Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Stable 10,000-hour operation of 20-W deep ultraviolet laser generation at 266 nm

Open Access Open Access

Abstract

We report 10,000-hour stable operation of a 266-nm picosecond laser with an average power of 20 W. We have developed a narrow-linewidth, high-peak-power 1064-nm laser source with a repetition rate of 600 kHz, an average power of 129 W, a linewidth of 0.15 nm, and a pulse duration of 14 ps using a gain-switched DFB-LD as a picosecond pulse seed source and a four-stage power amplifier with an Nd:YVO4 crystal. A 266-nm laser with a maximum average power of 25.4 W was generated by frequency conversion using LBO and CLBO crystals and had a pulse duration of 8 ps and beam quality factor of 1.5 at 20W. To the best of our knowledge, we also demonstrated that the average power and the beam quality can be maintained for 10,000 hours for the first time. We have confirmed the durability of the developed deep ultraviolet laser for industrial applications.

© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement

1. Introduction

High-power deep ultraviolet (DUV) short-pulse lasers with high photon energy have various applications, such as high-intensity gamma-ray generation [1], material processing [24], and semiconductor inspection [5]. DUV solid-state lasers, which use technologies of solid-state laser and frequency conversion, can achieve high beam quality which can be focused to a small spot size using lenses. In addition, DUV solid-state lasers are easier to handle and consume less electric power than conventional excimer lasers, which are currently used in industrial applications, such as precise and high-quality laser machining.

In the 80s and 90s, the development of borate-based nonlinear optical crystals, such as β-BaB2O4 (BBO) [6] and CsLiB6O10 (CLBO) [7], led to the development of high-power DUV solid-state lasers with wavelength below 300 nm. With the advancement of high power near-infrared lasers, the development of fourth harmonic light sources has been advanced [4,816]. As for the average output power of the fourth harmonic, since 2000 the average output power of 12 to 40 W using a multi-transverse-mode high-power nanosecond pulsed laser as the fundamental wave has been reported [811], and since 2009 the average output power of 10.5 to 14.8 W using a single-transverse-mode high-power nanosecond pulsed laser as the fundamental wave has been reported [1214]. Recently, the average output power of 11.4 to 20 W using a single-transverse-mode high-power picosecond pulsed laser as the fundamental wave has been reported [4,15,16]. In 2020, an average output power of 20 W at 258 nm was reported, with a repetition rate of 1 kHz using a fundamental wave that consists of a picosecond pulse oscillator with a Yb:YAG innoslab amplifier and a BBO crystal [16]. In 2021, an average output power of 20 W at 258-nm was reported using a fundamental wave that consists of a picosecond pulse oscillator and a Yb:YAG thin disk amplifier [4].

In addition to a high average power, stable long-term DUV generation is essential for industrial applications. In 2002, continuous DUV generation for 100 hours at 20 W was reported [17], however, power degradation was obvious, and 20 W was maintained for only 50 hours. Similar power degradation has been shown in other studies [13,18,19], and the realization of long-term stable operation remains a critical issue for high-power DUV solid-state lasers. In order to solve this issue, it was necessary to mitigate laser-induced damage in the nonlinear optical crystal that generates DUV light. This is done by not only to improve the quality of the nonlinear crystal, but also to suppress the DUV peak power density in the crystal to a point below 45 MW/cm2 [20]. The three acceptance bandwidths for phase-matching conditions (spectral, angular, and thermal) are narrow in DUV light generation for both BBO and CLBO [21], and conventional pulsed lasers can not satisfy all of them at the same time. Therefore, highly efficient frequency conversion could not be achieved, and power degradation in long-term operation was inevitable. In 2013, we demonstrated a fundamental laser source for DUV generation [22] that used a combination of a gain-switched LD and a hybrid amplifier with a linewidth of 0.1 nm and high-peak power 2.1 MW. The combination of the largely collimated high-peak-power picosecond pulse and the large aperture and long CLBO crystal satisfied the spectral and angular acceptance bandwidths in the DUV generation, while avoiding conversion efficiency roll-over caused by temporal and spatial walk-off. The frequency conversion efficiency from 532 to 266 nm exceeded 50%. The largely collimated beam also helped to suppress the temperature rise in the CLBO crystal, which is caused by linear absorption and two-photon absorption [2325], and maintained high conversion efficiency by satisfying the thermal acceptance bandwidth. In addition, beam divergence was much lower than that of the angular acceptance bandwidth of CLBO (spectral 0.13 nm, angular 0.49 mrad, thermal 6.2 K at 1 cm crystal length) [21]. This fundamental laser source enabled the DUV generation with high power and long-term stable operation. In our previous work [15], a fundamental laser with a 15-ps pulse, 46.5-W average power, 0.22-nm linewidth, and 17.2-MW peak power at a 200-kHz repetition rate was developed. The DUV had an average power of 14 W and a pulse duration of 10 ps using a 15-mm-long CLBO crystal with a conversion efficiency of 54% from 532 nm at a repetition rate of 200 kHz. Furthermore, we demonstrated long-term DUV generation at an average power of 10 W for 5000 hours with a peak power density of 50 MW/cm2 at 266 nm and showed that there was no power or beam profile degradation during operation.

The obtained maximum DUV power was limited by the available fundamental power, and there are demands for a much higher power DUV laser for industrial applications. In this study, the seed laser section and solid-state amplification section of the fundamental laser are improved to obtain a higher output power. By increasing the fundamental power to an average of 129 W with a linewidth of 0.15 nm and a peak power of 15.4 MW, the maximum generation of DUV reached an average power of 25.4 W at a repetition rate of 600 kHz using a 15-mm-long CLBO crystal, and a pulse duration of 8 ps at 266 nm was generated from 532 nm with a conversion efficiency of 28.9%. To the best of our knowledge, this is the first demonstration that an average power of 20 W and a beam quality factor (M2) of less than 1.5 can be maintained for 10,000 hours, with the 266-nm peak power density in the CLBO crystal being 34 MW/cm2.

2. Narrow-linewidth and high-peak power laser source development

In order to efficiently generate high-power DUV lasers, it was necessary to develop a high-power fundamental laser source with a narrow-linewidth and a high-peak power that satisfies all three acceptance bandwidths for the phase-matching conditions (spectral, angular, and thermal) of the nonlinear optical crystals.

2.1 Experimental setup of the laser source

Figure 1 shows the experimental setup of a laser source that consists of a seed laser section and a power amplifier section. The seed laser consists of a seed source, an amplified spontaneous emission (ASE) noise source, and a two-stage fiber pre-amplifier, and the power amplifier that consists of four-stage solid-state amplifiers and an acousto-optic modulator (AOM).

 figure: Fig. 1.

Fig. 1. Experimental setup of the laser source

Download Full Size | PDF

Although the hybrid amplifier that consists of fiber and solid-state amplifiers with a short- optical path length realized a robust and stable system compared with regenerative amplifiers and multi-pass thin disk amplifiers that use a mode-locked oscillator as a seed laser, instability of power and beam-pointing were observed in more than six-stage solid-state amplifier system.

A gain-switched distributed feedback-laser-diode (DFB-LD) [26] is used as the seed source in the seed laser section, and picosecond seed pulses are generated by injecting electrical pulses with a pulse duration of 100 ps into the DFB-LD at 600 kHz using a high-speed current driver developed in-house. The pulse duration and average power were 16 ps and 1.75 µW, respectively. A semiconductor optical amplifier (SOA) was used as a superluminescent diode for an intentional ASE noise source. In pulse-on-demand mode operation that triggerable any timing trigger signal, an ASE noise of around 1064 nm was injected to protect the solid-state amplifier in the later stages and to modulate the pulses. The time interval of the seed pulses was counted by the main control board. When the interval exceeded the specified time of 2100 ns in this system, the bias current calculated by the interval time was applied to the SOA to generate the ASE noise. By injecting the ASE noise continuously through the solid-state amplifiers, the gain for the pulses were controlled, and transition characteristics of the ON/OFF pulse train can be controlled as required by the pulse-on-demand system. Note that an ASE noise has negligible frequency conversion efficiency. In burst pulse mode operation, the height of the pulse train can be adjusted by modulating the electrical pulse current injected into the SOA. In this experiment, the bias current was 0 A because the abovementioned time interval was operated within the specified time, and a constant pulsed current of 400 mA with a pulse duration of 5 ns was injected into the SOA in synchronization with the seed pulse, resulting in an average output power of 16 µW. Note that when the pulse-on-demand and burst modes operation are not needed, the SOA can be replaced by a fiber amplifier. In the two-stage pre-amplifiers, both Pre-Amp 1 and Pre-Amp 2, core-pumped polarization-maintaining fiber amplifiers were used. The active fiber of Pre-Amp 1 was a 1-m-long 6-µm core (Yb-401-PM, Coractive) for forward pumping. After the active fiber, a hybrid device that consists of an optical isolator and a bandpass filter with a bandwidth of 2 nm was spliced. For Pre-Amp 2, a 1-m-long 10-µm core active diameter fiber (PLMA-YSF-10-125, Nufern) and passive fiber (FUD-3561, Nufern) were used. It was configured with backward pumping using a WDM hybrid device with a built-in 1064-nm optical isolator. In addition, a hybrid device that consists of an optical isolator and band pass filter was spliced to obstruct the return light from the solid-state amplifier.

Solid-state amplifiers were used for Amps 1 to 4 in the power amplifier to keep the linewidth narrow. The amplifying crystal is a 0.4 at.% doped neodymium-doped yttrium vanadate (Nd:YVO4) crystal [4 mm × 4 mm × 20 mm (α-cut) length] with a transmitted wavefront distortion of less than λ/6 at 633 nm (peak-to-peak). Since a single 40 mm crystal can not guarantee low-transmitted-wavefront distortion, two of these crystals are placed facing each other for a total length of 40 mm. An AR coating of 1064 nm and 888 nm was applied to both ends. To prevent parasitic oscillation, wedges angle of 1° were added, and the reflectance of the AR coating at 1064 nm was set to less than 0.1%. The four sides of Nd:YVO4 crystals were wrapped with 50-µm-thick indium foil and fused to a water-cooled copper heat sink. A wavelength-locked LD with a wavelength of 888 nm was used as the pump source, which is in-band pumping to the emitting level and less thermal effect in the crystal [2729]. The pump source was delivered through a 200-µm core diameter graded index fiber and launched into Nd:YVO4 crystals through a collimation lens and a focus lens module. The configuration of the power amplifier is shown in Fig. 2.

 figure: Fig. 2.

Fig. 2. Configuration of the power amplifier. M1: HR 1064-nm mirror; PBS1, 2: polarization beam splitter; FR: Faraday rotator; HWP1: half waveplate; DM1: HR1064/AR 1176-nm dichroic mirror; DM2: HR1064&1176/AR 888-nm dichroic mirror; DM3: HR1064/AR 888-nm dichroic mirror; AOM: acousto-optic modulator; L1-10: lenses.

Download Full Size | PDF

The fiber coupled collimator was connected to the delivery fiber from the fiber amplifier to launch the seed beam into free-space solid-state power amplifier section, and the beam diameter at the Nd:YVO4 crystal was adjusted using the lenses (L1-7). The coatings of DM1 and DM2 were HR1064/AR1176 and HR1064&1176/AR888, respectively. In order to protect the pump LD and the delivery fiber from the stimulated Raman-scattering (SRS) light generated by low-repetition-rate operation or other unforeseen circumstances, we added a AR1176-nm coating to DM1 and a HR1176-nm coating to DM2. The 1176-nm AR&HR coating on these mirrors is corresponding to the wavelength of the SRS light generated by Nd:YVO4, when the input peak power density of the 1064-nm laser is high. The laser pulses were collimated to a diameter of 2mm using lenses (L8– L10) and launched into an AOM with an active aperture diameter of 6.5mm, which enabled ON/OFF control of the laser pulses by adjusting the delay time of the seed pulses in synchronization with the Gate signal. The AOM device is made of crystal quartz with a 1064-nm AR coating.

2.2 Experimental results for the laser source

The DFB-LD and SOA were driven at 600 kHz, and the pumping power of Pre-Amp 1 and Pre-Amp 2 were 88.5 and 311 mW, respectively, for generating 3.4 mW as a seed laser. In Amp 1, the beam diameter was adjusted to φ0.6 mm at the Nd:YVO4 crystal using the lens (L1), and to prevent parasitic oscillation in the amplifier (Amp 1), PBS1, the Faraday rotator, and HWP1 were tilted 1 to 2° with respect to the optical axis. An amplified power of 10.1 W with a M2 = 1.25 was obtained by a pump power of 80 W. In Amp 2, the beam diameter at the Nd:YVO4 crystal was adjusted to φ0.7 mm using a lens (L2), and then a plano concave lens (L3) was installed and its focal length selected so that the amplified beam divergence angle was less than 2 mrad. An average output power of 47.9 W and M2 =1.35 were obtained when the pump power was 90 W. In Amps 3 and 4, the beam diameters at the Nd:YVO4 were adjusted by carefully monitoring the amount of SRS light and set to φ0.9 mm and φ1.0 mm, respectively, at a repetition rate of 600 kHz. For Amp 3, an average output power of 98.6 W and M2 = 1.40 were obtained at a pump power of 120 W, and for Amp 4, an average output power of 150 W and M2 = 1.48 were obtained at a pump power of 154 W. Table 1 shows the pump power, amplified power, and gain at each stage of the amplifier.

Tables Icon

Table 1. Pump power and amplified power of each stage of the fundamental laser source

Using a total pumping power of 444 W, the seed laser obtained from the DFB-LD was amplified by 79.3 dB to generate an average power of 150 W with an optical-to-optical efficiency of 33.8%. The average power characteristic of the power amplifier as a function of the repetition rate of Amps 3 and 4 and the beam profile at 600 kHz are shown in Fig. 3.

 figure: Fig. 3.

Fig. 3. Average power characteristic Amp3 and Amp4. The inset shows a beam profile at 600 kHz.

Download Full Size | PDF

In Amp 3, saturation amplification was observed and the amplified power did not change, even when the repetition rate was decreased to 500 kHz. On the other hand, in Amp 4, due to SRS effect, the 1064-nm power decreased when the repetition rate was lower than 600 kHz. When the peak power density exceeded 3 GW/cm2 [30], which is the threshold for SRS in a 40-mm long Nd:YVO4 crystal below 600 kHz, SRS light at 1176 nm transmitted through DM1 (HR1064/AR1176-nm) after Amp 4 was confirmed. Since a high-peak power is required for DUV generation, we decided to operate the system at 600 kHz. The inset in Fig. 3 shows the beam profile at 600 kHz, and it was confirmed that the beam maintained a good shape. In addition, the pulse duration was measured using an autocorrelator (FR-103XL, Femtochrome) and was determined to be 14 ps, with a peak power of 17.4 MW. We measured the optical spectrum of the low-power seed laser section using an optical fiber connected to the optical spectrum analyzer (AQ-6315A, ANDO) and the optical spectrum of the power amplifier using a free-space wavelength meter (SHR, SOLARLS). The results are shown in Fig. 4.

 figure: Fig. 4.

Fig. 4. Spectral waveform of (a) the seed laser and (b) the power amplifier.

Download Full Size | PDF

The DFB-LD generated a laser beam with a very narrow linewidth of about 20 MHz when driven into CW. However, when driven in pulsed operation by the gain-switching method, the spectrum became rather broad, and the linewidth (FWHM) was broadened to 0.44 nm when amplified with the SOA and Yb:doped fiber amplifiers. This linewidth was not sufficiently narrow for the acceptable bandwidth of the frequency conversion to DUV. Note that when amplified with an SOA and Yb:doped fiber amplifiers, the short-wavelength region was strongly amplified because the shorter wavelength of the optical pulses was emitted earlier during the transient lasing process in DFB-LD [31]. Here, spectral narrowing could be achieved by amplifying the seed light using an Nd:YVO4 crystal with narrow amplification bandwidth (FWHM) of 0.5 nm and a large stimulated emission cross-section of 25 × 1019 cm2 [22,32]. Taking advantage of this characteristic, we amplified the seed laser after Pre-Amp 2 by 46.4 dB with the 4-stage Nd:YVO4 amplifier to narrow the spectrum and obtained a linewidth of 0.15 nm, which is narrow enough taking account of the acceptance bandwidth in CLBO. In addition, an AOM was installed to realize ON/OFF of the laser pulse train, and 129 W first-order diffracted light was obtained with a diffraction efficiency of 86%.

With the above configuration, we have realized a fundamental laser with a narrow-linewidth and high-peak power of 129 W average power, 0.15 nm linewidth, and 15.4 MW peak power, which is suitable for high power deep ultraviolet generation.

3. Frequency conversion to the stable DUV

3.1 Experimental setup for second and fourth harmonic generation

To generate DUV light from 1064 to 266 nm, two nonlinear optical crystals were used to generate the second and fourth harmonics, as shown in Fig. 5.

 figure: Fig. 5.

Fig. 5. Configuration of the frequency convertor. A1: water cooled aperture; DM4: HR532/AR1064-nm dichroic mirror; HWP2: half waveplate; PBS3: polarization beam splitter; W1: AR532-nm window; DM5: HR532/AR266-nm dichroic mirror; DM6: HR266/AR532-nm dichroic mirror; BW: CaF2 Brewster window; L11-13: lenses; CDA: clean dry air.

Download Full Size | PDF

For the second harmonic generation (SHG), a LiB3O5 (LBO) crystal with an element size of 6 mm × 6 mm, 20-mm length (θ = 90.0°, φ = 10.4°, critically phase-matching), and an AR coating of 1064 nm and 532 nm on both ends was used as the nonlinear optical crystal. The LBO crystal was held in a copper holder with a built-in heater, and the holder temperature was set at 56 °C. 532-nm beam was generated by launching 1064-nm beam with a diameter of φ2.0 mm. A water-cooled aperture (A1) with a 5 mm aperture diameter was placed just before the LBO crystal. Due to the thermal lensing effect of the solid-state amplifier, the amplified 1064-nm beam had a higher-order profile at a few watts of power, slightly heating the LBO crystal holder and caused fluctuations in the 532-nm laser power. The installation of A1 generated a 532-nm beam without short-term power fluctuations when the AOM switch is on. The 532-nm laser was reflected by a dichroic mirror (DM4) coated with HR532/AR1064. The beam diameter was then expanded to φ7.8 mm using three lenses (L11, L12, L13), and collimated, and passed through a motor-driven half-waveplate (HWP2) and a polarizing beam splitter (PBS3). For the fourth harmonic generation, a CLBO crystal grown at Osaka University [33], with an element size of 16 mm × 16 mm, 15-mm length (θ = 62.0°, φ = 45.0°, critically phase-matching) was used as the nonlinear optical crystal. An AR532&266-nm coating on the input end only, and 1° wedges on both ends were applied. To avoid laser induced damage, AR coating was not applied on the output end. The crystals were placed in a chamber purged with clean dry air (CDA) to prevent deliquescent of the CLBO crystal and laser induced contamination. The CDA was generated from the atmosphere by an air purifier system which is specially designed in-house for DUV generation, because we recognize the conventional purge technics using nitrogen or argon gas is not sufficient to keep the performance of the DUV optics, including CLBO, in the long operation. The CLBO temperature was adjusted to 152 °C. and CDA was injected from the DM6 side into the chamber, which then exited from the DM5 side to purge BW. The 532 nm light was launched from W1, reflected by two HR532/AR266 nm dichroic mirrors (DM5), and launched into the CLBO crystal. The CLBO crystal was arranged so that the 266-nm polarization was S-polarized with respect to the rear HR266/AR532 nm dichroic mirror (DM6). The 266-nm laser reflected by the two DM6s was emitted out of the chamber through a CaF2 single crystal window (BW) placed at the Brewster angle. When collimated light is launched into the CLBO crystal to generate 266 nm, multiple reflections occurred inside the CLBO crystal. As a result, the 266-nm beam generated in a single pass contained multiple reflected stray light rays that were almost coaxial. We confirmed that stray light was problematic when the DUV light source was used for processing materials such as thin film patterning with a low processing threshold. In this experiment, we solved this problem by applying an AR coating to the incident surface and adding a wedge.

3.2 Experimental results

3.2.1 Second harmonic generation

An average power of 88.9 W at 532 nm was obtained with a conversion efficiency of 71.9% when a 1064-nm laser with an average power of 125 W was launched into the LBO crystal with a beam diameter of φ2.0 mm, and peak and average power densities of 947 MW/cm2 and 7962 W/cm2, respectively. Under these conditions, the thermal effects in LBO when the AOM is switched on and off were negligibly small. The SHG power characteristics are shown in Fig. 6, with no roll over in the conversion efficiency and no beam degradation showing in the intensity distribution. The linewidth and pulse duration were measured using a wavelength meter (SHR, SOLARLS) and a streak camera (C10910, Hamamatsu Photonics) to be 0.042 nm and 11 ps, respectively, with a peak power of 13.6 MW. The beam quality was measured using a beam profiler (SP920, Ophir) with a focal length of 1100 mm, and the beam quality factor was determined to M2 = 1.2.

 figure: Fig. 6.

Fig. 6. (a) Input and output characteristics of LBO crystal for SHG and (b) Beam propagation measurement results for the 532 nm laser source.

Download Full Size | PDF

3.2.2 Fourth harmonic generation

An average power of 25.4 W at 266 nm was obtained with a conversion efficiency of 28.9% when a 532-nm laser with an average power of 87.9 W was launched into the CLBO crystal with a beam diameter of φ7.8 mm, and peak and average power densities of 56 MW/cm2 and 368 W/cm2, respectively. The input-output characteristics of 532 nm when the input was varied using the waveplate HWP2 are shown in Fig. 7. Note that the CLBO temperature was not changed during the measurement. As in the case of 532 nm, the linewidth and pulse duration were measured to be 0.016 nm and 8 ps, respectively, and the peak power was 5.3 MW. The peak power density and average power density at 266 nm in the end of the CLBO crystal were 43 MW/cm2 and 206 W/cm2, respectively. Although the FHG process generally caused a larger thermal effect than the SHG process, under these conditions of low-power density, the thermal effects in CLBO when the AOM is switched on and off were tolerably small. A thermal sensor with a response time of 1.5 s was used for the measurement sampled in 0.1 seconds. We confirmed that the 10%−90% rise time at 20 W was 1.3 s when the AOM is switched on, which is less than the response time of the sensor and that the stability from 2 s to 120 s was 1% pp. This ON/OFF operation with suppressed thermal effects will be presented in detail in another paper. The beam quality was measured at 20 W using a lens with a focal length of 1016 mm, and the beam quality factor was determined to M2 = 1.5. The beam diameter and circularity were φ5.6 mm and 95%, respectively, and owing to the largely collimated light, satisfactory beam characteristics were achieved without any beam shaping optics after 266-nm generation. Note that the beam was measured with a beam profiler (L11050, Ophir) at 550 mm from the exit window.

 figure: Fig. 7.

Fig. 7. (a) Input and output characteristics of CLBO crystal for FHG and (b) Beam propagation measurement results of 266 nm at 20 W average power. The insets of (a) show the beam profile and (b) shows the beam profile at the focal position.

Download Full Size | PDF

Figure 8 shows the average power stability of the 266-nm laser during free-running after setting the 532-nm input power so that the average power was around 20 W. The laser generated stably for more than 24 hours, with an average power stability of 2.4% peak-to-peak and 0.45% rms over 30 hours. The power fluctuation dependent on the ambient temperature was 3% per degree Celsius. A critically phase-matched CLBO is sensitive to pointing stability of the 532-nm beam. Therefore, we consider the power fluctuation was caused by pointing deviation due to the mechanical deformation of the laser head and optical bench which is affected by thermal expansion of the materials.

 figure: Fig. 8.

Fig. 8. Short-term average power stability.

Download Full Size | PDF

3.2.3 Long-term operation

When a 532-nm laser with an average power of 76 W was launched into the CLBO crystal, an average power of 20 W at 266 nm was obtained with peak and average power densities at 266 nm of 34 MW/cm2 and 162 W/cm2, respectively. During the operation period, the 532-nm power was adjusted within the range of 76 W ± 6 W to keep the 266-nm average power at 20 W using a waveplate (HWP2) to compensate for power fluctuations caused by environmental temperature and other factors. The CLBO crystal was used in the same position, and the average power was measured only once per day for less than 10 minutes so as not to damage the surface of the power meter with the strong 266-nm laser pulse. Additionally, the shutter mirror was closed and dumped 266 nm, except during power measurement. Figure 9 shows the test results when measuring the average power.

 figure: Fig. 9.

Fig. 9. (a) Results of a 10,000-hour test, (b) 266-nm output characteristics for 532-nm input before and after the test, and (c) beam propagation measurement results after the test. The insets of (a) show the beam profiles before and after the test and the inset of (c) shows the beam profile at focal position.

Download Full Size | PDF

As shown in Fig. 9(a) and 9(b), 20 W was maintained for 10,000 hours and the maximum average power remained above 25 W after the test. We assume the reason for the higher power characteristics after the test is affected by the 532-nm beam diameter reduction about 4% by thermal lensing effect caused by the increased heat generation in the optical components, including Nd:YVO4 crystals and lenses in the fundamental wave and SHG section. The inset in Fig. 9(a) shows the beam profile at 550 mm from the exit before and after the test. The beam diameter was reduced from φ5.6 to φ5.4 mm, and the intensity distribution was slightly closer to top-hat, but it was confirmed that a satisfactory beam shape was maintained. Figure 9(c) shows the measured beam quality at 20 W after the test, and the inset shows the beam profile at the focal point. We consider that the higher-order spatial noise around the beam was reduced and the M2 factor improved to 1.2 compared with before the test. Additional verification of this phenomenon is underway.

As described above, we have developed a picosecond pulsed laser source with narrow- linewidth and high-peak power, which satisfies all three acceptance bandwidths of the CLBO crystal (spectral, angular, and thermal), reduces the effects of spatial and temporal walk-off, and improves the conversion efficiency. Furthermore, we reduced the power density by increasing the beam diameter in the CLBO crystal, which led to the solution of the output power degradation problem for high-power DUV solid-state lasers. This achievement is expected to lead to the widespread use of DUV solid-state lasers with an average power of 20 W in industrial applications.

4. Conclusions

We reported 10,000-hour stable operation of a 266-nm picosecond laser with an average power of 20 W. We have developed a narrow-linewidth, high-peak-power 1064-nm laser source with a repetition rate of 600 kHz, an average power of 150 W, a linewidth of 0.15 nm, a pulse duration of 14 ps, and a peak power of 17.4 MW using a gain-switched DFB-LD as a picosecond pulse seed source and a power amplifier with Nd:YVO4 crystals. In addition, 129 W and 15.4 MW of peak power were achieved in a configuration with an additional pulse ON/OFF function for processing applications using an AOM.

In the SHG, we generated a 532-nm laser with an average power of 88.9 W at a conversion efficiency of 71.9% using a 20-mm-long LBO crystal, and obtained a linewidth of 0.042 nm, a peak power of 13.5 MW, and M2 = 1.2. In the fourth harmonic generation, we generated a 266-nm laser with an average power of 25.4 W at a conversion efficiency of 28.9% using a 15-mm-long CLBO crystal, and obtained a linewidth of 0.016 nm, a peak power of 5.3 MW, and M2 = 1.5 at 20 W. The CLBO crystal was heated to 152 °C and placed in a chamber purged with CDA, including the dichroic mirrors. Furthermore, by using the developed fundamental, the peak and average power densities at 266 nm can be suppressed to 34 MW/cm2 and 162 W/cm2, respectively, and the beam quality factor with an average power of 20 W and M2 < 1.5 can be maintained for 10,000 hours. The obtained maximum DUV power was still limited by the available fundamental power. There is a prospect for much higher power DUV generation with further improvements in solid-state amplification. To the best of our knowledge, this is the first demonstration of long-term DUV operation in the range of 20 W. From these results, we confirmed that proposed DUV laser configuration can be suitable for industrial applications.

Funding

New Energy and Industrial Technology Development Organization (P16011).

Disclosures

The authors declare no conflicts of interest.

Data availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

References

1. N. Muramatsu, Y. Kon, S. Daté, Y. Ohashi, H. Akimune, J. Y. Chen, M. Fujiwara, S. Hasegawa, T. Hotta, T. Ishikawa, T. Iwata, Y. Kato, H. Kohri, T. Matsumura, T. Mibe, Y. Miyachi, Y. Morino, T. Nakano, Y. Nakatsugawa, H. Ohkuma, T. Ohta, M. Oka, T. Sawada, A. Wakai, K. Yonehara, C. J. Yoon, T. Yorita, M. Yosoi, and L. E. P. S. Collaboration, “Development of high intensity laser-electron photon beams up to 2.9 GeV at the SPring-8 LEPS beamline,” Nucl. Instrum. Methods Phys. Res., Sect. A 737, 184–194 (2014). [CrossRef]  

2. Y. Imamiya, S. Akama, Y. Fujita, and H. Niitani, “Development of Microfabrication Technology using DUV Laser,” Mitsubishi Heavy Industries Technical Review 53(4), 49–54 (2016).

3. Y. Kawasuji, Y. Adachi, A. Suwa, J. Fujimoto, K. Kakizaki, and M. Washio, “Pulse duration Dependence of Ablation Threshold and Ablation Rate,” Micro Session 1 of Laser Material Microprocessing of ICALEO2020, 0442_0638_000138, (2020)

4. S. Häfner, C. Wagner, B. Shnirman, M. Ginter, J. Brons, M. Sailer, A. Fehrenbacher, D. Grossmann, D. Flamm, K. Janami, S. Ruebling, U. Quentin, A. Budnicki, I. Zawischa, and D. H. Sutter, “Deep UV for materials processing based on the industrial TruMicro Series of ultrafast solid-state laser amplifiers,” Proc. SPIE 11670, 116701F (2021). [CrossRef]  

5. H. Nakao, M. Morita, Y. Kaneda, A. Miyamoto, T. Tago, T. Sasa, M. Sasaura, and Y. Furukawa, “High power 4th harmonic generation with optimized enhancement cavity,” in 2017 European Conference on Lasers and Electro-Optics and European Quantum Electronics Conference, (Optical Society of America, 2017), paper CA_P_16.

6. C. Chen, B. Wu, A. Jiang, and G. You, “A new type ultraviolet SHG crystal β-BaB2O4,” Science in China (Ser. B) 28(3), 235–243 (1985).

7. Y. Mori, I. Kuroda, S. Nakajima, T. Sasaki, and S. Nakai, “New nonlinear optical crystal: Cesium lithium borate,” Appl. Phys. Lett. 67(13), 1818–1820 (1995). [CrossRef]  

8. T. Kojima, S. Konno, S. Fujikawa, K. Yasui, K. Yoshizawa, Y. Mori, T. Sasaki, M. Tanaka, and Y. Okada, “20-W ultraviolet-beam generation by fourth-harmonic generation of an all-solid-state laser,” Opt. Lett. 25(1), 58–60 (2000). [CrossRef]  

9. M. Nishioka, S. Fukumoto, F. Kawamura, M. Yoshimura, Y. Mori, and T. Sasaki, “Improvement of laser-induced damage tolerance in CsLiB6O10 for high-power UV laser source,” in Conference on Lasers and Electro-Optics/Quantum Electronics and Laser Science Conference, Technical Digest (Optical Society of America, 2003), paper CTuF2.

10. S. Konno, Y. Inoue, T. Kojima, S. Fujikawa, and K. Yasui, “Efficient high-pulse-energy green-beam generation by intracavity frequency doubling of a quasi-continuous-wave laser-diode-pumped Nd:YAG laser,” Appl. Opt. 40(24), 4341–4343 (2001). [CrossRef]  

11. G. Wang, A. Geng, Y. Bo, H. Li, Z. Sun, Y. Bi, D. Cui, Z. Xu, X. Yuan, X. Wang, G. Shen, and D. Shen, “28.4 W 266 nm ultraviolet-beam generation by fourth-harmonic generation of an all-solid-state laser,” Opt. Commun. 259(2), 820–822 (2006). [CrossRef]  

12. Q. Liu, X. P. Yan, X. Fu, M. Gong, and D. S. Wang, “High power all-solid-state fourth harmonic generation of 266 nm at the pulse repetition rate of 100 kHz,” Laser Phys. Lett. 6(3), 203–206 (2009). [CrossRef]  

13. A. Diening, S. McLean, and A. Starodoumov, “High Average Power 258 nm Generation in a Nanosecond Fiber MOPA System,” Proc. SPIE 7195, 719506 (2009). [CrossRef]  

14. H. Xuan, C. Qu, S. Ito, and Y. Kobayashi, “High-power and high-conversion efficiency deep ultraviolet (DUV) laser at 258 nm generation in the CsLiB6O10 (CLBO) crystal with a beam quality of M2 < 1.5,” Opt. Lett. 42(16), 3133–3136 (2017). [CrossRef]  

15. K. Kohno, Y. Orii, H. Sawada, D. Okuyama, K. Shibuya, S. Shimizu, M. Yoshimura, Y. Mori, J. Nishimae, and G. Okada, “High-power DUV picosecond pulse laser with a gain-switched-LD-seeded MOPA and large CLBO crystal,” Opt. Lett. 45(8), 2351–2354 (2020). [CrossRef]  

16. K. Liu, H. Li, S. Qu, H. Liang, Q. J. Wang, and Y. Zhang, “20 W, 2 mJ, sub-ps, 258 nm all-solid-state deep-ultraviolet laser with up to 3 GW peak power,” Opt. Express 28(12), 18360–18367 (2020). [CrossRef]  

17. T. Kojima, S. Konno, S. Fujikawa, K. Yasui, T. Kamimura, M. Yoshimura, Y. Mori, T. Sasaki, M. Tanaka, and Y. Okada, “100-hour operation of an all-solid-state 20-W 266-nm UV laser by using high-quality CLBO crystal,” ASSL, 68, Trends in Optics and Photonics Series, paper WC2. (2002).

18. S. Kumar, J. Casals, J. Wei, and M. Ebrahim-Zadeh, “High-power, high-repetition-rate performance characteristics of β-BaB2O4 for single-pass picosecond ultraviolet generation at 266 nm,” Opt. Express 23(21), 28091–28103 (2015). [CrossRef]  

19. Q. Fu, N. Hanrahan, L. Xu, S. Lane, D. Lin, Y. Jung, S. Mahajan, and D. Richardson, “High-power, high-efficiency, all-fiberized-laser-pumped, 260-nm, deep-UV laser for bacterial deactivation,” Opt. Express 29(26), 42485–42494 (2021). [CrossRef]  

20. K. Takachiho, M. Yoshimura, Y. Takahashi, M. Imade, T. Sasaki, and Y. Mori, “Ultraviolet laser-induced degradation of CsLiB6O10 and β-BaB2O4,” Opt. Mater. Express 4(3), 559–567 (2014). [CrossRef]  

21. N. Umemura, K. Yoshida, T. Kamimura, Y. Mori, T. Sasaki, and K. Kato, “New data on the phase-matching properties of CsLiB6O10,” ASSL Vol. 26 of OSA Trends in Optics and Photonics (Optical Society of America, 1999), paper PD15.

22. Y. Orii, Y. Takushima, M. Yamagaki, A. Higashitani, S. Matsubara, S. Murayama, T. Manabe, I. Utsumi, D. Okuyama, and G. Okada, “High-energy 266-nm picosecond pulse generation from a narrow spectral bandwidth gain-switched LD MOPA,” Tech. Digest of CLEO 2013, JTh2A.64, (2013).

23. A. Dubietis, G. Tamošauskas, A. Varanavičius, and G. Valiulis, “Two-photon absorbing properties of ultraviolet phase-matchable crystals at 264 and 211 nm,” Appl. Opt. 39(15), 2437–2440 (2000). [CrossRef]  

24. T. Kamimura, R. Nakamura, H. Horibe, M. Nishioka, M. Yamamoto, M. Yoshimura, Y. Mori, T. Sasaki, and K. Yoshida, “Characterization of two-photon absorption related to the enhanced bulk damage resistance in CsLiB6O10 crystal,” Jpn. J. Appl. Phys. 44(No. 21), L665–L667 (2005). [CrossRef]  

25. G. Kurdi, K. Osvay, J. Klebniczki, M. Divall, E. J. Divall, Á. Péter, K. Polgár, and J. Bohus, “Two-photon-absorption of BBO, CLBO, KDP and LTB crystals,” in Advanced Solid-State Photonics, Technical Digest (Optical Society of America, 2005), paper MF18.

26. H. F. Liu, M. Fukazawa, Y. Kawai, and T. Kamiya, “Gain-switched picosecond pulse (<10ps) generation from 1.3 mm InGaAsP laser diodes,” IEEE J. Quantum Electron. 25(6), 1417–1425 (1989). [CrossRef]  

27. V. Lupei, A. Lupei, N. Pavel, T. Taira, I. Shoji, and A. Ikesue, “Laser emission under resonant pump in the emitting level of concentrated Nd:YAG ceramics,” Appl. Phys. Lett. 79(5), 590–592 (2001). [CrossRef]  

28. V. Lupei, N. Pavel, and T. Taira, “Efficient Laser Emission in Concentrated Nd Laser Materials Under Pumping Into the Emitting Level,” IEEE J. Quantum Electron. 38(3), 240–245 (2002). [CrossRef]  

29. L. McDonagh and R. Wallenstein, “High-efficiency 60W TEM00 Nd:YVO4 oscillator pumped at 888 nm,” Opt. Express 31(22), 3297–3299 (2006). [CrossRef]  

30. J. A. Piper and H. M. Pask, “Crystalline Raman Lasers,” IEEE J. Sel. Top. Quantum Electron. 13(3), 692–704 (2007). [CrossRef]  

31. Y. Kusama, Y. Tanushi, M. Yokoyama, R. Kawakami, T. Hibi, Y. Kozawa, T. Nemoto, S. Sato, and H. Yokoyama, “7-ps optical pulse generation from a 1064-nm gain-switched laser diode and its application for two-photon microscopy,” Opt. Express 22(5), 5746–5753 (2014). [CrossRef]  

32. V. N. Smiley, A. L. Lewis, and D. K. Forbes, “Gain and Bandwidth Narrowing in a Regenerative He–Xe Laser Amplifier,” J. Opt. Soc. Am. 55(11), 1552–1553 (1965). [CrossRef]  

33. R. Murai, T. Fukuhara, G. Ando, Y. Tanaka, Y. Takahashi, K. Matsumoto, H. Adachi, M. Maruyama, M. Imanishi, K. Kato, M. Nakajima, Y. Mori, and M. Yoshimura, “Growth of large and high quality CsLiB6O10 crystals from self-flux solutions for high resistance against UV laser-induced degradation,” Appl. Phys. Express 12(7), 075501 (2019). [CrossRef]  

Data availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (9)

Fig. 1.
Fig. 1. Experimental setup of the laser source
Fig. 2.
Fig. 2. Configuration of the power amplifier. M1: HR 1064-nm mirror; PBS1, 2: polarization beam splitter; FR: Faraday rotator; HWP1: half waveplate; DM1: HR1064/AR 1176-nm dichroic mirror; DM2: HR1064&1176/AR 888-nm dichroic mirror; DM3: HR1064/AR 888-nm dichroic mirror; AOM: acousto-optic modulator; L1-10: lenses.
Fig. 3.
Fig. 3. Average power characteristic Amp3 and Amp4. The inset shows a beam profile at 600 kHz.
Fig. 4.
Fig. 4. Spectral waveform of (a) the seed laser and (b) the power amplifier.
Fig. 5.
Fig. 5. Configuration of the frequency convertor. A1: water cooled aperture; DM4: HR532/AR1064-nm dichroic mirror; HWP2: half waveplate; PBS3: polarization beam splitter; W1: AR532-nm window; DM5: HR532/AR266-nm dichroic mirror; DM6: HR266/AR532-nm dichroic mirror; BW: CaF2 Brewster window; L11-13: lenses; CDA: clean dry air.
Fig. 6.
Fig. 6. (a) Input and output characteristics of LBO crystal for SHG and (b) Beam propagation measurement results for the 532 nm laser source.
Fig. 7.
Fig. 7. (a) Input and output characteristics of CLBO crystal for FHG and (b) Beam propagation measurement results of 266 nm at 20 W average power. The insets of (a) show the beam profile and (b) shows the beam profile at the focal position.
Fig. 8.
Fig. 8. Short-term average power stability.
Fig. 9.
Fig. 9. (a) Results of a 10,000-hour test, (b) 266-nm output characteristics for 532-nm input before and after the test, and (c) beam propagation measurement results after the test. The insets of (a) show the beam profiles before and after the test and the inset of (c) shows the beam profile at focal position.

Tables (1)

Tables Icon

Table 1. Pump power and amplified power of each stage of the fundamental laser source

Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.