UTA researcher aims to simplify chemical analysis
Researchers at The University of Texas at Arlington are developing a method to make automated chemical analysis more accessible to global industries.
Chemical analysis is essential to a range of common processes, such as medical diagnostics, environmental impact reporting, crime forensics and even nutritional value measurements for food. As technology has advanced, manufacturers have created smart instruments that use optimization and automation to make chemical analysis more efficient.
But these instruments require highly specialized operators, limiting the number of global users capable of operating them to slightly more than a handful.
Kevin Schug, Shimadzu Distinguished Professor of Analytical Chemistry, received a three-year, $325,702 grant from the National Science Foundation to design an application that will make one such example—the Nexera UC, a machine created by Shimadzu Scientific Instruments that automates chemical-compound sample preparation and analysis—easier to use.
“The instruments’ optimizations settings are too complex for the human mind to comprehend,” Schug said. “It could take months for a new user to set up and use the machine to achieve reliable results.”
Schug’s goal is to create a database where users can input the type of sample they want to analyze and then receive precise instructions on how to set up the instrument. This solution would allow Shimadzu Scientific Instruments, a leading manufacturer of precision and measuring instruments, to increase the marketability of its chemical analysis equipment.
“Shimadzu’s design makes the instrument broadly applicable. There is almost no small molecule that can’t be analyzed using this equipment,” Schug said. “This database would allow someone to buy the instrument, set it and go.”
The project will be a close collaboration between Schug’s laboratory and the Center on Stochastic Modeling, Optimization & Statistics (COSMOS) in UTA’s College of Engineering. COSMOS researchers specialize in the design and modeling of complex, real-world systems to develop new methods for making decisions.
Victoria Chen, professor of industrial engineering and the center’s director, said her team is uniquely equipped to integrate machine learning into analytical instrument development.
“In contrast to the popular big data scenario usually employed by machine-learning techniques, COSMOS has expertise in creating optimization methods when data collection is limited, as is the case with expensive laboratory experiments,” Chen said. “If successful, the database will have a broad impact, making this sophisticated instrument available for diverse applications in academic and professional settings.”