Classification of flotation frothers – a statistical approach

Abstract This paper presents a statistical study to classify non-ionic frothers based on seven frother characteristics, molecular weight (MW), hydrophile–lipophile balance (HLB), critical coalescence concentration (CCC), the minimum sauter mean diameter (DL), dynamic foam stability index (DFI), static foam stability index (SFI), and decay rate index (DRI). A principal component analysis (PCA) was conducted to transform the seven frother characteristics into uncorrelated variables called principal components. The first principal component already explained 64.4% of the variance, and the addition of the second and third principal components explained 92.1% of the variance. A hierarchical cluster analysis (HCA) was conducted using the first three principal components. The Euclidean distance, which is the square distance between the standardised coordinates of two frothers in a generalised parameter space, was calculated for all frothers. At a Euclidean distance of 4, frothers were grouped into four groups, in which Group 1 consisted of the most powerful frothers that produced stable foams and Group 4 was composed of weakest frothers.