Mechanistic modeling of twin screw wet granulation for pharmaceutical formulations: Calibration, sensitivity analysis, and model-driven workflow.

Wet granulation, a particle size enlargement process, can significantly enhance the critical quality attributes of powders and improve the ability to form tablets in pharmaceutical manufacturing. In this study, a mechanistic-based population balance model is applied to twin screw wet granulation. This model incorporated a recently developed breakage kernel specifically designed for twin screw granulation, along with nucleation, layering, and consolidation. Calibration and validation were performed on Hydrochlorothiazide and Acetaminophen formulations, which exhibit different particle size and wettability characteristics. Utilizing a compartmental experimental dataset, a comprehensive global sensitivity analysis identified critical inputs impacting quality attributes. The study revealed that the nucleation rate process model, effectively represented particle size distributions for both formulations. Adjustments to nucleation and breakage rate parameters, influenced by material properties and screw configuration, improved the model's accuracy. A model-driven workflow was proposed, offering step-by-step guidelines and facilitating PBM model usage, providing essential details for future active pharmaceutical ingredient (API) formulations.

A novel mixing rule model to predict the flowability of directly compressed pharmaceutical blends.

In the pharmaceutical industry, powder flowability is an essential manufacturability attribute to consider when selecting the suitable manufacturing route and formulation. The selection of the formulation is usually based on the physical and chemical properties of the Active Pharmaceutical Ingredient (API) under consideration. Current industrial practice heavily relies on experimental work, which often results in significant labor and API consumption that results in higher costs. In this study we describe the development of a mixing rule to predict powder blend flowability from the flow properties of the individual components for industrial formulations manufactured via Direct Compression (DC). The mixing rule assumes that the granular solids' interactions are dominated by cohesive forces but are pragmatic to calibrate from the perspective of the typical data collated in an industrial environment. The proposed model was validated using a range of different APIs and the results show that the model can effectively predict the flowability properties of any formulation across the space of DC-relevant formulation compositions. Finally, a connection between the model and APIs properties (shape and size) was investigated via a linear correlation between the API particle properties and interparticle forces.

Modelling the effect of L/S ratio and granule moisture content on the compaction properties in continuous manufacturing.

The pharmaceutical field is currently moving towards continuous manufacturing pursuing reduced waste, consistency, and automation. During continuous manufacturing, it is important to understand how both operating conditions and material properties throughout the process affect the final properties of the product to optimise and control production. In this study of a continuous wet granulation line, the liquid to solid ratio (L/S) and drying times were varied to investigate the effect of the final granule moisture content and the liquid to solid ratio on the properties of the granules during tabletting and the final tensile strength of the tablets. Both variables (L/S and granule moisture) affected the tablet tensile strength with the moisture content having a larger impact. Further analysis using a compaction model, showed that the compactability of the granules was largely unaffected by both L/S and moisture content while the compressibility was influenced by these variables, leading to a difference in the final tablet strength and porosity. The granule porosity was linked to the L/S ratio and used instead for the model fitting. The effect of moisture content and granule porosity was added to the model using a 3d plane relationship between the compressibility constant, the moisture content and porosity of the granules. The tablet tensile strength model, considering the effect of moisture and granule porosity, performed well averaging a root mean squared error across the different conditions of 0.17 MPa.

Model driven design for integrated twin screw granulator and fluid bed dryer via flowsheet modelling.

This paper presents a flowsheet modelling of an integrated twin screw granulation (TSG) and fluid bed dryer (FBD) process using a Model Driven Design (MDD) approach. The MDD approach is featured by appropriate process models and efficient model calibration workflow to ensure the product quality. The design space exploration is driven by the physics of the process instead of extensive experimental trials. By means of MDD, the mechanistic-based process kernels are first defined for the TSG and FBD processes. With the awareness of the underlying physics, the complementary experiments are carried out with relevance to the kinetic parameters in the defined models. As a result, the experiments are specifically purposeful for model calibration and validation. The L/S ratio (liquid to solid ratio) and inlet air temperature are selected as the Critical Process Parameters (CPPs) in TSG and FBD for model validation, respectively. Global System Analysis (GSA) is further performed to assess the uncertainty of CPPs imposed on the Critical Quality Attributes (CQAs), which provides significant insights to the exploration of the design space considering both TSG and FBD process parameters.

Exploring the role of crystal habit in the Ostwald rule of stages.

The crystallization of calcium carbonate is shown to be dictated by the Ostwald rule of stages (ORS), for high relative initial supersaturations ( S CaCO 3 = [ C a 2 + ] [ CO 3 2 - ] / K SP, Calcite > 2500 ), under sweet (carbon dioxide saturated) and anoxic (oxygen depleted) solution conditions. Rhombohedral calcite crystals emerge after the sequential crystallization and dissolution of the metastable polymorphs: vaterite (snowflake-shaped) and aragonite (needle-shaped). However, the presence of certain cations, which can form trigonal carbonates (e.g. Fe2+ and Ni2+), in concentrations as low as 1.5 mM, triggers the emergence of calcite crystals, with a star-shaped crystal habit, first. These star-shaped crystals dissolve to yield needle-shaped aragonite crystals, which in turn dissolve to give the rhombohedral calcite crystals. The star-shaped crystals, formed at high SCaCO3 , possess higher surface free energy (therefore higher apparent solubility) than their rhombohedral counterparts. This sequence of dissolution and recrystallization demonstrates that the ORS does not only drive the crystal towards its thermodynamically most stable polymorph but also towards its most stable crystal habit.

Model driven design for twin screw granulation using mechanistic-based population balance model.

This paper presents a generic framework of Model Driven Design (MDD) with its application for a twin screw granulation process using a mechanistic-based population balance model (PBM). The process kernels including nucleation, breakage, layering and consolidation are defined in the PBM. A recently developed breakage kernel is used with key physics incorporated in the model formulation. Prior to granulation experiments, sensitivity analysis of PBM parameters is performed to investigate the variation of model outputs given the input parameter variance. The significance of liquid to solid ratio (L/S ratio), nucleation and breakage parameters is identified by sensitivity analysis. The sensitivity analysis dramatically reduces the number of fitting parameters in PBM and only nine granulation experiments are required for model calibration and validation. A model validation flowchart is proposed to elucidate the evolution of kinetic rate parameters associated with L/S ratio and screw element geometry. The presented MDD framework for sensitivity analysis, parameter estimation, model verification and validation can be generalized and applied for any particulate process.

Tableting model assessment of porosity and tensile strength using a continuous wet granulation route.

This paper presents a comprehensive assessment of the most widely used tablet compaction models in a continuous wet granulation tableting process. The porosity models, tensile strength models and lubricant models are reviewed from the literature and classified based on their formulations i.e. empirical or theoretical and applications, i.e. batch or continuous. The majority of these models are empirical and were initially developed for batch tabletting process. To ascertain their effectiveness and serviceability in the continuous tableting process, a continuous powder processing line of Diamond Pilot Plant (DiPP) installed at The University of Sheffield was used to provide the quantitative data for tablet model assessment. Magnesium stearate (MgSt) is used as a lubricant to investigate its influence on the tensile strength. Whilst satisfactory predictions from the tablet models can be produced, a compromise between the model fidelity and model simplicity needs to be made for a suitable model selection. The Sonnergaard model outperforms amongst the porosity models whilst the Reynolds model produces the best goodness of fitting for two parameters fitting porosity models. An improved tensile strength model is proposed to consider the influence of powder size and porosity in the continuous tableting process.

Τhe role of surface energy in the apparent solubility of two different calcite crystal habits.

The interplay between polymorphism and facet-specific surface energy on the dissolution of crystals is examined in this work. It is shown that, using cationic additives, it is possible to produce star-shaped calcite crystals at very high supersaturations. In crystallization processes following the Ostwald rule of stages these star-shaped crystals appear to have higher solubility than both their rhombohedral counterparts and needle-shaped aragonite crystals. The vapour pressures of vaterite, aragonite, star-shaped calcite and rhombohedral calcite crystals are measured using thermogravimetric analysis and the corresponding enthalpies of melting are obtained. Using inverse gas chromatography, the surface energy of the aforementioned crystals is measured as well and the surface energy of the main crystal facets is calculated. Combining the effect of facet-specific surface energies and the enthalpies of melting on a modified version of the classical solubility equation for regular solutions, it is proved that the star-shaped calcite crystals can indeed have higher apparent solubility than aragonitecrystals.

Roller compaction: Infrared thermography as a PAT for monitoring powder flow from feeding to compaction zone.

Roller compaction is a continuous dry granulation process, in which powder is compressed by two counter-rotating rollers. During this process, the powder feeding to the compaction zone has a significant effect on product quality. This work investigates the flow of powder from the feeding zone to the compaction zone using online infrared thermography as Process Analytical Technology (PAT) which is achieved via a specially built cheek plate (side-sealing). The powder undergoes increasing stress from the rollers when it is approaching the minimum gap of the compaction zone, which can be indirectly monitored by measuring the powder temperature. The online monitoring of the powder flow during the roller compaction helps locate the nip region and identify the effect of different roller forces on the temperature of the feeding powder. The results show that the nip region can be identified by analysing the temperature profiles from the feeding to the compaction zone. The increase of roller force results in an increasing slope of the powder temperature profile. In addition, offline X-ray CT measurement results show the increase of density along the feeding to the compaction direction, which is compared with Johanson theory under different roller forces in the roller compaction process.

Application of feeding guiders to improve the powder distribution in the two scales of roller compactors.

Roller compaction is a continuous dry granulation process, where the powder is compressed between two counter-rotating rollers and compacted into ribbons. The quality and homogeneity of the granulate is determined by the uniformity and porosity of the ribbon, which depends on the feeding process of the primary powder to the rollers, the flow properties of the primary powder and process parameters such as roller forces. Previous work was conducted to improve the powder flow and distribution in the feeding zone by developing new feeding guiders, which are located in the feeding zone close to the rollers on the lab-scale roller compactor Alexanderwerk WP120 Pharma (Yu et al., 2018). These new feeding guiders were used to reduce the amount of powder that is delivered to the centre of the rollers and increase the amount of powder that is delivered to the sides of the rollers, in comparison to the original feeding guiders. This modified concept using new feeding guiders has been applied to the large-scale roller compactor Alexanderwerk WP200 Pharma in the present work. In order to evaluate the homogeneity of the ribbon properties across the ribbon width, the temperature profile and porosity distribution across the ribbon width were measured. The new feeding guiders resulted in ribbons being produced with a more uniform temperature profile and porosity distribution across the ribbon width when using the small and large scale roller compactors at different process parameters.

Characteristics of multi-component formulation granules formed using distributive mixing elements in twin screw granulation.

This work examines the influence of pharmaceutical powder formulation characteristics on granule properties formed using distributive mixing elements (DMEs) in twin screw granulation. High and low drug dose formulations with three different active pharmaceutical ingredients (APIs) were considered. The type and concentration of the API in the formulation significantly affected the dry blend particle size distribution and the wet blend dynamic yield strength. However, despite the differences in blend properties, the granule size distributions were not significantly affected by the type of API used. The granule size distributions were solely the functions of the liquid-to-solid ratio and screw element geometry. However, the granule porosities were observed to be dependent on both the liquid-to-solid ratio and the dynamic yield strength of the blends. This work is the first to study the influence of drug loading and API type on the granule attributes produced using distributive mixing elements.

Assessment of Intragranular and Extragranular Fracture in the Development of Tablet Tensile Strength.

When a tablet is compacted from deformable granules and then broken, the fracture plane may cleave granules in 2 (intragranular fracture) or separate neighboring granules (extragranular fracture). In this study, a novel method was developed to quantify the extent of intragranular versus extragranular fracture by compacting tablets from multicolored ideal granules and evaluating fracture surfaces. The proportions of intragranular and extragranular fracture were quantified and modeled in light of a new metric; the deformation potential, Δ, reflecting the solid fraction increase as an initial granule bed is compressed into a final tablet. Results show that a measurable tablet strength is achieved at Δ > 0.18, but intragranular fracture is not observed until Δ > 0.21. At very large Δ, tablets experience almost exclusively intragranular fracture, yet the tablet tensile strength is considerably lower than that of a tablet compacted from raw powders versus precompacted granules. Thus, secondary compaction of granules appears to weaken the granule matrix, leading to reduced tablet tensile strength even in the presence of strong extragranular bonding.

Improving feeding powder distribution to the compaction zone in the roller compaction.

In the roller compaction process, powder flow properties have a significant influence on the uniformity of the ribbon properties. The objective of this work was to improve the powder flow in the feeding zone by developing novel feeding guiders which are located in the feeding zone close to the rollers in the roller compactor (side sealing system). Three novel feeding guiders were designed by 3D printing and used in the roller compactor, aiming to control the amount of powder passing across the roller width. The new feeding guiders were used to guide more powder to the sides between the rollers and less powder to the centre comparing to the original feeding elements. Temperature profile and porosity across the ribbon width indicated the uniformity of the ribbon properties. Using the novel feeding guiders resulted in producing ribbons with uniform temperature profile and porosity distribution across the ribbon width. The design of the feeding guiders contributed to improving the tensile strength of the ribbons produced from the compaction stage as well as reducing the fines produced from the crushing stage.

Statistical analysis and comparison of a continuous high shear granulator with a twin screw granulator: Effect of process parameters on critical granule attributes and granulation mechanisms.

This study is concerned with identifying the design space of two different continuous granulators and their respective granulation mechanisms. Performance of a continuous high shear granulator and a twin screw granulator with paracetamol formulations were examined by face-centered cubic design, which focused on investigating key performance metrics, namely, granule size, porosity, flowability and particle morphology of granules as a function of essential input process parameters (liquid content, throughput and rotation speed). Liquid and residence time distribution tests were also performed to gain insights into the liquid-powder mixing and flow behavior. The results indicated that continuous high shear granulation was more sensitive to process variation and produced spherical granules with monomodal size distribution and distinct internal structure and strength variation. Twin screw granulation with such a particular screw configuration showed narrower design space and granules were featured with multimodal size distribution, irregular shape, less detectible porosity difference and tighter range of strength. Granulation mechanisms explored on the basis of nucleation and growth regime maps revealed that for most cases liquid binder was uniformly distributed with fast droplet penetration into the powder bed and that granule consolidation and coalescence mainly took place in the nucleation, steady growth and rapid growth regimes.

Investigation of an 11mm diameter twin screw granulator: Screw element performance and in-line monitoring via image analysis.

As twin screw granulation (TSG) provides one with many screw element options, characterization of each screw element is crucial in optimizing the screw configuration in order to obtain desired granule attributes. In this study, the performance of two different screw elements - distributive feed screws and kneading elements - was studied in an 11 mm TSG at different liquid-to-solid (L/S) ratios. The kneading element configuration was found to break large granules more efficiently, leading to narrower granule size distributions. While pharmaceutical industry shifts toward continuous manufacturing, inline monitoring and process control are gaining importance. Granules from an 11 mm TSG were analysed using the Eyecon™, a real-time high speed direct imaging system, which has been used to capture accurate particle size distribution and particle count. The size parameters and particle count were then assessed in terms of their ability to be a suitable control measure using the Shewhart control charts. d10 and particle count were found to be good indicators of the change in L/S ratio. However, d50 and d90 did not reflect the change, due to their inherent variability even when the process is at steady state.

Effect of relative humidity on onset of capillary forces for rough surfaces.

Atomic force microscopy (AFM) was used to investigate the effect of relative humidity (RH) on the adhesion forces between silicon nitride AFM probes, hydrophilic stainless steel, and hydrophobic Perspex® (polymethylmethacrylate, PMMA). In addition, AFM-based phase contrast imaging was used to quantify the amount and location of adsorbed water present on these substrates at RH levels ranging from 15% to 65% at 22°C. Both the adhesion forces and the quantities of adsorbed moisture were seen to vary with RH, and the nature of this variation depended on the hydrophobicity of the substrate. For the Perspex®, both the adhesion force and the amount of adsorbed moisture were essentially independent of RH. For the stainless steel substrate, adsorbed moisture increased continuously with increasing RH, while the adhesion force rose from a minimum at 15% RH to a broad maximum between 25% and 35% RH. From 35% to 55% RH, the adhesion force dropped continuously to an intermediate level before rising again as 65% RH was approached. The changes in adhesion force with increasing relative humidity in the case of the stainless steel substrate were attributed to a balance of effects associated with adsorbed, sub-continuum water on the cantilever and steel. Hydrogen bonding interactions between these adsorbed water molecules were thought to increase the adhesion force. However, when significant quantities of molecular water adsorbed, these molecules were expect to decrease adhesion by screening the van der Waals interactions between the steel and the cantilever tip, and by increasing the separation distance between these solid surfaces when they were 'in contact'. Finally, the slight increase in adhesion between 55% and 65% RH was attributed to true capillary forces exerted by continuum water on the two solid surfaces.

Osmotic second virial cross-coefficient measurements for binary combination of lysozyme, ovalbumin, and α-amylase in salt solutions.

Interactions measurement is a valuable tool to predict equilibrium phase separation of a desired protein in the presence of unwanted macromolecules. In this study, cross-interactions were measured as the osmotic second virial cross-coefficients (B23 ) for the three binary protein systems involving lysozyme, ovalbumin, and α-amylase in salt solutions (sodium chloride and ammonium sulfate). They were correlated with solubility for the binary protein mixtures. The cross-interaction behavior at different salt concentrations was interpreted by either electrostatic or hydrophobic interaction forces. At low salt concentrations, the protein surface charge dominates cross-interaction behavior as a function of pH. With added ovalbumin, the lysozyme solubility decreased linearly at low salt concentration in sodium chloride and increased at high salt concentration in ammonium sulfate. The B23 value was found to be proportional to the slope of the lysozyme solubility against ovalbumin concentration and the correlation was explained by preferential interaction theory.

Modeling the granule formation mechanism from single drop impact on a powder bed.

Granule formation from drop impact on a powder bed can occur by either Tunneling or Spreading/Crater Formation. The governing regime can be specified by the experimentally determined modified Bond number (Bo(g)*), which is a ratio of the capillary force to the gravitational force acting on a particle. It was hypothesized that Tunneling would occur when the capillary and surface tension forces exceeded the weight of a powder aggregate in contact with the drop. To confirm this hypothesis, force balances were derived for a drop in contact with a single particle and separately for a drop in contact with an aggregate to predict when a particle or aggregate will be sucked into the drop. The force ratios derived for each case were compared to the Bo(g)* force ratio used in a previously published regime map that separates Tunneling from Spreading/Crater Formation. The force balance model correctly predicts the trends of the impact of powder and liquid properties on the governing regime. However, the single particle model does not quantitatively predict the critical Bond number for regime change in Tunneling. The aggregate model gave a better prediction of the Tunneling boundary than the single particle model, but it still under predicts the experimentally determined Tunneling criterion given by the Bond number. Potential reasons for this discrepancy are discussed.

Evaluation of three approaches for real-time monitoring of roller compaction with near-infrared spectroscopy.

Three different approaches have been evaluated for monitoring ribbon density through real-time near-infrared spectroscopy measurements. The roll compactor was operated to produce microcrystalline cellulose (MCC) ribbons of varying densities. The first approach used the slope of the spectra which showed a variation through the ribbon that could be attributed to density. A second qualitative approach was also developed with a principal component analysis (PCA) model with spectra taken in-line during the production of ribbons in an ideal roll pressure range. The PCA (i.e., real-time) density scans show that the model was able to qualitatively capture the density responses resulting from variation in process parameters. The third approach involved multivariate partial least squares (PLS) calibration models developed at wavelength regions of 1,120-1,310 and 1,305-2,205 nm. Also, various PLS models were developed using three reference methods: caliper, pycnometer, and in-line laser. The third approach shows a quantitative difference between the model-predicted and the measured densities. Models developed at high-wavelength region showed highest accuracy compared with models at low-wavelength region. All the PLS models showed a high accuracy along the spectra collected throughout the production of the ribbons. The three methods showed applicability to process control monitoring by describing the changes in density during in-line sampling.

Correlation of second virial coefficient with solubility for proteins in salt solutions.

In this work, osmotic second virial coefficients (B(22)) were determined and correlated with the measured solubilities for the proteins, α-amylase, ovalbumin, and lysozyme. The B(22) values and solubilities were determined in similar solution conditions using two salts, sodium chloride and ammonium sulfate in an acidic pH range. An overall decrease in the solubility of the proteins (salting out) was observed at high concentrations of ammonium sulfate and sodium chloride solutions. However, for α-amylase, salting-in behavior was also observed in low concentration sodium chloride solutions. In ammonium sulfate solutions, the B(22) are small and close to zero below 2.4 M. As the ammonium sulfate concentrations were further increased, B(22) values decreased for all systems studied. The effect of sodium chloride on B(22) varies with concentration, solution pH, and the type of protein studied. Theoretical models show a reasonable fit to the experimental derived data of B(22) and solubility. B(22) is also directly proportional to the logarithm of the solubility values for individual proteins in salt solutions, so the log-linear empirical models developed in this work can also be used to rapidly predict solubility and B(22) values for given protein-salt systems.