A high performance face balm built from the lipid classes skin actually uses

A premium anhydrous barrier balm with ceramide NP at 0.5 to 1 percent, phytosterols at 2 and squalane at 35, plus solubility temperatures and the cost per gram.

Almost every balm on this site is built from triglycerides and wax, a lipid class the stratum corneum does not contain. This one does something different: it puts ceramide, sterol and free fatty acid into an anhydrous base in known quantities, keeps them dissolved, and is exact about what that does and does not buy. It is the most expensive formula published here, and the section on cost is not an afterthought.

• Ceramide NP 0.75%
• Squalane 35%
• Wax only 3%
• Hold 75 to 80 C
• 15 ml jar, 13.8 g

The problem this formula sets out to solve

A conventional balm works by occlusion. It lays a hydrophobic film over the skin, water loss falls while the film is present, and the skin repairs underneath. That mechanism is well evidenced, it does not require any particular oil, and it is set out in moisturisers and the skin barrier. Nothing here improves on it.

What this formula adds is composition. The intercellular lipid matrix of the stratum corneum is roughly equal thirds ceramides, cholesterol and free fatty acids by mole, with essentially no triglycerides, as stratum corneum and barrier lipids describes. If there is a case for delivering those classes rather than more sunflower oil, this is the vehicle it needs: high in a saturated, non-oxidising liquid so the actives have somewhere to sit, low in wax so the crystal network does not exclude them, and built so nothing crystallises out over the product's life.

That last requirement is the hard one. Ceramide NP melts near 95 to 110 C and phytosterols near 135 to 140 C, so neither is molten at any temperature you should take a balm to. You are not melting them, you are dissolving them, and a solution made at 80 C is not necessarily a solution at 20 C. Everything below is arranged around keeping them in.

The formula

Percentages are weight percent and total exactly 100. The gram columns are for a 90 g batch filling six 15 ml glass jars at 13.8 g (82.8 g) with 7.2 g of overage, and for a 450 g batch filling thirty.

Squalane at 35 percent is the largest single decision. It is a fully saturated C30 hydrocarbon with an iodine value near zero, so it cannot go rancid, it absorbs without residue, and it is the nearest thing to an inert liquid that still feels like an oil. Where the actives are the point, a third of the formula that will never oxidise is worth what it costs; the grades are compared on the squalane page. Jojoba and meadowfoam fill out the liquid phase for the same reason: both are unusually oxidation-resistant for plant-derived oils.

Why 3 percent wax and 2 percent behenyl alcohol

Three percent beeswax is the least that keeps a 72 percent liquid phase from slumping in a wide jar. Crystals exclude dissolved solutes as they grow, pushing the ceramide into the shrinking liquid between them and towards its solubility limit, so more wax means more exclusion. The wax is therefore held to the minimum the container needs, and the container is a squat 15 ml jar used with a spatula rather than a tube that would demand a self-supporting stick.

The body that the missing wax would have provided comes from behenyl alcohol instead. It is the C22 fatty alcohol, melting at roughly 65 to 73 C, and at 1 to 3 percent it thickens an oil phase and adds a cushioned, non-draggy feel without behaving like a wax; the comparison across cetyl, stearyl and behenyl is in fatty alcohols. It is worth being clear that behenyl alcohol is the C22 alcohol, not the C22 acid, so it contributes nothing to the free fatty acid side of the ledger no matter how convenient that would be.

The lipid ratio argument, and what it actually supports

The claim you will meet everywhere is that a barrier product should supply ceramides, cholesterol and free fatty acids in a 3:1:1 ratio. The work behind it is real: on acetone-disrupted and tape-stripped skin, incomplete two-lipid mixtures delayed recovery, complete equimolar mixtures allowed normal recovery, and recovery accelerated when any one of the three was raised up to threefold. Not the ceramide specifically. Any of them. The detail, and the weight conversions, are on ceramides and barrier lipids.

Three things stop this formula from claiming that result. The lipids in those experiments were dosed in a volatile propylene glycol and ethanol vehicle onto deliberately stripped skin, not in a wax and triglyceride film onto intact skin. The sterol here is a phytosterol, not cholesterol, and they are different molecules with different behaviour in a lamellar phase. And the free fatty acid is whatever the shea brings, which is neither controlled nor the right chain length.

So the balm is sterol-dominant on a molar basis, roughly 1 to 3.8 to 1.7, and the fatty acid figure carries an error bar wide enough to drive through: unrefined shea runs about 1 to 8 percent free fatty acid depending on kernel handling and storage, which is 0.2 to 1.6 percent of this balm, and refined shea would take it near zero. That is the real limitation of using a butter as the acid source, and shea butter explains why the number moves.

Getting crystalline actives into solution

Ceramide NP is a white crystalline powder that suppliers describe as dispersible rather than soluble in triglycerides and mineral oil. Dispersible means it will sit in the oil as fine solid particles, which on the skin reads as grit. To get it dissolved you need a solvent with more polarity than a triglyceride, heat, time and agitation, and this is the general behaviour set out in solubility in anhydrous formulas.

The premix does that work. Weigh the 13 percent polar ester into a small beaker, add the ceramide and the phytosterols, and hold at 78 to 82 C with continuous stirring for 20 to 30 minutes. What you are waiting for is optical clarity: lift the stirring rod and look for haze against a light. A premix that goes into the batch hazy will be gritty in the jar. Do not push above 85 C to force it, which degrades an expensive material rather than dissolving it, and do not shorten the hold because the surface looks clear at ten minutes. Dissolution of a high-melting crystal is rate-limited by diffusion away from the particle surface, so time does most of the work.

The controlled cool, and the cure that decides whether it grains

Once the premix is combined into phase B, the batch is a solution of two crystalline solids in a mixture that is itself about to crystallise. Cooling is where you decide the outcome.

Cool with slow continuous stirring at roughly 1 to 2 C per minute, which for a 90 g batch means out of the bath and stirred in air, not plunged into cold water. At 62 to 65 C the beeswax and behenyl alcohol nucleate and the batch turns faintly hazy. That haze is wanted: the crystallising wax builds a network that immobilises the dissolved ceramide, so it cannot diffuse to a nucleation site and grow. Left to stand through this range rather than stirred through it, you get fewer, larger wax crystals and far more freedom of movement for everything dissolved between them, which is the mechanism in crystallisation and cooling rate.

Pour at 55 to 60 C, working continuously so nothing thickens in the jug, then move the filled jars to 4 to 8 C for 30 minutes to lock the network fast. Not a freezer: contraction that quick cracks the surface and pulls the balm off the glass. Then 48 hours at 18 to 22 C before capping and judging texture. The pour window logic generalises across the site at pour temperatures.

None of that is the acceptance test. Ceramide recrystallisation in an anhydrous base characteristically appears at week two or three, not on day one, which means a batch that passes at the bench can fail at the customer. Retain one jar from every batch, keep it at 20 to 22 C, and rub a fingertip of it on the inner wrist at day 21. Run a second retained jar through five cycles of 24 hours at 4 C and 24 hours at 40 C, which is the standard screen described in accelerated ageing, and check that one too. Only then is a batch releasable.

Method

1. Sanitise and dry. Jars, jug, beaker, spatula, thermometer. This is a waterless product with no preservative and a very high material cost per jar, so a contamination failure is expensive twice over. See anhydrous formulation.
2. Make the phase A premix. Ester, ceramide NP and phytosterols in a small beaker. Hold at 78 to 82 C (172 to 180 F), stirring, for 20 to 30 minutes until optically clear. Keep it warm.
3. Melt phase B in a water bath at 75 to 80 C (167 to 176 F). Hold ten minutes after the last solid vanishes so no shea crystal survives to seed grain later.
4. Combine. Pour the warm premix into phase B, rinse the beaker with a little of the warm batch, and hold the combined batch at 75 to 80 C for five minutes with stirring.
5. Cool to below 60 C (140 F) with slow continuous stirring, then add phase C.
6. Pour at 55 to 60 C (131 to 140 F), 13.8 g per 15 ml jar, in one continuous movement per jar.
7. Chill 30 minutes at 4 to 8 C, uncapped, then cure 48 hours at room temperature. Cap, label, and hold a retain for the day 21 check.

What it costs per gram, and what that buys

Cost is a design constraint here rather than a footnote, so it is worth putting numbers on it. The table below uses arbitrary cost units per gram, anchored so that refined high oleic sunflower oil is 0.01. Substitute your own supplier figures; the arithmetic is linear, and the cost calculator will do it for a real price list.

Read the table twice. The base without the ceramide costs 1.73 units per jar, of which squalane is 61 percent. The ceramide line is 2.08 units at blend-equivalent pricing and 20.8 units if you buy pure powder at the top of the range, which takes the jar to 22.5. A plain shea, beeswax and sunflower balm of the same weight costs about 0.28 units. So the base alone is roughly six times a conventional balm, and the ceramide can multiply that again by six.

Set that against a supermarket ceramide cream and the comparison is uncomfortable, because in most markets the retail price per gram of a large tub is below the raw material cost per gram of this formula. That is not a failure of your sourcing. A cream is 60 to 80 percent water, bought at tonne scale, with a ceramide content frequently at the label-facing end of the 0.01 to 1 percent range. What you buy by making this instead is a known composition and a fill weight of pure lipid, which is a narrower proposition than the price gap suggests: see do expensive balms work better.

Shelf life, and the temptation that ruins it

This blend is deliberately hard to oxidise. Squalane has no double bonds, jojoba is a wax ester rather than a triglyceride, meadowfoam is dominated by very long chain monounsaturates, and shea is largely oleic and stearic. Two years unopened in a cool, dark place is realistic, and the 0.15 percent mixed tocopherols with 0.10 percent deodorised rosemary extract are there as insurance rather than as the main defence; the working ranges are in vitamin E and antioxidants.

The temptation is to make an expensive balm sound more expensive by adding rosehip, sea buckthorn, evening primrose or borage. All are polyunsaturated with induction periods measured in months, and 5 percent of any of them replaces a two-year blend with a six-month one. You would then have oxidised the costliest product you make, and the cardboard note that follows gets blamed on the ceramide rather than on the linolenic acid that caused it.

What goes wrong

What you can claim

Ceramides and phytosterols have no over-the-counter monograph anywhere, so any claim to treat or prevent a skin disease makes the product an unapproved drug in the United States regardless of level. In the EU and UK, claims must satisfy the common criteria of Commission Regulation (EU) No 655/2013, and the criterion that bites is evidential support: evidence for the claim as made about the product as sold, not about the ingredient in someone else's study on stripped skin.

Wording that stays cosmetic: an anhydrous facial balm containing 0.75 percent Ceramide NP; rich in skin-identical lipids; contains ceramide, sterol and fatty acid fractions; leaves skin feeling comfortable and less tight. Wording that does not: repairs the skin barrier, restores the skin's ceramides, rebuilds the lipid matrix, treats eczema or rosacea. A moisture-loss claim sits in between and becomes defensible only if you hold instrumental data on the finished product, which means measuring it the way transepidermal water loss describes rather than inferring it. The jurisdictional detail is in cosmetic versus drug claims.

When this balm is worth building

The honest position is that no one has demonstrated a wax balm to be a good vehicle for delivering barrier lipids into the stratum corneum. A high-melting crystalline solid held in a fatty film has little thermodynamic reason to leave it for the skin, and the trial evidence that moisturisers help compromised skin does not distinguish lipid-containing products from plain ones. If you build this and it works well, the base is doing most of the work, and the base would do most of that work at a sixth of the cost.

So the decision rule is narrow. Build it if you want a specific, declarable composition and you are willing to run a 21-day retained-jar check on every batch, because the failure mode is invisible on the day you pour. Build it at 0.5 to 1 percent ceramide or not at all, since below 0.3 percent the inclusion is decorative and above 1 percent the grit is close to certain. And if either of those conditions is uncomfortable, make the same base without phase A: 35 percent squalane, a stable oil blend, 3 percent wax and 2 percent behenyl alcohol is an excellent face balm on its own terms, and it will still be perfect in a jar at week three.

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