Efficient Stereoselective Carbocyclization to cis-1,4-Disubstituted Heterocycles Enabled by Dual Pd/Electron Transfer Mediator (ETM) Catalysis

An efficient Pd/ETM (ETM = electron transfer mediator)-cocatalyzed stereoselective oxidative carbocyclization of dienallenes under aerobic oxidation conditions has been developed to afford six-membered heterocycles. The use of a bifunctional cobalt complex [Co(salophen)-HQ] as hybrid ETM gave a faster aerobic oxidation than the use of separated ETMs, indicating that intramolecular electron transfer between the hydroquinone unit and the oxidized metal macrocycle occurs. In this way, a class of important cis-1,4-disubstituted six-membered heterocycles, including dihydropyran and tetrahydropyridine derivatives were obtained in high diastereoselectivity with good functional group compatibility. The experimental and computational (DFT) studies reveal that the pendent olefin does not only act as an indispensable element for the initial allene attack involving allenic C(sp3)–H bond cleavage, but it also induces a face-selective reaction of the olefin of the allylic group, leading to a highly diastereoselective formation of the product. Finally, the deuterium kinetic isotope effects measured suggest that the initial allenic C(sp3)–H bond cleavage is the rate-limiting step, which was supported by DFT calculations.

To a solution of S3 (368 mg, 1 mmoL) in acetone (5 mL) were added K 2 CO 3 (207 mg, 1.5 mmoL) and allyl bromide (0.17 mL, 1.40 g/mL, 2 mmol) at rt. After the mixture was stirred at 50 o C for 12 h, the reaction was cooled down to rt, and Et 2 O (10 mL) was added. The mixture was filtrated via eluation through a short pad of silica gel (5 cm, eluent: Et 2 O, 20 mL) to remove the inorganic salts. The solution was evaporated, and the residue was purified by column chromatography on silica gel (eluent: petroleum ether/ethyl acetate = 40/1 to 30/1) to afford the desired product 1n (305 mg, 75% yield from S3): colorless oil. 1

Synthesis of d 6 -1a
To a solution of ethynyltrimethylsilane S4 (2.8 mL, d = 0.69 g/mL, 20 mmoL) in dry THF (30 mL) was added n BuLi (9.6 mL, 2.5 M in hexane, 24 mmoL) at -78 o C under Ar atmosphere. After the mixture was stirred at -78 o C for 15 min, d 6 -acetone (2.9 mL, d = 0.87 g/mL, 40 mmol) was added at -78 o C. The reaction was then allowed to warm to room temperature, and stirred at rt for 2 h. After the reaction was carefully quenched with H 2 O (0.5 mL) at 0 o C, the reaction mixture was filtrated via eluation through a short pad of silica gel (5 cm, eluent: Et 2 O, 50 mL). The solution was evaporated, and the residue was dissolved in THF (20 mL). TBAF (Tetra-n-butylammonium fluoride, 24 mL, 1.0 M in THF, 24 mmoL) was added to the solution, and the reaction was stirred at rt for 10 min. The reaction was carefully S11 quenched with H 2 O (20 mL), and extracted with Et 2 O (3 × 20 mL). The combined organic layers were dried over Na 2 SO 4 , filtered, and evaporated. The residue was purified by column chromatography on silica gel (eluent: petroleum ether/diethyl ether = 15/1) to afford the desired product d 6 -S5 (3.030 g, 78% yield from S4): colorless oil. 1 For the preparation of d 6 -1a from d 6 -S5, see the reported procedure for the synthesis of allenic alcohol 1 , and the typical procedure for the synthesis of 1 from allenic alcohol. Slightly yellow liquid, 95% yield from allenic alcohol. 1

Synthesis of chiral dienallene (S)-1a
Chiral allenic alcohol (S)-5a was prepared via enzymatic kinetic resolution according to reported literature. 1 The ee values of (R)-6a and (S)-5a were determined by chiral GC analysis.

S12
To a dry round flask were added a solution of (S)-5a (180 mg, 1 mmol) in DMF (2 mL) and NaH (60%, 1.1 equiv. to (S)-5a, 44 mg, 1.1 mmol) sequentially at 0 o C. After the mixture was stirred at rt for 5 min, allyl bromide (0.17 mL, d = 1.40 g/mL, 2 mmol) were added sequentially at room temperature. After the reaction was stirred at room temperature for 16 h, H 2 O (10 mL) was carefully added to quench the reaction. The organic layer was separated, and the aqueous layer was extracted with Et 2 O (3 × 20 mL). The combined organic layers were dried over Na 2 SO 4 , filtered, evaporated, and purified via column chromatography on silica gel to afford the desired product (S)-1a (196 mg, 89% yield): colorless oil, which was used for the palladium-catalyzed carbocyclization reaction.

Synthesis of arylboronic acid neopentylglycol esters 2
For the preparation of arylboronic acid neopentylglycol esters 2, see the reported procedure. 2 S13

Typical procedure for the synthesis of products 3
A microwave tube (10 mL) was charged with Pd(OAc) 2

S25
The reaction mixture was stirred at 25 o C under atmosphere of O 2 (equipped with an O 2 balloon). After a specific time, the reaction mixture was diluted with Et 2 O (5 mL), and quickly filtered via a short column of silica gel (5 cm, eluent: 30 mL of Et 2 O). The collected filtrate solution was evaporated and the yield of 3a was determined by 1 H-NMR using anisole as the internal standard.

ETMs employing 5 mol% Co(salophen) and 10 mol% HQ:
A microwave tube (10 mL) was charged with Pd(OAc) 2 (5 mol%), Co(salophen) ( The results from the above experiments are summarized in Table S1. The reaction with Co(salophen)-HQ resulted in a much higher reaction rate than that of the reaction using Co(salophen) and HQ (or BQ) as separate ETMs (Table S1, see also Figure 1 of the Article). These results indicate a higher efficiency in the overall reaction by using the catalyst of Co(salophen)-HQ under aerobic conditions, implying that intramolecular electron transfer occurs between the hydroquinone unit and the oxidized metal macrocycle in the bifunctional catalyst.

Determination of the relative configuration in 3a
1 H NMR data for 3a:

Intermolecular KIE Experiments (Separate experiments)
To an NMR tube with a solution of Pd(OAc) 2 2 and 3). These two reactions were running at room temperature in the NMR machine, and the online detections were recorded at different time (see Table S2 and S3, respectively). The yields were determined by 1 H NMR measurement using CH 3 NO 2 as the internal standard. Due to the nature of the experiment, plots to determine the KIE were taken for 1a ( Figures   S1 and S2).   Due to the nature of the experiment, plots to determine the KIE were taken for d 6 -1a ( Figures S3 and S4).

Computational Methods
The geometries of all stationary points were fully optimized with the dispersion-corrected B3LYP functional 4 using the Gaussian 09 program. 5 For the geometry optimizations, the LANL2DZ basis set with effective core potential 6 was used for the Pd atom, and the 6-

TS-1' is the allenic C(sp 3 )-H bond cleavage with the coordination of the distal olefin instead
of the pending olefin. The barrier of TS-1' (Figure S6) is calculated to be 32.4 kcal/mol relative to Pd(II) acetate trimer.

TS-1A
is calculated to be 2.5 kcal/mol higher in energy than TS-1. The agostic interaction, which occurs in TS-1 but not in TS-1A, is one of the factors that stabilize TS-1 compared to TS-1A.