One-pot multistep synthesis of 3-aminoindolizine derivatives

Article abstract of DOI:10.1016/j.tetlet.2011.01.087

We have developed an efficient one-pot multistep sequence to the synthesis of 3-aminoindolizine derivatives by using Hantzsch ester (9) as a mild hydride transfer agent. The reaction scope was investigated and the effect of substrates on reaction outcomes was discussed.

Full text of DOI:10.1016/j.tetlet.2011.01.087

Tetrahedron Letters 52 (2011) 1392–1394
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
One-pot multistep synthesis of 3-aminoindolizine derivatives
⇑
Lianhai Li , Waepril Kimberly S. Chua
Department of Medicinal Chemistry, Merck Frosst Centre for Therapeutic Research, 16711 Trans Canada Hwy., Kirkland, Québec, Canada H9H 3L1
a r t i c l e i n f o
a b s t r a c t
Article history:
We have developed an efficient one-pot multistep sequence to the synthesis of 3-aminoindolizine deriv-
atives by using Hantzsch ester (9) as a mild hydride transfer agent. The reaction scope was investigated
and the effect of substrates on reaction outcomes was discussed.
Received 27 September 2010
Revised 8 January 2011
Accepted 18 January 2011
Available online 25 January 2011
Ó 2011 Published by Elsevier Ltd.
Keywords:
One-pot
Multistep
Synthesis
Indolizine
Improving efficiency is one of the main pursuits of modern or-
ganic synthesis research.1,2 This is further reinforced by the
increasing concerns of environmental protection and sustainability
in past decades. To date the design, development, and utilization of
efficient environmentally benign synthetic process has become a
conscientious choice of synthetic chemists.3,4 One attractive strat-
egy is to design and develop a novel one-pot multistep synthesis
that helps simplify reaction handling and product purification, im-
prove synthetic efficiency, and reduce solvent consumption and
disposal. This will ultimately help reduce consumption of natural
resources, minimize the potential, harmful impact of various
chemicals on environment, and increase sustainability.5
In one of our medicinal chemistry programs, we needed 1a
(Scheme 1) to prepare more complex analogs containing indolizine
core. Indeed, indolizine can serve as an isosteric replacement for
indoles6 and other heterocycles,7 and is found to possess interest-
ing biological activities.8 To the best of our knowledge, the synthe-
sis of 1a had previously been reported by two groups using a three-
step sequence (Scheme 1). The first step involves the Knoevenagel
condensation between 2-pyridylcarboxaldehyde (2a) and ethyl
cyanoacetate (3a) to form product 4a.9 When 4a was exposed to
DIBAL reduction,9 Raney-Ni catalyzed hydrogenation,9 or electro-
chemical reduction in the presence of chlorotrimethylsilane,10 1a
was obtained in a moderate 45%, 43% or 47% yield from 4a, respec-
tively. We hypothesized that the formation of 1a could proceed
through intermediate 10. The subsequent intramolecular cycliza-
tion reaction of 10 that involves the addition of pyridine nitrogen
to cyanide followed or accompanied by some rearrangement
processes would deliver 1a. Based on these reports and our reac-
tion pathway analysis, we envisioned that under suitable reduction
conditions it should be possible to run the whole sequence, includ-
ing the Knoevenagel condensation between 2-pyridylcarboxalde-
hyde (2a) and ethyl cyanoacetate (3a), and the following
transformation of 4a into 1a, in an environmentally benign
one-pot manner. Literature search showed that Hantzsch ester
(9), a versatile and mild hydride transfer reagent11,12 used for the
promotion of reductive alkylation reactions,13 might serve this
purpose.
O
O
O
ref 9
NC
H
+
OEt
4a
OEt
3a
N
2a
N
CN
85%
a, b, c
or d
reduction
O
OEt
N
OEt
O
N
NH2
1a
N
10
O
O
EtO
OEt
N
H
Hantzsch ester (9)
⇑
Corresponding author at present address: 4201 Hugo, Pierrefonds, Quebec,
Canada H9H 2V6. Tel.: +1 514 4283837.
Scheme 1. Reagents and conditions: (a) DIBAL, Ref. 9, 45%; (b) Raney-Ni in AcOH,
Ref. 9, 43%; (c) e-, Me3SiCl, ref 10, 47%; (d) 9, toluene, 105 °C, 3 h, 70%.
E-mail addresses: lianhai_li@merck.com, lianhai.li@gmail.com (L. Li).
0040-4039/$ - see front matter Ó 2011 Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2011.01.087

L. Li, Waepril Kimberly S. Chua / Tetrahedron Letters 52 (2011) 1392–1394
1393
Table 2
As per our initial proof of concept, we heated 4a with 1.1 equiv
Reaction scope exploration of 2-carbonylpyridine
of Hantzsch ester (9) in toluene at 105 °C for 3 h and the desired
compound 1a was obtained in 70% yield (Scheme 1). This result
suggested that Hantzsch ester (9) was an effective hydride transfer
agent for the transformation and prompted us to investigate the
possibility of running the sequence in a one-pot multistep manner.
In an initial run, we simply heated a mixture of equimolar
amount of 2a and 3a with 1.1 equiv of Hantzsch ester (9) at
105 °C in toluene for 3 h and obtained compound 1a in 12% yield
(Table 1, entry 1). Interestingly, when the same reaction was per-
formed again in the presence of 0.02 equiv of piperidinium acetate,
a widely used Knoevenagel condensation catalyst, the formation of
1a was increased to 56% yield (Table 1, entry 2). Encouraged by
these results, we further explored the effect of the amount of pipe-
ridinium acetate used on the reaction, and found that with 0.05
and 0.1 equiv of the catalyst and we obtained 1a in 74% and 71%
yield (Table 1, entries 3 and 4), respectively. Compared to the
70% yield we obtained from the direct reduction of 4a by Hantzsch
ester (9), it seemed that the one-pot multistep reaction had an
overall improved yield. Another commonly used Knoevenagel con-
NH2
N
CN
R'
R
6
[Pip][OAc]
(0.05 equiv)
9 (1.1 equiv)
toluene, 105 o
NC
NC
R'
R'
O
or
R
+
N
CN
CN
C
5
7
3
or
R
N
8
Entry
Carbonyl
Yielda
Product
O
88%
Br
H
CN
1
2
N
6a
N
Br
5a
NH2
Br
Br
O
61%
31%
H
CN
NH2
N
N
densation catalyst, L-proline, was also employed at 0.05 equiv in
5b
6b
the one-pot multistep process. However, the yield of 1a was de-
creased to 45% (Table 1, entry 5). Based on these results, we
decided to explore the reaction scope with a few other nitriles
using our optimized conditions (Table 1, entry 3). The aim is to de-
velop the method into a general approach for the synthesis of syn-
thetically useful substituted 3-aminoindolizine analogs.14 As
O
H
5c
3
4
CN
N
N
O
6c
NH2
Me
49%
53%
Table 1
Me
CN
Reaction conditions optimization and reaction scope exploration of nitriles
N
N
5d
6d NH2
catalyst,
EWG
N
9
(1.1 equiv)
2a
+
NC
EWG
CN
NH2
toluene, 105 o
C
O
N
NH2
3
5
6
N
1
N
H
N
5e
6e
Entry 3, EWG
Catalysta (equiv) Product
Yieldb (%)
OEt
CN
O
87%
86%
N
N
N
1
3a, EtOC(O)–
None
12
O
H
NH2
5f
NH2
6f
1a
O
2
3
4
5
3a, EtOC(O)–
3a, EtOC(O)–
3a, EtOC(O)–
3a, EtOC(O)–
A (0.02)
A (0.05)
A (0.1)
1a
1a
1a
1a
56
74
71
45
CN
H
N
7
8
N
B (0.05)
NH2
Me
5g
6g
t-Bu
Me
O
N
N
O
6
7
3b, t-BuC(O)– A (0.05)
53
34
CN
CN
NH2
1b
H
97%
N
N
O
7a
OMe
OMe
S
Me
Ph
8a
3c, MeS(O)2–
A (0.05)
O
O
NH2
1c
CN
CN
H
86%
64%
9
O
S
N
N
7b
N
N
Br
8b
8
9
3d, PhS(O)2–
3e, NC–
A (0.05)
A (0.05)
42
44
Br
F
O
NH2
1d
O
CN
CN
H
10
CN
N
F
N
7c
8c
NH2
1e
a
Isolated yield.
a
A = piperidinium acetate; B =
Isolated yield.
L-proline.
shown in Table 1 (Table 1, entries 6–9), electron-withdrawing
groups (EWGs) such as a ketone, sulfone, and cyano group were
b

1394
L. Li, Waepril Kimberly S. Chua / Tetrahedron Letters 52 (2011) 1392–1394
tolerated under these conditions but the corresponding products
were isolated with lower yields. It is noteworthy to point out that
it is possible to further improve the yield if reaction conditions are
carefully optimized for each of these EWGs.15
quence has a wide reaction scope with respect to nitrile substrates;
all the nitriles containing electronwithdrawing groups such as car-
boxylic acid ester, sulfone, ketone, and cyanide tested in our study
provided corresponding 3-aminoindolizine derivatives. However,
the reaction scope with various substituted 2-carbonylpyridines
was dependent on their electronic properties. While most of the
pyridine substrates used in our study provided 3-aminoindolizine
derivatives, those groups with electron inductive effect substituted
at 6-position of pyridine ring led to formation of only uncyclized
products 8.
Moving forward, we chose malononitrile (3e) as our standard
nitrile partner to explore the reaction scope with different 2-car-
bonylpyridines. The results are summarized in Table 2. Overall,
11 carbonyl compounds were subjected to this study and only
one aldehyde, isoquinoline-3-carbaldehyde (not listed in Table 2),
gave a complicated mixture products. However, its regioisomers
quinoline-2-carbaldehyde (5f) and isoquinoline-1-carbaldehyde
(5c) were able to participate in the reaction and provide the corre-
sponding 3-aminoindolizine derivative 6f and 6c in 87% and 31%
yield (Table 2, entries 3 and 6), respectively. Compared to quino-
line-2-carbaldehyde (5f), the replacement of a carbon with nitro-
gen as in 5e led to a lower reaction yield (87% vs 53%, entry 6 vs
entry 5, Table 2). Simple aldehydes such as 5a and 5b are good sub-
strates for this one-pot sequence to 3-aminoindolizine derivatives
6a and 6b (in 88% and 61% yield, respectively, entries 1 and 2 of Ta-
ble 2). It would appear that the intrinsic electronic property of the
substituent at the 6-position of pyridyl ring played a critical role in
controlling the product formation. While a 6-methyl group as rep-
resented by 5g led to the formation of 3-aminoindolizine deriva-
tive 6g in 86% yield (Table 2, entry 7), three other aldehydes
with a substituent at its 6-position, such as 6-methoxy in 7a, 6-
bromo in 7b, or 6-fluoro in 7c, afforded only the uncyclized prod-
uct 8a, 8b, or 8c in 97%, 86% or 64% yield, respectively (Table 2, en-
tries 8–10).16 What in common with these three substituents is
that they can all exert an electron inductive effect in 2-pyridylcar-
boxaldehyde system. There is a possibility that an electronic clash
between the electron cloud of lone-pair electrons of oxygen, bro-
mine or fluorine and of the cyanide prevented the cyclization from
occurring. However, this was ruled out based on the fact that we
could still obtain cyclized product 6e from aldehyde 5e in 53% yield
(Table 2, entry 5). Finally, a methyl ketone 5d was used to explore
the possibility of using the process to synthesize 3-aminoindoli-
zine derivative bearing a substituent at its 1-position and the de-
sired cyclized product 6d was obtained in 49% yield. This
represents a slight improvement on product yield in comparison
to the 44% yield of cyclized product 1e obtained from the corre-
sponding reaction of aldehyde 1a with malononitrile.
References and notes
1. Trost, B. M. Science 1991, 254, 1471–1477.
2. Trost, B. M. Science 1983, 219, 245–250.
3. Song, J. J.; Reeves, J. T.; Fandrick, D. R.; Tan, Z.; Yee, N. K.; Senanayake, C. H.
Green Chem. Lett. Rev. 2008, 1, 141–148.
4. Anastas, P. T.; Beach, E. S. Green Chem. Lett. Rev. 2007, 1, 9–24.
5. Bienayme, H.; Hulme, C.; Oddon, G.; Schmitt, P. Chem. Eur. J. 2000, 6, 3321–
3329.
6. Lehmann, T.; Hubner, H.; Gmeiner, P. Bioorg. Med. Chem. Lett. 2001, 11, 2863–
2866.
7. Jennings, A.; Tennant, M. J. Chem. Inf. Model. 2007, 47, 1829–1838.
8. For references of biological activities involves indolizine core, see references in
the following reference: Bai, Y.; Zeng, J.; Ma, J.; Gorityala, B. K.; Liu, X.-W. J.
Comb. Chem. ACS ASAP.
9. Flitsch, W.; Kahner-Groene, S. Chem. Ber. 1982, 115, 871–877.
10. Troll, T.; Beckel, H.; Lentner-Boehm, C. Tetrahedron 1997, 53, 81–90.
11. For earliest reaction discovery, see: Braude, E. A.; Hannah, J. J. Chem. Soc. 1960,
3268–3270; Norcross, B. E.; Klinedinst, P. E., Jr.; Westheimer, F. H. J. Am. Chem.
Soc. 1962, 84, 797–802.
12. For recent applications, see: Huang, Y.-B.; Cai, C. J. Chem. Res. 2009, 11, 686–
688; Rueping, M.; Tato, F.; Schoepke, F. R. Chem. Eur. J. 2010, 16, 2688–2691;
Ramachary, D. B.; Mondal, R.; Venkaiah, C. Org. Biomol. Chem. 2010, 8, 321–325.
13. Ramachary, D. B.; Vijayendar Reddy, Y. J. Org. Chem. 2010, 75, 74–85.
14. For recent reports of synthesis of 3-aminoindolizine, see: Ref. 8, and Yan, B.;
Liu, Y. Org. Lett. 2007, 9, 4323–4326.
15. At the end of our study, we re-examined more closely the formation of 1e from
2a and 3e (Table 1, entry 9). We found that under the original reaction
conditions as described in Table 1, entry 9, an enamine 11 side product was
formed (Fig. A) in ca. 20% yield. Compound 11 is stable to flash
chromatography over silica gel column
CN
N
N
N
11
Figure A
The general procedure for the synthesis of the 3-aminoindoli-
zine derivatives 1a–e, 6a–g, and uncyclized product 8a–c is de-
scribed below: To
a reaction mixture of cyano substrate 3
(2.0 mmol), 2-carbonyl pyridine derivative (2.0 mmol), and pipe-
ridinium acetate (15 mg, 0.10 mmol) in toluene (6 ml), was added
diethyl 1,4-dihydro-2,6-dimethyl-3,5-pyridinedicarboxylate (9)
(Hantzsch ester, 557 mg, 2.2 mmol) in one portion at room temper-
ature. The resulting mixture was degassed with a stream of nitro-
gen and then stirred at 105 °C for 3 h. After cooling to room
temperature, a minimum amount (ca 0.3-0.5 mL) of DMSO was
added to the reaction mixture and the resulting solution was di-
rectly loaded to a silica gel column and purified by column chro-
matography using Teledyne Isco Combiflash system.
CN
O
+
9 (1.1 eq)
toluene
105oC
N
NC
CN
N
2a
NH2
1e
3e
Scheme A
.
We then found that without the use of Knoevenagel condensation catalyst
piperidinium acetate, the formation of 11 was reduced but the reaction yield of
1e was hardly effected (46% without vs 44% with piperidinium acetate). In
order to further suppress the formation of 11, we used 1.4 instead of 1.0 equiv
of 3e in the absence of a catalyst, and found that the yield of 1e was improved
from 46% to 59% (Scheme A).
16. Attempts to cyclize 8a under a variety of conditions failed to provide any
cyclized product. Further exploration needs to be done.
In summary, we have developed an efficient one-pot multistep
sequence for the synthesis of 3-aminoindolizine derivatives by
using Hantzsch ester (9) as a mild hydride transfer agent. The se-