We compare the performance characteristics of Tm:YAP and Ho:YAG in ultrashort pulse regenerative amplification. Both systems follow the same amplification concept and use nearly the same experimental setup reaching similar output energies of >700 µJ.
© 2016 Optical Society of America
Until now, only few high energy, ultrashort pulse laser systems have been reported in the 2 µm wavelength region. Typical applications for these laser systems are the nonlinear frequency conversion in the 3 – 10 µm range [1, 2], high harmonic generation  or atmospheric sensing . Although up to 120 µJ pulse energy with a pulse duration of 540 fs was demonstrated by a thulium-doped large-pitch fiber CPA system , a further increase of pulse energy is limited due to a thermal degradation of the beam profile induced by the propagation in air [5,6].
Therefore, regenerative amplifiers (RA) with bulk materials have gained an increased interest recently, which generate sub-ps pulse duration and energy up to the mJ range [7–9]. In 2013, Malevich et al. used a Ho:YAG RA operating at 2.1 µm to generate pulses with a duration of 530 fs and energy of 990 µJ. This short pulse duration was achieved by spectral filtering of a broadband OPA seed inside the grating stretcher to compensate for the strong spectral gain shaping inside the RA ring cavity. When this system was seeded by a more compact Tm-Ho-doped ultrashort pulse fiber laser with a narrower seed spectrum, only a minimum pulse duration of 1 ps could be generated . Although high pulse energies were achieved with Ho:YAG, it is typically pumped by high power Tm:fiber lasers because high brightness laser diodes are not available. However, Tm:YAP offers a broader gain spectrum which is highly suitable for the amplification of ultrashort pulses < 500 fs and can be directly pumped with high power, high brightness laser diodes . Another promising candidate for regenerative amplification is Ho:YLF which demonstrated up to tens of mJ pulse energy, but its limited spectral gain bandwidth supports longer pulse durations in the multiple ps-range [10, 11]. Nevertheless, for many of the mentioned applications, a short pulse duration in the sub-ps range is crucial. Other typical and commercially available materials like Tm:YAG, Tm:YLF or Ho:YVO showed lower emission cross sections and hence Tm:YAP and Ho:YAG were chosen as best candidates for the experiments [12–14]. Table 1 summarizes the main optical properties of both materials.
Above all, the laser materials Ho:YAG and Tm:YAP are until now the only candidates which showed sub-ps pulse durations and high energy pulses in the 2 µm wavelength range, but a direct comparison between the two materials is not possible as the demonstrated laser systems based on different seed sources, seed energies, and especially cavity designs. In this contribution, we present a direct comparison of the performance aspects between Tm:YAP and Ho:YAG which delivers new insights in the amplification behaviour and gain dynamics. This challenging task has not been done before and might be potentially interesting for future investigations, in particular as this regenerative amplifier layout can be regarded as unique systems which can be used for a variety of other active materials operating around 2 µm and thus offering tailored features for above mentioned applications. Both materials perform under the same conditions utilizing nearly the same seed energy, seed spectral profile, and cavity design. Therefore, a direct comparison is feasible and delivers a fundamental understanding for the choice of the amplifying material. Both materials can achieve similar pulse energies in this configuration (>700 µJ) but different pulse durations.
2. Experimental setup and results
The two materials are evaluated separately in successive experiments starting at first with the thulium setup depicted in Fig. 1 (black squares). In general, the setup of both systems consists of an ultrashort pulse fiber oscillator, a stretcher unit, a fiber amplifier, and a regenerative amplifier. When the system is modified to operate with holmium-doped media shown in Fig. 1 (green squares), the fiber oscillator, stretching unit, fiber amplifier, and laser crystal are exchanged as well as the pump lasers and the two dichroic mirrors (DM) around the crystal for the pump coupling into the RA cavity. All other optical elements of the setup (mirrors, thin film polarizers, Faraday rotators, Pockels cells, and gratings) are coated broadband to support a wavelength range from 1.85 µm – 2.15 µm and are used for both setups. Therefore, the conditions for a feasible comparison between the two RA materials Tm:YAP and Ho:YAG are as similar as possible by using the same RA cavity, similar seed pulse energy, and spectral seed profile.
The ultrashort pulse fiber oscillator (setup comparable to the one described in ) generates 120 fs-pulses at 1940 nm in case of thulium (463 fs at 2100 nm in case of holmium). The pulses are stretched in a fiber stretcher (grating stretcher in case of holmium is used due to higher losses in silica fibers) to 90 ps pulse duration, corresponding to stretching ratios of 750 and 194, respectively. The stretched pulse duration is measured after the subsequent fiber preamplifier which amplifies the negatively chirped pulses to more than 25 nJ pulse energy. The optical spectra after preamplification are both Gaussian-like shaped and have a FWHM of 23 nm in the case of thulium and 15.5 nm in the case of holmium. The pulse picker picks the pulses to achieve 1 kHz repetition rate before injection into the RA cavity. A thin film polarizer (TFP) and a Faraday rotator (FR) separate the amplified output pulses of the RA from the seed pulses.
A detailed description of the RA cavity can be found in , but it consists in general of a linear cavity with two curved mirrors (CCM1 and CCM2) to generate a mode field diameter (MFD) of 320 µm inside the laser crystal. For the calculation of the MFD inside the cavity, a thermal lens of 80 mm in the Tm:YAP crystal was used which was experimentally investigated at a pump power of 9.45 W, while Ho:YAG showed no influence on the incident pump power due to the low quantum defect. The crystal lengths are 4 mm for the c-axis cut Tm:YAP (4 at. % doping concentration) and 10 mm for Ho:YAG with a doping concentration of 1.6 at. %. These values result in a similar calculated pump absorption of 89 % for Tm:YAP and 94 % for Ho:YAG, respectively. Due to the difference in crystal length and thermal lens, the distance of CCM1 in Fig. 1 to the Ho:YAG crystal is slightly shortened compared to the Tm:YAP setup to achieve the same MFDs.
In case of Tm:YAP, a fiber coupled multi-mode laser diode delivers up to 35 W pump power at a central wavelength of 793 nm out of a multi-mode fiber with a core diameter of 105 µm and a NA of 0.22. The pump light is focused through one of the dichroic mirrors into the crystal to produce a pump spot diameter of 500 µm over the whole length of the crystal. The pump light for the Ho:YAG crystal produces a spot of 320 µm and is generated by a homebuilt single-mode thulium-fiber laser system operating at 1908 nm with a maximum output power of 12.7 W and diffraction limited beam quality. Due to the strong thermal lens in Tm:YAP, a larger pump spot is chosen to reduce this effect. Therefore, the absorbed pump peak intensity values are used in order to keep the comparability between both systems at different pump power level when the system is varied in round trip numbers, repetition rate, and seed energy. The RA cavity is enclosed to purge the cavity with inert gas (argon, nitrogen etc). After amplification, the pulses are compressed with a Martinez-type compressor  with an efficiency of 50 % which is mainly limited by the low diffraction efficiency of the grating. This issue can easily be resolved by using gratings with higher diffraction efficiency.
The emission cross section spectrum of Ho:YAG exhibits two main peaks at 2090 nm and 2096 nm with the maximum of 16.3 × 10−25 m2 at 2090 nm which is 3.3 times higher than the maximum emission cross section of Tm:YAP (4.9 × 10−25 m2) at 1940 nm [12,15]. Unlike for Ho:YAG, the emission cross section spectrum of Tm:YAP is much broader, centered around 1940 nm with a FWHM of 39 nm and therefore highly suitable for the generation of ultrashort pulses. Taking emission and absorption cross sections in Table 1 into account, the calculated saturation fluence of Tm:YAP is 18J/cm2 at the main emission wavelength which is more than 3.5 times higher than the saturation fluence of Ho:YAG (5.1J/cm2) at its main emission wavelength. This lower saturation fluence results in a lower seed energy necessary to extract most of the energy of the crystal or lower round trip numbers compared to Tm:YAP. Own measurements of the fluorescence lifetime revealed a value of 10.6 ms for Ho:YAG, which is 1.7 times higher than the one of Tm:YAP, so more energy can be stored inside the Ho:YAG crystal when only considering the fluorescence lifetime. Both host materials exhibit similar physical properties like density, thermal diffusivity, thermal conductivity or thermal coefficient of refractive index , but YAP is less affected by thermal depolarization because of its birefringent nature.
In both cases, pulse energies > 700 µJ at 1 kHz repetition rate are achieved after 34 (Tm:YAP) and 10 round trips (Ho:YAG), respectively, which is shown in Fig. 2. The limiting factors are the damage thresholds of both laser crystals at a power density of 2.0–2.1J/cm2, which were investigated in earlier experiments. Therefore, we limit the power density below 1.77J/cm2 in all following experiments. The laser cavity has to be readjusted for every pump power step in case of Ho:YAG to find the maximum output power. This is not induced by a transversal mode hopping but by a slight misalignment of the crystal’s angle to the incoming pump light to avoid back reflections into the non-isolated pump laser. The output beam of the RA was perfectly round in both cases, which is depicted as insets in Fig. 2, and showed an averaged beam propagation parameter M2 of 1.59 for Tm:YAP and 1.63 for Ho:YAG, respectively. Nonetheless, Ho:YAG shows a higher slope efficiency of 36.5 % (14.5 % Tm:YAP), which can be explained by the higher emission cross section and the better pump mode overlap for Ho:YAG.
Figure 3(a) shows the optical spectrum of Tm:YAP in the purged case at 709 µJ output energy as black curve after regenerative amplification (blue dashed: seed pulses). The unpurged spectrum (not shown here) is strongly structured, particularly by atmospheric absorption lines which are present at this wavelength and could be partly reduced by purging the RA cavity with nitrogen gas. Therefore, the spectrum width is 18.5 nm at − 10 dB level. The spectrum of Ho:YAG at 714 µJ output energy is depicted in Fig. 3(b) which shows a strong deformation of the Gaussian-shaped seed spectrum. Mainly visible in the spectrum are the main emission peaks of Ho:YAG at 2090 nm and 2096 nm with a combined width of 7.8 nm at −10 dB which are well known by the emission cross section spectrum. In order to precompensate for the strong spectral evolution into these two peaks, the seed spectrum is hard-cut filtered inside the grating stretcher by the use of two pins. The resulting seed spectrum is shown in Fig. 3(c) as blue dashed curve with two dips at the main emission peaks of Ho:YAG. Due to the precompensation, the main peaks are suppressed to generate a broader optical spectrum (10.5 nm at −10 dB) around 2093 nm.
The autocorrelation (AC) trace of the compressed output pulses of Tm:YAP is shown in Fig. 4(a). After purging, the long ps-pedestal generated by the atmospheric absorption vanishes nearly completely. The pulses can be compressed to a pulse duration of 410 fs at a maximum output energy of 709 µJ. A satellite pulse with a relative peak intensity of 6.2 % in a temporal distance of 3.2 ps remains. Nonetheless, around 76 % of energy is confined in the main peak. The generation of the satellite pulse can be attributed at least partly to a technical imperfection of the PC. One explanation could be residual water absorption inside the dielectric AR coatings of the RTP-crystals [19,20]. As already mentioned, it is necessary to purge the RA cavity with nitrogen gas to lower the amount of water vapor in the surrounding air and therefore reduce the strong water absorption which affect the pulse structure (by a long ps-pedestal). These pulse quality degradation effects have been observed in the mid-infrared wavelength range between 6 µm and 8 µm at long ps pulses by Seilmeier et al. already in 1988 . Very recently, these effects have also been investigated in the 2 µm range . As spectral absorption and refractive index (and therefore dispersion) are linked together by the Kramers-Kronig relations , a sharp absorption line at a specific wavelength causes a Fano-shaped feature in the refractive index profile. This leads to a strong spectral phase distortion across the spectrum (higher dispersion values) which forms post-pulses and a long pulse tail (broadband ps-pedestal).
As RTP itself is non-hygroscopic , it can be excluded as possible cause. However, the AR coatings on the PC are manufactured by Electron Beam Evaporation , which are more porous than coatings fabricated by Ion Assisted Deposition or Ion Beam Sputtering . These porous layers adsorb water from the ambient air after the coating chamber is vented and cannot be released. This incorporated water causes high losses which have already been measured in the 2.9 µm water window . Therefore, switching to another wavelength region (e.g. 2.1 µm) without strong water absorptions should result in a satellite-free AC trace. This is proven when the whole setup is changed to holmium-doped materials. The compressed AC trace of the unshaped Ho:YAG pulses is depicted in Fig. 4(b), whereas Fig. 4(c) shows the trace of the shaped pulses in which no satellite pulses are visible. The advantage of the aforementioned spectral shaping is clearly visible, because the multi-pulse like structure of the AC trace in the unshaped case vanishes and a smooth AC trace in the shaped case appears with a compressed pulse duration of 1.19 ps at an energy of 711 µJ. The achieved pulse duration is only slightly longer compared with the work by Malevich et al. who additionally used an acousto-optic programmable dispersive filter for phase correction . Anyway, no satellite pulse nor a broadband ps-pedestal occurs in this wavelength region. The pulse duration is nearly a factor of 3 larger than for Tm:YAP, which results in a 2.34 times lower peak power. Obviously, the narrow gain spectrum of Ho:YAG is responsible for the longer compressed pulse duration of Ho:YAG. On the other hand, no atmospheric absorption take place at 2.1 µm, therefore purging as in the case of Tm:YAP is not necessary.
To determine the scaling potential of both materials, the round trips of the RA are varied at a repetition rate of 1 kHz. The results are depicted in Fig. 5(a). With similar pump intensities of 83.1MW/m2 (Tm:YAP) and 86.6MW/m2 (Ho:YAG) applied and at same seed energy of ~ 25 nJ, Ho:YAG amplifies much more efficient. High output energies of 492 µJ are achieved already after 15 round trips. A further increase leads to the onset of bifurcation instabilities which can be suppressed efficiently with a higher pump intensity of 98.2MW/m2. With this higher pump intensity, an output energy of 694 µJ is reached after 16 round trips which is simultaneously the upper limit due to the damage threshold of the laser crystal. Anyway, at these values, most of the energy of the Ho:YAG crystal is extracted and the pulse energy saturates. However, Tm:YAP achieves a pulse energy of 703 µJ after 44 round trips and shows no bifurcation behavior at all. Furthermore, a saturation of the output energy is not present, so it is assumed that even higher pulse energy can be extracted by overcoming the damage threshold of the crystal.
Investigations on output energy in terms of repetition rate are carried out for both materials under roughly the same conditions. First of all, the seed energy in both cases is set to ~25 nJ. Secondly, the pump intensity for both materials is set to the same level (Tm:YAP: 75.8MW/m2, Ho:YAG: 76.8MW/m2). Finally, the round trip numbers are set to a point, where the RA operates in the linear regime to eliminate any saturation or bifurcation effects, particularly 10 round trips for Ho:YAG and 34 round trips for Tm:YAP are chosen. Starting at a repetition rate of 3 kHz and reducing it to 100 Hz, both materials show a similar behavior and course of lines as depicted in Fig. 5(b). The achieved pulse energy of both materials is ca. 100 µJ at 3 kHz and follows a nearly exponential increase with lower repetition rates up to ca. 700 µJ at 100 Hz.
The seed energy for both amplifier materials is varied to explore the RA behavior at 1 kHz, which is shown in Fig. 6. For Tm:YAP (34 round trips) depicted in (a), the pulse energy increases linearly from 491 µJ at a seed energy of 12.3 nJ to a maximum pulse energy of 700 µJ at a seed energy of 48.7 nJ. The compressed pulse duration (blue curve) stays nearly constant during amplification between 380 fs and 402 fs for the different seed pulse energies and is shortest at the highest output power. For the seed variation of Ho:YAG shown in Fig. 6(b), a slightly different oscillator with a Gaussian-like spectrum centered around 2090 nm and a FWHM of 17.3 nm is used. In direct comparison to Tm:YAP, Ho:YAG shows a higher pulse energy of 526 µJ at a lower seed energy of 7.1 nJ after 10 round trips. It increases linearly up to an output energy of 705 µJ at maximum seed energy of 31.2 nJ. The compressed pulse duration (blue curve) varied between 1.20 ps and 1.12 ps for the different seed pulse energies and was shortest at 20.6 nJ pulse energy. Due to the different oscillator with spectral precompensation, the compressed AC traces are slightly shorter, but have stronger sidelobes compared to Fig. 4(c).
In conclusion, we carried out a detailed comparison between Thulium:YAP and Holmium:YAG in ultrashort pulse regenerative amplification. Both materials are investigated under similar conditions (seed pulse energy, seed spectrum profile etc.) in the same cavity design. Both materials are suitable to achieve uncompressed pulse energies of more than 700 µJ, limited in this configuration by the damage threshold of the crystals. Besides that, Ho:YAG shows a higher amplification efficiency which is caused by the better pump mode overlap and higher emission cross section, but a spectral deformation during amplification occurs. The seed pulse spectrum can be shaped to reduce this effect, but the pulse duration is limited to 1.19 ps at highest output power. In case of Tm:YAP, the amplified pulses can be compressed to 410 fs, but the cavity has to be purged to reduce the atmospheric absorption. Furthermore, an investigation concerning scaling behavior and gain dynamics while varying round trips, repetition rate and seed energy has been carried out which shows that no bifurcation or saturation effects are visible for Tm:YAP but are both present for Ho:YAG.
We gratefully thank Dr. Lars Jensen for the helpful discussion about the water adsorption in dielectric e-beam layers. This work has been funded by the German Federal Ministry of Education and Research (BMBF) under contract 13N12079 “NEXUS”.
References and links
1. D. Sánchez, M. Hemmer, M. Baudisch, S. L. Cousin, K. Zawilski, P. Schunemann, V. Smirnov, H. Hoogland, R. Holzwarth, O. Chalus, C. Simon-Boisson, and J. Biegert, “Broadband 7 µm OPCPA pumped by a 2 µm picosecond Ho:YLF CPA system,” in CLEO: 2015, OSA Technical Digest (online) (Optical Society of America, 2015), paper SW4O.5.
2. N. Leindecker, A. Marandi, R. L. Byer, K. L. Vodopyanov, J. Jiang, I. Hartl, M. Fermann, and P. G. Schunemann, “Octave-spanning ultrafast OPO with 2.6–6.1 µm instantaneous bandwidth pumped by femtosecond Tm-fiber laser,” Opt. Express 20, 7046–7053 (2012). [CrossRef]
3. K.-H. Hong, C.-J. Lai, J. P. Siqueira, P. Krogen, J. Moses, C.-L. Chang, G. J. Stein, L. E. Zapata, and F. X. Krtner, “Multi-mJ, kHz, 2.1m optical parametric chirped-pulse amplifier and high-flux soft x-ray high-harmonic generation,” Opt. Lett. 39, 3145–3148 (2014). [CrossRef] [PubMed]
4. P. Kadwani, R. Sims, J. Chia, F. Altat, L. Shah, and M. Richardson, ”Atmospheric Propagation Testing Using Broadband Thulium Fiber Systems,” in Advances in Optical Materials, OSA Technical Digest (CD) (Optical Society of America, 2011), paper FWB3. [CrossRef]
5. F. Stutzki, C. Gaida, M. Gebhardt, F. Jansen, C. Jauregui, J. Limpert, and A. Tünnermann, “Tm-based fiber-laser system with more than 200 MW peak power,” Opt. Lett. 40, 9–12 (2015). [CrossRef]
6. M. Gebhardt, C. Gaida, F. Stutzki, S. Hädrich, C. Jauregui, J. Limpert, and A. Tünnermann, “Impact of atmospheric molecular absorption on the temporal and spatial evolution of ultra-short optical pulses,” Opt. Express 23, 13776–13787 (2015). [CrossRef] [PubMed]
7. P. Malevich, G. Andriukaitis, T. Flöry, A. J. Verhoef, A. Fernández, S. Ališauskas, A. Pugžlys, A. Baltuška, L. H. Tan, C. F. Chua, and P. B. Phua, “High energy and average power femtosecond laser for driving mid-infrared optical parametric amplifiers,” Opt. Lett. 38, 2746–2748 (2013). [CrossRef] [PubMed]
8. P. Malevich, T. Kanai, H. Hoogland, R. Holzwarth, A. Baltuska, and A. Pugzlys, “Millijoule 1-ps Pulses from a kHz Ho:YAG Regenerative Amplifier Seeded with a Tm,Ho-Fiber Laser,” in CLEO: 2015, OSA Technical Digest (online) (Optical Society of America, 2015), paper SM1P.4.
10. L. von Grafenstein, M. Bock, D. Ueberschaer, U. Griebner, and T. Elsaesser, “Picosecond 34 mJ pulses at kHz repetition rates from a Ho:YLF amplifier at 2 m wavelength,” Opt. Express 23, 33142–33149 (2015). [CrossRef]
11. P. Kroetz, A. Ruehl, G. Chatterjee, A.-L. Calendron, K. Murari, H. Cankaya, P. Li, F. X. Kärtner, I. Hartl, and R. J. D. Miller, “Overcoming bifurcation instability in high-repetition-rate Ho:YLF regenerative amplifiers,” Opt. Lett. 40, 5427–5430 (2015). [CrossRef] [PubMed]
12. S. A. Payne, L. L. Chase, L. K. Smith, W. L. Kway, and W. F. Krupke, “Infrared Cross-Section Measurements for Crystals Doped with Er3+, Tm3+, and Ho3+,” IEEE J. Quant. Electron. 28, 2619–2630 (1992). [CrossRef]
13. B. M. Walsh, N. P. Barnes, and B. di Bartolo, “Branching ratios, cross sections, and radiative lifetimes of rare earth ions in solids: Application to Tm3+ and Ho3+ ions in LiYF4,” J. Appl. Phys. 83, 2772–2787 (1998). [CrossRef]
14. G. Li, B. Q. Yao, P. B. Meng, X. M. Duan, Y. L. Ju, and Y. Z. Wang, “Diode-pumped efficient laser operation and spectroscopy of Tm,Ho:YVO4,” Opt. Mater. 33, 937–941 (2011). [CrossRef]
15. J. Kwiatkowski, J. K. Jabczynski, L. Gorajek, W. Zendzian, H. Jelínková, J. Šulc, M. Němec, and P. Koranda, “Resonantly pumped tunable Ho:YAG laser,” Laser Phys. Lett. 6, 531–534 (2009). [CrossRef]
16. A. Wienke, F. Haxsen, D. Wandt, U. Morgner, J. Neumann, and D. Kracht, “Ultrafast, stretched-pulse thulium-doped fiber laser with a fiber-based dispersion management,” Opt. Lett. 37, 2466–2468 (2012). [CrossRef] [PubMed]
17. O. E. Martinez, J. P. Gordon, and R. L. Fork, “Negative group-velocity dispersion using refraction,” J. Opt. Soc. Am. A 1, 1003–1006 (1984). [CrossRef]
18. R. L. Aggarwal, D. J. Ripin, J. R. Ochoa, and T. Y. Fan, “Measurement of thermo-optic properties of Y3Al5O12, Lu3Al5O12, YAIO3, LiYF4, LiLuF4, BaY2F8, KGd(WO4)2, and KY(WO4)2 laser crystals in the 80300K temperature range,” J. Appl. Phys. 98, 103514 (2005). [CrossRef]
19. H. Ehlers, T. Gross, M. Lappschies, and D. Ristau, “Ion-assisted deposition processes for precision and laser optics,” Proc. SPIE 5250, 519–527 (2004). [CrossRef]
20. L. Jensen, M. Jupé, H. Mädebach, H. Ehlers, K. Starke, D. Ristau, W. Riede, P. Allenspacher, and H. Schroeder, “Damage threshold investigations of high-power laser optics under atmospheric and vacuum conditions,” Proc. SPIE 6403, 64030U (2006). [CrossRef]
21. A. Seilmeier, M. Wörner, H.-J. Hübner, and W. Kaiser, “Distortion of infrared picosecond pulses after propagation in atmospheric air,” App. Phys. Lett. 53, 2468–2470 (1988). [CrossRef]
22. Kronig R. de L., “On the theory of dispersion of X-Rays,” J. Opt. Soc. Am. 12, 547–556 (1926). [CrossRef]
23. A. Hildenbrand-Dhollande and F. R. Wagner, “Materials for Lasers: Frequency Conversion, Q-Switching, and Active Materials,” in Laser-Induced Damage in Optical Materials,E.D. D. Ristau, ed. (CRC Press, 2014).
24. K. S. Sree Harsha, Principles of Physical Vapor Deposition of Thin Films (Elsevier, 2006).