Tris-o-phenylenedioxycyclotriphosphazene (TPP) inclusion compounds containing a dipolar molecular rotor

Article abstract of DOI:10.1021/cg4013608

A rod-shaped molecular rotor consisting of a p-terphenyl shaft attached to p-carborane whose antipodal position carries a dipolar 2,3-dichlorophenyl rotator forms an inclusion compound with hexagonal tris-o- phenylenedioxycyclotriphosphazene (TPP). Results of solid-state NMR spectroscopy, X-ray powder diffraction, dielectric loss spectroscopy, and density functional theory calculations lead us to propose that the whole molecule inserts into the TPP channels, with the rotator located in the outermost surface layer. Although the placement and alignment of the dipoles at the surface appear favorable, the sample does not exhibit collective behavior even at 7 K, presumably due to the relatively large barrier to rotation (～8.6 kcal/mol). In incompletely annealed samples of the inclusion compound, some of the rotators protrude outside the surface and have a rotational barrier of ～3.4 kcal/mol. In the inclusion compound of an analog in which the rotator is replaced with a methyl group, some of the methyl substituents are located inside the surface layer of TPP and others protrude above it.

Full text of DOI:10.1021/cg4013608

Article
pubs.acs.org/crystal
Tris‑o‑phenylenedioxycyclotriphosphazene (TPP) Inclusion
Compounds Containing a Dipolar Molecular Rotor
Lukas Kobr,† Ke Zhao,‡ Yongqiang Shen,‡ Richard K. Shoemaker,† Charles T. Rogers,‡
́
̌
and Josef Michl*,†,§
†Department of Chemistry and Biochemistry and ‡Department of Physics, University of Colorado, Boulder, CO 80309, United States
§Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic, 16610 Prague 6, Czech Republic
S
* Supporting Information
ABSTRACT: A rod-shaped molecular rotor consisting of a p-
terphenyl shaft attached to p-carborane whose antipodal
position carries a dipolar 2,3-dichlorophenyl rotator forms an
inclusion compound with hexagonal tris-o-phenylenedioxycyclo-
triphosphazene (TPP). Results of solid-state NMR spectrosco-
py, X-ray powder diffraction, dielectric loss spectroscopy, and
density functional theory calculations lead us to propose that
the whole molecule inserts into the TPP channels, with the
rotator located in the outermost surface layer. Although the placement and alignment of the dipoles at the surface appear
favorable, the sample does not exhibit collective behavior even at 7 K, presumably due to the relatively large barrier to rotation
(∼8.6 kcal/mol). In incompletely annealed samples of the inclusion compound, some of the rotators protrude outside the surface
and have a rotational barrier of ∼3.4 kcal/mol. In the inclusion compound of an analog in which the rotator is replaced with a
methyl group, some of the methyl substituents are located inside the surface layer of TPP and others protrude above it.
and an axle that carries a dipolar rotator and is supposed to
orient itself normal to the surface on its outside. Presently, we
work with the nonpolar rotor 1 and polar rotor 2 (Scheme 1),
where the shaft is p-terphenyl, the stopper is p-carborane, and
the rotator is either the nonpolar methyl (1) or the dipolar 2,3-
dichlorophenyl (2; the dipole moment of o-dichlorobenzene is
2.51 D24).
TPP offers many advantages, in particular a small
interchannel distance, but also has two disadvantages: (i) in
addition to existing in the desirable hexagonal form, it is capable
of existing in a nearly equally stable monoclinic form that has
no channels, accepts no guest molecules, and whose formation
needs to be carefully avoided and (ii) being a molecular rather
than a covalent crystal, the hexagonal form has the ability to
adjust the size of its channels to accommodate even guest
molecules that are larger than the empty channels, complicating
the choice of the stopper part of the molecular rotor.
The principal tool that we use to characterize the rotation of
dipolar rotors is dielectric spectroscopy, the measurement of
capacitor dielectric loss peaks as a function of temperature.
These loss peaks show dispersion with measurement frequency,
indicative of a situation where the dipole moments move in a
potential with activation barriers. For temperatures sufficiently
high, thermal motion over the barriers allows the dipole
moment to align itself with applied electric field essentially
instantaneously and these dipoles then do not contribute to the
INTRODUCTION
■
Artificial molecular rotors1−7 have received considerable
attention. Collections of dipolar rotors have been examined
in three-dimensional (3D) crystals8−10 and on two-dimensional
(2D) surfaces as individual rotors11−14 or as random
assemblies15−18 but not in regular 2D arrays, which are of
particular interest to us. Such arrays should be capable of
exhibiting collective behavior, and ferroelectricity19 would be
particularly desirable. A ferroelectric ground state requires a
suitable regular array of rotors with low barrier to rotation (low
Debye temperature) and strong ferroelectric inter-rotor
interaction (high Curie temperature).
One possible approach to regular arrays of molecular rotors
is based on the separation of the design of the rotor molecules
themselves (high dipole, low rotational barrier) from the design
of the lattice that defines their locations (small lattice constant,
triangular array). This has led us to examine surface inclusion
compounds of tris-o-phenylenedioxycyclotriphosphazene
(TPP).20−22 The hexagonal phase of this layered material23
(Figure 1) tends to form platelets whose large faces are parallel
to the ∼0.5 nm thick layers and perpendicular to a set of
roughly hexagonal channels that run through the crystal. The
channels have a triangular cross-section within each layer, with
the triangles rotated by 60° between neighboring layers. They
are arranged in a triangular pattern, about 1.1 nm apart. Their
internal diameter is ∼0.5 nm, and they can easily accommodate
alkane chains or polyphenylenes.
The molecular rotors that are to be inserted into the
channels consist of a shaft with affinity for channel interior, a
stopper that is meant to be too large to fit inside the channel,
Received: September 11, 2013
Revised: December 1, 2013
Published: January 24, 2014
© 2014 American Chemical Society
559
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
Figure 1. Crystal structure of hexagonal TPP. (a) Chemical structure and (b) space-filling model of a TPP molecule; two successive layers of a
crystal viewed from the (c) top and (d) side; two anticipated modes of (e) rotor insertion.
dielectric loss. For low enough temperatures, the dipoles are
frozen into wells and cannot contribute to the dielectric loss
either. At intermediate temperatures, where the measurement
frequency is equal to the typical thermally activated rate, dipole
motion phase lags behind the applied field and leads to energy
dissipation. For a simple thermal activation model of dielectric
loss in a system with a density n of dipole moments of
magnitude p in a double well with a barrier height EB and an
asymmetry s, the loss tangent is expected to be given by25−27
measurement frequency allows us to extract estimates of
average rotational barrier height Ea and prefactor τ0. Detailed
fitting of the peak shape versus temperature and frequency
allows us to extract the barrier asymmetry s and to estimate the
distribution of activation energies.
RESULTS
■
Synthesis. 1-Methyl-12-(4-terphenylyl)-p-carborane (1)
was prepared in 79% yield by deprotonation and methylation
of the known22 1-(4-terphenylyl)-p-carborane (3). Deprotona-
tion of 3 followed by transmetalation with cuprous chloride and
Ulmann coupling with 1,2-dichloro-3-iodobenzene afforded 1-
(2,3-dichlorophenyl)-12-(4-terphenylyl)-p-carborane 2 in 96%
yield. The preparation of 2-d13 proceeded similarly starting
from 4-iodo-p-terphenyl-d13 (4), which was obtained by
iodination of p-terphenyl-d14.
−2
tan δ = (4p 2n/9ε0kBT)cosh (s/2kBT)[ωτ/(1 + ω2τ2)]
0
In this expression, kB is the Boltzmann constant and τ is the
inverse of the attempt frequency,
τ = τ0 exp(EB/kBT)
These equations predict a peak in dielectric losses whenever
the choice of measurement frequency ω and temperature T is
such that ωτ = 1. Thus, a plot of the inverse of observed
temperature where losses peak versus the logarithm of the
Inclusion Compounds. A mixture of the molecular rotor
28
and empty hexagonal TPP-d12 was subjected to alternating
ball milling and annealing to 70−75 °C, yielding the desired
560
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
Scheme 1. Synthesis of 1 and 2
inclusion compounds, which we label with symbols such as
15%1@TPP-d12 (a 15% molar mixture of 1 in fully deuteriated
TPP). The inclusion compound het-15%2@TPP-d12 was
prepared by ball milling empty hexagonal TPP-d12 and 2
(17:3 molar ratio) without subsequent annealing.
Figure 2. X-ray powder diffraction patterns (X-ray counts vs
magnitude of the scatter wave vector q, black) and fitting to a
hexagonal structure (red) for (a) 15%1@TPP-d12 and (b) 15%2@
TPP-d12.
Differential Scanning Calorimetry (DSC). The DSC
traces of 1 and 15%1@TPP-d12 are shown in Figure S1 of the
Supporting Information. Neat 1 (Figure S1a of the Supporting
Information) shows a single endotherm at 262 °C, whereas
15%1@TPP-d12 shows two endotherms at 233 and 275 °C and
a sharp exotherm at 235 °C (Figure S1b of the Supporting
Information). The DSC trace of hexagonal TPP-d12 (Figure S2a
of the Supporting Information) closely resembles the
reported29 DSC curve of hexagonal TPP and consists of a
broad exotherm centered around 148 °C (first order transition
of hexagonal to monoclinic TPP) and endotherms at 222 °C
(first order transition of monoclinic to hexagonal TPP) and 252
°C (melting). The neat rotor 2 (Figure S2b of the Supporting
Information) shows a single endotherm at 234 °C, and 15%2@
TPP-d12 (Figure S2c of the Supporting Information) shows a
broad exotherm around 200 °C and an endotherm at 286 °C.
Other than a small endotherm associated with melting TPP-d12,
the endotherms in the DSC traces of the inclusion compounds
are due to their melting. The trace of 15%2@TPP-d12 also
shows a very small endotherm at 236 °C that is associated with
the first order transition of monoclinic TPP-d12 into its
hexagonal form.
X-ray Powder Diffraction. Figure 2 shows X-ray powder
patterns taken with Cu Kα1 radiation for 15%1@TPP-d12 and
15%2@TPP-d12. The data sets are fitted to a hexagonal model
that includes a scattering background from the capillary sample
holder and line width broadening due to the finite powder
particle size. Both sets are consistent with an overall hexagonal
structure. Mixtures containing less than the 15% molar
concentration of the guest generally display powder patterns
that show a complicated superposition of expanded hexagonal,
empty hexagonal, and monoclinic structure of TPP.
nearly identical expanded hexagonal structure with an in-plane
lattice parameter of 1.2122
0.0004 nm, bilayer spacing of
1.0044 0.0004 nm, and particle size of 20 0.3 nm (for a
homogeneous 10%2@TPP-d12 sample, which contained a
significant amount of monoclinic TPP, the in-plane lattice
parameter in the hexagonal phase was 1.2169 0.0004 nm).
Solid-State NMR (ssNMR). The 31P cross-polarization
magic angle spinning (CP MAS) ssNMR spectra (1: Figure S3a
of the Supporting Information, 2: Figure S3c of the Supporting
Information) and the 31P single-pulse excitation magic angle
spinning (SPE MAS) ssNMR spectra (1: Figure S3b of the
Supporting Information, 2: Figure S3d of the Supporting
Information) of 15%1@TPP-d12 and 15%2@TPP-d12 all
contain a single resonance at 34.0 ppm. This is compatible
with the pure hexagonal structure found by XRD as monoclinic
TPP exhibits three distinct resonances in its 31P ssNMR
spectrum at 32.0, 35.1, and 38.3 ppm.30
The solid-state 13C CP MAS NMR of neat 1 (5 ms contact
time, Figure 3a) shows five resolved quaternary aromatic
carbon peaks between 133 and 140 ppm, unresolved
protonated aromatic carbon signals between 123 and 130
ppm, a peak for each carborane carbon (76.2 and 79.5 ppm),
and a peak of the methyl carbon (26.4 ppm). The solution 13
C
NMR spectrum of 1 in CDCl3 (Figure 3b) shows all the signals
resolved. The CP MAS ssNMR spectrum of 15%1@TPP-d12
obtained with 5 ms contact time, dipolar dephasing (DPD),
and total sideband suppression (TOSS) (Figure 3c) greatly
attenuates protonated carbons, making quaternary carbon
signals stand out, while the 13C CP MAS ssNMR spectrum
of 15%1@TPP-d12 obtained with a 5 ms contact time (Figure
3d) shows both protonated and quaternary carbons.
The five quaternary aromatic carbons in 1 give rise to three
peaks in the 13C CP MAS ssNMR spectra of 15%1@TPP-d12
(C-1, C-2, and C-3, Figure 3, panels c and d). The carborane
carbons give rise to two peaks (C-5 and C-6) in all spectra. All
the aromatic and carborane signals in the 13C ssNMR spectra of
15%1@TPP-d12 show a strong upfield shift of about 1.0 to 1.9
561
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
Figure 3. NMR spectra of 1: (a) 13C CP MAS, neat solid; (b) 13C
NMR in CDCl3 solution; (c) 13C CP MAS solid 15%1@TPP-d12 with
DPD and TOSS; and (d) 13C CP MAS solid 15%1@TPP-d12.
ppm relative to the solution spectrum (Table 1). The
protonated aromatic carbons of 15%1@TPP-d12 also exhibit a
strong upfield shift, and they overlap with the carbon C−B of
TPP-d12. The methyl signal in the spectrum of 15%1@TPP-d12
exhibits two separate broad peaks. One of them is shifted
upfield by 3.2 ppm and the other one downfield by 4.8 ppm
with respect to the solution spectrum of 1.
Figure 4. NMR spectra of 2: (a) 13C CP MAS neat solid, (b) 13C in
CDCl3 solution, (c) 13C CP MAS solid 15%2@TPP-d12, and (d) 13C
CP MAS solid 9%2-d13@TPP-d12.
chemical shifts of the p-terphenylyl carbons in solutions of 1
and 2 also display remarkable similarity. The spectrum of
15%2@TPP-d12 exhibits additional aromatic signals and two
resolved p-carborane peaks with chemical shifts that are
different from those in the spectrum of 15%1@TPP-d12. The
spectrum shows a similar signal pattern as the solid-state 13C
spectrum of neat 2, yet the spectra are distinctly different. The
signals in the spectrum of 15%2@TPP-d12 are better resolved,
and they are shifted upfield by 1−2 ppm relative to those in the
13C CP MAS ssNMR spectrum of neat 2.
1
The NMR signals in 2 were assigned using H −1H gCOSY
(Figure S4 of the Supporting Information, top), 1H −13C
gHSQC (Figure S4 of the Supporting Information, bottom),
1H NMR, boron-decoupled 1H −13C gHMBC NMR (J = 8 Hz:
Figure S5a of the Supporting Information, J = 18 Hz: Figure S5,
panels b and c, of the Supporting Information), 13C NMR, and
DEPT-135.
The 13C NMR CP MAS ssNMR spectrum of neat 2 (Figure
4a) was recorded using a 5 ms contact time, and it shows a
series of broad largely overlapping peaks in the aromatic region,
and a distinct peak for each p-carborane carbon at 82.7 and 87.5
ppm. The solution 13C NMR spectrum of 2 (Figure 4b) shows
all the peaks resolved and the p-terphenyl resonances have
nearly the same chemical shifts as in the solution 13C NMR
spectrum of 1.
In order to assign the signals in the 13C spectrum of 15%2@
TPP-d12, several additional solid-state experiments were
performed: 13C SPE MAS with a 1 s (Figure S6a of the
Supporting Information) and 60 s (Figure S6b of the
Supporting Information) relaxation delay, 13C CP MAS NMR
with a short contact time (0.2 ms) (Figure S6c of the
Supporting Information), and 13C CP MAS NMR spectrum
with dipolar dephasing (DPD) using a 5 ms contact time
(Figure S6d of the Supporting Information). The 13C CP MAS
NMR experiment with DPD shows only carbons that carry no
protons, while the 13C CP MAS with a short contact time
enhances the signals of protonated carbons relative to those of
In addition to the three intense aromatic singlet resonances
associated with hexagonal TPP-d12, the 13C CP MAS ssNMR
spectrum of 15%2@TPP-d12 (Figure 4c) displays a series of
aromatic resonances associated with the p-terphenyl unit and
identical to those in the 13C spectrum of 15%1@TPP-d12. The
Table 1. NMR Signal Assignments for 1 (δ/ppm)
C-1
C-2
139
137.26
C-3
C-4
C-5
C-6
C-7
25.83
22.62, 30.64
1 in CDCl3
140.73−140.59
138.88
135.49
133.7
128.97−126.70
124.58
79.2
75.96
73.03
15%1@TPP-d12
78.15
562
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
Table 2. Summary of Solid-State 13C NMR Assignments for 2, and the Changes in the Chemical Shift from the NMR Spectrum
in CDCl3 Solution to the 15%2@TPP-d12 CP MAS ssNMR Spectrum
a
solid
C-1
C-2
C-3
C-4
C-5
b,d,f(?)
−2 to −3(?)
C-6
C-7
C-8
b
solution
l,p,q
−1.7 to −2.2
m
c,i
−1.4 to −1.9
a
j,k,n,o,r,s,t
?
h
g
c
Δδ/ppm
−1.7
b
−1.6
−1.4
−1.6
a
c
Peak assignment in ssNMR; See Figure 3. Peak assignment in solution NMR; See Scheme 1. Difference in chemical shifts; solid minus solution.
quaternary carbons. The two experiments reveal that the peaks
C-1 and C-2 belong to quaternary carbons. The same peaks
with identical chemical shifts are also found in the solid-state
13C NMR spectrum of the inclusion compound 15%1@TPP-
d12 (C-1 and C-2, Figure 3d), and they therefore belong to the
four p-terphenyl carbons. The signal of the fifth quaternary p-
terphenyl carbon C-3 in 15%2@TPP-d12 overlaps with that of
one of the quaternary carbons of the 2,3-dichlorophenyl
residue. The signals of carbons C-4 and C-5 contain
contributions from the two other quaternary carbons of 2,3-
dichlorophenyl, and the peak C-5 also contains at least one of
the three signals of protonated carbons of 2,3-dichlorophenyl.
The solid-state 13C SPE MAS NMR spectrum of 15%2@TPP-
d12 (Figure S6b of the Supporting Information) was recorded
with a relaxation delay of 5T1 of the slowest relaxing carbon (12
s), and it is therefore quantitative. It shows that the peaks C-1
and C-2 contain altogether four carbons, peaks C-3 and C-4
together integrate to three, and peak C-5 also to three. The
integration of the signal C-5 is not accurate due to an overlap
with a TPP signal.
Figure 5. 13C CP MAS NMR spectrum of het-15%2@TPP-d12.
to 15%2@TPP-d12 and a similar set of signals shifted downfield
relative to those found in 15%2@TPP-d12.
The carbon atoms of the carborane unit appear doubled in
this heterogeneous sample, at 79.9 and 82.5 ppm, and at 85.2
and 87.6. Heating the het-15%2@TPP-d12 sample results in an
inclusion compound with the same sets of 13C signals as the
homogeneous 15%2@TPP-d12 sample, while the heating of a
mixture of finely ground 2 and TPP-d12 does not yield an
inclusion compound. This suggests that the inclusion
compound het-15%2@TPP-d12 is not contaminated with
detectable amounts of the neat rotor 2.
The solid-state 13C CP MAS NMR spectrum of 9%2-d13
TPP-d12 (Figure 4d) only contains the signals of the 2,3-
dichlorophenyl unit, as the deuteriated p-terphenyl carbons in
TPP-d12 channels do not get significantly cross-polarized by any
protons in the sample and are missing in the spectrum. The
comparison with the 15%2@TPP-d12 13C CP MAS ssNMR
spectrum confirms the assignment of the carbons C-1, C-2, and
part of C-3 to p-terphenyl and C-4, C-5, and part of C-3 to 2,3-
dichlorophenyl. The C−B peak in the spectrum of 9%2-
d13TPP-d12 is sharper than the C−B signal in the solid-state
spectrum of 15%2@TPP-d12 13C, which has a downfield
shoulder. This indicates that a significant portion of the signal
from the protonated carbons of the p-terphenyl lies in this
shoulder. The remaining aromatic signals (C-4, C-5, and part of
C-3) belong to the 2,3-dichlorophenyl unit.
Dielectric Loss Spectroscopy. In Figure 6, we show
dielectric spectra taken for several samples from 8 to 300 K
with applied frequencies in the audio range from 120 Hz to 12
The comparison with the solution 13C NMR spectrum of 2
suggests that the signal C-1 corresponds to the carbons Cl, Cp,
and Cq, while the signal C-2 corresponds to Cm and a part of
signal C-3 belong to Ci. Most of the protonated carbon signals
are not resolved (C-6) and they overlap with the TPP peak C−
B. The remaining part of peak C-3 belongs to the carbon Ca, C-
4 corresponds to the carbon Cc, and a part of the peak C-5
corresponds to the carbon Cb.
All the aromatic and p-carborane carbons in the solid-state
13C NMR spectra of 15%2@TPP-d12 display a large upfield
shift with respect to solution 13C NMR spectrum of 2. The
assignments and changes in the chemical shift relative to the
solution spectrum are summarized in Table 2.
Additional information is available for an incompletely
annealed sample het-15%2@TPP-d12 (Figure 5). In the solid-
state 13C CP MAS NMR spectrum, the four downfield
quaternary p-terphenyl carbon signals C-1 and C-2 all show
similar upfield change of the chemical shifts as in the
homogeneous sample 15%2@TPP-d12. The remaining aromatic
signals are a superposition of peaks with chemical shifts similar
Figure 6. (a) Dielectric loss spectra at 0.12 (black), 1.2 (red), and 12
(blue) kHz. Full lines: (a and b) het-15%2@TPP-d12. Dashed lines:
(a) 15%1@TPP-d12 and (b) 15%2@TPP-d12.
563
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
Figure 7. BLYP+D/3-21G geometry optimization of a single TPP channel with 2 inserted. The x axis is the displacement of 2 into (R > 0) or out of
(R < 0), the channel along the channel axis.
kHz. The unannealed sample het-15%2@TPP-d12 shows two
dispersing dielectric loss structures, one near 150 K, consistent
with thermal activation over a barrier of 8.75 0.5 kcal/mol
with an attempt frequency of 1 × 1014 Hz and an asymmetry in
the wells of 0.22 0.02 kcal/mol, and the second near 70 K,
consistent with a barrier of 3.39 0.3 kcal/mol and an attempt
frequency of 1013 Hz and a more symmetric well with only 0.09
0.01 kcal/mol asymmetry. The annealed homogeneous
sample 15%2@TPP-d12 spectra show only the high temper-
ature dielectric loss structure with a barrier of 8.52 0.5 and an
asymmetry of 0.26 0.01 kcal/mol.
Calculations. The geometry of 2 inserted to different
degrees into a channel of four layers of TPP was optimized
using the BLYP+D method31 and the 3-21G basis set (the
channel was allowed to expand or contract). The lowest
minimum obtained by this method shows most of the guest
molecule 2 contained inside the channel, except for a small part
of the dichlorophenyl rotator (Figure 7).
surface inclusion compound, with the guest molecules inside
the TPP channels but abutting the surface; (ii) the unannealed
het-15%2@TPP-d12 sample is heterogeneous and contains a
fraction of the guest molecules in the near-surface position and
a fraction in a surface location with the rotator extending
outside the surface; and (iii) the rotator-free 15%1@TPP-d12
sample in which the rotor only carries a methyl group is also
heterogeneous, with about half of the guest molecules in an
unknown position inside the TPP channels and the other half
in a surface location with the methyl group extending outside
the surface. In the following, we first discuss the supporting
structural evidence that is provided by the complementary
types of measurement, and second, we examine the
compatibility of the proposals with the rotational barriers
deduced from the observed dielectric spectra.
15%2@TPP-d12. First, we note that a 15% molar mixture in
TPP-d12 is a very close match to what one expects for a
platelike crystal of TPP with rotors populating the top and
bottom surfaces fully and with plate thickness just sufficient for
the inserted tails to touch along the plate midplane (similar to
what is shown for 1 in Figure 8d). Second, the very similar
lattice parameters observed with the methyl carrying rotor
15%1@TPP-d12 and the dipolar rotator carrying 15%2@TPP-
d12 suggests that the rotators are not important in setting lattice
parameters. Such a result is expected if the rotators are in the
surface layer and partly outside the crystallites. Previously
reported bulk inclusion compounds of a p-carborane with 2,3-
dichlorophenyl moieties on both of its carbons, which clearly
have the 2,3-dichlorophenyl groups fully within the TPP
channels,21 have a much expanded in-plane lattice parameter of
1.258 nm, substantially larger than 15%1@TPP-d12 and 15%2@
TPP-d12, again consistent with the 2,3-dichlorophenyl rotator
being at least partly outside the bulk channels in the latter two.
The solid-state 13C CP MAS NMR spectrum of 15%2@TPP-
d12 only displays upfield changes in the chemical shifts relative
to the solution 13C NMR spectrum, indicating that the 2,3-
dichlorophenyl rotator is located inside the TPP-d12 channel. If
the dichlorophenyl rotator were freely rotating, the T1
DISCUSSION
■
In the samples examined, 1 and 2 are undoubtedly included as
guests in the channels of TPP: (i) the XRD results show the
exclusive presence of the hexagonal phase of TPP, with
expanded lattice parameters in the plane of the layers. (ii) The
CP MAS experiments exhibit cross-polarization of 31P and 13C
atoms of fully deuteriated TPP, which indicates close proximity
1
to H atoms of the guest molecules. The same signals are also
observed for these atoms in SPE MAS experiments, which do
not depend on cross-polarization, indicating homogeneous
hexagonal structure throughout the samples, and the absence of
the monoclinic form of TPP (within the sensitivity of this
method). (iii) Most chemical shifts of the carbon atoms of the
guest molecules observed in the inclusion compounds are
displaced upfield. The following discussion depends on the firm
assignments of 13C NMR signals to individual carbons that
were described above.
We propose that (i) the annealed 15%2@TPP-d12 sample is
homogeneous and its structure can be characterized as a near-
564
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
15%1@TPP-d12. The near identity of the XRD results for
this inclusion compound and for 15%2@TPP-d12 suggests that
1 inserts its p-terphenyl shaft and the attached carborane cage
inside the TPP channel, and this agrees perfectly with the
upfield shift of all its carbon atoms in the NMR spectrum. The
only exception is the methyl carbon signal, which is split into
two peaks, showing that the methyl group is in two distinct
environments. One of the peaks is shifted upfield, by 3.2 ppm
with respect to the solution spectrum of 1 and has the methyl
group inside the TPP channel. We cannot tell if it is a true bulk
inclusion or if these molecules of 1 reside near the surface. The
other peak is shifted strongly downfield by 4.8 ppm, and we
attribute this to a surface inclusion with the methyl groups just
above the surface, in the deshielding zone of the channel
(Figure 8a). It could be argued that the molecules of 1 can
possibly fill the channel with three different relative orientations
as shown in Figure 8 and that the remarkable 4.8 ppm
downfield change in the chemical shift of the methyl carbon is
due to deshielding by the terminal p-terphenyl phenyl group of
a neighboring guest inside the channel (Figure 8c), but we
consider it highly unlikely. Many solid-state 13C NMR spectra
of bulk inclusion compounds are known30,32,33 and no such
strong effect on the chemical shift has ever been observed. It is
remarkable that application of dipolar dephasing (Figure 3c)
suppressed both methyl signals despite the fast rotation that
would be expected for such a group. In order to remove first
order dipolar coupling, any motion must be isotropic, as well as
faster than the magnitude of the decoupling energy. In addition,
since this methyl group is rather isolated, the C−H dipolar
coupling could be behaving as an isolated dipolar spin system,
in which the coherence nutates as a function of dephasing
delay. It is possible that the delay chosen in this experiment was
at the zero-crossing of this nutation of the CH dipolar coupling.
In spite of the much smaller size of its rotator, 1 thus appears
to be more inclined to form a true surface inclusion with the
rotator outside than 2 does. We suggest that this is due to a
favorable π−π stacking interaction of the 2,3-dichlorophenyl
rotator of 2 with the channel walls, which is absent in case of
the methyl in 1.
Figure 8. Possible relative orientations of 1 in a TPP-d12 channel.
relaxation times of these carbons should be lower than those of
carbons located inside the channels. The 13C SPE MAS NMR
spectrum with short relaxation delay (Figure S6 of the
Supporting Information) does not show any decrease in 13C
intensity of the shaft carbons versus the carbons of the
dichlorophenyl rotator, which is also consistent with the
insertion of the rotator into the channel.
A way to reconcile the apparently contradictory conclusions
from XRD and the NMR methods is to postulate that 15%2@
TPP-d12 is an inclusion compound in which the 2,3-
dichlorophenyl rotators are inserted in the first layer, in
agreement with our DFT calculations (Figure 7). The insertion
of all the rotor parts, including the rotator, is likely due to a
stabilizing π-stacking interactions of the 2,3-dichlorophenyl
group with an expanded TPP-d12 channel.
Dielectric Spectroscopy. To help isolate the behavior of
the dipolar rotors from the general dielectric response, we
compare the dielectric loss spectra of 15%2@TPP-d12 and het-
15%2@TPP-d12 with those of the nonpolar 15%1@TPP-d12
(Figure 4). This material shows low levels of dielectric loss and
serves as a good indication of overall contributions of TPP and
the carborane plus terphenyl moieties in the guest. Both
15%2@TPP-d12 and het-15%2@TPP-d12 show substantial new
dielectric loss peaks, attributed to the presence of 2,3-
dichlorophenyl dipolar rotators. While the former has only a
barrier peak at 8.52 0.5 kcal/mol, the latter has dipoles in two
distinct environments, with barriers to rotation of 8.75 0.5
kcal/mol and 3.39 0.3 kcal/mol.
Het-15%2@TPP-d12. Figures 4 and 5 show NMR spectra
for samples of neat 2, het-15%2@TPP-d12, and 15%2@TPP-
d
12, which has received additional annealing at 70 °C to lead to
The considerable ∼8.6 kcal/mol barrier observed in both
samples suggests that the 2,3-dichlorophenyl rotator faces
significant steric hindrance to rotation. Given its size, this
would be expected if it is included in the first layer of TPP. The
value is consistent with the rotational barriers found for the 2,3-
dichlorophenyl rotator in a bulk TPP inclusion compound of
di(1,12-dichlorophenyl)-p-carborane,21 which carries the 2,3-
dichlorophenyl rotator on both antipodal carbons. Their
heights are ∼5.4, ∼6.7, and ∼9.3 kcal/mol (attributed to
three different local environments). Here, the TPP in-plane
lattice parameter is expanded to 1.254 nm to fully
a more homogeneous NMR structure. The signals associated
with the carbons of the p-carborane cage appear as two peaks in
the spectrum of 15%2@TPP-d12 and are both shifted upfield
relative to the solution spectrum of 2. In het-15%2@TPP-d12,
however, these two peaks are split into pairs of shifted and
unshifted peaks, indicating the presence of two populations of
rotor molecules. One of them is the same as in 15%2@TPP-d12,
with the carborane fully inserted into TPP, and the other has
the carborane and therefore also the rotator outside. Annealing
to 70 °C produces a homogeneous population where all
carboranes are inserted at least as far as the first TPP layer.
565
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
accommodate the carborane and the rotator. The dielectric
measurement thus provides further evidence for a near-surface
inclusion of the rotor 2.
The het-15%2@TPP-d12 sample has an additional population
of dipoles with a much lower barrier of ∼3.4 kcal/mol. These
must be the rotators observed by NMR to be located further
outside the TPP surface. The height of this barrier is still clearly
in excess of the intrinsic intramolecular rotational barrier, most
likely because one of the chlorine substituents on the rotator
remains too close to the TPP surface.
Mechanism of Formation of the Inclusion Com-
pounds. The expanded lattice parameters in these and other
inclusion compounds formed by ball milling suggest a simple
explanation of how self-assembly may occur: any surface
perpendicular to the c axis direction of an empty TPP crystallite
that becomes populated with rotors, a necessary step for
formation of either a surface or bulk inclusion compound, will
be subjected to a substantial strain. The strain may naturally
cause the crystallite to fracture along a plane perpendicular to
the c axis into a populated and an unpopulated part. This
process will take place on both sides of a crystallite plate, and
sequential fracturing and population during ball milling may
then lead to the observed structure.
EXPERIMENTAL PART
■
X-ray Diffraction. The X-ray powder patterns were taken with a
system based on a Rigaku Ultrax18 rotating anode generator. It uses a
curved silicon multilayer monochromator to produce Cu Kα radiation
at wavelength λ = 0.15418 nm. Powder samples, loaded into 1.0 mm
diameter borosilicate glass capillaries with a wall thickness of 10 μm,
were mounted on a Huber four-circle goniometer. The scattered X-
rays were measured with a sodium iodide scintillator point detector
that was moved in a horizontal plane, by an angle 2θ with respect to
the direction of the incident X-rays, to scan the Bragg scattering
profile. The resolution of the instrument, in its usual configuration, is
qres ≈ 0.003 Å−1.
DSC. The DSC measurements were performed with a TA
Instruments DSC Q200 apparatus.
NMR. Solution spectra were obtained with Varian INOVA-400 and
INOVA-500 and Bruker Avance-III 300 NMR spectrometers. Solid-
state NMR measurements were performed with a Varian INOVA-400
(400 MHz) NMR spectrometer equipped with 5 and 4 mm solid-state
NMR probes. Magic-angle sample spinning (MAS) used spinning
frequencies of 12.5−14 kHz for CPMAS and SPE-MAS experiments.
For dipolar dephasing experiments, spinning speeds of 5−6 kHz were
used to allow adequate time for dipolar dephasing and chemical shift
refocusing within one rotor period. Cross-polarization experiments
were performed using spin-lock fields of 72−75 kHz, and ramped-
amplitude cross-polarization (RAMP-CP) transfer of magnetization
was applied. For both cross-polarization and single-pulse excitation
1
CONCLUSIONS
(SPE) experiments, broadband TPPM-modulated H decoupling was
■
used.
The molecular rotors 1 and 2 form inclusion compounds with
TPP. Heterogeneous samples with two dominant environments
for the guest species were obtained from both, and a
homogeneous sample with a single dominant guest environ-
ment was obtained from 2. In the homogeneous sample,
15%2@TPP-d12, the entire rotor molecule is inserted but
remains at the surface with the rotator in the first TPP layer,
forming what we term a near-surface inclusion compound. This
same structure was derived independently from DFT modeling.
As a result of rotator embedding, its rotation is considerably
hindered (∼8.6 kcal/mol).
Chemical shifts for solid-state 13C and 31P nuclei were calibrated
using the absolute referencing method, relative to the absolute
34
1
frequency of 0.0 ppm in H NMR. As a check for this referencing
method, external samples of hexamethylbenzene (HMB, 13C) and
triphenylphosphine (31P) were acquired and calibrated using this
referencing method, yielding 13C chemical shifts of HMB of 132.2 and
17.4 ppm. The observed chemical shift of pure, solid triphenylphos-
phine using the absolute referencing method was −8.5 ppm.
Dielectric Spectroscopy. The powder samples were loaded onto
the surface of coplanar interdigitated electrode capacitors fabricated by
gold metallization on fused silica substrates (each with 50 fingers
separated by ∼10 μm gaps). Nominal capacitor values for the empty
capacitors were near 1 pF and dielectric loss tangents were at or below
the 10−5 level.35 An Andeen−Hagerling capacitance bridge was used to
measure the dielectric loss and capacitance with a sinusoidal
alternating voltage (1.8−5 V and 0.12−12 kHz) applied on the
capacitor. Data were taken continuously with the temperature slowly
changing between 7 K and room temperature. The curves were
averaged over heating and cooling steps and smoothed.
Calculations. The geometry of a single channel composed of four
layers of TPP with a guest molecule 2 (Figure 7) was optimized using
BLYP with dispersion and the 3-21G basis set using Terachem 1.5.31
The channel was only partially constrained during the optimization to
allow for expansion or contraction. The initial geometries were
generated from the R = 0 geometry shown in Figure 7 by moving the
guest by multiples of 0.1 nm either inside or outside the channel along
its axis, and they were further optimized into the nearest local
minimum.
Synthesis. General Methods. Air-sensitive reactions were carried
out under argon, using standard Schlenk techniques. Tetrahydrofuran
was distilled from Na/K under argon. Pyridine was distilled from
calcium hydride under argon. Triethylamine was distilled from calcium
hydride or from sodium/benzophenone under argon. Dimethylforma-
mide was distilled from calcium hydride under reduced pressure.
Cuprous chloride was purchased from Sigma-Aldrich in sealed
ampules and handled in a glovebox. Only pure white cuprous chloride
was used in reactions. All chemicals except for the p-carborane
(Katchem) were purchased from Sigma-Aldrich and used as received.
Standard procedures were used to characterize all new compounds.
IR spectra were recorded with a Nicolet Avatar 360 FT-IR
spectrophotometer using KBr that had been dried at 200 °C for 24
The heterogeneous sample obtained from 2 contains this
same structure as one dominant component. The other main
component has the rotator protruding outside the TPP surface
but apparently not far enough to avoid all friction against the
surface, and it turns with a barrier of ∼3.4 kcal/mol.
The inclusion compound 15%1@TPP-d12 shows an interest-
ing behavior: it has about equal populations of a structure with
the methyl inserted and a structure with the methyl outside.
Clearly, although 1 has the same shaft and the stopper as 2, its
methyl group does not show as much affinity for the channel as
the 2,3-dichlorophenyl group of 2 does. This is attributed to a
weaker stabilizing interaction between the TPP channel and the
methyl group, which is too small for a van der Waals contact
with the channel walls, as opposed to the 2,3-dichlorophenyl
group, which nominally is too large. This observation provides
an important lesson about the design of the rotator group in
designing rotors for surface inclusion compounds.
In summary, we have prepared TPP inclusion samples with
interesting structures and have learned more about the way the
inclusion works, but we have not produced a homogeneous
sample with a sufficiently small barrier to rotation to show
collective dielectric behavior. Apparently, the stoppers need to
be even larger than p-carborane, or we need to work with larger
crystal facets that might make it harder for the guest molecules
to expand the lattice and force their way inside the TPP
channels. It might also be helpful to choose rotators that have
less affinity for the inside of the TPP channel.
566
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
h. UV−vis spectra were recorded with a Hewlett-Packard 8452A
spectrometer. A Hewlett-Packard 5989B mass spectrometer was used
to record mass spectra. Elemental analyses were performed by
Columbia Analytical Services, Tucson, AZ.
Inclusion Compound Preparation. The inclusion compounds were
prepared by ball milling of empty hexagonal TPP-d12 with the neat
rotor followed by annealing at 70 to 75 °C and repetition of the two
steps (15%1@TPP-d12: ball milling for 80 min, annealing for 1 day,
ball milling for 30 min, and annealing for 3 days; 15%2@TPP-d12: ball
milling for 120 min, annealing for 1 day, ball milling for 30 min, and
annealing for 3 days).
2H), 7.08 (dt, J = 8.1 Hz, 1H), 3.63−1.99 (m, 10H). 13C{1H} NMR
(75 MHz, CDCl3): δ 141.08 (s), 140.69 (s), 140.67 (s), 138.91 (s),
135.93 (s), 135.39 (s), 134.20 (s), 131.48 (s), 131.05 (s), 130.71 (s),
128.98 (s), 127.70 (s), 127.63 (s), 127.60 (s), 127.51 (s), 127.18 (s),
126.82 (s), 126.67 (s), 86.52 (s), 81.80 (s). 11B{1H} NMR (96 MHz,
CDCl3): δ −12.28 (s). IR (KBr, cm−1): 490, 528, 534, 565, 606, 616,
660, 697, 713, 739, 754, 763, 787, 919, 843, 865, 870, 905, 914, 931,
966, 981, 1002, 1015, 1037, 1047, 1061, 1082, 1122, 1132, 1164, 1194,
1204, 1255, 1267, 1314, 1330, 1358, 1397, 1450, 1485, 1504, 1558,
1581, 1594, 1729, 1920, 1950, 1965, 2606, 2663, 2852, 2922, 2955,
3029, 3048, 3055, 3110. HRMS (ESI−): 552.2105 (calcd for MCl−:
552.2083). UV−vis (CH2Cl2, nm) λmax (εmax): 290 (46840). Elemental
analysis: calcd C, 60.34%; H, 5.06%. Found: C, 60.13%; H, 5.17%.
Synthesis of 4-Iodo-p-terphenyl-d13 (4). A 5 mL glass vial was
charged with p-terphenyl-d14 (250.0 mg, 1.023 mmol, 1.0 equiv), silver
nitrate (700.0 mg, 4.250 mmol, 4.15 equiv), iodine (1.05 g, 4.132
mmol, 4.0 equiv.), anhydrous acetonitrile (0.2 mL), and a stir bar. The
vial was closed with a lid with a Teflon liner, and the resulting paste
was shaken vigorously for 12 h. It was then allowed to sit at room
temperature for 48 h. The mixture was transferred into a separatory
funnel with dichloromethane (100 mL), washed with 10% ammonium
hydroxide (2 × 80 mL), 5% sodium thiosulfate (2 × 50 mL), and
water (80 mL). The organic layer was dried over anhydrous sodium
sulfate, filtered, and the solvents were removed under reduced
pressure. The crude product was purified by two Kugelrohr
sublimations at 80 to 120 °C (50 to 80 mTorr) to yield 245.9 mg
(65%) of pure white crystalline product. 1H NMR (300 MHz,
CDCl3): δ 7.81 (s), 7.70 (s), 7.66 (s), 7.48 (s), 7.41 (s), 7.39 (s).
Synthesis of 3-d13. The product was prepared by the same
procedure as the nondeuteriated version 3,22 starting with 4.
Kugelrohr sublimation (140 °C/30 mTorr) was used in place of the
crystallization. The yield of the product was 89%. 1H NMR (300 MHz,
CDCl3) δ 7.65 (s), 7.63 (s), 7.60 (s), 7.46 (s), 7.45 (s), 7.36 (s), 7.30
(s), 3.68−0.75 (m).
1-Methyl-12-(2,3-dichlorophenyl)-p-dicarba-closo-dodecaborane
(1). A dry Schlenk flask (25 mL) was equipped with a septum and a
magnetic stir bar. It was charged with 322 (200.0 mg, 0.537 mmol, 1.0
equiv), evacuated, and filled with argon (3×). Anhydrous THF (20
mL) was added through the septum, and the mixture was cooled in an
acetone/dry ice bath. After 5 min, 1.6 M solution of n-BuLi in hexanes
(0.40 mL, 0.644 mmol, 1.2 equiv) was added to the solution, which
was then allowed to warm to room temperature. After 30 min of
stirring at room temperature, the mixture was again placed in an
acetone/dry ice bath, and methyl iodide (152.3 mg, 1.074 mmol, 2.0
equiv) was added. The resulting mixture was allowed to warm to room
temperature, and it was stirred overnight (16 h). It was quenched with
a few drops of water, and the solvents were removed under reduced
pressure. The solid residues were dissolved in dichloromethane (100
mL), washed with 5% hydrochloric acid (20 mL) and water (2 × 20
mL), and the organic layer was dried over anhydrous sodium sulfate,
filtered, and the solvents were removed under reduced pressure.
Crystallization from cyclohexane provided 163.7 mg (79%) of white
crystals, mp (CH2Cl2) 261 °C. 1H NMR (300 MHz, CDCl3): δ 7.68−
7.55 (m, 6H), 7.48−7.41 (m, 4H), 7.39−7.32 (m, 1H), 7.28 (dt, J =
7.1, 1.3 Hz, 2H), 1.42−3.5 (m, 10H), 1.48 (s, 3H). 13C{1H} NMR (75
MHz, CDCl3): δ 140.73 (s), 140.68 (s), 140.59 (s), 139.00 (s), 135.49
(s), 128.96 (s), 127.83 (s), 127.66 (s), 127.56 (s), 127.49 (s), 127.17
(s), 126.70 (s), 79.22 (s), 75.99 (s), and 25.83 (s). 11B{1H} NMR (96
MHz, CDCl3): δ −11.97 (s). IR (KBr, cm−1): 429, 445, 510, 528, 566,
579, 600, 620, 666, 700, 719, 730, 767, 813, 837, 864, 882, 979, 1004,
1032, 1069, 1115, 1133, 1169, 1196, 1253, 1309, 1318, 1331, 1361,
1387, 1396, 1412, 1445, 1482, 1496, 1503, 1526, 1542, 1558, 1568,
1580, 1595, 1953, 2600, 2609, 2873, 2939, 2987, 3027, 3047, 3075,
3085. HRMS (ESI−): 422.2718 (calcd for MCl−: 422.2718). UV−vis
(CH2Cl2, nm) λmax (εmax): 286 (43980). Elemental analysis: calcd C,
65.25%; H, 6.78%. Found: C, 64.86%; H, 7.18%.
Synthesis of 2-d13. The same procedure as for synthesis of 2 was
1
employed, starting with 3-d13. The yield was 94%. H NMR (300
MHz, CDCl3): δ 7.59 (dd, J = 8.4, 1.4 Hz, 1H), 7.41 (dd, J = 7.9, 1.4
Hz, 1H), 7.07 (t, J = 8.1 Hz, 1H), 2.93 (s, 5H), 2.70 (s, 5H).
ASSOCIATED CONTENT
* Supporting Information
DSC traces, NMR assignments, relaxation times. This material
is available free of charge via the Internet at http://pubs.acs.org.
■
S
1-(4-p-Terphenylyl)-12-(2,3-dichlorophenyl)-p-dicarba-closo-do-
decaborane (2). A pressure Schlenk flask (25 mL), equipped with a
pressure valve, a magnetic stir bar, a septum, and a hose connector, was
charged with 3 (300 mg, 0.804 mmol, 1.0 equiv) and with cuprous
chloride (103.4 mg, 1.045 mmol, 1.3 equiv) under argon. THF (10
mL) was added through the septum under argon, and the resulting
mixture was cooled in an acetone/dry ice bath. After 5 min of stirring,
n-BuLi (1.6 M solution in hexanes, 0.603 mL, 0.965 mmol, 1.2 equiv)
was added dropwise under argon. The reaction mixture was stirred for
15 min at −78 °C and then for 30 min at room temperature.
Anhydrous pyridine (1 mL) was added to the mixture, and it was
stirred until all the cuprous chloride dissolved. The solvents were
removed under reduced pressure and the flask was filled with argon.
After that, a suspension/solution of the 1,2-dichloro-3-iodobenzene
(329 mg, 1.206 mmol, 1.5 equiv) in anhydrous DMF (8 mL) was
added under argon. The resulting mixture was heated in a closed vessel
at 120 °C for 24 h and then at 140 to 150 °C for 24 to 48 h. After the
reaction was complete, the solvents were removed under reduced
pressure and the solids were dissolved in dichloromethane (100 mL),
washed with 5% hydrochloric acid (50 mL), with 10% ammonium
hydroxide (2 × 50 mL), and with water (50 mL). The organic layer
was dried over anhydrous sodium sulfate, filtered, and concentrated
under reduced pressure. The product was purified by chromatography
in pentanes and toluene (3/1 to 2/1) on silica gel to yield 399 mg
AUTHOR INFORMATION
Corresponding Author
*E-mail: michl@eefus.colorado.edu.
Notes
■
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported in Prague by the European Research
Council under the European Community’s Seventh Framework
Programme (FP7/2007-2013) ERC Grant 227756 and by the
Institute of Organic Chemistry and Biochemistry, Academy of
Sciences of the Czech Republic RVO: 61388963, and in
Boulder by the National Science Foundation under Grant CHE
0848663.
REFERENCES
■
(1) Kottas, G. S.; Clarke, L. I.; Horinek, D.; Michl, J. Chem. Rev.
2005, 105, 1281.
(2) Khuong, T.-A. V.; Nunez, J. E.; Godinez, C. E.; Garcia-Garibay,
M. A. Acc. Chem. Res. 2006, 39, 413.
(3) Horansky, R. D.; Magnera, T. F.; Price, J. C.; Michl, J. In
Controlled Nanoscale Motion, Lecture Notes in Physics; Linke, H.;
1
(96%) of pure product (white crystals), mp (CH2Cl2): 234 °C. H
NMR (500 MHz, CDCl3): δ 7.69−7.57 (m, 7H), 7.50−7.44 (m, 4H),
7.41 (dd, J = 7.9, 1.5 Hz, 1H), 7.38 (d, J = 7.4 Hz, 1H), 7.35−7.31 (m,
Mansson, A., Eds., Springer: Berlin, 2007; Vol. 711, p 303.
̊
567
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568

Crystal Growth & Design
Article
(4) Crowly, J. D.; Kay, E. R.; Leigh, D. A. Intelligent Materials;
Shahinpoor, M., Schneider, H.-J., Eds.; 2008; Vol 1, p 1−47.
(5) Garcia-Garibay, M. A. Nat. Mater. 2008, 7, 431.
(6) Michl, J.; Sykes, E. C. H. ACS Nano 2009, 3, 1042.
(7) Augulis, R.; Klok, M.; Feringa, B. L. Phys. Status Solidi C 2009, 6,
181.
(8) Winston, E. B.; Lowell, P. J.; Vacek, J.; Chocholousova, J.; Michl,
J.; Price, J. C. Phys. Chem. Chem. Phys. 2008, 10, 5188.
(9) Horansky, R. D.; Clarke, L. I.; Price, J. C.; Khuong, T-A. V.;
Jarowski, P. D.; Garcia-Garibay, M. A. Phys. Rev. B 2005, 72, 014302.
(10) Rodriguez-Molina, B.; Ochoa, M. E.; Farfan, N.; Santillan, R.;
Garcia-Garibay, M. A. J. Org. Chem. 2009, 74, 8554.
(11) Gimzewski, J. K.; Joachim, C.; Schlittler, R. R.; Langlais, V.;
Tang, H.; Johannsen, I. Science 1998, 281, 531.
(12) Zheng, X.; Mulcahy, M. E.; Horinek, D.; Galeotti, F.; Magnera,
T. F.; Michl, J. J. Am. Chem. Soc. 2004, 126, 4540.
(13) Tierney, H. L; Murphy, C. J.; Jewell, A. D.; Baber, A. E.; Iski, E.
V.; Khodaverdian, H. Y.; McGuire, A. F.; Klebanov, N.; Sykes, E. C. H
Nat. Nanotechnol. 2011, 6, 625.
(14) Delden, R. A.; ter Wiel, M. K. J.; Pollard, M. M.; Vicario, J.;
Koumura, N.; Feringa, B. L. Nature 2005, 437, 1337.
(15) Clarke, L. I.; Horinek, D.; Kottas, G. S.; Varaksa, N.; Magnera,
T. F.; Hinderer, T. P.; Horansky, R. D.; Michl, J.; Price, J. C.
Nanotechnology 2002, 13, 533.
(16) Carroll, G. T; Pollard, M. M.; van Delden, R.; Feringa, B. L.
Chem. Science 2010, 1, 97.
(17) Rodriguez-Molina, B.; Farfan, N.; Romero, M.; Mendez-Stivalet,
J. M.; Santillan, R.; Garcia-Garibay, M. A. J. Am. Chem. Soc. 2011, 133,
7280.
(18) Mulcahy, M. E.; Magnera, T. F.; Michl, J. J. Phys. Chem. C 2009,
113, 20698.
(19) Rozenbaum, V. M.; Ogenko, V. M.; Chuiko, A. A. Phys.-Usp.
1991, 34, 883.
(20) Kobr, L.; Zhao, K.; Shen, Y.; Comotti, A.; Bracco, S.;
Shoemaker, R. K.; Sozzani, P.; Clark, N. A.; Price, J. C.; Rogers, C.
T.; Michl, J. J. Am. Chem. Soc. 2012, 134, 10122.
(21) Kobr, L.; Zhao, K.; Shen, Y.; Polívkova, K.; Shoemaker, R. K.;
́
Clark, N. A.; Price, J. C.; Rogers, C. T.; Michl, J. J. Org. Chem. 2013,
78, 1768.
(22) Kobr, L.; Zhao, K.; Shen, Y.; Shoemaker, R. K.; Rogers, C. T.;
Michl, J. Adv. Mater. 2013, 25, 443.
(23) Allcock, H. R.; Siegel, L. A. J. Am. Chem. Soc. 1964, 86, 5140.
(24) Hurdis, E. C.; Smyth, C. P. J. Am. Chem. Soc. 1942, 64, 2212.
(25) Debye, P. Polar Molecules; Dover Publications, Inc.: New York,
1929.
(26) Frohlich, H. Theory of Dielectrics: Dielectric Constant and
̈
Dielectric Loss, 2nd ed.; Clarendon Press: Oxford, 1958.
(27) Daniel, V. V. Dielectric Relaxation; Academic Press: New York,
1967.
(28) Comotti, A.; Simonutti, R.; Catel, G.; Sozzani, P. Chem. Mater.
1999, 11, 1476.
(29) Allcock, H. R.; Levin, M. L.; Whittle, R. R. Inorg. Chem. 1986,
25, 41.
(30) Comotti, A.; Simonutti, R.; Stramare, S.; Sozzani, P.
Nanotechnology 1999, 10, 70.
(31) Umfitsev, I. S.; Martínez, T. J. J. Chem. Theory Comput. 2009, 5,
2619.
(32) Brustolon, M.; Barbon, A.; Bortolus, M.; Maniero, A. L.;
Sozzani, P.; Comotti, A.; Simonutti, R. J. Am. Chem. Soc. 2004, 126,
15512.
(33) Sozzani, P.; Comotti, A.; Bracco, S.; Simonutti, R. Angew. Chem.,
Int. Ed. 2004, 43, 2792.
(34) Harris, R. K.; Becker, E. D.; Cabral de Menzes, S. M.; Grainger,
P.; Hoffman, R. E.; Zilm, K. W. Pure Appl. Chem. 2008, 80, 59.
(35) Underwood, J. M.; Price, J. C. Rev. Sci. Instrum. 2008, 79,
093905.
568
dx.doi.org/10.1021/cg4013608 | Cryst. Growth Des. 2014, 14, 559−568