Microwave-assisted three component one-pot synthesis of pyrimido-oxazepines

Article abstract of DOI:10.1016/j.tetlet.2006.12.058

A synthetic approach toward pyrimido-oxazepine analogs was developed through the use of microwave heating. Certain analogs can be made in one step, which make this a valuable tool in the investigation of this therapeutically relevant scaffold.

Full text of DOI:10.1016/j.tetlet.2006.12.058

Tetrahedron Letters 48 (2007) 1489–1492
Microwave-assisted three component one-pot synthesis
of pyrimido-oxazepines
Colleen Hudson,a V. Srinivasa Murthy,a Kimberly G. Estepb and Gary Gustafsona,*
aArQule Inc., 19 Presidential Way, Woburn, MA 01801, United States
bPfizer Inc., Eastern Point Road, Groton, CT 06340, United States
Received 22 June 2006; revised 11 December 2006; accepted 12 December 2006
Available online 12 January 2007
Abstract—A synthetic approach toward pyrimido-oxazepine analogs was developed through the use of microwave heating. Certain
analogs can be made in one step, which make this a valuable tool in the investigation of this therapeutically relevant scaffold.
ꢀ 2007 Elsevier Ltd. All rights reserved.
It has become a widely accepted view that many classical
reactions perform better under microwave irradiation
than not.1 In attempts to synthesize pyrimido[4,5-b]-
[1,4]benzoxazepin-4-amines, it was discovered that such
compounds could be accessed in a one-pot synthesis
from 4,6-dichloropyrimidine, salicylaldehydes, and
amines through the use of microwave heating. This find-
ing is especially relevant given a series of recent publica-
tions that have revealed pyrimidio-oxazepines to be
potent inhibitors of the epidermal growth factor recep-
tor (EGFR),2,3 which is the biological target of such
cancer therapeutics as Iressa and Tarceva.4,5 Our one-
pot approach could be a valuable tool in studying the
SAR of this chemotype.
the aniline displaced the chloride to give the final prod-
uct. This pathway eliminated the need for the imine
reduction. However, the substitution pattern of the phen-
ols capable of being used in this pathway was limited
due to regioselectivity issues in the Mannich reaction.
In approach 3,6 the formation of the benzoxazepine pro-
ceeded through imine intermediate 7. In this approach,
electron deficient aldehydes worked best. Substitution
at the 6-position of the salicylaldehydes was not sup-
ported presumably due to the inhibition of imine
formation.
In trying to access the pyrimido[4,5-b][1,4]benzoxazepin-
4-amines, it was discovered that in the case of primary
and secondary aliphatic amines, the product could be
produced in one step by mixing amine, pyrimidine,
aldehyde and heating using a SmithCreator microwave
from Personal Chemistry (Scheme 2, approach A).9
Upon synthesis of 9, 2 equiv of NaBH4 were added to
the vial along with 1 mL of ethanol for solubility and
the reaction was stirred overnight to afford product 10.
Similar to previously published reports,2,6–8 a stepwise
approach using acid catalysis was needed when anilines
were coupled to the chloropyrimidine (approach B).10
The reaction could still be performed in one pot, how-
ever. After the aniline was reacted with the pyrimidine,
excess Cs2CO3 was added to quench the acid and make
the reaction basic for the addition of the aldehyde.
Imine 9 was reduced with NaBH4 as it was in the case
of approach A to the final amine 10. No transfers of
the reaction solution were necessary between the start
and the completion of the synthesis and all the com-
pounds discussed in this paper were made in less than
Published synthetic routes to the pyrimido[4,5-b][1,4]-
benzoxazepines proceeded through several path-
ways,2,6–8 all of which had limitations (Scheme 1). In
approach 1,2 anilines were reacted with dichloropyrimi-
dines to give intermediate 2. Benzoxazepine ring 3 was
formed in a condensation between the chloropyrimidine
and the salicylaldehyde. Finally, the imine was reduced
with sodium borohydride to give 4,5-dihydro compound
4. A self-acknowledged shortcoming of this pathway
was that only a narrow range of anilines performed well.
In approach 2,7 phenols were reacted with the dichloro-
pyrimidine to provide ether 5. Benzoxazepine 6 was
formed via a Mannich cyclization with the ether. Then,
Keywords: Microwave; Three component reaction; Oxazepines.
*
Corresponding author. Tel.: +1 914 347 4040x212; fax: +1 914 347
3091; e-mail: ggustafson@caratherapeutics.com
0040-4039/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.12.058

1490
C. Hudson et al. / Tetrahedron Letters 48 (2007) 1489–1492
HO
NH2
Cl
N
Cl
R2
(HCHO)n, TFA
MgSO4, CH2Cl2
H
N
R1
NH2
N
N
K2CO3
HCl, DMF
150 oC
microwave
40 oC
O
5
R2
N
O
6
R2
APPROACH 2
O
NH2
H
Cl
R1
R2 R1
R1
R1
NH2
HO
NH
NH
NH
N
N
H
N
N
O
NH2
N
NaBH4
N
N
NaH
HCl, Heat
N
1
Cl
APPROACH 1
N
N
3
Cl
O
4
R2
R2
2
O
NH2
H
R2
R1
Cl
H
N
Cl
N
HO
R2
N
N
N
NaBH4
TsOH/toluene
azeotrope
HCl, DMF
150 oC
OH
N
O
Cl
R2
microwave
8
7
APPROACH 3
Scheme 1. Previous approaches.
O
H
R2
R3
N
H
R4
HO
APPROACH A
5
Amine
N
Amine
4
3
Cs2CO3, iPrOH
150 °C
H
6
N
N
Cl
N
7
microwave
NH2
NaBH4
RT
8
N
N
O
N
1
O
10
R4
R4
O
9
N
Cl
9
NH2
1
H
R1
R1
R4
NH
N
HO
APPROACH B
NH2
Cl
N
HCl, iPrOH
Cs2CO3, iPrOH
150 °C
150 °C
microwave
microwave
2
Scheme 2. One-pot approaches.
24 h. All reactions were purified using HPLC–MS
purification.11
to be 8%.2 Using our method, 10f was obtained in a
35% yield. A separate publication6 indicates yields for
molecules of the type 10e–i were made in overall yields
ranging between 8% and 18%. We were able to make
the compounds in the yields of 30–41%. Additionally,
this approach allows access to analogs not accessible
from the current methods. Notable among these is the
ability to introduce substitution(s) at the 7-position.
Some of the existing methods have been unable to pro-
duce analogs with substitution at this position.6 The
authors suggested that steric hindrance may prevent
imine formation. However, the analogs with substitu-
tion at this position have appeared in a subsequent pub-
lication,3 although no yields were given.
The one-pot approach is fairly robust in both the variety
of amines and salicylaldehydes capable of being incor-
porated and the availability of both imines 9 and amines
10 as the final products. The examples of the products
made are shown below (Table 1). The products are
available in as little as 10 min for the imines and in less
than 24 h for the amines. This could be very valuable
especially with the iterative synthesis necessary during
the development of structurally related therapeutics.
Typical reaction times in the literature take 3–4 days
and three chromatographies. The yields shown in Table
1 are of purified compounds and they compare very
favorably to those obtained in the prior art.2,6–8 For
example, the literature reports the overall yield of 10f
It is interesting to note that the yields obtained for the
reduced products 10 were lower when aryl amines were

C. Hudson et al. / Tetrahedron Letters 48 (2007) 1489–1492
1491
Table 1. Examples of products made from one-pot method
Compound Amine
Salicylaldehyde Approach Yield11
8. Smith, L.; Piatnitski, E.; Kiselyov, A.; Ouyang, X.; Chen,
X.; Burdzovic-Wizemann, S.; Xu, Y.; Wang, Y.; Rosler,
R.; Patel, S.; Chiang, H.-H.; Milligan, D.; Columbus, J.;
Wong, W.; Doody, J.; Hadari, Y. Bioorg. Med. Chem.
Lett. 2006, 16, 1643–1646.
9a
9b
Aniline
Benzylamine R4 = H
Piperidine
Aniline
Benzylamine R4 = H
Piperidine
Aniline
3-F–aniline
3-CN–aniline R4 = 5-OMe
4-Cl–aniline
Aniline
R4 = H
B
A
A
B
A
A
B
B
B
B
B
B
38
71
63
42
98
74
41
30
35
31
31
41
9c
R4 = H
R4 = H
9. A typical procedure would be to add 400 lL of a 0.25 M
(100 lmol, 1.0 equiv) solution of 1 in anhydrous isopro-
panol to a microwave vial with stirbar followed by 100 lL
of a 1.0 M (100 lmol, 1.0 equiv) solution of benzyl amine
in anhydrous isopropanol. The vial was vortexed. 200 lL
of a 0.5 M (100 lmol, 1.0 equiv) solution of salicylalde-
hyde in anhydrous isopropanol was added followed by
100 mg (excess) of Cs2CO3. The vial was capped and then
vortexed. The reaction was heated for 10 min at 150 ꢁC
using the normal setting on the microwave to produce a
compound of type 9. If the imine was the desired product,
it proceeded to workup. If reduction of the imine to the
amine was required, the vial was uncapped and 2 equiv of
dry NaBH4 was added to the vial followed by 1 mL of
ethanol. The vial was capped and vortexed. The vial was
heated to 50 ꢁC overnight with stirring. The vial was
uncapped. After concentration in vacuo and aqueous
workup by portioning between water and DCM, the
organic layer was concentrated in vacuo. The residue was
suspended in 1 mL of DMSO and purified by HPLC. This
yielded 23 mg (74%) of 10c, a white solid. (M+H) = 321.
1H NMR (300 MHz, CDCl3): d 8.11 (s, 1H), 7.33–7.10 (br
m, 4H), 4.45 (d, J = 3.2 Hz, 2H), 4.09 (br s, 1H), 3.09 (br s,
4H), 1.64 (br m, 6H). 13C NMR (75 MHz, CDCl3): d 160,
155.8, 154.8, 147.3, 130.8, 129.6, 128.7, 124.9, 123.3, 121.4,
49.6, 48, 26.3, 24.6.
10. A typical procedure would be to add 400 lL of a 0.25 M
(100 lmol, 1.0 equiv) solution of 1 in anhydrous isopro-
panol to a microwave vial with stirbar followed by
100 lL of a 1.0 M (100 lmol, 1.0 equiv) solution of
aniline in anhydrous isopropanol. The vial was vortexed.
Concentrated HCl (5 lL) was added by pipette. The vial
was capped and vortexed. The reaction was heated for
10 min at 150 ꢁC using the normal setting on the
microwave. The vial was uncapped and 200 lL of a
0.5 M (100 lmol, 1.0 equiv) solution of 5-OMe–salicyl-
aldehyde in anhydrous isopropanol was added followed
by 100 mg (excess) of Cs2CO3. The vial was capped and
then vortexed. The reaction was heated for 10 min at
150 ꢁC using the normal setting on the microwave to
produce a compound of type 9. If the imine was the
desired product, the reaction proceeded to workup. If
reduction of the imine to the amine was required, the vial
was uncapped and 2 equiv of dry NaBH4 was added to
the vial followed by 1 mL of ethanol. The vial was
capped and vortexed. The vial was heated to 50 ꢁC
overnight with stirring. The vial was uncapped. After
concentration of the reaction in vacuo and aqueous
workup by portioning between water and DCM, the
organic layer was concentrated in vacuo. The residue was
suspended in 1 mL of DMSO and purified by HPLC.
10a
10b
10c
10d
10e
10f
10g
10h
10i
R4 = H
R4 = 5-OMe
R4 = 5-OMe
R4 = 5-OMe
R4 = 5-F
R4 = 6-OMe
Aniline
used in the syntheses. With these less reactive amines, it
was difficult to find conditions that drove the reactions
to completion without resulting in side products or
decomposition. Isopropanol was used as the reaction
solvent and produced much cleaner reaction mixtures
than the other solvents explored (DMF, DMSO, and
ACN). Since isopropanol is a modest microwave absor-
ber, it was difficult to push the temperature higher than
150 ꢁC. Side products varied according to the reagents
used but included di-addition of the amine in the first
step and reaction of the amine with the salicylaldehyde.
Optimization of the method is ongoing. Also, yields of
some of imines 9 were lower than amines 10 that would
result from their reduction. Certain compounds exhib-
ited a small amount of instability under the purification
conditions and methods are currently being optimized to
minimize decomposition.
In conclusion, the one-pot approach provides a flexible
and facile pathway to the pyrimido[4,5-b][1,4]benzox-
azepin-4-amines. By taking advantage of the speed of
microwave chemistry, the fact that no transfers of the
reaction mixture need to take place, and no workups
of the intermediates are necessary, the products
discussed here were produced in less than 24 h. Such a
rapid process should allow for an accelerated investiga-
tion of the SAR of this scaffold as a cancer therapeutic.
References and notes
1. Lange, T.; Lindell, S. Comb. Chem. High Throughput
Screening 2005, 8, 595–606.
2. Hadari, Yaron; Smith, Leon M., II. EGFR Tyrosine
Kinase Inhibitors and Therapeutic Compositions Con-
taining Them. WO 2005009384, 2005; 208 pp.
3. Pan, W.; Liu, H.; Xu, Y.-J.; Chen, X.; Kim, K.; Milligan,
D.; Columbus, J.; Hadari, Y.; Kussie, P.; Wong, W.;
Labelle, M. Bioorg. Med. Chem. Lett. 2005, 15, 5474–
5477.
4. Tamura, K.; Fukuoka, M. Expert Opin. Pharmacother.
2005, 6, 985–993.
5. Blackhall, F.; Rehman, S.; Thatcher, N. Expert Opin.
Pharmacother. 2005, 6, 995–1002.
6. Xu, Y.-J.; Liu, H.; Pan, W.; Chen, X.; Wong, W.; Labelle,
M. Tetrahedron Lett. 2005, 46, 7523–7526.
7. Duncton, M.; Smith, L., II; Burdzovic-Wizeman, S.;
Burns, A.; Liu, H.; Mao, Y.; Wong, W.; Kiselyov, A. J.
Org. Chem. 2005, 70, 9629–9631.
This yielded 12 mg (41%) of 10d,
a yellow solid.
(M+H) = 283. 1H NMR (300 MHz, CDCl3): d 8.79 (br
s, 1H), 8.01 (s, 1H), 7.46 (m, 2H), 7.26 (m, 2H), 7.11 (d,
J = 8.6 Hz, 2H), 6.71 (m, 2H) 4.37 (s, 2H), 3.75 (s, 3H).
13C NMR (75 MHz, CDCl3): d 157.3, 157.2, 148.8, 145.1,
137.4, 132.2, 129.2, 125.7, 122.4, 121.8, 115.1, 114.1, 114,
55.9, 47.5.
11. All reactions were purified using HPLC–MS purification.
Equipment: A Maccel semi-prep SH-C18 50 · 20 mm
column was used. All compounds were purified using a
Shimadzu HPLC system consisting of two LC-8A prep
pumps, LC-10ADvp pump, 2 SPD-10A UV detectors,
and an SCL-10A system controller. Two Gilson liquid

1492
C. Hudson et al. / Tetrahedron Letters 48 (2007) 1489–1492
handlers were used for injection and collection and a
A) and acetonitrile (solvent B) with 0.1% TFA and a
2.5 min run time. Molecular weights were detected using
positive electrospray ionization mode on the mass spec-
trometer with a cone voltage of 20 V. These conditions
were optimized for individual compounds using varying
acetonitrile/water gradients.
Waters ZQ mass spectrometer was used for detection. The
Masslynx version 4.0 platform controlled the injections
and tracked all data outputs.
Experimental conditions: Purification was done by RP-
HPLC with MS trigger using a gradient of water (solvent