(T1030-11-65) Formulation in a Droplet – a Material Sparing Approach to Wet Granulation Formulation Screening

Purpose: Wet granulation proceeds through three rate processes: wetting and nucleation, growth, and breakage. Wetting and nucleation is the present focus, where we apply a micro-scale approach called the drop penetration test (DPT) (Hapgood et al., 2002, Wade et al., 2024), where a liquid droplet is measured as it penetrates a static powder bed. The present study applied DPT to pharmaceutically relevant formulations, piloted an image analysis technique for measuring dynamic contact angle and drop diameter, and characterized the granule compaction properties. Variables were: Drug:diluent ratio (0-50%w/w acetaminophen (APAP), 50-100%w/w microcrystalline cellulose PH102 (MCC)), Polymeric binder (hydroxypropyl cellulose (HPC) or polyvinylpyrrolidone (PVP)), Aqueous binder concentration (6 and 8%w/w HPC, and 25 and 35%w/w PVP), Pre-wetting level (0, 5, 10%w/w).

Methods: Binary mixtures of 0–50%w/w APAP and MCC were prepared. MCC was added to a food processor followed by aqueous pre-wetting at 0, 5, or 10%w/w, and mixed for 1min, APAP was incorporated and mixed for 2min. DPT was performed in a petri dish (84.4mm diameter, 13.687mm height, 76.57cm³ volume). The empty dish was weighed, formulation added, surface levelled with a spatula and a flat disk (126.7g, 68mm diameter) applied to create a consistent surface. 20-30 drops were applied using a 10mL syringe with 1” 20-gauge needle. Each droplet was recorded with a high-speed camera at 86fps. The dish was dried at 60°C for 1h and contents sieved to isolate granules. DPT was determined from the videos. A sub-set were selected for dynamic drop diameter and contact angle analysis. Sessile drop diameter was estimated using a two-step Harris Corner Detection-Gaussian method. The contact angle was measured using a texture analysis Gabor wavelet-based approach. The harvested DPT granules were combined and compressed into 12mm diameter tablets. Force-displacement profiles were generated and tablet weight and thickness recorded.

Results: Figure 1 shows that increasing APAP drug load resulted in longer penetration times, regardless of the binder type or concentration. This is likely due to the combination of poor APAP wettability, and smaller particle size which improves the packing efficiency and reduces pore space, combining to reduce penetration time. Conversely, pre-wetting of the powder bed decreased penetration time. When comparing the polymer type, HPC results in a longer drop penetration time than PVP, potentially due to the more hydrophilic nature of PVP relative to HPC. Similarly, as the binder viscosity is increased for both HPC and PVP, the droplet penetration time increases, though the magnitude of the increase is less for PVP than for HPC. Figure 2 illustrates dynamic contact angle and droplet diameter responses for experiments where PVP was used as the binder. In the 25% PVP case (Figure 2a), a lower viscosity meant that irrespective of the degree of pre-wetting, no significant difference in the drop penetration time (~3sec) was observed. Slight spreading followed by a constant drop diameter were observed, followed by a slight reduction in drop diameters. The effect of pre-wetting was clearly observed at higher viscosities with 35% PVP (Figure 2b), where 5% pre-wetting resulted in a significant reduction in drop penetration time, from ~15sec (dry) to ~7sec. Figure 3 presents the springback (elastic recovery) at the compression endpoint (~30MPa) modelled in terms of dried composition and DPT processing parameters. A statistically significant trend was seen in the binder type and concentration, pre-wetting relative to MCC, droplet-penetration time, and a marginal effect of APAP% concentration in the powder blend. A possible explanation is that prewetting plasticizes MCC, reducing the elastic response during compaction.

Conclusion: Overall, the drop penetration is positively affected by minimizing the APAP drug load due to its poorly wettability, the use of hydrophilic PVP polymer binder, pre-wetting the powder bed, and minimizing the viscosity of the binder liquid. Exploratory high-speed video analysis to characterize dynamic contact angle and drop diameter showed promise to shed light on the mechanism of droplet spreading and penetration mechanisms and may inform future quantitative studies to further aid formulation selection. Granule compaction analysis showed that increased APAP drug load resulted in increased springback, however increased pre-wetting and predicted DPT both resulted in reduced springback. Furthermore, the polymer binder type also influenced compaction with PVP producing granules with lower springback than HPC, though this may be due to the increased mass of polymer in the formulation compared to HPC, though these compaction differences may be obscured at greater industrial scale compaction stresses. These insights into minimizing drop penetration time provide the basis for a rational selection of drug load, polymer binder, binder concentration, and the use of pre-wetting, in early development of wet granulation formulations.

References: Hapgood, K. P., Litster, J. D., Biggs, S. R., & Howes, T. (2002). Drop Penetration into Porous Powder Beds. Journal of Colloid and Interface Science, 253(2), 353–366.
Medina Monsanto, P., Koleilat, L., Gupta, S., Butikofer, S., Mort, P. & Wade. J., (2024) Formulation in a drop – templated granulation and compaction analysis, Powder Technology, 438 119611.
Acknowledgements: This research was funded by Eli Lilly & Company


Figure 1. Sequence of penetration times: Prediction profiler for (a) 10% APAP, no pre-wetting, and 25% PVP binder solution; (b) 50% APAP, no pre-wetting, and 25% PVP; (c) 50% APAP, no pre-wetting, and 35% PVP binder solution, and (d) 50 % APAP, pre-wetting 20% of MCC in powder mix, and 35% PVP binder solution.


Figure 2. Effect of pre-wetting with (a) 25% PVP and (b) 35% PVP binder liquid: Comparison of average contact angle (solid lines corresponding to left y-axis) and droplet contact diameter (dashed lines corresponding to right y-axis) for dry (red), 5% pre-wetting (blue), and 10% pre-wetting cases (green)


Figure 3. Summary of tablet springback ANOVA (R2 0.71, RMSE 4.81).