Carboxymethyl cellulose protects algal cells against ... - CiteSeerX

Enzyme and Microbial Technology 29 (2001) 602– 610

www.elsevier.com/locate/enzmictec

Carboxymethyl cellulose protects algal cells against hydrodynamic stress F. Garcıá Camacho*, E. Molina Grima, A. Sańchez Miroń, V. Gonza´lez Pascual, Yusuf Chisti1 Department of Chemical Engineering, University of Almerıá, E-04071 Almerıá, Spain Received 13 February 2001; received in revised form 2 August 2001; accepted 21 August 2001

Abstract The harmful effect of direct air sparging on Phaeodactylum tricornutum microalgal cultures was investigated in bubble columns and airlift photobioreactors with various superficial air velocities and two types of spargers which generated different sizes of bubbles. Small bubbles bursting at the surface of the culture were apparently the main cause of cell damage in batch cultures in laboratory-scale bubble columns. Other mechanisms of cell damage also were a contributing factor to the observed cell loss in outdoor pilot-scale bubble columns. Supplementation of the microalgal culture medium with carboxymethyl cellulose at concentrations of 0.02% and greater is shown to protect the algal cells against aeration-induced hydrodynamic stress. © 2001 Elsevier Science Inc. All rights reserved. Keywords: Microalgae; Shear damage; Photobioreactors

1. Introduction

2. Materials and methods

Microalgae and photosynthetic bacteria are potential sources of novel high-value compounds [1]. Production of these compounds will require controlled culture of microalgae in photobioreactors. Bubble columns and airlift photobioreactors have proved useful for photosynthetic culture [2–5] but certain algae are damaged by hydrodynamic stress in these gas-agitated culture systems [2,3,6,7]. Compared to aeration, mechanical agitation can be even more detrimental to algal cells [7–10]. Shear-sensitivity of algal cultures can severely restrict attainable productivity in bioreactors. This work demonstrates that the detrimental effects of hydrodynamic forces on suspended algal cells can be reduced greatly by supplementing the culture medium with a modifier of interfacial properties such as carboxymethyl cellulose.

2.1. Organism and culture medium

* Corresponding author. Tel. & fax: ⫹1-34-9-50-2154-84. E-mail address: [email protected] (F. Garcia). 1 Y. Chisti is now with the Institute of Technology and Engineering, Massey University, Private Bag 11 222, Palmerston North, New Zealand.

The alga Phaeodactylum tricornutum UTEX 640 and the culture medium used have been described previously [4,5]. Carboxymethyl cellulose (CMC) sodium salt (Sigma C-5678, low viscosity) was used as a shear protective additive. 2.2. Analytical measurements The biomass concentration was estimated by optical density measurements (625 nm, 1 cm light path) in a Hitachi U-1000 spectrophotometer. The biomass concentration (C, g 䡠 L–1) and the culture absorbance (OD) were linearly related: C ⫽ 0.388⫻ OD625. The optical density measurements were periodically checked by gravimetry. Cell viability was determined by the trypan blue due exclusion method. The viscosity of algal suspensions was measured using a Cannon-Fenske viscometer. Surface tension was measured with a Kru¨ss tensiometer K10T (Kru¨ss, Hamburg, Germany) using the Wilhelmy plate method. 2.3. Static cultures Control experiments in static culture were done in unsparged shake flasks for verifying that CMC does not affect

0141-0229/01/$ – see front matter © 2001 Elsevier Science Inc. All rights reserved. PII: S 0 1 4 1 - 0 2 2 9 ( 0 1 ) 0 0 4 4 2 - 2

F.G. Camacho et al. / Enzyme and Microbial Technology 29 (2001) 602– 610

the cellular metabolism, as measured by the growth rate and cell viability. Low-speed agitation (200 rpm) of the flasks prevented the cells from settling. Different flasks (50 ml) in a given run contained either CMC (0.05%, 0.10%, or 0.20%) or no additive. Exponentially growing cells from 500 ml sparged flasks (no additive) were used for the inoculum Prior to inoculation, the flasks had been autoclaved (120°C, 1 h). The cooled flasks were inoculated and placed on an orbital shaker under fluorescent light (Phillips TLD W154) that provided a photosynthetically active photon flux density at the culture surface of 90 ␮E 䡠 m–2 s–1. Samples (0.5 ml) taken at various intervals were used to measure the biomass concentration and cell viability. Each experiment was conducted in triplicate and the results were analyzed by one-way ANOVA. 2.4. Assessment of the protective effect of CMC (laboratory cultures) Four identical 4 liter bubble columns were used as photobioreactors for algal culture. The columns were 0.495 m tall and 0.08 m in diameter. The working volume was 1.9 L. The columns were jacketed for temperature control. The culture pH could be controlled automatically by injecting CO2 as needed. Filter-sterilized compressed air was injected into each column for agitation. The gas flow could be controlled and the gas sparger was located 0.06 m above the bottom of the column. Three ring-shaped fluorescent lamps were placed around each column such that the column was concentric with the lamp. A mean scalar irradiance of 350 ␮ E 䡠 m–2 s–1 was measured (quantum scalar irradiance sensor QSL-100; Biospherical Instrument, San Diego, CA) at the center of the bioreactor in the absence of cells. The columns were sterilized before culture by washing with a 5% solution of sodium hypochlorite. 2.4.1. Establishment of hydrodynamic stress conditions Several authors have suggested that gas-liquid interfacial area in a gas-sparged reactor is somehow linked with the hydrodynamic damage to suspended cells [3,10, 11]. To investigate this issue, two kinds of gas spargers were used in the present work: (a) a perforated pipe sparger made of stainless steel and drilled with 24 holes of 1 mm diameter (sparger A); and (b) a glass sintered disc aerator of 60 ␮m pore size (sparger B). The spargers produced bubbles of quite different diameters. The cell damaging effect of the aeration rate was examined for superficial air velocities (UG) between 0.01 and 0.05 m䡠s–1. 2.4.2. Protective effect of CMC Once the superficial air velocity and the type of sparger for which cell-damage was observed had been identified, experiments were done using the CMC pro-

603

tective additive. These experiments consisted of four simultaneous cultures in the laboratory bubble columns. Two of the cultures were controls with and without hydrodynamic stress and the other two were carried out under hydrodynamic stress with various amounts of CMC added. The four reactors were inoculated simultaneously and aerated initially at a non-damaging low aeration velocity until the biomass concentration reached ⬃0.34 g 䡠 L–1 (⬃ 25 h of culture). The aeration rate was then increased to a value that caused cell damage. 2.5. Pilot-scale outdoor cultures Three outdoor placed gas-sparged photobioreactors were used to culture the microalga. The reactors included a bubble column, a split-cylinder airlift device, and a draft-tube sparged airlift bioreactor. All vessels were made of 3.3 mm thick transparent ploy(methyl methacrylate), except for the lower 0.25 m sections that were made of stainless steel and jacketed. The vessels were 0.193 m in internal diameter. The riser-to-downcomer cross sectional area ratio was unity for the split-cylinder and 1.24 for the draft-tube airlift vessel. The internal diameter of the draft tube was 0.144 m. The draft tube and the baffle were located 0.091 and 0.096 m from the bottoms of the reactors, respectively. The heights of the baffle and the draft tube were 1.85 and 1.88 m, respectively. The gas-free culture height was about 2 m in all cases, corresponding to a working volume of 60 L. All reactors were sparged through perforated pipe gas spargers (1 mm hole diameter). The aeration rate varied, as detailed later. The reactors were located in Almerıá (36° 50⬘ N, 2° 27⬘ W), Spain. The other geometric details of the reactors have been published [5]. The inoculum for the photobioreactors was grown indoors under artificial light (230 ␮ E 䡠 m–2 s–1 light flux at the vessel’s surface) in a 20 liter bubble column. The outdoor cultures in the three reactors were carried out simultaneously. The biomass concentration at inoculation was about 0.1 g 䡠 L–1. The cultures started in batch mode until a cell concentration of about 2.5 g 䡠 L–1 was attained. At this point, the operation switched to continuous culture at a constant dilution rate of 0.03 h–1 (a feed rate of 1.8 L 䡠 h–1 for the 10 h daylight period). In some cases, the culture medium was supplemented with CMC at concentrations of 0.02 or 0.04% by wt. The cultures were carried out during 29 March and 30 May, 2000. The daily average outdoor irradiance during the culture period varied between 690 and 1100 ␮ E 䡠 m–2 s–1. The culture temperature was controlled at 21°C by circulating chilled water through the jacket that surrounded the lower steel portion of the reactors. The pH was controlled at 7.7 by injecting carbon dioxide, as needed. The cell viability was measured daily at 13:30 h.

604


Fig. 1. Growth curves of P. tricornutum cultured in unsparged shake flasks. Each data point is the average of triplicate runs. Vertical bars show the standard deviation.

Fig. 2. Influence of the superficial air velocity (UG) on growth kinetics of P. tricornutum cultures in laboratory-scale bubble columns using sparger A. Each point is the average of duplicate runs. Vertical bars show the standard deviation. C is instantaneous biomass concentration and Co is the initial concentration of the biomass.

3. Results and discussion 3.1. Effect of CMC in static cultures The growth curves of P. tricornutum in shake flask cultures at low agitation intensity are shown in Fig. 1 for various concentrations of the CMC additive. Each data point in Fig. 1 is the average of three replicate cultures. The data show little scatter of statistical significance and there is no obvious impact of the CMC on growth profiles under low-intensity agitation. Therefore, CMC has no physiological effect on the cells. The cell viability was close to 100% in all cases in Fig. 1 and no cell damage was observed at the low-intensity agitation used. In the past CMC addition to culture media has been shown to protect the alga Dunaliella from hydrodynamic stress in Roux bottles and an externalloop airlift bioreactor. The mechanism of protection was said to be purely physical [2] but no control experiments were done to show that a physiological effect did not occur. 3.2. Establishment of hydrodynamics stress conditions in laboratory cultures Experiments were carried out in laboratory bubble columns with two different types of spargers to discern the harmful effects of cell– bubble interactions. The effect of gas flow rate on biomass growth in the bubble column with sparger A is shown in Fig. 2. Up to about 75 h of culture, the growth curves at the various aeration rates were virtually identical (Fig. 2) and the maximum specific growth rate values for the three cases were statistically indistinguishable. Obviously, therefore, the changed hydrodynamic conditions because of the changed aeration rate did not affect cell growth. The experiments at the different aeration rates were carried out at different times and possible differences in the physiological state of the inoculum affected the max-

imum biomass concentration attained (Fig. 2). In view of the results (Fig. 2), aeration rate values of up to 0.05 m 䡠 s–1 do not cause hydrodynamic damage to algal cells in laboratory bubble columns aerated with sparger A In the following set of experiments, algal cultures were carried out in the laboratory bubble column with sparger B. In view of the results discussed above, any culture at a superficial air velocity of between 0.01 and 0.05 m 䡠 s–1 and using sparger A could be used as control for the series of runs with sparger B. A culture at a moderate UG value of 0.02 m 䡠 s–1 (sparger A) was used as control. The cells were grown in two bubble columns in parallel. Initially, a low aeration rate of 0.005 m 䡠 s–1 was used in both columns to start the cultures in conditions free of hydrodynamic stress. At midexponential growth, the UG value in one of the columns was increased. The results are shown in Fig. 3. At an aeration velocity of 0.005 m 䡠 s–1, the culture profiles in the two columns were not affected by the step change in aeration rate to 0.02 m 䡠 s–1 (Fig. 3a). Similarly, the cell viability profiles were not affected by the change in aeration rate (Fig. 3a inset). Microscopic examination did not reveal any cell damage. In contrast, when the superficial air velocity was increased to 0.01 m 䡠 s–1 in the culture with sparger B (Fig. 3b), the cell viability and the biomass concentration began to decline within hours of the step change in the aeration rate. Sparger B generated visually smaller bubbles than sparger A, although the difference in size was not quantified exactly. Either the generation of small bubbles at the sparger was responsible for the damaging effect, or more likely the rupture of the smaller bubbles at the surface of the fluid was the source of cell damage. Other accumulated evidence proves that rupture of small bubbles is much more damaging to fragile cells than the rupture of larger bubbles [7,12]. The energy release associ-


605

Fig. 3. Set of cultures in the laboratory bubble columns using sparger B at superficial air velocity (UG) of: (a) 0.005 m 䡠 s–1; and (b) 0.01 m 䡠 s–1. The control culture was carried out at 0.02 m 䡠 s–1 aeration velocity and using sparger A. In (a) the step increase in air velocity was from 0.005 to 0.05 m 䡠 s–1 at the instance indicated; in (b) the step change in air flow was from 0.005 to 0.01 m 䡠 s–1. Square symbols in the inset figures refer to viability.

ated with bubble rupture is highly sensitive to bubble size in the 0.1 to 0.4 cm size range [7], as observed in this work. Higher aeration velocities (UG ⬎ 0.01 m 䡠 s–1) were tested but the kinetics of cell damage could not be measured because of the rapid rate of damage right from the instance of increasing the aeration rate. Because the bubble size from the sintered glass sparger was not affected by the increase in aeration rate, the increased rate of cell damage at the higher

aeration velocity is most likely explained by the increased frequency of bubble burst at the culture surface. Similar results concerning the effect of bubble size on cell damage have been reported in the past for animal cells [7,13–15] and for cultures of microalgae [10,11]. Rupture of small bubble has proved to be a more energetic process than the rupture of large bubbles [7] and this difference explains the greater damage seen with the smaller bubbles in this work.

606


size was always greater than the size of the cells, suggesting that the cause of cell damage is apparently not the turbulence in the fluid. 3.3. Protective effect of CMC in laboratory bubble columns

Fig. 4. The mean microeddy length scale versus superficial gas velocity in seawater in bubble columns.

The turbulence environment experienced by algal cells in the bulk medium can be potentially harmful [3,10]. Turbulent forces can be transmitted to cells by the fluid eddies when the eddy size (or length scale) is equal to, or smaller than the dimensions of the cells. P. tricornutum cells may be up to 35 ␮ m long and 3 ␮ m wide [16], however the maximum cell dimensions measured in our cultures were 12 ␮m length and 3 ␮m width (see electron micrographs in [17]). As shown in Fig. 4, the calculated microeddy size declined rapidly as the aeration velocity was increased. The mean length scale ␭ of the microeddies was calculated with the equation:

␭⫽

冉冊 ␮L ␳L

3/4

E ⫺1/4

(1)

where ␮L is the viscosity of the medium, ␳L is the density of the fluid, and E is the mean energy dissipation rate per unit mass. The latter depended on the aeration velocity, as follows: E ⫽ gu G

(2)

where g is the gravitational acceleration and UG is the superficial gas velocity based on the total cross sectional area of the reactor [7,12]. The measured viscosity, density, and interfacial tension of the culture media are noted in Table 1. The values in Fig. 4 indicate the calculated eddy

Table 1 Properties of media Fluid

Bulk viscosity (Pa 䡠 s)

Density (kg 䡠 m⫺3)

Surface tension (N 䡠 m⫺1)

Seawater Seawater ⫹ 0.05% CMC Seawater ⫹ 0.10% CMC Seawater ⫹ 0.20% CMC

1.2 ⫻ 10⫺3 2.98 ⫻ 10⫺3 3.13 ⫻ 10⫺3 4.02 ⫻ 10⫺3

1060 1062 1064 1065

72.5 ⫻ 10⫺3 62.7 ⫻ 10⫺3 57.2 ⫻ 10⫺3 51.9 ⫻ 10⫺3

The results of experiments carried out for assessing the protective effect of CMC in laboratory bubble columns are shown in Fig. 5. In all cultures, the cells assumed exponential growth immediately after inoculation (Fig. 5). The CMC containing cultures and the control were not affected by a step increase in the aeration rate (Fig. 5); however, the CMC-free culture (sparger B) showed a massive decline in cell viability and the biomass concentration. Because the cell viability and the biomass concentration values in cultures supplemented with 0.1% and 0.2% CMC were virtually identical, a CMC concentration of 0.1% provides essentially complete protection under the hydrodynamic conditions used. In the past, methyl cellulose additives have been associated with reduced attachment of cells to bubbles [18 –21]. These additives reduce dynamic interfacial tension (see Table 2) and this leads to a bubble surface that is less ridged and less capable of attaching and carrying cells to the surface. Consequently, the protective effect of CMC observed in this work may be linked to a fewer algal cells being transported to the surface where the damage associated with bubble rupture occurs. 3.4. Effects of hydrodynamic stress on the cellular morphology The polymorphic diatom P. tricornutum has three main morphotypes: oval, fusiform and triradiate [16]. Occasionally, the morphology of P. tricornutum can change suddenly and dramatically without an apparent environmental cause. However, in our cultures, only the fusiform morphotype existed. The fusiform morphology is the most common in laboratory and mass cultures [22]. Whereas the oval morphotype has a silica shell, the other morphotypes have largely organic cell walls, except for some siliceous bands in the girdle region [23]. Because it lacks a silica skeleton, the fusiform morphology is more susceptible to hydrodynamic stress than some of the other morphotypes. The observed response of cells to hydrodynamic stress was a diminished length and width, and significant disruption. The reduced size seemed to be a cellular adaptation to mitigate the damaging effect, as a small cell can better withstand turbulence than a large cell for otherwise fixed conditions. Changes in morphology of microorganisms in response to hydrodynamic stress have been observed in many cases. In most cases such changes are the result of physical disruption


607

Fig. 5. Growth curves and cell viability of cultures grown in laboratory bubble columns under conditions of hydrodynamic stress with and without CMC. The control culture (sparger A) was carried out at 0.02 m 䡠 s–1 aeration velocity. In cultures with sparger B, the gas flow rate shifted from 0.005 m 䡠 s–1 to 0.05 m 䡠 s–1, at the instance noted.

of the chains of cells [7,24]. For example, Ma¨ rkl et al. [25] showed that hydrodynamic stress reduced the size of the cyanobacterial trichomes of Spirulina platensis. However, in other cases, the cells respond and adapt, to some extent, to their environment. For example, hybridoma cells cultured with low fluid stress levels were more sensitive to shear than cells from rapidly agitated cultures [26].

3.5. CMC in pilot-scale outdoor cultures The steady-state biomass concentration evolution in the three outdoor reactors is shown in Fig. 6. In region a (Fig. 6) the superficial gas velocity UG was 0.01 m 䡠 s–1, there was no CMC in the culture broth, and all reactors attained a stable biomass concentration of ⬃ 1 g 䡠 L–1. At the instance

608


Fig. 6. Steady-state biomass concentration (except Cb2) versus culture time in outdoor cultures. (a) UG of 0.01 m 䡠 s–1 and no carboxymethyl cellulose (CMC); (b) UG increased to 0.02 m 䡠 s–1, no CMC; (c) UG ⫽ 0.02 m 䡠 s–1, CMC added at 0.02%; (d) UG raised to 0.04 m 䡠 s–1, and CMC added at 0.04%.

of the arrow A, the aeration rate was increased to 0.02 m 䡠 s–1 (Fig. 6). The biomass concentration began to decline because of destruction of the cells. By day 49, the mean biomass concentration had dropped to ⬍ 0.7 g 䡠 L–1 in all reactors and no steady-state was attained. If the culture had been allowed to remain at these conditions, the reactors would have washed out. The decline in cell concentration was associated with the damaging effect of a high aeration velocity. Thus, in the same continuous culture and at the instance of the arrow B, the feed medium was supplemented with 0.02% CMC but the gas velocity remained at 0.02 m 䡠 s–1 (Fig. 6). All three reactors attained a new steady-state at ⬃ 64 days and the stable biomass concentration once again exceeded 1 g 䡠 L–1 in all cases (Fig. 6). The increase

in the steady-state biomass concentration in moving from region b to region c was because of a cell protective effect of the CMC additive. At the instance marked by arrow C, the aeration velocity was increased to 0.04 m 䡠 s–1 but the concentration of CMC in the medium was also increased (Fig. 6). Despite a high aeration rate, the cells could be stably cultured at a concentration of ⬎ 1 g 䡠 L–1 because of the protection afforded by the CMC. At the first steady state (Fig. 6, region a), the cell viability was 100% in all three reactors. In region b (Fig. 6), the cell viability was between 40 and 55% in all cases. In region c, the viability at steady state was about 94%. In region d, the cell viability in all reactors was 95% or greater. The increase in aeration velocity from 0.01 to 0.04 m 䡠 s–1 had no effect on the measured mixing time and the liquid circulation velocity in any given reactor, hence the observed changes in cell survival could not be ascribed to these factors. Similarly, gas–liquid mass transfer (oxygen removal, carbon dioxide supply) did not limit over the entire range of the gas velocities tested, as previously proved [5]. Because, experiments in static flasks with and without CMC supplementation revealed no effect of CMC on cell growth, the behavior in Fig. 6 cannot be associated to a possible metabolic effect of CMC. Also, P. tricornutum is not known to produce cellulose or to possess cellulose degrading enzymes. Notably, the CMC concentration needed to produce a protective effect in outdoor cultures was lower than the minimum tested in the shake flask control experiments (0.05% CMC). Papoutsakis also reported that low concentrations of CMC in animal cell cultures were as effective as higher viscosity grades of CMC in protecting the cells [27]. Thus, it is difficult to conclude that the effect of CMC is related to the turbulence dampening associated with an enhanced viscosity. However, changes in viscosity and interfacial tension can dramatically alter the bubble size and the dynamics of bubble rupture at the surface of the culture. 3.6. Cell-damage mechanism and protective effect of CMC in outdoor bubble columns In contrast to the results obtained in the laboratory bubble columns, the bubbles generated with stainless-steel sparger (1 mm hole diameter) did damage the cells in outdoor pilot-scale bubble columns. In the past, several authors have reported that damage to animal cells is reduced by increasing the aspect ratio of bubble columns [7,14]. The aspect ratio of our outdoor columns was ⬃ 10 whereas the aspect ratio of indoor columns was ⬃ 6. Thus, the expected effect of the column geometry was not confirmed in our work with algal cells. Possibly, different mechanisms of cell damage also contributed to the observed damage in the larger columns. Turbulence per se as the cause of cell damage could be disregarded because the length scale of microeddies in different zones of photobioreactors was always greater than the


dimensions of the cells. The microeddy length scale was estimated using approaches described by others [28–30]. The energy dissipation at the gas injection holes in the larger columns and bubble breakup/coalescence inside the culture were generally greater than in the smaller columns for any given superficial aeration velocity. This could have contributed to the observed higher cell damage in the larger columns. The differing results obtained for small and large bubble columns using similar spargers demonstrate the well-known complexity of scale-up of bubble columns. Models such as that of Meier et al. [31] have attempted to unify the sometimes contradictory observations relating to sparging-associated cell damage, but those models apply only to noninteracting bubbles (i.e. those that do not break-up or coalesce) and not to the kind of systems used in this study.

[5]

[6]

[7]

[8]

[9]

[10]

4. Conclusions This work confirms that the microalga P. tricornutum is sensitive to hydrodynamic stress. Apparently, the break-up of small bubbles on the surface of the medium is the cause of cell damage. Cell damage can be suppressed by adding a small amount (ⱖ0.02% by wt) of CMC to the culture medium. The CMC additive has a purely physical protective effect and no physiological effect on the algal cells. The hydrodynamic conditions that did not promote cell damage in the laboratory-scale bubble columns could not be extrapolated to the pilot-scale bioreactors. In view of this, more work is needed to assess the impact of factors such as the bubble column diameter and height in producing a cell damaging environment. Reliable scale-up of bubble column and airlift photobioreactors is difficult unless the effects of reactor aspect ratio (height-to-diameter ratio) on cell damage can be clarified. Acknowledgments This research was supported by the Comision Interministerial de Ciencia y Tecnologıá (CICYT) (BIO98 – 0522), Spain. Cristobal Sa´ nchez Martıń is thanked for technical assistance.

[11] [12] [13]

[14]

[15]

[16] [17]

[18]

[19]

[20]

[21]

References [22] [1] Borowitzka MA. Pharmaceuticals and agrochemicals from microalgae. In: Chemicals from microalgae. London: Taylor & Francis, 1999. p. 313–52. [2] Silva HJ, Cortin˜ as T, Ertola RJ. Effect of hydrodynamic stress on Dunaliella growth. J Chem Technol Biotechnol 1987;40:41–9. [3] Contreras Go´ mez A, Garcıá Camacho F, Molina Grima E, Merchuk JC. Interaction between CO2-mass transfer, light availability, and hydrodynamic stress in the growth of Phaeodactylum tricornutum in a concentric tube airlift photobioreactor. Biotechnol Bioeng 1998;60: 317–25. [4] Sa´ nchez Miro´ n A, Contreras Go´ mez A, Garcıá Camacho F, Molina Grima E, Chisti Y. Comparative evaluation of compact photobiore-

[23]

[24]

[25] [26]

609

actors for large-scale monoculture of microalgae. J Biotechnol 1999; 70:249 –70. Sa´ nchez Miro´ n A, Garcıá Camacho F, Contreras Go´ mez A, Molina Grima E, Chisti Y. Bubble column and airlift photobioreactors for algal culture. AIChE J 2000; 46:1872– 87. Suzuki T, Matsuo T, Ohtaguchi K, Koide K. Gas–sparged bioreactors for CO2 fixation by Dunaliella tertiolecta. J Chem Technol Biotechnol 1995;62:351– 8. Chisti Y. Shear sensitivity. In: Encyclopedia of bioprocess technology: fermentation, biocatalysis, and bioseparation, vol. 5. New York: Wiley, 1999. p. 2379 – 406. Jaouen P, Vandanjon L, Quemeneur F. Stress of microalgal cell suspensions (Tetraselmis suecia) in tangential flow filtration systems: The role of pumps. Bioresource Technol 1999;68:149 –54. Vandanjon L, Rossignol N, Jaouen P, Roberts JM, Que´ me´ neur F. Effects of shear on two microalgae species. Contribution of pumps and valves in tangential flow filtration systems. Biotechnol Bioeng 1999;63:1–9. Garcıá Camacho F, Contreras Go´ mez A, Mazzuca Sobezuk T, Molina Grima E. Effects of mechanical and hydrodynamic stress in agitated, sparged cultures of Porphyridium cruentum. Process Biochem 2000; 35:1045–50. Yang JD, Wang NS. Cell inactivation in the presence of sparging and mechanical agitation. Biotechnol Bioeng 1992;40:41–9. Chisti Y. Animal-cell damage in sparged bioreactors. Trends Biotechnol 2000;18:420 –32. Handa-Corrigan A, Emery AN, Spier RE. On the evaluation of gas–liquid interfacial effects on hybridoma viability in bubble column bioreactors. Develop Biol Standard 1987;66:241–53. Handa-Corrigan A, Emery AN, Spier RE. Effect of gas–liquid interfaces on the growth of suspended mammalian cells: mechanisms of the cell damage by bubbles. Enzyme Microb Technol 1989;11:230–36. Wu J, Goosen MFA. Evaluation of the killing volume of gas bubbles in sparged animal cell culture bioreactors. Enzyme Microb Technol 1995;17:1036 – 42. Lewin JC, Lewin RA, Philpott DE. Observations on Phaeodactylum tricornutum. J Gen Microbiol 1958;18:418 –26. Molina Grima E, Garcıá Camacho F, Acie´ n Ferna´ ndez FG. Production of EPA from Phaeodactylum tricornutum. In: Chemicals from microalgae, London: Taylor & Francis, 1999. p. 57–92. Michaels JD, Nowak JE, Malik AK, Koczo K, Wasan DT, Papoutsakis ET. Analysis of cell-to-bubble attachment in sparged bioreactors in the presence of cell-protecting additives. Biotechnol Bioeng 1995;47:407–19. Michaels JD, Nowak JE, Malik AK, Koczo K, Wasan DT, Papoutsakis ET. Analysis of cell-to-bubble attachment in sparged bioreactors in the presence of cell-protecting additives. Biotechnol Bioeng 1995;47:420 –30. Chattopadhyay D, Rathman JF, Chalmers JJ. The protective effect of specific medium additives with respect to bubble rupture. Biotechnol Bioeng 1995;45:473– 80. Chattopadhyay D, Rathman JF, Chalmers JJ. Thermodynamic approach to explain cell adhesion to air–medium interfaces. Biotechnol Bioeng 1995;48:649 –58. Borowitzka MA, Volcani BE. The polymorphic diatom Phaeodactylum tricornutum: ultrastructure of its morphotypes. J Phycol 1978;14: 10 –21. Borowitzka MA, Chiappino ML, Volcani BE. Ultrastructure of a chain-forming diatom Phaeodactylum tricornutum. J Phycol 1977;13: 162–70. Thomas CR. Problems of shear in biotechnology. In: Chemical engineering problems in biotechnology. critical reports on applied chemistry, vol. 29. London: Elsevier, 1990. p. 23–93. Ma¨ rkl H, Bronnenmeier R, Wittek B. The resistance of microorganisms to hydrodynamic stress. Int Chem Eng 1991;31:185–97. Papoutsakis ET. Fluid-mechanical damage of animal cells in bioreactors. Trends Biotechnol 1991;9:427–37.

610


[27] Papoutsakis ET. Media additives for protecting freely suspended animal cells against agitation and aeration damage. Trends Biotechnol 1991;9:316 –25. [28] Merchuk JC, Berzin I. Distribution of energy dissipation in airlift reactors. Chem Eng Sci 1995;50:2225–33. [29] Contreras A, Garcıá F, Molina E, Merchuk JC. Influence of sparger on energy dissipation, shear rate, and mass transfer to sea water in a

concentric-tube airlift bioreactor. Enzyme Microb Technol 1999;25: 820 –30. [30] Molina E, Chisti Y, Moo-Young M. Characterization of shear rates in airlift bioreactors for animal cell culture. J Biotechnol 1997;54:195–210. [31] Meier SJ, Hatton TA, Wang DIC. Cell death from bursting bubbles: role of cell attachment to rising bubbles in sparged reactors. Biotechnol Bioeng 1999;62:468 –78.