Pd-catalyzed Cross-Coupling of α-(Acyloxy)-tri-N-butylstannanes with Alkenyl, Aryl, and Heteroaryl Electrophiles

Abstract

Racemic and scalemic α-(acyloxy)-tri-N-butylstannanes undergo Pd-catalyzed cross-couplings with alkenyl/aryl/heteroaryl iodides, bromides, and triflates in moderate to good yields in THF at 45 °C. Simple aryl iodides and unprotected aza-arenes, two classes of electrophiles that typically react sluggishly, are also good substrates. Cross-couplings proceed with retention of configuration at the alkenyl and stannyl-substituted stereocenters.

The Stille reaction,¹ i. e., the transition metal-mediated cross-coupling of organostannanes with organic electrophiles, has achieved wide acceptance² as an exceptionally mild and efficient method for the creation of C-C bonds, especially between sp- and/or sp²-hybridized centers.³ Not surprisingly, the apparent advantages that would accrue from expanding the traditional scope and structural confines of the Stille reaction have attracted much interest.⁴ In the early 1990’s, this laboratory⁵ and others⁶ explored the utility of tri-n-butylstannanes for the transfer of stereogenic carbons bearing heteroatoms and reported the stereospecific palladium/copper co-catalyzed cross-coupling of scalemic α-alkoxy- and α-aminoalkylstannanes with acid chlorides.^7,8 Subsequent studies led to copper mediated cross-couplings with reactive electrophiles such as allylic and propargylic halides.⁹ The utility of this methodology for the construction of chiral ethers and alcohols was cogently demonstrated during asymmetric total syntheses of the anticancer agent (+)-goniofufurone¹⁰ and the potent endothelium-derived vasodilator 11,12,15-THETA.¹¹ On the other hand, comparable unions between α-heteroatom-substituted triorganostannanes and alkenyl/aryl electrophiles were conspicuously absent¹² and suitable methodology has been elusive.¹³ To address this methodological gap, we conducted an extensive survey of alternative reaction parameters including oxygen substituents¹⁴ and herein describe a practical, stereospecific cross-coupling capable of using a broad range of sp²-hybridized iodides/triflates/bromides (Scheme 1).

Scheme 1.

Proposed Cross-Coupling

Since esters and alkenyl iodides were identified as the most promising pairing in our initial evaluations, α-(acetyloxy)stannane 1a and (E)-β-iodostyrene (2) were selected as the test system. Screening an extensive collection of transition metals salts and complexes, tested individually or in combination, revealed palladium complexes were especially efficacious, and in particular freshly recrystallized Pd(dppe)Cl₂.¹⁵ The yield of adduct 3a increased proportionately with the amount of Pd(dppe)Cl₂ up to 10 mol % (62%; Table 1), but thereafter neither more catalyst nor alterations in mode of addition improved the outcome.¹⁶ Yields of 3a were patently better in THF as well as the reaction rate compared to other common solvents, inter alia, toluene, DME, and DMF; there was no cross-coupling in NMP, acetone, EtOAc, and CH₃CN. In contrast with the experience of others,¹⁷ sources of fluoride (LiF, TBAF, KF, AlF₃, CsF) did not have a beneficial effect nor did Lewis acids [MgBr₂, AlCl₃, FeCl₃, ZnBr₂, BF₃·Et₂O, Sc(OTf)₃].

Table 1.

Cross-Coupling of α-(Acyloxy)stannanes with (E)-β-Iodostyrene^a

compd	R	time (h)	yield (%)	compd	R	time (h)	yield (%)
a		12	62	f		12	68
b		10	71	g		12	72
c		10	72	h		10	72
d		12	70	i		8	81
e		10	68	j		8	81

^aReaction conditions: α-(acyloxy)stannane 1 (1 equiv), 2 (1.5 equiv), Pd(dppe)Cl₂ (10 mol %), THF, 45 °C, 8–12 h.

Replacement of the acetate of 3a with a benzoate, i.e., 3b, resulted in a modest improvement in yield. Electron donating substituents (3c–d) had little influence above that of the parent aromatic 3b and neither did electron-rich heterocycles like 2-carboxyfuran 3f and 2-carboxythiophene 3g despite the potential for intramolecular coordination in the transition state. In contrast, electron-withdrawing substituents (3h–j) offered an additional improvement in adduct yield and a modest rate enhancement. Because of its better handling characteristics and lower cost than the polyfluorinated aryls 3h,i, p-trifluoromethylbenzoate 3j was used in subsequent optimization studies.

Having standardized the reaction conditions, we next explored the scope of the cross-coupling of p-trifluoromethylbenzoate (TFMB) 1j with a panel of representative sp²-hybridized iodides, triflates, and bromides (Table 2). 3-Iodopropenoate 4, despite its proclivity towards loss of HI and/or isomerization, smoothly cross-coupled using Pd(dppe)Cl₂ (catalyst A) to furnish 5 (entry 1) in excellent yield without complication or loss of the Z-configuration at the alkenyl center (>98% by ¹H/¹³C NMR), thus precluding a conjugate addition mechanism. Under the same reaction conditions, 2-iodo-2-cyclohex-2-enone (6), (Z)-iodostyrene (9), (E)-triflate 12, and unconjugated alkenyl iodide 15 led to adducts 8 (entry 2), 11 (entry 3), 14 (entry 4), and 16 (entry 5), respectively, although the reaction rates were somewhat slower than for 4. TFMB-benzyl alcohols 19 (entry 6), 22 (entry 7), and 24 (entry 8) were obtained analogously from aryl/heteroaryl iodides 17, 20, and 23, respectively, in useful yields.

Table 2.

Scope of Cross-Coupling using α-(TFMB)stannane 1j^a

entry	electrophile	catalystb	time (h)	adduct	yield (%)
1	4	A	6	5	86
2	6: X = I 7: X = Br	A B	10 12	8	61 61
3	9: X = I 10: X = Br	A B	10 12	11	61 52
4	12: X = OTf 13: X = Br	A	10 12	14	60 72
5	15	A	10	16	66
6	17: X = I 18: X = Br	A B	8 12	19	61 64
7	20: X = I 21: X = Br	A B	10 12	22	65 62
8	23	A	10	24	62

^aReaction conditions: 1j (1 equiv), electrophile (1.5 equiv), Pd catalyst (10 mol %), THF, 45 °C, 8–12 h.

^bCatalyst A = recrystallized Pd(dppe)Cl₂, catalyst B = chloro(2-di-tert-butylphosphino-2′,4′,6′-tri-iso-propyl-1,1′-biphenyl)[2-(2-aminoethyl)phenyl] palladium(II).

Comparable couplings of alkenyl and aryl bromides using Pd(dppe)Cl₂ proved more challenging and only poor yields of adduct could be obtained. Fortunately, we discovered that the Buchwald pallacyclic catalyst¹⁸ chloro(2-di-tert-butylphosphino-2′,4′,6′-tri-iso-propyl-1,1′-biphenyl)[2-(2-aminoethyl)phenyl] palladium(II) (t-BuXphos) restored efficacy (Table 2). While reaction times were modestly increased for bromides 10, 13, 18, and 21, adduct yields were generally comparable or even slightly better than those seen with alkenyl iodides/triflates. The exception was (Z)-bromostyrene (10) which likely reflects this electrophile’s greater steric hindrance.

To ascertain the stereochemistry of the C-C bond formation, enantioenriched benzoate 25 (98% ee), readily prepared from the corresponding aldehyde via the one-pot procedure of He and Falck,¹⁹ was added to phenyl iodide using the standard conditions (Figure 1). Saponification of the resultant benzoate 26 gave rise to the known²⁰ 1,3-diphenylpropan-1(R)-ol (98% ee via chiral HPLC), thus demonstrating complete retention of configuration at the stannyl-substituted stereogenic center. This is consistent with prior experience using other classes of electrophiles, but opposite to the inversion of configuration observed by Kells and Chong⁸ using scalemic α-(sulfonamido)organostannanes and Pd/Cu cocatalysis. Analogous coupling of the diastereomeric glyceryl stannane 27²¹ to give 28 (Figure 1) was also instructive and revealed the adjacent stereocenter had no apparent influence on the outcome.

Figure 1.

Cross-Coupling of Enantioenriched Stannanes

In conclusion, we have demonstrated the stereospecific Pd-mediated cross-coupling TFMB-protected α-hydroxystannanes with alkenyl/aryl/heteroaryl iodides/triflates/bromides including secondary cycloalkenyl and unactivated aryl iodides in THF under mild conditions.²² When combined with the recently introduced methodology for the synthesis of racemic and scalemic α-hydroxystannanes,¹⁹ we anticipate the foregoing methodology will find wide utility in the synthesis of heteroatom-substituted stereogenic centers.