Researchers from Texas Christian University (TCU) and Lamar University are using two computational approaches to tackle a challenge at the interface of medicine and materials science: the chemistry behind near-infrared emission in biocompatible graphene quantum dots (GQD).
Their project was inspired by the work of Anton Naumov, a physicist at TCU, who has been developing graphene quantum dots that chemically absorb and emit near-infrared light. Skin and bones are transparent in this light, giving these devices many possible applications in detecting and treating disease. However, it is unclear exactly how and why the dots behave as they do.
“Stable and bright near-infrared fluorescence is a special property, and there are still important questions about why these graphene quantum dots can do it,” said Dr. Janesko, TCU professor and chemistry department chair. “It can be very challenging to make sense of the chemistry of these relatively large and disordered structures.”
The project integrates AI-driven uncertainty quantification in “cyber twins,” leveraging Masud Rana’s research, with Dr. Janesko’s atomistic simulation of chemical structure-property relationships. It uses computer simulations to help predict the chemical details of how and why these nanomaterials emit near-infrared light.
“The goal is to give our experimental colleagues a place to start in tuning the properties of their quantum dots,” Dr. Janesko said.
His team began by searching the literature to identify graphene-type structures previously made by other researchers capable of absorbing and emitting near-infrared light. They found 17 chemical structures for Dr. Rana’s group to use as a basis for their work.
The Lamar team designed a three-step process with AI machine learning to develop ensembles of chemically realistic GQD chromophore structures (the part of a molecule that absorbs and reflects specific light wavelengths). These structures are predicted to reproduce the size, composition, infrared absorption, and near-infrared absorption and emission of real graphene quantum dot samples.
“We are very close to our goal of creating 10,000 structurally diverse and chemically reasonable candidate structures, and hopefully even more by the end of the pilot project,” said Dr. Rana, Lamar University Assistant Professor of Computer Science.
“We have been working closely together with weekly team calls and periodic virtual meetings. We are starting to develop hypotheses as to why these systems perform as they do,” Dr. Janesko added. “Once we wrap up this work, we plan to share our AI-driven simulations Dr. Naumov and work with his team to design experiments to test our hypotheses. Ultimately, we hope to help design new ways to turn these quantum dots into clinically useful tools.”
The approach also can have broader applications, Dr. Rana reported. For example, he is using AI machine learning to analyze other systems, with another of his team hoping to improve methods for biomedical applications such as diagnosing lung cancer. A potential joint future project between the two WelchX researchers will involve graphene quantum dots’ use in energy applications such as batteries.
“The WelchX experience was so special,” said Dr. Rana. “Dr. Benjamin and I started comparing notes early on at the WelchX retreat as we are both in computation, although in very difference disciplines. Together, we are able to expand the impact of our work.”
“WelchX is catalyzing new kinds of interdisciplinary research across Texas,” Dr. Janesko said. “Our research program here at TCU was built with Welch support. Welch continues to help these programs grow by catalyzing cross-disciplinary and cross-institutional collaborations.”