Simulated ordered powder mixture

Abstract

A computer simulation of ordered powder mixing has been carried out in order to ascertain the properties of ordered mixtures independently of errors associated with sampling and assay. The effect of sample size on the standard deviation of sample concentration has been ascertained for a variety of simulated mixes of various carrier particle sizes and particle size distributions. Excepting the case where the carrier particles are monosized, an inverse relationship between sample standard deviation and the square-root of sample size is found in all cases. This is identical to the situation when a random powder mix is formed. Thus distinguishing between ordered and random mixtures must be based on other factors, i.e. interparticulate interaction, degree of homogeneity and segregation tendencies. The standard deviation sample size relationship is not, as previously thought, a useful criterion for ordered mixtures. Only when the carrier is monosized is the theoretical independent relationship of standard deviation with sample size observed as demonstrated with ideal ordered mixtures. The results showed that for real ordered powder mixtures, the carrier particle size should be as large as possible within other technological constraints and of narrow size range in order to prepare homogeneous dose forms.