Time-Resolved Studies of Nd:YAG Loser-Induced Breakdown

Plasma Formation, Acoustic Wave Generation, and Cavitation

Jomes G. Fujimoro,* Wei Z. Liaf Erich P. Ippen,* Carmen A. PuliQfiro4 and Roger F. Sreinerr^:

The use of high intensity ultrashort pulsed radiation to produce optical breakdown is an important approach for the surgical treatment of intraocular structures. We have investigated the transient properties of Nd:YAG laser induced breakdown in a saline model using time-resolved spectroscopic techniques. Spatially resolved pump and probe techniques are applied to study the dynamic behavior of the plasma formation, acoustic wave generation, and cavitation processes which accompany the optical breakdown. Measurements of plasma shielding and luminescence indicate that the laser induced plasma forms on a subnanosecond time scale and has a lifetime of several nanoseconds. An acoustic transient is generated at the breakdown site and propagates spherically outward with an initial hypersonic velocity, then loses energy and propagates at sound velocity. Transient heating following the plasma formation produces a liquid-gas phase change and gives rise to cavitation or gas bubble formation. This gas bubble expands rapidly for several microseconds, then slows to reach its maximum size and finally collapses. Invest Ophthalmol Vis Sci 26:1771-1777, 1985

Short pulsed Q-switched and modelocked Nd:YAG important role in protecting the retina from the inci- have recently achieved widespread use in dent laser energy,10"12'25"28 while the clinically useful ophthalmic surgery. The ability of high intensity short effects are primarily produced by the acoustic wave laser pulses to induce optical breakdown and produce and cavitation.10111923'28 lesions from acoustic transient and cavitation effects In this paper we describe the experimental investi- has been especially important for the surgical treat- gation of the transient phenomena associated with ment of intraocular structures in the anterior eye which laser-induced optical breakdown through the applica- are nominally transparent to the laser wavelength.1"12 tion of time-resolved spectroscopic techniques. In par- The transient physical phenomena which accompany ticular we emphasize pump-probe measurements as a the optical breakdown process include laser-induced method for characterizing the transient temporal and plasma formation, acoustic wave generation and prop- spatial evolution of the breakdown process. The pump- agation, as well as cavitation or gas bubble forma- probe technique relies on the use of two laser beams: tion.13"28 These processes are relevant to the clinical one beam, the pump, is used to produce a change in effectiveness of optical breakdown and have direct im- the sample, while the other beam, the probe, measures plications on its physiological and pathological effects the effects produced by the pump. In this investigation in the eye. The increased absorption of the laser-in- a Q-switched Nd:YAG is used as the pump laser to duced plasma has been shown to play a potentially induce optical breakdown in a saline sample while a HeNe laser is used to measure the transient changes in optical properties. The measure- From the Department of Electrical Engineering and Computer ment of both spatial and temporal behavior allows the Science and Research Laboratory of Electronics,* Massachusetts In- characterization of plasma formation, acoustic wave stitute of Technology, Cambridge, Massachusetts; the Howe Labo- generation, and cavitation which accompany the ratory of Ophthalmology, Massachusetts Eye and Ear Infirmary,! breakdown as well as a determination of the physical Boston, Massachusetts; Visiting from Zhongshan University,! dimensions and propagation velocities associated with Guangzhou, People's Republic of China. Supported in part by the Joint Services Electronics Program under these phenomena. While previous researchers have ex- Grant DAAG 29-83-K-003. amined laser-induced breakdown using a variety of Submitted for publication: January 22, 1985. techniques including high speed photography and ho- Reprint requests: James G. Fujimoto, PhD, Department of Elec- lography,1516'20'21'24 transient luminescence and pulse trical Engineering and Computer Science and Research Laboratory transmission detection,17'26'27 and acoustic transducer of Electronics, Massachusetts Institute of Technology, 77 Massachu- measurements,131419'23 the spatially resolved pump- setts Avenue, Cambridge, MA 02139.

1771

Downloaded from iovs.arvojournals.org on 09/26/2021 1772 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / December 1985 Vol. 26

SCHEMATIC OF PUMP - PROBE MEASUREMENT stability. The pulse energy delivered to the sample was varied using beamsplitters and calibrated neutral den- sity filters. Pulse energy measurements were performed with a Scientech Model 361 (Boulder, CO) pyroelectric He-Ne PROBE energy meter. The laser output beam was 10 mm in diameter and focused into the sample cell with a 25- mm focal length lens. Since the Nd:YAG laser em- ployed a diffraction coupled resonator, the spatial mode profile of the output beam was a "doughnut" rather Nd< YAG than a Gaussian mode, and diffraction limit focusing PUMP was not expected. The transient properties of the laser-induced break- down were probed by a low intensity, 1 mW, contin- SAMPLE CELL uous wave helium-neon (HeNe) laser (X = 632.8 nm). In order to obtain a spatial resolution of better than FILTER 10 nm, the probe beam was first expanded to a diameter of 6 mm before being focussed into the sample ceil with a 32-mm focal length lens. The focal point of the FAST PHOTODIODE HeNe probe laser was carefully adjusted to coincide with the center of the breakdown region with the probe beam orthogonal to the pump beam. Since the break- Fig. 1. Schematic diagram of experimental configuration for pump probe time resolved measurements of laser-induced breakdown. Nd: down region extends along the axis of the pump beam YAG laser pump pulses are focused into a sample cell containing with increasing pump pulse energy, the probe align- saline. The transient optical properties are probed by a HeNe laser ment was adjusted after changes in the pump energy. and a high speed photodiode. Spatially resolved measurements are Spatially resolved pump-probe measurements of the performed by displacing the pump and probe laser beams in a di- rection perpendicular to the plane of the figure. breakdown process were obtained by varying the rel- ative separation of the probe laser beam focus from the center of the breakdown region. This was accom- plished by translating the pump beam steering and fo- probe technique is one of the few approaches which cusing optics in a direction orthogonal to the plane of permits the investigation of a wide variety of physical Figure 1 using a l-/um resolution stepping motor me- processes in a simple and consistent experimental chanical stage (Klinger Scientific; Richmond Hill, NY). framework. The center of the breakdown region was determined by noting that the probe beam behavior should be Materials and Methods symmetric for positive and negative displacements. The use of this spatially resolved technique facilitates the The experimental configuration for pump-probe measurement of physical dimensions as well as the dy- measurements of laser-induced optical breakdown is namic behavior of processes involving transient prop- shown schematically (Fig. 1). For our investigations, agation away from the center of the breakdown region. optical breakdown was studied in a simple model sys- The transmitted HeNe probe beam was focussed tem consisting of physiological (0.9%) saline solution. onto a small area, high speed, avalanche photodiode The saline was contained in a 10 X 20-mm quartz flu- detector (Telefunken BPW28; Sommerville, NJ). A orimeter cell. In order to prevent the accumulation of narrowband dielectric interference filter centered at gas bubbles or other impurities produced by the optical 632.8 nm was used to block luminescence and scattered breakdown process, the solution was circulated and light from the optical breakdown. The small area of filtered with a 3-nm filter. the detector combined with this detection geometry A frequency doubled Q-switched Nd:YAG laser makes the probe laser sensitive not only to changes in (Quanta-Ray DCR1; Mountain View, CA) was used transmission arising from induced absorption or scat- as the pump laser to produce optical breakdown in the tering, but also to changes in the index of refraction sample. The Nd:YAG laser generates ~10 ns long which will influence the imaging of the probe beam pulses at 532 nm with a repetition rate of 10 Hz. The onto the detector. The temporal behavior of the laser- second harmonic of the Nd.YAG laser was used in induced plasma luminescence was also measured by order to facilitate accurate measurements of beam spa- turning off the probe laser and using a low pass, red tial profile and focusing characteristics. The laser was filter in front of the detector to block the scattered pump adjusted to optimize temporal pulse shape and energy laser.

Downloaded from iovs.arvojournals.org on 09/26/2021 No. 12 ND:YAG LASER-INDUCED BREAKDOWN / Fujimoro er ol. 1773

The output of the photodiode was measured with a high speed oscilloscope (Tektronix 7104; Beaverton, (a) OR) with a 1 GHz bandwidth. Measurements which Nd=YAG PULSE did not require as high temporal resolution were per- formed using a high speed storage oscilloscope (Tek- 5 ns/div tronix 7834). The temporal resolution was ~4 ns as limited by the duration of the Nd:YAG laser pulses. Experimentally, the effective temporal measurement resolution was limited by fluctuations in the laser-in- duced breakdown processes which occurred on a shot (b) to shot basis rather than bandwidth restrictions in the electronic instrumentation. LUMINESCENCE 5 ns/div Results

A sharp threshold for the onset of optical breakdown is observed with increasing Nd:YAG pump laser in- tensity. For our experimental conditions, using purified (c) saline solution, threshold was ~-6 mJ per pulse. The breakdown threshold was dependent on the purity of HeNe PROBE the saline sample and was reduced in the presence of 5 ns/div impurities or gas bubbles in the solution. The break- down process was evidenced by a dramatic increase in the scattering of the incident Nd:YAG laser and an audible noise from acoustic wave generation. Plasma Fig. 2. High speed oscilloscope traces showing the transient behavior luminescence and gas bubble formation were also vi- of the laser-induced plasma. All of the traces have identical time sually observable. scales, (a) Incident Nd:YAG laser pulse, (b) Temporal behavior of Figure 2 shows the experimentally measured time- the plasma luminescence, (c) Decrease in the transmission of the HeNe probe laser showing the rapid onset of the plasma absorption. resolved plasma luminescence as well as pump probe Zero transmission is one division from the bottom. measurements of induced absorption and scattering following the initiation of optical breakdown. Data was obtained for an incident Nd:YAG pulse energy of 20 mJ per pulse. Figure 2a shows an oscilloscope trace of minescence. Transmission is reduced by >50% in less the incident Nd:YAG laser pulse shape. The pulse du- than ~3 ns. The risetime of the plasma luminescence ration is approximately 10 ns with a 4-ns temporal as well as the onset of the induced absorption is limited substructure. Figure 2b shows the temporal behavior in this experiment by the duration of the incident Nd: of the plasma luminescence. The time scale for Figures YAG pump pulse. 2a and 2b are identical, indicating that the plasma lu- The rapid decrease in probe transmission is probably minescence develops rapidly within ~4 ns after the the result of absorption and scattering from the laser incident pump pulse intensity reaches a level sufficient induced plasma and has been termed plasma shield- for its generation. The plasma luminescence does not ing.10"12'25"28 It is important to note, however, that the have an identical temporal structure to the incident plasma shielding should last only for several nano- pump pulse, but rather exhibits a smoothed profile seconds, a time comparable to the plasma lifetime. In which resembles the envelope of the pump pulse with contrast, Figure 2c shows that the probe transmission a ~10 ns exponential decay. If the plasma lumines- does not recover on this time scale. Additional exper- cence behavior is an indicator of the presence of the imental measurements indicate that the probe trans- plasma, then these results indicate that the plasma life- mission in fact remains reduced for almost 100 /is. This time is of the order of 10 ns. Figure 2c shows the tem- effect may be attributed to cavitation or gas bubble poral behavior of the transmitted HeNe probe laser on formation. The probe is scattered by a gas bubble which a time scale identical to the previous photographs. The develops at the breakdown site. Transmission occurs HeNe probe beam focal point was made coincident only after the bubble has collapsed or broken up and with the central portion of the breakdown region. The the saline solution has returned to its unperturbed state. decrease in the probe transmission begins within ~4 The details of these physical processes may be further ns after the peak of the pump pulse and plasma lu- elucidated by performing spatially resolved pump-

Downloaded from iovs.arvojournals.org on 09/26/2021 1774 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / December 1985 Vol. 26

On a microsecond time scale (Figure 3b), the probe (a) beam transmission is observed to become almost com- pletely zero for times between 8 us and ~90 fis after ACOUSTIC the initiation of the breakdown. This is the result of a TRANSIENT gas bubble which is generated at the breakdown site, expands radially outward, then collapses. When the 50 ns/div optical breakdown occurs, the gas-liquid phase front is generated and begins expansion. The probe beam transmission is unaffected until this phase front crosses the probe beam focal point at 8 us. For times between (b) 8 fxs and -~90 us the gas bubble radius is larger than 250 /tim and the gas-liquid phase front scatters the probe CAVITATION beam. The bubble then collapses, and probe transmis- 20 ^s/div sion is restored when the bubble radius becomes less than 250 nm. A systematic set of time-resolved measurements as a function of spatial separation between the pump and Fig. 3. Temporal behavior of HeNe probe laser transmission at a probe beams allows the detailed characterization of the point 250 fim away from the center of the breakdown region. The transient temporal and spatial properties of the plasma, breakdown commences at a time corresponding to the lefthand origin acoustic wave, and cavitation which accompany the of the oscilloscope traces, (a) The probe beam experiences a transient laser induced breakdown process. The physical dimen- change produced by an acoustic pressure wave arriving after —140 sions and propagation velocities associated with these ns. (b) the probe beam is scattered by a gas bubble which expands and collapses, crossing the probe beam path for times between 8 /is phenomena may be measured to construct a detailed and ~90 /*s. dynamical description. The behavior of the acoustic transient was investi- gated by measuring the arrival time of the acoustic probe measurements. The spatial properties of the transient at given distances from the breakdown origin. laser-induced breakdown may be characterized by Figures 4a and 4b show experimental measurements varying the separation between the probe beam focal of acoustic wave propagation obtained with Nd:YAG point and the center of the breakdown region. This pump laser pulse energies of 10 and 20 mJ. The acous- allows the investigation of effects which propagate away tic wave is generated at the breakdown site and prop- from the breakdown site as well as the measurement agates radially outward as a shock wave with an initial of physical dimensions and propagation velocities. hypersonic velocity. As the wave front expands it loses Figure 3 illustrates experimental measurements of energy and propagates at a constant velocity. For a probe transmission obtained with the beam focal point pump pulse energy of 10 mJ, the initial velocity of the a distance of 250 /itn from the center of the breakdown shock wave is —2100 m/s. After propagating approx- volume. Figure 3a shows the behavior of the probe imately 100 jum in ~-50 ns, the acoustic wave velocity transmission on a nanosecond time scale with the decreases to ~ 1700 m/s and remains constant there- breakdown occurring at a time corresponding to the after. This is in reasonable agreement with the velocity left hand origin of the oscilloscope trace. The probe of sound in water of ~ 1500 m/s. Increasing the pump remains unaffected for the first ~~140 ns after the laser pulse energy results in a more rapid initial shock breakdown, then experiences a sharp transient reduc- wave velocity as well as an increase in the size of the tion lasting ~20 ns. This may be attributed to a scat- nonlinear propagation or shock wave region. Mea- tering of the probe beam by a propagating acoustic surements of the plasma absorption performed at initial wave. The acoustic wave is generated by heating and/ times of 10-20 ns as a function of distance indicate or plasma expansion at the breakdown site, then prop- that the initial radius of the breakdown region is ~ 10 agates spherically outward. When the acoustic transient nm and ~-25 nm with 10 mJ and 20 mJ pulses re- crosses the probe beam focal point, the index of re- spectively. fraction changes associated with the compressional Experimental measurement of the cavitation or gas wave cause a lensing of the probe beam which affects bubble formation and collapse may be performed by its imaging onto the photodiode detector. Since the determining the onset and recovery times of the probe acoustic transient travels a distance of ~-250 ^m in beam scattering on a microsecond time scale. Figures ~~ 140 ns, this implies an average propagation velocity 5a and 5b show data obtained for pump pulse energies of ~ 1800 m/s. of 10 mJ and 20 mJ respectively. For a pulse energy

Downloaded from iovs.arvojournals.org on 09/26/2021 No. 12 ND:YAG LASER-INDUCED BREAKDOWN / Fujimoro er ol. 1775

of 10 mJ, the gas-liquid phase front expands radially CAVITATION outward with an initial velocity greater than 50 m/s. 20 mJ/ PULSE -T 600 The expansion velocity then slows rapidly after the first V > 50m/s 10 MS. The maximum bubble radius of ~500 /im is reached at ~60 /us and is followed by a collapse of the W 400 bubble with a velocity of ~ 11 m/s. The expansion o velocity, maximum size, and temporal duration of the 200 bubble increase with increasing pump pulse energy. V) The error bars on the data are the result of fluctu- Q ations in the cavitation process which occur with suc- cessive laser shots and are not due to electronic or in- 20 40 60 80 100 120 strumental limitations in time resolution. The dynam- TIME (/xS) ics of the bubble collapse are relatively unstable and precise measurements are difficult; however, the ex- pansion phase of the cavitation may be well charac- lOmJ/PULSE V > 50 m/s terized. The use of unfiltered saline solution which had

ACOUSTIC TRANSIENT

250

200 .y I > yS v ~ 1700 m/s 20 40 60 80 100 120 -' 150 TIME (/xS) LLJ o z Fig. 5. Spatial and temporal behavior of the cavitation or gas bubble < 100 formation and collapse for incident Nd:YAG pulse energies of 10 to and 20 mJ. Plots are obtained by taking systematic measurements

Q of the type shown in Fig. 3b at several distances from the breakdown 50 V ~ 2400m/s origin. The error bars indicate pulse to pulse fluctuations in the cav- 20mJ/PULSE • itation process.

i i 0 20 40 60 80 100 120 140 TIME (ns) a higher concentration of small particle impurities and gas bubbles resulted in a decrease in the threshold for 250 r y laser-induced breakdown as well as a decrease in the shot to shot fluctuations in the dynamics of the break- y down process. The dimensions and propagation ve- %y V~ 1700 m/s s" locities associated with the acoustic transient cavitation *"' 150 - » also increase for unfiltered or impure solutions. LJJ o z Discussion < 100 in These results define some of the clinically relevant o .y* v ~ 2lOOm/s damage mechanisms involved in the photodisruption 50h lOmJ/PULSE of ocular structures with short pulsed Nd:YAG lasers. The dynamic behavior of the processes which accom- 1 i i 0 20 40 60 80 100 120 140 pany laser-induced optical breakdown may be sum- TIME (ns) marized as follows: When the laser pulse interacts Fig. 4. Spatial and temporal behavior of the acoustic transient for with the material, it first generates a laser-induced incident Nd:YAG pulse energies of 10 and 20 mJ. Plots are obtained plasma via local heating and/or nonlinear absorp-

by measuring the propagation distance of the acoustic transient as a tion 10-12,16,20,21,24,25,28 This plasma causes an increased function of time. The arrows indicate the initial radius of the break- down region. The acoustic transient propagates with an initial hy- absorption and scattering of the incident laser energy personic velocity, then loses energy after expansion and travels at which gives riset o the plasma shielding effect. The dy- sound velocity. namics of the plasma generation occur on a subnano-

Downloaded from iovs.arvojournals.org on 09/26/2021 1776 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / December 1985 Vol. 26

second time scale and are too rapid to be resolved in associated with laser-induced breakdown not only is this experiment. The dimensions of the plasma may important to ophthalmic applications but also is rele- be determined by measuring the absorption region. As vant in the context of recent studies which indicate the expected, the size of the plasma increases with increas- presence of acoustic and localized vaporization mech- ing laser pulse energy or the presence of impurities in anisms for pulsed laser interaction with pigmented the solution. Measurements of the temporal behavior structures, cells, or organelles.29"31 We believe that time- of the plasma luminescence indicate that the plasma resolved spectroscopic techniques will become in- decays by radiative emission, spatial expansion, and creasingly important in the experimental investigation heating of the surrounding solution on a time scale of of laser tissue interactions in general and will prove several nanoseconds. This expansion of the plasma especially useful with the development of laser medical and/or the heating of the solution generates a pressure techniques using pulsed laser sources. wave which propagates radially away from the break- down region. The wave exhibits an initial nonlinear Key words: Nd:YAG, laser-induced breakdown, plasma behavior and propagates with a hypersonic velocity, shielding, photo disruption then expands and loses energy to enter the linear re- gime. Thereafter it propagates with the velocity of sound. The heating of the solution also produces a gas- References liquid phase change and gives rise to cavitation or gas 1. Van-der Zypen E, Bebie H, and Fankhauser F: Morphologic bubble formation. The gas bubble expands rapidly for studies about the efficiency of laser beams upon the structures several microseconds, then slows to reach its maximum of the angle of the anterior chamber. Int Ophthalmol Clin 1, 2: size, and finally collapses. 109, 1979. 2. Van-der Zypen E, Fankhauser F, Bebie H, and Marshall J: The hypersonic acoustic transient velocity as well as Changes in the ultrastructure of the iris after irradiation with the velocity of the gas bubble and its dimensions in- intense light. Adv Ophthalmol 39:59, 1979. creases with increasing laser pulse energy. This obser- 3. Fankhauser F, Roussel P, Steffan J, Van-der Zypen E, and vation explains, in part, the increasing severity of tissue Chernkova A: Clinical studies on the efficiency of a high power alterations observed in the eye with photodisruption laser radiation upon some structures of the anterior segment of 8 9 the eye. Int Ophthalmol Clin 3, 3:129, 1981. at higher pulse energies. ' The presence of impurities 4. Fankhauser F, Loertscher HP, and Van-der Zypen E: Clinical in the target solution causes an increase in the shot to studies on high and lower power laser irradiation upon some shot stability of the laser induced breakdown process structures of the anterior and posterior segments of the eye. Int as well as a decrease in the breakdown threshold and Ophthalmol Clin 5:15, 1982. an enhancement of the hypersonic acoustic transient 5. Fankhauser F and Van-der Zypen E: Future of the laser in oph- thalmology. Trans Ophthalmol Soc UK 102:159, 1982. and gas bubble formation. This is in agreement with 6. Aron-Ross D: Use of a pulsed neodymium-YAG laser for anterior previous investigations which hypothesize that the capsulotomy before extracapsular extraction. J Am Intraocul breakdown process is initiated by the heating of im- Implant Soc 7:332, 1981. purities in the liquid.16-20-21-24 7. Aron-Rose D, Aron J, Griesemann J, and Thyzel R: Use of the In summary, this research has demonstrated the ap- neodymium-YAG laser to open the posterior capsule after lens implant surgery. J Am Intraocul Implant Soc 6:352, 1980. plication of time-resolved spectroscopic techniques for 8. Puliafito CA and Steinert RF: of the lens: experi- the investigation of Nd:YAG laser-induced optical mental studies. Ophthalmology 90:1007, 1983. breakdown in a saline model. The use of a spatially 9. Puliafito CA, Wasson PJ, Steinert RF, and Gragoudas ES: Neo- resolved pump-probe technique has been shown to al- dymium YAG laser surgery on experimental vitreous mem- low the measurement of both spatial and temporal be- branes. Arch Ophthalmol 102:843, 1984. 10. Taboada J: Interaction of short laser pulses with ocular tissues. havior so that a description of the physical dimensions, In YAG Laser Ophthalmic Microsurgery, Trokel S, editor. Nor- propagation velocities, and dynamical behavior of the walk, CT, Appleton-Century-Crofts, 1983, pp. 15-38. laser-induced breakdown may be obtained. Thus, 11. Loertscher H: Laser-induced breakdown for ophthalmic appli- plasma formation and shielding, acoustic transients, cations. In YAG Laser Ophthalmic Microsurgery, Trokel S, ed- itor. Norwalk, CT. Appleton-Century-Crofts, 1983, pp. 39-66. and cavitation phenomena have been measured within 12. Steinert RF and Puliafito CA: The Nd-YAG Laser in Ophthal- a simple and consistent experimental context. This ca- mology: Principles and Clinical Applications of Photodisruption. pability is useful for comparative or quantitative inves- Philadelphia, PA, WB Saunders, 1985. tigations of laser-induced breakdown. For example, 13. Carome EF, Clark NA, and Moeller CE: Generation of acoustic further studies may be able to resolve questions re- signals in liquids by ruby laser-induced thermal stress transients. garding differences in the physical processes and dam- Appl Phys Lett 4:95, 1964. 14. Carome EF, Moeller CE, and Clark NA: Intense ruby-laser-in- age mechanisms associated with Q-switched versus duced acoustic impulses in liquids. J Acoust Soc Am 40:1462, modelocked Nd:YAG ophthalmic laser surgery. An 1966. increased understanding of the transient phenomena 15. Carome EF, Carreira EM, and Prochaska CJ: Photographic

Downloaded from iovs.arvojournals.org on 09/26/2021 No. 12 ND:YAG LASER-INDUCED BREAKDOWN / Fujimoro er ol. 1777

studies of laser-induced pressure impulses in liquids. Appl Phys 25. Mainster MA, Sliney DH, Belcher CD, and Buzney SM: Laser Lett 11:64, 1967. photodistrupters: damage mechanisms, instrument design and 16. Bell CE and Landt JA: Laser-induced high pressure shock waves safety. Ophthalmology 9:973, 1983. in water. Appl Phys Lett 10:46, 1967. 26. Steinert RF, Puliafito CA, and Trokel S: Plasma formation and 17. Barnes PA and Rieckhoff KE: Laser-induced underwater sparks. shielding by three ophthalmic Nd-YAG lasers. Am J Ophthalmol Appl Phys Lett 13:282, 1968. 96:427, 1983. 18. Hu C-L: Spherical model of an acoustic wave generated by rapid 27. Steinert RF, Puliafito CA, and Kittrell C: Plasma shielding by laser heating in a liquid. J Acoust Soc Am 46:728, 1969. Q-switched and modelocked Nd-YAG lasers. Ophthalmology 19. Cleary SF and Hamrick PE: Laser-induced acoustic transients 90:1003, 1983. in the mammalian eye. J Acoust Soc Am 46:1037, 1969. 28. Puliafito CA and Steinert RF: Short-pulsed Nd:YAG laser mi- 20. Felix MP and Ellis AT: Laser-induced liquid breakdown—a step- crosurgery in the eye: biophysical considerations. IEEE J Quan- by-step account. Appl Phys Lett 19:484, 1971. tum Electronics QE-20:1442, 1984. 21. Lauterborn W: High-speed photography of laser-induced break- down in liquids. Appl Phys Lett 21:27, 1972. 29. Parrish JA, Anderson RR, Harrist T, Paul B, and Murphy GF: 22. Orlov RY, Skidan IB, and Telegin LS: Investigations of break- Selective thermal effects with pulsed irradiation from lasers: for down produced in dielectrics by ultrashort laser pulses. Soviet organ to organelle. J Invest Dermatol 8O(Suppl):75s, 1983. Physics JETP 34:418, 1972. 30. Paul BS, Anderson RR, Jarve J, and Parrish JA: The effect of 23. Cleary SF: Laser pulses and the generation of acoustic transients temperature and other factors on selective microvascular damage in biological material. In Laser Applications in Medicine and caused by pulsed . J Invest Dermatol 81:333, 1983. Biology, Worlbarsht ML, editor. Vol. 3. New York, Plenum Press, 31. Murphy GF, Shepard RS, Paul BS, Menkes A, Anderson RR, 1977, pp. 175-219. and Parrish JA: Organelle-specific injury to melanin-containing 24. Lauterborn W and Ebeling KJ: High-speed holography of laser- cells in human skin by pulsed laser irradiation. Lab Invest 49: induced breakdown in liquids. Appl Phys Lett 31:663, 1977. 680, 1983.

Downloaded from iovs.arvojournals.org on 09/26/2021