6812
M. Palucki et al. / Tetrahedron Letters 42 (2001) 6811–6814
The Chichibabin reaction is an efficient, well docu-
mented method for the synthesis of aminopyridine
derivatives.3,4 A review of the literature reveals that
most research efforts, both synthetic and mechanistic,
have focused on intermolecular amination of pyridine
derivatives with NaNH2 or KNH2.6a This may be due,
in part, to the observation that increasing substitution
of the nucleophilic amine typically affords a decrease in
the desired product yield (NH3>1°>aniline, 2°ꢀ3°). By
comparison, reports describing the study and develop-
ment of Chichibabin aminations involving alkylamines
or intramolecular Chichibabin aminations are few in
number, despite its potential in producing nitrogen
heterocycles directly.5,6 Nonetheless, we felt the
Chichibabin cyclization of compound 2 should be feasi-
ble since it would rely on the higher reactivity of the
pyridine carbon at the 6-position relative to the other
carbons towards nitrogen substitution to achieve high
regioselectivity, and on the favorable entropic factor to
ensure an intramolecular amination.
(diphenylphosphino)ferrocene (DPPF) in DMF. The
resulting mixture was heated at 70°C for 6–10 h to
provide the double Suzuki–Miyaura cross coupling
product 5 in 84% isolated yield. Faster reaction rates
and higher yields of the desired product were obtained
when the active Pd-catalyst was generated separately
via heating a solution of Pd(OAc)2 and DPPF in DMF
for 30 minutes compared to adding Pd(OAc)2, DPPF
and DMF directly to the reaction mixture. Removal of
both phthalimide protecting groups in one-pot was
accomplished using aqueous hydrazine in refluxing
EtOH to afford 2 in 98% isolated yield.
Issues of concern in the Chichibabin cyclization of
substrate 2 to the desired product 1 include regioselec-
tivity and intramolecular versus intermolecular reactiv-
ity. A survey of the literature reveals that the carbons at
the 2- and 6-position of the pyridine ring are consider-
ably more susceptible to amination under Chichibabin
conditions than the carbon at the 4-position, which in
turn is significantly more reactive than the carbons at
the 3-position.6 Substrate 2 contains two primary
amines with similar pKas. However, only the primary
amine connected at the 5-position of the pyridine ring
can aminate at a favorable 4- or 6-position, whereas the
primary amine connected at the 2-position of the pyri-
dine ring can only aminate at the unfavorable 3-posi-
tion of the pyridine ring. Finally, because of the high
effective molarity of the appended amine,8 entropic
factors should favor the intramolecular reaction over
the intermolecular reaction.
Our approach for the preparation of 2 via Pd-catalyzed
cross coupling of 3 with a common partner involved
investigation of three different strategies: (1) double
Heck reaction with acrylonitrile followed by hydro-
genation, (2) double Sonogashiro reaction with propar-
gyl amine followed hydrogenation and (3) double
Suzuki–Miyaura reaction via an in situ hydroboration
of an allyl amine derivative followed by deprotection.
Attempts at a double Heck reaction with acrylonitrile
resulted in only low-moderate yields of the mono-Heck
product (substitution at the 2-position) despite screen-
ing a variety of solvents/Pd-source/ligands. Similarly,
attempts at a double Sonogashiro reaction yielded at
best a 4:1 ratio of mono-Sonogashiro product (substitu-
tion at the 2-position) to starting material. By contrast,
the first attempt at a double Suzuki–Miyaura reaction
via 9-BBN hydroboration of a phthalimide-protected
allyl amine followed by coupling resulted in a 77% yield
of the desired product.7
A variety of reaction conditions for the Chichibabin
cyclization of 2 to 1 were screened. The optimization
studies include choice of base, solvent, and reaction
temperature. The results of these studies are shown in
Table 1.
Hawes and Davis obtained a 30% isolated yield for the
Chichibabin cyclization of 3-(3-pyridyl)propylamine to
1,2,3,4-tetrahydronaphthyridine using the following
optimized conditions: refluxing the substrate in toluene
in the presence of 2 equivalents of sodium for 72 h.8
Using these conditions as a starting point, the cycliza-
tion of 2 to 1 was examined by heating a toluene
solution of 2 in the presence of either sodium or sodium
hydride. Moderate yields 1 were obtained (entries 1 and
2), with sodium hydride affording slightly higher yields.
Similar to the results obtained by Hawes and Davis,
long reactions times were required for complete conver-
sion of the starting material. The use of other metal
hydride reagents did not increase product yield (entries
As illustrated in Scheme 2, synthesis of substrate 2 was
accomplished in three steps and in 80% overall yield.
Reaction of allylamine with phthalic anhydride in
DMF in the presence of molecular sieves afforded the
crystalline phthalimide-protected allylamine 4 in 97%
isolated yield. Hydroboration of 4 using 1.2 equiv.
9-BBN (0.5 M THF) resulted in complete conversion to
the desired alkyl borane. This mixture was charged with
0.45 equivalents of 2,5-dibromopyridine, 2.0 equiva-
lents K2CO3, and a 70°C solution of 0.045 equivalents
Pd(OAc)2 and 0.054 equivalents of 1,1%-bis-
O
O
PhthN
1) 9-BBN, THF
H2NNH2
NPhth
NPhth
O
2
NH2
N
2) Pd(OAc)2, DPPF
K2CO3, DMF, 3
(84%)
98%
DMF, MS
97%
4
5
Scheme 2.