The classic viscous-liquid failure is a short fill, and the usual explanation is that the liquid did not have time to flow into the tip before the plunger stopped. That is true, but it is only half the story. The other half is that the space the liquid did not fill often did not stay empty. It filled with air. A viscous liquid dragged upward too fast leaves a partial vacuum behind it that the surrounding air rushes to fill, and the bubble that forms rides in the tip pretending to be volume you drew. It is worth separating this from surface foam, the froth you get from whipping a detergent or a protein solution, because the fix is different: surface foam is about how you dispense and agitate, while the trouble here is air pulled into the column on the way in.
Why thick liquids swallow air
Air entrainment during aspiration is a direct consequence of viscosity and speed. When the plunger rises, it lowers the pressure above the liquid, and the liquid climbs to follow. A thin liquid keeps up and fills the tip as a continuous column. A thick liquid cannot climb fast enough, so for the moment the plunger is moving there is a pressure deficit that something has to relieve, and the easiest thing to move is air, which slips up the tip wall past the slow liquid and collects as a bubble above or within the column. The faster you aspirate, the larger the deficit and the more air you pull in. The narrower the tip, the harder the liquid has to work to climb, and the more readily air takes the shortcut instead.
Over-immersing the tip makes it worse in a subtler way. Push the tip deep and the hydrostatic pressure changes what the liquid does at the orifice; lift it during aspiration and you can gulp air directly. The tip position during the draw is part of the problem, not a detail.
How a bubble lies to you
A bubble in the tip is expensive because it corrupts more than one transfer. On the current dispense it is the short fill you expected: the bubble occupies volume the plunger counted as liquid, so you deliver less than the target, and because the bubble size varies with speed and wetting, the shortfall is imprecise as well as systematic. But it also poisons what comes next. A bubble that clears on dispense sprays the last of the liquid unpredictably. A bubble that sits at the orifice throws off liquid-level detection on the next aspiration, so the channel misjudges where the surface is and either plunges too deep or stops in air. One entrained bubble can therefore turn into a run of wrong transfers, each caused by the last, which is why air in a viscous column is worth designing against rather than tolerating.
The settings that keep air out
Everything that prevents entrained air comes back to giving the liquid time and not asking it to climb faster than it can.
- Aspiration flow rate: lower it hard. This is the single largest lever, because the pressure deficit that pulls air in scales with how fast you try to move the liquid. Slow the draw until the column comes up solid.
- Post-aspiration settling delay: hold the tip in the liquid for a second or more after the plunger stops, so a small bubble has a chance to rise and merge or so the column can finish equalizing before the tip lifts.
- Submerged, stable tip position: keep the orifice below the surface throughout the draw and do not let it lift into air mid-aspiration. Follow the liquid level down if the instrument does it reliably at this viscosity, and keep the immersion shallow but never zero.
- Tip bore: a wide-bore tip lowers the pressure the liquid needs to climb, so it entrains less air at a given speed. For thick liquids this is a genuine fix, not a convenience.
- Pre-wet the tip: a dry tip wall is where the first bubble loves to form, so pre-wetting settles the first transfer, which is otherwise the worst offender.
Reverse pipetting helps here too, because the over-aspirated reserve gives any small bubble somewhere to sit above the target volume rather than inside it, and because the gentler dispense does not spit a bubble out under pressure.
Confirm with mass, and look before you trust
You cannot always see a bubble through a tip, and you certainly cannot see the run of downstream errors it causes, so the check is the usual one: weigh dispensed volumes and watch the precision, not just the average. A liquid entraining air shows up as a low mean and, tellingly, a wide spread, because the bubble size is not constant. When you can, watch a few aspirations directly during setup; a channel that comes up with a clear column at your chosen speed is telling you the air problem is solved, and a channel that repeatedly shows a gap at the top is telling you to slow down further before you trust a single number.
A short fill in a thick liquid is often not missing liquid, it is present air. Slow the draw, keep the tip under and still, and give the column time, because a bubble does not just cost this transfer, it costs the next one too.