between values of about 4 and 6. The stability of foam, once formed, is further enhanced
by the presence of ECPs.
As with all fermentation processes, it is preferable to prevent foam formation rather
than to destroy it, once formed. Efforts to prevent the onset of foaming include reducing
aeration and agitation rates (without jeopardising system mixing and mass transfer) or,
where practicable, employing surface aeration or bubble free-aeration. The most common
approach is the use of antifoam, added directly to the medium prior to inoculation and/or
as required during cultivation. Studies of the biocompatability and effectiveness of a
range of antifoams and other surfactants have been performed in shake flask cultures and,
more valuably, under sparged conditions (e.g. Zhong et al., 1992a; Li et al., 1995). While
the interaction between antifoaming agents and mass transfer is complex, experience with
plant cells suggests that addition of antifoam is likely to reduce kLa values (Smart and
Fowler, 1981; Wongsamuth and Doran, 1994). Mechanical foam breakers have
traditionally been avoided with plant cell systems, for reasons of cost, complexity of
operation and possible shear damage.
Once foam formation commences, it is difficult to prevent wall growth. As a general
rule, it is advisable to minimise the internal surfaces available for wall growth by, for
example, employing jackets rather than internal coils for temperature control. Removal of
baffles has also been reported (Treat et al., 1989) although, as this strategy will reduce
bulk mixing efficiency for most agitation systems and is conducive to vortexing, it is not
generally recommended. There is experimental evidence to show that substantially
reduced Ca2+ levels in the medium (0.1–0.5 mM, as opposed to 3 mM in control medium)
can significantly reduce foaming and wall growth in both shake flask (Takayama et al.,
1977) and chemostat (STR) cultures (Sahai and Shuler, 1982).
Data on foaming in large scale systems are sparse. Hashimoto and Azechi (1988)
reported foam, rising to a height of 2 m during the exponential phase of a batch
fermentation in a 20 m^3 vessel (10–15 m^3 wv). The performance of a 300 L, external loop
ALR for the cultivation of P. somniferum (Park et al., 1992) was limited by foaming. It is
possible that the absence of more abundant data may reflect the fact that in commercial
scale STRs, foaming may be less severe, due primarily to increased vessel aspect ratios
(H/T) and lower aeration rates than those commonly employed in laboratory systems.
Bioreactor Configuration and Operation ModeThe design of bioreactors for microbial systems is well-established (e.g. Winkler, 1990;
Charles and Wilson, 1993). The same basic objectives apply to bioreactors for plant cell
suspensions, namely the extended provision of reproducible conditions, supporting
optimal system performance, evaluated in terms of biomass and/or product formation.
To date, most plant suspensions have been cultivated in either STRs or ALRs.
Strategies for optimising the hydrodynamic environment in STRs have focused primarily
on modifications to standard agitation (Section on Mixing and Mass Transfer) and
aeration (Section on Aeration in Plant Bioreactors) regimes. Examples of variations on
basic STR and ALR designs as well as rotating drum reactors, employed for plant cell
suspensions, are summarised by Su (1995).
Assuming that a product market exists, there are two main routes to enhancing the
commercial viability of plant cell-based processes, namely increasing system productivity
Multiphase bioreactor design 440