Reinvestigation of alternative method for the preparation of dibenz[b,f][1,4]oxazepine

Article abstract of DOI:10.1002/jhet.59

(Chemical Equation Presented) Two distinct alternative methods using different starting materials for the preparation of dibenz[b,f] [1,4]oxazepine (7, CR) were reinvestigated. The possibility of trans-cis conversion of the Schiff base 5 (produced from 1 and 2) is considered the favorable orientation that leads to cyclization (production of 7). Fluoro derivative of 1 afforded excellent and more convenient conditions for the one pot preparation of high yield pure 7. The presence of only trans configuration for the imine 6 (produced from 3 with 4) and the impossibility of its conversion to cis, makes it inadequate for the preparation of 7.

Full text of DOI:10.1002/jhet.59

988
Reinvestigation of Alternative Method for the
Preparation of Dibenz[b,f][1,4]oxazepine
Vol 46
Hossein Fakhraian* and Yaser Nafary
Department of Chemistry, Imam Hossein University, Tehran, Iran
*E-mail: fakhraian@yahoo.com
Received December 4, 2007
DOI 10.1002/jhet.59
Published online 25 August 2009 in Wiley InterScience (www.interscience.wiley.com).
Two distinct alternative methods using different starting materials for the preparation of dibenz[b,f]
[1,4]oxazepine (7, CR) were reinvestigated. The possibility of trans-cis conversion of the Schiff base 5
(produced from 1 and 2) is considered the favorable orientation that leads to cyclization (production of 7).
Fluoro derivative of 1 afforded excellent and more convenient conditions for the one pot preparation of
high yield pure 7. The presence of only trans configuration for the imine 6 (produced from 3 with 4) and
the impossibility of its conversion to cis, makes it inadequate for the preparation of 7.
J. Heterocyclic Chem., 46, 988 (2009).
INTRODUCTION
icyl-aldehyde with 2-nitrochlorobenzene (Scheme 2).
We have recently reported the effect of temperature and
catalyst on the yields and rate of the etherification reac-
tion between sodium salt of salicylaldehyde and 2-nitro-
chlorobenzene and have emphasized the harsh condi-
tions necessary for the preparation of 7 by this method
[22]. We have observed that the etherification reaction
of 2-nitrophenol and 2-chlorobenzaldehyde does not
occur under the same conditions and requires more diffi-
cult circumstances to be performed. This can be ration-
alized by the electrophilic effect of NO2 group, which
reduces the nucleophilic character of the hydroxy oxy-
gen atom in 2-nitrophenol.
Dibenz[b,f][1,4]oxazepine (7, CR) is an incapacitating
lachrymatory agent [1–6] and is an intermediate in the
synthesis of loxapine (antipsychotic), amoxapine (anti-
depressant) [7], and others substituted dibenzoxazepine
compounds that are analgesic or useful for the treatment
of different diseases [8–11].
Moreover, reduction of the imine group of dibenz-
[b,f][1,4]oxazepine and/or the oxidation of the same
group to the imid afforded intermediates in the synthesis
of several other compounds effective in the prevention
and treatment of circulatory disease and osteoporosis
[9,11].
The formation of 7 via 1 and 2 (Scheme 1) has been
reported recently [19,20], and the kinetic data have been
investigated [17,18]. Other attempts to prepare 7 via
cyclization of 2-[(2-fluorobenzylidene)amino] phenol
(fluoro derivative of 5) in ethanol with excess triethyl-
amine were unsuccessful even after prolonged reflux. In
contrast, the preparation of 1,2,3,4-tetrafluorobenzo[b,f]
[1,4]oxazepine (10) by cyclization of 2-[(2,3,4,5,6-penta-
flurobenzylidene)amino] phenol (8) (Scheme 3) using
the same conditions was successful, whereas the
attempts to cyclize 9 into 11 failed [23].
There are different methods for the preparation of 7
[12–24]. Two pathways starting from different raw
materials can be considered for the preparation of this
compound (Schemes 1 and 2).
These methods are differentiated in the occurrence of
either the imine reaction or etherification in the first
step. Thus, in Scheme 1, dibenz[b,f][1,4]oxazepine is
formed by the imine reaction followed by the etherifica-
tion (cyclization), whereas in Scheme 2, etherification is
followed by the imine reaction (cyclization). The two
etherification and imine reactions each have limitations
that influence the yield of the final product.
The difference in reactivity between 8 and 9 has been
justified by the necessity for the electrophilic arene to
be a benzylidene and not a phenyl imine, that is, the
Firstly, the method proposed by Tambute [12,13] for
the preparation of 7 involves the condensation of sal-
C
V
2009 HeteroCorporation

September 2009
Reinvestigation of Alternative Method for the Preparation of
Dibenz[b,f][1,4]oxazepine
989
Scheme 1
electron-withdrawing ACH¼¼N substituent enhances the
susceptibility of C6F5 to nucleophilic attack, whereas
the electron-donating AN¼¼CH substituent has the oppo-
site effect [23].
tal Section). It is noteworthy that the X-ray crystallogra-
phy of 5 indicates a unique trans configuration [25].
The retention time in GC and GC-MS, and fragmenta-
tion pattern of 5trans and 6trans were nearly similar else
the more abundant peak (m/z (100%); 120 for 5trans and
196 for 6trans). The occurrence of 5cis increases at high
temperature (> 100ꢀC) and with a catalytic amount of
acid (H2SO4 98%).
The last report on the synthesis of 7 was consisted in
the reaction of 2-fluorobenzaldehyde (fluoro derivative
of 1) with 2-aminophenol using K2CO3 in polyethylene
glycol (PEG-400) at 100ꢀC, affording 89% of 7 after
8 h [24].
The effect of solvent and UV irradiation in the trans-
cis transformation were also investigated, but complete
conversion of 5trans to 5cis was never observed. In con-
trast, 6trans was not converted to 6cis under any condi-
tions. The target product (7) was never obtained by
cyclization of imine 6. Thus, we presumed that apart
from electronic parameters, the prevention of trans-cis
conversion in 6 is another reason that the target product
(7) was not obtained by cyclization of imine 6. In fact,
trans-cis isomerization can occur through the hemiami-
nal resulting from attack of the phenolic oxygen to the
imine carbon, which is much easier in 5 (five-membered
ring) than in 6 (four-membered ring).
The preparation of 7 via 1 and 2, performed following
several procedures, afforded different yields (Scheme 4).
Reaction of 1 with 2 in DMSO at 150ꢀC (the proce-
dure B1) afforded 7 in 25% yield. Following procedure
B3, the mixture of 1, 2, and KOH in CCl4 are refluxed
for 5 h and then after evaporation of solvent, DMSO
was added, and the mixture was refluxed for 6 h to
afford 50% of 7.
This contribution discussed the feasibility of two dis-
tinct pathway methods (Scheme 1) using different start-
ing materials and passing by different intermediates for
the preparation of dibenz[b,f][1,4]oxazepine (7, CR).
RESULTS AND DISCUSSION
The preparation of 7 via 2-chlorobenzaldehyde (1)
and 2-aminophenol (2) or salicylaldehyde (3) and 2-
chloro-aniline (4) was reinvestigated (Scheme 1). Imine
formation following both methods was nearly quantita-
tive, and the water released in the course of the reaction
did interfere with imine formation.
1
We have reported the results of GC, GC-MS, and H
NMR analysis of the imine products [25], according to
which we have presumed the formation of 5cis and 5trans
and only 6trans in solution.
The 13C NMR spectrum 5cis and 5trans were similar
and their mixture presented 13 resonances for carbon,
but the chemical shift of imine proton (in 1H NMR
spectrum) was different. However, their MS fragmenta-
tion patterns were completely different (see Experimen-
According to the procedure B2, when we used xylene
as solvent, 5cis and 5trans were formed with proportional
amounts of 12 and 88%, respectively. Heating the
Scheme 2
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

990
H. Fakhraian and Y. Nafary
Vol 46
Scheme 3
The time of reaction in the second step determined
the yield of the final product. In fact, we have found
that compound 7 is very temperature sensitive and at
temperature higher than 100ꢀC, the target compound
decomposed and the yield decreased. Thus the time of
reaction and the temperature in the second step should
be controlled carefully.
The workup of the reaction mixture to obtain pure 7
was very simplified following this procedure. Effec-
tively, the extraction with ether and several washing of
organic phase with water, drying over CaCl2 and strip-
ing of solvent afforded very pure product without need
of further purification procedure.
solution at 120ꢀC for 24 h changed the proportional
amount of the two isomers in favor of cis, which
afforded 75% of 7 after addition of Na and refluxing the
reaction mixture. Without heating the solution, the
attend product (7) was not formed. While when we used
DMSO as solvent, prior conversion of trans to cis was
not indispensable and the salt of 5 rearrange for the pro-
duction of 7.
Preparation of sodium salt of 2 and its subsequent
reaction with 1 in DMSO at 150ꢀC for 6 h (the proce-
dure C) has yielded 47% of 7.
The reaction conditions following the procedure A are
more convenient (the time of reaction and the tempera-
ture), and pure products were formed after the first and
second steps. In addition to good yield (70%), the for-
mation of 7 following this method can be performed via
a one-pot procedure (procedure B).
The reaction of 1 with 2 conducted under same condi-
tions with PEG(300) has not afforded 7. However, the
usage of catalytic amount of KF (10 mol%) in the sec-
ond step (after formation of imine in the first step)
favors the formation of 7 (50%) but other experiments
should be performed to optimize this method.
In summary, two distinct alternative methods (Scheme
1) using different starting materials for the preparation
of dibenz[b,f][1,4]oxazepine (7) were reinvestigated.
The imine reaction, according to these methods, occurs
readily and yields of Schiff bases (5, its fluoro deriva-
tive, and 6) are nearly quantitative. The possibility of
trans-cis conversion of the Schiff base 5 were consid-
ered as favorable conditions that caused its cyclization
(production of 7). This reaction was performed follow-
ing different procedures and various yields are obtained.
The pathway A afforded more convenient and soft con-
ditions for the production of pure 7 and its intermedi-
ates. Fluoro derivative of 1 afforded excellent, more
convenient and soft conditions for the one-pot prepara-
tion of target compound (7). The presence of only trans
configuration for the imine 6 (produced from 3 with 4)
Our last tentative in the preparation of 7 was based
on the usage of 2-fluorobenzaldehyde (a fluoro deriva-
tive of 1) in conjunction with polyethylene glycol as
was outlined recently [24]. 2-fluorobenzaldehyde is
more expensive that 1 (7 fold) but the procedure offered
soft conditions to obtain high-yield pure product. Simul-
taneous mixing of 2-fluorobenzaldehyde, 2-aminophenol
(2), K2CO3 and PEG(300) (as was reported [24]) has
not conducted us to the considered results. It seems that
simultaneous addition of K2CO3 with other reactants
causes formation of potassium salt of 2 which alter
imine formation and its subsequent cyclization to 7. In
another attempt, 2-amino phenol was first dissolved in
PEG(300) at 50ꢀC and after addition of 2-fluorobenzal-
dehyde, the solution was stirred for 10 h at 50ꢀC to
complete imine (schiff base) formation. The production
of 7 was accomplished after addition of K2CO3 and con-
tinuing the reaction for 10 h at 100ꢀC (Scheme 5).
Thus, this procedure is slightly different from which
previously proposed [24]. In this manner, the prepara-
tion of 7 was performed in one-pot two-step procedure
following which preparation of imine in the first step
was followed by the potassium salt formation of imine
and its cyclization to produce 7 in the second step.
Scheme 4
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

September 2009
Reinvestigation of Alternative Method for the Preparation of
Dibenz[b,f][1,4]oxazepine
991
Scheme 5
pressure autoclave equipped with a magnetic stir bar. The mix-
ture was stirred at 150ꢀC for 6 h. Then the mixture was
washed with water (2 ꢁ 10 mL) and extracted by toluene (3 ꢁ
10 mL). Vacuum stripping of solvent and recrystalization from
benzene afforded 7 as an orange solid mp. 67–70ꢀC, lit. [14–
and the impossibility of its conversion to cis, makes it
inadequate for the preparation of 7.
1
16] mp. 71–72ꢀC, lit. [12] mp. 68–74ꢀC. H NMR (CDCl3): d
EXPERIMENTAL
7–7.5 (m, 8H, CH), 8.52 (s, 1H, CHN). 13C NMR (CDCl3): d
119.6, 120.3, 124, 124.6, 126.3, 127.7, 128.2, 129, 132.3,
139.5, 151.6, 159.4, 159.5, in agreement with literature [26].
Anal. Calcd for C13H9NO: C, 80.00; H, 4.61; N, 7.18. Found:
C, 79.23; H, 4.42; N, 7.04. GC-MS: retention time: 10.2 min;
m/z (intensity (%)): 50 (11), 51 (16), 63 (16), 139 (27), 140
(13), 166 (29), 167 (55), 195 (100), 196 (24) (lit.[27] 139 (25),
167 (52), 195 (100)).
One-pot preparation of 7 via fluoro derivative of 1. 2-
aminophenol (5.5 g, 0.05 mol) and PEG(300) (100 mL) were
placed in a 200-mL, one-necked flask equipped with a me-
chanical stirrer. The solution was stirred at 50ꢀC for 30 min.
After complete dissolution of 2-aminophenol in PEG(300), 2-
fluorobenzaldehyde (6.82 g, 5.77 mL, 0.055 mol) was added
and the mixture was stirred for 10 h at 50ꢀC. Finally, K2CO3
(6.9 g, 0.05 mol) was added and the mixture was stirred again
for 10 h at 100ꢀC. Extraction with ether (4 ꢁ 50 mL), several
washing of organic phase with water (5 ꢁ 200 mL), drying the
organic phase over CaCl2 and evaporation of solvent afforded
very pure product 7 (7.8 g, 80% yield) as an orange solid.
NMR spectra were obtained on a Bruker DPX-250 instru-
ment (250 MHz for 1H and 62.5 MHz for 13C), and CDCl3
was used as solvent; chemical shifts are reported in d (ppm)
from TMS. Electronic ionization GC-MS spectra were
recorded on a Varian (SATURN 4D) spectrometer with capil-
lary column (DB-5MS, 0.1 micron, 30 m ꢁ 0.250 mm). Only
m/z values having intensities of more than 10% are given and
retention times are reported using temperature programming
(100–250ꢀC, 10ꢀC/min) with He flow rate of 10 mL/min.
Melting points were obtained on a Mettler FP61 apparatus.
Preparation of 5. 2-Aminophenol (22 g, 0.2 mol) and
MeOH (200 mL) were placed in a 300-mL, one-necked flask
equipped with a mechanical stirrer. The solution was stirred at
50ꢀC for 30 min. After complete dissolution of 2-aminophenol
in MeOH, 2-chlorobenzaldehyde (28 g, 22.5 mL, 0.2 mol) was
added, and the mixture was stirred for 2 h at RT. The precipi-
tates formed were filtrated at 0ꢀC and dried to offered 39 g of
5 (84% yield) as an orange solid, mp. 97–99ꢀC.1H NMR
4
(CDCl3): d 6.89–7.42 (m, 8H, CH), 8.20 (d, JHAH ¼ 2.5 Hz,
1H, CHNtrans), 8.23 (m, 1H, CHNcis), 9.15 (s, 1H, OH). 13C
NMR (CDCl3): d 115.2, 116.2, 120.2, 127.2, 128.3, 129.5,
130.2, 132.4, 132.9, 135.4, 136.2, 152.6, 153.4. Anal. Calcd
for C13H10NOCl: C, 67.39; H, 4.32; N, 6.05. Found: C, 68.38;
H, 4.19; N, 6.18. GC: retention time: 28 min (5cis) and 38 min
(5trans). GC-MS of 5cis: retention time: 12.3 min; m/z (intensity
(%)): 50 (11), 63 (51), 64 (25), 75 (11), 92 (13), 166 (19), 201
(25), 229 (100), 230 (20), 231 (38). GC-MS of 5trans: retention
time: 13.6 min; m/z (intensity (%)):50 (13), 51 (16), 63 (19),
65 (29), 75 (11), 89 (11), 93 (16), 102 (10), 120 (100), 196
(24), 230 (18), 231 (33), 232 (25), 233 (18), 234 (12).
RERERENCES AND NOTES
[1] Siliwczuk, U.; Hibinger, H. Ger Offen DE 4, 104, 427;
Chem Abstr 1992, 117, 206963j.
[2] Norman, A. S Afr ZA 86, 08, 051; Chem Abstr 1989, 111,
161294z.
[3] Brown, H. A., Jr., US Pat Appl 69,034; Chem Abstr 1980,
93, 75016p.
[4] Okafor, C. O. Heterocycles 1977, 7, 391.
[5] Bandmann, A. L.; Savateev, N. V. Voen-Med Zh 1977, 3,
84; Chem Abstr 1977, 87, 96824s.
Preparation of 6. 2-Chloroaniline (12.8 g, 10.6 mL, 0.1
mol) salicylaldehyde (12.2 g, 10.4 mL, 0.1 mol) and xylene
(10 mL) were placed in a 100-mL, one-necked flask equipped
with a magnetic stir bar. The mixture was stirred at RT for
12 h. Evaporation of solvent and water afforded 21.9 g of
6trans (95% yield) as a yellow-orange solid, mp. 86.5–87ꢀC.
1H NMR (CDCl3): d 6.9–7.5 (m, 8H, CH), 8.61 (s, 1H, CHN),
13.2 (s, 1H, OH). 13C NMR (CDCl3): d 117.5, 119.1, 119.2,
127.7, 127.8, 129.6, 130.2, 132.6, 133.7, 144.8, 161.4, 163.2.
GC: retention time: 38 min. GC-MS: retention time: 13.5 min;
m/z (intensity (%)): 50 (13), 51 (21), 63 (12), 75 (22), 77 (12),
111 (14), 167 (16), 168 (14), 196 (100), 197 (15), 230 (11),
231 (57), 232 (40), 233 (25), 234 (13).
[6] Olajos, B. J.; Salem, H. J Appl Toxicol 2001, 21, 355.
[7] Coccaro, B. F.; Siever, L. J Clin Pharmacol 1985, 25, 241.
[8] Warawa, B. J.; Migler, B. M.; Ohnmacht, J. C.; Needles, A.
L.; Gatos, C. G.; McLaren, F. M.; Nelson, C. L.; Kirkland, K. M.
J Med Chem 2001, 44, 372.
[9] Hallinan, B. A.; Husa, R. K.; Peterson, K. B. EP 0512399
(or U.S. Pat. 5,283,240 (1994)); Chem Abstr 1993, 118, 102003a.
[10] Chandrakumar, N. S.; Hagen, T. J.; Hallinan, E. A.; Husa,
R. K. WO 9307132 (or U.S. Pat. 5,449,673 (1995)); Chem Abstr 1993,
119, 117284n.
[11] Ito, K.; Koizumi, M.; Murakami, Y.; Akima, M.; Aono, J.;
Ohba, Y.; Yamazaki, T.; Sakai, K.; Hata, S.; Takanashi, S. EP
0054951 (US Pat. 4,379,150 (1983)); Chem Abstr 1982, 97, 182469t.
[12] Tambute, A. Fr Demande 2,351,970; Chem Abstr 1979, 90,
23130q.
Preparation of 7 via 5. The sodium salt of 5 (25 g, 0.1
mol)—prepared by addition of 5 (23.2 g, 0.1 mol) to a metha-
nolic solution of KOH (5.6 g, 0.1 mol in 70 mL of MeOH)
followed by evaporation of solvent and drying the precipitate
formed—and DMSO (100 mL) were placed in a 250-mL, glass
[13] Tambute, A. Fr Demande 2,348,205; Chem Abstr 1978, 89,
24376r.
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

992
H. Fakhraian and Y. Nafary
Vol 46
[14] Narasimhan, N. S.; Chandrachood, P. S. Synthesis 1979, 589.
Amination Reaction, 16th Annual Research Symposium, Oklahoma
State University, 2005 (http://gradcollege.okstate.edu/events/ressymp/
abstracts/2005_poster_abstracts.pdf).
[15] Hunziker, F.; Kuenzel, F.; Schindler, O.; Schmutz, J Helv
Chem Acta 1964, 47, 1163.
[16] Higginbottom, R.; Suschitzky, H. J Chem Soc 1962, 2367.
[17] Bag, B. C.; Kaushik, M. P. Indian J Chem Technol 2002, 9,
484; Chem Abstr 2003, 138, 237689n.
[22] Fakhraian, H.; Nafary, Y.; Yarahmadi, A.; Hadj Ghanbary,
H. J Heterocycl Chem 2008, 45, 1469.
[23] Allaway, C. L.; Daly, M.; Nieuwenhuyzen, M.; Saunders,
G. C. J Fluor Chem 2002, 115, 91.
[18] Bag, B. C.; Kaushik, M. P. Indian J Chem Technol 2002, 9,
415; Chem Abstr 2003, 138, 273249c.
[24] Jorapur, Y. R.; Rajagopal, G.; Saikia, P. J.; Pal, R. R. Tet-
rahedron Lett 2008, 49, 1495.
[19] Malhotra, R. C.; Gutch, P. K.; Pal, V.; Ramachandran, P. K.;
Swamy, R. IN Pat. 180624 (1998); Chem Abstr 2004, 140, 375197t.
[20] Gutch, P. K.; Acharya, J. Heterocycl Commun 2007, 13,
393.
[25] Fakhraian, H.; Nafary Y.; Chalabi, H. Res Chem Intermed-
iat, to appear.
[26] Mesilaakso, M. Magn Reson Chem 1996, 34, 989.
[27] D’Agostino, P. A.; Provost, L. R. J Chromatogr A 1995,
695, 65.
[21] Patman, R.; Bunce, R. A. Dibenzo[b,f]1,4-oxazepines and
their 10,11-Dihydro Derivatives by a Consecutive Reduction-Reductive
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet