How much wax does it take to turn liquid oil into a solid?

Sunflower and rice bran wax gel oil at 0.5 to 1 percent, beeswax at 2 to 4, carnauba at 4. Why balms still use 8 to 20 percent, and what the gap buys you.

Very little. In published tests, sunflower wax stopped soybean oil flowing at 0.5 percent by weight, rice bran and sunflower wax gelled canola oil at 1 percent, candelilla at 2 percent, carnauba at 4 percent, and beeswax gelled camellia, soybean and sunflower oil at 3 percent. That is the critical gelling concentration: the least wax that holds an oil still in an upturned jar at room temperature. A balm uses far more, 7 to 12 percent wax in a tin and 10 to 20 percent in a twist-up tube on this site's formulas, because a gel that barely stops flowing at 25 C is nowhere near firm enough to survive a pocket, a warm car or a stick being pushed across a lip. The first few percent of wax make the oil a solid. Everything above that buys firmness and heat resistance, and it buys them quickly: in one study, doubling beeswax from 3 to 6 percent made a sunflower oil gel about seventeen times firmer.

• Sunflower wax gels at 0.5-1%
• Rice bran wax: 1%
• Candelilla: 1-2%
• Beeswax: 2-4%
• Carnauba: 4%
• Tin balm: 7-12% wax
• Tube: 10-20% wax

What an oleogel is, and why every balm is one

An oleogel is a liquid oil immobilised by a small amount of a solid that builds a network through it. Scharfe and colleagues, in a 2022 study from TU Berlin, describe the wax version directly: waxes dissolved in edible oils "at sufficient concentrations" may "form a 3D crystalline network that can immobilize the liquid". The oil is still liquid. It has not reacted with anything or changed into a fat. It is held in place by a scaffold of crystals that touch each other, and if you break that scaffold, by melting it or by shearing it hard enough, the oil runs again.

That is a fair description of most anhydrous skincare. A salve on this site is "typically 70 to 90% liquid oil, 8 to 15% wax and little or no butter", in the words of the herbal salve formula, so the solid that holds the whole tin together is a minor ingredient by weight. Butters complicate the picture, because shea or cocoa butter adds a second, fat crystal network of its own, and that is covered in what a balm is. But the wax and oil skeleton underneath is an oleogel, and the numbers food scientists have measured for oleogels are the numbers that explain why a balm sets, how much wax it needs, and why it sometimes weeps oil.

The critical gelling concentration, wax by wax

The critical gelling concentration, CGC, is the lowest concentration of a gelator that turns a liquid into a gel. The usual test is crude and practical. Pang and colleagues, in a 2020 study of beeswax in four vegetable oils, melted the wax into the oil at 85 C, let it set overnight, stood the samples at 25 C in 60 ml glass bottles and turned them upside down after 24 hours: "The concentration observed when the mixture did not flow was the critical concentration for oleogelation." The table gathers the figures from five studies that used food grade waxes in food oils.

Read down the middle column and an order appears that every study agrees on in outline. Sunflower wax is the most efficient gelator, rice bran wax is close to it in most oils, candelilla sits next, beeswax needs a little more, and carnauba needs the most. Scharfe and colleagues, using a more precise method described below, found the same ranking from the other end: "the rheological CGC increases in the order SFX<CLX<RBX<BWX", that is sunflower, candelilla, rice bran, beeswax. Rice bran and candelilla swap places between studies; the two ends of the list do not move.

The rice bran oil row is the exception worth noticing. Doan and colleagues found that rice bran wax, Brazilian carnauba and a fruit wax "showed weak gelling behavior in rice bran oil (prepared at concentrations as high as 5 % w/w)", while beeswax, candelilla, sunflower wax, berry wax and a wild carnauba grade formed "strong gels at concentration of <2 % w/w". Rice bran wax, which gels canola and olive oil at 1 percent, struggles in the oil it was refined out of. The rice bran wax page carries the practical consequence: check its performance in your own carrier before you trust a figure from a canola study.

Why the published numbers disagree, and which to believe

Anyone who searches for the gelling level of a wax finds figures that differ by a factor of five for the same wax, and it is tempting to conclude the research is unreliable. The spread has three identifiable causes, and once you know them you can read any figure correctly.

The method is coarse. The upside-down jar test only reports the first concentration in a series that holds, so its answer can be no finer than the step between samples. Scharfe and colleagues compiled the published values for four waxes and noted that the CGC "is often determined arbitrarily by increasing the wax concentration gradually, using large increments and different procedures", with steps of 0.5, 1 and even 2 percent in the studies they collected. Their compilation lists beeswax at 2 to 3 percent in soybean oil, 1 percent in high oleic sunflower, 1.5 percent in rice bran oil and 4 percent in both medium and long chain triglyceride oil. Their own method lowered the wax in steps of 0.05 percent and called it a gel once the storage modulus exceeded the loss modulus at 20 C, the rheologist's definition. By that method, "for all waxes studied, the data gathered here indicate a lower CGC" than the literature values. A jar test is a fair test of what a maker sees; it simply overstates the true minimum.

The wax is not one material. Hwang and colleagues tested waxes from several suppliers. Sunflower wax from all three suppliers gelled soybean oil at 0.5 percent, but candelilla and rice bran wax "also showed good gelation properties, which, however, varied with different suppliers", and the authors concluded that "gelation ability of a wax is significantly dependant on its purity and detailed composition". A refined, deodorised wax and a crude one sold under the same name are different products, which is part of why waxes compared gives ranges rather than single values.

The oil is not one material either. Pang's four oils needed 3 percent beeswax, except flaxseed, which needed 4. Scharfe's group stripped the minor components out of canola and sunflower oil by column chromatography and found that "the presence of PCs significantly reduces the CGC of all waxes but SFX", PCs being polar components such as free fatty acids and oxidation products. The effect "is more pronounced in BWX", beeswax. An ordinary unrefined or slightly aged oil therefore gels at a lower wax level than a highly purified one, and beeswax is the wax most sensitive to the difference.

Which figure to believe, then? For a jar on a workbench, the jar test figures from food oils in the table above are the right order of magnitude. Treat the lowest value for each wax as a floor that you are unlikely to beat and the highest as what a fussy oil or an unusual wax grade can demand.

How little of the wax is actually doing the work

Even the critical gelling concentration overstates how much solid it takes, because part of the wax at that level is not solid at all. Waxes dissolve in warm oil and come out of solution as it cools, and some remains dissolved at room temperature. Scharfe and colleagues used calorimetry on gels at two wax levels to estimate how much wax stays dissolved, and from that how much solid is present at the gel point in stripped oils. "The values generated are very low, 0.075-0.722% w/w."

Under one percent of crystalline material, and in the best case under a tenth of a percent, is enough to stop a jar of oil flowing. The reason is geometry. Wax crystals in oil grow as long needles or thin platelets rather than compact lumps. Pang's beeswax gels formed crystals that "were needle shaped", and Dassanayake and colleagues saw "thin long needle-shaped crystals" of rice bran wax in olive and canola-soybean oil, finely dispersed through the liquid. A long, thin crystal reaches across far more space per gram than a cube, so a small mass of them touches enough neighbours to form a continuous, space-filling scaffold. The oil sits in the spaces between.

Crystal shape also explains how fast a gel forms. Dassanayake and colleagues followed setting with calorimetry and viscosity and found that the rice bran wax gel "formed soon after RBX crystallized, whereas a time delay was observed between the organogel formation and wax crystallization of CRX and CLX", carnauba and candelilla. With rice bran wax, crystals appear and the network exists almost at once. With carnauba and candelilla, crystals form first and take time to connect.

Gelling power is not melting point

The most persistent mistake in balm making is to treat the hardest, highest melting wax as the strongest structurant. The table says otherwise. Carnauba has the highest melting point of the common natural waxes, with the carnauba wax page giving an 80 to 86 C drop point, and it needs the most wax to gel an oil: 4 percent in both canola and olive oil, against 1 percent for rice bran wax in the same oils.

The composition column comes from the table Scharfe and colleagues adapted from Doan's chemical profiling of the same four waxes, and it carries the explanation. Their conclusion is that a solid-like structure is promoted by "(1) high WEs content and (2) a limited difference in WE chain length (SFX), and (3) more homogenous composition with high concentrations of components that co-crystallized (HC, predominantly C31 in CLX)". Sunflower wax is almost pure wax ester of a narrow range of chain lengths. Candelilla is dominated by a single hydrocarbon, about 60 percent C31 according to the same paper. Both are uniform materials that crystallise as one population of long, thin crystals. Beeswax is a mixture of esters, hydrocarbons, free acids and alcohols, and Scharfe's summary is that the CGC "increases with wax inhomogeneity and WE chain length difference". Melting point tracks chain length; gelling power tracks crystal shape and how uniformly the wax crystallises. The two are separate properties, and the full argument for carnauba specifically is in candelilla wax vs carnauba wax.

Why a balm uses two to seven times the gelling minimum

Here is the confusion behind the search. Oleogel papers say a few percent of wax solidifies an oil. Balm recipes use 8, 15 or 25 percent. Both are right, because they describe different thresholds. The gelling concentration is the point at which the oil stops flowing under its own weight in a jar at room temperature. A balm has to do much more than that: hold a shape when pressed, survive temperatures well above 25 C, and in a tube, stand as an unsupported stick that is pushed sideways across a lip. Each of those needs more network than a jar test does.

The tin and tube windows come from lip balm ratios, which also converts the popular part-based recipes into percentages and shows a firmness scale from 5 to 25 percent wax. A cold reader who wants a starting number should take it from there: 15 percent beeswax for a tube in a temperate climate, and the middle of the 7 to 12 percent window for a tin. The gelling minimum is the floor below which no amount of cooling or stirring will make a solid, not a target.

Firmness climbs much faster than wax content

The reason a few extra points of wax change a balm so much is that firmness does not rise in proportion to wax. Pang and colleagues measured firmness with a 12.7 mm cylindrical probe pushed 10 mm into each gel at 25 C, from the gelling minimum upwards in 1 percent steps.

Doubling the wax multiplied firmness by roughly seventeen to twenty-six times in the three oils that gelled at 3 percent. At the gelling point itself the network is barely connected: Pang notes that camellia oil at 2 percent beeswax "was mostly a flowing liquid", and at 3 percent its network "may not have been well developed" compared with soybean and sunflower. Each extra percent of wax adds crystals to a network that already spans the jar, so it adds connections rather than simply more solid, and stiffness climbs steeply. The practical consequence is that in the range a maker works in, small changes matter. One or two points of wax are the difference between a tin that dents and a tin that resists, which is why balm too hard or too soft adjusts in single percentage points rather than fives.

Dilution pulls the melting point down

The second reason balms use so much wax is heat. A wax in oil does not melt at the wax's own melting point. Some of it dissolves in the oil as it warms, and the network falls apart well below the temperature at which the pure wax would melt. Pang's calorimetry put the melting peak of the beeswax gels at 46.17 to 49.88 C across 3 to 6 percent wax in camellia, soybean and sunflower oil, rising steadily with wax content. Beeswax itself melts at 62 to 65 C by the USP method cited on the beeswax page. At gelling levels, then, the network gives way roughly 12 to 19 degrees below the wax's own melting range.

Hwang's sunflower wax gels showed the same pattern over a wider span: the soybean oil gels had a "melting point range from about 47 to 65 °C with 0.5-10 wt% SW". More wax keeps the network intact to a higher temperature, which matters because a balm in a car or a sunny window sits far above 25 C.

Holding the oil: why some gels weep

A gel can be firm and still lose oil. Gelling power and oil binding are different measurements, and the wax that gels at the lowest level is not necessarily the one that holds its oil best under stress. Blake, Co and Marangoni measured how much oil canola oleogels of four plant waxes lost when stressed, and reported that "gels containing candelilla wax exhibited the greatest oil-binding capacity, as they retained nearly 90 % of their oil". They attributed it to "the small crystal size as well as the spatial distribution of these crystals", and related oil binding to crystal size, mass distribution and porosity rather than to the gelling minimum. In their study candelilla needed twice as much wax to gel canola oil as rice bran or sunflower wax, and still held its oil better.

That separation explains a common fault. A balm can be firm on day one and bead oil on the surface after a week, which is the problem diagnosed in balm sweating. The wax is present and the network spans the tin, but the spaces in it are too coarse, or the crystals coarsen with time and temperature swings, so the network no longer holds all of the oil. Raising the wax helps; so does switching part of it to a wax whose crystals are finer.

Cooling rate changes the network, differently for each wax

Because the network is made of crystals, how fast it is cooled changes it. Hwang found that "the dependence of firmness on cooling rate was so significant that the desired texture of an organogel could be achieved by controlling the cooling rate in addition to controlling the amount of gelling agent" in sunflower wax gels. Scharfe and colleagues tested all four waxes at 10 percent in purified canola and sunflower oil, cooled either at 10 C per minute or at 0.05 C per minute, and found the effect depends on the wax:

• Sunflower wax: the hardest gels of the four at both rates, but "cooling at 0.05 °C min-1 resulted in significantly softer gels".
• Rice bran wax: the opposite, "RBX-based gels became harder at the low cooling rate", though with large variation across the dish that the authors attribute to uneven crystallisation.
• Beeswax and candelilla: "no considerable difference" with cooling rate in the purified oils, and beeswax gave the softest gels of the four.

The beeswax result needs one qualification. When the oils still carried their natural polar components, the same paper found that beeswax gel hardness "appears to be sensitive to both cooling rate and the content of polar components, while the type of oil is negligible". So in an ordinary oil, cooling still matters for beeswax. For balm makers the general advice in cooling rate holds, and for good reason in any formula with a butter, but the size of the effect on the wax network itself depends on which wax you use. A sunflower wax balm left to cool slowly on a bench will be noticeably softer than the same balm chilled; a beeswax and candelilla balm will show less difference from the wax alone.

What this means at the bench

The research boils down to four working rules for a balm with wax and liquid oil as its skeleton.

1. Start from the product, not the gelling minimum. Use the ranges in the product table above and the wax ratio calculator. The gelling minimum tells you where a product stops being a liquid, not where it becomes a balm.
2. Expect a few points of wax to matter a great deal. Firmness climbed seventeen to twenty-six fold between 3 and 6 percent beeswax in Pang's oils. Adjust in steps of one or two percent and judge after three days.
3. Do not choose a wax for its melting point. If the goal is to gel oil with as little solid as possible, sunflower and rice bran wax do it at a third of the beeswax level, and carnauba is the least efficient of the five. Use carnauba at 1 to 3 percent for heat resistance and a glossy surface, as the carnauba page sets out, alongside a better gelator.
4. Choose for oil binding separately. If the fault is weeping rather than softness, a finer crystal wax such as candelilla, or a little ozokerite, can fix it where extra beeswax only makes the balm harder.

For vegan formulas the gelling data are directly useful, because the aim there is to replace beeswax with waxes that behave differently. The vegan beeswax substitutes page works through the swap.

Where these numbers stop applying

Every figure on this page comes from a single wax in a single liquid oil, almost all of them food oils: canola, soybean, sunflower, olive, camellia, flaxseed and rice bran oil. None of the studies tested castor oil, jojoba, squalane, a medium chain triglyceride used in cosmetics, or a blend containing shea, cocoa or mango butter. The Scharfe results show that minor components of an oil shift the gelling minimum, so a castor-heavy lip balm cannot be assumed to gel at the same wax level as sunflower oil. Jojoba is a liquid wax rather than a triglyceride, and no gelling data for it are cited here.

The gelling minima are room temperature values, measured at 20 or 25 C. The melting and firmness figures are from gels made in laboratory beakers and Petri dishes with controlled cooling, then measured at 24 hours. None of them describes a finished product in a tin, and none tells you how a balm changes after weeks of storage. For that, the questions are crystal coarsening and polymorphism, covered in balm changed after shipping. And none of these studies measured how anything feels on skin; for slip, drag and payoff, start with balm texture science.

What the numbers do settle is the order of magnitude and the order of the waxes. A few percent of wax is enough to make a liquid oil solid. Balms use two to seven times that to make the solid useful, and the hardest wax on the shelf is the least efficient at the first job.

Reviewed and updated 16 October 2026. Spotted an error? Tell us and we will fix and log it.