Crystal structure of a highly efficient clarifying agent for isotactic polypropylene

Article abstract of DOI:10.1021/cg300151s

Using a combined approach based on X-ray powder diffraction and solid-state NMR spectroscopy, we were able to determine the crystal structure of 1,3,5-tris(2,2-dimethylpropionylamino)benzene, an efficient clarifying agent for isotactic polypropylene. The XPRD data and 1D solid-state NMR experiments allowed to exclude most of the possible primitive orthorhombic space groups with the exception of 11. The structure solution was carried out using real space methods including a close-contact penalty. Four space groups lead to reasonable wRp values below 10%. ¹³C¹³C double quantum (DQ) experiments of the labeled carbonyl group were measured with a supercycled symmetry based dipolar recoupling sequence. Taking into account the wRp values after Rietveld refinement as well as simulations of DQ build-up curves based on nine spin systems, the space group P2₁2₁2₁ is clearly favored. The largest dipolar coupling within these spin systems was about 60 Hz corresponding to a distance of 5 A. 1,3,5-Tris(2,2- dimethylpropionylamino)benzene crystallizes in an orthorhombic metric (a = 14.91(5) A, b = 24.05(8) A, c = 6.80(2) A). The molecules are arranged in a pseudohexagonal rod packing with medium-strong hydrogen bonds and π-stacking. The antiferroelectric arrangement of neighbored rods leads to an averaging of the net dipolar moment in the whole crystal.

Full text of DOI:10.1021/cg300151s

Article
pubs.acs.org/crystal
Crystal Structure of a Highly Efficient Clarifying Agent for Isotactic
Polypropylene
Marko Schmidt,† Johannes J. Wittmann,† Roman Kress,‡ Denis Schneider,§ Stefan Steuernagel,§
Hans-Werner Schmidt,*,‡ and Jurgen Senker*,†
̈
†Inorganic Chemistry III, University of Bayreuth, 95440 Bayreuth, Germany
‡Macromolecular Chemistry I, University of Bayreuth, 95440 Bayreuth, Germany
§Bruker Biospin GmbH, Silberstreifen 4, 76287 Rheinstetten, Germany
S
* Supporting Information
ABSTRACT: Using a combined approach based on X-ray
powder diffraction and solid-state NMR spectroscopy, we were
able to determine the crystal structure of 1,3,5-tris(2,2-
dimethylpropionylamino)benzene, an efficient clarifying
agent for isotactic polypropylene. The XPRD data and 1D
solid-state NMR experiments allowed to exclude most of the
possible primitive orthorhombic space groups with the
exception of 11. The structure solution was carried out using
real space methods including a close-contact penalty. Four
space groups lead to reasonable wRp values below 10%. 13C13C
double quantum (DQ) experiments of the labeled carbonyl
group were measured with a supercycled symmetry based
dipolar recoupling sequence. Taking into account the wRp values after Rietveld refinement as well as simulations of DQ build-up
curves based on nine spin systems, the space group P212121 is clearly favored. The largest dipolar coupling within these spin
systems was about 60 Hz corresponding to a distance of 5 Å. 1,3,5-Tris(2,2-dimethylpropionylamino)benzene crystallizes in an
orthorhombic metric (a = 14.91(5) Å, b = 24.05(8) Å, c = 6.80(2) Å). The molecules are arranged in a pseudohexagonal rod
packing with medium-strong hydrogen bonds and π-stacking. The antiferroelectric arrangement of neighbored rods leads to an
averaging of the net dipolar moment in the whole crystal.
density of nucleation sites and inducing crystallization of low
light-scattering rod-like crystals.27 A detailed insight into the
state of the art of nucleation and clarification of i-PP is
published by Gahleitner et al.28 Among the different classes of
nucleating and clarifying agents the class of 1,3,5-benzene-
trisamide attracts more and more attention.23−26
1. INTRODUCTION
The class of 1,3,5-benzenetrisamides (BTAs) has evolved to
one of the most versatile motifs in supramolecular chemistry.
First, Matsunaga et al. reported on thermotropic liquid
crystalline behavior of alkyl substituted BTAs.1,2 Experimental
studies on self-assembly3−9 revealed the preferred formation of
supramolecular columnar aggregates. Stabilized by strong
hydrogen bonds between three amide groups per molecule
the formed nanofibers are able to build up hydrogels10,11 and
organogels,12−15 depending on the nature of the substituents.
BTAs were used to study concepts of chiral amplification in
helical self-assembly.3,16−19 Kreger et al. showed that
azobenzene substituted BTAs are interesting for holographic
volume gratings.20 They are applied successfully as additives to
improve the electret performance of polypropylene21,22 and as
nucleating agents for polyvinylidenfluoride23 and isotactic
polypropylene (i-PP).23−26 In this regard, the significantly
shorter cycle times in injection molding caused by the increased
crystallization temperature of nucleated i-PP are of highly
economic interest. New subclasses are the so-called “clarifying
agents” which render the normally turbid i-PP highly
transparent and thus opens the application for transparent
packaging. These additives are capable of providing an ultrahigh
Most important for the nucleation and clarification is the
epitaxial matching to induce crystallization and growth of the
crystal lamellae in the mesoscale regime. Nevertheless, only a
few studies exist with respect to the crystallographic structure of
BTAs. Lightfoot et al. described the columnar structure of a 2-
methoxyethyl substituted BTA,29 Jimen
́
ez et al. reported on n-
alkyl substituted BTAs,30 and Kristiansen et al. solved the
structure of a tert-butyl substituted BTA (all based on trimesic
acid).31 In the publication of Blomenhofer et al. the best
reported clarifying agent is 1,3,5-tris(2,2-dimethyl-
propionylamino)benzene (Figure 1).24 The optical properties
of a 1.1 mm thick sample of an i-PP homopolymer were
improved from 79% clarity to 98% and from a haze of 64% to
Received: February 1, 2012
Revised: March 10, 2012
Published: March 26, 2012
© 2012 American Chemical Society
2543
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551

Crystal Growth & Design
Article
13CO2 (0.0225 mol) was added 30 mL of dried and cold Et2O
dropwise, followed by 0.04 mol of t-butyl magnesium chloride (2.0 M
in Et2O; used in excess due to the age of the solution). The mixture
was stirred for several hours at −80 °C (acetone/dry ice) and
defrosted overnight. Et2O was evaporated under reduced pressure, and
a white solid precipitated. This solid (3.62 g, 0.0225 mol) was solved
in 35 mL of dichloromethane (DCM) and 3.9 mL of oxalyl chloride
(0.04 mol, equimolar to t-butyl magnesium chloride) to gain 13C
enriched 2,2-dimethylpropanoyl chloride and MgCl2. The solids were
solved in 80 mL of N-methyl-2-pyrrolidone (NMP), and potassium
stearate was added dropwise (to react excess oxalyl chloride) until no
further gas evolution was observed. The 13C labeled product was used
Figure 1. Left: Chemical structure of 1,3,5-tris(2,2-
dimethylpropionylamino)benzene 1 (if 13C labeled at all three
carbonyl groups, the compound will then be shorten as 1*). Right:
Section of compound 1 highlighting the main motional degrees of
freedom; the black arrows assign the structure directing torsion angles
CArCArNHCO (left) and CMeCqCOO (right).
1
without further purification. All resonances in the H NMR could be
unequivocally assigned to the solvents used during the synthesis and
the 13C labeled acid chloride. Yield: 1.6 g (13 mmol, 51.8%). 1H NMR
(250 MHz, CDCl3): δ = 1.14 ppm (s, 9H, −CH3).
1,3,5-Triaminobenzene was freshly synthesized by hydrogenation of
the nitro compound: 3,5-dinitroaniline (8.0 g, 44 mmol) was stirred in
a mixture of THF (400 mL), MeOH (50 mL), and palladium (0.6 g,
10% on activated carbon) in an autoclave at 35 °C for 12 h using a H2
pressure of 3.5 bar. The palladium catalyst was filtered off under argon
with Alox N, the solvents were evaporated under reduced pressure,
and the product was subsequently dried under high vacuum. Yield: 5.3
g (43.1 mmol, 98%). 1H NMR (250 MHz, [d6]DMSO): δ = 4.30 ppm
(s, 6H, −NH); 5.13 ppm (s, 3H, Ar−H).
10%. In the present work we focus on elucidating the structure
of 1,3,5-tris(2,2-dimethylpropionylamino)benzene 1 by means
of NMR crystallographic strategies.
The main problem with solving the crystal structure of 1,3,5-
tris(2,2-dimethylpropionylamino)benzene is its microcrystallin-
ity. As a consequence, all attempts to prepare suitable single
crystals failed which is often the case also for many other
substance classes like pharmaceuticals, dyes and pigments,
biochemical macromolecules, or inorganic network com-
pounds.32−34 Coping with powder data only complicates ab
initio structure solutions.35−38 NMR crystallographic strategies
help to overcome this problem by complementing powder
diffraction data with solid-state NMR experiments at various
steps of the crystal structure determination.39 Many of these
experiments are based on the three main interactions in solid-
state NMR, namely the chemical shift, the dipolar interaction,
and the quadrupolar splitting.40,41 The chemical shift and the
quadrupolar splitting help to determine the correct space
group, the asymmetric unit, the local symmetry of molecular
units in the crystal structure, and dynamical disorder of
individual building units.39,42−51 Furthermore, using adequate
pulse sequences, the dipolar coupling allows to extract
distances52−59 between homo- and heteronuclear nuclei or
even torsion angles. Thus, it is possible to validate or falsify
structure models, and the structure determination of ambivalent
diffraction data becomes less difficult. However, it is hardly
possible to get sufficient structural information based on nuclei
with a low natural abundance. Therefore, isotope enrichment,
for example 13C or 15N, is often inevitable. During the past
years, these complementary methods have shown that NMR
crystallography is a powerful tool to solve substructures and
even complete crystal structures of molecular compounds and
inorganic network compounds, respectively.60−69
Compound 1* was synthesized by dropping 13C enriched acid
chloride (0.46 g, 3.8 mmol) to a mixture of 1,3,5-triaminobenzene
(0.15 g, 1.2 mmol), NMP (30 mL), pyridine (5 mL), and LiCl (0.05
g) at 0 °C under argon atmosphere. The reaction was stirred overnight
at 60° C, and the reaction mixture was precipitated into an excess of
ice water. The precipitate was filtered off, dried, and recrystallized from
1
methanol. Yield: 0.174 g (0.46 mmol, 38%). H NMR (250 MHz,
3
[d6]DMSO): δ = 1.21 ppm (d, J(1H,13C) = 3.5 Hz, 27H, −CH3);
7.62 ppm (s, 3H, Ar−H); 9.2 ppm (d, 2J(1H,13C) = 3 Hz, 3H, −NH).
MS (70 eV), m/z (%): 378 (M+, 67); 321 (8); 293 (63); 236 (7); 208
(16); 123 (6); 57 (100).
2.2. Powder X-ray Diffraction. The powder pattern was recorded
in transmission using a STOE StadiP with a germanium (111)
monochromator (Cu Kα1 radiation, 1.5406 Å). The sample was filled
in a capillary tube with a diameter of 0.7 mm and measured in the 2θ
range 5−40° at room temperature. The data were fully handled using
the module Reflex Plus of the program package Accelerys MS
Modeling71 (version 5.0). The structure solution step was performed
by real-space methods using the simulated annealing algorithm
including a close-contact penalty. As starting model four molecules
were placed into the unit cell which were geometry optimized before
using DFT methods (see section 2.5). During the structure solution
these molecules were allowed to rotate and translate freely as rigid
bodies with two torsion angles in each side chain as additional degrees
of freedom. The first Rietveld refinement step was performed
including energy considerations using the COMPASS forcefield
where each atom was allowed to relax freely. The additional energy
weighting was set to 50%, and for the energy window the default value
of 40 kcal/mol was used.
2.3. Solid-State NMR. 1H and 13C shifts are reported with respect
to TMS, 15N shifts were referenced on liquid NH3. Compared to
nitromethane all values are shifted by 380.5 ppm. 1H spectra of
compound 1 were recorded on a Bruker Avance III spectrometer
operating at a proton frequency of 500 MHz. The sample was filled in
a 1.3 mm ZrO2 MAS rotor. The MAS frequency and the 90° pulse
length were set to 65 kHz and 2.5 μs, respectively. The 15N and 13C
spectra of the nonlabeled compound were recorded on a Bruker
Avance II 300 spectrometer using triple resonance probes (7 mm and
4 mm) at a MAS frequency of 5 kHz and 12.5 kHz, respectively.
Cross-polarization (CP) from proton to nitrogen and carbon was used
to enhance the signal-to-noise ratio. The 90° pulse on the proton
channel was set to 4.8 μs (3.0 μs), and the contact time was adjusted
to 2 and 5 ms, respectively. 13C 1D spectra and double quantum build-
up curves of compound 1* were recorded on a Bruker Avance III 700
in a 3.2 mm ZrO2 rotor. For 1D spectra, cross-polarization
2. EXPERIMENTAL SECTION
2.1. Synthesis and Characterization of 1*. All solvents were
purified and dried using standard procedures. H NMR spectra were
1
recorded on a Bruker AC250 spectrometer at room temperature and
referenced on TMS. Mass spectra were recorded on a VARIAN MAT
7 instrument with direct probe inlet and electron impact ionization.
The labeled Ba13CO3 was purchased from Deutero GmbH and used
without further purification. 3,5-Dinitroaniline was obtained from
Aldrich and used as received.
The 13C enriched 2,2-dimethylpropanoyl chloride (labeled at the
carbonyl group) was synthesized from Ba13CO3.70 A 15 mL portion of
concentrated sulfuric acid was dropped slowly to 5 g of enriched
barium carbonate (0.025 mol). The generated 13CO2 was caught in a
second flask that was cooled with a mixture of acetone, dry ice, and
liquid nitrogen to a temperature of roughly −120 °C. To 0.99 g of
2544
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551

Crystal Growth & Design
Article
a
Scheme 1. Synthetic Route to 13C Labeled 1,3,5-Tris(2,2-dimethylpropionylamino)benzene 1*
a
Precursors are numbered with I and II. The asterisk marks the 13C labeled carbonyl groups.
experiments at a MAS frequency of 12.5 kHz were performed. The 1H
3. RESULTS AND DISCUSSION
90° pulse length was adjusted to 3.0 μs by applying a contact pulse of 5
ms. The recycle delay was set to 5 s (not fully relaxed for the −NH
and aromatic protons). For the double quantum build-up curves, the
symmetry-based pulse sequence SR26141 was used.72 The performance
of this sequence was checked by recording and evaluating DQ build-up
curves for 13C labeled glycine (γ-polymorph). Here, the intermolecular
13C−13C distances could be reproduced with high accuracy possessing
rms values below 0.2 (compare to section 2.4).
3.1. Synthesis and Characterization of 13C Labeled
1,3,5-Tris(2,2-dimethylpropionylamino)benzene. The
synthesis we here present is well established,26 and the starting
material, 3,5-dinitroaniline (I), is commercially available. The
first step is the reduction of I in order to gain 1,3,5-
triaminobenzene (II) in almost quantitative yields of 98%.
1
The formation of II was characterized by H solution-state
NMR spectroscopy. The spectrum displayed two singlets at δ =
4.30 and 5.13 ppm. The signal at δ = 4.30 ppm can be assigned
to the amino groups while the signal at δ = 5.13 ppm belongs to
the aromatic protons. The upfield shift of the aromatic
resonance is caused by a strong decreasing of the shielding of
the protons due to the electronegativity of the three nitrogen
atoms of the NH2-groups. The expected intensity ratio of 2:1
clearly shows the complete formation as well as the purity of
the product. For the formation of the target molecule (1*) 13C
labeled 2,2-dimethylpropanoyl chloride was synthesized from
commercialized Ba13CO3 as described by Parnes et al.70 with a
The MAS frequency was set to 5.5 kHz corresponding to a nutation
frequency of roughly 36 kHz. The R-block was implemented as a 90°−
270° composite pulse. The filtering corresponding coherence pathway
0
2 0 −1 was performed with the conventional 16-fold phase cycle.52
Additionally, a standardization experiment was collected with the
coherence pathway 0 0 0 −1. The number of repetitions for each 1D
experiment was set to 32. Cross-polarization from proton to 13C with a
1H 90° pulse length of 3.0 μs and a contact time of 5 ms was used. The
read-out 90° pulse on 13C was set to 3.0 μs. During excitation and
reconversion CW decoupling with a nutation frequency of roughly 105
kHz was performed. For all experiments, a broadband SPINAL64
proton decoupling with a nutation frequency of about 85 kHz was
used during data collection.
1
yield of 51.8%. The product was characterized by H solution-
state NMR where only one signal corresponding to the methyl
group of the enriched acid chloride was observed reflecting the
complete conversion to the labeled product. In a third step, the
conversion of II with an excess of 13C labeled 2,2-
dimethylpropanoyl chloride gives the product with a yield of
38% after recrystallization from MeOH (Scheme 1). The 13C
labeled product 1* was characterized by 1H solution-state
NMR, solid-state NMR, mass spectrometry, and powder X-ray
diffraction.
The formation of compound 1* leads to three signal groups
in the 1H solution-state NMR spectrum. The resonances at δ =
1.209 ppm and δ = 9.205 ppm show an additional doublet
splitting while the signal at δ = 7.625 ppm occurs only as singlet
which can be assigned as the aromatic resonance. The signal at
δ = 1.209 ppm belongs to the CH3-group, and the signal at δ =
9.205 ppm can be assigned to the NH-group. Both proton
resonances show a splitting due to the coupling to the
neighbored 13C labeled carbonyl group over three bonds and
two bonds, respectively. The corresponding coupling constants
amount to 3.5 Hz for the methyl protons (3J-coupling) and 3.0
Hz (2J-coupling) for the amide protons. All three signal groups
show an integral ratio of 1:1:8.99 which fits very well to the
theoretical exception of a ratio of 1:1:9 resulting in a high purity
of the product.
2.4. Solid-State NMR Simulations. Simulations for 13C13C build-
up curves with the SR26 experiments were performed with the
simulation package SIMPSON.73 Magnitude and relative orientation
of the carbon−carbon dipole tensors were extracted from the structure
solutions after Rietveld refinement. The parameters for the 13C
chemical shift tensors were estimated from a static 1D spectrum and
included in the spin system. We first performed the simulations within
a seven spin system and completed in a nine spin system to reduce
finite size effects. The Euler angles for the CSA were found to have
only a minor contribution to the shape of the build-up curves. Due to
strong proton decoupling in the experiments, the influence of further
proton interactions could be neglected. The simulations were carried
out using a crystal file of 144 orientations and seven γ angles.
For each simulated build-up curve of the four space groups the root-
mean-square deviation (rms) was calculated with MATLAB74
according to
∑ (simi − exp)2
i
i
rms =
∑i exp2
(1)
i
where simi are the simulated and expi the experimental data points.
The choice of the correct space group was then made based on the
minimal rms values.
2.5. Computational Methods. The geometry optimization of the
molecule before the ab initio structure solution was performed by DFT
methods with the module DMol3 from MS Modeling.71 Within the
DNP (double numerical plus polarization) basis set the GGA
minimization was performed using the PW91 functional. The SCF
(self-consistent field) density convergence was set to 1.0e−6 eV per
atom.
3.2. Symmetry Considerations from 1D Solid-State
1
NMR Spectra. The H, 15N, and the 13C solid-state NMR
spectra of compound 1 are shown in Figure 2a−c. The proton
spectrum displays three different signal groups at roughly δ =
2.0, 6.5, and 9.5 ppm clearly corresponding to the methyl
protons, the aromatic protons, and the protons of the amide
2545
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551

Crystal Growth & Design
Article
described as quasihelical structure accompanying with a partial
loss of the coplanarity leading to a splitting of the aromatic
protons.
The 15N CP spectrum (Figure 2b) displays one single
resonance at δ = 127.5 ppm corresponding to the amide
nitrogen atom. The appearance of only one signal, on one
hand, indicates the entire conversion from 1,3,5-triaminoben-
zene to the product due to an absence of the resonance being
typical for an amino group. On the other hand, the single
resonance suggests that all amide groups depict similar torsion
angles with respect to the aromatic core and the t-butyl groups.
This in turn suggests that only one molecule or one-third is
located in the asymmetric unit. Both aspects display useful hints
with regard to the structure solution which will be discussed at
a later point. Finally, the narrow signal in the 15N CP spectrum
is in very good agreement with the results from the proton
signal of the amide group with respect to the hydrogen bonding
situation.
Complementary results can be obtained from the 13C CP
spectrum of 1 (not labeled, Figure 2c) where five signal groups
can be obtained. At roughly δ = 27 ppm the resonance for the
methyl carbon atoms occurs. Interestingly, two different
chemical shifts can be identified in an integral ratio of 2:1
suggesting a staggered conformation of the methyl groups
concerning the adjacent OC−NH bond. The resonance at δ =
39 ppm belongs to the quaternary carbon atom of the t-butyl
group, and the two aromatic carbon atoms are located at δ =
112 ppm and δ = 138 ppm, respectively, whereas the carbon
atom of the carbonyl group has a chemical shift of δ = 177 ppm.
The existence of only five resonances shows the high purity of
1,3,5-tris(2,2-dimethylpropionylamino)benzene without any
residual educts as well as high order within the structure.
Moreover, the narrow line shape of the signals of the
quaternary carbon atom and the upfield shifted aromatic carbon
atom indicates a high symmetry in the packing of the molecules
with uniformly distributed C−C distances between the benzene
rings in the packing pattern. In the case of a tilted conformation
of the benzene rings, more signals with a broader distribution of
chemical shifts are expected. The 13C CP signal of the carbonyl
group for the enriched BTA measured at a high external
magnetic field (B0 = 16.4 T), where the 14N interaction (I = 1)
is eliminated due to its ω−2 dependence, depicts a marginally
structured resonance due to the slightly different environment
of the carbonyl atoms within one molecule. The signal was
deconvoluted into three resonances (δ = 176.6, 177.0, and
177.2 ppm) possessing equal FWHM with an intensity ratio
1.02(2):1:00(2):1.01(2) (Figure 3). The FWHM of roughly 0.5
ppm pictures a typical value for an α-carbon signal of 13C-
enriched amino acids. This clearly shows that only one
molecule is arranged within the asymmetric unit.
3.3. Crystal Structure of 1,3,5-Tris(2,2-dimethyl-
propionylamino)benzene. For developing a structure
model, compound 1 was investigated by X-ray powder
diffraction and solid-state NMR. Indexing of the powder
pattern leads unequivocally to an orthorhombic metric which
could subsequently be refined using the Pawley algorithm (a =
14.79(3) Å, b = 23.85(5) Å, c = 6.75(1) Å, V = 2381(13) Å3,
wRp = 2.6%). Assuming that carbon, nitrogen, and oxygen
atoms possess a volume of roughly 18 Å3, the molecular volume
of compound 1 can be estimated to 490 Å3. Including the
hydrogen atoms would even increase this volume. This size
restriction just allows at maximum four molecules within the
unit cell. The reflection conditions, however, provide
Figure 2. 1D MAS spectra of 1,3,5-tris(2,2-dimethylpropionylamino)-
benzene: (a) H spectrum (νrot = 65 kHz, B0 = 11.7 T), with signal
assignment for 1 given in Figure 1, the asterisk marks an impurity; (b)
15N CP-MAS spectrum (νrot = 5 kHz); (c) 13C CP-MAS spectrum
(νrot = 12.5 kHz, with signal assignment on top).
1
groups. In comparison to the isotropic chemical shift in
solution-state, the resonance of the amide protons is shifted
downfield of about 0.3 ppm. The aromatic protons, on the
other hand, show an upfield shift of roughly 1 ppm (Figure 2a).
Both aspects indicate the existence of packing effects like
hydrogen bonds and π−π interactions due to aromatic ring
9
1
currents. Furthermore, the single sharp H resonance of the
NH-group suggests a quite narrow distribution of the hydrogen
bond lengths. For similar systems, i.e., symmetric 1,3,5-
benzenetrisamides (“N-centered”- based on 1,3,5-triaminoben-
zene and “C-centered” - based on trimesic acid), such effects
have recently been reported and verified with solid-state NMR9
and quantum chemical calculations9 as well as temperature
depending infrared spectroscopy.75
The upfield shift of the aromatic protons of 1.2 ppm is in the
1
same range as the H chemical shift presented in ref 9 for the
aromatic protons which points to a parallel arrangement of
benzene rings within a columnar packing. The π−π stacking
within these columns with intermolecular distances of roughly
3.5 Å leads to an additional shielding of the protons due to
strong ring currents of neighboring molecules. In our 1H
spectrum, we find one resonance for the aromatic protons.
Therefore, compound 1 seems to possess a coplanar arrange-
ment of the benzene rings, ideally with a helical rod stacking
which would maintain C3 symmetry of the molecules including
a high symmetry within the crystal structure. However, this is in
contrast to the N-centered tricarboxamide recently investigated
by Wegner and co-workers.9 Indeed, they reported about
significant upfield shifts for the aromatic proton signals due to
the ring current effect like in our case, but in contrast, they
observed three distinct signals for the aromatic resonances for
N-BTA 3 (N-centered benzene-1,3,5-tricarboxamide with (S)-
2,6-dimethylheptyl side chains). They explain their observance
with a symmetry-breaking within the local packing motif. For
their system it is possible that one carbonyl group of a molecule
points in the opposite direction. This new stacking order can be
2546
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551

Crystal Growth & Design
Article
Figure 3. 13C CP-MAS signal of the carbonyl group of 13C enriched
1,3,5-tris(2,2-dimethylpropionylamino)benzene (νrot = 12.5 kHz; B0 =
16.4 T) with deconvolution into three distinct 13C resonances with an
intensity ratio of 1.02(2):1.00(2):1.01(2). For all three resonances the
same restriction for the FWHM was used.
Figure 4. Rietveld profile plot of the X-ray powder diffraction pattern
for the solution in space group P212121 collected at room temperature
in a 2θ range 5−40°.
reasonable results for 19 primitive and six body-centered space
groups. On the basis of the nonplanarity of the molecule itself
and the absence of an intrinsic C2 symmetry, they cannot be
placed on special positions in the unit cell. Therefore, only
space groups possessing a multiplicity of 4 on general positions
have to be taken into consideration for a further investigation.
Nevertheless, 11 primitive space groups remain for a possible
structure elucidation.
carbonyl groups were taken into consideration (see Table 2).
While the intramolecular distances for all four space groups
essentially show the same values, the intermolecular ones can
For all these space groups an ab initio structure solution with
real-space methods was performed using a combination of a
close-contact penalty and wRp as cost functions. The DFT
geometry optimized molecule was taken as rigid body with the
rotational and translational degrees of freedom during the
solution step. Additionally, the six most important and structure
directing torsion angles (two in each side chain, see Figure 1,
right) were refined during the optimization process. At the end,
only four space groups remain possessing a comparable low
wRp value well below 10% (see Table 1), whereas all other
Table 1. wRp Values of the Four Most Probable Space
Groups after the Powder Solution Step and Rietveld
Refinement
space group
solution wRp/%
refinement wRp/%
P212121
Pna21
P2an
5.73
5.96
7.49
7.16
4.18
4.57
5.19
4.37
P21cn
structure solutions possess wRp values far above 10%. An initial
Rietveld refinement of these space groups leads to similar wRp
values which are listed in Table 1. Exemplarily, the Rietveld plot
for space group P212121 is depicted in Figure 4.
Selecting the correct structure model just from the observed
structures from the powder data is difficult since the wRp values
are almost equal. All four structures show a similar behavior
concerning their topology where the molecules build a
pseudohexagonal rod packing along the c-axis. The major
difference is the arrangement of the molecules within the
packing pattern that can be distinguished in hydrogen-bonding
(P212121 and Pna21) and non-hydrogen-bonding (P2an and
P21cn) patterns (see Figure 5).
Figure 5. Top view on one rod of the different packing patterns where
the NH-protons are emphasized. Top: pattern for the space groups
P2an and P21cn without hydrogen bonding. Bottom: pattern for the
space groups P212121 and Pna21 with a hydrogen bonding network. All
other protons are omitted for clarity.
To further distinguish the structure models the intermo-
lecular distances between the 13C labeled carbon atoms of the
2547
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551

Crystal Growth & Design
Article
Table 2. Intramolecular and Intermolecular 13C−13C
Distances between Labeled Carbonyl Groups from a Nine
Spin System Used as Input for the Simulations
therefore badly reproduce the experiment. Due to the shorter
intermolecular distances (see Table 2) the slope of the
simulated curves is stronger leading to a maximum excitation
time of roughly τ = 14 ms. In fact, the simulated curves for
P212121 and Pna21 match the experimental build-up curve very
well on the basis of the longer 13C−13C distances and the
narrower distribution of the distances between the 13C atoms in
the spin system. However, the best fit could be achieved for
space group P212121 (see rms values, Figure 6a). On the basis
of the higher wRp and rms values as well as the fact that
peculiar gliding planes (like in Pna21) are uncommon for
molecular compounds which prefer crystallizing in rather
simple lattices and point groups, the structure model in space
group Pna21 can definitely be excluded.
1,3,5-Tris(2,2-dimethylpropionylamino)benzene therefore
crystallizes in the orthorhombic noncentrosymmetric space
group P212121 (a = 14.91(5) Å, b = 24.05(8) Å, c = 6.80(2) Å)
containing four molecules in the unit cell, all on general
positions. The wRp of 4.18% is a very good value for an ab
initio structure solution picturing the high order of the sample
despite the presence of only light atoms leading to weak
scattering.
distance/Å
P212121
Pna21
P21cn
P2an
intramolecular 6.622−6.628 6.586−6.601 6.634−6.715 6.633−6.654
intermolecular 4.927−5.259 4.983−5.229 4.616−5.740 4.189−6.105
clearly be distinguished. For the space groups possessing a
hydrogen bond network the minimum and maximum distances
are almost equal (∼ 4.9 and ∼5.2 Å), and moreover, the length
distribution is quite narrow. On the other hand, both space
groups without H-bonds show a significantly different behavior.
The minimum distances amount to 4.19 and 4.62 Å,
respectively, while the maximum distance is drastically extended
up to 6.1 Å. Thus, the length distribution is much broader than
in the first case.
To support the structure selection we therefore performed
quantitative solid-state NMR experiments. For this purpose we
carried out 13C13C DQ experiments taking advantage of the
dipolar coupling between the carbonyl atoms using the SR26411
pulse sequence (see section 2.3). The constant thereby is
proportional to r−3 where r represents the 13C−13C distance.
The normalized experimental build-up curve of the carbonyl
resonance is depicted in Figure 6b.
It shows a very slow increase with a quite broad maximum at
about τ = 17 ms, and the double quantum efficiency reaches
roughly 50%. For simulating the experimental build-up curve
we cut out a characteristic element of one rod of each crystal
structure after the Rietveld refinements. Hence, the generated
spin system had to be sufficiently large to avoid finite size
effects and small enough to be handled with reasonable
computation time. A nine spin system (Figure 6c) turned out
to be the best choice. Here, each triangle represents one
molecule within one rod of the structure, while the black circles
illustrate the labeled carbonyl groups. The nine spin system
used for our simulations is hence spanned by the three carbonyl
spins of each molecule.
The molecular structure after the Rietveld refinement can be
seen in Figure 7. The CArCArNHCO torsion angles seem to play
a major role for the formation of the structure. Two side arms
exhibit almost identical values for this torsion angle (31.36° and
31.46°, respectively), while the third one is a little bit higher
(34.56°). This fact in addition to the deconvoluted 13CO
signal of 1* (Figure 3) slightly breaks the perfect regularity;
1
however, a pseudo-C3 symmetry remains as assumed from H
and 15N solid-state NMR experiments. The HNCO torsion
angles picturing the amide bonds are 178.78°, 178.93°, and
179.66°. This is in very good agreement with the expected
value of 180° due to the behavior as partial double bonds. The
oxygen atoms of the three carbonyl groups in one molecule
point in the same direction out of the plane of the central
aromatic ring system which is a precondition to form hydrogen
bond patterns within the structure (Figure 8).
In Figure 6a, the root-mean-square errors (rms) between
experiment and simulation (see section 2.4) are given for the
four remaining space groups. It can be distinguished between
space groups containing hydrogen bonds and without hydrogen
bonds. While the latter possess rms values of 1.1 and 2.9,
respectively, the former shows significantly lower values of 0.25
and 0.35. The simulated build-up curves for P2an and P21cn
́
In 2009, Jimenez et al. reported the structure of a “C-
centered” BTA (connected with the carbonyl carbon atom to
the benzene ring) carrying a propyl moiety in each side arm
instead of a t-butyl group.30 The molecules within the crystal
packing build up a supramolecular 3D H-bond network within
a primitive cubic metric. However, in contrast to 1, π-stackings
are avoided with the linear side arms. Also in 2009, Kristiansen
Figure 6. (a) Histogram with root-mean-square deviation between experimental and simulated symmetric build-up curve of all four space groups.
(b) Experimental (orange plus) and simulated (solid lines) symmetric 13C13C SR26 build-up curves of 1* after Rietveld refinement. (c) Scheme of a
nine spin system for all simulations where each triangle represents one molecule within the packing of the structures. The 13C spins (three in each
molecule) are marked as circles; intermolecular distances are highlighted with dotted lines.
2548
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551

Crystal Growth & Design
Article
1
from H and 15N solid-state NMR. The corresponding N−O
distances account for 2.97−3.06 Å. In addition to the H-bond
network the helical structure is characterized by sandwich
stacked π-systems within one rod. The aromatic rings and
therefore the molecules are stacked in a coplanar arrangement
where the aromatic rings are tilted less than 1°. Typical
distances of the benzene carbon atoms of two molecules are
between 3.37 and 3.42 Å which is similar to the distances in π-
stacked discotic liquid crystals.76
In comparison, the structure presented in ref 31 crystallizes
in a hexagonal unit cell where two molecules are sited on
special positions. The structure is also characterized by an
extensive hydrogen bonding network within the rod packing as
well as a helical arrangement. The N−O distance with 2.65 Å is
roughly 0.3 Å shorter than for compound 1 indicating stronger
hydrogen bonds. This is also supported by the increased jut of
the oxygen atoms from the aromatic ring system in comparison
to compound 1. Due to the 63 screw axis the molecules are also
twisted around 60°. The sandwich-like π−π contacts picture
the second structure directing interaction whereupon the
distances between two benzene rings amount to 3.45 and
3.47 Å, respectively, which is in the same range as that observed
for compound 1. Surprisingly, the synchrotron X-ray pattern of
this “C-centered” trisamide reveals an immense diffuse
scattering. This pictures a structural disorder where the two
rods within the unit cell can be slid against each other. For the
“N-centered” trisamide, on the other hand, no diffuse scattering
was observed during investigation indicating a high order
within the crystal structure.
Figure 7. Molecular structure of 1,3,5-tris(2,2-dimethyl-
propionylamino)benzene with distances in angstroms as derived
from the X-ray powder diffraction experiment.
The molecules of compound 1 and the ones from ref 31
remain nearly C3 symmetric where all carbonyl groups point in
one direction. Due to this out-plane arrangement of the oxygen
atoms the molecules itself possess a quite large dipolar moment
leading to the formation of macrodipoles for individual rods of
the packing. One common mechanism of choice to reduce this
moment within the unit cell and with it the potential energy is
to form an antiferroelectric arrangement.77 For compound 1,
indeed, the molecules in neighbored columns show in the
opposite direction (Figure 9) canceling out the net dipolar
Figure 8. Section of the crystal structure of 1,3,5-tris(2,2-
dimethylpropionylamino)benzene with view along the c-axis. The
unit cell contains four molecules on general positions.
et al. reported about the crystal structure of a similar 1,3,5-
benzenetrisamide (compared to compound 1) on the basis of
transmission electron microscopy and electron and X-ray
diffraction studies,31 carrying the same moieties in the side arms
(t-butyl), but, in contrast, with a “C-centering” of the amide
group. Here the molecules are reported to possess a complete
C3 symmetry, and moreover, all three oxygen atoms of the
carbonyl groups show in one direction with CArCArNHCO
torsion angles of 53.21° each of which is roughly 20° higher
than for our benzene based trisamide.
The crystal structure of 1 can be described as hexagonal rod-
packing where the molecules build long rods parallel to the c-
axis (see Figure 8). Due to the 21 screw axis and the additional
pseudo-C3 symmetry, the molecules are twisted around 60°
within one rod leading to a helical arrangement which has also
been reported for similar benzene based supramolecular
systems.31 An extensive hydrogen bond network is built
between the amide protons and the oxygen atoms of the
neighboring molecules in one rod. The NH···O distances with
2.04−2.14 Å, picturing medium strong H-bonds, are narrowly
distributed which is in very good agreement with the results
Figure 9. Part of the crystal structure of 1 with view along the b-axis.
The oxygen atoms of the carbonyl groups point in the opposite
direction leading to an averaging of the net dipolar moment. All
protons and the carbon atoms of the t-butyl groups were omitted for
clarity.
moment. Due to avoid frustration while canceling the dipole
moment, compound 1 undergoes an orthorhombic distortion
(“pseudohexagonal”) in contrast to the BTA reported by
Kristiansen et al. where all rods are aligned parallel.
Consequently, the net dipolar moment increases drastically
leading to a complete ferroelectric arrangement. Interestingly,
2549
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551

Crystal Growth & Design
Article
the structure of compound 1 in P212121 could easily be
transferred into the one of Pna21 by twisting the CArCArNHCO
torsion angle (see Figure 1) in each side chain of every second
rod by roughly 60° leading also to a ferroelectric arrangement
analogous to that of ref 31.
ASSOCIATED CONTENT
■
S
* Supporting Information
Spin system extracted from the structure solution as input file
for SIMPSON used for the simulation of the DQ build-up
curve as well as a CIF file of 1 including the crystallographic
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
4. CONCLUSIONS
We have presented an approach for the determination of the
crystal structure of 1,3,5-tris(2,2-dimethylpropionylamino)-
benzene, an efficient clarifying agent for isotactic polypropy-
lene. The structure was solved in the orthorhombic space group
P212121 using X-ray powder diffraction in combination with
multinuclear solid-state NMR experiments and computer
simulations. During structure elucidation, only four space
groups led to reasonable structure models. DQ experiments
allowed excluding the three remaining space groups. Using the
AUTHOR INFORMATION
■
Corresponding Author
*Fax: (+49) 921-55-3206 (H.-W.-S.); (+49) 921-55-2788
(J.S.). E-mail: hans-werner.schmidt@uni-bayreuth.de (H.-
W.S.); juergen.senker@uni-bayreuth.de (J.S.).
Notes
The authors declare no competing financial interest.
symmetry-based SR2611 pulse sequence, 13C−13C distances of
4
ACKNOWLEDGMENTS
■
nearly 5 Å could be reproduced with high precision. The
simulation of the build-up curves fitted best for the space group
P212121 with a root-mean-square deviation of 0.25. The c-axis
with a length 6.80(2) Å is in this connection in the similar
range with the c-axis of the crystalline part of i-PP (∼6.50 Å)
which possesses a helical structure as well.81 The additive
crystallites within the polymer melt therefore might serve as a
surface which allows the polymer to grow epitactically. This
remarkable fact may cause i-PP to appear highly transparent.
Therefore, further investigations in the additive/polymer
composite systems are mandatory and will be discussed in a
forthcoming publication.
The authors want to thank the DFG (SFB 840, Project B4) for
financial support of this work, Sandra Ganzleben for her careful
assistance with the synthesis of the carbon-enriched compound,
and Dr. Wolfgang Milius for the fruitful discussions with
respect to crystallographic issues.
REFERENCES
■
(1) Matsunaga, Y.; Nakayasu, Y.; Sakai, S.; Yonenaga, M. Mol. Cryst.
Liq. Cryst. 1986, 141, 327.
(2) Matsunaga, Y.; Miyajima, N.; Nakayasu, Y.; Sakai, S.; Yonenaga,
M. Bull. Chem. Soc. Jpn. 1988, 61, 207.
(3) Brunsveld, L.; Schenning, A. P. H. J.; Broeren, M. A. C.; Janssen,
H. M.; Vekemans, J. A. J. M.; Meijer, E. W. Chem. Lett. 2000, 292.
(4) Xue, C.; Ilhan, F.; Cheng, S. Z.; Meador, M. A.; Eby, R. Polym.
Prepr. 2004, 45, 820.
(5) Stals, P. J. M.; Smulders, M. M. J.; Martin-Rapun, R.; Palmans, A.
R. A.; Meijer, E. W. Chem. Eur. J. 2009, 15, 2071.
(6) Stals, P. J. M.; Everts, J. C.; de Bruijn, R.; Filot, I. A. W.;
Smulders, M. M. J.; Martin-Rapun, R.; Pidko, E. A.; de Greef, T. F. A.;
Palmans, A. R. A.; Meijer, E. W. Chem. Eur. J. 2010, 16, 810.
(7) Kluge, D.; Abraham, F.; Schmidt, S.; Schmidt, H.-W.; Fery, A.
Langmuir 2010, 26, 3020.
(8) Besenius, P.; van den Hout, K. P.; Albers, H. M. H. G.; de Greef,
T. F. A.; Olijve, L. L. C.; Hermans, T. M.; de Waal, B. F. M.; Bomans,
P. H. H.; Sommerdijk, N. A. J. M.; Portale, G.; Palmans, A. R. A.; van
Genderen, M. H. P.; Vekemans, J. A. J. M.; Meijer, E. W. Chem. Eur. J.
2011, 17, 5193.
(9) Wegner, M.; Dudenko, D.; Sebastiani, D.; Palmans, A. R. A.; de
Greef, T. F. A.; Graf, R.; Spiess, H. W. Chem. Sci. 2011, 2, 2040.
(10) Shi, N.; Dong, H.; Yin, G.; Xu, Z.; Li, S. Adv. Funct. Mater. 2007,
17, 1837.
(11) Bernet, A.; Albuquerque, R. Q.; Behr, M.; Hoffmann, S. T.;
Schmidt, H.-W. Soft Matter 2012, 8, 66.
(12) Yasuda, Y.; Iishi, E.; Inada, H.; Shirota, Y. Chem. Lett. 1996, 25,
575.
(13) Hanabusa, K.; Koto, C.; Kimura, M.; Shirai, H.; Kakehi, A.
Chem. Lett. 1997, 26, 429.
(14) van Gorp, J. J.; Vekemans, J. A. J. M.; Meijer, E. W. J. Am. Chem.
Soc. 2002, 124, 14759.
(15) de Loos, M.; van Esch, J. H.; Kellogg, R. M.; Feringa, B. L.
Tetrahedron 2007, 63, 7285.
(16) Ogata, D.; Shikata, T.; Hanabusa, K. J. Phys. Chem. B 2004, 108,
15503.
The structure itself can be described as a hexagonal rod
packing where the pseudo-C3 symmetric molecules are twisted
by 60° due to the 21 screw axis. The rods are stabilized via
moderate hydrogen bonds with NH···O distances of roughly
2.0 Å and, moreover, via π-interactions due to a sandwich
stacking of the aromatic cores with distances of about 3.4 Å.
Each molecule possesses an intrinsic dipolar moment due to
the fact that all oxygen atoms point in one direction. Therefore,
the dipolar moment in one rod strongly increases. This H-bond
mediated growth within one rod is in principle possible since
the increasing potential energy is counterbalanced by the
energy advantage based on the formation of hydrogen bonds.
Moreover, the occurrence of such macrodipoles is not
completely unknown within the BTA systems. Such phenom-
ena have recently been demonstrated by Kulkarni et al.80 on the
basis of quantum chemical calculations. In addition, exper-
imental proof was given by Sakamoto et al. where benzene-
1,3,5-tricarboxamides with long side chains formed one-
dimensional macrodipoles in solution of n-decane leading to
highly viscous liquids.78 In thin films it was even possible to
generate macrodipoles in a three-dimensional ordered system.79
However, a ferroelectric arrangement could not be achieved per
se rather than by applying an external electrical field. Obtaining
ferroelectricity by crystallization from solution seems to be
improbable. Once a macrodipole in one rod is built, the second
one must be arranged antiparallel since parallel formation
strongly increases the potential energy of the system that
cannot be compensated by the 1D H-bond network. In this
case only small nanocrystals would be observed. Therefore, an
antiferroelectric arrangement with antiparallel aligned rods as
observed for compound 1 leading to a net dipolar moment of
zero represents the driving force for the fast needlelike crystal
growth of this class of materials.
(17) Smulders, M. M. J.; Schenning, A. P. H. J.; Meijer, E. W. J. Am.
Chem. Soc. 2008, 130, 606.
(18) Smulders, M. M. J.; Stals, P. J. M.; Mes, T.; Paffen, T. F. E.;
Schenning, A. P. H. J.; Palmans, A. R. A.; Meijer, E. W. J. Am. Chem.
Soc. 2010, 132, 620.
2550
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551

Crystal Growth & Design
Article
(19) Smulders, M. M. J.; Filot, I. A. W.; Leenders, J. M. A.; van der
Schoot, P.; Palmans, A. R. A.; Schenning, A. P. H. J.; Meijer, E. W. J.
Am. Chem. Soc. 2010, 132, 611.
(20) Kreger, K.; Wolfer, P.; Audorff, H.; Kador, L.; Stingelin-
Stutzmann, N.; Smith, P.; Schmidt, H.-W. J. Am. Chem. Soc. 2010, 132,
509.
(54) Wirnhier, E.; Doblinger, M.; Gunzelmann, D.; Senker, J.;
̈
Lotsch, B. V.; Schnick, W. Chem. Eur. J. 2011, 17, 3213.
(55) auf der Guenne, J. S. J. Magn. Reson. 2003, 165, 18.
(56) Carravetta, M.; Eden, M.; Johannessen, O. G.; Luthman, H.;
Verdegem, P. J. E.; Lugtenburg, J.; Sebald, A.; Levitt, M. H. J. Am.
Chem. Soc. 2001, 123, 10628.
(57) Graf, R.; Demco, D. E.; Gottwald, J.; Hafner, S.; Spiess, H. W. J.
Chem. Phys. 1997, 106, 885.
(58) van Rossum, B. J.; de Groot, C. P.; Ladizhansky, V.; Vega, S.; de
Groot, H. J. M. J. Am. Chem. Soc. 2000, 122, 3465.
(59) Goward, G. R.; Schnell, I.; Brown, S. P.; Spiess, H. W.; Kim, H.-
D.; Ishida, H. Magn. Reson. Chem. 2001, 39, 5.
(60) Dutour, J.; Guillou, N.; Huguenard, C.; Taulelle, F.; Mellot-
́
Draznieks, C.; Ferey, G. Solid State Sci. 2004, 6, 1059.
(61) Brouwer, D. H.; Kristiansen, P. E.; Fyfe, C. A.; Levitt, M. H. J.
Am. Chem. Soc. 2005, 127, 542.
(62) Harris, R. K.; Ghi, P. Y.; Hammond, R. B.; Ma, C. Y.; Roberts,
K. J. Chem. Commun. 2003, 2834.
(63) Elena, B.; Pintacuda, G.; Mifsud, N.; Emsley, L. J. Am. Chem.
Soc. 2006, 128, 9555.
(64) Pickard, C. J.; Salager, E.; Pintacuda, G.; Elena, B.; Emsley, L. J.
Am. Chem. Soc. 2007, 129, 8932.
(65) Manolikas, T.; Herrmann, T.; Meier, B. H. J. Am. Chem. Soc.
2008, 130, 3959.
(66) Loquet, A.; Bardiaux, B.; Gardiennet, C.; Blanchet, C.; Baldus,
M.; Nilges, M.; Malliavin, T.; Boeckmann, A. J. Am. Chem. Soc. 2008,
130, 3579.
(67) Middleton, D. A.; Peng, X.; Saunders, D.; Shankland, K.; David,
W. I. F.; Markvardsen, A. J. Chem. Commun. 2002, 1976.
(68) Brouwer, D. H.; Darton, R. J.; Morris, R. E.; Levitt, M. H. J. Am.
Chem. Soc. 2005, 127, 10365.
(69) Fyfe, C. A.; Brouwer, D. H. J. Am. Chem. Soc. 2006, 128, 11860.
(70) Parnes, H.; Dekeczer, S. J. Labelled Compd. Radiopharm. 1985,
22, 287.
(71) MS Modeling v5.0.0.0; Accelrys Software Inc., 2009.
(72) Kristiansen, P. E.; Carravetta, M.; Lai, W. C.; Levitt, M. H.
Chem. Phys. Lett. 2004, 390, 1.
(21) Mohmeyer, N.; Muller, B.; Behrendt, N.; Hillenbrand, J.;
̈
Klaiber, M.; Zhang, X.; Smith, P.; Altstadt, V.; Sessler, G. M.; Schmidt,
̈
H.-W. Polymer 2004, 45, 6655.
(22) Erhard, D. P.; Lovera, D.; von Salis-Soglio, C.; Giesa, R.;
Altstaedt, V.; Schmidt, H.-W. Adv. Polym. Sci. 2010, 228, 155.
(23) Abraham, F.; Schmidt, H.-W. Polymer 2010, 51, 913.
(24) Blomenhofer, M.; Ganzleben, S.; Hanft, D.; Schmidt, H.-W.;
Kristiansen, M.; Smith, P.; Stoll, K.; Maeder, D.; Hoffmann, K.
Macromolecules 2005, 38, 3688.
(25) Kristiansen, P. M.; Gress, A.; Smith, P.; Hanft, D.; Schmidt, H.-
W. Polymer 2006, 47, 249.
(26) Abraham, F.; Ganzleben, S.; Hanft, D.; Smith, P.; Schmidt, H.-
W. Macromol. Chem. Phys. 2010, 211, 171.
(27) Bernland, K. Nucleating and Clarifying Polymers. Ph.D. Thesis,
ETH, Zurich, 2010.
̈
(28) Gahleitner, M.; Grein, C.; Kheirandish, S.; Wolfschwenger, J.
Int. Polym. Proc. 2011, 26, 2.
(29) Lightfoot, M. P.; Mair, F. S.; Pritchard, R. G.; Warren, J. E.
Chem. Commun. 1999, 1945.
(30) Jimenez, C. A.; Belmar, J. B.; Ortiz, L.; Hidalgo, P.; Fabelo, O.;
Pasan, J.; Ruiz-Perez, C. Cryst. Growth Des. 2009, 9, 4987.
(31) Kristiansen, M.; Smith, P.; Chanzy, H.; Baerlocher, C.;
Gramlich, V.; McCusker, L.; Weber, T.; Pattison, P.; Blomenhofer,
M.; Schmidt, H.-W. Cryst. Growth Des. 2009, 9, 2556.
(32) Cerny, R. Croat. Chem. Acta 2006, 79, 319.
(33) Harris, R. K. J. Pharm. Pharmacol. 2007, 59, 225.
(34) David, W. I. F.; Shankland, K. In Structure Determination from
Powder Diffraction Data; McCusker, L., Baerlocher, C., Eds.; IUCr
Monographs on Crystallography; Oxford University Press: Oxford,
U.K., 2002.
(35) Harris, K. D. M.; Tremayne, M.; Kariuki, B. M. Angew. Chem.,
Int. Ed. 2001, 40, 1626.
(73) Bak, M.; Rasmussen, J. T.; Nielsen, N. C. J. Magn. Reson. 2000,
147, 296.
(74) The Language of Technical Computing. MATLAB; Math Works
Inc.: Natick, MA, 1984−2012.
(75) Timme, A.; Kress, R.; Albuquerque, R. Q.; Schmidt, H.-W.
Chem. Eur. J. 2011, accepted.
(36) Harris, K. D. M. Cryst. Growth Des. 2003, 3, 887.
(37) Tremayne, M. Phil. Trans. R. Soc., A 2004, 362, 2691.
(38) Harris, K. D. M.; Cheung, E. Y. Chem. Soc. Rev. 2004, 33, 526.
(39) Harris, R. K. Analyst 2006, 131, 351.
(76) Chandrasekhar, S.; Sadashiva, B. K.; Suresh, K. Pramana 1977,
9, 471.
(40) Hirsemann, D.; Koster, T. K.-J.; Wack, J.; van Wullen, L.; Breu,
J.; Senker, J. Chem. Mater. 2011, 23, 3152.
(41) Duer, M. J. Introduction to Solid-State NMR Spectroscopy;
Blackwell Publishing: Oxford, U.K., 2004.
(42) Karau, F. W.; Seyfarth, L.; Oeckler, O.; Senker, J.; Landskron,
K.; Schnick, W. Chem. Eur. J. 2007, 13, 6841.
(43) Lotsch, B. V.; Senker, J.; Kockelmann, W.; Schnick, W. J. Solid
State Chem. 2003, 176, 180.
(44) Senker, J.; Seyfarth, L.; Voll, J. Solid State Sci. 2004, 6, 1039−
1052.
̈
̈
(77) Senker, J.; Ludecke, J. Z. Naturforsch. 2001, b, 1089.
̈
(78) Sakamoto, A.; Ogata, D.; Shikata, T.; Urakawa, O.; Hanabusa, K.
Polymer 2006, 47, 956.
(79) Fitie, C. F. C.; Roelofs, W. S. C.; Kemerink, M.; Sijbesma, R. P.
́
J. Am. Chem. Soc. 2010, 132, 6892.
(80) Kulkarni, C.; Reddy, S. K.; George, S. J.; Balasubramanian, S.
Chem. Phys. Lett. 2011, 515, 226.
(81) Phillips, R.; Wolkowicz, M. D. In Polypropylene Handbook, 2nd
ed.; Pasquini, N., Ed.; Hanser: Cincinnati, 2005; pp 147−264.
(45) Senker, J.; Sehnert, J.; Correll, S. J. Am. Chem. Soc. 2005, 127,
337−349.
(46) Frydman, L. In Fundamentals of Multiple-Quantum Magic-Angle
Spinning NMR on Half-Integer Quadrupolar Nuclei; Grant, D., Ed.; John
Wiley & Sons. Ltd.: Chichester, U.K., 2002; Vol. 9, pp 262−274.
(47) King, I. J.; Fayon, F.; Massiot, D.; Harris, R. K.; Evans, J. S. O.
Chem. Commun. 2001, 1766.
(48) Bestaoui, N.; Ouyang, X.; Fredoueil, F.; Bujoli, B.; Clearfield, A.
Acta Crystallogr. Sect. B 2005, 61, 669.
(49) Taulelle, F. Solid State Sci. 2004, 6, 1053.
(50) Kentgens, A. Geoderma 1997, 80, 271.
(51) Smith, M.; van Eck, E. Prog. Nucl. Magn. Reson. Spectrosc. 1999,
34, 159−201.
(52) Seyfarth, L.; Senker, J. Phys. Chem. Chem. Phys. 2009, 11, 3522.
(53) Seyfarth, L.; Seyfarth, J.; Lotsch, B. V.; Schnick, W.; Senker, J.
Phys. Chem. Chem. Phys. 2010, 12, 2227.
2551
dx.doi.org/10.1021/cg300151s | Cryst. Growth Des. 2012, 12, 2543−2551