Automated Measurement and Monitoring of Bioprocesses: Key Elements of the M3C Strategy - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 7

experience), and the manipulated variables may be the stirring power or speed,
aeration rate, pressure, or fresh gas composition (or combination thereof).
The classical pO 2 electrodes are membrane-covered amperometric electrodes,
so-called Clark(-type) electrodes. The membrane separates the measuring solution
from the electrode. The flux of oxygen through the membrane is diffusion con-
trolled but decisive for the sensitivity and response time. This is also why the
electrode reflects the partial pressure rather than the dissolved gas concentration.
The membrane is also responsible for the selectivity towards oxygen, since it
permits only gases and volatiles to diffuse. Membranes are usually made of silicon
or Teflon, or are a sandwich of both. The inner electrode consists of an anode made
from Ag and a cathode made from Pt, and the potential between them is usually
tuned to between 600 and 700 mV, but kept constant. Under these conditions,
oxygen is consumed (namely reduced) at the cathode, while four electrons per O 2
come from the anode, which converts to 4 Ag + (and finally, in a chloride-con-
taining electrolyte, AgCl, which must be removed from time to time in order to
keep the Ag accessible). The electron flux is low (typically in the range of nA) and
must be amplified to yield a useful signal. The rate-limiting step is the diffusion
through the membrane, which makes the electrode slow at low pO 2 .
Optical alternatives have been commercially available for several years. They
consist of an immobilized ruthenium or platinum complex that fluoresces in the
absence of oxygen; this fluorescence, however, is quenched in the presence of
oxygen. Oxygen is not consumed by these sensors, nor are parts of the sensor itself
consumed (such as the Ag anode in the amperometric type). Furthermore, there is
no more need to change the inner electrolyte of the electrode once in a while and,
probably, the patch-holding electrode tip is mechanically more resistant than the
membrane and needs less frequent exchange. The very thin patch of the immo-
bilized complex, which must be biocompatible and sterilizable, can be glued onto
a (glass) window, a shaking flask, a translucent plastic container, or directly on top
of fiber optics; the patches are inside while the optics are outside the sterile barrier
[ 19 ]. Contrary to the amperometric sensor, the signal is maximal at minimal pO 2 ,
and therefore the signal-to-noise ratio is best under oxygen-limited conditions. We
propose to use both amperometric and optical pO 2 sensors at the same time
(mounted close to each other) to take full advantage of their different sensitivity
regimes [ 20 ]. More detailed information about optical sensors can be found in the
review by Lam and Kostov [ 21 ].
2.2.4 Carbon Dioxide Partial Pressure, pCO 2
pCO 2 electrodes are membrane-covered pH electrodes. The membrane separates
the pH electrode from the measuring solution. The flux of carbon dioxide through
the membrane is diffusion controlled, as in pO 2 electrodes. CO 2 , an acidic gas,
dissolves in the inner electrolyte buffer and provokes a pH drop which is detected
by the pH electrode, whose signal is proportional to the (decadic) logarithm of
pCO 2 . This provides an elegant means to estimate the exponential character of a