We demonstrate efficient amplification of few-optical-cycle mid-IR pulses in single-pass continuously pumped laser amplifiers based on polycrystalline Cr2+:ZnS and Cr2+:ZnSe. The 1.7 W output of a Kerr-lens mode-locked master oscillator at 2.4 µm central wavelength, 79 MHz repetition rate was amplified to 7.1 W and 2.7 W in Cr2+:ZnS and Cr2+:ZnSe, respectively. High peak power of the input pulses (0.5 MW) and high nonlinearity of the amplifiers’ gain media resulted in a significant shortening of the output pulses and in spectral broadening. Transform-limited 40 fs pulses of the master oscillator were compressed to about 27–30 fs. The spectrum of the pulses was broadened from 136 nm to 450 nm (at −3 dB level); the span of the spectra exceeds 600 nm at −10 dB level.
© 2016 Optical Society of America
Middle-infrared (mid-IR) lasers are in great demand for a variety of applications in life sciences and technology. Availability of femtosecond (fs) mid-IR laser sources with high average power and broad spectral span is an important prerequisite for high dynamic range molecular spectroscopy and hyperspectral imaging with the applications that range from fundamental science to environmental sensing, medicine, and industrial process control.
Transition-metal-doped II-VI semiconductors (TM:II-VI), introduced in the late 1990s , represent a viable route to high-power mid-IR lasing. TM:II-VI materials feature a favorable blend of parameters: a four-level energy structure, an absence of excited state absorption, close to 100% quantum efficiency of fluorescence (for Cr doped II-VI media), broad vibronic absorption and emission bands. Cr2+ and Fe2+ doped ZnS and ZnSe are the typical and most well-studied representatives of TM:II-VI family. Recent achievements in TM:II-VI laser technology include pulsed Fe2+:ZnSe laser with 35 W average power at 4.1 µm wavelength and continuous wave (cw) Cr2+:ZnSe laser with power in excess of 50 W at 2.4 µm wavelength [2, 3].
Cr2+ doped ZnS and ZnSe enable efficient room-temperature lasers with broad tuning over 1.9–3.3 µm range [2, 4]. Very conveniently, Cr2+:ZnS/ZnSe can be pumped by efficient and reliable erbium (Er) and thulium (Tm) fiber lasers. Cr2+:ZnS and Cr2+:ZnSe have similar spectroscopic and laser parameters. Cr2+:ZnS is regarded superior for high power laser applications due to better physical properties, while Cr2+:ZnSe has higher nonlinearity. Both materials are available in single crystal and in polycrystalline forms. In most cases, polycrystalline laser gain elements allow to obtain superior laser parameters due to their better optical quality and higher dopant concentration .
Broad emission bands of TM:II-VI materials are favorable for generation of ultra-short mid-IR pulses. Ability to operate at room-temperature and availability of high power fiber lasers for optical pumping stimulated rapid progress of ultra-fast Cr2+:ZnS/ZnSe lasers, as reviewed in . Femtosecond Cr2+:ZnS and Cr2+:ZnSe lasers based on all major mode-locking techniques have been successfully implemented over the past decade including SESAM [6, 7], Kerr-lens [8, 9], and graphene [10, 11] mode-locked lasers. Recent demonstration of Kerr-lens mode-locked (KLM) lasers based on polycrystalline Cr2+:ZnS and Cr2+:ZnSe  has led to significant improvements in the output parameters of ultrafast mid-IR lasers in terms of average power (2 W) , pulse energy (24 nJ) , and pulse duration (≤29 fs) . Thus, mode-locked Cr doped ZnSe and ZnS oscillators have come of age, and, arguably, represent an appealing alternative to complex and inefficient ultra-fast mid-IR sources based on down-conversion of near-IR lasers, e.g. synchronously pumped optical parametric oscillators (OPO), optical parametric amplifiers (OPA) and difference frequency generation setups (DFG).
Zinc-blende semiconductors exhibit rather high second-order nonlinear susceptibility: 16 pm/V and 4 pm/V in ZnSe and ZnS, respectively, vs 15 pm/V in LiNbO3 at 2.4 µm wavelength . Polycrystalline Cr2+:ZnS and Cr2+:ZnSe gain media consist of a multitude of microscopic single-crystal grains. The broad distribution of grain sizes and orientations results in so-called random quasi-phase-matching (RQPM). The main distinctive features of RQPM are a linear dependence of the conversion yield with length of the medium and ultra-wide bandwidth [16, 17].
Another set of opportunities arises from high third order nonlinearly of ZnSe and ZnS. Nonlinear refractive indexes of these materials are approximately 50 times larger than that of CaF2. The critical power for self-focusing is about 400 kW in bulk ZnSe and 430 kW in ZnS at 2.4 µm wavelength. Filamentation of mid-IR fs pulses in ZnSe and ZnS has attracted recent attention. The formation of filaments has been observed in ZnSe pumped well above the critical power by a kilohertz OPA tunable from 0.4 µm to 2.5 µm with 6–20 µJ pulse energy and ~30 fs pulse duration . It was shown that the long-wavelength pumping is very favorable for the formation of filaments in ZnSe due to higher multiplicity of the multi-photon ionization (2.7 eV bandgap of ZnSe corresponds to six-photon ionization at 2.4 µm pump wavelength). The spectral broadening of about 450 nm was observed in ZnSe at 1.9–2.0 µm pump wavelength in five-photon ionization regime. More recently, 3-octave-spanning continuum has been obtained from filaments in ZnS (3.9 eV bandgap) pumped at 2.1 µm by a kilohertz 11 µJ, 27 fs OPCPA at about 1000 times the critical power for self-focusing .
Thus, the superb ultra-fast laser capabilities of Cr2+:ZnS and Cr2+ZnSe in combination with high second- and third-order nonlinearity of ZnS and ZnSe offer a number of unique possibilities. RQPM in polycrystalline Cr2+:ZnS and Cr2+ZnSe results in efficient second harmonic generation (SHG) directly in the gain element of KLM laser. Obtained SHG power (up to 250 mW) is high enough for some real-world applications . The broad spectrum of SHG pulses (up to 22 THz ,) suggests that they can be compressed to rather short durations. Other observed three-wave mixing effects include generation of third and fourth optical harmonics and sum frequency generation between mid-IR pulses and a cw pump beam. Obtained MW-level optical power inside the resonators of polycrystalline Cr2+:ZnS/ZnSe KLM lasers opens an avenue for extension of ultra-fast laser oscillations beyond 3 µm range, e.g. via DFG or optical parametric oscillations in the gain element of KLM laser. On the other hand, high third-order nonlinearity of Cr2+:ZnS and Cr2+ZnSe may facilitate the implementation of KLM lasers with an octave-spanning spectra, e.g. strong spectral broadening arising from χ(3) was reported for polycrystalline Cr2+:ZnS KLM laser with 1.3 MW peak power inside the resonator, which is 3 times higher than the critical power for self-focusing .
2. Experimental setup and results
In this paper we report high average power, high repetition rate fs laser MOPAs based on polycrystalline Cr2+:ZnS and Cr2+ZnSe gain/nonlinear media. The experimental setup is illustrated in Fig. 1(a). As a master oscillator (MO) we used polycrystalline Cr2+:ZnS KLM laser, which is described in detail in . MO is optically pumped at 1567 nm by 8.4 W linearly polarized Er- doped fiber laser (EDFL) radiation. MO was operated at 79 MHz repetition rate with 22 nJ pulse energy, 1.7 W average power, and 40 fs pulse duration. The spectrum of MO pulses was sech2 transform limited and centered at 2380 nm.
The amplifier stage is based on Cr2+:ZnSe (Cr2+ZnS) optically pumped at 1567 nm by another EDFL. AR coated polycrystalline Cr2+:ZnS and Cr2+:ZnSe gain elements were mounted at normal incidence between two curved mirrors and were cooled with room temperature water. Both gain elements were 9 mm long with ~5% low-signal transmission at the pump wavelength. The pump beam (Ø3.8 mm) was focused in the gain element by a lens (L) with 80 mm focal length (focal spot Ø50 µm). MO and pump emissions were superimposed on a curved mirror HR(1) with high reflectivity in 2100–2700 nm range and high transmission at 1567 nm. The same mirror was used for focusing of MO emission in the gain element. The output beam was collimated by a curved mirror with protected gold coating and separated from the residual pump by another HR(1). CaF2 wedges on a flip mounts were used to split a part of the output for spectral and autocorrelation measurements.
Dispersion of the MOPA was controlled within 2200–2700 nm range. Input pulses were pre-chirped to compensate for dispersion in the master laser’s output coupler (OC, 3.2 mm thick ZnSe substrate). The pre-chirping optics included: the high-reflector HR(1) with negative group delay dispersion GDD ≈-200 fs2 and TOD compensator HR* (TOD ≈-3000 fs2). To compensate for dispersion in the amplifier’s gain element, we de-chirped output pulses prior to their delivery to an autocorrelator, as shown in Fig. 1(b). Output dispersion control included second dispersive high reflector HR(1); dichroic mirror (DM) to separate SHG radiation, second TOD compensator HR*, and a combination of plane-parallel YAG plates.
Dispersion metrology in 2–3 µm range was not available. Therefore, dispersion of the gain elements, OC was calculated using the standard Sellmeier equations for undoped ZnS, ZnSe and YAG; the theoretical GDD spectra for the mirrors were provided by the coaters. Net GDD spectra of the MOPAs based on Cr2+:ZnS and Cr2+ZnSe are shown in Fig. 2.
The spectra are calculated taking into account the pre-chirping components as well as the de-chirping components, which were installed before the autocorrelator. Figure 2 also shows dispersion curves for 9 mm thick Cr2+:ZnS/ZnSe gain elements and for 3.2 mm thick ZnSe substrate of the OC.
Spectral parameters of the MOPA were characterized using a 0.15 m dual grating monochromator. The spectra were acquired using a lock-in amplifier and a PbSe detector (Thorlabs PDA20H). We did not post-process the spectra to account for wavelength dependences of the detector’s responsivity and of the grating diffraction efficiency. Temporal parameters of the MOPA were evaluated using an interferometric autocorrelator (A·P·E GmbH). The experiments were carried out in standard lab environment at about 60% air humidity.
During the preliminary experiments we evaluated spectral broadening in bulk polycrystalline Cr2+:ZnSe and Cr2+:ZnS in passive regime, i.e. without optical pumping of the gain elements. Pulses from the master laser with 22 nJ energy and 40 fs duration (∼480 kW peak power) were tightly focused (~Ø20 µm spot) in 9-mm long samples. The focusing was adjusted to obtain broadest spectra of output pulses. Results of the experiment are summarized in Fig. 3.
As can be seen, pumping of polycrystalline Cr2+:ZnSe and Cr2+:ZnS by fs mid-IR pulses at just above the critical power results in appearance of the wings in pulses spectra due to self-phase modulation. The broadening is about 10–20 nm (7–15%) at −3 dB and exceeds 150 nm (60%) at −10 dB. We explain the red-shift of the output spectra by absorption on Cr2+ centers.
In the consecutive experiments we engaged the optical pumping and characterized the output parameters of the Cr2+:ZnSe and Cr2+:ZnS amplifiers at different pump power levels. The size of the focal spot of the input beam from MO was increased to about Ø80 µm for better mode-matching with the pump beam.
Measured spectra of output pulses are shown in Fig. 4. At first we measured the spectra without pumping (dashed curves) and then we gradually increased the pump power (continuous curves). Maximum pump power was about 10 W for Cr2+:ZnSe amplifier (was limited by optical damage of AR coatings of the gain element) and about 20 W for Cr2+:ZnS amplifier. The spectra obtained at maximum pump power, are shown by thick solid lines. Numbers near the spectra correspond to measured average power at the amplifiers’ output.
As can be seen, cw optical pumping of Cr2+:ZnSe and Cr2+:ZnS results in simultaneous amplification of fs pulses (like in conventional laser amplifier) and in their spectral broadening. Remarkably, the spectra of pulses became broader with an increase of cw pump power (e.g. an increase of the amplifier’s gain). Thus, gain narrowing in the amplifier is overwhelmed by nonlinear interaction of pulses with the amplifier’s gain medium. As a result, we obtained amplified pulses with 450 nm broad spectra (at full width half maximum, FWHM); the spectral span exceeds 600 nm at −10 dB.
Measured autocorrelations (ACs) of output pulses of Cr2+:ZnSe and Cr2+:ZnS MOPAs are shown in Fig. 5(a) and Fig. 5(b), respectively. Left and right graphs compare the ACs of ‘cold’ and ‘hot’ amplifiers, i.e. measured without pumping and at maximum pump power. Top and bottom graphs compare the ACs acquired with and without output dispersion compensation. The background noise in the ACs of ‘hot’ amplifiers is due to residual signal at pump wavelength. Duration of output pulses Δτ was estimated by sech2 fit of the ACs using the autocorrelator’s control software.
Comparison of the ACs of ‘cold’ amplifiers, which were obtained with and without output dispersion compensation, shows that the dispersion management of the MOPAs is fairly good but not perfect: 40 fs input pulses were re-compressed to 45 fs output pulses. Comparison of the ACs of ‘cold’ and ‘hot’ amplifiers allow us to conclude that the amplification of fs pulses in single-pass Cr2+:ZnSe/ZnS amplifier is supplemented by significant decrease of pulse duration from 45 fs to about 30 fs. Most likely, the actual pulses are shorter: (i) some temporal broadening is introduced by the beam splitter of our autocorrelator (see Fig. 6 and discussion in ); (ii) dispersion control of the MOPA can be further improved; (iii) one can estimate 16 fs pulses taking into account 450 nm broad spectra at 2300 nm central wavelength and assuming, e.g., 0.4 time-bandwidth product.
Parameters of Cr2+:ZnSe and Cr2+:ZnS MOPAs as functions of pump power PPUMP are illustrated in Fig. 6. Figures 6(a)–6(c) show respectively the amplifier’s gain, spectral bandwidth at −10 dB level, and approximate pulse duration. The gain was defined as G = POUT(PPUMP)/POUT(0) there POUT is output power of the MOPA. The gain was measured in mode-locked regime of the master oscillator (shown in Fig. 6(a) by symbols) and in cw regime (shown in Fig. 6(a) by dashed lines). As can be seen, the gains of fs and cw MOPAs almost coincide in a range of pump powers. Increase of the pump power within this range results in gradual broadening the spectra and decrease of pulse duration. At certain pump power the spectral width reaches maximum as shown in Fig. 6(b) by asterisks (P*PUMP = 6.5 W for Cr2+:ZnSe, P*PUMP = 17.5 W for Cr2+:ZnS). Pumping of the fs amplifier above this level results in: (i) roll-off of the gain (if compared to cw amplifier); (ii) some decrease and then constancy of the spectral width; (iii) constancy and then some increase of the pulse duration.
Cr2+:ZnSe and Cr2+:ZnS gain elements were tested in similar conditions, which allows for some comparisons. Figure 6 shows that spectral broadening in Cr2+:ZnSe is more pronounced, which can be attributed to higher nonlinearity of the material. On the other hand, the gain roll-off in Cr2+:ZnSe occurs at relatively low pump power and limits the output power of the amplifier in fs regime. Probably, lower roll-off threshold in Cr2+:ZnSe can be explained by relatively narrow bandgap (2.7 eV in ZnSe vs 3.6 eV in ZnS) and, hence, lower multiplicity of the multi-photon ionization. Further study is required to confirm the relationship between the multi-photon ionization in Cr2+:ZnSe/ZnS and observed roll-off effects.
Output characteristics of Cr2+:ZnSe and Cr2+:ZnS mid-IR fs MOPAs are summarized in Table 1. Two sets of parameters in the table correspond to the broadest spectra and to the highest average power of output pulses. Pumping of Cr2+:ZnSe amplifier was limited to 10.3 W due to the damage of AR coatings. Pump power of Cr2+:ZnS amplifier was limited by available EDFL power.
In summary, we implemented simple and robust single-pass fs mid-IR amplifiers based on polycrystalline Cr2+:ZnSe and Cr2+:ZnS. The amplifiers are pumped by low-cost cw EDFL, operate at full repetition rate of the master oscillator, and feature simultaneous amplification, spectral broadening, and compression of input pulses. We obtained output pulses with 7.1 W average power (at 79 MHz repetition rate), 574 nm broad spectrum (at −10 dB), and 33 fs pulse duration (conservative estimate) from the single-pass Cr2+:ZnS amplifier. Conversion efficiency of cw pump radiation at 1.57 µm to few-optical-cycle mid-IR pulses at 2.4 µm was about 26% (31% of the absorbed pump power). Broader spectrum (> 600 nm) and shorter pulses (≤ 27 fs) with 2.8 W average power were obtained from the Cr2+:ZnSe amplifier. To the best of our knowledge these are the shortest pulses with the highest average power reported to-date in the mid-IR.
There is little doubt that the output characteristics of the single-pass Cr2+:ZnSe/Cr2+:ZnS fs amplifiers can be further significantly improved by optimization of the input pulses (size of the beam, pre-chirping) and of the gain element (length, Cr2+ concentration). Very likely, better dispersion control of the amplifier will lead to rather short output pulses: 16 fs pulse duration (two optical cycles) can be estimated from the spectra assuming 0.4 time-bandwidth product. Synchronous pumping of single-pass Cr2+:ZnSe/Cr2+:ZnS amplifiers by available 1.5 μm picosecond and femtosecond fiber lasers may result in even shorter pulses with broader spectral span and higher pulse energy. Rigorous model of the amplifier, which takes into account an interplay between laser and nonlinear interactions of fs pulses with the gain medium, would be very helpful for mentioned above optimizations and improvements.
Availability of high power fs lasers at 2 – 3 µm is an important prerequisite for the development of mid-IR sources that span the spectrum over 3 – 10 µm. Approaches to high power coherent sources, which cover the whole mid-IR range include multi-octave continuum generation in bulk dielectrics and semiconductors , synchronously pumped OPO  and DFG  setups.
Sergey Mirov declares competing financial interests.
References and links
1. L. D. DeLoach, R. H. Page, G. D. Wilke, S. A. Payne, and W. F. Krupke, “Transition metal-doped zinc chalcogenides: spectroscopy and laser demonstration of a new class of gain media,” IEEE J. Quantum Electron. 32(6), 885–895 (1996). [CrossRef]
2. S. Mirov, V. Fedorov, D. Martyshkin, I. Moskalev, M. Mirov, and S. Vasilyev, “Progress in mid-IR lasers based on Cr and Fe doped II-VI chalcogenides,” IEEE J. Sel. Top. Quantum Electron. 21(1), 1601719 (2015). [CrossRef]
3. S. Mirov, V. Fedorov, D. Martyshkin, I. Moskalev, M. Mirov and S. Vasilyev, “High average power Fe:ZnSe and Cr:ZnSe mid-IR solid state lasers,” in Advanced Solid-State Lasers, OSA Technical Digest (online) (Optical Society of America, 2015), paper AW4A.1.
4. E. Sorokin, I. T. Sorokina, M. S. Mirov, V. V. Fedorov, I. S. Moskalev, and S. B. Mirov, “Ultrabroad continuous-wave tuning of ceramic Cr:ZnSe and Cr:ZnS lasers,” in Lasers, Sources and Related Photonic Devices, OSA Technical Digest Series (CD) (Optical Society of America, 2010), paper AMC2.
5. I. T. Sorokina and E. Sorokin, “Femtosecond Cr2+-based lasers,” IEEE J. Sel. Top. Quantum Electron. 21(1), 1601519 (2015). [CrossRef]
6. T. Sorokina, E. Sorokin, and T. Carrig, “Femtosecond pulse generation from a SESAM mode-locked Cr:ZnSe laser,” in Conference on Lasers and Electro-Optics/Quantum Electronics and Laser Science Conference and Photonic Applications Systems Technologies, Technical Digest (CD) (Optical Society of America, 2006), paper CMQ2. [CrossRef]
9. E. Sorokin, N. Tolstik, and I. T. Sorokina, “1 Watt femtosecond mid-IR Cr:ZnS laser,” Proc. SPIE 8599, 859916 (2013).
10. M. N. Cizmeciyan, J. W. Kim, S. Bae, B. H. Hong, F. Rotermund, and A. Sennaroglu, “Graphene mode-locked femtosecond Cr:ZnSe laser at 2500 nm,” Opt. Lett. 38(3), 341–343 (2013). [CrossRef] [PubMed]
13. S. Vasilyev, M. Mirov, and V. Gapontsev, “Mid-IR Kerr-lens mode-locked polycrystalline Cr2+:ZnS laser with 0.5 MW peak power,” in Advanced Solid State Lasers, OSA Technical Digest (online) (Optical Society of America, 2015), paper AW4A.3.
14. S. Vasilyev, I. Moskalev, M. Mirov, S. Mirov, and V. Gapontsev, “Three optical cycle mid-IR Kerr-lens mode-locked polycrystalline Cr2+:ZnS laser,” Opt. Lett. 40(21), 5054–5057 (2015). [CrossRef] [PubMed]
15. H. P. Wagner, M. Kühnelt, W. Langbein, and J. M. Hvam, “Dispersion of the second-order nonlinear susceptibility in ZnTe, ZnSe, and ZnS,” Phys. Rev. B 58(16), 10494–10501 (1998). [CrossRef]
16. E. Yu. Morozov, A. A. Kaminskii, A. S. Chirkin, and D. B. Yusupov, “Second optical harmonic generation in nonlinear crystals with a disordered domain structure,” JETP Lett. 73(12), 647–650 (2001). [CrossRef]
17. M. Baudrier-Raybaut, R. Haïdar, P. Kupecek, P. Lemasson, and E. Rosencher, “Random quasi-phase-matching in bulk polycrystalline isotropic nonlinear materials,” Nature 432(7015), 374–376 (2004). [CrossRef] [PubMed]
19. H. Liang, P. Krogen, R. Grynko, O. Novak, C.-L. Chang, G. J. Stein, D. Weerawarne, B. Shim, F. X. Kärtner, and K.-H. Hong, “Three-octave-spanning supercontinuum generation and sub-two-cycle self-compression of mid-infrared filaments in dielectrics,” Opt. Lett. 40(6), 1069–1072 (2015). [CrossRef] [PubMed]
20. S. Vasilyev, M. Mirov, and V. Gapontsev, “High power Kerr-lens mode-locked femtosecond mid-IR laser with efficient second harmonic generation in polycrystalline Cr2+:ZnS and Cr2+:ZnSe,” in Advanced Solid State Lasers, OSA Technical Digest (online) (Optical Society of America, 2014), paper AM3A.3.
21. V. O. Smolski, S. Vasilyev, P. G. Schunemann, S. B. Mirov, and K. L. Vodopyanov, “Cr:ZnS laser-pumped subharmonic GaAs optical parametric oscillator with the spectrum spanning 3.6-5.6 μm,” Opt. Lett. 40(12), 2906–2908 (2015). [CrossRef] [PubMed]
22. I. Pupeza, D. Sánchez, J. Zhang, N. Lilienfein, M. Seidel, N. Karpowicz, T. Paasch-Colberg, I. Znakovskaya, M. Pescher, W. Schweinberger, V. Pervak, E. Fill, O. Pronin, Z. Wei, F. Krausz, A. Apolonski, and J. Biegert, “High-power sub-two-cycle mid-infrared pulses at 100 MHz repetition rate,” Nat. Photonics 9, 721–724 (2015).