1. Introduction

976-nm-diode-pumped Er³⁺/Yb³⁺ 1.55 μm laser is eye-safe and can be used in many applications, such as lidar, satellite laser ranging, three-dimensional imaging, and environmental sensing [1–3]. Among numerous gain media for 1.55 μm lasers [3], Er³⁺:Yb³⁺:RAl₃(BO₃)₄ (Er:Yb:RAB, R = Y, Gd and Lu) crystals have been demonstrated to be excellent ones, due to their low upconversion loss, efficient energy transfer from Yb³⁺ to Er³⁺, and high thermal conductivity [4–6].

Due to the existence of quantum defects caused by energy differences between pump and laser photons, as well as multi-phonon nonradiative transitions between some related multiplets of the doped rare earth ions, pump-induced heat is inevitably generated in the gain medium of diode-pumped solid-state lasers [7]. For a 976-nm-diode-pumped Er³⁺/Yb³⁺ 1.55 μm laser, its quantum defect is 37% and larger than those for a 807-nm-diode-pumped Nd³⁺ 1.06 μm laser (24%) and for a 976-nm-diode-pumped Yb³⁺ 1.04 μm laser (6%), respectively. Furthermore, approximately 7% fluorescence quantum efficiency of the upper laser level for the Er:Yb:RAB crystals is considerably lower than those of the Nd³⁺:YAG and Yb³⁺:YAG crystals (about 90%) [4,5,7], which implies that multi-phonon nonradiative transitions in the Er:Yb:RAB crystals are stronger. Therefore, a large amount of pump power is converted to heat for the 976-nm-diode-pumped Er:Yb:RAB 1.55 μm laser. Considering the narrower splitting of the lower laser level (308 cm⁻¹ of the ⁴I_15/2 multiplet of Er³⁺ in the YAB crystal versus 785 cm⁻¹ of the ²F_7/2 multiplet of Yb³⁺ in the YAG crystal) [8,9], the influence of thermal effects on the quasi-three-level Er:Yb:RAB 1.55 μm laser becomes a serious issue. Thus, the 1.55 μm laser performance can be significantly enhanced when thermal effects in the Er:Yb:RAB crystals are reduced by the adoption of some cooling scheme, such as through the close contact of an Er:Yb:RAB laser crystal with a sapphire crystal [6,10].

A composite crystal can be fabricated by thermal diffusion bonding of a rare earth ion doped crystal to a pure crystal. Using such a composite crystal as a gain medium has become an effective method to reduce thermal effects, and thus improve laser performance [11,12]. Compared to the doped crystal, the pure crystal has a higher thermal conductivity, which allows it to work as an excellent heat sink to dissipate the heat generated in the pump region of the gain medium. The investigation has shown that the peak temperature and thermal stress in a diffusion-bonded Nd³⁺:YAG/YAG composite rod can be dramatically decreased to less than 70% and 60%, respectively, of those in a Nd³⁺:YAG rod [13]. At present, diffusion-bonded Nd³⁺:YAG/YAG, Nd³⁺:YAG/Cr⁴⁺:YAG, and Nd³⁺:YVO₄/YVO₄ composite crystals are commercialized. Furthermore, some novel diffusion-bonded materials, such as Yb³⁺:KY(WO₄)₂/KY(WO₄)₂, Nd³⁺:YVO₄/Nd³⁺:GdVO₄, and Er³⁺:Yb³⁺:glass/Co²⁺:MgAl₂O₄, are fabricated and investigated [14–16]. In this work, we report on the first successful fabrication of an Er:Yb:YAB/YAB composite crystal through the thermal diffusion bonding method, as well as on its 1.55 μm continuous-wave (cw) laser performance.

2. Material fabrication and characterization

A 0.4-mm-thick, c-cut Er(1.1 at.%):Yb(25 at.%):YAB crystal and a 1.0-mm-thick, c-cut YAB crystal, both with the same cross section of 5 × 5 mm², were cut from single crystals. The front and rear endfaces of both crystals were precisely polished to achieve a high surface quality with scratch/dig specification of 10/5 (MIL-PRF-13830B), flatness of less than one-eighth wave at 633 nm, and parallelism of better than 10 arcsec. Then, one endface of each crystal was cleaned and close optical contact was formed through precise alignment of the crystalline a axis direction of both components. No inorganic or organic bonding aids or adhesives were used. Finally, after heat treatment at an elevated temperature, a high quality Er:Yb:YAB/YAB composite crystal was successfully fabricated by the thermal diffusion bonding method. A photo of the composite crystal is shown in Fig. 1(a), where the direction of the arrow indicates the Er:Yb:YAB gain medium. Another 0.4-mm-thick, c-cut Er:Yb:YAB crystal with a same cross section was cut from the same as-grown crystal and polished for comparative studies.

 

Fig. 1 (a) Transmission spectra in the range of 250–850 nm of the Er:Yb:YAB/YAB composite and Er:Yb:YAB crystals. Photo of the Er:Yb:YAB/YAB composite crystal is also shown. (b) Absorption coefficient spectra in the range of 875–1075 nm of both crystals.

Download Full Size | PDF

Transmission spectra in the range of 250–850 nm of both the Er:Yb:YAB/YAB composite and Er:Yb:YAB crystals were recorded by an UV-VIS-NIR spectrophotometer (Lambda-950, Perkin-Elmer) and are also shown in Fig. 1(a). Transmission characteristics of both crystals are similar in the 500–850 nm range, and transmissivities within this waveband are 84–85%, which is close to the theoretical limiting transmissivity of 86% originating from the interface reflection between the air and crystal [17], when the refractive index 1.75 of the YAB crystal is taken into account [18]. However, the transmissivity of the composite crystal is slightly lower than that of the Er:Yb:YAB crystal at wavelengths shorter than 450 nm, and the UV absorption edge of the composite crystal is red-shifted to ~300 nm. This implies that the slight lattice mismatch at the interface between the Er:Yb:YAB and YAB crystals still exists. Absorption spectra in the range of 875–1075 nm of both crystals were also recorded and are shown in Fig. 1(b). The absorption characteristics at the pump waveband of both crystals are similar, and the peak absorption coefficients at 976 nm are about 40 cm⁻¹. Thus, 80% of the incident pump power can be absorbed by a 0.4-mm-thick Er:Yb:YAB gain medium, which makes it suitable for using as a microchip medium.

3. Arrangement of laser experiment

An end-pumped linear resonator was adopted and the setup is depicted in Fig. 2. The uncoated Er:Yb:YAB/YAB composite crystal was mounted in a copper holder, which was cooled by water at 20 °C. There is a hole with radius of 1 mm in the center of the holder to permit the passing of laser beams. A cw 976 nm fiber-coupled laser diode (LD) with a core diameter of 100 μm (Dilas Inc.) was used as the pumping source. After passing a telescopic lens system (TLS) consisting of two convex lenses with the same focal length of 45 mm, a pump laser beam with a diameter of 100 µm was focused in the gain medium. An input mirror (IM) with 90% transmissivity at 976 nm and 99.8% reflectivity in 1.5-1.6 μm was used. Three output mirrors (OMs) with the same radius of curvature (100 mm) but different transmissivities (0.6%, 1.8% and 2.6%) in 1.5-1.6 µm were used. The resonator length was close to 100 mm. Laser spectrum was recorded by a monochromator (Triax550, Jobin-Yvon) with a Ge detector. Using a convex lens with a focal length of 50 cm to focus the output beam, the spatial profile of the focused beam at various distances from the focusing lens was recorded with a Pyrocam III camera (Ophir Optronics Ltd.). The beam radius was calculated by the 4-sigma method and the beam quality factor M² can be estimated by fitting these data to the Gaussian beam propagation expression [19].

 

Fig. 2 Experimental setup of the cw 976-nm-diode-pumped 1.55 μm Er:Yb:YAB/YAB laser.

Download Full Size | PDF

4. Results and discussion

Figure 3(a) shows the output power realized in the composite crystal versus absorbed pump power for different OM transmissivities T. When T was 1.8%, a maximum output power of 780 mW was obtained at an absorbed pump power of 3.14 W, where the slope efficiency η was 28%. The absorbed pump threshold was 0.37 W. The spectra of output lasers are shown in Fig. 3(b). Spectrum at a certain experimental condition was repeatedly recorded and the positions of spectral lines were generally fixed, although the intensity ratios between lines may be changed. The laser wavelength was blue-shifted from 1.55 to 1.52 μm when T was increased from 0.6% to 2.6%. This phenomenon is attributed to the gain characteristic of the Er³⁺/Yb³⁺ 1.55 μm laser with quasi-three-level [4–6]. For the Er:Yb:YAB crystal, the wavelength with maximal gain is blue-shifted from 1.602 to 1.522 μm with increment of inversion density (or intracavity losses) [20]. However, 1.602 μm laser was not observed in this experiment, which may be caused by the high intracavity losses of the Er:Yb:YAB/YAB laser. A nearly circular output beam with ellipticity of 0.96 was observed. The beam quality factors M² in the horizontal and vertical directions were similar and about 1.42 at an absorbed pump power of 3.14 W and OM transmissivity of 1.8%, as shown in Fig. 3(c).

 

Fig. 3 (a) Output power realized in the Er:Yb:YAB/YAB composite crystal as a function of absorbed pump power. (b) Laser spectra for different OM transmissivities T at an absorbed pump power of 3.14 W. (c) Squared beam radius ω² of the output laser as a function of the distance Z from the focusing lens at an absorbed pump power of 3.14 W.

Download Full Size | PDF

Under identical experimental conditions, the laser performance of a lone 0.4-mm-thick, c-cut Er:Yb:YAB crystal with the same doping concentrations was also investigated. For the 1.8% OM transmissivity, output powers of both the Er:Yb:YAB/YAB composite and lone Er:Yb:YAB crystals as a function of absorbed pump power are compared in Fig. 4. Upon increasing the pump power, saturation of the output power was observed, and a maximum output power of 350 mW was obtained in the lone Er:Yb:YAB crystal at an absorbed pump power of 2.0 W. The data demonstrate that thermal effects in the composite crystal were effectively reduced, giving rise to an enhanced output laser performance. The room-temperature thermal conductivity of YAB crystal has been measured to be 11–12 Wm⁻¹K⁻¹ [21], and is higher than that of the Er:Yb:YAB crystal (4.7 Wm⁻¹K⁻¹). Therefore, in the composite crystal, the YAB crystal works as an effective heat sink to dissipate the heat generated in the Er:Yb:YAB gain medium. By comparing the pump thresholds at different OM transmissivities [22], optical losses of the Er:Yb:YAB/YAB composite and lone Er:Yb:YAB crystals originating from defects, impurities, and reabsorption, etc., were estimated to be 1.6% and 0.7%, respectively. Thus, combined with the lower transmissivity at wavelengths shorter than 450 nm shown in Fig. 1(a), the optical quality of the composite crystal needs to be further improved. For improving the optical quality of the composite crystal, the alignment of the crystalline a axis direction of the Er:Yb:YAB and YAB components must be more precise and the heat treatment temperature in the bonding process needs be further optimized. Furthermore, better laser performance (output power of 1 W and slope efficiency of 35%) of a lone Er:Yb:YAB crystal has been reported in [4]. These discrepancies may be due to differences in experimental conditions, such as the design of the resonator structure, rare earth doping concentration, crystal thickness, and the optical quality of the gain medium. It can also be expected that when the Er³⁺ and Yb³⁺ doping concentrations reported in [4] were adopted in the composite crystal in the future, the performance of the Er:Yb:YAB/YAB laser may be enhanced.

 

Fig. 4 Output power as a function of absorbed pump power at OM transmissivity of 1.8% when the Er:Yb:YAB/YAB composite crystal and a lone Er:Yb:YAB crystal are used as gain media under identical experimental conditions, respectively.

Download Full Size | PDF

When a flat OM made of a 1.0-mm-thick sapphire crystal with a cross section of 5 × 5 mm² was closely contacted with the composite crystal through the use of screws, a micro-laser was constructed and the relevant schematic diagram is shown in Fig. 5(a). The OM has a transmissivity of 2% at 1.52 μm. The resonator length was close to 3 mm. End-pumped by the 976 nm LD, a maximum output power of 1.04 W was obtained at an absorbed pump power of 3.14 W, where the slope efficiency was 38%, as shown in Fig. 5(b). The laser wavelength was 1.52 μm. The spacing between the adjacent spectral lines is about 0.62 nm and in agreement with the theoretical line spacing ∆λ (0.66 nm) caused by the etalon effect associated to the 1.0-mm-thick YAB crystal, which is calculated by Δλ = λ²/2nL [22]. Here λ is the laser wavelength, n is refractive index (1.75) of the crystal at λ, and L is the thickness of the crystal. At an absorbed pump power of 3.14 W, a nearly circular output beam with a M² value of 1.51 was obtained, as shown in Fig. 5(c). As the thermal diffusion bonding of an Er:Yb:YAB crystal to a sapphire crystal is difficult, using a double-end diffusion bonded YAB/Er:Yb:YAB/YAB composite crystal as a microchip gain medium may be an alternative method to further enhance the laser performance of the Er:Yb:YAB crystal.

 

Fig. 5 (a) Experimental setup of the 976-nm-diode-pumped 1.52 μm Er:Yb:YAB/YAB micro-laser. (b) Output power realized in the micro-laser as a function of absorbed pump power. The inset shows the spectrum of the micro-laser. (c) Squared beam radius ω² of the output laser as a function of the distance Z from the focusing lens at an absorbed pump power of 3.14 W.

Download Full Size | PDF

5. Conclusion

An Er:Yb:YAB/YAB composite crystal was successfully fabricated by the thermal diffusion bonding method. Although a slight lattice mismatch at the interface between the Er:Yb:YAB and YAB crystals still exists in the composite crystal, a higher output power can be realized in the composite crystal than in the lone Er:Yb:YAB crystal, which is attributed to the YAB crystal in the composite crystal acting as an effective heat sink to reduce the thermal effects in the Er:Yb:YAB gain medium. In order to further enhance the 1.55 μm laser performance, the thicknesses of the Er:Yb:YAB gain medium and the YAB crystal in the composite crystal, as well as the optical quality of the composite crystal, will be optimized in the future.