Welcome to SPUR Research Showcase 2025 Students are presenting their research in a variety of disciplines, and we are excited for you to see their work. Please note that as a research centered university, we support research opportunities in a wide array of areas; some content may not be appropriate for all ages or may be upsetting. Please understand that the views and opinions expressed in the presentations are those of the participants and do not necessarily reflect UCLA or any policy or position of UCLA. By clicking on the "Agree" button, you understand and agree to the items above.

Week 10 Summer Undergraduate Research Showcase SURP 5- 3:30PM

Wednesday, August 27 3:30PM – 5:00PM

Location: Online - Live

The Zoom event has ended.

Presentation 1
Rohan Soni, Alexander Graening, Puneet Gupta
Generating and Benchmarking 3D Integrated Circuit Design Descriptions

Chiplet-based systems, including three-dimensional stacked integrated circuits (3D ICs), leverage advanced packaging to modularize designs for higher yields, flexibility, and scalability. However, the IC design process relies on tools with different system definition requirements. Fabrication tools require detailed physical data (chiplet sizes, orientations, electrical bump grids, etc.) in formats like the industry-standard 3Dblox, while high-level design space exploration tools, such as the CATCH cost-analyzer, use more abstract parameters (costs of assembly processes, test processes, die material, etc.). A key challenge is the lack of automated tools to translate between these formats. To bridge this gap, we present a free, open-source tool for bidirectional translation between the CATCH format and 3Dblox, preserving unsupported data in separate log files. We also present a suite of benchmark 3D IC designs to validate our tool. Accuracy is verified through a round-trip conversion test: design costs are calculated in CATCH, converted to a legal 3Dblox format, converted back, and the final cost is compared with the original. Our tool functions correctly on these benchmarks. This work enables a workflow where designers can use the industry-standard 3Dblox for physical design, then assess the cost of high-level choices (i.e. degree of chiplet partitioning) within the CATCH framework. By facilitating early-stage cost analysis of 3D IC designs, our tool helps reduce time and expense in the IC design process. The conversion tool is available at https://github.com/nanocad-lab/3dblox_testcase_generator/tree/main.


Presentation 2
JUSTIN T. KAMEI, Kaushal H. Patel, Michael J. Rose, Katie S. Kubiatko, Elle K. McCue, Alexandra S. Voloshina, Tyler R. Clites
Design and Validation of an Exoskeleton Following Ankle Arthrodesis

Ankle arthritis is caused by cartilage wear during rotation of the ankle. To reduce pain, the tibia and talus can be surgically fused to prevent such rotation in a procedure known as ankle arthrodesis. However, this treatment results in overloading adjacent joints, especially the metatarsals, which can lead to additional pain. Furthermore, a fused ankle produces less power than a healthy ankle, reducing the propulsive force generated when walking and increasing compensation at the knee and hip. Hence, we aimed to develop an exoskeleton that exerts propulsive force during gait to restore healthy kinetics. To assist individuals with ankle arthrodesis, our system needed to remain in contact with the ground throughout stance. We created preliminary models in MATLAB to evaluate component geometries, and used SolidWorks to 3D model the system. Here, we designed and built an exoskeleton leveraging cable tension to exert a propulsive force via plunger actuation against the ground. We provided actuation via a Humotech, a device that connects a programmed controller with offboard motors to pull on various cables. By pairing our exoskeleton with a curved bottom boot, we provided a platform over which the affected leg can be easily propelled. Manual testing of the exoskeleton confirmed both smooth actuation during stance and plunger restoration during swing. Walking with manual actuation validated that the exoskeleton remained in contact with the ground during stance. Future experiments will seek to implement gait detection and characterize exoskeleton responsiveness to an assistive force profile.


Presentation 3
ADEMOLE A.X. TURNER, Alexandra Grishchenko, Brendan Russ, Carissa N. Eisler
Optimizing the Self-Assembly of Monolayer Lead Halide Perovskite Nanocrystal Films

Lead halide perovskite (CsPbX3) nanocrystals are highly favorable for optoelectronics because they are bright and solution processible. However, creating uniform, ordered films to fabricate high quality optoelectronics remains a significant challenge. In this project, we developed a method to assemble lead halide perovskite nanocrystals into dense, ordered monolayer films via a liquid–liquid interface stamping technique. We synthesized colloidal nanocrystals using ligand- controlled room-temperature chemistry. The nanocrystals were 10 nm in diameter as characterized by transmission electron microscopy, with a peak emission wavelength of 518nm and a full width at half maximum of 14nm in solution. We assemble nanocrystal films by leveraging interparticle forces on an immiscible heptane–glycerol triacetate interface. We varied solvent volume, nanocrystal concentration, and assembly time between trials to control evaporation rate and particle packing. The optimized process resulted in using an optical density of 2.5 at 335 nanometers, 1mL of the colloidal solution, and one hour assembly time to produce films with a monolayer area coverage of 96.4%, which was confirmed via atomic force microscopy and wide field fluorescence microscopy. These high-quality perovskite nanocrystal films provide a robust platform for studying optoelectronic performance and integrating perovskites into next-generation light-emitting and photovoltaic devices.


Ray Lu
Presentation 4
Dachi Mkheidze
Presentation 5
Presentation 6
Justin Luc, Michela Mattia, Benjamin S. Williams
Streamlining the Collection, Correction, & Analysis of THz QCL Data

Terahertz quantum cascade lasers (THz QCLs), first demonstrated in 2002, are relatively young compared to their conventional mid-infrared (mid-IR) counterparts. As this technology continues to evolve, spectro- scopic applications are of great interest to scientists in the field. The purpose of this project is to develop a graphical user interface (GUI) that collects and corrects THz QCL spectra, streamlining the data acquisition process in our lab. The program establishes a connection to the equip- ment (detectors, NI DAQ, delay line stage) and initializes relevant scan parameters. The setup consists of two Michelson Interferometers, ter- minated by two detectors which capture the interference patterns of two lasers, a THz QCL and a Helium-Neon (HeNe) laser. The THz QCL inter- ferogram (IFG) appears jagged due to noise and detector non-idealities, a problem solved by using the HeNe laser as a reference for data resa- mpling, which makes the THz IFG’s features clearer. After the computer completes the corrections, the resulting IFG appears more sinusoidal, similar to the HeNe IFG. After saving the interferometric data, the GUI allows the user to adjust parameters to perform a Fast Fourier Transform (FFT). These parameters include Apodization, which applies a windowing function to the interferometric data to improve FFT results, Zero Padding, used to optimize the FFT algorithm efficiency and speed, and Phase Cor- rection, used to compensate for phase shifts in optical signals. The user can then plot the resulting spectrum in either linear or logarithmic scale and save figures and numerical data as needed.