Will the pack survive the product? Compatibility testing for balm containers

Fragrance attacking polystyrene, oil wicking into paperboard, stress cracking in polypropylene and liner failures, with a twelve week compatibility protocol to run.

A compatibility failure never shows up on the day you fill. It shows up at week six in a customer's bag: fine cracks around the base of a tube, a greasy halo on a paperboard sleeve, a cap that will not seal, an orange freckle under the balm in a tin. Choosing the pack is a materials decision, and this page gives the polymer ranking, the four mechanisms behind those failures, and a twelve week test small enough to run on a kitchen shelf.

• Default: PP or HDPE
• Avoid: PS and PLA
• Protocol: 12 weeks at 4, 25 and 40 C
• Fail line: above 2 percent weight loss
• EU 10/2011 is food contact, not cosmetic

The four things that actually go wrong

"Incompatible" is used loosely. There are four distinct mechanisms, they have different symptoms, and a test designed for one will miss the others.

1. Migration out of the pack, into the product. Plasticisers, slip agents, antioxidants, colourants and residual monomer leave the polymer and dissolve into the balm. Fat is a far more aggressive solvent for these additives than water is. The symptom is an off odour or a plastic taste in a lip product that was clean in the jug.
2. Sorption into the pack, out of the product. The reverse traffic. Polyolefins absorb terpenes and fragrance molecules into the wall, so the scent weakens with nothing visible happening. It is the main mechanism behind scent fading in balm, and why a fragrance correct at 0.8 percent smells thin after two months in a thick-walled tube.
3. Attack on the polymer itself. Swelling, softening, crazing, environmental stress cracking or outright dissolution. This is the failure that ends with product on a shelf.
4. Permeation and leakage. Volatiles pass through the wall or past the closure. Nothing looks wrong; the pack gets lighter and the product duller. Weighing is the only way to see it.

A balm loads all four harder than a lotion does, because the continuous phase is oil. Water-based products are protected by the fact that most polymer additives are not water soluble; a balm has no such protection, and if it carries essential oils it arrives with terpenes that are used industrially as solvents. Usage levels are on essential oils in balms.

Polymer ranking for balms

The useful ceiling column matters twice: once at filling and once in a delivery van. Filling windows by container are on pour temperatures, and summer transit temperatures are on shipping in hot weather. A PLA tube that survives a 60 C fill will still slump at 55 C on a doorstep in July.

Stress cracking, and why an empty tube looks fine

Environmental stress cracking is the failure people misdiagnose most often, because it needs three things at once and removing any one of them makes it disappear.

• One

  A susceptible polymer

  Semi-crystalline polyolefins and amorphous clear plastics both crack, by different routes. PS and PETG are the worst of the common choices; PP is the most resistant but is not immune.

• Two

  Tensile stress in the wall

  Residual stress frozen in at the moulding gate, a snap-fit cap, an interference fit around an elevator disc, or the hoop stress of a slightly overfilled pack.

• Three

  An active agent

  Limonene, linalool, pinene, menthol, or a surfactant. These do not attack the chemistry; they lower the surface energy at a microcrack and let it propagate under a stress the polymer would otherwise carry indefinitely.

That is why an unfilled tube lasts for years on a shelf, and why cracks appear in a pattern rather than at random: around the gate mark, along the parting line, at the thin base of the barrel, and around any moulded thread. In polypropylene the sign is fine white crazing rather than a clean split, typically four to ten weeks in with a limonene-rich fragrance at 0.5 to 1 percent. Citrus oils are largely limonene, often above 90 percent for cold pressed orange, so a nominally light citrus dose is a heavy solvent dose.

The standardised methods worth knowing exist and are named in supplier data sheets: ASTM D543 for general chemical resistance of plastics to reagents, ASTM D1693 for the bent-strip environmental stress cracking test on ethylene plastics, and ISO 22088 for stress cracking generally. You will not run these at small scale, but you can ask a moulder whether they have.

Paperboard tubes: wicking and the 60 C seam

Kraft board push-up tubes are the most requested plastic-free format and the most frequently returned. Two things fail.

Oil wicks. Paper is a fibre network with capillaries running through it, and liquid oil travels them by capillary action alone. An unlined board tube develops a translucent stain within days and a greasy exterior within weeks, destroying the print and lifting any label. The only fix is a barrier: a polyethylene inner layer, foil, or a separate greaseproof liner. If a supplier says the board is "coated", ask what with, because a clay or wax surface coating is a print surface rather than an oil barrier.

The seam softens. Wound board tubes are held by a spiral or butt seam adhesive, usually a hot melt or a PVA dispersion. Hot melts soften in the 60 to 80 C region and oil plasticises them further, so the practical ceiling for a filled board tube is below any data sheet figure. That is why paperboard sits at the bottom of the pour window at 60 to 65 C, and why a board tube left in a hot car unwinds at the seam while a PP tube merely goes soft. Fill cool, keep them out of heat, and say so on the label. The recyclability argument is on sustainable packaging.

Tinplate and aluminium: the seam corrodes first

Tinplate is low carbon steel with a thin tin coating, and the coating is what protects it. Corrosion therefore begins wherever the coating is thinnest or absent: the cut edge of a welded side seam, the inside of a rolled edge, and any scratch from the stamping line. What drives it in a balm is the combination named in the heading of every corrosion table: free fatty acids plus trace moisture. Neither alone does much. Together they form iron soaps, which appear as orange freckles under the product and a dark stain on the balm above them.

Free fatty acid content rises over shelf life as triglycerides hydrolyse, so a tin that was fine at three months can freckle at eighteen. Three controls follow: specify internally lacquered tins, usually epoxy phenolic or acrylic, rather than bare tinplate; keep introduced water out of the product, which is the subject of moisture control; and control storage humidity, because the outside rusts too. Diagnosis is on rust spots on balm tins.

Aluminium behaves differently. It is not attacked by neutral triglycerides but is attacked by free fatty acids and by some acidic actives, which pit and darken the metal and can grey the product. Lacquered aluminium is excellent, unlacquered aluminium is a gamble, and a supplier who cannot tell you whether their tins are lined is telling you something.

Liners and wads compared

The inverted samples in the protocol below exist to test exactly this component. A liner that seals under gravity will still weep when the product is molten and pressed against it, which is what produces leaking lip balm tubes after a warm delivery.

A twelve week compatibility protocol

This is the smallest test that answers the question honestly. It needs a shelf, a fridge, a warm cupboard or incubator, and a scale reading to 0.01 g, calibrated as on weighing and calibration.

1. Fill 18 units plus controls. Three units per condition per orientation: 4 C, 25 C and 40 C, each upright and inverted. Add three units of the same batch in glass jars at 25 C as the product control, so you can tell a product change from a pack interaction.
2. Fill and cap them exactly as production would. Same fill weight, same headspace, same cap torque, same cooling before capping. A test done with hand-tightened caps proves nothing about a torqued line.
3. Weigh every unit at time zero and record the individual mass against a unit number. Batch averages hide the one leaking unit, which is the one you need to find.
4. Hold 40 C within about 2 degrees. A cupboard above a boiler swinging between 30 and 50 C is a different experiment. If you cannot hold it, say what you actually ran.
5. Inspect and weigh at weeks 2, 4, 8 and 12. Return each unit to the same orientation afterwards.
6. Record eight things each time. Mass, crazing or cloudiness in the wall, distortion or panelling, cap removal effort, label lift, product on the outside, colour and odour against the glass control, and firmness by thumb pressure.
7. Read the result against the control, not against memory. A balm changes over twelve weeks on its own, and only the difference between pack and glass is a compatibility finding.

The 40 C arm is an accelerated condition, not a separate product. Under a Q10 of 2, twelve weeks at 40 C corresponds very roughly to eight months at 25 C, but that arithmetic breaks wherever a phase change or a glass transition sits between the two temperatures, which for PLA and for a soft balm is exactly the case. The assumptions and where they fail are set out on accelerated ageing, and the oxidation half of the same study belongs with shelf life testing.

Reading the weight loss number

Weight loss above 2 percent in twelve weeks means there is an evaporation route you have not closed. Either the closure is not sealing, the liner is wrong, or the wall itself is permeable. Work it through on a 4.5 g lip balm: 2 percent is 0.09 g. If the formula carries 1 percent flavour and fragrance, that is 0.045 g of volatiles in the whole unit, so a 0.09 g loss cannot be fragrance alone. Something structural is leaving, and the product will be measurably harder and duller by the time a customer opens it.

Two honest qualifications. First, the threshold is a working line, not a standard: it is chosen because 2 percent is comfortably above weighing noise and below the point where a formula's properties visibly shift. Second, it is sensitive to pack size, because loss scales with surface area while the denominator is mass. A 4.5 g tube has far more surface per gram than a 100 g jar, so an identical wall and closure will read as a higher percentage in the small pack. Compare like with like, and if you sell one formula in three sizes, set the threshold per size.

Weight gain is also a result. A unit that gets heavier at 40 C is taking up moisture, which points at a hygroscopic ingredient or a pack that breathes both ways. Softening or oil beading on the surface at the 40 C condition without weight change is usually the product rather than the pack, and belongs with balm sweating or balm melting in the heat.

Why food contact data does not settle a cosmetic case

Suppliers routinely answer a compatibility question with a food contact declaration. Regulation (EU) No 10/2011 covers plastic materials and articles intended to come into contact with food, in the EU, and Great Britain retains an equivalent regime. It sets an overall migration limit of 10 mg per dm² of contact surface, or 60 mg per kg of food, tested against specified simulants under specified conditions, with vegetable oil as simulant D2 for fatty foods and a standard long-term condition of 10 days at 40 C. The converter issues a Declaration of Compliance for the stated conditions of use.

That document is genuinely useful, and it does not answer your question, for five reasons.

• It measures the wrong direction. Migration testing asks what leaves the plastic. Your tube is cracking because of what the product does to the plastic, which 10/2011 does not assess at all.
• The simulant is not your product. Simulant D2 is refined vegetable oil. It contains no limonene, no linalool, no menthol and no fragrance load, and terpenes are far more aggressive to polymers than triglycerides are.
• The exposure route is different. Food contact limits are derived for ingestion. A cosmetic is assessed for dermal exposure, and for a lip product for incidental ingestion at a completely different daily amount.
• The duration is different. Ten days at 40 C stands in for room temperature storage of a food. A balm may sit in a pack for a thirty month period after use opening, plus shelf time before sale.
• It is silent on everything mechanical. Stress cracking, wicking, seam adhesive, closure torque retention, label adhesion and scent scalping are all outside its scope.

What does apply in the EU and the UK is Regulation (EC) No 1223/2009. Annex I Part A requires the Cosmetic Product Safety Report to include information on the packaging material, specifically its purity and its relevant characteristics, and Article 3 requires the finished product to be safe under reasonably foreseeable use. Your compatibility data is what the safety assessor uses to discharge that, as covered on safety assessment and CPSR. In the United States there is no premarket packaging approval for cosmetics, but section 601 of the Federal Food, Drug, and Cosmetic Act deems a cosmetic adulterated if its container is composed of a poisonous or deleterious substance that may render the contents injurious to health, and MoCRA did not change that.

The decision rule, and what nobody can decide for you

Default to polypropylene or lacquered metal, get an EPE or PE foam faced liner, and if the formula carries citrus or any terpene-rich fragrance treat polystyrene, PETG and PLA as unavailable rather than as options to test. That much you can decide from the table above without running anything.

Everything after that has to be measured, because the variables sit outside anyone's published data. Resin grade, additive package, regrind content, moulding conditions and residual stress all differ between suppliers and between lots of the same part number, and your fragrance is a mixture nobody else has tested. This site cannot tell you whether your fragrance at your dose in your supplier's tube will craze it at week seven, and neither can a data sheet. What it can tell you is that eighteen units, a fridge, a warm cupboard and four weighings will find out before your customers do. If you are still choosing the format rather than validating it, start from the packaging guide and the comparison in lip balm tube vs tin, then test the winner.

Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.