Mitigating clogging in a microfluidic array

The present review describes how Brian Dincau, Alban Sauret from Alban Sauret’s lab and Emilie Dressaire’s team at UC Santa Barbara created controlled pulsatile flow using a pressure-driven flow controller to mitigate clogging in a microfluidic array.

          Brian Dincau

          Alban Sauret

          Emilie Dressaire

Introduction | Investigating clogging in microfluidic arrays

Previous studies qualitatively indicated that adding pulsatile flow to microfluidic and millifluidic systems, including microfluidics arrays, can delay clogging. In a recent study, however, Brian Dincau and colleagues described the relevant physical mechanisms at play in clogging dynamics and how to establish pulsatile flow control in a microfluidics array.

According to the authors, dissecting clogging dynamics helps understand how particle deposition in microchannels may contribute to complete or partial clogging of individual channels, resulting in significant implications in the operational life of microfluidic arrays.

The researchers experimentally investigated clogging mechanisms at the pore and system scale by combining flow rate measurements with direct visualizations. Their study highlights the critical pulsatile parameters for delaying pore-clogging or removing existing clogs in an array of parallel microchannels.

Picture 1: “After running an experiment and adding a small amount of backpressure, it’s interesting to see where the flow emerges and how the filter cake shifts. This image was acquired during that process, and then adjusted for aesthetic reasons” (courtesy of Brian Dincau).

Experiment setup | Establishing flow control in a microfluidic array

To generate precise pulsatile flow, the authors combined the Elveflow OB1 MK3+ pressure controller with the Bronkhorst Coriolis flow sensor.

The Elveflow OB1 MK3+  delivers nearly any pressure profile to a system. Here, it was used to generate a tightly controlled sinusoidal pressure at the inlet of the fluidic circuit, leading to a pulsatile flow rate in the microfluidic chip.

Picture 2: Schematics of the experimental setup

For their experiments, both outlet and inlet reservoirs were pressurised  (average driving pressure set to 150 mbar) to limit bubble formation by improving gas solubility, enhancing the system´s performance overall. This also allowed them to change the flow direction when investigating the effects of flow reversal.

They used the in-series Bronkhorst Coriolis flow sensor to measure the flow rate across devices. This way, the researchers indirectly monitor clogging by correlating reductions in flow rate with specific clogging events.

The microfluidic chip used in the experiments contained 40 microchannels – each with 20 constrictions, where the width of the channel is reduced from 50 µm to 10 µm,- and a channel depth of 13 to 15 µm. Thus, each array had 800 potential clog sites. The channels connected two reservoirs – the inlet and the outlet – for the suspension flow (Figure 1 – schematic of experiment setup).

Picture 3: Picture of the PDMS microfluidic chip used in the study (courtesy of Brian Dincau)..

The researchers recorded the flow rate and video footage simultaneously, starting from the injection of the particle suspension (described under Materials session) until all the 40 channels were clogged or 10 hours had passed – whichever came first.

They used the visual observations over time with flow rate recordings in both steady and pulsatile flows to find the best combination of amplitude and frequency to delay clogging in this particular system.

Materials

• PDMS microfluidic chip
• Suspension of pure Millipore Mili-Q water, 100 mM NaCl, and 2 µm latex beads at a volume fraction of Φ = 0.3%.
• Elveflow OB1 MK3+ pressure controller
• Bronkhorst coriolis flow sensor
• Nikon eclipse microscope
• IDS imaging USB camera

Results | Influences of pulsatile amplitude and frequency

Pulsatile amplitudes determined the clogging and unclogging mechanisms at play.

High shear conditions can erode particles and aggregates and rearrange filter cakes, preventing clogging or unclogging channels. Choosing ideal pulsatile parameters, however, is essential.

When pulsatile pressure is higher than the average pressure (in this particular system, when the amplitude pressure is set to 125% of 150 mbar, the average pressure), suspension flow is reversed, resuspending particles in filter cakes. Once forward-flow resumes, some of these particles flow through and clog adjacent microchannels. Thus, compared to other amplitude settings, this parameter accelerates clogging at the beginning of the experiment.

The pulsatile frequency determines the probability of clogging and unclogging. The authors highlight that “it is critical that the frequency of pulsation is compared to the average clogging rate for a given system” to find the ideal frequency for each array.

Conclusions | Pulsatile flows improve filter half-life

Overall, adding weak pulsations to microfluidics arrays can significantly mitigate clogging. “Pulsation can achieve nearly 100% improvement in filter half-life compared to a steady flow”, conclude the researchers.

In their system, adding an amplitude of 50%, the average pressure (150 mbar) at a frequency of 0.1 Hz nearly doubled the filter half-life.

Ideally, the authors advise avoiding flow reversal for parallel microchannels as it showed no significant improvement in filter half-life compared to steady flows in their multichannel system. However, they still caution that this might be different for other microfluidic arrays.

Preventing clogging is still the best measure for prolonging and improving the operational life of microfluidic systems and diminishing costs with pieces replacement.

“The aim to delay clogging would be to reach a larger volume process over the lifetime of the device”, researchers highlight.

Creating systems that precisely control pressure and measure flow rate is ideal for finding the best settings for clogging mitigation in a microfluidic array.

This article has been written by Thais Langer.