Electropolymerized Molecularly Imprinted Polymer Synthesis Guided by an Integrated Data-Driven Framework for Cortisol...

Area of Science:

• Materials Science
• Analytical Chemistry
• Biomedical Engineering

Background:

• Molecularly imprinted polymers (MIPs) are synthetic receptors mimicking antibodies for biosensor applications.
• Electropolymerization offers a cost-effective method for *in situ* MIP synthesis on electrodes.
• Optimizing electropolymerized MIPs (e-MIPs) is complex due to numerous synthesis parameters, often relying on trial-and-error.

Purpose of the Study:

• To develop a data-driven framework using machine learning to optimize e-MIP synthesis for enhanced biosensor performance.
• To establish quantitative relationships between e-MIP synthesis parameters and their sensing capabilities.
• To demonstrate a novel, generally applicable approach for accelerating biosensor material development.

Main Methods:

• Fabrication of cortisol-imprinted polypyrrole e-MIPs using 72 parameter sets.
• Performance evaluation using a 12-channel potentiostat.
• Application of Gaussian process (GP) modeling for surrogate modeling and uncertainty quantification.
• Global sensitivity analysis using Sobol indices to identify key synthesis parameters.

Main Results:

• A GP surrogate model was built to predict e-MIP sensing performance.
• Sensitivity analysis revealed critical synthesis parameters and their interrelations.
• Optimized synthesis parameters, guided by GP predictions and local sensitivity analysis, led to a 1.5-fold increase in sensitivity.
• Experimental validation confirmed the enhanced performance of the optimized e-MIPs.

Conclusions:

• A machine learning framework, integrating GP modeling and sensitivity analysis, effectively optimizes e-MIP synthesis for biosensors.
• This approach significantly improves biosensor sensitivity and reduces development time compared to traditional methods.
• The framework is expandable and applicable to the development of various other sensing materials and biosensor platforms.