Exploitation of the menshutkin reaction for the controlled assembly of halogen bonded architectures incorporating 1,2-diiodotetrafluorobenzene and 1,3,5-Triiodotrifluorobenzene

Article abstract of DOI:10.1021/cg201017r

1,4-Diazabicyclo[2.2.2]octane (DABCO) forms well-defined co-crystals with 1,2-diiodotetrafluorobenzene (1,2-DITFB), [(1,2-DITFB)₂DABCO], and 1,3,5-triiodotrifluorobenzene, [(1,3,5-TITFB)₂DABCO]. Both systems exhibited lowerthan-expected supramolecular connectivity, which inspired a search for polymorphs in alternative crystallization solvents. In dichloromethane solution, the Menshutkin reaction was found to occur, generating chloride anions and quaternary ammonium cations through the reaction between the solvent and DABCO. The controlled in situ production of chloride ions facilitated the crystallization of new halogen bonded networks, DABCO-CH ₂Cl[(1,2-DITFB)Cl] (zigzag X-bonded chains) and (DABCO-CH ₂Cl)₃[(1,3,5-TITFB)₂Cl₃] ·CHCl3 (2D pseudo-trigonal X-bonded nets displaying Borremean entanglement), propagating with charge-Assisted C-I...Cl-halogen bonds. The method was found to be versatile, and substitution of DABCO with triethylamine (TEA) gave (TEA-CH₂Cl)₃[(1,2-DITFB)Cl₃] ·4(H₂O) (mixed halogen bond hydrogen bond network with 2D supramolecular connectivity) and TEA-CH₂Cl[(1,3,5-TITFB)Cl] (tightly packed planar trigonal nets). The co-crystals were typically produced in high yield and purity with relatively predictable supramolecular topology, particularly with respect to the connectivity of the iodobenzene molecules. The potential to use this synthetic methodology for crystal engineering of halogen bonded architectures is demonstrated and discussed.

Full text of DOI:10.1021/cg201017r

Article
pubs.acs.org/crystal
Exploitation of the Menshutkin Reaction for the Controlled Assembly
of Halogen Bonded Architectures Incorporating 1,2-
Diiodotetrafluorobenzene and 1,3,5-Triiodotrifluorobenzene
Published as part of the Crystal Growth & Design virtual special issue on Halogen Bonding in Crystal
Engineering: Fundamentals and Applications
†
‡
†
†
‡
Michael C. Pfrunder, Aaron S. Micallef, Llewellyn Rintoul, Dennis P. Arnold, Karl J. P. Davy,
,
†
and John McMurtrie*
†
Chemistry, Queensland University of Technology, Brisbane, Queensland, 4001, Australia
‡
Australian Institute for Bioengineering and Nanotechnology and Centre for Advanced Imaging, University of Queensland, St. Lucia,
Queensland, 4072, Australia
*
S Supporting Information
ABSTRACT: 1,4-Diazabicyclo[2.2.2]octane (DABCO) forms
well-defined co-crystals with 1,2-diiodotetrafluorobenzene (1,2-
DITFB), [(1,2-DITFB) DABCO], and 1,3,5-triiodotrifluoroben-
2
zene, [(1,3,5-TITFB) DABCO]. Both systems exhibited lower-
2
than-expected supramolecular connectivity, which inspired a search
for polymorphs in alternative crystallization solvents. In dichloro-
methane solution, the Menshutkin reaction was found to occur,
generating chloride anions and quaternary ammonium cations
through the reaction between the solvent and DABCO. The
controlled in situ production of chloride ions facilitated the
crystallization of new halogen bonded networks, DABCO−
CH Cl[(1,2-DITFB)Cl] (zigzag X-bonded chains) and
2
(
DABCO−CH Cl) [(1,3,5-TITFB) Cl ]·CHCl (2D pseudo-trigonal X-bonded nets displaying Borremean entanglement),
2
3
2
3
3
−
propagating with charge-assisted C−I···Cl halogen bonds. The method was found to be versatile, and substitution of DABCO
with triethylamine (TEA) gave (TEA-CH Cl) [(1,2-DITFB)Cl ]·4(H O) (mixed halogen bond hydrogen bond network with
2
3
3
2
2
D supramolecular connectivity) and TEA-CH Cl[(1,3,5-TITFB)Cl] (tightly packed planar trigonal nets). The co-crystals were
2
typically produced in high yield and purity with relatively predictable supramolecular topology, particularly with respect to the
connectivity of the iodobenzene molecules. The potential to use this synthetic methodology for crystal engineering of halogen
bonded architectures is demonstrated and discussed.
INTRODUCTION
bond interactions which have been successfully used for the
construction of supramolecular assemblies, anionic acceptors
■
In the continuing maturation of the field of crystal engineering,
halogen bond (or X-bond) interactions are developing into a
common tool for the design and construction of supra-
(
such as halides and thiocyanate) are known to be even more
5
effective in this context due to charge assistance. The strength
and directionality of halogen bonds can be further enhanced by
the incorporation of strongly electron withdrawing groups
attached to Y, and for this reason iodofluoroalkanes and
iodofluorobenzenes have been widely used as X-bond donors
1
molecular assemblies. Halogen bonds are of the form B···X−Y
and derive predominantly from electrostatic interaction
between the lone pair of electrons of a Lewis base (B, the
electron donor is called the halogen bond acceptor) and the
electrophilic site of a halogen atom (X = I, Br, Cl (but not F),
6
for the construction of halogen bonded assemblies. The
2
growing interest in halogen bonds for supramolecular chemistry
is due largely to their strength (similar to, and in some cases
greater, than hydrogen bonds), their strong directionality, and
their chemical compatibility with a range of different functional
groups.
called the X-bond donor). The electrophilic site on the
halogen resides in the region trans to the X−Y bond, and as a
result the contact is typically linear (or very close to it) across
3
B···X−Y. Halogen bond interactions have energies between 5
kJ/mol for weakly basic acceptors (e.g., ether-O atoms) and less
polarizable halogen donors (Cl) to 180 kJ/mol in the extreme
case of triiodide ion which incorporates a strong Lewis base
Received: August 5, 2011
Revised: October 26, 2011
Published: December 20, 2011
−
4
acceptor (I ) with a highly polarizable donor (I in I ). While
2
neutral amine and pyridyl bases produce moderately strong X-
©
2011 American Chemical Society
714
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
We have been exploring the potential of halogen bonds for
the construction of functional materials containing organic spin
7
systems, in particular, nitroxides. We recently reported the
construction of a cyclic tetramer comprising two molecules of
1,2-diiodotetrafluorobenzene (1,2-DITFB) and two molecules
of the isoindoline nitroxide 1,1,3,3-tetramethylisoindolin-2-
8
yloxyl (TMIO). In this tetramer, the 1,2-DITFB molecules
act as ditopic linkers to each of the two nitroxide-O atoms,
which form two halogen bonds each in a bifurcated
arrangement. Our investigations raised a concern about the
current dearth of crystallographic information on the potential
for 1,2-DITFB as well as for the potentially tritopic halogen
bond acceptor 1,3,5-triiodotrifluorobenzene to produce supra-
molecular architectures with predictable topological connectiv-
ity, particularly when compared to the more widely employed
para-substituted halogen bond donor 1,4-diiodotetrafluoroben-
zene (1,4-DITFB).
The molecules 1,4-DITFB, 1,2-DITFB, and 1,3,5-TITFB are
illustrated in Figure 1a along with their potential geometric
Figure 2. Predicted supramolecular topologies for combinations of
molecular nodes and linkers. (a) Linear X-bond donor nodes with
linear acceptor linkers result in 1D chains with linear topology. (b)
Ditopic 60° X-bond donor nodes (such as 1,2-DITFB) with linear
acceptor linkers can give rise to 0D supramolecular triangles or (c)
zigzag chains that propagate with 60° vertices. (d) Trigonal X-bond
donor nodes with linear acceptor linkers are predicted to form 2D
trigonal or pseudotrigonal networks.
In contrast, the predictability of supramolecular topologies
for 1,2-DITFB and 1,3,5-TITFB is not well established. At the
time of writing there are only six unique crystal structures in the
CSD incorporating 1,2-DITFB in halogen bonded co-crystals.
In crystals of [(1,2-DITFB)(4,4′-bipy)] the molecules are
arranged, as might be expected, in halogen bonded chains that
propagate with zigzag topology due to the disposition of the I
12
atoms of 1,2-DITFB. In other structures, with relatively
strongly X-bonding N-acceptor molecules such as thiomorpho-
Figure 1. (a) Structures of the halogen bond donors 1,4- and 1,2-
diiodotetrafluorobenzene (DITFB) and 1,3,5-triiodotrifluorobenzene
13
14
(
TITFB). (b) Predicted geometric topologies of the respective
line and thiocyanate, the 1,2-DITFB is also ditopic. In the
two structures with weak X-bonding S-acceptor molecules, the
iodofluorobenzenes when incorporated as molecular nodes in halogen
bonded structures. (c) Organic halogen bond acceptors 1,4-
diazabicyclo[2.2.2]octane (DABCO) and 4,4′-bipyridine (4,4′-bipy)
and a schematic representation of their linear ditopic topology when
incorporated as supramolecular linkers in halogen bonded structures.
15
1,2-DITFB is monotopic even though there are no obvious
steric inhibitors to the formation of two X-bonds. There are no
crystal structures incorporating 1,2-DITFB with DABCO, a fact
that prompted the first component of this research. Given that
1
,4-DITFB is ditopic and forms linear chains with DABCO as
do similarly basic X-bond acceptors, our hypothesis was that
,2-DITFB would also act as a ditopic linker between DABCO
topologies in halogen bonded networks (linear for 1,4-DITFB,
bent with a 60° angle of propagation for 1,2-DITFB and the
tritopic trigonal arrangement for 1,3,5-DITFB) as shown in
Figure 1b. At the time of writing, there are over 90 structures in
1
molecules with the result being formation of either discrete
supramolecular triangles (Figure 2b) or 1D zigzag chains with
6
9
the Cambridge Structural Database (CSD) in which 1,4-
0° angular propagation (Figure 2c).
The situation is even less clear for 1,3,5-TITFB. It might be
DITFB is an X-bonding component, and in the vast majority of
these it acts as a ditopic linear linker through the formation of
two X-bond interactions involving the I atom acceptors. 4,4′-
Bipyridine (4,4′-bipy) and 1,4-diazabicyclo[2.2.2]octane
expected that when 1,3,5-TITFB forms three X-bonds to linear
ditopic linkers that the result would be formation of a two-
dimensional (2D) honeycomb-like network as illustrated in
Figure 2d. While there are more structures in the CSD (21
unique co-crystals), most of them (16) incorporate anionic
acceptors (halide or thiocyanate). The 1,3,5-TITFB is tritopic,
forming halogen bonds at all three I atoms with just one
(
DABCO) are potentially ditopic linear X-bond linkers due
to the diametrically opposed directionality of the Lewis basic
lone pairs of the N atoms (as illustrated in Figure 1c). The
geometric disposition of the halogen bonding components
leads to predictable supramolecular arrangements as exempli-
fied in crystals of the supramolecular complex [(1,4-DITFB)-
16
exception in which two X-bonds form. It is reasonably clear
then that with strong anionic X-bond acceptors the preferred
topological connectivity for 1,3,5-TITFB is tritopic, with the
result being formation of 2D trigonal nets. Of the remaining
five structures, which involve weaker neutral X-bond acceptors
with 1,3,5-TITFB donors, less can be deduced. In co-crystals
comprising 1,3,5-TITFB and 4,4′-bipy, and similar ditopic
(
4,4′-bipy)] in which the molecules are arranged in one-
10
dimensional (1D) chains with linear propagation (Figure 2a).
In crystals of [(1,4-DITFB)(DABCO)], the molecules are also
11
(predictably) arranged in chains with linear topology.
7
15
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
pyridyl X-bond acceptors, the 1,3,5-TITFB consistently forms
29.30, 29.89, 30.36, 31.43, 31.89, 32.72, 34.00, 34.56, 35.81, 37.03,
17
3
7.68, 38.39, 39.23.
two rather than the maximum possible three X-bonds. As an
interaction with a charge transfer component, it is possible that
each subsequent X-bond that forms, lowers the capacity of the
remaining I atoms to also form X-bonds, and there is
computational evidence provided by the group of Bruce that
DABCO−CH Cl[(1,2-DITFB)Cl] (3). Crystals were formed in a
2
sealed vessel containing a solution of 1,2-DITFB (10.0 mg, 24.9 μmol)
and DABCO (2.8 mg, 25.0 μmol) in dichloromethane (0.10 cm )
3
approximately 1 h after mixing. Yield 11.9 mg, 85%; mp 120 °C (dec.);
Found C 26.24, H 2.51, N 4.80. Calc. for 1:1 complex
C H Cl F I N : C 26.07, H 2.36, N 4.68%. Raman (solid): ν
1
8
this is the case. Notably, this recent study reports the first,
and to date only, example of a co-crystal of 1,3,5-TITFB in
which all three I atoms form X-bonds to a neutral X-bond
acceptor (in this case dimethylaminopyridine, DMAP). This
result indicates that if the crystal packing is favorable, 1,3,5-
TITFB is a potential tritopic X-bond linker with neutral X-bond
acceptors. As with 1,2-DITFB, there were no published
structures of 1,3,5-TITFB with DABCO, prompting us to
investigate the topological connectivity for this pair of X-
bonding molecules.
1
3
14
2
4 2
2
̅
−
1
(cm ) = 2984, 2954, 2885, 1617, 1462, 1289, 1256, 1008, 806, 761,
683, 470, 357, 225 (C−I), 149; Powder XRD (Cu Kα 1.540598 Å):
1
°2θ = 8.092, 8.520, 10.27, 12.09, 13.13, 13.74, 15.22, 15.60, 16.96,
1
2
3
7.32, 18.01, 18.67, 21.11, 21.58, 22.01, 22.79, 24.46, 25.37, 25.59,
6.17, 27.39, 27.77, 28.27, 29.77, 30.48, 31.25, 31.71, 31.94, 33.10,
1
3.88, 34.26, 35.92, 36.52, 37.38, 39.39; H NMR (400 MHz, D -
6
1
3
DMSO, D O): δ = 5.377 (s, 2H), 3.40 (t, 6H), 3.08 (t, 6H); C NMR
2
(
6
101 MHz, D -DMSO): δ = 147.05 (dm), 139.07 (dm), 93.75 (m),
7.49, 50.66, 44.39.
6
(
DABCO−CH Cl) [(1,3,5-TITFB) Cl ]·CHCl (4). Crystals formed
2 3 2 3 3
To that end, we have investigated the supramolecular X-
bonding preferences for 1,2-DITFB and 1,3,5-TITFB with
DABCO and the results are reported herein. In the process of
our investigation, we discovered a new and highly efficient
method for the production of X-bonded networks in high
in a sealed vessel containing a solution of 1,3,5-TITFB (13.0 mg, 25.5
μmol) and DABCO (4.3 mg, 38.3 μmol) in a 50:50 dichloromethane/
3
chloroform mixture (0.70 cm ) over a period of approximately 18 h.
Yield 16.0 mg, 84%; mp 122 °C (dec.); Found C 23.25, H 2.87, N
5
.26. Calc. for 2:3 complex C H Cl F I N ·CHCl : C 23.60, H 2.50,
33 42 6 6 6 6 3
1
−
19
N 4.86%. Raman (solid): ν (cm ) = 3000, 2954, 2890, 1562, 1460,
1
̅
purity through application of the Menshutkin reaction, which
occurs between amines and alkyl halides (in this case
dichloromethane). The results and potential for expansion of
this methodology are also presented and discussed.
1
395, 1077, 792, 685, 647, 574, 374, 223, 167 (C−I); H NMR (400
MHz, DMSO-d , D O): δ = 5.24 (s, 2H), 3.35 (t, 6H), 3.07 (t, 6H);
6
2
13
C NMR (101 MHz, DMSO-d ) δ = 161.78 (dt), 68.00 (td), 50.65,
6
4
4.42.
(
TEA-CH Cl) [(1,2-DITFB)Cl ]·4(H O) (5). Crystals formed in a
2 3 3 2
EXPERIMENTAL SECTION
,2-Diiodotetrafluorobenzene (1,2-DITFB) (99%), 1,4-diiodotetra-
fluorobenzene (1,4-DITFB) (98%), and 1,4-diazabicyclo[2.2.2]octane
DABCO) (99%) were purchased from Sigma-Aldrich. Triethylamine
sealed vessel containing a solution of 1,2-DITFB (10.0 mg, 24.9 μmol)
and triethylamine (3.8 mg, 37.3 μmol) in dichloromethane (0.1 cm )
over a period of 5 days. Yield 2.2 mg, 14%; mp 170 °C (dec.); Found
■
3
1
(
(
C 27.62, H 3.79, N 2.99. Calc. for 2:3 complex C33
(H O): C 27.64, H 4.15, N 2.93%. Raman (solid): ν
H
51Cl
6
F
8
I
4
N
3
·4
20
−1
99%) was purchased from MERCK. 1,3,5-TITFB was prepared by a
2
̅
(cm ) = 2991,
previously reported method. Elemental microanalyses were deter-
mined on a Carlo Erba NA1500 Elemental microanalyser. Raman
spectra were obtained on a Renishaw System 1000 Raman microscope,
with a laser of wavelength 632.8 nm giving a maximum power of
approximately 5mW radiation at the sample. Spectra were recorded
2950, 1613, 1458, 1288, 1251, 1158, 1100, 1008, 810, 760, 690, 469,
415, 357, 339, 297, 226 (C−I), 207, 152; Powder XRD (Cu Kα
1
1.540598 Å): °2θ = 6.417, 6.714, 9.014, 9.539, 13.37, 14.50, 15.23,
15.77, 16.26, 16.51, 17.38, 18.04, 18.45, 19.20, 20.51, 21.38, 22.69,
22.90, 23.25, 23.50, 24.70, 25.13, 25.86, 25.94, 26.74, 27.13, 28.17,
−1
with a resolution better than 4 cm . The spectrometer was calibrated
28.62, 29.27, 29.99, 30.69, 31.14, 32.22, 32.93, 33.96, 34.30, 35.25,
−
1
1
by reference to the silicon wafer band of 520.5 cm . NMR spectra
were recorded on a Bruker Avance 400 MHz spectrometer, fitted with
a 5 mm multinuclear probe. 13C Spectra were collected using a
36.11, 37.77, 38.46, 38.98; H NMR (400 MHz, DMSO-d
6
, D
2
O): δ =
5.23 (s, 2H), 3.34 (q, 6H), 1.21 (t, 9H); 13C NMR (101 MHz,
DMSO-d
) δ = 147.06 (dm), 138.97 (dm), 94.02 (m), 62.62, 52.03,
TEA-CH Cl[(1,3,5-TITFB)Cl] (6). Crystals formed in a sealed vessel
2
6
standard proton decoupled carbon pulse sequence. D O was added to
7.23.
2
1
samples for H spectrum collection in order to shift the residual water
signal downfield, eliminating its overlap with sample proton signals.
containing a solution of 1,3,5-TITFB (10.0 mg, 19.6 μmol) and
triethylamine (19.9 mg, 195.3 μmol) in dichloromethane (0.1 cm )
over a period of 3 days. Yield 5.8 mg, 45%; mp 184 °C (dec.); Found
C 22.35, H 2.21, N 2.13. Calc. for 1:1 complex C13H17Cl F I N: C
22.44, H 2.46, N 2.01%. Raman (solid): ν (cm ) = 3116, 3053, 2994,
2947, 2887, 2148, 1717, 1560, 1455, 1393, 1073, 1042, 677, 645, 579,
375, 223, 164 (C−I); Powder XRD (Cu Kα 1.540598 Å): °2θ =
8.645, 12.21, 13.54, 14.72, 17.45, 18.45, 20.33, 20.78, 21.87, 22.58,
3
[
(1,2-DITFB) DABCO] (1). Crystals of the X-bonded complex
2
[
(1,2-DITFB) DABCO] were obtained by slow evaporation of solvent
2
from a solution of 1,2-DITFB (10.0 mg, 24.9 μmol) and DABCO (1.4
2 3 3
3
3
−1
mg, 12.5 μmol) in ethanol (0.40 cm ) and water (0.10 cm ). Yield 8.6
mg, 75%; mp 127.5−129 °C; Found C 23.57, H 1.31, N 3.13. Calc. for
̅
2
:1 complex C H F I N : C 23.60, H 1.32, N 3.06%. Raman (solid):
1
1
8
12 8 4
2
−1
ν
1
(cm ) = 2945, 2876, 1485, 1469, 1443, 1344, 1304, 1287, 1253,
̅
max
1
172, 1103, 974, 812, 801, 763, 619, 469, 356, 228, 217 (C−I), 148;
25.39, 26.25, 26.68, 27.29, 28.34, 30.39, 32.64, 33.43, 35.27, 37.16; H
Powder XRD (Cu Kα 1.540598 Å): °2θ = 8.09, 8.52, 10.27, 12.09,
NMR (400 MHz, DMSO-d , D O): δ = 5.21 (s, 2H), 3.33 (q, 6H),
6 2
1
1.21 (t, 9H); 13C NMR (101 MHz, DMSO-d
) δ = 161.76 (dt), 67.91
6
1
2
3
3
3.13, 13.74, 15.22, 15.60, 16.96, 17.32, 18.01, 18.67, 21.11, 21.58,
2.01, 22.79, 24.46, 25.37, 25.59, 26.17, 27.39, 27.77, 28.27, 29.77,
0.48, 31.25, 31.71, 31.94, 33.10, 33.88, 34.26, 35.92, 36.52, 37.38,
9.39.
(td), 62.61, 52.03, 7.23.
Single crystal X-ray diffraction data were collected for 1−6 at
2
1
173(2) K under the software control of CrysAlis CCD on an Oxford
Diffraction Gemini Ultra diffractometer using Mo−Kα radiation
generated from a sealed tube. Data reduction was performed using
[
(1,3,5-TITFB) DABCO] (2). Crystals were obtained by slow
2
evaporation of solvent from a solution of 1,3,5-TITFB (13.0 mg,
5.5 μmol) and DABCO (1.4 mg, 12.5 μmol) in a 2:7:7
2
1
2
CrysAlis RED. Multiscan empirical absorption corrections were
applied using spherical harmonics, implemented in the SCALE3
ABSPACK scaling algorithm, within CrysAlis RED and subsequent
3
ethanol:acetone:water mixture (0.80 cm ). Yield 12.4 mg, 86%; mp
74−176 °C; Found C 19.38, H 0.99, N 2.43. Calc. for 2:1 complex
2
1
1
−1
C H F I N : C 19.10, H 1.07, N 2.48%. Raman (solid): ν (cm ) =
computations were carried out using the WinGX-32 graphical user
18
12 6 6
2
̅
2
2
3
1
(
1
2
122, 2955, 2942, 2875, 2539, 2150, 1563, 1468, 1453, 1399, 1361,
346, 1075, 1046, 973, 802, 654, 644, 619, 575, 374, 228, 218, 167
interface. The structures were solved by direct methods using
23 24
SIR97 and refined with SHELXL-97. Full occupancy non-
hydrogen atoms were refined with anisotropic thermal parameters.
C−H hydrogen atoms were included in idealized positions and a
riding model was used for their refinement. Crystal data and
C−I); powder XRD (Cu Kα 1.540598 Å): °2θ = 10.12, 11.17, 13.01,
1
3.57, 17.17, 17.63, 18.23, 18.93, 19.18, 19.80, 20.20, 20.86, 21.46,
1.73, 23.08, 23.51, 24.49, 25.42, 26.35, 27.36, 27.91, 28.52, 29.10,
7
16
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
Table 1. Crystal and Refinement Data for 1−6
(
1,2-DITFB)
(
1,2-DITFB)2
(1,3,5-TITFB)2
(DABCO) (2)
[DABCO−CH Cl]Cl
(1,3,5-TITFB) ([DABCO−
(1,2-DITFB) ([TEA-
1,3,5-TITFB ([TEA-
2
2
2
(
DABCO) (1)
(3)
CH Cl]Cl) ·CHCl (4)
CH Cl]Cl) ·4H O (5)
CH Cl]Cl) (6)
2
3
3
2
3
2
2
formula
C H F I N
C H F I N
18 12 6 6
C H Cl F I N
C H Cl F I N
C H Cl F I N O
C H Cl F I N
1
8
12 8 4
2
2
13 14
2
4 2
2
34 43
9
6 6
6
33 59
6
8 4
3
4
13 17
2 3 3
M
915.90
1131.70
598.96
1730.19
1434.13
695.88
crystal
system
triclinic
monoclinic
triclinic
monoclinic
monoclinic
trigonal
space group P1
̅
C2/c
P1
̅
P2 /c
1
C2/c
̅
R3c
a/Å
b/Å
c/Å
α/°
β/°
γ/°
7.5384(4)
12.7514(6)
13.4479(5)
92.749(4)
104.919(4)
102.130(4)
1214.07(9)
2.505
15.7938(5)
9.6819(2)
18.2902(5)
12.3666(4)
12.5432(4)
14.0845(4)
83.083(2)
72.525(3)
60.817(3)
20.2337(9)
12.6399(4)
22.2082(8)
27.3744(9)
7.0166(2)
26.3653(9)
20.2571(3)
25.5490(6)
107.577(3)
110.871(5)
93.360(3)
V/Å3
2666.26(13)
2.819
1818.52(12)
2.188
5307.1(3)
2.165
5055.4(3)
1.884
9079.4(3)
2.291
−
3
D /g cm
c
Z
2
4
4
4
4
18
color
habit
colorless
prism
colorless
prism
colorless
plate
colorless
multifaced
colorless
plate
colorless
prism
dimensions
0.30 × 0.14 ×
.12
0.17 × 0.07 ×
0.06
0.34 × 0.17 × 0.05
0.48 × 0.34 × 0.24
0.44 × 0.18 × 0.09
0.24 × 0.12 × 0.10
0
μ (Mo Kα)
5.203
7.042
3.788
4.019
2.849
4.933
−
1
/
mm
Tmin,max
0.686, 1
8181 (0.032)
6467
0.647, 1
3096 (0.034)
2486
0.657, 1
8424 (0.029)
6538
0.723, 1
0.714, 1
5891 (0.027)
4752
0.736, 1
2467 (0.031)
1685
Nind (Rint)
12239 (0.027)
9439
Nobs − (I >
2σ(I))
N
289
0.0289
146
415
550
297
106
var
a
R1
0.0228
0.0249
0.0343
0.0753
0.03, 6
1.080
0.0293
0.0726
0.04, 0
1.031
0.0305
a
wR2
0.0548
0.0343
0.0415
0.0713
A, B
GOF
Δρ
0.008, 2.3
1.071
/e− −1.174, 1.015
0.005, 1
1.013
0.01, 0.5
1.024
0.035, 0
1.054
−0.541, 0.605
−0.689, 0.678
−1.351, 1.568
−0.876, 0.794
−0.701, 1.432
min,max
−
3
Å
a
2
2
c
2
2
c
2
1/2
Reflections with [I > 2σ(I)] considered observed. R = Σ||F | − |F ||/Σ|F | for F > 2σ(F ) and wR
= {Σ[w(Fo − F ) ]/Σ[w(F ) ]} where w
1
o
c
o
o
o
2(all)
2
2
2
2
2
=
1/[σ (F ) + (AP) + BP], P = (F + 2F )/3.
o o c
1
refinement details for 1−6 are supplied in Table 1. The CIF files have
been deposited with the Cambridge Structural Database (CCDC
reference numbers 837965−837970) and can be obtained free of
charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from the
Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge
CB2 1EZ, UK; fax: (+44) 1223−336−033; or deposit@ccdc.cam.ac.
uk).
the H NMR spectrum and absence of the expected CHCl carbon
3
atom peak in the 13C NMR spectrum. In this process the crystals also
lose crystallinity. Consequently, a powder pattern was not recorded for
this sample. However, in an attempt to confirm structural purity of the
sample, single crystal X-ray data sufficient for unit cell determination
were collected for 20 individual crystals. In each case, the unit cell
dimensions and Bravais lattice type of 4 were confirmed indicating a
high degree of structural purity for this batch of co-crystals. Elemental
microanalysis results for 4 match within accepted values for the
structure solvated with chloroform. The sample analyzed was removed
from the mother liquor and placed directly in a small sealed vial prior
to analysis, which appeared to slow the loss of solvent thus providing
reasonable elemental analysis results.
Powder X-ray diffraction (XRD) data were collected for 1, 2, 3, 5,
and 6 on an XPERT-PRO diffractometer with graphite-monochro-
mated Cu radiation, Kα = 1.540598 Å. The powder XRD patterns
1
were compared to patterns simulated from the single crystal X-ray
structures in order to confirm structural purity of the samples. The
powder patterns for the bulk samples were in excellent agreement with
simulated patterns indicating structural purity, except in the case of 5.
The powder pattern for 5 showed the presence of some peaks that
could not be attributed to reflections simulated from the crystal
structure of 5. The crystals were of similar size and habit. Seven large
crystals (which made up a significant proportion of the sample) were
selected and their Raman spectra recorded. In each case, the C−I
stretch, which is diagnostic of X-bonding, was identical. The powder
XRD spectra of these seven crystals closely match the pattern
simulated from 5. The elemental microanalysis for sample 5 was
determined for the entire batch prior to manual separation of crystals.
The close match indicates that the structural impurity has the same or
similar stoichiometry as 5, but to date we have been unable to identify
crystals of the impurity. As such, the composition and structure of this
additional phase (or phases) remain an enigma. The crystals of 4 lose
chloroform from the structure on removal from the mother liquor as
RESULTS AND DISCUSSION
■
As 1,2-DITFB and DABCO are both potentially ditopic linkers,
it was anticipated that they would co-crystallize in a 1:1 ratio to
form a discrete supramolecular 3 + 3 complex with triangular
topology or, alternatively, a 1D supramolecular polymer with
zigzag topology propagating with 60° vertices. Accordingly,
initial co-crystallization experiments were carried out using a
:1 ratio of the reagents. However, these experiments yielded
crystals of a 2:1 DITFB/DABCO complex. Co-crystallization of
,2-DITFB with DABCO from a 2:1 solution of the reagents
dissolved in an ethanol/water mixture gave the discrete 2:1
1
1
supramolecular complex [(1,2-DITFB)
chemical and structural purity.
DABCO] (1) in high
2
indicated by the absence of a significant peak for the CHCl proton in
3
7
17
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
In crystals of [(1,2-DITFB) DABCO] (1), the component
would form X-bonds. However, as was the case for the 1,2-
DITFB/DABCO combination, this experiment yielded crystals
of a 2:1 TITFB/DABCO X-bonded complex. A subsequent
crystallization experiment using a 2:1 stoichiometric ratio of
1,3,5-TITFB/DABCO resulted in formation of crystals of
2
molecules are connected by halogen bonds in discrete
supramolecular trimers comprising ditopic DABCO which
forms halogen bonds to one of the iodine atoms of each 1,2-
DITFB molecule as shown in Figure 3. The complex has
[
(1,3,5-TITFB) DABCO] (2) in high chemical and structural
2
purity. The X-bonded complex in 2 is a supramolecular trimer,
with DABCO forming N···I halogen bonds to two 1,3,5-TITFB
molecules (Figure 5).
Figure 3. A pair of halogen bonded supramolecular trimers in [(1,2-
DITFB) DABCO] (1) (C gray, N blue, I purple, F off-white, I···N
2
halogen bonds are shown in black and white). Halogen bond details:
I2···N2 2.74 Å, C−I2···N2 170.3°; I3···N1 2.80, C−I3···N1 171.0°.
pseudo C2v symmetry with a bow-like shape due to the slight
deviation from colinearity of the two N···I halogen bonds.
Somewhat curiously, and perhaps even counterintuitively, the I
atoms not involved in halogen bonds (I1 and I4) project on the
same side of the molecule. Interestingly, this arrangement of
the I atoms in the structure is conducive to the formation of
discrete 3 + 3 supramolecular triangles. In the crystal, the
supramolecular complexes stack via π-interactions between 1,2-
DITFB molecules as illustrated in Figure 4. This results in
distinct fluorocarbon and hydrocarbon domains in the crystal.
Figure 5. In the structure of 2, the DABCO molecule forms N···I X-
bonds to two DITFB molecules making a trimeric [(1,2-
DITFB) DABCO] supramolecular complex with 2-fold crystallo-
2
graphic symmetry (C gray, N blue, I purple, F off-white, I···N halogen
bonds are shown in black and white). Adjacent trimers are connected
by I···I halogen bonds as indicated. Halogen bond details: I1···N1 2.76
Å, C−I1···N1 173.9°; I2···I1 3.84 Å, C−I2···I1 164.2°, C−I1···I2
107.6°.
The basic (2:1) structural unit in 2 bears a striking similarity
to that in 1, but there is additional intercomplex X-bonding
through I···I interactions. Atom I1, which is the X-bond donor
in the interaction with the DABCO-N atom, is also the X-bond
acceptor in an I···I interaction with I2. Thus, atom I1 is
involved in two X-bonds (as both donor and acceptor), and I2
is an X-bond acceptor while I3 is not involved in X-bonding.
The I···I X-bonds connect adjacent trimers into 2D tapes that
propagate parallel to the ab plane as illustrated in Figure 6a. π-
Stacking interactions between the 1,3,5-DITFB molecules of
adjacent 2D tapes result in their interdigitation (but not
interpenetration) as described in Figure 6b.
Both structures 1 and 2 could be described as “unsaturated”
in a supramolecular sense. As noted earlier, in X-bonding
structures containing para-substituted 1,4-DITFB, the DITFB
molecules almost always act as ditopic X-bond connectors with
both I atoms participating as X-bond donors. While there are
significantly fewer examples of X-bonded structures containing
Figure 4. Supramolecular trimers in 1 pack along the crystallographic
a axis through π-stacking of 1,2-DITFB molecules (C gray, N blue, I
purple, F off-white, I···N halogen bonds are shown in black and white).
1
,2-DITFB and 1,3,5-TITFB, there are examples of each in
It is possible that steric interference due to ortho-substitution
of the X-bond donor may be partly responsible for preventing
formation of X-bonds involving both I atoms of 1,2-DITFB in
which the I atoms are all involved in X-bonds. On the basis of
these examples, it was hypothesized that 1,2-DITFB would also
form two X-bonds and that with three peripheral I atoms, 1,3,5-
TITFB could form three X-bonds when challenged with
DABCO as the X-bond acceptor. While it is possible that
potential steric interference between DABCO molecules may
inhibit the formation of X-bonds at both I atoms of 1,2-DITFB,
in 1,3,5-TITFB, the I atoms are well spaced and project at 120°
1. The co-crystallization of the less sterically hindered 1,3,5-
TITFB with DABCO was then investigated. The initial
crystallization experiment was carried out using a 2:3
stoichiometric ratio of 1,3,5-TITFB:DABCO in line with the
hypothesis that all I atoms of TITFB and N atoms of DABCO
7
18
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
The crystallization mixtures containing DCM gave intriguing
results.
A chemical process, often referred to as the Menshutkin
19
reaction, occurs between tertiary amines and alkyl halides.
The accepted mechanism is one of nucleophilic substitution
and results in the formation of a salt comprising a quaternary
ammonium cation and the displaced chloride counteranion.
Dichloromethane (DCM) readily participates in the Men-
25
shutkin reaction, which is illustrated in Figure 7a. (While well
Figure 7. (a) The Menshutkin reaction between a tertiary amine and
dichloromethane gives a quaternary ammonium cation and the
chloride anion as products. (b) The Menshutkin reaction between
DABCO and DCM under ambient conditions results in formation of a
+
quaternary ammonium monocation (DABCO−CH Cl ) and chloride.
2
documented in the literature, this reaction is not widely
appreciated by the broader chemistry community. Many
procedures involve the use of tertiary amines in DCM,
particularly the widely used combination of TEA/DCM as an
eluent for column chromatography.) Here, in the crystallization
mixtures containing DCM, the Menshutkin reaction between
DABCO and DCM resulted in the in situ formation of a
+
−
tertiary amine salt of the form (DABCO−CH Cl )Cl (Figure
2
7
b), which has recently been shown to participate in the
formation of a novel coordination polymer. The chloride
anion is a well-known X-bond acceptor, and in these mixtures
Figure 6. (a) I···I halogen bond interactions between (1,2-
26
DITFB) DABCO units in 2 results in the formation of 2D tape-like
2
4
network (C gray, N blue, I purple, F off-white, I···N halogen bonds are
shown in black and white). (b) Adjacent tapes (one black, the other
green) interdigitate via formation of π-interactions.
halogen bonded networks were produced but with chloride as
the X-bond acceptor in place of DABCO.
Crystals of DABCO−CH Cl[(1,2-DITFB)Cl] (3) formed
2
from each other. Thus, there is no obvious reason why the 2:1
TITFB/DABCO complex 2 is preferred over the predicted 2:3
TITFB/DABCO complex. As stated above, lower than
predicted topological connectivity involving 1,3,5-TITFB has
from a 1:1 mixture of DABCO and 1,2-DITFB in DCM. The
chloride anion acts as a linear ditopic X-bond acceptor linking
1,2-DITFB molecules and both iodine atoms of the 1,2-DITFB
form X-bonds to chloride as shown in Figure 8. As predicted,
the combination of a linear ditopic X-bond acceptor with 1,2-
DITFB produced 1D zigzag chains that propagate with ca. 60°
vertices due to the supramolecular metrics of 1,2-DITFB.
Adjacent chains overlap forming π-interactions between 1,2-
DITFB molecules (Figure 9a). This arrangement produces a
rhombic cavity between the chains (Figure 9b). The π-stacking
propagates parallel to the c axis producing rhombic channels
17a
been commented on by the group of van der Boom. The
exact reason for this was not established, but the likely reason
was postulated to be the favorable packing arrangements
combined with the possibility that charge donation from X-
bond interaction at two iodine atoms of the 1,3,5-TITFB has a
detrimental effect on the capacity of the third iodine atom to
accept a pair of electrons to make the third X-bond. Our results
for 2 do not confirm or deny either possibility but lend weight
to the argument that with neutral nitrogen based X-bond
acceptors there appears to be a definite preference for fewer
than three X-bonds when combined with 1,3,5-TITFB.
occupied by the DABCO−CH Cl (1-(chloromethyl)-4-aza-1-
2
azonia-bicyclo[2.2.2]octane) cations. The cations no doubt play
an important role in lattice stabilization due to strong ionic
interactions with the chloride ions; however, they are not
directly involved in halogen bonding, despite the presence of
The lower than possible topological connectivity displayed in
13
1
1
and 2 piqued our interest, and attempts were made to
the Lewis basic N atom on the cations. C NMR and H NMR
spectra of single co-crystals were used to confirm the formation
of the chloromethyl quaternary ammonium cation formed by
the in situ Menshutkin reaction in this and subsequent
structures 4−6. The NMR spectra of the co-crystals are simple
superpositions of the spectra of the individual building blocks,
which is consistent with the complete dissociation of the single
crystals in solution and the absence of appreciable interaction
between the components under these conditions.
crystallize 1,2-DITFB with DABCO and 1,3,5-TITFB with
DABCO from a range of different solvent combinations
(
reagents co-crystallized from evaporating solutions of
methanol/water, acetone/water, acetonitrile/water, and di-
chloromethane (DCM)) in the search for polymorphs and in
particular those with the topologies initially predicted. Except
for reagent mixtures containing DCM, co-crystallization either
did not occur, or when it did, the products were as per 1 or 2.
7
19
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
Figure 8. In DABCO−CH Cl[(1,2-DITFB)Cl] (3), chloride acts as a
2
linear ditopic X-bond acceptor linking 1,2-DITFB molecules into 1D
zigzag chains that propagate with 60° vertices (C gray, N blue, I
−
purple, F off-white, I···Cl halogen bonds are shown in black and
green). Halogen bond details I1···Cl1 3.22 Å, C−I1···Cl1 173.8°,
I4···Cl1 3.08 Å, C−I4···Cl1 176.1°, I1···Cl1···I4 165.1°, I2···Cl2 3.11
Å, C−I2···Cl2 172.1°, I3···Cl2 3.06 Å, C−I3···Cl2 174.8°, I2···Cl2···I3
167.°
Initial attempts to crystallize 1,3,5-TITFB with DABCO in
neat DCM were frustrated by rapid precipitation of a fine
crystalline product of insufficient quality for single crystal X-ray
analysis. Given that the mechanism for the Menshutkin reaction
is nucleophilic substitution, it is intuitive that chloroform
should be more reactive than dichloromethane. However,
27
kinetic studies conducted by Gainza et al. indicate that at
ambient temperatures and pressures chloroform is essentially
unreactive toward tertiary amines. This is presumably due to
steric interference resulting from the presence of the additional
Cl atom on chloroform and made chloroform the ideal solvent
with which to dilute the dichloromethane for the purposes of
crystallization. Crystals of (DABCO−CH Cl) [(1,3,5-
Figure 9. (a) Adjacent zigzag chains in 3 overlap via π-interactions
creating space for the DABCO−CH Cl cations (C gray, N blue, I
2
−
purple, F off-white, I···Cl halogen bonds are shown in black and
green). (b) The X-bonded architecture combined with π-interactions
results in formation of rhombic channels in the crystal that run parallel
to the crystallographic c axis that are occupied by the cations.
2
3
TITFB) Cl ]·CHCl (4) were obtained from a mixture of
(Figure 11b), while chloroform molecules reside in the space
between the layers.
2
3
3
1
5
,3,5-TITFB and DABCO (in a 2:3 molar ratio) dissolved in a
0:50 mixture of dichloromethane and chloroform. As was to
The use of the Menshutkin reaction to generate chloride ions
from DCM in 3 and 4 is a convenient method for growing
halogen bonded co-crystals. The in situ generation of one of the
halogen bonding components, namely, chloride ion, facilitates
relatively slow crystallization from a nonevaporating solvent
thus producing high quality single crystals while minimizing
simultaneous crystallization of the pure reagents, which is a
common problem when attempting to produce co-crystals from
evaporating solvent. The obvious pathway for further work was
to expand the methodology by incorporating a different tertiary
amine to determine the viability of this methodology for the
preparation of a wider variety of co-crystals. This is particularly
important as the ability to alter the cation by using a range of
different tertiary amine starting compounds would provide a
further tool for engineering crystals of these materials.
Triethylamine (TEA) was the obvious first choice and as
such was used in co-crystallization mixtures containing either
1,2-DITFB or 1,3,5-TITFB.
be expected, dilution of the DCM with chloroform resulted in
slower crystallization (18 h) and produced crystals of sufficient
quality for X-ray structure determination in high yield (80%).
In 4, as in 3, the chloride ions form ditopic linear X-bond
bridges between iodofluorobenzene molecules. Each iodine of
the 1,3,5-TITFB is involved in one X-bond with the result
being formation of a 2D halogen bonded framework with 6,3-
network topology (Figure 10). Three such 2D nets inter-
penetrate in a rare example of Borromean entanglement as
shown in Figure 11a. While we predicted the formation of 2D
trigonal networks (in this case pseudotrigonal) by combination
of 1,3,5-TITFB with linear ditopic linkers, the observed
Borromean entanglement of the layers was a serendipitous
outcome. Discussion of the preference for Borromean
entanglement for this system is deferred to a later section of
this report. The triply interpenetrated nets are arranged such
that there are pairwise π-stacking interactions between TITFB
molecules creating cavities that are occupied by the cations
7
20
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
Figure 10. In (DABCO−CH Cl) [(1,3,5-TITFB) Cl ]·CHCl (4),
2
3
2
3
3
the TITFB molecules are tritopic nodes linked by chloride ions with
ditopic linear topology to create a pseudotrigonal 2D network (C gray,
−
N blue, I purple, F off-white, I···Cl halogen bonds are shown in black
and green).
S i n g l e c o - c r y s t a l s o f ( T E A - C H C l ) [ ( 1 , 2 -
2
3
DITFB) Cl ]·4H O (5) were obtained by the reaction of
2
3
2
TEA with DCM in the presence of 1,2-DITFB over a period of
days. In this structure the 1,2-DITFB molecules are arranged
5
in what look like zigzag chains in the ac plane linked by a
−
combination of Cl ···I halogen bonds as well as H O···HOH
2
−
and HOH···Cl hydrogen bonds as shown in Figure 12. One of
the chloride ions directly bridges two DITFB molecules
through C−I···Cl···I−C X-bond interactions. The other
chloride ion is the acceptor in an X-bond with an iodine
atom of DITFB and the acceptor of two hydrogen bond
interactions to water molecules. Each of these water molecules
participates in hydrogen bonding to two other water molecules
which are themselves involved in H-bonds to chloride ions
which in turn halogen bond to another 1,2-DITFB iodine atom.
Adjacent zigzag chains overlap through π-stacking of the 1,2-
DITFB molecules creating channels in the crystal occupied by
the cations (Figure 13a). The combination of π-stacking, and
mixed H-bond and X-bond connectivity among DITFB, water,
and chloride is illustrated in Figure 13b. The X-bond/H-bond
network propagates with 1D connectivity in the b direction and
as such connects the zigzag chains of DITFB to make an overall
network with 2D supramolecular connectivity. Adjacent 2D
sheets interdigitate in a zipper-like fashion through the π-
stacking of 1,2-DITFB molecules from contiguous sheets.
As predicted, the 1,2-DITFB acts as a ditopic X-bonding
node in the structure of 5; however, the inclusion of water and
the resulting hydrogen bond network highlight the potential for
cross-contamination of the interactions present, leading to
somewhat less than desirable levels of structure prediction
Figure 11. (a) In the structure of 4, three 2D pseudotrigonal networks
(
shown in black, green and red) interpenetrate in Borromean fashion
to create a supramolecular “tri-net” layer. (b) The spaces between the
Cl cations and the
individual 2D nets are occupied by DABCO−CH
spaces between the “tri-net” layers are occupied by chloroform
2
molecules. The overall lattice has pseudohexagonal appearance as
−
illustrated (C gray, N blue, I purple, F off-white, I···Cl halogen bonds
are shown in black and green).
such the potential exists for the creation of mixed hydrogen/
halogen bond supramolecular structures. Our attempts to
obtain anhydrous co-crystals comprising 1,2-DITFB and Cl−
have so far been unsuccessful, but we predict that such crystals
would display a zigzag chain topology linked solely through X-
bonds. The powder diffraction pattern obtained for the bulk
sample from this crystallization batch showed evidence of the
presence of a second phase (see Experimental Section for
(
bad) while at the same time expanding the potential structural
complexity (good). While chloride is an excellent halogen bond
acceptor, it is also an excellent hydrogen bond acceptor and as
7
21
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
Figure 12. In (TEA-CH Cl) [(1,2-DITFB) Cl ]·4H O (5), the 1,2-
2
3
2
3
2
DITFB molecules are connected into zigzag chains via two different
motifs. There is a direct ditopic linear I···Cl···I X-bond link through
Cl1. The second chloride ion (Cl2) is a tritopic linker forming one
halogen bond (to I2) and two hydrogen bonds to water molecules
(
associated with O1 and O2). Hydrogen bonding between water
molecules around a 2 screw axis connects water and chloride ions in
1
an infinite 1D chain that propagates in the b direction with 21
Figure 13. (a) Adjacent zigzag chains of 1,2-DITFB molecules in 5
overlap through π-interactions with cations occupying the space in
between. (b) Transverse view of the π-stacked DITFB molecules also
highlights the continuous connectivity of the hydrogen/halogen bond
network between I2 of 1,2-DITFB, chloride Cl2 and water molecules
crystallographic symmetry (shown in detail in Figure 13(b)) (C
−
gray, N blue, I purple, F off-white, I···Cl halogen bonds are shown in
−
black and green, HO−H···Cl hydrogen bonds are green and white,
HO−H···OH hydrogen bonds are red and white). Halogen bond
2
details: I1···Cl1 3.11 Å, C−I1···Cl1 177.6°, I1···Cl1···I1 170.5°,
I2···Cl2 3.22 Å, C−I2···Cl2 173.8°.
O1 and O2 which propagates with 2 symmetry in the b direction.
1
Hydrogen bond details: O1···Cl2 3.17 Å, O2···Cl2 3.20 Å, O···O
distances for hydrogen bonds 2.75 Å (with O1 donor) and 2.79 Å
details), which could possibly be anhydrous; however, we were
not able to obtain crystals of sufficient quantity or quality for
further analysis.
(with O2 donor).
ammonium and phosphonium iodides ((NEt )I and (PEt )I), it
4
4
When 1,3,5-TITFB was crystallized from DCM in the
presence of TEA, the Menshutkin reaction facilitated formation
is not surprising that the same overall structure (illustrated in
Figure 15b) forms for this suite of co-crystals.
of trigonal crystals with formula TEA-CH Cl[(1,3,5-TITFB)-
The contrasting structures of 4 and 6 are a particular point of
interest. Both halogen bonded networks incorporate 1,3,5-
TITFB acting as a tritopic connecting node. Both structures
2
Cl] (6). In these crystals, the TITFB molecule has crystallo-
graphic 3-fold symmetry and the iodine atoms are involved in
halogen bonds to chloride. Unlike structures 3, 4, and 5
however, the chloride ion is tritopic, simultaneously acting as
−
incorporate Cl as the halogen bond acceptor. Both display 6,3
−
network topology, but in 4 the Cl is a ditopic (essentially)
−
−
the halogen bond acceptor in three C−I···Cl halogen bonds to
linear connector while in 6 the Cl is a tritopic connector.
−
make 2D nets with trigonal symmetry and 6,3 topology with
Given that both structures comprise 1,3,5-TITFB and Cl as
−
Cl and TITFB alternating as connecting nodes in the network
the supramolecular connectors, it is immediately obvious that,
in principle, the same structure is possible for both. A question
arises as to why the molecules form interpenetrating nets with
Borromean entanglement in 4 but form nets with no
interpenetration in 6. The difference between the two
(Figure 14). The cations are completely enclosed in cavities
between contiguous 2D trigonal halogen bonded nets of 1,3,5-
TITFB and Cl− (Figure 15a). This crystal structure is
isostructural with a range of related structures with the general
+
−
+
−
formula (cation )[(1,3,5-TITFB)I ] (where cation = NEt ,
structures are the counter cations to the Cl (DABCO−
4
+
+
+
28
+
+
NPr , PEt , and SMe ). This lattice therefore appears
CH Cl in 4 and TEA-CH Cl in 6) which play no direct role
4
4
3
2
2
robust enough to resist disruption resulting from small changes
in the size of the cation. Given the similarity of the Menshutkin
in the halogen bonding but no doubt make very important
contributions through electrostatic charge balance and space
filling. The likely cause of the difference between the structures
product of TEA and DCM ([(TEA-CH Cl)Cl]) and the alkyl
2
7
22
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
Figure 15. (a) The combination of π-stacking of layers with rotation
of alternate layers in 6 (illustrated in Figure 14) produces cavities in
Figure 14. (a) The structure of TEA-CH Cl[(1,3,5-TITFB)Cl] (6)
the crystal that are occupied by the TEA-CH Cl cations. (b) The
crystal structure viewed down the c axis shows the resulting 3-fold
symmetry of the lattice.
2
2
has trigonal crystallographic symmetry. The 1,3,5-TITFB molecules
occupy 3-fold special positions. The chloride ions provide tritopic X-
bond links to the trigonal 1,3,5-TITFB nodes creating 2D nets
propagating in the ab plane. (b) Contiguous nets stack through π-
interactions between TITFB molecules and are rotated 60° around the
c axis. Halogen bond details: I1···Cl1 3.24 Å, C−I1···Cl1 173.9°,
I2···Cl1 3.31 Å, C−I2···Cl1 180.0°, I1···Cl1···I1 121.7°, I1···Cl1···I2
−
network in 4 (where Cl is tritopic). The resulting holes in the
network are too large to house just the cations as in 4, and thus
the requirements for space filling are further alleviated by
interpenetration. This whole arrangement is complemented by
the π-stacking of the TITFB molecules in the interpenetrated
nets. While it is not easy to reverse engineer such an
arrangement, it is clearly possible to create the conditions
under which it is more likely to occur. We note that the
structure of 4 also contains chloroform molecules. These
molecules reside in the space between adjacent triply linked
networks and not within those networks and so appear to play a
role in optimizing the filling of space between the layers rather
than a templating effect on the formation of the layers or their
interpretation.
119.1°.
results from the difference in the size, shape, and flexibility of
the counterions. The lattice described in 6 contains well-
defined cavities that are occupied by the cations. As stated
+
4
earlier, this lattice is stable to changes in the cation for NEt ,
+
+
+
28
PEt , NPr and SMe . However, it was found that the
4
4
3
+
lattice was disrupted when larger cations such as NBu4 or
+
+
PPh4 were incorporated. It appears that DABCO−CH Cl
2
which is larger and less flexible than the small alkyl ammonium
ions is unsuited to the cavity in 6 and thus demands a different
halogen bonded topological framework in which to reside.
Resnati and Metrangolo showed that by carefully controlling
the size of the holes in 6,3 halogen bonded networks it is
possible to create the right conditions for Borromean
CONCLUSIONS
■
1,2-DITFB and 1,3,5-TITFB demonstrated low topological
connectivity when co-crystallized with the linear ditopic base
DABCO. This was somewhat surprising, particularly for 1,2-
DITFB, given that the para-isomer 1,4-DITFB most commonly
acts as a linear ditopic X-bond linker with nitrogen based
2
9
interpenetration. In the case of 6, it appears that the spatial
demands of the cation lead to the formation of the 6,3 nets with
−
Cl as a linear linker in preference to the tightly connected
7
23
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724

Crystal Growth & Design
Article
halogen bond acceptors, including with DABCO. Lower than
expected topological connectivity for 1,3,5-TITFB has been
observed previously, and thus our results confirm a preference
for fewer than three halogen bond interactions at least as far as
neutral nitrogen based X-bond acceptors are concerned. The
Menshutkin reaction between dichloromethane and tertiary
amine bases was used to create chloride ions during
crystallization of 1,2-DITFB and 1,3,5-TITFB. The X-bonded
(12) Liantonio, R.; Metrangolo, P.; Pilati, T.; Resnati, G. Acta
Crystallogr., Sect. E 2002, 58, o575.
(
7
(
13) Cinci
524−7525.
14) Cauliez, P.; Polo, V.; Roisnel, T.; Llusar, R.; Fourmigue, M.
CrystEngComm 2010, 12, 558−566.
15) (a) Jay, J. I.; Padgett, C. W.; Walsh, R. D. B.; Hanks, T. W.;
̌ ́ ̌ ̌ ́
c, D.; Friscic, T.; Jones, W. J. Am. Chem. Soc. 2008, 130,
(
Pennington, W. T. Cryst. Growth. Des. 2001, 1, 501−507. (b) Arman,
H. D.; Gieseking, R. L.; Hanks, T. W.; Pennington, W. T. Chem.
Commun. 2010, 46, 1854−1856.
−
networks in the resulting crystals incorporated Cl as X-bond
acceptors with the iodofluorobenzene donors. This method-
ology was found to produce high quality crystals of new X-
bonded architectures including a rare example displaying
Borromean entanglement. The nature of the countercation
(16) Cavallo, G.; Biella, S.; Lu, J.; Metrangolo, P.; Pilati, T.; Resnati,
G.; Terraneo, G. J. Fluorine Chem. 2010, 131, 1165−1172.
(17) (a) Lucassen, A. C. B.; Karton, A.; Leitus, G.; Shimon, L. J. W.;
Martin, J. M. L.; van der Boom, M. E. Cryst. Growth Des. 2007, 7, 386−
392. (b) Vartanian, M.; Lucassen, A. C. B.; Shimon, L. J. W.; van der
Boom, M. E. Cryst. Growth Des. 2008, 8, 786−790.
+
+
(
DABCO−CH Cl or TEA-CH Cl ) resulting from the
2
2
reaction of the tertiary amine was found to influence the
crystal packing motifs. Given the wide variety (size, structure,
functionality) of possible tertiary amines that could be
employed to produce X-bonded networks by the same
procedure, there is significant potential to subtly engineer the
resulting crystal packing. We propose to do this through
systematic changes of the amine and this will be a major focus
of our continuing research.
(
18) Roper, L. C.; Prasang, C.; Kozhevnikov, V. N.; Whitwood, A. C.;
Karadakov, P. B.; Bruce, D. W. Cryst. Growth Des. 2010, 10, 3710−
720.
3
(
19) (a) Menshutkin, N. Z. Phys. Chem. 1890, 6, 41−57. (b) Abboud,
J. L. M.; Notario, R.; Bertran, J.; Sola, M. Prog. Phys. Org. Chem. 1993,
19, 1−182.
(20) Wenk, H. H.; Sander, W. Eur. J. Org. Chem. 2002, 3927−3935.
(21) Oxford Diffraction, CrysAlis CCD and CrysAlis RED; Oxford
Diffraction Ltd: Abingdon, Oxfordshire, England, 2007.
22) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837−838.
(
ASSOCIATED CONTENT
■
(23) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.;
Giocovazzo, C.; Guagliardi, A.; Moliterni, A. G. C.; Polidori, G.;
Spagna, S. J. J. Appl. Crystallogr. 1999, 32, 115−119.
*
S
Supporting Information
X-ray crystallographic information files (CIF) are available for
−6. Simulated and experimental powder diffraction patterns
(
24) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112−122.
1
1
13
(25) (a) Almarzoqi, B.; George, A. V.; Isaacs, N. S. Tetrahedron 1986,
are provided for 1, 2, 3, 5, and 6. H- and C NMR spectra for
4
2, 601−607. (b) Nevstad, G. O.; Songstad, J. Acta Chem. Scand. 1984,
3
−6 are provided along with Raman spectra for 1−6. This
B38, 469−477. (c) Rudine, A. B.; Walter, M. G.; Wamser, C. C. J. Org.
Chem. 2010, 75, 4292−4295. (d) Matsumoto, K.; Hashimoto, S.;
Ikemi, Y.; Otani, S. Heterocycles 1984, 22, 1417−1420. (e) Iwamoto,
E.; Yokoyama, T.; et al. Bull. Chem. Soc. Jpn. 1991, 64, 523−527.
(f) Winter, R. F.; Wolmershauser, G. J. Organomet. Chem. 1998, 570,
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
2
(
(
01−218.
Corresponding Author
26) Xiao, J.-M. Acta Crystallogr. 2010, C66, m348−m350.
*
Phone: +61 7 3138 1220. Fax: +61 7 3138 1804. E-mail j.
mcmurtrie@qut.edu.au.
27) (a) Gainza, A. H.; Konyeaso, R. I.; Asenjo, R. A. J. Pharm. Sci.
1994, 32, 226−232. (b) Gainza, A. H. Int. J. Chem. Kinet. 2004, 36,
00−509.
28) Metrangolo, P.; Meyer, F.; Pilati, T.; Resnati, G.; Terraneo, G.
Chem. Commun. 2008, 1635−1637.
29) Liantonio, R.; Metrangolo, P.; Meyer, F.; Pilati, T.; Navarrini,
W.; Resnati, G. Chem. Commun. 2006, 1819−1821.
5
(
REFERENCES
■
(
1) (a) Legon, A. C. Phys. Chem. Chem. Phys. 2010, 12, 7736−7747.
(
(
b) Fourmigue,
́
M. Curr. Opin. Solid State Mater. Sci. 2009, 13, 36−45.
(
c) Metrangolo, P.; Meyer, F.; Pilati, T.; Resnati, G.; Terraneo, G.
Angew. Chem., Int. Ed. 2008, 47, 6114−6127. (d) Rissanen, K.
CrystEngComm 2008, 10, 1107−113. (e) Metrangolo, P.; Resnati, G.;
Pilati, T.; Biella, S. Struct. Bonding (Berlin) 2008, 6, 105−136.
(
2) Metrangolo, P.; Pilati, T.; Resnati, G. CrystEngComm 2006, 8,
9
46−947.
3) Politzer, P.; Lane, P.; Concha, M. C.; Ma, Y.; Murray, J. S. J. Mol.
Model. 2007, 13, 305−311.
4) Metrangolo, P.; Neukirch, H.; Pilati, T.; Resnati, G. Acc. Chem.
Res. 2005, 38, 386−395.
5) Metrangolo, P.; Pilati, T.; Tarraneo, G; Biella, S.; Resnati, G.
CrystEngComm 2009, 11, 1187−1196.
6) Metrangolo, P.; Resnati, G.; Pilati, T.; Liantonio, R.; Meyer, F. J.
Polym. Sci., Part A: Polym. Chem. 2007, 45, 1−15.
7) Hanson, G. R.; Jensen, P.; McMurtrie, J.; Rintoul, L.; Micallef, A.
S. Chem.Eur. J. 2009, 15, 4156−4164.
8) Davy, K. J. P.; McMurtrie, J.; Rintoul, L.; Bernhardt, P. V.;
(
(
(
(
(
(
Micallef, A. S. CrystEngComm 2011, 13, 5062−5070.
(
(
9) Allen, F. Acta Crystallogr., Sect. B: Struct. Sci. 2002, 58, 380−388.
10) Walsh, R. B.; Padgett, C. W.; Metrangolo, P.; Resnati, G.;
Hanks, T. W.; Pennington, W. T. Cryst. Growth Des. 2001, 1, 165−
75.
11) Cinci T.; Jones, W. Chem.Eur. J. 2008, 14, 747−
53.
1
(
̌
c,
́
D.; Frisc
̌
i
̌
c,
́
7
7
24
dx.doi.org/10.1021/cg201017r | Cryst. Growth Des. 2012, 12, 714−724