Introduction to Crystallization of Fine Chemicals and Pharmaceuticals - Molecules: Nucleation, Aggregation and Crystallization - page 164

seeding of unwanted polymorphs. In the pharmaceutical industry, crystal-
lizations are generally carried out using the same equipment as is used for
other unit operations and processes, i.e. an agitated tank in a multipurpose,
multi-product plant. For the production of fine chemicals, there is usually
some margin for adapting the batch equipment to a particular crystallization.
The product from a crystallizer is generally filtered, washed and
dried. A significant degradation of the product size may occur during
these steps (Jancic, Grootscholten, 1984) and it is necessary to take this
into consideration when planning and designing crystallization.
The role of agitation
The purpose of stirring in a crystallizer is, among other things, to suspend
the crystals from the bottom in order to promote crystal growth. Crystals
piled up in stagnant zones in the equipment tend to agglomerate. In addi-
tion, stirring affects the distribution of supersaturation in the tank, perhaps
mainly through the distribution of supersaturation-consuming crystals.
Stirring is also important for the transport of heat from cooling surfaces:
gradients can be reduced and hence the risk of incrustation. Mixing condi-
tions affect the mass transfer between the fluid and the particles, and exert
a strong influence on nucleation. Normally, in cooling or evaporation crys-
tallization, agitation should be limited to the level where all crystals are just
suspended from the bottom. As the stirring is increased, secondary nucle-
ation increases faster than crystal growth, resulting in a lowering of the
product mean size. Depending on the stirrer type, the fractions of the energy
supplied as pumping and as shear and turbulence, respectively, differ
(Oldshue, 1983). Suspension of particles is best achieved with an axial flow
stirrer pumping downwards. The mixing energy is used more efficiently if
so-called baffles are used. Some good reviews of particle suspending have
been done by Nienow (1968), Mersmann et al . (1975) and Chapman et al .
(1983). The equation of Zwietering (1958) is still very useful for calculat-
ing the necessary stirring for complete off-bottom suspension.
For reaction crystallizations, the situation is somewhat different. In
the first place, the two reactant solutions need to be brought into contact