Water-soluble dendritic architectures with carbohydrate shells for the templation and stabilization of catalytically active metal nanoparticles

Article abstract of DOI:10.1021/ma0510791

Hyperbranched poly(ethylenimine) (PEI) with amino groups or carbohydrate terminal groups have been used as support materials for metal nanoparticles (i.e., Cu, Ag, Au, and Pt) in water. Various parameters have been optimized, such as pH, concentration of the polymer in solution, and [metal ions]/ [polymer] ratio in order to obtain stable metal nanoparticles with a narrow size distribution. TEM measurements revealed that particles with a diameter as low as 1.4 nm were obtained. An increase of stability was obtained after functionalization of PEI with glycidol, gluconolactone, or lactobionic acid. In the case of Pt the catalytical activity of the corresponding nanoparticles was evidenced.

Full text of DOI:10.1021/ma0510791

8308
Macromolecules 2005, 38, 8308-8315
Water-Soluble Dendritic Architectures with Carbohydrate Shells for the
Templation and Stabilization of Catalytically Active Metal
Nanoparticles
Michael Kr a1 mer,† Nelly P e´ rignon, Rainer Haag,* Jean-Daniel Marty,*
,‡
†
,‡,§
,†
‡
†
†
Ralf Thomann, Nancy Lauth-de Viguerie, and Christophe Mingotaud
IMRCP, UMR CNRS 5623, Universit e´ Paul Sabatier, 118 route de Narbonne, 31062 Toulouse, France;
Freiburger Materialforschungszentrum und Institut f u¨ r Makromolekulare Chemie, Universit a¨ t
Freiburg, Stefan-Meier-Str. 21, 79104 Freiburg, Germany; and Department of Chemistry, Free
University of Berlin, Takustrasse 3, 14195 Berlin, Germany
Received May 26, 2005; Revised Manuscript Received July 18, 2005
ABSTRACT: Hyperbranched poly(ethylenimine) (PEI) with amino groups or carbohydrate terminal groups
have been used as support materials for metal nanoparticles (i.e., Cu, Ag, Au, and Pt) in water. Various
parameters have been optimized, such as pH, concentration of the polymer in solution, and [metal ions]/
[
polymer] ratio in order to obtain stable metal nanoparticles with a narrow size distribution. TEM
measurements revealed that particles with a diameter as low as 1.4 nm were obtained. An increase of
stability was obtained after functionalization of PEI with glycidol, gluconolactone, or lactobionic acid. In
the case of Pt the catalytical activity of the corresponding nanoparticles was evidenced.
cles.2
,19-26
In contrast to those perfectly branched mono-
Introduction
disperse dendrimers, randomly branched polymers can
Metal nanoparticles are of special interest because
they show physical properties owing to quantum-me-
chanical rules that are neither those of the bulk metal
nor those of molecular compounds. Encapsulated metal
nanoparticles have found potential application in
be easily accessible.2
7,28
These hyperbranched polymers
have already been successfully used as templates for
metal nanoparticles.1
5,29-32
In the literature, it was
suggested that hyperbranched polyamines, e.g. poly-
ethylenimine (PEI), are better suited for stabilization
of metal nanoparticles due to the more globular struc-
1
-5
6
catalysis
or material sciences due to their large
7
surface/volume ratio as discussed in the literature.
Furthermore, they have also found applications in
15
ture as compared to linear polymers. So far poly-
1
5,16
8
(ethylene glycol)-functionalized PEIs
and alkyl-
medicine and biology as labeling agents, for transfer
modified PEIs2
9,33,34
have been reported as stabilizers
9
of drugs including genetic material to the nucleus,
1
0
for metal nanoparticles in organic solvents.
detection of metal ions in solution, single molecule
11
In this paper, we report for the first time on readily
accessible water-soluble dendritic core-shell architec-
tures and study the influence of the attached carbohy-
drate shell on the formation and stabilization of metal
nanoparticles in water. The control of the 3D architec-
ture of the hyperbranched polymer (via choice of the Mw
of the polymer and functionalization) might allow size
control of the nanoparticles and the enhancement of
their stability. For this purpose, three different carbo-
hydrates acting as a shell, i.e., glycidol, gluconolactone,
and lactobionic acid, were attached to hyperbranched
PEI of different molecular weights (Mw) to obtain the
corresponding PEI-glycol, PEI-gluconamide, and PEI-
lactobionamide. Various parameters such as pH, con-
centration, [metal ions]/[polymer] ratio, and the nature
of the metal precursors (HAuCl4, AgNO3, CuSO4, and
H2PtCl6) have been optimized to obtain stable nanopar-
ticle systems. In the case of stabilized Pt nanoparticles,
we have also briefly investigated their catalytic activity.
spectroscopy, and as organic-inorganic semiconductor
composites with defined sizes that are promising can-
1
2
didates for nanoelectronic devices.
In the present state of the research, the preparation
of nanosized water-soluble analogues of heterogeneous
catalysts for the use as more efficient and environmental
friendly catalytic systems is still a challenge,13 and new
efficient approaches are required to obtain structures
that are stable toward shear forces and dilution. To
reach such a goal, the requirements are (a) the develop-
ment of nanoreactors in which the catalytic nanopar-
ticles can be formed in a controlled fashion (i.e., with
an average diameter usually less than 10 nm) and (b)
the use of stabilizers which prevent the thermodynami-
cally favored aggregation of the nanoparticles in the
solution without any loss of the catalytic properties due
to a too strong adsorption of the stabilizers on the active
1
4
sites of the particles. To overcome these problems,
various molecular systems were reported to stabilize
metal nanoparticles in aqueous media, such as poly-
1
5-18
electrolytes or block copolymers.
Derivatives of
Results and Discussion
polyamidoamine (PAMAM) dendrimers were often used
as nanoreactor and stabilizer for metal nanoparti-
Synthesis of Dendritic PEIs with Carbohydrate
Shells. Hyperbranched PEIs can be obtained in a single
cationic polymerization step of aziridine.3
5-37
Slow
†
Universit e´ Paul Sabatier.
monomer addition allows control over the degree of
‡
Universit a¨ t Freiburg.
38
branching (DB) as discussed in the literature. These
§
Free University of Berlin.
Corresponding authors. E-mail: marty@chimie.ups-tlse.fr;
polymers are available in a multigram scale and high
molar mass with a polydispersity index typically below
*
haag@chemie.fu-berlin.de.
1
0.1021/ma0510791 CCC: $30.25 © 2005 American Chemical Society
Published on Web 08/27/2005

Macromolecules, Vol. 38, No. 20, 2005
Metal Nanoparticles 8309
Scheme 1. Structure of Hyperbranched Polymers. Functionalization of PEI with (a) Glycidol (PEI
Gluconic Acid (PEI GLU), and (c) Lactobionic Acid (PEI LAC)
x
GLY), (b)
x
x
Table 1. Properties of Investigated Polymers PEIx,
2
.0 (Supporting Information).28 In the following, modi-
a
PEI25GLY, PEIxGLU, and PEI25LAC
3
fied PEIs are labeled as PEIx with ×10 ) Mw. In a
second reaction step, PEIx was modified with a co-
valently attached hydrophilic shell to create water-
soluble core-shell architectures with PEIx as core and
the carbohydrate acting as a shell (Scheme 1). The
density and size of the shell were varied by functional-
ization of the amine groups with a controlled amount
of glycidol (PEIxGLY), gluconolactone (PEIxGLU), or
lactobionic acid (PEIxLAC).
Mw
Mn
[
g mol-1]
[g mol-1]
MWD
DPnc
DF [%]
PEI0.8
800
5 000
21 000
25 000
1 800
11 900
51 200
59 800
47 800
78 000
600
3 600
1.3
1.4
2
2.6
1.3
1.4
2
2.6
2.6
2.6
14
84
PEI5
PEI21
PEI25
10 500
244
223
14
b
9 600
1 400
8 500
25 600
23 000
18 400
30 000
PEI0.8GLU
PEI5GLU
PEI21GLU
PEI25GLU
PEI25GLY
PEI25LAC
33
33
33
33
53
27
84
244
223
223
223
Functionalization with glycidol, leading to PEI25GLY
(PEI-glycol, Scheme 1a), can be achieved via addition
of DL-glycidol to a solution of PEI25 in methanol, result-
ing in complete conversion of the reactive epoxide group
after 24 h at room temperature. The degree of function-
alization (DF ) number of shell molecules/number of
monomer unit of PEI, given in %) was determined via
a
GLU ) functionalization with gluconic acid. GLY ) function-
alization with glycidol. LAC ) functionalization with lactobionic
acid. Mn ) number-averaged molar mass. Mw ) mass-averaged
molar mass. MWD ) molecular weight dispersity. DPn ) number
degree of polymerization. DF ) degree of functionalization of all
b
-1
1
monomer units (Supporting Information). Mn ) 9600 g mol
integration of the respective H NMR signals (Support-
-
1
calculated from Mw and MWD is too low; a Mn of 12 000 g mol
ing Information). The attachment of D-glucono-1,5-
lactone to PEIx with different Mw (x ) 0.8, 5, 21, or 25),
leading to PEIxGLU (PEI-gluconamide, Scheme 1b), can
be realized in DMSO as reported in the literature with
can be estimated from GPC due to existence of a shoulder in the
c
diagram. Represents also the number of primary and tertiary
amine groups of the hyperbranched structure.
3
9,40
PAMAM dendrimers
or via a newly developed
istics of the water-soluble polymers are summarized in
Table 1.
pathway by a simple melt reaction of glucono-1,5-lactone
with PEIx at 140 °C (see Experimental Section). The
advantage of the latter reaction is that only stoichio-
metrical amounts of water have to be added (glucono-
lactone:water ) 1:1) to obtain a soluble polymer with a
precise stoichiometry control. Inverse gated 13C NMR
in DMSO indicates that about 70% of the gluconolactone
is covalently attached and about 30% is linked via an
ionic bridge. Functionalization of the reactive primary
and secondary nitrogen atoms of PEIx with gluconolac-
tone could be realized up to 50% as determined by
elemental analysis (Supporting Information). The de-
gree of functionalization controlled by the stoichiometry
of the reaction in the case of the PEIxGLU was con-
firmed by elemental analysis. A higher degree of func-
tionalization via the melt reaction route resulted in a
dark brown and insoluble polymer. In the case of the
functionalization of PEI25 with lactobionic acid, leading
to PEI25LAC (PEI-lactobionamide, Scheme 1c), 1-ethyl-
Preparation of Metal Nanoparticles. Polymer-
stabilized nanoparticles were prepared in a two-step
process. After complexation of metal ions with the
respective polymer in a first step, a chemical reduction
with sodium borohydride was performed in a second
step to obtain the metal nanoparticles.
Complexation with Metal Cations. In the case of
dendrimers, it was shown in the literature that they can
only encapsulate a maximum number of metal ions per
molecule.21 The maximum metal loading of the den-
drimer and the value of the [metal ions]/[dendrimer]
ratio determines the size of the formed nanoparticles.23
To evaluate the ability of the dendritic architecture of
various PEIs to act in the same fashion as dendrimers,
complexation properties of PEI5 and PEI GLU were
determined in the case of copper, gold, and platinum
5
ions. (Because of rapid self-reduction of silver ions in
the presence of polymers prior to the addition of a
reduction agent, complexation studies could not be
rigorously performed in the case of silver.) All the
complexation results obtained are summarized in Table
2. In the case of Pt(IV) and Au(III), slow kinetics have
3
-[3-(dimethylamino)propyl]carbodiimide hydrochloride
(
EDC) was used as coupling reagent in water, and the
degree of functionalization was determined via elemen-
tal analysis (Supporting Information). The character-

8310 Kr a¨ mer et al.
Macromolecules, Vol. 38, No. 20, 2005
Table 2. Complexation Properties of PEI5 and PEI5GLUa
solution exhibits a strong absorption band at 226 nm
and a shoulder at 290 nm due to charge transfer
between the metal and the chloro ligands (Supporting
Information). After addition of HAuCl4 to a solution of
PEI5, the absorbance at 290 nm increased linearly with
PEI5
PEI5GLU
CuSO4
H2PtCl6
HAuCl4
40
45
35
40
40
40
III
III
the quantity of added Au ions until the [Au ]/[PEI]
ratio reaches ca. 35. The change in the slopes suggests
also some interactions between the nitrogen atoms of
the polymer and the gold cations. As already described
a
Precision on values is (5.
1
been reported for complexation with dendrimers. To
avoid any problems, the incubation time for metal
uptake was chosen to be equal to 48 h for all metals,
although due to the lower degree of branching, com-
plexation with hyperbranched polymers should occur
more rapidly. (No significant evolution of the spectra
was observed 24 h after the beginning of the complex-
ation process.) All samples were stable on this time scale
if protected from light exposure.
41
in the case of chitosan, formation of ion pairs between
-
-
AuCl4 or hydrolyzed forms (AuCl3OH , etc.) and the
protonated amino groups of PEI could be responsible
at least partially for this phenomenon. Whatever the
-
nature of the interaction between AuCl4 and the
hyperbranched polymer may be, this experiment indi-
cates that PEI5 can be loaded with a maximum number
III
Initially, an aqueous solution of the respective poly-
mer (at various concentrations calculated with the Mw
value) was mixed with an aqueous solution of metal ions
of ca. 35 Au ions per macromolecule and PEI5GLU
III
with ∼40 Au ions per macromolecule.
In the absence of PEI5 or PEI5GLU, the UV-vis
(
CuSO4, HAuCl4, H2PtCl6, or AgNO3) to obtain a con-
-4
absorption spectrum of a 10
M H2PtCl6 solution
trolled molar ratio [metal ions]/[polymer] (in the range
consists of a strong absorption peak at 196 nm arising
from a ligand-to-metal charge-transfer transition (Sup-
porting Information). If PEI5 or PEI5GLU is added to
this solution, a new band at 230 nm emerges while the
band at 196 nm shifts slightly to 199 nm. The absor-
bance at 230 nm is proportional to the number of
0
-100). Figure 1 shows the results obtained from
complexation of copper ions with PEI5. In the absence
II
of a polymer with complexation capabilities, Cu exists
2
+
primarily as [Cu(H2O)6] in aqueous solutions, which
gives rise to a broad, weak absorption band at 810 nm
associated with a d-d transition (ꢀ ∼ 10). In the
IV
complexed Pt ions by the polymer over the range of
II
presence of hyperbranched PEI5, λmax for the Cu d-d
IV
[
Pt ]/[PEI5GLU] ) 0-40, which indicates that it is
transition was shifted to 605 nm (ꢀ ∼ 30). In addition,
a strong ligand-to-metal charge-transfer (LMCT) transi-
tion appeared at 300 nm (Figure 1A). This change in
possible to control the extent of the complexation by the
IV
IV
[
Pt ]/[PEI5] and [Pt ]/[PEI5GLU] ratio.
II
These results demonstrate that loading of the den-
the UV-vis spectrum allows to follow the Cu ion
dritic architectures by metal cations is possible in a well-
defined range, as it was already observed for dendrim-
ers. The maximum number of loaded metal ions was
found to be around 40 for PEI5 and is close to the
number evaluated for a fourth-generation PAMAM
dendrimer (i.e., 50),32 whereas the molecular weight of
the hyperbranched polymer is only half of it. This
suggests, as already observed with hyperbranched
polymers derivated from PAMAM dendrimers,32 that
the open structure of the hyperbranched polymer com-
pared to the dendrimer architecture facilitates the
interactions of the metal ions with the internal chemical
functions, increasing therefore the maximum load of
ions per polymer. Moreover, the substitution of amino
groups of PEI5 by gluconamide did not significantly
modify the complexation properties. This suggested a
possible interaction between substituted amino groups
and metal ions (or only the dendritic N atoms are
involved in the complexation process).
complexation with different polymers. The absorbance
II
at λmax increased with the [Cu ]/[polymer] ratio (Figure
1
B) until a critical value is reached, above which the
absorbance increased only slowly. The change in the
slopes suggests different environments for the copper
cations before and after this critical value and therefore
some interactions between the nitrogen atoms of the
polymer and the metal cations. Extrapolating the two
linear regions of the absorbance at this wavelength
allows to evaluate the maximum copper loading of the
polymer. This analysis indicates that PEI5 (hyper-
-
1
branched polyethylenimine with Mw ) 5000 g mol
)
II
can complex up to 40 Cu ions. Functionalization of the
primary or secondary amino groups with gluconamide
results in similar observations, indicating that the two
polymers (PEI5 and PEI5GLU) possess the same com-
plexation capabilities.
In the case of HAuCl4, besides a very weak band
around 640 nm, the absorption spectrum of a HAuCl4
-
4
-1
; 10-5 mol L-1 of PEI
-4
-1
Figure 1. (A) Absorption spectra of aqueous solutions: 2 × 10 mol L of CuSO
4
5
; 2 × 10 mol L of
-
5
-1
-5
-1
II
CuSO
4
, and 10 mol L of PEI
5
. (B) Absorption spectra of a solution 5 × 10 mol L of PEI
5
with various amounts of Cu (the
optical path length is 0.5 cm). (C) Absorbance at 605 nm as a function of [copper ion]/[polymer] ratio. Determination of capacity
was based on the highest achieved absorbance.

Macromolecules, Vol. 38, No. 20, 2005
Metal Nanoparticles 8311
Formation of Nanoparticles: Conditions of Sta-
bilization. To obtain the respective metal nanopar-
ticles, reduction of the cations complexed by the polymer
was carried out with a 1-100 fold excess of NaBH4. If
not mentioned, a 100-fold excess was chosen to induce
a fast reduction of the metal cations. Numerous publica-
tions deal with the formation, stability, and size of metal
nanoparticles in dependence of the [metal ions]/[poly-
Table 3. Gold Nanoparticles Obtained from Different
III
a
Metal Loading [Au ]/[Polymer] Ratios
Mw
Mn
ratio [AuIII]/
[polymer]
sample
[g mol-1] [g mol-1]
stability
PEI0.8
800
600
precipitation
precipitation
precipitation
precipitation
precipitation
precipitation
0.67
3.3
6.7
0.74
3.7
7.4
6.8
6.8
5.0
5.0
PEI0.8GLU
1 800
1 400
15,16,19,20
15,42
mer] ratio,
pH of the solution,
concentration
concentration of metal
cations in solution, and the reduction agent ap-
1
7-20
of the polymer in solution,
b
PEI5
5 000
6 900
21 000
51 200
3 600
8 500
10 500
25 600
stable
1
7
c
PEI5GLU
PEI21
stable
stable
stable
1
5,16
plied.
In the following part, parameters that influ-
PEI21GLU
ence the stability of nanoparticles/PEI systems, such as
the [metal ions]/[polymer] ratio, the molecular weight
of the polymer core, pH of the solutions, and the
concentration of the polymer in solution, were investi-
gated.
a
pH of all solutions was 12 adjusted with phosphate buffer;
-1
concentration of HAuCl was 0.29 mmol L ; reduction was carried
4
out with NaBH4; concentration of the polymer was between 0.4
-
1 b
and 0.06 mmol L
. Partial precipitation was observed im-
c
mediately. Partial precipitation was observed after 1 week.
Stability Depending on the Concentration of the
Polymer and the [Metal ions]/[Polymer] Ratio. In
a first set of experiments, the influence of the polymer
concentration in solution on the stability of metal
nanoparticles and the maximum possible [metal ions]/
Stability in Dependence of the Molecular Weight
of the Polymeric Core. To investigate the influence
of the molecular weight of the polymer on the stabiliza-
[polymer] ratio was evaluated in order to obtain systems
tion properties, three different PEIs (PEI0.8, PEI , and
5
that show no sign of immediate precipitation after
reduction by sodium borohydride. For concentrated
PEI ) functionalized with gluconic acid were synthe-
21
sized. Additionally, the stability of gold nanoparticles
-
2
-3
-1
polymer solutions (10 and 10 mol L ) immediate
complete or partial precipitation was observed regard-
less of the [metal ions]/[polymer] ratio. In the case of
was investigated with unfunctionalized PEI (Table 3).
x
Higher molecular weight polymers have been found to
be better suited to stabilize higher [metal ions]/
[polymer] ratios independent of the nature of the
attached shell. No stabilization was possible with PEIs
-
4
-5
lower concentrations of the polymers (10 -10 mol
-
1
L
), immediate precipitation was observed only for
-1
[
metal ions]/[polymer] ratio above the maximum com-
with a molecular weight of 800 g mol regardless if they
were functionalized with gluconic acid or not. However,
gold nanoparticles can be stabilized with high molecular
weight unfunctionalized PEI (see Table 3, PEI ). For
plexation numbers given in Table 2. Under these
conditions, copper and silver nanoparticles were less
stable than their gold and platinum counterparts:
partial or even complete reoxidation of metal nanopar-
ticles was observed when exposed to atmospheric oxygen
21
the intermediate case (i.e., PEI ), the solution was
5
filtrated after reduction due to partial precipitation of
2
2
as expected (Supporting Information). With these
results in mind, hyperbranched polymer solutions with
gold which occurred immediately. In contrast, the
PEI GLU showed only partial precipitation after 1
5
-
4
-5
-1
concentrations in the range of 10 -10 mol L should
week.
result in the most stable metal nanoparticles with
metal ions]/[polymer] ratios below 40 if high molecular
Influence of the Attached Carbohydrate Chain
on the Stability. To enhance the stability of nanopar-
ticles, the influence of an attached carbohydrate chain
on the stabilization properties of the polymers on silver
nanoparticles at a pH of 12 was examined. Previous
experiments revealed that silver nanoparticles were less
stable than their gold counterparts and therefore more
sensitive to the nature of the applied stabilizer. Func-
tionalization of the linear and terminal monomer units
of PEI25 with glycidol (53%), gluconolactone (33%), and
lactobionic acid (27%) results in the desired polymers
with PEI core and carbohydrate shell. Experimentally,
even the higher degree of functionalization of PEI with
glycidol (53%) results in reduced stability compared to
PEI functionalized with gluconolactone (33%). The
rather short glycol units are therefore not sufficient to
enhance stability of the silver nanoparticles. The at-
tachment of the more bulky lactobionic acid (27%)
results in a further increase of stability as compared to
the gluconamide. The degree of functionalization with
lactobionic acid is comparable to the PEI-gluconamide.
However, none of the polymers were able to stabilize
silver nanoparticles for more than 2 weeks regardless
of their functionalization. Nevertheless, a clear increase
in stability with increasing shell size was revealed.
These experiments demonstrate that the stability in-
creased in the series glycol < gluconamide < lactobiona-
mide.
[
weight polymers are used. The observation of increased
stability of nanoparticles at lower concentrations of the
stabilizer in aqueous solution is in agreement with the
4
3
literature.
Stability of Metal Nanoparticles in Dependence
of the pH. To evaluate the influence of the pH on the
stability of metal nanoparticles, the more stable Au
nanoparticles were investigated to minimize the influ-
ence of the polymer-metal system. The primary effect
of excess borohydride was to reduce the metal ions, but
at the same time it raises the pH of the solution, so it
was necessary to adjust the pH with buffer solution. The
III
Au compound and PEI21GLU were mixed together to
obtain an aqueous solution with a concentration of 230
-
1
and 9.1 µmol L , respectively, resulting in a [metal
ions]/[polymer] ratio of 25. The pH of the solution was
adjusted with a buffer solution to obtain pH values of
2
, 5, 8, and 12. Only in the case of pH ) 12 were stable
nanoparticles obtained; in all other cases precipitation
occurred after reduction. The increased stability of gold
nanoparticles after reduction with NaBH4 at a pH of
1
5,16
1
2 is in agreement with the literature
and was also
observed for other metal nanoparticles such as Pd, Pt,
and Au and PEI-PEG stabilizers. All further experi-
ments were therefore carried out in buffered solution
at a pH of 12.

8312 Kr a¨ mer et al.
Macromolecules, Vol. 38, No. 20, 2005
Table 4. Size of Different Metal Nanoparticles Stabilized with Different Polymers in Dependence of the [Metal Ions]/
[
Polymer] Ratioa
size [nm] ( standard deviation ([metal ions]/[polymer] ratio)
PEI
functionalization
Ag
Au
Ptb
PEI5
13.7 ( 12.6 (8.5)
6.4 ( 5.6 (8.5)
8.8 ( 3.0 (6.8)
7.7 ( 2.6 (6.8)
6.3 ( 5.3 (5)
1.3 ( 0.3 (5)
4.1 ( 2.3 (25)
4.2 ( 3.3 (25)
3.8 ( 1.6 (25)
5.6 ( 4.5 (25)
40 ( 10 (20)c
8.6 ( 1.6 (20)
PEI5
GLU
GLU
PEI21
PEI21
PEI25
PEI25
PEI25
PEI25
10.7 ( 7.6 (6.3)
1.9 ( 0.6 (6.4)
15.7 ( 5.6 (23.6)
13.6 ( 6.4 (23.6)
12.8 ( 6.5 (25.5)
12.6 ( 9.2 (25.7)
GLU
GLY
LAC
a
GLU ) functionalization with gluconolactone (33%). GLY ) functionalization with glycidol (53%). LAC ) functionalization with
b
c
lactobionic acid (27%). In contrast to the Ag and Au nanoparticles, the pH was controlled via the addition of NaBH4. Very polydisperse
sample with large clusters (more than 100 nm).
Concerning stability, the experiments performed so
far revealed that the use of high pH conditions and
[
metal ions]/[polymer] ratio below maximum load num-
bers are necessary conditions to stabilize metal nano-
particles with PEI derivatives in water. Nevertheless,
when considering other parameters, differences in
stability were apparently dependent on the nature of
the metal nanoparticle. No stabilization was possible
in conjunction with copper particles. Reoxidation occurs
immediately in the presence of oxygen, which is in
agreement with the literature.21 Silver nanoparticles
were less stable than their gold counterparts. The
experiments revealed that the stability of silver nano-
particles depends on the polymer applied for stabiliza-
tion and can be extended with functionalized PEIx
(PEIxGLU and PEIxLAC). Gold nanoparticles could be
stabilized with unfunctionalized PEIx; however, smaller
particles were obtained if the PEIxGLU was used. Pt
particles were stable for months if PEIxGLU was used.
They could not be stabilized with unfunctionalized PEI.
Formation of Nanoparticles: Size Control of the
Nanoparticles Obtained. The effects of [metal ions]/
[polymer] ratio, molar mass, and nature of the carbo-
hydrate shell on the size of the nanoparticles formed
are investigated. PEIx with different Mw (PEI5, PEI21,
and PEI25) was functionalized with various carbohy-
drate shells. The resulting polymers PEI5GLU,
PEI21GLU, PEI25GLU, PEI25GLY, and PEI25LAC were
used to create the respective nanoparticles. The results
obtained from the analysis of TEM measurements are
summarized in Table 4.
Figure 2. TEM histograms of Au particles of (a) PEI
filtration (due to immediate partial precipitation after reduc-
tion) and (b) PEI GLU (without filtration, partial precipitation
occurred after 1 week after reduction); mean diameter of PEI
and PEI GLU is 8.8 and 7.8 nm, respectively, at a ratio of
Au ]/[polymer] of 6.8. Black bar in the TEM pictures equals
0 nm.
5
after
5
5
(a) Influence of the Molar Mass of the Polymeric
5
III
[
5
Core. The average particle size of gold and silver
nanoparticles decreased with increasing Mw of the
polymeric core (8.8 and 6.3 nm diameter for gold
particles stabilized by PEI5 and PEI21, respectively, and
of the size of silver nanoparticles stabilized with three
different polymers functionalized with carbohydrate
chains (PEI GLU, PEI GLY, and PEI LAC) exhibited
7
.7 and 1.3 nm diameter for gold particles stabilized by
PEI5GLU and PEI21GLU, respectively). Similar results
were observed in the case dendrimer encapsulated
25
25
25
no significant differences (14, 13, and 13 nm, respec-
tively, at a ratio of 25). It seems that the nature of the
attached carbohydrate chain influences rather the
stability properties than the particle size.
4
4
nanoparticles.
b) Influence of the Substitution of Amino
(
Groups. Functionalization of the amino groups with
carbohydrate chains leads to nanoparticles with nar-
rower size distribution and smaller mean particle
diameter as compared to PEIx (Figures 2 and 3). For
example, in the case of PEI21GLU, Ag and Au nanopar-
ticles smaller than 2 nm were formed (1.9 and 1.4 nm
diameter at ratios of 6 and 5, respectively) as compared
to unfunctionalized PEI21 (10.8 and 6.3 nm diameter at
ratios of 6 and 5, respectively). In this case, the resulting
particle sizes are comparable to the ones obtained with
(c) Influence of the [Metal]/[Polymer] Ratio. It
III
was reported in the literature that a Au /[G5.5]-
PAMAM ratio of 10 (Mn [G5.5]-PAMAM ) 50 864 g
-
1
mol ) exhibited particle sizes of 5.6 nm, whereas at a
ratio of 1/10 the synthesized particles were smaller (2.3
nm).19 We have observed the same behavior and ob-
tained smaller particle sizes after lowering the [metal
ions]/[polymer] ratio. In the case of the PEI21GLU the
mean particle diameter of gold nanoparticles was 1.3
nm at a ratio of 5. A higher molecular weight PEI25-
19,20,23,44
PAMAM dendrimers.
Moreover, the comparison

Macromolecules, Vol. 38, No. 20, 2005
Metal Nanoparticles 8313
Figure 4. GC chromatograms of products of the hydrogena-
5 5
tion reaction in the presence of PEI and PEI GLU stabilized
platinum nanoparticles for similar conditions (see Experimen-
tal Section).
effect, i.e., the possibility to control the size of nanopar-
ticles just by changing specific parameters: (i) molecular
weight and functionalization and/or (ii) the [metal ions]/
[
polymer] ratio.
Catalytical Applications. Oxidation of copper and
silver nanoparticles (vide supra) has indicated the
accessibility of the surface of metal nanoparticles and
potential catalytic activity. As a first probe for the
applicability of these polymer-stabilized water-soluble
nanoparticles in catalysis, the hydrogenation of isophor-
one with hyperbranched polymer-encapsulated plati-
num nanoparticles was investigated in water under
mild conditions (Scheme 2, low pressure of H2 ) 2 bar,
room temperature). The product of the hydrogenation
reaction was analyzed by monitoring the disappearance
of isophorone by gas chromatography (commercial prod-
ucts were used as standards). Conversion rate was
measured as well as the turnover number (TONs)
defined as the moles of isophorone consumed by moles
of Pt and the turnover frequencies (TOFs) defined as
the moles of isophorone consumed by moles of Pt and
by hour. The hydrogenation activities of nanoparticles
stabilized by PEI5, PEI25, PEI5GLU, PEI25GLU,
PEI25GLY, and PEI25LAC are given in Table 5 as well
as the ones of PAMAM of third and fourth generation
([G3]-PAMAM and [G4]-PAMAM, respectively) mea-
sured under the same conditions for comparison.
Catalytic activities for the chosen substrate isophor-
one are quite low, but in good agreement with the
literature45 considering the mild conditions that were
used here. Moreover, they are calculated with respect
to the entire amount of Pt present in the clusters; i.e.,
activity of the significant surface atoms will be higher.
Nevertheless, the results obtained allow to describe the
main characteristics of the catalytic activity of our
systems. No reduction of the ketone function is observed.
The only product resulting from hydrogenation is 3,3,5-
Figure 3. TEM histograms of Au particles of (a) PEI21 and
(
b) PEI21GLU; mean diameter of PEI21 and PEI21GLU is 6.3
III
and 1.4 nm, respectively, at a ratio of [Au ]/[polymer] of 5.
Black bar in the TEM pictures equals 50 nm.
Scheme 2. Hydrogenation Reaction of Isophorone
Catalyzed by Platinum Stabilized Nanoparticles
GLU led to a particle diameter of 4.1 nm at a ratio of
2
5. It is noteworthy that gold nanoparticles were much
smaller than silver ones at the same ratio of 25 (4.2,
3
.8, and 5.6 nm, respectively).
We were able to show that the dendritic architectures
can be used as templates to obtain nanoparticles of
defined size. In view of the observed cluster sizes and
roughly estimating that the PEIxGLU or PEIx gyration
radius is around 4.5 nm, a diameter of about 13 nm can
only be explained if more than one macromolecule
stabilizes a nanoparticle as suggested in the literature
in the case of gold nanoparticles stabilized by low-
1
9
generation PAMAM dendrimers. However, high mo-
lecular weight polymers at low [metal ions]/[polymer]
ratios (that leads to the smallest nanoparticles) might
be able to stabilize nanoparticles in the interior of the
polymer. Our studies demonstrate a clear template
Table 5. Results Concerning Hydrogenation of Isophorone Catalyzed by Platinum-Stabilized Nanoparticles
[
PtIV]/[polymer]
stability
conversion [%]
TONc
TOFc,d
[
[
G3]-PAMAM
G4]-PAMAM
20
20
20
20
20
20
20
20
stable
stable
46.0
50.0
0.2
23.0
25.0
0.1
2.9
3.1
<0.1
4.0
<0.1
<0.1
<0.1
0.4
a
a
a
PEI5
partial precipitation
b
PEI5GLU
PEI25
PEI25GLU
PEI25GLY
PEI25LAC
stable
64.0
0.2
32.0
0.1
partial precipitation
b
stable
0.6
0.3
partial precipitation
0.2
0.1
b
stable
17.0
8.5
a
Observed immediately. b Stable during the time of experiment; however, partial precipitation was observed after several weeks.
c
Calculated with respect to the entire amount of Pt present in the clusters; i.e., activity of the significant surface atoms will be higher.
d
Moles of isophorone consumed by moles of Pt and by hour (mol of isophorone (mol of Pt)-1
-1
h
).

8314 Kr a¨ mer et al.
Macromolecules, Vol. 38, No. 20, 2005
trimethylcyclohexan-1-one. As expected, no significant
catalytic activity was observed for systems that are not
stable on the time scale of the experiment (PEI5, PEI25,
PEI25GLY): existence of large clusters not dispersed in
aqueous solutions induced a decrease of the Pt surface
accessible to the substrate and therefore poor catalytic
activity. The catalytic activities of PEI5GLU are com-
parable to [G3]-PAMAM and [G4]-PAMAM encapsu-
lated platinum nanoclusters and demonstrate good
accessibility of the colloidal metal for the substrate. The
hydrogenation activity could be controlled in two ways.
First, as already observed for dendrimers, the hydro-
genation reaction rate was lowered if a hyperbranched
polymers with a high Mw was used (compare PEI5GLU
and PEI25GLU) due to reduced accessibility of the
substrates to the metal nanoparticles located in the
interior of the polymer. In addition, modification of the
outer shell (compare PEI25GLU and PEI25LAC) allows
a similar control. Very poor catalytic activities of
PEI25GLY emphases the crucial role of stability for
systems to be used as catalyst. After the hydrogenation
reactions were run for up to 8 h, the solutions of stable
nanoparticles remained clear, and there was no evidence
of agglomeration (only after weeks partial precipitation
occurred). The recycling of the catalyst (by ultracen-
trifugation and isolation as a powder) does not lead to
a significant decrease in activity after redissolving it in
water.
units), gluconolactone (1.466 g, 8.23 mmol), and stoichiometri-
cal amounts of water (142 mg, 7.89 mmol) were mixed together
and heated to 140 °C oil bath temperature for 10 min. About
50 wt % of the water was evaporated under these conditions
to yield a light brown and brittle solid (yield: 99%). Inverse
1
3
gated C NMR (d-DMSO) revealed that about 30% of the
-1
gluconolactone is bonded via a salt bridge. IR: v˜ (cm ) ) 1640
1
(
CdO, covalent), 1600 (CdO, ionic). H NMR (300 MHz,
2 2
D O): δ (ppm) ) 2.3-2.9 [m, PEI-CH -NHCO-CHOH-
(CHOH) -CH OH], 3.2-3.3 [m, PEI-CH -NHCO-CHOH-
3
3
3
2
2
2
2
2
2
(CHOH)
(CHOH)
-CH
-CH
OH], 3.5-3.8 [m, PEI-CH
OH], 3.8-4.0 [m, PEI-CH
-NHCO-CHOH-
-NHCO-CHOH-
4
6
(CHOH) -CH OH], 4.19 [s, PEI-CH -NHCO-CHOH-
3
2
2
1
3
(
CHOH)
3
-CH
2
OH]. C NMR (75.4 MHz, D
2
O): δ (ppm) )
3
4
6
7
1
8.7, 39.5, 40.4, 41.0 [PEI-CH
2
-NHCO-(CHOH)
-NHCO-(CHOH)
4
-CH
-CH
2
OH],
OH],
7.8, 49.6, 53.2, 54.7 [PEI-CH
2
4
2
2 4 2
5.0 [PEI-CH -NHCO-(CHOH) -CH OH], 72.7, 73.5, 74.5,
4.9, 75.8, 76.4 [PEI-CH
2
-NHCO-(CHOH)
4 2
-CH OH],
76.7 [PEI-CH -NHCO-(CHOH)
2
4
-CH OH], 180.9 [PEI-
2
+
-
CH -NH
2
4
4 2
OOC-(CHOH) -CH OH]. Elemental analysis
(found): C, 44.27%; N, 12.02%; H, 8.07%.
Attachment of Lactobionic Acid to PEI25 To Obtain
-
1
PEI25LAC. A solution of PEI25 (500 mg, 15.8 mmol g T and
L units) and lactobionic acid (1.423 g, 3.97 mmol) in water (12
mL) was adjusted to pH ) 5 with hydrochloric acid, and
1
-ethyl-3-[3-(dimethylamino)propyl]carbodiimide hydrochlo-
ride (EDC, 776.5 mg, 4.05 mmol) was added at room temper-
ature. The reaction mixture was stirred for 3 days at room
temperature. The crude product was dialyzed in water to
obtain a yellow powder (yield: 69%). About 50% of lactobionic
13
acid is covalently bonded according to inverse gated C NMR.
1
H NMR (300 MHz, D
NHCO-lact], 3.4-4.5 [PEI-CH
MHz, D O): δ (ppm) ) 36.2, 37.4, 39.0 [PEI-CH
lact], 45-55 [PEI-CH -NHCO-lact], 61.6, 62.4 [PEI-CH
NHCO-lact(CH OH)], 69.1, 70.8, 71.6, 72.2, 73.0, 75.8 [PEI-
CH -NHCO-lact], 81.3, 82.1 [PEI-CH -NHCO-lact(CH-
O-C(acetal)-O-CH)], 103.9 [PEI-CH -NHCO-lact(acetal-
C)], 174.8, 175.8 [PEI-CH -NHCO-lact], 179.0 [PEI-CH -
2
O): δ (ppm) ) 2.4-3.4 [PEI-CH -
2
13
Conclusion
2
-NHCO-lact]. C NMR (75.4
-NHCO-
2
2
In summary, we have reported on a simple protocol
to prepare dendritic core-shell architectures with car-
bohydrate shells. For the first time, hyperbranched PEI
2
2
-
2
2
2
(
(
polyethylenimine) or functionalized PEI with glycidol
PEI-GLY), gluconolactone (PEI-GLU), or lactobionic
2
2
2
+
-
acid (PEI-LAC) molecules have been used as support
materials for metal nanoparticles in water. Various
parameters such as the concentration of the polymer
in solution as well as the [metal ions]/[polymer] ratio
and pH have been optimized. We were able to show that
these dendritic polymers can act as colloid stabilizers
to control the size of the nanoparticles. (In the case of
PEI21GLU a template effect was clearly evidenced.) The
attached carbohydrate shell influences rather the sta-
bility of the nanoparticles than their size. The prepara-
tion of these innovative materials involves anchory of
Pt(IV) to the support prior to reduction with NaBH4 in
water. The TEM micrographs revealed that the dimen-
sions of the clusters are greatly influenced by the nature
of the polymeric stabilizer. The platinum nanoparticles
stabilized with PEI derivatives proved to be as effective
and robust for the selective hydrogenation of isophorone
in water as the dendrimer derivatives. Moreover, they
permit simple workup by phase separation, and the
catalytical results demonstrate the ability to control the
reaction rate by adjusting the chemical properties of the
stabilizer (size of the core or nature of the shell).
Furthermore, improvement of the biocompatibility in-
duced by the carbohydrate shell will allow the additional
usage of these particles in biological systems.
NH
H, 7.46%.
4
OOC-lact]. Elemental analysis: N, 8.91%; C, 39.89%;
Attachment of DL-Glycidol to PEI25 To Obtain PEI25
-
-
1
GLY. To a solution of PEI25 (1.0 g, 23.3 mmol g N-H) in
methanol (10 mL) a solution of DL-glycidol (11.7 mmol for 50%
functionalization) in methanol (10 mL) was added in a drop-
wise fashion at room temperature. The reaction mixture was
stirred for 1 day. Subsequent removal of the solvent yielded a
1
2
colorless viscous oil (yield: 96%). H NMR (300 MHz, D O): δ
(ppm) ) 2.5-2.8 [PEI-CH -CHOH-CH OH], 3.35 (traces of
2
2
CH
3
OH), 3.55 [PEI-CH
2
-CHOH-CH
2
2
OH], 3.82 [PEI-CH -
1
3
CHOH-CH
2
2
OH]. C NMR (75.4 MHz, D O): δ (ppm) ) 40-
6
0 [PEI-CH
2
-CHOH-CH OH), 64.5
2
OH], 49.3 (traces of CH
3
[
PEI-CH
2
-CHOH-CH OH], 69.8, 70.8 [PEI-CH -CHOH-
2
2
CH
2
OH].
Typical Procedure for the Complexation Studies. For
these experiments no buffered solutions were used. The
solutions were equilibrated for at least 1 h before measure-
ment.
Typical Procedure for the Hydrogenation of Isophor-
one with Platinum Nanoparticles. 10 mL of an aqueous
-
4
solution containing 2 × 10 mol of H
2 6
PtCl complexed with
-
5
10 mol of PEI25GLU reduced by a 10-fold excess of NaBH
relative to Pt) was stirred under H for 1 h prior to the
addition of 10 mol of isophorone. The progress of the reaction
was followed by sampling at different time and extracting
these samples with dichloromethane. The reaction product was
analyzed by gas chromatography.
4
(
2
-2
Gas chromatography measurements were performed with
Varian Chrompack (CP-3800) apparatus equipped with a SGE
nonpolar capillary column (BPX5 30 m × 0.25 mm) (injector
temperature: 280 °C; splitless injection mode; carrier gas:
Experimental Section
GPC Measurement of PEI. GPC (gel permeation chro-
matography) of the PEIs have been performed by BASF AG
with pullulan standard (Supporting Information). The molec-
ular weights were confirmed by light scattering.
-1
helium at 1 mL min ; initial oven temperature: 60 °C, 1 min;
-
1
rate 1: 20 °C min ; final temperature: 100 °C; rate 2: 20 °C
-
1
Attachment of Gluconolactone to PEI
x
To Obtain
min ; final temperature: 250 °C, 2 min; detector: 300 °C;
solvent: dichloromethane). The retention time were found to
-
1
PEI GLU. For example, PEI21 (1 g, 16.2 mmol g T and L
x

Macromolecules, Vol. 38, No. 20, 2005
Metal Nanoparticles 8315
be ∼10 min for the isophorone and 6.9 min for the correspond-
(21) Zhao, M.; Sun, L.; Crooks, R. M. J. Am. Chem. Soc. 1998,
120, 4877-4878.
ing hydrogened product.
(
(
(
22) Balogh, L.; Tomalia, D. A. J. Am. Chem. Soc. 1998, 120,
7
355-7356.
Supporting Information Available: Experimental sec-
tion consisting of figures showing NMR and IR spectra,
calculation of the degree of functionalization of the polymers
discussed, TEM micrographs, UV-vis data, GPC graphs and
table with characteristics of PEI polymers, table illustrating
stability of nanoparticles. This material is available free of
charge via the Internet at http://pubs.acs.org.
23) Esumi, K.; Suzuki, A.; Aihara, N.; Usui, K.; Torigoe, K.
Langmuir 1998, 14, 3157-3159.
24) Li, Y.; El-Sayed, M. A. J. Phys. Chem. B 2001, 105, 8938-
8943.
(25) Rahim, E. H.; Kamounah, F. S.; Frederiksen, J.; Christensen,
J. B. Nano Lett. 2001, 1, 499-501.
(26) Ooe, M.; Murata, M.; Mizugaki, T.; Ebitani, K.; Kaneda, K.
Nano Lett. 2002, 2, 999-1002.
(
27) Fr e´ chet, J. M. J.; Henmi, M.; Gitsov, I.; Aoshima, S.; Leduc,
M. R.; Grubbs, R. B. Science 1995, 269, 1080-1083.
References and Notes
(
28) Frey, H.; Haag, R. In Encyclopedia of Materials, Science and
Technology; Buschow, K. H. J., R. H. C., Flemings, M. C.,
Ilschner, B., Kramer, E. J., Majahan, S., Eds.; Elsevier
Science Ltd.: Oxford, 2001; pp 3997-4000.
(
(
(
(
(
(
1) Scott, R. W. J.; Wilson, O. M.; Crooks, R. M. J. Phys. Chem.
B 2005, 109, 692-704.
2) Crooks, R. M.; Zhao, M.; Sun, L.; Chechik, V.; Yeung, L. K.
Acc. Chem. Res. 2001, 34, 181-190.
(29) Aymonier, C.; Schlotterbeck, U.; Antonietti, L.; Zacharias, P.;
3) Maye, M. M.; Lou, Y.; Zhong, C.-J. Langmuir 2000, 16, 7520-
Thomann, R.; Tiller, J. C.; Mecking, S. Chem. Commun. 2002,
7523.
3018-3019.
4) Cuenya, B. R.; Baeck, S.-H.; Jaramillo, T. F.; McFarland, E.
W. J. Am. Chem. Soc. 2003, 125, 12928-12934.
(
(
(
(
(
(
(
(
30) Mecking, S.; Thomann, R.; Frey, H.; Sunder, A. Macromol-
ecules 2000, 33, 3958-3960.
5) Jaramillo, T. F.; Baeck, S.-H.; Cuenya, B. R.; McFarland, E.
W. J. Am. Chem. Soc. 2003, 125, 7148-7149.
31) Sablong, R.; Schlotterbeck, U.; Vogt, D.; Mecking, S. Adv.
Synth. Catal. 2003, 345, 333-336.
6) Hall, S. R.; Shenton, W.; Engelhardt, H.; Mann, S. Chem-
PhysChem 2001, 3, 184-186.
32) P e´ rignon, N.; Mingotaud, A.-F.; Marty, J.-D.; Rico-Lattes, I.;
Mingotaud, C. Chem. Mater. 2004.
(
7) Daniel, M.-C.; Astruc, D. Chem. Rev. 2004, 104, 293-346.
8) Zheng, J.; Dickson, R. M. J. Am. Chem. Soc. 2002, 124,
33) Haag, R.; Kr a¨ mer, M.; Stumb e´ , J.-F.; Krause, S.; Komp, A.;
Prokhorova, S. Polym. Prepr. 2002, 43, 328.
(
13982-13983.
34) Garcia-Bernab e´ , A.; Kr a¨ mer, M.; Ol a` h, B.; Haag, R. Chem.s
Eur. J. 2004, 10, 2822-2830.
(
9) Tkachenko, A. G.; Xie, H.; Coleman, D.; Glomm, W.; Ryan,
J.; Anderson, M. F.; Franzen, S.; Feldheim, D. L. J. Am.
Chem. Soc. 2003, 125, 4700-4701.
35) Griessbach, R.; Meier, E.; Wassenegger, H. Farbenindustrie,
I. G. US 2,223,930, 1940.
(
(
(
10) Kim, Y.; Johnson, R. C.; Hupp, J. T. Nano Lett. 2001, 1, 165-
36) Steuerle, U.; Reuther, W.; Harder, W. AG, B. WO 97/21760,
167.
1997.
11) Dragnea, B.; Chen, C.; Kwak, E.-S.; Stein, B.; Kao, C. C. J.
Am. Chem. Soc. 2003, 125, 6374-6375.
37) Dick, C. R.; Potter, J. L.; Coker, W. P. Chemical, D. US
,565,941, 1968.
3
12) Donners, J. J. J. M.; Hoogenboom, R.; Schenning, A. P. H.
J.; van Hal, P. A.; Nolte, R. J. M.; Meijer, E. W.; Sommerdijk,
N. A. J. M. Langmuir 2002, 18, 2571-2576.
(
38) H o¨ lter, D.; Frey, H. Acta Polym. 1997, 48, 298-309.
39) Schmitzer, A.; Perez, E.; Rico-Lattes, I.; Lattes, A.; Rosca, S.
Langmuir 1999, 15, 4397-4403.
(
(
(
13) Lewis, L. N. Chem. Rev. 1993, 93, 2693-2730.
14) Kiwi, J.; Gr a¨ tzel, M. J. Am. Chem. Soc. 1979, 101, 7214-
(
40) Schmitzer, A. R.; Franceschi, S.; Perez, E.; Rico-Lattes, I.;
Lattes, A.; Thion, L.; Erard, M.; Vidal, C. J. Am. Chem. Soc.
7217.
2
001, 123, 5956-5961.
(
15) Sidorov, S. N.; Bronstein, L. M.; Valetsky, P. M.; Hartmann,
J.; C o¨ lfen, H.; Schnablegger, H.; Antonietti, M. J. Colloid
Interface Sci. 1999, 212, 197-211.
(
(
(
(
(
(
41) Yonezawa, Y.; Kawabata, I.; Sato, T. Ber. Bunsen-Ges. Phys.
Chem. 1996, 100, 39.
42) Zheng, J.; Stevenson, M. S.; Hikida, R. S.; van Patten, P. G.
J. Phys. Chem. B 2002, 106, 1252-1255.
(
16) Bronstein, L. M.; Sidorov, S. N.; Gourkova, A. Y.; Valetsky,
P. M.; Hartmann, J.; Breulmann, M.; C o¨ lfen, H.; Antonietti,
M. Inorg. Chim. Acta 1998, 280, 348-354.
43) Turkevich, T.; Stevenson, P. C.; Hillier, J. Discuss. Faraday
Soc. 1951, 11, 55-75.
44) Esumi, K.; Suzuki, A.; Yamahira, A.; Torigoe, K. Langmuir
(
(
(
(
17) Zhou, Y.; Wang, C. Y.; Zhu, Y. R.; Chen, Z. Y. Chem. Mater.
1999, 11, 2310-2312.
2
000, 16, 2604-2608.
18) Mandal, M.; Ghosh, S. K.; Kundu, S.; Esumi, K.; Pal, T.
Langmuir 2002, 18, 7792-7797.
45) Sipos, E.; Tungler, A.; Bitter, I. J. Mol. Catal. A 2003, 198,
67-173.
46) Wells, M.; Crooks, R. M. J. Am. Chem. Soc. 1996, 118,
988-3989.
1
19) Esumi, K.; Kameo, A.; Suzuki, A.; Torigoe, K. Colloids Surf.
A: Physiochem. Eng. Aspects 2001, 189, 155-161.
20) Esumi, K.; Torigoe, K. Prog. Colloid Polym. Sci. 2001, 117,
3
80-87.
MA0510791