Synthesis of 2-alkylbenzimidazoles via TiO2-mediated photocatalysis

Full text of DOI:10.1021/jo970253j

5222
J . Org. Chem. 1997, 62, 5222-5225
contains both a reduction center (conduction band elec-
Syn th esis of 2-Alk ylben zim id a zoles via
tron) and an oxidation center (valence band electron
hole). Intermediates generated from one site could be
the substrates at another site. The integrated use of both
reaction sites could complete a sophisticated multistep
synthesis in “one pot”. In addition, most semiconductor
particles are inexpensive and nontoxic. The use of these
nanoparticles could provide environmental benign alter-
natives for organic synthesis. It has been reported that,
in an alcoholic solution of a primary amine, the alcohol
can be photocatalytically oxidized to an aldehyde or
ketone, which couples with the amine to form an imine.
The imine is then reduced to a secondary amine.10,11
Other examples include the photocatalytic conversion of
1,5-pentanediamine to piperidine on Pt/TiO2 (eq 3). The
transformation involves a single-electron transfer from
one of the amino functional groups to the electron hole
on the semiconductor and its subsequent hydrolysis to
an aldehyde. The aldehyde terminal is then intramo-
lecularly coupled with the remaining amino group to form
an imine, which then reduces to the secondary amine on
TiO2.12,13 Semiconductor particle-mediated syntheses of
1,2,4-triazolines and 2-phenylindazoles from azoben-
zene14 and tetrahydroquinoline from nitrotoluene15 also
involve a combined redox scheme.
TiO2-Med ia ted P h otoca ta lysis
Hongyu Wang, Richard E. Partch, and Yuzhuo Li*
Department of Chemistry, Clarkson University,
Potsdam, New York 13699-5810
Received February 11, 1997X
In tr od u ction
Benzimidazole and its derivatives have been the
subject of much research due to their importance in
various applications and its widespread biochemical
significance.1 For example, substituted 2-butylbenzim-
idazoles bearing a biphenylylmethyl moiety at the 1-posi-
tion have been used as Angiotensin II receptor antago-
nists.2 A number of benzimidazoles have been utilized
to prevent the corrosion of iron and steel in acid media.3
The classic study of the formation of 2-substituted
benzimidazoles was conducted by Philips. In Philips’
synthesis, o-phenylenediamine is used as a starting
material that reacts with carbonyl-containing com-
pounds, such as aldehydes, carboxylic acids, esters, or
amides, to form benzimidazole products4,5 (eq 1). Aque-
ous hydrochloric or polyphosphosphoric acid under reflux
conditions is usually used to assist the final ring-closure
step that involves dehydration.5,6
(2)
(1)
(3)
A variety of organic functional transformations medi-
ated by irradiation semiconductors have been reported.7
For example, it has been well established that a primary
alcohol can be easily oxidized on the semiconductor
surface to its corresponding aldehyde.8 We have reported
that irradiation of a primary alcoholic solution of a nitro
compound with suspended TiO2 particles gives an amino
compound in a high yield as the reduced product and an
aldehyde as the oxidized product (eq 2).9 For simple
oxidation or reduction needs, a photocatalyst does not
offer an overall advantage over conventional chemical
oxidizing or reducing agents and is no different from a
high surface area electrode. For multistep synthesis,
however, an illuminated semiconductor does offer a
unique feature. The charge separation on a semiconduc-
tor particle creates a unique microenvironment that
The objective of this work was to investigate the
feasibility of using this type of combined photocatalytic
redox reactions for the synthesis of benzimidazoles via
the reduction of an o-dinitro benzene and the oxidation
of an alcohol. In this paper, the formation of benzimid-
azole on the surface of semiconductor particles is first
described. The reaction mechanism and steps taken to
validate the reaction mechanism are discussed.
Resu lts a n d Discu ssion
Irradiation of a suspension of titanium dioxide in
ethanol containing o-dinitrobenzene gave 2-methylbenz-
imidazole in 96% yield in 1 h (eq 4). In a control
experiment, irradiation of an ethanolic solution of 1,2-
dinitrobenzene alone led to no reaction product, indicat-
ing that TiO2 was essential to the reaction. On the other
X Abstract published in Advance ACS Abstracts, J uly 1, 1997.
(1) Pozharskii, A. F.; Garnovskii, A. D.; Simonov, A. M. Russ. Chem.
Rev. 1966, 35(2), 122.
(2) Kubo, K.; Inada, Y.; Kohara, Y.; Sugiura, Y.; Ojima, M.; Itoh,
K.; Furukawa, Y.; Nishikawa, K.; Naka, T. J . Med. Chem. 1993, 36,
1772.
(3) Raicheva, S. N.; Aleksiev, B. V.; Sokolova, E. I. Corros. Sci. 1993,
34(2), 343.
(4) Philips, M. A. J . Chem. Soc. 1928, 2393.
(5) Hein, D. W.; Alheim, R. S.; Leavitt, J . J . J . Am. Chem. Soc. 1957,
79, 427.
(6) Alcalde, E.; Dinares, I.; Perez-Garcia, L.; Roca, T. Synthesis 1992,
4, 395.
(7) Bahnemann, D. W. Photochem. Convers. Storage Sol. Energy,
Proc. Int. Conf., 8th. 1991, 251.
(10) Ohtani, B.; Osaki, H.; Nishimoto, S.; Kagiya, T. J . Am. Chem.
Soc. 1986, 108, 308.
(11) Ohtani, B.; Goto, Y.; Nishimoto, S.; Inui, T. J . Chem. Soc.,
Faraday Trans. 1996, 92(21), 4291-4295.
(12) Nishimoto, S.; Ohtani, B.; Yoshikawa, T.; Kagiya, T. J . Am.
Chem. Soc. 1983, 105, 7180-7182.
(13) Ohtani, B.; Tsuru, S.; Nishimoto, S.; Kagiya, T. J . Org. Chem.
1990, 55, 5551-5553.
(14) Shibata, K.; Mimura, T.; Matsui, M.; Sugiura, T.; Minoura, H.
J . Chem. Soc., Chem. Commun. 1988, 1318-1320.
(15) Park, K. H.; J oo, H. S.; Ahn, K.; J un, K. Tetrahedron Lett. 1995,
36(33), 5943-5946.
(8) Hussein, F. H.; Pattenden, G.; Rudham, R.; Russell, J . J .
Tetrahedron Lett. 1984, 25(31), 3363.
(9) Mahdavi, F.; Bruton, T. C.; Li, Y. J . Org. Chem. 1993, 58, 744.
S0022-3263(97)00253-3 CCC: $14.00 © 1997 American Chemical Society

Notes
J . Org. Chem., Vol. 62, No. 15, 1997 5223
F igu r e 2. o-Diaminobenzene under photocatalytic condition.
the photocatalyzed reaction of a 3,4-dinitrotoluene in
ethanol with intentionally added 1-propanal successfully
gave 2-ethyl-6-methylbenzimidazole as the product. This
result provided supporting evidence for the involvement
of an aldehyde in the formation of benzimidazole. Ac-
cording to the proposed mechanism, the reductions of the
two nitro groups are not simultaneous. A time-depend-
ent study of the photocatalytic reaction of o-dinitroben-
zene showed that 2-nitroaniline was an intermediate. In
addition, irradiation of a suspension of TiO2 containing
2-nitroaniline instead of o-dinitrobenzene produced benz-
imidazole as well. This is consistent with the fact that
2-nitroaniline was found in the formation of benzimida-
zole as a transient product.
Irradiation of an alcoholic solution of p-dinitrobenzene
gave a high yield of p-diaminobenzene as the only product
via a p-aminonitrobenzene intermediate (eq 5). The
F igu r e 1. Photocatalytic formation of benzimidazole on TiO2
particles.
(5)
(4)
p-aminonitrobenzene intermediate can be isolated in a
high yield if the irradiation time is carefully controlled.
This is consistent with the fact that a ring-closure
reaction is needed to drive the condensation between an
aniline and an aldehyde toward the imine as shown in
Figure 1.
Irradiation of o-phenylenediamine in ethanol in the
presence of TiO2 failed to produce 2-methylbenzimidazole.
Although monoanil and dianil might have been formed
in this reaction mixture, they could not proceed further
to form benzimidazole compounds (Figure 2) as in the
Philips synthesis,4,16 due to the lack of oxidizing reagents.
Although the electron hole in the valence band is often
used in the oxidation of organic functional groups, the
monoanil or dianil clearly cannot compete efficiently with
the alcohol solvent. This result indicated that the second
nitro group is not reduced all the way to an amino group
in the pathway to benzimidazole.
hand, without irradiation under UV light, no reaction
product was detected either.
A mechanism for the production of benzimidazole is
proposed as illustrated in Figure 1. First, the solvent
ethanol is oxidized to aldehyde I while the dinitro
compound is reduced to 2-nitroaniline, II. The condensa-
tion between aldehyde I and the amino group of II forms
an imine, III. At room temperature and in the presence
of a considerable amount of water, III is in equilibrium
with the aniline and aldehyde. However, the further
reduction of III into hydroxylamine, IV, pushes the
equilibrium away from III. The nitrogen atom in the
hydroxylamine group attacks the imino carbon to close
the ring. Further dehydration leads to a benzimidazole.
The stiochiometry of the overall reaction demands a 10-
electron reduction for each dinitro compound and a two-
electron hole oxidation for each ethanol. In addition,
some conduction band electrons are also used to reduce
surface protons. As a consequence, therefore, some
residual acetaldehydes are found in a postreaction analy-
sis of the suspension. As all protons generated from
oxidation are consumed in the reduction of nitro com-
pounds or into hydrogen, no protons should be ac-
cumulated during the reaction. It is consistent with our
observation that the pH of the suspension remained
relatively constant throughout the reaction. The surface
hydroxyl groups are likely to serve as a cache-buffer zone
for the generated protons before being used in the
reduction sequences.
We have previously demonstrated that irradiation of
nitrobenzene with TiO2 gives a high yield of aniline while
the same reaction using ZnO particles instead generates
hydroxylamine as the final product (eq 6).9,17 Irradiation
(6)
of o-nitroaniline in ethanol with ZnO for ca. 1 h gave
2-methylbenzimidazole. In this case, ethanol solvent
A number of experiments were conducted to verify the
reaction mechanism. For the involvement of the solvent,
(16) Latif, N.; Mishriky, N.; Assad, F. N. Recl.: J . R. Neth. Chem.
Soc. 1983, 102(2), 73.

5224 J . Org. Chem., Vol. 62, No. 15, 1997
Ta ble 1. Syn th esis of Ben zim id a zoles via TiO2-Med ia ted P h otoca ta lysis
Notes
GC/MS
m/z (%)
1H NMR (CCl3-d)δ,
J (Hz)
yield
(%)
reactant/solvent
product
1,2-dinitrobenzene/ethanol
1,2-dinitrobenzene/1-propanol
2-methylbenzimidazole
2-ethylbenzimidazole
132.0 (100) 2.6 (s, 3H); 6.77-7.6 (m,4H)
96
92
146.0 (70)
146.0 (90)
160.0 (67)
1.46 (t, 3H, J ) 7.5); 2.94-3.03
(q, 2H, J ) 7.5); 7.21-7.58 (m, 4H)
2.43 (s, 3H); 2.53 (s, 3H); 6.33
(br, 1H); 7.01-7.43 (m, 3H)
3,4-dinitrotoluene/ethanol
2,6-dimethylbenzimidazole
95
97
3,4-dinitrotoluene/1-propanol
2-ethyl-6-methylbenzimidazole
1.41 (t, 3H, J ) 7.5); 2.44 (s, 3H);
2.91-3.01 (q, 2H, J ) 7.5);
7.01-7.45 (m, 3H), 8.55 (br, 1H)
4-ethoxy-2-nitroaniline/ethanol
4-ehtoxy-2-nitroaniline/1-propanol
1-chloro-3,4-dinitrobenzene/ethanol
5-ethoxy-2-methylbenzimidazole 176.0 (100) 1.40 (t, 3H, J ) 5.0); 2.60 (s, 3H);
94
96
3.96-4.05 (q, 2H, J ) 7.5);
6.83-7.44 (m, 3H)
1.37-1.43 (m, 6H); 2.90-2.99
(q, 2H, J ) 7.5); 3.77-4.04
5-ethoxy-2-ethylbenzimidazole
6-chloro-2-methylbenzimidazole
190.0 (62)
(q, 2H, J ) 7.5); 6.82-7.44 (m, 3H)
2.63 (s, 3H); 7.16-7.52 (m, 3H)
1.43 (t, 3H, J ) 7.5); 2.93-2.99
(q, 2H, J ) 7.5); 7.16-7.52 (m, 3H)
1.40 (t, 3H, J ) 7.5); 2.67 (s, 3H);
4.35-4.43 (q, 2H, J ) 7.5);
166 (100)
180 (54)
89
84
1-chloro-3,4-dinitrobenzene/1-propanol 6-chloro-2-ethylbenzimidazole
ethyl 3,4-dinitrobenzoate/ethanol
ethyl 3,4-dinitrobenzoate/1-propanol
6-(ethoxycarbonyl)-2-methyl-
benzimidazole
204 (43)
218 (69)
78
70
7.53-8.28 (m, 3H)
6-(ethoxycarbonyl)-2-ethyl-
benzimidazole
1.37-1.48 (m, 6H); 2.97-3.02
(q, 2H, J ) 7.5); 4.35-4.43
(q, 2H, J ) 7.5); 7.53-8.29 (m, 3H)
F igu r e 4. Involvement of o-nitrosoaniline.
F igu r e 3. Irradiation of o-nitroaniline with ZnO.
ylbenzimidazole compounds were formed, whereas 2-eth-
ylbenzimidazole compounds were produced when 1-pro-
panol was used. When a secondary alcohol such as
2-propanol was used as a solvent, no benzimidazole
compounds were obtained. The oxidation of 2-propanol
generated acetone instead of acetaldehyde. Compared
with acetaldehyde, acetone is much less apt, due to the
steric effect, to undergo condensation with nitroaniline
to form an imine. For the same reason, with a bulky
primary alcohol such as 1-butanol, the reaction was also
slow. The condensation reaction between the amino
compound and the aldehyde is clearly the rate-limiting
step of the entire reaction. Therefore, the larger the
alcohol the slower the reaction proceeds. Methanol,
however, is an exception. The reactions using methanol
as a solvent were slower than those using ethanol as a
solvent. For formaldehyde, the formation of hemiacetal
with solvent was predominant. The amount of free
formaldehyde was, therefore, small.
The substitutents on the benzene ring also have a
significant impact on the reaction kinetics. Electron-
withdrawing groups that lower the electron density on a
nitro group facilitate the conversion of the nitro group
into an amino group. The electron-withdrawing groups
also reduce the nucleophilicity of the resulting aniline,
hence lowering the reactivity for the condensation with
an aldehyde. The overall quantum efficiency, therefore,
depends upon the balance of these two effects. Prelimi-
nary results reveal that 3,4-dinitrotoluene has a higher
overall quantum efficiency than o-dinitrobenzene, which
in turn has a higher quantum efficiency than a reactant
could still be oxidized to aldehyde by ZnO particles, and
the nitro group of o-nitroaniline was reduced only to
hydroxylamine after condensation of the amino group
with aldehyde. This result confirmed that, after the
formation of III, it is not necessary for the second nitro
group to be reduced into an amino group (Figure 3).
When o-dinitrobenzene was used, 2-methylbenzimidazole
was also found on the irradiated ZnO particles. This
seems to be inconsistent with the reaction mechanism
illustrated in Figure 1 since it is the amino group that is
required for the first step, and only hydroxylamine can
be formed on ZnO particles. A possible explanation is
that the o-(dihydroxylamino)benzene from o-dinitroben-
zene may undergo an intramolecular redox reaction in
which one hydroxylamino group is oxidized into a nitroso
group while the other is reduced into an amino group.
The generated o-nitrosoaniline is then converted to
2-methylbenzimidazole as shown in Figure 4.
Irradiation of substituted dinitrobenzenes in the pres-
ence of alcohol solvents gave the corresponding substi-
tuted benzimidazoles in high yields (Table 1). The
products were isolated and purified via thin-layer chro-
matography. The structure of each product was con-
1
firmed with its mass and H NMR spectra (Table 1). The
reaction product and reaction rate were strongly solvent
dependent. When ethanol was used as a solvent, 2-meth-
(17) Li, Y.; Wang, L. Photoinduced Organic Functional Group
Transformations on Semiconductor Particles. In Nanocrystalline Semi-
conductor Materials; Kamat, P.; Meisel, D., Eds.; Elsevier: New York,
1996.

Notes
J . Org. Chem., Vol. 62, No. 15, 1997 5225
Gen er a l P r oced u r e for Ir r a d ia tion w ith TiO2. An a lyti-
ca l Sca le. In a 5 mL shell vial equipped with a magnetic stir
bar were mixed 2.0 mL of a 0.005 M alcoholic solution of a
reactant and 2.0 mL of a TiO2 suspension (1.0 g/L) in the same
solvent. The vial was sealed with a rubber septum and purged
with nitrogen for 5 min. The mixture was irradiated while
bearing an electron-withdrawing group, such as 1-chloro-
3,4-dinitrobenzene. A time-dependent study for the
irradiation of a TiO2 suspension containing 1-chloro-3,4-
dinitrobenzene revealed that the benzimidazole was
undetectable until all the starting material was com-
pletely converted into its nitroaniline intermediate. The
reaction of 3,4-dinitrobenzene gave a similar result.
These results indicated that the first reaction step, in
which the reactant was converted into an amine inter-
mediate, was faster than the second step, in which the
nitroaniline intermediate was converted into a benzim-
idazole product. This is consistent with the fact that a
dinitrobenzene compound is already electron deficient
enough and the basicity of nitroaniline or aniline is not
very high.
stirring in
a Rayonet photochemical reactor for a specific
duration (typically 15-30 min). After the irradiation, the
mixture was filtered with a 0.2 µm HPLC syringe filter to remove
TiO2 particles. The resulting solution was injected into the GC-
MS for analysis.
Yield Deter m in a tion . The experimental procedure was
basically the same as that of the analytical scale with the
following modifications. A 20.0 mL sample was placed in a shell
vial. After the irradiation and filtration, the solvent was
removed with a rota-evaporator. The resulting reaction mixture
was dried on a vacuum line before being dissolved in CCl3D for
1H NMR analysis. A known amount of trioxane was added as
an internal standard. The yield was determined on the basis of
the integrated area ratio between a product peak and the peak
from trioxane.
Con clu sion
Substituted benzimidaloes have been prepared via a
semiconductor-mediated photocatalytic process in which
the reduction product of an o-dinitrobenzene and the
oxidation product of an alcoholic solvent are coupled.
Although the coupled reactive intermediate can be easily
hydrolyzed via an equilibrium with water, the ring-
closure reaction pushes the reaction further to the final
benzimidazole product. The utilization of both oxidation
and reduction aspects of the semiconductor charge sepa-
ration is proven to be useful.
Sm a ll P r ep a r a tive Sca le. The reaction chamber has an
outside tube with a 33.0 mm inner diameter and an inside tube
with a 28.0 mm inner diameter. Both are 400 mm long. The
suspension of the semiconductor particles with an organic
substrate in solvent is pumped through the jackets of the
reaction chambers during the irradiation. A typical flow rate
is ca. 20 mL/min. After the reaction, the suspension is filtered
and the semiconductor particles can be reused. This type of
batch-flow reactor can handle up to ca. several liters of reaction
mixture. In a typical run, a 100 mL three-neck round-bottom
flask equipped with a magnetic stir bar was employed as a
reservoir for the reaction mixtures. A reaction suspension was
prepared by mixing 100.0 mL of 0.01 M reactant solution and
100.0 mL of 1.0 g/L TiO2 suspension. The suspension was
pumped through the flow reactor. The reaction mixture was first
purged with nitrogen for 30 min and then irradiated for 4 h.
After the irradiation, the mixture was filtered to remove TiO2
particles. The solvent was then removed with a rota-evaporator.
For a photochemical flow reactor with immobilized TiO2, a
similar workup procedure was employed. The resulting oily
mixture was placed on a TLC plate (SIL G-200 UV254, 20 mm ×
20 mm) for separation using chloroform as the eluent. After
the separation, the components were analyzed and characterized
with 1H NMR and GC-MS.
Exp er im en ta l Section
Ma ter ia ls. 3,4-Dinitrotoluene, 1,2-dinitrobenzene, 1-chloro-
3,4-dinitrobenzene, and 4-ethoxy-2-nitroaniline were obtained
from Aldrich. o-Nitroaniline and trioxane were obtained from
Eastman Kodak. 3,4-Dinitrobenzoic acid was obtained from
Lancaster. Titanium dioxide (P-25) was obtained as a gift from
Degussa Corp. All reagents were used upon receipt without
further purification. 3,4-Dinitro-1-ethoxycarbobenzene was pre-
pared from 3,4-dinitrobenzoic acid by reacting with ethanol
under acidic condition. GC-MS spectra were obtained on a
Hewlett-Packard Model 5890 series II gas chromatograph with
a Model 5917A mass spectrometer. 1H NMR spectra were
obtained on
a Bruker AC250 250 MHz spectrometer. All
1
products were identified by GC-MS and confirmed by H NMR
and MS. Photolysis experiments were conducted with a Rayonet
photochemical reactor equipped with 350 nm bulbs or a home-
made flow photochemical reactor that was built based on setups
described in the literature.17-20 The photon flux in the home-
made photochemical reactor was ca. 1.9 × 1017 quanta/min‚cm2
or ca. 2.6 × 1018 quanta/min‚cm2 in a Rayonet photochemical
reactor.
Ack n ow led gm en t. Financial support from the Na-
tional Science Foundation and the Petroleum Research
Fund, administered by the American Chemical Society,
is gratefully acknowledged.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
new compounds in lieu of combustion analysis (8 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
(18) Matthews, R. W. AU Patent B 76028, 1987.
(19) Turchi, C. S.; Mehos, M. S. Chemical Oxidation: Technologies
for Nineties; 2nd International Symposium, Vanderbilt University,
Nashville, TN, Feb 19-21, 1992.
(20) Hofstadler, K.; Bauer, R.; Novalic, S.; Heisier, G. Environ. Sci.
Technol. 1994, 28, 670-674.
J O970253J