Hydrogen bonding control of molecular self-assembly: Aggregation behavior of acylaminopyridine-carboxylic acid derivatives in solution and the solid state

Article abstract of DOI:10.1016/S0040-4020(00)00733-X

In this paper we report the design, synthesis and investigation of a family of acylaminopyridine-carboxylic acid monomers that can aggregate through bidentate hydrogen bonding interactions between the carboxylic acid and the aminopyridine sites. We show that in solution the nature of the aggregate depends on the substitution pattern of the monomer and present evidence from NMR and mass spectrometry that in some cases cyclic association occurs. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00733-X

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 8419±8427
Hydrogen Bonding Control of Molecular Self-Assembly:
Aggregation Behavior of Acylaminopyridine±Carboxylic Acid
Derivatives in Solution and the Solid State
a
Abdullah Zafar, Steven J. Geib, Yoshitomo Hamuro, Andrew J. Carr
a
a
b
b,
*
and Andrew D. Hamilton
a
Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260, USA
Department of Chemistry, Yale University, New Haven, 225 Prospect Street, Connecticut, CT 06520, USA
b
Received 6 July 2000; accepted 22 August 2000
AbstractÐIn this paper we report the design, synthesis and investigation of a family of acylaminopyridine±carboxylic acid monomers that
can aggregate through bidentate hydrogen bonding interactions between the carboxylic acid and the aminopyridine sites. We show that in
solution the nature of the aggregate depends on the substitution pattern of the monomer and present evidence from NMR and mass
spectrometry that in some cases cyclic association occurs. q 2000 Elsevier Science Ltd. All rights reserved.
5
Introduction
and in certain cases 2:2 aggregates can be formed. As part
of a program aimed at using hydrogen bonding to control
self-assembly, we were interested in probing the design of
components containing both carboxylic acid and amino-
pyridine groups within the same framework. Our hope
was that such molecules would self-assemble in non-polar
solvents (as in Fig. 1b) with the shape and size of the aggre-
gate controlled by the relative positioning of the two hydro-
gen bonding groups. The simplest derivative in this class
would involve substituting a carboxylic acid group on 2-
acylaminopyridine at different positions, as in 1 and 2.
The advantage in this design is the simplicity of the mono-
mer subunit and the relatively easy functionalization of the
amine nitrogen. The resulting self-complementary deriva-
tives can self-associate through bidentate hydrogen bonding
interactions to form dimers (3 and 4) with very different
dispositions of the carboxylic acid and acylaminopyridine
The development of small organic components that can
form large molecular aggregates through non covalent inter-
actions has been an active area of research over the past
1
decade. Potential applications for these types of controlled
assemblies lie in such areas as non-linear optical materials,
2
liquid crystals and catalysis. A key to the development of
such systems is to control the structure of the ®nal supra-
molecular assembly. Another important requirement is the
possibility of functionalization on the self-assembling
3
template without disturbing the supramolecular structure.
We have shown that bis-aminopyridines can bind to bis-
carboxylic acids through a double bidentate hydrogen bond-
ing interaction to form 1:1 complexes (as in Fig. 1a). The
4
length of the spacer determines the nature of the complex
Figure 1. (a) Host±guest interaction between bis-carboxylic acid and bis-acylaminopyridine. (b) Potential aggregation of acylaminopyridine±carboxylic acid
derivatives.
Keywords: self-assembly; hydrogen bonding; acylaminopyridine±carboxylic acid.
*
Corresponding author. Tel.: 11-203-432-5570; fax: 11-203-432-3221; e-mail: andrew.hamilton@yale.edu
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00733-X

8
420
A. Zafar et al. / Tetrahedron 56 (2000) 8419±8427
Scheme 1.
functionalities. These in turn can potentially form a range of
linear or cyclic molecular aggregates. In this paper we
report the synthesis, X-ray crystal structures and aggre-
gation properties of a family of acylaminopyridine±
carboxylic acids and show that the nature of the aggregate
is determined by the relative orientation of the two inter-
acting groups.
Results and Discussion
Synthesis
The acylaminopyridine±carboxylic acid derivatives could
be prepared by two primary routes. The ®rst involved acyl-
ation of the aminopicoline followed by oxidation of the
Scheme 2.
Figure 2. (a) Space ®lling representation of the X-ray crystal structure of 1a showing the hydrogen bonding pattern within a layer. (b) Three dimensional
packing in 1a.

A. Zafar et al. / Tetrahedron 56 (2000) 8419±8427
8421
Figure 3. (a) Acetic acid dimer, (b) Intermolecular complex between 8 and 9.
6
3a. Fig. 2b shows the stacking of layers to form the lattice
methyl group to the corresponding carboxylic acid. For
example, acylaminopyridine 1a was synthesized by acyl-
ating 2-amino-6-picoline with acetic anhydride followed
by oxidation with potassium permanganante (Scheme 1).
Alternatively, direct acylation of the aminopicolinic acid
or ester could be used in certain cases. Acylaminopyridine
Ê
in 1a. The distance between the layers is 4.9±5.1 A.
Acid 2a was found to be sparingly soluble in most organic
solvents. However, we were able to grow X-ray quality
crystals from a solution of 2a in water at 388C. The X-ray
crystal structure shows a bidentate hydrogen bonding
pattern between the aminopyridine and the carboxylic acid
1b was synthesized in three steps via esteri®cation of 6-
aminopicolinic acid followed by acylation with decanoyl
chloride and basic hydrolysis (Scheme 1). Aminopyridines
Ê
groups that form linear tapes (NH´´´Oˆ2.03 A,
Ê
2a and 2b were synthesized in a single step by acylating 6-
OH´´´Nˆ1.75 A) (Fig. 4a). The structure is nearly planar
aminonicotinic acid with acetic anhydride and decanoyl
chloride respectively (Scheme 2).
and the out of plane bending of both the amide and the
carboxylic acid group is within 58. There is close packing
between the two adjacent antiparallel linear tapes with close
positioning of the amide carbonyl and the hydrogen at C-4
of the pyridine ring (CO´´´HC-4ˆ2.38±2.47 AÊ ). Fig. 4b
Solid state structures
shows the three dimensional packing in which the linear
tapes stack over each other. The interlayer distance ranges
Acid 1a was not soluble in chloroform. However, we were
able to grow X-ray quality crystals from a mixture of metha-
nol and water. The most striking feature of the crystal struc-
ture is the absence of the expected bidentate hydrogen
bonding pattern between the aminopyridine and the
carboxylic acid groups. The aminopyridines pack in a linear
fashion by forming hydrogen bonds between the amide car-
Ê
from 4.9±5.1 A.
Solution studies
Acid 1b. It is known that amides are self-complementary
through NH´´´OC interactions and this feature has been used
bonyl and the carboxylic acid OH (CvO´´´H±O, dist.
Ê
.82 A) as shown in Fig. 2a. The linear ribbons face each
7
to form hydrogen-bonding aggregates. The absence of any
1
other in an antiparallel arrangement and the individual units
are connected together by two water molecules. The water
dimer forms four hydrogen bonds to the two amino-
bidentate hydrogen bonding in the crystal structure of 1a
raised the need to establish the nature of self-association in
solution. NMR dilution experiments on 1b were performed
Ê
Ê
pyridine units (NH´´´OH , 2.06 A; CvO´´´HOH, 2.03 A).
2
in CDCl over a concentration range of 0.1 mM to 100 mM.
3
Another interesting feature is the CH´´´O interaction
Ê
The amide N±H shifted down®eld from 7.8 ppm at
0.16 mM to 10.36 ppm at 95.9 mM (Fig. 5). A correspond-
ing dilution experiment with ester 7, which is incapable of
bidentate hydrogen bonding, showed a small down®eld shift
of the amide N±H resonance indicating little self-associa-
tion in solution.
(
CH´´´O, 2.33 A) between the acetyl CH and the acid car-
3
bonyl groups. This type of bidentate hydrogen bonding inter-
action between acetyl and carboxylic acid groups has also
been found in the crystal structure of acetic acid (OH´´´OC,
Ê Ê
1.64 A, CH´´´OC, 2.4 A), as shown schematically in Fig.
Figure 4. X-ray crystal structure of 2a. (a) Hydrogen bonding layer arrangement. (b) The three dimensional packing in 2a.

8
422
A. Zafar et al. / Tetrahedron 56 (2000) 8419±8427
Figure 5. The amide N±H NMR shift in CDCl
3
against Log[Conc. mM] for 1b (W), 7 (V) and 8 and 9 (1:1, O).
The difference of more than 2 ppm in the down®eld shift of
the amide N±H groups of 1b and 7 at 100 mM clearly
indicates that the presence of carboxylic acid is essential
for the aggregation of 1b. The size of the down®eld shift
The molar ratio of the monomer, XˆM/M , is obtained using
0
n
n
0
the equation: M ˆXM 1nKX M , where M is the original
0
0
0
concentration of the monomer. The chemical shift is then
calculated by using: dcalcdˆd X1d (12X), where d is the
M
N
M
(
.2.5 ppm) and the sigmoidal shape of the curve for 1b
chemical shift of the monomer and the d is the chemical
N
suggest the formation of a hydrogen-bonded complex at
high concentration with dissociation occurring at or below
shift of the oligomer. Non-linear curve ®tting can then be
carried out for different values of n, which represents the
number of monomers involved in the most stable oligomeric
1
mM. The nature of the self-associated species can be
investigated by using the Saunders and Hyne analysis of
complex. Finally, for each curve ®t the chi square value is
i
P
1
3
2
2
molecular aggregation. This model assumes that at equi-
librium only two dominant species, the monomer (M) and
the most stable oligomeric species (N) are present in
solution:
calculated by using the expression: x ˆ ꢀdcalcd 2 dobs† :
Saunders and Hyne analysis of the dilution curve for 1b
provides the best ®t at aggregation number `n'ˆ3. The
curve ®t for the monomer±trimer model is shown in Fig.
6. Detailed results of the curve ®tting analysis for different
aggregation numbers are listed in Table 1.
K
nM O N
N
K ˆ Mn
1
Figure 6. H NMR dilution curve obtained for 1b in CDCl
3
, (W) experimental; (Ð) curve ®tting for nˆ3.
Table 1. Results of non-linear curve ®tting
2
23
Aggregation number `n'
Aggregation constant (K
n
)
Monomer shift d
m
(ppm)
Oligomer shift d
n
(ppm)
x £10
21
2
3
4
5
27 M
4474 M
7.75
7.88
7.92
7.94
11.95
10.95
10.70
10.59
101.9
2
2
7.46
37.3
82.6
5
23
7.28£10 M
8
24
1.2£10 M

A. Zafar et al. / Tetrahedron 56 (2000) 8419±8427
8423
1
Figure 9. The H chemical shifts vs. log[2b] for H2 (V); H4 (S), H5 (W)
and the amide-N±H (o) in CDCl .
3
bonds are formed in the latter but 2n in the former. Also, in
the cyclic aggregate a planar arrangement of the compo-
nents is possible whereas in the linear form steric interaction
of the alkyl substituents forces non-planarity.
Figure 7. ESMS of 1b in methanol.
Further support for the cyclic complex comes from a
comparison (Fig. 5) of the chemical shift dependence on
concentration between 1b (containing both hydrogen bond-
ing groups within the same molecule) and a 1:1 mixture of
benzoic acid 8 and acyl aminopyridine 9 (in which the inter-
acting groups are in separate molecules; Fig. 3b). In the
latter situation no cyclization is possible and the chemical
shift changes would be expected to mirror those of the linear
aggregate rather than the cyclic form. Fig. 5 shows differ-
ences in both the shape of the dilution curve and the size of
the chemical shift changes for 1b compared to 819. At
Electrospray mass spectrometry in positive ion mode was
performed using methanol as a solvent but gave no [M11]
peaks. However, in the negative ion mode there were three
distinct peaks corresponding to M21 (m/e 291.28), 2M21
1
(
m/e 583.31) and 3M21 (m/e 875.39) (Fig. 7). The absence
of any higher oligomers in the ESMS spectrum strongly
suggests that the trimeric and dimeric aggregates are formed
preferentially in the gas phase.
Nature of the trimeric aggregate
1
00 mM concentrations the Dd for the amide-NH of 1b is
The NMR experiments described above point to a bidentate
hydrogen bonding interaction for the association of 1b, as in
twice that of 819, suggesting the formation of a more stable
entity in the case of 1b. This increased stability is also
re¯ected in the sigmoidal shape of the dilution curve for
1b which contrasts with the steady change for 819. While
not conclusive, the different behavior of 1b compared to
819 indicates the formation of a cyclic aggregate in
solution.
3
. However, further oligomerization to the trimeric species
indicated by the curve ®tting and ESMS experiments can
occur to give either a cyclic or linear aggregate. Energy
minimized structures for cyclic and linear forms of the
trimer are shown in Fig. 8. In both cases undistorted biden-
tate hydrogen bonds are observed between the amino-
Ê
pyridine and the carboxylic acid groups (N±H´´´O, 1.7 A;
Acid 2b. A dilution experiment for 2b in CDCl over a
3
concentration range of 0.16 mM to 96 mM is shown in
Fig. 9. The amide N±H showed a small up®eld shift (Fig.
Ê
OH´´´N, 1.9±2.0 A). Cyclic aggregates should be favored
8
over the corresponding linear form, since 2n22 hydrogen
9
Figure 8. Energy minimized structures of two possible trimers of 1b. (a) Cyclic trimer. (b) Linear tape.

8
424
A. Zafar et al. / Tetrahedron 56 (2000) 8419±8427
in the more polar conditions of 2% DMSO-d /98% CDCl3
6
(
8
V/V). In this case the amide-N±H resonance shifted from
.6 ppm at 0.17 mM to 10.16 ppm at 100.9 mM (Fig. 12).
The shift of more than 1.5 ppm as well as the sigmoidal
shape of the curve indicated the formation of a well de®ned
hydrogen bonded aggregate at high concentration. Again,
up®eld shifts were seen for the pyridine-H resonances above
1
0 mM (Fig. 10). However, unlike the N±H shifts, the
aromatic-H shifts did not show saturation at high concen-
tration, indicating that they are due to a higher order
p-stacking aggregation and not as a result of the bidentate
hydrogen bonding.
The presence of both hydrogen bonding and p-stacking
aggregation with 2b renders Saunders±Hyne analysis of
the amide N±H dilution curve imprecise, as the single equi-
librium model no longer applies. The results of curve ®tting
to different aggregation numbers are summarized in Table 2.
The best ®t (Fig. 12) was obtained for the monomer±pen-
tamer model. Although, the quality of the ®t is poor due to
the multiple aggregation phenomena the results point to
aggregation properties of 2b that are very different from
those of 1b. In particular, a hydrogen bonded aggregate,
larger than the trimer found with 1b, appears to be favored
with 2b, where the only difference is the angular displace-
ment of the two complementary groups. This assertion is
supported by the ESMS spectrum of 2b (in methanol; nega-
tive ion mode) which showed six major peaks from m/e
(M21) to m/e (6M21).
1
Figure 10. The H chemical shifts vs log[2b] for H2 (W); H4 (V), and H5
S) in 2% DMSO-d /98% CDCl (V/V).
(
6
3
9) with increasing concentration in sharp contrast to the
large changes that occurred with the isomeric 2,6-disubsti-
tuted derivative 1b.
1
0
At 96 mM the chemical shift of the amide-N±H in 2b
matches exactly that of 1b. Therefore, it appears that the
amide N±H in 2b is strongly hydrogen bonded over the
entire concentration range and that the hydrogen bonded
aggregate is very stable in chloroform. There is also an
up®eld shift of the protons on the pyridine ring with increas-
ing concentration above 10 mM (Fig. 9). These aromatic
proton shifts in the absence of amide N±H shifts indicate
higher order association of the stable hydrogen bonded
aggregate presumably through p-stacking interactions (no
pyridine-H shifts are seen with 7 or 819). Saunders and
Hyne analysis of the pyridine-2H shift (as described
above) gave a best ®t for the monomer±trimer model and
Two possible forms of the larger aggregate with 2b are
shown in Fig. 13. Five molecules linked through 10 hydro-
gen bonds in a head-to-tail manner can form a cyclic aggre-
gate with a pseudo-planar arrangement of components
around a central cavity (Fig. 13a). A similar planarity can
be maintained in a linear aggregate (Fig. 13b) resulting in an
alternating projection of the alkyl substituents from the
central ribbon. The large shifts and sigmoidal shape of the
NMR dilution curve for 2b point to a signi®cant population
of the cyclic aggregate that can then undergo a multi-aggre-
gate p-stacking aggregation at higher concentration. In
contrast, eliminating the possibility of cyclization (as in 8
and 9) leads to weaker aggregates and little sigmoidal
behavior.
3
22
an association constant of 22.49£10 M (Fig. 11). This
type of p-stacking aggregation has previously been
observed by Moore and Zhang in the dimerization of
hexa-(phenylacetylene) macrocycles in chloroform and
also by us, in both the solid state and solution, with
1
1
1
2
5
-alkoxyisophthalic acid derivatives.
In order to reduce the stability of the hydrogen bonded
aggregate of 2b, we performed another dilution experiment
Figure 11. Curve ®tting for pyridine-2H in 2b, with nˆ3.

A. Zafar et al. / Tetrahedron 56 (2000) 8419±8427
8425
1
Figure 12. Experimental (W) and calculated (solid line: for nˆ5) H NMR chemical shift of amide-N±H vs log[2b] in 2% DMSO-d
6 3
/98% CDCl (V/V).
Table 2. The results of curve ®tting obtained for the amide N±H at various aggregation numbers (n)
2
23
Aggregation number (n)
Aggregation constant (K
n
)
Monomer shift d
m
(ppm)
Oligomer shift d
n
(ppm)
x £10
21
2
3
4
5
6
13 M
1376 M
8.52
8.61
8.64
8.65
8.66
11.89
10.88
10.65
10.55
10.49
210.08
86.55
58.96
55.76
59.44
2
2
5
23
24
25
1.37£10 M
7
1.40£10 M
9
1.48£10 M
8
Figure 13. Energy minimized structures of two key aggregation modes in 2b. (a) Cyclic pentamer. (b) Linear pentamer.
Conclusion
Siemens P4 or Rigaku AFC5R spectrometer using Siemens
SHELXTL PLUS (PC Version) or Siemens SHELXTL
IRIS. The re®nement was done by using direct methods
on full matrix least-squares ®t.
In summary, we have investigated the self-assembling
properties of a family of acylaminopyridine±carboxylic
acid derivatives. By changing the position of the two hydro-
gen bonding groups in the molecule we can dramatically
change the nature of the aggregate formed in solution.
-Acylaminopyridine-2-carboxylic acid shows a preference
for the formation of trimeric aggregates in solution while
-acylaminopyridine-3-carboxylic acid forms larger aggre-
NMR dilution studies were performed in deuterated solvents
at room temperature. The dilution data was analyzed by
1
3
6
using the Saunders and Hyne model. A computer program
was used to determine the best ®t for the dilution data at
different aggregation numbers. Chi-square values were used
as an indicator of the goodness of the ®t. The solvents were
dried using literature methods. All the chemicals were
purchased from Aldrich except 6-aminonicotinic acid
which was purchased from Lancaster Chemicals. Proton
6
1
gates. H NMR and ESMS experiments are used to support
the formation of well de®ned cyclic aggregates in solution.
1
3
Experimental
and C NMR spectra were recorded on a Bruker AM-300
300 MHz) instrument. NMR chemical shifts are reported in
ppm down®eld from tetramethylsilane (TMS). EI and FAB
(
X-ray crystallography was performed on a Siemens R3m/V,

8
426
A. Zafar et al. / Tetrahedron 56 (2000) 8419±8427
mass spectra (MS) were obtained using a Varian MAT CH-5
or VG 7070 mass spectrometer under the direction of Dr
Kasi V. Somayajula in the Department of Chemistry at the
University of Pittsburgh. Melting points were determined
using an Electrothermal capillary melting point apparatus.
Elemental analysis was carried out by Atlantic Microlab,
Inc. Norcross, GA.
suspended in a mixture of water (25 mL) and saturated
sodium bicarbonate (20 mL). The aqueous suspension was
then extracted with diethyl ether (6£15 mL). The combined
ether extract was dried over anhydrous magnesium sulfate.
The solution was concentrated by rotary evaporation and
after 12 h, the crystalline product was ®ltered and washed
with a minimum amount of cold (2208C) diethyl ether
(
1
0.50 g, 84%): mp 70±718C; H NMR (300 MHz, CDCl )
3
2
2
-Acetylamino-6-methylpyridine (5). To a solution of
-amino-6-picoline (5.0 g, 46.23 mmol) in dichloromethane
d 7.43 (1H, t, Jˆ7.6 Hz, Pyr. C±H at C-4), 7.36 (1H, d,
Jˆ6.8 Hz, Pyr. C±H at C-3), 6.59 (1H, d, Jˆ7.9 Hz, Pyr.
(
5
100 mL) at 08C was added acetic anhydride (4.9 mL,
0.86 mmol). The solution was stirred for 12 h at room
C±H at C-5), 5.20 (2H, s, NH ), 4.33 (2H, q, Jˆ7.1 Hz,
2
1
3
OCH ), 1.32 (3H, t, Jˆ7.1 Hz, ester CH ); C NMR
2
3
temperature and then the solvent was evaporated under
reduced pressure. The resulting light yellow solid was
dried under vacuum for 18 h. The dry product was dissolved
in potassium hydroxide solution (3.2 g, 57.03 mmol). The
solution was extracted with dichloromethane (2£40 mL)
and the combined organic extract was dried over anhydrous
magnesium sulfate. The solvent was evaporated under
reduced pressure and the residue was dried under vacuum
to obtain a light yellow crystalline solid as the desired
(75 MHz, CDCl ) d 166.3, 159.6, 147.1, 138.9, 116.0,
3
113.5, 62.3, 15.1; MS m/e calcd for C H N O : 166.0742,
8
10
2
2
found 166.0744.
6-Decanoylaminopyridine-2-carboxylic acid ethyl ester
(7). To a solution of 6 (0.76 g, 4.58 mmol) in dichloro-
methane (50 mL) was added decanoyl chloride (0.95 mL,
4.58 mmol). The reaction mixture was stirred at room
temperature for 18 h and then water (15 mL) and saturated
sodium bicarbonate were added (2 mL). The organic layer
was separated and the aqueous layer was extracted again
with dichloromethane (25 mL). The combined organic
extract was dried over anhydrous magnesium sulfate. The
solvent was then removed on a rotary evaporator and the
residue was dried under vacuum to give the crude product.
The crude material was puri®ed by silica gel chroma-
tography using a mixture of ethyl acetate and hexane (1:1)
1
product (6.3 g, 90.7%): mp 86±888C; H NMR (300 MHz,
CDCl ) d 8.25 (1H, br s, amide N±H), 7.97 (1H, d,
3
Jˆ8.1 Hz, Pyr. C±H at C-3), 7.57 (1H, t, Jˆ 7.8 Hz, Pyr.
C±H at C-4), 6.87 (1H, d, 7.5 Hz, Pyr. C±H at C-5), 2.42
1
3
(
(
3H, s, Pyr±CH ), 2.15 (3H, s, acetyl CH ); C NMR
3 3
75 MHz, CDCl ) d 169.5, 156.3, 150.8, 138.4, 118.9,
3
1
1
11.0, 24.2, 23.7; HRMS m/e calcd for C H N O:
8 10 2
50.0793, found 150.0789.
1
to afford the desired product (1.15 g, 78%): mp 42±448C; H
NMR (300 MHz, CDCl ) d 8.47±8.41 (1H, m, Pyr. C±H at
6-Acetylaminopyridine-2-carboxylic acid (1a). 2-Acetyl-
3
amino-6-picoline (2.86 g, 19 mmol) was dissolved in water
50 ml) and the solution was heated to 708C. Potassium
permanganate (6.32 g, 39 mmol) was added in small incre-
ments over a period of 30 min. The solution was re¯uxed for
C-3), 8.28 (1H, s, amide N±H), 7.85±7.80 (2H, m, Pyr.
C±H at C-4 and C-5), 4.46 (2H, q, Jˆ7.1 Hz, ester CH ),
(
2
2.38 (2H, t, Jˆ7.4 Hz, N±COCH ), 1.77±1.61 (2H, m,
2
N±COCH ±CH ), 1.42 (3H, t, Jˆ7.1 Hz, ester CH ),
2
2
3
5
min and ®ltered. The residue was washed several times
1.31±1.26 (12H, br m, alkyl chain), 0.86 (3H, t, Jˆ
6.2 Hz, alkyl chain CH ); C NMR (75 MHz, CDCl ) d
3 3
1
3
with boiling water. The washings and the ®ltrate were
combined and the volume was decreased to 50 ml on the
rotary evaporator. The solution was acidi®ed to pH ,1.5 by
using concentrated hydrochloric acid and then the pH was
adjusted to ,4.5 by adding saturated sodium bicarbonate
solution. After 2 h of refrigeration the precipitate was
172.4, 164.8, 151.7, 146.2, 139.5, 121.0, 117.8, 62.2, 37.9,
32.0, 29.4, 25.5, 22.8, 14.4, 14.3; MS m/e calcd for
C H N O : 320.2099, found 320.2100.
1
8
28
2
3
6-Decanoylaminopyridine-2-carboxylic acid (1b). To a
suspension of 7 (0.79 g, 2.48 mmol) in a mixture of water
and ethanol (31 mL, 1:4) was added lithium hydroxide
monohydrate (0.10 g, 2.48 mmol). The reaction mixture
was stirred at room temperature for 30 h. In order to
complete the reaction another portion of lithium hydroxide
monohydrate (0.04 g, 0.95 mmol) was added. The reaction
mixture was then re¯uxed for 1 h and then the solvent was
removed on a rotary evaporator. The residue was then
dissolved in water (20 mL) and ®ltered. The pH of aqueous
solution was then adjusted to 4.9 and the white precipitate
were ®ltered, washed thoroughly with water and air dried to
isolate the crude product. The crude material was puri®ed by
silica gel chromatography using 10% methanol in dichloro-
®
yellow powder as the desired product (0.88 g, 26%): mp
ltered, washed with water and air dried to obtain a light
1
15±2178C; H NMR (300 MHz, DMSO-d ) d 11.80 (1H,
2
6
br, acid OH), 10.79 (1H, s, amide N±H), 8.26 (1H, d,
Jˆ8.4 Hz, Pyr. C±H at C-3), 7.92 (1H, t, Jˆ8.1 Hz, Pyr.
C±H at C-4), 7.71 (1H, d, Jˆ7.5 Hz, Pyr. C±H at C-5), 2.09
1
3
(
3H, s, acetyl CH ); C NMR (75 MHz, DMSO-d ) d
3 6
169.8, 165.9, 152.0, 146.9, 139.3, 120.1, 116.8, 23.9;
HRMS m/e calcd for C H N O : 180.0534, found 180.0524.
8
8
2
3
6-Aminopyridine-2-carboxylic acid ethyl ester (6). A
100 mL round bottom ¯ask, ®tted with a re¯ux condenser
and a drying tube (drierite), was charged with 6-amino-
pyridine-2-carboxylic acid (0.50 g, 3.61 mmol) and an
excess of oxalyl chloride (10 mL). The reaction mixture
was stirred at room temperature for 12 h and then at 458C
for 1 h. The solvent was evaporated on a rotary evaporator.
The residue was dried under vacuum for about 1/2 h.
Absolute ethanol (70 mL) was then added slowly (caution).
The resulting suspension was re¯uxed for 1 h. The solvent
was removed on a rotary evaporator and the residue was
methane as the eluant to afford shiny thin needles as the
1
product (0.25 g, 35%): mp 102±1038C;
(300 MHz, THF-d ) d 10.90 (1H, s, acid OH), 9.94 (1H, s,
H NMR
8
amide N±H), 8.44 (1H, d, Jˆ7.8 Hz, Pyr. C±H at C-3), 7.81
(1H, t, Jˆ7.5 Hz, Pyr. C±H at C-4), 7.73 (1H, d, Jˆ7.5 Hz,
Pyr. C±H at C-5), 2.40 (2H, t, Jˆ6.9 Hz, N±COCH ), 1.72
2
(embedded under the solvent peak), N±CH CH ), 1.34±
2
2
1.28 (br m, 14H, alkyl chain), 0.88 (3H, t, Jˆ6.9 Hz, alkyl

A. Zafar et al. / Tetrahedron 56 (2000) 8419±8427
8427
1
CH ); C NMR (75 MHz, DMSO-d ) d 172.7, 166.1, 151.8,
3
References
3
6
139.2, 119.8, 116.4, 36.0, 31.3, 28.7, 24.9, 22.1, 13.8; MS
m/e calcd for C H N O : 292.1786, found 292.1776.
1
6
24
2
3
1. (a) Klok, H.-A.; Jolloffe, K. A.; Schauer, C. L.; Prins, L. J.;
Spatz, J. P.; M oÈ ller, M.; Timmerman, P.; Reinhoudt, D. N. J. Am.
Chem. Soc. 1999, 121, 7154±7155. (b) Kolotuchin, S. V.; Zimmer-
man, S. C. J. Am. Chem. Soc. 1998, 120, 9092±9093. (c) Conn,
M. M.; Rebek, Jr., J. Chem. Rev. 1997, 97, 1647±1668. (d) Marsh,
A.; Silvestri, M.; Lehn, J. M. Chem. Commun. 1996, 1527±1528.
6
-Acetylaminopyridine-3-carboxylic acid (2a). To a solu-
tion of 6-aminonicotinic acid (1.0 g, 7.24 mmol) in pyridine
30 mL) was added acetic anhydride (0.81 mL, 7.96 mmol)
(
and the reaction mixture was heated at 1308C for 24 h. The
reaction mixture was then cooled to room temperature and
the white precipitate was ®ltered and washed with pyridine
and excess water. The white solid was dried under vacuum
(
e) Fredericks, J. R.; Hamilton, A. D. In Comprehensive Supra-
molecular Chemistry, 1st ed.; Atwood, J. L., Davies, J. E. D.,
MacNicol, D. D., Vogtle, F., Eds.; Elsevier Science: New York,
to afford the desired product (1.2 g, 92%): mp 2818
(
1
996; Vol. 9, pp 565±593. (f) Lawrence, D. S.; Jiang, T.; Levett,
1
decomp.); H NMR (300 MHz, DMSO-d ) d 10.85 (1H,
6
M. Chem. Rev. 1995, 95, 2229±2260, (g) 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±54.
s, amide N±H), 8.79 (1H, d, Jˆ2.1 Hz, Pyr. C±H at C-2),
8.21±8.14 (2H, m, Pyr. C±H at C-4 and C-5), 2.11 (3H, s,
acetyl CH ); C NMR (75 MHz, DMSO-d ) d 169.8, 166.0,
1
3
3
6
2
3
4
. Rebek Jr., J. Chem. Soc. Rev. 1996, 225±264.
. Cram, D. J. Angew. Chem. Int. Ed. Engl. 1988, 27, 1009±1112.
. Garcia-Tellado, F.; Geib, S. J.; Goswami, S.; Hamilton, A. D.
155.0, 149.7, 139.4, 121.9, 112.3, 24.0; MS m/e calcd for
C H N O : 180.0534, found 180.0540.
8
8
2
3
J. Am. Chem. Soc. 1991, 113, 9265±9269.
6
6
-Decanoylaminopyridine-3-carboxylic
acid
(2b).
-Aminonicotinic acid (0.50 g, 3.61 mmol) was dissolved
5
1
6
7
. Yang, J.; Fan, E.; Geib, S. J.; Hamilton, A. D. J. Am. Chem. Soc.
993, 115, 5314±5315.
in pyridine at 708C under nitrogen. Decanoyl chloride
1.6 mL, 7.2 mmol) was added over a period of 45 min.
. Jones, R. E.; Templeton, D. H. Acta Crystallogr. 1958, 11, 484.
. Clark, T. D.; Buriak, J. M.; Kobayashi, K.; Isler, M. P.;
(
The solution was heated to 1108C for 6 h. Water (50 mL)
was added to the solution and it was acidi®ed to pH ,4 by
using hydrochloric acid (6 M). The precipitate was ®ltered
and washed thoroughly with water and air dried. The solid
was washed with hexane and dichloromethane and dried
under vacuum to obtain a white powder as the desired
McRee, D. E.; Ghadiri, M. R. J. Am. Chem. Soc. 1998, 120, 8949±
8
8
2
9
962.
. Zimmerman, S. C.; Duerr, B. F. J. Org. Chem. 1992, 57, 2215±
217.
. Molecular modeling was done using the Amber force®eld and a
1
continuum dielectric for chloroform within the MacroModel
program. Still, W.C., Columbia University.
product (0.59 g, 57%): mp 2058C (Decomp.); H NMR
(
(
(
300 MHz, DMSO-d ) d 12.00 (1H, br, acid OH), 10.79
6
1
0. We were unable to detect the amide N±H below 1 mM due to
1H, s, amide N±H), 8.79 (1H, s, Pyr. C±H at C-2), 8.21
2H, m, Pyr. C±H at C-4 and C-5), 2.39 (2H, t, Jˆ7.2 Hz,
broadening.
1
9
1
1. Zhang, J.; Moore, J. S. J. Am. Chem. Soc. 1992, 114, 9701±
702.
NCOCH ), 1.56 (2H, m, NCOCH ±CH ), 1.22 (12H, br,
2
2
2
alkyl chain CH ), 0.83 (3H, t, Jˆ6.9 Hz, alkyl chain CH );
2
3
1
3
2. Yang, J.; Marendez, J. L.; Geib, S. J.; Hamilton, A. D.
C NMR (75 MHz, DMSO-d ) d 172.8, 165.9, 155.1, 149.6,
6
1
1
42.9, 139.3, 121.6, 112.3, 36.1, 31.3, 28.7, 28.6, 24.8, 22.1,
3.9; MS m/e calcd for C H N O : 292, found 292.
Tetrahedron Lett. 1994, 3665±3668.
13. Saunders, M.; Hyne, J. B. J. Chem. Phys. 1958, 20, 1319±
1
6
24
2
3
1
323.
Acknowledgements
This work was supported by a grant from the National
Science Foundation.