Impact of Hot-Melt-Extrusion on Solid-State Properties of Pharmaceutical Polymers and Classification Using Hierarchical Cluster Analysis

• Main Authors: Ioannis Partheniadis, Miltiadis Toskas, Filippos-Michail Stavras, Georgios Menexes, Ioannis Nikolakakis
• Format: Article
• Language: English
• Published: MDPI AG 2020-09-01
• Series: Processes
• Subjects: factor analysis; classification; dendrograms; hot-melt extrusion; polymers; powders
• Online Access: https://www.mdpi.com/2227-9717/8/10/1208

The impact of hot-melt extrusion (HME) on the solid-state properties of four methacrylic (Eudragit^® L100-55, Eudragit^® EPO, Eudragit^® RSPO, Eudragit^® RLPO) and four polyvinyl (Kollidon^® VA64, Kollicoat^® IR, Kollidon^® SR, and Soluplus^®) polymers was studied. Overall, HME decreased Tg but increased electrostatic charge and surface free energy. Packing density decreased with electrostatic charge, whereas Carr’s and Hausner indices showed a peak curve dependency. Overall, HME reduced work of compaction (Wc), deformability (expressed as Heckel P_Y and Kawakita 1/b model parameters and as slope S′ of derivative force/displacement curve), and tablet strength (TS) but increased elastic recovery (ER). TS showed a better correlation with S′ than P_Y and 1/b. Principal component analysis (PCA) organized the data of neat and extruded polymers into three principal components explaining 72.45% of the variance. The first included Wc, S′ and TS with positive loadings expressing compaction, and ER with negative loading opposing compaction; the second included P_Y, 1/b, and surface free energy expressing interactivity with positive loadings opposing tap density or close packing. Hierarchical cluster analysis (HCA) assembled polymers of similar solid-state properties regardless of HME treatment into a major cluster with rescaled distance Cluster Combine Index (CCI) < 5 and several other weaker clusters. Polymers in the major cluster were: neat and extruded Eudragit^® RSPO, Kollicoat^® IR, Kollidon^® SR, Soluplus^®, and extruded Eudragit^® L100-55. It is suggested that PCA may be used to distinguish variables having similar or dissimilar activity, whereas HCA can be used to cluster polymers based on solid-state properties and pick exchangeable ones (e.g., for sustain release or dissolution improvement) when the need arises.