Efficient Third-Harmonic Generation with a Ti:Sapphire Laser

Alex Yu. Dergachev, Bhabana Pati and Peter F. Moulton

Q-Peak, Inc., 135 South Road, Bedford, MA 01730

dergachev@qpeak.com

Abstract: We report the highest efficiency (to the best of our knowledge) third-harmonic generation from a gain-switched Ti:sapphire laser. Using Type I LBO crystals, we obtained ~45 % conversion in the sum-frequency stage and ~35% overall conversion from the fundamental. The diode-laser-injection-seeded Ti:sapphire laser operated in 920-960 nm range to provide 307-320-nm UV energy.

OCIS codes: (140.3590) Lasers, titanium; (190.2620) Frequency conversion

Introduction

The primary objective of this work was the development of an all-solid-state source that provides nanosecond-duration pulses in 300-320 nm range. The specific application is for ozone differential absorption lidar (DIAL).

Atmospheric studies of ozone have been conducted successfully from both ground-based and aircraft-based systems employing a variety of sources that emit in the 285- to 315-nm region. The sources used to date for ozone DIAL work have included Nd:YAG-pumped, frequency-doubled dye lasers; Raman-shifted, quadrupled, Nd:YAG lasers; excimer lasers; and Raman-shifted excimer lasers. Existing aircraft and future space-based ozone measurements would benefit from coherent sources that are more efficient, compact, robust, reliable, and low-maintenance than those currently available.

We describe here an all-solid-state laser transmitter, based on third-harmonic generation (THG) of a titanium-sapphire laser, that provides 35 mJ output pulses with 500-ms temporal separation in the 307-320 nm wavelength range.

Ti:sapphire laser

As the tunable oscillator, we used an unstable-resonator, high-energy, gain-switched Ti:sapphire laser [1]. The laser was pumped by a high-energy, flashlamp-pumped, Q-switched, frequency-doubled, Nd:YLF laser configured for a 5-Hz, double-pulsed format. We achieved dual-wavelength operation of the Ti:sapphire laser by incorporating a mirror mounted on a high-speed galvanometer scanner that can rapidly switch the laser to any wavelength in the 920-960-nm range in a time less than the double-pulse separation of 500 ms.

In order to reduce the Ti:sapphire laser linewidth and increase harmonic conversion efficiency [2] we injection-seeded the laser using two external-grating diode lasers (EOSI, Inc) tuned to regions around 925 nm and 945 nm, respectively. The linewidth of the Ti:sapphire laser was ~ 3-4 cm-1 in unseeded regime, and less than 0.007 cm-1 in seeded regime of operation.

The main parameters of the pump Nd:YLF laser and unseeded Ti:sapphire tunable oscillator are given in Table 1. (When the Ti:sapphire laser was seeded the output energies were ~10-15% smaller).

 

Table 1. Main parameters of the Nd:YLF and Tisapphire lasers

 

Wavelength,

nm

Beam Quality

Beam Diameter, mm

Min.Pulsewidth, ns

Max. Energy per pulse, mJ

Nd:YLF pump laser

527

Multimode,

M2 ~10

~8

~10

~525

925

TEM00

M2 ~ 1.3

~2

~20

155

945

TEM00

M2 ~ 1.3

~2

~30

125

Harmonic Generation

In order to select the optimum approach for efficient frequency tripling of the Ti:sapphire laser we performed a theoretical evaluation on a number of nonlinear materials suitable for the sum-frequency-generation stage of tripling. The results for calculation of "power threshold" NLO figure-of-merit [3] for BBO (Type I and II), LBO (Type I and II) and CLBO (Type I) are presented in Fig.1 as a function of third-harmonic wavelength. The high "threshold" for BBO results from its small angular acceptance.

thgti1a.gif (13283 bytes)

Figure 1. THG "power threshold" versus THG wavelength in borate materials.

It follows from calculations that CLBO (Type I) and LBO (Type II) exhibit the lowest thresholds. However, LBO (Type II) does not have phase matching for fundamental wavelengths shorter than 945 nm and CLBO has significant unsolved problems related to its hygroscopic nature. Therefore LBO (Type I) is the best practical material for the THG stage when the fundamental wavelength is less than 945 nm. We employed it for our system.

Figure 2. Schematic of LBO THG configuration.

A schematic of the set-up for THG generation is shown in Fig.2. We used a LBO (Type I), 20-mm long, AR coated crystal for second harmonic generation. The efficiency of second harmonic generation was ~50% in the seeded or unseeded regime of operation. We evaluated performance of the LBO (Type I) sum-frequency crystal with and without a 2X telescope, which was used to reduce the diameter of the first- and second-harmonic beams. The laser energy per pulse was set to ~ 100 mJ in the seeded regime of operation, which corresponds to ~0.1 GW/cm2 intensity on the SHG crystal.

The THG conversion was characterized by two numbers: (1) overall efficiency, defined as the ratio of the pulse energy (1w) at the input of the SHG crystal to the THG pulse energy (3w) at the output from the THG crystal, and (2) THG crystal efficiency, defined as the ratio of the pulse energy at the input of the THG crystal (1w+2w) to the THG pulse energy (3w) at the output from the THG crystal. Losses from less-than-ideal coatings on the THG crystal faces and telescope were not taken into account for efficiency calculations.

We measured the following THG efficiencies (seeded operation of Ti:sapphire laser):

Efficiency No telescope With telescope
THG Crystal ~40 ~45
Overall ~30-35 ~35

The dependence of THG output energy and THG crystal efficiency versus the input pulse energy in seeded and unseeded regimes of operation (using 2X telescope) are shown in Fig. 3. The THG conversion efficiency was ~ 3 times larger for the seeded regime than for unseeded. We measured a maximum conversion efficiency of ~ 45 % (~35% overall). The conversion efficiency stayed approximately the same when the wavelength was tuned in 925-945-nm range. Damage limitations in some of the external optics limited the maximum energy input, but we did not observe any damage to the nonlinear crystals.

thgti3.gif (7393 bytes)

Figure 3. THG output energy and THG crystal efficiency versus the combined input energy at the THG crystal  (with telescope, fundamental wavelength – 925 nm).

With improved optics and coatings on the nonlinear crystals we expect that the overall THG efficiency could approach 50%. The large bandgap of LBO serves to reduce or eliminate losses and dephasing related to two-photon absorption of the THG energy, but this issue requires further study.

In conclusion, we have demonstrated highly efficient frequency tripling of a TEM00-mode, injection-seeded Ti:sappphire laser operated in 925-945 nm range using LBO (Type I) crystals. While this effort emphasizes the ozone application with operation in the 308- to 320-nm region, the same approach could be applied for OH and SO2 sensing at 308 nm and 300 nm, respectively, as well for 290-nm excitation of biological species.

Acknowledgments

This work was supported by the NASA Langley Research Center under Contract NAS1-20574.

References

G.A.Rines and P.F.Moulton, "Performance of gain switched Ti:Al2O3 unstable resonator lasers," Opt.Lett. 15, 434-436 (1990).

G.A.Rines, H.H.Zenzie, R.A.Schwarz, Y.Isyanova, P.F.Moulton, "Nonlinear conversion of Ti:sapphire laser wavelengths," IEEE J. of Selected Topics in Quantum Electron. 1, 50-57 (1995).

D.A.Roberts, "Simplified characterization of uniaxial and biaxial nonlinear optical crystals: a plea for standartization of nomenclature and conventions," IEEE J. Quantum Electron. 28, 2057-2074 (1992).


WB01624_.gif (281 bytes) If you got here from a search engine, click the icon to go to Recent Technical Papers at Q-Peak