β-alanine-based dendritic β-peptides: Dendrimers possessing unusually strong binding ability towards protic solvents and their self-assembly into nanoscale aggregates through hydrogen-bond interactions

Article abstract of DOI:10.1002/1521-3765(20010202)7:3<686::AID-CHEM686>3.0.CO;2-Z

A series of poly(β-alanine) dendrimers 1-4 with Boc-carbamate as the surface functionality, β-alanine as the dendritic branch, 3,5-diaminobenzoic acid as the branching agent, and 1,2-diaminoethane as the interior core has been synthesized by a solution-phase peptide-coupling method. The structural identities and purities of the products have been fully characterized by spectroscopic and chromatographic methods. ¹H NMR studies on the dendrimers indicated that the Boc-carbamate surface groups exist as a mixture of syn and anti rotamers in solution, and that the dendrimers adopt an open structure in polar solvents; this allows the free interaction of the interior core functionality with solvent molecules. Due to the cooperative effect of a large number of carbamate and amide groups, the dendrimers exhibit an unusually strong binding ability towards protic solvents and behave as H-bond sponges. As a result, the H/D exchange rates of the N - H protons are significantly enhanced in such dendritic structures, as compared to those of nondendritic carbamates and amides. These dendritic peptide dendrimers also exhibit a strong tendency to form nanoscopic aggregates in nonpolar or polar aprotic solvents through intermolecular H-bond interactions.

Full text of DOI:10.1002/1521-3765(20010202)7:3<686::AID-CHEM686>3.0.CO;2-Z

FULL PAPER
b-Alanine-Based Dendritic b-Peptides: Dendrimers Possessing Unusually
Strong Binding Ability Towards Protic Solvents and Their Self-Assembly into
Nanoscale Aggregates through Hydrogen-Bond Interactions
Tony K.-K. Mong,[a] Aizhen Niu,[a] Hak-Fun Chow,*[a] Chi Wu,[a]
Liang Li,[b] and Rui Chen[b]
Abstract: A series of poly(b-alanine)
dendrimers 1 ± 4 with Boc-carbamate as
the surface functionality, b-alanine as
the dendritic branch, 3,5-diaminobenzo-
ic acid as the branching agent, and 1,2-
diaminoethane as the interior core has
been synthesized by a solution-phase
peptide-coupling method. The structural
identities and purities of the products
have been fully characterized by spec-
troscopic and chromatographic meth-
ods. 1H NMR studies on the dendrimers
indicated that the Boc-carbamate sur-
face groups exist as a mixture of syn and
anti rotamers in solution, and that the
dendrimers adopt an open structure in
polar solvents; this allows the free
interaction of the interior core function-
ality with solvent molecules. Due to the
cooperative effect of a large number of
carbamate and amide groups, the den-
drimers exhibit an unusually strong
binding ability towards protic solvents
and behave as H-bond sponges. As a
result, the H/D exchange rates of the
N H protons are significantly enhanced
in such dendritic structures, as compared
to those of nondendritic carbamates and
amides. These dendritic peptide den-
drimers also exhibit a strong tendency to
form nanoscopic aggregates in nonpolar
or polar aprotic solvents through inter-
molecular H-bond interactions.
Keywords: aggregation ´ amides ´
dendrimers ´ peptides
We have been interested in using synthetic dendritic
systems to model the properties of redox proteins[4] and
enzyme molecules.[5] Prompted by the impressive work on
dendritic peptides by Zimmerman[1d] and the recent surge of
interest in the conformational and biological properties of b-
peptides,[6] we began to look into the preparation of dendritic
macromolecules constructed from b-amino acid residues and
to investigate their peptide-like properties. In fact, b-peptide
dendrimers, apart from poly(amidoamine) (PAMAM) den-
drimers,[7] have been less well studied than a-peptide den-
drimers. Herein, we report a facile solution-phase convergent
synthesis of a series of poly(b-alanine) dendrimers 1 ± 4, in
which the b-alanine units are connected by amide linkages by
using aromatic branching units, and show that they form
nonspecific yet discrete self-assembling aggregates of differ-
ent nanoscopic sizes in various solvent systems. Furthermore,
due to the presence of a large number of amide functionalities
inside the dendritic structure, they exhibit unusually high
cooperative binding abilities towards protic solvents as a
result of H-bond interactions relative to their nondendritic
counterparts.
Introduction
There is an increasing interest in the synthesis of peptide-
based[1] (i.e. those in which the amino acid units are directly
connected to each other) and a-amino acid based (i.e. those in
which the amino acid units are not directly linked to each
other) dendrimers[2] owing to their structural resemblance to
globular proteins. Such biomimetic dendrimers represent
potentially novel drug-delivery agents or bio-drugs with
enhanced biocompatibilities. They have also been used as
biological mimics for the exploration and understanding of
the biological functions of proteins.[3] However, the synthesis
of such peptide- or polyamide-based compounds also presents
a substantial synthetic challenge because of their high polar-
ity. To facilitate their preparations, amino acid based den-
[2a]
[2c]
drimers with nonpolar linkages such as C C
bonds to connect the amino acid residues together have been
reported.
or C O
[a] Prof. H.-F. Chow, T. K.-K. Mong, A. Niu, Prof. C. Wu
Department of Chemistry, The Chinese University of Hong Kong
Shatin, NT, Hong Kong (PR China)
Although the formation of well-defined aggregates from
carefully pre-designed dendritic subunits has been well
documented,[8] the preparation and self-assembly mechanism
of nonspecific dendritic aggregates from simple dendritic
fragments has been less well studied. This approach deserves
Fax : (‡852)26035057
E-mail: hfchow@cuhk.edu.hk
[b] Prof. L. Li, R. Chen
Department of Chemistry, University of Alberta
Edmonton T6G 2G2 (Canada)
686
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0686 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 3

686±699
further attention because it allows the construction of nano-
systems with novel functions and properties without resorting
to elaborate syntheses of highly functionalized subunits. It
therefore represents an alternative and efficient strategy for
the preparation of large nanoscopic systems. For example,
cylindrical and spherical nanostructures have been generated
through the self-assembly of flat tapered and conical poly-
ether monodendrons, respectively, presumably as a result of
hydrophobic interactions.[9] Highly spherical metallodendri-
mers of 200 nm in diameter have also been assembled by
virtue of the coordination chemistry of AB2 organopalladium
monomers.[10] Aggregates with long fibrous rod-like structures
have been prepared from arborols, again through hydro-
phobic interactions.[11] In this report, we wish to demonstrate
that large nanoscopic aggregates may also be generated by
hydrogen bonding from this series of poly(b-alanine) den-
drimers.
Results and Discussion
Structural features of the poly(b-alanine) dendrimers 1 ± 4:
The target poly(b-alanine) dendrimers consist of b-alanine as
the dendritic branch, 3,5-diaminobenzoic acid 5 as the
branching agent, and 1,2-diaminoethane 6 as the central core
unit. With this design, both the amino and carboxylic acid
moieties in the various building blocks could be coupled to b-
alanine solely through amide/peptide linkages.[12] Further-
more, we wished to develop a synthetic route that could be
extended to allow the incorporation of different amino acid
residues into the dendritic architecture if we were to replace
the b-alanine residue with different amino acids in the future.
The synthetic operation should also be amenable to large-
scale preparation of the dendrimers in order to facilitate
investigation of their properties. A convergent synthetic
scheme[13] was chosen because it provided a better control
over the placement of different amino acid residues in
different dendritic layers and a better chance of producing
structurally perfect dendrimers. To avoid uncontrolled cou-
pling between 3,5-diaminobenzoic acid 5 and b-alanine,
selective protection of the amino and carboxylic acid func-
tions was necessary. Our initial goal was to prepare a series of
Abstract in Chinese:
Chem. Eur. J. 2001, 7, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0687 $ 17.50+.50/0
687

H.-F. Chow et al.
FULL PAPER
carboxylic acid dendrons 7 ± 9, which were subsequently
coupled to the central 1,2-diaminoethane core to furnish the
target dendrimers.
ethoxycarbonyl-1,2-dihydroquinoline (EEDQ)[14] to provide
the G1-ester 11 as the key AB2 monomer unit (Scheme 1).
Base hydrolysis of the ester group of 11 and subsequent work-
up afforded the carboxylic acid 7 as a crystalline solid.
Alternatively, the Boc-protecting groups could be removed by
treatment of 11 with HCl (1.2m) to furnish the diammonium
salt 12.
Synthesis: Ethyl 3,5-diaminobenzoate 10, prepared by acid-
catalyzed esterification from the acid 5, was treated with two
equivalents of Boc-b-alanine in the presence of 2-ethoxy-1-
Scheme 1. i) H2SO4, EtOH; ii) Na2CO3; iii) Boc-b-alanine, EEDQ, THF; iv) NaOH, H2O, MeOH, THF; v) HCl, EtOH, H2O; vi) DCC, HOBt, DIPEA,
DMF, 12.
688
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0688 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 3

Dendritic b-Peptides
686±699
Reaction of 7 and 12 in the presence of EEDQ and
triethylamine failed to produce the desired G2-ester dendron
13. After several attempts, it was found that effective coupling
between these two fragments could be realized through the
use of dicyclohexylcarbodiimide (DCC), hydroxybenzotri-
azole (HOBt), and diisopropylethylamine (DIPEA) in DMF,
providing the G2-ester 13 in 70% yield. Compound 13,
despite having eight amide and four carbamate functionali-
ties, proved to be soluble in a number of organic solvents such
as alcohols, THF, and DMF. Ester 13 was then subjected to
base hydrolysis to furnish the Boc-protected G2-(carboxylic
acid) 8, which could be obtained in a pure state (74% yield)
after flash chromatography on silica gel. By using the same
iterative reaction sequence, the Boc-protected G3-ester 14
was synthesized as a white amorphous solid in 70% yield from
the carboxylic acid 8 and the diammonium salt 12. Base
hydrolysis of the ester 14 gave the Boc-protected G3-
(carboxylic acid) 9 in 70% yield. The overall yield of the
Boc-protected G3-ester 14 based on Boc-b-alanine was 23%.
The target dendrimers 1 ± 3 could, in principle, be assem-
bled by treating the respective Boc-protected Gn-(carboxylic
acid) dendrons 7 ± 9 with 1,2-diaminoethane. In practice,
however, a multivalent central core unit 15 was synthesized in
order to shorten the synthetic route (Scheme 2). This core unit
dendron for use in the preparation of the G4-dendrimer 4.
The tetraammonium salt 15 was then converted in situ to
the corresponding free tetraamine by reaction with four
equivalents. of DIPEA, which was then coupled to the
respective carboxylic acid dendrons 7 ± 9 in the presence of
DCC and HOBt to furnish the G2- to G4-dendrimers 2 ± 4.
The target compounds were purified by flash chromatography
on either neutral alumina or silica gel to remove the by-
products. Due to the extremely high polarity of these
compounds, substantial loss of material, especially for the
higher generation dendrimers, was noted. Hence, the per-
centage yield of the G2-dendrimer 2 was 70%, while those of
the G3- and G4-dendrimers 3 and 4 were both 40%.
Characterization
1H NMR spectroscopy: All of the poly(b-alanine) dendrimers
and the relevant intermediates were characterized by 1H and
13C NMR spectroscopy. Because the poly(b-alanine) den-
drimers possess a large number of polar amide bonds, they
tend to entrap solvent molecules. Generally, the sample was
placed under high vacuum at elevated temperature (ꢀ808C)
for two days to remove the residual solvents. This operation
was essential in order to obtain a satisfactory NMR spectrum
free of solvent interference. Due to the poor solubility of the
dendrimers in chloroform, their NMR spectra were recorded
in [D6]DMSO. Because the residual proton signals and the
signals of the dissolved water in [D6]DMSO partially over-
lapped with the resonance signals of the methylene protons of
the b-alanine moiety and those of the central 1,2-diamino-
ethane core, the relative integrals in this region could not be
determined accurately.
The 1H NMR spectra of the various dendrons and
dendrimers are shown in Figures 1 and 2, respectively. Despite
their large molecular size, the symmetrical architecture of
these products simplified the task of peak assignment. Hence,
the aromatic proton signals were located in the range d ˆ 7.5 ±
8.5, while the resonance signals of the methylene protons of
the b-alanine were identified in the range d ˆ 2.4 ± 3.6. Three
groups of N H signals were noted. A set of downfield signals
Scheme 2. i) DCC, HOBt, DMF, 1,2-diaminoethane 6; ii) HCl, EtOH,
H2O; iii) DCC, HOBt, DIPEA, DMF, DMSO, 15.
15 was prepared in two steps
from the Boc-protected G1-
(carboxylic acid) 7. First, two
equivalents of the dendron 7
were treated with 1,2-diamino-
ethane 6 in the presence of
DCC and HOBt to yield the
G1-dendrimer 1 as a solid in
80% yield. Secondly, the four
Boc-protecting groups were re-
moved by treatment of 1 with
HCl (1.2m) in aqueous ethanol
to produce the tetraammonium
salt 15 (80%). The use of this
'hypercore' had the advantage
of bypassing the need to syn-
Figure 1. 1H NMR spectra of the various Gn-ester dendrons 11, 13, 14.
thesize the G4-(carboxylic acid)
Chem. Eur. J. 2001, 7, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001 0947-6539/01/0703-0689 $ 17.50+.50/0
689

H.-F. Chow et al.
FULL PAPER
dendrimer generation. The rela-
tive integral of this small peak
was found to be approximately
one-tenth of that of the carba-
mate signal. A similar finding was
also reported by Schlüter et al. in
his study on Boc-terminated den-
dritic building blocks, in which
this small signal was proposed as
being due to the intermolecular
H-bonding dendrimer aggre-
gates.[17] This rationale was sup-
ported by the immediate disap-
pearance of the minor signal
upon addition of H-bonding sol-
vents such as methanol, which
effectively broke-up the intermo-
Figure 2. 1H NMR spectra of the various Gn-dendrimers 1 ± 4.
lecular H-bond interaction.
In the case of the poly(b-alanine) dendritic molecules, the
1H NMR spectra in [D6]DMSO were found to consist of sharp
signals. This suggested that there was little or no aggregate
formation, at least in DMSO. Although a dynamic laser light
scattering (DLLS) study (see later) on the G2-dendrimer 2
did reveal the presence of aggregates in DMSO, they were
found to be present at much lower concentration (<2%) than
would account for the observed 1H NMR signal intensity
(ꢀ10%). On the basis of literature data and the results of our
own experiments, we conclude that the dominant Boc-
carbamate N H resonance signal at d ˆ 6.87 is due to the
anti rotamer, while the smaller upfield signal is caused by the
corresponding syn rotamer.
at d ꢁ 10 was attributed to the amide protons adjacent to the
aromatic ring, while a second set of signals (d ˆ 8.4 ± 8.6) could
be assigned to the amide protons of the b-alanine unit. On the
other hand, the signals of the carbamate N H appeared at
around d ˆ 6.87. Finally, the tert-butyl proton signal of the Boc
group appeared as a sharp singlet, invariably at d ˆ 1.37.
1
Further examination of the H NMR data revealed addi-
tional information regarding the gradual change of the
dendritic microenvironment with increasing dendrimer gen-
eration. First, functional groups of the same nature located on
the surface sector of the various generations were found to
have almost identical chemical shift values (Table 1). Thus,
the resonance signals of the two different aromatic protons
located on the surface sector of the various dendrimers and
dendrons appear at d ꢁ 7.68 and 8.06. Likewise, the surface
anilide N H protons of all the compounds resonate at d ˆ
10.08. Secondly, 1H nuclei exhibit a gradual shift of d value on
moving towards the interior core of the dendrimer. In general,
1H nuclei located in the middle layer of the dendrimer are
shifted downfield by about 0.03 ± 0.07 ppm when compared to
1H nuclei of the same nature located on the surface sector.
This downfield shift is more pronounced for nuclei situated in
the inner core (ꢀ0.07 ± 0.15 ppm). Similar shifts of NMR
signals have also been reported by Meijer et al. and Stoddart
The characteristics of syn/anti rotamers of carbamates in
solution has been thoroughly studied by Nudelman et al.[18]
Due to the poor solubility of the dendritic poly(b-alanine)
molecules in chloroform, we were unable to monitor their
1H NMR characteristics in CDCl3. Hence, an aliphatic
carbamate, Boc-heptylamine 16, was prepared and used as a
model compound to examine its NMR spectroscopic behavior
in solvent systems (ꢀ3% w/v solution) in which the poly(b-
alanine) dendrimers are soluble. The 1H NMR spectral
characteristics of compound 16 were found to be similar to
1
the spectra obtained by Nudelman et al. Thus, its H NMR
1
et al.[15] in their H NMR studies and by Seebach et al. in 19F
spectrum in CDCl3 at room temperature exhibited one N H
proton signal at d ˆ 4.54. Meanwhile, addition of the donor
solvent [D6]DMSO to the CDCl3 solution resulted in a
downfield shift and a splitting of the N H signal. Hence, in
CDCl3/[D6]DMSO (3:1, v/v) solution, two N H proton signals
could be identified at d ˆ 5.54 and 5.84. In pure [D6]DMSO
NMR studies.[16]
Syn and anti rotamers: Upon close scrutiny of Figures 1 and 2,
a small signal (labeled *) appears at about 0.35 ppm upfield
from the carbamate N H signal (d ˆ 6.87), irrespective of the
1
Table 1. Selected H NMR chemical shifts for the Gn-dendrimers 2 ± 4, Gn-(carboxylic acid)s 8, 9, and Gn-esters 13, 14.
Gn
ArNH
middle
ArCONH
middle
ArH
middle
BocNH
surface
tBu
surface
surface
inner
surface
inner
surface
inner
2
3
4
8
9
G2
G3
G4
G2
G3
G2
G3
10.08
10.09
10.09
10.08
10.09
10.09
10.08
±
10.15
10.16
10.16
10.18
10.18
10.23
10.23
8.46
8.47
8.46
8.46
8.47
8.48
8.46
±
8.54
8.57
8.57, 8.61
±
8.48
±
7.68, 8.06
7.69, 8.06
7.69, 8.07
7.67, 8.07
7.69, 8.07
7.68, 8.06
7.68, 8.06
±
7.73, 8.08
7.74, 8.10
7.72, 8.10
7.94, 8.18
7.95, 8.16
7.97, 8.22
7.97, 8.21
6.86
6.86
6.86
6.86
6.86
6.87
6.86
1.37
1.37
1.37
1.37
1.37
1.37
1.37
10.16
10.16
±
10.15
±
8.51
8.51
±
±
±
7.71, 8.10
7.72, 8.10
±
7.71, 8.10
±
13
14
10.15
±
8.50
7.71, 8.08
690
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0690 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 3

Dendritic b-Peptides
686±699
solution, the two N H signals were further shifted downfield
to d ˆ 6.40 and 6.75 with a relative integrals of 1:11. The
smaller hump was therefore assigned accordingly to the N H
signal of the syn rotamer, whilst the larger peak was attributed
to the corresponding signal of the anti rotamer. Furthermore,
addition of acetic acid to the above CDCl3/[D6]DMSO (3:1,
v/v) solution resulted in a large downfield shift of the syn-
rotamer N H protons (Figure 3), which is consistent with the
formation of stable intermolecular dimers between the syn
rotamer of compound 16 and acetic acid.[18]
For the series of poly(b-alanine) dendrimers, we chose the
Boc-protected G1-ester 11 as a representative in our study
due to its lower tendency to form aggregates in solution; the
observed NMR behavior could thus be assumed not to be the
consequence of intermolecular aggregation. In either CDCl3/
[D6]DMSO (3:1, v/v) or [D6]DMSO, two carbamate N H
signals were observed for the G1-ester 11. In both cases, the
predominant N H signal due to
Figure 3. Plot of the chemical shift values (d) of the syn-carbamate N H
protons of compounds 11 and 16 against the % concentration of acetic acid
in CDCl3/[D6]DMSO (3:1, v/v).
the anti rotamer appeared
about 0.35 ppm further down-
field than that of the syn ro-
tamer. Titration of a [D6]DMSO
solution of the G1-ester 11 with
acetic acid resulted in a gradual
but noticeable downfield shift
of the syn-carbamate N H sig-
nal. Thus, in all cases the
1H NMR spectroscopic patterns
of the G1-ester 11 were found
to be similar to those of the
model compound 16. Hence, we
conclude that the small peak
adjacent to the N H signal is in
Figure 4. 13C NMR spectra of the various Gn-ester dendrons 11, 13, 14.
fact due to the minor syn ro-
tamer, and not due to aggre-
gates.
13C NMR spectroscopy: The
13C NMR resonance signals of
the various functional entities
of the poly(b-alanine) dendri-
mers were located in distinct
spectral regions. Stacked plots
of 13C NMR spectra for the
poly(b-alanine) dendrons and
dendrimers are shown in Fig-
ures 4 and 5. Two types of
amide carbonyl could be differ-
entiated; one originates from
the b-alanine moiety and the
other is due to the benzamide
brancher. The former gave rise
to 13C signals in the range d ˆ
169 ± 171, the latter at d ˆ 167.
Figure 5. 13C NMR spectra of the various Gn-dendrimers 1 ± 4.
in the range d ˆ 110 ± 142. In the aliphatic region, the tertiary
and primary carbons of the prominent tert-butyl group
resonate at d ˆ 77 and 28, respectively. The two methylene
carbons of the b-alanine branch were observed in the range
d ˆ 33 ± 38. For the G1- to G4-dendrimers 1 ± 4, the 13C signal
By expanding the spectrum in this region, the b-alanine amide
moieties situated at the different locations (i.e., inner, middle,
or surface) in the dendrimer could be revealed. On the other
hand, the 13C signal of the Boc-carbamate carbonyl was seen
at d ˆ 155. The 13C signals of the aromatic carbons were found
Chem. Eur. J. 2001, 7, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0691 $ 17.50+.50/0
691

H.-F. Chow et al.
FULL PAPER
of the core ethylene unit was found to be obscured by the 13
C
signals of [D6]DMSO. However, its presence in the G1- and
G2-dendrimers 1 and 2 could be revealed by heteronuclear
magnetic correlation resonance (HMQC) spectroscopy.
Mass spectrometry: The development of modern ionization
techniques has allowed polymers of high molecular weight
(MW) to be conveniently characterized by mass spectrometry.
For the poly(b-alanine) dendrimers, the fast-atom bombard-
ment (FAB) technique has proved to be more useful for
species with MW <1000 amu, while secondary ion mass
spectrometry (L-SIMS) method has been found to provide
more satisfactory results for those with MW >3000 amu
Unfortunately, both methods failed to detect the molecular
ions of poly(b-alanine) dendrimers with MW >4000 amu. For
dendrimers for which a mass spectrum could satisfactorily be
Figure 7. GPC chromatogram (detector: LR) of the Gn -dendrimers 1 ± 4.
(Table 2). The deviation from the theoretical value was found
to be less than 7% for the G2- to G4-dendrimers 2 ± 4, which
is within the acceptable error range for the GPC/MALLS
technique. The larger deviation in the Mw value for the
smallest G1-dendrimer 1 probably stemmed from the rela-
tively weak scattering signal of this small-size molecule.
‡
‡
obtained, molecular ions in the form of [M] , [M‡H] ,
‡
‡
[M‡Na] , or [M‡Ag] could always be detected. Further-
more, a characteristic fragmentation pattern involving the
successive loss of Boc groups (100 a.m.u.) from the molecular
ion could be noted. This was seen most clearly in the L-SIMS
mass spectrum of the G2-dendrimer 2 (Figure 6), in which the
‡
abundance of the molecular ion [M‡Na] was relatively low
Table 2. Weight-average molar mass Mw, polydispersity, and hydrodynam-
ic radii of dendrimers 1 ± 4.
and was followed by a number of ion fragments with m/z
‡
corresponding to [M‡H (100  n)] , in which n is the
[a]
Calcd Mw
Measured Mw
Polydispersity[a]
Hydrodynamic
radius [nm][b]
number of Boc groups lost.
1
2
3
4
1013
2519
5530
1.3 Â 103
2.8 Â 103
5.6 Â 103
1.1 Â 104
1.03
1.02
1.01
1.01
±
1.6[c]
2.1[d]
2.8[e]
11553
[a] Determined by GPC with a MALLS detector in DMF as the eluent.
[b] Determined by DLLS in DMF solution. [c] Concn ˆ 4.4mm, a small
aggregate peak at around 16 nm was noted. [d] Concn ˆ 1.9mm, a small,
broad aggregate peak at around 60 nm was noted. [e] Concn ˆ 0.9mm, no
higher aggregate was detected.
Properties: Some natural proteins and peptides are able to
fold spontaneously from a denatured state to a unique
conformation. The driving force behind this folding process
involves the participation of various noncovalent interactions.
These interactions, in turn, originate from the various
structural units residing on the linear amino acid sequence
of the native peptide. Among these interactions, hydrogen
bonding plays a substantial role in dictating the secondary
structure of peptides.
Figure 6. The L-SIMS mass spectrum of the G2-dendrimer 2 showing
successive loss of Boc groups.
Gel-permeation chromatography and weight-average molar
mass determination: Although all the synthesized dendrimers
were isolated and purified by flash chromatography, it was still
necessary to assess their purities before studying their
physicochemical properties. One of the most convenient
methods was to subject the dendrimers to gel-permeation
chromatographic analysis (GPC) (solvent: DMF) with laser
refractometer (LR)/multi-angle laser light scattering
(MALLS) detectors. Both the LR (Figure 7) and MALLS
chromatograms of all the dendrimers exhibited one single
sigmoidal peak. No defective products or unreacted starting
materials could be detected. Furthermore, no peak of lower
retention time than the main peak was seen in any of the
chromatograms, suggesting the absence of aggregates in DMF.
Due to the lack of mass spectral data for the G3- and G4-
dendrimers 3 and 4, the weight-average molar masses (Mw) of
the target dendrimers 1 ± 4 were determined by MALLS
ˆ
Due to the presence of a large number of N H and C O
functionalities in our poly(b-alanine) dendrimers, we were
intrigued to know whether they could also form intra- and
intermolecular H-bonding networks, thereby possibly giving
rise to some higher-order secondary structures. In this section,
we wish to describe our efforts aimed at probing the presence/
absence of inter/intramolecular hydrogen bonding in these
dendrimers; we considered this to be of great importance in
the elucidation of their solution structures and exploration of
their physical and chemical properties.
Solution structure
Proton ± deuterium exchange (H/D) experiment: In peptides,
hydrogen atoms attached to heteroatoms are sufficiently
692
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0692 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 3

Dendritic b-Peptides
686±699
labile to undergo exchange with deuterium provided by
deuterated solvents. Lenormant and Blout noted that the rate
of exchange was dependent on the local environment of the
N H group.[19] Generally speaking, perseverant or H-bonded
N H protons were found to exchange with protic solvents at a
much slower rate than N H protons exposed to the solvent.
To assess the effect of dendrimerization on the N H H/D
exchange behavior, four model compounds were studiedÐthe
previously mentioned carbamate 16, an aliphatic benzamide
17, an anilide 18, and a G1-benzamide dendron 19. The
indicates that the H/D exchange rates were slightly enhanced
in this partially dendrimer-like structure. For the more
complex G1- to G4-dendrimers 1 ± 4, the t1/2 values of the
N H protons were found to be further reduced, regardless of
their location in the dendrimer (i.e. interior or surface). This
indicated that the H/D exchange of the N H protons was
further enhanced in a poly(b-alanine) dendritic framework.
For example, the t1/2 value for the carbamate N H in the
dendrimers decreased to 300 min, while those of the anilide
and benzamide N H protons were significantly reduced to
<20 min. The fast H/D exchange rates of the N H protons,
including those near the interior core of the dendrimer, also
suggested that these poly(b-alanine) dendrimers have an open
structure with the interior functionalities being fully exposed
to the solvent medium.
The enhancement of the H/D exchange rates of the various
N H protons in a dendritic environment, especially those
inside the dendrimer core, was unexpected. We attribute this
finding to the highly polar internal environment of the
dendrimers. Thus, the presence of a large number of
carbamate and amide functionalities in a confined space
serves as an H-bonding sponge for polar protic molecules such
as water. As a result of the strong H-bonding interactions
between water molecules and the dendrimers, the local
concentration of water in the vicinity and, in particular, in
the interior of the dendrimer could be much higher than that
in the bulk solvent, thereby accounting for the enhancement
in the H/D exchange rates as compared to those of non-
dendritic N H amides and carbamates. This strong water-
binding property is also consistent with the elemental analysis
data for these poly(b-alanine) dendrimers. It was found that
although the analytical data were reproducible for different
batches of the same compound, the carbon contents were
consistently lower than the expected values due to the
inclusion of water. Typically, one to seven water molecules
per dendrimer were included, with the higher generation
compounds incorporating the largest numbers. The inclusion
of water molecules in polar polyamide dendrimers has also
been reported by Rosowsky, Voit, and Schlüter.[21]
exchange half-lives (t1/2) of the various N H protons of the
model compounds as well as those of the G1- to G4-
dendrimers 1 ± 4 are given in Table 3. The data were computed
by plotting the relative integrals as a function of time,
assuming that the H/D exchange process obeyed a pseudo
first-order rate law.[20]
For the model compounds, three different types of N H
protons (i.e. the Boc-carbamate, the anilide, and the benz-
amide N H) could be identified. Examination of the H/D
exchange behavior of the nondendritic models 16 ± 18 showed
that the t1/2 values of the various N H protons decreased in
the following order: Boc-carbamate N H (ꢀ1000 min) >
benzamide N H (ꢀ500 min) > anilide N H (ꢀ10 min). The
rate of exchange of the anilide N H was comparatively fast,
possibly due to the electronic stabilization effect of the
adjacent aromatic ring. The NMR spectra of these compounds
consisted of sharp signals, an indication that there was no
aggregate formation at the sample concentration range
(<5%, w/v) in [D6]DMSO. Therefore, these t1/2 values should
reflect the ªintrinsicº exchange properties of the respective
N H protons in the absence of external factors such as intra-
and intermolecular H-bonding effects.
Temperature coefficients of N H: The change in the chemical
shift value (d) of N H protons as a function of temperature
(i.e., the temperature coefficient jDd/DTj ) has been em-
ployed as a criterion to differentiate intra- from intermolec-
ular hydrogen bonding. Generally, a jDd/DTj value of less
1
than 2.0 Â 10 3 ppmK in [D6]DMSO can be taken to
For the G1-benzamide dendron 19, a hybrid structure that
served to bridge the structural features of the simple
compounds 16 ± 18 and the more complex G1- to G4-
dendrimers 1 ± 4, the t1/2 values of all the N H protons, apart
from that of the anilide, were found to be slightly lower than
the corresponding figures for the model compounds; this
indicate the presence of intramolecular hydrogen bonding.
When this coefficient is greater than 4.0 Â 10 3 ppmK 1, the
existence of intramolecular hydrogen bonds can be exclud-
ed.[22] The temperature coefficients of the various N H
protons of the G2-dendrimer 2 in DMSO were therefore
determined. Their values were all found to be greater than
Table 3. Exchange half-lives of N H protons of the Gn-dendrimers 1 ± 4, aliphatic carbamate 16, benzamide 17, anilide 18, and G1-dendron 19.
1
2
3
4
16
17
18
19
t1/2 of carbamate N H (min)
t1/2 of anilide N H (min)
t1/2 of benzamide N H (min)
300
< 10
20
300
< 10
20
300
< 10
20
200
< 10
10
1000
±
±
500
±
< 10
400
80
400
±
±
±
Chem. Eur. J. 2001, 7, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0693 $ 17.50+.50/0
693

H.-F. Chow et al.
FULL PAPER
6.0 Â 10 3 ppmK 1 (Figure 8), indicating the absence of intra-
molecular hydrogen bonding in this solvent and, thus, ruling
out the existence of a higher-order, thermodynamically stable
secondary structure for these poly(b-alanine) dendrimers in
such highly polar solvents.
low intensity was seen in the hydrodynamic radius distribution
diagram for the poly(b-alanine) dendrimers. Thus, the G2-
dendrimer 2 in pure DMF solution (4.4mm) was shown to
consist of two narrowly distributed populations at 258C
(Figure 9). The first peak, located at 1 ± 2 nm, corresponded to
Figure 9. Temperature dependence of the hydrodynamic radius distribu-
tion f(Rh) of G2-dendrimer 2 (4.4mm) in DMF (scattering angle ˆ 158).
Figure 8. Temperature coefficients (ppmK 1) of G2-dendrimer 2.
Based on the results of the NMR experiments, the poly(b-
alanine) dendrimers are highly solvated in polar solvents and
have an open structure that allows the free diffusion of solvent
molecules into the interior. The cooperative effect of a large
the expected monomeric species. The second peak, located at
10 ± 20 nm, was due to some higher molar mass intermolecular
aggregates, which disappeared at slightly elevated temper-
ature (358C). Judging from the relative areas of the peaks, the
amount of aggregate was very small. The scarcity of the
aggregates at 258C also accounts for the fact that they were
not detectable by the less sensitive GPC on-line MALLS/RI
technique. Likewise, an aggregation peak with an even
broader distribution centered at around 60 nm was seen in
the distribution diagram for the G3-dendrimer 3. However, no
aggregation peak was found for the G4-dendrimer 4, possibly
due to a higher polydispersity and a lower population of the
aggregates.
ˆ
number of highly polar N H and C O functionalities within
the dendrimer also facilitates the formation of intermolecular
hydrogen bonds with polar protic solvents such as water,
resulting in an enhancement of the H/D exchange rates.
Aggregation behavior
Dynamic laser light scattering: Recently, several reports
dealing with the aggregation and self-association of dendri-
mers have appeared.[23] The large size and controllable
functionalities of dendritic molecules make them ideal build-
ing blocks for assembling into large nano- and mesoscopic
structures in solutions. On the basis of the aforementioned
NMR and GPC studies, both the poly(b-alanine) dendrons
and dendrimers appear to be monomeric species in polar
aprotic solvents, such as DMSO and DMF. However, such
polar dendrimers may be able to form aggregates in less polar
solvents. Besides solvent effects, the formation of intermo-
lecular aggregates is also a temperature-dependent process.
Here, we wish to report on a study of the aggregation behavior
of the poly(b-alanine) dendrimers in various solvent systems
and at different temperatures by the more sensitive, dynamic
laser light scattering (DLLS) method. This technique has also
enabled us to determine the average hydrodynamic radii
(hRhi) of the dendritic species.
Solvent dependence: The aggregation behavior of the poly(b-
alanine) dendrimers was also investigated in different solvent
systems. Upon addition of a small amount of protic solvent
such as methanol (7% v/v in DMF), the aggregate peak of the
G2-dendrimer 2 gradually disappeared (Figure 10). There
The experimental hRhi values of the G2- to G4-dendrimers
2 ± 4 in DMF solution are presented in Table 2. Because of the
small size of the G1-dendrimer 1, its hRhi value could not be
determined with accuracy. Each dendrimer exhibited a major
peak with an hRhi value slightly larger than that of the
PAMAM dendrimer of the same generation.[24] Due to the
higher sensitivity of the laser beam used in the DLLS study, in
addition to the monomeric peak, an aggregate peak of very
Figure 10. Time dependence of the hydrodynamic radius distribution f(Rh)
after addition of methanol to a DMF solution of the G2-dendrimer 2
(4.4mm) at 258C (scattering angle ˆ 158).
694
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0694 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 3

Dendritic b-Peptides
686±699
fore,H-bond donor solvents effectively break-up the inter-
molecular H-bonding network and stabilize the monomeric
state. On the other hand, the DLLS distribution diagram
changed dramatically when a nonpolar solvent such as
chloroform was added to the DMF solution of the G2-
dendrimer 2 (Figure 11). In a 33% (v/v) CHCl3/DMF
well as on the dendrimer generation. The aggregates broke up
at elevated temperatures or in the presence of H-bond donor
solvents. Work is underway aimed at studying the properties
of the Boc-deprotected native poly(b-alanine) dendrimers.
Experimental Section
Melting points were measured on an Electrothermal 9100 melting point
apparatus and are uncorrected. 1H (300 MHz) and 13C NMR (75.5 MHz)
spectra were acquired on a Bruker Advance DPX spectrometer and were
recorded in [D6]DMSO unless otherwise stated. Mass spectrometry was
performed either by fast atom bombardment (FAB) on
a Hewlett
Packard 5989B mass spectrometer or by the liquid secondary ionization
mass spectrometry (L-SIMS) method on a Bruker APEX47EFTMS model
spectrometer. The reported molecular mass (m/z) values refer to mono-
isotopic masses. Elemental analyses were carried out at MEDAC (UK).
Unless otherwise stated, all chemicals were purchased from either Acros or
Aldrich and were used without further purification. DMF and DMSO were
stirred over calcium hydride overnight and then distilled in vacuo. THF was
distilled from sodium and benzophenone. Thin-layer chromatography was
performed on Merck silica gel 60F254 plates and spots were visualized by
UV illumination. Flash chromatography was carried out on Macherey
Nagel silica gel (230 ± 400 mesh) or active Merck aluminium oxide 90,
neutral grade (70 ± 200 mesh).
Figure 11. Solvent dependence of the hydrodynamic radius distribution
f(Rh) after addition of chloroform to a DMF solution of the G2-dendrimer
2 (2.8mm) at 258C (scattering angle ˆ 158).
Ethyl 3,5-diaminobenzoate (10): A suspension of 3,5-diaminobenzoic acid
5 (7.0 g, 46 mmol) in ethanol (300 mL) and 98% sulfuric acid (14 mL) was
heated to reflux for 18 h. The organic solvent was then removed on a rotary
evaporator and the residual syrup was diluted with iced water (200 mL) and
made slightly alkaline by the addition of powdered Na2CO3. The resulting
mixture was extracted with EtOAc (3 Â 150 mL) and the combined organic
phases were washed with water (2 Â 300 mL), dried (Na2SO4), and
concentrated to dryness in vacuo. The residue was purified by flash
chromatography on silica gel (eluent: EtOAc/hexane 1:1) to give the ester
solution, the aggregate peak originally situated at 16 nm
disappeared, while another aggregate peak with a larger hRhi
value (ꢀ56 nm) emerged when the amount of CHCl3 was
increased to 45% (v/v). In other words, the poly(b-alanine)
dendrimers self-assembled into larger aggregates in a non-
polar aprotic solvent. This phenomenon could also be
1
3
10 (7.5 g, 90%) as an amber liquid. H NMR: d ˆ 1.27 (t, J(H,H) ˆ 7 Hz,
3H; CH3), 4.21 (q, 3J(H,H) ˆ 7 Hz, 2H; CH2), 5.00 (s, 4H; NH2), 6.03 (t,
4J(H,H) ˆ 2 Hz, 1H; ArH), 6.44 (d, 4J(H,H) ˆ 2 Hz, 2H; ArH); 13C NMR:
d ˆ 14.5, 60.2, 103.8, 131.1, 149.6, 167.0; MS (FAB): m/z (%): 181 (100)
1
confirmed by a H NMR study. Solvent titration of solutions
of G2- to G4-dendrimers 2 ± 4 (ꢀ2 to 8mm range) in
‡
[M ‡ H] ; elemental analysis calcd (%) for C9H12N2O2 (180.2): C 59.99, H
[D6]DMSO or [D7]DMF with CDCl3 resulted in significant
6.71, N 15.55; found C 59.82, H 6.66, N 15.76.
1
broadening of H NMR signals, indicating the formation of
Boc-protected G1-ester dendron (11): A mixture of Boc-b-alanine (10.5 g,
56 mmol) and EEDQ (14.0 g, 56 mmol) was added to a solution of the ester
10 (5.0 g, 28 mmol) in THF (200 mL) at 208C. After 24 h, the solvent was
removed on a rotary evaporator and the residue was taken up in ethyl
acetate (300 mL). The product 11 slowly precipitated from the solution and
was subsequently collected by filtration and washed with EtOAc/diethyl
ether (1:1) to give a white solid (11.7 g, 80%). M.p. 189 ± 1908C; 1H NMR:
d ˆ 1.32 (t, 3J(H,H) ˆ 7 Hz, 3H; CH3), 1.38 (s, 18H; tBu), 2.48 (t, 3J(H,H) ˆ
7 Hz, 4H; CH2CO), 3.22 (q, 3J(H,H) ˆ 7 Hz, 4H; NCH2), 4.31 (q,
3J(H,H) ˆ 7 Hz, 2H; OCH2), 6.89 (t, 3J(H,H) ˆ 7 Hz, 2H; carbamate-
NH), 7.95 (s, 2H; ArH), 8.22 (s, 1H; ArH), 10.17 (s, 2H; NHAr); 13C NMR:
d ˆ 14.4, 28.4, 36.6, 36.9, 61.0, 77.8, 114.1, 114.6, 130.6, 139.9, 155.7, 165.7,
intermolecular aggregates in nonpolar solvents.
Conclusion
We have developed a convenient method for the synthesis of a
series of highly polar poly(b-alanine) dendrimers from the
first to the fourth generation based on a solution-phase
coupling procedure. This synthetic operation has proved to be
more efficient than the solid-phase method in terms of
‡
169.9; MS (FAB): m/z (%): 522 (25) [M] ; elemental analysis calcd (%) for
C25H38N4O8 (522.6): C 57.46, H 7.33, N 10.72; found C 57.30, H 7.26, N 10.73.
1
product yield and homogeneity. Based on H NMR studies,
General procedure I-±hydrolysis of ethyl esters: An aqueous solution of
NaOH (1M, 20 mL, 20 mmol) was added to a solution of the Boc-protected
Gn-CO2Et (4 mmol) in MeOH/THF (1:1, 100 mL) at 208C. The progress of
the reaction was monitored by TLC. When the hydrolysis was complete
(12 ± 36 h), the excess solvents were removed on a rotary evaporator, and
the residue was taken up in a mixture of 10% aqueous citric acid (200 mL)
and EtOAc (200 mL) with vigorous stirring. The organic phase was
separated and then washed with water (2 Â 200 mL). The Boc-protected
carboxylic acid dendron Gn-CO2H subsequently precipitated from the
organic phase and was collected by filtration. Purification by flash
chromatography (for details, see below) was required for the G2- and
G3-CO2H dendrons 8 and 9.
these poly(b-alanine)-based dendritic species have been
shown to have an open structure in polar aprotic solvents,
which allows the free interaction of the interior functionalities
with solvent molecules. Due to the presence of a large number
of carbamate and amide groups, the dendrimers act as
H-bonding sponges, which strongly bind protic solvents such
as water. As a consequence, the H/D exchange rates of the
N H protons are significantly enhanced in such dendritic
structures as compared with those of nondendritic carbamates
and amides. These peptide dendrimers also form intermolec-
ular aggregates of nanoscopic size in both nonpolar and polar
aprotic solvents. Furthermore, the size of the aggregates has
been shown to be dependent on the nature of the solvent as
Boc-protected G1-(carboxylic acid) dendron (7): General procedure I:
Starting from Boc-protected G1-ester 11 (10.0 g, 20 mmol), after 12 h
under hydrolytic conditions the target compound 7 was isolated as a white
solid (9.0 g, 90%) after filtration. M.p. 199 ± 2008C; 1H NMR (COOH not
Chem. Eur. J. 2001, 7, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0695 $ 17.50+.50/0
695

H.-F. Chow et al.
FULL PAPER
observed): d ˆ 1.38 (s, 18H; tBu), 2.48 (t, 3J(H,H) ˆ 7 Hz, 4H; CH2CO),
3.22 (q, 3J(H,H) ˆ 7 Hz, 4H; NCH2), 6.88 (t, 3J(H,H) ˆ 7 Hz, 2H;
carbamate-NH), 7.93 (s, 2H; ArH), 8.17 (s, 1H; ArH), 10.13 (s, 2H;
NHAr); 13C NMR: d ˆ 28.4, 36.7, 37.0, 77.8, 113.9, 115.0, 131.6, 139.8, 155.7,
8 mmol), after 12 h of reaction the diammonium salt 12 (2.5 g, 80%) was
isolated as a white amorphous solid. 1H NMR: d ˆ 1.31 (t, 3J(H,H) ˆ 7 Hz,
3
3
3H; CH3), 2.79 (t, J(H,H) ˆ 7 Hz, 4H; CH2CO), 3.07 (t, J(H,H) ˆ 7 Hz,
3
4H; NCH2), 4.31 (q, J(H,H) ˆ 7 Hz, 2H; CH2), 8.02 (s, 2H; ArH), 8.05 ±
‡
‡
167.3, 169.9; MS (FAB): m/z (%): 517 (17) [M ‡ Na] ; HRMS (L-SIMS):
8.20 (brs, 6H; N H3), 8.20 (s, 1H; ArH), 10.60 (s, 2H; NHAr); 13C NMR:
m/z: 495.2411 (C23H35N4O8 requires 495.2446).
d ˆ 14.4, 33.5, 35.0, 61.1, 114.2, 115.0, 130.7, 139.7, 165.7, 168.9; MS (FAB):
‡
m/z: 323 (100) [M 2HCl ‡ H] ; HRMS (L-SIMS): m/z: 323.1722
Boc-protected G2-(carboxylic acid) dendron (8): General procedure I:
Starting from Boc-protected G2-ester 13 (5.0 g, 4.0 mmol), after 24 h under
hydrolytic conditions the target compound 8 (3.7 g, 74%) was obtained as a
white glassy solid after purification by flash chromatography on silica gel
(eluent: EtOAc/EtOH, 12:1, gradient to 8:1). 1H NMR (COOH not
observed): d ˆ 1.37 (s, 36H; tBu), 2.45 ± 2.51 (m, 8H; surface CH2CO), 2.62
(t, 3J(H,H) ˆ 7 Hz, 4H; inner CH2CO), 3.21 (q, 3J(H,H) ˆ 7 Hz, 8H;
surface NCH2), 3.52 (q, 3J(H,H) ˆ 7 Hz, 4H; inner NCH2), 6.86 (t,
3J(H,H) ˆ 7 Hz, 4H; carbamate-NH), 7.67 (s, 4H; surface ArH), 7.94 (s,
2H; inner ArH), 8.07 (s, 2H; surface ArH), 8.18 (s, 1H; inner ArH), 8.46 (t,
3J(H,H) ˆ 7 Hz, 2H; ArCONH), 10.08 (s, 4H; surface NHAr), 10.18 (s,
2H; inner NHAr); 13C NMR: d ˆ 28.5, 36.1, 36.4, 36.7, 36.9, 77.9, 112.8,
113.3, 113.8, 115.0, 136.0, 139.5, 139.8, 155.7, 166.9, 167.3, 169.8, 169.9; MS
(C15H23N4O4 requires 323.1714).
Tetraammonium core (15): General procedure II: Starting from the Boc-
protected G1-dendrimer 1 (2.0 g, 2.0 mmol), after 24 h of reaction the
product 15 (1.2 g, 80%) was obtained as a white hygroscopic amorphous
solid. 1H NMR: d ˆ 2.80 (t, 3J(H,H) ˆ 7 Hz, 8H; CH2CO), 3.07 (q,
3J(H,H) ˆ 7 Hz, 8H; NCH2), 3.40 (brs, 4H; core NCH2), 7.74 (s, 4H;
‡
ArH), 8.10 (s, 2H; ArH), 8.16 (brs, 12H; N H3), 8.53 (s, 2H; ArCONH),
10.52 (s, 4H; NHAr); 13C NMR: d ˆ 33.5, 35.1, 113.1, 113.8, 136.1, 139.3,
‡
167.0, 168.7; MS (FAB): m/z (%): 613.32 (100) [M 4HCl ‡ H] ; HRMS
(L-SIMS): m/z: 613.3207 (C28H41N10O6 requires 613.3202).
General procedure IIIÐsynthesis of Boc-protected ester dendrons: A
solution of the diammonium salt 12 (1.0 g, 2.5 mmol) and DIPEA (0.87 mL,
5.0 mmol) in DMF (25 mL) was added in one portion to a stirred solution
of the appropriate carboxylic acid dendron (5.0 mmol) and HOBt (0.74 g,
5.5 mmol) in DMF (150 mL) at 08C. DCC (1.13 g, 5.5 mmol) was then
added and the reaction mixture was stirred at 08C for 12 h and thereafter at
room temperature for a further 36 ± 60 h. The reaction mixture was then
filtered and the filtrate was concentrated to dryness in vacuo. The residue
was dissolved in EtOAc/EtOH (12:1, 200 mL) and the resulting solution
was washed successively with saturated NaHCO3 (2 Â 200 mL), 5%
KHSO4 (2 Â 200 mL), and water (2 Â 200 mL), dried (Na2SO4), filtered,
and concentrated on a rotary evaporator. The concentrate was further
purified by flash chromatography.
‡
(L-SIMS): m/z (%): 1269.6 (60) [M ‡ Na] ; elemental analysis calcd (%)
for C59H82N12O18 (1247.4): C 56.81, H 6.63, N 13.47; found C 56.47, H 6.79, N
13.12.
Boc-protected G3-(carboxylic acid) dendron (9): General procedure I:
Starting from Boc-protected G3-ester 14 (8.0 g, 2.9 mmol), after 36 h under
hydrolytic conditions the target product 9 (5.5 g, 70%) was obtained as a
white amorphous solid after purification by flash chromatography on silica
gel (eluent: EtOAc/EtOH, 10:1, gradient to 6:1). 1H NMR (COOH not
observed): d ˆ 1.37 (s, 72H; tBu), 2.45 ± 2.53 (m, 16H, surface CH2CO),
2.63 (t, 3J(H,H) ˆ 7 Hz, 12H; middle and inner CH2CO), 3.21 (q,
3J(H,H) ˆ 7 Hz, 16H; surface NCH2), 3.50 ± 3.55 (m, 12H; middle and
inner NCH2), 6.86 (t, 3J(H,H) ˆ 7 Hz, 8H; carbamate-NH), 7.69 (s, 8H;
surface ArH), 7.71 (s, 4H; middle ArH), 7.95 (s, 2H; inner ArH), 8.07 (s,
4H; surface ArH), 8.10 (s, 2H; middle ArH), 8.16 (s, 1H; inner ArH), 8.47
(t, 3J(H,H) ˆ 7 Hz, 4H; surface ArCONH), 8.48 (t, 3J(H,H) ˆ 7 Hz, 2H;
inner ArCONH), 10.09 (s, 8H; surface NHAr), 10.15 (s, 4H; middle
NHAr), 10.18 (s, 2H; inner NHAr); 13C NMR: d ˆ 28.4, 36.1, 36.4, 36.7,
36.9, 77.8, 112.8, 113.3, 136.0, 139.5, 155.7, 166.8, 169.7, 169.8, 169.9; MS
Boc-protected G2-ester dendron (13): General procedure III: Starting
from G1-(carboxylic acid) dendron 7 (2.5 g, 5.0 mmol), after 48 h of
reaction the crude product was purified by flash chromatography on silica
gel (eluent: EtOAc/EtOH/THF, 20:1:1, gradient to 12:1:1) to give the
product 13 (2.2 g, 70%) as a white amorphous solid. 1H NMR: d ˆ 1.31 (t,
3J(H,H) ˆ 7 Hz, 3H; CH3), 1.37 (s, 36H; tBu), 2.47 (t, 3J(H,H) ˆ 7 Hz, 8H;
surface CH2CO), 2.62 (t, 3J(H,H) ˆ 7 Hz, 4H; inner CH2CO), 3.21 (q,
3J(H,H) ˆ 7 Hz, 8H; surface NCH2), 3.52 (q, 3J(H,H) ˆ 7 Hz, 4H; inner
‡
(L-SIMS): m/z (%): 2774 (60) [M ‡ Na] ; elemental analysis calcd (%) for
C
131H178N28O38 ´ 2H2O (2789.1): C 56.41, H 6.58, N 14.06; found C 56.05, H
3
3
NCH2), 4.31 (q, J(H,H) ˆ 7 Hz, 2H; OCH2), 6.87 (t, J(H,H) ˆ 7 Hz, 4H;
carbamate-NH), 7.68 (s, 4H; surface ArH), 7.97 (s, 2H; inner ArH), 8.06 (s,
2H; surface ArH), 8.22 (s, 1H; inner ArH), 8.48 (t, 3J(H,H) ˆ 7 Hz, 2H;
ArCONH), 10.09 (s, 4H; surface NHAr), 10.23 (s, 2H; inner NHAr);
13C NMR: d ˆ 14.4, 28.4, 36.1, 36.4, 36.7, 36.9, 61.0, 77.8, 112.8, 113.3, 114.1,
114.7, 130.7, 136.0, 139.5, 139.9, 155.7, 165.7, 166.9, 169.7, 170.0; MS
6.75, N 13.88.
Boc-protected G1-dendrimer (1): DCC (1.13 g, 5.5 mmol) was added to a
stirred mixture of the G1-(carboxylic acid) dendron 7 (2.5 g, 5.0 mmol),
HOBt (0.74 g, 5.5 mmol), and 1,2-diaminoethane (0.17 mL, 2.5 mmol) in
DMF (150 mL) at 08C. The mixture was stirred at 08C for 12 h and then at
room temperature for an additional 24 h. It was then filtered and the
solvents were removed in vacuo. The residue was redissolved in EtOAc/
EtOH (12:1, 200 mL) and the resulting solution was washed successively
with 5% KHSO4 (2 Â 200 mL), water (2 Â 200 mL), and brine (2 Â
200 mL), dried (Na2SO4), filtered, and concentrated to dryness on a rotary
evaporator. The residue was chromatographed on silica gel (eluent:
EtOAc/EtOH/THF, 14:1:1) to give the target G1-dendrimer 1 (2.0 g,
80%) as a white amorphous solid. 1H NMR: d ˆ 1.38 (s, 36H; tBu), 2.47 (t,
3J(H,H) ˆ 7 Hz, 8H; CH2CO), 3.21 (q, 3J(H,H) ˆ 7 Hz, 8H; NCH2), 3.35 ±
3.43 (m, 4H; core NCH2), 6.87 (t, 3J(H,H) ˆ 7 Hz, 4H; carbamate-NH),
7.69 (s, 4H; ArH), 8.06 (s, 2H; ArH), 8.50 (t, 3J(H,H) ˆ 7 Hz, 2H;
ArCONH), 10.09 (s, 4H; NHAr); 13C NMR (core NC overlapped with
solvent signals): d ˆ 28.4, 36.7, 36.9, 77.8, 112.8, 113.3, 136.0, 139.5, 155.7,
‡
(L-SIMS): m/z (%): 1297.6 (50) [M ‡ Na] ; elemental analysis calcd (%)
for C61H86N12O18 ´ H2O (1293.4): C 56.65, H 6.86, N 12.99; found C 56.77, H
6.77, N 13.31.
Boc-protected G3-ester dendron (14): General procedure III: Starting
from the G2-(carboxylic acid) dendron 8 (6.2 g, 5.0 mmol), after 72 h of
reaction the crude product was purified by flash chromatography on silica
gel (eluent: EtOAc/EtOH/THF, 14:1:1, gradient to 6:1:1) to yield the final
product 14 (4.8 g, 70%) as a white amorphous solid. 1H NMR: d ˆ 1.31 (t,
3J(H,H) ˆ 7 Hz, 3H; CH3), 1.37 (s, 72H; tBu), 2.47 (t, 3J(H,H) ˆ 7 Hz, 16H;
surface CH2CO), 2.62 (brs, 12H; middle and inner CH2CO), 3.21 (q,
3J(H,H) ˆ 7 Hz, 16H; surface NCH2), 3.52 (q, 3J(H,H) ˆ 7 Hz, 12H; middle
3
3
and inner NCH2), 4.30 (q, J(H,H) ˆ 7 Hz, 2H; OCH2), 6.86 (t, J(H,H) ˆ
7 Hz, 8H; carbamate-NH), 7.68 (s, 8H; surface ArH), 7.71 (s, 4H; middle
ArH), 7.97 (s, 2H; inner ArH), 8.06 (s, 4H; surface ArH), 8.08 (s, 2H;
middle ArH), 8.21 (s, 1H; inner ArH), 8.46 (t, 3J(H,H) ˆ 7 Hz, 4H; surface
ArCONH), 8.50 (t, 3J(H,H) ˆ 7 Hz, 2H; inner ArCONH), 10.08 (s, 8H;
surface NHAr), 10.15 (s, 4H; middle NHAr), 10.23 (s, 2H; inner NHAr);
13C NMR: d ˆ 14.4, 28.4, 36.1, 36.4, 36.6, 36.7, 36.9, 37.1, 61.0, 77.9, 112.8,
113.3, 114.1, 114.7, 130.7, 136.0, 139.5, 140.0, 155.7, 165.7, 166.9, 169.7, 169.8,
‡
167.0, 169.8; MS (L-SIMS): m/z (%): 1035.5 (100) [M ‡ Na] ; elemental
analysis calcd (%) for C48H72N10O14 (1013.2): C 56.90, H 7.16, N 13.82;
found C 56.30, H 6.70, N 13.99.
General procedure IIÐremoval of the Boc group: An ethanolic solution of
HCl (1.2M, 20 mL, 24 mmol) was added to a solution of the Boc-protected
compound (4 mmol) in EtOH/H2O (2:1, 100 mL) at 408C. The progress of
the reaction was monitored by withdrawing aliquots and examining them
for complete disappearance of the tert-butyl signal by 1H NMR spectros-
copy. Upon complete cleavage of the Boc group, the solvent was removed
in vacuo and the ammonium salt was precipitated from an EtOH/H2O
mixture (8:1, v/v).
‡
170.0; MS (L-SIMS): m/z (%): 2802.2 (100) [M ‡ Na] ; elemental analysis
calcd (%) for C133H182N28O38 ´ H2O (2799.1): C 57.07, H 6.63, N 14.01; found
C 56.72, H 6.77, N 13.62.
General procedure IVÐsyntheses of Boc-protected G2- to G4-dendrimers
(2 ± 4): A solution of the tetraammonium salt 15 (1 equiv) and DIPEA
(5 equiv) in DMSO (20 mL) was added in one portion to a mixture of the
Gn-(carboxylic acid) dendron (5 equiv) and HOBt (5 equiv) in DMF
Ethyl 3,5-di(b-aminopropionamido)benzoate dihydrochloride (12): Gen-
eral procedure II: Starting from Boc-protected G1-ester dendron 11 (4.0 g,
696
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0696 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 3

Dendritic b-Peptides
686±699
(150 mL). DCC (5 equiv) was then added and the reaction mixture was
stirred at room temperature for 48 ± 72 h. The precipitate was filtered off
and the solvents were removed in vacuo to leave a syrup, which was
redissolved in EtOAc/EtOH (10:1, 200 mL). The organic phase was then
washed with 5% aqueous KHSO4 (2 Â 200 mL) and water (2 Â 200 mL),
dried (Na2SO4), filtered, and concentrated on a rotary evaporator. The
concentrate was then purified by flash chromatography.
22.2, 26.4, 28.5, 28.6, 29.7, 31.5, 77.4, 155.7; HRMS (L-SIMS): m/z: 216.1950
(C12H26NO2 requires 216.1957).
Heptyl benzamide (17): 1-Heptylamine (1.48 mL, 10 mmol) and triethyl-
amine (1.2 mL, 15 mmol) were added to a stirred solution of benzoyl
chloride (1.75 mL, 15 mmol) in dry THF (30 mL). The mixture was stirred
at 08C for 6 h and then the excess solvents were removed in vacuo. The
residue was redissolved in EtOAc (2 Â 50 mL) and washed with saturated
NaHCO3 (2 Â 50 mL), 5% aqueous KHSO4 (2 Â 200 mL), and water (2 Â
200 mL), dried (Na2SO4), filtered, and concentrated to dryness on a rotary
evaporator. The residue was chromatographed on silica gel (eluent:
EtOAc/hexane, 1:8) to give the product 17 (1.5 g, 70%) as a white solid.
M.p. 34 ± 358C; 1H NMR: d ˆ 0.85 (t, 3J(H,H) ˆ 7 Hz, 3H; CH3), 1.20 ± 1.35
(m, 8H; CH2), 1.45 ± 1.60 (m, 2H; CH2), 3.24 (q, 3J(H,H) ˆ 7 Hz, 2H;
NCH2), 7.40 ± 7.52 (m, 3H; ArH), 7.80 ± 7.88 (m, 2H; ArH), 8.44 (t,
3J(H,H) ˆ 7 Hz, 1H; NH); 13C NMR (NC overlapped with solvent signals):
d ˆ 14.1, 22.3, 26.7, 28.7, 29.3, 31.5, 127.3, 128.4, 131.1, 134.9, 166,2; HRMS
(L-SIMS): m/z: 220.1699 (C14H22NO requires 220.1696).
Boc-protected G2-dendrimer (2): General procedure IV: Starting from the
G1-(carboxylic acid) dendron 7 (2.5 g, 5 mmol), after 48 h of reaction the
target product 2 (1.8 g, 70%) was obtained as a white amorphous solid after
flash chromatography on silica gel (eluent: EtOAc/EtOH/THF, 12:1:1,
gradient to 6:1:1). 1H NMR: d ˆ 1.37 (s, 72H; tBu), 2.47 (t, 3J(H,H) ˆ 7 Hz,
16H; surface CH2CO), 2.63 (t, 3J(H,H) ˆ 7 Hz, 8H; inner CH2CO), 3.21 (q,
3J(H,H) ˆ 7 Hz, 16H; surface NCH2), 3.39 (brs, 4H; core NCH2), 3.52 (q,
3J(H,H) ˆ 7 Hz, 8H; inner NCH2), 6.86 (t, 3J(H,H) ˆ 7 Hz, 8H; carbamate-
NH), 7.68 (s, 8H; surface ArH), 7.73 (s, 4H, inner ArH), 8.06 (s, 4H; surface
ArH), 8.08 (s, 2H; inner ArH), 8.46 (t, 3J(H,H) ˆ 7 Hz, 4H; surface
ArCONH), 8.54 (t, 3J(H,H) ˆ 7 Hz, 2H; inner ArCONH), 10.08 (s, 8H;
surface NHAr), 10.15 (s, 4H; inner CONHAr); 13C NMR (core NC
overlapped with solvent signals): d ˆ 28.4, 36.1, 36.4, 36.7, 36.9, 77.8, 112.8,
Ethyl 3,5-di(acetamido)benzoate (18): A mixture of acetic anhydride
(4.0 mL) and the diamine 10 (0.4 g, 2.2 mmol) was stirred at 408C for 12 h.
The volatiles were then evaporated in vacuo and the residue was extracted
with EtOAc (2 Â 30 mL). The organic extracts were washed with saturated
NaHCO3 (2 Â 50 mL), 5% aqueous KHSO4 (2 Â 200 mL), and water (2 Â
200 mL), dried (Na2SO4), filtered, and concentrated to dryness on a rotary
evaporator. The residue was chromatographed on silica gel (eluent:
EtOAc/hexane, 3:1) to give the product 18 (0.4 g, 70%) as a white solid:
M.p. 142 ± 1438C; 1H NMR: d ˆ 1.31 (t, 3J(H,H) ˆ 7 Hz, 3H; CH3), 2.04 (s,
6H; CH3CO), 4.30 (q, 3J(H,H) ˆ 7 Hz, 2H; CH2), 7.90 (s, 2H; ArH), 8.18 (s,
1H; ArH), 10.15 (s, 2H; NH); 13C NMR: d ˆ 14.4, 24.2, 61.0, 113.8, 114.4,
113.2, 136.0, 139.5, 155.7, 166.8, 166.9, 169.7, 169.8; MS (L-SIMS): m/z (%):
‡
2540.2 (10) [M ‡ Na] ; elemental analysis calcd (%) for C120H168N26O34
´
H2O (2536.8): C 56.82, H 6.75, N 14.36; found C 56.51, H 6.66, N 14.79.
Boc-protected G3-dendrimer (3): General procedure IV: Starting from the
G2-(carboxylic acid) dendron 8 (6.2 g, 5 mmol), after 72 h of reaction the
target compound 3 (2.2 g, 40%) was obtained as a white amorphous solid
after flash chromatography on alumina (eluent: EtOAc/EtOH/H2O, 7:1:0,
gradient to 7:1:0.02) followed by precipitation from MeOH/EtOAc.
1H NMR: d ˆ 1.37 (s, 144H; tBu), 2.47 (t, 3J(H,H) ˆ 7 Hz, 32H; surface
CH2CO), 2.63 (brs, 24H; middle and inner CH2CO), 3.21 (q, 3J(H,H) ˆ
7 Hz, 32H; surface NCH2), 3.40 ± 3.60 (m, 28H; middle, inner, and core
NCH2), 6.86 (t, 3J(H,H) ˆ 7 Hz, 16H; carbamate-NH), 7.69 (s, 16H; surface
ArH), 7.71 (s, 8H; middle ArH), 7.74 (s, 4H; inner ArH), 8.06 (s, 8H;
surface ArH), 8.10 (s, 6H; middle and inner ArH), 8.47 (t, 3J(H,H) ˆ 7 Hz,
8H; surface ArCONH), 8.51 (t, 3J(H,H) ˆ 7 Hz, 4H; middle ArCONH),
8.57 (t, 3J(H,H) ˆ 7 Hz, 2H; inner ArCONH), 10.09 (s, 16H; surface
NHAr), 10.16 (s, 12H; middle and inner NHAr); 13C NMR (core NC
overlapped with solvent signals): d ˆ 28.4, 36.1, 36.4, 36.7, 36.9, 77.9, 112.8,
113.3, 136.0, 139.5, 155.7, 166.9, 169.8, 169.9; elemental analysis calcd (%)
‡
130.7, 140.1, 165.7, 168.8; MS (FAB): m/z (%): 264 (90) [M] ; HRMS
(L-SIMS): m/z: 265.1188 (C13H17N2O4 requires 265.1184).
Heptyl
3,5-di[b-(N-tert-butyloxylcarbonylamino)propionamido]benz-
amide (19): A mixture of pentachlorophenol dicyclohexylcarbodiimide[25]
(0.8 g, 0.8 mmol) and G1-(carboxylic acid) dendron 7 (0.4 g, 0.8 mmol) in
DMF (20 mL) was stirred at 208C for 12 h. The reaction mixture was then
filtered and diluted with THF (20 mL). 1-Heptylamine (0.13 mL, 0.9 mmol)
and DIPEA (0.15 mL, 0.9 mmol) were added and the resulting mixture was
stirred at 208C for a further 12 h. The solvents were then evaporated on a
rotary evaporator, and the residue was taken up in ethyl acetate (30 mL).
The organic phase was washed with saturated NaHCO3 (2 Â 50 mL), 5%
aqueous KHSO4 (2 Â 200 mL), and water (2 Â 200 mL), dried (Na2SO4),
filtered, and concentrated to dryness on a rotary evaporator. The residue
was chromatographed on silica gel (eluent: EtOAc/hexane, 3:1) to give the
product 19 (0.35 g, 75%) as a white solid. M.p. 92 ± 938C; 1H NMR: d ˆ
0.84 (t, 3J(H,H) ˆ 7 Hz, 3H; CH3), 1.20 ± 1.40 (m, 10H; CH2), 1.38 (s, 18H;
tBu), 2.47 (t, 3J(H,H) ˆ 7 Hz, 4H; CH2CO), 3.20 ± 3.35 (m, 6H; NCH2),
6.85 (t, 3J(H,H) ˆ 7 Hz, 2H; carbamate-NH), 7.65 (s, 2H; ArH), 8.05 (s,
1H; ArH), 8.32 (t, 3J(H,H) ˆ 7 Hz, 1H; ArCONH), 10.05 (s, 2H; NHAr);
13C NMR (one NC obscured by solvent signals): d ˆ 14.2, 22.3, 26.6, 28.4,
28.7, 29.3, 31.5, 36.7, 36.9, 77.8, 112.6, 113.2, 136.4, 139.4, 155.7, 166.7, 169.7;
elemental analysis calcd (%) for C30H49N5O7 (591.7): C 60.89, H 8.53, N
11.84; found C 60.73, H 8.46, N 11.69.
for C264H360N58O74 ´ 3H2O (5584.2): C 56.78, H 6.61, N 14.55; found C 56.54,
1
H 6.56, N 14.50; Mw (MALLS) ˆ 5600 Æ 200 gmol
.
Boc-protected G4-dendrimer (4): General procedure IV: Starting from the
G3-(carboxylic acid) dendron 9 (6.9 g, 2.5 mmol), after 72 h of reaction the
product 4 (2.3 g, 40%) was obtained as a white amorphous solid after flash
chromatography on alumina (eluent: EtOAc/EtOH/H2O, 6:1:0, gradient to
6:1:0.02) followed by precipitation from MeOH/EtOAc. 1H NMR: d ˆ 1.37
3
(s, 288H; tBu), 2.48 (t, J(H,H) ˆ 7 Hz, 64H; surface CH2CO), 2.52 ± 2.75
(m, 56H; middle and inner CH2CO), 3.21 (q, 3J(H,H) ˆ 7 Hz, 64H; surface
3
NCH2), 3.53 (brs, 60H; middle, inner, and core NCH2), 6.86 (t, J(H,H) ˆ
7 Hz, 32H; carbamate-NH), 7.69 (s, 32H; surface ArH), 7.72 (s, 28H;
middle and inner ArH), 8.07 (s, 16H; surface ArH), 8.10 (s, 14H; middle
and inner ArH), 8.46 (t, 3J(H,H) ˆ 7 Hz, 16H; surface ArCONH), 8.51 (t,
3J(H,H) ˆ 7 Hz, 8H; middle ArCONH), 8.57 (t, 3J(H,H) ˆ 7 Hz, 4H; inner
Proton ± deuterium exchange experiments: The experimental methodology
of the proton exchange study was an adaptation of that reported by
Krishna.[26] [D6]DMSO was treated with 3 Š molecular sieves for 2 days
and the dendrimer samples were stored at 608C under high vacuum
(0.01 mmHg) for 3 weeks to remove the residual solvents and moisture. In a
typical example, a 4% (w/v) solution of G2-dendrimer 2 in [D6]DMSO
(0.50 mL) was prepared in a pre-dried NMR tube. D2O (50 mL) was then
added by means of a microsyringe and a series of 1H NMR spectra was
recorded at regular time intervals (ꢀ10 min, acquisition time ꢀ1 min per
spectrum) until all N H signals had vanished. From the relative integrals
thus obtained, a pseudo first-order rate constant for the PDX process of a
particular proton ± deuterium exchangeable functionality was computed,
from which an exchange half-life t1/2 (time for 50% of the protons in such a
functionality to be exchanged) was evaluated.[20]
3
ArCONH), 8.61 (t, J(H,H) ˆ 7 Hz, 2H; inner ArCONH), 10.09 (s, 32H;
surface NHAr), 10.16 (s, 28H; middle and inner NHAr); 13C NMR (core
NC overlapped with solvent signals): d ˆ 28.4, 36.1, 36.4, 36.7, 36.9, 77.9,
112.8, 113.3, 136.0, 139.5, 155.7, 166.9, 169.7, 169.8; elemental analysis calcd
(%) for C552H744N122O154 ´ 7H2O (11678.8): C 56.77, H 6.54, N 14.63; found C
1
56.33, H 6.73, N 14.53; Mw (MALLS) ˆ 11000 Æ 600 gmol
.
(N-tert-Butyloxycarbonyl)heptylamine (16): A mixture of 1-heptylamine
(2.1 g, 10 mmol) and di(tert-butyl) carbonate (2.2 g, 10 mmol) in THF
(20 mL) was stirred at 208C for 12 h. The solvent was then removed in
vacuo and the residue was extracted with EtOAc (2 Â 50 mL). The organic
extracts were washed with saturated NaHCO3 (2 Â 50 mL), 5% KHSO4
(2 Â 50 mL), and water (2 Â 50 mL), dried (Na2SO4), and filtered. The
filtrate was concentrated to dryness on a rotary evaporator and the residue
was purified by column chromatography on silica gel (eluent: EtOAc/
hexane, 1:20) to give the target product 16 (1.5 g, 70%) as a colorless liquid.
1H NMR: d ˆ 0.85 (t, 3J(H,H) ˆ 7 Hz, 3H; CH3), 1.20 ± 1.24 (m, 10H; CH2),
Variable-temperature NMR experiments: A 3% (w/v) solution of the G2-
dendrimer 2 in [D6]DMSO was prepared in an NMR tube, tetramethylsi-
lane was added as an internal standard, and the tube was sealed. The VT-
NMR experiment was performed on a Varian 400 MHz NMR spectrom-
eter. Spectra were recorded at 298, 303, 313, and 323 K, as set by a
3
3
1.36 (s, 9H; tBu), 2.88 (q, J(H,H) ˆ 7 Hz, 2H; NCH2), 6.76 (t, J(H,H) ˆ
7 Hz, 1H; NH); 13C NMR (NC overlapped with solvent signals): d ˆ 14.1,
Chem. Eur. J. 2001, 7, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0697 $ 17.50+.50/0
697

H.-F. Chow et al.
FULL PAPER
temperature programmer. The temperature coefficient jDd/DTj was
obtained by measuring the slope of a graph of chemical shift (d) versus
temperature (K).
(ADLAS DPY425II, output power 400 mW at l0 ˆ 532 nm) as light source.
The effects of varying the temperature and solvent on the hydrodynamic
radius distribution function were studied for the G2-dendrimer 2 by raising
the temperature from 258C to 358C and by titrating the initial DMF sample
solution with the solvent of interest.
Determination of dendrimer purity and Mw by gel-permeation chromatog-
raphy with multi-angle laser light scattering/refractive index detectors
(GPC-MALLS/RI)
General: Polystyrene standards were purchased from Aldrich and used for
the calibration of the laser refractometer, normalization of the MALLS
instrument, and as test samples for the GPC-MALLS/RI system. For
calibration of the MALLS detector, HPLC grade toluene was used.
Distilled (over CaH2) DMF was used as the mobile phase for both GPC-
MALLS/RI and off-line dn/dc measurements. For the preparation of
standard and sample solutions, the distilled DMF was further clarified by
passage through an Anotop 10 membrane filter (purchased from What-
man) of 0.2 mm pore size. The GPC-MALLS/RI system was configured by
connecting the various detector modules with polyether ether ketone
(PEEK) tubing of 0.01 inch internal diameter to minimize the dead volume.
A pressure damper was installed just before the injector port to ensure a
smooth mobile-phase delivery. As mentioned above, DMF was employed
as the mobile phase and was delivered by means of a Waters 510 HPLC
pump. Two styragel HR-1 and HR-3 GPC columns (purchased from
Waters) were connected in series for sample separation. The static light
scattering and refractive index signals were measured with a multi-angle
DAWNꢁ DSP laser photometer (He-Ne laser, output power ˆ 5 mW at
l0 ˆ 632.8 nm, purchased from Wyatt Technology) and an LR40 refrac-
tometer (purchased from Viscotek), respectively.
Acknowledgements
This research was financially supported by the Chemistry departmental
allocation of the CUHK and the Research Grants Council of HKSAR
(CUHK4123/98P). We also thank Dr. K. H. Sze for performing the VT-
NMR experiment for us.
[1] a) R. G. Denkewalter, J. Kolc, W. J. Lukasavage, U.S. Pat. 4410688,
1981 [Chem. Abs. 1984, 100, 103907p]; b) D. N. Posnett, H. McGrath,
J. P. Tam, J. Biol. Chem. 1988, 263, 1719; c) T. M. Chapman, G. L.
Hillyer, E. J. Mahan, K. A. Shaffer, J. Am. Chem. Soc. 1994, 116,
11195; d) Y. Kim, F. Zeng, S. C. Zimmerman, Chem. Eur. J. 1999, 5,
2133.
[2] a) S. J. E. Mulders, A. J. Brouwer, R. M. J. Liskamp, Tetrahedron Lett.
Â
à Â
1997, 38, 3085; b) D. Prevote, S. Le Roy-Gourvennec, A.-M. Cami-
nade, S. Masson, J.-P. Majoral, Synthesis 1997, 1199; c) B. Kayser, J.
Altman, W. Beck, Chem. Eur. J. 1999, 5, 754.
dn/dc measurement: The dn/dc measuring system was set up by connecting
the HPLC pump directly to the LR40 refractometer and the dn/dc values of
the dendrimers were measured in off-line mode. Stock solutions (20.0 mL,
ꢀ3.4 mgmL 1) of known concentrations of the G1- to G4-dendrimers 1 ± 4
were prepared in clarified DMF and were allowed to stand overnight to
ensure complete dissolution. For each of the stock solutions, a series of
dilutions was made so as to obtain five sample solutions of different
concentrations (at approximately 0.4, 0.9, 1.4, 1.9, and 2.4 mgmL 1), each
with a volume of 5.0 mL. A 1.0 mL sample loop was used for sample
loading. At least 2.5 mL of the sample solution was injected for each
[3] D. K. Smith, F. Diederich, Chem. Eur. J. 1998, 4, 1353.
[4] H.-F. Chow, I. Y.-K. Chan, D. T. W. Chan, R. W. M. Kwok, Chem.
Eur. J. 1996, 2, 1085.
[5] a) C. C. Mak, H.-F. Chow, Macromolecules 1997, 30, 1228; b) H.-F.
Chow, C. C. Mak, J. Org. Chem. 1997, 62, 5116.
[6] a) D. Seebach, J. L. Matthews, Chem. Commun. 1997, 2015; b) U.
Koert, Angew. Chem. 1997, 109, 1922; Angew. Chem. Int. Ed. Engl.
1997, 36, 1836.
[7] a) D. A. Tomalia, H. Baker, J. Dewald, M. Hall, G. Kallos, S. Martin, J.
Roeck, J. Ryder, P. Smith, Macromolecules 1986, 19, 2466; b) D. A.
Tomalia, V. Berry, M. Hall, D. M. Hedstrand, Macromolecules 1987,
20, 1164.
[8] For a review, see: F. Zeng, S. C. Zimmerman, Chem. Rev. 1997, 97,
1681.
[9] V. Percec, W.-D. Cho, P. E. Mosier, G. Ungar, D. J. P. Yeardley, J. Am.
Chem. Soc. 1998, 120, 11061.
[10] W. T. S. Huck, F. C. J. M. van Veggel, B. L. Kropman, D. H. A. Blank,
E. G. Keim, M. M. A. Smithers, D. N. Reinhoudt, J. Am. Chem. Soc.
1995, 117, 8293.
[11] G. R. Newkome, G. R. Baker, S. Arai, M. J. Saunders, P. S. Russo, K. J.
Theriot, C. N. Moorefield, L. E. Rogers, J. E. Miller, T. R. Lieux,
M. E. Murray, B. Phillips, L. Pascal, J. Am. Chem. Soc. 1990, 112, 8458.
[12] Dendrons of similar structure have previously been synthesized in
poor yield and poor homogeneity by solid-phase synthesis. However,
no physical data are available for these compounds; see K. E. Uhrich,
analysis. The sample in the loop was delivered by a DMF mobile phase at a
1
flow rate of 0.6 mLmin
.
The temperature of the refractive index
measurement was maintained at 218C. The recorded RI signal was
converted to the refractive index by multiplying the value by the calibration
constant of the laser refractometer. A graph of refractive index values
plotted against the corresponding dendrimer concentrations was obtained.
The slope of the graph thus gave the dn/dc of the dendrimer. The measured
dn/dc values of the G1- to G4-dendrimers 1 ± 4 were 0.100, 0.121, 0.132, and
0.134 mLg 1, respectively.
M
w determination: Six sample solutions were prepared from the G1- to G4-
dendrimers 1 ± 4 (ꢀ40 mg) and two polystyrene standards (ꢀ7 mg) by
individually dissolving the samples in 2.0 mL aliquots of clarified DMF;
these solutions were allowed to stand overnight to ensure complete
dissolution. A 200 mL sample loop was used for sample loading and the flow
rate of the mobile phase was set at 1.0 mLmin 1. The temperature of the
MALLS and RI detectors was kept at 218C. The dn/dc values of the
dendrimers obtained from the off-line measurement were entered man-
ually into an ASTRA program (purchased from Wyatt Technology). Data
from the MALLS and RI detectors were processed by the ASTRA
software to generate the Mw of the injected dendrimer. The molar masses of
the G1- to G4-dendrimers 1 ± 4 were determined by injecting 1.5 ± 2.5%
(w/v) sample solutions in DMF into the GPC-MALLS/RI system. Two
polystyrene standards (Mw ˆ 3.0  103 and 3.0  104) were also analyzed
under same experimental conditions to check the reliability of the method.
The measured Mw values of the dendrimers 1 ± 4 are given in Table 2.
Â
S. Boegeman, J. M. J. Frechet, S. R. Turner, Polym. Bulletin 1991, 25,
551.
Â
[13] a) C. Hawker, J. M. J. Frechet, J. Chem. Soc. Chem. Commun. 1990,
Â
1010; b) C. Hawker, J. M. J. Frechet, J. Am. Chem. Soc. 1990, 112,
7638.
[14] B. Belleau, G. Malek, J. Am. Chem. Soc. 1968, 90, 1652.
[15] a) S. Stevelmans, J. C. M. van Hest, J. F. G. A. Jansen, D. A. F. J.
van Boxtel, E. M. M. de Brabander van den Berg, E. W. Meijer, J.
Am. Chem. Soc. 1996, 118, 7398; b) P. R. Ashton, D. W. Anderson,
C. L. Brown, A. N. Shipway, J. F. Stoddart, M. S. Tolley, Chem. Eur. J.
1998, 4, 781.
[16] G. Greiveldinger, D. Seebach, Helv. Chim. Acta 1998, 81, 1003.
[17] R. Klopsch, S. Koch, A. D. Schlüter, Eur. J. Org. Chem. 1998, 1275.
[18] D. Marcovici-Mizrahi, H. E. Gottlieb, V. Marks, A. Nudelman, J. Org.
Chem. 1996, 61, 8402.
Aggregation study by dynamic laser light scattering (DLLS): Details of the
DLLS instrumentation and theory have been reported previously.[27]
1
Sample solutions of 10.0 mgmL of the G2- to G4-dendrimers 2 ± 4 were
prepared by dissolving the samples in clarified DMF. The solutions were
allowed to stand overnight to ensure complete dissolution and then
clarified by passage through a 0.2 mm pore size filter to achieve a dust-free
state. The hydrodynamic radii of the dendrimers were measured on a
modified commercial LLS spectrometer (ALV/SP-125) equipped with a
multi-t digital time correlation unit (ALV5000) and a solid-state laser
[19] H. Lenormant, E. R. Blout, Nature 1953, 172, 770.
[20] D. Seebach, S. Abele, K. Gademann, G. Guichard, T. Hintermann, B.
Jaun, J. L. Matthews, J. V. Schreiber, L. Oberer, U. Hommel, H.
Widmer, Helv. Chim. Acta 1998, 81, 932.
698
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0698 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 3

Dendritic b-Peptides
686±699
[21] a) R. A. Forsch, A. Rosowsky, J. Org. Chem. 1984, 49, 1305; b) D.
Wolf, B. I. Voit, Tetrahedron 1997, 53, 15535; c) I. Neubert, A. D.
Schlüter, Macromolecules 1998, 31, 9372.
[22] H. Kessler, Angew. Chem. 1982, 94, 509; Angew. Chem. Int. Ed. Engl.
1982, 21, 512.
[23] a) A. W. Bosman, H. M. Janssen, E. W. Meijer, Chem. Rev. 1999, 99,
1665; b) G. R. Newkome, E. He, C. N. Moorefield, Chem. Rev. 1999,
99, 1689; c) A. J. Berresheim, M. Müller, K. Müllen, Chem. Rev. 1999,
99, 1747.
[24] D. A. Tomalia, H. Baker, J. Dewald, M. Hall, G. Kallos, S. Martin, J.
Roeck, J. Ryder, P. Smith, Polym. J. 1985, 17, 117.
[25] M. Bodanszky, A. Bodanszky, The Practice of Peptide Synthesis, 2nd
ed., Springer, Germany, 1994, p. 101.
[26] N. R. Krishna, D. H. Huang, J. D. Glickson, R. Rowan, R. Walter,
Biophys. J. 1979, 26, 345.
[27] C. Wu, Adv. Polym. Sci. 1998, 137, 103.
Received: May 09, 2000 [F2475]
Chem. Eur. J. 2001, 7, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0703-0699 $ 17.50+.50/0
699