Dca

of free cells. With respect to brewing processes using such yeast, some of these alterations have beneficial effects and some are disadvantageous. The precise causes of the changes are obscure and many factors appear to be influential. A proper elucidation of cause and effect is difficult because of the heterogeneous nature of immobilised systems. In addition, it is difficult to unravel those effects due solely to immobilisation of cells and those which relate to restricted mass transfer. It is an important area of study since a proper understanding of the physiology of immobilised yeast is clearly a prerequisite to optimising the design and operation of bio-reactors.

Factors influencing yeast physiology fall into two loose groupings, which roughly coincide with the positive and negative responses to immobilisation. Thus, as Ryder et al. (1995) have commented, yeast cells in nature are usually found attached either to each other and/or to a surface (see Section 8.1.4.1 Biofilms). Therefore, immobilisation may be viewed as a protected and indeed preferred habitat. In the discussion to the same paper, Iserentant makes the point that cell-to-cell contact has profound but as yet incompletely understood effects on yeast physiology and that immobilised cells may react as a colony and not as individuals. Again, the implication is that this is a natural state for yeast cells to be in, and, as a corollary, freely suspended cells are in an abnormal and disadvantageous situation.

The negative responses relate to the stresses that immobilisation imposes on yeast cells. The most important of these are undoubtedly the effects of restricted mass transfer of supply of nutrients, particularly oxygen, and the removal of potentially toxic products such as ethanol and carbon dioxide. The severity of these effects will depend on the biomass loading, the type of support and the degree of agitation present in the reactor. Thus, it would be predicted that surface attached cells in a rapidly stirred reactor would be the least affected. In contrast, cells entrapped in the core of large beads in a packed bed type reactor are subject to the greatest degree of stress and will exhibit the most negative response. In between these two conditions, a whole continuum of varying restrictions to mass transfer is possible. For example, Pilkington et al. (1999) using lager yeast entrapped in carrageenan gel beads noted that after six months' continuous fermentation the viability of the entrapped cells decreased to approximately 50%. Cells in the core of the beads were noted as having fewer bud scars and had an altered morphology compared with those at the outer surface.

Superimposed on these effects are responses, which are due to the use to which the reactor is put. Where it is being used for primary fermentation, and, consequently, where a complete aerobic growth medium is supplied, the response of the immobilised yeast would be expected to be different to that exhibited by yeast in a reactor used for continuous maturation, where the feed stock is beer with little or no nutritive content. Similarly, as with conventional continuous fermenters, stirred homogeneous reactors used for primary fermentation are likely to elicit different responses in yeast to those which use unstirred packed beds.

Bearing in mind the complexities which arise from the preceding discussion some general comments may be made. There is a consensus that overall metabolic rates are enhanced in immobilised yeast cells adding support to the contention that this is an advantageous state. For example, increased rate of glucose uptake, increased rate of ethanol formation, increased yields of ethanol, lower intracellular pH and alterations in the activities of glycolytic enzymes have all been reported (Doran & Bailey, 1986; Tyagi & Ghose, 1982; Aires Barros et al., 1987; Galazzo & Bailey, 1989, 1990). In contrast, specific growth rates are reduced, as are biomass yields, specific rates of oxygen uptake and carbon dioxide evolution (Tyagi & Ghose, 1982; Doran & Bailey, 1986). These alterations can be dramatic; thus, in the report of Doran and Bailey (1986) specific rates of ethanol production in immobilised cells were 40-50% greater than in free cells. The differential in rate of glucose consumption was two-fold in favour of the immobilised yeast, yet biomass yield only 30% of the suspended cell value.

In addition to increased ethanol production, both glycogen and trehalose pools are elevated compared to free cells (Doran & Bailey, 1986). These reserve polysaccharides are implicated in resistance to starvation and other stresses and perhaps provide a part explanation for the observation that immobilised cells are more resistant than free cells. For example, Jirko (1989) concluded that immobilised yeast was better able to withstand starvation than free cells and this was due to the former being able to maintain a high adenylate energy charge for longer periods - presumably at the expense of dissimilation of the larger pool of glycogen.

Increased resistance to starvation in immobilised cells is accompanied by improved ethanol tolerance (Holcberg & Margalith, 1981; Dror et al., 1988). The precise mechanism remains obscure. Dror et al. (1988) proposed that cross-linked gels could form a thin layer on the cell surface and this provided stabilisation in the presence of ethanol. Hilge-Rotmann and Rehm (1991) concluded that under anaerobic conditions and in the presence of ethanol, immobilised cells contained greater concentrations of long-chain saturated fatty acids, compared to free cells. A positive correlation between fermentation rate and the degree of fatty acid saturation in immobilised cells was demonstrated.

This is contrary to the more usual assertion that ethanol tolerance and the proportion of unsaturated fatty acids in membranes are directly related (Beavan et al., 1982). However, Alexandre et al. (1994) concluded that membrane fluidity is the key to ethanol tolerance and that different cells have varying ways of manipulating this parameter by altering membrane lipid composition. It is conceivable, therefore, that the micro-environment to which immobilised yeast is exposed may produce a response to ethanol different from that of free cells. This aspect of immobilised yeast physiology is intriguing in that such cells may have restricted access to dissolved oxygen and of course, this is required for the de novo synthesis of both unsaturated fatty acids and sterols. It might be expected, therefore, that immobilised yeast would have reduced tolerance to ethanol but this is apparently not the case.

A further characteristic of the immediate environment of immobilised cells may be reduced water activity, which represents yet another stress that the cells must cope with. Typically, yeast cells react to low water activity by excreting osmoprotective metabolites such as glycerol. Galazzo and Bailey (1989) observed that the rate of production of this metabolite by alginate-entrapped yeast was approximately double that of suspended cells, which perhaps supports this view. Of course, it is also true that glycerol production has a redox-balancing role and this could be implicated in these changes. It is known that when yeast is subject to sub-lethal heat stress a response is elicited, termed the heat shock response in which increased thermo-

tolerance is acquired (Lindquist & Craig, 1988). The response is complex and involves the simultaneous induction of a large number of so-called heat shock proteins (Schlesinger, 1990). Stresses other than heat can also trigger the response and it has been demonstrated that increased tolerance to these other stresses is also acquired. For example, heat shock leads to enhanced ethanol tolerance (Watson & Caviccioli, 1983) and increase in levels of trehalose (van Laere, 1989). It is tempting therefore, to speculate that some of the physiological changes seen in immobilised yeast may be stimulated by applied stresses and manifest in a similar way to the heat shock response.

Several authors have reported altered morphology of immobilised cells, which suggest that cell proliferation and bud separation may be affected. Jirko et al. (1980) noted that yeast cells covalently linked to a modified hydroxyalkyl methacrylate gel proliferated in nutrient media without separation of daughter cells and new cells were of an elongated shape. It was assumed that the altered morphology was due to attachment disrupting normal growth. Koshcheyenko et al. (1983), also using gel entrapment, reported that cells in the surface layers were apparently of normal morphology but those deep in the matrix showed many abnormalities. With prolonged incubation these cells lysed, presumably due to nutrient starvation. Cryptogenic growth of the outer viable cells on the lysis products was also evident. Doran and Bailey (1986) showed that immobilised yeast contained increased concentrations of structural polysaccharide and had higher ploidy than free cells, indicating that the normal cell cycle had been disrupted.

It would appear that growth, as measured by cell proliferation and/or increase in mean cell size, is restricted in some immobilised systems at least for some of the time. This is of particular relevance to application to primary fermentation since growth and the concomitant assimilation of wort nutrients other than sugars, is necessary for the production of an appropriate spectrum of flavour-active metabolites. The pertinent question is whether growth and nutrient assimilation are restricted by immobilisation per se, or can the limitation be overcome by increasing rates of mass transfer?

The evidence suggests that restricted mass transfer is the more significant factor, particularly the supply of oxygen. Thus, most growth occurs during the initial start-up period at a rate governed by the type of reactor, physical parameters such as temperature and the nature of the process fluid. In this initial period the carrier is not fully loaded and as growth proceeds the new daughter cells occupy the spare capacity. This may be via attachment to unoccupied surface or by colonisation of a matrix, depending on the type of carrier. When surface attachment carriers are completely occupied new daughter cells are released into the medium, suggesting that growth is not limited by restricted mass transfer, providing sufficient agitation. For example, van Dieren (1995) described use of DEAE-cellulose immobilised yeast for the production of alcohol-free beer. The reactor was operated at a temperature of 0 to 1 °C under anaerobic conditions using filtered acidified wort. The initial cell count was 50 x 106 cells per gram carrier, and during the initial start-up period this doubled. However, the biomass concentration slowly increased throughout the entire operation to the extent that periodically excess yeast had to be removed to prevent blockages occurring. With this system, it would appear that a stable steady state was never achieved.

Yamauchi and Kashihara (1995) working with a primary fermentation system investigated the effect of varying the dissolved oxygen concentration of wort introduced into a packed bed reactor containing yeast immobilised in calcium alginate. They reported that increasing the oxygen concentration from less than 0.3 ppm to 6.0 ppm had no effect on the bound yeast concentration (c. 1 x 107 cells ml *) but the suspended yeast count increased from 1 x 106 to 1 x 107 cells ml 1. It was assumed that the increase in free yeast count was due to greater liberation of new daughter cells from the carrier. This was consistent with the observation that during prolonged use the viability of the bound yeast gradually declined. Again, this indicates that steady-state conditions are not achieved with these types of continuous reactors.

It is, of course noteworthy that yeast cells have a finite life span and can produce only a certain number of daughters. It would be predicted, therefore, that the mean age of the immobilised cells would gradually increase and eventually a state of general senescence would pertain. On the other hand, non-viable cells could either detach from surface supports or lyse within porous types leaving space for further colonisation. At present, these possibilities are poorly researched.

Where reactors have been designed to have high mass transfer characteristics, growth and assimilation of wort components such as amino nitrogen are comparable with that seen in conventional batch fermentations, as is flavour development in the resultant beers. See, for example Mensour et al. (1995) and Krikilion et al. (1995). However, this is perhaps an over-simplistic view of what are very complex interactions between yeast and the external environment. The influences of mass transfer, residence times and type of reactor are all intertwined. It has been discussed already (Section 5.6.1) that in a plug flow type continuous reactor there is an opportunity for ordered assimilation and metabolism of wort nutrients as in a batch culture. The opportunity for this to occur exists in a packed bed reactor with no back mixing. The extent and rate of uptake of metabolites will depend on flow rate, temperature, provision of oxygen and the efficiency of mass transfer in the reactor.

In a homogeneous reactor mass transfer may be improved by efficient stirring but the continued in-feed of fresh wort may inhibit or repress some processes. For example, in the same way that free glucose represses maltose uptake, the presence of some amino acids inhibits the uptake of others. With stirred immobilised yeast reactors, there is the added complication that diffusion gradients will exist between the fluid and matrix phases. The severity of these effects will depend upon the type of support, the degree of agitation and the flow rate amongst other factors. Bearing in mind these complexities it is unsurprising that it is difficult to predict how any combination of wort, yeast and reactor type will interact and produce beer of a given quality. However, it would be suspected that for primary fermentation, relatively long residence times with reactors that allow considerable re-cycling, offer the best chance of production of balanced beers.

Others have chosen to overcome the problems of restricted mass transfer by using multi-stage systems consisting of a combination of stirred vessels and freely suspended yeast and bioreactors containing immobilised yeast. Yamauchi et al. (1994, 1995a, b) studied primary fermentation using a 'continuous stirred tank reactor' (CSTR) the outflow of which was linked to the in-feed of an immobilised yeast 'packed bed reactor' (PBR). The CSTR was supplied continuously with wort. The yeast loading was reduced in the process flow issuing from the CSTR by continuous centrifugation. Conditions in the CSTR were aerobic to encourage yeast growth and anaerobic in the PBR to promote ethanol formation. Using this process, beer quality and composition was comparable with those made by conventional batch fermentation.

It was observed that biomass yield in the CSTR was approximately ten-fold that of the PBR. Higher alcohol formation occurred largely in the CSTR, whereas ester levels were much higher in the PBR. This pattern was ascribed to the relatively high levels of growth in the CSTR allowing the formation of higher alcohols. These, together with ethanol were then available for esterification in the PBR where growth was restricted. The significant influence was considered to be the ratio of assimilation of fermentable extract to amino nitrogen in each vessel. Virtually no sulphite was generated in the CSTR but comparatively high concentrations were formed in the PBR. Levels in the latter continued to increase over a period of 30 days' continuous use. The observations were ascribed to the differential effects of growth on uptake of amino acids in each vessel. Thus, methionine inhibits sulphite production and this was present in the CSTR because of the continual in-feed of fresh wort. However, in the PBR, the concentration of methionine gradually fell to a non-inhibitory level, thereby allowing reduction of wort sulphate to sulphite.

The production of organic acids differed in each vessel and from a conventional batch fermentation. Overall, higher concentrations of succinate and lower levels of acetate were formed in the multi-vessel system compared to batch fermentation. Virtually no acetate was formed in the CSTR but succinate was produced in both vessels. Elevated succinate in beers produced using immobilised reactors has also been reported by Shindo et al. (1993a). These authors concluded that this was produced through the methylcitric acid pathway due to an increased assimilation of isoleucine compared to freely suspended yeast systems.

Contrary to multi-stage continuous reactors, it has been suggested that primary fermentation can best be achieved by an immobilised yeast reactor followed by a stirred fermenter containing free suspended yeast (Masschelein & Andries, 1995; Andries et al., 1997b). In this scenario, the immobilised reactor provides a constant inoculum of young yeast cells for the stirred reactor. There is no need for intermediary removal of yeast and growth in the stirred reactor is limited by the degree of wort attenuation achieved in the first vessel. Theoretically, this approach should allow better control over ester production since higher alcohol formation may proceed in the immobilised reactor via the biosynthetic route but with limited amino acid uptake. In the subsequent stirred vessel, amino acid assimilation may occur to provide substrates for esterification using the already present higher alcohols and ethanol.

Viewed as a whole these results suggest that a two-stage bioreactor used for primary fermentation offers a practical means of maintaining the volumetric productivity gains of immobilisation and producing standard beer. However, some anomalies remain, suggesting that further development is required.

With regard to continuous flavour maturation and low-alcohol or alcohol-free beer, the physiological effects of yeast immobilisation are to some extent less problematic. The simplest arrangement is that in which primary fermentation is performed in a standard batch fermenter. Green beer is then passed through an immobilised yeast reactor, which has the task of simply reducing diacetyl to a sub-flavour threshold concentration. In practice, there is an added complication in that it is necessary to ensure that all a-acetolactate in the green beer is converted to diacetyl prior to application onto the immobilised bioreactor. Otherwise diacetyl may be formed in finished beer when no yeast is present to remove it. The conversion is carried out by subjecting beer to a controlled heat treatment (5-10 minutes at 90°C). Strictly anaerobic conditions should be observed to remove the possibility of beer oxidation and the free cell concentration should be reduced to as low a level as possible, by efficient centrifugation, to avoid the formation of yeast thermal decomposition products.

A more complex system is where both primary and secondary fermentation are carried out in separate immobilised bioreactors. Ryder et al. (1995) discuss the observation that green beer issuing from immobilised yeast bioreactors used for primary fermentation has elevated levels of a-acetolactate. This may be related to the relatively low rates of assimilation of amino nitrogen compared to sugar utilisation or specific alterations in the pattern of uptake of valine and isoleucine might be implicated. The former would seem the most plausible explanation. Thus, where glycolytic flux rates are high and uptake of amino nitrogen is restricted it would be predicted that the yeast would seek to synthesise amino acids de novo. In this case, rates of synthesis of a-acetolactate from pyruvate may also be enhanced. Where primary fermentation only is carried using an immobilised yeast bioreactor care must obviously be exercised to ensure that the relatively high levels of a-acetolactate are removed in subsequent warm conditioning. If a second immobilised yeast reactor is used for maturation, no problems should arise as long as the intermediate heat step is performed correctly, since yeast has a very high capacity for reduction of the resultant diacetyl.

An elegant approach was described by Kronlof and Linko (1992) in which primary fermentation was performed using a single immobilised cell reactor with a genetically modified yeast strain containing a-acetolactate decarboxylase (see Section 4.3.4). This modification allows the conversion of a-acetolactate directly to acetoin without the formation of diacetyl. Pilot scale trials with this yeast allowed the production of beer with standard analysis in 2-6 days. No maturation period was needed. Of course, the question of public acceptance of beer made with genetically modified yeast remains a commercial bar to such approaches.

For alcohol-free or reduced-alcohol beer production by fermentation, there are three requirements. First, limited alcohol production; second, formation of a normal spectrum of beer flavour compounds other than ethanol; and, third, reduction of carbonyls which contribute to undesirable 'worty' characters. Debourg et al. (1994) have demonstrated that the reduction of wort carbonyls by yeast is complex and involves several enzymes. However, immobilisation had no effect on the ability of yeast to perform this task. Yeast with a repressed physiology, as would be the case in limited wort fermentation, was the most efficient at reduction of wort carbonyls.

Conversion of sugar to alcohol may be limited by use of a combination of low temperature and anaerobiosis. Several reports have indicated that these conditions may be readily achieved using immobilised yeast bioreactors (Breitenbucher & Mistier, 1995; van de Winkel et al., 1995; van Dieren, 1995; Mieth, 1995). The first of these reports compared the sugar spectrum of the in-feed wort and exiting beer. This showed that, during contact with the yeast, there was a reduction in the concentration of sucrose, roughly a doubling of fructose, a slight increase in glucose and no change in levels of maltose and maltotriose. As only sucrose hydrolysis and some uptake of glucose had occurred, limited fermentation was assured since the continued presence of glucose would repress assimilation of maltose.

Production of other flavour metabolites in low- or zero-alcohol beer fermentations, particularly esters, is subject to the same restrictions as discussed already with regard to applications of immobilised yeast to primary fermentation. Indeed, because of limited ethanol and higher alcohol formation it would be predicted that it would be an even more serious problem. Using a reactor with a high mass transfer rate and limited aerobiosis, ester levels were found to be low compared to alcohol-free beers produced from a fully fermented beer with subsequent de-alcoholisation (van de Winkel, 1995). A possible compromise is to use a combination of slightly increased fermentation in the immobilised yeast bioreactor to generate flavour metabolites and reduce wort carbonyls followed by de-alcoholisation. Such an approach has been tested successfully at pilot scale as part of a fully continuous process for alcohol-free beer production (Dziondziak & Seiffert, 1995). However, this does lack the elegance and simplicity of the single pass immobilised yeast reactor approach. In general, it would appear that processes for alcohol-free beer by limited fermentation using immobilised yeast are likely to be most successful where beers with low ester levels are acceptable. For more flavoursome beers, this approach clearly has limitations.

5.7.1.3 Reactor types. There is no reason why immobilised yeast cannot be used in conventional batch fermentation (Atkinson & Taidi, 1995). The advantages afforded would be rapid process times due to the high biomass concentration and ease of separation of green beer. However, undoubtedly the most gains are made using a continuous process. A plethora of fermenter designs have been suggested for use with immobilised yeast for fuel alcohol production. However, for commercial beverage processes the two most common are packed and stirred bed types. Some typical configurations are shown in Fig. 5.42(a) and (b).

In a packed bed reactor, the liquid flow passes through an unstirred bed of the support matrix together with immobilised yeast cells. Therefore, it is a fully closed two-phase system. This method has the advantage of simplicity and the plug flow arrangement may ensure that substrates do not fall to limiting levels and avoids cells being exposed to inhibitory concentrations of metabolites (Masschelein & Andries, 1995). Theoretically, maintenance of ideal plug flow conditions in a fixed bed reactor would allow the various stages of a batch fermentation to be mimicked, and therefore this approach should be the most suitable for carrying out primary fermentation.

In practice, these ideal conditions are difficult to achieve. Fixed bed reactors are prone to channelling, difficulties are encountered in allowing escape of carbon dioxide and some support media, particularly the soft-bead types, are liable to compression, thereby restricting flow and causing high cross-bed pressures to develop. The efficiency of transfer of metabolites and nutrients between cells and their surrounding environment is low in fixed bed reactors since fluid flow is linear and of low velocity. This effect is exacerbated with soft-bead type supports where the

Outlet

In-feed

In-feed

In-leed

Outlet

Outlet t

Re-circulation pump

Top seal t

Re-circulation pump

Top seal

Bottom seal

In-feed Exit

Fig. 5.42 Configurations of immobilised yeast bioreactors. (a) Fixed bed. (b) Fluidised bed. (c) Gas lift, (d) Loop reactor. (Re-drawn from Andries et al.. 1997b.)

Bottom seal

In-feed Exit

Fig. 5.42 Configurations of immobilised yeast bioreactors. (a) Fixed bed. (b) Fluidised bed. (c) Gas lift, (d) Loop reactor. (Re-drawn from Andries et al.. 1997b.)

relatively slow rate of fluid flow may be insufficient to generate a satisfactory diffu-sional gradient through the bead. The generally low mass transfer also makes dissipation of heat difficult and this further mitigates against the use of packed bed reactors for primary fermentation where exothermy is considerable and attempera-tion is critical to product and yeast quality. Some of these disadvantages can be minimised by using a combination of upward flow and a solid support medium, which is not readily compressed. Axelsson (1988) described a novel horizontal packed bed reactor with baffles, which significantly reduced carbon dioxide hold-up and much increased ethanol productivity. However, this approach has not been applied to beer production and fixed bed reactors are most suitable for either partial primary fermentation as in alcohol-free beer production or rapid maturation processes.

Stirred bed reactors are provided with a means of increasing mass transfer rates by forced agitation. This may simply be achieved by providing the vessel with a motorised impeller; however, care must be taken to ensure that the support is not damaged. More commonly, liquid fluidised bed reactors are used. These vessels have a continuous bottom in-feed and top located exit point. A proportion of the process flow exiting from the top is continuously recirculated back into the bottom of the reactor. This arrangement prevents compaction of the bed and thereby ensures relatively high rates of mass transfer between the fluid and cell support medium and presents little restriction to removal of carbon dioxide. The efficiency of mixing can be regulated by control of the relative rates of recirculation and medium addition. Any support medium can be used with fluidised bed reactors although those with a density slightly greater than the suspending medium are the most appropriate since this avoids the possibility of wash-out at high flow rates. Cho and Choi (1981) compared the efficiency of ethanol production from glucose in similar sized packed bed and fluidised bed laboratory reactors containing yeast entrapped in calcium alginate gel. They concluded that the fluidised bed reactor was twice as efficient as the packed bed type-

Other approaches for increasing yeast metabolic activity using bioreactors with improved mass transfer characteristics have been suggested. Mensour et al. (1995) described a gas lift fermenter containing yeast immobilised in K-carrageenan. This type of fermenter has been widely used with filamentous fungi because it has good mixing characteristics but avoids damage to hyphae, which may occur with conventional stirred reactors. Use of such vessels with comparatively fragile gel beads represents a similar application. It comprises a cylindrical vessel with a central draft tube. Wort is continuously fed from the bottom and product from the top. Mixing is achieved by introducing a stream of air and carbon dioxide into the base of the draft tube. This encourages vertical flow up the draft tube and downward movement in region between draft tube and the outer wall of the vessel. Removal of carbon dioxide was via a flared vessel top. K-carrageenan beads were chosen since they have a density close to that of wort, thereby facilitating ease of mixing. Overall, productivity of the vessel was controlled by a combination of regulating flow rate, temperature and the proportion of air in the gas stream. Although used only at pilot scale this vessel was used for primary fermentation with a residence time of 20 hours. If used in conjunction with an immobilised yeast reactor for accelerated maturation a total process time of two days was claimed. No data was provided regarding beer analysis.

Andries et al. (1997b) described a reactor in which yeast was immobilised in silicon carbide. The support consists of a number of ceramic elements, containing silicon carbide matrix, which is colonised by yeast cells, and a number of open channels for circulation of process fluid. When used for primary fermentation, wort is oxygenated and continuously fed into the bottom of the reactor. Fermenting wort is recycled through the silicon carbide elements and the void between the elements and the vessel wall via an external loop. A proportion is withdrawn from the external loop and fed into a second conical vessel. The contents of this vessel, which are inoculated with yeast issuing from the immobilised yeast reactor, are also mixed by circulation through an external loop. Claimed advantages for this system are high rates of mass transfer, reduced risk of clogging, ease of cleaning, ease of carbon dioxide removal and suitability for use with unfiltered wort. It has been used for the production of both ales and lagers apparently of perfectly acceptable flavour at a semi-industrial scale.

5.7.2 Commercial systems

Commercial scale immobilised yeast reactors have been developed for the production of low- or zero-alcohol beers and for continuous maturation. Application at production scale for primary fermentation has yet to be realised; however, a few systems have been developed to semi-industrial scale. For completeness, representative examples of these are described here.

5.7.2.1 Alcohol-free beer. The layout of a typical plant for the production of alcohol-free beer is shown in Fig. 5.43 and is as described by van Dieren (1995) and Mieth (1995). This system uses yeast immobilised on DEAE-cellulose (Spezyme GDC®) and contained within a specifically designed vessel termed the 'Immocon' reactor (Fig. 5.44).

The wort used is of low fermentability and experiences an extensive boiling stage to reduce as fully as possible the concentrations of volatile aldehydes and carbonyls. To avoid oxidation careful handling is required to minimise oxygen pick-up. After cooling, the wort is acidified with lactic acid. This is also produced by a bioreactor, in

Immobilised yeast bioreactors

Holding tank Filter Heat exchanger pH

adjustment

■H

Healthy Chemistry For Optimal Health

Healthy Chemistry For Optimal Health

Thousands Have Used Chemicals To Improve Their Medical Condition. This Book Is one Of The Most Valuable Resources In The World When It Comes To Chemicals. Not All Chemicals Are Harmful For Your Body – Find Out Those That Helps To Maintain Your Health.

Get My Free Ebook


Post a comment