The onset of parasitic oscillations limits the extraction efficiency and therefore energy scaling of Q-switched lasers. A solid-state laser was end pumped with a fiber-coupled diode laser and operated in q-cw as well as in passively Q-switched operation. For Q-switched operation, we demonstrate the suppression of parasitic oscillations in a core-doped ceramic Nd:YAG laser by Sm:YAG cladding.
©2010 Optical Society of America
For end pumped lasers polycrystalline ceramic Nd:YAG exhibits many advantages compared to grown single crystals, because it enables a variety of novel rod designs with respect to dopant concentration, dopant gradient, distribution, size and geometry [1–5]. Comparable spectroscopic and thermo-optical properties of ceramic and crystalline material were reported [4, 6–9] as well as the use of core-doped ceramic Nd:YAG laser rods for diode-pumped solid-state lasers [10–12]. For core-doped laser rods, the concentration of the pump light absorption around the rod’s axis leads to a brightness conversion . With the aid of polished barrel surfaces guidance of the pump light due to total reflection is possible. Therefore, core-doped rods seem to be promising candidates as gain media using pump sources with low brightness, e.g. diode stacks.
Because of these advantages, we investigated the appropriateness of core-doped Nd:YAG ceramic laser rods for pulse energy scaling of Q-switched lasers. We compared the behaviour of core-doped Nd:YAG ceramics to a homogeneously doped crystalline Nd:YAG rod in mode selective and divergent longitudinal pumping schemes. The latter concept allows the use of compact, low brightness pump sources such as diode stacks with simple optics consisting of lenses, lens ducts or optical concentrators. First, we investigated the laser performance in q-cw mode, where the pump light is continuously converted into laser light, and secondly in passively Q-switched operation, where the pump energy is stored until pulse emission, to observe the influence of parasitic oscillations on the laser output, which limit further energy scaling. For a divergent pumping scheme, waveguiding in the core-doped laser rod is required for efficient pump light absorption and polished barrel surfaces are necessary which can lead to parasitic oscillations. With the aid of a Sm:YAG cladding for core-doped laser rods, a suppression of parasitic oscillations seems to be possible without noticeable losses of the pump light due to its spectroscopic properties. Sm:YAG has a high absorption at the laser wavelength of 1064 nm, but a high transmission for the pump light of 808 nm . The main focus was to analyze the impact of Sm:YAG cladding for core-doped ceramic Nd:YAG lasers.
2. Experimental setup
We used two composite core-doped laser rods made of polycrystalline ceramic Nd:YAG (Baikowski, USA , ) for our investigations. The cladding of the first rod was undoped YAG, the cladding of the second rod was doped with 4.0 at. % Sm (Table 1 ). The barrel surfaces of the core-doped rods were polished to optical quality in order to act as a waveguide for the pump light.
To analyze the influence of the doped core on the laser performance, a homogeneously doped Nd:YAG crystal with the same outer dimensions was used. The unpolished barrel surface of the homogeneously doped Nd:YAG rod has a surface roughness Ra of 0.3 µm, which does not allow for waveguiding of the pump light, but its rough surface can suppress parasitic oscillations perpendicular to the optical axis due to diffuse reflexions.
In Fig. 1 the experimental setup is shown. As pump source a fiber-coupled q-cw diode laser (DILAS GmbH, N7F-806.7-1000Q-H207) with a a peak output power of 1 kW, a pump pulse duration of 200 µs and a repetition rate of 10 Hz was used. The q-cw diode module with a spectral width of 2.5 nm (FWHM) was temperature stabilized. To maximize pump light absorption in the laser rods, the pump wavelength was matched by temperature tuning of the pump diode module to 806 nm. The fiber tip with 800 µm diameter was imaged into the laser rod, which was mounted in a temperature-controlled holder. The dichroic mirror of the laser oscillator was directly coated on the Nd:YAG rods. The other endface of the rods had an antireflection coating for 1064 nm.
To analyze the laser performance of the rods two different pumping schemes were used. With the first scheme a mode selectively pumping scheme was choosen to compare the behaviour of the core-doped ceramic rods with the homogeneously doped crystal, as the latter allows no waveguiding. This idea is shown as raytrace in Fig. 2a . To investigate the pump light guidance behavior for low brightness pump sources such as diode stacks with simple optics (lenses, lens ducts or optical concentrators) where the laser rod acts as waveguide, we additionally used a divergent incoupling of the pump radiation. Therefore, we used a lens with a short focal length (f = 8.0 mm) and focussed the pump light in front of the rod surface to achieve multiple internal reflexions at the barrel surface (Fig. 2b).
The integrated pump light absorption along the rod’s axis with a fixed pump power, a center wavelength of 806 nm and a spectral width (FWHM) of 2.5 nm of the diode is calculated by the raytrace software Zemax and shown for both pumping schemes in Fig. 3 . The pump light absorption corresponds directly to the integrated fluorescence in the rod and shows for both rod designs, homogeneously doped as well as core-doped, a maximum in the center. For mode selective pumping, there is no difference between the different rod types (Fig. 3a), because nearly no pump light reaches the outer regions of the rods. With a divergent pumping scheme, it is possible to concentrate the maximum gain for highly divergent pumping in the central region around the rod’s axis of the core-doped laser ceramics for brightness conversion . The outer aperture of the core with 2.0 mm of the core-doped ceramic rod can be seen clearly in Fig. 3b (the square insets have a size of 4 mm x 4 mm). The maximum of the integrated pump light absorption of the core-doped rod is significantly higher than that of the homogeneously doped rod. Here, no pump light is absorbed in the outer regions of the rod’s axis (transversal position) due to the undoped YAG cladding.
With the aid of the divergent pump configuration, we investigated the waveguiding behaviour of the polished core-doped rod barrels with respect to the pump light absorption and the suppression of parasitic oscillations. In Fig. 4 the simulated pump light absorption along the optical axis is shown for both rod types, homogeneously doped (a) and core-doped (b) rod, in the divergent pumping scheme (Fig. 2b). A second maximum inside the rod at a distance of about 6 mm can be observed, which is due to the first total reflexion at the barrel surface. The interface between core and clad can be identified quite clear in Fig. 4.
For mode selective pumping nearly no difference between the two rod types is observed. In case of the core-doped rod the pump light absorption in the divergent pumping scheme is concentrated and increased around the rod’s center axis (Fig. 4b) in comparison to the homogeneously doped crystal where most of the pump light is absorbed in the first 2 mm of the rod (Fig. 4a).
3. Experimental results
3.1 Q-cw laser operation
To compare the lasing properties of the different rods, the laser performance was measured with a 35 mm resonator setup and a plane output coupler with optimized transmission to achieve maximum multimode output power. A transmssion of 10% was found to maximize the output energy for all above mentioned rods (Table 1). Both pumping schemes (Fig. 2), the mode selective and the divergent, were used to investigate the performances of the core-doped rods in comparison to the homogeneously doped rod. The measured values at a pump wavelength of 806 nm and a pump duration of 200 µs are listed in Table 2 .
The slopes of the q-cw laser output for the mode selective and the divergent pumping scheme are plotted in Fig. 5 . With the mode selective pumping scheme the optical-to-optical efficiency of 62.7% of the core-doped rod with Sm:YAG cladding is the highest measured, with a peak output power of 636 W respectively, followed by the Nd:YAG crystal with 59.3%. The lasing performances of all rods in the mode selective pumping scheme are nearly the same (Fig. 5).
In the divergent pumping scheme the core-doped ceramic Nd:YAG rod with Sm:YAG cladding has the lowest laser threshold (dashed lines) and also the best lasing performance. Maximum peak output power here was 446 W, which corresponds to an optical-to-optical efficiency of 44%. The laser threshold of the homogeneously doped Nd:YAG crystal is higher due to a lower inversion around the center axis of the rod due to absorption of the pump light in the outer regions (Fig. 3b). The threshold of the core-doped rods is in both pumping schemes nearly the same but the slope of the YAG cladded rod is smaller. This might be an effect of parasitic oscillations perpendicular to the opical axis, which reduce the available inversion for the laser output.
3.2 Q-switched laser operation
To determine the energy storage capacity of the different types of laser rods and to analyze the losses due to parasitic effects, we investigated their performance in Q-switched operation. Thus, we changed the laser setup by inserting a passive Q-switch (Cr4+:YAG) in the cavity comparable to the master oscillator introduced in Ref . Here, a mode selective pumping scheme was chosen to obtain passive Q-switching with a good beam quality. The plano-concave cavity was set to a length of 90 mm. Cr4+:YAG crystals with various initial transmissions were placed as saturable absorber into the cavity and the output coupler transmission was subsequently optimized for each configuration. In Table 3 the pulse energies and optical-to-optical efficiencies for different configurations are listed.
At higher initial transmission (60% and 40%) of the Cr4+:YAG, passive Q-switching was demonstrated with all rods. For the core-doped ceramic Nd:YAG with undoped YAG as cladding media, no pulsed operation was achieved at lower initial transmissions in the configurations with 20% and 10%, which were used for energy scaling (Tab. 3). The threshold for bleaching out the saturable absorber was not reached at a maximum peak pump power of 1000 W (200 mJ for 200 µs pump pulses). In contrast to this, nearly the same output pulse energy of 3.3 mJ was generated with the homogeneously doped Nd:YAG crystal and the core-doped Nd:YAG ceramic with Sm:YAG as cladding media at pump energies around 32 mJ. This corresponds to an optical-to-optical efficiency of around 10% respectively. With an initial Cr4+:YAG transmission of 10%, a pulse duration (FWHM) of 2.6 ns was measured in both setups with a fast Si-PIN photodiode (alphalas, UPD-200-SP: bandwidth > 2.0 GHz and risetime < 175 ps) and a digital oscilloscope (LeCroy, Waverunner 6100 A: 1 GHz bandwidth).
These results show that the use of core-doped ceramic Nd:YAG rods with Sm:YAG as cladding media does not affect the laser performance because it is similar to the homogeneously doped rods.
To elucidate the reasons for the absence of pulsed operation in case of the undoped YAG cladded laser rod and saturable absorbers with low initial transmission, we removed the top of the crystal holder and monitored the fluorescence with an indium gallium arsenide photodiode (Thorlabs, Det410) at the polished barrel surface of the optically pumped rods. To block the pump light, we placed a dichroic mirror in front of the photodiode, which was coated highly transmissive for the laser wavelength of 1064 nm and highly reflective for the pump light at 808 nm. The pump pulse duration was set to 200 µs. The observed time-resolved fluorescence signals of both core-doped ceramic Nd:YAG laser rods are shown in Fig. 6 .
The time-resolved fluorescence signals of the rods with Sm:YAG and undoped YAG cladding are different. With Sm:YAG cladding, the expected exponential energy storage curves were measured (Fig. 6a). In contrast to this, a plateau indicating a steady state was detected for the YAG cladded rods (Fig. 6b). The onset of the plateau occurs earlier with increasing peak pump power. In Fig. 7 the detected fluorescence signals of both ceramic core-doped rods, with undoped YAG cladding and with Sm:YAG cladding, are normalized for direct comparison of the signal shape.
At the end of the 200 µs pump pulse the expected exponential fluorescence decay was observed for the Sm:YAG cladded rod, whereas the YAG cladded rod showed a steeper decay (Fig. 6 and 7). This observed behavior suggests parasitic oscillations in the YAG cladded rod. The slopes in q-cw mode of all rods, where the inversion is continuously depleted, are nearly identical (Fig. 5). But in Q-switched operation, where the stored energy is released at the end of the pump pulse, the performance of the YAG cladded rod differs from the homogeneously doped Nd:YAG crystal and the Sm:YAG cladded Nd:YAG ceramic and shows a decayed behaviour (Tab. 3). With a time-resolved fluorescence measurement shown in Fig. 6 and 7 an onset of a steady-state oscillation was detected in the YAG cladded rod unlike to the expected energy storage curves with the followed exponential fluorescence decay observed with the Sm cladded rod. This parasitic effect prevents energy scaling to a higher laser output.
In comparison to undoped YAG, Sm:YAG as cladding media in core-doped ceramic Nd:YAG rods can suppress the onset of parasitic oscillations. Sm:YAG does not affect the laser performance in q-cw and passively Q-switched mode compared to a homogeneously doped Nd:YAG crystal. Therefore, Sm3+ is a useful doping candidate for a cladding of core-doped rods when using highly divergent pump sources like diode stacks with simple pump optic configurations such as lens ducts and concentrators relying on waveguiding inside the laser rods.
Core-doped laser rods enable a concentration of the absorbed pump light around the rod’s axis and allow the use of pump light sources with low brightness such as diode laser stacks with simple pump optics like lens ducts. A polished barrel surface is needed to act as waveguide for the divergent pump light, but it favors the onset of parasitic oscillations perpendicular to the rod’s axis. Parasitic oscillations cause a lower q-cw laser output power and can even inhibit Q-switched operation. With Sm:YAG cladding, these restrictions of barrel-polished core-doped laser rods can be compensated. As experimentally shown, core-doped Nd:YAG ceramic with Sm:YAG cladding exhibits the same optical-to-optical efficiency as the homogeneously doped Nd:YAG laser crystal of more than 60% in q-cw operation and of 10% in passive Q-switching. Parasitic oscillations, which limit extraction efficiency and energy scaling of lasers are significantly suppressed for Sm:YAG cladding compared to undoped YAG cladding.
In further investigations the pump source will be changed from a fiber-coupled pump diode to a diode stack with a lower brightness. To realize an efficient propagation of the pump light with waveguiding at the polished barrel surfaces and simultanously suppress parasitic oscillations, a core-doped Nd:YAG laser rod with Sm:YAG cladding will be used.
The authors thank the German Research Foundation (DFG) for funding the Cluster of Excellence Centre for Quantum Engineering and Space-Time Research QUEST.
References and links
1. J. Lu, M. Prabhu, J. Song, C. Li, J. Xu, K. Ueda, A. A. Kaminshii, H. Yagi, and T. Yanagitani, “Optical properties and highly efficient laser oscillation of Nd:YAG ceramics,” Appl. Phys. B 71(4), 469–473 (2000). [CrossRef]
2. L. Jianren, T. Murai, K. Takaichi, T. Uematsu, K. Misawa, M. Prabhu, J. Xu, K. Ueda, H. Yagi, T. Yanagitani, A. A. Kaminskii, and A. Kudryashov, “72 W Nd:Y3Al5O12 ceramic laser,” Appl. Phys. Lett. 78(23), 3586–3588 (2001). [CrossRef]
3. L. Jianren, M. Prabhu, X. Jianqiu, K. Ueda, H. Yagi, T. Yanagitani, and A. A. Kaminskii, “Highly efficient 2% Nd:yttrium aluminum garnet ceramic laser,” Appl. Phys. Lett. 77(23), 3707–3709 (2000). [CrossRef]
4. D. Kracht, M. Frede, R. Wilhelm, and C. Fallnich, “Comparison of crystalline and ceramic composite Nd:YAG for high power diode end-pumping,” Opt. Express 13(16), 6212–6216 (2005). [CrossRef] [PubMed]
5. Y. Qi, X. Zhu, Q. Lou, J. Ji, J. Dong, and Y. Wei, “Nd:YAG ceramic laser obtained high slope-efficiency of 62% in high power applications,” Opt. Express 13(22), 8725–8729 (2005). [CrossRef] [PubMed]
6. J. B. Gruber, D. K. Sardar, R. M. Yow, T. H. Allik, and B. Zandi, “Energy-level structure and spectral analysis of Nd3+ (4f 3) in polycrystalline ceramic garnet Y3Al5O12,” J. Appl. Phys. 96(6), 3050–3056 (2004). [CrossRef]
7. V. Lupei, A. Lupei, S. Georgescu, T. Taira, Y. Sato, and A. Ikesue, “The effect of Nd concentration on the spectroscopic and emission decay properties of highly doped Nd:YAG ceramics,” Phys. Rev. B 64, 092102/1–4 (2001).
8. G. A. Kumar, J. Lu, A. A. Kaminskii, K. Ueda, H. Yagi, T. Yanagitani, and N. V. Unnikrishnan, “Spectroscopic and Stimulated Emission Characteristics of Nd3+ in Transparent YAG Ceramics,” IEEE J. Quantum Electron. 40(6), 747–758 (2004). [CrossRef]
9. V. Lupei, A. Lupei, S. Georgescu, B. Diaconescu, T. Taira, Y. Sato, S. Kurimura, and A. Ikesue, “High resolution spectroscopy and emission decay in concentrated Nd:YAG ceramics,” J. Opt. Soc. Am. B 19(3), 360–368 (2002). [CrossRef]
11. T. Denis, S. Hahn, S. Mebben, R. Wilhelm, C. Kolleck, J. Neumann, and D. Kracht, “Compact diode stack end pumped Nd:YAG amplifier using core doped ceramics,” Appl. Opt. 49(5), 811–816 (2010). [CrossRef] [PubMed]
13. H. Yagi, J. F. Bisson, K. Ueda, and T. Yanagitani, “Y3Al5O12 ceramic absorbers for the suppression of parasitic oscillation in high-power Nd:YAG lasers,” J. Lumin. 121(1), 88–94 (2006). [CrossRef]
14. U. S. A. Baikowski, 6601 NorthPark Blvd., S. H. Charlotte, NC 28216, www.baikowski.com.
15. J. Neumann, S. Hahn, R. Huß, R. Wilhelm, M. Frede, D. Kracht, and P. Peuser, “Compact, Highly Efficient, Passively Q-Switched Nd:YAG MOPA for Spaceborne Bepi Colombo Laser-Altimeter, “ in Technical Digest on CD-ROM of Conference on Lasers and Electro-Optics/Quantum Electronics and Laser Science, (Long Beach, Calif., 2006), paper CWF1.