# What Are BCI Electrodes Made Of, and Why Does It Matter?

The material an electrode is made from determines how well a [brain-computer interface](https://bciintel.com/glossary/brain-computer-interface) works over months and years — and right now, the field is splitting across at least three competing chemistries: silicon-based arrays, platinum-tipped carbon fiber, and graphene. A [BrainGate](https://bciintel.com/companies/braingate) study published in *Nature Medicine* (DOI: 10.1038/s41591-026-04414-6) this past June crystallized what is at stake: a man with [amyotrophic lateral sclerosis (ALS)](https://bciintel.com/glossary/als) used an intracortical BCI to communicate independently for close to two years and returned to full-time work. That result is the ceiling any electrode technology must support — stable, decodable signal across a multi-year implant horizon in a patient who cannot afford device failure.

The Utah array, the most clinically validated intracortical platform, packs up to one hundred metal tines onto a surface roughly half the size of a fingernail. It penetrates cortex like a pushpin, capturing action potentials near individual neurons. The problem, well-documented in the literature, is that silicon provokes a sustained immune response. Glial scarring encapsulates the tines over time, progressively attenuating signal and compressing the viable recording window. That biological reality is what is driving electrode materials research across academia and industry simultaneously.

---

## The Utah Array Is Still the Clinical Benchmark

The penetrating Utah array remains the reference platform for high-resolution intracortical recording precisely because it has the longest clinical track record. The BrainGate consortium's June 2026 *Nature Medicine* report — involving a man with ALS who communicated independently for nearly two years — is among the most compelling longitudinal demonstrations of what this architecture can sustain.

Principal investigator David Brandman described the functional outcome directly: "You have a man who's paralyzed. He can't move his arms and his legs. He can't be understood when he speaks. And using this technology, he's gone back to work full-time. Using this technology, he can have a conversation with a 6-year-old daughter."

That is a single-patient feasibility report, not a controlled trial, and the BCI field should resist extrapolating population-level efficacy from it. But as a proof-of-concept for multi-year intracortical decoding stability, it sets a meaningful benchmark — one that places implicit pressure on any emerging electrode material to demonstrate comparable longevity.

---

## Silicon's Biocompatibility Problem Is Driving Material Alternatives

The fundamental physics of electrode placement are not in dispute: closer proximity to target neurons yields stronger local field potential and single-unit signals, improving decoding accuracy. The clinical challenge is keeping electrodes close to neurons when the brain's immune response is actively trying to wall them off.

Cindy Chestek, a BCI researcher at the University of Michigan quoted in the *Chemical & Engineering News* feature, is developing carbon-fiber electrodes tipped with platinum as a potential solution. Carbon fiber's case rests on three material properties: it does not fracture like silicon, it does not deform plastically like metal, and Chestek characterizes it as the closest currently available approximation to true [biocompatibility](https://bciintel.com/glossary/biocompatibility). Whether that translates to meaningfully reduced glial scarring over multi-year implant windows in human subjects remains an open question that requires clinical data the field does not yet have at scale.

---

## Graphene: The Bidirectional Advantage

[INBRAIN Neuroelectronics](https://bciintel.com/companies/inbrain-neuroelectronics), a Barcelona-based startup, is making a distinct material argument for graphene — one that extends beyond passive recording into active stimulation. The company's electrodes are designed both to record cortical activity and to modulate it via intracortical microstimulation (ICMS), targeting conditions including Parkinson's disease.

INBRAIN's scientific rationale centers on what its team calls graphene's "charge injection limit" — the capacity to deliver electrical stimulation pulses without degrading the electrode material or triggering redox reactions that could damage surrounding tissue. For a closed-loop BCI that must both read neural signals and write stimulation back to the brain, this is a meaningful engineering constraint. Metal electrodes capable of stimulation can corrode or produce harmful electrochemical byproducts at therapeutic current densities. Graphene's electrochemical stability, if it holds up in chronic implant conditions, would represent a genuine material advantage for bidirectional applications.

The company's broader biocompatibility claim for graphene aligns with published preclinical data, though chronic human implant data from INBRAIN's platform is not yet publicly available at the scale needed to confirm clinical translation.

---

## The Industry Architecture Debate: Penetrating vs. Surface

While material science debates play out in the lab, the commercial BCI field is simultaneously running a parallel architectural argument between penetrating arrays and surface-contact approaches — a distinction that compounds the materials question.

[Precision Neuroscience](https://bciintel.com/companies/precision-neuroscience), co-founded by Ben Rapoport after he departed [Neuralink](https://bciintel.com/companies/neuralink), is developing surface [electrocorticography (ECoG)](https://bciintel.com/glossary/electrocorticography) electrodes that sit on the cortical surface without penetrating brain tissue. The design trades single-unit resolution for reduced procedural risk and avoidance of the penetration-induced scarring that afflicts arrays like the Utah. Whether surface electrodes can match the decoding bandwidth of intracortical arrays for demanding communication and motor restoration tasks — particularly in patients with rapidly progressing ALS — remains the central unresolved question for that platform.

[Synchron](https://bciintel.com/companies/synchron), by contrast, has taken a third path: endovascular deployment via the jugular vein into the superior sagittal sinus, positioning its Stentrode electrode array against the vessel wall overlying motor cortex without open-brain surgery. This approach avoids craniotomy entirely, but it accepts significant signal attenuation relative to intracortical recording by placing electrodes outside the dura.

These three architectural approaches — penetrating intracortical, ECoG surface, and endovascular — are not converging on a consensus. They are competing in parallel clinical programs with different risk-benefit profiles, regulatory pathways, and patient populations.

---

## What This Means for Clinical Translation and Patient Access

The electrode materials and architecture choices being made in 2026 will determine which BCI platforms reach commercial approval and at what scale within the next decade. Several vectors matter for anyone tracking clinical translation:

**Signal longevity vs. surgical risk tradeoff.** The BrainGate ALS case demonstrates that multi-year intracortical recording is achievable, but the Utah array's silicon-driven scarring trajectory remains a known degradation risk. Carbon fiber and graphene alternatives are still establishing their own chronic performance profiles in human subjects.

**Stimulation capability.** Platforms targeting closed-loop neurostimulation — for Parkinson's, epilepsy, or somatosensory feedback in motor prosthetics — face materially different electrode chemistry requirements than pure recording devices. Graphene's charge injection properties position INBRAIN specifically for this market segment.

**Regulatory pathway complexity.** Each new material introduced into an implantable device adds biocompatibility testing requirements under ISO 10993, extending the IDE and PMA timeline. Carbon fiber and graphene devices are earlier in this regulatory pipeline than silicon-based arrays with established safety records.

**Surgical access and patient volume.** Endovascular approaches like Synchron's could dramatically expand the eligible patient population by removing neurosurgical implantation as a prerequisite. If the decoding performance gap versus intracortical arrays narrows — aided by improved machine learning decoders — the procedural advantage could outweigh the signal disadvantage for a broad ALS population.

For researchers working at the intersection of BCI and neuroprosthetic limb control, where motor cortex decoding drives robotic systems, the electrode longevity question has direct implications for the durability of human-robot interfaces — a topic tracked closely at [humanoidintel.ai](https://humanoidintel.ai).

---

## Key Takeaways

- A BrainGate study published in *Nature Medicine* (DOI: 10.1038/s41591-026-04414-6) in June 2026 documented an ALS patient communicating independently with a BCI for nearly two years — the most compelling multi-year intracortical longevity demonstration in the public literature to date. This is a single-patient feasibility report, not a controlled trial.
- Silicon Utah arrays carry the deepest clinical validation record but provoke glial scarring that attenuates signal over time — the core problem driving alternative material research.
- Carbon-fiber electrodes tipped with platinum (University of Michigan, Chestek lab) are positioned as a high-biocompatibility alternative to silicon; human implant longevity data at scale is not yet available.
- INBRAIN Neuroelectronics is using graphene specifically for its charge injection capacity in closed-loop stimulation applications, targeting Parkinson's disease among other indications.
- Three competing BCI architectures — intracortical penetrating, ECoG surface (Precision Neuroscience), and endovascular (Synchron) — are advancing in parallel with distinct risk profiles and regulatory timelines, with no clinical consensus emerging yet.
- Material selection is a regulatory variable, not just a biological one: new electrode chemistries add ISO 10993 biocompatibility testing requirements that extend device approval timelines.

---

## Frequently Asked Questions

**What is the Utah array and why is it still used in BCIs?**
The Utah array is a silicon-based intracortical electrode array with up to one hundred penetrating tines packed onto a surface roughly half the size of a fingernail. It remains the most clinically validated platform for high-resolution intracortical recording because of its long track record in human implant studies, including multi-year BrainGate trials. Its main limitation is that silicon triggers an immune response that causes glial scarring around the electrodes over time, progressively degrading signal quality.

**Why does electrode material matter for brain-computer interfaces?**
The material an electrode is made from governs two critical variables: how close it can stay to target neurons over a multi-year implant horizon, and whether it can safely deliver electrical stimulation without degrading. Silicon causes scarring that pushes electrodes away from neurons. Alternatives like carbon fiber and graphene are being developed specifically to reduce that immune response and, in graphene's case, to handle stimulation currents without electrochemical breakdown.

**What is graphene's advantage over silicon in neural electrodes?**
Graphene's primary claimed advantage for neural stimulation applications is its high charge injection limit — the amount of electrical charge it can deliver per pulse without triggering degradation reactions at the electrode-tissue interface. This matters specifically for closed-loop BCIs that must both record brain signals and deliver therapeutic stimulation, such as those targeting Parkinson's disease. INBRAIN Neuroelectronics is the leading company commercializing graphene-based neural electrodes, though chronic human implant data at scale is not yet publicly available.

**How does Synchron's approach differ from Neuralink's or Precision Neuroscience's?**
Synchron deploys its Stentrode electrode array endovascularly — threading it through the jugular vein into a blood vessel overlying motor cortex — avoiding open-brain surgery entirely. Neuralink uses penetrating intracortical electrodes requiring a craniotomy and robotic insertion. Precision Neuroscience uses surface ECoG electrodes placed on the cortex without penetrating it, also requiring craniotomy but avoiding the tissue damage of penetrating arrays. Each approach trades signal resolution against surgical invasiveness differently.

**Can BCI electrode technology support long-term use in ALS patients?**
The BrainGate June 2026 *Nature Medicine* case report documents an ALS patient using an intracortical BCI to communicate independently for nearly two years — the strongest published evidence of multi-year usability in a rapidly progressing neurodegenerative disease. However, this represents a single-patient feasibility study. Broader evidence from controlled trials across larger patient populations is needed before population-level conclusions about longevity can be drawn. The material degradation challenge — glial scarring around silicon arrays — remains an active research problem that carbon fiber and graphene platforms are specifically trying to solve.