1. Introduction

High-power diode-pumped solid-state lasers are widely used in scientific research and industrial applications. For high-power solid-state lasers, thin and wide laser gain media are preferred. A thin shape is suited for efficient cooling and reducing thermal effects and a wide area is required for mitigating laser-induced damage. For example, active-mirror lasers [1,2], multi-slab lasers [3,4], and thin-disk lasers [5,6] have been studied as high-power solid-state lasers. However, one of the problems with these designs is that parasitic oscillation (PL) and amplified spontaneous emission (ASE) are more likely to arise. The thin shape leads to a short gain length along the laser amplification and a wide area means a long gain length which is oriented transverse to the laser amplification. Thus, the gain of PL or ASE may exceed that of the laser amplification, in which case the depopulation loss would be large and the laser gain would be small. Reducing the depopulation losses caused by PL or ASE is an important design factor [7].

To tackle these problems, the use of an absorption material as cladding has been proposed [3,8–10]. A cladding that has absorption at the laser wavelength is placed around the gain material so as to absorb the ASE. The refractive index of the cladding material must match that of the gain material for low reflection at the boundary. That means the host material for cladding must be the same as the gain material.

Recently, ceramic materials such as Yb-doped YAG and Nd-doped YAG have been used more commonly for solid-state lasers [1–3,9,11–13]. Ceramic materials offer larger sizes at lower cost compared to single crystals, while at the same time performing as well as single crystals [13]. Thus, ceramic materials are suitable for high-power lasers as well as for cladding. Another advantage is that composite ceramics, in which the gain media and cladding are bonded together, can be made. The mechanical properties of composite ceramics are the same as those of non-composite ceramics [14]. Therefore, we focus on ceramic cladding for high-power lasers, such as Yb:YAG and Nd:YAG.

Such lasers are used at various temperatures, including low or cryogenic temperatures. Low-temperature operations offer certain advantages, such as higher thermal conductivity and lower reabsorption losses [15,16].

Present-day high-power lasers such as DiPOLE [4] employ ceramic materials under low-temperature operation. In order to utilize the cladding technique in such high-power lasers, the absorption characteristics of the ceramic materials used must be known over a wide temperature range.

The characteristics of single-crystalline Co:YIG, Co:YAG, and Co:YGG at low and room temperature are reported in [17]. The absorption temperature dependence of Cr:YAG and V:YAG was first reported in [18], while it is not clear whether those samples were made of ceramics. For ceramic YAG materials, the absorption temperature dependence of Sm:YAG was reported in [19]. An earlier work [8] measured the absorption of Cr:YAG, Sm:YAG, and Co:YAG ceramics at room temperature, but no temperature dependence was reported.

In this paper, we measured and assessed the performance of YAG ceramic absorbers doped with Cr⁴⁺, Co³⁺, Co²⁺, and Sm³⁺ under various temperature conditions. The specifications of the samples are listed in Table 1. All samples were made by Konoshima Chemical. The doping concentration of each sample is same as that in the report by the manufacturer [8]. The doping concentration is determined by checking components of the acid-dissolved ceramics compact, which is done by the manufacturer. According to the manufacturer, uncertainty for the doping concentration is ± 1% of each concentration. The valence of doping ion affects the absorption spectrum as seen in [8] (Cr⁴⁺/Cr³⁺ and Co³⁺/Co²⁺) or in [17] (Co³⁺/Co²⁺). As to our samples, the valence of the doping ion was primarily controlled by choosing the appropriate sintering aids [8]. To the best of our knowledge, the present work is the first to report the temperature dependence or robustness of Co-doped YAG ceramic absorbers.

Table 1. Specifications of Samples

View Table

2. Experimental setup

We measured the temperature dependence of the absorption coefficient as follows. We placed a collimated white light source (400–1800 nm) into a cryostat. We used an AQ4303B (ANDO) as the light source, coupled into a multi-mode fiber (SI, NA0.22, core 50 μm), and collimated with an achromatic lens (f = 10 mm). The transmitted beam was coupled to an optical fiber, and its spectrum was measured with an optical spectrum analyzer (ADVANTEST Q8381A). The resolution setting of the optical spectrum analyzer was 5 nm for full-scale measurements and 0.5 nm for high-resolution measurements. The number of averaging counts was set to 50. We checked the optical spectrum analyzer with an AR-1 Argon Calibration Light Source (Ocean Optics). The maximum wavelength error between 900 nm and 1150 nm was 0.12 nm. We calculated the absorption coefficient from the transmittance of the beam with and without the sample. The sample was bonded to a copper holder in the cryostat. To cool the cryostat, a Gifford-McMahon (GM) cooler, capable of bringing the temperature down to as low as 4 K, was installed on top of the cryostat. The holder temperature was controlled in the 25 to 300 K range by a temperature controller (LakeShore Model 331). The temperature fluctuation was typically less than 0.5 K. To prevent bedewing, we drew a vacuum on the cryostat to a pressure of about 1 Pa (abs.).

3. Results and discussion

3.1 Absorption coefficient

We measured the absorption spectra of Cr⁴⁺, Co³⁺, Co²⁺, and Sm³⁺ YAG ceramics (Table 1) and calculated the absorption coefficients according to the Lambert-Beer law. The results are shown in Fig. 1—Fig. 4. Since the wavelength range of interest for high-power lasers was around 1 μm, we measured the spectrum around that range at a higher resolution (0.5 nm), shown as (b) in each figure. As for Co²⁺, a higher-resolution measurement was not conducted since it had almost no absorption around 1 μm. The figure (b) of Sm³⁺ is complicated, thus we simplified the figure (b) to make another figure (c). For technical reasons, some of the absorption coefficients saturated at around 15 cm⁻¹ in Fig. 1 and Fig. 4. In Fig. 2, the measurement around 1700 nm was not stable and deleted. Because the optical spectrum analyzer we used changes its detector automatically at a wavelength of 1040 nm, there is some discontinuity at that wavelength. All spectral shapes at room temperature were similar to the results in [8].

 

Fig. 1 Absorption spectrum of Cr⁴⁺:YAG.

Download Full Size | PDF

 

Fig. 2 Absorption spectrum of Co³⁺:YAG.

Download Full Size | PDF

Generally, the absorption spectrum depends on the temperature of the material. Thus, the absorption at a given wavelength can be sensitive to temperature in some case. In the case of Cr⁴⁺, for example, the absorption coefficient at around 1000 nm (1150 nm) increased (decreased) as the temperature decreased. That means the absorption at these wavelengths is sensitive to temperature. On the other hand, the coefficient at around 900 nm changed little as the temperature decreased. That indicates the absorption at this wavelength is less sensitive, or more stable in other words, to the temperature. As we discuss later, we think the stability is beneficial to cladding use.

3.2 ASE absorber performance

In this section, we discuss the performance of the absorber materials. For an absorber, one critical requisite is that the material have a sufficiently high absorption coefficient at the laser wavelength. By contrast, a low absorption coefficient is required at the pump wavelength for the side pumping layout [8]; however, it depends on the laser layout. Thus, we put no limitation on the value of the absorption coefficient at the pump wavelength in this study. Another important point is the temperature dependence of the absorption. If the absorption of a cladding material is sensitive to temperature changes, a laser with that cladding cannot be used over a wide temperature range. In addition, a weak temperature dependence or robustness would be helpful in designing laser media with cladding. ASE absorption induces heat losses and increases the temperature of the absorber itself. Thus, if the absorption coefficient were to change dramatically, the absorptance of ASE would be affected.

We determined the absorption coefficients and their temperature dependence at wavelengths of 1030 nm and 1064 nm. During this process, we measured the absorption spectra with a resolution of 0.5 nm. First, we investigated whether the specimens have enough absorption. At 1030 nm, Co²⁺:YAG and Sm³⁺:YAG show almost no absorption and thus cannot be used for cladding. The other two specimens have absorption. The room-temperature coefficients for Cr⁴⁺:YAG and Co³⁺:YAG are ~4.2 cm⁻¹ and ~1.9 cm⁻¹, respectively. However, this does not necessarily mean that Cr has greater absorption, because the concentrations of the samples differ by about one order of magnitude. Higher doping densities would degrade the thermal conductivity [20]; thus, a low concentration is desirable. We evaluated only one concentration due to availability. According to the manufacturer, lower concentration than that of our samples can be made without difficulty, while much higher concentration ones could be difficult to produce with no loss of quality. We expect that if we had used a lower concentration sample, we would have a result of proportionally decreased absorption coefficient.

The characteristics of Sm³⁺:YAG change at 1064 nm. While Co²⁺:YAG shows no absorption at 1064 nm, either, Sm³⁺:YAG exhibits an absorption of ~2.9 cm⁻¹ at room temperature, which is comparable to the value of 2.8 cm⁻¹ in [19]. Thus, Sm³⁺:YAG is used as cladding in the Nd:YAG laser [10].

It depends on laser scheme what level of absorption is needed for suppressing ASE. In addition to the performance of the cladding material, parameters of the laser gain medium such as small signal gain, gain length, shape, laser material lifetime will affect that. A computational work [21] on this, for example, shows that they can suppress ASE with Cr⁴⁺:YAG layers, predicting maximum output energy of over 1 kJ. Their optimization result describes that initial absorption coefficient of Cr⁴⁺ is equivalent to 2.09 cm⁻¹. The results of our samples (~4.2 cm⁻¹ in Cr⁴⁺:YAG, ~1.9 cm⁻¹ in Co³⁺:YAG) are above or close to this value. Thus we think our material can be used for this laser scheme (the absorption coefficient of Cr⁴⁺ can be optimized easily by changing doping concentration).

Next, we discuss the temperature dependence. We compared the absorption coefficient at different temperatures to that at 300 K. The results are shown in Fig. 5. The red dashed line with open squares shows the temperature dependence in [18]. We interpolated the value at 300 K for comparison. The green dashed line with open triangles represents the values taken from [19]. The other lines with solid symbols are measured data obtained in the present work.

At 1030 nm, Cr⁴⁺ has a negative dependence on temperature. The temperature dependence of Cr⁴⁺ shows a similar trend to that in [18]. Co³⁺ exhibits a weaker temperature dependence than Cr⁴⁺. At 1064 nm, both Cr⁴⁺ and Co³⁺ show weaker temperature dependences than Sm³⁺. The behavior of Cr⁴⁺ in the present work is not identical, but similar, to that in [18].

The present behavior of Sm³⁺, however, clearly differs from that in [19], while both coefficients are nearly the same at room temperature. This may be because of the difference between the two materials. When a sample is cooled to low temperature, absorption spectrum becomes narrower and sharper shape as seen in our result (e.g. Fig. 4) as well as in their report (Fig. 2, Fig. 3 in [19]). 1064 nm is at the shorter side of a local absorption peak around 1065.5 nm. Thus characteristics at 1064 nm would be significantly affected by narrowing effect of the peak as the temperature decreases. In this case, their sample shows narrower spectrum at lower temperature than our sample, making their sample is more temperature dependent than ours. In their work, they used ceramics Sm³⁺:YAG produced by solid-state synthesis [19] while our sample is made by wet chemical method [22]. We think the difference in fabrication process affected the characteristics at low temperature.

 

Fig. 3 Absorption spectrum of Co²⁺:YAG.

Download Full Size | PDF

 

Fig. 4 Absorption spectrum of Sm³⁺:YAG.

Download Full Size | PDF

In summary, Co³⁺ is less temperature-dependent than Cr⁴⁺ for 1030-nm absorber use. Co³⁺ and Cr⁴⁺ are less thermally dependent than Sm³⁺ at 1064 nm. Co²⁺ cannot be used as an absorber at these wavelengths.

 

Fig. 5 Temperature dependence of absorption coefficient at (a) 1030 nm and (b) 1064 nm.

Download Full Size | PDF

4. Conclusion

We determined the temperature dependence of Cr⁴⁺:YAG, Co³⁺:YAG, Co²⁺:YAG, and Sm³⁺:YAG ceramics as ASE absorbers from 25 K to 300 K. Cr⁴⁺ and Co³⁺ have absorption and can be used for 1030-nm laser absorbers. In addition to Cr⁴⁺ and Co³⁺, Sm³⁺ can be used for 1064 nm; however, its performance declines at lower temperatures.

Compared to Cr⁴⁺:YAG, Co³⁺:YAG exhibited higher temperature robustness, i.e., a smaller variation in absorption coefficient with temperature. This means that the changes in the ASE absorber temperature, due to absorbed light converted into waste heat, would have a smaller impact in the case of Co³⁺:YAG and that designing such absorbers would be easier than in the case of Cr⁴⁺:YAG.