Peptide-coated silver nanoparticles: Synthesis, surface chemistry, and ph-triggered, reversible assembly into particle assemblies

Article abstract of DOI:10.1002/chem.200802329

Simple tripeptides are scaffolds for the synthesis and further assembly of peptide/silver nanoparticle composites. Herein, we further explore peptide-controlled silver nanoparticle assembly processes. Silver nanoparticles with a pH-responsive peptide coat

Full text of DOI:10.1002/chem.200802329

FULL PAPER
DOI: 10.1002/chem.200802329
Peptide-Coated Silver Nanoparticles: Synthesis, Surface Chemistry, and pH-
Triggered, Reversible Assembly into Particle Assemblies
Philipp Graf,[a] Alexandre Mantion,*[a, e, f] Annette Foelske,[b] Andriy Shkilnyy,[c, d]
Admir Masˇic´,[d] Andreas F. Thꢀnemann,[e] and Andreas Taubert*[a, c, d]
Abstract: Simple tripeptides are scaf-
folds for the synthesis and further as-
sembly of peptide/silver nanoparticle
composites. Herein, we further explore
peptide-controlled silver nanoparticle
assembly processes. Silver nanoparti-
cles with a pH-responsive peptide coat-
ing have been synthesized by using a
one-step precipitation/coating route.
The nature of the peptide/silver inter-
action and the effect of the peptide on
the formation of the silver particles
have been studied via UV/Vis, X-ray
photoelectron, and surface-enhanced
Raman spectroscopies as well as
through electron microscopy, small
angle X-ray scattering and powder X-
ray diffraction with Rietveld refine-
ment. The particles reversibly form ag-
gregates of different sizes in aqueous
solution. The state of aggregation can
be controlled by the solution pH value.
At low pH values, individual particles
are present. At neutral pH values,
small clusters form and at high pH
values, large precipitates are observed.
Keywords: hybrid materials · nano-
particles · oligopeptides · pH · silver
Introduction
markers and sensors[2] or in antibiotic applications.[3,4] Now,
metal-nanoparticle synthesis is well established.[5,6] Electro-
magnetic radiation,[7–11] dimethylformamide (DMF),[12] ascor-
bate,[13–15] and citrate[16–19] are common reducing agents for
silver salts. Ascorbate and citrate have become popular for
their “green” connotation, but they also lead to uniform
nanoparticles,[20] including nanorods,[21] core–shell struc-
tures,[22,23] and polyamine-[24] and polysulfonate-stabi-
lized[14,25] nanoparticles. Current research focuses on factors
that control the shape, size, and size distribution of the par-
ticles by using ascorbate.[26,27] Reduction of silver ions with
DMF is less studied, but also yields well-defined silver nano-
particles.[12,28]
Metal nanoparticles, in particular silver nanoparticles, have
attracted attention because of their interesting optical prop-
erties.[1] Silver nanoparticles have also been studied as bio-
[a] P. Graf, Dr. A. Mantion, Prof. Dr. A. Taubert
Department of Chemistry, Klingelbergstrasse 80
University of Basel, 4056 Basel (Switzerland)
Fax : (+49)331-977-5773
E-mail: alexandre.mantion@bam.de
ataubert@uni-potsdam.de
For many applications, nanoparticle dispersions in water
are desirable. Among others, nanoparticle dispersions have
been stabilized with water-soluble polymers,[29–32] dendritic
stabilizers,[33,34] thiols,[35] cyclodextrins,[36] and genetically en-
gineered proteins.[37] Proteins are interesting stabilizers be-
cause they can combine a wide range of properties in one
molecule, such as communication with biology, stimuli re-
sponsiveness, and tailored interaction with a nanoparticle
surface. As a result, protein/nanoparticle hybrid structures
are potentially interesting for the fabrication of (biomimet-
ic) advanced functional materials. However, proteins are
often difficult to process, modify, and obtain in large quanti-
ties.
[b] Dr. A. Foelske
Electrochemistry Laboratory, Paul Scherrer Institute
5232 Villigen (Switzerland)
[c] Dr. A. Shkilnyy, Prof. Dr. A. Taubert
Institute of Chemistry, University of Potsdam
14476 Golm (Germany)
ˇ ´
[d] Dr. A. Shkilnyy, Dr. A. Masic, Prof. Dr. A. Taubert
Max-Planck-Institute of Colloids and Interfaces
14476 Golm (Germany)
[e] Dr. A. Mantion, Prof. Dr. A. F. Thꢀnemann
BAM Federal Institute for Materials Research and Testing
Richard-Willstaetter-Strasse 11, 12489 Berlin (Germany)
[f] Dr. A. Mantion
Current address:
BAM Federal Institute for Materials Research and Testing
Richard-Willstaetter-Strasse 11, 12489 Berlin (Germany)
Fax : (+49)30-8104-1137
Peptides can be viewed as simple variants of proteins. De-
pending on the amino acid sequence, they can be acidic,
Chem. Eur. J. 2009, 15, 5831 – 5844
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5831

basic, hydrophilic, or hydrophobic. In contrast to proteins,
they are usually available on the gram scale. As a result,
peptides are attractive building blocks for advanced materi-
als.[38–40]
Results
Pure peptide: Figure 1 shows representative IR spectrosco-
py data of the pure peptide 4. IR spectra show overlapping
Despite these interesting properties, however, little is
known on the synthesis and properties of metal nanoparti-
cle/peptide hybrids. Naik et al. have described the growth of
nano-objects in the presence of small peptides.[41] The pep-
tides are often based on l-tryptophane[42,43] and l-tyrosine[44]
because of their propensity to form organogels and to be
redox active.
Similarly, little is known on how peptides control the as-
sembly of nanoparticles. The assembly of thiol- or polymer-
stabilized metal nanoparticles is rather well understood,[45–50]
but there are comparatively few data on how peptides con-
trol the self-assembly of metal particles. Correspondingly,
little is known on how the peptides affect the properties of
the resulting peptide/inorganic hybrid material.[51–54] Under-
standing and controlling the response of hybrid nanoparti-
cles and being able to regulate their assembly is, however,
crucial for the design of novel materials with specific proper-
ties such as a well-defined optical response.
Figure 1. Deconvoluted IR spectrum (diagnostic region) of the pure pep-
tide. The sum fit overlaps with the experimental curve.
As a result, there is a need to further study the synthesis,
structure, and behavior of peptide-functionalized metal
nanoparticles as building blocks for complex materials. The
current study presents an in-depth investigation of a pep-
tide/metal nanoparticle model system. The model system
consists of silver nanoparticles coated with a pH-responsive
hexapeptide (Scheme 1). The goal of the study is to 1) deter-
mine the exact nature of the silver nanoparticle surface,
2) the type of interaction between the silver nanoparticle
surface and the peptide, and 3) the self-assembly behavior
versus the pH of the nanoparticle dispersion.
bands in the region between 1550 and 1800 cmꢀ1, which is
the diagnostic region for the secondary structure in peptides.
Peak deconvolution indicates amide I bands nACTHNUTRGNEUG(N C=O) at
1663 cmꢀ1 and 1692 cmꢀ1 and an amide II band at 1550 cmꢀ1.
IR suggests that in the solid state (the powder), the peptide
adopts a 310 helix[55–57] as suggested by the amide I and am-
AHCTUNGTRENNUNG
ide II bands. A b-sheet turn-type secondary structure[58]
cannot be excluded, but the signals around 1631 cmꢀ1 are
[59]
ꢀ
obscured by the presence of n
G
tected by X-ray diffraction or 2D IR spectroscopy. Besides
ꢀ
the amide I and II bands, symmetric and asymmetric n
(N
H) bending bands at 1630 cmꢀ1 and 1524 cmꢀ1, a trifluoro-
acetic acid (TFA) carboxylate nACHTUNGTRNEUNG
(C=O) band at 1597 cmꢀ1,
carboxylic acid bands (terminal carboxylic acid) at
1716 cmꢀ1 and 1782 cmꢀ1,[59] are also observed. A shoulder
at 3280 cmꢀ1 is due to amide A bands and indicates the pres-
ence of a b sheet.[58]
Figure 2 shows an X-ray diffraction (XRD) pattern of the
peptide powder. It shows two peaks at 20.118 (d spacing of
4.41 ꢂ) and at 55.128 (d spacing of 1.66 ꢂ). The first peak is
broad. It is attributed to a b-sheet motive or, more general-
ly, an intermolecular b turn.[58] The second peak is consistent
with an intramolecular motive of a 310 structure.[60] XRD
thus supports IR spectroscopy in that in both cases, b-turn
motives are found.
Peptide-modified silver particles: Sample P1 was prepared
at pH 3, P2 at pH 7, and P3 at pH 9 (see the Experimental
Section for details). Figure 3 shows a representative XRD
pattern of P1. The reflections can be assigned to cubic face-
centered metallic silver (Joint Committee of Powder Dif-
fraction Standards (JCPDS) file 04–0783). The reflections
are broad, indicating the presence of small particles. Riet-
veld refinement (Table 1) further confirms this as for all hkl
Scheme 1. Peptide 4 used in this study. All amino acids are l-amino acids.
The neutral form of the peptide is shown. The peptide sequence is Lys-
Lys-Cys (KKC). The two sequences are connected through disulfide
ꢀ
bridge (-S S-). Peptide 4 was prepared from precursors 1, 2, and 3 (see
the Experimental Section).
5832
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5831 – 5844

Assembly of Peptide-Coated Silver Nanoparticles
FULL PAPER
to silver. Comparison with Figure 2 shows that this reflec-
tion can be assigned to the peptide.
Table 2 shows the shift of the (311) and (420) reflections
compared with the particle preparation conditions with re-
spect to their position in bulk, single crystalline silver
Table 2. Shift of (420) and (311) position in 2q [8] as a function of the
synthesis pH value.[a]
Reflection
(331)
(420)
P1
0.28
ꢀ0.31
P2
P3
G
0.48
0.02
ꢀ0.07
ꢀ0.12
AHCTUNGTRENNUNG
[a] Shifts are given with respect to the bulk silver structure (JCPDS 04–
0783).
Figure 2. XRD pattern of the pure peptide.
(JCPDS 04–0783). Shifts can be up to 0.58 2q, as revealed
by Rietveld refinement. These shifts are rather large. The
shifts are larger for (331) than (420). Overall, the shifts are
most prominent for P1, which suggests that the crystal struc-
ture of P1 is the least close to bulk silver. This is interesting,
because the crystallite sizes (Table 1) of P1 and P2 are es-
sentially the same. As a result, this suggests that, possibly,
there are surface effects contributing to a more distorted
structure in P1. A more detailed discussion of this behavior
is, however, complicated by the fact that the shifts in Table 2
are also affected by the amount of ligand on the surface and
by the fact that there is also evidence for residual ascorbate
on the particle surface (see below).
The crystal structure of P3 is more closely related to bulk
silver than the others, as the shifts are rather limited. By
using the microfacet[61] and step notation, the (331) reflec-
Figure 3. XRD pattern of P1 and corresponding Rietveld refinement.
Arrow denotes the extra reflection at 54.948 as mentioned in the text.
tion can be rewritten as 3
A
a three-atom (111) terrace and one-atom (11ꢀ1) step in
face-centered cubic crystal structures. Similarly, the (420) re-
flection can be rewritten as Ag(S)-2ꢃ(100)+1ꢃ(111)+1ꢃ
(11ꢀ1), which corresponds to a kinked terrace. Both are in-
dicating that high index lattice planes interact with the pep-
tide disulfide bond or a thiol generated in situ. A similar be-
havior has been reported for gold nanoparticles that consist
of 102 gold atoms and 44 p-mercaptobenzoic acid mole-
cules.[62] Furthermore, Gaultier and Buergi[63] showed that
chiral inversion of CD spectra of (thiol-modified) gold nano-
particles only happens if the gold crystal structure adapts so
that it can present the ligand for maximal chemical interac-
tion, namely adopts the chirality of the substrate. As a
result, our data suggest that already a very simple peptide
can significantly shift the fine structure of the crystal at its
interface.
Table 1. Crystallite sizes from Rietveld analysis.[a]
A
P1
P2
P3
A
55
54
73
43
76
85
74
43
78
80
A
30
31
A
55
50
A
61
58
A
55
54
A
30
31
A
57
53
A
57
54
average size
(anisotropy)
50 (11.44)
48 (10.20)
69 (15.44)
[a] Sizes are in ꢂngstroms.
values, particle sizes below 15 nm are found. The ratio of
the intensities (200)/(111) and (220)/(111) of our particles is
A
ACHTUNGTRENNUNG
between 0.35 and 0.44 or 0.22 and 0.26, respectively. This is
roughly consistent with the JCPDS values, which are 0.40
and 0.25, respectively. This suggests that the particles are ap-
proximately spherical, but may be distorted somewhat. The
bump in the residual curve of the Rietveld refinement be-
tween 388 and 458 is presumably due to an amorphous con-
tribution of the peptide on the particle surface. Finally,
there is also an extra reflection at 54.948, which is not due
Aside from the silver reflections, XRD (Figure 3) also
shows an additional peak at 54.948 (d spacing=1.66 ꢂ). This
reflection is due to the helical pitch in the 310 structure that
is adopted by the peptide present on the particle. Interest-
ingly, XRD suggests that the b-sheet component is lost
when the peptide is adsorbed on the metallic surface. This
can be concluded from the loss of the broad peak that is ob-
served in the neat peptide at 20.118 (Figure 2).
Chem. Eur. J. 2009, 15, 5831 – 5844
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5833

A. Mantion, A. Taubert et al.
Figure 4 shows representative transmission electron mi-
croscopy (TEM) images of the particles. Sample P1 (pre-
pared at pH 3) and P2 (prepared at pH 7) are essentially
identical, that is, in both cases roughly spherical particles
Although XRD finds sizes around 2 to 5 nm, TEM finds
particle sizes in the range of approximately 10–50 nm. Fur-
thermore, the reconstructed primary particle shapes (crystal-
lite shapes, see insets in Figure 4) only match the shapes of
P1 and P2, but not P3.
Figure 5 shows representative IR spectra of the pure pep-
tide and P1. The spectra of P1, P2, and P3 have very simple
Figure 5. IR spectra of neat peptide 4 as a TFA salt and P1, respectively.
The spectra for P1, P2, and P3 are virtually indistinguishable.
structures when compared with the original peptide. The
only visible bands in the carbonyl region are at 1735 cmꢀ1
and 1212 cmꢀ1. They can be assigned to n
G
N
ꢀ
O), respectively, from residual ascorbic acid adsorbed on the
surface, remains of oxidized ascorbic acid, or the peptide
carboxylic acid. Other vibrations are due to the aliphatic
chain of the peptide and, again, some residues of the ascor-
ꢀ
bic acid and its oxidized form, mainly d
N
(C H), and
ꢀ1
n
(N H) at around 1200 to 1300 cm and 3000 cmꢀ1, respec-
ꢀ
tively.
The lack of further signals is not surprising as the silver
surface interferes with the IR spectra of bound molecules.
The absence of further signals indicates that the peptide is
close to the particle surface and the dipolar transition mo-
ments are parallel to the surface for the amide-bond transi-
tion. This in turn suggests that the amide bonds are parallel
to the particle surface, which indicates that the peptides
form a brush-like structure on the particle surface. They are
thus not able to interact with the electric field present at the
surface, which is perpendicular to the silver nanoparticle sur-
Figure 4. TEM images of a) P1 prepared at pH 3, b) P2 prepared at pH 7,
and c) P3 prepared at pH 9. Insets are primary particle shapes (oblates)
generated from XRD and Rietveld data via GFourier.[64] The scale bar
applies to all images.
with diameters of 18.9ꢁ3.9 nm (P1) and 15.4ꢁ5.0 nm (P2)
form. Both samples contain a few particles at the larger end
of the size distribution that appear triangle-like. Unlike P1
and P2, P3 (prepared at pH 9) contains larger particles with
a variety of shapes, including rod-like or elongated particles.
These findings are different from results by Sondi et al.,[14]
who showed that the higher the acidity of the starting silver
solution, the bigger the particles are. However, as their pro-
tocol does not involve peptides, it is difficult to draw further
conclusions.
face.[28] The n
ACTHNUTRGENNUG(C S) and nACHTUNGTERN(NUGN S S) bands cannot be detected
ꢀ
ꢀ
because of the low-lying frequency of these bands and their
relative weakness for infrared spectroscopy. Therefore, the
chemical nature of the sulfur species cannot be determined.
Nevertheless, IR suggests the presence of the peptide on the
surface of the particle.
Figure 6 shows surface-enhanced Raman spectroscopy
(SERS) data of P1. All SERS bands are relatively broad,
which indicates that the peptides interact strongly with the
surface. SERS detects two strong bands at 480 and 670 cmꢀ1.
ꢀ
Comparison of the XRD and TEM data suggests that, in
spite of their small sizes, the particles (P3 in particular) are
not single crystals but consist of a few smaller crystallites.
The former can be attributed to nACTHUNTGRNEUNG(S S) in the l-cystine di-
5834
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5831 – 5844

Assembly of Peptide-Coated Silver Nanoparticles
FULL PAPER
Figure 6. Surface-enhanced Raman spectrum of P1. Inset: Magnification
of the 200 cmꢀ1–500 cmꢀ1 region. Arrows indicate bands described in the
text.
[65]
ꢀ
sulfide and the latter to n
G
shows that both the original dimeric peptide 4 (the disul-
fide) and the cleaved species (the thiol) are simultaneously
present on the silver-particle surface. However, much like in
our earlier study,[28] the weak signal at 240 cmꢀ1 is a superpo-
As a result, SERS
cannot unambiguously identify the nature of the Ag S bond
[66]
ꢀ
sition of n
N
and n
N
on the metal surface.
The band at 356 cmꢀ1 can be attributed to a n
band, which is indicative of a partially oxidized silver sur-
face.[67] Peptide bands from 1169 to 1651 cmꢀ1 are indicative
of a variety of structural orientations that are present at the
Figure 7. XPS spectra of the Ag and S edges of P2 (excitation source:
MgKa).
Table 3. Surface composition of P1 to P3 determined by using the mono-
source.[a]
ꢀ
ꢀ
surface. They can be attributed to dACTHNUTRGENNGU(C H) and nACHUTNGTRENN(UGN N H) of
the peptide, similar to the assignments given for the decon-
voluted IR spectrum in Figure 1.
Signal
P1
P2
P3
Ag 3d5/2
C 1s
N 1s
23.5
61.7
5.4
37.9
42.6
6.1
46.2
31.0
6.4
There is no clear evidence of the 310 structure because the
ꢀ
amide I and amide II bands are superimposed with the nACTHNUTRGNEUNG(N
O 1s
9.4
13.4
16.4
H), and a clear interpretation of the amide bond orientation
at the metal surface with this spectral resolution is not possi-
ble. However, the signal at around 2950 cmꢀ1 can be attrib-
ꢀ
S 2p
G
G
ACHTUNGTRENNUNG(1.0)
[a] The beam of the monosource has been directed on a rectangular
powder rich surface area of 800 mmꢃ300 mm and therefore a contribution
from the sample support (flexible graphite) is almost excluded. The
sulfur content (in brackets) has been determined from measurements
with the twin anode and is given with respect to all other species detected
from an average surface area of 0.8 cm2. All values are given in at%.
uted to amide A vibrations and nACTHUNTGRNEUNG(C H) present in the pep-
tide. In summary, SERS measurements show that the pep-
tide is connected to the nanoparticle surface and that both
disulfide and thiol groups are bound to the silver surface.
Moreover, SERS also shows that the silver is partially oxi-
ꢀ
dized as we also detect a n
(Ag O) band.
oxygen (see Table 3) at the silver nanoparticle surface indi-
cates the presence of oxidized silver, as Ag2O, and/or that
some reducing agent is also adsorbed on the nanoparticle
surface. Table 3 summarizes the surface composition of the
three samples. It clearly shows that the metal/carbon ratio
increases from P1 to P3.
Figure 8 shows that from P1 to P3, the amount of carbon
species at higher binding energies increases. This indicates a
change in the organic layer towards more-oxidized carbon
species from P1 to P3. At the same time, the oxygen content
increases from 9.4 to 16.4 at% and the oxygen to nitrogen
ratio increases as well. XPS therefore suggests that particles
prepared at pH 9 are partly still covered with ascorbate or
Figure 7 shows representative X-ray photoelectron spec-
troscopy (XPS) spectra of the particles. For all samples, the
Ag 3d5/2 signals exhibit broad peaks with full widths at half
maximum (FWHM) of around 1.3 eV by using the twin
anode and 0.8 eV by using the monsource. The silver 3d5/2
and 3d3/2 are located at binding energies of 368.4 eV and
374.3 eV, respectively. These findings correlate with former
studies of similar systems.[28,68] The broadened shape is due
to the variety of conformations of the peptide on the parti-
cle, and a further contribution arises from the silver particle
size.[68–71] This silver binding energy is representative of an
oxidized metallic silver surface.[72–74] The high amount of
Chem. Eur. J. 2009, 15, 5831 – 5844
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5835

A. Mantion, A. Taubert et al.
values rather than at basic pH values. Overall, the bond
types (dithiol, thiol, sulfoxide, etc.) between the silver parti-
cle and the peptide are most uniform in sample P2. That is,
in P2 we predominantly observe thiol molecules.
Figure 9 shows titration data for the peptide and P1 by
using hydrochloric acid. The titration curve of the pure pep-
Figure 8. Energy evolution of the C 1s edge from P1 to P3 (excitation:
monochromatic AlKa). B.E.=binding energy.
dehydroascorbate rather than with peptide. As a result, the
P3 silver particles that were prepared are less efficiently
capped by the peptide and expected to be less stable.
Table 4 shows that depending on the preparation pH, dif-
ferent signals are observed for sulfur S2p in XPS. The signal
Table 4. Peak deconvolution of the S 2p peak.
Sample
Bonding energy of sulfur S 2p peaks [eV]
P1
P2
P3
161.0; 161.7; 163.8; 167.7; 169.4
161.6; 163.9, 165.4, 167.95, 169.7
161.3; 162.0; 164.9; 167.0; 168.5; 170.5
at 161.7 eV is indicative of a thiol at the silver surface. Be-
cause of the high reactivity of thioethers and dithiols to-
wards silver surfaces, the presence of a thiol is not surpris-
ing.[65,75] The signal at 163.9 eV is indicative of a cystine
bridge chemisorbed on the silver surface. Higher energies
are indicative of sulfoxide and cystine disulfoxide, probably
from air exposure of the nanoparticles.[76] XPS spectra of P1
exhibit an additional signal at 161 eV caused by sulfide
S2ꢀ.[77] This may indicate a different mechanism of thiol
cleavage, possibly through the direct cystine bridge reduc-
Figure 9. a) pH as a function of titrant added. b) Granꢄs treatment of ti-
tration data.[78]
tide is typical for molecules with two different, pH-respon-
sive functional groups. The first equivalence point at a pH
value of 5.1 corresponds to the deprotonation of the carbox-
ylic acid groups at the C terminus of the peptide. The
second equivalence point at a pH value of 8.5 is attributed
to the deprotonation of the side chain ammonium groups.
The titration curve of the nanoparticles only exhibits one
equivalence point around pH 7. This can be explained by
the fact that the titration was performed by using an excess
of hydrochloric acid, that is, the peptide was back-titrated.
As the concentration of peptide is low compared with the
HCl, essentially only the titration of the excess HCl is visi-
ble. To obtain the same titration curve as for the pure pep-
tide, the particle concentration would need to be 20.83 times
larger. This corresponds to 4.17 g of modified SNP/40 mL of
solution, which is not possible owing to solubility problems.
Figure 9 also shows the results of Granꢄs analysis[78] that
includes strong-acid, weak-acid or -base, and strong-base
fractions. The difference between the strong-base and
strong-acid fractions gives the amount of weakly acidic or
[65,75]
ꢀ
tion and cleavage of a C S bond,
which is similar to
one of our previous studies.[28]
Overall, XPS confirms SERS in the sense that we also
detect a variety of chemical species including a mixture of
disulfides and thiols on the silver surface and some oxida-
tion of the particle surface. In addition, P1 shows a non-per-
fect disulfide bridge cleavage where presumably a significant
ꢀ
fraction of the C S bonds is broken. P2 particles do not
show the signal corresponding to the S2ꢀ binding energy.
They do, however, show about the same degree of silver oxi-
dation (as shown by similar amount of oxygen on the sur-
face) as P1. P3 exhibits high-energy sulfur peaks that suggest
the presence of a large fraction of oxidized sulfur species. In
summary, XPS clearly demonstrates the presence of peptide
on the metal surface and also suggests that the peptide
covers the particle more efficiently at neutral and acidic pH
5836
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5831 – 5844

Assembly of Peptide-Coated Silver Nanoparticles
FULL PAPER
basic groups and allows determination of the concentration
of peptide on the particle surface, which in the current case
is 57.7 mmolgꢀ1. This corresponds to (on average) 1300 pep-
tides per silver nanoparticle, assuming that the particle di-
ameter of the silver core is 18.9 nm (average particle size,
from TEM) and the volume of one silver nanoparticle is
94.4 nm3 (1Ag =10.59 gcmꢀ3). It is assumed that the peptide
coverage is complete, that is, there is only peptide on the
surface.
Figure 10 shows thermogravimetric analysis (TGA) data
of P1 to P3. TGA curves show three distinct weight losses,
1718C, 2488C, and 3708C. These can be assigned to the suc-
Figure 11. UV/Vis spectra of P1, P2, and P3 at different pH values. All
spectra were recorded in solutions with a nanoparticle concentration of
0.2 mgmLꢀ1. The bands below 300 nm are from the peptide.
Figure 10. TGA curves of P1, P2, and P3.
cessive thermal destruction of the peptide layer on the parti-
cles. However, a precise mechanism of decomposition could
not be formulated, as mass spectrometry data were not
available. Above approximately 4008C, silver reacts with
oxygen, which causes the slight increase in mass observed in
all cases at higher temperatures. P1 contains approximately
3.7% of organic, P2 contains around 2.8%, and P3 only con-
tains approximately 1.4%. In the case of P1, this represents
around 1200 peptides per particle, which corresponds well
with the findings from the titration experiments. TGA thus
shows that the pH value of preparation has some influence
on the amount of organic compound present on the surface.
TGA is also consistent with titration data in which a maxi-
mum peptide fraction of 5% for P1 was found. However,
direct comparisons of P1 and P2 with P3 are not possible be-
cause of their difference in size.
Figure 12. Representative deconvolution of a UV/Vis spectrum of P1 at
pH 3. The sum fit overlaps with the experimental data and can only
barely be seen at the top of the experimental peak at around 400 nm.
Responsive behavior and aggregation of peptide-modified
silver particles: Metal nanoparticle shapes and assembly in
solution can easily be characterized by using UV/Vis spec-
troscopy.[79–81] Figure 11 shows typical UV/Vis spectra of P1,
P2, and P3 in aqueous solutions of different pH. Figure 12
shows a representative example of a deconvoluted UV/Vis
spectrum and Table 5 summarizes the deconvolution data
for P1.
Table 5. UV/Vis spectral deconvolution of P1.[a]
Peak maximum [nm]
pH 9
pH 7
pH 3
347, 370, 405, 477, 625, 754
350, 372, 402, 463, 575, 696
351, 376, 414, 470
[a] Absorption bands were modeled by using Gaussian curves.
Figure 11 shows that all bands in the UV/Vis spectra are
broad and often exhibit shoulders caused by overlapping
bands. This indicates that 1) the particle-size distribution is
broad and 2) that in some cases, not only spherical particles
Chem. Eur. J. 2009, 15, 5831 – 5844
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5837

A. Mantion, A. Taubert et al.
are present in the sample. Bands at around 350, 370, 400,
and 470 nm have been assigned to scattering and plasmon
resonance of single particles.[80,81] Bands around 500 nm
have been assigned to bent or distorted particles,[80] bands at
around 600 nm are due to triangular nanoparticles or to
nanoparticle aggregates,[80] and bands at 700 nm and longer
have been assigned to aggregates, raspberry-like shapes, or
silver nanowires.[81] Only P3 shows a significant absorption
at and above 700 nm, which is assigned to the various
shapes observed in this particular sample (Figure 4).
Besides a preliminary shape assignment, UV spectroscopy
provides a means to study particle aggregation in solution.
P1 and P2 show a clear shift of the absorption maximum
from approximately 400 nm to 700 and 750 nm, respectively,
as the pH value changes from 3 to 11. As the particle shapes
do not change, this is a clear indication of a stepwise aggre-
gation. UV/Vis spectroscopy suggests that at each pH a spe-
cific type of aggregation occurs, as a different but broad ab-
sorption band is always detected. Furthermore, as the sur-
face plasmon band of individual particles at 400 nm does
not completely disappear, UV/Vis spectroscopy also sug-
gests that in all cases some particles remain as isolated
spheres. Moreover, the surface chemistry (and thus peptide
loading) affects the shapes of the plasmon bands. For exam-
ple, P1 at pH 7 is not equivalent to P2 at pH 7. However, on
a qualitative level, the behavior is correlated in the sense
that P1 and P2 (but not P3) follow the same trends. For sim-
plicity, only the behavior of P1 particles, which have been
shown to be the most uniform sample, will be discussed.
Figure 13 shows representative TEM images of samples
taken at different pH. TEM confirms UV/Vis spectroscopy
and shows that at pH 3, essentially all particles are present
as isolated spheres and that virtually no aggregation occurs.
At pH 7, smaller aggregates with 10 to 30 particles per ag-
gregate are found, although there are also some isolated
particles. At pH 11, virtually all particles have assembled
into large structures and only very few isolated particles can
be observed. Both TEM and UV/Vis spectra show that the
process is reversible and that a lowering of the pH again
leads to individual particles.
To more precisely describe the size of the aggregates and
to rule out drying artifacts in the TEM images, the particles
were analyzed by using dynamic light scattering (DLS). Al-
though the analysis of the DLS data cannot be correlated di-
rectly to the TEM images owing to charge and absorption
effects, especially at low pH in which the particles are highly
charged, DLS provides evidence of aggregation that is con-
sistent with TEM and UV/Vis spectra.
Table 6 shows the size of the aggregates versus the solu-
tion pH value. Clearly, an increase in the pH value leads to
an increase in the hydrodynamic diameter dh of the aggre-
gates. The transition between the different sizes (i.e. pH) is
not abrupt. For example, at pH 6.5, there are at least two
populations with significant intensity. This finding further
supports the TEM observation of the presence of multiple
types of aggregates; it also suggests a dynamic growth pro-
cess.
Figure 13. TEM images of P1 at a) pH 3, b) pH 7, and c) pH 11. Scale bar
applies to all images.
Table 6. Aggregate dimensions extracted from DLS.[a]
pH
dh
[%]
dh
[%]
2.3
6.3
7.3
8.4
10.3
23.2ꢁ13.0
22.5ꢁ2.5
100
55
55
100
100
60.2ꢁ14.4
45
45
60.8ꢁ11.2
684.0ꢁ77.0
874.7ꢁ108.3
352.6ꢁ83.5
[a] Volume weighted data are given, the hydrodynamic diameter dh is in
nm
A simple calculation taking an average particle size of
18.9 nm (from TEM, Figure 4) indicates that at pH 2.3, the
particles are present as monomers and possibly dimers, at
pH 6.3 the particles are present as monomers (possibly
dimers) and, predominantly, tetramers. At pH 10.3, the ag-
gregates are mainly approximately 50-mers, but some very
big aggregates (>300-mers) are also present. Larger aggre-
gates may sediment; thus the size distribution at pH 10.3 is
approximate. In summary, even with a simple analysis, DLS
shows that aggregation takes place in solution, and this pro-
5838
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5831 – 5844

Assembly of Peptide-Coated Silver Nanoparticles
FULL PAPER
cess is pH dependent. DLS thus supports TEM and UV/Vis
experiments (Figure 11 to Figure 13).
Small-angle X-ray scattering (SAXS) further supports
DLS and provides both size-distribution quantification and
an interaction-potential estimation. Figure 14 shows typical
Figure 15. SAXS intensity of P1 at pH 2.28 and a model curve of polydis-
perse spheres with Schulz distribution fitted to the data beyond q=
0.3 nmꢀ1 (thick arrow). The inset displays the frequency of the radii of
the Schulz distribution corresponding to the SAXS fit curve.
Figure 14. SAXS intensities of P1 at pH 2.28 (a), 6.30 (b), and 10.30 (c).
Arrows indicate regions where interparticle interactions affect the scat-
tering curves (q<0.3 nmꢀ1) and where interactions can be neglected (q>
0.3 nmꢀ1). The Porod region (I(q)1qꢀ4) is indicated by a straight line. (b)
and (c) are vertically shifted by a factor of 102 and 104, respectively, for
clarity. SAXS data were obtained at a concentration of 1% (w/w) at a
pH value between 2.3 to 10.3.
SAXS curves. The shape of all curves is identical at q values
larger than 0.3 nmꢀ1 that display a characteristic minimum
around 0.4 nmꢀ1 and a Porod behavior (I(q)1qꢀ4) for q>
0.6 nmꢀ1. These characteristics indicate both a moderate
polydispersity and a sharp interface between nanoparticles
and their surroundings. A fractal particle surface can there-
fore be excluded and the peptide shell is practically invisible
for SAXS. The increasing intensity and steepness of the
SAXS curve in the lowest q region shows that attractive in-
terparticle interactions increase with increasing pH values.
As a result, the form factor of P1 cannot be derived from a
whole curve fitting by using a model function form factor,
but the observation of increasing attraction with increasing
pH is consistent with DLS and TEM (Figure 13 and
Table 6).
Figure 16. Structure factors of P1 at pH 2.28 (a), 6.30 (b), and 10.30 (c).
Inset: Experimental structure factor at pH 2.28 (thin line) and approxi-
mation (thick line) with a square well potential with a depth of U0 =5 kT
and a width of l=0.75.
Experimental estimation of S(q) by using a square well
model (see the Experimental Section) resulted in interaction
potentials of 5, 10, and 12 kT (12.3, 24.6, and 29.5 kJmolꢀ1
at 293 K) for pH values of 2.28, 6.30, and 10.30, respectively.
As a result, Figure 16 shows an increasing interparticle inter-
action with increasing pH values. This again confirms that
the particle stickiness is pH dependent, although the as-
sumption of a simple square well potential can certainly be
optimized.
The scattering intensities at high q values (0.3 nmꢀ1 <q<
4 nmꢀ1) were used to determine the radius and polydispers-
ity of the particles (Figure 15). In this range, particle interac-
tions are negligible. The mean radius of P1 is 10.1ꢁ0.1 nm
and the polydispersity is 0.18ꢁ0.01.These data are in good
agreement with TEM and DLS (Figure 13 and Table 6).
Figure 15 also shows that there is a deviation between the
experimental SAXS curve and the theoretical curve of non-
interacting particles in the low q region below 0.3 nmꢀ1. This
deviation can be assigned particle–particle interaction ef-
fects. Quantitatively, the interaction is expressed in terms of
the structure factor S(q)=I(q)/P(q), where I(q) is the inten-
sity and P(q) is the particle form factor (Figure 16).
Discussion
Particle synthesis and peptide coating: It has been pointed
out previously that the formation of metal nanoparticles is
strongly affected by the pH value of the solution.[13–15,25] It is
also well known that thiols are efficient growth modifiers
for metal nanoparticles.[82,83] In the current study, we have
Chem. Eur. J. 2009, 15, 5831 – 5844
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5839

A. Mantion, A. Taubert et al.
used a peptide, which at the same time acts as a template
for the formation of well-defined (P1 and P2) nanoparticles
and also imparts pH-responsiveness to the silver particles.
The effect on the particle growth is rather strong because
both the cystine and the lysine moieties of our peptide can
interact with AgI.[84] Therefore, the size and shape of the re-
sulting particles are a result of the combined influence of
the pH value and peptide chemistry.
tion of the disulfide and an accumulation of silver atoms in
the vicinity of the peptide could contribute to the resulting
particle sizes, shapes, and distortions. Further studies into
this topic are, however, necessary to elucidate the growth
mechanism and specific interactions of the peptide with the
growing particles in more detail.
XRD (Table 1) also shows the effect of the peptide on the
crystal structure of the silver particles. Similar to a few earli-
er studies[62,63,85–87] in which small molecules were investigat-
ed, the current study reveals that also peptides influence the
crystallographic structure of the SNPs in a specific fashion.
TEM (Figure 3) shows that for the controlled formation
of uniform particles, protonated lysine residues (P1 and P2,
pH 3 and 7, respectively) are favorable. In this case, the
lysine residues interact less strongly with the growing parti-
cle than if they are not protonated. The sulfur atoms pro-
vide the dominant interaction with the particle. The pres-
ence of negatively charged carboxylate groups and neutral
amines (P3, pH 9) is less favorable, as large, irregularly
shaped particles form (Figure 4). This is intriguing because
intuitively, one would expect that with the combined adsorp-
tion of the disulfide and the free amine groups onto growing
Ag particles there should be a more efficient interaction be-
tween the growing particle and the peptide and the particles
should therefore be smaller. However, TEM and UV/Vis
spectroscopy clearly show that positively charged ammoni-
um groups on the lysine side chains are more favorable. We
hypothesize that with too many adsorbing groups (four
amine groups and one disulfide per peptide, 1200 to 1300
peptides per particle) there are too many competing interac-
tions, which overall prevent an ordered particle growth.
TGA, titration, XPS, IR, and SERS show that the pep-
tides are stably bound to the particle surfaces. XPS and
SERS also show that 1) some of the peptide is present as a
disulfide, 2) some is present as a thiol, and 3) that the pep-
tides most likely form a brush on the particle surfaces.
There is also evidence for Ag oxidation, which is similar to
our earlier study.[28]
Reversible particle aggregation: TEM, DLS, and UV/Vis
spectroscopy (Figures 11–13, Table 5, and Table 6) show that
the aggregation of the peptide-modified nanoparticles can
be controlled by variation of the pH value and that particle
aggregates with discrete sizes (although a broad size distri-
bution) form as a function of the solution pH value. At
pH 3, the lysine side chains and the C-terminal carboxylic
acid groups (Scheme 1) are protonated, which leads to a
positively charged shell. As a result of the high positive
charge, particles exist as individual species and no aggrega-
tion occurs. At pH 7, the C-terminal carboxylic acid is de-
protonated and the lysine residues are still protonated.
Owing to the resulting zwitterionic particle surface, some
aggregation occurs and small clusters of particles are ob-
served in the TEM (Figure 13, Table 6).
At pH 11, the C-terminal carboxylate (Scheme 1) is de-
protonated and carries a negative charge, but the lysine resi-
dues are not charged anymore. At this pH value, the parti-
cles assemble into large aggregates. Presumably, the repul-
sive forces exerted by the single terminal carboxylate groups
are not sufficient to prevent aggregation. Furthermore, as
the deprotonated lysine is less hydrophilic, hydrophobic in-
teractions and hydrogen bonding can occur, leading to the
large aggregates.
XRD suggests that the peptide is not, like regular thiols,
simply sitting on the surface of the nanoparticles, but (at
least on the dry particles) adopts a certain periodic order.
This is clearly evidenced by the additional peak observed in
the XRD patterns of the particles (Figure 3). Moreover,
XRD shows that the peptide induces additional effects in
the silver nanoparticles, as seen by the shifts of the (311)
and (402) reflections (Table 2). Although the particle
growth mechanism has not been investigated yet, XRD
shows that the peptide acts as a simple and very mild cap-
ping agent, which is in contrast with the classically used
thiols of the Brust–Schiffrin technique, which traps gold
nanoparticles at early grow stages.[45,46] It is possible that the
peptide used in the current study acts in a similar manner to
a growth mechanism suggested by Naik et al.[41] These au-
thors suggest that particles can grow by accumulation of
silver atoms, which is mediated by the presence of oligopep-
tides. Typically, in these cases, the lowest energy faces are
favored and particles with large (111) faces grow. Although
this is not the case in the current study, we do observe a
slight crystallographic anisotropy along (200) and (400)
(Table 1). This suggests that both effects, the strong interac-
SAXS qualitatively supports TEM, DLS, and UV/Vis
spectroscopy in that an approximate interaction potential
can be estimated, which increases with increasing pH value.
Because the interaction potential is attractive at all pH
values, it supports the observation that the particles can
form small, weakly bound aggregates, which reversibly form
and disassemble with variation of the pH value.
Potential of the peptide-functionalized particles: The aggre-
gation process is reversible and can easily be controlled by
the pH value. Furthermore, the particles contain further
(chiral) information that is not lost in the solid state (XRD,
Figure 3). Our approach therefore enables the preparation
of silver nano-objects that can be viewed as building blocks
for biomimetic or bioinspired hybrid materials, which can be
reversibly assembled and disassembled. They are therefore
interesting for materials where, for example, the optical
properties need to change upon an external stimulus or
where a response upon a change in a biological environment
is desired.
5840
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5831 – 5844

Assembly of Peptide-Coated Silver Nanoparticles
FULL PAPER
Our model system is just one example of a stimulus-re-
sponsive material. Modification of the peptide sequence and
the metal core will lead to a wide variety of other particles.
For example, one can envision materials that are composed
of two or more components, such as peptide-coated gold
and silver particles. The resulting materials should have in-
teresting optical properties, in particular, because the pep-
tide coating provides additional chiral information.
On the experimental side, disulfides are convenient be-
cause they avoid the additional steps of thiol protection and
deprotection without losing efficiency in the peptide attach-
ment. This work therefore extends our previous study in
which it has been shown that thioether- (and not disulfide)
bearing peptides are efficient growth modifiers for SNPs.[28]
A further advantage of our current approach is that a
ligand-exchange reaction (which are not always effective) is
not used,[52,88,89] but rather a mild reduction process for par-
ticle synthesis. The approach is simpler than the approach
by Levy et al.,[52] and the results are similar to data by Si,
Mandal and co-workers with a water-soluble peptide,[42,43]
but are very different from Xie et al.,[90] who prepared
plate-like particles.
Experimental Section
Peptide synthesis: Amino acids were purchased from Bachem AG (Bu-
bendorf, Switzerland) and chemicals from Fluka (Buchs, Switzerland).
All chemicals were used as received. All amino acids are l-amino acids.
Chemical shifts are shown in ppm compared with TMS.
Synthesis of Z-Lys
ACHUTGTNRNEUG(N Boc)-LysACHUTNGTRENN(NGU Boc)-OtBu (1): To an ice-cooled solution of
Z-Lys(Boc)-OH (7.3 g, 19.2 mmol, 1.3 equiv; Boc=tert-butoxycarbonyl)
in chloroform (70 mL), 4-methylmorpholine (2.43 mL, 22.15 mmol,
1.5 equiv) was added and the solution was stirred under Ar for 5 min.
Then, isobutyl chloroformate (2.7 mL, 20.7 mmol, 1.4 equiv) were added
AHCTUNGTRENNUNG
and the solution was stirred for 2 min. To this solution, H-LysACTHNUGTRNEUNG(Boc)-OtBu
HCl salt (5 g, 14.75 mmol, 1 equiv) was added, followed by careful addi-
tion of 4-methylmorpholine (1.62 mL, 14.75 mmol, 1 equiv). After
15 min, the cooling was removed and the solution was allowed to warm
up to room temperature and stirred overnight. Then, chloroform
(200 mL) was added and the solution was washed with sodium hydroge-
nocarbonate solution (3ꢃ100 mL), saturated brine (100 mL), 10% citric
acid solution (3ꢃ100 mL), saturated brine (100 mL), and deionized water
(2ꢃ100 mL). The organic phase was dried over sodium sulfate and
evaporated to dryness. The solid 1 was crystallized from chloroform/di-
ethyl ether and dried in a vacuum oven at 408C, yielding a white powder
(4.3 g, 49%). 1H NMR (400 MHz, [D6]DMSO): d=8.05 (m, 0.82H,
Hamide), 7.31 (m, 5.5H, Hphenyl + 1Hphenyl carbamate), 6.73 (m, 1.73H, 1.73ꢃ
Htert-butyl carbamate), 6.38 (m, 0.15H, 0.15ꢃHtert-butyl carbamate), 4.99 (m, 2.04H,
H
H
phenyl), 4.00 (m, 1.86H, Ha-Lysine1 + Ha-Lysine2), 2.90 (m, 3.95H, Hb-Lysine1
b-Lysine2), 1.50 (m, 4.25H, Hg-Lysine1 + Hg-Lysine2), 1.30 ppm (m, 35H, Ht-butyl
+
+ HBoc
Lysine2).
+ HBoc
+ Hd-Lysine1 + Hd-Lysine2 + He-Lysine1 + He-
Lysine2
ester
Lysine1
N
8.37; 13C NMR (100 MHz, [D6]DMSO): d=172.95, 172.02,
Conclusion
156.77, 156.41, 137.91, 129.17, 128.61, 128.55, 81.22, 78.17, 66.18, 55.19,
53.50 , 32.57, 31.54, 30.10, 29.93, 29.13, 28.46, 23.71, 23.46 ppm; ATR-IR
(neat): n˜ =3334, 2973, 2932, 2867, 1688, 1514, 1452, 1244, 1159, 1043, 853,
777, 739, 695, 617 cmꢀ1; FAB-MS: m/z: calcd for [MꢀH]+ =665; mea-
sured [MꢀH]+ =665; elemental analysis calcd (%): C 61.42, H 8.49, N
8.43, O 21.66; found: C 61.35, H 8.25.
Herein, we show that peptides based on a disulfide bridge
can be efficiently used for the synthesis of peptide-coated
nanoparticles. At low pH values, individual particles are
present. At high pH values, large aggregates form. These
can be disintegrated again by lowering the pH value. There
is also evidence for the formation of different aggregate
sizes, depending on the pH value. The main asset of the cur-
rent system is its synthetic simplicity, with two identical tri-
peptides connected by a disulfide bond. The disulfide simpli-
fies the synthesis while maintaining full responsiveness of
the peptide in aqueous solution. To our best knowledge, this
is the first study of a stimulus-responsive silver nanoparticle/
peptide hybrid material in an aqueous environment. Some-
what simpler than an earlier gold-nanoparticle-based
system, it nevertheless keeps its responsiveness.[52]
In summary, we present a simple, yet flexible system for
the synthesis of complex materials. The most intriguing fea-
ture is probably the fact that the structure of the nanoparti-
cle aggregates depends only on the solution pH value, but in
a very interesting way: instead of a simple precipitation—
dissolution, variation of the pH value leads to individual
nanoparticles, small, or large aggregates. Overall, our study
suggests that metal nanoparticle/peptide hybrid materials
could be interesting for applications in which a reversible re-
sponse to an external stimulus (in this case, pH value) is de-
sired, such as medicine, biotechnology, or optical devices.
Synthesis of NH2-LysACTHUNGTERN(NNUG Boc)-LysAHCTUNGTRENN(GUN Boc)-OtBu (2): Compound 1 (4.3 g,
6.47 mmol, 1 equiv) was dissolved in methanol (50 mL) and the solution
with degassed with argon. To this solution, Pd/C catalyst (108 mg) was
added. The solution was hydrogenated at room temperature and at
50 bar overnight. The resulting black solution was filtered over sea sand
and the sea sand was carefully washed with portions of methanol. The re-
maining catalyst was removed by centrifugation, the solvents were evapo-
rated to dryness and the residue was left overnight in a vacuum oven at
408C to yield 2 (3.25 g; 95%). 1H NMR (400 MHz, [D6]DMSO,): d=8.4
(m, 0.1H, Hamide), 8.03 (m, 0.9H, Hamide), 6.74 (m, 1.86H, HBoc + HBoc),
6.37 (m, 0.14H, HBoc + HBoc), 4.06 (m, 1H, Ha-Lysine1), 3.12 (m, 1H, Ha-
Lysine2), 2.87 (m, 4H, He-Lysine1+2), 1.63 (m, 1.42H, Hb-Lysine1), 1.53 (m,
2.51H, Hb-Lysine1 + Hb-Lysine2), 1.3 ppm (m, 35H, HBoc Lysine1 + HBoc Lysine2
+
Ht-butyl ester + HgLysine1 + Hg-Lysine2); 13C NMR (100 MHz, [D6]DMSO,): d=
175.89, 172.16, 156.42, 156.40, 81.37, 78.17, 78.15, 55.04, 53.08, 49.45,
35.61, 31.89, 30.30, 29.94, 29.13, 28.48, 23.39, 23.37, 23.36 ppm; ATR-IR
(neat): n˜ =3358, 2975, 2938, 2867, 1720, 1686, 1655, 1519, 1502, 1457,
1392, 1365, 1338, 1252, 1160, 1061, 1037, 996, 867, 846, 778, 747, 696,
617 cmꢀ1
;
FAB-MS: m/z: calculated for [MꢀH]+ =531; measured
[MꢀH]+ =531; elemental analysis calcd (%): C 58.84, H 9.50, N 10.56, O
21.10; found: C 57.67, H 9.20, N 10.45.
Synthesis of (Boc-Cys-LysACTHNUTRGENN(GU Boc)-LysACHTUNTGRNE(NUGN Boc)-OtBu)2 (3): (Boc-Cys-OH)2
(cystine bridge) (1 g, 2.55 mmol, 0.9 equiv) was dissolved in THF (9 mL)
under argon, and the resulting solution was cooled with an ice bath. 4-
Methylmorpholine (2.7 mL, 12.3 mmol, 4.4 equiv) was added to this solu-
tion and the solution was stirred for 5 min. Then, isobutyl chloroformate
(1.5 mL, 12.3 mmol, 2.2 equiv) were added and the slurry was stirred for
13 min. To this solution, 2 (3 g, 5.6 mmol, 1 equiv) dissolved in DMF
(40 mL) was carefully added, and the mixture was stirred for 30 min with
ice cooling. Then, the ice bath was removed and the slurry stirred for
48 h. Chloroform (200 mL) was then added and the solution was washed
with sodium hydrogen carbonate solution (3ꢃ100 mL), saturated brine
(100 mL), 10% citric acid solution (3ꢃ100 mL), saturated brine
Chem. Eur. J. 2009, 15, 5831 – 5844
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5841

A. Mantion, A. Taubert et al.
(100 mL), and deionized water (2ꢃ100 mL). The organic phase was dried
over sodium sulfate and evaporated to dryness. The solid 3 was crystal-
lized from chloroform/pentane as a white powder and was dried in a
Hastings pseudo-Voigt profile that was convoluted with an axial diver-
gence asymmetry function.[91] The instrumental resolution function was
determined according to Louꢅr.[92] Anisotropic strain broadening was
used as implemented,[93] but the anisotropic size broadening was modeled
by using the Scherrer formula written as a linear combination of spheri-
cal harmonics.[94] GFourier[64] was used to generate average crystallite
shapes.
1
vacuum oven at 408C (2.7 g, 65%). H NMR (400 MHz, [D6]DMSO): d=
8.30 (0.23H, m, Hamide), 8.16 (1.23H, m, Hamide), 8.00 (0.71H, m, Hamide),
7.80 (0.91H, m, H), 7.12 (1.4H, m, HBoc), 6.72 (4.0H, HBoc), 6.30 (0.4H,
m, H), 4.28 (1.65H, m, Ha-Lysine1), 4.15 (0.9H, m, Ha-Lysine2), 4.00 (3H, m,
Ha-Lysine3
+ Ha-Lysine4 + Ha-Cystine2), 3.69 (1.28H, m, Hb Cystine1), 3.30
X-Ray photoelectron spectroscopy (XPS): XPS measurements have been
performed by using a VG ESCA Lab 220iXL (Thermo VG Scientific)
spectrometer equipped with a twin anode (MgKa and AlKa line) and an
X-ray source providing monochromatic AlKa light for excitation of the
photoelectrons. The take-off angle was 908 to the surface and the pres-
sure during the measurements was approximately 10ꢀ10 mbar. Silver
nanoparticles were deposited on a roughened flexible graphite surface to
avoid charging effects. For the quantification of the detected elements
the X-ray spot of the monosource, which is 500mm2 in size, was directed
on powder rich regions of the sample. To obtain S 2p signals with reason-
able intensity for quantification and interpretation of the binding state of
the sulfur, the twin anode (MgKa (1253.6 eV)) was used. In this case, the
complete surface of the sample was excited and the analyzed area was
defined by the settings of the analyzer which was 0.8 cm2.
NMR, IR, and UV/Vis spectroscopy: 1H and 13C NMR spectra were re-
corded on an Avance 400 MHz NMR spectrometer. Infrared spectra
were obtained from the neat samples on a Shimadzu FTIR 8300 with a
Golden Gate ATR unit. Spectra were recorded from 300 cmꢀ1 to
4000 cmꢀ1 with a resolution of 1 cmꢀ1. FAB-MS spectra were taken on a
Finnigan MAT 312. MALDI-TOF spectra were recorded on a Voyager-
DE Pro (Applied Bioscience) by using a-cyano-hydroxycinnamic acid as
the matrix. UV/Vis spectroscopy was performed in a quartz cuvette with
an optical path length of 1 cm on a Perkin Elmer Lambda. Data were de-
convolution was performed by using Fityk.
(0.62H, m, Hb Cystine2), 3.20 (0.26H, m, Hb-Cystine1), 3.06 (1.48H, m, Hb-Cys-
+ Hb-Cystine2), 2.84 (9.8H, m, Hb-Lysine), 1.60 (4H, m, Hg-Lysine1 + Hg-
tine1
Lysine2), 1.51 (5.3H, m, Hg-Lysine3 + Hg-Lysine4) , 1.33 (71H, m, HBocꢃ6 + He-
+ He-Lysine2 + He-Lysine3 + He-Lysine4 + Hd-Lysine1), 0.8 ppm (6H, m,
Lysine1
Hd-Lysine1
+
Hd-Lysine2
+
Hd-Lysine3); MALDI-TOF: m/z: calculated
[MꢀH]+ =1465, measured [MꢀH]+ =1465; elemental analysis calcd
(%): C 55.72, H 8.53, N 9.59, O 21.83, S 4.37; found: C 55.61, H 8.38, N
9.36.
Synthesis of (TFA.NH2-Cys-LysACTHUNRGTENNG(U NH2TFA)-LysACHUTNGTRNE(NUGN NH2TFA)-OH)2 (4): Iced
cooled 95% TFA/water (v/v) solution (20 mL) was added to peptide 3
(2.7 g). The solution was stirred for 2 h. Solvents and reagents were re-
moved under vacuum and the residual solvents were azeotropically re-
moved by using chloroform (3ꢃ100 mL). The resulting, slightly yellow
solid peptide 4 was dried overnight in a vacuum oven at 408C (2.63 g,
1
99%). H NMR (400 MHz, [D6]DMSO): d=8.8 (m, 1.0H, Hamide), 8.5 (m,
3H, Hamide), 8.25 (m, 0.5H, Hamide), 8.12 (m, 0.5H, Hamide), 8.08 (m, 0.75H,
Hamide), 7.9 ppm (m, 8H, 2Hlysinecarboxylicacid + 6HTFAcarboxylicacid), 4.34 (m,
1.4H, Ha-Lysine1), 4.13 (m, 3.24H, Ha-Lysine2 + Ha-Lysine3 + Ha-Lysine4), 3.9 (m,
1.18H, Hb1-Cystine1), 3.7 (m, 1.37H, Hb1-Cystine2), 2.9 (m, 1.4H, m, 1.18H,
H
b2-Cystine1), 2.7 (m, 8.5H, Hb-Lysine1 + Hb-Lysine2 + Hb-Lysine3 + Hb-Lysine4), 1.9
(m, 1.34H, Hb2-Cystine1), 1.74 (m, 3.74H, Hb2-Cystine2), 1.50 (m, 4H, He-Lysine1
+ He-Lysine2), 1.30 (m, 6H, He-Lysine4 + He-Lysine3 + Hd-Lysine1) 0.8 ppm (m,
6H, Hd-Lysine2 + Hd-Lysine3 + Hd-Lysine4); 13C NMR (100 MHz, [D6]DMSO):
d=174.28, 174.20, 174.14, 172.98, 171.91, 167.64, 167.46, 167.24, 160.00,
159.67, 159.35, 159.01, 157.07, 119.00, 116.12, 70.70, 55.00, 53.82, 52.82,
52.50, 52.38, 51.90, 32.48, 32.10, 31.10, 28.48, 28.41, 27.53, 27.47, 27.37,
23.29, 23.23, 23.18, 23.14, 22.90 ppm; ATR-IR (neat): n˜ =3269, 3061,
3942, 2877, 2652, 2542, 2877, 2652, 2542, 1659, 1635, 1525, 1474, 1430,
1392, 1331, 1181, 1126, 836, 795, 716 cmꢀ1. MALDI-TOF: m/z: calcd for
[MꢀH]+ =753, found [MꢀH]+ =753; calcd [MꢀNa]+ =775, found 775;
elemental analysis calcd (%): C 35.10, H 4.63, N 9.75, O 22.75, S 4.46, F
23.80; found: C 35.78, H 4.93, N 8.69.
Transmission electron microscopy (TEM): TEM images were taken by
using an FEI Morgani 268D operated at 80 kV. Samples were deposited
on carbon-coated copper grids and directly imaged after drying in air.
Some samples were diluted prior to imaging to allow for better imaging
conditions. Particle sizes were determined by measuring over 500 parti-
cles per sample from several images.
Surface-enhanced Raman spectroscopy (SERS): Silver nanoparticles
were investigated as neat powders with a confocal Raman microscope
(CRM300, WITec, Germany) equipped with a piezo-scanner (P-500,
Physik Instrumente, Germany), a 60ꢃobjective, and a 532 nm Nd:YAG
laser. The spectra were acquired by using an air-cooled CCD detector
(DU401-BV, Andor, UK) with 600 gratings/mm (UHTS 300, WITec, Ger-
many). ScanCtrlSpectroscopyPlus (version 1.38, WITec) was used for
data acquisition and processing.
Nanoparticle synthesis: To 0.2m silver(I) nitrate solution (5 mL, 1 mmol),
a pH X (X=3, 7, 9) solution of peptide 4 (0.059 mmol, 5 mL) was added.
The mixture was stirred for 12 h in the dark. Then 1.5m aqueous sodium
ascorbate solution (2 mL, 3 mmol) and an aqueous solution of the pep-
tide 4 (5 mL, 0.059 mmol) at pH 3 were added. The mixture was stirred
for 12 h in the dark. The nanoparticles were purified during for sequen-
ces of centrifugation (14000 rpm, 3 mL Teflon centrifugation tubes) and
washing with ultrapure water befor lyophilization. Sample 1 (denoted P1)
was prepared at pH 3, sample 2 (P2) at pH 7, and sample 3 (P3) at pH 9.
TGA shows a capping of the nanoparticles with the peptide of around
5% for P1 and P2, and a capping of 2% for P3.
Dynamic light scattering (DLS): DLS experiments were performed on a
Zetasizer ZS (Malvern Instruments, UK) on particle dispersions in dis-
tilled water in a disposable plastic cuvette. Experiments were performed
at 20ꢁ18C. Samples were not filtered before measurements. The laser
wavelength was 633 nm and data were recorded in backscattering mode
at 2q=1738. At least ten 10 s measurements were made and data were
averaged. A regularized inverse Laplace transformation based on the
Contin algorithm as implemented in the Malvern Software DTS 5.02 was
used for data analysis. The concentration was adjusted to avoid strong
particle–particle repulsion through electrostatic interactions and to ac-
count for an optimum count rate (around 200 kcps). Optical properties of
bulk metallic silver[95] (n=0.56 and k=4.27 at 633 nm, where n is the real
part and k is the imaginary component of the silver refractive index)
were used for data analysis.
Thermogravimetric analysis: TGA was performed with a Mettler Toledo
TGA/SDTA 851e from 25 to 10008C with a heating rate of 108C.minꢀ1 in
N2.
Potentiometry: Potentiometric titrations were performed on a Mettler
Toledo T 50 automatic titrator with a DG-115 SC glass electrode at 258C.
Silver nanoparticles (197.7 mg) were suspended in a mixture of 0.03m hy-
drochloric acid and 0.5m sodium chloride solution (40 mL). During titra-
tion, 0.25m sodium hydroxide (10 mL) was delivered in aliquots of
0.1 mL from the titrator. The time interval between additions was 90 s.
Prior to titration, the electrode was calibrated by titrating a mixture of
0.1m ammonia acetate and 0.1m hydrochloric acid with 0.5m sodium hy-
droxide. Data were analyzed by using Microsoft Office 2003 Excel soft-
ware employing Granꢄs method.[78]
Small-angle X-ray scattering (SAXS): SAXS measurements were per-
formed with a Kratky-type instrument (SAXSess from Anton Paar, Aus-
tria). The SAXSess has a low sample-to-detector distance which is suita-
ble for investigation of low scattering intensities. The measured intensity
was corrected by subtracting the intensity from a capillary filled with
pure water. The scattering vector is defined in terms of the scattering
angle q and the wavelength l of the radiation (l=0.154 nm): thus q=
X-ray diffraction (XRD) and Rietveld refinement: XRD was done on a
Nonius PDS 120 with a position sensitive detector (1 to 1208 2q by using
CuKa radiation. Rietveld refinement was performed by using Fullprof ver-
sion 4.00 (May 2008).[64] The peaks were fitted with a Thompson–Cox–
4p/l sin
ACHTUNGTREN(NUNG q/2). Deconvolution (desmearing) of the SAXS curves was per-
formed with the SAXS-Quant software (version 2.0) from Anton Paar.
5842
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5831 – 5844

Assembly of Peptide-Coated Silver Nanoparticles
FULL PAPER
An analytic solution for Schulz-type polydisperse spheres[96] was used for
determination of mean particle radius of hRi and polydispersity of s/hRi.
S(q) was modeled with a square well potential.[97] The square well poten-
tial U(r) is identical to the hard sphere interaction for distances shorter
than the particle radius, that is, U(r)=1 (hard sphere interaction), ꢀU0
for distances hRi<r<lhRi (attractive) and zero for r>lhRi.
[25] D. Andreescu, T. K. Sau, D. V. Goia, J. Colloid Interface Sci. 2006,
298, 742–751.
[26] A. Henglein, M. Giersig, J. Phys. Chem. B 1999, 103, 9533–9539.
[27] Z. S. Pillai, P. V. Kamat, J. Phys. Chem. B 2004, 108, 945–951.
[28] A. Mantion, A. G. Guex, A. Foelske, L. Mirolo, K. M. Fromm, M.
Painsi, A. Taubert, Soft Matter 2008, 4, 606–617.
[29] E. E. Foos, A. W. Snow, M. E. Twigg, M. G. Ancona, Chem. Mater.
2002, 14, 2401–2408.
[30] S.-W. Kim, S. Kim, J. B. Tracy, A. Jasanoff, M. G. Bawendi, J. Am.
Chem. Soc. 2005, 127, 4556–4557.
[31] S. KØki, J. Tçrçk, G. Deµk, L. Daróczi, M. Zsuga, J. Colloid Inter-
face Sci. 2000, 229, 550–553.
Acknowledgements
[32] A. H. Latham, M. E. Williams, Langmuir 2006, 22, 4319–4326.
[33] A. Manna, T. Imae, K. Aoi, M. Okada, T. Yogo, Chem. Mater. 2001,
13, 1674–1681.
[34] U. Oktay, H. Ying, V. Ayush, C. Suelin, C. Andrea, S. Francesco,
Chem. Commun. 2007, 196–198.
[35] D. I. Gittins, F. Caruso, ChemPhysChem 2002, 3, 110–113.
[36] C. H. B. Ng, J. Yang, W. Y. Fan, J. Phys. Chem. C 2008, 112, 4141–
4145.
[37] R. M. Kramer, C. Li, D. C. Carter, M. O. Stone, R. R. Naik, J. Am.
Chem. Soc. 2004, 126, 13282–13286.
We thank Prof. E. C. Constable for access to his IR spectrometer, Prof.
B. Giese for access to his MALDI-TOF apparatus, and D. De Bruyn for
the MALDI-TOF measurement. A. Mantion thanks the Adolf-Martens-
Fonds e.V. for an Adolf Martens Fellowship. a. Masic is grateful for sup-
port by the Alexander von Humboldt Foundation and the Max Planck
Society in the framework of the Max Planck Research Award by the Fed-
eral Ministry of Education and Research. The Swiss National Science
Foundation, the University of Potsdam, the Federal Institute for Materi-
als Research and Testing, and the Max-Planck-Institute of Colloids and
Interfaces are acknowledged for financial support.
[38] S. Banta, Z. Megeed, M. Casali, K. Rege, M. L. Yarmush, J. Nano-
sci. Nanotechnol. 2007, 7, 387–401.
[39] K. Channon, C. E. MacPhee, Soft Matter 2008, 4, 647–652.
[40] S. S. Santoso, S. Zhang, Encycl. Nanosci. Nanotechnol. 2004, 459–
471.
[41] R. R. Naik, S. J. Stringer, G. Agarwal, S. E. Jones, M. O. Stone, Nat.
Mater. 2002, 1, 169–172.
[1] D. D. Evanoff, G. Chumanov, ChemPhysChem 2005, 6, 1221–1231.
[2] K.-S. Lee, M. A. El-Sayed, J. Phys. Chem. B 2006, 110, 19220–
19225.
[3] M. Kang, R. Jung, H.-S. Kim, J. H. Youk, H.-J. Jin, J. Nanosci.
Nanotechnol. 2007, 7, 3888–3891.
[42] S. Si, E. Dinda, T. Mandal, Chem. Eur. J. 2007, 13, 9850–9861.
[43] S. Si, T. K. Mandal, Chem. Eur. J. 2007, 13, 3160–3168.
[44] S. Ray, A. K. Das, A. Banerjee, Chem. Commun. 2006, 2816–2818.
[45] M. Brust, J. Fink, D. Bethell, D. J. Schiffrin, C. Kiely, J. Chem. Soc.
Chem. Commun. 1995, 1655–1656.
[4] A. Panacek, L. Kvꢆtek, R. Prucek, M. Kolar, R. Vecerova, N. Pizfflro-
va, V. K. Sharma, T. Nevecna, R. Zboril, J. Phys. Chem. B 2006, 110,
16248–16253.
[5] M.-C. Daniel, D. Astruc, Chem. Rev. 2004, 104, 293–346.
[6] A. Henglein, Chem. Rev. 1989, 89, 1861–1873.
[7] P. Chen, L. Song, Y. Liu, Y.-e. Fang, Radiat. Phys. Chem. 2007, 76,
1165–1168.
[46] M. Brust, M. Walker, D. Bethell, D. J. Schiffrin, R. Whyman, J.
Chem. Soc. Chem. Commun. 1994, 801–802.
[47] M. Haupt, S. Miller, R. Glass, M. Arnold, R. Sauer, K. Thonke, M.
Moller, J. P. Spatz, Adv. Mater. 2003, 15, 829–831.
[48] R. R. Naik, S. E. Jones, C. J. Murray, J. C. McAuliffe, R. A. Vaia,
M. O. Stone, Adv. Funct. Mater. 2004, 14, 25–30.
[8] Y.-O. Kang, S.-H. Choi, A. Gopalan, K.-P. Lee, H.-D. Kang, Y. S.
Song, J. Non-Cryst. Solids 2006, 352, 463–468.
[9] F.-K. Liu, Y.-C. Hsu, M.-H. Tsai, T.-C. Chu, Mater. Lett. 2007, 61,
2402–2405.
[49] A. Semenov, J. P. Spatz, M. Moller, J.-M. Lehn, B. Sell, D. Schubert,
C. H. Weidl, U. S. Schubert, Angew. Chem. 1999, 111, 2701–2705;
Angew. Chem. Int. Ed. 1999, 38, 2547–2550.
[50] J. P. Spatz, S. Mossmer, M. Moller, T. Herzog, A. Plettl, P. Ziemann,
Polym. Mater. Sci. Eng. 1997, 77, 420–421.
[10] S. Remita, P. Fontaine, E. Lacaze, Y. Borensztein, H. Sellame, R.
Farha, C. Rochas, M. Goldmann, Nucl. Instrum. Methods Phys. Res.
Sect. B 2007, 263, 436–440.
[11] A. Taubert, I. Arbell, A. Mecke, P. Graf, Gold Bull. 2006, 39, 205–
211.
[51] Q. Dong, H. Su, D. Zhang, J. Phys. Chem. B 2005, 109, 17429–
17434.
[12] I. Pastoriza-Santos, L. M. Liz-Marzan, Pure Appl. Chem. 2000, 72,
83–90.
[52] R. LØvy, K. T. Thanh Nguyen, D. R. Christopher, H. Irshad, R. J.
Nichols, D. J. Schiffrin, M. Brust, D. G. Fernig, J. Am. Chem. Soc.
2004, 126, 10076–10084.
[53] J. Slocik, D. Wright, Biomacromolecules 2003, 4, 1135–1141.
[54] Z. Wang, R. LØvy, D. G. Fernig, M. Brust, Bioconjugate Chem. 2005,
16, 497–500.
[13] S. P. Mushran, M. C. Agrawal, R. M. Mehrotra, R. Sanehi, J. Chem.
Soc. Dalton Trans. 1974, 1460–1462.
[14] I. Sondi, D. V. Goia, E. Matijevic, J. Colloid Interface Sci. 2003, 260,
75–81.
[15] L. Suber, I. Sondi, E. Matijevic, D. V. Goia, J. Colloid Interface Sci.
2005, 288, 489–495.
[55] J. Kong, S. Yu, Acta Biochim. Biophys. Sin. 2007, 39, 549–559.
[56] H. Maekawa, C. Toniolo, A. Moretto, Q. Broxterman, N.-H. Ge, J.
Phys. Chem. B 2006, 110, 5834–5837.
[16] P. C. Lee, D. Meisel, J. Phys. Chem. 1982, 86, 3391–3395.
[17] V. Privman, D. V. Goia, J. Park, E. Matijevic, J. Colloid Interface
Sci. 1999, 213, 36–45.
[57] G. Martinez, G. Millhauser, J. Struct. Biol. 1995, 114, 23–27.
[58] A. Mantion, A. Taubert, Macromol. Biosci. 2007, 7, 208–217.
[59] S. Y. Venyaminov, N. N. Kalnin, Biopolymers 1990, 30, 1243–1257.
[60] L. Pauling, R. B. Corey, H. R. Branson, Proc. Natl. Acad. Sci. USA
1951, 37, 205–211.
[61] M. A. Van Hove, G. A. Somorjai, Surfactant Sci. Ser. 1980, 92, 489–
518.
[62] P. D. Jadzinsky, G. Calero, C. J. Ackerson, D. A. Bushnell, R. D.
Kornberg, Science 2007, 318, 430–433.
[63] C. Gautier, T. Buergi, J. Am. Chem. Soc. 2008, 130, 7077–7084.
[64] J. Rodrꢆguez-Carvajal, Physica B+C 1993, 192, 55–69.
[65] C. J. Sandroff, D. R. Herschbach, J. Phys. Chem. 1982, 86, 3277–
3279.
[18] J. Turkevich, P. C. Stevenson, J. Hillier, J. Phys. Chem. 1953, 57,
670–673.
[19] J. Turkevich, P. C. Stevenson, J. Hillier, Discuss. Faraday Soc. 1951,
11, 55–75.
[20] J. A. Dahl, B. L. S. Maddux, J. E. Hutchison, Chem. Rev. 2007, 107,
2228–2269.
[21] H. Chih-Ching, Y. Zusing, C. Huan-Tsung, Langmuir 2004, 20,
6089–6092.
[22] Y.-F. Huang, Y.-W. Lin, H.-T. Chang, Nanotechnology 2006, 17,
4885–4894.
[23] H. Yu-Fen, H. Kuan-Ming, C. Huan-Tsung, J. Colloid Interface Sci.
2006, 301, 145–154.
[24] I. Halaciuga, D. V. Goia, J. Mater. Res. 2008, 23, 1776–1784.
Chem. Eur. J. 2009, 15, 5831 – 5844
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5843

A. Mantion, A. Taubert et al.
[66] B. Pettinger, H. Wetzel, Chem. Phys. Lett. 1981, 78, 398–403.
[67] B. A. Sexton, R. J. Madix, Chem. Phys. Lett. 1980, 76, 294–297.
[68] M. Schnippering, M. Carrara, A. Foelske, R. Kçtz, D. J. Fermꢆn,
Phys. Chem. Chem. Phys. 2007, 9, 725–730.
[69] G. Faraci, E. Costanzo, A. R. Pennisi, Y. Hwu, G. Margaritondo, Z.
Phys. D: At. Mol. Clusters 1992, 23, 263–267.
[70] L. Guczi, G. Peto, A. Beck, Z. Paszti, Curr. Top. Catal. 2004, 29,
129–138.
[71] H. S. Shina, H. C. Choia, Y. Junga, S. B. Kim, H. J. Song, H. J. Shin,
Chem. Phys. Lett. 2004, 383, 418–422.
[82] M. Brust, C. J. Kiely, Colloids Surf. A 2002, 202, 175–186.
[83] Y.-S. Shon, D. Choi, Curr. Nanosci. 2007, 3, 245–254.
[84] L. C. Gruen, Biochim. Biophys. Acta 1975, 386, 270–274.
[85] N. Naoki, Y. Hiroshi, K. Keisaku, Langmuir 2008, 24, 2759–2766.
[86] J. I. Pascual, J. V. Barth, G. Ceballos, G. Trimarchi, A. De Vita, K.
Kern, H.-P. Rust, J. Chem. Phys. 2004, 120, 11367–11370.
[87] G. Shemer, O. Krichevski, G. Markovich, T. Molotsky, I. Lubitz,
A. B. Kotlyar, J. Am. Chem. Soc. 2006, 128, 11006–11007.
[88] C. M. Niemeyer, B. Ceyhan, S. Gao, L. Chi, S. Peschel, U. Simon,
Colloid Polym. Sci. 2001, 279, 68–72.
[72] K. Ashavani, J. Hrushikesh, P. Renu, M. A. B. S. Murali, J. Colloid
Interface Sci. 2003, 264, 396–401.
[89] M. Wanner, D. Gerthsen, S.-S. Jester, B. Sarkar, B. Schwederski,
Colloid Polym. Sci. 2005, 283, 783–792.
[73] S. Kang, K. Kim, Langmuir 1998, 14, 226–230.
[74] C. D. Wagner, G. E. Muilenberg, Handbook of X-ray photoelectron
spectroscopy: A reference book of standard data for use in x-ray pho-
toelectron spectroscopy, Perkin-Elmer, Eden Prairie, 1979, p. 122.
[75] C.-J. Zhong, M. D. Porter, J. Am. Chem. Soc. 1994, 116, 11616–
11617.
[76] L. D. Setiawan, H. Baumann, D. Gribbin, Surf. and Interface Anal.
1985, 7, 188–195.
[77] J. C. Munro, C. W. Frank, Polymer 2003, 44, 6335–6344.
[78] L. M. Sigg, W. Stumm, Aquatische Chemie, Teubner, Stuttgart, 1994.
[79] K. Kelly, E. Coronado, L. Zhao, G. Schatz, J. Phys. Chem. B 2003,
107, 668–677.
[90] J. Xie, J. Y. Lee, D. I. C. Wang, Y. P. Ting, ACS Nano 2007, 1, 429–
439.
[91] P. Thompson, D. E. Cox, J. B. Hastings, J. Appl. Crystallogr. 1987,
20, 79–83.
[92] D. Louꢅr, J. I. Langford, J. Appl. Crystallogr. 1988, 21, 430–437.
[93] J. Rodriguez-Carvajal, M. T. Fernandez-Diaz, J. L. Martinez, J. Phys.
Condens. Matter 1991, 3, 3215–3234.
[94] M. Jarvinen, J. Appl. Crystallogr. 1993, 26, 525–531.
[95] P. B. Johnson, R. W. Christy, Phys. Rev. B 1972, 6, 4370–4379.
[96] D. Pontoni, S. Finet, T. Narayanan, A. R. Rennie, J. Chem. Phys.
2003, 119, 6157–6165.
[97] R. V. Sharman, K. C. Sharman, Physica 1977, 89, 213–217.
[80] J. J. Mock, M. Barbic, D. R. Smith, D. A. Schultz, S. Schultz, J.
Chem. Phys. 2002, 116, 6755–6759.
[81] I. Sosa, C. Noguez, R. Barrera, J. Phys. Chem. B 2003, 107, 6269–
6275.
Received: November 9, 2008
Revised: February 6, 2009
Published online: April 15, 2009
5844
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 5831 – 5844