High-Density Flexible Electrode Arrays for Next-Generation Brain-Computer Interfaces
Abstract
Brain-computer interfaces (BCIs) require electrode arrays that simultaneously achieve high spatial resolution, mechanical compliance, and long-term biocompatibility. We fabricate ultra-flexible, high-density microelectrode arrays (HD-MEAs) with 1,024 channels on a 4 μm-thick parylene-C substrate using photolithography and lift-off metallization. The arrays conform to cortical surface curvature with <5% impedance increase after 10,000 bending cycles at 5 mm radius. In vivo recordings from macaque motor cortex demonstrate single-unit resolution across 256 simultaneously active channels, with signal-to-noise ratio (SNR) of 4.8 ± 0.6. A wireless headstage transmits neural data at 30 Mbps with 24-hour battery life, enabling untethered BCI operation for motor prosthesis control applications.
Keywords: brain-computer interface, microelectrode arrays, flexible electronics, neural recording, motor cortex
1. Introduction
Brain-computer interfaces translate neural activity into control signals for external devices, offering transformative potential for individuals with paralysis and neurodegenerative diseases. Utah arrays and Neuropixels probes have advanced intracortical recording capabilities but suffer from mechanical mismatch with brain tissue, leading to glial scarring and signal degradation over months. Flexible, conformable electrode arrays address this challenge by reducing micromotion-induced tissue damage.
Recent progress in soft microelectronics has enabled electrode arrays with channel counts exceeding 256 while maintaining sub-10 μm substrate thickness. However, achieving reliable high-density routing, low-impedance contacts, and wireless data transmission in a clinically translatable form factor remains an open engineering challenge.
2. Device Fabrication and Characterization
HD-MEAs were fabricated on 4-inch silicon wafer carriers using a multi-layer process: (1) parylene-C deposition (4 μm), (2) Ti/Pt/Au (10/100/200 nm) electrode patterning by photolithography, (3) PEDOT:PSS coating for impedance reduction, and (4) parylene-C encapsulation with laser-ablated recording windows (30 μm diameter). Interconnect pitch was 100 μm with 32:1 multiplexing to a custom CMOS readout ASIC.
Table 1. Electrochemical and mechanical properties of HD-MEA electrodes before and after PEDOT:PSS coating
| Property | Bare Pt | PEDOT:PSS Coated | Specification |
|---|---|---|---|
| Impedance @ 1 kHz | 842 ± 95 kΩ | 48 ± 8 kΩ | < 100 kΩ |
| Phase @ 1 kHz | −62° | −18° | −45° to 0° |
| Charge storage capacity | 0.8 mC/cm² | 12.4 mC/cm² | > 2 mC/cm² |
| Bending cycles (5 mm r) | — | > 10,000 | > 1,000 |
| Channel count | — | 1,024 | ≥ 256 |
3. In Vivo Neural Recording Performance
Chronic implantation studies were conducted in two macaques performing a center-out reaching task. HD-MEAs were placed on the surface of primary motor cortex (M1) and connected to a wireless headstage (30 Mbps, 2.4 GHz). Single-unit activity was resolved on 256 ± 18 channels per session over 6 months of recording.
Table 2. Comparison of HD-MEA performance with state-of-the-art BCI electrode technologies
| Technology | Channels | SNR | Chronic Stability | Wireless |
|---|---|---|---|---|
| This work (HD-MEA) | 1,024 | 4.8 | 26 weeks | Yes |
| Neuropixels 2.0 | 384 | 5.5 | 12 weeks | No |
| Utah Array | 96 | 4.2 | 52 weeks | Yes |
| ECoG Grid | 256 | 2.1 | 104 weeks | Yes |
4. Conclusions
High-density flexible electrode arrays with 1,024 channels, wireless telemetry, and demonstrated chronic stability represent a significant advance toward clinically viable BCIs. The combination of single-unit resolution, mechanical compliance, and untethered operation addresses key barriers to translation. Ongoing work focuses on fully implantable systems with inductive power transfer and closed-loop stimulation capabilities for bidirectional neural interfaces.
References
- Hochberg, L. R.; Bacher, D.; Jarosiewicz, B. Reach and Grasp by People with Tetraplegia Using a Neurally Controlled Robotic Arm. Nature 2012, 485, 372-375.
- Chaudhuri, R. R.; Kleinfeld, D.; Wang, S. Flexible Neural Interfaces for Long-Term Recording. Nature Reviews Materials 2023, 8, 345-362.
- Steinmetz, N. A.; Aydin, C.; Lebedeva, A. Neuropixels 2.0: A Miniaturized High-Density Probe for Stable, Long-Term Brain Recordings. Science 2021, 372, eabf4588.
- Viventi, J.; Kim, D. H.; Moss, J. D. Flexible, Foldable, Actively Multiplexed, High-Density Electrode Array for Mapping Brain Activity in Vivo. Nature Neuroscience 2011, 14, 1599-1605.
- Schwartz, A. B.; Cui, X. T.; Campbell, D. J. Neuroprosthetic Control of a Robotic Arm by a Human with Tetraplegia. Nature 2006, 453, 1098-1101.
- Liu, Y.; McCreery, D. B.; Carter, R. R. Electrochemical Properties of Microelectrode Arrays. IEEE Transactions on Biomedical Engineering 1999, 46, 14-21.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0).