Stephan Link, PhD, Rice University
The proposal's goal is to understand the relationship between the plasmon resonance of individual metal nanoparticles and their environment. In particular, the aim of this project is to characterize the linewidth of the surface plasmon resonance in addition to the maximum, which is generally used while the linewidth is mostly ignored. Progress made on this project in year 2 has involved two main approaches. We have further characterized one-photon luminescence in gold nanorods and determined the luminescence quantum yield as a function of nanorod aspect ratio for excitation of the transverse plasmon resonance. We found that the competition between internal relaxation of hot electrons and creation of a longitudinal plasmon that then emits is the likely reason for the observed decrease in luminescence quantum yield with increasing aspect ratio. We have also characterized super- and sub-radiant plasmon modes in linear chains of gold nanoparticles. Strong plasmon coupling due to close interparticle distances leads to new plasmon modes that are absent in the individual nanoparticles. Of particular interest are sub-radiant modes that have narrow linewidths and are therefore great candidates for sensing applications. These projects present important steps toward a better understanding about the influence of the environment on surface plasmon resonances.
Project 1: One-photon luminescence of single gold nanorods
We have investigated the one-photon photoluminescence of gold nanorods with different aspect ratios. We measured photoluminescence and scattering spectra from 82 gold nanorods using single-particle spectroscopy. We found that the emission and scattering spectra closely resemble each other independent of the nanorod aspect ratio. We assigned the photoluminescence to the radiative decay of the longitudinal surface plasmon generated after fast interconversion from excited electron-hole pairs that were initially created by 532 nm excitation. The emission intensity was converted to the quantum yield and was found to approximately exponentially decrease as the energy difference between the excitation and emission wavelength increased, for gold nanorods with plasmon resonances between 600 and 800 nm. This decrease is interpreted as a change in coupling efficiency between excited electronic states and the surface plasmon, i.e. the collective electron motion of the longitudinal surface plasmon resonance. Furthermore, we found a small but reproducible spectral blue-shift of the photoluminescence compared to the dark-field scattering independent of the gold nanorod aspect ratio. The magnitude of the blue-shift was determined to be 11 ± 3 nm as the average of all measured gold nanorods. An excitation power dependence showed that the maximum of the photoluminescence spectrum further decreased with increasing laser power. A laser induced melting into smaller aspect ratio nanorods can be excluded because the dark-field scattering spectra recorded before and after measuring the photoluminescence were the same. A thermal effect due to a change in the refractive index because of continued laser excitation is also unlikely because the temperature increase was estimated to be smaller than 10 K under our experimental conditions. We hypothesize therefore that the blue-shift of the photoluminescence spectrum was caused by charging of the gold nanorods through reactions of the environment with the hot electrons and holes as effective higher electron densities cause blue-shifts of the longitudinal surface plasmon resonance. To address this hypothesis, we collected dark-field scattering spectra under simultaneous 532 nm laser illumination as a function of laser power. While the dark-field scattering spectra were also blue-shifted with laser excitation, the blue-shift of the corresponding photoluminescence peak was always more pronounced. Although these results are consistent with the suggested dynamic, photo-induced charging process, further investigations are needed to completely resolve the underlying mechanism. If indeed charge transfer reactions of the hot electrons with the environment are involved, the photoluminescence could be used in the future as a sensitive tool to investigate plasmon mediated surface reactions.
Project 2: Sub- and super-radiant plasmon modes resulting from strong near-field coupling
The interaction between adjacent metal nanoparticles within an assembly induces interesting collective plasmonic properties. We have investigated linear chains of 50 nm gold nanoparticles. These chains were fabricated using a combination of top-down and bottom-up approaches resulting in the smallest possible interparticle separations and therefore strongest plasmon coupling. Gold nanoparticles were assembled into trenches of variable length and width created in a polymer resist by an electron beam. Lift-off of the unexposed polymer resist removed nanoparticles that were non-selectively deposited onto the polymer surface instead of into the trenches, yielding free-standing nanoparticle chains. The electron-beam generated templates successfully defined the dimensions of the overall assembly, but the nanoparticles did not pack uniformly within the trenches. By adding a unique identification pattern to the substrate, we were able to locate individual nanoparticle chains for detailed structural characterization by SEM and correlated optical spectroscopy using polarization-sensitive broadband single-particle extinction spectroscopy. Comparison between the experimental extinction spectra and electrodynamics calculations using generalized Mie theory allowed us to assign the different resonance maxima to the super-radiant as well as sub-radiant plasmon modes. The former is typically the lowest energy collective mode of the assembly and for a linear chain of nanoparticles corresponds to the plasmon oscillation where all dipoles of the constituent nanoparticles are in phase with each other so that the total optical response collectively adds up to give a super-radiant response. The sub-radiant modes couple less to the incident electromagnetic field because of a smaller overall dipole moment as dipole-like domains along the chains alternate in opposite directions and effectively cancel each other. We demonstrated that this smaller coupling to far-field radiation can be exploited to achieve long propagation distances in gold nanoparticle chain waveguides. In addition to smaller extinction efficiencies these sub-radiant modes also have narrower linewidths which make them excellent candidates for environmental sensing applications. By tuning the geometry of the nanoparticle cluster geometry one can effectively shape the optical response and hence the spectral position of sub-radiant plasmon modes. Although the assembly method we used here yields nanoparticle arrangements containing defects with respect to nanoparticle size and positioning, such disorder has a relatively small effect on the collective plasmonic response and is off-set by the stronger plasmon coupling for close interparticle spacing.