Preparation of substituted 1,2,3,4-tetrahydroquinoxalines and 2,3,4,5-tetrahydro-1H-benzo[b][1,4]diazepines from catalytic Cp*Ir hydrogen transfer N-heterocyclization of anilino alcohols

Article abstract of DOI:10.1016/j.tetlet.2006.07.033

The [Cp^*IrCl₂]₂/K₂CO₃ catalyzed hydrogen transfer N-heterocyclization on a series of anilino alcohols has been investigated. The catalyst (20% loading) converts anilino alcohols to 1,2,3,4-tetrahydroquinoxalines and 2,3,4,5-tetrahydro-1H-benzo[b][1,4]diazepines in 30-84% isolated yield.

Full text of DOI:10.1016/j.tetlet.2006.07.033

Tetrahedron Letters 47 (2006) 6899–6902
Preparation of substituted 1,2,3,4-tetrahydroquinoxalines
and 2,3,4,5-tetrahydro-1H-benzo[b][1,4]diazepines
from catalytic Cp*Ir hydrogen transfer
N-heterocyclization of anilino alcohols
C. Todd Eary* and Dane Clausen
Array BioPharma, 3200 Walnut, Boulder, CO 80301, USA
Received 16 June 2006; revised 6 July 2006; accepted 8 July 2006
Available online 4 August 2006
Abstract—The [Cp*IrCl2]2/K2CO3 catalyzed hydrogen transfer N-heterocyclization on a series of anilino alcohols has been investi-
gated. The catalyst (20% loading) converts anilino alcohols to 1,2,3,4-tetrahydroquinoxalines and 2,3,4,5-tetrahydro-1H-
benzo[b][1,4]diazepines in 30–84% isolated yield.
Ó 2006 Published by Elsevier Ltd.
We recently became interested in new methods for the
synthesis of substituted 1,2,3,4-tetrahydroquinoxalines
as part of a medicinal chemistry program. Compounds
containing these heterocyclic cores have demonstrated
a wide range of biological activities. Some 1,2,3,4-tetra-
hydroquinoxaline containing structures (Fig. 1) have
been pursued as prostaglandin D2 receptor1 antagonists
(1) and vasopressin V2 receptor antagonists (2).2
priately substituted so that it is reduced to the desired
tetrahydro species and does not allow for the formation
of quaternary centers. An alternate strategy involves the
metal-mediated reaction of 1,2 dianilines with 1,4 butene
diol and acetate derivatives to form 2-vinyl 1,2,3,4-tetra-
hydroquinoxalines.4 Although the 2-vinyl 1,2,3,4-tetra-
hydroquinoxalines are formed in high yield, this
method is limited to symmetrically substituted dianilines
to avoid mixtures of regioisomers. Additional methods
to prepare 1,2,3,4-tetrahydroquinoxalines involve inter-
molecular Michael additions,5 tandem reduction–reduc-
tive amination reactions6 and the reduction and SN2
cyclization of nitroarenes with leaving groups.7,8
At present, there are a limited number of methods for
preparing 1,2,3,4-tetrahydroquinoxalines. A common
strategy involves the reduction of quinoxalines.3 This re-
quires that the starting aromatic quinoxaline be appro-
Recent reports of the Cp*Ir-complex hydrogen transfer
N-heterocyclization9 to form indoles,4 tetrahydroquino-
lines,10 and lactams11 prompted us to investigate this
transformation as a new method to form 1,2,3,4-tetra-
hydroquinoxalines from anilino alcohols (Scheme 1).
CO2H
OH
H
N
N
N
O
O
O
N
N
O
Cl
Cl
Cl
N
Ir
Ir
H
Cl
H
N
1
2
NH2
OH
R1
3
R4
R1
R4
R3
Figure 1. Biologically active 1,2,3,4-tetrahydroquinoxalines.
R3
N
R2
Base, Δ
N
R2
A
B
*
Corresponding author. Tel.: +1 303 386 1101; fax: +1 303 381
6652; e-mail: teary@arraybiopharma.com
Scheme 1.
0040-4039/$ - see front matter Ó 2006 Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2006.07.033

6900
C. Todd Eary, D. Clausen / Tetrahedron Letters 47 (2006) 6899–6902
This approach would allow direct preparation of
1,2,3,4-tetrahydroquinoxalines (B) from anilino alcohols
(A) in a catalytic and regioselective fashion that would
allow for the incorporation of quaternary centers. The
required substrates are easily prepared from substituted
ethanolamines and ortho-nitro aryl fluorides via SNAr
reactions12 and the catalyst (3) is now commercially
available. The proposed mechanism9,10 for N-hetero-
cyclization is thought to begin with initial Oppenauer-
type oxidation of the alcohol followed by cyclization
and reduction of the imine intermediate by the cata-
lyst system.
conditions failed to produce the desired 1,2,3,4-tetrahy-
droquinoxaline even after several days of heating. Cata-
lyst (3) loading was increased to 25% and the solvent
was switched to xylenes, in order to increase the temper-
ature of the reaction, and the product was formed in
30% yield after five days.
Since the hydrogen atom transfer reaction on substrate 4
was sluggish compared to the simpler published sys-
tems,10 we decided to study the effect of substitution
on N1. The extra heteroatom in our substrates may
allow bis coordination with iridium and render catalyst
turnover less efficient. The N-methylated substrate 5
produced the desired 1,2,3,4-tetrahydoquinoxaline in
modest yield after eight days with 10% of 3. Increasing
the catalyst loading to 25% also increased the yield to
80% with the reaction being complete after 17 h. We also
investigated N-heterocyclization for N-benzyl substrate
6, which was converted to product in low yield with
One of the first systems studied was 3-(2-aminophenyl-
amino) ethanol (4), which should undergo hydrogen
transfer to produce 1,2,3,4-tetrahydroquinoxaline (Ta-
ble 1). Conditions similar (10% 3, 10% K2CO3, toluene
sealed tube 110 °C) to those used by Fujita and Yama-
guchi10 were utilized to study the transformation. These
Table 1. a
Substrate
NH2
Product
Catalyst (3) (%)
Time
Yieldb (%)
H
N
10
25
4 d
5 d
NR
30c
OH
OH
N
4 H
N
H
NH2
H
N
10
25
20
20
8 d
17 h
4 d
4 d
52
80
19
NR
5 R = Me
6 R = Bn
7 R = Ts
N
R
N
R
a Reactions performed in a sealed tube at ꢀ120 °C in toluene at 0.25 M with K2CO3 content equal to catalyst.
b All yields based upon isolated material after chromatography.
c Xylene was used as a solvent and the reaction was run at 140 °C.
Table 2.
Substrate
Product
Catalyst (%)
20
Time
Yield (%)
64
H
N
NH2
N
2 d
OH
OH
N
8
H
N
NH2
N
20
2 d
59
N
9
H
N
Br
NH2
Br
F
20
20
3 d
62
64
OH
N
N
10
H
F
NH2
N
N
3 d
OH
N
11

C. Todd Eary, D. Clausen / Tetrahedron Letters 47 (2006) 6899–6902
6901
Table 2 (continued)
Substrate
Product
Catalyst (%)
20
Time
3 d
Yield (%)
61
H
N
O
NH2
N
O
OH
OH
N
12
13
H
N
NH2
N
20
3 d
42
O
O
O
N
H
N
O
18
84
N
H
NH2
N
20
20
1 d
7 d
OH
N
H
N
H
14
H
N
NH2
N
68
OH
N
15
some debenzylated material produced. No desired
product was isolated when N-heterocyclization was
attempted on tosyl protected aniline 7. After several
days, ꢀ80% of the starting material was recovered.
2,3,4,5-tetrahydro-1H-benzo[b][1,4]diazepines directly
from anilino alcohols using catalytic amounts of 3.
This reaction allows for the incorporation of quaternary
centers in a regioselective manner. Reaction times
vary from one to seven days using 20% catalyst loading
with yields averaging 63%. Substrates that are sterically
hindered at the 2-position, or alkylated at N1,
produce higher N-heterocyclization yields with shorter
reaction times.
After our initial experiments, we decided to investigate a
series of substrates with 20% catalyst (3) loading under
our standard conditions.13 We chose a series of N1
methyl substrates (8–13) with various aromatic substitu-
tions along with a 2,2-dimethyl substrate (14) and
homologated substrate (15). In general, yields were
ꢀ60% regardless of substitution with electron donating
or withdrawing groups. The weakly donating methyl
groups for substrates 8 and 9 produced marginally faster
reactions. The strong electron donating groups in 12 and
13 did not increase the reaction rate. Substrate 13 with
the methoxy para to the aniline nitrogen underwent a
side reaction to produce significant amounts of the
3,4-dihydroquinoxalin-2(1H)-one (18%). This byprod-
uct was observed in low (1–5%) amounts for several of
the substrates studied.11 The 2,2-branched free aniline
14 produced the highest yield of product with the short-
est reaction time. Since 14 is sterically hindered around
N1, the nitrogen’s ability to bis coordinate iridium
may be attenuated producing a faster transformation
and/or the dimethyl substituents may produce a more
favorable conformation for cyclization.14 We were
pleased to observe cyclization of 1515 to the 7-membered
1-methyl-2,3,4,5-tetrahydro-1H-benzo[b][1,4]diazepine.
The extended reaction time needed is likely due to the
less favorable 7-endo cyclization mode (Table 2).
References and notes
1. Torisu, K.; Kobayashi, K.; Iwahashi, M.; Nakai, Y.;
Onoda, T.; Nagase, T.; Sugimoto, I.; Okada, Y.; Mat-
sumoto, R.; Nanbu, F.; Ohuchida, S.; Nakai, H.; Toda,
M. Bioorg. Med. Chem. 2004, 5361–5378.
2. Ohtake, Y.; Naito, A.; Hasegawa, H.; Kawano, K.;
Morizono, D.; Taniguchi, M.; Tanaka, Y.; Matsukawa,
H.; Naito, K.; Oguma, T.; Ezure, Y.; Tsuriya, Y. Bioorg.
Med. Chem. 1999, 1247–1254.
3. For some representative examples of quinoxaline reduc-
tion to 1,2,3,4-tetrahydroquinoxalines see: (a) Cavagnol, J.
C.; Wiselogle, F. Y. J. Am. Chem. Soc. 1947, 69, 795–799;
(b) Pitts, M. R.; Harrison, J. R.; Moody, C. J. J. Chem.
Soc., Perkin Trans. 1 2001, 955–977; (c) Smith, R. F.;
Rebel, W. J.; Beach, T. N. J. Org. Chem. 1966, 39, 205–
207; (d) Murata, S.; Sugimoto, T.; Matsura, S. Hetero-
cycles 1987, 26, 763–766.
4. For some representative examples for the formation of
1,2,3,4-tetrahydroquinoxalines from dianilines and diols,
see: (a) Yang, S. C.; Shue, Y. J.; Liu, P. C. Organometallics
2002, 21, 2013–2016; (b) Yang, S. C.; Liu, P. C.; Feng, W.
H. Tetrahedron Lett. 2004, 45, 4951–4954; (c) Massacret,
M.; Lhoste, P.; Sinou, D. Eur. J. Org. Chem. 1999, 129–
134.
In summary, we have described a new and convenient
method to prepare 1,2,3,4-tetrahydroquinoxalines and

6902
C. Todd Eary, D. Clausen / Tetrahedron Letters 47 (2006) 6899–6902
Substrate was prepared from 4-methyl-N-(2-nitro-
5. Bunce, R. A.; Herron, D. M.; Ackerman, M. L. J. Org.
7
Chem. 2000, 65, 2847–2850.
6. Bunce, R. A.; Herron, D. M.; Hale, L. Y. J. Heterocyclic
Chem. 2003, 40, 1031–1039.
phenyl)benzenesulfonamide and 2-bromethanol utilizing
sodium hydride as base (43%). The nitro group was then
reduced with Zn as shown above (77%). Substrate 14 was
prepared as shown above in Scheme 1 using methyl 2-
amino-2-methylpropanoate (77%). The resulting ester was
then reduced with LiBH4 (22%) to the alcohol followed by
nitro reduction Pd/C, H2 (60%) as shown above.
´
´
7. Krchnˇak, V.; Smith, J.; Vagner, J. Tetrahedron Lett. 2001,
42, 2443–2446.
8. LaBarbera, D. V.; Skipbo, E. B. Bioorg. Med. Chem. Lett.
2005, 13, 387–395.
9. Fujita, K.; Yamaguchi, R. Synlett 2005, 560–571.
10. Fujita, K.; Yamamoto, K.; Yamaguchi, R. Org. Lett.
2002, 16, 2691–2694.
11. Fujita, K.; Takahashi, Y.; Owaki, M.; Yamamoto, K.;
Yamaguchi, R. Org. Lett. 2004, 6, 2785–2788.
12. Substrates 4–6 and 8–11 were prepared as shown in
Scheme 1. Substrates 12 and 13 were prepared as described
in Scheme 2.
13. General experimental: Substrate (1.0 equiv), 3 (0.2 equiv),
and base (0.2 equiv) were suspended in toluene (0.25 M) in
a pressure tube under a nitrogen atmosphere. The vessel
was sealed and heated to 110 °C with magnetic stirring.
The reaction was followed by TLC, HPLC, and LCMS.
The reactions generally took between 1 and 4 days
depending on the substrate. Once complete, the cooled
reaction mixture was purified directly via flash chroma-
tography (1:1 hexanes/EtOAc to 100% EtOAc).
Scheme 1
R1
14. Beesley, R. M.; Ingold, C. K.; Thorpe, J. F. J. Chem. Soc.
1915, 1080.
OH
N
H
NO2
NO2
F
R
K2CO3
DMF
R
15. Hussenether, T.; Hubner, H.; Gmeiner, P.; Troschutz, R.
¨
¨
OH
N
R1
Bioorg. Med. Chem. 2004, 12, 2625–2637, Substrate 15 was
prepared from LAH carboxylic acid reduction (29%)
and subsequent nitro reduction (Pd/C, 94%) of 3-(methyl-
(2-nitrophenyl)amino)propanoic acid.
60-98%
R1 = H, Me, Bn
Scheme 2
NH2
Pd/C, H2 or
Zn/HOAc
R
OH
OH
N
R1
45-85%
N
H
NO2
NO2
CsOAc
CuI
R
R
OH
N
DMF
I
14-26%