5 min read
Updated September 9, 2025
A NeuroNexus silicon probe is a microfabricated electrode array. Site positions are defined by photolithography, not by hand assembly. The geometry is the same from probe to probe and from animal to animal. That reproducibility is the reason to use one.
Research use only. NeuroNexus products and the procedures described here are for non-clinical research use, and must not be used in human or veterinary medical procedures.
What fabrication buys you#
Three properties follow from making the probe on a wafer:
- Site positions are known. Depth, spacing and shank geometry are design parameters, so a laminar profile means the same thing in every animal in the cohort.
- The geometry is repeatable across batches. Between-animal variability comes from the surgery and the biology, not from the electrode.
- Sites can be dense. Adding sites does not add shank width the way adding wires adds bundle diameter.
What that spans, measured across the catalog:
| Range in the catalog | |
|---|---|
| Base designs | 213 |
| Recording sites | 4 to 1,024 (4, 16, 32, 64, 128, 256, 1,024 and intermediate counts) |
| Shanks | 1 to 16 — 82 single-shank designs, 130 multi-shank, 32 three-dimensional arrays |
| Shank length | 1.5 to 20 mm, in 16 standard lengths |
| High-density designs | 71 |
| Substrate | Silicon, on every design |
| Site metal | Iridium on 210 designs, platinum on 3 |
Two numbers in that table set expectations. 198 designs have shanks of 10 mm or less, and only thirteen reach further — the catalog is built around the rodent brain, so reaching past 10 mm means one of those thirteen or a custom geometry. And the 1,024-site designs are active CMOS (SiNAPS), not passive: above a few hundred sites the multiplexing moves onto the probe.
Choosing a site layout#
Layout is the decision that matters most, and it follows from what you intend to measure.
| Layout | Sites are | Choose it when | What it costs you |
|---|---|---|---|
| Linear | Evenly spaced along one shank | Laminar work — cortical depth profiles, hippocampal layers, current source density | Nothing structural; the risk is choosing spacing from the highest available density rather than from the layer thickness you need to resolve |
| Edge | At the shank edge rather than the face | Chronic preparations and small nuclei — less tissue displaced relative to the recording surface | Sites sample one side; orientation at insertion becomes part of the protocol |
| Tetrode | Clustered in fours | Unit isolation is the point — four views of one neuron let a sorter separate overlapping waveforms by amplitude ratio | Coverage. You get excellent isolation at a few discrete depths, not a profile |
| Polytrode | Dense two-dimensional arrangements on one shank | You need isolation and depth coverage from a single penetration | Channel count per millimetre of depth span |
| Multi-shank | Several shanks on one probe | Cross-regional and circuit-level work needing lateral as well as depth sampling | Shank pitch is fixed at fabrication and must match the anatomy — a rat column pitch puts two shanks in two columns in a mouse |
Acute and chronic#
The same design can be packaged for either, and the package differs more than the probe does.
Acute packages prioritize handling and repeated use within a session. Chronic packages prioritize mechanical anchoring, cable strain relief and connector durability over weeks to months.
Decide acute or chronic before selecting a design. Retrofitting a chronic implant onto an acute package is the most common reason a first chronic cohort fails.
Site coating#
Site impedance sets part of the noise floor and limits how much charge a site can deliver. Z-Coat™ lowers impedance on supported designs, which helps both recording noise and stimulation headroom. Coating support is a property of the design; check it when you choose the probe, not afterwards. See Z-Coat.
Packages, connectors and compatibility#
The package determines the connector, the cable, and therefore which headstage you can use. Probe, package, headstage and DAQ have to agree, and the constraint propagates in that order.
Package options, channel counts, shank lengths and materials are catalog data. Read them from the product catalog, and confirm the full chain before ordering.
Custom designs#
Where no catalog design matches the target, NeuroNexus designs to specification: site layout, shank count and length, thickness, materials and packaging. Custom work runs on the same fabrication process as the catalog, so the reproducibility argument above still holds.
Selecting a probe#
- Target — depth, structure size, and whether the approach is one shank or several.
- Measurement — single units, population activity, LFP, or stimulation as well.
- Duration — acute session, or chronic implant.
- Subject — species and the weight the animal can carry.
- System — the headstage and DAQ already on the rig.
Steps 1 and 2 select the layout. Steps 3 to 5 select the package.