"Preservative-free" is a thermodynamic claim, not a marketing one. Microorganisms do not respond to how much water a product contains - only to the energy status of that water, expressed as water activity aw. Two formulas with identical water content can sit on opposite sides of microbial safety; the number that decides it is aw.
The thresholds every claim leans on
| Organism group | Minimum aw for growth |
|---|---|
| Gram-negative bacteria | 0.91 |
| Gram-positive bacteria | 0.86 |
| Yeasts (practical limit) | 0.88 |
| Molds (practical limit) | 0.70-0.80 |
| Xerophilic molds | 0.62-0.65 |
| Absolute floor for any growth | 0.60 |
A conventional O/W emulsion sits at aw 0.95-0.98 - it supports everything on the list, which is why it needs a full preservative system and challenge testing. Below aw 0.70 mold growth stops; below 0.60 nothing cosmetic-relevant grows at all. ISO 29621:2017 uses exactly this logic to define microbiologically low-risk products, which can be exempted from full microbial testing in many markets. The threshold table above follows the Novasina cosmetics review and the underlying microbiology (Buechat 1983 adaptation).
The humectant arithmetic
- Small doses do almost nothing. At the 3-10% glycerin typical of emulsions, aw barely moves - the formula stays at 0.97+. Glycerin at leave-on dosages is a humectant, not a preservative.
- The math only works at high solids. Single-humectant suppression of aw to safe levels needs concentrations of 50% or more - unusable as a skin feel. That is why the workable designs are anhydrous or ultra-low-water systems built on a polyol matrix.
- Workable design point: glycerin + propanediol + 1,2-hexanediol combined at 40-60% in an anhydrous or waterless-format product compresses aw to roughly 0.50-0.60 - inside the absolute no-growth floor. Kerdudo et al. (2015) demonstrated glycerin-based aw control for "preservative-free" cosmetics, and the approach has since become the backbone of waterless bars and balm formats.
- Measure, do not estimate. Benchtop aw meters resolve to about +/-0.003; build an aw reading into the release protocol of every low-preservative SKU, and re-measure after accelerated humidity storage.
The chemical trade-off nobody quotes
Microbial safety improves monotonically as aw falls, but chemical stability does not. The 0.40-0.70 aw window is where lipid oxidation and hydrolysis rates peak - precisely the range most "waterless" products live in. Low-aw products still need antioxidant protection (tocopherol-class) and packaging that blocks moisture pickup: a waterless bar stored in a humid bathroom absorbs water from the air, its surface aw climbs, and mold finds the niche. aw < 0.60 is a maintained property, not a one-time certificate.
Where the multifunctional glycols fit
Glycols earn their place in low-aw systems twice over: they bind water (lowering aw) and they contribute antimicrobial boost at the formulation margins. The selection logic by carbon-chain position is in our multifunctional diols guide, and the comparative case for 1,2-hexanediol as a preservative booster is in 1,2-hexanediol vs phenoxyethanol.
1,2-Hexanediol (6920-22-5), 1,3-Butylene Glycol (107-88-0) and 1,2-Pentanediol (bio-based) are available from Shanghai Better Chemical with COA, TDS and samples on request.