This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
The present work shows a global vision of gas absorption in aqueous solutions with Tween80. The study includes a characterization of the absorption process based on hydrodynamics and a different one based on mass transfer. The effect caused by surfactant concentration upon gas hold-up and gas-liquid interfacial area has been analysed. On the other hand, the analysis of the influence of the liquid-phase contamination on the absorption process has been carried out on the basis of the liquid-film mass-transfer coefficient (kL), removing the effect caused by the presence of surfactant and the gas flow rate on the interfacial area (a) and, thereby, on the volumetric mass-transfer coefficient (kLÂ·a).
Keywords: absorption, surfactant, interfacial area, mass transfer coefficient.
Several industrial processes are based on the use of gas-liquid equipment and for this reason, a suitable gas-liquid mass transfer rate is important to different chemical and biochemical processes. The gas phase is placed in the contactor by means of small bubbles in order to supply a large interfacial area and an efficient mass transfer between the gas and liquid phases. The liquid phases commonly found in the chemical industry are very complex, due to the presence of several compounds and to the operating conditions. Consequently, an important research studies have been performed to understand the influence of the liquid phase properties on the bubble formation phenomenon. Nevertheless, whilst the influence of the liquid density and viscosity has been widely studied, the liquid surface tension and its effects are largely an unknown factor. Taken into account the results and conclusions concerning the effect of the liquid surface tension, its influence is not really separated from the effect caused by density and viscosity [1, 2].
In the last few years, several studies have increased the knowledge about the absorption processes in systems that involve the presence of surface active substances. It is generally recognized that small amounts of surface-active additives or contaminants can markedly affect the mass transfer parameters kLÂ·a and kL [3, 4]. These studies have deepened on the influence of different operation variables upon the global mass transfer and, more specifically, upon the gas hold-up , the bubbles diameter and gas-liquid interfacial area , the interfacial turbulence  and the mass transfer coefficient . Other researchers have shown some interest in analysing the influence of the surfactants nature, taking into account the molecules size  as well as their ionic character .
Nowadays, the presence of surface agents in gas-liquid systems has reached great importance, mainly due to the presence of this kind of substances in bioreactors: (i) as a product of the bioreaction (production of biosurfactants) , and also (ii) as a stabilizer substance in the two-phase partition bioreactors (TPPB) to maintain the emulsion formed in the liquid phase .
The aim of the present work is increasing the knowledge of the absorption process behaviour in complex systems, being involved the use of a surfactant commonly used in numerous processes. The system employed in this study has been analysed taking into account hydrodynamic parameters (such as gas hold-up, bubble diameter and interfacial area) and mass transfer based on the analysis of mass transfer coefficient.
2. Material and methods
Tween80 was supplied by Sigma-Aldrich (CAS number: 9005-65-6). Commercial grade carbon dioxide of 99.998% purity, supplied by Carburos Metálicos (Spain), was also used in this work. The solutions were prepared by mass using a balance with a precision of Â±10-7 kg, and bi-distilled water has been employed to prepare the absorbent phases.
The studies of carbon dioxide mass transfer to liquid phases were carried out using a bubble column contactor similar to other employed in former studies related to the absorption processes . The gas/liquid contactor used in these studies has been a square bubble column (side length = 6 cm; height = 114 cm), made in methacrylate and working with a 3-litre liquid volume. The gas sparger has been a glass capillary with only one orifice to produce a small number of bubbles, which allowed us a careful analysis of the operation conditions influence on the bubbles size.
The gas to be absorbed, pure carbon dioxide, was passed through two "humidifiers" at 25°C to prepare the gas phase. This procedure removed the resistance to the mass transfer in the gas phase, and it only allowed us the evaluation of the liquid phase resistance to the gas transfer. The gas flow-rate was measured and controlled with two mass flow controllers (5850 Brooks Instruments). The mass flow controllers employed in the present study for the gas flow-rate and the pressures were calibrated by the supplier.
The pressure drop was measured between the column's inlet and outlet, using a Testo 512 digital manometer. The operational regime was continuous in relation to the gas phase and batch as regards the absorbent liquid.
In this work we have analysed the effect caused by the typical operational variables used in the present contact device (gas flow-rate and liquid phase composition) upon hydrodynamic parameters (gas hold-up and interfacial area), as well as the absorption kinetics of the carbon dioxide mass-transfer process to liquid phases. The liquid phases chosen for the present paper have been aqueous solutions of Tween80, using different surfactant concentrations. Three gas flow-rates were employed (18, 30 and 40 LÂ·h-1) to analyse the influence of this variable upon the gas/liquid absorption process.
The bubble diameter was measured using a photographic method based on images of the bubbles taken along the height of the column, from bottom to top. A Sony (DCR-PC330E) video camera was used to obtain the images. A minimum number of 100 well-defined bubbles along the bubble column were used to evaluate the size distribution of the bubbles in the liquid phase employed, and for each gas flow-rate that has been used. The Image Tool v3.0 software was used to carry out the necessary measurements of the bubbles geometric characteristics.
The overall gas hold-up is an important parameter to determine the gas-liquid interfacial area and it was measured using the volume expansion method. The calculation of the volume change in the bubble column was based on the change observed on the liquid level and on the increase of this value after gassing using the cylindrical bubble column (see equation 1).
where VL is the ungassed liquid volume and DV is the volume expansion after gas dispersion, calculated from the liquid level change and the cross sectional area. The change in the bubble column volume was calculated based on the change observed on the liquid level and on the increase of this value after gassing.
The images of the bubbles we obtained in the liquid phases employed show an ellipsoid shape. For this reason, major (E) and minor (e) axes of the projected ellipsoid (in two dimensions) were determined. The diameter of the equivalent sphere (equation 2) was taken as the representative bubble dimension.
Different authors recommend using the Sauter mean diameter (d32), which can be determined using the data calculated for the equivalent diameter:
where ni is the number of bubbles that have an equivalent diameter (di).
The Sauter mean diameter and the gas hold-up values allow us the calculation of the specific interfacial area using equation 4.
3. Results and discussion
The present work has been focused on the effect caused by the presence of different concentrations of a surfactant in aqueous solution upon the absorption process. The surfactant employed, Tween80, is commonly present in different processes that imply gas-liquid mass transfer processes . One of the studies developed in this work is related to the hydrodynamic behaviour, based on the analysis of the gas hold-up and the gas-liquid interfacial area produced in the contactor. In relation to the first parameter (gas hold-up), the behaviour observed for this experimental system is shown in figure 1. The influence of the gas flow-rate (or superficial gas velocity) upon the gas hold-up is shown in this figure, where we can observe that an increase in the gas flow-rate produces an increase in the gas hold-up value. This kind of behaviour shown in figure 1 indicates that a change in the bubbling regime is not produced in the studied range. The bubbling regime is pseudo-homogeneous in all cases and, then, the coalescence process is not observed. The experimental data shown in figure 1 corresponds to pure water (without Tween80 addition) but the experiments developed with different surfactant concentrations do not show significant changes with regard to the values obtained for carbon dioxide - water system. This behaviour is in agreement with previous studies that analyse the influence of surfactants upon this variable [14, 15].
Also, the hydrodynamic characterization of the gas-liquid systems employed in present work includes the analysis of the bubbles size distribution produced in the bubble column. The bubble size distribution and gas hold-up could be used to gas-liquid interfacial area determination. An example of bubble size distribution determined in present work is shown in figure 2. This figure allows analyse the influence of surfactant concentration upon bubble size distribution. The experimental results indicate that a higher value in surfactant concentration produces a decrease in the bubble size. This behaviour is in agreement with the results obtained for systems with small chain length surfactant [6, 9]. Also, figure 2 shows that an increase in surfactant concentration produces a narrower size distribution.
This behaviour is related to the influence caused by the surfactant presence upon the surface tension value. This kind of substances produces an important decrease in the value of the surface tension, and this behaviour has a high influence upon the bubble size produced in the contactor. Figure 2 shows the decrease observed in the bubbles size distribution produced by the surfactant.
Using the bubble size distribution for each experimental conditions and the gas hold-up produced in the bubble column, the gas-liquid interfacial area was calculated using equation 3. Figure 3 summarized the determined data for the interfacial area under the experimental conditions analysed in this work. Regarding the influence of the surfactant concentration upon the interfacial area, an increase in this parameter was observed when the surfactant concentration also increases in the liquid phase. This behaviour is due to the presence of this solute, that produces a decrease in the bubble size value (vide supra) with no influence upon the gas hold-up.
On the basis of experimental data, we conclude that the largest change in the value of interfacial area is caused by the addition of small quantities of surfactant. Higher values than 0.03% of surfactant concentration produce practically constant interfacial area values with slight changes, in spite of the fact that important quantities of surfactant were added to the liquid phase. A similar behaviour has been observed for other systems that have employed surfactants in aqueous solution [8, 16].
This behaviour has been observed for all the gas flow-rates employed in this work. In relation to the influence of this operation variable, an increase in this value produces an increase in the interfacial area, although the reason is different to the one previously commented, when the surfactant concentration was varied. Regarding the gas flow rate effect, an increase in this variable does not produce significant changes in the bubble size, but on the other hand, this variable produces an increase in the gas hold-up (see figure 1).
The experimental data obtained for the gas-liquid interfacial area has been fitted using an equation based on different operation variables. The correlation used is shown in equation 5, and similar equations based on potential effects of variables have been proposed by other authors [17, 18].
where Cs is the surfactant concentration, cmc is the critical micelle concentration of Tween80 and Qg is the gas flow-rate.
Equation 5 includes the surfactant concentration and the gas flow-rate as important operation variables, but this equation also includes the value of the critical micelle concentration. This parameter provides information about the chain length and other properties related to hydrofobicity. Figure 4 shows a comparison between the experimental values and the calculated ones using the simple correlation (equation 5). Only one surfactant has been employed in this work and then the value of the critical micelle concentration is constant, but the expression used in equation 5 includes this value to preserve the equation general formulation. The use of the same equation allows us to compare the value of the fit parameters with previous and future studies that use other surfactants. Figure 4 shows a comparison between experimental and calculated gas-liquid interfacial area under different experimental conditions, observing a good agreement between the experimental values and the corresponding ones calculated using equation 5.
The present work also includes gas-liquid mass transfer studies corresponding to the absorption of carbon dioxide in Tween80 aqueous solutions. The operation regime was semicontinuous and then, the liquid phase was placed into the contactor and the gas phase was fed continuously to the bubble column. The absorption kinetics was obtained and this experimental data has been employed to calculate the volumetric mass transfer coefficient. The mass transfer coefficient (kL) was calculated taking into account the values of the volumetric mass transfer coefficient and the gas-liquid interfacial area. This way, our aim in this work is analysing the influence of surfactant presence and concentration upon the mass transfer coefficient and the gas-liquid interfacial area, as well as analysing the influence upon each parameter individually.
Figure 5 shows an example of the evolution of the carbon dioxide absorbed concentration in the liquid phase. Being more specific, two experiments are compared in figure 5 using the same surfactant concentration and different gas flow-rate values. The experimental data indicate that an increase in the value of the gas flow-rate fed to bubble column produces a higher increase in the carbon dioxide concentration. This behaviour indicates that the mass transfer rate increases with the gas flow-rate. This behaviour is complemented with the fact previously commented: that an increase in the gas flow-rate produces an increase in the interfacial area that, at the same time, produces an increase in the mass transfer rate.
The carbon dioxide concentration in the liquid phase has been calculated throughout time, based on the experimental results obtained for the absorption rate. In this kind of absorption processes, working in a semi-continuous regime and using a pure gas phase, equation 6 is used to determine the volumetric mass transfer coefficient obtained, using a gas phase mass balance.
where KLÂ·a is the volumetric mass transfer coefficient, and C* and C are the solubility and carbon dioxide concentration, respectively. In the present work, the carbon dioxide solubility value  has been considered equal to the value corresponding to pure water, due the low surfactant concentration employed in the liquid phase. Under the experimental conditions (without gas phase resistance to mass transfer) employed, the individual mass transfer coefficient of the liquid phase is considered equal to the global mass transfer coefficient.
The same experimental procedure has been carried out for the different experiments performed in the present work, and then the volumetric mass transfer coefficient has been calculated for each experimental condition. Figure 6 shows the obtained behavior and the influence of different operation variables, such as gas flow-rate and surfactant concentration. In relation to the influence of surfactant concentration upon the mass transfer coefficient value, an important decrease in this parameter was produced when small quantities of Tween80 were added to the liquid phase, reaching a constant value of mass transfer coefficient and not dependent of surfactant concentration. This behavior is opposite to the previous one commented about the influence of surfactant concentration upon the value of gas-liquid interfacial area (the increase of surfactant concentration produced an increase in interfacial area until a constant value). The influence of surfactant concentration upon the interfacial area must be an increase in the value of the volumetric mass transfer coefficient; however, the obtained behavior is the opposite. Taking this fact into account, a priori conclusion is that the presence of surfactant in the liquid phase has a very important negative effect upon the mass transfer coefficient, and this important effect is higher that the positive influence caused upon the interfacial area.
On the other hand, in relation to the influence of gas flow-rate fed to the bubble contactor, an increase in the value of the volumetric mass transfer coefficient has been observed when the gas flow-rate increases. Taking into account the previously analysed influence of the gas flow-rate upon the interfacial area and the results for volumetric mass transfer coefficient, we can conclude that the gas flow-rate affects positively upon both parameters (mass transfer coefficient and interfacial area).
Using the experimental values of volumetric mass transfer coefficient and the previously determined interfacial area, we can calculate the mass transfer coefficient values for each experimental condition by means of equation 7.
The experimental data shown in figure 7 indicates a similar behaviour to the previous one obtained for the volumetric mass transfer coefficient. A decrease in the mass transfer coefficient value is observed when low additions of surfactant are added to the liquid phase. Therefore, the conclusions proposed a priori are certain. The effect of the gas flow-rate upon the mass transfer coefficient (shown in figure 7) indicates that this variable loses importance, and the mass transfer coefficient values for the different gas flow-rates are similar. This behaviour is in agreement with previous studies that conclude the non- influence of the gas flow-rate upon the mass transfer in this kind of contactors .
Previous studies employing similar systems indicate that the presence of surfactant in the liquid phase reduces the mass transfer coefficient until a plateau for higher concentrations than the critical micelle concentration . This reduction has been assigned to different reasons in relation to the increment in the transport resistance caused by the presence of surfactant molecules. These molecules produce a reduction in the liquid elements renewal near the interface. Then, it produces a decrease in the driving force that is directly related to the gas mass transfer rate to the liquid phase.
Other studies have concluded that a low surfactant concentration produces an enhancement of mass transfer that produces an increase in the value of the mass transfer coefficient [20, 21]. In this work, this increase or enhancement is not observed and this behaviour is assigned to the size of surfactant molecule, compared to the experimental systems that show the enhancement behaviour .
The decrease in the mass transfer coefficient by the presence of surfactant molecules in the liquid phase is assigned to different modifications caused by the accumulation of surfactant molecules at the gas-liquid interface. This accumulation causes a reduction in the renewal of the liquid elements and then, a decrease in the value of the driving force. These effects produce a decrease in the mass transfer rate. Different studies  have concluded that this reduction in the mass transfer rate is observed until the surfactant concentration reaches the value corresponding to the critical micelle concentration. When this concentration is reached, it is impossible to increase the surfactant concentration in a gas-liquid interface because the micelle formation has already been produced.
For gas-liquid systems involving the presence of different quantities of surfactants in the liquid phase, a previous work has developed a model that shows a good behavior in the mass transfer coefficient determination and, taking into account the special changes produced by this kind of substance, upon the dynamics of gas-liquid systems as well. This model predicts, under the experimental conditions employed in the present work, that the mass transfer coefficient must be included between two limits called: : mass transfer coefficient for free surface (Se = 0) and : mass transfer coefficient for a saturated surface (Se = 1). On the basis of these conditions, this model for mass transfer coefficient estimation has the expression shown in equation 8.
Determining under present experimental conditions could be carried out using Higbie's equation . On the other hand, the corresponding value to a saturated interface, , has additional difficulties since there are not models in the literature that allow us the calculation of this coefficient. Frössling equation in particular can not be used directly for this calculation. This value depends on the surfactant nature, so Sardeing et al  suggest the use of equation 9 for this mass transfer coefficient.
where K is the adsorption equilibrium constant. This constant is very important because high values imply that the surfactant molecules reach the gas-liquid interface quickly. Then, the clean bubbles that are fed to the bubble column accumulate contamination in the surface in a low operation time, producing a decrease in the mass transfer. Equation 9 also includes the mass transfer coefficient determined by the Frössling model .
Figure 8 shows the obtained behaviour for the Higbie, Frössling and Sardeing et al (using equation 9) models. The comparison with the experimental values of mass transfer coefficient (see figure 8) indicates that the Higbie's model overestimate the values of mass transfer coefficient except for the system in the absence of surfactant. On the other hand, Frössling's model takes lower values than the experimental ones. At a high surfactant concentration, the experimental values are closer to the corresponding ones to Frössling's model, due to the increase in the surfactant concentration at the gas-liquid interface. The last model, developed by Sardeing et al, allows the calculation of the mass transfer coefficient with better results when it is compared with the experimental data. But the values contributed by this last model when the surfactant concentration increases are very close to Frössling's model, and the experimental data shows a plateau with a constant value of mass transfer coefficient higher than the value calculated using Frössling equation. Due to the behaviour of Sardeing's model, a modification of this model has been performed in this work, by means of changing the constant (1.744) of equation 9, since this value is related to the surfactant nature . This constant has been determined in the present study using the experimental data of mass transfer coefficient because the surfactant we have employed is very different (in molecular weight and size) to the substances used in previous works that had used this model. Then, equation 9 has been modified, obtaining the expression shown in equation 10. This modification in the Sardeing et al model allows fitting, with better results, the influence of Tween80 concentration upon the experimental values of mass transfer coefficient, in comparison with the other models analysed (see figure 8).
The present work has analyzed the effect caused by the presence of Tween80 upon different parameters (gas hold-up, bubble size distribution, interfacial area and mass transfer coefficient) related with mass transfer rate in a bubble column contactor. The presence of this substance produces an important increase in gas-liquid interfacial area produced in the bubble column, and caused by an important decrease in the value of bubble diameter, because non-influence of Tween80 upon gas hold-up was detected. On the other hand, an increase in the gas flow-rate produces also an increase in interfacial area due to an increase in gas hold-up.
The presence of Tween80 produces the opposite behavior upon mass transfer coefficient producing a high decrease with the presence of low surfactant concentration. The effect of gas flow-rate upon mass transfer coefficient was considered negligible.
A modification of Sardeing model allows fit the experimental data taken into account the values corresponding to mobile a rigid bubbles, and the special characteristics (in relation with its surface activity) of Tween80.