Collaborative Research: 2D Chiral Fingerprint Metasensor Empowered by Machine Learning for Chiral Molecule Sensing

About This Grant

Nontechnical description: Chiral molecule sensing plays an important role in many areas such as pharmaceutical industry, biomedical diagnostics, and food analysis. It is challenging to accurately quantify the chirality of chiral medium due to the weak intrinsic circular dichroism signal of chiral molecules. Recently, chiral metasurfaces have been used to amplify the circular dichroism signal and improve the sensitivity in circular dichroism spectroscopy for the vibrational transitions of chiral molecules. However, the detection sensitivity is quite limited by the chiral metasurface design with certain structural geometries and optical properties. In this project, a new type of 2D chiral fingerprint metasensors with high sensitivity based on ultrathin 2D material chiral metasurfaces will be rapidly designed through a machine learning framework and demonstrated for the detection and identification of various kinds of chiral molecules with high sensing performance. This research will benefit many biomedical and photonic applications in point-of-care healthcare, food analysis, environment monitoring, quantum spectroscopy, and quantum communication. This project also includes educational activities for training graduate students, recruiting students and promoting broad participation, and mentoring high school students through outreach activities. Technical description: Chiral metasurfaces have been utilized to enhance the circular dichroism signal and improve the detection sensitivity for the vibrational transitions of chiral molecules. However, the detection sensitivity is quite limited by the chiral metasurface design with certain structural geometries and optical properties, which are insufficient to cover the whole design space to achieve the optimal sensitivity in chiral molecule sensing. The goal of this project is to study a new type of 2D chiral fingerprint metasensors based on ultrathin 2D material chiral metasurfaces which will be rapidly designed through a machine learning framework and demonstrated for the detection and identification of the mid-infrared vibrational fingerprints of chiral molecules with high sensitivity and selectivity. In this project, a new machine learning algorithmic methodology will be developed for rapid and precise inverse design of 2D chiral fingerprint metasensors with the maximized sensitivity. The underlying mechanism of high sensitivity in 2D chiral fingerprint metasensors will be revealed. Nanofabrication processes will be developed to fabricate the designed 2D material chiral metasurfaces. The 2D chiral fingerprint metasensors will be characterized for demonstrating the detection and identification of various kinds of chiral molecules with high sensitivity and selectivity. The project will advance the rapid design and integration of future 2D material-based photonic and optoelectronic metadevices. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.