- Email: [email protected]
Electrocoalescence in non-uniform electric fields: An experimental study
Chemical Engineering and Processing 96 (2015) 28–38
Contents lists available at ScienceDirect
Chemical Engineering and Processing: Process Intensification journal homepage: www.elsevier.com/locate/cep
Electrocoalescence in non-uniform electric fields: An experimental study Sameer Mhatre, Rochish Thaokar ∗ Department of Chemical Engineering, Indian Institute of Technology Bombay, Mumbai, Maharashtra 400076, India
a r t i c l e
i n f o
Article history: Received 30 May 2015 Received in revised form 24 July 2015 Accepted 25 July 2015 Available online 29 July 2015 Keywords: Electrocoalescence Phase separation Crude oil demulsification Electrohydrodynamics Non-uniform electric field Dielectrophoresis
a b s t r a c t The method of electrostatic force-assisted phase separation is widely used in the chemical industry. The fast and clean separation capabilities make electrocoalescence a favorite among the other competing phase separation techniques. In the present study, we experimentally investigated the relative merits of symmetric and asymmetric non-uniform and uniform electric fields in electrocoalescence. Three different types of emulsions are studied: (i) a perfectly conducting (PC) phase dispersed in a perfectly dielectric (PD) medium, (ii) a leaky dielectric (LD) phase dispersed in another LD phase of higher electrical conductivity, and (iii) a LD phase dispersed in another LD phase of lower electrical conductivity. The coalescence behavior of each emulsion in the non-uniform electric fields is analyzed and the results are compared with that in a uniform field. The observations show that a non-uniformity as well as asymmetry of electric field can affect electrocoalescence in a nontrivial way and is sensitive to the drop-medium system, specifically pin-plate, quadrupole and annular electrode configurations are found to be advantageous over uniform field in PC drops in PD medium, less conducting LD drops in more conducting LD medium and more conducting LD drops in less conducting LD medium, respectively. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Stable emulsions have been an issue of concern in many chemical processes. In the petroleum industry, naturally stable emulsion of water in crude oil can pose several problems during its processing. In liquid–liquid extraction, a stable mixture of extract and raffinate phases take a long time to separate by gravity. There exist many techniques of phase separation; e.g. gravitational or buoyancy separation, centrifugation, chemical treatment, membrane separation, etc. [1–3]. However each of these conventional techniques has some disadvantages; e.g. gravity settling takes long time for dispersed phase to settle if the drops are very small in size, chemical treatment affects the quality of separated water and also causes pollution, mechanical energy and thereby shear may not be very efficient in coalescence, etc. The use of electrostatic energy has been considered as one of the most effective and clean methods of phase separation. In addition to separation of water to a desired level, low energy consumption and lack of necessity of further purification of the separated phases make the electrostatic method worthy. Electrocoalescence is the de-facto method in
∗ Corresponding author. E-mail address: [email protected] (R. Thaokar). http://dx.doi.org/10.1016/j.cep.2015.07.025 0255-2701/© 2015 Elsevier B.V. All rights reserved.
desalting operations wherein minute water droplets of size below 50 m are separated from crude oil [4]. Started in early 20th century, extensive research has resulted in fundamental understanding of many complex problems in electrocoalescence that have helped to make the technology more efficient, faster and the devices more compact [5]. Applying an electric field induces free charge separation in a conducting water drop resulting in a dipole and the consequent dipole–dipole interaction between neighboring drops in an emulsion enhances the coalescence rate. In addition to the dipolar attraction, an applied electric field can induce macroscopic flow currents due to movement of drops or due to tangential stresses in leaky dielectric systems which increase the probability of inter-drop contact. The coalescence of two drops in an electric field is assume to occur in three stages [5,6]; in the first stage, two drops are attracted towards each other on application of the field. The second stage involves squeezing of the medium fluid from the film between leading faces of the approaching drops. When such an interstitial film becomes microscopically thin, its break up and consequent coalescence of two drops occurs in the third stage. There are several parameters that determine the coalescence dynamics in the presence of an electric field. The factors such as the magnitude and type of an applied field, electrical and physical properties of the fluids, impurities in the emulsion, flow or turbulence, drop size, dispersity, etc., govern the kinetics of electrocoalescence. The role of such parameters has
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
been a topic of investigation in the majority of scientific studies in electrocoalescence [7–10]. Application of moderate electric field to an emulsion leads to several phenomena such as chain formation, partial coalescence, non-coalescence and receding of about to coalesce drops, drop breakup, etc., which are known to adversely affect the rate of electrocoalescence [5]. In an electrocoalescing emulsion, drops can arrange into chains aligned in the direction of the applied field [11,12]. The interfacial stability of the drop interface often promotes chain formation. Such chains of droplets not only slow down the coalescence rate, but can also extend to the electrodes resulting in to short circuiting. The formation of fine progeny droplets during drop-interface or drop–drop coalescence is termed as partial coalescence [13]. This phenomenon introduces new drops in the emulsion which are difficult to separate owing to their very small sizes. The application of an electric field induces drop–drop approach, but cannot guarantee coalescence. It has been observed that in certain conditions (especially under strong applied electric fields), coalescing drops retreat after their contact [14]. In another undesired effect, after coalescence, if the field at the resulting drop interface exceeds the critical field limit, Ec = 0.648 /2am , a drop can further break into smaller droplets [15]. Here, is interfacial tension, a is drop radius and m is permittivity of medium fluid. Such adverse effects need to be overcome to speedup the electrocoalescence process.
Glass Slides
29
In addition to water separation from crude oil, electric field can also be employed for phase separation in systems involving a leaky dielectric (LD) fluid dispersed in to another LD medium. The examples include fractionation of mixed oils [16], electrorheology of electro-magneto responsive fluids, polymer blends, etc. Electrorheological (ER) fluids are LD–LD suspensions which display a change in rheological properties e.g. shear thinning or thickening under the influence of an electric field [17,18]. Such fluids are used in the active control devices such as dampers, shock absorbers, clutches, brakes, etc. The preparation of a polymer blend involves dispersing one polymer in to another in order to acquire the desired properties from the resulting new polymer. Control over the droplet size in a blend plays a key role in deciding the properties of the resulting polymer. Electric fields can also be used to precisely regulate the dispersed phase drop-size distribution. Furthermore, during recycling, electrocoalescence can be used to separate a dispersed polymer phase from the molten blend. An in-depth understanding of the electrocoalescence in an oil-in-oil emulsion can help to control the size distributions in the ER fluids as well as in the polymer blends. Electrocoalescence has been construed as separation of two liquid phases by applying uniform electric field. The electric field employed in the previous electrocoalescence studies was predominantly of uniform kind. Very few studies in the literature discuss the use of non-uniform electric fields for phase separation. Pearce
Glass Slides d
d
Spacer Electrodes
(a) Parallel plate electrodes
(b) Quadrupole electrode system V
d d
Glass Slides d
Spacer Electrodes
(c) Pin-plate electrode system
(d) Four-pin electrode system
d
(e) Annular electrode system Fig. 1. Schematics of various electrode systems used to induce the uniform and non-uniform electric fields. Electrode-electrode separation, d = 5 mm in all the electrode configurations.
30
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
[11] used the electric field generated by the coaxial cylindrical electrodes for breaking a water-in-oil emulsion. In another study, Eow and Ghadiri [19] suggested five different electrode designs to study the effect of spatial distribution of an applied field on coalescence of two drops in a flowing medium. They also studied the effect of applied field and frequency on the oscillation of a moving drop in the suggested electrode systems. In a pilot plant-scale study, Noik et al. [20] designed a compact coalescer employing two annular cylindrical electrodes and centrifugal flow of the emulsion. The roles of electric field as well as flow field in electrocoalescence were investigated. The shear force or turbulence due to the hydrodynamic conditions induce drop break-up or disturb the dipolar interactions between the drops [20]. Non-uniform electric fields have been extensively used in the segregation and separation of rigid- as well as bio-particles. Depending upon the polarizability of the phases involved, the non-uniformity of an applied field forces the suspending particles to assemble in a certain region of the field. This phenomenon, known as dielectrophoresis, can also be utilized to expedite the approach of two droplets which is the first and crucial step of the coalescence in an emulsion. The purpose of the present study is to investigate the prospects of the applied non-uniform electric fields in electrocoalescence over a uniform field. Different kinds of electrode systems were used to generate the non-uniform electric fields, viz. quadrupole, pin-plate, four-pin and annular electrodes. The non-uniform fields induced in the quadrupole, four-pin and annular electrode systems are essentially symmetric; whereas, the pin-plate system generates a highly asymmetric electric field. All the electrode systems used in electrocoalescence experiments were of two dimensional type
with uncoated electrodes of 1 mm thickness. A variety of emulsions were investigated in this work, namely water-in-oil, oil-in-oil with more conducting dispersed phase and oil-in-oil with less conducting dispersed phase.
2. Experimental method In the present study the in situ measurement of temporal drop size distribution during electrocoalescence was done using video microscopy. An emulsion in an electric field was continuously recorded. Since both the fluids (dispersed and continuous) in all the emulsions studied are transparent, the droplets in the front plane do not becloud those in the focal plane. Moreover, the thickness of the electrode cell was kept small enough so that the light could pass through the emulsion, making it easier to visualize the maximum number of coalescing droplets; moreover, focusing at different planes allowed access to drops in the inner planes. A variety of methods have been used in the coalescence studies in the literature to estimate the progress of phase separation. A frequently used method involves sampling a small amount of a coalescing emulsion and analyzing it under a microscope [21]. This method can be erroneous as there are chances of further coalescence or breakup of droplets during sampling and off-line imaging [22]. The use of non-uniform electric field in the present study necessitated the visualization at every corner of the field owing to spatial variation of the coalescence rate. Also, taking samples at regular intervals causes interruptions in the electrocoalescence process as well as redistribution of the droplets between the electrodes. A disturbance in the spatial distribution of drops in an electrocoalescing
Initial Size Distribution
Initial Size Distribution V%
N%
10 5 0 0 10
1
10
2
10
0 0 10
3
10
0 0 10
V%
N%
Uniform Field, t = 30 s
5
1
10
2
10
10
3
10
3 2 1 0 0 10
5 0 0 10
V%
N%
Pin−Plate Field, t = 30 s
1
10
2
10
3
10
5
0 0 10
V%
N%
Uniform Field, t = 600 s
10
1
10
2
10
3
5 0 0 10
V%
N%
Pin−Plate Field, t = 600 s
1
2
10 10 Drop Size [µm]
3
10
3 2 1 0 0 10 3 2 1 0 0 10 3 2 1 0 0 10
1
10
2
10
3
10
Uniform Field, t = 30 s
1
10
2
10
3
10
Pin−Plate Field, t = 30 s
1
10
2
10
3
10
Uniform Field, t = 600 s
10
1
10
2
10
3
Pin−Plate Field, t = 600 s
1
2
10 10 Drop Size [µm]
3
10
Fig. 2. Number (N) and volume (V) density based drop size distribution in an electrocoalescing water-in-oil emulsion at times t = 0 s, t = 30 s and t = 600 s. Applied potential, V0 = 400 V and electrode-to-electrode separation, d = 5 mm.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
In the present work, three kinds of emulsions were studied in different types of electrode systems. The emulsions studied and their methods of preparation are given below: i. Water-in-oil emulsion: 10% (v/v) Milli-Q water (density w = 1000 kg/m3 , viscosity w = 1 ×10−3 Pa.s, dielectric constant w = 80 and electrical conductivity w = 20 ×10−6 S/m) dispersed in silicone oil (Merck, density s = 970 kg/m3 , viscosity s = 0.34 Pa.s, dielectric constant s = 2.65 and electrical conductivity s = 0.8 ×10−12 S/m). A water-in-silicone oil emulsion without surfactant is unstable and it was therefore stabilized using a non-ionic, oil soluble surfactant, Span-80 (0.0012 M). The surfactant was first dissolved in to the oil by stirring for 2 min, water was then added slowly and the mixture stirred for another 5 min. The average droplet size in the emulsion was found to be 6 ± 4 m. ii. Oil-in-oil emulsion with dispersed phase less conducting: 10% (v/v) silicone oil emulsified in castor oil (S D Finechem, density c = 940 kg/m3 , viscosity c = 0.78 Pa.s, dielectric constant c = 4.7 and electrical conductivity c = 1 × 10−11 S/m) is essentially a stable system. No stabilizing reagent was used. Silicone oil was added slowly in castor oil while stirring. The average droplet size in the emulsion was 7 ± 4 m. iii. Oil-in-oil emulsion with dispersed phase more conducting: The emulsion was prepared by dispersing 10% (v/v) castor oil in silicone oil using Triton X-100 as a stabilizing reagent. The emulsion with 1 mM surfactant was found to be stable; however, higher concentrations of surfactant resulted in the segregation of droplets. The emulsion was found to be stable for more than 24 h which was also ascertained earlier by Jaitley et al. [23]. The average droplet size in the emulsion was 9 ± 4.5 m.
2.2. Electrode configurations and electric fields The schematics of the electrode systems studied in this work are shown in Fig. 1. The uniform electric field was generated with two copper plates (50 mm (L) × 10 mm (W) × 1 mm (T)) placed horizontal, 5 mm apart and separated by the Plexiglas spacers (Fig. 1(a)).
Arithmetic Mean Diameter [ µm]
2.1. Preparation of emulsions
phase before the addition of dispersed phase. The reproducibility of drop size distribution in a freshly prepared emulsion was ensured.
50
40
30
20
10
0
100
200
300
400
500
600
Time [s] (a) Arithmetic Mean Diameter
Volume Mean Diameter [µm]
emulsion introduces severe errors, particularly in the non-uniform electric fields. Moreover, the estimation of change in the spatial droplet size distribution with time is not possible with the sampling method. These difficulties can be obviated using in situ drop size distribution measurement methods.
31
All the emulsions were prepared using a homogenizer (ULTRATURRAX, IKA) at a speed of 3400 rpm for 5 min. In case of the water-in-oil and oil-in-oil emulsions (with dispersed phase more conducting) surfactants were thoroughly mixed in a continuous
140 120 100 80 60 40 20
0
100
200
300
400
500
600
Time [s]
Sauter Mean Diameter [ µm]
(b) Volume Mean Diameter 300 250 200 150 100 50 0 0
100
200
300
400
500
600
Time [s] (c) Sauter Mean Diameter Fig. 3. Spatially varying drop growth in a water-in-oil emulsion under the pin-plate electric field at time, t = 30 s. Applied potential, V0 = 400 V, electrode-to-electrode separation, d = 5 mm. Electrodes are not visible in the picture.
Fig. 4. Kinetics of electrocoalescence in a water-in-oil emulsion under the uniform (–䊏–) and pin-plate (- -•- -) electric fields. Applied potential, V0 = 400 V, electrodeto-electrode separation, d = 5 mm.
32
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
Uniform Field, t = 30 s
Uniform Field, t = 30 s
6 1.5 V%
N%
4 2 0 0 10 6
1 0.5
1
10
2
10
0 0 10
3
10
Pin−Plate Field, t = 30 s
1
10
2
10
3
10
Pin−Plate Field, t = 30 s 1.5 V%
N%
4 2 0 0 10 6
1 0.5
1
10
2
10
0 0 10
3
10
1
10
2
10
3
10
Four Pin, t = 30 s @ 150 V
Four Pin, t=30s @ 150 V 1.5 V%
N%
4 2 0 0 10
1 0.5
1
10
2
10
3
10
Drop Size [µm]
0 0 10
1
10
2
10
3
10
Drop Size [µm]
Fig. 5. Drop size distribution in a water-in-oil emulsion at t = 30 s in the uniform, pin-plate and four-pin electric fields. Applied potential for uniform and pin-plate electrodes is 400 V over 5 mm separation while for four-pin electrodes V0 = 150 V over 5 mm.
The cell thickness was kept very small (1 mm) so that the maximum number of coalescing drops remain in the depth of field of microscope. The space between the electrodes was covered with two glass slides and emulsion was then poured in to the cavity. In the case of water-in-oil emulsion, when the field was applied for longer time, the coalesced droplets spread over the covering glass slides. To overcome this problem, the glass slides were treated with a siliconizing reagent Sigmacote to make them hydrophobic. The planar quadrupole electrodes (Fig. 1(b)) were made hyperbolic following the equations [24], (r 2 /r02 ) − (2z 2 /r02 ) = 1 and (r 2 /2z02 ) − (z 2 /z02 ) = 1. Here r and z are the radial and axial coor√ dinates, respectively; while, z0 (=2.5 mm) and r0 (= 2z0 ) are the minimum distances of the center from a live electrode and a ground electrode, respectively. The pin-plate electrode system (Fig. 1(c)) was made by replacing a plate in the uniform electrode system by a metal strip with a semicircular end of 1 mm radius. The distance between the plate and pin was kept d = 5 mm. The pin was connected to the power supply while the plate was grounded. In the four-pin electrode setup four pin electrodes of radii 1 mm were arranged in a fashion as shown in Fig. 1(d). Two opposite pin electrodes were connected to the power supply while the remaining two were grounded. In annular electrode systems (Fig. 1(e)) a circular plate (diameter 1 cm) was placed at the center of a disc with a hole at the middle in such a way that the gap between two plates was 5 mm. All the electrodes were 1 mm thick and made of copper. The coalescing emulsion was observed through a glass slide using a Nikon stereoscopic zoom microscope (Nikon, USA). To record the coalescence of the emulsion, a high speed camera (Phantom V12.1, Vision Research, USA) was mounted on the microscope. The emulsion was illuminated using Nikon C-FI230 Fiber illuminator (Nikon, USA). The electric power supply was controlled using a function generator (33220A, Agilent, USA) and the potential was
amplified using the Trek High Voltage Amplifier (Model 5/80, Trek Inc. USA). The images were analyzed using an image processing software, ImageJ 1.46r.
3. Results and discussion The demulsification efficiency in non-uniform electric fields for different kind of emulsions was investigated and the results were compared with that in a uniform electric field.
Fig. 6. Spatially varying drop growth, water bridging and hydrolysis in the gaps between adjacent electrodes in the four-pin electrode system. Time, t = 30 s, applied potential, V0 = 400 V and electrode-to-electrode separation, d = 5 mm. Pin electrode 1 and pin electrode 3 are not visible in the picture.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
Uniform Field, t = 10 minute
33
Uniform Field, t = 10 minute 3
4
V%
N%
6
2 0 0 10
1
10
2
10
2 1 0 0 10
3
10
Pin−Plate Field, t = 10 minute
1
10
2
10
3
10
Pin−Plate Field, t = 10 minute 3
4
V%
N%
6
2 0 0 10
1
10
2
10
2 1 0 0 10
3
10
Quadrupolar Field, t = 10 minute
1
10
2
10
3
10
Quadrupolar Field, t = 10 minute 3
4
V%
N%
6
2 0 0 10
1
10
2
10
2 1 0 0 10
3
10
Drop Size [µm]
1
10
2
10
3
10
Drop Size [µm]
Fig. 7. Drop size distribution in silicone oil-in-castor oil emulsion at t = 10 min under the uniform, pin-plate and quadrupole electric fields. Applied potential is 1 kV over 5 mm separation.
3.1. Coalescence in water-in-oil emulsion The pin-plate and four-pin electrode systems were used to investigate the effect of non-uniformity of the applied electric field on coalescence dynamics of a water-in-oil emulsion. The results thus obtained were compared with electrocoalescence in a uniform electric field. Fig. 2 shows the drop size distribution in a water-in-oil emulsion under uniform and pin-plate electric fields at times t = 30 s and
t = 600 s, respectively. The number density (N) and volume density (V) functions are plotted. The shift of the mean towards the right at t = 30 s suggests the progress in electrocoalescence indicating coalescence of the larger drops and thereby reduction in their number fraction. The histograms for pin-plate and uniform fields at t = 30 s clearly indicate the larger average drop size in a non-uniform field than that in a uniform field. In an electric field, droplets near an electrode coalesce faster due to stronger effective field at the electrode surface. However,
Uniform Field, t = 30 minute
Uniform Field, t = 30 minute 3 V%
N%
4 2 0 0 10
2 1
1
10
2
10
0 0 10
3
10
Pin−Plate Field, t = 30 minute
V%
N%
2
1 1
10
2
10
0 0 10
3
10
1
10
2
10
3
10
Quadrupolar Field, t = 30 minute 3 V%
4 N%
3
10
2
Quadrupolar Field, t = 30 minute
2 0 0 10
2
10
3
4
0 0 10
1
10
Pin−Plate Field, t = 30 minute
2 1
1
10
2
10
Drop Size [µm]
3
10
0 0 10
1
10
2
10
3
10
Drop Size [µm]
Fig. 8. Drop size distribution in silicone oil-in-castor oil emulsion at t = 30 min under the uniform, pin-plate and quadrupole electric fields. Applied potential is 1 kV over 5 mm separation.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
Arithmetic Mean Diameter [ µm]
Fig. 5 shows the normalized droplet size distributions for the water-in-oil emulsion in the uniform, pin-plate and four-pin electrode systems at time t = 30 s after application of an electric field. In the uniform and pin-plate fields, applied electric potential V0 = 400 V; however, in four-pin system V0 = 150 V as the higher applied potential shorts the neighboring pin electrodes (as shown in Fig. 6). The shapes of the histograms of uniform and four-pin electric fields indicate similar drop size distributions, and the average drop size, aavg ≈ 32 m. Although the use of a non-uniform field has an advantage over a uniform field in demulsification of water-in-oil emulsion, care must be taken to avoid certain adverse effects. In the pin-plate electrode system when an applied potential is high, a drop undergoes cyclic motion near the pin electrode. At very strong electric fields the oscillating droplets can break in the tip-streaming mode [26]. Moreover, the strong electrohydrodynamic flows enhance the breakup in high field strength (near pin) region [25]. Similar drop disintegration can be observed in the four-pin electrode setup,
80
60
40
20
0 0
5
10 15 20 Time [minute]
25
30
(a) Arithmetic Mean Diameter
Volume Mean Diameter [ µm]
in the absence of convective flows, resulting large drops shield the droplets in the inner core from the field. As a result coalescence rate varies spatially between the electrodes even in a uniform electric field. In the pin-plate electric field, the coalescence rate varies spatially due to the non-uniformity of applied field (shown in Fig. 3). The droplets near the pin electrode coalesce at a faster rate on account of the high field intensity than those near the plate electrode. Recirculatory electrohydrodynamic flows were observed from right below the pin electrode to the plate and back to the pin. The flow currents generated due to the non-uniformity of applied field continuously replace the bigger drops near the pin electrode with the finer droplets. The flow currents thus keep the drop size distribution fairly uniform throughout the cell. Such kind of flow currents and the continuous displacement of drops were not observed in a uniform electric field. However, at strong electric fields (E0 > 0.4 kV/cm), the coalesced drops near an electrode assume cyclic motion. Such a drop motion near an electrode induces flow currents which resist the fine droplets from reaching to the electrodes in a uniform field. The plots of drop size distribution at a longer time after application of electric field (at t = 600 s in Fig. 2) suggest similar trends. It should be noted that the bin sizes for histograms at t = 600 s are different. The criteria for the bin size selection as well as the definitions of number and volume densities are given in an earlier publication [25]. In an electric field the coalescence rate of an emulsion varies with time. Although, immediately after applying the electric field the coalescence rate is high, it diminishes with time. The initial high rate can be attributed to the high drop density and hence the low inter-drop separation. However, increased inter-drop separation after initial fast coalescence decreases the probability of drop–drop contact. Fig. 4 shows the kinetics of the water-in-oil emulsion coalescence in two types of electric fields i.e. uniform and pin-plate. The arithmetic mean diameter curves are inconclusive (Fig. 4(a)) in demonstrating the relative rate of coalescence in pin-plate and uniform fields. However, the average diameter is certainly higher in pin-plate as compared to uniform field. Fig. 4(b), shows an increase in the volumetric mean diameter of the droplets for both the uniform and pin-plate systems. However, the rate of coalescence is higher in the case of non-uniform electric field. This is also suggested by the normalized distribution functions at different time stages; which indicate a larger fraction of bigger droplets in the pin-plate field as compared to that in a uniform field (Fig. 2). Further, Sauter mean diameter shows a continuous increase with time (Fig. 4(c)) and demonstrates the effectiveness of the pin-plate electrode system in water-in-oil demulsification. In addition to the electrohydrodynamic flow currents, the nonuniform field-induced mobility of droplets is responsible for the higher coalescence rate in the later part of the process. Unlike in uniform electric field, a water drop at any point in the pin-plate electrode field tries to move towards the region of stronger electric field. Dielectrophoretic mobility, thus assists the fast segregation of droplets in the high electric field region. Such a segregation helps to reduce the inter-drop separation and increases the probability of the drop–drop contact. The effectiveness of a symmetric non-uniform electric field in the water-in-oil coalescence is also investigated by using a four-pin electrode setup. The coalescence rate was observed to be higher in the gaps between two adjacent pin electrodes. However, unlike the pin-plate electrode system, electrohydrodynamic flow currents are restricted within the gaps between neighboring pin electrodes. As a result the droplets in this region grow faster while those at the central part coalesce at a slower rate. It has also been observed that at an applied potential above 200 V, the resulting bigger drops extend to the adjacent electrodes and cause short circuit immediately after the application of field.
140 120 100 80 60 40 20 0 0
10 20 Time [minute]
30
(b) Volume Mean Diameter
Sauter Mean Diameter [ µm]
34
350 300 250 200 150 100 50 0 0
10 20 Time [minute]
30
(c) Sauter Mean Diameter Fig. 9. Electrocoalescence kinetics of silicone oil-in-castor oil emulsion in the uniform (–䊏–), pin-plate (- -•- -) and quadrupole (- -- -) electric fields. Applied potential is 1 kV over 5 mm separation.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
35
Fig. 10. Spatial droplets distribution in silicone oil-in-castor oil emulsion under quadrupolar electric field. (a) Shortly after application of field. (b) Segregation at the center, long time after field application.
Fig. 11. Chain formation in silicone oil-in-castor oil emulsion under; (a) quadrupolar electric field and, (b) pin-plate electric field. Chains are encircled.
Initial Size Distribution
Initial Size Distribution
10
10
V%
N%
15
5 0 0 10 6
1
2
10 10 Uniform Field, t = 60 s
0 0 10
3
10
V%
N%
2
6
1
2
10 10 Annular Field, t = 60 s
1
2
1
2
3
10
1 10 10 Annular Field, t = 60 s
3
10
3 V%
N%
2
2
0 0 10
3
10
4 2 0 0 10
1
10 10 Uniform Field, t = 60 s
3
4
0 0 10
5
1
2
10 10 Drop Size [µm]
3
10
2 1 0 0 10
10 10 Drop Size [µm]
3
10
Fig. 12. Drop size distribution of castor oil-in-silicone oil emulsion at t = 60 s under uniform and annular electric fields. Applied potential is 400 V over 5 mm electrode separation.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
An oil-in-oil emulsion with less conducting dispersed phase needs stronger electric field and longer time to break compared to a water-in-oil emulsion due to a weaker dipolar interaction force [6]. In this work, uniform, pin-plate and quadrupole electric fields were used to coalesce the less conducting silicone oil droplets dispersed in a more conducting castor oil medium. Fig. 7 suggests nearly normal drop size distributions in uniform and quadrupolar electric fields; however, emulsion in the pin-plate field is more polydispersed. The right skewness of the distribution in pin-plate electric field indicates the presence of a large number of small size droplets with some larger drops in the electrocoalescing emulsion. The symmetric non-uniform electric field in the quadrupole electrode system favors electrocoalescence in an oil-in-oil emulsion with less conducting dispersed phase over the uniform field. However, the asymmetric electric field generated by the pin-plate electrode system is not much efficient. The drop size distribution at t = 30 min (Fig. 8) suggests similar trends at a long time after field application. The lower rate of coalescence of a silicone oil-in-castor oil emulsion in the pin-plate field compared to the uniform and symmetric non-uniform electric fields is evident from the kinetics curves in Fig. 9. Similar to the water-in-oil emulsion, electrocoalescence rate immediately after the application of electric field is higher in the case of oil-in-oil emulsion with dispersed phase less conducting. As coalescence progresses, increase in the inter-drop separation results in asymptotic slowing down in the rate. It is evident that, the coalescence is faster in the quadrupolar field followed by uniform and pin-plate electrode systems. The lower electrocoalescence in a pin-plate electrode system is attributed to the segregation of droplets near the plate electrode (a low electric field strength region) and the absence of electrohydrodynamic flow currents. The emergence of big drops suggests electrocoalescence in the high field strength region near the pin electrode. However, unlike the water-in-oil emulsion, negative dielectophoretic leaky dielectric system of silicone oil drops in castor oil does not admit strong electrohydrodynamic flow currents. The dielectrophoretic motion of the coalesced drops is unidirectional (pin to plate) and that too is not strong enough to keep the drop size distribution uniform. In the quadrupole electrode system, silicone oil droplets migrate towards the central region. Moreover, such a drop movement is assisted by the circulatory flows generated between two neighboring electrodes. However, low electric field strength at the center leads to a lower coalescence in the beginning. This results in a toroidal region between the center and the electrodes where the coalescence rate is observed to be high (Fig. 10(a)). As the coalescence progresses, larger drops accumulate at the center of the quadrupolar electric field as shown in Fig. 10(b). In an electrocoalescing emulsion of silicone oil-in-castor oil, droplets try to arrange themselves in chains. Such chains were observed in both uniform as well as non-uniform electric fields. However, the phenomenon is more pronounced in the non-uniform electric fields. Fig. 11(a) shows such a chain formation in the silicone oil-in-castor oil emulsion under a quadrupolar electric field. The chains are short in length and do not extend to the electrodes. In quadrupolar electric fields, the chains are curved and arranged along the isopotential lines. However, at a higher applied
3.3. Coalescence in oil-in-oil emulsion with dispersed phase more conducting A castor oil drop dispersed in the silicone oil exhibits positive dielectrophoresis owing to a positive Clasius-Mossotti factor [26]. We used uniform and annular electric fields to demonstrate the effect of non-uniformity of an applied field on the coalescence rate in a positive dielectrophoretic LD–LD fluid system. The use of pin-plate electrode system was avoided based on our earlier observations of the castor oil drop breakup and emulsification on and around the pin electrode [25]. Arithmetic Mean Diameter [ µm]
3.2. Coalescence in oil-in-oil emulsion with dispersed phase less conducting
potential, flow circulations between the two adjacent electrodes prevent chain formation and also push the coalesced drops towards the center. In a pin-plate electric field, the silicone oil droplets form small chains away from the electrodes (Fig. 11(b)); whereas, in a uniform field, chain formation was observed at much longer times after application of field.
50 40 30 20 10 0 0
50
100
150
200
250
300
Time [s]
(a) Arithmetic Mean Diameter Volume Mean Diameter [µm]
either due to the tip-streaming near an electrode or bridging of the adjacent electrodes. These types of drop breakup not only lower the coalescence rate but also introduce tiny droplets in the emulsion, which are difficult to coalescence even after application of electric field for a very longer time.
80 70 60 50 40 30 20 10
0
50
100
150
200
250
300
Time [s]
(b) Volume Mean Diameter 200
Sauter Mean Diameter [µm]
36
150
100
50
0 0
50
100
150
200
250
300
Time [s]
(c) Sauter Mean Diameter Fig. 13. Electrocoalescence kinetics of castor oil-in-silicone oil emulsion under uniform (–䊏–) and annular (––) electric fields. Applied potential V0 = 400 V over 5 mm electrode separation.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
37
Fig. 14. Drop distribution and oil layer formation in castor oil-in-silicone oil emulsion during coalescence under annular electric field. Applied potential V0 = 400 V over 5 mm electrode separation.
The drop size distributions in an electrocoalescing castor oilin-silicone oil emulsion at time t = 60 s in the uniform and annular electric fields are shown in Fig. 12. The nature of histograms suggests that an electrocoalescing emulsion in an annular electric field is more polydispersed compared to that in a uniform field. The average of the distribution falls in the same range in charts for both types of fields. The average droplet size of an emulsion at the different times after application of field in the uniform as well as annular electric fields is plotted in Fig. 13. Similar to the water-in-oil and oil-in-oil (with dispersed phase less conducting) emulsions, in a castor oil-insilicone oil emulsion droplets coalesce at a faster rate immediately after the field is applied; but the rate of coalescence subsides with time. The gradual decrease in the coalescence rate can be attributed to the coalescence and spreading of castor oil drops on the electrode surface. The coalesced drops near an electrode cover a large part of its surface with a thick oil layer. The oil layer acts as an insulation, which shields the emulsion drops from the electric field. In a uniform field such type of oil layer is observed on the live electrode. However, similar layer is observed in the annular electric field, but over the outer i.e. ground electrode. As shown in Fig. 14, in an annular field, emulsion droplets coalesce faster near the electrodes and the coalesced big drops remain adhered to the electrode surface. A layer of oil on the ground electrode and big drops adhering to the live electrode diminish the field in the inner parts of the annular field resulting in slower coalescence after the initial faster rate.
4. Conclusions The experimental observations of coalescence behavior in uniform and non-uniform electric fields are reported for the water-in-oil and oil-in-oil emulsions. In the water-in-oil emulsion, asymmetric non-uniform field generated using the pin-plate electrodes is found to be advantageous as compared to the uniform and symmetric quadrupole non-uniform electric fields. Seemingly, the electrohydrodynamic flows generated due to the applied non-uniform field as well as the dielectrophoretic drop motion promote the coalescence rate. However, the strong electrohydrodynamic flows and drop instability at the high applied electric fields adversely affect the coalescence dynamics. The use of coated electrodes to induce an asymmetric non-uniform electric field below critical strength is suggested to acquire the optimum coalescence rate. In an oil-in-oil emulsion with less conducting dispersed phase, symmetric non-uniform field generated using quadrupolar electrodes results in the faster coalescence than the uniform and asymmetric non-uniform electric fields in pin-plate electrodes.
Whereas, in an oil-in-oil emulsion with more conducting dispersed phase, the annular electric field is more promising than the uniform field. We thus demonstrate that the electrical properties of the dispersed and continuous phases decide the efficacy of an electrode system in the electrocoalescence of an emulsion.
References [1] K. Klasson, P. Taylor, J. Walker, S. Jones, R. Cummins, S. Richardson, Modification of a centrifugal separator for in-well oil-water separation, Sep. Sci. Technol. 40 (2005) 453–462. [2] D. Sun, S. Jong, X. Duan, D. Zhou, Demulsification of water-in-oil emulsion by wetting coalescence materials in stirred- and packed-columns, Colloid Surf. A 150 (1999) 69–75. [3] J. Wu, Y. Xu, T. Dabros, H. Hamza, Effect of demulsifier properties on destabilization of water-in-oil emulsion, Energy Fuel 17 (2003) 1554–1559. [4] P. Atten, Electrocoalescence of water droplets in an insulating liquid, J. Electrostat. 30 (1993) 259–270. [5] S. Mhatre, V. Vivacqua, M. Ghadiri, A. Abdullah, M. Al-Marri, A. Hassanpour, B. Hewakandamby, B. Azzopardi K.B., Electrostatic phase separation: a review, Chem. Eng. Res. Des. 96 (2015) 177–195. [6] S. Mhatre, S. Deshmukh, R.M. Thaokar, Electrocoalescence of a drop pair, Phys. Fluid. (forthcoming). [7] T. Hano, T. Ohtake, K. Takagi, Demulsification kinetics of w/o emulsion in an AC electric field, J. Chem. Eng. Jpn. 21 (4) (1988) 345–351. [8] C. Lesaint, W.R. Glomm, L.E. Lundgaard, J. Sjoblom, Dehydration efficiency of ac electrical fields on water-in-model-oil emulsions, Colloid Surf. A 352 (2009) 63–69. [9] G. Berg, L.E. Lundgaard, N. Abi-Chebel, Electrically stressed water drops in oil, Chem. Eng. Process. 49 (2010) 1229–1240. [10] P. Suemar, E.F. Fonseca, R.C. Coutinho, F. Machado, F. Fontes, L.C. Ferreira, E.L. Lima, P.L. Melo, J. Pinto, M. Nele, Quantitative evaluation of the efficiency of water-in-crude-oil emulsion dehydration by electrocoalescence in pilot-plant and full-scale units, Ind. Eng. Chem. Res. 51 (2012) 13423–13437. [11] C.A.R. Pearce, The mechanism of the resolution of water-in-oil emulsions by electrical treatment, Br. J. Appl. Phys. 5 (1953) 136–143. [12] C. Bezemer, G.A. Croes, Motion of water droplets of an emulsion in a nonuniform field, Br. J. Appl. Phys. 6 (1955) 224. [13] H. Aryafar, H. Kavehpour, Drop coalescence through planar surfaces, Phys. Fluid 18 (2006) 072105–072106. [14] P. Atten, Electrohydrodynamics of dispersed drops of conducting liquid: from drop deformation and interaction to emulsion evolution, in: Proceedings of the International Symposium on Electrohydrodynamics, Gdansk, Poland, 2012. [15] G.I. Taylor, Disintegration of water drops in an electric field, Proc. R. Soc. Lond. Ser. A 280 (1964) 383–397. [16] T.C. Scott, R.M. Wham, An electrically driven multistage countercurrent solvent extraction device: the emulsion-phase contactor, Ind. Eng. Chem. Res. 28 (1989) 94–97. [17] H. Block, J.P. Kelly, Electro-rheology, J. Phys. D: Appl. Phys. 21 (1988) 1661–1677. [18] C.F. Zukoski, Material properties and the electrorheological response, Annu. Rev. Mater. Sci. 23 (1993) 45–78. [19] J.S. Eow, M. Ghadiri, Drop–drop coalescence in an electric field: the effects of applied electric field and electrode geometry, Colloids Surf. A: Physicochem. Eng. Asp. 219 (2003) 253–279. [20] C. Noik, J. Trapy, A. Mouret, G. Laborie, Design of a crude oil dehydration unit, in: SPE Annual Technical Conference and Exhibition, 2002. [21] H. Wang, R. Davis, Experiments on phase separation of dilute dispersions of coalescing drops, JCIS 181 (1996) 93–98.
38
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
[22] N.V.T. Opedal, G. Sorland, J. Sjoblom, Methods for droplet size distribution determination of water-in-oil emulsions using low-field NMR, diffusionfundamentals.org 9 (2009) 1–29. [23] V. Jaitely, T. Sakthivel, G. Magee, A. Florence, Formulation of oil in oil emulsions: potential drug reservoirs for slow release, J. Drug Deliv. Sci. Technol. 14 (2004) 113–117. [24] R.D. Knight, The general form of the quadrupole ion trap potential, IJMSIP 51 (1983) 127–131.
[25] S. Mhatre, R. Thaokar, Pin-plate electrode system for emulsification of a higher conductivity leaky dielectric liquid into a low conductivity medium, Ind. Eng. Chem. Res. 53 (2014) 13488–13496. [26] S. Mhatre, R.M. Thaokar, Drop motion, deformation, and cyclic motion in a non-uniform electric field in the viscous limit, Phys. Fluids 25 (2013) 072105.
Contents lists available at ScienceDirect
Chemical Engineering and Processing: Process Intensification journal homepage: www.elsevier.com/locate/cep
Electrocoalescence in non-uniform electric fields: An experimental study Sameer Mhatre, Rochish Thaokar ∗ Department of Chemical Engineering, Indian Institute of Technology Bombay, Mumbai, Maharashtra 400076, India
a r t i c l e
i n f o
Article history: Received 30 May 2015 Received in revised form 24 July 2015 Accepted 25 July 2015 Available online 29 July 2015 Keywords: Electrocoalescence Phase separation Crude oil demulsification Electrohydrodynamics Non-uniform electric field Dielectrophoresis
a b s t r a c t The method of electrostatic force-assisted phase separation is widely used in the chemical industry. The fast and clean separation capabilities make electrocoalescence a favorite among the other competing phase separation techniques. In the present study, we experimentally investigated the relative merits of symmetric and asymmetric non-uniform and uniform electric fields in electrocoalescence. Three different types of emulsions are studied: (i) a perfectly conducting (PC) phase dispersed in a perfectly dielectric (PD) medium, (ii) a leaky dielectric (LD) phase dispersed in another LD phase of higher electrical conductivity, and (iii) a LD phase dispersed in another LD phase of lower electrical conductivity. The coalescence behavior of each emulsion in the non-uniform electric fields is analyzed and the results are compared with that in a uniform field. The observations show that a non-uniformity as well as asymmetry of electric field can affect electrocoalescence in a nontrivial way and is sensitive to the drop-medium system, specifically pin-plate, quadrupole and annular electrode configurations are found to be advantageous over uniform field in PC drops in PD medium, less conducting LD drops in more conducting LD medium and more conducting LD drops in less conducting LD medium, respectively. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Stable emulsions have been an issue of concern in many chemical processes. In the petroleum industry, naturally stable emulsion of water in crude oil can pose several problems during its processing. In liquid–liquid extraction, a stable mixture of extract and raffinate phases take a long time to separate by gravity. There exist many techniques of phase separation; e.g. gravitational or buoyancy separation, centrifugation, chemical treatment, membrane separation, etc. [1–3]. However each of these conventional techniques has some disadvantages; e.g. gravity settling takes long time for dispersed phase to settle if the drops are very small in size, chemical treatment affects the quality of separated water and also causes pollution, mechanical energy and thereby shear may not be very efficient in coalescence, etc. The use of electrostatic energy has been considered as one of the most effective and clean methods of phase separation. In addition to separation of water to a desired level, low energy consumption and lack of necessity of further purification of the separated phases make the electrostatic method worthy. Electrocoalescence is the de-facto method in
∗ Corresponding author. E-mail address: [email protected] (R. Thaokar). http://dx.doi.org/10.1016/j.cep.2015.07.025 0255-2701/© 2015 Elsevier B.V. All rights reserved.
desalting operations wherein minute water droplets of size below 50 m are separated from crude oil [4]. Started in early 20th century, extensive research has resulted in fundamental understanding of many complex problems in electrocoalescence that have helped to make the technology more efficient, faster and the devices more compact [5]. Applying an electric field induces free charge separation in a conducting water drop resulting in a dipole and the consequent dipole–dipole interaction between neighboring drops in an emulsion enhances the coalescence rate. In addition to the dipolar attraction, an applied electric field can induce macroscopic flow currents due to movement of drops or due to tangential stresses in leaky dielectric systems which increase the probability of inter-drop contact. The coalescence of two drops in an electric field is assume to occur in three stages [5,6]; in the first stage, two drops are attracted towards each other on application of the field. The second stage involves squeezing of the medium fluid from the film between leading faces of the approaching drops. When such an interstitial film becomes microscopically thin, its break up and consequent coalescence of two drops occurs in the third stage. There are several parameters that determine the coalescence dynamics in the presence of an electric field. The factors such as the magnitude and type of an applied field, electrical and physical properties of the fluids, impurities in the emulsion, flow or turbulence, drop size, dispersity, etc., govern the kinetics of electrocoalescence. The role of such parameters has
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
been a topic of investigation in the majority of scientific studies in electrocoalescence [7–10]. Application of moderate electric field to an emulsion leads to several phenomena such as chain formation, partial coalescence, non-coalescence and receding of about to coalesce drops, drop breakup, etc., which are known to adversely affect the rate of electrocoalescence [5]. In an electrocoalescing emulsion, drops can arrange into chains aligned in the direction of the applied field [11,12]. The interfacial stability of the drop interface often promotes chain formation. Such chains of droplets not only slow down the coalescence rate, but can also extend to the electrodes resulting in to short circuiting. The formation of fine progeny droplets during drop-interface or drop–drop coalescence is termed as partial coalescence [13]. This phenomenon introduces new drops in the emulsion which are difficult to separate owing to their very small sizes. The application of an electric field induces drop–drop approach, but cannot guarantee coalescence. It has been observed that in certain conditions (especially under strong applied electric fields), coalescing drops retreat after their contact [14]. In another undesired effect, after coalescence, if the field at the resulting drop interface exceeds the critical field limit, Ec = 0.648 /2am , a drop can further break into smaller droplets [15]. Here, is interfacial tension, a is drop radius and m is permittivity of medium fluid. Such adverse effects need to be overcome to speedup the electrocoalescence process.
Glass Slides
29
In addition to water separation from crude oil, electric field can also be employed for phase separation in systems involving a leaky dielectric (LD) fluid dispersed in to another LD medium. The examples include fractionation of mixed oils [16], electrorheology of electro-magneto responsive fluids, polymer blends, etc. Electrorheological (ER) fluids are LD–LD suspensions which display a change in rheological properties e.g. shear thinning or thickening under the influence of an electric field [17,18]. Such fluids are used in the active control devices such as dampers, shock absorbers, clutches, brakes, etc. The preparation of a polymer blend involves dispersing one polymer in to another in order to acquire the desired properties from the resulting new polymer. Control over the droplet size in a blend plays a key role in deciding the properties of the resulting polymer. Electric fields can also be used to precisely regulate the dispersed phase drop-size distribution. Furthermore, during recycling, electrocoalescence can be used to separate a dispersed polymer phase from the molten blend. An in-depth understanding of the electrocoalescence in an oil-in-oil emulsion can help to control the size distributions in the ER fluids as well as in the polymer blends. Electrocoalescence has been construed as separation of two liquid phases by applying uniform electric field. The electric field employed in the previous electrocoalescence studies was predominantly of uniform kind. Very few studies in the literature discuss the use of non-uniform electric fields for phase separation. Pearce
Glass Slides d
d
Spacer Electrodes
(a) Parallel plate electrodes
(b) Quadrupole electrode system V
d d
Glass Slides d
Spacer Electrodes
(c) Pin-plate electrode system
(d) Four-pin electrode system
d
(e) Annular electrode system Fig. 1. Schematics of various electrode systems used to induce the uniform and non-uniform electric fields. Electrode-electrode separation, d = 5 mm in all the electrode configurations.
30
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
[11] used the electric field generated by the coaxial cylindrical electrodes for breaking a water-in-oil emulsion. In another study, Eow and Ghadiri [19] suggested five different electrode designs to study the effect of spatial distribution of an applied field on coalescence of two drops in a flowing medium. They also studied the effect of applied field and frequency on the oscillation of a moving drop in the suggested electrode systems. In a pilot plant-scale study, Noik et al. [20] designed a compact coalescer employing two annular cylindrical electrodes and centrifugal flow of the emulsion. The roles of electric field as well as flow field in electrocoalescence were investigated. The shear force or turbulence due to the hydrodynamic conditions induce drop break-up or disturb the dipolar interactions between the drops [20]. Non-uniform electric fields have been extensively used in the segregation and separation of rigid- as well as bio-particles. Depending upon the polarizability of the phases involved, the non-uniformity of an applied field forces the suspending particles to assemble in a certain region of the field. This phenomenon, known as dielectrophoresis, can also be utilized to expedite the approach of two droplets which is the first and crucial step of the coalescence in an emulsion. The purpose of the present study is to investigate the prospects of the applied non-uniform electric fields in electrocoalescence over a uniform field. Different kinds of electrode systems were used to generate the non-uniform electric fields, viz. quadrupole, pin-plate, four-pin and annular electrodes. The non-uniform fields induced in the quadrupole, four-pin and annular electrode systems are essentially symmetric; whereas, the pin-plate system generates a highly asymmetric electric field. All the electrode systems used in electrocoalescence experiments were of two dimensional type
with uncoated electrodes of 1 mm thickness. A variety of emulsions were investigated in this work, namely water-in-oil, oil-in-oil with more conducting dispersed phase and oil-in-oil with less conducting dispersed phase.
2. Experimental method In the present study the in situ measurement of temporal drop size distribution during electrocoalescence was done using video microscopy. An emulsion in an electric field was continuously recorded. Since both the fluids (dispersed and continuous) in all the emulsions studied are transparent, the droplets in the front plane do not becloud those in the focal plane. Moreover, the thickness of the electrode cell was kept small enough so that the light could pass through the emulsion, making it easier to visualize the maximum number of coalescing droplets; moreover, focusing at different planes allowed access to drops in the inner planes. A variety of methods have been used in the coalescence studies in the literature to estimate the progress of phase separation. A frequently used method involves sampling a small amount of a coalescing emulsion and analyzing it under a microscope [21]. This method can be erroneous as there are chances of further coalescence or breakup of droplets during sampling and off-line imaging [22]. The use of non-uniform electric field in the present study necessitated the visualization at every corner of the field owing to spatial variation of the coalescence rate. Also, taking samples at regular intervals causes interruptions in the electrocoalescence process as well as redistribution of the droplets between the electrodes. A disturbance in the spatial distribution of drops in an electrocoalescing
Initial Size Distribution
Initial Size Distribution V%
N%
10 5 0 0 10
1
10
2
10
0 0 10
3
10
0 0 10
V%
N%
Uniform Field, t = 30 s
5
1
10
2
10
10
3
10
3 2 1 0 0 10
5 0 0 10
V%
N%
Pin−Plate Field, t = 30 s
1
10
2
10
3
10
5
0 0 10
V%
N%
Uniform Field, t = 600 s
10
1
10
2
10
3
5 0 0 10
V%
N%
Pin−Plate Field, t = 600 s
1
2
10 10 Drop Size [µm]
3
10
3 2 1 0 0 10 3 2 1 0 0 10 3 2 1 0 0 10
1
10
2
10
3
10
Uniform Field, t = 30 s
1
10
2
10
3
10
Pin−Plate Field, t = 30 s
1
10
2
10
3
10
Uniform Field, t = 600 s
10
1
10
2
10
3
Pin−Plate Field, t = 600 s
1
2
10 10 Drop Size [µm]
3
10
Fig. 2. Number (N) and volume (V) density based drop size distribution in an electrocoalescing water-in-oil emulsion at times t = 0 s, t = 30 s and t = 600 s. Applied potential, V0 = 400 V and electrode-to-electrode separation, d = 5 mm.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
In the present work, three kinds of emulsions were studied in different types of electrode systems. The emulsions studied and their methods of preparation are given below: i. Water-in-oil emulsion: 10% (v/v) Milli-Q water (density w = 1000 kg/m3 , viscosity w = 1 ×10−3 Pa.s, dielectric constant w = 80 and electrical conductivity w = 20 ×10−6 S/m) dispersed in silicone oil (Merck, density s = 970 kg/m3 , viscosity s = 0.34 Pa.s, dielectric constant s = 2.65 and electrical conductivity s = 0.8 ×10−12 S/m). A water-in-silicone oil emulsion without surfactant is unstable and it was therefore stabilized using a non-ionic, oil soluble surfactant, Span-80 (0.0012 M). The surfactant was first dissolved in to the oil by stirring for 2 min, water was then added slowly and the mixture stirred for another 5 min. The average droplet size in the emulsion was found to be 6 ± 4 m. ii. Oil-in-oil emulsion with dispersed phase less conducting: 10% (v/v) silicone oil emulsified in castor oil (S D Finechem, density c = 940 kg/m3 , viscosity c = 0.78 Pa.s, dielectric constant c = 4.7 and electrical conductivity c = 1 × 10−11 S/m) is essentially a stable system. No stabilizing reagent was used. Silicone oil was added slowly in castor oil while stirring. The average droplet size in the emulsion was 7 ± 4 m. iii. Oil-in-oil emulsion with dispersed phase more conducting: The emulsion was prepared by dispersing 10% (v/v) castor oil in silicone oil using Triton X-100 as a stabilizing reagent. The emulsion with 1 mM surfactant was found to be stable; however, higher concentrations of surfactant resulted in the segregation of droplets. The emulsion was found to be stable for more than 24 h which was also ascertained earlier by Jaitley et al. [23]. The average droplet size in the emulsion was 9 ± 4.5 m.
2.2. Electrode configurations and electric fields The schematics of the electrode systems studied in this work are shown in Fig. 1. The uniform electric field was generated with two copper plates (50 mm (L) × 10 mm (W) × 1 mm (T)) placed horizontal, 5 mm apart and separated by the Plexiglas spacers (Fig. 1(a)).
Arithmetic Mean Diameter [ µm]
2.1. Preparation of emulsions
phase before the addition of dispersed phase. The reproducibility of drop size distribution in a freshly prepared emulsion was ensured.
50
40
30
20
10
0
100
200
300
400
500
600
Time [s] (a) Arithmetic Mean Diameter
Volume Mean Diameter [µm]
emulsion introduces severe errors, particularly in the non-uniform electric fields. Moreover, the estimation of change in the spatial droplet size distribution with time is not possible with the sampling method. These difficulties can be obviated using in situ drop size distribution measurement methods.
31
All the emulsions were prepared using a homogenizer (ULTRATURRAX, IKA) at a speed of 3400 rpm for 5 min. In case of the water-in-oil and oil-in-oil emulsions (with dispersed phase more conducting) surfactants were thoroughly mixed in a continuous
140 120 100 80 60 40 20
0
100
200
300
400
500
600
Time [s]
Sauter Mean Diameter [ µm]
(b) Volume Mean Diameter 300 250 200 150 100 50 0 0
100
200
300
400
500
600
Time [s] (c) Sauter Mean Diameter Fig. 3. Spatially varying drop growth in a water-in-oil emulsion under the pin-plate electric field at time, t = 30 s. Applied potential, V0 = 400 V, electrode-to-electrode separation, d = 5 mm. Electrodes are not visible in the picture.
Fig. 4. Kinetics of electrocoalescence in a water-in-oil emulsion under the uniform (–䊏–) and pin-plate (- -•- -) electric fields. Applied potential, V0 = 400 V, electrodeto-electrode separation, d = 5 mm.
32
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
Uniform Field, t = 30 s
Uniform Field, t = 30 s
6 1.5 V%
N%
4 2 0 0 10 6
1 0.5
1
10
2
10
0 0 10
3
10
Pin−Plate Field, t = 30 s
1
10
2
10
3
10
Pin−Plate Field, t = 30 s 1.5 V%
N%
4 2 0 0 10 6
1 0.5
1
10
2
10
0 0 10
3
10
1
10
2
10
3
10
Four Pin, t = 30 s @ 150 V
Four Pin, t=30s @ 150 V 1.5 V%
N%
4 2 0 0 10
1 0.5
1
10
2
10
3
10
Drop Size [µm]
0 0 10
1
10
2
10
3
10
Drop Size [µm]
Fig. 5. Drop size distribution in a water-in-oil emulsion at t = 30 s in the uniform, pin-plate and four-pin electric fields. Applied potential for uniform and pin-plate electrodes is 400 V over 5 mm separation while for four-pin electrodes V0 = 150 V over 5 mm.
The cell thickness was kept very small (1 mm) so that the maximum number of coalescing drops remain in the depth of field of microscope. The space between the electrodes was covered with two glass slides and emulsion was then poured in to the cavity. In the case of water-in-oil emulsion, when the field was applied for longer time, the coalesced droplets spread over the covering glass slides. To overcome this problem, the glass slides were treated with a siliconizing reagent Sigmacote to make them hydrophobic. The planar quadrupole electrodes (Fig. 1(b)) were made hyperbolic following the equations [24], (r 2 /r02 ) − (2z 2 /r02 ) = 1 and (r 2 /2z02 ) − (z 2 /z02 ) = 1. Here r and z are the radial and axial coor√ dinates, respectively; while, z0 (=2.5 mm) and r0 (= 2z0 ) are the minimum distances of the center from a live electrode and a ground electrode, respectively. The pin-plate electrode system (Fig. 1(c)) was made by replacing a plate in the uniform electrode system by a metal strip with a semicircular end of 1 mm radius. The distance between the plate and pin was kept d = 5 mm. The pin was connected to the power supply while the plate was grounded. In the four-pin electrode setup four pin electrodes of radii 1 mm were arranged in a fashion as shown in Fig. 1(d). Two opposite pin electrodes were connected to the power supply while the remaining two were grounded. In annular electrode systems (Fig. 1(e)) a circular plate (diameter 1 cm) was placed at the center of a disc with a hole at the middle in such a way that the gap between two plates was 5 mm. All the electrodes were 1 mm thick and made of copper. The coalescing emulsion was observed through a glass slide using a Nikon stereoscopic zoom microscope (Nikon, USA). To record the coalescence of the emulsion, a high speed camera (Phantom V12.1, Vision Research, USA) was mounted on the microscope. The emulsion was illuminated using Nikon C-FI230 Fiber illuminator (Nikon, USA). The electric power supply was controlled using a function generator (33220A, Agilent, USA) and the potential was
amplified using the Trek High Voltage Amplifier (Model 5/80, Trek Inc. USA). The images were analyzed using an image processing software, ImageJ 1.46r.
3. Results and discussion The demulsification efficiency in non-uniform electric fields for different kind of emulsions was investigated and the results were compared with that in a uniform electric field.
Fig. 6. Spatially varying drop growth, water bridging and hydrolysis in the gaps between adjacent electrodes in the four-pin electrode system. Time, t = 30 s, applied potential, V0 = 400 V and electrode-to-electrode separation, d = 5 mm. Pin electrode 1 and pin electrode 3 are not visible in the picture.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
Uniform Field, t = 10 minute
33
Uniform Field, t = 10 minute 3
4
V%
N%
6
2 0 0 10
1
10
2
10
2 1 0 0 10
3
10
Pin−Plate Field, t = 10 minute
1
10
2
10
3
10
Pin−Plate Field, t = 10 minute 3
4
V%
N%
6
2 0 0 10
1
10
2
10
2 1 0 0 10
3
10
Quadrupolar Field, t = 10 minute
1
10
2
10
3
10
Quadrupolar Field, t = 10 minute 3
4
V%
N%
6
2 0 0 10
1
10
2
10
2 1 0 0 10
3
10
Drop Size [µm]
1
10
2
10
3
10
Drop Size [µm]
Fig. 7. Drop size distribution in silicone oil-in-castor oil emulsion at t = 10 min under the uniform, pin-plate and quadrupole electric fields. Applied potential is 1 kV over 5 mm separation.
3.1. Coalescence in water-in-oil emulsion The pin-plate and four-pin electrode systems were used to investigate the effect of non-uniformity of the applied electric field on coalescence dynamics of a water-in-oil emulsion. The results thus obtained were compared with electrocoalescence in a uniform electric field. Fig. 2 shows the drop size distribution in a water-in-oil emulsion under uniform and pin-plate electric fields at times t = 30 s and
t = 600 s, respectively. The number density (N) and volume density (V) functions are plotted. The shift of the mean towards the right at t = 30 s suggests the progress in electrocoalescence indicating coalescence of the larger drops and thereby reduction in their number fraction. The histograms for pin-plate and uniform fields at t = 30 s clearly indicate the larger average drop size in a non-uniform field than that in a uniform field. In an electric field, droplets near an electrode coalesce faster due to stronger effective field at the electrode surface. However,
Uniform Field, t = 30 minute
Uniform Field, t = 30 minute 3 V%
N%
4 2 0 0 10
2 1
1
10
2
10
0 0 10
3
10
Pin−Plate Field, t = 30 minute
V%
N%
2
1 1
10
2
10
0 0 10
3
10
1
10
2
10
3
10
Quadrupolar Field, t = 30 minute 3 V%
4 N%
3
10
2
Quadrupolar Field, t = 30 minute
2 0 0 10
2
10
3
4
0 0 10
1
10
Pin−Plate Field, t = 30 minute
2 1
1
10
2
10
Drop Size [µm]
3
10
0 0 10
1
10
2
10
3
10
Drop Size [µm]
Fig. 8. Drop size distribution in silicone oil-in-castor oil emulsion at t = 30 min under the uniform, pin-plate and quadrupole electric fields. Applied potential is 1 kV over 5 mm separation.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
Arithmetic Mean Diameter [ µm]
Fig. 5 shows the normalized droplet size distributions for the water-in-oil emulsion in the uniform, pin-plate and four-pin electrode systems at time t = 30 s after application of an electric field. In the uniform and pin-plate fields, applied electric potential V0 = 400 V; however, in four-pin system V0 = 150 V as the higher applied potential shorts the neighboring pin electrodes (as shown in Fig. 6). The shapes of the histograms of uniform and four-pin electric fields indicate similar drop size distributions, and the average drop size, aavg ≈ 32 m. Although the use of a non-uniform field has an advantage over a uniform field in demulsification of water-in-oil emulsion, care must be taken to avoid certain adverse effects. In the pin-plate electrode system when an applied potential is high, a drop undergoes cyclic motion near the pin electrode. At very strong electric fields the oscillating droplets can break in the tip-streaming mode [26]. Moreover, the strong electrohydrodynamic flows enhance the breakup in high field strength (near pin) region [25]. Similar drop disintegration can be observed in the four-pin electrode setup,
80
60
40
20
0 0
5
10 15 20 Time [minute]
25
30
(a) Arithmetic Mean Diameter
Volume Mean Diameter [ µm]
in the absence of convective flows, resulting large drops shield the droplets in the inner core from the field. As a result coalescence rate varies spatially between the electrodes even in a uniform electric field. In the pin-plate electric field, the coalescence rate varies spatially due to the non-uniformity of applied field (shown in Fig. 3). The droplets near the pin electrode coalesce at a faster rate on account of the high field intensity than those near the plate electrode. Recirculatory electrohydrodynamic flows were observed from right below the pin electrode to the plate and back to the pin. The flow currents generated due to the non-uniformity of applied field continuously replace the bigger drops near the pin electrode with the finer droplets. The flow currents thus keep the drop size distribution fairly uniform throughout the cell. Such kind of flow currents and the continuous displacement of drops were not observed in a uniform electric field. However, at strong electric fields (E0 > 0.4 kV/cm), the coalesced drops near an electrode assume cyclic motion. Such a drop motion near an electrode induces flow currents which resist the fine droplets from reaching to the electrodes in a uniform field. The plots of drop size distribution at a longer time after application of electric field (at t = 600 s in Fig. 2) suggest similar trends. It should be noted that the bin sizes for histograms at t = 600 s are different. The criteria for the bin size selection as well as the definitions of number and volume densities are given in an earlier publication [25]. In an electric field the coalescence rate of an emulsion varies with time. Although, immediately after applying the electric field the coalescence rate is high, it diminishes with time. The initial high rate can be attributed to the high drop density and hence the low inter-drop separation. However, increased inter-drop separation after initial fast coalescence decreases the probability of drop–drop contact. Fig. 4 shows the kinetics of the water-in-oil emulsion coalescence in two types of electric fields i.e. uniform and pin-plate. The arithmetic mean diameter curves are inconclusive (Fig. 4(a)) in demonstrating the relative rate of coalescence in pin-plate and uniform fields. However, the average diameter is certainly higher in pin-plate as compared to uniform field. Fig. 4(b), shows an increase in the volumetric mean diameter of the droplets for both the uniform and pin-plate systems. However, the rate of coalescence is higher in the case of non-uniform electric field. This is also suggested by the normalized distribution functions at different time stages; which indicate a larger fraction of bigger droplets in the pin-plate field as compared to that in a uniform field (Fig. 2). Further, Sauter mean diameter shows a continuous increase with time (Fig. 4(c)) and demonstrates the effectiveness of the pin-plate electrode system in water-in-oil demulsification. In addition to the electrohydrodynamic flow currents, the nonuniform field-induced mobility of droplets is responsible for the higher coalescence rate in the later part of the process. Unlike in uniform electric field, a water drop at any point in the pin-plate electrode field tries to move towards the region of stronger electric field. Dielectrophoretic mobility, thus assists the fast segregation of droplets in the high electric field region. Such a segregation helps to reduce the inter-drop separation and increases the probability of the drop–drop contact. The effectiveness of a symmetric non-uniform electric field in the water-in-oil coalescence is also investigated by using a four-pin electrode setup. The coalescence rate was observed to be higher in the gaps between two adjacent pin electrodes. However, unlike the pin-plate electrode system, electrohydrodynamic flow currents are restricted within the gaps between neighboring pin electrodes. As a result the droplets in this region grow faster while those at the central part coalesce at a slower rate. It has also been observed that at an applied potential above 200 V, the resulting bigger drops extend to the adjacent electrodes and cause short circuit immediately after the application of field.
140 120 100 80 60 40 20 0 0
10 20 Time [minute]
30
(b) Volume Mean Diameter
Sauter Mean Diameter [ µm]
34
350 300 250 200 150 100 50 0 0
10 20 Time [minute]
30
(c) Sauter Mean Diameter Fig. 9. Electrocoalescence kinetics of silicone oil-in-castor oil emulsion in the uniform (–䊏–), pin-plate (- -•- -) and quadrupole (- -- -) electric fields. Applied potential is 1 kV over 5 mm separation.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
35
Fig. 10. Spatial droplets distribution in silicone oil-in-castor oil emulsion under quadrupolar electric field. (a) Shortly after application of field. (b) Segregation at the center, long time after field application.
Fig. 11. Chain formation in silicone oil-in-castor oil emulsion under; (a) quadrupolar electric field and, (b) pin-plate electric field. Chains are encircled.
Initial Size Distribution
Initial Size Distribution
10
10
V%
N%
15
5 0 0 10 6
1
2
10 10 Uniform Field, t = 60 s
0 0 10
3
10
V%
N%
2
6
1
2
10 10 Annular Field, t = 60 s
1
2
1
2
3
10
1 10 10 Annular Field, t = 60 s
3
10
3 V%
N%
2
2
0 0 10
3
10
4 2 0 0 10
1
10 10 Uniform Field, t = 60 s
3
4
0 0 10
5
1
2
10 10 Drop Size [µm]
3
10
2 1 0 0 10
10 10 Drop Size [µm]
3
10
Fig. 12. Drop size distribution of castor oil-in-silicone oil emulsion at t = 60 s under uniform and annular electric fields. Applied potential is 400 V over 5 mm electrode separation.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
An oil-in-oil emulsion with less conducting dispersed phase needs stronger electric field and longer time to break compared to a water-in-oil emulsion due to a weaker dipolar interaction force [6]. In this work, uniform, pin-plate and quadrupole electric fields were used to coalesce the less conducting silicone oil droplets dispersed in a more conducting castor oil medium. Fig. 7 suggests nearly normal drop size distributions in uniform and quadrupolar electric fields; however, emulsion in the pin-plate field is more polydispersed. The right skewness of the distribution in pin-plate electric field indicates the presence of a large number of small size droplets with some larger drops in the electrocoalescing emulsion. The symmetric non-uniform electric field in the quadrupole electrode system favors electrocoalescence in an oil-in-oil emulsion with less conducting dispersed phase over the uniform field. However, the asymmetric electric field generated by the pin-plate electrode system is not much efficient. The drop size distribution at t = 30 min (Fig. 8) suggests similar trends at a long time after field application. The lower rate of coalescence of a silicone oil-in-castor oil emulsion in the pin-plate field compared to the uniform and symmetric non-uniform electric fields is evident from the kinetics curves in Fig. 9. Similar to the water-in-oil emulsion, electrocoalescence rate immediately after the application of electric field is higher in the case of oil-in-oil emulsion with dispersed phase less conducting. As coalescence progresses, increase in the inter-drop separation results in asymptotic slowing down in the rate. It is evident that, the coalescence is faster in the quadrupolar field followed by uniform and pin-plate electrode systems. The lower electrocoalescence in a pin-plate electrode system is attributed to the segregation of droplets near the plate electrode (a low electric field strength region) and the absence of electrohydrodynamic flow currents. The emergence of big drops suggests electrocoalescence in the high field strength region near the pin electrode. However, unlike the water-in-oil emulsion, negative dielectophoretic leaky dielectric system of silicone oil drops in castor oil does not admit strong electrohydrodynamic flow currents. The dielectrophoretic motion of the coalesced drops is unidirectional (pin to plate) and that too is not strong enough to keep the drop size distribution uniform. In the quadrupole electrode system, silicone oil droplets migrate towards the central region. Moreover, such a drop movement is assisted by the circulatory flows generated between two neighboring electrodes. However, low electric field strength at the center leads to a lower coalescence in the beginning. This results in a toroidal region between the center and the electrodes where the coalescence rate is observed to be high (Fig. 10(a)). As the coalescence progresses, larger drops accumulate at the center of the quadrupolar electric field as shown in Fig. 10(b). In an electrocoalescing emulsion of silicone oil-in-castor oil, droplets try to arrange themselves in chains. Such chains were observed in both uniform as well as non-uniform electric fields. However, the phenomenon is more pronounced in the non-uniform electric fields. Fig. 11(a) shows such a chain formation in the silicone oil-in-castor oil emulsion under a quadrupolar electric field. The chains are short in length and do not extend to the electrodes. In quadrupolar electric fields, the chains are curved and arranged along the isopotential lines. However, at a higher applied
3.3. Coalescence in oil-in-oil emulsion with dispersed phase more conducting A castor oil drop dispersed in the silicone oil exhibits positive dielectrophoresis owing to a positive Clasius-Mossotti factor [26]. We used uniform and annular electric fields to demonstrate the effect of non-uniformity of an applied field on the coalescence rate in a positive dielectrophoretic LD–LD fluid system. The use of pin-plate electrode system was avoided based on our earlier observations of the castor oil drop breakup and emulsification on and around the pin electrode [25]. Arithmetic Mean Diameter [ µm]
3.2. Coalescence in oil-in-oil emulsion with dispersed phase less conducting
potential, flow circulations between the two adjacent electrodes prevent chain formation and also push the coalesced drops towards the center. In a pin-plate electric field, the silicone oil droplets form small chains away from the electrodes (Fig. 11(b)); whereas, in a uniform field, chain formation was observed at much longer times after application of field.
50 40 30 20 10 0 0
50
100
150
200
250
300
Time [s]
(a) Arithmetic Mean Diameter Volume Mean Diameter [µm]
either due to the tip-streaming near an electrode or bridging of the adjacent electrodes. These types of drop breakup not only lower the coalescence rate but also introduce tiny droplets in the emulsion, which are difficult to coalescence even after application of electric field for a very longer time.
80 70 60 50 40 30 20 10
0
50
100
150
200
250
300
Time [s]
(b) Volume Mean Diameter 200
Sauter Mean Diameter [µm]
36
150
100
50
0 0
50
100
150
200
250
300
Time [s]
(c) Sauter Mean Diameter Fig. 13. Electrocoalescence kinetics of castor oil-in-silicone oil emulsion under uniform (–䊏–) and annular (––) electric fields. Applied potential V0 = 400 V over 5 mm electrode separation.
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
37
Fig. 14. Drop distribution and oil layer formation in castor oil-in-silicone oil emulsion during coalescence under annular electric field. Applied potential V0 = 400 V over 5 mm electrode separation.
The drop size distributions in an electrocoalescing castor oilin-silicone oil emulsion at time t = 60 s in the uniform and annular electric fields are shown in Fig. 12. The nature of histograms suggests that an electrocoalescing emulsion in an annular electric field is more polydispersed compared to that in a uniform field. The average of the distribution falls in the same range in charts for both types of fields. The average droplet size of an emulsion at the different times after application of field in the uniform as well as annular electric fields is plotted in Fig. 13. Similar to the water-in-oil and oil-in-oil (with dispersed phase less conducting) emulsions, in a castor oil-insilicone oil emulsion droplets coalesce at a faster rate immediately after the field is applied; but the rate of coalescence subsides with time. The gradual decrease in the coalescence rate can be attributed to the coalescence and spreading of castor oil drops on the electrode surface. The coalesced drops near an electrode cover a large part of its surface with a thick oil layer. The oil layer acts as an insulation, which shields the emulsion drops from the electric field. In a uniform field such type of oil layer is observed on the live electrode. However, similar layer is observed in the annular electric field, but over the outer i.e. ground electrode. As shown in Fig. 14, in an annular field, emulsion droplets coalesce faster near the electrodes and the coalesced big drops remain adhered to the electrode surface. A layer of oil on the ground electrode and big drops adhering to the live electrode diminish the field in the inner parts of the annular field resulting in slower coalescence after the initial faster rate.
4. Conclusions The experimental observations of coalescence behavior in uniform and non-uniform electric fields are reported for the water-in-oil and oil-in-oil emulsions. In the water-in-oil emulsion, asymmetric non-uniform field generated using the pin-plate electrodes is found to be advantageous as compared to the uniform and symmetric quadrupole non-uniform electric fields. Seemingly, the electrohydrodynamic flows generated due to the applied non-uniform field as well as the dielectrophoretic drop motion promote the coalescence rate. However, the strong electrohydrodynamic flows and drop instability at the high applied electric fields adversely affect the coalescence dynamics. The use of coated electrodes to induce an asymmetric non-uniform electric field below critical strength is suggested to acquire the optimum coalescence rate. In an oil-in-oil emulsion with less conducting dispersed phase, symmetric non-uniform field generated using quadrupolar electrodes results in the faster coalescence than the uniform and asymmetric non-uniform electric fields in pin-plate electrodes.
Whereas, in an oil-in-oil emulsion with more conducting dispersed phase, the annular electric field is more promising than the uniform field. We thus demonstrate that the electrical properties of the dispersed and continuous phases decide the efficacy of an electrode system in the electrocoalescence of an emulsion.
References [1] K. Klasson, P. Taylor, J. Walker, S. Jones, R. Cummins, S. Richardson, Modification of a centrifugal separator for in-well oil-water separation, Sep. Sci. Technol. 40 (2005) 453–462. [2] D. Sun, S. Jong, X. Duan, D. Zhou, Demulsification of water-in-oil emulsion by wetting coalescence materials in stirred- and packed-columns, Colloid Surf. A 150 (1999) 69–75. [3] J. Wu, Y. Xu, T. Dabros, H. Hamza, Effect of demulsifier properties on destabilization of water-in-oil emulsion, Energy Fuel 17 (2003) 1554–1559. [4] P. Atten, Electrocoalescence of water droplets in an insulating liquid, J. Electrostat. 30 (1993) 259–270. [5] S. Mhatre, V. Vivacqua, M. Ghadiri, A. Abdullah, M. Al-Marri, A. Hassanpour, B. Hewakandamby, B. Azzopardi K.B., Electrostatic phase separation: a review, Chem. Eng. Res. Des. 96 (2015) 177–195. [6] S. Mhatre, S. Deshmukh, R.M. Thaokar, Electrocoalescence of a drop pair, Phys. Fluid. (forthcoming). [7] T. Hano, T. Ohtake, K. Takagi, Demulsification kinetics of w/o emulsion in an AC electric field, J. Chem. Eng. Jpn. 21 (4) (1988) 345–351. [8] C. Lesaint, W.R. Glomm, L.E. Lundgaard, J. Sjoblom, Dehydration efficiency of ac electrical fields on water-in-model-oil emulsions, Colloid Surf. A 352 (2009) 63–69. [9] G. Berg, L.E. Lundgaard, N. Abi-Chebel, Electrically stressed water drops in oil, Chem. Eng. Process. 49 (2010) 1229–1240. [10] P. Suemar, E.F. Fonseca, R.C. Coutinho, F. Machado, F. Fontes, L.C. Ferreira, E.L. Lima, P.L. Melo, J. Pinto, M. Nele, Quantitative evaluation of the efficiency of water-in-crude-oil emulsion dehydration by electrocoalescence in pilot-plant and full-scale units, Ind. Eng. Chem. Res. 51 (2012) 13423–13437. [11] C.A.R. Pearce, The mechanism of the resolution of water-in-oil emulsions by electrical treatment, Br. J. Appl. Phys. 5 (1953) 136–143. [12] C. Bezemer, G.A. Croes, Motion of water droplets of an emulsion in a nonuniform field, Br. J. Appl. Phys. 6 (1955) 224. [13] H. Aryafar, H. Kavehpour, Drop coalescence through planar surfaces, Phys. Fluid 18 (2006) 072105–072106. [14] P. Atten, Electrohydrodynamics of dispersed drops of conducting liquid: from drop deformation and interaction to emulsion evolution, in: Proceedings of the International Symposium on Electrohydrodynamics, Gdansk, Poland, 2012. [15] G.I. Taylor, Disintegration of water drops in an electric field, Proc. R. Soc. Lond. Ser. A 280 (1964) 383–397. [16] T.C. Scott, R.M. Wham, An electrically driven multistage countercurrent solvent extraction device: the emulsion-phase contactor, Ind. Eng. Chem. Res. 28 (1989) 94–97. [17] H. Block, J.P. Kelly, Electro-rheology, J. Phys. D: Appl. Phys. 21 (1988) 1661–1677. [18] C.F. Zukoski, Material properties and the electrorheological response, Annu. Rev. Mater. Sci. 23 (1993) 45–78. [19] J.S. Eow, M. Ghadiri, Drop–drop coalescence in an electric field: the effects of applied electric field and electrode geometry, Colloids Surf. A: Physicochem. Eng. Asp. 219 (2003) 253–279. [20] C. Noik, J. Trapy, A. Mouret, G. Laborie, Design of a crude oil dehydration unit, in: SPE Annual Technical Conference and Exhibition, 2002. [21] H. Wang, R. Davis, Experiments on phase separation of dilute dispersions of coalescing drops, JCIS 181 (1996) 93–98.
38
S. Mhatre, R. Thaokar / Chemical Engineering and Processing 96 (2015) 28–38
[22] N.V.T. Opedal, G. Sorland, J. Sjoblom, Methods for droplet size distribution determination of water-in-oil emulsions using low-field NMR, diffusionfundamentals.org 9 (2009) 1–29. [23] V. Jaitely, T. Sakthivel, G. Magee, A. Florence, Formulation of oil in oil emulsions: potential drug reservoirs for slow release, J. Drug Deliv. Sci. Technol. 14 (2004) 113–117. [24] R.D. Knight, The general form of the quadrupole ion trap potential, IJMSIP 51 (1983) 127–131.
[25] S. Mhatre, R. Thaokar, Pin-plate electrode system for emulsification of a higher conductivity leaky dielectric liquid into a low conductivity medium, Ind. Eng. Chem. Res. 53 (2014) 13488–13496. [26] S. Mhatre, R.M. Thaokar, Drop motion, deformation, and cyclic motion in a non-uniform electric field in the viscous limit, Phys. Fluids 25 (2013) 072105.