A Dynamic Kinetic Asymmetric Heck Reaction for the Simultaneous
Total Page:16
File Type:pdf, Size:1020Kb
Page 1 of 11 Journal of the American Chemical Society 1 2 3 4 5 6 7 A Dynamic Kinetic Asymmetric Heck Reaction for the Simultaneous 8 Generation of Central and Axial Chirality 9 10 José Alberto Carmona,§ Valentín Hornillos,*, §,∫ Pedro Ramírez-López,§ Abel Ros,§,∫ Javier Iglesias-Sigüenza,∫ En- 11 § ,∫ ,§ 12 rique Gómez-Bengoa, Rosario Fernández,* José M. Lassaletta* 13 § Instituto de Investigaciones Químicas (CSIC-US) and Centro de Innovación en Química Avanzada (ORFEO-CINQA), Avda. Amé- 14 rico Vespucio, 49, 41092 Sevilla, Spain. 15 ∫ 16 Departamento de Química Orgánica, Universidad de Sevilla and Centro de Innovación en Química Avanzada (ORFEO-CINQA), 17 C/ Prof. García González, 1, 41012 Sevilla, Spain. 18 § Departamento de Química Orgánica I, Universidad del País Vasco, UPV/EHU, Apdo. 1072, 20080 San Sebastián, Spain 19 20 21 ABSTRACT: A highly diastereo- and enantioselective, scalable Pd-catalyzed dynamic kinetic asymmetric Heck reaction of heterobiaryl sul- 22 fonates with electron-rich olefins is described. The coupling of 2,3-dihydrofuran or N-boc protected 2,3-dihydropyrrole with a variety of quin- oline, quinazoline, phthalazine and picoline derivatives takes place with simultaneous installation of central and axial chirality, reaching excellent 23 diastereo- and enantiomeric excesses when in situ formed [Pd0/DM-BINAP] was used as the catalyst, with loadings reduced down to 2 mol% 24 in large scale reactions. The coupling of acyclic, electron-rich alkenes can also be performed using a [Pd0/Josiphos ligand] to obtain axially 25 chiral heterobiaryl a-substituted alkenes in high yields and enantioselectivities. Products from Boc-protected 2,3-dihydropyrrole can be easily 26 transformed into N,N ligands or appealing axially chiral, bifunctional proline-type organocatalysts. Computational studies suggest that a b- 27 hydride elimination is the stereocontrolling step, in agreement with the observed stereochemical outcome of the reaction. 28 29 30 31 Scheme 1. Asymmetric intermolecular Heck reactions. 32 INTRODUCTION 33 The Heck reaction is a fundamental palladium-catalyzed Previous work: generation of carbon stereogenic centers 34 C–C bond-forming transformation with numerous applica- A) Cyclic alkenes [Pd], L* 2 2 1 R R 35 tions in the synthesis of natural products and valuable syn- R X + 2 or R Z 1 Z 1 36 1 Z High ee’s R R thetic intermediates. Since the first intermolecular asymmet- X = OTf, Br, Cl R2 = H, alkyl 37 ric version reported in 1991 by the group of Hayashi,2 these R1 = Ar, alkenyl, Bn Z = O, NR, CH 38 couplings have become a benchmark to validate the design of 2 39 novel chiral ligands and catalysts rather than finding suitable B) Acyclic alkenes 2 [Pd], L* 2 40 3 R R1 R applications in stereoselective organic synthesis. Only very Ar X + OH 1 O 41 recently, the synthetic utility of the asymmetric Heck reaction R n High ee’s Ar n X = N +BF −, B(OH) 42 has been expanded thanks to the outstanding performance of 2 4 2 43 chiral mixed phosphine/phosphine oxide ligands in several This work: 44 C) Dynamic kinetic asymmetric Heck Reaction representative transformations.4 Additionally, methods to en- 45 5 W W − W Y X able the use of benzylic electrophiles and previously elusive 2 R2 Y Y 46 R + R2 aryl halides6 have also appeared. Moreover, conditions for the N N L Z N 47 Pd [Pd], L* highly enantioselective construction of quaternary stereocen- X L' 48 ?? Z ters from trisubstituted dihydrofurans have also been re- 49 3 7 R 1 R3 R3 50 ported (Scheme 1A). The redox relay Heck-Matsuda reac- R R1 R1 8 tions of acyclic alkenyl alcohols have also significantly ex- (±)-1: X = OTf, ONf IOA Z = O, NBoc 51 W, Y = CH, N panded the synthetic value of the reaction, a strategy that has Stereogenic axis & center 52 generated at once 53 been also extended to oxidative Heck reactions using boronic acids as reactants (Scheme 1B).9,10 Mention is also due to the 54 12 55 dynamic kinetic resolution of atropoisomeric o-iodoacryl- The direct asymmetric cross-coupling approach to axially 56 anilides.11 chiral biaryls has failed so far for the synthesis of heterobiaryls 57 such as QUINAP and related derivatives. Among a handful of 58 59 60 ACS Paragon Plus Environment Journal of the American Chemical Society Page 2 of 11 alternative approaches,13 our group has recently described Pd- promising 82% ee for the major product 3b (Table 1, entry 1). 1 catalyzed dynamic kinetic asymmetric C–C, C–P and C–N An excess of dihydrofuran was used (8 equiv.), due to the vol- 2 cross-coupling reactions of racemic heterobiaryl sulfonates atility of this compound. A screening of ligands and condi- 3 1.14 On the other hand, more complex ligands/catalysts com- tions was then performed to optimize this reaction [Table 1 4 bining central and axial chirality have found a number of re- and Table S1 (supporting information)]. At lower tempera- 5 15 6 cent applications in asymmetric catalysis, but their synthesis ture (60 °C, entry 2) the conversion dropped substantially but 7 usually require tedious multi-step procedures. In fact, exam- the enantioselectivity raised up to 88% ee. No improvements 8 ples of catalytic procedures for the simultaneous generation of were observed using related P,P ligands such as L2 (entry 3) 16 9 axial and central chirality are rare, and none of them were di- or L3 (entry 4), while Josiphos derivative L4 was unproduc- 10 rected to the synthesis of catalysts/ligands with practical ap- tive (entry 5). Not surprisingly, hemilabile ligand L5 afforded 11 plications. In this context, the dynamic kinetic asymmetric the non-isomerized Heck product 3'b exclusively with high 12 Heck reaction from 1 appears as an appealing strategy for the diastereoselectivity, although in low ee (entry 6). Remarka- 13 synthesis of bifunctional heterobiaryls with central and axial bly, though, ligands L6 or L7 were ineffective (entries 7 and 14 stereogenic elements (Scheme 1C). This transformation, 8); a faster alkene insertion may be favored with less σ-donat- 15 however, is particularly challenging for two main reasons: ing P,N and P,O ligands, since they form more electrophilic 16 First, in contrast with the common cationic Heck reaction cationic aryl palladium(II) centers. However, a stronger bind- 17 pathway,17 the oxidative addition intermediate IOA has no va- ing of the pyridine/isoquinoline nitrogen is also expected in 18 cant coordination sites; the relatively good isoquinoline/pyr- this case, preventing further coordination of the olefin. 19 idine N ligand must be displaced by a neutral olefin on the PdII 20 center. Second, the reaction with internal olefins can afford up Table 1. Screening of Reaction Conditions and Ligands. 21 to 8 different Heck stereoisomers considering the formation Pd(dba)2 (10 mol%) 22 N L (11 mol%) N N of two stereogenic elements and the possible double bond mi- + 2 + 23 ONf gration. Therefore, the chiral ligand will be required to per- DIPEA (5 eq.), toluene, 18h O O 24 form an exquisite control of both the generation of the stere- 25 rac-1b 3b 3 b ogenic axis and the migratory insertion event in order to af- 26 ford a satisfactory result. 27 P(tBu) PPh PPh PPh2 2 2 2 MeO Ph2P Fe 28 PPh PPh2 RESULTS AND DISCUSSION 2 MeO PPh2 Me 29 In a preliminary experiment, it was observed that the reac- L4: SL-J002 30 L1: (R)-BINAP L2: (R)-H8-BINAP L3: (R)-MeO-BIPHEP 31 tion of 2-(pyridin-2-yl)phenyl nonaflate 1a and 2,3-dihydro- 32 furan 2 fails to give any reaction product under common Heck PAr2 PPh P(O)Ph O PAr conditions [Pd(dba)2/(R)-BINAP (cat.), DIPEA, toluene, 80 2 2 2 33 OMe PPh2 Ph2P N 34 °C], thus confirming our initial concerns (Scheme 2). It was tBu OA L8: (R)-Tol-BINAP (Ar = p-Tolyl) 35 assumed that the formation of a very stable I (R = R’ = H) L5: (R)-MOP L6: (R)-BINAP(O) L7 (S)-tBu-phox L9: (R)-DM-BINAP (Ar = 3,5-Xylyl) 36 intermediate prevents coordination of the olefin and further b b c 37 migratory insertion. Despite this discouraging result, we en- Entry L T (°C) conv (%) 3b:3'b dr ee 38 visaged that in more strained (less stable) IOA palladacycles (R 1 L1 80 50 10:1 >20:1 82 39 = R’ ≠ H; e.g. from 1-(isoquinolin-1-yl)naphthalene-2-yl no- 2 L1 60 28 6:1 >20:1 88 40 naflate 1b), dissociation of the N–Pd bond would be facili- 3 L2 60 27 11:1 >20:1 82 41 tated by release of steric strain. 42 4 L3 60 11 n.d. n.d. n.d. Scheme 2. Preliminary experiments. 43 5 L4 60 <5 n.d. >20:1 n.d. 44 6d L5 80 40 <1:20 >20:1 13e 45 N Pd(dba)2 (10 mol%), (R)-BINAP (11 mol%) 46 + NO REACTION 7 L6 60 <5 n.d. n.d. n.d. ONf DIPEA (5 eq.), Toluene, 80 °C, 48h 47 O 2 8 L7 60 <5 n.d. n.d. n.d. 48 1a 49 + + 9 L8 60 16 8:1 >20:1 96 R = R 50 10 L9 60 22 6:1 >20:1 99 N P N Difficult N−Pd bond dissociation R R P 51 Pd Pd R P R R = R 11 L9 80 >99 >20:1 >20:1 97 52 P steric f repulsion N−Pd bond dissociation facilitated 12 L9 80 97 >20:1 >20:1 97 53 IOA 54 aConditions: 0.1 mmol of rac-1b, 0.8 mmol of dihydrofuran in 0.5 mL of 55 toluene.