Magnesium balm: why it goes gritty, and three formulas that actually hold together

Magnesium chloride is deliquescent, which is why magnesium balms turn clumpy and weepy. The mechanism, three routes with full percentage formulas, and the honest evidence.

Magnesium balm is one of the few craft formats where the recipes circulating online fail for a reason nobody explains. People follow the method exactly, get a smooth balm on the day, and open the jar three weeks later to find a clumpy, sandy, slightly wet mess. That is not bad technique. It is chemistry the recipe never mentioned.

What a magnesium balm is, and what "magnesium oil" is not

Almost every magnesium skin product is built on magnesium chloride, usually as hexahydrate flakes (MgCl₂·6H₂O) or as the ready-made concentrated solution sold as "magnesium oil".

Magnesium oil is not an oil. It contains no lipid at all. It is a concentrated water solution of magnesium chloride, named for the slippery, faintly viscous feel of a strong chloride brine. The distinction is not pedantry. Stir it into a melt as though it were an oil and you are adding water to an anhydrous product, at which point everything on the anhydrous formulation page stops applying.

Deliquescence: the mechanism behind every failed batch

A deliquescent salt does not merely attract water. Above a threshold called the deliquescence relative humidity, it absorbs enough water vapour from the air to dissolve itself into a saturated solution, and then keeps going. That threshold is low for magnesium chloride.

The clean way to cite it is the humidity-calibration standard. A saturated magnesium chloride solution is one of the reference solutions used to hold a fixed relative humidity for calibrating hygrometers, and it sits at about 33% at 25 C (ASTM E104). Saturated sodium chloride, by comparison, holds about 75%. So any room above about 33% relative humidity, which is essentially every room you will work in, is wet enough for magnesium chloride to start pulling water in.

Now follow the water. Fat cannot dissolve an ionic salt, so a particle in a balm is not solvated, merely surrounded. It draws vapour through the wax network, which is permeable enough over weeks, and it draws water out of anything nearby that has some: an unrefined butter, a botanical extract, moisture in a tin, the humid air in the headspace each time the lid comes off. A droplet of brine forms around each particle. Brine and fat do not mix and no emulsifier is present, so droplets coalesce and surface as beading, while the partly dissolved salt cements neighbouring particles into hard grains.

The result reads as three separate faults, which is why it is rarely diagnosed correctly. It looks like balm sweating, it feels like grain, and it smells fine, so people blame the shea butter or the pour temperature. Antioxidants have no bearing on it either, which is worth saying because tocopherol and rosemary extract get thrown at every balm fault as a reflex.

The three routes, and what each one costs you

There is no fourth option in which dry magnesium chloride stays smooth in an open jar. Recipes promising it describe a balm that has not yet aged.

Making the brine, with the arithmetic nobody publishes

Routes A and B both start with a solution, and the number that matters is that the hexahydrate brings its own water: the crystal is 53% water by mass, so you need far less added water than a recipe written for anhydrous salt suggests. Near saturation at room temperature is roughly 35 g of magnesium chloride per 100 g of solution, so 100 g of near-saturated brine is about 75 g of flakes plus 25 g of water, carrying about 9 g of elemental magnesium. That is the same order of concentration as the products sold as magnesium oil.

Route A: water-in-oil magnesium balm, 200 g

All percentages are weight percent of the finished batch and total exactly 100. This is the version to make if you want something that feels like a balm rather than wet sand. Electrolytes help rather than hinder here: salt in the water phase of a water-in-oil emulsion reduces droplet coalescence, which is why industrial water-in-oil systems routinely add some.

Method: dissolve salt and glycerin in warm water, melt the oil phase at 75 to 80 C, bring both to 70 to 75 C, then add the brine to the fat in a thin stream under a stick blender. Cool with gentle stirring, adding preservative and tocopherol below 45 C. The emulsion is the fragile part, not the salt.

Route B: minimum-water version, 250 g

Same conclusion, different balance. Use only enough water to hold the flakes in solution, which pushes magnesium loading up and free water down without pretending the product is anhydrous. It still needs a preservative: the water activity remains high enough for microbial growth despite the salt.

Route C: the honestly anhydrous version, 200 g

If you want a genuinely water-free product with no preservative, this is what it looks like, and this is where most sites go quiet. Milled salt suspended in fat is perceptible on the skin, and that is not a fault you can formulate away. Mill the flakes fine in a dry grinder, pre-wet the powder with the castor oil and the dispersant to break up agglomerates, and add it below 60 C so the hexahydrate does not start releasing its own water.

Package this one properly or do not make it. It needs a genuinely airtight container with a gasket or a wadded liner, minimal headspace, no wet fingers and no bathroom shelf. The packaging guide covers closures; the point here is that a slide-lid tin is not airtight and will let this product fail within a season. Expect grittiness from the start and more with age, and say so on the label.

Stinging, tingling and itch: why it happens

Many users report tingling, prickling, itching or a definite sting, particularly in the first weeks. No exotic mechanism is needed. Physiological fluid sits near 0.9% sodium chloride; a magnesium brine at 25 to 35% salt is dozens of times more concentrated. On skin it sets up a steep osmotic gradient, draws water out of the surface layers, and delivers a concentrated divalent chloride to nerve endings well able to report it.

Three things make it worse: recently shaved skin, cracked skin where the barrier is compromised, and application straight after a hot shower. The responses are to halve the magnesium loading, apply to intact skin only, and stop if redness persists. The broader logic is in balm stings or burns.

What the tingling is not is evidence of absorption or efficacy. That is the same error dermatologists warn about with menthol lip products, where a sensation gets read as the product working when it is the product irritating.

Transdermal magnesium: the honest position

The claim that magnesium crosses skin in nutritionally meaningful amounts is asserted far more often than it is tested.

First, the physical chemistry is unfavourable. The barrier is the lipid lamellae of the stratum corneum, which passes small, uncharged, moderately lipophilic molecules best. A magnesium ion is doubly charged and carries a tightly held shell of water, about as poor a candidate for passive lipid diffusion as a common ion can be. Routes through hair follicles and sweat ducts exist in the literature as hypotheses, not demonstrated pathways. What applied products mostly do instead is covered under occlusive, emollient and humectant.

Second, the human evidence is a small number of small studies with weak outcome measures. Serum magnesium is tightly regulated and holds only a fraction of body magnesium, so a real change in stores moves it very little. A 2017 pilot study of a transdermal magnesium cream in PLoS One reported changes that did not reach statistical significance in the whole group, and a 2017 review in Nutrients titled "Myth or Reality: Transdermal Magnesium?" concluded the evidence was insufficient to settle the question. Neither is a refutation and neither is a confirmation.

Trebalm's line is narrow. A magnesium balm is a cosmetic. It puts a salt on skin, and a salt on skin has local effects people notice. Anything beyond that is not established, and building a product around it is a regulatory and an honesty problem.

Diagnosing a magnesium balm that has gone wrong

What to make instead, if you want the format without the problem

Many people who set out to make a magnesium balm actually want a rich occlusive body product that feels substantial and stays put. If that is the goal, an ordinary body balm does it with none of the deliquescence, preservation work or sting, and a foot and heel balm does it with more structure for thicker skin. Adding a hygroscopic salt buys a marketing word and a formulation problem.

If you do want the magnesium, make route A, preserve it properly, test it in a warm room for a month before making a second batch, and write the label to describe what it is rather than what you hope it does. The formulation is not difficult once you accept that it is an emulsion. The failures come almost entirely from pretending otherwise.

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