Sieve-based deterministic particle displacement for suspension separation: Bridging the gap between microfluidics and microfiltration

Jaap Dijkshoorn

Research output: Thesisinternal PhD, WU


Large industries such as for food and chemicals production, and for wastewater treatment always strive to improve their processes. Separation processes are amongst the most common and most energy intensive processes and therefore subjected to continuous improvement efforts. In this thesis, the aim is to improve the industrial scale separation of dispersed particles or droplets in suspension by scaling up a microfluidic separation principle. Microfluidic (micron-sized) separation techniques are extremely effective and precise, but unfortunately, these systems only process very small volumes. To apply these systems on an industrial scale their throughput must be increased by several orders of magnitude. For this reason, we investigated methods to increase the throughput of microfluidic techniques in order to function as alternative to existing separation techniques, such as microfiltration.

In chapter 2, several microfluidic techniques were identified, compared to cross-flow microfiltration and evaluated for their potential to process larger volumes. We discussed the current state-of-the-art of microfiltration and of microfluidic techniques, and their advantages and challenges for use on industrial scale. Three promising systems were selected with potential for industrial-scale use: fluid skimming microfiltration, sparse lateral displacement arrays and inertial spiral microchannel. These three systems were evaluated on four important aspects required for industrial use. Conceptual large scale designs were proposed.

The conceptual design of the sparse lateral displacement array was selected and further investigated in chapter 3. This system was selected because it separates particles that are smaller than the gaps throughout the system, which lowers the pressure drop and the risk of (irreversible) internal fouling. The throughput was increased by replacing the traditional obstacles by sieves. Initially the introduction of sieves adversely affected the separation because of the inhomogeneous pressure difference across the sieve over its length. This pressure difference was stabilized by optimizing the outflow conditions of the outlets, which improved the performance of the system. This demonstrated that deterministic displacement of particles is possible in a sieve-based lateral displacement system as long as the flow conditions are adapted to it.

This concept was found to work well for displacing large particles (Dp = 785µm) in chapter 3. However, decreasing the critical particle size was not straightforward, because traditional deterministic lateral displacement (DLD) scaling guidelines do not apply to the asymmetric sieve-based lateral displacement system. Therefore, in chapter 4 we present an analysis of the influence of the geometric DLD parameters on the hydrodynamics and particle displacement in sieve-based lateral displacement systems. This analysis led to different guidelines to scale the critical particle diameter than are valid for original DLD systems.

The analysis of the relation between geometric parameters, hydrodynamics and particle displacement showed the large influence of the hydrodynamics on displacement and thus separation. In chapter 5 we investigated the hydrodynamics in a sieve-based lateral displacement system both experimentally and numerically. This was done by visualizing the flow lanes with high speed imaging and subsequent analysis of the velocity components for different inflow velocities. The experimental observations were confirmed with two dimensional numerical simulations. Thorough analysis of both experimental and numerical results revealed the underlying fundamentals of the flow lanes and the hydrodynamic requirements to change the critical particle diameter. With this understanding on the hydrodynamic requirements, we proposed a simplified design that would allow deterministic displacement of particles at high throughputs.

A cross-flow module with a microsieve was designed and constructed to evaluate the findings in chapter 5, which was discussed in chapter 6. Computational Fluid Dynamics (CFD) simulations were done to obtain the required hydrodynamic conditions, which were confirmed by experimental visualization of the flow field. Experiments verified that at the right hydrodynamic conditions, there is significant retention of particles and/or oil droplets that are smaller than the pores in the sieve. These results demonstrated that the microfluidic DLD separation can be applied in a microfiltration-like system to displace particles that are smaller than the pores on a much larger scale, which is not possible with conventional microfiltration. The main findings and conclusions are discussed in chapter 7. This discussion is followed by a short evaluation of the feasibility of deterministic displacement on industrial scale compared to microfiltration. The chapter is concluded by an outlook for future research. As stated before, the principle of deterministic displacement of particles on industrial scales can be very interesting because of the lower pressure loss, lower risk of (irreversible) fouling, and the possibility to effectively concentrate deformable particles and droplets. However, to achieve increased throughput, scale-up of the technique is required. In this thesis, a large-scale cross-flow microsieve (CFM) module is proposed. Its relatively simple design makes its manufacturing well feasible. Even though deterministic displacement in a CFM system has benefits compared to microfiltration, the novelty of the technique also brings its questions and risks. In the outlook we discuss several aspects that still need additional research, such as: the optimum industrial scale design, the minimum particle diameter that can be displaced, how the concentration polarisation affects displacement and fouling mechanisms. Concluding, an established, off-the-shelf technology like microfiltration may seem an easy and safe option, but we believe that the benefits of deterministic displacement may outweigh its current, initial risks, leading to a new generation of separation processes.

Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wageningen University
  • Boom, Remko, Promotor
  • Schutyser, Maarten, Co-promotor
  • Wagterveld, R.M., Co-promotor, External person
Award date12 Oct 2018
Place of PublicationWageningen
Print ISBNs9789463434973
Publication statusPublished - 2018

Fingerprint Dive into the research topics of 'Sieve-based deterministic particle displacement for suspension separation: Bridging the gap between microfluidics and microfiltration'. Together they form a unique fingerprint.

Cite this