Synthesis of pyridines and pyrido[2,3-d]pyrimidines by the Lewis acid catalysed Bohlmann-Rahtz heteroannulation reaction

Article abstract of DOI:10.1055/s-2001-17447

Lewis acids catalyse the Bohlmann-Rahtz heteroannulation reaction to generate highly functionalised pyridines from enamino esters and alkynones in a single synthetic step. Of the catalysts studied, ytterbium(III) trifluoromethanesulfonate and zinc(II) bromide are the two most efficient for the synthesis of pyridines and pyrido[2,3-d]pyrimidines, from ethyl β-aminocrotonate or 2,6-di-aminopyrimidin-4-one respectively, in up to 94% yield.

Full text of DOI:10.1055/s-2001-17447

LETTER
1523
Synthesis of Pyridines and Pyrido[2,3-d]pyrimidines by the Lewis Acid
Catalysed Bohlmann-Rahtz Heteroannulation Reaction
Synthesis of Pyridines
a
and Pyrido[
r
2,3-
d
]py
k
rimidines C. Bagley,*a James W. Dale,a David D. Hughes,a Maren Ohnesorge,a Nathan G. Phillips,a Justin Bowerb
a
Department of Chemistry, Cardiff University, PO Box 912, Cardiff, CF10 3TB, UK
b
Ribotargets Ltd., Granta Park, Abington, Cambridge, CB1 6GB, UK
Fax +44(0)2920874030; E-mail: bagleymc@cf.ac.uk
Received 6 August 2001
In order to improve the utility of the Bohlmann-Rahtz py-
ridine synthesis, and facilitate the transfer of this technol-
ogy to the solid phase, we demonstrated that this method,
which traditionally requires high cyclodehydration tem-
Abstract: Lewis acids catalyse the Bohlmann-Rahtz heteroannula-
tion reaction to generate highly functionalised pyridines from
enamino esters and alkynones in a single synthetic step. Of the cat-
alysts studied, ytterbium(III) trifluoromethanesulfonate and zinc(II)
bromide are the two most efficient for the synthesis of pyridines and peratures, may be affected in a single synthetic step for the
pyrido[2,3-d]pyrimidines, from ethyl -aminocrotonate or 2,6-di-
aminopyrimidin-4-one respectively, in up to 94% yield.
synthesis of tetrasubstituted pyridines 5 using either acetic
acid or Amberlyst 15 ion exchange resin at 50 °C
(Scheme 2).6 These modified conditions were for the most
Key words: pyridines, Bohlmann-Rahtz heteroannulations, hetero-
cycles, Lewis acids
part unsuccessful for the synthesis of pyrido[2,3-d]pyrim-
idine 7 from 2,6-diaminopyrimidin-4-one 6, although the
traditional two-step Bohlmann-Rahtz heteroannulation
The synthesis, reactions and biological properties of pyri- reaction,2 that proceeds by Michael addition in DMSO
dine-containing derivatives is an ever-expanding subject and subsequent cyclodehydration at 180 °C, proved to be
in heterocyclic chemistry. This structural motif may be a highly expedient method for the preparation of these
found in a large number of pharmaceutical agents,1 as a heterocycles (Scheme 3).7
pharmacophore of considerable historical importance,
and so new methods for their preparation have received a
great deal of attention in the literature. The synthesis of
pyridine 4 in two steps by Michael addition-cyclodehy-
dration of ethyl -aminocrotonate 1 and but-3-yn-2-one 2
was first reported by Bohlmann and Rahtz in 1957 and
proceeds via aminoheptadienone intermediate 3, that is
isolated and heated to high temperature to facilitate the
cyclisation (Scheme 1).2 Since its discovery, it is notewor-
thy that very few applications of this reaction have ap-
peared in the literature. A synthesis of the modified core
of promothiocin antibiotics has been reported,3 using the
Bohlmann-Rahtz heteroannulation reaction for the prepa-
ration of clusters of pyridine-containing heterocycles that
were elaborated to the natural product.4,5
O
R4
R4
R3
R2
R3
R2
R6
NH2
Acetic acid or amberlyst 15
50 °C, 5-24 h (65-95%)
N
R6
5
Scheme 2
O
i.
O
O
N
R1
N
R1
HN
R2
HN
H2N
N
NH2
ii. 180 °C (62-98%)
H2N
R2
6
7
Scheme 3
EtO2C
EtO2C
EtOH
50 °C
+
In order to establish a new one-step general method for the
synthesis of pyridines and pyrido[2,3-d]pyrimidines that
avoids the use of high temperatures, we have investigated
the use of Lewis acid catalysts in the Bohlmann-Rahtz
heteroannulation reaction. To test the validity of this ap-
proach, Bohlmann-Rahtz intermediate aminodienone 3
was prepared by standard procedures2 and treated with a
Lewis acid to see if cyclodehydration could be conducted
at a lower reaction temperature. The use of BF3 OEt2
failed to generate even a trace of pyridine 4, whereas
zinc(II) bromide (15 mol%) or ferric chloride (10 mol%)
in dichloromethane did promote the cyclisation. Full con-
version was achieved when a solution of aminodienone 3
and zinc(II) bromide (15 mol%) in toluene was heated at
Me
H2N
Me
Me
NH2
O
O
Me
2
3
1
EtO2C
Me
120-160 °C
N
Me
4
Scheme 1
Synlett 2001, No. 10, 28 09 2001. Article Identifier:
1437-2096,E;2001,0,10,1523,1526,ftx,en;D17401ST.pdf.
© Georg Thieme Verlag Stuttgart · New York
ISSN 0936-5214

1524
M. C. Bagley et al.
LETTER
reflux for 5 hours, to give pyridine 4 as the only product
in 59% yield (Scheme 4).
O
SiMe3
EtO2C
Me
Me3Si
9a
EtO2C
Me
Me
NH2
O
Lewis acid, CH2Cl2 or PhMe,
reflux, 18 h (26-90%)
N
Me
EtO2C
EtO2C
H2N
ZnBr2 (15 mol%)
PhMe (59%)
1
5a
Me
Me
N
Me
Scheme 5
Table 1 Lewis Acid Catalysed Synthesis of Pyridine 5a
Me
3
4
Scheme 4
Lewis acid
None
Cat. mol%
-
Solvent
Toluene
CH2Cl2
CH2Cl2
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Yielda%
0b
The success of this reaction catalysed by zinc(II) bromide
demonstrated that Lewis acids facilitate the cyclodehy-
dration at a lower reaction temperature. To examine if
these conditions were suitable for the synthesis of pyri-
do[2,3-d]pyrimidines, a solution of pyrimidine 8a or 8b
and zinc(II) bromide (20 mol%) in DMSO was stirred
overnight at 110 °C to give the heteroannulated product,
7a or 7b respectively, in quantitative yield (Figure).
ZnBr2
100
20
26b,c
29b
53
ZnBr2
ZnBr2
10
ZnBr2
15
90d
42
ZnCl2
100
20
O
O
O
ZnCl2
44
HN
HN
R
ZnI2
100
20
35
H2N
N
N
R
H2N
N
NH2
ZnI2
66
8a R = Me, 8b R = CO2Et
7a R = Me, 7b R = CO2Et
Sc(OTf)3
Tb(OTf)3
Tb(OTf)3
Yb(OTf)3
Yb(OTf)3
20
60
Figure
100
20
75
Since it was clear that this method had the potential to
generate either pyridine or pyrido[2,3-d]pyrimidine het-
erocycles in a single synthetic step, a range of different
Lewis acids was screened in a known Bohlmann-Rahtz
heteroannulation reaction.6 A solution of ethyl -amino-
crotonate 1 and 4-trimethylsilylbut-3-yn-2-one 9a in ei-
ther dichloromethane or toluene was heated at reflux
overnight in the presence of 10–100 mol% of a Lewis acid
(Scheme 5). In all of these experiments the Lewis acid
promoted the formation of pyridine 5a, catalysing both
the Michael addition and subsequent cyclodehydration.
Aminoheptadienone intermediate 3 or its trimethylsilyl
derivative was not isolated from any of these reactions,
leading to the conclusion that, in the presence of a Lewis
acid, cyclodehydration occurred spontaneously under the
reaction conditions. Furthermore it was also apparent that
the trimethylsilyl moiety was stable to these conditions as
pyridine 4 could not be detected and this was in marked
contrast to similar reactions catalysed by Amberlyst 15
ion exchange resin.6 Although the yields were variable,
for the most part, only a single identifiable product was
formed. However, out of the Lewis acids that were inves-
tigated, only zinc(II) bromide or ytterbium(III) trifluo-
romethanesulfonate (ytterbium(III) triflate) gave
consistent results, generating pyridine 5a in 90% yield in
both instances (Table 1).
45
100
20
70
90
a Isolated yield after column chromatography.
b Only unreacted starting material 1 was isolated.
c Crude yield.
d Reaction was heated at reflux for only 5 h.
esters and ethynyl ketones was submitted to our Lewis
acid catalysed heteroannulation conditions (Scheme 6).
Enamino esters, prepared by standard procedures,8 were
reacted with a number of alkynones9 in toluene in the
presence of either zinc(II) bromide or ytterbium(III) tri-
flate (Table 2).10 In general, reactions catalysed by
zinc(II) bromide were faster (complete after 5 hours as op-
posed to 24 hours) and more efficient than those conduct-
ed in the presence of ytterbium(III) triflate, with a few
notable exceptions (such as the synthesis of pyridine 5b
and 5f). In all cases the Lewis acid catalysed Bohlmann-
Rahtz heteroannulation reaction was successful to give,
following spontaneous cyclodehydration, tri- and tetra-
substituted pyridine 5a-h.11
O
R4
R4
R3
R2
R3
R2
R6
This study illustrated that Lewis acids promote the Bohl-
mann-Rahtz reaction. In order to demonstrate the utility
and versatility of this new method for the synthesis of py-
ridines and pyrido[2,3-d]pyrimidines, a range of enamino
5a-h
R6
NH2
Yb(OTf)3 or ZnBr2 (15 mol%)
PhMe, reflux, 5-18 h (32-90%)
N
Scheme 6
Synlett 2001, No. 10, 1523–1526 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Synthesis of Pyridines and Pyrido[2,3-d]pyrimidines
1525
Table 2 Heteroannulation Catalysed by Yb(OTf)3 or ZnBr2
R2
R3
R4
R6
Yb(OTf)3 Yield% ZnBr2 Yield%
5a
5b
5c
5d
5e
5f
Me
Me
Me
Me
Ph
EtO2C
EtO2C
EtO2C
EtO2C
EtO2C
EtO2C
EtO2C
EtO2C
Me3Si
Et
Me
90a
83
55
33
32
68
44
58
90
67
85b
44c
72
62
65
70
Me
Ph
CO2Et
CO2Et
Me
Me3Si
Et
Ph
Ph
Me
5g
5h
Ph
Ph
CO2Et
Me
Ph
H
a 20 mol% catalyst was used.
b 10 mol% catalyst was used. c Desilylated pyridine (R4 = H) was also isolated.
Following the success of the zinc(II) bromide catalyst,
compared with previous results,6 a solution of 2,6-diami-
nopyrimidin-4-one 6 and either 4-trimethylsilylbut-3-yn-
2-one 9a or ethyl 2-oxo-4-(trimethylsilyl)but-3-ynoate 9b
in DMSO was stirred at 110 °C in the presence of either
zinc(II) bromide (20 mol%) or ytterbium(III) triflate (20
mol%). After 72 hours and standard work up12 the zinc(II)
catalysed reactions gave pure pyrido[2,3-d]pyrimidine
7a,b in 92% and 89% yield, respectively, whereas the yt-
terbium(III) catalysed process gave a slightly impure
product 7a,b in 94% and 91% respective yield
(Scheme 7). In contrast to the behaviour of pyrimidinone
6 in the uncatalysed Bohlmann-Rahtz heteroannulation
reaction,7 in the presence of a Lewis acid Michael addition
and cyclodehydration were facilitated in a single synthetic
step in excellent yield at a lower reaction temperature. No-
tably, in this case, desilylation occurred in the course of
the reaction in an analogous fashion to the two-step syn-
thesis of pyrido[2,3-d]pyrimidine 7.
Acknowledgements
We thank the E.P.S.R.C. (CNA to James), B.B.S.R.C. (grant to
DDH) and Royal Society for support of this work and the
E.P.S.R.C. Mass Spectrometry Service, Swansea for high resoluti-
on spectra.
References and Notes
(1) Roth, H. J.; Kleemann, A. Pharmaceutical Chemistry.
Volume 1: Drug Synthesis; John Wiley & Sons: New York,
1988.
(2) Bohlmann, F.; Rahtz, D. Chem. Ber. 1957, 90, 2265.
(3) Moody, C. J.; Bagley, M. C. Synlett 1998, 361.
(4) Moody, C. J.; Bagley, M. C. Chem. Commun. 1998, 2049.
(5) Bagley, M. C.; Bashford, K. E.; Hesketh, C. L.; Moody, C.
J. J. Am. Chem. Soc. 2000, 122, 3301.
(6) Bagley, M. C.; Dale, J. W.; Bower, J. Synlett 2001, 1149.
(7) Bagley, M. C.; Hughes, D. D.; Lloyd, R.; Powers, V. E. C.
Tetrahedron Lett. 2001, 6585.
(8) Prepared by reaction of -ketoester with ammonium acetate
in toluene-acetic acid according to the procedure mentioned
in: Baraldi, P. G.; Simoni, D.; Manfredini, S. Synthesis 1983,
902.
(9) 4-Phenylbut-3-yn-2-one, 4-(trimethylsilyl)but-3-yn-2-one
and hex-3-yn-2-one were commercially available. Ethyl 2-
oxo-4-phenylbut-3-ynoate and ethyl 2-oxo-4-(trimethyl-
silyl)but-3-ynoate were prepared by the addition of the
corresponding lithium acetylide to the Weinreb amide, see:
Chiu, C. C.; Jordan, F. J. Org. Chem. 1994, 59, 5763.
(10) In a typical experimental procedure, a solution of ethyl -
aminocrotonate (0.13 g, 1.0 mmol), 4-(trimethylsilyl)but-3-
yn-2-one (0.15 g, 1.1 mmol) and ytterbium(III)
O
O
O
R
9a R = Me
9b R = CO2Et
Me3Si
HN
HN
H2N
N
NH2
H2N
N
N
R
Yb(OTf)3 or ZnBr2 (20 mol%)
DMSO, 110 °C , 18 h (89-94%)
6
7a R = Me,
7b R = CO2Et
Scheme 7
In summary, the Lewis acid catalysed Bohlmann-Rahtz
heteroannulation reaction was successful in both reducing
the temperature of this transformation and, more impor-
tantly, affecting the synthesis in a single preparative step
using a single experimental procedure for both the synthe-
sis of pyridines and pyrido[2,3-d]pyrimidines. Work is
now underway to transfer this technology to the solid
phase and apply our new general procedure to the synthe-
sis of a number of biologically active heterocycles.
trifluoromethanesulfonate (0.12 g, 0.2 mmol) in toluene (6
mL) was heated at reflux for 5 hours. Water (6 mL) was
added, the solution was heated at reflux for 20 minutes and
allowed to cool. The aqueous phase was separated and
extracted with ethyl acetate (2 10 mL) and the organic
extracts were combined, washed with brine, dried (MgSO4)
and evaporated in vacuo. The residue was purified by flash
chromatography on silica, eluting with light petroleum-ethyl
acetate (3:1), to give ethyl 2,6-dimethyl-4-
(trimethylsilyl)pyridine-3-carboxylate (0.23 g, 90%) as a
colourless oil.
Synlett 2001, No. 10, 1523–1526 ISSN 0936-5214 © Thieme Stuttgart · New York

1526
M. C. Bagley et al.
LETTER
(11) All compounds exhibited spectroscopic properties that
agreed with literature data according to reference 6.
(12) In a typical experimental procedure, a solution of 4-
(trimethylsilyl)but-3-yn-2-one 9a (0.17 g, 1.2 mmol), 2,4-
diamino-6-hydroxypyrimidine 6 (0.16 g, 1.2 mmol) and
zinc(II) bromide (56 mg, 0.24 mmol) in DMSO (5 mL) was
stirred at 110 °C for 72 h. H2O (30 mL) was added and the
precipitated solid was filtered, washed with water and dried
to give 2-amino-7-methyl-5-trimethylsilyl-3H-pyrido[2,3-
d]pyrimidin-4-one 7a (0.20 g, 92%) as a pale yellow solid;
spectroscopic properties agreed with literature data
according to reference 7.
Synlett 2001, No. 10, 1523–1526 ISSN 0936-5214 © Thieme Stuttgart · New York