Engineering the solid state with 2-benzimidazolones

Article abstract of DOI:10.1021/ja952836l

Six derivatives of 2-benzimidazolone, disubstituted in the 4 and 5 positions, have been synthesized, and their structures have been determined in the solid state. Four of these compounds crystallize as molecular tapes. Compounds 1-(CH₃)₂, 1-CI₂, and 1-Br₂ form tapes that pack with their long axes parallel; compounds 1-H₂ and 2-H₂ form tapes that pack with their long axes at an angle to one another. Compounds 1-F₂ and 1-I₂ crystallize with a three-dimensional network of hydrogen bonds. The packing arrangement of molecular tapes is rationalized on the basis of closest packing and electrostatic interactions between aromatic rings. The occurrence of the network motif rather than the tape motif for 1-F₂ and 1-I₂ is rationalized on the basis of secondary interactions involving the halogen atoms.

Full text of DOI:10.1021/ja952836l

4018
J. Am. Chem. Soc. 1996, 118, 4018-4029
Engineering the Solid State with 2-Benzimidazolones
Kathryn E. Schwiebert, Donovan N. Chin, John C. MacDonald, and
George M. Whitesides*
Contribution from the Department of Chemistry, HarVard UniVersity,
Cambridge, Massachusetts 02138
ReceiVed August 17, 1995. ReVised Manuscript ReceiVed February 6, 1996X
Abstract: Six derivatives of 2-benzimidazolone, disubstituted in the 4 and 5 positions, have been synthesized, and
their structures have been determined in the solid state. Four of these compounds crystallize as molecular tapes.
Compounds 1-(CH3)2, 1-Cl2, and 1-Br2 form tapes that pack with their long axes parallel; compounds 1-H2 and
2-H2 form tapes that pack with their long axes at an angle to one another. Compounds 1-F2 and 1-I2 crystallize with
a three-dimensional network of hydrogen bonds. The packing arrangement of molecular tapes is rationalized on the
basis of closest packing and electrostatic interactions between aromatic rings. The occurrence of the network motif
rather than the tape motif for 1-F2 and 1-I2 is rationalized on the basis of secondary interactions involving the
halogen atoms.
Introduction
bond energies ≈ 1-5 kcal/mol) and fairly directional.6,7,11,12
The design of molecular aggregates using hydrogen bonds to
control structure in the solid state has yielded a number of
potentially useful structural motifs, including three-dimensional
diamondoid networks,13-15 two-dimensional layers,16-19 and
one-dimensional molecular tapes or rods.1-5,20-22
This work continues our exploration of the hydrogen-bonded
molecular tape as a motif in designing the structures of organic
crystals.1-5 Our objective in this work is to rationalize and
predict the structure of molecular crystals from the structures
of their constituent molecules. To achieve this goal, we must
simplify the structure of organic molecular solids sufficiently
that we can understand and control them. Our strategy is to
limit the possible orientations of molecules with respect to one
another by designing non-covalent interactions into the mol-
ecules. We have used molecular tapessstructures in which non-
covalent interactions hold molecules in a linear array that is
infinite in one dimensionsas the central motif in our studies.
By constraining molecular aggregates to be tapes, the problem
of predicting the packing of molecules simplifies to the problem
of predicting the packing of tapes.
Hydrogen-bonding,6,7 charge-transfer,8,9 and weak dipole-
dipole interactions10 are non-covalent intermolecular interactions
that have been used in efforts to assemble tapes and other
structures in the solid state. Hydrogen bonds, in particular, are
well suited for building molecular aggregates with specific
shapes and sizes because their energies are roughly comparable
to thermal energies (kbT ≈ 0.6 kcal/mol at 300 K; hydrogen
In previous work, we determined the structures of cocrystals
of a series of para- and meta-substituted diphenylmelamines
with diethylbarbital (Bar‚Mel(PhX)2).1-3 The strategy in this
work was to make a controlled set of modifications of molecular
structure and to try to rationalize the differences in the resulting
crystalline structures. These molecules cocrystallized in a 1:1
ratio and formed three hydrogen-bonded motifs: “linear” tapes,
“crinkled” tapes, and “rosettes” (Figure 1). In this system, we
established a correlation between steric interactions between the
substituents in the para positions and the crystalline motif: as
the size of the substituents increased, there was a change from
linear to crinkled to rosette motifs. It was difficult to study the
effect of the substituents on the packing of tapes, however,
because substitution could induce a switch from one tape motif
to another, rather than simply inducing a change in the packing
of the tapes. Polymorphism (the existence of more than one
(11) Aakeroy, C. B.; Seddon, K. R. Chem. Soc. ReV. 1993, 22, 397.
(12) Jeffrey, G. A.; Saenger, W. Hydrogen Bonding in Biological
Structures; Springer-Verlag: Berlin, 1991.
(13) Ermer, O.; Eling, A. J. Chem. Soc., Perkin Trans. 2 1994, 925.
(14) Wang, X.; Simard, M.; Wuest, J. D. J. Am. Chem. Soc. 1994, 116,
12119.
(15) Zaworotko, M. J. Chem. Soc. ReV. 1994, 23, 283.
(16) Zhao, X.; Chang, Y.-L.; Fowler, F. W.; Laugher, J. W. J. Am. Chem.
Soc. 1990, 112, 6627.
(17) Chang, Y.-L.; West, M.-A.; Fowler, F. W.; Laugher, J. W. J. Am.
Chem. Soc. 1993, 115, 5991.
X Abstract published in AdVance ACS Abstracts, March 15, 1996.
(1) Zerkowski, J. A.; Mathias, J. P.; Whitesides, G. M. J. Am. Chem.
Soc. 1994, 116, 4305.
(2) Zerkowski, J. A.; MacDonald, J. C.; Seto, C. T.; Wierda, D. A.;
Whitesides, G. M. J. Am. Chem. Soc. 1994, 116, 2382.
(3) Zerkowski, J. A.; Whitesides, G. M. J. Am. Chem. Soc. 1994, 116,
4298.
(4) Zerkowski, J. A.; Seto, C. T.; Whitesides, G. M. J. Am. Chem. Soc.
1992, 114, 5473.
(5) Zerkowski, J. A.; Seto, C. T.; Wierda, D. A.; Whitesides, G. M. J.
Am. Chem. Soc. 1990, 112, 9025.
(18) Hollingsworth, M. D.; Brown, M. E.; Santarsiero, B. D.; Huffman,
J. C.; Goss, C. R. Chem. Mater. 1994, 6, 1227
(19) Hollingsworth, M. D.; Santarsiero, B. D.; Oumar-Mahamat, H.;
Nichols, C. J. Chem. Mater. 1991, 3, 23.
(6) (a) Chin, D. N.; Zerkowski, J. A.; MacDonald, J. C.; Whitesides, G.
M. In Organised Molecular Assemblies in the Molecular Soid State;
Whitesell, J. T., Ed.; John Wiley & Sons: London, 1995, in press. (b)
Desiraju, G. R. Crystal Engineering: The Design of Organic Solids;
Elsevier: New York, 1989.
(7) Etter, M. C. J. Phys. Chem. 1991, 95, 4601.
(8) Fagan, P. J.; Ward, M. O.; Calabrese, J. C. J. Am. Chem. Soc. 1989,
111, 1698.
(20) Leiserowitz, et al. have demonstrated that it is possible to modify
the morphologies of hydrogen-bonded crystals by selectivity inhibiting the
rate of growth along a given face of a crystal: Addadi, L.; Berkovitch-
Yellin, A.; Weissbuch, I.; Mil, J. V.; Shimon, L. J. W.; Lahav, M.;
Leiserowitz, L. Angew. Chem., Int. Ed. Engl. 1985, 33, 466.
(21) (a) Lehn, J. M. Angew. Chem., Int. Ed. Eng. 1990, 29, 1304. (b)
Lehn, J.-M.; Mascal, M.; DeCian, A.; Fischer, J. J. Chem. Soc., Perkin
Trans. 2 1992, 461.
(9) Ward, M. D.; Fagan, P. J.; Calabrese, J. C.; Johnson, D. C. J. Am.
Chem. Soc. 1989, 111, 1719.
(10) Reddy, D. S.; Panneerselvan, K.; Pilati, T.; Desiraju, G. R. J. Chem.
Soc., Chem. Commun. 1993, 661.
(22) Hosseini, M. W.; Ruppert, T.; Schaeffer, P.; Decian, A.; Kyritsakas,
N.; Fischer, J. J. Chem. Soc., Chem. Commun. 1994, 2135.
(23) Fan, E.; Yang, J.; Geib, S. J.; Stoner, T. C.; Hopkins, M. D.;
Hamilton, A. D. J. Chem. Soc., Chem. Commun. 1995, 1251.
S0002-7863(95)02836-8 CCC: $12.00 © 1996 American Chemical Society

Engineering the Solid State with 2-Benzimidazolones
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4019
Figure 2. Tapes of substituted derivatives of 2-benzimidazolones.
2-benzimidazolones (Figure 2). The benzimidazolones are rigid,
so conformational polymorphism is not possible. Polymorphism
based on torsions associated with the substituents can be
prevented by choosing only axially symmetrical substituents.
We also hypothesized that using a molecule that could form
two hydrogen bonds, rather than the three of the (Bar‚Mel-
(PhX)2) system, would make kinetics of growth along the
hydrogen-bond axis less favored. Slower growth along the
hydrogen-bond axis might promote three-dimensional crystals
rather than the very thin needles observed in the (Bar‚Mel-
(PhX)2) system.23
A primary objective of this work was to identify factors that
influence the packing of tapes of benzimidazolones in the solid
state. Ideally, the tapes will pack in the crystal with a minimum
of free energy. Molecules have to be close together for
favorable interactions to occur, but they must not be so close
that unfavorable interactions overwhelm the favorable ones.
Kitaigorodskii described the ideal packing as one in which “no
molecules are suspended in empty space and none overlap.” 25
This relationship of lowest free energy to closest packing
suggests that tapes should leave as little empty volume in the
crystal as possible. Thus, we hoped that we might be able to
predict the packing arrangement on the basis of closest packing
of tapes.
Since tapes of benzimidazolones are planar, substitution of
the aromatic ring primarily affects the profile of the edge of
the tape. To fill as much space as possible, tapes with scalloped
edges (small substituents) should pack in a nonparallel arrange-
ment, while tapes with smooth edges (large substituents) should
pack in a parallel arrangement (Figure 3). We also expected
electrostatic interactions between aromatic rings to influence
the packing. Desiraju and Gavezzotti have suggested that
polynuclear aromatic hydrocarbons adopt orientations that
maximize two specific interactions: (i) van der Waals interaction
between carbon atoms in the rings (that is, geometries in which
the aromatic rings are parallel) or (ii) electrostatic attraction
between a hydrogen and carbon in the ring (that is, geometries
in which aromatic rings pack at an angle to one another).26 In
some cases, the packing arrangement of an unsubstituted
aromatic compound shifts from nonparallel to parallel for a
heterocyclic analog. Desiraju and Gavezzotti argue that this
shift occurs because there are fewer hydrogen atoms around
the rim of the ring in the heterocycle (and hence fewer C-H
bonds to interact with the proximate aromatic π orbitals) than
in the unsubstituted aromatic compound.
Figure 1. Interactions between neighboring phenyl groups in a
developing linear tape (a) can force adoption of a crinkled tape (b);
further steric stresses, now with barbituric acid substituents, lead to
the rosette (c).
solid form) also complicated the comparison of crystalline
structures in this series. We detected polymorphism in two cases
(X ) Br and CF3). To predict crystal structure on the basis of
molecular structure, each perturbation of molecular structure
should lead to only one type of crystal structure. We hypoth-
esized that polymorphism resulted from conformational isomer-
ism around the phenyl-nitrogen bond. Since many torsions
of the phenyl rings were possible, there was no well-defined
minimum of free energy in the crystal. Obtaining single crystals
was an additional problem in this series because growth along
the long axis of the tape was fast relative to growth in other
directions; thus, the cocrystals were usually very fine needles.
To avoid the problems associated with the (Bar‚Mel(PhX)2)
series, we searched for systems that could form only one tape
motif and had no possibility for conformational polymorphism.
We identified derivatives of 2-benzimidazolone as a class of
molecules with potential to form linear tapes in the solid state.24
Unlike the (Bar‚Mel(PhX)2) system, there is only one orientation
of the hydrogen-bond donor/acceptor pair that permits tapes in
The unsubstituted 2-benzimidazolone 1-H2sthe only 2-ben-
zimidazolone that has been previously characterized in the solid
statesforms tapes that pack at an angle to one another.27 This
nonparallel packing arrangement is consistent with electrostatic
interactions between a hydrogen atom of an aromatic C-H bond
(25) Kitaigorodskii, A. I. Molecular Crystals and Molecules; Academic
Press: New York, 1973.
(26) Desiraju, G. R.; Gavezzotti, A. J. Chem. Soc., Chem. Commun. 1989,
621.
(24) MacDonald, J. C.; Whitesides, G. M. Chem. ReV. 1994, 94, 2303.
(27) Herbstein, F. H.; Kapon, M. Z. Kristallogr. 1985, 173, 249.

4020 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
Schwiebert et al.
Table 1. Crystallographic Data for Compounds 1 and 2
space
class group
density
crystal
melting
b
c
urea
a (Å)
b (Å)
c (Å)
b (deg)a
R1
R2
(g/cm3)d solvente
shape point (°C) Ck* f
1-(CH3)2
1-Cl2
1-Br2
1-H2
2-H2
1-F2
|
|
|
P21/c
P21/c
P21/a 14.051(6)
P21/n
P21/n
8.0616(9) 7.186(1)
13.9442(9)
13.723(3)
7.967(4)
23.440
96.300(7) 0.048 0.223
1.34
1.75
2.39
1.42
1.43
1.65
2.77
DMF
DMF
blocks 395
blocks 398
0.71
0.71
7.800(2)
7.193(4)
7.267(2)
5.738
92.94(2)
93.38(4)
94.50
0.055 0.191
0.099 0.316
0.052
CH3OH
CH3OH
CH3OH
THF/Et2O plates 332
CH3OH
blocks 386 (dec) 0.72
×
5.296
blocks 316g
blades 394
0.70
0.72
0.69
×
6.0397(8) 4.0704(9) 34.706(5)
8.2319(6) 7.3405(4) 11.3087(5)
90.90(1)
0.051 0.220
3-D P21/c
90.530(5) 0.034 0.099
108.19(2) 0.045 0.124
1-I2
3-D P21/a 14.156(6)
4.780(6)
14.391(3)
plates 377 (dec) 0.68
a R and γ are constrained to be 90° by the symmetry of the space group. b This value is the crystallographic Reliability Index, R1 ) ∑|Fo -
Fc|/∑Fo for Fo > 2σ. c R2 ) ∑|Fo2 - Fc2|/∑Fo2 for all σ. d Calculated using the program Cerius2.34 e Solvent used for the growth of single crystals
for X-ray crystallograhy. f Ck* is the packing fraction: Ck* ) N(Vm/Vc), where N is the number of molecules in the unit cell, Vm is the volume of
the molecules in the unit cell (calculated with the program Platon35 rather than using tables of incremental volumes, as done by Kitiagorodski),25
and Vc is the total volume of the unit cell. g Lit.36 309-310 °C.
Scheme 1. Synthesis of the Benzimidazolones 1-X2 and 2-H2
hydrogen would provide smooth “edges” along the tape and
promote parallel packing.
In this work, we investigated a series of 4,5-disubstituted
derivatives of 2-benzimidazolone to determine how substituents
on the aromatic ring influenced the formation and packing of
tapes. We demonstrate that substituents in these positions direct
both the hydrogen-bonding motif adopted in the solid state and
the packing arrangement of hydrogen-bonded tapes. To mini-
Figure 3. Schematic showing the packing arrangements for hydrogen-
mize free volume in a crystal and maximize interactions between
bonded tapes. (a) Tapes with substituents larger than hydrogen have
aromatic carbons and hydrogens, tapes with scalloped edges
“smooth” edges. These tapes pack with their long axes parallel and
pack in a nonparallel arrangement. Tapes with smooth edges
form sheets that are infinite in two dimensions. (b) Tapes substituted
and few aromatic hydrogens pack in a parallel arrangement.
only with hydrogen have “scalloped” edges and pack with their long
The presence of weak electrostatic interactions between sub-
axes at an angle to one another. This arrangement fills the hollows of
stituents on the 2-benzimidazolones can shift the hydrogen-bond
pattern from a tape motif to a three-dimensional network motif.
These results provide a basis for prediction of the structure of
other derivatives of benzimidazolones in the solid state.
the scalloped edges (bottom) and increases the filling of space. The
parallel packing arrangement leaves empty space between adjacent
tapes.
Results and Discussion
Synthesis and Crystallization of 4,5-Disubstituted Benz-
imidazolones. We prepared a series of benzimidazolones by
the reaction of the appropriate derivative of o-phenylenediamine
with 1,1′-carbonyldiimidazole (Scheme 1). Compound 1-H2 has
been previously synthesized by several alternative routes.28-32
The best solvents for crystallizing 1-H2, 1-(CH3)2, 1-Cl2, 1-Br2,
1-I2, and 2-H2 were polar protic and aprotic solvents: hot
methanol, warm dimethylformamide, pyridine, and N-meth-
Figure 4. Close contact (2.86 Å) between a hydrogen on one aromatic
ring and a carbon on an adjacent aromatic ring in 2-benzimidazolone
(1-H2). This contact is 0.04 Å shorter than the sum of the van der Waals
radii.
(28) Agrawal, V. K.; Sharma, S. Indian J. Chem., Sect. B 1982, 21B,
967.
(29) Almeida, P. S.; Lobo, A. M.; Prabhakar, S. Heterocycles 1989, 28,
653.
(30) Crotti, C.; Cenini, S.; Ragaini, F.; Porta, F.; Tollari, S. J. Mol. Catal.
1992, 72, 283.
(31) Heyman, D. A. J. Heterocycl. Chem. 1978, 15, 573.
(32) Yoshida, T.; Kambe, N.; Murai, S.; Sonoda, N. Bull. Chem. Soc.
Jpn. 1987, 60, 1793.
on one molecule and a carbon atom in the ring of a neighboring
molecule (Figure 4). It is also the arrangement we would expect
on the basis of the scalloped edges. This analysis of the packing
as a combination of closest packing and local electrostatic
interactions suggested a strategy for manipulating crystal
structure by perturbing molecular structure: substitution in the
4 and 5 positions of the aromatic ring with atoms larger than

Engineering the Solid State with 2-Benzimidazolones
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4021

4022 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
Schwiebert et al.
Figure 6. Orientations of molecules: (a) θ is the dihedral angle between the planes of opposing molecules in the tape. Tapes of 1-H2 and 2-H2
are planar (θ ) 0°). Tapes of 1-(CH3)2, 1-Cl2, and 1-Br2 are “buckled” (θ < 0°). (b) θ is the dihedral angle between the planes of opposing
molecules joined by a two-point contact. æ is the dihedral angle between the planes of opposing molecules joined by a one-point contact.
ylpyrolidinone. The initial dissolution of the compounds usually
required a large amount of hot solvent (1-3 L/g of solute).
Subsequent evaporation of low-boiling solvents such as metha-
nol under ambient conditions yielded single crystals as needles,
blocks, or plates. For higher boiling solvents, slow cooling of
the solutions, sometimes after dilution with diethyl ether, gave
the best quality crystals. The presence of solvents with good
hydrogen acceptor and donor sites did not interfere with
formation of the tapes: none of the crystals incorporated
molecules of solvent into their lattice. The solubility of 1-H2,
1-(CH3)2, 1-Cl2, 1-Br2, 1-I2, and 2-H2 in polar solvents such
as THF, CH2Cl2, acetone, and water was negligible. In contrast,
1-F2 was soluble in THF, CH2Cl2, and acetone was crystallized
from concentrated solutions of these solvents upon dilution of
diethyl ether by vapor diffusion at room temperature, and gave
thick rectangular or hexagonal plates. Compound 1-F2 also
crystallized from mixtures of THF or acetone with water upon
slow evaporation of the organic solvent.
Characterization of the Structures of 4,5-Benzimidazolo-
nes in the Solid State. We solved the crystal structures of
1-(CH3)2, 1-Cl2, 1-Br2, 2-H2, 1-F2, and 1-I2 (Table 1) to
evaluate patterns of hydrogen bonds and packing arrangements.
The crystal structure of 1-H2 has been reported,27 and we have
included the data for comparison with the substituted benzimi-
dazolones. All the crystals were monoclinic and members of
the centric family of P21 space groups. Their packing fractions,
(Ck*,33 Table 1) fell in the narrow range of 68-72%, typical
for organic compounds.25 Most of the 2-benzimidazolones
formed tapes, but there were also instances of a three-
dimensional network of hydrogen bonds (3-D). The tapes
packed in two different arrangements, parallel (|) and nonparallel
(×). We discuss the structures in the sections that follow.
1-H2 and 2-H2 Form Tapes That Pack in with Their Long
Axes Nonparallel (×). Like tapes of 1-H2, tapes of 2-H2 have
scalloped edges; thus, we predicted that tapes of 2-H2 would
pack in a nonparallel arrangement to fill the empty spaces at
the edges of the tape. The structure of 2-H2 was indeed closely
related to the known structure of 1-H2 (Figure 5). The tapes of
both compounds were planar (Figure 6), the hydrogen bonds
had similar lengths, and the hydrogen-bond angles were nearly
linear (Table 2). In both compounds, the tapes stacked with
the aromatic rings in one tape offset from those of the tapes
above and below, presumably to minimize repulsion between
π-systems.37-39 The stacks of tapes packed at an angle to one
another with molecules from one tape occupying some of the
space between molecules in an adjacent tape.
The relative orientations of molecules in adjacent stacks of
2-H2 were different than in those in adjacent stacks of 1-H2
(Figure 5a). We did not identify specific interactions responsible
for the difference in relative orientations; the orientations might
be dictated purely by closest packing. In both 1-H2 and 2-H2,
there were interactions between an aromatic hydrogen and an
aromatic carbon (C‚‚‚H). The C‚‚‚H contact in 1-H2 was a short
contact (slightly less than the sum of the van der Waals radii)41
and thus suggested an electrostatic attraction between carbon
and hydrogen (Table 3). The C‚‚‚H contact in 2-H2 was longer
than that in 1-H2 (Table 3). This lengthening suggested that
the C‚‚‚H interaction was less important in the 2-H2 than in
1-H2. These data are consistent with the hypothesis that closest
packing is the most important determinant of the packing for
benzimidazolones with only hydrocarbon substituents.
(34) Cerius2 1.0 molecular modeling program, MSI: Burlington, MA,
1994.
(35) Pluton-92 (Version 28) and Platon-92 (Version 25): A. L. Spek,
Bijvoet Center for Biomolecular Research, Vakgroep Kristal- en Structure-
chemie, University of Utrecht.
(36) Hershenson, F. M.; Bauer, L.; King, K. F. J. Org. Chem. 1968, 33,
2543.
(37) Hunter, C. A.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112,
5525.
(38) Cozzi, F.; Cinquini, M.; Annunziata, R.; Dwyer, T.; Siegel, J. S. J.
Am. Chem. Soc. 1992, 114, 5729.
(39) Cozzi, F.; Cinquini, M.; Annuziata, R.; Siegel, J. S. J. Am. Chem.
Soc. 1993, 115, 5330.
(40) Taylor, R.; Kennard, O. Acc. Chem. Res. 1984, 17, 320.
(41) We used the van der Waals radii given by Bondi (Å): 1.47 (F),
1.75 (Cl), 1.85 (Br), 1.98 (I). See: Bondi, A. J. Phys. Chem. 1964, 68,
441.
(33) We have calculated the packing fractions (Ck*) using molecular
volume with the program Platon (see ref 35). The packing fraction (Ck) is
traditionally calculated using tables of average volume increments. We prefer
the former approach because some functional groups do not have tabulated
values and because molecular volumes depend on the environment.

Engineering the Solid State with 2-Benzimidazolones
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4023
Table 2. Summary of Hydrogen-Bond Distances and Angles
O(1)‚‚‚N(1)a
O(1)‚‚‚N(2)a
angle (deg)b
O(1)‚‚‚H(1)‚‚‚N(1)
angle (deg)b
N(2)‚‚‚H(2)‚‚‚O(1)
angle (deg)c
H(2)‚‚‚O(1)‚‚‚H(1)
class
(Å)
(Å)
1-H2
2-H2
1-(CH3)2
1-Cl2
1-Br2
1-F2
×
×
|
|
|
3-D
3-D
2.84
2.84
2.79
2.78
2.87
2.89
2.85
2.84
2.80
2.86
2.85
2.79
2.82
2.85
176.2
167.4
167.7
171.8
140.8
163.1
164.2
176.2
174.1
167.2
170.9
156.4
161.5
161.8
127.5
124.1
118.2
126.7
117.2
91.2
1-I2
124.1
a The atom in bold belongs to the central molecule highlighted with the box. The atoms in plain text belong to the adjacent molecules. We give
only the distance between the N and O atoms on adjacent hydrogen-bonded molecules, which we believe is subject to less error than distances
involving the hydrogen atoms. The positions of the hydrogen atoms were fixed at idealized positions in some structures and refined in others,
introducing variability in distances. b These values are the angles of three atoms involved in one hydrogen bond. The mean value for amides is
161.2°.40 c These values are the angles between the amide N-H, the carbonyl oxygen, and a second amide N-H.
Table 3. Summary of Selected Intermolecular Close Contact
Distances between the Asymmetric Molecule and Its Neighbors
1-(CH3)2, 1-Cl2, and 1-Br2 Form Tapes That Pack with
Their Long Axes Parallel (|). The crystal structures of
1-(CH3)2, 1-Cl2, and1-Br2 were effectively isostructural (Figure
7). The hydrogen bonds were essentially the same length for
all three tapes (2.78-2.89 Å) (Table 2).42 Unlike the × tapes,
these | tapes were not completely planar: the tapes buckled so
that opposing hydrogen-bonded molecules within a tape lay at
an angle to one another (θ, Figure 6). The buckling may
distance ( sum
alleviate steric interaction between substituents on adjacent
molecules on the same side of the tape or, perhaps, unfavorable
steric interaction between tapes in adjacent layers.43
(Å)a
of VDW atom i‚‚‚atom jb notec electrostaticd VDW
1-H2
2-H2
2.86
2.86
3.07
3.09
3.09
3.07
2.95
2.95
2.53
2.74
2.84
2.53
2.86
2.74
3.43
3.46
3.48
3.46
3.76
3.59
3.54
3.88
3.92
4.07
3.88
4.07
-0.04
-0.04
0.17
0.19
0.19
0.17
0.05
0.05
-0.14
0.07
-0.10
-0.14
0.19
0.07
-0.07
0.19
0.03
0.19
0.06
-0.11
-0.01
-0.09
-0.04
0.11
C(6)‚‚‚H(6)
H(6)‚‚‚C(6)
C(6)‚‚‚H(5)
C(5)‚‚‚H(6)
H(6)‚‚‚C(5)
H(5)‚‚‚C(6)
C(8)‚‚‚H(8)
H(8)‚‚‚C(8)
F(1)‚‚‚H(4)
F(1)‚‚‚H(7)
F(2)‚‚‚F(2)
H(4)‚‚‚F(1)
H(1)‚‚‚F(2)
H(7)‚‚‚F(1)
Cl(2)‚‚‚Cl(2)
Cl(2)‚‚‚O(1)
C(3)‚‚‚Cl(2)
O(1)‚‚‚Cl(2)
Br(1)‚‚‚Br(1)
Br(2)‚‚‚Br(2)
C(3)‚‚‚Br(2)
I(1)‚‚‚I(2)
<
<
‚
‚
Tapes composed of 1-(CH3)2, 1-Cl2, and 1-Br2 packed with
their long axes parallel, in sheets that were infinite in two
dimensions (Figure 7a). The substituents on the edges of one
tape packed in between the substituents on an adjacent tape.
The substituents packed particularly closely in 1-Cl2 and 1-Br2,
where the distances between halogen atoms in adjacent tapes
are slightly shorter than the sum of the isotropic van der Waals
(VDW) radii (Table 3).41,44-46 The sheets stacked on top of
one another (Figure 7b). Compared with × tapes, the | tapes
were offset significantly along their short axes relative to the
tapes above and below (Figure 7c). This offset presumably
prevents unfavorable steric interaction between substituents on
molecules in one layer with substituents on molecules in the
layers above and below. The offset suggests another reason
why benzimidazolones with substituents other than hydrogen
‚
‚
‚
‚
e
1-F2
<
<
<
‚
‚
‚
‚
‚
1-Cl2
1-Br2
<
‚
‚
‚
‚
‚
<
<
<
<
‚
(42) The R factor for 1-Br2 is high because the crystals were twinned.
The errors in distances associated with 1-Br2 are expected to be higher
than distances for the other compounds in Table 2.
f
1-I2
‚
‚
‚
‚
‚
I(1)‚‚‚I(1)
I(1)‚‚‚I(2)
(43) We observed a similar buckling in the rosette (cyclic hexamer)
structures in the (Bar‚Mel(PhX)2) system. Modeling of this system also
indicated that buckling was not a steric requirement for the motif itself,
and we believe that buckling occurs only within the crystal.
(44) Short contacts between chlorines or between bromines in halogenated
aromatics have been reported in many instances and are apparently the
primary influence on crystal structure in some cases (see refs 45 and 46)
In the present series, the fact that the methyl-substituted cyclic urea adopts
the same structural motif as the chloro- and bromo-substituted benzimida-
zolones suggests that chlorine-chlorine and bromine-bromine interactions
do not influence the packing of parallel tapes.
-0.09
0.11
I(2)‚‚‚I(1)
<
I(2)‚‚‚I(1)
a Atoms in bold involve halogen-halogen contacts; atoms in italics
are other halogen-non-halogen contacts and close contacts between
aromatic rings. b Atom i is an atom from the asymmetric molecule.
Atom j is an atom from a neighboring molecule. c Symbols: < denotes
contacts less than the sum of the van der Waals radii. These
calculations used values of the radii from Bondi (Å): 1.47 (F), 1.75
(Cl), 1.85 (Br), 1.98 (I). d The partial charges were derived from
semiempirical calculations (ESP charges using MOPAC6.0) as de-
scribed in the text. e The asymmetric molecule of 1-F2 and its nearest
neighbors are shown in Figure 12. f The asymmetric molecule of 1-I2
and its nearest neighbors are shown in Figure 13.
(45) Desiraju, G. R.; Sarma, J. A. R. P. Proc. Indian Acad. Sci. Chem.
Sci. 1986, 96, 599.
(46) Desiraju, G. R.; Parthasarathy, R. J. Am. Chem. Soc. 1989, 111,
8725.

4024 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
Schwiebert et al.

Engineering the Solid State with 2-Benzimidazolones
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4025
Figure 8. Three views of the packing arrangement for each compound: (a) a simplified schematic showing the relationship of the tape motif to
the network motif, (b) a stereoview of the network, emphasizing the three-dimensional nature of the connectivity of hydrogen bonds, and (c) a
space-filling model of a portion of the network.
cannot form × tapes: tapes of these compounds cannot stack
directly on top of one another to form the stacks observed for
1-H2 and 2-H2.
Benzimidazolones 1-F2 and 1-I2 Form a Network (3-D)
Motif, Rather Than a Tape Motif. In the network motif, each
molecule is hydrogen-bonded through a two-point contact (cyclic
dimer) to one molecule and through one-point contacts to two
other molecules (Figure 8). These networks may be viewed as
“broken tapes,” in which every other set of hydrogen bonds is
broken and rejoined to two different molecules outside the plane
of the tape (Figure 8a). The hydrogen bonds in 1-F2 and 1-I2
were similar in length to the hydrogen bonds in both parallel
and nonparallel tapes (Table 2). The molecules involved in the
two-point contact (dimers) were coplanar (θ ) 0°) while the
molecules involved in the one-point contact lay at angle to one
another (æ, Figure 6).
Figure 9. Energies of the carbonyl stretches in the infrared spectra of
1-X2 and 2-H2. Compounds that crystallize in the 3-D network have
two carbonyl stretches because the environment of the carbonyl is
unsymmetricalsa two-point contact on one side and two one-point
contacts on the other. The energies of the carbonyl stretches of
compounds bearing electron-withdrawing substituents are similar to the
energy of a compound bearing an electron-donating substituent (1-
(CH3)2). This similarity indicates that the influence of the substituent
on the energy of the carbonyl stretch is small compared to the influence
of structural type.
Infrared Spectroscopy Suggests That the Energy of the
Hydrogen Bonds Is Comparable in Tape and 3-D Motifs.
We expected the energy of the hydrogen bonds to be similar
for compounds 1-X2 and 2-H2: the number of hydrogen bonds
is conserved in each structural type, and the lengths and angles
of the hydrogen bonds are similar for all the compounds studied
(Table 2). We further analyzed the strengths of the hydrogen
bonds by infrared spectroscopy (Nujol mulls of crystalline
samples) (Figure 9). The carbonyl stretch has been used as an
indicator of the strength of hydrogen bonds: stronger hydrogen
bonds shift the stretch to lower energy.47 By the criterion of
the carbonyl stretch, hydrogen bonds in the 3-D motif are
stronger than those in the × tapes but comparable to those in
the | tapes. Thus, tapes are not necessarily constituted of
stronger hydrogen bonds than are the 3-D motifs.
Comparison of the Calculated Lattice Energies Also
Indicates That Energies of Hydrogen Bonds Are Similar in
Both the Tape and 3-D Motifs. Analyzing the components
of the lattice energy (Ulatt) of the crystals using computations
is an alternative way of comparing the strengths of the hydrogen
(47) Susi, H. Methods Enzymol. 1972, 26, 381.

4026 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
Schwiebert et al.
Figure 10. Calculated lattice energies (Ulatt) for the benzimidazolones, including the van der Waals and Coulombic components. The portion of
the Coulombic term that results from hydrogen bonds is marked with hatches.
bonds in the crystals. In these calculations, the coordinates of
each molecule from the single crystal data served as the starting
point. The geometry of a single molecule was first optimized
until the gradient in its energy was below 0.01 kcal/mol using
the semiempirical AM1 Hamiltonian in MOPAC6,48 electrostatic
potential (ESP) partial charges were assigned,49 and this final
geometry was used for further computations. To determine Ulatt,
we used the Cerius2 modeling program34 and the Drieding force
field (with the ESP charges from MOPAC6.0).50 The van der
Waals interactions were truncated at 8 Å, and the Ewald
summation technique was used for Coulombic interactions in
the crystal.51,52 The appropriate crystalline system for each
molecule was built in Cerius2, and the energy of the crystal
was minimized until the gradient in energy was below 0.01 kcal/
mol. The values of Ulatt were calculated from this minimized
system, and the contributions from van der Waals and electro-
statics analyzed (Figure 10). The results of Ulatt suggest that
the contribution from the hydrogen bondssa component of the
electrostaticssis similar for all the molecules regardless of the
motif. These data from computations, and those from infrared
spectroscopy, suggest that the preference of a given molecule
for a particular motif is based on interactions involving
substituents, rather than strengths of hydrogen bonds.
Figure 11. Bulky substituents might interact unfavorably in the tape
motif and consequently favor the network motif. The graph shows (a)
the total interaction energy (IE) between a central cyclic urea (black)
and its four closest hydrogen-bonded neighbors (hatched) within a tape
and (b) the van der Waals component of the IE between a central cyclic
urea and its two adjacent neighbors.
Computations Suggest That Tapes of 1-F2 and 1-I2 Are
Not Disfavored Outside the Crystalline Environment. We
originally speculated that the failure of 1-F2 and 1-I2 to form
tapes was due to unfavorable interactions within the tape. For
example, electrostatic interactions between C-F bond dipoles
and steric interactions between adjacent iodine atoms in different
molecules might both be sufficiently unfavorable that, in
principle, they destabilized the tapes. To address the question
of stability of the tapes outside the environment of the crystalline
lattice, we have carried out molecular mechanics calculations
on fragments of tapes. In QUANTA 3.353 five molecules of
1-H2 were built and arranged into a tape. The molecules were
constrained to be coplanar, with an initial value of 2.8 Å for
the N‚‚‚OdC distance, and the energy of this tape was
minimized until its gradient reached 0.01 kcal/mol. Using this
energetically minimized tape as a template, we added the various
substituents of interest, adjusted the geometries appropriately,
and calculated the total interaction energy (IE) of the central
cyclic urea with the four neighboring molecules composing the
rest of the tape (Figure 11). These IE data showed that a tape
of 1-F2 is as favorable as a tape of 1-H2. Moreover, a tape of
1-I2 is more favorable than tapes of any other compound.54 These
data suggest that tapes of 1-F2 and 1-I2 are not destabilized
relative to other tapes but that these compounds are influenced
by interactions outside of a tape motif.
(48) MOPAC6.0: J. J. P. Stewart, QCPE, 455 (V. 6.0). Mopac-ESP:
Besler, B. H.; Merz, K. M.; Kollman, P. A. J. Comput. Chem. 1990, 11,
431.
(49) We calculated the ESP charges using MOPAC6.0: -0.01e (F),
+0.16e (Cl), +0.16e (Br), +0.17e (I).
(50) Dreiding II: Mayo, S. L.; Olafson, B. D.; Goddard, W. A. J. Phys.
Chem. 1990, 94, 8897.
(51) Karasawa, N.; Goddard, W. A. J. Phys. Chem. 1989, 93, 7320.
(52) Ewald, P. P. Ann. Phys. (Leipzig) 1921, 64, 253.
(53) QUANTA 3.3 molecular modeling program, MSI: Burlington, MA,
1992.

Engineering the Solid State with 2-Benzimidazolones
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4027
of 1-F2. We note that the interaction between fluorine and
hydrogen need not be very strong, since the preference for the
tape motif (if it exists at all) is weak.
Molecules of 1-I2 Have Five Short Iodine-Iodine Contacts
in the 3-D Motif. Three of these five short contacts (I‚‚‚I) are
less than the sum of the van der Waals radii (Table 3, Figure
13). The 3-D motif observed for 1-I2 has more short contacts
between halogens than does the | tape motif observed for 1-Cl2
and 1-Br2. The parallel tapes can only form contacts between
halogen atoms in the plane of the tape. Halogen-halogen
interactions become more important as the polarizability
increases: 1-Cl2 has one contact, 1-Br2 has two, and 1-I2 has
five (three of which are short contacts). We propose that van
der Waals interaction between iodine atoms is sufficiently strong
to be an important factor in the selectivity of the motif. The
van der Waals interactions between chlorine atoms and bromine
atoms are not strong enough to overcome what is otherwise a
preference for tapes.
We Have Not Detected Polymorphism in This Series of
Benzimidazolones. We wished to know whether the observed
structures represented a global minimum in free energy for the
system or whether there were polymorphs of equal or lower
energies. To determine whether polymorphism occurred in this
series of benzimidazolones, we examined crystals from each
structural class using X-ray powder diffraction (XPD). The
XPD pattern provides a fingerprint for a specific crystalline
lattice: all powders having the same XPD pattern have identical
crystal structures.61,62 We deliberately attempted to induce the
formation of polymorphs by crystallizing selected compounds
from each structural class (1-(CH3)2 (|), 1-H2 (×), and 1-F2
(3-D)) from four different solvents. The compounds crystallized
upon slow evaporation of the solvent under ambient conditions,
diffusion of ether into the solution at room temperature, or rapid
cooling of the solution. We obtained crystals of varying size
and shape depending on the conditions of crystallization; XPD
showed, however, that the structures of these crystals were the
same.
For all compounds, the agreement between the position of
the peaks in the experimental traces and the calculated trace
was good and indicated that only one solid phase crystallized
under a variety of conditions.63 For 1-(CH3)2, the relative
intensities of the calculated and measured peaks varied consid-
erably for peaks corresponding to certain hkl planes. This
variation in relative intensity of the peaks suggests that a
preferred orientation may occur in the crystalline sample used
for XPD.64
We were unable to induce a change of phase in the
benzimidazolones upon heating. A powdered sample of a
member of each class (1-H2, 1-F2, and 1-(CH3)2) was heated in
an open glass vial for 3 days at 160 °C. During this time, some
crystals grew on the sides of the vial in the samples of 1-F2
and 1-H2 as a result of sublimation, but the sample of 1-(CH3)2
remained unchanged. After cooling, the samples were reground
and the XPD pattern recorded again: the XPD patterns remained
unchanged. The XPD pattern of these three molecules did not
change for samples grown by sublimation (150 °C, 10-3 Torr).
The absence of polymorphism in this series of benzimidazolones
under the conditions we have tested suggests that these
Figure 12. asymmetric molecule of 1-F2 with H‚‚‚F contacts. The
fluorine atoms are shaded.
The Selectivity for the 3-D Motif over That of the Tape
Motif in 1-F2 and 1-I2 Seems To Be Due to Interactions
Involving the Substituents. We sought to identify secondary
interactions for 1-F2 and 1-I2 that stabilized a 3-D network
relative to a tape. The calculated lattice energies (Figure 10)
indicate that Coulombic interactions are more important than
van der Waals interactions for molecules with small substituents
(1-H2 and 1-F2). In contrast, van der Waals interactions are
more important than Coulombic interactions for molecules with
large substituents (1-(CH3)2, 1-Cl2, 1-Br2, and 1-I2). These
observations are consistent with a number of analyses of the
Cambridge Structural Database.46,55-58 These analyses indicate
that halogens can form short contacts to other halogens and to
polar hydrogens. These short contacts have been rationalized
in terms of specific intermolecular interactions. Interactions
between halogens and polar hydrogens are electrostatic, so the
strength of the interaction increases with the ratio of charge to
mass (I < Br < Cl < F). The nature of the interactions between
two halogens is a subject of debate.46,59 It is known, however,
that the relative order of favorable halogen-halogen interactions
is F < Cl < Br < I.46 The polarizability of iodine, and hence
its dispersion energy, is significantly greater than those of the
other halogens.58,60 Therefore, short contacts between iodine
atoms should increase the van der Waals interaction contribution
to the lattice energy.
Crystals of 1-F2 Show Short Contacts between Fluorine
and Hydrogen Atoms. Each molecule of 1-F2 forms two short
contacts (less than the sum of the VDW radii) between fluorine
atoms and aromatic hydrogens (F‚‚‚H) (Figure 12). None of
the other benzimidazolones have short contacts between halo-
gens and aromatic hydrogens. Semiempirical calculations of
the partial charges on the atoms in the benzimidazolones suggest
that only fluorine is sufficiently electronegative to bear a partial
negative charge;48,49 therefore, only fluorine can interact favor-
ably with an aromatic C-H bond. We propose that the
favorable electrostatic interaction between fluorine and hydrogen
is a factor that favors the 3-D motif over the tape for crystals
(54) The assertion that 1-I2 is not too large laterally to allow the formation
of a tape is further supported by the following observation: the distance
between bromine atoms on the same side of a tape of 1-Br2 is 4.0 Å. The
I-I contacts in the observed network structure are shorter (3.98 Å) than
the 4.0 Å required, so replacing the bromine atoms in 1-Br2 with iodine
atoms is not precluded by unfavorable steric interactions.
(55) Kumar, V. A.; Begum, N. S.; Venkatesan, K. J. Chem. Soc., Perkin
Trans. 2 1993, 463.
(56) Murray-Rust, P.; Motherwell, W. D. S. J. Am. Chem. Soc. 1979,
101, 4374.
(57) Murray-Rust, P.; Stallings, W. C.; Monti, C. T.; Preston, R. K.;
Glusker, J. P. J. Am. Chem. Soc. 1983, 105, 3206.
(61) Bish, D. L., Post, J. E., Ed. Modern Powder Diffraction; Mineralogi-
cal Society of America: Washington, DC, 1989.
(62) Klug, H. P.; Alexander, L. E. X-Ray Diffraction Procedures; John
Wiley & Sons: New York, 1954.
(63) The supporting information includes a figure showing the X-ray
powder diffraction patterns for 1-(CH3)2 (|), 1-H2 (×), and 1-F2 (3-D).
Each compound was crystallized under six conditions and the powder
patterns were recorded and compared with the calculated pattern.
(64) Preferred orientation is well-known to cause variations in relative
intensities (see refs 61 and 62).
(58) Ramasubbu, N.; Parthasarathy, R.; Murray-Rust, P. J. Am. Chem.
Soc. 1986, 108, 4308.
(59) Price, S. L.; Stone, A. J.; Lucas, J.; Rowland, R. S.; Thornley, A.
E. J. Am. Chem. Soc. 1994, 116, 4910.
(60) Nyburg, S. C. Acta Crystallogr. 1979, A35, 641.

4028 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
Schwiebert et al.
Figure 13. Stereoview of the asymmetric molecule of 1-I2 with I‚‚‚I contacts. The iodine atoms are shaded.
compounds crystallize in a well-defined minimum of free
energy; we acknowledge, however, that failure to detect
polymorphism does not preclude the possibility that polymorphs
of these benzimidazolones exist.
morphism in the 2-benzimidazolones by eliminating the pos-
sibility for conformational isomerism. The lack of polymor-
phism in this system allows us to compare only structures of
minimum free energy. Finally, benzimdazolones tend to
crystallize as blocks or plates, rather than as the thin needles
that plagued the (Bar‚Mel(PhX)2) system. The hydrogen bonds
in the benzimidazolones are weaker and fewer in number than
those in the (Bar‚Mel(PhX)2); thus, the rate of growth along
the long axis of the tape is comparable to that in other
dimensions.
One complication in this system relative to (Bar‚Mel(PhX)2)
is that the formation of tapes is less reliable. Various interac-
tions introduced through substituents can promote a switch from
one hydrogen-bonded motif to another. In the (Bar‚Mel(PhX)2)
system, a switch in motifs occurs to relieve steric interactions
between molecules within the tape. In the benzimidazolones,
a switch in motifs occurs to maximize electrostatic or van der
Waals interactions between molecules within the crystalline
enVironment. That is, there is nothing inherently unstable about
the tape itself, but interactions between tapes are sometimes less
favorable than interactions between molecules in the 3-D motif.
For the benzimidazolones, the interactions that shift the motif
from tape to 3-D can be weak and thereby complicate the
prediction of the motif.
We believe that the benzimidazolones are a useful model
system with which to approach structural design in the solid
state. Derivatives of benzimidazolones with a wide range of
substituents are relatively easy to synthesize (although the
diamines from which they are derived can be synthetically
challenging). Careful crystallization yields single crystals,
although the low solubility of some of the compounds makes
crystallization nonroutine. The simplicity (planarity, confor-
mational rigidity, and lack of conformational polymorphism)
of these molecules also makes them more amenable to com-
putational analysis than the (Bar‚Mel(PhX)2) system.65 Our
results emphasize the importance of identifying the effects of
secondary interaction on the structure of organic molecules in
crystals.
Conclusions
This work demonstrates that this series of disubstituted
benzimidazolones pack in a limited set of crystalline forms (|,
×, and 3-D). Our results suggest that there is a propensity for
the benzimidazolones to form tapes in the solid state, although
the 3-D network is preferred in some cases. In this system, the
tape and the 3-D network are competing motifs that are
comparable in energy; thus, influences such as weak electrostatic
interactions and interactions between halogens seem to be
important in determining the motif. Although this situation
complicates prediction of the motif, it has allowed us to identify
some secondary intermolecular interactions (I‚‚‚I and F‚‚‚H
contacts) that seem to influence the crystal structure. On the
basis of our results, we expect that molecules with other
substituents that are highly polarizable (permitting favorable
VDW interactions) or polarized (permitting favorable electro-
static interactions) will crystallize in a 3-D motif.
For the benzimidazolones that do form tapes, we can
rationalize the packing of tapes in three dimensions only on
the basis of molecular structure: the substitution at the edge of
the tape determines whether the tapes crystallize in a parallel
or nonparallel arrangement. Tapes of benzimidazolones without
substituents at the edges tend to crystallize in the nonparallel
arrangement because it fills the spaces between adjacent
molecules in the tape. For the parent compound 1-H2, the
nonparallel arrangement also permits a favorable electrostatic
interaction between a carbon in the aromatic ring and an
aromatic hydrogen from an adjacent tape. Substitution of the
aromatic ring with atoms larger than hydrogen provides smooth
edges that cannot intercalate in a nonparallel arrangement. This
substitution also replaces aromatic hydrogens involved in the
electrostatic interaction favoring the nonparallel arrangement.
Thus, we have begun to define a relationship between the
molecular structure and the crystal structure of benzimidazo-
lones. Our results suggest that analysis of interactions at the
edges of tapes provides the best chance for predicting the
packing arrangement of the tapes.
Experimental Section
General Methods. Dinitroaniline, 1,2-phenylenediamine, 4,5-
dichloro-1,2-phenylenediamine, 4,5-dimethyl-1,2-phenylenediamine,
4,5-difluoro-2-nitroaniline, and 1,2-diaminonaphthalene (Aldrich), and
1,1′-carbonyldiimidazole (Sigma) were used as received without
purification. Reagent-grade THF was distilled from sodium benzophe-
none ketyl; all other solvents were reagent grade and were used without
Compared with the melamine/barbituric acid system, the
benzimidazolone system offers several important improvements
in designing the solid state. Most significantly, we believe that
we can rationalize the packing of tapes of benzimidazolones
on the basis of the substitution at the edge of the tapes. Whether
we can predict the packing of new molecules still remains to
be tested. Furthermore, we have successfully avoided poly-
(65) The many torsions of the phenyl rings in the Bar‚Mel(PhX)2 made
modelling difficult because there were multiple conformations that gave
nearly isoenergetic structures. We expect computations to be most successful
in systems having well-defined minima of free energy.

Engineering the Solid State with 2-Benzimidazolones
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4029
2-Naphthimidazolone (2-H2). From 0.982 of 1,2-di-
purification. Melting points were determined on a Mel-Temp apparatus
and are uncorrected.
g
1
aminonaphthalene: 0.913 g, 80%; H NMR (100 MHz, DMSO-d6) δ
10.79 (br s, 2H), 7.79 (m, 2H), 7.29 (m, 2H), 7.26 (m, 2H); 13C NMR
(400 MHz, DMSO-d6) δ 156.22, 131.09, 129.24, 126.85, 123.26,
103.60; HRMS (M+) calcd for C11H8N2O 184.0637, found 184.0640.
Crystallization of Single Crystals of Benzimidazolones for X-ray
Crystallography. 1-H2, 1-Br2, 1-I2, and 2-H2: crystals were grown
by evaporation of a solution of the compound in methanol in a beaker
covered with filter paper. 1-H2 gave colorless blocks (approximately
0.2 × 0.2 × 0.1 mm). 1-Br2 gave colorless plates (approximately 0.02
× 0.10 . 0.02 mm). 1-I2 gave colorless plates (approximately 0.35 ×
0.15 × 0.03). 2-H2 gave gold rods (approximately 0.15 × 0.02 ×
0.02 mm). 1-(CH3)2 and 1-Cl2: crystals were grown by cooling warm
solutions in DMF. 1-(CH3)2 gave gold blocks (approximately 0.25 ×
0.22 × 0.10 mm). 1-Cl2 gave gold blocks (approximately 0.25 × 0.07
× 0.11 mm). 1-F2: thick gold plates were grown by diffusion of diethyl
ether into a solution in THF (approximately 0.35 × 0.25 × 0.08).
Determination of Crystal Structure by X-ray Crystallography.
The details of X-ray data collection and structure solution and
refinement are provided in the supporting information. Data were
collected by Molecular Structure Corp., The Woodlands, TX, on a
Rigaku AFC5R diffractometer equipped with a rotating anode generator.
All structures were solved and refined at Harvard using the SHELX-
PLUS package of programs.
X-ray Powder Diffraction. The powder diffraction patterns were
collected on a Scintag XDS 2000 diffractometer using CuKR1 λ 1.5405
radiation. The samples were prepared by grinding the crystals with a
mortar and pestle and spreading the powder on a microscope slide.
Computational Methods. All energy calculations used the Drieding
II parameters50 for van der Waals interactions and electrostatic potential
charges (ESP) derived from MOPAC6. Starting from the geometries
of the crystal structures, MOPAC6 was used to optimize the bonds
and angles. Calculations of the lattice energies were done using
Cerius2. CHARMm 2270 was used to calculate the interaction energies
in these tapes using values of the VDW parameter and partial charges
described above.
Preparation of 4,5-Substituted 1,2 Phenylenediamines. 4,5-
Diiodo-1,2-phenylenediamine was prepared by nitration of dinitroa-
niline, followed by reduction with Sn/HCl.66 4,5-Difluoro-1,2-
phenylenediamine was prepared by reduction of 4,5-difluoro-2-
nitroaniline with SnCl2/HCl.67 4,5-Dibromo-1,2-phenylenediamine was
prepared by bromination of N,N′-bis(p-toluenesulphonyl)-o-phenylene-
diamine.68
General Procedure for Carbonylation of 4,5-Substituted 1,2-
Phenylenediamines.69 A solution of 1,1′-carbonyldiimidazole (1.2
equiv, 0.4 M in THF) was added in drops with stirring at room
temperature to a solution of the appropriate diamine (1 M in THF),
and the mixture was stirred for 3 days. Compounds 1-H2, 1-Cl2,
1-(CH3)2, 1-Br2, 1-I2, and 2-H2 precipitated from solution as they were
formed and were isolated by filtration. The product was washed with
THF and then dried in Vacuo. Solutions of 1-Br2, 1-I2, and 2-H2 in
DMF were treated with decolorizing charcoal prior to recrystallization.
Compound 1-F2 was soluble in THF and did not precipitate. It was
isolated by crystallization from THF/water, upon slow evaporation of
THF.
2-Benzimidazolone (1-H2).36 From 5.0135 g of 1,2-phenylene-
diamine: 2.0181 g, 47%; 1H NMR (400 MHz, DMSO-d6) δ 10.60 (br
s, 2H), 6.91 (s, 4H (two overlapping signals)); 13C NMR (100 MHz,
DMSO-d6) δ 160.57, 134.88, 125.70, 113.78; HRMS (M) calcd for
C7H6N2O 134.0480, found 134.0482.
4,5-Dimethyl-2-benzimidazolone (1-(CH3)2). From 4.9924 g of
1
4,5-dimethyl-1,2-phenylenediamine: 1.8445 g, 44%; H NMR (400
MHz, DMSO-d6) δ 10.33 (br s, 2H), 6.69 (s, 2H); 13C NMR (100 MHz,
DMSO-d6) δ 155.41, 127.75, 127.74, 109.54, 19.39; HRMS (M +
NH4+) calcd for C9H10N2O 180.1137, found 180.1147.
4,5-Dichloro-2-benzimidazolone (1-Cl2). From 4.5999 g of
4,5-dichloro-1,2-phenylenediamine: 2.4791 g, 52%; 1H NMR (400
MHz, DMSO-d6) δ 10.91 (br s, 2H), 7.09 (s, 2H); 13C NMR (100 MHz,
DMSO-d6) δ 155.11, 129.80, 122.30, 109.70; HRMS (M) calcd for
C7H4Cl2N2O 201.9701, found 201.9697.
4,5-Difluoro-2-benzimidazolone (1-F2). From 4,5-difluoro-1,2-
phenylenediamine: 1H NMR (400 MHz, DMSO-d6) δ 10.78 (br s, 2H),
6.98 (t, JH-F ) 8.8 Hz, 2H); 13C NMR (100 MHz, DMSO-d6) δ 156.05,
145.05 (d, JC-F ) 219.9), 125.59, 98.25; HRMS (M+) calcd for
C7H4F2N2O 170.0291, found 170.0295.
4,5-Diiodo-2-benzimidazolone (1-I2). From 352 mg of
4,5-diiodo-1,2-phenylenediamine: 197 mg, 54%; 1H NMR (400 MHz,
DMSO-d6) δ 10.79 (br s, 2H), 7.39 (s, 2H); 13C NMR (100 MHz,
DMSO-d6) δ 154.59, 131.45, 118.00, 96.41; HRMS (M+) calcd for
C7H4I2N2O 385.8413, found 385.8410.
Acknowledgment. We acknowledge the support of the
National Science Foundation through Grant CHE-91-22331.
J.C.M. wishes to acknowledge Merck for a postdoctoral
fellowship. We thank Professor Andrew Barron for helpful
discussions, especially concerning X-ray powder diffraction.
Supporting Information Available: Crystallographic detail
including tables of atomic positional parameters and bond
lengths and angles (33 pages). This material is contained in
many libraries on microfiche, immediately follows this article
in the microfilm version of the journal, can be ordered from
the ACS, and can be downloaded from the Internet; see any
current masthead page for ordering information and Internet
access instructions.
4,5-Dibromo-2-benzimidazolone (1-Br2). From 1.035
g of
4,5-dibromo-1,2-phenylenediamine: 0.628 g, 55%; 1H NMR (100 MHz,
DMSO-d6) δ 10.91 (br s, 2H), 7.21 (s, 2H); 13C NMR (400 MHz,
DMSO-d6) δ 154.99, 130.65, 113.92, 112.69; HRMS (M+) calcd for
C7H4Br2N2O 289.8690, found 289.8701.
(66) Arotsky, J.; Butler, R.; Darby, A. C. J. Chem. Soc. (C) 1970, 1480.
(67) Uchida, M.; Morita, S.; Chihiro, M.; Kanbe, T.; Yamasaki, K.;
Yabuuchi, Y.; Nakagawa, K. Chem. Pharm. Bull. 1989, 37, 1517.
(68) Cheesman, G. W. H. J. Chem. Soc. (C) 1962, 1170.
JA952836L
(70) Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States, D. J.;
(69) Felix, A. M.; Fryer, R. I. J. Heterocycl. Chem. 1968, 5, 291.
Swaminathan, S.; Karplus, M. J. Comput. Chem. 1983, 4, 187.