Open Access
Add to BibSonomy
Abstract
We have succeeded in bonding a Nd:YAG crystal and an anti-reflection coated diamond crystal to realize a composite laser which minimizes the Fresnel-reflection loss at the bonded interface for the first time. This newly developed composite laser has realized higher output power at higher pump power than a non-composite Nd:YAG laser while keeping nearly the same slope efficiency.
© 2017 Optical Society of America
1. Introduction
Thermal effects such as thermal lens and thermally induced birefringence are sources of degrading the beam quality and preventing higher-power operations in solid-state lasers [1]. Composite lasers, which consist of laser-ion-doped and undoped materials, have been successful in reducing those thermal effects because the heat generated in the doped material can be effectively removed into the undoped region [2]. Especially for end-capped composite lasers in the case of end pumping, direct removal of heat is possible from the input and/or output facets of the doped materials [3,4]. Composite structures have been fabricated mainly with the diffusion-bonding [5–7], and composite hybrid- or all-ceramic lasers have been also recently reported [8,9]. However, the diffusion-bonding is a high-temperature process so that it is difficult to bond materials with different thermal-expansion coefficients, while laser ceramics are usually limited to isotropic materials.
In this paper we propose a new technique, the room-temperature bonding (RTB), to develop new types of composite lasers. The RTB technique enabled us to fabricate directly bonded composite structures comprising laser crystals such as Nd:YAG and Nd:YVO4, and a diamond crystal as a heat spreader. Moreover, we have succeeded in bonding a Nd:YAG crystal and an anti-reflection coated diamond crystal to realize a composite laser without the Fresnel-reflection loss at the bonded interface for the first time. This newly developed composite laser has realized higher output power at higher pump power than a non-composite Nd:YAG laser with keeping nearly the same slope efficiency.
2. Fabrication of composite structures using RTB technique
Room-temperature bonding (RTB), which is also called as surface-activated bonding (SAB), is a versatile technique for bonding variety of materials including metals, semiconductors, dielectrics, and even different materials [10–12]. We have recently fabricated wavelength-conversion devices using RTB [13, 14]. It is processed under high vacuum (∼ 10−5 Pa) at room temperature. Figure 1(a) shows the RTB process. First, the two plates to be bonded are set in a vacuum chamber and irradiated by Ar atom beams. The oxidized layers, the adsorbed molecules, and/or surface-reconstructed atoms are then etched, so that the surface atoms of the plates have dangling bonds and are chemically activated (Fig. 1(b)). After that, when the two plates are touched and pressed (Fig. 1(c)), they are atomically bonded (Fig. 1(d)).
Fig. 1 (a) The room-temperature-bonding (RTB) process. (b) Surface activation by irradiation of Ar atom beams in the RTB process. (c) Two plates are touched and pressed. (d) Bonded plates are pulled up. (e) Fabricated Nd:YAG/diamond composite.
We previously fabricated composite structures comprising laser-ion-doped and undoped same host materials, Yb:YAG/YAG and Nd:YAG/YAG, using the RTB technique, and confirmed that those composite lasers exhibit higher-power and higher slope efficiency than the non-composite ones [15,16]. Since one of the advantages of RTB is to enable us to bond different materials with different thermal expansion coefficients, we then fabricated composite structures using an ideal material as a heat spreader, i.e. diamond. Diamond has an extremely high thermal conductivity of 2000 W/mK which is 200 times higher than that of YAG. On the other hand, its linear expansion coefficient (1 × 10−6 /K) is much smaller than that of YAG (8 × 10−6/K), which makes it difficult to apply the diffusion bonding. Although capillary bonding has been used for assembling laser crystals and diamond heat spreaders [17], those devices are not atomically bonded. Then the RTB is a superior method to fabricate atomically bonded composites comprising such different materials.
First, we fabricated a directly bonded composite of 1 at.% Nd:YAG and diamond. The size of the Nd:YAG crystal was 3 mm × 3 mm × 3 mm. On the other hand, the diamond we used was a single-crystal synthetic diamond grown by chemical vapor deposition (CVD) at Element Six, the size of which was 3 mm × 3 mm × 1.5 mm. The input and output faces of the crystals were polished with the surface flatness and roughness of less than λ/10 and Ra < 1 nm, respectively, which were measured with a 3D optical surface profiler (Zygo, NewView 7200). Crystals with worse flatness and roughness than those values resulted in failure of bonding. We also confirmed that there was no dust particle observed on the polished surface before bonding.
Room-temperature bonding of Nd:YAG and diamond were processed under the degree of vacuum of 4.4 × 10−5 Pa. The acceleration voltage, the electric current, and the irradiation time of Ar atom beams to the Nd:YAG and the diamond surfaces were 1.2 kV, 15 mA, and 30–70 seconds and 5–15 minutes, respectively. We did not observe any change of the surface flatness and roughness after the irradiation. The pressure between the Nd:YAG and diamond plates during the bonding process was 30 to 50 kg, which did not cause the deformation of the plates because of the high mechanical strength of Nd:YAG and diamond. The fabricated composite is shown in Fig. 1(e).
3. Depolarization measurement
We first measured the depolarization caused by the thermally induced birefringence in order to directly evaluate the reduction of the thermal effects. Since Nd:YAG is isotropic, a linearly polarized beam passes through its crystal as it is under unpumped condition. When the crystal is pumped and heat is deposited in the crystal, however, the photoelastic effect causes the thermally induced birefringence, generating depolarized components in the transmitted beam. Depolarization is defined as the ratio of depolarized power to the total probe power given by
where P⊥ is the depolarized power and P‖ is the undepolarized power. The experimental setup for the depolarization measurements is shown in Fig. 2. We performed the pump-probe experiment [18]. A fiber-coupled laser diode (LD) oscillating at the wavelength of 808 nm was used as the pump source. The pump beam was focused onto the sample with a radius of 200 μm. On the other hand, a linearly polarized red LD beam at the wavelength of 635 nm was used as the probe. After it passed through the sample, the probe beam was reflected at the plane surface of the focusing lens and went back through the sample again. Part of the beam was reflected at the beam splitter, and only the depolarized component of the probe beam transmitted the analyzer, which was in the crossed-Nicol configuration with the polarizer. The depolarized beam passed through a filter that absorbed the pump beam, the power of which was measured with a photodetector.Figure 3 show the dependence of the depolarization on the absorbed pump power for the Nd:YAG/diamond composite and the non-composite 1 at.% Nd:YAG crystal, respectively. We measured at the absorbed pump power lower than 15 W in order to prevent the samples from being damaged; the thermal load in the Nd:YAG crystals under the non-lasing condition in the depolarization measurement is much larger than that under the lasing condition in the laser oscillation measurement. It is found that the depolarization of the composite is nearly half that of the non-composite Nd:YAG, which indicates improvement of the heat-removal efficiency owing to the diamond heat spreader, although we did not directly measure the temperature of the samples in the present study.
Fig. 3 Dependence of the depolarization on the absorbed pump power for the Nd:YAG/diamond (squares) composite and the non-composite Nd:YAG (circles).
4. Laser oscillation measurement
Next, we performed the laser oscillation measurements for the directly bonded Nd:YAG/diamond composite. The experimental setup is shown in Fig. 4. A fiber-coupled laser diode at the wavelength of 808 nm was used as the pump source. The pump beam radius was about 200 μm. The laser cavity, the length of which was 50 mm, consists of a plane input mirror with high-reflection (HR) coating for the wavelength of 1064 nm and a concave output mirror with the curvature of 100 mm with the reflectivity of 80 %. The input face of the diamond was anti-reflection (AR) and high-transmission (HT) coated for the wavelengths of 1064 and 808 nm, respectively, while the output face of the Nd:YAG was AR and HR coated for the wavelengths of 1064 and 808 nm, respectively. Figure 5 shows the laser characteristics. The output power of the non-composite 1 at.% Nd:YAG single crystal (circles) began to decrease at 7.54 W for the pump power of 19 W, and the crystal was finally broken as a result of the severe thermal effect. On the other hand, the Nd:YAG/diamond composite (squares) did not show any saturation even at 30 W pump power, and achieved the maximum output power of 9.22 W. We did not observe any degradation such as occurrence of gap at the bonded interface at this power level.
Fig. 5 Laser characteristics of the non-composite Nd:YAG single crystal (circle), the directly bonded Nd:YAG/diamond composite (squares), and the Nd:YAG/diamond composite with AR coating at the bonded interface (diamonds).
This result confirmed that the diamond heat spreader worked effectively in the composite structure. However, the slope efficiency of the directly bonded Nd:YAG/diamond composite became smaller than that of the non-composite Nd:YAG from 46.3 % to 35.5 %. This is mainly caused by the Fresnel reflection at the bonded interface. Since the refractive indices of Nd:YAG and diamond are 1.82 and 2.39, respectively, the reflectivity at the interface is calculated to be 1.8 %, which corresponds to the round-trip loss of 3.7 % in the laser cavity. This value is comparable to the cavity loss arising from the composite, 7 %, which is estimated from the difference of the slope efficiencies between the composite and non-composite in Fig. 5.
In order to increase the efficiency of the composite laser, we newly fabricated a Nd:YAG/diamond composite which has an AR-coating layer at the bonded interface. The schematic is shown in Fig. 6(a). First, the surface of a diamond to be bonded was AR coated for the wavelengths of 808 and 1064 nm. Then the coated surface was room-temperature bonded to the surface of a 1 at.% Nd:YAG crystal. By optimizing the RTB process including the duration of the Ar beam irradiation and the pressure applied between the Nd:YAG and diamond, we have succeeded in bonding them. The fabricated composite is shown in Fig. 6(b).
Fig. 6 (a) Schematic of the Nd:YAG/diamond composite with AR coating at the bonded interface. The surface of the diamond to be bonded was AR coated for the wavelengths of 808 and 1064 nm. (b) The fabricated Nd:YAG/diamond composite with AR coating at the bonded interface.
After AR and HT coating for the wavelengths of 1064 and 808 nm, respectively, on the input face of the diamond, and AR and HR coating for the wavelengths of 1064 and 808 nm, respectively, on the output face of the Nd:YAG, we performed the laser oscillation measurement of the composite with AR coating at the bonded interface. Figure 5 shows the laser characteristics for the composite with its interface AR coated (diamonds), the directly bonded composite (squares), and the non-composite Nd:YAG (circles). The composite with its interface AR coated has achieved 44.9 % slope efficiency, which is almost the same with that of the non-composite Nd:YAG. Moreover this composite has obtained the output power of 11.4 W which is still higher than that of the directly bonded composite. This result shows that the Fresnel-reflection loss at the bonded interface was minimized by the AR coating.
We also performed the beam-quality (M2) measurements. Typical examples of the beam profiles at different pump power are shown in Fig. 7 for the non-composite Nd:YAG ((a) ∼ (c)) and the composite with its interface AR coated ((d) ∼ (f)). The M2 factors for both the samples increased as the pump power increased. However, better beam quality was obtained for the composite than the non-composite: although the beam profile of the non-composite Nd:YAG had two peaks at 12.7 W pump power and M2 = 6.4, the composite Nd:YAG/diamond had a Gaussian profile even at 14.8 W pump power with M2 = 2.0. This indicates the reduction of the thermal-lens effect in the Nd:YAG/dimond composite.
Fig. 7 Beam profiles for the non-composite Nd:YAG laser at the pump power of (a) 1.0, (b) 8.8, and (c) 12.7 W, and for the composite Nd:YAG/diamond laser with its interface AR coated at the pump power of (d) 1.0, (e) 14.8, and (f) 22.0 W, respectively.
5. Conclusion
In conclusion, we have successfully fabricated a Nd:YAG/diamond composite structure with an AR coating at the bonded interface for the first time. This composite laser realizes highly efficient and high-power operation at the same time. The RTB technique can be applied to develop more sophisticated highly efficient and high-power composite lasers. One example is thin disk lasers which have HR mirror coatings at the boded interfaces between laser crystals and diamond. In this case polycrystalline diamond may also be used as the heat spreader since the laser beam does not propagate into the diamond so that the scattering loss inside the polycrystalline diamond, if any, does not affect the efficiency.
Funding
Nippon Sheet Glass Foundation for Materials Science and Engineering; Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Number JP17K05081.
References and links
1. W. Koechner, Solid-State Laser Engineering, 6th ed. (Springer, 2006) Chap. 7.
2. F. Hanson, “Improved laser performance at 946 and 473 nm from a composite Nd:Y3Al5O12 rod,” Appl. Phys. Lett. 66, 3549–3551 (1995). [CrossRef]
3. E. C. Honea, R. J. Beach, S. B. Sutton, J. A. Speth, S. C. Mitchell, J. A. Skidmore, M. A. Emanuel, and S. A. Payne, “115-W Tm:YAG Diode-Pumped Solid-State Laser,” IEEE J. Quantum Electron. 33, 1592–1600 (1997). [CrossRef]
4. R. Zhou, E. Li, H. Li, P. Wang, and J. Yao, “Continuous-wave, 15.2W diode-end-pumped Nd:YAG laser operating at 946 nm,” Opt. Lett. 31, 1869–1871 (2006). [CrossRef] [PubMed]
5. H. C. Lee, P. L. Browlie, H. E. Meissner, and E. C. Rea, “Diffusion bonded composites of YAG single crystals,” Proc. SPIE 1624, 2–12 (1991). [CrossRef]
6. A. Sugiyama, H. Fukuyama, T. Sasuga, T. Arisawa, and H. Takuma, “Direct bonding of Ti:sapphire laser crystals,” Appl. Opt. 37, 2407–2410 (1998). [CrossRef]
7. I. Mukhin, E. Perevezentsev, and O. Palashov, “Fabrication of composite laser elements by a new thermal diffusion bonding method,” Opt. Mater. Express 4, 266–271 (2014). [CrossRef]
8. M. Tsunekane and T. Taira, “300 W continuous-wave operation of a diode edge-pumped, hybrid composite Yb:YAG microchip laser,” Opt. Lett. 31, 2003–2005 (2006). [CrossRef] [PubMed]
9. M. Tsunekane and T. Taira, “High-power operation of diode edge-pumped, composite all-ceramic Yb: Y3Al5O12 microchip laser,” Appl. Phys. Lett. 90, 121101 (2007). [CrossRef]
10. T. Suga, Y. Takagi, B. Gibbesch, and G. Elssner, “Structure of AL-AL and Al-Si3N4 interfaces bonded at room temperature by means of the surface activation method,” Acta Metall. Master. 40, S133–S137 (1992). [CrossRef]
11. H. Takagi, K. Kikuchi, R. Maeda, T. R. Chung, and T. Suga, “Surface activated bonding of silicon wafers at room temperature,” Appl. Phys. Lett. 68, 2222–2224 (1996). [CrossRef]
12. T. R. Chung, L. Yang, N. Hosoda, and T. Suga, “Room temperature GaAs-Si and InP-Si wafer direct bonding by the surface activated bonding method,” Nucl. Instrum. Methods Phys. Res., Sect. B 121, 203–206 (1997). [CrossRef]
13. K. Hara, S. Matsumoto, T. Onda, W. Nagashima, and I. Shoji, “Efficient Ultraviolet Second-Harmonic Generation from a Walk-Off-Compensating β-BaB2O4 Device with a New Structure Fabricated by Room-Temperature Bonding,” Appl. Phys. Express 5, 052201 (2012). [CrossRef]
14. T. Kubota, H. Atarashi, and I. Shoji, “Fabrication of quasi-phase-matching stacks of GaAs plates using a new technique: room-temperature bonding,” Opt. Mater. Express 7, 932–938 (2017). [CrossRef]
15. K. Takayanagi, K. Hara, T. Ishikawa, K. Imura, and I. Shoji, “Fabrication of composite Yb:YAG lasers by use of the room-temperature-bonding technique,” Tech. Dig. Advanced Solid-State Photonics 2011, paper AMB22.
16. I. Shoji, T. Ishikawa, T. Yamauchi, K. Hara, and S. Matsumoto, “Fabrication and characterization of room-temperature-bonded composite lasers,” Tech. Dig. Conference on Lasers and Electro-Optics Europe 2013, paper CE-6.4.
17. P. Millar, A. J. Kemp, and D. Burns, “Power scaling of Nd:YVO4 and Nd:GdVO4 disk lasers using synthetic diamond as a heat spreader,” Opt. Lett. 34, 782–784 (2009). [CrossRef] [PubMed]
18. I. Shoji, Y. Sato, S. Kurimura, V. Lupei, T. Taira, A. Ikesue, and K. Yoshida, “Thermal-birefringence-induced depolarization in Nd:YAG ceramics,” Opt. Lett. 27, 234–236 (2002). [CrossRef]
