A typical kitchen cleaner contains multiple components including surfactants, solvents, and other additives. These ingredients can interact with each other, with the soil to be cleaned, and with the surface to be cleaned, affecting formulation performance. This makes an a priori prediction of performance based on ingredients challenging. Additionally, the complex nature of the soils and impurities in formulation ingredients imposes limitations on computational chemistry approaches. To overcome some of these limitations, we take a machine learning approach. As surface tension plays an important role in the cleaning process, here we use Bayesian optimization to optimize surface tension of a model formulation having multiple surfactants. This approach may be applied to other cleaning factors to help develop and predict effective formulations.

Machine learning to better understand hard surface cleaning formulations

255th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.