How to perform nanobubble generation?

This application note will focus on the generation of bulk nanobubbles by pressure-driven flow controlled microfluidics based on the PhD Thesis entitled “Microfluidics investigation of foam stability” Leslie Labarre from the University of Birmingham.

Introduction to microfluidic nanobubble generation

In nature, there are two kinds of nanobubbles in aqueous solutions: surface and bulk nanobubbles.

• Surface nanobubbles defined as gas-filled spherical caps of 10 to 100 nm height and a contact line radius between 50 to 500 nm present on hydrophobic surfaces.
• Bulk nanobubbles described as gas-filled spherical bubbles with a diameter below 1 micron located in the bulk of an aqueous solution (Alheshibri et al., 2016).

Both kinds present an extraordinary stability: the evidence of stable nanobubbles with a diameter below 1 micron generated by shear in brine was reported for the first time by Johnson and Cooke (1981).

Their gaseous nature was demonstrated because their size was increasing under vacuum and decreasing under pressure.

Advantages

• Accurate control over gas & liquid flows
• Low consumption generation technique
• Straightforward generation technique

Applications

Bulk nanobubbles take already part in a wide range of application:

• Biomedical applications (Peyman et al., 2012).
• Froth flotation (Fan et al., 2010).
• Cleaning of surfaces (Liu et al., 2008; Chen et al., 2009; Liu and Craig, 2009; Alheshibri and Craig, 2019).

Bulk nanobubbles can be generated via a wide number of techniques as listed below:

• Hydrodynamic cavitation (Oliveira et al., 2018)
• Ultrasonication (Cho et al., 2005)
• Solvent-water exchange (Zhang and Ducker, 2007; Zhang et al., 2017)
• Electrolysis of aqueous solutions (Yang et al., 2009)
• Strong mechanical agitation (Ohgaki et al., 2010).

However, all these methods are energetically-costly and often imply the generation of contaminants. Consequently, Peyman et al. (2012, 2016) developed a microfluidic method to generate high throughput contrast agents for ultrasound imaging. Yet, no attempt was made to evaluate the parameters influencing bulk nanobubble generation by microfluidics.

Microfluidic nanobubble generation

The microfluidic device employed consists of a rectangular cross-section microchannel of 25 μm (d 1) depth in a first part of the device that further develops with a 55 μm (d2) depth expansion channel. A 50 μm wide gas inlet meets at a flow-focusing junction with two 100 μm wide liquid inlets as shown. The nanobubbles are generated and collected at the exit of the channel.

One reservoir of air is connected via a pressure controller (OB1 MK3, Elveflow) to the gas inlet via 0.020″ x 0.060″ OD tubing to accurately control the gas inlet pressure. A 10 mL plastic syringe filled with the sample is connected to the liquid inlet via the same tubing, and its flow rate is finely controlled via a syringe pump.

Three nozzle widths (10, 15, 20 μm) are chosen to study the effect of the nozzle size on the bulk nanobubble generation.

One of the main peculiarities of this design is the change in depth between the nozzle and the expansion channel. Peyman et al. (2012, 2016) explained that to produce the atomization-like spray, the depth is doubled (from d1 = 25 μm to d2 = 55 μm) so that the pressure drop is highly increased. Unlike common microfluidics application, a turbulent flow is created due to the increased pressure drop within the device as illustrated in Fig. 2.

Nanobubbles are generated by injecting simultaneously pure distilled water previously filtered and air into a 10 to 20 μm width nozzle two-depth (d1 = 25 μm, d2 = 55 μm) flow-focusing device. The liquid flow rate is set at 90 μL min-1 and the gas inlet pressure is varied from 1000 to 1800 mbar.

OB1 pressure-driven flow controller

PTFE tubing

2x 50 mL Falcon reservoirs

Fittings

Hardware:

• OB1 flow controller with at least two channels 0/2000 mbar
• Kit starter pack Luer Lock + 1/32 tubings + 1/32 sleeves + 23G needles
• 2x 50 mL Falcon reservoirs
• Two-depth flow-focusing microfluidic chip (hydrophobic channels)
• Optical microscope for observation

Chemicals:

• Distilled water
• Air

Results & discussions

Figure 1: Bulk nanobubbles size distribution time evolution by intensity (a) & number (b) in pure water (30 min, blue – 1 week, orange – 1 month, green) via the 15 μm nozzle microfluidic device. The results represent the average of two experiments. Courtesy of Leslie Labarre. 

Figure represents the result of one generation cycle via the 15 μm and its evolution after 30 minutes, 1 week and 1 month after generation. It was observed that smaller bubbles were merging into bigger bubbles. Indeed, the intensity annumber bimodal distribution evolved over time: 30 minutes after collection, the number of smaller nanobubbles was greater than the 200 nm bubbles. Then, after one week, two clear peaks in the number distribution were observed. This result suggested that the number of smaller bubbles had decreased compared to the number of 200 nm bubbles in solution. It is possible also that the coalescence occurred in solution leading to the increase in the number of 200 nm bubbles in solution. After one month, the two bubble populations had decreased slightly but persisted still. Thereby, these evolutions showed that unlike residual impurities or nanoparticles that could have sedimented or agglomerate over time, nanobubbles were stable in solution.

The effect of the size of three different nozzle diameters (10, 15 and 20 μm) on nanobubble size distribution for one cycle of generation via microfluidics, 30 minutes after collection, for a gas inlet pressure of 1000 mbar and a liquid flow rate of 90 μL min-1 is represented in Figure 2.

Figure 2: Effect of nozzle diameter on bubble size distribution by intensity for one cycle of generation with the 10, 15 and 20 μm nozzle distribution represented as blue, red and purple lines respectively. The results represent the average of two to three measurements per nozzle. Courtesy of Leslie Labarre. 

From the intensity distribution, it appeared that each nozzle provided a range of bubble size varying from 50 nm up to 400 nm. The 10 μm nozzle distribution showed a monodisperse distribution centred at 200 nm. The 15 μm nozzle distribution presented a bimodal distribution with a smaller peak below 100 nm and a second peak centred at 250 nm. Lastly, the 20 μm nozzle distribution depicted a single peak centred at 300 nm. Each nozzle and resulting bubble distribution presented the same stability up to one month as illustrated in Figure 2 summarising the results obtained for two to three repeats of the same experiment.

If you’re interested in reproducing what the author has performed, feel free to contact our team of experts! 

References