substrate concentration (0.6 mM) and the reaction rate (0.14 μmol min−^1 ml−^1 ) predicted
by the aqueous phase bulk reaction model and the measured values indicating that the
reaction occurs in the bulk of the aqueous phase. The Lewis cell may also be used to
measure the partitioning of substrates and products between phases. The generic use of
the Lewis cell lies in the ability it gives to expose biocatalysts to defined amounts of
interface and consequently the Lewis cell has a role in determining interfacial effects not
only upon biocatalyst kinetics as illustrated here but also upon biocatalyst stability.
This Lewis cell-based design method gives an estimate of the reactant mass transfer
coefficient required for a given reactor productivity which then enables a preliminary
selection of reactor type. Liquid-liquid contacting equipment, suitable for biocatalytic
reactions, may be characterised by the range of mass transfer duties (i.e. minimum and
maximum mass transfer coefficients) achievable in that particular design (e.g.
Doraiswamy and Sharma, 1984). Knowledge of the mass transfer coefficient required
will therefore eliminate some of these possibilities.
DOWNSTREAM PROCESSING
Reactor selection, design and operation all have an impact on subsequent product
recovery and the potential to recycle either the biocatalyst or the organic phase. It is
important therefore that design methods incorporate such reasoning. Figure 5.6 shows a
schematic integrated process design strategy. Target productivity, together with
preliminary selection of the biocatalyst form define the required mass transfer coefficient.
This then sets guidelines for initial reactor selection which subsequently set operating
parameters (for example, in a stirred-tank reactor: stirrer speed and phase volume ratio).
The effects of these decisions on the rest of the process must then be considered and if
there are problems for product recovery then either the catalyst selection or reactor
selection or (operation or some combination of these) may need to be re-examined. The
figure also shows that this is an iterative procedure and indicates where molecular genetic
methods may be applied to overcome design constraints associated with the biocatalyst.
Using graphical techniques (Woodley and Titchener-Hooker, 1996) many of the “what
if” studies can be done prior to experimentation so as to guide the development effort. A
number of experimental design tools can also be used here (Woodley and Lilly, 1992)
and preliminary studies indicate that interfacial tension measurements can be used to
predict the likelihood of emulsification.
To date only a limited number of authors have addressed the downstream processing
issues of two-phase transformations carried out in stirred-tank reactors. Whole process
flowsheets for the conversion of a poorly water-soluble substrate into either a poorly
water-soluble product or a water-soluble product are shown in Figure 5.7 (options for
recycling of the phases and the biocatalyst are also indicated). In both cases the key initial
Multiphase bioreactor design 134