Open Access
Add to BibSonomy
Abstract
We report on a Yb:LuPO4 microchip laser that is able to be passively Q-switched with few-layer WS2 or MoS2 deposited on its sapphire etalon output coupler. With 11.9 W of pump power absorbed, an average output power of 2.34 W is produced at a pulse repetition rate of 1.43 MHz with a slope efficiency of 31%, in WS2 passive Q-switching; the pulse energy, duration, and peak power are respectively 1.64 μJ, 34 ns, and 48.2 W. While passively Q-switched by MoS2, the laser can generate an average output power of 1.57 W at a repetition rate of 1.27 MHz, with the shortest pulse duration of 39 ns. Our work provides a novel simple way of making compact, reliable, high-repetition-rate pulsed lasers capable of producing multi-watt output power with several tens ns of pulse duration.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Two-dimensional (2D) transition metal dichalcogenides (TMDs), such as MoS2, WS2, MoSe2, WSe2, etc., a class of layered semiconductors possessing thickness-dependent band-gaps, have recently been recognized as promising broadband saturable absorbers for passive Q-switching or mode-locking [1, 2]. Much research work has been carried out, to explore the potential of these 2D saturable absorbers for passive Q-switching in solid-state lasers based on various rare-earth active ions. In studies reported so far on passive Q-switching of Yb-ion lasers, most work was concentrated on MoS2 [3–7]. Among these studies the best results were achieved with an Yb:LGGG mixed garnet laser, from which 0.6 W of average output power was produced with shortest pulse duration of 182 ns [3]. Apart from MoS2, other two-dimensional TMD saturable absorbers have not been widely utilized for passive Q-switching of Yb-ion lasers; the only known work was on an Yb:LSO/WS2 and an Yb:YAG/MoSe2 lasers, with the maximum output power limited to 0.3 W and the minimum pulse duration to 250 ns [8, 9].
Very recently, we demonstrated an Yb:LuPO4 miniature crystal laser passively Q-switched by a few-layer MoS2 saturable absorber, which was capable of producing output power at 1.5−2 W level, with shortest pulse duration of 61 ns [10]. In addition to higher output power and shorter pulse duration, the Yb:LuPO4/MoS2 laser also proved to be advantageous over other TMD passively Q-switched Yb-ion lasers in that it could be operated under very high output couplings. In our previous work on the Yb:LuPO4/MoS2 laser, output couplings of up to T = 60% were allowed to be utilized to generate stable passive Q-switching laser action [10], in sharp contrast to other lasers in which the output couplings employed never exceeded T = 10% [3–9].
It is desirable for a Q-switched laser to operate under high output coupling; this cannot only lead to high pulse energy and short pulse duration, but can also prevent optical damage to intracavity elements. The fact that the Yb:LuPO4/MoS2 laser can operate under high output couplings also suggests the possibility of building a very compact pulsed laser by removing the output coupler of the resonator, with an uncoated dielectric etalon serving as both the output coupler and the substrate on which the 2D TMD is deposited.
In this work, we demonstrated such an Yb:LuPO4 microchip laser that was passively Q-switched by few-layer MoS2 or WS2 deposited on a sapphire substrate, which also served as an etalon output coupler. With MoS2/WS2 acting as saturable absorber, an output power of 1.57/2.34 W could be produced at pulse repetition rate of 1.27/1.43 MHz, with pulse duration of 39/34 ns. To our knowledge, this was the first time a passively Q-switched microchip laser with 2D saturable absorber was realized without using additional mirror as the output coupler.
2. Description of experiment
The Yb:LuPO4 microchip laser was built employing a 2.7 mm long plane-parallel cavity, which was formed by a plane reflector mirror and a 0.35 mm thick sapphire etalon output coupler. The plane reflector was coated for high reflectance at 1010−1200 nm (≥ 99.8%) and for high transmittance at 975 nm (> 95%). The Yb:LuPO4 crystal was uncoated, 1.0 mm thick (along a crystallographic a axis), having an Yb-ion concentration of 1.85 × 1021 cm−3 (15 at. %); it was fixed on a thin copper heat-sink and was positioned close to the plane reflector. The few-layer MoS2 or WS2 was deposited by use of CVD technique onto one surface of the sapphire etalon coupler. A 975-nm fiber-coupled diode laser (fiber core diameter of 105 μm and NA of 0.22) was used to pump the microchip laser, its emitting beam was focused by a re-imaging unit and then was coupled through the reflector mirror into the Yb:LuPO4 crystal, with a pump beam spot radius of approximately 70 μm.
3. Results and discussion
To understand the output characteristics of the current laser having an etalon output coupler, we first discuss the transmission properties of the sapphire etalon output coupler. Taking n = 1.75 for sapphire in the 1-μm spectral region, one obtains the Fresnel reflectance R = 7.5%. With the help of a simple coupled-cavity theory [11], one can calculate the transmittance of the sapphire etalon coupler (Tet) as a function of wavelength (λ). The calculated results are plotted in Fig. 1 as Tet versus 2nL/λ, with L denoting the etalon thickness. One sees that depending on λ, the magnitude of Tet varies periodically from the minimum of 74.0% to the maximum of 100%. Clearly, laser oscillation could only occur at those wavelengths for which the quantity 2nL/λ equals half integers, and hence Tet reaches its minimum.
Fig. 1 Transmittance (Tet) versus 2nL/λ, calculated for the sapphire etalon output coupler. The quantity q on the horizontal axis represents some integer.
With the few-layer MoS2 or WS2 acting as saturable absorber, passively Q-switched laser action usually occurred at multiple wavelengths, covering a range of roughly 1001−1010 nm. This laser emission region remained almost unchanged, independent of pumping level. At low pumping levels just above threshold, the laser action would occur at a single longer wavelength; an example of lasing spectrum is shown in Fig. 2(a) for the case of WS2, measured at an absorbed pump power of Pabs = 4.0 W. With the pump power increased and higher gain available, shorter wavelength oscillation would be allowed to build up. The triple-wavelength oscillation began to appear at Pabs = 5.39 W in the case of WS2, the resulting emission spectrum is also presented in Fig. 2(a). Such kind of triple-wavelength oscillation could be maintained over the high-power operational region. The lower part of Fig. 2(a) shows a typical lasing spectrum, measured at Pabs = 11.1 W in this high-power region, for both cases of WS2 and MoS2.
Fig. 2 (a) Lasing spectra measured for the MoS2 or WS2 passively Q-switched Yb:LuPO4 microchip laser. (b) π-polarized gain cross-section curve calculated for Yb:LuPO4 crystal, showing the predicted lasing wavelength region. The grey line represents the σ-polarized gain curve for β = 0.487.
Through an estimation of the amount of gain required for reaching passive Q-switching laser threshold, one can understand the observed spectral behavior. In doing this one needs to know the characteristic parameters of these two 2D TMD saturable absorbers, including the modulation depth (ΔT) and the unsaturable absorption losses (Ans). These parameters are listed in Table 1, which were measured in our previous work [10, 12]. Another important parameter that is also needed to characterize a saturable absorber, the saturation intensity (Isat), is also listed in Table 1.
Table 1. Characteristic Parameters for the 2D TMD Saturable Absorbers and Gain-Related Parameters for the Yb:LuPO4 Microchip Laser
For the laser to reach passive Q-switching threshold, the round-trip gain must be able to balance the overall resonator losses [13], viz., 2σgNtl = −2ln(1−Li) − 2ln(1−Ans−ΔT) −ln(1−Tet). In this equation σg represents π-polarized gain cross-section of Yb:LuPO4 crystal; Nt is the Yb-ion concentration; l is the length of laser crystal; while Li denotes the single-pass internal dissipative losses.
The magnitude of Li is determined mainly by the absorption and scattering losses in the Yb:LuPO4 crystal, independent of the losses caused by the WS2 or MoS2 absorber; the latter have been included in the term of 1−Ans−ΔT. In order to evaluate the amount of Li, we carried out a Findlay-Clay analysis for the Yb:LuPO4 laser operating in continuous-wave (cw) mode, by use of the relation between the threshold pump power (Pth) and the reflectivity of output coupler (R), Pth = Pth0 + (Pth0/2Li)ln(1/R) [14]. To do this, a set of threshold pump powers were measured, with the output reflectivity changed from R = 97% to R = 30%. For this measurement, a 0.35 mm thick sapphire plate was used to replace the saturable absorber, making the resonator remain as same as possible. From this analysis, we obtain Li = 14.7%. Such a fairly large dissipative loss results primarily from the resonant absorption caused by the Yb ions in the lower level.
Using the values for the parameters, Li = 14.7%, Nt = 1.85 × 1021 cm−3, l = 0.10 cm, Tet = 74.0%, and taking the values for ΔT and Ans listed in Table 1, one can calculate the magnitude of gain cross-section: σg = 0.649 × 10−20 cm2 for the case of MoS2 saturable absorber; and σg = 0.514 × 10−20 cm2 for the WS2 saturable absorber. Plotting the gain cross-section versus wavelength, σg(λ) = βσem(λ) − (1−β)σabs(λ), by use of the emission and absorption cross-section spectra for π polarization, σem(λ) and σabs(λ) [15], one finds that to get lasing around λ = 1010 nm, the fraction of excited Yb ions should be at least βmin = 0.569 for MoS2; and βmin = 0.487 for WS2, as illustrated in Fig. 2(b). These gain related parameters are also listed in Table 1. From Fig. 2(b) one sees that the longest lasing wavelength (~1010 nm) nearly coincides with a gain peak located at 1009.1 nm; shifting toward short-wavelength side, the gain will become increased, reaching its maximum at 1001.2 nm. So at the pumping level corresponding to βmin, laser action could occur over the wavelength region of 1001.2−1009.1 nm, in agreement with the measured lasing spectrum. It should be pointed out that the reflector mirror of the resonator was coated for high reflectance only at wavelengths of λ > 1010 nm; for shorter wavelengths, the transmission loss would become increased, thus benefiting oscillation at longer wavelengths. Also because of this reason, laser action at wavelengths shorter than 1001.2 nm was not observed.
Along with the π-polarized gain cross-section curve for β = 0.487, the σ-polarized curve is also plotted for the same excitation level in Fig. 2(b), as represented by the grey line. One can see that over the lasing wavelength region predicted by the calculation, the gain for π polarization is always higher than that for σ polarization, suggesting π-polarized laser action. This proved to be in agreement with the experimental results; no σ-polarized laser oscillation was observed in the operation of the Q-switched Yb:LuPO4 laser.
The output characteristics of the passively Q-switched Yb:LuPO4 microchip laser are illustrated in Fig. 3. With WS2 serving as saturable absorber, the threshold for passive Q-switching was reached at Pabs = 3.79 W; above which the average output power increased linearly with Pabs, with a slope efficiency of 31%. A maximum output power of 2.34 W was measured at Pabs = 11.9 W, resulting in an optical-to-optical efficiency of 19.7%. In excess of this pumping level, the Q-switched operation became unstable, probably arising from thermal instability of the 2D saturable absorber, for the thermal conductivities of such materials are rather low [16]. For passive Q-switching induced by the MoS2 saturable absorber, the microchip laser arrived at threshold at Pabs = 4.85 W, considerably higher than in the case of WS2. The higher lasing threshold was attributed to the greater insertion losses, which were due to the much lower initial transmission (or much larger sum of modulation depth and unsaturable absorption loss), and might be also connected with the saturation intensity and relaxation time of the few-layer MoS2 absorber. The highest output power, attainable prior to instability setting in, was measured to be 1.57 W, with an optical-to-optical efficiency of 13.2%. A typical laser beam pattern, measured at an output power of 1.5 W, is presented as an inset to Fig. 3. The beam quality factor was measured to be M2 = 1.46 (horizontal), and M2 = 1.52 (vertical).
Fig. 3 Average output power versus Pabs, measured for the passively Q-switched Yb:LuPO4 microchip laser with few-layer WS2 or MoS2 acting as saturable absorber.
Figure 4(a) shows the variation of pulse repetition rate with pump power. For WS2 passive Q-switching, the repetition rate increased from 337 kHz, measured just above threshold, to 1.43 MHz at the highest pumping level of Pabs = 11.9 W. In the case of MoS2 passive Q-switching, the repetition rate amounted to 481 kHz at Pabs = 5.39 W, which was not far from threshold; and reached 1.27 MHz at the highest pump power. Figure 4(b) depicts the pulse energy versus Pabs, calculated from the measured average output power (Fig. 3) and pulse repetition rate. The pulse energy increased with Pabs over a low pumping range; its increase would become slower exceeding certain pumping level. At the highest pump power of Pabs = 11.9 W, the pulse energy generated was 1.64 μJ (WS2), and 1.24 μJ (MoS2).
Fig. 4 Pulse repetition rate (a) and pulse energy (b) versus Pabs, measured (calculated) for the passively Q-switched Yb:LuPO4 microchip laser.
The pulse duration was also dependent on the pumping level. As shown in Fig. 5(a), the pulse duration became shortened with increasing pump power, rapidly at the initial stage and then progressively. The shortest pulse duration was measured to be 39 ns at Pabs = 11.5 W for passive Q-switching with MoS2; and 34 ns at Pabs ≥ 11.5 W for the case of WS2 passive Q-switching. From the pulse energy and duration, the peak power can be calculated. Figure 5(b) shows the variation of peak power with pumping level. One sees that the peak power increased with Pabs, reaching, at the highest pump power, 48.2 W (WS2) and 31.0 W (MoS2).
Fig. 5 Pulse duration (a) and peak power (b) versus Pabs, measured (calculated) for the passively Q-switched Yb:LuPO4 microchip laser.
Figure 6 shows a typical pulse train generated in passively Q-switched operation with WS2 or with MoS2. The measurement was made at Pabs = 11.5 W, at which the shortest laser pulse was obtained in each case. The pulse amplitude fluctuations (rms) were calculated to be 3.6% (WS2)/3.8% (MoS2); while the timing jitters (rms) were 4.6% (WS2)/6.0% (MoS2). Over most part of the pump power range (Fig. 3), fairly stable pulsed operation could be obtained. However, the passively Q-switched laser action near threshold was found to be less stable, which could probably result from the very limited degree of absorber bleaching. Approaching the upper limit of pump power, the stability of passive Q-switching would also deteriorate, arising from the strengthened thermal effects in the few-layer WS2 or MoS2 absorber.
Fig. 6 Pulse train measured at Pabs = 11.5 W for WS2 passive Q-switching (a) and for MoS2 passive Q-switching (b). The inset in each figure shows the temporal profile of an individual pulse.
Table 2 lists the main parameters characterizing the passive Q-switching performance for the current Yb:LuPO4/WS2 (MoS2) laser and the Yb:LuPO4/GaAs, Yb:LuVO4/Cr4+:YAG lasers that were previously reported [17, 18]. One notes that while the output power (Pmax), slope efficiency (ηs), and pulse duration (tp) are comparable to some extent, the pulse repetition rate (PRR) achieved here proves to be much higher than obtainable with GaAs or with Cr4+:YAG (but much lower pulse energy (Ep) and peak power (Pp)). The capability of inducing high-repetition-rate passive Q-switching action turns out to be the major advantage of such 2D TMD materials over the traditional GaAs and Cr4+:YAG saturable absorbers. In addition to high repetition rates, the pulse duration might be further shortened to several nanoseconds by optimization of the 2D TMD saturable absorbers, given the fact that pulse durations of 4−5 ns have been obtained in a Nd:YVO4 laser with MoS2 or WS2 solution saturable absorber [19].
Table 2. Comparison of Passive Q-switching Performance between the Current Yb:LuPO4/WS2 (MoS2) Laser and the Previously Reported Yb:LuPO4/GaAs or Yb:LuVO4/Cr4+:YAG Laser [17, 18]
4. Summary
In conclusion, we have demonstrated a high-gain Yb:LuPO4 microchip laser that could be passively Q-switched by few-layer WS2 or MoS2 deposited on the sapphire etalon output coupler. With 11.9 W of pump power absorbed by the 1.0 mm thick Yb:LuPO4 crystal, an average output power of 2.34 W was generated at a repetition rate of 1.43 MHz with a slope efficiency of 31%, in the case of WS2 passive Q-switching; the pulse energy, duration, and peak power were 1.64 μJ, 34 ns, and 48.2 W, respectively. For passive Q-switching with MoS2, a maximum output power of 1.57 W was produced at 1.27 MHz with a slope efficiency of 23%; the shortest pulse duration measured in this case was 39 ns. Our work provides a novel simple way of making, based on high-gain laser media and 2D saturable absorbers, compact, reliable, high-repetition-rate pulsed lasers capable of producing multi-watt output power with several tens nanoseconds of pulse duration.
Funding
National Natural Science Foundation of China (11574170).
References and links
1. R. I. Woodward, R. C. T. Howe, G. Hu, F. Torrisi, M. Zhang, T. Hasan, and E. J. R. Kelleher, “Few-layer MoS2 saturable absorbers for short-pulse laser technology: current status and future perspectives,” Photon. Res. 3(2), A30–A42 (2015). [CrossRef]
2. R. I. Woodward, R. C. T. Howe, T. H. Runcorn, G. Hu, F. Torrisi, E. J. R. Kelleher, and T. Hasan, “Wideband saturable absorption in few-layer molybdenum diselenide (MoSe2) for Q-switching Yb-, Er- and Tm-doped fiber lasers,” Opt. Express 23(15), 20051–20061 (2015). [CrossRef] [PubMed]
3. F. Lou, R. Zhao, J. He, Z. Jia, X. Su, Z. Wang, J. Hou, and B. Zhang, “Nanosecond-pulsed, dual-wavelength, passively Q-switched ytterbium-doped bulk laser based on few-layer MoS2 saturable absorber,” Photon. Res. 3(2), A25–A29 (2015). [CrossRef]
4. Y. Zhan, L. Wang, J. Y. Wang, H. W. Li, and Z. H. Yu, “Yb:YAG thin disk laser passively Q-switched by a hydro-thermal grown molybdenum disulfide saturable absorber,” Laser Phys. 25(2), 025901 (2015). [CrossRef]
5. Y. Sun, J. Xu, S. Gao, C. Lee, H. Xia, Y. Wang, Z. You, and C. Tu, “Wavelength-tunable, passively Q-switched Yb3+:Ca3Y2(BO3)4 solid state laser using MoS2 saturable absorber,” Mater. Lett. 160, 268–270 (2015). [CrossRef]
6. Y. Sun, J. Xu, Z. Zhu, Y. Wang, H. Xia, Z. You, C. Lee, and C. Tu, “Comparison of MoS2 nanosheets and hierarchical nanospheres in the application of pulsed solid-state lasers,” Opt. Mater. Express 5(12), 2924–2932 (2015). [CrossRef]
7. J. M. Serres, P. Loiko, X. Mateos, H. Yu, H. Zhang, Y. Chen, V. Petrov, U. Griebner, K. Yumashev, M. Aguiló, and F. Díaz, “MoS2 saturable absorber for passive Q-switching of Yb and Tm microchip lasers,” Opt. Mater. Express 6(10), 3262–3273 (2016). [CrossRef]
8. J. Liu, J. Tian, H. Guoyu, R. Xu, K. Li, Y. Song, X. Zhang, L. Su, and J. Xu, “Passively Q-switched dual-wavelength Yb:LSO laser based on tungsten disulphide saturable absorber,” Chin. Phys. B 25(3), 034207 (2016). [CrossRef]
9. J. Ma, W. Jiang, C. Shen, and S. Yuan, “Passively Q-switched Yb:YAG laser based on a MoSe2 saturable absorber,” Appl. Opt. 57(8), 1958–1961 (2018). [CrossRef] [PubMed]
10. X. Dou, J. Yang, M. Zhu, H. Xu, W. Han, D. Zhong, B. Teng, and J. Liu, “Watt-level passively Q-switched Yb:LuPO4 miniature crystal laser with few-layer MoS2 saturable absorber,” Opt. Express 26(11), 14232–14240 (2018). [CrossRef] [PubMed]
11. A. E. Siegman, Lasers (University Science Books, 1986).
12. X. Dou, Y. Ma, M. Zhu, H. Xu, D. Zhong, B. Teng, and J. Liu, “Multi-watt sub-30 ns passively Q-switched Yb:LuPO4/WS2 miniature laser operating under high output couplings,” Opt. Lett. 43(15), 3666–3669 (2018). [CrossRef]
13. O. Svelto, Principles of Lasers (Springer, 2010).
14. J. Liu, J. Wang, Y. Liu, J. Wei, C. Wang, Z. Shao, and M. Jiang, “Performance of CW NYAB laser at 1.06 μm end-pumped by a high-power diode-laser-array,” Opt. Laser Technol. 32(3), 183–186 (2000). [CrossRef]
15. J. Liu, W. Han, X. Chen, D. Zhong, B. Teng, C. Wang, and Y. Li, “Spectroscopic properties and continuous-wave laser operation of Yb:LuPO4 crystal,” Opt. Lett. 39(20), 5881–5884 (2014). [CrossRef] [PubMed]
16. B. Chen, X. Zhang, K. Wu, H. Wang, J. Wang, and J. Chen, “Q-switched fiber laser based on transition metal dichalcogenides MoS2, MoSe2, WS2, and WSe2.,” Opt. Express 23(20), 26723–26737 (2015). [CrossRef] [PubMed]
17. X. Dou, L. Wang, W. Han, H. Xu, D. Zhong, B. Teng, and J. Liu, “Near-IR 1-μm high-repetition-rate pulsed radiation generated with an Yb:LuPO4 miniature crystal rod laser,” Opt. Commun. 420, 90–94 (2018). [CrossRef]
18. X. D. Dou, J. N. Yang, Y. J. Ma, W. J. Han, H. H. Xu, and J. H. Liu, “Passive Q-switching of a Yb:LuVO4 laser with Cr4+:YAG: approaching the intrinsic upper limit of repetition rate,” Chin. Phys. Lett. 35(6), 064201 (2018). [CrossRef]
19. H. Liu, Z. Sun, X. Wang, Y. Wang, and G. Cheng, “Several nanosecond Nd:YVO4 lasers Q-switched by two dimensional materials: tungsten disulfide, molybdenum disulfide, and black phosphorous,” Opt. Express 25(6), 6244–6252 (2017). [CrossRef] [PubMed]
