Process Control and scale-up

of Pharmaceutical Wet
Granulation Processes:
A Review (A. Faure et al.)

Mandeep Singh
2014A1PS0751G
Introduction
Widely used process in the pharmaceutical industry
At low drug concentrations, it provides drug content uniformity
At high drug concentrations, it controls product bulk density and
compactibilty
Types of Granulating Sysytems:
High-shear mixers
Fluid bed granulators
Fluid Bed Granulation
Powder Mix is maintained as a fluidized bed by a flow of air injected
upwards through bottom screen of granulator
Binding solution is sprayed above the powder bed (in a direction
opposite to air flow)
The granules result from the adhesion of solid particles to the liquid
droplets that hit the bed.
Partial drying by the fluidizing air occurs continuously during
granulation.
The process continues until all the powder has been agglomerated
Fluid Bed Granulation (cont.)
Moisture Balance needs to be maintained
The equilibrium may not be constant, however, as
The moisture content of the granules could be increasing slightly throughout
the process.
Trajectories of the particles may change with changes in the density of the
agglomerated powder bed.
Complete drying is quickly achieved in the hot air stream when binder
spraying is stopped.
High-Shear Granulation
An impeller maintains the powder in agitation in a closed vessel and a
binder solution is sprayed from the top.
As the liquid droplets disperse in the powder, they form the first
nuclei of future granules.
The agitation forces prevent the development of large agglomerates,
because they would be too fragile to sustain the shear.
However, as mixing and spraying proceed, the existing agglomerates
undergo densification and the internalized binder is squeezed out to
the surface of the wet agglomerates.
It makes the agglomerates harder, and their surface more adhesive,
and hence granule growth enters a new, more efficient phase.
High-Shear Granulation (cont.)
The process is stopped somewhere in this phase before an excess of
liquid leads to formation of a slurry.
The drying step traditionally takes place after transferring the damp
mass into another piece of equipment (fluid bed dryer), but the use
of single-pot technology (drying in place) is now spreading.
The granules formed are denser than those obtained in fluid bed
granulation.
Control Parameters
Material Parameters
Granulation Conditions
Drying Conditions
Granule Properties
Tablet Properties
Material Parameters
It includes
Powder particle size distribution
Wettability of the solid by the liquid
Solid solubility and degree of swelling in binder liquid
Binder concentration and viscosity
The nature and characteristics of the ingredients entering the
formulation are almost always fixed.
Granulation Conditions (high-shear mixing)
It includes:
Mixing/collision generation levels
Process time
Fill level
Liquid spray rate
Quantity of solvent
Temperature
Spraying conditions are not much critical due to fast dispersion of
binder.
The main parameters are mixing conditions and proportion of liquid
used.
The principal difficulty here is the control of end-point or to decide
when to stop
Granulation Conditions (fluid bed)
It includes:
Spray droplet size
Spraying surface and rate
Quantity of solvent
Bed fluidity/air flow rate
Inlet air temperature and Relative Humidity
Equilibrium temperature and Relative Humidity in bed
Process time
The spreading of the binder liquid droplets in the powder bed
controls most of the agglomeration.
Droplet size, contact with the powder bed and humidity in the bed
are key parameters.
Variables listed above are inter-related which makes the problem of
controlling the process more complex.
Drying Condition
In case of high shear mixing, parameters include:
Extent of mixing
Mode of drying (air stripping, microwave, infrared)
Energy input
Process time
In case of fluid bed granulation, parameters include:
Inlet air temperature and Relative Humidity
Air flow rate
The drying end-point is detected by a sudden rise in bed and outlet
air temperature, and an equalization of the outlet air dew point to
that of the inlet air.
Granule properties
It includes:
Particle size distribution
Bulk density and porosity
Moisture content
Drug content uniformity across the particle size distribution
Binder distribution
Granule strength/friability
Tablet properties
It includes:
Tableting conditions(compaction force and speed)
Extra-granular additions: e.g. lubrication, extragranular disintigrant
Process control and scale-up in high-shear
mixers
Using power consumption as a monitoring parameter
Monitoring parameters other than power consumption
Experimental designs for use in process control
Use of population balance models
Using power consumption as a monitoring
parameter
Variations in power consumption during granulation can be attributed
to the evolution of the strength of the wet agglomerates.
Monitoring of the impeller amperage, power consumption, torque or
indirect torque are affiliated techniques.
In an experiment, the power consumption was found to be related to
the extent of densification of the wet mass:
is the intragranular porosity.
There is a relation between intragranular porosity
and S, saturation level which is the ratio of pore
volume occupied by liquid to total pore volume [H is mass ratio of
liquid to solids, density of particle is relative to that of liquid ]
Saturation Level
As power is related to intragranular porosity, so power is also related
to S.
While the relationships between the granule diameter and the
porosity of the wet granules remain equipment and scale-dependent
,the saturation level provides the missing parameter to establish the
link from one machine to another.
This parameter can be more easily monitored following the power
consumption with time rather than by retrieving samples for the
determination of their saturation level, a measurement that remains
somewhat delicate and time consuming.
Wet Mass Consistency
Wet mass consistency depends on the deformability of the wet granules
which is a function of the distribution of the liquid in the available pores.
A mixer torque rheometer is used to quantify the WMC of granules.
WMC is easier to determine than S . Also relationship between WMC
and dry granule properties has been demonstrated in an experiment.
There is a dimensionless relationship established between WMC and
power consumption which lets us use the relationship for various scales.
Optimal WMC, is found and a dimensionless relation between Power
and WMC is found by performing series of granulation experiments.
Optimal WMC can be found at small scale and can then be applied at
large scale to get power consumption.
Cont.
Power consumption curves are notoriously unpredictable in high-shear
granulation.
Conditions for binder spraying and irregular wall adhesion in time make the
curve of power consumption versus time fluctuate and decrease in
reproducibility.
Some authors have tried treating the signal (e.g. by Fast Fourier Transform)
Otherwise it is possible but tedious to analyze the dry granules obtained at
different power consumption levels to determine at which level the best
granule and the best yield are obtained
There are no rules for scale-up, except trial and error.
Monitoring parameters other than power
consumption
Vibration probes
The vibration probes tested were able to record a signal that reflected the
level of densification achieved by the wet granules
Moisture sensors
Conductivity probe
Capacitance sensors
Infrared sensors
Experimental designs for use in process
control
The design of experiments approach can be a powerful tool to model
processes.
In the most common type of experimental designs, a number of input
variables are selected among the process conditions and material
properties. ( impeller speed, binder concentration, mixing time, etc).
A number of output variables are also selected , usually the granule and
sometimes also the tablet properties.
The principle is to determine for each output variable, a model involving
each input variable and the interactions between variables.
Variables and interactions which do not contribute significantly to the
model are withdrawn from the model.
Use of population balance models
The population balance models focus only on the PSD of the granules.
The models are extremely complex and require several assumptions
to be made for simplification and solving.
They require great understanding of the high-shear mixer to make a
reasonable estimate for the assumptions made.
The entire granule size distribution is divided in small intervals
At the core of the problem lies the estimation of the probability of
coalescence when a granule from size interval i collides with a granule
of size interval j.
The results are compared to experimental data, thereby checking the
validity of the models.
At the moment, the models are used as a learning tool to understand
processes and mechanisms taking place.
Process control and scale-up in fluid bed
granulators
Specificities of the fluid bed granulation process
Methods used for process control and scale-up
Specificities of the fluid bed granulation process
Liquid distribution slow, so difficult to stabilize.
Shear forces are minimal, so liquid less likely to squeeze out , hence
limited coalescence.
Initial spreading of binder very crucial in determining future granular
size distribution.
Viscosity of binder solution also has great effect on final granules.
When a droplet hits the bed, then liquid extends to entire surface of
particle wetted, hence giving rise to bridges between neighbouring
particles. When agglomerate formed is wet enough, it collides with
other particles.
Granule drying takes place in the bed itself.
Predicting and modelling how bed will evolve, drying and
agglomerate circulation , density changes is a very complex task.
Cont.
In scale-up, initial wetting of powder can be easily modelled.
Spraying zone depends on spray angle, and its position above bed.
When nozzle is low, gives rise to clogging, when high, leads to wetting
of walls. Number of nozzles is multiplied keeping the size of droplet
constant in scale up.
Evaporation needs to be kept in check, spray drying of binder
undesirable. Interval of Renewal of powder bed also important.
Initial nucleation is followed by fast agglomerate growth rate which is
followed by slower granule growth phase, a transition phase.
In this phase, there is very less ungranulated powder. Due to process
being slow, easier to control and end point is found in this phase.
Process Monitoring done through image processing. It can also be
done though monitoring of bed moisture and related parameters.
Trial and error experiments are needed to determine amount and
time of binder required to get desired granule size.
Cont.
The granules not completely dried are sent back to wetting zone
Humidity needs to be kept in check using relative humidity probes.
Bed temperature can also be monitored as it is related to inlet air
temp. and evaporation rate of binder solution.
Nowadays, infrared probes are used . Output from these probes can
help to adjust the binder spray rate manually or automatically.
We can modify inlet air temperature to control evaporation rate.
Although drying capacity of air depends on dew point but due to use
of dehumidification systems , air humidity is generally under control.
Air flow rate could also be adjusted as the granulation proceeds
Methods used for process control and scale-up
Due to many interaction between the input variables ,it is extremely
hard to optimize process by one parameter at a time concept.
Fuzzy logics and neural networks have been used to control such
processes..
In the computer models generated, the output variables are the PSD
of the granulate, bulk density and shape factors of the agglomerates.
Some self-organizing maps have been developed to represent in two
dimensions the effect of the different process variables over the
evolution of the granulation process.
With this technique we can visually detect deviations from the
reference, and can take action to bring back the process on its normal
pathway within the self-organizing map
Conclusion
Quality control at early stages of drug product development such as
wet granulation process has become extremely important.
In macro scale, Empirical techniques have been developed for process
control and scale-up in high shear mixers, while computing
techniques (e.g. neural networks, fuzzy logic, self-organizing maps)
are being generated for monitoring the more complex fluid bed
granulation process. There is a concept of agitation fluid bed
granulators where there is an impeller at the bottom of the bed
improving densification.
In micro scale, fundamental knowledge of the processes is
progressing, with developments in areas such as granulation regime
mapping and population balance modelling.
Thank You