could be increased proportionally without complete consumption of Ni(II) before its turnover. Indeed, doubling both the current and the reaction scale, (i.e., maintaining the same concentrations of all reaction components), gave a nearly identical yield and selectivity and went to completion within the same 2.0 F/mol (Table 2, Entries 1 and 2). From the current screen described above, 20 mA at 1.3 mmol scale is too high of a current leading to poor yield, selectivity, and current efficiency (Table 1, Entry 8). However, the higher current is acceptable when the reaction components 3756

DOI: 10.1021/acs.orglett.7b01598 Org. Lett. 2017, 19, 3755−3758

Letter

Organic Letters Table 2. Tuning of Electrolysis Current with Reaction Scale and Catalyst Loading

entry

mmol of 1

currenta (mA)

mol% Ni

yieldb 3 (%)

Ar−Hb 4 (%)

Ar−Arb 5 (%)

1 2 3 4

1.3 2.6 6.5 1.3

10 20 50 5

10 10 10 5

85 85 80 79

11 10 16 14

4 5 3 2

Table 3. Reaction Scope for Electrochemical CrossElectrophile Couplings

a

The ratio of current to catalyst is a constant value of 77 mA/mmol Ni. bNMR yield.

are scaled by the same factor. Similarly, 50 mA could be applied on an even larger 1.5 g scale while maintaining current efficiency and with only a small decrease in selectivity (Table 2, Entry 3). Finally, it was found that the catalyst loading could be lowered to 5 mol% if the current was proportionally decreased to 5 mA at 1.3 mmol scale (Table 2, Entry 4). Again, high current efficiency was maintained with only a minor decrease in selectivity, in contrast to the much lower yield and selectivity observed when the current is dropped to 5 mA at a 10% catalyst loading (Table 1, Entry 9). Based on these trends, it appears that the relevant factor for tuning the current is the current per mole of catalyst, as the current corresponds directly to the amount of nickel reduced per unit time. Scaling this ratio with reaction scale leads to reintroducing the same concentration of nickel back into the catalytic cycle per unit time for a given catalyst loading. For a lower catalyst loading, the current needs to be scaled back such that the rate of reduction does not exceed the rate at which the Ni(II) species is recycled. With a set of working conditions in hand, we sought to demonstrate that the scope of the electrochemical crosselectrophile couplings was comparable to those using metal reductants through representative examples. Indeed, the electrochemical conditions successfully gave cross-coupling products for a variety of aryl-bromide substrates (Table 3). In most cases, yields under electrochemical conditions were comparable to those using zinc powder under similar conditions with the same optimized ligand.13 In the case of 8, however, better cross-coupling selectivity and a significantly higher yield were obtained electrochemically than with zinc, similar to the case of 3 previously described. This again demonstrates that the removal of Zn powder alongside reduction rate control of electrochemistry can be beneficial in select cases. We were also pleased to find that the electrochemical conditions were compatible not only with the commercially available 4-4′-dimethoxy-2-2′-bipyridine ligand (dmbpy) but also with amidine-based ligands (L1−L3) which have recently been demonstrated to give high yield and selectivity for these reactions.14 For example, the reactions to form ortho-substituted aryl 9 and aza-indole 12 proceed in significantly higher yields with the amidine-based ligands than with bipyridines, making compatibility of the electrochemistry with these ligands vital for the method’s substrate diversity. While the general electrochemical conditions developed for 3 (Table 1) applied directly to some substrate pairs, other substrate combinations required some minor changes to these conditions. Reactions of the heterocyclic aryl bromides (10−

a

Nonelectrochemical reaction, 2.0 equiv of Zn powder reductant activated with 0.1 equiv trifluoroacetic acid. b2.0 equiv alkyl bromide used. cRef 13. d0.075 M NaI electrolye. e0.2 M Bu4NBF4 electrolyte. f No NaI added.

12) as well as that with the secondary alkyl bromide (8) required lower currents for high selectivity and current efficiency. The need for lower currents for these reactions coincided with slower reaction rates observed for these substrates under metal powder reductive conditions. Thus, the optimal current for a given reaction is based on both the catalyst loading as previously described, but also on the rate at which the catalyst is turned over through the catalytic cycle. It was also found that the reaction to form 12 proceeds poorly both with metal and electrochemical reduction methods in the presence of NaI, producing significant amounts of bialkyl. In the case of the metal reductant, no NaI additive was added. Alternatively, in the electrochemical case, NaI was replaced as the electrolyte with the commonly used Bu4NBF4 to similar effect. Finally, we turned our attention to cross-couplings of substrates we have found in the past to be incompatible with metal powder reductants. Electron-poor pyrimidine 13 and pyrazole 14 formed in low yields under Zn powder conditions, with decomposition of and direct Zn insertion into the aryl bromide competing with Ni-mediated cross-coupling. In both 3757

DOI: 10.1021/acs.orglett.7b01598 Org. Lett. 2017, 19, 3755−3758

Letter

Organic Letters

2017, 19, 2150. (c) García-Domínguez, A.; Li, Z.; Nevado, C. J. Am. Chem. Soc. 2017, 139, 6835. (4) Xu, H.; Zhao, C.; Qian, Q.; Deng, W.; Gong, H. Chem. Sci. 2013, 4, 4022. (5) (a) Zhang, P.; Le, C.; MacMillan, D. W. C. J. Am. Chem. Soc. 2016, 138, 8084. (b) Duan, Z.; Li, W.; Lei, A. Org. Lett. 2016, 18, 4012. (c) Paul, A.; Smith, M. D.; Vannucci, A. K. J. Org. Chem. 2017, 82 (4), 1996. (6) (a) For a review of mediated electrolysis reactions, see Steckhan, E. Angew. Chem., Int. Ed. Engl. 1986, 25, 683. Deronzier, A.; Moutet, J.C. From Comprehensive Coordination Chemistry II; McCleverty, J. A., Meyer, T. J., Eds.; 2004; Vol. 9, pp471−507. (b) For recent examples of mediated electrochemical palladium catalysis, see Jiao, K.-J.; Zhao, C.-Q.; Fang, P.; Mei, T.-S. Tetrahedron Lett. 2017, 58, 797. (7) For homocoupling of aryl halides, see: (a) Rollin, Y.; Troupel, M.; Tuck, D. G.; Perichon, J. J. Organomet. Chem. 1986, 303 (1), 131. (b) Meyer, G.; Rollin, Y.; Perichon, J. J. Organomet. Chem. 1987, 333, 263. (c) Cassol, T. M.; Demnitz, F. W. J.; Navarro, M.; De Neves, E. A. Tetrahedron Lett. 2000, 41, 8203. (d) De Franca, K. W. R.; Navarro, M.; Leonel, E.; Durandetti, M.; Nedelec, J.-Y. J. Org. Chem. 2002, 67, 1838. For cross-coupling of aryl halides, see: (e) Gosmini, C.; Lasry, S.; Nedelec, J.-Y.; Perichon, J. Tetrahedron 1998, 54, 1289. (f) Sengmany, S.; Vitu-Thiebaud, A.; Le Gall, E.; Condon, S.; Leonel, E.; Thobie-Gautier, C.; Pipelier, M.; Lebreton, J.; Dubreuil, D. J. Org. Chem. 2013, 78, 370. (8) Durandetti, M.; Perichon, J. Synthesis 2004, 3079. (9) For a review of the use of sacrificial anodes, see Chaussard, J.; Folest, J.-C.; Nedelec, J.-Y.; Perichon, J.; Sibille, S.; Troupel, M. Synthesis 1990, 1990, 369. (10) Common inorganic additives include NaI, MgCl2, NaBF4, FeBr2, and CuI. Pyridine and others have been used as organic additives. (11) (a) Poremba, K. E.; Kadunce, N. T.; Suzuki, N.; Cherney, A. H.; Reisman, S. E. J. Am. Chem. Soc. 2017, 139 (16), 5684. (12) While it is not impossible that some Zn2+ in solution is reduced to Zn metal at the cathode, this is unlikely based on cyclic voltammetry studies that suggest that under the reaction conditions the Ni(II) complex for all ligands tested are considerably easier to reduce than the Zn2+. Cyclic voltammetry data is provided in the Supporting Information. (13) Ligands were optimized using Zn metal as the reductant. (14) (a) Hansen, E. C.; Pedro, D. J.; Wotal, A. C.; Gower, N. J.; Nelson, J. D.; Caron, S.; Weix, D. J. Nat. Chem. 2016, 8, 1126. (b) Hansen, E. C.; Li, C.; Yang, S.; Pedro, D.; Weix, D. J. Coupling of Challenging Heteroaryl Halides with Alkyl Halides via NickelCatalyzed Cross-Electrophile Coupling. J. Org. Chem. 2017, 82, 7085−7092.

of these cases, yields were significantly increased by switching to electrochemical reduction conditions. Similar results have been reported previously for these aryl bromides using the organic reductant tetrakis(dimethylamino)ethylene (TDAE),3 however this reductant is air sensitive, gives amine byproducts that can interfere with the reaction, and in our hands was found to be limited in substrate scope. In summary, we have developed an electrochemical method for nickel catalyzed reductive sp2-sp3 couplings between aryl and alkyl bromides. This method offers the ability to tune the rate of reduction via the current passed, which was found to be vital to reaction optimization. The developed conditions reliably initiate and drive the reactions while giving yields comparable to or in some cases much higher than those obtained using Zn powder as the reductant. With these advantages, the use of electrochemistry broadens the applicability of nickel-catalyzed reductive cross-couplings.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.orglett.7b01598. Experimental procedures, characterization data, and copies of NMR spectra (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*eric.hansen@pfizer.com ORCID

Eric C. Hansen: 0000-0002-5057-4577 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was influenced by the ongoing efforts of the NonPrecious Metal Catalysis Alliance between Pfizer, BoehringerIngelheim, Abbvie, and Asymchem. We also thank Daniel Weix (Univ. of Rochester) for his input.

■

REFERENCES

(1) (a) Everson, D. A.; Jones, B. A.; Weix, D. J. J. Am. Chem. Soc. 2012, 134 (14), 6146. (b) Wang, S.; Qian, Q.; Gong, H. Org. Lett. 2012, 14, 3352. (c) Biswas, S.; Weix, D. J. J. Am. Chem. Soc. 2013, 135 (43), 16192. (d) Knappke, C. E. I.; Grupe, S.; Gartner, D.; Corpet, M.; Gosmini, C.; von Wangelin, A. J. Chem. - Eur. J. 2014, 20, 6828. (e) Everson, D. A.; Weix, D. J. J. Org. Chem. 2014, 79, 4793. (f) Everson, D. A.; Buonomo, J. A.; Weix, D. J. Synlett 2014, 25 (2), 233. (g) Liu, H.; Liang, Z.; Qian, Q.; Lin, K. Synth. Commun. 2014, 44 (20), 2999. (h) Molander, G. A.; Wisniewski, S. R.; Traister, K. M. Org. Lett. 2014, 16 (14), 3692. (i) Molander, G. A.; Traister, K. M.; O’Neill, B. T. J. Org. Chem. 2014, 79, 5771. (j) Lu, X.; Yi, J.; Zhang, Z. Q.; Dai, J.-J.; Liu, J.-H.; Xiao, B.; Fu, Y.; Liu, L. Chem. - Eur. J. 2014, 20, 15339. (k) Weix, D. J. Acc. Chem. Res. 2015, 48 (6), 1767. (l) Kadunce, N. T.; Reisman, S. E. J. Am. Chem. Soc. 2015, 137 (33), 10480. (m) Cai, D.-J.; Lin, P.-H.; Liu, C.-Y. Eur. J. Org. Chem. 2015, 2015, 5448. (2) Durandetti, M.; Gosmini, C.; Perichon, J. Tetrahedron 2007, 63, 1146. (3) (a) Anka-Lufford, L. L.; Huihui, K. M. M.; Gower, N. J.; Ackerman, L. K. G.; Weix, D. J. Chem. - Eur. J. 2016, 22, 11564. (b) Suzuki, N.; Hofstra, J. L.; Poremba, K. E.; Reisman, S. E. Org. Lett. 3758

DOI: 10.1021/acs.orglett.7b01598 Org. Lett. 2017, 19, 3755−3758