Sparger design as key parameter to define shear conditions in pneumatic bioreactors

Highlights

• Airlift and stirred tank bioreactors exhibited similar shear environments.

• Maximum shear rate was observed in the bottom region, around sparger holes.

• Sparger design plays a key role in determining the maximum shear rate.

• Maximum shear rate and shear frequency must be used for proper bioreactor comparison.

Abstract

The average shear rate (${\dot{\gamma }}_{av}$) is a parameter used to characterize the shear environment in bioreactors, enabling comparison of the performances of different bioreactor models in terms of microorganism morphology and viability, and consequently bioproduct formation. Based on this approach, pneumatic bioreactors have been classified as low shear devices. However, the shear behavior cannot be generalized over a wide range of operating conditions, suggesting that the maximum shear rate (${\dot{\gamma }}_{max}$) may be more suitable for the purpose of bioreactor performance comparison. Therefore, the aim of this work was to evaluate average and maximum shear rates in pneumatic bioreactors (bubble column and airlift), based on computational fluid dynamics (CFD) simulations. Concentric-duct and split airlift bioreactors exhibited higher ${\dot{\gamma }}_{av}$ values, compared to the bubble column design, due to the nature of the liquid circulation patterns. The pneumatic bioreactors exhibited a significant (order of magnitude) difference between ${\dot{\gamma }}_{av}$ (11.0–27.3 s⁻¹) and ${\dot{\gamma }}_{max}$ 4555 to 25,040 s⁻¹), reflecting a non-uniform spatial distribution. The ${\dot{\gamma }}_{max}$ values occurred close to the sparger holes and presented a linear relationship with gas injection velocity, which is dependent on the sparger geometry. In this way, sparger characteristics (number and diameter of sparger holes) defined ${\dot{\gamma }}_{max}$ values in pneumatic bioreactors, showing that sparger should be properly designed in order to avoid excessive local shear rates.

Introduction

The selection of a bioreactor model for application in an aerobic bioprocess should consider not only the oxygen transfer capability, but also the shear environment, since biochemical processes are extremely sensitive to the shear intensity when fragile animal cells, plant cells, and filamentous microorganisms are used [1]. Excessive shear can cause cell damage, leading to viability loss and cell disruption or disintegration, so bioreactors should provide moderate or low shear environments, in order to avoid such effects [1,2].

The traditional parameter applied for evaluation of shear conditions is the average shear rate (${\dot{\gamma }}_{av}$), which is extensively used for comparison of the performances of different bioreactor models. Several studies have evaluated this parameter, with the development of methodologies and correlations for bioreactor types including the bubble column [[3], [4], [5], [6], [7]], airlift [2,6,[8], [9], [10], [11]], and stirred tank [7,[12], [13], [14], [15], [16]]. Based on this approach, the effect of ${\dot{\gamma }}_{av}$ on microbial growth, enzyme activity, and biomolecule production has been evaluated [17]. A general observation is that pneumatic bioreactors provide environments with lower shear, compared to stirred tank bioreactors [1,10,18,19].

However, there are exceptions to this general behavior. Studies have evaluated cellulase and xylanase production using different strains of Aspergillus sp., comparing the performance of pneumatic and stirred tank bioreactors [[20], [21], [22]]. Typically, higher enzyme yields were achieved using the pneumatic bioreactors (airlift and bubble column), which could be attributed to the low shear environments in pneumatic bioreactors. However, enzyme production by filamentous fungi is favored by a dispersed hyphal morphology. Consequently, higher enzyme production may be induced by the higher shear rate in pneumatic bioreactors, because exposure of the fungi to the shear environment leads to a predominantly dispersed morphology. On the other hand, when Aspergillus sp. grows with pelleted morphology, production of citric acid may increase [23].

The exceptions mentioned above regarding the effects of shear levels in pneumatic and stirred tank bioreactors have also been observed in other studies. Siedenberg et al. [24] cultivated Aspergillus awamori in stirred tank and airlift bioreactors, using wheat bran as carbon source and inductor of xylanase production. Analysis of fungal morphology showed pelleted growth in the conventional stirred tank bioreactor, while filamentous mycelium formation was observed in the airlift bioreactor, which could have been due to a higher shear environment in the latter device.

Cerri and Badino [17] correlated the production of clavulanic acid by Streptomyces clavuligerus with the average shear rate in 4 L stirred tank and 6 L concentric-duct airlift bioreactors. For the same initial oxygen transfer condition, the maximum value of the fermentation broth consistency index (K_max), which is a rheological parameter related to the morphological structure of the mycelium, was lower for cultivation in the concentric-duct airlift bioreactor, compared to the stirred tank bioreactor. It was found that a higher shear rate was associated with greater fragmentation of the bacterial hyphae and a lower consistency index of the broth, indicating higher shear rates in the pneumatic bioreactor.

Jesus et al. [25] evaluated the hydrodynamics and oxygen transfer in stirred airlift and stirred tank bioreactors operated with xanthan gum solution, at specific air flow rates (φ_air) ranging from 0.5–1.5 vvm. At the same agitation speed (N), the values of the volumetric oxygen transfer coefficient (k_La) and average shear rate (${\dot{\gamma }}_{av}$) were higher for the stirred airlift bioreactor, due to the additional liquid circulation flow pattern observed in this bioreactor. However, the shear level depended as much on the presence or absence of mechanical agitation as on the operating conditions (N and φ_air).

These previous findings question the established protocol for evaluation of the operation and performance of bioreactors, since they suggested that ${\dot{\gamma }}_{av}$ may not be an ideal parameter to use for characterization of shear conditions in bioreactors. In fact, adoption of a maximum shear rate (${\dot{\gamma }}_{max}$) would seem to be more suitable for the purpose of comparison among different bioreactor models, since it describes the worst condition to which a microorganism could be exposed. Depending on the ${\dot{\gamma }}_{max}$ value, a single exposure to this condition could be sufficient to cause irreversible damage to the cells. The ${\dot{\gamma }}_{max}$ represents a local value associated with a particular location inside the bioreactor. In stirred tank bioreactors, the region around the impeller exhibits the highest shear rate [26], while in airlift bioreactors, this worst shear condition is found in the bottom region, around the sparger holes [[27], [28], [29]]. Nonetheless, despite the effect that ${\dot{\gamma }}_{max}$ has on cells, several other factors may affect the resistance of cells to shear, such as cell size and the mechanical resistance of the cell wall.

The shear conditions in different pneumatic bioreactors were evaluated by Thomasi et al. [6], who performed Streptomyces clavuligerus cultivations for clavulanic acid production in 5 L bubble column (BC), concentric-duct airlift (CDA), and split airlift (SA) bioreactors. Using the relation between the maximum value of the fermentation broth consistency index (K_max) and the average shear rate (${\dot{\gamma }}_{av}$), proposed by Cerri and Badino [17], Thomasi et al. [6]. obtained the following order for the average shear rate: ${\dot{\gamma }}_{av,CDA}>{\dot{\gamma }}_{av,SA}>{\dot{\gamma }}_{av,BC}$. The value of ${\dot{\gamma }}_{av}$ appeared to be related to the capacity of the bioreactor to produce an internal circulation, since the lowest ${\dot{\gamma }}_{av}$ values were found for the bubble column bioreactor, which had no well-defined liquid circulation pattern.

The evaluation of ${\dot{\gamma }}_{max}$ can be performed using computational fluid dynamics (CFD) to solve conservative equations (continuity and momentum), with calculation of the flow fields and velocity profiles enabling estimation of local and average shear rates for different fluids and operating conditions. CFD has been extensively applied for evaluation of the performance of pneumatic bioreactors, mainly by analyzing variables such as liquid velocity [[27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42]], global and local gas hold-up [[27], [28], [29], [30],32,33,[35], [36], [37], [38], [39], [40], [41],[43], [44], [45]], and volumetric oxygen transfer coefficient [28,31,38,43]. However, few studies using CFD to evaluate the shear conditions in pneumatic bioreactors are reported in the literature [[27], [28], [29],31,37,44]. Recently, Esperança et al. [46] showed that CFD was a suitable tool for estimation of the average shear rate (${\dot{\gamma }}_{av}$) in pneumatic bioreactors, obtaining values for this parameter within an expected range. In the present work, the aim was to evaluate and compare the average shear rate (${\dot{\gamma }}_{av}$) and maximum shear rate (${\dot{\gamma }}_{max}$) in airlift and bubble column bioreactors, as well as to obtain a relation between ${\dot{\gamma }}_{max}$ and the sparger characteristics of pneumatic bioreactors.