10:15-11:45 AM Oral Competition Presentations
B5: Materials Science and Engineering (Mulder Hall 221)
10:15-11:45 AM Oral Competition Presentations
B5: Materials Science and Engineering (Mulder Hall 221)
10:15-10:27 TimeGPT-Driven Predictive and Interpretable Modeling for Critical Materials in Renewable Energy Applications
MD Shafikul Islam (LSU A&M)
Shafikul Islam, Mahathir M Bappy, Jeimy Martinez De La Hoz
The transition to a low-carbon economy, essential for reducing global greenhouse gas emissions, depends on clean energy technologies reliant on critical materials. Among these, gallium, indium, and cobalt are vital for solar photovoltaics, electric vehicles, and energy storage systems but face significant global supply-chain uncertainties. Disruptions in their availability can hinder manufacturing and deployment, making reliable forecasts crucial for mitigating risks in policymaking. Existing machine learning methods often struggle with scarce data and limited integration of external factors. To address these challenges, we developed a transformer-based foundation model (TimeGpt) to forecast global production trajectories for critical materials through 2035. Additionally, we integrated SHAP (Shapley Additive exPlanations) for model interpretability, enabling the identification and quantification of critical market drivers. Empirical evaluations show that the proposed approach outperforms traditional methods across multiple error metrics, projecting a moderate rise in production, with cobalt showing the greatest growth. SHAP analysis reveals that lithium-ion battery prices and EV market share significantly influence outcomes. The framework enhances resilience by forecasting demand and identifying key market drivers, supporting the steady availability of critical materials, and aiding policymakers and manufacturers in transitioning to a low-carbon economy.
10:30-10:42 Surface tension on the nanoparticle synthesis by laser ablation in liquid
Ka'Tra Winchester (GSU)
Haeyeon Yang, Ka’Tra Winchester, Noble Agyeman-Bobie, Tatenda Kasirori, Enoch Owoade
Laser ablation has emerged as a sustainable and green method to synthesize nanoparticles (NPs) without the heavy use of chemicals. The process involves irradiating a target with nanosecond laser pulses to generate NPs. Upon the impact of laser pulse, atoms absorb energy and excited, leading to the ejection of ions and atoms from the affected surface. These ejected elements expand into the surrounding liquid, form a bubble or laser plume. The interface area gets larger when the surface tension of liquid is smaller. As the plasma plume expands and contracts, it reaches a supersaturated state, at which nuclei form. Eventually, the bubble collapses, releasing NPs into the surrounding liquid, forming colloidal NPs. In this talk, we present how the control of the bubble volume at supersaturation affects the NP size, assuming the volume depends on the temperature of liquid. When bubble volume is reduced under controlled conditions, precursors are in closer proximity due to the higher concentration at a smaller volume, creating higher supersaturation. This may yield larger NPs. Notably, bubble volume is influenced by the liquid's surface tension, which in turn is affected by temperature. By examining the impact of temperature on the surface tension, we aim to determine how variations in temperature influence the volume of bubbles, and inherently, NP size. This study highlights the importance of temperature control in fine-tuning NP production and achieving desired size distributions.