Effect of pH and Concentrated Salt on the Droplet Size and Mass

Jan 26, 2018 - To ensure the reliability of the data, over 200 glass bead images were obtained, and the sizes were measured using the Adobe Photoshop ...
0 downloads 9 Views 3MB Size
Subscriber access provided by UNIV OF DURHAM

Article

Effect of pH and concentrated salt on the droplet size and mass transfer coefficient in a stirred liquid-liquid reactor Jeil Park, Shinbeom Lee, and Jae W Lee Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.7b04793 • Publication Date (Web): 26 Jan 2018 Downloaded from http://pubs.acs.org on February 1, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Effect of pH and concentrated salt on the droplet size and mass transfer coefficient in a stirred liquid-liquid reactor

Jeil Park†, Shinbeom Lee‡ and Jae W. Lee*,†



Department of Chemical and Biomolecular Engineering, KAIST, 291 Daehak-ro, Yuseonggu, Daejeon 34141, Republic of Korea



Hanwha Chemical R&D Institute, 76 Gajeong-ro, Yuseong-gu, Daejeon 34128, Republic of Korea

*

Corresponding author: [email protected]

1

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT This work addressed liquid-liquid mass transfer for actual industrial conditions of high pH with coexisting salt. The Sauter mean diameter was determined from the direct measurement of a chlorobenzene droplet size using a borescope system. At a high NaOH concentration, the droplet size decreased significantly due to the reduced density difference between two phases, and it further decreased when NaOH and NaCl coexisted. A Sauter mean diameter correlation with a density difference term was proposed, and it successfully predicted the experimental results. To determine the mass transfer coefficient, the solubility and the concentration gradient of chlorobenzene in the aqueous phase were measured under various conditions. As the NaOH and NaCl concentration increased, the solubility of chlorobenzene and the rate of change of chlorobenzene concentration in the aqueous phase both decreased, but the decrease in solubility was more pronounced. Thus, the mass transfer coefficient increased with the NaOH and NaCl concentration.

2

ACS Paragon Plus Environment

Page 2 of 48

Page 3 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

1. INTRODUCTION The liquid-liquid (L-L) stirred reactor has been widely used in many industrial processes, such as nitration, sulfonation, and catalytic extractive reaction, due to its high yield and selectivity.1-4 Thus, many studies focused on the mass transfer in a L-L stirred reactor. When two immiscible liquids are subjected to mechanical agitation, one phase is dispersed in the other phase in the form of droplets, and L-L mass transfer occurs at the surface of the droplet. The change in the interfacial area and the mass transfer rate, which can be represented by the specific interfacial area (a) and mass transfer coefficient (kL), respectively, are both important parameters for understanding the mass transfer behavior in a stirred L-L system. The product of these two parameters is known as a volumetric mass transfer coefficient (kLa), and it is an essential parameter for reactor design. Since the interfacial area is related to the Sauter mean diameter (d32), better design of the stirred L-L system can be achieved if the effects of various factors on the Sauter mean diameter and the mass transfer coefficient are understood.

Many studies have investigated the effect of temperature and pH on the Sauter mean diameter,5-9 and the effect of the electrolyte has also been studied.10-12 Some studies conducted experiments at nearly neutral conditions of pH 5-8,7, 8 and others at pH = 13.6 However, actual industrial processes are run at much higher pH levels with NaOH concentrations higher than 0.5 M (pH > 13). For example, 2-4wt.% of aqueous NaOH solution (approximately 0.5-1.0 M) is used for a catalyst for the Aldol condensation process,13 and 10wt.% of NaOH solution (approximately 2.7 M) is used in the dehydrochlorination process of 1,1,2-trichloroethane.14 In addition, NaCl is produced as a result of the reaction in the latter case, so NaOH and NaCl coexist in the aqueous phase. To our best knowledge, no 3

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 48

experimental study has been conducted in such conditions. In case of the mass transfer coefficient, previous studies have focused on the effect of temperature and pH.8,

15, 16

However, most experiments have been conducted at neutral or acidic conditions. Therefore, the effect of a highly basic condition on L-L mass transfer has not been investigated.

This study investigated the effect of high concentrations of NaOH and NaCl solution on the Sauter mean diameter and aimed at understanding how the coexistence of NaOH and NaCl influences the Sauter mean diameter. To measure the droplet size, some studies took droplet pictures from the outside of the reactor.17-19 In this case, the droplet images could be distorted due to the reflection and refraction of light. Sampling or encapsulation has also been frequently used,20-24 but there still remained a possibility of droplet coalescence. On the other hand, a borescope system has been proposed6, 9, 25 because it can minimize the distortion of the image and hydrodynamic effect in a reactor. We utilized the borescope system to directly measure the drop size for the calculation of the Sauter mean diameter. For each salt concentration of the aqueous phase, the temperature and impeller speed were also varied to analyze which factor was dominant in the variation of the droplet size. A new Sauter mean diameter correlation was derived for practical prediction. In addition, we analyzed the effect of temperature, pH, and salt concentration on the mass transfer coefficient. A fundamental understanding of the L-L mass transfer at high pH and salt concentration will provide a general framework for reactor design in many industrial applications.

4

ACS Paragon Plus Environment

Page 5 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

2. EXPERIMENTAL SECTION 2.1 Experimental setup Measurements of the droplet size were conducted in a stainless steel reactor with a diameter of 13 cm and a height of 19.5 cm. We used a four-blade 45°-pitched paddle impeller with a 6.5 cm diameter and a 1.3 cm width, so the ratio of the impeller diameter to the reactor diameter (D/T) was 0.5. A heating jacket was used for controlling the reactor temperature, and the difference between the set point and the actual temperature was within ±1°C throughout the experiment. To insert a borescope into the reactor, a hole was punched in the upper plate of the reactor, and a 3/16″ stainless tube was passed through the hole to support the borescope as illustrated in Figure 1(a). At the end of the stainless tube, a lens with a 5 mm diameter and 1.75 mm thickness (Edmund Optics, BOROFLOAT® Borosilicate Windows, USA) was attached to prevent back flow. A borescope with a small diameter (0.110″ diameter, 10″ length; Gradient Lens Corporation, Hawkeye Pro Superslim, USA) was used to minimize its effect on the droplet size. To take a picture of a droplet, a CCD camera (Sony, XCD-U100, Japan) and a xenon lamp (Excelitas Technologies, X-1500, USA) were used, and they were connected with a light guide. In addition, a personal computer was connected to the CCD camera and the xenon lamp to control shooting and saving of images.

Measurements of the mass transfer coefficient were conducted in a 1 L glass stirred cell with a diameter of 11 cm. An overhead stirrer unit (DAIHAN Scientific Corporation, HT50DX, Korea) was installed on the upper side of the reactor, and a stainless steel four-blade straight paddle was used for homogenization of each phase. The details of each apparatus are illustrated in Figure 1(b). The temperature inside the reactor was monitored and controlled 5

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

using the temperature indicator/controller of the heating mantle, and the difference between the actual temperature and the set point was within the range of ±0.3°C.

2.2 Calibration of the borescope system To calculate the actual size of the droplet, we used the known size of a reference material. The calibration of the borescope system was conducted using two glass bead sets with diameters of 1.0 mm (Glastechnique, Germany) and 0.25 mm (SUPELCO, USA). Prior to taking the pictures, we adjusted a focus of the borescope so that sharp images were formed when a glass bead was located just in front of the lens. We put 100 g of glass beads into 1 L of distilled water and set the impeller speed to 500 rpm to prevent the glass beads from sinking. To ensure the reliability of the data, over 200 glass bead images were obtained and the sizes were measured using the Adobe Photoshop program. After averaging these values, the actual size of 1 pixel was calculated using simple proportionality based on the actual size of the glass bead. Two sizes obtained by each glass bead experiment were averaged and used as a standard for further experiments.

2.3 Calculation of Sauter mean diameter To determine the Sauter mean diameter, a guaranteed grade chlorobenzene (99.5%, JUNSEI, Japan) was selected as an organic phase. No reaction occurred between chlorobenzene with NaOH or NaCl at our experimental conditions. For the aqueous phase, a mixture of distilled water and a guaranteed grade NaOH (95%, JUNSEI, Japan) and an ACS reagent grade NaCl (99%, Sigma-Aldrich, USA) was used. The properties of the aqueous phases and organic phase in the experiment conditions are listed in Table 1 and Table 2, 6

ACS Paragon Plus Environment

Page 6 of 48

Page 7 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

respectively. The densities of the NaOH and NaCl solutions were calculated by the equation of Laliberte and Cooper,26 and the density of chlorobenzene was estimated from ASPEN Properties V8.8. The viscosities of NaOH and NaCl solutions were obtained from Laliberte,27 and the viscosity of chlorobenzene was obtained from Poling et al.28 The interfacial tensions were estimated using the empirical correlation of Good and Elbing29 from the surface tension values of two different liquids. We obtained the surface tensions of the NaOH and NaCl solutions from Dutcher et al.,30 and the surface tension of chlorobenzene was obtained by interpolating the values of Rankin.31

The droplet size was measured at 20-60°C with a 0-10wt.% NaOH solution, 0-6wt.% NaCl solution and an impeller speed of 300-500 rpm. Furthermore, the droplet size was also measured when NaOH and NaCl coexisted in the aqueous phase. The binary salt solution contained 2wt.% NaOH and 6wt.% NaCl based on the weight of the distilled water. For all experimental conditions, the aqueous phase of 1 L and the organic phase of 0.1 L were used. The two liquid phases were added into the reactor and the temperature was set using the temperature controller. When the temperature was stabilized, agitation was started and droplet pictures were taken 10 min after stabilization. To acquire a reliable droplet size, over 200 droplet images were obtained from over 50 pictures, and the sizes represented by pixels were measured using the Adobe Photoshop program. These values were converted to actual sizes using the size of 1 pixel we obtained in section 2.2. Then, the Sauter mean diameter was calculated from the number of droplets ( ) with diameter ( ) using Equation 1,

7

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

 =

6 ∑   = ∑  

Page 8 of 48

(1)

where and are the organic phase holdup and the specific interfacial area, respectively.

2.4 Estimation of the Sauter mean diameter Since density, viscosity, and interfacial tension of two liquid phases in a stirred L-L system have a strong influence on the Sauter mean diameter, several correlations using physical properties have been made to estimate the Sauter mean diameter. Equation 2 shows the most popular correlation, which estimates the maximum stable droplet diameter (  ) in a dilute solution.32 Here,  and  are the impeller diameter and the impeller Weber number,

respectively. In some studies,32, 33 the Sauter mean diameter was substituted for the maximum stable diameter on the assumption that they have a linear proportionality.

  =  . 

(2)

Doulah34 revised Equation 2 into Equation 3 by adding an organic phase holdup term to reflect the effect of coalescence.

 =  (1 +  ). 

8

ACS Paragon Plus Environment

(3)

Page 9 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

We used various Sauter mean diameter correlations20, 24, 35-39 that have a reactor/impeller geometry or an organic phase holdup similar to those in our study. The estimated Sauter mean diameters were compared with the measured ones, and the correlations were revised by multiple regressions. 2.5 Measurement of chlorobenzene solubility in the aqueous phase To calculate the mass transfer coefficient, the chlorobenzene solubility in the aqueous phase is necessary. Chlorobenzene is always located beneath the aqueous phase due to its high density, which is an advantage for sampling. If the aqueous phase is located at the bottom of the reactor, the concentration of the organic phase in the aqueous sample may be changed because the sampling syringe should pass through the upper organic phase layer. However, a change in the organic phase concentration in the sample can be prevented when chlorobenzene is used.

Various concentrations of the NaOH and NaCl solutions were used for the aqueous phase. However, 10wt.% NaOH solution and a binary system of NaOH and NaCl could not be used because the solution density was too close to the density of chlorobenzene. In these cases, a flat interface could not be maintained even at a low stirring speed, and small droplets were generated. For the organic phase, chlorobenzene was used in the same way as in the Sauter mean diameter measurement. Chlorobenzene has been used for the organic phase in various mass transfer studies,16,

40-42

and the mass transfer of chlorobenzene at extreme basic

conditions can be compared to the work of Bandaru et al.16 because they used the other extreme of acidic conditions.

9

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

A mixture of aqueous phase (200 mL) and chlorobenzene (10 mL) was poured into a 250 mL glass vessel, and the whole mixture was intensely stirred (500 rpm) for 6 h to obtain the equilibrium concentration. The glass vessel was placed in a heating mantle to control the temperature. After stirring, 10 mL of supernatant was collected and the phase separation was conducted (5000 rpm, 10 min) using a centrifuge (Eppendorf, Centrifuge 5804, Germany). After centrifuging, the two immiscible liquid phases were completely separated. Then 3 mL of the aqueous phase was collected, and chlorobenzene in the aqueous phase was extracted (3 min, 3300 rpm) with 3 mL of hexane (96%, JUNSEI, Japan) using a vortex mixer (DAIHAN Scientific Corporation, VM-10, Korea). Afterwards, 1 mL of the hexane layer was collected and 10 µL of chloroform (99%, JUNSEI, Japan) was added as an internal standard solution. This mixture was analyzed by gas chromatography (Agilent, 7890B GC, USA). In the GC analysis, a HP-5 column (30.0 m × 0.32 mm × 0.25 µm) was used, and the FID detector was run at 280°C with helium as a carrier gas. The flow rate of the helium was 2.1 mL/min. The oven temperature was increased from 40°C to 240°C at a rate of 25°C/min. The concentration of chlorobenzene in the aqueous phase was calculated using the ratio between chloroform and chlorobenzene.

2.6 Calculation and estimation of the mass transfer coefficient Measurements for estimating the mass transfer coefficient were also conducted at the flat interface. Once the two liquid phases reached a desired temperature, 400 mL of the aqueous phase was poured into the glass stirred cell. Then, 400 mL of chlorobenzene was carefully added to the bottom of the reactor using a funnel to prevent droplet formation. When the two phases were stabilized, sampling was started. For the homogenization of each phase, an 10

ACS Paragon Plus Environment

Page 10 of 48

Page 11 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

overhead stirrer unit and a double impeller were used, and each impeller was placed at the midpoint of each phase. Two different impeller speeds (50 rpm and 100 rpm) were used, which were slow enough to maintain a flat interface and prevent droplet formation. However, the flat interface could not be maintained at the impeller speed over 100 rpm. The temperature of the reactor was maintained using a heating mantle. At a specific time interval, 3 mL of the aqueous phase was collected and the concentration of chlorobenzene in the aqueous phase was measured by GC, as described in section 2.5. Using Equation 4 and the rate of change of chlorobenzene concentration in the aqueous phase, the mass transfer coefficient (  ) was calculated.16 Here, !

! and """"" are the concentration of 

chlorobenzene in the aqueous phase and the equilibrium concentration of chlorobenzene in )

'( the aqueous phase, respectively. If we plot − ln &1 − """""" ) * against +, the volumetric mass

'(

transfer coefficient is obtained from the slope of the graph. By dividing the volumetric mass transfer coefficient over the specific interfacial area, the mass transfer coefficient was obtained. The whole experiment was repeated three times to assure the reliability of the data.

− ln ,1 −

! - =  + """"" !

(4)



For the estimation of the experimental results, a new mass transfer correlation was proposed. Equation 5 shows the correlation of Bandaru et al.,16 which contains Sherwood number, impeller Reynolds number, Schmidt number, and impeller Weber number. Since their mass transfer coefficient measurement was conducted at the flat interface condition and their experimental setup was quite similar to ours, Equation 5 was selected as the basic form

11

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 48

of the new correlation. With multiple regression, four constants  (. = 1–4) were determined.

/ℎ! =  (1! + 12 )3 (/4! )5 ( )6

(5)

3. RESULTS AND DISCUSSION 3.1 Calibration of the borescope system To calibrate the borescope system, images of glass beads used (for reference sizes) were obtained and the number of pixels of each glass bead was measured using the Adobe Photoshop program. As presented in Figs. 2(a) and 2(c), the static images of the glass beads were successfully captured despite the relatively high impeller speed. In addition, we confirmed small circles at the center of the glass beads. These circles are already known as reflections of the light supply,25 and we used the circle as a standard of the center point of each glass bead during the size measurement.

The size distribution of the glass beads is illustrated in Figures 2(b) and 2(d). Although the actual sizes of the glass beads are equal, the difference in measured sizes ranged from few to dozen pixels due to the location of the glass bead. If the glass bead was located slightly away from the focal length of the CCD camera, its image is enlarged. To minimize this error, we carefully chose glass bead pictures with clear borderlines. As the actual size of the glass bead was known, a size of 1 pixel was calculated from each glass bead set. It was highly reproducible because there was no significant difference between two values (3.644 µm/pixel for a 0.25 mm glass bead, and 3.547 µm/pixel for a 1 mm glass bead), and the average of 12

ACS Paragon Plus Environment

Page 13 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

these two was used.

3.2 Calculation of the Sauter mean diameter The actual organic droplet pictures taken in pure water are shown in Figure 3. The static images were obtained at a relatively high impeller speed, and we could also specify the organic droplet even in high NaOH concentrations (2 – 10wt.%) and NaCl concentrations (6wt.%). A total of 50-100 pictures were taken and over 200 droplet sizes were measured to obtain a reliable Sauter mean diameter.25 Figures 4 – 6 show the droplet size distributions for pure water, high NaOH concentration (10wt.%), and high NaCl concentration (6wt.%), respectively. In addition, Figure 7 shows the droplet size distribution when NaOH and NaCl coexisted in the aqueous phase. Using these droplet sizes, the Sauter mean diameter for each condition was calculated and the results are presented in Figure 8.

The Sauter mean diameter had a negative relationship to the impeller speed. This is obvious because the droplet breakup can easily occur when the impeller speed is high. Notably, the Sauter mean diameter becomes less sensitive to the NaOH and NaCl concentration at high impeller speeds. In particular, when the speed was 500 rpm, there was almost no difference between Sauter mean diameters although the NaOH and NaCl concentrations in the aqueous phase were different. It seems that the other factors are overwhelmed when the impeller speed is sufficiently high.

A decrease in the droplet size due to a high pH was reported in prior studies.5-8 In addition, 13

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

some studies claimed that the droplet size decreased at high pH conditions because the interfacial tension between the two liquid phases became lower.7, 8 Like the previous studies, we observed a drastic decrease in the droplet size when the concentration of NaOH solution was high. However, as presented in Table 1, the interfacial tension slightly increased with the rise of the NaOH concentration. This implies that the change of the interfacial tension was too small to affect droplet size. Thus, there should be another factor, which has stronger effect on the Sauter mean diameter than interfacial tension does.

One possible explanation is the effect of density change in the aqueous phase. When the NaOH concentration was low and the density difference between the two phases became large, it is difficult for the heavy phase chlorobenzene droplet to rise toward the rotating impeller. With high impeller speeds (around 500 rpm), the density difference is overcome because there is sufficient power to draw chlorobenzene droplets into the upper position. Otherwise, the breakup of droplets is hard to occur because the droplets were located at the bottom, which is away from the impeller. On the other hand, if the concentration of the NaOH solution was as high as 10wt.%, its density was almost identical to the density of chlorobenzene as presented in Table 1. Therefore, sinking of droplets occurs less frequently.

The size of a droplet became smaller when NaCl was added to the NaOH solution. However, it was reported that the droplet size increased as the concentration of the electrolyte increased.5 It was claimed that the thickness of the diffusive layer decreased as the ionic strength increased, so the droplet coalescence was more likely to occur. However, they also mentioned that the effect of the electrolyte on the droplet size was not that noticeable compared to the effect of pH. 14

ACS Paragon Plus Environment

Page 14 of 48

Page 15 of 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

It was also reported that the size of a droplet increased when the electrolyte was present in the aqueous phase.12 They used NaCl and a surfactant together as an aqueous phase and analyzed the effect of their concentration on the droplet size. When the surfactant concentration was low, the droplet size increased with the NaCl concentration. However, with high surfactant concentrations, the droplet size became slightly smaller at higher NaCl concentrations. Thus, it is hard to say that the presence of an electrolyte always leads to a decrease in droplet size. In addition, the prior study12 used a higher organic phase fraction (0.28) than the current study (0.09), so droplet coalescence might easily have occurred in their experiment. At a lower organic phase holdup, the probability of droplet coalescence may become lower, so the droplet size would not increase even at a thinner electrical diffusive layer. The droplet size decrease can be explained by the increase in the aqueous phase density, as previously mentioned.

The droplet size also decreased with an increasing temperature. As presented in Table 1, interfacial tension between the two phases decreased as the temperature increased. Since the change of the interfacial tension was more sensitive to temperature than pH, it could make the droplet size smaller. In addition, density change could also reduce the droplet size, because the density difference between the two phases was lessened as temperature increased. However, the effect of temperature was relatively small compared to the concentration of NaOH and NaCl solution.

3.3 Estimation of the Sauter mean diameter

15

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 48

Many Sauter mean diameter correlations have been reported in the literature. Among these correlations, we chose some correlations20, 24, 35-39 that have organic phase holdups and reactor geometry similar to our study. The summary of the selected correlations and the estimated results are presented in Table 3 and Figure 9, respectively. However, the measured Sauter mean diameters cannot be well predicted using the prior correlations. Since the phase density difference is a crucial parameter for the droplet size, it was implemented in Equation 6 as a correction term, while the organic phase holdup term was excluded because it always remained constant in the experiment. Equation 6 shows our proposed Sauter mean diameter correlation obtained by the multiple regressions, where 7! and Δ7 are the density of the aqueous phase and the density difference between the two phases, respectively.

 7! . = 0.105.;