Sandy grit that comes from carnauba, candelilla or beeswax rather than from butter

Hard waxes crystallise as needles a lip can feel, especially above 3 percent carnauba. Telling wax grit from butter grain, and the blend that fixes it.

Not all grit is shea. A balm made with no butter at all can still come off the tin as fine sand, and when it does, the cause is the wax network itself rather than a segregated triglyceride. The two faults look identical on the lip and need opposite corrections, so the first job is to tell them apart, and there is a five minute test that does it.

• Carnauba ceiling 1-3%
• Candelilla ceiling 6-10%
• Growth window 55-40 C
• Felt above about 25 um

Three different faults, one sensation

A lip detects a particle at somewhere around 20 to 30 micrometres. That threshold comes from the confectionery industry, where chocolate is refined below about 30 micrometres precisely because the tongue starts to register particles above it, and lips are at least as sensitive as the palate. Any solid inclusion coarser than that will be felt, whatever it is made of. Three quite different things reach that size in a balm.

The distinction is not academic. Butter grain is corrected by a hotter, longer melt and a faster cool, as set out in grainy shea butter. Grown wax crystals are corrected by changing the wax blend or the cooling profile. Undissolved specks are corrected by melting properly in the first place, which is the separate problem of undissolved wax specks. Applying the shea remedy to a carnauba problem simply reproduces the fault.

The 45 C smear test

This takes five minutes and needs a glass microscope slide or the back of a stainless teaspoon, a probe thermometer and a water bath.

1. Take a representative sample. Dig a smear from the middle of the tin, not the surface. Surface material has cooled differently and is not the batch.
2. Spread it thin on the slide. A film of a few tenths of a millimetre, wide enough to rub with a fingertip.
3. Rub it cold and record what you feel. Note whether the grit is fine and even, or coarse and occasional. Even sandiness points at crystal growth; occasional hard points at undissolved material.
4. Warm the slide to 45 C and hold two minutes. Float it on a water bath and check with the probe. Do not guess: 50 C would start melting nothing useful, and 60 C would confuse the result.
5. Rub again. If the grit has gone, it was butter. If it is still there, it is wax, and it will still be there at 60 C too.
6. Look at what is left against the metal. Grown crystals disperse into a faint even haze under pressure. Undissolved flake fragments stay as discrete white points you can count.

Why hard waxes grow crystals a lip can feel

Waxes do not all crystallise into the same shapes. The habit a wax adopts depends on what it is made of, and the balm waxes differ more in composition than their similar-sounding melting points suggest. Work on plant wax oleogels, notably by Blake, Co and Marangoni, sorts them into morphology classes that track their oil binding behaviour, and the classes explain the texture faults directly.

Two things follow from that table. First, carnauba's mixture of aromatic esters, diesters and free alcohols is chemically diverse, and diverse mixtures crystallise into coarse, poorly matched networks. That is the same reason it is the hardest solid wax and the weakest oil structurant, a contradiction worked through in carnauba wax. Second, beeswax gets blamed constantly and deserves it rarely: its microcrystalline habit is why a plain beeswax lip balm almost never feels sandy unless a butter is misbehaving. If your gritty balm contains beeswax and nothing else hard, look at the butter first.

The network itself is a separate question from the individual crystal, and the relationship between the two is set out in oleogels and wax networks. A wax can build a strong network out of crystals too small to feel, which is what candelilla does, or a weak network out of crystals large enough to scratch, which is what carnauba does at the wrong level.

The levels at which it starts

Grit risk does not rise smoothly with wax percentage. It is negligible up to a threshold and then climbs quickly, because once the wax exceeds what the oil can hold in a fine dispersion, the excess has to go somewhere and it goes into larger crystals.

The last row is the one people miss. Two percent carnauba plus three percent rice bran wax is a five percent high melting stack, and it behaves like an overload even though neither material is individually above its own ceiling. Sum every wax melting above 70 C before you decide the formula is within limits. The candelilla figure also needs a word: the usage window on candelilla wax runs to 13 percent, and that is a window for firmness. The top of it is where texture complaints begin, so treat 10 percent as the sensory limit even though the structural limit is higher. Comparative hardness and gelling data for all of these sit in waxes compared, and the wax substitution calculator will recalculate a blend if you need to move weight between them.

Cooling: the window between 55 and 40 C

Two batches of identical composition can differ entirely in texture, and the difference is decided in the minutes after the pour. Crystals form by nucleation and then grow, and the two processes compete for the same material. Cool fast and you get a burst of nuclei, many crystals, all small. Cool slowly and you get few nuclei, each of which grows large on the material the others did not take. The general principle is set out in crystallisation and cooling rate; the specific temperatures for wax are different from the ones that matter for shea.

For balm waxes the growth window sits roughly between 55 and 40 C. Above 55 C most of the wax is still dissolved in the oil and nothing much is happening. Below 40 C the liquid is viscous enough that molecules no longer travel far, so crystals stop growing appreciably. In between, the wax is supersaturated and the oil is still fluid, which is exactly the condition in which a small number of nuclei can grow into long needles. A tin left on a warm worktop under a lid can take twenty minutes to cross that window. A tin cooled on a cold metal tray crosses it in three.

Pour temperature interacts with this. Pouring too hot means the container itself becomes the slow cooling vessel, and a thick walled tin holds heat for a long time. The windows in pour temperatures exist partly for this reason: they get the balm into the container at a temperature from which it can cool quickly rather than one that guarantees a leisurely descent.

Rescuing the batch on the bench

Wax grit is fully reversible, because nothing has degraded. The crystals are the same molecules that were always there, and dissolving them puts the formula back exactly where it started.

1. Combine the whole batch. Reprocessing tin by tin gives inconsistent results and doubles the work.
2. Melt to 85 C and hold 15 minutes. Time at temperature does the dissolving. A wax dissolves in oil below its own melting point given long enough, which is why a hold beats a brief peak.
3. Check for survivors before you cool. Draw a little onto a stainless spoon and look at it in good light at an angle. Glinting points mean carnauba is still undissolved, and the batch needs 88 to 90 C, or the split melt described on the carnauba page, where the wax goes into a quarter of the oil on its own first.
4. Cool fast, with agitation. Stir steadily from 80 C down to visible thickening, standing the vessel in cold water. Agitation multiplies nucleation sites, which is the whole mechanism for getting many small crystals instead of a few large ones.
5. Add heat sensitive materials during the cool. Essential oil, flavour and tocopherol below 50 to 60 C, after the hold.
6. Pour into the window and cool the containers, not just the pot. Stand filled tins on a cold metal tray, uncovered, until set hard.
7. If it comes back, cut the carnauba rather than repeating the process. Two failed rescues means the formula is over its ceiling, not that the technique was wrong.

The compositional route is often quicker than the thermal one. Dropping carnauba from 4 percent to 2 and returning the difference to beeswax or candelilla usually removes the fault outright, and the finished balm loses only a couple of degrees of heat tolerance. If the balm also scrapes rather than gliding, both faults have the same cause and the same fix, which is set out in drag and poor glide.

Designing the grit out

The prevention rule is a single sentence: use hard wax as a seasoning inside a base of a well behaved wax, never as the base itself. Carnauba at 1 to 2 percent inside 8 percent candelilla, or inside 12 percent beeswax, buys most of the heat resistance and none of the grit, because the base wax nucleates first and provides a dense population of small crystals for the carnauba to deposit onto rather than growing alone. The vegan combinations that work this way are covered in vegan beeswax substitutes and applied in the vegan lip balm formula.

Two smaller levers help. Buying carnauba as flakes or powder rather than as a block removes the commonest source of undissolved fragments, since hand broken lumps are difficult to dissolve and easy to lose track of in hot oil. And a genuinely polar oil in the blend, castor being the obvious one, improves the solubility of the more polar waxes and shortens the time needed at temperature; the practical detail is in melting methods. If you are checking whether a change has actually worked, a penetrometer reading of the sort described in measuring balm hardness will tell you what the reformulation cost you in firmness.

Why it comes back after a hot delivery van

A batch that left the bench smooth can arrive sandy, and this is the most common reason for a complaint about a formula that tested perfectly. Nothing melted and resolidified in the obvious sense. What happened is subtler and depends on solubility rather than melting.

The amount of wax an oil can hold in solution rises steeply with temperature. A parcel sitting at 45 to 55 C in a van or a warehouse redissolves a meaningful part of the wax network, even though every wax in it melts far higher. The crystals that survive act as seeds. Then the parcel cools slowly, over hours, through exactly the 55 to 40 C window described above, and the dissolved wax comes back out of solution onto those few surviving seeds instead of nucleating afresh. Growth without nucleation is the definition of a coarse crystal population, and repeated day and night cycling drives it further by Ostwald ripening, in which large crystals grow at the expense of small ones.

The consequences are practical. Judge a heat sensitive formula after a deliberate temperature challenge, not after a week on a cool shelf. Hold a retained sample at 30 C, and a second through five cycles between 20 and 40 C, then run the smear test on both. Shipping practice matters as much as formula here, and both are covered in shipping in hot weather and balm changed after shipping. A formula with a large heat tolerance margin and a low hard wax load survives transit better than a formula built to resist heat with carnauba, which is an irony worth absorbing.

When to reformulate instead of reprocessing

Grit is a defect of feel, not of safety. Nothing in a gritty balm has spoiled, the crystals are the same wax the formula always contained, and remelting restores the original material completely. That means the decision is commercial rather than technical, and it comes down to three questions.

Has the batch already been rescued once? If so, the formula is over its ceiling and reprocessing will only postpone the fault. Cut the carnauba instead. Does the product have to survive a summer supply chain? If it does, a formula that only tests smooth when cooled quickly on a cold tray is too fragile to sell, because the customer's storage will undo it. And is the batch worth the time? Reprocessing fifty tins is an afternoon; reprocessing a dozen four gram tubes costs more than the ingredients.

The honest limit on all of this is that no published work measures cosmetic wax graininess as such. The morphology classes, gelling concentrations and melting ranges are real and sourced, but the ceilings in the tables above are bench limits from formulating and wear testing rather than measurements, and the threshold at which a particular person notices grit varies. Test on the product's actual users, blind, after a heat challenge, and trust that over any percentage printed anywhere, including here.

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