Detergents are the liquids that misbehave in the opposite direction from the ones people warn you about. A viscous liquid is slow and heavy and reluctant to move, and you learn to coax it. A detergent solution is the reverse: it is eager, it goes everywhere, and its whole problem is that it will not stay where a well-behaved aqueous liquid would. If you have ever watched a wash buffer climb the outside of a tip, or turned a clean-looking dispense into a head of foam with one too many mix cycles, you have met the real enemy here. It is not viscosity. It is low surface tension.
This is worth separating cleanly from the viscous-liquids problem, because the settings that help pull in almost the opposite direction from what a thick liquid needs. Surfactants are common enough, in wash buffers, in bead protocols, in every ELISA plate on the deck, that treating them as ordinary aqueous liquids is a slow, quiet way to lose accuracy. They deserve a class of their own.
What low surface tension actually does
Surfactants exist to lower surface tension, and that single property drives every downstream headache. Water beads and pulls itself into compact drops because its surface wants to be small. Add a detergent and that restraint is gone. The liquid now wets almost any surface it touches, including the outside of your tip, and it spreads into a thin film rather than pulling back into a droplet. That film is the source of the most insidious error, because it drains slowly. When you aspirate, liquid clings to the tip exterior and interior as a film that keeps creeping and draining after the mechanism thinks it is done, and that lagging film ends up as over-delivery you never see on a single dispense but that shows up as drift across a plate.
Then there is foam. A low-surface-tension liquid entrains air readily, and once air is in there the bubbles are stable and slow to collapse, because the very thing that would normally pop them has been suppressed. The slightest agitation, an aggressive aspiration, a hard dispense from a height, one mix cycle too many, and you have a layer of foam. Foam is not a cosmetic annoyance. It wrecks volume accuracy because it occupies space that liquid should, and it wrecks level detection because it presents a false surface to any sensor looking for the top of the liquid.
The settings that tame them
The class for a detergent is built almost entirely around not entraining air and giving films time to settle. Almost every setting is a variation on slow down and stay gentle.
- Slow aspiration and especially slow dispense: speed is what entrains air and whips up foam, so the dispense in particular wants to be gentle, because that is where a fast jet turns into bubbles.
- Submerged or contact dispense below the surface: releasing the liquid under the existing surface, or against the wall just below it, avoids the splash and free fall that beat air into the sample.
- Minimal or no blowout: a blowout blast at the end of a dispense drives air straight through the liquid and is a reliable foam generator, so trim it hard or drop it entirely.
- No over-mixing: each mix cycle is another chance to entrain air, so use the fewest that do the job and resist the instinct to add a couple for safety.
- A drain delay: pausing at the end lets the wetting film drain and equalize instead of dribbling into the next well, which directly attacks the over-delivery problem.
- Pre-wetting the tip: cycling the liquid once before the real transfer equilibrates the film on the tip so the first dispense is not systematically different from the rest.
- Dispensing to the wall: running the liquid down the side of the well rather than dropping it into a pool keeps the delivery smooth and bubble-free.
None of these is exotic, and that is the point. The detergent class is mostly the aqueous class with the aggression removed and a couple of delays added, tuned so the liquid never gets the chance to foam or to leave a film behind on the fast path.
Level detection is the trap
The part that catches people out is not volume at all. It is liquid-level detection. Both capacitive and pressure-based sensing assume a clean, well-defined surface, and a detergent gives them anything but. A creeping film changes the capacitance the sensor sees before the tip has reached the real liquid, and a raft of foam gives a pressure sensor a surface that is not the surface. The instrument locks onto the wrong height, dips too shallow or too deep, and now your submerged dispense is not submerged, or your aspiration pulls air.
This is one of the rare cases where the clever feature is the liability. When level detection is being fooled by films and foam, a fixed-height strategy, where you tell the instrument exactly where the liquid is based on known geometry and volume, is often safer than trusting a sensor that a detergent is actively deceiving. It feels like a step backward. It is usually a step toward reproducibility.
Tips and the film
Tip choice matters more than for a plain aqueous liquid because the film behavior lives on the tip surface. Filtered tips guard against foam and film creeping up into the pipette body, which is a real risk when the liquid wants to climb. Beyond that, the honest advice is to hold tip type constant once you have tuned a detergent class, because the wetting film that drove your drain delay and your pre-wet is specific to that tip surface, and swapping to another wets and drains differently enough to move your volumes.
The failure modes to watch
When a detergent class is wrong, it fails in a small and recognizable set of ways, and knowing them tells you which setting to reach for.
- Foam: a visible head on the liquid, telling you the aspiration or dispense is too fast, the blowout is too aggressive, or you are mixing too much.
- Drifting volumes: doses that creep up or wander across a plate, the fingerprint of a wetting film that is not being given time to drain and equilibrate.
- Level-detection misfires: the tip diving too deep or stopping short, meaning a sensor has locked onto a film or a foam raft instead of the true surface.
Read backward, that list is a diagnostic. Foam points at speed and blowout, drift points at drain delay and pre-wetting, and a misfiring probe points at your detection strategy. The settings and the symptoms map onto each other cleanly, which is exactly why a detergent deserves its own deliberately tuned class rather than a borrowed aqueous one.
A detergent will do everything an aqueous liquid does, only faster and messier, so the whole art is subtraction: less speed, less blowout, less mixing, and a little patience while the films settle.