Developments in Pharmaceutical Process Modeling

Published on: 

Equipment and Processing Report

Equipment and Processing Report, Equipment and Processing Report-06-19-2013, Volume 0, Issue 0

Model-predictive design is applied to solid-dosage processes.

Model-predictive design—used to define, predict, and control a process—is well established in many industries and is beginning to take hold in pharmaceutical process-development and manufacturing. Modeling offers a way to improve process understanding and control and to speed development. In response to this trend, software tools that have improved capabilities and increased user-friendliness are being designed to meet the needs of pharmaceutical-industry users. Although modeling is relatively established in API-production processes, modeling of solid-dosage processes is less mature; software tools for modeling solids behavior are at the cutting edge. A wide range of tools can be used to model solids behavior, including computational fluid dynamics, population-balance models, and discrete-element modeling (DEM).

DEM discretizes a granular domain into elements and computes the interaction forces between the elements, which results in a detailed description of solid material movement, explains Richard LaRoche, PhD, vice-president of Engineering at DEM Solutions. DEM is being used to model solid-dosage unit operations including mixing, coating, powder feeding, granulation, and compaction. "A well-calibrated model can provide useful process information that is difficult or impossible to measure experimentally. For example, simulation and analysis can show shear inside a mixer, real-time forces in a tablet coater, segregation within a feed frame, and change in flow properties as the material is mixed or operated upon by devices (e.g., mills)," says LaRoche. DEM Solutions recently made available a searchable DEM literature database containing over 3000 abstracts of papers involving DEM modeling, including many topics relevant to pharmaceutical solid-dose production.

As with any new tool, there are challenges in adapting DEM to the needs of the pharmaceutical industry. One challenge is for the industry to better accept modeling in general and to understand the power, scalability, and adaptability of DEM, says LaRoche, who adds that he does see, however, an increasing acceptance of DEM-based modeling methods within the pharmaceutical industry.


Another challenge is that modeling has traditionally been the tool of chemical engineers rather than chemists (who make up a large segment of the pharmaceutical industry). New modeling software interfaces, however, are more straightforward and incorporate online training that is designed to make the software more user-friendly for a wider range of users, notes Jonathan Kadane, director of product marketing for Pharmaceuticals/Life Sciences at AspenTech. Other improvements include a special search tool in the software that allows users to efficiently find and access models and data and a feature that allows users to identify ways to change designs to reduce energy or capital cost. Most recently, the company added a web-based user interface for its aspenONE software, which empowers more users to work with the software without specialized training using touch-enabled devices (i.e., through simple touch and swipe gestures on a tablet).  

For further discussion on modeling that includes additional comments from AspenTech and DEM Solutions as well as input from other modeling software providers, such as Accelrys and CD-adapco, and pharmaceutical industry members, see "Using Model-Predictive Design in Solid-Dosage Manufacturing Processes" in the June issue of Pharmaceutical Technology. Also in the June issue, read “Model-Predictive Design, Control, and Optimization,” written by members of the Engineering Research Center for Structured Organic Particulate Systems at Rutgers, the State University of New Jersey, which describes application of model-predictive methods and a continuous process-control framework to a continuous tablet-manufacturing process.