Supramolecular polymer chemistry: Self-assembling dendrimers using the DDA·AAD (GC-like) hydrogen bonding motif

Article abstract of DOI:10.1021/ja0202006

Heterocyclic unit 2 containing complementary donor-donor-acceptor (DDA) and acceptor-acceptor-donor (AAD) hydrogen bonding arrays at an angle of about 60° was designed to self-assemble into a hexamer. To investigate whether this unit could self-assemble d

Full text of DOI:10.1021/ja0202006

Published on Web 10/26/2002
Supramolecular Polymer Chemistry: Self-Assembling
Dendrimers Using the DDA‚AAD (GC-like) Hydrogen Bonding
Motif
Yuguo Ma, Sergei V. Kolotuchin, and Steven C. Zimmerman*
Contribution from the Department of Chemistry, UniVersity of Illinois at Urbana-Champaign,
Urbana, Illinois 61801
Received February 11, 2002
Abstract: Heterocyclic unit 2 containing complementary donor-donor-acceptor (DDA) and acceptor-
acceptor-donor (AAD) hydrogen bonding arrays at an angle of about 60° was designed to self-assemble
into a hexamer. To investigate whether this unit could self-assemble dendrimers, the 2,8-diamino-2-N-
ethylpyrimido-(4,5-b)(1,8)naphthyridine-3H-4-one subunit was synthesized with a first (2a), second (2b),
and third generation (2c) Fre´chet-type dendron attached to the 8-amino group. The synthesis of 2a-c
was accomplished in 11 steps beginning with 2,6-diaminopyridine and the corresponding dendron bromide.
Studies using 1H NMR, size exclusion chromatography (SEC), and dynamic light scattering (DLS) support
the cooperative formation of cyclic hexamers in apolar solvents. The stability of the self-assembled
dendrimers is dependent on the size of the attached dendron, and mixing studies with 2a-c indicate their
usefulness in constructing dynamic combinatorial libraries.
Scheme 1
Introduction
Engineering specific noncovalent interactions between pre-
formed polymer molecules is emerging as a powerful strategy
for modulating the next higher level of organizational complex-
ity of macromolecules.1 Beyond simply increasing the cohesive
forces between polymers, such binding contacts may lead to
specific higher-level architectures through programmed self-
assembly. In turn these architectures may exhibit new properties
unattainable by conventional methods. Hydrogen bonding is
particularly useful in this regard because it is a strong and
directional force that has been used to generate a variety of novel
supramolecular structures through self-assembly.2 For example,
we recently reported a 34 kDa self-assembled dendrimer (1d)6
formed by the specific pairing of 24 carboxylic acid groups at
the core of 1 (Scheme 1).3 This first example of a discrete
aggregate of polymers had limitations, in particular, a low
stability in even moderately competitive solvents such as
* To whom correspondence should be addressed. E-mail: sczimmer@
uiuc.edu.
(1) For selected recent reviews see: (a) Supramolecular Polymers; Ciferri, A.,
Ed.; Marcel Dekker: NewYork, 2000. (b) Zimmerman, N.; Moore, J. S.;
Zimmerman, S. C. Chem. Ind. (London) 1998, 604-610. (c) Zeng, F.;
Zimmerman, S. C. Chem. ReV. 1997, 97, 1681-1712. (d) Brunsveld, L.;
Folmer, B. J. B.; Meijer, E. W.; Sijbesma, R. P. Chem. ReV. 2001, 101,
4071-4097. (e) Percec, V.; Cho, W. D.; Ungar, G.; Yeardley, D. J. P.
Angew. Chem., Intl. Ed. 2000, 39, 1598-1602.
(2) Selected reviews: (a) Lehn, J.-M. Angew. Chem., Int. Ed. Engl. 1990, 29,
1304-1319. (b) Fredericks, J.; Yang, J.; Geib, S. J.; Hamilton, A. D. Proc.
Indian Acad. Sci., Chem. Sci. 1994, 106, 923-935. (c) 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. (d) Lawrence, D. S.;
Jiang, T.; Levett, M. Chem. ReV. 1995, 95, 2229-2260. (e) Conn, M. M.;
Rebek, J., Jr. Chem. ReV. 1997, 97, 1647-1668. (f) Zimmerman, S. C.;
Corbin, P. S. Struct. Bonding 2000, 96, 63-94. (g) Prins, L. J.; Reinhoudt,
D. N.; Timmerman, P. Angew. Chem., Int. Ed. 2001, 40, 2382-2426. (h)
Archer, E. A.; Gong, H.; Krische, M. J. Tetrahedron 2001, 57, 1139-
1159.
tetrahydrofuran (THF). Furthermore, the information content in
the isophthalic acid units was ambiguous with 1b-d forming
discrete hexamers, whereas 1a formed oligo- and polymeric
aggregates.4
Our early investigations5 of the stability of DNA base pairs
and their heterocyclic analogues focused attention on the DDA‚
(3) (a) Zimmerman, S. C.; Zeng, F.; Reichert, D. E. C.; Kolotuchin, S. V.
Science 1996, 271, 1095-1098. (b) Thiyagarajan, P.; Zeng, F.; Ku, C. Y.;
Zimmerman, S. C. J. Mater. Chem. 1997, 7, 1221-1226. (c) Zeng, F.;
Zimmerman, S. C.; Kolotuchin, S. V.; Reichert, D. E. C.; Ma, Y.
Tetrahedron 2002, 58, 825-843.
9
10.1021/ja0202006 CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 13757-13769
13757

A R T I C L E S
Ma et al.
AAD hydrogen-bonding motif found in the guanine-cytosine
(G‚C) base pair. This motif can form in only one way and
exhibits an association constant (Kassoc) in chloroform6 that is
at least 102-fold higher than that of the benzoic acid dimer and
the commonly used DAD‚ADA motif.5 Thus, we designed,
synthesized, and studied 2a7 whose DDA and AAD hydrogen-
bonding arrays are fixed at a 60° angle and serve, therefore, as
an information code for cyclic assembly of a hexamer (Figure
1).8,9 The strength of the DDA‚AAD motif allowed hexamer
(2a)6 to form even in 15% aqueous THF.7 Herein, we provide
a detailed account of the synthesis and assembly studies of 2a
as well as its higher generation analogues 2b and 2c. Whereas
experimental problems associated with the isophthalic acid based
system prevented determination of the relative stability of the
hexamers 1a-d, studies with their heterocyclic counterparts
2a-c allowed a relationship between dendrimer size and
aggregate stability to be established. The results of the studies
involving pairwise mixing of 2a-2c indicate that 2 is a useful
building block for creating supramolecular, dynamic combina-
torial libraries.10
Results and Discussion
Synthesis. The synthesis of compounds 2a-c followed the
established synthetic route,5b,7 with the minor modifications
outlined below (Scheme 2). Thus, lithiation of 3 with BuLi
followed by dimethylformamide (DMF) quench, aqueous work-
up, and deprotection of the crude aldehyde 4 with lithium
hydroxide (LiOH) gave 5 in a 62% yield without the need for
column chromatography. Friedla¨nder condensation of 5 with
diethyl malonate afforded 7-amino-3-carbethoxy-1,8-naphthy-
ridin-2(1H)-one (6) in a 92% yield.5b In the previously reported7
conversion of 6 to BOC-protected 10, the butanoyl protecting
group in 7a was found to give superior yields to the acetyl group.
Increasing the protecting group chain length even more gives
further improvements in yield. Thus, treatment of 6 with valeric
Figure 1. Self-assembling monomer 2 and the corresponding aggregate
(2)6: D, hydrogen bond donor; A, acceptor; R, Gn, n ) 1-3, a-c in Scheme
1.
anhydride in the presence of pyridine gave pentanoyl derivative
7b in 93% yield. Naphthyridinones 7a and 7b react with POCl3
to give gram quantities of chloronaphthyridines 8a and 8b.
Protection of the amino group of 8a and 8b with (BOC)2O
occurred in 75 and 97% yield, respectively, and chemoselective
cleavage of the amide groups gave naphthyridine 10 in 82 and
93% yield, respectively (Scheme 2).
The required dendron bromides 11a-c were synthesized3 by
a convergent approach similar to that used by Fre´chet.11 Starting
from methyl 3,5-dihydroxybenzoate and 3,5-di-tert-butylbenzyl
bromide, dendron bromide 11a-c were prepared in three, five,
and seven total steps with overall yields of 47, 30, and 7%,
respectively. N-Alkylation of 10 with the appropriate dendron
bromide in the presence of potassium carbonate and 18-crown-6
produced compounds 12a (80%), 12b (47%), and 12c (70%).
The change in the 7-amino protecting group from pentanoyl to
BOC was necessitated by the observation that alkylation of 8
was unselective, producing a mixture of the desired N-alkylated
and undesired O-alkylated isomers. Beyond reducing the yield,
the O-alkylation byproduct proved difficult to remove chro-
matographically. Changing the protecting group to BOC resulted
in clean and regioselective N-alkylation of 10. Chemoselective
deprotection of 12 with trifluoroacetic acid (TFA) afforded
amines 13a (86%), 13b (45%), and 13c (64%) without affecting
the ether groups of the dendrons. Finally, substitution and
cyclization12 of 13a-c with N-ethyl guanidine, which was
generated in situ from an excess of its sulfate salt and sodium
hydride, afforded compounds 2a (64%), 2b (30%), and 2c
(64%), respectively (Scheme 2).
(4) Other examples of main-chain supramolecular polymers: Castellano, R.
K.; Rudkevich, D. M.; Rebek, J., Jr. Proc. Natl. Acad. Sci. U.S.A. 1997,
94, 7132-7137. Sijbesma, R. P.; Beijer, F. H.; Brunsveld, L.; Folmer, B.
J. B.; Hirschberg, J. H. K. K.; Lange, R. F. M.; Lowe, J. K. L.; Meijer, E.
W. Science 1997, 278, 1601-1604.
(5) (a) Murray, T. J.; Zimmerman, S. C. J. Am. Chem. Soc. 1992, 114, 4010-
4011. (b) Fenlon, E. E.; Murray, T. J.; Baloga, M. H.; Zimmerman, S. C.
J. Org. Chem. 1993, 58, 6625-6628. (c) Zimmerman, S. C.; Murray, T. J.
Philos. Trans. R. Soc. London, Ser. A 1993, 345, 49-56. (d) Murray, T.
J.; Zimmerman, S. C. Tetrahedron Lett. 1995, 36, 7627-7630.
(6) Jorgensen, W. L.; Pranata, J. J. Am. Chem. Soc. 1990, 112, 2008-2010.
Pranata, J.; Wierschke, S. G.; Jorgensen, W. L. J. Am. Chem. Soc. 1991,
113, 2810-2819. Kra´l, V.; Sessler, J. L. Tetrahedron 1995, 51, 539-554.
Sessler, J. L.; Wang, R. J. Org. Chem. 1998, 63, 4079-4091.
(7) Kolotuchin, S. V.; Zimmerman, S. C. J. Am. Chem. Soc. 1998, 120, 9092-
9093.
(8) For selected examples of other heterocycles that form cyclic assemblies.
(a) Trimers: Zimmerman, S. C.; Duerr, B. F. J. Org. Chem. 1992, 57,
2215-2217. Boucher, E.; Simard, M.; Wuest, J. D. J. Org. Chem. 1995,
60, 1408-1412. (b) 2 + 2 Tetramer: Yang, J.; Fan, E.; Geib, S. J.;
Hamilton, A. D. J. Am. Chem. Soc. 1993, 115, 5314-5315. (c) Isoguanine
tetramers: Tirumala, S.; Davis, J. T. J. Am. Chem. Soc. 1997, 119, 2769-
2776. (d) Isoguanine pentamer-cesium complex: Shi, X.; Fettinger, J. C.;
Cai, M.; Davis, J. T. Angew. Chem., Int. Ed. 2000, 39, 3124-3127. (e) 3
+ 3 Hexamer: Zerkowski, J. A.; Seto, C. T.; Whitesides, G. M. J. Am.
Chem. Soc. 1992, 114, 5473-5475. Vreekamp, R. H.; van Duynhoven, J.
P. M.; Hubert, M.; Verboom, W.; Reinhoudt, D. N. Angew. Chem., Int.
Ed. Engl. 1996, 35, 1215-1218. (f) Higher order aggregates based on
cyanuric acid-melamine system: refs 2c,g.
(9) Lehn, Mascal, and Fenniri have independently developed a bicyclic
heterocycle which forms a hexamer via the DDA‚AAD hydrogen-bonding
motif: Marsh, A.; Silvestri, M.; Lehn, J.-M. Chem. Commun. 1996, 1527-
1528. Mascal, M.; Hext, N. M.; Warmuth, R.; Moore, M. H.; Turkenburg,
J. P. Angew. Chem., Int. Ed. Engl. 1996, 35, 2204-2206. Fenniri, H.;
Mathivanan, P.; Vidale, K. L.; Sherman, D. M.; Hallenga, K.; Wood, K.
V.; Stowell, J. G. J. Am Chem. Soc. 2001, 123, 3854-3855.
(11) Hawker, C. J.; Fre´chet, J. M. J. J. Am. Chem. Soc. 1990, 112, 7638-7647.
Wooley, K. L.; Hawker, C. J.; Fre´chet, J. M. J. J. Am. Chem. Soc. 1993,
115, 11496-11505.
(10) Lehn, J.-M. Chem. Eur. J. 1999, 5, 2455-2463. Albrecht, M. J. Inclusion
Phenom. Macrocyclic Chem. 2000, 36, 127-151.
(12) Taylor, E. C.; Wong, G. S. K. J. Org. Chem. 1989, 54, 3618-3624.
9
13758 J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002

Supramolecular Polymer Chemistry
A R T I C L E S
Scheme 2
Table 1. 1H NMR Chemical Shifts (ppm) of NH Groups in 2a-c
and Model Compounds 14-18a
Self-Assembly Studies. Establishing and characterizing solu-
tion assembly remains a challenging endeavor. Generally, no
single method is sufficient to determine the solution structure
of an aggregate, but a strong circumstantial case can be built
by collecting data from several techniques. In the case of 2, 1H
NMR, size exclusion chromatography (SEC), and dynamic light
scattering (DLS) were used in conjunction with covalently
synthesized size standards and analogues of 2 that serve as
chemical shift standards.
1H NMR Analysis. The 1H NMR chemical shifts of the NH
groups of 2a-c (see Figure 2 for 2b) recorded in different
solvents are shown in Table 1. In each case the NH signals are
sharp, downfield, and, in the case of 2a, essentially insensitive
to concentration in apolar solvents. All three compounds exhibit
a small downfield shift in aromatic solvents, and there is a
noticeable change in the NH shifts in moving to the third
generation dendrimer (i.e., 2c). The aromatic solvent effect is
most likely a general solvent effect although stronger hydrogen
bonding cannot be ruled out. It is notable that the spectra for
2a and 2b exhibit a relatively small dependence on solvent and
dendrimer size and are nearly superimposable. Most importantly,
the three NH signals are far downfield of analogous non-
hydrogen-bonded NH protons such as those in 14 and 15, and
they appear in the same region as the corresponding NH groups
in complexes 16‚17 and 17‚18 (Figure 3 and Table 1). These
data indicate that the NH groups of 2 are engaged in strong
hydrogen-bonding contacts.
proton obsd
compd/complex solvent
NH-2
NH-3
NH-8
2a
CDCl3
THF-d8
toluene-d8
dioxane-d8
10% DMSO-d6/CDCl3 10.66
8% H2O/THF-d8
CDCl3
THF-d8
toluene-d8
CDCl3
THF-d8
10.89
11.00
11.18
10.94
13.96
14.04
14.28
13.94
13.89
14.00
14.39
14.10
14.31
14.34
14.39
14.68
10.33
10.48
10.87
10.38
10.26
10.40
10.06
10.51
10.84
10.11
10.29
10.72
11.00
10.93
11.00
11.21
10.91
11.15
11.55
5.04
2b
2c
toluene-d8
CDCl3
14
15
16‚17a
17a‚18a
CDCl3
CDCl3
CDCl3
4.80
13.90b (17a)
13.27b (18a)
13.3 (18b)
17b‚18b CDCl3
11.1 (17b)
a Not all NH protons were seen in complexes. Compound 14 was
prepared from 7-amino-2,4-dimethyl-1,8-naphthyridine and 11a. Compounds
15-18 were available from a previous study (see ref 3). b Value represents
∆δmax obtained by curve fitting 1:1 binding isotherms.
Two-dimensional nuclear Overhauser enhancement spectros-
copy (NOESY) was employed to probe structural details of
aggregate (2b)6 in solution. In the NOESY spectra of 2b in
CDCl3 and toluene-d8 (Figure 4), the cross-peaks from NH-8
to both NH-2 and NH-3 indicate their spatial proximity
consistent with pairing of the DDA and AAD binding sites. In
the structure of the monomer, the NH-8 to NH-3 distance based
on molecular modeling (Macromodel) is >8.0 Å, which is much
larger than the required distance to see a NOE correlation
between the two protons.
Size Exclusion Chromatography Studies. Size exclusion
chromatography with appropriate size comparisons is now
established as a powerful method of determining the aggregation
state of supramolecular assemblies.2c,3,8b,13 Thus, whereas the
1H NMR data indicated the formation of hydrogen bonds, SEC
of 2a-c provided molecular weight (MW) and size data
Figure 2. 1H NMR signals of NH groups of 2b in CDCl3.
9
J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002 13759

A R T I C L E S
Ma et al.
Figure 3. Structure of compounds 14-18, analogues of 2.
Figure 5. (A) SEC traces of 2a (3.83 mM), 2b (1.02 mM), and 2c (4.47
mM) in toluene and (B) change in SEC derived MW with injected
concentration of 2a (2), 2b ([), and 2c (9) in toluene.
Table 2. Molecular Weight Determination by SEC
compd calcd MW
MALDI
SEC (toluene) SEC (THF) calcd MW of hexamer
2a
2b
2c
782.49
1430.9
2727.7
784.4
1430.4
2730.3
5660a
8800c
4800b
8870d
2650f
4698
8592
16378
11500e
a 0.0036-3.6 mM. b 4.4 mM. c 0.007-7.0 mM. d 1.0-5.0 mM. e 0.7-
4.4 mM. f 0.01-5.0 mM.
The nature of the aggregates formed by 2a-c can be inferred
from comparison of the SEC retention times to appropriate size
standards. Thus, molecular weights were determined in both
toluene and THF by SEC using PS standards for calibration
(Table 2). The MWPS values for 2a and 2b are ca. 25 and 2%
larger than that calculated for the corresponding hexamers,
whereas the MWPS for 2c is ca. 30% smaller than that calculated
for the hexamer. The deviations from the theoretical MW
calculated for hexamer can be explained by a low-density but
rigid heterocyclic core holding six densely packed dendrons
around a ca. 22.7 Å diameter disk. In 2a the core dominates,
whereas in 2c the dendrons contain most of the mass. It is well-
established that dendritic structures, being denser and more
compact than linear polymers, give PS determined molecular
weight values that underestimate their actual molecular weights,
especially in the case of higher-generation dendrimers.11,16
Figure 4. 1H NMR NOESY of 2b in toluene-d8 (8.7 mM).
indicating their aggregation state in solution. As seen in Figure
5A, 2a-c each gave a sharp and symmetrical peak in toluene
with a polydispersity index (PDI) MW/Mn ) 1.02 that is in the
range typically found for monodisperse, Fre´chet-type dendrim-
ers.11,14 Concentration-independent SEC retention can be taken
as prima fascia evidence in support of a discrete, closed (i.e.,
cyclic) aggregate.15 As was reported for 2a,7 the polystyrene
(PS) based molecular weights (MWPS) of 2b and 2c in toluene
changed minimally across a 103-fold dilution in SEC injection
concentration (Figure 5B). These results indicate that 2b and
2c also form discrete, cyclic aggregates in toluene.
(13) Michelsen, U.; Hunter, C. A. Angew. Chem., Int. Ed. 2000, 39, 764-767.
Haycock, R. A.; Hunter, C. A.; James, D. A.; Michelsen, U.; Sutton, L. R.
Org. Lett. 2000, 2, 2435-2438.
(14) Tyler, T. L.; Hanson, J. E. Chem. Mater. 1999, 11, 3452-3459.
(15) Corbin, P. S.; Lawless, L. J.; Li, Z.; Ma, Y.; Witmer, M. J.; Zimmerman,
S. C. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 5099-5104.
Compound 19 was synthesized as a semirigid covalent size
standard for proposed hexamer (2a)6. As seen in Figure 6 the
9
13760 J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002

Supramolecular Polymer Chemistry
A R T I C L E S
Figure 6. Molecular modeling of (2a)6 (A) and 19 (B).
sizes of (2a)6 and 19 obtained by molecular modeling are very
similar. Indeed, the two compounds have identical SEC retention
computationally (see Experimental Section for details). Both
highly extended and folded-back conformations were examined
to provide an indication of the uncertainty in the calculated Rg
values. As seen in Figure 7, a reasonably linear correlation is
seen between the calculated Rg values and the experimentally
determined SEC retention times.
times in both toluene and THF and coelute when co-injected.
For compounds 2b and 2c comparisons were made to known
hexamers 1b and 1c and to recently described hexamers (20b)6
and (20c)6.15 Thus, the elution times followed the order 1c >
20c > 20b > 2c > 2b > 2a = 19, paralleling their modeled
sizes, which, in turn, is consistent with ordering of the core
diameters as determined by molecular modeling: (1)6 (41.1 Å)
> (20)6 (30.6 Å) > (2)6 (22.7 Å).
The SEC retention times are determined by the hydrodynamic
volume of the aggregate or compound examined.17 Thus, to put
the comparison of retention times on a more quantitative footing,
the radius of gyration (Rg) of size standard 19 and self-assembled
dendrimers (1c)6, (20b,c)6, and (2a-c)6 were determined
Dynamic Light Scattering. The aggregates formed by 1c
and 2a-c were also studied by dynamic light scattering. In some
runs a trace amount (ca. 0.1 wt %) of a very large particle was
observed. Nonetheless, and consistent with the SEC results, all
compounds gave a narrow polydispersity value (Table 3),
indicating formation of a discrete aggregate in toluene solution.
Most importantly, the relative magnitude of the hydrodynamic
radii (Rh) measured by DLS correlated well with the observed
by SEC retention times and the aggregate sizes determined by
molecular modeling. Thus, the Rh values decreased in the order
(1c)6 > (2c)6 > (2b)6 > (2a)6 = 19 and were found to be within
(16) Tomalia, D. A.; Naylor, A. M.; Goddard, W. A., III. Angew. Chem., Int.
Ed. Engl. 1990, 29, 138. Aharoni, S. M.; Crosby, C. R.; Walsh, E. K.
Macromolecules 1982, 15, 1093.
(17) Grubisic, Z.; Rempp, P.; Benoit, H. Polym. Lett. 1967, 5, 753-759.
9
J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002 13761

A R T I C L E S
Ma et al.
Table 3. Hydrodynamic Radius (Rh) Determined by Dynamic Light Scattering
19
(2a)6
(2b)6
(2c)6
(1c)6
calcd Rha (nm)
DLS Rhb (nm)
polydispersity (%)
MW (starburst)
calcd MW
1.78 ( 0.14
1.79
12
4670
4098
1.68 ( 0.08
1.90
13
6000
4698
1.88 ( 0.15
2.20
15
9000
8592
2.34 ( 0.12
2.30
14
11000
16378
3.24 ( 0.24
3.40
15
40000
19200
a Determined by molecular modeling (see text and Experimental Section). Calcd Rh ) Rg (5/3)1/2 assuming a spherical, compact structure. b Average Rh
from DLS on sample solutions of 5.0 mg/mL at 25 °C.
Figure 7. Correlation between SEC retention time and modeled size of
Figure 8. Upfield spectral regions from a representative 1H NMR dilution
covalent size standard 19 and self-assembled dendrimers (1c)6, (20b,c)6,
and (2a-c)6.
study of the (2b)6 in (A) CDCl3 at (a) 0.588 mM, (b) 98.0 µM, (c) 16.3
µM, and (d) 2.72 µM and (B) CDCl3/DMSO-d6 ) 1:1 (v/v) at 1.77 mM.
15% of the calculated Rh values18 obtained from molecular
modeling (Table 3). Although molecular weights could be
derived from the DLS data by applying a starburst polymer
model, these values deviated by approximately 50% in the case
peak ratio of hexamer to monomer also decreased. The same
result is observed by increasing the temperature in a variable
1
temperature (VT) H NMR study in CDCl3. Likewise, upon
addition of a polar solvent (e.g., DMSO-d6) to a solution of 2b
in CDCl3, the hexamer peak decreases in intensity, being
of 1c and 2c.
Both the SEC and DLS experiments are consistent with the
replaced by the monomer. As seen in Figure 8B, in a 1:1 (v/v)
formation of discrete, hexameric aggregates from 2, but the data
CDCl3-DMSO-d6 mixture a single sharp peak was observed
for the monomer tert-butyl groups.
do not exclude the formation of pentamers or heptamers.
Nonetheless, such structures cannot be made with conventional
CPK models, and computational studies indicate that such
aggregates, if formed, would have highly distorted hydrogen
The slow exchange between monomer and hexameric ag-
1
gregate observed in H NMR spectra of 2b allowed the effect
of the solvent on the aggregation process to be studied more
systematically. Thus, for 2b, the concentrations which afford a
1:1 ratio of the monomer and hexamer tert-butyl peaks was
bonds.
Monomer-Hexamer Equilibrium. The SEC of compound
2c in THF shows a single peak corresponding to monomer. At
∼10-3 M in THF-d8, ∼10-4 M in CDCl3 and ∼10-6 M in
most concentrations of 2a the SEC in THF shows a single sharp
benzene-d6. These results and those described above indicate
peak corresponding to hexamer, but at very high dilution some
that the relative stability of the hexamer increases with decreas-
monomer is observed. For 2b the SEC exhibits intermediate
ing solvent polarity: benzene-d6 > CDCl3 > THF-d8 > DMSO-
d6.
behavior with monomer and hexamer present at most concentra-
tions; with the former increasing at the expense of the latter
1H NMR dilution and VT studies were also carried out on
upon dilution. These data suggest that (1) the larger dendrons
2a and 2c with similar results. For example, the tert-butyl region
destabilize the hexamer and (2) the hexamer forms cooperatively
1
of the H NMR spectra of 2a shows a single tert-butyl singlet
with monomer and hexamer in slow exchange on the time scale
(hexamer) at a concentration greater than 0.1 mM in CDCl3
of the SEC experiment (minutes).
Similar observations were made in the H NMR of 2b in
but the presence of monomer at ca. 15 µM and below (Figure
9). All of these observations are consistent with a cooperative,
cyclic aggregation process driven by hydrogen bonding. Thus,
decreasing the concentration, increasing the temperature, or
1
chloroform-d (CDCl3), THF-d8, and benzene-d6, with two sets
of proton signals assigned to monomer and hexamer in slow
exchange. For example, as shown in Figure 8A, the singlets at
adding a competitive solvent destabilizes the hexamer, shifting
δ 1.28 and δ 1.33 ppm in CDCl3 are assigned, respectively, to
the equilibrium toward monomer in solution.
Dependence of Hexamer Stability on Dendron Size. The
the tert-butyl groups of the hexamer and monomer of 2b. As
the concentration decreased from 1.0 mM to 1.0 µM, the relative
well-separated tert-butyl signals seen for the monomer and
hexamer in the 1H NMR were used to provide semiquantitative
data on the relative stability of the hexamers formed from 2a-
(18) Trollsås, M.; Atthof, B.; Wu¨rsch, A.; Hedrick, J. L.; Pople, J. A.; Gast, A.
P. Macromolecules 2000, 33, 6423-6438.
9
13762 J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002

Supramolecular Polymer Chemistry
A R T I C L E S
a secondary hydrogen bonding interaction5,6 that is not present
in the 17b‚18b complex. In a related system the additional
secondary hydrogen bond was found to contribute about 0.7
kcal mol-1 20
On the other hand, the N-H donor in 17b is an
.
amide group, whereas the corresponding N-H donor in 2 is an
amine suggesting stronger primary hydrogen bonding in the
complex. In short, no definitive conclusion can be drawn
regarding the relative strength of the hydrogen bonding in the
two systems, but the hexamer does contain six additional
secondary hydrogen bonds.
The second factor that is certainly operating is a cooperativity
originating in the cyclic assembly process. This same higher
stability for a cyclic aggregate relative to an analogous complex
was most clearly demonstrated for a trimerization process.8a
Although the cooperativity manifests itself in the thermodynam-
ics of assembly, its origin can best be thought of in terms of
the kinetics of hexamer dissociation. Thus, upon breaking one
set of DDA‚AAD hydrogen bonds in (2)6 the barrier to
re-forming the hexamer is very low in comparison to the barrier
to dissociating a molecule of 2 giving the pentamer.
Figure 9. Upfield spectral region from a representative 1H NMR dilution
study of the (2a)6 in CDCl3 at (a) 6.90 mM, (b) 0.144 mM, and (c) 15.4
µM.
Dynamic Combinatorial Library from 2a + 2c. The rapid
and economical generation of large libraries of compounds
continues to be a significant challenge to organic chemists. One
approach that has attracted considerable attention recently uses
a dynamic combinatorial library in which constituent members
may interconvert.10 Thus, an external stimulus may shift the
equilibrating library toward one or a subset of members. Often
the interconversion occurs through the breaking and formation
of covalent bonds, but the library may also contain supramo-
lecular complexes or assemblies that equilibrate through non-
covalent exchange processes.21
As a logical extension of the research described above, the
possibility of generating a dynamic library from 2 was inves-
tigated by mixing first- and third-generation dendritic monomers
2a and 2c. In toluene, 2a and 2c exhibit discrete, symmetrical
SEC peaks with retention times of 15.1 and 14.2 min,
respectively, but only one sharp peak with a retention time of
14.3 min was observed after equimolar amounts of 2a and 2c
were mixed in solution for several hours before injection. The
measured PDI (1.02) was comparable to the PDI of each
individual hexamer. The kinetics of mixing, which were
monitored by SEC, clearly show that the subunits in each
aggregate (2a)6 and (2c)6 combine to generate new aggregates
but these ultimately convert to a discrete structure (Figure 13).
This recombination process was fully complete in about 16 min.
This rapid mixing may seem inconsistent with the slow
interconversion of monomer and hexamer, wherein monomers
2a and 2b do not exchange significantly with their respective
hexamers (2a)6 and (2b)6 on the time scale of the SEC runs
(ca. 16 min). However, the solvent in the two experiments is
different, and the mechanism for exchange need not proceed
through monomer. For example, the mixing of (2a)6 and (2c)6
might occur by an end-unit exchange between linear aggregates
as opposed to a full dissociation-recombination.
c. Thus, CDCl3 solutions of 2a-c were titrated with DMSO-d6
until the hexamer signal was no longer visible. Figure 10 shows
representative data for hexamer (2a)6 whose tert-butyl and NH
group signal intensity regularly decreases upon addition of
DMSO-d6 and then fully disappears at 35% (v/v) DMSO-d6-
CDCl3. The same experiments were carried out for 2b and 2c,
but in this case the amount of DMSO-d6 required to fully
dissociate the hexamer in CDCl3 was 19 and 9% (v/v),
respectively. These results, which are fully consistent with the
SEC dilution studies (vide supra), indicate that the stability of
the hexameric aggregates decreases with increasing dendrimer
size: (2a)6 > (2b)6 > (2c)6.
Cooperativity in the Cyclic Assembly Process. The obser-
vation that monomer and hexamer are in equilibrium, absent
any obvious observation of the intermediate n-mers (n ) 2-5),
suggested strongly that a type of cooperativity is present in the
cyclic assembly. To probe this aspect in more detail, the stability
of hexamer (2a)6 was compared to that of complex 17b‚18b,
which utilizes the same DDA‚AAD hydrogen bonding motif.19
Because of the high Kassoc value for the 17‚18 complex in CDCl3
(Kassoc > 104 M-1),5 1H NMR dilution studies were carried out
in 5% (v/v) DMSO-d6-CDCl3. At a concentration of 0.08 mM
17b‚18b was mostly dissociated, while NH proton signals of
(2a)6 showed a minimal upfield shift. Full dilution curves are
shown in Figure 11, where it is seen that the half-dissociation
point for 2a ([(2a)6]1/2
) 0.125 mM) comes at lower
dissoc
concentration than it does for 17b‚18b ([17b‚18b]1/2
)
dissoc
0.725 mM).
The cohesive force that stabilizes both species is the DDA‚
AAD hydrogen-bonding motif. The obvious, yet striking,
conclusion that can be drawn from these data is that (2a)6 is
more stable than complex 17b‚18b, even though the former,
being comprised of six molecules, requires a considerably higher
loss in entropy. The greater stability of (2a)6 may be attributable
to two factors. First, there are subtle differences in the hydrogen
bonding contacts used. As seen in Figure 12, N-1 in 2 provides
(20) Zimmerman, S. C.; Murray, T. J. Tetrahedron Lett. 1994, 35, 4077-4080.
(21) Huc, I.; Lehn, J.-M. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 2106-2110.
Calama, M. C.; Hulst, R.; Fokkens, R.; Nibbering, N. M. M.; Timmerman,
P.; Reinhoudt, D. N. Chem. Commun. 1998, 1021-1022. Hof, F.; Nuckolls,
C.; Rebek, J., Jr. J. Am. Chem. Soc. 2000, 122, 4251-4252. Furlan, R. L.
E.; Cousins, G. R. L.; Sanders, J. K. M. Chem. Commun. 2000, 1761-
1762. Cousins, G. R. L.; Furlan, R. L. E.; Ng, Y.-F.; Redman, J. E.; Sanders,
J. K. M. Angew. Chem., Int. Ed. 2001, 40, 423-428.
(19) For a more quantitative treatment of cyclic vs linear aggregation see:
Bielejewska, A. G.; Marjo, C. E.; Prins, L. J.; Timmerman, P.; de Jong,
F.; Reinhoudt, D. N. J. Am. Chem. Soc. 2001, 123, 7518-7533.
9
J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002 13763

A R T I C L E S
Ma et al.
Figure 10. 1H NMR titration of (2a)6 in CDCl3 by DMSO-d6.
Figure 12. Comparison of hydrogen bonding contacts in complex 17‚18
and (2)6.
lower stability of (2c)6, it is reasonable to hypothesize that the
destabilizing steric interactions between the dendritic side chains
can be minimized in the fully mixed and alternating (2a2c)3
hexamer (see 21).
Figure 11. Stability comparison of (2a)6 (2) with G‚C base-pair analogue
17b‚18b (9).
Conclusion
In addition to (2a)6 and (2c)6, monomers 2a and 2c can form
11 distinct hexameric aggregates (Figure 14). Thus, mixed
hexamers (2a)1(2c)5 and (2a)5(2c)1 have a single structure
possible, whereas (2a)2(2c)4 and (2a)4(2c)2 each have three
different structures as does (2a)3(2c)3. The broadness of the SEC
signal at intermediate mixing times (Figure 13) suggest that
many, if not all, of these of aggregates are sampled but that
ultimately a single aggregate is formed. Given the significantly
Compound 2 was designed with complementary DDA and
AAD hydrogen-bonding arrays fixed at a ca. 60° angle, thus
serving as an information code for cyclic assembly of a hexamer.
Extensive 1H NMR, SEC, and DLS studies on 2a-c and various
related compounds indicate that these compounds self-assemble
cooperatively into stable hexameric aggregates in apolar organic
solvents. The equilibrium is shifted from hexamer to monomer
9
13764 J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002

Supramolecular Polymer Chemistry
A R T I C L E S
Figure 13. Mixing of (2a)6 with (2c)6 monitored by SEC in toluene (UV
detector, 390 nm).
Figure 14. Possible mixed aggregates formed from mixing of (2a)6 and
(2c)6.
by increasing the temperature, adding polar solvents, or reducing
the concentration. The 1H NMR titration and SEC studies
demonstrate that the first generation dendritic monomer 2a forms
the most stable hexamer, with the stability of hexamers
decreasing upon increasing the dendron size: (2c)6 < (2b)6 <
(2a)6.
Experimental Section
General Methods. The following solvents were freshly distilled
under N2 prior to use: tetrahydrofuran from sodium and benzophenone;
acetonitrile (CH3CN) and triethylamine from calcium hydride. Dry tert-
butyl alcohol was obtained by storing the reagent grade tert-butyl
alcohol over molecular sieves overnight. All other solvents and reagents
were of reagent grade quality from Aldrich and used without further
purification. All reactions were run under a nitrogen atmosphere unless
otherwise noted.
Analytical thin-layer chromatography (TLC) was performed on 0.2
mm silica gel coated plastic sheets (Merck) with F254 indicator. Flash
chromatography was performed on Merck 40-63 µm silica gel.22
Preparative size exclusion chromatography was performed on poly-
styrene Bio-Beads S-X Beads (Bio-Rad: 400-14 000) in toluene.
Melting points were determined on a Thomas-Hoover melting point
apparatus without calibration.
Nuclear magnetic resonance (NMR) spectra were recorded on Varian
Unity U-400 (400 MHz), Varian Unity U-500 (500 MHz), Varian Unity
Inova 500 NB, or Varian Unity Inova (Keck) 750 (750 MHz)
instruments in chloroform-d (CDCl3) unless otherwise noted. Chemical
shifts are reported in parts per million (ppm) with referencing to the
solvent used, and coupling constants are reported in hertz (Hz).
Elemental analyses were performed by the Microanalysis Lab at the
University of Illinois School of Chemical Sciences. Mass spectra (MS)
were obtained on a Micromass ZAB-SE (FAB) mass spectrometer
and a PerSeptive Biosystems Voyager-DE-STR (MALDI-TOF) mass
spectrometer.
The dendron bromides 11a-c were synthesized3 by a convergent
approach similar to that used by Fre´chet.11 For the1H NMR assignments
of the dendron units in 2, 12, and 13, the protons of the nonfused phenyl
rings were given a prime (′), and their number increases proceeding
from the core to the phenyl rings bearing tert-butyl groups.
Ethyl 7-butanoylamino-1,8-naphthyridin-2(1H)-one-3-carboxylate
(7a) was synthesized by using a procedure similar to that was used to
make 7b.
2,6-Diaminopyridine-3-carboxaldehyde (5). This is a modification
of the previously reported procedure.5b To a stirred suspension of 10 g
(36 mmol) of protected pyridine 3 in 160 mL of THF was added
dropwise 60 mL of BuLi (2.2 mol/ L in hexane), maintaining the
A library of hexameric aggregates was obtained by recom-
bination of subunits within a mixture of preformed hexamers
(2a)6 and (2c)6. The dynamic nature of the supramolecular
assemblies is apparent, as the mixture appears to change over
time ultimately producing a hexamer with first- and third-
generation dendrons alternating. Given that there are at least
13 possible structures in the mixture, it is striking that a single,
discrete aggregate is self-selected. This use of steric hindrance
to direct the formation of a specific mixed aggregate should
prove to be particularly useful for creating nanoscale assemblies
with novel functions.
(22) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923-2925.
9
J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002 13765

A R T I C L E S
Ma et al.
internal temperature < -50 °C. The solid dissolved upon complete
addition. The flask was placed in a refrigerator (0 °C) overnight to
yield a yellow solid. The yellow suspension was stirred at 0 °C for 1
h, cooled to -78 °C, and quenched with a solution of 8.5 mL (110
mmol) of freshly distilled DMF in 20 mL of THF. The flask was
removed from the cooling bath, warmed to room temperature, and
placed in a 30 °C water bath. The resulting yellow solution was poured
onto 150 mL of ice water. The aqueous layer was extracted with 3 ×
100 mL of dichloromethane (CH2Cl2). The combined organic layers
were evaporated in vacuo. The resulting yellow oil was redissolved in
250 mL of CH2Cl2, and the solution was washed with 5% (w/w)
aqueous HCl and then saturated aqueous sodium bicarbonate (NaHCO3).
The organic layer was dried over sodium sulfate (Na2SO4), and the
solvent was removed in vacuo. The product was recrystallized from
petroleum ether-ethyl acetate (PE-EtOAc) to afford 7.1 g (65%) of
4 as a light yellow solid: mp 138-140 °C.
CO), 1.74 (pentet, J ) 7.5, 2H, CH3CH2CH2CH2CO), 1.45 (t, J ) 7.2,
3H, OCH2CH3), 1.42 (m, 2H, CH3CH2CH2CH2CO), 0.95 (t, J ) 7.3,
3H, CH3CH2CH2CH2CO); 1H NMR (DMSO-d6): δ 11.25 (s, 1H, NH),
8.92 (s, 1H, H-4), 8.58 (d, J ) 9.1, 1H, H-5), 8.48 (d, J ) 9.1, 1H,
H-6), 4.38 (q, J ) 7.1, 2H, OCH2CH3), 2.49 (t, J ) 7.8, 2H, CH3CH2-
CH2CH2CO), 1.59 (pentet, J ) 7.5, 2H, CH3CH2CH2CH2CO), 1.36 (t,
J ) 7.0, 3H, OCH2CH3), 1.33 (m, 2H, CH3CH2CH2CH2CO), 0.87 (t,
J ) 7.3, 3H, CH3CH2CH2CH2CO); 13C NMR (DMSO-d6): δ 174.2,
164.4, 157.4, 155.3, 150.1, 143.0, 141.2, 123.6, 118.8, 116.6, 62.5,
36.7, 27.4, 22.4, 14.6, 13.4. MS (EI, 70 eV): m/z 335.1 (20, M+), 306.1
(48), 293.1 (27), 251.1 (100), 206.1 (58). MS (FAB): 336.0 (M +
H)+. HRMS (FAB). Calcd for C16H19ClN3O3: 336.1115. Found:
336.1115. Anal. Calcd for C16H18ClN3O3: C, 57.23; H, 5.40; N, 12.51;
Cl, 10.56. Found: C, 56.93; H, 5.31; N, 12.25; Cl, 10.45.
2-Chloro-3-carbethoxy-7-amino-N-tert-BOC-N-butanoyl-1,8-naph-
thyridine (9a). To a suspension of 7.4 g (23 mmol) of 8a in 170 mL
of reagent grade dichloromethane were added 5.43 g (24.9 mmol) of
BOC-anhydride followed by 0.54 g (4.4 mmol) of 4-N,N-(dimethyl-
amino)pyridine (DMAP). After the addition of DMAP all solid
dissolved to result in a brown solution. The solution was stirred at room
temperature for 4 h at which time a TLC indicated complete consump-
tion of 8a. Then the brown solution was diluted with 100 mL of
dichloromethane and 20 mL of ethyl acetate and passed through a 50
mL plug of silica gel eluting with 10% EtOAc-CH2Cl2. The fractions
containing product were combined and the solvent was removed under
vacuum. The resulting brown oil was chromatographed over 250 mL
of silica gel (10% PE-CH2Cl2 to 10% EtOAc-CH2Cl2). The fractions
containing the product were combined, and the resulting yellow oil
was crystallized from a mixture of hexane and a small amount of ethyl
acetate to afford 8.0 g (75%) of the desired 9a as a white solid. An
analytically pure sample of 9a was obtained by second recrystallization
To a warm light yellow solution of 7.1 g (23.3 mmol) of 4 in 100
mL of methanol (MeOH) was added in one portion 2.0 g (47.6 mmol)
of LiOH‚H2O, and the resulting mixture was heated to reflux for 1 h.
The resulting yellow solution was cooled to room temperature and
neutralized with concentrated aqueous HCl to pH ) 6. The solvent
was evaporated in vacuo, and the solid was collected by vacuum
filtration. The solid was suspended in 80 mL of Et2O and filtered. The
crude solid was recrystallized using EtOAc to give 3.05 g (95%) of 5
as yellow powder, which was identical to that described previously.5b
Ethyl 7-Pentanoylamino-1,8-naphthyridin-2(1H)-one-3-carbox-
ylate (7b). A suspension of 6 (2.6 g, 11.2 mmol), 10 mL of valeric
anhydride, and 5 mL of pyridine was heated to reflux for 24 h, during
which time the granular solid disappeared forming a homogeneous
brown solution which deposited silver and platelike crystals. The
suspension was cooled to room temperature, diluted with petroleum
ether (30 mL), stirred, and filtered. The yellowish plates were collected,
suspended in 50 mL of dichloromethane, and filtered. Recrystallization
of crude product from DMF produced 3.3 g (93%) of product as a
1
from the same solvent: mp 113-115 °C. H NMR: δ 8.67 (s, 1H,
H-4), 8.29 (d, J ) 8.5, 1H, H-5), 7.46 (d, J ) 8.5, 1H, H-6), 4.46 (q,
J ) 7.0, 2H, OCH2CH3), 2.95 (t, J ) 7.2, 2H, CH2CH2CH3), 1.73 (m,
J ) 7.3, 2H, CH2CH2CH3), 1.41 (t, J ) 7.0, 3H, OCH2CH3), 1.34 (s,
9H, tert-Bu), 0.97 (t, J ) 7.5, 3H, CH2CH2CH3); 13C NMR: δ 175.8,
163.9, 157.1, 154.8, 151.4, 151.2, 141.4, 138.9, 126.3, 123.4, 119.6,
84.3, 62.4, 39.6, 27.7, 18.0, 14.1, 13.6. Anal. Calcd for C20H24N3O5Cl:
C, 56.94; H, 5.73; N, 9.96; Cl, 8.40. Found: C, 56.90; H, 5.72; N,
9.76; Cl, 8.48.
1
yellow needles: mp 321-322 °C. H NMR (DMSO-d6): δ 12.17 (s,
1H, NH), 10.67 (s, 1H, NH), 8.44 (s, 1H, H-4), 8.19 (d, J ) 8.6, 1H,
H-5), 7.98 (d, J ) 8.5, 1H, H-6), 4.25 (q, J ) 7.1, 2H, OCH2CH3),
2.46 (t, J ) 7.7, 2H, CH3CH2CH2CH2CO), 1.55 (pentet, J ) 7.5, 2H,
CH3CH2CH2CH2CO), 1.30-1.26 (m, 5H, CH3CH2CH2CH2CO +
OCH2CH3), 0.87 (t, J ) 7.4, 3H, CH3CH2CH2CH2CO); 13C NMR
(DMSO-d6): δ 173.8, 164.7, 160.4, 155.5, 150.7, 143.5, 140.9, 121.8,
109.8, 106.1, 61.2, 36.5, 27.6, 22.3, 14.8, 14.3. MS (EI, 70 eV): m/z
317.2 (36, M+), 233.1 (52), 188.1 (32), 161.1(100). HRMS (EI). Calcd
for C16H19N3O4: 317.1375. Found: 317.1372. Anal. Calcd for
C16H19N3O4: C, 60.56; H, 6.03; N, 13.24. Found: C, 60.27; H, 5.89;
N, 13.12.
2-Chloro-3-carbethoxy-7-amino-N-tert-BOC-N-pentanoyl-1,8-
naphthyridine (9b). Using the procedure described for 9a, a suspension
of 2.79 g (8.31 mmol) of 8b, 2.0 g (9.15 mmol) of BOC-anhydride,
and 0.22 g (1.8 mmol) of DMAP afforded a crude brown oil, which
was purified by column chromatography over silica gel (5-10%
EtOAc-CH2Cl2) to afford 3.52 g (97%) of 9b as white needles: mp
1
104-106 °C. H NMR: δ 8.70 (s, 1H, H-4), 8.30 (d, J ) 8.5, 1H,
Ethyl 2-Chloro-7-butanoylamino-1,8-naphthyridine-3-carbox-
ylate (8a). Using the published procedure,5b an analytically pure sample
was obtained by column chromatography over silica gel (1-3%
MeOH-CH2Cl2) followed by recrystallization from EtOAc-heptane
H-5), 7.48 (d, J ) 8.4, 1H, H-6), 4.49 (q, J ) 7.2, 2H, OCH2CH3),
3.00 (t, J ) 7.6, 2H, CH3CH2CH2CH2CO), 1.70 (m, 2H, CH3CH2CH2-
CH2CO), 1.46 (t, J ) 7.2, 3H, OCH2CH3), 1.42 (m, 2H, CH3CH2CH2-
CH2CO), 0.94 (t, J ) 7.3, 3H, CH3CH2CH2CH2CO). 13C NMR: δ
176.0, 164.0, 157.2, 154.8, 151.2, 141.4, 138.9, 127.5, 126.4, 123.5,
119.6, 84.3, 62.5, 37.6, 27.8, 26.6, 22.2, 14.2, 13.9. MS (FAB): 436.1
(M + H)+. HRMS (FAB). Calcd for C21H27ClN3O5: 436.1639.
Found: 436.1640. Anal. Calcd for C21H26ClN3O5: C, 57.86; H, 6.01;
N, 9.64; Cl, 8.13. Found: C, 57.73; H, 6.05; N, 9.51; Cl, 8.41.
1
(10% (v/v)) as short, white needles, mp: 202-204 °C. H NMR: δ
9.72 (s, 1H, NH), 8.67 (m, 2H, H-4, H-5), 8.27 (d, J ) 8.8, 1H, H-6),
4.44 (q, J ) 7.0, 2H, OCH2CH3), 2.55 (t, J ) 7.5, 2H, CH2CH2CH3),
1.74 (sextet, J ) 7.4, 2H, CH2CH2CH3), 1.43 (t, J ) 7.0, 3H,
OCH2CH3), 0.95 (t, J ) 7.3, 3H, CH2CH2CH3); 13C NMR: δ 173.2,
164.0, 156.5, 154.7, 151.5, 141.6, 139.8, 123.9, 118.2, 116.8, 62.2,
39.5, 18.4, 14.1, 13.6. Anal. Calcd for C15H16N3O3Cl: C, 55.99; H,
5.01; N, 13.06; Cl, 11.02. Found: C, 55.91; H, 4.95; N, 12.89; Cl,
11.20.
2-Chloro-3-carbethoxy-7-amino-N-tert-BOC-1,8-naphthyridine (10).
A suspension of 3.52 g (8.08 mmol) of 9b, 2.23 g (16.2 mmol) of
potassium carbonate in a mixture of 30 mL of reagent grade dichlo-
romethane, and 40 mL of reagent grade ethanol was stirred for 3 h at
room temperature. An additional 2.23 g (16.2 mmol) of potassium
carbonate were added. The suspension became yellow over time and
was stirred at room temperature for 24 h at which time TLC showed
complete consumption of 9b. The suspension was filtered to give a
yellow solid and the solvent removed from the filtrate under vacuum
retaining the residue. The yellow solid was partitioned between 80 mL
Ethyl 2-Chloro-7-pentanoylamino-1,8-naphthyridine-3-carbox-
ylate (8b). Using the published procedure,5b reaction of 7b (2.69 g,
8.5 mmol) with 25 mL of POCl3 afforded 2.17 g (76%) of the desired
product as light yellow flakelike crystals after recrystallized from DMF/
H2O: mp 192-193 °C. 1H NMR (CDCl3): δ 8.83 (brs, 1H, NH), 8.65
(s, 1H, H-4), 8.64 (d, J ) 9.0, 1H, H-5), 8.27 (d, J ) 9.0, 1H, H-6),
4.47 (q, J ) 7.1, 2H, OCH2CH3), 2.53 (t, J ) 7.4, 2H, CH3CH2CH2CH2-
9
13766 J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002

Supramolecular Polymer Chemistry
A R T I C L E S
of dichloromethane and 60 mL of water. The organic layer was
combined with the retained residue and washed with 100 mL of a
solution prepared by saturation with sodium bicarbonate and addition
of 1 mL of glacial acetic acid. The collected organic layer was dried
over Na2SO4, and the solvent was removed under reduced pressure.
The crude product was purified by column chromatography over silica
gel (5-10% EtOAc-CH2Cl2) to afford 2.40 g (93%) of the desired
0.52 g (13 mmol) of sodium hydride as a 60% (w/w) suspension in oil
and the mixture was heated to reflux for 10 min at which time hydrogen
gas evolution ceased. To the resulting turbid mixture was added 2 g
(7.35 mmol) of bis(N-ethyl guanidinium) sulfate and the mixture was
heated to reflux for 20 min. The resulting turbid solution was separated
from the insoluble salts by pipet and added to 1.7 g (2.2 mmol) of 13a
and then sealed into a glass tube. The resulting yellow mixture was
heated in a 160 °C oil bath for 24 h. All the initial 13a dissolved after
ca. 20 min, and a new solid formed after ca. 12 h. The resulting yellow
suspension was cooled to room temperature, and the solvent was
removed in vacuo. The resulting yellow residue was partitioned between
150 mL of dichloromethane and 120 mL of a saturated aqueous solution
of sodium bicarbonate and 2 mL of a 6 N aqueous solution of
hydrochloric acid. The collected organic layer was dried over Na2SO4,
and the solvent was removed in vacuo. The product was further purified
by column chromatography over silica gel (5% methanol-dichloro-
methane to 10% methanol-dichloromethane) followed by preparative
size exclusion chromatography in toluene over S-X beads (Bio-Rad;
400-14 000 Da) to afford 1.1 g (64%) of the desired 2a as a yellow
1
product as white needle crystals: mp 125-126 °C. H NMR: δ 8.63
(s, 1H, H-4), 8.42 (d (br), 1H, H-5), 8.20 (d, J ) 9.2, 1H, H-6), 7.78
(s, 1H, NH), 4.46 (q, J ) 7.4, 2H, OCH2CH3), 1.55 (s, 9H, tert-Bu),
1.44 (t, J ) 7.1, 3H, OCH2CH3); 13C NMR: δ 168.5, 156.4, 154.9,
151.8, 151.6, 141.6, 139.4, 123.6, 117.6, 114.4, 82.5, 62.6, 28.1, 14.2.
MS (FAB): 352.10 (M + H)+. HRMS (FAB). Calcd for C16H19-
ClN3O4: 352.1064. Found: 352.1062. Anal. Calcd for C16H18N3O4Cl:
C, 54.63; H, 5.16; N, 11.94; Cl, 10.08. Found: C, 54.45; H, 5.13; N,
11.80; Cl, 10.32.
2-Chloro-3-carbethoxy-7-amino-N-tert-BOC-N-(3′,5′-bis(3′′,5′′-di-
(tert-butyl)benzyloxy)benzyl)-1,8-naphthyridine (12a). A suspension
of 3.66 g (10.4 mmol) of chloronaphthyridine 10, 5.8 g (9.7 mmol) of
11a, 9.8 g (71 mmol) of potassium carbonate, and 136 mg (0.6 mmol)
of 18-crown-6 in a mixture of 60 mL of THF and 110 mL of acetonitrile
was heated to reflux for 4 h at which time TLC indicated consumption
of starting materials. The resulting yellow suspension was cooled to
room temperature, and the solvent was removed at reduced pressure.
The resulting yellow residue was partitioned between 200 mL of water
and 260 mL of dichloromethane. The organic layer was washed with
200 mL of water and dried over Na2SO4, and the solvent was removed
in vacuo. The resulting yellowish solid was recrystallized from
heptane-toluene to afford 6.8 g (80%) of 12a as a white solid. An
analytically pure sample of 12a was obtained by preparative thin-layer
chromatography over silica gel (30% petroleum ether-dichloromethane
to dichloromethane to 10% ethyl acetate-dichloromethane): mp 210-
1
solid: mp 218-220 °C. H NMR (DMSO-d6): δ 10.75 (br, 1H, NH-
3), 8.50 (s, 1H, H-5), 8.02 (br s, 1H, NH-8), 7.93 (br d, 1H, H-7), 7.33
(t, J ) 1.7, 2H, H-4′′), 7.24 (d, J ) 1.7, 4H, H-2′′, H-6′′), 6.76 (d, J
) 8.0, 1H, H-6), 6.68 (d, J ) 1.9, 2H, H-2′, H-6′), 6.61 (t, 1H, H-4′),
6.50 (br s, 1H, NH-2), 5.01 (s, 4H, OCH2Ar), 4.63 (d, J ) 5.1, 2H,
8-N-CH2), 3.40 (quintet, J ) 5.9, 2H, 2-N-CH2), 1.27 (s, 36H, tert-
1
Bu), 1.16 (t, J ) 7.1, 3H, NCH2CH3); H NMR (deacidified CDCl3):
δ 13.95 (s, 1H, NH-3), 10.89 (s, 1H, NH-2), 10.33 (s, 1H, NH-8), 8.21
(s, 1H, H-5), 7.33 (s br, 2H, H-4′′), 7.28 (d, J ) 8.7, 1H, H-6), 7.20
(br s, 4H, H-2′′, H-6′′), 6.76 (d, J ) 8.7, 1H, H-7), 6.68 (d, J ) 1.9,
2H, H-2′, H-6′), 6.61 (t, J ) 1.9, 1H, H-4′), 4.93 (AB q, J ) 11, 4H,
OCH2Ar), 4.72 (br d, J ) 10.9, 1H, 8-N-CH2), 4.42 (br d, J ) 13.8,
1H, 8-N-CH2), 4.33 (br s,1H, 2-N-CH2), 4.19 (br s, 1H, 2-N-CH2), 1.47
(br, 3H, NCH2CH3), 1.26 (s, 36H, tert-Bu); 13C NMR: δ 165.4, 162.6,
162.4, 160.5, 160.0, 154.2, 150.9, 141.4, 139.1, 137.6, 135.6, 122.3,
122.2, 113.4, 109.6, 107.1, 105.7, 100.6, 70.9, 46.6, 35.4, 34.8, 31.4,
16.6; IR (deacidified CHCl3): 1688. MS (MALDI): 784.4 (M + H)+.
MS (FAB): 783.5 (M)+. Anal. Calcd for C49H62N6O3: C, 75.16; H,
7.98; N, 10.73. Found: C, 75.16; H, 7.79; N, 10.77.
1
211 °C. H NMR: δ 8.60 (s, 1H, H-4), 8.28 (d, J ) 8.8, 1H, H-5),
8.09 (d, J ) 8.8, 1H, H-6), 7.38 (t, J ) 1.5, 2H, H-4′′), 7.24 (d, J )
1.5, 4H, H-2′′, H-6′′), 6.67 (d, J ) 2.1, 2H, H-2′, H-6′), 6.53 (t, J )
2.1, 1H, H-4′), 5.48 (s, 2H, NCH2Ar), 4.94 (s, 4H, OCH2Ar), 4.45 (q,
J ) 7.1, 2H, OCH2CH3), 1.48 (s, 9H, tert-BuO), 1.44 (t, J ) 7.1, 3H,
OCH2CH3), 1.32 (s, 36H, tert-Bu); 13C NMR: δ 164.2, 164.0, 158.8,
154.7, 153.9, 151.1, 150.9, 141.3, 140.9, 137.5, 135.8, 123.9, 122.2,
121.1, 119.8, 117.6, 106.6, 100.5, 82.9, 70.8, 62.1, 49.4, 34.8, 31.4,
28.0, 14.1. Anal. Calcd for C53H68N3O6Cl: C, 72.45; H, 7.80; N, 4.78;
Cl, 4.04. Found: C, 72.65; H, 7.88; N, 4.61; Cl, 4.21.
2-Chloro-3-carbethoxy-7-amino-N-tert-BOC-N-(3′,5′-bis(3′′,5′′-
bis(3′′′,5′′′-di(tert-butyl)benzyloxy)benzyloxy)benzyl)-1,8-naphthyr-
idine (12b). Using the procedure described for 12a, a suspension of
0.352 g (1.0 mmol) of chloronaphthyridine 10, 1.255 g (1.0 mmol) of
11b (G2-Br), 0.966 g of potassium carbonate, and 14 mg of 18-crown-6
in a mixture of 6 mL of THF and 11 mL of acetonitrile afforded
yellowish oil, which was purified twice by column chromatography
2-Chloro-3-carbethoxy-7-amino-N-(3′,5′-bis-(3′′,5′′-di-tert-butyl)-
benzyloxy)benzyl-1,8-naphthyridine (13a). To a clear solution of 1.3
g (1.5 mmol) of 12a in 50 mL of reagent grade dichloromethane were
added 0.8 mL of trifluoroacetic acid. The solution was stirred overnight
at room temperature at which time it became pale yellow. The solution
was washed with 40 mL of a saturated aqueous solution of sodium
bicarbonate, collected, and dried over Na2SO4, and the solvent was
removed at reduced pressure. The resulting yellowish solid was
recrystallized from heptane-toluene to afford 1 g (86%) of the desired
13a as a white solid. An analytically pure sample of 13a was obtained
by column chromatography over silica gel (1-5% methanol-
1
(CH2Cl2-hexane ) 2:1) to give 0.72 g of 12b (47%). H NMR: δ
8.58 (s, 1H, H-4), 8.26 (d, J ) 9.0, 1H, H-5), 8.05 (d, J ) 9.0, 1H,
H-6), 7.40 (t, J ) 1.7, 4H, H-4′′′), 7.27 (d, J ) 1.7, 8H, H-2′′′, H-6′′′),
6.69 (d, J ) 2.4, 4H, H-2′′, H-6′′), 6.62 (t, 2H, H-4′′), 6.60 (d, 2H,
H-2′, H-6′), 6.49 (t, 1H, H-4′), 5.46 (s, 2H, NCH2), 4.98 (s, 8H,
Ar′′′CH2OAr′′), 4.93 (s, 4H, Ar′′CH2OAr′), 4.44 (q, J ) 7.0, 2H, OCH2-
CH3), 1.43 (s, 9H, tert-BuO), 1.42(t, J ) 7.0, 3H, OCH2CH3), 1.33(s,
72H, tert-Bu); 13C NMR: δ 164.0, 160.2, 160.1, 159.7, 154.6, 153.7,
151.0, 150.9, 141.3, 141.0, 139.0, 137.5, 135.5, 123.8, 122.2, 122.1,
117.5, 106.4, 106.2, 106.1, 101.3, 100.5, 82.8, 70.9, 69.9, 62.0, 34.7,
31.3, 27.9, 14.1; MS (MALDI): m/z 1527.65 (M + H)+; MS (FAB):
m/z 1528.0 (M + H)+. Anal. Calcd for C97H124ClN3O10: C, 76.27; H,
8.18; N, 2.75. Found: C, 76.69; H, 8.54; N, 2.31.
1
chloroform): mp 229-231 °C. H NMR: δ 8.44 (s, 1H, H-4), 7.77
(d, J ) 8.8, 1H, H-5), 7.42 (t, J ) 1.7, 2H, H-4′′), 7.27 (d, J ) 1.7,
4H, H-2′′, H-6′′), 6.72 (d, J ) 8.8, 1H, H-6), 6.68 (d, J ) 2.1, 2H,
H-2′, H-6′), 6.53 (t, J ) 2.1, 1H, H-4′), 4.82 (d, J ) 5.6, 2H, NCH2-
Ar), 4.98 (s, 4H, OCH2Ar), 4.42 (q, J ) 7.0, 2H, OCH2CH3), 1.43 (t,
J ) 7.1, 3H, OCH2CH3), 1.34 (s, 36H, tert-Bu); 13C NMR: (THF) δ
164.8, 161.9, 161.5, 158.6, 151.4, 150.9, 142.3, 141.9, 137.7, 137.6,
122.9, 122.4, 120.3, 116.3, 115.7, 107.8, 101.3, 71.3, 61.8, 46.0, 35.4,
31.8, 14.6. Anal. Calcd for C48H60N3O4Cl: C, 74.06; H, 7.77; N, 5.40;
Cl, 4.55. Found: C, 74.19; H, 7.89; N, 5.42; Cl, 4.73.
2-Chloro-3-carbethoxy-7-amino-N-(3′,5′-bis(3′′,5′′-bis(3′′′,5′′′-di-
tert-butyl)benzyloxy)benzyloxy)benzyl-1,8-naphthyridine (13b). Us-
ing the procedure described for 13a, a solution of 0.72 g of 12b in 20
mL of reagent grade CH2Cl2 and 0.5 mL of trifluoroacetic acid afforded
yellowish oil, which was purified by column chromatography (2%
1
2,8-Diamino-2-N-ethyl-8-N-(3′,5′-bis(3′′,5′′-di-tert-butyl)benzyloxy)-
benzylpyrimido(4,5-b)(1,8)naphthyridine-3H-4-one (2a). To a mix-
ture of 40 mL of THF and 40 mL of tert-butyl alcohol were added
MeOH-CHCl3) to give 0.30 g (45%) of 13b. H NMR: δ 8.44 (s,
1H, H-4), 7.76 (d, J ) 9.2, 1H, H-5), 7.40 (t, J ) 1.6, 4H, H-4′′′), 7.27
(d, J ) 1.7, 8H, H-2′′′, H-6′′′), 6.71 (d, J ) 2.4, 4H, H-2′′, H-6′′), 6.67
9
J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002 13767

A R T I C L E S
Ma et al.
(d, J ) 9.0, 1H, H-6), 6.64 (t, J ) 2.3, 2H, H-4′′), 6.63 (d, J ) 2.5,
2H, H-2′, H-6′), 6.58 (t, J ) 2.2, 1H, H-4′), 4.99 (s, 8H, OCH2′′Ar),
4.98 (s, 4H, OCH2′Ar), 4.78 (s, 2H, 8-N-CH2), 4.43 (q, J ) 7.1, 2H,
OCH2CH3), 1.43 (t, J ) 7.1, 3H, OCH2CH3), 1.33 (s, 72H, tert-Bu);
13C NMR: δ 167.9, 160.7, 160.5, 151.3, 141.7, 139.1, 135.8, 124.9,
122.6, 122.2, 106.7, 106.6, 101.7, 101.4, 70.4, 57.8, 35.1, 31.7, 14.4;
MS (MALDI): 1426.79 (M + H)+. Anal. Calcd for C92H116ClN3O8:
C, 77.41; H, 8.19; N, 2.94. Found: C, 77.37; H, 8.35; N, 2.76.
120.4, 115.6, 106.7, 106.6, 101.9, 101.8, 71.3, 61.9, 57.8, 35.1, 31.7,
14.4; MS (MALDI): 2724.83; MS (FAB): 2725.0 (M + H)+.
2,8-Diamino-2-N-ethyl-(8-N-(3′,5′-bis(3′′,5′′-bis(3′′′,5′′′-bis(3′′′′,5′′′′-
di-tert-butyl)benzyloxy)benzyloxy)benzyloxy)benzyl)pyrimido(4,5-
b)(1,8)naphthyridine-3H-4-one (2c). Using the procedure described
for 2b, 0.8 g of 13c (0.29 mmol) was used to react with N-ethyl-
guanidinium generated from 0.27 g of bis(N-ethylguanidinium) sulfate
and sodium hydride to give 0.51 g (64%) of 2c as a yellow solid: mp
1
120-122 °C. H NMR (deacidified CDCl3): δ 14.40 (s, 1H, NH-3),
2,8-Diamino-2-N-ethyl-(8-N-(3′,5′-bis(3′′,5′′-bis(3′′′,5′′′-di-tert-bu-
tyl)benzyloxy)benzyloxy)benzyl)pyrimido(4,5-b)(1,8)naphthyridine-
3H-4-one (2b). Using the procedure described for 2a, in a glass sealed
tube placed in a 160 °C oil bath, the reaction of 0.3 g (0.21 mmol) of
13b with N-ethylguanidine which was generated by refluxing a solution
of 0.2 g (0.735 mmol) of bis(N-ethylguanidinium) sulfate and 52 mg
(1.3 mmol) of sodium hydride in a mixture of 4 mL of dry THF and
4 mL of dry tert-butyl alcohol afforded yellowish residue, which was
purified by column chromatography (2-5% MeOH-CH2Cl2), followed
by preparative size exclusion chromatography in toluene over S-X
beads (Bio-Rad: 400-14 000 Da) to afford 80 mg of 2b (30%) as a
yellow solid: mp 138-140 °C. 1H NMR: δ (deacidified CDCl3) 14.40
(s, 1H, NH-3), 10.90 (s, 1H, NH-2), 10.10 (s, 1H, NH-8), 8.65 (s, 1H,
H-5), 7.72 (d, J ) 8.7, 1H, H-6), 7.39 (t, 4H, H-4′′′), 7.24 (d, 8H,
H-2′′′, H-6′′′), 7.18 (d, J ) 8.7, 1H, H-7), 6.71 (d, 4H, H-2′′, H-6′′),
6.62 (t, 2H, H-4′′), 6.56 (d, 2H, H-2′, H-6′), 6.54 (t, 1H, H-4′), 5.00 (s,
12H, -CH2OAr), 4.60 (s, br, 2H, 8-N-CH2), 4.16 (s, br, 2H, 2-N-CH2),
1.45 (br, 3H, NCH2CH3), 1.29 (s, 72H, tert-Bu); 13C NMR: δ 172.8,
160.9, 160.6, 151.3, 139.2, 138.8, 135.9, 125.1, 122.6, 118.1, 106.8,
105.9, 101.8, 97.2, 71.3, 54.3, 35.1, 31.8, 29.3, 16.2; MS (MALDI):
1430.4 (M)+; MS (FAB): 1431.8 (M + H)+. Anal. Calcd for
C93H118N6O7: C, 78.00; H, 8.31; N, 5.87. Found: C, 77.81; H, 8.41;
N, 5.56.
10.90 (s, 1H, NH-2), 10.10 (s, 1H, NH-8), 8.64 (s, 1H, H-5), 7.68 (brs,
1H, H-6), 7.38 (brs, 8H, H-4′′′′), 7.34 (brs, 8H, H-2′′′, H-6′′′), 7.22 (d,
J ) 2.5, 16H, H-2′′′′, H-6′′′′), 7.17 (brs, 1H, H-7), 6.70 (brs, 4H, H-4′′′),
6.68 (brs, 1H, H-2′′, H-6′′), 6.63 (brs, 2H, H-2′, H-6′), 6.60 (brs, 2H,
H-4′′), 6.55 (brs, 1H, H-4′), 5.0 (brs, 26H, -OCH2Ar), 4.80 (d, br,
2H, -CH2N-8), 4.55 (d, br, 2H, -CH2N-2), 1.43 (br, 3H, -NCH2CH3),
1.31 (s, 144H, tert-Bu); 13C NMR: δ 160.6, 160.4, 157.7, 151.3, 151.2,
139.2, 135.9, 122.6, 122.5, 122.4, 106.7, 101.7, 71.2, 57.8, 35.1, 31.7,
31.4, 16.8; MS (MADLI): 2730.34 (M + H)+. Anal. Calcd for
C
181H230N6O15: C, 79.64; H, 8.49; N, 3.08. Found: C, 79.67; H, 8.54;
N, 3.04.
Size Exclusion Chromatography. Analytical SEC was performed
on a Hitachi SEC system including a Model L-6000 HPLC pump,
Model L-4000H UV detector, Model D-2520 GPC integrator, and a
Waters 410 differential refractometer. Two Waters Styragel HR3 (7.8
× 300 mm; effective molecular weight range, 500-30 000) SEC
columns were connected in series. The flow rate was 1 mL/min unless
otherwise noted.
Dynamic Light Scattering. Performed on a DynaPro-MS/X from
Protein Solutions, Inc., (Charlottesville, VA) equipped with He-Ne
Laser (10 mW) operating at 632.8 nm. All samples are measured at
90° scattering angle. Solutions were prepared in HPLC grade toluene
(Sigma, 99.8% pure) and were filtered with 0.2 µm Nylon filter from
Millipore.
The autocorrelation function of the intensity was analyzed by the
method of cumulants analysis to obtain the average diffusion coefficient,
D, of the particles and the polydispersity. The hydrodynamic diameter,
Dh, was calculated by means of the Stokes-Einstein equation (Dh )
kBT/3πηD, where kBT is the thermal energy and η is the viscosity of
the continuous phase).
Molecular Modeling. Modeling was performed using Macromodel
and Cerius2 (Accelrys) on a Silicon Graphics workstation. The radius
of gyration (Rg) of all models was calculated by Cerius2. The
DREIDING2.21 force field was used with the energy of monomers
and covalent standard 19 minimized first. The intramolecular hydrogen
bond in monomer 20 was preformed. Six molecules of monomers were
then arranged to form the corresponding hexamer containing 18 primary
hydrogen bonds for 2, 24 for both 1 and 20. The threshold for a primary
hydrogen bond was set to a 2.1 Å.9 The energy was minimized with
the resulting hexamers each maintaining all of the initial hydrogen
bonds. During the entire simulated annealing process of the dynamics
simulation, the hexamer core was held fixed while leaving the dendron
substituents to change conformation. Both extended and extensively
folded back (compact) starting structures were generated by manual
bond rotations. With both starting structures dynamics simulations were
performed 10 times with a heating cycle from 300 to 1000 K with
increments of 50 K in 1000 steps (1 step ) 1 fs). The overall process
generated 20 frames of minimized structures for each starting compound
structure. A separate value for the radius of gyration (Rg) of the extended
and compact structures was taken as an average of these 20 minimized
structures.
2-Chloro-3-carbethoxy-7-amino-N-tert-BOC-N-(3′,5′-bis(3′′,5′′-
bis(3′′′,5′′′-bis(3′′′′,5′′′′-di-tert-butyl)benzyloxy)benzyloxy)benzyloxy)-
benzyl-1,8-naphthyridine (12c). Using the procedure described for
12b, reaction of 11c (G3-Br, 1.22 g, 0.5 mmol) with 10 (0.183 g, 0.5
mmol) in the presence of potassium carbonate and 18-crown-6 in CH3-
CN/THF afforded 0.95 g (70%) of 12c as yellow oil. 1H NMR: δ 8.54
(s, 1H, H-4), 8.24 (d, J ) 9.0, 1H, H-5), 8.03 (d, J ) 9.2, 1H, H-6),
7.39 (t, J ) 1.8, 8H, H-4′′′′), 7.27 (d, J ) 1.9, 16H, H-2′′′′, H-6′′′′),
6.72 (d, J ) 2.1, 8H, H-2′′′, H-6′′′), 6.66 (d, J ) 2.2, 4H, H-2′′, H-6′′),
6.64 (t, J ) 2.2, 4H, H-4′′′), 6.60 (d, J ) 2.2, 2H, H-2′, H-6′), 6.57 (t,
J ) 2.2, 2H, H-4′′), 6.51 (t, J ) 2.3, 1H, H-4′), 5.44 (s, 2H, NCH2-
Ar′), 5.00 (s, 16H, OCH2Ar′′′′), 4.97 (s, 8H, OCH2Ar′′′), 4.92 (s, 4H,
OCH2Ar′′), 4.40 (q, J ) 7.2, 2H, COOCH2CH3), 1.41 (s, 9H, tert-
butyl), 1.38 (t, J ) 7.1, 3H, COOCH2CH3), 1.32(s, 144H, tert-Bu);
13C NMR: δ 164.4, 160.7, 160.4, 160.1, 159.1, 154.9, 151.4, 151.3,
141.7, 139.6, 139.2, 137.9, 135.9, 122.6, 122.5, 122.4, 119.9, 117.9,
106.7, 106.6, 101.8, 71.2, 70.3, 62.3, 57.8, 35.1, 31.7, 28.3, 14.4; MS
(MALDI): 2847.92 (M + Na)+; MS (FAB): 2725.8 (M - BOC)+.
Anal. Calcd for C185H236ClN3O18: C, 78.47; H, 8.51; N, 1.29. Found:
C, 78.65; H, 8.42; N, 1.49.
2-Chloro-3-carbethoxy-7-amino-N-(3′,5′-bis(3′′,5′′-bis(3′′′,5′′′-bis-
(3′′′′,5′′′′-di-tert-butyl)benzyloxy)benzyloxy)benzyloxy)benzyl-1,8-
naphthyridine (13c). Using the procedure described for 13b, a solution
of 2.26 g of 12c (0.8 mmol) and 0.5 mL of trifluoroacetic acid (TFA)
in 30 mL of CH2Cl2 afforded 1.40 g (64%) of 13c as a pale yellow
1
solid: mp 108-111 °C. H NMR: δ 8.40 (s, 1H, H-4), 7.71 (d, J )
9.2, 1H, H-5), 7.39 (t, J ) 1.9, 8H, H-4′′′′), 7.27 (d, J ) 1.9, 16H,
H-6′′′′, H-2′′′′), 6.72 (d, J ) 2.2, 8H, H-6′′′, H-2′′′), 6.67 (d, J ) 2.2,
4H, H-6′′, H-2′′), 6.64 (t, J ) 2.2, 4H, H-4′′′), 6.59 (t, J ) 2.2, 2H,
H-4′′), 6.58 (d, J ) 2.2, 2H, H-6′, H-2′), 6.57 (t, J ) 2.2, 1H, H-4′),
6.55 (d, J ) 8.9, 1H, H-6), 4.98 (s, 16H, OCH2Ar′′′′), 4.97 (s, 8H,
OCH2Ar′′′), 4.95 (s, 4H, OCH2Ar′′), 4.95 (d, 2H, NCH2Ar), 4.72 (s,
br, 1H, NHCH2Ar′), 4.40 (q, J ) 7.1, 2H, COOCH2CH3), 1.41 (t, J )
7.1, 3H, COOCH2CH3), 1.32 (s, 144H, tert-Bu); 13C NMR: δ 164.7,
160.7, 160.5, 160.4, 151.3, 141.6, 139.3, 139.2, 135.9, 122.6, 122.5,
1H NMR Dilution and Titration Study. Chloroform-d was dea-
cidified by running through a column of activated basic aluminum
oxide. DMSO-d6 was used without further purification. NMR tubes
and microsyringes were dried in oven and cooled under nitrogen.
Dilution Studies. A stock solution of sample with known concentra-
tion was prepared in a chosen solvent (CDCl3, 5% or 10% of DMSO-
9
13768 J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002

Supramolecular Polymer Chemistry
A R T I C L E S
d6 in CDCl3 (v/v)), and a 1H NMR spectrum was recorded. The
temperature was maintained at 20 °C. Half of the sample was removed
and replaced by solvent. After sufficient mixing, the spectrum was
recorded for the second sample whose concentration is half of the first
sample. This procedure was repeated until the sample was too dilute
to give a detectable spectrum (4 h at 750 MHz). For 2a, the ratio of
hexamers to monomers was calculated on the basis of the integration
values of two sets of tert-butyl proton signals under the assumption of
slow exchange between hexamers and monomers on the NMR time
scale. For complex 17‚18, the observed chemical shifts represent the
weighted average of the free and complexed monomers due to their
fast exchange on the NMR time scale. The ratio of complexed species
maintained at 20 °C. Then 0.01 mL (0.05 mL for 2a) of DMSO-d6
was added to the solution. After sufficient mixing, the 1H NMR
spectrum was recorded. This procedure was repeated until no hexamer
signal could be observed in the spectrum. Once again, the ratio of
hexamers to monomers was calculated on the basis of the integration
values of two sets of tert-butyl proton signals under the assumption of
slow exchange between hexamers and monomers on an NMR time
scale.
Acknowledgment. This work was supported by the National
Institutes of Health (Grant GM39782). The authors thank Dr.
Ilya Zharov for assistance with the artwork and Dr. Gabriel
Lemcoff for help on molecular modeling.
could be calculated by the following equation: complex % ) ∆δ/∆δmax
,
while ∆δ ) δ - δmonomer and ∆δmax ) ∆δcomplex - ∆δmonomer
.
Titration Studies. 1H NMR spectrum of 0.65 mL of stock solution
of 2 in deacidified CDCl3 was recorded. The temperature was
JA0202006
9
J. AM. CHEM. SOC. VOL. 124, NO. 46, 2002 13769