Biochemical Reactions - A First Look
67
It may seem strange that a metabolic product is unexpectedly being produced, but the complexity
of the metabolic network of microorganisms makes it quite possible to miss a product unless the
biochemical potential of the particular organism used in the experiment has been thoroughly
investigated. Thus, the unexpected conversion to gluconic acid of a large fraction of the glucose
used as feed in penicillium production was for a long time puzzling (Nielsen
et al.,
1994). The
appearance of oxalic acid at moderately high pH in citric acid fermentation (Example 3.6) can if
not detected and thereafter avoided lead to a large economic loss.
Stripping of some ethanol (and the even lower boiling acetaldehyde) from both aerobic and
anaerobic fermentations where N2 is sparged through the medium to give a completely 0 2 free
environment is a common cause of error in the mass balances. Even at low liquid phase mole
fraction ethanol has a high gas phase mole fraction and a substantial loss of ethanol results,
especially at low
D
since the production rate is small, while the rate of stripping is more or less
independent of
D.
Refluxing the medium, even with 1-2° C cooling water is of little help since
the heat transfer coefficient is small on the mostly dry gas phase side of the heat exchanger.
Feeding of substrate through the reflux condenser may be a practical way of avoiding this
problem since in that case a wet surface is present on the vapor side of the heat exchanger
(D ubocandvon Stockar, 1998).
Finally instruments, especially flow meters may give rise to systematic errors. When the gas flow
is low, e.g. 0.1-0.5 w m (volume gas per volume medium per minute), in laboratory bioreactors
calibration of flow meters can be difficult and the rate of gas flow vK
is in error. Here a high
accuracy of the instrument that determines the concentration of the gas-phase reactant in the gas
feed is of course of no help.
Examples 3.5 and 3.6 will illustrate how consistency tests of data may be of great help. These
tests should clearly be made during the experimental program and not retrospectively when all
the data have been collected and the experimental equipment is dismantled or is being used for
other purposes.
Example 3.5. Consistency analysis of yeast fermentation
One of the first experimental investigations of continuous aerobic yeast fermentation where the data was
of such generally high quality that they could be used - and still can be used - for quantitative
physiological studies was by von Meyenburg (1969) who worked in professor Fiechters group at ETH,
Zürich. Many later papers have used the data, e.g., Bijerk and Hall (1977) to set up a structured kinetic
model for aerobic yeast growth. The data are shown in Fig. 3.5. The stirred tank bioreactor was operated
as a chemostat, and the feed was 28 g L'1
sterile glucose solution with sufficient NHj (or rather NH 4 )
and other substrates to make the culture glucose limited. At low values of
D
no ethanol is produced and
the biomass concentration is high and approximately constant at 14 g L"1, except perhaps for a slight drop
for the lowest
D-
values. At these
D
-values the metabolism is purely respiratory, and the yeast obtains
sufficient ATP for growth by complete oxidation of glucose to CO:. The critical dilution rate
D„n
is in
the vicinity of 0.25 h'1
and above Dcri, the yeast starts to produce ethanol due to a bottleneck in the TCA
pathway or in the respiration (see also Example 7.3). Above Z)crit the mode of fermentation is called
respiro-fermentative. RQ increases rapidly and Tss drops sharply.
previous page 90 Bioreaction Engineering Principles, Second Edition  read online next page 92 Bioreaction Engineering Principles, Second Edition  read online Home Toggle text on/off