Tablet strength is determined by inter-particle bonding strength and inter-particle bonding area. In a mixture, the type and prevalence of interparticle contacts between different materials can vary significantly depending on the particle size of the constituents while mass or volume contributions remain unchanged. Accordingly, mixture models which use mass or volume as a weighting term to determine the strength of pharmaceutical powders are fundamentally deficient. To address this deficiency, a particle size dependent tabletability mixture model was derived from consideration of inter-particle contact. The model was applied to binary mixtures of common pharmaceutical excipients and for binary mixtures of acetaminophen which was used as a model pharmaceutical active. The effects of particle size on tabletability were demonstrated and implications for tablet development were discussed. Interestingly, the particle size dependent mixture model reduces to a volume weighted mixture model for materials of equivalent particle size. Exploring this relationship between the two, it was shown that a volume weighted approach may be inappropriate under some circumstances not uncommonly encountered in tablet development.

Fluidized bed coating was investigated as a method to coat aqueous suspensions of nanocellulose onto acetaminophen (APAP) for the purpose of tabletability enhancement. Effects of suspension concentration and fluidized bed processing conditions were evaluated. Nanocellulose coatings of up to 2.0% w/w were achieved. This study found that the nanocellulose coating significantly improved the tabletability of APAP. A 2.0% w/w coating increased tablet tensile strength by nearly 200%. With respect to tablet formulation, it was found that APAP coated with 2.0% w/w nanocellulose allowed a 60% reduction in filler to produce tablets acceptable for manufacturing compared to APAP without nanocellulose. The effectiveness of fluidized bed coating was also compared to a twin-screw extrusion process which was previously used to coat nanocellulose onto APAP. It was concluded that process conditions and the process method had no effect on the extent of tabletability enhancement provided by nanocellulose. Improvement to tabletability was only dependent on the concentration of nanocellulose coated. Fluidized bed coating may be more advantageous than twin-screw processing due to the inherent ability of the former to better control liquid addition and removal. Accordingly, fluidized bed coating was able to achieve a higher coating concentration than the twin-screw process. Overall, the use of small amounts of nanocellulose to substantially improve tabletability of APAP and to reduce the need for filler excipients is evident.

Nanocellulose, a sub-micron cellulose powder, was investigated as a potential filler excipient to enhance the tabletability of acetaminophen (APAP). Due to its high specific surface area, nanocellulose was expected to outperform the tabletability enhancement provided by microcrystalline cellulose (MCC), a common tablet filler used in pharmaceutical tablets. Results showed that nanocellulose is inferior to MCC when used as a conventional filler. This is due to the difficulty in deaggregating and dispersing aggregates of nanocellulose using a typical dry blending process. To improve its effectiveness, nanocellulose prepared as a suspension was blended or "co-processed" with APAP using a twin-screw process. Microscopy images show that the twin-screw process coats the APAP with nanocellulose. The tabletability of APAP was significantly improved by low concentrations of nanocellulose up to 1.10% w/w. Such remarkable improvement allowed acetaminophen to be processed into tablets without any additional excipients. This study shows that nanocellulose can be used as a highly functional additive to enhance tabletability at low concentrations.

The adhesion between powders and surfaces is influenced by their topography. We investigate the relationships between the topography on stainless steel, the humidity, and the adhesion of three pharmaceutical powders. These relationships are captured within 'effective' work of adhesion parameters tuned using the enhanced centrifuge method (ECM). The parameters describe the effects of the size, shape, humidity, and topography on the adhesion. The powders' ability to nestle into features on the steel drove the work of adhesion. When the powders absorbed moisture, the balance of moisture-driven deformation and increased mass dictated the powder removal from the surface. The irregularly shaped hypromellose (HPMC) adhered more strongly to the stainless steel at low humidity than at high. The lactose monohydrate and ABT-089 showed the opposite behavior. The HPMC adhered more strongly to the rough steel than to the smooth or untreated steel. The lactose monohydrate and ABT-089 again showed the opposite behavior.

Glidant used for the purpose of powder flowability enhancement is well known within the pharmaceutical industry to improve tablet manufacturing. Despite the widespread use of glidant for this purpose, the effect of glidant on the effect of tableting behavior is not well studied. To address this deficiency, the effect on glidant on the tabletting behavior of seven common excipients was investigated. Tabletability, compressibility, compactability, and tablet expansion were studied. It was shown that glidant increased the tabletability for excipients known to exhibit plastic behavior. Based on compressibility and compactability, it was inferred that an increase in total bonding strength due to the presence of glidant was responsible for the improvement in tabletability. Results suggest this may be due to an increase in inter-particle bonding area. Conversely, glidants did not affect the tabletability of brittle excipients. The effect of glidant on die-fill bulk density was also studied to examine the impact associated with air entrapment and tablet expansion. While glidant significantly increased the die-fill powder density, this did not have an observable effect on tablet expansion or any other tabletting behavior examined.

The conical screen mill (comill) is investigated as a dry-coating process for flowability and bulk density enhancement of pharmaceutical powders. In this study, the effectiveness of the comill is improved by using modified screens with reduced open area. In comparison to the screens provided by the comill manufacturer, the modified screens increase mean residence time of the process and improve the extent of flowability and bulk density enhancement. The effectiveness of the comill as a dry-coating process is demonstrated using Avicel PH 105, a fine grade of microcrystalline cellulose, as a model cohesive powder. The process is evaluated thoroughly using a lab scale comill and scalability is demonstrated using a manufacturing scale model. The use of the modified screens is also compared against the so-called "multi-pass" approach in which material is passed through the comill, collected, and passed through once or several times. While the "multi-pass" approach is offered as a simple method to increase mean residence time and to improve process effectiveness, the use of the modified screens is shown to be the superior approach. Due to the ubiquitous use of the comill and the improvement in effectiveness attained in this study, dry-coating is shown to be a practical and readily implemented process for the pharmaceutical industry.

Most challenges during the development of solid dosage forms are related to the impact of any variations in raw material properties, batch size, or equipment scales on the product quality and the control of the manufacturing process. With the ever pertinent restrictions on time and resource availability versus heightened expectations to develop, optimize, and troubleshoot manufacturing processes, targeted and robust science-based process modeling platforms are essential. This review focuses on the modeling of unit operations and practices involved in batch manufacturing of solid dosage forms by direct compaction. An effort is made to highlight the key advances in the past five years, and to propose potentially beneficial future study directions.

The ultimate goal of the pharmaceutical industry has always been bringing safe and efficacious drug therapies to market as soon as possible and maintaining its robust supply for patients. Owing to key progress in the related sciences and engineering, the pace and innovative pathways built to achieve this goal continue to advance. Nevertheless, material and manufacturing costs are still significant for solving complex technical challenges during such fast-tracked science-based development as well as post-commercialization improvements. Process models that accurately incorporate the necessary detail of the underpinning science(s) are key tools to ensure that development and investigational studies are targeted, rapid, and effective. It is our pleasure, therefore, to present this special issue on the theme of Pharmaceutical process modeling. Part I of the issue presents contributions that address multiple unit operations, from a holistic product and process perspective. Parts II and III present contributions that describe advances in modeling approaches for drug substance and drug product processes, respectively. Overall, these contributions highlight the remarkable capabilities of the modeling approaches and tools used to impact the development of drug products. The issue is multi-disciplinary in pharmaceutical sciences and engineering, with contributions originating around the globe from academia, industry, and regulatory agency. We thank the authors for their contributions and the reviewers for their comments that have helped the scientific quality of these publications. We thank the Editor-in-Chief for making this issue possible. Considering the unprecedented way all have been impacted by COVID-19, we can’t thank enough everyone involved with this special issue. We hope this compilation helps the readers to advance their manufacturing processes via modeling, and move a step closer towards true process control with reduction in material and manufacturing costs.

This study investigated the Comil as a dry-coating process to enhance the flowability of pharmaceutical powders. As a major novelty, the mean residence time was measured and the effect of processing conditions on the extent of flowability enhancement was evaluated. As opposed to previous work which used a "multi-pass" approach to increase residence time and to improve the effectiveness of the Comil, this study reduces the open area of the screen. This simple modification to the screen increases residence time and allowed the Comil to improve flowability of a model cohesive powder, Avicel PH 105, as effectively as the LabRAM blender which was used as a benchmarking dry-coating device. The comilling process was applied to four fine and poorly flowing APIs which were subsequently used in high drug loading (50% w/w) formulations. In comparison to conventional blending, it was shown that comilling can significantly enhance the manufacturability of the formulations.

The effect of particle surface area and adhesion force on sticking behavior for a model pharmaceutical blend was studied. Various lots of an active pharmaceutical ingredient (API) differing in particle size distribution and surface area were blended with commercial grades of microcrystalline cellulose ranging in size from 20 to 110 μm. A dry-coating technique was also used to modify the surface of microcrystalline cellulose to reduce its adhesion force. This allowed study of sticking behavior due to effects associated with particle adhesion force independently from effects associated with surface area. Using a removable-tip experiment to quantify the mass of adhered material to a tablet punch, this study concludes that both particle surface area and adhesion force significantly affect sticking behavior. Tablets with higher tensile strength comprised of API with lower surface area relative to the excipient surface area resulted in less sticking. This study found that the difference between the tablet tensile strength of the blend and that of the API normalized by the surface area fraction of API in the blend correlates well with the rate of mass adhered to the punch. This quantity, referred to as the sticking index, can be used to assess sticking propensity for a pharmaceutical blend.

The effect of particle size on the dissolution behavior of the particles of amorphous solid dispersions (ASDs) of griseofulvin (GF), with 0%-50% Kollidon® VA 64 as a crystallization inhibitor is investigated. Both the final dissolved GF concentration and the dissolution rate of GF ASDs were found to be inversely proportional to the particle size. The solution concentrations for the smallest (45-75 μm) size group with different polymer loadings were significantly higher than those for the largest (250-355 μm) group regardless of the initial GF amount. Specifically, the dissolution rate of GF ASDs with 50% polymer loading for the finest group was 2.7 times higher than for the largest group under supersaturating conditions. The rates of dissolution and recrystallization were assessed through surface concentration (Cs) and Avrami recrystallization rate kinetics, where the solid-state recrystallization was confirmed using Raman spectroscopy. Outcomes indicated that particle size reduction enhanced ASD drug loading by reducing the amount of polymer necessary as finest size ASDs initially dissolve faster, negating their higher recrystallization rate. Kollidon® VA 64 at 30% loading was sufficient to inhibit the GF recrystallization. Overall, the combination of particle size reduction and recrystallization inhibition is effective for improved dissolution behavior of GF ASDs.

This study investigated the effect of material properties, primarily particle size and surface energy, on the effectiveness of glidants used for the purpose of flowability enhancement. Three pharmaceutical grade glidants (Aerosil 200, Aerosil R972, and Cab-O-Sil M5P) were evaluated and blended with various pharmaceutical actives as well as cohesive excipients common to capsule and tablet formulation. Flowability enhancement was characterized by the flow function coefficient (ff c ). An industry-relevant mixer (Turbula mixer) and a highly efficient and effective mixer (LabRAM vibratory mixer) were used to further understand the effect of material properties on glidant effectiveness. While concepts of inter-particle cohesion and interaction strength were applied to evaluate their usefulness in understanding and predicting flowability enhancement, theoretical expectations did not fully explain the behavior of all three glidants. However, the study suggests that the low surface energy and optimal particle size of Aerosil R972 relative to the other glidants results in lower inter-particle force and consequently better flowability. Aerosil R972 was also shown to be more effectively utilized in the Turbula mixing process particularly for larger (d50 > 40 μm) and less cohesive (ff c  > 3) materials. This may be due to its lower surface energy and hydrophobic surface which allows it to disperse easily. Overall, this study provides useful insight into the material properties which influence the effectiveness of glidants used in formulation development.

This study modeled the particle size distribution (PSD) of pharmaceutical extrudates after milling by developing a so-called time-discrete population balance model (PBM). The PBM, which models size reduction as a series of breakage events, was formulated so that the model parameters separate the effect of material properties and milling process conditions. Because of this novel aspect, the PBM should have excellent predictive capability with specific applications in technology transfer and scale-up. To investigate this application, copovidone extrudate produced by the hot-melt extrusion process was milled using a lab-scale continuous impact mill (Fitz Mill). The effect of impeller speed and classification screen size on PSD of the extrudate was investigated. The PBM with parameters obtained by fitting lab-scale PSD data was then applied to model the PSD of the extrudate following milling by a pilot-scale continuous impact mill (Hosokawa mill). The study found that the parameters determined at the lab-scale can be used to model PSDs at the pilot-scale and may be generally applied to similar classification-type impact mills. Since technology transfer and scale-up can be material and time consuming, this approach may offer significant benefits to the pharmaceutical industry for the development of milling processes. (C) 2018 The Society of Powder Technology Japan. Published by Elsevier B.V. and The Society of Powder Technology Japan. All rights reserved.

A multi-scale modeling approach is formulated which combines the use of the discrete element method (DEM) and population balance model (PBM) to simulate the evolution of particle size distribution (PSD) in dry milling. As a major novelty, a pseudo-coupled DEM–non-linear PBM approach is proposed, in which the parameters of a non-linear PBM were calibrated using the DEM simulations of the microdynamic environment in a mill by using the DEM intermittently. To demonstrate the application of this approach, breakage dynamics of silica particles in a vibrating cylindrical vessel containing an alumina grinding ball was simulated numerically. DEM simulations were performed to obtain the collision frequency and impact energy spectra, calculate the first-order breakage rate parameters from short milling of mono-sized feeds, and calibrate the parameters of the effectiveness factor of the non-linear PBM. As another novelty, the form of the effectiveness factor chosen accounted for both acceleration effects and deceleration effects simultaneously, which have been observed in various experimental milling studies. By incorporating DEM input, the non-linear PBM with the effectiveness factor was used to simulate the evolution of the PSD. This study demonstrates that a non-linear PBM, whose parameters were calibrated using DEM at early milling times, could predict the PSD evolution during prolonged milling and potentially obviate the need for using computationally expensive DEM simulations for the whole milling duration.

In a recent study, it was demonstrated that improving flow of a model poorly flowing and poorly compactable drug substance, acetaminophen, via dry coating while using fine excipients, may promote direct compression. To validate this novel strategy, particularly for high drug-loading formulations, this study investigates the effect of microcrystalline cellulose (MCC) particle size and dry coating on powder tabletability and flowability. It was determined that blends containing fine-sized MCC (20 μm) resulted in the highest tablet tensile strength and best tabletability because it provides higher interparticle contact area compared with coarse-sized MCC. Although tabletability can be improved using fine MCC, flowability is poor but can be improved through dry coating, a process that coats glidants (nano-sized silica) onto particle surfaces. To retain the tabletability, which was adversely affected because of the presence of glidant in the blend, while simultaneously enhancing flowability via dry coating, separately blending the drug substance with glidant is shown to be the best method of processing. The combined use of fine excipients and selective dry coating offers a novel and advantageous formulation strategy in comparison with the conventional use of coarse excipients, such as Avicel PH 102, that have been designed and marketed for direct compression.

This study investigates the relationship between particle interactions dominated by the cohesive van der Waals force and powder flowability for materials commonly used by the pharmaceutical industry in oral solid dosage formulation. This study first sought to correlate the granular Bond number, defined as the ratio of the inter-particle cohesion force to particle weight, to the flow function coefficient, a metric commonly used to assess powder flowability. However, the granular Bond number which strictly quantifies inter-particle cohesiveness was found to correlate poorly with powder flowability due to the complexity associated with particle assemblies. To account for the multitude of interactions between particles of different sizes within a powder and to more precisely predict bulk powder behavior, a population-dependent granular Bond number was proposed. The population-dependent granular Bond number which explicitly accounts for particle size distribution and described herein as a quantification of powder cohesiveness (instead of inter-particle cohesiveness) was shown to correlate well with the flow function coefficient for a wide variety of materials including four active pharmaceutical ingredients (APIs) and fourteen common pharmaceutical excipients. Due to the success of the population-dependent granular Bond number, it was extended to predict the flowability of powder blends. This so-called population-dependent multi-component granular Bond number takes into account relevant material properties and particle interactions and was used to predict the flowability of 6-component powder blends containing acetaminophen as a model cohesive active pharmaceutical ingredient. Prediction of bulk powder behavior from individual material properties as accomplished here may be highly useful in formulation development.

A novel solvent-less dry-polymer coating process employing high-intensity vibrations avoiding the use of liquid plasticizers, solvents, binders, and heat treatments is utilized for the purpose of controlled release. The main hypothesis is that such process having highly controllable processing intensity and time may be effective for coating particularly fine particles, 100 μm and smaller via exploiting particle interactions between polymers and substrates in the dry state, while avoiding breakage yet achieving conformal coating. The method utilizes vibratory mixing to first layer micronized polymer onto active pharmaceutical ingredient (API) particles by virtue of van der Waals forces and to subsequently mechanically deform the polymer into a continuous film. As a practical example, ascorbic acid and ibuprofen microparticles, 50-500 μm, are coated with the polymers polyethylene wax or carnauba wax, a generally recognized as safe material, resulting in controlled release on the order of seconds to hours. As a novelty, models are utilized to describe the coating layer thickness and the controlled-release behavior of the API, which occurs because of a diffusion-based mechanism. Such modeling would allow the design and control of the coating process with application for the controlled release of microparticles, particularly those less than 100 μm, which are difficult to coat by conventional solvent coating methods.

In order to improve fundamental understanding of powder flow behavior which is essential to the success of many pharmaceutical processes, this study investigates the relationship between particle-scale interactions dominated by the cohesive van der Waals force and the flow function coefficient. This study finds that the granular Bond number, defined as the ratio of the inter-particle cohesion force to particle weight, correlates well to the flow function coefficient, a metric used to assess powder flow performance and defined as the ratio of consolidation stress to unconfined yield strength. The inter-particle cohesion force was calculated by the so-called multi-asperity model which is a modification of the well-known Rumpf equation. As a major novelty, a granular Bond number is defined for multi-component mixtures (i.e. powder blends) and used to predict the flow function coefficient of binary, ternary, and quinary mixtures of a model API, acetaminophen, and two common pharmaceutical excipients, microcrystalline cellulose and pregelatinized starch. Surface modification via dry-coating was also used to alter the inter-particle force and more thoroughly investigate the effect of particle interactions on powder flow performance. Since the multi-component granular Bond number takes into account particle properties and particle interactions of all components in the powder blend, this novel approach shows good predictability for powder mixtures. Although the flow function coefficient alone is not a stringent prediction of powder flow, the modeling effort put forth in this study can be used to better guide formulation development.