Alkene Borylation-Hydrogenation Enables Highly Active, Site-Selective Cobalt-Catalyzed Borylation

Abstract

A method for promoting highly active and site-selective cobalt-catalyzed C(sp²)–H arene borylation is described. Addition of tert-butyl ethylene (TBE) increased the activity of cobalt-catalyzed borylation of electron-rich arenes. With monosubstituted anisoles and anilines, synthetically useful site-selectivities favoring the meta-position of the ring were observed. Monitoring the catalytic reaction in situ by NMR spectroscopy established a borylation-hydrogenation sequence of tert-butyl ethylene as responsible for the increased catalytic activity where borylation of the alkene preceded functionalization of the arene. Added or in situ generated trans-^tBuCH=CHBPin served as the active H₂ acceptor to overcome the inhibitory effect of HBPin and enabled both HBPin and B₂Pin₂ to be effective reagents for generating the active catalyst. Normal primary deuterium isotope effects of 5.0 and 6.0 in parallel and 3.1(1) and 3.7(3) in competition for meta and para borylation respectively were measured at 23 °C for the catalytic borylation of N-phenyl morpholine, supporting irreversible and rate-determining oxidative addition of the C(sp²)–H bond during the catalytic reaction. The combination of the kinetic isotope effects, in situ reaction monitoring and stoichiometric experiments support the origin of meta selectivity as arising from irreversible oxidative addition of the meta-C(sp²)–H bond to cobalt(I).

Graphical Abstract

INTRODUCTION

Arenes are prevalent subunits in pharmaceuticals with 49% of FDA approved drugs containing at least one substituted benzene ring. ¹ Among the possible substitution patterns, those accessible by aromatic substitution chemistry (1-, 1,2-, 1,2,4) account for 76% of the structures.¹ Conversely, 1,3- and 1,3,5- substitution patterns appear in just 4% and 1% of active pharmaceutical ingredients (APIs) respectively, owing primarily to the lack of reliable methods for their preparation.¹ Though a number of approaches for meta-selective C–H functionalization have been reported including: copper-² and palladium-catalyzed arylations,³ rhodium-catalyzed vinylation,⁴ and ruthenium catalyzed alkylation,⁵ among others,⁶ iridium-catalyzed C–H borylation has emerged as an effective method for converting 1,3- to 1,3,5- substituted arenes owing to the presence of a single, sterically accessible C–H bond.⁷

Despite these advances, preparation of 1,3-disubstituted benzenes remains a challenge and methods for the direct synthesis of these compounds from monosubstituted arenes are attractive to broaden molecular space (Scheme 1).⁸ Strategies for meta-selective borylation of arenes with state-of-the-art iridium catalysts have typically relied on interaction of the substrate with the secondary coordination sphere of the metal through appropriately modified ligands. Although general solutions for meta-selective borylation of 1,2-disubstituted arenes to produce the corresponding 1,2,4-trisubstituted compounds have been reported,⁹ the corresponding methods using monosubstituted arenes to prepare the 1,3-substituted products remain underdeveloped and pose a formidable challenge.^9n,r,s,t,v

Scheme 1.

Representation of 1,3- and 1,3,5-substituted arenes in active pharmaceutical ingredients.

While iridium catalysts for C(sp²)–H borylation are highly active, exhibit broad functional group tolerance, and proceed with sterically controlled selectivity,¹⁰ cobalt catalysis has demonstrated potential for orthogonal and complementary site selectivity.¹¹ A distinguishing feature of pyridine bis(phosphine) (PNP) cobalt catalysts is electronically-driven site selectivity, where the differences between M–C and C–H bond strengths rather than directing groups influence the regioselectivity of arene borylation.¹² This concept has been experimentally demonstrated in the borylation of fluorinated arenes as introduction of ortho-fluorine substituents strengthens M–C bonds and provides the thermodynamic driving force for ortho-to-fluorine selective reactions.¹³ Mechanistic and computational studies have shown that C(sp²)–H oxidative addition of the fluorinated arene is fast and reversible relative to C–B bond formation and hence the “ortho-to-fluorine” effect dominates and determines the outcome of the reaction (Scheme 2A).

Scheme 2.

Electronically-controlled, site-selective cobalt-catalyzed C(sp²)–H borylation of fluorinated arenes.

Altering the electronic properties of the pincer ligand from strongly donating phosphine chelates to weaker pyridine donors in terpyridines promoted meta-to-fluorine selective reactions.¹⁴ The weaker pyridine donors decreased electron density at the cobalt, slowing C(sp²)–H oxidative addition such that it is irreversible and selectivity determining, resulting in the formation of kinetically preferred, meta-to-fluorine organoboronate ester products. These principles offer a potentially powerful, general, and orthogonal method for regioselective borylation reactions through catalyst control. However, the use of a weakly donating pincer resulted in poor activity and narrow scope and functional group tolerance (Scheme 2B).

Additional pincer modifications through introduction of N-heterocyclic carbene donors increased the activity of the cobalt catalyst while maintaining kinetic meta-to-fluorine site selectivity.¹⁵ The increased activity translated onto broader substrate scope and a host of (poly)fluorinated arenes and pyridines underwent site-selective borylation using B₂Pin₂ (Scheme 3A). Mechanistic studies have established that the meta-to-fluorine selective borylation arose from a composite of meta-to-fluorine selective C(sp²)–H oxidative addition as well as a ground state destabilized cobalt(I) meta-to-fluorine aryl complex that underwent more facile reaction with B₂Pin₂.¹⁶ Although (^iPrACNC)CoMe has proven to be effective for the switchable site-selective borylation of fluorinated arenes in the absence of directing groups, extending these concepts beyond fluorinated arenes and other types of activated substrates was limited by the poor activity of the cobalt catalyst.

Scheme 3.

(a) Site-selective C(sp²)–H borylation with (^iPrACNC)CoMe and B₂Pin₂. (b) Formation of (^iPrACNC)CoH₂BPin resulted in catalyst deactivation. (c) Cobalt-catalyzed borylation of electron-rich arenes suffers from poor activity.

More in-depth examination of the substrate scope and in situ reaction monitoring revealed that C(sp²)–H oxidative addition of more electron rich arenes was slowed relative to fluorinated arenes and that the cobalt borohydride complex, (^iPrACNC)CoH₂BPin accumulated during catalytic borylation with B₂Pin₂, a consequence of the formation of HBPin as a stoichiometric byproduct.¹⁶ While (^iPrACNC)CoH₂BPin promoted the reversible C–H activation of selected, activated fluorinated arenes, it was persistent in the presence of non-fluorinated arenes such as benzene-d₆ (Scheme 3B).¹⁶ The stability of (^iPrACNC)CoH₂BPin and inhibition of the catalytic reaction by HBPin accumulation prompted exploration of methods to overcome catalytic deactivation and expand the scope of cobalt-catalyzed borylation to include electron-rich arenes (Scheme 3C). Open questions include whether these reactions would be site selective and what processes govern the origin of selectivity.

Here we describe that addition of tert-butyl ethylene (TBE) increased the activity of borylation reactions promoted by (^iPrACNC)CoMe and expands the scope to include electron-rich arenes. This approach has enabled the meta-selective borylation of a host of anisoles and anilines that lack traditional directing groups. Monitoring the reactions in situ established that TBE underwent C(sp²)–H borylation prior to the arene and that the resultant trans-^tBuCH=CHBPin serves as a hydrogen acceptor to regenerate the active cobalt catalyst and overcome inhibition by HBPin. Deuterium kinetic isotope effects supported that the observed site-selectivity derives from irreversible oxidative addition of the C(sp²)–H bond to the putative cobalt(I) boryl complex, (^iPrACNC)CoBPin.

RESULTS AND DISCUSSION

Studies on the cobalt-catalyzed borylation of electron-rich arenes commenced with 1,2-dimethoxybenzene (veratrole). This substrate was selected to assess catalytic activity and contains only one sterically accessible C(sp²)–H bond for borylation. Standard catalytic conditions employed 4 mol% of (^iPrACNC)CoMe as the precatalyst with three equiv. of B₂Pin₂ in a 0.5 M THF solution of the arene for 48 hours at 23 °C. Three equivalents B₂Pin₂ and longer reaction times were used to assay the impact of reaction variables. Under these conditions, only 22% of the aryl boronate ester was obtained. Repeating the experiment with addition of one equivalent of HBPin reduced the yield to 2%, demonstrating the inhibitory effect of the borane byproduct. Given that alkenes are known to improve the performance of C–H borylation reactions by scavenging HBPin¹⁷ and that [(A)CNCCo] complexes are known to catalyze hydroboration reactions,¹⁸ one equivalent of TBE was added to the catalytic borylation and improved the yield of the organoboronate ester to 58% along with formation of trans-^tBuCH=CHBPin. Increasing the amount of TBE to 2.5 equivalents increased the yield to 92% (Scheme 4A).

Scheme 4.

Effect of HBPin and TBE on the yield of the cobalt-catalyzed borylation of veratrole with B₂Pin₂. NMR yields using 1,3,5-trimethoxybenzene internal standard are reported and the major product is shown.

To probe the generality of the enhanced activity observed upon addition of TBE, a series of arenes with one sterically accessible C(sp²)–H bond was examined (Scheme 4B). Oxygenated substrates such as 1,3 dimethyoxybenzene and 2,3-dimethoxymethylbenzoate were borylated in high yields (>99% and 91% respectively) demonstrating the compatibility of the method with electron rich arenes. A reduced yield of 40% was observed with meta-xylene, owing to competitive borylation of the benzylic C(sp³)–H bonds. Sluggish reactivity was also observed with 1,3 di-tert-butylbenzene (15% yield with large amounts of recovered arene and trans-^tBuCH=CHBPin), likely owing to the steric demands of arene substituents. Borylation of this substrate has also proven challenging with [(dtbpy)Ir] and [(tmphen)Ir] catalysts and only recently has highly activity been achieved with a catalyst bearing a specialized spirobipyridine ligand.¹⁹

With a method for increasing the activity of the cobalt catalyst in hand, the borylation of anisole was studied in more detail. Using 2.5 equiv TBE (the optimal loading for veratrole borylation), the equivalents of the diboron reagent were varied from 1 to 3 equivalents. With 1 equivalent B₂Pin₂, the desired product was obtained in 52% yield and 76:16:08 (3–:3,5–:4–) selectivity. Although monoborylation is the major products, the yield is modest owing to competitive diborylation. Addition of a slight excess of B₂Pin₂, furnished higher yields of the diborylated produced where three equivalents were optimal. Under these conditions, the desired product was isolated in 91% yield, 19:75:06 (3–:3,5–:4–) selectivity (Table 1).

Table 1.

The effect of B₂Pin₂ loading on the catalytic borylation of anisole.

a series of monosubstituted benzenes was examined to evaluate the selectivity of the catalytic borylation reaction as a function of substituent. Arenes with electron withdrawing substituents underwent borylation in quantitative conversion and high yield with selectivities that ranged from near statistical to modestly para selective. For example, both PhBPin and methyl benzoate were borylated in >99% yield with 27:73 and 22:78 m:p selectivity, respectively.^15,20 While the site selectivities are indistinguishable from the TBE-free method,¹⁵ the activities are significantly improved as only 58 and 64% yields were obtained in the absence of olefin. Arenes with N,N-dimethylsulfonamide, N,N-dimethylamide, and trifluoromethyl groups were tolerated and the desired products were obtained in quantitative conversions as mixture of 4- and 3,5- borylated products with near statistical selectivity. Because of the highly active cobalt catalyst generated upon treatment with TBE, monoborylated products were not obtained due to the presence of a remaining sterically accessible C(sp²)–H bond which underwent additional borylation under these conditions (Scheme 5A).

Scheme 5.

Scope of the cobalt-catalyzed borylation of mono-substituted arenes in the presence of TBE. Yields determined by NMR spectroscopy using 1,3,5-trimethoxybenzene internal standard.

To address the outstanding challenge of the preparation of 1,3- and 1,3,5-substituted arenes by directing group free catalytic C(sp²)–H functionalization, the site-selective borylation of a series of anilines and aryl ethers was studied. With aryl ethers, uniformly high yields (91 to >99%) of the desired 3,5-disubstituted organoboronate esters were obtained. These products arise from an initial meta-selective borylation of the monosubstituted arene followed by a second borylation at the remaining sterically available C(sp²)–H bond owing to the highly active cobalt catalyst. In each case, high selectivity (90:10 to 92:08 3,5:4) was observed and varied little as a function of the oxygen substituent (Scheme 5B). This represents a notable advance in arene functionalization as it allows direct access to synthetically challenging 1,3,5 trisubstituted arenes from inexpensive and abundant monosubstituted aryl ethers and provides an alternative to iridium-catalyzed methods.⁸

The cobalt-catalyzed borylation of anilines also proceeded with modest activities and meta site-selectivity (Scheme 5C). Significantly, aniline and N-Me-aniline that contain N–H bonds were tolerated, though reduced yields (54 and 47%) and a mixture of 3- and 3,5-borylated products were obtained even when additional TBE and B₂Pin₂ (4 and 4.5 equivalents respectively) were used. The excess reagents were needed to achieve synthetically useful yields and conversions. The added B₂Pin₂ serves as a means to protect the nitrogen as free N–H bonds are deleterious for the cobalt catalyst. The excess boron reagents and deprotection of the nitrogen were achieved upon filtration through silica during work-up. While reduced yields were obtained in these cases, the direct meta-selective C–H borylation of anilines is rare.²¹ In general, across the series of monosubstituted arene substrates examined, the method proceeds with orthogonal site-selectivity to electrophilic aromatic substitution and therefore provides an alternative protocol for C–H functionalization of abundant monosubstituted benzenes.

The interaction of TBE with (^iPrACNC)CoH₂BPin was examined to determine what role the added olefin plays in activation of the cobalt catalyst. Addition of excess TBE to a benzene-d₆ solution of (^iPrACNC)CoH₂BPin at 23 °C followed by analysis by ¹H and ¹¹B NMR spectroscopies established an immediate reaction and complete conversion to (^iPrACNC)CoPh-d₅ (Scheme 6). The TBE was converted to a mixture of the alkane, arising from hydrogenation as well as ^tBuCH₂CH₂BPin from hydroboration. The rapid formation of (^iPrACNC)CoPh-d₅ likely arises from reversible dissociation of HBPin and insertion of the alkene to form the corresponding cobalt alkyl. This complex promotes the oxidative addition of benzene to yield the deuterated alkane and (^iPrACNC)CoPh-d₅. The observed hydroboration of TBE results from the interception of the cobalt alkyl with HBPin.

Scheme 6.

Synthesis of (^iPrACNC)CoPh-d₅ from (^iPrACNC)CoH₂BPin and TBE.

With experimental support for the hydroboration-hydrogenation of TBE as a mode of increasing the activity of the cobalt catalyst, efforts were devoted to understanding the role of TBE under catalytic conditions. N-phenyl morpholine was selected as a representative arene for these studies owing to ease of handing and accessibility of deuterated variants for determination of kinetic isotope effects. To account for the fate of TBE and N-phenyl morpholine over the course of the catalytic borylation, the progress of the reaction was monitored by ¹H NMR spectroscopy. Using the optimized conditions and 8 mol% (^iPrACNC)CoMe to facilitate observation of the cobalt intermediates, rapid and near complete (89% conversion) consumption of TBE was observed after 10 minutes at ambient temperature. A 55:45 mixture of trans-^tBuCH=CHBPin and ^tBuCH₂CH₂BPin was observed prior to any detectable borylation of the arene. This experiment demonstrated that TBE was not itself directly relevant during arene borylation but instead was first converted to trans-^tBuCH=CHBPin, which serves as the true acceptor of H₂ during arene borylation. This observation is likely a consequence of vinylic borylation occurring faster than arene borylation but slows once the boron substituent is introduced. These observations also support that both HBPin and B₂Pin₂ are effective in generating the active cobalt catalyst, removing the inhibitory effect of the former. At this stage of the reaction (10 minutes reaction time), the cobalt pre-catalyst was converted to (^iPrACNC)CoH₂BPin. Following the induction period during which TBE is borylated, the C(sp²)–H borylation of N-phenyl morpholine commenced and then proceeded over the course of 48 hours and was accompanied by conversion of trans-^tBuCH=CHBPin to ^tBuCH₂CH₂BPin (Figure 1).

Figure 1.

Reaction time course for the cobalt-catalyzed borylation of N-phenyl morpholine. A parallel kinetic isotope effect was measured using N-phenyl-d₅ morpholine.

It is significant to note that at early conversion (25 minutes reaction time, 18% conversion), a detectable quantity of 3,5-diborylation was observed, indicating that borylation of the intermediate monoborylated product is competitive with that of the starting material. The amount of meta-borylated product steadily increased over the course of an hour and reached a maximum between 1 and 3 hours at which point diborylation outcompeted the borylation of the arene. The meta-borylated product was steadily converted to 3,5-diborylated product for the duration of the reaction. The combined 3-+3,5-:4 selectivities were constant (76:24 3-+3,5-:4) over the course of the reaction from 18 to 100% conversion, consistent with the same selectivity determining step throughout the reaction (Table 1). Analysis of the observable cobalt complexes showed conversion of (^iPrACNC)CoH₂BPin, the resting state during TBE borylation, (10 mins reaction time) to two or more previously unidentified cobalt compounds at all reaction time points after TBE borylation was complete (25 min to 48h).

Because trans-^tBuCH=CHBPin was the true acceptor during catalytic arene borylation, it was hypothesized that this alkene converts (^iPrACNC)CoH₂BPin to the active cobalt catalyst. To probe this possibility, an equimolar mixture of (^iPrACNC)CoD₂BPin and trans-^tBuCH=CHBPin was prepared in frozen THF-d₈, thawed and the reaction monitored by ¹H NMR spectroscopy. Formation of [(^iPrACNC)CoH]₂ along with ^tBuCHDCHDBPin occured immediately upon thawing (Scheme 7). These observations are consistent with the intermediacy of (^iPrACNC)CoBPin as a previous study has shown that (^iPrACNC)CoBPin is exceptionally reactive and forms [(^iPrACNC)CoH]₂ or (^iPrACNC)CoH₂BPin depending on the amounts of HBPin or B₂Pin₂ present.¹⁶

Scheme 7.

Stoichiometric regeneration of (^iPrACNC)CoBPin from (^iPrACNC)CoD₂BPin and trans-^tBuCH=CHBPin.

To provide additional support for the ability of HBPin to regenerate the active cobalt borylation catalyst in the presence of trans-^tBuCH=CHBPin and to further support the assertion that trans-^tBuCH=CHBPin serves as an H₂ acceptor during borylation, four arenes from each substrate class were evaluated. The optimized catalytic conditions (4 mol% of (^iPrACNC)CoMe, 0.5M THF, 48h, 23 °C) were slightly modified to use 2.5 equivalents trans-^tBuCH=CHBPin and HBPin which provided the desired products in 48% yield (borylated veratrole), >99% yield (23:77 3,5:4 borylated methyl benzoate), >99% yield (93:07 3+3,5:4, 22:78 3:3,5 borylated anisole), and 91% yield (76:24 3+3,5:4, 49:51 3:3,5 borylated N-phenyl morpholine) respectively. Notably, in the absence of trans-^tBuCH=CHBPin, HBPin alone generated no productive borylation and is consistent with inhibition and ultimately deactivation of the catalyst by formation of (^iPrACNC)CoH₂BPin (Scheme 8).

Scheme 8.

Scope of borylation using trans-^tBuCH=CHBPin and HBPin. NMR yields using 1,3,5-trimethoxybenzene internal standard are reported and the major product is shown. No borylation products were observed in the absence of added alkene.

To gain insight into the origin of the site selective borylation of anilines, a parallel deuterium kinetic isotope effect was measured for the borylation of N-phenyl morpholine and N-phenyl morpholine-d₅ at 23 °C. By monitoring the reaction progress by gas chromatography, primary KIEs of k_H/k_D = 5.0 and 6.0 were measured for meta and para borylation respectively. The observation of a normal primary parallel kinetic isotope effect is consistent with turnover-limiting and selectivity-determining cleavage of the C(sp²)–H bond.²² Therefore, the selectivity of the catalytic arene borylation derives from the kinetic selectivity of C(sp²)–H oxidative addition to cobalt(I). Consistent with this result, a competition kinetic isotope effect was measured and furnished primary KIEs of k_H/k_D = 3.1(1) and 3.7(3) for meta and para borylation, respectively. This is similar to the behavior of observed with [(terpy)Co] catalysts during the borylation of activated fluoroarenes, where parallel and competitive KIE values of 2.9(2) and 2.6(3) were reported and are consistent with turnover and selectivity determining C–H bond cleavage.^14b This pathway contrasts with the one operative in the electronically-controlled borylation of fluorinated arenes where ortho or meta-to-fluorine site selectivity is observed depending on the catalyst. With ortho-to-fluorine selective [(PNP)Co] catalysts, borylation of fluorinated arenes proceeds with a parallel KIE value of 1.1(1) at 50 °C, signaling reversible C–H activation and thermodynamic control of borylation enabled by arene isomerization.^12d With meta-to-fluorine selective [(^iPrACNC)Co], parallel and same-flask KIEs of 1.3(2) and 2.9(2) were measured.¹⁶ These values support a pathway where the observed selectivity is derived from a composite of oxidative addition of the C(sp²)–H bond and the reaction of the resulting cobalt(I)-aryl resting state with B₂Pin₂.^15,16

Previous studies have established that (^iPrACNC)CoBPin is the likely intermediate responsible for C(sp²)–H activation during the meta-selective borylation of fluorinated arenes. To support the intermediacy of this compound in the borylation of N-phenyl morpholine, a stoichiometric experiment with 20 equivalents of B₂Pin₂ relative to an equimolar mixture of (^iPrACNC)CoMe and N-phenyl morpholine was performed. Excess B₂Pin₂ was used to minimize complications arising from HBPin byproduct of (^iPrACNC)CoBPin C–H activation. The observed ratio of the resulting isomeric organoboronate ester products is therefore an approximation of the selectivity of C(sp²)–H oxidative addition with (^iPrACNC)CoBPin.¹⁶ The selectivity of stoichiometric borylation of N-phenyl morpholine (76:24) was indistinguishable from the catalytic reaction (75:25), supporting C(sp²)–H oxidative addition by (^iPrACNC)CoBPin (Scheme 9).

Scheme 9.

Selectivity of stoichiometric borylation of N-Phenylmorpholine by putative (^iPrACNC)CoBPin. Assay yields were determined by NMR spectrosopy using a 1,3,5-trimethoxybenzene internal standard and the major product is shown.

DFT studies were also performed to provide additional insight into the origin of meta selectivity. The oxidative addition of both N-phenyl morpholine and methyl benzoate to the cobalt(I) boryl was studied. The results and full details of the computational studies are reported in the Supporting Information. As with fluorinated arenes, the barrier for oxidative addition of the meta C(sp²)–H bond was favored by 1.4 kcal/mol over the para site in both cases, with the overall barrier for methyl benzoate lower as expected for a more electron deficient arenes. As with other computational studies on [(CNC)Co(I) compounds,^13c,16 the barrier for C–B reductive elimination from Co(III) was lower than that for oxidative addition consistent with the observation of cobalt(I) but not cobalt(III) complexes.

The experimental and computational data are consistent with the proposed catalytic cycle presented in Figure 2. Monitoring the reaction in situ revealed an induction period whereby TBE was converted to trans-^tBuCH=CHBPin prior to arene borylation. Once trans-^tBuCH=CHBPin is formed, it serves as a hydrogen acceptor from (^iPrACNC)CoH₂BPin to reactivate the cobalt catalyst. The HBPin liberated as a byproduct of TBE borylation forms (^iPrACNC)CoH₂BPin which is observed as the resting state during the phase of the reaction where the alkene is borylated (Figure 1). Once (^iPrACNC)CoBPin is generated from dehydrogenation of the corresponding cobalt borohydride, it promotes turnover limiting and selectivity-determining C(sp²)–H oxidative addition of the arene. The resulting cobalt(III) intermediate undergoes C–B reductive elimination to form the observed aryl organoboronate ester and the cobalt(I)-hydride. This compound captures HBPin formed from catalytic borylation of TBE and generates (^iPrACNC)CoH₂BPin. Reversible dissociation of HBPin and interaction with trans-^tBuCH=CHBPin enables transfer of H₂ to the alkene and regeneration of (^iPrACNC)CoBPin.

Figure 2.

Proposed mechanism for active catalyst regeneration by TBE.

CONCLUSION

Overcoming inhibition by HBPin has enabled the generation of high-activity cobalt C(sp²)–H borylation catalysts with activities that compete with those of state-of-the art precious metal-catalyzed C–H borylation methods. This increased performance has enabled the direct synthesis of underrepresented but pharmaceutically privileged 1,3,5 trisubstituted anisoles and anilines from readily available monosubstituted arenes through site-selective catalysis. The improved activity derives from a borylation-hydrogenation sequence of the TBE additive where intermediate trans-^tBuCH=CHBPin reacts with (^iPrACNC)CoH₂BPin to regenerate (^iPrACNC)CoBPin. These lessons establish general strategies to improve activity and impart site selectivity in cobalt–catalyzed C(sp²)–H borylation and provide critical insights for catalyst and additive design.

Supplementary Material

ASSOCIATED CONTENT

Supporting Information

The Supporting Information is available free of charge at http://pubs.acs.org.

General experimental considerations, description of catalytic and mechanistic experiments, NMR spectra and results of DFT calculations.

Table 2.

Time course for the catalytic borylation of N phenylmorpholine: Product selectivity ratios as a function of time.

Reaction Time (h)	% Arene Conversion	3-:4-:3,5-
0.17	0	N/A
0.42	18	72:22:6
0.68	36	66:24:10
1	62	51:25:24
3	85	36:24:40
6	95	22:24:50
12	98	12:26:62
24	100	6:24:70
36	100	4:24:72
48	100	4:23:73