3-PARAMETER DISTRIBUTION FUNCTION FIT TO GROWTH-RATE DISPERSION AMONGSMALL CRYSTALS

This paper considers the advantages of using a three-parameter fit to both log-normal and gamma distributions rather than the two parameter fit normally used for the statistical treatment of growth rate dispers ion among small crystals. It discusses the distribution functions, the procedure of searching for location parameters and the optimal number of bins for the best log-normal and gamma fits. As an example, the th ree-parameter fit is applied to results of the growth rate distributio ns of small paracetamol crystals. It is shown that the application of three-parameter distributions yields both a better fit to the experime ntal data than the two-parameter ones and the potential for a compound statistical treatment of growth, zero-growth and dissolution among sm all particles.