Characterization and measurement

What the eye can and cannot tell you

Turbidity shows microbubbles, not nanobubbles. When generated, microbubbles (typically 10–50 µm) scatter enough light to make water look milky or cloudy; because they rise quickly and burst at the surface, the milkiness fades soon after Agarwal et al., 2011.

Nanobubbles are much smaller, with reported sizes from under 200 nm up to 1 µm, and have negligible buoyancy. They do not scatter enough light to make water opaque, so a nanobubble suspension looks transparent even at high concentrations Etchepare et al., 2017; Babu and Amamcharla, 2023. Clear water is therefore consistent with nanobubbles, but it does not confirm them; that takes instruments.

Nanoparticle tracking analysis: tracked particles per mL

Nanoparticle tracking analysis (NTA) is the preferred method in the cited studies because it counts as well as sizes. A laser beam illuminates particles in a liquid sample, and a microscope with a high-sensitivity camera records the light each one scatters. Software follows the Brownian motion of every particle frame by frame and, using the Stokes–Einstein equation, converts its diffusion speed into a hydrodynamic diameter: smaller particles move faster Nirmalkar et al., 2018; Azevedo et al., 2016.

NTA reports a number concentration, in tracked particles per mL, alongside the size distribution. It resolves individual particles even in polydisperse samples, where ensemble methods bias results toward larger particles. Studies of nanobubble existence and stability have used it for this reason Alheshibri et al., 2016; Ushikubo et al., 2010; Azevedo et al., 2019.

What NTA counts is scattering objects, not bubbles specifically. A gas nanobubble, a solid nanoparticle and an oil droplet of the same size can all be tracked, so an NTA result is a count of particles until other evidence shows they are gas-filled. That is why the studies above pair NTA with the tests described below Nirmalkar et al., 2018.

Dynamic light scattering: size, not concentration

Dynamic light scattering (DLS) estimates an average size from fluctuations in scattered light intensity caused by Brownian motion. It is widely used because it is easy to operate and covers sizes from below a nanometer up to the micron range Gurung et al., 2016; Li and Zhang, 2022.

Its main limitation for nanobubbles is weighting. Scattered intensity is proportional to the sixth power of diameter (d⁶), so a few large microbubbles or contaminants can mask the signal of millions of nanobubbles and push the reported mean size upward. DLS also cannot provide the number concentration of bubbles, a key parameter for evaluating a nanobubble generator Nirmalkar et al., 2018; Li and Zhang, 2022.

Telling bubbles from impurities

Researchers use stress tests to separate gas-filled nanobubbles from solid nanoparticles or oil droplets, because the counting instruments alone cannot.

• Compression. Nanobubbles are compressible: under external pressure they shrink or dissolve, whereas solid nanoparticles keep their size. The response is tracked with DLS or NTA under pressure Alheshibri and Craig, 2018; Ahmed et al., 2018.
• Freeze–thaw. Freezing destroys nanobubbles, while solid contaminants remain detectable. A large drop in particle count after thawing indicated that the original population was mostly nanobubbles Nirmalkar et al., 2018; Li et al., 2021.
• Resonant mass measurement. RMM separates particles by buoyancy. Nanobubbles have negligible mass and are positively buoyant in liquid, whereas solid particles are negatively buoyant, which distinguishes gas from solid phases Alheshibri and Craig, 2018; Azevedo et al., 2019.

Controls before calling particles bubbles

A particle count becomes a bubble count only when controls rule out the other explanations. Alongside the stress tests above, three controls address the most common ones:

• Filtered-water blank. Run the water used to make the sample, filtered and without gas treatment, through the same instrument and settings. Its count is the background that treatment has to exceed.
• Degassed control. Measure a sample processed the same way with the gas removed or omitted. Particles that appear regardless of gas are unlikely to be bubbles.
• Stability over time. Measure the same sample repeatedly after generation, and the controls the same way. A single reading cannot show whether the population is stable, growing or disappearing.

Results should then be reported as measured: particles per mL, with the instrument, settings, temperature and any dilution stated.