Materials of nanoscale size exhibit properties that macroscopic materials often do not have. The same holds for bubbles on the nanoscale: nanoscale gaseous domains on a solid-liquid interface have surprising properties. These include the shape, the long life time, and even superstability. Such so-called surface nanobubbles may have wide applications. This prospective article covers the basic properties of surface nanobubbles and gives several examples of potential nanobubble applications in nanomaterials and nanodevices. For example, nanobubbles can be used as templates or nanostructures in surface functionalization. The nanobubbles produced in situ in a microfluidic system can even induce an autonomous motion of the nanoparticles on which they form. Their formation also has implications for the fluid transport in narrow channels in which they form.
I. INTRODUCTION TO THE SCIENTIFIC CHALLENGE
Surface nanobubbles (see Fig. 1) are gaseous domains with nanoscale thickness on immersed substrates.1–4 They were first speculated to exist about 20 years ago,5 based on stepwise features in force curves between two hydrophobic surfaces, eventually leading to the first atomic force microscopy (AFM) image in 2000.6,7 While in the early years, it was suspected that they may be an artefact caused by interactions between the substrate and the AFM tip; meanwhile, their existence has been confirmed with various other methods such as infrared spectroscopy,8 quartz crystal microbalance,9,10 neutron reflectometry,11,12 x-ray reflectivity,13–15 and most recently, through direct optical observation with interference enhanced reflection microscopy and total internal reflection fluorescence microscopy.16,17 Their existence seemed to be paradoxical,18 as a simple classical theory estimates that a bubble with a 100 nm radius should dissolve in microseconds, due to the large Laplace pressure inside these nanoscopic objects.19 Yet, they are known to survive for days.20
AFM image of nanobubbles on HOPG; the scale of the image is 2 μm × 2 μm × 40 nm.21 The typical lateral extension of the nanobubbles is 100 nm, their typical height 10 nm, and one would expect that the bubbles dissolve due to the Laplace pressure. Yet, they survive for days. Reprinted with permission from Borkent et al., Langmuir 26, 260 (2010). Copyright 2010 American Chemical Society.
AFM image of nanobubbles on HOPG; the scale of the image is 2 μm × 2 μm × 40 nm.21 The typical lateral extension of the nanobubbles is 100 nm, their typical height 10 nm, and one would expect that the bubbles dissolve due to the Laplace pressure. Yet, they survive for days. Reprinted with permission from Borkent et al., Langmuir 26, 260 (2010). Copyright 2010 American Chemical Society.
Surface nanobubbles have attracted major attention from the scientific and even general media, as they are not only interesting from a fundamental point of view but as they also have major application potential, e.g., in providing enhanced slippage in nanochannels as first pointed out by de Gennes,22 in (photo)catalysis and heterogeneous cavitation,23,24 in surface cleaning,25 and in flotation.26 Next, surface nanobubbles are also interesting from a conceptional point of view, as the flow and the mass transfer on a nanoscale have macroscopic consequences. The question arises: How do these vastly different length- and time-scales couple—and how can they be coupled in a theoretical and/or numerical approach? How to connect molecular dynamics (MD) simulations27 with continuum mechanics approaches? Finally, we view surface nanobubbles as manifestation that our knowledge on surface science still has major gaps and as challenge to fill these gaps.
II. SEVERAL EXOTIC FEATURES OF SURFACE NANOBUBBLES
Surface nanobubbles can be either spherical caps (see Fig. 1), or quasi two-dimensional so-called micropancakes (quasi two-dimensional gaseous domains, see Fig. 2), or even the composite of nanobubbles sitting on top of micropancakes (see Fig. 3). This is in analogy to the wetting of a small amount of liquid at a solid-vapor or solid-gas interface:28 The liquid on the surfaces comes in the form of either a liquid drop or a liquid pancake directly sitting on the surface with a sharp molecular tip and a three-phase contact line, or in the form of a liquid drop sitting on a liquid pancake which itself is sitting on a solid surface, as both depicted by Brochardwyart et al. and reproduced in Fig. 3. And there is another analogy to the inverse situation of droplets on surfaces: Just as the shape of nanodroplets,28–31 also the shape of surface nanobubbles is particularly sensitive to physical and chemical heterogeneities of the substrate even down to the nanoscale. On a surface with nanoroughness, e.g., imperfectly silane-coated silicon, the surface nanobubbles can be pinned by the heterogeneities and exhibit irregular three-phase contact lines,32 so that they were described as nanoscale gas networks in the literature.33,34
AFM image of micropancakes produced by solvent exchange, the same procedure that produces nanobubbles. The light domains are micropancakes, and the dark blue area is HOPG.
AFM image of micropancakes produced by solvent exchange, the same procedure that produces nanobubbles. The light domains are micropancakes, and the dark blue area is HOPG.
Top: Sketch of a liquid droplet or pancake on a solid-vapor interface, with a three-phase contact line and a sharp molecular tip. Middle: Sketch of a liquid drop sitting on a liquid micropancake, which itself is sitting on a solid-vapor or solid-gas interface. Reprinted with permission from P. G. de Gennes, Rev. Mod. Phys. 57, 827 (1985). Copyright 1991 American Chemical Society. Bottom: AFM image of the composite of nanobubbles sitting on micropancakes, produced by solvent exchange, the same procedure that produces nanobubbles directly sitting on the surface. The dark blue area is HOPG, the bright dots are nanobubbles, and the light blue disks are micropancakes. The left panel is the height image, and the right panel is the phase image. Reprinted with permission from Zhang et al., Langmuir 23, 1778 (2007). Copyright 2007 American Chemical Society.
Top: Sketch of a liquid droplet or pancake on a solid-vapor interface, with a three-phase contact line and a sharp molecular tip. Middle: Sketch of a liquid drop sitting on a liquid micropancake, which itself is sitting on a solid-vapor or solid-gas interface. Reprinted with permission from P. G. de Gennes, Rev. Mod. Phys. 57, 827 (1985). Copyright 1991 American Chemical Society. Bottom: AFM image of the composite of nanobubbles sitting on micropancakes, produced by solvent exchange, the same procedure that produces nanobubbles directly sitting on the surface. The dark blue area is HOPG, the bright dots are nanobubbles, and the light blue disks are micropancakes. The left panel is the height image, and the right panel is the phase image. Reprinted with permission from Zhang et al., Langmuir 23, 1778 (2007). Copyright 2007 American Chemical Society.
For the surface nanobubbles, their nanoscopic contact angle has been carefully examined in order to quantify their shape. The puzzling result is that the nanoscopic contact angle (on the gas side) is always much smaller than that of a macroscopic bubble sitting on the same substrate surrounding by the same liquid phase.6,32 To explain the difference between nanoscopic and macroscopic contact angle, size effects on the contact angle were often invoked, such as line tension effects.35 However, the change of the contact angle with the bubble size (base radius) is far too large to be explained by the theoretically predicted36 and numerically calculated27 value of the line tension.35,37
Micropancakes preferentially form on atomically flat substrates such as HOPG with cleavage steps, as shown in Fig. 2. They are only a few nanometer high, but spread up to several micrometer wide. The cross-sectional profile of micropancakes is flat on the top with the curvature at the boundary. So far, micropancakes have only been observed on crystalline substrates in water including HOPG, talc, MoS2.32,38
Up to now, no chemical characterisation was done to prove that the chemical nature of micropancakes and nanobubble-on-micropancake composites is exclusively gaseous. Molecular spectra of the micropancakes will be very helpful to provide convincing evidence for their gaseous nature. It was also reported39 that nitrogen gas molecules can self-assemble on a surface as a monolayer with high regular patterns, which presumably is different from the case of micropancakes. The existence of such gas-enriched layers was proposed also based on the interaction between the tip of atomic force microscopy and the substrate.40
Another interesting feature of surface nanobubbles is their mechanical property. The stiffness of surface nanobubbles was investigated by AFM measurements (see Fig. 4), see Refs. 41 and 42. Stronger applied AFM forces lead to flatter nanobubbles. In fact, the nanobubbles behave like a harmonic spring under the load applied by the AFM tip,42 i.e., with a linear relationship between applied force and deformation. The stiffness of surface nanobubbles is size dependent: The smaller the nanobubbles are, the stiffer they are. A quantitative understanding of this size-dependance of the stiffness has not yet been achieved.
Illustration showing that the gas-water interface of nanobubbles behaves like a simple spring under the load applied by the AFM tip. The spring constant is dependent on the bubble size. Reprinted with permission from Zhao et al., Soft Matter 9, 8837 (2013). Copyright 2013 Royal Society of Chemistry.
Illustration showing that the gas-water interface of nanobubbles behaves like a simple spring under the load applied by the AFM tip. The spring constant is dependent on the bubble size. Reprinted with permission from Zhao et al., Soft Matter 9, 8837 (2013). Copyright 2013 Royal Society of Chemistry.
The most celebrated and on first sight most surprising feature of surface nanobubbles maybe their stablity, as they survive at least for hours under ambient conditions. This is in contrast to the prediction of the classic theory on the bubble stability of spherical bubbles in the bulk of (partially) degassed water.43 In addition, surface nanobubbles can also sustain very large pressure reductions down to −6 MPa,44 in contrast to the expectation from nucleation theory.45,46 This is why they were called super-stable.
Nanobubbles can also sustain high temperature rises.47 The stability of surface nanobubbles was examined by coupling AFM and temperature control.48,49 The size of nanobubbles shows a temperature dependence with the maximal bubble size around 35–40 °C, where the solubility of major atmospheric gases in water reaches the minimum.50 Optical microscopic fast imaging was used to characterize the giant surface nanobubbles with lateral diameter up to several micrometers when the substrate was highly hydrophobic. It was found that nanobubbles can survive at temperatures even up to the boiling point of water.47
Fig. 5 illustrates the boiling events when surface nanobubbles were present on the substrate, showing an artist's impression of the boiling process of a liquid on a surface decorated with surface nanobubbles. At temperatures close to the boiling point, normal vapour microbubbles have already formed, expanding across the surface. During the expansion of the vapour microbubble, the three-phase contact line collides with surface nanobubbles. These are stable enough to pin the three-phase contact line for a while. Finally, when the three-phase contact line snaps off, a microdroplet is nucleated on top of the nanobubble. These microdroplets then grow by vapor condensation and remain at exactly the original positions of the surface nanobubbles even after the bulk water has retreated from the surface. As all water evaporate in the system and the condensation ceases, the nanobubbles inside the microdroplets finally burst. Some snapshots of this process are shown in Fig. 6.
Artist's impression of the boiling process of a liquid (background) with immersed surface nanobubbles: The three-phase contact line moves backwards. Once it passes nanobubbles, these do not open towards the gaseous phase as one would expect, but survive, nucleating a microdroplet in the gaseous phase. In the long term—once the vapour saturation become less—these nanobubble-filled microdroplets pop, and the gas is released.
Artist's impression of the boiling process of a liquid (background) with immersed surface nanobubbles: The three-phase contact line moves backwards. Once it passes nanobubbles, these do not open towards the gaseous phase as one would expect, but survive, nucleating a microdroplet in the gaseous phase. In the long term—once the vapour saturation become less—these nanobubble-filled microdroplets pop, and the gas is released.
Bursting of nanobubbles at 97–100 °C. (A)–(D) show top view snapshots of nanobubbles through which the three-phase contact line passes. It first remains pinned, then a microdroplet with an entrapped nanobubble emerges, which finally bursts. (E)–(H) show schematic drawings of a finally bursting nanobubble, which occurs under low humidity conditions. Reprinted with permission from Zhang et al., Phys. Rev. Lett. 112, 144503 (2014). Copyright 2014 American Physical Society.
Bursting of nanobubbles at 97–100 °C. (A)–(D) show top view snapshots of nanobubbles through which the three-phase contact line passes. It first remains pinned, then a microdroplet with an entrapped nanobubble emerges, which finally bursts. (E)–(H) show schematic drawings of a finally bursting nanobubble, which occurs under low humidity conditions. Reprinted with permission from Zhang et al., Phys. Rev. Lett. 112, 144503 (2014). Copyright 2014 American Physical Society.
We attribute the remarkable stability of surface nanobubbles—both with respect to dissolution, to massive pressure reduction, and to boiling conditions—to the strong pinning towards the substrate. It will, therefore, be interesting to find out whether the stability of nanobubbles depends on the physical and chemical heterogeneities of the substrate, which provides the pinning forces. We predict that this will be the case. In particular, we do not expect a long life-time of surface nanobubbles on ultra-smooth substrates such as slippery liquid-infused porous surfaces, in which a micro- or nano-structured surface is infused with a lubricant.51,52
III. IMPLICATIONS OF SURFACE NANOBUBBLES
The presence of surface nanobubbles alters a range of physical and chemical properties of the solid-liquid interfaces. In some cases, they isolate the substrate from the surrounding materials or they act as physical barrier, partly blocking the way. On the one hand, surface nanobubbles should also enhance the slip at the wall, leading to a detailed balance which of these two effects will prevail, just as on the microscale, on which one can control the slippage through the morphology of microbubbles covering the surface.53 Surface nanobubbles influence the adsorption of salts, proteins, or nanoparticles.54–56 In other cases, surface nanobubbles can be applied as soft templates in producing hollow nanostructures, for example, gold nanoparticles with optical properties.57 Here, we give a few examples of the applications of nanobubble in heterogenous bubble nucleation with ultrasound,24,58,59 nanopattern formation,60 and the design of nanodevices.
A. Nanobubbles and heterogeneous cavitation by ultrasound
It has been suspected for a long time that the presence of gas at the solid-liquid interfaces should have significant effects on the heterogeneous nucleation of bubbles under ultrasonic pressure reduction. The effect of nanobubbles on heterogenous nucleation of macroscopic bubbles was elegantly demonstrated in the work of Belova et al.58,59 In order to reveal the effect of surface properties on the heterogeneous nucleation of cavitation bubbles, these authors patterned some soft substrate with hydrophobic and hydrophilic regions, immersed it in water with a controlled amount of dissolved gas, and then applied strong ultrasound to make emerging bubbles cavitate. Strongly collapsing bubbles close to the surface develop a jet directed towards the surface.61 Due to the softness of the substrate, the jet was strong enough to damage it. The resulting pits on the substrate were then simply counted, giving a measure of the cavitation activity and how it depends on the gas type and concentration. An example is shown in Fig. 7. The cavitation rate was lowest in degassed water, and did not show significant differences in nitrogen-saturated water as compared to standard water. However, the cavitation rate was significantly higher in argon-saturated water. This relationship between the gas type and the cavitation rate was explained by the adsorption of the different gases on the surface: The solubility of argon in water is much higher than that of nitrogen and, due to the high volume of the adsorbed gas, bubble cavitation starts much earlier than in the case of sonication under standard conditions with air-saturated water.59
Plot of the pit density as a function of sonication time. The pits form due to the impact of collapsing cavitation bubbles. The substrate was (soft) aluminium, patterned with hydrophobic and hydrophilic stripes with the width of tens of micrometers. The pit density refers to that on the hydrophobic areas. The reference area was 230 μm × 167 μm. The impact is most pronounced when surface nanobubbles or micropancakes were pre-formed by applying the solvent exchange process prior to sonication. Note that for the comparison of the different procedures, the shorter sonications times are the relevant ones, as—after too long sonication times—the cumulative effects of the collapsing bubbles damage the surface too heavily. Reprinted with permission from Belova et al., Chem. Sci. 4, 248 (2013). Copyright 2013 Royal Society of Chemistry.
Plot of the pit density as a function of sonication time. The pits form due to the impact of collapsing cavitation bubbles. The substrate was (soft) aluminium, patterned with hydrophobic and hydrophilic stripes with the width of tens of micrometers. The pit density refers to that on the hydrophobic areas. The reference area was 230 μm × 167 μm. The impact is most pronounced when surface nanobubbles or micropancakes were pre-formed by applying the solvent exchange process prior to sonication. Note that for the comparison of the different procedures, the shorter sonications times are the relevant ones, as—after too long sonication times—the cumulative effects of the collapsing bubbles damage the surface too heavily. Reprinted with permission from Belova et al., Chem. Sci. 4, 248 (2013). Copyright 2013 Royal Society of Chemistry.
The effect of the interfacial gases on the cavitation process became even more evident when nanoscale gaseous domains (nanobubbles and micropancakes) were pre-formed by solvent exchange on the surface. AFM images (Figs. 8(a) and 8(b)) showed that after solvent exchange, nanobubbles formed in both hydrophilic and hydrophobic areas. However, the gas volume was larger in the hydrophobic areas, and, after sonication, more pits formed there, as shown in Figures 8(c) and 8(d). The suggested explanation is that the cavitation processes were facilitated due to the large amount of gas in the form of nanobubbles or micropancakes accumulated at the interface by the solvent exchange.24 The implication of the finding is that surface nanobubbles may be applied to accelerate heterogeneous cavitation, which, e.g., is required in the removal of coatings or to clean surfaces.
AFM images of the patterned substrate after the ethanol/water solvent exchange procedure. (b) is an enlargement of (a). SEM images of the pits after 3 min (c) and 10 min (d) of sonication, which was performed after the formation of surface nanobubbles or micropancakes by the ethanol-water solvent exchange process. Reprinted with permission from Belova et al., Chem. Sci. 4, 248 (2013). Copyright 2013 Royal Society of Chemistry.
AFM images of the patterned substrate after the ethanol/water solvent exchange procedure. (b) is an enlargement of (a). SEM images of the pits after 3 min (c) and 10 min (d) of sonication, which was performed after the formation of surface nanobubbles or micropancakes by the ethanol-water solvent exchange process. Reprinted with permission from Belova et al., Chem. Sci. 4, 248 (2013). Copyright 2013 Royal Society of Chemistry.
B. Surface nanobubbles for nanomaterials
Surface nanobubbles can potentially provide a convenient template for fabrication of nanopatterns and nanostructures. Fig. 9 demonstrates the formation of nanorings around the nanobubble boundary during the evaporation of a drop of a gold nanoparticle suspension. The nanobubbles formed at the interface between the substrate and the suspension of gold nanoparticles. As the droplet evaporated, the nanoparticles accumulated around the nanobubbles. When the system was dried, the nanobubbles bursted, and the nanoparticles were pushed to the boundary of the nanobubbles and formed rings. The nanobubble rupture is also a familiar process at the end of boiling at high temperature, as discussed above.47
Scheme depicting the three main steps of the evaporation process of a nanocolloid suspension confining immiscible nanobubbles and leading to the formation of isolated rings of Au nanoparticles. (a) A drop of Au nanoparticle suspension is deposited on the surface. Nanobubbles form on the surface inside the drop. (b) The drop evaporates, and the nanoparticles coat the nanobubbles. (c) Finally, nanobubbles rupture, and the nanoparticles deposit on the rim and form nanorings. Reprinted with permission from Darwich et al., Nanoscale 3, 1211 (2011). Copyright 2011 Royal Society of Chemistry.
Scheme depicting the three main steps of the evaporation process of a nanocolloid suspension confining immiscible nanobubbles and leading to the formation of isolated rings of Au nanoparticles. (a) A drop of Au nanoparticle suspension is deposited on the surface. Nanobubbles form on the surface inside the drop. (b) The drop evaporates, and the nanoparticles coat the nanobubbles. (c) Finally, nanobubbles rupture, and the nanoparticles deposit on the rim and form nanorings. Reprinted with permission from Darwich et al., Nanoscale 3, 1211 (2011). Copyright 2011 Royal Society of Chemistry.
Nanobubbles were also used as templates to prepare crystals and nanoscale containers. Nanobubbles may be directly involved in the formation of tube-shaped CaCO3 crystals. By nanobubble-templated crystal growth, one could achieve the formation of crystals with round external shape without intervening chemicals.62 Surfactant-coated submicron bubbles can act also as templates for the formation of conducting polymer (polypyrrole) microcontainers with morphology like bowls, cups, and bottles. The morphological features of the conducting polymer containers can be simply controlled by the electrochemical polymerization conditions.63
Fig. 10 shows the deposition process of a conductive polymer film on a surface with pre-formed nanobubbles. The in-situ AFM images show the morphology of the film. The nanopores on the film were attributed to the presence of electrochemically generated hydrogen nanobubbles on the substrate. The size and the number of the nanopores in the film were simply tuned by the nanobubble formation under the applied electric potentials and reaction times.64
Scheme depicting the steps for nanobubble-templated nanoporous film and AFM images of the nanobubbles (left) and the film (right). Reprinted with permission from Hui et al., Electrochem. Commun. 11, 639 (2009). Copyright 2009 Elsevier.
Scheme depicting the steps for nanobubble-templated nanoporous film and AFM images of the nanobubbles (left) and the film (right). Reprinted with permission from Hui et al., Electrochem. Commun. 11, 639 (2009). Copyright 2009 Elsevier.
C. Nanobubbles in micro- and nano-fluidic systems
Already, de Gennes22 speculated that a very thin layer of gas at the solid-liquid interface would be sufficient to lubricate the fluid and reduce the drag in the fluid transport. This application of nanobubbles, although appealing, is currently hindered by the low and uncontrollable production rate of nanobubbles over a large surface area. A more successful example of the application of surface nanobubbles in nanodevices is the propulsion of a nanorod:65 A Janus nanorod—e.g., a rod with one more hydrophobic and one more hydrophilic side—can self-propel in a directional motion, thanks to nanobubble formation on one segment (Fig. 11). In the example of Paxton et al.,65 the Janus nanorod consisted of one Pt and one Au segment can move autonomously in a 2%–3% aqueous solution of hydrogen peroxide, due to the catalytic formation of oxygen at the Pt end. The force along the rod axis was generated by an oxygen concentration gradient, which leads to an interfacial tension force. More recently, several other types of nanomotors were designed to achieve self-propelled motion due to bubble formation from catalytic reactions.66–68
Schematic diagram of an autonomously moving Pt/Au nanorod in a peroxide solution, due to the shown catalytic reaction, generating oxygen nanobubbles on the nanorod. Reprinted with permission from Paxton et al., J. Am. Chem. Soc. 126, 13424 (2004). Copyright 2004 American Chemical Society.
Schematic diagram of an autonomously moving Pt/Au nanorod in a peroxide solution, due to the shown catalytic reaction, generating oxygen nanobubbles on the nanorod. Reprinted with permission from Paxton et al., J. Am. Chem. Soc. 126, 13424 (2004). Copyright 2004 American Chemical Society.
IV. MORE CHALLENGES AND OUTLOOK
Various theories have been suggested to account for the remarkable stability of surface nanobubbles: Among them, contamination on the surface, hindering gas exchange and reducing the surface tension,69 a dynamic equilibrium theory,70–72 postulating that the gas outflux is balanced by some gas influx, and finally pinning, together with cooperative effects of the nanobubbles and their diffusive interaction with the liquid in the far-field.73,74 These theories have meanwhile been made quantitative, leading to predicted phase diagrams and temperature dependences. However, it is fair to say that all of these theories have problems and that none of them is generally accepted.
Presently, we are in a phase in which incidental information on surface nanobubbles is more and more replaced by systematic and quantitative experimental, theoretical, and numerical studies. While in the early years, progress came mainly from colloidal science; in recent years, it became clear that the fluid dynamics of and around the surface nanobubbles is crucial for their understanding.
In spite of all the progress in recent years, it is clear that the field has still a long way to go to fully understand surface nanobubbles and to explore possible applications. We are now in need of standardized procedures to reproducibly produce surface nanobubbles, without any trouble from contamination. We are now also in need of new experimental imaging and detection methods, complementary to AFM with all its limitations with respect to time resolution and difficulties in applying this technique in water, with less ambiguity in the interpretation of the data. With such techniques, controversial observations should be reproduced—or falsified. We are in need of numerical models which couple nanoscale MD simulations with fluid dynamics approaches. And finally, we are in need of a theoretically framework which can account for the many puzzling findings.
Our present lack of understanding of surface nanobubbles reveals that there is a major gap in our knowledge of surface science and, in particular, of hydrophobic surfaces and their interaction with water. Filling this gap is a major challenge. Nanobubbles bring together neighbouring disciplines, namely physics of fluids, colloidal science, surface chemistry, soft matter, optical and imaging sciences, nano-technology, and perhaps an even broader group of scientists who might be key to understanding this puzzle.
ACKNOWLEDGMENTS
We would like to thank all of our coworkers and colleagues for the joint work and the many discussions we had on nanobubbles over the years. X.H.Z. and D.L. gratefully acknowledge financial support by the Australian Research Council (FT120100473) and by an ERC Advanced Grant, respectively.