Search for: All records

In porous hollow fiber membrane (HFM) based anti-solvent crystallization (AsCr) technique, the anti-solvent is injected from HFM bore into the active pharmaceutical ingredient (API) containing solution flowing in the shell side. Conventionally, shell-side feed solution flows along the HFM. In our compact module design, the shell-side liquid flows perpendicular to the HFMs; the mean flow length is ~ 3 cm reducing considerably the residence time (tres) which strongly influences crystal growth. We investigated AsCr of griseofulvin (GF) from its acetone solution using anti-solvent water injected through HFM pores. The range of tres was 5–30 s; the volume flow rate ratio (FRR) of acetone/water ranged between 0.35 and 1.66; griseofulvin concentration was varied between saturation (3.73 g/100 g acetone) to 0.4 g GF in 100 g of solution having 50 wt% acetone in the total solution. Low tres yielded nanocrystals < 100 nm. The average API crystal size (Davg) remained around 230 nm in a three hour- long experiment. At a fixed mass FRR of 1.34 ±0.1, Davg increased linearly from 48 to 280 nm as tres increased from 5 to 28 s. For two HFM modules in series with the nanocrystal suspension from one module fed to the next module independently fed with anti-solvent water, GF crystallization yields as high as 98 % were achieved. AsCr of L-glutamic acid in water with anti-solvent acetone in the same device illustrated continuous nanocrystal production and similar behaviors vis-`a-vis variations of tres and FRR. The cross-flow HFM based continuous AsCr technique can continuously produce nanocrystals of APIs. 

Abstract PurposePredicting powder blend flowability is necessary for pharmaceutical manufacturing but challenging and resource-intensive. The purpose was to develop machine learning (ML) models to help predict flowability across multiple flow categories, identify key predictive features, and arrive at formulations with improved flow properties. MethodsA dataset of 410 blends, composed of 9 active pharmaceutical ingredients (APIs) and 18 excipients with varying silica dry-coating parameters, was analyzed. Supervised ML models were trained to predict various flowability categories (very cohesive, cohesive, semi-cohesive, well-flowing, and free-flowing). Particle size, morphology, surface properties, and coating parameters were used as features. Classification algorithms, including Random Forest (RF) and Extreme Gradient Boosting (XGBoost), were evaluated. Unsupervised clustering identified natural groupings within flowability data. ResultsThe best-performing models achieved up to 85% accuracy for predicting flowability regimes of individual components and 87% for blends. Individual components generally showed higher accuracy than blends, except in the uncoated scenario with 2 flow regimes, where blends outperformed with 94.67%. SHapley Additive exPlanations (SHAP) and Feature Importance analysis indicated dry coating parameters as the most influential factors, followed by particle size and morphology. ML models effectively identified category transitions between flow regimes, offering insights into blend optimization. ConclusionIntegrating ML with mechanistic approaches effectively predicted powder blend flowability across diverse categories and elucidated feature-property relationships. These outcomes can facilitate the rational design of blends having enhanced flow properties at reduced experimental effort through judiciously selected dry coating of a blend constituent; making this approach promising for advancing pharmaceutical process and product development. 

Purpose: To investigate the effect of dry coating the amount and type of silica on powder flowability enhancement using a comprehensive set of 19 pharmaceutical powders having different sizes, surface roughness, morphology, and aspect ratios, as well as assess flow predictability via Bond number estimated using a mechanistic multi-asperity particle contact model. Method: Particle size, shape, density, surface energy and area, SEM-based morphology, and FFC were assessed for all powders. Hydrophobic (R972P) or hydrophilic (A200) nano-silica were dry coated for each powder at 25%, 50%, and 100% surface area coverage (SAC). Flow predictability was assessed via particle size and Bond number. Results: Nearly maximal flow enhancement, one or more flow category, was observed for all powders at 50% SAC of either type of silica, equivalent to 1 wt% or less for both the hydrophobic R972P or hydrophilic A200, while R972P generally performed slightly better. Silica amount as SAC better helped understand the relative performance. The power-law relation between FFC and Bond number was observed. Conclusion: Significant flow enhancements were achieved at 50% SAC, validating previous models. Most uncoated very cohesive powders improved by two flow categories, attaining easy flow. Flowability could not be predicted for both the uncoated and dry coated powders via particle size alone. Prediction was significantly better using Bond number computed via the mechanistic multi-asperity particle contact model accounting for the particle size, surface energy, roughness, and the amount and type of silica. The widely accepted 200 nm surface roughness was not valid for most pharmaceutical powders. 

The impact of mixer type and critical process parameters (CPPs) on critical quality attributes (CQAs), including the drug content uniformity (CU) of slurry-cast polymer films loaded with micro-sized poorly water-soluble drugs were investigated. Previously untested hypothesis was that the best mixer at suitable CPPs promotes uniform drug dispersion within film precursors leading to acceptable dried-film CU at low, ~0.6 wt% drug concentrations. Taguchi design was utilized to select the best of three mixers; low-shear impeller, high-shear planetary, and high-intensity vibrational, for dried-film drug concentration of ~23 wt%. As-received fenofibrate, a model poorly water-soluble drug (~6 µm) was directly mixed with the hydroxypropyl methylcellulose (HPMC) and glycerin aqueous solution. Impeller and planetary mixers yielded desirable film relative standard deviation (RSD), while vibrational mixer could not. For the lowest dried-film drug concentration of ~0.6 wt%, only planetary mixer yielded RSD <6%. The precursor drug homogeneity was a sufficient but not a necessary condition for achieving dried-film RSD <6%. Thus, proper selection of mixer and its CPPs assured desirable film CQAs. However, minor drug particle aggregation was identified via re-dispersion testing which also led to incomplete drug release.