little work has been carried out on the subsequent purification of the target solute from
the phase in which it is contained. Given the nature of the phases and solutes used,
distillation would appear to be a favoured operation (Mathys et al., 1998a; 1998b). Here
the influence of biological materials in the feed to the columns has, under certain
conditions, led to colouring of the product stream and poor separations due to
unfavourable gas-liquid hydrodynamics within the column. Clearly more research is
required in this area.
PROCESS SCALE-UP
Considerations for Scale-upThe mass transfer-reaction model previously shown in Figure 5.4 provides some insight
into the factors which might influence the performance of a stirred-tank reactor at various
scales of operation. In order to reliably scale-up a two-phase process an understanding of
how the individual parameters within the model will vary as a function of scale is
required. Generally it would be expected that the kinetic constants of the biocatalyst and
substrate/product partition coefficients would be scale independent. While the
substrate/product mass transfer coefficients may vary somewhat, it would be expected
that the most significant parameter with respect to scale-up would be the maintenance of
a constant interfacial area per unit volume. This will be critical in determining solute
mass transfer rates (flux being proportional to area per unit volume) and hence the
maintenance of substrate/product concentrations in the vicinity of the biocatalyst below
toxic levels. Rules for the scale-up of two-phase biocatalytic process are currently
lacking. Current work in our laboratory is examining how the droplet size distribution in
stirred-tank bioreactors changes as a function of scale (3–75 L vessels) in order to
rationally specify a scale-up basis; either constant power per unit volume or constant
impeller tip speed.
Scale-up will also require examination of several other factors. Figure 5.8 is a design
chart (plot of agitation rate in a stirred-tank against biocatalyst concentration) for a two-
liquid phase biocatalytic reaction. One operational limit occurs at high fractions of
aqueous phase substrate saturation. Above this concentration (defined by a particular
ratio of agitation rate to biocatalyst concentration) a loss of activity is observed over a
period of time. This loss of catalyst stability may be explained by the high liquid-liquid
interfacial area (relative to catalyst concentration) or by a time-dependent effect of a
relatively high aqueous phase reactant concentration, both of which occur at a high
operational agitation rate to biocatalyst concentration ratio. This, and the other bounds
shown in Figure 5.8 therefore define a process operating window (Woodley and
Titchener-Hooker, 1996). In scale-up, further bounds will be introduced
Advances in the selection and design of two-liquid phase biocatalytic reactors 137