You are on page 1of 6

Pressurized bioreactor

Marlene Lopes
Manuel Mota

Research Article

Isabel Belo

Oxygen Mass Transfer Rate in a Pressurized
Lab-Scale Stirred Bioreactor

University of Minho, Centre
of Biological Engineering,
IBB – Institute for
Biotechnology and
Bioengineering, Braga,


Oxygen mass transfer from air to the liquid phase in bioreactors with aerobic
cultures has long been a serious impairment to the productivity of various bioprocesses. An increase of the oxygen mass transfer rate (OTR) can be the key to overcome oxygen limitation. The influence of higher air pressure on OTR was measured and a significantly enhanced OTR could be obtained. The oxygen volumetric
mass transfer coefficient (kLa) was described by a function of the air pressure in a
stirred lab-scale pressurized bioreactor. The correlation obtained proved that kLa
slightly decreased with higher air pressure, following a power function.
Keywords: Air pressure, Oxygen mass transfer, Oxygen transfer coefficient,
Pressurized bioreactor
Received: January 31, 2013; revised: March 22, 2013; accepted: July 10, 2013
DOI: 10.1002/ceat.201300082



In aerobic bioprocesses, oxygen is a key substrate. Due to its
low solubility in aqueous solutions, it is important to ensure
an adequate delivery of oxygen from a gas stream to the culture broths. The oxygen mass transfer rate (OTR) and volumetric oxygen mass transfer coefficient (kLa) must be known,
and if possible predicted to achieve an optimum design operation and scale-up of bioreactors. OTR and kLa are influenced
by a numerous parameters like physical properties of gas and
liquid, operating conditions, and geometrical parameters of
the bioreactor, and also by the presence of biomass, i.e., the
consumption of oxygen by the cells [1, 2]. Both parameters
can be related and stated mathematically as:
OTR ˆ kL a…C*


where C* is the solubility of oxygen in the liquid, C is the dissolved oxygen concentration in the liquid, and kLa is made up
of the mass transfer coefficient (kL) and the interfacial area (a).
In this equation, the term …C* C † is considered to be the
driving force which causes oxygen to transfer from the gas
phase to the liquid phase [3].
Several empirical correlations have been proposed to estimate the kLa in mechanically agitated bioreactors (STR), with
the most well-known function: 
…vs †c
kL a ˆ a

Correspondence: Dr. I. Belo (, University of
Minho, Centre of Biological Engineering, IBB – Institute for Biotechnology and Bioengineering, Campus de Gualtar, 4710-057 Braga, Portugal.

Chem. Eng. Technol. 2013, 36, No. 10, 1779–1784

where Pg represents the power input to the aerated bioreactor, V
is the bioreactor working volume, and vs is the superficial gas
velocity. The parameters a, d, and c are dimensionless constants.
To calculate the power input to the aerated system (Pg), the
Reynolds number (Re) is determined by Eq. (3) and the power
number (Np) by Eq. (4).
Re ˆ

D2i Nq


Np ˆ

qN 3 D5i


where q represents the liquid density, N the agitation rate, v
the liquid viscosity, and Di the impeller diameter.
According to Cheremisinoff and Gupta [4], if the flow regime
inside the system is turbulent (19 070 < Re < 38 141), Np is not a
function of Re when the vessel is fully baffled. Consequently, Pg
without aeration (Pg0 ) can be determined by Eq. (5).
Pg0 ˆ KT D5i N 3 q


where KT is a constant dependent on the impeller used.
Finally, to determine Pg in an aerated system, Eq. (6) can be
Pg0 ND3i
Pg ˆ c


where c is a constant dependent on the impeller and Fg is the
volumetric gas flow rate.
In order to overcome the oxygen limitation in aerobic
microbial cultures, selection of adequate, normally high OTR

© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

. 10. (8): OTR ˆ kL aC* …8† Some authors have demonstrated the applicability of pressurized bioreactors in microbial cultures. According to the reaction stoichiometry (Eq. 1779–1784 . a temperature probe. Thus. Special aeration systems. (2). Va – check valve. aeration using oxygen-enriched air and increased reactor pressure are techniques applied to increase oxygen availability [5–8]. the volumetric oxygen mass transfer can be determined by Eq. F – air filter.063 m diameter (total area 0. The gas flow rate was measured with a calibrated mass flow controller (Alicat scientific. Schematic diagram of the stainless steel stirred-tank bioreactor (PARR 4563. Also. e. At regular times.1780 I. 2 Materials and Methods 2. the oxidation rate is controlled by the rate of mass transfer and measures the overall rate. and the stirring rates. the OTR is estimated as half of the rate of sulfite consumption. PARR Instruments. the use of in situ production of oxygen [9] or the use of a second liquid phase of various organic compounds such as perfluorodecalin [10] or n-hexadecane [11] in the culture medium can increase the availability of oxygen to the microorganisms.00312 m2 and the ratio H/D is 3) equipped with www. with the dissolved oxygen to produce sodium sulfate in the presence of a catalyst (usually a divalent cation of Cu2+ or Co2+). The operating pressure was set by manipulation of the pressure of the inlet compressed air and the regulatory valve position in the exit gas an impeller with two turbines of four pitched blades (0. V3. 2013.g.04–1 N) for which the oxygen concentration can be assumed as zero.cet-journal. Technol. and superficial gas velocity was attempted. Parr Instruments.2 mol L–1) is oxidized to sodium sulfate in the presence of a catalyst (CuCl2. No. The parameters studied were stirring rate (200. 1. it seems important to describe OTR and kLa in such bioreactors. power input of the aerated bioreactor. This technique is based on the reaction of sodium sulfite. 2 vvm. experimental values of OTR were obtained in a labscale pressurized bioreactor. V4 – valves. values is crucial. 600 rpm). MFC – mass flow controller. Thus. The techniques vary according to the accuracy required and have advantages and disadvantages depending on the availability of the necessary analytical instruments and material and labor costs [12]. The amount of residual sulfite can be estimated by:  2  0:05VI SO3 ˆ © 2013 WILEY-VCH Verlag GmbH & Co. Weinheim …VI ‡ Vs †I2 Vs …9† Chem. Therefore. Based on Eq. measured under standard temperature and pressure conditions). USA) to monitor the total internal pressure. the OTR can be estimated by the oxygen absorption rate of a sodium sulfite solution [13]. Since these improvements could be related to an improvement in oxygen mass transfer due to the increase of oxygen solubility with pressure. T – pressure transducer. CW – cooling water.1 Experimental Procedure A 600-mL stainless steel stirred-tank bioreactor (PARR 4563. knowing OTR and oxygen solubility. The amount of iodine that not reacted with sulfite ion was determined by spectrophotometric absorbance at 595 nm and converted to molar concentration using a previous calibration. 400. V5 – regulatory valve. samples of known volumes were collected and mixed with an excess of iodine solution (0. and total air pressure inside the bioreactor (1–5 bar). The reaction rate is much faster than the oxygen transfer rate and there is a concentration range of sodium sulfite (0.001 mol L–1). 36. M – motor. A number of methods have been developed to determine the oxygen transfer rate in bioreactors.2 Oxygen Transfer Rate (OTR) OTR in bioreactors operating under different conditions was estimated in blank assays by the sulfite oxidation method [13] at 30 °C.035 m of diameter). data fitting to an empirical correlation for the prediction of the kLa as a function of air pressure. KGaA. aeration rate (0. Belo et al. 0. V1. V2.05 mol L–1). USA) with 400 mL of operating volume was used (Fig. In the absence of microbial cells. 1). Model MC-5SLPM-D).5. 2. and a sparger tube for aeration. A sodium sulfite solution (0. The reactor was equipped with a pressure transducer (PARR 4842. The bioreactor is a cylinder of 0. The reaction can be expressed as: 1 Na2 SO3 ‡ O2 2 Cu =Co 2‡ 2‡ ! Na2 SO4 …7† Compressed air MFC CW Temperature probe Heating Sparger Control device Figure 1. Parr Instruments. with enhancements in biomass and product yields [14–16]. D – security disk. Eng. by varying the air pressure. (7)). a reducing agent. USA): A – agitator. the aeration.

cet-journal. In order to predict the effect of air pressure on the oxygen solubility. the a and consequently kLa values [18]. at 0. and (C) 2 vvm. The dimensionless parameters a. This effect was attributed to the rapid breakage of the gas bubbles into smaller sizes with higher impeller speed. it enables to predict the effect of pressure increase on the oxygen mass transfer capacity of the system. several experiments were carried out. No. the Krichevsky-Kasarnovsky equation [17] was applied. Separate analyses of each factor indicated that the increment of aeration and stirring rates and total air pressure inside the bioreactor led to an enhancement of the OTR value. a 4. 1779–1784 B OTR (mg O2/L.. b. the increase in agitation proved to be more efficient for OTR enhancement than the increase in aeration. (8). Thus. and consequently enhancement in the gas-liquid interfacial area for mass transfer [1].h) Pressurized bioreactor 4000 3000 2000 1000 0 A correction of Eq. As this method measures the rate of O2 absorption by a Na2SO3 solution. agreement with the results of Amaral et al. and I2 is the molar concentration of the iodine solution. 2. the OTR value also increases. d. At 1 bar of air pressure.5 vvm.where Vs is the sample volume. Juárez and Orejas [21]. using the Solver tool of Microsoft Excel 2007 software. The air flow rate inside the reactor was corrected from the measured one under standard conditions to the values of temperature and pressure of the assay using the ideal state gas equation.h) 4000 200 400 600 Srring rate (rpm) 3000 2000 1000 0 1 2 3 P (bar) C 5000 4000 4 5 200 5 200 400 600 Srring rate (rpm) 3000 2000 1000 0 To evaluate the effect of total air pressure on OTR values. [22] who showed that kLa depends more strongly on agitation than on aeration rates for STR. Impeller speed is the major factor that affects kLa values of a stirred-tank bioreactor as it determines the overall power dissipation for any specific impeller design. This lower improvement in OTR with higher aeration rate as compared to the agitation rate was similar for all values of pressure. (10) were calculated with the help of the equations presented in the introduction. 3 1781 1 2 3 P (bar) 4 400 600 Srring rate (rpm) Figure 2. (2) was made in order to predict the effect of absolute pressure. Weinheim www. VI is the iodine solution volume. The change in gas flow rate affects the fractional gas holdup. For all the conditions studied. 2). changing stirring and aeration rates under increased air pressure up to 5 bar (Fig. and hence. Eng. in kLa: 1  d Pg kL a ˆ a …v s †c P b V P (bar) 3. (B) 1 .3-fold enhancement was achieved with the increase of air flow rate up to 2 vvm. The OTR improvement was more pronounced with the increment of the stirring rate or the rise of air pressure than with the increase of aeration rate. Gomes et al. 10. The result proved that Henry’s law was still valid for the air pressure range used in this work (up to 5 bar). 36. and c were estimated by minimizing the sum of least squares of the difference between the experimental and modeled value of kLa. 2013. varying the stirring rate from 200 to 600 rpm. the OTR value augments by ten times. leading to a higher surface area of bubbles which in turn increases the kLa values. [10]. Pg and vs in Eq.h) Results and Discussion 2 …10† kLa values were obtained by dividing the experimental OTR data by the oxygen solubility at 30 °C. At higher gas flow rates.e. KGaA. This behavior is in Chem. Gómez-Díaz and Navaza [20]. [19]. according to Eq.1 OTR Measurement 3 4 5 5000 OTR (mg O2/L. i. P. Experimental OTR values under various experimental conditions and aeration rates of (A) 0. Technol. converting the aeration rate to real Fg. the gas holdup in the bioreactor increases. In turn. Raising the stirring rate from 200 to 600 rpm led to an improvement of the OTR value more significantly at 1 bar air pressure and © 2013 WILEY-VCH Verlag GmbH & Co.5 vvm of aeration rate.3 kLa Modeling A 5000 OTR (mg O2/L. and Chen et al.

3-fold) was achieved with the rise of air pressure from 1 to 5 bar and with the increase of the stirring rate from 200 to 600 rpm. (11) with estimated parameters for increased air pressure up to 5 bar.70.6 and for the specific power consumption the coefficient is a function of QG. the increased air pressure from 1 to 5 bar led to improved OTR values. to the power V of 0. Belo et al. and aeration rates from 0.48.1-fold improvement of the OTR value at 0. [6] found similar values for coefficients of Predicted kLa (h-1) 0. and decreases with the air pressure.and 3-fold improvements were found at 1 bar and 2 vvm and with 5 bar and 2 vvm. Maier et al. of 0. www. This effect was even more pronounced at lower stirring and aeration rates. Yang and Wang [24] observed that the bioreactor pressurization up to 2. . No.5 vvm. This effect can be attributed to the use of a constant gas flow rate (measured under standard conditions) which in fact led to a decrease of the true gas flow rate inside the bioreactor with pressure. once the coefficient of vs was lower than the   Pg . This result proves that the increased air pressure is an alternative to the stirring rate increase which is particularly important for high-cell-density cultures and when the cells are sensitive to shear stress that limits the increase of stirring. the total surface area for mass transfer decreases with pressure. it is costly and requires special handling [24]. This causes a reduction in the number of air bubbles and consequently in the gas holdup [27]. kLapredicted = Knoll et al. 3. A   Pg V and vs. as demonstrated by the coefficient of P.5 vvm and 200 rpm was achieved when air pressure varied from 1 to 5 bar. predicted versus experimental kLa values are plotted with a deviation of 4 % of a unitary slope. and cost efficiency of oxygen transfer can be greatly enhanced by employing elevated reactor pressures of up to 11 bar. (10) were estimated as indicated in Eq. Weinheim Chem. the values of a. © 2013 WILEY-VCH Verlag GmbH & Co. Yang and Wang [24] found a 2. [6] reported that the oxygen transfer capacity. respectively. In order to predict bioreactor performance when using models that account for the effect of the increased air pressure. Technol. (11): kL a ˆ 535  0:70 Pg …vs †0:48 P V 0:13 200 150 100 50 …11† From Eq. in spite of the specific interfacial area increasing with pressure due to the air bubbles’ compression. at 1 bar and 0.5 vvm aeration.1782 I. The rise of total air pressure had a small coefficient of V negative effect on kLa. Using the experimental data obtained in experiments with increased air pressure up to 5 bar.2 kLa Modeling The determination of kLa in bioreactors is essential to establish aeration efficiency and to quantify the effects of the operating variables on the provision of oxygen. and the observed pressure effects have to be attributed to changes in the interfacial area a [28]. Belo and Mota [23] observed that Escherichia coli TB1 cells proved to be more sensitive to high shear stress caused by stirring than to air total pressures up to 4 bar. d. Knoll et al. (11) indicate that the kLa dependence was higher on the specific power input than the superficial gas velocity.42. This means that the increase of air pressure slightly decreases the volumetric mass transfer coefficient. energy efficiency. to the power of 0.72 bar had little effect on kLa. 2013. the same improvement in oxygen mass transfer (4. The global effect of pressure in kLa obtained in this work was more influenced by the negative effect on gas holdup than by the positive effect of the air bubble compression. 10.5 to 2 vvm.13). A 10-fold improvement in OTR was observed by changing the stirring rate from 200 to 600 rpm. the authors did not take into account the air pressure in the kLa mathematical correlation. The oxygen transfer rate raise promoted by the increased air pressure inside the bioreactor is based on the fact that the equilibrium oxygen solubility in the nutrient broth increases linearly with the total air pressure according to Henry’s law.g. KGaA. 1779–1784 . The values of coefficients in Eq. P. 0 0 50 100 150 Experimental kLa (h-1) 200 Figure 3. b. [26] also reported that raising oxygen solubility through the increase in total air pressure enhanced OTR in the pressurized bioreactor and decreased kLa. an empirical correlation (Eq. vs..06 to 2. In Fig. Zhang et al. In the experiment conducted at 2 vvm aeration rate. Eng.5 to 2 vvm. stirring rates from 200 to 600 rpm. e. whereas at 2 vvm and 600 rpm the OTR at 1 bar was four times higher than at 5 bar.74 and 0. and c coefficients from Eq. Independently of the stirring and aeration rates tested. to the power of (–0. 3. Moreover. Thus. However. aeration rates ranging from 0. [25] proposed a correlation to predict the volumetric mass transfer coefficient in a rotating-drum bioreactor. only 4. [29] reported that the kLa values in a stirred-tank reactor remain constant irrespective of the reactor pressure if the superficial gas velocity is kept constant.72 bar. (10)) for the kLa in a pressurized bioreactor was proposed.5-fold increase in OTR by applying air pressures from 1. It seems that the stirring rate effect is more significant at low kLa values. and the coefficient of volumetric aeration rate (QG) was 0. to the superficial gas velocity. Belo et al. respectively.cet-journal. Comparatively. Although oxygen-enriched air can achieve the same result. Correlation between the experimental and predicted kLa values using Eq. 36. and stirring rates from 200 to 600 rpm. (11) it can be concluded that the kLa in measured  Pg creases according to the specific power input. it is accepted that the liquid mass transfer coefficient kL is independent of the headspace pressure P. Other authors have reported the enhancement in OTR values due to the increased air pressure inside the bioreactor.

Herein. E. Eng. I. M. 2008. Coutinho. Ann Arbor Science. Lopes. J. Lett. The kLa increase turned out to be higher with the rise of specific power input than with the superficial gas velocity whereas a higher total air pressure had only a small negative effect on kLa. Buchs. M. Büchs. I. Seletzky. 2nd ed. 2009. M. N. Tscherrig. V. 63. 2003. 77. Biotechnol. Suresh. 1068. J. [8] I. Biochem. J. Cooper. M. Maie. S. Sinclair. 36. Sonnleitner. Appl. P. I. Ribeiro. J. 588. A. . 2003. 17. R. Biotechnol. The proposed correlation for kLa prediction could be valuable for further work on the development of strategies for optimization and scale-up of processes where oxygen transfer is a limiting factor. Microbiol. J. Nielsen. Biotechnol. Srivastava.. M. Belo. Azevedo. A. Chem. Teixeira. Bioeng. [10] P. Biotechnol. Technol. K. The authors have declared no conflict of interest. [16] R. N. Rocha-Leão. 26 (10). Biochem. B. c [–] m vs q [kg m–1s–1] [m s–1] [kg m–3] non-linear models parameters to be fitted liquid viscosity superficial gas velocity liquid density References [1] S. M. [12] M. J. Klekner. Gómez-Díaz. 1994. the effects of increased air pressure on the oxygen transfer rate were investigated. M. Bartsch. Gomes. 65. Cheremisinoff. F. Schroer. 1944. 243. C. Eng. Gonçalves. V. Belo. Maier. M. 1984. 255.96 kLaexperimental which indicates a good approximation between real kLa values and the values calculated by the correlation. R. [6] A.911). Stöckmann. I. Vidal-Tato. Biotechnol. Adv. 504. I. Belo. 2009. C. 735. 34 (11). 84. Ind. Gómez-Díaz. Molecular Thermodynamics of Fluid-Phase Equilibria. B. 83. 1790. b. R. P. 159. R. J.Pressurized bioreactor 0. Technol. J. 10. Chem. I. J. Freire. this pressure can serve as an alternative to avoid shear stress caused by the increased stirring rates that might be harmful to cells. Adv. 99 (3). G. 27. Gupta. A. U. S. Mota. 132. Husemann. Coelho. Eng. Ann Arbor. 438. Buchs. Appl.. despite the dispersion of the values (R2 = 0. Gupta). Eng. optimization of the bioreactor performance concerning the oxygen mass transfer requirement is a crucial task in industrial bioprocesses. [7] M. Mota. A. Garcia-Ochoa. Technol. Gomez. Technol. [13] C. 2004. Belo. A. Belo. McLellan. Chem. M. Weinheim www. M. Coelho. Microbiol. [20] D. A. G. [3] C. NJ 1986. 46. [9] B. 153. Losen. R. and superficial gas velocity was established. Biochem. J. Thus. [2] F. Acknowledgment The authors acknowledge the financial support provided by the Fundação para a Ciência e Tecnologia (Grant SFRH/BD/ 47371/2008). Lett. Aguedo. 2007. Engel. 155. [17] J. J. Tech. 2005. M. Knoll. Lichtenthaler. Mishra. Biotechnol. García-Abuín. 2013. Navaza. 1779–1784 P Pg Pg0 [bar] [W] [W] QG Re V VI Vs [vvm] [–] [m3] [mL] [mL] 1783 absolute pressure power input to the aerated system power input to the non-aerated system volumetric aeration rate Reynolds number working volume iodine solution volume sample volume Greek letters a. MI 1983. 92. particularly for low kLa values as obtained in the experiments conducted at 200 rpm stirring rate. G. P. 36. [11] D. [5] M. P. Gomes. Cheremisinoff. 2008. I. Eng. J. Miller. Symbols used a c [m–1] [–] C [mgO2 L–1] C* D Di Fg H kL kLa [mgO2 L–1] [m] [m] [m3s–1] [m] [m s–1] [h–1] KT [–] N Np OTR [rpm] [–] [mgO2 L–1h–1] specific interfacial area constant dependent on the impeller used dissolved oxygen concentration in the liquid solubility of oxygen in the liquid bioreactor vessel diameter impeller diameter volumetric gas flow rate bioreactor vessel height liquid side mass transfer coefficient oxygen volumetric mass transfer coefficient constant dependent on the impeller used agitation rate power number oxygen mass transfer rate Chem. A. 2009. 35. Lopes. Mota. Prausnitz. E. F. Biotechnol. Belo. An empirical correlation to predict the kLa value as a function of pressure. 1988. Prog. 665. Hahnemann. 37. H. The application of increased air pressure up to 5 bar proved to be a successful means to improve OTR. B. PrenticeHall. Pinheiro. G. Lett. Fernstorm. 19. N. Biotechnol. [18] I. 167. 4 Conclusions Mass transfer between gas and liquid phases in stirred-tank reactors is a key process in the chemical and biochemical industry. 46. Novak. Daugulis. J. P. J. J. Navaza. 380. M. Appl. © 2013 WILEY-VCH Verlag GmbH & Co. KGaA. Englewood Cliffs. D. 2011. Marrucho. Mota. Biochem. No. Chem. Knoll./ Biotechnol. Amaral.cet-journal. 2. 2007. 2003. [4] N. Pinheiro. H. Biotechnol. M. Eng. M. [14] M. [15] A.e. 2003. 38. [19] N. d. M. 6 (2). Z. C. Biotechnol. i. Eng. Z. M. in Handbook of Fluids in Motion (Eds: N. 1091. Bioeng. power input.

Wen. 1842. Eng. Whang. 152. M. Bos. 1217. 36. J. 8. S. © 2013 WILEY-VCH Verlag GmbH & Co. 2009. Gómez-Díaz. Dietrich. Inst. No. Technol. 35 (10). 79 (C). Cong. Sci. Maier. 2001. Biotechnol. Oyevaar. [21] P. Mota. Bioeng. Res. 451. 244. [24] J. Liu. 433. R. J. 10. S. J. Eng.cet-journal.-D. 1998. Belo. [23] I. M. R. [25] Q. C. [29] B. Eng. Orejas. [28] M. Belo et al. X. Technol. I. Eng. N. 31. Wu. 517. Chem. 1999. Juárez. 311. Appl. Microbiol. 53. Wen. J. Chen. R.-L. 18. 2000. A. KGaA. Büchs. www. 2001. Chem.-Y. Z. Eng.-M. Pinheiro. Eng. Yang. 2012. T. Chem. Bioprocess. W. Mota. 1779–1784 . N. Prog.1784 I. 1991. C. Zhang. Biotechnol. [22] J. 2013. 62 (3). Chem. Teixeira. Belo. [26] I. [27] D. Lat. 46 (5/6). Wang. 107. Weinheim Chem. Belo. Trans. 1992. Am. Chen. Gomes. Appl. G. Westerterp. Biotechnol.