(T0930-11-65) Quantitative Prediction of Segregation Risk in Multicomponent Pharmaceutical Blends

Purpose: Segregation of particles in mixtures is a common problem in the pharmaceutical industry. This phenomenon risks the quality of the final product (tablet). Segregation largely depends on the physical properties of blends, such as particle size distribution (PSD), texture, flowability, shape and density. From QbD perspective, understanding a blend’s segregation potential relative to its physical properties allows one to decide the most suitable API physical properties and formulation composition in early stage development, thereby reducing the segregation risk at the late stages of the manufacturing line.

Methods: In this study, the effect of physico-chemical properties on the segregation potential of ternary pharmaceutical blends was investigated using the Jenike and Johanson Sifting Segregation Tester, a bench-scale device. To explore the materials properties influence on the segregation, especially the density and PSD effect, 5 materials were selected from the materials library. The experiments utilized acetaminophen (APAP) and sodium ibuprofen (Na-IBU) as active pharmaceutical ingredients (APIs), combined with excipients such as microcrystalline cellulose (MCC), mannitol (MN), flowLac (Lac), and starch. Twenty experiments were conducted, varying the combinations of APIs and excipients to maximize differences in physical properties among the mixtures. Several properties of the powder blends, including particle size distribution (PSD), flowability, density, and permeability, were assessed. After segregation test, particle size of blends was carried out on Sympatech-Rodos by laser diffraction method and Near Infrared (NIR) spectroscopy was employed as the analytical tool to determine the change in API concentration indicative of segregation risk.

Results: The segregation tendency was determined by the ratio of API concentration between the first and last samples in the segregation test. A careful selection of APIs and excipients led to apparent segregation phenomenon that ranged from 90% to 105%. Due to the varying excipient ratio (2 to 4) and API concentration (2% to 10%), the segregation tendency increased with excipient ratio but decressed with API concentration. To explore the intrinsic reasons for the segregation, a data-driven analysis regarding the material property of blends was conducted. It was observed that the segregation tendency correlated not only with particle size distribution but also strongly with the blend’s basic flow energy, arching dimension, and differences in API properties relative to the blend. The predictive capability of these correlations achieved an R-squared value of up to 0.92, incorporating the fraction of API’s properties -- f(API) = (API property * API%)/(Blend property).

Conclusion: In this study, the segregation tendency of ternary pharmaceutical blends was examined. The predictive model, developed based on various material properties, mechanistically supported the segregation results. It was revealed that, although a significant particle size difference between the API and other components is known to enhance segregation, the interactions between the API and excipients also played a crucial role in influencing the segregation risks.


Effect of drug load and excipient ratio on segregation tendency


Predictability of segregation index from material intrinsic properties.