Mass transfer in sparged and stirred reactors: influence of carbon particles and electrolyte

https://doi.org/10.1016/j.ces.2003.05.004Get rights and content

Abstract

Mass transfer in multiphase systems is one of the most studied topics in chemical engineering. However, in three-phase systems containing small particles, the mechanisms playing a role in the increased rate of mass transfer compared to two-phase systems without particles, are still not clear. Therefore, mass transfer measurements were carried out in a 2D slurry bubble column reactor (0.015×0.30×2.00m3), a stirred tank reactor with a flat gas–liquid interface, and in a stirred tank reactor with a gas inducing impeller. The rate of mass transfer in these reactors was investigated with various concentrations of active carbon particles (average particle size of 30μm), with electrolyte (sodium gluconate), and with combinations of these. In the bubble column, high-speed video recordings were captured from which the bubble size distribution and the specific bubble area were determined. In this way, the specific mass transfer area agl was determined separately from the mass transfer coefficient kl. Mechanisms proposed in literature to describe mass transfer and mass transfer enhancement in stirred tank reactors and bubble columns are compared. It is shown that the increased rates of mass transfer in the 2D bubble column and in the stirred tank reactor with the gas inducing impeller are completely caused by an increased gas–liquid interfacial area upon addition of carbon particles and electrolyte. It is suggested that an increased level of turbulence at the gas–liquid interface caused by carbon particles accounts for a smaller effective boundary layer thickness and an enhancement of mass transfer in the flat gas–liquid surface stirred tank reactor. However, for the carbon particles used in this study, it is rather unlikely that mass transfer enhancement takes place due to the well-known shuttle or grazing effect.

Introduction

In industry, three-phase systems are applied in the Fischer–Tropsch synthesis, biological wastewater treatment, and in many production processes in the fine chemicals sector. Mass transfer is one of the key parameters determining the performance of these three-phase systems. A good understanding of mass transfer is relevant to obtain adequate reactor designs. Most mass transfer phenomena can be well described with the two-film model of Whitman (1923). Because the two-film model is commonly used as a steady-state model, various dynamic models were developed, to model non-stationary mass transfer phenomena as well. These models are generally addressed as penetration models or surface renewal models. Depending on the hydrodynamics of a certain system, several age distribution functions for the liquid elements at the surface have been developed, of which the models of Higbie (1935) and Danckwerts (1951) are the most well known. These models have been used to describe mass transfer phenomena in many gas–liquid contactors. Alper, Wichtendahl, and Deckwer (1980) and Alper and Ozturk (1986) suggested that for three-phase systems with small particles, suspended in the liquid, the two-film model and the penetration model could not describe the observed mass transfer phenomena. Therefore, Alper et al. introduced the concept of enhancement of mass transfer, which is due to the presence of the small particles in three-phase systems. The particles are supposed to adsorb an additional amount of the transferred component at or near the gas–liquid interface within the mass transfer boundary layer, after which this adsorbed gas desorbs from the particles, in the liquid bulk. This effect has been described as the so-called “shuttle” or “grazing” effect. Alper et al. (1980), Alper and Ozturk (1986) and many others (Dagaonkar, Beenackers, & Pangarkar 2001a, Dagaonkar, Beenackers, & Pangarkar 2001b; Demmink, Mehra, & Beenackers, 1998; Zahradnik, Kuncova, & Fialova, 1999; Tinge and Drinkenburg 1992, Tinge and Drinkenburg 1995) have measured an increased rate of mass transfer in several three-phase systems (Quicker, Alper, & Deckwer 1987, Quicker, Alper, & Deckwer 1989). However, as shown in the review of Beenackers and van Swaaij (1993), the exact cause of this mass transfer enhancement is not clear.

The mechanism of mass transfer enhancement by the shuttle or grazing effect is closely related to the mechanism of mass transfer as described by the penetration theory: the refreshment of small particles adsorbing gas at the gas–liquid interface after which the gas is desorbed in the liquid bulk, is very similar to the refreshment of liquid phase elements at the gas–liquid interface. Holstvoogd, van Swaaij, and van Dierendonck (1988) attempted to model the mass transfer enhancement as described by Alper et al. (1980) and Alper and Ozturk (1986) with the penetration model. They modelled the increased rate of mass transfer by assuming a decreased effective diffusion layer at the gas–liquid interface, caused by adsorption of gas by the particles in the diffusion layer. They concluded that only a very high adsorption capacity of the particles could account for the mass transfer enhancement as suggested by Alper et al. (1980) and Alper and Ozturk (1986). More recently, Van der Zon, Hamersma, Poels, and Bliek (1999) used the two-film model to calculate mass transfer enhancement during reaction in a three-phase system. They also found that mass transfer was enhanced by the catalyst particles and that the enhancement was a function of the hydrophobicity of the particles used. Although the rate of mass transfer is well described by the models of both Holstvoogd et al. (1988) and Van der Zon et al. (1999), the exact cause of the mass transfer enhancement is still not understood. Both models neglect other possible mechanisms leading to an increased rate of mass transfer. Other mechanisms have been published in many articles and reviews like those of Lee and Foster (1990) and Beenackers and van Swaaij (1993) and are also supported by recent studies in our laboratory (Kluytmans, van Wachem, Kuster, & Schouten, 2001). From these studies three mechanisms can be identified, which may account for an increased rate of mass transfer in three-phase systems. In this work, three different reactors, a 2D slurry bubble column, a stirred tank reactor with a gas inducing impeller, and a stirred tank reactor with a flat gas–liquid interfacial area, were used to study these mechanisms.

Mechanism 1: Shuttle or grazing effect: This mechanism has been described by Alper et al. (1980) and Alper and Ozturk (1986): the particles are supposed to transport an additional amount of gas to the liquid bulk through adsorption in the gas–liquid diffusion layer and desorption in the liquid bulk. As mentioned before, the shuttle or grazing effect is very similar to the penetration theory. Therefore, it is expected that with increasing carbon particle concentration and with increasing stirrer speed in a stirred tank reactor, the refreshment rate of carbon particles at the gas–liquid interface will increase, leading to an increased transport of gas from the gas–liquid interface to the liquid bulk, which will result in a larger mass transfer coefficient kl.

Mechanism 2: Hydrodynamic effects in the gas–liquid boundary layer: The presence of particles can affect the hydrodynamic behavior of three-phase systems. Particles can collide and interact with the gas–liquid interface or may induce turbulence at the gas–liquid interface, leading to a smaller effective diffusion layer. Diffusion of gas into the liquid film, and mixing of gas into the bulk liquid can therefore be increased by the presence of particles, leading to an increase in the mass transfer coefficient kl. Increasing the stirrer speed in a stirred tank reactor or increasing the superficial gas velocity in a bubble column, will increase the shear stress in the system. Eventually, the shear stress in the system will be much higher compared to the forces induced by the small particles. The relative effect of carbon particles on the increase of the gas–liquid mass transfer will therefore decrease, if the shear stress in the system becomes higher. The number of collisions of carbon particles with the gas–liquid interface or the degree of induced turbulence at the gas–liquid interface, is not necessarily dependent on the concentration of the particles in the bulk liquid. More important in this case are the number of particles present at the interface and the nature of the particle interactions with the interface, which are mainly determined by the affinity of the particles for the gas–liquid interface, as expressed by the hydrophobicity or hydrophilicity of the particles.

Mechanism 3: Changes in the specific gas–liquid interfacial area: The increase of gas–liquid mass transfer is generally expressed by an increase in the combined mass transfer coefficient klagl. The increase in mass transfer can thus be due to a change in the mass transfer coefficient kl or due to a change in the specific gas–liquid interfacial area agl. Previous studies (Kluytmans et al., 2001) showed that carbon particles and electrolyte affect the gas hold-up and therefore the gas–liquid interfacial area in a 2D bubble column. Besides changes in the specific gas–liquid interfacial area, no additional increase in the gas–liquid mass transfer is expected upon changing the carbon particle concentration, stirrer speed or superficial gas velocity, if only this mechanism is present. Also for this mechanism, with increasing superficial gas velocity or stirrer speed, the shear stresses in the system increase, decreasing the effect of electrolyte and carbon particles on the increased gas–liquid interfacial area, and thus decreasing the effect on the rate of gas–liquid mass transfer (Kluytmans et al., 2001).

The objective of this work is to clarify which mechanism leads to the observed increased rate of gas–liquid mass transfer in sparged and stirred three-phase reactors. By definition, mass transfer enhancement as considered in this article, is defined as the increased rate of gas–liquid mass transfer due to an increase in the mass transfer coefficient kl (mechanisms 1 and 2).

Section snippets

Experimental setup and procedures

Experiments were carried out at ambient conditions. From previous work (Kluytmans et al., 2001) it was found that the type of gas (nitrogen, oxygen, air) did not influence mass transfer significantly. In all three reactors, experiments were carried out with distilled water, carbon particles, electrolyte, and combinations of carbon particles and electrolyte solutions. Sodium gluconate was used as electrolyte in concentrations of 0.05–0.5M. Carbon particles (Engelhard Q500-130) with a mean

Mass transfer in a 2D bubble column

The rate of mass transfer was measured at different superficial gas velocities and in different carbon particles slurries and electrolyte solutions. Measurements in solutions containing both carbon particles and electrolyte did not give reliable data, because of severe foaming. As shown in Fig. 1, the rate of mass transfer increases with increasing superficial gas velocity and under influence of addition of carbon particles and electrolyte. The klagl values were calculated from the oxygen

Discussion

From the experiments, it is not straightforward to appoint one of the mechanisms described in Section 1.2, as being responsible for the observed increased rate of mass transfer in all three reactors. The observed phenomena in each reactor should therefore be compared with the expected phenomena for each mechanism as described in Section 1.2.

Conclusions

This article shows that finding the cause of an increased rate of mass transfer in three-phase systems is not straightforward. In most cases, it is not possible to measure the parameters well to clarify the exact mechanisms, like the effective diffusion layer, the adsorption and desorption rates of oxygen on a carbon interface in a liquid, the oxygen transport by particles, and the specific gas–liquid interfacial area. However, we have shown that by combining the results of experiments in three

Notation

aglspecific gas–liquid interfacial area, m2m−3
Cl,bulkconcentration in the liquid bulk, molm−3
Cl,iconcentration at the gas–liquid interface in the liquid, molm−3
Csensorconcentration measured by sensor, molm−3
deffeffective diffusion layer thickness, m
Ddiffusion coefficient, m2s−1
feffratio between effective diffusion layers, dimensionless
kloverall mass transfer coefficient, ms−1
kl,largemass transfer coefficient of the large bubbles, ms−1
ksensorsensor constant, s−1
kl,smallmass transfer coefficient

References (21)

There are more references available in the full text version of this article.

Cited by (146)

  • Boosting photosynthetic biogas upgrading via carbon-coated zero-valent iron nanoparticle addition: A pilot proof of concept study

    2023, Sustainable Chemistry and Pharmacy
    Citation Excerpt :

    Moreover, the addition of polymeric nanofibers containing Fe2O3 NPs has also been validated as a promising tool to enhance CO2 biofixation by Chlorella fusca (Vaz et al., 2019a, 2019b, 2020). Additionally, Kluytmans and co-workers demonstrated that the addition of carbon particles enhanced the interfacial area of the gas-liquid phases in a 2D bubble column (Kluytmans et al., 2003). Despite the fact that the gas-liquid mass transfer of CO2 is complex and depends on the type of photobioreactor, agitation and gas sparging device, there are three potential mechanisms of NP interaction: 1) gas bubble breaking effect, where NPs promote smaller gas bubble diameters and the gas-liquid diffusion area increases; 2) shuttle effect, where CO2 is rapidly adsorbed to the NPs surface and then slowly released into the aqueous medium; 3) hydrodynamic effect, where NPs collide, inducing turbulence and refreshing the liquid-gas boundary layer (Choi et al., 2015).

View all citing articles on Scopus
1

Present address: Department of Thermo and Fluid Dynamics, Chalmers University of Technology, SE-412 96 Göteborg, Sweden.

View full text