8420
Ind. Eng. Chem. Res. 2007, 46, 8420-8430
Liquid-Liquid Slug Flow in a Capillary: An Alternative to Suspended Drop or Film Contactors M. N. Kashid, Y. M. Harshe, and D. W. Agar* Institute of Reaction Engineering, UniVersity of Dortmund, Emil-Figge Strasse 66, 44227 Dortmund, Germany
The widely used separation process of liquid-liquid extraction is performed in a variety of contactors. The interfacial area in these conventional contactors is often poorly defined, because of the complex hydrodynamics involved, and the intensity of mass transfer is limited by the constraints imposed by the underlying buoyancy or gravitational effects being exploited. Similar shortcomings are apparent in most of the laboratory equipment presently used for investigating extraction and biphasic reactions. In the present work, a new contactor concept, liquid-liquid slug flow in a capillary, is presented as an alternative to conventional equipment. Experiments were performed to investigate the effect of operating conditions on mass-transfer coefficients for different nonreacting systems. In addition, a flow splitter was developed for downstream separation of two liquid phases. The combination of this flow splitter with the capillary provides miniature mixer-settler modules, which can be networked in a wide variety of configurations. Finally, the results obtained were compared with the literature data, and it was determined that such a microextractor-reactor offers superior performance and greater efficiency, in comparison to conventional equipment for liquid-liquid extraction. The results also show that such equipment can be exploited to enhance mass-transfer- and heat-transfer-limited liquidphase reactions. 1. Introduction
Table 1. Advantages and Limitations of Different Types of Contactors
Liquid-liquid extraction is a widely used separation process in the laboratory and chemical, petroleum, pharmaceutical, hydrometallurgical, and food industries. This is mainly because of its cost-effectiveness, in comparison with alternative separations processes. Extraction involves three basic steps: (i) mixing of the two liquid phases, (ii) maintaining the droplets or films of the dispersed phase, and (iii) subsequently separating the two phases from each other. Further solvent recovery and raffinate clean-up operations may also be required. The performance of an extraction process is dependent on many factors, such as selection of solvent, operating conditions, mode of operation, extractor type, and an assortment of design criteria. 1.1. Conventional Contactors. A wide variety of extraction equipment is available: mixer-settlers, centrifugal extractors, and columns. A mixer-settler consists of a mixer (agitated tank) in which the aqueous and organic liquids are contacted, followed by gravity separation in a shallow basin called a settler, where the liquids disengage into individual layers and are discharged separately. In centrifugal extractors, the two immiscible liquids of different densities are rapidly mixed in the annular space between a rotor and the stationary housing. The separation efficiency is very much higher in the centrifugal extractor than in a mixer-settler. The third class of extraction equipment, columns, is usually used industrially in its countercurrent mode. Columns may be divided into two types: static columns (e.g., spray column, sieve plate column, and packed column) and agitated columns (e.g., rotating disk contactor, Scheibel column, Kuhni column, Karr column, and pulsed column). The advantages and limitations of these contactors are summarized in Table 1. A common drawback of conventional equipment is the inability to condition the drop size precisely and the non* To whom correspondence should be addressed. Tel.: +49 231 755 2697. Fax: +49 231 755 2698. E-mail address:
[email protected].
advantages
disadvantages
Mixer Settler simplicity and low maintenance high investment costs low number of stages required high operation costs good contacting design from first principle is not feasible Centrifugal Contactor works at low density difference difficult to scale-up between two liquids low solvent volume required mechanical complexity and high maintenance cost rapid mixing and separation can enhance product recovery and quality Static Columns performance completely relies on the packing/internals satisfactory performance good performance in limited at lower cost range of flow rates easy to operate
Agitated Columns difficult to provide right amount of mixing better performance due to difficult to separate small decrease in the mass-transfer density differences resistance with increased good performance for limited agitation flow ratio low investment cost
uniformities that result because of the complexities of the underlying hydrodynamics. This leads to uncertainties in extractor design and often imposes severe limitations on the optimal performance that can be achieved. 1.2. Process IntensificationsMicroextractors. Process intensification is often summed up in the mnemonic “safer, cleaner, and smaller”. Various techniques are used for process intensification in the chemical process industries, such as miniaturization, dynamic operation, special reaction media, and the use of non-conventional energy sources, to name but a few. In particular, the reduction of characteristic plant dimensions in microreactors offers a powerful tool for overcoming bottle-
10.1021/ie070077x CCC: $37.00 © 2007 American Chemical Society Published on Web 11/13/2007
Ind. Eng. Chem. Res., Vol. 46, No. 25, 2007 8421 Table 2. Different Types of Microextractors contactors partially overlapping channels [2] wedge-shaped flow contactors [3] microchannel separated by a micromachined membrane [4] microchannels separated by a sievelike walls[1] micromixer-settler systems [1]
systems studied xylene-Fe(III)-water octan-1-ol-phenol-water water-kerosene (visual study) and nitration of benzene water-cyclohexanol-cyclohexane
water-acetone-toluene water-acetone-n-butyl acetate water-succinic acid-n-butanol water-D,L-panto-lactone-methyl-i-butyl-ketone
necks in heat and mass transfer. The well-defined flow patterns and temperature conditions in microstructured systems result in an extremely uniform processing environment. While microscale processes are fundamentally unsuitable for some unit operations, such as distillation, because of their inherent suppression of temperature gradients, extraction can benefit considerably. Some of the microextractors which have been developed and successfully tested in the laboratory are listed in Table 2. A detailed review of such contactors can be found in Ehrfeld et al.1 It can be seen that a variety of microscale approaches have been evaluated as alternatives to conventional extraction equipment. In this work, we seek to demonstrate that liquid-liquid slug flow in a capillary, together with a wettability-based flow splitter, is superior to both conventional contactors and the microscale options previously proposed. Liquid-liquid slug flow in a capillary microreactor provides two independent and tunable transport mechanisms: convection through the internal circulations within each slug and diffusion between two adjacent slugs, as depicted in Figure 1. The shear between capillary walls and slug axis generates an intense internal circulation (Taylor-like vortices) within the slug, which, in turn, reduces the thickness of interfacial boundary layer and thereby augments the diffusive penetration. Furthermore, the low residence times and hold-ups of such microscale extraction processes make them an attractive alternative, both technically and economically, to conventional arrangements. For large-scale production, numbering-up, rather than scale-up, reduces the risk involved when transferring the technology from the laboratory. Temperature profiling during the course of extraction, which is difficult in conventional system, is facilitated at the microscale, because of the high specific surface areas. The potential for enhancing the performance of rapid mass-transfer-limited liquid-liquid reactions using such slug flow capillary microreactors has already been demonstrated (Burns and Ramshaw5 and Dummann et al.6). The high interfacial area and intensity of internal circulation have important roles in dictating masstransfer rates. Both are governed by the slug geometry, which can be varied in a controllable manner by adjusting the operating conditions. 1.3. Literature Review. Since extraction is a well-established process, many studies, research papers, review articles, and books have been published on the hydrodynamics, mass transfer, and biphasic reaction in different contactors. Recently, for example, Dehkordi has published a series of articles on impinging stream contactors and compared their superior masstransfer performance with conventional contactors for various liquid-liquid systems.7-9 When one considers the attention that has been lavished on gas-liquid microreactors, the paucity of publications on microextractor technology is somewhat surprising. The liquidliquid slug flow regime, on the other hand, which is alternatively
developed/used at Central Research Laboratories (CRL), Middlesex, U.K. University of Newcastle and British Nuclear Fuels, Preston, U.K. Pacific Northwest National Laboratory, Richland, WA Institut fu¨r Mikrotechnik, Mainz, Germany Institut fu¨r Mikrotechnik, Mainz and BASF, Ludwigshafen, Germany
referred to as segmented or liquid train flow, has already been the subject of extensive research. Hodgson and Charles10 studied slug flow for its use in the transportation of oil, with water as the carrier phase. They conducted experiments in a 1-in.diameter horizontal pipe and reported flow regimes, slug velocities, and pressure gradients for different operating conditions. Furthermore, measurements of the slug flow velocity and pressure gradient for laminar and turbulent flow conditions were described by a model that was developed by Charles.11 Three decades later, this flow scheme was exploited for reaction engineering applications by Professor Ramshaw’s group at the University of Newcastle. They developed a multiphase microreactor based on the use of liquid-liquid slug flow and obtained mass-transfer performance data for the extraction of acetic acid from kerosene slugs.5 A numerical model was later developed to describe the internal flow patterns within the fluid segments12 and the transfer of a dissolved chemical species within and across the segments for liquid-liquid slug flow was simulated. The model was validated with various experimental results and yielded good predictions of the flow field and mass transfer. During the same period, Dummann et al.6 performed experiments for the production of nitroaromatics in a capillary microreactor using the same concept. The reaction was performed in the slug-flow regime, and the work focused on reducing the formation of by-products. Some computational fluid dynamics (CFD) simulations were performed, which led to the conclusion that the enhancement of mass transfer can be interpreted in terms of an intense internal circulation flow within the individual slugs. In our previous studies,13 we have characterized internal circulations in liquid-liquid slug flow using simplified CFD simulations and visualized them with the help of PIV experiments and CFD particle tracing. The results from both these visualization techniques exhibited good qualitative agreement. The mechanism of slug-flow generation was studied with the help of free surface calculations.14 Furthermore, the extent of the slug-flow regime, the slug size, and the pressure drop were measured experimentally,15 and a model was developed to characterize mass transfer with and without chemical reaction.16 Recently, Tanthapanichakoon et al.17 conducted two-dimensional (2D) and three-dimensional (3D) simulations to study the mixing behavior in a liquid slug of liquid-liquid slug flow and have proposed a modified Peclet number for calculating the mixing in the liquid slugs. 1.4. Present Work. Previous work has, thus, clearly demonstrated the potential of microreactor technology and shown that the flow patterns that arise in the slug-flow capillary extractor/microreactor can be described very well using CFD tools, even down to the slight complications that result because of the presence of a thin film of the better-wetting liquid on the capillary wall. An a priori prediction of mass transfer in such
8422
Ind. Eng. Chem. Res., Vol. 46, No. 25, 2007
Figure 1. Experimental snapshot of liquid-liquid slug flow and a schematic illustration of slug flow showing internal circulation within each slug and interfacial diffusion between two adjacent slugs.
Figure 3. Flow splitter showing splitting of two phases: an aqueous phase (water + brilliant blue dye) and an organic phase (n-butyl formate).
Figure 2. Diagram showing the experimental setup. (1,2) Reservoirs containing aqueous and organic liquids, respectively; (3,4) feed pumps; (5) Y-junction mixing element; (6) slug-flow capillary; (7) flow splitter; and (8,9) sampling bottles.
systems becomes feasible. In the present work, the liquid-liquid extraction of three non-reactive systems was considered: iodine from its aqueous solution into kerosene (system I), succinic acid from its aqueous solution into n-butanol (system II), and acetic acid from kerosene into distilled water (system III). Similar systems were used in the impinging streams contactor by Dehkordi.7 The objective was to compare the performance of liquid-liquid slug flow with the impinging stream contactor and thus other contactors. In addition, a microscale flow splitter based on the wetting properties of the liquid on solid walls was developed to separate the phases following extraction. The effect of various operating conditions such as flow rate, capillary diameter, and flow ratio was studied. The power requirement of the system was investigated for one of the non-reacting systems. Finally, the values of the mass-transfer coefficients and power input were compared with those for conventional contactors. 2. Experimentation 2.1. Experimental Setup. The experimental setup used in the present work is depicted schematically in Figure 2. The figure illustrates that the two immiscible liquids (aqueous and organic, from reservoirs 1 and 2 in Figure 2) are introduced by continuously operating high-precision piston pumps (3 and 4 in Figure 2) to a symmetric 120° Y-piece mixing element made of Teflon (5 in Figure 2). The pumps (with a floe range of 1-999 mL/h each) are controlled with the help of a computer, which permits precise tuning of the flow rate down to 1 mL/h. The capillary contactor (6 in Figure 2) is made of polytetrafluoroethylene (PTFE) and was attached directly downstream of the Y-piece. The length of the capillary microreactor can be adjusted easily; thus, the setup was very flexible with respect to the residence time used. At the end of the contacting stage, a flow splitter (7 in Figure 2), which is discussed in detail in the next section, was used to separate the phases. The samples from the two outlets of the splitter were collected in the sampling bottles (8 and 9 in Figure 2) for further analysis.
2.2. Flow Splitter. We have developed a Y-piece splitting element with one inlet and two outlets to separate the two immiscible liquid phases (aqueous and organic), as illustrated in the photograph in Figure 3 and in the schematic shown in Figure 4. The splitter was comprised of Teflon with a steel needle having an internal diameter equal to the Y-junction internal diameter, being fitted into one of the outlets. The splitter works on the principle of preferential wettability of a liquid on a solid material. The aqueous phase has strong affinity toward steel, whereas the organic phase has affinity toward Teflon. This difference in the affinity can be harnessed for the separation of the two phases. The technique is especially suitable for use in microscale equipment, in which surface tension forces, rather that gravitational body forces, dominate. Furthermore, excellent separation can be achieved over a wide range of flow rates and phase ratios. 2.3. Operating Conditions and Measurements. The operating conditions used for the experiments are given in Table 3. Various capillary sizes (diameters of 0.5, 0.75, and 1 mm) were used with the corresponding Y-junction dimensions. All the experiments were performed under the slug-flow regime with both equal and unequal flow ratios of the two phases. For each set of operating conditions, three samples were taken from each outlet stream of the splitter, to ensure the repeatability of the results. Quantitative analysis of the samples was performed by titration, using a standard sodium hydroxide (NaOH) as a titrant for succinic acid and acetic acid samples and sodium thiosulfate for iodine samples. The samples from each set of experiments were analyzed and a mean value was recorded. The accuracy of the analytical methods was checked using known calibration samples. Before taking final measurements, several orientation experiments were performed to identify the appropriate lengths and residence times. Two sets of measurements were taken for each system: varying the flow rate at constant capillary length (i.e., varying residence time) and varying the flow rates in parallel with the length of the capillary (i.e., at constant residence time). Various capillary and Y-junction sizes were used at different flow velocities and phase ratios, to examine the dependence of microextractor performance on the operating conditions. The interfacial area was estimated using a snapshot technique.15 The interfacial area thus determined was used to calculate the mass-transfer coefficients under all operating conditions. Two parameterssthe extraction efficiency and the mass-transfer coefficientswere applied to evaluate the performance of the capillary contactor. The flow splitting experiments were performed separately with pure liquid phases. The operating parameters and conditions
Ind. Eng. Chem. Res., Vol. 46, No. 25, 2007 8423
Figure 4. Flow splitters used in the present study. Table 3. Parameters and Operating Conditions parameter
value/comment Mass Transfer
systems initial concentration of succinic acid in aqueous solution initial concentration of iodine in aqueous solution initial concentration of acetic acid in kerosene flow rate, each phase capillary ID Y-junction ID capillary length
I, II, and III 10 kg/m3 0.94 kg/m3 2 kg/m3 5-60 mL/h 0.5, 0.75, and 1 mm 0.5, 0.75, and 1 mm 0.1-1 m
Flow Splitting systems
water-kerosene water-n-butanol 5-60 mL/h 0.5, 0.75, and 1 mm 0.5, 0.75, and 1 mm 100 mm
flow rate, each phase capillary ID Y-junction ID capillary length
Figure 5. Extraction efficiency as a function of slug-flow velocity for different capillaries in the extraction of iodine from its aqueous solution into kerosene (capillary length ) 100 mm).
Table 4. Physical Properties of the Liquids Used liquid
density [kg/m3]
viscosity [N/(m2 s)]
surface tension with water [N/m]
water n-butanol kerosene
1000 780 810
0.001 0.003 0.0022
∼0.05 ∼0.05
used for the experiments are shown in Table 3, and the physical properties of the liquids used are given in Table 4. Various capillary and Y-splitter sizes were used for the experiments. For a given flow rate, the liquid from each outlet was collected over a certain time and then the fraction of the total flow in the two outlets (steel and Teflon) and its composition (aqueous and organic phase) were determined. 3. Results and Discussion 3.1. Extraction Efficiency. The extraction efficiency is the ratio of the amount of material transferred to the maximum amount transferable. For a solute transferring from one phase to another, the extraction efficiency can be written as follows:
E)
C2,out - C2,in C2,sat - C2,in
(1)
where C2,in, C2,out, and C2,sat are the inlet, outlet, and saturation concentrations of any species in phase 2 (solvent), respectively. The extraction efficiency for the extraction of iodine from its aqueous solution into kerosene, the extraction of succinic acid from its aqueous solution into n-butanol, and the extraction of acetic acid from kerosene into water for three different capillary diameters is plotted in Figures 5-7. For systems I and III, the capillary length was 100 mm, whereas for system II, the capillary length was 300 mm. The results show that an extraction efficiency of >90% can be achieved in 40 mL/h) (see Figure 21). This is due to the fact that, at high flow rates with unequal flows, the wetting forces for the lower flow rate liquid (e.g., the aqueous phase in the steel outlet branch) are overridden by the inertial flow forces and therefore, the lower-flow-rate outlet of the splitter receives part of the highflow-rate liquid.
Ind. Eng. Chem. Res., Vol. 46, No. 25, 2007 8429
Figure 23. Volume fraction in steel and Teflon outlets for various capillaries (flow ratio ) 1).
Table 7. Volumetric Mass-Transfer Coefficient, Specific Interfacial Area, and Mass-Transfer Coefficient for Different Capillary Internal Diameters (ID) at Equal Flow Rates of Both Phases for the Cases with and without Film in the Liquid-Liquid Slug-Flow Capillary Microreactor (Water-Iodine-Kerosene) Specific Interfacial Area, a [m2/m3]
Mass-Transfer Coefficient, kL [× 10-4 m/s]
ID [mm]
volumetric mass-transfer coefficient, kLa [× 10-4 1/s]
without film
with film
without film
with film
0.5 0.75 1
0.31-0.98 0.29-0.64 0.13-0.32
1080-1970 870-1686 590-780
4500-4800 2980-3190 2510-2760
2.75-6.29 4.70-7.90 2-4.64
0.66-2 0.98-2 0.5-1.2
4.3. Effect of Capillary Size. The effect of capillary internal diameter on the flow splitting is shown in Figure 22. It can be observed that the capillary diameter has no significant influence on the degree of flow splitting. For all capillaries studied, more than half of the flow goes to the steel outlet branch (∼5%). For the capillary with an internal diameter of 0.75 mm, the fraction of the organic phase in the steel was determined to be as high as 12% (see Figure 23). However, in all cases, the Teflon outlet yielded a pure organic phase, except at very low flow rates for the capillary with an internal diameter of 0.5 mm. The minor problems observed with phase cross-contamination could be resolved by cascading additional splitters downstream of the initial one. 5. Conclusions and Future Work Experiments were performed to investigate the mass-transfer coefficients in the liquid-liquid slug-flow capillary microreactor for three different non-reacting chemical systems. The masstransfer coefficients were compared with those of conventional contactors, and it was observed that such liquid-liquid microextractor-reactors offer superior performance and greater efficiency in comparison to conventional equipment. The capillary microreactor provides very large specific interfacial areas in comparison with other contactors, which enhances the mass-transfer rates. Moreover, the internal circulations in the slugs, which occur because of the shear between the slug axis and the continuous phase film or capillary wall, improve the mass-transfer rate by surface renewal at the phase interface. The ability to (i) tune the mass-transfer contributions precisely, (ii) impose temperature profiles, (iii) monitor the progress of extraction and reaction, together with the excellent accessibility for process modeling, and (iv) separate the downstream liquids easily suggests that the liquid-liquid slug flow in capillaries is a powerful laboratory technique for elucidating extraction and biphasic reactions. The results reported in the present work can also be used to identify the asymptotic performance limits in such systems, by permitting a precise evaluation of the masstransfer effects. Future work will be devoted to exploiting the newly developed flow splitter to produce a microscale cascade for countercurrent extraction and to extending the CFD models to include interface curvature and wall film behavior, to provide
a reliable a priori prediction of concentration and flow fields, which should permit the liquid-liquid slug flow capillary contactor to fulfill its potential as a pwoerful laboratory tool for studying extraction and biphasic reaction. Acknowledgment The authors acknowledge the assistance of Mr. Rohan Karande in performing the experiments described. Literature Cited (1) Ehrfeld, W.; Hessel, V.; Lo¨we, H. Microreactors; Wiley-VCH Verlag GmbH: Weinheim, Germany, 2000. (2) Robins, I.; Shaw, J.; Miller, B.; Turner, C.; Harper, M. Solute transfer by liquid/liquid exchange without mixing in micro-contactor devices. In Microreaction Technology; Ehrfeld, W., Ed.; Proceedings of the 1st International Conference on Microreaction Technology (IMRET 1); SpringerVerlag: Berlin, 1997; p 35. (3) Burns, J. R.; Ramshaw, C. Developement of a microreactor for chemical production. Trans. Inst. Chem. Eng. 1999, 77 (A), 206211. (4) TeGrotenhuis, W. E; Cameron, R. J; Butcher, M. G; Martin, P. M; Wegeng, R. S. Microchannel devices for efficient contact of liquids in solvent extraction. In Process Miniaturization: 2nd International Conference on Microreaction Technology; Tropical Conference Preprints; Ehrfeld, W., Rinard, I. H., Wegeng, R. S., Eds.; AIChE: New Orleans, LA, 1998; p 329. (5) Burns, J. R; Ramshaw, C. The intensification of rapid reactions in multiphase systems using slug flow in capillaries. Lab Chip 2001, 1, 10. (6) Dummann, G.; Quittmann, U.; Groschel, L.; Agar, D. W; Worz, O. Morgenschweis, K. The capillary microreactor: a new reactor concept for the intensification of heat and mass transfer in liquid-liquid reactions. Catal. Today 2003, 79-80, 433. (7) Dehkordi, A. M. Novel type of impinging streams contactor for liquid-liquid extraction. Ind. Eng. Chem. Res. 2001, 40, 681. (8) Dehkordi, A. M. Liquid-liquid extraction with an interphase chemical reaction in an air-driven two-impining-streams reactor: effective interfacial area and overall mass transfer coefficient. Ind. Eng. Chem. Res. 2002, 41, 4085. (9) Dehkordi, A. M. Experimental investigation of an air-operated twoimpinging-streams reactor for copper extraction processes. Ind. Eng. Chem. Res. 2002, 41, 2512. (10) Hodgson, G. W.; Charles, M. E. The pipeline flow of capsules, Part 1: Theoretical analysis of the concentric flow of cylindrical forms. Can. J. Chem. Eng. 1963, 43-45. (11) Charles, M. E. The pipeline flow of capsules, Part 2: The concept of capsule pipelining. Can. J. Chem. Eng. 1963, 46.
8430
Ind. Eng. Chem. Res., Vol. 46, No. 25, 2007
(12) Harries, N.; Burns, J. R.; Barrow, D. A.; Ramshaw, C. A numerical model for segmented flow in a microreactor. Int. J. Heat Mass Transfer 2003, 46, 3313. (13) Kashid, M. N.; Gerlach, I.; Goetz, S.; Franzke, J.; Acker, J. F.; Platte, F.; Agar, D. W.; Turek, S. Internal circulation within the liquid slugs of liquid-liquid slug flow capillary microreactor. Ind. Eng. Chem. Res. 2005, 44 (14), 5003. (14) Kashid, M. N.; Platte, F.; Agar, D. W.; Turek, S. Computational modelling of slug flow in a capillary microreactor. J. Comput. Appl. Math. 2007, 203 (2), 487. (15) Kashid, M. N.; Agar, D. W. Hydrodynamics of liquid-liquid slug flow: flow regimes, slug size and pressure drop. Chem. Eng. J. 2007, 131 (1-3), 1. (16) Kashid, M. N.; Agar, D. W.; Turek, S. CFD modelling of mass transfer with and without chemical reaction in a liquid-liquid slug flow capillary microreactor. Chem. Eng. Sci. 2007, 62 (18-20), 5102.
(17) Tanthapanichakoon, W.; Aoki, N.; Matsuyama, K.; Mae, K. Design of mixing in microfluidic liquid slugs based on a new dimensionless number for precise reaction and mixing operations. Chem. Eng. Sci. 2006, 61, 4220. (18) Bico, J.; Quere, D. Liquid trains in a tube. Europhys. Lett. 2000, 51 (5), 546. (19) Hoettges, K. F.; Stevenson, D.; Homewood, K. P.; Gwilliam, R. M. Liquid-liquid separation. International Patent WO 03/082429 A2, 2003.
ReceiVed for reView January 12, 2007 ReVised manuscript receiVed October 10, 2007 Accepted October 11, 2007 IE070077X
