Publications

Authors: A. Dan and R. Ramachandran

Paper link: Link

Journal: Chemical Engineering Research and Design

Abstract: Milling is an important unit operation used as a particle design process in solid drug manufacturing processes. The size reduction process, where granules break into smaller fragments, governs the operation of the mill. The breakage behavior in a mill can be described through its breakage mechanisms, which are described using the critical screen size, breakage mode, and breakage rate. Breakage mechanisms are important because they determine the output quality attributes of the mill, such as the particle size distribution (PSD) and mass throughput. This study investigates the effect of process parameters and material properties on breakage mechanisms and determines its correlation to the output quality attributes. Additionally, dynamic data was collected and analyzed to determine the effect of process parameters and material properties on the transition of regimes and the evolvement of breakage modes. The findings in this study indicate that stronger granules undergo solely impact-attrition breakage mode, whereas weaker granules first undergo impact breakage mode and then transition into impact-attrition mode. Mass throughput is observed to be higher for impact mode than impact-attrition mode. Moreover, impact mode produces a wide and unimodal PSD, while impact-attrition mode produces a bimodal distribution.

Authors: Chaitanya Sampat, Lalith Kotamarthy, Rohit Ramachandran

Paper Link: Link

Journal: Pharmaceutics

Abstract: Twin screw granulation (TSG) is a continuous wet granulation technique that is used widely across different solid manufacturing industries. The TSG has been recognized to have numerous advantages due to its modular design and continuous manufacturing capabilities, including processing a wide range of formulations. However, it is still not widely employed at the commercial scale because of the lack of holistic understanding of the process. This study addresses that problem via. the mechanistic development of a regime map that considers the complex interactions between process, material, and design parameters, which together affect the final granule quality. The advantage of this regime map is that it describes a more widely applicable quantitative technique that can predict the granule growth behavior in a TSG. To develop a robust regime map, a database of various input parameters along with the resultant final granule quality attributes was created using previously published literature experiments. Missing data for several quality attributes was imputed using various data completion techniques while maintaining physical significance. Mechanistically relevant non-dimensional X and Y axis that quantify the physical phenomena occurring during the granulation were developed to improve the applicability and predictability of the regime map. The developed regime map was studied based on process outcomes and granule quality attributes to identify and create regime boundaries for different granule growth regimes. In doing so breakage-dominant growth was incorporated into the regime map, which is very important for TSG. The developed regime map was able to accurately explain the granule growth regimes for more than 90% of the studied experimental points. These experimental were generated at vastly different material, design, and process parameters across various studies in the literature, this further increases the confidence in the developed regime map.

Authors: Subhodh Karkala and Rohit Ramachandran

Paper Link: Link

Journal: International Journal of Pharmaceutics

Abstract: A discrete element method (DEM) model was developed for a co-rotating twin-screw mixer (TSM) consisting of conveying and kneading elements. This model was used to simulate the mixing of a cohesive binary system of particles (P1 and P2). The particles were calibrated using angle of repose and dynamic yield strength simulations. Properties of P1 were kept constant throughout the study while properties of P2 were varied. The TSM model was used to study the influence of particle size, density, and flowability of P2 on process outputs such as powder holdup, mean residence time (MRT), and degree of mixing along the length of the TSM. All the three particle properties were varied at three-factor levels with flowability ranging from good flow (GF), poor flow (PF), and very very poor flow (VVPF) based on Carr’s classification. The model provides fundamental insights into the differences in mixing in the conveying sections of the mixer for materials with different flowability. Due to the co-rotating nature of the screws, the majority of the material in the conveying sections was carried on one side of the TSM for GF and PF particles, while both the screws carried VVPF particles. The total steady-state holdup and the mean residence time (MRT) of particles inside the TSM were heavily influenced by particle flowability compared to the other two tested particle parameters. The holdup in the conveying compartments ranged from 11%–15% of the total holdup, while the kneading compartment comprised 40% of the total holdup in the mixer. The simulations found that VVPF particles mixed the fastest and that the GF particles had the highest tendency to demix. The model was also used to suggest and test other screw design configurations to improve mixing. A screw design with kneading elements closer to the mixer outlet resulted in better mixing of GF particles. A screw design consisting only of conveying elements achieved the same degree of mixing as the original screw design for VVPF particles.

Authors: Ahmed Zidan, Lalith Kotamarthy, Rohit Ramachandran, Muhammad Ashraf, Thomas O’Connor

Paper Link: Link

Journal: International Journal of Pharmaceutics

Abstract: This study aimed at understanding the effect of screw design on the critical characteristics of granules and tablets of an extended-release (ER) formulation for twin screw granulation process. The screw design parameters assessed included number of kneading elements (KEs) per kneading zone, distance separating kneading zones, staggering angle (SA) of kneading elements and number of sizing elements (SEs). These input variables were varied using a design of experiment (DoE) approach to manufacture granules. Particle size distribution (PSD), flow and bulk properties of the granules, breaking strength and dissolution of tablets manufactured using these granules were characterized.

The results of least square fitting showed that KEs, SA, and SEs of the screws significantly (p -values < 0.05) affected the PSD, cohesion, compressibility (CPS), conditioned bulk density (CBD) and permeability of the granules. The KEs and SEs significantly (p -value < 0.05) affected the dissolution, which was attributed to their effects on CPS and CBD of the granules. The distance between kneading zones had no significant effect on granules and tablet characteristics. These results may be used to further study the interaction of the identified critical screw design parameters with other processing parameters for continuous manufacturing of this ER matrix-based tablet formulation.

Authors: Chaitanya Sampat, Lalith Kotamarthy, Pooja Bhalode, Yingjie Chen, Ashley Dan, Sania Parvani, Zeal Dholaki, Ravendra Singh, Benjamin J. Glasser, Marianthi Ierapetritou, Rohit Ramachandran.

Paper Link: Link

Journal: Journal of Advanced Manufacturing and Processing

Abstract:  The global pharmaceuticals industry is a trillion-dollar market. However, the pharmaceutical sector often lags in manufacturing innovation and automation which limits its potential to maximize energy efficiency. The integration of techno-economic analysis (TEA) with advanced process models as part of an overarching smart manufacturing (SM) platform, can help industries create business models, which can be adapted for manufacturing to reduce energy consumption and operating costs while ensuring product quality which can further enable a more sustainable process operation. In this study, a rational design of experiment (DOE) on three unit-operations (wet granulation, drying and milling) was performed on a batch (case 1) and continuous (case 2) pharmaceutical process to obtain experimental data. Process models for predicting product quality and energy efficiency of each of the three-unit operations were developed. The experimental data were used to validate the models and good agreement was observed. The energy consumption of each unit operation was calculated using statistical models relating the power consumption and the process parameters. The developed process models and energy models were further integrated into a TEA framework, which quantified the energy and monetary cost of manufacturing for both batch.

Authors: C. Sampat and R. Ramachandran

Journal: Pharmaceutical Research

Paper Link: Link

Abstract: Quality risk management is an important task when it pertains to the pharmaceutical industry, as this is directly related to product performance. With the ICH Q9 guidelines, several regulatory bodies have encouraged the pharmaceutical industry to implement risk management plans using scientific and systemic approaches such as quality-by-design to asses product quality. However, the implementation of such methods has been challenging as assessment of risks requires accurate quantitative models to predict changes in quality when variations occur. This study describes a framework that quantitatively assesses risk for a twin screw wet granulation process. This framework consists of a physics-constrained autoencoder system, whose outputs are constrained using physics-based boundary conditions. The latent variables obtained from the auto-encoder are used in a support vector machine-based classifier to understand the granule growth behavior occurring within the system. This framework is able to predict the process outcomes with $~86\%$ accuracy and classify the granule growth regimes with a true positive rate of ~0.73. Based on the classification the risk associated with the process can be estimated.

Authors: Chaitanya Sampat and Rohit Ramachandran

Paper Link: Link

Abstract: With the advancement of digitization of industrial manufacturing, there has been an increase in the application of machine learning methods to model these processes. These data-driven models are multivariate in nature and on occasion may not deliver the accuracy that can be obtained from first-principle models. The statistical approach in data-driven models is completely data-dependent and may give erroneous or undesired results due to a noisy and incomplete database. Though accurate, first-principle models are often slow to simulate and lack the ability to predict data in real-time (C he n e t a l., 2020). Thus, to obtain real-time process predictions with accuracy similar to first-principle models, there is a need to develop data-driven models with first principle-based process constraints within their framework. In this study, several experimental datasets for twin-screw granulators (TSG) were considered. The data for 13 different TSGs were collected from previously published studies. The collected data was sorted for process parameters, material properties, and geometric conditions of the study. An autoencoder neural network was developed to model these processes. The output from this model not only predicted the data well but also showed granule growth characteristics with the output properties obeying first-principle laws. The encoding section of the neural network helped find correlated inputs creating a reduced-order model and capturing information about the underlying physics of the process.

Authors: L. Kotamarthy, X. Feng, A. Alayoubi, P. K. Bolla, R. Ramachandran, M. Ashraf, T. O’Connor, A. Zidan

Publication Page: Link

Abstract: Continuous manufacturing (CM) has been used to produce several immediate-release drug products. No extended-release (ER) product manufactured employing CM technology has been approved yet. This study investigated the critical aspects of switching from the batch mode of high shear granulation to the continuous operation of twin-screw granulation for extended-release tablets. Metoprolol succinate ER tablets were used as a model ER formulation for this purpose. A central composite design (CCD) was employed to determine the effects of high shear granulator (HSG) parameters, namely impeller speed, granulation time, and binder liquid feeding rate, on the critical granulation characteristics important for product performance. These critical granulation characteristics served as a guide for switching from batch processing to continuous operation for achieving the same breaking strength and dissolution for these ER metoprolol tablets.

Authors: S. V. Muddu, R. Ramachandran

Paper Link: Link

Abstract: This work is concerned with the incorporation of semi-mechanistic residence time metrics into population balance equations for twin screw granulation processes to predict key properties. From the historical residence time and particle size data sourced, process parameters and equipment configuration information were fed into the system of equations where the input flow rates and model compartmentalization varied upon the parameters. Semi-mechanistic relations for the residence time metrics were employed to predict the particle velocities and dispersion coefficients in the axial flow direction of the twin screw granulation. The developed model was then calibrated for several experimental run points in each data-set. The predictions were evaluated quantitatively through the parity plots. The root mean square error (RMSE) was used as a metric to compare the degree of goodness of fit for different data-sets using the developed semi-mechanistic relations. In summary, this paper presents a more mechanistic but simplified approach of feeding residence time metrics into the population balance equations for twin screw granulation processes.

Authors: Y. Baranwal, A. D. Roman-Ospino, J. Li, S.M. Razavi, F.J. Muzzio, R. Ramachandran.

Paper Link: Link

Abstract: Process analytical technology in the pharmaceutical industry requires the monitoring of critical quality attributes (CQA) through calibrated models. However, the development, implementation, and maintenance of these quantitative models are both resource and time-intensive. This study proposes the implementation of a non-linear iterative optimization technology (IOT) to study the magnitude of analytical errors when the calibration tablet used to extract the λ vector deviates physically and chemically from the test samples. IOT is based on mathematical optimization of excess spectral absorbance. It requires minimum calibration effort and allows simultaneous prediction of the entire formulation instead of only the active pharmaceutical ingredient (API), with just one standard and pure component spectral data. Unlike Partial Least Squares (PLS), which requires the development of standards to incorporate variations in the process, this non-destructive methodology minimizes significant calibration effort by developing a mathematical model that uses only one standard and spectral information of pure powders present in the tablet. The method described in this study allows a fast re-calculation to include factors such as change of spectroscopic instruments, variations in raw materials, environmental conditions, and methods of tablet preparation. The robustness of the proposed approach for variation in compaction (physical changes) and variation in composition (chemical changes) was evaluated for correlated and uncorrelated formulations. For uncorrelated formulation a PLS model was also constructed to compare the robustness of the proposed methodology. The RMSEP of API in target formulation predicted using non-linear IOT method was varied from 0.17 to 1.50 depends on compaction of tablet chosen to compute λ vector. On the other hand, the RMSEP of API in target formulation predicted using PLS-based model was varied from 0.13 to 0.57 depending on compaction of tablet. The additional accuracy achieved in PLS based model required significant calibration effort of preparing 84 tablets compared to just one in proposed non-linear IOT method.