given by Kioukia et al. (1992), who suggest that, although the mean bubble diameter was
equal in both situations, the distribution of bubble sizes might have been different. More,
smaller bubbles in the case of the stainless-steel distributor would result in higher death
rates.
In conclusion, there is only one reported case in the literature indicating cell death at
the sparger site. It seems unlikely that this is due to liquid flows in this region, although
there is no clear proof for this. Furthermore, if adsorption of protective additives cannot
keep up with the rapid expansion of the bubble surface at the sparger, bubble surface is
created that is not or only partially covered with surfactants. If this surface comes into
contact with a cell, it may lead to direct cell death in the case of surfactant-free surface or
stable adsorption of the cell in the case of partially covered surface. However, to date
there is no experimental proof for this mechanism either.
Cell Death During Bubble RiseMost studies on overall theories suggest that the specific death rate is proportional to the
reciprocal height of the reactor or, in other words, the killing volume is independent from
the residence time of a bubble (Tramper et al. 1988, Jöbses et al., 1991, Martens et al.,
1992).
Hülscher and Onken (1992) derived an equation assuming that during the rise of a
bubble, surface-active components adsorb to the bubble interface. Cells that come into
contact with uncovered surface are eventually killed. They obtained the following
equation to describe the first-order death-rate constant:
(13)
where vb (m.s−^1 ) is the rise velocity of the bubble, vl (m.s−^1 ) is the liquid velocity, Asb (m^2 )
is the cross-sectional area of the bubble, κ (–) is a model parameter containing the
collision frequency and the tendency of a cell to interact with the bubble, r (s−^1 ) is the
surfactant adsorption rate, which is proportional to the surfactant concentration and Vb
(m^3 ) is the bubble volume. This equation again compares well to equation (3) of the
hypothetical-killing-volume theory with
(14)
The model is in accordance with the results of Jordan et al. (1992) and Michaels et al.
(1995), who showed that cells may interact with surfactant-free or partially saturated
bubbles, which either leads to direct cell death or to cell attachment. Both groups also
showed that surfactant adsorption to bubbles is very rapid. Jordan et al. (1992) showed
that for medium containing one percent serum, bubbles were already saturated before
they reached a height of 2.5 cm above the sparger. For equation (14) this would mean that
for columns higher than this 2.5 cm Vd becomes height independent and the death-rate
constant becomes proportional to the reciprocal column height. Since most sparging
experiments are done in columns higher than 2.5 cm, this is in accordance with the
finding that the killing volume is independent from reactor height. Furthermore,
Multiphase bioreactor design 468