(M0930-01-02) Predicting Compression Behaviors of Mixtures from Those of Individual Components

Purpose: The current approach to tablet formulation in the pharmaceutical industry largely follows empirical, design-of-experiments (DOE) based approaches for optimization. This approach is both time and materials intensive, making it unfit for tablet formulation development early in drug product development due to the limited availability of active pharmaceutical ingredients (API). In contrast, a materials science-based approach to tablet formulation development following quality-by-design (QbD) principals is fast and material-sparing. In this approach, the ability to predict the compression behavior of a mixture from that of the individual components and the composition of the mixture is necessary.¹ This work aims to quantitatively characterize the tabletability of common pharmaceutical excipients with different particle sizes and their mixtures to establish a mixing model that can predict the tabletability behaviors of mixtures from those of individual components.

Methods: Microcrystalline cellulose Avicel PH102 (MCC PH102), Pharmatose 350M and 110M (LM 350M and LM 110M), and mixtures of MCC PH102 with LM 350M and MCC PH102 with LM 110M in 20% (w/w) increments were compressed on a compaction simulator (Styl’One Evolution, MedelPharm, Beynost, France) simulating a Korsch XL100 compression cycle at 20 rpm with a 10 MPa precompression step. The main compaction pressure varied between 25 and 500 MPa. Round, flat-faced (11.28 mm in diameter) punches were used for all mixtures except 100% MCC, and 60%/40% and 80%/20% mixtures of MCC with LM, which were made with 6.0 mm round, flat-faced tooling to lower the forces required to break the tablets. External lubrication (magnesium stearate spray, Styl’One MIST) was used on the die wall and punch tips before each compression. Tablets were broken with a texture analyzer, and tensile strength was calculated using the recorded breaking force along with tablet dimensions measured with calipers. Tabletability data (tensile strength versus main compaction pressure) was fitted with the Vreeman-Sun equation² (eq. 1) by nonlinear regression. 𝜎=𝜎_maxexp(𝛼W(-exp(-P/𝛽-1))) (1). The three fitted parameters in this equation, 𝜎_max, 𝛼, and 𝛽, were evaluated as a function of composition (C). Comparison of each parameter (P) to a linear (P_1,2=P₁C₁+P₂C₂), power (P_1,2=P₁^C1P₂^C2), and inverse (P_1,2=(C₁/P₁+C₂/P₂)^-1) mixing rule was evaluated.

Results: The tabletability of the mixtures of MCC and LM varied between those of MCC and LM as compositions varied (Figure 1 a & b). The smaller LM (350M, d₅₀=30 µm) had a slightly higher tabletability than the larger LM (110M, d₅₀=130 µm). For both particle sizes of LM, the power mixing rule provided overall better predictions of all three tabletability parameters of the mixtures than linear and inverse mixing rules (Figure 2, a-f). Although deviations from the experimental tabletability parameters are observed, the predicted tabletability profiles using only the parameters from the pure components and the power mixing rule are in good agreement with the experimental data of all four mixtures (Figure 3 a & b), suggesting robustness of this approach. The generality of this finding in the MCC-LM systems requires further studies using other binary mixture systems, particularly hard materials mixed with other hard materials and soft materials mixed with other soft materials, as the mixing behaviors may depend on the deformation properties of the pure components. Additionally, separation of contributions from compactibility and compressibility components to tabletability by considering compact porosity will also be useful. Such detailed knowledge in mixing behaviors may help inform other unit operations, such as roller compaction, in the tablet formulation development process.

Conclusion: A power mixing model satisfactorily described all three tabletability parameters of mixtures from those of pure components for the MCC-LM system with two LM particle sizes. Tabletability profiles for mixtures obtained using the predicted parameters are in good agreement with the experimental data. Future work aims to examine the generality of this observation with additional materials and particle sizes.

References: 1. Ramaswamy, C. M., Varma, Y. B. G. & Venkateswarlu, D. Compaction of mixtures of materials. Chem. Eng. J. 1, 168–171 (1970).
2. Vreeman, G. & Sun, C. C. A powder tabletability equation. Powder Technol. 408, 117709 (2022).

Acknowledgements: This work was supported by the Center for Integrated Materials Science and Engineering for Pharmaceutical Products (CIMSEPP), which is funded by the National Science Foundation (grant IIP-2137264). GV thanks the National Science Foundation (IIP-1919037), the American Foundation for Pharmaceutical Education (2022 Dr. Paul B. Myrdal Memorial Pre-Doctoral Fellowship), and the UMN Department of Pharmaceutics (David and Marilyn Grant Fellowship in Physical Pharmacy, 2022-2023) for partial support.

Tabletability data fitted with the Vreeman-Sun equation for (a) MCC PH102, LM 350M (d50=30 µm), and their mixtures in 20% increments and (b) MCC PH102, LM 110M (d50=130 µm), and their mixtures in 20% increments.

Tabletability parameters, 𝜎max, 𝛼, and 𝛽 as a function of LM composition for mixtures of MCC PH102 with LM 350M (a-c) and LM 110M (d-f).

Comparison of predicted tabletability profiles (solid lines) based on the parameters obtained from the pure components (dashed lines) and the power mixing rule to experimental data (points) for (a) MCC PH102 and LM 350M mixtures and (b) MCC PH102 and LM 110M mixtures.