Directed assembly of gold nanorods through the use of dithiolated molecular linkers is one of the most efficient methodologies for the morphologically controlled tip-to-tip assembly of this type of anisotropic nanocrystals. However, in a direct analogy to molecular polymerization synthesis, this process is characterized by difficulties in chain-growth control over nanoparticle oligomers. In particular, it is nearly impossible to favor the formation of one type of oligomer, making the methodology hard to use for actual applications in nanoplasmonics. We propose here a light-controlled synthetic procedure that allows obtaining selected plasmonic oligomers in high yield and with reaction times in the scale of minutes by irradiation with low fluence near-infrared (NIR) femtosecond laser pulses. Selective inhibition of the formation of gold nanorod n-mers (trimers) with a longitudinal localized surface plasmon in resonance with a 800 nm Ti:sapphire laser, allowed efficient trapping of the (n - 1)-mers (dimers) by hot spot mediated photothermal decomposition of the interparticle molecular linkers. Laser irradiation at higher energies produced near-field enhancement at the interparticle gaps, which is large enough to melt gold nanorod tips, offering a new pathway toward tip-to-tip welding of gold nanorod oligomers with a plasmonic response at the NIR. Thorough optical and electron microscopy characterization indicates that plasmonic oligomers can be selectively trapped and welded, which has been analyzed in terms of a model that predicts with reasonable accuracy the relative concentrations of the main plasmonic species.
Directed assembly of gold nanorods through the use of dithiolated molecular linkers is one of the most efficient methodologies for the morphologically controlled tip-to-tip assembly of this type of anisotropic nanocrystals. However, in a direct analogy to molecular polymerization synthesis, this process is characterized by difficulties in chain-growth control over nanoparticle oligomers. In particular, it is nearly impossible to favor the formation of one type of oligomer, making the methodology hard to use for actual applications in nanoplasmonics. We propose here a light-controlled synthetic procedure that allows obtaining selected plasmonic oligomers in high yield and with reaction times in the scale of minutes by irradiation with low fluence near-infrared (NIR) femtosecond laser pulses. Selective inhibition of the formation of gold nanorod n-mers (trimers) with a longitudinal localized surface plasmon in resonance with a 800 nm Ti:sapphire laser, allowed efficient trapping of the (n - 1)-mers (dimers) by hot spot mediated photothermal decomposition of the interparticle molecular linkers. Laser irradiation at higher energies produced near-field enhancement at the interparticle gaps, which is large enough to melt gold nanorod tips, offering a new pathway toward tip-to-tip welding of gold nanorod oligomers with a plasmonic response at the NIR. Thorough optical and electron microscopy characterization indicates that plasmonic oligomers can be selectively trapped and welded, which has been analyzed in terms of a model that predicts with reasonable accuracy the relative concentrations of the main plasmonic species.
Research on gold nanoparticles (AuNPs) is of increasing interest in
connection with their unique optical properties, which originate from
the interaction of light with free conduction electrons. This phenomenon
is known as localized surface plasmon resonance (LSPR)[1,2] and has shown great potential for applications in sensing,[3,4] photovoltaics, and photocatalysis,[5,6] as well as
in waveguides and metamaterials,[7,8] to cite a few examples.
In this context, many of such applications are based on the collective
optical properties of AuNPs that are in close proximity of each other.[2] When the distance between two AuNPs is sufficiently
short, new hybridized plasmon modes appear through LSPR coupling,[9] in such a way that the intensity of the resulting
interaction is controlled by the interparticle distance.[10−12] In the case of anisotropic AuNPs, such as gold nanorods (AuNRs),
dimer structures display hybridized bonding and antibonding resonance
modes, which are highly sensitive to the relative orientation of the
nanocrystals and the separation distance.[11,13] For tip-to-tip AuNR assemblies, the bonding longitudinal mode registers
a significant redshift and an increase of the effective polarizability
with respect to the monomer (single AuNR) state, as the distance is
reduced.[12] Moreover, the magnitude of the
near field enhancement at the interparticle gap depends directly on
the distance between the nanocrystals, being greater as the gap gets
smaller.[2,11,12] Therefore,
AuNP ensembles act as nanolenses that are able to confine light at
subwavelenght dimensions, giving rise to electromagnetic field enhancements
that are several orders of magnitude larger than those of the incident
field at the interparticle gap space, known as hot spots.[14]Regardless of nanocrystal morphology,
control over the directed assembly of AuNPs is a crucial factor to
build up plasmonic nanomaterials with tailored functionalities.[15] For instance, a number of organizations based
on molecular concepts,[16] such as spherical
clusters,[17] linear and branched chains,[18,19] polymers,[20] and supercrystals,[21] can be produced depending on the AuNPs surface
functionalization. These attempts to assemble AuNPs in a controlled
fashion generally include the use of supramolecular complexes,[20] surfactants,[21] (bio)macromolecules,[22] amino acids,[23] and
thiol-functionalized ligands.[13,24] In the case of AuNRs,
the inherent structural anisotropy gives rise to a different reactivity
of the tips as compared to the lateral facets of the nanocrystals,
thus allowing their preferential functionalization.[25] Notably, most assemblies are controlled by means of chemical
parameters (stabilizer, linker and nanoparticle concentrations, solvent,
evaporation rates and temperature),[16−24] and/or photochemical conditions (light activation of the reactants).[26,27] In this respect, we questioned ourselves whether electromagnetic
field enhancements at hot spots, which are directly associated with
a high temperature increase,[28] could be
used to control the assembly of AuNPs via a photothermal process.The strong interaction of light, particularly of laser radiation
with AuNPs, can result in irreversible morphological changes of the
nanocrystals.[1] Ultrafast femtosecond (fs)
irradiation can induce melting of AuNPs, whereas nanosecond (ns) laser
light produces both photothermal melting as well as fragmentation.[29,31] All these effects have been ascribed to the relaxation dynamics
of the localized surface plasmon electron oscillations in resonance
with the laser wavelength. In the case of AuNP irradiation with a
fs laser, a thermal equilibration process takes place after irradiation,
whereas for the ns laser the electrons continue absorbing photons
when the nanocrystal lattice is still “hot”, ultimately
resulting in AuNPs fragmentation.[30,32] In all these
investigations, reshaping and fragmentation of single AuNPs of various
sizes and geometries have been described using relatively high fluence
laser irradiation.[29−32] Under such conditions, Baumberg and co-workers employed a fs laser
to efficiently weld spherical AuNP assemblies through the generation
of hot spots at interparticle gaps.[33] Inspired
by these examples, we decided to exploit not only the AuNP melting
capability of fs lasers but also the possibility of using lower fluence
fs laser irradiation to control the assembly of AuNPs. This kind of
control over the nanostructures’ morphology by means of laser
pulses may be exploited in the scalable manufacturing of electronic
and optoelectronic devices.[16]For
this purpose, we reasoned that irradiation with 800 nm Ti:sapphire
low fluence 50 fs laser pulses during tip-to-tip assembly by monomer
addition of AuNRs with initial longitudinal LSPR bands well below
800 nm for the monomers can selectively inhibit the formation of longer
oligomers (larger clusters) with longitudinal LSPRs in resonance with
the laser wavelength. The inhibition process is proposed to occur
by photothermal decomposition of interparticle molecular linkers.
This hypothesis relies on the relatively strong hot spots generated
at the interparticle gaps, where a dithiolated molecule (1,8-octanedithiol)
is preferentially located as the molecular linker during the assembly.[13,25] We selected the AuNR aspect ratio to ensure that the respective
maxima of the LSPRs for the monomer and dimer species are located
at 600 and ca. 700 nm, respectively, that is, far away from the wavelength
of the fs pulses, which are in resonance with the trimer structure
(ca. 800 nm). Hence, field enhancement is negligible for the shorter
species but not for the trimer, selectively disrupting formation of
the latter due to temperature increase at the interparticle gaps.
This leads to a significant increase of the population of dimers beyond
what is attainable through the nonirradiated reaction. Finally, we
explored the possibility of welding intermediate species, such as
dimers and trimers, during the AuNRs assembly by increasing the fluence
of the fs laser pulses. Those welded nanostructures form high-aspect
ratio nanorods with a great potential for applications in electronics
and plasmonics, which are hard to produce by chemical or physical
means.[16] Moreover, welding can be attained
at relatively low fluences due to the extreme field enhancement in
the interparticle gaps.Finite-difference time-domain (FDTD)
calculations using the free software MEEP (see the Supporting Information for details, Figure S1)[34] were performed for guidance in predicting the
ideal conditions for the experiments (Figure ). An interparticle separation of 1.5 nm
was considered for the calculation, which is in agreement with the
length of the linking molecule assuming a fully outstretched conformation
of the alkyl chain. We found that for AuNRs with an aspect ratio of
1.8 (see Supporting Information, Figure
S3), the longitudinal LSPR of the trimer was in resonance with the
laser wavelength whereas shorter species were not affected. Because
the dithiols may allow some degree of flexibility at the linking point,
we also analyzed the dependence of the LSPR with the angle between
the AuNRs in the dimer. Our results show that for a wide angular range
(∼60–180°) the LSPR position is barely affected,[11,35] whereas the plasmon intensity varies significantly with the angle
(Figure a) being maximum
at 180°. We additionally studied the field enhancement (|E|/|E0|) in the surroundings of the
AuNRs (Figure –e)
for the particular case of parallel polarization. Maximum field enhancements
of 15, 180, and 530 were obtained for single AuNRs, dimers, and trimers,
respectively (Figure b).
Figure 1
(a) Optical extinction spectra of single AuNRs, rod dimers at various
angles (0, 90, 120, and 180°) and linear trimers, calculated
using FDTD. (b) Longitudinal field enhancement profiles for single
AuNRs (black line), AuNR dimers (red line) and AuNR trimers (green
line). Local field enhancement contour plots in the middle plane for
(c) single AuNRs, (d) AuNR dimers, and (e) AuNR trimers.
(a) Optical extinction spectra of single AuNRs, rod dimers at various
angles (0, 90, 120, and 180°) and linear trimers, calculated
using FDTD. (b) Longitudinal field enhancement profiles for single
AuNRs (black line), AuNR dimers (red line) and AuNR trimers (green
line). Local field enhancement contour plots in the middle plane for
(c) single AuNRs, (d) AuNR dimers, and (e) AuNR trimers.It can be readily seen that in particular for the
trimer the field enhancement is so large that even for illumination
with moderate intensity the energy concentrated in the gap may be
sufficient to decompose the linking molecules and even to melt the
AuNR tips at higher laser intensities. For instance, AuNRs illuminated
with fs laser pulses have been reported to reach gold melting temperatures
(1337 K for the bulk material) for irradiation fluences above 500
μJ/cm2 per pulse.[28] Illumination
with a fluence of 100 μJ/cm2 per pulse would be translated
into temperature increments of ca. 550 K for single AuNRs, and very
likely far larger values around the interparticle gaps for dimers
and trimers,[33] due to the enhanced electric
field in these regions. It is therefore clear that the largest fluence
(650 μJ/cm2) used in our work should be high enough
to melt AuNRs, thus favoring the formation of welded structures during
tip-to-tip assembly. On the other hand, we expect that intermediate
fluences (130 μJ/cm2) may lead to a temperature increase
high enough to decompose the organic linker but not to melt AuNRs
species, and that the temperature produced at lower fluence (13 μJ/cm2) is so small that the AuNR assembly is barely affected by
such fs laser pulses.The use of dithiols as molecular linkers
required, as a first step, the transfer of AuNRs into ethanol, which
is a more suitable solvent for the organic linkers. Solvent transfer
was achieved by addition of a high molecular weight branched polyethylenimine
polymer (see Supporting Information, Figure
S4), which provided sufficient stability to AuNRs in ethanol via steric
hindrance effects.[36] Optimized linker concentrations
were then used to control both initiation and assembly rates (see Supporting Information, Figure S5). Termination
of the reaction was achieved by blocking the free thiols of the linker
by means of a “click” thiol-maleimide reaction.[37] The redshift of the longitudinal LSPR band resulting
from tip-to-tip AuNR assembly was monitored to follow the reaction
process (Figure a),
and the resulting products were analyzed by transmission electron
microscopy (TEM) (Figure b,c). Initiation occurs along with a decrease of the LSPR
band intensity at 600 nm and formation of a new intense band centered
at ca. 700 nm, likely due to the longitudinal LSPR of dimers. After
8 min, the LSPR bands corresponding to AuNR monomers and dimers showed
similar intensities, while after 10 min the reaction was stopped when
the LSPR of trimers at ca. 800 nm reached an analogous intensity.
Figure 2
(a) Extinction
spectra during AuNRs (∼109 M) tip-to-tip assembly
in ethanol, using 1,8-octanedithiol (0.5 mM) as linker, acquired at
time intervals of 90 s. Typical TEM micrographs of AuNRs (b) before
and (c) after assembly (10 min of reaction). (d–f) Extinction
spectra after the assembly of AuNRs exposed for 10 min to 800 nm 50
fs laser pulses: (d) equal frequency (1 kHz) at different pulse fluences,
13, 130, and 650 μJ/cm2; (e) equal fluence (650 μJ/cm2) at different frequencies, 10 Hz, 200 Hz, and 1 kHz; and
(f) the same average power fluence (130 mW/cm2) attained
with different pulse fluences and frequencies, 130 μJ/cm2 at 1 kHz and 650 μJ/cm2 at 200 Hz.
(a) Extinction
spectra during AuNRs (∼109 M) tip-to-tip assembly
in ethanol, using 1,8-octanedithiol (0.5 mM) as linker, acquired at
time intervals of 90 s. Typical TEM micrographs of AuNRs (b) before
and (c) after assembly (10 min of reaction). (d–f) Extinction
spectra after the assembly of AuNRs exposed for 10 min to 800 nm 50
fs laser pulses: (d) equal frequency (1 kHz) at different pulse fluences,
13, 130, and 650 μJ/cm2; (e) equal fluence (650 μJ/cm2) at different frequencies, 10 Hz, 200 Hz, and 1 kHz; and
(f) the same average power fluence (130 mW/cm2) attained
with different pulse fluences and frequencies, 130 μJ/cm2 at 1 kHz and 650 μJ/cm2 at 200 Hz.A systematic analysis of the influence
of 50 fs (fwhm) laser pulses on AuNR assembly was carried out by varying:
(i) pulse fluence (13, 130, and 650 μJ/cm2) at a
pulse frequency of 1 kHz (Figure d); (ii) pulse frequency (10 Hz, 200 Hz, and 1 kHz)
at a pulse fluence of 650 μJ/cm2 (Figure e); and (iii) both laser fluence
and pulse frequency (130 μJ/cm2 at 1 kHz, and 650
μJ/cm2 at 200 Hz) at an equal average power fluence
of 130 mW/cm2 (Figure f). When the pulse frequency was maintained constant
at 1 kHz, no significant differences were observed on the assembly
process at 13 μJ/cm2, resulting in a final extinction
spectrum similar to that of the reaction in the absence of laser irradiation.
By contrast, at 650 μJ/cm2 the final extinction spectrum
showed a new broad band ranging from 900 to 1400 nm, which may be
attributed to the welding of AuNRs, which then effectively behave
as longer rods.[33] Moreover, the peak at
800 nm is considerably damped after irradiation with an intermediate
pulse fluence of 130 μJ/cm2, indicating the coexistence
of AuNR monomers and dimers as the main reaction products. Therefore,
at this irradiation regime the concentration of trimers is significantly
reduced with respect to the nonirradiated case.Interestingly,
irradiation with 650 μJ/cm2 while varying the pulse
frequency from 10 Hz to 1 kHz led to similar results (Figure e) in which welded plasmonic
AuNRs register LSPR maxima in the NIR between 900 and 1400 nm. This
indicates that the pulse frequency is a rather irrelevant parameter,
probably because relaxation dynamics of LSPR oscillations and pulse
irradiation occur at very different time scales. In other words, when
AuNR trimers are exposed to such ultrafast laser pulses, free electrons
absorb the energy of incident photons, thereby increasing their kinetic
energy. Highly energetic electrons with an initial energy distribution
out of equilibrium are then relaxed through electron–electron
scattering on the order of 10–50 fs.[38,39] Within these time scales, no energy exchange occurs between electrons
and phonons. The AuNRs lattice temperature starts increasing as a
result of electron–phonon scattering, reaching thermal equilibrium
between the electrons and the lattice within a time scale of tens
of ps, depending on the initial rise in electron temperature. As the
AuNR temperature increases, energy exchange between the particle and
its surrounding medium occurs through phonon–phonon coupling.
Finally, thermal equilibrium between the AuNR and the aqueous solution
is achieved within 100 ps to 1 ns, depending on particle size and
laser pulse intensity.[40,41] Interestingly, an experiment
in which both the pulse fluence and frequency were modified while
keeping constant the average power fluence (130 mW/cm2)
showed significant differences in the final UV–vis-NIR spectrum,
thus revealing that the average power fluence is not a significant
parameter regarding reaction control. Hence, this series of experiments
showed that the most important parameter toward controlling the reaction
kinetics of AuNRs is pulse fluence.We investigated the resulting
assembled products by using transmission electron microscopy (TEM)
on AuNR oligomers, both assembled (Figure a) and welded (Figure b), through their tips. Figure c shows the distribution of
AuNR oligomers upon irradiation for 10 min at 130 and 650 μJ/cm2 (1 kHz), as compared to the nonirradiated process (see Supporting Information, Figure S6). An analysis
of the number of nanocrystals (either linked or welded) per ensemble
showed that although the proportion of long chains (>6 AuNRs) is
∼50% at 650 μJ/cm2, this value decreased to
∼10% at 130 μJ/cm2, which is significantly
lower than the ∼25% observed in absence of irradiation. Interestingly,
the yield of dimers at 130 μJ/cm2 doubled those obtained
both without irradiation and in the presence of high pulse energies,
meaning that the intermediate dimer assembly can be trapped during
the assembly via fs laser irradiation at the selected pulse fluence.
The proportion of monomers is similar (∼50%) in the absence
of irradiation and at 130 μJ/cm2, which indicates
that both monomers and dimers are formed under the photothermal decomposition
of the linker molecule. Although the fraction of dimers is limited
to 25% under irradiation at 130 μJ/cm2, it should
be noted that this value corresponds to the number of assembled dimers,
which means that the number of AuNRs with a tip-to-tip dimer configuration
is the double of such fraction. We attribute this limit to the natural
polydispersity of gold nanorods obtained from the seed-mediated growth
method, which increases upon gold nanorod tip-to-tip assembly: (i)
dimers with large aspect ratios (with LSPRs close to 800 nm) may be
disrupted by the laser pulses increasing the population of monomers,
which is in contrast to the trend of the reaction; and (ii) trimers
with large aspect ratios (with LSPRs above 800 nm) can elude the effect
of the laser pulses, forming larger oligomers.
Figure 3
Representative TEM micrographs
of AuNRs (∼109 M) assemblies after 10 min in ethanol,
using 1,8-octanedithiol (0.5 mM) as molecular linker: (a) 130 μJ/cm2 per pulse at 1 kHz and (b) 650 μJ/cm2 per
pulse at 1 kHz. (c) Statistical distribution of AuNRs in assembled
species after 10 min of reaction without laser irradiation (black),
130 μJ/cm2 per pulse at 1 kHz irradiation (red) and
650 μJ/cm2 per pulse at 1 kHz irradiation (blue).
(d) Angular distribution of the AuNRs dimers obtained after irradiation
with 130 mJ/cm2 laser pulses irradiation. TEM micrographs
of a dimer (e) and a tetramer (g) obtained by irradiation at 130 μJ/cm2 per pulse. (f,h) Magnifications of the interparticle region
within (e,g), respectively. (i) HAADF-STEM image of a trimer with
the CBED patterns shown as insets. (j,k) High-resolution HAADF-STEM
images of the connection points between the AuNRs in the trimer.
Representative TEM micrographs
of AuNRs (∼109 M) assemblies after 10 min in ethanol,
using 1,8-octanedithiol (0.5 mM) as molecular linker: (a) 130 μJ/cm2 per pulse at 1 kHz and (b) 650 μJ/cm2 per
pulse at 1 kHz. (c) Statistical distribution of AuNRs in assembled
species after 10 min of reaction without laser irradiation (black),
130 μJ/cm2 per pulse at 1 kHz irradiation (red) and
650 μJ/cm2 per pulse at 1 kHz irradiation (blue).
(d) Angular distribution of the AuNRs dimers obtained after irradiation
with 130 mJ/cm2 laser pulses irradiation. TEM micrographs
of a dimer (e) and a tetramer (g) obtained by irradiation at 130 μJ/cm2 per pulse. (f,h) Magnifications of the interparticle region
within (e,g), respectively. (i) HAADF-STEM image of a trimer with
the CBED patterns shown as insets. (j,k) High-resolution HAADF-STEM
images of the connection points between the AuNRs in the trimer.High-resolution high-angle annular
dark field scanning transmission electron microscopy (HAADF-STEM)
and TEM analysis confirmed that at pulse fluences of 130 μJ/cm2, AuNRs were assembled tip-to-tip, mainly into dimers (Figure a,e) with and angular
distribution centered at 60–120°, and interparticle gaps
of 1.0 ± 0.5 nm (Figure f), which is in good agreement with the length of 1,8-octanedithiol
in a fully outstretched conformation of the alkyl chain. Additionally,
other types of larger ensembles are also obtained in low yield (Figure a,g) without significant
differences of the interparticle distances (Figure h). In Figure i, an HAADF-STEM image of a trimer is presented with
the convergent beam electron diffraction (CBED) patterns from the
AuNRs shown as insets. From the latter, it can be concluded that the
three nanocrystals are not oriented along the same crystallographic
orientation, an observation that suggests the absence of structural
connection between the AuNRs. In Figure j,k, high resolution HAADF-STEM images of
the connection points between the AuNRs are presented and the presence
of a layer of ligands can be observed (low contrast regions between
the rods). By contrast, irradiation under pulse fluences of 650 μJ/cm2 led to the formation of welded oligomers (Figure b). High-resolution HAADF-STEM
images of a dimer and a trimer are shown in Figure a,b, respectively, for the case of welded
AuNRs. Figure c,d
show high-resolution HAADF-STEM images of the connection between two
AuNRs in a dimer and a trimer, respectively. A grain boundary is always
present at the connection point of two rods. From the CBED patterns,
shown as insets in Figure b for the case of the trimer, it is clear that the three nanocrystals
are in almost the same crystallographic orientation, but yield a slight
misorientation with respect to each other. It should be noted, however,
that in other cases no common crystallographic orientations were found
to be apparent for welded AuNRs. This rather surprising effect may
be due to the extreme concentration of energy on this region and/or
to the fact that the energy deposition occurs in such short times
that the welding is essentially a nonthermal process.[33] Of course, we also should keep in mind that the HAADF-STEM
images are 2D projections of 3D objects and therefore electron tomography
experiments might be able to yield a further understanding concerning
the connection mechanism. As such experiments should be carried out
with atomic resolution, they are outside the scope of this work.
Figure 4
HAADF-STEM
images of a dimer (a) and a trimer (b) of AuNRs obtained at 650 μJ/cm2. The insets in figure (b) show the respective CBED patterns
of the AuNRs in the trimer conformation. (c,d) High-resolution HAADF-STEM
images of the connection points between the rods within (a,b), respectively.
HAADF-STEM
images of a dimer (a) and a trimer (b) of AuNRs obtained at 650 μJ/cm2. The insets in figure (b) show the respective CBED patterns
of the AuNRs in the trimer conformation. (c,d) High-resolution HAADF-STEM
images of the connection points between the rods within (a,b), respectively.To gain insight into the way fs
laser affects AuNRs tip-to-tip assembly, we repeated a number of experiments
acquiring in situ extinction spectra (see Supporting Information, Figure S2) at time intervals of 2 s (Figure –c). The
relative concentration of the main species was determined by fitting
the experimental spectra with those calculated using FDTD (Figure a and Supporting Information). Unfortunately, the possible
configurations for the AuNRs oligomers are rather numerous, far more
than can be simulated, which introduces an error in the concentrations
predicted by the model. We assume that the AuNRs are preferentially
bonded by the tips, according to TEM analysis,[13,25] which considerably reduces the possible number of configurations.
As discussed above, the angle between AuNRs does not change the position
of the LSPR but only its intensity.[35] Because
the only configuration without tip-to-tip links that was found in
meaningful amounts during TEM analysis was the side-by-side dimer
(see Supporting Information, Figure S7),
we also included the corresponding extinction spectrum in the fit.
We additionally considered the spectra for the main species (single
AuNRs, dimers, and trimers), whereas all larger-chain oligomers were
approximated by a Gaussian curve. Despite its simplicity, the model
is able to yield reasonable results for the relative concentrations
of the main species (see Supporting Information, Figure S8).
Figure 5
Evolution of AuNRs tip-to-tip assembly for different irradiation
conditions. (a–c) Extinction spectra at 20 s intervals for
10 min (a) without laser irradiation, (b) 130 μJ/cm2 per pulse at 1 kHz, and (c) 650 μJ/cm2 per pulse
at 1 kHz. Arrows point to the spectral region at the LSPR maxima for
the monomer (orange), dimer (red), and trimer (black). Concentration
of side-by-side (SBS, green lines) dimers yielded by the fit is very
close to zero. (d–f) Concentration of single AuNRs, dimers,
and trimers obtained from the fits for (d) no laser and pulse fluences
of (e) 130 μJ/cm2 and (f) 650 μJ/cm2.
Evolution of AuNRs tip-to-tip assembly for different irradiation
conditions. (a–c) Extinction spectra at 20 s intervals for
10 min (a) without laser irradiation, (b) 130 μJ/cm2 per pulse at 1 kHz, and (c) 650 μJ/cm2 per pulse
at 1 kHz. Arrows point to the spectral region at the LSPR maxima for
the monomer (orange), dimer (red), and trimer (black). Concentration
of side-by-side (SBS, green lines) dimers yielded by the fit is very
close to zero. (d–f) Concentration of single AuNRs, dimers,
and trimers obtained from the fits for (d) no laser and pulse fluences
of (e) 130 μJ/cm2 and (f) 650 μJ/cm2.Concentration of the main species
as a function of time are shown in Figure –f for the nonirradiated reaction,
as well as those for medium and large fluences. Data for the low fluence
was omitted because they were virtually identical to those of the
nonirradiated reaction. The rate of AuNR assembly is slightly decreased
by laser irradiation with respect to the nonirradiated reaction, which
is apparent for a fluence of 130 μJ/cm2. This suggests
that fs irradiation not only affects trimers but also a small fraction
of dimers due to the natural polydispersity of AuNRs in aspect ratio,
resulting in an increase of monomer concentration at early reaction
stages. Our model predicts an inversion of the concentrations of dimers
and trimers due to fs irradiation (Figure d,e), which is in good agreement with the
proportion of oligomers determined by TEM (Figure c). We additionally found a significant increase
of the rate at which monomers vanish at the early stages of reaction
at the fluence of 650 μJ/cm2, suggesting that an
alternative reaction pathway based on the activation of chemical reactants
by light[26,27] takes place for larger fluences. It should
be noted that by the end of the reaction there is still a significant
remaining concentration of monomers (∼1010 particles/cm3) because the reaction rate stagnates from this point onward.
This is probably due to the polydispersity of AuNRs from which dimers
with large aspect ratios are affected by the laser pulses, increasing
the population of nonreactive monomers.In summary, we have
shown that femtosecond laser irradiation is a powerful tool to control
the assembly of AuNRs. Laser pulse fluence was found to be the most
important parameter affecting tip-to-tip assembly kinetics. This effect
was demonstrated using AuNRs that were tailored in such a way that
the longitudinal LSPR of the rod trimers is in resonance with the
wavelength of a Ti:sapphire fs laser (800 nm). We were thus able to
use fs pulses with a fluence of ∼100 μJ/cm2 to considerably reduce the formation of trimers and longer oligomers,
thereby increasing the relative population of AuNR dimers. We also
found that if the fluence is increased beyond 500 μJ/cm2, the increase of temperature at the interparticle gaps is
large enough to melt AuNR tips, thus producing new species: welded
AuNRs with LSPR bands at the NIR. Preliminary results in this direction
indicate that it can be used to produce nanostructures with tailored
LSPR modes in the IR region. Although we have only illustrated the
use of laser irradiation for controlling the assembly kinetics of
gold nanorods in high yields and short reaction times, this technique
can be generalized to a variety of shapes, limited only by the wavelength
of available fs lasers. Therefore, we show that under irradiation
of femtosecond laser pulses with wavelengths in resonance with the
LSPR of larger tip-to-tip AuNR oligomers, plasmonic polymers with
relatively high monodispersities may be obtained.
Authors: Jer-Shing Huang; Johannes Kern; Peter Geisler; Pia Weinmann; Martin Kamp; Alfred Forchel; Paolo Biagioni; Bert Hecht Journal: Nano Lett Date: 2010-06-09 Impact factor: 11.189
Authors: Yuan Zhao; Liguang Xu; Luis M Liz-Marzán; Hua Kuang; Wei Ma; Ana Asenjo-Garcı́a; F Javier García de Abajo; Nicholas A Kotov; Libing Wang; Chuanlai Xu Journal: J Phys Chem Lett Date: 2013-02-07 Impact factor: 6.475
Authors: Andrés Guerrero-Martínez; Baptiste Auguié; José Lorenzo Alonso-Gómez; Zoran Džolić; Sergio Gómez-Graña; Mladen Žinić; M Magdalena Cid; Luis M Liz-Marzán Journal: Angew Chem Int Ed Engl Date: 2011-04-19 Impact factor: 15.336
Authors: Shaunak Mukherjee; Florian Libisch; Nicolas Large; Oara Neumann; Lisa V Brown; Jin Cheng; J Britt Lassiter; Emily A Carter; Peter Nordlander; Naomi J Halas Journal: Nano Lett Date: 2012-12-05 Impact factor: 11.189
Authors: Ovidio Peña-Rodríguez; Pablo Díaz-Núñez; Vladimir Rodríguez-Iglesias; Luis Montaño-Priede; Antonio Rivera; Umapada Pal Journal: Nanoscale Res Lett Date: 2017-01-05 Impact factor: 4.703