Continuous direct compression of a commercially batch-manufactured tablet formulation with two different processing lines

The transfer from batch-based to continuous tablet manufacturing increases the quality and efficiency of processes. Nonetheless, as in the development of a batch process, the continuous process design requires optimization studies to ensure a robust process. In this study, processing of a commercially batch-manufactured tablet product was tested with two continuous direct compression lines while keeping the original formulation composition and tablet quality requirements. Tableting runs were conducted with different values of process parameters. Changes in parameter settings were found to cause differences in tablet properties. Most of these quality properties could be controlled and maintained within the set limits effortlessly already at this stage of studies. However, the API content and content uniformity seemed to require more investigation. The observed content uniformity challenges were traced to individual tablets with a high amount of API. This was suspected to be caused by API micro-agglomerates since tablet weight variability did not explain the issue. This could be solved by adding a mill between two blenders in the process line. Overall, this case study produced promising results with both tested manufacturing lines since many tablet properties complied with the test result limits without optimization of process parameter settings.

Impact of blend properties and process variables on the blending performance

In this study, quantitative relationships were established between blend properties, process settings and blending responses via multivariate data-analysis. Four divergent binary blends were composed in three different ratios and processed at various throughputs and impeller speeds. Additionally, different impeller configurations were tested to see their impact on the overall blending performance. During each run, feeder mass flows were compared with the API concentration (BU) in order to investigate the dampening potential of the blender. The blender hold-up mass (HM), mean residence time (MRT), strain on the powder (#BP) and BU variability (RSD_BU) were determined as blending descriptors and analyzed via PLS-regression. This elucidated the correlation between process settings (i.e. throughput and impeller speed) and blending responses, as well as the impact of blend properties on MRT and RSD_BU. Furthermore, the study revealed that HM does not need to be in steady state conditions to assure a stable BU, while it became clear that long/large feeder deviations can only be dampened by the blender when using dedicated impeller configurations. Overall, this study demonstrated the generic application of the blender, while the developed PLS models could be used to predict the blender performance based on the blend properties.

Two dimensional modeling of sewage sludge flow in a double-axis continuous paddle dryer

A two-dimensional model based on the theory of Markov chain has been developed to determine the Residence Time Distribution (RTD) of municipal sewage sludge in a double-axis continuous paddle dryer. Based on the experiments, the paper proposes further corrections to simulate the transition. To characterize the sludge transition in the dryer, the parameter of internal recirculation based on one dimensional model is further developed. In addition, two parameters, the internal forward coefficient between two axes and the internal forward coefficient in one axis are introduced to characterize the new model. In absence of available correlation, solid hold-up of each cell in dryer and both the recirculation parameters are identified by fitting the model to experiments. Through analysis, the model demonstrates its ability to describe the sludge flow in a double-axis continuous paddle dryer by the experimental and simulation RTD curve. Finally, an analysis of influence factors highlights that recirculation coefficients are critical for the model while solid hold-up H_u of each cell controls the mean residence time and the final moisture content. In addition, the geometric residence time of sludge flow has a negligible effect on sludge flow for it has no effect on dimensionless variance. Moreover, compared with the recirculation coefficient between different axes R, the recirculation coefficient in one axis r has a negligible effect. In addition, recirculation parameters have no effect on mean residence time of sludge flow.

Integrating sensors for monitoring blend content in a pharmaceutical continuous manufacturing plant

In a pharmaceutical manufacturing process, Critical Quality Attributes (CQAs) need to be monitored not only for the final product but also for intermediates. Blend uniformity of powders is one such attribute that needs to be measured to ensure the quality of the final product. Multiple in-line sensors were implemented within a Direct Compaction (DC) continuous tablet manufacturing line to monitor the blend content of the powders. In most cases, since the primary ingredient of interest is the active pharmaceutical ingredient (API), the concentration (potency) of the API was monitored/predicted over the course of manufacturing. For the calibration model building process, a unique setup involving dynamic powder spectral acquisition method was used. This setup was aimed at mimicking the powder flow characteristics within the manufacturing line, while at the same time utilizing a relatively small amount of powder. A Raman probe and a portable NIR were used concurrently at the exit of the blending process before the tableting stage. The performance of the two sensors and their respective models were evaluated in terms of accuracy, precision, operating range, measurement frequency, placement, reliability, robustness, and compared to predictions using gravimetric feed rates. Additionally, their performances were validated by off-line traditional analytical measurements.

Mixing Cell: a Device to Mimic Extent of Lubrication and Shear in Continuous Tubular Blenders

Continuous manufacturing (CM) has clear potential for manufacturing solid oral dosages. It provides several advantages that may aid the manufacturing and quality of drug products. However, one of the main challenges of this technology is the relatively large amount of knowledge required and the amounts of material needed to develop the process during the early stages of development. Early process development evaluations of continuous manufacturing equipment often require larger amounts of material compared with batch, which hinder CM prospect for drugs during the early stages of process development. In this work, a small-scale evaluation of the mixing process occurring in a continuous mixing system was performed. The evaluation involved the use of a small-scale "mixing cell" which was able to replicate the lubrication process of a continuous mixer. It is worth mentioning that we designed the mixing cell by reconfiguration of an existing continuous tubular blender. The extent of lubrication evaluation was performed for three example formulations and was done by mimicking the amount of shear provided to a formulation by means of matching the number of paddle-passes that a formulation experiences within a continuous blending process in the batch mixing cell. The evaluation showed that the small-scale mixing cell was able to replicate the extent of lubrication-evaluated by measuring the tensile strength of compacts being made with both the continuous and mixing cell experiments-occurring in the continuous mixer using a fraction of the amount of materials needed to perform the same evaluation in the continuous blending process.

Robust state estimation of feeding-blending systems in continuous pharmaceutical manufacturing

State estimation is a fundamental part of monitoring, control, and real-time optimization in continuous pharmaceutical manufacturing. For nonlinear dynamic systems with hard constraints, moving horizon estimation (MHE) can estimate the current state by solving a well-defined optimization problem where process complexities are explicitly considered as constraints. Traditional MHE techniques assume random measurement noise governed by some normal distributions. However, state estimates can be unreliable if noise is not normally distributed or measurements are contaminated with gross or systematic errors. To improve the accuracy and robustness of state estimation, we incorporate robust estimators within the standard MHE skeleton, leading to an extended MHE framework. The proposed MHE approach is implemented on two pharmaceutical continuous feeding-blending system (FBS) configurations which include loss-in-weight (LIW) feeders and continuous blenders. Numerical results show that our MHE approach is robust to gross errors and can provide reliable state estimates when measurements are contaminated with outliers and drifts. Moreover, the efficient solution of the MHE realized in this work, suggests feasible application of on-line state estimation on more complex continuous pharmaceutical processes.

Using Residence Time Distributions (RTDs) to Address the Traceability of Raw Materials in Continuous Pharmaceutical Manufacturing

Continuous processing in pharmaceutical manufacturing is a relatively new approach that has generated significant attention. While it has been used for decades in other industries, showing significant advantages, the pharmaceutical industry has been slow in its adoption of continuous processing, primarily due to regulatory uncertainty. This paper aims to help address these concerns by introducing methods for batch definition, raw material traceability, and sensor frequency determination. All of the methods are based on established engineering and mathematical principles, especially the residence time distribution (RTD). This paper introduces a risk-based approach to address content uniformity challenges of continuous manufacturing. All of the detailed methods are discussed using a direct compaction manufacturing line as the main example, but the techniques can easily be applied to other continuous manufacturing methods such as wet and dry granulation, hot melt extrusion, capsule filling, etc.