Gain-switched by a 1.914-µm Tm:YLF crystal laser, a two-stage Tm3+ fiber laser has been achieved 100-W level ~2-µm pulsed laser output with a slope efficiency of ~52%. With the 6-m length of Tm fiber, the laser wavelength was centered at 2020 nm with a bandwidth of ~25 nm. Based on an acousto-optic switch, the pulse repetition rate can be modulated from 500 Hz to 50 kHz, and the laser pulse width can be tuned between 75 ns and ~1 µs. The maximum pulse energy was over 10 mJ, and the maximum pulse peak power was 138 kW. By using the fiber-coiling-induced mode-filtering effect, laser beam quality of M2 = 1.01 was obtained. Further scaling the pulse energy and average power from such kind of gain-switched fiber lasers was also discussed.
©2010 Optical Society of America
Owing to its advantageous operating wavelength, ~2-µm solid-state lasers have attracted much attention in recent years. They can be used in wide areas, such as military remote detection, medical surgery, and scientific experiments, industrial machining, etc. ~2-µm laser output can be directly obtained from Ho3+-doped or Tm3+-doped crystal (or ceramic) lasers, or from the same ions doped into fibers. With crystal as host material, continuous-wave (CW) laser output of 150 W at ~2 µm  and 190 W at 1.9 µm  have been achieved. The combination of double-clad pumping technique with particular fiber configuration (relative large transverse area and high numerical aperture) has provided great potential to scale ~2 µm laser output power from Tm3+-doped fiber lasers. Since the first reported high-power double clad Tm3+-doped silica fiber laser , power scaling of the ~2-µm fiber laser has achieved great progress [4–7]. Recently, over 1000-W 2-µm laser output has been reported from a Tm3+-doped fiber laser .
For many applications, pulsed laser is preferred to CW mode owing to the advantages of high precision, high sensitivity, and high response speed from the former. High pulse-energy and high peak-power 2-µm lasers have been achieved from Tm- or Ho-doped crystal lasers or fiber lasers [9–11]. With Tm- and Ho- doped fiber as the gain element, 2-µm pulsed output from watt-level to over 10 W has also been achieved [12–15]. High-power 2-µm pulsed laser can find important application in pumping optical parametric oscillators to achieve 3~4-µm mid-infrared laser pulse , and in environmental detecting . For these applications, high laser power is usually required. However, high average-power pulsed 2-µm laser output has not been successfully accomplished due to thermal lensing effect or thermal damage in crystal systems, and immature Q-switching techniques combined with parasitic oscillation in fiber amplifier systems.
For obtaining 2-µm laser pulse, gain-switched fiber lasers can be excellent candidates owing to the well controlled pump pulse width and various available pump wavelengths [18–21]. Gain-switched operation of Tm3+ fiber lasers have achieved pulse energy of more than 10 mJ [18–20] and pulse width as narrow as 10 ns . Compared with fiber amplifier systems, which often include multiple stages, gain-switched devices can provide a much more compact configuration, greatly decreasing the complexity of the system. It is known that actively Q-switched crystal lasers can provide moderate-power 2-µm laser pulse with stably controlled characteristics. At the same time, Tm3+ fiber lasers can provide high gain with high operation efficiency by taking advantage of the cross relaxation (CR) process and minimizing the energy transfer up-conversion (ETU) process [22–24]. Therefore, combining these two kinds of systems can scale pulsed 2-µm laser to a new power level. Besides, the advantages of high-doping concentration, efficient pump absorption, long exited-state lifetime, and broad emission band make Tm3+ fiber laser extremely suitable for achieving high-power pulsed 2-µm laser output.
In this study, we report a high-power pulsed 2-µm Tm fiber laser, pumped by 793-nm diode lasers and, at the same time, gain switched by a Tm:YLF laser. This gain-switched system is different from conventional fiber amplifiers in that the high-power ‘seed’ laser acts just as a pulse switch, while the spectral characteristics are decided by the gain fiber. It is also different from conventional gain-switched lasers in that the gain and the switch are separated. We defined this laser system a CGSFL (combined gain-switched fiber laser) system. With this system, 100-W-level 2-µm pulsed laser has been realized from Tm3+ fibers for the first time. The generation of high-power 2-µm pulsed laser was contributed to the combination of high-power switch laser, damage-threshold improvement of the fiber-end facts, and appropriate system configurations. The maximum pulse energy was larger than 10 mJ, and maximum pulse peak power was >100 kW. Our experimental results indicate that in the nanosecond pulsed regime, Tm3+ fiber devices can be as efficient as, and even better than Yb3+ fiber systems.
The experiment setup for the CGSFL system is shown in Fig. 1 . The switch laser was constructed with a Tm:YLF slab crystal (grown with the Czochralski method by Shanghai Institute of Optics and Fine Mechanics). The slab crystal had an atomic-number doping concentration of 2%, and dimension of 1.5 × 6 × 20 mm, which was pumped by a 793-nm laser diode (LD). A plane-concave cavity configuration was adopted for the Tm:YLF seed laser to compensate the passive lensing effect  of the crystal. For Q-switching operation, a quartz acousto-optic Q-switch (QS027-4M-AP1, Gooch & Housego Co.) was adopted with a modulation loss of ~55%. In the output path, a plane mirror (R~99.5%@793 nm) was placed to block residual pump light.
The 1914-nm laser beam from the Tm:YLF crystal was launched by an aspheric lens (L1) and a 45° dichroic mirror (M2) into the Tm fiber. The double-clad Tm3+-doped silica fibers (both in the first-stage and the second-stage system) had a ~25-µm diameter, 0.1-NA core doped with ~2wt.% Tm3+. Large core of the fiber can help reduce laser power density, thus provide higher power scalability in pulsed operation. The octagonal pure-silica inner cladding, coated with a low-index polymer, had a 400-µm diameter and a NA of 0.46. The absorption coefficient at the pump wavelength (~793 nm) was measured with the cut-back method to be ~3dB/m. Between the switch laser and the first-stage fiber, as well as between the first-stage fiber and the second-stage fiber, Faraday optical isolators were used to prevent counter propagation of the laser beam.
In the first-stage CGSFL, the pump source was a 120-W LD module at 793 nm. Two aspheric lenses (L1) were used to couple the 793-nm pump light into the gain fiber, and a dichroic-mirror (M2: T = 97%@793 nm and R = 99.5%@1910 nm) was used to transmit the pump beam and reflect the switch beam. The total coupling efficiency of the 793-nm pump power was near 80%. In the first stage, a 6-m long fiber was adopted with total pump absorption of ~18 dB. In the second-stage CGSFL, two aspheric lenses (L1) and a dichroic mirror (M2) were used to couple the laser beam into the second Tm3+ gain fiber for further power scaling. In this stage, the pump source was a ~230-W 793-nm LD module. At the output end of the fiber, an aspheric lens (L1) was used to collimate the ~2-µm output laser, and a dichroic (R>99.9%@793nm; AR@1910nm) mirror (M1) was used to filter un-absorbed 793-nm pump light. Due to the quasi-three-level nature of Tm3+ fibers, cooling the fiber was critical for achieving high slop efficiencies. In both stages, the fiber ends was clipped between copper sinks, which in turn were water cooled at 16 °C. The central part of the Tm3+ fiber was wrapped on a 10-cm-diameter copper drum, which in turn was cooled by 18-°C circulating water. The laser output power was measured with a thermal power meter (Chinese Metrology College) and the laser spectrum was tested with a spectrometer (Zolix Instruments Co.). The pulse characteristics were detected with an InAs detector (Thorlabs Co.) combined with a 1-GHz Tektronix oscilloscope.
3. Results and discussion
The Tm:YLF laser can provide a maximum average output power of ~4.4 W (1914 nm) and repetition rate from 500 Hz to 50 kHz. The pulse width can be varied between ~80 ns and ~1.2 μs by tuning the repetition rate together with pump power. Firstly, bare fiber ends (10°-cleaved) were used as the output facet. Inclusion of optical isolators in the system was to amplify the 1914-nm switch laser pulse in advance for the sake of efficiently gain switching the Tm fiber laser (detailed description is presented in later spectral analysis).
In the first-stage CGSFL system (6-m Tm fiber), we launched ~1-W switch (1914 nm) laser beam into the fiber and kept the repetition rate at 10 kHz. In this case, the switch laser provided a pulse width of ~400 ns. The average output power and pulse energy of the first-stage system are shown in Fig. 2 . Under maximum pump power, the pulsed laser output reached ~40 W with a slope efficiency of 50% and threshold pump of 12 W. The maximum pulse energy was about 4 mJ, corresponding to a peak power of ~10 kW. At this level, no fiber facet damage was observed. For this stage, increasing the switch power to 4 W increased the output power to 41.3 W. Further increasing the switch power could not scale the average output any longer.
To scale the average power to a higher level, we incorporated a second-stage part (4-m Tm3+ fiber) to the CGSFL system. The laser beam launched into the second-stage Tm3+ fiber was tuned to 15 W and the pulse repetition rate was kept at 50 kHz. The laser output characteristics from the two-stage Tm3+ fiber are shown in Fig. 3 . The output increases linearly with the 793-nm pump power, and the maximum output is ~105 W, corresponding to a pulse energy of 2.1 mJ. The slope efficiency is 52.8%, comparable to the results achieved in the continuous-wave power amplifier system . We kept the pump power constant, and decreased the pulse repetition rate. When the repetition rate was decreased to 40 kHz, the output power dropped down to ~100 W, and the fiber end facet was damaged. At this time, the pulse energy is about 2.5 mJ, which is higher than the amplified pulse energy achieved in high-power Yb fiber amplifiers . Based on the pulse width of ~600 ns at 40 kHz, the peak power is ~4.2 kW.
In order to further improve the laser pulse energy, the fiber end facet damage threshold must be enhanced. Therefore, we fusion spliced a short piece (~2 mm) of passive silica fiber with a diameter of 1 mm to both ends of the active Tm fiber. The passive fiber ends were also angle-cleaved at ~10 degree. Adoption of the passive endcaps can greatly improve the damage threshold of the output facets. At this time, only the first-stage CGSFL system (6-m fiber) was employed. The pulse repetition rate was kept at 500 Hz, and the switch power launched into the fiber core was about 200 mW. Then, we increased the pump power until the facet of the endcap was damaged. Under these conditions, the maximum ~2-μm output power was ~5.2 W, corresponding to a pulse energy of 10.4 mJ. The 1-mm-diameter endcaps significantly decreased the maximum optical fluence in the gain fiber of >1000 J/cm2 to <5 J/cm2 at the output facet, which is less than the measured surface-damage fluence for nanosecond pulses in silica . According to the empirical damage threshold for fused silica  of >22tP 0.4 J/cm2 for 1064-nm laser pulses (tP is the pulse width in ns), the damage threshold with the 2-μm pulse should be even higher. With this relationship, the damage threshold for the 25-μm-core Tm fiber will be ~1.4 mJ, which is in reasonable agreement with the measured 2.5 mJ damage value (provided that the 2-μm pulse has a higher damage threshold than the 1064-nm pulse). With the 1-mm endcap fusioned, coarse calculation provides a damage threshold of hundreds of mJ for the endcap facet. Therefore, the damage to the endcap facet in experiment was probably caused by imperfectly managing the end surface or slight contamination of the facet. With our present CGSFL system, scaling the ~ 2-μm laser pulse energy to larger than 100 mJ should be possible.
Under different repetition rates, the maximum average power and pulse energy achieved with this CGSFL system are shown in Fig. 4 . With increasing repetition rate, the maximum achieved average output power increase near linearly. Over 40 kHz, the roll-over of the average output was owing to the limited pump. On the other hand, the maximum pulse energy decreases sharply first and then steadily with the repetition rate. At higher repetition rates, the limited stored energy in the gain fiber leaded to the decrease of pulse energy. The inset is the pulse shape measured at the pulse energy of ~10 mJ (repetition rate of 500 Hz) with a FWHM (full width at half maximum) width of 75 ns. Dividing the pulse energy by the pulse width, we obtained a maximum pulse peak power of >138 kW, which is believed to be the highest peak power from gain-switched Tm3+ fiber lasers.
In order to study the evolution of the 2-μm pulse shape and pulse width with pump power, the two-stage CGSFL system was adopted, and the pulse repetition rate was kept at 50 kHz. The pulse shape of the switch laser and the fiber laser at various power levels are shown in Fig. 5 , among which the Figs. 5(b)-5(d) were measured after the first-stage system and Fig. 5(e) was measured after the second-stage system. The output power of the switch laser was kept at 1 W, corresponding to a pulse energy of 20 μJ. The switch pulse had a pulse width of 900 ns, and a Gaussian pulse shape (as shown in Fig. 5(a)). In the first-stage system, the pulse width increased first and then narrowed with increased output power. At 35-W power level, leading edge of the pulse was steepened. After the second-stage system (~100 W), the pulse width was further reduced to 750 ns and steepening of the pulse leading edge was enhanced. The pulse broadening at low power levels was probably originated from the switching effect induced by the switch pulse. The coexistence of the switch pulse and the ~2020-nm laser pulse (see Fig. 6 ) caused the pulse broadening. Besides, increased gain may also augmented this effect. The pulse width narrowing and pulse steepening at high power levels was attributed to gain saturation (leading edge of the pulse deplete most of the stored energy) and self-phase modulation [25,29,30] in the power scaling process. The laser pulse obtained enough high energy so that it began to saturate the system, and the energy was primarily extracted by the pulse leading edge. Compared with the switch pulse width, the pulse narrowing is about 17%.
Based on pulsed operation, the spectral characteristics of the CGSFL system was studied. The operating conditions were the same as in the study of pulse evolution, 50 kHz repetition rate and 1 W switch power. The spectrum of the switch pulse and the fiber laser beam at different power levels are indicated in Fig. 6, among which the Figs. 6(b)-6(f) were measured after the first-stage CGSFL system and Fig. 6(g) was measured after the second-stage CGSFL system. The fluorescence spectrum of the Tm3+-doped fiber is also shown in Fig. 6(a) as the magenta curve. As shown in Fig. 6 (a), the Tm:YLF switch laser had a spectrum centered at 1914 nm with a spectral width (FWHM) of 2.5 nm. At the power level of 2 W (as shown in Fig. 6b), the 1914-nm switching laser beam was amplified. This amplification process continued until the switch laser reached ~5 W. Augmenting the switch laser to a high value is critical to effectively switch the gain of the fiber laser at high power levels. Further increasing the pump, the ~2-µm laser pulse was stimulated, and thereafter more and more stored energy was extracted by the 2020-nm laser pulse (as shown in Figs. 6(c)-6(e)). At even higher power levels, all energy was included in the 2020-nm laser beam, so only this wavelength spectrum was observed. This is a unique characteristic of the CGSFL system, significantly different from fiber amplifiers and singly gain-switched devices. The detailed process can be described as follows. As indicated in Fig. 6(a), the fluorescence spectrum of our Tm fiber covers the spectral range of 1920-2040 nm. The wavelength of the switch pulse (1914 nm) lies in the wing of the gain spectrum of the Tm fiber. When the switch laser was launched into the fiber core, more than 90% of the switch beam was absorbed. At the same time, the Tm fiber was also pumped by the 793-nm LD and achieved population inversion. The unabsorbed 1914-nm laser pulse will be amplified, while the absorbed 1914-nm laser pulse will modulate the gain of the system as a switch for the ~2-µm laser emission. There was a gain competition between the 1914-nm amplification and the stimulation of the ~2-µm laser emission. At low pump levels, the amplification process dominated over the latter. With increase of the 793-nm pump, the ~2-µm laser emission obtained much more gain than the 1914-nm beam. At the same time, the amplified 1914-nm pulse can be further reabsorbed by the Tm fiber. The 1914-nm laser was consumed step by and step, and finally all the stored energy was extracted by the ~2-µm laser emission. Therefore, at high pump levels, the 1914-nm laser functions just as a switch to trigger the 2020-nm laser pulse, and the spectral characteristics of the CGSFL system will be completely decided by the gain fiber. The deviation of the laser wavelength (2020 nm) from the gain spectrum center was owing to laser re-absorption in the slightly long Tm fiber (~6 m). This shows the quasi-three-level characteristics of Tm3+ ions, a broad Stark splitting of the energy levels. At 100-W level, the spectral width of the ~2-µm laser pulse was ~25 nm. In the inset of Fig. 6(g), the laser spectrum in the unit of dB is indicated, showing that the background of the laser pulse was ~30dB lower than the signal. For comparison, we also observed the spectral evolution of a 2-m Tm3+ fiber laser (50 kHz repetition rate and 1 W switch power) with one-stage CGSFL configuration, and the results are shown in Fig. 7 . At this time, the laser wavelength was 1982 nm, showing excellent overlap with the gain spectrum center due to reduced re-absorption of laser light with shorter fibers. In the Tm3+-doped CGSFL, shortening the fiber length will lead to blue shift of the laser wavelength, providing a wide wavelength tuning range . With the 2-m fiber, the ~2-µm laser pulse appeared (Fig. 7(b)) and dominated (Fig. 7(d)) at higher power levels (10 W and 20 W, respectively) than that with the 6-m fiber. Therefore, shorter fibers probably need stronger pump to switch on the CGSFL system.
The near-field fiber laser beam image was recorded by burning a piece of fax paper, which is indicated as the inset of Fig. 8 , showing a Gaussian mode field profile. Beam quality of the fiber laser light was measured by focusing the output with an f = 25 mm convex lens and using the 86.5/13.5 scanning-knife-edge method. From the focused point, the beam radius as a function of axial displacement is shown in Fig. 8. The M 2 factor of the 2-µm laser beam at the 100-W level was best-fitted to be 1.01 ± 0.02, confirming the nearly diffraction-limited beam quality and fundamental mode operation of the CGSFL system. The fundamental mode operation was attributed to high-order-mode suppression by the coiling-filtering effect . In our experiment, the Tm fiber was coiled on a copper drum with a diameter of ~10 cm to suppress high-order modes.
In this paper, a new kind of gain-switched fiber laser (CGSFL) was constructed, and high-power 2-µm pulsed laser has been achieved from such a system with Tm3+ fibers as the gain material. Operated at 50-kHz repetition rate, the Tm3+ fiber CGSFL provided 105-W average power with a slope efficiency of 52%. At 500-Hz repetition rate, the pulse energy and pulse peak power reached 10.4 mJ and 138 kW, respectively. High average power and pulse energy were attributed to high cavity gain, large gain-fiber core, and improving the damage threshold of the fiber-end facets. The laser spectrum of the CGSFL had a center wavelength of 2020 nm with a bandwidth of ~25 nm. The high-power laser pulse possessed a beam quality of M 2 = 1.01 ± 0.02 due to the mode filtering effect. Such a high-average-power, large-pulse-energy CGSFL system with good beam quality is competitive in the ‘eye-safe’ application areas.
With a high-power pulsed slab laser as the switch, and large-mode-area Tm fibers as the gain element, such CGSFL system has a compact configuration and can provide high-power high repetitive 2-µm laser pulses in the nanosecond regime. This kind of high-power pulsed fiber laser sources can find wide applications and spreading potentials in areas such as ranging, remote sensing and medical surgery. Further scaling the average power and pulse energy of such a CGSFL system requires further increasing the fiber-core area and improving the damage threshold of the fiber facets.
This work was supported by the “863” Program of China under Grant No. 2009AA03Z441 and the Natural Science Foundation of Shanghai under Contract 10ZR1433700.
References and links
1. K. Lai, W. Xie, R. Wu, Y. Lim, E. Lau, L. Chia, and P. Phua, “A 150- 2-micron diode-pumped Tm:YAG laser,” OSA Trends Opt. Photo. 68, paper WE6, 2002.
2. M. Schellhorn, S. Ngcobo, C. Bollig, M. Esser, D. Preussler, and K. Nyangaza, “High-power diode-pumped Tm:YLF slab laser,” in Lasers and Electro-Optics 2009 and the European Quantum Electronics Conference. CLEO Europe - EQEC pp. 1–1 (2009).
3. S. D. Jackson and T. A. King, “High-power diode-cladding-pumped Tm-doped silica fiber laser,” Opt. Lett. 23(18), 1462–1464 (1998). [CrossRef]
4. S. Jiang, J. Wu, Zh Yao and J. Zong, “104 W Highly Efficient Thulium Doped Germanate Glass Fiber Laser,” Adv. Solid-State Photon. MF3 (2007).
5. E. Slobodtchikov and P. F. Moulton, “Efficient, High-Power, Tm-doped Silica Fiber Laser,” Adv. Solid-State Photon. MF2 (2007).
6. G. D. Goodno, L. D. Book, and J. E. Rothenberg, “600-W, Single-Mode, Single-Frequency Thulium Fiber Laser Amplifier,” Proc. SPIE7195, 71950Y–1~10 (2009).
7. P. F. Moulton, G. A. Rines, E. V. Slobodtchikov, K. F. Wall, G. Frith, B. Samson, and A. L. G. Carter, “Tm-Doped Fiber Lasers: Fundamentals and Power Scaling,” IEEE J. Sel. Top. Quantum Electron. 15(1), 85–92 (2009). [CrossRef]
9. J. Yu, B. C. Trieu, E. A. Modlin, U. N. Singh, M. J. Kavaya, S. Chen, Y. Bai, P. J. Petzar, and M. Petros, “1 J/pulse Q-switched 2 µm solid-state laser,” Opt. Lett. 31(4), 462–464 (2006). [CrossRef] [PubMed]
10. B. C. Dickinson, S. D. Jackson, and T. A. King, “10 mJ total output from a gain-switched Tm-doped fibre laser,” Opt. Commun. 182(1-3), 199–203 (2000). [CrossRef]
11. G. Imeshev and M. E. Fermann, “230-kW peak power femtosecond pulses from a high power tunable source based on amplification in Tm-doped fiber,” Opt. Express 13(19), 7424–7431 (2005). [CrossRef] [PubMed]
12. M. Eichhorn and S. D. Jackson, “Actively Q-switched Tm3+-doped and Tm3+,Ho3+-codoped Silica Fiber Lasers,” in Conference on Lasers and Electro-Optics and Quantum Electronics and Laser Science Conference 2008, (San Jose, CA, 2008).
15. D. Creeden, P. Budni, P. A. Ketteridge, T. M. Pollak, E. P. Chicklis, G. Frith, and B. Samson, “High Power Pulse Amplification in Tm-Doped Fiber,” in Conference on Lasers and Electro-Optics/Quantum Electronics and Laser Science Conference and Photonic Applications Systems Technologies, OSA Technical Digest (CD), Optical Society of America, paper CFD1(Washington DC, 2008).
16. D. Creeden, P. A. Ketteridge, P. A. Budni, S. D. Setzler, Y. E. Young, J. C. McCarthy, K. Zawilski, P. G. Schunemann, T. M. Pollak, E. P. Chicklis, and M. Jiang, “Mid-infrared ZnGeP2 parametric oscillator directly pumped by a pulsed 2 microm Tm-doped fiber laser,” Opt. Lett. 33(4), 315–317 (2008). [CrossRef] [PubMed]
18. B. C. Dickinson, S. D. Jackson, and T. A. King, “10 mJ total output from a gain-switched Tm-doped fibre laser,” Opt. Commun. 182(1-3), 199–203 (2000). [CrossRef]
20. S. D. Jackson and T. A. King, “Efficient Gain-Switched Operation of a Tm-Doped Silica Fiber Laser,” IEEE J. Quantum Electron. 34(5), 779–789 (1998). [CrossRef]
22. Y. L. Tang and J. Q. Xu, “Effects of excited-state absorption on self-pulsing in Tm3+-doped fiber lasers,” J. Opt. Soc. Am. B 27(2), 179–186 (2010). [CrossRef]
23. G. Frith, D. G. Lancaster, and S. D. Jackson, “85 W Tm3+-doped silica fibre laser,” Electron. Lett. 41(12), 687–688 (2005). [CrossRef]
24. S. D. Jackson, “Cross relaxation and energy transfer upconversion processes relevant to the functioning of 2 μm Tm3+-doped silica fibre lasers,” Opt. Commun. 230(1-3), 197–203 (2004). [CrossRef]
25. 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. Sel. Top. Quantum Electron. 33(9), 1592–1600 (1997). [CrossRef]
26. J. Limpert, S. Hofer, A. Liem, H. Zellmer, A. Tunnermann, S. Knoke, and H. Voelckel, “100-W average-power, high-energy nanosecond fiber amplifier,” Appl. Phys. B 75(4-5), 477–479 (2002). [CrossRef]
27. B. C. Stuart, M. D. Feit, S. Herman, A. M. Rubenchik, B. W. Shore, and M. D. Perry, “Nanosecond-to-femtosecond laser-induced breakdown in dielectrics,” Phys. Rev. B Condens. Matter 53(4), 1749–1761 (1996). [CrossRef] [PubMed]
28. W. Koechner, Solid-state laser engineering, 5th ed., (Springer-Verlag, Berlin, 1995), p. 685.
29. C. D. Brooks and F. Di Teodoro, “1-mJ energy, 1-MW peak-power, 10-W average-power, spectrally narrow, diffraction-limited pulses from a photonic-crystal fiber amplifier,” Opt. Express 13(22), 8999–9002 (2005). [CrossRef] [PubMed]
30. F. Di Teodoro and C. D. Brooks, “1.1 MW peak-power, 7 W average-power, high-spectral-brightness, diffraction-limited pulses from a photonic crystal fiber amplifier,” Opt. Lett. 30(20), 2694–2696 (2005). [CrossRef] [PubMed]
31. J. Q. Xu, M. Prabhu, J. R. Lu, K. I. Ueda, and D. Xing, “Efficient double-clad thulium-doped fiber laser with a ring cavity,” Appl. Opt. 40(12), 1983–1988 (2001). [CrossRef]
32. M.-Y. Cheng, Y.-Ch. Chang, A. Galvanauskas, P. Mamidipudi, R. Changkakoti, and P. Gatchell, “High-energy and high-peak-power nanosecond pulse generation with beam quality control in 200-µm core highly multimode Yb-doped fiber amplifiers,” Opt. Lett. 30(4), 358–360 (2005). [CrossRef] [PubMed]