Particle migration and trapping in 3D microbubble streaming flows

This short review article is originally based on the research paper titled “Size-dependent particle migration and trapping in 3D microbubble streaming flows”, authored by Massimiliano Rossi, Andreas Volk, Bhargav Rallabandi, Christian J. Kahler, Sascha Hilgenfeldt & Alvaro Marin. The paper was published in the journal, Physical Review Fluids in 2020. It demonstrates the migration and trapping of finite-sized microparticles using acoustically-driven microfluidics. The pressure controller’s precision played a crucial role in carrying out this study successfully.

Abstract

This study explores the use of acoustically actuated sessile semi-cylindrical microbubbles. It is demonstrated with the help of an experiment that finite-sized microparticles undergo size-sensitive migration and trapping. This study was conducted using pressure-driven flow control in microfluidics in a PDMS-based microfluidic channel. The principle uses the flow generated by acoustic-driven oscillations of a microbubble located in the channel to trap the particles. This occurs in specific spatial positions in 3D with high reproducibility. The trajectories of these particles are also recreated by passive advection of the particles in a steady 3D streaming flow field. For particles of different sizes, volume exclusion leads to three regimes of migration. This behaviour suggests applications of this study in separation, trapping and sorting of particles in the 3D space.

Introduction

The use of oscillatory fluid motion to produce powerful steady flows goes back to the 19th century. [1] Recent developments in the study of microparticles, micro-organisms and other microscopic objects [2] has revitalized research on acoustic rectification, streaming and trapping. The high frequency oscillations and non-linear properties of fluid dynamics are used to create streaming flow and particle migration.

A well-known procedure was developed to create massive rectification involves driving a microbubble’s interface at ultrasound frequencies. The resulting net flow, known as steady streaming [3], or is also known as microbubble streaming. Although the particles instead, they migrate onto specific positions and/or undergo entrapment in certain areas of the bubble’s vicinity.

Observations of such entrapment were made by Miller et al. [5]. They explored how human platelets formed aggregates around gas-filled micropores when the liquid was exposed to ultrasound. Their study made it possible for methods to be created for trapping, suspending and manipulating particles or cells. In recent times, with advancements in microfabrication technology & microfluidic experiments, Lutz et al. [7] and Lieu et al. [8] generated microvortices  or microeddies. These microvortices were formed around solid microcylinders by performing low-frequency liquid oscillations, and thus showing the procedure to trap large particles.

But, the physical mechanism behind the particle entrapment phenomena is mostly unclear. The controlled experiments in the above mentioned research were exclusively designed to confine particle motion to a plane for easier analysis. Marin et al. [4] recently revealed the strong 3D character of these streaming flows primarily due to confinement in the 3rd dimension.

Aim & objectives

The aims of the current work are twofold:

1. To examine and describe the size-sensitive migration & trapping of particles in such a 3D bubble streaming flow.
2. The researchers also intend to understand how to employ 3D structure to separate particles along qualitatively different paths.

Materials & methods

The experiments were performed in a PDMS (polydimethylsiloxane) microchannel, which had a rectangular cross-section of 500 μm x 72 μm and a blind side channel with width w =90 μm and depth l = 350 μm. Soft lithography was used for the microchannel fabrication and bonded to a 1mm thick film of PDMS. This is to ensure that all the channel walls are made of PDMS. The bonding was finalized by functionalizing the surfaces with a plasma treater (Elveflow, France). A schematic diagram of the experimental setup used for this study is demonstrated in Fig 1.

Of crucial importance is the microfabrication of the dead-end channel where the cylindrical bubble is located. This needs to be precisely controlled to yield reproducible results, which were successfully achieved using the Elveflow OB1 flow controller.

Although the fluid flow was acoustically generated by the bubble’s oscillations, the system needed to be refreshed with new liquid and particles for every new measurements and was fully automatized with a pressure-driven flow controller (Elveflow, OB1).

Key findings

Flow structure around the bubble

Tracer particles with a diameter of 2 μm were used to investigate the flow structure around the bubble. Fig 2(a) shows the measured trajectories of four tracer particles. Theoretically, each particle trajectory is confined to one of the four regions defined by the two symmetry planes of the system : z=0 & x=0. Away from the bubble, the particles move in almost planar trajectories, but as soon as the particles approach the bubble, they accelerate and jump to a different z-position, with the direction and size of the jump depending on the loop size and position. Fig 2(b) shows the repetition in the patterns, with trajectories lying on a torus-shaped surface. Tori of various sizes create a nested structure, with the small torus contained inside a large torus (Fig 2c). Extensive research and simulations have been performed on these structures, showcased in previous work [4, 5].

Trajectories of different particle sizes – Experimental & Numerical

This section of the study shows how particles are restricted to certain regions owing to their size and delves deeper into how their trajectories deviate from the fluid pathlines. Each particle starts the experiment, at rest and distributed randomly across the whole channel. After the ultrasound is turned on, and a streaming flow is generated, particles are made mobile. Fig 3 shows the z-position of different particle sizes as a function of time.

The swift migration and entrapment of the larger particles is shown for ones having a diameter of 15 μm. After migration to the eye of the toroidal flow vortex, end up in contact with the top or bottom channel wall. For a much more detailed understanding of the processes that lead to migration & entrapment, the particle trajectories are modeled and computed in the upcoming section.

For the numerical analysis of the particle trajectories, a computational study developed for 1D & 2D flow trajectories is replicated for the particles in a 3D flow. A similar type of displacement is applied whenever a particle touches a side wall or the top/bottom of the microfluidic channel.

Conclusion

In the course of this experimental study, three regimes are identified according to the type of migration and/or trapping experience and shown in the trend-line (Fig 6). It is plotted for different particle sizes and considering the initial conditions. For a more detailed understanding of each of these regimes please refer to the original research paper at Marin et al.

This study successfully shows the final position of a particle that is introduced into a streaming flow and can be well understood when taking into account the 3D flow structure. The numerical analysis performed by the researcher team demonstrates that the final position of the particles strongly depends on the diameter of the particle. Small particles behave as passive particles, while large particles not only migrate along the xy-plane but also along the z-axis.

The 3D migration of particles explored in this short review paves the way for development of microfluidic devices specifically for the particle trapping and size-sensitive sorting applications. Future work will focus on the understanding of the role that inertial forces plan in long-term migration & the role of density mismatch.

These promising results were achieved by the researchers with the help of a highly accurate pressure driven flow control equipment in Elveflow OB1.

Authors’ contribution

Alvaro Marin and Massimiliano Rossi designed the research and the experimental setup. Andreas Volk performed the experiments and build the PDMS devices. The data was analyzed by Marin, Rossi and Volk. The work on simulations and modelling correspond to Bhargav Rallabandi and Sascha Hilgenfeldt. Alvaro Marin and Andreas Volk wrote the paper.

References