Regioselective synthesis of 1-benzyl- and 1-methyl-4- vinylimidazole and their reactions with N-phenylmaleimide

Article abstract of DOI:10.3987/com-02-9598

The regioselective synthesis of 1-benzyl- and 1-methyl-4- vinylimidazole from 4,5-diiodoimidazole is described. Their Diels-Alder reactions with N-phenylmaleimide provide a variety of adducts, including the anticipated enamine and the corresponding aromatized isomer. However, additional products including ene adducts, a bis Diels-Alder adduct and oxidation products were isolated.

Full text of DOI:10.3987/com-02-9598

HETEROCYCLES, Vol. 60, No. 1, 2003
1
HETEROCYCLES, Vol. 60, No. 1, 2003, pp. 1 - 7
Received, 12th August, 2002, Accepted, 13th November, 2002, Published online, 15th November, 2002
REGIOSELECTIVE SYNTHESIS OF 1-BENZYL- AND 1-METHYL-4-
VINYLIMIDAZOLE AND THEIR REACTIONS WITH N-
PHENYLMALEIMIDE
Carl J. Lovely,* Hongwang Du, and H. V. Rasika Dias†
Department of Chemistry and Biochemistry, The University of Texas at
Arlington, Arlington, TX 76019, USA
Abstract-The regioselective synthesis of 1-benzyl- and 1-methyl-4-
vinylimidazole from 4,5-diiodoimidazole is described.
Their Diels-Alder
reactions with N-phenylmaleimide provide a variety of adducts, including the
anticipated enamine and the corresponding aromatized isomer. However,
additional products including ene adducts, a bis Diels-Alder adduct and oxidation
products were isolated.
Without question, the Diels-Alder reaction is one of the most powerful synthetic transformations
available to organic chemists as a result of the broad substrate scope and the generally predictable regio-
1
and stereochemical course of this transformation. Despite extensive research in this area, the application
of the Diels-Alder reaction to
Ph
N
N
imidazole derivatives has yet to
N
O
O
N
N
R
N
R
2,3
O
O
receive significant attention.
X
N
R
N
PhH or CH2Cl2
Reflux
N
except R = Tr
Ph
O
Ph
O
Recently, we became interested in
the application of the Diels-Alder
reaction to 4-vinylimidazoles as a
3
1
2
R = Tr, Ts, MOM, SO2NMe2
Scheme 1
potentially efficient means to prepare a number of imidazole containing natural products and analogs. As
part of this effort, our group has reported the preparation and Diels-Alder reactions of 4-vinylimidazoles
4
possessing electron deficient substituents on N1 (Scheme 1). These studies demonstrated that these
dienes engage in the Diels-Alder reaction to provide the enamine (2) in good yield, without isomerizing to
the thermodynamically more stable aromatic system (3), and as a single stereoisomer (endo). The ability
to isolate these enamines was attributed to the conjugation present in the product, the reduction of the
electron density on the imidazole N1 (R = Ts, MOM, SO2NMe2), and the reduction in steric compression
4
between the N1 protecting group and the pyrrole ring (R = Tr) compared to the aromatic form. As a
continuation of these studies, we now describe the regioselective preparation and more complex reactivity
of 4-vinylimidazoles with electron donating substituents on N1 (R = Bn, Me).
* To whom correspondence should be addressed; e-mail: lovely@uta.edu
† Corresponding author for information regarding the X-Ray structures reported in this paper; e-mail: dias@uta.edu

2
HETEROCYCLES, Vol. 60, No. 1, 2003
Prior to the studies reported in this paper only non-regiocontrolled approaches to these vinylimidazole
had been reported. It was our intent to utilize 4-iodoimidazoles as precursors to these vinylimidazoles and
employ transition metal cross-coupling reactions to introduce the vinyl group. The motivation for this
approach lies in the divergency that can be obtained through the use of various cross-coupling partners. In
order to effect this approach, a regioselective synthesis of 1-alkyl-4-iodoimidazoles was required.
Although literature precedent was not particularly encouraging with regard to chemoselectivity, we
5
initially explored the alkylation of 4(5)-iodoimidazole (6). Our studies commenced with the synthesis of
6, which was prepared according to literature procedures via polyiodination and reductive deiodination
5
(Scheme 2). Protection of N1 was accomplished by application of Pilarski’s procedure for imidazole
alkylation, thus treatment of 6 with benzyl chloride and aqueous NaOH in DMSO provided a mixture of
6
the 4- and 5-iodoimidazoles in good overall yield. This reaction was not chemoselective and a ca. 2:1
mixture of the two isomers (4- vs. 5-) was formed during the course of this reaction. Although the two
regioisomers could be separated with reasonable efficiency by chromatography, the chemoselectivity of
this alkylation reaction with BnCl suggested that attempts to react 4(5)-iodoimidazole with other small
alkylating agents, e.g., MeI, would be less selective and in addition may lead to inseparable mixtures of 4-
7
and 5-substituted imidazoles, therefore, an alternate route was developed. Utilization of a symmetrical
substrate, such as 4,5-diiodoimidazole, would circumvent the problem of regioisomer formation during
the alkylation step, however the chemoselective removal of the 5-iodo substituent then became an issue.
However, it is well-known that polyhalogenated imidazoles undergo halogen-metal exchange (Br/I→Li
or Br/I→MgX) in a specific sequence under the appropriate conditions, i.e., C2>C5>C4, this ample
5,8-11
precedence encouraged us to employ this approach.
Thus, 4,5-diiodoimidazole (7) was treated with
NaH and then reacted with BnCl, or MeI to provide the corresponding protected imidazoles (8a-b) in
12
excellent yields.
The 4,5-diiodoimidazoles were then treated with 1.1 eq. of EtMgBr in THF, after
Lit.5
Lit.5
I
I
N
N
N
I
N
H
N
H
N
H
I
4
5
6
NaOH, H2O,
BnCl, DMSOa
Lit.12
EtMgBrb
I
I
I
I
N
N
N
N
SnBu3
Pd2dba3, PPh3
DMF, 90 o
NaH, THF
then RX
then H2O
N
H
N
R
N
R
N
R
I
C
8a: R = Bn, 90%
8b: R = Me, 93%
9a: R = Bn, 60%a or 93%b
10a: R = Bn, 90%
10b: R = Me, 30%
7
9b: R = Me, 87%
Scheme 2
iodine-magnesium exchange, the resulting Grignards were quenched with water. In each case analysis of
1H-NMR spectra of the crude products indicated the formation of a single iodoimidazole. After
purification and analysis through NOE experiments, it was clear that the iodide at C5 had been removed

HETEROCYCLES, Vol. 60, No. 1, 2003
3
selectively providing 9a or 9b, in accord with expectation. With the 4-iodoimidazoles (9a-b) in hand,
they were subjected to Stille reactions with tributylvinyltin to provide the corresponding 4-
13-15
vinylimidazoles (10a-b) in moderate to good yield.
In the studies previously reported by our group, the Diels-Alder reactions were conducted in either
CH2Cl2 or benzene at reflux and so for comparative purposes, these same solvents were employed in this
4
work. Generally, a 1:2.5 mixture of the 4-vinylimidazole (10a or 10b) and N-phenylmaleimide (NPM) in
the appropriate solvent was heated in a sealed tube at the indicated temperature in the Table. The outcome
of these reactions is discussed below.
O
1'
Ph
OH
N
N
N
Ph
N
3'
O
N
Ph
N
H
3
O
O
O
O
H
N
O
N
5
N
Ph
N
R
1
N
R
Ph
O
16
Solvent, heat
N
R
O
7
O
OH
N
Ph
N
O
O
Ph
N
N
10a-b
11a-b
12a-b
a: R = Bn
b: R = Me
O
1,3-H
Shift
N
-H2
Ph
Ph
O
17
Ph
N
N
N
O
O
O
O
N
N
N
O
N Ph
N
Ph
X
N
O
O
N
N
Ph
R
Ph
N
Ph
N
Ph
O
O
O
Ph
N
O
O
14a
18
Ph
13a-b
15a
Scheme 3
Table: Yields and Product Distribution for the Reactions of 4-Vinylimidazoles (10a-b) with N-Phenylmaleimide
Entry
Ra
Solvent
Time/ 11/% 12/% 13/% 15/%
Temp/°
C
h
14
10
1
2
3
4
5
6
7
8
9
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Me
Me
Me
Me
10a CH2Cl2
10a CH2Cl2
10a CH2Cl2
10a PhH
10a PhH
10a PhH
50
50
50
90
90
90
90
50
80b
88
-
-
-
-
-
-
-
9
-
-
-
-
-
18c
21d
21e
21f
21g
3.5
24
-
18
-
19
-
21
36
-
80
18
75
74
32
39h
21
76
46
-
28
-
-
19
-
10a PhH
10b CH2Cl2
10b CH2Cl2
10b CH2Cl2
10b PhH
50
50
90
10
11
13
-
-
-
-
6d
15
(a). A solution containing 1.0 eq. of 10a-b and 2.5 eq. of N-phenylmaleimide in PhH or CH2Cl2 was heated at the indicated
temperature for the indicated time. (b). Contaminated with a few percent of the aromatized product. (c) 30 mol% p-TsOH (d).
The reactions in benzene were allowed to proceed until all of the enamine had been consumed. This was done to assist
product isolation. (e). The reaction was run in the presence of catechol (30 mol%). (f). The reaction was run in the presence of
BHT (30 mol%). (g) Oxygen was bubbled through the reaction prior to heating. (h) Overall yield from 10b after treatment of
the “purified” enamine with p-TsOH.

4
HETEROCYCLES, Vol. 60, No. 1, 2003
1
1-Benzyl-4-vinylimidazole: Examination of the H-NMR spectrum of the crude reaction mixture of the
experiments performed in CH2Cl2 indicated that two new products had formed (Table, Entry 1). The
resulting adducts were separated by column chromatography to provide the enamine (11a) and ene
product (12a) in 80% and 9% yield respectively (Scheme 3). In contrast to the enamines prepared
previously (2, Scheme 1), 11a was much less stable to silica gel, on purification, even with Et3N treated
silica gel, a small amount of isomerization to the aromatized product (13a) was observed. Although it
should be pointed out that none of this aromatized product was observed by TLC during the reaction, or
in the 1H-NMR spectrum of the crude product. A NOESY experiment was performed on the Diels-Alder
adduct (11a) and it was clear from this spectrum that the addition had occurred via an endo transition
state. The second product (12a) isolated from this reaction, is assumed to arise from an ene reaction
16
between 11a and excess NPM. It was subsequently discovered that if the reaction was not allowed to
quite go to completion, then only the enamine and trace amounts of the ene product were formed (TLC
analysis). Concentration and trituration of the crude product allowed the isolation of the pure enamine in
88% yield (Table, Entry 2). It also proved possible to
isolate the aromatic product as the sole product when the
reaction was conducted in the presence of 30 mol% p-
TsOH (Table, Entry 3). A 1H NMR spectrum of the crude
material from this reaction indicated that the only product
was 13a. When the reaction was repeated in benzene a
different set of products was obtained, this was partly due
to the fact that the reaction was allowed to run until the
Figure 1: X-Ray Structure of 15a
17
enamine (11a) had been consumed (Table, Entry 4).
Under these conditions, three compounds were isolated. The aromatized adduct (13a) was isolated in 18%
yield along with a comparable quantity of the ene product (12a), however the major product, isolated in
28% yield, was a compound that had incorporated two moles of NPM. It was clear from the NMR
spectral data that this compound was not an isomeric ene product. Subsequently, X-Ray analysis of this
material revealed that this adduct had a completely different origin, it was in fact a bis Diels-Alder adduct
(15a, Figure 1). This compound results from the cycloaddition of NPM to vinylimidazole (14a) that arises
18,19
from oxidation of the initial adduct (11a).
Control experiments indicate that the enamine undergoes
oxidation, since the reaction of 13a with NPM does not produce the bis adduct, but resubmission of 11a
to the reaction conditions leads to the formation of 15a. Further support for this conclusion was obtained
by heating a sample of 11a under conditions utilized for the Diels-Alder reaction except in the absence of
NPM leading to the formation of the aromatic adduct (13a, 58%), vinylimidazole (14a, 9%), and a trace
of the fully aromatized adduct (18, 3%). Surprisingly, in addition to these three products, alcohols (16 and
17) were isolated as an inseparable 2:1 mixture of stereoisomers in 20% yield. The isolation of

HETEROCYCLES, Vol. 60, No. 1, 2003
5
oxygenated products suggested that dissolved oxygen in the reaction solvent was the source of these
alcohols, and so the reactions were conducted in the presence of anti-oxidants (catechol or
butylatedhydroxytoluene - BHT). In the presence of 30 mol% of catechol only the aromatized product
was obtained (Table, Entry 5), whereas when the reaction of 10a was performed in the presence of 30
mol% of BHT, the aromatized product and the ene adduct were obtained (Table, Entry 6). To further
probe this issue, the reaction was repeated after bubbling oxygen through the solvent under otherwise
standard conditions for the cycloaddition (Table, Entry 7), on purification of the crude reaction mixture,
the bis Diels-Alder adduct (15a) is obtained in 19% yield, along with the aromatic compound (13a, 32%).
In addition to these adducts, a mixture of the alcohols (16 and 17) was obtained as a 2:1 mixture of
stereoisomers in 28% yield that presumably results from the oxidation of 11a.
1-Methyl-4-vinylimidazole: Similar experiments were performed with this diene in both CH2Cl2 and
1
benzene as reaction solvents. In CH2Cl2 two compounds were observed in the H NMR spectrum of the
20
crude product, the endo enamine (11b) and the ene product (12b, 11b:12b = 3:1). In this case, the
enamine was extremely sensitive to purification on silica gel, proving difficult to obtain in pure form and
so it was converted to the aromatic isomer (13b) in 57% yield (39% overall) by treatment with p-TsOH
21
for the purpose of characterization. Crystals suitable for analysis by X-Ray crystallography were
obtained from the ene product (12b) in this case
(Figure 2). Not only did this analysis provide the
relative stereochemistry of the C3’-stereocenter, it
confirmed the regiochemistry of the 4-vinylimidazole
(10b) and suggested that the Diels-Alder product has
the indicated endo stereochemistry. The additional
Figure 2: X-Ray Structure of 12b
steps could be circumvented by allowing the reaction
to proceed for 24 h, at which time 11b was
completely consumed (Table, Entry 9), or by performing the reaction in the presence of p-TsOH (Table,
Entry 10). In benzene, the ene and aromatized adducts were isolated in 15% and 46% yield respectively.
Although the reaction was allowed to run until all of the enamine was consumed, no bis Diels-Alder
adduct was isolated in this case. We attempted to obtain 11b free from contamination with the aromatized
product by running the reaction to less than completion, analogous to the benzyl case described above.
However in this instance, it did not prove possible to isolate the enamine by trituration.
In summary, we have prepared 1-alkyl-4-vinylimidazoles in a regiodefined manner and these dienes
participate in efficient Diels-Alder reactions. Under the appropriate reaction conditions, the initial
enamine adduct can be obtained in good yield with a benzyl protecting group, although this enamine is

6
HETEROCYCLES, Vol. 60, No. 1, 2003
4
less stable than those which possess electron deficient substituents on N1. Under more forcing
conditions, benzene at reflux or prolonged reaction times, additional adducts are obtained that result from
the incorporation of a second equivalent of NPM, either via an ene reaction or through an oxidation-
Diels-Alder reaction sequence. These 2:1 adducts were not isolated in reactions involving more electron
4
deficient 4-vinylimidazoles. The methyl-substituted 4-vinylimidazole appears to be more reactive than
the benzyl protected derivative, although it did not prove possible to isolate the pure enamine. The 4-
regioisomers appear to be more reactive than the corresponding 5-regioisomer, and afford higher yields of
22
the cycloadduct.
We are currently investigating the Diels-Alder chemistry of these and other 4-
vinylimidazoles with other dienophiles and the application of these reactions to the synthesis of several
imidazole containing natural products. These results will be reported in due course.
ACKNOWLEDGEMENTS
This work was supported by the Robert A. Welch Foundation (CJL: Grant Y-1362; HVRD: Grant Y-
1289), the Texas Higher Education Coordinating Board-Advanced Research Program (003656-0004-
1999), and The University of Texas at Arlington (Start-up Funds). The NSF (CHE-9601771) is thanked
for partial funding of the purchase of a 500 MHz NMR spectrometer. Prof. Stephen Pyne (University of
Wollongong, Australia) is gratefully acknowledged for advice regarding the reduction of compound 5.
REFERENCES
1.
W. Oppolzer, Comprehensive Organic Synthesis, ed. by B.M. Trost and I. Fleming, Pergamon Press, Vol. 5, p 315,
1991.
2.
(a) A. S. Rothenburg, D. L Daulplaise, and H. P. Panzer, Angew. Chem., Int. Ed. Engl., 1983, 22, 560. (b) B. Abarca-
Gonzalez, R. A. Jones, M. Medio-Simon, J. Quilez-Pardo, J. Sepulveda-Arques, and E. Zaballos-Garcia, Synth.
Commun., 1990, 20, 321. (c) C. Yamazaki, K. Katayama, and K. Suzuki, J. Chem. Soc., Perkin Trans 1, 1990, 3085. (d)
K. Kosaka, K. Maruyama, H. Nakamura, and M. Ikeda, J. Heterocycl. Chem., 1991, 28, 1941. (e) M. A. Wuonola and J.
M. Smallheer, Tetrahedron Lett., 1992, 33, 5697. (f) M. A. Walters and M. D. Lee, Tetrahedron Lett., 1994, 35, 8307.
(g) Z. Wan and J. K. Synder, Tetrahedron Lett., 1997, 38, 7495. (h) C. Neipp, P. B. Ranslow, Z. Wan, and J. K. Snyder,
Tetrahedron Lett., 1997, 38, 7499. (i) I. Kawasaki, N. Sakguchi, N. Fukshima, N. Fujioka, F. Nikaido, M. Yamashita,
and S. Ohta, Tetrahedron Lett., 2002, 43, 4377. For related studies using 4-vinyl-2-imidazolidinones as dienes see A.S.
Dilley, and D. Romo, Org. Lett., 2001, 3, 1535.
3.
For a review of the Diels-Alder reactions of various vinyl heterocycles see: J. Sepulveda-Arques, B. Abarca-Gonzalez,
and M. Medio-Simon, Adv. Heterocycl. Chem., 1995, 63, 339.
4.
5.
6.
7.
C. J. Lovely, H. Du, and H. V. R. Dias, Org. Lett., 2001, 3, 1319.
B. Iddon and B. L. Lim, J. Chem. Soc., Perkin Trans 1, 1983, 735.
B. Pilarski, Liebigs Ann. Chem., 1983, 1078.
For a selective approach to 5-iodoimidazoles see F. B. Panosyan and I. W. Still, Can. J. Chem., 2001, 79, 1110. See also
P. Benjes and R. Grimmett, Heterocycles, 1994, 37, 735.
8.
9.
B. H Lipshutz and W. Hagen, Tetrahedron Lett., 1992, 33, 5865.
M. Abarbri, J. Thibonnet, L. Bérillon, F. Dehmel, M. Rottländer, and P. Knochel, J. Org. Chem., 2000, 65, 4618.
10. M. P. Groziak and L. Wei, J. Org. Chem., 1991, 56, 4296.
11. I. Kawasaki, M. Yamashita, and S. Ohta, Chem. Pharm. Bull., 1996, 44, 1831.
12. D. S. Carver, S. D. Lindell, and E. A. Saville-Stones, Tetrahedron, 1997, 53, 14481.
13. (a) M. D Cliff and S. G. Pyne Tetrahedron, 1996, 52, 13703. (b) H. Yashioka, T. Choshi, E. Sugino, and S. Hibino,
Heterocycles, 1995, 41, 161.
14. J. M. Kokosa, R. A Szafasz, and E. Tagupa, J. Org. Chem., 1983, 48, 3605.
15. C. G. Overberger and T. W. Smith, Macromolecules, 1975, 8, 401.
16. The relative stereochemistry of the C3’-stereocenter has not been established unequivocally for 12a. However, the
similarity of the NMR spectra of 12a with that of 12b, for which the stereochemistry was determined via X-Ray
crystallography, suggests that they possess similar relative stereochemistry. Further support for this assumption comes
from an X-ray determination of a closely related ene product derived from a 2-methyl substituted benzyl protected 4-

HETEROCYCLES, Vol. 60, No. 1, 2003
7
vinylimidazole, which possessed the same relative stereochemistry as 12b. H. Du, C.J. Lovely, and H.V.R. Dias
unpublished results.
17. The reaction was allowed to proceed until all of the enamine had been converted into other products in order to simplify
purification.
18. Noland has demonstrated that similar bis adducts can be obtained in the Diels-Alder reactions of 3-vinylindoles, although
these substrates contained substituents, e.g., -NO2, that either were readily eliminated (-HNO2) or activate oxidation (-H2)
to provide the new diene. W. E Noland, M. J. Konkel, M. S. Tempesta, R. D. Cink, D. M. Powers, E. O. Schlemper, and
C. L. Barnes, J. Heterocycl. Chem., 1993, 38, 183. See also L Pfeuffer and U. Pindur, Helv. Chem. Acta, 1987, 70, 1419.
19. This X-ray determination confirms the regiochemical assignment of the 4-vinylimidazole (10a).
20. The stereochemistry of this adduct was not determined experimentally, but inferred from the similarities of its 1H NMR
spectrum to that of other related enamines prepared in these laboratories, see ref. 4. Further support for the indicated
stereochemistry comes from the X-Ray analysis of the ene adduct, which indicates that it is derived from reaction with
the endo Diels-Alder adduct.
21. Our own observations regarding the isolation of this enamine are in accord with those reported by Walters and Lee for
the corresponding 5-isomer. See ref. 2f.
22. We have performed comparable experiments with 1-methyl-5-vinylimidazole and have determined that these substrates
are less reactive, requiring 9 h for complete consumption of starting material.