Microelectrode arrays (MEAs) are widely used to study the behavior of cells noninvasively and in real time. While the design of MEAs focuses mainly on the electrode material or its application-dependent modification, the passivation layer, which is crucial to define the electrode area and to insulate the conducting paths, remains largely unnoticed. Because often most cells are in direct contact with the passivation layer rather than the electrode material, biocompatible photoresists such as SU-8 are almost exclusively used. However, SU-8 is not without limitations in terms of optical transmission, optimal cell support, or compatibility within polymer-based microfluidic lab on chip systems. Here, we established a silicon nitride (SiN) passivation by physical vapor deposition (PVD), which was optimized and evaluated for impedance spectroscopy-based monitoring of cells. Surface characteristics, biocompatibility, and electrical insulation capability were investigated and compared to SU8 in detail. To investigate the influence of the SiN passivation on the impedimetric analysis of cells, HEK-293 A and MCF-7 were chosen as adherent cell models and measured on microelectrodes of 50–200 μm in diameter. The results clearly revealed an overall suitability of SiN as alternative passivation. While for the smallest electrode size a cell line dependent comparable or slightly decreased cell signal could be observed in comparison with SU-8, a significant higher cell signal was observed for microelectrodes larger than 50 μm in diameter. Furthermore, a high suitability for the bonding of PEGDA and PDMS microfluidic structures on the SiN passivation layer without any leakage could be demonstrated.

Miniaturization and integration of chemical reactions into fluidic systems in combination with product purification or buffer exchange can reduce the amount of solvents and reactants required while increasing synthesis efficiency. A critical step is the regulation of flow rates to realize optimal synthesis conditions and high purification rates, so real-time, label-free monitoring is required in methods such as free-flow electrophoresis. Optical detection methods are widely used, but they often have complex excitation and detection setups that are disadvantageous for point-of-care applications. The method we have chosen is electrochemical impedance spectroscopy for detecting charged compounds in aqueous buffers with low ionic strength. Propranolol was selected for proof of concept and was separated from the organic solvent and the precursor oxirane by free-flow electrophoresis. For this purpose, electrode structures were fabricated in microfluidic channels by photolithographic lift-off technique and optimized in terms of positioning, electrode size and distance for sensitive detection, and quantification of propranolol in the nanomolar range. It is also noteworthy that the organic solvent dimethyl sulfoxide (DMSO) could be detected and quantified by an increased impedance magnitude. Subsequently, the optimized interdigital electrode structures were integrated into the outlet channels of the electrophoretic separation chamber to monitor the various outgoing fluidic streams and provide inline control of the fluidic flows for the purification step. In conclusion, we can provide a microfluidic chip to monitor the separation efficiency of a substance mixture during free-flow electrophoresis without the need of complex analytical techniques using electrochemical impedance spectroscopy.