Dull, matte or frosted tops, where surface roughness is scattering the light away

Gloss needs a surface smoother than about 0.1 micron, and coarse crystals destroy it. Cooling, wax choice and castor level, plus a reflow that rescues it.

A balm can leave the pan clear and glassy and set with a top like matt emulsion. Pour the same batch into two identical tins, cool one on a cold worktop and lid the other while it is still warm, and the difference in shine will be visible across a room. Gloss is a property of the top few tenths of a millimetre, decided in the last minutes of cooling, which is why it can usually be fixed without remelting anything.

• Smoothness for gloss: under about 0.1 micron
• Castor for shine: 10 to 25%
• Carnauba ceiling: 2%
• Reflow: 70 to 75 C, top 1 to 2 mm
• Lid below 25 C

Gloss is a surface property, not an ingredient

A glossy surface reflects light specularly, in one direction, so you see a sharp image of the window or the lamp in it. A matte surface scatters the same light diffusely into many directions and the image dissolves into a pale haze. Nothing has been lost from the formula in between. The balm is not less oily, no ingredient has migrated out, and the material a millimetre down is identical in both cases.

How much light is available to reflect at all is set by the refractive index step at the interface. For a plant oil at about n = 1.47 meeting air, the Fresnel equation at normal incidence gives ((1.47 minus 1) divided by (1.47 plus 1)) squared, which is 0.036. Only 3.6 percent of the light striking the surface bounces off it, and that 3.6 percent is your entire gloss budget. Everything else goes into the balm and comes back out scattered and dulled by whatever is suspended in it. Raising the refractive index to 1.49 with a high index ester lifts that figure to about 3.9 percent, a real gain of roughly a tenth, but a small one. Roughness decides whether you keep the 3.6 percent in one direction or lose it in all of them, and it can take the specular component to nearly zero.

The threshold has a name. The Rayleigh smoothness criterion says a surface behaves as a mirror while its root mean square roughness stays below about the wavelength divided by eight times the cosine of the incident angle. At 550 nm, the middle of the visible band, with light arriving near normal, that is about 0.07 micron, and in practice a surface reads glossy up to around 0.1 micron, satin from there to a few tenths, and flat beyond. Fat crystals are far larger than that. Microscopy of slowly crystallised fats routinely shows aggregates from a few microns to tens of microns across, so a surface carrying grown crystals sits one to two orders of magnitude past the criterion. That is the whole mechanism in one line: gloss fails when crystals at the free surface get big enough to see with light.

Instrument gloss is measured by ASTM D523 and ISO 2813 at fixed geometries of 20, 60 and 85 degrees, against a polished black glass primary standard of refractive index 1.567 defined as 100 gloss units. Nobody has published gloss unit values for handmade balm, so this page gives you the mechanism and the process levers rather than a number to hit. The wider optics of shine, tint and pearl are set out in lip colour and gloss physics.

Why the top goes dull when the inside is fine

A filled tin loses heat two ways: sideways and downward into the metal and whatever it is standing on, and upward from the open face by convection and radiation. Put a lid on and you close the second path and trap a pocket of hot air directly over the surface you care about. The top few tenths of a millimetre then become the last part of the balm to cross the crystallisation window, and they cross it slowly.

Slow crossing means few nuclei form, and the few that do grow large before the melt runs out, which is the same argument that governs texture everywhere in the product and is set out in crystallisation and cooling rate. At a free surface the consequence is geometric rather than tactile. Crystals inside the balm are hemmed in by their neighbours; crystals at the air interface have nothing above them and grow outward into the air, leaving facets, edges and gaps between aggregates. That texture is the roughness. The network below is doing exactly what balm texture science describes, and the same slow cooling that coarsens the surface also softens the set, so a dull top and a slightly slumped feel usually arrive together.

There is a second, slower route to the same appearance. Fat bloom, studied in detail in chocolate, is liquid fat migrating to a surface and recrystallising there in coarse, high melting form, turning a glossy face white and dull weeks after manufacture with nothing microbiologically wrong (Lonchampt and Hartel, 2004). Any warm and cool cycle in storage drives it, and the polymorphic transitions that make it irreversible are covered in fat crystal polymorphism. A balm that left the bench glossy and arrived matte after a week in a hot van is bloomed, not badly poured.

Work out which fault you have before you touch it

Half the batches described as matte are not gloss faults at all. The distinguishing question is whether the whole face is uniformly changed or whether there are discrete marks on an otherwise reflective surface. Uniform means surface crystals. Discrete means something has arrived or grown.

The levers that add gloss, in order of size

Process beats formula here, which is unusual for a balm defect and worth stating plainly: cooling rate at the surface moves gloss more than any single raw material. Once cooling is under control, these are the compositional levers, and they work by two different mechanisms. Some make the liquid film continuous and thick, some raise the refractive index, and only the first group is capable of a large effect.

Castor oil is the one that carries the load. Its ricinoleic acid hydroxyl group makes it viscous and polar enough to hold a thick, continuous, unbroken film where a thin oil breaks up into patches over lip texture. The general working band is 5 to 15 percent; when gloss is the point of the product rather than a nice-to-have, 10 to 25 percent is where shine becomes obvious, and tack sets the ceiling. Lanolin does something similar by sheer viscosity. The refractive index route, using high index esters or polyisobutene and gloss polymers, is real but arithmetically small on its own, as the Fresnel numbers above show. It earns its place in combination, by making the film that castor is already holding reflect a little harder. If shine is the point of the product rather than a rescue, build the formula around castor from the start instead of trying to add gloss to a finished stick.

The four things that dull a balm

Each of these adds scattering centres at or near the surface, and each has a threshold rather than a linear effect.

1. Carnauba above about 2 percent. At 1 to 2 percent carnauba improves gloss, because it crystallises fine and hard and raises the melting point of the surface layer. Above 2 to 3 percent the picture reverses: there is enough of a very high melting wax to crystallise early and coarsely while the rest of the melt is still fluid, and it also drags on application, which reads as dullness even when it is not.
2. Silica, starch and clay. These are matting agents by design. Their whole function is to sit at the surface as particles of a size that scatters, and even 1 percent will visibly flatten a shine. See starches and silica, and treat any of them in a formula that is meant to be glossy as an error rather than a variable to tune.
3. A heavy load of high stearic hard fat. Cocoa butter, shea stearin and free stearic acid all crystallise into large, high melting platelets that reach the surface. Stearic acid in particular is used deliberately in some products for exactly this effect. Above roughly 20 to 25 percent total hard butter in a lip formula, gloss becomes difficult regardless of how carefully you cool.
4. Pigment and pearl loading. Every solid particle you disperse is a scattering centre. Mica and pearl pigments give directional sparkle rather than specular gloss, and iron oxides at the levels a coloured stick needs will flatten a surface noticeably. This is why lipstick formulas carry so much castor and polybutene: they are buying back gloss the pigment took away.

Rescue the batch in front of you

A surface fault needs a surface fix. Remelting the whole batch to correct 0.5 mm of crystal habit is wasted heat, and it exposes the flavour, the essential oils and the polyunsaturated fraction to a second thermal cycle for no reason.

1. Bring the containers to room temperature and stand them level. A tin reflowed on a tilted surface sets as a wedge, and you will not notice until the lid goes on.
2. Reflow the top 1 to 2 mm at 70 to 75 C. A heat gun on its lowest setting, held 15 to 20 cm away and kept moving, is the usual tool. Watch for the moment the surface goes wet and mirror-like across its whole face and stop there. If you can see the melt front travelling downwards as a translucent band, you have gone too deep.
3. Cool the reflowed surface fast. Move the containers straight onto a cold metal tray or a cool worktop, uncovered, ideally with gentle air movement across them. Ten minutes in a fridge is fine. This is the step that decides whether you get gloss or a repeat of the fault.
4. Leave the lids off until the surface is below 25 C. Touch the outside of the tin near the rim rather than the balm.
5. Reflow tubes from above, not around. Only the exposed dome needs treating, and heating the barrel softens the column and can slump the whole stick. Wind the balm down slightly first if it stands proud.
6. Check one unit after two hours and again at 48 hours. Gloss assessed while the surface is still warm is a lie, and any bloom you failed to address will start to come back within days.

When you do need a full remelt, and whether that is safe

Reflow the surface when the fault is uniform, shallow and cosmetic. Remelt the batch when the dullness is a symptom of something structural: grain running all the way through, undissolved wax, layer separation, or a formula that is simply too hard-fat heavy for the finish you want. Remelting itself is safe for the material in the sense that nothing is chemically consumed, which is the same argument made at length on grainy shea butter. What it costs is volatiles and oxidative headroom, so add flavour, essential oils and tocopherol back at the end of the second melt rather than assuming the originals survived.

Two cases where a remelt makes things worse. A balm containing infused botanical material picks up more colour and more extracted matter on every heating, and it will darken. And a batch already smelling sharp or crayon-like is oxidising, in which case a remelt gives you glossy rancid balm; that one belongs on rancid balm, not here.

Pour hot, cool naked, lid cold

Three rules prevent nearly all of it, and they conflict with the instinct to tidy up as you go.

Pouring too cool is the commonest single cause, because a melt poured near its setting point is already carrying crystals and lays them straight onto the surface. The full windows and what each boundary costs are on pour temperatures,. For tubes, the pour and cooling sequence that gives a clean top face is set out in filling lip balm tubes.

Where gloss cannot be recovered

Some products will never be glossy and the honest answer is to change the target rather than the process. A balm above roughly 25 percent hard butter, a heavily pigmented stick without a gloss oil to compensate, and anything containing a matting filler are all matte by construction. Beeswax-dominant formulas set to a soft satin rather than a shine, which is a characteristic of the wax and not a fault. A product that has to survive 40 C in transit needs the high melting waxes that coarsen a surface, so heat resistance and gloss genuinely trade against each other.

The decision rule is short. If the dullness is uniform and the balm underneath feels right, reflow the surface and fix the cooling, because that is a five minute job with no formulation change. If the dullness returns within a fortnight in normal storage, it is bloom and the answer is compositional or logistical, not thermal. If the dullness came with a change in feel, hardness or smell, stop treating it as a gloss problem: the surface is telling you about something happening in the bulk, and reflowing it only hides the evidence.

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