Development of Tabletop Femtosecond Vacuum Ultraviolet Laser Source Based on Four-Wave Mixing Techniques
Li Zijing, Yan Lili, Zuo Peng, Xie Liangyue, Li Zhiqiang, and Jin Bing
Significance Femtosecond laser technology plays an important role in the study of ultrafast dynamics of light-induced reactions. Many ultrafast spectroscopy techniques, such as transient absorption spectroscopy, ultrafast Raman spectroscopy, and ultrafast photoelectron spectroscopy/imaging, are widely used in scientific research in the fields of physics, chemistry, biology, and materials science.At present, the laser wavelength range produced by mature commercial femtosecond lasers is mainly limited to infrared, visible, and ultraviolet (UV) bands. When the absorption spectrum or ionization energy of a sample is in the vacuum ultraviolet (VUV) band below the wavelength of 200 nm (~6 eV), commercial femtosecond laser pulses are insufficient for achieving single-photon excitation/ionization of a sample. The two-photon or multiphoton absorption of long-wavelength lasers leads to low excitation/ionization efficiency compared with the single-photon process.In the past two decades, the technology for developing a miniaturized tabletop femtosecond VUV laser source using commercial femtosecond lasers (such as Ti:sapphire laser) in laboratories has advanced rapidly. This review briefly introduces four-wave mixing (FWM) techniques, which are widely used in a tabletop femtosecond VUV laser source. This work mainly focuses on the development of FWM in gas-filled hollow fibers and filaments.Progress An early femtosecond VUV laser system was capable of producing tunable femtosecond VUV pulses by two-photon near-resonant four-wave difference-frequency mixing in argon (Fig. 3). To generate VUV pulses using the four-wave difference-frequency mixing scheme, high-intensity femtosecond laser pulses are required as a driving source. Moreover, only a part of the incident spectrum can contribute effectively to the frequency-mixing process, thereby leading to spectrally narrowed and temporally lengthened VUV pulses. Therefore, this near-resonant requirement limits the phase-matching bandwidth, tunability, pulse width, and conversion efficiency.FWM is achieved by converting the frequency of ultrashort-pulse Ti:sapphire laser pulses from visible light into deep UV light using a hollow-fiber geometry (Fig. 4). Collinear phase matching using off-resonant χ(3) processes in a hollow fiber to generate VUV light is more efficient and generates broader bandwidths than past schemes. It is confirmed that the conversion efficiency can be significantly improved by exciting higher-order transverse modes and coating the inner surface of the hollow fiber with aluminum.Another method for producing ultrafast VUV pulses is developing FWM in a filament (Fig. 6, Fig. 7, and Fig. 9). Typically, considerably more energy-driving pulses can be used in a filament than in a hollow fiber. Furthermore, the alignment of two laser beams, e.g., the third harmonic and fundamental of a Ti:sapphire laser, in a gas cell is considerably easier than in a narrow hollow fiber. Intensity clamping and mode-cleaning effects of filamentation provide stable and spatially clean output pulses.In contrast to the above-mentioned FWM schemes, VUV pulses with remarkable high pulse energies can be generated via a third-harmonic generation process (Fig. 10). However, the conversion efficiency of the high-harmonic generation process from fundamental radiation is low, and it is desirable to avoid such a significant loss of VUV pulse energy. A short-wavelength driving laser eases this difficulty.To acquire tunable femtosecond VUV pulses, the use of optical parametric amplifier (OPA) or noncollinear optical parametric amplifier (NOPA) systems is considered (Fig. 11). Although it is convenient to produce continuously tunable femtosecond pulses using OPA or NOPA systems, the wavelength range covered by continuously tunable tabletop VUV lasers is limited.Conclusions and Prospects In the past two decades, the development of tabletop femtosecond VUV laser sources has made great progress. The demand for developing femtosecond VUV laser sources is increasing in tandem with the advancement of scientific research and application. In the future, it is critical to improve the frequency up-conversion efficiency of the high-harmonic generation/FWM process and continuously investigate the development and application of new nonlinear media.
  • Jun. 11, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201007 (2021)
  • DOI:10.3788/CJL202148.1201007
Development of High-Power Ultrafast Fiber Laser Technology
Liu Yizhou, Qiao Wenchao, Gao Kong, Xu Rong, Feng Tianli, Zhang Meng, Li Xun, Liang Yangyang, and Li Tao
Significance In 1960, after the invention of the first ruby laser, fast-developed solid-state, fiber, gas, and semiconductor lasers provided great support for the research and development of multiple applications, such as optical communication, industrial processing and manufacturing, military and national defense, and state-of-the-art scientific research. Fiber lasers with good heat dissipation characteristics, excellent transverse mode, high amplification efficiency, compact laser construction, and less costs have become the first choice in developing next generation high-power ultrafast lasers. Fiber lasers can achieve long-term operation stability with good beam quality under above-average power because of their waveguide characteristics and large specific gain fiber surface area. High-power ultrafast fiber lasers usually contain four modules, ultrafast fiber oscillators, optical parameters management, ultrafast fiber amplifiers, and nonlinear compression. Ultrafast fiber oscillators provide seed lasers to achieve high-power ultrafast fiber lasers. A qualified mode-locked fiber oscillator has long-term stability and a proportional repetition shared rate corresponding to the requirements of high-power fiber amplifications. Optical parameters management plays a key role in inhibiting uncompensated nonlinear effects and enabling high-energy pulse output with good pulse quality after optical pulse stretching, high power fiber amplification, and optical pulse compression. The ultrafast fiber amplifiers are key modules to scale up the average power of the stretched-signal pulses. Unfortunately, the uncompensated nonlinear phase introduced by the high-peak power of the signal pulse distorts the pulse profile during its propagation in the fiber system. Based on the well-managed optical parameters of fiber lasers, the well-known fiber amplification methods, such as chirped-pulse, divided-pulse, and pre-chirp managed amplifications are making a significant breakthrough in achieving high-power ultrafast fiber lasers. The pulse duration after high-power fiber amplification is hundreds of femtoseconds limited by the gain-narrowing effect. Therefore, a further cascaded nonlinear compression stage is needed for shortening the amplified pulses, which can realize single/few optical cycle pulse duration to fulfill the requirements of the state-of-the-art physical experiments. With their excellent optical characteristics, the fast-developing high-power fiber lasers can play an increasingly important role in multiple applications.Progress Progress in developing ultrafast fiber oscillators, optical parameters management, ultrafast fiber amplifiers, and nonlinear compression are summarized in this paper, and latest published results are discussed by illustrating the advantages and disadvantages of different methods. The highest repetition rate of fiber oscillators reported using the method of nonlinear polarization rotation is 1 GHz provided to be useful in astronomical optical frequency comb, pulse stacking, and the cavity-enhanced high harmonic generation. The highest average output power and pulse energies are 1.98 W and 684 nJ, which are achieved with the nonlinear loop mirror mode-locking scheme, respectively. Applying a semiconductor saturable absorber mirror to the mode-locked fiber laser can generate an output mode-locked laser with the repetition rate range of 10 kHz--1 GHz and sub-μJ pulse energy. As a newly invented mode-locked method, Mamyshev mode-locked fiber laser has attracted attention for its broadband optical spectrum, high-pulse energy output, and high-peak power. As the seeder for a high-power ultrafast fiber laser system, further efforts need to be taken in developing a more stable fiber oscillator with better parameters.Relying on optical parameter management, current ultrafast fiber amplifiers are realized with different amplification methods, such as chirped-pulse, divided-pulse, and pre-chirp managed amplifications. The highest average output power of 830 W at 1 μm was reported by applying the chirped-pulse amplification. Limited by the transverse mode instability and thermal damage threshold, there is one research direction for further improvement that can be realized by searching for new gain materials with better optical performances. Combining the chirped-pulse and multi-channel divided-pulse amplifications, the highest average output power of 10.4 kW was obtained in a 12-channel fiber laser amplifier. 36 fs mode-locked pulses with 100 W average power were achieved with the method of pre-chirp managed amplification, avoiding adding a cascaded nonlinear compression stage. Apart from the aforementioned amplification methods, coherent pulse stacking method is also an efficient way in realizing ultrafast fiber laser with high-pulse energy. Pulse energy of 10 mJ was achieved with the coherent pulse stacking based on the high-power ultrafast fiber laser source.It is difficult to realize sub-100 fs or even shorter pulse durations in a high-power fiber chirped pulse amplification system due to the gain-narrowing effect. Therefore, a further nonlinear compression stage is necessary to satisfying the state-of-the-art applications, requiring short pulse duration. Multipass cells with quartz sheet/noble gas and noble-gas-filled hollow-core fibers are two common constructions in building the nonlinear compression stage, which are illustrated in the nonlinear compression section of this paper. The pulse duration can be compressed to 22 fs, and a pulse energy of 15.6 μJ was realized in the multipass cell construction. Using the noble-gas-filled hollow-core fibers, pulse duration was shortened to approximately 4.3 fs corresponding to a 1.6 optical cycle with a pulse energy of 1 mJ.Conclusions and Prospect In this paper, the high-power ultrafast fiber laser systems are introduced. Research and development status of high-power ultrafast fiber lasers are illustrated along with introducing principles and internal relations of four fundamental modules of ultrafast fiber oscillators, optical parameters management, ultrafast fiber amplifiers, and nonlinear compression. Depending on the fast-developing requirements from multiple state-of-the-art applications, more efforts need to be taken. Further research directions in developing high-power ultrafast fiber lasers have prospected. One promising way is investigating new fiber materials with promising better optical parameters compared to fused silica. Further, making contributions in developing the aforementioned fiber amplification methods is also an efficient way in developing fiber lasers with above-average power, higher-pulse energy, and shorter pulse duration. Newly designed optical fiber amplification methods still need to be invented by carefully considering the optical characteristics of fiber gain material and theoretical nonlinear optical conditions. High-power ultrafast fiber lasers can play a key role in multiple state-of-the-art applications relying on the development of searching for more functional fiber gain materials, optimizing aforementioned amplification techniques, and inventing new methods in amplifying high-power ultrafast fiber lasers.
  • Jun. 11, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201003 (2021)
  • DOI:10.3788/CJL202148.1201003
Physical Limitations of High-Power, High-Energy Lasers
Zhang Xiaomin, Hu Dongxia, Xu Dangpeng, Wang Jing, Chen Xinbin, Liu Jun, Han Wei, Li Min, and Li Mingzhong
Significance High-power lasers enable us to peer deeper into the outer frontiers of the physical world. Since the demonstration of the first pulsed laser in 1960, pushing the limits of accessible laser power has been one of the themes in optical engineering. In this article, we reviewed the progress in developing high-power solid-state lasers and discussed the design issues that determine the performance of these systems.Progress The more one works with a given technology, the more one becomes aware of its limitations—in the case of solid-state lasers, these are primarily the simultaneous availability of high peak and average powers, combined with excellent beam quality in space domain and pulse quality in time domain. In general, the output capability and beam quality of high-power solid-state lasers are essentially limited by five physical limitations categories—gain capability, beam transformation, thermal load, power load, and fluence load. Priority orders of these five limitations largely depend on the application scenario, operational mechanisms, and technical routes of specific laser facilities. For example, for high-power continuous lasers, the main challenge arises from the thermal load limit, while for high-power pulsed lasers, the critical challenge lies in the power load limit. Thus, detailed knowledge of the physics underlying these limitations and their interactions is crucial to the generation of high-quality, high-power lasers.We compiled some recent experimental and theoretical works on the understanding, avoidance, and breakthrough of these physical limitations, as well as relevant enabling developments for high-power solid-state lasers, including novel materials, geometries, and techniques. This paper consists of an introduction, five body sections, and a conclusion. Each section discusses the necessary ingredients for fighting against one of the five physical limitations. These are accompanied by numerous ideas and tips on how to improve the ef?ciency to make maximum use of pump energy.Conclusions and Prospects In conclusion, the core of breaking the gain capability limitation is fighting against the diverse “losses.” The chock point in breaking the limitation of beam and pulse quality is fighting against the diverse “noise” in all the domains of space, time, and spectrum. The key to overcome the limitation of thermal load is combating the thermal effects. Pushing the limit of power load prevents diverse nonlinear optical effects that accompany the propagation of high-power lasers. Furthermore, breaking the deadlock of the fluence load limit helps counteract the inevitable defects in optical elements. During the long struggle of physical limitations with these five categories, a series of novel laser materials, methods, optical techniques, techniques for optics processing, and geometries were correspondingly developed. In addition, theories on the dynamic properties of laser pumping and amplification, propagation, damage, and thermal control were deepened and consummated.We are now on the threshold to reach a new realm of high-power lasers—developing triple-high lasers with high-peak-power, high-energy (i.e., high-average-power), and high-repetition simultaneously. This is a new territory for laser engineering, which requires us to balance conflicting performance parameters. For example, the simultaneous availability of high-peak, average power (high-energy), presents a contradiction because increasing the peak power typically necessitates raising the laser bandwidth, causing an increase in the quantum defect and subsequent ef?ciency loss. This paper intends to be the beginning of a discussion, not the final word, to pave the way for “triple-high lasers.”
  • Jun. 11, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201002 (2021)
  • DOI:10.3788/CJL202148.1201002
Research on the Design Criteria of Laser Weapons
Liu Zejin, Yang Weiqiang, Han Kai, and Xu Xiaojun
  • Jun. 11, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201001 (2021)
  • DOI:10.3788/CJL202148.1201001
A 976-nm Single-Frequency Laser Based on the Yb∶YAG Crystal-Derived Fiber
Xie Yongyao, Cong Zhenhua, Zhao Zhigang, Zhang Xingyu, Zhao Xian, Shao Xianbin, Zhao Wei, and Liu Zhaojun
Objective Single-frequency fiber lasers (SFFLs) are widely used in areas of coherent beam combination, gravitational wave detection, lidar, and nonlinear frequency conversion because of their excellent performance. In particular, SFFLs operating at 976 nm are highly demanded for nonlinear wavelength conversion to generate coherent blue light. SFFLs use either a ring- or linear-cavity configuration. The ring-cavity setup is complicated because many additional components must be inserted to enable a single-frequency output, which unavoidably introduces insertion loss. In addition, the stable single-frequency operation of a ring-cavity fiber laser is susceptible to environmental changes and vibrations, thereby resulting in mode hopping. In comparison, linear-cavity construction, such as the distributed Bragg reflector (DBR) scheme, is more compact, which creates a large longitudinal mode spacing, helping to maintain lasing on a stable single longitudinal and hop-free mode. The cavity length of DBR SFFL is limited to only a few centimeters. Therefore, high-gain fibers are demanded to enable sufficiently high gain. A novel Yb∶YAG crystal-derived fiber (YDSF) that exhibits some unique properties in fiber lasers has been developed. The YDSF was fabricated based on a molten core method (MCM) and shows advantages such as high doping levels and high stimulated Brillouin scattering threshold. In addition, the pure silica cladding of the YDSF makes it highly compatible with commercially available silica fiber devices. All the above mentioned characteristics make the YDSF suitable for high-power single-frequency lasers. Based on these fibers, single-frequency lasers emitting at 1 μm have been demonstrated recently. In 2019, we demonstrated a 110-mW single-frequency YDSF laser at 1064 nm. However, to the best of our knowledge, single-frequency YDSF lasers below 1 μm have never been reported.Methods A commercially available 10% (atomic number fraction)Yb∶YAG crystal was used to prepare a YDSF. In the experiment, the entire preparation process was divided into two steps to maintain the uniformity of the optical fiber. First, a rod fiber having a diameter of ~1.7 mm was fabricated using a 1.6-mm YAG crystal and pure silica tube (Dinner=2 mm, Dexternal=10 mm). The drawing temperature was controlled at ~2000 ℃. Second, the YDSF was fabricated based on the rod fiber. A short piece of rod fiber was inserted into a different silica tube with the same specification to constitute a new preform, which was drawn into the fiber at 1940 ℃. Next, the physical and optical properties of the YDSF were measured using some devices and methods, such as an optical microscope, energy dispersive spectrometer, fiber refractometer, and cut-back method. Afterward, a homemade all-fiber amplifier was used to measure the gain coefficient of the YDSF at 976 nm. Then, the laser performance of the YDSF was investigated by optimizing the gain-fiber length and reflectivity of fiber Bragg grating (FBG). In addition, a DBR SFFL based on an 8-mm-long YDSF was built to further verify the performance of the YDSF.Results and Discussions The mass fraction of SiO2 and Yb2O3 in the core region of the YDSF were measured to be 58.83% and 5.25%, respectively (Fig. 1). As expected, interdiffusion occurred between the Yb∶YAG core and silica cladding during the drawing process. The refractive index profile of the fiber cross section was measured; the numerical aperture (NA) of the core with a diameter of 8.7 μm was 0.5 (Fig. 1), indicating that the YDSF was a multimode fiber. The absorption peaks of the YDSF were located at 915 nm and 976 nm, corresponding to the transitions from the ground state 2F7/2 to higher states of 2F5/2 of Yb 3+. The peak absorption coefficients were 6 dB/cm and 30 dB/cm for 915 nm and 976 nm, respectively (Fig. 1). For a signal power of 0 dBm and pump power of 181 mW, the net gain coefficient of the YDSF reached 12.6 dB/cm (Fig. 2), which indicated that the YDSF could be used as a gain medium for a 976-nm laser. By optimizing the gain-fiber length and reflectivity of FBG, a maximum output power of 37.2 mW was obtained with a slope efficiency of 24.3% (Fig. 3). In addition, using the 8-mm-long YDSF as the gain medium, a 976-nm DBR SFFL was demonstrated. A maximum output power of 17.8 mW with a signal-to-noise ratio (SNR) of >45 dB was obtained at a launched pump power of 203 mW, and no output power saturation was observed. The corresponding slope efficiency was 15.1% (Fig. 5), which was low because of the mode mismatch. More efforts should be made for reducing the NA and improving Yb 3+ doping concentration. The linewidth of the laser was measured to be less than 41 kHz, which was limited by the measurement setup (Fig. 6). The beam quality of the laser output was also measured using a charge-coupled device (Thorlabs, BC106N-VIS); the beam quality factor was measured to be 1.01 and 1.02 in the horizontal and vertical directions, respectively (Fig. 5). Conclusions A YDSF with 5.25% Yb2O3 doping concentration(mass fraction) was fabricated using MCM. The transmission loss of the YDSF with a core diameter of 8.7 μm was measured to be 1.29 dB/m at 1550 nm. The gain coefficient of the YDSF was 12.6 dB/cm at 976 nm with a pump absorption coefficient of 6 dB/m at 915 nm. Using the DBR linear cavity, a 17.8-mW single-frequency laser at 976 nm was achieved with an 8-mm-long YDSF, exhibiting a slope efficiency of 18.5%. To the best of our knowledge, this is the first demonstration of a single-frequency YDSF laser below 1 μm. The SNR was measured to be >45 dB with a linewidth of less than 41 kHz. Results indicate that the YDSF is a promising candidate material for the SFFL operating in the 976-nm wavelength region.
  • Jun. 07, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201010 (2021)
  • DOI:10.3788/CJL202148.1201010
Recent Trend of High Repetition Rate Ultrashort Laser Pulse Generation and Frequency Conversion
Zheng Jiaqi, Cong Zhenhua, Liu Zhaojun, Wang Shang, and Zhao Zhigang
Significance Ultrafast lasers with pulse durations on the orders of picosecond and femtosecond are widely used in various fields, such as supercontinuum generation, photoelectron microscopy, and material micromachining. The traditional high-power ultrafast lasers with repetition rates of kHz-MHz exhibit a large heat-affected zone during material micromachining, leading to unpleasant damage to the materials. The emergence of pulse lasers with ~GHz repetition rate can effectively solve this problem. Combining the very high repetition rate of ~GHz and novel burst mode processing technique, the GHz “burst-mode” femtosecond lasers have been used to ablate the target materials before the residual heat deposited by previous pulses diffuses away from the processing region, which can not only improve the ablation efficiency, but also ensure excellent processing quality.Due to its short wavelength, high resolution, and high photon energy, deep ultraviolet (DUV) lasers are widely used in chip defect detection and photoelectron spectroscopy experiments. In order to obtain DUV lasers with high beam quality, high coherence and high repetition rate, near-infrared all-solid-state lasers are usually used as the fundamental drivers to DUV lasers through the nonlinear optical crystals-based multi-stage frequency conversion technique. Due to the high peak powers and high wavelength conversion efficiencies of the near-infrared pulsed lasers with repetition rates of kHz and MHz, it is easy to obtain high-power DUV lasers for lasers with those repetition rates. At present, the repetition rates of industrial high-power ultraviolet lasers are usually in kHz and MHz range. There are very few research results on DUV lasers with ~GHz repetition rate, which greatly limits the application potential of DUV lasers in the above aspects.In recent years, various methods have been proposed to achieve DUV laser pulses with repetition rates of ~GHz. However, these methods still face a series of challenges. Therefore, it is necessary to summarize recent development tendency of technology of high repetition rate ultrashort laser pulse generation and frequency conversion.Progress There are many methods for producing GHz bursts of laser pulses. Femtosecond pulses at multi-GHz repetition rates can be obtained directly from the oscillators with harmonic mode-locking technique, semiconductor saturable absorber mirror and Kerr lens based passive mode-locking techniques. Typical pulse repetition rates of pulse trains generated by mode-locked fiber oscillators are in the range from tens up to hundreds of MHz. The GHz pulses can be obtained through repetition rate multiplication techniques. In this study, we briefly illustrate their pros and cons and review their recent developments. The emergence of multi-stage amplification systems has increased the average power of ~GHz femtosecond pulses in the near-infrared band to the order of hundreds of watts (Table 1).There are many methods for producing DUV lasers. For the method of nonlinear crystal frequency conversion, the research of 266/258 nm DUV nanosecond lasers (Table 3), picosecond lasers (Table 4) and femtosecond lasers (Table 5), as well as 213/206 nm (Table 6) and 193 nm DUV lasers (Table 7) in the past two decades are summarized. Nowadays, the average powers of high power 355 nm ultraviolet lasers have reached hundreds of watts, and the market is relatively mature. Although commercialization of DUV lasers with wavelength below 300 nm is still not mature, the current 266 nm laser developed in the laboratory can achieve an output average power of 50.1 W, which is about to enter the order of 100 W, and has passed the stability test for more than 5000 hours of continuous operation.For high power GHz repetition rate near-infrared femtosecond pulse lasers, the difficulty lies in the generation of GHz seed. For GHz repetition rate amplifier, it is relatively easy to obtain higher average powers due to low single pulse energy and low peak power. For GHz repetition rate DUV femtosecond pulse laser source, the difficulty is not in the generation of the fundamental frequency laser, but in the low peak power of the fundamental frequency laser and the thin nonlinear medium used, which leads to low nonlinear frequency conversion efficiency, and it is difficult to obtain GHz femtosecond pulse laser in the DUV band (Fig. 6).Conclusion and Prospect In recent years, the French company Amplitude has put forward the idea of “GHz Revolution”, which mainly refers to the development of ultra-short pulse laser sources with pulse repetition rate in GHz. The emergence of multi-stage amplification systems has increased the average power of GHz femtosecond pulses in the near-infrared band to the order of hundreds of watts, which successfully solves the problem of the GHz pulse in industrial processing. Therefore, the development of high-power near-infrared band GHz repetition rate pulse lasers is relatively mature at present. Coupled with the continuous improvement of nonlinear frequency conversion technology, DUV laser repetition rate has entered the GHz. Although the industrialization and commercialization of DUV laser techniques still face some problems, such as easily damaged crystal coating, low wavelength conversion efficiency of DUV lasers, and long-term unstable operation of high-power DUV lasers, these problems have been gradually solved in practice. With the further maturity of frequency conversion and power amplification techniques, perhaps kilowatt-level DUV lasers will appear in the next 5-10 years, all of which will certainly make a breakthrough in the secondary laser source based on ultraviolet laser and DUV laser.
  • Jun. 07, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201008 (2021)
  • DOI:10.3788/CJL202148.1201008
Microstructure Lasers Based on Parity-Time Symmetry and Supersymmetry
Fu Ting, Wang Yufei, Wang Xueyou, Chen Jingxuan, Zhou Xuyan, and Zheng Wanhua
Significance Conventional semiconductor lasers typically use gratings, such as distributed feedback (DFB), distributed Bragg reflector (DBR), and slotted surface, to select longitudinal modes and microstructures to select lateral modes, such as narrow ridge, chirped waveguide array, and angled cavity. Even though these technologies are mature, their practicality is limited by output power or complex fabrication processes. For example, a narrow ridge can suppress the high-order lateral modes of the edge-emitting semiconductor laser, thereby limiting the laser's output power due to the small area of current injection. Therefore, new physical effects should be explored to offer new insights into the designs of lasers. Recently, because of the similarity between quantum and optical systems, some physical terminologies of the former are introduced to the latter such as parity-time (PT) symmetry and supersymmetry (SUSY).The PT symmetry can be used to control the laser's spectral and spatial characteristics. The optical system obeying PT symmetry requires that its complex refractive index satisfies the relation, n(x)=n*(-x), which means that the distributions of the real and imaginary parts of the complex refractive index are even and odd functions, respectively. One specified pair of modes of the system can evolve from the PT-symmetric phase to the broken PT-symmetric phase by varying the gain/loss contrast of the PT-symmetric system [Figs. 1(b), (c)]. Especially when the modes stay in the broken phase, the mode field distribution of the amplified mode will be in the gain area and the lossy mode will be in the loss area [Fig. 1(d)], allowing the realization of a single-mode laser. SUSY can also control the optical modes of lasers, making it an excellent candidate for single lateral mode laser arrays. For a waveguide array, the superpartner of the array can be constructed by supersymmetric transformation to couple the high-order modes of the original array but do not influence the fundamental mode (Fig. 3), which can increase the lasing threshold difference between high-order and fundamental modes. Then, the SUSY laser array can exclusively achieve fundamental mode lasing, which improves the laser array's lateral beam quality. Therefore, it is critical to review recent works on the two new methods of optical mode control in lasers.Progress PT symmetry can be realized in the lateral direction of the lasers (Figs. 4--6). Here, the lasing of a single lateral mode can be achieved due to the selective PT symmetric breaking of the fundamental mode, which results from the smaller coupling constant of the fundamental mode than that of high-order mode. When the optical system is PT symmetric, the increased gain threshold between the centered longitudinal modes in the gain spectrum and neighboring longitudinal modes aid the realization of a single longitudinal mode lasing. Furthermore, PT symmetry can be applied to the longitudinal direction (direction along the cavity length). The longitudinally PT-symmetric laser can also realize single-mode lasing because of the PT symmetric breaking of the specified modes (Fig. 7). In addition, coherent perfect absorber (CPA)-laser (Fig. 8) and orbital angular momentum (OAM) laser (Fig. 9) are realized on the basis of longitudinally PT-symmetric microstructures. The double mode spacing of the CPA-laser is observed compared with that of the common Fabry-Perot laser, showing that the neighboring lasing modes move in the opposite direction of the complex plane of frequency when the laser stays in the broken PT-symmetric phase [Fig. 2(f)]. The OAM laser can use the orbital angular momentum of light to transfer information, increase the density of data transmission, and pave the way to develop a multidimensional OAM-spin angular momentum (SAM)-wavelength division multiplexing.Similarly, SUSY can control the optical modes of non-Hermitian systems. The SUSY transformation is used to determine the profile of the refractive index distribution of the SUSY laser array so that the modes are selectively confined in the original array. Simultaneously, the chirped energy pumping increases lasing threshold difference between the selectively confined modes and other modes. If the fundamental mode is confined in the original array and other modes extend to the lossy superpartners, single lateral mode lasing can be realized with higher output power than the single-ridge laser under the same energy pumping density [Figs. 10(a)--(h)]. Furthermore, the second-order SUSY micro-ring laser array is also reported [Figs. 10(i)--(k)], which greatly simplifies large-scale laser array engineering because the superpartner and original array possess identical elements. Also, the second-order SUSY micro-ring laser array emits light in a single lateral mode.Conclusions and Prospect In summary, PT-symmetric lasers that can not only be pumped optically and electrically are realized. However, the methods to suppress the influences of nonlinear effects on the stability of PT-symmetric laser operation should be explored eagerly. Compared with the PT-symmetric lasers, SUSY lasers are still pumped optically. Electrically injected SUSY lasers with multiple coupling terminals are promising candidates for high output power single lateral mode lasers.
  • Jun. 07, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201005 (2021)
  • DOI:10.3788/CJL202148.1201005
High Repetition-Rate Optical Parametric Chirped-Pulse Amplifiers
Shang Jingcheng, Liu Yizhou, Zhao Shengzhi, Feng Tianli, Yang Kejian, Qiao Wenchao, Zhao Yuantao, and Li Tao
Significance Optical parametric chirped-pulse amplifier (OPCPA), which relies on the optical parametric process existing in nonlinear crystals to realize pulse amplification, can further enhance a yielded peak power by avoiding the gain narrowing and thermal effects that usually exist in a chirped-pulse amplifier. The generated ultrashort pulse with high peak power from OPCPA can considerably extend the ultrafast pulse to X-ray and infrared regimes, thereby bringing a new revolution to ultrafast science. OPCPA system pumped ultrafast X-ray desktop light source has potential applications in medicine, biology, and materials science. In this article, we review the development of OPCPA systems, including its main characteristics and research progresses in different wavelength regimes (~0.8, ~1.5, ~2, ~3, and >4 μm).Progress In ~0.8-μm OPCPA systems, the pumping laser pulse (0.515 μm) is usually generated via the second harmonic generation process from a Yb-doped laser amplifier. The seed pulses can be generated using a mode-locking few-cycle Ti: Sapphire oscillator directly or via the supercontinuum generation process. The energy/power of seed pulses is usually scaled up inside the BBO nonlinear crystals. The dispersion compensation for the amplified seed pulses is realized by the chirped mirror or glass block. Currently, the highest pulse repetition rate of 11.5 MHz, shortest pulse duration of 5 fs, maximum average output power of 112 W, largest pulse energy of 54 mJ, and highest peak power of 5.5 TW have been realized in the ~0.8-μm OPCPA systems ( Fig. 6). In ~1.5-μm OPCPA systems, the pumping source is a Ti: Sapphire laser amplifier or Yb-doped laser amplifier. The seed pulses are generated via the supercontinuum generation process in noble gas or transparent medium, which is stimulated by the Ti: Sapphire laser oscillator and Er- or Yb-doped fiber/solid-state laser oscillators. Different crystals, such as BIBO, DSTMS (organic crystal), KTA, and LBO, have been employed to amplify the seed pulses. The dispersion of the amplified seed pulses is compensated by the fused quartz, Si, or chirped mirrors. A 350-kHz pulse repetition rate, 6-fs pulse duration, 106-W average output power, 3-mJ pulse energy, and 263-GW peak power have been achieved in the ~1.5-μm OPCPA systems (Fig. 7).For ~2-μm OPCPA systems, the Yb-doped disk laser or Ti: Sapphire laser amplifiers are mainly used as the pumping sources. The seed pulses are generated via the difference frequency generation after the supercontinuum generation process. The seed pulse energy/power is enhanced in a crystal, such as PPLN, LiNbO3, BIBO, or YCOB. The amplified seed pulses are compressed by a high-transmittance crystal, such as Si, ZnSe, or quartz. The optimal output parameters achieved from the ~2-μm OPCPA systems are 100-kHz pulse repetition rate, 10.5-fs pulse duration, 33-W average output power, 3.3-mJ pulse energy, and 132-GW peak power (Fig. 8).A ~3-μm OPCPA system is usually pumped with the Ti: Sapphire laser or Yb-doped lasers. A 3-μm ultrafast pulse can be directly amplified using the ~3-μm OPCPA system or can be the idler pulse from a ~1.5-μm OPCPA system. The employed nonlinear crystals for amplification are PPLN, KNbO3, and MgO∶LiNbNO3. The pulse is compressed with grating pairs or solid medium (Si). The highest pulse repetition rate of 160 kHz, shortest pulse duration of 20 fs, highest average output power of 21 W, largest pulse energy of 5.8 mJ, and highest peak power of 290 GW have been reported from ~3-μm OPCPA systems (Fig. 9).In OPCPA systems operating beyond 4-μm wavelength region, the pump lasers are 1-μm Yb-doped or 2-μm Ho-doped laser amplifiers. In particular, the Ho-doped laser amplifiers are beneficial to realize a high-efficient long wavelength from the OPCPA system. The seed pulses realized from the difference frequency process are amplified by the ZGP, KTA, or LGS(@1 μm). The dispersion management is performed by CaF2, Ge, or grating pair. At present, OPCPA systems can deliver a laser pulse with the longest wavelength of up to 9 μm (Table 2).Conclusion and Prospect Although the performances of OPCPA systems in different wavelength regimes have remarkably improved, there is still scope for further improvement. With the progress of high-power pump laser sources and high-quality nonlinear crystals, the OPCPA system is heading toward achieving shorter pulse duration, larger pulse energy, higher peak power, and longer output wavelength than the existing ones.
  • Jun. 07, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201004 (2021)
  • DOI:10.3788/CJL202148.1201004
100 Hz High-Energy KTA Dual-Wavelength Optical Parametric Oscillator
Meng Jun, Cong Zhenhua, Zhao Zhigang, Wang Shang, Qi Yunxuan, Zhang Xingyu, and Liu Zhaojun
Objective Laser sources at 1.5 μm, providing high pulse energies and short pulse durations are used in various applications, such as electro-optical countermeasures and high precision ranging. Lasers at 3--5 μm are used as lighting sources for active remote sensing and gas detection, which show important potential applications. Therefore, lasers with high energy at both wavelength bands have become research hotspots. The optical parametrical oscillators (OPO) are effective ways to generate lasers with wavelengths at 1.5 μm and interval 3--5 μm due to their compactness, wavelength-tunable property, and potential for generating high energy and short pulse width. Presently, the major nonlinear crystals with high-quality include biaxial crystals such as KTiOAsO4 (KTA), KTiOPO4 (KTP), ZnGeP2 (ZGP), and periodically poled crystals such as PPKTP, PPLN, PPLT, etc. KTP crystals are used to obtain lasers at 1.5 μm, which is affected by severe absorption in the mid-infrared region. To obtain lasers at 3--5 μm, ZGP crystals have been under investigation for a long time. However, 2 μm pump sources are more in need, which is technically more difficult than their 1-μm counterpart. PPLN crystals are used to obtain mid-infrared lasers. Compared with crystals such as KTP, the damage threshold of PPLN crystals is lower. KTA and KTP crystals belong to the same crystal group and have a high damage threshold (>600 MW/cm 2), large nonlinear coefficient (d24=3.2 pm/V), large acceptance angle, a wide temperature range, and stable physical and chemical properties. The transmission performance of KTA crystals in the mid-infrared band (3--5 μm) must be better than that of KTP crystals. These characteristics make KTA crystals suitable for high energy mid-infrared laser applications. In this study, we report a 100 Hz high energy KTA crystal-based OPO system. Methods The 100 Hz high energy KTA-OPO system is composed of 1064 nm Nd: YAG main oscillator power amplifier (MOPA) and KTA crystal-based OPO. The Nd∶YAG MOPA laser at 1064 nm served as the pump source. To obtain high beam quality, the Nd∶YAG MOPA system adopted the “unstable cavity oscillator + two-stage amplifiers” scheme. Both the oscillator and the two-stage amplifier used a double rod structure connected in series, and a 90° quartz rotator between the two Nd∶YAG crystal rods to compensate for the thermal depolarization effect. To prevent self-excited oscillation and spontaneous radiation between the stages while protecting the optical components of each stage, isolators are placed between each stage. The X-cut KTA crystal is used in the experiment, and the dimension of the KTA crystal is 10 mm×10 mm×33 mm. The cavity is designed as a signal resonant oscillator with a cavity length of 65 mm. The input mirror is coated to be highly reflective for the signal and high transmittance for the pump light. The output mirror is coated with a partial reflectivity of 50% for the signal and high transmittance for the idler. The pump light passed the OPO twice. An isolator protects the pump laser from the remaining pump light that comes back from the OPO cavity.Results and Discussions A homemade 1064 nm Nd∶YAG MOPA with a pulse energy of 580 mJ at 100 Hz repetition rate is employed as the pumping source. After two-stage amplification, 580 mJ of 1064 nm laser is obtained with the extraction efficiency of the primary amplifier and secondary amplifier at 6.7% and 10.8%, respectively (Fig. 3). The beam quality factor of the 1064 nm laser is Mx2=4.6 and My2=3.7 (Fig.4). The pulse width of the laser from the oscillator and primary amplifier and secondary amplifier are 15.3, 16.9, and 18.0 ns, respectively (Fig.5). In the OPO experiment, the optical-to-optical conversion efficiency is optimized by increasing the cavity length and KTA crystals length. The output energy and conversion efficiency of the KTA crystal with a length of 33 mm are higher than that of the KTA crystal with a length of 38 mm (Fig.6). Then, experiments with different OPO cavity lengths are performed on the 33-mm KTA crystal. The results indicated that the output energy and conversion efficiency are higher for short cavity length (Fig.6). The threshold of the OPO is about 20 mJ. When the pump energy is 580 mJ, 64 mJ idler is obtained at 3.47 μm and associated signal at 1.54 μm is 178 mJ (Fig.7). The OPO efficiency is 46.3% high, and OPO output stability is 1.2% rms (Fig.7). The pulse width of the output laser at 1.54 and 3.47 μm is 13.7 and 11.8 ns, respectively (Fig.8). The beam quality factor of the 1.54 μm laser is Mx2=30.5 and My2=28.2 (Fig.9). In addition, the center wavelength of the signal laser is 1.535 μm (Fig.10).Conclusions A 100-Hz, high-energy KTA crystal-based OPO system is reported. A homemade 1064 nm Nd∶YAG MOPA with a pulse energy of 580 mJ at a 100 Hz repetition rate is used as the pumping source. We adopted plane-plane cavity configuration for the OPO, and an X-cut KTA crystal as the nonlinear crystal. The obtained pulse energies at a signal wavelength of 1.53 μm and idler wavelength of 3.47 μm are 178 and 64 mJ at a pulse repetition rate of 100 Hz, respectively. Furthermore, the pulse durations are 13.7 and 11.8 ns, respectively, and the optical-to-optical conversion efficiency is 43.6%.
  • Jun. 07, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 12 1201009 (2021)
  • DOI:10.3788/CJL202148.1201009
Optical Compensation Method for Pigtail Module in High-Brightness Laser Diodes
Jiang Quanwei, Zhang Lanping, Guo Linhui, Tan Hao, Wu Hualing, Fu Bo, Zhang Nisha, Zhang Fengqun, Du Weichuan, Wu Deyong, and Gao Songxin
Objective The laser diode(LD) coupled with the fiber is an effective and high-quality pump source of fiber and solid lasers. Although this technology is relatively mature, its efficiency, structure, and optical design need to be optimized. The optical alignment and adjustment directly influence the efficiency of the LD module, which has become a restriction in the development of high-efficiency and high-brightness LDs. To improve the efficiency of the LD module, without improving the LD chip itself and electric to optical (e-o) efficiency, reducing the optical loss by the maximum possible extent is a research key. In the optical assembly and adjustment engineering practice, beam collimation has an important impact on LD modules to achieve high brightness, high power, and pigtail output. Furthermore, to ensure smaller divergence, the directivity of the optical axis should also be improved, so as to realize precise beam coupling. In direct LD applications, controlling the direction of the laser beam based only on mechanical alignment is limited by high volume, high cost, and poor stability. Thus, the optical axis should be optically controlled.Methods The pointing error of the fast axis is mainly influenced by the ultraviolet glum solidification technology; however, in practice, we cannot ensure the pointing error accuracy because there exists inherent shrinkage stress when the UV (ultraviolet) glue solidifies. Through an optical method based on the principle of laser refraction propagation in a medium, calibration of fast-axis directivity via a bizarre slow-axis collimation mirror was studied. When light is normally incident upon a medium, the directions of the transmitted light and source light are the same, but at oblique incidence, the directions of the two light beams differ. We employed this basic principle to design a slow-axis collimation mirror with a certain inclination angle according to the relationship, based on Eqs. (1)--(3), between the inclination angle and pointing error of the fast axis to correct the pointing error.Results and Discussions We designed four kinds of slow-axis collimation mirrors with a tilt angle of 0.23°, a lens height of approximately 2 mm, and an edge-thickness difference of approximately 0.009 mm (Table 2). We also measured the central offset along the fast axis with and without the bizarre slow-axis collimation mirror (Fig. 8). The central point coordinate decreased from 1.07 mm to 0.15 mm when the focal length of the Fourier transformation lens was 510 mm. After measurement, the original average deviation of fast-axis directivity was approximately 2.1 mrad; after calibration by our designed slow axis collimation mirror matched with the pointing error of the fast axis, the directivity of the fast axis was maintained at 290 μrad. In this way, the power entering the fiber was maximized, and the efficiency of the coupled fiber was increased. We found that the average pointing error was reduced by an order of magnitude based on the seven-step LD module with pigtail output (Fig. 9). The laser spot was not completely filled with the coupled fiber when the pointing error was high,and some of the laser spot overflowed to the fiber outside (Fig. 10). The fiber would be filled with the laser spot only when the pointing error was lower. Thus, the efficiency of the 60-W fiber-output LD module increased from about 53% to 55%. The power and e-o efficiency of the LD module after calibration were measured (Fig. 11).Conclusions In this research, we proposed a novel method to correct the pointing error of the fast axis by employing a bizarre slow-axis collimation mirror (SAC). This method has applications in high-brightness and high-efficiency LDs with pigtail outputs, and its advantages include not requiring extra devices to compensate for the pointing error of fast axis in existing conditions, so precise compensation can be achieved. The material and machining costs of the SAC are reduced because complex processing is not needed. Globally, domestic advantageous units have reported e-o efficiencies of approximately 51%--53%, whereas the efficiency is approximately 51% overseas. Our technical specifications contribute to the domestic leading efficiency. Our research provides a novel method for high-efficiency LDs with fiber output. In terms of the LD’s direct application, this study has numerous potential applications in laser fuse, radar, ranging, and illumination. We have achieved the goal of applying tens of thousands produced LD modules to a pump source.
  • Jun. 04, 2021
  • Chinese Journal of Lasers
  • Vol.48 Issue, 11 1101003 (2021)
  • DOI:10.3788/CJL202148.1101003