Droplet based microfluidics is a rapidly growing interdisciplinary field of research combining soft matter physics, biochemistry and microsystems engineering. Its applications range from fast analytical systems or the synthesis of advanced materials to protein crystallization and biological assays for living cells. Precise control of droplet volumes and reliable manipulation of individual droplets such as coalescence, mixing of their contents, and sorting in combination with fast analysis tools allow us to perform chemical reactions inside the droplets under defined conditions. In this paper, we will review available drop generation and manipulation techniques. The main focus of this review is not to be comprehensive and explain all techniques in great detail but to identify and shed light on similarities and underlying physical principles. Since geometry and wetting properties of the microfluidic channels are crucial factors for droplet generation, we also briefly describe typical device fabrication methods in droplet based microfluidics. Examples of applications and reaction schemes which rely on the discussed manipulation techniques are also presented, such as the fabrication of special materials and biophysical experiments.

https://iopscience.iop.org/article/10.1088/0034-4885/75/1/016601/pdf

Droplet based microfluidics

Precise, tunable emulsions and foams produced in microfluidic geometries have found wide application in biochemical analysis and materials synthesis and characterization. Superb control of the volume, uniformity, and generation rate of droplets and bubbles arises from unique features of the microscale behavior of fluid interfaces. Fluid interfaces confined within microfluidic channels behave quite differently than their counterparts in unbounded flows. Confinement inhibits capillary instabilities so that breakup occurs by largely quasi-static mechanisms. The three-dimensional flow near confined interfaces in rectangular geometries and feedback effects from resistance changes in the entire microfluidic network play important roles in regulating the interfacial deformation. Timescales for transport of surfactants and particles to interfaces compete with flow timescales at the microscale, providing further opportunity for tuning the interfacial coverage and properties of individual droplets and bubbles.

Droplets and Bubbles in Microfluidic Devices

There is a recognized and growing need for rapid and efficient cell assays, where the size of microfluidic devices lend themselves to the manipulation of cellular populations down to the single cell level. An exceptional way to analyze cells independently is to encapsulate them within aqueous droplets surrounded by an immiscible fluid, so that reagents and reaction products are contained within a controlled microenvironment. Most cell encapsulation work has focused on the development and use of passive methods, where droplets are produced continuously at high rates by pumping fluids from external pressure-driven reservoirs through defined microfluidic geometries. With limited exceptions, the number of cells encapsulated per droplet in these systems is dictated by Poisson statistics, reducing the proportion of droplets that contain the desired number of cells and thus the effective rate at which single cells can be encapsulated. Nevertheless, a number of recently developed actively-controlled droplet production methods present an alternative route to the production of droplets at similar rates and with the potential to improve the efficiency of single-cell encapsulation. In this critical review, we examine both passive and active methods for droplet production and explore how these can be used to deterministically and non-deterministically encapsulate cells.

The Poisson distribution and beyond: methods for microfluidic droplet production and single cell encapsulation

Two-phase flow microfluidics is emerging as a popular technology for a wide range of applications involving high throughput such as encapsulation, chemical synthesis and biochemical assays. Within this platform, the formation and merging of droplets inside an immiscible carrier fluid are two key procedures: (i) the emulsification step should lead to a very well controlled drop size (distribution); and (ii) the use of droplet as micro-reactors requires a reliable merging. A novel trend within this field is the use of additional active means of control besides the commonly used hydrodynamic manipulation. Electric fields are especially suitable for this, due to quantitative control over the amplitude and time dependence of the signals, and the flexibility in designing micro-electrode geometries. With this, the formation and merging of droplets can be achieved on-demand and with high precision. In this review on two-phase flow microfluidics, particular emphasis is given on these aspects. Also recent innovations in microfabrication technologies used for this purpose will be discussed.

https://www.mdpi.com/1422-0067/12/4/2572/pdf

Droplets Formation and Merging in Two-Phase Flow Microfluidics

Industrial lab-on-a-chip: Design, applications and scale-up for drug discovery and delivery

Vortex formation by a vibrating cantilever

We describe the use of two passive components to achieve controllable and alternating droplet generation in a microfluidic device The approach overcomes the problems associated with irregularities in channel dimensions and fluid flow rates, and allows precise pairing of alternating droplets in a high-throughput manner We study droplet generation and self-synchronization in a quantitative fashion by using high-speed image analysis

Passive self-synchronized two-droplet generation.

This paper addresses a comparison of numerical stall simulations with experimental data at 60% (subsonic) and 95% (supersonic) of the design speed in a modern transonic fan rig. The unsteady static pressures were obtained with high frequency Kulite transducers mounted on the casing upstream and downstream of the fan. The casing pressure variation was clearly visible in the measurements when a stall cell passed below the transducers. Numerical stall simulations were conducted using an implicit, time-accurate 3D compressible RANS solver. The comparisons between the experiment and simulation mainly cover performance curves and time-domain pressure traces of Kulites during rotating stall. At two different fan speeds, the stall characteristics such as the number and rotating speed of the stall cells were well-matched to the experimental values. The mass flow rate and the loading parameter under the fully-developed rotating stall also showed good agreement with the experiment. In both numerical and experimental results, a large stall cell was eventually formed after stall inception regardless of the fan speed. Based on the validation, the detailed flow has been evaluated to understand rotating stall in a transonic fan. In addition, it was found that the mass flow measurement using casing static pressure might be wrong during transient flow if the Kulites were mounted too close to the fan blade.Copyright © 2011 by Rolls-Royce plc

Validation of Numerical Simulation for Rotating Stall in a Transonic Fan

An implicit, time-accurate 3D compressible RANS solver is used to simulate rotating stall inception and recovery, the so-called rotating stall hysteresis, in a modern fan geometry. In the first instance, rotating stall was simulated for 70%, 80% and 90% fan speeds using a whole-annulus fan model with a variable-area nozzle downstream. As the fan speed is increased, the stall cell also increases in size but the number of stall cells decreases. One large stall cell is predicted to rotate along the annulus at 80% and 90% speeds, while there are three smaller cells at 70% speed. In all cases, the reverse flow is confined to the near-tip region and the rotating stall does not develop into a full-span stall because of the fan blade’s high-aspect ratio. To simulate stall recovery, the nozzle area was increased gradually at 70% and 90% speeds and the flow was seen to recover from rotating stall to reach an unstalled operating condition. The recovery process was found to be affected by the fan speed. At 70% speed, the large disturbances decay first to form almost symmetric stall cells. Thereafter, the stall cells shrink into smaller ones as the mass flow rate decreases further. At 90% fan speed, a single stall cell rotates along the annulus, the disappearance of which results in recovery. An attempt has been made to explain the dependence of the stall inception and recovery patterns on the fan speed.Copyright © 2010 by Rolls-Royce plc

Minsuk Choi

Papers

Vortex formation by a vibrating cantilever

Passive self-synchronized two-droplet generation.

Validation of Numerical Simulation for Rotating Stall in a Transonic Fan

Effects of Fan Speed on Rotating Stall Inception and Recovery

Effects of recessed blade tips on stall margin in a transonic axial compressor