- Email: [email protected]
Hollow fiber bioreactors with internal aeration circuits
JOURNAL OF FERMENTATION AND BIOENGINEERING
Vol. 69, No. 3, 175-177. 1990
Hollow Fiber Bioreactors with Internal Aeration Circuits BONG HYUN CHUNG 1. AND HO NAM CHANG 2 Genetic Engineering Center, Korea Institute of Science and Technology, l and Department of Chemical Engineering, Korea Advanced Institute of Science and Technology, 2 p. 0. Box 131, Cheongryang, Seoul, Korea Received 13 December 1989/Accepted 10 January 1990
New hollow fiber bioreactors for aerobic culture were introduced and Aspergillus niger for citric acid production was cultivated as a model system. These reactors consisted of a bundle mixed of hydrophilic membranes for liquid nutrient transport and hydrophobic membranes for gaseous nutrient transport. The cells were successfully cultivated. However, the polymeric hollow fiber membranes were compressed and blocked by excessive fungal cell growth. Citric acid was produced with a high volumetric productivity compared with that of shake-flask fermentation, but the long-term operation was not successful due to a rapid decrease of the production rate.
Immobilization of enzymes, microbial cells, and animal and plant cells in a hollow fiber reactor has several merits compared with other methods (1). However, supply and removal of gasses are rather difficult in conventional hollow fiber bioreactors, which makes them unsuitable for aerobic microbial cell cultures requiring high oxygen tension. Inloes et al. (2) studied fl-lactamase production by an aerobic cell culture of Escherichia coli C600 (pBR 322) in a hollow fiber bioreactor. However, due to the leakage of the liquid nutrient and the cells by ultrafiltration into the shell side, reactor operation was rather difficult. Recently, dual hollow fiber bioreactors suitable for aerobic microbial cell culture were developed (3-8). Rifamycin B was produced for longer than 50 d in a dual hollow fiber bioreactor system where Nocardia mediterranei was immobilized and citric acid was successfully produced with immobilized Aspergillus niger. However, the reactor design was so complicated that it could not be easily scaled up. Thus, a novel aerobic hollow fiber bioreactor approach with a simpler structure is necessary for easy scaleup. In this study we introduce a new hollow fiber bioreactor system with a simpler structure compared with a dual hollow fiber bioreactor and report the possibility of its application to aerobic fermentation by cultivating a strictly aerobic fungus, A. niger, as a model system.
the reactor body and the two types of membranes were separately sealed with silicone rubber at each end of the reactor body. Liquid and gaseous nutrients were passed through the lumen sides of the hydrophilic and hydrophobic hollow fiber membranes, respectively, and A. niger cells were cultivated within the shell side. Figure 2 shows the detailed structure and the dimensions of the reactor. In this study two reactors with different kinds of hollow fiber membranes were used. The detailed specifications of these two reactors are shown in Table 1. The polypropylene hollow fiber membranes with hydrophobicity were used for gaseous nutrient transport in reactor A, and also these were used for liquid nutrient transport in reactor B after wetting with 50%0 (v/v) ethanol as described in our previous report (4). Bioreactor operation The reactor and the medium flow lines were sterilized by recirculation of 5 ~ formalin solution. After washing out the formalin solution with ample sterilized distilled water, the medium was pumped from the reservoir to the reactor. The experimental set-up was similar to that described in our previous report (5) except that the newly-designed bioreactors were used. After 7 d of culture in the shake flask, the broth including cell debris was inoculated into the reactor by a syringe. During the continuous reactor operation the temperature was
~ G L A S S
MATERIALS AND METHODS Microorganism and media The microorganism used in this study was A. niger B60 obtained from Rohr and Kubicek of the Technische Universitat Wien (Austria), who used this strain for citric acid production (9-11). The composition of the growth and production media used in the shake-flask culture and reactor operation in g/l of deionized water as follows: sucrose 60, NH4NO3 2.5, KH2PO4 1, and MgSO4.7H20 0.25. The medium pH was adjusted to 3.1 before sterilization at 121°C for 15 min. Bioreactor design Figure 1 shows the conceptual design of the reactor. A bundle mixed of hydrophilic and hydrophobic hollow fiber membranes was inserted into
# #/~ ~
•
f,~O~
I
~-J,V. 0 i i ~ J .• V * ~ ,
, T..L
\,l,,I,
~
TUBE HYDROPHOBI MEMBRANE C MICROBES
T/..
/
~
./
u
H
~ I /
YDROPHILIC
MEMBRANE
u,o
NU,RIENT
GASEOUS NUTRIEN]
FIG. ]. Conceptualdesignof the aerobichollow fiberbioreactor.
* Corresponding author. 175
176
CHUNG AND CHANG
J. FERMENT. BIOENG.,
INOCULUM GASEOUS NUTRIENT (~
_j
silicone tubing
L__/ con.ector
_J I___
l
~ = ~ ~ ~
L I Q U I D NUTRIEI~ IN
IN
"---'~)'O TLIQUID UNUTRIENT
m
t
GASEOUS
s i l i c o n e rubber sealing
C
NUTRIENTI
OUT
20 cm
<
FIG. 2.
Detailed structure and dimension of the reactor.
maintained at 30°C, and the liquid nutrient and pure oxygen flowed at 2 ml/h and 100 ml/min, respectively. Analytical methods The citric acid concentration was assayed by the method of Marier and Boulet (12). RESULTS AND DISCUSSION
Figure 3 shows the citric acid concentration and pH in the effluent during the continuous operation of reactor A. During the first two days of operation, little citric acid was TABLE 1.
Detailed specifications of reactor A and B Reactor A
Reactor body Material of phil. No. of phil. Size of phil. Manufacturer of phil. Pore size of phil. Material of pho. No. of pho. Size of pho. Manufacturer of pho. Pore size of pho.
Reactor B
Glass (0.8 cm i.d.) Polyamide 16 0.06 cm i.d. 0.11 cm o.d. Berghof, FRG Mol. cut-off, 50,000 Polypropylene 32 0.033 cm i.d. 0.063 cm o.d. Enka, FRG 0.4-0.6/~m
Glass (0.8 cm i.d.) Polypropylene 40 0.033 cm i.d. 0.063 cm o.d. Enka, FRG 0.4-0.6/~m Teflon 5 O.l cm i.d. 0.2cm o.d. Sumitomo, Japan 1 pm
phil. and pho. denote the hydrophilic and hydrophobic hollow fiber membranes, respectively.
~3
2.5
~
2.0
2
1 0
/ n
0
produced, but decreasing effluent pH indicated cell growth in the reactor. The citric acid concentration increased dramatically from the 4th d to the 8th d while the effluent pH remained relatively constant. At the 8th d citric acid concentration reached about 3.5 g/l with a volumetric productivity of 0.7g//.h, based on the inner glass tube volume (10 ml). This productivity corresponds to a 12-fold increase over that (0.06 g/l. h) of the shake flask fermentation (7). However, after 8 d of operation citric acid production rapidly decreased and the citric acid concentration reached a constant level of about 1 g/l. After the reactor operation, the medium inlet part (C in Fig. 2) was sectioned with a razor blade and the cross section was photographed for more detailed observation (Fig. 4). The reactor was full of cells and, surprisingly, most of the polyamide hollow fiber membranes were compressed and blocked by excessive fungal cell growth. A similar phenomenon was also observed with fungal cultures of A. niger and Humicola sp. in the dual hollow fiber bioreactors (7, 8), where the flexible silicone tubes were expanded by excessive fungal cell growth, and it ultimately blocked the air flow path and restricted the substrate flow rate. It was believed that the rapid decrease of the citric acid production shown in Fig. 3 was caused by the nonuniform flow of liquid nutrient resulting from the blockage of polyamide hollow fiber membranes. Figure 5 shows the citric acid and pH in the effluent dur-
15
~b--/'--
I
I
I
I
I
2
~,
6
8
10
12
1.0 lZ,
Days FIG. 3. Citric acid concentration (©) and pH ( • ) in the effluent during reactor A operation.
FIG. 4. Photograph showing the cross section of reactor A after operation. PP, Polypropylene hollow fiber membrane; PA, polyamide hollow fiber membrane; AN, Aspergillus niger cells.
VOL 69, 1990
HOLLOW FIBER BIOREACTORS
3.5 3.0 ~3
2.5 -r
- 2.0
-~' -~._- 21
0
0
g.=ojl.5 n
2
t~
6
8 Days
10
12
FIG. 5. Citric acid concentration (©) and pH ( • ) in the effluent during reactor B operation (Q denotes the flow rate of liquid nutrient). ing the c o n t i n u o u s operation of reactor B. Profiles of citric acid concentration and pH were similar to those of Fig. 3. At the 7th d the m a x i m u m concentration of citric acid was produced at a concentration of 2 g / l , which was much lower than that of Fig. 3 (3.5 g//). One of reasons for the lower production of citric acid can be explained by the difference of gas-liquid interfacial areas in these two reactors which are equal to the outer surface areas of hydrophobic hollow fiber membranes used for aeration. The outer surface area (126.7 cm 2) of polypropylene membranes used in reactor A for aeration was larger than that (62.8 cm 2) of the Teflon membranes used in reactor B, and thus the oxygen transfer rate in reactor A should be larger than that in reactor B. In general, citric acid fermentation requires a high oxygen tension because oxygen is used directly for the product formation as well as the maintenance and growth of the cells. Thus, we concluded that the increased oxygen transfer rate resulted in a significant increase of the volumetric productivity. At the 14th d the flow rate of liquid nutrient suddenly dropped to zero, and thus further reactor operation had to be discontinued. It was believed that this was caused by the complete
177
blockage of the polypropylene hollow fiber membranes used for the liquid nutrient transport. Figure 6 shows the cross sectional view of reactor B, where complete blockage of the polypropylene membranes was not observed. But it would be certain that the polypropylene membranes in other regions of reactor B were blocked, which consequently restricted the liquid flow. These operational problems occurred due to the excessive fungal cell growth were also observed in the dual hollow fiber bioreactor operation for the purpose of producing citric acid by the same fungus, where we (7) solved these problems by repressing excessive cell growth using a nitrogen-deficient medium. In this study we could not try this experiment because blockage of the hollow fiber membranes happened suddenly. However, it would be possible to carry out a successful long-term operation if the cell growth can be appropriately controlled. In conclusion, aerobic fungal cell cultures using these hollow fiber bioreactor systems were possible, but longterm operation was not successful because of excessive fungal cell growth in the reactor that resulted in deformation of the polymeric hollow fiber membranes. Controlled cell growth in the reactor and use of stronger hollow fibers that can resist deformation would facilitate successful long term operation of these reactors. REFERENCES 1. Vick Roy, T.B., Blanch, H . W . , and Wilke, C.R.: Microbial
hollow fiber bioreactors. Trends in Biotechnol., 1, 135-139 (1983). 2. Inloes, D . S . , Smith, W . J . , Taylor, D . P . , Cohen, S . N . , Michaels, A. S., and Robertson, C. R.: Hollow-fiber membrane
3. 4. 5. 6.
7.
bioreactors using immobilized E. coli for protein synthesis. Biotechnol. Bioeng., 25, 2653-2681 (1983). Robertson, C.R. and Kim, I.H.: Dual aerobic hollow-fiber bioreactor for cultivation of Streptomyces aureofaciens. Biotechnol. Bioeng., 27, 1012-1020 (1985). Chang, H. N., Chung, B. H., and Kim, I. H.: Dual hollow-fiber bioreactor for aerobic whole cell immobilization. ACS (Am. Chem. Soc.) Symp. Ser., 314, 32-42 (1986). Chang, H. N., Kyung, Y. S., and Chuog, B. H.: Glucose oxidation in a dual hollow fiber bioreactor with a silicone tube oxygenator. Biotechnol. Bioeng., 29, 552-557 (1987). Chung, B. H., Chang, H. N., and Kim, I. H.: Rifamycin B production by Nocardia mediterranei immobilized in a dual hollow fiber bioreactor. Enzyme Microb. Technol., 9, 345-349 (1987). Chung, B. H. and Chang, H. N.: Aerobic fungal cell immobilization in a dual hollow fiber bioreactor: continuous production of citric acid. Biotechnol. Bioeng., 32, 205-212 (1988).
8. Hwang, Y. B., Chung, B . H . , Chang, H. N., and Han, M. H.:
Enzymatic conversion of rifamycin B by live Humicola sp. immobilized in a dual hollow-fiber bioreactor. Bioprocess Eng., 3, 159-163 (1988). 9. Rohr, M., Stadler, P . J . , Salzbrunn, W . O . J . , and Kubicek,
FIG. 6. Photograph showing the cross section of reactor B after operation. PP, polypropylene hollow fiber membrane; TF, Teflon hollow fiber membrane; AN, Aspergillus niger cells.
C. P.: An improved method for characterization of citrate production by conidia of Aspergillus niger. Biotechnol. Lett., 1, 281-286 (1980). 10. Rohr, M., Zehentgruber, O., and Kubicek, C.P.: Kinetics of biomass formation and citric acid production by Aspergillus niger on pilot plant scale. Biotechnol. Bioeng., 23, 2433-2445 (1981). 11. Misehak, H., Kubicek, C, P., and Rohr, M.: Citrate inhibitionof glucose uptake in Aspergillus niger. Biotechnol. Lett., 6, 425-430 (1984). 12. Marier, J. R. and Boulet, M. L.: Direct determination of citric acid in milk with an improved pyridine-acetic anhydride method. J, Dairy Sci., 41, 1683-1692 (1958).
Vol. 69, No. 3, 175-177. 1990
Hollow Fiber Bioreactors with Internal Aeration Circuits BONG HYUN CHUNG 1. AND HO NAM CHANG 2 Genetic Engineering Center, Korea Institute of Science and Technology, l and Department of Chemical Engineering, Korea Advanced Institute of Science and Technology, 2 p. 0. Box 131, Cheongryang, Seoul, Korea Received 13 December 1989/Accepted 10 January 1990
New hollow fiber bioreactors for aerobic culture were introduced and Aspergillus niger for citric acid production was cultivated as a model system. These reactors consisted of a bundle mixed of hydrophilic membranes for liquid nutrient transport and hydrophobic membranes for gaseous nutrient transport. The cells were successfully cultivated. However, the polymeric hollow fiber membranes were compressed and blocked by excessive fungal cell growth. Citric acid was produced with a high volumetric productivity compared with that of shake-flask fermentation, but the long-term operation was not successful due to a rapid decrease of the production rate.
Immobilization of enzymes, microbial cells, and animal and plant cells in a hollow fiber reactor has several merits compared with other methods (1). However, supply and removal of gasses are rather difficult in conventional hollow fiber bioreactors, which makes them unsuitable for aerobic microbial cell cultures requiring high oxygen tension. Inloes et al. (2) studied fl-lactamase production by an aerobic cell culture of Escherichia coli C600 (pBR 322) in a hollow fiber bioreactor. However, due to the leakage of the liquid nutrient and the cells by ultrafiltration into the shell side, reactor operation was rather difficult. Recently, dual hollow fiber bioreactors suitable for aerobic microbial cell culture were developed (3-8). Rifamycin B was produced for longer than 50 d in a dual hollow fiber bioreactor system where Nocardia mediterranei was immobilized and citric acid was successfully produced with immobilized Aspergillus niger. However, the reactor design was so complicated that it could not be easily scaled up. Thus, a novel aerobic hollow fiber bioreactor approach with a simpler structure is necessary for easy scaleup. In this study we introduce a new hollow fiber bioreactor system with a simpler structure compared with a dual hollow fiber bioreactor and report the possibility of its application to aerobic fermentation by cultivating a strictly aerobic fungus, A. niger, as a model system.
the reactor body and the two types of membranes were separately sealed with silicone rubber at each end of the reactor body. Liquid and gaseous nutrients were passed through the lumen sides of the hydrophilic and hydrophobic hollow fiber membranes, respectively, and A. niger cells were cultivated within the shell side. Figure 2 shows the detailed structure and the dimensions of the reactor. In this study two reactors with different kinds of hollow fiber membranes were used. The detailed specifications of these two reactors are shown in Table 1. The polypropylene hollow fiber membranes with hydrophobicity were used for gaseous nutrient transport in reactor A, and also these were used for liquid nutrient transport in reactor B after wetting with 50%0 (v/v) ethanol as described in our previous report (4). Bioreactor operation The reactor and the medium flow lines were sterilized by recirculation of 5 ~ formalin solution. After washing out the formalin solution with ample sterilized distilled water, the medium was pumped from the reservoir to the reactor. The experimental set-up was similar to that described in our previous report (5) except that the newly-designed bioreactors were used. After 7 d of culture in the shake flask, the broth including cell debris was inoculated into the reactor by a syringe. During the continuous reactor operation the temperature was
~ G L A S S
MATERIALS AND METHODS Microorganism and media The microorganism used in this study was A. niger B60 obtained from Rohr and Kubicek of the Technische Universitat Wien (Austria), who used this strain for citric acid production (9-11). The composition of the growth and production media used in the shake-flask culture and reactor operation in g/l of deionized water as follows: sucrose 60, NH4NO3 2.5, KH2PO4 1, and MgSO4.7H20 0.25. The medium pH was adjusted to 3.1 before sterilization at 121°C for 15 min. Bioreactor design Figure 1 shows the conceptual design of the reactor. A bundle mixed of hydrophilic and hydrophobic hollow fiber membranes was inserted into
# #/~ ~
•
f,~O~
I
~-J,V. 0 i i ~ J .• V * ~ ,
, T..L
\,l,,I,
~
TUBE HYDROPHOBI MEMBRANE C MICROBES
T/..
/
~
./
u
H
~ I /
YDROPHILIC
MEMBRANE
u,o
NU,RIENT
GASEOUS NUTRIEN]
FIG. ]. Conceptualdesignof the aerobichollow fiberbioreactor.
* Corresponding author. 175
176
CHUNG AND CHANG
J. FERMENT. BIOENG.,
INOCULUM GASEOUS NUTRIENT (~
_j
silicone tubing
L__/ con.ector
_J I___
l
~ = ~ ~ ~
L I Q U I D NUTRIEI~ IN
IN
"---'~)'O TLIQUID UNUTRIENT
m
t
GASEOUS
s i l i c o n e rubber sealing
C
NUTRIENTI
OUT
20 cm
<
FIG. 2.
Detailed structure and dimension of the reactor.
maintained at 30°C, and the liquid nutrient and pure oxygen flowed at 2 ml/h and 100 ml/min, respectively. Analytical methods The citric acid concentration was assayed by the method of Marier and Boulet (12). RESULTS AND DISCUSSION
Figure 3 shows the citric acid concentration and pH in the effluent during the continuous operation of reactor A. During the first two days of operation, little citric acid was TABLE 1.
Detailed specifications of reactor A and B Reactor A
Reactor body Material of phil. No. of phil. Size of phil. Manufacturer of phil. Pore size of phil. Material of pho. No. of pho. Size of pho. Manufacturer of pho. Pore size of pho.
Reactor B
Glass (0.8 cm i.d.) Polyamide 16 0.06 cm i.d. 0.11 cm o.d. Berghof, FRG Mol. cut-off, 50,000 Polypropylene 32 0.033 cm i.d. 0.063 cm o.d. Enka, FRG 0.4-0.6/~m
Glass (0.8 cm i.d.) Polypropylene 40 0.033 cm i.d. 0.063 cm o.d. Enka, FRG 0.4-0.6/~m Teflon 5 O.l cm i.d. 0.2cm o.d. Sumitomo, Japan 1 pm
phil. and pho. denote the hydrophilic and hydrophobic hollow fiber membranes, respectively.
~3
2.5
~
2.0
2
1 0
/ n
0
produced, but decreasing effluent pH indicated cell growth in the reactor. The citric acid concentration increased dramatically from the 4th d to the 8th d while the effluent pH remained relatively constant. At the 8th d citric acid concentration reached about 3.5 g/l with a volumetric productivity of 0.7g//.h, based on the inner glass tube volume (10 ml). This productivity corresponds to a 12-fold increase over that (0.06 g/l. h) of the shake flask fermentation (7). However, after 8 d of operation citric acid production rapidly decreased and the citric acid concentration reached a constant level of about 1 g/l. After the reactor operation, the medium inlet part (C in Fig. 2) was sectioned with a razor blade and the cross section was photographed for more detailed observation (Fig. 4). The reactor was full of cells and, surprisingly, most of the polyamide hollow fiber membranes were compressed and blocked by excessive fungal cell growth. A similar phenomenon was also observed with fungal cultures of A. niger and Humicola sp. in the dual hollow fiber bioreactors (7, 8), where the flexible silicone tubes were expanded by excessive fungal cell growth, and it ultimately blocked the air flow path and restricted the substrate flow rate. It was believed that the rapid decrease of the citric acid production shown in Fig. 3 was caused by the nonuniform flow of liquid nutrient resulting from the blockage of polyamide hollow fiber membranes. Figure 5 shows the citric acid and pH in the effluent dur-
15
~b--/'--
I
I
I
I
I
2
~,
6
8
10
12
1.0 lZ,
Days FIG. 3. Citric acid concentration (©) and pH ( • ) in the effluent during reactor A operation.
FIG. 4. Photograph showing the cross section of reactor A after operation. PP, Polypropylene hollow fiber membrane; PA, polyamide hollow fiber membrane; AN, Aspergillus niger cells.
VOL 69, 1990
HOLLOW FIBER BIOREACTORS
3.5 3.0 ~3
2.5 -r
- 2.0
-~' -~._- 21
0
0
g.=ojl.5 n
2
t~
6
8 Days
10
12
FIG. 5. Citric acid concentration (©) and pH ( • ) in the effluent during reactor B operation (Q denotes the flow rate of liquid nutrient). ing the c o n t i n u o u s operation of reactor B. Profiles of citric acid concentration and pH were similar to those of Fig. 3. At the 7th d the m a x i m u m concentration of citric acid was produced at a concentration of 2 g / l , which was much lower than that of Fig. 3 (3.5 g//). One of reasons for the lower production of citric acid can be explained by the difference of gas-liquid interfacial areas in these two reactors which are equal to the outer surface areas of hydrophobic hollow fiber membranes used for aeration. The outer surface area (126.7 cm 2) of polypropylene membranes used in reactor A for aeration was larger than that (62.8 cm 2) of the Teflon membranes used in reactor B, and thus the oxygen transfer rate in reactor A should be larger than that in reactor B. In general, citric acid fermentation requires a high oxygen tension because oxygen is used directly for the product formation as well as the maintenance and growth of the cells. Thus, we concluded that the increased oxygen transfer rate resulted in a significant increase of the volumetric productivity. At the 14th d the flow rate of liquid nutrient suddenly dropped to zero, and thus further reactor operation had to be discontinued. It was believed that this was caused by the complete
177
blockage of the polypropylene hollow fiber membranes used for the liquid nutrient transport. Figure 6 shows the cross sectional view of reactor B, where complete blockage of the polypropylene membranes was not observed. But it would be certain that the polypropylene membranes in other regions of reactor B were blocked, which consequently restricted the liquid flow. These operational problems occurred due to the excessive fungal cell growth were also observed in the dual hollow fiber bioreactor operation for the purpose of producing citric acid by the same fungus, where we (7) solved these problems by repressing excessive cell growth using a nitrogen-deficient medium. In this study we could not try this experiment because blockage of the hollow fiber membranes happened suddenly. However, it would be possible to carry out a successful long-term operation if the cell growth can be appropriately controlled. In conclusion, aerobic fungal cell cultures using these hollow fiber bioreactor systems were possible, but longterm operation was not successful because of excessive fungal cell growth in the reactor that resulted in deformation of the polymeric hollow fiber membranes. Controlled cell growth in the reactor and use of stronger hollow fibers that can resist deformation would facilitate successful long term operation of these reactors. REFERENCES 1. Vick Roy, T.B., Blanch, H . W . , and Wilke, C.R.: Microbial
hollow fiber bioreactors. Trends in Biotechnol., 1, 135-139 (1983). 2. Inloes, D . S . , Smith, W . J . , Taylor, D . P . , Cohen, S . N . , Michaels, A. S., and Robertson, C. R.: Hollow-fiber membrane
3. 4. 5. 6.
7.
bioreactors using immobilized E. coli for protein synthesis. Biotechnol. Bioeng., 25, 2653-2681 (1983). Robertson, C.R. and Kim, I.H.: Dual aerobic hollow-fiber bioreactor for cultivation of Streptomyces aureofaciens. Biotechnol. Bioeng., 27, 1012-1020 (1985). Chang, H. N., Chung, B. H., and Kim, I. H.: Dual hollow-fiber bioreactor for aerobic whole cell immobilization. ACS (Am. Chem. Soc.) Symp. Ser., 314, 32-42 (1986). Chang, H. N., Kyung, Y. S., and Chuog, B. H.: Glucose oxidation in a dual hollow fiber bioreactor with a silicone tube oxygenator. Biotechnol. Bioeng., 29, 552-557 (1987). Chung, B. H., Chang, H. N., and Kim, I. H.: Rifamycin B production by Nocardia mediterranei immobilized in a dual hollow fiber bioreactor. Enzyme Microb. Technol., 9, 345-349 (1987). Chung, B. H. and Chang, H. N.: Aerobic fungal cell immobilization in a dual hollow fiber bioreactor: continuous production of citric acid. Biotechnol. Bioeng., 32, 205-212 (1988).
8. Hwang, Y. B., Chung, B . H . , Chang, H. N., and Han, M. H.:
Enzymatic conversion of rifamycin B by live Humicola sp. immobilized in a dual hollow-fiber bioreactor. Bioprocess Eng., 3, 159-163 (1988). 9. Rohr, M., Stadler, P . J . , Salzbrunn, W . O . J . , and Kubicek,
FIG. 6. Photograph showing the cross section of reactor B after operation. PP, polypropylene hollow fiber membrane; TF, Teflon hollow fiber membrane; AN, Aspergillus niger cells.
C. P.: An improved method for characterization of citrate production by conidia of Aspergillus niger. Biotechnol. Lett., 1, 281-286 (1980). 10. Rohr, M., Zehentgruber, O., and Kubicek, C.P.: Kinetics of biomass formation and citric acid production by Aspergillus niger on pilot plant scale. Biotechnol. Bioeng., 23, 2433-2445 (1981). 11. Misehak, H., Kubicek, C, P., and Rohr, M.: Citrate inhibitionof glucose uptake in Aspergillus niger. Biotechnol. Lett., 6, 425-430 (1984). 12. Marier, J. R. and Boulet, M. L.: Direct determination of citric acid in milk with an improved pyridine-acetic anhydride method. J, Dairy Sci., 41, 1683-1692 (1958).