We demonstrated a high-peak-power orthogonally polarized multi-wavelength laser at 1.6-1.7 µm based on the intracavity cascaded nonlinear optical frequency conversion (CNOFC). This CNOFC can be characterized by a configuration of “series-parallel optical circuit”, where the OPO pumped Raman conversion operates in the “series-mode” and orthogonally-polarized waves are simultaneously converted in the “parallel-mode”. The fundamental wave at 1064 nm was first simultaneously converted to orthogonally-polarized 1534 and 1572 nm by the x-cut KTA and KTP optical parametric oscillation (OPO), respectively. Then the x-cut KTA and KTP acted as the Raman crystal to each other with the X(ZZ)X Raman configuration, converting the OPO signals to multi-order Raman emissions with the wavelengths at 1601, 1632, 1673, 1697, 1752 and 1767nm. A maximum total average output power of 1.2 W and a minimum pulse width of 10.5 ns were achieved from this CNOFC-based laser, corresponding to a pulse peak power of 28.5 kW and a pulse energy of 0.3 mJ.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1.6-1.7 µm waveband lasers have attracted a great deal of attention due to their wide range of potential applications in biological engineering, range finding, remote sensing, laser therapy, etc. This waveband lasers have low water absorption in biological tissue, and can enhance imaging contrast at deep penetration in biological imaging and promote glycosylation of serum protein in the treatment of inflammatory arthritis [1–3]. Remote sensing with the 1.6-1.7 µm lasers plays an important role in CO2 and CH4 detection due to the negligible Rayleigh scattering effect caused by water vapor in the atmosphere [4,5]. Furthermore, water absorption peaks near 1.7 µm decrease towards a minimum in human corneas. So this waveband laser can be used to improve the success rate of edematous corneas surgery . In recent years, various methods have been developed to generate laser emissions at 1.6-1.7 µm, such as the bismuth-doped fiber lasers (BFLs) , Er-doped solid-state lasers [8–10], and semiconductor lasers [11,12].
Converting the well-developed 1 µm laser to 1.6-1.7 µm by the cascaded nonlinear optical frequency conversions (CNOFCs) is also an effective technical means to obtain coherent light source with high peak power and cost-effectiveness. In 2011, Zhao et al. demonstrated a 1.8 µm laser based on the KTiOPO4 (KTP) optical parametric oscillator (OPO) pumped with a SrWO4 Raman laser at 1180 nm . Our group had designed a 1625 nm laser with a noncritically phase-matched (NCPM) KTP OPO intracavity pumped by a 1092 nm KTA Raman laser . Recently, Duan et al. reported that the signal light of 1742nm was achieved in an intracavity KTiOAsO4 (KTA) OPO driven by a laser diode (LD) end-pumped acousto-optic (AO) Q-switched Nd:YVO4 self-Raman laser at 1176 nm . These CNOFC-based lasers as reported above can be characterized by the configuration of OPO pumped by a Raman laser. In order to achieve the NCPM OPO, the emitting wavelength from the front Raman conversion should be elaborately selected. Moreover, this kind of CNOFC always operates at single-wavelength limited to the ultimate OPO operation.
In this paper, we demonstrated a novel CNOFC-based laser that used an OPO to pump the Raman conversion, generating high-peak-power orthogonally polarized multi-wavelength laser emissions at 1.6-1.7 µm. OPO as a pre-process can provide high-intensity pumping light for subsequent Raman conversion. Raman lasers do not require birefringent- or quasi-phase matching, allowing for better wavelength compatibility with the front OPO. This CNOFC also offers a greater flexibility in terms of output wavelength via the cascaded Raman conversions. Furthermore, a concept of “series-parallel optical circuit” was further integrated into this CNOFC, where the cascaded OPO and Raman conversion operated in a “series-mode” to achieve target wavelength conversion and orthogonally-polarized waves were simultaneously converted in the “parallel-mode” to avoid competition between multiple wavelengths. For the experimental design, the 1064 nm fundamental waves were first intracavity converted to orthogonally-polarized 1534 and 1572 nm by the x-cut KTA and KTP OPO, respectively. Then the x-cut KTA and KTP acted as the Raman crystal to each other with the X(ZZ)X Raman configuration using the strongest Raman shifts of 271 cm-1 and 234 cm-1 respectively for KTP and KTA. The OPO signals were converted to multi-order Raman emissions with the wavelengths at 1601, 1632, 1673, 1697, 1752 and 1767nm. A maximum total average output power of 1.2 W, a minimum pulse width of 10.5 ns, a pulse peak power of 28.5 kW and a pulse energy of 0.3 mJ were obtained accordingly. This CNOFC-based multi-wavelength laser at 1.6-1.7 µm will offer great potential applications in the medical imaging and atmospheric sensing. Especially, the 1767nm laser can play a crucial role in the formaldehyde detection . The multi-wavelength property may be advantageous for other applications. These multiple closely spaced wavelengths may be of use in terahertz radiation generation based on difference frequency mixing (DFM), such as 1.45 THz for the DFM of 1752 and 1767nm, and 2.54 THz for the DFM between 1673 and 1697 nm. Moreover, multi-wavelength lasers at 1.6-1.7 µm can also be used in the simultaneous multi-component trace gas detection, such as CO2 and CH4.
2. Design and experiments setup
The starting wavelength for the CNOFC-based laser was set at 1064 nm. Both the x-cut KTA and KTP crystal were used the nonlinear frequency convertors. With a 1064 nm pumped NCPM OPO configuration, the signal wavelengths of x-cut KTA and KTP would be at 1534 and 1572 nm, respectively. KTA and KTP also exhibit a large χ(3)-nonlinearity to realize efficient Raman conversion. Under the X(ZZ)X Raman configuration, the most prominent Raman shifts of KTA and KTP are respectively 234 cm−1 and 272 cm−1, enabling to convert the OPO signals to the Stokes in the 1.6-1.7 µm via multi-order Stokes shifts. Due to the cavity dumping characteristics of intracavity OPO, the pulse width of generated signals can be greatly reduced than the 1064 nm pumping pulse . In combination with the advantages of NCPM, KTA-OPO and KTP-OPO can provide high-peak-power pumping for the following KTP and KTA Raman conversion.
In the OPO operation, the 1534 nm (1572 nm) radiation with a polarization being parallel to the Y-axis of KTA (KTP) crystal could be generated by consuming the Y-polarized 1064 nm waves. Therefore, KTP and KTA were designed to be placed coaxially in the resonator with their Y-axis perpendicular to each other. The fundamental wave at 1064 nm would be first simultaneously intracavity converted to orthogonally-polarized 1534 and 1572 nm by the x-cut KTA and KTP OPO, respectively. Then the x-cut KTA and KTP acted as the Raman crystal to each other with the X(ZZ)X Raman configuration, converting the OPO signals to Raman emissions. Moreover, the output polarization of KTA Raman laser pumped by the 1572 nm KTP-OPO is orthogonal to that of the reciprocal KTP Raman laser pumped by the 1534 nm KTA-OPO. Therefore, this designed CNOFC can be characterized by the configuration of “series-parallel optical circuit”. For the series-optical-circuit, the OPO pumped Raman conversion enables the efficient wavelength conversion from 1064 nm to 1.6-1.7 µm. While in the parallel-mode, orthogonally-polarized fundamental waves and OPO signals can be simultaneously converted, avoiding competition between the multiple wavelengths.
The schematic diagram of the CNOFC with the configuration of OPO pumped Raman laser is shown in Fig. 1. The entire cavity was composed of the rear mirror (RM), AO Q-switch, Nd:YAG module, intermediate input mirror (IM), KTA crystal, KTP crystal and the output coupler (OC). The length of the cavity was 238 mm. The RM was a flat concave mirror with a concave radius of 1000 mm and was coated for high-reflection (HR) at 1064 nm (R > 99.8%). The 45-mm-long AO Q-switch had anti-reflection (AR) coatings on both faces at 1064 nm (T > 97%) and was driven at 27 MHz center frequency with the radio-frequency power of 50 W. The Nd:YAG rod (1.1 at.%, ∅3 mm × 65 mm) was placed in a water-cooled quartz glass tube and side-pumped by three laser-diode (LD) array modules. The transmittances of IM and OC were displayed in Fig. 2. IM was a plano-plane mirror which was HR coated at 1500-1800nm (R > 99%) and high-transmission (HT) coated at 1064 nm (T > 99.5%). The plane-plane OC mirror was HR coated at 1064 nm and 1500-1600 nm (R > 99.2%). The transmittances of OC mirror at the Stokes of 1601, 1632, 1673, 1697, 1752 and 1767nm were 0.1%, 0.1%, 0.5%, 3.1%, 45.4% and 47.8%. Therefore, the 1064 nm resonance could be established between RM and OC, while OPO and Raman conversion shared the same resonator comprised of IM and OC. Both the x-cut (θ = 90°, φ = 0°) KTA and KTP crystal have two dimensions of 4 × 4 × 30 mm3 and 4 × 4 × 20 mm3. Both end faces of the KTA and KTP crystals were AR coated at 1064 nm and 1500-1600 nm (R < 1%). The KTA and KTP crystals were wrapped by thin indium and placed in the water-cooled copper blocks.
3. Experimental results and discussions
The output performance of this CNOFC-based laser was initially evaluated by tuning the repetition rate of AO Q-switch. An optimized repetition rate of 4 kHz was obtained. The emission spectrum of the CNOFC-based laser at different LD pump power was measured by a spectrometer (model Q6376, Yokogawa), with the results shown in Fig. 3. An additional IM was placed behind the OC with a slight misalignment respect to the resonator axis. According to the coatings presented in Fig. 2, the fundamental wave at 1064 nm will transmit through the additional IM. As can be seen, multi-wavelengths at 1601, 1632, 1673, 1697, 1752 and 1767 nm were observed in the reflected beam. No spectral signal at 1064, 1534 and 1572 nm was observed, indicating the good spectral purity in the 1.6-1.7 µm of the measured output power. 1601 nm corresponds to the first Stokes generated by the 1534 nm pumped KTP Raman laser using the 272 cm−1 Raman shift. 1673 and 1752 nm are the second-order and third-order Stokes produced by the cascaded Raman conversion in KTP. Similarly, 1632, 1697 and 1767 nm correspond to the first, second and third Stokes generated by the 1572 nm pumped KTA Raman laser using the 234 cm−1 Raman shift. The LD pump power thresholds for the first-order, second-order, and third-order Raman conversions were 125W, 135W, and 152W, respectively. These achieved multiple wavelengths definitely indicate successful realization of the designed CNOFC. Considering the polarization of interacting waves in CNOFC, the polarization of the group of 1601, 1673, 1752 nm was orthogonal to that of group of 1632, 1697, 1767 nm. It was also found that the emitting wavelengths were dependent on the output power. Under the average output power of 0.3 W, as shown in Fig. 3(a), the first-order Raman lights at 1601 and 1632 nm dominated the output. The intensities for the two wavelengths were comparable, illustrating the advantage of the designed “parallel-mode”. While the LD pump power was increased from 153 to 184 W, the second-order Raman lights at 1673 and 1697 nm would dominate the output gradually, as shown in Fig. 3(b). When the LD pump power was 229 W, the average output power reached the maximum and the third-order Raman lights at 1752 and 1767nm dominated the output, as shown in Fig. 3(c).
Figure 4 shows the comparison of average output power for two cases, one case for the combination of 20-mm-long KTA and 20-mm-long KTP (denoted as 20-mm crystals), and the other case for the combination of 30-mm-long KTA and 30-mm-long KTP (denoted as 30-mm crystals). Due to the little wavelength separation between multi-wavelengths, it was hard for us to accurately measure the average output power at a single wavelength. So the average output powers shown in Fig. 4 were the total average output power. The case of 30-mm crystals has a higher output power and a lower threshold than that case of 20-mm crystals. A maximum average output power of 1.2 W was obtained at a LD pump power of 229 W and a pulse repetition rate (PRF) of 4 kHz in the case of 30-mm crystals. In addition, the average output power versus the PRFs at the pump power of 229 W is also shown in the inset of Fig. 4. Under the maximum average output power, the instabilities of the 1752 and 1767nm line were measured to be 1.64% and 2.49%, respectively. The average output powers of the orthogonally-polarized waves were further measured by using a linear polarizer. The corresponding power ratio between vertical- and horizontal-polarization components was 1:1.27. The weak competition between the two wavelengths further demonstrated the advantages of the configuration of “series-parallel optical circuit”, illustrating that orthogonally-polarized fundamental waves and OPO signals can be simultaneously converted to the multiple wavelengths.
The poor optical conversion efficiency from 808 nm to 1.6-1.7 µm Raman light is mainly caused by the poor spatial mode-matching between the 808 nm and 1064 nm beams in the LD side-pumping configuration. The performance of the Nd:YAG module at 1064 nm was further studied in continuous-wave (CW) and Q-switched mode, with the OC replaced by an optimized 1064 nm output coupler with a transmittance of 20%. The KTA, KTP and mirror IM were kept in the cavity to maintain the same linear losses for 1064 nm with that in the CNOFC. Under the same LD pump power with that for achieving maximum average output power in Fig. 4, the 1064 nm output powers of 15.4 and 11.6 W were obtained in the CW and Q-switching (4 kHz) mode, respectively. So an effective optical conversion efficiency of 10.3% was obtained in this CNOFC-based laser, which can be defined as the ratio between the achievable average output powers of the CNOFC and the fundamental 1064 nm Nd:YAG laser.
The pulse characteristics at different average output powers were recorded using a 1 GHz digital oscilloscope (model DSO-S104A, Keysight) and a photodetector (modle DET10D2, Thorlabs). Figure 5 shows the dependence of pulse width and corresponding peak power on the LD pump power for the case of 30-mm crystals. As the LD pump power increased, the pulse width was decreased and the corresponding peak power was increased, reaching the optimum values of 10.5 ns and 28.5 kW, respectively.
Figure 6 shows the average output powers versus the cooling temperature of nonlinear crystals at different pump powers for the case of 30-mm crystals. When the temperature was 290.8 K, the average output powers all reached the maximum. Due to the inter-coupling characteristics of cascaded nonlinear optical frequency conversion, the KTP and KTA were taken as a whole for the measurement of sensitivity of average output power on the temperature of cooling water. As shown in Fig. 6, the measured temperature tuning curves have the similar variation trend for different average output powers. When the temperature was 290.8 K, the average output powers all reached the maximum for the two cases. Since the Stokes Raman conversion does not require phase-matching, the obtained temperature tuning curves should be mainly caused by the phase-matching characteristic of OPO. For the 1064 nm pumped NCPM OPO, the 30-mm-long KTA and KTP theoretically have the phase-matching temperature width of 9 and 11 K, respectively. But the configuration of intracavity nonlinear optical frequency conversion will result in a different temperature tuning curve, as demonstrated in Ref. . The intracavity optical frequency conversions can be considered as the equivalent nonlinear loss for the fundamental wave. Then the temperature-induced phase-mismatching effect will be coupled into the dynamical behavior of fundamental laser. The varying parameters of fundamental wave and phase-matching properties of used nonlinear optical crystal will jointly determine the ultimate average output power. This cascaded nonlinear optical frequency converter involves OPO and multi-order Raman conversions. So the dependence of average output power on the temperature will present more different behavior in comparison with the extra-cavity-based OPO conversion.
Figures 7(a) and 7(c) display the results achieved under an average output power of 0.3 W for the case of 30-mm crystals, with a pulse width of 48.5 ns and corresponding pulse train at 4 kHz. When the average output power reached its maximum value of 1.2 W, a pulse width of 10.5 ns was obtained, as shown in Figs. 7(b) and 7(d). The pulse-to-pulse intensity instability was less than 6% in the pulse trains. At the maximum average output power, the single pulse energy and peak power were calculated to be 0.3 mJ and 28.5 kW, respectively. Compared to the 1.6-1.7 µm BaWO4 and KGd(WO4)2 Raman lasers [19–21], this orthogonally polarized multi-wavelength laser has a higher peak power, a narrower pulse width and a higher pulse energy.
In conclusion, we reported a high-peak-power orthogonally polarized multi-wavelength laser at 1.6-1.7 µm using a novel cascaded OPO and Raman conversion. The 1064 nm fundamental waves from Nd:YAG were first intracavity converted to orthogonally-polarized 1534 and 1572 nm by the x-cut KTA and KTP OPO, respectively. Then the x-cut KTA and KTP acted as the Raman crystal to each other with the X(ZZ)X Raman configuration, converting the OPO signals to Raman emissions. The rear Raman conversion do not require phase matching, allowing for better wavelength compatibility with the front OPO. Therefore this cascaded conversion offers a greater flexibility in terms of output wavelength via multi-order Stokes shifts. A maximum total average output power of 1.2 W, a signal pulse width was 10.5 ns, a corresponding pulse peak power of 28.5 kW and a PRF of 4 kHz were achieved.
National Natural Science Foundation of China (61605068, 61875077, 61911530131, U1730119); Applied Basic Research Programs of Xuzhou (KC17085); The Natural Science Foundation of the Jiangsu Higher Education Institutions of China (18KJA510001); Priority Academic Program Development of Jiangsu Higher Education Institutions
1. M. Tanaka, M. Hirano, K. Murashima, H. Obi, R. Yamaguchi, and T. Hasegawa, “1.7-µm spectroscopic spectral-domain optical coherence tomography for imaging lipid distribution within blood vessel,” Opt. Express 23(5), 6645–6655 (2015). [CrossRef]
2. M. Muniyappa, “Glycosylation as a marker for inflammatory arthritis,” Cancer Biomar. 14(1), 17–28 (2014). [CrossRef]
3. U. Sharma, E. W. Chang, and S. H. Yun, “Long-wavelength optical coherence tomography at 1.7 µm forenhanced imaging depth,” Opt. Express 16(24), 19712–19723 (2008). [CrossRef]
4. M. QueiBer, D. Granieri, M. Burton, A. La Spina, G. Salerno, and R. Avino, “Intercomparing CO2 amounts from dispersion modeling, 1.6 µm differential absorption lidar and open path FTIR at a natural CO2 release at Caldara di Manziana, Italy,” Atmos. Meas. Tech. 8(4), 4325–4345 (2015). [CrossRef]
5. P. Prasad, S. Rastogi, R. Singh, and S. Panigrahy, “Spectral modelling near the 1.6µm window for satellite based estimation of CO2,” Spectrochim. Acta, Part A 117, 330–339 (2014). [CrossRef]
6. F. Morin, F. Druon, M. Hanna, and P. Georges, “Microjoule femtosecond fiber laser at 1.6 µm for corneal surgery applications,” Opt. Lett. 34(13), 1991–1993 (2009). [CrossRef]
7. S. Firstov, S. Alyshev, M. Melkumov, K. Riumkin, A. Shubin, and E. Dianov, “Bismuth-doped optical fibers and fiber lasers for a spectral region of 1600-1800 nm,” Opt. Lett. 39(24), 6927–6930 (2014). [CrossRef]
8. V. Fromzel, N. Ter-Gabrielyan, and M. Dubinskii, “Acousto-optically Q-switched, resonantly pumped, Er:YVO4 laser,” Opt. Express 21(13), 15253–15258 (2013). [CrossRef]
9. N. W. Chang, N. Simakov, D. J. Hosken, J. Munch, D. J. Ottaway, and P. J. Veitch, “Resonantly diode-pumped continuous-wave and Q-switched Er:YAG laser at 1645 nm,” Opt. Express 18(13), 13673–13678 (2010). [CrossRef]
10. A. Aubourg, J. Didierjean, N. Aubry, F. Balembois, and P. Georges, “Passively Q-switched diode-pumped Er:YAG solid-state laser,” Opt. Lett. 38(6), 938–940 (2013). [CrossRef]
11. B. W. Tilma, M. S. Tahvili, J. Kotani, R. Notzel, and E. A. J. M. Bente, “Measurement and analysis of optical gain spectra in 1.6 to 1.8 µm InAs/InP (100) quantum-dot amplifiers,” Opt. Quantum Electron. 41(10), 735–749 (2009). [CrossRef]
12. Q. Gong, P. Chen, S. G. Li, Y. F. Lao, C. F. Cao, and C. F. Xu, “Quantum dot lasers grown by gas source molecular-beam epitaxy,” J. Cryst. Growth 323(1), 450–453 (2011). [CrossRef]
13. F. Bai, Q. P. Wang, Z. J. Liu, X. Y. Zhang, W. J. Sun, X. B. Wan, P. Li, G. F. Jin, and H. J. Zhang, “Efficient 1.8 µm KTiOPO4 optical parametric oscillator pumped within an Nd:YAG/SrWO4 Raman laser,” Opt. Lett. 36(6), 813–815 (2011). [CrossRef]
14. H. T. Huang, D. Y. Shen, and J. L. He, “Compact 1625-nm noncritically phase-matched KTiOPO4 optical parametric oscillator intracavity driven by the KTiOAsO4 Raman laser,” IEEE Photonics Technol. Lett. 25(4), 359–361 (2013). [CrossRef]
15. H. Y. Zhu, J. H. Guo, Y. M. Duan, J. Zhang, Y. C. Zhang, C. W. Xu, H. Y. Wang, and D. Y. Fan, “Efficient 1.7 µm light source based on KTA-OPO derived by Nd:YVO4 self-Raman laser,” Opt. Lett. 43(2), 345–348 (2018). [CrossRef]
16. H. Barry, L. Corner, and G. Hancock, “Cross sections in the 2ν5 band of formaldehyde studied by cavity enhanced absorption spectroscopy near 1.76 µm,” Phys. Chem. Chem. Phys. 4(3), 445–450 (2002). [CrossRef]
17. H. T. Huang, H. Wang, S. Q. Wang, and D. Y. Shen, “Designable cascaded nonlinear optical frequency conversion integrating multiple nonlinear interactions in two KTiOAsO4 crystals,” Opt. Express 26(2), 642–650 (2018). [CrossRef]
18. H. T. Huang and J. L. He, “A new view on the temperature insensitivity of intracavity SHG configuration,” Opt. Express 20(8), 9079–9089 (2012). [CrossRef]
19. M. Jelinek Jr., O. Kitzler, H. Jelinkova, J. Sulc, and M. Nemec, “CVD-diamond external cavity nanosecond Raman laser operating at 1.63 µm pumped by 1.34 µm Nd:YAP laser,” Laser Phys. Lett. 9(1), 35–38 (2012). [CrossRef]
20. H. N. Zhang, X. H. Chen, Q. P. Wang, X. Y. Zhang, J. Chang, L. Gao, H. B. Shen, Z. H. Cong, Z. J. Liu, X. T. Tao, and P. Li, “High-efficiency diode-pumped actively Q-switched ceramic Nd:YAG/BaWO4 Raman laser operating at 1666 nm,” Opt. Lett. 39(9), 2649–2651 (2014). [CrossRef]
21. V. I. Dashkevich and V. A. Orlovich, “Raman laser based on a KGd(WO4)2 crystal: generation of stokes components in the 1.7–1.8 µm range,” J. Appl. Spectrosc. 79(6), 975–981 (2013). [CrossRef]