Enzyme and Microbial Technology 32 (2003) 553–559
Xylitol production by Ca-alginate entrapped cells: comparison of different fermentation systems Walter Carvalho a , Silvio S. Silva a,∗ , Julio C. Santos a , Attilio Converti b a
Department of Biotechnology, Faculty of Chemical Engineering of Lorena, Rodovia Itajubá-Lorena, km 74.5, 12.600-000 Lorena, S.P., Brazil b Department of Chemical and Process Engineering, University of Genova, Via Opera Pia 15, 16145 Genova, Italy Received 5 September 2002; received in revised form 24 December 2002; accepted 24 December 2002
Abstract Candida guilliermondii cells were entrapped in Ca-alginate beads and used for xylitol production from concentrated sugarcane bagasse hemicellulosic hydrolysate in three different fermentation systems, namely 125-ml Erlenmeyer flasks (EF), 2.4-l stirred tank reactor (STR) and 2.4-l basket-type stirred tank reactor (BSTR). The EF system provided a xylitol production of 21.0 gP l−1 , a product yield based on xylose consumption of 0.54 gP gS −1 , and an overall production rate of 0.44 gP l−1 h−1 after 48 h of fermentation. By the STR system, 23.5 gP l−1 was produced after 60 h of fermentation, corresponding to a yield of 0.58 gP gS −1 and a production rate of 0.39 gP l−1 h−1 . As the average volume of the beads decreased by 10.7% (p < 0.05) during this fermentation, a basket was fitted into the reactor vessel to prevent the beads abrasion and thus maintain their integrity. However, the xylitol yield (0.46 gP gS −1 ), production rate (0.21 gP l−1 h−1 ) and concentration (15.0 gP l−1 ) attained by this system were affected by mass-transfer limitations. © 2003 Elsevier Science Inc. All rights reserved. Keywords: Xylitol; Sugarcane bagasse hydrolysate; Immobilized cells; Erlenmeyer flasks; Stirred tank reactor; Basket-type stirred tank reactor
1. Introduction Xylitol is a polyalcohol as sweet as sucrose. Unlike the other sweeteners, xylitol is anticariogenic and has a pleasant taste and no unpleasant aftertaste [1]. In addition, this compound has several clinical applications: it can be used as a sugar substitute in low energy and diabetic foods [2], is very well received in post-surgery infusions by patients with difficulty in metabolising sugar [3], improves the biomechanical properties of bones in case of osteoporosis [4] and prevents acute medium otitis [5]. The biotechnological production of xylitol could be a promising alternative to the present chemical reduction, since the downstream processing is expected to be cheaper [6]. However, the concentrations, yields and production rates obtained from fermentation media consisting in lignocellulosic hydrolysates are still the bottlenecks of a large-scale process, although they can be improved by selecting the right fermentation system, operation mode and cultivation conditions [7].
Immobilized cell systems have been traditionally considered as an alternative for increasing the process overall productivity and for minimizing production costs, while offering advantages over free cell fermentation systems [8]. Among the various matrices available for immobilization, Ca-alginate beads stand out, because their production method does not require drastic conditions and, in regard to food applications, utilizes ingredients that are accepted as food additives [9]. The present study discusses and compares the results of three different fermentation systems employed for xylose-to-xylitol bioconversion by Ca-alginate entrapped Candida guilliermondii cells: Erlenmeyer flasks (EF), stirred tank reactor (STR) and basket-type stirred tank reactor (BSTR).
2. Materials and methods 2.1. Preparation and treatment of sugarcane bagasse hydrolysate
Abbreviations: EF, Erlenmeyer flasks; STR, stirred tank reactor; BSTR, basket-type stirred tank reactor ∗ Corresponding author. Tel.: +55-12-553-3209; fax: +55-12-553-3133. E-mail address:
[email protected] (S.S. Silva).
Sugarcane bagasse was hydrolyzed in a 250-l steel reactor as described in a previous study [10]. The liquid fraction of the hydrolysate was separated by centrifugation and
0141-0229/03/$ – see front matter © 2003 Elsevier Science Inc. All rights reserved. doi:10.1016/S0141-0229(03)00007-3
554
W. Carvalho et al. / Enzyme and Microbial Technology 32 (2003) 553–559
Nomenclature Go PF qP QP qS QS So SF VI VM VR XI XM XR XRI XRM YI YP/S YS YX/C
initial glucose concentration (gG l−1 ) final xylitol concentration (gP l−1 ) specific xylitol production rate (gP gx −1 h−1 ) overall xylitol production rate (gP l−1 h−1 ) specific xylose consumption rate (gS gx −1 h−1 ) overall xylose consumption rate (gS l−1 h−1 ) initial xylose concentration (gS l−1 ) final xylose concentration (gS l−1 ) volume of Ca-alginate beads (l) volume of fermentation medium (l) reactor volume (l); VR = VI + VM immobilized cell concentration in the Ca-alginate beads (gx l−1 ) free cell concentration in the fermentation medium (gx l−1 ) total cell concentration in the reactor (gx l−1 ) immobilized cell concentration in the reactor (gx l−1 ) free cell concentration in the reactor (gx l−1 ) yield of immobilization (dimensionless) xylitol yield based on xylose consumption (gP gS −1 ) yield of xylose consumption (dimensionless) cell yield based on xylose and glucose consumption (gx gC −1 )
Subscripts C total carbon source (xylose and glucose) G glucose o starting value P xylitol S xylose x biomass
Fig. 1. Schematic diagram of the stirred tank reactor (STR) used for the xylose-to-xylitol bioconversion.
speed of 500 rpm and an air flowrate of 1.7 l min−1 ; 2.4-l BSTR (Fig. 2) containing 200 g of beads and 1.3 l of medium was operated at an agitation speed of 1300 rpm and an air flowrate of 0.01 l min−1 . As shown in Fig. 2, the basket reactor consisted in a fixed bed arranged in the centre of the conditioning vessel. A peristaltic pump was used for recirculating the medium from the conditioning vessel through the fixed bed (from the centre of the bed to the outer radius) and back to the vessel at a flowrate of 350 ml min−1 . 2.3. Micro-organism, inoculum cultivation and cell immobilization C. guilliermondii FTI 20037, described by Barbosa et al. [12], was used in the experiments. The inoculum was
concentrated three-fold at 70 ◦ C under vacuum. To minimize the amounts of the main fermentation inhibitors, the concentrated hydrolysate was treated according to the method proposed by Alves et al. [11]. 2.2. Medium and fermentation conditions The treated hydrolysate was heated to 110 ◦ C for 15 min and supplemented with ammonium sulphate (3.0 g l−1 ), calcium chloride (0.1 g l−1 ) and rice bran extract (20% v/v) before being used as the fermentation medium. Three different fermentation systems, namely EF, STR and BSTR, were used for xylose-to-xylitol bioconversion as follows: 125-ml EF containing 7.4 g of beads and 40 ml of fermentation medium were maintained in a rotatory shaker at 30 ◦ C and 200 rpm; 2.4-l STR (Fig. 1) containing 200 g of beads and 1.3 l of medium was operated at an agitation
Fig. 2. Schematic diagram of the basket-type stirred tank reactor (BSTR) used for the xylose-to-xylitol bioconversion.
W. Carvalho et al. / Enzyme and Microbial Technology 32 (2003) 553–559
cultivated and the cells were immobilized in Ca-alginate beads following a procedure previously described [13]. The average diameter of the resulting beads was 2.70 mm. Fermentations were performed in all reactor configurations using the same starting biomass concentration in the beads (12–13 gx l−1 ), corresponding to a total biomass concentration in the reactor of 1.6–1.7 gx l−1 . 2.4. Analytical methods Xylose and xylitol concentrations were determined by high performance liquid chromatography as already described [10]. Free and immobilized cell concentrations were determined by absorbance at 600 nm and correlated with the cell dry-weight through a corresponding calibration curve. The liquid phase of the samples taken during the fermentation runs was centrifuged (2000 × g, 15 min) and the cells were resuspended in water for the determination of the free cell concentrations. A known mass (accuracy within 0.01 g) of Ca-alginate beads taken during the fermentation runs and previously dried with an absorbent paper was dissolved in a 2.0% potassium citrate solution under agitation. The resulting suspension was centrifuged (2000 × g, 15 min) and the cells were resuspended in water for the determination of the immobilized cell concentrations. The density of the Ca-alginate beads, assumed to be constant during the fermentation runs, was 1.0 g ml−1 . The cellular growth was considered either as the concentration of free cells in the fermentation medium (XM ) and of immobilized cells in the Ca-alginate beads (XI ), or as the total cell concentration in the reactor (XR ). The following material balance was used to determine the total cell concentration in the reactor (for symbols see Nomenclature): VR XR = VI XI + VM XM
(1)
The free cell concentration in the fermentation medium (XM ) and the immobilized cell concentration in the Ca-alginate beads (XI ) were related to the reactor working volume (VR ) by the following equations: XRM = XRI =
X M VM VR
XI VI VR
(2) (3)
The stability of the Ca-alginate beads was estimated by monitoring their solubilization during the batch fermentations. The average diameter of the beads was measured (accuracy within 0.05 mm) at the beginning and at the end of each run. The average volumes of the beads were then calculated assuming a spherical geometry. The beads solubilization was finally expressed as the percentage of volume reduction.
555
2.5. Calculation of fermentation parameters The yield of immobilization (YI ) was considered as the ratio of immobilized cells (XRI ) to total cells in the reactor (XR ) at the end of each run; the yield of xylose consumption (YS ) was considered as the ratio of xylose consumption at the end of each run (So − SF ) to starting xylose level (So ); the xylitol yield (YP/S ) was considered as the ratio of final xylitol concentration (PF ) to (So − SF ); the yield of growth (YX/C ) was considered as the ratio of total biomass produced (XR ) to total carbon source consumption (So − SF + Go ), including glucose (Go ). The overall xylose consumption rate (QS ) was defined as the amount of xylose consumed in the overall fermentation time per unit reactor volume, while the specific xylose consumption rate (qS ) was the ratio of QS to the final total cell concentration in the reactor (XR ). Analogously, the overall xylitol production rate (QP ) was defined as the amount of xylitol produced in the overall fermentation time per unit reactor volume, while the specific xylitol production rate (qP ) was the ratio of QP to the final total cell concentration in the reactor (XR ).
3. Results and discussion The possibility of producing xylitol with Ca-alginate entrapped cells in synthetic xylose solutions [14–16] and in lignocellulosic hydrolysates [17,18] has been previously demonstrated. By adjusting the immobilization conditions [10], it was possible to reuse the immobilized cells during five successive batches in a concentrated sugarcane bagasse hydrolysate without losses in the bioconversion yields and rates, product concentration and Ca-alginate beads integrity [13]. In the present study, the xylose-to-xylitol bioconversion was performed using three different fermentation systems, namely EF, STR, and BSTR. As can be seen in Fig. 3, free cells grew significantly in the fermentation medium in all the experiments, thus characterizing the fermentation systems as a combination of immobilized and free cells instead of only immobilized cells. Such a significant free cell growth has, likewise, been reported in synthetic medium [15] and sugarcane bagasse hydrolysate [19]. It has been proposed that the cell leakage from the alginate beads occurs naturally due to the appearance of crater-like pores on the gel surface determined by the growth of the immobilized cells [20]. However, as can be seen in Table 1, the lower yield of immobilization (YI ) observed at the end of the fermentation carried out in STR suggests that the beads abrasion resulting from direct contact with the agitation turbines has contributed to the increase in cell leakage from the alginate beads. Table 1 shows the performance of the fermentation systems employed for the xylose-to-xylitol bioconversion. The EF system provided a xylitol production of 21.0 gP l−1 with
556
W. Carvalho et al. / Enzyme and Microbial Technology 32 (2003) 553–559
Fig. 3. Xylose-to-xylitol bioconversion in: (a) EF; (b) STR; (c) BSTR. Immobilized cell concentration in the beads, XI (䊏), and in the reactor, XRI (䊐); free cell concentration in the medium, XM (䊉), and in the reactor, XRM (䊊); total cell concentration in the reactor, XR ().
a yield based on xylose consumption of 0.54 gP gS −1 , the overall production rate reaching 0.44 gP l−1 h−1 after 48 h of fermentation. As can be seen in Table 2, these results are very close to those obtained with a similar system [13] and better than those obtained with a system characterized by higher mass-transfer limitations [17]. However, the use of synthetic xylose solutions instead of hemicellulosic hydrolysates resulted in final product concentrations and overall production
rates at least four times higher than ours [15,16]. These increments attained with synthetic media can reasonably be ascribed to the higher xylose concentrations at the beginning of the fermentation runs as well as to the absence of fermentation inhibitors. As is well known, the bioconversion of xylose-containing hemicellulosic hydrolysates is hindered by the presence of toxic compounds, which limit the increase in xylose concentration by vacuum evaporation
W. Carvalho et al. / Enzyme and Microbial Technology 32 (2003) 553–559 Table 1 Performance of the fermentation systems employed for xylose-to-xylitol bioconversion in sugarcane bagasse hydrolysate Fermentation systems
PF (gP l−1 ) QP (gP l−1 h−1 ) qP (gP gx −1 h−1 ) YP/S (gP gS −1 ) QS (gS l−1 h−1 ) qS (gS gx −1 h−1 ) YI YS YX/C (gx gC −1 ) a b
EFa
STRb
BSTRb
21.0 0.44 0.082 0.54 0.82 0.152 0.70 0.90 0.089
23.5 0.39 0.047 0.58 0.68 0.082 0.62 0.89 0.149
15.0 0.21 0.043 0.46 0.45 0.092 0.69 0.71 0.089
So = 43.4 gS l−1 ; XRo = 1.6 gx l−1 . So = 45.5 gS l−1 ; XRo = 1.7 gx l−1 .
[7]. Therefore, the combined effects of xylose and inhibitors concentrations have to be taken into account in determining the adequate hydrolysate concentration for xylitol production from hemicellulosic hydrolysates. The oxygen availability within the beads is probably the discriminating factor responsible for the large variations observed in the xylitol yields (YP/S ) reported in the literature (Table 2), since different conditions have been used for immobilizing and cultivating the cells. In this context, it should be noted that a xylitol yield close to the maximum theoretical value could only be reached with the use of a recombinant Saccharomyces cerevisiae strain [14]. On the other hand, increases in xylitol yield have been obtained by recycling the immobilized cells through repeated batch fermentation systems [13,16]. As previously observed [13], these increments in xylitol yield could have resulted from the limited growth of the immobilized cells after the first fermentation batch as well as from cell adaptation to the toxic compounds present in the hydrolysate medium. Stirred tank reactors have been used to cultivate entrapped cells even though the catalysts are thus exposed to a rather high degree of shear. Natural polymers used to immobilize the cells include pectin [21], carrageenan [22] and alginate [22–24]. The STR has the advantage of
557
an easy control of temperature and pH and its continuous operation may be useful in the case of substrate inhibition [25]. As shown in Table 1, EF and STR gave similar values of final xylitol concentration (PF ), yield (YP/S ) and overall production rate (QP ). However, the specific rates of xylose consumption (qS ) and xylitol production (qP ), which reached 0.152 gS gx −1 h−1 and 0.082 gP gx −1 h−1 respectively with the EF system, dropped to 0.082 gS gx −1 h−1 and 0.047 gP gx −1 h−1 when STR was employed. It is possible that the better oxygen transfer in the STR bulk might have favored the formation of bigger immobilized cell colonies in the outer bead surface, limiting the oxygen transfer to the inner cells. Therefore, possible anaerobic zones inside the beads could have led to a decrease in the rate of xylose transport across the plasma membrane. According to Spencer-Martins [26], the shift to anaerobic conditions can alter the xylose transport mechanism across the plasma membrane in xylose-fermenting yeasts, leading to a comparatively lower rate of its entry into the cells. A reduction of 10.7% (p < 0.05) in the average volume of the Ca-alginate beads was observed at the end of the fermentation carried out in STR. It has been reported that fitting a basket (fixed bed) into the reactor vessel allows the maintenance of the beads integrity and offers some advantages over the traditional packed bed reactors [27]. Basket-type stirred tank reactors containing Ca-alginate entrapped cells were successfully used to produce ethanol [28] and lactic acid [29]. In the present study, however, the use of a BSTR did not produce good results. In fact, as can be seen in Table 1, only 15.0 gP l−1 was produced after 72 h of fermentation, corresponding to an overall production rate of 0.21 gP l−1 h−1 . Roca et al. [14], using a packed-bed reactor to produce xylitol in continuous mode, came to excellent values of xylitol yield (1.03 gP gS −1 ) and productivity (5.80 gP l−1 h−1 ), in spite of having utilized synthetic xylose medium instead of hemicellulosic hydrolysate. Fig. 3c shows the growth profiles of free and immobilized cells during fermentation in BSTR. Unlike EF (Fig. 3a) and STR (Fig. 3b), the Ca-alginate beads were apparently full of cells only after 60 h of fermentation, thus suggesting a poor transfer of oxygen and nutrients by perfusion to the immobilized cells. Moreover, it was necessary to use a quite high
Table 2 Data from the literature on the use of Ca-alginate entrapped cells for xylitol production from synthetic media and lignocellulosic hydrolysates Operation mode
Medium
Reactor
So (gS l−1 )
PF (gP l−1 )
QP (gP l−1 h−1 )
YP/S (gP gS −1 )
Reference
C B RB
Synthetic Synthetic Synthetic
PB EF EF
50 170 120
Eucalyptus hydrolysate Sugarcane hydrolysate
EF EF
12 50
5.80 2.16 0.31a 1.97b 0.016 0.50a 0.58b
1.03 0.50 0.37a 0.74b 0.50 0.49a 0.56b
[14] [15] [16]
B RB
16.0 86.2 37.6a 94.3b 3.4 23.8a 27.6b
B = batch; C = continuous; EF = Erlenmeyer flasks; PB = packed bed; RB = repeated batch. a First batch. b Fifth batch.
[17] [13]
558
W. Carvalho et al. / Enzyme and Microbial Technology 32 (2003) 553–559
agitation speed to keep the free cells homogeneously suspended in the fermentation medium. Due to the fact that the reactor height-to-diameter ratio was high and the basket was placed right over the agitation turbine, a poor mixing at the top of the vessel was observed even at 1300 rpm. Below this agitation speed, unhomogeneity of the mixing pattern took place and the free cells tended to adhere to the upper walls of both the vessel and the basket, thus hindering the xylose-to-xylitol bioconversion (data not shown). It is noteworthy that the cell growth was different in all the three fermentation systems (Fig. 3), leading to variations in the bioconversion rates and yields. While a better oxygen transfer to the STR bulk favored the growth of both immobilized and free cells, the limited transfer of oxygen and nutrients in the BSTR system may have been responsible for the slow growth of the immobilized cells. As a consequence, the specific rates of xylose consumption (qS ) and xylitol production (qP ) in both systems were much lower than in the EF system (Table 1). In fact, an adequate semiaerobic condition is a fundamental requisite to ensure a good xylitol production by the yeast C. guilliermondii FTI 20037, since anaerobic or aerobic conditions increase the fraction of xylose consumed in alternative pathways [19]. Therefore, the results of the present work not only show that the rate of oxygen transfer to the cells have a strong influence on the xylose-to-xylitol bioconversion, but also suggest that, depending on the system employed, the degree of limitation in the oxygen diffusion into the beads affects the immobilized cells growth. From the literature it can be concluded that the use of immobilized cell systems would offer the opportunity of performing a continuous xylose-to-xylitol bioconversion using dilution rates higher than the maximum specific growth rate of the micro-organism [30], provided that a rigorous control of oxygen supply could be ensured. Besides, the xylose-to-xylitol bioconversion rates and yields attained with the use of immobilized-cell systems, although lower than those observed for free cell cultivation systems, could be improved by selecting the right fermentation system, operation mode, cultivation conditions and mainly by controlling the oxygen supply. Future efforts will be directed to evaluating other immobilization supports and reactor configurations in order to improve the oxygen distribution in the reactor. Acknowledgments We gratefully acknowledge the financial support of FAPESP and CNPq (Brazil). We thank Ms. M.E.M. Coelho for revising this manuscript. References [1] Mäkinen KK. Dietary prevention of dental caries by xylitol—clinical effectiveness and safety. J Appl Nutrition 1992;44:16–28.
[2] Bakr AA. Application potential for some sugar substitutes in some low energy and diabetic foods. Nahrung 1997;41:170–5. [3] Mäkinen KK. Can the pentitol–hexitol theory explain the clinical observations made with xylitol? Med Hypotheses 2000;54:603– 13. [4] Mattila PT, Svanberg MJ, Jamsa T, Knuuttila MLE. Improved bone biomechanical properties in xylitol-fed aged rats. Metabolism 2002;51:92–6. [5] Uhari M, Tapiainen T, Kontiokari T. Xylitol in preventing acute otitis media. Vaccine 2000;19:144–7. [6] Winkelhausen E, Kuzmanova S. Review: microbial conversion of d-xylose to xylitol. J Ferment Bioeng 1998;86:1–14. [7] Parajó JC, Dom´ınguez H, Dom´ınguez JM. Biotechnological production of xylitol. Part 3: operation in culture media made from lignocellulose hydrolysates. Biores Technol 1998;66:25–40. [8] Ramakrishna SV, Prakasham RS. Microbial fermentations with immobilized cells. Curr Sci 1999;77:87–100. [9] Champagne CP, Blahuta N, Brion F, Gagnon C. A vortex-bowl disk atomizer system for the production of alginate beads in a 1500-liter fermentor. Biotechnol Bioeng 2000;68:681–8. [10] Carvalho W, Silva SS, Converti A, Vitolo M, Felipe MGA, Roberto IC, et al. Use of immobilized Candida yeast cells for xylitol production from sugarcane bagasse hydrolysate: cell immobilization conditions. Appl Biochem Biotechnol 2002;98/100:489– 96. [11] Alves LA, Felipe MGA, Silva JBA, Silva SS, Prata AMR. Pretreatment of sugar cane bagasse hemicellulose hydrolysate for xylitol production by Candida guilliermondii. Appl Biochem Biotechnol 1998;70/72:89–98. [12] Barbosa MFS, Medeiros MB, Mancilha IM, Schneider H, Lee H. Screening of yeasts for production of xylitol from d-xylose and some factors which affect xylitol yield in Candida guilliermondii. J Ind Microbiol 1988;3:241–51. [13] Carvalho W, Silva SS, Vitolo M, Felipe MGA, Mancilha IM. Improvement in xylitol production from sugarcane bagasse hydrolysate achieved by the use of a repeated-batch immobilized cell system. Z Naturforsch 2002;57c:109–12. [14] Roca E, Meinander N, Hahn-Hägerdal B. Xylitol production by immobilized recombinant Saccharomyces cerevisiae in a continuous packed-bed bioreactor. Biotechnol Bioeng 1996;51:317–26. [15] Yahashi Y, Hatsu M, Horitsu H, Kawai K, Suzuki T, Takamizawa K. D-glucose feeding for improvement of xylitol productivity from d-xylose using Candida tropicalis immobilized on a non-woven fabric. Biotechnol Lett 1996;18:1395–400. [16] Dom´ınguez JM. Xylitol production by free and immobilized Debaromyces hansenii. Biotechnol Lett 1998;20:53–6. [17] Dom´ınguez JM, Cruz JM, Roca E, Dom´ınguez H, Parajó JC. Xylitol production from wood hydrolysates by entrapped Debaryomyces hansenii and Candida guilliermondii cells. Appl Biochem Biotechnol 1999;81:119–30. [18] Carvalho W, Silva SS, Vitolo M, Mancilha IM. Use of immobilized Candida cells on xylitol production from sugarcane bagasse. Z Naturforsch 2000;55c:213–7. [19] Carvalho W, Silva SS, Converti A, Vitolo M. Metabolic behavior of immobilized Candida guilliermondii cells during batch xylitol production from sugarcane bagasse acid hydrolyzate. Biotechnol Bioeng 2002;79:165–9. [20] Quirós C, Rendueles M, Garcia LA, Diaz M. Diffusion of microorganisms in calcium alginate beads. Biotechnol Tech 1995;9: 809–14. [21] Kesava SS, Panda T. Ethanol production by immobilized whole cells of Zymomonas mobilis in a continuous flow expanded bed bioreactor and a continuous flow stirred tank bioreactor. J Ind Microbiol 1996;17:11–4. [22] Lamboley L, Lacroix C, Champagne CP, Vuillemard JC. Continuous mixed strain mesophilic lactic starter production in supplemented whey permeate medium using immobilized cell technology. Biotechnol Bioeng 1997;56:502–16.
W. Carvalho et al. / Enzyme and Microbial Technology 32 (2003) 553–559 [23] Qureshi N, Manderson GJ. Ethanol production from sulphuric acid wood hydrolysate of Pinus radiata using free and immobilized cells of Pichia stipitis. J Ind Microbiol 1991;7:117–22. [24] Hernández H, Asanza ML, Alcaraz I, Hidalgo AM, Máximo MF, Bastida J. Production of optically pure l-alanine in batch and continuous stirred tank reactors using immobilized Pseudomonas sp. Biotechnol Lett 2001;23:887–91. [25] Fukuda H. Immobilized microorganism bioreactors. In: Asenjo JA, Merchuk JC, editors. Bioreactor system design. New York: Marcel Dekker; 1994. p. 339–75. [26] Spencer-Martins I. Transport of sugars in yeasts: implications in the fermentation of lignocellulosic materials. Biores Technol 1994;50:51–7.
559
[27] Baron GV, Willaert RG, Backer L. Immobilized cell reactors. In: Willaert RG, editor. Immobilized living cell systems: modelling and experimental methods. London: Wiley; 1996. p. 67–95. [28] Gamarra JA, Cuevas CM, Lescano G. Production of ethanol by a stirred catalytic basket reactor with immobilized yeast cells. J Ferment Technol 1986;64:25–8. [29] Abelyan VA, Abelyan LA. Production of lactic acid by immobilized cells in stirred reactors. Appl Biochem Microbiol 1996;32:495–9. [30] Bailey JE, Ollis DF. Biochemical engineering fundamentals. 2nd ed. New York: McGraw Hill; 1986.
