Five orders-of-magnitude enhancement of two-photon absorption for dyes on silver nanoparticle fractal clusters

Article abstract of DOI:10.1021/jp014675f

Strong enhancement of the two-photon absorption of organic molecules near silver nanoparticle fractal clusters has been observed and has been exploited to yield composite materials with very strong two-photon absorption and two-photon-excited fluorescence properties. Measurements on cluster films coated with chromophoric polymer or with thiol-bound chromophores give spatially-averaged enhancements of 1000 and 20000, respectively. Two-photon fluorescence microscopy studies show that the enhancements are spatially inhomogeneous, with peak-enhancement factors of a?￥ 10000 (polymer/cluster) and a?￥ 160000 (thiol chromophore/cluster), and excitation frequency dependent. These results are in accord with theoretical predictions of local-field effects due to strong localization of collective plasmon modes in fractal metal clusters, and demonstrate an approach to ultrasensitive two-photon processes.

Full text of DOI:10.1021/jp014675f

J. Phys. Chem. B 2002, 106, 6853-6863
6853
Five Orders-of-Magnitude Enhancement of Two-Photon Absorption for Dyes on Silver
Nanoparticle Fractal Clusters
Wim Wenseleers,† Francesco Stellacci, Timo Meyer-Friedrichsen, Timo Mangel,
Christina A. Bauer, Stephanie J. K. Pond, Seth R. Marder,* and Joseph W. Perry*
Department of Chemistry, The UniVersity of Arizona, Tucson, Arizona 85721
ReceiVed: December 31, 2001 ; In Final Form: March 7, 2002
Strong enhancement of the two-photon absorption of organic molecules near silver nanoparticle fractal clusters
has been observed and has been exploited to yield composite materials with very strong two-photon absorption
and two-photon-excited fluorescence properties. Measurements on cluster films coated with chromophoric
polymer or with thiol-bound chromophores give spatially-averaged enhancements of 1000 and 20 000,
respectively. Two-photon fluorescence microscopy studies show that the enhancements are spatially
inhomogeneous, with peak-enhancement factors of g 10 000 (polymer/cluster) and g 160 000 (thiol
chromophore/cluster), and excitation frequency dependent. These results are in accord with theoretical
predictions of local-field effects due to strong localization of collective plasmon modes in fractal metal clusters,
and demonstrate an approach to ultrasensitive two-photon processes.
Introduction
of the local electric field distribution at a gold tip give an
intensity enhancement factor of 1000. Theoretical studies by
Composite systems based on metal nanoparticles and surfaces
offer possibilities for enhancement of optical responses, such
as surface-enhanced Raman scattering (SERS) and second
harmonic generation, due in part to large local field effects.1-3
It is thought that giant local fields, leading to strongly enhanced
nonlinear optical responses, can be generated in the vicinity of
fractal clusters of metal nanoparticles.4 A nonlinear optical
process of current interest is two-photon absorption (TPA),
wherein two quanta of light are absorbed simultaneously, which
provides a means for three-dimensionally resolved excitation
of materials. Applications of TPA being actively pursued include
three-dimensional fluorescence microscopy,5 optical data stor-
age,6 microfabrication,7 and optical limiting.8 TPA is typically
a weak process that requires high intensity for excitation, even
with chromophores specifically designed to have a large TPA
cross-section.
Local field enhancement of two-photon-excited fluorescence
(TPF) has been obtained for molecules on smooth metal
surfaces9 and on evaporated silver island films,10 resulting in a
90 and 150-fold increase, respectively. Enhancement of TPF
has also been realized by illuminating a gold tip,11 and this was
used for ultrahigh resolution (30 nm) near-field fluorescence
imaging. Photodecomposition of pyridine and other aromatic
molecules on roughened Ag(110) surfaces has also been
attributed to an enhanced two-photon absorption process.12 Much
larger enhancements of nonlinear optical effects and TPA, in
particular, are expected to occur near fractal clusters of metal
nanoparticles.4
Shalaev et al.4,13,14 have predicted an even stronger enhancement
of optical processes near semi-continuous metal films and metal
nanoparticle clusters with a random fractal geometry.15 Whereas
a periodic lattice (1, 2, or 3D) of metal particles can support
traveling plasmon waves, aggregates of metal particles pos-
sessing a fractal geometry (noninteger dimension) lack trans-
lational symmetry and therefore exhibit strongly localized
plasmon modes. Importantly, interaction of the particles in the
fractal cluster leads to a redistribution of the plasmon amplitude
from a large number of particles into a much larger amplitude
localized on a few particles, producing giant local fields in sub-
wavelength sized “hotspots”. The coupled plasmon resonances
of fractal clusters extend to long wavelengths relative to the
single particle plasmon resonance, and span a broad region of
the spectrum (visible to IR), allowing for strong resonance
effects over a wide wavelength range. Due to the inhomogen-
eous nature of the coupled resonances, the locations of the
hotspots are predicted to be excitation wavelength dependent.
Direct evidence for localization of optical fields near metal
nanoparticle fractal clusters has been obtained using near-field
optical microscopy.16 Because optical responses scale with a
power of the local field, molecules residing in hotspots can
exhibit strongly enhanced responses. This picture offers one way
to explain the electromagnetic contribution to SERS.2,13,17
Nonlinear optical responses of molecules near fractal metal
clusters are expected to be enhanced by many orders of
magnitude,18 as they are proportional to a high order function
of the local field. However, the use of metal fractal composites
for the enhancement of nonlinear optical effects has been
limited. Observations of reduction in lasing thresholds for dyes
in colloidal silver suspensions19,20 and enhanced degenerate four
wave mixing (DFWM)21 in clustered metal nanoparticles have
been reported.
Because an applied electromagnetic field can excite collective
oscillations of free electrons (plasmons) of metal nanostructures,
large local fields can be generated. For example, calculations11
* To whom correspondence should be addressed: E-mail: jwperry@
u.arizona.edu (JWP), smarder@u.arizona.edu (SRM).
† Current address: Physics Department, University of Antwerp, 2610
Wilrijk-Antwerpen, Belgium.
There is potential for strong enhancement of multiphoton
absorption effects because simultaneous absorption of n photons
10.1021/jp014675f CCC: $22.00 © 2002 American Chemical Society
Published on Web 06/06/2002

6854 J. Phys. Chem. B, Vol. 106, No. 27, 2002
Wenseleers et al.
peak wavelength of 890 nm. The fluorescence quantum yield
(η) of polymer A is 0.83, as measured relative to 9,10-bis-
(phenylethynyl)anthracene26. The large ηδ product of the
polymer allows for sensitive detection of TPA by monitoring
the TPF signal. The TPF excitation spectrum (Figure 2) of the
polymer shows strong TPA over a broad wavelength region
spanning from <710 to >920 nm. Amorphous polymer films
of 10 to 200 nm thickness were cast onto the nanoparticle cluster
film by solvent evaporation from dilute dichloromethane solu-
tions.
We investigated the TPF signals from polymer coated cluster
films, neat polymer films, and uncoated cluster films (the latter
two serving as control samples) using ∼100 fs mode-locked
Ti:sapphire laser pulses with a wavelength of ∼800 nm.
Intensity-dependent and spectroscopic measurements provide
strong evidence that the observed emission from the composite
was due to TPA and was dominated by the emission from the
polymer. The fluorescence signals from the composites and the
control samples were observed to depend quadratically on
incident laser intensity (see Figure 3) and no fluorescence was
detected when a continuous-wave laser beam (with the same
average intensity but peak intensity lower by ∼105) was used,
confirming an instantaneous two-photon process for all samples.
The TPF emission spectrum of the composite film is different
from that of the cluster film27 but very similar to that of the
neat polymer, as shown in Figure 2, indicating that the emission
of the composite originates from the organic chromophores.
Figure 1. Structures of the chromophoric polymer (A) and thiol-
functionalized chromophore (B).
scales as In, where I is the light intensity. Because the
concentration of intensity into hotspots creates large peaks of
local intensity, the net multiphoton absorption (proportional to
the average value of In over a given volume) can be very large
compared with that for a uniform intensity distribution. Fur-
thermore, applications of multiphoton absorption effects are not
significantly hampered by slight background absorption and
scattering effects involved in fractal cluster systems, as com-
pared to nonlinear refractive index processes such as DFWM.
Here we report an investigation of the two-photon-excited
fluorescence of both polymer-coated and dye-anchored fractal
clusters of silver nanoparticles, and study the magnitude of the
enhancement as well as its spatial variation and excitation
frequency dependence.
Much stronger TPF signals were observed for the composite
films than for the control films. An enhancement factor for the
TPF signal can be defined as (Scomp - Sc)/Sp, where Scomp, Sc,
and Sp are the signals for the composite, the cluster film and
the neat polymer, respectively. We obtained a large spatially
averaged enhancement factor for a series of composite films,
with values ranging from 50 to 250 (as measured with an
unfocused beam with a cross-sectional area of ∼10 mm2) for
composite films with an ∼100 nm thick polymer coating. The
exact value was found to depend on the surface coverage of
the silver clusters, as expected. To study the dependence on
polymer film thickness a series of substrates with the same silver
coverage was produced by simultaneous deposition in the same
sedimentation receptacle. For polymer films with thicknesses
of 10 nm, the spatially averaged enhancement was 1000. For
thicker films, the enhancement factor (always determined
relative to a neat polymer film of the same thickness) varied
inversely with thickness, e.g., an enhancement factor of 90 was
observed for a 100 nm thick film. This falloff for thicknesses
g 10 nm can be explained as follows: for all the composite
films the TPF signal is dominated by the TPF from the first
<10 nm layer and the additional polymer at greater distance
from the silver clusters does not contribute significantly to the
total signal, whereas the signal for the neat polymer films grows
linearly with thickness. The apparent enhancement factor
therefore drops off inversely with polymer thickness. This
observation is consistent with the expectation based on theory
that the local field enhancement occurs only very close to the
metal clusters.28
Results and Discussion
Our experiments focused on TPA by organic chromophores
at silver nanoparticle cluster surfaces. Essentially, no enhance-
ments of TPA were observed for rhodamine 6G dye solutions
(10-7 mol/L as in ref 19) containing colloidal silver fractal
clusters, or for dye-doped (10% by wt) poly(methyl methacry-
late) (PMMA) coated silver fractal films. We recognized that it
is necessary to have a high concentration of chromophores near
the surface of the fractals to effectively populate the hotspots,
while minimizing the number of chromophores far from the
silver that would generate a background fluorescence signal.
As a first approach to achieve this, a thin film of a polymer
comprised entirely of chromophoric units was coated onto a
silver cluster film. Silver nanoparticle fractal films were prepared
by sedimentation of clusters onto a glass substrate as described
in the literature, and were characterized by various techniques
(see the Experimental Section). We estimated that these films
have a fractal dimension of 1.45-1.65 from optical and scanning
electron microscopy (imaged perpendicular to film), using the
area-perimeter method.22
To test whether the observed phenomenon is spatially
inhomogeneous, as predicted by theory,14 and if so, to estimate
the magnitude of the peak enhancement factor, we imaged the
TPF from the composite films by using laser-scanning fluores-
cence microscopy. As shown in Figure 4A, most of the TPF
indeed originates from small spots, whose size is limited by
the instrumental resolution of 250 nm (fwhm).29 Comparison
of TPF images with transmission optical microscopy images
The chromophoric polymer (A), whose structure is shown in
Figure 1, contains donor-acceptor-donor units23,24 in the main
chain that exhibit a large TPA cross section and high fluores-
cence quantum efficiency. It was synthesized via a Heck
coupling of a divinyltriarylamine, 3, and a dicyano-substituted,
diiodobis(stryryl)benzenze, 6, as shown in Scheme 1. Measure-
ments of the TPA cross section25 (δ) for this polymer in toluene
give δ ) 1020 × 10-50 cm4 s/photon per repeat unit at the

Dyes on Silver Nanoparticle Fractal Clusters
J. Phys. Chem. B, Vol. 106, No. 27, 2002 6855
SCHEME 1
of the same film shows a strong correlation between the
positions of fractal clusters and of the enhanced TPF (Figure 4
B, D). The TPF enhancement are observed only from distinct
spots within a cluster, not from the whole cluster, as will be
discussed below. The peak TPF signal in these spots is at least
an order-of-magnitude larger than the spatially averaged signal.
A quantity of interest is the peak enhancement factor of the
TPF intensity generated by a molecule residing in a hotspot
relative to that of a molecule in the absence of the metal clusters.
Combining the spatially resolved imaging results with the
thickness dependent measurements on the composite films, we
deduce that the peak enhancement factor of the TPF is g 10 000
for the polymer coated nanoparticle cluster composites.
performed two-photon laser-scanning fluorescence microscopy
to obtain images of the same region of a composite film using
different laser wavelengths. Indeed, we observed a distinct and
reproducible pattern of hotspot positions for each excitation
wavelength used (Figure 4B). The spatially resolved and clearly
visible colors in the false color overlay image arise from
resonances being excited with different wavelengths, and
confirm the frequency dependence of the plasmon localization.
With the ability to perform excitation frequency-dependent
spatial imaging, TPF can be a useful tool for studying localiza-
tion of plasmons, which is of interest in the context of Anderson
localization.30
Enhancement factors obtained using collimated excitation
beams at different angles of incidence, as well as under tight
focusing conditions (numerical aperture ) 1.4, cone-half-angle
Another prediction of theory is that the positions of the
hotspots are dependent on the excitation wavelength.14 We

6856 J. Phys. Chem. B, Vol. 106, No. 27, 2002
Wenseleers et al.
To investigate whether the large enhancements depend on
the coupling of particles in the fractal clusters, we compared
the TPF of dyes attached to isolated nanoparticles and fractal
clusters. In the sedimented silver films, the formation of fractal
clusters is intrinsic to the deposition process. However, by using
silver nanoparticles coated with alkylthiol ligands, aggregation
of nanoparticles is suppressed, and isolated nanoparticles can
be studied in solution. We synthesized nanoparticles in which
a fraction of the alkylthiol coating was exchanged with a thiol-
functionalized two-photon dye (B; see Figure 1). The thiol
containing two-photon chromophore, B, was synthesized in six
steps from 4-methoxyphenol as shown in Scheme 2. We
compared the one- and two-photon-excited fluorescence from
such nanoparticles with those of the free dye, both in toluene
solution and in PMMA films.
The fluorescence signal of the dye molecules bound to the
nanoparticles was reduced by a factor of 4.5 compared to that
of the free dye, as observed using one- or two-photon excitation
(at 450 and 730 nm, respectively) of toluene solutions with the
same total dye concentration (see also the Experimental Section).
It should be noted that there was no significant emission from
analogous particles without dye. One possible interpretation of
these results is that neither the one- nor the two-photon
absorption cross sections are significantly changed when the
dye is bound to the particles, but the emission quantum yield
of the dye is reduced from 75% to 17%. Alternatively, it could
be that both the one- and the two-photon absorptions are
enhanced, and the fluorescence quantum yield is reduced by a
larger factor. However, as one would expect the fluorescence
quantum yield to be the same in both cases, this would imply
that the one- and the two-photon absorption have the same
average enhancement factor, which is very unlikely given the
different powers of the local fields involved in the two
absorption processes.
Figure 2. One-photon and two-photon spectra of polymer (A),
composite, and metal cluster films. One-photon absorption spectrum
of polymer in toluene solution (blue). Two-photon fluorescence
excitation spectrum of polymer in toluene solution (femtosecond data
in red, nanosecond data in green), calibrated against Coumarin 307 in
methanol (ref 25) and plotted against half the excitation wavelength.
One-photon-excited fluorescence spectrum of polymer in toluene
solution (solid black line). Two-photon-excited emission spectra (λexc
) 800 nm) of the neat polymer film (dashed black line), the polymer/
cluster composite film (dash-dotted black line), and the neat metal
cluster film (orange).
For the PMMA films, the ratio of two-photon-excited
fluorescence to the one-photon excited fluorescence was found
to be the same for the films containing dye-coated nanoparticles
and for PMMA films containing free dye. Thus, within
experimental error, no enhancement of the two-photon absorp-
tion is observed for these isolated particles, whereas large
enhancements are observed for the coupled particles in the
fractal clusters, confirming the importance of the particle
interactions in the enhancement mechanism.
As an intermediate case to the fractal clusters and the isolated
nanoparticles, we investigated neat films of the dye-attached
ligand-coated nanoparticles. It is well established31 that, upon
casting by solvent evaporation, alkylthiol-coated nanoparticles
form partially ordered arrays, with a particle spacing of about
one ligand length, due to the interdigitation of alkyl chains.
Reference films comprising dye coated nanoparticles dissolved
in a PMMA matrix were used for comparison to the neat nano-
particle films. However, it was difficult to prepare reference
films with the same number of dye molecules within the
interaction volume of the excitation beam compared to the neat
nanoparticle film. Therefore, for this experiment, we used a
normalization scheme based on the one-photon excited fluo-
rescence intensity. In this way, an estimate of the enhancement
of the TPA itself is obtained (see the Experimental Section).
We observed that dye-attached nanoparticle films (thickness ≈
20 nm) showed a measurable but relatively small enhancement
of TPA by a factor of 5, relative to the same nanoparticles
dispersed in a PMMA film (see also the Experimental Section).
In contrast, the same dye bound to a Ag fractal cluster film
showed an enhancement in TPA (spatially averaged) by a factor
Figure 3. Laser power dependence of the two-photon-excited fluo-
rescence intensity from a polymer/silver-cluster composite film (circles),
neat polymer film (squares), and silver cluster film (triangles).
Experiments were conducted using a 3 mm diameter collimated laser
beam. Inset shows the same data on a double logarithmic plot. The
solid lines are quadratic fits to the data.
of 68°), were similar, showing that the enhancement is relatively
insensitive to the angle of incidence. This is rather different
behavior from that observed for two-photon-excited fluorescence
enhancement due to coupling to surface-plasmon modes of
smooth metal films,9 which occurs only at a particular angle of
incidence, as dictated by wavevector matching conditions. We
rule out a significant contribution of multiple scattering to the
observed enhancement in the composite films because the total
optical density of the silver cluster films, including absorptive
and scattering losses, was only in the range of 0.05 to 0.3 (90%
to 50% transmission).

Dyes on Silver Nanoparticle Fractal Clusters
J. Phys. Chem. B, Vol. 106, No. 27, 2002 6857
Figure 4. Two-photon laser-scanning fluorescence, scanning electron, and transmission optical microscopy images of polymer/silver-cluster
composites. (A) Plot of TPF intensity excited at 890 nm versus x,y position in the film, showing the presence of the hotspots. (B) False-color
overlay of TPF images obtained at three different excitation wavelengths: 720 nm (shown in blue), 820 nm (green), and 890 nm (red), showing
the frequency dependence of the hotspot positions (scale bar: 10µm). (C) Scanning electron microscopy image of the Ag fractal clusters deposited
on indium tin oxide (scale bar: 5µm). (D) Transmission optical microscopy image of the same area of the Ag cluster/polymer composite of Figure
4B (scale bar: 10 µm).
of ∼20 000. These results further emphasize the importance of
the particle spacing and spatial distribution in the enhancement
mechanism. Using TPF microscopy, we again observed an
8-fold increase of the peak to the average TPF signals and thus
deduced a peak TPA enhancement factor of >160 000 for the
chromophores attached to the fractal clusters. These TPA
enhancements are also much larger than those observed on
evaporated silver island films.10 This is consistent with the fact
that the nanoparticle aggregates formed in colloidal solution
through the mechanism of cluster-cluster aggregation (CCA)
have a fractal geometry32 that closely approximates that of the
simulated CCA aggregates for which extremely large enhance-
ments have been calculated.33 In contrast, the evaporated films
have a far more regular “droplet-like” structure.34
The large enhancement of two-photon absorption by metal
nanoparticle fractal clusters has potential for application in
ultrasensitive detection of molecular binding,35 imaging, and
localized activation of photochemical processes. An important
consequence of the enhancement of the effective TPA cross-
section by five orders-of-magnitude is that it allows for the use
of a several hundred-fold lower laser intensity to produce a given
fluorescence signal, which can be of significance in biological
applications. We believe that metal nanoparticle fractal clusters
are interesting systems with potential to increase the impact of
two-photon processes in photonics, materials processing, chemi-
cal sensing, and biological applications.
Conclusions
In conclusion, our studies show that organic chromophores
near the surface of clusters of silver nanoparticles with a random
fractal geometry exhibit large enhancements in two-photon
absorption and two-photon-excited fluorescence, relative to
isolated chromophores. In polymer coated films, chromophores
within j10 nm of the surface show an enhancement in TPF of
1000 and we estimate the peak values in the hotspots of the
clusters to be >10 000. For dyes that are chemically bound
within a distance of j15 Å, the spatially averaged enhancement
of TPA is ∼20 000, whereas the peak values are J160 000.
By comparing results on isolated ligand-coated nanoparticles
with attached dyes, partially ordered films of such particles,

6858 J. Phys. Chem. B, Vol. 106, No. 27, 2002
Wenseleers et al.
SCHEME 2
and fractal clusters with attached dyes, we conclude that
interactions of silver nanoparticles in fractal cluster arrangements
are critical to the magnitude of the enhancement, consistent with
the theoretical results of Shalaev et al. Furthermore, our studies
show that the enhancements are localized on the clusters, with
the locations of strong enhancement depending on excitation
wavelength, also in qualitative agreement with theory. Further
studies aimed at the optimization of these effects, including the
effect of distance and orientation of the chromophores relative
to the silver surface, and the use of other metals are in progress.
using a General Electric QE-300 spectrometer (300 MHz for
1H and 75 MHz for 13C), or using a Varian 500 spectrometer
1
(500 MHz for H and 125 MHz for 13C). If microanalyses are
not reported, the purity of the compounds were judged to be
1
>90% by H NMR and 13C NMR. All the reported yields are
isolated yields unless otherwise indicated. High-resolution mass
spectroscopy was performed by the Department of Chemistry,
University of California at Irvine
Silver Nanoparticle Cluster Film Preparation. Silver
nanoparticles were prepared by reduction of silver nitrate with
sodium citrate in aqueous solution.36 Cluster formation was
induced by adding 15 mL of 0.026 M aqueous fumaric acid to
500 mL of silver colloid solution.18 The amount of fumaric acid
was chosen so that cluster formation was clearly observable by
monitoring the optical absorption spectrum of the colloid (see
the Supporting Information). Precipitated clusters were deposited
on substrates over one to three days in a receptacle filled to a
height of 2.5 cm with the colloid. The substrates were then
carefully taken out of the solution, dried, and then immersed in
pure water again to remove any residual reagents. After the films
had been dried once, the silver was attached strongly enough
to the glass for further processing, such as solvent casting or
rinsing with solvents. The films were characterized by optical
absorption spectroscopy, optical microscopy, TPF microscopy,
scanning and transmission electron microscopy, and atomic force
microscopy. The particle mean diameter was found to be 10
nm from TEM images. The optical spectra of the silver fractal
films were very similar to those of the aqueous colloids, showing
the typical cluster band2 with an optical density in the range of
0.05 to 0.3. Upon sedimentation, as well as upon drying of the
Experimental Section
All solvents and reagents were purchased from Aldrich as
reagent grade and used as received. Spectrophotometric grade
solvents were used for optical measurements, cleaning, and film-
casting. Optical absorption spectra were measured using either
a diode array spectrometer (HP 8453) or a Varian Cary 5E UV-
vis-NIR spectrometer. Fluorescence spectra for fluorescence
quantum yield determinations were recorded using a Spex
Fluorolog 2. Fluorescence quantum yields were determined
relative to 9,10-bis-(phenylethynyl)-anthracene.26 Laser excited
emission spectra were recorded using a spectrograph (SpectraPro
150, Acton Research Corp.) with liquid nitrogen cooled CCD
detection (Princeton Instruments LN/CCD-1100-PB). Scanning
electron microscopy was performed on a Hitachi S-2460N. Gas
chromatography detected by mass spectrometry (GC-MS) was
performed using an HP 6890 gas chromatograph equipped with
a mass selective detector HP 5973. Elemental analyses were
performed by Atlantic Microlab Inc., Norcross, GA or by Desert
Analytics Laboratory, Tucson, AZ. NMR spectra were recorded

Dyes on Silver Nanoparticle Fractal Clusters
J. Phys. Chem. B, Vol. 106, No. 27, 2002 6859
films after sedimentation, the cluster band shifts to longer
wavelengthssconsistent with further aggregation.
NMR (300 MHz, CDCl3) δ 7.26 (d, J ) 7.6 Hz, 4H), 7.06 (d,
J ) 8.5 Hz, 2H), 7.04-6.96 (m, 6H), 6.64 (dd, J ) 17.6 Hz, J
) 10.9 Hz, 2H), 5.62 (dd, J ) 17.6 Hz, J ) 0.6 Hz, 2H), 5.14
(d, J ) 10.9 Hz, 2H), 2.56 (t, J ) 7.6 Hz, 2H), 1.63-1.51 (m,
2H), 1.44-1.30 (m, 2H), 0.93 (t, J ) 7.3 Hz, 3H). 13C NMR
(75 MHz, C3D6O) δ 148.4, 145.9, 139.3, 137.3, 132.9, 130.3,
128.1, 126.0, 124.2, 112.5, 35.7, 34.6, 23.1, 14.3. GC-MS: 353
(M+). El. Anal. Calcd. for C26H27N: C, 88.34; H, 7.70, N, 3.86.
Found: C, 86.08; H, 7.62; N, 3.81.
Coating Cluster Films with Thiolated Dye, B. The thiol-
substituted dye, B, was coated onto silver cluster films by
solvent evaporation from dichloromethane solution, followed
by repeated immersion in neat dichloromethane to remove
unbound dye. The complete removal of the unbound dye was
established by checking that after the last two immersions the
solvent showed no fluorescence.
4-Iodo-2,5-dihexylbenzaldehyde (4). To a solution of 1,4-
diiodo-2,5-dihexylbenzene (4.98 g, 10 mmol) in THF (80 mL)
at -78 °C under nitrogen was added t-BuLi (13.8 mL, 1.45 M
solution in pentane, 20 mmol) and the resultant solution was
stirred at this temperature for 30 min. Then, dimethylformamide
(2.3 mL, 30 mmol) was added to the solution and the mixture
was stirred at -78 °C for 1 h and at room temperature for 3 h.
The solvent was evaporated and the residue was separated
between ether and water layers. The organic layer was dried
with MgSO4 and the solvent removed using rotary evaporation
to yield 3.43 g of the crude product. Crystallization from ethanol
gave a small amount of relatively pure 4, whereas the compound
Synthesis of TPA Polymer, A. The following starting
materials were prepared according to literature methods: 1,4-
diiodo-2,5-dihexylbenzene37 and 2,5-dicyano-1,4-bis(bromom-
ethyl)benzene.38
4-Butyl-N,N-diphenyl-aniline (1). To a solution of tris-
(dibenzylideneacetone)dipalladium (Pd2dba3) (1.099 g, 1.2
mmol) and bis(diphenylphosphino)ferrocene (dppf) (0.998 g,
1.8 mmol) in toluene (180 mL) under nitrogen atmosphere was
added bromobenzene (28.264 g, 180 mmol) and the resultant
solution was stirred for 10 min at ambient temperature. To this
solution were then added sodium tert-butoxide (14.417 g, 150
mmol) and 4-butylaniline (8.954 g, 60 mmol). The resultant
solution was stirred at 90 °C for 60 h under nitrogen atmosphere.
After 36 h an additional amount of catalyst (Pd2dba3 (366 mg,
0.4 mmol) and dppf (333 mg, 0.6 mmol)) and base (sodium
tert-butoxide (4.806 g, 50 mmol)) was added to the solution.
The reaction mixture was cooled to room temperature and
separated between ether and water layers. The combined organic
layer was concentrated using rotary evaporation, and purified
using flash column chromatography (SiO2/hexane) to obtain
16.84 g (93%) of the product as a viscous, slightly yellow oil.
1H NMR (300 MHz, CDCl3) δ 7.27 (d, J ) 7.4 Hz, 2H), 7.25
(d, J ) 7.4 Hz, 2H), 7.15-6.96 (m, 10 H), 2.62 (t, J ) 7.7 Hz,
2H), 1.70-1.58 (m, 2H), 1.48-1.37 (m, 2H), 0.99 (t, J ) 7.3
Hz, 3H). 13C NMR (75 MHz, CDCl3) δ 148.0, 145.3, 137.7,
129.1, 129.0, 124.7, 123.6, 122.2, 35.0, 33.7, 22.4, 14.0.
1
seems to decompose during chromatography on silica gel. H
NMR (300 MHz, CDCl3) δ 10.19 (s, 1H), 7.73 (s, 1H) 7.57 (s,
1H), 2.88 (t, J ) 7.8 Hz, 2H), 2.70 (t, J ) 7.9 Hz, 2H), 1.56
(m, 4H), 1.30 (m, 12H), 0.87 (m, 6H).
Tetraethyl-(2,5-dicyano-r,r′-p-xylylenediphosphonate) (5).
314 mg (1 mmol) of 2,5-dicyano-1,4-bis(bromoethyl)benzene
and 500 mg (3 mmol) of triethyl phosphite were heated to 160
°C for 18 h. Then the excess of triethyl phosphite was removed
using rotary evaporation and the residue was crystallized from
hexane/ethyl acetate (1:1). 280 mg of beige crystals were
1
collected, yield: 65% H NMR (500 MHz, CDCl3) δ 7.80 (s,
2 H), 4.13 (m, 8 H, OCH2), 3.38 (d, J ) 21 Hz, 4 H, CH2),
1.30 (t, J ) 7.25 Hz, 12 H, CH3). 13C NMR (125 MHz, CDCl3)
δ 16.24, 31.83 (d, J ) 139 Hz), 62.71, 115.87, 117.45, 134.77,
135.64. El. Anal. Calcd. for C18H26N2O6P2 :C, 50.47, H, 6.12,
N, 6.54, P, 14.46, O, 22.41. Found: C, 50.76, H, 6.34, N, 6.65,
P, 14.22, O, 22.03.
4-Butyl-N,N-bis(4-bromophenyl)-aniline (2). To a solution
of 4-butyl-N,N-bis(phenyl)aniline (1) (6.024 g, 20 mmol) in
DMF (60 mL) was added N-bromosuccinimide (NBS) (7.112
g, 40 mmol). The resultant solution was stirred for 20 h at
ambient temperature. The reaction mixture was separated
between ether and water layers and the organic layer was
concentrated using rotary evaporation to yield 8.45 g (92%) of
the product as a slight brown, very viscous oil. Further
purification was achieved by using column chromatography
E,E-1,4-Bis(4-iodo-2,5-di-n-hexylstyryl)-2,5-dicyano-ben-
zene (6). To a stirred solution of the bisphoshonate (5) (535.2
mg, 1.25 mmol) and 4-iodo-2,5-dihexyl-benzaldehyde (4) (1.02
g, 2.54 mmol) in THF (10 mL) under nitrogen was slowly added
a suspension of KO-t Bu (291.2 mg, 2.6 mmol) in THF (5 mL).
After 2.5 h, the reaction was quenched with water and the
solution was separated between water and ether layers. The
organic layer was dried with MgSO4 and the solvent was
removed using rotary evaporation. The residue was purified
using column chromatography (SiO2/hexane/methylenchloride
(2:1)) to yield 580 mg (50% of the pure all-trans product 6).
1H NMR (300 MHz, CD2Cl2) δ 8.06 (s, 2 H), 7.70 (s, 2 H),
7.55 (d, J ) 16.1 Hz, 2 H), 7.49 (s, 2 H), 7.29 (d, J ) 16.1 Hz,
2 H), 2.70 (m, 8 H), 1.64-1.52 (m, 16 H), 1.45-1.26 (m, 16
H), 0.93-0.86 (m, 12 H). 13C NMR (75 MHz, CD2Cl2) δ 144.2,
141.7, 141.2, 139.8, 135.0, 132.8, 131.1, 127.0, 124.1, 117.2,
115.7, 102.1, 41.0, 33.0, 32.3, 31.9, 31.0, 29.7, 29.6, 23.3, 23.2,
14.4. El. Anal. Calcd. for C48H62I2N2: C, 62.61; H, 6.79; I,
27.56; N, 3.04. Found: C, 62.49; H, 6.98; I, 27.30; N, 2.84.
1
(SiO2/20% dichloromethane in hexane). H NMR (300 MHz,
CDCl3) δ 7.30 (dd, J ) 8.9 Hz, J ) 2.5 Hz, 4H), 7.07 (d, J )
8.5 Hz, 2H), 6.96 (d, J ) 8.5 Hz, 2H), 6.90 (dd, J ) 8.5 Hz, J
) 2.5 Hz, 4H), 2.56 (t, J ) 7.8 Hz, 2H), 1.63-1.51 (m, 2H),
1.42-1.30 (m, 2H), 0.93 (t, J ) 7.3 Hz). 13C NMR (75 MHz,
CDCl3) δ 146.7, 138.8, 132.2, 129.4, 125.0, 124.9, 114.9, 35.0,
33.6, 22.4, 14.0. HR-MS: calcd. for C22H21NBr2 457.0041 found
457.0039.
4-Butyl-N,N-bis(4-ethenylphenyl)-aniline (3). To a solution
of 4-butyl-N,N-bis(4-bromophenyl)aniline (2) (1.225 g, 2.660
mmol), Pd(PPh3)4 (61.69 mg, 0.053 mmol) and a few crystals
of 2,6-di-tert-butyl-4-methylphenol in toluene (7 mL) was added
tributyl(vinyl)tin (2.093 g, 6.6 mmol). The resulting solution
was heated to reflux for 20 h. After cooling the reaction mixture
was diluted with ether (100 mL) and washed with 2 N HCl and
water. The organic layer was dried with MgSO4 and the solvent
was removed using rotary evaporation. The residue was purified
using column chromatography (SiO2/20% dichloromethane in
hexane) to yield 453 mg (48%) of a slightly yellow impure oil,
TPA-Polymer Derived from 3 and 6 (A). 4-Butyl-N,N-bis-
(4-ethenylphenyl)-aniline (3) (176.6 mg, 0.5 mmol), E,E-1,4-
Bis(4-iodo-2,5-di-n-hexylstyryl)-2,5-dicyano-benzene (6) (460.2
mg, 0.5 mmol), Pd(OAc)2 (6.3 mg, 0.025 mmol), P(o-tolyl)3
(39 mg, 0.25 mmol) were dissolved in DMF (8 mL), toluene
(5 mL) and triethylamine (0.29 mL) under nitrogen atmosphere.
Then the solution was heated at 80 °C for 20 h. After this time
1
that was used in the next step without further purification. H

6860 J. Phys. Chem. B, Vol. 106, No. 27, 2002
Wenseleers et al.
4-butyl-N,N-bis(4-ethenylphenyl)-aniline (25 mg, 0.07 mmol)
was added, because thin-layer chromatography still showed
some 6, but no 3. After 44 h Pd(OAc)2 (4.7 mg, 0.019 mmol),
P(o-tolyl)3 (30 mg, 0.19 mmol) and an additional amont of
4-butyl-N,N-bis(4-ethenylphenyl)-aniline (29 mg, 0.08 mmol)
were added and the solution was heated for further 22 h at 100
°C. Then the reaction mixture was added dropwise into methanol
(500 mL) and the precipitate was collected by filtration. The
polymer was redissolved in a small amount of dichloromethane
and added dropwise into hexane (500 mL). After filtration the
product was dried under vacuum. The yield of the orange
phosphite (2.5 mL, 14.5 mmol) were heated to 160 °C for 8 h.
After removing the volatiles, the product was chromatographed
on silica 60 (ethyl acetate) to give 3.34 g (81%) of a white
1
solid. H NMR (500 MHz, CDCl3) δ: 6.89 (s, 1H), 6.87 (s,
1H), 4.04-3.96 (m, 10H), 3.89 (t, J ) 6.5 Hz, 2H, -OCH2-),
3.77 (s, 3H, -OCH3), 3.21 (d, JP-H ) 21.0 Hz, 2H), 3.18 (d,
JP-H ) 21.0 Hz, 2H), 2.02 (s, 3H, -CH3), 1.73 (m, 2H), 1.59
(m, 2H), 1.41 (m, 2H), 1.26 (m, 12H), 1.21 (t, J ) 7.0 Hz,
12H, P-OCH2CH3) ppm; 13C NMR (125 MHz, CDCl3) δ:
171.49, 151.12, 150.68, 119.80, 119.47, 115.24, 114.13, 69.22,
64.87, 62.14, 62.09, 62.04, 56.33, 29.80, 29.74, 29.69, 29.47,
28.81, 27.23, 27.03, 26.35, 26.13, 25.93, 21.26, 16.60, 16.56
ppm.
1
polymer was 400 mg (79%). H NMR (300 MHz, CDCl3) δ
8.03 (bs, 2 H), 7.63-6.97 (m, 24 H), 2.85-2.65 (m, 8 H), 2.65-
2.52 (m, 2 H), 1.71-1.52 (m, 10 H), 1.50-1.26 (m, 26 H),
0.95-0.82 (m, 15 H) 13C NMR (75 MHz, CD2Cl2) δ 144.2,
141.7, 141.2, 139.8, 135.0, 132.8, 131.1, 127.0, 124.1, 117.2,
115.7, 102.1, 41.0, 33.0, 32.3, 31.9, 31.0, 29.7, 29.6, 23.3, 23.2,
14.4. GPC: Mn ) 12030, Mw ) 31220, determined relative to
a polystyrene standard.
11-(2,5-bis{(E)-2-[4-(diethylamino)phenyl]ethenyl}-4-meth-
oxyphenoxy)-1-undecanol (10): To 11-{2,5-bis[(diethoxy-
phosphoryl)methyl]-4-methoxyphenoxy}undecyl acetate (9) (5.57
g, 8.75 mmol), 4-(N,N-diethylamino)benzaldehyde (3.102 g,
17.5 mmol) in THF (125 mL),was added KOtBu (22.75 mL of
1M solution in THF) at room temperature. The reaction was
stirred with NaOH solution (30 mL, 3M solution) for 1 day.
CH2Cl2 was added, the organic layer separated, dried over
MgSO4, filtered and evaporated. Chromatography (ethyl ac-
etate: hexanes 1:3) and recrystallization from toluene gives 1.42
g (25%) product. 1H NMR (500 MHz, CDCl3) δ: 7.40 (d, J )
8.5 Hz, 2H), 7.39 (d, J ) 8.5 Hz, 2H), 7.25 (d, J ) 16.0 Hz,
1H, E-CHdCH-), 7.22 (d, J ) 16.5 Hz, 1H, E-CHdCH-), 7.09
(s, 1H), 7.07 (s, 1H), 7.02 (d, J ) 16.5 Hz, 1H, E-CHdCH-),
6.98 (d, J ) 16.5 Hz, 1H, E-CHdCH-), 6.65 (d, J ) 8.0 Hz,
2H), 6.64 (d, J ) 9.0 Hz, 2H), 4.01 (t, J ) 6.5 Hz, 2H, -OCH2-
), 3.88 (s, 3H, -OCH3), 3.60 (t, J ) 6.0 Hz, 2H, -CH2O-),
3.37 (t, J ) 6.5 Hz, 4H, -NCH2-), 3.36 (t, J ) 6.5 Hz, 4H,
-NCH2-), 1.84 (m, 2H, -CH2-), 1.53 (m, 4H), 1.38 (m, 2H),
1.29 (m, 10H), 1.17 (m, 12H, -NCH2CH3) ppm; 13C NMR (125
MHz, CDCl3) δ: 151.41, 150.96, 150.45, 147.41, 128.75 128.59,
128.06, 128.00, 127.11, 126.77, 125.75, 125.66, 118.81, 118.48,
111.92, 111.89, 110.54, 108.78, 69.95, 63.31, 56.68, 44.61
(-NCH2-), 33.03, 29.84, 29.79, 29.76, 29.66, 26.49, 25.94,
12.89 (-NCH2CH3) ppm. El. Anal. Calcd. for C42H60N2O3:
C: 78.70 H: 9.44 N: 4.37 Found: C: 78.05, H: 9.59 N: 4.28.
Synthesis of the Thiol Functionalized Chromophore. 11-
(4-methoxyphenoxy)undecan-1-ol (7): NaH (4.8 g, 200 mmol)
was suspended in dry THF (250 mL) and 4-methoxyphenol
(24.7 g, 199 mmol) was added portionwise. The reaction was
cooled with an ice bath and after the gas evolution had ended
11-bromoundecanol (50 g, 199 mmol) was added. The reaction
was stirred for 1 h at room temperature and then heated to reflux.
The progress of the reaction was followed by GC-MS. After
19 h, the reaction was completed. Water and CH2Cl2 were added
and the organic layer was separated. The water layer was washed
two times with CH2Cl2. The combined organic layers were dried
over MgSO4, were filtered, and the solvent was evaporated.
Recrystallization from CH2Cl2 gives 51.81 g white solid
1
(88.4%). H NMR (500 MHz, CDCl3) δ 6.81 (br s, 4H), 3.88
(t, J ) 6.5 Hz, 2H, -OCH2-), 3.75 (s, 3H, -OCH3), 3.62 (t, J
) 6.5 Hz, 2H, HOCH2-), 1.73 (m, 2H), 1.54 (m, 2H), 1.42 (m,
2H), 1.27 (m, 12H) ppm;13C NMR (125 MHz, CDCl3) δ:153.62,
153.27, 115.40, 114.59, 70.98, 68.63, 63.10, 55.74, 32.81, 29.56,
29.55, 29.51, 29.43, 29.40, 26.20, 26.05, 25.73 ppm. GC-MS:
(19.11 min) m/z (%): 294 (17) M, 124 (100) M - C11H22O.
El. Anal. Calcd. for C18H30O3: C: 73.43 H: 10.27 Found: C:
73.55, H: 10.38.
4-[(E)-2-(2-[(11-bromoundecyl)oxy]-4-{(E)-2-[4-(diethyl-
amino)phenyl]ethenyl}-5-methoxyphenyl)ethenyl]-N,N-di-
ethylaniline (11): 11-(2,5-bis{(E)-2-[4-diethylamino)phenyl]-
ethenyl}-4-methoxyphenoxy)-1-undecanol (10) (252 mg, 0.393
mmol) and CBr4 (163 mg, 0.492 mmol) were dissolved in dry
CH2Cl2 (2 mL) and cooled with an ice bath. Triphenylphosphine
(155 mg, 0.590 mmol) in CH2Cl2 (2 mL) was added dropwise
and the resulting solution was stirred for 5 min. Water was added
and the solution was made basic with K2CO3 solution. The
organic layer was separated and washed with water twice, dried
over MgSO4, filtered and evaporated. Flash chromatography
(ethyl acetate:hexanes 2:7) gave a yellowish oil that was taken
up in toluene from which slightly brown crystals grow upon
cooling. 290 mg of the product were collected and the slightly
impure product was used without further purification.
1H NMR (500 MHz, CDCl3) δ: 7.40 (d, J ) 8.0 Hz, 2H),
7.39 (d, J ) 8.5 Hz, 2H), 7.08 (s, 1H), 7.07 (s, 1H), 6.64 (m,
4H), 4.01 (t, J ) 6.5 Hz, 2H), 3.88 (s, 3H, -OCH3), 3.37 (t, J
) 7.0 Hz, 8H, -NCH2-),1.83 (m, 4H), 1.52 (m, 4H), 1.38 (m,
4H), 1.28 (m, 6H), 1.16 (m, 12H, -NCH2CH3) ppm.
11-(2,5-bis{(E)-2-[4-(diethylamino)phenyl]ethenyl}-4-meth-
oxyphenoxy)undecan-1-thiol (B): To a solution of 4-[(E)-2-
(2-[(11-bromoundecyl)oxy]-4-{(E)-2-[4-(diethylamino)phenyl]-
ethenyl}-5-methoxyphenyl)ethenyl]-N,N-diethylaniline (11) (276
mg, 0.39 mmol) in dry THF (5 mL) cooled to -10 °C was
11-[2,5-Bis(bromomethyl)-4-methoxyphenoxy]undecyl ac-
etate (8): To a suspension of 11-(4-methoxyphenoxy)undecan-
1-ol (7) (51.8 g, 177 mmol) and p-formaldehyde (10.57 g, 352
mmol) in CH3COOH (250 mL) was added HBr (100 mL of a
33% HBr solution in CH3COOH) at once. The suspension
became clear and the reaction was kept at 80 °C for 30 h, and
the progress was followed by GC-MS. After pouring into water
(300 mL), the white precipitate was filtered off, washed with
water and dried using rotary evaporation. The product contained
slight impurities but was used without further purification steps.
Crude yield: 28.14 g (30.5%). 1H NMR (500 MHz, CDCl3) δ:
6.84 (s, 1H), 6.83 (s, 1H), 4.51 (s, 2H, -CH2Br), 4.50 (s, 2H,
-CH2Br), 4.03 (t, J ) 6.5 Hz, 2H, -OCH2-), 3.96 (t, J ) 6.0
Hz, 2H, -OCH2-), 3.84 (s, 3H, -OCH3), 3.38 (t, J ) 7.0 Hz,
2H), 2.02 (s, 3H, -CH3), 1.79 (m, 2H), 1.59 (m, 2H), 1.40 (m,
2H), 1.27 (m, 12H) ppm;13C NMR (125 MHz, CDCl3) δ:
171.48, 151.29, 150.99, 127.74, 127.51, 114.91, 113.90, 69.17,
64.87, 56.42, 29.74, 29.69, 29.63, 29.51, 29.46, 29.43, 28.93,
28.88, 28.80, 26.27, 26.12, 21.26 ppm. El. anal.calcd. for C22H34-
Br2O4: C: 50.59 H: 6.56 found C: 50.57, H: 6.82.
11-{2,5-bis[(diethoxyphosphoryl)methyl]-4-methoxyphen-
oxy}undecyl acetate (9): 11-[2,5-bis(bromomethyl)-4-meth-
oxyphenoxy]undecyl acetate (8) (3.40 g, 6.5 mmol) and triethyl

Dyes on Silver Nanoparticle Fractal Clusters
J. Phys. Chem. B, Vol. 106, No. 27, 2002 6861
added ((CH3)3Si)2S (100 µL, 0.47 mmol). After stirring for 5
min. tetra-n-butylammonium fluoride (TBAF) (430 µL 1M
solution in THF, 0.43 mmol) was added at once, and the reaction
was allowed to warm to room temperature. After stirring for
another 30 min, CH2Cl2 was added and the solution was washed
with saturated NH4Cl, dried over MgSO4, filtered and evapo-
rated. Flash chromatography (ethyl acetate:hexanes 1:5) gave
giving ∼100 fs pulses at a repetition rate of 82 MHz, a
wavelength tunable over the range from ∼700 to 1050 nm, and
an average power up to ∼500 mW, and using an optical
parametric oscillator (OPO) system (Spectra Physics Pro250/
MOPO combination) giving ∼6 ns pulses at a repetition rate
of 10 Hz.
For the measurements of the two-photon excitation spectrum
of the polymer in solution, the laser beam was expanded to a
diameter of ∼1 cm and then focused using a 76 mm focal length
lens into a 1 cm path length cell. The TPF was collected in a
right angle geometry and separated from any scattered laser light
using a combination of short wavelength passing filters and a
monochromator. The TPF was detected at 570 nm for both the
sample and a reference solution with known ηδ (Coumarin 307
in methanol41), to obtain ηδ of the sample, after correction for
the different emission spectra of the sample and reference
standard, which are taken to be the same as the one-photon
excited emission spectra. TPF excitation measurements with the
nanosecond OPO system were performed according to methods
described previously.24
One- and two-photon-excited fluorescence measurements on
the dye-coated particles and free dye in toluene were performed
at similar dye concentrations and were normalized for differ-
ences in concentration. The weight percentage of the dye in
the dye coated nanoparticles was determined from the elemental
analysis. The absorption spectrum of the dye coated nanopar-
ticles closely matched the sum of the spectra of the nanoparticles
without dye and that of the free dye (all in toluene), suggesting
that the dye bound to the nanoparticles has essentially the same
extinction coefficient as the free dye.
1
160 mg (62%) product. H NMR (500 MHz, CDCl3) δ: 7.40
(d, J ) 9.0 Hz, 2H), 7.39 (d, J ) 8.5 Hz, 2H), 7.25 (d, J )
16.0 Hz, 1H, E-CHdCH-), 7.22 (d, J ) 16.0 Hz, 1H, E-CHd
CH-), 7.09 (s, 1H), 7.07 (s, 1H), 7.03 (d, J ) 16.5 Hz, 1H,
E-CHdCH-), 6.98 (d, J ) 16.5 Hz, 1H, E-CHdCH-), 6.65 (d,
J ) 9.0 Hz, 2H), 6.64 (d, J ) 9.0 Hz, 2H), 4.02 (t, J ) 6.5 Hz,
2H, -OCH2-), 3.89 (s, 3H, -OCH3), 3.37 (m, 8H, -NCH2-),
2.49 (q, J ) 7.5 Hz, 2H, -CH2SH), 1.84 (m, 2H), 1.58 (m,
2H), 1.54 (m, 2H), 1.32 (m, 12H), 1.17 (m, 12H, -NCH2CH3);
13C NMR (125 MHz, CDCl3) δ: 151.41, 150.96, 147.41,
147.38, 128.75, 128.60, 128.06, 128.00, 127.11, 126.77, 125.72,
125.65, 118.80, 118.48, 111.89, 110.53, 108.78, 69.93, 56.68,
44.61 (-NCH2-), 34.30, 29.93., 29.84, 29.77, 29.67, 29.31,
28.62, 28.07, 26.50, 24.90, 12.90 (-NCH2CH3). El. Anal. Calcd.
for C42H50N2O2S: C: 76.78, H: 9.20, N: 4.26, S: 4.88 Found
C: 76.61, H: 9.59, N: 4.41, S: 5.01.
Synthesis of Solubilized, Chromophore Bound Silver
Nanoparticles. Place exchange reactions39 on nanoparticles were
performed by dissolving the thiol-functionalized chromophore
in a solution of silver nanoparticles (mean diameter ) 10 nm)
coated with a 3:1 mixture of dodecyl- and octyl-thiol (synthe-
sized using a modification of a literature method40) in CH2Cl2
and stirring for 4 days at room temperature.
TPF measurements on films were performed using the
collimated ∼3 mm diameter beam, at incident angles between
0° and 45°, and collecting the emission at an angle between
60° and 90° to the incident beam (chosen to avoid specular
reflection of the incident beam through the collecting lens). The
emission was analyzed using a spectrograph with liquid nitrogen
cooled CCD detection. For the measurements on the chro-
mophoric polymer coated silver cluster films, the enhancement
factors were determined relative to the TPF signal from neat
polymer films of the same thickness. The mass of polymer per
surface area in these films was determined gravimetrically from
the known concentration and amount (1-2 mL) of the solutions
cast on the given surface area of the substrate (e.g., 19 cm2).
Nominal polymer thicknesses were subsequently obtained by
assuming a density of 1200 kg/m3, as for similar polymers. Thus,
being a ratio of TPF signals for the same amount of polymer,
the enhancement factor determined for these samples reflects
the enhancement of the two-photon-excited fluorescence, in other
words the effective enhancement of the product ηδ.
For all the other measurements on films, enhancement factors
were still determined by comparison of the TPF signals of films
under study with TPF signals of control samples (dye-doped
PMMA films) without silver and containing approximately the
same amount of dye, with normalization performed using the
one-photon excited fluorescence signals as a relative measure
of concentration. In this alternative referencing scheme, the
enhancement factor is thus calculated as
Octylthiol/dodecylthiol Protected Silver Nanoparticles:
AgNO3 (3.06 g, 18 mmol) and ethanol (2700 mL) were placed
in a 4L threeneck round-bottom flask with a mechanical stirrer.
After complete dissolution of the silver salt, octylthiol (670 µL,
4 mmol) and dodecyl dodecylthiol (479 µL, 2 mmol) were
injected simultaneously with vigorous stirring resulting in a
milky white dispersion. The solution was stirred for 5 min. Then
NaBH4 (1000 mL saturated solution in EtOH) was added very
slowly. After addition of a few drops of NaBH4, the solution
became purple and then the rate of addition was slightly
increased. Upon complete addition, after ca. 4 h, the reaction
mixture was stirred for another 1.5 h and then kept in a
refrigerator overnight, in order to allow the nanoparticles to
precipitate from solution. The solution was filtered carefully
over quantitative filter paper, and the filtrate was washed with
water and acetone, and dried under vacuum. Yield: 1.98 g
(68%).
Bis(diethylamino)bis(styryl)benzene Functionalized Nano-
particles: Octylthiol/dodecylthiol protected silver nanoparticles
(150 mg) and 11-(2,5-bis{(E)-2-[4-(diethylamino)phenyl]ethe-
nyl}-4-methoxyphenoxy)undecan-1-thiol, B, (50 mg, 0.076
mmol), were dissolved in CH2Cl2 (300 mL) and stirred for 4
days. After evaporation of the solvent the nanoparticles were
suspended in acetone, filtered, and washed extensively with
acetone until excess dye was removed completely, as checked
by monitoring the fluorescence of the extract. Isolated yield:
80.5 mg. The nanoparticles were characterized by UV-vis and
fluorescence spectroscopy, elemental analysis (found (%) C:
19.43 H: 2.81 N: 0.62 S: 3.15 Ag: 71.08). The molar ratio
dye:octylthiol:dodecylthiol was determined to be 1.2:4.2:1 from
the elemental analysis.
(2)
comp
(2)
(1)
(S
- S
)/S
control comp
(2)
dye dye
S
/S (1)
Two-Photon-Excited Fluorescence Measurements. TPF
measurements were conducted using a cw mode-locked Ti:
sapphire laser (Spectra-Physics Millennia/Tsunami combination)
where Scomp, Scontrol, and Sdye are the signals for the composite,
the background signal (if any) from the control film without
dye, and the signal from the reference dye doped PMMA film,

6862 J. Phys. Chem. B, Vol. 106, No. 27, 2002
Wenseleers et al.
(12) Goncher, G. M.; Parsons, C. A.; Harris, C. B. J. Phys. Chem. 1984,
88, 4200-4209.
respectively. The superscripts 1 and 2 refer to fluorescence
signals obtained under one- and two-photon excitation, respec-
tively. No significant emission from the control samples was
detected under one-photon excitation, nor was there any
significant background for PMMA films without dye. This
alternative referencing scheme may be justified by the fact that
the theoretical models of the plasmon localization in fractal
clusters predict no net enhancement (spatially averaged) for
linear optical effects because the fraction of nanoparticles
involved in the localized plasmon modes is inversely propor-
tional to the peak enhancement factor.42 However, a small
enhancement of the one-photon excited fluorescence cannot be
excluded. Because of the normalization using the one-photon
excited fluorescence signals in this alternative referencing
scheme, any changes in the fluorescence quantum yield are also
eliminated. In other words, these measurements are effectively
normalized for the concentration as well as the fluorescence
quantum yield η, and for any enhancement of the linear
absorption. The enhancement factors obtained for these film
samples reflect an effective enhancement of the ratio δ/σ, where
σ is the linear absorption cross-section. Thus, neglecting
enhancement of the linear absorption, the enhancement factors
reported for these samples can be interpreted as an effective
two-photon absorption enhancement (effective δ; as opposed
to the effective TPF enhancement, or effective ηδ, reported for
the polymer coated fractal cluster films). As a smaller enhance-
ment of the linear absorption cannot be excluded, these TPA
enhancement factors represent lower limits.
(13) Shalaev, V. M.; Stockman, M. I. Zh. Eksp. Teor. Fiz. 1987, 92,
509-522 [SoV. Phys. JETP 1987, 65, 287-294].
(14) Shalaev, V. M.; Botet, R.; Butenko, A. V. Phys. ReV. B 1993, 48,
6662-6664.
(15) Mandelbrot, B. M. The Fractal Geometry of Nature (Freeman, San
Francisco, 1982).
(16) Tsai, D. P.; Kovacs, J.; Wang, Z. H.; Moskovits, M.; Shalaev, V.
M.; Suh, J. S.; Botet, R. Phys. ReV. Lett. 1994, 72, 4149-4152.
(17) Pettinger, B.; Picardi, G.; Schuster, R.; Ertl, G. Electrochem. 2000,
68, 942-949.
(18) Butenko, A. V.; Shalaev, V. M.; Stockman, M. I. Z. Phys. D 1988,
10, 81-92.
(19) Kim, W.; Safonov, V. P.; Shalaev, V. M.; Armstrong, R. L. Phys.
ReV. Lett. 1999, 82, 4811-4814.
(20) Reduction of lasing threshold has also been reported in titania fractal
colloids [Wang, H. Z.; Zhao, F. L.; He, Y. J.; Zheng, X. G.; Huang, X. G.;
Wu, M. M. Opt. Lett. 1998, 23, 777-779].
(21) Butenko, A. V.; Chubakov, P. A.; Danilova, Y. E.; Karpov, S. V.;
Popov, A. K.; Rautian, S. G.; Safonov, V. P.; Slabko, V. V.; Shalaev, V.
M.; Stockman, M. I. Z. Phys. D. 1990, 17, 283-289.
(22) Sa´nchez-Lo´pez, J. C.; Ferna´ndez, A. Act. Mater. 2000, 48, 3761-
3771.
(23) Albota, M.; Beljonne D.; Bre´das, J.-L.; Ehrlich, J. E.; Fu, J. Y.;
Heikal, A. A.; Hess, S. E.; Kogej, T.; Levin, M. D.; Marder, S. R.; McCord-
Maughon, D.; Perry, J. W.; Ro¨ckel, H.; Rumi, M.; Subramaniam, C.; Webb,
W. W.; Wu, X. L.; Xu, C. Science 1998, 281, 1653-1656.
(24) Rumi, M.; Ehrlich, J. E.; Heikal, A. A.; Perry, J. W.; Barlow, S.;
Hu, Z. Y.; McCord-Maughon, D.; Parker, T. C.; Ro¨ckel, H.; Thayumanavan,
S.; Marder, S. R.; Beljonne, D.; Bre´das, J.-L. J. Am. Chem. Soc. 2000,
122, 9500-9510.
(25) Xu, C.; Webb, W. W. J. Opt. Soc. Am. B 1996, 13, 481-491.
(26) Berlman, I. B. Handbook of fluorescence spectra of aromatic
molecules (Academic Press: New York, 1971)
(27) A background emission from the neat Ag cluster film was observed.
This emission was much weaker and broader than the TPF from the Ag/
polymer composites, and also showed a quadratic power dependence. A
broadband background, but with a different spectral distribution and higher
power dependence on incident intensity, has also been observed in studies
of surface enhanced second-harmonic generation from roughened silver
surfaces (Chen, C. K.; de Castro, A. R. B.; Shen, Y. R. Phys. ReV. Lett.
1981, 46, 145-148). “White-light” generation and second harmonic
generation has also been previously observed for fractal aggregates of Ag
particles (Ducourtieux, S.; Gresillon, S.; Boccara, A. C.; Rivoal, J. C.;
Quelin, X.; Gadenne, P.; Drachev, V. P.; Bragg, W. D.; Safonov, V. P.;
Podolosky, V. A.; Ying, Z. C.; Armstrong, R. L.; Shalaev, V. M. J.
Nonlinear Opt. Phys. Mat. 2000, 9, 105-116). Other possibly related
phenomena include the background often observed in SERS (Michaels, A.
M.; Nirmal, M.; Brus, L. E. J. Am. Chem. Soc. 1999, 121, 9932-9939),
the photoinduced fluorescence observed from metal nanoparticles (Peyser,
L. A.; Vinson, A. E.; Bartko, A. P.; Dickson, R. M. Science 2001, 291,
103-106; Mohamed, M. B.; Volkov, V.; Link, S.; El-Sayed, M. A. Chem.
Phys. Lett. 2000, 317, 517-523). Further study is needed to understand
the origin of this background.
Acknowledgment. Support of this work by the NSF Bio-
photonics Program (Grant No. BES0086944), the Office of
Naval Research (through the CAMP MURI program, Grant No.
N0001401106), and the National Institutes of Health (Grant No.
5R01RR14312-02) is gratefully acknowledged. TMF acknowl-
edges support from the DFG. The authors thank Mr. Phil
Anderson for technical assistance with the TEM imaging.
Supporting Information Available: Optical absorption
spectra of silver colloids and sedimented films. Transmission
electron microscopy images of nanoparticle films. This material
is available free of charge via the Internet at http://pubs.acs.org.
References and Notes
(28) Markel, V. A.; Shalaev, V. M.; Zhang, P.; Huynh, W.; Tay, L.;
Haslett, T. L.; Moskovits, M. Phys. ReV. B 1999, 59, 10 903-10 909.
(29) Perhaps surprisingly, this resolution is comparable to the resolution
obtained in some near field optical imaging studies of the local intensity
distribution near metal fractal cluster surfaces [see e.g., refs 14, 28, and
(Bozhevolnyi, S. I.; Markel, V. A.; Coello, V.; Kim, W.; Shalaev, V. M.;
Phys. ReV. B. 1998, 58, 11 441-11 448.)]. Although the resolution of near-
field optical techniques can in general be higher, it is limited when applied
to rough surfaces because of the finite tip size and especially, when using
the constant intensity mode of controlling the tip height, because of the
large spatial intensity variations which cause the tip to move further from
the sample when approaching a hotspot. In contrast, the resolution of two-
photon microscopy, albeit generally lower than that of near-field techniques,
is not dependent on surface topography and even allows imaging through
an arbitrarily thick polymer film. Another advantage of two-photon
microscopy over near-field techniques is that no tip is involved. Scanning
a near-field tip, which is typically metal coated or made entirely out of
metal, in close proximity to the metal fractal clusters would itself be expected
to significantly change the coupled plasmon modes as the image is acquired.
(30) Sarychev, A. K.; Shubin, V. A.; Shalaev, V. M. Phys. ReV. B 1999,
60, 16 389-16 408.
(1) Gehr, R. J.; Boyd, R. W. Chem. Mater. 1996, 8, 1807-1819.
(2) Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Chem.
ReV. 1999, 99, 2957-2975.
(3) Boyd, G. T.; Rasing, T.; Leite, J. R. R.; Shen, Y. R. Phys. ReV. B
1984, 30, 519-526.
(4) Shalaev, V. M. Phys. Rep. 1996, 272, 61-137.
(5) Denk, W.; Strickler, J. H.; Webb, W. W. Science 1990, 248, 73-
76.
(6) Parthenopoulos, D. A.; Rentzepis, P. M. Science 1989, 245, 843-
845.
(7) Cumpston, B. H.; Ananthavel, S. P.; Barlow, S.; Dyer, D. L.;
Ehrlich, J. E.; Erskine, L. L.; Heikal, A. A.; Kuebler, S. M.; Lee, I.-Y. S.;
McCord-Maughon, D.; Qin, J.; Ro¨ckel, H.; Rumi, M.; Wu, X.-L.; Marder,
S. R.; Perry, J. W. Nature 1999, 398, 51-54.
(8) Bhawalkar, J. D.; He, G. S.; Prasad, P. N. Rep. Prog. Phys. 1996,
59, 1041-1070.
(9) Kano, H.; Kawata, S. Opt. Lett. 1996, 21, 1848-1850. As the
enhancement on these smooth films is spatially homogeneous, there is no
distinction between a “peak” or “spatially averaged” enhancement factor.
However, the enhancement only occurs at a very specific incident angle,
and the factor of 90 mentioned represents the peak enhancement occurring
for an incident angle of 56.4°.
(10) Glass, A. M.; Wokaun, A.; Heritage, J. P.; Bergman, J. G.; Liao,
P. F.; Olson, D. H. Phys. ReV. B 1981, 24, 4906-4909. The factor of 150
mentioned is the area-averaged enhancement factor.
(11) Sanchez, E. J.; Novotny, L.; Xie, X. S. Phys. ReV. Lett. 1999, 82,
4014-4017.
(31) Sandhyarani, N.; Pradeep, T.; Chakrabarti, J.; Yousuf, M.; Sahu,
H. K. Phys. ReV. B 2000, 62, R739-R742.
(32) Jullien, R.; Botet, R., Aggregation and Fractal Aggregates; World
Scientific: Singapore, 1987.
(33) Shalaev, V. M.; Poliakov, E. Y.; Markel, V. A. Phys. ReV. B. 1996,
53, 2437-2449.
(34) Sennett, R. S.; Scott, G. D. J. Opt. Soc. Am. 1950, 40, 203-211.

Dyes on Silver Nanoparticle Fractal Clusters
J. Phys. Chem. B, Vol. 106, No. 27, 2002 6863
(35) See for example: Liebermann, T.; Knoll, W.; Sluka, P.; Hermann,
R. Colloids & Surfaces A, 2000, 169, 337-350.
(39) Ingram, R. S.; Hostetler, M. J.; Murray, R. W. J. Am. Chem. Soc.
1997, 119, 9175-9178.
(36) Lee, P. C.; Meisel, D. J. Phys. Chem. 1982, 86, 3391-3395.
(37) Mangel, T.; Eberhardt, A.; Scherf, U.; Bunz, U. H. F.; Mulllen, K.
Macromol. Rapid Comm. 1995, 16, 571.
(40) Kang, S. Y.; Kim, K. Langmuir 1998, 14, 226-230.
(41) Xu, C.; Williams, R. M.; Zipfel, W.; Webb, W. W. Bioimaging
1996, 4, 198-207.
(38) Xiao, Y.; Yu, W. L.; Chua, S. J.; Huang, W. Chem. Eur. J. 2000,
6, 1318-1321.
(42) Shalaev, V. M.; Botet, R.; Tsai, D. P.; Kovacs, J.; Moskovits, M.
Phys. A 1994, 207, 197-207.