The Electrode Design Space

The electrode design space is the set of parameters a thin-film electrode is specified by — site count, shank count, shank length, site span, site pitch, shank separation, site area, substrate thickness and site layout — and the range each takes across the 216 silicon probe and 53 thin-film grid designs in the catalog. This reference lists each parameter, its catalog range, what it decides for a recording, and where it appears in a probe name.

The design space is large because the thin-film process fixes geometry on the wafer: anything that can be drawn can be made, reproducibly. That is what lets a probe be designed against a target rather than chosen from what assembly happens to allow, and it is why the parameters below are worth knowing as a vocabulary. The interactive design space plots every catalog design on these axes; this page is the definitions behind it. Each parameter has its own glossary entry; the derived parameters the design-space page also publishes are listed at the end.

How to read a probe name#

A NeuroNexus probe name encodes the design space in order. A1x32-5mm-100-177 reads: A-Series, 1 shank × 32 sites, 5 mm shank, 100 µm site pitch, 177 µm² site area — a single-shank linear probe. A4x8-tet-5mm-100-200-177 reads: A-Series, 4 shanks × 8 groups, tetrode layout, 5 mm shank, 100 µm between groups, 200 µm shank separation, 177 µm² sites. A1x32-Poly2-5mm-50s-177 is a two-column polytrode with 50 µm staggered spacing. V1x32-Edge-10mm-100-177 is a Vector Array with an edge layout on a 10 mm silicon array. Buz32-sp-5mm is a 32-site Buzsaki design.

The fields, left to right: family letter (A, V, M, Q, E, Buz) · shanks × sites or groups per shank · layout token (empty for linear; Edge, tet, Poly2 to Poly5, Buz) · shank length · site pitch (an s suffix means staggered) · shank separation (multi-shank only) · site area. The leading count is shanks by groups, and what a group is depends on the layout token, so read the site count from the specification, never from the name: A2x2-tet has 4 sites on one design and 16 on another. Grid names follow their own pattern — E16-300-CL5-25 is a 16-site ECoG grid at 300 µm pitch with a 5 mm cable and 25 µm sites.

Site count#

Site count is the number of electrode sites — the channels the design records or stimulates on at once. Catalog silicon probes run from 4 to 1,024: 32 sites on 85 designs, 64 on 42, 16 on 37, 128 on 15, 256 on 11, and 1,024 on the two largest SiNAPS active-pixel designs. Grids run from 3 to 256, with 32 on 16 designs. Count sets coverage, not quality — more sites reach more neurons only when the layout places them in new tissue — and it fixes the package and headstage: 16 and 32 channels on one connector, 64 on one or two, 128 and above on several or on a high-density package.

Shank count#

Shank count is the number of penetrating shanks: 1 on 81 catalog designs, 2 on 21, 4 on 74, 8 on 22, up to 16. One shank samples a single track and reads as a depth profile; several sample laterally separated tracks in one insertion and read as a map. The cost is displacement — every shank is a track — which is why multi-shank designs are more often chosen for acute mapping than for the longest chronic implants. A grid has one shank by definition.

Shank length#

Shank length is tip to base, in millimetres, and caps how deep the sites can be placed. Catalog silicon designs run from 1.5 to 60 mm, with 5 mm on 79 designs and 10 mm on 50; 198 designs are 10 mm or shorter, which covers the rodent brain, and the longest belong to the Vector and rDBSA families for deep targets in large animals. Length is not span: a 10 mm shank with 32 sites at 100 µm has 3.1 mm of sites near the tip and 7 mm of bare shank behind them. A longer shank needs a thicker substrate or a support body to resist buckling.

Site span#

Site span is the distance from the first site to the last along a shank — the tissue one placement samples — and follows from site count and pitch: 16 sites at 50 µm cover 0.75 mm, 32 at 100 µm cover 3.1 mm. Catalog spans run from 20 µm (a single tetrode cluster) to about 10.5 mm (a Matrix 3D array), with 0.75 mm the median. Span is chosen against the structure: about 1 mm for mouse cortex, 2 mm for rat, 3 mm or more for a hippocampal profile through CA1 and the dentate gyrus. Its lateral counterpart, site span (width), is the distance between the outermost sites across the probe — up to 3.5 mm on catalog silicon designs.

Site pitch#

Site pitch is the centre-to-centre distance between neighbouring sites along a shank. Catalog linear pitches are 50, 100, 150 and 200 µm; tetrode and polytrode clusters space sites at about 20 to 25 µm; a few dense edge designs run at 20 µm. Pitch is chosen against the detection radius of about 100 µm: sites farther apart than that each see their own neurons, sites closer than that see the same neuron from several positions, which is what a spike sorter separates units on. The across-shank pitch between columns of a polytrode is 17 to 43 µm on catalog designs.

Shank separation#

Shank separation is the centre-to-centre distance between neighbouring shanks on a multi-shank design: 125 to 500 µm on most catalog designs, 200 µm on about half, and up to 1.8 mm on the Matrix platform's arrays. At 200 µm each shank samples its own cylinder of tissue about a detection radius wide; closer separation lets neighbouring shanks see some of the same neurons, wider places them in different structures. Separation times one less than the shank count is the width of the site span.

Site area#

Site area is the exposed surface of one site in square micrometres, and it is the parameter that couples geometry to electronics because it sets the site's impedance. Catalog silicon sites run from 92 to 1,250 µm²: 177 µm² (about 15 µm across) on 110 designs, 703 µm² (about 30 µm) on 41, 121 µm² on 15, with 160 and 165 µm² on the dense edge and Buzsaki designs. A 177 µm² site typically reads 200 kΩ to 1 MΩ at 1 kHz. Grid sites are quoted as diameters, 25 µm to 1 mm. Smaller sites isolate spikes more sharply at a higher noise floor; larger sites give a cleaner field potential and more charge capacity for stimulation.

Substrate thickness#

Substrate thickness is the thickness of the silicon or polyimide the array is built on: 15 or 50 µm for silicon, with 138 catalog designs offered at 15 and 154 at 50 and most at both; 12 µm for every polyimide grid. Thinner displaces less tissue and provokes less of the response that raises impedance over a chronic implant; thicker resists buckling on a long trajectory or through intact dura. It is chosen at order and is the one parameter in the design space that has no field in the probe name.

Site layout#

Probe geometry — the shanks and the spatial layout of the sites on them — is the qualitative parameter the others are arranged by. Linear, one evenly spaced column down the shank, is the default and the depth-profiling layout, on 108 catalog silicon designs. Edge, on 28, moves the column to the boundary of the substrate for an incremental gain in spike amplitude; Buzsaki, on 16, is an edge variant that staggers closely spaced sites at the tip for dense unit isolation. Tetrode, on 18, clusters four sites inside one detection radius; polytrode, on 38, runs two to five close columns the length of the shank. Multi-shank composes any of these across 2 to 16 shanks, and the Matrix platform stacks 2D arrays into a 3D volume at a slot spacing of 300 to 1,000 µm. Grids add their own layouts — regular ECoG and EEG grids, accordion and segmented cardiac arrays, pre-curved nerve cuffs.

Materials#

Materials are a design selection alongside geometry. Iridium sites on 209 of the 216 catalog silicon designs, with sputtered iridium oxide (SIROF) as a site option; platinum sites on the SiNAPS designs and on every grid; gold and platinum traces; silicon-carbide dielectrics; silicon or polyimide substrate. Z-Coat is a site coating applied at order that lowers impedance and raises charge capacity inside the same footprint. What each decides is set out in What Makes a Thin-Film Electrode Good.

Derived parameters#

The design-space page publishes five parameters computed from the site positions rather than specified in the design. The shallowest and deepest site are the positions of the first and last sites from the tip, which bracket the depth range a placement covers. The site footprint is the rectangle the sites span, width times depth, in square millimetres — the cross-section of tissue covered, not the area of one site. The neural interface volume is the volume enclosed by the sites in three dimensions, 0.0004 to 4.8 mm³ across the catalog. Site density is sites per square millimetre of footprint and brain coverage density is sites per cubic millimetre of interface volume — the second is the fairer comparison across shank counts, because a single shank's footprint is nearly a line.

Beyond the catalog#

The catalog is the part of the design space that is drawn already. The same process draws a new design against a target the catalog does not reach — geometry, materials and package are all customisable — and a custom design begins with the same parameters as a probe name: how many sites, on how many shanks, how long, at what pitch and area, in what layout, on what substrate. Use the design space to find the nearest catalog design, choose against the target with Choosing an Electrode for Your Experiment, and ask Oaks or request a quote with the parameters when the nearest design is not near enough. NeuroNexus products are for research use only.