Role of Benzylic Deprotonation in Nickel-Catalyzed Benzylic Dehydrogenation

Abstract Alkylarenes are readily functionalized via the corresponding benzylic anions. Benzylic anions have been used for a range of catalytic reactions, including Ni-catalyzed dehydrogenation. Interestingly, the employment of Zn(TMP)2 for slow and incomplete deprotonation of the benzylic position was observed. This manuscript describes a preliminary investigation into the deprotonation of heteroarenes and its relationship to Ni-catalyzed benzylic dehydrogenation.

A central point of reaction development in the area of nickel catalysis has been to uncover new reactivity that has not been accessible with more well-studied second-and third-row metals. 1 One of the major obstacles in this area is that the greater Lewis acidity of nickel compared to palladium has made the development of methods that tolerate a broad range of functionality challenging. For example, catalytic methods that tolerate heteroarenes have been difficult to develop with palladium and the direct analogues of those methods with nickel have suffered even more from functional group compatibility limitations. We recently disclosed a nickel-catalyzed method for benzylic dehydrogenation 2 that was successful with heteroarenes and did not encounter the classic exclusion of substrates containing imine-type nitrogen bases. One underlying reason for this apparent departure may be that the reaction mechanism involves benzylic deprotonation to form benzyl anions that are stronger nucleophiles and react more readily with nickel. The strategic use of benzyl anions may prove to be a general means to broaden the scope of first-row transitionmetal-catalyzed reactions with benzylic substrates.
Benzylic positions can undergo deprotonation with a strong base, and the resulting benzylic anion nucleophiles can engage a range of electrophiles, which provides an inexpensive and effective method for benzylic functionalization. These reactions usually utilize lithium, potassium, or zinc bases to generate cation-stabilized intermediates that react with electrophiles, similar to those that react with enolates. 3,4 Benzylic deprotonation is a powerful means to functionalize a generally inert position. 5

Scheme 1 Benzylic functionalization via deprotonation
Aside from classic polar reactions that take advantage of consonant relationships, transition-metal catalysis is beginning to make headway to allow functionalization at benzylic positions. In particular, benzylic substrates have been a gateway to C-sp 3 cross-coupling reactions. 5 These methods include reaction at the benzylic position enabled by prefunctionalization, such as benzyl halides or pseudohalides, or through their direct deprotonation. Catalysis with benzylic anions as inputs has been particularly fruitful and enabling (Scheme 1A

Cluster Synlett
asymmetric allylic alkylation (AAA) reactions of 2-picoline and toluene derivatives, respectively. More recently, Walsh and coworkers 7c realized the Ni-catalyzed arylation and allylation of diarylmethanes. In those transformations, excess LiHMDS or KHMDS is necessary for the deprotonation of the benzylic proton. The Oshima group 8 realized the Pdcatalyzed arylation of aryl(azaaryl)methanes at the benzylic position in the presence of a hydroxide base (CsOH) to yield triarylmethanes. Hagadorn and coworkers 9 demonstrated that Zn(TMP) 2 (TMP = 2,2,6,6-tetramethyl-piperidinyl anion) is an effective base for the mild deprotonation of 2-methylpyridine to form functionalized organozinc intermediates, which are conveniently coupled with aryl bromides using Pd-catalyzed coupling methods. Knochel and co-workers 10 realized Pd-catalyzed benzylic cross-couplings of pyridines with aryl bromides by employing TMPZnCl·LiCl. In addition to Pd and Ni catalysts, 11 the You group 12 realized Ir-catalyzed benzylic allylation using LiHMDS or LiOt-Bu as base for benzylic deprotonation.
We were interested in using benzyl anions as intermediates for dehydrogenation to form heterocyclic styrenes inspired by the importance of styrenes in polymerization and large-scale industrial conversion of ethyl benzene to styrene. Studies conducted by our group in 2020 yielded a heteroarene benzylic dehydrogenation with wide scope and high yields at practical temperatures. 2 The conditions developed utilize nickel as a catalyst, with Zn(TMP) 2 as base and an aryl oxidant (Scheme 1B). In that study, it was suggested that an equilibrium may exist for the benzylic deprotonation with Zn(TMP) 2 due to the high levels of dehydrogenation product formed in spite of low deprotonation levels. It is thought that this equilibrium contributes to the broad scope and functional group tolerance of the reaction as high concentrations of the benzyl zinc species might have led to various off-cycle decomposition pathways.
In order to more thoroughly understand the limiting factors and general mechanism of the reaction, we performed deuteration studies on several electron-deficient heteroarenes, including pyridines, quinolines, pyrazines, pyridazines, and triazines (Scheme 1C). An investigation with different bases, temperatures, and reaction times to explore the relationship of deprotonation on nickel-catalyzed dehydrogenation is described here, with the goal of promoting additional development of methods that involve benzyl anions.
To begin, a base screen was conducted with 2-phenethyl-5-(trifluoromethyl)pyridine (1). 13 different bases were first added to the starting material in 1,4-dioxane at room temperature for 30 minutes and then heated to 85 °C for 4 hours, followed by quenching with D 2 O at 85 °C. 13 These parameters were investigated because they closely resemble the optimized conditions for heteroarene dehydrogenation. As Table 1 shows, inorganic bases LiOH, Cs 2 CO 3 , K 3 PO 4 failed to deprotonate the benzylic proton to any extent observable by 1 H NMR spectroscopy, and instead resulted in 8-13% decomposition (entries 1-3). Entry 4 shows that the extent of degradation with LiOt-Bu was even more significant (52%). Other stronger organic Li bases, such as n-BuLi, LDA, LiCyan, and LiHMDS, gave 6-15% deuteration with more than 70% degradation (entries 5-9). These results highlight the challenge in employing such bases at elevated temperatures. As expected, the use of Zn bases resulted in less, but significant, decomposition (entries 10-12). Taken both deuteration and degradation percentages into account, Zn(TMP) 2 displays observable deuteration with modest degradation, which was shown previously to be the best base for Ni-catalyzed dehydrogenation. It is thought that the low levels of reversible deprotonation with Zn(TMP) 2 prevent decomposition, while providing a sufficient supply of the anion for productive catalysis. We next conducted a temperature-dependent comparison of yields between deuteration and dehydrogenation. As shown in Table 2 in entry 1, at room temperature when 2 was subjected to basic conditions and quenched with D 2 O, 7% of 2a-d 1 was detected (conditions A), but at that temperature the dehydrogenation product 2b was not formed under otherwise standard Ni-catalyzed dehydrogenation conditions (conditions B). 14 These results illustrate the need for higher temperatures for dehydrogenation. As temperature is increased, both deuteration and dehydrogenation

Cluster Synlett
yields increase (entries 2 and 3); 80 °C was determined to be optimal, with 80% dehydrogenation and 16% deuteration observed. However, when temperature was increased to 120 °C, comparable levels of deuteration were observed with a slight decrease in yield, presumably due to product decomposition. Deuteration of tetrahydroquinoline was also investigated in relation to temperature. Similar findings revealed that deuteration increased with temperature, from 6% deuteration at 0 °C to 32% at 80 °C (see the Supporting Information). Thus, it is not surprising that the optimal temperature for Ni-catalyzed dehydrogenation is around 80°C . These results strongly suggest the existence of an equilibrium for the benzylic deprotonation with Zn(TMP) 2 , and that the position of equilibrium is affected by temperature. 15

Table 2 Temperature-Dependent Comparison of Yields between Deuteration and Dehydrogenation
To further probe the equilibrium effects on Ni-catalyzed dehydrogenation via deuteration, time-dependent deuteration studies were performed on the phenethyl-pyrazine 3 ( Table 3) Identical pairs of reactions were compared at different time points: one without the catalytic system and simply quenched with D 2 O, and the other with the reagents used to induce catalysis. It is striking that after 20 minutes (entry 1), only 2% of 3a-d 1 is formed and yet the yield of 3b over the same time is 45%. This further suggests deprotonation occurs as a pre-equilibrium, as Le Chatelier's principle would allow for dehydrogenation to proceed to a greater extent than the observed deprotonation. The trend persisted at longer reaction times (entries 2 and 3). At the point that the starting material is consumed in the dehydrogenation reaction (360 minutes, entry 3), the extent of deuteration of 3a is only 36% along with moderate degradation. The extent of deprotonation did not significantly increase after 600 minutes (entry 4), suggesting that under these conditions the position of equilibrium is approximately 2:1 (3/3a-d 1 ). The generality of which dehydrogenation proceeded to a greater extent than deprotonation was investigated by examining deprotonation on a series of substrates under conditions similar to the reported dehydrogenation conditions. 2 As shown in Scheme 2, an array of 2-phenethylpyridines with both electron-donating (-OMe) and electronwithdrawing substituents (-CF 3 , -F, -CN) were treated with Zn(TMP) 2 . Interestingly, compound 4 with a methoxy group at the 5-position showed no deuteration, and the substrates with electron-withdrawing groups (5-7) also showed less deprotonation than compound 2. For nitrile 6, considerable decomposition occurred instead, which provides an explanation for the low yield of dehydrogenation of this substrate. This decomposition is in contrast to the fact that deuteration was not detected and limited decomposition was observed for 2-alkylpyridines 8 and 9, which proceed with high yield under the catalytic conditions (conditions B). The apparently low levels of deprotonation is consistent with the notion that only minimal deprotonation is required for a smooth dehydrogenation.
The examination of the aryl fluoride 5 led to a surprising result: this substrate predominantly underwent deprotonation of the pyridine C6 position to a significant extent (65% with only 5% deprotonation at benzylic position). Such a deprotonation with Zn(TMP) 2 has previously been observed by Wang and co-workers in their report on coppercatalyzed electrophilic amination of heteroarenes. 16 This pathway is particularly surprising considering that nickel-

Cluster Synlett
catalyzed dehydrogenation works well for this substrate (66% yield) under the standard reaction conditions. This divergent reactivity suggests that the benzylic position is the site of kinetic deprotonation and dehydrogenation rapidly follows under the catalytic conditions thus avoiding productive pathways via the C6-deprotonated intermediate.
Other heterocyclic systems, such as quinoline (10, 17 11), pyrazine (3), pyridazine (12), and triazine (13) substrates were also tested for the deprotonation. Consistent with expectations, the more nitrogen atoms within the ring, the higher percentage of deuteration observed. 6c For example, pyrazine compound 3 and pyridazine compound 12 showed 36 and 29% deprotonation, respectively. However, triazine (13) gave obvious degradation, and deuteration is hard to identify through NMR before column chromatography. Meanwhile, Ni-catalyzed dehydrogenation of 13 only gave 30% of the desired product and significant decomposition, which can be explained by the instability of the benzyl Zn anion.
We also investigated the deprotonation of substrates that failed to undergo dehydrogenation. Deuteration was not observed for 6-substituted pyridine (14), 4-substituted pyridine (15), 18 and 5,6,7,8-tetrahydroisoquinoline (16). Our working hypothesis to understand the lack of reactivity of 15 and 16 is that an adjacent heteroatom is necessary to facilitate deprotonation and dehydrogenation through coordination to Zn or Ni.
In summary, this report describes the role of deprotonation in Ni-catalyzed dehydrogenation of heteroarenes. Although the exact mechanism has not yet been elucidated, Ni-catalyzed dehydrogenation with slow and minimal deprotonation by Zn(TMP) 2 is generally the most successful. The experiments described herein indicate deprotonation slowly arrives at an equilibrium position while some decomposition occurs. Over the same period of time, dehydrogenation is rapid. The mechanistic understanding provided by this study is expected to assist future reaction development that involves deprotonated intermediates as inputs for catalytic manifolds.

Conflict of Interest
The authors declare no conflict of interest.

Funding Information
We are grateful for financial support from Yale University, Amgen, Genentech, Boehringer-Ingelheim Pharmaceuticals, the Sloan Founda-