Case study 1 · Charge

Calcium bridging: a preservative blend that curdled an anionic cream

A facial cream that was stable at a low preservative dose curdled within hours at 40 °C when the dose was raised. Nothing in the oil phase or the emulsifier had changed; the cause was a divalent cation inside the preservative blend.

Published 6 November 2026 · Primary reference: Pensive Beauty® Nanoscience Labs (2026), Zenodo

The problem

An oil-in-water facial cream carried roughly 13% oil plus 5% shea butter, 0.2% xanthan gum and an emulsifier system built from glyceryl stearate, cetearyl alcohol and sodium stearoyl lactylate. The lactylate is the anionic component: its carboxylate head groups give the droplet surfaces a negative charge.

Preservation came from a ready-made blend of gluconolactone, sodium benzoate and calcium gluconate. At 0.85% of the blend the cream was stable. When the level was raised to 1.2–1.5%, to give more margin in preservative-efficacy testing, the cream developed a cheese-like curd, white specks and visible separation within hours at 40 °C.

Raising the dose scales every component of the blend, calcium included. Going from 0.85% to 1.2% is a factor of about 1.41 (1.2 ÷ 0.85); going to 1.5% is a factor of about 1.76. The cream therefore received roughly 40–75% more calcium.

What an HLB-based reading checks

An HLB-based review asks whether the emulsifier blend matches the required HLB of the oil phase. Between the stable and the unstable versions, neither the oil phase nor the emulsifiers changed, so the arithmetic returns the same answer for both. HLB describes the interface as a balance between hydrophilic and lipophilic character. It has no term for a divalent ion arriving inside a preservative, so this reading has nothing to flag. That is a limit of scope, not an error in the calculation.

What a placement (TDA Logic) reading asks

The placement checklist asks for the charge of every ingredient at the finished pH, and it reads supplier blends line by line rather than as a single trade name [1]. Applied to this conventional cream, three of its questions point at the failure.

Is there a multivalent cation facing an anionic surface?

Yes. Ca²⁺ can bind two carboxylate head groups at once, including head groups on neighbouring droplets. That bridges droplets together, which is the curd. The Schulze–Hardy rule describes how steeply coagulating power rises with counter-ion valence; in the idealised high-potential limit of DLVO theory the critical coagulation concentration scales as z⁻⁶, so a divalent ion is about 64 times as effective as a monovalent one (2⁶ = 64) [2], [3]. Calcium stearoyl lactylate is also far less water-dispersible than the sodium salt, which accounts for the white specks. Heat speeds ion exchange and droplet encounters, which is why 40 °C exposed the problem in hours.

Under TDA Logic, multivalent metal ions are minimised or excluded whatever their destination, because they interact directly with an anionic surface. In this cream, that single check would have marked the calcium gluconate at the design stage.

What does the ionic-strength budget contain?

Every salt counts, including those supplied inside a blend: calcium gluconate and sodium benzoate here. Raising the blend dose raises the ionic load at the same time as the calcium.

What does each ingredient become over time?

A lactone counts as the acid it hydrolyses to. Gluconolactone opens slowly in water to gluconic acid (pK_a ≈ 3.9), so the pH drifts downward during storage. A falling pH protonates part of the lactylate head groups and changes the charge the interface carries, adding a slower second source of drift.

Practical fix

Chelate the calcium. Disodium EDTA or sodium phytate at 0.1–0.2%, then re-run the 40 °C bench test. Chelation works mole for mole (EDTA binds calcium 1:1), so ask the supplier for the calcium gluconate content of the blend and check that the chelator dose covers it. Sodium phytate is itself highly charged and adds to the ionic load.
Or remove the salt. Move to a nonionic, salt-free preservative system, such as phenoxyethanol with ethylhexylglycerin, or caprylhydroxamic acid with glyceryl caprylate, and confirm by preservative-efficacy testing. Caprylhydroxamic acid binds iron and can develop colour with trace iron, so keep a chelator in place.
Or change the interface. The public thread records that the formula was moved to a nonionic glucoside emulsifier and then passed microbial testing.
Track pH over storage while gluconolactone remains in the formula.

Takeaway

A multivalent ion in a supplier blend is a charge decision, not just a preservative decision. HLB arithmetic has no term for it; the placement checklist asks the question that catches it, because it reads every line for charge at the finished pH.

Discussed publicly

The formula problem behind this analysis was discussed in a public forum thread. See In the field for the link and context. Other cases: case-study index.

References

Pensive Beauty® Nanoscience Labs (2026). TDA Logic: The Tri-Domain Placement Logic Succeeding 1940s HLB Logic in Nanoscale Cosmetic Formulation (Version 1.0) [Preprint]. Zenodo. doi:10.5281/zenodo.23384911
Verwey EJW, Overbeek JThG (1948). Theory of the Stability of Lyophobic Colloids. Elsevier.
Israelachvili JN (2011). Intermolecular and Surface Forces, 3rd ed. Academic Press.