First Week of the Program
Left to right: Flor Nardone, Daniel Ziabicki, Emily DeVeyra, Abdullah Barlas, Jamir Riffas, Hannah Lias, Fiona Kovisto, Charley Davis
Dinner in Paris
Left to right: Jamir Riffas, Fiona Kovisto, Daniel Ziabicki, Emily DeVeyra, Flor Nardone, Charley Davis, Hannah Lias, Abdullah Barlas
After Dinner in Paris
Left to right: Fiona Kovisto, Charley Davis,, Flor Nardone, Jamir Riffas, Hannah Lias, Abdullah Barlas, Daniel Ziabicki, Emily DeVeyra
2025 Projects
Hannah Lias: Thin Film Micro LEDs
Whiteboard Video:
This summer I worked on the fabrication and characterization thin film micro-LEDs for an optical cochlear implant prototype.
The larger goal of this project is to create a prototype for a new kind of cochlear implant that would utilize optical stimulation of the hair cells in the cochlea rather than electrical stimulation. My job was to help create and test a thin, flexible strip of LEDs designed for the cochlea of a live mouse. These LEDs needed to be small (~100 micrometers in width), flexible (1 millimeter radius of curvature), and emit bright light (around 450 nanometer frequency) in order to work for the cochlea - a small, spiral-shaped structure in the inner ear.
We made our LEDs out of a Gallium Nitride (GaN) crystal. When it came to us it had already been doped with a PIN structure, with multiple quantum wells in between the p and n-doped layers of GaN. The crystal sits upon a sapphire substrate, with a two-dimensional layer of hexagonal Boron Nitride (h-BN) in between the GaN and the sapphire. The h-BN acts as a mechanical release layer, allowing the GaN LEDs to be easily picked from the sapphire substrate once completed, and placed on a flexible substrate.
We began by establishing a p-contact for our LEDs. We used UV lithography to create a pattern on the p-GaN surface, and then evaporated metal onto it. Once the excess metal was removed, we had a metal p-contact. After that we etched down to the n-contact, and evaporated metal onto it as well.
At this point, I had the chance to do some characterization. To do this, I used two microscopic probes attached to a power source. I used code to run voltage and current sweeps through the probes (placed on the p and n contacts of the LEDs) and an optical fiber to measure the wavelength and optical power of the LEDs. I found that the LEDs were functional, but not quite as powerful as we had hoped (operating at about one tenth of the optical power we would like).
After this we used a plasma source to etch down to the sapphire substrate, isolating the individual LEDs. Unfortunately, we etched a little too far and ended up destroying the metal contacts, rendering the LEDs inert. We didn’t have to time to start again from scratch before my internship was up, but the team will continue without me.
I had an amazing time in my internship this summer. I learned so much about the world, myself, and my future. I fell in love with nanotech, and with the city of light, and I can’t wait to come back and have more incredible adventures. I would like to thank Jules Duraz and Sophie Bouchoule for being amazing research mentors, Steve Yalisove and the University of Michigan team for organizing this internship, and everyone in the cohort for being incredible friends and partners in crime this summer.
Top
Daniel Ziabicki: Building the EICHEL Autocorrelator
Whiteboard Video:
This past summer, I worked in the Laser R&D group at the Laboratoire pour l'utilisation des lasers intenses (LULI) to develop a scanning autocorrelator to measure the pulse length of the picosecond beam at the LULI2000, one of the highest energy lasers in the world.
An autocorrelator is an optical device that can be used to characterize laser pulses shorter than conventional electronics can react to. Scanning autocorrelators function by splitting a laser pulse into two paths and time delaying one with respect to the other. When the pulses are recombined in a nonlinear crystal, the resulting signal is correlated with the overlap between the pulses. By scanning across a range of offsets, information about the length of a single pulse can be obtained from several measurements.
I began my project by connecting the necessary hardware to the control software and learning how the two components work together. This involved analyzing design schematics, testing hardware, and debugging the control software with the help of a collaborator at GSI who is an expert on the software. The process of calibrating the software was never recorded, so we sought to understand where these seemingly magical values came from so that we could reproduce them on our end. Together, we were able to successfully interface the electronics with the control software, allowing us to proceed further. I then miniaturized all of the hardware to fit on the optical table and soldered new connectors for more flexibility.
The second phase of this project involved placing and aligning the optics onto the optical table. The first complication here was that, with an approximately 2-meter beam line, the tolerances for any component were extremely low. The beam tended to become elevated and angled after leaving the custom components. To adjust for these offsets, we had to reorient the mirrors in the custom mounts to neutralize this effect. The second major complication was generating the third harmonic light from the two pulses. The two paths must have an identical length to maximize overlap, the pulses must be focused precisely onto the BBO crystal to generate any nonlinear effect, and the crystals are extremely sensitive to orientation.
Due to time constraints, I was unable to produce the third harmonic light and, thus, was unable to make any measurement. I left behind detailed notes of the ongoing issues with the project, and I hope that it will soon be operational.
I had an incredible summer here in Paris, and I would like to thank Loïc Meingien and the LULI team for being so helpful and welcoming. I’d also like to thank the NSF and the University of Michigan for funding and organizing this program.
Top
Emily DeVerya: Generation of Structured Light for Free Space Quantum Communications
Whiteboard Video:
This summer I worked with Dr. Juan Rafael Álvarez on the “Generation of Structured Light for Free Space Quantum Communications” at Télécom Paris. I used a spatial light modulator (SLM) for generating spatial distributions of light.
I used the SLM to shape the beam into arbitrary transverse forms (Laguerre-Gaussian, 5x5 beam grid, “Merci Beaucoup”, anything!) using a He-Ne laser, a beam expander for collimation and to increase the spot size, and a half waveplate for rotating the incoming polarization. The SLM is tilted at a small angle and displays a computer generated hologram (CGH) on the screen, which is the phase (or angular spectrum) of the Fourier transform for the image. The image is retrieved and resized onto a camera with a 4f system.
After being able to generate arbitrary forms with the SLM, I set up a double pass configuration where one half of the SLM is used to imprint a phase onto the other half, mimicking the effect of using two optical elements. The large active area of the SLM, combined with the control over the number and size of beams incident on the screen, opens up many possibilities for simulating a number of various optical components with a single device. To do this for two components, the 0th order mode reflected off the SLM first pass is redirected using mirrors such that a lens is placed one focal distance away from the SLM first half where the SLM shapes a beam in frequency using a mask. The lens takes the spatial Fourier transform of the phase mask displayed, then the Fourier plane is positioned onto the other half of the SLM, allowing for changes to be made in the frequency domain and beam shaping. We chose to have one half of the SLM to be a grating or N-slit aperture with user defined dimension, and the other half be the phase mask to simulate either propagation or a lens varying in focal length. This is to observe the Talbot effect. This occurs when a planewave passes through a diffraction grating, and the grating self images at certain distances. This effect has useful applications in quantum communications, electron microscopy, optical metrology, and more. Here, I use the SLM to replicate the conditions needed to observe this effect without extra components and needing to move the camera in various positions.
I used the API from ThorCam and the SDK from HOLOEYE to generate a script in Python that allows the user to define the shape of the grating or aperture on on side, then capture hundreds of images in a minute or two while looping through the phase mask for beam propagation or variable focal length on the other side of the SLM. This makes using the camera and SLM experimentally automated and efficient in data acquisition for multiple SLM configurations. I wrote another script to process the images and measure the intensity profiles. The intensity profiles are then plotted against the position x on the camera and the simulated propagation distance or variable focal length. I retrieved data on the Talbot effect, and the results will be refined in the upcoming months. Shaping the gratings and proper configuration for the SLM will continue to be studied, but my time here has come to an end.
The experiment will continue to be used in the team, as the results appear to be promising for an educational publication. I would like to thank Dr. Juan Rafael Álvarez, Dr. Nicholas Fabre, and Dr. Tanya Sharma for their help as well as NSF and the University of Michigan for this amazing opportunity!
Top
Flor Nardone: Behavior of a gold nanorod placed within a hybrid optical cavity
Whiteboard Video:
This summer, I conducted research on a gold nanorod placed within a hybrid optical cavity. While I participated in some experimental work, I mostly focused on building simulations that could model and predict the systems behaviors.
Experimental Work
I became familiar with the general setup and gained hands-on experience with a variety of optical tools. I learned how to properly set up and align the entire system. However, my experimental work was limited due to persistent issues with the laser. This includes fluctuation in the power and problems with the dispersion. Issues we measured and assessed using the proper optical tools. There were further issues including the fact that the alignment would change depending on the pulse picker value we give it, which should not be the case. Not only this, but the temperature in the lab fluctuate due to air conditioning problems in the lab. After these issues, and consultation with my lab group we concluded that a better use of my time would be to continue my simulation work.
Simulation Work
My script is an adaptation of a script provided to me by one of the graduate students in my lab (Benedict S. Morris). I received help from him as well as from Luis Santos and guidance from Professeur Bruno Palpant to write this script.
My Lumerical FDTD script models a hybrid optical cavity that contains a gold rectangular nanorod (AuNR) placed in a defect layer of Alumina, embedded in a multi-layer Bragg reflector. My goal is to probe how the cavity’s geometry influences light–matter coupling, with special attention to the relationship between coupling strength and cavity order (the number of standing‑wave antinodes spanned by the defect layer). Additionally, this simulation can characterize a variety of other properties observed such as the interaction of the E-field on the nanorod, changes in the background index of the sample, etc.
If interested in my script please refer to my Script Report: https://docs.google.com/document/d/1wVCQY_ZhIXRIcG-K8UWoA6hed8heGPjIQCCJRoDht2w/edit?usp=sharing
Grad school and France: Working alongside graduate students full-time has made me open up to the idea of perusing graduate studies to the PhD level. I really enjoyed doing research and was very lucky to form part of such a welcoming lab and work side by side with such amazing grad students. Moreover, the French emphasis on work-life balance definitely took some getting used to, but ultimately allowed me to get to know people in my lab at a more personal level. I learned about their life in and out of work, which made the idea of grad school less scary and intimidating. I am now seriously looking into graduate studies to apply to, and even considering grad school in France.
Top
Abdullah Barlas: Experimental Study of Thermal Effects in Optical Nanofibers
Whiteboard Video:
My participation in the Optics in the City of Light iREU program 2025 provided the opportunity to intern at the Charles Fabry Laboratory, conjoined with Institut d’Optique, working under researcher Dr. Sylvie LeBrun. The project we worked on involved the use of fiber optics, specifically studying the thermal effects on nanofibers.
To become a nanofiber, we stretch an optic fiber until its diameter is near visible light wavelengths. The uniform section is the nanofiber, which is linked to the unstretched portion by two conical tapers. When we do so, the core is so thin that we disregard it, and the cladding becomes the new “core.” When we run a laser through this stretched fiber, the light actually escapes through one of the tapers; it then travels tightly along the nanofiber then reenters through the returning taper.
The evanescent field probes the surrounding medium (such as air or water). The interaction of this field with “pollutants” can alter the original spectrum of the laser by adding new wavelengths. These new wavelengths can be read on a spectrometer — they are the “spectral signature” of the pollutants. The change in energy for each wavelength peak indicates the pollutant present. The goal of this study is for nanofibers to be able to be used to detect pollutants in water.
In optical fibers, there is no energy absorbed due to total internal reflection. However, the stretching of the fiber creates dangling silica bonds at the surface. Light interacts with the dangling silica bonds, causing an absorption of energy, which in turn causes the nanofiber to warm. This warming can cause detrimental effects on the nanofiber. My role during the internship was to study the thermal effects of the fiber by performing a great number of measurements.
The process of stretching begins with an ordinary optic fiber. For our project, we used SMF-28 fiber. To stretch it, we use the “Pull-and-Brush” technique. After cutting a portion of fiber, we stripped the ends and center of the protective coating, as this layer is flammable. Using ethanol we clean the stripped portions to remove any remaining coating. Using software, we specified the final parameters that we would like the nanofiber to have: diameter, length and reduction ratio. We set the parameter on the stretching stage, open the flow of gas to ignite a flame, and began the stretching process.
Once the fiber is done stretching, we carefully transport it to a mount that has magnets to hold it in place. We then position the mount on a translation stage, which is under the view of a microscopic camera. The purpose of the camera is to simultaneously record the diameter of the fiber during measurement. To measure the temperature, a thermocouple (~200 micron diameter) is in a fixed position to record through contact.
We found a symmetrical relationship between the fiber diameter and the temperature across the nanofiber when using a 1480nm infrared laser. With a 532nm green laser, we found the fiber would heat more closer to the side which the laser enters. We tried to change certain variables, such as the type of fiber used, to test whether a more symmetrical pattern was possible, though we did not find this by the end of the program.
I am very grateful for the incredible opportunity and experience that I was provided with by my participation in this REU program. I would like to thank everyone involved from the University of Michigan, École Polytechnique, Institut d’Optique, and the NSF. Please continue to provide such an experience for future cohorts in the years to come!
Top
Jamir Riffas: Cell Biomechanics with Optical Tweezers and FRET Microscopy
Whiteboard Video:
This summer, I had the incredible opportunity to take part in the Optics REU in the City of Light—Paris. I spent two months living at Cité Universitaire, specifically in Maison du Portugal, which became more than just a place to stay—it was a community. I met people from all over the world and quickly built friendships with the other tenants. We shared meals, stories, and late-night conversations that I’ll never forget. If I had to complain about anything, it would be the shower, which was barely shoulder-width. Honestly, though, even that became something to laugh about. During the day, I worked at the Institut d’Optique in the Charles Fabry Lab, under the mentorship of Nathalie Westbrook. Working in the lab was an experience I’ll always treasure. I learned so much—not just about optics or research—but about problem-solving, patience, and how to embrace the process of discovery. Whether it was running experiments, analyzing fluorescence images, or figuring out why something didn’t work (and then finally getting it right), every day challenged me in the best way. In the lab, I worked on projects studying focal adhesions and the protein vinculin, using fluorescence microscopy to understand how cells attach and interact with surfaces. I spent hours using ImageJ to process images, analyze data, and calculate quantum efficiency of different cameras. It was detailed, hands-on work, and it reminded me why I love science. Being part of this program didn’t just teach me technical skills—it strengthened my desire to pursue a PhD and build a career in physics. It confirmed for me that I want to be a scientist, and that this field is where I belong. I’m especially grateful to NSF for supporting programs like this; these REU opportunities are truly priceless, and I don’t take any part of it for granted. Living in Paris, doing real research, and being part of an amazing cohort of seven other students made this summer unforgettable.
Top
Charley Davis: Photografting with Diazonium Salts
Whiteboard Video:
This summer, I worked under Professor Vitor Brasiliense on investigating two-photon photografting with diazonium salts. Photografting with diazonium salts is when a diazonium salt molecule interacts with a photon of significant enough energy, an N2 molecule is released, and a radical is left in its place. This radical then allows this molecule to bond to another diazonium salt or a surface. This reaction generates large structures using only lasers and diazonium salt. The first process I looked at was using a Spatial Light Modulator (SLM). An SLM works similarly to a liquid crystal display (LCD) TV, where crystals are suspended in a liquid, and by applying an electric field, we can get an “image” to form. By reflecting light off an SLM, the intensity wavefront equals the intensity of the image displayed on the SLM. The SLM allows us to graft any image we want onto diazonium salts. The second process I looked at, and focused on for the majority of the summer was just grafting points by focusing the laser directly on the diazonium sample. Here, we used a laser with half the energy required to cause a grafting reaction. Therefore, the only way to cause grafting was to have two photons hit the diazonium salt simultaneously, hence the two-photon process.
The main advantage of this process is that it allows us to break the diffraction limit normally present in photografting. The diffraction limit prevents the points we graft from being smaller than half the wavelength of the laser. So, for our typical 400 nm laser, we could not create points smaller than 200 nm. However, in a two-photon system, we can beat the wavelength divided by two limits since grafting does not occur wherever the laser shines and has a significantly higher probability of happening closer to the center of the beam. So this was how we would prove a two-photon process happened, by observing points with a width less than 400 nm, since we were using an 800 nm beam.
Proving this meant at first quantifying how the power and exposure time of the laser affect the geometry of the points created. To do this, we grafted grids of 40 by 40 using control software that I wrote. The 40 by 40 grids would be split into 25 point groups, each with its own power and exposure time, to have good statistics. Then we tried to find an expression for the “dose”, or the relation between power and time, such that two different power and time pairings can lead to the same point geometry. This process was complicated because our solution was diazonium salts in water so we would heat the water with a high-power laser. The extra energy from the heated water can cause a grafting reaction to occur as well. So for low-dose, but high laser power points, we had to account for this thermal process. These grids of 40 by 40 points would be split up, and each point would be fitted with a Gaussian added to a bump function. The Gaussian, since that was the laser profile, and a bump function to account for thermal processes. We then observed points with a width of less than 400 nm, proving that we saw a two-photon process.
I want to thank Vitor Brasiliense, Jeff Audibert, and Clément Lafargue for welcoming me into their labs and for all of their help this summer. I also want to thank everyone at PPSM for being incredibly kind, helpful, and welcoming.
Top
Fiona Kovisto: Donut Beam for LOPA Direct Laser Writing for Gold Nanoparticle Fabrication
Whiteboard Video:
This summer I worked with Quang Truong Pham, a 3rd year PhD student, in Professor Ngoc Diep Lai’s lab within the LuMIn lab at ENS Paris Saclay on a project for using a donut beam laser for nanoparticle fabrication.
My lab uses the low one photon absorption (LOPA) method of direct laser writing (DLW) for nanoparticle fabrication. DLW is a process where we use a laser to create nanoparticles within a film. We coat a glass slide in a film of a solution of the photoresist SU-8 and gold salt, and when we apply a tightly focused laser beam to it, the SU-8 forms polymers and crystallizes the gold ions into neutral nanoparticles. The specific method is called Low One Photon Absorption because we use a wavelength of light that is on the edge of the photoresist’s absorption band, so the absorbed light can have the precision to do 3D fabrication without being so weak that we need two photons to reach the excited state (Two Photon Absorption). Because our wavelength is still within the absorption band, our laser can have far less power than those used for the other method of 3D microstructure fabrication, Two Photon Absorption.
I began my research by learning how to fabricate using this process that my lab created and printed some structures of interlocking circles and also a micro-scale print of an image of my face, made of nanoparticles. Then, I began working on a Matlab simulation of the intensity profile of the laser I would be using, adapting some previous code created for other versions of the beam, made by a previous graduate student.
My project was to begin working with a donut beam, because Professor Lai is hoping to use a donut beam to fabricate even smaller nanoparticles. Previously, students in my lab have used a gaussian beam, which has a gaussian intensity distribution radially. A donut beam instead has a zero intensity hole in the center, with the most intense region forming a ring around that hole. To make a donut beam we used an optical vortex mask, which creates a radial 0 to 2π phase shift using increasingly thicker steps. By that, I mean that the mask’s thickness at the bottom of the mask is a π phase shift from that of the top of the mask, which we take to be 0, likewise, waves entering the right side of the mask experience a π/2 phase shift whereas waves entering at the left side experience a 3π/2 phase shift. Where such pairs of waves meet, they have total destructive interference, cancelling each other out and leaving a 0 intensity hole. The hope is that when we fabricate with this beam, the SU-8 will polymerize along the brightest intensity ring, pushing the nanoparticles to the central hole, making them smaller than they would be under the gaussian beam.
The second part of my research was taking scans of pre-fabricated gold nanoparticles to get a better understanding of the intensity patterns they created when we scanned them with the donut beam laser, and to optimize the beam profile. This involved perfecting the angle and placement of the mask with respect to the beam, and optimizing it with the pinhole of the laser set up which precisely focuses the beam. When a future student tries DLW fabrication with the donut beam, they’ll be able to use the information and images of the scans I took to better understand the size and shape of the nanoparticles they’re creating.
Top
Whiteboard Video:
This summer I worked on the fabrication and characterization thin film micro-LEDs for an optical cochlear implant prototype.
The larger goal of this project is to create a prototype for a new kind of cochlear implant that would utilize optical stimulation of the hair cells in the cochlea rather than electrical stimulation. My job was to help create and test a thin, flexible strip of LEDs designed for the cochlea of a live mouse. These LEDs needed to be small (~100 micrometers in width), flexible (1 millimeter radius of curvature), and emit bright light (around 450 nanometer frequency) in order to work for the cochlea - a small, spiral-shaped structure in the inner ear.
We made our LEDs out of a Gallium Nitride (GaN) crystal. When it came to us it had already been doped with a PIN structure, with multiple quantum wells in between the p and n-doped layers of GaN. The crystal sits upon a sapphire substrate, with a two-dimensional layer of hexagonal Boron Nitride (h-BN) in between the GaN and the sapphire. The h-BN acts as a mechanical release layer, allowing the GaN LEDs to be easily picked from the sapphire substrate once completed, and placed on a flexible substrate.
We began by establishing a p-contact for our LEDs. We used UV lithography to create a pattern on the p-GaN surface, and then evaporated metal onto it. Once the excess metal was removed, we had a metal p-contact. After that we etched down to the n-contact, and evaporated metal onto it as well.
At this point, I had the chance to do some characterization. To do this, I used two microscopic probes attached to a power source. I used code to run voltage and current sweeps through the probes (placed on the p and n contacts of the LEDs) and an optical fiber to measure the wavelength and optical power of the LEDs. I found that the LEDs were functional, but not quite as powerful as we had hoped (operating at about one tenth of the optical power we would like).
After this we used a plasma source to etch down to the sapphire substrate, isolating the individual LEDs. Unfortunately, we etched a little too far and ended up destroying the metal contacts, rendering the LEDs inert. We didn’t have to time to start again from scratch before my internship was up, but the team will continue without me.
I had an amazing time in my internship this summer. I learned so much about the world, myself, and my future. I fell in love with nanotech, and with the city of light, and I can’t wait to come back and have more incredible adventures. I would like to thank Jules Duraz and Sophie Bouchoule for being amazing research mentors, Steve Yalisove and the University of Michigan team for organizing this internship, and everyone in the cohort for being incredible friends and partners in crime this summer.
Top
Daniel Ziabicki: Building the EICHEL Autocorrelator
Whiteboard Video:
This past summer, I worked in the Laser R&D group at the Laboratoire pour l'utilisation des lasers intenses (LULI) to develop a scanning autocorrelator to measure the pulse length of the picosecond beam at the LULI2000, one of the highest energy lasers in the world.
An autocorrelator is an optical device that can be used to characterize laser pulses shorter than conventional electronics can react to. Scanning autocorrelators function by splitting a laser pulse into two paths and time delaying one with respect to the other. When the pulses are recombined in a nonlinear crystal, the resulting signal is correlated with the overlap between the pulses. By scanning across a range of offsets, information about the length of a single pulse can be obtained from several measurements.
I began my project by connecting the necessary hardware to the control software and learning how the two components work together. This involved analyzing design schematics, testing hardware, and debugging the control software with the help of a collaborator at GSI who is an expert on the software. The process of calibrating the software was never recorded, so we sought to understand where these seemingly magical values came from so that we could reproduce them on our end. Together, we were able to successfully interface the electronics with the control software, allowing us to proceed further. I then miniaturized all of the hardware to fit on the optical table and soldered new connectors for more flexibility.
The second phase of this project involved placing and aligning the optics onto the optical table. The first complication here was that, with an approximately 2-meter beam line, the tolerances for any component were extremely low. The beam tended to become elevated and angled after leaving the custom components. To adjust for these offsets, we had to reorient the mirrors in the custom mounts to neutralize this effect. The second major complication was generating the third harmonic light from the two pulses. The two paths must have an identical length to maximize overlap, the pulses must be focused precisely onto the BBO crystal to generate any nonlinear effect, and the crystals are extremely sensitive to orientation.
Due to time constraints, I was unable to produce the third harmonic light and, thus, was unable to make any measurement. I left behind detailed notes of the ongoing issues with the project, and I hope that it will soon be operational.
I had an incredible summer here in Paris, and I would like to thank Loïc Meingien and the LULI team for being so helpful and welcoming. I’d also like to thank the NSF and the University of Michigan for funding and organizing this program.
Top
Emily DeVerya: Generation of Structured Light for Free Space Quantum Communications
Whiteboard Video:
This summer I worked with Dr. Juan Rafael Álvarez on the “Generation of Structured Light for Free Space Quantum Communications” at Télécom Paris. I used a spatial light modulator (SLM) for generating spatial distributions of light.
I used the SLM to shape the beam into arbitrary transverse forms (Laguerre-Gaussian, 5x5 beam grid, “Merci Beaucoup”, anything!) using a He-Ne laser, a beam expander for collimation and to increase the spot size, and a half waveplate for rotating the incoming polarization. The SLM is tilted at a small angle and displays a computer generated hologram (CGH) on the screen, which is the phase (or angular spectrum) of the Fourier transform for the image. The image is retrieved and resized onto a camera with a 4f system.
After being able to generate arbitrary forms with the SLM, I set up a double pass configuration where one half of the SLM is used to imprint a phase onto the other half, mimicking the effect of using two optical elements. The large active area of the SLM, combined with the control over the number and size of beams incident on the screen, opens up many possibilities for simulating a number of various optical components with a single device. To do this for two components, the 0th order mode reflected off the SLM first pass is redirected using mirrors such that a lens is placed one focal distance away from the SLM first half where the SLM shapes a beam in frequency using a mask. The lens takes the spatial Fourier transform of the phase mask displayed, then the Fourier plane is positioned onto the other half of the SLM, allowing for changes to be made in the frequency domain and beam shaping. We chose to have one half of the SLM to be a grating or N-slit aperture with user defined dimension, and the other half be the phase mask to simulate either propagation or a lens varying in focal length. This is to observe the Talbot effect. This occurs when a planewave passes through a diffraction grating, and the grating self images at certain distances. This effect has useful applications in quantum communications, electron microscopy, optical metrology, and more. Here, I use the SLM to replicate the conditions needed to observe this effect without extra components and needing to move the camera in various positions.
I used the API from ThorCam and the SDK from HOLOEYE to generate a script in Python that allows the user to define the shape of the grating or aperture on on side, then capture hundreds of images in a minute or two while looping through the phase mask for beam propagation or variable focal length on the other side of the SLM. This makes using the camera and SLM experimentally automated and efficient in data acquisition for multiple SLM configurations. I wrote another script to process the images and measure the intensity profiles. The intensity profiles are then plotted against the position x on the camera and the simulated propagation distance or variable focal length. I retrieved data on the Talbot effect, and the results will be refined in the upcoming months. Shaping the gratings and proper configuration for the SLM will continue to be studied, but my time here has come to an end.
The experiment will continue to be used in the team, as the results appear to be promising for an educational publication. I would like to thank Dr. Juan Rafael Álvarez, Dr. Nicholas Fabre, and Dr. Tanya Sharma for their help as well as NSF and the University of Michigan for this amazing opportunity!
Top
Flor Nardone: Behavior of a gold nanorod placed within a hybrid optical cavity
Whiteboard Video:
This summer, I conducted research on a gold nanorod placed within a hybrid optical cavity. While I participated in some experimental work, I mostly focused on building simulations that could model and predict the systems behaviors.
Experimental Work
I became familiar with the general setup and gained hands-on experience with a variety of optical tools. I learned how to properly set up and align the entire system. However, my experimental work was limited due to persistent issues with the laser. This includes fluctuation in the power and problems with the dispersion. Issues we measured and assessed using the proper optical tools. There were further issues including the fact that the alignment would change depending on the pulse picker value we give it, which should not be the case. Not only this, but the temperature in the lab fluctuate due to air conditioning problems in the lab. After these issues, and consultation with my lab group we concluded that a better use of my time would be to continue my simulation work.
Simulation Work
My script is an adaptation of a script provided to me by one of the graduate students in my lab (Benedict S. Morris). I received help from him as well as from Luis Santos and guidance from Professeur Bruno Palpant to write this script.
My Lumerical FDTD script models a hybrid optical cavity that contains a gold rectangular nanorod (AuNR) placed in a defect layer of Alumina, embedded in a multi-layer Bragg reflector. My goal is to probe how the cavity’s geometry influences light–matter coupling, with special attention to the relationship between coupling strength and cavity order (the number of standing‑wave antinodes spanned by the defect layer). Additionally, this simulation can characterize a variety of other properties observed such as the interaction of the E-field on the nanorod, changes in the background index of the sample, etc.
If interested in my script please refer to my Script Report: https://docs.google.com/document/d/1wVCQY_ZhIXRIcG-K8UWoA6hed8heGPjIQCCJRoDht2w/edit?usp=sharing
Grad school and France: Working alongside graduate students full-time has made me open up to the idea of perusing graduate studies to the PhD level. I really enjoyed doing research and was very lucky to form part of such a welcoming lab and work side by side with such amazing grad students. Moreover, the French emphasis on work-life balance definitely took some getting used to, but ultimately allowed me to get to know people in my lab at a more personal level. I learned about their life in and out of work, which made the idea of grad school less scary and intimidating. I am now seriously looking into graduate studies to apply to, and even considering grad school in France.
Top
Abdullah Barlas: Experimental Study of Thermal Effects in Optical Nanofibers
Whiteboard Video:
My participation in the Optics in the City of Light iREU program 2025 provided the opportunity to intern at the Charles Fabry Laboratory, conjoined with Institut d’Optique, working under researcher Dr. Sylvie LeBrun. The project we worked on involved the use of fiber optics, specifically studying the thermal effects on nanofibers.
To become a nanofiber, we stretch an optic fiber until its diameter is near visible light wavelengths. The uniform section is the nanofiber, which is linked to the unstretched portion by two conical tapers. When we do so, the core is so thin that we disregard it, and the cladding becomes the new “core.” When we run a laser through this stretched fiber, the light actually escapes through one of the tapers; it then travels tightly along the nanofiber then reenters through the returning taper.
The evanescent field probes the surrounding medium (such as air or water). The interaction of this field with “pollutants” can alter the original spectrum of the laser by adding new wavelengths. These new wavelengths can be read on a spectrometer — they are the “spectral signature” of the pollutants. The change in energy for each wavelength peak indicates the pollutant present. The goal of this study is for nanofibers to be able to be used to detect pollutants in water.
In optical fibers, there is no energy absorbed due to total internal reflection. However, the stretching of the fiber creates dangling silica bonds at the surface. Light interacts with the dangling silica bonds, causing an absorption of energy, which in turn causes the nanofiber to warm. This warming can cause detrimental effects on the nanofiber. My role during the internship was to study the thermal effects of the fiber by performing a great number of measurements.
The process of stretching begins with an ordinary optic fiber. For our project, we used SMF-28 fiber. To stretch it, we use the “Pull-and-Brush” technique. After cutting a portion of fiber, we stripped the ends and center of the protective coating, as this layer is flammable. Using ethanol we clean the stripped portions to remove any remaining coating. Using software, we specified the final parameters that we would like the nanofiber to have: diameter, length and reduction ratio. We set the parameter on the stretching stage, open the flow of gas to ignite a flame, and began the stretching process.
Once the fiber is done stretching, we carefully transport it to a mount that has magnets to hold it in place. We then position the mount on a translation stage, which is under the view of a microscopic camera. The purpose of the camera is to simultaneously record the diameter of the fiber during measurement. To measure the temperature, a thermocouple (~200 micron diameter) is in a fixed position to record through contact.
We found a symmetrical relationship between the fiber diameter and the temperature across the nanofiber when using a 1480nm infrared laser. With a 532nm green laser, we found the fiber would heat more closer to the side which the laser enters. We tried to change certain variables, such as the type of fiber used, to test whether a more symmetrical pattern was possible, though we did not find this by the end of the program.
I am very grateful for the incredible opportunity and experience that I was provided with by my participation in this REU program. I would like to thank everyone involved from the University of Michigan, École Polytechnique, Institut d’Optique, and the NSF. Please continue to provide such an experience for future cohorts in the years to come!
Top
Jamir Riffas: Cell Biomechanics with Optical Tweezers and FRET Microscopy
Whiteboard Video:
This summer, I had the incredible opportunity to take part in the Optics REU in the City of Light—Paris. I spent two months living at Cité Universitaire, specifically in Maison du Portugal, which became more than just a place to stay—it was a community. I met people from all over the world and quickly built friendships with the other tenants. We shared meals, stories, and late-night conversations that I’ll never forget. If I had to complain about anything, it would be the shower, which was barely shoulder-width. Honestly, though, even that became something to laugh about. During the day, I worked at the Institut d’Optique in the Charles Fabry Lab, under the mentorship of Nathalie Westbrook. Working in the lab was an experience I’ll always treasure. I learned so much—not just about optics or research—but about problem-solving, patience, and how to embrace the process of discovery. Whether it was running experiments, analyzing fluorescence images, or figuring out why something didn’t work (and then finally getting it right), every day challenged me in the best way. In the lab, I worked on projects studying focal adhesions and the protein vinculin, using fluorescence microscopy to understand how cells attach and interact with surfaces. I spent hours using ImageJ to process images, analyze data, and calculate quantum efficiency of different cameras. It was detailed, hands-on work, and it reminded me why I love science. Being part of this program didn’t just teach me technical skills—it strengthened my desire to pursue a PhD and build a career in physics. It confirmed for me that I want to be a scientist, and that this field is where I belong. I’m especially grateful to NSF for supporting programs like this; these REU opportunities are truly priceless, and I don’t take any part of it for granted. Living in Paris, doing real research, and being part of an amazing cohort of seven other students made this summer unforgettable.
Top
Charley Davis: Photografting with Diazonium Salts
Whiteboard Video:
This summer, I worked under Professor Vitor Brasiliense on investigating two-photon photografting with diazonium salts. Photografting with diazonium salts is when a diazonium salt molecule interacts with a photon of significant enough energy, an N2 molecule is released, and a radical is left in its place. This radical then allows this molecule to bond to another diazonium salt or a surface. This reaction generates large structures using only lasers and diazonium salt. The first process I looked at was using a Spatial Light Modulator (SLM). An SLM works similarly to a liquid crystal display (LCD) TV, where crystals are suspended in a liquid, and by applying an electric field, we can get an “image” to form. By reflecting light off an SLM, the intensity wavefront equals the intensity of the image displayed on the SLM. The SLM allows us to graft any image we want onto diazonium salts. The second process I looked at, and focused on for the majority of the summer was just grafting points by focusing the laser directly on the diazonium sample. Here, we used a laser with half the energy required to cause a grafting reaction. Therefore, the only way to cause grafting was to have two photons hit the diazonium salt simultaneously, hence the two-photon process.
The main advantage of this process is that it allows us to break the diffraction limit normally present in photografting. The diffraction limit prevents the points we graft from being smaller than half the wavelength of the laser. So, for our typical 400 nm laser, we could not create points smaller than 200 nm. However, in a two-photon system, we can beat the wavelength divided by two limits since grafting does not occur wherever the laser shines and has a significantly higher probability of happening closer to the center of the beam. So this was how we would prove a two-photon process happened, by observing points with a width less than 400 nm, since we were using an 800 nm beam.
Proving this meant at first quantifying how the power and exposure time of the laser affect the geometry of the points created. To do this, we grafted grids of 40 by 40 using control software that I wrote. The 40 by 40 grids would be split into 25 point groups, each with its own power and exposure time, to have good statistics. Then we tried to find an expression for the “dose”, or the relation between power and time, such that two different power and time pairings can lead to the same point geometry. This process was complicated because our solution was diazonium salts in water so we would heat the water with a high-power laser. The extra energy from the heated water can cause a grafting reaction to occur as well. So for low-dose, but high laser power points, we had to account for this thermal process. These grids of 40 by 40 points would be split up, and each point would be fitted with a Gaussian added to a bump function. The Gaussian, since that was the laser profile, and a bump function to account for thermal processes. We then observed points with a width of less than 400 nm, proving that we saw a two-photon process.
I want to thank Vitor Brasiliense, Jeff Audibert, and Clément Lafargue for welcoming me into their labs and for all of their help this summer. I also want to thank everyone at PPSM for being incredibly kind, helpful, and welcoming.
Top
Fiona Kovisto: Donut Beam for LOPA Direct Laser Writing for Gold Nanoparticle Fabrication
Whiteboard Video:
This summer I worked with Quang Truong Pham, a 3rd year PhD student, in Professor Ngoc Diep Lai’s lab within the LuMIn lab at ENS Paris Saclay on a project for using a donut beam laser for nanoparticle fabrication.
My lab uses the low one photon absorption (LOPA) method of direct laser writing (DLW) for nanoparticle fabrication. DLW is a process where we use a laser to create nanoparticles within a film. We coat a glass slide in a film of a solution of the photoresist SU-8 and gold salt, and when we apply a tightly focused laser beam to it, the SU-8 forms polymers and crystallizes the gold ions into neutral nanoparticles. The specific method is called Low One Photon Absorption because we use a wavelength of light that is on the edge of the photoresist’s absorption band, so the absorbed light can have the precision to do 3D fabrication without being so weak that we need two photons to reach the excited state (Two Photon Absorption). Because our wavelength is still within the absorption band, our laser can have far less power than those used for the other method of 3D microstructure fabrication, Two Photon Absorption.
I began my research by learning how to fabricate using this process that my lab created and printed some structures of interlocking circles and also a micro-scale print of an image of my face, made of nanoparticles. Then, I began working on a Matlab simulation of the intensity profile of the laser I would be using, adapting some previous code created for other versions of the beam, made by a previous graduate student.
My project was to begin working with a donut beam, because Professor Lai is hoping to use a donut beam to fabricate even smaller nanoparticles. Previously, students in my lab have used a gaussian beam, which has a gaussian intensity distribution radially. A donut beam instead has a zero intensity hole in the center, with the most intense region forming a ring around that hole. To make a donut beam we used an optical vortex mask, which creates a radial 0 to 2π phase shift using increasingly thicker steps. By that, I mean that the mask’s thickness at the bottom of the mask is a π phase shift from that of the top of the mask, which we take to be 0, likewise, waves entering the right side of the mask experience a π/2 phase shift whereas waves entering at the left side experience a 3π/2 phase shift. Where such pairs of waves meet, they have total destructive interference, cancelling each other out and leaving a 0 intensity hole. The hope is that when we fabricate with this beam, the SU-8 will polymerize along the brightest intensity ring, pushing the nanoparticles to the central hole, making them smaller than they would be under the gaussian beam.
The second part of my research was taking scans of pre-fabricated gold nanoparticles to get a better understanding of the intensity patterns they created when we scanned them with the donut beam laser, and to optimize the beam profile. This involved perfecting the angle and placement of the mask with respect to the beam, and optimizing it with the pinhole of the laser set up which precisely focuses the beam. When a future student tries DLW fabrication with the donut beam, they’ll be able to use the information and images of the scans I took to better understand the size and shape of the nanoparticles they’re creating.
Top