Soft on the day, hard a month later, because the wax network is still crystallising
Wax and butter crystallisation carries on for weeks after the pour, stiffening the network. How to age test properly and formulate for the hardness you ship.
The complaint arrives in a predictable form. A balm was judged perfect on the bench, signed off, filled and labelled, and a month later a customer says it skates over the lips and deposits nothing. Nothing was contaminated and no ingredient changed. The wax network simply carried on assembling itself after the product looked finished, and the batch was assessed at the one moment it was guaranteed to be softest.
Firmness keeps climbing after a balm sets. Most of the rise happens in the first 24 to 72 hours, but crystallisation, polymorphic transition and sintering carry on for 7 to 28 days. Measured by penetration, a same day reading commonly sits 20 to 40 percent softer than the same batch at day 28. Formulate to the day 28 number, never the pour day one.
What is still happening a week after it set
A balm is firm enough to handle ten to thirty minutes after pouring. Three separate processes continue past that point, and only one of them is what anyone means by setting.
Continued crystallisation. Fat that is thermodynamically solid at room temperature but still dissolved in the liquid phase keeps coming out of solution for hours. It comes out slowly because the network that has already formed restricts diffusion, so the last few percent of solid fat arrives long after the surface stopped being tacky. Beeswax is the extreme case among common structurants because it is not one compound. Its monoester, hydrocarbon and free acid fractions crystallise across a broad range rather than at a point, which is what makes beeswax forgiving to pour and also what leaves material still coming out of solution days later.
Polymorphic transition. The crystal forms that appear first are the least stable and lowest melting, and they convert over hours to weeks into tighter, denser, higher melting packings. Shea stearin held at 20 C is largely beta prime after a single day and beta after about a week. Lauric butters set into the alpha form and move to beta prime within hours to days. Cocoa butter is slowest of all, drifting from form V to form VI over weeks. The mechanism is set out under fat crystal polymorphism, and the denser packings occupy less volume and build stiffer contacts, so the network gets harder as they arrive.
Sintering. Crystals touching one another fuse at their contact points over time, turning a loose assembly of particles into a rigid solid without any change in composition or solid fat content at all. This is the process most often missed, because it cannot be predicted from the recipe. Two balms with identical formulas and identical solid fat content can differ in stiffness purely by how well connected their networks have become, which is the core finding of the fat crystal network literature and the reason oleogels and wax networks treats connectivity as a separate variable from quantity.
A fourth process runs alongside and pulls the other way. Ostwald ripening coarsens the crystal population, big crystals growing at the expense of small ones, which reduces the number of contact points. Over long enough periods that can flatten or even reverse the firmness climb while making the product grainier and worse at holding oil. Post hardening is therefore not a straight line to infinity. It is a rise that flattens somewhere between week two and week four.
How long each structurant keeps moving
| Structurant | What continues after set | Practically stable at | Direction of travel |
|---|---|---|---|
| Beeswax | Broad crystallisation range, so material keeps joining the network, plus sintering | 7 to 14 days | Firmer, then steady |
| Candelilla, carnauba, rice bran wax | Fast, well ordered set with little left to convert | 3 to 7 days | Slightly firmer, then steady |
| Shea butter | Beta prime by day 1, beta by about day 7, coarsening thereafter | 21 to 28 days | Firmer, then grainier |
| Cocoa butter | Form V drifting to form VI | 28 days and beyond | Firmer, more brittle, bloom risk |
| Lauric butters and coconut | Alpha to beta prime, quick but pronounced | 7 to 10 days | Noticeably firmer |
| Palm oil and palm stearin | Diglycerides slow nucleation and prolong crystallisation | 7 to 14 days | Firmer, occasionally gritty |
| Hydrogenated castor wax | Network builds yield stress rather than bulk hardness | 3 to 7 days | Holds shape better, penetration barely moves |
The last row is the one that catches people out in the opposite direction. A castor wax balm can stop slumping and stop weeping over the first week while a penetrometer reads almost no change, so the product feels different without being harder in any measurable sense. Two rows in the middle carry a second consequence: palm stearin and the lauric butters both post-harden more than a beeswax and liquid oil balm does, so a formula containing either follows the slower band.
How big the climb is
Expressed as a change in penetration depth or in the force needed to push a probe in, the working band is a 20 to 40 percent rise between a same day check taken two to four hours after pouring and the same batch at day 28. Most of it, roughly 10 to 30 percent, lands in the first 24 to 72 hours, which is the figure quoted under crystallisation and cooling rate. The remainder accrues slowly over the following three to four weeks.
Be honest about what that band is. It is a working expectation assembled from the polymorph timelines above and from repeated workshop observation, not a measured constant, and it moves with the wax, the butter fraction, the cooling schedule and the storage temperature. A balm structured entirely by candelilla will sit near the bottom of it. A shea and cocoa butter balm poured slowly in a warm room will exceed the top of it.
The band also has to be read against a limit. Post hardening will not rescue a batch that was genuinely too soft on the day it was poured, because a 30 percent rise on a penetration reading is worth roughly one to two points of wax, not five. The ladder of what each point of wax actually buys is set out in balm too hard or too soft, and the honest reading of the two pages together is this: the bulk of a balm's hardness is decided by composition and fixed within a day, and post hardening is the last fifth of the journey. That last fifth is still enough to move a lip balm from good payoff to drag, which is why it is worth measuring rather than assuming.
Measure it instead of arguing about it
The diagnosis is a retained sample series. You need three units from the same batch, filled at the same time into the same container, stored undisturbed in the dark at 18 to 22 C, and a rig that gives the same reading twice.
The rig can be crude as long as it is consistent. A weighted probe test works: a rounded rod of fixed diameter, or the bowl of a teaspoon, loaded to a fixed mass, lowered onto the surface without a drop, held for a fixed dwell of five or ten seconds, and the depth read with a digital calliper. The published methods do the same thing with better hardware, and their conventions are worth borrowing: cone penetration of petrolatum under ASTM D937 and ISO 2137, and needle penetration of waxes under ASTM D1321, both reporting depth in tenths of a millimetre with lower meaning harder. More options, including texture analyser probes, are in measuring balm hardness.
Whatever rig you build, fix five things and write them on the record: probe shape, load, dwell time, sample temperature, and depth of product below the probe. Change any one of them and the numbers stop being comparable. Take three readings at different points on the surface and average them, and never re-probe a spot you have already broken.
| Check | Condition | Typical change vs same day | What it tells you |
|---|---|---|---|
| Same day, 2 to 4 hours | 20 C, undisturbed | Baseline | Almost nothing. Useful only as the zero for everything after it |
| Day 1 | 20 C, undisturbed | 10 to 30% firmer | Most of the climb is already in. First honest look at the formula |
| Day 7 | 20 C, undisturbed | 15 to 35% firmer | Butters through their first polymorph transition |
| Day 28 | 20 C, undisturbed | 20 to 40% firmer | The number to formulate to and to sign the batch off on |
| Day 28 plus 30 C challenge | 24 hours at 30 C, then 2 hours back at 20 C | Should return close to the day 28 value | Whether a warm delivery van resets or worsens the structure |
| Day 28 at 5 C | 2 hours at 5 C, read cold | Substantially firmer, reversible | What a customer meets in a coat pocket in January |
Cold storage, heat cycling and the 30 C bound
Cold storage is often described as an accelerator. Be precise about what it does, because two effects get merged.
Holding a sample at 5 C raises firmness immediately and substantially, because more of the fat is solid at that temperature. That is a real and useful reading, but it is reversible and it is a bound rather than a forecast: it tells you how the product behaves in a cold pocket, not how it will behave in a month. A sample taken out of a fridge and left two hours at 20 C reads close to where it would have been anyway, so a fridge is not a time machine for the network.
What genuinely accelerates the underlying transitions is temperature, but not at either extreme. Polymorphic conversion needs molecular mobility, so it runs fastest somewhat below the melting range of the fraction that is converting and stalls when the sample is very cold. Steady warm storage at 25 to 30 C therefore ages a shea or palm system faster than a fridge does. Cycling is faster still, since every warm and cool swing coarsens the crystal population, which is the mechanism behind balm changed after shipping.
The 30 C challenge is the opposite bound and it answers a different question: it is the softest realistic condition the product will meet, and post hardening cannot help you there. A stick that is only just firm enough at day 28 and 20 C will not hold a column at 30 C. Hold sample units 24 hours at 30 C, return them to 20 C for two hours, then read. A formula that survives both bounds, the cold pocket and the warm van, has a working window. The wider ageing framework, including why a 40 C oven can pass a formula that fails in a customer's drawer, is in accelerated ageing.
Rescue the batch in front of you
Remelting is safe here. There is no water phase to break, no emulsion to fail, and nothing has changed chemically in the fats, so a rescued batch is not a compromised batch. The real costs are three: heat sensitive materials get a second exposure, the batch takes another oxidation cycle, and the clock resets, so the rescued product needs another 28 days before you can judge it.
- Confirm the diagnosis first. Compare a retained sample against a freshly poured control from the same recipe. If the old one is measurably harder and both are smooth, it is post hardening. If the old one is gritty, you have a different fault.
- Reprocess the whole lot in one vessel. Scrape every tin and tube back together. Rescuing pot by pot guarantees units that do not match each other.
- Melt fully and hold. 70 to 80 C for 20 to 30 minutes with a probe thermometer in the fat. Melting only until it looks liquid, around 40 C, leaves high melting seed crystals intact and they will rebuild the same structure within days.
- Correct by removing wax, not by adding oil, if you can. Taking 1 to 2 points of wax out of the recipe and remaking is the clean fix. One point is perceptible, two points is a clear step, which is the right size of correction for a 20 to 40 percent firmness climb.
- If you must dilute the batch you have, do the arithmetic. Adding 5 g of liquid oil to 100 g of a 15 percent wax balm gives 15 divided by 105, or 14.3 percent, a drop of only 0.7 points. To take the same balm to 13 percent needs 15.4 g of oil and grows the batch by 15 percent. The wax ratio calculator will run the numbers for any starting point.
- Add heat sensitive materials at the end. Flavour, essential oils and tocopherol below 50 to 60 C, during the controlled cool.
- Cool fast and pour at the right temperature. Tins at 60 to 68 C, tubes at 65 to 72 C, as set out under pour temperatures, then get through the crystallisation window quickly rather than letting the filled units drift down overnight.
- Retain three units and start the schedule again. A rescued batch signed off on the day of the rescue is the original mistake repeated.
Do not correct a hardness complaint on a batch you have not aged. The most expensive version of this fault is a maker who removes two points of wax on day one, pours the replacement batch, and discovers at day 28 that the second batch is now too soft, has poor oil binding and starts to weep. One correction, then four weeks of waiting, then a decision.
Prevent it: formulate to the day 28 value
The whole of prevention is a single discipline. Decide the formula on aged samples, never on the day it was poured.
In practice that means aiming slightly soft. A balm that feels exactly right two hours after pouring will be too hard in a month. It should feel very slightly under-structured at day 1, with the correct feel arriving somewhere in the second or third week. For a beeswax and butter balm, building the recipe one to two points of wax below the number that felt right on the day is the usual allowance. Handle it more carefully for a tube than for a tin, because a tube must support a protruding column and has a narrower usable window.
- Same day reading sits 20 to 40 percent softer than day 28.
- 10 to 30 percent of that arrives in the first 24 to 72 hours.
- The formula allowance that corresponds to it: 1 to 2 points of wax.
- Sign-off point: day 28 at 20 C, on a retained sample in final packaging.
- Bounds to test: 2 hours at 5 C, and 24 hours at 30 C.
- Adding 5 percent liquid oil to a poured batch is worth about 0.7 points of wax.
Three supporting habits make the discipline cheap. Keep three retained units from every batch in final packaging, which costs almost nothing and doubles as the evidence behind your best-before date under shelf life testing. Record the cooling method and the wax lot number alongside the percentages in the batch record, because both move the result as much as a point of wax does. And pour a small control alongside any reformulation so that you always have a same-age comparison rather than a memory.
Where post hardening gets confused with something else
| Observation | Likely cause | Confirm by | What to do |
|---|---|---|---|
| Firmer over a month, still smooth and uniform | Post hardening | Penetration on retained samples at day 1 and day 28 | Reformulate to the day 28 value, 1 to 2 points less wax |
| Firmer and now gritty to the fingertip | Butter crystals transformed and coarsened | Rub a smear hard: grain smears out under warmth, leaving no particle | Follow grainy shea butter, not this page |
| Dull white or grey film on the surface, firmness unchanged | Bloom, a surface crystal layer | Warm the surface briefly: bloom clears, hardness does not change | Cosmetic. Fix cooling rate and storage temperature |
| Rock hard in the workshop in January, fine indoors | Ambient temperature, fully reversible | Equilibrate two hours at 20 C and read again | Nothing, or a seasonal formula two to three points lower |
| Harder and weeping oil at the same time | Coarse network with poor oil binding, not maturation | Hold at 30 C for 24 hours and inspect the surface | See balm sweating and fix the cooling schedule |
| Harder than the last batch of the same recipe, from day one | Different wax or butter lot | Check lot numbers, remake with the retained old lot if you kept it | Adjust by one point, and read batch to batch inconsistency |
| Firm and draggy from the day it was poured | High melt point wax, not elapsed time | Make a control with the same total wax as beeswax only | Reduce the hard wax, and see drag and poor glide |
| Tube stops advancing cleanly after a few weeks | Hardened column shearing against the tube wall | Warm the tube in the hand and try again | Soften the formula, and see lip balm will not twist up |
The decision rule, and what these numbers are not
The rule is short. Do not sign off a formula, print a best-before date, or correct a hardness complaint on a batch younger than 28 days, and hold a retained sample at 20 C so that there is always something to measure against. If a customer complaint arrives about a batch you assessed on the day it was poured, assume post hardening until a retained sample says otherwise.
Three limits on everything above. The 20 to 40 percent band is a working expectation compiled from polymorph timelines and workshop observation, not a controlled measurement of your formula, and a maker who needs it precisely must generate it with a consistent rig. Second, none of these processes is uniform across a container: the top few millimetres cool differently from the middle, so a probe reading from an opened tin and one from a sealed tube are not comparable even within the same batch. Third, the direction is not guaranteed to be upward forever. Coarsening eventually reduces the number of crystal contacts, so a very old balm can be simultaneously harder to the probe, worse at holding its oil and grainier to the finger. The connectivity argument underneath all of this is developed properly in balm texture science.
Frequently asked questions
Why is my lip balm harder than when I made it?
Because setting is not finishing. Fat that is still dissolved keeps crystallising for hours, the crystals that formed first convert to denser and higher melting forms over days to weeks, and crystals touching one another fuse at their contact points. Together these commonly raise firmness by 20 to 40 percent between a same day check and day 28, with most of it in the first three days.
How long does a balm take to reach its final hardness?
For a plain beeswax and liquid oil balm, roughly 7 to 14 days. Add shea and it stretches to 21 to 28 days, and cocoa butter can still be drifting past a month. Lauric butters such as murumuru and coconut move quickly but noticeably, over 7 to 10 days. When a formula contains several of these, plan for the slowest component and judge at day 28.
Is post hardening the same as curing?
No. Curing implies a chemical reaction completing, as in soap, and nothing of the kind happens in an anhydrous balm. What continues is physical: continued crystallisation, polymorphic transition and sintering of the crystal network. The distinction matters because it sets the limit. Post hardening is worth one to two points of wax, so it will not rescue a batch that was badly too soft on the day.
Can I remelt a balm that has gone too hard?
Yes, safely. There is no water phase to break and nothing has degraded, so remelting to 70 to 80 C for 20 to 30 minutes and correcting the ratio gives a sound product. Three costs apply: heat sensitive flavour and essential oils take a second exposure, the batch takes another oxidation cycle, and the ageing clock resets, so wait 28 days again before judging it.
Does putting balm in the fridge make it harden faster?
Not in the way people mean. Cold raises firmness immediately because more of the fat is solid, but that effect reverses when the sample warms up. The underlying polymorphic transitions actually need molecular mobility, so they run faster in steady warm storage around 25 to 30 C than in a fridge. A cold hold is a useful worst case reading, not an accelerator.
How much wax should I take out to allow for it?
One to two percentage points below the level that felt right on the pour day is the usual allowance for a beeswax and butter balm. One point is a perceptible change and two points is a clear step. Make the change, pour a small control alongside, and compare both at day 28 rather than deciding again on the day of the remake.
My balm is harder and grainy. Is that the same problem?
Related but not the same, and the fix differs. Post hardening is a uniform rise in firmness with no change in texture. Grain is a minority high melting fraction, usually shea, crystallising separately and coarsening into particles large enough to feel. Rub a smear hard: grain smears out under pressure and warmth. If it does, treat it as a graininess problem instead.
Sources and further reading
- Ray, J. and colleagues, Polymorphic behaviour of shea stearin, European Journal of Lipid Science and Technology, 115:1094, 2013.
- Narine, S. S. and Marangoni, A. G., Relating structure of fat crystal networks to mechanical properties, Food Research International, 32:227-248, 1999.
- Ghazani, S. M. and Marangoni, A. G., Molecular origins of polymorphism in cocoa butter, Annual Review of Food Science and Technology, 12:567, 2021.
- Doan, C. D. and colleagues, Internal and external factors affecting the crystallisation of wax oleogels, Journal of the American Oil Chemists' Society, 92:801-811, 2015.
- ASTM International, ASTM D937, Standard test method for cone penetration of petrolatum, West Conshohocken.
- ASTM International, ASTM D1321, Standard test method for needle penetration of petroleum waxes, West Conshohocken.
- International Organization for Standardization, ISO 2137, Petroleum products and lubricants: determination of cone penetration of lubricating greases and petrolatum, Geneva.
Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.