Chemical Engineering September 2010 - 56

TABLE 1. KEY PAT TOOLS AND APPLICATIONS
PAT Tool
Pristine Processing
flow cell or transmission probe that
requires neither sample preparation
nor laboratory analysis [3]. The process
can be more tightly controlled,
problems detected earlier, and overall
product quality and yields improved.
Many chemical reactions used in
the manufacture of APIs use metal
catalysts. In some cases, these metal
catalysts are doped with carbon black
particles to increase the catalyst surface
area. The high sensitivity and
throughput of a Fourier-transform
near-infrared (FT-NIR) system can analyze
these solutions with carbon-black
concentration levels of 0.05-0.5%.
Catalytic hydrogenation reactions are
potentially sensitive to matrix effects.
Inline analysis provides kinetic and
mechanistic information that can be
used to study the sensitivity of the reaction
to matrix effects.
Optimal mixing
An agitator impeller can be designed
and operated such that agitator power
input is used to maximize flow with
minimum shear, maximize shear with
minimal flow, or to balance shear and
flow at a designated point in between.
However, solid-solid, liquid-liquid or
solid-liquid mixing or combinations
thereof are complex. For these applications,
high-shear mixers are typically
used because they can combine immiscible
fluids to produce emulsions and
dispersions through the controlled
formation and integration of droplets.
They are also used for deagglomeration
and particle size reduction to improve
the blending of solid ingredients
into liquids.
Many powders are difficult to disperse
efficiently and the resulting
mixtures are prone to persistent clogging,
which tends to drive up costs and
drive down throughput. Early indication
of these problems can control or
eliminate them.
Tomography involves taking measurements
around the periphery of
an object, such as a process vessel or
impeller, to determine what is going
on inside [4]. Each image (tomogram)
is composed of an array of derived
conductivity measurements. Process
vessels with all types of mixers and
accessories can be retrofitted with a
linear tomography probe, typically a
Near Infrared Spectroscopy (NIRS)
Attenuated Total Reflectance (ATR), Particle Video
Microscopy (PVM), X-ray diffraction (XRD),
Raman Spectroscopy (RS)
Heat Balance Calorimetry
Laser Diffraction (LD)
Digital Image Processing(DIP)
Dynamic Light Scattering (DLS)
Tomography
Turbidity Metering
Conductivity Metering
Thermodynamic Lyophilization Control (TLC)
Acoustic-Resonance Spectrometry (ARS)
baffle, to measure axial mixing and
solids concentration levels. The drug
and the diluents can be dry-mixed in a
double cone mixer, which gives ease of
cleaning and complete discharge. The
homogeneity of the mixture can be
confirmed using NIR measurements
with an FT-NIR spectrometer.
Solvent recovery
The solvent recovery process involves
the distillation of a feed or crude solvent
from a low purity of about 60%,
to a pure product that can be reused in
the manufacturing process. This process
is commonly analyzed by offline
GC, requiring an operator to collect a
sample from the process and take it to
a laboratory for analysis. This process
can take longer than 1 h and result in
the distillation of a product that does
not meet specifications.
Process FT-NIR has many benefits
over GC in this application, and, if implemented
correctly, can give realtime
analysis of the feed streams and the
product [5]. High-quality NIR spectra
can be collected usually in less than
1 min and the quantitative analysis
of multiple components performed.
This information is then used as a key
component of the control strategy to
change column temperatures or reflux
ratios in either a feedback or feed
forward control. This results in better
product quality and yield.
Crystallization
Pharmaceutical crystallization processes
are extremely complex and are
not baseline scalable. The key aspects
of a crystallization process are the rate
of nucleation and growth rate of the
crystals. Both of these rates depend on
the type of flow, mixing, heat and mass
transfer rates of the process in an unpredictable
manner. Impurities and
additives can have a significant effect
on these rates by altering the interfa48
CHEMICAL ENGINEERING WWW.CHE.COM SEPTEMBER 2010
Application in Unit Operations
Raw material testing, Drying, Powder
blending, Coating
Fourier Transform Near Infrared Spectroscopy (FT -NIR) Mixing, Reaction, Solvent recovery
Crystallization, Drying, Powder
blending, Coating
Reaction, Crystallization
Crystallization, Milling, Granulation
Crystallization, Milling, Granulation
Crystallization, Milling, Granulation
Mixing, Filtration, Drying
Crystallization, Filtration
Filtration
Lyophilization
Tableting
cial energy of particular crystal faces.
The main measures of the outcome of
a crystallization process, listed here,
are difficult to reproduce and achieve.
* Particle size distribution, or the relative
amounts of particles present,
sorted according to size
* Crystal shape (needle, cube and
plates)
* Polymorphic form - polymorphism
is the tendency of some compounds
to exist in more than one stable
crystal form
The size and shape of the crystals
formed has a strong effect on processability.
Smaller-sized and nonspherical
particles make flow, filtration and processing
more difficult. Fine particles
have greater surface area and can
collect more impurities at the surface.
The solubility of different polymorphic
forms may differ, and dissolution rates
will depend on crystal size as well as
polymorphic form.
Supersaturation measurement is
the key process variable that affects
crystallization performance [6]. To
control supersaturation, the solubility
curve and meta-stable zone width
(MSZW) need to be known.
Focused-beam reflectance measurement
(FBRM) can be used to measure
the solubility curve and MSZW by
identifying the point of dissolution
and point of nucleation at different
solute concentrations. FBRM provides
information about yield, suitable seeding
locations and optimal zones of operation
to get uniform particle size.
MSZW depends on cooling rate, agitation
and solid concentration.
Uneven or different particle-size distributions
may require further powder
processing, such as milling or compaction.
In some cases, the additional processing
costs can be reduced or eliminated
through realtime observation of
crystal size distribution using laser diffraction
(LD), and the progress of crys
http://WWW.CHE.COM
Chemical Engineering September 2010

Table of Contents for the Digital Edition of Chemical Engineering September 2010
