D90: The Strongest Contributor to Setting Time in Mineral Trioxide Aggregate and Portland Cement

•Commercial mineral trioxide aggregates (MTAs) and their setting times are compared.•The particle size distributions of commercial MTAs are assessed.•The correlation between MTA and particle size and particle size distribution is determined.•Linear relationships between setting time and particle size are seen with Portland cement and MTA.•The largest particles (d90) proportionally influence setting time more than the median (d50) or the smallest particles (d10).

IntroductionThe setting times of commercial mineral trioxide aggregate (MTA) and Portland cements vary. It was hypothesized that much of this variation was caused by differences in particle size distribution.MethodsTwo gram samples from 11 MTA-type cements were analyzed by laser diffraction to determine their particle size distributions characterized by their percentile equivalent diameters (the 10th percentile, the median, and the 90th percentile [d90], respectively). Setting time data were received from manufacturers who performed indentation setting time tests as specified by the standards relevant to dentistry, ISO 6786 (9 respondents) or ISO 9917.1 (1 respondent), or not divulged to the authors (1 respondent). In a parallel experiment, 6 samples of different size graded Portland cements were produced using the same cement clinker. The measurement of setting time for Portland cement pastes was performed using American Society for Testing and Materials C 191. Cumulative heat release was measured using isothermal calorimetry to assess the reactions occurring during the setting of these pastes. In all experiments, linear correlations were assessed between setting times, heat release, and the 3 particle size parameters.ResultsParticle size varied considerably among MTA cements. For MTA cements, d90 was the particle size characteristic showing the highest positive linear correlation with setting time (r = 0.538). For Portland cement, d90 gave an even higher linear correlation for the initial setting time (r = 0.804) and the final setting time (r = 0.873) and exhibited a strong negative linear correlation for cumulative heat release (r = 0.901).ConclusionsSmaller particle sizes result in faster setting times, with d90 (the largest particles) being most closely correlated with the setting times of the samples.