A branched, hydrogen-bonded heterodimer: a novel system for achieving high stability and specificity

Article abstract of DOI:10.1016/j.tet.2007.04.043

Presented is a description of the design, synthesis, and characterization of a novel, branched, six hydrogen-bonded heterodimer termed 'trident'. Two branched oligoamide molecules, 1 and 2, with complementary hydrogen-bonding sequences 3(AD) and 3(DA), re

Full text of DOI:10.1016/j.tet.2007.04.043

Tetrahedron 63 (2007) 5460–5469
A branched, hydrogen-bonded heterodimer: a novel system for
achieving high stability and specificity
*
Jason B. Bialecki, Li-Hua Yuan and Bing Gong
*
Department of Chemistry, University at Buffalo, State University of New York, Buffalo, NY 14260-3000, USA
Received 9 February 2007; revised 10 April 2007; accepted 12 April 2007
Available online 20 April 2007
Abstract—Presented is a description of the design, synthesis, and characterization of a novel, branched, six hydrogen-bonded heterodimer
termed ‘trident’. Two branched oligoamide molecules, 1 and 2, with complementary hydrogen-bonding sequences 3(AD) and 3(DA), respec-
tively, were found to form a very stable (Kaꢀ9ꢁ106 Mꢂ1) heterodimer in chloroform. Confirmation of the high stability of the heterodimer was
obtained through 1D and 2D 1H NMR spectroscopy, thin-layer chromatography (TLC), and vapor pressure osmometry (VPO). The stability of
the trident occurs through the cooperative action of six programmed hydrogen-bonding interactions, which are facilitated by the pre-organi-
zation of the individual strands through intramolecular hydrogen bonding.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
DNA is the molecule with the most readily predictable and
programmable intermolecular interactions.
Recently, there has been intense interest by chemists to con-
struct self-assembling aggregates that will lead to supramo-
lecular structures.1,2 In nature supramolecular systems are
made up of multi-component architectures that assemble
through highly specific intermolecular interactions.3 These
highly specific interactions are from the cooperative action
of numerous non-covalent forces such as Van der Waals
interactions, steric effects, and, most prominently, hydrogen
bonding. Nanoscale structures are commonplace throughout
nature. However, the preparation of artificial nanostructures
presents one of the most daunting challenges to modern
chemists and biochemists alike. Based on these non-covalent
forces, numerous artificial self-assembly systems have been
developed.4–11
Seeman14 has taken advantage of the programmable se-
quence-specificity of DNA and has reported an elegant
strategy for the construction of a variety of DNA-based
nanostructures. Conjugating these DNA-based nanostruc-
tures to other structural units such as peptides and unnatural
structures could implement specific intermolecular inter-
actions but synthetically this would be a great challenge.
Besides, DNA must be used in aqueous media. Many appli-
cations, particularly in materials science, are not compatible
with aqueous media. Unnatural H-bonding units with high
sequence-specificity should greatly facilitate the design of
a variety of nanostructures. In fact, the design of unnatural
H-bonding structural motifs that demonstrate DNA-like pro-
grammable sequence-specificity has been an area of intense
interest. The development of a diverse set of molecular
subunits as associating units for supramolecular assembly
represents a continuing challenge.15
The most well known and most intensely studied example of
self-assembly through molecular recognition is the DNA
double helix.12 The pairing of complementary nucleobases
through ‘sticky ended’ association is the key driving force
for the formation of the DNA double helix.13 The formation
of this secondary structure is entropically unfavorable.
However, this energy loss is made up for by the cooperative
action of many enthalpically favorable H-bond interactions
between complementary nucleobases. This sequence-speci-
ficity between nucleobase pairs is so powerful that duplex
Meijer et al.16 and Zimmerman and Corbin17 have reported
quadruply hydrogen-bonded heterocyclic complexes with
arrays of H-bond donors (D) and acceptors (A). The dimer-
ization between the reported heterocycles shows high
specificity. However, these systems are often complicated
by tautomerism, which is usually associated with nitrogen-
based heterocycles.
Keywords: Hydrogen-bonding; Heterodimer; Self-assembly; Oligoamide;
NMR.
* Corresponding authors. Fax: +1 716 645 6963 (B.G.); fax: +1 201 216
8240 (J.B.B.); e-mail addresses: crotalus97@yahoo.com; bgong@
buffalo.edu
Another feature previously reported and commonly ob-
served in H-bonded complexes is secondary electrostatic in-
teractions.18,19 Such interactions can display either repulsive
or attractive forces between adjacent H-bonding sites in
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.04.043

J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
5461
host–guest complexes. These non-covalent forces can dras-
tically change the binding strength in complexes with the
same number of H-bonding sites. On the basis of differences
in secondary electrostatic interactions, triply hydrogen-
bonded complexes of DDA$AAD arrays showed roughly
a 10-fold increase in stability over triply hydrogen-bonded
complexes of DAD$ADA arrays.
This pre-organization is driven by intramolecular H-bonds
such as the S(6) type, which occurs adjacent to the aromatic
ring, and the five-membered type, which occurs within the
glycine linker.
The rigidity and flatness of each oligoamide strand facilitate
the alignment of the programmed donor/acceptor sites and
maintain that they work cooperatively. It is possible that ei-
ther strand (A or B) could self-dimerize, producing at best,
a four hydrogen-bonded homodimer. However, due to the
cooperative nature of the hydrogen bonds, the association
constant of the six hydrogen-bonded duplex is five orders
of magnitude stronger than that determined for the four
H-bonded homodimers. The six H-bonded duplex (A$B) is
so greatly favored that no homodimers (A$A or B$B) were
detectable.23
A new class of H-bonded duplexes free of secondary interac-
tions and tautomeric effects has been developed.20 These
H-bonded duplexes have unnatural backbones and are
characterized by adjustable strength and programmable
sequence-specificity. This is the first systematic design of
unnatural molecular recognition systems with programmable
strength and specificity similar to those demonstrated by
DNA. These duplexes are synthesized by linking residues de-
rived from 1,3-diaminobenzene, 1,3-benzenedicarboxylic
acid, and 3-aminobenzoic acid with one or more amino acid
linkers (usually glycine) to generate oligoamides with vari-
ous numbers and sequences of H-bonding sites. Ether and es-
ter side chains are incorporated into the molecular structure
to adjust solubility. Intramolecular H-bonds, especially the
S(6) type21 and the unconventional aromatic H-bond
(Fig. 1), are engineered into the oligoamide strands to pre-
organize the molecules and facilitate binding. An example
of this system is displayed in Figure 2. A duplex strand is
designed to sequence-specifically pair with another strand of
its complementary sequence. Depending on the H-bonding
sequence, either homodimers or heterodimers will form in
solution. The binding strength of duplexes greatly increases
with an increase in the number of programmed intermolecu-
lar hydrogen bonds.22 Duplexes demonstrate high specificity
in their formations by including all possible arrays that are
allowable by the amino acid linkers.
The design of the new branched heterodimer ‘trident’ rests
on the same design of the duplex, with the replacement of
the rigid glycine linker with more flexible cyclohexane-
based linkers. Figure 3 displays the structure and expected
3D arrangement for trident 1$2. Although some rigidity
will likely be lost due to the less rigid linker, the branched
oligoamides should still pre-organize in such a way to facil-
itate binding. The benefits of the new linker system will al-
low for a third dimension of stability to be added and for the
potential of designing an all donor/acceptor complementary
heterodimer, which has not yet been obtained with this
motif. Also, the design of these branched oligoamides will
further discourage self-dimerization. Figure 4 shows that
branched oligoamides 1 and 2 can at best self-dimerize
to produce two H-bonded homodimers such as 1$1 or
2$2. Two H-bonded homodimers similar to this have
been reported to have very low association constants
A six hydrogen-bonded molecular duplex (Fig. 2), which
formed from a 1:1 ratio of oligoamides A and B, was re-
ported23 to be extremely stable in chloroform. For high
stability to be achieved, each oligoamide strand must pre-or-
ganize into a ridged, flat structure prior to duplex formation.
g
f
H
CH3
O
e
2
j
H h
H
H
H
CH3H
O
R
O
i
H
H
N
C6H13
N
O
N
H
R
d
C6H13
C7H15
H
c
R
R
=
N
H
O
b
a
O
H
H
H
H
H
O
H
N
H
H
1 2
H
N
H
O
l
k
O
H
m
H
O
O
H
n
'
=
R'
N
H
N
H
C7H15
r
N
H
CH3
O
R'
H
H
O
o
O
Hs
OH
H
O
H
q
C7H15
p H
1
C7H15
Figure 1. (a) S(6) type H-bonding in salicylic acid (15–20 kJ/mol); (b)
unconventional aromatic H-bonding in N-phenylacetamide (2–4 kJ/mol).
H
O
2
C8H17
O
H
O
O
C7 H15
C7H15
H
O
O
O
O
H
H
H
H
O
N
O
O
H
N
O
H
O
O
1 2
N
H
N
H
N
H
N
H
H
O
H
A B
O
O
H
N
O
O
O
O
H
N
H
O
1
N
H
N
H
N
H
N
H
O
O
H
H
O
O
C7H15
C7 H15
C7H15
C7H15
Figure 2. The six H-bonded heterodimer duplex A$B with a Ka ofw109 Mꢂ1
in CHCl3.
Figure 3. The chemical structure and spatial model of the six H-bonded
trident 1$2.

5462
J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
C7H15
The complementary sequences of 1 and 2 are 3(AD) and
p
H
H
O
q
3(DA). We asked if the branched complementary strands 1
and 2 would dimerize effectively and if they would bind
more or less tightly than the six hydrogen-bonded duplex
A$B.
H
s
O
H
o
H
H
R'
r
N
C7H15
R'
N
H
H
n
H
H
O
m
k
l
1•1
C7H15
k
H
H
O
mH
H
N
H
l
2. Results and discussion
2.1. Synthesis
n
R'
N
H
r
H
R'
o
O
q
H
H
O
s
H
H
p
C7H15
Oligoamides 1 and 2 were synthesized by replicate coupling
steps by conversion of the corresponding carboxylic acid to
an acid chloride and coupling it to the appropriate amine.
Synthesis of oligoamide 1 (Scheme 1) began with the alkyl-
ation of methyl salicylate with 1-bromooctane to afford
product 1a, which was then hydrolyzed to the acid 1b and
then nitrated with ammonium nitrate and sulfuric acid to
yield the 5-nitro acid 1c. The 5-nitro acid 1c was converted
into the corresponding acid chloride using trimethylacetyl
chloride, and then treated with octylamine to yield the ni-
trated amide 1d, which was then catalytically hydrogenated
to produce amine 1e. Amine 1e of 3 equiv was coupled to the
triacid chloride of cis,cis-1,3,5-cyclohexanetricarboxylic
acid to produce oligoamide 1. Oligoamide 2 (Scheme 2)
was produced through a different approach. First a Curtius
rearrangement was performed on cis,cis-1,3,5-cyclohexane-
tricarboxylic acid to yield the tricarbamate 2a, which was
then deprotected to cis,cis-1,3,5-triaminocyclohexane 2b
via catalytic hydrogenation. The triamine 2b was then
treated with 3 equiv of anisoyl chloride to produce the tri-
amide 2c, which was then nitrated with ammonium nitrate
and sulfuric acid to give the tris(5-nitro) compound 2d in
surprisingly good yield. The tris(5-nitro) compound 2d
was catalytically hydrogenated to produce triamine 2e,
which was then finally treated with octanoyl chloride to
yield oligoamide 2.
g
f
CH3
O
e
j
H
H
h
R
H
O
H
i
C6H13
N
N
H
R
H
d
c
H
H
O
b
a
H
H
2•2
a
b
c
H
H
O
H
N
d
H
R
N
H
C6H13
H
i
O
R
h
H
H
e
O
j
g
f
H
CH3
Figure 4. Likely configurations for homodimers 1$1 and 2$2. Homodimers
were undetectable via 2D NOESY studies. For 1$1 no inter-strand NOE con-
tacts were observed between protons n and k, and n and l. For 2$2 no inter-
strand NOE contacts were observed between protons a and c, and a and d.
(w25 Mꢂ1).20 Therefore, we conclude that trident 1$2 will
not be in equilibrium with homodimer formation(s). In
fact, as mentioned in Figure 4 homodimers were un-
detectable through 2D NMR studies.
The linkers adopted for the synthesis of branched oligo-
amides 1 and 2 were cis,cis-1,3,5-cyclohexanetricarboxylic
acid and cis,cis-1,3,5-triaminocyclohexane, respectively.
BrC8H17
Na / MeOH
OH
OC8H17
OC8H17
1M. NaOH / MeOH
90%
NH3NO3 H SO
/
2 4
71%
84%
COOCH3
COOCH3
COOH
1a
1b
1).
(CH3)3COCl
OC8H17
OC8H17
COOH
/ DCM
H
2). NH2C8H17 / TEA
H2, Pd/C
94%
N
O2N
C8H17
N
O2
70%
O
1d
1c
OC8H17
HN
C8H17
COOH
COOH
SOCl2
O
HOOC
O
O
N
C8H17
O
H
CO
Cl
H
N
O
N
H
ClOC
C8H17
CO
Cl
N
OC8H17
O
TEA / DCM
H
H
46%
N
1
OC8H17
H2 N
C8 H17
O
1e
O
NH
C8H17
Scheme 1. Synthesis of branched oligoamide 1.

J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
5463
O
,
H2 Pd / C
DPPA, Benzyl alchohol
Ph
COOH
COOH
O
EtOH
90%
TEA / Benzene
52%
O
HOOC
NH
O
Ph
O
N
H
N
H
O
2a
Ph
OCH3
COCl
H3CO
HN
1)
2)
NH4NO3 / H2SO4
75%
NH2
NH2
TEA / DMSO / THF
H2N
OCH3
OCH3
O
H2, Pd / C
EtOH
95%
59%
H
N
NH
2b
O
O
2c
O
C7H15
NH2
CH3
N
C7H15COCl
TEA / DMSO / DCM
H
H3CO
HN
H3CO
HN
CH3
O
40%
O
OCH3
O
OCH3
O
NH
H
N
NH
H
N
O
O
O
O
NH2
N
H2N
O
N
O
2e
C7H15
C7H15
2
H
H
Scheme 2. Synthesis of branched oligoamide 2.
2.2. Characterization of the hydrogen-bonded
heterodimer
However, the pre-organization of 2 appears to be much better
as reflected by the upfield shift of Hc prior to the mixing of 1
and 2. Here the S(6) intramolecular H-bonds are toward the
center of the molecule where structural rigidity is most
important.
Trident 1$2 was prepared by mixing a 1:1 molar equivalent
of strand 1 and strand 2 in CDCl3 (1 mL). Strand 2 had a poor
solubility of <5 mM in CDCl3 while strand 1 was quite sol-
uble (w50 mM). A 1:1 mixture of single strands 1 and 2 both
became very soluble (>>50 mM) in CDCl3, which was sim-
ilar to the results previously reported for duplex A$B.23 This
observation indicated that the heterodimer formed. The for-
mation of the heterodimer 1$2 shields the polar amide
groups of each strand from the non-polar solvent and thus al-
lows the solubility to drastically increase. Efforts to obtain
dimeric crystals for X-ray crystallography were unsuccess-
ful. Thus far, crystal structures have only been reported for
homodimer duplexes.20
To determine the association constant for 1$2 1H NMR bind-
ing studies were carried out in CDCl3, by diluting a 1:1 mix-
ture of 1$2 from 10 to 0.01 mM. However, no significant
changes in the chemical shifts of aniline NH signals upon di-
lution (Hl moved 0.012 ppm upfield, Hc moved 0.004 ppm
upfield) were observed. Assuming a 10% dissociation of
1$2 at 0.01 mM,16 the association constant was conserva-
tively estimated leading to a lower limit of ꢀ9ꢁ106 Mꢂ1
for the trident 1$2. Our findings are comparable to other as-
sociation constants that have been reported for duplexes with
up to six hydrogen bonds.22
2.3. 1H NMR spectroscopy
1H NMR binding studies were also carried out in 5%
DMSO-d6/CDCl3 and 10% DMSO-d6/CDCl3 for 1$2. Table
1 shows a list of the various association constants determined
for trident 1$2. The trident is substantially weakened by the
addition of small amounts of DMSO suggesting that it would
not exist in polar solvents such as methanol or water. This
trend, however, is consistent for molecular duplexes.
1H NMR spectra of 1, 2, and 1$2 in CDCl3 are shown in
Figure 5. All signals were first assigned and then confirmed
by assigning NOE intramolecular contacts obtained from
2D NMR (NOESY) (data not shown). Figure 5 reveals sig-
nificant downfield shifts of the aniline NH signals. At
2 mM, Hl in 1 appeared at 9.01 ppm, the same proton in
1$2 at 24 ꢃC appeared at 9.71 ppm. At 2 mM, Hc in 2 ap-
peared at 7.38 ppm, whereas Hc in 1$2 at 24 ꢃC appeared
at 9.80 ppm. It is worth noting that the signal of Hl is shifted
significantly downfield before the mixing of 1 and 2. This
suggests that 1 may not acquire a ridged conformation prior
to dimerization and thus may be forming small aggregates. A
closer look at the structure of 1 reveals that the stabilizing
S(6) intramolecular H-bonds are on the outer portions of
the branched molecule. The interior alignment of the mole-
cule may not be completely achieved because it relies on the
weak interactions of the unconventional aromatic H-bonds.
The 2D NMR (NOESY, CDCl3, 500 MHz) studies on 1$2
provided the most conclusive evidence for formation of
the trident in solution. The formation of the trident should
lead to the observation of seven NOE inter-strand contacts
as represented in Figure 6. The 2D NMR studies were per-
formed at variable temperatures. The first study was per-
formed at room temperature (24 ꢃC), which led to the
following results. In Figure 7 the NOESY spectra of 1$2 at
24 ꢃC shows cross-strand NOE contacts between protons
a and k, a and l, c and n, and d and n. Figure 8 shows the

5464
J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
Figure 5. 1H NMR (500 MHz, CDCl3) spectra: (a) 2 mM of 2 at 24 ꢃC; (b) 2 mM of 1 at 24 ꢃC; (c) 2 mM of 1$2 at 24 ꢃC; (d) 2 mM of 1$2 at 45 ꢃC.
Table 1. 1H NMR dilution studies of trident 1$2 in CDCl3 and in DMSO-d6/
CDCl3
Concn CDCl3
range
5% DMSO-d6/CDCl3 10% DMSO-d6/CDCl3
0.05–10.0 mM
0.01–10.0 mM 0.05–5.0 mM
Unit
Hl
Hc
Ka (Mꢂ1
ꢀ9ꢁ106
ꢀ9ꢁ106
)
Ka (Mꢂ1
5.0ꢁ104
1.1ꢁ104
)
Ka (Mꢂ1
1.1ꢁ103
1.2ꢁ103
)
CHg3
O
f
H
e
j
h
H
O
H
H
R
H
i
C6H13
N
N
R
d
H
c
b
a
H
H
O
H
H
H
1 2
k
H
O
O
m
H
N
H
H
l
H
n
R'
C7H15
N
H
r
R'
o
O
H
H
s
q
H
p
C7H15
Figure 7. NOESY spectra of 1$2 (2 mM in CDCl3 at 24 ꢃC, 500 MHz, mix-
ing time 0.5 s) showing cross-strand contacts between protons a and k, a and
l, c and n, and d and n.
Figure 6. The seven proposed and observed inter-strand NOE contacts for
trident 1$2.
cross-strand NOE contact between protons b and l. The
NOESY spectra of 1$2 at 24 ꢃC gives substantial evidence
for the formation of the trident dimer, but two predicted
cross-strand NOE contacts between protons, c and m, and
b and m could not be observed because of signal overlap.
Because of this signal overlap, variable temperature 1H
NMR (500 MHz) experiments of 1$2 were performed in
CDCl3 from ꢂ10 to ꢂ2 ꢃC at 10 mM and again from 24 to
50 ꢃC at 2 mM. Figure 9 shows the partial 1D spectra corre-
sponding to the aromatic region of 1$2 obtained from NMR

J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
5465
Figure 8. NOESY spectra of 1$2 (2 mM in CDCl3 at 24 ꢃC, 500 MHz, mix-
ing time 0.5 s) showing cross-strand contacts between protons b and l.
variable temperature experiments. As temperature increased,
the resolution of the 1D NMR spectra improved. At 45 ꢃC,
sufficient resolution of all signals was obtained in the 1D
spectrum.
At 45 ꢃC, some of the newly resolved signals in the aromatic
region looked unfamiliar and had to be assigned using their
observed intramolecular NOE contacts. All 10 signals in the
aromatic region were found to have at least one NOE contact
with a previously confirmed 1H signal.
Results of the 2D NMR (NOESY, CDCl3, 500 MHz, 45 ꢃC,
mixing time 0.5 s) showed cross-strand NOE contacts
between protons m and c (Fig. 10), and protons b and m
(Fig. 11) as well as the five previous cross-strand NOE
contacts that were observed at 24 ꢃC. In total, all seven
cross-strand NOE contacts that were proposed in Figure 6
for trident 1$2 were observed.
Another interesting feature that was observed in the 1H NMR
variable temperature study was the dramatic change in broad-
ness of signal Hl. Hl was unexpectedly broad at room temper-
ature and continued to broaden with increased temperature
(Fig. 7). When temperatures were lowered Hl would sharpen.
The explanation for this observation is the interconversion of
the chair conformation of the cyclohexane rings. The most
stabile conformation of the cyclohexane ring is the chair con-
formation. However, it is known that even at room tempera-
ture the cyclohexane ring undergoes rapid interconversion
between chair conformations (ring-flip). When temperature
increases, this process becomes more rapid. This conforma-
tional change would affect the interior H-bonds, i.e., the three
H-bonds that are close to the cyclohexane ring, more than it
does to the other three exterior H-bonds. This conformational
flexibilitywouldoscillatetheinteriorNHgroupsandcarbonyl
O atoms, which would cause the corresponding NMR signal
to broaden. We would then expect the Ka value based on
Figure 9. 1H NMR (500 MHz, CDCl3) spectra of the aromatic region of 1$2
at variable temperatures. All signals in the aromatic region were resolved at
45 ꢃC.
the concentration-dependent changes of Hl to be lower than
that obtained from Hc. In fact the difference is about five-
fold in 5% DMSO-d6/CDCl3 (Table 1).
2.4. TLC studies
The dimerization of 1$2 was further confirmed by straight-
phase thin-layer chromatography (TLC) analysis (silica

5466
J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
Figure 10. NOESY spectra of 1$2 (2 mM in CDCl3 at 45 ꢃC, 500 MHz,
mixing time 0.5 s) showing cross-strand contacts between protons m and c.
Figure 12. Thin layer chromatography (TLC) of 2, 1$2, and 1 from left to
right. TLC developing conditions were 25% acetone in DCM. The TLC
plate was viewed under short wavelength UV light.
gel, 3:1 DCM/acetone). As shown in Figure 12, the presence
of trident 1$2 is demonstrated by the different Rf values of 1
(Rf¼0.12), 2 (Rf¼0.11), and 1$2 (Rf¼0.46). Dimerization of
1$2 substantially increases the Rf value by shielding the po-
lar amide groups from the solid phase. The reason that 1$2
displays tailing on the TLC plate indicates that it is partially
dissociated under the relatively polar assaying conditions.
The fact that 1$2 could be detected under such polar condi-
tions is a direct confirmation of the very high stability shown
by this trident.24 Trident 1$2 had to be developed using
slightly less polar solvent conditions than for duplex A$B
(10% DMF/CHCl3).23 This is evidence that 1$2 has a lower
binding constant than duplex A$B.
2.5. VPO studies
Using polystyrene as a molecular weight standard, vapor
pressure osmometry (VPO) studies of a 1:1 mixture of 1
and 2 at ambient temperature in CHCl3 were performed
over the concn range of 2–20 mM. The average value ob-
tained over this concn range was a molecular weight
(MW) of 2260.25 This value is consistent with the calculated
MW of 1+2 (2247 g/mol), which suggests the formation of
a tightly bound dimer between 1$2.
3. Conclusion
Experimental results conclude that the designed comple-
mentary strands 1 and 2 were successful in forming the
highly stable dimer 1$2 in chloroform. Although consider-
able flexibility exists in the structures of 1 and 2—the six
separate points of free rotation (N(H)–Ccyclohexane and
C(]O)–Ccyclohexane bonds) and the ring-flip of the cyclohex-
ane linkers—these two molecules were still able to undergo
molecular recognition of one another and bind at all six pro-
grammed hydrogen-bonding sites. Evidence from TLC and
1H NMR studies suggests that branched dimer 1$2 does
not bind as tightly as the six H-bonded duplex A$B men-
tioned above. The varying factor between these two is the
unit linker. The binding strength of the duplex and trident re-
lies on two aspects. First, the number of programmed H-
bonding donor/acceptor sites and second, the rigidity and
pre-organization of each strand. Overall, we conclude that
glycine works better than cyclohexanetricarboxylic acid to
pre-organize and rigidify the strands and therefore is the
main reason why duplex A$B binds more tightly than trident
Figure 11. NOESY spectra of 1$2 (2 mM in CDCl3 at 45 ꢃC, 500 MHz,
mixing time 0.5 s) showing cross-strand contacts between protons b and m.

J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
5467
C7H15
C7H15
H
4
were recorded from a Varian VXR 400 spectrometer
(400 MHz). H NMR data are reported in parts per million
H
1
O
O
O
O
C7H15
C7H15
H
H
H
H
H
R''
downfield from tetramethylsilane (TMS). All coupling con-
stants (J¼values) are reported in hertz (Hz). All 13C NMR
spectra are decoupled and were recorded on a Varian VXR
400 spectrometer (100.6 MHz). Chemical shifts are reported
relative to the central peak of CDCl3 at 77.0 ppm or the cen-
tral peak of DMSO-d6 at 39.5 ppm. CDCl3 (99.8% D) and
DMSO-d6 (99.8% D) were purchased from Cambridge Iso-
tope Laboratories.
O
O
O
O
N
H
H
H
H
N
H3C
H
R''
N
R'' =
R''' =
CH3
H
H
H
3
4
H
R'''
R'''
H
N
H
N
H
N
H
N
H
C4H9
C4H9
O
O
O
H
O
H
O
O
H
H
H
3
C8H17
O
O
C8H17
4.2. Synthesis
O
4
O
H
H
4.2.1. Methyl 2-octyloxybenzoate (1a). A mixture of methyl
salicylate (3.61 g, 16 mmol), K2CO3 (4.1 g, 30 mmol), octyl
bromide (6.80 g, 35 mmol), and DMF (40 mL) was heated at
90 ꢃC for 24 h. The solution was diluted with water
(125 mL) and extracted with EtOAc (3ꢁ20 mL). The com-
bined organic extracts were dried over MgSO4, and evapo-
rated under reduced pressure to afford the product (3.0 g,
O
H
O
3
4
H
O
H
O
3
1
71%). H NMR (400 MHz, CDCl3) d 8.09 (q, 1H, J¼8.8,
H
2.4), 7.54 (q, 1H, J¼8.8, 2.4), 6.94 (t, 1H, J¼8.8), 6.87 (t,
1H, J¼8.8), 4.24 (d, 2H, J¼6.8), 3.87 (s, 3H), 2.06 (m,
2H), 1.58 (m, 2H), 1.37 (m, 8H), 0.89 (t, 3H, J¼6.8). 13C
NMR (100.6 MHz, CDCl3) d 165.0, 157.3, 135.8, 133.9,
123.3, 117.0, 111.6, 70.1, 58.2, 31.8, 29.0, 28.7, 25.7,
22.5, 19.1. Anal. Calcd for C16H24O3: C, 72.69; H, 9.15.
Found: C, 72.63; H, 9.25.
Figure 13. The chemical structure and proposed spatial model of six
H-bonded trident 3$4.
1$2. At this point it is difficult to deduce whether the added
dimension of the trident helps to significantly increase the
binding strength.
4.2.2. 2-Octyloxybenzoic acid (1b). The ester 1a (3.09 g,
10 mmol) was dissolved in hot MeOH (40 mL), to which
1 M NaOH (20 mL, 20 mmol) was added. The mixture
was refluxed for 2 h and diluted with water (100 mL). The
aqueous solution was acidified with concd HCl (2 mL) to
precipitate the crude product, which was collected and
recrystallized from MeOH to yield a pure white solid
(2.66 g, 90%). 1H NMR (400 MHz, CDCl3) d 10.38 (s,
1H), 8.09 (q, 1H, J¼9.0, 2.8), 7.53 (q, 1H, J¼9.0, 2.8),
6.92 (t, 1H, J¼9.0), 6.87 (t, 1H, J¼9.0), 4.31 (d, 2H,
J¼6.8), 1.98 (m, 2H), 1.59 (m, 2H), 1.36 (m, 8H), 0.89 (t,
3H, J¼6.8). 13C NMR (100.6 MHz, CDCl3) d 166.7,
157.8, 134.9, 134.0, 123.0, 117.6, 113.1, 69.9, 31.7, 29.0,
28.6, 25.7, 22.6, 14.1. Anal. Calcd for C15H22O3: C, 71.97;
H, 8.86. Found: C, 71.73; H, 8.98.
Although the cyclohexane linkers do not enforce as much
rigidity as glycine they do incorporate enough rigidity to pre-
ventthe individual strands from folding prior to dimerization.
Ideas for future studies are the synthesis of a pair of all donor/
acceptor strands, which should yield a novel, six H-bonded
heterodimer as shown in Figure 13. Also it would be interest-
ing to incorporate adamantane (cis,cis-1,3,5-adamantane-
tricarboxylic acid and cis,cis-1,3,5-triaminoadamantane)
linkers emplace of the cyclohexane (cis,cis-1,3,5-cyclohexa-
netricarboxylic acid and cis,cis-1,3,5-triaminocyclohexane)
linkers to gauge the effects the cyclohexane ring-flip has on
the binding strength of the trident. The success of the 1$2
trident may potentially lead to a whole new avenue of molec-
ular design and alternative strategies for generating new
sequence-specificity.
4.2.3. 2-Octyloxy-5-nitrobenzoic acid (1c). The product
was prepared according to the literature procedure.26 The
acid 1b (2.01 g, 8.00 mmol) was dissolved in concd H2SO4
(30 mL) and cooled to 0 ꢃC in an ice bath. NH4NO3
(0.71 g,8.81 mmol)wasaddedinsmallportionsoveraperiod
of 40 min. The solutionwas pouredinto cold water (125 mL).
Awhite precipitate formed was collected, washed with water,
and dried to give 1c as a white solid (1.98 g, 84%). 1H NMR
(400 MHz, CDCl3) d 10.63 (s, 1H), 9.06 (q, 1H, J¼9.2,
2.8), 8.45 (d, 1H, J¼9.2), 7.19 (d, 1H, J¼9.2), 4.38 (t, 2H,
J¼6.8), 2.00 (m, 2H), 1.54 (m, 2H), 1.33 (m, 8H), 0.91 (t,
3H, J¼6.8). 13C NMR (100.6 MHz, CDCl3) d 165.4, 162.9,
140.0, 128.4, 126.30, 121.6, 113.3, 70.1, 31.8, 29.2, 28.9,
25.8, 22.6, 14.0. Anal. Calcd for C15H21NO5: C, 61.00; H,
7.17; N, 4.74. Found: C, 60.88; H, 7.21; N, 4.79.
4. Experimental
4.1. General
All chemicals were purchased from Aldrich or Fluka unless
specified. All products were detected as one spot on thin-
layer chromatographic analysis (pre-coated 0.25 mm silica
plates purchased from Fischer), and further characterized
by 1H NMR, 13C NMR, and elemental analysis or mass spec-
tral (MS) analysis. All elementary analyses were performed
by Atlantic Microlab, Inc. High-resolution mass spectra
were measured on a Waters-Micromass Q-TOF API-US
spectrometer equipped with a nanoelectrospray ion source.
Signals were acquired in the W-mode operation. Ion series
generated by water clusters charged by the ammonium ion
4.2.4. N-Octyl-2-octyloxy-5-nitrobenzamide (1d). The
acid 1c (3.21 g, 10.9 mmol) was dissolved in DCM
1
were used as reference mass peaks. All H NMR spectra

5468
J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
(25 mL), then TEA (2.12 g, 21.0 mmol) and trimethylacetyl
chloride (1.33 g, 11.0 mmol) were added, and the solution
was stirred for 1 h at room temperature. A second solution
of octylamine (1.70 g, 13.2 mmol) and TEA (2.12 g,
21.0 mmol) in DCM (25 mL) was placed into an ice bath
to which the first solution was added. After stirring over-
night, the solution was washed with 2 M HCl (200 mL),
dried over MgSO4, and the solvent was removed. The resi-
due was recrystallized from MeOH to afford pure 1d
was vigorously stirred and heated at 80 ꢃC until the solids
dissolved (w2 h). Benzyl alcohol (13.34 g, 123.5 mmol)
was added and the solution was refluxed for 18 h during
which a precipitate formed. After cooling to ambient tem-
perature, the product was collected under vacuum filtration,
washed with a minimal amount of cold benzene, and dried to
1
yield 6a (10.2 g, 52%). H NMR (400 MHz, DMSO-d6)
d 7.34 (m, 15H), 5.01 (s, 6H), 3.38 (m, 2H), 1.89 (d, 2H,
J¼12.0), 1.06 (q, 2H, J¼12.0). 13C NMR (100.6 MHz,
DMSO-d6) d 146.8, 128.7, 120.4, 119.9, 119.3, 113.7,
111.4, 56.7, 38.2, 37.1. ESIHRMS Calcd for C30H34N3O6
[M+H]+: 532.2442. Found: 532.2432.
1
(3.17 g, 70%). H NMR (400 MHz, CDCl3) d 9.00 (d, 2H,
J¼2.8), 8.49 (q, 2H, J¼9.2, 2.8), 7.95 (d, 2H), 7.25 (d, 2H,
J¼9.2), 4.22 (t, 2H, J¼6.6), 3.68 (q, 2H, J¼6.6), 1.83 (t,
2H, J¼6.6), 1.59 (t, 2H, J¼6.6), 1.47 (m, 20H), 0.94 (m,
6H). 13C NMR (100.6 MHz, CDCl3) d 163.4, 161.7, 142.3,
128.9, 128.2, 123.2, 113.0, 70.8, 40.6, 32.3, 32.3, 23.0,
29.8, 29.8, 29.8, 29.7, 29.6, 27.7, 26.7, 23.1, 23.1, 14.6.
Anal. Calcd for C23H38N2O4: C, 67.95; H, 9.42; N, 6.89.
Found: C, 67.80; H, 9.44; N, 6.81.
4.2.8. cis,cis-1,3,5-Triaminocyclohexane (2b). A mixture
of carbamate 2a (0.4 g, 0.76 mmol), 10% Pd/C (0.30 g),
and 95% EtOH (100 mL) was heated and shaken at 35 ꢃC
under 5 atm of hydrogen for 12 h. The Pd/C was filtered off
and the solvent was removed under reduced pressure to
give 2b as a white solid (0.09 g, 90%). 1H NMR (400 MHz,
D2O) d 3.26 (t, 3H, J¼11.6), 2.28 (d, 3H, J¼11.6), 1.44 (q,
3H, J¼11.6). 13C NMR (100.6 MHz, D2O) d 45.5, 32.5.
ESIHRMS Calcd for C6H16N3 [M+H]+: 130.1339. Found:
130.1338.
4.2.5. N-Octyl-5-amino-2-octyloxybenzamide (1e). Com-
pound 1d (1.21 g, 2.91 mmol), 10% Pd/C (0.11 g), and
MeOH (25 mL) were stirred at room temperature under
1 atm of hydrogen for 4 h. The solvent was filtered and
evaporated to afford the aryl amine (1.03 g, 94%). This
amine was used in the next step without further purification.
1H NMR (400 MHz, CDCl3) d 8.14 (s, 1H), 7.74 (s, 1H), 6.98
(s, 1H), 6.85 (s, 1H), 4.06 (t, 2H, J¼6.4), 3.46 (q, 2H, J¼6.4),
1.85 (t, 2H, J¼6.4), 1.61 (t, 2H, J¼6.4), 1.49 (t, 2H, J¼6.4),
1.32 (m, 18H), 0.90 (m, 6H). 13C NMR (100.6 MHz, CDCl3)
d 165.8, 150.7, 140.9, 122.8, 119.5, 119.0, 114.3, 70.1, 40.3,
32.3, 32.2, 29.8, 29.7, 29.6, 29.6, 29.5, 29.5, 27.7, 26.6, 23.1,
23.0, 14.4. Anal. Calcd for C23H40N2O2: C, 73.36; H, 10.71;
N, 7.44. Found: C, 73.19; H, 10.85; N, 7.36.
4.2.9. N,N0,N00-Tris(anisoyl) cis,cis-1,3,5-triaminocyclo-
hexane (2c). Compound 2b (0.07 g, 0.5 mmol) was dis-
solved in dry DMSO (5 mL) and TEA (0.5 mL). An ice
cold solution of anisoyl chloride (0.31 g, 1.8 mmol) in dry
THF (10 mL) was added. The reaction mixture was stirred
overnight. Distilled water (50 mL) was added to the reaction
mixture and a precipitate formed within minutes. The solid
was collected and washed with distilled water to yield 2c
as a white solid (0.17 g, 59%). 1H NMR (400 MHz,
CDCl3) d 8.19 (d, 3H, J¼8.0), 7.88 (d, 3H, J¼6.8), 7.43 (t,
3H, J¼8.0), 7.07 (t, 3H, J¼8.0), 6.95 (d, 3H, J¼8.0), 4.15
(m, 3H), 3.94 (s, 9H), 2.58 (d, 3H, J¼12.0), 1.59 (q, 3H,
J¼12.0). 13C NMR (100.6 MHz, CDCl3) d 164.7, 157.6,
133.0, 132.5, 121.8, 121.6, 111.6, 56.2, 46.7, 38.3.
ESIHRMS Calcd for C30H34N3O6 [M+H]+: 532.2442.
Found: 532.2431.
4.2.6. Oligoamide (1). cis,cis-1,3,5-Cyclohexanetricarb-
oxylic acid (0.12 g, 0.54 mmol) was dissolved in thionyl
chloride (4 mL) and refluxed for 3 h. The solvent was
removed under reduced pressure to yield a residue, which
was dissolved in DCM (10 mL) and added drop wise to a so-
lution of amine 1e (0.64 g, 1.70 mmol), DCM (15 mL), and
TEA (0.35 g, 3.4 mmol) and stirred overnight. The reaction
mixture was washed with 2 M HCl (20 mL), dried over
MgSO4, and evaporated to yield an amber solid, which was
recrystallized twice from MeOH and washed with EtOAc
4.2.10. N,N0,N00-Tris(2-methoxy-5-nitrobenzoyl) cis,cis-
1,3,5-triaminocyclohexane (2d). Compound 2c (0.17 g,
0.32 mmol) was dissolved into concd H2SO4 (6 mL) and
placed in an ice bath. Solid NH4NO3 (0.086 g, 1.1 mmol)
was added in small portions to the mixture over a period
of 30 min. After the addition was completed the mixture
was stirred for 15 min and poured into cold water
(100 mL). The white precipitate that formed was collected,
washed with water, and dried to give 2d as a pale gray solid
1
to afford pure 5 as a white solid (0.34 g, 46%). H NMR
(400 MHz, CDCl3) d 8.84 (s, 3H), 8.01 (t, 3H, J¼6.0), 7.97
(s, 3H), 7.87 (d, 3H, J¼9.0), 6.83 (d, 3H, J¼9.0), 4.00 (t,
2H, J¼6.0), 3.39 (q, 2H, J¼6.0), 2.45 (s, 3H), 2.13 (d, 3H,
J¼12.2), 1.91 (q, 3H, J¼12.2), 1.79 (m, 6H), 1.52 (m, 6H),
1.42 (m, 6H), 1.27 (m, 48H), 0.88 (m, 18H). 13C NMR
(100.6 MHz, CDCl3) d 174.0, 165.2, 153.7, 132.0, 126.7,
124.8, 121.6, 113.0, 69.5, 44.3, 40.3, 32.1, 32.1, 29.7, 29.7,
29.7, 29.6, 27.5, 26.5, 22.9, 14.3, 14.3. Anal. Calcd for
C78H126N6O9: C, 72.50; H, 9.85; N, 6.51. Found: C, 72.26;
H, 9.90; N, 6.26. ESIHRMS Calcd for C78H127N6O9
[M+H]+: 1291.9659. Found: 1291.9673.
1
(0.16 g, 75%). H NMR (400 MHz, CDCl3) d 8.89 (d, 3H,
J¼2.8), 8.33 (q, 3H, J¼9.2, 2.8), 7.81 (d, 3H, J¼7.2), 7.17
(d, 3H, J¼9.2), 4.18 (m, 3H), 4.10 (s, 9H), 2.52 (d, 3H,
J¼12.0), 1.59 (q, 3H, J¼12.0). 13C NMR (100.6 MHz,
CDCl3) d 163.1, 162.2, 141.3, 127.9, 126.6, 124.9, 113.0,
57.5, 46.4, 38.8. ESIHRMS Calcd for C30H31N6O12
[M+H]+: 667.1994. Found: 667.2002.
4.2.7. N,N0,N00-Tris(benzylcarbamyl) cis,cis-1,3,5-triami-
nocyclohexane (2a). The product was prepared according
to a literature procedure.27 Under nitrogen, cis,cis-1,3,5-
cyclohexanetricarboxylic acid (8.0 g, 37 mmol) was added
to a solution of benzene (400 mL), TEA (15.6 mL,
112 mmol), and DPPA (30.85 g, 112 mmol). The mixture
4.2.11. N,N0,N00-Tris(5-amino-2-methoxybenzoyl) cis,cis-
1,3,5-triaminocyclohexane (2e). Compound 2d (0.11 g,
0.17 mmol) was dissolved in DMF (0.1 mL) and EtOH
(0.25 mL), mixed with 10% Pd/C (0.02 g), and reduced un-
der 4 atm of hydrogen at 35 ꢃC for 3 h. The Pd/C was

J. B. Bialecki et al. / Tetrahedron 63 (2007) 5460–5469
5469
6. Seebach, D.; Matthews, J. L.; Meden, A.; Wessels, T.;
Baerlocher, C.; McCusker, L. B. Helv. Chim. Acta 1997, 80,
173–182.
7. Ghadiri, M. R.; Granja, J. R.; Milligan, R. A.; McRee, D. E.;
Khazanovich, N. Nature 1993, 366, 324–327.
8. (a) Castellano, R. K.; Rebek, J. J. Am. Chem. Soc. 1998, 120,
3657–3663; (b) Castellano, R. K.; Clark, R.; Craig, S. L.;
Nuckolls, C.; Rebek, J. Proc. Natl. Acad. Sci. U.S.A. 2000,
97, 12418–12421; (c) Castellano, R. K.; Nuckolls, C.;
Eichhorn, S. H.; Wood, M. R.; Lovinger, A. J.; Rebek, J.
Angew. Chem., Int. Ed. 1999, 38, 2603–2606.
removed by filtration and the solvent was removed under
a high-pressure vacuum to afford (0.90 g, 95%) amine 2e,
which was used without further purification. 1H NMR
(400 MHz, DMSO-d6) d 8.06 (d, 3H, J¼7.6), 7.09 (d, 3H,
J¼2.8), 6.88 (d, 3H, J¼9.0), 6.74 (d, 3H, J¼2.8, 9.0), 3.97
(m, 3H), 3.77 (s, 9H), 2.11 (d, 3H, J¼12.0), 1.36 (q, 3H,
J¼12.0). 13C NMR (100.6 MHz, DMSO-d6) d 165.7,
150.8, 140.8, 122.8, 119.6, 118.9, 114.5, 57.1, 46.3, 38.9.
ESIHRMS Calcd for C30H37N6O6 [M+H]+: 577.2769.
Found: 577.2775.
9. (a) Szabo, T.; Hilmersson, G.; Rebek, J. J. Am. Chem. Soc.
1998, 120, 6193–6194; (b) Rebek, J. J. Acc. Chem. Res.
1999, 33, 278–286.
10. Boucher, E.; Simard, M.; Wuest, J. D. J. Org. Chem. 1995, 60,
1408–1412.
4.2.12. Oligoamide (2). Amine 2e (0.16 g, 0.27 mmol) was
dissolved in dry DMSO (3 mL), then DCM (7 mL) and
TEA (0.32 g, 3.2 mmol) were added to the solution. Octanoyl
chloride (0.26 g, 1.6 mmol) in DCM (10 mL) was added to
the solution. The mixture was stirred overnight. Water
(50 mL) was added to the solution and the bottom layer
was extracted with DCM. The solvent was removed under
high-pressure vacuum to afford crude 2. The crude product
was purified by column chromatography (silica gel) using
11. Sanford, A. R.; Yuan, L.; Gong, B. Eur. J. Biochem. 2004, 271,
1416–1425.
12. (a) Saenger, W. Principles of Nucleic Acid Structure; Springer:
New York, NY, 1984; (b) Blackburn, G. M. Nucleic Acids in
Chemistry and Biology; Gait, M. J., Ed.; IRL: Oxford, 1990.
13. Cohen, S. N.; Chang, A. C. Y.; Boyer, H. W.; Helling, R. B.
Proc. Natl. Acad. Sci. U.S.A. 1973, 70, 3240–3244.
14. Seeman, N. C. Acc. Chem. Res. 1997, 30, 357–363.
15. Zimmerman, S. C.; Corbin, P. S. Struct. Bonding (Berlin) 2000,
96, 63–94.
16. Beijer, F. H.; Sijbesma, R. P.; Kooijman, H.; Spek, A. L.;
Meijer, E. W. J. Am. Chem. Soc. 1998, 120, 6761–6769.
17. Corbin, P. S.; Zimmerman, S. C. J. Am. Chem. Soc. 1998, 120,
9710–9711.
1
1:20 MeOH/DCM as the eluant (0.11 g, 40%). H NMR
(400 MHz, CDCl3) d 8.16 (q, 3H, J¼9.2, 2.0), 7.90 (d,
3H, J¼7.2), 7.78 (d, 3H, J¼2.0), 7.33 (s, 3H), 6.93 (d, 3H,
J¼9.2), 4.16 (d, 3H, J¼7.2), 3.92 (s, 9H), 2.56 (d, 3H,
J¼11.6), 2.34 (d, 6H, J¼7.2), 1.71 (m, 6H), 1.44 (q,
3H, J¼11.6), 1.28 (m, 24H), 0.88 (t, 9H, J¼7.2). 13C NMR
(100.6 MHz, DMSO-d6) d 172.6, 164.4, 153.7, 133.4,
125.0, 123.5, 121.9, 112.4, 56.7, 46.1, 39.1, 37.5, 32.1,
29.6, 29.4, 26.1, 23.0, 14.5. Anal. Calcd for C54H78N6O9:
C, 67.88; H, 8.25; N, 8.80. Found: C, 67.67; H, 8.13; N,
8.71. ESIHRMS Calcd for C54H79N6O9 [M+H]+: 955.5903.
Found: 955.5922.
18. Murray, T. J.; Zimmerman, S. C. J. Am. Chem. Soc. 1992, 114,
4010–4011.
19. (a) Jorgenson, W. L.; Pranata, J. J. Am. Chem. Soc. 1990, 112,
2008–2010; (b) Pranata, J.; Wierschke, S. G.; Jorgenson, W. L.
J. Am. Chem. Soc. 1991, 113, 2810–2819.
Acknowledgements
20. (a) Gong, B.; Yan, Y.; Zeng, H.; Skrzypczak-Jankunn, E.; Kim,
Y. W.; Zhu, J.; Ickes, H. J. Am. Chem. Soc. 1999, 121, 5607–
5608; (b) Gou, H.; Li, S.; Xie, D.; Gong, B. Chem. Phys.
Lett. 2005, 410, 264–268.
21. (a) Berstein, J.; Davis, R. E.; Shimoni, L.; Chang, N. L. Angew.
Chem., Int. Ed. Engl. 1995, 34, 1555–1573; (b) Etter, M. C.
Acc. Chem. Res. 1990, 23, 120–126.
We thank Athula B. Attygalle (Stevens Institute of Technol-
ogy) for granting access to his Micromass/Waters Q-TOF-
API-US mass spectrometer. This project was supported by
the National Institute of Health (NIH).
References and notes
22. (a) Zeng, H.; Ickes, H.; Flowers, R. A.; Gong, B. J. Org. Chem.
2001, 66, 3574–3583; (b) Zeng, H.; Yang, X.; Brown, A. L.;
Martinovic, S.; Smith, R. D.; Gong, B. Chem. Commun.
2003, 1556–1557.
1. Philip, D.; Stoddart, J. F. Angew. Chem., Int. Ed. 1996, 35,
1154–1196.
23. Zeng, H.; Miller, R. S.; Flowers, R. A.; Gong, B. J. Am. Chem.
Soc. 2000, 122, 2635–2644.
24. Seto, C. T.; Whitesides, G. M. J. Am. Chem. Soc. 1990, 120,
6409–6411.
2. Whitesides, G. M.; Simanek, E. E.; Mathias, J. P.; Seto, C. T.;
Chin, D. N.; Mammen, M.; Gordon, D. M. Acc. Chem. Res.
1995, 28, 37–44.
3. Lawrence, D. S.; Jiang, T.; Levett, M. Chem. Rev. 1995, 95,
2229–2260.
4. Kolotuchin, S. V.; Zimmerman, S. C. J. Am. Chem. Soc. 1998,
120, 9092–9093.
5. Klok, H. A.; Joliffe, K. A.; Schauer, C. L.; Prins, L. J.; Spatz,
J. P.; Moller, M.; Timmerman, P.; Reinhoudt, D. N. J. Am.
Chem. Soc. 1999, 121, 7154–7155.
25. The values of three separate determinations were MW¼2260,
2280, 2250.
26. De Paulis, T.; Janowski, A.; Kessler, R. M.; Clanton, J. A.;
Smith, H. E. J. Med. Chem. 1988, 31, 2027–2033.
27. Planalp, R. P.; Brechbiel, M. W. Bioorg. Med. Chem. Lett.
1996, 6, 807–810.