Speaker: Dr. Yaoli Zhao, Tufts University
Title: "Nanomechanical Photothermal Spectroscopy for Molecular-Level Sensing"
Abstract: One of the critical barriers to addressing pressing environmental challenges, from plastic pollution to pervasive “forever chemicals,” is the measurement and characterization of chemicals. In many real-world systems, we lack molecularly specific, real-time data needed to track contaminants, understand their transformations, and verify circularity at scale. Conventional chemical sensors, however, often face fundamental trade-offs between sensitivity, selectivity, reversibility, and scalability, particularly at micro- and nanoscale dimensions.

In this seminar, I will present a nanomechanical photothermal spectroscopy platform that integrates infrared vibrational spectroscopy with micro- and nano-mechanical transducers to enable highly sensitive, receptor-free molecular detection. By harnessing the photothermal effect generated during infrared absorption and converting it into measurable mechanical or thermoelectric signals, this approach achieves attogram-level mass sensitivity under ambient conditions while preserving molecular specificity. I will discuss the fundamental mechanisms underlying photothermal nanomechanical transduction, device design strategies that mitigate sensor-to-sensor variability, and interface engineering approaches that enable robust chemical identification without relying on fragile receptor coatings.

I will further demonstrate how these principles are translated into deployable sensing systems for real-world applications, including parts-per-trillion detection of PFAS compounds, real-time pre-concentration strategies that overcome the limited surface area of micro- and nano-sensors, and multimodal platforms for plastic identification to support circular economy initiatives. In particular, by integrating feature extraction with machine learning-based classification models, the platform achieves plastic type identification accuracies exceeding 99%, even for chemically similar polymers and mixed or contaminated samples. Together, these works illustrate how coupling molecular spectroscopy, nanomechanical transduction, and AI-enabled data analytics enables scalable, field-relevant sensing technologies that bridge fundamental science and practical applications.