Effect of surfactants on liquid-side mass transfer coefficients in gas-liquid systems: A first step to modeling

This paper focuses on the effect of surfactants on the mass transfer parameters (volumetric mass transfer coefficient k_La and liquid-side mass transfer coefficient k_L). Tap water and aqueous solutions with surfactants (anionic, cationic and non-ionic at concentrations up to are used as liquid phases. The bubbles are generated into a small-scale bubble column having an elastic membrane with a single orifice as gas sparger. To understand the effects of the surfactants on the mass transfer, not only the static surface tension is used, but also the characteristic adsorption parameters like the surface coverage ratio at equilibrium Se. The liquid-side mass transfer coefficient is obtained from the ratio of the volumetric mass transfer coefficient (measured by a chemical method) and the specific interfacial area. These two parameters are obtained simultaneously. The methods used to obtain these parameters are described in Painmanakul et al. [2005. Effects of surfactants on liquid-side mass transfer coefficients. Chemical Engineering Science 60, 6480-6491].Whatever the liquid phase, three zones are found on the liquid-side mass transfer coefficient variation with the bubble diameter. For bubble diameters less than 1.5 mm, whatever the liquid phases, the k_L values are roughly constant at . For bubble diameters greater than 3.5 mm, the k_L values do not vary much with the bubble diameter, but depend on the surfactant concentration. For bubble diameters between 1.5 and 3.5 mm, the k_L values increase from to the value reached at 3.5 mm. This increase depends on the surfactants. Higbie's model does not represent the k_L values for bubble diameters greater than 3.5 mm, even though there is a small amount of surfactant in the liquid phase. Thus, a model is proposed for each zone described above. Explanations are also proposed for the effect of the surfactant on the k_L values for each of the above zones.     

Multicomponent phase equilibria of thin liquid films and their vapour

Subject of the paper are multicomponent phase equilibria between extremely thin liquid films and their vapour phase on curved or planar solid surfaces. The solid surface is either heated or cooled, so that the liquid films are evaporating or the vapour is condensing.The curvature of the film surface and long range molecular forces due to the van der Waals attraction acting over distances of about 0.2- between solid and liquid film can lead to a composition shift in liquid and vapour phase compared to the composition that would be observed at planar and not extremely thin liquid films in equilibrium with their vapour phase.The Kelvin equation for the interfacial pressure is derived, as well as the equations for the composition shift. As a numerical example shows the van der Waals forces considerably influence the pressures at the liquid-vapour interface. Responsible for the composition shift is a dimensionless thermodynamic quantity . For small values D_i→0, liquid-vapour phase equilibria become identical with those of planar surfaces. For D_i?0 the vapour phase contains less and for D_i?0 more of the lighter volatiles than planar interfaces.     

Lateral mixing in trickle bed reactors

Knowledge of lateral mixing is essential to understand heat and momentum transfer parameters in both single-phase liquid and two-phase gas-liquid co-current down flow through packed bed columns. The reactors through which gas and liquid concurrently flow downwards through a bed of catalytic packing are called trickle bed reactors. Experimental data on lateral mixing coefficients from both the heat transfer and radial liquid distribution studies are obtained over a wide range of flow rates of gas and liquid using glass spheres (4.05 and 6.75 mm), ceramic spheres (2.59 mm), and ceramic raschig rings (4 and 6.75 mm) as packing materials covering trickle flow, pulse flow, and dispersed bubble flow regimes. In the present work, an expression for estimation of lateral mixing coefficient (αβ)_L is derived using the data on radial liquid distribution studies. The agreement between the values of (αβ)_L obtained from heat transfer studies and from radial liquid distribution studies using the experimental data shows that there exists an analogy between the heat transfer and radial liquid distribution in packed beds. Since (αβ)_L is an important variable for estimation of various heat and mass transfer parameters, a correlation for (αβ)_L based on present heat transfer study is proposed. The agreement between the (αβ)_L values estimated from the proposed correlation and experimental values is satisfactory with a standard deviation (s.d.) of 0.119.     

A comparative study of PtCo, PtCr, and PtCoCr catalysts for oxygen electro-reduction reaction

A comparative study of carbon supported Pt₃₀Co₇₀, Pt₃₀Cr₇₀, and Pt₃₀Co₃₀Cr₄₀ catalysts for oxygen electro-reduction reaction (ORR) activity was performed. In alloy catalysts synthesized via NaBH₄ reduction, more than a 3-fold improvement was observed in ORR specific activity compared with that of Pt/C catalyst, while mass activities did not show significant improvement. After annealing at 900 °C under reducing conditions, the ORR specific activities of the alloy catalysts increased to give a relative ORR catalytic activity ordering of PtCo/C-900 (2800 μA ) > PtCoCr/C-900 (1770 μA ) > PtCr/C-900 (871 μA ) > Pt/C-900 (393 μA ) > Pt/C (334 μA ). On the other hand, the ORR mass activity followed an order of PtCr/C-900 (140 mA ) > PtCoCr/C-900 (111 mA ) > PtCo/C-900 (84.1 mA ). Cyclic voltammetry results suggest that incorporation of Cr resulted in a large electrochemically active surface area producing higher mass activity in the PtCr/C-900 catalyst although it showed the lowest specific activity among the alloy catalysts. The intermediate EAS and ORR activity values of the PtCoCr/C-900 catalyst suggest that the characteristics of the PtCo/C-900 (low mass and high specific activities) and PtCr/C-900 (high mass and low specific activities) are combined by alloying of Pt with both Co and Cr.     

The intensification of gas-liquid flows with a periodic, constricted oscillatory-meso tube

The experimental study of gas dispersion in a vertical periodically, constricted, oscillatory meso-tube (OMT) is herein presented. Water was continuously pumped through the OMT in the laminar flow regime along with an oscillatory flow component superimposed into the net flow in a range of fluid oscillation frequency (f) and centre-to-peak amplitude (x₀) of and 0-3 mm, respectively, in the presence of a very low superficial gas velocity . Bubble images were recorded with a CCD camera and analysed with Visilog^® software. A bimodal distribution of bubble size was in general observed but the bubble size was found strongly dependent on the oscillatory flow mixing conditions imposed into the fluid. A number fraction of micro-bubbles (with an equivalent diameter, D_eq, equal or bellow 0.2 mm) up to 60% was generated with increasing values of x₀ (i.e. 3 mm) and values of f in the range . Furthermore, it is demonstrated that the Sauter mean diameter, D₃₂, and the specific interfacial area, a, can be fined tune by setting both f and x₀ in this studied range. The high number fraction of micro-bubbles was concluded to have a positive impact in enhancing the liquid-side mass transfer coefficient, k_L. Globally, the differences in bubbles sizes were found to play a marginal effect in the global enhancement of the k_La in the meso-tube in comparison with the intensive contact experimented by the bubbles rising in the oscillatory flow. The higher order of magnitude of the k_L values found in this work (up to ) is promising for running numerous industrial gas-liquid flows processes through smaller and better, while aeration of biotransformations can be run more efficiently, as supported by our recent proof-of-concept studies carried out in the platform.     

Mechanism of enhanced gas absorption in presence of fine solid particles. Effect of molecular diffusivity on mass transfer coefficient in stirred cell

Desorption of oxygen and hydrogen from various liquids (water, 0.8 molar sodium sulphate solution) containing suspended particles of activated carbon at various solid loading was investigated. The desorption was used to avoid supersaturation effect which was observed during oxygen and hydrogen absorption into liquid saturated with nitrogen. Experiments were carried out in a stirred cell with flat gas-liquid interface at and atmospheric pressure. An increase of k_L upon addition of the particles was observed. Enhancement factor increases with increasing contact time of the particles with liquid reaching maximum steady-state value of approx. 3 after sufficiently long time (a few hours) regardless of solid loading , agitator frequency and solute gas (O₂,H₂). The results fit the correlation (e is specific power dissipated by agitator in liquid and D is molecular diffusivity of gas absorbed) with the exponent for liquids without and for the liquids with the particles. It indicates that the interface is rigid in absence of particles and hinders the motion of liquid along the interface forming boundary layer while in the presence of particles the interface is completely mobile and surface renewal proceeds according to the penetration model. These results confirm a finding of Kaya and Schumpe (2005) that the enhancement of mass transfer in the cell at the presence of hydrophobic solids is due to clean-up of the interface from surfactants by their adsorption on hydrophobic solids rather than by a “shuttle mechanism” exerted by particles with a high adsorption capacity for the transfer component.     

Mixing in bubble columns: a new approach for characterizing dispersion coefficients

Use of bubble columns as photobioreactors requires a quantitative knowledge of radial mixing in these columns. A complete model of liquid-phase dispersion was used to simultaneously characterize axial and radial mixing in a relatively large (0.06 m³, 2.3 m tall, 0.193 m in diameter) bubble column photobioreactor. Axial and radial dispersion coefficients and mixing times were determined in tap water and sea water for superficial aeration velocities of up to . The measured axial dispersion coefficients (D_z) were generally consistent with the predictions of the well established correlations, thus validating the complete dispersion model used in the analysis. The D_z values ranged from ∼150 to and were highly reproducible. There was evidence that the existing literature data on D_z in bubble columns are slightly underestimated, as consistent underestimation was found to be a characteristic of the widely used dispersion model that disregards radial dispersion. The value of the radial dispersion coefficient was typically about 1% of the D_z value under any given condition. Except at incipient aeration, the radial dispersion coefficient was not as sensitive to the magnitude of the aeration rate as was the axial dispersion coefficient. The mixing time data were generally consistent with the existing correlations.     

Application of the off-gas method to the measurement of oxygen transfer in biofilters

To determine the oxygen mass transfer in clean water in biofilters, a method based on the follow up of the oxygen fraction in the off gas during the oxidation of sulphite in excess has been evaluated and applied to a pilot-scale unit (250 L, superficial gas velocity from 0 to , superficial liquid velocity from 0 to ). Tests performed on a two-phase reactor showed that, without any cobalt addition, standard oxygen transfer efficiencies (SOTE) obtained from the proposed method are not statistically different from those issued from the standardised method. A relationship has been proposed to express SOTE values as a function of the conductivity, and the influence of the gas and liquid velocities on SOTE and k_La has been investigated.     

Determination of mass transfer coefficients for packing materials used in biofilters and biotrickling filters for air pollution control. 1. Experimental results

Gas film mass transfer coefficients (k_Ga_t,k_Ga_w) and liquid film mass transfer coefficients (k_La_w) for packing materials used in biofilters and biotrickling filters for air pollution control were determined experimentally. Lava rock, polyurethane foam cube (PUF), Pall ring, porous ceramic beads, porous ceramic Raschig rings and compost-woodchips mixtures were investigated. The experiments were performed at gas velocities ranging from 100 to and liquid velocities of , i.e., a wide range that covers most biofilters and biotrickling filters. k_Ga_t in biofilter packings ranged from about 500 to , while k_Ga_w and k_La_w in biotrickling filters ranged from 100 to , and 1 to , respectively, depending on the packings and the conditions. This is markedly lower than mass transfer coefficients usually observed for conventional wet scrubbing. The gas film mass transfer coefficient (k_Ga_t) of 50:50% vol compost-woodchips mixture, a common biofilter packing, was greater than this of a 20% vol compost and 80% woodchips mixture, though the mass transfer was not increased by increasing further the volume fraction of compost. All compost mixtures exhibited a greater gas film mass transfer coefficient than lava rock or other synthetic materials. The mass transfer coefficients of compost mixtures was also influenced by packing method and it was directly proportional to the surface area of the bulking agents added. The gas film mass transfer coefficient (k_Ga_w) of five biotrickling filter packing materials increased linearly with gas velocity. The effect of liquid on the gas film mass transfer coefficient was not significant. Of all the biotrickling filter packings, the porous ceramic beads had the highest gas and liquid film mass transfer coefficients followed by lava rock, porous ceramic rings, 1 in Pall ring and PUF cubes. The liquid film mass transfer coefficient (k_La_w) was directly proportional to liquid velocity and the effect of gas velocity was negligible. Several correlations allowing prediction of mass transfer coefficients are presented in Part 2 of this paper.     

Heavy crude oil viscosity reduction and rheology for pipeline transportation

Different methods of reducing the viscosity of heavy crude oil to enhance the flow properties were investigated. Experimental measurements were conducted using RheoStress RS100 from Haake. Several factors such as shear rate, temperature and light oil concentration on the viscosity behavior have been studied. This study shows that the blending of the heavy crude oil with a limited amount of lighter crude oil provided better performance than the other alternatives. Experimental measurements in terms of shear stress τ-shear rate and yield stress τ₀ were conducted on the mixture of heavy crude oil-light crude oil (O-light). The results showed a significant viscosity reduction of 375 mPa s at a room temperature of 25 °C. This study shows that the heavy crude oil required a yield stress of 0.7 Pa, whereas no yield stress was reported for the heavy crude oil-light crude oil mixture.     

Static mixers with a gas continuous phase

The aim of this work was to characterise hydrodynamics and mass transfer in a gas-liquid contactor containing static mixers (SMs). The originality of this study lies in the fact that these mixing organs are used with a gas continuous phase. Two types of SM were implemented in co-current flows, Statiflo and Lightnin. The pressure drop ΔP, the volumic interfacial area a and the volumic mass transfer coefficient k_La were measured in several configurations: horizontal flow, vertical up-flow and vertical down-flow. The influences of position and flow rates were studied in order to understand the behaviour of these contactors, and to optimise the operating conditions. As expected, the pressure drop was found to increase mainly with gas velocity but also with liquid velocity, and to reach 3300 Pa in the range of velocities studied (the gas flow rate varied between 4 and and the liquid flow rate between 0 and 100 L/h), far less than Sülzer SM. The volumic interfacial area and the volumic mass transfer coefficient showed the same changes, a varying between 100 and , and k_La reaching 0.07 L/s. This is interesting compared with other classical absorption processes: indeed, even if packing towers can provide the same range of values, the operating conditions are more drastic or the dimensions of the apparatuses are far larger than SM ones. The position was also found to have an influence on the hydrodynamic and mass transfer parameters (ΔP, a and k_La).