FOAMING AND ITS CONTROL IN BIOPROCESSES F ... - Springer Link

0 downloads 0 Views 3MB Size Report
Dept. of Chemical Engineering ... property which distinguishes a foaming solution from a non-foaming solution ..... It must also be present in a sufficient quantity to main ..... some knowledge of the structure-dependence of the chemical, physical.
FOAMING AND ITS CONTROL IN BIOPROCESSES

F. VARDAR-SUKAN Ege University Faculty of Engineering Dept. of Chemical Engineering 35100 Bornova, Izmir, Turkey ABSTRACT. Control of foaming in bioprocesses presents a dilemna, since a highly surface active heterogeneous system has to be adequately aerated and agitated while the bioreaction is proceeding without interruption. This paper deals with mechanisms of foam formation and destruction, causes and effects of foaming in bioprocesses and different methods of foam destruction and prevension. Special emphasis is given to chemical anti foaming agents and to the determination of their efficiencies. Important criteria for the choice of a suitable anti foaming agent are discussed. It is concluded that foam suppression or collapse is a result of a fine balance between the effects of var ious surface active agents. Therefore, in choosing and employing an anti foaming agent, the aim, foam suppression or collapse, should be well defined and the interactions between biomedia and the anti foaming agent should be experimentally determined. Finally, a few suggestions for future research are mentioned. 1.0. INTRODUCTION Foaming is a serious problem often encountered in bioprocess industries, particularly when the system has to be adequately aerated and agitated, while foam formation is kept under control. The complexi ties of biosystems make it di fficult to relate their foaming characteristics to individual factors and qualitative differences exist between foaming abilities of liquids and types of foam produced. Many factors contribute to the stability and pattern of foam formation. In submerged culture, foaming is associated with hydrodynamic conditions which in turn are affected by the introduction of gas, the composition of the medium, the presence of growing cells, the formation of metabolites and surface-active substances. Foam production is usually autocatalytic in its generation. Small amounts of foam can create conditions that may promote lysis of cells and this may in turn lead to heavier foaming. The problems created by 113

F. Vardar-Sukan and S. S. Sukan (eds.), Recent Advances in Biotechnology, 113-146. @ 1992 Kluwer Academic Publishers.

114

excessi ve foaming fall into two classes; those which are caused by its appearance wi thin the reactor and those which are caused by its escape if some control is not exercised. Although foam control systems are used extensively in industry, preventing the formation of foam is preferable if this can be attained without any deleterious effects on the productivity of the process. However, this is indeed a formidable endeavour since foaming is a result of very complex interactions amongst the initial biomedia components, products of biochemical reactions and all the other surface active substances that may be present as well as the operating conditions of the bioreactor. Extensive research is still required to elucidate the exact mechanism of these relationships so that realistic predictions of the foaming behaviour of bioprocesses can be carried out. 2.0. THEORY 2.1. Definition Foams are comprised of thousands of tiny bubbles of mechanical or chemical origin and are generated wi thin a liquid. I f these bubbles rise and accumulate at the liquid surface faster than they decay, foaming occurs. Gas makes up the larger volume portion of such a foam, therefore the bulk density of the system approaches that of a gas rather than a liquid. The bubbles are separated only by a thin liquid film. True foaming only occurs when the intervening liquid between two bubbles thins down to a lamella, instead of rupturing at the point of closest approach. Although foam is defined as a dispersion of gas in liquid, it di ffers from the more common gas dispersions in liquid because the distance between individual bubbles are extremely small and the volume ratio of gas to liquid is rather large. The presence of foam, and more particularly its extent, represents a balance between the processes creating it and the forces causing its destruction. The high speci fic surface areas of foams make them thermodynamically unstable as in the separated state, the gas and liquid have a lower surface energy. The complexity of foaming systems makes them unsusceptible to mathematical analysis and treatments starting from fundamentals have generally needed considerable number of simplifying assumptions before a reasonable correlation with experimental observations could be obtained (Thomas and Winkler, 1977). It is generally accepted that pure liquids do not foam. The property which distinguishes a foaming solution from a non-foaming solution is its ability to resist local thinning of a liquid film whilst allowing general thinning to proceed. In a foaming solution the liquid films must possess elasticity.

115

2.2. Types of Foam When a solution containing a surfactant is aerated, bubbles rise towards the surface and two distinct regions can be seen, the top layer where the bubbles are densely packed is usually called the foam and the region beneath, where the liquid volume fraction is much greater, is termed the froth. At high gas rates, especially if the bubbles generated are small, (less than 1 mm in diameter) these regions may merge into one frothy mass called "fluid foam". Based on their characteristics, foams are classified into various categories (Table 1). In general fluid foams are encountered in submerged processes and these can be unstable, metastable, transient or persistent. Unstable foam continuously approaches the equilibrium state, constantly breaks down as the liquid dries between the bubbles. Its lifetime depends on the concentration of the solution. Metastable foam is characterised by the fact that the process of drying of the liquid between the bubbles can stop and the foam can persist indefinitely, if absolutely protected from disturbing influences (Berovic and Cimerman, 1979). TABLE 1: Classification of Foams (Ghildyal et al., 1988) Type True Fluid Unstable Metastable Transient Persistent

Characteristics Predominantly gaseous dispersion Predominantly liquid dispersion with enhanced holdup of gas in a large portion of the liquid Equilibrium state is continuously approached Progress to the equilibrium state is arrested Lifetime of seconds Lifetime of hours or days if undisturbed

Metastabili ty may be conferred on the foam by the presence of a solute that is positively adsorbed at the surface and requires work to remove it from there to the bulk. Generally, the physicochemical character of the interfacially adsorbed film of surface active agent in air-water systems depends on: -Types of surface active agent used, -Number of C atoms in the adsorbed molecule, -Molecular configuration, -The mode of adsorption and orientation, -The compressibility and spreading of the adsorbed film, -The presence or absence of hydrogen bonding either in the adsorbed film itself or between the polar groups (Mancy and Okun, 1960).

116

2.3. Theory of Foam Formation and Destruction Being gas dispersions in liquids, foams are affected by liquid properties such as surface tension, viscosity and ionic strength etc. There are qualitative di fferences between the foaming abilities of liquids and the types of foam produced. The foaming ability of liquids shows an obvious correlation with the surface activity of the solutes. Persistent foams arise only with solutes which lower the surface tension strongly, ie: the highly surface active substances. Surface activity in aqueous solutions arises from the presence of a molecule with both hydrophobic and hydrophilic properties. The surfactant hydrophobe is squeezed out of solution by the attraction between the water molecules being greater than the combined attraction of water to hydrophilic and water to hydrophobic portions of the surfactant. This positive adsorption at the vapor liquid interface resul ts in a surface tension lower than that of the pure sol vent. The surface activi ty of a solute can thus be defined as the ability to lower the surface tension of a solution by transfer of solute molecules from the bulk solution to the surface. The mechanism to explain the stability of persistent foams are those of film elasticity and the formation of gelatinous surface layers. Film elasticity is expressed as the ability of liquid films to resist localised thinning while general thinning proceeds. As the area of potential rupture is stretched further, the surfactant concentration at the surface decreases and the surface tension rises. The resulting imbalance of forces causes the surface surrounding that region to move towards the thinned spot to equalize the surface tensions. The movement of the surface layer drags along layers of the underlying bulk liquid, hence preventing thinning at the incipient weakness. This self-heal effect, which is largely independent of equilibrium surface tension is often referred to as Gibbs-Marangoni elasticity. Alternatively, the surface tensions may be equalized by migration of surfactant molecules from the bulk solution rather than from the adjacent surface. If this occurs, there is no movement to restore the thinned portion and nothing to prevent further thinning and eventual rupture. Film restoration will occur only if the rate of attaining equlibrium surface tension by surfactant adsorption from the bulk solution is slower than surface migration. Foam life is also enhanced by high bulk and which result from the interaction of forces between molecules in liquid surface. These retard drainage There is a good correlation between high surface stability, but the development of very rigid detrimental to foam life.

surface viscosities adjacent surfactant from film lamellae. viscosity and foam monolayers can be

117

The surface layers of natural surfactants such as proteins have very high surface viscosities while the surface layers of saponins exhibit viscoelastic behaviour. (Evans and Hall, 1971). It has been suggested that protein solutions have surface layers that are "plastic", meaning that they remain rigid under stress until the stress exceeds a yield value. Positive adsorption at the liquid surface implies that work must be done to transfer solute molecules from the surface to the bulk of the solution, which represents the free energy change due to the adsorption. If this free energy gain balances the free energy loss of the surface due to a change in surface tension or surface area; then the surface and the foam of which it is a part, is thermodynamically stable. The "surface activity" of a solute can now be defined as the ability to lower the surface tension of a solution by transfer of solute molecules from the bulk solution to the surface (Thomas and Winkler, 1977). There are a variety of papers on the stability of foams. Before the foam bubbles can burst, the liquid must drain from the foam lamellae. This draining occurs under the influence of gravity and capillary forces. However, its rate depends on rheological properties (surface viscosity and elastiCity) of the film which in turn are functions of the type and concentration of the surfactant. Another reason for the break up and coalescence of gas bubbles is gas diffusion which depends on the nature of the gas and on the bubble size distribution (Zlokarnik, 1986). Di fferent gases were compared using a sample of linseed oil as surfactant and the foam half-life was found to be 36, 36, 31 and 22 minutes for air, nitrogen, hydrogen and carbondioxide, respectively. The reasons behind this effect are not clear (Bikerman, 1953). 3.

FOAMING IN BIOPROCESSES

The foaming tendency of a bioprocess depends on the systems and operating variables. The broth is usually non-coalescing due to the presence of certain types of surface active materials which tend to stabilize foams when they are formed (Kawase and Moo-Young, 1987). Microbial cultures contain a variety of surface active molecules and particles. Surface activity which is defined as the ability to adsorb at interfaces is a consequence of both hydrophilic and hydrophobic (lipophilic) moieties being present simultaneously on the molecule or particle. The interaction of different surface active agents and different interfaces can give rise to a variety of surfactant functions such as emulsification, deemulsification, foaming, defoaming, spreading, etc. There are several phenomena that are responsible for the presence

118

of surface activity during the course of a bioprocess (Kosaric et a1. 1983): i) the normal metabolic activities of exponentially growing cells, ii) enzyme catalysed degradation during autolysis and release of biosurfactants during stationary and endogenous metabolism phase of older cultures, iii) physical processes resulting from culture agitation, including the shearing of the cell wall, iv) physico-chemical processes such as: -non specific leakage of intracellular lipids from cells, -extraction of intracellular and/or cell surface surfactart~ due to water insoluble substrates used for microbial growth. -adsorption/desorption phenomena involving the c surface v) uptake and metabolic degradation of previously released surfactants. vi) highly hydrophobic nature of certain microbial species (eg: ~ amarae), (Blackull and Marshall, 1989). 3.1. Factors Contributing to Foaming in Bioprocesses The stability of foam formed during a bioprocess and its foaming pattern is affected by di fferent factors. The formation of foam in submerged culture is associated with hydrodynamic conditions which in turn are affected by the introduction of gas, the nature and composi tion of the medium (pH, concentration of salts, proteins and sugars, presence of alcohols, etc.), the presence of growing cells, the formation of metabolites and surface active substances, operating conditions such as temperature, rheological properties of the solution, condi tions of sterilization and the composition of gas making up the gas bubbles CDuitschaever et al. , 1988). The physico-chemical properties of the medium affect surface phenomena. Furthermore, vessel pressure as well as capillary pressure in foam cells indirectly determine the strength of the foam. Foaming tendency changes during the course of a bioprocess. This suggests some stabilizing effect resulting from microbial acti v i ty, in addition to that conferred by the original ingredients of the broth (Evans and Hall, 1971). Because of the complexities of biosystems, it is a formidable task to relate the foaming characteristics of biomedia to individual components. However, several attempts have been made by various researchers. In a series of studies using di fferent biomedia, SchUgerl and his collegues (Bumbullis et a1., 1979, 1981; Bumbullis and SchUgerl 1979, 1981; Kotsaridu et al., 1983, a,b) have investigated the effects of, different salt concentrations, pH, alcohol, surface viscosity and viscoelasticity on foaminess and foam stability. They have concluded that foaminess and foam stability are complementary properties.

119

The foaminess of solutions increases in the presence of salts, but their relative foam stability diminishes. This influence of different salts on foam formation of protein solutions can be explained by the changes of protein solubility in aqueous solutions due to the variation of the water structure caused by salts (Schugerl, 1985). The solubili t y of proteins in water increases at low salt concentrations, causing an increase in foam stability. The converse occurs at higher salt concentrations. With the addition of pure salts to pure water, no foaming occurs at all, whereas bubble stability increases. On the other hand, the dependence of foaminess on pH is complex. Proteins have their lowest solubility and their highest foam formation capacity at their isoelectric point. Therefore, the foam formation capacity is influenced significantly by the pH of the medium. It has been observed that some substances cannot foam at a pH other than pI because of poor foam stability (Uraizee and Narsimhau, 1990). The effects of anti foam agents can also be pH-dependent. Bumbullis et a1. (1979) to influence foaminess:

point out that two factors are expected

i) the interaction of the additives with the protein molecules, ii) the effect of the additives on the solvent structure. In the case of salt addi ti ves, the interaction with the protein molecules also influences the foaminess. The influence of alcohols on foam formation is more complex, because it is caused by direct alcohol-protein interaction as well as by indirect interaction through the water structure • Foaminess is increased by short chain alcohols. The maximum effect occurs at a concentration of 1-2~~ (v/v) of alcohol added to an aqueous protein solution. Experimental results indicate that, in the presence of alcohols such as ethanol, methanol, propanol, the apparent concentration of the protein increase, causing a higher foaming at lower alcohol concentrations (Bumbullis and Schugerl, 1979). A relationship exists between the foam stability and the drainage rate of a liquid from lamellae (Bikerman, 1953). The drainage rate is diminished with increasing viscosity of the bulk liquid. However, it is not known if the drainage rate is influenced by the surface viscosity of the liquid. On the other hand, it is possible that the surface influences the gas diffusion rate across the lamallae ie: changes the foam structure with time (Bumbullis et al., 1981). Foaminess is inversely proportional to temperature. This may be due to an increase in drainage, resulting from a decrease in viscosity. Sometimes, increased evaporation of volatile surface active components can occur. At higher temperatures, the denaturation of proteins increases the foaminess of biomedia. Complex nutrient media have a particularly high foaming capacity that can increase considerably during sterilization (SchUgerl, 1985). Heat seems mainly

120

to affect the behaviour of substances which act as nitrogen sources in biomedia because these become hydrolysed or degraded to some degree and thus the products of the Maillard reaction between reducing sugars and aminoacids, proteins and peptides, are released into the solution further enhancing foam formation (Szarka and Magyar, 1969; Vardar-Sukan, 1987, 1988). Figure 1 (a) and (b) clearly show the effect of sterilization. 30

30

25

25

70

]

20

15

15

10

o

"

70

time

40 (8)

(a)

60

80

40

120

80

time

160

200

(5)

(b)

Figure 1. Effect of sterilization and the presence of solid components on the foaming behaviour of soybean flour (a) and sugar beet cosette (b) Model Media. (-0--0-) first foaming up - unsterilized medium, (~~) second foaming up - unsterilized medium, (~) first foaming up - sterilized medium, (~) second foaming up - sterilized medium. Two semi-soluble substrates, soybean flour and sugar beet cosette were chosen for their different foaming characteristics and 5~'; solutions of each were prepared using tap water to simulate fermentation media. Ster ilized solutions were obtained by autoclaving at 121 0 C for 15 minutes. A temperature controlled cylindrical glass foaming apparatus, 60 mm in diameter, was em§lloyed and foaming capacities were determined at 25 0 C, using 200 cm of model medium adjusted to pH 7. The system was agitated with a magnetic stirrer and aerated by means of a porous cylindrical diffuser, 15 mm in diameter and 30 mm in height, subm~2ged _\nto the solution. The airflow rate was kept constant at 1.33xlO ms . Foaming was induced by connecting the air pump (t=O) and continued for 10 minutes or until the total height (liquid+foam) reached the 350 mm level in the foaming apparatus - whichever occurred first. At this point (t="Z 1)' the air supply was cut off and the foam formed was allowed to collapse naturally.

121

This procedure was used in all of the experiments conducted with the four model media namely, sterilized and unsterilized solutions of soybean flour (Model Media II and I) or sugar beet cossette (Model Media IV and III). This effect is enhanced with increasing temperature, reaction time and pH. Maillard reaction is partially reversible, therefore, a reduction in foaminess is observed after storage at a low temperature (Kotsaridu et a1., 1983, b). Szarka and Magyar (1969) have called this phenomenon, "aging" and reported that enhancement of this favorable effect could be achieved by passing through sterilized air during standing. It was found that in three phase systems, the stability of the foam is a ffected by the solid particles. In the first foaming up, part of the solid components adhere to the walls of the reactor. Thus the stability of the foam subsequently generated in changed.

Vardar-Sukan (1988, 1991) investigated this effect in soybean flour and sugar beet cosette Model Media. FolloWing the first foaming up, foaming was again initiated in the foaming apparatus, by reconnecting the air supply. Thus, foam formation and foam collapse profiles were obtained for first and second foaming up processes. The rate of foam formation (r f) and the rate of foam collapse (r ) for the first and second foaming up experiments were calculated fof the different Model Media. These were found by dividing the final foam height attained (h L ) by the time period required to reach it ('l f) or by the time penod required to drain it (? ), respectively. Histograms in Figures 2(a) and (b), clearly indicatl that the faster the foam forms, the faster it will disappear, even though the rate of formation is always significantly higher than the rate of collapse. rf (e../s)

foaming up data

collapse data rc (em/s)

016 - 0.6

..

--

.-- ----

-

-------

-"-

04 -

~I---

OG - ' - -

-

--

--

, --

-

- 1---'-

---

~-

-n

J] 11--II

III (a)

~­ 0-

IV

II

III

IV

(b)

Figure 2. Rate of foam formation (a) and foam collapse (b) in different Model Media, for first ( I!lI ) and second ( 0 ) foaming up. When the ratio r /r f is considered it is found higher for second foaffiing up in soybean flour media and II) and lower in sugar beet cosette media (Model IV). This demonstrates the dependence of foaming on

that, r /r f is (Model Medla I Media III and the nature of

122

biomedia (Table 2). The decrease observed in Rl and R2 values for sterilized media is due to extremely low collapse rates ln soybean flour but due higher foam formation rates in sugar-beet cosette medium. (Fig 2 a and 2 b ). This quantitatively indicates that during sterilization foam stabilizing components are formed in soybean flour medium while foam enhancing components are produced in sugar beet cosette medium. TABLE 2: Ratio of rate of collapse over rate of formation for first and second foaming up experiments. 1'Iode1 l'Iedia I

II

III

IV

R1=(r/r f )1

0.19

0.021

0.95

0.33

R2=(r/r f )2

0.25

0.029

0.B5

0.20

% change

24%

27%

-10%

-39%

The presence of solids tends to stabilize liquid films if the solids are wetted, as is the case with microorganisms. One explanation for this is that surface active materials are also adsorbed into the solid particles, with non-polar ends oriented towards the water phase. This imparts a hydrophobic character to the particles, so that air bubbles adhere to them, resulting in a stabilization of the bubble and a longer residence time (Hall et a1. 1971). However, little is known about the influence of cells on foaming since cells are always accompanied by proteins and other solutes and also viscosity changes are proportional to cell densities. In general, a very dense suspension show less foaming acti v i ty than a very dilute solution, however this might as well be a viscosity effect (Prins and Van't Riet, 1987). It has also been observed that anti foaming agents too high in concentration enhance foam formation as well as the presence of fine solid particles. Generally, foam stability decreases up to a maximum surfactant concentration and from there on, further increases in concentration only increase foam stability. However, foam stability is not always additive. The addition of an agent itself capable of foaming can destroy existing foam if it is of opposite charge.

Operating conditions of the reactor such as air flow rate and agi tat ion also influence foaming. As the gas flow rate increases I the height of the foam layer increase, since more bubbles reach the surface and are converted into foam. Yet, in some cases, the thickness of the foam layer may decrease with increasing gas flow rate after a maximum thickness has been reached. The reasons for this are not

123

clear, but it may be due to a decrease in monodispersity of the bubbles or to other changes in bubble diameter. Mechanical disturbance of the foam at high gas flow rates might be another explanation. 3.2. Effects of Foaming on Bioprocesses The problems created by excessi ve foaming in bioprocesses fall into two classes; those which are caused by its appearance wi thin the bioreactor and those which are caused by its escape if some control is not exercised. Nutrient media used in bioindustries are of mUlti-component nature. Therefore, special care should be given to select an appropriate composition and concentration of medium components, so that the foaming tendency of the solution is kept at a minimum while the qualitative characteristics of the process (product yield, efficiency etc.) are not deteriorated. There are numerous problems, physical or biological in nature, created by foaming in bioprocesses (Table 3). When foaming occurs, the effective volume of the liquid in the bioreactor increases. This leads to a loss of culture liquid and microorganisms through air exhaust and seepage into bearings and other attachments creating sterility and containment problems. TABLE 3: Problems Created by Foaming in Bioprocesses Physical Problems .Increased effective reactor volume .Reduction in the working volume .Enhanced gas hold-up .Changes in air bubble size and composition .Reduction in app. viscosity I. Decreased power dissipation .Decreased circulation rate .Changed pattern of dissolved gases due to heterogeneous i dispersion .Lower mass and heat transfer rates I.Invalid process data due to interferance at the I electrodes i.Incorrect monitoring and control

I i

I

Biological Problems .Deposition of cells on upper parts of the bioreactor .Loss of culture fluid from exit lines causing product ond bio catalyst loss .Microbial lysis .Changes in microbial metabolism due to nutrient limitations .Froth flotation and foam separation causing preferential removal of surface active agents .Protein denaturation in the foam layer .Problems in sterile operation .Risk of environmental contamination

124

Concentration of proteins and microorganisms by froth flotation is a well known technique. Thus, enzymes or microorganisms, carried by the foam layer, may be deposited on the walls and/or lid of the reactor, where they could be no longer useful. Further problems are observed in continuous culture due to the bleed of an unrepresentative sample. All of these result in a considerable decrease in effective biocatalyst concentration. Foam generation is also autocatalytic in some cases. Small amounts of foam can create conditions in cultures that promote lysis of some cells and this in turn leads to heavier foam production. Excessive foaming also causes a reduction in the working volume of the reactor. Detrimental effects are also observed in the mass and heat transfer patterns of foaming processes. The percentage of microorganisms in foam/culture liquid depend on the composition of the cell wall affecting its hydrophilic /hydrophobic balance and its floatability. In the foam layer, the microorganisms are reported to adhere at the foam bubble surface and respire using gaseous oxygen. Due to such consumption of oxygen, the rigid bubble of foam becomes depleted of oxygen and cause oxygen limitation in the foam layer. A definite decrease of enzyme activity was found in the foam layers, which may be attributed to DOT limitation and/or protein denaturation at the gas-liquid interface. (Sukan and GUray, 1985). The enhanced gas hold up will decrease the apparent viscosity of the liquid, resulting in a decrease in power dissipation and circulation rate. The presence of tiny bubbles within the liquid also affect the transport properties and coefficients of the liquid, leading to lower mass, momentum and heat transfer within the reactor. These limitations will create additional heterogeneities within the reactor, interfering with process monitoring and control of both on-line and off-line parameters. The denaturation of proteins or enzymes due to the stresses associated with bubble formation can also be a serious problem. When a surface is created in a solution containing surface active material, the surface active molecules diffusing to the surface will tend to remain there, increasing their concentration near the surface. Subsequent to the establishment of an equilibrium, the surface may start "aging" as the molecules at the surface re-orientate themselves or undergo distortion or unfolding caused by the stresses between hydrophilic and hydrophobic groups. This has lead to the opinion that proteins are denatured by foaming. Since unwanted foams are wasteful and uncontrolled foams are often hazardous, elimination of foaming is imperative in bioprocesses. The general practice is to prevent foaming rather than destroy the foam that has already formed. However in some cases, foam formation is inevitable, thus different foam breaking methods are employed.

125

4. DIFFERENT METHODS OF FOAM DESTRUCTION AND PREVENTION IN BIOPROCESSES Even though foaming is a very wide spread problem in bioindustries, the methods adopted to combat it are as variable as the processes themselves. They range from the purely empirical approach of adding the nearest available anti foam, to a comprehensive investigation of the underlying causes and a complete and systematic evaluation of the available control methods (Hall et al. 1971). Foams can be controlled mechanically, physically or chemically. 4.1. Mechanical Methods Mechanical foam breaking is largely based on subjecting the foam lamellae to shear stress. The employed equipment include (Sfokarnik, 1986): i) Injectors, ejectors, orifices where a sudden pressure drop is occasionally superimposed on the shear stress, causing the bubbles to burst. ii) Revolving disks, impellers, stirrers where the shear stress is increased by rapidly alternating pressure fields (Ohkawa et a1., 1985; Furchner and Mersmann, 1990). iii) Centrifuges and cyclones where the rotational force is superimposed on the centrifugal force and the special design features enhance the twisting effect of foam strands (Moller, 1988). Mechanical foam breakers are preferred to overcome the disadvantages associated with the use of chemical anti foam agents. However, mechanical devices alone are generally not totally effective and in many cases, their action is enhanced by simultaneously using chemical anti foam agents at the lowest possible concentration. Mechanical foam breakers especially become indispensible in processes where even the most efficient chemical anti foam agents are not able to cope. The disadvantages of mechanical foam breakers include high running cost, complicated designs, possible damage to the product or microorganisms (Vrana and Seichert, 1988), risk of disturbances to the unit operations and their effectiveness with only light foam, under limited foaming formation. 4.2. Physical Methods of Foam Breaking Various physical devices to break foams are based on physical changes in the foaming region. Physical methods based on ultrasound, thermal or electrical treatments are employed in some bioprocesses. However, these methods are not widely used since microorganisms are highly sensitive to most physical factors (Viesturs et al., 1982). The acoustic

destruction of foam by sonic defoamers is attributed to pressure, undirected radiation pressure, induced reason ant

126

v ibrations in the bubble, high internal pressure in foam bubbles as compared to that in surrounding particles, vacuum caused by sonic energy, and turbulence produced by sonic waves (Ghildyal et al., 1988). The mechanism of collapse of the foam by the thermal method is based on the intensive expansion of the bubbles, evaporation of moisture and solvent causing foam, decrease in surface viscosity, thermal degradation of the foam producing material, freezing, and reductio~ in surface tension. Electrical foam breakers are based on passing an electric current through the foamy region to break up the foam. Eventhough the exact mechanism of foam breakage by this method is not known, the effect is probably based on the apperance of forces which act di fferently on liquid and gas (Viesturs et al., 1982). 4.3. Chemical Foam Breaking The action of chemical agents for controlling foam is fundamentally di fferent from that of mechanical foam breakers. Mechanical devices only destroy foam after it has been formed whereas chemical anti foam agents can prevent foam formation as well as destroy the existing foam. 4.4. Evaluation of the Various Methods Al though the use of chemical anti foam agents offers advantages such as simplicity, ease of operation, and acceptable economics in most cases, its disadvantages are sometimes serious. Most of the efficient mechanical foam breakers are too complicated, consume a lot of power, and do not provide ease of operation. While, the simpler mechanical devices are not efficient enough to be used alone. Thus, a single method may not be effective enough to eliminate the foam problem, and so the combined action of more than one method has been employed. These devices can be operated simultaneously or in a predetermined order. Aternatively, the least damaging device could be used alone, with another device to be put into service for supplementing the action of the former whenever it is not able to cope with the job. On the other hand, although foam control systems are used widely in practice, the inhibition of foam formation in the reactor i tsel f is preferable i f it can be achieved without any negative effects on the process. For example, foam formation can be minimized by the use of lower rates of aeration and agitation, but this usually has a detrimental effect on productivity. It may also be possible to reduce foaming by employing shorter periods of sterilization. Another solution may be the utilisation of special mutants and tailored biomedia that prevent the formation of foam. Growth of mixed

127

cuI tures of microorganisms is also a promlslng al ternati ve for foam control. The conditions may be tailored so that, the foam forming substances produced by one culture could be assimilated by the other culture. The foaming can also be reduced by controlling aeration rates and CO 2 pressure and also by lowering the incubation temperature from 32 0 to 28 0 C. It was reported that foam suppression by this aerodynamic control was more effective than with chemical antifoam agents (Ghildyal et a1. 1988). 5.

ANTIFOAMING AGENTS AND THEIR MODE OF ACTION

Chemical anti foam agents are surface active substances which decrease the surface elasticity of liquids and prevent metastable foam formation. The foam breaks as a result of the tendency to attain the equilibrium between the surface elasticity of the liquid and the surface active substances. 5.1. Mechanism of Foam Destruction by Antifoaming Agents Antifoaming agents (AFA) used to destabilize the foam in bioprocesses can be composed of oils, fatty acids, esters, polyglycols and siloxanes (Prins and Van't Riet, 1987). The surface properties of anti foaming agents are not constant and considerable variation in their efficiency is often observed. Droplets of anti foaming agents can destabilize protein films by different mechanisms: i) acting as hydrophobic bridges between two film surfaces: Obviously, the prerequisite for foam control is that the oil droplet can enter the film, thus the droplet diameter should be smaller than the film thickness. Therefore, the oil should be added as an emulsion. However, from the described mechanisms, it is plausible that very small droplets will cause only a small amount of surface movement and that the film will remain stable, Thus, there is an optimum droplet size for film collapse (Van't Riet et al., 1984). However, the mode of action of a chemical anti foam agent, in the destruction of foam, varies with the nature of the compound, the type of the foam, and nature of the substances causing foam formation. Generalized mechanisms of foam breakage by chemical anti foam agents clear ly demonstrate the di versi ty and speci fity of these mechanisms to the nature of the chemical anti foaming agent. When a hydrophobic element contacts both surfaces of a film, a convex surface is formed in the liquid film. Liquid flows away from the hydrophobic element due to increased pressure forces occuring under the curved surface. Thus, the film collapses.

128

ii) displacing the adsorbed protein on the film surface, thus disturbing its stabilizing action: Low molecular weight surfactants are more active than proteins. Therefore, they tend to displace the adsorbed protein from the surface. This affects the equilibrium on the film, possibly leading to localized drainage from the film and subsequent thinning and collapse. iii) rapidly spreading on the surface of the film, causing the liquid to be squeezed away and the film to thin and collapse: This mechanism will only operate if the surface tension of the film is larger than the sum of the surface tension of the droplet and the oil/water interfacial tension. 5.2.

Properties of an Ideal Antifoaming Agent

The properties three classes:

of

an anti foam agent used in a bioprocess fall

into

i) properties needed for it to function as an anti foaming agent, ii) properties dictated by the fact that it is used with a living system, iii) properties that facilitate its utilisation in the presence of other analytical systems such as pH probes or DOT electrodes. The anti foaming agent is added into the reactor in a sterile manner. Thus, sterilization process should not deteriorate its properties or promote the formation of corrosive products. Since the anti foam agent is added on demand when the reactor contents are foaming vigorously, instantaneous action is of great importance. To ensure a large positive spreading coefficient, the interfacial tension between anti foam and foaming medium must also be low. This requires some hydophilic character in the anti foam molecule, which is in some opposition to the necessity for low solubility. It must also be present in a sufficient quantity to main a high surface concentration even under the dynamic conditions found in a reactor, so a low solubility is advantageous. A typical antifoam: i) May have a hydrophilic character built into the molecule ii) a suitable hydrophilic emulsifier may be physically mixed with the hydrophobe. iii) should have low intermolecular cohesive forces so that i t does not itself contribute to surface viscosity or rigidity. The desired characteristics of an ideal anti foaming agent have been described by different investigators (Bryant, 1970; Evans and Hall, 1971, Ghildyal et aI., 1988). Table 4, clearly demonstrates that some of these properties contradict with each other and it is definitely not possible to find a single anti foaming agent possessing all of the desired properties.

129

TABLE 4: Desired Characteristics of Antifoams Fast foam breaking Long-lasting action Nonexplosiveness Nonvolatility Noncorrosiveness Low cost Non- or low flammability Nontoxicity to microorganisms, humans, and animals Lower BOD and COD values Nonmetabolizability by microbial strain under use Inability to form corrosive products during sterilization/fermentation Low surface tension Low interfacial tension Effectiveness at low concentration

Stability during sterilization Insolubility in foaming medium No chemical action with finished product No imparting of color or odor to the product Inability to affect O2 transfer rates adversely Better ability to destroy surface elasticity/surface viscosity Nonresistance for relatively faster biooxidation/biodegradation High entering and spreading coefficients Low intramolecular cohesive forces Presence of some hydrophobic character

5.3. Natural oils as Antifoaming Agents Natural oils belong to the group of water insoluble organic substances. They consist of triglycerides ie: esters of glycerol with long chain saturated and unsaturated fatty acids together with some free fatty acids. Existing in abundance as agricultural metabolizable and giving better product yields, advantage over other chemical anti foam agents.

products, being they possess an

However, the efficiency of natural oils is limited due to their insufficient dispersion capacity, high viscosity and metabolizable nature. Thus, they differ widely in their effectiveness as foam suppressors. Nine different natural oils; namely, castor, corn, cotton seed, linseed, olive, poppy seed, sesame, soybean and sunflower oils were examined regarding 3 their foam suppressing capacities (Vardar-Sukan, 1988) . 0.2 cm of one of the natural oils was added to simulated medium amounting to 0.1% (v/v) natural oil concentration. Foam formation and collapse profiles as well as final foam heights reached at the end of 10 minutes, were determined. (Figure 3). This procedure was repeated while the natural oil concentration was increased in increments of 0 .l~~, to a final concentration of 1. m~ (v/v). The final foam heights, which correspond to steady-state values in the foaming profiles similar to the one presented in Figure 3, were determined for each natural oil concentration and model medium. The steady-state foam heights obtained for unsterilized and sterilized soybean flour solutions (Model Media I and II) with the six best

130

natural oil,__are shown in Figure 4 a and 4 b, respectively. - r________________________________________' '2.0~

25. E

]

...

i

19.2

12.9

..

...~

'

6.4

o

150

100 ti'lle (5)

600

450

Figure 3. A typical foam formation profile.

17

12

10

10

E +-'

....~ .c '" ....~ ~ .,.,

....

6

6

4

4

o

o

0.2

0.6

0.4

0.8

\ natural oil adde:l (a)

1.0

o

0.2 \ natural oil adde:l (b)

Figure 4. Steady state foam heights obtained for Model Media I (a) and II (b), using cotton seed (-0-0-), linseed (-'7~), olive (o-o-),poppy seed (---.) , sesame (-&-.) , soybean (-......... ) , and sunflower (-.....-'f'-) oils. It appears that the effectiveness of a natural oil in foam suppression greatly varies with the type of medium. The various surface active components of different media cause variations between their responses to di fferent anti foams. For instance, sugar beet cosette is known to contain an emulsi fy ing agent (saponin) which enhances the dispersion of the natural oil in aqueous solution, increasing

131

its efficiency. This may partially account for the higher efficiencies observed with sugar beet cosette. Naturally, the efficiency of an anti foaming agent is directly proportional to its ability to suppress foam and inversely proportional to the amount consumed to attain that effect. An economical evaluation also involves its unit cost. Efficiency coefficients for each natural oil in different model media were calculated. Table 5 provides a quantitative comparison of these potential anti foaming agents, with respect to the final minimum foam height, consumption and cost. TABLE 5: Efficiency coefficients of different natural oils in various Model Media. Natural Oil

Soybean flour media unsterilized sterilized

Castor Corn Cotton Seed Linseed Olive Poppy Seed* Sesame Soybean Sunflower

75 Bl 427 2B5 9B 11 19 947 65

Sugar best cosette media unsterilized sterilized

21 2 472 16 106 45 3 47 31

lBl 210 1449 119 1539 400 1460 144 7294

139 B6 327B 354 414 lB4 500 279 1111

* since a market cost was not available, C was taken as 1.0 c

When the cost factor is also taken into account, cot ton seed oil appears to be the most favorable one, giving considerably good results with all four of the model media tested. These natural oils were similarly compared with respect to their foam distruption capacities (Vardar-Sukan,1991) and similar profiles of collapse time against natural oil concentrations were obtained (Figure 5 a-d) • 26

.,-

.or·~IHD~I--

a

20

10 6

0

02

---..

0.4

0.8

......... ...... ........

o.e

-..... - ......

011 concenlrallOl18

~b .

:26

16

0

________________________-,

,2

:~ oL~~~~~ o

42

0."

48

0..

-....... -t-.. ............. __ . . . ., _.....011 conoenlraHOI18

132 6r·~I~~-------------------------d~

lr·~I~~--------------------------, o

c

6

2

oL-__ 0.4

oil

- - _....

o.e

0.0

0.

concant,.t......

-+-aarn

~

____

~

~

__

~

~

____

M

~

__

~

__- - J

M

U

011 concantral......

- - ...............a-.

Figure 5. Foam collapse profiles for Model Media I (a), II (b), III (c), and IV (d), against varying concentrations of natural oils. In fastest compared in Table

all cases an optimum natural oil concentration resulting in collapse (C l) could be determined. When these data were with those a for foam formation (C a2 ), the results presented 6 were obtained.

TABLE 6: Optinum natural oil concentrations resulting in minimum foam height and minimum foam collapse time and maximum foam collapse rate for diffirent Model Media.

Castor Corn Linseed Poppyseed Soybean

1IIod. JIIed. I C Ca2 Ca3 al

Plod. JIIed. II Cal Ca2 Ca3

iliad. iliad. III Cal Ca2 Ca3

Plod. JIIed. IV Cal Ca2 Ca3

0.2 0.2 0.5 0.5 0.1

0.2 0.9 0.5 0.4 0.4

0.2 0.6 0.3 0.1 0.9

0.4 0.6 0.2 0.1 0.8

0.2 0.3 0.5 0.5 0.1

0.3 0.3 0.1 0.5 0.1

0.2 0.9 0.5 0.5 0.4

0.6 0.9 0.6 0.1 0.4

0.5 0.5 0.5 0.1 0.2

0.6 0.5 0.8 0.1 0.2

0.5 0.4 0.3 0.1 0.2

0.5 0.4 0.4 0.1 0.2

On the other hand, when individual collapse rates for each concentration were calculated and plotted, i t was observed that general tendency is towards increasing foaln collapse rates as natural oil concentration increases. Figures 6 (a-d) indicate foam collapse rates of the most vigorously foaming medium (II) the presence of different natural oils. Ur""'::::;'-:::::.." ::..:::;1""""=..;::.)__________________-, ;.=' ...

a

b

,-.~------~'-------"r

0.

0.2

Q.4

0.0

____

0.0

all concentrations

-------1

o.

o..~-f-----------------oL--~---~--~--~

oil the the the in

~_~

\2

0-4

04

(l'

oil concentrations

\2

133

;=-:::·:..:':::"':..!',,",,=-=:...}- - - - - - - - - - ,

,.2 001

c

~.~~'~~.~'~.,.~'~~~~)~----------. 0.6 --_.- -- -

--.. --~

~8r____;.----

n.

0.3

.-----

~'r_---

oL--~-~--~-~--~-~

o

~

~

M

M

011 concentrations

g

oL--~-~--~-~--~-~

o

0.2

0."

0.0

0.$

011 concentrations

t2

Figure 6. Foam collapse rates against natural oil concentrations for Model Medium II in the presence of corn oil (a), linseed oil (b), poppy seed oil (c), soybean oil (d). A maximum collapse rate was usually present for each model medium, following which the curves either levelled off or started to decline. The natural oil concentration corresponding to each maximum foam collapse rate value (C a3 ) for each Model Medium are also presented in Table 6. The best natural oil with respect to foam collapse behaviour is undoubtedly the one that is effective at the lowest concentration, exhibiting a maximum collapse rate and smallest final foam heights. When the results of these two studies are compared (Vardar-Sukan, 1988 and 1991), it can be concluded, foam suppression is a result of a fine balance between the effect of various surface active components including those of the chemical antifoam. An agent capable of destroying foam in one case may well act as a foam stabilizer in another one. Moreover, the effect of natural oils on foam suppression is significantly different from that of foam collapse. Thus, in choosing and employing a natural oil as an anti foaming agent, the aim suppression or collapse should be well defined and the interactions between biomedia and natural oils should be experimentally determined. 6.

EFFECT OF ANTIFOAMING AGENTS ON BIOPROCESSES

Merits of eliminating foaming and thus improving the productivity of bioprocesses are well known and at present, nearly all submerged processes are being carried out in the presence of chemical anti foaming agents. However, the additional dimension introduced by the addition of a surface active agent should not be overlooked. Positive and negati ve effects of anti foaming agents on microbial metabolism, mass transfer, downstream operations and quality of the find product have been reported by various workers (Hall et al., 1971, Viesturs et al., 1982, Ghildyal et al., 1988). Unfortunately,

the

observed

effects

are

usually

interrelated,

134

preventing the relationships.

identification

of

individual,

specific

effect-cause

Anti foam agents belonging to different groups of surface active agents exert different effects on the process. 6.1. Effects on Microbial Metabolism Certain types are known to be toxic to microorganisms, while some exert a favourable effect on growth and product formation. However, this effect cannot be ascribed entirely on their antifoaming activity (Sukan et al., 1984). The anti foam may also have a physiological effect through being metabolized by the organism. Consequently the pH of the medium is affected due to the release of fatty acids into the medium by lipase action and the constituents of the oil are utilized as carbon sources and have a marked efffect on the overall metabolism. This has been reported to increase or to decrease microbial growth, product formation and substrate utilization. Enzyme systems of the fermenting microorganisms may be damaged by some of the oils used as defoamer or carrier, causing rates of sugar utilization to decrease and production of desired metabolites such as antibiotics to be inhibited. Therefore, for enzyme production, inert anti foam agents which cannot be metabolized by the microorganisms are preferred. Anti foam agents which are metabolized during bioprocesses are often used for secondary metabolite production. In the case of metabolizable AF A, cells often exhibit diauxic behaviour, especially if the AFA concentration is high enough. Because the anti foam effect of an agent also depends on the microorganisms, medium composition and operational conditions, an empir ical test of di fferent agents is necessary for each individual system (SchUgerl, 1985). For media on carbohydrate basis, like molasses, the addition of an anti foaming agent makes the foam denser and thus impair oxygen transfer rates. However, for hydrocarbon based media, the picture in entirely different. The addition of AFA increases mass transfer rates resulting in an increase in productivity. Therefore, in some cases, the observed effect on microbial metabolism could be the result of a not-sa-obvious effect on mass transfer. Similarly, when a water dispersible silicone anti foam was used in continuous culture of P. aeruginosa grown on n-heptane, growth yield was increased 10-15%. Further studies have shown that the presence of an anti foaming agent reduced the "stripping" of the hydrocarbon by the sparged air, thereby causing and increase in the concentration of carbon source. Thus, the observed metabolic effect is in actual fact a surface effect at the aqueous-organic phase interface (Hall et al., 1971). In some cases the delay in the onset of antibiotic production was again explained by oxygen limitation due to the presence of the AFA.

135

6.2. Effects on Mass Transfer Another factor associated with anti foam additions which often causes concern is their possible effect on oxygen transfer. This is a complex phenomenon which is dependent on the interaction of many variables such as, aeration and agitation, turbulence, viscosi ty, oxygen gradient, concentration and morphology of the microorganism, contact time and surface parameters of the system. Addition of anti foaming agents alters the surface tension, surface viscosity and ionic strength, affecting the surface area, coalescence behaviour and rigidity of the bubbles. The net effect of the addition of AFA is difficult to predict since it depends on the limiting stage in the oxygen supply process. This may be divided into three stages: i) transfer across the interface, ii) transfer through the bulk liquid, iii) transfer across the liquid boundary layer from bulk to the surface of the microorganism. When an anti foaming agent is added to a bioreactor, the first change results from the breakdown of the foam layer on the surface of the aerated liquid. If surface aeration is used, aeration will be improved by the disappearance of this barrier between the bulk liquid and the atmosphore. Elimination of the fluid foam and reduction in the air hold-up causes a reduction in the contact time. When air hold-up decreases, the power drawn by the system increases up to 2m~ (Benedek and Heideger, 1971; Bungay et al., 1960). In a system in which the liquid to cell mass transfer is the limiting step, the increased power will improve aeration by improving mixing. This effect is enhanced by the changed rheology of the broth. At the gas-liquid and liquid-cell interfaces, accumulation of the anti foam agent on the cell surface increases resistance to mass transfer. These effects are more enhanced in complex biomedia, due to interactions between various media components and anti foaming agents. Thus, there are contradictory reports on the effects of various surface active agents on mass transfer. It is also reported that the presence of antifoaming agents suppress the effects of the other components on overall mass transfer coefficient (k La). In the range investigated (0.5-1.5?~) by Adler et al. (1980 a) kLa values were only slightly affected by soy oil concentration, especially in the presence of proteins. This suggests an interrelated influence of additives on kLa. to

The values of k a in the presence of anti foams were reported be larger than t~e corresponding values in the absence of the

136

antifoam at high driving forces, however observed kL values were lower in the presence of anti foaming agent at low Jriving forces (Ghildyal et a1., 1988). The behavioural properties of anti foaming agents responsible for lowering or increasing oxygen transfer rates are presented in Table 7. TABLE 7: Properties of on Oxygen Mass Transfer Increasing Mass Transfer Displacement of the barrier by a permeable silicon monolayer Release of oxygen-depleted air from the bubbles and the entry of fresh air Increased diffusibility of gases in liquids Increased interfacial area Progressive decrease in concentration of the antifoam in the medium due to its separation on the walls of the reactor Increases in the power drawn by the system

Anti foams

Having

Effect

Decreasing Mass Transfer Barrier for gas diffusion across the interface Decreased effective contact between culture and air trapped in foam Formation of large bubbles with a smaller surface-to-volume ratio Decreased (air hold-up) contact time of gas bubbles with liquid medium Additional gas-liquid interfacial resistance due to surfaceactive nature of anti foams Gas diffusion across the interface to the extent determined by its monolayer

Both foam stability and oxygen transfer from the gaseous to the liquid phase depend on surface phenomena. Conditions which cause the collapse of bubbles in foam also favour the coalescence of bubbles wi thin the liquid phase, resulting in larger bubbles with reduced surface to volume ratios. Resolving these two opposing phenomena to produce a non-foaming bioprocess with high mass transfer characteristics is a formidable task requiring a deeper understanding of surface effects (Prins and Van't Riet, 1988). The mass transfer coefficient k a is strongly influenced through "a" by the coalescence behaviour of the medium. The observed decrease in kLa by an AFA is attributed to a decrease of "a". The equilibrium bubble size in a dispersion depends on coalescence and dispersion processes. In a series of experiments, using model media containing Na-caseinate and soy-oil, Van' t Riet et al (1984) have observed that in a coalescing liquid, the equilibrium bubble diameter is around 6 mm. In a noncoalescing liquid, the bubble size is determined by the size of the original bubbles and in stirred

137

vessels the diameter can be much smaller. Stabilization (ie: by proteins) of the drainage between colliding bubbles can change a coalescing liquid into a non-coalescing one in which the bubble size may be markedly reduced. This will improve the mass transfer from bubbles to liquid because the interfacial area increases linearly wi th decreasing bubble size. Therefore, up to an extent a foaming liquid have a higher kLa. When too much AF is added to the liquid, the foam will be destroyed completely, but the liquid will become strongly coalescing and mass transfer will be reduced. However, by selecting AF droplets of an appropriate size, thin foam films can be forced to coalesce whereas the thicker films in the broth remain stable (Prins and Van't Riet, 1987). The actual mechanism appears to be more complex. Koide et al. (1985) observed that the volume to surface mean diameter of bubbles in the presence of an anti foam agent is smaller than that in its absence, except at low gas flow rates. The reduction in the size of the new bubbles causes an increase in the interfacial area, but these tiny bubbles are unable to create much turbulence in their passage through the medium and they are surrounded by rather thick liquid films. 80

v

70

"-:. ....

bll

50

'"'

40

'"

30

....'"

a

b

~ 211 10

0 U

10

20

30

~0 50 60 tim .. (min)

70

8 (J

Figure 7. DOT profile of a penicillin fermentation upon the discrete addition of 0.05~~ lard oil, while the DOT is around 70?~ (a) and 20% (b) air saturation. In a penicillin production system studied, (Vardar and Lilly, 1982) discrete additions of anti foaming agent (lard oil) created disturbances in the system. Below 25~~ DOT, addition of 0.05~~ lard oil caused a peak in DOT, whilst above 25~~ air saturation, a drop

138

was observed. This was accompanied by apparent large increases in carbondioxide production and oxygen uptake rates (Figure 7). Similarly, the pH of the system immediately increased and then decreased to a value well below the pH before anti foam addition. The duration of the disturbance in the system was roughly proportional to the amount of anti foaming agent added, and ranged between 10 and 2Q minutes. Thus, the situation is more complex in a real fermentation process. In the penicillin production system studied, the observed changes in the DOT profile can be explained by the various effects of anti foam agents on mass transfer. If the DOT is below 25~~, then the gas-liquid mass transfer rate is the limiting step and the favourable effect of the antifoam agent on bubble size, air hold-up, rheology and power drawn by the system will enhance mass transfer resulting in an increase in the DOT of the broth. If the DOT is high, the deleterious effect of the anti foam agent is more dominant causing a decrease in the DOT. When the foam breaks up, the CO 2 entrapped inside the bubbles is released changing the percentage composition of the exit gas, also affecting the value of % O2 consumed. Nyiri and Lengyel (1965) and Lengyel and Nyiri (1966) reported a decrease in DOT and a peak increase in oxygen uptake rate following the addition of anti foam and reasoned that this might either be a physical effect or an increase in respiration due to the elimination of carbon dioxide influencing metabolism. They observed a similar phenomenon upon the addition of sulphuric acid into the broth for pH adjustment, causing CO 2 liberation and an apparent increase in respiration. On the other hand, the changes in pH are basically a reflection of the change in the equilibria: H2C0 3 HCO 3 CaC0 3

~

-----"" --""

H+

+ HCO;

H2O + CO 2 Ca++

+

CO .. 3

CO 2 release creates an instantaneous increase in pH followed by a fast decrease as a result of a series of shifts in these equilibria. Since these effects can also be observed in uninoculated fermentation media, it is reasonable to conclude that the observed disturbance is mainly due to changes in solubility and mass transfer rates. Therefore it can be concluded that, the net changes observed are due to the competative effects on the individual terms, I kL I and I a I , and that the effect on the overall mass transfer depends on the limiting stage in the process and the nature of the anti foaming agent. The effect of the variation of biomedia composition on bubble coalescence behaviour is difficult to evaluate since several parameters can influence this process. The composition variation due to metabolites produced by the cells or dissolved proteins from dead microorganisms which have become liberated by cell lysis can have

139

significant effects on the coalescence. The effects of surface active substances on the coalescence rate are well known and consumption of these substrates by the microorganisms can vary the bubble size distribution significantly. Antifoaming agents usually increase coalescence rate and reduce oxygen transfer rate. However, no quantitative relationships between these factors are known (Adler et al., 1980 b). Moreover, it should be noted that the addition of an anti foaming agent may replace or destroy the foam stabilizing monolayer or the electrical double layer between bubbles and could cause a decrease in surface viscosity thereby enhancing bubble coalescence. On the other hand, the composition of the solution to which the anti foaming agent is added influences the observed result on mass transfer. If the anti foaming agent is added to a salt solution kL a diminishes significantly due to the increase in coalescence rate. By adding protein to this system, the coalescence rate is slightly reduced. It is possible to partially displace AFA by protein on the gas-liquid interface. Addition of ethanol increases the KLa value because of coalescence repression and further addition of protein reduces it (SchUgerl, 1985, Bumbullis and SchUgerl, 1979). Vardar-Sukan (1990) carried out a comparative study on the effects of different natural oils on kLa. Figure 8 compares the effects of two different natural oils, poppy seed and corn, chosen for their di fferent physical properties. It was found that generally kL a decreases down to a minimum natural oil concentration and then gradually increases with increasing concentration. The same pattern was observed with other natural oils tested. In case of poppy seed oil, it was not possible to test higher concentrations since a good oil-water dispersion could not be maintained above 0.3% using realistic stirrer speeds. 0.05 0.04

I

O.OJ 'n ",../ :G



0.02

0 0

0

0.01

o

0.1

0.20.3

o.~

0.;

0.bO.7

o.tio.'}

1.0

l.. natllr.J J "j J

Figure 8. Effect of poppy seed oil (.......) and corn oil (-0-0) of an aqueous solution

on kL a

140

The results obtained were in agreement with those of some early researchers (Mancy and Okun, 1960, Deindoerfer and Gaden, 1955) recent studies, anti foaming agents in aqueous and nutrient salt solutions were investigated in bubble column reactors for a number of parameters such as, media composition, air flow rate, air hold-up and type of anti foaming agent. The effects were found to increase or decrease kL a depending on the interraction of these parameters (Adler et a1., 1980 a,b). Since dispersion of natural oils in solution is important with respect to their effectiveness as anti foaming agents, emulsi fication by a surfactant should increase their efficiency. Therefore most anti foaming agents are mixed with a sui table carrier substance, so that their anti foaming properties are utilized fully. The carrier seems to act as a reservoir from which the anti foaming agent is liberated. A carrier may either accelerate defoaming by an anti foam agent and prolong its duration of action or may show the opposite effect (Ghosh and Pirt, 1959). Production of a "good" natural anti foam will depend upon the extent to which the natural oil is emulsified to provide good dispersion within an aqueous system. Selection and tailoring of appropriate surfactants for a particular application will depend on some knowledge of the structure-dependence of the chemical, physical and interfacial properties of the surfactant. The addition of the surfactant (Tween 20) increases the gas hold-up due to enhancement of the non-coalescing tendency of the bubbles. In fact, it was clear from visual observations that the number of very small bubbles increased considerably in the presence of the surfactant and the addition of Tween 20 to water without AFA caused formation of a foam (Kawase and MooYoung, 1987). 1500 360

~

1

"

'0 c

-=t5 u

If

0

...,.....

i ... 0

~

i"l

7

.,§ . .

.

t

~ c::

i a

{a}

,?: "0

~ ; ~~

c § on

"0

u

u

au

a

100

.:~

.a II

E

:;>

~

~ "Z 0

~

.:

. . .!

\:

.;"

200

.

~

c 0

c

~u

...~8.

.. i.. a; E 0 on

on ~

i .2

..

0

8 'E > Ii !!u " :g ~

.

{b}

Figure 9. Effect of 0.4~~ natural oil on foaming time ( Iii! ) and foam collapse time ( 0 ) of 72-hour-old T. reesei cultures in the absence (a), and in the presence (b) of O.l~~ Tween 80.

141

Our observations are also along similar lines. Tween 80 was found to be a good emulsi fy ing agent for all natural oils tested (Sukan et al. 1984). Figures 9(a) and (b) demonstrate the increased efficiency upon the addition of the emulsifier, in the fermentation media. However, when natural oils (soybean, poppy and sunflower) were emulsi fied by a high speed homogeniser using 3~~ Tween 80, a considerable decrease in their anti foaming capacities were observed in different Model Media (Vardar-Sukan, 1988 b). This effect was extended to a total loss of foam breaking ability at concentrations higher than 0.1% AFA. The effect of an AFA on mass transfer rate also varies as a function of the aeration rate of the reactor. Schugerl et a1. (1978) found that at high gas flow rates (v :> 0.033 ml s) values of kL a in the presence of an AFA were lower tH'an the corresponding values of kL a in the absence of AFA but at low gas flow rates (v < 0.033 m/s) was higher in the presence of AFA. At low aeration Srates, where coalescence does not play an important role, addition of an AFA reduces kL a to about one third of the value in the absence of AFA. At high aeration rates, the coalescence promoting effect becomes effecti ve. The specific interfacial area is reduced significantly. The kLa value passes a maximum with increasing anti foam concentration, the gas holdup on the other hand diminishes, passes through a minimum and then a maximum (Schugerl, 1985). Coalescence is not the only reason for the decrease of k a. It is generally known that a surface active material tends to be adsorbed and accumulate near the rear of a single rising bubble, retard internal circulation, therefore decrease the terminal bubble rising velocity and mass transfer rate (Kawase and Moo-Young, 1987). Furthermore, the bubble rising velocity is signi ficantly reduced in the presence of surface acti ve materials due to the existance of surface tension gradient at the bubble interface (Clift et al., 1978). These observations lead to an expected increase in the gas hold-up. However, observations indicate that al though the bubble size in pure water is rather uniform, the presence of AFA causes a less uniform bubble size distribution due to bubble coalescence and bubble break-up. Anti foaming agents may enhance not only bubble coalescence but also bubble break up. AF A lower surface tension by adsorption on the bubble surfaces. Since the addition of an anti foaming agents lowers surface tension, it enhances the tendency for bubble break up thus decreasing the average bubble size. "kL" can also decrease considerably in biomedia in comparison to pure water. This effect can be partially explained by the surface renewal theory. A continuously renewing interface with respect to the bubble increase kL' which is a constant dependent on the diffusion coefficient of the gas at a stationary interface. In the presence of surfactants the mobility of the bubble surface decrease eventually leading to a rigid surface as the concentration of the AFA increases

142

(Prins and Van't Riet, 1988). In the two-phase system (air-water), the molecules of a soluble component exhibit a higher concentration at the interface than in either water or air. Such molecules pass through the surface of water and they thicken the interfacial region, thus/reducing the interfacial tension (Mancy and Okun, 1960). This results in an additional resistance to mass transfer causing a further decrease in 'k L '. 6.3. Effects on Process and Unit Operations There is the risk that the anti foam agent induces considerable changes in the physical properties of the culture broth and may even cause various operational characteristics of the bioreactor to deteriorate (Ohkawa et al.1985). ProductR and installations become contaminated by highly surface active agents. In some cases they may be forced into shutdown. Excessive AFA could adversely affect many of the unit operations involved in product recovery, separation and isolation. They can foul membranes used in broth clarification, cell harvesting, filtrate concentration during downstream processing, especially causing flux reduction (Mc Gregor et a1., 1988). However, there is evidence that this observed detrimental effect is the resultant of several factors including the molecular mass of solutes, type and material of membranes and anti foaming agents (Kloosterman IV et al., 1988). Negative effects were also observed in unit processes such as adsorption, extraction, electrophoresis and crystallization. Even if the anti foam does not interfere with downstream unit operations, its presence in the final product may in some cases create serious problems with respect to product quality, toxicology and legislation. In this case, the advantages of metabolisable anti foaming agents are undeniable. If additions are well-timed during the early or middle stages of the process, all traces of the anti foam is removed through microbial activity prior to separation and purification. The right use of the antifoam agent, in minimal possible quanti ties, is vital not only because it reduces the adverse effects but also because it can lead to improved economics in view of the high cost of defoamers. In addition, it prevents fine emulsion formation in the fermentation medium, which is highly desirable for maintaining a large surface area between the phases in many cases. Finally, the use of too low or too high a concentration of the anti foam agent can lead to stabilization of the existing foam (Schugerl, 1985).

143

7.0. CONCLUSIONS

Foam control in industry is still an empirical art. The best method established in one plant for a particular process is not necessarily the best for the same process carried out in a di fferent site. The reasons for this are that, natural products of complex nature are generally used in the preparation of industrial biomedia. Their properties are not particularly constant nor completely defined. Moreover, variations occur between batches. Similarly, the design and operating parameters of a bioreactor may affect the properties and extent of the foam formed. Often, improper choice and inaccurate dosage of anti foaming agents create unforseen problems and expenses. The selection of an appropriate anti foaming agent must be treated wi th caution and understanding. The economic factors governing the choice of anti foaming agent vary from plant to plant, thus no universal set of rules can be defined. The efficiency of an anti foaming agent is generally determined by two parameters; the minimum volume necessary and maximum yield of product or absence of any antimicrobial activity. Prior to the application of an anti foaming agent, comparative experiments should be conducted with respect to their minimum influence on the physiological characteristics and maximum efficiency on foam suppression and collapse in the given medium. In addition, the cost of the anti foaming agent and its effects on the end product should be taken into consideration. Possibilities of increasing its efficiency with emulsification should be investigated. Lastly, the employment of combined methods together with tailoring the process conditions to minimize the occurrance of foaming should be considered. Further research in this neglected area is undoubtedly imperative especially to elucidate the complex interractions amongst biomedia components, biocatalysts, products and anti foaming agents. In the future, it will be a necessity to replace the currently used trial-and-error method in the selection of the sui table foam control method. However, extensive research is required for the successful prediction of foaming and defoaming phenomena in bioprocesses.

144

8.0. REFERENCES: Adler, I., Buchholz, H., Voigt, J., Wittler, R., and SchUgerl, K. (1980, a) • 'Bubble Coalescence Behav iour in Biological Media: I. Hansenula Polymorpha Cultivation Broths', Europ. J. Appl. Microbiol. Biotechnol. 9, 249260. Adler, I, Diekmann, J., Hartke, W., Hecht, V., Rohn, F., and SchUger 1, K. (1980, b) 'Bubble Coalescence Behav iour in Biological Media: I!. Effect of Antifoam Additives', Europ. Appl. Microbiol. Biotechnol. 10, 171-186. , The Atlas HLB System', Atlas Chemical Industries, INC. Chemicals Division, Wilmington 99, Delaware, 1963. Berovic, M. and Cimerman, A. (1979) 'Foaming in Submerged Citric Acid Fermentation on Beet Molasses' , Europ. J. Appl. Microbial. Biotechnol., 7, 313-319. Blackall, L. L., and Marshall, K. C. (1989) 'The Mechanism of Stabilization of Actinomycete Foams and the Prevention of Foaming Under Laboratory Conditions', Ind. Microbial. 4, 181-188. Bryant, J. (1970) 'Anti-Foam Agents', in N. Ribbon and R. Norris (eds.), Methods in Microbiology, pp. 187-203. Bumbullis, W., Kalischewski K., and SchUgerl K. (1979) 'Foam Behavior of Biological Media: II Salt Effects', Europ. J. Microbiol. and Biotechnol. 7, 147-154. Bumbullis, W., Kalischewski, K., and SchUgerl, K. (1981) 'Foam Behavior of Biological Media: VII. Surface Viscosity and Viscoelasticity', Europ. J. Appl. Microbiol. Biotechnol. 11, 110-115. Bumbullis, W. and SchUgerl, K. (1979) 'Foam Behavior of Biological Media: V Alcohol Effects', Europ. J. Appl. Microbial. Biotechnol. 8, 17-25. Bumbullis, W. and SchUgerl, K. (1981) 'Foam Behaviour of Biological Media: VI. Foam Stability. Salt Effects', Europ. J. Appl. Microbial. Biotechnol. 11, 106-109. Deindoerfer, F. and Gaden, E. (1955) 'Effects of Liquid Physical Properties on Oxygen Transfer in Penicillin Fermentations', Appl. Microbinl. 3, 253-257. Duitschaever, C.L., Buteau, C., and Kamel, B.S. (1988) 'An Investigation on the Efficiency of Anti foaming Agents in Aerobic Fermentation', Process Biochem., December, 163-165. Evans, J.I. and Hall, M.J. (1971) 'Foams and Antifoams in Fermentation', Process Biochem., April, 23-26. Furchner, B. and Mersmann, A. (1990) 'Foam Breaking by High Speed Roters', Chem. Eng. Technol. 13, 86-96. Ghildyal, N.P., Lonsane, B.K. and Karanth, N.G. (1988) 'Foam Control in Submerged Fermentation State of the Art', Adv. in Appl. Microbial. 33, 173-222. Ghosh, D. and Pirt S.J. (1954) 'Antifoam Agents in Aerobic Fermentations. Evaluation of the Activities of Antifoam Preparations', Ren. Inst. Sup. di San. (En 7, 149-167. Hall, M.J., Dickinson, S.D., Pritchard, R. and Evans J.I. (1971) 'Foams and Foam Control in Fermentation Processes', Prog. Ind. Microbial. 12, 171-231.

145

Hancock, R.I. (1984) 'Macromolecular Surfactants' ii, Tadros, Th.F. (ed.), Surfactants, Academic Press, London. Kalischewski, K., Bumbullis, W. and SchUgerl, K. (1979) 'Foam Behaviour of Biological Media: I. Protein Foams', Europ. J. Appl. Microbiol. 8iotechnol. 7, 147. Kawase, Y. and Moo-Young, M. (1987) 'Influence of Antifoam Agents on Gas Hold-up and Mass Transfer in Bubble Columns with Non-Newtonian Fluids', Appl. Microbiol. and Biotechnol. 27, 159-167. Kloosterman IV, J., Van Wassenaar, P.D., Slater, N.K.H., and Baksteen, H. (1988) 'The Effect of Anti-foam Agents on the Ultrafiltration of a Protease Solution', Bioprocess Eng. 3, 181-185. Komarova, A.B., Dubyaga, E.G., Gladkovskii, G.A., Ryzherhova, E.V. and Tarakanov, O. G. (1984) 'Foam Formation Characteristics of Oligoethers, Copolymers of Ethylene and Propylene Oxides in Water', Colloid J. 46, 511. Kosaric, N., Gray, N.C.C., Cairns, W.L. (1983) 'Microbial Emulsifiers and De-emulsifiers', in H.-J.Rehm and G. Reed (eds.), Biotechnology: A Comprehensive Treatise'. Volume 3, Verlag Chemie, Weinheim. Kotsaridu, M., Gehle, R., and SchUgerl, K. (1983,a) 'Foam Behaviour of Biological Media: IX. pH and Salt Effects', Europ. J. Appl. Microbiol. Biotechnol. 18, 60-63. Kotsaridu, M., MUller, B., Pfanz, V., and SchUgerl, K. (1983,b) 'Foam Behav iour of Biological Media: X. Influence of the Sterilization Conditions on the Foaminess of PPL Solutions', Europ. J. Appl. Microbiol. Biotechnol. 17, 258-260. Lee, J.C. and Tynan, K.J. (1988) 'Antifoams and Their Effects on Coalescence Between Protein Stabilised Bubbles', Bioreactor Fluid Dynamics, Conference paper, 353-77. Lengyel, Z.L. and Nyiri, L. (1966) 'The Inhibitory Effect of CO 2 on the Penicillin Biosynthesis', Biotechnol. Bioeng. 8, 337-352. Mancy, K.H., and Okun, D.A. (1960) 'Effects of Surface Active Agents on Bubble Aeration', Journal WPCF 32, April, 351-364. Mazumder, T.K., Nishio, N.B and Nagai S. (1985) 'Carbon Monoxide Conversion to Formate by Methanosarcina Barker i " Biotechnol. Lett. 7, 377-382. McGregor, W.C., Weaver, J.F., and Tansey, P.S. (1988) 'Antifoam Effects on Ultrafiltration', Biotechnol. Bioeng. 31, 385-389. Moller, Von O. (1988) 'Entgasung von FlUssigkeiten im Fliehkraft feld-Das CENDEGA- Verfahren " Fat Sci. Technol., 90, Mai, 529-531. Nyiri, L. and Lengyel, Z.L. (1965) 'Studies on the Elimination of C0 7 Gas Containing Foam Formed in Penicillium chrysoqenum Deep CuItures', Antibiot. Advan. Res., Prod. Clin. Use, Proc. Congr. Prague, 729-732. Ohkawa, A., Ueda, Y., and Sakai N. (1985) 'Effect of Impeller Design on Foam Breaking and Power Characteristics of an Aerated Stirred Tank Fitted with a Rotating Disk Mechanical Foam-Breaker', Process Biochem. August, 109116. Prins, A., and Van't Riet, K. (1987) 'Proteins and Surface Effects in Fermentation: Foam, Antifoam and Mass Transfer', Trends in Biotechnol. 5, 296-301.

146

SchUgerl, K. (1985) 'Foam Formation, Foam Suppression, and the Effect of Foam on Growth', Process Biochem., August, 122-123. Sukan, S. and GUray, A. and Vardar-Sukan, F. (1989) 'Effects of Natural Oils and Surfactants on Cellulase Production and Activity', J. Chem. Tech. Biotechnol. 46, 179-187. Sukan, S., Vardar, F., and GUray, A. (1984) 'Evaluation of Natural Oils as Anti foam Agents in Bioconversion of Cellulosic Substrates', Third Europ. Congr. Biotechnol. MUnchen, Germany, September, 10-15, 2, 203-21l. Szarka, L. and Magyar, K. (1969) 'The Foams of Fermentation Broths: I. Some Parameters of the Foaming of Fermentation Media', Biotechnol. Bioeng. II, 701-710. Thomas, A. and Winkler, M.A. (1977) 'Foam Separation of Biological Materials', in A. Wiseman (eds.), Topics in Enzyme and Fermentation Biotechnology, Ellis Horwood Ltd., Chichester, pp. 43-71. Uraizee, F. and Narsimhan, G. (1990) 'Foam Fractionation of Proteins and Enzymes: II. Performance and Modelling', Enzyme Microbial. Technol. 12, 315-316. Van't Riet, K., Prins, A. and Nieuwenhuijse, J.A. (1984) 'Some Effects of Foam Control by Dispersed Natural Oil on Mass Transfer in a Bubble Column' Third Europ. Congr. Biotechnol., MUnchen, Germany, Sept., 10-14. 3~ 521-525. Vardar, F. and Lilly, M.D. (1982) 'Effect of Cycling Dissolved Oxygen Concentrations on Product Formation in Penicillin Fermentations', Eur. J. Appl. Microbial. Biotechnol. 14, 203-211. Vardar-Sukan, F. (1988,a) 'Efficiency of Natural Oils as Antifoaming Agents in Bioprocesses', J. Chem. Technol. Biotechnol. 43, 39-47. Vardar-Sukan, F. (1988,b) Unpublished Results. Vardar-Sukan, F. (1990) 'Effects of Natural Oils on Oxygen Mass Transfer Rates in Biomedia', Asia-Pasific Biochem. Eng. Conf.' 90, Apr. 22-25, Seoul, Korea. Vardar-Sukan, F. (1991) 'Effects of Natural Oils on Foam Collapse in Bioprocesses' Biotechnol. Lett. 13(2), 107-112. Viesturs, U.E., Kristapsons, M.Z., and Levitans, E.S. (1982) 'Foam in Microbiological Processes', Adv. Biochem. Eng. 21, 169-224. Vrana, D., and Seichert, L. (1988) 'Cytomorphological Comparison of Mechanical and Chemical Defoaming of a Yeast Culture', Folia Microbiol. 33, 144-147. Zlokarnik, M. (1986) 'Design and Scale-up of Mechanical Foam Breakers', German Chem. Eng. 9, 314-320.