Surface treatment for culture devices: plasma, coatings, and how long each one lasts

A culture device is not finished when the geometry is right. The surface has to be wettable enough to fill, biocompatible enough to seed, and either adhesive or anti-adhesive depending on what you are growing — and the treatment that achieves that has a shelf life measured in hours to weeks, not years.

This is the part of a device programme that is invisible in the quote and decisive at the bench. A part that arrives with a treated surface and no statement of when it was treated is a part whose surface you cannot trust.

Native surfaces and where they start

The contact-angle bands worth knowing, since every treatment paper reports in them: below 10° is superhydrophilic, 10–90° hydrophilic, 90–150° hydrophobic, above 150° superhydrophobic. Untreated PDMS sits comfortably in the hydrophobic band. A filled microchannel needs it in the hydrophilic one.

Oxygen plasma: what it does and how fast it undoes itself

Plasma treatment is the default for PDMS because it does two jobs at once — it makes the surface hydrophilic and it enables irreversible bonding to glass or to a second PDMS layer.

The mechanism. Oxygen plasma introduces polar functional groups, principally silanol (Si–OH), through oxidation, and removes hydrocarbon groups by surface corrosion. The effect is large: one published microchannel study reduced the water contact angle from 120° to 17° with a 300-second treatment — hydrophobic to firmly hydrophilic in five minutes.

The problem is that it reverts. Plasma-treated PDMS exhibits recovery of hydrophobicity within minutes of exposure, and the effect is accelerated by bonding and heat treatment. The mechanism is understood: low-molecular-weight oligomers migrate from the bulk to the surface, and polar groups reorient inward. Nothing about the treatment is permanent, because the bulk material is a reservoir of untreated polymer sitting directly behind the treated skin.

Two published levers extend the window materially:

• Treat longer. At a constant 70 W plasma power, treatment longer than 5 minutes produced a PDMS surface that maintained its hydrophilicity for more than 6 hours, against the minutes typical of a brief treatment. The same study found that longer oxygen plasma exposure produced a smoother surface by AFM, so the longer treatment is not simply more damage.
• Store under water. Storing plasma-treated devices in deionised water maintained their hydrophilicity for weeks. This is the single highest-value, lowest-cost intervention in the whole area, and it is routinely omitted from device handover instructions.

The limit. Prolonged plasma treatment causes surface cracks that compromise device integrity. There is a genuine optimum, it is equipment-specific, and it has to be found on your own chamber — which is a real argument for having the fabricator do the treatment and state the parameters, rather than doing it in-house on a benchtop cleaner with unlogged settings.

UV-ozone: slower, deeper, no cracks

The published alternative behaves differently in a way that matters for some geometries. UV-ozone oxidises the chains at the surface and in more internal layers of the PDMS, and does so free of cracks, in contrast to the oxygen plasma approach. It has been used specifically for mass conversion of PDMS by deep penetration, and for complete oxidation of thick membranes.

The trade is time — UV-ozone is slow relative to plasma — and it does not give you the same bonding behaviour. Choose it when crack-free surfaces or through-thickness oxidation matters more than turnaround.

Bulk modification: mixing the fix into the polymer

If reversion is caused by untreated bulk migrating to the surface, one answer is to change the bulk. Two published approaches:

• Pluronic F127 added to the PDMS pre-polymer before curing. When the channel is filled with water, the incorporated F127 migrates to the water/PDMS interface to minimise surface energy; the PPO segments interact hydrophobically with the PDMS while the hydrophilic PEO segments extend outward. The reported contact angle fell from 99° to 63° after 24 hours of immersion, against 104° for native PDMS. Note that this is a hydrophilic surface generated by the presence of water, which is a fundamentally more stable arrangement than a plasma skin.
• Silwet L-77 bulk-mixed at increasing weight percent, which decreased contact angle monotonically with concentration. A related study with Triton X-100 reported that the contact angle after 7 days showed little difference, which the authors described as good durability.

The catch is that anything mixed into the bulk can leach out of it. A surfactant that migrates to the interface to lower surface energy is by definition mobile, and mobile additives end up in your medium. For a device intended for pharmacology or for any assay with a sensitive readout, treat bulk surfactant modification as a change requiring its own cytotoxicity and leachables check — not as a free improvement.

Coatings, and what each one is for

Surface treatment and coating are different steps solving different problems, and both usually happen.

The absorption point deserves emphasis because coatings are so often sold as the fix. Across eight polymers, absorption differed by more than 1000-fold with PDMS the most absorptive, and more than 99% of chlorpromazine partitioned into the PDMS bulk within 24 hours. A coating that halves absorption on a material that absorbs 99% of your compound leaves you absorbing 98%. If small molecules are central to the assay, the answer is a different material, not a better coating — the full argument is on the material selection guide.

The order of operations problem

Surface treatment, bonding and sterilisation interact, and the sequence is not free to choose:

1. Plasma bonding consumes the plasma treatment. The activated surface is what bonds. You cannot bond first and treat the interior of a sealed channel afterwards by plasma, because plasma is line-of-sight.
2. Treating an enclosed channel requires a flow-through method — chemical treatment, a solution-phase coating, or corona/UV where geometry permits. This is why treatment strategy has to be decided at design time, not after the part exists.
3. Sterilisation can undo the treatment. Autoclaving is a heat treatment, and heat accelerates hydrophobic recovery. If a device is plasma-treated and then autoclaved, assume the treatment is gone. See sterilising a custom culture device.
4. ECM coating is nearly always last and nearly always yours. It is applied in your laboratory, with your protein lot, shortly before seeding.

The practical consequence is that a device specification has to state the sequence, not just the steps. “Plasma treated, bonded, sterile” is three operations in an order that determines whether the surface you receive is usable.

What to specify, and what to ask on receipt

Put these into the RFQ. eNUVIO, as an example of a fabricator that documents this area, lists surface modification by chemical means, vapour deposition, UV exposure and plasma, alongside bonding to glass and other materials — so these are answerable questions, not unreasonable ones.

In the specification:

• Which surfaces are treated — all wetted surfaces, or only some, and if patterned, where.
• Treatment method and its parameters (for plasma: power, time, gas).
• Where treatment sits in the sequence relative to bonding and sterilisation.
• Required wettability at the point of use, stated as a contact angle or as an unambiguous functional test (“channel fills by capillary action without air entrapment”).
• Storage condition and the expected effective window under it.
• Whether any coating is applied by the vendor, what it is, and how it is bonded.
• Whether the part carries any bulk additive — surfactant, plasticiser, mould release — and whether leachables data exists.

On receipt:

• The date of treatment, not the date of shipping. For a plasma-treated part these can differ by weeks and the difference is the whole shelf life.
• Storage medium in transit. Dry, or under water? This determines whether the “weeks” figure or the “hours” figure applies.
• A functional acceptance test you can run in ten minutes — typically a fill test with dyed medium, checked for air entrapment. Run it on arrival, before the cells exist.

The uncomfortable summary

Surface treatment is the least durable thing about a culture device and the least documented. The geometry you commissioned will still be correct in two years; eNUVIO publishes a two-year shelf life on the OMEGA-MP stored at room temperature away from UV, which tells you the part is stable. The surface state is a different quantity with a different clock, and a vendor who states a shelf life for the part without stating one for the surface has answered a different question from the one you asked.

Ask both. Then store the parts the way the treatment requires, which for plasma-treated PDMS means under water, and plan the experiment around the shorter of the two clocks.

How we can help

We do not treat or coat devices. What we do is make sure the specification we send to fabricators contains the surface questions, so the quotes that come back are comparable on this axis rather than silent about it — which is the default. If you have a device that fills badly, seeds unevenly, or worked for the first batch and not the second, the surface is the first place to look, and we can put that question to the shops that have the process capability to answer it. Send us the details.