Glycerin, honey and hyaluronic acid in a product that contains no water

Glycerin is insoluble in oils and separates above about 3 percent. Hygroscopicity data, hyaluronic acid molecular weights, and the humectants that do work in a balm.

Glycerin is the best evidenced moisturising ingredient in cosmetics and one of the least useful things you can put in a wax and oil stick. Both statements are true, and the reason is physical rather than a matter of quality. This page covers what glycerin will and will not dissolve in, how much water each common humectant actually holds at a stated humidity, what it takes to carry one in an anhydrous base, and where the honest ceiling sits before a lip balm turns tacky.

• INCI: Glycerin
• Practically insoluble in fatty oils
• Beads above 2-3%
• Density 1.26 g/mL

What glycerin dissolves in, and what it does not

The pharmacopoeias settle this in one line each. The United States Pharmacopeia describes glycerin as miscible with water and with alcohol and insoluble in fixed and volatile oils. The European Pharmacopoeia monograph for glycerol gives the same picture in its own vocabulary: miscible with water and with ethanol, practically insoluble in fatty oils. Neither is hedging. Glycerol carries three hydroxyl groups on a three carbon backbone and hydrogen bonds to itself far more strongly than a triglyceride can compete with.

Nobody publishes a solubility of glycerol in sunflower oil in milligrams per gram, because the trade that cares about the number is biodiesel rather than cosmetics. In fatty acid ester systems the measured figures sit at roughly 0.06 to 0.2 percent by weight at ambient temperature, and solubility in the parent triglyceride is lower still. Take 0.1 percent as the working order of magnitude for a balm oil, rising a little with temperature and a little more with polarity. Castor oil is the outlier worth knowing: a hydroxyl group on around 90 percent of its fatty acid chains makes it the most polar common balm oil, and it tolerates more glycerin before separation than any other. Tolerates, not dissolves. The two are still not miscible.

That means every glycerin containing balm you have ever used is a suspension. The glycerin is present as droplets, and it stays put only because the wax crystal network has enough yield stress to stop them moving. The mechanics of that network, and the reason a solid looking balm still permits slow migration, are covered in oleogels and wax networks. The general case, which takes in salt, sugar, urea and hyaluronic acid as well, is solubility in anhydrous formulas.

The beading threshold, and what you actually see

Suspensions fail by coalescence. Two droplets meet, merge into a bigger one, and a bigger droplet is both more mobile and more visible. In practice a balm holds glycerin invisibly up to about 2 percent, becomes marginal between 2 and 3, and shows visible separation above 3. The threshold is not a constant. It moves up with castor oil content, with a genuine emulsifier, and with a fine crystal wax that builds a tighter network. It moves down with heat cycling, with a slow cool that grows coarse crystals, and with any pigment or clay that gives droplets a surface to gather on.

What you see, in the order people notice it: a slick, watery first swipe from a tube that felt normal last week; clear beads standing proud on the surface of a tin, sometimes mistaken for oil sweating; a wet meniscus at the rim where the stick meets the tube wall; and, in a pigmented balm, sticky patches that pick up dust. Glycerin has a density of about 1.26 g/mL against roughly 0.91 for molten balm, so in a warm jug the droplets sink. If the last tubes off a batch are stickier than the first, you have watched that happen. The look alike faults are worth ruling out before you blame the humectant: oil sweating from a wax network under strain is balm sweating, and a clean horizontal split is balm separated in layers.

How much water a humectant actually holds

Hygroscopicity is measurable, and the honest way to state it is grams of water taken up per 100 g of humectant at equilibrium with air of a stated humidity, at a stated temperature. For the liquid polyols the figures below are read off published water activity curves for their aqueous solutions, which is the same measurement approached from the other side: an 80 percent glycerol solution sits at a water activity near 0.5, so glycerol left in 50 percent humidity air ends up around 80 percent glycerol and 20 percent water.

Sorbitol is the instructive row. It is a crystalline solid with a critical relative humidity of roughly 74 to 78 percent at 25 C, form dependent. Below that it picks up almost nothing, because a dry crystal has no surface layer of solution to grow. Above it the crystal deliquesces and takes up water without any obvious limit. There is no gentle middle. A powder humectant in a lip balm therefore spends most of its life doing nothing at all, and the rest of it turning wet.

Now scale the glycerin numbers to a real product. A 4.25 g lip balm at 2 percent glycerin contains 85 mg of glycerin. Fully equilibrated at 50 percent humidity, that holds about 20 mg of water, roughly a fortieth of a millilitre spread across both lips, and only if atmospheric water can reach it through the occlusive film that the rest of the formula exists to create. The mechanism is real. The quantity is not the point of the product.

Hyaluronic acid, kilodalton by kilodalton

Hyaluronic acid is sold to makers as a family of grades distinguished by molecular weight, and the grades genuinely behave differently in water. None of that survives the move into an anhydrous base, but the differences are worth knowing so you can read a supplier sheet properly.

The right hand column is the entire finding. Hyaluronic acid works by hydrating: the chains uncoil in bulk water, entangle, and hold a hydrated network in place. A dry powder dispersed in oil and wax has nothing to uncoil into. It is a solid particle in a lipid, no different in kind from a grain of sand, and molecular weight changes nothing about that. Typical label levels of 0.01 to 0.1 percent put 0.4 to 4 mg of it in a lip tube. The consumer facing version of this argument is in ingredients for chapped lips.

What actually carries a humectant in an anhydrous base

There are four honest routes, and each costs something. You can emulsify, which means adding a water in oil emulsifier that wraps the polar droplets and keeps them dispersed. You can absorb, using a high surface area powder to hold the polar phase so it cannot pool. You can substitute an oil compatible derivative for the water soluble parent. Or you can stop pretending the product is anhydrous, build a proper water in oil emulsion, and preserve it.

Two of those rows deserve a note. The absorbent route works because the polar phase wets the powder rather than pooling; details of the materials are in starches and silica. The derivative route is the only one that keeps the formula genuinely water free, and it is why panthenol appears in anhydrous sticks more often than glycerin does. Urea is sometimes proposed as a humectant here and should not be: it is a crystalline solid in a lipid base, and its main topical role is keratolytic rather than humectant, which is a different job with a different dose.

When a humectant makes chapping worse

A humectant does not create water, it moves it. In humid air it takes water from the atmosphere. In dry air, roughly below 40 to 50 percent relative humidity, there is little atmospheric water available, so a humectant film on the surface equilibrates against the only reservoir it can reach, which is the skin underneath. Water is drawn from the deeper epidermis to the surface, where it evaporates. The three roles and how they interact are laid out in occlusive, emollient and humectant, and the outward flux they are all working against is transepidermal water loss.

The humidities that matter are lower than most people assume. A heated room in winter commonly runs 20 to 30 percent. An aircraft cabin at cruise is frequently 10 to 20 percent. Those are conditions in which a naked humectant layer has nothing to draw on but you, which is why the pairing rule exists: a humectant needs an occlusive above it, always. In a balm the humectant is at least embedded within an occlusive matrix, which mitigates the problem while it stays put. A glycerin bead that has migrated to the surface of a stick is a naked humectant film, sitting on the lip with nothing over it. That is the mechanism behind one of the several explanations examined in does lip balm dry your lips, and it is why the practical advice in cold weather skin puts the occlusive last in the routine rather than first.

Honey, aloe and the water they bring

Three popular humectant inputs are not dry, whatever the marketing implies. Honey is 15 to 20 percent water by the Codex standard, which caps moisture at 20 percent for most types, and its own water activity sits at roughly 0.50 to 0.65. Aloe juice is around 99 percent water at a water activity near 0.99. Pharmacopoeial glycerin is the only genuinely dry one, at 99 percent or better with the balance mostly water, low enough to be irrelevant.

Do the arithmetic on a formula. Five percent honey at 18 percent water adds 0.9 percent water to the batch. That sounds negligible, and the number itself is not the risk. The risk is that the water is not dispersed molecularly through the lipid, it is sitting in discrete droplets, and each droplet keeps its own local water activity of about 0.6. Osmophilic yeasts grow at 0.61. Two percent aloe juice is worse in kind rather than degree: those droplets are at 0.99, which supports anything. A bulk measurement averaged over the whole product tells you nothing useful, because the organism lives in the droplet, not in the average. The specific inputs are covered in honey in balms, and what happens when a wet finger or a steamy bathroom adds the water instead is in water contamination in balm.

The tack ceiling in a lip balm

Assume you have solved the separation problem with an emulsifier. There is a second ceiling, and it is sensory. Glycerin is viscous and hygroscopic, so it reads on the lip as cling. At 1 percent it is imperceptible in a waxy base. At 2 percent it gives a slight cushioned cling that most people read as pleasant. At 3 percent it is unmistakably tacky and starts picking up hair and dust. At 5 percent it is unpleasant, and in a warm pocket it will bead out no matter what emulsifier you used.

Honey has a lower ceiling still, because its sugars stay tacky after the water has gone: 1 to 3 percent in a tin is workable, and above 5 percent the product grabs. Total polar phase across all humectants is the number to watch, and about 5 percent is the practical maximum for a stick that people will use twice a day. Beyond the feel, tack costs you gloss, because a surface that grabs and collects dust is not a smooth reflecting one. If a formula has passed the ceiling, work through sticky, tacky balm before you assume the wax is at fault.

What the evidence actually supports

Glycerin has one of the strongest evidence bases of any moisturising ingredient. It raises stratum corneum hydration, it accelerates barrier recovery after damage, and there is credible mechanistic work on its role in corneocyte maturation and desquamation, most of it summarised in the dermatological review literature. It is used at 2 to 20 percent in creams and lotions and it earns its place there.

Every one of those studies used a vehicle containing water. That is not an incidental detail: the mechanism requires water, both to deliver the glycerin into the surface layers and to be the thing the glycerin holds. There is no controlled evidence that adding 1 or 2 percent glycerin to a wax stick improves lip hydration over the same stick without it, and given the quantities calculated above there is no strong reason to expect one. Where lip products have been tested, the ingredients with demonstrated performance are occlusives.

So the marketing claim and the science are pointing at different products. "With glycerin" on a balm is accurate as a statement of contents and close to meaningless as a statement of function. "With hyaluronic acid" is weaker still. Neither is dishonest in the legal sense, and neither tells you anything about how the balm will perform.

The decision rule

Decide what you are building before you reach for the glycerin bottle. If you want humectant action, the humectant belongs in a water phase, and the balm belongs on top of it: a hydrating serum or a damp skin application followed by an occlusive stick delivers what the marketing promises, and costs nothing to specify. If you want a genuinely water free product, keep glycerin at or below 2 percent with a water in oil emulsifier at a similar level, accept that you are buying label copy and a little cling rather than measurable hydration, and screen the batch warm before you commit. If you want more than that, build a preserved water in oil emulsion and test it properly, following anhydrous formulation for what you gain and lose by leaving the anhydrous category.

The cheapest version of all this is not a formula change. Apply the balm to skin that is still damp, within about three minutes of washing, and the occlusive holds surface water that is already there. That is the same physics the humectant was supposed to provide, done by the instructions instead of the ingredient list, and it is the whole premise of an overnight lip mask.

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