Synthesis and structural analysis of triphenylmethane-based alkanecarboxamides and their assembly into nanometer-size fibrous objects

Article abstract of DOI:10.1246/bcsj.82.730

A series of C₃-symmetric trialkanecarboxamido derivatives was synthesized via a triphenylmethane derivative. The triamides with short alkyl chains (16 carbons) crystallized as either prisms or plates, whereas those with long alkyl chains (711 c

Full text of DOI:10.1246/bcsj.82.730

730
Bull. Chem. Soc. Jpn. Vol. 82, No. 6, 730–736 (2009)
© 2009 The Chemical Society of Japan
Synthesis and Structural Analysis of Triphenylmethane-Based
Alkanecarboxamides and Their Assembly into
Nanometer-Size Fibrous Objects
Hirohiko Houjou,* Tatsuya Koga, Midori Akiizumi, Isao Yoshikawa, and Koji Araki
Institute of Industrial Science, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505
Received January 7, 2009; E-mail: houjou@iis.u-tokyo.ac.jp
A series of C3-symmetric trialkanecarboxamido derivatives was synthesized via a triphenylmethane derivative. The
triamides with short alkyl chains (1-6 carbons) crystallized as either prisms or plates, whereas those with long alkyl
chains (7-11 carbons) formed gel-like objects composed of fibrous crystals with 100-nm diameters. A discontinuity as a
function of alkyl chain length was also observed for the melting points, the vibrational frequencies of the amide groups,
and the powder X-ray diffraction patterns of the triamides. On the basis of analysis of the crystal structures of two of the
compounds, we make some proposals about the molecular packing and the local structure of the molecular assembly. Our
results suggest that there were critical differences in inter-core interactions and in the hydrogen-bonding network,
depending on the length of the alkyl chains. We interpret the drastic change in morphology brought about by the tuning of
the chain length as a consequence of a change in the mode of molecular assembly.
Systems of multiple hydrogen bonds serve as reversible and
could control inter-conversion between the core-core interac-
tion mode and the hydrogen-bonded mode.14
flexible cross links in supramolecular polymers and supra-
molecular elastomers as well as in highly organized biomo-
lecular structures such as proteins and nucleic acids.1,2 The
reversibility and moderate strength of hydrogen-bonded cross
links in these structures are responsible for such characteristic
properties as thermal plasticity, viscoelasticity, and the dielec-
tric property.3 Introduction of a network of hydrogen bonds
provides supramolecular polymers with practical properties,
such as mechanical strength, thermal durability, and non-
Newtonian fluidity.4-7
Molecular building blocks that have the potential to form
multiple hydrogen bonds are known as supramolecular syn-
thons.8 The greater the number of hydrogen bonds a synthon
can form, the firmer the resulting supramolecular architecture
will be, as a result of the formation of multiple cross links.
Many of the supramolecular synthons successfully used to date
in, for example, columnar liquid crystals, organogels, fibers,
and films,9-12 have a common feature: they are Cn-symmetric
molecules composed of a rigid core with self-complementary
hydrogen-bonding sites and flexible side chains. Because of its
energetic stability, hydrogen bonding plays a predominant role
in the mode of molecular assembly, and inter-core and inter-
chain interactions assist in arranging the architecture.
Systems with aromatic cores prefer an inter-core distance of
3.3-3.5 ¡ owing to ³-³ stacking interactions, such as those
typically found in graphite. However, a mismatch in distance
of stacking cores sometimes arises, because the ³-³ stacking
distance does not always permit the hydrogen-bonding sites to
connect to one another. Takasawa et al. utilized a non-aromatic
cyclohexane moiety as a core to fabricate supramolecular yarns
in which the inter-core and hydrogen-bonding distances were
optimized simultaneously.13 With an aromatic core, pyrene, the
same research group obtained a bistable system, in which they
The exploration of new molecules with various aromatic and
non-aromatic cores bearing multiple hydrogen-bonding sites is
a fascinating area of research. In this study, we designed and
synthesized C3-symmetric molecules composed of a triphenyl-
methane core, three amide groups, and alkyl chains with 1 to 11
carbon atoms. The unsubstituted triphenylmethane framework
is not planar but parasol shaped, owing to the central sp3 carbon
atom. However, a triphenylmethane with electron-donating
groups easily loses its methyne hydrogen by oxidation, and this
oxidation results in a virtually planar methylium ³-system,
which is responsible for the vivid colors of synthetic dyes such
as crystal violet and malachite green.15 The use of a redox-
active moiety in a supramolecular system has been shown to
lead to such interesting functions as sol-gel conversion that is
controlled by oxidation state.16
Although triphenylmethane and its related methylium frame-
work can be obtained easily from the above-mentioned
artificial dyes, the use of triphenylmethanes in supramolecular
chemistry has been rather limited,17-23 and little is known about
their molecular assembly. Recently, we found that triphenyl-
methane triamides with long alkyl chains can form a gel-like
precipitate in certain solvent systems and that the precipitate is
actually a collection of micrometer-thick fibrous objects. In this
paper, we describe the synthesis and structural analyses of a
series of triphenylmethane trialkanecarboxamides, focusing on
the how the structures of their molecular assemblies depended
on the length of the alkyl chains.
Results and Discussion
Triphenylmethane-based trialkanecarboxamides 1a-1k were
prepared in moderate yield and then characterized (Scheme 1).
The products were roughly categorized into groups by
Published on the web June 12, 2009; doi:10.1246/bcsj.82.730

H. Houjou et al.
Bull. Chem. Soc. Jpn. Vol. 82, No. 6 (2009)
731
morphology: 1a-1f (1-6 carbons per chain) were obtained as
either prismatic or platelet microcrystals, whereas 1g-1k (7-11
carbons per chain) were obtained as thin needles. By examining
several recrystallization conditions for these compounds, we
found that 1g-1k, the group with the longer alkyl chains,
formed gel-like objects when recrystallized from mixtures of
moderately polar (e.g., THF and ethyl acetate) and protic (e.g.,
methanol and ethanol) solvents. Similar objects also formed
when solutions of 1g-1k in certain hydrophobic solvents (e.g.,
chloroform and dichloromethane) were emulsified by vigorous
shaking with a water phase. Figures 1a and 1b show typical
optical and scanning electron microscope images, respectively,
of these objects. The images indicate that the gel-like objects
were actually micrometer-thick collections of tangled fibers,
with each fiber consisting of a bunch of thinner fibers with
thicknesses of 100-200 nm.
The drastic change in morphology observed as a result of the
transition from a short side chain to a long side chain suggests
differences in the structures of the molecular assemblies
originating in hydrogen bonding among the amide moieties
and in inter-core interactions between the triphenylmethane
moieties. To obtain insight into the structure of the molecular
assemblies, we measured melting points, infrared spectra, and
powder X-ray diffraction patterns of 1a-1k. In the preparation
of the samples for these measurements, we standardized the
recrystallization system (DMF-ethanol-water) to minimize the
effects of environmental differences during crystal growth.
Using this procedure, we obtained samples of 1a-1g as crystals
with micro- to millimeter dimensions, and we obtained 1h-1k
as gel-like objects.
Differential scanning calorimetry (DSC) was employed
to determine melting points. We identified each endothermic
peak as either a melting point or a solid-to-solid phase-
transition temperature by referring to the optical microscopy
images. For 1e, DSC showed two endothermic peaks (at 151
and 194 °C) that corresponded to the melting points of two
different polymorphs (platelets and prisms, respectively).
Triamide 1f also showed two endothermic peaks (at 186 and
206 °C); the former was a solid-solid transition (prisms to
platelets) point, and the latter was the melting point of the
platelets. For 1g-1i, a small endothermic peak (¦H = 2.3-
4.1 J g¹1) was recorded at about 20 °C lower than the melting
points of the triamides, but no morphological change was
observed. In view of the relatively small enthalpy changes, we
attributed these small peaks to the partial melting of the alkyl
chains. The phase-transition temperatures are summarized in
Figure 2.
350
300
250
200
150
Cl-
NH2
NH2
NaBH4
MeOH
H2N
NH2
H2N
NH2
R
1a: R = CH3
1b: R = CH2CH3
O
NH
1c: R = (CH2)2CH3
1d: R = (CH2)3CH3
1e: R = (CH2)4CH3
1f: R = (CH2)5CH3
1g: R = (CH2)6CH3
1h: R = (CH2)7CH3
1i: R = (CH2)8CH3
1j: R = (CH2)9CH3
1k: R = (CH2)10CH3
RCOCl
Et3N
1
3
5
7
9
11
Number of carbon atoms in alkyl chain
CH2Cl2
O
Figure 2. Plot of phase-transition temperatures of 1a-1k as
a function of the number of carbon atoms in the alkyl
chains. Solid circles denote melting points, and open
circles denote solid-solid transitions (for 1e and 1f) and
undetermined transitions (for others).
HN
N
R
H
R
O
Scheme 1.
(a)
(b)
10 µm
Figure 1. (a) Optical micrograph and (b) scanning electron micrograph of 1i obtained from a dichloromethane-water emulsion.

732
Bull. Chem. Soc. Jpn. Vol. 82, No. 6 (2009)
Triphenylmethane-Based Alkanecarboxamides
For 1a-1f, the melting point varied from 190 to 350 °C
depending on the length of the alkyl chains, whereas for 1g-1k
the melting point was almost constant at around 160 °C
(¦mH = 76-85 J g¹1). This uniformity implies that the pre-
dominant contributions to the intermolecular forces were
similar for 1g-1k and were independent of the chain length.
We can assume that the melting enthalpies of 1g-1k were
determined mostly by the dissociation of intermolecular
hydrogen bonds, whereas those of 1a-1f involved some other
intermolecular forces such as inter-core CH/³ interactions.
The profiles of the IR spectra of 1a-1k were quite similar to
one another, with slight differences in the wavenumbers of the
characteristic peaks. In Figure 3, the wavenumbers of the
amide I (C=O stretch) and amide II (N-H bend) bands are
plotted as a function of the length of the alkyl chains. The
wavenumbers of the amide II bands differed clearly between
groups 1a-1f (1537-1541 cm¹1) and 1g-1k (1526-1531 cm¹1).
The trend for the amide I bands was basically the same as
that for the amide II bands, deviating only for a few of the
compounds. Similarly, the trend for the N-H stretch vibra-
tion bands (µ3290 cm¹1, not shown) showed a discontinuity
(slightly high wavenumber shift) at around 1e and 1f. The
clear difference in the resonances of the two amide bands
indicates that the modes of intermolecular hydrogen bonding
were different for 1a-1f and 1g-1k. On the basis of the sim-
ple relationship between the resonance of the C=O stretch
oscillation and the strength of hydrogen bonding, we can
assume that the molecular arrangement in 1g-1k was such that
the amide groups formed stronger hydrogen bonds than those
in 1a-1f. On the other hand, the N-H bond oscillation shows
quite an opposite trend. Thus, we should confine the inter-
pretation to a qualitative change in hydrogen-bonding mode.
The X-ray diffraction patterns of 1b-1d were quite similar to
one another (Figure 4), which suggests that they had the same
crystal system and similar packing. The peak at around 8° of 2ª
was characteristic of this group. The patterns for 1h-1k were
also similar to one another, but their crystal system was
apparently different from that of 1b-1d. The patterns for 1h-1k
under the same measurement conditions showed low S/N
ratios, which suggests low crystallinity, although there were
characteristic peaks in the region of 7° > 2ª. The pattern of 1e,
which was not similar to the patterns of any of the other
triamides, seemed to be a mixture of the patterns of 1d and 1f.
We succeeded in obtaining single-crystal X-ray structures of
1c and 1d, which had essentially the same packing structure,
although their lattice constants (a = 18.85 ¡ and c = 13.93 ¡
for 1c; a = 19.99 ¡ and c = 13.50 ¡ for 1d) differed slightly.
Powder patterns simulated from the crystallographic data of
1c and 1d agreed sufficiently well with the patterns shown
in Figure 4.24 In addition, from the observed powder pattern
of 1b, we estimated that it had a rhombohedral crystal system
with lattice constants of a = 17.7 ¡ and c = 14.0 ¡. The
ORTEP drawings of 1c are shown in Figure 5. Additionally,
a schematic view of the molecular packing is shown in
Figure 6a. The molecule was shaped like a wide-open parasol
centered around the central sp3 carbon atom (the H-C-C(ipso)
and C(ipso)-C(H)-C(ipso) angles were 107.2 and 111.7°,
respectively). The depth of the molecular parasol was roughly
1680
1670
1660
1650
1640
1630
1570
1560
1550
1540
1530
1520
1
3
5
7
9
11
Number of carbon atoms in alkyl chain
Figure 3. Plot of the wavenumber of the amide I ( , left
axis) and II ( , right axis) vibrational bands as a function
of the number of carbon atoms in the alkyl chains.
1f
1e
1k
1d
1c
1b
1j
1i
1h
1a
1g
0
5
10
15
20
25
30
0
5
10
15
20
25
30
2θ
/ degree
2θ / degree
Figure 4. X-ray diffraction patterns for the powder samples of 1a-1k.

H. Houjou et al.
Bull. Chem. Soc. Jpn. Vol. 82, No. 6 (2009)
733
(a)
(a)
(b)
(b)
(c)
(c)
Figure 5. ORTEP drawings of the crystal structure of 1c:
(a) parallel to the c axis, (b) parallel to the a axis of two
proximal molecules, and (c) molecular packing in the C-
plane. One-third of the molecule is crystallographically
independent.
Figure 6. The packing structures of 1b-1d, where each
triple-ribbed parasol represents the unit molecule, and each
edge (deep blue) of the rib is a hydrogen-bonding site: (a)
bird’s-eye view, (b) top view of the hydrogen-bonding
sites, and (c) side view of the 31-helical hydrogen-bonded
chain of 1c. For clarity, only one-third of the molecule is
shown.
3 ¡, and the C=O and N-H groups were nearly parallel to the c
axis (Figure 5b).
Two proximal parasols of 1c formed a dimer with alternating
ribs; in the dimer, there were some close contacts between the
phenyl rings. The shortest distance between H and C atoms was
3.026 ¡ (denoted by a line in Figure 5b), which suggests a CH/
³ interaction.25 A sextuple CH/³ interaction between two
molecules may have contributed to the relatively high melting
points of 1b-1d. The dimer was hexagonally packed with its
parasol’s “ferrule” coincident with the c axis, and consequently
the amide groups were arranged in a hydrogen-bonded 31-helix
parallel to the c axis (Figures 6b and 6c). The N£O distance in
each proximal pair was 2.78 ¡, which is in the normal range for
an amide-amide hydrogen bond.
The alkyl chains formed a vortex in a space surrounded by
six molecules of 1c (Figure 5c). Because of restriction by the
hydrogen bonds, the volume of the space that accommodated
the alkyl chains could not freely expand. Thus, this type of
crystal packing can be expected to become unstable when the
number of carbon atoms of the chains is increased to 5 or 6.
The polymorphism observed for 1e and 1f implies that these
compounds were situated at a critical point between the

734
Bull. Chem. Soc. Jpn. Vol. 82, No. 6 (2009)
Triphenylmethane-Based Alkanecarboxamides
(a)
packing structure of 1b-1d and another structure suitable for
1g-1k.
Although relating crystal structure to macroscopic crystal
habit is difficult, we can at least assume that the prismatic or
platelet habit observed for 1a-1f was a consequence of the rates
of crystal growth that were well-balanced between the c axis
direction (hydrogen bonding and CH/³ interaction) and the C-
plane direction (hydrogen bonding).
Compound 1g was obtained as needles under the given
conditions. In contrast, recrystallization of compounds 1h-1k
from all the solvent systems we examined afforded very thin,
fibrous objects. This result suggests that for these compounds
the rates of crystal growth in the axial direction and the radial
direction were markedly unbalanced. Because we did not
obtain single-crystal X-ray structures of all the compounds, we
attempted to estimate the structure of the molecular assemblies
of 1g-1k based on the analytical results described above.
First, the trend in the melting points of 1g-1k indicate the
compounds had a molecular packing in which a certain
intermolecular force was the predominant contributor to
melting enthalpy. Second, the trend of the IR spectra suggests
that 1g-1k formed stronger (stiffer) hydrogen bonds than did
the other triamides. Third, the X-ray powder patterns suggest
that the molecular packing of 1g-1k was definitely different
from that of 1b-1d.
(b)
Figure 7. A plausible packing structure of 1g, where each
triple-ribbed parasol represents the unit molecule and the
ends of the ribs (dark blue) are hydrogen-bonding sites:
(a) bird’s-eye view and (b) top view of the hydrogen-
bonding sites.
attributable to a difference in the mode of molecular assembly.
Taking account of the precipitation conditions adopted, we
assumed the following mechanism for the formation of the
needles. In a moderately polar medium, the hydrogen-bonded
and non-hydrogen-bonded triphenylmethane molecules were in
equilibrium with each other. Upon addition of a protic solvent
or water, hydrophobic interactions between the alkyl chains
promoted crystal growth along the c axis, and growth in the
direction perpendicular to the c axis was suppressed owing to
accelerated dissociation of hydrogen bonding. As a result of the
difference in growth rate, 1g formed an assembly with a high
axis-to-radius ratio and eventually separated out of solution as
needles. This assumption on the growth mechanism is also in
agreement of the cylindrical symmetry of the presumed
molecular packing.
The X-ray diffraction patterns (Figure 4) of the powder
samples of 1h-1k were similar to one another, which suggests
that their packing structures were similar. However, the
diffraction patterns showed a low S/N ratio, which indicates
that the order of the molecular arrangement was considerably
disturbed. For these compounds, as well as for 1g, the
formation of a 31-helical hydrogen-bonded chain was impos-
sible, owing to their long alkyl chains. Thus, we can suggest
that the predominant force for assembly of the molecules
themselves was hydrophobic interactions. It is probable that an
extreme difference in the rates of crystal growth resulted in the
formation of fibrous objects.
These findings suggest that the local structure of the
molecular assembly in 1g is a key to the estimation of the
structure of 1h-1k. By analyzing the powder pattern of 1g
under various boundary conditions, we provisionally deter-
ꢀ
mined its crystal system as trigonal (P3), with lattice constants
of a = 15.4 and c = 7.8 ¡, which best reproduces the observed
pattern.25 Although a Rietlvelt analysis of 1g has not been
completed to give the convergence of a structure, we attempted
to fit the molecule into this lattice. A triangle with a base of
15.4 ¡ could just accommodate a triphenylmethane triamide
core, although the height of 7.8 ¡ was not in agreement with
the alternating assembly of the dimer (µ13.5 ¡) observed for
1b-1d. Therefore, we suggest that the parasol of triphenyl-
methane stacked in columnar form and that the neighboring
pairs of columns ran anti-parallel to each other (Figure 7a).
Inside the column, there was sufficient volume for the alkyl
chains to fold between the stacked parasols. Packing two
molecular units in this cell resulted in a calculated density of
1.39 g cm¹3, which was in good agreement with the density of
common organic compounds, albeit slightly larger than the
values for 1c (1.162 g cm¹3) and 1d (1.155 g cm¹3).
In this model, each lattice point assembled six ribs in S6
symmetry, by virtue of which the amide groups formed a cyclic
hydrogen-bonding network, although we did not determine the
detailed structure (Figure 7b). A possible model for such a
cyclic hexamer has been reported for the crystal structure of a
cyclohexane-1,3,5-tricarboxamide.26 Notably, the hydrogen-
bonding interactions in the above model contributed to crystal
growth in the lateral (i.e., perpendicular to the c axis) direction,
whereas for 1b-1d, the hydrogen bond served as a linker in
both axial (i.e., parallel to the c axis) and lateral directions.
Consequently, we suggest that 1g, growth in the axial direction
was dominated by interactions among alkyl chains and that the
obvious difference in crystal habit between 1b-1d and 1g is
Conclusion
We synthesized a series of triphenylmethane triamides with
alkyl chains of various lengths as new synthons for supra-
molecular systems. The morphology of the compounds
depended on the length of the alkyl chains. Upon recrystalliza-
tion, the triamides with short alkyl chains (1-6 carbons) gave
prismatic or platelet microcrystals, whereas those with long
alkyl chains (7-11 carbons) gave gel-like objects composed of

H. Houjou et al.
Bull. Chem. Soc. Jpn. Vol. 82, No. 6 (2009)
735
74.52; H, 7.46; N, 8.41%. Found: C, 74.43; H, 7.62; N, 8.69%;
1H NMR (DMSO-d6): ¤ 0.90 (t, J = 7.4 Hz, CH3, 9H), 1.59 (sext,
J = 7.4 Hz, -CH2-, 6H), 2.25 (t, J = 7.4 Hz, -CH2-, 6H), 5.41 (s,
Ar3CH, 1H), 6.99 (d, J = 8.6 Hz, ArH, 6H), 7.51 (d, J = 8.6 Hz,
ArH, 6H), 9.84 (s, NH, 3H); 13C NMR (DMSO-d6): ¤ 14.6, 19.6,
39.2, 55.1, 119.8, 129.9, 138.2, 139.4, 171.8.
1d: Yield 90%; mp 283-284 °C; IR (KBr): 3288 (¯NH), 1663
(¯amideI), 1541 (¯amideII) cm¹1; FAB(+) MS m/z 542.61 (542.34
calcd for M + H+); elemental analysis Calcd for C34H43N3O3: C,
75.38; H, 8.00; N, 7.76%. Found: C, 75.12; H, 8.17; N, 7.63%;
1H NMR (DMSO-d6): ¤ 0.89 (t, J = 7.4 Hz, CH3, 9H), 1.31 (sext,
J = 7.4 Hz, -CH2-, 6H), 1.56 (quint, J = 7.4 Hz, -CH2-, 6H),
2.27 (t, J = 7.4 Hz, -CH2-, 6H), 5.41 (s, Ar3CH, 1H), 6.99 (d,
J = 8.6 Hz, ArH, 6H), 7.50 (d, J = 8.6 Hz, ArH, 6H), 9.84 (s, NH,
3H); 13C NMR (DMSO-d6): ¤ 15.4, 23.4, 28.9, 37.7, 55.1, 120.5,
130.6, 138.9, 140.2, 172.6.
1e: Yield 95%; mp 194-195 °C; IR (KBr): 3297 (¯NH), 1660
(¯amideI), 1538 (¯amideII) cm¹1; FAB(+) MS m/z 584.55 (584.39
calcd for M + H+); elemental analysis Calcd for C37H49N3O3¢
(H2O)1/3: C, 75.35; H, 8.49; N, 7.12%. Found: C, 75.73; H, 8.52;
N, 7.14%; 1H NMR (CDCl3): ¤ 0.87 (t, J = 7.0 Hz, CH3, 9H),
1.30-1.33 (m, -CH2-, 12H), 1.67-1.71 (m, -CH2-, 6H), 2.33 (t,
J = 7.6 Hz, -CH2-, 6H), 5.36 (s, Ar3CH, 1H), 6.93 (d, J = 8.4 Hz,
ArH, 6H), 7.39 (d, J = 8.4 Hz, ArH, 6H), 7.81 (s, NH, 3H);
13C NMR (CDCl3): ¤ 14.3, 22.8, 25.8, 31.8, 38.0, 55.4, 120.2,
129.9, 136.5, 139.8, 172.0.
1f: Yield 80%; mp 205-206 °C; IR (KBr): 3298 (¯NH), 1662
(¯amideI), 1537 (¯amideII) cm¹1; FAB(+) MS m/z 626.68 (626.43
calcd for M + H+); elemental analysis Calcd for C40H55N3O3: C,
76.76; H, 8.86; N, 6.71%. Found: C, 76.48; H, 8.86; N, 6.52%;
1H NMR (CDCl3): ¤ 0.85 (t, J = 6.8 Hz, CH3, 9H), 1.25-1.35 (m,
-CH2-, 18H), 1.68 (quint, J = 7.5 Hz, -CH2-, 6H), 2.33 (t, J =
7.5 Hz, -CH2-, 6H), 5.36 (s, Ar3CH, 1H), 6.93 (d, J = 8.4 Hz,
ArH, 6H), 7.39 (d, J = 8.4 Hz, ArH, 6H), 7.83 (s, NH, 3H);
13C NMR (CDCl3): ¤ 14.4, 22.9, 26.1, 29.3, 31.9, 38.1, 55.4,
120.2, 129.9, 136.6, 139.8, 172.0.
1g: Yield 78%; mp 182-183 °C; IR (KBr): 3285 (¯NH), 1654
(¯amideI), 1526 (¯amideII) cm¹1; FAB(+) MS m/z 668.82 (668.48
calcd for M + H+); elemental analysis Calcd for C43H61N3O3: C,
77.32; H, 9.20; N, 6.29%. Found: C, 77.12; H, 9.30; N, 6.79%;
1H NMR (CDCl3): ¤ 0.87 (t, J = 6.9 Hz, CH3, 9H), 1.26-1.32 (m,
-CH2-, 24H), 1.70 (quint, J = 7.4 Hz, -CH2-, 6H), 2.33 (t, J =
7.4 Hz, -CH2-, 6H), 5.40 (s, Ar3CH, 1H), 6.97 (d, J = 8.4 Hz,
ArH, 6H), 7.40 (d, J = 8.4 Hz, ArH, 6H), 7.47 (s, NH, 3H);
13C NMR (CDCl3): ¤ 14.5, 23.0, 26.1, 29.5, 29.7, 32.1, 38.2, 55.4,
120.1, 130.0, 136.5, 139.9, 171.8.
1h: Yield 85%; mp 173-174 °C; IR (KBr): 3289 (¯NH), 1658
(¯amideI), 1527 (¯amideII) cm¹1; FAB(+) MS m/z 710.94 (710.53
calcd for M + H+); elemental analysis Calcd for C46H67N3O3: C,
77.81; H, 9.51; N, 5.92%. Found: C, 77.58; H, 9.60; N, 6.09%;
1H NMR (CDCl3): ¤ 0.86 (t, J = 6.8 Hz, CH3, 9H), 1.25-1.32 (m,
-CH2-, 30H), 1.69 (quint, J = 7.6 Hz, -CH2-, 6H), 2.33 (t, J =
7.6 Hz, -CH2-, 6H), 5.39 (s, Ar3CH, 1H), 6.95 (d, J = 8.4 Hz,
ArH, 6H), 7.40 (d, J = 8.4 Hz, ArH, 6H), 7.61 (s, NH, 3H);
13C NMR (CDCl3): ¤ 14.5, 23.0, 26.1, 29.5, 29.7, 29.7, 32.2, 38.1,
55.4, 120.1, 129.9, 136.5, 139.8, 171.9.
100-nm fibrous crystals. The melting points, the vibrational
frequencies of the amide groups, and the molecular packing in
the solid state also clearly depended on chain length. The
crystal structures of some of the compounds suggested that the
compounds with short chains preferred a structure composed of
stacked dimers in which a 31-helical hydrogen-bonded chain
was formed. However, this structure appeared to be destabi-
lized in compounds with alkyl chains with more than 6
carbons. We propose that these compounds preferred a structure
in which the alkyl chains were accommodated between stacked
parasols of triphenylmethane. We propose that molecular
assembly with a high axis-to-radius ratio resulted from the
predominance of hydrophobic interactions during crystal
growth. In summary, in this study we demonstrated that a
drastic change in macroscopic morphology can be brought
about by the tuning of the length of the alkyl side chains, which
resulted in a change in the mode of molecular assembly.
Experimental
General.
All the chemicals were purchased from Tokyo
Chemical Industry, and used without further purification. Thermo-
metric analyses were performed using a Rigaku TG-DTA 8120 and
a Perkin-Elmer Prys1 DSC (sample weight µ20 mg, heating rate
¹1
5 K min under N2 flow 20 mL min¹1). IR spectra were recorded
using a JEOL WINSPEC100 spectrometer, with wavenumber
1
resolution of 0.5 cm¹1. H and 13C NMR spectra were measured
1
using a JEOL JNM-AL400 (400 MHz for H nuclei). Powder X-
ray diffraction patterns were recorded using a Rigaku RINT2100
¹1
diffractometer, -(Cu K¡) = 1.5418 ¡, with scan rate of 4° min
and scan step of 0.02°.
,
Synthesis. Pararosaniline hydrochloride salt (3.24 g, 10 mmol)
was reduced to the corresponding leuco form, 4,4¤,4¤¤-triamino-
triphenylmethane (2), using sodium tetrahydroborate in 100 mL of
ethanol solution. The usual work up of the solution afforded
grayish white crystalline solid (y. 52%). Three equivalents of an
alkanoyl chloride (CnH2n+1COCl, n = 1-10) were allowed to react
with 2 in dichloromethane solution in the presence of three
equivalents of triethylamine. The reaction mixture was worked up
in the usual way, and colorless solid of 1a-1j was isolated. The
1
products were identified by H and 13C NMR, IR, FAB(+) MS,
and elemental analysis.
1a: Yield 77%; mp 206-207 °C; IR (KBr): 3307 (¯NH), 1673
(¯amideI), 1537 (¯amideII) cm¹1; FAB(+) MS m/z 415.67 (415.19
calcd for M+); elemental analysis Calcd for C25H25N3O3¢H2O: C,
69.27; H, 6.28; N, 9.69%. Found: C, 69.19; H, 6.25; N, 9.61%;
1H NMR (DMSO-d6): ¤ 2.01 (s, CH3, 9H), 5.41 (s, Ar3CH, 1H),
6.99 (d, J = 8.6 Hz, ArH, 6H), 7.47 (d, J = 8.6 Hz, ArH, 6H), 9.90
(s, NH, 3H); 13C NMR (DMSO-d6): ¤ 24.8, 80.0, 119.7, 129.8,
138.1, 139.4, 168.8.
1b: Yield 93%; mp 294-295 °C; IR (KBr): 3288 (¯NH), 1666
(¯amideI), 1541 (¯amideII) cm¹1; FAB(+) MS m/z 458.61 (458.25
calcd for M + H+); elemental analysis Calcd for C28H31N3O3¢
(C3H7NO)1/3: C, 72.27; H, 6.97; N, 9.69%. Found: C, 72.21; H,
1
7.06; N, 9.64%; H NMR (DMSO-d6): ¤ 1.06 (t, J = 7.6 Hz, CH3,
9H), 2.29 (q, J = 7.6 Hz, -CH2-, 6H), 5.40 (s, Ar3CH, 1H), 6.98
(d, J = 8.6 Hz, ArH, 6H), 7.50 (d, J = 8.6 Hz, ArH, 6H), 9.83 (s,
NH, 3H); 13C NMR (DMSO-d6): ¤ 10.5, 30.3, 55.0, 119.7, 129.8,
138.1, 139.3, 172.5.
1c: Yield 79%; mp 325-326 °C; IR (KBr): 3287 (¯NH), 1663
(¯amideI), 1541 (¯amideII) cm¹1; FAB(+) MS m/z 500.44 (500.29
calcd for M + H+); elemental analysis Calcd for C31H37N3O3: C,
1i: Yield 92%; mp 177-178 °C; IR (KBr): 3292 (¯NH), 1658
(¯amideI), 1527 (¯amideII) cm¹1; FAB(+) MS m/z 752.94 (752.58
calcd for M + H+); elemental analysis Calcd for C49H73N3O3: C,
78.25; H, 9.78; N, 5.59%. Found: C, 78.18; H, 9.91; N, 5.86%;
1H NMR (CDCl3): ¤ 0.87 (t, J = 6.9 Hz, CH3, 9H), 1.25-1.32 (m,
-CH2-, 36H), 1.70 (quint, J = 7.6 Hz, -CH2-, 6H), 2.33 (t, J =

736
Bull. Chem. Soc. Jpn. Vol. 82, No. 6 (2009)
Triphenylmethane-Based Alkanecarboxamides
7.6 Hz, -CH2-, 6H), 5.41 (s, Ar3CH, 1H), 6.99 (d, J = 8.4 Hz,
ArH, 6H), 7.35 (s, NH, 3H), 7.41 (d, J = 8.4 Hz, ArH, 6H);
13C NMR (CDCl3): ¤ 14.5, 23.1, 26.1, 29.7, 29.7, 29.8, 29.9, 32.3,
38.2, 55.4, 120.1, 130.1, 136.5, 139.8, 171.7.
S. Sivakova, S. J. Rowan, Chem. Soc. Rev. 2005, 34, 9; L.
Brunsveld, B. J. B. Folmer, E. W. Meijer, R. P. Sijbesma, Chem.
Rev. 2001, 101, 4071; C. Schmuck, W. Wienand, Angew. Chem.,
Int. Ed. 2001, 40, 4363.
1j: Yield 80%; mp 178-179 °C; IR (KBr): 3290 (¯NH), 1655
(¯amideI), 1528 (¯amideII) cm¹1; FAB(+) MS m/z 795.01 (794.62
calcd for M + H+); elemental analysis Calcd for C52H79N3O3: C,
78.64; H, 10.03; N, 5.29%. Found: C, 78.68; H, 10.22; N, 5.46%;
1H NMR (CDCl3): ¤ 0.87 (t, J = 6.9 Hz, CH3, 9H), 1.24-1.30 (m,
-CH2-, 42H), 1.68 (quint, J = 7.6 Hz, -CH2-, 6H), 2.33 (t, J =
7.6 Hz, -CH2-, 6H), 5.42 (s, Ar3CH, 1H), 7.00 (d, J = 8.5 Hz,
ArH, 6H), 7.25 (s, NH, 3H), 7.41 (d, J = 8.5 Hz, ArH, 6H);
13C NMR (CDCl3): ¤ 14.5, 23.1, 26.2, 29.7, 29.7, 29.8, 29.9, 30.0,
32.3, 38.1, 55.4, 120.2, 129.9, 136.6, 139.8, 172.1.
3
C. Hilger, R. Stadler, L. de Lucca Freitas, Polymer 1990,
31, 818; M. Müller, U. Seidel, R. Stadler, Polymer 1995, 36, 3143;
M. Müller, R. Stadler, F. Kremer, G. Williams, Macromolecules
1995, 28, 6942.
4
5
6
M. Chino, K. Ashiura, Macromolecules 2001, 34, 9201.
C.-C. Peng, V. Abetz, Macromolecules 2005, 38, 5575.
Y. Zhao, J. B. Beck, S. J. Rowan, A. M. Jamieson,
Macromolecules 2004, 37, 3529.
L. Leibler, M. Rubinstein, R. H. Colby, Macromolecules
1991, 24, 4701.
7
1k: Yield 89%; mp 171-172 °C; IR (KBr): 3290 (¯NH), 1659
(¯amideI), 1528 (¯amideII) cm¹1; FAB(+) MS m/z 836.84 (836.67
calcd for M + H+); elemental analysis Calcd for C55H85N3O3: C,
78.99; H, 10.24; N, 5.02%. Found: C, 79.12; H, 10.48; N, 5.18%;
1H NMR (CDCl3): ¤ 0.87 (t, J = 6.9 Hz, CH3, 9H), 1.24-1.31 (m,
-CH2-, 48H), 1.69 (quint, J = 7.6 Hz, -CH2-, 6H), 2.33 (t, J =
7.6 Hz, -CH2-, 6H), 5.38 (s, Ar3CH, 1H), 6.95 (d, J = 8.4 Hz,
ArH, 6H), 7.40 (d, J = 8.4 Hz, ArH, 6H), 7.63 (s, NH, 3H);
13C NMR (CDCl3): ¤ 14.5, 23.1, 26.2, 29.7, 29.8, 29.8, 29.9, 30.0,
30.0, 32.3, 38.0, 55.4, 120.3, 129.8, 136.6, 139.8, 172.3.
8
2311.
9
G. R. Desiraju, Angew. Chem., Int. Ed. Engl. 1995, 34,
F. J. M. Hoeben, P. Jonkheijm, E. W. Meijer, A. P. H. J.
Schenning, Chem. Rev. 2005, 105, 1491.
10 D. Pasini, A. Kraft, Curr. Opin. Solid State Mater. Sci.
2004, 8, 157.
11 C. Bao, R. Lu, M. Jin, P. Xue, C. Tan, T. Xu, G. Liu, Y.
Zhao, Chem.®Eur. J. 2006, 12, 3287; C. Bao, M. Jin, R. Lu, Z.
Song, X. Yang, D. Song, T. Xu, G. Liu, Y. Zhao, Tetrahedron
2007, 63, 7443.
X-ray Crystal Structure Analysis. For X-ray diffraction of
single crystals, data were collected on a MacScienceDIPLabo
Imaging Plate diffractometer, -(Cu K¡) = 1.5418 ¡. The structure
was solved by direct methods and expanded using Fourier
techniques. All calculations were performed with the crystallo-
graphic software package SHELX-97.27 Crystallographic data have
been deposited with Cambridge Crystallographic Data Centre:
Deposition numbers CCDC-723047 and -723048 to compounds
No. 1c and 1d. Copies of the data can be obtained free of charge
via http://www.ccdc.cam.ac.uk/conts/retrieving.html (or from
the Cambridge Crystallographic Data Centre, 12, Union Road,
Cambridge, CB2 1EZ, U.K.; Fax: +44 1223 336033; e-mail:
deposit@ccdc.cam.ac.uk).
12 For recent examples: X.-Q. Li, V. Stepanenko, Z. Chen, P.
Prins, L. D. A. Siebbeles, F. Würthner, Chem. Commun. 2006,
3871; B. El Hamaoui, L. Zhi, W. Pisula, U. Kolb, J. Wu, K.
Müllen, Chem. Commun. 2007, 2384; K. Sakajiri, T. Sugisaki, K.
Moriya, Chem. Commun. 2008, 3447.
13 R. Takasawa, K. Murota, I. Yoshikawa, K. Araki, Macro-
mol. Rapid Commun. 2003, 24, 335.
14 Y. Sagara, T. Mutai, I. Yoshikawa, K. Araki, J. Am. Chem.
Soc. 2007, 129, 1520.
15 D. F. Duxbury, Chem. Rev. 1993, 93, 381.
16 J. Liu, J. Yan, X. Yuan, K. Liu, J. Peng, Y. Fang, J. Colloid
Interface Sci. 2008, 318, 397.
17 F. Osterod, A. Kraft, Chem. Commun. 1997, 1435; A. Kraft,
F. Osterod, J. Chem. Soc., Perkin Trans. 1 1998, 1019.
18 W. Heilmayer, R. Smounig, K. Gruber, W. M. F. Fabian, C.
Reidlinger, C. O. Kappe, C. Wentrup, G. Kollenz, Tetrahedron
2004, 60, 2857.
1c: C31H37N3O3, Mr = 499.64, rhombohedral, a = 18.8490(6),
¹3
c = 13.927 ¡, V = 4285.14(14) ¡3,
Dcalcd = 1.162 g cm , T =
¹1
ꢀ
233 K, space group R3 (#143), Z = 6, ®(Cu K¡) = 5.9 cm
,
12848 reflections measured and 1385 unique (2ªmax = 146.1°,
int = 0.033), which were used in all calculations. R = 0.080,
R
19 M. S. Shchepinov, V. A. Korshun, Chem. Soc. Rev. 2003,
32, 170.
Rw = 0.168.
1d: C34H43N3O3, Mr = 541.71, rhombohedral, a = 19.9890(5),
20 V. Rudzevich, D. Schollmeyer, D. Braekers, J. F. Desreux,
R. Diss, G. Wipff, V. Böhmer, J. Org. Chem. 2005, 70, 6027.
21 J. Otsuki, S. Shimizu, M. Fumino, Langmuir 2006, 22,
6056.
22 C. A. Papamicaël, O. Mongin, A. Gossauer, Chem. Mon.
2007, 138, 791.
¹3
c = 13.500 ¡, V = 4671.39(12) ¡3,
Dcalcd = 1.155 g cm , T =
¹1
ꢀ
298 K, space group R3 (#143), Z = 6, ®(Cu K¡) = 5.8 cm
,
13799 reflections measured and 1677 unique (2ªmax = 146.7°,
int = 0.029), which were used in all calculations. R = 0.083,
Rw = 0.207.
R
23 H. Meier, S. Kim, Eur. J. Org. Chem. 2001, 1163.
24 We utilized a Rietan2000 program package for the purpose
of powder pattern simulation and estimation of lattice constants: F.
Izumi, T. Ikeda, Mater. Sci. Forum 2000, 321-324, 198.
25 M. Nishino, M. Hirota, Y. Umezawa, The CH/³ Inter-
action. Evidence, Nature, and Consequences, Wiley-VCH, New
York, 1998.
This research was partially supported by the Ministry of
Education, Culture, Sports, Science and Technology, Grant-in-
Aid for Scientific Research, 18710094, 2006-2007.
References
1
Supramolecular Polymers, 2nd ed., ed. by A. Ciferri,
Taylor & Francis, New York, 2005.
H. Hofmeier, U. S. Schubert, Chem. Commun. 2005, 2423;
26 B. K. Saha, Acta Crystallogr., Sect. C 2007, 63, o591.
27 G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112.
2