Key takeaways
- Most systems pair a primary preservative with a booster such as ethylhexylglycerin.
- pH, water activity and oil load decide how much preservative is actually active.
- A challenge test on the finished product is the only proof that a system works.
A preservation system is not a preservative. It is a combination of a primary agent, one or more boosters, a pH strategy and a pack decision that together keep a product free of microbial growth for its shelf life. Most failures in the category come from treating the preservative as a checkbox rather than as a system.
The system is designed against four variables: the water content, the pH, the oil and surfactant load that binds preservative, and the pack that determines how much air and finger contact the product sees. Change any one and the system needs re-testing.
1. The primary agent
Phenoxyethanol at 0.8 to 1.0 percent is the default for a neutral formula. Sodium benzoate at 0.1 to 0.5 percent suits a formula below pH 5. Benzyl alcohol at 0.5 to 1.0 percent is effective and odorous. Each primary agent has a spectrum gap, and the gap is what the booster fills.
2. Boosters
Ethylhexylglycerin at 0.05 to 0.1 percent improves the activity of phenoxyethanol and allows a lower dose. Caprylyl glycol, 1,2-hexanediol and pentylene glycol work similarly and add a light emollient feel. A system without a booster usually needs a higher primary dose to pass the same test.
3. Water activity
High glycerin or sugar content lowers water activity and slows microbial growth, which is why a honey-based product can pass with a lighter system. This is a real effect and it should be measured rather than assumed, because a glycerin dose below 20 percent does not change water activity much.
4. pH strategy
Weak-acid preservatives need a pH below their pKa to be active. Benzoate is largely inactive above pH 5; sorbate above pH 6.5. A formula designed at pH 5.5 with benzoate is a formula designed for failure, and the fix is either a lower pH or a different primary agent.
5. Partitioning
In an emulsion, preservative distributes between water and oil, and only the aqueous fraction is active. A 30 percent oil phase can bind a substantial share of a lipophilic preservative. This is why a formula that passes in a water blank fails in a rich cream at the same label dose.
6. The pack
An airless pump limits air ingress; a jar exposes the surface every time it is opened. A jar product needs a stronger system than a pump product with the same formula. Test the challenge in the actual pack, filled and closed, not in a beaker.
7. Challenge testing
ISO 11930 inoculates with S. aureus, E. coli, P. aeruginosa, C. albicans and A. brasiliensis and measures log reduction at 7, 14 and 28 days. Pass criteria are defined per product category. A product that passes at release but fails at twelve months has a stability problem, not a preservation problem.
8. Reformulation triggers
Any change to fragrance, active, oil load, pH or pack invalidates the previous test. Brands routinely change a fragrance for cost and skip the retest, which is how a stable product becomes a recall. Treat the challenge test as part of the change control process.
9. Natural systems
“Preservative-free” usually means a blend of glycols and organic acids that performs the same function under a different name. These systems are legitimate and often weaker, and they work best in low-pH, low-water-activity formulas. The label story improves; the test requirement does not.
10. Record keeping
The batch record should name the preservative lot, the pH at addition and the final pH, and the challenge test reference. When a complaint arrives two years later, that record is the only way to reconstruct what happened. It costs one line per batch.
11. Where it fits
Preservation runs across the cream range; the single-agent detail continues in the phenoxyethanol entry.
Reference: Preservation protocols and challenge test results, 2025
