A New Powder, the Same Formula? Not Necessarily

What does an equal replacement actually preserve?

One powder leaves the formula, another takes its place, and the mass stays unchanged. The batch may look familiar during mixing yet feel different when spread. It is tempting to start by adjusting viscosity. But first ask: which performance are we trying to recover?

Liquid retention, response to deformation and application behaviour are separate questions. The three studies below do not rank the same materials in the same formula. Their value is in showing why one matching number may leave several important questions unanswered.

More water retained does not mean a better substitute

The 2025 talc-substitution study examined stiff zinc oxide pastes, not lightweight cosmetic emulsions. Its experimental composition was 25% zinc oxide, 25% talc or starch and 50% petrolatum. These are study conditions, not recommended use levels. Whole pastes were mixed with a liquid, centrifuged and weighed; the result is not a universal absorption value for dry powder or a skin oil-control endpoint.

Tapioca-starch paste retained the most water and talc paste the least. With paraffin oil, talc paste exceeded all tested starch pastes; with olive oil, the statistically significant pairwise comparisons were not identical. A change of liquid therefore changes the comparison that can be made. Muzi's first prediction is conditional: in a target formulation, differences attributed to powder identity may depend on which liquid and test conditions are used. That prediction remains to be tested.

Would restoring viscosity restore the formula?

The paste study also measured oscillatory rheology. Continuous-shear testing could not be completed because of sample climbing. Complex viscosity from an oscillatory experiment cannot simply be renamed steady-shear viscosity, and a stronger solid-like response does not by itself establish easier spreading or shelf stability.

A 2025 spheroidal-cellulose cream study separately used a rotor viscosity measurement at 22 degrees Celsius and 10 rpm, probe texture testing and a sensory group of ten trained women aged 25-45. The speed is not a transferable shear rate. Nor can a single rotor reading rank materials against the paste study's oscillatory data. The paper also contains unresolved inconsistencies in a viscosity increment's units and the interpretation of signed adhesion values. Muzi does not use those passages to claim a quantitative or sensory advantage. What remains useful is the separation of measurement methods.

A particle can take on a different job in an emulsion

The 2019 cork-based Pickering study brings the oil-water environment into the discussion. The authors measured wetting of powder-coated plates and varied particle amount and oil fraction while examining droplet size. Wetting in a single liquid on a powder layer is not a direct measurement of an individual particle's three-phase contact angle inside a formulation.

The reported water contact angle led the authors to expect water-in-oil emulsions, but they reported oil-in-water emulsions. They proposed high water-phase content and phase inversion as an explanation; the complete inversion pathway was not directly established by these observations. Three batches followed at 0, 14 and 30 days also do not establish a general shelf life. This study adds a question about particle role and phase conditions. It does not show that the starches or cellulose in the other two studies stabilised an interface in the same way.

What would distinguish the competing explanations?

The strongest alternative to an interface-led explanation is simpler: much of the change may come from bulk mechanics and dispersion. If matching those conditions restores the chosen application endpoint, the bulk explanation gains support. If a reproducible difference remains and varies with liquid or phase conditions, an additional explanation becomes worth testing. A residual difference alone does not prove particle adsorption at an interface.

This is a comparison of evidence weights, not a claim of direct disagreement among the papers. The paste study supplies system-specific retention and rheology observations. The cellulose study helps distinguish measurement endpoints, subject to its unresolved reporting issues. The cork study supplies an example where wetting expectations and the reported emulsion type do not coincide. None performs the matched experiment needed to choose between Muzi's two explanations.

Which predictions should a formulation team test first?

Predict a condition-dependent result, not an industry revolution. First, changing the specified liquid may change the apparent substitution advantage. Second, matching structure and dispersion may reduce a difference in a predeclared application endpoint. Third, any remaining difference must survive independent preparation before it can guide a substitution decision.

The practical change, if these predictions hold, would be to specify what a replacement must retain rather than treating equal addition or one viscosity reading as equivalence. These are candidate decision rules derived from cross-study reasoning. They are not reported experimental findings, validated commercial specifications or forecasts of market adoption.

Four proposed experiments, four ways to change our mind

Compare liquid conditions: retain the original formula control, change only the predefined powder and liquid factors, and standardise mixing, centrifugation and weighing. Record whole-formula retention and uncertainty across independent preparations. Reversal with liquid conditions would support a context-dependent comparison; reproducible equivalence across the tested liquids would narrow that prediction. Neither outcome establishes skin oil control.

Match bulk structure: compare an equal-mass substitution with a second, deliberately adjusted variant while controlling dispersion and measuring a relevant flow or oscillatory protocol. Separately measure the intended spreading or application endpoint. Convergence after matching would strengthen the bulk explanation; a reproducible residual difference would justify testing additional variables, not automatically prove an interface mechanism.

Test particle role in the actual emulsion: use matched base and processing controls, confirm emulsion type by complementary methods, and follow droplet distributions and particle location with methods capable of answering that question. Failure to detect the proposed location, or unchanged behaviour in an appropriate particle-free control, would weaken an interface-led interpretation. A powder-layer wetting test alone cannot settle it.

Test independent preparation and time: repeat the selected comparison in independent batches with predefined sampling times, storage conditions and acceptance limits for the chosen endpoints. A batch-dependent reversal or drift outside the predefined limits would undermine the proposed substitution rule. These experiments have not been performed by Muzi. If a powder carries an active material, constituent identity, reliable quantification, effective amount, batch consistency, retention after formulation and dose-response evidence require their own review.