(M1530-07-47) Predicting the Tabletability of Powder Mixtures Lubricated with Magnesium Stearate

Purpose: Accurately predicting the tabletability, i.e., the relationship between tensile strength and compaction pressure, of powder mixtures could streamline tablet formulation development, saving both time and materials. ¹ A robust strategy for predicting the tabletability of binary mixtures has been developed based on the Vreeman-Sun tabletability equation and the power law mixing rule. ¹2 However, this method does not account for the effects of varying levels of magnesium stearate (MgSt), a widely used lubricant known to weaken tablet mechanical strength in a concentration-dependent manner. ³ This study aims to address this gap by developing a generalizable strategy for predicting the tabletability of powder mixtures containing any desired amount of MgSt, based on the tabletability of lubricated individual components obtained by interpolation.

Methods: The tabletability of MCC (Avicel PH102, d₅₀ = 100 µm) lubricated with 0.5%, 1%, 2%, 5%, and 10% of MgSt and LM (Pharmatose 350M, d₅₀ = 30 µm) lubricated with 0.5%, 1%, 2%, 3%, and 5% of MgSt was collected over the pressures up to 550 MPa. All powders were mixed for 2 min using a Turbula mixer (GlenMills Inc., Clifton, NJ) operated at 49 rpm. Each set of tabletability data was fitted to the Vreeman–Sun equation to obtain three tabletability parameters: 𝜎ₘₐₓ, 𝛼, and 𝛽. ² The relationship between each of these parameters and MgSt% was obtained by non-linear regression using the power law function, from which these tabletability parameters for powders with any desired MgSt% were predicted. Formulations consisting of 60 g MCC and 40 g LM with 0.5 g, 1 g, 2 g, and 5 g of MgSt were prepared to validate the predicted tabletability profile. All tablets were prepared using a compaction simulator (Styl’One, Evolution; Medelpharm, Beynost, France) with 8 mm flat-faced tooling, and tablet tensile strength was calculated from tablet dimensions and breaking force using the standard method. ³

Results: Hypothesis 1: Compared to MCC, LM is much less sensitive to MgSt-induced tabletability reduction as expected. ⁴ Therefore, we initially hypothesized that the reduction in tabletability of MgSt-lubricated mixtures is primarily driven by MgSt-MCC interactions. Thus, the tabletability of lubricated mixtures can be predicted from those of MgSt-MCC and pure LM by applying the power mixing rule. For example, a 1:1 MCC-LM mixture lubricated with x% MgSt is equivalent to a binary mixture of (MCC + 2x% MgSt) and LM in a 1:1 ratio. Different sets of 𝜎ₘₐₓ, 𝛼, and 𝛽, for MCC lubricated with varying concentrations of MgSt were obtained (Fig. 1a). Each of these tabletability parameters was correlated with MgSt levels using power law fitting (Fig. 1b, c, d), from which the tabletability parameters of MCC lubricated with any desired concentration of MgSt were determined by interpolation. The predicted tabletability parameters of lubricated MCC were then combined with those of pure LM to predict tabletability of the mixtures. However, the predicted tabletability profiles at the four MgSt levels deviated significantly from the corresponding experimental profiles (Fig. 2a), indicating flaws in this hypothesis. Hypothesis 2: Given MgSt coats the surfaces of both MCC and LM in their blends, we modified the hypothesis by taking the MCC lubrication level the same as that in the final blend, for tabletability predictions, while assuming LM tabletability remains insensitive to MgSt. For example, tabletability of 1:1 MCC-LM mixture lubricated with x% MgSt is predicted from a binary mixture of (MCC + x% MgSt) and LM in a 1:1 ratio. This approach improved the agreement between the predicted and experimental tabletability profiles (Fig. 2b). However, the remaining discrepancies suggest that further refinement of the prediction approach is needed. Hypothesis 3: We subsequently hypothesized that effect of MgSt on the tabletability of LM, although much less pronounced than that on MCC, cannot be ignored. Thus, predicting tabletability of a lubricated mixture needs to use both components lubricated with MgSt at the same level of the final blend. For example, tabletability of a 1:1 MCC-LM mixture lubricated with x% MgSt is predicted from a binary mixture of (MCC + x% MgSt) and (LM + x% MgSt) in a 1:1 ratio. To evaluate Hypothesis 3, tabletability parameters of lubricated LM were predicted using the same approach as that for MCC.  This refined approach results in excellent predictions across all four MgSt levels, with the experimental tabletability profiles falling within the 95% confidence interval of the predicted profiles (Figs. 2c, 3).

Conclusion: We developed a generalizable strategy to accurately predict the tabletability of MgSt-lubricated powder blends by first predicting the tabletability of the individual components lubricated at the target MgSt level. This approach enables accurate tabletability predictions for any given MgSt-lubricated LM-MCC mixture. More importantly, it lays the foundation for predicting the tabletability of realistic formulations, providing a promising framework for advancing the digital design of tablet formulations.

References: (1) Bhagali, V. G. Predicting Tabletability of Powder Mixtures. 2023.
(2) Vreeman, G.; Sun, C. C. A Powder Tabletability Equation. Powder Technol. 2022, 408, 117709.
(3) Fell, J.; Newton, J. Determination of Tablet Strength by the Diametral-Compression Test. J. Pharm. Sci. 1970, 59 (5), 688–691.
(4) Paul, S.; Sun, C. C. Lubrication with Magnesium Stearate Increases Tablet Brittleness. Powder Technol. 2017, 309, 126–132.

Acknowledgements: CCS thanks the National Science Foundation for support through the Industry University Collaborative Research Center (IUCRC) grant IIP-2137264, Center for Integrated Materials Science and Engineering for Pharmaceutical Products (CIMSEPP).


Figure 1. a) Tabletability profiles of MCC lubricated with varying concentrations of MgSt (0, 0.5%, 1%, 2%, 5%, and 10%), fitted with the Vreeman–Sun equation. Relationship between MgSt concentration and b) 𝜎ₘₐₓ, c) 𝛼, and d) 𝛽 for MCC.


Figure 2. a) Predicted (dashed line) vs. experimental (triangles) tabletability profiles based on a) Hypothesis 1, b) Hypothesis 2, c) Hypothesis 3.


Figure 3. Comparison between tabletability profiles predicted based on Hypothesis 3 and experimental profiles. All data points fall within the 95% confidence interval of predicted profiles (shaded region) for a) 0.5%, b) 1%, c) 2%, and 5% MgSt.