- Email: [email protected]
Determination of phase separation and mass transfer in complex micellar three phase systems
G Model
ARTICLE IN PRESS
CEP-6631; No. of Pages 6
Chemical Engineering and Processing xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Chemical Engineering and Processing: Process Intensification journal homepage: www.elsevier.com/locate/cep
Determination of phase separation and mass transfer in complex micellar three phase systems N. Paul ∗ , J.M. Schulz, M. Kraume Chair of Chemical and Process Engineering, Technische Universität Berlin, Ackerstraße 76, D-13355 Berlin, Germany
a r t i c l e
i n f o
Article history: Received 13 March 2015 Received in revised form 11 June 2015 Accepted 4 July 2015 Available online xxx Keywords: Separation process Liquid/liquid mass transfer Liquid three phase systems
a b s t r a c t Combining the advantages of the homogeneous and heterogeneous catalysis is the main goal of tunable solvent systems. Homogeneous conditions shall be achieved throughout the reaction period and heterogeneous conditions shall be adjusted within the separation process. Therefore, high and specific reaction rates are coupled with a simplified separation process. For the design of the separation process the phase separation itself and the liquid/liquid mass transfer must be quantified. Therefore, a test cell is designed in this work. For the experimental investigations a model test system was applied (water/1-dodecene/C4 E2 ). The separation process of the liquid three phase system was observed by an image analysis. A hysteresis effect was observed which also influenced the separation process. Within the three phase system the coalescence rate is faster than under two phase conditions in the observed temperature range. A diffusion coefficient through the C4 E2 rich middle phase could be determined experimentally, which was in good agreement with theoretical calculations. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Tunable solvent systems can be applied to combine the advantages of the homogeneous and heterogeneous catalysis. Mild reaction conditions as well as high and selective yields are combined with a simplified separation processes. In homogeneous systems mass transfer processes can be neglected. Nevertheless, the catalyst/product separation is the great challenge of these systems. Therefore, the main idea of tunable solvent systems is to provide homogeneous conditions for the reaction and heterogeneous conditions for the separation. Liquid/liquid reaction systems are a special type of the homogeneous catalysis. The separation process is simplified in liquid/liquid systems if product and catalyst are solved in different phases. But due to the additional phase transport phenomena must be taken into consideration. Micellar multiphase systems are examples for tunable solvent systems. These systems fulfill many principles of the green chemistry, e.g., using water as a solvent, reducing waste production, etc. Due to the interfacial active molecules (surfactants) which are used in these systems the phase behaviour is complex [1]. Four different phase conditions are possible for ternary systems of the type
∗ Corresponding author. E-mail address: [email protected] (N. Paul).
water/organic solvent/non-ionic surfactant. These conditions were classified by Winsor [2]: • Type I consists of an organic and aqueous phase; micelles solubilize oil within the aqueous phase. Therefore, an O/W mircoemuslion occurs [3]. • Type II consists of an organic and aqueous phase; inverse micelles solubilize water in the organic phase; therefore, a W/O microemulsion is created [2]. • Type III consists of three coexisting phases an organic phase, an aqueous phase and a surfactant rich middle phase. Low amounts of water and oil are solubilized within the surfactant rich phase, which is also a microemuslion. • Type IV: One phase condition. Within these conditions anisotropic structures can be observed at discrete conditions. In these cases micelles agglomerate to large micelle clusters and form liquid crystalline structures [4]. For a clearer understanding only the Winsor III system is called a microemulsion in this paper. The best performance for some reactions and separation processes were achieved under three phase conditions [5]. For some reactions (e.g., the hydroformylation of long chained alkenes) it is beneficial if the catalyst systems are solved within the microemulsion phase [6]. In case of the hydroformylation syngas must be transported into the microemulsion phase [7]. In this phase all other reactants are present; hence,
http://dx.doi.org/10.1016/j.cep.2015.07.010 0255-2701/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model CEP-6631; No. of Pages 6 2
ARTICLE IN PRESS N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
mass transfer phenomena must be taken into consideration. Furthermore, the distribution coefficient and the mass transfer of the transferred component must be known to design the separation process. After the reaction the products have to be transported back into the organic phase to realize a simplified separation process. Therefore, the phases need to coalesce in a reasonable period of time. In presence of surfactants the coalescence behavior is mostly hindered due to the adsorption of surfactants at the liquid/liquid interface [8]. This work focuses on the determination of fundamental understanding of the liquid/liquid transport processes in liquid three phase systems which is needed for the design of the separation process. Furthermore, the coalescence must be taken into consideration to understand the separation of these complex systems in detail. Often single drop experiments are carried out to understand transport processes [9] and coalescence processes fundamentally [10]. In the presence of non-ionic surfactants an additional mass transfer resistance was observed. With exceeding the critical micelle concentration (CMC) a change of the phase behavior at the liquid/liquid interface was observed. This lead to an increase of the interfacial viscosity and caused a reduction of the mass transfer rates [11,12]. Especially, for micellar three phase systems these additional mass transfer resistances are expected. Under micellar three phase conditions single drop experiments fail due to low interfacial tension. Nevertheless, the mass transfer and the distribution coefficients, etc. must be quantified in order to design the extraction process. Different test cells need to be taken into consideration. Stirred test cells are also commonly used to determine mass transfer rates across liquid/liquid interfaces. Here, two phases are overlaid with each other; hence a defined area of contact arises. The first test cell was presented by Lewis [13]. In the “Lewis test cell” the stirrer speed can only be adjusted in both phases equally. This results in different Reynolds numbers in both phases due to different physical properties (density, viscosity). Hence, the fluid dynamic properties within the phases cannot be chosen equally. An improvement of the “Lewis test cell” was achieved by Nitsch [14,15]. In the “Nitsch test cell” both liquid phases are stirred separately. A broad overview of methods determining the liquid/liquid mass transfer is given in [16]. The test cells which are given in this overview were modified throughout the last
years and have been adjusted to their applications. Lee and Varma [17] used a stirred cell to quantify the kinetics of a biphasic aldol condensation within the mass transfer controlled regime. In this work low stirrer speeds were applied to ensure a stable interface. Another development is shown in the work of Wang et al. [18]; here the extraction kinetics of Thorium were determined within a constant interfacial area test cell. Two stirrers were applied to stir the phases independently. Nevertheless, these stirrers were installed in a cubic test cell and were not installed above each other. They were installed shifted to ensure a stable interface. Datta and Kumar [19] presented a test cell in which two stirrers were installed above each other, but both were attached to one stirrer shaft. Therefore, both stirrers could not be operated independently and further the interface is disturbed. Bertakis et al. [20] optimized the Nitsch test cell based on the Fisher information matrix to improve the quality of the measurements. In this work other requirements on the test cell arise. Due to the low interfacial tension the stirring has to be carried out with great care to neither destabilize the interface nor to create an emulsion, which ensures the quantification of the mass transfer across a defined cross section. Therefore, the influence of the amphiphilic molecules can be identified to optimize the extraction process.
2. Materials and methods 2.1. Experimental setup In Fig. 1 the test cell is shown which was designed to determine the separation process and the mass transfer in liquid three phase systems. The test cell is designed based on the model of the “Nitsch test cell”. It was modified for the challenges in micellar three phase systems in which very low interfacial tensions occur. A double-walled glass reactor (1) with an inner diameter of D = 65 mm and a planar bottom was applied. Therefore, an optical access and an exact determination of the volume was ensured. Two different stirrers (2/3) were installed to avoid concentration and temperature gradients in the aqueous and organic phase. The microemulsion phase was not stirred. Both stirrers were operated independently to provide similar fluid dynamic conditions in each phase. The volume of each
Fig. 1. Experimental setup for the determination of the phase separation and the liquid/liquid mass transfer in micellar three phase systems: 1 double-walled glass reactor; 2 and 3 stirrers; 4 scale; 5 camera; 6a,b,c Hamilton modules PSD/3 with three way valves; 7 Lauda E-200 thermostat; 8 computer.
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model
ARTICLE IN PRESS
CEP-6631; No. of Pages 6
N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
1 mg/L PADA in water pure diluted w. C4E2 (cC4E2 = 0.3 mol/L)
1 mg/L PADA in 1-dodecene pure diluted w. C4E2 (cC4E2 = 0.3 mol/L)
0,15 0.15
0,15 0.15 0.15
λ max= 406 nm
λ max= 470 nm
Absorption A [-]
0,12 0.12
Absorption A [-]
3
0,09 0.09 0,06 0.06 0.03 0,03
0
a)
300
0,12 0.12 0.12 0.09 0,09 0.09 0.06 0,06 0.06 0.03 0,03 0.03
b) 400
500
600
0 300
Wavelength λ [nm]
400
500
600
Wavelength λ [nm]
Fig. 2. Absorption behavior of PADA in water (a) and 1-dodecene (b) in pure systems and in presence of C4 E2 .
phase was observed by a camera (5). Furthermore, the camera was used to determine the separation process. By recording pictures with a camera and an image analyzing software it was possible to determine the heights of the interfaces respectively the volume of each phase. This was the basis for the determination of the separation process. As soon as the volumes of each phase reached a constant value the separation process was treated as completed. For an exact determination a scale (4) was installed. For taking samples automatically and for injecting a specific amount of the tracer substance three different syringe pumps (Hamilton PSD/3-modules) were used. These modules were connected to small pipes, which were adjustable in their heights. Therefore, it was possible to take clean samples out of the organic and the aqueous phase. Furthermore, the syringe pumps were equipped with three-way valves; hence a simple cleaning process and sample taking was provided. The test cell was working semi-automatically and was controlled by a Labview program (8). A Lauda thermostat (E-200) was installed to control the temperature. For the determination of the transport and separation processes in complex micellar three phase systems the test cell shown in Fig. 1 was qualified due to its ability to setup the stirrer speed independently at low stirrer speeds. The optical accessibility allowed the determination of the separation process. Furthermore, the pipes for taking samples could be adjusted to the phase heights to ensure clean probes. The syringe pumps, which were controlled by a Labview program provided a simplified and semi-automatic method to take clean samples of the organic and the aqueous phases.
As transferred component an azo dye pyridine-2-azodimethyl aniline (PADA) was chosen. Hence, the concentration could be quantified by an UV/vis-photometer (Specord 210 by Jena Analytik). PADA changes the absorption behavior due to the pH-value of its environment [21]. Therefore, the pH value was determined before and after the measurements with a constant result. To neglect interactions between PADA and C4 E2 the absorption behavior of PADA in an aqueous solution and in 1-dodecene is given in Fig. 2a and b. In both figures the absorption behavior of PADA is given in the pure phases and in the same solution which was diluted with pure C4 E2 . This was done to provide a better clarity of the figures. In case of the same PADA concentration the curves were congruent. Due to the dilution of the pure solutions with C4 E2 a concentration of 0.3 mol/L was adjusted within both phases and the absorption of PADA was reduced. Nevertheless, the absorption behavior remained the same. The absorption maximum for PADA in pure water is determined at max = 470 nm which is in good agreement with the work of Klotz and Ming [21]. In presence of 0.3 mol/L C4 E2 the absorption behavior remained the same; hence there were no interactions between PADA and C4 E2 which might affect the analysis of the mass transfer measurements. Within 1-dodecene the results were similar. The absorption behavior did not change due to the presence of C4 E2 . The absorption maximum for PADA in pure 1-dodecene was max = 406 nm. Based on a serial solution of PADA in water and 1-dodecene the analysis of PADA could be carried out.
2.2. Materials
3. Results
In this work the test system water/1-dodecene was used. For the validation of the designed test cell a pure model amphiphilic substance was applied; instead of surfactants at technical grade which consist of large distribution of molecules. Therefore, a short chain glycerol 1-monoether was used as an amphiphilic component. These so called solvo-surfactants [22] are available in purest grades. In this work diethylene glycol butyl ether (C4 E2 ) provided by Sigma Aldrich at an analytical grade of 99+ % is applied. The composition of the ternary system was kept constant. The mass ratio between oil and water ˛ was 0.5: moil ˛= = 0.5. (1) moil + mwater
The separation process of the ternary system water/1dodecene/C4 E2 was observed in the test cell shown in Fig. 1. After stirring for 15 min at a stirrer speed of 750 rpm the stirrer was stopped and the separation process was captured by a camera. The separation process was considered as completed when the deviation between the phase volume at steady state and the dynamic phase volume was 5%. This criterion was defined to avoid the long time which is needed that the system reaches the thermodynamic equilibrium which would result in a constant phase volume. Nevertheless, liquid/liquid separation processes are usually not designed to separate the phases based on the thermodynamic equilibrium. Instead the height of the liquid phases respectively the separation efficiency are the basis for the design. Additionally, the coalescence is a stochastical process. Hence, the separation time depends significantly on the behaviour of last drops that coalesce with the bulk phase. Therefore, the 5% criterion was assumed to be suitable to quantify the separation time.
The mass fraction of the short chain glycerol 1-monoether was 0.2: =
mC4 E2 moil + mwater + mC4 E2
= 0.2
(2)
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model CEP-6631; No. of Pages 6
ARTICLE IN PRESS N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
4
Fig. 3. Dynamic separation process of the ternary system water/1-dodecene/C4 E2 at a temperature of 74 ◦ C.
Fig. 3 shows the results of the dynamic separation process at Due to the stirring the overall volume is slightly higher at the beginning. After two seconds a constant volume in the reactor was detected. The coalescence of the observed emulsion was fast. Within 52 s both phases separated from each other. The separation times which are observed by applying the 5%criterion are given as a function of the temperature in Fig. 4. The separation time increases with the temperature until the third phase appears. In the three phase system the coalescence of the system becomes faster. Furthermore, the organic phase becomes the largest phase by volume. In Fig. 5 the phase volumes under steady state conditions are given as a function of the temperature. With exceeding temperatures of 79 ◦ C three liquid phases arise and the organic phase becomes the largest phase by volume and forms the continuous phase. Therefore, a phase inversion occurs: the oilin-water-emulsion is inverted into a water-in-oil-emulsion. This was proven by conductivity measurements. During the heating process a certain conductivity was determined, therefore, the aqueous (polar) phase was the continuous phase. At 79 ◦ C the organic phase became the continuous phase and the conductivity approached a value of zero. The water-in-oil-emulsion is stable and does not invert to an oil-in water-emulsion by simply cooling down the ternary system. Unless the stirrer is not stopped the conductivity remained at a neglectable value near zero; hence, the emulsion is 74 ◦ C.
Fig. 5. Phase volumes under steady state conditions for various temperatures.
stable. This also affects the separation process. After cooling down the water-in-oil emulsion the separation process is faster. Every measurement was terminated after 60 min and steady state conditions were assumed. The distribution of the phase volumes under steady state conditions are given in Fig. 5 for various temperatures. This information of the volume distribution respectively the heights of the phases are of great importance for the design of the separator. The middle phase which consists mainly of the surfactant/amphiphilic molecules has the highest viscosity (see Table 1) and hence, the mass transfer through this phase is slow. Especially, for non-ionic surfactants which are applied at technical grade. These molecules form structures (microemulsion layers) at the interface which result in additional mass transfer resistances at high concentrations [11,12]. Hence, for the quantification and the fundamental understanding of the transport processes in micellar three phase systems the diffusion coefficient is of great importance. Based on this fundamental knowledge and further the distribution equilibrium the product extraction can be designed and optimized. In this work the mass transfer through the middle phase was observed. To lower the complexity the pure substitute C4 E2 was applied to validate the test cell which was designed for this task. The temperature was kept constant. Otherwise the size of the middle phase would have changed (see Fig. 5). The mass transfer was observed at 85 ◦ C. The dynamic concentration of PADA in all three phases of the ternary system water/1-dodecene/C4 E2 is given in Fig. 6. After homogenising the system in its composition and temperature for 2 h at a stirrer speed of 500 rpm the stirrers were stopped to separate the phases from each other. After the separation process both stirrers were switched back on. The stirrer speed of the upper stirrer was reduced to a value of 28 rpm. The stirrer at the bottom of the test cell was set up to a speed of 100 rpm. Therefore, in both phases a Reynolds number of 260 was realized. Due the low interfacial tension of 0.05 mN/m that was the highest acceptable stirring rate, otherwise the interface was destabilized and an emulsion Table 1 Physical properties of the separated phases of the ternary system water/1dodecene/C4 E2 (˛ = 0.5 ; = 0.2) at 85 ◦ C.
Fig. 4. Separation time as a function of the temperature of the ternary system water/1-dodecene/C4 E2 .
Phase
Density, [kg/m3 ]
Viscosity, [mPa s]
Organic phase Microemulsion phase Aqueous phase
730 932 964
1.12 1.55 0.62
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model
ARTICLE IN PRESS
CEP-6631; No. of Pages 6
N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
5
aq. phase
micro-emulsion
org. phase
Concentration cPADA
xmicro c*aq/mic
c*org/mic
cPADA,org cPADA,aq Characteristic length z Fig. 7. Schematic concentration profile under steady state conditions. Fig. 6. Time dependent concentration of PADA in the three phases of the ternary system water/1-dodecene/C4 E2 (˛ = 0.5 ; = 0.2) at 85 ◦ C.
was created. The azo dye PADA (transferred component) was injected into the aqueous phase after 30 min of slow stirring to avoid any temperature gradient. The concentration of PADA was determined in the aqueous and in the organic phase. The initial concentration of PADA was 2 mg/L. By calculating the mass balance the concentration within the microemulsion phase was calculated. The experiments were repeated three times and the deviations were less than 5%. The PADA concentration in the aqueous phase decreased within the first 60 minutes to a value of approximately 0.01 mg/L. The PADA concentration in the microemulsion increased within this time period to a value of 6.4 mg/L. The PADA concentration in the organic phase under steady state conditions was determined at 0.1 mg/L. These PADA concentrations under steady state conditions were used to calculate the distribution ratio between the three phases. The distribution coefficient K* between the aqueous phase and the microemulsion phase resulted into: ∗ Kaq,micro
=
∗ cmicro ∗ caq
= 668.
(3)
From the high distribution coefficient between the aqueous and the microemulsion phase a high interfacial concentration can be derived. Whereas, the distribution coefficient between the microemulsion and organic phase was low: ∗ Korg,micro =
∗ corg ∗ cmicro
= 0.014.
(4)
The results of the determination of the PADA concentration in all three phases are summarized in the schematic concentration profile (under steady state conditions) given in Fig. 7. Under these conditions the diffusion of PADA out of the aqueous phase into the microemulsion phase is described by: ∗
Vaq
cPADA,micro dc PADA,aq , = −DPADA,mirco · A xmicro dt
(5)
respectively for the organic phase: ∗
Vorg
cPADA,micro dc PADA,org = −DPADA,mirco · A . xmicro dt
(6)
The height of the micro emulsion xmicro is determined by the ∗ analysis given in Fig. 5. cPADA,micro is the concentration difference
within the microemulsion phase, which is given by: ∗ ∗ ∗ ∗ cPADA,micro = caq/micro − corg/micro = Kmicro/aq · cPADA,aq −
cPADA,org . ∗ Korg/micro (7)
By combing the mass balances (Eqs. (5) and (6)) and substituting the concentration difference within the microemulsion phase by Eq. (7) a differential equation is derived: ∗ dcPADA,micro
dt = −DPADA,mirco ·
A
∗ Kmicro/aq
xmicro
Vaq
+
1 ∗ Korg/micro · Vorg
∗ cPADA,micro .
(8) On this way the diffusion coefficient is experimentally determinable. For the observed system at the temperature of 85 ◦ C a diffusion coefficient of PADA through the microemulsion phase of 6.94 · 10−10 m2 /s was determined. In the work of Paul et al. [11] a diffusion coefficient of PADA in water was given with a value of 5.54 · 10−10 m2 /s at 25 ◦ C. This value was calculated by applying the group theory developed by Nakanishi [23]. By applying the Stokes–Einstein-Equation [24] to consider the change of the viscosity (see Table 1) in the microemulsion phase and the change of temperature to 85 ◦ C a diffusion coefficient of 6.29 · 10−10 m2 /s is obtained. This is in excellent agreement with the value determined experimentally by the test cell developed in this work. 4. Conclusion and outlook For the determination of the separation and transport processes in complex liquid three phase systems a modified test cell was developed. The separation process was recorded by a camera and quantified by an image analyzing software. It was shown that the coalescence rate is a function of the phase conditions, respectively the temperature. Furthermore, a hysteresis effect was observed. By observing the steady state conditions the phase volume distribution was measured. In combination with conductivity measurements the phase volume distribution is an important tool for the determination of the continuous phase, which again affects the coalescence behavior. In this work the diffusion of PADA through the microemulsion phase was determined in good
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model
ARTICLE IN PRESS
CEP-6631; No. of Pages 6
N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
6
agreement to the predicted values applying the group theory. After the successful validation of the modified test cell to determine the separation process and the liquid/liquid mass transfer in complex three phase systems technical surfactants will be tested where the complexity increases due to superimposed interfacial phenomena. Nomenclature
Latin letters A area, m2 A absorption c concentration, mg/L d stirrer diameter, m reactor diameter, m D D molecular diffusivity, m2 /s g gravitational acceleration, m/s2 h height, m t time, s temperature, K T V volume, m3 Greek letters ˛ ratio of organic and aqueous phase ˛ = mO /(mO + maq ) mass % of surfactant = ms /(mO + maq + ms ) ˇ mass transfer coefficient, m/s conductivity, mS/m wavelength, nm dynamic viscosity, Pa s density, kg/m3 interfacial tension, N/m Subscripts aq aqueous phase mic microemulsion phase organic phase org C4 E2 diethylene glycol butyl ether Dimensionless numbers Re
Reynolds number
Re =
n · d2 · R
Acknowledgments This work is part of the Collaborative Research Center “Integrated Chemical Processes in Liquid Multiphase Systems” coordinated by the Technische Universität Berlin. Financial support by the Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged (TRR 63).
References [1] M. Kahlweit, R. Strey, Phasenverhalten ternrer Systeme des Typs H2 O-lnichtionisches Amphiphil, Angew. Chem. 97 (1985) 655–669 (in German). [2] P.A. Winsor, Hydrotropy, solubilisation and related emulsification processes. Part I, Trans. Faraday Soc. 44 (1947) 376–398. [3] A.S.C. Lawrence, The metal soaps and the gelation of their paraffin solutions, Trans. Faraday Soc. 34 (1938) 660–667. [4] R.G. Larson, Self-assembly of surfactant liquid crystalline phases by Monte Carlo simulation, J. Chem. Phys. 91 (1989) 2479–2488. [5] M. Schwarze, T. Pogrzeba, K. Seifert, T. Hamerla, R. Schomäcker, Recent developments in hydrogenation and hydroformylation in surfactant systems, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.06.026 [6] T. Hamerla, A. Rost, Y. Kasaka, R. Schomäcker, Hydroformylation of 1-dodecene with water-soluble Rhodium catalysts with bidentate ligands in multiphase systems, ChemCatChem 7 (2013) 1854–1862. [7] T. Hamerla, N. Paul, M. Kraume, R. Schomäcker, Mass transfer phenomena in micellar multiphase systems investigated for a hydroformylation reaction, Chem. Ing. Tech. 85 (2013) 1530–1539. [8] S. Maaß, N. Paul, M. Kraume, Influence of the dispersed phase fraction on experimental and predicted drop size distributions in breakage dominated stirred systems, Chem. Eng. Sci. 76 (2012) 140–153. [9] M. Wegener, N. Paul, M. Kraume, Fluid dynamics and mass transfer at single droplets, Int. J. Heat Mass Transfer 75 (2014) 475–495. [10] J. Villwock, F. Gebauer, J. Kamp, H.-J. Bart, M. Kraume, Systematic analysis of single droplet coalescence, Chem. Eng. Technol. 37 (2014) 1103–1111. [11] N. Paul, P. Schrader, S. Enders, M. Kraume, Effects of phase behaviour on mass transfer in micellar liquid/liquid systems, Chem. Eng. Sci. 115 (2014) 148–156. [12] N. Paul, S. Schön, R. von Klitzing, M. Kraume, Transport processes at single droplets in micellar liquid/liquid systems, AIChE J. 61 (2015) 1092–1104. [13] J.B. Lewis, The mechanism of mass transfer of solutes across liquid-liquid interfaces: Part I: The determination of individual transfer coefficients for binary systems, Chem. Eng. Sci. 3 (1954) 248–259. [14] W. Nitsch, K. Hillekamp, Zur Kinetik der Zinkionen Extraktion aus Wasser in Dithizon-beladene Solventen, Chem. Ztg. 96 (1972) 254 (in German). [15] W. Nitsch, B. Kruis, The influence of flow and concentration on the mass transfer mechanism in chelating liquid/liquid-extractions, J. Inorg. Nucl. Chem. 40 (1978) 857–864. [16] G.J. Hanna, R.D. Noble, Measurement of liquid–liquid interfacial kinetics, Chem. Rev. 85 (1985) 583–598. [17] S. Lee, A. Varma, Kinetic study of biphasic aldol condensation of nbutyraldehyde using stirred cell, Chem. Eng. Sci. 104 (2013) 619–629. [18] Y. Wang, Y. Li, D. Li, W. Liao, Kinetics of thorium extraction with di-(2-ethylhexyl) 2-ethylhexyl phosphonate from nitric acid medium, Hydrometallurgy 140 (2013) 66–70. [19] D. Datta, S. Kumar, Equilibrium and kinetic studies of the reactive extraction of nicotinic acid with Tri-n-octylamine dissolved in MIBK, Ind. Eng. Chem. Res. 52 (2013) 14680–14686. [20] E. Bertakis, M. Kalem, A. Pfennig, Model-based geometry optimization of a Nitsch cell using the Fisher information matrix, Chem. Eng. Sci. 63 (2008) 4881–4887. [21] I.M. Klotz, W.-C. Loh Ming, Stability constants for some metal chelates of pyridine-2-azo-p-dimethylaniline, J. Am. Chem. Soc. 7 (1953) 4159–4162. [22] S. Queste, P. Baudin, D. Touraud, W. Kunz, J.-M. Aubry, Short chain glycerol 1monoethers – a new class of green solvo-surfactants, Green Chem. 8 (2006) 822–830. [23] K. Nakanishi, Prediction of diffusion coefficient of nonelectrolytes in dilute solution based on generalized Hammond–Stokes plot, Ind. Eng. Chem. Fundam. 17 (1978) 253–256. [24] A. Einstein, Über die von der molekularkinetischen Theorie der Wärme geforderte Bewegung von in ruhenden Flüssigkeiten suspendierten Teilchen, Ann. Phys. 322 (1905) 549–560 (in German).
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
ARTICLE IN PRESS
CEP-6631; No. of Pages 6
Chemical Engineering and Processing xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Chemical Engineering and Processing: Process Intensification journal homepage: www.elsevier.com/locate/cep
Determination of phase separation and mass transfer in complex micellar three phase systems N. Paul ∗ , J.M. Schulz, M. Kraume Chair of Chemical and Process Engineering, Technische Universität Berlin, Ackerstraße 76, D-13355 Berlin, Germany
a r t i c l e
i n f o
Article history: Received 13 March 2015 Received in revised form 11 June 2015 Accepted 4 July 2015 Available online xxx Keywords: Separation process Liquid/liquid mass transfer Liquid three phase systems
a b s t r a c t Combining the advantages of the homogeneous and heterogeneous catalysis is the main goal of tunable solvent systems. Homogeneous conditions shall be achieved throughout the reaction period and heterogeneous conditions shall be adjusted within the separation process. Therefore, high and specific reaction rates are coupled with a simplified separation process. For the design of the separation process the phase separation itself and the liquid/liquid mass transfer must be quantified. Therefore, a test cell is designed in this work. For the experimental investigations a model test system was applied (water/1-dodecene/C4 E2 ). The separation process of the liquid three phase system was observed by an image analysis. A hysteresis effect was observed which also influenced the separation process. Within the three phase system the coalescence rate is faster than under two phase conditions in the observed temperature range. A diffusion coefficient through the C4 E2 rich middle phase could be determined experimentally, which was in good agreement with theoretical calculations. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Tunable solvent systems can be applied to combine the advantages of the homogeneous and heterogeneous catalysis. Mild reaction conditions as well as high and selective yields are combined with a simplified separation processes. In homogeneous systems mass transfer processes can be neglected. Nevertheless, the catalyst/product separation is the great challenge of these systems. Therefore, the main idea of tunable solvent systems is to provide homogeneous conditions for the reaction and heterogeneous conditions for the separation. Liquid/liquid reaction systems are a special type of the homogeneous catalysis. The separation process is simplified in liquid/liquid systems if product and catalyst are solved in different phases. But due to the additional phase transport phenomena must be taken into consideration. Micellar multiphase systems are examples for tunable solvent systems. These systems fulfill many principles of the green chemistry, e.g., using water as a solvent, reducing waste production, etc. Due to the interfacial active molecules (surfactants) which are used in these systems the phase behaviour is complex [1]. Four different phase conditions are possible for ternary systems of the type
∗ Corresponding author. E-mail address: [email protected] (N. Paul).
water/organic solvent/non-ionic surfactant. These conditions were classified by Winsor [2]: • Type I consists of an organic and aqueous phase; micelles solubilize oil within the aqueous phase. Therefore, an O/W mircoemuslion occurs [3]. • Type II consists of an organic and aqueous phase; inverse micelles solubilize water in the organic phase; therefore, a W/O microemulsion is created [2]. • Type III consists of three coexisting phases an organic phase, an aqueous phase and a surfactant rich middle phase. Low amounts of water and oil are solubilized within the surfactant rich phase, which is also a microemuslion. • Type IV: One phase condition. Within these conditions anisotropic structures can be observed at discrete conditions. In these cases micelles agglomerate to large micelle clusters and form liquid crystalline structures [4]. For a clearer understanding only the Winsor III system is called a microemulsion in this paper. The best performance for some reactions and separation processes were achieved under three phase conditions [5]. For some reactions (e.g., the hydroformylation of long chained alkenes) it is beneficial if the catalyst systems are solved within the microemulsion phase [6]. In case of the hydroformylation syngas must be transported into the microemulsion phase [7]. In this phase all other reactants are present; hence,
http://dx.doi.org/10.1016/j.cep.2015.07.010 0255-2701/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model CEP-6631; No. of Pages 6 2
ARTICLE IN PRESS N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
mass transfer phenomena must be taken into consideration. Furthermore, the distribution coefficient and the mass transfer of the transferred component must be known to design the separation process. After the reaction the products have to be transported back into the organic phase to realize a simplified separation process. Therefore, the phases need to coalesce in a reasonable period of time. In presence of surfactants the coalescence behavior is mostly hindered due to the adsorption of surfactants at the liquid/liquid interface [8]. This work focuses on the determination of fundamental understanding of the liquid/liquid transport processes in liquid three phase systems which is needed for the design of the separation process. Furthermore, the coalescence must be taken into consideration to understand the separation of these complex systems in detail. Often single drop experiments are carried out to understand transport processes [9] and coalescence processes fundamentally [10]. In the presence of non-ionic surfactants an additional mass transfer resistance was observed. With exceeding the critical micelle concentration (CMC) a change of the phase behavior at the liquid/liquid interface was observed. This lead to an increase of the interfacial viscosity and caused a reduction of the mass transfer rates [11,12]. Especially, for micellar three phase systems these additional mass transfer resistances are expected. Under micellar three phase conditions single drop experiments fail due to low interfacial tension. Nevertheless, the mass transfer and the distribution coefficients, etc. must be quantified in order to design the extraction process. Different test cells need to be taken into consideration. Stirred test cells are also commonly used to determine mass transfer rates across liquid/liquid interfaces. Here, two phases are overlaid with each other; hence a defined area of contact arises. The first test cell was presented by Lewis [13]. In the “Lewis test cell” the stirrer speed can only be adjusted in both phases equally. This results in different Reynolds numbers in both phases due to different physical properties (density, viscosity). Hence, the fluid dynamic properties within the phases cannot be chosen equally. An improvement of the “Lewis test cell” was achieved by Nitsch [14,15]. In the “Nitsch test cell” both liquid phases are stirred separately. A broad overview of methods determining the liquid/liquid mass transfer is given in [16]. The test cells which are given in this overview were modified throughout the last
years and have been adjusted to their applications. Lee and Varma [17] used a stirred cell to quantify the kinetics of a biphasic aldol condensation within the mass transfer controlled regime. In this work low stirrer speeds were applied to ensure a stable interface. Another development is shown in the work of Wang et al. [18]; here the extraction kinetics of Thorium were determined within a constant interfacial area test cell. Two stirrers were applied to stir the phases independently. Nevertheless, these stirrers were installed in a cubic test cell and were not installed above each other. They were installed shifted to ensure a stable interface. Datta and Kumar [19] presented a test cell in which two stirrers were installed above each other, but both were attached to one stirrer shaft. Therefore, both stirrers could not be operated independently and further the interface is disturbed. Bertakis et al. [20] optimized the Nitsch test cell based on the Fisher information matrix to improve the quality of the measurements. In this work other requirements on the test cell arise. Due to the low interfacial tension the stirring has to be carried out with great care to neither destabilize the interface nor to create an emulsion, which ensures the quantification of the mass transfer across a defined cross section. Therefore, the influence of the amphiphilic molecules can be identified to optimize the extraction process.
2. Materials and methods 2.1. Experimental setup In Fig. 1 the test cell is shown which was designed to determine the separation process and the mass transfer in liquid three phase systems. The test cell is designed based on the model of the “Nitsch test cell”. It was modified for the challenges in micellar three phase systems in which very low interfacial tensions occur. A double-walled glass reactor (1) with an inner diameter of D = 65 mm and a planar bottom was applied. Therefore, an optical access and an exact determination of the volume was ensured. Two different stirrers (2/3) were installed to avoid concentration and temperature gradients in the aqueous and organic phase. The microemulsion phase was not stirred. Both stirrers were operated independently to provide similar fluid dynamic conditions in each phase. The volume of each
Fig. 1. Experimental setup for the determination of the phase separation and the liquid/liquid mass transfer in micellar three phase systems: 1 double-walled glass reactor; 2 and 3 stirrers; 4 scale; 5 camera; 6a,b,c Hamilton modules PSD/3 with three way valves; 7 Lauda E-200 thermostat; 8 computer.
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model
ARTICLE IN PRESS
CEP-6631; No. of Pages 6
N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
1 mg/L PADA in water pure diluted w. C4E2 (cC4E2 = 0.3 mol/L)
1 mg/L PADA in 1-dodecene pure diluted w. C4E2 (cC4E2 = 0.3 mol/L)
0,15 0.15
0,15 0.15 0.15
λ max= 406 nm
λ max= 470 nm
Absorption A [-]
0,12 0.12
Absorption A [-]
3
0,09 0.09 0,06 0.06 0.03 0,03
0
a)
300
0,12 0.12 0.12 0.09 0,09 0.09 0.06 0,06 0.06 0.03 0,03 0.03
b) 400
500
600
0 300
Wavelength λ [nm]
400
500
600
Wavelength λ [nm]
Fig. 2. Absorption behavior of PADA in water (a) and 1-dodecene (b) in pure systems and in presence of C4 E2 .
phase was observed by a camera (5). Furthermore, the camera was used to determine the separation process. By recording pictures with a camera and an image analyzing software it was possible to determine the heights of the interfaces respectively the volume of each phase. This was the basis for the determination of the separation process. As soon as the volumes of each phase reached a constant value the separation process was treated as completed. For an exact determination a scale (4) was installed. For taking samples automatically and for injecting a specific amount of the tracer substance three different syringe pumps (Hamilton PSD/3-modules) were used. These modules were connected to small pipes, which were adjustable in their heights. Therefore, it was possible to take clean samples out of the organic and the aqueous phase. Furthermore, the syringe pumps were equipped with three-way valves; hence a simple cleaning process and sample taking was provided. The test cell was working semi-automatically and was controlled by a Labview program (8). A Lauda thermostat (E-200) was installed to control the temperature. For the determination of the transport and separation processes in complex micellar three phase systems the test cell shown in Fig. 1 was qualified due to its ability to setup the stirrer speed independently at low stirrer speeds. The optical accessibility allowed the determination of the separation process. Furthermore, the pipes for taking samples could be adjusted to the phase heights to ensure clean probes. The syringe pumps, which were controlled by a Labview program provided a simplified and semi-automatic method to take clean samples of the organic and the aqueous phases.
As transferred component an azo dye pyridine-2-azodimethyl aniline (PADA) was chosen. Hence, the concentration could be quantified by an UV/vis-photometer (Specord 210 by Jena Analytik). PADA changes the absorption behavior due to the pH-value of its environment [21]. Therefore, the pH value was determined before and after the measurements with a constant result. To neglect interactions between PADA and C4 E2 the absorption behavior of PADA in an aqueous solution and in 1-dodecene is given in Fig. 2a and b. In both figures the absorption behavior of PADA is given in the pure phases and in the same solution which was diluted with pure C4 E2 . This was done to provide a better clarity of the figures. In case of the same PADA concentration the curves were congruent. Due to the dilution of the pure solutions with C4 E2 a concentration of 0.3 mol/L was adjusted within both phases and the absorption of PADA was reduced. Nevertheless, the absorption behavior remained the same. The absorption maximum for PADA in pure water is determined at max = 470 nm which is in good agreement with the work of Klotz and Ming [21]. In presence of 0.3 mol/L C4 E2 the absorption behavior remained the same; hence there were no interactions between PADA and C4 E2 which might affect the analysis of the mass transfer measurements. Within 1-dodecene the results were similar. The absorption behavior did not change due to the presence of C4 E2 . The absorption maximum for PADA in pure 1-dodecene was max = 406 nm. Based on a serial solution of PADA in water and 1-dodecene the analysis of PADA could be carried out.
2.2. Materials
3. Results
In this work the test system water/1-dodecene was used. For the validation of the designed test cell a pure model amphiphilic substance was applied; instead of surfactants at technical grade which consist of large distribution of molecules. Therefore, a short chain glycerol 1-monoether was used as an amphiphilic component. These so called solvo-surfactants [22] are available in purest grades. In this work diethylene glycol butyl ether (C4 E2 ) provided by Sigma Aldrich at an analytical grade of 99+ % is applied. The composition of the ternary system was kept constant. The mass ratio between oil and water ˛ was 0.5: moil ˛= = 0.5. (1) moil + mwater
The separation process of the ternary system water/1dodecene/C4 E2 was observed in the test cell shown in Fig. 1. After stirring for 15 min at a stirrer speed of 750 rpm the stirrer was stopped and the separation process was captured by a camera. The separation process was considered as completed when the deviation between the phase volume at steady state and the dynamic phase volume was 5%. This criterion was defined to avoid the long time which is needed that the system reaches the thermodynamic equilibrium which would result in a constant phase volume. Nevertheless, liquid/liquid separation processes are usually not designed to separate the phases based on the thermodynamic equilibrium. Instead the height of the liquid phases respectively the separation efficiency are the basis for the design. Additionally, the coalescence is a stochastical process. Hence, the separation time depends significantly on the behaviour of last drops that coalesce with the bulk phase. Therefore, the 5% criterion was assumed to be suitable to quantify the separation time.
The mass fraction of the short chain glycerol 1-monoether was 0.2: =
mC4 E2 moil + mwater + mC4 E2
= 0.2
(2)
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model CEP-6631; No. of Pages 6
ARTICLE IN PRESS N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
4
Fig. 3. Dynamic separation process of the ternary system water/1-dodecene/C4 E2 at a temperature of 74 ◦ C.
Fig. 3 shows the results of the dynamic separation process at Due to the stirring the overall volume is slightly higher at the beginning. After two seconds a constant volume in the reactor was detected. The coalescence of the observed emulsion was fast. Within 52 s both phases separated from each other. The separation times which are observed by applying the 5%criterion are given as a function of the temperature in Fig. 4. The separation time increases with the temperature until the third phase appears. In the three phase system the coalescence of the system becomes faster. Furthermore, the organic phase becomes the largest phase by volume. In Fig. 5 the phase volumes under steady state conditions are given as a function of the temperature. With exceeding temperatures of 79 ◦ C three liquid phases arise and the organic phase becomes the largest phase by volume and forms the continuous phase. Therefore, a phase inversion occurs: the oilin-water-emulsion is inverted into a water-in-oil-emulsion. This was proven by conductivity measurements. During the heating process a certain conductivity was determined, therefore, the aqueous (polar) phase was the continuous phase. At 79 ◦ C the organic phase became the continuous phase and the conductivity approached a value of zero. The water-in-oil-emulsion is stable and does not invert to an oil-in water-emulsion by simply cooling down the ternary system. Unless the stirrer is not stopped the conductivity remained at a neglectable value near zero; hence, the emulsion is 74 ◦ C.
Fig. 5. Phase volumes under steady state conditions for various temperatures.
stable. This also affects the separation process. After cooling down the water-in-oil emulsion the separation process is faster. Every measurement was terminated after 60 min and steady state conditions were assumed. The distribution of the phase volumes under steady state conditions are given in Fig. 5 for various temperatures. This information of the volume distribution respectively the heights of the phases are of great importance for the design of the separator. The middle phase which consists mainly of the surfactant/amphiphilic molecules has the highest viscosity (see Table 1) and hence, the mass transfer through this phase is slow. Especially, for non-ionic surfactants which are applied at technical grade. These molecules form structures (microemulsion layers) at the interface which result in additional mass transfer resistances at high concentrations [11,12]. Hence, for the quantification and the fundamental understanding of the transport processes in micellar three phase systems the diffusion coefficient is of great importance. Based on this fundamental knowledge and further the distribution equilibrium the product extraction can be designed and optimized. In this work the mass transfer through the middle phase was observed. To lower the complexity the pure substitute C4 E2 was applied to validate the test cell which was designed for this task. The temperature was kept constant. Otherwise the size of the middle phase would have changed (see Fig. 5). The mass transfer was observed at 85 ◦ C. The dynamic concentration of PADA in all three phases of the ternary system water/1-dodecene/C4 E2 is given in Fig. 6. After homogenising the system in its composition and temperature for 2 h at a stirrer speed of 500 rpm the stirrers were stopped to separate the phases from each other. After the separation process both stirrers were switched back on. The stirrer speed of the upper stirrer was reduced to a value of 28 rpm. The stirrer at the bottom of the test cell was set up to a speed of 100 rpm. Therefore, in both phases a Reynolds number of 260 was realized. Due the low interfacial tension of 0.05 mN/m that was the highest acceptable stirring rate, otherwise the interface was destabilized and an emulsion Table 1 Physical properties of the separated phases of the ternary system water/1dodecene/C4 E2 (˛ = 0.5 ; = 0.2) at 85 ◦ C.
Fig. 4. Separation time as a function of the temperature of the ternary system water/1-dodecene/C4 E2 .
Phase
Density, [kg/m3 ]
Viscosity, [mPa s]
Organic phase Microemulsion phase Aqueous phase
730 932 964
1.12 1.55 0.62
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model
ARTICLE IN PRESS
CEP-6631; No. of Pages 6
N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
5
aq. phase
micro-emulsion
org. phase
Concentration cPADA
xmicro c*aq/mic
c*org/mic
cPADA,org cPADA,aq Characteristic length z Fig. 7. Schematic concentration profile under steady state conditions. Fig. 6. Time dependent concentration of PADA in the three phases of the ternary system water/1-dodecene/C4 E2 (˛ = 0.5 ; = 0.2) at 85 ◦ C.
was created. The azo dye PADA (transferred component) was injected into the aqueous phase after 30 min of slow stirring to avoid any temperature gradient. The concentration of PADA was determined in the aqueous and in the organic phase. The initial concentration of PADA was 2 mg/L. By calculating the mass balance the concentration within the microemulsion phase was calculated. The experiments were repeated three times and the deviations were less than 5%. The PADA concentration in the aqueous phase decreased within the first 60 minutes to a value of approximately 0.01 mg/L. The PADA concentration in the microemulsion increased within this time period to a value of 6.4 mg/L. The PADA concentration in the organic phase under steady state conditions was determined at 0.1 mg/L. These PADA concentrations under steady state conditions were used to calculate the distribution ratio between the three phases. The distribution coefficient K* between the aqueous phase and the microemulsion phase resulted into: ∗ Kaq,micro
=
∗ cmicro ∗ caq
= 668.
(3)
From the high distribution coefficient between the aqueous and the microemulsion phase a high interfacial concentration can be derived. Whereas, the distribution coefficient between the microemulsion and organic phase was low: ∗ Korg,micro =
∗ corg ∗ cmicro
= 0.014.
(4)
The results of the determination of the PADA concentration in all three phases are summarized in the schematic concentration profile (under steady state conditions) given in Fig. 7. Under these conditions the diffusion of PADA out of the aqueous phase into the microemulsion phase is described by: ∗
Vaq
cPADA,micro dc PADA,aq , = −DPADA,mirco · A xmicro dt
(5)
respectively for the organic phase: ∗
Vorg
cPADA,micro dc PADA,org = −DPADA,mirco · A . xmicro dt
(6)
The height of the micro emulsion xmicro is determined by the ∗ analysis given in Fig. 5. cPADA,micro is the concentration difference
within the microemulsion phase, which is given by: ∗ ∗ ∗ ∗ cPADA,micro = caq/micro − corg/micro = Kmicro/aq · cPADA,aq −
cPADA,org . ∗ Korg/micro (7)
By combing the mass balances (Eqs. (5) and (6)) and substituting the concentration difference within the microemulsion phase by Eq. (7) a differential equation is derived: ∗ dcPADA,micro
dt = −DPADA,mirco ·
A
∗ Kmicro/aq
xmicro
Vaq
+
1 ∗ Korg/micro · Vorg
∗ cPADA,micro .
(8) On this way the diffusion coefficient is experimentally determinable. For the observed system at the temperature of 85 ◦ C a diffusion coefficient of PADA through the microemulsion phase of 6.94 · 10−10 m2 /s was determined. In the work of Paul et al. [11] a diffusion coefficient of PADA in water was given with a value of 5.54 · 10−10 m2 /s at 25 ◦ C. This value was calculated by applying the group theory developed by Nakanishi [23]. By applying the Stokes–Einstein-Equation [24] to consider the change of the viscosity (see Table 1) in the microemulsion phase and the change of temperature to 85 ◦ C a diffusion coefficient of 6.29 · 10−10 m2 /s is obtained. This is in excellent agreement with the value determined experimentally by the test cell developed in this work. 4. Conclusion and outlook For the determination of the separation and transport processes in complex liquid three phase systems a modified test cell was developed. The separation process was recorded by a camera and quantified by an image analyzing software. It was shown that the coalescence rate is a function of the phase conditions, respectively the temperature. Furthermore, a hysteresis effect was observed. By observing the steady state conditions the phase volume distribution was measured. In combination with conductivity measurements the phase volume distribution is an important tool for the determination of the continuous phase, which again affects the coalescence behavior. In this work the diffusion of PADA through the microemulsion phase was determined in good
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010
G Model
ARTICLE IN PRESS
CEP-6631; No. of Pages 6
N. Paul et al. / Chemical Engineering and Processing xxx (2015) xxx–xxx
6
agreement to the predicted values applying the group theory. After the successful validation of the modified test cell to determine the separation process and the liquid/liquid mass transfer in complex three phase systems technical surfactants will be tested where the complexity increases due to superimposed interfacial phenomena. Nomenclature
Latin letters A area, m2 A absorption c concentration, mg/L d stirrer diameter, m reactor diameter, m D D molecular diffusivity, m2 /s g gravitational acceleration, m/s2 h height, m t time, s temperature, K T V volume, m3 Greek letters ˛ ratio of organic and aqueous phase ˛ = mO /(mO + maq ) mass % of surfactant = ms /(mO + maq + ms ) ˇ mass transfer coefficient, m/s conductivity, mS/m wavelength, nm dynamic viscosity, Pa s density, kg/m3 interfacial tension, N/m Subscripts aq aqueous phase mic microemulsion phase organic phase org C4 E2 diethylene glycol butyl ether Dimensionless numbers Re
Reynolds number
Re =
n · d2 · R
Acknowledgments This work is part of the Collaborative Research Center “Integrated Chemical Processes in Liquid Multiphase Systems” coordinated by the Technische Universität Berlin. Financial support by the Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged (TRR 63).
References [1] M. Kahlweit, R. Strey, Phasenverhalten ternrer Systeme des Typs H2 O-lnichtionisches Amphiphil, Angew. Chem. 97 (1985) 655–669 (in German). [2] P.A. Winsor, Hydrotropy, solubilisation and related emulsification processes. Part I, Trans. Faraday Soc. 44 (1947) 376–398. [3] A.S.C. Lawrence, The metal soaps and the gelation of their paraffin solutions, Trans. Faraday Soc. 34 (1938) 660–667. [4] R.G. Larson, Self-assembly of surfactant liquid crystalline phases by Monte Carlo simulation, J. Chem. Phys. 91 (1989) 2479–2488. [5] M. Schwarze, T. Pogrzeba, K. Seifert, T. Hamerla, R. Schomäcker, Recent developments in hydrogenation and hydroformylation in surfactant systems, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.06.026 [6] T. Hamerla, A. Rost, Y. Kasaka, R. Schomäcker, Hydroformylation of 1-dodecene with water-soluble Rhodium catalysts with bidentate ligands in multiphase systems, ChemCatChem 7 (2013) 1854–1862. [7] T. Hamerla, N. Paul, M. Kraume, R. Schomäcker, Mass transfer phenomena in micellar multiphase systems investigated for a hydroformylation reaction, Chem. Ing. Tech. 85 (2013) 1530–1539. [8] S. Maaß, N. Paul, M. Kraume, Influence of the dispersed phase fraction on experimental and predicted drop size distributions in breakage dominated stirred systems, Chem. Eng. Sci. 76 (2012) 140–153. [9] M. Wegener, N. Paul, M. Kraume, Fluid dynamics and mass transfer at single droplets, Int. J. Heat Mass Transfer 75 (2014) 475–495. [10] J. Villwock, F. Gebauer, J. Kamp, H.-J. Bart, M. Kraume, Systematic analysis of single droplet coalescence, Chem. Eng. Technol. 37 (2014) 1103–1111. [11] N. Paul, P. Schrader, S. Enders, M. Kraume, Effects of phase behaviour on mass transfer in micellar liquid/liquid systems, Chem. Eng. Sci. 115 (2014) 148–156. [12] N. Paul, S. Schön, R. von Klitzing, M. Kraume, Transport processes at single droplets in micellar liquid/liquid systems, AIChE J. 61 (2015) 1092–1104. [13] J.B. Lewis, The mechanism of mass transfer of solutes across liquid-liquid interfaces: Part I: The determination of individual transfer coefficients for binary systems, Chem. Eng. Sci. 3 (1954) 248–259. [14] W. Nitsch, K. Hillekamp, Zur Kinetik der Zinkionen Extraktion aus Wasser in Dithizon-beladene Solventen, Chem. Ztg. 96 (1972) 254 (in German). [15] W. Nitsch, B. Kruis, The influence of flow and concentration on the mass transfer mechanism in chelating liquid/liquid-extractions, J. Inorg. Nucl. Chem. 40 (1978) 857–864. [16] G.J. Hanna, R.D. Noble, Measurement of liquid–liquid interfacial kinetics, Chem. Rev. 85 (1985) 583–598. [17] S. Lee, A. Varma, Kinetic study of biphasic aldol condensation of nbutyraldehyde using stirred cell, Chem. Eng. Sci. 104 (2013) 619–629. [18] Y. Wang, Y. Li, D. Li, W. Liao, Kinetics of thorium extraction with di-(2-ethylhexyl) 2-ethylhexyl phosphonate from nitric acid medium, Hydrometallurgy 140 (2013) 66–70. [19] D. Datta, S. Kumar, Equilibrium and kinetic studies of the reactive extraction of nicotinic acid with Tri-n-octylamine dissolved in MIBK, Ind. Eng. Chem. Res. 52 (2013) 14680–14686. [20] E. Bertakis, M. Kalem, A. Pfennig, Model-based geometry optimization of a Nitsch cell using the Fisher information matrix, Chem. Eng. Sci. 63 (2008) 4881–4887. [21] I.M. Klotz, W.-C. Loh Ming, Stability constants for some metal chelates of pyridine-2-azo-p-dimethylaniline, J. Am. Chem. Soc. 7 (1953) 4159–4162. [22] S. Queste, P. Baudin, D. Touraud, W. Kunz, J.-M. Aubry, Short chain glycerol 1monoethers – a new class of green solvo-surfactants, Green Chem. 8 (2006) 822–830. [23] K. Nakanishi, Prediction of diffusion coefficient of nonelectrolytes in dilute solution based on generalized Hammond–Stokes plot, Ind. Eng. Chem. Fundam. 17 (1978) 253–256. [24] A. Einstein, Über die von der molekularkinetischen Theorie der Wärme geforderte Bewegung von in ruhenden Flüssigkeiten suspendierten Teilchen, Ann. Phys. 322 (1905) 549–560 (in German).
Please cite this article in press as: N. Paul, et al., Determination of phase separation and mass transfer in complex micellar three phase systems, Chem. Eng. Process. (2015), http://dx.doi.org/10.1016/j.cep.2015.07.010