Tunable near- and mid-infrared (1.36–1.63 µm and 3.07–4.81 µm) optical vortex laser source

The generation of near- and mid-infrared vortex mode is demonstrated from a 1 µm nanosecond optical vortex pumped optical parametric oscillator using a multi-grating MgO-doped periodically poled lithium niobate crystal with five grating domains. This system enables the orbital angular momentum between the signal and idler outputs to be exchanged simply by controlling the cavity Q-factor, and a vortex output in the wavelength ranges of 1.36–1.63 µm or 3.07–4.81 µm could be obtained. A maximum signal (idler) vortex output energy of 4.3 mJ (2.2 mJ) was achieved at a pump energy of 21 mJ, which corresponds to an optical–optical conversion efficiency of over 20% (10%).

External spatial phase modulation elements, such as azimuthally continuous or segmented spiral phase plates (SPP) [25][26][27], q-plates [28][29][30], spatial light modulators [31,32], and cholesteric liquid crystal chiral superstructures [33] are commonly used as mode converters to generate an optical vortex beam. However, these elements are designed for a specified wavelength, and they inherently constrain the wavelength versatility of optical vortex beams. A widely tunable optical vortex source with a lasing wavelength that can be matched to the absorption bands of target materials is strongly desired for applications. An optical parametric oscillator (OPO) is a promising way to develop tunable solid-state lasers. In particular, a tunable optical vortex source in the near-(~1.5 µm) and mid-infrared (3-5 µm) region, in which many molecules have eigenfrequencies and their overtones originating from vibration modes [34][35][36], will open the door towards next-generation molecular sciences and applications, such as super-resolution molecular spectroscopy with a high spatial resolution beyond the diffraction limit, and organic materials processing without the destruction of chemical structures.
Here, we extend this tunable mid-infrared laser source to fill in the frequency gap (3-5 µm) of the mid-infrared vortex source. Interestingly, the system also enables the topological charges between the idler (low energy photon) or signal (high energy photon) outputs to be exchanged simply by adjusting the reflectivity of the output coupler, thereby generating a vortex output in the wavelength ranges of 1.36-1.63 µm or 3.1-4.8 µm. Such topological charge exchange based on the Q-factor control of the cavity has been never discussed, so far. Figure 1 shows a schematic diagram of the developed midinfrared optical vortex source formed of a 1 µm optical vortex pumped PPLN-OPO. A conventional Q-switched Nd:YAG laser (LS-2136LP; pulse duration, 25 ns; wavelength, 1.064 µm; pulse repetition frequency (PRF), 50 Hz; Gaussian spatial profile) was used as a pump source, and its output was converted into a first-order optical vortex with a topological charge ℓ of 1 with a SPP. The optical vortex pump beam was loosely focused onto the OPO by a lens to form an annular spot with a diameter of 1 mm.

Experiments and discussion
A nonlinear crystal used was a 5 mol% MgO-doped PPLN transversely segmented into five grating domains with periods of 26-30 µm and a step of 1 µm; the dimensions were 40 mm long, 10 mm wide and 2 mm thick. Both end faces of the crystal were anti-reflection (AR)-coated for 1.064 µm (pump), 1.3-1.7 µm (signal), and 3-5 µm (idler). The polarization of the pump beam was then aligned parallel along a crystallographic axis of the crystal to achieve type-0 (e → e + e) phase matching among the pump, signal and idler outputs. The crystal was placed in an oven, so as to control the temperature within 25 • C-200 • C with an accuracy of 0.1 • C. The crystal was also mounted onto a transverse translator to ensure the pump beam passed through an individual grating domain.
A single resonant high-Q cavity for the signal was made from a flat input mirror (IM) with high reflectivity for the signal and idler, and high transmission for the pump, and a flat output mirror (OC1) with high reflectivity for the signal output and pump beam (for double-pass pumping) and high transmission for the idler; its physical length was fixed to ca. 100 mm. The undesired pump beam was also removed by employing an AR-coated filter with a high transmission of >95% for 1.9 µm.
The nonlinear gain for the signal in this singly resonant cavity is determined by the spatial overlapping efficiency η, between the resonant signal and pump fields (non-resonant idler always exhibits the almost same beam size as that of pump beam), given as follows: The signal beam is here assumed to lase at a vortex mode in the cavity, and it can then be given by a following expression: where E p and E s are the electric fields of the pump and signal beams, respectively, and ω p and ω s are the beam radii of the pump and signal beams. Therefore, the general relationship for η is given by This formula suggests that the spatial overlapping efficiency η decreases significantly as the increase of the signal mode radius ω s . The plane-parallel cavity used in this experiment is classified into a stable-unstable resonator, and it prevents the signal output to lase at a higher-order mode, such as the vortex mode, with an infinite beam radius due to severe diffraction loss (η is almost zero). Figure 2 shows the experimental spatial forms and wavefronts of the signal and idler outputs observed using a pyroelectric camera (Spiricon Pyrocam III; spatial resolution: 75 µm) and a self-reference interferometer. The idler output exhibits an annular spatial form, and its topological charge is identical with that of the pump beam, as evidenced by a pair of Y-or fork-shaped fringes (figures 2(a) and (b)). In contrast, the signal output shows a Gaussian profile without any phase singularities (figures 2(c) and (d)). These results indicate that the OAM of the pump beam is selectively transferred into the idler output. The idler output can also be tuned within the wavelength range of 3.07-4.81 µm by controlling the crystal temperature and selecting the grating domain ( figure 3(a)). A simulated tuning curve by employing the Smelleier equation can also support well the experiments [51,52]. This tuning range is the widest, to the best of our knowledge, obtained by a mid-infrared vortex source based on an OPO. The system enables the frequency gap of the mid-infrared vortex sources to be filled. The maximum idler output energy was measured to be 2.2 mJ at a wavelength of 3.42 µm. The incident pump energy was then 21 mJ. The corresponding optical-optical conversion efficiency was estimated to be >10% ( figure 3(b)).
The idler output typically had a narrow spectral band width (FWHM) of ∆λ i ≈ 4.3 nm (ca. 1.85 cm −1 ); however, it became extremely broadband (∆λ i ≈ 23 nm, ca. 24.4 cm −1 ) near a wavelength of 3.1 µm due to the relatively wider phase matching acceptance of the crystal at high temperature ( figure 4).
The OC1 was replaced by a flat partial reflective OC2 with 50% reflectivity for the signal, high reflectivity for the pump, and high transmission for the idler. Such a low-Q cavity should act as a near optical parametric generator and significantly impact the signal to be confined in the cavity as an eigenmode (in fact, the signal output only makes one or two round trips in the cavity), so as to ensure that the signal beam holds the OAM of the pump beam. The signal output did lase at the vortex mode, as evidenced by its annular spatial form and Y-forked fringes (figures 5(a) and (b)). The signal output was also tuned into the wavelength range of 1.36-1.63 µm. The idler output had a Gaussian profile without any phase singularities (figures 5(c) and (d)). Thus, control of the cavity Q-factor enables us to exchange the OAM between the signal and idler outputs. Selective control of the OAM will therefore  become available with an electric Q-switching element, such as an acoustic-optical modulator, as a future work. It should also be noted that a maximum signal vortex output energy of 4.3 mJ was obtained at 1.51 µm with an optical-optical conversion efficiency of over 20% ( figure 6). The energy and spatial form of the vortex output also remained unchanged during a long observation time of over 3 h. Further improvement of the optical-optical conversion efficiency of this OPO system will be possible by optimization of the output mirror reflectivity.

Conclusions
We have demonstrated a widely tunable, near-(~1.5 µm) and mid-infrared (3-5 µm) vortex laser by employing an optical vortex pumped singly resonant OPO formed of MgO:PPLN crystal with five grating domains. This system enables the frequency gap of the mid-infrared optical vortex sources to be filled and it also allows the OAM between the signal and idler outputs to be exchanged simply by controlling the cavity Qfactor. Such a tunable near-and mid-infrared optical vortex source will open the door towards next-generation molecular sciences and applications, including super-resolution molecular spectroscopy, and laser materials processing of polymeric materials.