SLAS plate format: the dimensions your custom device must hit, and what breaks when it misses

A custom culture device that does not fit the instruments you already own is not a cheaper experiment. It is a second capital purchase you did not budget for. The commonest version of this mistake is a beautifully designed chip that images perfectly on a manual inverted scope and cannot be loaded into the high-content imager the laboratory bought to run the screen.

The dimensions that decide this are public, free and short. There are five ANSI/SLAS microplate standards, and between them they specify the footprint, the height, the bottom flange, the well positions and the well bottom elevation. A fabricator will build to whatever you specify. If you do not specify these, you will get something that is nearly right, and nearly right is the failure mode that only appears after tooling is cut.

The five numbers

Two details in that table are quietly load-bearing.

Tolerance is not uniform across the plate. SLAS 1 gives a tighter band near the corners (± 0.25 mm) than along the sides (± 0.5 mm), because the corners are what a gripper and a nest actually reference. A part that meets the loose figure everywhere and misses the tight figure at the corners is out of specification in the only place that matters for automation.

Height is measured against a stacking surface, not the bench. SLAS 2 allows a 0.76 mm projection above the top-stacking surface and requires 1 mm of clearance in the well area. This is the clause that custom devices break most often, because a designer adds a port, a Luer, a lid boss or a tubing barb, and the plate stops stacking, stops nesting in the hotel and stops passing under the imager’s plate cover.

Note also the citation rule SLAS publishes: because SLAS is a society rather than a standards body, the standards must be referenced as “ANSI/SLAS 1-2004: Microplates — Footprint Dimensions” and so on. If you write the requirement into an RFQ, write it in that form so the fabricator’s quality department can find it.

What conformance does not buy you

This is the part that costs money. A plate can meet all five standards and still fail in your laboratory, because the standards deliberately do not cover the following:

• Bottom thickness and material. High-NA objectives are corrected for a specific coverslip thickness. InSphero publishes 188 µm for the continuous COP membrane on the Akura Twin; MIMETAS publishes 150 µm microscope-grade glass on the OrganoPlate. Both are conformant plates with different optics. If your objective expects #1.5 (170 µm), neither is exactly it, and the spherical aberration that follows is a real loss of resolution, not a rounding error.
• Bottom flatness across the plate. Autofocus on a high-content imager tracks a surface. A plate that bows by tens of microns across 127 mm costs you a focus map per plate, or focus failures at the edges.
• Optical clarity and autofluorescence of the bottom. Covered on the material selection guide; thermally bonded polystyrene is the trap.
• Lid geometry and condensation. Not standardised. A lid that fits and seals differently from the plate the assay was developed on changes evaporation, which changes osmolality, which changes the assay.
• Skirt friction and gripper force. A conformant skirt in a slippery material can still be dropped by a gripper tuned on polystyrene.
• Barcode position and quiet zone. Every plate hotel and tracking system assumes one. Custom devices routinely ship without a flat area to put one on.
• Bottom clearance for transmitted light. A device with a thick base or an opaque manifold under the culture area is conformant and unimageable in transmitted light.

The practical consequence: conformance is necessary and not sufficient. Ask the fabricator to declare conformance to the five standards by name, and separately ask for the four numbers the standards do not cover — bottom thickness, bottom material, flatness across the plate, and total height including any port that protrudes.

The three device shapes, and what each one costs you

Custom culture hardware arrives in one of three form factors, and the choice determines your instrument compatibility before any biology is designed.

Note where the Sphericalplate 5D sits: it is a 24-well plate on a standard SBS footprint whose height is 20.15 mm, well above the SLAS 2 typical 14.35 mm. That is common and legitimate for deep-well and reservoir plates — but it means every downstream instrument has to be checked for z-clearance, and a plate sealer or a stacker tuned to 14.35 mm will not take it. Publishing the number, as Kugelmeiers does, is exactly the right behaviour; a custom device that does not publish its height is a compatibility question you cannot answer without a caliper.

Well bottom elevation, and why it is a separate standard

ANSI/SLAS 6 exists because “how high is the well bottom above the bench” is a different question from “how tall is the plate”, and both matter to a different instrument. Well bottom elevation sets the working distance an objective must reach and the z-position a dispense tip must find.

Corning publishes a well bottom elevation of 1.3 mm for the HTS Transwell-96 permeable support, which is the kind of number custom devices almost never state. If your device holds cells on a membrane suspended in a well, the elevation of that membrane — not the elevation of the well bottom — is what the objective must reach, and a long-working-distance objective is a purchase, not a setting.

For a device intended to be imaged, put this in the specification directly: the distance from the outside bottom surface of the device to the cell plane, and the material and thickness of everything in between. That single sentence resolves most objective-compatibility arguments before they happen.

The liquid-handling constraints nobody writes down

Automated liquid handling adds requirements the plate standards do not:

• Tip access geometry. A tip descends vertically. Any overhang, any lip that narrows the well opening, and any port that shares the well’s footprint can collide. InSphero publishes an explicit tip position for the Akura Twin — a 0.7 mm horizontal offset from well centre and a 9.86 mm z-distance to the plate top — precisely because the well narrows into an asymmetric cavity at the bottom. A custom device with a non-obvious tip path and no published tip position is a device your automation team has to reverse-engineer.
• Residual volume and aspiration geometry. Round-bottom and tapered wells hold a residual volume that a flat tip cannot reach. If the assay depends on a complete medium exchange, this is a specification, not a detail.
• Working volume versus maximum volume. These are different numbers and vendors that publish both are being helpful. Kugelmeiers publishes a maximum of 3 mL and a working range of 0.5–2 mL per well on the Sphericalplate 5D 24-well; InSphero publishes a 40–50 µL working volume on the Akura Twin. The gap between maximum and working is the evaporation and spill headroom, and a custom device quoted with only one of the two numbers is under-specified.

What to write in the RFQ

Copy this block into the specification. It costs nothing and removes the most expensive class of device failure.

1. Footprint: conform to ANSI/SLAS 1-2004, ANSI/SLAS 3-2004 and ANSI/SLAS 4-2004 for the stated well count, and state the achieved tolerance.
2. Height: state total height including any protruding feature, and declare whether the part conforms to ANSI/SLAS 2-2004 or deliberately exceeds it.
3. Bottom: material, thickness in µm, and flatness across the plate.
4. Cell plane: distance from the outside bottom face to the cell plane, and everything in the optical path between them.
5. Volumes: maximum and working volume per well or chamber, separately.
6. Tip path: the xy offset and z-depth at which a standard tip can safely aspirate to the bottom.
7. Stacking, sealing, barcode: whether plates stack, whether a standard sealer engages, and where a barcode label can be placed.
8. Lid: supplied or not, and its contribution to total height.

A fabricator who can answer all eight has understood the part. A fabricator who answers only the first is quoting on geometry alone, and geometry alone is what gets delivered.

When to break the standard deliberately

There are good reasons to leave the SLAS envelope, and they are all the same reason: the instrument you care about is not a plate instrument.

• MEA work is governed by the recording system’s plate geometry, not by SLAS. Electrode count and well count are set by the platform, and the MEA platform compatibility matrix is the relevant constraint set.
• Force-readout muscle devices are governed by the reader’s optical or magnetic pickup geometry.
• Anything destined for a bespoke rig should be designed around that rig and given an adapter for everything else.

If you do break the standard, break it knowingly and write down which instrument you have chosen to be incompatible with. The failure that hurts is the unintentional one discovered after tooling.

How we can help

We do not manufacture devices. What we do is send the same specification — including the eight fields above — to several fabricators, so the differences that come back are differences in capability rather than differences in what each shop assumed you meant. If you have a sketch, a well count and the make and model of the instrument the part has to work in, that is enough to start. Send it to us.