Preservation of anhydrous products
At the "Walker Formulation Academy" school, we regularly analyse topics where cosmetic chemists make the most mistakes — and the preservation of anhydrous products is one of them.
Is it necessary, and how to do it correctly — a methodological breakdown for the Formulation Club.
In a nutshell. An anhydrous product — a balm, butter, solid body butter, or lip balm — in most cases does not need a preservative. However, it has its own separate enemy that is often forgotten. Let's break it all down: why a preservative is usually not needed, what is needed instead, and the only case where a preservative is actually justified — and how to choose it correctly.
1. Why microbes do not live in an anhydrous product
Microorganisms — bacteria, yeast, and mold — need free water to grow and multiply. Not just "moisture in general," but specifically available water, which is characterized by the water activity (aw) index — this is the portion of water that a microbe can "take" and use.
In a true anhydrous product (oils, butters, waxes), the water activity is so low that no bacteria, yeast, or mold can grow in it. There is simply nothing to take. Therefore, there is essentially nothing to preserve there — there is no aqueous phase in which a preservative would need to work.
Hence the first important conclusion: if you add a preservative to a dry balm "just in case," it simply dissolves in the oil and sits idle. Microbes don't need it there — they have nothing to feed on.
2. What actually threatens a balm: oxidation, not microbes
A balm has its own real enemy, and it is not microbiology, but chemistry — rancidity (oxidation) of oils.
Unsaturated fatty acids in oils oxidize over time under the influence of oxygen, light, heat, and traces of metals. A characteristic "old," rancid smell appears, the color changes — the product spoils. This is a completely different process, and it is solved with a different tool:
- against oxidation — an antioxidant (vitamin E / tocopherol, rosemary extract, etc.);
- against microbes — a preservative.
These are different tasks. In an anhydrous product, you almost always need an antioxidant, not a preservative. Confusing them is a common beginner's mistake.
3. The only real microbial risk: water introduced by the consumer
So why does anyone preserve balms at all? Because of one specific scenario — water introduced by the consumer during use:
- wet fingers used to scoop the product out of an open jar;
- condensation and steam in a warm bathroom;
- a drop of water accidentally falling onto the surface.
This drop creates a tiny local zone with high water activity — a micro-area where mold can easily take hold. This is exactly how those fuzzy spots appear, which are sometimes found on old lip balm or body butter. The balm itself hasn't "gone bad" in the usual sense — water has simply gotten into it.
It is important to understand: the threat is not that microbes are growing throughout the bulk of the balm. The threat is that specific introduced drop. And everything else follows from this: if we are preserving an anhydrous product, the preservative has one task — to reach this introduced drop and protect it.
4. Principle #1 — pH independence
This is the first thing that eliminates half of the usual options.
The water introduced by the consumer is uncontrolled: tap water with a pH of about 7–8, splashes from the shower, condensation. You cannot acidify it — it is already neutral.
And most "natural" preservatives are organic acids: sorbic, benzoic, dehydroacetic, anisic, levulinic. They only work in their undissociated (protonated) form, which exists only at a pH below their pKa — that is, in an acidic environment (usually pH < 5). In a neutral introduced drop, such acids are practically "switched off," even if they are present in the formula by percentage.
Conclusion: to protect against introduced water, you need a preservative that works over a wide pH range, up to neutral. This immediately makes phenoxyethanol and benzyl alcohol favorites (both are alcohols, not acids, and do not depend on pH as much) and pushes acid-based systems aside.
5. Principle #2 — solubility: the preservative must reach the water
To protect the introduced drop, the preservative must distribute into it — that is, it must have sufficient water solubility. And here, a beautiful trap with parabens arises, which is worth discussing separately because intuition fails us here.
The paraben chain length trap
Parabens are esters of p-hydroxybenzoic acid, and they differ by the length of their alkyl chain: methyl- → ethyl- → propyl- → butylparaben. As the chain lengthens, two things happen simultaneously:
- the preservative becomes stronger against mold and yeast — which sounds ideal, since mold is our threat;
- but it also becomes increasingly oil-soluble and less water-soluble.
And here is the problem:
- Butylparaben — water solubility ~0.02%, log P ~3.6. It "buries" itself in the oil phase and barely reaches the water drop.
- Methylparaben — solubility ~0.25%, it gets into the water better, but it is the weakest member of the family against mold.
It turns out to be a paradox: the paraben that is strong against mold cannot reach the water; and the one that reaches the water is weak against mold. For our task — an introduced drop where mold is the specific threat — this is a dead end.
Why phenoxyethanol bypasses the trap
Phenoxyethanol — water solubility ~2.7% (about 100 times higher than butylparaben), log P ~1.1. It is amphiphilic: it dissolves in oil and maintains a sufficient concentration in water. Therefore, it can actually migrate into the introduced drop and work there. Plus, it does not depend on pH. Two key properties from points 4 and 5 — in one substance.
And one more thing — the regulatory wall
Even if you wanted to use the longest-chain, most effective mold-fighting parabens, you cannot. In the EU and UK, isobutyl-, isopropyl-, pentyl-, benzyl-, and phenylparaben are prohibited (they have been added to the list of banned substances due to a lack of safety data and concerns regarding endocrine activity). Propyl- and butylparaben are limited to a total of 0.14% — too little to rely on them.
Note: Regulations vary by region. The EAEU follows TR CU 009/2011 with its own standards; always check the requirements of the market where you are selling.
Summary: where each preservative stands
| Preservative | Water sol. | log P | Volatility | Works at neutral pH |
|---|---|---|---|---|
| Benzyl alcohol | ~4% | ~1.1 | high | yes |
| Phenoxyethanol | ~2.7% | ~1.1 | very low | yes |
| Methylparaben | ~0.25% | ~2.0 | low | yes, but weak against mold |
| Butylparaben | ~0.02% | ~3.6 | low | yes, but "locked" in oil |
| Acids (sorbic, etc.) | depends | — | usually low | no — requires pH < pKa |
6. Principle #3 — volatility: preservatives can evaporate
The third property that is often forgotten is volatility.
Some preservatives are noticeably volatile. If a product sits in an open jar, especially in a warm, humid bathroom, a volatile preservative can gradually evaporate from the surface throughout the product's shelf life. This means the dosage you incorporated is not necessarily what will remain after a few months.
- Phenoxyethanol — practically non-volatile (vapor pressure is very low). Reliable in this regard.
- Benzyl alcohol — noticeably more volatile (vapor pressure is approximately 100 times higher). This doesn't matter for a sealed tube, but for an open jar, it is a real drawback.
Rule of thumb: add the preservative during the cooling phase (usually < 40 °C or according to the supplier's TDS) and use airtight packaging.
7. Charge and compatibility (briefly, but important)
Phenoxyethanol, benzyl alcohol, and parabens are non-ionic, so they do not have charge compatibility issues: they work perfectly fine alongside anionic components.
However, cationic preservatives (PHMB, benzalkonium chloride) are inactivated by anionic surfactants and anionic thickeners. This is not relevant for most anhydrous balms, but if your "balm" contains anionic components, keep this in mind.
8. What to choose in the end: effective systems
Basic choice: phenoxyethanol + ethylhexylglycerin
This is a workhorse. The value of this combination is not just that it is "two preservatives together," but in their clever synergy.
Ethylhexylglycerin is a mild surfactant (an amphiphilic glyceryl ether). It reduces interfacial tension, and in a balm, this provides two useful benefits:
- helps the oil system "reach" the introduced water droplet and wet it — that is, it delivers phenoxyethanol to where the microbes are, rather than leaving the oil and water on opposite sides;
- disrupts microbial cell membranes, making them more permeable, so that phenoxyethanol works more effectively at a lower dosage. This is true synergy, not just simple addition.
In other words, ethylhexylglycerin is a weak preservative on its own — it is a "delivery driver" that brings the active substance to its target.
Since phenoxyethanol's weak spot is mold and yeast, strengthen the antifungal flank with caprylyl glycol (it is oil-compatible and works against fungi). Many ready-made blends already include caprylyl glycol.
Natural / COSMOS route: benzyl alcohol + dehydroacetic acid
If you are working with certified natural cosmetics where phenoxyethanol is not permitted, the closest relative is benzyl alcohol. A useful property: it is amphiphilic — soluble in oil and at the same time ~4% soluble in water (log P ~1.1). This dual solubility is exactly what is needed: it distributes evenly throughout the oil base of a balm and delivers a useful dose to the introduced droplet — and it actually reaches the water even slightly better than phenoxyethanol.
Two nuances:
- it shares phenoxyethanol's weakness against mold — which is why it is used in a pair with an antifungal acid (the classic is benzyl alcohol + dehydroacetic acid, commercial blends like Geogard 221, Cosgard, Mikrokill);
- it is noticeably volatile (see point 6) — it can evaporate from an open jar. Therefore: use a sealed tube, add during the cooling phase.
And one more thing: benzyl alcohol is a declared allergen (fragrance allergen); it must always be listed in the INCI ingredients list if it exceeds the threshold. For those selling products, it is important to take this into account on the label.
9. The main point: a preservative is the last line of defense, not the first
Perhaps the most important conclusion of the entire article. For an anhydrous product with an "accidental" water risk, protection is built in the following order — in descending order of real benefit:
- Packaging that keeps water out — a tube, airless bottle, pump, or squeezable bottle instead of an open jar. This choice alone does more than any change of preservative.
- Antioxidant for oils — a separate task, but it is what keeps the base itself healthy.
- Instructions for use — dry hands or a spatula, do not store in the shower area.
- And only then — preservative, as a "safety net" for that drop that manages to get through anyway.
Do not hang all your protection on a single molecule.
10. The most important distinction: accidental contact vs. expected contact
This distinction defines the entire strategy.
If contact with water is accidental (an occasional wet finger) — everything described above applies: packaging plus a safety preservative.
But if contact with water is expected and regular — for example, a cleansing balm that is emulsified on the skin with wet hands every time, or a shower scrub — then this is no longer an "anhydrous product with low risk." In terms of behavior, it is a water-containing product, and it must be preserved fully, like an aqueous system, and verified with a challenge test like an aqueous system. Do not deceive yourself with the phrasing "but my product is anhydrous."
11. The truth about challenge testing for anhydrous products
The standard preservative efficacy test (ISO 11930) is designed for water-based products: they are inoculated with microbial cultures to observe the dynamics. It cannot be performed on a pure anhydrous product "as is" — there is nothing for the microbes to grow in.
There are modified protocols where the product is intentionally "challenged" by introducing water containing microorganisms — effectively simulating the exact scenario of misuse that we fear. This is a reasonable approach, but it is important to be honest about the limitations and not claim a dry balm is "tested" if such a test has not been conducted.
12. Practical details without which everything above won't work
- Dosage. Under-dosing is the #1 cause of failure. Follow the supplier's TDS range, do not measure "by eye."
- Addition during cooling. Due to both volatility and heat sensitivity, add the preservative at the end, at a low temperature.
- Production hygiene. A preservative protects against contamination during use, but it will not save a product made with dirty hands. Use clean equipment and clean raw materials.
- Chelating agents — case by case. In a purely anhydrous system, a chelating agent (EDTA, sodium phytate) has nothing to work in — it is needed in the water phase. However, if the product is intended to come into contact with water, a chelating agent in a water-containing version will enhance protection.
Summary
A true anhydrous balm is of no interest to microbes — it needs an antioxidant, not a preservative. A preservative is only justified against one scenario: water introduced by the consumer. In that case, what matters is not the "percentage in the formula," but the preservative's ability to reach the introduced droplet (water solubility), work in a neutral environment (pH independence), and not evaporate over time (low volatility). Based on these criteria, phenoxyethanol + ethylhexylglycerin (plus caprylyl glycol against mold) is a functional basic choice, while benzyl alcohol + dehydroacetic acid is a natural alternative. But the main protection is not a molecule, but packaging that keeps water out.
If you want to delve deeper into preservation, selecting antioxidants, and other key aspects of formulation — with practice and analysis of real formulas — read other materials on our blog and keep an eye out for new publications: https://.
