- What this coversWhere a cold chain label failure actually happened, and what to hand your printer so the next run does not repeat it.
- Who this is forBrand owners with frozen, refrigerated or freeze-thaw products who have had labels lift, wrinkle or fall off after distribution. Also brand owners specifying cold chain labels for the first time.
- Core findingA cold chain label fails in one of two places. On the line where it went on, or in the chain it had to survive afterwards. The fix for one is not the fix for the other.
- Bottom lineFind out where it failed, and what changed on your line, before you ask anyone for a different material.
The labels went on fine. Two weeks later the retailer calls to say they are lifting at the corners in the freezer case, or lying in the bottom of it.
Three things get reached for first. Ask the printer for a stronger adhesive. Assume the batch was bad. Say that nothing on the line has changed. Any of the three can turn out to be right, and any of the three can send you looking in the wrong place for a month.
The useful first move is narrower than choosing a material.
The two places a cold chain label fails
The first is the line, in the seconds and minutes around the applicator. The bond never came up to strength. The pack surface was colder than the adhesive was rated to be applied at. The surface was wet, and moisture can interfere with the adhesive reaching the pack. Or the pack went into the cold before the bond had built.
A label that failed here was failing from the moment it went on. It looked right because the applicator pressed it flat, and it held long enough to leave the building.
The second is everything after that. The bond formed properly and then lost the argument with what it met. A stiff facestock on a pack that contracts in the cold. A paper face taking up moisture through a thaw. Repeated freezing and thawing working at the edges of a label on a tight curve.
More than one cause can contribute inside either stage.
Timing helps, as long as you look early. Labels that lift or slide within hours, or that are already flagging at the corners on the staging floor before the pallet moves, point straight at the line. Labels that are still firmly down after a day and only start lifting after thaw cycles or handling point at what the pack met afterwards.
Late discovery proves nothing on its own. A bond that never formed can sit still inside a sealed case for a fortnight and only let go when someone unpacks the unit in a freezer aisle. Timing narrows it. It does not settle it.
If you are not sure which, or the failure does not look like either, the general method for reading a label failure after the run is in how to diagnose a label failure after the run, and it works for any job rather than just cold ones.
What happens at the applicator
Three things have to happen for a label to hold, and a pack can leave the line looking right when one of them did not.
THREE THINGS THAT HAVE TO HAPPEN
- ContactThe adhesive has to conform to the surface it meets, which is called wet-out. A smooth, clean, flat panel gives it the best chance. A ribbed tray, a recessed panel or a tight curve gives it less. The industry’s own training for converters puts it as the greater the curvature, the higher the tack needed to hold a label in position. How stiff the label is matters too, because a stiff face on a small diameter spends its whole life trying to spring flat again.
- GrabInitial tack is what holds a label in place on contact, before the bond has had time to build. Cold can reduce initial tack, especially in an adhesive not designed for cold application. That is what cold application grades are for, and they are a separate product rather than a variant of a standard one.
- TimeThe bond is not finished when the label looks applied. It goes on building afterwards, which the trade calls dwell, and how long it takes depends on the material.
Surface moisture works against the first of the three. A chilled pack in a warm room can pick up a film of water, and that film can interfere with the adhesive reaching the surface underneath. Some constructions are made for damp application and some are not, and the two are not the same purchase.
Any of the three can be short on its own, and more than one can be short at once.
The trap is that an applicator presses a label flat whatever is going on underneath it. At the end of the line, a bond that is finished and a bond that has barely started look the same.
What changed
Before you phone anyone, work out what is different.
A label that ran for a year and then started coming off is a reason to investigate what changed before assuming the material was always wrong.
A new container supplier, or a new mold or a new lot from the same one. A different cleaning agent, or a change to the washdown before the line. Line speed. Where the applicator sits relative to the chiller, and how long a pack now spends between the two. The season, because plant humidity moves with the weather and so does how much a cold pack sweats. And the roll itself, if this batch sat on a dock over a weekend or came off a truck in February.
None of that shows up on a purchase order. The artwork is the same. The adhesive code on the quote is the same. The supplier is the same. Everything on paper says nothing changed, and the thing the adhesive actually meets has changed twice.
It is the cheapest place to start. It goes before any conversation about material, because if something on the line moved, a different construction can mask the change rather than resolve it, and you will be paying more per thousand without knowing what you fixed.
Sometimes the conditions have moved beyond what the original material could do, and a different construction is the right answer. You cannot tell which until you know what moved.
Why nobody can give you the number
The next instinct, once you know there are two temperatures, is to go and find the right ones. There is no table that will give them to you, and the people who publish the figures say so themselves.
Look at the foot of an adhesive datasheet. The numbers are described as typical values and not a specification. Some say outright that they should not be used for specification purposes. That is not legal throat clearing. It is the manufacturer telling you the figure was measured under conditions that are not yours.
Three things make that true.
The figures were measured on something else. A sheet states the conditions behind its figures, and those conditions can include a reference facestock rather than the construction being sold. The numbers then describe a reference material rather than what is on your quote, and the sheet is where you check which.
A label is not an adhesive. The industry’s own training for converters puts it directly: peel adhesion and tack are properties of the whole label, not of the adhesive alone. Change the face and you change the figure.
And the number you need is not there. A datasheet prints minimum application temperature, service temperature range, loop tack and peel adhesion. How long a bond needs to build before the pack goes cold is not a field on the sheet.
So a sheet tells you what a material is rated for. It does not tell you what will happen on your line.
The construction was checked against the storage temperature and never against the pack or the line. Those are two different approvals, and only one of them happened.
What to tell your printer before they quote
A converter is supposed to ask before quoting. The industry’s own training for converters says the information should be obtained from the customer in advance and agreed between the two sides before the order goes ahead, and the list it gives covers the container and its shape, the conditions at the point of application, whether the surface is clean and dry, how the labels are applied, where they are stored beforehand and what happens to the pack immediately afterwards. If nobody asked you any of it, that is worth noticing. It is also easy to fix by sending it anyway.
Six things, and you can answer all six without knowing anything about adhesives.
WHAT TO SEND WITH THE BRIEF
- The containerWhat it is made of and what shape it is. A thermoformed PET clamshell, a polypropylene deli tub, a coated carton and a glass jar are four different surfaces to bond to, and ribs, recessed panels and tight curves all cut how much of the label is actually in contact.
- What the surface reads at the applicatorMeasured on the pack, not taken off the room thermostat. A pack coming out of a chiller is not the same temperature as the air around it.
- Whether it is dry, damp or frosted at that momentYou are the one standing in front of it, and it moves with the weather.
- How long between the label going on and the pack going coldMinutes or hours. It is not a field on the sheet.
- What the pack meets afterwardsStorage temperature, how many times it thaws and refreezes on the way to a shelf and how long the label has to stay readable.
- How the labels go on and where the roll livesHand or automatic, and whether the roll sits on a dock or in an unheated store before it reaches the machine.
That is your side of it. Which facestock, which adhesive and which liner is theirs. Handing a converter a material to quote instead of a job to solve is how a house default ends up on your pack.
Then one question worth more than the rest of the list. Have you run this before, on a line like mine, on this material. A converter who has will tell you what happened and what they changed afterwards. A converter who has not may still have run something close, and what to listen for is whether they say which part is new to them and how they would prove the construction before you commit a run.
What settles it
Three kinds of evidence get used to settle a question like this, and they answer three different questions. None of them answers the other two.
THREE KINDS OF EVIDENCE
- The datasheetTells you what a material is rated for. Minimum application temperature, service range, tack and peel. What it cannot tell you is how that material behaves on your pack. The figures were measured on something else, and there is no field for the time a bond needs before the pack goes cold.
- Your converterTells you how a material behaves on a setup like yours. A converter with relevant production experience may know what has worked on similar jobs and what came back. That is evidence you cannot generate on your own and it costs nothing to ask for.
- A trialTells you whether the whole thing works on your pack. Filled containers, your line, your surface temperature, through the real cycle rather than an afternoon in a freezer.
Of the three, the trial is the one to insist on, and the people who make the material point the same way. Look at the foot of the sheets again. One says all pressure sensitive materials should be tested thoroughly under end use conditions. Another says the customer is solely and exclusively responsible for testing. A third says performance should always be tested in the actual application conditions.
Those lines read like legal cover, but they are also the honest answer. Three manufacturers saying the sheet took the question as far as a sheet can take it, and the rest gets settled on the job itself.
Which makes the sequence simple enough. Find out what changed. Send your conditions. Ask what they have run. Then test what they propose on filled packs, before you commit a run.
BEFORE YOU BRIEF A SUPPLIER
- What you are solvingPutting the container, its surface temperature at application and the time before it goes cold into the brief, so the quote prices your job instead of a house default.
- Read nextWhat a label supplier brief should contain before you request a quote
WHEN LABELS HAVE ALREADY COME OFF
- What you are solvingWorking out whether the bond was short at application or gave way to what came after, because those point at different fixes.
- Read nextHow to diagnose a label failure after the run
FAQ
What is the difference between application temperature and service temperature?
They are two separate fields on the same sheet. Minimum application temperature is the coldest surface the adhesive is rated to be applied to. Service temperature range is the range the maker recommends once the label is on. A construction can be rated far below freezer temperature in service and still need a surface above freezing to bond to in the first place. A service rating for the freezer does not tell you the minimum temperature the label can be applied at.
Will a stronger adhesive fix it?
Sometimes. It does not change what sits between the adhesive and the pack, so a film of water is still a film of water, and it does not fix a stiff label fighting a tight curve or a pack that went cold before the bond had built. If the surface was wet or the dwell was short or something on the line changed, you may pay more per thousand for the same failure. Find out which before you ask for it.
My labels passed the supplier’s test and failed in distribution. Why?
Published figures are measured under stated test conditions, and those can include a reference facestock rather than the construction being sold. Peel and tack are properties of the whole label rather than the adhesive alone, so a figure measured on one face does not describe another. Ask which facestock and which surface the figures came from, then test the construction on filled packs through your own cycle.
ON THE SOURCES
Label constructions are specified per job and vary by manufacturer, facestock and region. Lines quoted here come from published manufacturer datasheets and from industry training material for label converters. The manufacturers describe their own published figures as typical values rather than a specification, and say the buyer should test under end use conditions. Confirm anything that matters against the sheet for the exact construction you are quoted. This article does not contain affiliate links.
