Introduction
I worked with Philippe Guyot-Sionnest on an experimental project involving propagating surface plasmons which are are formed by the coupling of an incident light wave with the quantized oscillations of the free electrons. Dr. Guyot-Sionnest's chemistry group specializes in the synthesis and analysis of colloidal quantum dots and other nanoparticles and the goal of my project was to use surface plasmons to study various responses in these nanoparticles. Specifically we looked into making measurements of the transverse magneto-optical Kerr effect and examining how properties such as the intensity and direction of quantum dot emission and quantum dot photostability are affected by surface plasmons.
My work included constructing an optical table for the generation of surface plasmons, using existing apparatus (e.g. an evaporator to fabricate thin film samples), and computer programming for comparison between theory and results.
The project was primarily experimental in focus (involving data taking and analysis), but it also drew a bit on theoretical means for the comparison of results to expectations and in the interpretation of results.
Surface Plasmon Propagation
Plasmons are quantized oscillations of free electrons on metallic or dielectric surfaces. Under certain conditions, light incident on such a surface can couple to this plasmon mode to create propagating surface plasmons, also known as polaritons. Free plane waves incident on a metallic surface cannot generate surface plasmons however, because they do not satisfy the proper dispersion relation. Methods exist to create a condition where the relation is satisfied, and for this project, we used the attenuated total reflection method where light passes through a glass prism and is bent before encountering the surface. We used a white light source and studied various angles of incidence.
As examples of some of the results we obtained, below I give two reflectivity plots. Plasmons are induced on the film's surface when the incident light's wavenumber matches the plasmon resonance wavenumber. For a particular incident angle, this resonance is achieved at a single frequency/wavelength. (One may also do the experiment in reverse. If a single frequency source (e.g., a laser) is used, the resonance will only occur at a specific angle.) In the plots below, the frequencies away from the plasmon resonance show normal reflection, however near the resonance, reflectivity drops significantly as the incident light couples to the surface as a plasmon and is not reflected.
Figure 1: This is a sample of silver (50nm) at two different incident angles. Note that the experimental dips at resonance do not go as low as the theory predicts. The reason is that the incident beam is not infinitely sharp, but has some small angular width (~0.16 degrees based on the geometry of our setup.) When the theory takes into account this width, the agreement with experiment improves.
Figure 2: This is a sample of gold (50nm) over chromium (1.2nm) at an incident angle of 43.45 degrees. Note here the significant change in the resonance wavelength with the addition of a monolayer (~5.7nm) of quantum dots.
Magneto-Optical Kerr Effect
We had hoped to use surface plasmon propagation as a method to enhance what is known as the magneto-optical Kerr effect. When light hits a magnetic material, the reflected wave can undergo several changes, for example a rotation of polarization angle or a change in the magnitude of the reflectivity. Such an effect is strongly dependant upon the properties of the material and it can be used to measure the dielectric constant of a material.
Our initial idea that the enhancement of local fields due to propagating plasmons would also enhance this effect ultimately were incorrect. Our own theoretical studies showed that changes in reflectivity, for example, would be on the order of 500-1000 times smaller than the absolute reflectivity, a result consistent with conventional Kerr effect techniques. Published experimental results studying this approach also reported similar results.
We did design an experimental apparatus to test the effect, but time constraints coupled with our low expectations for the experiment caused us to scrap the idea.
Quantum Dot Fluorescence
Quantum dots are small nanoparticles which exhibit discrete energy level separation as in individual atoms. Perhaps the most well known feature of these dots is their fluorescence. When illuminated by light, electrons can be excited to higher energy levels in the dot and, upon de-excitation, will give off photons of a particular frequency. Two features of quantum dot fluorescence which we were interested in looking at were the direction of dot emission and their photostability.
In solution, quantum dots emit isotropically. If placed on a metallic surface where emission into a surface plasmon mode is possible, how would the spatial distribution of the emission change? Using similar theoretical techniques to those of reflection employed above, we looked at the predicted emission of light from dots on metal films. We found that when the dot emission was directed toward/through the metal film it was concentrated in a cone whose opening angle was that same given above in the reflectivity studies. We saw this (to a limited extent) in lab experiments.
Figure 3: This is the theoretical emission distribution for quantum dots on a layer of silver (57.5nm) on both the excitation (front) side and the glass/metal (back) side. The front is rather featureless as expected, but the back shows a clear peak. When rotated azimuthally, this peak forms a hollow cone. Red is for dots with polarizations parallel to the surface while blue is for polarizations perpendicular.
To study photostability, we built a fluorescence microscope to observe fluorescing dots. It had been observed previously that quantum dots will "blink" on and off when illuminated. We wished to study this phenomenon and to see if the blinking would change in the presence of plasmon modes. We were able to observe individual dots on glass slides and saw blinking. On silver films, we again saw blinking, but we also saw a dramatic (~10x) increase in overall brightness and a greater distinguishability among the different colors of the dots.
For those interested in more information:
|
