Heat Transfer Evaluation During Twin-Screw Wet Granulation in View of Detailed Process Understanding.

During the last decade, the pharmaceutical industry has shown a growing interest in continuous twin-screw granulation (TSG). Despite flourishing literature on TSG, limited studies focused on fundamental process understanding on its mechanisms. In current study, granule quality attributes along the length of the TSG barrel were evaluated together with heat transfer in order to achieve a more fundamental understanding of the granulation process. An experimental setup was developed allowing the collection of granules at the different TSG compartments. In addition to the determination of typical granule attributes, mechanical energy, barrel and granule temperature (measured using an in-line implemented infra-red camera) were measured to evaluate heat transfer occurring at the different compartments and to relate them to granulation mechanisms. Collected data identified wetting enthalpy and friction forces as the main sources of heat along the granulator length. Wetting occurred in the wetting zone and generated temperature increase depending on liquid-to-solid ratio and powder wettability. In the kneading zones, granule temperature increase was proportional to mechanical energy. While it is usually admitted that granule consolidation and reshaping are the consequence of the high shear experienced by the granules, it was highlighted that most of the mechanical energy is converted into thermal energy with no correlation between mechanical energy and granule size distribution. Combined mass and energy balance of the granulation process are therefore necessary to capture the interaction between granule properties and physico-chemical and mechanical phenomena occurring in each compartment.

A framework for the in silico assessment of the robustness of an MPC in a CDC line in function of process variability.

The shift from batch manufacturing towards continuous manufacturing for the production of oral solid dosages requires the development and implementation of process models and process control. Previous work focused mainly on developing deterministic models for the investigated system. Furthermore, the in silico tuning and analysis of a control strategy are mostly done based on deterministic models. This deterministic approach could lead to wrong actions in diversion strategies and poor transferability of the controller performance if the system behaves differently than the deterministic model. This work introduces a framework that explicitly includes the process variability which is characteristic of powder handling processes and tests it on a novel continuous feeding-blending unit (i.e., the FE continuous processing system (CPS)), followed by a tablet press (i.e., the FE 55). It employs a stochastic model by allowing the model parameters to have a probability distribution. The performance of a model predictive control (MPC), steering the feed rate of the main excipient feeder to compensate for the feed rate deviations of the active pharmaceutical ingredient (API) feeder to keep the API concentration close to the desired value, is evaluated and the impact of process variability is assessed in a Monte Carlo (MC) analysis. Next to the process variability, a model for the prediction error of the chemometric model and realistic feed rate disturbances were included to increase the transferability of the results to the real system. The obtained results show that process variability is inherently present and that wrong conclusions can be drawn if it is not taken into account in the in silico analysis.

Effect of process parameters and formulation properties on the lead-lag between in-line NIR tablet press feed frame and off-line NIR tablet measurements.

The use of in-line near-infrared (NIR) measurements for tablet potency monitoring and diversion was studied. First, the optimal sample size for in-line NIR measurements inside the feed chute and the dosing and filling chamber of the tablet press feed frame was determined to allow proper comparison between these different measurement positions. Because of the considerably longer measurement time needed to obtain the same sample size inside the feed chute compared to the feed frame, the possibility of powder segregation inside the feed chute and the additional powder mixing inside the feed frame, the latter is preferred over the feed chute for in-line blend potency monitoring. Next, a design of experiments (DoE) was performed to evaluate the effect of paddle speed, turret speed, overfill level and formulation properties upon the lead-lag and the time it takes before the powder blend that is expelled at the dosing station is measured by the NIR inside the dosing chamber. Lead-lag is defined as the difference in time and API concentration between the measured in-line NIR response inside the filling chamber of the feed frame and the off-line NIR tablet response. Paddle speed and turret speed were the only compression parameters affecting lead-lag. Lead-lag decreased with increasing paddle speed for the first formulation. For the second formulation, lead-lag decreased with decreasing paddle speed and/or increasing turret speed. Formulation properties did not have an effect on the lead-lag. The in-line NIR response inside the dosing chamber of the feed frame was found to be closely following the tablet NIR response. Therefore, the dosing chamber could be used as an additional in-line NIR position for tablet potency monitoring and diversion. It can provide an extra layer of confidence about the final tablet quality. To demonstrate this potential benefit of simultaneous in-line NIR measurements inside the filling and dosing chamber of the feed frame, a tableting experiment was performed where a surrogate API spike was introduced into the product stream to mimic a potential process disturbance. The in-line NIR measurements inside the filling chamber allow diverting tablets in-time when the blend potency crosses the predefined control limits. And because the NIR response inside the dosing chamber closely follows the tablet NIR response, tablet diversion can discontinue when the blend potency inside the dosing chamber is again within the control limits. This could increase the yield of the tableting process by avoiding a longer than needed wash-out period and rejecting tablets that meet the release limits.

T-shaped partial least squares for high-dosed new active pharmaceutical ingredients in continuous twin-screw wet granulation: Granule size prediction with limited material information.

This work presents a granule size prediction approach applicable to diverse formulations containing new active pharmaceutical ingredients (APIs) in continuous twin-screw wet granulation. The approach consists of a surrogate selection method to identify similar materials with new APIs and a T-shaped partial least squares (T-PLS) model for granule size prediction across varying formulations and process conditions. We devised a surrogate material selection method, employing a combination of linear pre-processing and nonlinear classification algorithms, which effectively identified suitable surrogates for new materials. Using only material properties obtained through four characterization methods, our approach demonstrated its predictive prowess. The selected surrogate methods were seamlessly integrated with our developed T-PLS model, which was meticulously validated for high-dose formulations involving three new APIs. When surrogating new APIs based on Gaussian process classification, we achieved the lowest prediction errors, signifying the method's robustness. The predicted d-values were within the range of uncertainty bounds for all cases, except for d90 of API C. Notably, the approach offers a direct and efficient solution for early-phase formulation and process development, considerably reducing the need for extensive experimental work. By relying on just four material characterization methods, it streamlines the research process while maintaining a high degree of accuracy.

Exploring the effect of raw material properties on continuous twin-screw wet granulation manufacturability.

Twin-screw wet granulation (TSWG) stands out as a promising continuous alternative to conventional batch fluid bed- and high shear wet granulation techniques. Despite its potential, the impact of raw material properties on TSWG processability remains inadequately explored. Furthermore, the absence of supportive models for TSWG process development with new active pharmaceutical ingredients (APIs) adds to the challenge. This study tackles these gaps by introducing four partial least squares (PLS) models that approximate both the applicable liquid-to-solid (L/S) ratio range and resulting granule attributes (i.e., granule size and friability) based on initial material properties. The first two PLS models link the lowest and highest applicable L/S ratio for TSWG, respectively, with the formulation blend properties. The third and fourth PLS models predict the granule size and friability, respectively, from the starting API properties and applied L/S ratio for twin-screw wet granulation. By analysing the developed PLS models, water-related material properties (e.g., solubility, wettability, dissolution rate), as well as density and flow-related properties (e.g., flow function coefficient), were found to be impacting the TSWG processability. In addition, the applicability of the developed PLS models was evaluated by using them to propose suitable L/S ratio ranges (i.e., resulting in granules with the desired properties) for three new APIs and related formulations followed by an experimental validation thereof. Overall, this study helped to better understand the effect of raw material properties upon TSWG processability. Moreover, the developed PLS models can be used to propose suitable TSWG process settings for new APIs and hence reduce the experimental effort during process development.

Partial least squares regression to calculate population balance model parameters from material properties in continuous twin-screw wet granulation.

In the pharmaceutical industry, twin-screw wet granulation has become a realistic option for the continuous manufacturing of solid drug products. Towards the efficient design, population balance models (PBMs) have been recognized as a tool to compute granule size distribution and understand physical phenomena. However, the missing link between material properties and the model parameters limits the swift applicability and generalization of new active pharmaceutical ingredients (APIs). This paper proposes partial least squares (PLS) regression models to assess the impact of material properties on PBM parameters. The parameters of the compartmental one-dimensional PBMs were derived for ten formulations with varying liquid-to-solid ratios and connected with material properties and liquid-to-solid ratios by PLS models. As a result, key material properties were identified in order to calculate it with the necessary accuracy. Size- and moisture-related properties were influential in the wetting zone whereas density-related properties were more dominant in the kneading zones.

Mechanistic modeling of semicontinuous fluidized bed drying of pharmaceutical granules by incorporating single particle and bulk drying kinetics.

In this work, a mechanistic fluidized bed drying model computing the granule moisture content in function of granule size, drying time, process settings and formulation properties is developed. Modeling the moisture content distribution concerning the granule size is essential for tabletability and drug product quality. This work combines a mechanistic bulk model and a single-particle drying kinetics model in a semicontinuous mode. The added model complexity allows physical approximations of drying phenomena at both the drying system level and the granular level. This includes quantifying the variations in moisture content by taking into account the specific dryer design and the variations in granule size. The model performance was quantified through industrially relevant case studies. It was revealed that the proposed model structure accurately predicts the drying behavior of the yield fraction. However, systematic model biases were observed for the fine and coarse fractions of the granule size distribution. In addition, discrepancies in the predicted outgoing air properties (relative air humidity and air temperature) were obtained. Further enhancement of the model complexity, e.g. complete incorporation of fluidization and segregation phenomena, is likely to improve the model performance. Notwithstanding, the developed model forms a step towards a formulation-generic fluidized bed drying model as interacting mechanisms on different levels of the drying system are considered.

Validation of model-based design of experiments for continuous wet granulation and drying.

This paper presents an application case of model-based design of experiments for the continuous twin-screw wet granulation and fluid-bed drying sequence. The proposed framework consists of three previously developed models. Here, we are testing the applicability of previously published unit operation models in this specific part of the production line to a new active pharmaceutical ingredient. Firstly, a T-shaped partial least squares regression model predicts d-values of granules after wet granulation with different process settings. Then, a high-resolution full granule size distribution is computed by a hybrid population balance and partial least squares regression model. Lastly, a mechanistic model of fluid-bed drying simulates drying time and energy efficiency, using the outputs of the first two models as a part of the inputs. In the application case, good operating conditions were calculated based on material and formulation properties as well as the developed process models. The framework was validated by comparing the simulation results with three experimental results. Overall, the proposed framework enables a process designer to find appropriate process settings with a less experimental workload. The framework combined with process knowledge reduced 73.2% of material consumption and 72.3% of time, especially in the early process development phase.

Determination and understanding of lead-lag between in-line NIR tablet press feed frame and off-line NIR tablet measurements.

The influence of different tableting process parameters on lead-lag was studied by collecting in-line near-infrared (NIR) spectra in the filling chamber of the tablet press feed frame and off-line NIR tablet data. Lead-lag is defined as the difference in time and API concentration between the measured in-line feed frame NIR response and the off-line NIR tablet data. Lead-lag results from the product formulation blend undergoing additional mixing after passing the NIR probe inside the feed frame, before being filled into the dies of the tablet press. A design of experiments (DoE) was performed to evaluate the effect of the tableting process factors paddle speed, turret speed, overfill level, paddle speed ratio and feed frame type upon lead-lag. Paddle speed and turret speed were identified as the only tableting parameters affecting lead-lag. Lead-lag decreased with increasing paddle speed or turret speed and became negligible at high paddle speed and high turret speed. Overfill level, paddle speed ratio and feed frame type did not affect lead-lag, suggesting that the amount and the trajectory of the recirculating powder in the feed frame did not significantly vary and hence influence the lead-lag within the examined process factor ranges. Finally, a methodology was developed using the in-line feed frame NIR measurements for the continuous monitoring and control of blend potency and tablet content uniformity. Tablet diversion should start when the in-line feed frame monitored blend potency exceeds the predefined control limits and can discontinue when this blend potency is again within the control limits for a duration equal to the lead-lag time. A combination of continuous blend potency monitoring inside the feed frame and in-process tablet weight control allows real-time tablet content uniformity assurance. Although the findings of this study are restricted to the specific equipment, tableting parameter ranges and product formulation used, the suggested approach for lead-lag determination and continuous tablet content uniformity monitoring can be applied to any rotary tablet press and product formulation.

An extended 3-compartment model for describing step change experiments in pharmaceutical twin-screw feeders at different refill regimes.

Residence time distributions (RTDs) are a valuable tool for product tracking in the unit operations of a continuous line for manufacturing pharmaceutical oral solid dosage (OSD) and the integrated system itself. The first unit operation in such a continuous line in which extended intermixing can occur, is typically a feeder. The RTD of a feeder can be obtained by performing tracer experiments with a tracer material. A physical interpretation can be given to the observed tracer concentration responses by fitting a tanks-in-series (TIS) or compartmental model to it. Consequently, the internal mixing behaviour inside the feeder hopper can be rationalized. However, typically, a constant volume is assumed for the tanks or compartments in these models. This has led to several publications where the experimental set-up does not violate the constant volume assumption, i.e. one performs refills at a high hopper fill level. Here, we step away from this assumption and develop a set of differential equations for a 3-compartment model in order to account for a non-constant volume of the compartments. Moreover, the model distinguishes between a bypass trajectory formed by the agitator inside the feeder and an inner mixing volume, in which the tracer concentration lags on the tracer concentration in the bypass volume. This compartmentalization was inspired by the results obtained in a previous study using a spatial sampling method to assess the tracer concentration throughout the feeder hopper for different experimental runtimes. The developed model successfully describes the step responses for different refill regimes: the standard smooth first order plus dead time response (FOPDT) for a high refill regime and the more complex step response, including dips in the rising phase of the curve, for the low refill regime. As a consequence, a more thorough understanding of the complex mixing behaviour inside the feeder is obtained, which allows for an improved traceability. Next to that, the model delivers enhanced knowledge on the interaction between the residence time and the refill regime. The developed model was fitted to a data set, containing step change experiments for different pharmaceutical materials (Tablettose 80 (T80), Microcelac 100 (MCL), and Avicel PH101 (MCC)), different mass flow rates, and refill regimes. The experimentally observed phenomena could be reliably described by the proposed model. The model showed an improved transferability compared to typical TIS models.

Development of a tablet press feed frame lead lag determination model using in-line and off-line NIR measurements.

For continuous pharmaceutical manufacturing of oral solid dosages, it is essential that product quality is measured inline. In this application, a continuous rotary tablet press is used. The goal is a model-based assessment of the quality of the blend in the feed frame to determine whether the concentration of the active pharmaceutical ingredient (API) will be within the prescribed limits. This is to achieve a better quality assurance than by offline testing of a small sample of tablets. In this way, product quality for real-time release (RTR) could be implemented. With a near-infrared (NIR) probe, the concentration of the API in the feed chute and the feed-frame were measured, as well as the API concentration of the tablets by an offline NIR measurement. These different data sets are connected and used for the residence time distribution characterization of the mixing dynamic of the tablet press. A residence time distribution model is fitted to the data, and is further used to compute the lead-lag time. This yields information on how long it takes for a quantity of product to go from being measured in the feed frame until ending up in tablets. Further, it gives information on the occurrence of mixing in the feed-frame itself. These models allow making accurate predictions of whether tablets fall within specified concentration range in real-time. The real-time prediction can be used in combination with a control system both to maintain the quality of the blend as well as to know which tablets to discard. This real-time quality assurance will lead to less material waste and fewer declined batches of tablets.

A multivariate formulation and process development platform for direct compression.

The efficient development of robust tableting processes is challenging due to the lack of mechanistic understanding on the impact of raw material properties and process parameters on tablet quality. The experimental determination of the effect of process and formulation parameters on tablet properties and subsequent optimization is labor-intensive, expensive and time-consuming. The combined use of an extensive raw material property database, process simulation tools and multivariate modeling allows more efficient and more optimized development of the direct compression (DC) process. In this study, key material attributes and in-process mechanical properties with a potential effect on tablet processability and tablet properties were identified. In a first step, an extensive characterization of 55 raw materials (over 100 material descriptors) (Van Snick et al., 2018) and 26 formulation blends (31 material descriptors) (Dhondt et al., 2022) was performed. These blends were subsequently compacted on a compaction simulator under multiple process conditions through a design of experiments (DoE) approach. A T-shaped partial least squares (T-PLS) model was established which correlates tablet quality attributes with process settings, raw material properties and blend ratios. During future development of the DC formulation and process for a new active pharmaceutical ingredient (API), this model can then be used to provide a preliminary formulation and compaction process settings as starting point to be further optimized during development trials based on well-defined raw material characteristics and compaction tests. This study hence contributes to a better understanding on the impact of raw material properties and process settings on a DC process and final properties of the produced tablets; and provides a platform allowing a more efficient and more optimized development of a robust tableting process.

Improvement of a 1D Population Balance Model for Twin-Screw Wet Granulation by Using Identifiability Analysis.

Recently, the pharmaceutical industry has undergone changes in the production of solid oral dosages from traditional inefficient and expensive batch production to continuous manufacturing. The latest advancements include increased use of continuous twin-screw wet granulation and application of advanced modeling tools such as Population Balance Models (PBMs). However, improved understanding of the physical process within the granulator and improvement of current population balance models are necessary for the continuous production process to be successful in practice. In this study, an existing compartmental one-dimensional PBM of a twin-screw granulation process was improved by altering the original aggregation kernel in the wetting zone as a result of an identifiability analysis. In addition, a strategy was successfully applied to reduce the number of model parameters to be calibrated in both the wetting zone and kneading zones. It was found that the new aggregation kernel in the wetting zone is capable of reproducing the particle size distribution that is experimentally observed at different process conditions as well as different types of formulations, varying in hydrophilicity and API concentration. Finally, it was observed that model parameters could be linked not only to the material properties but also to the liquid to solid ratio, paving the way to create a generic PBM to predict the particle size distribution of a new formulation.

Predicting Pharmaceutical Particle Size Distributions Using Kernel Mean Embedding.

In the pharmaceutical industry, the transition to continuous manufacturing of solid dosage forms is adopted by more and more companies. For these continuous processes, high-quality process models are needed. In pharmaceutical wet granulation, a unit operation in the ConsiGma TM -25 continuous powder-to-tablet system (GEA Pharma systems, Collette, Wommelgem, Belgium), the product under study presents itself as a collection of particles that differ in shape and size. The measurement of this collection results in a particle size distribution. However, the theoretical basis to describe the physical phenomena leading to changes in this particle size distribution is lacking. It is essential to understand how the particle size distribution changes as a function of the unit operation's process settings, as it has a profound effect on the behavior of the fluid bed dryer. Therefore, we suggest a data-driven modeling framework that links the machine settings of the wet granulation unit operation and the output distribution of granules. We do this without making any assumptions on the nature of the distributions under study. A simulation of the granule size distribution could act as a soft sensor when in-line measurements are challenging to perform. The method of this work is a two-step procedure: first, the measured distributions are transformed into a high-dimensional feature space, where the relation between the machine settings and the distributions can be learnt. Second, the inverse transformation is performed, allowing an interpretation of the results in the original measurement space. Further, a comparison is made with previous work, which employs a more mechanistic framework for describing the granules. A reliable prediction of the granule size is vital in the assurance of quality in the production line, and is needed in the assessment of upstream (feeding) and downstream (drying, milling, and tableting) issues. Now that a validated data-driven framework for predicting pharmaceutical particle size distributions is available, it can be applied in settings such as model-based experimental design and, due to its fast computation, there is potential in real-time model predictive control.

Model-based NIR spectroscopy implementation for in-line assay monitoring during a pharmaceutical suspension manufacturing process.

The implementation of Process Analytical Technology (PAT) instruments is generally achieved stochastically. Sub-optimal PAT locations could introduce variation in the measurements which is not related to the analyte of interest. For this reason, rational approaches should be considered to establish an optimal sensor placement where relevant measurements are possible and the impact of disturbances is minimized. The aim of this paper is to demonstrate how mechanistic modelling can support appropriate sensor implementation by means of a case study. A PAT method was developed for a bottle filling process of a pharmaceutical formulation with the goal of increasing the yield of the process by gaining process understanding and redefining the endpoint of the process. To ensure proper measurements, an advanced measuring interfacing was assembled. The design of this device was rationalized with the help of a model-based approach using three-dimensional Computational Fluid Dynamics modeling. This allows to maximize the performance of the PAT method and exploit its full benefits.

Application of iterative robust model-based optimal experimental design for the calibration of biocatalytic models.

The aim of model calibration is to estimate unique parameter values from available experimental data, here applied to a biocatalytic process. The traditional approach of first gathering data followed by performing a model calibration is inefficient, since the information gathered during experimentation is not actively used to optimize the experimental design. By applying an iterative robust model-based optimal experimental design, the limited amount of data collected is used to design additional informative experiments. The algorithm is used here to calibrate the initial reaction rate of an ω-transaminase catalyzed reaction in a more accurate way. The parameter confidence region estimated from the Fisher Information Matrix is compared with the likelihood confidence region, which is not only more accurate but also a computationally more expensive method. As a result, an important deviation between both approaches is found, confirming that linearization methods should be applied with care for nonlinear models. © 2017 American Institute of Chemical Engineers Biotechnol. Prog., 33:1278-1293, 2017.

In-depth experimental analysis of pharmaceutical twin-screw wet granulation in view of detailed process understanding.

Twin-screw wet granulation is gaining increasing interest within the pharmaceutical industry for the continuous manufacturing of solid oral dosage forms. However, limited prior fundamental physical understanding has been generated relating to the granule formation mechanisms and kinetics along the internal compartmental length of a twin-screw granulator barrel, and about how process settings, barrel screw configuration and formulation properties such as particle size, density and surface properties influence these mechanisms. One of the main reasons for this limited understanding is that experimental data is generally only collected at the exit of the twin-screw granulator barrel although the granule formation occurs spatially along the internal length of the barrel. The purpose of this study is to analyze the twin-screw wet granulation process using both hydrophilic and hydrophobic formulations, manufactured under different process settings such as liquid-to-solid ratio, mass throughput and screw speed, in such a way that the mechanisms occurring in the individual granulator barrel compartments (i.e., the wetting and different conveying and kneading compartments) and their impact upon granule formation are understood. To achieve this, a unique experimental setup was developed allowing granule characteristic data-collection such as size, shape, liquid and porosity distribution at the different compartments along the length of the granulator barrel. Moreover, granule characteristic information per granule size class was determined. The experimental results indicated that liquid-to-solid ratio is the most important factor dictating the formation of the granules and their corresponding properties, by regulating the degree of aggregation and breakage in the different compartments along the internal length of the twin-screw granulator barrel. Collecting appropriate and detailed experimental data about granule formation along the internal length of the granulator barrel is thus crucial for gaining fundamental physical understanding of the twin-screw wet granulation process.