(52f) Application of DEM Simulations and Experimental Studies to Improve the Uniformity of An Active Coating Process

Toschkoff, G. - Presenter, Research Center Pharmaceutical Engineering
Just, S., Heinrich Heine University
Knop, K., Heinrich Heine University
Kleinebudde, P., Heinrich Heine University
Funke, A., Bayer Pharma AG
Djuric, D., L.B. BOHLE Maschinen + Verfahren GmbH
Scharrer, G., RCPE GmbH
Khinast, J. G., Graz University of Technology

Introduction: Pan coating is commonly
used to apply a functional layer to tablet cores. Tablets are placed in a
rotating drum, and a coating liquid is sprayed onto the moving tablet bed. For
most applications, the uniformity of coating (the equal distribution of coating
over a single tablet and/or between different tablets of a batch) is of great
importance. This necessitates a high degree of process understanding. A number
of investigations were done both experimentally [1] and in recent years also
computationally using the Discrete Element Method (DEM) [2], [3].

Objective: In this work, the application
of an active coating was investigated: Gastrointestinal therapeutic systems
(Adalat® GITS, Bayer Pharma AG, Germany) were coated with an aqueous
suspension containing candesartan cilexetil as active pharmaceutical ingredient
(API). As the variation of this API between tablets has to be small, a high
inter-tablet coating mass uniformity is vital. The central aim therefore was to
improve the inter-tablet coating uniformity.

Methods: To study how different
process parameters influence the uniformity, both experiments and DEM
simulations were performed. A lab pan coater (BFC 5, L.B. Bohle, Germany) with
two or four spray nozzles (Düsen Schlick GmbH, Germany) was used. The process parameters
were set according to the same statistical plan in experiments and simulations.
The results of both are compared, and guidelines for optimization are

Experiments: The factors pan load, pan
speed, and spray rate were investigated in two 23 full factorial DoEs
with three centre points. For each DoE, either two or four spray nozzles were
used. The response variable was the coefficient of variation (CoV) of the API
amount in the coating layer at the coating endpoint. To get the API amount, 20
active coated tablets per trial run sampled at the endpoint were analysed by
HPLC (Elite La Chrom, VWR Hitachi, Germany).


The Discrete
Element Method (DEM) simulations were done using commercial software (EDEM 2.4,
DEM Solutions, UK). The geometry of the coating apparatus was provided by the
manufacturer. The material properties came from measurements [4]. Two methods for the
modelling of a spray in DEM simulations were used (during run-time and in
post-processing). They gave detailed information on the spray process, such as the
coating mass of each tablet or the cycle and spray zone residence times.

23 full factorial DoE (the same as in the experiments) were
performed; for each run 90 seconds of process time were simulated. The response
variable was the CoV calculated from the individual coating mass on the
tablets. Further, additional simulations were done to provide more data points
and to study the impact of e.g. axial tilt or spray nozzle setup.

& Conclusion:
the presented work, DEM simulations and experiments were performed to
investigate active coating. The results are compared, and the insights
concerning the optimization of uniformity are presented. For a deeper
understanding, the outcome is also interpreted in the light of known analytic expressions
for e.g. the time development of coating uniformity (cf. Fig. 1). It was
possible to decrease the variation between tablets to values well below 6% as
required from regulatory authorities, and to reduce the total process time.

1: CoV of coating mass with increasing time for different process parameter
settings from the DEM simulations. Random mixing models predict a decrease
inversely proportional to the square root of coating time. (line of slope of
-1/2 in the figure). After ~10 seconds, tablet movement is randomized, and the
DEM results agree well with model expression.



[1]          D. Brock, J.
A. Zeitler, A. Funke, K. Knop, and P. Kleinebudde, ?A comparison of quality
control methods for active coating processes.,? International Journal of Pharmaceutics,
vol. 439, no. 1?2, pp. 289?95, Dec. 2012.

[2]          B.
Freireich, J. Li, J. Litster, and C. Wassgren, ?Incorporating particle flow
information from discrete element simulations in population balance models of
mixer-coaters,? Chemical Engineering Science, vol. 66, no. 16, pp.
3592?3604, 2011.

[3]          A.
Dubey, F. Boukouvala, G. Keyvan, R. Hsia, K. Saranteas, D. Brone, T. Misra, M.
G. Ierapetritou, and F. J. Muzzio, ?Improvement of tablet coating uniformity
using a quality by design approach.,? AAPS PharmSciTech, vol. 13, no. 1,
pp. 231?246, Mar. 2012.

[4]          S.
Just, G. Toschkoff, A. Funke, D. Djuric, G. Scharrer, J. G. Khinast, K. Knop,
and P. Kleinebudde, ?Experimental Analysis of Tablet Properties for Discrete
Element Modeling of an Active Coating Process,? AAPS PharmSciTech, 2013.