Simulation-Aided Optimal Microfluidic Sorting for Monodispersed Microparticles

Introduction

Particulate technology has widely applied in pharmaceutical areas for drug delivery, diagnostics, and target manipulation (Arifin, 2006). Conventional methods for the preparation of microparticles, such as solvent evaporation, phase separation, and spray drying (van der Lubben, 2001), have common drawbacks. These drawbacks include a wide particle-size distribution, difficulty in precisely controlling the particle size, and poor batch-batch reproducibility of microparticles. There are many disadvantages associated with putting non-uniform microparticles to use. For example, in the field of controlled release, the target release behavior of different microparticle batches will result in poor repeatability, and the particle size and size distribution will influence the necessary drug dosage and clearance rate by the body (Berkland, 2001). Although sieving or sorting methods can solve the non-uniformity problem in microparticles, the additional procedures, apart from the preparation of microparticles, are tedious and uneconomical. Therefore, it is important to be able to control the particle size and size distribution for preparing microparticles for use in various applications.

Recently, microfluidic methods have received considerable attention because of their widespread use in applications such as analyses, syntheses, and microreactors (Stone, 2004; Vilkner, 2004). In order to obtain uniform microparticles, microfluidic devices are ideal for droplet formation because they have the advantage of continuous, reproducible, and scalable production (Jahn, 2004; Nie, 2006; Zhang, 2007; Yang, 2009; Huang, 2011). However, satellite droplets are common byproducts, which result from the breakup of immiscible liquids, in the droplet formation process. Therefore, there are often two or more peaks in the droplet size distribution owing to the liquid/liquid breakup. In order to obtain uniform microparticles, droplet sorting is needed for isolation and purification.

The methods for sorting droplets in microfluidic chips can be active or passive (Andersson, 2003; Teh, 2008). In general, active sorting needs an additional supplement such as an electrical, an optical, a mechanical, or a magnetic control system (MacDonald, 2003; Yi, 2006; Cho, 2007; Li, 2007; Roman, 2007; Wu, 2008), which is detrimental to the miniaturization of the system required for producing uniform droplets. Passive sorting, which is achieved by either gravity or channel geometry (Huh, 2007; Tan, 2004; Tan, 2007), usually requires a great deal of effort by a trial-and-error approach to adjust the flow conditions in order to obtain a high sorting efficiency. Recently, a double T-junction hybrid was proposed to generate uniform droplets at the 1^st T-junction and to efficiently separate the desired droplets away from the byproduct droplets at the 2^nd T-junction (Yang, 2009; Huang, 2011). However, how the channel geometry affects the separation efficiency was not clear.