1. Introduction

Increasing attention has been focused on Yb³⁺-based laser systems since the rapid development of high power and high brightness laser diodes emitting at 900-980-nm in latter 1990s. Compared to Nd³⁺-doped materials, Yb³⁺-doped ones have broader absorption and emission spectra owing to the strong electron-phonon coupling [1]. In addition, the only two electronic multiplets of Yb³⁺ (the ground state ²F_7/2 and the excited state ²F_5/2) give rise to a simple electronic-level scheme, and contribute to a low intrinsic quantum defect, a weak thermal load, an absence of luminescence quenching, and an enhanced laser action. Also, the longer radiative lifetimes of the upper laser manifolds of Yb³⁺-doped materials, i.e. increased energy-storage property, are favorable to enhancing the economic utilization of the diode pumps and hence the development of high-energy diode pumped sources. Therefore, Yb³⁺-based laser systems have been expected to be the most potential alternatives to the Nd³⁺-doped ones in the near-IR spectral range.

A variety of interesting results have been reported for cw or mode-locked operations based on the diode pump with Yb³⁺-doped materials during the past decade, such as garnet Yb:YAG [2,3], vanadate Yb:YVO⁴ [4], tungstates Yb:KGW and Yb:KYW [5–7], oxyorthosilicates Yb:LSO and Yb:YSO [8,9], and fluorides Yb:CaF₂ and Yb,Na:CaF₂ [10–12]. However, a main drawback of current Yb³⁺-doped materials results from its quasi-three-level operating scheme. In the three-level system, the pumping rate must exceed a minimum value before any inversion at all can be obtained [13], in addition, the crystal has some reabsorption at the operating wavelengths due to the thermal populating of the terminal laser level, which result in a high pumping threshold power. Therefore, a potential Yb³⁺ doped laser medium with sufficiently large splitting of the fundamental manifold ²F_7/2 is pursued for researchers. Recently, efficient mode-locked laser actions of a new Yb³⁺-doped oxyorthosilicate crystal with a large splitting of 1067 cm^-1, Yb:GSO, was described, and cw laser slope efficiency of 46% was achieved [14,15]. In this paper, we present a new material, Yb:GYSO, for the purpose of eliminating cleavage of Yb:GSO, and demonstrate its cw tunable laser action. The cw tunable laser performance of Yb:GYSO was studied for the first time, and high slope efficiency was achieved.

2. Properties of Yb:GYSO

Yb:GSO exhibits lots of superiorities in laser operation, however, the examinations by X-ray diffractometer (XRD) (Cu target, Kα) show that the Yb:GSO crystal still maintains the primitive monoclinic structure with space group of P2₁/c. It is composed of a two-dimensional network of corner linked (OGd₄) tetrahedra into which the (SiO₄) tetrahedra are packed. Although the property of cleavage can be effectively mitigated by adjusting the orientation of GSO seed during the growth, the relatively weak bonding between these layers is still prone to cleavage along the (100) plane to some extent. In comparison, the crystal structure of YSO belongs to the monoclinic C2/c space group. Hereby, the (SiO₄) and (OY₄) tetrahedra share edges and form chains interconnected by isolated (SiO₄) tetrahedra, and this arrangement results in a more rigid and isotropic structure [16]. Analogous alloyed crystals, such as Ce:LGSO and Ce:GYSO [17,18] have been grown successfully without exhibiting significant cleavage, and good scintillation properties were demonstrated. We can speculate that Yb-doped GSO alloyed with YSO may avoid cleavage, unlike the pure Yb:GSO crystal, and combine the excellent laser performance of Yb:GSO with good mechanical properties of Yb:YSO. To our knowledge, this is the first report on Yb:GYSO crystal until now.

 

Fig. 1 The room-temperature unpolarized absorption and emission spectra of Yb:GYSO laser crystal.

Download Full Size | PDF

The alloyed Yb:GYSO crystal was grown by Czochraski method from a 50/50 solution of GSO and YSO with the doping level of 5-at.% Yb³⁺. The pulling rate was 2.5 mm/h and the rotation rate of the seed was 30 rpm. The Yb:GYSO boule was up to 70mm in length and 35 mm in diameter. The Yb:GYSO crystal is smooth and crystallographically perfect, which indicates that Yb:GYSO with appropriate contents of Y₂O₃ can be easily grown to have a higher degree of crystalline perfection than Yb:GSO. As expected, no cleavage phenomena emerge and good mechanical properties are exhibited. As mentioned above, the GSO crystal has cleaving property, and is likely to suffer from cracking during the cooling for crystal growth. In addition, new cracks are likely to occur during the machining such as cutting and polishing, thereby disadvantageously lowering the specimen yield. Yb-doped GSO alloyed with 50% YSO exhibits C2/c structure and has no cleaving property, which can prevent the crystal from cracking under thermal or mechanical shocking. The fluorescence lifetime value of the excited manifold ²F_5/2 of Yb³⁺ has measured to 1.92 ms by exciting the samples with a xenon lamp and detected by an S-1 photomultiplier tube. This favorable lifetime could increase the economic utilization of the costly pump power. The splittings of ²F_5/2 and ²F_7/2 for Yb:GYSO are 865cm^-1 and 995 cm^-1, respectively, which is comparable to the splitting of 1067 cm^-1 for ²F_7/2 of Yb:GSO. The unpolarized absorption and emission spectra of the Yb:GYSO along b-axis are shown in Fig. 1. We can observe that the absorption spectrum is mainly composed of four bands around 900, 918, 950 and 976 nm in Fig. 1. Apparently, the absorption peak around 976 nm belongs to the zero-line transition between the lowest levels of ²F_7/2 and ²F_5/2 manifolds, and the broad absorption bandwidth of about 16 nm is suitable for high efficiency diode-pumped operation. The IR fluorescence spectrum of 5-at.% Yb³⁺-doped GYSO crystal excited under the InGaAs LD source with the wavelength of 940nm at room temperature is also presented in Fig. 1. The broad fluorescence curve mainly includes four bands around 1004, 1039, 1056, and 1080 nm, corresponding to the transitions from the lowest level of ²F_5/2 manifold to the other levels in the ²F_7/2 manifold except the lowest. Among the four emission bands, the emission band at 1004 nm is relatively high, but the laser performance is inefficient due to the strong reabsorption losses. However, the emission band around 1080 nm is most easily to get efficient laser action because of few-populated terminal laser level.

3. Experiments

The laser experiments were performed with a stable three-mirror folded cavity supporting only TEM₀₀ mode, as shown in Fig. 2. The resonator consisted of one dichroic input coupler M1 (highly transmissive at 976 nm and highly reflective at 1030-1170 nm), one folding mirror M2 (highly transmissive at 976 nm and highly reflective at 1030-1170 nm), and one output coupler (OC). The input mirror M1 and output coupler are both flat, and the curvature radius of the folding mirror M2 is 300 mm. In order to realize the laser operation in TEM₀₀ mode and result in high conversion efficiency, the length of two arms were configured to keep the mode matching in crystal between the pump beam and the fundamental resonant mode. The 5×5×3 mm³ 5-at. % Yb:GYSO sample was wrapped with indium foil and mounted in a water-cooled copper block. The water temperature was maintained at 16°C to prevent thermal fracture, and our sample was uncoated and polished with parallel end faces. A fiber-coupled diode laser with a core-diameter of 200-μm and a numerical aperture of 0.22, emitting at the wavelength range of 975–978 nm was used as the pump source. To keep the maximum absorption, the operating wavelength was tuned by temperature of diode to match the peak absorption of the crystal. The pump beam was imaged relay to the crystal with a ratio of 1:1 and the pump radius in the crystal is ~100-μm.

 

Fig. 2 Configuration of the cw Yb:GYSO laser.

Download Full Size | PDF

4. Results and discussions

For Yb:GYSO laser, output couplers with different transmissions (3%, 8%, 12% and 14%) were used to obtain the optimum output. And the dependence of the laser outputs on the absorbed pump power is illustrated in Fig. 3. At absorbed pump power of 5.54 W, maximum laser output power of 2.44 W at 1081.5 nm without tuning was obtained with a 12% transmission output coupler, and the corresponding threshold was about 1.2 W. Under lasing condition and at maximum power, the uncoated crystal absorbed about 75% of the incident pump power, and the slope efficiency arrived at 57%. Good beam quality was achieved, with measured M² < 1.2 at all pump level. It is worth noting that with a 3% transmission output coupler, the laser threshold was approximately only 0.5 W. As described in [14], large splitting of the fundamental manifold promises laser operation with low pump threshold, for reabsorption losses at emission wavelengths can be decreased due to low thermal population of the terminal laser level. That is to say, the splitting of the ground-state manifold ²F_7/2 of 995 cm-1 in the alloyed Yb:GYSO crystal, which is somewhat larger than that of the Yb:YSO, is more potential to generate a low threshold value. We speculate that the large splitting was due to the introduction of Yb:GSO into Yb:YSO. The explanation in the present work is just qualitative, and more precise should be obtained by further detailed studies in future.

 

Fig. 3. Output power versus absorbed pump power with 5-at.% Yb:GYSO for different output couplers at 1081.5 nm.

Download Full Size | PDF

The wavelength tuning for Yb:GYSO was fulfilled by inserting an SF10 dispersive prism in the collimated arm of the laser cavity. In order to get a reasonable judge of this crystal, a Yb:GSO crystal of the same size (5 at. % doped) was used, instead of Yb:GYSO, in the same cavity. Because the insertion of the prism increased the intracavity losses, we chose the output coupler with 10% transmission to achieve efficient tuning output, other than the optimum output transmission without tuning of 12%. At 4 W absorbed pump power, the Yb:GYSO crystal supported a broad continuous range of 59 nm, from 1030 to 1089 nm. The wavelength tuning of Yb:GYSO is illustrated in Fig. 4. For Yb:GSO, its tuning range extended from 1028 nm to 1093nm for absorbed power of 4.22 W, as shown in the inset of Fig. 4. There is an unfavorable prominent emission peak around 1089 nm in the wavelength tuning curve of Yb:GSO, because the few populated terminal laser level and the strongest emission cross section around 1089 nm exist simultaneously. Flat tuning curve, just like those of its silicate family members Yb:YSO and Yb:LSO [9], must be obtained by enhancing the losses around 1089 nm. Compared with that of Yb:GSO, the wavelength tuning curve of Yb:GYSO is flatter and smoother with fewer fluctuations, and it seems like that of Yb:YSO. This is possibly by reason that some of the gadolinium atoms have been replaced by the yttrium atoms in Yb:GSO. And this result makes more promising for the use of Yb:GYSO in the development of all-solid-state tunable cw and femtosecond lasers. Further tuning on shorter wavelength is limited by the coating of input coupler, which is difficult to maintain high reflection at this range and high antireflection around 980 nm. Therefore, it always resonated at its vicinal emission peak of 1030 nm with worse beam quality when the laser was tuned to shorter wavelength.

 

Fig. 4. Tuning curve of Yb:GYSO with an SF10 dispersive prism and a 10% transmission output coupler. The inset is the wavelength tuning curve for Yb:GSO with transmission output coupler of 10%.

Download Full Size | PDF

Although the study of Yb:GYSO is still in the period of tentative, plenty of its advantages have appeared in experiments. Possessing of appropriate content of Y₂O₃, Yb:GYSO crystal is easily grown with a higher degree of crystalline perfection than Yb:GSO, and it has a more rigid and isotropic structure like that of Yb:YSO. There is no significant cleavage emerged as we expected, which is hard to avoid in the case of pure Yb:GSO crystal. At the same time, good laser properties are demonstrated for Yb:GYSO. Its laser threshold is relatively lower, and it has broad wavelength tuning range with fewer fluctuations than that of Yb:GSO. Although something incomplete existed in our experiments, experiment results indicate that this novel crystal is promising in achieving efficient, low-threshold, broadly tunable diode-pumped lasers, longer upper laser manifold lifetime, and low-cost growth. In a word, Yb:GYSO can integrate the high laser performance of Yb:GSO and good mechanical properties of Yb:YSO simultaneously.

5. Conclusions

To maintain the excellent laser performance of Yb:GSO crystal and improve its imperfect structure, we grew Yb:GYSO crystal and investigated its laser action for the first time. With the 5-at.% Yb:GYSO sample, 2.44 W output power at 1081.5 nm was achieved with a slope efficiency of 57%, and its threshold is relatively low. The laser wavelength of Yb:GYSO crystal could be tuned from 1030nm to 1089 nm continuously. Owing to its smooth and broad wavelength tuning, it is suitable to be used for a mode locked laser operation. Although the study for Yb:GYSO is still tentative, this result confirms that the method of combining the high laser performance of Yb:GSO with good mechanical properties of Yb:YSO simply is effective. Currently we are working on the growth a series of Yb:Gd_2(1-x)Y_2xSiO₅ (x=0~1) crystals and their laser actions. We believe that, as long as x is chosen to be suitable, Yb:Gd_2(1-x)Y_2xSiO₅ could be one of the most excellent laser crystals for achieving high-power diode-pumped broadly tunable cw or mode locked lasers.