The 96 deep well plate knowledge base
The problems that actually cost libraries and runs, written down: seals in the freezer, evaporation in tall wells, shaking discipline, and thawing without mixing up the archive.
A deep well plate's failures are slow ones: a seal that lifts in month four, a library that loses volume at the edges, a shaker run that shares wells. These are the failure modes that account for most spoiled deep well work, and the working rules that prevent them.
Freezer storage: the seal is the system
At -80 C the polypropylene is fine and stays fine for years; the archive lives or dies on its seal. Adhesives stiffen and shrink at freezer temperatures, and a film that was tight in October lifts at a corner by spring, after which the corner wells sublime away even in a freezer. Working rules: use silicone cap mats or freezer-rated seals for anything stored past a few weeks, press mats home with a roller or press rather than thumbs, leave headspace because aqueous contents expand, and spot-check archive plates by weight or by eye on a schedule. A library audit that finds volume loss is reading the seal's diary.
Evaporation and the edge in tall wells
Deep wells slow evaporation compared with shallow plates, because the vapor path is long, but the perimeter wells still lose fastest, and in storage the loss is concentration change in exactly the wells a screen reads first. Thirty-six of 96 wells are perimeter wells. For libraries, the moat strategies of culture work do not apply; the defenses are a real seal, consistent fill volumes, and freezer discipline (minimize door-open time and plate-out time, and let a plate equilibrate sealed before opening it, so room humidity condenses on the seal instead of in the wells).
Shaking deep wells without sharing them
Deep well plates are shaken for culture, extraction, and resuspension, and the failure is cross-talk: liquid thrown from well to well under a loose seal. Three controls decide it. Fill: past about half of brim, vigorous shaking starts exchanging contents, and pyramid bottoms throw liquid higher than round ones. Seal: a cap mat pressed home, never a film, which creeps under hours of orbital motion. Orbit: a small orbit at higher speed mixes deep wells better than a large orbit that sloshes, and square wells baffle liquid in ways that help aeration for cultures. Set the three once, write them on the protocol, and shaking becomes boring, which is the goal.
Thawing a library without hurting it
Frozen DMSO plates thaw fastest at room temperature on the bench, sealed, and slowest in a cold room where condensation management is harder. Keep the seal on until the plate reaches room temperature, spin it down briefly before opening so droplets on the seal rejoin their wells, and work from the thawed plate promptly: every freeze-thaw cycle is a small tax on compound integrity, which is why working aliquot plates exist. The mother plate's map, and the record of how many times it has thawed, are part of the library's value; the plate map tools on this site keep the first, and a line on the map template is the cheapest place to keep the second.
Related guides
Common questions
Why are my corner wells losing volume in the freezer?
Seal failure at the perimeter, almost always. Adhesive films lift at corners as adhesives stiffen at -80 C, and perimeter wells sublime through the gap. Switch the archive to silicone cap mats or freezer-rated heat seals, press them home mechanically, and check stored plates periodically by weight.
How full can I fill a deep well plate for shaking?
About half of brim for vigorous shaking, up to roughly 70 percent for gentle mixing, always under a cap mat pressed home on every well. Pyramid bottoms throw liquid higher than round bottoms at the same fill, so give square pyramid plates the conservative end of the range.
How many freeze-thaw cycles can a compound library take?
Fewer than it usually gets. Each cycle risks compound precipitation and water uptake by the DMSO, and the damage is compound specific, so there is no universal number. The standard defense is structural: keep a mother plate that rarely thaws and make working aliquot plates for routine screening, so cycles accumulate on the copies, not the archive.