Why not worst-case corrosion tests?

Written on: October 1, 2021 by W. Stephen Tait

Hello, everyone. Supply shortages, changing market demographics and faster-to-market pressures tend to make formula chemical compositions change and evolve over time. These pressures often make worst-case corrosion tests seem attractive for new products and derivative products (line extensions) in particular.

The reasoning for worst-case corrosion testing is often that a product that passes a worst-case test will be OK under normal conditions. Worst-case corrosion testing is also often viewed as a way to reduce the number of corrosion tests, number of variables in a corrosion test and time needed to complete corrosion tests.

However, worst-case testing often provides misleading results. For example, would you conduct a consumer fragrance-preference study with a group that only likes unfragranced products? (The worst-case scenario indicates that if this group likes it, everyone will like it!) Of course you wouldn’t.

The same is true for container corrosion testing. Results from corrosion test parameters  that don’t reflect actual use by the majority of consumers and results that don’t accurately predict spray package corrosion by commercial products lead to surprise corrosion.

















For example, you might have a product where the majority of consumers store packages in the upright position between uses with a small portion of consumers (e.g., 5%) who store packages on their sides in a dresser drawer or an athletic bag.

The corrosion of packages stored on their sides is typically different from the corrosion in packages stored in the upright position. Hence, conducting only corrosion tests with packages on their sides—as a worst-case test—often lead to inaccurate conclusions for corrosion of the majority of the packages that are stored in the upright position. In other words, worst-case testing for corrosion when packages are on their sides has a high likelihood of providing incorrect conclusions about corrosion (along with surprise corrosion) for the majority of the packages purchased by consumers.

Worst-case corrosion testing doesn’t work because package corrosion is a complex interaction between a large number of formula and package factors. Several of these factors are:

Formula factors
• Formula and contaminant water pH
• Types of formula ingredients
• Ingredient concentrations
• Water (both as part of the formula and as contamination of an anhydrous formula)
• The physical form of a product inside the spray package (e.g., water-out or oil-out emulsion, single phase or multiple phases)
• Dissolved oxygen concentration
• Carbon dioxide propellant
• Formula ingredients that are carboxylic acids
• Formula ingredients whose molecules have unsaturated bonds
• Surfactants/detergents
• Fragrance types
• Consumer perception enhancers (e.g., herbs and other botanicals)
• Corrosion inhibitors whose concentrations are above or below the optimum concentration range
• Copper ion contamination

Package factors
• Type of package (e.g., aerosol containers and bag-on-valve)
• Type of internal package surface treatment (e.g., packages with or without internal coatings)
• Internal container morphology/geometry (e.g., welds, double seams, valve clench)

As you can see, there are many corrosion-causing and corrosion contributing factors. In addition, there is also often synergy between two or more of these different factors that produces significantly more serious corrosion than the corrosion contributed to or caused by individual factors. In other words, a combination of two or more formula/package factors could lead to serious, unexpected corrosion with a very short package service lifetime.

Only worst-case corrosion testing on a base formula—and/or assuming a derivative formula is non-corrosive because its base formula is not corrosive—could also lead to the wrong conclusions with surprise corrosion in the marketplace. Consequently, I recommend avoiding worst-case corrosion testing for spray packaging with both new formulas and derivatives of base formulas.

The next edition of Corrosion Corner will provide a more detailed discussion on how the various formula and factors could contribute to, or cause, spray package corrosion.

Thanks for your interest and I’ll see you in November. Please contact me at 608-831-2076, rustdr@pairodocspro.com or from one of our two websites pairodocspro.com and aristartec.com. SPRAY