Synthesis and transport properties of new dendritic core-shell architectures based on hyperbranched polyglycerol with biphenyl-PEG shells

Article abstract of DOI:10.1039/c1nj20466a

A new core-shell type of nano-architectures based on hyperbranched polyglycerol (hPG) has been designed by attaching a mono(methoxy)polyethylene glycol (mPEG) shell either directly or through a hydrophobic biphenyl spacer to the hPG scaffold. Alternativel

Full text of DOI:10.1039/c1nj20466a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
NJC
Cite this: New J. Chem., 2012, 36, 371–379
www.rsc.org/njc
PAPER
Synthesis and transport properties of new dendritic core–shell architectures
based on hyperbranched polyglycerol with biphenyl-PEG shellswz
a
b
a
a
¨
Indah N. Kurniasih, Hua Liang, Vicki D. Moschwitzer, Mohiuddin A. Quadir,
Micha$ Radowski,a Jurgen P. Rabeb and Rainer Haag*a
¨
Received (in Montpellier, France) 1st June 2011, Accepted 15th July 2011
DOI: 10.1039/c1nj20466a
A new core–shell type of nano-architectures based on hyperbranched polyglycerol (hPG) has been
designed by attaching a mono(methoxy)polyethylene glycol (mPEG) shell either directly or
through a hydrophobic biphenyl spacer to the hPG scaffold. Alternatively the hPG core was
decorated with hydrophobic segments specifically located around the hPG and mPEG as the shell.
The constructed structures were compared and contrasted for their ability to solubilize guest
molecules of different polarity indices to their corresponding non-solvent for possible drug
delivery applications. UV/Vis spectroscopy and Scanning Force Microscopy (SFM) techniques
have been used to characterize the host–guest complex. Highly hydrophilic nanocarriers composed
of an hPG–mPEG arrangement were found to be very efficient in transporting hydrophilic
molecules to an organic environment with almost no encapsulation of the hydrophobic guests.
Introduction of biphenyl fragments as hydrophobic spacers between hPG and mPEG, or near the
hPG core, substantially increased the hydrophobic guest encapsulation efficiency of the resulting
system. The encapsulation and transport properties were found to critically depend on the Mn of
hPG, degree of functionalization with hydrophilic and/or hydrophobic fragments and length of
mPEG chains, either alone or in combination with each other. SFM images revealed that the size
of the nanocarriers is within the range of 10 nm as single particles and 50 nm as aggregates, with
the sizes substantially increased upon interaction with the guest species.
monomer/initiator ratio to obtain narrow polydispersities
Introduction
(typically o 2.0). It is now possible to develop hPG architectures
within the size range of 1 nm to several hundred nanometres.
The degree of branching in hPG is only 60% compared to
fully branched perfect glycerol dendrimers.19,20 So far as
biocompatibility issues are concerned, similar, or even
improved, profiles have been observed with hPG compared
to existing aliphatic polyether polyols, e.g., polysaccharides or
polyethylene glycols.21–23 Dendritic architectures based on hPG are
therefore well-suited for the generation of spherical amphiphilic
macromolecules for applications in drug solubilization and
delivery.24
The use of dendritic macromolecules as drug delivery1–11
vehicles is a field of substantial scientific research. However,
due to the synthetic complexity to achieve bulk quantities of
high generations of dendrimers at low prices,12 hyperbranched
polymers emerged as an alternative to many novel applications
of dendritic scaffold.8,13–17
Hyperbranched polyglycerol (hPG) is a highly branched
macromolecule that has been prepared from anionic ring-
opening polymerization of glycidol (Fig. 1).18 Extensive
research on hPG architectures has been carried out in the last
decade by several groups around the globe including ours,
revealing multiple unique characteristics of hPG. Initially
hPGs with molecular weights of 1–20 kDa were prepared with
a degree of polymerization (DP) that can be tailored by the
Post-polymerization modification is a powerful gateway to a
diversified macromolecular architecture. Incorporation of
hydrophilic and/or hydrophobic segments randomly or to a
selective location within the preformed hPG scaffold renders
the parent molecule with novel and exciting solubility and
host–guest interaction properties.25,26 One of the primary
requisites of such post-modification is the conversion of
existing functional groups of the polymer into reactive ones
for further chemical changes. Such chemical modification can
be easily performed on hPG using classical hydroxyl group
chemistry thereby changing the hPG hydroxyl groups namely
to azides, alkynes, amines, and to many others.27 Unlike
a Institute of Chemistry and Biochemistry, Freie Universitat Berlin,
¨
Takustr. 3, 14195 Berlin, Germany. E-mail: haag@chemie.fu-berlin.de;
Fax: +49 30 83853357
b Department of Physics, Humboldt-Universitat zu Berlin, Newtonstr. 15,
12489 Berlin, Germany
¨
w Electronic supplementary information (ESI) available. See DOI:
10.1039/c1nj20466a
z This article is part of the themed issue Dendrimers II, guest edited by
Jean-Pierre Majoral.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 371–379 371

View Article Online
Fig. 1 Core–shell architectures with the hPG scaffold; [x = 5 kDa or 10 kDa; y = 7, 12, 16, 24, 45; z = 0.4, 0.6, 0.9].
dendrimers hPGs show no distinguishable interior or periphery.
Instead they possess two types of hydroxyl functionalities
arising from linear and terminal hydroxyl units (Fig. 1).
Conceptually, these linear hydroxyl groups are in proximity
to the core compared to the terminal ones which are closer to
the periphery of the molecule. The so-called ‘‘selective chemical
differentiation’’ strategy enables one to selectively and differentially
modify these two types of hydroxyl groups to generate core–
shell-type architectures within the hPG scaffold.19,25,28 To this
end, the 1,2-diols of the terminal glycerol units were selectively
transformed into the corresponding ketals/acetals in order to
distinguish between the interior (close to the focal unit) and
periphery (distant from the focal unit) of the macromolecule,
which was possible as the remaining linear glycerol units are
unaffected by this transformation. A subsequent reaction of
the linear units and selective deprotection of 1,2-ketals are
generally performed to yield core-functionalized hPGs. The
procedure allows selective tailoring of the hPG scaffold to
contain hydrophobic substituents (e.g. aromatic rings or
fluorinated chains) in the interior, thereby modulating the
distribution co-efficient of the generated structure between
organic and aqueous phases.
This paper reports the synthetic procedure to generate
core–shell architectures of the hPG scaffold containing
hydrophilic (PEG) and/or amphiphilic functionality in a
random or location-specific manner. The host–guest interaction
properties of the synthesized architectures in terms of transport
of hydrophilic/hydrophobic guest molecules to corresponding
non-compatible media have been compared and contrasted.
Different chain lengths of mPEG have been attached to hPG
either directly (1) or through a biphenyl spacer (2) to generate
fully hydrophilic or amphiphilic constructs (Fig. 1 and 2).
Synthetically such molecules are generated by attaching either
mPEGs alone or mPEG pre-coupled with biphenyl spacers to
hPG using CuI mediated ‘‘click chemistry’’29–31 after conversion
of the hPG hydroxyl groups to the corresponding azide
functionality. Location specific functionalization of an hPG
carrier has been attained by attaching biphenyl rings to the hPG
core using a selective ‘‘chemical differentiation’’ strategy,
followed by attachment of the PEG chains as the shell to the
c
372 New J. Chem., 2012, 36, 371–379
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Fig. 2 Model for polymer architectures 2 and 3 with non-polar guest molecules.
peripheral glycerol units (3). The chemical difference between
the two categories of constructs lies in the fact that in the case of
compounds 1 and 2, the hydrophilic mPEG units or amphiphilic
PEG-biphenyl units are randomly distributed throughout the
hPG molecule, while for compound 3, the hydrophobic and
hydrophilic segments are specifically located within the focal
core and the peripheral shell of the hPG respectively, The
constructs have been evaluated for their transport capacity as
a function of structural diversity, molecular weight of the hPG
core, and length of the hydrophilic PEG chain. Efficiency of
the synthesized architectures to encapsulate and transport
different guest molecules (i.e. b-carotene, Congo red, nimodipine
and Rose Bengal) to their non-compatible environment has been
assessed. The resulting supramolecular assemblies were
studied by UV-Vis spectroscopy and SFM measurement.
synthetic scheme for the preparation of the core–shell
structures is presented in Fig. 3. For detailed synthetic
information and analytical data see ESI.w
Synthesis of core–shell architectures 1: hPG–mPEG system
Alkynylation of poly(ethylene glycol) monomethylether
(mPEG). 1 eq. of mPEG (7, 12, 16, 24, 45) was dissolved in
tetrahydrofuran (THF) and sodium hydride (2 eq.) was added.
The mixture was stirred for one hour and propargylbromide
(2 eq.) was added dropwise at 0 1C. The reaction mixture was
allowed to warm up to room temperature and stirred for 24 h.
The crude product was purified by column chromatography
Mesylation of hPG. This reaction was carried out according
to the previously published procedure with slight modification.
Briefly, hPG (Mn of 5 kDa and 10 kDa) was dissolved in abs.
pyridine and mesyl chloride (0.5, 0.7, or 1.0 eq. per OH group
depending on the desired degree of functionalization) in abs.
pyridine was added dropwise under inert gas. The mixture was
stirred at room temperature for 22 h and the crude products
were purified by dialysis in acetone to achieve a 40, 60 and
90% functionalized O-mesylpolyglycerol.
Experimental part
General strategy for the synthesis of hPG-based core–shell
architectures (1–3)
The conceptual scheme of the synthesized polymers and their
abbreviations are presented in Fig. 1.
The hPG-based core–shell architecture containing a randomly
distributed PEG chain (1) or a biphenyl unit spaced PEG chain
(2) was synthesized utilizing CuI mediated azide-alkyne
cycloaddition, widely known as ‘‘click chemistry.’’ The
synthesis of construct 2 was carried out first by alkynylation
of 4,40-dihydroxyl biphenyl resulting in the bisalkyne-terminated
biphenyl derivative. In the following steps, azide terminated
mPEG of different lengths was clicked to one of the termini of
the biphenyl units. Subsequently, this PEG attached biphenyl
moiety was again clicked, this time onto an azide-terminated
hPG scaffold. The core–shell system 3 with a biphenyl moiety
attached to the hPG core and mPEG units in the periphery
(shell) was constructed in a step-wise pattern. Firstly, the
biphenyl 4-methyl groups were selectively coupled to the
hPG core through etherification by a previously published
procedure.25 In the following steps, clickable PEG chains of
different lengths were attached to remaining hydroxyl groups
via the click addition protocol after conversion of these groups
to corresponding azide or alkyne groups. The generalized
Azidation of mesylated hPG. Azidation of O-mesylpolyglycerol
was carried out by dissolving mesylated hPG in dimethyl-
formamide and reacting with sodium azide (4 eq. per OM
group). The mixture was stirred at 120 1C for 4 h. After
cooling to room temperature the residual sodium azide was
removed via filtration. The orange filtrate was concentrated
in vacuo and the crude product was purified by dialysis in
chloroform (48 h).
Coupling of alkyne terminated mPEG to hPG. hPG-azide
(1 eq. of azide group) and mPEG-alkyne were mixed in
THF/water solution (1 : 1, 20 ml). Next CuSO4ꢀ5H2O
(5 mol%) and sodium ascorbate (10 mol%) were added and
the mixture was stirred for 24 h at room temperature. The
solvent was removed in vacuo and the residue was dissolved in
methanol. After filtration, the crude product was purified by
dialysis in methanol following dialysis in 5% ethylenediamine-
tetraacetic (EDTA) acid/water solution to yield the core–shell
structure (1).
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 371–379 373

CH. 4,40-
View Article Online
Fig. 3 Synthesis of core–shell architectures (1–3).
R
Synthesis of core–shell architecture 2: mPEG attached to hPG
through a biphenyl spacer
Synthesis of mPEG7-biphenyl-O–CH2–C
Bis(prop-2-ynyloxy)biphenyl (3.0 eq.) and mPEG7–N3
(1.0 eq.) were mixed with THF. To the resulting suspension
in water, CuSO4ꢀ5H2O (5 mol%), ascorbic acid (10 mol%) and
NaOH (10 mol%) were added and the brown mixture was
stirred at room temperature for 48 h. After removal of the
solvent in vacuo THF (20 ml) was added and the mixture was
filtered to remove the solid residues. The crude product was
purified by column chromatography using chloroform to
chloroform/methanol 10 : 1 as an eluent.
Tosylation of poly(ethylene glycol) monomethylether. mPEG7
and triethylamine (2.1 eq.) were dissolved in THF and the
resulting solution was cooled down to 0 1C. p-Toluene
sulfonylchloride (1.9 eq.) in THF was added dropwise for
1 h. After stirring for 20 h at room temperature the mixture
was filtered to remove the precipitated triethylamine hydro-
chloride salts. The filtrate was concentrated in vacuo and the
crude product was purified by column chromatography.
Coupling of alkyne terminated mPEG-biphenyl to hPG. hPG-
azide (1 eq. of the azide group) and mPEG7-biphenyl-
O–CH2–CRCH (1.3 eq.) were mixed in THF/water solution
(1 : 1, 20 ml). Next CuSO4ꢀ5H2O (5 mol%) and sodium
ascorbate (10 mol%) were added and the mixture was stirred
for 24 h at room temperature. The solvent was removed in
vacuo and the residue was dissolved in methanol. After
filtration, the crude product was purified by dialysis in methanol
following dialysis in 5% EDTA/water solution.
Azidation of O-tosyl poly(ethylene glycol) monomethylether.
mPEG7–OTs (1 eq.) was dissolved in dimethylformamide and
sodium azide (2 eq.) was added. After 28 h of stirring at room
temperature the excess of sodium azide was removed by
filtration. Removal of the solvent was done in vacuo and the
crude product was purified by column chromatography using
ethyl acetate as an eluent.
Alkynylation of 4,40-dihydroxy-biphenyl. To a solution of
4,40-dihydroxy-biphenyl (1 eq.) in dimethylformamide was
added sodium hydride (2 eq.). After stirring for 2 h the
solution was cooled down to 0 1C and propargylbromide
(2.5 eq.) was added slowly via a syringe. The reaction
temperature was allowed to rise to room temperature and to
run for 49 h. Afterwards, addition of water resulted in
precipitation of the product. The mixture was filtered and
the product was washed with water to obtain 4,40-bis(prop-2-
ynyloxy)biphenyl.
Synthesis of core–shell architecture 3: hPG with biphenyl units
at the core and mPEG at the shell
hPG core functionalized with biphenyl-4-methyl ether group.
The reaction was performed according to a previously
published procedure.25
Mesylation and azidation of core-functionalized hPG. This
reaction was carried out according to the protocol described
c
374 New J. Chem., 2012, 36, 371–379
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
for core–shell system 1, after deprotection of the terminal
SFM
hydroxyl groups.
A droplet of polymer solution in chloroform and in water was
deposited on a freshly cleaved mica surface and spun off after
5 seconds. The surface was dried and imaged by SFM in
tapping-mode under ambient conditions, with a Nanoscope 3a
(Veeco, USA), using silicon cantilevers (Olympus, Japan) with
a typical resonance frequency of 300 kHz and a spring
constant of about 42 N mꢁ1. Both height and phase images
were recorded.
Coupling of alkyne terminated mPEG to core-functionalized
hPG. Core-functionalized hPG-azide (1 eq. of the azide group)
and mPEG-alkyne (1.3 eq.) were mixed in THF/water solution
(1 : 1, 20 ml). Next CuSO4ꢀ5H2O (5 mol%) and sodium
ascorbate (10 mol%) were added and the mixture was stirred
for 24 h at room temperature. The solvent was removed
in vacuo and the residue was dissolved in methanol. After
filtration, the crude product was purified by dialysis in
methanol following dialysis in 5% EDTA/water solution to
yield core–shell system 3.
Results and discussion
Synthesis of a new core–shell architecture
Three different types of core–shell nanoparticles based on the
hPG scaffold have been developed where the constructs varied
in terms of specific arrangement of hydrophobic and
hydrophilic segments. For both classes of structures, biphenyl
moieties and PEG chains represent the hydrophobic and
hydrophilic units respectively. While architecture 1 is fully
hydrophilic with different chain lengths of mPEG units
attached to the hPG scaffold, in core–shell architecture 2,
the hydrophobic and hydrophilic segments are arranged
side-by-side and functionalized throughout the hPG scaffold.
In core–shell architecture 3, the hydrophobic biphenyl units are
specifically located within the hPG core and the hydrophilic
PEG units are located in the shell (Fig. 1). For generating
compound 1, alkyne terminated mPEGs have been attached to
UV/VIS Spectroscopy
UV/VIS spectra were measured with a SCINCO (S-3100)
spectrometer (Jasco co.) in CHCl3 or water.
Determination of the transport capacity
The hPG-based core–shell architectures were tested for their
capacity to encapsulate and transport guest molecules of
varying polarity to non-compatible media. The transport
capacity of the carrier systems was determined in
a
monophasic system with the solvent being either chloroform
or water. Polar (Congo red, Rose Bengal) and non-polar
(nimodipine, b-carotene) guest molecules (Fig. 4) were used
for the transport experiment. For the polar guest molecules
chloroform was selected as the non-polar solvent and for the
non-polar guest molecules water was used. To the solution of
the core–shell architecture (1 g lꢁ1), the guest compound was
added as a solid (insoluble in the solution) and the mixtures
were rigorously stirred for 16–18 h. After subsequent removal
of the non-encapsulated guest molecules via a syringe-filter
(Rotilabo, 13 mm in pore size) the absorption of the solution
was measured with UV/VIS spectroscopy and evaluated via
comparison to a calibration curve.
azide functionalized hPGs by
a straightforward ‘‘click
reaction’’. Architecture 2 was synthesized by bisalkynylation
of the biphenyl moiety, followed by coupling to mPEG and
attachment of a bisphenyl-mPEG segment to hPG. For
generating compound 3, the biphenyl units are first attached
to the core region of hPG by Williamson etherification reaction.
Prior to this reaction, the terminal hydroxyl groups were
blocked by the formation of acetal and selectively cleaved off
after the core-functionalization generating free hydroxyl
groups. Conversion of these linear hydroxyl groups to clickable
azide or alkyne functionality enabled the immobilization of the
alkyne-terminated PEG shell by CuI assisted click reaction.
Both alkyne and azide terminated hPGs were tested with a
orthogonally terminated PEG chain to find the best approach.
No differences in yields and the functionalization level were
observed (yield B60%, functionalization B90%) regardless of
azide or alkyne functionalities present in the hPG or in the
PEG. The azide groups can be introduced in hPG under
ambient reaction conditions in a moderate to high functionalization
level. The first step to insert the azide groups is the mesylation of the
hPG hydroxyl groups. The extent of mesylation reaction can be
stoichiometrically controlled by the amount of applied mesyl
chloride. Upon azidation the mesyl groups convert quantitatively
to azide groups which can be followed by IR spectroscopy.
The alkyne groups, on the other hand, can also be generated
within the hPG scaffold, although a higher degree of alkyne
functionalization is difficult to achieve, probably for steric
reasons. Additionally densely distributed alkyne functionality
over a hPG architecture increases the chance of internal side
reactions involving the CRC bond. To this end, the route of
preferential azidation of hPG has been undertaken to generate
Fig. 4 Structures of polar (Congo red, Rose Bengal) and non-polar
(nimodipine, b-carotene) guest molecules.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 371–379 375

View Article Online
clickable hPG. A series of the substructures have been
synthesized following this general protocol by varying the
molecular weight of the core, chain length of PEG, and degree
of functionalization (ESIw).
Transport capacity for hPG–mPEG polymers 1
In order to fully understand the influence of each domain of
the macromolecules on the encapsulation efficiency, molecular
weight of hPG, the length of the PEG chain, and the level of
functionalization were varied to generate
a library of
substructures of 1 (ESIw). hPG–mPEG polymers containing
PEG chains of different lengths attached to hPG with no
hydrophobic units transported only Congo red and Rose Bengal
as guest molecules from water to non-polar solvent (chloroform).
These structures were not able to encapsulate and transport
non-polar guest molecules (nimodipine, b-carotene) in water.
A similar result has been reported for the unfunctionalized
hyperbranched polyglycerol.32,33 The reason lies in the fact
that these polymers possess a polar hPG core and a highly
hydrophilic mPEG periphery that generates a highly polar
environment within the molecule. The transport capacity of
hPG–mPEG series (hPG core of 5 kDa) for Congo red to the
chloroform phase was determined by UV-VIS spectroscopy
via comparison to calibration curves and is presented in Fig. 5.
With the increase of mPEG length no significant changes in
the mg/g transport ratio were observed. Increasing the degree
of functionalization on the other hand with mPEGs of
different chain lengths caused a generalized reduction of
transport capacity probably due to increased steric blockade.
For the polymers with hPG 10 kDa core size, four different
PEG lengths (mPEG7, mPEG12, mPEG16 and mPEG24) were
tested as the shell moiety (see ESIw). In contrast to molecules
with smaller cores the transport ratio initially increases with
prolongation of mPEG chains from 7 to 12 glycol units, and
then diminishes continuously for longer PEGs with 16 and 24
monomer units. This tendency is best visible for series with
50% level of core functionalization with mPEGs where the
transport capacity increases initially from B50 mg gꢁ1 for
mPEG7 to B65 mg gꢁ1 for mPEG12 and then decreases to
B45 mg gꢁ1 and B35 mg gꢁ1 for mPEG16 and mPEG24
respectively (ESIw).
Fig. 6 Transport-structure dependences of core size (hPG5 kDa
,
hPG10 kDa) and core functionalization (conversion) demonstrated on
polymers with mPEG7-shells (transport capacities a) in [mg guest/g
polymer] and (b) in [mol guest/mol polymer]).
The level of core functionalization was found to impart a
substantial impact on the transport efficiency of hPG–mPEG
polymer series. Increased core functionalization reduces
transport capacity both in mg guest/g host or mol guest/mol
host level, particularly for the hPG 10 kDa system (Fig. 6). At a
high functionalization level of 70% to 90%, the transport
capacity remains almost constant or changes irregularly.
Modification of the core size increased the mol/mol transport
ratio by approximately a factor of three for polymers with a
lower functionalization level (40–70%) and by a factor of two
for highly functionalized polymers. Considering the higher
molecular masses of architectures with a hPG 10 kDa core
than those with a hPG 5 kDa core, the change of transport
capacity is not substantial in the mg/g ratio. The transport
increased only by 20–25% with the increase of the core size for
functionalization levels within the range of 40–70% and even
slightly decreased for polymers with a 90% functionalized core
(Fig. 6). For core–shell architectures with hPG 5 kDa, the effect
of functionalization was less prominent than for polymers with
a bigger core, and stayed below a factor of two between 90%
and 40% core functionalization. The transport result of Rose
Bengal encapsulated in a hPG–mPEG polymer was found to be
associated with oversaturation phenomena of the carrier
molecule with the dye. With increasing amounts of Rose Bengal
the encapsulation values rose to a maximum saturation level
that led to precipitation of the Rose Bengal–polymer
complexes, which were subsequently removed from the solution
via filtration. Therefore the results of Rose Bengal were not
reliable and could not be used for interpretation.
Transport capacity for core–shell architecture 2: mPEG
attached to hPG through a biphenyl spacer
Incorporation of a nonpolar element into the core–shell
architecture can be used in order to achieve the transport of
Fig. 5 Transport-structure dependences of polymer with a hPG5 kDa
core (transport capacities) for Congo red in [mg guest/g polymer].
c
376 New J. Chem., 2012, 36, 371–379
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
non-polar guest molecules in polar solvent.25 The insignificant
transport of non-polar guest molecules by the hPG–mPEG
systems 1 motivated us to design core–shell architectures
where the mPEG units are attached to the hPG core through
biphenyl spacers, thereby increasing the amphiphilicity of the
molecule. The biphenyl spacer was chosen as a linker because
of its interaction with aromatic scaffolds present in different
guest molecules (i.e. drugs and dyes).
As expected core–shell system 2 solubilized non-polar guest
molecule nimodipine in the water phase (Fig. 7). The transport
capacity was found to vary directly with increasing levels of
functionalization (40 to 90%) of hPG with biphenyl
spaced-mPEG units (ESIw). The reported role of the biphenyl
spacer to interact with an aromatic p-electron system of
nimodipine molecules forming the p–p interaction motif is a
major contributing factor for such encapsulation enhancement.25
In the case of b-carotene, the transport capacities of the
core–shell system 2 were significantly lower (0.08 mg gꢁ1 to
0.50 mg gꢁ1) probably due to the extended structure of
b-carotene that is difficult to accommodate within the
dendritic scaffold. For Congo red as guest molecules, changes
in the encapsulation efficiency of the guest molecule to the host
molecule [mg guest/g host] decreased stepwise from 61.76 to
34.81 mg gꢁ1 of a polymer with an increasing functionalization
level of 40–90% (ESIw).
Fig. 8 UV spectra of the complex of nimodipine with polymers 2b
and 3.
region. It is envisioned that such hybrid structures will include
the high non-polar guest transport efficiency of structure 2 and
at the same time can carry polar guest molecules to a non-polar
environment.
The transport results of nimodipine with such architectures
are presented in Fig. 8. The result showed that polymer 3 can
transport more nimodipine than core–shell structure 2 at the
same level of functionalization with mPEG of equivalent
length. Polymer 3’s transport capacity was also higher than
compound 2b’s which had an even larger number of biphenyl
groups (60%). We assumed that the biphenyl spacer at the
core facilitated intermolecular aggregation of the hPG
scaffolds due to p–p interactions and thereby generated larger
aggregates that incorporated more non-polar guest molecules.
Furthermore, architecture 3 generated a nonpolar core that
increased the hydrophobicity of the system as well as the
solubility contribution of nimodipine in the polymer.
Transport capacity for core–shell architecture 3: hPG with
biphenyl units at the core and mPEG at the shell
Introduction of biphenyl spacers between mPEG and hPG
caused a drastic change in the polarity profile of the resulting
molecule. In our previous report we have also shown that
introduction of a biphenyl fragment into the hPG core via a
chemical differentiation strategy can change the encapsulated
and transport property of hPG. Such molecules with
hydrophobic cores and hydrophilic shells were able to solubilize
non-polar guest molecules in water.25,28 However, this system
has the limitation that the conversion of the linear hydroxyl
groups of hPG must be lower than 50%, in order to keep the
systems soluble in water. Furthermore, the system could not
transport polar guest molecules to nonpolar solvent such as
chloroform. Combining these two phenomena, we synthesized
core–shell architecture 3 by a selective chemical differentiation
strategy where the biphenyl units are located near the foci of the
hPG scaffold and mPEG units are distributed in the peripheral
For b-carotene as guest molecules, the transport capacity
for core–shell architecture 3 was also higher (0.30 mg gꢁ1) than
for 2 probably due to increased hydrophobicity and higher
propensity of the systems to form aggregates. For Congo red
as guest molecules, the transport capacity for compound 3 was
lower than for 1 and 2 mostly due to the polarity reduction of
the core. However, when Rose Bengal was introduced as the
guest molecules to this host polymer, the solubility was higher
in 3 than in 2 but lower compared to hPG–mPEG systems
with an equivalent level of PEG functionalization.
Host–guest complexes
To reveal the properties of the polymer and the drug/dye
complexes (host–guest interaction), scanning force microscopy
(SFM) was used. The study was performed by preparing the
solution of the polymer in chloroform or water.
Fig. 9 demonstrates hPG5 kDa(biphenyl)0.2core(mPEG7)0.6shell
(compound 3) on mica deposited from a chloroform solution
of 0.1 mg mlꢁ1. Single round shaped particles with an average
size of around 10 nm were detected on the surface. Particles
with sizes up to 40 nm were observed (Fig. 9a). Zooming in
on one of the particles reveals a double layer structure both
in height and phase images: a center core and a surrounding
shell (Fig. 9b and c). A height profile across the center of
the particle indicated by the dotted line shows the height
difference between the central core, the shell, and the substrate
Fig. 7 Transport-structure dependences of Polymer 2a–c and 3 for
nimodipine.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 371–379 377

View Article Online
Fig. 10 (a, b) AFM height images of hPG5kDa(biphenyl)0.2 core
(mPEG7)0.6 shell (polymer 3) in water (0.1 mg mlꢁ1), (c, d) 0.1 mg mlꢁ1
polymer 3 solution with nimodipine.
Fig. 9 AFM images of hPG5 kDa(biphenyl)0.2 core(mPEG7)0.6
shell
(polymer 3) in chloroform (0.1 mg mlꢁ1): (a) height image,
(b) zoom-in height, (c) phase image, (d) height profile indicated by
the dotted line in (b).
moieties have also been differently arranged within the hPG
scaffold to increase the amphiphilicity of the architecture.
Depending on the structural features these nanocarriers were
found to encapsulate and transport a wide range of guest
molecules. The defined core–shell arrangement in compound 3
respectively (Fig. 9d). Numbers and arrows in the height profile
indicate the corresponding positions in the image.
We attribute the single particles to single polymer
molecules. The size difference in single particles is attributed
to the size distribution of the polymer. The different contrast
in both height and phase images may be caused by different
chemical components in the core and the shell, which confirms
the core–shell structure of the molecule.
showed
a
significantly higher transport capacity for
hydrophobic guest molecules indicating that the stepwise
modification of the core and shell leads to more defined systems.
In general both nanotransporters 2 and 3 contain aromatic
residues and are capable of carrying hydrophobic molecules,
which is in clear contrast to hPG–mPEG 1. The molecular
weight of the core, degree of functionalization, and presence of
hydrophilic and hydrophobic moieties within the nanocarrier
structure were found to critically govern their encapsulation
properties. The biocompatibility of the core–shell architectures
discussed here is now under investigation for their possible
application as drug delivery vehicles.
In the case of the solution of core–shell structure 3 in water
(0.1 mg mlꢁ1), we observed single round shaped small particles of
around 10 nm as well as aggregates of around 50 nm (Fig. 10a).
Zooming in onto one of the aggregates revealed that the aggregate
was formed by single small particles (Fig. 10a and b). The same
polymer solution codissolved with nimodipine led to formation
of larger aggregates of around 200 nm on the surface (Fig. 10c
and d). Again, a higher resolution height image reveals that the
aggregates were formed by many small particles.
Acknowledgements
We attribute the small round shaped particles to single
polymer molecules. The aggregates have been shown to form by
single molecules through hydrophobic core–core interactions.26
The hydrophilic shell–shell interaction also plays a role in the
formation of the aggregates, which is confirmed by the fact that no
aggregates formed in a chloroform solution shown in Fig. 9. The
presence of hydrophobic organic compounds such as nimodipine
seems to enhance the interaction between the molecules, which
results in the formation of larger aggregates.
The authors thank Dr Andreas Mohr for the discussion and
Meta Mentari for providing hPG-azide. This work was
supported by the Bundesministerium fur Bildung und
¨
Forschung (Nano Future Award for R.H.) and the German
Science Foundation (DFG) within the collaborative research
center SFB 765.
Notes and references
1 J. F. G. A. Jansen, E. M. M. de Brabander-van den Berg and
E. W. Meijer, Science, 1994, 265, 1226; M. W. P. L. Baars,
R. Kleppinger, M. H. J. Koch, S.-L. Yeu and E. W. Meijer, Angew.
Chem., Int. Ed., 2000, 39, 1285.
Conclusion
2 C. C. Lee, J. A. MacKay, J. M. J. Fre
Biotechnol., 2005, 23, 1517.
´
chet and F. C. Szoka, Nat.
Core–shell type architectures based on hPG have been
designed where PEG shells are either directly coupled to hPG
or through a biphenyl spacer. The hydrophilic and hydrophobic
3 R. Duncan and L. Izzo, Adv. Drug Delivery Rev., 2005, 57, 2215.
4 R. Esfand and D. A. Tomalia, Drug Discovery Today, 2001, 6, 427.
c
378 New J. Chem., 2012, 36, 371–379
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
5 S. J. Guillaudeu, M. E. Fox, Y. M. Haidar, E. E. Dy, F. C. Szoka
and J. M. J. Frechet, Bioconjugate Chem., 2008, 19, 461.
21 M. Calderon, M. A. Quadir, S. Sharma and R. Haag, Adv. Mater.,
2010, 22, 190.
´
6 A. W. Bosman, H. M. Janssen and E. W. Meijer, Chem. Rev., 1999,
99, 1665.
22 D. Howes, R. Wilson and C. T. James, Food Chem. Toxicol., 1998,
36, 719; R. Wilson, B. J. Van Schie and D. Howes, Food Chem.
Toxicol., 1998, 36, 711.
7 F. Zeng and S. C. Zimmerman, Chem. Rev., 1997, 97, 1681.
8 C. Gao and D. Yan, Prog. Polym. Sci., 2004, 29, 183.
9 R. Haag, Angew. Chem., Int. Ed., 2004, 43, 278.
10 C. Liu, C. Gao and D. Yan, Angew. Chem., Int. Ed., 2007, 46, 4128.
11 C. Liu, C. Gao and D. Yan, Macromolecules, 2006, 39, 8102.
12 C. J. Hawker, Advanced Computer Simulation Approaches for Soft
Matter Science, Spinger, Berlin, 1999, vol. 147, p. 113.
13 A. Sunder, J. Heinemann and H. Frey, Chem.–Eur. J., 2000,
6, 2499.
23 R. Haag, A. Sunder and J.-F. Stumbe
122, 2954.
24 R. Haag and F. Kratz, Angew. Chem., Int. Ed., 2006, 45, 1198.
´
, J. Am. Chem. Soc., 2000,
25 H. Turk, A. Shukla, P. C. A. Rodrigues, H. Rehage and R. Haag,
¨
Chem.–Eur. J., 2007, 15, 4187.
26 E. Burakowska and R. Haag, Macromolecules, 2009, 42, 5545.
27 S. Roller, H. Zhou and R. Haag, Mol. Diversity, 2005, 9, 305–316.
28 I. N. Kurniasih, H. Liang, J. P. Rabe and R. Haag, Macromol.
Rapid Commun., 2010, 31, 1516.
14 R. Haag, Chem.–Eur. J., 2001, 7, 327.
15 N. Karak and S. Maiti, Dendrimers and Hyperbranched Polymers
Synthesis to Applications, M D Publications Pvt. Ltd, New Delhi,
2008.
16 B. Voit and A. Lederer, Chem. Rev., 2009, 109, 5924.
17 Y. Zhou, W. Huang, J. Liu, X. Zhu and D. Yan, Adv. Mater.,
2010, 22, 4567.
29 V. V. Rostovtstev, L. G. Green, V. V. Folkin and K. B. Sharpless,
Angew. Chem., Int. Ed., 2002, 41, 2596; C. W. Tornoe,
C. Christensen and M. Meldal, J. Org. Chem., 2002, 67, 3057.
30 V. O. Rodinov, V. V. Fokin and M. G. Finn, Angew. Chem., 2005,
117, 2250; V. O. Rodinov, V. V. Fokin and M. G. Finn, Angew.
Chem., Int. Ed., 2005, 44, 2210.
18 A. Sunder, R. Hanselmann, H. Frey and R. Mulhaupt, Macro-
¨
H. Frey, Adv. Mater., 2000, 12, 235.
19 R. Haag, J.-F. Stumbe, A. Sunder, H. Frey and A. Hebel, Macro-
´
31 V. D. Bock, H. Hiemstra and J. H. v. Maarseveen, Eur. J. Org.
Chem., 2006, 51; F. Himo, T. Lovell, R. Hilgraf, V. V. Rostovtsev,
L. Noodlemann, K. B. Sharpless and V. V. Folkin, J. Am. Chem.
Soc., 2005, 127, 210.
¨
molecules, 1999, 32, 4240; A. Sunder, R. Mulhaupt, R. Haag and
molecules, 2000, 33, 8158.
20 D. Holter, A. Burgath and H. Frey, Acta Polym., 1997, 48, 30.
32 T. Ooya, J. Lee and K. Park, J. Controlled Release, 2003, 93, 121.
33 T. Ooya, J. Lee and K. Park, Bioconjugate Chem., 2004, 15, 1221.
¨
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 371–379 379