Improved synthesis of functionalized 5,6-dihydro-pyrimido[4,5-b][1,4] oxazepines

Article abstract of DOI:10.1016/j.tetlet.2005.08.161

Novel synthetic approaches toward 5,6-dihydro-pyrimido[4,5-b][1,4] oxazepines were reported that led to successful introduction of poorly reactive anilines and various 2-hydroxybenzaldehydes to this therapeutically relevant scaffold. More extensive SAR st

Full text of DOI:10.1016/j.tetlet.2005.08.161

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 7523–7526
Improved synthesis of functionalized
5,6-dihydro-pyrimido[4,5-b][1,4]oxazepines
Yong-Jiang Xu, Hu Liu,* Weitao Pan, Xin Chen, Wai C. Wong and Marc Labelle
Department of Chemistry, ImClone Systems Incorporated, 710 Parkside Avenue, Brooklyn, NY 11226, USA
Received 11 July 2005; revised 31 August 2005; accepted 31 August 2005
Available online 15 September 2005
Abstract—Novel synthetic approaches toward 5,6-dihydro-pyrimido[4,5-b][1,4]oxazepines were reported that led to successful intro-
duction of poorly reactive anilines and various 2-hydroxybenzaldehydes to this therapeutically relevant scaffold. More extensive
SAR studies on this scaffold hence became possible.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Recently, the 5,6-dihydro-pyrimido[4,5-b][1,4]oxaze-
pines 5 were described by us as potent EGFR inhibi-
tors.1 Additional attractive features of this scaffold
also include its good PK profile. Prior to our work in
this field, there was only one reported synthesis of the
5,6-dihydro-pyrimido[4,5-b][1,4]oxazepines,2 and that
route was adopted by us for the initial SAR work on
the oxazepines (Scheme 1, approach A). Although it
could be applied to a number of aldehydes (3), the
synthesis was not convergent enough to allow facile
variation of the anilines. This limitation was soon over-
come by a Mannich cyclization approach developed by
us (Scheme 1, approach B).3 Nevertheless, this approach
worked only with phenol 6 that bore electron donating
groups, and certain restrictions on substitution pattern
of 6 were also applied. Moreover, both approaches
required reasonable nucleophilicity of the anilines to
ensure the thermal or microwave assisted chloride
O
H
Cl
R1
R1
R2
R1
NH2
Cl
R1
NH
N
NH
HO
N
NH2
NH
N
N
O
H
N
NH2
Cl
NaBH4
3
N
N
N
heat
Approach A
N
N
NaH
1
R2
O
2
R2
5
4
R1
heat
NH2
R2
Cl
N
H
N
Cl
HO
NH2
TFA, MgSO4
formaldehyde
N
6
N
O
N
O
R2
R2
Approach B
7
8
Scheme 1.
Keywords: Oxazepines; Reductive amination.
*
Corresponding author. Tel.: +1 917 755 5856; fax: +1 718 633 0856; e-mail: huliu205@yahoo.com
0040-4039/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.08.161

7524
Y.-J. Xu et al. / Tetrahedron Letters 46 (2005) 7523–7526
Table 1. Formation of 8/9 using reductive amination conditions
O
H
R2
H
N
1
R2
1
1
NH2
9
N
HO
NaBH4
N
N
Cl
OH
3
N
Cl
8
3
Cl
N
O
3
N
Cl
3
TsOH
toluene
azeotrope
R2
N
Cl
7
6
1
(1-Cl)
8 (1-Cl)
9 (3-Cl)
10 (3-Cl)
11
Compound
Cl
R2
Conditions
% Yielda
9a
8a
8b
8c
8d
8e
8e
8f
3-Cl
1-Cl
1-Cl
1-Cl
1-Cl
1-Cl
1-Cl
1-Cl
1-Cl
1-Cl
1-Cl
8-NO2
8-NO2
H
6-OH
6-OMe
7-OH
7-OH
7-OMe
8-OMe
8-F
Ab
A
Bc
B
B
A
B
B
B
B
B
40
80
33
34
36
0
8
25
55
56
0
8g
8h
8i
9-OMe
a Isolated yield.
b Conditions A, detailed procedure in Ref. 5.
c Conditions B, detailed procedure in Ref. 6.
displacement in either 1 or 8, and synthesis of 5 with het-
eroaryl amines remained unexplored.
tron deficient aldehydes such as 2-hydroxy-5-nitro-benz-
aldehyde, 8a was obtained in a 80% yield when the imine
formation was carried out in toluene under normal azeo-
tropic conditions (conditions A).5 On the other hand,
aldehydes with methoxy or hydroxyl substitutes
required several rounds of azeotropic processes to
maximize the imine formation (conditions B).6 It is
noteworthy to point out that 8e, which arose from the
reaction of a highly electron enriched ortho- and para-
dihydroxy substituted benzaldehyde, was obtained in a
8% yield using conditions B. However, 8e was previously
not obtainable by using conditions A. Neither condi-
tions A nor B produced 8i. We assumed that the steric
hindrance of the aldehyde moiety prevented the imine
formation completely.
Our interests in developing the benzoxazepines as kinase
inhibitors necessitated a more general preparation of
intermediate 8 and its 3-Cl regioisomer 9 (Table 1).4 Pre-
vious attempts to make the oxazepine 8/9 via approach
A (Scheme 1) all failed. Furthermore, the overall substi-
tution patterns of R2, for example, hydroxy or nitro,
excluded the possibilities of using the Mannich type
approach (Scheme 1, approach B) due to either regio-
selectivity issue during the addition step (1 to 7, Scheme
1), or the nitroÕs deactivating effect on the cyclization
(e.g., 7 to 8, Scheme 1). We hence envisioned a new syn-
thetic route to those compounds via reductive amination
reactions between amines 1/10, and aldehydes 3.
As an example of further chemical elaboration, the nitro
in 9a was reduced with indium in the presence of ammo-
nium chloride to afford aniline 12 (Scheme 2).7 The chlo-
ride in 12 was then displaced by a morpholine under
microwave conditions to provide 13, that served as a
useful template for the preparation of therapeutically
relevant molecules.
In the present study, imine 11 was first formed under
acid catalyzed condition in toluene with azeotropic re-
moval of water and a brief survey of the acids showed
no difference among p-toluene sulfonic acid (p-TsOH),
10-camphor sulfonic acid, and oxalic acid. The formed
imines were then reduced with sodium borohydride fol-
lowed by cyclization to afford 8/9 after standard work-
up. The yield of the reaction was mainly determined
by the extent of the imine formation, which in turn
was largely affected by the electronic and steric property
of the aldehyde functionality on 3. As expected, for elec-
With 8 in hand, we then shifted our attention to the dis-
placement of the chloride in 8 with amine 14 to give 15
(Table 2). Since initial attempts to generate 15 by direct
substitution under thermal or microwave conditions all
O
H
HN
HN
N
N
H
In/NH4Cl
NO2
N
NH2
N
N
O
NH2
80 oC
(55%)
N
Microwave,
15 minutes
(50%)
O
N
N
N
O
Cl
Cl
O
9a
12
13
Scheme 2.

Y.-J. Xu et al. / Tetrahedron Letters 46 (2005) 7523–7526
Table 2. Formation of 15 via palladium catalyzed coupling
7525
N
N
R1
4'
4'
R1
NH2
H
N
HN
O
N
Cl
N
H
N
O
N
N
9
NH
H
N
14a: R1 = H
14b: R1 = 4'-F
8
O
N
9
R2
6
7
Pd(dba)2, BINAP, NaOtBu
8
N
O
R2
7
6
8
15
Compound
R1
R2
% Yielda
15a
15b
15c
15d
15e
15f
15g
15h
15i
H
H
H
6-OH
6-OMe
7-OMe
7-OMe
8-OMe
8-OMe
8-F
36
32
7
27
53
14
35
22
14
40-F
H
H
H
40-F
H
40-F
H
a Isolated yield.
failed, the palladium catalyzed coupling conditions8 be-
came the method of choice. The optimized conditions
utilized a catalyst/ligand combination of Pd(dba)2 and
rac-BINAP, and sodium tert-butoxide as the base.9
The isolated yields of 15, though modest and variable,
were partially due to chromatographic purification fol-
lowed by recrystallization to ensure purity for biological
testing.
tial displacement of the chlorides on 16 by hydroxyl
aldehyde 3f, followed by 14a, afforded crude 18. A
one-pot reduction of the nitro group on 18, followed
by triethyl orthoformate promoted cyclization, gave
imine 20 in a 51% overall yield. Standard reduction of
20 with NaCNBH3 gave 15e, with its spectroscopic
properties identical to those obtained previously.
In conclusion, the synthetic methods reported herein led
to the successful introduction of poorly reactive anilines
and various 2-hydroxybenzaldehydes to the 5,6-dihydro-
pyrimido[4,5-b][1,4]oxazepines scaffold, whereby allow-
ing a more extensive SAR study on this scaffold. The
biological results for some of the compounds prepared
in this study will be reported in due course.
In order to get access to such imines as 20, a further
modified synthesis of the oxazepines was developed that
exploited the high reactivity of 4,6-dichloro-5-nitro-
pyrimidine 16 (Scheme 3).10 Its synthetic utility was
demonstrated by using 2-hydroxy-4-methoxy-benzalde-
hyde 3f as a standard substrate. As expected, a sequen-
Ph
H
N
Cl
N
O
N
O
O
Cl
O
N
N
NO2
N
N
Ph
N
N
O
N
Cl
NO2
NH NO2
H
MeO
16
O
N
O
NH2
HO
OMe
N
iPr2NEt, DMF
14a
3f
18
MeO
17
H2/Pd(C), DME
O
H
H
N
N
Ph
Ph
N
N
Ph
NH2
NH
N
N
N
H
HC(OEt)3
O
N
N
NaCNBH3
(66%)
O
N
NH
O
O
NH
H
N
N
O
N
N
N
N
N
MeO
O
OMe
OMe
20
(51% from 3f)
15e
19
Scheme 3.

7526
Y.-J. Xu et al. / Tetrahedron Letters 46 (2005) 7523–7526
Y.; Yamada, K.; Chida, N. Org. Lett. 2000, 2, 1137; (d)
References and notes
Harwood, E. A.; Sigurdsson, S. T.; Edfeldt, N. B. F.;
Reid, B. R.; Hopkins, P. B. J. Am. Chem. Soc. 1999, 121,
5081.
1. Smith, L. M., II; Hadari, Y. Patent WO 2005009384, 200.
Chem. Abstr. 2005, 142, 172179.
9. Compound 15e: In a sealed tube were placed sodium
t-butoxide (0.28 mmol, 0.027 g), bis(dibenzylideneacetone)
palladium (0.07 mmol, 0.041 g), and rac-(2,20-bis(diphen-
ylphosphino))-1,10-binaphthyl (0.21 mmol, 0.131 g). The
tube was then purged with argon. Then, 8f (0.2 mmol,
0.053 g) in anhydrous 1,4-dioxane (1 mL) was added,
followed by 14a (0.2 mmol, 0.043 g) in 1,4-dioxane (1 mL).
The tube was then sealed and the reaction was stirred
at 75 ꢁC for 10 h, then cooled down to rt. Dichlorometh-
ane (5 mL) was added and the mixture was stirred for
another 20 min before filtration of the mixture. The
insoluble was washed with methanol (5 mL) followed by
acetone (4 mL), and the filtrate was concentrated to give a
residue that was purified with TLC (MeOH/CH2Cl2: 5/
100) to give a green solid, which upon recrystallization
from dichloromethane gave 15e as a white solid (0.046 g,
53%). 1H NMR (300 MHz, DMSO-d6): d 3.77 (s, 3H), 4.40
(d, J = 4.2 Hz, 2H), 5.35 (t, J = 4.2 Hz, 1H), 6.73–6.80 (m,
2H), 7.30 (d, J = 8.4 Hz, 1H), 7.48–7.62 (m, 3H), 7.92 (s,
1H), 7.95–7.98 (m, 2H), 8.72 (br s, 1H), 8.96 (s, 2H), 10.91
(br s, 1H). 13C NMR (300 MHz, DMSO-d6): d 45.7, 53.2,
102.5, 104.8, 112.0, 126.3, 127.6, 129.8, 130.9, 131.3, 132.5,
133.2, 143.8, 144.3, 153.2, 156.1, 157.3, 158.1, 160.7, 163.2.
LC–MS (M+1): 442. Anal. Calcd for C23H19N7O3: C,
62.58; H, 4.34; N, 22.21. Found: C, 62.33; H, 4.10; N,
22.29.
10. Compound 15e (Scheme 3): In a 15 mL round bottomed
flask were placed 3f (152 mg, 1 mmol), and 16 (193 mg,
1 mmol) in DMF (2 mL). The solution was cooled to 0 ꢁC,
and diisopropylethylamine (0.35 mL, 2.0 mmol) was
added dropwise. The reaction was stirred at 0 ꢁC for 1 h,
warmed to rt for 14 h. The reaction was then cooled to
0 ꢁC, 14a (214 mg, 1 mmol) in 2 mL of DMF was added
slowly, followed by diisopropylethylamine (0.35 mL,
2.0 mmol). The mixture was stirred at 0 ꢁC for 1 h, at rt
for additional 3 h, then quenched with water (5 mL). The
precipitation formed was filtered, washed with ethyl ether
(2 · 10 mL), and dried to afford 366 mg of crude 18.
In a 100 mL round bottomed flask equipped with a
hydrogen balloon were placed crude 18 (366 mg) and Pd/C
(20 mg) in ethylene glycol dimethyl ether (DME, 10 mL).
Hydrogen was then connected and the reaction was stirred
at rt for 12 h. The hydrogen was disconnected, and triethyl
orthoformate (5 mL) was added to the reaction. The
mixture was stirred for 5 h and then filtered. The filtrate
was concentrated under vacuum and the residue was
purified by preparative TLC (5% MeOH/CH2Cl2) to
afford 20 (226 mg, 51% from 3f) as a solid. 1H NMR
(300 MHz, DMSO-d6): d 10.96 (s, 1H), 9.82 (s, 1H), 9.11
(s, 2H), 8.73 (s, 1H), 8.36 (s, 1H), 8.10–7.96 (m, 3H), 7.62–
7.50 (m, 4H), 6.97 (d, 1H, J = 6.3 Hz), 3.92 (s, 3H); LC–
MS (M+1): 439.88. 13C NMR (300 MHz, DMSO-d6): d
55.2, 101.6, 107.1, 115.1, 122.2, 126.3, 129.3, 130.8, 131.3,
132.9, 133.6, 142.6, 151.4, 152.8, 158.7, 160.7, 161.8, 163.2,
164.3, 166.0. Anal. Calcd for C23H17N7O3: C, 62.87; H,
3.90; N, 22.31. Found: C, 62.97; H, 4.12; N, 22.61.
In a round bottomed flask was placed 20 (70 mg,
0.16 mmol) in 4 mL of THF. Sodium cyanoborohydride
(30 mg, 0.48 mmol) was added in one portion. The
reaction was stirred at rt for 1 h, cooled to 0 ꢁC, and
quenched slowly with aq HCl (3 M, 3 mL). Standard
workup followed by preparative TLC (5% MeOH/
CH2Cl2) afforded 15e (46 mg, 66%) as a white solid with
spectroscopic properties identical to those obtained from
the reductive amination procedure.
2. Levkovskaya, L. G.; Sazonov, N. V.; Grineva, N. A.;
Mamaeva, I. E.; Serochkina, L. A.; Safonova, T. S.
Khimiya Geterotsiklicheskikh Soedinenii 1985, 1, 122.
3. Duncton, M. A. J.; Smith, L. M., II; Burdzovic-Wizeman,
S.; Burns, A.; Liu, H.; Mao, Y. J. Org. Chem., in
press.
4. For biological results on compounds prepared in this
study, please refer to Pan, W.; Liu, H.; Xu, Y.-J.; Chen,
X.; Milligan, D. L.; Columbus, J.; Balagtas, C.; Hadari,
Y.; Wong, W. C.; Kim, K. H.; Labelle, M. Bioorg. Med.
Chem. Lett., in press.
5. Compound 8a (conditions A): In a 200 mL round-
bottomed flask equipped with a Dean–Stark trap and a
condenser were placed 5-amino-2,4-dichloropyrimidine
(20 mmol,
3.26 g),
2-hydroxy-5-nitrobenzylaldehyde
(20.5 mmol, 3.50 g) and p-toluenesulfonic acid (40 mg) in
100 mL anhydrous toluene. The mixture was refluxed till
no water was brought to Dean–Stark trap (ca. 100 min).
The toluene was removed and the solid obtained was
collected and dried in vacuum oven.Reduction of the
crude imine (0.77 g, 2.5 mmol) similarly to that described
in condition B6 gave 8a as yellowish solid (0.55 g, 80%). 1H
NMR (300 MHz, DMSO-d6): d 4.71 (d, J = 4.0 Hz, 2H),
6.12 (br s, 1H), 7.52 (m, 1H), 8.19 (s, 1H), 8.29 (m, 2H).
13C NMR (300 MHz, DMSO-d6): d 42.4, 116.8, 120.3,
121.0, 126.8, 139.9, 143.8, 142.2, 148.3, 152.0, 159.1. LC–
MS (M+1): 279. Anal. Calcd for C11H7ClN4O3: C, 47.41;
H, 2.53; N, 20.11. Found: C, 47.22; H, 2.33; N, 20.38.
6. Compound 8g (conditions B): In a 100 mL round-
bottomed flask equipped with a Dean–Stark trap and a
condenser were placed 5-amino-4,6-dichloropyrimidine
(10 mmol, 1.63 g), 2-hydroxy-5-methoxylbenzylaldehyde
(10 mmol, 1.52 g) and 10-camphorsulfonic acid (20 mg)
in 30 mL anhydrous toluene. The mixture was refluxed
under argon for 30 min, then the toluene collected in the
Dean–Stark trap was released. The toluene collecting and
releasing process was repeated until about a total of 20 mL
of toluene was removed from the reaction mixture. Then,
another 20 mL of anhydrous toluene was added to the
reaction mixture and the azeotrope continued. The
removal-addition procedure was repeated three more
times. Then, the toluene was removed under vacuum to
give the crude imine, which was dissolved in 1,4-dioxane
(40 mL). To that solution was added sodium borohydride
(10 mmol, 0.40 g), and the mixture was stirred at 44 ꢁC
overnight under argon. The reaction was then cooled
down, and ethyl acetate (40 mL) and brine (40 mL) were
added. The layers were separated, and the aqueous layer
was extracted with ethyl acetate (40 mL). The combined
organic layers were dried over MgSO4, filtered, and
concentrated under vacuum to give the crude product.
The residue was washed with ether (2 · 15 mL) to give the
product 8g as a yellowish solid (1.43 g, 55%). 1H NMR
(300 MHz, DMSO-d6): d 3.75 (s, 3H), 4.47 (d, J = 4.2 Hz,
2H), 6.37 (t, J = 4.2 Hz, 1H), 6.87–6.96 (m, 2H), 7.18 (d,
J = 8.7 Hz, 1H), 7.97 (s, 1H). 13C NMR (300 MHz,
DMSO-d6): d 43.7, 51.2, 113.0, 113.2, 126.7, 133.4,
135.2, 140.2, 142.3, 148.2, 151.7, 159.8. LC–MS (M+1):
264. Anal. Calcd for C12H10ClN3O2: C, 54.66; H, 3.82; N,
15.94. Found: C, 54.36; H, 4.02; N, 15.64.
7. Banik, B. K.; Nanik, I.; Becker, F. F. Org. Synth. 2004, 81,
188.
8. (a) Muci, A. R.; Buchwald, S. L. Top. Curr. Chem. 2002,
219, 132; (b) Hartwig, J. F. Angew. Chem., Int. Ed. 1998,
37, 2046; (c) Suzuki, T.; Tanaka, S.; Yamada, I.; Koashi,