A Chemoselective Polarity‐Mismatched Photocatalytic C(sp3)−C(sp2) Cross‐Coupling Enabled by Synergistic Boron Activation

Abstract We report the development of a C(sp3)−C(sp2) coupling reaction using styrene boronic acids and redox‐active esters under photoredox catalysis. The reaction proceeds through an unusual polarity‐mismatched radical addition mechanism that is orthogonal to established processes. Synergistic activation of the radical precursor and organoboron are critical mechanistic events. Activation of an N‐hydroxyphthalimide (NHPI) ester by coordination to boron enables electron transfer, with decomposition leading to a nucleofuge rebound, activating the organoboron to radical addition. The unique mechanism enables chemoselective coupling of styrene boronic acids in the presence of other alkene radical acceptors. The scope and limitations of the reaction, and a detailed mechanistic investigation are presented.

[11] The mechanistic requirements of these processes are relatively well understood, specifically with respect to radical philicity (Scheme 1a): [12,13] aligning radical and alkene polarity is key to successful bond formation.Generally, nucleophilic radicals engage electrophilic alkene SOMOphiles and vice versa, [12][13][14] with the main exception arising from ambiphilic radicals, which can engage both electron-rich and electron-poor alkene SOMOphiles. [12,13]ertinent examples of these reactivity principles can be seen in radical additions to well-known electron-rich π-systems, such as vinyl boron compounds where ipso-substitution takes place using the electrophilic radicals generated from αhalocarbonyls, as developed by Leonori, [15] or the Togni reagent, as developed by Koike and Akita (Scheme 1b). [16,17]xamples of polarity-mismatched radical addition to alkenes are rare.This has inspired development of strategies for formal polarity-mismatched bond formations.For example, a recent approach by Silvi and co-workers used vinyl sulfonium salts as vinyl halide equivalents, allowing preparation of products that display the same downstream reactivity profiles (i.e., use as an electrophile for subsequent alkylations) but without the reactivity issues of, for example, vinyl bromide (Scheme 1c). [18]This innovative approach allows formal access to the polarity-mismatched products; however, from a radical philicity perspective, this reaction remains a polarity-matched addition.Yu has developed cascade processes based on single electron transfer (SET) and energy transfer (EnT, e.g., Scheme 1d) and hydrogen atom transfer (HAT) where radical intermediates are intercepted by organoboron reagents. [19]Alternative approaches have been developed using transition metals, for example, the Pd(I)mediated methods developed by Wu and Loh [20] and Gevorgyan. [21]ere we show a direct polarity-mismatched radical addition of a nucleophilic radical to an electron-rich SOMOphile (Scheme 1d).
Based on our previous work in nucleophile-nucleophile couplings using organoboron compounds under transition metal catalysis, [29][30][31][32][33] we were intrigued by the prospect of identifying a nucleophile-nucleophile coupling using organoboron compounds with nucleophilic radicals.We selected NHPI ester 2 as a precursor to a nucleophilic alkyl radical.A directed organoboron screening campaign identified a hit reaction where commercial styrene boronic acid 1 a delivered the desired product 3 in 63 % yield upon treatment with Ru(bpy) 3 (PF 6 ) 2 under irradiation with blue LEDs (Scheme 2); however, while noteworthy, this result was found to be irreproducible.Thorough purification of 1 a gave consistent results but very low yields.The main impurity in commercial 1 a is catechol; addition of catalytic quantities of catechol to the reaction restored reactivity and importantly, gave consistent and reproducible results.
Since catechol has an oxidation potential similar to some common additives used as electron shuttles in photoredox reactions, [34][35][36] while unusual, we hypothesized that catechol had the same function.Indeed, based on this initial hit reaction, the system was modified to use the cyclohexyl NHPI ester 4 and optimized to deliver the standard conditions shown in Table 1, which delivered excellent conversion to the expected cross-coupled product 5 (entry 1).
Selected optimization data are worth noting.Firstly, the reaction operates under photoredox conditions -removal of the photocatalyst or light source completely inhibits the reaction (entries 2 and 3).Next, the reaction requires an electron shuttle, with PhNMe 2 the most effective (see ESI for full details of the additive screen).Removal of the electron shuttle significantly diminished efficiency, although a small background reaction was noted.Finally, boron speciation was a key component -the boronic acid was the most effective, with Bpin and Bcat esters less effective (entries 5 and 6).Contrary to previous radical additions using vinyl organoboron reagents, [15][16][17] the BF 3 K and BMIDA were completely unreactive (entries 7 and 8).This, along with other evidence, revealed important information on the mechanism of the reaction (vide infra).Finally, the reaction was incompatible with arylboronic acids (see ESI).
The generality of the reaction was explored by application to a panel of substrates (Scheme 3).A broad scope was observed in the NHPI component, with primary, secondary, and tertiary alkyl radicals, including well-known nucleophilic α-amino radicals (e.g., 23, 24, 30, 31) and α-oxo radicals (e.g., 33, 34) accommodated.Functional groups that are potentially reactive with radical species were compatible, including an alkyl bromide (19), alkenes (20, 21), and alkynes (16,  22).[18] Allyl radicals were unsuccessful (41).Similarly, the styrene component was broadly tolerant of functional group electronic and regiochemical variation.A dienyl example underwent the expected ipsosubstitution in low yield (58). Interception of the putative benzylic radical and/or carbocation (vide infra) was not observed (e.g., 51, 59, 60).39] Control experiments also suggested in situ isomerization (see ESI).The unusual reactivity of this system compelled deeper investigation (Scheme 4).Firstly, support for the formation of the expected alkyl radical from the precursor NHPI ester was achieved by using radical clock NHPIs (Scheme 4a).Under standard conditions, use of alkene substituted NHPI 61 delivered a ca. 1 : 3 mixture of 62 and ring closed analogue 63.Likewise, cyclopropyl NHPI 64 led exclusively to ring opened product 20.
Addition of TEMPO to the standard reaction completely inhibited the reaction (Scheme 4b).Greater insight was found in the reactivity profiles of the organoboron component (Scheme 4c).Styrene boronic acid 1 a undergoes the reaction to deliver 5 in high conversion.Substitution of the alkene was tolerated but only in the α-position.Styrene 65 delivered the expected product in 67 % yield; however, substitution on the β-carbon (i.e., 66) gave no conversion.The origin of this reactivity arises from an increase in LUMO energy for 66 due to a sterically induced deconjugation of the π-system.[39] In contrast to polarity-matched processes, [15][16][17] this leads to exclusive selectivity for styrene boronic acids, with alkyl substituted alkenyl boronic acids, such as 67, unreactive, and establishes a previously unknown chemoselectivity element in these radical addition processes.Alkylboronic acids (e.g., 68) were also unreactive.
The unique chemoselectivity characteristics of this reaction were further demonstrated by variation of the alkene

Communications
(Scheme 4d).The nucleophilic alkyl radical liberated from NHPI ester 4 would be expected to undergo effective reaction with ambiphilic radical acceptors, such as styrenes, and electrophilic acceptors, such as acrylate esters; [12,13] however, this was not the case.While styrene boronic acid 1 a delivers 5 in good yield, styrene 69, ethyl acrylate 70, and ethyl cinnamate 71 were completely unreactive.
In the reactions above where no product was observed (i.e., using alkenes 66-71), the majority of the NHPI ester was recovered, suggesting the reaction failed to initiate or turnover in the absence of the styrene boronic acid.Increased quantities of photocatalyst and PhNMe 2 did not induce any product formation with 66-71 (see ESI).We therefore hypothesized that the styrene boronic acid was involved in radical generation from the NHPI ester.
The role of the styrene boronic acid in radical initiation was established in a series of competition experiments (Scheme 4e).While styrene 69 is unreactive under standard conditions, in the competition reaction between 1 a and 69, a ca. 2 : 1 ratio of 5 and 75 was observed.Conversion of styrenes to ketones such as 75 is a known reaction under photoredox conditions using nucleophilic radicals, proceeding via radical addition to the styrene, oxidation of the benzylic radical to the carbocation, and interception by DMSO leading to Kornblum-type oxidation; [40,41] however, under the developed conditions, the styrene is only reactive in the presence of the styrene boronic acid.The same effect was observed using styrene boronic acid 66 (Scheme 4c), which was unreactive to the cross-coupling, but enabled similar levels of conversion of 69 to ketone 75.Boric acid and arylboronic acids (which were inert to the coupling process) also promoted formation of 75; however, the same experiments using 70 and 71 did not lead to 75, 73 a/b, 74 a/b, or any oxidized derivatives (see ESI), surprisingly indicating that the nucleophilic radical does not react with these electron-poor alkenes.Similarly, electron-deficient styrenes did not react with the nucleophilic radical in the presence and absence of B(OH) 3 or PhB(OH) 2 (see ESI).Collectively, these data suggested that the boronic acid activated the NHPI ester to photocatalytic reduction, similar to the activation of B 2 Cat 2 proposed by Aggarwal for decarboxylative borylation, [42] and Chen on radical addition to ketoacids, [43] and Indeed, UV studies suggested a Lewis pairing interaction between 1 a and 4 (Scheme 4f).Since the reaction works moderately with the equivalent styrenyl Bpin and Bcat (see Table 1, entries 5 and 6), H-bonding activation of the NHPI is unlikely. [19]In addition, since the reaction does not operate with the equivalent styrenyl BF 3 K and BMIDA, there is a clear requirement for a neutral organoboron reagent (see Table 1, entries 7 and 8).Accordingly, a Lewis pairing interaction between NHPI is proposed, which is consistent with both Aggarwal's and Chen's proposed activation model. [42,43]tern-Volmer studies confirmed that only PhNMe 2 quenches the excited photocatalyst (Scheme 4g).A range of other photocatalysts were assessed, including those capable of energy transfer, but delivered no reaction (see ESI). 11 B NMR studies supported boronate formation from 1 a and phthalimide anion (Scheme 4h).Light ON/OFF experiments supported a lack of chain process and the role of PhNMe 2 as catalytic electron shuttle (see ESI).
Based on all the above data, we propose a reaction mechanism as shown in Scheme 5. Consistent with established oxidation/reduction potentials (indicated), [5,34,36,44,45] excitation of [Ru(bpy The nucleofuge rebound removes the possibility of uncontrolled activation of the boronic acid as the boronate derivative (i.e., 86).Interception of the benzylic carbocation 88 with free boronate, such as the alkene difunctionalization reactions developed by Molander, Rueping, and others [46][47][48][49] is therefore avoided, further highlighting the distinct mechanism of this polarity-mismatched coupling.
In summary, we have developed a rare polarity-mismatched radical addition reaction.The developed conditions allow the straightforward arylalkenylation of nucleophilic radicals with nucleophilic styrene boronic acids under photoredox conditions using an electron shuttle to mediate a radical-polar crossover process.Mechanistic investigations support a synergistic activation pathway where the NHPI ester is activated by the organoboron towards electron transfer.The resulting decomposition allows a nucleofuge rebound of phthalimide anion, which activates the boronic acid to radical addition.This unique mechanism displays unusual reactivity, enabling chemoselective coupling of styrene boronic acids in the presence of other well-known alkene radical acceptors.

Supporting Information
The research data supporting this publication can be accessed at https://doi.org/10.17630/33380f7f-28cd-45b8-b55b-dd6b83651e89.The authors have cited additional references within the Supporting Information.
) 3 ] 2 + (78) leads to [Ru(bpy) 3 ] 2 + * (79), which is reduced by PhNMe 2 to give radical anion 80 and aminium radical 81.Activation of NHPI 82 by coordination to 83 facilitates reduction by 80 to regenerate the ground state photocatalyst 78.Decarboxylation of 84 and nucleofuge rebound of phthalimide anion leads to alkyl radical 85 and simultaneous activation of 84 as boronate 86, which is now primed for radical addition giving benzylic radical 87.Radical-polar crossover, via oxidation of 87 with radical cation 81, regenerates the electron shuttle and benzyl carbocation 88.Elimination then delivers the cross-coupled product 89.