Optics in the City of Light

2016 Projects

Melissa Guidry: Polymer-based microlasers

Whiteboard Video:


I analyzed the lasing spectra emitted from polymer-based microlasers with Prof. Mélanie Lebental at École Normale Supériore de Cachan in the Laboratoire de Photonique Quantique et Moléculaire. Microlasers are highly relevant to the cutting-edge of many fields of research, including telecommunications, chemical and bio-sensing, and quantum chaos studies.
My research concerned the emissions from three-dimensional Fabry-Perot (FP) microlasers. The stripe-shaped cavity of a FP resonator is one of the simplest to study: in particular, when the FP cavity thickness is on the order of a wavelength (i.e., when the cavity is two-dimensional), light only resonates in a simple back-and-forth orbit between the cavity walls. This trajectory is known as a Fabry-Perot periodic orbit (FP PO).
When the cavity thickness is on the order of the cavity length (i.e., when the cavity is three-dimensional), the resonant states become more complex, and their behavior is largely unexplored. Both the cavity dimensions and surrounding media affect which modes are formed. Using UV lithography, we fabricated three-dimensional FP cavities on two different substrates, glass and silicon. Part of my work involved developing a new protocol for fabricating cavities thicker than 60 microns.
We pumped our cavities perpendicular to the substrate, using a frequency doubled Nd:YAG laser. Emissions were collected in the far-field using a lens and then sent to a spectrometer using a fiber. Our first tests were limited to detection in the plane of the cavity. From the Fourier transform of the lasing spectrum, we extracted the optical path length of the light in the cavity. For all varied parameters (i.e., thickness, width, and substrate), we observed path lengths longer than that of the simplest FP PO, suggesting more complex lasing modes. To predict which modes would dominate at different parameters, we calculated lasing thresholds for simple geometric orbits. For both substrates, the threshold model was not predictive of the observed path lengths.
After calculating emissions for various modes at different angles out of the cavity plane, we measured emissions using a 3D goniometer. Every measured cavity emitted light primarily 5 to 10 degrees outside of the plane. For each cavity thickness tested, one spectrum was seen at all emission angles; the spectrum at each intensity peak had the same measured optical length after a Fourier transform. In addition, the predicted emission angles for simple geometric orbits were much larger (>20 degrees) than the angles of the measured intensity maxima. We explored a ray interference model to describe the emissions. According to the model, as the cavity thickness becomes much larger than the wavelength of the light, the cavity emits primarily inside of the plane. Experimentally, the greatest emission intensity is slightly outside of the plane.
In summary, three-dimensional FP lasing modes are not localized on the simple FP PO. At this time, our theory cannot predict a specific PO for each cavity. We intend to explore a semiclassical quantization of the cavity cross-section, which may explain the roughly in-plane emission.
I am grateful to my advisor, Prof. Mélanie Lebental, for invaluable guidance and support. I am also thankful toward TianLun Li for her assistance with measurements and data analysis, as well as Dr. Rasta Ghasemi for her endless patience while instructing me on cavity fabrication.

Top

Jacqueline Remmel: Fabrication and characterization of gold nanostructures

Whiteboard Video:

Using Direct Laser Writing (DLW), researchers in LPQM fabricate many kinds of nanostructures, such as polymeric structures made from photoresist; most of my work this summer focused on gold structures. Previous work in LPQM has involved using a hot plate to thermally anneal a gold layer into nano-islands, and this summer I worked on experiments using DLW to thermally anneal the gold instead. We started by using a magnetron sputtering machine to deposit a thin layer of gold onto a glass substrate. The setup I used to fabricate nanostructures includes a 532nm cw laser which is tightly focused by a high numerical aperture objective lens onto the sample. When the sample is placed into the fabrication setup, the high intensity of the laser beam induces a local thermal annealing effect (aka. dewetting) in the gold, creating a patterned nanostructure whose size and shape can be manipulated by changing various parameters, including the thickness of the gold layer, the laser’s output power, the delay time of the laser at each point, and the periodic spacing of the laser’s movement. So, as the name implies, DLW allows us to use our laser to draw patterns in the gold layer and create visibly colored images or letters, but at nanoscale.

A second optical setup was then used to characterize the optical properties of the gold nanostructures, allowing us to investigate the presence of a plasmonic effect. After directing a light source which produces a wide range of wavelengths (eg. a broadband laser or a halogen lamp) through the gold, a spectrometer allows us to measure the transmission spectrum. After normalizing the spectra transmitted by the nanostructure with that transmitted by the regular gold layer, we can observe the presence of a plasmonic effect in the form of a dip in transmission (corresponding to a peak in absorption). When the plasmon resonance frequency of the gold structure matches the frequency of the external electromagnetic field, the free electrons in the metal are driven to oscillate at this frequency, causing a peak in absorption. The peak can be manipulated not only by changing the fabrication parameters, but also by changing the refractive index of the surrounding environment, which occurs, for example, when covering the gold structure with a protective polymer layer.

Being able to control the optical properties of gold nanostructures is useful for various applications including data storage, sensors, optical filers, and color printing. The results we obtained this summer are particularly useful for data storage because of the sub-micrometer size of the structures, which is useful for enhancing the spatial efficiency of optical disks. In general, the use of a continuous wave laser (rather than a femtosecond pulse laser) in LPQM has allowed ENS Cachan researchers to develop methods of nanostructure fabrication that are comparatively simple and inexpensive, and thereby industrially advantageous.

I’d like to express my sincerest gratitude to all of LPQM, and in particular, PhD student Quang Cong Tong, Master’s student Mai Hoang Luong, and Associate Professor Ngoc Diep Lai, all of whom were fantastic teachers and colleagues to me this summer.

Top

David Bishel: XRD Analysis of Piezoelectrics

Whiteboard Video:

This summer I worked under Professor Laurent Daniel of Centrale-Supélec on his investigation of local strain in piezoelectric ceramics. In his current research, a typical experiment involves recording the evolution of the x-ray diffraction (XRD) profiles produced by different compositions of lead zirconium titanate (PZT, a material with strong piezoelectric properties) while changing the applied electric field or stress. Thus, I was tasked with furthering a MATLAB program that would extract from such a profile the necessary information to calculate the local strain ϵ of the material. The program accepts the experimentally observed value pairs (2θ, intensity) as input, requests the user to select a single- or double-peak to analyze, and then fits the profile according to the pseudo-Voigt function.

Being initially unfamiliar with MATLAB, I spent the first few weeks learning the fundamentals of MATLAB coding while also understanding the workflow of the program that I was given to develop. Once I was functionally literate in the core MATLAB functions used, I began adding the necessary elements to extend the analysis of the local strain. I constructed and included a general formulation of the multiples of random distribution (MRD), which is a measure of the relative intensities of the two peaks in a double-peak pair formed by a given crystal lattice set {hkl}. The MRD is used as a sort of weight by which to average the individual strains ϵ(hkl) of the two peaks, giving the composite strain ϵ{hkl} of the crystal lattice set. Along with these calculations, I propagated the respective errors from the original fit parameters.

Towards the end of my traineeship, Professor Daniel provided me with a new data set that had yet to be analyzed. This data contained experiments run on five PZT samples, each having a different zirconim:titanium ratio. Each trial was subjected to an electric field loading composed of multiple Rayleigh loops; the electric field was essentially “looped” about some reference magnitude, and the size of this “loop” is increased with each cycle. During the time that I was able to work with this data, I wrote a MATLAB function that aids in identifying the profile peaks (center location, lattice spacing, and whether it is a single- or double-peak), included code that loads the experimental conditions directly into the program as variables, and modified the output procedure to account for the new experimental conditions and thus produce meaningful output plots. I have been able to run only preliminary analyses on this data, but there appears to be an accumulation of residual stress that depends on the orientation with respect to the applied electric field. These initial tests still need to be validated, and the full set of data remains to be analyzed.

Because of this program, I was able to use MATLAB programming as a tool for data analysis and witness first-hand what is required at the back end of the research cycle, while also being immersed in the current state of crystallography and piezoelectric technology. I am grateful to the University of Michigan for creating this opportunity, the National Science Foundation for sponsoring it, and Professor Laurent Daniel for allowing and aiding me to contribute to his research.
Top


Christopher Ayala: Supercontinuum Light and Fluorescence Lifetime Measurements

Whiteboard Video:

This summer I had the pleasure of working with Robert Pansu at École Normale Supérieure de Cachan. During my internship in the PPSM laboratory, I was introduced to a side of optics pertaining to chemistry. I was dealt the task of using a supercontinuum laser to excite a sample and cause to fluoresce in a sample. The fluorescence lifetime of a sample can be very helpful to researchers in that we can use this information to gather data about changes in the nano-environment, the size and sometimes shape of the molecules, the molecular interactions and many other aspects. Since the supercontinuum laser has pulses with different wavelengths, we use many optical devices such as a monochromator and dichroic mirrors to select wavelengths and remove undesired ones. To get an accurate lifetime, we use a photodiode and photomultiplier tubes to calculate the fluorescence delay for each photons. But for supercontimuum laser the pulse intensity is random. The trig by the photodiode is tricky, at the limit of the current electronic. We also use an auto correlator which has a sort of Michelson Interferometer inside to help determine the pulse shape. I was able to accomplish creating a system from scratch which can efficiently send the beam at the selected wavelength and was able to send the light into the microscope to the sample. One of the toughest parts was the careful aligning of the lasers as well as learning about all the new equipment within a matter of a few weeks. Through this I can say I was able to gain invaluable optics experience. Be the immense patience and focus required in this field or the lesson that asking for help isn’t going to make you look stupid but benefit you more than you think; this summer has been a great push forward in my career as a physicist and will continue to push me to strive to be a better physicist as I pursue a PhD.
Top

Michael Tripepi: Measuring Nanoradian Faraday Rotations in the Polarization of Light

Whiteboard Video:

I worked under Dr. Alistair Rowe at École Polytechnique with his graduate student Indira Zhaksylykova and fellow REU researcher, Jill Antonishen. Our research group aimed to measure nanoradian Faraday rotations in semiconductors. The purpose of which is to develop magnetometers that can measure the spin dynamics in more complicated semiconductors such as silicon. Faraday rotations are rotations in the polarization angle of linearly polarized light as it propagates through a material placed in a magnetic field.
There are two primary optical setups that are proposed to measure these rotations: a Sagnac interferometer and an optical bridge. The Sagnac is a common-path interferometer in which two beams travel in opposite directions around a loop and interfere with each other, adding constructively or destructively depending on their respective phases. While the setup involves many components, it can be shown to be operationally equivalent to two partially crossed polarizers. The optical bridge works by dividing the beam using a polarized beam splitter and measuring the difference in the signals entering two separate detectors. In this way, the beam is separated into two orthogonal polarizations whose intensities are based on the initial polarization of the entering beam. This effectively transforms a phase measurement of the polarization into an intensity measurement. The optical bridge operates best at what is called the “balanced condition” where the difference in the signals between the two detectors is approximately zero.
We were tasked with measuring the noise characteristics of both setups at low intensities. There are two kinds of noise in the system: source and shot. Source noise is the noise due to defects in the setup, laser, etc… It is linearly proportional to the intensity of the laser beam. Shot noise is intrinsic to the system and caused by the particle nature of photons. Hence it is desirable to be in this range. After measuring this trend at various laser power ranges on the two setups, we demonstrated a superior performance in the optical bridge configuration. It was able to perform shot noise limited measurements at higher intensities than the Sagnac interferometer. The operation at higher intensities allows for a higher signal-to-noise ratio, which is also desirable a measurement apparatus. This is unexpected since phase measurements (such as the Sagnac) are generally considered more sensitive than intensity measurements. However, the optical bridge at the balanced condition allows one to measure with higher laser intensities since taking the difference of the signals cancels various types of noise.
Nevertheless, the Sagnac interferometer is still of interest for its ability to isolate rotations in the polarization of light due to different magneto-optical effects. In fact, a Sagnac can be configured only to measure Faraday rotations or only rotations due to birefringence. We were able to briefly explore this area by testing one particular configuration sensitive to magneto-electric rotations according to theory.
Top

Jill Antonishen: Microradian faraday rotations to observe optical pumping in semiconductors

Whiteboard Video:

I spent the summer working in the PMC lab (Laboratoire de Physique de la Matière Condensée) at École Polytechnique with the electrons-photons-surfaces research group. Fellow REU student Michael Tripepi and I worked with our advisor, Alistair Rowe, and grad student Indira Zhaksylykova to detect microradian faraday rotations for the ultimate goal of observing optical pumping in semiconductors. This breakthrough would allow for advancements in the field of spintronics and help fill in the gaps with information like spin relaxation times in silicon. The typical polarized luminescence methods do not work in indirect band gap materials which must account for phonon emission and absorption. Additionally, silicon contains spin polarizations that are immeasurably small and does not luminesce very well.
The technique we worked with and perfected over the summer, instead, involves the faraday effect – rotation of the linear polarization of light as it passes through a magnetized medium. We worked with two magnetometers, a Sagnac interferometer and an optical bridge, designed to measure these rotations. High and low power tests were performed on each configuration to decipher the experimental noise limitations and help decide which technique can be best used in future experiments. While Indira was setting up these future experiments with Silicon, we worked with a TGG crystal because it has a high Verdet constant which results in the Faraday effect.
Several variable density filters in combination with a neutral density filter were placed in front of the 532nm laser source in an effort to limit the power as we continually measured the shot and source noise dependencies. Shot noise is the intrinsic noise in our system which comes from the particle nature of light. This type of noise varies as the square root of intensity. Source noise is the extrinsic noise in our system due to the laser cavity or defects in the set-up and depends linearly on intensity. Initially, it was unexpected that the Sagnac could operate in the shot noise limit at any intensity. We proved the hypothesis wrong and showed that as the source intensity is reduced to the microwatt range the noise vs. intensity curves closely resemble a square root function. However, the shot noise limit can only be obtained at the expense of the signal to noise ratio.
Although it is common to believe that interferometers provide the most sensitive kind of measurements, the optical bridge looks more promising because it contains a mechanism designed to take an intensity difference measurement. This means that if the half-wave plate in the configuration is aligned at π/4, the beams will split equally through a polarizing beam splitter and the back-to-back photodiode detector will read a signal which is approximately zero. Thus, any extrinsic noise in the system should be cancelled out. Results proved the bridge is superior at performing shot noise limited measurements because it can do so with three orders of magnitude more photons than the Sagnac.
While these results make the Sagnac interferometer seem, for lack of a better word, useless, this is not entirely true. In fact, Sagnac’s hold other special properties which make them extremely unique. The nature of the interferometer – allowing two beams to pass through an object in both directions – introduces the effects of time (T) and parity (P) symmetries. The last few weeks of the summer were spent investigating different Sagnac configurations and working with Jones Calculus (a way to describe polarized light) and Jones Matrices (a way to describe optical elements). This matrix manipulation provides insight into which combinations are sensitive to different materials with specific time and parity symmetries like linear birefringence (TP), chirality (TP’), magneto-optical materials (T’P’), and faraday rotations (T’P). Understanding the reasoning behind why certain combinations work and some don’t is still very unclear to us and our advisor. However, we enjoyed the opportunity to experimentally test these theoretical results whether or not our immediate results were totally satisfying. Top

Michael Dominguez: Modeling an induction heating system

Whiteboard Video:

This summer, I worked at GeePs laboratory at CentraleSupelec. I worked under Prof. Laurent Daniel and our project dealt with modeling an induction heating system. Induction heating is caused when a changing magnetic field induces eddy currents with in a material and therefore induction heating deals with both electricity/magnetism and heat transfer. We uncouple the induction heating problem into Electromagnetic and Heat Transfer parts to study the eddy current and heat transfer behavior independently.
We used Finite Element Method modeling to simulate the heating system. FEM is a numerical technique which takes a complicated geometry and reconfigures it into numerous simple geometries, and then solves the partial differential equations across these simple geometries. We used the software COMSOL and simplified the geometry in order to quickly obtain numerical results. Some of the simplifications include decreasing the size of the geometry, focusing only on heat conduction, and using one material throughout the geometry.
Due to the computational performance required, we calculated results for a 2D and 3D model to verify that the 2D model can be used as an approximation.
We studied and confirmed that the eddy current distribution is frequency dependent. This behavior in the 3D model is identical to the 2D model and on the same order of magnitude for the 3D and 2D models. Due to the similar eddy current behavior, joule heating of the 3D model is also identical to the 2D results. After we have validated that the 2D calculation is a suitable approximation for 3D calculations of this geometry. The next step is to further explore heating behavior using the 2D models with variable currents, frequencies, conductor size, and different 3D inductor configurations.
Top

Rachel Odessey: Introducing Metallic Nanoparticles to 2D and 3D Photonic Structures

Whiteboard Video:
This summer, I’ve worked with Professor Ngoc Diep Lai’s research group at ENS Cachan, particularly on the project of Trang Nguyen, a PhD student. This summer’s project has centered on fabricating 3D polymeric photonic structures and introducing nanoparticles to a center structural defect. Work for the summer has fallen in three categories: attempting structures with Probes nanoparticles, refining our design for structures without using any nanoparticles, and fabricating 2D arrays with silver nanoflowers.
Photonic structures with a microcavity defect localize and enhance the light at the site of the defect; since we know that nanoparticles can be spontaneous emitters of light, we attempt to couple a nanoparticle to a defect in a photonic structure. We predict that this coupling will enhance the optical properties of the defect such as its spectrum and lifetime.
The first project of the summer, attempting 3D structures with Probes nanoparticles, involved spin coating a layer of SU8 (a commercial photoresist) on a glass substrate, followed by a layer of Probes nanoparticles in solution, then a second layer of photoresist. We would then scan with a laser at low power (so as not to polymerize the SU8) to find the fluorescent signal from the nanoparticles in order to fabricate a structure around each one. This ran into several problems. In early trials we used SU8 2005 followed by SU8 2002 in order to get thin 5 μm and 2μm layers respectively and facilitate creation of structures. However, the thinner 2002 layer ate away at the 2005 layer, rendering the samples useless. Next, we tried a layer of SU8 2005 followed by a layer of SU8 2025 (to produce a 25 μm layer), but the signal from the nanoparticles wasn’t strong enough to find them. Deciding to regroup, we moved on.
The next project was fabricating the structures to which we would couple a nanoparticle, refining their structure without the added complication of dealing with nanoparticles for the moment. To that end, we used Low One-Photon Absorption (LOPA) Direct Laser Writing (DLW) to develop structures for square and hexagonal hole arrays with a period of 700 nm and a central defect, suspended by fabricated legs. We tuned the parameters such as power, scanning velocity and defect size for the different structures to achieve a fine, clean mesh and a stable 3D structure.
Having achieved a desirable 3D structure, we returned to nanoparticles, this time incorporating Ag@Fe3O4 nanoflowers sandwiched between layers of SU8 photoresist on a substrate. These nanoflowers produced a stronger signal than the Probes nanoparticles and we successfully incorporated single nanoparticles at the center of 2D nanoparticle arrays.
Top