Synthesis and structural investigation of new triptycene-based ligands: En route to shape-persistent dendrimers and macrocycles with large free volume

Article abstract of DOI:10.1021/jo701297q

(Figure Presented) Triptycene was used to build a series of ligands containing pyrazine groups for use in forming coordination frameworks and as model compounds toward target shape-persistent dendrimers and macrocycles. The phenylenediamine precursors are formed by controlled double nitration/reduction of triptycene, followed by condensation with triptycenyl o-quinone to form the ligands. Solid-state structures show that the ligands favor packing in layered structures with intermolecular π stacking. Increasing the length of the wings still allows for effective packing, but increasing the number of pyrazine groups or adding more triptycenyl moieties reduces packing efficiency. To demonstrate the potential of these molecules to function as ligands, a coordination complex of 15 with copper(I) iodide was obtained and found to form dimers that pack into a layered arrangement.

Full text of DOI:10.1021/jo701297q

Synthesis and Structural Investigation of New Triptycene-Based
Ligands: En Route to Shape-Persistent Dendrimers and
Macrocycles with Large Free Volume
Jonathan H. Chong and Mark J. MacLachlan*
Department of Chemistry, UniVersity of British Columbia, 2036 Main Mall,
VancouVer, BC, V6T 1Z1, Canada
mmaclach@chem.ubc.ca
ReceiVed June 18, 2007
Triptycene was used to build a series of ligands containing pyrazine groups for use in forming coordination
frameworks and as model compounds toward target shape-persistent dendrimers and macrocycles. The
phenylenediamine precursors are formed by controlled double nitration/reduction of triptycene, followed
by condensation with triptycenyl o-quinone to form the ligands. Solid-state structures show that the ligands
favor packing in layered structures with intermolecular π stacking. Increasing the length of the wings
still allows for effective packing, but increasing the number of pyrazine groups or adding more triptycenyl
moieties reduces packing efficiency. To demonstrate the potential of these molecules to function as ligands,
a coordination complex of 15 with copper(I) iodide was obtained and found to form dimers that pack
into a layered arrangement.
Introduction
macrocycles, and most involve the incorporation of conjugated
components. This typically results in a planar framework where
the orbitals involved in conjugation are directed perpendicularly
above and below the plane of the macrocycle. Another approach
to shape-persistent macrocycles involves belt-shaped molecules
such as cyclodextrins and cucurbiturils.2 Belt-shaped molecules
can also be designed to incorporate extended conjugation, as
seen in the proposed [n]cyclacene and cyclo[n]phenacene classes
of molecules; the conjugated orbitals in these molecules are
radially directed.3
Rigid shape-persistent macrocycles such as porphyrins, ph-
thalocyanines, and phenyleneethynylenes have emerged as
important substances for the study of aggregation and
liquid crystallinity.1 Over the past three decades, many strategies
have been developed to synthesize large shape-persistent
(1) (a) Balaban, T. S. Acc. Chem. Res. 2005, 38, 612-623. (b)
Elemans, J. A. A. W.; van Hameren, R.; Nolte, R. J. M.; Rowan, A. E.
AdV. Mater. 2006, 18, 1251-1266. (c) Eichhorn, H. J. Porphyrins
Phthalocyanines 2000, 4, 88-102. (d) Hu, J.-S.; Guo, Y.-G.; Liang, H.-P.;
Wan, L.-J.; Jiang, L. J. Am. Chem. Soc. 2005, 127, 17090-17095. (e)
Castella, M.; Lopez-Calahorra, F.; Velasco, D.; Finkelmann, H. Chem.
Commun. 2002, 2348-2349. (f) Jasat, A.; Dolphin, D. Chem. ReV. 1997,
97, 2267-2340. (g) van Nostrum, C. F. AdV. Mater. 1996, 8, 1027-1030.
(h) Zhang, J.; Moore, J. S. J. Am. Chem. Soc. 1992, 114, 9701-9702. (i)
Inabe, T. J. Porphyrins Phthalocyanines 2001, 5, 3-12. (j) Lahiri, S.;
Thompson, J. L.; Moore, J. S. J. Am. Chem. Soc. 2000, 122, 11315-11319.
(k) Ho¨ger, S.; Bonrad, K.; Mourran, A.; Beginn, U.; Mo¨ller, M. J. Am.
Chem. Soc. 2001, 123, 5651-5659. (l) Rosselli, S.; Ramminger, A.-D.;
Wagner, T.; Lieser, G.; Ho¨ger, S. Chem.-Eur. J. 2003, 9, 3481-3491.
The organization of shape-persistent macrocycles may create
molecule-based nanotubes with advantages over other types of
(2) (a) Szejtli, J. Pure Appl. Chem. 2004, 76, 1825-1845. (b) Harada,
A. J. Polym. Sci., Part A: Polym. Chem. 2006, 44, 5113-5119. (c) Harada,
A. Acc. Chem. Res. 2001, 34, 456-464. (d) Kim, K.; Selvapalam, N.; Ko,
Y. H.; Park, K. M.; Kim, D.; Kim, J. Chem. Soc. ReV. 2007, 36, 267-279.
(e) Lee, J. W.; Samal, S.; Selvapalam, N.; Kim, H.-J.; Kim, K. Acc. Chem.
Res. 2003, 36, 621-630.
10.1021/jo701297q CCC: $37.00 © 2007 American Chemical Society
Published on Web 10/11/2007
J. Org. Chem. 2007, 72, 8683-8690
8683

Chong and MacLachlan
FIGURE 1. Target giant beltlike macrocycle 1.
SCHEME 1. Synthesis of 2,3-Diaminotriptycene (2)
TABLE 1. Product Distributions of Nitrated Triptycenes Formed
under Various Reaction Conditionsa
peptides, which have found use as transmembrane channel
mimics.5 With this molecule-based approach to nanotubes, the
size of the tube may be controlled, opening new possibilities
for host-guest applications such as selective adsorption,
catalysis, and nanowire templation. Furthermore, controlled
macrocycle aggregation and de-aggregation would be useful in
controlling the adsorption and release of guests and would allow
for their eventual removal when used in template-directed
synthesis.
We are targeting shape-persistent ligands such as the giant
beltlike macrocycle 1 illustrated in Figure 1 for assembly into
tubular structures. Phenazine groups on each side of the hexagon
could coordinate to metals to link the molecules into a tubular
structure or a highly porous 3-D framework.6 Central to this
design of these ligands is the use of triptycene as a thermally
robust component. Triptycene is an ideal building block for
macrocycle 1 because the constrained 120° angle between the
phenyl rings provides the required geometry to form a hexagon.
b
KNO3
dinitroc
mononitroc
triptycenec
1
1.1
1.2
1
1
1
7.79
6.58
5.12
2.09
1.60
0.81
a
All reactions were performed using 0.25 g of triptycene and 1.5 equiv
of p-toluenesulfonic acid in 10 mL of acetic anhydride, stirring overnight
b
at room temperature. Equivalents of KNO3 added per equivalent of
triptycene used. c Ratios of nitrotriptycenes were determined by integration
of 1H NMR spectra and are normalized to dinitrotriptycene. Traces of trinitro
products evident were not integrated.
nanotubes. For example, although carbon nanotubes possess
interesting physical and electronic properties, they are limited
by costly and nonintuitive synthetic methods and are difficult
to separate.4 These difficulties have hindered the commercial
development of nanotubes for wide-ranging applications. The
organization of shape-persistent macrocycles into 1-D tubes
offers the possibility of creating uniform, monodisperse nano-
tubes with homogeneous properties. This approach has been
illustrated with the hydrogen-bonded assemblies of cyclic D,L-
(4) (a) Ajayan, P. M. Chem. ReV. 1999, 99, 1787-1800. (b) Tasis, D.;
Tagmatarchis, N.; Bianco, A.; Prato, M. Chem. ReV. 2006, 106, 1105-
1136.
(5) (a) Ghadiri, M. R.; Granja, J. R.; Milligan, R. A.; McRee, D. E.;
Khazanovich, N. Nature 1993, 366, 324-327. (b) Khazanovich, N.; Granja,
J. R.; McRee, D. E.; Milligan, R. A.; Ghadiri, M. R. J. Am. Chem. Soc.
1994, 116, 6011-6012. (c) Ghadiri, M. R.; Granja, J. R.; Buehler, L. K.
Nature 1994, 369, 301-304. (d) Granja, J. R.; Ghadiri, M. R. J. Am. Chem.
Soc. 1994, 116, 10785-10786. (e) Bong, D. T.; Ghadiri, M. R. Angew.
Chem., Int. Ed. 2001, 40, 2163-2166.
(3) (a) Tahara, K.; Tobe, Y. Chem. ReV. 2006, 106, 5274-5290. (b)
Ashton, P. R.; Brown, G. R.; Isaacs, N. S.; Giuffrida, D.; Kohnke, F. H.;
Mathias, J. P.; Slawin, A. M. Z.; Smith, D. R.; Stoddart, J. F.; Williams,
D. J. J. Am. Chem. Soc. 1992, 114, 6330-6353. (c) Cory, R. M.; McPhail,
C. L. Tetrahedron Lett. 1996, 37, 1987-1990. (d) Neudorff, W. D.; Lentz,
D.; Anibarro, M.; Schlu¨ter, A. D. Chem.-Eur. J. 2003, 9, 2745-2757. (e)
Nakamura, E.; Tahara, K.; Matsuo, Y.; Sawamura, M. J. Am. Chem. Soc.
2003, 125, 2834-2835.
(6) Campbell, K.; Kuehl, C. J.; Ferguson, M. J.; Stang, P. J.; Tykwinski,
R. R. J. Am. Chem. Soc. 2002, 124, 7266-7267.
8684 J. Org. Chem., Vol. 72, No. 23, 2007

Synthesis of New Triptycene-Based Ligands
Extension of triptycene’s phenyl rings would then provide an
electronic environment where the orbitals involved in conjuga-
tion have a similar direction to other conjugated beltlike
molecules. The shape of triptycene also hinders packing into
dense structures, particularly when appended to larger structures.
Along with its derivatives, triptycene has been used to inhibit
aggregation in conjugated polymers and to align liquid crystals.7
Chen has also used triptycene to form crown ether hosts that
can complex a variety of ammonium cations.8
Munakata has demonstrated that silver(I) can be π-coordi-
nated with triptycene to form extended frameworks.9 Apart from
this, the use of triptycene and its derivatives as ligands for the
generation of porous structures is virtually unexplored. We have
demonstrated the use of triptycene-based ligands to prepare
highly porous 2-D metal-organic frameworks.10 These struc-
tures, formed with CuI, exhibit high chemical and thermal
stability and can be used to remove organic impurities from
contaminated water. They undergo reversible guest exchange
without decomposition of the framework. Coordination of metals
to the apical nitrogen atoms of macrocycle 1 could allow for
their assembly into stable metal-organic nanotubes with similar
properties.
In this paper, we describe the use of triptycenyl o-quinone
to form a family of shape-persistent ligands. Among the
structures prepared, we present a first generation shape-persistent
dendrimer and several components of macrocycle 1 that we are
targeting. Synthesis, characterization, and structures of the novel
phenazine- and quinoxaline-containing compounds are de-
scribed. These ligands, which incorporate one to four triptycenyl
groups, are shape-persistent structures themselves and are
promising candidates for the development of robust coordination
frameworks with large free volume.
KNO3 reduced the amount of unreacted starting material but
led to increased amounts of the undesired dinitrotriptycene; the
most desirable method is to use only stoichiometric KNO3.
However, we found that it was faster to use nitric acid in acetic
acid combined with heating as this reaction was complete after
6 h.
Compound 4 is difficult to purify by recrystallization, and
chromatography is not a suitable technique for scaled-up
reactions because the mobility of 4 on silica gel requires dry
loading. It is preferable to first reduce the mixture of compounds
and then separate 5 from triptycene and diaminotriptycene as
all these compounds have significant differences in their
mobilities on silica. In situ conversion of the amino group to
an acetamide followed by selective nitration at the other â
position at low temperature with stoichiometric nitric acid
(generated in situ using potassium nitrate and para-toluene-
sulfonic acid) gave compound 6. Compound 6 was isolated as
a yellow solid, and the acetate protecting groups were easily
removed by basic hydrolysis to give 7 as an orange-yellow solid.
The colors of 6 and 7 arise from the push-pull chromophore
between the nitro and amino substituents. Reduction of 7 then
afforded air-sensitive 2 as a white crystalline solid in 53%
yield.10
Hexaaminotriptycene (3) has been reported in the literature,
prepared by nitration of trinitrotriptycene (45% yield) followed
by reduction.11 The low yield of the nitration is due to the
formation of byproducts resulting from reaction at undesired
locations under the harsh conditions of the reaction. This is also
problematic because the byproducts are similar and are difficult
to separate, even by chromatography. To avoid these problems,
we synthesized 3 by a sequential double nitration/reduction route
analogous to that used to prepare compound 2 (Scheme 2).
Triple nitration of triptycene leads primarily to a mixture of
two isomers (8a,b) in an approximate ratio of 3:1 (8a:8b), which
shows that the location of nitration is statistically controlled
and not significantly affected by electronic effects. These
isomers are difficult to separate by chromatography on a large
scale because dry loading is also required and they have similar
Rf values. Because both are suitable precursors to 3, it is possible
to carry them through the entire synthetic route without
intermediate purification, but this makes purification of the final
product more difficult. Therefore, it is preferable to reduce them
to the corresponding amines (9) and separate the two isomers
at this stage by column chromatography. Each amine isomer
can then be taken separately through the remainder of the
synthetic route to 3.10
Results and Discussion
Precursors. With the long-term goal of building a shape-
persistent macrocycle that could be assembled into a nanotube,
we planned to synthesize several quinoxaline- and phenazine-
type model compounds that represent portions of the macro-
cycle. This required us to develop synthetic routes to new
triptycenyl precursors such as 2,3-diaminotriptycene (2) and
2,3,6,7,14,15-hexaaminotriptycene (3).
The best route to compound 2 appeared to be sequential
nitration and reduction of one triptycene ring as shown in
Scheme 1. Prior kinetic studies of triptycene have shown that
the most reactive sites toward electrophilic aromatic substitution
are the desired â positions. Nitration on one phenyl ring
deactivates it toward further substitution; however, the other
two rings are not significantly deactivated, and this leads to a
mixture of products. We attempted to optimize the nitration
conditions in acetic anhydride to maximize the yield of
mononitrotriptycene 4, with the other byproducts being dini-
trotriptycene and unreacted starting material. Table 1 shows the
yields of the compounds from several reactions. Adding more
Although all of the quinoxaline-type compounds can be
synthesized by Schiff base condensation with glyoxal, obtaining
the phenazine-type compounds by this synthetic strategy requires
ortho-quinones. Quinones are typically encountered as the stable
para species because o-quinones are prone to dimerization
through a Diels-Alder reaction. o-Quinones can be isolated at
low temperatures or by incorporating bulky substituents that
prevent close approach of two molecules in the orientation
required for dimerization to take place.12 The o-quinones we
require for our target phenazines must be stable because Schiff
base condensations with ketones occur under more forcing
conditions with higher temperatures. An o-quinone bearing the
(7) (a) Yang, J.-S.; Swager, T. M. J. Am. Chem. Soc. 1998, 120, 5321-
5322. (b) Long, T. M.; Swager, T. M. AdV. Mater. 2001, 13, 601-604.
(8) (a) Peng, X.-X.; Lu, H.-Y.; Han, T.; Chen, C.-F. Org. Lett. 2007, 9,
895-898. (b) Zong, Q.-S.; Chen, C.-F. Org. Lett. 2006, 8, 211-214. (c)
Zhu, X.-Z.; Chen, C.-F. J. Am. Chem. Soc. 2005, 127, 13158-13159.
(9) Munakata, M.; Wu, L. P.; Sugimoto, K.; Kuroda-Sowa, T.; Maekawa,
M.; Suenaga, Y.; Maeno, N.; Fujita, M. Inorg. Chem. 1999, 38, 5674-
5680.
(11) (a) Shalaev, V. K.; Skvarchenko, V. R. Vestn. Mosk. UniV., Ser. 2:
Khim. 1974, 15, 726-730. (b) Shalaev, V. K.; Getmanova, E. V.;
Skvarchenko, V. R. Zh. Org. Khim. 1976, 12, 191-197.
(12) Minisci, F.; Citterio, A.; Vismara, E.; Fontana, F.; De Bernardinis,
S.; Correale, M. J. Org. Chem. 1989, 54, 728-731.
(10) Chong, J. H.; MacLachlan, M. J. Inorg. Chem. 2006, 45, 1442-
1444.
J. Org. Chem, Vol. 72, No. 23, 2007 8685

Chong and MacLachlan
SCHEME 2. Synthesis of 2,3,6,7,14,15-Hexaaminotriptycene (3)
SCHEME 3. Synthesis of Triptycenyl Quinoxaline Ligands
13 and 14
SCHEME 4. Synthesis of Triptycenyl Phenazine Ligands
15-17
TABLE 2. 1H NMR Resonances of the Bridgehead Protons for
Quinoxalines 13 and 14 and Phenazines 15-17
compound
chemical shift (ppm)a
13
14
15
16
17
5.64
6.13
5.66
5.61
5.59,b 6.05c
a Calibrated internally to residual CHCl3. b Resonance of outer bridgehead
c
protons. Resonance of inner bridgehead protons.
byproduct. This is surprising because the expected polymer
arising from reversible Schiff base condensation should shift
toward the thermodynamically stable aromatic quinoxaline. The
increased number of amino groups on 3 also increases the
possibility for forming a cross-linked polymer that would be
difficult to degrade once it precipitates out of solution.
Fortunately, it was possible to separate the desired compounds
by column chromatography using polar solvent mixtures.
We synthesized the phenazine-type ligands 15 and 16 from
quinone 12 (Scheme 4). Ligand 15 is a model compound
comprising a vertex and a complete side of macrocycle 1, and
16 represents a side and two vertices of 1 (Figure 2). THF or
THF/ethanol mixtures were initially used as we believed they
were necessary for solubility, but this resulted in low product
yields. Because ketones react less rapidly than aldehydes, the
reaction time was increased but did not produce a corresponding
increase in the yield. However, switching to ethanol resulted in
sterically bulky triptycenyl structure (12) was recently reported
by Chen.13 Compound 12 is indeed stable at ambient temper-
atures for months and can be recovered after heating to reflux
overnight in ethanol or THF, making it suitable for our use.
Ligands. The quinoxaline-based ligands (13 and 14) were
synthesized by Schiff base condensation with a glyoxal equiva-
lent that was chosen because of its availability in an anhydrous
form (Scheme 3). Ligand 13 is a model compound that
represents a vertex and part of a side of the target macrocycle
1. We expected that these reactions would be essentially
quantitative, but the yields were lowered by the formation of
unidentified byproducts. In particular, the reaction to form 14
generated a large amount of a poorly soluble oligomeric
(13) Gong, K.; Zhu, X.; Zhao, R.; Xiong, S.; Mao, L.; Chen, C. Anal.
Chem. 2005, 77, 8158-8165.
8686 J. Org. Chem., Vol. 72, No. 23, 2007

Synthesis of New Triptycene-Based Ligands
FIGURE 2. Representation of ligands 13-17 as model components of macrocycle 1.
FIGURE 3. Packing arrangement of quinoxaline 13 in the solid state. (a) View along the a-axis showing a layered structure. (b) View along the
b-axis showing interdigitated cofacial assembly. (c) View of π stacked dimers.
reaction mixtures that were easy to purify and gave reasonable
yields of 61% and 68% for 15 and 16, respectively.
mass spectrometry for 17 shows the expected [M + H]+ ion,
without any observable oligomeric or polymeric impurities (see
Supporting Information).
Initial attempts to form 17 were particularly problematic as
only trace quantities of the desired product were detected.
Interestingly, the use of ethanol did not produce a significant
improvement as with 15 and 16. To deal with the possibility of
insolubility of the intermediates, we tried THF and THF/ethanol
mixtures, but this produced no significant improvement and only
resulted in the formation of polymers. Such a polymeric
byproduct may be more problematic when a ketone such as 12
is used instead of an aldehyde because the resulting imines
would be more stable, making the system less reversible and
more difficult to reach the desired thermodynamic product.
Fortunately, adding a catalytic amount of piperidine improved
the yield of 17 sufficiently to 35%, so that it could be purified
and isolated. The base serves as a catalyst for the reaction,
making it easier to reverse the formation of unwanted kinetic
byproducts. Purification of 17 was facilitated by chromatography
with an alumina stationary phase. NMR spectroscopy shows
the expected signals and also confirms that there are no
impurities present. Notably, the 1H NMR spectrum shows two
signals for the bridgehead, at 6.05 and 5.59 ppm for the inner
and outer bridgeheads, respectively (Table 2). MALDI-TOF
The chemical shift of the triptycenyl bridgehead protons is a
useful diagnostic feature for all triptycene compounds as it is
sensitive to changes on the phenyl rings. For the ligands in this
study, the chemical shifts are dependent on the number of
substituted phenyl rings adjacent to the bridgehead. Bridgeheads
with three adjacent quinoxalines or phenazines have protons
that resonate downfield by ca. 0.4 ppm compared to those that
only have a single adjacent quinoxaline or phenazine. This can
be seen in the chemical shifts for 17; the protons on its inner
bridgehead resonate 0.46 ppm downfield of those on its outer
bridgeheads. Although sensitive to the number of substitutions,
the chemical shift is not significantly affected by the presence
of a quinoxaline versus a phenazine moiety.
Structures. To study how changes in the structure of the
ligands affected their packing in the solid state, we undertook
single-crystal X-ray diffraction experiments of 13-16. In the
solid state, quinoxaline 13 packs into a layered structure, and
within each layer the molecules show interdigitated cofacial
assembly (Figure 3). The pyrazine ring lies over top of the
phenyl ring on the other molecule with a separation of 3.7 Å;
J. Org. Chem, Vol. 72, No. 23, 2007 8687

Chong and MacLachlan
FIGURE 4. Packing arrangements of trisquinoxaline 14 and phenazine 16 in the solid state. Guest solvent molecules have been omitted for clarity.
(a) View of 14 crystallized from MeCN showing orthogonal layers and π stacking. (b) View of 14 crystallized from THF showing the presence of
layers with two different orientations. (c) View of 16 along the a-axis. (d) View of 16 along the b-axis.
this allows for weak π stacking between pairs of molecules
within each layer. This appears to be a relatively efficient
packing arrangement, as there are no guest solvent molecules
in the lattice, and the structure has a calculated density of 1.35
g/cm3.
Crystals of trisquinoxaline 14 were obtained from two
different solvents, MeCN and THF. The structure of 14
crystallized from MeCN contains 12 guest acetonitrile molecules
per unit cell (1.5 molecules of MeCN per molecule of 14 as
one MeCN lies on a mirror plane) (Figure 4a). This structure
contains two layers orientated orthogonally to each other. A
similar π stacking arrangement, with the phenyl ring being over
top of the pyrazine ring separated by 3.6 Å, can be seen between
pairs of adjacent molecules within the same layer. This packing
arrangement of the quinoxaline is sufficiently poor that there is
space to accommodate MeCN in the lattice. Interestingly, the
solid-state structure of 14 crystallized from THF is completely
different (Figure 4b). In this structure, there is no guest solvent
in the lattice. This solvent-free structure has two layers that are
twisted with respect to each other by roughly 30°, allowing for
a more efficient filling of space. However, this geometry does
not allow for the π stacking interactions that may have been
required to stabilize the more loosely packed form obtained from
MeCN. The structure of 14 crystallized from MeCN has a
calculated density of 1.20 g/cm3 (1.38 g/cm3 with the presence
of solvent guests), whereas crystallizing from THF gives a
higher density of 1.42 g/cm3.
FIGURE 5. First (17) and second generations of the proposed
triptycenyl dendrimer.
Information). Again, there is weak π stacking between pairs of
molecules in the same layer with a separation of 3.8 Å. Ligand
15 differs from 13 as the pyrazine rings in 15 stack directly on
top of each other and not over top of the adjacent phenyl rings.
This form of packing leads to minimal free space in the lattice
(calculated density of 1.34 g/cm3), with no guest solvent,
demonstrating that lengthening one of the wings has little effect
on solid-state packing efficiency.
However, adding another triptycenyl moiety resulting in a
molecule with two sets of triptycenyl wings, as in 16, produces
a significant decrease in the solid-state packing efficiency. It is
clear that the structure of this compound would restrict close
packing. The molecules are arranged into two layers that are
twisted with respect to each other, presumably to fill space more
efficiently (Figures 4c and 4d). However, the presence of guest
solvent molecules shows that this arrangement is not as efficient,
with a calculated density of 1.26 g/cm3 (1.31 g/cm3 with the
The smallest triptycenyl phenazine (15) packs into a layered
structure with interdigitated cofacial assembly within each layer,
which is similar to the structure for 13 (see Supporting
8688 J. Org. Chem., Vol. 72, No. 23, 2007

Synthesis of New Triptycene-Based Ligands
FIGURE 6. Solid-state structure of the Cu-15 coordination dimer. (a) Asymmetric unit. Ellipsoids are shown at the 50% probability level. Carbon,
black; nitrogen, blue; copper, green; iodine, red. (b) View along the a-axis showing a herringbone layered structure. Guest solvent molecules have
been omitted for clarity. (c) View along the b-axis showing solvent-filled voids.
presence of solvent guests). There is also no observed π
stacking, as the extra triptycenyl moiety prevents the molecules
from approaching each other closely enough with the correct
orientation.
from quinoxaline 13 formed with copper(I) iodide.10 To
demonstrate the potential of the new molecules to coordinate
to metals, we investigated the reaction of 15 with copper salts.
Reaction of 15 with CuI in benzonitrile gave dark red crystals
of Cu-15. Surprisingly, the structure in this case is not an
extended framework but instead is a dimer with two molecules
of 15 bridged by Cu2I2 (Figure 6). The solid-state packing of
Cu-15 shares some characteristics with its ligand, as it forms a
herringbone layered structure with π stacking. The stacked
phenazines are separated by 3.6 Å, and they also have the
pyrazine rings over top of each other and the phenyl rings on
top of each other. Despite the π stacking, the dimers do not
pack together very well, with large solvent-filled voids, similar
to what is observed for Cu-13. The similarity suggests that if
the coordination can be extended the expected framework should
have a structure similar to Cu-13.
AM1 calculations of the target macrocycle 1 show that it
should have an inner diameter of 17 Å. While this is likely too
large for use in rotaxane chemistry, its size makes it ideal for
forming large shape-persistent nanotubes when stacked on top
of each other. These nanotubes would possess hydrophobic
interiors that should selectively adsorb nonpolar guests. Fur-
thermore, the π stacking interactions from the numerous
aromatic rings in 1 would be useful for selectively binding a
large number of aromatic guests within the pores. Calculations
suggest the interior of macrocycle 1 may be able to bind a single
molecule of C60 as the intermolecular distances are around 3.7
Å, allowing for π-π interactions; this behavior has been seen
Unfortunately, attempts to crystallize 17 have so far been
unsuccessful. Compound 17 can also be considered as the first
generation of a triptycenyl-based dendrimer (Figure 5); subse-
quent generations could be synthesized divergently by alternat-
ing condensations of aminotriptycenes with triptycenyl quinones.
Such dendrimers would also be expected to exhibit poor packing
in the solid state, and if the expected void spaces are guest-
free, these dendrimers may be useful as low-κ dielectric
materials for electronic applications. With a high accessible
surface area that is constructed with aromatic rings, the
dendrimers may also be promising hydrogen storage materials.
Ligands 13-15 were stable to high temperatures of 341, 442,
and 378 °C, respectively, while 16 and 17 showed mass losses
beginning at 169 and 163 °C, respectively. It appears that
thermal stability of these compounds decreases as the number
of triptycenyl wings increases. Interestingly, compounds 13 and
15, with only a single set of wings, essentially have a single
decomposition where the entire mass is lost in one step. As the
number of triptycenyl wings increases, so does the number of
decomposition steps, with 17 appearing to have three stages of
mass loss.
The new triptycenyl quinoxaline and phenazine ligands here
are potential precursors for forming highly porous, rigid
frameworks. We previously found that coordination frameworks
J. Org. Chem, Vol. 72, No. 23, 2007 8689

Chong and MacLachlan
for similar geometrical motifs.14 Their assembly into nanotubes
could be facilitated by coordination of CuI to the apical nitrogen
atoms on the top and bottom edges of the macrocycle, similar
to what has been observed for Cu-13, Cu-14, and Cu-15.10
Incorporation of CuI is expected to maintain the hydrophobicity
of the macrocycle, ideal for nonpolar guests. Moreover, metal
coordination is expected to assist in holding the nanotubes
together and should increase their rigidity and make them more
robust.
Synthesis of Triptycene Phenazine 15. To a flask containing
o-phenylenediamine (0.050 g, 0.46 mmol) and 12 (0.150 g, 0.53
mmol) was added degassed ethanol (30 mL). The solution was
refluxed under N2 for 1.5 h, and the solvent was removed by rotary
evaporation. Column chromatography of the residue on silica gel
with CH2Cl2 yielded the desired product as a pale yellow solid
1
(0.101 g, 61% yield). H NMR (300 MHz, CDCl3) δ 8.17-8.14
(m, 2H), 8.07 (s, 2H), 7.75-7.72 (m, 2H), 7.50-7.48 (m, 4H),
7.10-7.08 (m, 4H), 5.66 (s, 2H) ppm; 13C NMR (75.4 MHz,
CDCl3) δ 146.6, 143.2, 143.0, 142.9, 129.8, 129.4, 126.2, 124.1,
122.4, 53.5 ppm; MS (EI, 70 eV) m/z 356 (M+); IR ν (KBr) )
3430, 3072, 3039, 2954, 1526, 1461, 1429, 1357, 1319, 1203, 1180,
1159, 1130, 1025, 1002, 886, 859, 801 cm-1; mp > 300 °C. Anal.
Calcd for C26H16N2: C, 87.62; N, 7.86; H, 4.52. Found: C, 87.44;
N, 7.81; H, 4.61.
Conclusions
We have successfully synthesized a series of triptycene-based
ligands containing pyrazine groups for use in forming coordina-
tion frameworks and as model compounds toward target shape-
persistent dendrimers and macrocycles. Their packing arrange-
ment in the solid state has been studied by X-ray crystallography,
and it was found that an increase in the number of triptycene
groups disrupted intermolecular interactions that lead to efficient
packing. We have also synthesized a rigid, shape-persistent first
generation dendrimer based on this chemistry. We will continue
to expand our library of triptycenyl ligands and further
investigate their use in forming coordination frameworks.
Synthesis of Triptycene Phenazine 16. To a flask containing 2
(0.100 g, 0.35 mmol) and 12 (0.110 g, 0.39 mmol) was added
degassed ethanol (50 mL). The solution was refluxed under N2 for
1.5 h, and the solvent was removed by rotary evaporation. Column
chromatography of the residue on silica gel with CH2Cl2 yielded
the desired product as a pale yellow solid (0.128 g, 68% yield). 1H
NMR (300 MHz, CDCl3) δ 8.00 (s, 4H), 7.47-7.44 (m, 8H), 7.07-
7.05 (m, 8H), 5.61 (s, 4H) ppm; 13C NMR (75.4 MHz, CDCl3) δ
145.8, 143.4, 142.3, 126.1, 124.1, 122.4, 53.5 ppm; MS (EI, 70
eV) m/z 532 (M+); IR ν (KBr) ) 3071, 3037, 2955, 1530, 1461,
1425, 1347, 1315, 1245, 1210, 1201, 1166, 1157, 1071, 1026, 878,
799, 763, 754, 739, 724 cm-1; mp > 300 °C. Anal. Calcd for
C40H24N2: C, 90.20; N, 5.26; H, 4.54. Found: C, 89.87; N, 5.66;
H, 4.73.
Experimental Section
Synthesis of Triptycene Quinoxaline 13. To a solution of
compound 2 (0.814 g, 2.86 mmol) in THF (150 mL) was added
2,3-dihydroxy-1,4-dioxane (0.391 g, 3.26 mmol). The solution was
stirred under N2 overnight to give a yellow solution. The solvent
was removed by rotary evaporation, and the red-orange residue was
chromatographed on silica gel with 1:9 ethyl acetate:dichlo-
romethane to give the product as an off-white solid (0.587 g, 67%
Synthesis of Triptycene Phenazine 17. To a flask containing 3
(0.025 g, 0.07 mmol) and 12 (0.100 g, 0.35 mmol) was added a
degassed mixture of ethanol (40 mL) and 6 drops of piperidine.
The solution was refluxed overnight under N2, and the solvent was
removed by rotary evaporation. Chromatography of the residue on
an alumina column with 1:9 EtOAc:CH2Cl2 gave the desired product
1
1
yield). H NMR (300 MHz, CDCl3) δ 8.70 (s, 2H), 7.99 (s, 2H),
as a pale yellow solid (0.028 g, 35% yield). H NMR (300 MHz,
7.47-7.44 (m, 4H), 7.07-7.04 (m, 4H), 5.64 (s, 2H) ppm; 13C
NMR (75.4 MHz, CDCl3) δ 146.7, 144.2, 143.7, 142.0, 126.0,
124.1, 122.9, 53.6 ppm; MS (EI, 70 eV) m/z 305 (M+); IR ν (KBr)
) 3423, 3056, 2959, 1470, 1459, 1433, 1364, 1174, 1157, 1069,
1029, 940, 899, 769, 742, 629, 577, 496 cm-1; mp > 300 °C. Anal.
Calcd for C22H14N2: C, 86.25; H, 4.61; N, 9.14. Found: C, 85.90;
H, 4.78; N, 9.41.
CDCl3) δ 8.23 (s, 6H), 7.99 (s, 6H), 7.45-7.41 (m, 12H), 7.06-
7.03 (m, 12H), 6.05 (s, 2H), 5.59 (s, 6H) ppm; 13C NMR (75.4
MHz, CDCl3) δ 146.4, 143.2, 142.75, 142.69, 142.4, 126.2, 124.1,
124.0, 122.4, 53.5, 52.9 ppm; MS (MALDI, dithranol) m/z 1089.7
([M + H]+); IR ν (KBr) ) 3438, 3072, 3043, 3024, 2961, 1526,
1492, 1459, 1422, 1348, 1271, 1223, 1160, 883, 733 cm-1; mp >
300 °C. Anal. Calcd for C80H44N6‚2H2O: C, 85.39; N, 7.47; H,
4.30. Found: C, 85.59; N, 7.52; H, 4.48.
Synthesis of Coordination Dimer Cu-15. A mixture of 15 and
copper(I) iodide in benzonitrile was refluxed for 2 h. Slow cooling
to room temperature afforded dark red crystals of suitable quality
for X-ray diffraction. Anal. Calcd for C73H47N2Cu2I2 (Cu-15‚3PhCN
[composition confirmed by TGA]): C, 62.49; N, 6.99; H, 3.38.
Found: C, 62.48; N, 6.92; H, 3.48.
Synthesis of Triptycene Quinoxaline 14. To a solution of
compound 3 (0.319 g, 0.93 mmol) in THF (120 mL) was added
2,3-dihydroxy-1,4-dioxane (0.334 g, 2.78 mmol). The solution was
stirred under N2 overnight to give an orange solution. The solvent
was removed by rotary evaporation, and the brown residue was
chromatographed on silica gel with 1:9 methanol:dichloromethane
1
to give the product as an off-white solid (0.112 g, 29% yield). H
NMR (300 MHz, CDCl3) δ 8.76 (s, 6H), 8.21 (s, 6H), 6.13 (s,
2H); 13C NMR (75.4 MHz, CDCl3) δ 144.9, 143.7, 142.2, 124.4,
52.8 ppm; MS (EI, 70 eV) m/z 410 (M+); IR ν (KBr) ) 3433,
3047, 1474, 1430, 1366, 1176, 1078, 1025, 934, 901, 747, 577
cm-1; mp > 300 °C. Anal. Calcd for C26H14N6: C, 76.09; H, 3.44;
N, 20.48. Found: C, 75.87; H, 3.57; N, 20.36.
Acknowledgment. We thank NSERC of Canada for funding.
J.H.C. thanks NSERC for a PGSD graduate fellowship. We also
thank Brian Patrick for assistance with XRD.
Supporting Information Available: General experimental
details, NMR spectra, TGA curves, and X-ray diffraction data (CIF).
This material is available free of charge via the Internet at
http://pubs.acs.org.
(14) (a) Kawase, T.; Kurata, H. Chem. ReV. 2006, 106, 5250-5273. (b)
Veen, E. M.; Feringa, B. L.; Postma, P. M.; Jonkman, H. T.; Spek, A. L.
Chem. Commun. 1999, 1709-1710.
JO701297Q
8690 J. Org. Chem., Vol. 72, No. 23, 2007