Document Type



Doctor of Philosophy


Electrical Engineering

First Adviser

Tatic-Lucic, Svetlana

Other advisers/committee members

Berdichevsky, Yevgeny; Hatalis, Miltiadis; Hwang, James; Ou-Yang, Daniel; Perry, Susan


Multi-Electrode Array (MEA) systems have been widely used, for decades, in neuroscience research, such as the detection of neuroactive compounds and the study of neuronal electrophysiology and communication. Accurate positioning of neurons on electrodes enables the recording from and stimulation of specified individual neurons on a MEA. Various cell patterning techniques, integrated with MEAs, have been developed to ensure neuron-electrode correspondence, which is the capability that most conventional MEAs lack, because of the random distribution of neurons with respect to electrodes.A novel multi-electrode array system has been designed and developed for active recruitment of neurons on electrodes and the formation of mechanically confined neural networks. In this system, positive dielectrophoresis (DEP) is applied to actively recruit hippocampal neurons to the electrodes of the MEA, where polymer microstructures, such as microchambers and microtrenches are created to effectively define a patterned neuronal network. The dielectrophoresis theoretical calculation and simulation is first examined to prove the feasibility of active positioning of hippocampal neurons using positive DEP. The MEA device, referred to as DEP MEA, is designed, modeled, and fabricated on a quartz glass substrate. The fabricated DEP MEA chip (8 mm by 8 mm) is packaged, and the functionality of our MEA system is verified by active recruitment of embryonic mouse hippocampal neurons, the formation of precisely patterned hippocampal neuronal networks, as well as successful recording of spontaneous and stimulated neuronal potentials, including the propagation of evoked neuronal bursts between electrodes. The cytocompatibility of the top microstructure layer on the DEP MEA, which is a layer of thin cured SU-8 epoxy, is investigated to improve the viability of cultured primary hippocampal neurons. We then investigate the selective trapping (separation) of mouse hippocampal neurons from glial cells using positive DEP, based on their different dielectric and physical properties. The DEP movement of neurons and glial cells in the targeted suspension medium is analyzed. By comparing the experimentally measured DEP crossover frequencies of neurons and glial cells with the simulated values, new, refined neuron and glial dielectric and physical properties are predicted that better reflect the DEP experimental results. Finally, the design and development of the electronic circuit system and LabView control interface is introduced. We conclude with potential improvements and future work to make this MEA system a more precise, efficient, reliable and versatile BioMEMS platform for neural engineering research.