Correlation between 13C chemical shifts and the halogen bonding environment in a series of solid para-diiodotetrafluorobenzene complexes

Article abstract of DOI:10.1039/c3ce26750d

The co-crystallization of para-diiodotetrafluorobenzene (p-DITFB) with ammonium and phosphonium halide (Cl^- and Br^-) salts afforded four new compounds, [(n-Bu₄PCl)(p-DITFB)] (2), [(n-Bu ₄NBr)(p-DITFB)] (3), [(n-Bu₄PBr)(p-DITFB)] (4), and [(EtPh₃PBr)₂(p-DITFB)] (5), that exhibit moderately strong halogen bonding interactions. They have been characterized by single-crystal X-ray diffraction and ¹³C solid-state nuclear magnetic resonance (SSNMR) spectroscopy in magnetic fields of 9.4 and 21.1 T. The X-ray crystallography shows that in 2, 3, and 4, the halide is ditopic and forms long polymeric zigzag chains, whereas the bromide in 5 forms a dianionic species when involved in halogen bonding interactions. The NMR data, when combined with zeroth-order regular approximation density functional theory (ZORA-DFT) calculations, provide insight into the relationship between the strength of the halogen bond and the ¹³C isotropic chemical shift. When the carbon-iodine bond length increases, the ¹³C chemical shift also increases. Further insights into the relationship between halogen bonding and the ¹³C chemical shifts are obtained through additional systematic ZORA-DFT calculations as a function of the halogen bonding environment.

Full text of DOI:10.1039/c3ce26750d

CrystEngComm
View Article Online
View Journal | View Issue
PAPER
13
Correlation between C chemical shifts and the
halogen bonding environment in a series of solid para-
diiodotetrafluorobenzene complexes3
Cite this: CrystEngComm, 2013, 15,
3168
Jasmine Viger-Gravel, Sophie Leclerc, Ilia Korobkov and David L. Bryce*
The co-crystallization of para-diiodotetrafluorobenzene (p-DITFB) with ammonium and phosphonium
halide (Cl2 and Br2) salts afforded four new compounds, [(n-Bu4PCl)(p-DITFB)] (2), [(n-Bu4NBr)(p-DITFB)]
(3), [(n-Bu4PBr)(p-DITFB)] (4), and [(EtPh3PBr)2(p-DITFB)] (5), that exhibit moderately strong halogen
bonding interactions. They have been characterized by single-crystal X-ray diffraction and 13C solid-state
nuclear magnetic resonance (SSNMR) spectroscopy in magnetic fields of 9.4 and 21.1 T. The X-ray
crystallography shows that in 2, 3, and 4, the halide is ditopic and forms long polymeric zigzag chains,
whereas the bromide in 5 forms a dianionic species when involved in halogen bonding interactions. The
NMR data, when combined with zeroth-order regular approximation density functional theory (ZORA–
DFT) calculations, provide insight into the relationship between the strength of the halogen bond and the
13C isotropic chemical shift. When the carbon–iodine bond length increases, the 13C chemical shift also
increases. Further insights into the relationship between halogen bonding and the 13C chemical shifts are
obtained through additional systematic ZORA–DFT calculations as a function of the halogen bonding
environment.
Received 23rd October 2012,
Accepted 17th January 2013
DOI: 10.1039/c3ce26750d
www.rsc.org/crystengcomm
drug design due to its appealing potential to generate
intermolecular stabilizing forces between electron-rich atoms.4
Introduction
Hydrogen bonds form a very important class of intermolecular
interaction which plays a key role in diverse areas of chemistry
and biochemistry, including molecular recognition, folding,
and stability of biological macromolecules. Halogen bonds
(XB) are in many ways analogous to hydrogen bonds, and the
former have also been shown to be important in forming
complexes between halogen-substituted ligands and their
biological substrates.1 Although there is ongoing discussion
of the nature of halogen bonding, to a first approximation,
both XB and hydrogen bonding can be treated as directional
and electrostatic interactions between a Lewis acid and base.
Importantly, the directionalities of the two types of bonding
interactions tend to differ.2 Halogen bonding interactions
were first described many years ago as charge transfer bonds,3
but it is only in the last decade or so that they have taken on a
more prominent role in crystal engineering, partly because
they can be stronger and more directional than hydrogen
bonds. The XB has also been highlighted for its potential in
In light of these applications, it is advantageous to have a
clear definition of this interaction. It has proven difficult to
formulate a definition that applies to such a broad spectrum of
situations. The International Union of Pure and Applied
Chemistry (IUPAC) sponsored a workshop last year on this
subject5 and has since released a provisional recommendation
on a general definition for the phenomenon.6,7 Briefly, there
must be a net attractive interaction between the electrophilic
region present at a halogen and an electron-rich moiety of
another molecule. More clearly, the halogen, X, is typically part
of a molecule, RX, and interacts with an electron-rich atom or
group, D (e.g., Lewis base, halide, p-electrons). The electro-
philic region of the halogen, known as a s-hole,8 arises from
the covalent bond formed between R and X. The presence of
electronegative substituents on R increases the electron-
withdrawing capacity of R and this creates a region of positive
electrostatic potential on the halogen, opposite the covalent
bond which is surrounded by a negative electrostatic potential.
Nuclear magnetic resonance (NMR) has proven to be a very
good tool for the characterization of hydrogen bonding in
solution and in solids.9 Following suit, 1H (I = 1/2, natural
abundance (n.a.) 99.98%) NMR has also been used for the
study of halogen bond formation. This was observed by a
chemical shift (CS) change for the methylene protons of
CH2Cl2, CH2Br2, and CH2I2 and the haloformic protons
Department of Chemistry and Centre for Catalysis Research and Innovation
University of Ottawa, 10 Marie Curie Private, Ottawa, Ontario, Canada K1N 6N5.
E-mail: dbryce@uottawa.ca; Fax: +1 613 562 5170; Tel: +1 613 562 5800 ext. 2018
Electronic supplementary information (ESI) available: CIF files; further
3
crystallographic details; figures depicting the unit cells and crystal packing; full
13C SSNMR spectra; complete list of calculated chemical shifts. CCDC 907312–
907315. For ESI and crystallographic data in CIF or other electronic format see
DOI: 10.1039/c3ce26750d
3168 | CrystEngComm, 2013, 15, 3168–3177
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
(CHCl3, CHBr3, and CHI3) in the presence of various amines,
ethers, and non-polar electron-rich solvents.10,11 Resnati and
co-workers also used multinuclear solution NMR to demon-
strate the presence of XB between quinuclidine and 1-iodoper-
fluoropropane.12 The 1H and 13C (I = 1/2, n.a. 1.07%) NMR
spectra showed changes in the CS of the interacting
compounds relative to the isolated components. Additional
and covalently-bonded iodine, and any possible competing
effects of hydrogen bonding are absent. This will facilitate the
elucidation of
a relationship between halogen bonding
environment and NMR parameters.
The proposed IUPAC definition of halogen bonding states
that changes in magnetic resonance parameters may accom-
pany the formation of a halogen bond.6 We are particularly
interested in understanding what drives the change in
chemical shift when XB occurs. In solution, 13C NMR has
been used to probe XB and there have been strong correlations
between the electron-donor ability of the solvent and the
chemical shift.17–19 In the solid state, it is more difficult to
observe the resonance of the covalently-bonded carbon
involved in halogen bonding interactions due to the strong
residual dipolar interactions with the halogen nucleus20,21 and
long relaxation times. Our present approach involves a
combination of X-ray diffraction, multi-field SSNMR, and
computational analysis. In this study, we report the single
crystal X-ray structures and 13C SSNMR spectra of five halogen
bonded compounds constructed from para-diiodotetrafluor-
obenzene (p-DITFB) and ammonium or phosphonium halide
(X = Cl, Br) salts (n-Bu4PCl, n-Bu4NCl, n-Bu4PBr, n-Bu4NBr, and
EtPh3PBr). The compounds depicted in Fig. 1 have the
following formulas: [(n-Bu4NCl)(p-DITFB)] (1), [(n-Bu4PCl)
(p-DITFB)] (2), [(n-Bu4NBr)(p-DITFB)] (3), [(n-Bu4PBr)
(p-DITFB)] (4), [(EtPh3PBr)2(p-DITFB)] (5). Compounds 1 and
3 were previously reported by Abate et al.; however, the
structure solved by our group for 3 is slightly different (vide
infra).22 The crystallographic structure of each compound will
be discussed with an emphasis on the subtle differences in the
various halogen bonding networks. We will also compare the
13C cross-polarization magic-angle-spinning (CP MAS) SSNMR
spectra of the XB compounds to that of p-DITFB. Properly
describing the NMR parameters in these compounds involves
many challenges which will be discussed (vide infra); we
employ the zeroth-order regular approximation density func-
…
evidence of a C–I N electrostatic interaction originated from
19F (I = 1/2, n.a. 100%) and 14N (I = 1, n.a. 99.63%) NMR
spectra, where chemical shifts changed by several ppm upon
complexation. Metrangolo and co-workers used 19F NMR to
quantify the halogen bonding interaction in various halo-
perfluorocarbons and heteroatom-containing hydrocarbons by
correlating the solvent’s electron donating ability to chemical
shifts.13 The Taylor group has also used 19F NMR in titration
experiments to determine the strength of the halogen bonding
interaction in solution.14 Our group is interested in the study
of the XB via solid-state NMR (SSNMR), where solvent effects
are absent, and complete NMR interaction tensors may afford
additional insight into the XB phenomenon. Our initial
studies probed the halogen bonding between selenocyanates
and diiodotetrafluorobenzene using 77Se (I = 1/2, n.a. 7.63%)
NMR.15 We observed that, in the presence of a XB, the smallest
component of the 77Se CS tensor (d33) decreases substantially.
Another study looked at halides involved in the XB directly
using chlorine, bromine, and iodine SSNMR in a series of
haloanilium halides.16 The 81Br (I = 3/2, n.a. 49.31%, nuclear
electric quadrupole moment (Q) = 261.5 mb) quadrupolar
interaction (QI), isotropic CS (diso), and CS span (V) all proved
to vary slightly as a function of the strength of the halogen
bonding interaction. One difficulty encountered in that study
was the simultaneous presence of halogen bonds and
hydrogen bonds to halide ions, thereby complicating the
interpretation of results solely in terms of the halogen bonding
environment. Furthermore, the halogen bonds were quite
weak. In the present work, the compounds have been designed
such that only halogen bonding is present between halides
Fig. 1 Halogen bonding contacts in compounds 1 to 5 with the corresponding formula and compound number. In compounds 1 to 4, the halide interacts with two
crystallographically distinct iodines, whereas in compound 5 the bromide interacts with one. Cations are omitted for clarity.
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 3168–3177 | 3169

View Article Online
Paper
CrystEngComm
Table 1 Crystallographic data and selected data collection parameters
Compound
2
3
4
5
Empirical formula
C22H36ClF4I2P
696.730
C22H36BrF4I2N
724.23
C44H72Br2F8I4P2
1482.38
C23H20BrF2IP
572.17
Formula weight/g mol21
Crystal size/mm
Crystal system
Space group
0.17 6 0.15 6 0.14
Monoclinic
P21/c
0.18 6 0.13 6 0.11
Orthorhombic
Pccn
0.17 6 0.12 6 0.11
Monoclinic
C2/c
0.31 6 0.19 6 0.19
Monoclinic
P21/n
Z
4
8
4
4
a/Å
b/Å
c/Å
14.4599(2)
14.9417(2)
15.0638(2)
90
117.683(1)
90
2882.1(1)
1.606
2.364
1368
48.3095(10)
8.7022(2)
13.6477(3)
90
90
90
5737.5(2)
1.677
3.620
2816
49.7447(10)
8.4785(2)
14.1541(3)
90
104.119(1)
90
5789.3(2)
1.701
3.641
2880
10.8351(3)
11.8074(3)
17.0179(4)
90
92.3430(10)
90
2175.35(10)
1.747
3.405
a/u
b/u
c/u
Volume/Å3
Calculated density/Mg m23
Absorption coefficient/mm21
F(000)
1116
h range for data collection/u
2.05 to 28.31
219 ¡ h ¡ 19
219 ¡ k ¡ 19
220 ¡ l ¡ 19
62 401/7118
0.0186
1.69 to 28.26
264 ¡ h ¡ 53
211 ¡ k ¡ 11
218 ¡ l ¡ 18
58 424/7074
0.0297
1.69 to 28.33
266 ¡ h ¡ 64
0 ¡ k ¡ 11
0 ¡ l ¡ 18
15 069/7049
0.0000
2.10 to 28.33
214 ¡ h ¡ 14
214 ¡ k ¡ 15
222 ¡ l ¡ 19
23 686/5372
0.0233
Limiting indices
Reflections collected/unique
Rint
Completeness to h = 28.32/%
Max and min transmission
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I . 2a(I)]
R indices (all data)
99.1
99.2
97.6
98.9
0.7331 and 0.6894
7118/75/298
1.027
R1 = 0.209, wR2 = 0.0492
R1 = 0.269, wR2 = 0.0537
0.671 and 20.714
0.6916 and 0.5619
7074/244/319
1.122
R1 = 0.0668, wR2 = 0.1844
R1 = 0.0742, wR2 = 0.1888
1.001 and 21.481
0.6902 and 0.5765
7049/221/347
1.002
R1 = 0.0429, wR2 = 0.1057
R1 = 0.0570, wR2 = 0.1136
1.007 and 21.448
0.5639 and 0.4183
5372/0/254
1.045
R1 = 0.0167, wR2 = 0.0452
R1 = 0.0190, wR2 = 0.0464
0.346 and 20.650
Largest diff peak/hole/e Å23
their halogen bonding environments. Compounds 1 and 3
have been reported previously and discussed thoroughly in the
context of halogen bonding by Abate et al.22 Similar
compounds have previously been discussed in terms of
halogen bonding by Triguero et al., where 1,3,5-trifluoro-
2,4,6-triiodobenzene is co-crystallized with n-Bu4NX or
n-Bu4PX (X = Cl2, Br2).27 Grebe et al. report donor–acceptor
compounds involving Ph4PX or Me4NX (X = Cl2, Br2) and
p-DITFB.28 Hence, in this section our new crystal structures are
briefly described and analyzed in terms of how they compare
with previously reported crystallographic data. The halogen
bonding motifs observed here are reminiscent of some of
those reported for artificial anion receptors and related
complexes.29
Compound 1 exhibits the same unit cell parameters and
halogen bonding environment as previously reported;22 the
chlorides enable the formation of long chains by interacting
with iodines from two different p-DITFB molecules (see Fig. 1).
As for 3, Abate et al. have solved a different structure that
crystallizes in the monoclinic space group C2/c. Presently, co-
crystallization of n-Bu4NBr and p-C6F4I2 in a different solvent
(dichloromethane) afforded 3, but which packs in the
orthorhombic Pccn space group. The main difference between
this polymorph and the previously reported one is disorder of
tional theory (ZORA–DFT) approach which allows for the
treatment of scalar and spin–orbit relativistic effects.23
Results and discussion
X-ray crystal structures
Single-crystal X-ray crystallography was used to characterize
the halogen bonded compounds 2, 3, 4, and 5. Relevant
crystallographic data are presented in Table 1. Halogen
bonding interactions are considered to be present when the
distance between the halogen (i.e., iodine) and the halide (i.e.,
X2 = Cl2, Br2), dI–X, is shorter than the sum of their Van der
Waals (VdW) radii (1.75, 1.85, 1.98 Å for Cl, Br, I,24 and 1.81,
1.96, 2.20 Å for Cl2, Br2, and I2, respectively25). The ratio of
the short intermolecular contact distance and the sum of the
VdW radii (dVdW) is presented as the normalized distance
parameter, RXB = dI–X/SdVdW, and is useful to describe the
short contacts (i.e., RXB = 1 is a weak to non-existent XB).26
Furthermore, the carbon–iodine bond length, dC–I, will change
for each halogen bonded compound largely as a consequence
of the non-covalent interaction. As described in the proposed
IUPAC definition of XB, the length of the R–X covalent bond
usually increases upon XB formation.6 A further point is that
the XB interaction has been observed to be almost linear in the
case of a nucleophile interacting with an RX molecule. The
…
the aromatic ring in the ac plane (Fig. S5, see ESI ). Compound
3
2 packs in the same space group and crystal system as
compound 1 (P21/c). Compound 4 also packs in a monoclinic
crystal system, but in a different space group (C2/c). Similarly
to compound 3, the aromatic ring in the ac plane in 4 exhibits
halide will often align itself at an angle, hC–I , that varies
X2
between 160 and 180u. These intermolecular contact distances
and angles are reported in Table 2 for the compounds of
interest as well as other geometrical information pertaining to
disorder (see Fig. S6 in ESI ).
3
3170 | CrystEngComm, 2013, 15, 3168–3177
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
Table 2 Selected intermolecular contact distances and anglesa
observed previously by Grebe et al. in the compounds
(Ph4P)2[(p-C6F4I2)Cl2]?4CH2Cl2
and (Ph4P)2[(p-C6F4I2)Br2]?
#I dI–C/Åb dI X/Å RXB hC–I X/u hI
I/u
c
4CH2Cl2.28 Rows of EtPh3P+ cations alternate with rows of
…
…
… …
X
Compound
22
…
…
[Br I–C6F4–I Br]
moieties along the b axis (Fig. S7, see
1
2
3
4
[(n-Bu4NCl)(p-C6F4I2)]d
[(n-Bu4PCl)(p-C6F4I2)]
[(n-Bu4NBr)(p-C6F4I2)]
[(n-Bu4PBr)(p-C6F4I2)]
[(EtPh3PBr)2(p-C6F4I2)]
1
2
1
2
1
2
1
2
1
2.107 2.988 0.79 178.1
109.1
155.6
139.2
140.9
—
ESI ). Also, a view of the unit cell of 5 shows that the cations
3
2.096 3.104 0.82 170.2
2.095 3.038 0.80 175.3
2.102 2.976 0.79 176.9
2.112 3.168 0.80 175.3
2.096 3.232 0.82 177.9
2.105 3.189 0.81 176.8
2.099 3.196 0.81 177.7
2.108 3.148 0.80 175.5
are associated two by two into an inversion-centered phenyl
embrace motif. Triguero et al. observed the same pattern in
(Ph4P)[(sym-C6F3I3)Br]27 and mention that it is a recurrent
motif for Ph3PR derivatives.30,31 Again, the bromide adopts
2
…
characteristic XB by forming systematic strong y180u I Br
5
…
Br2
contacts (R = 0.80 and hC–I
= 175.5u).
a
b
See also models in Fig. 1. Experimental values determined by
X-ray crystallography. RXB, normalized distance parameter, RXB
dI X/SdVdW, where dI is the shortest contact distance between the
Solid-state NMR spectroscopy: 13C chemical shifts
c
=
Presented in Fig. 2 are sections of the 13C CP MAS NMR
spectra corresponding to the covalently bonded C–I regions for
solid powdered samples of 1 to 5, and for p-DITFB. Such
spectra are challenging to acquire with a good signal-to-noise
ratio given the lack of directly bonded protons on the carbons
of interest, the directly bonded iodine, the nearby fluorines,
and long relaxation times. The 13C chemical shifts for each of
the compounds considered in this work are summarized in
Table 3. These were obtained by a simultaneous simulation of
each of the spectra acquired at both 9.4 and 21.1 T magnetic
field strengths, shown in Fig. 2(g–l) and (a–f), respectively. The
full 13C CP MAS SSNMR spectra for compounds 1 to 5 and for
p-DITFB acquired at both fields may be found in Fig. S10 and
…
…
X
halogen and the halide and dVdW is the sum of their Van der Waals
radii (1.98 Å for I, and 1.81, 1.96 Å for Cl2 and Br2, respectively).
This compound was previously reported by Abate et al.22
d
In all of the halogen bonded compounds reported here with
the exception of 5, the two iodines of p-DITFB act as acceptors
2
…
of electron density, and they form almost linear C–I
X
arrangements with the donor of electron density (i.e., Cl2 or
2
2
…
Br ). The C–I
X
angles range from 170.2 to 178.1u for
compounds 1 to 4, and their normalized distance parameter,
RXB, values range between 0.79 and 0.82 (see Table 2). Hence,
these compounds can be classified as having moderately
strong halogen bonding interactions, since the short linear
S11, ESI. The assignments of their chemical shifts are
3
tabulated in Table S1, ESI.
3
…
contacts, dI X, observed are y20% shorter than the sum of
Several important points can be made concerning the 13C CP
MAS solid-state NMR spectra. The use of the 21.1 T instrument
allows for improved precision of the 13C chemical shifts since
the residual dipolar coupling (RDC) between 13C and quad-
rupolar halogen nuclides (i.e., 127I) decreases significantly at
such a field,32 and this is observed by a decrease in the line
widths (80 to 200 Hz) of the resonances compared to the ones
observed at 9.4 T (137 to 289 Hz). Residual dipolar coupling
was satisfactorily accounted for with the use of apodization, as
no specific asymmetric broadening or splitting patterns
attributable to RDC were observed. This approach was
validated by the consistent field-independent chemical shift
values obtained in both applied magnetic fields. From the
crystal structures, two resonances are expected in 1, 2, 3, and 4
for the carbons bonded to iodine whereas in 5 only one is
anticipated. The spectra acquired at 21.1 T are consistent with
the expected number of C–I resonances in all compounds with
an increase in spectral resolution when compared to data
acquired at 9.4 T. One particularly clear and unambiguous
their VdW radii. These XB compounds form polymeric anionic
zigzag chains with halide bridges and have different coordina-
2
…
…
tion angles. The I Cl I angles are 109.1u and 155.6u for 1
2
…
…
and 2 respectively, and the I Br I angles are 139.2u and
140.9u for 3 and 4, respectively (see Fig. 1). In these
compounds, the two iodines which interact with the halide
are crystallographically non-equivalent. Similar halogen bond-
ing environments were also observed by Grebe et al.,
where the zigzag chains for [(Me4NCl)(p-DITFB)] and
(Me4N)[(p-C6F4I2)Br]?CH3CN had, however, more acute angles,
hI
I = 77.2u and hI
I = 74.3u.28 Furthermore, Abate et al.
…
Cl2
…
…
Br2
…
2
…
…
found an almost identical I Br I angle for compound 3
(138.4u).22
The crystal packing is similar for compounds 1 and 2. The
rows of polymeric chains between the halide and the p-DITFB
molecules are separated by a row of cations. This is observed
along the a axis for compounds 1 and 2, as depicted in Fig. S1
and S3, respectively (ESI ). Also, the butyl chains in compound
3
aspect of the spectra in Fig. 2 is that an increase in diso(
13C) is
2 exhibit some disorder. In contrast to 1 and 2, the polymeric
chains in 3 and 4 formed by the two molecules of p-DITFB
bridged by a bromide are not in the same plane; the non-
disordered aromatic ring p-DITFB is in the ab plane, oriented
almost perpendicular to the disordered aromatic ring. The
cations, n-Bu4N+ or n-Bu4P+, and bromide anions alternate in
observed for the C–I resonance in the presence of halogen
bonding relative to pure p-DITFB (76.50(0.50) ppm). These
range from 80.64(0.02) ppm for 1 to 84.72(0.10) ppm for 3. The
fact that these chemical shifts are substantially lower than
those typically observed for aromatic carbons is due to a well-
known relativistic spin–orbit-induced heavy atom substituent
effect caused by the covalently bonded iodine.33
It is of interest to further understand why changes in the
value of diso(13C) occur for the C–I carbons in the presence of
halogen bonding, relative to p-DITFB alone, where such an
interaction is absent. It was mentioned in the previous section
columns along the b axis, (Fig. S5 and S6, see ESI ).
3
Finally, compound 5 is formed between EtPh3PBr and
p-C6F4I2, resulting in [(EtPh3PBr)2(p-DITFB)] which crystallizes
in a monoclinic system with a P21/n space group. The bromide
…
anion acts as a terminal donor, forming a dianionic [Br I–
22
…
C6F4–I Br]
species (see Fig. 1). This scaffold has been
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 3168–3177 | 3171

View Article Online
Paper
CrystEngComm
Fig. 2 Experimental (solid line) and simulated (dashed line) 13C CP MAS SSNMR spectra for the region of the carbon covalently bonded to iodine. The spectra on the
left were recorded at 21.1 T with a MAS speed of 18 kHz and correspond to p-DITFB (a), 1 (b), 2 (c), 3 (d), 4 (e), and 5 (f). The right column corresponds to the spectra
recorded at 9.4 T with a spinning speed of 8 kHz, where the spinning sidebands are indicated by a red asterisk and carbon atoms from the butyl chains are cut with a
red dash. The spectra correspond to p-DITFB (g), 1 (h), 2 (i), 3 (j), 4 (k), and 5 (l).
that the geometry of the halogen bond is different in each of
the compounds studied. Firstly, the value of diso for the 13C
directly bonded to the iodine involved in halogen bonding was
plotted as a function of the strength of the XB by using the
normalized distance parameter, RXB. No clear correlation
could be distinguished for the series of XB compounds
presented here. Secondly, the C–I distances, dC–I, are different
for each of the halogen bonded compounds, and this distance
is one possible indicator of halogen bonding.6 When the
values of diso(13C) are plotted as function of their respective C–I
as well (Fig. 3). It is critical to note here that the assignments
of the two values of diso(13C) to the two different crystal-
lographic sites in a particular compound were determined
independently from ZORA–DFT calculations (vide infra). There
is no obvious way to do this experimentally, and therefore the
correlation presented in Fig. 3 should be considered as an
experimental correlation established with the aid of computa-
tion. Small cluster models of the halogen bonding environ-
ment were used for the calculations, as shown in Fig. 1. The
calculated values of diso(13C) are presented in Table 3. In most
cases it can be seen that when the C–I distance increases, the
value of diso(13C) will also increase, perhaps since the spin–
orbit effect is reduced. It is also interesting to mention that,
consistent with this interpretation, p-DITFB has the most
shielded carbon (76.50(0.50) ppm) and shortest dC–I (blue
square in Fig. 3).
distances, a good exponential correlation is observed (R2
=
0.9613, see Fig. 3). An exponential is used to fit the data, since
in principle as the bond length becomes infinitely long, the CS
value observed would be constant (however, the physical
meaning of this value is unclear). A linear fit which excludes
the datum for the non-halogen bonded p-DITFB is reasonable
3172 | CrystEngComm, 2013, 15, 3168–3177
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
Table 3 Calculated and experimental 13C isotropic chemical shifts of carbon
described previously in the literature. Efforts by Bouchmella
and co-workers on the acquisition of 13C CP MAS NMR spectra
for halogen bonded compounds of imidazolyl-containing
haloalkenes and haloalkynes have been reported.20
Unfortunately, they were unable to observe the chemical shifts
of the carbons bonded to iodine experimentally. Our observa-
tions are consistent with previous reports of 13C chemical
shifts in solution. Goroff and coworkers demonstrated that
there is a measurable change in carbon resonances associated
with the carbons bonded to iodine due to halogen bonding
interactions with the solvent in diiodohexatriyne and diio-
dooctatetrayne.17 These changed from 0.9 ppm in CDCl3 to
14.6 ppm in DMSO for C6I2 and from 1.9 ppm in CDCl3 to 17.9
ppm in DMSO for C8I2. This phenomenon is well known for
iodoalkynes, were the carbon in iodophenylacetylene has a
chemical shift of 6.1 ppm in CDCl3 and 20 ppm in pyridine-
d5.18,19 However, Glaser, Kaupp and co-workers have elo-
quently described the challenges in understanding such
changes, particularly in solution, where spin–orbit effects
and coordination shifts may nearly cancel each other out,
making trends difficult to identify experimentally.34
covalently bonded to iodine
Compound
diso( (
13C) exp./ppma diso 13C) calc./ppmb
p-C6F4I2
[(n-Bu4NCl)(p-C6F4I2)]
76.50(0.50)
83.92(0.01)
80.64(0.02)
81.75(0.45)
83.65(0.15)
84.72(0.10)
81.84(0.01)
83.80(0.06)
83.00(0.02)
72.12
114.17
85.22
94.74
98.21
92.95
69.53
117.66
67.54
95.62
1
2
3
4
[(n-Bu4PCl)(p-C6F4I2)]
[(n-Bu4NBr)(p-C6F4I2)]
[(n-Bu4PBr)(p-C6F4I2)]
5
[(EtPh3PBr)2(p-C6F4I2)] 84.50(0.25)
a
Experimental 13C isotropic chemical shift of the carbon covalently
bonded to iodine. In parentheses are the errors assessed from
b
WSolids1 simulations. DFT calculated diso(
13C) including scalar and
spin–orbit relativistic effects using ZORA as implemented in ADF.
The revPBE exchange-correlation functional was used with the ZORA/
TZP basis set, plus extra diffuse functions on the halide ion (AUG/
ATZP).
To our knowledge, this is the first time this trend correlating
the values of diso(13C) to the C–I distance in halogen bonds has
been reported in the solid state. Indeed, difficulties in
observing such shifts and interpreting them have been
ZORA DFT calculations of 13C magnetic shieldings and
chemical shifts
The challenges associated with chemical shift calculations for
13C bonded to heavy atoms have been discussed in the
literature.33,34 ZORA–DFT computed 13C–I chemical shifts for
the compounds of interest are reported in Table 3. While the
calculations provide chemical shifts which are significantly
lower than those of aromatic carbons lacking a bonded iodine,
the computed values of diso(13C) generally deviate from the
experimental data by several ppm. Calculations of these data
are quite challenging for various reasons, and there are several
possible explanations for the deviation between experiment
and theory. For example, the cluster models used here do not
allow for a full treatment of the effect of the ions in the crystal
lattice on the computed NMR parameters. The gauge-includ-
ing projector-augmented wave (GIPAW) DFT method35 does
allow for calculations using periodic boundary conditions;
however, we have found that the inclusion of relativistic effects
via ZORA is more important for the present compounds (see
Table S11, ESI ). It would be desirable to be able to include
3
periodic boundary conditions as well as relativistic effects;
while this is an ongoing area of research,36 it is beyond our
current computational capabilities. Furthermore, in general it
is known that the treatment of relativistic effects as well as the
relative roles of dispersion and electrostatics in the XB are
challenging for current computational methods.37 On balance,
we believe that our approach is a reasonable one given the
various competing factors involved and the computational
tools available.
Given the relatively poor absolute agreement between the
experimental and calculated shifts shown in Table 3, we
sought to confirm the experimentally observed trend (Fig. 3) by
attempting to reproduce the trend computationally, without
focus on the precise magnitudes of the computed shifts.
Additional calculations of 13C chemical shifts were therefore
performed on cluster models involving a molecule of p-DITFB
halogen-bonded through one iodine to a chloride ion. Two
Fig. 3 Plot of experimental values of diso
carbon–iodine distance, dC–I, for carbons directly bonded to iodine. The values
of diso
13C) are taken from Table 3 and the values of dC–I are tabulated in
(
13C) as a function of the corresponding
(
Table 2. The black diamonds represent the halogen bonded compounds 1, 2, 3,
and 5 and the blue square represents p-DITFB. The red diamonds represent
compound 4, where the assignment is ambiguous; they could be very well be
assigned in reverse order from the ZORA DFT cluster model calculations. Some
of the experimental vertical error bars are within the size of the symbols. The
best fit represented by a black line is an exponential function: diso(
13C) = 8.5629
6 (1 2 exp(2128.96Dd)) + 76.5, Pearson correlation coefficient R2 = 0.9613,
where Dd = dC–I(XB compound) 2 dC–I(p-DITFB). A linear fit to the data for the halogen
bonded complexes (blue square excluded from fit) is described by: diso
220.15d 2 379.71, R2 = 0.859.
(
13C) =
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 3168–3177 | 3173

View Article Online
Paper
CrystEngComm
Fig. 4 Calculated 13C isotropic chemical shift of the carbon covalently bonded to iodine in a cluster model involving p-DITFB and chloride, using ZORA DFT
implemented in ADF (GGA PBE and ZORA/TZP basis set). Plot (a and c) corresponds to diso(
13C) vs decreasing halogen bond strength (from RXB = 0.79 to 0.98), where
…
the value of hC–I Cl is fixed to 180u. The shaded areas represent the range of observed RXB values for the halogen bonded compounds reported herein (RXB = 0.79 to
0.82). In (a), scalar and spin–orbit relativistic effects are calculated using ZORA and in (c) no relativistic effects are included. Plots (b and d) correspond to diso(
13C) as a
…
function of increasing distance between the carbon and covalently bond iodine (dC–I), where the halogen bond strength is held constant at 0.81 and hC–I Cl is linear.
Scalar and spin–orbit relativistic effects are implemented using ZORA in (b) and not in (d). Solid lines are of best linear fit, except in the case of (c) where a second-
13C) = 277.961RXB + 168.43, (b) diso
(
13C) = 60.75(dC–I) 2 22.814, (c) diso
(
13C) = 64.21RXB
2
2
order polynomial is used. The lines of best fit are as follows: (a) diso
(
148.25RXB + 207.56 and the linear fit would be diso
(
13C) = 233.96RXB + 156.92, Pearson correlation coefficient R2 = 0.9902; (d) diso
(
13C) = 76.75(dC–I) 2 31.585.
different systematic studies were performed. First, the short
contact between the chloride and iodine was increased
systematically from RXB = 0.79 to 0.98, while the C–I distance
…
was performed where the C–I distance was increased from
2.094 to 2.110 Å (representing the range of C–I bond lengths
observed by X-ray diffraction in our series of XB compounds)
by increments of 0.002 Å and the normalized distance
Cl2
and the halogen bonding angle, hC–I
were kept fixed at
2.096 Å and 180u, respectively. Shown in Fig. 4 is an inverse
linear correlation between the increasing strength of the XB
interaction and the calculated values of diso(13C) of the carbon
covalently bonded to iodine. In Fig. 4(a), scalar and spin–orbit
relativistic effects are included in the calculations using ZORA,
since spin–orbit-induced effects are expected in the presence
of iodine.33 In Fig. 4(c), the same plot is presented without the
incorporation of relativistic effects. There, a second-order
polynomial is used to best fit the data (R2 = 1). The general
trend is the same in both cases: an increase in XB strength is
accompanied by an increase in chemical shift. The shaded
areas in the plots are representative of the strength of the XB
in the compounds presented in this study (RXB = 0.79 to 0.82).
A 4 ppm change is observed over this range when relativistic
effects are incorporated in the calculations, and a 2 ppm
change is observed for the non-relativistic calculations. Both
sets of data are consistent with experiment, where a 4 ppm
range is observed for diso(13C). Recall, however, that a clear
correlation between diso(13C) and RXB was not observed
experimentally. Therefore, a second series of calculations
parameter, RXB, was fixed at 0.81 with a linear C–I Cl2 angle.
…
Linear correlations are observed between the values of diso(13C)
and the C–I bond length (Fig. 4 with (b) and without (d) scalar
and spin–orbit relativistic effects). In both cases, there is a y1
ppm increase in diso(13C) when increasing the C–I bond length,
and the same general trends are observed whether relativistic
effects are included or not. However, the inclusion of scalar
and spin–orbit relativistic effects in the calculations results in
a decrease in the value of diso(13C) by 25 ppm relative to the
non-relativistic results (y105 ppm vs. y130 ppm). Relativistic
ZORA–DFT calculations reproduce better the observed magni-
tudes of the chemical shifts of carbon covalently bonded to
iodine.
ZORA–DFT calculations and experiments both suggest that
an increase in the C–I bond length correlates with an increase
in the 13C chemical shift for C–I groups involved in halogen
bonds. While experimentally it was observed that increases in
the C–I bond length correlate best in an inverse exponential
manner with diso(13C), a linear trend is observed using ZORA–
DFT. It is reasonable to suggest that this might be due to the
3174 | CrystEngComm, 2013, 15, 3168–3177
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
constraints of the model. For example, the XB strength is fixed
to RXB = 0.81 in the calculations whereas experimentally, when
the halogen bond strength increases, the C–I bond length
simultaneously increases. This might also help to explain the 1
ppm change observed computationally versus the 4 ppm
change observed experimentally for diso(13C) over the same
range of C–I bond lengths. However, even though the
calculated chemical shifts are overestimated, the overall trend
is the one observed experimentally: an increase in the C–I
bond length will result in a higher chemical shift of the carbon
resonance involved in the XB interaction.
Experimental
Co-crystallization
All ammonium and phosphonium salts (n-Bu4PCl, n-Bu4NCl,
n-Bu4PBr, n-Bu4NBr, and EtPh3PBr) were purchased from
Aldrich and used without further purification. Compounds 1
and
3 were prepared according to previously reported
literature22 but using a different solvent, dichloromethane.
Compounds 2, 4, and 5, were all prepared by dissolving
equimolar amounts of the appropriate ammonium or phos-
phonium halide salt with p-DITFB, in a minimum amount of
dichloromethane. Vials containing the solutions were then
placed into a closed jar containing mineral oil and left to
evaporate until the desired compound crystallized.
Conclusions
X-ray crystallography
We have reported the preparation of a series of compounds
exhibiting halogen bonding between p-diiodotetrafluoroben-
zene and halide anions, characterized by close linear contacts,
which have been studied by X-ray diffraction and 13C solid-
state magnetic resonance spectroscopy. X-ray diffraction
revealed similarities and differences in the halogen bonding
motifs for halides described by Abates et al.,22 Triguero et al.,27
Data collection results for compounds 1, 2, 3, 4, and 5
represent the best data sets obtained in several trials for each
sample. The crystals were mounted on thin glass fibers using
paraffin oil. Mounted crystals were cooled by a stream of dry
air to 200.15 K prior to data collection. Data were collected on
a Bruker AXS KAPPA single crystal diffractometer equipped
with a sealed Mo tube source (wavelength 0.71073 Å) and APEX
II CCD detector. Raw data collection and processing were
performed with the APEX II software package from BRUKER
AXS.38 Diffraction data for 1, 2, 4, and 5 were collected with a
sequence of 0.5u v scans at 0, 120, and 240u in Q. Diffraction
data for 3 were collected with a sequence of 0.5u v scans at 0,
90, 180, and 270u in Q due to the lower quality of the crystal
and weaker diffraction intensity to provide adequate data
coverage. Initial unit cell parameters were determined from 60
data frames collected for different sections of the Ewald
sphere. Semi-empirical absorption corrections based on
equivalent reflections were applied.39 Systematic absences in
the diffraction data set and unit cell parameters were
consistent with the monoclinic C2/c (no. 15) space group for
compound 4, monoclinic P21/n (no. 14) for compound 5,
monoclinic P21/c (no. 14) for compounds 1 and 2, and
orthorhombic Pccn (no. 56) for compound 3. Solutions in the
centrosymmetric space groups for all compounds yielded
chemically reasonable and computationally stable results of
refinement. The structures were solved by direct methods,
completed with difference Fourier synthesis, and refined with
and Grebe et al.28 In particular, the halides are ditopic and
2
…
X
…
form long polymeric chains with I
I angles ranging from
109.1 to 140.9u for compounds 1 to 4. The bromide anion in
compound 5 forms a halogen bond with one iodine, forming
22
…
…
[Br I–C6F4–I Br] moieties.
For compounds 1 to 5, the solid-state 13C chemical shifts
increase by several ppm relative to the pure non-halogen-
bonded p-DITFB component. Recalling the provisional IUPAC
…
definition of the XB interaction (R–X D), the halogen bond
may lead to characteristic changes in the NMR signals of R–X.
In this study, we were able to establish a correlation between
the value of diso(13C) and the halogen bonding environment, as
quantified by the carbon–iodine bond length. As the C–I
distance increases, the chemical shift also increases.
Correlations with the value of the normalized distance
parameter, RXB, were not observed experimentally; this may
be in part due to the fact that both chloride and bromide
electron donors are considered within the same data set.
ZORA–DFT calculations of the 13C chemical shift carried out
on a cluster model reproduced the experimental trend as the
carbon–iodine distance was varied. The inclusion of scalar and
spin–orbit relativistic effects in the calculations resulted in a
full-matrix least-squares procedures based on F2. See ESI for
further information.
All hydrogen atoms were treated as idealized contributions.
All scattering factors are contained in several versions of the
SHELXTL program library, with the latest version used being
v.6.12.40 Crystallographic data and selected data collection
parameters are reported in Table 1.
3
decrease in the computed values of diso(
13C) of about 25 ppm.
It is clear that there is further room for improvement in the
calculation of magnetic shielding tensors in halogen-bonded
systems where relativistic effects are important. This com-
bined experimental–theoretical study has demonstrated that
carbon–iodine–halide halogen bonding interactions are
reflected in the 13C chemical shifts, thereby providing an
example of the utility of NMR methods in characterizing
halogen bonding interactions in the solid state.
13C solid-state NMR spectroscopy
All compounds were ground into fine powders and packed in 4
mm o.d. zirconia rotors. Data were acquired at the University
of Ottawa using a 9.4 T (v0(13C) = 100.6 MHz) wide-bore magnet
equipped with a Bruker Avance III 400 console and a 4 mm
Bruker triple resonance MAS probe. Data were also acquired at
the National Ultrahigh-Field NMR Facility for Solids in Ottawa
using a 21.1 T (v0(13C) = 226.3 MHz) standard-bore magnet
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 3168–3177 | 3175

View Article Online
Paper
CrystEngComm
equipped with a Bruker AVANCE II 900 console and a double-
resonance MAS probe.
exchange-correlation functional was used with the ZORA/TZP
basis set. Some calculations were done with additional diffuse
functions on the halide ions. For compounds 3 and 4, disorder
in the aromatic rings was eliminated by using an average
structure or optimized C and F coordinates. The calculations
were performed using the High Performance Computing
Virtual Laboratory (HPCVL). Magnetic shielding tensors were
converted to chemical shift tensors using the following
9.4 T data. 13C solid-state NMR spectra were collected using
cross-polarization and magic angle spinning (CP MAS).
Experimental setup and pulse calibration were performed
using the recommended IUPAC standard for 13C (I = K), solid
glycine (d(13CLO) = 176.4 ppm with respect to TMS).41 For
compounds 1, 2, 5 and 3, 4 the recycle delays and contact
times were optimized to 13 s and 5 s, and 3 ms and 5 ms,
respectively. The proton p/2 pulse was set to 3.50 ms in all
cases, and the MAS frequencies were 8 kHz with the exception
of 5 where 7.5 kHz was used. The number of scans varied for
each compound, from a few hundred to 4 k, and the
experiments were run until a reasonable signal-to-noise ratio
was obtained. Total experimental times ranged from 64 min to
15 h. In the case of p-DITFB, a rotor-synchronized Hahn-echo
(i.e. p/2-t1-p-t2-acq)42 MAS experiment was used. Pulse
calibration was performed on the secondary reference ada-
mantane (d(13C) = 38.55 and 29.50 ppm relative to TMS). The
spinning speed was set to 8 kHz, while the recycle delay and
sref {sij
equation: dij~
(where sref = 184.1 ppm for carbon,48 dij
1{sref
is the chemical shift tensor component, and sij is the magnetic
shielding tensor component).
Acknowledgements
´ ´
We thank Kevin Burgess and Frederic Perras for helpful
comments. D. L. B. thanks the Natural Sciences and
Engineering Research Council (NSERC) of Canada for funding.
Dr. Victor Terskikh, Dr. Eric Ye, and Dr. Glenn Facey are
thanked for technical support. Access to the 900 MHz NMR
spectrometer was provided by the National Ultrahigh-Field
NMR Facility for Solids (Ottawa, Canada), a national research
facility funded by the Canada Foundation for Innovation, the
1
p/2 pulse length were 20 s and 2.75 ms, respectively. TPPM H
decoupling was used.43 Line broadening was also used in the
processing of the data and varied from 25 to 100 Hz.
21.1 T data. The same method was used at this field for the
acquisition of the 13C SSNMR spectra as was done at 9.4 T. The
MAS frequency was set to 18 kHz. The following parameters
were used for the CP MAS experiments: recycle delay of 10 s,
contact time of 3 ms and p/2 pulse length of 2.50 ms. The
number of scans varied for each compound, ranging from 256
to 1024. As for the Hahn-echo MAS experiment (p-DITFB), a
total of 4 k scans were collected using a recycle delay equal to
10 s and a p/2 pulse length of 5 ms.
´
Ontario Innovation Trust, Recherche Quebec, the National
Research Council Canada, and Bruker BioSpin and is managed
by the University of Ottawa (www.nmr900.ca). NSERC is
acknowledged for a Major Resources Support grant.
References
All NMR data were processed with Bruker TopSpin 3.0
software. The FIDs acquired using a Hahn-echo experiment for
p-DITFB were left shifted an appropriate number of data
points to the top of the echo. The number of distinct diso(13C)
values for carbon atoms directly bonded to iodine atoms was
determined independently from the X-ray crystal structures
(one value for each magnetically distinct carbon site).
Simulations of the MAS NMR spectra were carried out using
WSolids144 through manual iteration using the appropriate
number of diso(13C) values, and apodization. The precision in
the reported diso(13C) values was assessed through simulta-
neous iterative fitting of the 13C MAS NMR spectra obtained at
both 21.1 and 9.4 T. Stack plots of the experimental and
simulated spectra were prepared using DMFIT (v.2011).45
1 E. Parisini, P. Metrangolo, T. Pilati, G. Resnati and
G. Terraneo, Chem. Soc. Rev., 2011, 40, 2267.
2 Z. P. Shields, J. S. Murray and P. Politzer, Int. J. Quantum
Chem., 2010, 110, 2823.
3 (a) F. Guthrie, J. Chem. Soc., 1863, 16, 239; (b) O. Hassel,
Science, 1970, 170, 497.
4 (a) P. Auffinger, F. A. Hays, E. Westhof and P. S. Ho, Proc.
Natl. Acad. Sci. U. S. A., 2004, 101, 16789; (b) A. R. Voth and
P. S. Ho, Curr. Top. Med. Chem., 2007, 7, 1336; (c) A.
R. Voth, F. A. Hays and P. S. Ho, Proc. Natl. Acad. Sci. U. S.
A., 2007, 104, 6188; (d) Y. Lu, T. Shi, Y. Wang, H. Yang,
X. Yan, X. Luo, H. Jiang and W. Zhu, J. Med. Chem., 2009,
52, 2854.
5 D. L. Bryce, Chem. Int., 2012, 34, 27.
6 G. R. Desiraju, P. S. Ho, L. Kloo, A. C. Legon, R. Marquardt,
P. Metrangolo, P. A. Politzer, G. Resnati and K. Rissanen,
Definition of the Halogen Bond, IUPAC Technical Reports and
Recommendations, 2012, task group 2009-032-1-100.
7 J. Kemsley, Chem. Eng. News, 2012, 90(35), 36.
8 T. Clark, M. Hennemann, J. S. Murray and P. Politzer, J.
Mol. Model., 2007, 13, 291.
9 (a) A. E. Aliev and K. D. M. Harris, Supramolecular Assembly
via Hydrogen Bonds, Structure and Bonding, 2004, 108, 1–53;
(b) S. Grzesiek, F. Cordier, V. Jaravine and M. Barfield, Prog.
Nucl. Magn. Reson. Spectrosc., 2004, 45, 275; (c) G. Wu, C.
J. Freure and E. Verdurand, J. Am. Chem. Soc., 1998, 120,
13187.
Computational details
Cluster models centered on the halide ion involved in halogen
bonding were generated using the X-ray crystal structures’
atomic coordinates (models used are depicted in Fig. 1). The
structure of Chaplot et al. was used for p-DITFB.46 The 13C
magnetic shielding tensors were calculated with the
Amsterdam Density Functional (ADF) software (version
2009.01).47 Two sets of calculations were carried out: one set
including scalar and spin–orbit relativistic effects using the
zeroth-order regular approximation (ZORA), and the other set
omitting relativistic effects. The Perdew, Burke, and Ernzerhof
(PBE) (or revPBE) generalized gradient approximation (GGA)
3176 | CrystEngComm, 2013, 15, 3168–3177
This journal is ß The Royal Society of Chemistry 2013

View Article Online
CrystEngComm
Paper
´
´
10 J. F. Bertran and M. Rodrıguez, Org. Magn. Reson., 1980, 14,
31 I. Dance and M. Scudder, J. Chem. Soc., Chem. Commun.,
1995, 1039.
244.
´
´
11 J. F. Bertran and M. Rodrıguez, Org. Magn. Reson., 1979, 12,
92.
32 V. V. Terskikh, S. J. Lang, P. G. Gordon, G. D. Enright and J.
A. Ripmeester, Magn. Reson. Chem., 2009, 47, 398.
12 M. T. Messina, P. Metrangolo, W. Panzeri, E. Ragg and
G. Resnati, Tetrahedron Lett., 1998, 39, 9069.
13 P. Metrangolo, W. Panzeri, F. Recupero and G. Resnati, J.
Fluorine Chem., 2002, 114, 27.
¨
33 M. Kaupp, O. L. Malkina, V. G. Malkin and P. Pyykko,
Chem.–Eur. J., 1998, 4, 118.
34 R. Glaser, N. Chen, H. Wu, N. Knotts and M. Kaupp, J. Am.
Chem. Soc., 2004, 126, 4412.
14 M. G. Sarwar, B. Dragisic, L. J. Salsberg, C. Gouliaras and
M. S. Taylor, J. Am. Chem. Soc., 2010, 132, 1646.
15 J. Viger-Gravel, I. Korobkov and D. L. Bryce, Cryst. Growth
Des., 2011, 11, 4984.
16 R. J. Attrell, C. M. Widdifield, I. Korobkov and D. L. Bryce,
Cryst. Growth Des., 2012, 12, 1641.
17 K. Gao and N. S. Goroff, J. Am. Chem. Soc., 2000, 122, 9320.
18 P. D. Rege, O. L. Malkina and N. S. Goroff, J. Am. Chem.
Soc., 2002, 124, 370.
19 J. A. Webb, J. E. Klijn, P. A. Hill, J. L. Bennett and N.
S. Goroff, J. Org. Chem., 2004, 69, 660.
20 K. Bouchmella, S. G. Dutremez, B. Alonso, F. Mauri and
C. Gervais, Cryst. Growth Des., 2008, 8, 3941.
21 R. K. Harris and A. C. Olivieri, Prog. Nucl. Magn. Reson.
Spectrosc., 1992, 24, 435.
22 A. Abate, S. Biella, G. Cavallo, F. Meyer, H. Neukirch,
P. Metrangolo, T. Pilati, G. Resnati and G. Terraneo, J.
Fluorine Chem., 2009, 130, 1171.
23 (a) J. Autschbach and T. Ziegler, J. Chem. Phys., 2000, 113,
936; (b) J. Autschbach and T. Ziegler, J. Chem. Phys., 2000,
113, 9410; (c) R. M. Dickson and T. Ziegler, J. Phys. Chem.,
1996, 100, 5286; (d) J. Khandogin and T. Ziegler, Spectrochim.
Acta, Part A, 1999, 55, 607.
35 (a) C. J. Pickard and F. Mauri, Phys. Rev. B: Condens. Matter,
2001, 63, 245101; (b) J. R. Yates, C. J. Pickard and F. Mauri,
Phys. Rev. B: Condens. Matter Mater. Phys., 2007, 76, 024401;
(c) M. Profeta, F. Mauri and C. J. Pickard, J. Am. Chem. Soc.,
2003, 125, 541; (d) S. J. Clark, M. D. Segall, C. J. Pickard, P.
J. Hasnip, M. I. J. Probert, K. Refson and M. C. Payne, Z.
Kristallogr., 2005, 220, 567.
36 D. Skachkov, M. Krykunov and T. Ziegler, Can. J. Chem.,
2011, 89, 1150.
37 (a) K. E. Riley and P. Hobza, Cryst. Growth Des., 2011, 11,
4272; (b) K. E. Riley and P. Hobza, J. Chem. Theory Comput.,
2008, 4, 232.
38 Bruker AXS, Bruker Suite, version 2008/3, Bruker AXS Inc.,
Madison, WI, USA, 2008.
39 G. M. Sheldrick, SADABS – Bruker AXS area detector scaling
¨
and absorption, version 2008/1, University of Gottingen,
Germany, 2008.
40 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found.
Crystallogr., 2008, 64, 112.
41 W. L. Earl and D. L. VanderHart, J. Magn. Reson., 1982, 48,
35.
42 S. R. Hartmann and E. L. Hahn, Phys. Rev., 1962, 128, 2042.
43 A. E. Bennett, C. M. Rienstra, M. Auger, K. V. Lakshmi and
R. G. Griffin, J. Chem. Phys., 1995, 103, 6951.
44 K. Eichele and R. E. Wasylishen, WSOLIDS NMR Simulation
Package, Version 1.20.15, Dalhousie University, Halifax,
2001.
24 A. Bondi, J. Phys. Chem., 1964, 68, 441.
25 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
Theor. Gen. Crystallogr., 1976, 32, 751.
26 J. P. M. Lommerse, A. J. Stone, R. Taylor and F. H. Allen, J.
Am. Chem. Soc., 1996, 118, 3108.
´
45 D. Massiot, F. Fayon, M. Capron, I. King, S. Le Calve,
´
27 S. Triguero, R. Llusar, V. Polo and M. Fourmigue, Cryst.
B. Alonso, J.-O. Durand, B. Bujoli, Z. Gan and G. Hoatson,
Magn. Reson. Chem., 2002, 40, 70.
46 S. L. Chaplot, G. J. McIntyre, A. Mierzejewski and G.
S. Pawley, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst.
Chem., 1981, 37, 2210.
Growth Des., 2008, 8, 2241.
28 J. Grebe, G. Geiseler, K. Harms and K. Dehnicke, Z.
Naturforsch., 1999, 54b, 77.
29 (a) M. G. Chudzinski, C. A. McClary and M. S. Taylor, J. Am. Chem.
´
Soc., 2011, 133, 10559; (b) E. Dimitrijevic, O. Kvak and M.
47 Amsterdam Density Functional Software ADF2009.01 SCM,
Theoretical Chemistry, Vrije Universiteit, Amsterdam, The
Netherlands, http://www.scm.com/, 2010.
48 A. K. Jameson and C. J. Jameson, Chem. Phys. Lett., 1987,
134, 461.
S. Taylor, Chem. Commun., 2010, 46, 9025; (c) M. G. Sarwar,
B. Dragisic, S. Sagoo and M. S. Taylor, Angew. Chem., Int. Ed.,
2010, 49, 1674; (d) P. Metrangolo, T. Pilati, G. Terraneo, S. Biella
and G. Resnati, CrystEngComm, 2009, 11, 1187.
30 I. Dance and M. Scudder, Chem.–Eur. J., 1996, 2, 481.
This journal is ß The Royal Society of Chemistry 2013
CrystEngComm, 2013, 15, 3168–3177 | 3177