Low power cw laser radiation at λ=1.32µm was coupled into a chemically etched, metalized Near-Field Scanning Optical Microscopy (NSOM) fiber probe to generate a stable microbubble in water as well as in other fluids. The microbubble, which was attached to the end face of the fiber probe, was used to trap, manipulate and mix micron sized glass, latex and fluorescent particles as well as biological material.
© 2004 Optical Society of America
Microparticles as well as biological material can be trapped and manipulated in fluids using highly focussed laser radiation, i.e., “optical tweezers” [1,2] as well as by using more conventional contact techniques such as tapered probes . Recently we have demonstrated that 5–10 mW of cw laser radiation coupled into a fiber probe with a partially metalized tip, similar to those used in Near-Field Scanning Optical Microscopy (NSOM) , can trap a glass microsphere in 3-D in water . The particle was trapped radially by an annular light field emerging from the probe tip and confined in the third dimension by the balance of an electrostatic force of attraction towards the tip and the light scattering force pushing the particle away from the tip. We remarked in this paper that if the laser power was raised beyond the threshold for trapping it was possible to initiate a stable microbubble attached to the probe tip. The microbubble was formed by means of highly localized superheating and spontaneous boiling of the liquid which occured in the last few microns of the tapered tip at the position of maximum light absorption and tip heating. Similar microbubbles have been formed using small resistive heating elements and used as controllable actuators to manipulate bioparticles . Indeed microbubbles have been studied for years since the ability to control the creation and collapse of the bubbles has led to interesting optical cross-connect switching , microinjector , bubble valve  and bubble-induced micromixing microfluidic applications .
In previous work with optical fibers transient microbubbles with lifetimes of hundreds of microseconds have been created at the fiber end face [11,12]. The microbubbles were used to displace highly absorbing blood  or water  so that pulsed near-infrared laser radiation could ablate tissue adjacent to the bubble for surgical applications. Recently microsecond duration, semiconductor diode laser radiation coupled into a single-mode fiber with a Ga metalized output face was used to generate a transient microbubble of a few microns in size in liquid nitrogen . In this paper we demonstrate that a modest cw laser power (10 mW) coupled into a fully metalized version of a selectively chemically etched, conically tapered NSOM probe tip can produce a stable microbubble in a number of liquids such as water. A microbubble formed in water contains heated water vapour and possibly some trapped air. It expands rapidly in a few seconds to a diameter which depends on the laser power. Stable microbubbles with diameters up to 400µm which can last up to several hours have been produced. To the best of our knowledge this is the first time that a microbubble with such a long lifetime has been generated at the tip of a fiber probe. We will show that once a microbubble has been created the laser power can be turned off and the microbubble attached to the end of the fiber probe can effectively be used to capture either individual microparticles or large numbers of particles attached to a surface in a vacuum cleaner mode of operation. Alternatively, if the laser power is allowed to continuously couple into the fiber, post bubble formation, strong temperature gradients are produced on the bubble surface. The temperature gradients result in significant surface tension gradients, which produce convective fluid flow to rapidly mix  large numbers of particles on the surface of the bubble via the Marangoni effect. Quantitative measurements of the growth and decay dynamics of the microbubbles as well as the importance of the probe geometry in making stable microbubbles will be described in a subsequent publication.
One of the long term objectives of our microbubble research is to use the mobile fiber-bubble probe to benignly trap and deliver a living cell to a desired location in similar fashion to what can be achieved with optical tweezers . A fiber optic based probe may be particularily advantageous at trapping and manipulating material in turbid media and/or in large volumes of fluid where conventional optical tweezers have great difficulty due to attenuation of the trapping beam by light scattering and difficulty in achieving a tightly focussed laser beam over distances >200–300µm . It is a further objective of our research to investigate the possiblity of analyzing the cell spectroscopically or by other means while it is immobilized on the bubble membrane.
As a first step towards these objectives we report the results of a preliminary study to determine whether stable microbubbles can be produced at the end of a fiber probe in weak acid, base, alcohol and salt solutions as well as in minimal essential medium (MEM) and gelatin. The results are encouraging and suggest that it may be possible to trap and manipulate biological material in a wide range of media. We will also show in this paper that the use of the Marangoni effect to mix biological material  results in thermal damage. However a novel probe design is also described which largely overcomes the thermal damage problem.
The experimental layout for the microbubble experiments is shown in Fig. 1. The light source consisted of a polarized Lightwave Electronics cw laser operating at a wavelength of 1.32 µm. The laser beam passed through a mechanical chopper to modulate the beam power, when required, followed by a half-wave plate to permit rotation of the polarization. The beam then passed through a polarizing cube beamsplitter and was coupled into the core of an approximately 1m long single-mode fiber using a lens with a numerical aperture NA=0.32. Laser powers of 5–20 mW were typically coupled into the fiber. The output end of the fiber was placed inside a few mm thick droplet of liquid. An overhead long working distance microscope objective (NA=0.5) together with a Panasonic video camera or a Nikon Coolpix model 995 digital camera were used to image the fiber tip region.
Some of the light traveling down the fiber core was retroreflected from the metalized probe tip and returned through the lens to be directed by the beamsplitter to a Newport model 1830-C powermeter for detection. The back-reflected signal at the detector consisted of a small depolarized component due to the polarization scrambling nature of the fiber. This signal was optimized to ensure that the laser light was efficiently coupled into the small diameter (4µm) core and then directed all the way to the fiber tip. The combination of the half-wave plate and the polarizing beamsplitter also acted as a variable attenuator to permit continuous variation of the laser power delivered to the fiber probe.
The fiber probes were made from a high GeO2 doped, single-mode (at λ=1.55µm) telecom fiber obtained from Fibercore Inc.. The fiber which had an initial diameter of 125µm was etched at room temperature in a solution of 10:1 buffered oxide etchant BOE (i.e. 10 volumes of ammonium fluoride NH4F to one volume of hydrofluoric acid HF) for 70 minutes to produce the tip. The fiber diameter after etching was ≈118 µm. In some cases the fiber was pre-etched in a weak BOE solution to further reduce the probe diameter.
It has been demonstrated previously that preferential etching of the fiber cladding relative to the core in BOE results in the growth of a tapered conical shaped structure whose base corresponds to the diameter of the fiber core [18,19]. However there are a number of telecom fibers such as the Fibercore Inc. fiber, which as a result of the manufacturing process, have a narrow low index of refraction center inside the high index core. Etching of these fibers in BOE produces a hollow region in the center of the conical structure . An etched Fibercore Inc. hollow core fiber probe tip is shown in Fig. 2. These probes when suitably metalized and subjected to focussed ion-beam milling have recently been used by the authors to generate an annular light distribution for optical trapping of glass microparticles . In those experiments it was mentioned that when the laser power coupled into the fiber was increased above the power level for particle trapping a microbubble was created centered on the conical tip. The experiments described in this paper used the etched Fibercore probes however the probes were entirely coated with a ≈100 nm thick coating of platinum in order to make them optically opaque. Platinum was chosen as the metal due to its very high melting point (2042°K) compared to aluminum (933°K) which is generally used to coat NSOM probes.
Other probe geometries such as cleaved fibers and pulled tapered fibers were investigated and their performance will be discussed in a follow-up paper. However by far the most reliable performance was obtained with the probe geometry shown in Fig. 2. The flat-topped, short stature of the probe tips made them very robust and these tips were essentially undamaged after direct contact with surfaces. The small diameter of the fiber core made it possible for modest laser powers to heat the conical tapered tip to temperatures, which spontaneously initiated a tiny microbubble. Presently it is not completely clear whether it is absolutely necessary to have a hole in the tapered conical tip. The metalized hole may result in enhanced light attenuation and therefore more efficient tip heating. However more experiments with very similar probe tips but without the hole are required to be sure.
In order to test the generality of the stable microbubble formation process various liquids were used in the experiments including room temperature tap water, distilled and deionized water, a 5% salt water solution (using commercial iodized salt), a 10% solution of hydrochloric acid HCl (nominal 37% by wt. from Anachemia), ethanol solution (95%) and a 10% by weight sodium hydroxide NaOH solution. No attempt was made to degas the liquids. We also tested minimum essential medium (MEM from Wisent), gelatin (from Jello™) and blood serum from a liver product sold commercially. In other experiments a number of different particles were added to the various media to establish whether they would readily adhere to the bubble membrane and therefore be trapped. The particles included 2µm diameter solid borosilicate glass microspheres from Structure Probe Inc., 0.5µm and 1.0µm latex microspheres from Duke Scientific and 100 nm diameter transfluorspheres (absorption/fluorescence=488nm/560nm) from Molecular Probes.
3.1 Particle trapping in water using a microbubble attached to an NSOM fiber probe
Once a stable bubble was initiated at the fiber tip the laser power could be shut off. The bubble which was firmly attached to the fiber end face could be used to pick up particles. Figure 3(a) shows the microbubble pulling on a number of 2µm glass spheres attached to the surface of an etched fiber. The significant distortion of the bubble shape indicates the strong surface tension force involved in the pulling action. The picture on the right side of Fig. 3 demonstrates that it was possible to delicately pluck a single microparticle from the surface. The small diameter fiber to which the particles were initially attached was suspended in the water over a distance of ≈1cm before being clamped at one end. In this configuration it acted as a thin lever sensitive to applied forces which we calculate to be in the nanoNewton range using the formulae from Ref.  for the spring constant of a cylindrical rod clamped at one end. We observed that it was possible to extract particles attached near the end of such a fiber or deposit particles onto the fiber without noticeably disturbing the position of the fiber indicating the sensitivity of the bubble trapping and deposition procedure.
After the particle was gently plucked from the surface it was absorbed into the bubble membrane and then over a period of seconds drifted under the influence of gravity to sit on the bottom portion of the bubble surface. A video of such a trapped 2µm solid glass microsphere showed that the particle was immobilized on the bottom surface of the bubble membrane with less than 1µm of movement over a period of many minutes. Sub-micron diameter particles such as the fluorospheres still exhibited some Brownian motion on the bubble membrane although less than what was observed when they were free to move in 3-D in the fluid.
The fiber probe with an attached bubble containing the trapped particle could be moved to a desired location then using a shearing action the bubble could be detached onto a surface. The bubble would then slowly collapse in a symmetric fashion depositing the particle or particles at a specific location on the surface usually with an accuracy of a few microns.
It was also possible to effortlessly sweep up many glass particles from the surface of a fiber in a vacuum cleaner mode of operation as shown in the video attached to Fig. 4. Hundreds of particles could be vacuumed in this manner and then deposited at a desired location as described above.
There was another mode of trapping which made it possible to trap particles dispersed in a fluid volume i.e. not adhered to a surface. In this mode the fiber probe without a bubble was moved into the vicinity of the objects to be trapped. At this point the laser beam was blocked and there was no laser power coupled into the probe. The beam was suddenly unblocked allowing a bubble to be created in a time <1s. The strong surface tension gradients that occurred during bubble formation produced long range (mm) convective currents, which caused material to jump to the surface of the bubble. Particles could therefore be drawn in from great distances and attached to the bubble. A demonstration of this trapping mode was made using 1µm diameter latex spheres and is shown in the video attached to Fig. 5.
3.2 Mixing of particles on the surface of a microbubble in water
We investigated two ways of mixing particles using a microbubble attached to a fiber probe. The first method utilized the fact that the bubble diameter during its expansion phase depended on the laser power delivered to the probe tip. By using a mechanical chopper to modulate the laser power it was possible to pulsate (e.g., at 5–10Hz) the bubble to produce a microstreaming mixing of particles trapped on the bubble surface.
However a much more powerful mixing method was to allow the laser radiation to heat the probe microtip post bubble formation to create large temperature and therefore large surface tension gradients on the surface of the bubble . As mentioned previously this resulted in cyclonic convective fluid flow via the Marangoni effect. In Fig. 6 a video shows tens of thousands of glass microspheres distributed throughout the water converging towards a heated microbubble. The convergence began with pulsating waves of particles which after a few seconds became a continuous focussed flow of particles at speeds of a few mm/s towards the probe tip. Some of these spheres (≈104) collided and stuck to the bubble and proceeded to assemble themselves in a monolayer on the top portion of the bubble surface. This self-assembly process was assisted by Marangoni convection continually trying to move the particles along the bubble surface allowing particles stacked on top to slip down onto the surface. Indeed as the laser power was raised stronger convection currents drove the particles into cyclonic flow patterns around the bubble surface. Cessation of the laser power by rapidly blocking the beam immediately stopped the convective flow. We also show in Fig.6 (long version of the video) that it was possible to penetrate the bubble membrane with a narrow fiber that could be moved into the bubble interior without disrupting the bubble integrity. This fiber could be used to pull off the bubble from the fiber probe. When a second bubble previously created and attached to the side of a fiber was brought near the bubble attached to the probe the flow pattern of microspheres was coupled from the heated bubble to the other bubble even if the bubbles were separated by many microns (Fig. 6 video clip#3).
Similar self assembly and mixing of particles was produced with latex spheres (early stage shown in Fig. 5) and 100 nm diameter fluorescent beads. The flow pattern on the bubble surface was often very complicated, however the pattern most often observed is shown in Fig.7.
It consisted of a northern hemisphere (clockwise) and southern hemisphere (counterclockwise) counterpropagating cyclonic flow pattern. A blow-up of this region using 1µm latex spheres for flow visualization is shown in the video attached to Fig. 8. Very fast particle speeds in excess of 100 µm/s were generated in the cyclonic flow regions. We also observed two symmetric regions of cyclonic flow located outside of the bubble and near the probe cladding, which were driven by the convective flow initiated on the bubble. As the bubble was heated and grew in size these two cyclonic flow regions were pushed backwards away from the bottom of the bubble.
3.3 Microbubble trapping of particles in various fluids
As a result of the success trapping particles in water with the fiber-bubble technique we performed a preliminary investigation into the possiblity of performing similar experiments in a wide range of fluids. Since the microbubble-particle trapping process, in which the laser beam was blocked, was very benign it may be possible to trap biological material without damage. Therefore the threshold laser power for bubble formation; the degree that particles stuck to the bubble surface and the stability of the bubble were evaluated for various liquids, serums and gelatin media relevant to both chemical and biological applications.
The most reliable bubble formation and particle trapping occurred in water whether it was distilled and deionized or tap water. There was a consistently ≈25% higher laser power threshold for bubble formation (12.5 mW coupled into the fiber) for distilled and deionized water compared to tap water presumably due to the lower concentration of trapped air or other nucleation centers. We also tested the bubble performance in dilute acid, base and salt solutions. A general observation was that although similar results to water could be obtained the stability and particle trapping capability of the solutions were usually poorer compared to water. For example, an ≈10% solution of HCl increased the power threshold for bubble formation to 23 mW. The bubbles were stable for a similar time period compared to water, however the bubble was less effective at trapping the 2 µm spheres and it was easier for the bubble to detach from the base of the fiber when trying to pluck a particle off a surface. In a 10% by weight NaOH crystals in a water solution the bubble formation process was unstable resulting in microjet formation, which will be described in a subsequent article. Dilution of the solution by a factor of 3 permitted stable bubble formation at a couped laser power of ≈23 mW. The glass microparticles could be trapped and mixed however there was still an increased tendency to multiple bubble formation. For a ≈5% salt solution i.e. comparable to the salt concentration of sea water, the power threshold was similar to that of distilled-deionized water (12.5 mW), however it was harder to grow large (>50µm) bubbles and there was a smaller window of laser powers (a few mW) for stable bubble formation and trapping.
The observations for 95% proof ethanol were profoundly different from the other liquids including water. The threshold for bubble formation was about one half that of water due to ethanol’s lower boiling point (80°C). The bubble decay data were very interesting. If the laser beam was suddenly blocked after bubble initiation the microbubble didn’t decay slowly it simply collapsed. For example, a 80µm diameter bubble collapsed in less than 33ms. The sudden collapse of the bubble was not observed with any of the other liquids. On the other hand if the laser power coupled into the fiber probe was slowly decreased (over a period of seconds) a 70µm initial bubble diameter didn’t collapse immediately but decayed taking ≈30s to disappear. This bubble lifetime is still some x20 shorter than the bubble lifetime which we obtained in water. Similar short decay times have been observed in Ref.  using cw laser radiation focussed onto an absorbing (black paint) surface submerged in alcohol to produce the bubble. It appears that bubbles produced in ethanol are fragile and can be disrupted by sudden changes in microtip heating conditions. In future experiments it might be possible to take advantage of the short decay time by mixing alcohol with water to decrease the bubble lifetime to more quickly deposit trapped material onto a surface.
Stable bubble formation was successfully obtained in gelatin (Jello™). The threshold was 27 mW and the bubble grew more slowly relative to bubbles formed in water. The microbubble lasted x15 longer in the Jello™ (e.g. 1 hour in Jello™ for a 50 µm diameter bubble) and the fiber-bubble could be translated quite easily in 3-D. Similar longevity of microbubbles in gelatin has been observed in Ref. .
A microbubble was also produced at the end of our probe in minimum essential medium (MEM) frequently used in biology experiments. The power threshold was 23 mW and similar to gelatin the bubble lifetime was longer than that observed in water. The bubble could be vigourously moved through the serum. However an interesting observation was made during attempts to trap glass microspheres. The bubble membrane became deformed by the 2 µm glass microsphere rather than being able to envelop and absorb the particle into the membrane. Of course this behaviour made it more difficult to trap particles.
A stable microbubble was also successfully produced in undiluted blood serum (from a liver product) although presently it has not been possible to use the bubble to trap blood cells as they failed to adhere to the surface of the bubble.
Clearly the use of fluids with different physical and thermal properties from water creates distinctive problems which can make it more difficult to achieve a stable microbubble which can trap particles or biological material. The most significant problem that must be addressed is the sticking power of the particles and biological material to the bubble surface. In future we will be investigating the role of different surfactants  either applied directly to the probe tip or mixed in solution at increasing the bubbles trapping capability.
There are however some encouraging signs of being able to use this microbubble technique to trap biological specimens moving about in water. Figure 9 shows that it was possible to trap a small (400 µm) crustacean (Daphnia often refered to as a water flea) which moved at speeds of a few mm/s in a sample of pond water. The creature was trapped during the bubble initiation stage due to long range convective fluid forces which brought it from approximately a mm away to get stuck on the membrane of the bubble. The laser power was shut off immediately after bubble initiation to avoid damaging the creature. After the bubble diameter had decayed to a point where it couldn’t hold the creature it was released and then scurried off apparently none the worst for wear. It was also possible to trap and release micron sized pond specimens. This technique may be useful for trapping and manipulating other moving specimens such as spermatazoa [25,26]. Once trapped the sperm can be moved and deposited at a desired location on an Oocyte.
A major concern with using a microheater based approach rather than an acoustic micromixing technique  to mix biological material is how to avoid thermal damage such as protein denaturing which can occur at even modest temperature rises (≈60°C). The severity of the problem is illustrated in Fig. 10(a) which shows that the conical tip of a fiber probe heated in air gets white hot when using laser power levels required for mixing in water (i.e., 20 mW). Heating the probe tip further melted the fused silica core (estimated to occur at T ≈1800°K). This figure also indicates that there is considerable heating well away from the conical structure. Blockage of the laser beam stopped the probe from glowing almost immediately due to the fast thermal response of the small heated zone. When the probe was immersed in water the tip temperature was estimated to be an order of magnitude lower due to efficient cooling by the water circulating around the tip region. However the temperature at the base of the microbubble was still sufficiently high to damage biological material.
Our attempts to mix blood by allowing laser power into the fiber probe post bubble formation ultimately resulted in coagulation of the blood which was in the vicinity of the base of the bubble but well away from the probe tip. This would seem to rule out using our fiber-bubble probe to mix biological material although such heating effects could be advantageous for mixing chemicals on a microscale. However we have recently developed a novel probe in which only the small conical tapered region gets hot. This was accomplished by removing the metal coating from around the base of the probe tip, using focussed ion beam nanomachining. This breaks the thermal conductivity pathway from the metalized hot tip to the remainder of the metalized probe. This pathway dominates in cooling down the tip compared to thermal conduction in the silica. Figure 10(b) shows a dramatic picture of the end face of the novel probe being laser heated in air and glowing white hot only in the tip region which is a few cubic microns in size.
Since the base of the bubble, which sits on the flat cladding surface, tends to seal-off the microtip region from the biological material and since the cladding end face is now “cold” biological material should not be heated to the same extent with this probe. Indeed preliminary tests with this novel design showed that there was a dramatic decrease in the thermal damage done to latex spheres or to biological material such as micron sized creatures found in common pond water during the bubble mixing procedure. In future we hope to report detailed experiments on successful mixing of biological material with minimal thermal damage.
One can think of the generation of a stable microbubble at the end of a fiber probe as the creation of a miniature mobile biochamber or as a mobile chemical micromixing chamber. The bubble might be used to trap and manipulate biological (e.g., a cell) or chemical (colloidal particles) samples. In a simple manner the bubble can be released at a desired location and another bubble created. This might be useful for the transport of toxic particles without the particles touching the fiber probe. We have shown in Fig. 6 that the bubble can be penetrated by a fiber probe and as a result of a self-sealing action around the regions where the probe enters the bubble there is no significant acceleration in the decay of the bubble. It should therefore be possible to insert a metallic electrode to sit behind some biological or chemical sample trapped on the bubble surface. A second electrode can be brought close to the outside of the bubble surface and a voltage applied across the material to study its electrical properties . Alternatively material trapped on the bubble can be probed using tapered probes  which can penetrate the bubble membrane, such as micropipettes for sampling and NSOM probes for high resolution spectroscopy.
Since the interior of the microbubbles are essentially optically transparent and large particles (>1µm) or possibly a cell can be immobilized on the bubble membrane it should be possible to perform spectroscopy on the trapped material using a single fiber probe to first generate a bubble then transmit light through the bubble to the trapped material and finally collect fluorescence light from the material back through the fiber for detection. In a separate experiment we have demonstrated that it was possible to make the probe optically transmitting by removing the metal from the top face of our probe tip and still be able to create a bubble. This technique would be particularily useful for sensing applications in highly turbid and/or absorbing media.
The experimental setup required to generate controlled microbubbles described in this paper is quite straightforward and can be simplified even further compared to that shown in Fig. 1. Since the heating mechanism is based upon the absorption of light in a high melting point metallic coating a low cost modest output power (20 mW), unpolarized, diode laser at visible to near infra-red wavelengths can be used. The input end of the fiber can be directly connected to the laser and the laser power adjusted electronically.
Low power cw laser radiation coupled into an etched optical fiber has been used to heat a metalized microtip to generate a stable bubble by spontaneous and highly localized boiling of the liquid. The microbubble, which is firmly attached to the end of the fiber probe, can be easily manipulated in 3-D and used to delicately extract small particles from a surface. During the microbubble formation stage it was also possible to trap material onto the bubble surface which was dispersed within the fluid. The microbubbles were remarkably resilient and could withstand penetration by an external probe. This opens up the possibility of positioning probes on either side of biological material trapped on the surface of the bubble. Mixing biological material on the surface of a bubble without thermal damage will require special control over the temperature distribution on the probe surface permitted by new probe designs like the focussed ion beam nanomachined probe described in this paper.
The hollow tip fiber probe design provides control over the microbubble both in space and in time and should be very useful for the further studies of bubble formation, growth and collapse including the role of trapped gases and surfactants on these processes. The microbubble experiments are interesting in their own right since a bubble membrane provides an alternative medium compared to glass slides to study immobilized biological material.
The authors would like to thank Dr. Linda Johnston of the Steacie Institute at NRC for providing us with samples of fluorescent particles. We would also like to thank Dr. Christophe Py and Dr. Karim Faid of the Institute for Microstructural Sciences at NRC for helpful discussions on microbubble mixing applications and Dr. Milan Kovar at the University of Western Ontario for his suggestion of using a microbubble as a mobile “biochamber”.
References and Links
3. D.T. Chiu, A. Hsiao, A. Gaggar, R.A. Garza-Lopez, O. Orwar, and R.N. Zare, “Injection of ultrasmall samples and single molecules into tapered capillaries,” Anal. Chem. 69, 1801–1807 (1997). [CrossRef] [PubMed]
4. M.A. Paesler and P.J. Moyer, “Near-Field Optics,” (J. Wiley and sons,New York1996).
5. R.S. Taylor and C. Hnatovsky, “Particle trapping in 3-D using a single fiber probe with an annular light distribution,” Opt. Express 11, 2775–2782 (2003). http://www.opticsexpress.org/abstract.cfm?URI=OPEX-11-21-2775 [CrossRef] [PubMed]
6. R.B. Maxwell, A.L. Gerhardt, M. Toner, M.L. Gray, and M.A. Schmidt, “A microbubble-powered bioparticle actuator,” J. of Microelectromechanical Systems 12,630–640 (2003). [CrossRef]
7. J.E. Fouquet, “Compact optical cross-connect switch based on total internal reflection in a fluid-containing planar lightwave circuit,” Trends Opt. Photon. 37,204–206 (2000).
8. Chang-Jin- Kim, “Microfluidics using the surface tension force in microscale,” Proc. of SPIE 4177,49–55 (2000). [CrossRef]
9. S.Z. Hua, F. Sachs, D.X. Yang, and H.D. Chopra, “Microfluidic actuation using electrochemically generated bubbles,” Anal. Chem. 74,6392–6396 (2002). [CrossRef]
10. R.H. Liu, J. Yang, M.Z. Pindera, M. Athavale, and P. Grodzinski, “Bubble-induced acoustic micromixing,” Lab Chip 2, 151–157 (2002). [CrossRef]
11. P.E. Dyer, M.E. Khosroshahi, and S.J. Tuft, “Studies of laser-induced cavitation and tissue ablation in saline using a fibre-delivered pulsed HF laser,” Appl. Phys. B.56, 84–93 (1993).
13. Ton G. van Leeuwen, M.J. van der Veen, R.M. Verdaasdonk, and C. Borst, “Noncontact tissue ablation by Holmium:YSGG laser pulses in blood,” Lasers in Surgery and Medicine 11, 26–34 (1991). [CrossRef] [PubMed]
14. H. Pratisto, M. Ith, M. Frenz, and H.P. Weber, “Infrared multiwavelength laser system for establishing a surgical delivery path through water,” Appl. Phys. Lett. 67, 1963–1965 (1995). [CrossRef]
15. K.F. MacDonald, V.A. Fedotov, S. Pochon, B.F. Soares, and N.I. Zheludev, “Oscillating bubbles at the tips of optical fibers in liquid nitrogen,” Phys. Rev. E 68,027301–3 (2003). [CrossRef]
16. D.W. Berry, N.R. Heckenberg, and H. Rubinsztein-Dunlop, “Effects associated with bubble formation in optical trapping,” J. Mod. Opt. 47, 1575–1585 (2000).
17. Y. Xin-Cheng, L. Zhao-Lin, G. Hong-Lian, C. Bing-Ying, and Z. Dao-Zhong, “Effects of spherical aberration on optical trapping forces for Rayleigh particles,” Chin. Phys. Lett. 18, 432–434 (2001). [CrossRef]
18. T. Pangaribuan, K. Yamada, S. Jiang, H. Ohsawa, and M. Ohtsu, “Reproducible fabrication technique of nanometric tip diameter fiber probe for photon scanning tunneling microscope,” Jpn. J. Appl. Phys. 31, L1302–L1304 (1992). [CrossRef]
19. P. Burgos, Z. Lu, A. Ianoul, C. Hnatovsky, M.-L. Viriot, L.J. Johnston, and R.S. Taylor, “Near-field scanning optical microscopy probes:a comparison of pulled and double-etched bent NSOM probes for fluorescence imaging of biological samples,” J. Microscopy 211, 37–47 (2003). [CrossRef]
20. R.S. Taylor and C. Hnatovsky, “High resolution index of refraction profiling of optical waveguides,” Proc. SPIE 4833, 811–819 (2003). [CrossRef]
21. D. Sarid, Scanning Force Microscopy, (Oxford University press,Oxford,1994).
22. Y. Otani, Y. Matsuba, and A. Osaka, “Microbubble actuator using photothermal effect,” Proc. SPIE 4902, 83–86 (2002). [CrossRef]
23. D.E. Yount, E.W. Gillary, and D.C. Hoffman, “A microscopic investigation of bubble formation nuclei,” J. Acoust. Soc. Am. 76, 1511–1521 (1984). [CrossRef]
24. W. Wang, C.C. Moser, and M.A. Wheatley, “Langmuir trough study of surfactant mixtures used in the production of a new ultrasound contrast agent consisting of stabilized microbubbles,” J. Phys. Chem. 100, 13815–13821 (1996). [CrossRef]
25. Y. Tadir, T. Ord, W.H. Wright, R.H. Asch, O. Vafa, and M.W. Berns, “Micromanipulation of sperm by a laser generated optical trap,” Fertility and Sterility 52, 870–873 (1989). [PubMed]
26. A. Clement-Sengewald, K. Schutze, A. Ashkin, G.A. Palma, G. Kerlen, and G. Brem, “Fertilization of bovine Oocytes induced solely with combined laser microbeam and optical tweezers,” J. of Assisted Reproduction and Genetics 13, 259–265 (1996). [CrossRef]
27. D.T. Chiu, C.F. Wilson, F. Ryttsen, A. Stromberg, C. Farre, A. Karlsson, S. Nordholm, A. Gaggar, B.P. Modi, A. Moscho, R.A. Garza-Lopez, O. Orwar, and R.N. Zare, “Chemical transformations in individual ultrasmall biomimetic containers,” Science 283(5409), 1892–1895 (1999). [CrossRef]