Pd-Catalyzed Dynamic Kinetic Asymmetric Cross-Coupling of Heterobiaryl Bromides with N-Tosylhydrazones.

A dynamic kinetic asymmetric Pd-catalyzed cross-coupling reaction of heterobiaryl bromides with ketone N-tosylhydrazones for the synthesis of heterobiaryl styrenes is described. The combination of Pd(dba)2 as a precatalyst with a TADDOL-derived phosphoramidite ligand provides the corresponding coupling products in good yields and high enantioselectivities under mild conditions. Racemization-free N-oxidation and N-alkylation of the products allowed us to obtain appealing functionalized axially chiral heterobiaryl derivatives.

Axially chiral biaryl atropisomers are fundamentally important in nature due to their presence in a large number of natural products and bioactive substances.[1] Moreover, they are also key structural frameworks in material sciences, supramolecular chemistry, and organic synthesis.[2] Remarkably, an axially chiral (hetero)biaryl constitutes the central core of many privileged chiral ligands, catalysts, and auxiliaries that are routinely employed in asymmetric synthesis.[3] Consequently, a great deal of effort has already been devoted to the efficient preparation[4] of these chiral structures, including the asymmetric coupling of two aryl groups by oxidative dimerization or cross-coupling,[5] asymmetric [2+2+2] cycloadditions,[6] asymmetric ring opening of bridged biaryl lactones,[7] stereoselective functionalization of prochiral biaryls, in particular by C–H functionalization,[8] (dynamic) kinetic resolutions,[9] and a growing number of organocatalytic approaches.[10] Our group reported in 2013 an alternative methodology for the synthesis of heterobiaryls (e.g., 2-arylpyridines or analogues) consisting of Pd-catalyzed dynamic kinetic asymmetric (DYKAT) coupling between aryl boroxines and racemic heterobiaryl triflates.[11] The resolution strategy is based on the formation of cationic oxidative addition diastereomeric intermediates (Scheme A) in which the configurational stability of the stereogenic axis is compromised by the widening of angles φ1 and φ2. This method was later extended to perform dynamic kinetic C–P,[12] C–N,[13] and other C–C cross-couplings[14] from diverse heterobiaryl electrophiles. On the contrary, catalytic processes initiated by formation of metal carbenoids followed by migratory insertion have rarely been applied to the synthesis of axially chiral compounds. Inspired by the work of Barluenga and Valdés,[15] the group of Gu reported on the use of 1-tetralone tosyl hydrazones as carbene precursors in the Pd-catalyzed coupling with substituted 1-naphthyl bromides, affording axially chiral vinyl arenes with large enantiomeric excesses (Scheme B).[16] More recently, a related Cu-catalyzed coupling of diazo compounds with isoquinoline or phthalazine N-oxides has been reported to obtain axially chiral QUINOX analogues, although in racemic form (Scheme C).[17] On the basis of the findings described above, we envisioned that the use of carbene precursors (e.g., hydrazones) as coupling partners in the DYKAT-based strategy should enable the synthesis of bifunctional heterobiaryl olefins via a palladium/carbene insertion, migration, and β-hydride elimination process (Scheme D). As a starting hypothesis, it was assumed that the low rotational barrier in carbenoid intermediate I increases significantly after the migratory insertion event as a result of the geometrical restrictions in the resulting intermediate II, a larger six-membered cycle with long N–Pd and Pd–C bonds.

Scheme 1

Antecedents and Our Synthetic Plan

The initial studies were carried out using the coupling between racemic bromide 1A and acetophenone tosylhydrazone 2a as the model reaction, with NaOBu as the base, anhydrous toluene as the solvent at 60 °C, 10 mol % Pd(OAc)2, and 12 mol % ligand as the catalyst system (Table ). Different ligands that proved to be successful in our previous DYKAT processes were screened (see the Supporting Information for complete ligand screening). Bidentate P,P and P,N ligands such as BINAP L1, QUINAP L2, Josiphos-type L3, and N,N-pyridine-oxazoline ligand L4 were not effective, and the desired product 3Aa was obtained in a nearly racemic form (entries 1–4). These results can be explained by considering that bidentate ligands result in the formation of coordinatively saturated oxidative addition intermediates that, consequently, are not capable of forming key intermediate I. As expected, monodentate ligands such as TADDOL-based L5–L10 and BINOL-derived L11–L13 phosphoramidites showed in general better performance (entries 5–13). In particular, TADDOL derivative L8, containing a pyrrolidine moiety on the phosphoramidite, proved to be a promising ligand affording the desired (R)-3Aa product in good conversion (83%) and a moderate enantioselectivity (67%) (entry 8). After an additional screening of a Pd source, solvents, and a base (entries 14–21), we found that the use of Pd(dba)2 in combination with LiOBu as the base and anhydrous 1,4-dioxane as the solvent (entry 18) allowed the formation of (R)-3Aa with 85% conversion and 95% ee. Increasing the reaction temperature (65–70 °C) allowed full conversion to be reached, although at the expense of the enantioselectivity (entries 19 and 20). Finally, using a slightly larger excess of 2a (1.5 equiv), the reaction also reaches full conversion while maintaining an excellent 95% ee (entry 21). Moreover, the amount of ligand could also be reduced to 10 mol % without erosion of the enantioselectivity or the catalytic activity (entry 22).

Table 1

Screening of Ligands and Reaction Conditionsa

	[Pd]	L	base	solvent	C (%)b	ee (%)c
1	Pd(OAc)2	L1	NaOtBu	toluene	95	0
2	Pd(OAc)2	L2	NaOtBu	toluene	22	3
3	Pd(OAc)2	L3	NaOtBu	toluene	9	5
4	Pd(OAc)2	L4	NaOtBu	toluene	32	0
5	Pd(OAc)2	L5	NaOtBu	toluene	90	57
6	Pd(OAc)2	L6	NaOtBu	toluene	72	21
7	Pd(OAc)2	L7	NaOtBu	toluene	82	57
8	Pd(OAc)2	L8	NaOtBu	toluene	83	67
9	Pd(OAc)2	L9	NaOtBu	toluene	82	51
10	Pd(OAc)2	L10	NaOtBu	toluene	58	51
11	Pd(OAc)2	L11	NaOtBu	toluene	20	7
12	Pd(OAc)2	L12	NaOtBu	toluene	24	9
13	Pd(OAc)2	L13	NaOtBu	toluene	36	5
14	Pd(TFA)2	L8	NaOtBu	toluene	85	67
15	Pd2(dba)3	L8	NaOtBu	toluene	48	70
16	Pd(dba)2	L8	NaOtBu	toluene	76	70
17	Pd(dba)2	L8	LiOtBu	toluene	82	92
18	Pd(dba)2	L8	LiOtBu	dioxane	85	95
19d	Pd(dba)2	L8	LiOtBu	dioxane	>99	89
20e	Pd(dba)2	L8	LiOtBu	dioxane	>99	91
21f	Pd(dba)2	L8	LiOtBu	dioxane	>99	95
22g	Pd(dba)2	L8	LiOtBu	dioxane	>99	95

Reaction conditions: 0.1 mmol of 1A in an anhydrous solvent (1.2 mL), 2a (0.12 mmol, 1.2 equiv), and 3 equiv of base.

Conversions were determined by 1H NMR spectroscopy.

The ee values were determined by HPLC on chiral stationary phases.

Reaction carried out at 70 °C.

Reaction carried out at 65 °C.

With 0.15 mmol (1.5 equiv) of 2a.

Reaction performed with 10 mol % ligand.

The coupling reaction of bromide 1A could also be extended to other aromatic tosylhydrazones (Scheme ). The reaction tolerates hydrazones 2b–d containing electron-donating (OMe and Me) or slightly electron-withdrawing (Cl) groups in the para position, affording products 3Ab–d in excellent yields and enantioselectivities of ≤96% ee. Additionally, the reaction also tolerates substrates containing different groups (F, OMe, and Me) in the ortho (2e), meta (2g), and ortho, meta (2f) positions, affording the desired products (R)-3Ae–g in excellent yields and excellent enantioselectivities (89–93% ee). A 1.5 mmol scale reaction (0.5 g) of rac-1A and 2a was performed, affording (R)-3Aa in a similar 82% yield and 95% ee.

Scheme 2

Scope of Hydrazones and Heterobiaryls

Reaction conditions: 0.1 mmol of 1A–D in anhydrous 1,4-dioxane (1.2 mL), 2a–k (0.15 mmol, 1.5 equiv), and 3 equiv of LiOtBu for 24 h at 60 °C. Yields given for isolated products after chromatographic purification. The ee values were determined by HPLC on chiral stationary phases.

Reaction performed on a 1.5 mmol (536 mg) scale.

Absolute configuration determined by X-ray single-crystal analysis.

Next, we examined the scope of other heterobiaryl bromides 1B–D. Their reactivity followed a similar pattern. Different naphthyl picoline 1B, isoquinoline 1C, and quinazoline 1D derivatives could be coupled with the model acetophenone tosylhydrazone 2a and with derivatives 2c–h containing substituents in the ortho, meta, or para positions to afford the desired products (R)-3B–D in excellent yields and enantioselectivities of >90% in most cases. The absolute configuration of product (R)-3Ac could be unambiguously assigned by X-ray diffraction analysis. The absolute configuration of other products (R)-3A–D was assigned by analogy assuming a uniform reaction pathway.

The nitrogen atom of the isoquinoline unit maintains its reactivity and can be used in quaternization reactions such as N-oxide formation with m-CPBA (→4Aa) and N-alkylation with BnBr (→5Aa) to yield interesting functionalized products for applications in asymmetric catalysis (Scheme ).

Scheme 3

Representative Derivatizations

In summary, we have developed a highly efficient methodology for the synthesis of axially chiral heterobiaryl styrenes based on a dynamic kinetic asymmetric coupling between readily available racemic heterobiaryl bromides and tosyl hydrazones. A broad scope, functional group tolerance, and excellent enantiomeric excesses were obtained using a chiral Pd(dba)2/TADDOL-derived phosphoramidite catalytic system.