Fig. 1. Schematic diagram of the experimental layout.

Fig. 2. Scanning electron microscope image of a selectively chemically etched conical tapered Fibercore Inc. probe tip showing a hollow central region.

Fig. 3. Left: Photograph showing the pulling capability of a microbubble firmly attached to the flat fiber cladding surface. A second fiber covered with 2µm diameter glass spheres is shown to the left in this figure, Right: Picture of a microbubble just after delicately picking up a single 2µm glass sphere.

Fig. 4. Video (970Kb) showing a ≈100µm diameter bubble “vacuum cleaning” 2µm diameter glass spheres stuck to a ≈20µm diameter fiber. The conical tip can be seen inside the bubble on the left. In this case the fiber probe was pre-etched to reduce its diameter. The picture on the right shows a number of 2µm glass particles which collected on the bottom surface of the bubble after vacuum cleaning.

Fig. 5. Photograph of patches of agglomerated 1µm latex spheres trapped on the surface of a microbubble. The width of the image is 120µm. The video (1.2Mb) shows the fiber probe launching a bubble in a water solution containing 1µm latex spheres and the subsequent condensation of a large number of the spheres onto the bubble surface.

Fig. 6. Video (short version 2.5Mb) showing a shower of solid 2µm diameter glass spheres converging in an initially pulsating manner from mms away onto the top surface of a ≈350µm diameter bubble which was being heated with ≈20mW of laser power coupled into the fiber probe. The spheres form a monolayer on the bubble surface. The first clip of the long version of the video (13.8Mb) shows cyclonic mixing of ≈10⁴ microspheres on the bubble surface. When the laser was blocked the mixing stopped almost immediately. The second clip shows a ≈20µm diameter fiber being inserted into the bubble and used to remove the bubble from the fiber probe. The last clip shows that when a second bubble with glass microspheres is brought close to the fiber probe containing a bubble which is mixing other microspheres the convective flow is transfered to the second bubble to mix those trapped particles as well.

Fig. 7. Schematic diagram showing the flow patterns which generally occured on and in the vicinity of a heated microbubble.

Fig. 8. Video (634Kb) shows an enlargement of the two counterpropagating cyclonic flow regions on the surface of a bubble in water made visible using 1µm diameter latex spheres. The width of the image is ≈50µm.

Fig. 9. Left: Image of a Daphnia trapped on the backside of a newly generated microbubble. Right: Image of the tiny crustacean moving away after release from the fiber probe.

Fig. 10. (a) Photograph of a Pt-coated hemispherically tipped large core pre-etched fiber heated in air to a white hot temperature using ≈25mW of λ=1.32µm laser light coupled into the fiber probe. (b) Top view of a metalized Fibercore Inc. selectively chemically etched fiber probe in air. Approximately 30mW of laser power was coupled into the probe. A focussed ion beam was used to remove the metal coating inside a 30 µm×30 µm square region but leaving the metal on the conical structure. The heated tip glows white hot while the regions near the base of the probe are relatively cold.