FY2010 - Oak Ridge National Laboratory

Director’s R&D Fund—
General
Results and Accomplishments
In this project, our goal was to achieve active control over the plasmonic response of metallic
nanostructures in the visible range using ferroelectric materials as a tunable substrate. The tuning was to
be achieved by creating extreme field gradients across the substrate, resulting in large changes in the
dielectric response at the metal/dielectric interface. For this goal, we prepared diffusion-aggregated gold
nanoparticle clusters on LuFe 2 O 4 ferroelectric substrates. An external electric field was applied to control
the plasmonic response at the metal/ferroelectric. It was found that the reflectance and surface-enhanced
Raman response of the composite could be altered by the application of an electric field, demonstrating
direct and active control of the plasmonic response, as proposed in the project.
LuFe 2 O 4 was selected as the ferroelectric substrate because it exhibits high dielectric tunability at fairly
low electric fields over a broad temperature range close to room temperature. Gold nanoparticle clusters
with individual nanoparticle diameters of 20 nm were prepared by a solvent method. Reflectance
measurements over a wavelength range from 200 to 3000 nm were obtained with a Cary 5000
spectrophotometer. Surface-enhanced Raman scattering (SERS) spectra at an excitation wavelength of
785 nm were obtained with a laser microscopic confocal Raman spectrometer, equipped with a
thermoelectrically cooled charge-coupled device detector with a spectral resolution of 1~2 cm -1 .
For applied electric fields as low as 50 V/cm, a clear change in the SERS intensity as well as reflectance
was observed. Below this value, the optical response showed little change with increase in applied field.
This threshold behavior, which originates from the unique charge ordering behavior of LuFe 2 O 4 , provides
the rare advantage of enabling extremely large changes in the dielectric constant for applied fields above
the 45 V/cm threshold. The maximum change in reflectivity observed at a wavelength of 1500 was ~10%
at an applied field of 75 V/cm, despite averaging effects due to the large beam size. The maximum change
in the SERS response at the same applied field was as large as 65%. These changes were repeatable over
multiple cycles, demonstrating the controllability and robustness of the response.
In order to simulate the experiment, numerical calculations based on the Finite Different Time domain
(FDTD) method were performed. The cross section of the Raman intensity estimated at a fixed
wavelength of 785 nm showed increases with the applied voltage bias. Good agreement was found
between experiment and theory for the Raman intensity increase with the application of a bias voltage
(within a factor of ~2). The simulations results also replicate the threshold behavior observed in
experiment. The qualitative and quantitative agreement between the experimental results and the
numerical estimates from FDTD suggest that the increase in the SERS signal at higher voltages is
represented accurately by the simulations. This implies that the application of an applied field results in a
local refractive index change in the multiferroic (LFO) substrate, and this index change, then, is the
primary mechanism for enabling the tuning or control of the optical response of the gold clusters, as
verified through the Raman and reflectance data.
We have demonstrated active tuning of the plasmonic response of gold nanostuctures by the use of
modest electrical fields on an LFO substrate for applications as tunable SERS templates. The estimated
refractive index change is of the order of magnitude of related work (~0.03) even at dramatically lower
applied electric fields. The SERS enhancement is significant at ~65%. Numerical calculations suggest
that the local refractive index change of the substrate accounts for the dominant portions of the optical
response of the plasmonic structure. This demonstration of repeatable and reversible changes in the
optical response of the nanostructures paves the way for real-time active control and tuning of SERS
templates for label-free chemical sensing. One can then conceive of sensing a wide variety of molecules
with the same template, an important advancement in the field of chemical sensing. The electronic
polarization characteristics of LFO indicate that one may also achieve high switching speeds for device
applications pertaining to the computing and communications industry. Enabling refractive index changes
164