Abstract

The controlled production of especially monodisperse emulsions is highly soughtafterin food, cosmetics, pharmaceutical, and chemical industries and is required toproduce well-structured multiphase systems (e.g. double emulsions, microcapsules,tailor-made structures with functional properties) as well as for lab-on-the-chipanalysis. Narrow droplet size distribution is difficult to achieve and therefore majoreffort has been put into development and optimization of various dispersing machines,such as high pressure homogenizers, rotor-stator toothed disk turbines, staticand dynamic membrane systems, or opposed jet micro fluidizer in order to producemonodisperse emulsions, however, without satisfying results. The drop size and sizedistribution, interfacial properties, and structuring elements in the bulk phase ofemulsions determine quality characteristics of the final product. Monodisperseemulsions are also very useful for fundamental studies of emulsion processes andproperties, because the interpretation of experimental results is simplified. In orderto design dispersing devices and droplet based reactors, a thorough understanding ofthe dispersing process mechanism and precise investigation of drop generation dynamicsare required. In laboratory research, microchannels are often utilized to analyzethe droplet formation kinetics in coflow, flow-focusing, or T-flow channel configurations.Precise control of drop size and formation dynamics, constant laminarflow field, high energy efficiency, significant reduction of the sample material, andthe simple manufacturing of the devices are some of the great advantages of thisrather new technology.The present work focuses on single w/o droplet formation and breakup at a capillarytip in coflow devices, where the disperse phase is injected parallel to a laminarcoflowing stream. The main objective was to investigate the effect of process andmaterial parameters on the drop formation dynamics at different processing lengthscales, as we varied the capillary size (0.03 - 0.11 mm), device material (glass,PDMS, stainless steel), cell diameter (0.1 - 20 mm) and length (3 - 100 cm), and theflow rate of both phases (0.001 ≤ Qc ≤ 1900 ml/min, 0.0002 ≤ Qd ≤ 0.9 ml/min). Ahomological row of food-grade Tween-surfactants and aqueous solutions containingsurface active biopolymer (HPGG; hydroxypropylether guar gum) were used for thedisperse phase illustrating in particular the influence of interfacial tension forces andelasticity on the drop formation into sunflower oil. We visualized and analyzed thedroplet dynamics systematically with high-speed imaging systems applying framerates from 1000 to 90’000 fps. Using microchannel, streak imaging (StrIm) and microparticle image velocimetry (μPIV) were employed to map and visualize the entireflow situation in and around the forming droplet, which is influenced by theflow geometry, processing conditions, and material parameters.The dripping droplet formation mechanism can be categorized into three stages: (i)start of the droplet forming called filling stage, (ii) necking stage in which the dropletis still continuously provided with liquid from the capillary, and (iii) pinching orbreaking-off after that the droplet starts to flow along the channel following thestream of the ambient fluid. The kinetics at the capillary tip in the coflow systemstrongly depends on the stresses acting on the droplet during generation. We foundthat at high velocity difference (or at lower velocity ratio, vd/vc) between bothphases, the drag forces imposed by the ambient fluid dominate interfacial tensionforces in macro as well as in mini scaled coflow devices. The presence of surfactanthas little influence under such flow conditions but the droplets are highly uniform.At lower velocity difference (or at higher vd/vc) in the same channel scales the interfacialforces strongly affect the drop filling, breakup and satellite droplet formation.In microchannel, however, we observed that the droplet production shows most pronounceddependency on the viscous drag forces due to the smaller geometricallengths scale and lower velocity differences in general. With the addition of HPGGpolymer into the disperse phase, additional differences in the droplet dynamics areclearly detected. As a result of the elastic extra stresses developed within the viscoelasticfluid, elongated drop creation and the generation of a thread, while pinchingoff, were visualized.The drop breakup experiments, including all available solutions and process parameters,were finally characterized in terms of Weber, Reynolds, Capillary, Ohnesorge,and S-numbers (surface tension number). The resulting Ca-We-phase diagram coversa wide range of parameters (2.96·10-09 ≤ We ≤ 5.9·103, 2.56·10-05 ≤ Ca ≤ 11.5,10-4 ≤ Re ≤ 2.5·104, 1.3·10-03 ≤ Oh ≤ 3.53, and 1.7·10-4 ≤ S ≤ 3.62·1004). The graphrepresents a significant contribution to predict the flow behavior of drop breakupdynamics in geometrically similar drop dispersing devices. Additionally, it is a generalizedillustration of the main forces acting in coflow channels at three differentlength scales under approximate similar flow conditions in the vicinity of the formingdroplet at the capillary tip. A preliminary force balance based on an analyticalmodel is discussed for the approximate prediction of the primary droplet size.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call