Organoid Chip Specifications & MEA Interfaces

Biophysics of the Bio-Electronic Boundary

The electronic-to-biological boundary is governed by electrochemical kinetics. When a metal electrode contacts a physiological electrolyte, it forms an electrical double layer. This boundary acts as a capacitor, storing charge and restricting the direct transfer of electrons. To record extracellular action potentials (which are typically small voltage deflections on the order of 10 to 100 microvolts), we must minimize the interface impedance.

We achieve this by utilizing Platinum Black or Titanium Nitride (TiN) coatings. These materials possess a highly porous, fractal micro-texture that increases the effective surface area by several orders of magnitude without increasing the geometric footprint. This fractal structure drops the electrode impedance at the key neural frequency of 1 kHz from several megaohms down to less than 50 kilohms.

To model this boundary, we apply the Randles equivalent circuit model, representing the interface as a solution resistance (R_s) in series with a parallel combination of double-layer capacitance (C_dl) and charge-transfer resistance (R_ct). During stimulation, we enforce charge-balanced biphasic current pulses. This prevents the accumulation of direct current (DC) offsets that would drive water electrolysis, generating toxic pH shifts and destroying the surrounding neural tissue.

High-Density Recording & Stimulation Electronics

Our hardware platform architecture routes 1,024 independent recording channels and 64 stimulation channels through a low-noise analog front-end (AFE). Each recording channel is equipped with a low-noise amplifier (LNA) featuring a programmable bandpass filter (typically configured from 100 Hz to 3 kHz for action potential spikes, and 1 Hz to 100 Hz for local field potentials).

Analog signals are digitized at the chip boundary using high-resolution 16-bit analog-to-digital converters (ADCs) operating at a sampling rate of 25 kHz per channel. This generates a aggregate data throughput of approximately 400 Megabits per second per chip. This raw digital stream is packaged into SPI frames and transmitted to a local Field Programmable Gate Array (FPGA) for real-time spike detection and closed-loop stimulation feedback.

Microfluidic and Environmental Engineering

An organoid is a living, breathing biological structure. Without continuous perfusion, waste products accumulate and oxygen gradients collapse, leading to core necrosis within hours. Our microfluidic chips are engineered to deliver a steady, laminar flow of oxygenated nutrient media.

We utilize Cyclic Olefin Copolymer (COC) as the structural substrate for our commercial chips. Unlike Polydimethylsiloxane (PDMS), which is notorious for absorbing small-molecule drugs and organic compounds, COC is chemically inert and has low water-absorption profiles. However, since COC is gas-impermeable, we integrate a thin, gas-permeable silicone membrane over the central cultivation chamber to allow direct diffusion of oxygen and carbon dioxide.

Environmental control is maintained using integrated sensors. Indium Tin Oxide (ITO) micro-heaters are patterned directly onto the glass substrate beneath the MEA. Because ITO is optically transparent, it allows simultaneous high-resolution brightfield imaging. A PID feedback loop reads from integrated platinum RTD sensors to stabilize the chamber temperature at 37.0°C with an accuracy of 0.1°C, mitigating thermal noise and preventing cellular stress.

Real-Time closed-Loop Telemetry

The primary computation loop is executed off-chip on a high-speed FPGA board. When a neural spike occurs, the extracellular potential drops, triggering our online threshold detector. The system performs real-time template matching to sort spikes from individual neurons, assigning events to specific channels within 100 microseconds of occurrence.

For biocomputing applications, latency is critical. To train a neural network using reinforcement learning (such as the classical Pong simulation), the stimulation feedback must occur within the biological plasticity window (under 10 milliseconds). Our hardware bypasses PC operating system kernels, transmitting stimulation vectors directly from the FPGA processing core back to the analog stimulator ports, closing the loop with a round-trip latency of just 1.2 milliseconds.

This seamless integration of electrochemical biophysics, low-noise analog engineering, microfluidics, and real-time FPGA computing represents the state-of-the-art in biocomputing infrastructure. By transforming living tissue into a reliable, interactive processing unit, we open the door to a new era of computational biology.