Palladium-catalyzed domino C-C/C-N coupling using a norbornene template: Synthesis of substituted benzomorpholines, phenoxazines, and dihydrodibenzoxazepines

Article abstract of DOI:10.1021/jo100408p

A rapid, four-step approach to alkyl- and aryl-substituted benzomorpholines is accomplished by a Pd-catalyzed domino C-C/C-N bond coupling using a norbornene template. Extension to phenoxazines and dihydrodibenzoxazepines is presented.

Full text of DOI:10.1021/jo100408p

pubs.acs.org/joc
an iodoarene can be functionalized with alkyl or aryl groups,
Palladium-Catalyzed Domino C-C/C-N Coupling
Using a Norbornene Template: Synthesis of
Substituted Benzomorpholines, Phenoxazines, and
Dihydrodibenzoxazepines
and the reaction is terminated with a palladium-catalyzed
coupling. The scope of this reaction has broadened consider-
ably since its discovery.5,6 Recently, we have reported a
highly modular approach to synthesize indolines 3 by a pal-
ladium-catalyzed domino C-C/C-N coupling (Scheme 1).7
Praew Thansandote, Eugene Chong, Kai-Oliver Feldmann,
and Mark Lautens*
SCHEME 1. Synthesis of Indolines by a Domino C-C/C-N
Coupling
Davenport Research Laboratories, Department of Chemistry,
University of Toronto, 80 St. George Street, Toronto, Ontario,
Canada M5S 3H6
mlautens@chem.utoronto.ca
Received March 4, 2010
Though good yields of indoline products were achieved, the
most successful examples of this approach suffered from the
requirement of using p-nitrophenyl as the group on nitrogen.
This report was the first example of the use of a reactive
heteroatom in the palladium-catalyzed, norbornene-
mediated domino sequence. We wished to further explore
the use of this novel feature within new substrate classes, and
herein we report our findings. A common byproduct of the
indoline methodology was the formation of an aziridine
from bromoalkylamine 2 or the formation of other products
from 2 (i.e., alkylation) that prevented the substrate from
reacting. Thus, we wished to explore substrates with separate
alkyl halide and amine moieties that when reacted would
provide a valuable approach to useful synthetic targets.
In this manner, we envisaged the formation of substituted
benzomorpholine 4 from 2-(2-iodophenoxy)ethanamine 5
(Scheme 2). The alkyl halide is now part of an intermolecular
component and the amine is part of a tethered chain that does
not bear a leaving group. The substrate 5 can be synthesized in
three straightforward steps using standard procedures.8 No-
tably, these steps combined with the key transformation can
produce substituted benzomorpholines 4 in only four steps.
A rapid, four-step approach to alkyl- and aryl-substituted
benzomorpholines is accomplished by a Pd-catalyzed
domino C-C/C-N bond coupling using a norbornene
template. Extension to phenoxazines and dihydrodiben-
zoxazepines is presented.
Benzomorpholines have been intensively studied as im-
portant scaffolds for pharmaceutically active compounds.1
However, there are only a handful of reports describing the
utility and synthesis of aryl-substituted benzomorpholines,
even though they are desirable constituents of pharmaceu-
ticals for the treatment of central nervous system disorders.2
A modern synthetic strategy toward these compounds would
be desirable to make them readily available for biological
studies.
SCHEME 2. Proposed Retrosynthesis of Alkylbenzomorpholines
The ability to control the formation of multiple bonds in a
single process is an efficient approach to maximize the value
of a synthetic operation.3 Our group has long been interested
in the development of one such process, the norbornene-
mediated palladium-catalyzed reaction initially reported by
Catellani4 which can form up to three C-C bonds in one syn-
thetic sequence. In this process, both of the ortho-positions of
We started our investigations by studying the synthesis of
1-alkylbenzomorpholine 6. The optimized reaction condi-
tions for the indoline synthesis7 were found to work well in
(5) (a) Catellani, M.; Motti, E.; Della Ca’, N. Acc. Chem. Res. 2008, 41,
1512. (b) Alberico, D.; Fang., Y.-Q.; Wilhelm, T.; Lautens, M. Pure Appl.
Chem. 2006, 78, 351.
(6) (a) Ca’, N. D.; Sassi, G.; Catellani, M. Adv. Synth. Catal. 2008, 350,
2179. (b) Mariampillai, B.; Alliot, J.; Li, M.; Lautens, M. J. Am. Chem. Soc.
2007, 129, 15372. (c) Candito, D. A.; Lautens, M. Angew. Chem., Int. Ed.
2009, 48, 6713.
(7) Thansandote, P.; Raemy, M.; Rudolph, A.; Lautens, M. Org. Lett.
2007, 9, 5255.
(8) See the Supporting Information.
(1) Ilas, J.; Anderluh, P. S.; Dolenc, M. S.; Kikelj, D. Tetrahedron 2005,
61, 7325.
(2) Stack, G. P.; Hatzenbuhler N. T.; Dahui, Z. International Patent
PCT/US2007/082421, 2007.
(3) Tietze, L. F.; Brasche, G.; Gericke, K. M. Domino Reactions in
Organic Synthesis; Wiley-VCH: Weinheim, 2006.
(4) Catellani, M.; Frignani, F.; Rangoni, A. Angew. Chem., Int. Ed. 1997,
36, 119.
DOI: 10.1021/jo100408p
r
Published on Web 04/27/2010
J. Org. Chem. 2010, 75, 3495–3498 3495
2010 American Chemical Society

Thansandote et al.
SCHEME 5. Byproduct 17 from Intermediate 16
JOCNote
SCHEME 3. Synthesis of Alkylbenzomorpholine 6
SCHEME 4. Mechanism of the Transformation
TABLE 1. Scope for the Synthesis of Benzomorpholines 18a
entry
R
product
yield (%)
1
2
3
4-CF3
3-CF3
2-CF3
18a
18b
18c
62
44
2b
ap-Methoxyphenyl abbreviated as PMP. bProduct identified by high-
resolution mass spectrometry.
seen. These protecting groups resulted mainly in the forma-
tion of the C-C/C-H product 17 (Scheme 5). The termi-
nal C-N coupling presumably happens from intermediate
16, and the required conformation for C-N coupling to
produce 6 may be achievable only in the case of aniline
substrates.10
Upon discovery of the importance of aniline-type sub-
strates for this transformation, we tested aryl iodide 5b in the
domino sequence. The use of the p-methoxyphenyl (PMP)
group on nitrogen adds value as it can be more readily
removed from the amine than the phenyl group. A switch
from ortho-alkylation to ortho-arylation was necessary to
avoid competing N-alkylation of the electron-rich amine.
Using similar reaction conditions, 1-arylbenzomorpholine
18 can be synthesized in 62% yield from iodide 5b (Table 1,
entry 1). In comparison with the alkylbenzomorpholine
synthesis, optimization studies showed that DMF was the
better solvent, P(m-ClPh)3 was the better ligand, and the
reaction time could be reduced. However, attempts to ex-
pand the scope of the synthesis of benzomorpholines such as
6 and 18 were met with limited success. We did, however,
study the effect of substituent position on the aryl bromide
(Table 1). Moving from a p-CF3 to a m-CF3 substituent
decreased the yield to 44% (entry 2). An o-CF3 substituent
gave only a trace of product (entry 3), illustrating the
negative effect of steric bulk for oxidative addition onto
the palladacycle 10 (Scheme 4).
this case. Benzomorpholine 6 can be synthesized from pre-
cursor 5a in 43% yield (Scheme 3). The major byproduct of
this benzomorpholine synthesis was the simple intramolecu-
lar amination product 7 in 28% yield. Given the ease of
formation of this byproduct, it was remarkable that ortho-
functionalization could successfully compete with intramo-
lecular amination.
The mechanism of this reaction involves palladium alter-
nating between three different transition states: Pd(0), Pd(II),
and Pd(IV) (Scheme 4).4,9 The mechanism begins with
palladium(0) oxidatively inserting into the aryl iodide to give
arylpalladium(II) species 8 (Scheme 4). Carbopalladation
with norbornene leads to 9, followed by intramolecular C-H
functionalization to give 10. Oxidative addition of 1-bromo-
butane leads to the palladium(IV) complex 11, which reduc-
tively eliminates to the palladium(II) species 12. Retro-
carbopalladation of norbornene, a result of the steric
constraints imposed by the two ortho substituents, leads to
the formation of 13 which undergoes an aromatic amination
to furnish the benzomorpholine 6.
Given the importance of aniline nitrogens in this domino
sequence, we turned our focus to substrate 5c (Scheme 6). We
had anticipated that the aromatic linker would reduce the
degrees of freedom to promote the final cyclization. In
The success of substrate 5a in the domino C-C/C-N
coupling suggests that aniline-type substrates are neces-
sary for the reaction. When Boc, Ts, Ac, Bz, and Cbz
protecting groups were tested, no desired products were
(10) Preliminary DFT calculations (B3LYP/SDD) of 5a, 5b, N-Ac-5, and
N-Bz-5 suggest that after ortho-alkylation, only aniline substrates 5a and 5b
possess the correct geometry for C-N coupling. N-Ac-5 and N-Bz-5 have the
Pd and N atoms too far apart for amination to occur, resulting instead in the
C-C/C-H product 17 from intermediate 16. Although substrate 5b pos-
sessed the correct geometry, the electron richness of the nitrogen caused
N-alkylation to occur faster than amination. See the Supporting Information
for more details.
(9) (a) Catellani, M.; Cugini, F. Tetrahedron 1999, 55, 6595. (b) Bocelli,
G.; Catellani, M.; Ghelli, S. J. Organomet. Chem. 1990, 390, 251. (c) Catellani,
M.; Mann, B. E. J. Organomet. Chem. 1993, 458, C12. (d) Catellani, M.;
Fagnola, M. C. Angew. Chem., Int. Ed. 1994, 33, 2421.
3496 J. Org. Chem. Vol. 75, No. 10, 2010

Thansandote et al.
JOCNote
SCHEME 6. Proposed Retrosynthesis of Arylphenoxazines
TABLE 2. Scope for the Synthesis of Phenoxazines 19
SCHEME 7. Optimized Conditions for Phenoxazine 19a
entry
R
product
yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
4-CF3
19a
19b
19c
19d
19e
19f
19g
19h
19i
47
4-CO2Me
2-NO2
4-NO2
4-C(O)Me
4-CN
34
60a
26b
30
40c
45d
27d
20d
15d
9d
3-F
3,4-diF
3-Me, 4-F
4-Me
4-OMe
2-F
2-CN
addition, the product of the C-C/C-N transformations
would be desirable phenoxazines.
19j
19k
19c
19d
Phenoxazines are ubiquitous pharmacophores in a variety
of important compounds, including anticancer,11 antiviral,12
and antimicrobial13 agents. In addition, they are also used as
dyes and pigments14 and biological probes.15 However,
1-arylphenoxazines have been largely unexplored, perhaps
due to the limited number of procedures available for their
preparation. To the best of our knowledge, there are only
two reported methods for the construction of 1-arylphenox-
azines, which both require multistep transformations and
were published over 30 years ago.16
0
0
aWith 1.5 equiv of aryl bromide. bWith 2.5 equiv of aryl bromide.
cWith 10 equiv of aryl bromide. dWith 1:1 aryl bromide/acetonitrile.
SCHEME 8. One-Pot Synthesis of Lactam 19l
Given the necessity for protecting groups in the previous
cases, we were delighted to find that the synthesis of 19 was
successful using unprotected anilines. Optimization of the
reaction of aniline 5c produced phenoxazine 19a in 47%
yield (Scheme 7). Palladium acetate and tri(m-chlorophenyl)-
phosphine provided the best catalyst system. Two equiva-
lents of norbornene was required for ortho-arylation as its
absence resulted in only amination products. A large excess
of base and high temperatures were required for the domino
sequence, and acetonitrile provided the best solvent medium.
Though the yield of 19a was modest, protection and depro-
tection steps were eliminated, and the nitrogen atom can be
directly further functionalized. In addition, substrate 5c can
be recovered from the reaction in some cases.
more problematic. Using 1-bromo-3-fluorobenzene resulted
in 45% yield of 19g (entry 7), but only with a very large excess
of aryl bromide (1:1 with the solvent). Similarly, using
1-bromo-3,4-difluorobenzene and 5-bromo-2-fluorotoluene
resulted in 27% of 19h and 20% of 19i, respectively, using a
1:1 ratio of aryl bromide to acetonitrile (entries 8 and 9).
Reactions with electron-neutral or electron-donating aryl
bromides also resulted in very low yields of phenoxazines
(entries 10 and 11). A noteworthy result occurred with the
use of methyl-2-bromobenzoate. The expected phenoxazine
further reacted with the ortho-ester moiety in situ to produce
lactam 19l in 60% yield (Scheme 8).
We explored the scope of this methodology by the reaction
of 5c with a variety of commercially available aryl bromides
(Table 2).
These results suggest that the substituent of the aryl
bromide largely influences these reactions. Initially in the
catalytic cycle, the aryl iodide 5c must be more reactive than
the aryl bromide toward Pd(0). However once C-H func-
tionalization of the aryl iodide has occurred, the resulting
Pd(II) palladacycle 10 (Scheme 4) must be selective for the
aryl bromide for the desired ortho-arylation to happen. The
full details governing this selectivity are unclear. However,
the insertion of the aryl bromide onto the Pd(II) palladacycle
has been shown to be aided by a directing group ortho to the
bromide.6 Thus, phenoxazines 19c and 19l are formed in
higher yields. Strongly electron-withdrawing groups on the
aryl bromides also fare well in this chemistry since they are
reactive enough for oxidative addition and subsequent ortho-
functionalization. In contrast, aryl bromides that are not
reactive for the Pd(II) palladacycle 10 are required in large
excess (entries 6-11). In these reactions, the aryl iodide 5c
Strongly electron-withdrawing aryl bromides were most
compatible with the reaction, giving between 47 and 60% of
the desired phenoxazine products (entries 1-6). The use of
1-bromo-2-nitrobenzene gave 60% yield of 19c (entry 3),
though most other ortho-substituted aryl bromides did not
react (entries 12 and 13). Less activated aryl bromides were
(11) Miyano-Kuraski, N.; Ikegami, K.; Kurosaki, K.; Endo, T.; Aoyagi,
H.; Hanami, M.; Yasumoto, J.; Tomoda, A. J. Pharmacol. Sci. 2009, 110, 87.
(12) Hayashi, K.; Hayashi, T.; Tomoda, A. J. Pharmacol. Sci. 2008,
106, 369.
(13) Foley, J. W.; Song, X.; Demidova, T. N.; Jilal, F.; Hamblin, M. R.
J. Med. Chem. 2006, 49, 5291.
(14) (a) Bruyneel, F.; Payen, O.; Rescigno, A.; Tirant, B.; Marchand-
Brynaert, J. Chem.;Eur. J. 2009, 15, 8283. (b) Porcal, G. V.; Previtali, C. M.;
Bertolotti, S. G. Dyes Pigm. 2009, 80, 206.
(15) Frade, V. H. J.; Coutinho, P. J. G.; Moura, J. C. V. P.; Gonc-alves, M.
S. T. Tetrahedron 2007, 63, 1654.
(16) (a) Sen, A. B.; Sharma, R. C. J. Indian Chem. Soc. 1956, 33, 671.
(b) Reid, W.; Stratz, A. Justus Liebigs Ann. Chem. 1972, 764, 11.
J. Org. Chem. Vol. 75, No. 10, 2010 3497

Thansandote et al.
JOCNote
SCHEME 9. Byproduct with Weakly Electron-Withdrawing
Arenes
amination sequence in 17-50% yields (Table 3). In general,
the yields of these products were lower than those of the
corresponding phenoxazines. However, 4-bromobenzaldehyde
was a successful substrate for this reaction (entry 3), which was
not true for the phenoxazine case. Notably, the use of methyl-2-
bromobenzoate did not produce a lactam similar to 19l, and no
desired dihydrodibenzoxazepines were obtained. The bypro-
ducts in these cases were not amenable to isolation and char-
acterization.
In summary, we reveal a novel synthesis of alkyl- and
arylbenzomorpholines using an extension of our previou-
sly reported palladium-catalyzed ortho-functionalization/
aromatic amination. A variety of arylphenoxazines and a
novel class of dihydrodibenzoxazepines can also be syn-
thesized in moderate yields via this approach. Notably,
unprotected anilines are compatible with the latter two
reactions.
TABLE 3. Scope for the Synthesis of Dihydrodibenzoxazepines 22
Experimental Section
entry
R
product
yield (%)
General Procedure for the Synthesis of Phenoxazines. Com-
pound 5a (62 mg, 0.20 mmol), aryl bromide (1.5-10 equiv
as specified), Pd(OAc)2 (4.5 mg, 10 mol %), tri(3-chloro)-
phosphine (16.0 mg, 22 mol %), norbornene (37.7 mg, 0.40 mmol),
cesium carbonate (6.0 equiv, 1.2 mmol), and dry CH3CN (2 mL,
0.10 M) were combined in a 0.5-2.0 mL microwave tube.
The tube was purged with argon for 30 s and sealed with a
Teflon cap. The resulting mixture was stirred at 135 °C for 18 h.
After being cooled to room temperature, the reaction mixture
was filtered through Celite and the solvent was removed under
reduced pressure. The residue was purified by flash chromato-
graphy on silica gel (25-50% DCM in hexanes).
1
2
3
4
5
4-CF3
22a
22b
22c
22d
22e
17
50
17
48
27
4-NO2
4-C(O)H
2-NO2
4-CN
becomes the more reactive ortho-arylation component for
oxidative addition onto the palladacycle and dimeric species
20 becomes the major product (Scheme 9).
Dihydrodibenzoxazepines are also important pharmaco-
phores for a variety of drugs including antipsychotic,17
anticancer,18 and anticonvulsant19 agents. Despite their
prevalence and numerous previously reported approaches,20
1-aryldihydrodibenzoxazepines 22 are unknown in the literature.
This new class of benzoxazepines can be synthesized from sub-
strate 21 using the analogous one-pot ortho-arylation/aromatic
Acknowledgment. We gratefully acknowledge the financial
support of the University of Toronto, the Natural Sciences
and Engineering Research Council of Canada (NSERC),
and Merck Frosst Canada for an IRC. We thank Professor
Vy Dong (University of Toronto) for the generous use of
her group’s GC-MS. We also thank Dr. Mima Staikova
(University of Toronto) for computational assistance. P.T.
and E.C. thank NSERC for CGSM/CGSD and USRA
scholarships.
ꢀ
(17) Ellenbroek, B. A.; Liegeois, J.-F. CNS Drug Rev. 2003, 9, 41.
(18) Mc Gee, M. M.; Gemma, S.; Butini, S.; Ramunno, A.; Zisterer,
D. M.; Fattorusso, C.; Catalanotti, B.; Kukreja, G.; Fiorini, I.; Pisano, C.;
Cucco, C.; Novellino, E.; Nacci, V.; Clive Williams, D.; Campiani, G.
J. Med. Chem. 2005, 48, 4367.
(19) Bajaj, K.; Archana; Kumar, A. Eur. J. Med. Chem. 2004, 39, 369.
(20) For examples of dihydrodibenzoxazepine synthesis, see: (a) Guo, L.;
Li, B.; Huang, W.; Pei, G.; Ma, D. Synlett 2008, 12, 1833. (b) Sekiyama, T.;
Matsuzawa, T.; Yamamoto, T.; Yatagai, M.; Ezaki, J. PCT Int. Appl. WO
2001017980, 2001. (c) Sakata, K.; Tsuji, T.; Sasaki, N.; Takahashi, K. PCT Int.
Appl. WO 9912925, 1999. (d) Hallinan, E. A.; Husa, R. K.; Peterson, K. B. Eur.
Pat. Appl. EP 512399, 1992.
Supporting Information Available: Specific experimental
details and characterization data for all unknown compounds.
Optimization tables and computational data are provided where
appropriate. This material is available free of charge via the
Internet at http://pubs.acs.org.
3498 J. Org. Chem. Vol. 75, No. 10, 2010