Review on Measurement Techniques for Drop Size Distribution in a Stirred Vessel

Transcription

1 pubs.acs.org/iecr on Measurement Techniques for Drop Size Distribution in a Stirred Vessel Mohd Izzudin Izzat Zainal Abidin, Abdul Aziz Abdul Raman,* and Mohamad Iskandr Mohamad Nor Department of Chemical Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur, Malaysia *S Supporting Information ABSTRACT: A literature review on measurements of drop size distribution in liquid liquid dispersion produced in a stirred vessel is presented in this work. The methods of measurement can be classified into in situ and external measurement. Two main groups of measurement techniques, namely, a laser system and image analysis, are reviewed. Several issues regarding the applications of the techniques and possible ways to overcome the problems are discussed. The suitability of different techniques depends on the operating conditions and properties of the drops. Laser systems provide fast in situ measurements which are useful for online monitoring and detecting process changes but unable to deliver reliable drop size and distribution values. In situ image analysis techniques could give accurate measurement of drop size, but a long time is required to analyze drops from a large number of images. However with development of automated image analysis, analysis time can be reduced. Therefore real-time monitoring and process control by image analysis techniques can be possible. 1. INTRODUCTION Liquid liquid dispersion is produced by mixing two immiscible liquids together with the aid of an agitator. An immiscible liquid liquid system refers to two or more mutually insoluble liquids which present as separate phases. 1 Those phases are classified as dispersed phase and continuous phase, where the dispersed phase which is usually smaller in volume will be dispersed in the continuous phase. 1 The dispersion process is very complex because it involves the simultaneous process of drop breakup and drop coalescence, where both phenomena have to be considered. 2 Stirred vessel, rotor stator mixer, inline mixer, static mixer, valve or jet homogenizers, and extraction columns are used in industrial processes to contact the liquid liquid system. 3 A stirred vessel is commonly used to produce liquid liquid dispersion in the chemical industry. Numerous investigations are conducted by researchers on the mixing mechanism and analysis of the flow pattern and its relation with liquid liquid dispersion in the stirred vessel. 2,4 6 Mass and heat transfer between the two immiscible fluids are involved in industrial processes. The interfacial area between the fluids will govern the amount of heat and mass transferred between its boundaries. 7 Thus, knowing the interfacial area in a dispersion is an important factor to determine the interphase reaction rate and the mass-transfer rate between the liquid phases. 8,9 Research has been done in the system to find optimum conditions in order to produce as large an interfacial area as possible to enhance the mass- and heat-transfer rates. 10 Since the drop size distributions are related to the interfacial area and transfer rate, techniques of drop size measurement have been developed on the basis of different physical principles. Having an accurate technique is important to produce reliable models of drop size and distributions, which can help in industrial scale-up. Accurate and reliable models which can help in the scale-up of reactors can save cost. 11 Existing data on drop size and distributions are limited for unstable dispersions and highly dispersed phase fractions caused by the difficulty of performing measurements under such conditions. 12 Available measurement techniques such as image analysis techniques are not suitable at high concentration because there is a tendency of drop overlapping in the images 13 and fast acting equipment are required to capture rapid changes in unstable dispersions. In this work, the suitability and limitations of laser systems and image analysis techniques on drop size measurements are focused. Understanding the mechanism and limitations of the measurement techniques is important in order to choose suitable techniques depending on the operating conditions and properties of the drops Drop Breakup and Coalescence. Drop breakup or deformation is caused by mechanical forces induced by the surrounding fluid while resisted by surface and internal viscous forces. Drop deformation occurs when the mechanical forces are bigger than the combined resisted force. 1 The main disruptive forces are turbulent pressure fluctuations, viscous stress due to velocity gradients in the surrounding continuous phase, 14 and interfacial stability, 15 while the main cohesive stress is interfacial tension and internal viscous stress. 8 The drop coalescence process involves drainage and eventual rupture of the intervening liquid film which depends on the physical properties of the fluids. 16 Factors controlling the drop breakup and coalescence processes such as viscosity, interfacial tension, dispersed phase fraction, and impeller speed will govern the rate of drop breakup and coalescence. The complex hydrodynamic behavior between the interacting phases can be described by population balance based modeling. It can be Received: May 15, 2013 Revised: October 18, 2013 Accepted: October 22, 2013 Published: October 22, American Chemical Society 16085

2 Figure 1. Classification of measurement techniques. done in two ways, which are stagewise and differential models. 17 In several processes, only drop breakup occurrences are preferred. Therefore several researchers only conducted studies involving drop breakup. For example, in suspension polymerization, only drop breakup occurrences are preferred to obtain a narrow size distribution of the polymer particles. 21 Thus, a suspending agent is used to minimize drop coalescence. The final particle size is determined by the drop size and drop size distribution during early stages of the suspension polymerization with the aid of suspending agents. 22 Liquid liquid dispersion can be characterized by parameters which can represent the entire dispersion such as the median drop size; the Sauter mean diameter, d 32 ; and the diameter of the largest drop, d max. 23 d 32 is commonly used in characterization of liquid liquid dispersion because it relates the area of the dispersed phase to its volume. Therefore, it can be related to mass transfer and the chemical reaction rate METHODS OF MEASUREMENT Different techniques are available to determine the drop size and distributions. The measurement techniques for drop size in a liquid liquid dispersion can be divided into in situ and external measurements. In in situ measurement, the drop size is measured directly inside the vessel where the dispersion is produced, while external measurement requires sampling of the dispersion and is analyzed outside of the vessel. Both measurement techniques can be further classified into different groups based on the principal of measurements such as laser systems and image analysis, 11 which are the most common techniques used in liquid liquid dispersion. Figure 1 shows the classifications of external and in situ measurement techniques In Situ Measurement Techniques. In situ measurement allows the measurement to be conducted at the temperature and pressure of the operating conditions. It can further be divided into direct and indirect measurements, which are laser systems and image analysis techniques. In situ measurement is important in order for online analysis to be conducted. It allows the transient drop size to be determined, and therefore the rate of drop deformation can be obtained. 25 Fast acting in situ instruments that could conduct measurements in very short time allow real-time monitoring and control of the dispersion process Laser Systems. Drop size measurement using laser systems is an indirect method, where the equipment measures the variation of some physical parameter of the dispersion instead of measuring the diameter of the drops. The physical parameter is then converted to the drop size and distributions. 25 Laser systems can further be classified into different groups that depend on their principle of measurements, which are Fraunhoffer diffraction and light backscattering Light BackScattering. Light backscattering techniques determine the chord length distribution (CLD) which is then transformed into particle size distribution (PSD). FBRM and ORM are the commonly used instruments in which this measurement principle is applied (a). Focus Beam Reflectance Measurement. The FBRM consists of a probe which can be installed easily, an electronic measurement unit, and a computer for data analysis. 27 Near-infrared light is transmitted through fiber optics to the probe tip where an optical lens which rotates at high velocity focuses the laser beam near the sapphire window. The focused beam scans a circular path at the interface between the probe window and the particle system where it will be reflected when it scans across the particle s surface. The reflectance time is then measured by the probe. Thus, the chord length of the scanned particles is determined as the product of reflectance time and the velocity of the laser scan (2 16 m/ s). 26 In other words, chord length is a straight line connecting two points on the edge of a particle. The measurements of CLD is adequate for monitoring process dynamic changes related to particle count, distribution, shape, concentration, and 16086

3 Figure 2. Illustrations of FBRM probe (right) and ORM probe (left). rheological behavior. 27 In this technique, no assumption on particle shape is made for the measurement. Therefore, it could avoid complex mathematical assumptions which could add significant errors. The FBRM conducts fast measurements where thousands of chord lengths are obtained every second. Thus, a robust CLD can be determined to illustrate changes in particle dimensions, population, and shape in time, 28 making it suitable for online measurement. In a study of hydrate formation from water-in-oil (W/O) emulsion, FBRM can successfully identify system changes, through detection, for example, of hydrate and ice nucleation, and could distinguish the extent of ice and hydrate agglomeration (b). Optical Reflectance Measurement. Another technique that applied the light backscattering principle is ORM, where a laser beam is used with a focal point of 0.6 μm in diameter at a distance of less than 1.0 mm from the instrument front. The short distance between focal point and the probe helps to reduce the distance traveled by the laser beam through the liquid mixture. Thus, it can measure the chord length in high-concentration dispersions (up to 50%). 28 The beam rotates at a known velocity, and every time it intercepts a drop, light is scattered back in the beam path and transmitted to a detector where it will be transformed into an electrical signal. The chord length is then determined from the electric signal duration and the speed of the rotating beam. Therefore the chord length distributions can be built up over a period of time. Validations of ORM measurements were conducted with a known particle size ranging from 20.6 to 230 μm, and the results shows a high deviation for a mean particle size of 20.6 μm (6.65%) compared to that for 58.5 μm (0.17%). 12 This is because for a particle size less than 20 μm, the light backscattering instruments may receive backscattered light instead of reflected light from particles which tends to oversize the particles. 29 The ORM was used successfully in analyzing system changes such as a change of drop size distributions (DSD) at different power inputs. It shows almost similar proportionality to the mean Sauter diameter over the power input when compared with image analysis techniques (endoscope). 11 The limitation of the ORM is the large size of the probe (30 mm) compared to FBRM and an endoscope (Figure 2), which may interfere with the drop size inside the vessel. However, if the probe is carefully positioned to be parallel to the flow stream going out from the impeller, the effect can be minimized. Multiple mathematical unfolding techniques are required to transform the CLD measured by the light scattering techniques into PSD. Certain inaccuracies might come from the use of CLD to represent the size distributions of the dispersion. 11 This is because the CLD obtained by FBRM is complex, which not only depends on the size distribution but also on the particles optical properties and shapes. 27 Therefore light scattering techniques are very sensitive to the surface structure of the particles and not suitable for measurements of particles with a smooth surface exterior. This is because the reflection of the laser beam from the smooth surface is not diffuse over the whole surface area but will be punctuated which may undersized particles. However, the result can be improved by producing synthetic roughness on the drop surface by introducing particles such as titanium dioxide (TiO 2 ). Adding synthetic roughness was proved to reduce the deviation between measurements conducted by FBRM and a reliable image analysis technique. 11 One of the limitations of a light scattering technique is the inability to determine the actual shape of the drops where it assumes that the drops are spherical Laser Diffraction. Laser diffraction technique is one of the recent techniques used by researchers 18,19,30 32 for drop size measurements in a stirred vessel. It can be used in situ or externally to measure the drop size and distribution. Laser diffraction technique is based on the measurement and interpretation of angular distribution of light diffracted by the droplets using Fraunhoffer diffraction theory. 25 Particle sizes are obtained by measuring the intensity of light scattered as a laser beam passes through the particles. A Malvern Mastersizer is a laser diffraction instrument which is commonly used by researchers. The analyzer is capable of measuring droplets in the range of μm with an accuracy of ±1% on volume median diameter. Calibration is not needed, but it requires the information on refractive index of the dispersed phase used in creating the dispersion. 19 It has a short measuring time, thus permitting online analysis for measurement of transient drop size distributions with minimal possible instrumental, sampling, and dispersion errors. Fast measurements also allow this technique to detect changes in drop size when there is a change in the dispersion process, such as varying impeller speed. 9 Online monitoring of drop size distributions in an agitated vessel using a Malvern Mastersizer 16087

4 Figure 3. Illustrations of endoscope probe (left) and PVM probe (right). has been conducted by Chatzi et al. 25 and Sis et al. 33 Sis et al. 33 applied the instrument for in situ measurement of dodecane in water (see Table 1) emulsion in turbulent flow. It is achieved by placing the flow-through cell in the path of the light beam of the light scattering device. The limitation of this method is that low concentrations of dispersion (<3%) are required, which means if offline measurement is applied, the sample has to be diluted first before analysis. Thus, it is unsuitable for a concentrated dispersion. 27 This technique also was used by several researchers 9 for external measurements. It uses a small volume recirculating cell sample for sample analysis. El-Hamouz and co-workers 18,19 applied this instrument to determine the drop size and drop size distribution of silicone oil in water (see Table 1) at different impeller speed, impeller type, and sample injection position. Another laser diffraction instrument is a laser granulometer, 30 which can be used to measure drop size in liquid liquid dispersion at high phase ratio (up to 30%). This equipment is based on light intensity pattern diffracted by drops at different angles in forward direction of their diameter. Thus, the drops have to be larger than the laser wavelength, and its size is calculated in terms of the volume percent distribution based on a Fraunhoffer approximation In Situ Image Analysis. Image analysis is a direct method where images of the drops are captured and the sizes are determine directly by analyzing the images. Thus, it can avoid errors which may occur during translation of other physical parameters into drop size, such as translating CLD into drop size in the light backscattering technique. In this technique, photographs of the drops are captured, and then they are analyzed by manual process or with the aid of image recognition software tools. It can be done by invasive method by using a probe or by a noninvasive method. A visual probe is commonly, used and it is trusted as a reliable technique wherein it allows direct visual observation of the dispersion process. 34 The main instruments in this technique are camera, microscope, and illumination setup. A different type of camera can be used such as a charge-coupled device (CCD) camera, a high-speed camera, or a video camera. Choosing a camera with a suitable framing rate is important to make sure that it could produce data within a process-relevant time frame, especially when characterizing changes during the early stage of dispersion where the rate of change of the drop size is high. A camera with a high framing rate could obtain enough images of drops to compute the drop size distribution in a very short time. 35 It also helps to acquire clear images of drops in motion. In order for the camera to be used for real-time monitoring, it should be able to capture a minimum of 250 drops within some seconds. A CCD camera is commonly used because it gives the optimum effort/cost ratio in most applications. 36 It could achieve a maximum speed of 50 fps which is equal to a recording time or also called data acquisition time (DAT) of 1/ 50 s/image. 37 When higher speed is required, a high-speed camera can be used in which a frame rate of 1000 fps could be achieved. 38 Since this technique works with image recognition to determine the size of the drops, having high image quality could contribute to higher accuracy. The quality of the images can also depend on the optical properties of the drops, which is refractive index. Drops with high refractive index will appear clearly in images, making the image analysis process easier and lowering the error. Using a low dispersed phase also will help in image analysis process by reducing the overlapping of drops. 37 From Table 1, one can see that most image analysis techniques are applied for a low to normal (up to 20%) dispersed phase fraction. Therefore errors during drops detection can be reduced especially if manual drops counting and measurement are conducted. However, some image analysis software could conduct accurate drop detection at high dispersed phase, thus allowing image analysis techniques to be applied for that condition Endoscope. In situ image analysis is conducted inside the vessel by inserting an endoscope connected to a camera as a microscope lens into the dispersion. A camera is placed at the upper end of the endoscope to capture images of drops in the vicinity of the glass window at the bottom of the endoscope. 39 Thus, drop size measurement could be done at a specific area of interest such as near the impeller zone where the rate of drop breakage is high. This technique allows realtime recording of a 2D image of particles, which allows the transient mean drop size and distribution to be determined. It is able to measure drop size in the range of μm even at high dispersed phase (45%). 2 Sharp images can be captured especially at the impeller area by integrating a light supply into the endoscope where the strobe flash is guided by a fiber optic cable which surrounds the endoscope. To avoid disturbances in front of the focal point by drops, a covering tube with a window can be placed at the tip of the endoscope lens. 11 This technique also allows single drop sizes to be analyzed via image analysis

5 Reliable and accurate values of drop sizes can be expected from this techniques, and the results provided by an endoscope are often used as a standard when comparing different probes Particle Video Microscope. PVM is also an invasive method of measurements, but the probe is different from an endoscope (see Figure 3). This technique has been applied to characterize particulate systems such as nucleation and dissolution behavior of crystallizing suspension, particle measurement in emulsion polymerization, and bubble size distributions in a gas liquid contactor. 20 Illumination is provided at the PVM probe tip by light produced from six near-ir lasers which are focused using a hexagonal array of lenses. Drops passing through the illuminated region will scatter the light, and the backscattered light will be collected by a lens system inside the probe. It is then relayed to a CCD array. 28 Digital images in the illuminated field area of 826 μm 619 μm will be recorded by the probe. 40 The focusing effect of its six lasers was shown to enhance the gray scale structure in the images which can enhance a fully automatic image recognition system. 28 The PVM probe favors the measurement of larger particles, where it provides images which can clearly characterize drops larger than 20 μm, but it is hard to distinguish individual drops below that scale. The repeatability in analyzing the images provided by PVM between analyzers is 5.1%. 40 It is able to produce consistent results where the difference between the corresponding Sauter mean diameter measurements does not exceed 4% in any case Stereomicroscope. Image analysis techniques can be applied for drop size measurement from outside of the stirred vessel by using a camera with a stereomicroscope, as shown in Figure 4. The camera can be synchronized with a flashing Figure 4. Noninvasive measurement technique: (1) camera, (2) stereomicroscope, (3) measurement zone, (4) stirred tank, and (5) light probe. stroboscopic light source to avoid capturing blurred images. Clear and sharp images can be obtained by using stroboscopic imaging or shutter speeds less than s, 36,42 and the shutter speed can be adjusted depending on the impeller speed. The stereomicroscope (Olympus SZ 1145ESD) allows images to be captured up to 20 mm inside the tank starting from the tank inner wall. The camera is attached to the stereomicroscope by a coupling device. 43 Better light illumination can be achieved by attaching mirrors behind the vessel wall near the position of photographing. This could help in photographing dense dispersions. 8 This technique only requires a very narrow stroboscopic light probe (less than 0.4 mm) to be inserted inside the vessel compared to other techniques that require a large probe to be in contact with the dispersion. Thus, distortion of the flow pattern will be minimized. However, there is a possibility that errors may exist due to curvature on the vessel wall which may affects the measurements. 36 The measurements however are limited to the drops located near the vessel wall only. Images captured by CCD (up to 30 fps) camera in this technique are clear for processes stirred below 0.2 kw/m 3, but as the power input increases, the images captured are poorly defined and blurred and have a low contrast as a result of higher velocity inside the vessel. The problem can be solved by replacing the camera with a high-speed video system (up to 5130 fps) which could acquire sequence of images at a power input up to 0.5 kw/m The maximum error in drop size measurement was estimated to be around 5%. 8 Image analysis techniques provide accurate drop size measurements as they are commonly used to validate the results obtained by laser systems. 9 Online image analysis allows the real-time monitoring of dispersion and the drop breakup and coalescence phenomenon to be observed. The information on the actual shape of the dispersed drops can be obtained by this technique. However, it is time-consuming because it requires processing of a large number of images and drops to build the drop size distribution for every dispersion; it is suggested that a minimum of 500 droplets is required to represent meaningful size distribution. 41 Therefore, there is a need for fully automatic image analysis to reduce the analysis time Automated Image Analysis. The main purpose of image analysis and processing is to modify the intensities and the structure of captured images so that their quality can be enhanced. Therefore, automated image analysis could contribute to accurate calculation of drop size distribution from the particles seen in the images. Its two main procedures are image processing and image analysis. In image processing, certain features of the images are enhanced by modifying its intensity distribution. It involves several processes such as image segmentation, contrast enhancement, and binary conversion. In image analysis, arc and circle were detected for radius measurement. 39 Earlier the images were analyzed by visual or manual techniques, which are high in cost, required intensive labor, and are prone to high error rates. Thus, developing a fully automatic image analysis could reduce the analysis time, making this technique suitable for online monitoring and control of a dispersion process. Different approaches can be taken to analyze the size of drops from the captured image automatically. However, in most approaches, Hough transformation is usually applied in which a simple technique is used to detect circles. Alban et al. 39 developed a drop measurement program which involves arc and circle center detection and template matching. A minimum acquisition time of 20 s was achieved for analysis of one image which contains around 100 drops. Another program was developed by Braś et al., 45 involving an edge detection process by evaluating the relation between the gradient and thickness of the drops. Hough transformation is then applied to identify the contour of the drops in the image for different values of the radius. Another approach by Khalil et al. 34 involves detection of drops circular patterns by use of Hough transformation and conversion into circles. The radii of the circles are calculated by computing the signature curve for each circle identified in the 16089

6 image and then matching it with a standard signature curve of a circle. The automatic image analysis system developed by Khalil et al. 34 has a mean error of 10.3% based on image resolution of 2.0 μm/pixel. The accuracy was considered to be satisfactory but the analysis process for 300 frames required about 40 min, which is too large to allow real-time image analysis and DSD calculation during an experiment. Therefore Maaß et al. 37 employed an automatic image analysis system which can reduce the time required for drops measurement analysis process so that real-time monitoring can be achieved. Since the success of automated analysis is greatly influenced by the quality of images, their image processing procedures involve contrast enhancement steps and the images are prefiltered to remove irrelevant and misleading image information. Their image analysis procedures consist of pattern recognition by correlation of prefiltered images with search patterns, preselection of plausible circle coordinates, and classification of the plausible circle by examining corresponding edges individually. The error of the program is below 1% with a detection rate of 250 particles/min. They reported that the measurement rate can be increased by parallelization on different cores in multicore PC. The automated image analysis program developed by researchers reduced the time required to analyze a large number of drops from the sample images. It also avoids errors in manual particle counting due to human bias during analysis, where, for a clear image, the deviation of measured Sauter mean diameter between analyzer is around ±5%. Thus, with the right tools, image analysis techniques combined with an automated analysis program could be used for online process observation control which provides real-time information, and therefore process control becomes possible Comparisons of Drop Size Measurement between Laser Systems and Image Analysis Techniques. Comparisons between techniques of measurements were performed by researchers to determine the suitable technique for their system and their area of interest to be studied. Different techniques have their own limitations and are suitable for use in different conditions. Thus, selecting the suitable techniques, which usually depends on the operating conditions and optical properties, is important to get reliable drop size and distributions. A comparison between in situ measurement using light scattering technique (3D ORM) and video recording technique (endoscope) have been done by Lovick et al. 12 for kerosene in water dispersion (ρ = 828 kg m 3, μ = 5.5 mpa s, σ = 44.7 mn m 1 ). Results of measurements between the two techniques are in reasonably good agreement (average standard deviation between d 32 is 2.5 μm). They conclude that both techniques have their own advantages, where ORM can be used for high dispersed phase (up to ϕ = 0.6) and can provide fast, in situ measurements. Therefore, it is suitable for studying unstable dispersion. Laser diffraction technique (Malvern Mastersizer) also was reported to give results very similar to those of image analysis technique for spherical particles. The mean diameter of glass beads measured by laser diffraction and image analysis are 35.2 and 35.9 μm, respectively. Thus, laser diffraction can be used to determine the size of drops in dispersion. 27 Drop size measured by FBRM does not agree well with image analysis techniques. The droplet size measured by PVM was vastly undersized by FBRM. 26,40 On the basis of comparisons between light scattering techniques (FBRM and ORM) and image analysis technique (endoscope) conducted by Maaß et al., 11 they reported that the main advantage of a laser system compared to image analysis techniques is a high number of particle counts per time unit for the laser system. It could analyze several thousand drops in a short time compared to only a few hundreds of drops for image analysis techniques. However, light backscattering technique exhibits significant errors occurring in the measurement which are caused by the drops smooth exterior surface. As a result, a large divergence was obtained when light scattering techniques were used to measure drop sizes wherein the techniques tend to undersize the drops. 27,46 Thus, although light scattering techniques are fast, which makes it successful in analyzing system changes related to the particle size in a liquid liquid dispersion, it could not be used to determine the exact size of the drops External Measurements. Although in situ techniques are available, a lot of researchers still analyzed their sample outside the system. External sampling is one of the earliest methods to measure drop size where the drops can be measured continuously within the sampling tube. 35 External measurement requires samples of the dispersion to be extracted out from the vessel and transferred to the equipment. Samples extracted from the vessel can be measured and analyzed directly under a microscope connected to a camera or by laser diffraction technique such as with a Malvern Mastersizer Sampling. During sampling, coalescence might occur which causes the measured drop size to be bigger than its actual size. 1 Hence, precautionary measures are needed to prevent errors during sampling. For example, the sampling and analysis time can be determined to make sure that analyses are done in the time interval to reduce error. 9 First, the relaxation and time of the emulsion samples were determined using light scattering technique. Then the sampling and analysis time were adjusted to be less than the determined relaxation time. When drawing out samples, it is important to minimize the amount of samples extracted from the system so that the whole dispersion system is not interrupted. Tobin et al. 47 extracted only a small volume ( μl) of samples out at a time so that it does not change or disrupt the content of the vessel. However, it is also important to make sure that the amount of sample extracted is adequate to provide a sufficient number of droplets to image. Most investigators suggested that a minimum of 500 droplets is required to represent meaningful size distribution Use of Stabilizer/Surfactants. During sampling, drops in dispersion will coalesce with each other causing the final drop size and drop size distributions to be larger than the actual size. Thus, it is important to minimize or prevent the coalescence of droplets during the sampling and external measurement process. There are two methods to minimize coalescence which are by adding surfactant 18,48 50 and by using low dispersed phase fractions. 23,25,51 55 Commonly used surfactants are poly(vinyl alcohol) (PVA) and sodium dodecyl sulfate (SDS). The surface active agent works by adsorbing on the drops interface and preventing other drops from approaching by the strong repulsion forces. Thus, immediate coalescence due to the increasing strength of the liquid film entrapped between the two colliding drops is reduced. The protective colloid will also increase the contact time for drop coalescence, increasing the probability of drop separation by agitation before it combined. 56 The stabilizing properties depend on the concentration of the stabilizer. 22 However, the stabilizer is more effective at low viscosity, and the influence of 16090

7 Table 1. Details of Liquid Liquid Dispersion Systems for in Situ and External Measurement a ref dispersion system φ measurement techniques γ (mn/m) μ d (mpa s) 2 organic solvents/water + PVA endoscope kerosene/water stereomicroscope toluene/water 0.2 FBRM, ORM, endoscope kerosene/water ORM, endoscope silicone oil/water + SLES 0.01 external Malvern glass beads (20 36 μm)/water 0.1 FBRM, Malvern, external microscope N/A N/A 30 HCl/NaCl 2 /organic solvent laser granulometer cyclohexane/toluene/water + PVA/Tween Malvern dodecane/water Malvern silicone oil/water stereomicroscope silicone oil/water + SDS 0.3 PVM n-butyl chloride/water + PVA Endoscope 5 N/A 58 sunflower oil/sugar syrup 0.03 External Microscope heavy oil/water External Malvern sunflower oil/sugar syrup + SDS 0.03 External Microscope polystyrene/styrene/water + PVA/Tween External Microscope N/A kerosene/water +SDS >0.6 External Malvern N/A a φ = dispersed phase fraction; γ = interfacial tension; μ d = viscosity. Table 2. Summary of Parameters and Characteristics of Different Measurement Techniques techniques measurement range (μm) accuracy dispersed phase fraction drops optical properties in situ invasive/ noninvasive determination of particle shape probe diameter (mm) FBRM low high external surface yes invasive no 25 ORM low high external surface yes invasive no 30 laser diffraction high low N/A yes noninvasive no N/A endoscope high high refractive Index yes invasive yes 7 PVM high high refractive Index yes invasive yes 19 stereomicroscope high high refractive Index yes noninvasive yes 0.4 (light probe) external microscope high medium refractive Index no noninvasive yes N/A stabilizer concentration in lowering the surface tension decays when the viscosity of the drops increases. 16 In a study conducted by Tobin et al. 47 a few drops of SDS solution were added to their sample in order to immobilize the drops. The SDS solution is able to efficiently stabilize the extracted sample from further coalescence. The effectiveness is also confirmed from visual observations by O Rourke and MacLoughlin. 41 According to El-Hamouz et al., 18 their sample which was stabilized by sodium laureth sulfate solution can be kept until 24 h without any changes in the drop size External Image Analysis: Microscope. Extracting the dispersion sample and observing it under a microscope is one of the earliest and simplest methods to determine the size of the drops and to determine the drop size distributions. a major advantage of using a microscope to determine the drop size is that it is a direct method with straightforward calibration. The limitations of this method are well-understood, being caused mainly by the nature of light wave and by optical aberrations. 57 External image analysis using a microscope is able to measure drops in the range of μm but varies according to the lens used in the microscope. 41 Usually, a microscope is used together with image analysis software which allows enhancement of the image quality, thus improving the precision of the drop size measurement. During analysis, the microscope can be adjusted to obtain clear and sharp images of the drop. Table 1 shows the details of liquid liquid dispersion systems for various in situ and external measurement techniques. From Table 1, one can see that laser diffraction and image analysis techniques are successfully applied at high-viscosity dispersed phase. 19,38,41,61,62 This is because drops with higher viscosity have higher refractive index value, since the viscosity of the dispersed phase is linearly related to the refractive index. 63 Therefore it is suitable to use laser diffraction or image analysis techniques to conduct measurement for drops with high viscosity Comparisons of In Situ and External Measurement. Although laser systems could conduct fast online measurements, several researchers still pair a laser technique with image analysis techniques to ensure the validity of their result. Comparison between external measurement using a microscope and in situ measurement using a particle vision measurement probe. which is essentially a video microscope, was conducted by O Rourke and MacLoughlin. 41 Precautionary steps were taken during sampling to minimize or eliminate the error during sampling such as using surfactants (SDS) to immobilize the dispersion sample and adding iodopentane to create a droplet phase with a density near the density of continuous phase. Hence, an almost neutral buoyant system is produced. They reported that both techniques were in agreement with each other wherein the difference in results yielded by both techniques does not exceed 10%. The external sampling method can provide more droplet images which are more than 500 droplets compared to the PVM method which only captured less than 400 measurable droplets. In the PVM technique, there is a tendency to underestimate a larger droplet size and overestimate a smaller 16091

8 drop size. 41 Thus, as the mean droplet size decreases, the probability of PVM measurement capturing and measuring the entire range of droplet size increases and consequently closer agreement between the microscope and PVM method is expected. From the work in ref 41 it was concluded that a satisfactory level of agreement was achieved by both techniques and the PVM technique could be improved by increasing the amount of measurable droplets. Therefore, although prone to errors, external image analysis can be used to produce reliable drop size and drop size distribution if proper care is taken during sampling and measurement. Comparisons between in situ laser techniques and external and image analysis technique were conducted 27 using light scattering (FBRM), light diffraction (Malvern Mastersizer), and image analysis techniques (microscope). They reported that laser diffraction technique gives results very similar to those with image analysis technique for spherical particles. The mean diameter of glass beads (20 36 μm by sieving) measured by laser diffraction and image analysis are 35.2 and 35.9 μm, respectively. However, the result provided by FBRM techniques shows very large deviation compared to other techniques. It shows that FBRM technique is not suitable to be used to determine the size of the drops. The parameters and characteristics of different in situ and external measurement techniques are shown in Table 2. From the comparisons, although external measurement techniques are prone to errors, it shows good agreement with in situ measurement techniques. Thus, with proper care to minimize errors during measurement such as adding surfactant and proper sampling techniques, external measurement techniques could also give reliable results for drop size and distribution. From Table 1, one can see that surfactants are used in most of the external measurements. However, external measurement techniques could not conduct real-time monitoring and online measurement of drop size. 3. CONCLUSION In situ or external measurement techniques have their own advantages and limitations which mostly depend on the operating conditions and optical properties of the working liquid. On the basis the review, several conclusions can be made: (1) In situ measurements which applied laser systems can provide fast measurements which are suitable for online monitoring and for unstable dispersions with a high rate of change. However, FBRM and ORM techniques are not suitable to determine the exact size of the drops. (2) In situ image analysis techniques could provide an accurate value of the drop sizes. With the development of camera and lenses, this technique could obtain a high number of images in a short time, but it requires a lot of work and time to extract data of drops from the amount of images. Since the measurement analysis time is much higher than its data analysis time, it is not suitable for online or real-time analysis. However, with the development of an automated image analysis program, the analysis time can be shortened, allowing real-time monitoring and process control to be possible. (3) Image analysis techniques should be used as secondary tools to validate the results obtained by light scattering techniques. It also can be used to calibrate the light scattering techniques. For example PVM can be used to calibrate FBRM techniques. (4) Although external measurement techniques are prone to errors, with proper precaution steps, such as adding surfactants, it could produce reliable result similar to in situ measurement techniques. (5) The measurement techniques depend on the optical properties of the drops to be measured, which are the refractive index and the smoothness of the exterior surface. ASSOCIATED CONTENT *S Supporting Information Table listing details of dispersion systems for coalescence prevention. This material is available free of charge via the Internet at AUTHOR INFORMATION Corresponding Author *Tel.: (603) Fax: (603) azizraman@um.edu.my. Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS We are grateful to the University of Malaya High Impact Research Grant (HIR-MOHE-D ) from the Ministry of Education Malaysia which financially supported this work. NOMENCLATURE CLD chord length distribution d 32 Sauter mean diameter (m) d max Maximum stable diameter (m) DAT data acquisition time FBRM focused beam reflectance measurement fps frames per second N/A not available ORM optical reflectance measurement PSD particle size distribution PVA poly(vinyl alcohol) PVM particle video microscope SDS sodium dodecyl sulfate SLES sodium laureth sulfate Greek Symbols μ dynamic viscosity (mpa s) ρ density (kg m 3 ) φ dispersed phase fraction REFERENCES (1) Leng, D. E.; Calabrese, R. V. Immiscible Liquid Liquid Systems. In Handbook of Industrial Mixing: Science and Practice; Paul, E. L., Atiemo-Obeng, V. A., Kresta, S. M., Eds.; John Wiley and Sons: Hoboken, NJ, USA, 2004; pp (2) Maaß, S.; Paul, N.; Kraume, M. Influence of The Dispersed Phase Fraction on Experimental and Predicted Drop Size Distributions in Breakage Dominated Stirred Systems. Chem. Eng. Sci. 2012, 76, (3) Azizi, F.; Al Taweel, A. M. Turbulently Flowing Liquid Liquid Dispersions. Part I: Drop Breakage and Coalescence. Chem. Eng. J. 2011, 166 (2), (4) Shinnar, R. On the Behaviour of Liquid Dispersions in Mixing Vessels. J. Fluid Mech. 1961, 10, 259. (5) Pacek, A. W.; Chamsart, S.; Nienow, A. W.; Bakker, A. The Influence of Impeller Type on Mean Drop Size and Drop Size 16092

9 Distribution in an Agitated Vessel. Chem. Eng. Sci. 1999, 54, (6) Zaccone, A.; Ga bler, A.; Maaß, S.; Marchisio, D.; Kraume, M. Drop Breakage in Liquid Liquid Stirred Dispersions: Modelling of Single Drop Breakage. Chem. Eng. Sci. 2007, 62 (22), (7) Eastwood, C. D.; Armi, L.; Lasheras, J. C. The Breakup of Immiscible Fluids in Turbulent Flows. J. Fluid Mech. 2004, 502, (8) Sechremeli, D.; Stampouli, A.; Stamatoudis, M. Comparison of Mean Drop Sizes and Drop Size Distributions in Agitated Liquid Liquid Dispersions Produced by Disk and Open Type Impellers. Chem. Eng. J. 2006, 117 (2), (9) Angle, C. W.; Dabros, T.; Hamza, H. A. Predicting Sizes of Toluene-Diluted Heavy Oil Emulsions in Turbulent Flow. Part 1 Application of Two Adsorption Kinetic Models for in Two Size Predictive Models. Chem. Eng. Sci. 2006, 61 (22), (10) Giapos, A.; Pachatouridis, C.; Stamatoudis, M. Effect of the Number of Impeller Blades on the Drop Sizes in Agitated Dispersions. Chem. Eng. Res. Des. 2005, 83 (12), (11) Maaß, S.; Wollny, S.; Voigt, A.; Kraume, M. Experimental Comparison of Measurement Techniques for Drop Size Distributions in Liquid Liquid Dispersions. Exp. Fluids 2010, 50 (2), (12) Lovick, J.; Mouza, A. A.; Paras, S. V.; Lye, G. J.; Angeli, P. Drop Size Distribution in Highly Concentrated Liquid Liquid Dispersions Using a Light Back Scattering Method. J. Chem. Technol. Biotechnol. 2005, 80 (5), (13) Simmons, M. J. H.; Azzopardi, B. J. Drop Size Distribution in Dispersed Liquid Liquid Pipe Flow. Int. J. Multiphase Flow 2001, 27, (14) Zhou, G.; Kresta, S. M. Correlation of Mean Drop Size and Minimum Drop Size with the Turbulence Energy Dissipation and the Flow in an Agitated Tank. Chem. Eng. Sci. 1997, 53 (11), (15) Liao, Y.; Lucas, D. A Literature on Mechanisms and Models for the Coalescence Process of Fluid Particles. Chem. Eng. Sci. 2010, 65 (10), (16) Jahanzad, F.; Sajjadi, S.; Brooks, B. W. Comparative Study of Particle Size in Suspension Polymerization and Corresponding Monomer Water Dispersion. Ind. Eng. Chem. Res. 2005, 44, (17) Attarakih, M. M.; Bart, H.; Faqir, N. M. Numerical Solution of the Spatially Distributed Population Balance Equation Describing the Hydrodynamics of Interacting Liquid Liquid Dispersions. Chem. Eng. Sci. 2004, 59, (18) El-Hamouz, A.; Cooke, M.; Kowalski, A.; Sharratt, P. Dispersion of Silicone Oil in Water Surfactant Solution: Effect of Impeller Speed, Oil Viscosity and Addition Point on Drop Size Distribution. Chem. Eng. Process. 2009, 48 (2), (19) El-Hamouz, A. Effect of Surfactant Concentration and Operating Temperature on the Drop Size Distribution of Silicon Oil Water Dispersion. J. Dispersion Sci. Technol. 2007, 28 (5), (20) Singh, K. K.; Mahajani, S. M.; Shenoy, K. T.; Ghosh, S. K. Representative Drop Sizes and Drop Size Distributions in A/O Dispersions in Continuous Flow Stirred Tank. Hydrometallurgy 2008, 90 (2), (21) Kichatov, B. V.; Korshunov, A. M.; Assorova, P. V. Particle Size Distribution of the Product of Suspension Polymerization. Theor. Found. Chem. Eng. 2003, 37 (3), (22) Yang, B.; Takahashi, K.; Takeishi, M. Styrene Drop Size and Size Distribution in an Aqueous Solution of Poly(vinyl alcohol). Ind. Eng. Chem. Res. 2000, 39, (23) Ruiz, M. C.; Lermanda, P.; Padilla, R. Drop Size Distribution in a Batch Mixer Under Breakage Conditions. Hydrometallurgy 2002, 63, (24) Liu, S.; Li, D. Drop Coalescence in Turbulent Dispersions. Chem. Eng. Sci. 1999, 54, (25) Chatzi, E. G.; Boutris, C. J.; Kiparissides, C. On-Line Monitoring of Drop Size Distributions in Agitated Vessels. 1. Effects of Temperature and Impeller Speed. Ind. Eng. Chem. Res. 1991, 30, (26) Greaves, D.; Boxall, J.; Mulligan, J.; Montesi, A.; Creek, J.; Dendy, S. E.; Koh, C. A. Measuring The Particle Size of a Known Distribution Using the Focused Beam Reflectance Measurement Technique. Chem. Eng. Sci. 2008, 63 (22), (27) Li, M.; Wilkinson, D.; Patchigolla, K. Comparison of Particle Size Distributions Measured Using Different Techniques. Part. Sci. Technol. 2005, 23 (3), (28) Maas, S.; Grunig, J.; Kraume, M. Measurement Techniques For Drop Size Distribution in Stirred Liquid-Liquid Systems. Chem. Process Eng. 2009, 30, (29) Li, M.; White, G.; Wilkinson, D.; Roberts, K. J. Scale up study of retreat curve impeller stirred tanks using LDA measurements and CFD simulation. Chem. Eng. J. 2005, 108 (1 2), (30) Desnoyer, C.; Masbernat, O.; Gourdon, C. Experimental Study of Drop Size Distributions at High Phase Ratio in Liquid Liquid Dispersions. Chem. Eng. Sci. 2003, 58 (7), (31) Jurado, E.; Bravo, V.; Camacho, F.; Vicaria, J. M.; Fernandez- Arteaga, A. Estimation of the distribution of droplet size, interfacial area and volume in emulsions. Colloids Surf., A 2007, 295 (1 3), (32) Lobry, E.; Theron, F.; Gourdon, C.; Le Sauze, N.; Xuereb, C.; Lasuye, T. Turbulent Liquid Liquid Dispersion in SMV Static Mixer at High Dispersed Phase Concentration. Chem. Eng. Sci. 2011, 66, (33) Sis, H.; Kelbaliyev, G.; Chander, S. Kinetics of Drop Breakage in Stirred Vessels under Turbulent Conditions. J. Dispersion Sci. Technol. 2005, 26 (5), (34) Khalil, A.; Puel, F.; Chevalier, Y.; Galvan, J.-M.; Rivoire, A.; Klein, J.-P. Study of Droplet Size Distribution During an Emulsification Process Using In Situ Video Probe Coupled With an Automatic Image Analysis. Chem. Eng. J. 2010, 165 (3), (35) Ribeiro, M. Non-Invasive System and Procedures for the Characterization of Liquid Liquid Dispersions. Chem. Eng. J. 2004, 97 (2 3), (36) Junker, B. Measurement of Bubble and Pellet Size Distributions: Past and Current Image Analysis Technology. Bioprocess Biosyst. Eng. 2006, 29 (3), (37) Maaß, S.; Rojahn, J.; Ha nsch, R.; Kraume, M. Automated drop detection using image analysis for online particle size monitoring in multiphase systems. Comput. Chem. Eng. 2012, 45, (38) Podgoŕska, W. Modelling of High Viscosity Oil Drop Breakage Process in Intermittent Turbulence. Chem. Eng. Sci. 2006, 61 (9), (39) Alban, F. B.; Sajjadi, S.; Yianneskis, M. Dynamic Tracking of Fast Liquid Liquid Dispersion Processes with a Real-Time in-situ Optical Technique. Chem. Eng. Res. Des. 2004, 82 (8), (40) Boxall, J. A.; Koh, C. A.; Sloan, E. D.; Sum, A. K.; Wu, D. T. Measurement and Calibration of Droplet Size Distributions in Waterin-Oil Emulsions by Particle Video Microscope and a Focused Beam Reflectance Method. Ind. Eng. Chem. Res. 2010, 49, (41) O Rourke, A. M.; MacLoughlin, P. F. A Comparison of Measurement Techniques Used in the Analysis of Evolving Liquid Liquid Dispersions. Chem. Eng. Process. 2005, 44 (8), (42) Andersson, R.; Andersson, B. On The Breakup of Fluid Particles in Turbulent Flows. AIChE J. 2006, 52 (6), (43) Galindo, E.; Larralde-Corona, C. P.; Brito, T.; Cordova-Aguilar, M. S.; Taboada, B.; Vega-Alvarado, L.; Corkidi, G. Development of advanced image analysis techniques for the in situ characterization of multiphase dispersions occurring in bioreactors. J. Biotechnol. 2005, 116 (3), (44) Guevara-Loṕez, E.; Sanjuan-Galindo, R.; Coŕdova-Aguilar, M. S.; Corkidi, G.; Ascanio, G.; Galindo, E. High-speed visualization of multiphase dispersions in a mixing tank. Chem. Eng. Res. Des. 2008, 86 (12), (45) Braś, L. M. R.; Gomes, E. F.; Ribeiro, M. M. M.; Guimaraẽs, M. M. L. Drop Distribution Determination in a Liquid Liquid Dispersion by Image Processing. Int. J. Chem. Eng. 2009,