Artificial Intelligence Algorithms Enable Automated Characterization of the Positive and Negative Dielectrophoretic...

Area of Science:

• Microfluidics and Nanotechnology
• Biophysics and Electrophysics
• Artificial Intelligence in Scientific Automation

Background:

• Dielectrophoresis (DEP) is an electrokinetic phenomenon utilized for manipulating micro- and nanoparticles.
• Traditional methods for determining DEP frequency ranges can be time-consuming and complex.
• Automated systems are needed for efficient and precise control in micro/nanoscale assembly.

Purpose of the Study:

• To develop and validate a phenomenological approach for automatically determining DEP frequency ranges.
• To implement an Artificial Intelligence (AI) feedback loop for real-time control of DEP experiments.
• To demonstrate the platform's capability in identifying positive DEP (pDEP) and negative DEP (nDEP) behaviors.

Main Methods:

• Utilized a microfabricated chip with gold microelectrodes connected to a function generator.
• Employed a digital camera and OpenCV for automated image analysis and bead tracking.
• Implemented an AI algorithm to analyze bead movement and adjust AC signal frequency.
• Tested with 3-μm latex microbeads suspended in a medium.

Main Results:

• The AI-guided platform successfully automated the determination of DEP frequency ranges.
• Positive DEP (pDEP) was observed below 500 kHz, and negative DEP (nDEP) above 1 MHz.
• The crossover frequency was identified to be between 500 kHz and 1 MHz.
• Experimental results align with previously published data for latex microbeads.

Conclusions:

• The developed phenomenological approach with AI feedback offers a fully automated method for DEP characterization.
• This automated platform can assist in active particle manipulation for desired outcomes, even without complex model-based predictions.
• The system shows promise for advancing micro- and nano-assembly applications through precise electrokinetic control.