Characterisation of a high concentration ionic bubble column using electrical resistance tomography

Flow Measurement and Instrumentation 31 (2013) 69–76

Contents lists available at SciVerse ScienceDirect

Flow Measurement and Instrumentation journal homepage: www.elsevier.com/locate/flowmeasinst

Characterisation of a high concentration ionic bubble column using electrical resistance tomography A.D. Okonkwo a,n, M. Wang a, B. Azzopardi b a b

School of Process, Environmental and Materials Engineering, University of Leeds, UK Process and Environmental Engineering Research Division, Faculty of Engineering, University of Nottingham, UK

a r t i c l e i n f o

abstract

Available online 26 October 2012

Attentions has been given to ionic liquids as an alternative physical solvent for carbon dioxide (CO2) absorption because of their potential for gas selectivity, absorption capacity and low desorption energy by tailoring the molecules. Ionic liquid normally have a high viscosity, which influences the performance of absorption processes, and therefore, efficiency. This study investigates the hydrodynamics of ionic liquids in a two-phase gas–liquid flow by determination of the bubble formation, distribution of gas and bubble velocity profiles. A dual plane electrical resistance tomography (ERT) system and an optical imaging device were applied to a bubble column reactor of 50 mm internal diameter for the study. The model ionic liquids were aqueous solutions of sodium chloride (NaCl) with conductivity adjusted by altering the concentration of NaCl. Gas holdup has been estimated by analyses of conductivity data obtained from ERT by application of Maxwell’s relationship which reveals significant increase in gas holdup as ionic concentration increases and is in good agreement with other studies. & 2012 Elsevier Ltd. All rights reserved.

Keywords: Ionic liquids Gas holdup ERT CO2 absorption and bubble column

1. Introduction Bubble column reactors are widely used as multiphase gas– liquid contacting devices. Some of its advantages include the ease by which they can be operated and controlled, low operating cost and high energy efficiency. Bubble column reactors have several industrial applications which are often applied to determine the hydrodynamics of bubble swarm characters such as gas holdup; bubble rise velocity; flow regimes identification or patterns and characteristics of flow structure occurring in various gas/liquid reactions. Several types of bubble column reactors commonly used today are shown in Fig. 1. They include: (a) simple bubble column; (b) cascade bubble column with sieve tray; (c) packed bubble column; (d) multi-shaft bubble column; and (e) bubble column with static mixers. The study [1] of bubble swarm characters are important for the determination of how they affect the specific gas–liquid interfacial area, residence time distribution, mass transfer rates and rates of reactions in most chemical processes. Accurate estimation of bubble size distribution is very pertinent in determination of hydrodynamics which has been met by the technique of Dynamic Gas Disengagement (DGD), first

n

Corresponding author. Tel.: þ44 7863793544. E-mail addresses: [email protected] (A.D. Okonkwo), [email protected] (M. Wang). 0955-5986/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.flowmeasinst.2012.10.005

introduced by Sriram and Mann [2] based on the disengagement rates of gas holdup and level of gas–liquid dispersion after the gas flow to the bubble column was shut off. The DGD technique is widely adopted in the study of gas holdup, bubble size distribution and bubble rise velocity profiles as shown by Wang et al. [3], pp. 459–464). Knowledge of the behaviour of such systems is imperative in improving the process efficiency of its industrial process applications which by nature are very complex [4]. Bubble column reactors are commonly applied in chemical processes involving several reactions such as oxidation, chlorination, polymerisation and hydrogenation. They are extensively used in the manufacture of synthetic fuels by gas conversion processes and in biochemical processes such as fermentation and biological wastewater treatment [5,6]. A very broad chemical application of bubble column is the Fischer–Tropsch process which is the indirect coal liquefaction process to produce transportation fuels, methanol synthesis and manufacture of other synthetic fuels which are environmentally friendly over petroleum derived fuels [7]. Further industrial applications of bubble column reactors include catalytic reactions, coal liquefaction and bio reactions. Liquids which are salts below 100 1C are generally classified as ionic liquids i.e. solvents which are often fluids at room temperature while consisting entirely of ionic species. Ionic liquids are generally made of ions and short lived ion-pairs which are sometimes referred to as ionic melts, liquid electrolytes, ionic

70

A.D. Okonkwo et al. / Flow Measurement and Instrumentation 31 (2013) 69–76

fluids, fused salts etc. Common examples of ionic liquids include 1,3-dialkylimidazolium chloride, ethyl ammonium nitrate, and 1-ethyl-3-methylimidazolium. Studies by Freemantle [9] reveal the primary driving forces behind various research works into ionic liquids which are the perceived benefits of replacing traditional industrial solvents most of which are volatile organic compounds with non-volatile ionic liquids. This would enhance reduction in the emission of volatile organic compounds which constitutes major sources of environmental pollution. Room temperature ionic liquids possess several advantages in an industrially relevant catalytic process over other solvents largely due to the lack of detectable vapour pressure, and thereby do not contribute to the volatile organic compound emissions into the atmosphere as described by Seddon [10]. There are presently very few industrial applications of these classes of liquids which include inorganic; organic and catalytic synthesis; biochemistry; electrochemistry; analytical chemistry; chemical engineering; and material sciences. Some specific applications include cellulose processing; dispersants; gas handling; gas treatment; nuclear industry; solar energy; high purity organometallic compounds; food and bio-products; waste recycling; batteries

etc. Recent and ongoing researches on ionic liquids (this inclusive) are focused on its suitability as alternative solvents for carbon dioxide (CO2) absorption due to their low vapour pressures [11]. Ionic liquid properties such as thermal stability, solubility of carbon dioxide and selectivity over nitrogen have been investigated by Styring et al. [13] and Tand et al. [12]; results obtained from their studies show high CO2 absorption capacity, selectivity and low solvent loss which potentially make polymers of ionic liquids suitable for carbon dioxide absorption over amine based fluids such as monoethanolamine (MEA) and diethanolamine (DEA). It is worthy to note that the selection of a suitable solvent for CO2 absorption is crucial for economic viability of any process [14]. This arises due to high cost of ionic liquids when large scale industrial applications are considered. The thermodynamics and kinetics of reactions carried out in ionic liquids are markedly different from those in the convectional molecular solvents, making the chemistry of most ionic liquids unpredictable at our current knowledge base. This is in fact, the major reason for this study. The limitation in the application of ionic liquids can largely be attributed to the limited understanding of the classical, physical and thermodynamic properties associated with ionic liquids [15].

Fig. 1. Types of bubble column reactors [8].

Fig. 2. An electrical resistance tomography data acquisition system [17].

A.D. Okonkwo et al. / Flow Measurement and Instrumentation 31 (2013) 69–76

Electrical resistance tomography (ERT) is an online measurement technique for obtaining information about the contents and conditions of a multiphase flow process vessel. ERT is one of the most promising techniques for multi-phase flow measurement due to its low cost, high speed and non-intrusive sensing method. A typical ERT structure is composed of three basic parts: sensors, data acquisition system (DAS) and image reconstruction system/host computer. Unlike electrical capacitance tomography (ECT) systems [16], the sensors in ERT systems must be in continuous electrical contact with

Gas outlet

550mm

150mm

1200mm

200mm

Flow meter

ERT sensors

Gas distributor

400mm

Gas inlet

50mm

Fig. 3. Schematics of the experimental set-up.

71

the electrolyte inside the process vessel. The sensors must be more conductive than the electrolyte in order to obtain reliable measurements. Most process applications use metallic electrodes fabricated from stainless steel, silver, gold, platinum or other suitable materials exhibiting certain properties. Some of the essential properties include low cost; ease of fabrication and installation; good electrical conduction; resistance to corrosion and abrasion. The DAS performs a series of functions such as measurements; de-modulation and control; waveform generation and synchronisation; multiplexer control; and power supply [17], Fig. 2. Data collection strategies are vital as it helps to resolve potentially misleading images for reconstruction. Some of the common data collection strategies suited to a single-currentsource/sink-drive stage includes: the adjacent strategy; the opposite strategy; the diagonal strategy and the conducting boundary strategy. A reconstruction algorithm is applied to determine the internal distribution of resistivities within the process vessel from measurements acquired from an array of electrodes mounted on its periphery. The algorithms reside in the host computer which is connected to the DAS and can be used both on-line and off-line, depending on the type of image required. ERT image reconstruction is an inverse problem to determine the conductivity distribution of an object by measuring the voltage between sets of electrodes placed around its periphery whilst calculating the measurements from a known image is known as the forward problem. In the application of ERT for online measurement, multiphase electrodes are arranged around the boundary of the vessel at fixed locations in such a way that they make electrical contact with the fluid inside the vessel but do not interfere or affect the flow of materials within the vessel. Specific successful applications of ERT include solid/liquid, liquid/gas, packed columns, hydro-cyclones, flotation columns, precipitation processes etc. In principle, ERT can be applied to investigate and monitor any process where the main continuous phase is at least slightly conducting and the other phases and components having differing conductivity values. An examination on the potentialities of an electrical resistance tomography shows that variable gas flow rates enable easy determination of time resolution of ERT in a bubble column reactor [18].

2. Experimental setup

Fig. 4. (a) Experimental setup and (b) enlarged view of ERT sensors.

The experiment has been carried out in a bubble column reactor of 50 mm internal diameter (ID) and 1200 mm high (H) made from transparent acrylic resin as shown in Fig. 3. A thermometer is inserted into the experimental set-up to provide a continuous monitoring of the ionic liquid temperature as temperature variation affects conductivity. The experimental setup was designed to conform to this specification to allow flexibility and maximise operating conditions, Fig. 4. Two planes of ERT sensors located 7 cm apart are uniformly spaced at the centre of the bubble column at planes 1 and 2 and connected to the P2000 electrical resistance tomography system (Industrial Tomography

Fig. 5. Schematic diagram of enlarged view of a gas distributor.

72

A.D. Okonkwo et al. / Flow Measurement and Instrumentation 31 (2013) 69–76

Systems Plc, UK). Each sensor is composed of 16 rectangular electrodes mounted into the inner walls of the column in a noninvasive fashion. The electrodes are made of stainless steel with a contact area of 6 mm (w) and 14 mm (h). The bottom ERT sensor ring is located 450 mm above the gas distributor. Modified Sensitivity Back Projection (MSBP) algorithm [19] has been applied to reconstruct the flow distribution from ERT measurement data. Aqueous solutions of sodium chloride of known concentration were used in the experiment in place of ionic liquids. Sixteen flow conditions were investigated using aqueous NaCl solutions of varying concentrations to obtain different conductivities ranging from 0.387 mS/cm to high ionic concentration of 4.10 mS/cm. The aqueous solution was used as the continuous phase and air as the dispersed phase. For each ionic concentration, conductivity data were collected from the ERT system at three different flow conditions for which superficial gas velocities were in the range of 0.85 cm/s–4.2 cm/s. The ionic liquid under investigation was placed in the column with the gas phase gradually fed into the column through a porous, sintered-glass sparger of 30 mm in diameter (pore index 40–100 mm, grade 2; Fig. 5). Gas superficial velocity, ug expressed as the volumetric flow rate divided by the columns cross-sectional area was estimated using the information obtained from the flow metre between the gas line and the bubble column. The gas superficial velocity, ug for the study was varied from 0.85 cm/s to 4.2 cm/s for each run with flow rate increased continually to obtain the actual effect of difference in gas superficial velocity on hydrodynamics. A fast speed camera was employed for taking high speed videos of the reactor for validation of the experimental results.

conductivity. Conductivity increases with an increase in temperature which is mainly affected by liquid viscosity and the nature of ions involved. The effects of viscosity between ionic liquids and tap water were considered. The experiment was conducted in a semi-batch manner at an ambient temperature of 22 71 1C and atmospheric pressure. Appropriate measures were taken during the experiments to forestall the effects of temperature variation on ionic liquid conductivity. Steady conductivity was achieved by taking references at points T1 and T3 with the actual measurement taken at T2, Fig. 6. A time interval of 20 min was allowed after agitation for entrained gas bubbles to escape from the system. Slight temperature variation detected was analysed using the recorded data for any anomalies’ but shows negligible effects. 2.2. Estimation of gas holdup Maxwell’s relation in Eq. (2.1) which relates conductivity with concentration was applied using the conductivity data obtained from ERT to determine the air concentration or gas holdup of the dispersed phase at different conductivities.

eg,i ¼

2s1 þ s2  2smc  smc s2 =s1 smc  s2 =s1 þ 2ðs1  s2 Þ

ð2:1Þ

where s1 is the conductivity of the first phase and s2 is the conductivity of the second phase. smc is the local value of mixture conductivity distribution. As the dispersed phase in the experiment is a non-conductive material, Eq. (2.1) is thus simplified into Eq. (2.2), where the local mixture conductivity, smc is determined from pixel conductivity of the ERT image. 2s1 2smc 2s1 smc

2.1. Temperature effects

eg,i ¼

The conductivity of ionic solutions which is temperature dependent occurs by means of ionic motion unlike metal

The local mixture conductivity smc was determined from the pixel conductivity of ERT image. As a result, the average value of mixture conductivity over the column cross-sectional area Ac is given by Eq. (2.3): R smc dAc smc ¼ Ac ð2:3Þ Ac

Reference Taken at T1

Measurement Taken at T2

Reference Taken at T3

Fig. 6. Temperature adjustment using reference.

where Ac is the cross-sectional area of the column.

Fig. 7. (a) Homogeneous regimes at 0.85 cm/s, (b) heterogeneous regime at 2.5 cm/s, and (c) slug flow regime at 4.2 cm/s.

ð2:2Þ

A.D. Okonkwo et al. / Flow Measurement and Instrumentation 31 (2013) 69–76

3. Results and discussions 3.1. Estimation of flow regimes In a laboratory/pilot-plant scale experiments where the walls of bubble column reactors are usually transparent, visual observation of the prevailing flow regimes are usually employed. Visual observation gives high degree of accuracy on flow conditions within a bubble column reactor. Flow regimes are easily distinguished based on their individual characteristics enabling easy identification of flow structures in transparent bubble columns. Visual observation was adopted for identification and analysis of

73

prevalent flow structures within the column. The gas was introduced at superficial gas velocities of 0.85 cm/s and 4.2 cm/s. Bubbly flow characterised by small bubbles of similar sizes and velocity dispersed throughout the liquid with little coalescence was observed at 0.85 cm/s superficial gas velocities (Fig. 7a). Increasing the superficial gas velocity to 2.5 cm/s, churnturbulent flow was observed (Fig. 7b and c) mainly characterised by variations in bubble sizes and rise velocities. At 4.2 cm/s superficial gas velocity, large Taylor bubbles were observed which was obviously absent at lower superficial gas velocity of 0.85 cm/s. The presence of Taylor bubbles covering the entire cross-sectional area of the column clearly characterises the flow

Fig. 8. Comparison of gas holdup distributions at different superficial gas velocities. (a) Gas holdup distributions at 0.85 cm/s, (b) gas holdup distributions at 2.5 cm/s and (c) gas holdup distributions at 4.2 cm/s. (For interpretation of the references to colour in this figure, the reader is referred to the web version of this article.)

74

A.D. Okonkwo et al. / Flow Measurement and Instrumentation 31 (2013) 69–76

0.16

0.14 0.85 cm/s 0.12

4.2 cm/s

0.12 0.1

0.08

εg

εg

0.1

0.85 cm/s 2.5 cm/s 4.2 cm/s

0.14

2.5 cm/s

0.06

0.08 0.06

0.04

0.04

0.02

0.02

0

0 -0.5

-0.3

-0.1

0.1

0.3

-0.5

0.5

-0.3

-0.1

r/D

0.1

0.3

0.5

r/D

0.18

0.25

0.85 cm/s

0.16

2.5 cm/s 4.2 cm/s

0.14

0.85 cm/s 2.5 cm/s 4.2 cm/s

0.2 0.15

0.1

εg

εg

0.12

0.08

0.1

0.06 0.04

0.05

0.02 0 -0.5

-0.3

-0.1

0.1

0.3

0.5

0 -0.5

-0.3

-0.1

r/D

0.1

0.3

0.5

r/D

Fig. 9. Comparison of gas holdup at different conductivities using ERT. (a) 100% tap water 0.387 mS/cm, (b) 0.055% NaCl solution, 2.25 mS/cm, (c) 0.07% NaCl solution, 3.30 mS/cm, and (d) 0.09% NaCl solution, 4.07 mS/cm.

structure as slug flow. Evidence of coalescence in ionic liquids was observed between Taylor bubbles which usually burst at the top of the column to form smaller uniform bubble. Another method of flow regime estimation involves the application of Reynolds number which is a dimensionless number. It is a measure of the ratio of inertial forces to viscous forces and is commonly applied to characterise different flow regimes within a process vessel, such as laminar or turbulent flow. Laminar flows occurs at low Reynolds number characterised by smooth, constant fluid motion while turbulent flow occurs at high Reynolds number which tends to produce random eddies and flow instabilities. The Reynolds number of a given system can be quantified using Eq. (3.1): Re ¼

rVD m

ð3:1Þ

where r is the density of fluid (kg/m3), V is the superficial gas velocity (m/s), D is the pipe diameter (m) and m is the dynamic fluid viscosity (cP).

3.2. Cross-sectional gas holdup Fig. 8(a)–(c) shows a cross sectional time-averaged image distribution of gas holdup obtained from ERT at different superficial gas velocities. The result shows increase in air concentration as superficial gas velocity increases. This is represented in different colour regions with the low concentration shown in high blue region to the highest concentration in yellow. It can be observed that the concentration is highest at the centre of the column giving rise to high gas/liquid interfacial area at the centre. Constant colour change is observed as the superficial gas velocity increases, giving rise to maximum holdup at the centre of the column.

3.3. Radial gas holdup The gas holdup profiles in Fig. 9(a)–(d) show effects of varying conductivity on gas holdup, which increases with increase in conductivity and is a strong function of gas superficial velocity. There is an appreciable increase in gas holdup as superficial gas velocity increases resulting in high holdup at the centre of the column and lower holdups at the boundaries. This reveals a high gas/liquid interaction resulting in high mass transfer rates. Comparing results obtained from ERT with that obtained from differential pressure (DP) method under the same operating conditions, it can be concluded that the ionic strength of the ionic liquids under investigation is not sufficient enough to produce a clear impact in the form of concentration variation, resulting in estimation of uniform holdup for all the liquids investigated using DP. This is due to the sensitivity of the resistance field to process conditions and operating temperature giving rise to non-linear dependence on gas holdups, as a result, calibration techniques of ERT are required with steady monitoring of operating conditions [20]. Fig. 10 shows gas holdup profile of ionic liquids of different conductivities at constant superficial gas velocity of 2.5 cm/s. Observations show radial gas expansion with an increase in superficial gas velocity which results in maximum holdup in the centre of the column cross-section. Conductivity effects are clearly shown in the curves which reveal significant increase in gas holdup as conductivity increases. The phenomenon is mainly based on the mechanism of the coalescence suppression behaviour for high ionic solutions of NaCl, as a result, gas holdup increases due to ions reinforcing the liquid film between bubbles against bubble coalescence. The result clearly highlights the effects of conductivity and viscosity on gas holdup, which increases the gas holdup as they increase and is in good agreement with the analysis of Khar and Joshi [21]. The high gas holdup at high conductivity of 4.07 mS/cm shows high gas–liquid interfacial area which may favour its selectivity for CO2 absorption.

A.D. Okonkwo et al. / Flow Measurement and Instrumentation 31 (2013) 69–76

3.4. Velocity profile The velocity profile of the dispersed phase in a two phase gas/ liquid flow in a bubble column reactor is largely dependent on the superficial gas velocity in the column. The velocity profile has been obtained from cross-correlation between the two ERT sensing planes attached to the column. At lower gas velocities, bubbles of uniform sizes and shapes rise along the column with uniform velocity. Fig. 11 shows ionic liquid velocity profile obtained from ERT. In homogeneous flow regimes, bubbles of 0.16 0.387 mS/cm 2.25 mS/cm 3.30 mS/cm 4.07 mS/cm

0.14 0.12

εg

0.1 0.08 0.06 0.04 0.02 0 -0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

r/D Fig. 10. Conductivity variations at constant superficial gas velocity of 2.5 cm/s.

0.25 0.85 cm/s 2.5 cm/s 4.2 cm/s

75

similar sizes and shape rise in the form of swarms distributed uniformly over the column cross-section. Increase in superficial gas velocity causes a significant change in flow regime leading to the formation of larger bubbles or agglomerates in addition to the already existing bubbles. At lower velocity of 0.85 cm/s, homogeneous flow was observed with bubbles of similar sizes moving with uniform velocity throughout the column cross-sectional area possessing a rise velocity of about 0.20 m/s. A decrease in rise velocity was observed when the superficial gas velocity was increased to 2.5 cm/s. This can be attributed to variation in bubble sizes resulting in variation in bubble rise velocity across the column cross sectional area. At 4.2 cm/s, continuous bubble recirculation was observed which is attributable in part to slug flow regime consisting of large bubble sizes filling the entire cross-sectional area of the column. The large bubbles were observed to move with different velocities while splitting into smaller bubbles at the top of the column. This causes a significant decrease in bubble rise velocity. Studies by Hills and Darton [22] have revealed that the velocity of large bubbles rising in vertical columns of liquid is influenced by the concentration of small bubbles in the liquid ahead of it. As a result, decreased velocity is observed at high superficial gas velocities. The curve highlights a decrease in swarm velocity as superficial gas velocity increases unlike in most solvents where large bubbles travel with relative high velocity than small bubbles. This shows the influence of ionic properties on gas–liquid interaction and possible effects of surface tension. Fig. 12 shows the averaged frames of the two sensing planes at 2.5 cm/s superficial gas velocity. The curves show a good correlation between the two sensing planes, P1 and P2, which are in good agreement. The distance between the sensing planes has an adverse effect on the quality of data obtained, with large distances resulting in poor correlations.

Velocity (cm/s)

0.2

4. Conclusion

0.15

0.1

0.05

0 -0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

r/D Fig. 11. Gas phase velocity profile.

0.3

0.4

0.5

Electrical resistance tomography (ERT) has been successfully applied for the investigation of hydrodynamics of various ionic liquids in a bubble column reactor. The basic properties of flow regimes, gas holdup and rise velocity profile have been estimated. Gas holdup has been shown to increase with an increase in ion concentration (or liquid conductivity) resulting in increase in viscosity which are some of the basic properties of ionic liquids. The increase in gas holdup as ionic concentration increases as shown in the experimental results shows great potentials of ionic liquids suitability for carbon dioxide (CO2) absorption for various industrial applications.

Fig. 12. Correlation between the mean values of resistivity at planes 1 and 2 when s1 ¼ 3.30 mS/cm at ug ¼ 2.5 cm/s.

76

A.D. Okonkwo et al. / Flow Measurement and Instrumentation 31 (2013) 69–76

Velocity profile was observed to decrease with an increase in superficial gas velocity due to turbulence or constant recirculation observed at high flow rates. This is due to the effects of surface tension on ionic liquids. Further work is recommended to be conducted on the rise velocity profile of other specific ionic liquids to help improve our current knowledge base. References [1] Jin H, Wang M, Williams RA. Analysis of bubble behaviours in bubble columns using electrical resistance tomography. Chemical Engineering Journal 2007;130:179–85. [2] Sriram K, Mann R. Dynamic Gas Disengagement: a new technique for assessing the behaviour of bubble columns. Chemical Engineering Science 1977;32:571–80. [3] Wang M, Jia X, Bennet M, Williams RA. Bubble column measurement and control using Electrical Resistance Tomography. In: Xie H, Wang Y, Jiang Y, editors. Computer application in the minerals industries. Rotterdam, Netherlands: A. A. Balkema; 2001. p. 459–64. [4] Dudukovic MP. Opaque multiphase reactors: experimentation, modelling and troubleshooting. Oil and Gas Science and Technology 2000;55:135–58. [5] Shia YT, Godbole SP, Deckwer WD. Design parameters estimation for bubble column reactors. AIChE Journal 1982;28:353–79. [6] Krishna R, De Stewart JWA, Ellenberger J, Martina GB, Marreto C. Gas hold-up in slurry bubble columns: effects on column diameter and slurry concentrations. AIChE Journal 1997;43:311–6. [7] Degaleesan S, Dudukovic M, Pan Y. Experimental study of gas induced liquid flow structures in bubble columns. AIChE Journal 2001;47:1913–31. [8] Deckwer WD. Bubble column reactors. John Wiley and Sons; 1992. [9] Freemantle M. Chemical Engineering 1998;76:32. [10] Seddon RK. In: Boghosian S, et al. (Eds.) Proceedings of the international George Papatheodorou symposium. Patras, Greece: Institute of Chemical Engineering and High Temperature Chemical Processes; 1999. p. 131–5.

[11] Styring P, Aschenbrenner O. Comparative study of solvent properties for carbon dioxide absorption. Energy and Environmental Science 2010;3:1101–13, http://dx.doi.org/10.1039/C002915G. www.rsc.org/ees. [12] Tand J, Tang H, Sun W, Radosz M, Shen Y.. Low Pressure CO2 sorption in ammonium based poly(ionic liquid)s. Polymer 2005; 46:12460. [13] Styring P, Aschenbrenner O, Supasitmongkol S. Preprints of Papers— American Chemical Society, Division of Fuel Chemistry 2009;54(1):002. [14] Notz R, Asprion N, Clausen I, Hasse H. Selection and pilot plant tests of new absorbents for post-combustion carbon dioxide capture. Chemical Engineering Research and Design 2007;85(A4):510–5. [15] Kaji R, Zhao D, Azzopardi BJ, Licence P. Studies of the interaction of ionic liquids and gas in a small diameter bubble column. Industrial and Engineering Chemistry Research 2009;48:7938–44. [16] Yang WQ, Xie CG, Gamio JC, Beck MS. In: Beck MS, et al., editors. Design of a capacitance tomography image sensor with uniform electric field process tomography—implementation for industrial processes. Manchester: UMIST publications); 1995. p. 266–75. [17] Fraser D, Wang M. Electrical resistance tomography for process applications. Measurement Science and Technology 1996;7:247–60. [18] Toye D, Fransolet E, Simon D, Crine M, L’homme G, Marchot P. Possibilities and limits of applications of electrical resistance tomography in hydrodynamics of bubble columns. Canadian Journal of Chemical Engineering 2005;83:4–10. [19] Wang M. Inverse solutions for Electrical Impedance Tomography based on conjugate gradient method. Measurement Science and Technology-IOP Science 2000;13:101–17. [20] Jin H, Han Y, Yang S. Gas–liquid flow characterisation in bubble columns with various gas–liquid using electrical resistance tomography. In: Proceedings of the 6th international symposium on measurement techniques for multiphase flows. Journal of Physics: Conference Series 147; 2009. http://dx.doi.org/10. 1088/1742-6596/147/012032. [21] Khar AS, Joshi JB. Effects of fine particles on gas holdup in Three-Phase Sparged Reactor. Chemical Engineering Journal 1990;44:11–25. [22] Hills JH, Darton RC. The rising velocity of a large bubble in a bubble swarm. Transactions of the IChemE 1976;54:258–64.