FRACTIONAL FACTORIAL-DESIGNS FOR OPTIMIZING EXPERIMENTAL CONDITIONS FOR HIESTANDS INDEXES OF TABLETING PERFORMANCE

The evaluation of drug substances and pharmaceutical excipients for th eir physico-mechanical properties is of prime importance in the develo pment of oral solid dosage forms. Some of these materials are brittle and consolidate by brittle fracture or fragmentation, some others are ductile and consolidate by plastic deformation, while others consolida te by both fragmentation and plastic flow. One method of characterizat ion is by determination of Hiestand's Indices of Tableting Performance (ITP), Bonding Index, Brittle Fracture Index and Strain Index, from m easurements of indentation hardness (P) and tensile strengths of large compacts with a hole (sigma(TO)) and without a hole in the center (si gma(T)). Three individual fractional factorial designs have been emplo yed for identifying experimental conditions that would maximize the di fferentiation between materials in terms of their ITP. Factors that ar e varied in the factorial designs include compact size, compact weight , compression pressure, indenter diameter, indenter release angle, sto rage temperature, storage time, strain rate during tensile break. Micr ocrystalline cellulose and acetaminophen were chosen as representative s of highly and poorly compatible materials, respectively. Factor leve ls were chosen to span the practical limitations of the experiments. T he Strain Index analysis indicates that the four main effects, indente r release angle, indenter diameter, storage temperature and compressio n pressure, and seven two-factor interactions, three of which involve the material characteristics, are significant. The Brittle Fracture In dex analysis indicates three main effects, material, compact size and storage time, and five two-factor interactions, two of which involve t he material properties, are significant. The analysis with porosity as a covariate makes no substantial reduction in the factors which influ ence the indices. Additional experiments are required to deconfound th e major effects, especially the 'material' effect. (C) 1998 Elsevier S cience S.A. All rights reserved.