.
Angewandte
Communications
Table 2: Substrate scope.[a]
efficiency to generate the active species plays a key role on
the rate of most cross-coupling reactions. Unfortunately, none
of the other palladium sources surveyed offered a similar
balance between activity and selectivity to that of our initial
choice. Neither did other solvents and bases (See the
Supporting Information). Under optimized conditions (see
Table 1), variation of the ligand structure was next performed
(entries 4–11). Replacement of the tert-butyl group at the
phosphorus atom in ligand 5e by a less-bulky and less-
donating phenyl ring, such as in 5a, led to a net increase in
enantioselectivity while the yield of 8a was substantially
reduced (86% ee, 17% yield, entry 4). The use of an ethyl or
a cyclohexyl substituent did not prove beneficial (entries 6
and 7). The presence of a methyl or a phenyl substituent at the
2-position of the pyridyl unit systematically affected the
activity and, more importantly, the selectivity of the in situ
generated catalyst (entries 5, 8, and 9). In an attempt to
combine the best features of ligand 5a (aromatic P substitu-
ent) and 5e (electron-rich phosphine, sterically demanding
substituent), ligands 5h and 5I, which both contain an ortho-
methoxy substituent and, in the case of 5I, an additional para-
methoxy group, were designed.[21] To our delight, both of
them delivered 8a with unprecedented levels of enantio-
selection, albeit with moderate yields (96% ee, 12% yield and
96% ee, 35% yield, respectively; entries 10 and 11). Suspect-
ing the solubility of the base may be crucial in obtaining
a satisfactory reaction rate, DMF was finally evaluated as
solvent for this cross-coupling reaction.[22] Gratifyingly, the
yield of 8a significantly improved under otherwise identical
conditions (entry 12). When the reaction time was prolonged
to 48 h, 8a was obtained in 75% yield and 96% ee (entry 13).
Reactions performed at higher temperatures led to improved
rates, but at the expense of the enantioselectivity
(entry 14).[23]
Entry
7
X
n
R1
R2
8
Yield [%][b]
ee [%][c]
1
2
3
4
5
6
7
8
a
b
c
d
e
f
g
h
i
Br
Cl
I
1
1
1
1
1
1
1
1
1
2
1
H
H
H
F
OMe
CF3
Me
H
H
H
Me
Me
Me
Me
Me
Me
Me
Et
a
a
a
b
c
d
e
f
75
96 (S)
98 (S)
98 (S)
95 (S)
95 (S)
92 (S)
87 (S)
94 (85)[f] (S)
75 (R)
20 (S)
35 (R)
6[d]
15
Br
Br
Br
Br
Br
Br
Br
Br
66[e]
83
91
52
80 (>99)[f]
9
10
11
iPr
Me
Ph
g
h
i
51
>99
99
j
k
H
[a] Reactions performed on a 0.5 mmol scale. Standard conditions:
Pd(OAc)2 (5 mol%), (Ra,R,Rp)-5i (10 mol%), cesium carbonate
(1.2 equiv), DMF solvent, 808C, 48 h. [b] Yield of isolated product.
[c] Determined by GC or HPLC analysis using a chiral-phase column.
Absolute configuration assigned by analogy with literature data.
[d] Along with 49% of 7b and 45% of hydrogenolysis product.
[e] Contains 34% of inseparable 7d. [f] These values refer to the reaction
carried out at 1108C.
tion of the aryl bromide moiety with electron-withdrawing or
electron-donating substituents was well-tolerated; products
8b–e were obtained in practical yields with good to excellent
enantioselectivities (entries 4–7). Whereas an a-ethyl sub-
stituent, as in 7h, delivered the product in good yield and high
enantioselectivity (entry 8), the use of a secondary alkyl
substituent (R2 = iPr) had a deleterious effect on both the
activity and the selectivity (entry 9). A six-membered ring
(entry 10) and a product with a phenyl substituent (entry 11)
could be formed quantitatively, albeit with much lower ee
values. Initial efforts to extend this method to an intermo-
lecular process have met with limited success.[26]
Lautens and co–workers have recently demonstrated the
ability of tBu3P to promote reversible oxidative addition into
aryl bromide bonds at the catalytic level in the context of the
palladium-catalyzed selective coupling of polyhalogenated
substrates.[27,28] Remarkably, when dibrominated substrate 7l
was subjected to the optimized reaction conditions, the
corresponding mono-brominated product 8j was obtained in
48% yield and greater than 99% ee (Scheme 2). No traces of
debrominated 7l or 8j were observed along with the
remaining unreacted starting material (50% of recovered
yield for 7l). Upon increasing the reaction temperature to
1108C, 8j could be obtained in 87% yield and 99% ee. This
initial result suggests that ligand 5i, like tBu3P, allows for
reversible oxidative addition into aryl bromide bonds.
To further validate our initial ligand design, a comparative
study was conducted using commercially available chiral
(P,N) ligands A–G that have proven successful in a number of
other catalytic asymmetric transformations (entries 15–
21).[15,24] Although the cyclization product was generally
obtained in good to excellent yields, none of these scaffolds
gave 8a with levels of enantioselectivity comparable to those
obtained with 5i. Interestingly, ligand G, which was identified
by Buchwald and co-workers as one of the most promising
candidates in their seminal contribution,[14] displayed both
reduced activity and enantioselectivity under our optimized
reaction conditions.
The scope and limitations of the intramolecular
a–arylation of a variety of a-branched aldehydes was next
investigated, employing the new optimal conditions (Table 2).
As one may expect from the reactivity trend of aryl halides in
ꢀ
C C bond-forming reactions, the use of aryl chloride
precursor 7b gave much lower yield of product while
maintaining a very high level of enantioselectivity (entry 2).
More surprisingly, the corresponding aryl iodide 7c led to
significantly diminished yield of the bicyclic aldehyde
(entry 3). Although further studies are required to elucidate
the origin of this phenomenon, it suggests that iodide may
To further investigate whether the aryl bromide bond in 8j
is indeed reactive towards palladium catalysis, cyclization
product rac-8j was prepared independently and subjected to
a series of Suzuki cross-couplings using phenylboronic acid as
a reaction partner. Under prototypical reaction conditions,
using either PCy3 or ligand 5i, the corresponding product 9j
inhibit the coupling reaction, as recently observed in related
[25]
ꢀ
Pd-catalyzed C N cross-coupling processes. Para substitu-
3828
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3826 –3831