An Ultra-Low-Cost Electronic Device for Amperometric Neuronal Signal Acquisition using Micro-Graphitic Single-Crystal Diamond Multi-Electrode Arrays

In recent years, cell culture electronic interfaces have become increasingly valuable for studying neuronal behaviour. However, research in electrochemistry is constrained by the high cost of existing instrumentation. This work addresses these limitations, helping neuroscientists explore new ways to slow neuronal degeneration and develop new therapies for neurodegenerative diseases. This thesis aims to design an ultra-low-cost, low-noise, high-gain electronic device for acquiring neural signals, particularly amperometric ones. Amperometry, an electrochemical technique, applies a static potential to an electrode to drive a redox reaction in the target analyte, producing a current proportional to its concentration. This project is a collaboration with two departments at the University of Turin: Drug Science and Technology and Physics. The former is involved in the growth of in vitro cell cultures and the testing phase of our device, and the latter contributes significantly by producing a 60-channel micro-graphitic single-crystal diamond multi-electrode array (μG-SCD-MEA), inside which the cell culture is placed. These graphitic electrodes are embedded in a diamond matrix, which offers high biocompatibility, durability and chemical inertness - ideal properties for precise electrochemical measurements. The device comprises two PCBs: a Single-Channel board and a Motherboard. The channel board acquires and amplifies the amperometric signal from µG-SCD-MEAs by 5 billion times. The first stage, responsible for the current-to-voltage conversion, is highly critical due to the very low amplitude of the input, which is in the tens of picoamperes range. For this purpose, the precision amplifier LMP7721 was chosen, mainly for its ultra-low input bias current (3 fA) and low voltage noise (6.5 nV/√Hz). The signal is then filtered to capture only relevant components in the [100 Hz, 8 kHz] band, reducing high-frequency noise and power line interference. After that, the signal goes to the motherboard, which contains a potentiostat circuit, responsible for controlling the potential to drive the redox reaction within the cell culture, and an Event-Driven System, the main innovation of this work. The Event-Driven System, an ultra-low-cost solution compared to market alternatives, aims to detect neural events converting spikes into events. Managing events - represented by a bit set to 1 - reduces technical and computational costs contrary to handling the raw signal, sampled at 30 kHz to capture the rapid rise time (fractions of a second), despite the low average firing frequency of [0.1 Hz, 15 Hz]. More, the cost-effectiveness improves as the number of channels increases. Although shape information is lost, the system remains reliable as most applications require only event detection. Moreover, another firmware version enables continuous raw data acquisition for morphology analysis. The signal is transmitted to a PC over USB protocol and displayed in real-time using a Matlab Graphical User Interface (GUI). The device’s electrical behaviour was first evaluated using LTspice simulations. Then laboratory tests with precision bench instruments were carried out to confirm the positive results. All circuits met the expected outcomes, except the TIA, due to a malfunction of the femtoammeter. Firmware debugging verified correct signal acquisition and transmission. Future goals include cell culture testing, multi-channel motherboard design, and simultaneous detection of amperometric and potentiometric spikes.