Differing patterns in the self-assembly of radially anchored imidazoles, 1,3,5-tris(1H-imidazol-2-yl)benzene and 1,3-Bis(1H-imidazol-2-yl)benzene

Article abstract of DOI:10.1007/s10870-007-9286-6

The novel 3-fold symmetric 1,3,5-tris(1H-imidazol-2-yl)benzene, having a propensity for a 3-fold salt formation by hydrogen bonding at the periphery, was prepared from oxidation of the hexahydro imidazole precursor. The transformation proved difficult and

Full text of DOI:10.1007/s10870-007-9286-6

J Chem Crystallogr (2008) 38:39–48
DOI 10.1007/s10870-007-9286-6
ORIGINAL PAPER
Differing Patterns in the Self-Assembly of Radially Anchored
Imidazoles, 1,3,5-Tris(1H-imidazol-2-yl)benzene
and 1,3-Bis(1H-imidazol-2-yl)benzene
Y. B. R. D. Rajesh Æ Subramania Ranganathan Æ
Richard D. Gilardi Æ Isabella L. Karle
Received: 2 August 2007 / Accepted: 15 October 2007 / Published online: 3 November 2007
Ó Springer Science+Business Media, LLC 2007
Abstract The novel 3-fold symmetric 1,3,5-tris(1H-
imidazol-2-yl)benzene, having a propensity for a 3-fold
salt formation by hydrogen bonding at the periphery, was
prepared from oxidation of the hexahydro imidazole
precursor. The transformation proved difficult and, of the
many reagents tried, only barium manganate and o-iodoxy
benzoic acid were successful. The oxidation to the
aromatic 1,3,5-tris imidazolyl benzene produces dramatic
changes, particularly when the X-ray structures of their
tris-TFA (trifluoroacetic acid) salts are compared. While
the hexahydro precursor exhibited a herringbone type
arrangement in the crystal, the oxidation resulted in
almost co-planar molecules that formed parallel stacks
which was confirmed by crystallization of the bis imidazole
from water.
Keywords Infinite parallel stacks of novel 1,3,5-tris
imidazole Á Unusual tetrameric assembly of 1,3-bis
imidazole Á Assembly mediated by water
Introduction
Self-assembly by non-covalent interactions is an elegant
approach to many kinds of large defined molecular
aggregates that circumvent time consuming chemical
synthesis that are often accompanied by various purifica-
tion problems. The efficiency of self-assembly leading to
supramolecules increases when the core module harbors
similar ligands. We have been interested in the radial self-
assembly having the constructs anchored on a benzene core
[1]. The present work relates to, inter alia, 1,3,5-tris
(1H-imidazol-2-yl)benzene 1 wherein three imidazoles are
arranged symmetrically around a benzene ring with ligand
attachments at two locations on each imidazole. Interest-
ingly in addition to 1 two additional isomers 2 and 3 are
possible. Of these only isomer 2 is known and has been
found to have interesting properties. In fact each of these
isomers can be expected to show different profiles. In
particular 1, the only member having the 3-fold symmetry
could be expected to have properties of self-assembly by
hydrogen bonding and stacking by virtue of its hydrogen
bonded aromatic profile. Both these expectations have been
realized.
˚
separated by 3.57 A between the atoms in separate planes.
1,3-Bis(1H-imidazole-2-yl)benzene is an attractive core
unit for assembly using a variety of auxiliaries. Several
attempts to complex the bis-imidazole with organic acids
failed. These efforts lead to the serendipitous discovery of
a novel tetrameric assembly, mediated by water molecules,
Richard D. Gilardi—Retired.
Y. B. R. D. Rajesh Á S. Ranganathan
Discovery Laboratory, Organic III, Indian Institute of Chemical
Technology, Hyderabad 500 007, India
R. D. Gilardi Á I. L. Karle (&)
Laboratory for the Structure of Matter, Naval Research
Laboratory, 4555 Overlook Avenue, S.W, Washington,
DC 20375-5341, USA
e-mail: isabella.karle@nrl.navy.mil
123

40
J Chem Crystallogr (2008) 38:39–48
Among the notable properties of 5 perhaps the most
attractive is its complexation with tetrazoles. The crystal
structure of which is shown in 6. A valuable property of
this complex is that the tetrazole allows ligand formation
with a variety of compounds to secure materials of desired
properties [3]. Compound 5 forms sandwiches by hydrogen
bonding with analogous chiral isooxazolines which then
form a supramolecular helical assembly [4]. The tetrahydro
pyrimidine analog of 5 has been found to be extraordinarily
versatile in the triple recognition of variety of carboxylic
acids having good solubility in non polar organic solvents
with an association constant of 61,700 : 11,900 M@1 in
CD3OD/CDCl3 (97:3) [5a]. Such salts derived from long
chain acids and 5 form supramolecular liquid crystals by
self-assembly around the tri basic core [5b]. It appears that
the encasement of the amidine group as part of a ring
system seems to be necessary to provide stable complexes
with a variety of carboxylic acids. In an open system the
stability of such salts vastly decreases [6]. An interesting
development is the replacement of the aromatic core with
methane that is triply liganded to 4-amino phenyl groups, a
very readily available compound, 7. This compound has
been successfully used in the design of a variety of den-
drimers whose profile ranges from hydrophobic to polar [7].
N
N
N
N
H
H
N
N
N
N
N
N
N
N
H
2
1
N
HN
N
N
HN
NH
NH
N
NH
N
4
3
In this context we felt that 1,3-bis(1H-imidazol-2-yl)ben-
zene 4 lacking one imidazole compared to 1 would have
advantages in self-assembly and molecular recognition.
The tris imidazole 2 having available only a tertiary
nitrogen for ligand formation self-assembles with Ag and
Cu ions to form interlinked metallocycles. In addition
complexes were secured from zinc and nickel. The struc-
tures were established by X-ray crystallography.
Additionally the analogous imidazolyl methyl compound
afforded the silver complex. A variety of such tris
N-imidazole compounds liganded to the benzene core have
been prepared and have been shown to form a range of
metal-organic frame works with specific individual topol-
ogies. Such compounds have also been demonstrated to
form selective guest inclusion and anion exchange [2].
The impetus for the present work to obtain the 3-fold sym-
metric 1 came largely from the ready availability of 1,3,5-tris
(4,5-dihydro-1H-imidazol-2-yl)benzene 5andthereportsonits
extensive chemistry [3]. In reality, the simple oxidative trans-
formation required extensive experimentation (vide infra).
NH
2
H
H N
NH
2
2
7
Radially designed heterocycles around the benzene core
are increasingly showing promise in the material domain.
W
HN
N
=
=
H
W
N
Water
W
=
H
N
N
H
N
=
NH
N
N
W
N
N
=
6
5
N
N
H
123

J Chem Crystallogr (2008) 38:39–48
41
For example such systems where the ligand is a substituted
oxadiazole are electroluminescent [8].
HR FABMS: Calculated for (C15H12N6+1)+: 277.120170;
Found: 277.119670.
Triply radially ordered designs could be simplified even
to methylene ammonium compounds. They form stable
duplexes with appropriately arranged partners. Such com-
plexes formed with the methylene phosphonate anions
have a ball shaped topology and are water soluble [9].
UV (kmax, MeOH): 280.5 nm (e = 30,487).
Fluorescence max (MeOH): 365 nm.
B. By oxidation of 5 with o—Iodoxy benzoic acid (IBX)
[11]: to a suspension of 5 (0.282 g; 1 mmole) in DMSO
(3 mL) was added 1.26 g (4.5 mmole) of IBX, the reaction
mixture was left stirred at 45 °C till the starting material
was consumed (TLC, 14 h), quenched by addition of
saturated Na2S2O3 (3 mL), the mixture made alkaline by
addition of 1 N KOH (2 mL) and extracted with EtOAc
(3 9 10 mL). The organic phase was washed with water
(2 9 10 mL), brine (15 mL), dried (MgSO4) and concen-
trated to afford 0.130 g of the crude product, which was
chromatographed on a column of silica gel. Elution with
CHCl3: MeOH (85:15) afforded 0.081 g (30%) of 1 Mp:
[300 °C.
Experimental
Synthesis
Materials and Methods
Melting points were recorded on a Fischer—John’s apparatus
and are uncorrected. Infrared spectra were recorded on a
Thermo Nicolet Nexus 760 spectrometer as KBr pellets and
prominent peaks expressed in cm@1. 1H and 13C NMR were
recorded on a Varian Gemini-200, Bruker-Avance 300 and
Inova 500 MHz NMR spectrometers. The chemical shifts are
expressed in d (ppm) with TMS at 0.00 as an internal refer-
ence. FAB and EI mass spectra were recorded on a VG
Autospec (with m-nitrobenzyl alcohol as matrix) and VG
7070H mass spectrometers. HR-MS were obtained on a
QSTAR XL instrument. UV measurements were carried out
on a Cintra 5 UV–Visible spectrometer. Fluorescence spectra
were recorded on a F-2500 fluorescence spectrophotometer.
1H NMR (200 MHz, CD3OD): d 7.15 (s, 6H, Imidazole
protons), 8.15 (s, 3H, aromatic protons).
1H NMR (200 MHz, DMSO-d6): d 7.38 (s, 6H, Imid-
azole protons), 8.30 (s, 3H, aromatic protons).
13C NMR (75.47 MHz, CD3OD): d 123.80, 124.40,
132.90, 146.90.
Preparation of 1,3,5-Tris(1H-imidazol-2-yl)benzene—Tris
Trifluoroacetate 9 A solution of 1 (0.01 g, 0.04 mmole)
in EtOH (1 mL) was admixed with a solution of TFA
(0.01 mL, 0.12 mmole) in ethanol (1 mL). The reaction
mixture was left aside for 12 h and the thin pale orange
needles precipitated were collected and dried. M.p: 250–
255 °C.
1,3,5-Tris(1H-imidazol-2-yl)benzene 1 A. By oxidation of
1,3,5-tris(4,5-dihydro-1H- imidazol-2-yl)benzene 5 [7]
with BaMnO4 [10]: a stirred suspension of 5 (1.5 g;
5.3 mmole) in CCl4 (100 mL) was admixed with BaMnO4
(13.5 g; 53 mmole) and the mixture refluxed while stirring.
The progress of the reaction was monitored by UV in
MeOH, which showed a shift from 228 nm (a = 0.3161) to
280.5 nm (a = 0.8191). At the end of 100 h of reflux,
solvents were removed and the resulting sludge was loaded
on a Soxhlet thimble and extracted with MeOH (100 mL).
The procedure was repeated two times and the combined
methanolic extracts were evaporated and the residue
(0.300 g) was chromatographed on silica gel. Elution with
CHCl3–MeOH (85:15) afforded 0.075 g (5%) of pale
orange 1 Mp: [300 °C.
1H NMR (200 MHz, CD3OD): d 7.66 (s, 6H, imidazole
protons), 8.52 (s, 3H, aromatic protons).
The structure was confirmed by X-ray crystallography.
1,3-Bis(1H-imidazol-2-yl)benzene 4 A. A solution of 1,
3-bis (4,5-dihydro-1H-imidazol-2-yl) benzene [12]
(0.428 g, 2 mmole) in dry DMSO (6 mL) was held at
120 °C for 48 h, cooled to room temperature, poured on to
water (100 mL), extracted with EtOAc (3 9 100 mL),
dried, evaporated to yield 0.320 g of the crude product
which was chromatographed on silica gel. Elution with
EtOAc: MeOH (90:10) gave 0.120 g (28%) of 4. Mp: 215–
220 °C (dec).
IR (KBr): 3,400, 3,084, 2,930, 1,546, 1,492, 1,453,
1,269, 1,107, 707 cm@1
.
1H NMR (500 MHz, CD3OD): d 7.22 (s, 6H, Imidazole
protons), 8.25 (s, 3H, aromatic protons).
1H NMR (200 MHz, DMSO-d6): d 7.28 (s, 6H, Imid-
azole protons), 8.30 (s, 3H, aromatic protons).
13C NMR (75.47 MHz, CD3OD): d 124.1, 124.4, 132.8,
146.8.
1H NMR (200 MHz, CD3OD): d 7.20 (s, 4H, imidazole
protons), 7.60 (t, 1 H, J = 7.93 Hz), 7.90 (d, 2H, J =
7.52 Hz), 8.22 (s, 1H) (aromatic protons).
EI-MS (m/z)%: 210 (M+) (100%).
B. A solution of 1,3-bis(4,5-dihydro-1H-imidazol-2-yl)
benzene (0.214 g, 1 mmole) in dry DMSO (3 mL) was
admixed with IBX (0.840 g, 3 mmole) and the reaction
FAB MS (m/z)%: 277 (M + 1)+ (100%).
123

42
J Chem Crystallogr (2008) 38:39–48
mixture left stirred at 45 °C till the starting material has
been consumed (TLC, 14 h). The reaction mixture was
quenched by addition of saturated Na2S2O3 (3 mL), made
alkaline with 1N KOH (2 mL) and extracted with EtOAc
(3 9 20 mL). The organic phase was washed with water
(2 9 15 mL), brine (15 mL), dried (MgSO4) and solvents
evaporated to afford 0.150 g of the crude product, which
was purified by column chromatography on silica gel.
Elution with EtOAc: MeOH (90:10) afforded 0.068 g
(32%) of 4. Mp: 215–220 °C (dec).
ordered core or backbone and a disordered periphery like
flexible side-chains and/ or disordered solvent. Direct
phase determination and partial structure reiterations
interspersed with least-squares refinements and difference
maps yielded the partial structure shown in Fig. 1a. Dif-
ference maps yielded indecisive information about the
placement of the nine fluorine atoms on the C2AA, C2BB
IR (KBr): 3,122, 3,035, 1,564, 1,418 cm@1
.
1H NMR (200 MHz, CD3OD): d 7.15 (s, 4H, imidazole
protons), 7.45 (t, 1H, J = 7.93 Hz), 7.75 (d, 2H, J =
7.52 Hz), 8.20 (s, 1H) (aromatic protons).
1H NMR (200 MHz, DMSO-d6): d 7.18 (s, 4H, imid-
azole protons), 7.50 (t, 1H, J = 7.93 Hz), 7.85 (d, 2H, J=
7.73 Hz), 8.55 (s, 1H) (aromatic protons).
13C NMR (75.47 MHz, DMSO-d6): d 121.90, 124.17,
129.20, 131.52, 145.50.
UV (kmax, MeOH): 274.0 nm (e = 20,018).
Fluorescence max (MeOH): 348 nm.
ESI-MS (m/z)%: 211 (M + H)+ (100%).
HR ESI-MS: Calculated for C12H11N4: 211.0983;
Found: 211.0978.
1,3-Bis(1H-imidazol-2-yl)benzene Dihydrate 10 A sus-
pension of 4 (0.02 g, 0.1 mmole) in water (15 mL) was
heated to *60 °C, the resulting clear solution filtered
through cotton and set aside. The brown rods that deposited
were harvested and dried. Mp: 310 °C.
IR (KBr): 3,207, 1,554, 1,452 cm@1
.
1H NMR (300 MHz, CD3OD): d 7.18 (s, 4H, imidazole
protons), 7.58 (t, 1H, J = 7.55 Hz), 7.85 (d, 2H, J =
7.74 Hz), 8.22 (s, 1H) (aromatic protons).
The structure was confirmed by X-ray crystallography.
X-Ray Structure Analysis
The pale orange crystals of complex 9, presented problems
for X-ray data collection. The tiny crystals were almost
always multiple. Several attempts to cool the crystals
resulted in shattering them. For the specimens that
appeared to be primarily a single crystal, data collections
were made at room temperature with Cu radiation on a
Bruker P4 four-circle diffractrometer and with Mo radia-
tion on the Bruker SMART APEX II CCD. The only data
set that yielded a partial structure was the one collected
with Cu radiation. The data was characterized by fairly
strong scattering at the lower scattering angles, followed by
a sudden steep decrease in the intensity of diffracted spots
at higher scattering angles. This kind of scattering pattern
is characteristic for molecules that have a reasonably
Fig. 1 (a) Partial strucutre of the TFA complex with 1,3,5-tris(1H-
imidazol-2-yl)benzene, 9, without the 9F atoms. Note that the CF3
groups in crystals appear often as two different conformers, or at least
with significant rotational disorder [13, 14]. (b) Anisotropic least
squares refinement for 9 at 50% probability for the thermal ellipsoids.
There are no non-positive definite values
123

J Chem Crystallogr (2008) 38:39–48
43
and C2CC atoms of the three TFA molecules complexed
with the core molecule. The behavior of the difference map
and the attenuation of the intensities of the scattered X-ray
data strongly suggested considerable rotational disorder in
the three CF3 groups. Rotational disorder in CF3 groups in
crystals is a frequent occurrence [13, 14]. The paucity of
the X-ray data necessitated the liberal use of distance
restraints on the better-known segments of the molecule,
that is, the phenyl ring, the connecting bonds to the phenyl
ring, the C–F, C–C–F and F–F distances. In addition,
groups of atoms (the three CF3 and three COO groups)
were divided into ‘‘part 2’’, ‘‘part 1’’ and ‘‘part 0’’ to be
treated separately for occupancy in the least-squares cycles.
Isotropic refinement, with each CF3 in two disordered
positions, produced an R factor of 22%. Anisotropic
refinement that involved 182 variables, 129 restraints and
703 data with |F0| [ 4r(F) resulted in an R factor of 15.1
%; however several of the F atoms (out of 18 positions) had
non-positive definite (NPD) thermal factors. In order to
avoid NPD thermal factors, only the CF3 on atom C2CC
could have two positions for each of the F atoms. The CF3
groups on C2AA and C2BB had to be described with one
position having large oscillations for the F atom as shown
in Fig. 1b. The best R factor under these circumstances is
0.168.
A disordered solvent molecule in 9, Fig. 1b, at positions
O1S and O2S, was found to exist in between separate
complexes. It is not clear whether O1S represents a water
molecule, or is part of an ethanol molecule where X1S is
one of the C atoms. A second C atom was not indicated in
the difference maps, although it could be disordered. Eth-
anol is a possible solvent molecule since the crystals of the
complex were grown from an ethanol solution, and there is
sufficient space in the crystal cell.
The structure of 9 obtained from X-ray data with a
˚
resolution of 1.5 A, obviously does not contain precise
Table 1 Crystallographic data and structure refinement
10
10LT
9
Identification
CCDC# 654599
CCDC# 662905
Same
CCDC# 654598
C21H15F9N6O6
618.39
Empirical formula
Molecular weight
Chemical, m.p., °C
Temperature, K
C
12H10N4 Á 2H2O
246.27
Same
310°
293°
1.5418
Same
250–255°
293°
1.5418
93°
0.71073
˚
Wavelength, A
Space group
C2/c
C2/c
C2/c
˚
a, A
34.757(3)
4.592(1)
17.018(2)
111.24(10)
2,531.6(4)
8
34.267(15)
4.571(2)
16.924(7)
111.468(7)
2,466.8(18)
8
30.339(12)
7.257(3)
˚
b, A
˚
c, A
26.363(16)
105.48(4)
5,593.9(48)
8
b, °
Volume A
3
˚
Z
Density (calc.(mg/mm3)
Abs, coeff., l
F(000)
1.292
1.326
1.469
0.755
0.094
1.322
1,040
1,040
2,496
Color
Brown
Brown
Pale orange
Needles
Shape
Rods
Rods
Cryst. size, mm
Limiting indices
1.1 9 0.14 9 0.14
@40 B h B 1
@5 B k B 1
@18 B ‘ B 20
2,163
0.54 9 0.18 9 0.1
@45 B h B 45
@5 B k B 6
@21 B ‘ B 22
14,637
0.065 9 0.050 9 0.040
@30 B h B 1
@1 B k B 7
@25 B ‘ B 26
2,835
Reflections collected
Reflections observed [ 4r(F)
Goodness-of-fit
1,860
2,556
709
1.065
1.060
1.004
Final R1 index [I [ 2r(I)]
Extinc. Coeff.
0.0596
0.0420
0.168
0.0060
0.00048
3
˚
Larg. diff. peak and hole, e/A
0.18/@0.26
188/0/1,860
0.29/@0.23
0.48/@0.39
441/129/703
Param./Restr./Data
182/6/3,141
123

44
J Chem Crystallogr (2008) 38:39–48
values for bond lengths and angles. Nevertheless the con-
formation of the molecule, the location of the hydrogen
bonds (Table 2) and the rotational disorder of the CF3
groups are established, as well as the most important
information for this investigation which is the nature of the
packing of the tri-substituted phenyl in 9. The structure of 9
is sufficiently well known to compare its major confor-
mational features with the structures of 8 and 10 (vide
infra).
N
NH
N
NH
BaMnO4-CCl4 / IBX-DMSO
H
H
N
N
N
N
HN
N
HN
N
5
1
Scheme 1
For the brown rods of 10, X-ray data were collected at
20 °C with Cu radiation on an automated four circle dif-
oxidized product by extensive Soxhlet extraction of the
BaMnO4 sludge with methanol. A very recent reagent
effective for the transformation of imidazolines and similar
compounds, namely, IBX (o—iodoxy benzoic acid) [11]
was found superior, not only in terms of better yields (30%)
but also the workup was far less cumbersome (Scheme 1).
In the initial stages where BaMnO4 was used as a reagent,
the course of oxidation was followed by UV irradiation in
methanol. On oxidation, the band at kmax 228.0 nm
(a = 0.3161) is shifted to kmax 280.5 nm (a = 0.8191). This
allowed workup when the maximum ratio is reached. The
structure of compound 1 m.p. [ 300 °C is supported by
extensive spectral data (IR, 1H NMR, 13C NMR, MS, HRMS,
UV and fluorescence).
fractometer (Bruker) equipped with
a
graphite
monochromator. Three reflections monitored after every 97
measurements remained constant within 3%. The structure
was solved by direct phase determination. Full matrix least-
squares refinement was performed on all atoms except the
H atoms. During the refinement H atoms were located in
the difference maps, added in idealized positions and
allowed to ride on the C or N atom to which they were
bonded. Hydrogen atoms on the water oxygens, O1S and
O2S were disordered among two positions. They were
weighted at half-occupancy in the refinement. A separate
data collection for 10, that is labeled 10LT, was repeated at
93 K on the CCD instrument with Mo radiation. Expected
The oxidation of 5 to 1 produced rather dramatic effects
in the NMR taken in CD3OD at 500 MHz. In the tris
imidazole 1, the complex aromatic peaks present at
8.40 ppm in 5, were replaced by a single one at 8.25 ppm.
A sharp peak at 7.22 ppm was seen for the imidazole
protons. The latter would indicate a hydrogen bonding
network that would make all nitrogens equivalent. The pale
orange color of 1 in contrast to the yellow color of pre-
cursor 5, denotes a more extensive p-bond network, which
is supported by a UV maximum at 280.5 nm and fluores-
cence maximum at 365 nm.
results were obtained, namely,
a lower R factor
(R1 = 0.0419) and a smaller cell (Table 1). The structure,
including the disorder of the H atoms on the water mole-
cules, was essentially the same. Data at both temperatures
are listed in Table 1 and filed at the CCDC.
Results and Discussion
The Tris Compound 1
The crystal structures of the TFA salts prepared from
the precursor 5 [18] (SOMHIL.RES (CCDC)) and the
oxidized product 1, respectively 8, 9, exhibited profound
differences.
The novel, symmetrical and radially organized 1,3,5-
tris(1H-imidazol-2-yl)benzene 1 with a hydrogen bonded
periphery appeared attractive for the study of self-assem-
bly, stacking and hydrogen bonding networks. Additional
impetus came from the fact that closely related systems
exhibited a wide range of properties, such as self-assembly,
molecular recognition, liquid crystal behavior, material
properties and electroluminescence. The synthesis of 1
seemed deceptively simple since the hexahydro precursor
1,3,5-tris(4,5-dihydro-1H-imidazol-2-yl)benzene 5 can be
readily prepared from mesitoic acid [7]. Actually, it took
more than two years of effort to acheieve this goal.
Crystals of the TFA salt of 1,3,5-tris(1H-imidazol-2-yl)
benzene 9 (m.p. [ 250–255 °C) were btained as very fine
pale orange needles by mixing imidazole 1 and trifluoro-
acetic acid in ethanol. The 1H NMR confirmed the
structure. Here again, the three aromatic protons and the six
olefinic protons appeared as sharp peaks at 8.52 and
7.66 ppm respectively, with the expected down shift
compared to 1. A major change occurred in the packing
motifs for salt 8 (SOMHIL.RES (CCDC)) (Fig. 2) and salt
9 (Fig. 3) that was directly caused by the aromatization of
5 to 1, reported in the present work. A herringbone motif is
observed for salt 8 (SOMHIL.RES (CCDC)) (Fig. 2). The
corresponding salt 9 arising from the oxidation (Fig. 3)
forms infinite stacks throughout the crystal with a
A major problem in the 5 to 1 change appeared to be the
extreme insolubility of 5 in almost all solvents. To date, of
the several reagents recommended for the oxidation of
dihydro imidazole to imidazole tried [10, 11, 15–17], initial
success was accomplished with BaMnO4 [10] that resulted
in yield of 5%. The work- up involved the removal of the
123

J Chem Crystallogr (2008) 38:39–48
45
O
F3C
+
CF3
O
O
N
N H
F3C
+
CF3
O
+
N
N H
H
H
O
H
O
O
H
O
H
N
H
N
N
+
N
+
+
N
H
O
N
N
H
O
N
H
H
O
O
CF3
CF3
8
9
˚
separation of only 3.57 A between the separate planes of
the cations. The pale orange color of the crystals is most
likely due to the close distance between the aromatic lay-
ers. The stacking of two modules of the complex is viewed
from the top in Fig. 4 where all the hydrogen bonds are
indicated by dotted lines. The complexation pattern with
trifluoroacetic acid of salts 8 [18] and 9 is quite similar
even though the packing is entirely different. The disorder
observed in the CF3 groups in Fig. 1 is fairly common in
the crystalline state [13, 14], where it is seen in nearly 25%
of the cases. The structure 1 is supported by all the data
cited. (With the recent ready availability of 1, it is planned
to study its complexation pattern and material properties).
The Bis Compound 10
The encouraging results from 1 made it attractive to
explore the complexation behavior of the 1,3-bis(1H-
imidazol-2-yl)benzene 4. This compound was readily pre-
pared by reaction of 1,3-dicyano benzene with H2S in DMF
[12] followed by oxidation with DMSO [15] or IBX to 4 in
Fig. 2 A herringbone packing
motif is seen for the TFA salt
of 1,3,5-tris(4,5-dihydro-1H
imidazol-2-yl)benzene, 8. The
imidazoline groups preclude
a planar arrangement
123

46
J Chem Crystallogr (2008) 38:39–48
acids did not succeed. However these efforts led to an
unusual pathway for the assembly to the basic module 10.
Methanolic solutions of 4 and p- bromo benzoic acid were
individually prepared and mixed. After one week the clear
solution was evaporated and the residue crystallized from
hot water to secure brownish rods mp 310 °C, compared to
215–220 °C (dec) for the precursor. Surprisingly an X-ray
crystallographic analysis revealed a unique molecular
assembly mediated with two water molecules, 4.2H2O
(Fig. 5a) leading to the basic assembly 44(H2O)8. These
non planar tetramers are further assembled in unique ways.
This unexpected result was confirmed by crystallization of
4 from water under the same conditions to secure precisely
the same hard crystals of 10.
N
N
N
N
H
H
O
Fig. 3 In contrast to Fig. 2, the cation of the TFA salt of the oxidized
product tris imidazole forms an infinite stack throughout the crystal
O
H
H
H
˚
with an interplane separation of 3.57 A
H
10
Fig. 4 The stacking of two cation layers, viewed perpendicular to
Fig. 3, a view from the top. All hydrogen bonds are indicated by
dotted lines. The two-fold disorder in CF3 groups is represented
by the cross-hatched atoms
an overall yield of 32%. The structure of 4 is fully sup-
ported by spectral data (IR, 1H NMR, 13C NMR, MS,
HRMS, UV and fluorescence). As could be seen from 4,
the proximate imidazoles have potential in the design of
tailored complexes and for specific molecular recognition.
All efforts to secure the linking of the two 1, 3 placed
imidazoles with a variety of mono and poly carboxylic
Fig. 5 (a) A single molecule of 1,3-bis(1H–imidazol-2-yl)benzene
10 drawn with thermal ellipsoids at 20 °C and 50% probability and
two water molecules. (b) An edge-on view. The two water molecules
are in the foreground. The molecule is not entirely planar
123

J Chem Crystallogr (2008) 38:39–48
47
An edge-on view of 10 with the two water molecules in
the foreground shows a nearly planar arrangement with a
small twist (Fig. 5b). An unusual feature for 10 is the
grouping of four neighboring molecules by hydrogen
bonds. The presence of a single NH in each imidazole
necessitates a disorder at the centers of symmetry shown
between N1bb and N1bc, Fig. 6. Each cannot coexist on
the N1b atoms. Each has half occupancy, in an apparent
random fashion. As a result, the hydrogen atom on N3b
occurs at half occupancy. When there is a hydrogen on
N1b, there is none on N3b, and vice-versa. This, in turn,
causes a disorder on O1s when it participates in hydrogen
bonding with N3b. There are three sites for hydrogen
atoms on O1s, only two being occupied at any one time.
All the hydrogen bonds are strong, see Table 2.
Table 2 Hydrogen bonds
˚
D—A, A
˚
H—A, A
Compound
Donor
Acceptor
9, tris TFA salta
N1a
O2c
2.65
2.67
2.78
2.67
2.61
2.55
2.80
2.874
2.784
2.855
2.78
2.72
N3a
O1a
N1b
N3b
N1c
O2a
O1b
O2b
O1c
O2sb
N1ba
N3b
O2sb
N3c
N3c
O1c
10, dihydrate
N1bc
O1sb
N1c
1.99
1.84
1.95
1.72
O1sbc
O1sbb
N3bbb
The tetrameric assembly shown in Fig. 6 is not coplanar
but the molecules are tilted with respect to each other. The
arrangement of the four molecules, when viewed from the
c-axis is shown in Fig. 7. The molecules can not stack.
Although in 4 the two imidazoles are disposed almost
exactly as in 1, the fact that they do not interlink through a
carboxylic group in spite of several trials was unexpected.
The formation of a stable salt with carboxylic acids
depends upon the relative importance of self-association
versus that with the donor. For several reasons the asso-
ciation of acids with 4 can be expected to be weaker
compared to the tris imidazole case. Thus in principle the
a
H atoms modeled, not found in difference map
Solvent (water)
b
c
Disordered H around a center
bis imidazole can associate only with a single carboxylic
unit, leaving sites for additional hydrogen bonding open.
On the other hand both in the case of 1 and 5 the three
carboxylic units snugly formed salts with protonated imi-
dazoles leading to a continuous array of strong ionic
interactions.
Conclusion
Radially arranged modules around a core unit has been
proved to be an effective vehicle for further elaboration to
macromolecules. In this context the synthesis of the 3-fold
symmetric 1,3,5-tris(1H-imidazol-2-yl)benzene (1), the
only isomer among three possible to have this property, is
noteworthy in several ways. The compound has potential to
be the base for crafting 3-fold symmetric designs that are of
current interest. The equivalence of the aromatic and the
six olefinic protons in 1 as well as its TFA salt 9, as seen by
1H and 13C NMR suggests also the equivalence of the six
nitrogen atoms by a hydrogen bonded periphery in 1. Such
modules form infinite stacks, as suggested by the X-ray
structure of its trifluoro acetate salt, and the resulting
p-interactions account for the pale orange color and
fluorescence of 1. The unprecedented water mediated tet-
ramerization of the 1,3-bis imidazole 4 and its assembly as
established by X-ray crystallography, makes it possible to
identify water as an agent for macromolecule formation of
molecules that can participate in hydrogen bonding.
Fig. 6 Four neighboring molecules of 10 that are connected by
hydrogen bonding. There is a disorder at the centers of symmetry,
shown between N1bb and N1bc, necessitated by sharing of a single
proton. All hydrogen bonds are strong
Acknowledgement SR and Rajesh are grateful to Department of
Biotechnology, New Delhi for financial assistance. Work performed
at the Naval Resesearch Laboratory was supported by the Office
of Naval Research.
123

48
J Chem Crystallogr (2008) 38:39–48
(b) Karle IL, Gilardi RD, Chandrashekar Rao Ch, Muraleedharan
KM, Ranganathan S (2003) J Chem Crystallogr 33:727; (c) Karle
IL, Rajesh YBRD, Ranganathan S (2005) J Chem Crystallogr
35:835
2. Fan J, Sun W-Y, Okamura T-A, Tang W-X, Ueyama N (2003)
Inorg Chem 42:3168 and references cited therein
3. Kraft A, Osterod F, Frohlich R (1999) J Org Chem 64:6425
4. Nam SR, Kim H-J, Sakamoto S, Yamaguchi K, Hong J-I (2004)
Tetrahedron Lett 45:1339
5. (a) Kraft A, Reichert A (1999) Tetrahedron 55:3923; (b) Kraft A,
Reichert A, Kleppinger R (2000) Chem Commun 1015
6. Kraft A (1999) Perkin Trans I 705
7. Kraft A, Osterod F (1998) Perkin Trans I 1019
8. Kraft A, Grimsdale AC, Holmes AB (1998) Angew Chem Int Ed
37:402
9. Grawe T, Schrader T, Zadmard R, Kraft A (2002) J Org Chem
67:3755
10. Hughey JL IV, Knapp S, Schugar H (1980) Synthesis 489
11. Nicolaou KC, Mathison CJN, Montagnon T (2004) J Am Chem
Soc 126:5192
12. Levesque G, Gressier J-C, Proust M (1981) Synthesis 963
13. A survey of disorders reported in the Cambridge Crystallographic
Data Centre shows that 25% of the reported structures containing
CF3 groups are disordered. ‘‘Twins, Disorders and other
Demons’’, Anderson JLF, Deschamps JR, Gilardi RD, George C
(2001) Crystal Eng 4:131
14. Karle JM, Karle IL (1989) Chemotherapy 1081
15. Anastassiadou M, Mouysset GB, Payard M (2000) Synthesis
1814
16. Amemiya Y, Miller DD, Hsu F-L (1990) Synth Commun
20(16):2483
17. Baltork IM, Zolfigol MA, Alibeik MA (2004) Tetrahedron Lett
45:8687
18. Kraft A, Frohlich R (1998) Chem Commun 1085
Fig. 7 Although it may appear in the projection when looking down
the b axis (Fig. 6) that the molecules are coplanar, they actually are
severely titled with respect to each other as seen in the projection
down the c-axis
References
1. (a) Ranganathan S, Muraleedharan KM, Chandrashekar Rao Ch,
Vairamani M, Karle IL, Gilardi RD (2001) Chem Commun 2544;
123