Magnetic-bead cleanup is one of the most common jobs on a liquid handler and one of the least forgiving, because its whole logic runs against what a pipette wants to do. A pipette wants to remove liquid completely. A bead cleanup needs you to remove liquid while leaving something precious behind, and the something is invisible, packed against a wall, and easily lost. Almost every bad cleanup traces back to a single act done badly: aspirating the supernatant off an engaged pellet without disturbing it.
The chemistry is simple enough to state in a line. Paramagnetic beads, the SPRI kind used across nucleic acid cleanup, bind your target under the right buffer conditions, hold onto it while you wash the contaminants away, and release it when you change conditions to elute. Bind, wash, elute. The difficulty is not the chemistry. It is that every wash and every supernatant removal is a chance to aspirate the beads themselves, and once they are in the tip, your yield left with them.
Aspirating around an engaged pellet
When the magnet is engaged, the beads are pulled out of suspension and collected against the side or bottom of the well as a pellet. That pellet is where all your bound target lives, so the supernatant above it is waste you want gone and the pellet is everything you want kept. The tip has to clear the one without touching the other, and doing that reliably is a matter of a few settings working together.
- Offset from the pellet: position the tip on the opposite side of the well from where the magnet has drawn the beads, so aspiration pulls from clear liquid rather than from the pellet face.
- Slow aspiration: draw slowly so the flow into the tip does not set up currents strong enough to peel beads off the wall and carry them along.
- Downward tracking: lower the tip as the liquid level drops so it stays just under the surface, taking liquid from the top down instead of parking at the bottom where the beads are.
- A deliberate residual: stop short and leave a small volume behind rather than chasing the last drop, because the last drop is exactly where touching the pellet costs you beads.
That residual is not sloppiness, it is insurance. The few microliters you leave are cheap; the beads you would have caught reaching for them are not. A class tuned to leave a consistent small residual, well offset from the pellet, moving slowly and tracking down, is the difference between a cleanup that holds its yield and one that bleeds a little every wash.
The ethanol wash is its own problem
The wash steps are usually ethanol, and ethanol behaves nothing like the aqueous liquids around it. It is volatile, so it evaporates from an open well while the deck works through a plate, and it wets surfaces readily, so it creeps and drips in ways water does not. Both facts push the class toward air gaps and gentle handling: a trailing air gap below the liquid keeps ethanol from dripping across the plate as the head travels between wells, which otherwise seeds cross-contamination and costs you volume.
After the final wash comes the balance that decides a lot of cleanups: how dry to let the beads get. Leave ethanol behind and it carries into the eluate, where it inhibits the very downstream reactions the cleanup was meant to feed. Over-dry the pellet and it cracks and refuses to resuspend fully, so the target stays stuck to beads you cannot get back into solution. The dry step is a window, not a target, and both walls of it cost you.
Elution needs the opposite instinct
Elution is where the class has to change its mind. Everywhere else you have been avoiding the beads; now you need to hit them. Adding elution buffer and mixing to actively resuspend the pellet is what releases the target back into solution, and a timid mix that leaves half the pellet packed against the wall leaves half your yield there too. Mix enough to break the pellet up and get the beads moving, then let the magnet re-engage and pull them back down so you can take a clean eluate off the top.
Notice that this is the mirror image of supernatant removal, and it is why one generic transfer class cannot serve a bead protocol. The same well, the same beads, and the same tip need aggressive mixing at elution and delicate offset aspiration at every wash, and those are different classes doing different jobs at different steps.
How bead cleanups fail
The failure modes are few and they share a family resemblance: each one is quiet, showing up not as an error on the deck but as a disappointing number two steps later.
- Aspirating beads: a tip too close to the pellet or moving too fast pulls beads into the waste, and the yield loss is silent because nothing on the run reports it.
- Residual ethanol: an under-dried pellet or a wash left too full carries ethanol into the eluate, where it inhibits the downstream reaction and looks like a failed library or a bad amplification rather than a cleanup error.
- Incomplete resuspension: a weak elution mix leaves target bound to beads, so the eluate is clean but thin, and you never see the material that stayed on the wall.
The unifying point is that all of this is tied to the specific magnet module and plate you are running. Where the pellet forms, how far to offset, how deep to reach, how much residual to leave: these are properties of a geometry, not universal numbers, and a class validated on one magnet-plate pairing is a starting guess on another.
A bead cleanup never tells you when it went wrong. The beads leave with the waste, the ethanol rides into the eluate, and the only report you get is a yield that came back low.