Isostructurality, polymorphism and mechanical properties of some hexahalogenated benzenes: The nature of halogen...halogen interactions

Article abstract of DOI:10.1002/chem.200500983

The nature of intermolecular interactions between halogen atoms, X...X (X = Cl, Br, I), continues to be of topical interest because these interactions may be used as design elements in crystal engineering. Hexahalogenated benzenes (C₆Cl_6-n

Full text of DOI:10.1002/chem.200500983

DOI: 10.1002/chem.200500983
Isostructurality, Polymorphism and Mechanical Properties of Some
Hexahalogenated Benzenes: The Nature of Halogen···Halogen Interactions
C. Malla Reddy,[a] Michael T. Kirchner,[a] Ravi C. Gundakaram,[b]
K. Anantha Padmanabhan,*[c] and Gautam R. Desiraju*[a]
Abstract: The nature of intermolecular
interactions between halogen atoms,
X···X (X=Cl, Br, I), continues to be of
topical interest because these interac-
tions may be used as design elements
in crystal engineering.Hexahalogenat-
ed benzenes (C6Cl6ÀnBrn, C6Cl6ÀnIn,
C6Br6ÀnIn) crystallise in two main pack-
ing modes, which take the monoclinic
space group P21/n and the triclinic
secutively in
through the crystal, similar to what has
been suggested for the motion of dislo-
a
snakelike motion
bumps in one layer fit into the hollows
of the next in a space-filling manner.
The triclinic crystals shear on applica-
tion of a mechanical stress only along
the plane of deformation.This shearing
arises from the sliding of layers against
one another.Nonspecificity of the
weak interlayer interactions here is
demonstrated by the structure of twin-
ned crystals of these compounds.One
of the compounds studied (1,3,5-tribro-
mo-2,4,6-triiodobenzene) is dimorphic,
adopting both the monoclinic and tri-
clinic structures, and the reasons for
polymorphism are suggested.To sum-
marise, both chemical and geometrical
models need to be considered for X···X
interactions in hexahalogenated ben-
zenes.The X ···X interactions in the
monoclinic group are nonspecific,
whereas in the triclinic group some
X···X interactions are anisotropic,
chemically specific and crystal-struc-
ture directing.
cations.The bending of C Cl6 crystals is
6
related to the weakness of the Cl···Cl
interactions compared with the stron-
ger p···p stacking interactions.The tri-
clinic packing is less common and is re-
stricted to molecules that have a sym-
metrical (1,3,5- and 2,4,6-) halogen sub-
stitution pattern.This packing type is
characterised by specific, polarisation-
induced X···X interactions that result
in threefold-symmetrical X3 synthons,
especially when X=I; this leads to a
layered pseudohexagonal structure in
which successive planar layers are in-
version related and stacked so that
¯
space group P1.The former, which is
isostructural to C6Cl6, is more common.
For molecules that lack inversion sym-
metry, adoption of this monoclinic
structure would necessarily lead to
crystallographic disorder.In C Cl6, the
6
planar molecules form Cl···Cl contacts
and also p···p stacking interactions.
When crystals of C6Cl6 are compressed
mechanically along their needle length,
that is, [010], a bending deformation
takes place, because of the stronger in-
teractions in the stacking direction.
Further compression propagates con-
Keywords: crystal engineering
crystal plasticity
intermolecular
·
·
·
·
halogens
interactions
supramolecular synthons
Introduction
It has long been noted that recognition between molecules
during crystallisation is governed by geometrical or chemical
factors, that is, because of shape complementarity and size
compatibility (short-range repulsion),[1,2] or specific aniso-
tropic interactions of electrostatic or polarisation origin
(long-range attraction).[3,4] In the process of minimisation of
the total free energy, a balance between these geometrical
and chemical factors is reached.[5] But, where exactly does
this balance lie? Do atoms approach one another more be-
cause of a need for a good geometrical fit or because of par-
[a] C.M.Reddy, Dr.M.T.Kirchner, Prof.G.R.Desiraju
School of Chemistry, University of Hyderabad
Hyderabad 500 046 (India)
Fax : (+91)40-2301-0567
E-mail: gautam_desiraju@yahoo.com
[b] Dr.R.C.Gundakaram
International Advanced Research Centre for Powder
Metallurgy and New Materials
Balapur P.O., Hyderabad 500 005 (India)
ticular chemical effects?[6] This question is especially diffi-
[c] Prof.K.A.Padmanabhan
School of Physics, University of Hyderabad
Hyderabad 500 046 (India)
Fax : (+91)40-2301-1777
[7]
À
À
cult to answer for C X···X C interactions (X=Cl, Br, I)
in crystals because they seem to be of two types based on
À
E-mail: ananthaster@gmail.com
the values of the two C X···X angles, q1 and q2.[7b,e] The
2222
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2222 – 2234

FULL PAPER
type-I interactions (q1 ꢀq2) represent close packing of X
atoms in a geometrical model because identical portions of
the halogen atoms make the nearest approach.The type-II
interactions (q1 ꢀ1808, q2 ꢀ908) are understood on the basis
of X-atom polarisation, X(d+)···X(dÀ), and represent a chemi-
cal model with each halogen atom polarised positively in
the polar region and negatively in the equatorial region
(Scheme 1).Type-II interactions are included in a larger cat-
egory of X(d+)···Y(dÀ) “halogen bonds”, so termed because
an electrophilic halogen is involved.[8]
Whether these layered structures arise on account of type-II
halogen interactions or the ubiquitous p···p stacking interac-
tions[14] is hard to say.However, there is a fundamental dis-
tinction between stacked (two-dimensional) and cross-linked
(three-dimensional, Kitaigorodskii type) structures with re-
spect to the nature of the intermolecular interactions.We
maintain that the default packing for all organic molecular
solids is the three-dimensional Kitaigorodskii packing.
Therefore when a stacked structure is obtained, there must
be specific reasons for its formation.[4]
In this paper, we report the structural chemistry and some
mechanical properties[15] in a series of hexahalogenated ben-
zenes.Our measurements on these crystals enable us to iden-
tify geometrical and chemical features of X···X interactions.
We show that for a realistic description of these crystals, both
chemical and geometrical factors need to be considered.
Results and Discussion
The monoclinic structure type—Hexachlorobenzene: We ex-
amined 16 hexahalogenated benzenes C6X6 (X=Cl, Br, I) in
this study (Scheme 2) and also some of their mixed crys-
tals.[16] These compounds adopt two broad packing modes,
triclinic and monoclinic (Table 1).The monoclinic forms are
Scheme 1.
Difficulties in understanding the nature of X···X interac-
tions also arise from the fact that the halogens are of low to
high electronegativity (I to F) and polarisability (F to I).
They act, depending on the circumstances, as either electro-
positive or electronegative entities in an intermolecular in-
teraction, or in some cases with no particular electrostatic
character.[9] Iodine is somewhat easier to understand in the
context of halogen bonding when compared to the other
halogens because it is more readily polarised as I(d+).Ac-
cordingly, contacts such as I···Cl and I···Br may be represent-
ed as I(d+)···Cl(dÀ) and I(d+)···Br(dÀ) and they generally have
the type-II geometry.Fluorine is very hard and nonpolarisa-
ble, and it is still not really possible to deduce the nature of
its interactions with other halogens;[10] we have stated else-
where that the F···F interaction is not really viable.[11] Per-
haps the I···F interaction is polarisation induced.Chlorine
and bromine belong to an intermediate region and various
researchers, subscribing to one of the two models given
above, have attributed the observed geometries of X···X
contacts to either a van der Waals (nonspherical atoms) or
polarisation (d+···dÀ) character of the interaction.[7] What is
possible is that the type-I contacts formed by Cl and Br are
of the van der Waals variety whereas the type-II contacts
are polarisation induced.If all contacts in a crystal were of
the van der Waals type, one would expect a greater degree
of isotropy in the packing—a Kitaigorodskii solid.[1] As the
intermolecular contacts acquire some distinctiveness, aniso-
tropy enters the crystal with a change in properties.[4] In a
one-dimensional structure involving halogen atoms, mole-
cules are held relatively strongly in this one direction.[12] In
a layered or two-dimensional crystal structure, the interac-
tions within a layer (intralayer) are stronger and more direc-
tional than the interactions between layers (interlayer).[13]
Scheme 2.
Chem. Eur. J. 2006, 12, 2222 – 2234
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
2223

K.A.Padmanabhan, G.R.Desiraju et al.
Table 1.Crystallographic data and structure refinement parameters.
C6Br6
C6Cl6
124B356I
124C356I
124B356C
123B456C
formula
crystal system
space group
a [Š]
b [Š]
c [Š]
C6Br6
monoclinic
P21/n
8.3262(16)
3.9491(8)
15.271(3)
90
C6Cl6
monoclinic
P21/n
7.967(3)
3.7609(14)
14.670(5)
90
C6Br3I3
monoclinic
P21/n
8.5007(6)
4.0812(3)
15.6911(10)
90
C6Cl3I3
monoclinic
P21/n
8.7025(7)
4.1087(3)
15.0989(12)
90
C6Br3Cl3
monoclinic
P21/n
8.1454(8)
3.8704(4)
15.0160(15)
90
C6Br3Cl3
monoclinic
P21/n
8.2578(17)
3.9429(8)
15.137(3)
90
a [8]
b [8]
92.929(3)
90
501.46(17)
2
92.459(6)
90
439.2(3)
2
92.6560(10)
90
543.79(7)
2
93.5890(10)
90
538.82(7)
2
92.385(2)
90
472.98(8)
2
92.34(3)
90
492.46(17)
2
g [8]
V [Š3]
Z
l [Š]
0.71073
3.653
492
0.71073
2.153
276
0.71073
4.229
600
0.71073
3.446
492
0.71073
2.936
384
0.71073
2.82
384
1calcd [gcmÀ3
F[000]
]
m [mmÀ1
q [8]
index ranges
]
23.957
2.73–26.07
À10ꢁhꢁ10
À4ꢁkꢁ4
À16ꢁlꢁ18
100(2)
0.0218
0.0519
0.0317
56
1.884
19.588
2.67–26.07
À10ꢁhꢁ10
À4ꢁkꢁ5
À19ꢁlꢁ19
100(2)
0.0199
0.0492
0.0253
68
9.388
13.575
2.72–26.03
À10ꢁhꢁ10
À4ꢁkꢁ4
À16ꢁlꢁ18
100(2)
0.0162
0.0397
0.0301
67
13.038
2.69–28.13
À10ꢁhꢁ10
À4ꢁkꢁ4
À19ꢁlꢁ9
293(2)
0.0765
0.1994
0.0759
73
2.78–26.03
À9ꢁhꢁ9
À4ꢁkꢁ4
À18ꢁlꢁ18
100(2)
0.0449
0.1097
0.0317
55
2.63–26.03
À10ꢁhꢁ10
À5ꢁkꢁ4
À18ꢁlꢁ18
100(2)
0.0195
0.0469
0.0159
68
T [K]
R1
wR2
Rmerge
parameters
GOF
1.098
1.15
1.093
1.105
1.09
1.013
reflns total
unique reflns
obsd reflns
CCDC[a]
4469
986
913
257165
3112
861
812
257166
5470
1059
1015
281004
4377
1051
1033
280886
6318
923
880
280885
2542
1075
822
280883
123C456I
14C2356I
14B2356I
14B2356C
1245C36I-M1
1245C36I-M2
formula
crystal system
space group
a [Š]
b [Š]
c [Š]
C6Cl3I3
monoclinic
P21/n
8.720(4)
4.1170(19)
15.582(7)
90
C6Cl2I4
monoclinic
P21/n
8.7829(19)
4.2989(9)
15.320(3)
90
C6Br2I4
monoclinic
P21/n
9.3109(19)
4.2609(9)
14.776(3)
90
C6Br2Cl4
monoclinic
P21/n
8.0578(6)
3.8484(3)
14.9122(11)
90
C6Cl4I2
monoclinic
P21/n
8.6484(6)
4.1259(3)
15.7424(11)
90
C6Cl4I2
monoclinic
P21/n
6.5276(6)
5.9682(5)
13.3913(11)
90
a [8]
b [8]
92.667(7)
90
558.8(4)
2
94.908(4)
90
576.3(2)
2
93.33(3)
90
585.2(2)
2
92.1900(10)
90
462.08(6)
2
92.1300(10)
90
561.34(7)
2
98.7720(10)
90
515.60(8)
2
g [8]
V [Š3]
Z
l [Š]
0.71073
3.323
492
0.71073
3.749
564
0.71073
4.197
636
0.71073
2.686
348
0.71073
2.767
420
0.71073
3.012
420
1calcd [gcmÀ3
F[000]
]
m [mmÀ1
q [8]
index ranges
]
9.053
11.225
2.58–27
À6ꢁhꢁ11
À5ꢁkꢁ5
À19ꢁlꢁ19
293(2)
0.0397
0.1055
0.0662
56
17.426
2.52–29.06
À12ꢁhꢁ12
0ꢁkꢁ5
À19ꢁlꢁ19
293(2)
0.0401
0.1031
0.0462
41
9.86
6.499
7.076
2.73–25.97
À7ꢁhꢁ10
À5ꢁkꢁ5
À19ꢁlꢁ19
293(2)
0.0536
0.1512
0.0721
74
2.73–26.20
À9ꢁhꢁ9
À4ꢁkꢁ4
À17ꢁlꢁ18
100(2)
0.0147
0.037
0.0183
68
1.092
2.59–26.01
À10ꢁhꢁ10
À5ꢁkꢁ5
À19ꢁlꢁ16
100(2)
0.034
0.0866
0.0204
59
3.08–26.03
À8ꢁhꢁ8
À7ꢁkꢁ7
À16ꢁlꢁ16
100(2)
0.0136
0.0324
0.0220
56
T [K]
R1
wR2
Rmerge
parameters
GOF
1.108
1.099
1.084
1.09
1.105
reflns total
unique reflns
obsd reflns
CCDC[a]
2733
1079
946
280884
2977
1243
1168
280882
2676
1340
1142
280881
5435
919
896
280880
5638
1089
1012
280892
9332
1007
1000
280893
12345B6C
135B246C
135C246I
135B246I-T
135B246I-M1
135B246I-M2
formula
C6Br5Cl
monoclinic
P21/n
C6Br3Cl3
monoclinic
P21/n
C6Cl3I3
triclinic
P1
C6Br3I3
triclinic
P1
C6Br3I3
monoclinic
P21/n
C6Br3I3
monoclinic
P21/n
crystal system
space group
a [Š]
¯
¯
8.2742(14)
8.1859(9)
7.7131(11)
7.9452(3)
8.5045(16)
9.3129(50)
2224
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2222 – 2234

Halogen···Halogen Interactions
FULL PAPER
Table 1.(Continued)
12345B6C
135B246C
135C246I
135B246I-T
135B246I-M1
135B246I-M2
b [Š]
c [Š]
a [8]
3.9205(7)
15.196(3)
90
92.729(2)
90
3.8619(4)
15.0355(16)
90
92.766(2)
90
9.4269(13)
9.4299(18)
60.213(2)
66.116(3)
85.575(2)
537.43(15)
2
9.4962(4)
9.5119(6)
60.1370(10)
66.202(2)
85.5120(10)
562.61(5)
2
4.0444(8)
15.608(3)
90
92.842(3)
90
4.1977(20)
14.7054(80)
90
93.296(6)
90
b [8]
g [8]
V [Š3]
Z
492.39(15)
2
474.77(9)
2
536.17(18)
2
573.9(5)
2
l [Š]
1calcd [gcmÀ3
F[000]
0.71073
3.42
465
0.71073
2.925
384
0.71073
3.455
492
0.71073
4.088
600
0.71073
4.289
600
0.71073
4.007
600
]
m [mmÀ1
q [8]
]
20.613
2.68–26.04
À10ꢁhꢁ10
À4ꢁkꢁ4
À18ꢁlꢁ16
100(2)
0.0209
0.0504
0.0319
67
13.524
2.78–26.01
À10ꢁhꢁ10
À4ꢁkꢁ4
À18ꢁlꢁ18
100(2)
0.0172
0.0409
0.0273
56
9.413
18.933
19.866
2.61–26.03
À10ꢁhꢁ9
À4ꢁkꢁ4
À19ꢁlꢁ19
100(2)
0.0242
0.0687
0.0256
55
18.56
2.52–26.01
À9ꢁhꢁ9
À11ꢁkꢁ11
À11ꢁlꢁ11
100(2)
0.0284
0.0754
0.0315
110
2.50–26.00
À9ꢁhꢁ9
À11ꢁkꢁ11
À11ꢁlꢁ11
100(2)
0.0213
0.0528
0.0297
110
2.52–26.02
À4ꢁhꢁ11
À5ꢁkꢁ5
À14ꢁlꢁ15
100(2)
0.1455
0.3945
0.1369
25
index ranges
T [K]
R1
wR2
Rmerge
parameters
GOF
1.051
1.091
1.224
1.102
1.107
1.697
reflns total
unique reflns
obsd reflns
CCDC[a]
4060
965
904
280894
4781
932
891
257163
7213
2116
2091
257164
8971
2195
2104
257162
5041
1049
993
257161
1470
1000
912
280887
135F246I
135C246I:135B246I
135B246I:135I246M135C246I
:135B246M135C246I
:135I246M
formula
C6F3I3
0.863(C6Cl3I3) 1.137(C6Br3I3) 0.922(C6Br3I3) 1.078(C9H9I3) 0.154(C6Cl3I3) 1.846(C9H9Br3) 0.154(C6Cl3I3) 1.846(C9H9I3)
crystal system monoclinic
triclinic
P1
triclinic
P1
triclinic
P1
triclinic
P1
¯
¯
¯
¯
space group
a [Š]
b [Š]
c [Š]
a [8]
P21/n
13.937(4)
4.7919(15)
15.488(5)
90
7.8534(8)
9.4683(9)
9.4810(9)
60.1420(10)
66.2530(10)
85.656(2)
552.69(9)
1
7.9146(5)
9.5156(6)
9.5282(6)
60.1830(10)
66.6090(10)
85.9680(10)
564.10(6)
1
7.6549(9)
9.0787(10)
9.1025(10)
60.0510(10)
67.6180(10)
85.0940(10)
502.49(10)
1
7.8342(8)
9.4966(10)
9.5031(10)
60.2820(10)
66.7010(10)
86.2450(10)
556.34(10)
1
b [8]
107.486(3)
90
986.6(5)
4
g [8]
V [Š3]
Z
l [Š]
0.71073
3.432
888
9.493
1.73–26.03
0.71073
3.814
553
14.887
2.51–26.04
0.71073
3.458
516
0.71073
2.461
348
0.71073
3.12
463
1calcd [gcmÀ3
F[000]
]
m [mmÀ1
q [8]
index ranges
]
13.147
11.837
8.666
2.50–26.03
À9ꢁhꢁ9
À11ꢁkꢁ10
À11ꢁlꢁ11
100(2)
0.0173
0.041
0.0152
123
2.61–26.04
À9ꢁhꢁ9
À11ꢁkꢁ11
À11ꢁlꢁ11
100(2)
0.0208
0.0518
0.0242
131
2.50–26.12
À9ꢁhꢁ9
À11ꢁkꢁ11
À11ꢁlꢁ11
100(2)
0.0177
0.0433
0.0167
123
À17ꢁhꢁ15 À9ꢁhꢁ9
À5ꢁkꢁ5 À11ꢁkꢁ11
À19ꢁlꢁ18 À11ꢁlꢁ10
T [K]
R1
wR2
Rmerge
parameters
GOF
reflns total
298(2)
0.0317
0.0799
0.0241
109
100(2)
0.0244
0.0641
0.0283
120
1.049
5283
1.175
7021
1.134
6198
1.028
7763
1.175
7617
unique reflns 1923
2178
2216
1991
2211
obsd reflns
1642
261303
2127
280889
2150
280888
1821
280890
2178
280891
CCDC[a]
[a] Deposition number.
dominant among the compounds we studied and one of
these, as exemplified by the classical hexachlorobenzene
structure, was found to be especially common.[17,18] The unit
cell of C6Cl6 at 100 K has the dimensions a=7.967(3), b=
3.7609(14), c=14.670(5) Š and b=92.459(6)8.The space
group is P21/n, with two molecules in the unit cell.The mol-
ecule lies on an inversion centre in the crystal.For mole-
cules that lack inversion symmetry (12345B6C, 124B356C,
124C356I, 124B356I, 135B246C, 123B456C, 123C456I),
adoption of this structure would lead to crystallographic dis-
order.Accordingly, there are three distinct sites for the hal-
ogen atoms, which we refer to as X1, X2 and X3
Chem. Eur. J. 2006, 12, 2222 – 2234
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
2225

K.A.Padmanabhan, G.R.Desiraju et al.
(Scheme 3).The planar mole-
cules form p···p stacks; within
these stacks the molecules are
3.44 Š apart (perpendicular dis-
tance) and tilted 63.38 to the
stack direction so that p···p in-
teractions are optimised (Fig-
ure 1a).The rest of the struc-
ture is close-packed in the usual
Scheme 3.
Figure 2.Hexachlorobenzene: a), b) and c) show the propagation of the
bend through the crystal on continuous stress application.Arrows show
the point and direction of the stress applied.
way.There are two categories of Cl ···Cl interaction in the
overall distance range of 3.44–3.67 Š. The type-I contacts at
atic study that has attempted to
correlate
these
mechanical
properties with crystal structure
and more particularly with the
nature of the intermolecular in-
teractions contained therein.[21]
Along these lines, we have re-
cently reported
a
structural
model for the bending of organ-
ic crystals.[22] In this paper, we
wish to use this model to corre-
late the bending of C6Cl6 crys-
tals with the nature of their
Cl···Cl interactions (Figure 4).
When a crystal with flat ex-
ternal faces bends, two parallel
opposite faces become curved.
We define these as the bending
Figure 1.Monoclinic form for hexahalogenated benzenes: a) corrugated layer in C 6Cl6; b) Cl···Cl distances [Š]
between stacks in C6Cl6.Cl atoms are shown in dark grey and C atoms in light grey.
sites X1, X2 and X3 (only the last lies on an inversion
centre) have Cl···Cl distances of 3.8165(18) (107.48(13)8),
3.6246(17) (119.52(13)8) and 3.6343(19) Š (124.51(13)8) and
are unremarkable.The second category does not appear to
be type-I and the distances are nominally shorter
(3.4434(15) Š, 175.05(15)8, 116.79(13)8; 3.4697(17) Š,
174.56(13)8,
124.47(13)8;
3.6664(16) Š,
171.25(13)8,
123.18(13)8) but the smaller angle (q2) is not so close to 908,
which would be more characteristic of type-II (Figure 1b).
No Cl···Cl interaction appears to be particularly important
and p···p stacking dominates this packing as it also does in
the isostructural C6Br6 and C6I6.[19,20] Not surprisingly, the
unit-cell volume increases uniformly in these three struc-
tures with the size of the halogen substituent (C6Cl6:
3
3
439.2(3) Š at 100 K; C6Br6: 501.46(17) Š at 100 K; C6I6:
618.642 Š3 at 298 K).
We found that when crystals of C6Cl6 (m.p. 500 K) are
compressed along the needle length, that is, [010], a bending
deformation occurs (Figure 2), which on further compres-
Figure 3.Hexachlorobenzene: a) crystal as mounted for face indexing;
b) SMART face indexing graphic; c) same crystal before bending;
d) crystal bent on (001) face.The occlusion on the crystal surface has no
specific role in its mechanical properties.
sion propagates continuously in
a consecutive motion
through the crystal (Figure 3).Although the typical organic
crystal is not mechanically robust and is susceptible to bend-
ing or breakage on application of stress, and crystallogra-
phers have surely noted crystals with unusually deformed
morphologies over the years, we are unaware of any system-
faces and the crystal bends when it is subject to a mechani-
cal stress perpendicular to this pair of opposite faces.Plastic
bending (no volume change) occurs under compression
2226
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2222 – 2234

Halogen···Halogen Interactions
FULL PAPER
like type-I and type-II might permit some sort of classifica-
tion but they cannot be used to draw definitive chemical in-
ferences about the nature of the respective intermolecular
interactions.
The feeble nature of the Cl···Cl interactions in C6Cl6 is
further indicated by the variation in the shape of the bent
specimens as a function of the mode of the applied force.
Figure 5a results from a compressive force applied at both
Figure 4.A model for bending (half-sectional view): a) undeformed crys-
tal (the spaces between stacks represent the weakest interactions);
b) bent crystal (note the pronounced deformation in some of the interfa-
cial angles).
when the structure is constituted of layers parallel to the
bending faces.[23] In other words, for bending to take place
there must be a strong type of interaction in one direction
and only weak interactions in a perpendicular direction
normal to the bending faces.[22] So, bending occurs in the di-
rection of the weak interaction as shown in Figure 4.This
figure also shows that there can be significant changes in the
interfacial angles upon bending.Whether this involves
breaking and making of intermolecular interactions, or rota-
tion coupled with sliding between molecules, or whether
some other processes are at work is being currently ex-
plored.As strong interactions are often associated with
faster crystal growth, interaction anisotropy is also accompa-
nied by morphology anisotropy.Therefore many crystals
that bend are also needle shaped and C6Cl6 is no exception.
The direction of bending shows that the strong interaction
in C6Cl6 is the p···p stacking.The fact that the crystal bends
on (001) suggests that the Cl···Cl contacts that emerge from
this face are weaker.Are these contacts type-I or type-II?
Ideal type-I and type-II geometries are distinctive but these
Cl···Cl geometries in C6Cl6 (3.4434(15) Š, 175.05(15)8,
116.79(13)8; 3.4697(17) Š, 174.56(13)8, 124.47(13)8) are in-
termediate and difficult to classify.The results on crystal
bending indicate, however, that these Cl···Cl interactions are
weak and nonspecific.They are easily deformed as the
stacks of molecules slide against each other during bending.
Distance and angle criteria are particularly poor indicators
of the nature of Cl···Cl interactions especially when the ge-
ometry is ambiguous and does not correspond strictly to
type-I or type-II.For a type-I interaction that lies on an in-
version centre, a decrease in the distance to below the van
der Waals separation would be repulsive.For a type-II ge-
ometry, however, such a decrease might be moderately at-
Figure 5.Bending of normal crystal into different shapes.a) A crystal of
C6Cl6 bent in four regions by a compressive force applied at both ends,
parallel to the longitudinal axis.b) Schematic depiction of the process de-
scribed in (a).c) Bending into an W shape.The two bends in the flat por-
tion of the crystal occurred without additional anchoring—the forcepsꢁ
tips were enough.d) Bending into an n shape by changing the points of
anchoring.Arrows indicate the point and direction of force applied.
ends, parallel to the longitudinal axis.Figure 5b shows the
situation when the crystal volume remains unchanged as in
plastic bending, if care is taken to prevent buckling.[24a] We
also subjected the crystal to three-point bending to obtain a
bent shape, the schematic of which is shown in Figure 5c.[24]
The cartoon depictions in Figure 5 show that the sliding of
molecules past each other is very easy, and that the internal
stresses between molecules in the stacks are dissipated with
these kinds of deformation.Again, the conclusion is clear
enough: interactions between halogen atoms between stacks
are weaker and/or less specific than the p···p interactions
within a stack.
tractive.By merely inspecting the crystal structure of C Cl6
The bending experiments on C6Cl6 crystals were also car-
ried out by using a nanoindenter.The crystal was placed on
a small iron washer, which had a hole in the middle, and
with the bending face (001) upwards, as shown in Figure 6a.
The middle part of the crystal was unsupported.The inden-
ter tip was used to press the crystal in the middle portion by
6
À
and noting the Cl···Cl distances and C Cl···Cl angles, it
would be very difficult to ascribe any particular chemical
character to these quasi type-II interactions.In contrast, the
bending results show that these interactions are weak and of
low importance in the packing.To summarise, terminologies
Chem. Eur. J. 2006, 12, 2222 – 2234
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
2227

K.A.Padmanabhan, G.R.Desiraju et al.
volume increments on going from 124B356C to 124C356I to
124B356I.Perhaps these structures have ordered domains of
I atom clusters that cannot be resolved in the X-ray experi-
ment.The 1,2,3-trisubstituted compounds also give the same
monoclinic structure but the two disordered orientations in
123B456C and 123C456I are equally occupied as shown in
Figure 7.This type of disorder could be linked to the molec-
ular shape.[25]
Figure 6.Bending of C 6Cl6 crystal by using a nanoindenter tip.a) Sche-
matic depiction of crystal as mounted on sample holder.b) Crystal after
bending.
applying a load of 10 g (a force of 98 mN).The direction of
the indenter movement was [001] with a constant rate of in-
denter tip displacement.The crystal underwent deformation
as shown in Figure 6b.Two crystals that were used in this
experiment had almost the same thickness (0.3 mm) and the
displacement (measured by using a graduated optical micro-
scope) was approximately 1 mm for both.Notably, when a
crystal of C6Cl6 was placed on the washer with the (100)
face pointing upwards, indentation led to immediate break-
age.This reveals the strongly anisotropic nature of the plas-
tic bending present/inherent in these molecular crystals.
The isostructural C6Br6 did not bend and this appeared,
initially, to be counterintuitive.However, by using the well-
known idea that the correct way of comparing the mechani-
cal response of materials is in terms of similar/identical ho-
mologous temperature[24] (equal to T/Tm in which T is the
temperature of deformation and Tm is the melting tempera-
ture), we repeated the bending experiment at 403 K.Bend-
ing took place, as it did for C6Cl6 at room temperature
Figure 7.Disorder in the monoclinic structures 124C356I (left) and
123C456I (right).Notice the differences in the relative positions of halo-
gen atoms and ring C atoms in the two cases.The heavier atoms are
shown darker in colour than the lighter atoms.
The 1,4-disubstituted compounds 14B2356C, 14C2356I,
14B2356I and 1245C36I show a slightly different behaviour.
The first is isostructural to C6Cl6 and (although the molecule
has Ci symmetry) is disordered like the 1,2,4-trisubstituted
compounds.[26] The second and the third take a variant of
the C6Cl6 structure but they are ordered (Figure 8).The
(300 K).The higher melting point of C Br6 relative to C6Cl6
6
is surely of enthalpic origin because the entropic change
upon melting is expected to be very similar for both com-
pounds.Accordingly, we may conclude that the Br ···Br inter-
actions are stronger than the Cl···Cl interactions in these
compounds, and this is not incompatible with their disper-
sive nature.
The 1,2,4-trisubstituted compounds 124B356C, 124B356I
and 124C356I also have the same monoclinic structure.
These molecules lie on the inversion centre and as men-
tioned above, they must be disordered.However, the disor-
der is not statistical over the three positions; the site occu-
pancy factors for Br at X1/X2/X3 are 0.76/0.34/0.40 in
124B356C, for I in 124B356I they are 0.76/0.35/0.39 and for
I in 124C356I they are 0.82/0.64/0.04. All three structures
prefer to have the biggest atom in the X1-position where it
can avoid a type-I contact with itself.There is a big increase
in cell volume going from 124B356C (473 Š3) to 124C356I
(539 Š3) whereas the exchange of Cl by Br in 124B356I
(544 Š3) has only a small effect on the volume.This is
unlike the uniform volume increases in C6Cl6, C6Br6 and
C6I6 and shows that in the 1,2,4-trisubstituted compounds
the biggest substituent determines the cell volume.A fully
statistical disorder would perhaps lead to more uniform cell
Figure 8.Ordered structure of 1,4-dichloro-2,3,5,6-tetraiodobenzene
(14C2356I).Notice the type-I Cl ···Cl interaction.
reader will note that the Cl (or Br) and I positions might
well have been disordered in this packing, but they are not.
The reason for the ordering is suggested by the fact that the
Cl···Cl (or Br···Br) interaction is exclusively type-I whereas
the I···I interactions are both type-I and quasi type-II.So,
even among these marginal X···X contacts, there seems then
to be some preference for the Cl···Cl (or Br···Br) interac-
tions to be more dispersive than the other interactions.
2228
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2222 – 2234

Halogen···Halogen Interactions
FULL PAPER
There is just a hint therefore that Cl (or Br) and I are chem-
ically distinguishable in these two structures.[27] 1245C36I,
however, has two polymorphs.For the first, which occurs as
needles, only a cell could be found that matches C6Cl6
(8.6484(6), 4.1259(3), 15.7424(11) Š; 90, 92.1300(10), 908)
but the structure could not be solved adequately.The
second crystallised from CCl4 as monoclinic rhombs with
cell dimensions of a=6.5276(6), b=5.9682(5), c=
13.3913(11) Š and b=98.7720(10)8.The packing is not do-
minated by p···p stacks; rather there is a herringbone ar-
rangement of molecules to form a zig-zag chain of type-II
I···I interactions (3.8957(3) Š, 170.52(6)8, 79.48(6)8;
Figure 9).In this respect, 1245C36I-M2 is more reminiscent
factors and varying volume increments and perhaps there
are ordered domains, especially when iodine is introduced.
The adoption of an ordered structure of the C6Cl6 type by
14C2356I and 14B2356I is surely suggestive of chemical dif-
ferences between Cl (or Br) and I.[30] In the end, however, it
is the bending experiments that provide the clinching evi-
dence; they show that most of the X···X interactions in
these structures are weak and nonspecific.
The triclinic structure type—Supramolecular synthons: We
now discuss the triclinic structures. 135B246I crystallises
from THF concomitantly in the C6Cl6 monoclinic form
135B246I-M (thin needles) and a triclinic form 135B246I-T,
which was obtained as thick blocks, some of which are boo-
merang shaped.[31] The molecules in the triclinic form are ar-
ranged in planar layers parallel to (100) and in a nearly hex-
agonal arrangement with clusters of three I atoms from
three neighbouring molecules and correspondingly, clusters
of three Br atoms (Figure 10).The I clusters are distinctive
3
with particularly short I···I distances of 3.7548(4) Š
(174.24(12),
119.68(13)8),
3.7762(4) Š
(176.56(12),
119.68(12)8) and 3.7979(5) Š (170.06(12), 119.06(12)8).[32]
The q2 values are all close to 1208 and, despite the short I···I
distances, it would once again not be possible to say whether
these contacts are type-II, which ideally has q1 ꢀ1808 and q2
ꢀ908.The I cluster or synthon is supposedly stabilised co-
3
operatively as I(d+)···I(dÀ) [33]
The Br3 clusters have Br···Br dis-
tances of 4.0660(7), 4.0937(7) and 4.0946(7) Š and are some-
what loosely packed because of the bigger size of the I
.
Figure 9.1,2,4,5-Tetrachloro-3,6-diiodobenzene ( 1245C36I).Notice the
herringbone arrangement of molecules to form a zig-zag chain of type-II
I···I interactions.
of 1,4-diiodobenzene rather
than any of the other hexahalo-
genated benzenes.Unfortunate-
ly, crystals of these four 1,4-dis-
ubstituted compounds were
very small and detailed bending
experiments could not be per-
formed on them.[28] Still, there
is an indication here that the in-
troduction of iodine causes
Figure 10.Planar layer in a) 135C246I, b) 135B246I-T and c) 135I246M.The three I atoms in the I 3 synthon are
in close contact in each case, but the Cl, Br and Me groups are not.This indicates the structural importance of
some structural perturbation.
The question now arises as to the I3 synthons.
why C6Cl6, C6Br6, C6I6 and the
above-mentioned mixed hexa-
halogenated benzenes adopt the same monoclinic structure.
When molecules that are of the same shape and size adopt
the same crystal structure, geometrical factors are presumed
to operate.[29] When molecules of different sizes and shapes
adopt the same crystal packing, chemical factors are suppos-
edly more dominant.However, these generalisations are not
very helpful here.Although Cl (181. Š 3), Br (24.4 Š3) and I
(33.0 Š3) are clearly of different sizes, all the experimental
observations on bending and crystallographic disorder in the
monoclinic group seem to indicate that geometrical factors
are more important.There is, however, a hint of the impor-
tance of chemical factors from the different site-occupation
atoms.Successive planar layers are inversion related and
stacked so that bumps in one layer fit into the hollows of
the next.The interlayer interactions are nonspecific in that
they are based on close packing of spheres in hollows.
An important aspect of chemically directed recognition is
the repeated appearance of specific supramolecular syn-
thons, which are substructural units containing directional
interactions.[34] The planar layer structure in 135B246I-T is
reproduced in other related hexasubstituted benzenes.In
the corresponding chloro derivative 135C246I, the intralayer
I···I distances are 3.7985(6), 3.8250(6) and 3.8646(8) Š
(nearly the same as in 135B246I-T).Because of this, the
Chem. Eur. J. 2006, 12, 2222 – 2234
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
2229

K.A.Padmanabhan, G.R.Desiraju et al.
Cl···Cl distances are pushed apart as far as 4.3360(17),
4.3667(17) and 4.3697(18) Š (longer than the Br···Br distan-
ces in 135B246I-T) and the Cl atoms are not even in con-
tact.In 135I246M,[35,36] the I···I distances in the I3 synthon
are 3.851, 3.897 and 3.933 Š and the Me groups are well
separated, a case of Cl/Me exchange.[8a] The I atom is crucial
in all these cases, and its size determines the overall layer
structure.We suggest that the Br clusters in 135B246I-T, the
Cl clusters in 135C246I and the Me clusters in 135I246M are
mere spectators in a layer structure that is determined by
the robustness of the I3 synthons.In 135F246I a simple geo-
metrical calculation showed that the F···F distances would
need to be at an unrealistic distance of 5.01 Š if the layered
structure of 135B246I-T were adopted with the I3 synthons
conserved.Instead, the experimental structure is three-di-
mensionally corrugated and not layered, with I and F atoms
at van der Waals separation (I···F; 3.53, 3.59 Š).
Many crystals of 135B246I-T and 135C246I were found to
have a definitely curved appearance.Curvature was also ob-
served when an undeformed crystal was held at one end and
disturbed with a needle at the other—a shearing movement
occurs easily.This phenomenon although not unprecedented
is certainly unusual;[37] as mentioned earlier, organic crystals
usually break when subject to mechanical deformation (by
the application of a bending moment or a shear force) and
do not shear in this way.Inspection of Figure 11 shows a
presence of three halogen atoms in a 1,3,5-arrangement.For
example, 135I246M and 135B246M also contain this pattern
and shear just as easily.Compounds that lack this substitu-
tion pattern take the monoclinic disordered structure (which
is not layered) and do not show shearing on application of
mechanical stresses.[39] They either break or bend.There is a
clear relationship between the appearance of a layered
structure and susceptibility to mechanical shearing of layers,
which is well known in plasticity of inorganic crystals.This
relationship is mediated by dislocation movements within
each layer, analogous to basal slip in hexagonal inorganic
crystals like Zn, Cd and Mg; this aspect is under investiga-
tion.This deformation process has other parallels in the in-
organic solid state and resembles the shearing of graphite[40]
(which has a layered structure) and the polytypism in the
layered CdI2.[41] In the context of X···X interactions, the sig-
nificant fact obtained from these experiments is that in
135B246I-T the intralayer I···I interactions are much more
important than the interlayer I···I and I···Br interactions.
These intralayer interactions may be considered to be syn-
thon forming and are chemically significant, arising as they
do from polarisation of the I atoms.
The nonspecificity of the weak interlayer interactions is
further demonstrated by the structure of the boomerang-
shaped (and other shaped) crystals of 135C246I and
135B246I-T (Figure 12).Crystals with these distinctive
Figure 11.Shearing of crystals of 135C246I: a) unsheared specimen elon-
gated along [100]; b) after shearing (notice the striations of slip along
(100)); c) broken specimen after an attempt to cut it in a direction per-
pendicular to the shear plane; d) and e) multiple shearing.
Figure 12.Twinning in 135B246I-T: a) rotation of 608 to achieve boomer-
ang shape; b) twinned crystal used for X-ray measurements; c) and d) in-
dexing of twin faces showing common plane (100) (note that this is the
plane that contains the structure-determining I3 supramolecular syn-
thons).
number of striations that correspond to the shearing move-
ment within a crystal of 135C246I.Notably, the shearing di-
rection corresponds to sliding of the bc planes past one an-
other.In contrast, when we attempted to shear or cut the
crystal in other directions, the crystal fractured.These direc-
tions would correspond to disrupting the layer structure—
the difficulty in doing so attests to the fact that the intralay-
er interactions are strong and directionally specific (synthon
forming).[38] When these crystals were subjected to a three-
point bend test they simply broke along the bc plane.
shapes were obtained for every compound for which shear-
ing was possible.X-ray data were collected for a boomer-
ang-shaped crystal of 135B246I-T.Twinning occurs at the
hinge plane and the common plane between the twins is
(100), which is the shearing layer mentioned above.A plau-
sible twinning model is one in which two consecutive invert-
A necessary (but not sufficient) condition for shearing of
the kind seen in this family of aromatic compounds is the
2230
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2222 – 2234

Halogen···Halogen Interactions
FULL PAPER
ed layers in the normal structure are twisted by 608 and
layers across the twin boundary are packed in a bumps-in-
hollows fashion.In general, the stacking of layers across the
twin boundary is not as satisfactory as for the inverted
layers in the untwinned crystal.In the 2:98 solid solution of
135B246M:135C246I grown from THF, multiple twinning
was observed within the same crystal (Figure 13).In addi-
length increases markedly with temperature compared with
the other parameters.Clearly, the interlayer interactions are
weaker than the intralayer ones.An analogous situation in
inorganic crystals is the anisotropic thermal expansion seen
in metals like Zn and U.
A series of solid solutions was prepared, using equimolar
concentrations in the crystallisation solvent, 1,4-dioxane.All
take the triclinic packing of
135B246I-T, with the heavier
atom forming the X3 synthon.
The
solid
solutions
135B246I:135C246I,
135B246I:135I245M
and
135C246I:135I245M are rather
close to equimolarity in the
solid state (57:43, 46:56 and
41:59, respectively) and have
nearly the same cell volumes
(553, 564 and 556 Š3, respec-
tively).By using the same
method,
135C246I:135B246M with
molar ratio of 8:92 was ob-
a solid solution of
Figure 13.Crystals of 135C246I:135B246M: triple hinged (w-type) crystal (left) and multiple-hinged (“alkyl-
chain” type) crystal (right).
a
tion to the usual single-hinged (V-shape) crystals, we ob-
served double-hinged (N-shape), triple-hinged (W-shape)
and multiple-hinged (“alkyl-chain-type zigzag”) crystals.All
these crystals were grown as such from solution and ob-
tained without mechanical deformation.The fact that twin-
ning is easy[42] suggests that normal and twinned stacking
are energetically comparable.In other words, the interlayer
I···I, Br···Br and I···Br interactions are nonspecific and com-
parable to the easily deformable Cl···Cl interactions in the
bent crystals of C6Cl6.To further confirm the anisotropy of
the 1,3,5-substituted compound, we determined the cell pa-
rameters of 135B246I-T and 135C246I at six different tem-
peratures between 100 and 300 K (Figure 14).The a-axis
tained.The crystallographically determined ratio was also
confirmed by HPLC.Perhaps the lack of I atoms in one of
the constituents prevents a more equimolar crystallisation
product, once again hinting at the importance of I in form-
ing these layered structures.The importance of I in this
regard is also shown by the fact that 135B246C adopts the
disordered monoclinic structure of C6Cl6.This crystal also
shows bending at 413 K, which matches the bending temper-
ature of C6Br6.
Polymorphism: A comment on the occurrence or non-occur-
rence of polymorphs in some of these compounds is in
order.Although 135B246I is trimorphic (triclinic layered,
monoclinic C6Cl6, monoclinic as
in 14B2356I), 135C246I gives
only the triclinic layered struc-
ture and 135B246C only the
C6Cl6 form.Despite repeated
and exhaustive attempts, we
were never able to obtain the
absent polymorphs 135B246C-T
and 135C246I-M.We believe
that the monoclinic C6Cl6 struc-
ture (and the attendant disor-
der) is favoured when the two
halogen atoms are similar in
size as Cl and Br are.The tri-
clinic structure is tolerated even
with large differences in halo-
gen size as in Cl and I.In the
combination Br/I both situa-
tions occur, but the cell volume
Figure 14.Variation of cell parameters with respect to temperature for compounds 135C246I-T (left) and
135B246I-T (right).Notice the a-axis length increases markedly with temperature compared with the other pa-
rameters.
of the triclinic form (563 Š3) is
higher than that of the mono-
Chem. Eur. J. 2006, 12, 2222 – 2234
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
2231

K.A.Padmanabhan, G.R.Desiraju et al.
clinic form (536 Š3) because of the empty spaces in the Br3
clusters in the former.
strument at 400 MHz.IR spectra were recorded on a Jasco 5300 spec-
trometer.All melting points were measured by using a Fisher–Jones
melting point instrument.
Synthesis: All the mixed halogenated compounds were synthesised by
either bromination or iodination of the corresponding halogenated start-
ing materials.The general bromination and iodination procedures used
to prepare these compounds are given for 135B246C and 135C246I.[43]
Conclusion
Halogen···halogen interactions (X···X) have been investigat-
ed in a series of hexahalogenated benzenes.These com-
pounds occur in two broad structural groups.The more
common one is a monoclinic packing that is isostructural
with C6Cl6.The less common one is a layered triclinic pack-
ing that is exclusive, although not mandated, to compounds
in which different halogen atoms occupy the 1,3,5- and
2,4,6-positions.Compounds C 6Cl(6Àn)Brn adopt the disor-
dered monoclinic structure irrespective of their substitution
pattern.The 1,3,5/2,4,6-arrangement of distinct halogen
atoms is compatible with a threefold-symmetrical X3-syn-
thon-based pseudohexagonal layer structure.X ···X interac-
tions have been traditionally classified by using geometrical
criteria as type-I and type-II.Although there might be a
general consensus that the symmetrical type-I interactions
are of the van der Waals type and the unsymmetrical type-II
contacts are polarisation induced, some of the interactions
in these structures do not lend themselves easily to this clas-
sification.Crystals of these hexahalogenated benzenes un-
dergo two types of deformation when subject to a mechani-
Bromination: A mixture of 1,3,5-trichlorobenzene (3 mmol, 556 mg) and
electrolytic Fe powder (19 mmol, 1 g) was placed in a round-bottomed
flask and Br2 (0.23 mol, 6 mL) was added dropwise at 273 K. After this,
the mixture was heated at 408–413 K for 1 h.The resulting mixture was
poured into a saturated NaHSO3 solution (500 mL), the precipitate fil-
tered off and then crystallised from THF to give 135B246C (900 mg).IR
(KBr): n˜ =1317, 1271, 623 cmÀ1; MS (70 eV): m/z (%): 414 [M+], 416
[M++2], 418 [M++4], 420 [M++6], 422 [M++8], 424 [M++10], 335,
258, 107, 77.
Iodination: H5IO6 (1.66 mmol, 380 mg) was dissolved in concentrated
H2SO4 (6 mL) and crushed I2 (5 mmol, 1.27 g) was added followed by
stirring for 0.5 h. Then 1,3,5-trichlorobenzene (1 mmol, 186 mg) was
added to the reaction mixture and followed by stirring for 24 h at room
temperature and then for 36 h at 333 K.The reaction mixture was cooled
and poured onto crushed ice.The solid was filtered and recrystallised
from THF to yield 135C246I (390 mg).IR (KBr): n˜ =1647, 1292, 559,
416 cmÀ1; MS (70 eV): m/z (%): 558 [M+], 560 [M++2], 562 [M++4],
431, 304, 234, 179, 142, 127, 107, 72.
Characterisation: 1H NMR (400 MHz, [D6]DMSO, 258C, TMS): No
peaks were found in any of the spectra of the hexahalogenated com-
pounds.IR (KBr): 135F246I: n˜ =1560, 1404, 1323, 1049, 702, 652 cmÀ1
;
1245C36I: n˜ =1309, 1282, 1246, 679, 582 cmÀ1; 14B2356C: n˜ =1325, 1286,
686, 623 cmÀ1; 135B246I: n˜ =1263, 1224, 1028 cmÀ1; 124C356I: n˜ =1539,
1516, 1288, 1269, 632, 555 cmÀ1; 135C246I:135B246I: n˜ =1292, 1255, 1222,
cal stress.The monoclinic C Cl6 type crystals undergo bend-
1026 cmÀ1
470 cmÀ1
; 135C246I:135B246M: n˜ =2948, 1120, 1020, 949, 644, 559,
6
.
ing but only along certain planes: this can happen only
when interactions in a direction orthogonal to these planes
are particularly weak.Because it is the X ···X interactions
that emerge at the bending faces, we conclude that these in-
teractions are weak and nonspecific despite their geometry,
which is more like type-II than type-I.The triclinic crystals
undergo shear along layers within which the X···X interac-
tions, mostly I···I, are much stronger than X···X interactions
between layers.The strong and specific intralayer I ···I inter-
actions assemble to form I3 supramolecular synthons, which
are the most important structural element in the triclinic
group.Accordingly, we conclude that X ···X interactions
(X=Cl, Br, I) are of several types and that it is sometimes
difficult to characterise them by using geometrical criteria
only.The use of an independent technique, like observation
of the mechanical behaviour of the crystal, offers a much
clearer insight into their nature.The mechanical response of
these molecular crystals resembles in some respects that of
certain types of inorganic crystals.As the latter class of crys-
tals has been investigated extensively, we conclude that
there is scope for understanding the mechanical behaviour
of these two different classes of crystals on a common basis.
Crystallisation: All the compounds were crystallised from either CCl4,
THF or 1,4-dioxane by slow evaporation at ambient temperature.
X-ray crystallography: Intensity data were collected on a Bruker Nonius
Smart Apex CCD with graphite monochromated MoKa radiation.Gaussi-
an face-indexed absorption corrections by Xprep were applied before
empirical data correction by Sadabs 2.10. The structures were solved by
direct methods and refined anisotropically by full-matrix least-squares
methods using the Shelxtl 6.14 software package. Crystal data are given
in Table 1.The supplementary crystallographic data for this paper can be
obtained free of charge from the Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
Nanoindentation experiments: Nanoindentation experiments were per-
formed by using a nanoindenter model XP, supplied by MTS Systems
Corporation, USA.The diamond indenter had a Berkovich (three-sided)
pyramidal geometry.The bending experiments were carried out at 293 K
using the load control mode.The maximum allowed drift was set to
0.1 nmsÀ1 and the indenter velocity was 10 nmsÀ1
.
Acknowledgements
We thank the CSIR (C.M.R.) for an SRF and the AvH Foundation
(M.T.K.) for a Lynen Fellowship. Financial assistance from the DST
(IRPHA) and the UGC (UPE, CMSD) is acknowledged. R.C.G. thanks
G.Sundararajan and R.Sundaresan for their encouragement.
Experimental Section
[1] A.I. Kitaigorodskii, Molecular Crystals andMolecules , Academic
Press, New York, 1973.
Materials: All reagents and solvents employed were commercially avail-
able (Lancaster) and were used as supplied without further purification.
All these compounds were characterised by means of NMR and IR spec-
troscopy.The 1H NMR spectra were recorded on a Bruker Avance in-
[2] J.L.Atwood, L.J.Barbour, A.Jerga,
Science 2002, 296, 2367–2369.
[3] H.A.Bent, Chem. Rev. 1968, 68, 587–648.
[4] G.R. Desiraju, Crystal Engineering: The Design of Organic Solids,
Elsevier, New York, 1989.
2232
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2222 – 2234

Halogen···Halogen Interactions
FULL PAPER
[5] To expand on this point, the final structure is the result of the mini-
misation of the internal energy, the entropy term and the electro-
static, polarisation and van der Waals interactions.
thylbenzene:
135B246M;
1,3,5-triiodo-2,4,6-trimethylbenzene:
135I246M; 1,2,3-tribromo-4,5,6-trichlorobenzene: 123B456C; 1,2,3-
trichloro-4,5,6-triiodobenzene: 123C456I; 1,4-dichloro-2,3,5,6-tet-
raiodobenzene: 14C2356I; 1,4-dibromo-2,3,5,6-tetraiodobenzene:
[6] A.Nangia, G.R.Desiraju,
944.
Acta Crystallogr. Sect. A 1998, 54, 934–
14B2356I;
1,2,4,5-tetrachloro-3,6-diiodobenzene: 1245C36I.
[17] a) G.M. Brown, O.A.W. Strydom, Acta Crystallogr. Sect. B 1974,
30, 801–804; b) J.A.R.P.Sarma, G.R.Desiraju,
1985, 117, 160–164.
[18] C6Cl6 has been reported to be dimorphic but has not been character-
1,4-dibromo-2,3,5,6-tetrachlorobenzene:
14B2356C;
[7] a) W. Jones, C.R. Theocharis, J.M. Thomas, G.R. Desiraju,
J.
Chem. Soc. Chem. Commun. 1983, 1443–1444; b) T.Sakurai, M.
Sundaralingam, G.A. Jeffrey, Acta Crystallogr. 1963, 16, 354–363;
Chem. Phys. Lett.
c) S.C.Nyburg, W.Wong-Ng,
367, 29–45; d) D.E. Williams, L.Y. Hsu, Acta Crystallogr. Sect. A
1985, 41, 296–301; e) GR. .Desiraju, R.Parthasarathy, J. Am.
Proc. R. Soc. London Ser. A 1979,
ised as such: K.Schaum, K.Schäling, F.Klausing,
Justus Liebigs
Chem. Soc. 1989, 111, 8725–8726; f) SL. .Price, AL. .Stone, J.
Lucas, R.S.Rowland, A.E.Thornley, J. Am. Chem. Soc. 1994, 116,
Ann. Chem. 1916, 411, 161–195.
[19] a) E.Baharie, G.S.Pawley, Acta Crystallogr. Sect. A 1979, 35, 233–
4910–4918; g) O.Navon, J.Bernstein, V.Khodorkovsky,
Angew.
235; b) RJ..Steer, SF. .Watkins, P.Woodward,
J. Chem. Soc. C
Chem. 1997, 109, 640; Angew. Chem. Int. Ed. Engl. 1997, 36, 601;
h) V.R.Pedireddi, D.S.Reddy, B.S.Goud, D.C.Craig, A.D.Rae,
G.R.Desiraju, J. Chem. Soc. Perkin Trans. 2 1994, 2353.
1970, 403–408.
[20] For
a general discussion of isostructurality, see: L.Fµbiµn, A.
Kµlmµn, Acta Crystallogr. Sect. B 1999, 55, 1099.
[8] a) P.Metrangolo, G.Resnati,
Chem. Eur. J. 2001, 7, 2511–2519;
[21] a) J.D. Wright, Molecular Crystals, Cambridge University Press,
Cambridge, 1996; b) A.R.West, SolidState Chemistry andIts Appli-
cations, Wiley, Chichester, 1984, pp.347–351.
[22] C.M.Reddy, R.C.Gundakaram, S.Basavoju, M.T.Kirchner, K.A.
Padmanabhan, G.R.Desiraju, Chem. Commun. 2005, 3945–3947.
[23] A.H. Cottrell, Dislocations andPlastic Flow , Claredon, Oxford,
London, 1953, p.29.
b) E.Corradi, S.V.Meille, M.T.Messina, P.Metrangolo, G.Resnati,
Angew. Chem. 2000, 112, 1852; Angew. Chem. Int. Ed. 2000, 39,
1782; c) T.Caronna, R.Liantonio, T.A.Logothetis, P.Metrangolo,
T.Pilati, G.Resnati, J. Am. Chem. Soc. 2004, 126, 4500–4501; d) P.
Metrangolo, H.Neukirch, T.Pilati, G.Resnati,
2005, 38(5), 386–395.
Acc. Chem. Res.
[9] Selected recent references on halogen atom interaction in the crystal
engineering context include: a) F.Zordan, L.Brammer, P.Sher-
wood, J. Am. Chem. Soc. 2005, 127(16), 5979–5989; b) A.C.B.Lu-
[24] a) G.E. Dieter (adapted by D. Bacon), Mechanical Metallurgy, SI
Metric Edition, McGraw-Hill, Singapore, 1988, pp.301, 433. See
p.652 for the different curved shapes that can be produced by
changing the method of support and location of anchor; b) K.A.
Padmanabhan, P.Mondal, H.Hahn, J. Mater. Sci. 2005, published
online: 8 September 2005, DOI: 10.1007/s10853-005-3171-5.
[25] These observations may be explained on the basis of order–disorder
changes as in metallic alloys, which were discussed by Cottrell a
long time ago.See: A.H.Cottrell, Theoretical Structural Metallurgy,
the English Language Book Society and Edward Arnold, London,
1964, pp.191–202.The effects involved are the decrease in internal
energy on ordering, and the increased importance of entropy at high
homologous temperatures.
[26] The site-occupancy factors for Br are 0.73/0.12/0.14 in positions X1/
X2/X3, respectively.We note that position X1 is again favoured for
the biggest substituent.A refinement was published in 1968 but did
not include the disorder.See: T.L.Khotsyanova, T.A.Babushkina,
G.K.Semin, Zh. Strukt. Khim. 1968, 9, 148.
[27] In 14C2356I there is a weak I···Cl (3.792 Š) type-II interaction in
which the more electronegative Cl atom is more acute (1258),
whereas the less electronegative I atom is less acute (1688) which
shows its correct polarisation nature.Again, in 14B2356C the inter-
action is acute more at the Cl atom (1178) than at the Br atom
(1738).However, in 14B2356I the more acute angle is at the less
electronegative I atom (1188) rather than at the Br atom (162.58).
Perhaps, with a larger electronegativity difference between the halo-
gen atoms there is a greater likelihood for a small angle (closer to
908) at the more electronegative halogen, but the Cambridge Struc-
tural Database (CSD) is equivocal on this issue possibly because of
lack of a sufficient number of examples.
cassen, M.Vartanian, G.Leitus, M.E.van der Boom,
Des. 2005, 5, 1671–1673; c) H.L. Nguyen, P.N. Horton, M.B.
Hursthouse, A.C.Legon, D.W.Bruce, J. Am. Chem. Soc. 2004, 126,
16–17; d) R.Glaser, N.Chen, H.Wu, N.Knotts, M.Kaupp, J. Am.
Chem. Soc. 2004, 126, 4412–4419; e) S.Berski, Z.Ciunik, K.Dra-
bent, Z.Latajka, J.Panek, J. Phys. Chem. B 2004, 108, 12327–
12332; f) A.Dey, G.R.Desiraju, CrystEngComm 2004, 6(104), 642–
Cryst. Growth
646; g) A. Dey, R.K.R. Jetti, R. Boese, G.R. Desiraju,
Comm 2003, 5(43), 248–252; h) F.Neve, A.Crispini,
CrystEng-
CrystEng-
Comm 2003, 5(46), 265–268; i) D.Das, RK. R. .Jetti, R.Boese,
G.R.Desiraju, Cryst. Growth Des. 2003, 3, 675–681; j) A.Crihfield,
J.Hartwell, D.Phelps, R.B.Walsh, J.L.Harris, J.F.Payne, W.T.
Pennington, T.W. Hanks, Cryst. Growth Des. 2003, 3, 313–320;
k) R.Liantonio, P.Metrangolo, T.Pilati, G.Resnati, A.Stevenazzi,
Cryst. Growth Des. 2003, 3, 799–803; l) S.V.Lindeman, J.Hecht,
J.K. Kochi, J. Am. Chem. Soc. 2003, 125(38), 11597–11606; m) R.
Thaimattam, D.S.Reddy, F.Xue, T.C.W.Mak, A.Nangia, G.R.
Desiraju, New J. Chem. 1998, 22, 143–148.
[10] J.D.Dunitz, R.Taylor, Chem. Eur. J. 1997, 3, 89–98.
[11] V.R.Thalladi, H-.C.Weiss, D.Bläser, R.Boese, A.Nangia, G.R.
Desiraju, J. Am. Chem. Soc. 1998, 120, 8702–8710.
[12] A.D.Bond, J.Griffiths, J.M.Rawson, J.Hulliger,
Chem. Commun.
2001, 2488–2489.
[13] J.A.R.P. Sarma, G.R. Desiraju,
228.
Acc. Chem. Res. 1986, 19, 222–
[14] a) C.Janiak, J. Chem. Soc. Dalton Trans. 2000, 3885–3896; b) H.I.
Süss, A.Neels, J.Hulliger, CrystEngComm 2005, 7(61), 370–371.
[15] a) A.H. Cottrell, The Mechanical Properties of Matter, Wiley, New
York, 1964; b) A.Kelly, Strong Solids, Claredon Press, Oxford,
1966; c) G.Kaupp, M.R.Naimi-Jamal, CrystEngComm 2005, 7(66),
[28] 14B2356C bends at 383–393 K, 1245C36I-M1 does not bend at
298 K and 1245C36I-M2 becomes waxy at 413–423 K.Specific bend-
ing properties for this last compound could not be distinguished
from waxy flow.
402–410; d) G.Kaupp, J.Schmeyers, U.D.Hangen,
J. Phys. Org.
Chem. 2002, 15, 131–138; e) P.Pavlides, D.Pugh, KJ..Roberts,
Acta Crystallogr. Sect. A 1991, 47, 846; f) G.M.Day, S.L.Price, M.
Leslie, Cryst. Growth Des. 2001, 1, 13; g) V.V. Boldyrev, Reactivity
of Solids: Past, Present and Future, Blackwell, Oxford, 1996, p.272.
[16] IUPAC names and codes of molecules: 1,2,3,4,5-pentabromo-6-
chlorobenzene: 12345B6C; 1,2,4-tribromo-3,5,6-trichlorobenzene:
124B356C; 1,2,4-trichloro-3,5,6-triiodobenzene: 124C356I; 1,2,4-tri-
bromo-3,5,6-triiodobenzene: 124B356I; 1,3,5-trifluoro-2,4,6-triiodo-
[29] A.Dey, RK. R. .Jetti, R.Boese, GR. .Desiraju,
CrystEngComm
2003, 5, 248–252.
[30] These seem to be the result of a complex interplay of free energy
considerations.
[31] A third polymorph was obtained from THF and has the same struc-
ture as 14B2356I.
[32] a) B.K.Saha, R.K.R.Jetti, L.S.Reddy, S.Aitipamula, A.Nangia,
Cryst. Growth Des. 2005, 5, 887–899; b) R.K.R.Jetti, P.K.Thalla-
benzene:
135F246I;
1,3,5-tribromo-2,4,6-trichlorobenzene:
pally, F.Xue, T.C.W.Mak, A.Nangia,
Tetrahedron 2000, 56, 6707–
135B246C; 1,3,5-trichloro-2,4,6-triiodobenzene: 135C246I; 1,3,5-tri-
bromo-2,4,6-triiodobenzene: 135B246I; 1,3,5-tribromo-2,4,6-trime-
6719; c) J.M.A.Robinson, B.M.Kariuki, K.D.M.Harris, D.Philp,
J. Chem. Soc. Perkin Trans. 2 1998, 2459–2469; d) M.B. Zaman,
Chem. Eur. J. 2006, 12, 2222 – 2234
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
2233

K.A.Padmanabhan, G.R.Desiraju et al.
K.A. Udachin, J.A. Ripmeester, Cryst. Growth Des. 2004, 4, 585–
589.
[38] Bending and shearing are related.Bending requires a pair of weak
and strong interactions perpendicular to each other.Shearing re-
quires a weak, nonspecific interaction perpendicular to a layer of
strongly interacting molecules.Bending crystals also do “shear”,
mostly along the needle axis.This corresponds to a shifting of stacks
against each other without actual bending deformation.
[39] Pleasingly, 135F246I, which has the same substitution pattern but
takes a different (ordered) monoclinic structure (Table 1) that is not
layered, also does not shear but bends.
[33] We distinguish here between “synthon” and “cluster”.The synthon
has chemical relevance (short intermolecular distances, favourable
geometries) and repeats in a number of related crystal structures.A
cluster, as understood by us in this paper, is any association of
atoms from different molecules in a crystal with no particular supra-
molecular significance.
[34] a) G.R. Desiraju, Angew. Chem. 1995, 107, 2541–2558; Angew.
Chem. Int. Ed. Engl. 1995, 34, 2311–2327; b) G.R.Desiraju in Stim-
ulating Concepts in Chemistry (Eds.: F. Vçgtle, J. F. Stoddart, M.
Shibasaki), Wiley-VCH, Weinheim, 2002, pp.293–306.
[40] V.A.Phillips, P.Cannon, Nature 1960, 187, 313–314.
[41] A.R.Verma, P.Krishna, Polymorphism andPolytypism in Crystals ,
Wiley, New York, 1966, pp.115–116.
[35] E.Bosch, C.L.Barnes, Cryst. Growth Des. 2002, 2, 299–302.
[36] 135C246M and 135B246M, wherein the Cl3 and Br3 synthons, re-
spectively, are structure determining, are also isostructural.
[37] A very similar demonstration of shearing caused by the sliding of
layers held by nonspecific interactions may be seen in single crystals
of metallic Cu.However, Cu is completely isotropic and the plane
of shearing is immaterial. See: R.P. Feynman, R.B. Leighton, M.
Sands, The Feynman Lectures on Physics, Vol. 2, Addison-Wesley,
USA, 1964, pp.30-8.
[42] Twinning boundaries always have a local minimum in free energy.
See: G.A.Chadwick, D.A.Smith,
Grain Boundary Structure and
Properties, Academic Press, London, 1976.
[43] D.L.Mattern, J. Org. Chem. 1984, 49, 3051–3053.
Received: August 13, 2005
Published online: December 29, 2005
2234
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2006, 12, 2222 – 2234