Freezer inventory and sample chain of custody

The four failures that fill a freezer

Unusable freezer contents almost always come from one of four causes:

• No position. The sample is listed, but not where it is — so finding it means opening boxes, which warms everything around it.
• No provenance. The vial is labelled with an initial and a date, and the person who made it left in 2021.
• No lineage. An aliquot exists but nothing connects it to the parent, so its history — passages, treatments, freeze–thaws — is unknown.
• No owner. Nobody is responsible for the box, so nobody is entitled to clear it.

Record the position, and make it a real position

"Freezer 2" is not a location. A workable inventory records the full path — the freezer, the shelf or rack, the box, and the slot within the box — because that is what lets someone retrieve a vial in under a minute with the door open once.

A visual box map is worth more here than it sounds. A grid you can look at answers "is this box full?", "where is the gap?" and "what is next to it?" instantly, and it makes a mis-filed vial visible rather than merely wrong in a list.

It also gives you an honest answer to a question that comes up in every sustainability assessment: how full is that freezer actually, and could two of them become one?

Lineage: what came from what

The single most valuable relationship in a sample inventory is parent-to-aliquot. It is also the one spreadsheets model worst, because a row cannot easily point at another row.

Recorded properly, lineage answers the questions that otherwise end an experiment: which stock did this come from, how many times has this line been through a freeze–thaw, what else was made from the same parent, and if the parent turns out to be contaminated or mis-identified, what else is affected. That last question is the one where a missing link becomes expensive.

Freeze–thaw counts, kept honestly

Freeze–thaw cycles degrade many sample types, and the count is the sort of number that is trivial to record at the time and impossible to reconstruct afterwards. Increment it when the vial is thawed, on the vial itself rather than on the group of vials, and let it be visible wherever the sample is.

The honest version of this is worth insisting on: if a count is unknown, it should read as unknown rather than as zero. A confident "0" on a vial that has been round the block three times is worse than a blank, because somebody will rely on it.

Chain of custody: who had it, and when

Chain of custody sounds like a forensic term, and in most research settings the requirement is milder — but the underlying record is the same. Every movement of a sample gets a line: who took it, when, why, and where it went. Transfers between people, shipments to a collaborator, and disposals all belong in the same append-only trail.

Two design rules make it trustworthy. The trail is appended to, never edited — a correction is a new entry — and a shipment manifest is fixed once it is sent, because a manifest that can be changed after dispatch is not evidence of anything.

If your work involves relevant human material, or samples covered by a specific consent or a material transfer agreement, the requirements around custody and retention are more demanding and are set by your governance framework rather than by convention. That is a conversation for whoever holds that responsibility in your institution — but the practical record they will ask you for is the same one described here.

Labels that survive the freezer

None of the above works if the label falls off at −80 °C or the handwriting frosts illegible. A machine-printed label with a code that can be scanned removes the two worst failure modes at once: it is readable after frosting, and scanning it means the position and identity are read rather than transcribed.

Print in sheets rather than one at a time. Labelling a freezer retrospectively is a job people abandon halfway if each label is a separate operation, and a half-labelled freezer is barely better than an unlabelled one.

Clearing a freezer you inherited

The pragmatic order for a drawer nobody will touch: photograph and map what is physically there before moving anything; record every box with an owner, even if the owner is "unknown"; give a deadline by which unclaimed boxes are disposed of, and publish it; and inventory what is kept properly on the way back in, so the exercise never has to happen twice.

The deadline is the part that works. Nothing else empties a freezer, because in the absence of a decision the safest individual choice is always to leave the box alone.