Page 265 - Cascade_Biocatalysis_Integrating_Stereoselective_and_Environmentally_Friendly_Reactions

10.6 Process Modeling 241

10.5.2
Process Engineering
Despite the power of biology, in many cases it cannot provide all the necessary
solutions for industrial process challenges. Many implemented examples of single
step biocatalytic processes have clearly shown the need for innovative process
engineering solutions, as well as highly sophisticated molecular biology to engineer
the biocatalyst. Indeed it is highly likely that integration of catalyst design and
process design will be part of the next paradigm in chemical engineering [34].
Particularly interesting is that no single objective (e.g., lowest production cost or
lowest development cost), development route (e.g., protein engineering or process
engineering), or solution (e.g., operating with enzyme immobilization for 100
recycles using microfiltration or operating with soluble enzyme for 5 recycles
using ultrafiltration) exists in any given case. A clear need from the perspective
of process engineers is to develop a means to navigate the solution space in an
effective way. In the pharmaceutical sector, the time limitations (as a result of a
limited patent lifetime) mean that the emphasis is on speed of development. It is
clear that automated, systematic methods of data collection, linked with design of
experiments and process models will have huge benefits in much the way they have
already in other sectors of the chemical industry. Process engineering strategies
such as feeding of substrates [28] and removal of products during the reaction
and in situ product removal (ISPR) [35–38] will also be required in multienzyme
processes. In such cases, an added degree of freedom can also come from spatial
or temporal changes in the reactor system, adding significantly to the complexity
of the problem.

10.6
Process Modeling

Process modeling is increasingly implemented as a means of mathematically
describing bioprocesses. It is of course easiest for enzyme-based biocatalysis [39],
but is also necessary for complex fermentations where population based models
are required. Two types of models need to be developed – those that describe the
reaction phase and those that describe the associated unit operations and process,
via mass balances. Much progress has already been made but more sophisticated
models are required to enable a more predictive approach for scale-up and design.
This will also be an important contribution from chemical engineers in the future
as we move from empirical to more mechanistically based models. Alongside this,
it will be necessary to build property databases of suitable feed-stocks, reagents,
and chemicals. In many cases, predictive tools for the properties of many of these
molecules would perhaps be even more useful, to save valuable experimental time.
The chemistry, in particular in many processes where biological catalysts can best
be exploited, is complex and the building of a suitable database and predictive tools
will be an important contribution. Kinetic, thermodynamic, and process models
for multienzyme processes are particularly valuable [40] because of the complexity