Passively Q-Switched Microchip Lasers and Applications John J. ZAYHOWSKI Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, MA 02420-9108, USA (Received October 8, 1998) Passively Q-switched Nd:YAG microchip lasers are robust, compact, economical, all-solid-state sources of coherent, subnanosecond, multikilowatt pulses at high repetition rates. These diminutive, quasi-monolithic devices produce pulses as short as 218 ps, pulse energies up to 250ƒÊJ,1.1, and peak powers up to 565 kw. The high output intensities enable the construction of extremely compact nonlinear optical systems. The short pulses are useful for high-precision ranging using time-of-flight techniques. The short pulse durations and ideal mode properties can also be used to advantage in the characterization of materials. When focused, the output intensities are sufficient to photoablate materials, with applications in laser-induced breakdown spectroscopy and micromachining. The ultraviolet harmonics of the microchip laser have been used to perform fluorescence spectroscopy for a variety of applications. Systems based on the passively Q-switched microchip lasers, like the lasers themselves, are small, efficient, robust, and potentially low cost, making them ideally suited for field use. Key Words: Solid-state laser, Diode-pumped laser, Q-switched laser, Nonlinear optics 1. Introduction Many applications of lasers require subnanosecond optical pulses with peak powers of several kilowatts and pulse energies of several microjoules. The most common method of producing subnanosecond pulses is to modelock a laser, generating a periodic train of short pulses with an interpulse period equal to the round-trip time of the laser cavity, typically 10 ns. Because of the large number of pulses produced each second, even lasers with high average powers (10 W or greater) do not produce much energy per pulse. Energetic pulses can be produced by Q switching. However, the size of conventional Q-switched lasers, along with their physics, precludes producing subnanosecond pulses.1) Extremely short, high-energy pulses can be obtained from Q- switched modelocked lasers or amplified modelocked lasers. Both of these approaches require complicated systems, typically several feet long and consuming several kilowatts of electrical power, and are therefore expensive. The short cavity lengths of Q-switched microchip lasers allow them to produce pulses with a duration comparable to that obtained with modelocked systems. At the same time, they take full advantage of the gain medium's ability to store energy. Actively Q-switched microchip lasers, pumped with a 0.5-W diode laser, have produced pulses as short as 115 ps with peak powers of tens of kilowatts and pulse energies of several microjoules.2) For proper Q switching, these lasers require high-speed, highvoltage electronics. A passively Q-switched microchip laser does not require any switching electronics,3) thereby reducing system size and complexity, and improving power efficiency. Pumped with a 1.2-W diode laser, passively Q-switched microchip lasers produce pulses as short as 218 ps with peak powers in excess of 25 kw at pulse repetition rates greater than 10 khz. More recently, passively Q-switched microchip lasers have been pumped with up to 15 W from a diode-laser array, and produced 380-ps pulses with peak powers in excess of 560 kw at pulse repetition rates up to 1 khz. All of these devices oscillate in a single, transform-limited longitudinal mode, and produce a diffraction-limited, linearly polarized, circularly symmetric Gaussian beam. The optical head of a passively Q-switched microchip laser can be mass produced in a robust package occupying a volume of between 1 cm3, for a 1.2-W-pumped infrared device, and 40 cm3, for a 15-W-pumped ultraviolet device. The only input to the optical head is a multimode optical fiber carrying the diodelaser pump radiation. The combination of short pulse duration, high peak power, and good beam quality, coupled with their simplicity, small size, and low cost, makes passively Q-switched microchip lasers attractive sources for a wide range of applications. This paper reviews the passively Q-switched microchip lasers developed at MIT Lincoln Laboratory, and discusses several applications for these enabling devices. 2. Concept The principle behind the operation of a passively Q-switched laser is that an intracavity saturable absorber prevents the onset of lasing until the average inversion density within the cavity reaches a critical threshold value. The onset of lasing, at that point, produces a high intracavity optical field that saturates the saturable component of the optical loss, increasing the cavity Q and resulting in a Q-switched output pulse. Increasing the pump power above threshold changes the pulse repetition rate, but leaves the rest of the pulse parameters unchanged, and the pulse amplitude and pulse width are extremely stable. In their simplest embodiment, the passively Q-switched mi- Vol.26, No.12 Passively Q-Switched Microchip Lasers and Applications
crochip lasers developed at MIT Lincoln Laboratory are constructed by diffusion bonding a thin, flat wafer of Nd3+:YAG gain medium to a similar wafer of Cr4+:YAG saturable absorber.3) More generally, flat wafers of undoped YAG may be bonded to both ends of the Nd:YAG/Cr4+:YAG structure, as shown in Fig.1. The composite structure is polished flat and parallel on the two faces normal to the optic axis. The pump-side face of the gain medium is coated dielectrically to transmit the pump light and to be highly reflecting at the oscillating wavelength. The output face is coated to be partially reflecting at the lasing wavelength and provides the optical output from the device. The infrared laser is completed by dicing the wafer into small squares, typically 1 to 2 mm on a side. The YAG cavity is then mounted on an appropriate heatsink and pumped with the output of a fibercoupled diode laser or diode-laser array. The simplicity of the passively Q-switched microchip laser and its small amount of material give it potential for inexpensive mass production; nearly monolithic construction results in robust devices. The fabrication technique described above is used in all of the MIT Lincoln Laboratory devices discussed below. Alternative approaches include the use of a single material that acts as both the gain medium and the saturable absorber,4'5) the growth of one material on the other,6) and the use of semiconductor saturable absorbers.7-8) 3. Infrared Passively Q-Switched Microchip Lasers The low-power passively Q-switched microchip lasers (LPMCL-1, LPMCL-2, LPMCL-3, LPMCL-4, and LPMCL-5) were designed to be pumped with a 1.2-W fiber-coupled semiconductor diode laser. These devices have the simplest possible cavity configuration, containing only the gain medium and the saturable absorber. A generalized schematic of these devices is shown in Fig.1, with the corresponding device parameters listed in Table 1. A photograph of a device is shown in Fig.2. The faces of the composite material system are polished flat, to a parallelism of better than 10 grad. The input coating, deposited on the Nd:YAG, is antireflecting at the pump wavelength (808 nm) and highly reflecting ( > 99.9 %) at the oscillating wavelength (1.064 lam). The output coating on LPMCL-1 has a reflectivity of 80 % at the oscillating wavelength; the output coating on the rest of the low-power 1.064-1,im devices has a reflectivity of 85 %. Cavity lengths vary from 0.75 to 2 mm. By keeping the cavity as short as possible, the pulse duration is minimized and the peak power is maximized. 1,3) LPMCL-1, LPMCL-2, and LPMCL-3 are pumped by bonding the microchip laser directly to the ferrule of an SMAconnectorized multimode optical fiber (fiber diameter = 100 NA = 0.2), as shown in Fig.2 - the pump diode is connected to the opposite end of the fiber. No intermediate optics are used between the output facet of the fiber and the microchip laser; the metal ferrule on the fiber serves as the heatsink. The pump diodes are typically operated CW. The shortest pulses we have obtained come from LPMCL-1, and have a 218-ps full width at half-maximum and an energy of 4 IA. at repetition rates below 5 khz. Devices using semiconductor saturable absorbers have demonstrated shorter pulses, with durations as short as 56 ps.8,9) These devices also have much less pulse energy, with a reported value of 62 nj. LPMCL-2 has a slightly longer output pulse Fig.1 Generalized schematic of passively Q-switched microchip lasers. Fig.2 Low-power passively Q-switched microchip laser bonded to the end of a multimode pump fiber. Table 1 Design parameters for passively Q-switched microchip lasers. The Review of Laser Engineering December 1998
than LPMCL-1 (280 ps) and a slightly higher pulse energy (4.7 J), and operates reliably at repetition rates in excess of 8 khz. u LPMCL-3, LPMCL-4, and LPMCL-5 were designed to maximize the pulse energy that could be obtained with a 1.2-W pump. LPMCL-3 represents the highest pulse energy that could be reliably obtained by bonding the microchip laser directly to the output of the pump fiber. In devices LPMCL-4 and LPMCL-5, a GRIN lens was used to image the output of the fiber into the microchip laser, with a typical magnification of - 1. In this way, the waist of the highly divergent pump light is located in the center of the gain medium, and more of the light is absorbed within the oscillating mode volume, thereby reducing the power required to reach threshold. LPMCL-5 produced 14ƒÊJ in 400-ps pulses, but had some problems with optical damage to the output coupler - a result of the high intensity of the internal optical field and the thermal stress induced by the pump. LPMCL- 4 has proven to be highly reliable, and operates at repetition rates in excess of 12 khz. The performance of the low-power devices is summarized in Table 2. All of these lasers oscillate in a single longitudinal mode with transform-limited spectral performance, in the fundamental transverse mode with diffraction-limited divergence, and in a linear polarization. The pulse-to-pulse amplitude fluctuations of these devices have been measured to be < 0.05 %. Pulse-to-pulse timing jitter tracks fluctuations in the output of the pump diode. In addition to devices operating at 1.064 ƒêm, we have demonstrated efficient operation of low-power lasers at 946 nm using Cr4+:YAG as the passive Q switch.10) Semiconductor saturable absorbers can be engineered for operation at many different wavelengths, and low-power passively Q-switched lasers using semiconductor saturable absorbers have been demonstrated at 1.3 and 1.5 ƒêm.9) The mid-power passively Q-switched microchip lasers (MPMCL-1, MPMCL-2, and MPMCL-3) were designed to be pumped with a 3-W fiber-coupled diode-laser array. They are constructed of Nd:YAG, Cr4+:YAG, and undoped YAG, diffusion bonded to each other in that order. The design parameters are listed in Table 1. The output coating on all of the mid-power devices has a reflectivity of 60 % at 1.064 ƒêm. The total cavity length for MPMCL-1, MPMCL-2, and MPMCL-3 is 6, 12, and 24 mm, respectively. To pump the mid-power lasers, the output of a 3-W fibercoupled diode-laser array (fiber diameter = 100 ƒêm, NA = 0.2) is imaging in the Nd:YAG with a pair of aspheric lenses, with a typical magnification of ` 1. The diode lasers are typically operated in pulsed mode, with each pulse from the diode lasers generating 1 Q-switched output pulse from the microchip laser. With 3 W of pump power, MPMCL-1 produces 30 ƒêj in a 700-ps pulse at repetition rates up to 16 khz; MPMCL-2 produces 40 tj in a 1200-ps pulse at repetition rates up to 3 khz. MPMCL- 3 requires a pump power that slightly exceeds the rating of the 3-W diode array. With 5 W of pump power, it produces 65 ƒêj in a 2200-ps pulse at repetition rates up to 2.5 khz. The output properties of the mid-power devices are summarized in Table 2. All of these devices operate in a linearly polarized, single longitudinal mode, with diffraction-limited output. The high-power microchip lasers (HPMCL-1, HPMCL-2, HPMCL-3, and HPMCL-4) were designed to be pumped with a 10-W fiber-coupled diode-laser array. They comprise a 4-crystal sandwich of undoped YAG, Nd:YAG, Cr4+:YAG, and undoped YAG, diffusion bonded to each other in that order, as shown in Fig.1. Because of the high optical intensities (up to 14 GW/cm2) and fluences (up to 4.4 J/cm2) present in these devices, special care is taken in the surface preparation and coating of the undoped-yag endcaps. The output coatings of HPMCL-1 and HPMCL-2 have a reflectivity of 40 % at 1.064 pm. The reflectivity of the output coatings for HPMCL-3 and HPMCL-4 is 26 %. As can be seen from Table 1, the cavity lengths of the high-power devices vary between 6.5 and 12 mm. To pump the high-power lasers, the output of a 10-W fibercoupled diode-laser array (fiber diameter = 250 ƒêm, NA = 0.2) is imaged in the Nd:YAG with a pair of aspheric lenses, with a typical magnification of ` 0.5. HPMCL-1 produces 130 ƒêj in a 390-ps pulse and operates at pulse repetition rates up to 5.5 khz. HPMCL-2 produces 225 ƒêj in a 700-ps pulse at repetition rates up to 2 khz. HPMCL-3 and HPMCL-4 produce 200 ƒêj in a 310-ps pulse, and 250ƒÊJ in a 380-ps pulse, respectively. These two devices required pump powers of 11 and 15 W, respectively (both in excess of the nominal rating for the diode-laser array), and have only been operated at repetition rates up to 1 khz. Both have peak powers in excess of 550 kw, with peak output intensities, at the output coupler, of ` 10 GW/cm2. Even at these high intensities, we have not had problems with damage to the coatings. The output properties of all of the high-power devices are included in Table 2. These devices also operate in a linearly polarized, single longitudinal mode, with diffraction-limited output. Table 2 Output properties of passively Q-switched microchip lasers. Vol.26, No.12 Passively Q-Switched Microchip Lasers and Applications 843
Table 3 Average power generated by LPMCL-4, MPMCL-1, and HPMCL-1 at fundamental and harmonics. 4. Nonlinear Frequency Conversion The output of several of the microchip lasers described above has been harmonically converted into the visible and ultraviolet. The high peak intensities of the passively Q-switched microchip lasers allow for efficient nonlinear frequency generation in very simple and robust devices.3,10-16) Frequency doubling is performed using type-ii phase matching in 5-mm-long pieces of KTP. The KTP is placed in the output of the microchip laser, without any intermediate optics. For the high-power devices, the KTP is placed more than 1.5 cm from the output facet of the laser so that beam divergence reduces the peak intensities below the damage threshold of the KTP; for the low- and mid-power devices, it is placed within 250 ƒêm of the output facet of the laser. Typical second-harmonic conversion efficiencies for all of the devices are between 50 % and 60 %. Third- and fourth-harmonic generation is performed in properly oriented, 5-mm-long pieces of BBO placed adjacent to the output of the KTP. A second piece of BBO has been used to generate fifth-harmonic radiation (first plus fourth) in one of the low-power devices. Harmonic systems built around low-power device LPMCL-4 generate up to 1.5 ƒêj/pulse at 355 nm, 0.9 p/pulse at 266 nm, and 50 nj/pulse at 213 nm, at repetition rates up to 12 khz. Slightly higher UV energies, at lower repetition rates, are obtained with systems built around LPMCL-5. A nonlinear system built around mid-power device MPMCL-1 generates 2.5 gj/pulse of 266-nm output at pulse repetition rates up to 15 khz. Nonlinear systems built around high-power device HPMCL-1 produce 19 ppulse of 355-nm output and 12 ƒêj / pulse of 266-nm output. The time-averaged output powers of these systems are summarized in Table 3. The optical head of the low-power UV systems is typically packaged in a 1-cm-diameter ~ 3-cm-long stainless-steel can,10-13) as shown in Fig.3; the optical head of the mid- and high-power systems is packaged in a 2.5-cm-diameter ~ 8-cm-long can, as shown in Fig.4. In all cases, the only input to the head is the multimode fiber carrying the pump radiation. Cascaded stimulated Raman scattering has been used to generate broadband continua in optical fibers.17) A continuum extending from 532 to 950 nm was obtained in a 100-m length of single-mode silica fiber pumped with 2 ƒêj of 532-nm radiation from a low-power frequency-doubled passively Q-switched mi- Fig.4 Optical head of a high-power UV microchip laser system packaged in a 2.5-cm-diameter ~ 8cm-long (3-inch-long) stainless-steel can. crochip laser.10,16) The long-wavelength end of the continuum is limited by the guiding properties of the fiber. Similar spectra can be generated in the mid-ir or near-uv-to-blue regions of the spectrum by starting with the fundamental or third harmonic of the microchip laser output, respectively. The mid- and high-power devices have been used to pump optical parametric amplifiers (OPAs) and optical parametric oscillators (OPOs). The unfocused 1.064-ƒÊm output of a 12-mmlong mid-power passively Q-switched microchip laser (MPMCL- 2) has been used to drive single-pass periodically poled lithium niobate (PPLN) OPAs at wavelengths between 1.4 and 4.3 ƒêm, with a peak conversion efficiency of nearly 100 %.10,14,16) The high-power 1.064-ƒÊm devices (HPMCL-1 and HPMCL-2) have pumped singly resonant monolithic KTP OPOs oscillating at 1.6 m. The frequency-doubled output of a mid-power device ƒê (MPMCL-3) has been used to pump doubly resonant monolithic KTP OPOs at wavelengths between 820 and 1515 nm.10,15,16) Optical parametric devices afford microchip laser systems enormous wavelength flexibility. Conservative calculations indicate that the frequency-doubled output of mid- and high-power passively Q-switched microchip lasers should be able to pump doubly resonant KTP OPOs operating at wavelengths (including both signal and idler) between 650 and 3000 nm.10,15,16) Ongoing work in periodically poled materials 18-20) promises to extend this range and greatly reduce the threshold of the OPOs, making microchip-laser-pumped singly resonant devices possible over a broad range of wavelengths. The output of the OPOs can be harmonically converted, or summed with harmonics of the microchip laser, to obtain visible and UV radiation with wavelengths as short as 190 nm. The entire spectrum from the mid IR through the deep-uv is accessible to these diminutive optical systems. 5. Ranging Applications Time-of-flight optical ranging is one application for the microchip laser. The resolution of such a system is one half of the speed of light multiplied by the pulse width. A 200-ps optical Fig.3 Optical head of a low-power UV passively Q- switched microchip laser system packaged in a 1- cm-diameter ~ 3-cm-long stainless-steel can. pulse can provide a range resolution (minimum separation between two resolvable objects) of 3 cm. When the shape of the optical pulse is repeatable, as is the case for the microchip laser, the accuracy of the system can be much better. At MIT Lincoln The Review of Laser Engineering December 1998
Laboratory, we have demonstrated a compact time-of-flight optical transceiver using a low-power frequency-doubled green microchip laser attenuated to the Class-II eye-safe level of 0.2 J per pulse. We were able to range to objects, including black ƒê felt, with a single-pulse range accuracy of 1 mm at distances up to 50 m. The system used a commercial 1-GHz detector and 5- cm-diameter collection optics. Coupled to a two-dimensional For compounds that do not fluoresce appreciably, such as chlorinated solvents, we have measured Raman spectra, using the time domain to distinguish the Raman lines from fluorescence due to interferents in the same spectral region. Work is under way to develop a laser-based NOx detector for the detection of energetic contaminants such as the explosives TNT, RDX, and HMX. scanning system, the high repetition rate of the laser makes it possible to obtain a high-resolution, three-dimensional image in 7. Detection of Biological Substances minutes. This system was developed in collaboration with Cyra Technologies, Inc. Cyra offers a commercial, tripod-mounted, battery-operated version of the system, CryaxTM, with software that can quickly convert the captured images to 3D CAD models.21) Applications for such a system include automated production, civil engineering, construction, and architecture. The system just described uses one of the least powerful of the microchip lasers discussed here, LPMCL-2, and attenuates its frequency-doubled output. Higher-power devices have the capability of ranging at much greater distances. The unamplified output of a high-power device, HPMCL-4, is capable of performing earth-to-satellite ranging with centimeter accuracy.22) The output of microchip-laser-pumped OPAs and OPOs can be We have also used UV laser-induced fluorescence spectroscopy for the detection of airborne biological particles. In this application, the spectral characteristics of the fluorescence of tryptophan, NADH, and flavins can be used to distinguish biological particles from nonbiological particles, and even to do some classification of biological particles. In recent field tests, our battery-operated, portable bioaerosol fluorescence sensor, built around a low-power microchip laser (LPMCL-4), proved to be effective in the detection of biological-warfare simulants. Other potential applications of this technology include monitoring the air in hospitals and public buildings to help control the spread of airborne communicable diseases. used to perform ranging at eye-safe wavelengths. 8. Laser-Induced Breakdown Spectroscopy 6. Remote Environmental Monitoring Passively Q-switched microchip lasers can be used to per- Laser-based techniques are highly sensitive methods for determining concentrations of chemical and biological species, including pollutants. For many applications, the optimal measurement wavelengths lie in the UV. In recent years, remote detection has been performed with UV light delivered to the remote area with an optical fiber. Unfortunately, optical fibers transmit UV poorly. Thus, powerful lasers are required to provide sufficient energy at the fiber's distal end to ensure adequate detection sensitivity. In addition, sensitivity is critically dependent on the fiber length. These limitations can be overcome by using a multimode fiber to deliver easily transmitted near-ir diodelaser pump radiation to a remote head containing a frequencyconverted passively Q-switched microchip laser.12,23) We have constructed a sensor head that is 2.5 cm in diameter by approximately 7 cm long for use in a cone penetrometer24-27) to characterize subsurface contamination at depths up to 50 m. The sensor head contains a frequency-quadrupled low-power passively Q-switched microchip laser (LPMCL-3) and collection optics. The laser output is filtered to remove the IR and visible light before the UV light is focused outside through a sapphire window. Fluorescence from material contacting the window is collected and focused into a 500-ƒÊm-core return fiber for spectral analysis. The short duration of the excitation pulse facilitates accurate measurements of the decay times of even the short-lived benzene, toluene, ethylbenzene, and xylene (BTEX) compounds (decay times from 2 to 60 ns). Measurements of fluorescence decay times offer greater chemical selectivity than that of spectra alone.28) form laser-induced breakdown spectroscopy (LIBS). Although the per pulse energy is relatively low compared to lasers conventionally used in this application, the diffraction-limited beam can be focused to a spot size as small as 1 ii,m in diameter. Even the low-power microchip lasers can be focused to intensities in excess of 1 TW/cm2. This is sufficient to break down metals and many other solids.12) In the resulting plasma, there are highly excited ions, atoms, and molecules - each emits a unique spectrum as it recombines. By examining the recombination spectra, it is possible to determine the composition of the material. We have demonstrated the detection of various metals in soil using a low-power microchip laser (LPMCL-3) and a compact diode-array-based spectrometer.29) Because the optical pulse length is short and the resulting plasma volume is small, the plasma continuum radiation decays rapidly (in -15 ns). This has allowed us to measure sensitivities of up to 100 ppm without any of the standard temporal or spatial gating that conventional LIBS systems employ.30-32) Potential applications include the identification of heavy-metal contaminants such as Pb, Hg, Cd, Cr, and Zn.33) As the power of the microchip laser increases, so too do the sensitivity of the resulting LIBS system and the variety of materials that can be examined. The mid-power devices easily break down transparent media, including glasses and water; the focused output of the highest-power devices (HPMCL-3 and HPMCL-4) is sufficient to break down clean air, with potential applications in the monitoring of effluents and closed-loop process control. This laser-probe technology has recently been field tested.29) 9. Micromachining and Microsurgery By examining the spectral and temporal fluorescence characteristics as the probe was pushed into the ground, we identified BTEX compounds as well as heavier aromatic hydrocarbons that resulted from contamination due to aviation and heating fuels. These tests demonstrated that the microchip-laser-based probe offers the potential for in situ, real-time characterization of soils and groundwater in a robust, compact, inexpensive package. It is apparent from the above discussion that the Q-switched output of a microchip laser can photoablate most materials, including metals, semiconductors, glasses, and biological tissues. We have used a low-power microchip laser (LPMCL-3) at both 1.064 ƒêm and 532 nm to cut clean 5-ƒÊm-wide lines in the metal- Vol.26, No.12 Passively Q-Switched Microchip Lasers and Applications
lization on semiconductor wafers and to drill holes through the substrate. Higher-power devices have been used to bore holes in glass, scribe alumina, etc. Applications in microsurgery are being investigated. 10. Other Applications Passively Q-switched microchip lasers are attractive devices for a wide range of applications, beyond those discussed in detail above. Active Impulse Systems has developed the InSite 300TM, an instrument designed to measure the mechanical and thermal properties of thin films, around a low-power passively Q-switched microchip laser.34) Using the technique of impulsestimulated thermal scattering, the InSite 300 can make nondestructive measurements of thin-film thickness to an accuracy of 5 nm, with a transverse resolution of 10 Đm. It can also measure the anisotropic elastic moduli and thermal diffusivity, as well as determine whether or not there is delamination of the film, all at about 1000 measurements per second. In addition to semiconductor manufacturing, potential applications of this technology include checking for defects in painted or laminated surfaces, and monitoring the curing of epoxies and resins. SPARTA, Inc. uses the green and UV output of a low-power microchip laser in their second-generation nanaligntm interferometer system.35) This system improves the ability to position the stage used to hold semiconductor wafers during the lithography process. The broad spectrum generated by microchip-laser-pumped cascaded stimulated Raman scattering in fibers has applications in absorption, reflection, and excitation spectroscopy; and active 3-dimensional hyperspectral imaging. The highpower 355-nm UV devices can be used for matrix-assisted laser desorption and ionization (MALDI) and UV stereolithography. The list goes on, and new applications continue to emerge as this technology becomes more readily available. 11. Conclusions Passively Q-switched microchip lasers are a family of highperformance devices, with capabilities that exceed those of more conventional laser technology. They are an enabling technology, with the proven potential to take pulsed laser applications from the laboratory to the field and make new applications commercially viable. Microchip-laser-based systems, like the lasers themselves, are small, efficient, robust, and potentially low cost. This technology is still in its infancy; low-power microchip lasers are just now becoming commercially available.36,37) Further work can be expected to lead to shorter pulses, higher peak power, increased pulse energy, and new wavelengths of operation. With advances in the technology, and increased availability, applications of passively Q-switched microchip lasers will continue to expand. Acknowledgements This work was sponsored by the U. S. Department of the Air Force under Air Force Contract #F19628-95C-0002. References 1) J. J. Zayhowski and P. L. Kelley: IEEE J. Quantum Electron. 27 (1991) 2220; 29 (1993) 1239. 2) J. J. Zayhowski and C. Dill III: Opt. Lett. 20 (1995) 716. 3) J. J. Zayhowski and C. Dill III: Opt. Lett. 19 (1994) 1427. 4) S. Zhou, K. K. Lee, Y. C. Chen, and S. Li: Opt. Lett. 18 (1993) 511. 5) P. Wang, S.-H. Zhou, K. K. Lee, and Y. C. Chen: Opt. Commun. 114 (1995) 439. 6) L. Fulbert, J. Marty, B. Ferrand, and E. Molva: Conf. Lasers Electro- Optics Tech. Dig. 15 (1995) 176. 7) B. Braun, F. X. Kartner, U. Keller, J.-P. Meyn, and G. Huber: Opt. Lett. 21 (1996) 405. 8) B. Braun, F. X. Kartner, G. Zhang, M. Moser, and U. Keller: Opt. Lett. 22 (1997) 381. 9) R. Fluck, B. Braun, U. Keller, E. Gini, and H. Melchior: Conf. Lasers Electro-Optics Tech. Dig. 11 (1997) 355. 10) J. J. Zayhowski: Conf. Lasers Electro-Optics Tech. Dig. 11 (1997) 463. 11) J. J. Zayhowski, J. Ochoa, and C. Dill III: Conf. Lasers Electro- Optics Tech. Dig. 15 (1995) 139. 12) J. J. Zayhowski: Laser Focus World 32 (April 1996) 73. 13) J. J. Zayhowski: Opt. Lett. 21 (1996) 588; 1618. 14) J. J. Zayhowski: Opt. Lett. 22 (1997) 169. 15) J. J. Zayhowski: IEEE Photon. Technol. Lett. 9 (1997) 925. 16) J. J. Zayhowski: Opt. Mater. 11 (1998) 255. 17) G. P. Agrawal: Nonlinear Fiber Optics, 2nd ed. (Academic Press, San Diego, 1995), Ch. 8. 18) L. E. Myers, R. C. Eckardt, M. M. Feyer, R. L. Byer, W. R. Bosenberg, and J. W. Pierce: J. Opt. Soc. Am. B 12 (1995) 2102. 19) L. E. Myers and W. R. Bosenberg: Conf. Lasers Electro-Optics Tech. Dig. 11 (1997) 17. 20) H. Karlsson and F. Laurell: Conf. Lasers Electro-Optics Tech. Dig. 11 (1997) 15. 21) Cyra Technologies, Inc., Oakland, California, USA, www.cyra.com. 22) J. J. Degnan and J. F. McGarry: Proc. SPIE 3218 (1997) 63. 23) J. J. Zayhowski and B. Johnson: Laser Applications to Chemical, Biological and Environmental Analysis, 1996 Tech. Dig. Series 3 (OSA, 1996) 37. 24) S. H. Lieberman, G. A. Theriault, S. S. Cooper, P. G. Malone, R. S. Olsen, and P. W. Lurk: 2nd International Symposium on Field Screening Methods for Hazardous Wastes and Toxic Chemicals, (1991) 57. 25) W. Chudyk, K. Pohlig, L. Wolf, and R. Fordiani: Proc. SPIE 1172 (1989) 123. 26) W. Schade and J. Bublitz: ES&T 30 (1996) 1451. 27) G. Bujewski and B. Rutherford: USA EPA Report #EPA/600/R-97/ 020 (1997). 28) R. W. St. Germain and G. D. Gillispie: Proc. SPIE 1637 (1992) 159. 29) J. Bloch, B. Johnson, N. Newbury, J. Germaine, H. Hemond, and J. Sinfield: Appl. Spectrosc. 52 (1998) 1299. 30) R. J. Radziemski and D. A. Cremers: Laser-Induced Plasmas and Applications (Marcel Dekker, Inc., New York, 1989). 31) G. A. Theriault and S. H. Lieberman: in Proc. of the European Symposium on Optics for Environmental and Public Safety (1995). 32) B. C. Castle, K. Visser, B. W. Smith, and J. D. Winefordner: Appl. Spectrosc. 51 (1997) 1017. 33) T. E. Bell: IEEE Spectrum, October 1994, p. 14. 34) Active Impulse Systems, Natick, Massachusetts, USA, www.activeimpulse.com. 35) SPARTA, Inc., Laguna Hills, California, USA, www.sparta.com/ Products.html. 36) Uniphase Lasers & Fiberoptics, a division of Uniphase Corp., San Jose, California, USA. 37) SYNOPTICS, a division of Litton Airtron, Charlotte, North Carolina, USA. The Review of Laser Engineering December 1998