Nickel-catalyzed asymmetric reductive arylation of α-chlorosulfones with aryl halides

We report an asymmetric Ni-catalyzed reductive cross-coupling of aryl/heteroaryl halides with racemic α-chlorosulfones to afford enantioenriched sulfones. The reaction tolerates a variety of functional groups under mild reaction conditions, which complements the current methods. The utility of this work was demonstrated by facile late-stage functionalization of commercial drugs.


Introduction
Sulfones are not only important functional units in synthetic chemistry 1 but also unique compounds that have been widely studied in medicinal chemistry. 2Among them, a number of alkyl sulfones are FDA-approved drugs (e.g., bicalutamide for prostate cancer and apremilast for psoriatic arthritis). 35][6] However, asymmetric methods are largely conned to Tsuji-Trost allylation and Michael addition reactions by virtue of formation of relatively stable a-sulfonyl carbanions. 6Fu and co-workers have reported the rst enantioselective Negishi coupling of a-bromosulfonamides and -sulfones with organozinc and organozirconium reagents to furnish secondary benzylic and allylic sulfonamides and sulfones in good yields and high levels of enantioselectivities. 7 Asymmetric alkylation of a-bromosulfonamides with alkenes was later found to be effective under Ni-H-alkene insertion/ alkylation conditions which afforded enantioenriched a-alkyl sulfones and sulfonamides. 8In both cases, the involvement of a-sulfonyl carbon radicals is thought to be key.Although this progress signicantly advances the asymmetric functionalization of a-sulfone derivatives, there are a few issues remaining to be addressed.First, asymmetric introduction of heteroaromatics to the a-carbon of sulfones remains elusive. 8,9econd, the substrate scope may be limited with the current methods due to the use of organometallic nucleophiles and terminal alkenes. 7,8We envision that a cross-electrophile coupling strategy that utilizes readily accessible aryl electrophiles and a-halosulfones may otherwise exhibit different substrate scope and functional group tolerance compared to the reported methods. 102][13] The reaction offers expeditious access to enantioenriched sulfones with no need of pre-generating organozinc reagents, and displays broad substrate scope including a set of heteroaromatics with excellent functional group tolerance.To the best of our knowledge, this work represents the rst asymmetric functionalization of a-sulfonyl compounds using the cross-electrophile coupling strategy that complements the reported Negishi protocol.

Optimization of the reaction conditions
Our investigations began with the coupling between 1 and 2 using chiral bis(oxazoline) ligands (Table 1). 14An extensive screening of reaction parameters revealed that the use of NiBr 2 (dme) (10 mol%) and L1 (20 mol%), Zn as the reductant and MgCl 2 as the additive in DMA delivered sulfone 3 in 73% yield and 91% ee (Table 1, entry 1).Under the latter conditions, replacing L1 with other chiral dibox ligands L2-L6 did not result in better outcomes in terms of both yields and ee's (entries 2-6).The use of Mn to replace Zn diminished the ee (entry 7).MgCl 2 appeared to be crucial for this transformation, whereas MgBr 2 is much less effective, and TBAB gave trace amounts of the coupling product (entries 8-9). 15Change of other parameters such as solvent and the ratio of the coupling partners was not satisfactory (Table S1 †). 14 The addition of the pre-coordinated L1$NiBr 2 complex (10 mol%) as the precatalyst to replace NiBr 2 (dme) (10 mol%) and L1 (20 mol%) led to slightly higher yield and the same ee (entry 10).
We tentatively speculate that L1 adopting a rigid cyclopropyl linker features a larger bond angle of the C(oxazoline)-C(methylene)-C(oxazoline) bond than those in L2-L4 (entries 1-4), which appears to have a crucial effect on the enantioselectivity.On the other hand, by comparing the results arising from L1, L2 and L5 (entries 1, 2 and 5), it is reasonable to conclude that the rigid tricyclic fused rings in L1 and L2 also fulll an optimal steric and rigidity requirement for enantioselective control.
Moreover, a broad scope of substituted a-chlorosulfones were investigated for the coupling with methyl 4-bromobenzoate (2) (Fig. 2).The coupling products (45-62) were produced in moderate to good yields and high enantioselectivities.Notable functional groups included vinyl and hydroxyl.The aryl groups attached to the sulfone moieties did not seem to have key effects on the enantioselectivities as evidenced in 54-60.Although a-chloro-arylsulfone electrophiles were the primary focus of our research, a-chlorosulfonamide (61) and a-chloroalkylsulfone (62) were also investigated.The yields were modest and the enantioselectivities were reasonably good.
The applicability of this asymmetric a-sulfonyl arylation method was showcased by late-stage diversication of a number of commercial drugs.To our delight, the coupling of a-chlorosulfone 1 with bromo/iodo-substituted indomethacin, aniracetam, fenobrate, procacine and clobrate derivatives delivered the enantioenriched products 63-67 in high ees.Next, we examined the arylation of celecoxib-derived a-chlorosulfone; compounds 68 and 69 were obtained in modest yields and high enantioselectivities (Fig. 3). 16e have also examined the coupling of a-chlorosulfone 1 with 4-bromoanisole, and 1 with methyl 4-iodobenzoate.Both reactions gave the products in low yields (15%).The mass balance for the reaction with less reactive 4-bromoanisole includes hydro-dechlorination of 1 and a large amount of recovered 4-bromoanisole.In the case of methyl 4-iodobenzoate, we recovered substantial amounts of 1 and the aryl iodide.The rst case (with 4-bromoanisole) can be easily attributed to the low reactivity of aryl bromide deactivated by the electron-donating group (MeO), leading to mismatched reactivity of the coupling partners with the Ni catalyst.The inhibition of the reaction when using more reactive methyl 4iodobenzoate is surprising.We reason that a different competing path may operate, which prevents recycling of the Ni catalyst for the coupling process.Such a phenomenon is oen seen in our previous studies, e.g., recovery of aryl-I from the aryl-Ni(II)-I complex. 17Thus, we speculate that the use of electron-poor aryl bromides and electron-neutral and -rich aryl iodides ensures matched reactivities of the coupling partners with the Ni catalyst.

Mechanistic consideration
To understand the possible mechanism of this transformation, several experiments were conducted.A cyclopropyl-containing substrate 70 was prepared and subjected to the standard reaction conditions.We did not detect the corresponding Fig. 1 The scope of aryl halides.a The standard reaction conditions; isolated yields are provided; ee is determined by chiral HPLC analysis.b Mn instead of Zn, 1.5 equiv.heteroaryl iodide used.c Mn instead of Zn, 1.5 equiv.heteroaryl bromide used.d Mn instead of Zn, 2.0 equiv.heteroaryl bromide used, dioxane is used as the solvent, rt.Edge Article Chemical Science cyclopropane-retaining product 71.Rather, only the ringopening coupling product 72 was obtained in 23% yield (Fig. 4), which indicated the involvement of a radical ringopening process.12c,d, 15 The radical nature of the coupling event was further veried by the inhibition of the reaction of 1 and 2 with externally added TEMPO (1 equivalent), the yield of 3 dropped to 5% with 90% ee; no TEMPO-trapping adducts were detected.
In general, the coupling of an aryl electrophile with an alkyl halide follows a radical-chain mechanism, as initially proposed by Weix. 18One of the key experiments therein showed that a stable Ar-Ni(II)-X complex can initiate radical formation by reacting with an alkyl iodide to generate the coupling product.The process involves halide abstraction to generate an alkyl radical and an Ar-Ni(III)X 2 intermediate.The radical then diffuses to bulk solution and nds Ar-Ni(II)-X to form Ar-Ni(III) X-alkyl, whose reductive elimination gives the product.Following this, we rst conducted a catalytic reaction of 1 with methyl 4-bromobenzoate which generated racemic 3 in 65% yield along with the hydrodechlorination product 1-H in 25% yield (eqn (1)).Next, we performed the stoichiometric reaction of Ar-Ni(II)-Br (complex 73) with 1 in DMF at 0 C, and the racemic product 3 was obtained in 23% yield associated with recovered 1 (eqn (2)).Addition of MgCl 2 and/or Zn increased the coupling yield of 3.With Zn, reduction of 1 was observed.
These results are in accordance with a radical-chain mechanism, in which the addition of an alkyl radical to Ar-Ni(II)X is featured. (1) (2) Thus, we propose a mechanism that the oxidative addition of an aryl halide to Ni(0) results in an Ar-Ni(II)-X complex, which intercepts an a-sulfonyl radical to give the Ar-Ni(III)X-alkyl Fig. 2 The standard reaction conditions; isolated yields are provided; ee is determined by chiral HPLC analysis.complex, and upon reductive elimination, the product is omitted along with a Ni(I)X complex or a dimeric Ni(I) species. 19he latter species reacts with a-chlorosulfone to generate the alkyl radical and Ni(II)ClX.The radical diffuses to the bulk solution to combine with the Ar-Ni(II)X complex, whereas the Ni(II)ClX is reduced to Ni(0) by Mn, allowing the catalytic cycle to continue (Scheme 1).

Conclusions
In summary, we have developed a Ni-catalyzed asymmetric reductive cross-coupling between a-chlorosulfones and aryl halides.This transformation enables the synthesis of enantioenriched a,a-disubstituted sulfones from simple organic halides under mild conditions, including easy introduction of a wide set of heteroaromatic moieties into the a-sulfonyl positions.The present method is tolerant of a variety of functional groups, affording the coupling products generally in modest to good yields and high enantioselectivities.Finally, this work provides a convenient tool for late-stage functionalization of bioactive compounds by facile introduction of chiral a-sulfonyl subunits.

Fig. 3
Fig. 3 Late-stage functionalization.The standard reaction conditions; isolated yields are provided; ee is determined by chiral HPLC analysis.a X ¼ Br. b X ¼ I. c Mn instead of Zn.

Table 1
Optimization for the formation of 3 a