Xuewu Jia,Weichao Hu,Xigang Yuan*,Kuotsung Yu

State Key Laboratory of Chemical Engineering(Tianjin University),School of Chemical Engineering and Technology,Tianjin University,Tianjin 300072,China

Keywords:

A B S T R A C T

1.Introduction

The gas liquid mass transfer is widely encountered in many industrial processes such as chemical,biochemical,and environmental processes[1-5].In the liquid phase of these processes,surfactants are commonly added as detergent[6],emulsifier[7],and wetting agent[8],and they are also inevitable in wastewater treatment[9]and production of biosurfactants[10].The presence of surfactant has a tremendous effect on gas liquid mass transfer by changing the solubility,mobility and especially surface tension of fluids.In the wastewater treatment,mass transfer is of great economic importance because the oxygen transfer process demands a great amount of electrical energy and requires a significant capital cost.However,mass transfer in the wastewater treatment processes is often affected by the surfactant that is inevitably produced in the course of treatment.

Many research works have been conducted to explore the effect of surfactants on mass transfer in the past few years[11-19].Moraveji et al.[12]investigated the effect of two kinds of surfactant and found that the liquid mass transfer coefficient(kL)could be greatly reduced.Painmanakul et al.[13]investigated the effects of sodium laurylsulfate and lauryl dimethyl benzyl ammonium bromine on the liquid-side mass transfer coefficient and pointed out that the surface coverage ratio(Se)was crucial for predicting the decrease of kLin aqueous solutions with surfactant.Sardeing et al.[14]focused on the effect of surfactants(anionic,cationic and nonionic at concentrations up to 5 × 10-3mol·L-1)on the mass transfer parameters and proposed a model taking into account the dynamic properties of surfactants and the surface coverage ratio to predict the mass transfer coefficient.In their study,the liquid side mass transfer coefficients for all surfactants were significantly smaller than those of pure water.However,Gómez-Díaz et al.[15]showed that the liquid mass transfer coefficient was increased in the presence of very low concentration of decyltrimethylammonium bromide(DTABr).Similar results were also obtained by Alvarez et al.[16]when sodium dodecyl sulfate(SDS)was added in pure water.The inconsistency in the aforementioned results indicates the complexity in the mechanism of the effect of a surfactant on the gas liquid mass transfer.Moreover,there are variety of surfactants,which have been categorized as cationic,anion and nonionic types,and they behave differently[14].However,it has been found that the influence of a surfactant on mass transfer is related closely with the surfactant micelles in the aqueous solution and the critical micelle concentration(CMC)plays a central role[13].So,it is important to reveal such a role,and this motivates the present work.

In the present work,three kinds of surfactant are investigated.They were so selected that they were different in terms of the critical micelle concentration(CMC)as shown in Table 1:n-octyltrimethylammonium bromide(OTABr),a cationic reagent with highest CMC,sodium dodecyl benzene sulfonate(SDBS),an anionic species with medium CMC,and Tween 80,a nonionic surfactant with lowest CMC.Aqueous solutions with low concentrations of the selected surfactants were used in our experiments as solvent to absorb CO2in a bubble column,in order to investigate the effect of surfactant concentration on the gas liquid interfacial area,a,and the liquid mass transfer coefficient,kL.

Table 1 Critical micelle concentration(CMC),adsorption equilibrium constant(K),surface tension at CMC(σCMC)and C/CMof different surfactants

2.Experimental Details

Measurements were carried out at 25°C in a cylindrical bubble column.The experimental setup is schematically shown in Fig.1.The bubble column was made of Plexiglas with an internal diameter of 70 mm and a height of 1000 mm.Outside the cylinder,a square column was installed with taps for thermostat water to maintain the liquid in the bubble column at 25°C.Another function of the square column was to eliminate the light refraction effect caused by the cylindrical column wall in order to ensure that the bubble images can be taken with no distortion.A capillary tube with an inner diameter of 1.5 mm located at the bottom of the bubble column was used to introduce gas bubbles into the column.

Fig.1.Experimental mass transfer apparatus for CO2absorption.(1)bubble column,(2)thermometer,(3)mass flow controller for gas inlet,(4)mass flow controller for gas outlet,(5)humidifiers,(6)gas cylinder,(7,8)inlet and outlet of thermostat liquid,(9)diffuser plate,(10)camera,(11)computer,(12)high power light.

Different quantities of OTABr(CAS No.2083-68-3),SDBS(CAS No.25155-30-0)and Tween 80(CAS No.9005-65-6),supplied by TCI(Tokyo,Japan,purity＞99%),were employed to prepare aqueous solutions(in concentration range of 0-1.4 × 10-3mol·L-1)using ultra pure water(conductivity of 0.2 μS·cm-1).This concentration range was selected to cover the CMC of SDBS,which had a medium CMC,to investigate the influence of surfactants with different CMCs on the mass transfer.The solutions were prepared with a precision of±10-7kg.The surface tension measurements were performed by the video-based optical contact angle measuring instruments(OCA15,Dataphysics Instruments,Germany)based on the pendant drop method.The viscosity of solutions was assumed to be equal to that of pure water,since the concentration was so small[23,24].Commercial grade CO2of 99.98%purity was used in this work as the gas phase.

In the experiment,the gas was saturated with water by passing through two bubbling humidifiers filled with ultra pure water at 25°C.This procedure prevented the water evaporation during the mass transfer in the bubble column and guaranteed that only CO2was transferred into the liquid.So the liquid phase mass transfer coefficient could be measured and considered as the global mass transfer coefficient.The experiment was operated in batches with regard to the liquid phase.The gas flow rate was measured and controlled with two mass flow controllers(Beijing Sevenstar Electronics Co.),which were calibrated by the supplier before the experiments.The gas flow rate has been set at 24 L·h-1.The corresponding gas phase Reynolds number[Reg=4ρgQg/(πd0μg),where Qgis the gas flow rate,d0is the orifice diameter and ρgand μgare the density and viscosity of the gas,respectively]is 634.This value is well within the range of the dynamic bubble regime[25].

TheCO2concentration and its variation with time in the liquid phase was calculated based on the mass flow rate difference between the inlet and outlet of the gas obtained from two mass flow controllers for the gas in the experiment.Then,the volumetric mass transfer coefficient was calculated from.

where kLa is the volumetric mass transfer coefficient;Cgis the concentration of CO2in the liquid at time t,which is assumed to be the same everywhere in the bubble column based on the assumption that the bubble column is well mixed;C*is the saturated concentration of CO2in water.In the present work,C*was considered to be equal to the value in pure water,since the surfactant concentration employed in the liquid phase is low[22].Owing that the liquid phase is homogeneous,and assuming Cg0=0(the CO2concentration in the liquid at t=0),the integration of Eq.(1)leads to.

Then the linear relationship between ln[C*/(C*-Cg)]and t was plotted according to the experimental data as shown in Fig.2,and the value of the volumetric mass transfer coefficient was estimated from the slope.All the experimental kLa data are the averages of measurements repeated three times.

Fig.2.Examples of linear fitting to determine volumetric mass transfer coefficient()Water,()SDBS(8.0 × 10-4mol·L-1)(Correlation coefficient R2=0.989).

The liquid phase mass transfer coefficient was calculated by

The gas-liquid interfacial area can be calculated by using the Sauter mean diameter d32and the gas holdup ε[26]:

The gas holdup was determined by using the volume expansion method[27],taking into account the change of the liquid volume observed after gassing:

The bubble diameter was measured using a photographic method[28]based on the images of bubbles taken from bottom to top of the column.A video camera(Canon 5D II)was used to obtain the images.As the examples shown in Fig.3,the bubbles were assumed to be oblate-ellipsoidal and characterized by the major axis(E)and the minor axis(e).The bubble dimension associated with an equivalent diameter of a sphere with the same volume as the ellipsoid was calculated by.

The Sauter mean diameter was used to get an adequate average diameter for each experiment:

where niis the number of bubbles which have an equivalent diameter di.

3.Results and Discussion

3.1.The influence of surfactant on a

Fig.3 shows the images of bubbles in pure water and in each aqueous solution of OTABr,SDBS and Tween 80,respectively.The presence of surfactant makes the bubbles smaller and more elliptical than the bubbles in pure water,and this effect is more pronounced at higher surfactant concentrations,while the presence of OTABr changes little the dimension of bubbles.

Fig.4 shows the results of the bubble diameter distribution produced in the bubble column.The values of the Sauter mean diameter obtained in this work are in the range of 2-11 mm,which is in agreement with other experimental report[28].The experimental results indicate that when a surfactant is employed in the liquid phase,the average bubble size turns to decrease,as can be seen from Fig.3 qualitatively.

Fig.4.Influence of surfactants on bubble size distribution.()Water,()OTABr(8.0 × 10-4mol·L-1),()SDBS(8.0 × 10-4mol·L-1),(○)Tween 80(8.0 × 10-4mol·L-1).

The decrease of the bubble diameter can be attributed to the decrease of surface tension when surfactants are added to pure water.Fig.5 illustrates the influence of surfactant concentration on surface tension,indicating the impact of low surfactant concentration of different surfactants.The surfactant produces a different effect in surface tension,thus affecting the bubble size produced in the contact or differently.The presence of surfactants on the surface of bubbles may reduce the tendency to coalescence and result in smaller bubbles in the system[29].

Fig.5.Influence of surfactant concentration on surface tension()OTABr;()SDBS;()Tween 80.

Fig.3.The pictures of bubbles in different solutions.(a)Water,(b)OTABr(8.0 × 10-4mol·L-1),(c)SDBS(8.0 × 10-4mol·L-1),(d)Tween 80(8.0 × 10-4mol·L-1).(The unit of ruler is cm.).

Fig.6 reports,for each aqueous solution of OTABr,SDBS and Tween 80,the interfacial area measured for the bubble column.It clearly illustrates that,whatever its nature is,the presence of the surfactants in water noticeably increase the interface area at low concentrations,and the interfacial area tends to be constant when the surfactant concentration is greater than a critical value.OTABr induces a less increase in the interfacial area as compared with SDBS and Tween 80.A similar behavior has been observed in other surfactant aqueous solutions[30,31],which was attributed to the decrease of the bubble diameter.

Fig.6.Influence of surfactant concentrations on gas liquid interfacial areas.()OTABr;()SDBS;()Tween 80.

3.2.The influence of surfactants on kLa

The volumetric mass transfer coefficient,kLa,is calculated for each experimental condition using Eq.(2).An important difference in kLa is generated when small quantities of different surfactants are added to the liquid phase,as can be seen in Fig.7.The addition of OTABr causes no significant changes on kLa,while a continuous decrease is observed when SDBS is added.In particular,a sharp decrease in kLa is generated even when the smallest concentration of Tween 80 is added to the liquid phase,and afterwards kLa reaches a constant value independent of the surfactant concentration.This behavior is opposite to the influence of surfactant concentration on the gas-liquid interfacial area(the increase of surfactant concentration results in an increase in the interfacial area until a constant value is reached).

Fig.7.Influence of surfactant concentrations on volumetric mass transfer coefficient.()OTABr;()SDBS;()Tween 80(error bars are relative errors).

3.3.The influence of surfactants on kL

Fig.8 shows the influence of the surfactant concentration on the liquid phase mass transfer coefficient kL.The variation of kLwith the surfactant concentration is shown to be similar to that on the surface tension and the volumetric mass transfer coefficient kLa,since the variation of interfacial area is rather insensitive to the surfactant type and its concentration.The decrease of kLcan be attributed to the fact that the presence of surfactant molecules in the liquid phase tends to stabilize the liquid surface and to reduce the liquid renewal at the interface[30].Different surfactants decrease kLdifferently.Some studies[31,32]assumed that this difference was related to the value of the critical micelle concentration(CMC).From Table 1,it can be seen that kLhas a significant relationship with the CMC of surfactants.OTABr has the highest CMC,and decreased kLvery weakly,which was different from the result of García-Abuín et al.[20].They found an enhancement of kLat very low concentrations.The mass transfer coefficient kLdecreased continuously with the increase of the SDBS concentration and tended to be a constant when the surfactant concentration approached its CMC.As for Tween 80,kLdecreased sharply even at a very low concentration,and kept at the low value as the concentration of the surfactant increased.It has the smallest CMC according to Table 1.

Fig.8.Influence of surfactant concentration on liquid phase mass transfer coefficient kL.()OTABr;()SDBS;()Tween 80.

The above observations indicate that the effect of any surfactant on mass transfer of CO2into water can be rated by a specific concentration of the surfactant,which is de fined as the ratio of the concentration to the CMC of the surfactant,namely C/CM.The C/CMvalues of the three surfactants in our experiments are given in Table 1.To insure effective mass transfer of CO2into water,the value of C/CMof a surfactant should be as low as possible,and must not exceed 1,as can be clearly seen by comparing Table 1 and Fig.8.

4.Correlation

For gas-liquid systems,a model has been developed for the liquid phase mass transfer coefficient kLin the presence of surfactants[14]:

where kL0is the liquid phase mass transfer coefficient without surfactant,which is calculated using the model developed by Higbie[33].kL1is the liquid phase mass transfer coefficient where the surface is saturated with the surfactant.Sardeing et al.[14]suggested an equation based on the Fr?ssling equation[34]to calculate kL1which incorporated the influence of surfactant characteristics:

where K is the surfactant adsorption equilibrium constant as also shown in Table 1,and kLFis the Fr?ssling coefficient that can be expressed as

In the present work,because the surfactants employed were different in molecular weight and size compared to that used in Sardeing et al.[14],and the concentration range was much lower,the parameter values in Eq.(9)were correlated as below:

Fig.9 illustrates the behavior of the new correlation for the three surfactants in contrast to the Sardeing's original Eq.(9).It is indicated that the new correlation improved significantly the predictability of the model with the average relative deviation within±10%for all three surfactants at low concentrations.

Fig.9.Comparison between experimental and predicted kL.()OTABr,()SDBS and()Tween 80.Solid marks:calculated using Sardeing's model with original parameters;open marks:calculated using new correlation.

5.Conclusions

The effects of low concentration surfactants:OTABr,SDBS and T ween 80 on gas-liquid interfacial area a and liquid phase mass transfer coefficient kLfor CO2absorption into aqueous solutions were experimentally investigated in a bubble column.The presence of these surfactants increased a,because of the decrease of the bubble diameter.The liquid phase mass transfer coefficient kLwas decreased by surfactants in the liquid,and the variations of kLwith the increase of the surfactant concentration behaved differently.It was shown that the variations of kLwith respect to the surfactant concentration depended directly on the CMC of surfactants.OTABr,having the highest CMC,caused little effect on kL;SDBS generated quicker decrease of kLwith the increase of surfactant concentration;and Tween 80 with the lowest CMC caused a strong decrease at very low concentration.It can be concluded that the mass transfer of CO2in water can be mainly deactivated by the micelle phenomenon of the surfactant in water.It is also demonstrated that the improved Sardeing's model can be used effectively to predict the behaviors of kLofCO2absorption in water with the presence of surfactants.

Nomenclature

a gas-liquid interfacial area,m2·m-3

C surfactant concentration,mol·L-1

Cgconcentration of CO2in liquid,mol·L-1

CMthe critical micelle concentration,mol·L-1

C* saturation concentration of CO2in liquid,mol·L-1

D diffusivity of CO2in the liquid,m2·s-1

G acceleration vector of gravity,m·s-2

K adsorption equilibrium constant of the surfactant,L·mol-1

kLliquid phase mass transfer coefficient,m·s-1

kLa volumetric mass transfer coefficient,s-1

kLFliquid phase mass transfer coefficient of saturated surface calculated by Fr?ssling equation,m·s-1

kL0liquid phase mass transfer coefficient of free surface,m·s-1

kL1liquid phase mass transfer coefficient of saturated surface,m·s-1

Sesurface coverage ratio

t time,s

vsslip velocity,m·s-1

υ kinematic viscosity,mm2·s-1

ρ liquid density,kg·m-3