.
Angewandte
Communications
Accordingly, we chose perfluorohexyl bromide (1a) as
our initial substrate based on the following considerations:
(5 mol%), and K2CO3 (2.0 equiv) in 1,4-dioxane at 808C,
39% yield of 3a was obtained without observation of the by-
products 6–9 (entry 1). Switching the pre-palladium catalyst
from [PdCl2(PPh3)2] to [Pd(PPh3)4] dramatically improved the
yield to 62% (entry 2). Encouraged by these results, several
reaction parameters were investigated to improve the reac-
tion efficiency further (for details, see the Supporting
Information). It was found that 1,2-dichloroethane (DCE)
was the optimal solvent, and a higher yield (68%) was
obtained when the reaction was catalyzed by [PdCl2(PPh3)2]
in DCE (entry 3). Other palladium sources also catalyzed the
reaction, but led to lower yields (entries 4–8). The choice of
base also influenced the reaction efficiency, and the best yield
(89% upon isolation) was obtained with Cs2CO3 as a base
(entry 10). However, no reaction occurred when other
phosphine ligands other than Xantphos were used, thus
demonstrating the essential role of Xantphos in the catalytic
cycle (for details, see the Supporting Information). It should
be mentioned that the diminished yields were observed by
reducing the ratio of Pd/Xantphos from 1:2 to 1:1 (entries 11
and 12). In addition, neither the product nor by-products were
observed in the absence of the palladium sources or the base
(entries 13 and 14). Thus, these findings implied that the
tetraphosphine-ligated Pd0 species plays a crucial role in the
promotion of the reaction. Whereas no 3a was observed when
perfluorohexyl iodide (1a’) was tested under optimized
reaction conditions, a mixture of oligomerized by-products
(7a; Path C, Scheme 1), the dimer 9a (Path D), and other
uncertain by-products were provided as major products (for
details, see the Scheme S1 in the Supporting Information).
Furthermore, only a small amount of benzyl iodide derivative
6a (Path B) was observed in this reaction (for details, see the
Scheme S1 in the Supporting Information). We reasoned that
it is probably because of the high reactivity of Pd0/Xantphos
complex which may generate perfluoroalkyl radical from
perfluoroalkyl iodides faster than that from perfluoroalkyl
bromides, thus initiating different side reactions.
A variety of styrene derivatives could be fluoroalkylated
with 1a through this method (Table 2). Versatile functional
groups, such as alkoxycarbonyl and formyl were quite well-
tolerated (3h,i). In particular, the heterocyle thiazole fur-
nished the corresponding product in a synthetically useful
yield (3j). Branched alkenes were also applicable to the
reactions. Good yield was obtained, when cyclic alkene 1,2-
dihydronaphthalene (2k) was examined (3k). As for a termi-
nal branched alkene bearing an alkyl group, the double bond
migrated product was the only product obtained (3l), thus
providing an alternative strategy to access fluoroalkylated
allylic compounds. It is noteworthy that the conjugated alkene
underwent the reaction smoothly, thus providing 3m in good
yield without observation of other by-products. The alkenes
were not restricted to styrenes, as an enamide was also
a suitable substrate, thus providing perfluorohexylated alkene
with 81% yield (3o). In addition, the successful fluoroalky-
lation of dihydropyran makes it possible to rapidly access
fluorinated gylcomimetics for carbohydrate-based studies
(3n).[12] However, aliphatic alkenes were not suitable sub-
strates. Internal linear alkenes also failed to provide products
because of steric effects.
À
À
1) The C Br bond is stronger than the C I bond; as a result,
the generation of the side products 6–9 through Paths B–D
(Scheme 1) from perfluoroalkyl bromides would be slower
than that from perfluoroalkyl iodides. Although perfluor-
oalkyl bromides have been demonstrated to be poor fluo-
roalkyl radical precursors,[7] we envisioned that if a suitable
phosphine ligand in combination with a palladium catalyst
could induce perfluoroalkyl bromides to generate fluoroalkyl
radicals and accelerate Path A, it will benefit the formation of
desired products. 2) Perfluoroalkyl bromides are one of the
cheapest and most widely available fluorinated sources.
However, the transition-metal-catalyzed fluoroalkylation
reactions with perfluoroalkyl bromides remain underdevel-
oped, and represent a challenge because of their inert
reactivity. Therefore, it is of great interest to develop new
and efficient catalytic system for their wide applications in life
and materials sciences.
Initially, the coupling of 1a with styrene (2a) was
examined by using bidentate phosphine Xantphos as
a ligand which binds palladium with a large bite angle[10]
(Table 1). The use of Xantphos was inspired by our very
Table 1: Representative results for optimization of palladium-catalyzed
Heck-type reaction of 1a with styrene (2a).[a]
Entry
[Pd]
Solvent
Base
Yield [%][b]
1
2
3
4
5
6
7
8
[PdCl2(PPh3)2]
[Pd(PPh3)4]
[PdCl2(PPh3)2]
Pd(OAc)2
1,4-dioxane
1,4-dioxane
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
DCE
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
none
39
62
68
35
49
39
39
30
79 (81)
(89)
63
34
n.r.
n.r.
PdCl2
[PdCl2(dppf)]
[PdCl2(MeCN)2]
[Pd(PPh3)4]
[PdCl2(PPh3)2]
[PdCl2(PPh3)2]
[PdCl2(PPh3)2]
[PdCl2(PPh3)2]
none
9
10[c]
11[d]
12[e]
13
14
[PdCl2(PPh3)2]
DCE
[a] Reaction conditions (unless otherwise specified): 1a (0.6 mmol,
2.0 equiv), 2a (0.3 mmol, 1.0 equiv), solvent (2 mL), 24 h. [b] Deter-
mined by 19F NMR spectroscopy using fluorobenzene as an internal
standard and yield within parentheses is that of the isolated product.
[c] DCE (3 mL). [d] 7.5 mol% Xantphos was used. [e] 5 mol% Xantphos
was used. DCE=1,2-dichloroethane, dppf=1,1’-bis(diphenylphosphi-
no)ferrocene, n.r.=no reaction, Xantphos=9,9-dimethyl-4,5-bis(diphe-
nylphosphino)xanthene.
recent investigations into the palladium-catalyzed direct
difluoroalkylation of organoborons, in which a single-electron
transfer (SET) from a Pd0/Xantphos complex to difluoroalkyl
bromides initiated the catalytic cycle and gave difluoroalkyl
radicals.[11] To our delight, when the reaction was carried out
with Xantphos (10 mol%) in the presence of [PdCl2(PPh3)2]
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!