Journal of the American Chemical Society
Article
albeit in a lower yield (Table 1, entry 8). Ligand L4 with 2,6-
dimethyl substituents further enhanced the reaction efficiency
and enantioselectivity (Table 1, entry 9). The 1-naphthyl- and
9-phenanthryl-subsituted N,N,P-ligands (L5 and L6) did not
give better results than L4 (Table 1, entries 10 and 11). At last,
we identified that L4 was the best one, affording product 3
with 89% ee at ambient conditions. Further varying the copper
salts resulted in a decreasing yield of product 3 with almost a
similar ee (Table 1, entries 12−14). Lowering the temperature
to −5 °C by adding a cosolvent further increased the
enantioselectivity, and the mixed solvent of DMSO/CH2Cl2
significantly enhanced the reaction efficiency (Table 1, entries
15−17). A trial with the more readily available boronate ester
2b under the optimized conditions indicated a very low
conversion of 2b (Table 1, entry 18). We finally identified the
optimal conditions as follows: the reaction of 1a and 2e in a
molar ratio of 1.5:1.0 in the presence of 10 mol % of CuI, 12
mol % of L4, 4.0 equiv of LiOtBu, and 2.0 equiv of H2O in
DMSO/CH2Cl2 (v/v = 2:1) generated 3 in 80% yield with
94% ee (Table 1, entry 15). Replacing L4 with its
pseudoenantiomer L7 gave rise to the enantiomer of 3 (ent-
3) in 72% yield with 95% ee under otherwise identical
conditions.
Table 2. Scope for Synthesis of Enantioenriched 1,1-
Diarylalkanes
a
With the optimal reaction conditions in hand, we then
investigated the scope of benzyl bromides (Table 2).
Substrates with either electron-donating or electron-with-
drawing functional groups at the para or meta positions of
the phenyl rings reacted smoothly in comparable yields with up
to 97% ee (4−13). The substituents at the ortho positions of
the phenyl rings significantly affected the reaction efficiency:
the 2-fluoro-substituted product 14 was obtained in very low
yield, while other substrates with methyl or methoxyl groups at
the ortho positions did not give rise to the desired products.
Notably, the aryl−X (X = Cl, Br, I, F) bonds did not interfere
with the reaction and the desired products 8−11 were
efficiently generated. The naphthyl bromide was also a suitable
substrate for the coupling reaction to deliver the desired
product 15 with 91% ee, albeit with a low yield. In addition,
the simple linear benzyl bromides also worked well to provide
16−27 in moderate to good yields with 89−96% ee. A myriad
of functional groups, such as terminal olefin (23), ester (24),
nitrile (25), and acetal (26) at different distances away from
the reaction site were all compatible with the reaction
conditions. Furthermore, good chemoselectivity was observed
for the reaction of secondary benzylic bromide in preference to
the primary bromide (27). We next turned our attention to
examining the scope of arylboronate esters. The electron-
donating methoxyl group at the meta/para position was
tolerated to give rise to 28 and 29 in 55−62% yields with 94%
ee. The absolute configuration of 29 was determined to be R
by comparing its HPLC spectrum and optical rotation with
those reported in literature,18 and those of other products were
determined in reference to 29. Besides, the phenyl-substituted
arylboronates were also amenable to the coupling reaction to
afford 30 and 31 in excellent enantioselectivity. More
importantly, the electron-withdrawing functional groups
(carbonyl, cyano) at the meta/para position of arylboronates
were also suitable for the reaction to furnish the coupling
products 32−35 with 83−93% ee. Overall, by variation of
functional groups at either benzylic halides or B(mac)-derived
arylboronate esters, this strategy provided a wide range of
enantioenriched 1,1-diarylalkanes.15
a
Reaction conditions: ( )-1 (1.5 equiv), 2 (0.20 mmol), CuI (10 mol
%), L4 (12 mol %), LiOtBu (4.0 equiv), and H2O (2.0 equiv) in
DMSO/DCM (2:1, 2.4 mL) at −5 °C for 48 h under argon.
Heterocycles are prevalent in many pharmacologically
important core structures, and we envisioned that the
installation of various heterocycles into the enantioenriched
1,1-driarylalkane skeletons might be potentially useful for drug
discovery. Thus, we investigated the scope toward the
synthesis of enantioenriched 1-aryl-1-heteroarylalkanes and
1,1-diheteroarylalkanes from heterobenzyl bromides and
heteroarylboronate esters (Table 3). As such, alkyl bromides
bearing heteroarenes such as pyridine (36 and 37) and
quinoline (38) worked well to yield the enantioenriched 1-
aryl-1-heteroarylalkanes, and the quinoline-substituted product
was generated in a higher yield and enantioselectivity.
Similarly, the boronate esters containing different types of
electron-rich heteroarenes, such as the furan and thiophene,
coupled smoothly to provide 39 and 40 in good yields and
excellent enantioselectivity. The electron-withdrawing 2-
chloropyridylboronate ester was also suitable for the reaction
to provide 41 in good yield and enantioselectivity. Unfortu-
nately, no product was observed for the Bpin-derived pyridyl
boronate ester without substituents ortho to the nitrogen.
C
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX