The simultaneous measurement of temperature and velocity plays an important role in research on turbulent combustion which is a complex problem coupled of fluid dynamics and chemical reaction dynamics.With the rapid development of laser combustion diagnosis technology, researchers have studied many schemes of simultaneous measurement of multi-parameters, such as temperature and velocity, by means of multi-technology combination and extending the multi-parameter measurement function of measurement technology. The principle of hydroxyl tagging velocimetry (HTV) is to dissociate with water molecules in the flow by lasers which produce ion OH^- as tagging particles (OH_P). Combined with the technique, plane laser-induced fluorescence (PLIF), OH_P can be displayed in certain positions at different time. It is unnecessary to inject extra tracer particles into the flow field without a particle-following problem. OH_P, the tagging molecule, is of long life in a high-temperature environment, which has the advantage of measuring velocity in flow with high temperature and velocity. The technique has been widely applied to obtain the velocity of various flows such as super-combustion ramjet engines. The technique of temperature acquiring through double-color PLIF and fluorescence intensity of OH_P is configured based on OH_P on the basis of HTV, making it possible to accomplish simultaneous measurement of temperature and velocity of flows. In this paper, the principle and the configuration of the set-up are fully illustrated, while an experiment to verify the technique used in the flow of temperature is conducted.

The key to simultaneously measuring velocity and temperature based on HTV is how to realize temperature measurement based on OH_P. In this paper, two methods of temperature measurement are studied, one of which is the two-line PLIF temperature measurement method. On the basis of the PLIF device used to display the image of OH_P in HTV technology, another set of PLIF devices is added to build the double-color PLIF, realizing the temperature measurement. The variation of OH_P fluorescence intensity with laser dissociation and fluorescence excitation time delay under different excitation lines (Fig. 1) is obtained by experiments. It can be concluded that OH_P has reached the state of thermal equilibrium with the surrounding environment when the delay time is more than 500 ns, and OH_P can be used as a probe to monitor temperatures in the flow. It is proven to be probable that temperature and velocity can be simultaneously acquired by OH_P. The other method is using OH_P fluorescence intensity to measure temperature. Analyzing the relationship among OH_P fluorescence intensity, temperature, and intensity of fluorescence is monotonically changeable as the temperature of the objects changes when appropriate fluorescence excitation is selected (Fig. 3). Therefore, temperature can be observed by fluorescence intensity.

The simultaneous measurement of velocity and temperature is tested and verified in the electric heating flow field and the flow field of the super-combustion ramjet engine. Compared with results obtained by thermocouple, in the range from room temperature to 900 K, the average standard deviation of temperature measurement based on PLIF is 12.1 K (Fig. 6). The maximum deviation of temperature measurement based on photolysis OH_P fluorescence intensity is 16.8 K, with the uncertainty of velocity measurement less than 1% (Fig. 10). In the flow of super-combustion ramjet engine, data of temperature measured by CARS serve as baseline to acquire the results of temperature and velocity on the tagging line simultaneously. The maximum deviation between them is 44 K (Fig. 11).

In this paper, simultaneous measurement of velocity and temperature based on HTV technology is presented, which extends the function of HTV technology and provides a new scheme for simultaneous measurement of temperature and velocity in flow fields with high speed and high temperature.

Unmanned aerial vehicles (UAVs) have emerged as a versatile platform for a wide range of civil applications, and offer flexibility and convenience in performing various tasks, like mapping, investigation, and patrolling. Accurate navigation information such as velocity and position is required to ensure the flight safety of UAVs. However, the global navigation satellite system (GNSS) is easily denied in urban environments due to the occlusion of tall buildings, which results in navigation information loss. Inertial navigation systems (INSs) are also hard to be relied on for long-scale autonomous navigation as its error accumulates over time. High-precision and independent velocity measurement methods will be beneficial for the navigation of UAVs. The laser Doppler velocimeter (LDV) has been applied to the integrated navigation of land vehicles improving localization accuracy. There is a bottleneck for LDV deployment on UAVs due to the limited working distance of LDV which is typically restricted to only a few meters. However, UAVs often operate at flight heights of dozens of meters, posing a challenge for LDV integration and utilization.

The carrier-to-noise ratio (CNR) of LDV is analyzed concerning coherent Doppler wind lidar since both of them are coherent detection systems. For the same LDV, the CNR can be improved by reducing the size of the probe beam spot on the target. On this basis, we propose a solution to the bottleneck of implementing airborne LDV by an optical transformation system to extend the working distance of LDV. The optical transformation system comprises a concave lens and a convex lens. By passing the Gaussian probe beam through this system, the size and location of the beam waist can be adjusted by varying the distance between the two lenses. Simulation and experiments show that the size of the transformed waist can be reduced without changing the transformed location of the waist by a probe beam with a larger waist size in the optical transformation system. Before the airborne LDV prototype is assembled, the parameters of the transformation system are optimized through simulation, while considering the size and weight of the LDV. The focuses of the concave and convex lenses are chosen to be -100 mm and 600 mm respectively. Before being input into the optical transformation system, the probe beam is expanded to 10 mm by an 8× expander. The entire LDV system is constructed with a sturdy cage structure, with four metal rods serving as the core skeleton to ensure the coaxial alignment of the optical transformation system. As the attitude of UAVs changes over time, a micro-electromechanical (MEMS) INS has been employed to measure and track these variations in UAV attitude. Additionally, the quality factor of the Doppler signal is defined as the ratio between the amplitude of the Doppler frequency and the mean amplitude of the base in the frequency domain, and it is adopted to represent the CNR in experiments.

After designing, a single-beam airborne LDV prototype is fabricated with a working distance of 50 m and a 10 m depth of field. The spot diameter of the probe beam at 50 m is 0.34 mm. The quality factor has been measured to be 3685 at the working distance of 50 m and remains above 800 throughout the entire depth of field. The depth of field is enough to prevent signal loss and a 110-second flight experiment is conducted with a UAV as the carrier. The velocity measured by the prototype is corrected for the pitch angle recorded by MEMS INS, and the corrected velocity is basically consistent with the velocity recorded by global position system(GPS) as a reference. The entire measurement maintains a high Doppler signal quality factor. However, the accuracy of MEMS INS is insufficient. Although the airborne LDV provides highly precise velocity components along the direction of the probe beam, the accuracy of velocity worsens after correction for pitch angle. The utilization of a two-beam LDV can alleviate this problem since the two beams have different angles for the ground. The velocity can be accurately determined by measuring the Doppler frequency and the angle between the two beams.

The airborne LDV is designed and tested through simulation and experiments. We verify the feasibility of airborne LDV which has promising applications in UAV-integrated navigation. Our study lays a foundation for the development of multi-beam onboard LDV in the future. The acquisition of all accurate velocity components can significantly improve the navigation and localization of UAVs.

Developing an early rapid and highly sensitive diagnostic technology is of great significance to the prevention and treatment of diseases. However, current detection methods require tedious steps, long analytical time, high cost, advanced instruments, and skillful personnel, even with high sensitivity and specificity. Immunochromatographic assay (ICA) is currently recognized as the most promising method for point of care testing (POCT) due to its portability, simple operation, low cost, and short detection time. However, traditional ICA is judged by the visual detection results produced by colloidal gold nanoparticles, and can merely achieve qualitative and semi-quantitative detection. To overcome the disadvantages of low sensitivity and non-quantitative detection, researchers apply PCR, fluorescent probes, and surface-enhanced Raman scattering (SERS) to immunochromatographic systems, greatly improving the sensitivity and quantitative detection properties of ICA. SERS is an excellent analytical method with high sensitivity, against photobleaching, narrow bandwidth, and multi-channel detection.

SERS-ICA is the research on cutting-edge technologies and has become a research hotspot in related fields recently, which combines the advantages of SERS including high throughput and sensitivity, and ICA featuring simpleness and rapid speed. SERS immuno-tags prepared by nanomaterials are employed to replace the traditional colloidal gold nanoparticles and can provide SERS signals for quantitative detection. SERS immuno-tags are mainly composed of three parts containing noble metal nanomaterials (e.g., gold, silver), Raman report molecules, and specific recognition elements (e.g., antibodies, aptamer, nucleic acids). Notably, Raman report molecules are adopted to provide the characteristic Raman signals, and their SERS signal intensities will be greatly enhanced while approaching the rough surface of the noble metal nanomaterials. Additionally, the specific recognition elements are applied to specifically recognize and capture the targets from the sample solutions. Quantitative detection of SERS-ICA is achieved by collecting and analyzing the SERS signals on the test line of the strips produced by the intercepted SERS immuno-tags.

High-performance SERS tags play a key role in SERS-ICA detection. Many studies concentrate on developing the nanomaterials with high density "hot spots", and improving the SERS signals by constructing multi-dimensional and high-density "hot spots" on the SERS substrates. To improve the detection sensitivity, in recent years, researchers have synthesized the SERS substrates with strong SERS enhanced performances by designing and optimizing the particle sizes, morphology and structures of the nanomaterials (Figs. 2-5). Raman report molecules and antibodies are successively conjugated on the SERS substrates to prepare functionalized SERS immuno-tags. The as-prepared SERS immuno-tags can specifically capture the target antigen to form the immunocomplex of tags-antigen and migrate on the ICA strips by capillary action towards the absorbent pad. In addition, the detection antibodies precoated on the test line of the ICA strips can specifically identify the target antigen and capture the immunocomplex of tags-antigen. Therefore, visual band and SERS signals can be found on the test line due to the formation of antibody-antigen-antibody sandwich composite structure. Then, the SERS signals on the test line are collected by the Raman detector, and the calibration curves of the SERS signal intensities and the corresponding concentrations of target antigen are plotted for quantitative analysis of the unknown target concentration in the sample. For improving the detection efficiency of SERS-ICA, researchers have set up multiple testing lines or testing dots on one ICA strip (Figs. 6-10). Meanwhile, based on the characteristic Raman fingerprint spectra, different Raman report molecules with uncrossed characteristic Raman shifts are modified on the SERS substrates to distinguish different targets on the same sites and this feature is applied to SERS-ICA (Figs. 8-9). Moreover, the as-reported integrated multi-channel immunochromatography reaction column greatly improves the multiplexing and automation detection properties of SERS-ICA (Fig. 10).

We briefly introduce the basic principles of SERS and ICA and summarize several different SERS substrates for SERS-ICA and the application of SERS-ICA in different detection fields. It is of significance for increasing the detection sensitivity of SERS-ICA to improve the antigen-capture ability of the SERS tags and SERS enhanced properties of the SERS substrates. Finally, the future development trend of SERS-ICA detection technology is prospected.

Polarization is one of the fundamental properties of light waves and an important carrier of information. The technique of measuring polarization is known as polarimetry. Compared with traditional intensity detection methods, polarimetry can significantly enhance the ability to acquire and analyze target information by fully utilizing the polarization characteristics of light waves. Due to its unique advantages, polarimetry has been widely used in various fields such as remote sensing, industrial inspection, biomedical, and environmental monitoring.

The key to polarimetry is to obtain the Stokes vector of the measured light waves or the Mueller matrix of the measured sample, which correspond to the Stokes polarimetry and Mueller matrix polarimetry, respectively. Several methods have been proposed for achieving Stokes polarimetry and Mueller matrix polarimetry, including multi-channel polarimetry, temporally modulated polarimetry, spectral polarization modulated polarimetry, and spatially modulated polarimetry. In the multi-channel method, the incident beam is split into several channels with different polarization optics for analyzing the polarization state. In the last three methods, the incident light is modulated in the time domain, spectral domain, or spatial domain for measuring polarization information. The multi-channel scheme is competent for real-time monitoring, but its configuration is usually complicated to adjust. The configuration of temporally modulated polarimetry is compact, but it is restricted by poor stability and slow measurement speed. The spectral polarization modulated polarimetry obtains polarization information at a single integration interval without rotating or active components, but its measurement accuracy is limited, and its wave band is narrow. In contrast, spatially modulated polarimetry modulates the polarization information at different spatial locations by using the spatial modulation components, and it has the advantage of stability, rapidness, and compactness, so it is a promising technique for polarization measurement.

Numerous review articles have provided comprehensive summaries of multi-channel polarimetry, temporally modulated polarimetry, and spectral polarization modulated polarimetry. However, little attention has been given to spatially modulated polarimetry. With the increasing maturity of micro-nano fabrication and optical field control technologies, various spatial modulation devices such as vortex retarders, azimuthal polarizers, and S-waveplates have been fabricated with high quality and have gained important applications in the field of polarization detection. Therefore, it is crucial and imperative to consolidate the current research about spatially modulated polarimetry to guide the future development of this field more rationally.

Firstly, the working principles and technical characteristics of various spatially modulated Stokes polarimeters and Mueller matrix polarimeters are analyzed and summarized. In these spatially modulated polarimeters, vector optical beams with spatially inhomogeneous polarization distributions or spatial polarization modulation components such as micro-polarizer arrays, polarization grating, azimuthal or radial polarizers, Savart polariscopes, and handmade axisymmetric quarter-wave plates have been utilized to modulate the spatial distribution of light intensity. This enables the measurement of polarization information in a stable, rapid, and compact way. However, it should be noted that the existing spatially modulated polarimetry is limited by the hard fabrication, poor modulation quality of the spatial polarization modulation components, complex processing procedures, and low accuracy. In particular, there is no configuration yet that can realize accurate Mueller matrix measurement experimentally with the spatial polarization modulation technique.

Secondly, in order to overcome the drawbacks of the traditional spatially modulated polarimeters, vortex retarders with the advantages of mature fabricating processes, good wavelength and temperature stability, high modulation quality, and low cost are utilized to construct the high-performance spatially modulated polarimeter. The vortex half-wave retarder-based Stokes polarimeter (Fig. 9) can achieve polarization measurement in a single shot, and it is fast, stable, and easy to implement. However, the measurement accuracy of the vortex retarder-based Stokes polarimeter is limited by various error sources. In order to reduce the measurement error, an efficient calibration method is proposed by analyzing the general effects of the different error sources on the intensity modulation curve of the incident waves with the Stokes-Mueller formalism. The error calibration method can effectively reduce the measurement error from about 0.05 to less than 0.01 (Fig. 12). The proposed vortex half-wave-based Stokes polarimeter lacks the capability of detecting circular polarization components because the fourth Stokes parameter is calculated indirectly, and the handedness of the input light cannot be recognized. In order to directly measure all the Stokes parameters, a vortex quarter-wave retarder is employed to substitute the vortex half-wave retarder, and experiments show that the measurement accuracy of the vortex quarter-wave retarder-based polarimeter is less than 0.035 (Fig. 13). Based on the vortex retarder-based Stokes polarimeter, a complete dual vortex retarder Mueller matrix polarimetry (Fig. 14) is proposed and experimentally verified by using two vortex quarter wave retarders with different orders in polarization state generation (PSG) arm and polarization state analyzer (PSA) arm, and the maximum absolute error is less than 0.04 (Fig. 16).

The spatially modulated polarimetry has the advantages of simple optical structure, good stability, fast measurement speed, and high accuracy, and it is promising in target detection and recognition, industrial and biochemical detection, and many other fields. As for the perspective of spatially modulated polarimetry, one direction may be smart polarimetry, which is expected to improve the speed and accuracy of image processing and reduce the prior information of the sample and the technical requirements for operators in polarimetry. In addition, the metasurfaces can be utilized to achieve polarization measurement in a compact size. Furthermore, spatially modulated polarimeters should overcome the limitations in spatial resolution and spectral measurement to expand their application fields.

Fiber lasers by laser diode (LD) pumping directly have power scaling due to stimulated Raman scattering (SRS) and transverse mode instability (TMI). Beam combining is an effective way to increase the output power of fiber lasers. Spectral combining is one of the structurally simple and proven effective means. Dichromatic mirror spectral combining technologies employing broad-spectrum laser sources can achieve high efficiency, high beam quality, and high reliability for medium output power. This technology can maintain the beam quality based on flexible structures and low costs. Spectral combining employing dichromatic mirrors requires expanding the central wavelength range. Currently, the output power in short-wavelength fiber lasers cannot meet the requirements of spectral combining. Ytterbium-doped fibers (YDFs) can provide continuous-wave (CW) high-power output mainly focusing on 1030-1100 nm. Short-wavelength fiber amplifiers are susceptible to the amplified spontaneous emission (ASE) effect. ASE and SRS impose serious constraints on the power scaling of short-wavelength fiber lasers. In this paper, a fiber amplifier of 3.5 kW with near-single-mode emitting at 1050 nm is demonstrated. The SRS and ASE suppression methods in short-wavelength fiber lasers are proposed.

The fiber laser amplifier of 1050 nm is constructed by using large-mode-area (LMA) double-cladding YDF in the counter-pump scheme. The core and cladding diameters of the YDF are 20 μm and 400 μm, respectively. The length of gain fiber is optimized to mitigate ASE in short-wavelength fiber amplifiers. The threshold of SRS is raised by optimizing the temporal stability of the seed. The wavelength-stabilized LDs of 976 nm are utilized as the pump source. In order to ensure beam quality, YDF is coiled on a water-cooled plate with a bending diameter of about 8 cm. Employing LDs of 976 nm in the counter-pump and choosing appropriate diameters of YDF can aid TMI mitigation.

Compared with output characteristics of different YDF lengths (Fig. 2), short YDF is beneficial to mitigate ASE. The 4.8 m-long-YDF of seed can suppress ASE and guarantee the absorption of pump power. The SRS threshold of the fiber amplifier adopting the main oscillator is related to temporal stability. The temporal signal stability of the seed is affected by the seed power. The output spectra of the fiber amplifier are recorded at different seed power (Fig. 3). From the spectra at an output power of 2 kW [Fig. 3(a)], the signal-to-Raman noise ratios (SRNRs) corresponding to the seed power of 133, 183, and 232 W are 31.78 dB, 44.66 dB, and 49.36 dB, respectively. As the seed power gets higher, the intensity of SRS gets lower. The spectra at the output power of 3 kW are compared with the seed power of 183 W and 232 W [Fig. 3(b)]. The corresponding SRNRs are 29.76 dB and 34.10 dB, respectively. From the comparison results, the SRS threshold is positively proportional to the seed power. During the increase in seed power, the bandwidth of 3 dB gradually broadens. The broadening of seed spectra indicates that the number of longitudinal modes gradually increases, which leads to an increase in temporal stability. Figure 4(a) shows output power and conversion efficiency corresponding to different pump power. The optical-to-optical conversion efficiency is 86.3% at the maximum output power of 3520 W. The spectrum at the maximum output power is shown in Fig. 4(b). The center wavelength is 1050 nm with an SRNR of 27.6 dB. Figure 4(c) shows the time-domain signal standard deviation (STD) at different output power. With the increase in the output power, the STD always remains at the level of about 0.003 without any features of TMI. Figure 4(d) shows the beam quality of the laser at the highest output power. The beam quality factors in X and Y directions are about 1.33 and 1.25, respectively.

Short-wavelength high-power fiber amplifiers have important applications in fields such as spectral combining. Limited by ASE and SRS, it is difficult to achieve higher power. In this manuscript, a monolithic fiber amplifier of 1050 nm is demonstrated in a counter-pump based on an LMA gain fiber of 20/400 µm. The suppression of ASE in 1070-1080 nm is achieved by shortening the gain fiber length. The temporal stability of the seed is optimized to mitigate the SRS effect. At the maximum output power of 3.5 kW, the optical-to-optical conversion efficiency is 86.3%, and the bandwidth of the output laser spectrum of 3 dB is 4.07 nm. The beam quality factors M² in the X and Y directions are about 1.33 and 1.25, respectively. The intensity of Raman Stokes light is about 27.6 dB lower than the signal light intensity. The fiber amplifier of 3.5 kW and 1050 nm meets the needs of spectral combining and has great potential for application in the field of high-power lasers.

Ring laser gyro (RLG) is an angular rate sensor based on the Sagnac effect and can measure the angular rate of a moving vehicle relative to inertial space. RLG inertial navigation system is widely applied in the defense military and commercial fields. In the defense military field, RLG inertial navigation systems are widely employed in launch vehicles, cruise missiles, ships, unmanned aerial vehicles, new fighter aircraft, transport aircraft, tanks, and the in-service transformation of armed systems. In the commercial field, different levels of RLG inertial navigation systems are also choosed in transportation, commercial spaceflight, deep sea survey, coal mine drilling, geological mapping, and other fields. RLG has the highest market share in medium and high precision gyro due to its high accuracy, high reliability, and mature engineering technology. RLG is characterized by high reliability and strong resistance to shock and acceleration without rotating parts. Additionally, it is unnecessary for RLG to turn at high speeds since the time required to reach a constant speed, with short start-up time. As RLG has a wide dynamic range, theoretically there is no upper limit to the angular velocity measurement range, and RLG has a long life of up to 100000 h or more. With the development of aviation, aerospace, navigation, and other fields, the RLG inertial navigation systems have put forward the requirements of high precision and long flight time. According to China's relevant research plan, the development of a high-precision, long-endurance RLG inertial navigation system is an important development direction in inertial technology.

To improve the accuracy of the inertial navigation system of RLG and meet the long-endurance requirements, we should carry out two aspects of the research. First, from the perspective of the inertial device itself, new material and improvement in the assembly process and sensitive characteristics should be employed to improve the performance of RLG and accelerometer. The second is inertial navigation system technology. From the perspective of the error propagation characteristics of the system, the error is comprehensively compensated and prevented according to the error model. The main technical approach is rotational modulation technology, in which the rotational mechanism drives the inertial measurement unit to rotate according to the proposed scheme to realize the error modulation function. After the accuracy of RLG reaches a high level, it is difficult to improve the measurement accuracy by enhancing the processing technology, with a long development period and high costs. The utilization of inertial navigation system technology to provide the system with higher accuracy and stronger self-sustaining power is a low-cost and efficient method to effectively improve the navigation capability of long endurance. The key technologies of the long-endurance RLG inertial navigation system can be summarized as five aspects in Fig. 5.

According to the five key technical directions listed in Fig. 4, the studies of relevant scholars in recent years are summarized. The future development of long-endurance RLG inertial navigation system technology can be mainly considered in the following aspects:

1) Further exploration of error types, models, and calibration methods for error calibration techniques

The existing error models can no longer meet the high-precision and long-endurance requirements. For the RLG inertial navigation system, more precise and perfect full-parameter error models should be built, and the corresponding error calibration method needs to be studied.

2) Further research on the error decoupling method of initial alignment technology

Since the existing initial alignment methods seldom consider the influence of errors in inertial devices, it is necessary to consider the influence of errors on the initial alignment accuracy based on a more refined and perfect full-parameter error model. Additionally, a reasonable compensation scheme, initial alignment scheme and decoupling of errors and misalignment according to the characteristics of each error should be designed.

3) Optimization of rotation modulation scheme

It is of great significance to optimize and design a reasonable tri-axis rotational scheme to overcome the limitations of single-axis and dual-axis rotational modulation techniques, eliminate the influence of earth rotation angular velocity on rotation modulation effect, and pursue high-precision navigation in long-endurance conditions.

4) Exploration and optimization of system level and device level redundancy scheme, fault diagnosis, isolation, and reconstruction method improvement

The real-time estimation method of device error parameters by joint rotary modulation of multiple inertial navigation system should be explored, and position error dispersion and long-time stability in the long-time autonomous navigation of multiple inertial guide systems need to be solved. Thus, it is also necessary to design a reasonable device-level redundant configuration to achieve a balance between mass, volume, and reliability. Considering the influence of failure signal and noise signal characteristics of the inertial navigation system and devices on the decision conclusion, the failure mode classification is extended in-depth, and the selection of detection threshold and adaptive adjustment need to be improved.

5) Building of high-precision earth gravity field model and earth rotation parameter mode

High-precision and long-endurance inertial navigation systems urgently need high-precision geophysical field parameter models. The geophysical field compensation of the inertial navigation system requires multidisciplinary integration, which is combined with the frontier knowledge of geophysics, astronomy, and other disciplines to build a more accurate and comprehensive model of the earth's gravity field and rotation parameter.

We introduce the working principle and research progress of RLG and RLG inertial navigation systems, and summarize the study of the six aspects in the long-endurance RLG inertial guidance system technology including error calibration technology, initial alignment technology, rotation modulation technology, high-reliability fault tolerance technology, multi-inertial guidance cooperative positioning technology, and geophysical field compensation technology. With the development of national defense science and technology, the demand for high-precision, long-endurance, and strong self-sustainability navigation systems is becoming increasingly urgent. With the maturity of laser gyro manufacturing and processing technology and further research and analysis on laser gyro error mechanism, the technology system of long-endurance RLG inertial navigation system will be developed more perfectly.

High-power fiber lasers have been widely used in industrial processing, automobile manufacturing, national defense, scientific research, and other fields. It has been ten years since IPG Photonics demonstrated the fiber laser output of 20 kW in 2013, which is an important milestone. However, it has also marked a dark cloud in the field of high-power fiber lasers. Although fiber laser is still in full swing, it was not until 2021 that very individual units tied the record, let alone made breakthroughs. The higher output power of the laser indicates more waste heat produced in the laser, which brings thorny transverse mode instability (TMI) and many nonlinear effects such as stimulated Raman scattering (SRS), Brillouin scattering. This is a strong negative feedback mechanism, which means that it will be harder to improve the output power. Under the current limited conditions, high stability fiber laser has gradually become a demand, and researchers have gradually carried out in-depth research on the influence of temperature on fiber lasers.

First, the research and results of the influence of temperature on optical fiber devices are discussed. The influence of temperature on LD is clear and direct, including the change of power and the drift of central wavelength. Optical fiber is different. The influence of small structural changes of optical fiber on devices such as gratings and combiners is difficult to be determined qualitatively. The physical parameters of optical fiber, such as thermal expansion coefficient, thermal diffusion coefficient, thermal conductivity, critical absorption, and emission cross sections, are difficult to be measured accurately. The experimental results of the central wavelength of laser diodes, the reflectivity of fiber grating, and the absorption and emission cross section of gain fiber with temperature are displayed, which is very helpful to simulate the behavior of laser at different temperatures.

Second, the research on the output efficiency of fiber lasers in a wide temperature range in China and abroad is reviewed. Among them, more studies focus on the temperature range higher than normal temperature. Under this condition, the efficiency of fiber lasers with longer wavelengths can be improved. On the contrary, some studies are based on cryogenic media such as liquid nitrogen to cool gain fibers, which often contribute to shorter-wavelength lasers. These studies are generally at a low level of output power, basically less than the order of 100 W, and the main body of temperature control is ytterbium-doped fiber (YDF).

Third, the experimental study on the effect of temperature on TMI and SRS of fiber laser is illustrated. As for TMI, the increase in the TMI threshold can be observed by fully strengthening YDF refrigeration. However, in some experimental studies, when the change in YDF cooling is not obvious, it is difficult to find the change in the TMI threshold. The influence of temperature on TMI is affected by laser structure, gain fiber type, fiber bending state, cooling condition, etc. The mechanism of the effect of operating temperature on the TMI threshold of fiber laser needs to be further studied. As for SRS, two experimental studies have demonstrated that lowering the temperature of YDF can inhibit SRS. At present, the preliminary analysis claims that the decrease in the working temperature of the optical fiber suppresses the spontaneous Raman noise and increases the stimulated Raman scattering threshold.

Forth, the research results of fiber lasers with high power and wide temperature operation in our research group are introduced. Due to the lack of research on the low temperature at present, our research group has carried out research on fiber laser which is at a wide temperature lower than room temperature based on the high-power fiber laser above kw level. In 2020, our research group demonstrated a fiber laser oscillator of 2 kW, which can maintain stable output in the temperature range of -10-20 ℃, and the power fluctuation is less than 7%. In 2021, our research group realized a kilowatt optical fiber oscillator with output power fluctuation of less than 7% in the range of -30-20 ℃.

The characteristics of compact structure, small size, light weight, and flexible operation contribute to the high ease of use of fiber lasers. Due to the stagnation of the maximum output power, the development of a wider temperature range and more stable output power will become one of the development trends of optical fiber laser technology in the coming period. Further expansion of the power and operating temperature range also requires in-depth study of the working mechanism of fiber lasers and the temperature characteristics of fiber devices. The development trend of wide-temperature operating fiber laser is prospected, and the research results provide a reference for the development of high-power wide-temperature operation fiber laser.

Fiber lasers feature sound beam quality, compact structure, and flexible transmission. In recent years, they have developed rapidly and have been widely applied in industrial fields, such as metal cutting, remote welding, 3D cutting, and laser marking. With the rapid development of high-power and high-brightness laser diode (LD), and fabrication technologies of double clad fibers, the output power of fiber lasers continues to increase. In addition to gaining fibers as the core raw material of fiber lasers, the performance indicators and research progress of fiber passive devices, including but not limited to fiber Bragg gratings, cladding optical filters, fiber end caps, end pump/signal combiners, and signal combiners, are also closely related to the development of fiber lasers. The cladding light filter is employed to filter the cladding light by breaking the total reflection condition of the outer boundary of the fiber cladding, which causes the cladding light to be transmitted outside the cladding through refraction, scattering, or absorption effects. The fiber end cap is a high-power device designed for processing the output end face of high-power fiber lasers and amplifiers. By expanding the core of the output fiber to reduce the optical power density at the output end, the fiber end face is protected from damage. Additionally, an anti-reflective film is applied to the output surface of the glass cone rod to avoid the back light from ruining the laser or amplifier. The function of the end pump/signal combiner is to efficiently couple several pump beams and one signal beam into the double clad fiber spontaneously. Therefore, the fusion loss and signal quality deterioration rate are two important optical performance indicators for this combiner. The signal combiner is designed to combine multiple medium power fiber lasers to obtain higher-power fiber laser output, with the advantages of compact structure, high reliability, low cost, and sound stability.

We introduce the latest research progress of the high-energy laser research team at the National University of Defense Technology (NUDT) on key passive devices adopted in high-power fiber lasers. 2 kW bidirectional cladding optical filter, 30 kW fiber end cap, end pump/signal combiner with low beam quality degradation rate, and 20 kW signal combiner with high beam quality are mainly included.

1) Cladding power filter: In 2022, the NUDT research team designed a new preparation scheme for a weak-strong-weak cladding power filter that can achieve bidirectional filtering. The weakest textured area on both sides of the double clad fiber is 1 cm, and then it enters the medium textured area by 3 cm on both sides. Finally, the strongest textured area is located in the middle area, with a length of 3 cm. The fabricated cladding power filter is tested under 2051 W input power, of which the temperature rise rate is 3.5 ℃/kW and the filtration efficiency is 20.1 dB.

2) Fiber end cap: The NUDT research team designed and built an end cap fusion system in 2014. The heating source is a hydrogen oxygen flame, and an automatic alignment fusion system is designed to achieve the fusion of different shapes of end caps and optical fibers. Based on this system, the development of 3 kW single mode fiber end caps and 6 kW multimode fiber end caps was achieved in 2015. Based on the key technologies and specially designed fiber end caps, the current fabricated end caps have been successfully applied in a 30 kW high-power fiber laser system.

3) End pump/signal combiner: By changing the fusion position of the pump arm on the signal arm, the NUDT research team improves the pump coupling efficiency of the side pump combiner under high-brightness fiber laser pumping. The optimized combiner single arm passes the pump power test of 2737 W, with a coupling efficiency greater than 99% and a temperature rise coefficient of only 6.5 °C/kW. In 2015, the NUDT research team reported a type of end face signal pump combiner that utilized heating and core expansion technology to reduce signal insertion loss. Through this technology, the combiner's pass rate was increased from 51% to 94%. In 2019, the NUDT research team utilized multi-mode field adaptive structures to achieve a transformation of core size from 10 to 50 μm. In 2022, the NUDT research team employed the corrosion-threading method to improve the performance of the signal pump combiner. The combiner prepared through this scheme maintains sound signal beam quality, the signal degradation ratio of which is only 2.2% for 25/400 μm fiber and 5.1% for 50/400 μm fiber.

4) Signal combiner: In 2018, the NUDT research team developed a 7×1 signal power combiner with the maximum output power of 14 kW and the beam quality of 5.37. To further improve the beam quality of the combined laser, the key preparation process has been optimized to produce a signal power combiner that changes the numerical aperture of the fiber core from 0.22 to 0.12. In 2019, based on a self-developed 3×1 signal combiner with an output signal fiber of 50/400 μm (NA=0.12/0.46), the combined beam quality M² was optimized to about 3.5 under 6 kW output power. In 2021, based on a self-developed 4×1 signal combiner with an output signal fiber of 50/400 μm (NA=0.12/0.46), the maximum output power was about 12 kW with the beam quality M² of less than 4, which is the ever-reported optimal beam quality for synthetic lasers greater than 10 kW. In 2022, the NUDT research team fabricated a 7×1 signal power combiner and achieved more than 20 kW output power with M² less than 4.5, which is the highest reported power in similar synthetic laser systems in the size of the 50 μm fiber core.

The NUDT research team has been engaged in research on high-power fiber optic devices for over a decade. Some of the developed devices have reached international advanced performance indicators and applied as core fiber optic passive devices in multiple major equipment and scientific research tasks. This review provides a detailed introduction to the latest research progress of high-power fiber optic passive devices, including key process methods, technical difficulties, and some considerations for the future development of fiber optic devices. There are two potential directions of fiber optic passive devices, containing further improvement in beam quality retention characteristics and implementation of integrated fiber optic devices.

High power fiber lasers with high beam quality have been widely adopted as laser sources in many applications, such as laser cutting, laser welding, manufacturing, and military defense. The advantages of high conversion efficiency, good beam quality, compact structure, and reliable stability, are favored by the consumers. There has been a remarkable increase in the output power of fiber lasers, and the business consumption of high power fiber lasers in the market shows a remarkable increase in recent years. High power and high beam quality are essential aspects of many manufacturing and defense applications.

There are two strategies to achieve high power laser beams, which are multi-laser beam combinations and power scaling of single fiber laser. The single fiber laser with high power can be directly employed as a laser source in the manufacture or serve as an element in multi-laser beam combinations. However, the power scaling of high brightness monolithic fiber laser oscillators is limited by the manufacturing techniques of fiber laser components, stimulated Raman scattering (SRS) effect, and transverse mode instability (TMI) effect. In order to achieve a monolithic fiber laser with high power and high beam quality, the SRS and TMI effects in the fiber lasers have to be effectively mitigated based on the current power handling ability of the laser components. In this paper, three strategies to achieve high power fiber lasers are respectively reviewed and discussed, which correspond to the fiber laser oscillators, the fiber laser amplifiers, and the oscillating-amplifying integrated fiber lasers.

First, the progress of kW-level fiber laser oscillators is reviewed. The published reports on high power fiber laser oscillators are reviewed in respective spatial configurations and all-fiber configurations. In the spatial configurations, reflection mirrors serve as cavities mirrors, and the coupling of pump and signal lasers are achieved by the lens. In the all-fiber configuration, the fiber Bragg gratings serve as cavity mirrors, and the pump/signal fiber combiners are packaged in all-fiber format. The all-fiber configurations have the advantages of compact structure and reliable stability. The research on high power fiber laser oscillators in recent years is outlined, and descriptions of the typical results are respectively reviewed. With the optimization of the pump scheme, fiber Bragg grating (FBG) parameters, and gain fiber parameters, the all-fiber laser oscillators are scaled to over 8 kW, and further power scaling is mainly limited by the fiber nonlinear effect and TMI effect. The development of femtosecond laser inscription of FBGs also enables the high quality FBGs for the high power fiber laser oscillators.

Second, the progress of high power fiber laser amplifiers is reviewed. Without the limitation components of FBGs, the fiber laser amplifiers can be scaled to kW-level with enough pump power and appropriate large mode area ytterbium-doped fiber. There have been many reports on kW-level fiber laser amplifiers in the last decade around the world. However, open reports on high power fiber laser amplifiers over 5 kW-level are still in a small quantity. We focus on reviewing the typical results of over 5 kW with good beam quality, which are outlined and described in detail. Output laser power of 13 kW with beam quality M² of 2.9 is reported in the few-mode region. While in the nearly single mode region, an output laser power of 6 kW with M² of about 1.3 is reported. Further power scaling is also mainly limited by the fiber nonlinear effect and TMI effect. Similarly, ytterbium-doped fibers with large mode areas and specially designed mode discrimination techniques are promising in further power scaling of the high power fiber laser amplifiers.

Third, a new conception of the combination of advantages of fiber laser oscillators and amplifiers is proposed, which is called oscillating-amplifying integrated fiber laser. Physically, it is a master oscillation power amplification (MOPA) structure with no cladding light stripper (CLS) or isolator between the seed and power amplifier. Technically, it enjoys the advantages of high anti-reflection ability and simple control logic, similar to typical fiber laser oscillators. Many researchers have investigated the new structure, and the results are diverse, especially in the application of narrow linewidth laser amplification. Reports show that the new structure also has encountered the limitations of SRS and TMI in the power scaling, which have to be mitigated in future studies.

In this work, three strategies for achieving high power fiber lasers are reviewed. The recent progress and typical results are outlined and described. In the section on fiber laser oscillators, the results are described in respective spatial and all-fiber configurations. In the section on fiber laser amplifiers, results with over 5 kW are outlined and discussed. In the section on oscillating-amplifying integrated fiber laser, recent results are described. No matter which strategy is chosen, the high power scaling encounters the limitations of fiber nonlinear effect and TMI effect. The most promising techniques to settle the limitations lie in the new design and manufacture of large-mode area ytterbium-doped fibers.

In the past few years, high-power lasers have shown great potential in the application of industrial processing, medical treatment, national defense, and other fields. Fiber lasers have become a mainstream technical solution to high-power lasers due to their unique advantages of excellent output beam quality, high energy conversion efficiency, good heat dissipation capacity, compact structure, and high robustness. Currently, a single fiber laser can reach the output power of 10 kW. However, due to energy loss, quantum defect, and other factors, the waste heat generated in high-power fiber lasers has seriously restricted the further improvement of laser output power. The performance of the high-power fiber laser is significantly affected by the high temperature inside the gain fiber, which reduces the stability of the fiber laser, and results in mode instability effect and degradation of the output beam quality. Theoretical studies have shown that the mode instability effect is accompanied by dynamic gratings caused by the thermo-optical effect in the fiber core. The high-order mode and fundamental mode are coupled with each other under the interaction of dynamic gratings, causing degraded output beam quality and further restricting the power improvement of high-power fiber lasers. Therefore, the thermal effect in high-power fiber lasers is an important factor that determines laser operation stability. It is very important to measure and monitor the temperature of the fiber core in the laser for improving the stability of the high-power fiber laser and avoid the thermal damage of the gain fiber.

The refractive index of the fiber core is affected by the temperature of the fiber material to change the phase or spectrum of the reflected/scattered light. By measuring the phase or spectrum change of the reflected/scattered light in the optical fiber, the temperature change in the optical fiber core can be detected. Fiber Bragg grating (FBG) is a reflector composed of short fiber with refractive index period modulation, and it can be employed to measure the fiber core temperature with discrete points. To study the thermal effect and photon darkening effect in the gain fiber, Leich et al. adopted four FBGs to measure the fiber core temperature in the gain fiber. The research group at the National University of Defense Technology also measured the fiber core temperature by the π-phase shifted FBG and composite grating in 2018 and 2020, respectively, which realized the real-time temperature monitoring of the gain fiber. However, FBG can only measure the fiber temperature at discrete locations and cannot achieve continuous distributed temperature measurements. It misses lots of temperature information at key locations and cannot comprehensively characterize the fiber core temperature of the high-power fiber laser. Therefore, researchers from the University of Jena measured the distributed temperature of the fiber core by optical frequency domain reflectometry (OFDR), which is a distributed temperature sensing method with extremely high spatial resolution. The temperature of fiber lasers with single mode output, several hundred-watt output, and kilowatt output power was characterized in 2015 and 2017, respectively. However, their study did not measure the absolute temperature of the fiber core in the fiber laser. Thus, researchers from the University of Defense Technology optimized the OFDR method to measure the temperature of high-power fiber lasers by decoupling the temperature and stress changes of the fiber, calibrating the temperature coefficient of the fiber, optimizing the fiber winding, and reducing the end reflection. The temperature of fiber laser amplifiers and oscillators with kilowatt output power was measured based on the optimized OFDR method. Since there are still few studies on the temperature measurement of the fiber core in high-power fiber lasers, the corresponding applications have not been fully demonstrated. The current applications include the quality evaluation of the splice point, and thermal management and nonlinear effect suppression in high-power fiber lasers, providing references for further research on high-power fiber lasers.

Fiber sensing method is an important tool to monitor the temperature of the fiber core in high-power fiber lasers. FBG can only measure the temperature at discrete locations, and the OFDR method can measure the distributed fiber core temperature with high spatial resolution. With several applications of thermal management and nonlinear effect suppression, the fiber core temperature characterization shows great potential in the performance improvement of high-power fiber lasers.

High-power fiber lasers are of application significance in scientific research and industrial processing. Stimulated Raman scattering (SRS) is a main factor limiting the power scaling of high-power fiber lasers. As a spectral filter component, chirped and tilted fiber Bragg grating (CTFBG) has been extensively studied to suppress SRS or filter Raman light in recent years. The traditional CTFBG fabrication method is the ultraviolet laser phase mask, but the fiber must be hydrogen-loaded and annealed before and after inscribing the grating, which leads to a long fabrication period. Additionally, if the annealing treatment is incomplete, the residual hydrogen molecules and hydroxyl compounds in the fiber would absorb near-infrared laser for heating to limit the power handling capability of CTFBG. To this end, the special annealing method, multiplexed inscribing technology, and efficient refrigeration packaging are proposed, but these methods and technologies greatly increase the fabrication period, cost, and complexity of high-power CTFBG. The development of femtosecond laser inscribing technology provides a new scheme for fabricating high-power CTFBG. As femtosecond lasers do not require the photosensitivity of the fiber, hydrogen loading, and annealing treatment are not needed, which shortens the fabrication period and avoids the heating caused by incomplete annealing. Meanwhile, since the fiber Bragg grating (FBG) written by femtosecond lasers feature high-temperature resistance, they have better tolerance to the temperature rise caused by a high-power laser.

Two FBGs and a CTFBG are inscribed in 20/400 μm large-mode-area double-cladding fiber by femtosecond laser phase mask method, as shown in Figs. 1 and 2. The central wavelengths of two FBGs are 1080 nm. The bandwidth and reflectivity of high reflective FBG (HRFBG) are 3.6 nm and more than 99% respectively, and those of low reflective FBG (LRFBG) are 1.2 nm and 10% respectively. The filtering band central wavelength of the CTFBG is 1133 nm with a 3 dB bandwidth of 17.3 nm and a filtering depth greater than 20 dB. The loss of CTFBG at 1080 nm signal power is less than 2% measured by the cut-off method. Two FBGs are employed to build a high-power fiber oscillator for testing CTFBG, and the CTFBG is inserted in the resonant cavity of the fiber oscillator, as shown in Fig. 3.

Figs. 4, 5, and 6 show the CTFBG testing results based on the high-power fiber oscillator. When the CTFBG is not inserted, the output power does not increase after the pump power exceeds 3500 W due to transverse mode instability (TMI). At maximum pump power, the output power is 2678 W, corresponding to the optical-to-optical conversion efficiency of 69.7% and the beam quality M² factor of 1.54. After inserting CTFBG into the resonant cavity, the SRS is suppressed with a Raman suppression ratio of ~13 dB, and the TMI is not observed. The maximum output power is increased to 2811 W, corresponding to the optical-to-optical conversion efficiency of 73.2%, and the beam quality M² factor is reduced to 1.43. During power scaling, the CTFBG is not packaged and cooled by a fan. The temperature slope of CTFBG is 7.9 °C/kW, and the maximum temperature is 44 °C。

High-power CTFBG is inscribed in large-mode-area double-cladding fibers based on the femtosecond laser phase mask method. To test the power handling capability, the CTFBG is introduced into the resonant cavity of the high-power fiber oscillator. The maximum handling power of CTFBG is 2.8 kW, and the insertion loss of CTFBG is less than 2%. The CTFBG is cooled by an air fan during the test, and the temperature slope of CTFBG is 7.9 °C/kW. This study shows that the femtosecond laser-written CTFBG has excellent power handling capability and temperature characteristics, which will promote the development and application of CTFBG. In the future, the CTFBG will be fabricated in larger core fibers by femtosecond lasers, and its performance will be further investigated in fiber lasers with higher output power.

Fiber laser has the advantages of high conversion efficiency, compact structure, and easy maintenance, and it has been widely used in industrial processing, material treatment, radar detection, and other fields. In some specific fields, requirements for the output power of fiber lasers have been further proposed, such as using high-power femtosecond laser filament principle to realize atmospheric multi-component detection, using narrow linewidth fiber lasers to achieve long-distance Doppler wind speed measurement, and using 532 nm green laser for copper material processing and 3D printing applications. By increasing the output power of fiber lasers, the detection distance and processing efficiency can be improved. However, due to the presence of nonlinear effects such as stimulated Brillouin scattering and mode instability in fibers, the output power of a single fiber is limited. Coherent beam combining (CBC) with multiple lasers is one of the effective ways to obtain high average output power fiber lasers. In a pulse fiber laser CBC system, a shorter pulse indicates wider spectral width. In a high-power continuous fiber laser CBC system, it is often necessary to apply phase modulation, so as to appropriately broaden the seed linewidth and suppress nonlinear effects in the fiber amplifier. A wider spectral width poses higher requirements on the optical path control of each laser in a CBC system. Meanwhile, there exists a coupling between the phase and the optical path. Therefore, simultaneous control of the optical path and phase in the CBC system is needed.

We investigated CBC of broadband light sources based on spectral filtering. Numerical simulations of phase and optical path simultaneous control in broadband laser interferometry were performed using a stochastic parallel gradient descent algorithm. A rectangular spectral light source with a 3 dB spectral width of 10 nm was obtained by using an amplification of spontaneous emission (ASE) light source and a fiber bandpass filter with a center wavelength of 1064 nm. Phase modulation, static optical path compensation, and real-time optical path control were achieved using phase modulators, fiber delay lines, and fiber stretchers, respectively. Two narrowband bandpass fiber bandpass filters were used to filter out laser light with center wavelengths of 1061 nm and 1066 nm and spectral widths of 2 nm, so as to achieve phase and optical path control across the entire spectrum range.

The numerical simulation results of phase and optical path simultaneous control are shown in (Fig. 5). The phase error is random, and the optical path difference is set to 50 μm. Figure 5(a) shows the light intensity change when the wavelength of (1066±1) nm laser is used as the phase control signal. Figure 5(b) shows the light intensity change when the wavelength of (1061±1) nm laser is used as the optical path control signal. Figure 5(c) shows the total light intensity variation. Before the 2100th iteration step, only phase control is performed without optical path control. After 2100 iterations, optical path control is performed once every 100 iteration steps to achieve simultaneous control of phase and optical path. After about five optical path iterations, the total light intensity can reach a normalized intensity of 1. At this time, the light intensities in Fig. 5(a) and Fig. 5(b) are also close to the corresponding normalized intensities of 1, realizing effective optical path control. In addition, an experimental system is set up to validate the results, as shown in Fig. 3. By separately enabling phase and optical path control, the interference light intensity curves from the open loop to the closed loop (Fig. 8) are obtained. Figure 8(a) shows the interference light intensity change after filtering at 1066 nm. Figure 8(b) shows the interference light intensity change after filtering at 1061 nm, and Fig. 8(c) shows the total light intensity change. From Fig. 8, it can be seen that when the system is in an open loop, the interference light intensity fluctuates randomly under the influence of noise due to the lack of phase and optical path control. After phase control is performed, similar to that in Fig. 5, the phase-controlled interference light intensity in Fig. 8(a) can reach a higher value, and the optical path-controlled interference light intensity and the total interference light intensity are stabilized around a fixed value. After optical path control is enabled, the optical path control cycle is about 1 s, and all the interference light intensities are effectively improved, indicating that effective compensation for the optical path has been performed. The conservative estimate of the optical path control range is over 0.1 ps, the phase residue error is less than λ/16, and the combined efficiency is about 90.3%.

We conduct research on the phase and optical path simultaneous control in CBC of broadband light sources based on spectral filtering. The combined laser is appropriately filtered to obtain different spectral components, and simultaneous control of phase and optical path across the entire spectrum range is achieved by separately controlling the phase of different spectral components. The feasibility of this method is verified theoretically and experimentally, and a detailed control process is provided. This method has the advantages of a clear theoretical model, high reliability, and good scalability, and it has important application value in the CBC of femtosecond pulse lasers and high-power continuous broadband lasers. Future work will focus on optimizing the control system, improving the control accuracy, using fiber gratings to achieve different bandwidths filtering, and clarifying the influence of filtering wavelength on the combining effect. This method will also be applied to ultra-short pulse CBC and the realization of phase and delay locking between different spectra, achieving more channels and wider spectral width in ultra-short pulse CBC.

Compared with the laser diode (LD)-pumped scheme, the tandem-pumped scheme has the advantages of high pump brightness, small quantum defect, and low thermal load, making it the main technical solution for ultra-high power ytterbium-doped fiber lasers (YDFLs). Nevertheless, due to the weak absorption of Yb-ions at 1018 nm, it is necessary to enlarge the core diameter or increase the length of YDF to improve the total pump absorption, resulting in more severe nonlinear effects [mainly stimulated Raman scattering (SRS)] and transverse mode instability (TMI). Despite the 20 kW YDFLs have been obtained based on YDF with a core diameter of about 50 μm and tandem-pumped scheme, the beam quality is poor for the lacking of efficient mode control. It is very challenging to maintain excellent beam quality with increasing power actually, due to the contradictions between the SRS and TMI suppression strategies. Although IPG photonics has announced a 10 kW single-mode YDFL tandem pumped by 1018 nm YDFLs as early as 2009, no other institutions have ever reported >5 kW YDFL with high beam quality (M²<2) as of 2021. We presented the research progress on tandem-pumped YDFLs achieved by National University of Defense Technology in the past three years. Possible approaches to enhance power and beam quality were also discussed.

To balance high power and high beam quality, we proposed a backward/bidirectional tandem-pumped solution. Although backward/bidirectional pumping schemes are widely applied in LD-pumped laser systems, investigations on their technicalities in tandem-pumped YDFL are very rare in open publications. The primary reason is that in the tandem-pumped YDFL, the 1018 nm fiber laser which functions as the pump source, is highly susceptible to the signal light. Even a small proportion of signal light coupled into the 1018 nm fiber laser might cause significant power decline, even leading to the destruction of the 1018 nm laser. By optimizing the 1018 nm laser oscillator and the backward combiner, we successfully reduced the adverse effects of signal laser on the pump source and ensured the stable operation of the backward/bidirectional pumping system. Then by employing the optimized 1018 nm fiber lasers as a pump source, the benefits of a backward tandem pump were first demonstrated with conventional 25/250 μm YDF. The signal power was boosted to 5 kW (M²=1.54) free of TMI or SRS. On the contrary, the SRS threshold of the YDFL was only 3.94 kW when the YDF was forward pumped. Afterward, the bidirectional pump scheme was also tested with 30/250 μm YDF. A 6.22 kW laser output with M²=1.53 was obtained by using a 30/250 μm YDF, but further power scaling was limited by SRS. For higher power, the backward pump scheme was applied and the maximum laser power reached 10.03 kW (M²=1.92) without SRS or TMI.

In addition to optimizing the pump scheme, special fiber designs were also considered. The geometric or optical structure in the transverse or longitudinal direction of the YDF was modified for SRS suppression and mode control. By using confine-doped YDF (CYDF) of which only part of the core was selectively doped, it was possible to tailor the gain of high-order modes for better beam quality. The influence of key parameters, including the doping ratio, core diameter of CYDF, and mode content of seed laser, on the beam cleanup effect of CYDF was numerically analyzed. According to the simulation, the CYDF with core/inner cladding diameter of 40/250 μm and a relative doping ratio of 0.75 was designed, fabricated, and applied in a backward tandem-pumped YDFL. 10.1 kW laser power at 1080 nm was realized with M²= 2.16. The beam quality was superior to that of conventional double-clad YDF with an equivalent core diameter. Besides, tapered YDF (TYDF) that has varied core and cladding diameter along the longitudinal direction was also fabricated and used in tandem pump for the first time. The core/inner cladding diameter of the input and output end of the TYDF was 30/250 μm and 48/400 μm respectively. The beam quality of the signal laser was well maintained during the high-power scaling process. The M² factor was measured to be about 2.2 at 10.13 kW, which was much better than that of the 48/400 μm YDF drawn from the same fiber preform. Our special structured fibers have demonstrated superior SRS suppression and mode control compared with conventional double-clad fibers.

Our team has taken a lead in conducting research on high-power backward/bidirectional tandem pumps in China, which has resulted in a significant improvement in the SRS threshold while keeping good beam quality of the YDFL. We have achieved fiber laser output of tens of kilowatts based on conventional YDF, CYDF, and STYDF, with significantly better beam quality than fiber lasers pumped by LD of the same power. However, it should be noted that further improvements to the beam quality and power of the tandem-pumped YDFL still pose significant challenges. Existing solutions have not been able to achieve >10 kW single-mode laser. Next, the team will continue to deepen the research on the evolution of SRS and TMI at extremely high-power levels. The focus of our future work is to improve the pump absorption of the gain fiber, reduce the NA of the core, and increase the loss of higher-order modes. With the help of high-performance gain fibers and advanced fiber optic devices, we hope to steadily improve the output power and beam quality of tandem-pumped YDFLs.

Excessive electromagnetic radiation not only affects the normal operation of precision electronic equipment but also poses a potential threat to human health. Traditional microwave absorbers such as iron powder possess high magnetic permeability for effectively dissipating electromagnetic waves, but the high density restricts the applications. Although carbon materials have lower density and better conductivity, the electromagnetic wave attenuation performance needs further improvement for unfavored impedance matching. Aiming at developing the microwave absorber with light weight, high efficiency, and wide band, recently metal/carbon composites prepared by pyrolyzing metal-organic framework(MOFs) for microwave absorption have become research hotspots for their light weight and high efficiency. However, the pyrolysis process is usually performed in muffle or tube furnaces, and the slow heating rate usually leads to aggregated metal nanocrystals, thus resulting in weak interface polarization. Additionally, the preparation takes a long time (a few hours) and the pyrolysis atmosphere is usually inert gases such as argon, which makes it difficult to achieve rapid preparation. Therefore, we take laser beams as the thermal heat resource instead of traditional heat resources to achieve rapid preparation, and the micromorphology could be precisely controlled by the fast heating-quenching process, with metal nanocrystal size at nanoscale of 5-20 nm. Finally, the dielectric loss ability is enhanced with effective electromagnetic wave absorption.

First,we prepare the ZIF-67 precursor. 0.24 g Co (NO₃)₂·6H₂O is dissolved into 30 mL water as solution A. 1.64 g 2-methylimidazole is dissolved in 30 mL water as solution B. Solution B is dropped into solution A dropwide with the reaction for 24 h. By centrifugation for three times, the purple powder ZIF-67 is obtained.

Firstly, laser beam irradiation is introduced as a thermal heat resource to pyrolyze ZIF-67, and Co/C composites are successfully prepared in just a few seconds. The micromorphology of Co/C could be precisely controlled just by modulating the irradiation power, with Co nanocrystal size of about 5-20 nm by 45-55 mW laser irradiation, as shown in Fig. 1(e)-(f). The Co nanocrystal deposited on thin carbon film guarantees the improved polarization loss ability. XRD spectra and XPS in Fig. 2 further confirm the composites phase of Co and CoO deposited on carbon. The laser power could influence the micromorphology and phase of the Co/C composites, thus affecting the complex permittivity and permeability, as shown in Fig. 3. The dielectric loss ability is improved by the higher power laser irradiation, but exorbitant laser power would lead to carbon oxidation and lowered dielectric loss. With benefits from the fast heating-quenching process, the nano-sized Co deposited on thin carbon film is prepared with satisfied microwave absorption ability (Fig. 4). The reflection loss of Co/C composites prepared by 35 mW, 45 mW, and 55 mW laser irradiation could reach -23.5 dB, -53.9 dB, and -46.3 dB respectively, which indicates excellent microwave absorption ability. Specially, the working band of Co/C composites by 45 mW laser irradiation could cover the X band (8.8-13 GHz) just at the thickness of 2.6 mm to exhibit the light weight and wide absorption properties. The microwave absorption mechanisms by Co/C composites have been discussed in Figs. 5-6, which confirms that rich electromagnetic loss mechanisms (polarization loss, conductive loss, ferromagnetic resonance, and eddy current) synergized with good impedance matching induce favorable microwave absorption.

We introduce laser beams as the heat source to pyrolyze ZIF-67 and the nano-sized Co/C composites are successfully prepared in just a few seconds. The prepared Co/C shows excellent microwave absorbing properties with the effective absorption band covering X to Ku band. The improvement of electromagnetic wave loss ability is mainly attributed to the synergistic effect of dielectric loss (interface polarization, dipole polarization, and conductive loss) and magnetic loss (resonance loss and eddy current loss) by Co/C composites. Compared with the traditional thermal treatment process, the laser irradiation process exhibits low request for facility and environmental, making it feasible for industrial production and ushering in a new way for preparing the lightweight and wide-band microwave absorbers with MOFs as the precursor.

Bioaerosols are colloid systems formed by tiny biological particles floating in the atmosphere, such as bacteria, fungi, viruses, and pollen. Bioaerosol particles can scatter and absorb solar radiation, communication signals, and optical signals, and can also be suspended in the air and spread long distances with the wind. Therefore, they exert an effect on many fields such as climate change, optical communication, and optical remote sensing. They have also attracted widespread attention in research areas such as functional materials, environmental protection, and disease transmission and prevention. Bioaerosol particles have diverse morphologies and rich compositions and can absorb and scatter incident light in multiple wavelengths. Additionally, their artificial cultivation techniques are relatively mature, with low cultivation costs, simple operation, and short cultivation periods. As a result, the artificially prepared bioaerosols have great development potential in the research on optical functional materials. Drying is an essential step in the artificial preparation of bioaerosols, and it can affect the broadband extinction performance of artificially prepared bioaerosols. However, the influence of different drying methods on the extinction performance of bioaerosols has not been studied. Our paper aims to investigate the influence of different drying methods on the extinction performance of artificially prepared bioaerosols and provide references for improving the extinction performance of bioaerosols through preparation.

BB0819 spores are one of the common entomopathogenic fungal spores and have been proven to have significant broadband extinction performance. Therefore, BB0819 spores are selected as the research object to investigate the effect of drying methods on the extinction performance of artificially prepared bioaerosols. In our study, two common microbial treatment methods, freeze drying and hot-air drying, are chosen to dry BB0819 spores. The preparation process consists of four main steps, including spore suspension preparation, solid culture medium cultivation, drying, and spore collection. To investigate the effect of the two drying methods on the broadband extinction performance of BB0819 bioaerosol, we build a model with detailed structural information on bio-particle morphology, and the Kramers-Kronig algorithm, discrete dipole approximation, and Monte Carlo simulation methods are adopted to calculate the extinction performance. Furthermore, Fourier transform infrared spectrum and two-dimensional correlation infrared spectrum are employed to analyze the internal composition and protein structure differences of bioaerosols after drying, explaining the changes in extinction performance. The reliability of the simulation results is confirmed through experimental validation in an aerosol chamber, where transmittance data is obtained for bioaerosols produced by different drying methods in the mid-infrared and far-infrared wavelengths.

The drying process can disrupt the internal composition of spores. This leads to changes in the absorption functional groups, protein secondary structures, and various component contents, which in turn affects the optical absorption of spores. Compared with freeze drying, hot-air drying results in a higher protein absorbance in BB0819 spores, about 7.19%. The increased protein content can enhance the optical absorption ability of spores in the 6-8 μm wavelength range. Although the protein absorbance in BB0819 spores is relatively higher after hot-air drying, the fitting results of protein secondary structure indicate that the protein absorbance of α-helix structure in spores decreases by about 5%. The α-helix is the main structure maintaining protein conformation, indicating that although the protein content is not reduced by hot-air drying, it has a significant influence on the stability of protein structure. Additionally, the rising temperature during hot-air drying leads to significant denaturation of polysaccharide substances (mainly peptide polysaccharide layer) and changes in the content of polysaccharide groups inside the material. Polysaccharides contain abundant C—O and C—C bonds, and the stretching and vibration of these chemical bonds have a strong absorption peak at around 10 μm, which may affect the optical absorption performance of spores around 10 μm. As shown in Figs. 7(c)-(f), in the 2.5-10.3 μm wavelength range, freeze-dried spores have a larger extinction cross-section and stronger extinction performance for individual spores, while in the 10.3-15.4 μm wavelength range, hot-air dried spores have a larger extinction cross-section and stronger extinction performance for individual spores. Both BB0819 bioaerosols dried by the two drying methods show significant optical attenuation ability. In the 2.5-10.3 μm wavelength range, both bioaerosols can attenuate the transmittance of incident light to below 20%, with most of them below 15%. In the 10.3-15.4 μm wavelength range, freeze-dried spores can attenuate the transmittance to between 20% and 35%, while hot-air dried spores can attenuate the transmittance below 30%.

Drying is an essential step in the artificial preparation of bioaerosols, and it affects the broadband extinction performance of artificially prepared bioaerosols. Research results show that bioaerosols after freeze drying have higher content of polysaccharides and more stable protein structures, which leads to better extinction performance in the far-infrared band. Bioaerosols obtained through hot-air drying contain more proteins for better extinction performance in the mid-infrared band. In the mid-infrared band, selecting bioaerosols obtained by freeze-dried spores can reduce the average transmittance from 11.95% to 9.14% within three minutes, while in the far-infrared band, adopting bioaerosols obtained by hot-air dried spores can reduce the average transmittance from 34.38% to 26.03% within three minutes. We clarify the effects of drying methods on the extinction properties of artificially prepared bioaerosols and provide references for improving their extinction properties through preparation.

Supercontinuum (SC) has experienced a boom in recent decades because of its rich spectral compositions and laser characteristics. At present, the studies of SC mainly focus on power scaling, spectrum extension, and spectrum flatness improvement, among which power scaling enables abundant potential applications, such as photoelectric countermeasures, optical coherence tomography, and hyperspectral lidars.

In the photoelectric countermeasure system, a high-power broadband source is employed to suppress and disturb enemy equipment. For example, the AN/AAQ-24(V) directional infrared countermeasure system jointly developed by the United States and Britain has the function of laser jamming, and the corresponding jamming range covers the entire near-infrared waveband. In the equipment of optical coherence tomography, the material sample is scanned by a broadband source, thus achieving the reconstruction of its two-dimensional or three-dimensional image. Scanning resolution, speed, and sensitivity are three key performance parameters in this equipment, in which the scanning sensitivity can be improved by a high-power source, and the axial resolution can be enhanced by a broadband source. With these factors considered, a high-power SC source is an appropriate choice. In the hyperspectral lidar system, the active hyperspectral detection covering broadband wavelength is attractive in long-range target identification. It has been reported that the active hyperspectral detection of diffusion target has a measurement range of several hundred meters and requires the utilization of quite expensive instruments to produce and detect infrared laser radiation. The laser power is one of the main factors to determine the measurement range limit in this system. The high-power SC source has great application prospects in remote hyperspectral sensing and lidar performance enhancement due to its unique characteristics, such as good direction, broadband wavelength range, and high spectral intensity.

We review three major schemes generating high-power visible to near-infrared SC on the main oscillating power amplification (MOPA) structure, random fiber laser structure, and multichannel incoherent combination. Specifically, in the MOPA structure scheme, there are two SC generation schemes according to the generated waveband. One is adopting a MOPA structure combined with a photonic crystal fiber (PCF) or a graded-index multimode fiber (GRINMMF) to achieve a visible SC output. Some typical PCFs for high-power SC generation are also introduced (Fig. 2). The other is that the near-infrared SC generates directly from a fiber amplifier. For the SC generation scheme in a random fiber laser, several typically experimental structures are reviewed in detail. In the multichannel incoherent combination scheme, the broadband power combinations of visible SC and near-infrared SC are listed respectively. The advantages and disadvantages of these schemes and their future development potential are comprehensively analyzed.

The visible SC output power reaches 300 W based on the MOPA structure scheme, the near-infrared SC output power has reached 3 kW based on the scheme of random fiber laser, and the emergence of new fibers and schemes brings new energy for the development of high-power SC sources.

In terms of high-power visible SC, PCF is the main nonlinear medium and the corresponding studies focus on its structure design. However, the mode field diameter of PCF is small, which means it has less potential for further SC output power scaling. With the further development of the GRINMMF, this fiber with a large core size, beam self-cleaning effect, and unique mechanism of short-wave expansion can promote the further development of high-power visible SC. At present, the generated SC spectral properties of GRINMMF are poor compared with those of PCF. Believing in the future that the SC output power and spectral performance based on GRINMMF can be further improved by optimizing the refractive index curve, doping concentration, and fiber structure. In addition, most of the reported high-power visible SC is achieved by the MOPA structure scheme, and the scheme of multichannel incoherent combination can also scale the visible SC output power effectively, which can be further improved by optimizing the design of the broadband power combiner in the future.

For high-power near-infrared SC, the MOPA structure scheme is complex, but it can provide high pump peak power under the premise of ensuring the average pump power, and the generated spectral performance of SC is excellent. For the scheme of the random fiber laser, the generated SC structure is simple with high obtained SC output power, which also needs to achieve more development theoretically and experimentally in the future. The scheme of multichannel incoherent combination has the potential to break the limit of SC output power in the single fiber, but the current development is relatively slow due to the small market demand at home and abroad. However, when the output SC power of the single fiber is close to the limit in the future, the scheme will show its advantages.

We select some representative studies of high-power visible to near-infrared SC at home and abroad in terms of the above three schemes and focus on demonstrating the research progress of the National University of Defense Technology in recent years. With the improvement in the fiber drawing technology and semiconductor laser output power, and the gradual application of SC sources in photoelectric countermeasures, optical coherence tomography, and hyperspectral lidars, high-power SC sources can be further developed.

Cavity optomechanical system enables a hybrid system where optical resonant mode and mechanical oscillation mode are coupled together. Benefiting from the compatibility of the manufacturing process with the semiconductor industry, a strong optomechanical interaction between the optical cavity and mechanical resonator can be achieved in nanoscale silicon photonics devices. The nanoscale optomechanical system can considerably improve the interaction between optics and mechanics mediated by the optical force, enabling the development of ultra-stable and ultra-narrow linewidth mechanical oscillation signals. Cavity optomechanics research on classical mechanics and quantum systems has been widely utilized to provide an unprecedented platform for high-precision sensing, cavity optomechanical devices and circuits, and high-efficiency microwave-optical conversion devices. We propose a nonlinear cavity optomechanical coupling scheme by introducing a nonlinear mechanical oscillator integrated with a linear mechanical oscillator to achieve the high-performance optomechanical frequency comb. This novel nonlinear optomechanical system can provide a solution for the high-performance cavity optomechanical frequency comb and hence effectively promote the development and application of photonic microwave source technology.

A nonlinear mechanical oscillator integrated with a linear mechanical oscillator is introduced into an optomechanical system, and the optomechanical frequency comb is experimentally observed with a series of equally spaced and discrete frequencies. In this optomechanical system, a zipper cavity consists of two photonic crystal nanobeam (PCN) cavities. One PCN is doubly clamped to work as a fixed PCN for laser pumping, and the other movable PCN is integrated with a 2-degree-of-freedom (2-DOF) vibration system to transform the optical force acting on two mechanical oscillator. In the 2-DOF oscillator system, a curved nonlinear mechanical oscillator is integrated with a linear mechanical oscillator. Based on the degenerated multimode mixing method, the mechanism of the optomechanical frequency comb is theoretically analyzed. These nonlinear mechanics are coupled to the optomechanical interaction, and the response of the light field can be eventually detected to observe the optomechanical frequency comb.

The optical performance of the zipper cavity is investigated for the following detection of mechanical spectra, and the optical spectrum is divided to identify the laser sweeping range for the observation of optomechanical frequency comb. In this coupled PCN cavity with a high Q_o, the even cavity mode has a high optomechanical coupling strength to detect the in-plane oscillation of the 2-DOF oscillator system. The characterization of the nonlinear optomechanical device, a power spectral density (PSD) map, is conducted to observe the mechanical oscillations by the wavelength swept through the resonance of TE_1,e and TE_2,e modes. When the laser is swept into the optomechanical frequency comb zone at the TE_2,e resonance mode, two sidebands of the mechanical resonant frequency are generated with some symmetrical side lobes with a frequency-spacing of 81.5 kHz, which behaves very similar to an optical frequency comb. At the TE_1,e resonance mode, cavity energy inside the fundamental optical mode is higher than that of TE_2,e mode, so the nonlinear optomechanical system is driven by a higher optical force. The optomechanical frequency comb at the TE_1,e resonance is observed with more excited side lobes with a frequency-spacing of 81.3 kHz due to the stronger optomechanical coupling between the 2-DOF vibrator and optical cavity induced by higher cavity energy. As a result, the measured optomechanical frequency comb in the TE_1,e cavity mode is clearer than that of TE_2,e cavity mode when the pump power is 25μW, which indicates that the generation of the optomechanical frequency comb is highly dependent on the energy in the intracavity.

In the paper, the optomechanical frequency comb is experimentally observed in a nonlinear optomechanical system. In this optomechanical system, a nonlinear oscillator is integrated with a linear oscillator to form a 2-DOF oscillator. In analogue to the generation of optomechanical frequency comb, a numerical model is theoretically proposed to analyze the nonlinear optomechanical coupling by introducing a nonlinear term. When the nonlinear force is acted on the 2-DOF oscillator, the degenerate four-wave mixing of mechanical modes can contribute to the generation of an optomechanical frequency comb. In the experiments, a demonstration of the optomechanical frequency comb is conducted using the TE_1,e and TE_2,e modes. From the analysis of the optomechanical frequency comb, it can be concluded that first, the optomechanical frequency comb always arises at the red-detuned sideband, and it is highly dependent on the optical power coupled into the cavity. Second, a remarkable optomechanical frequency comb is excited in the fundamental cavity mode much more visible than that of the second cavity mode. In future practical applications, the optomechanical frequency comb can be potentially used as a microwave photonic source, which can provide a reference clock signal in the frequency range of MHz.

Linearly polarized supercontinuum laser source has linear polarization characteristics, which provides a new dimension for the research and application of supercontinuum. In recent years, many linearly polarized supercontinuum laser sources have been reported, and some achievements have been made in high birefringence realizing, high coherence and low noise, power increasing, and spectrum broadening. However, there are still some factors that limit the development of linearly polarized supercontinuum laser sources, and it is meaningful to make a summary and propose prospects for linearly polarized supercontinuum laser sources.

In existing reports on linearly polarized supercontinuum, the phenomenon of polarization extinction ratio (PER) degradation of linearly polarized supercontinuum has been found. However, due to the inability of the existing method to reflect the polarization states at each wavelength of the linearly polarized supercontinuum, the reason for PER degradation cannot be explained. We introduce two PER measurement methods suitable for broadband linearly polarized supercontinuum laser sources.

We review the research results of linearly polarized supercontinuum fiber sources, mainly focusing on high PER realization, output power increasing, and spectrum broadening. The measurement methods of PER are introduced, especially two methods for linearly polarized supercontinuum laser sources.

The research progress of linearly polarized supercontinuum is mainly listed below.

1) The design of polarization-maintaining photonic crystal fiber (PM-PCF). Traditional polarization-maintaining fiber realizes birefringence through stress rods. The birefringence coefficient is usually on the order of 10^-4. PM-PCF can achieve birefringence through the irregular arrangement of air holes. The birefringence coefficient of optimally designed PM-PCF reaches the order of 10^-3 at the wavelength of above 1000 nm, resulting in a linearly polarized supercontinuum laser with PER of 21.2 dB.

2) In terms of coherence. Linearly polarized supercontinuum laser source generated in all-normal dispersion PM-PCF, which is essentially based on coherent nonlinear effects such as self-phase modulation and optical wave breaking, has a high degree of coherence. The coherence of linearly polarized supercontinuum lasers generated in all-normal dispersion PM-PCF is greater than 0.7 in the wavelength range of 650 nm.

3) Output power increasing. The output power of linearly polarized supercontinuum fiber lasers generated in PM-PCF is limited by the small core of the PM-PCF and its splice with the output fiber of the pump laser. By optimizing the splicing parameters, the coupling efficiency between PM-PCF and pump laser reaches 90%, and the output power of linearly polarized supercontinuum generated in PM-PCF reaches 93 W. Linearly polarized supercontinuum laser directly generated in polarization maintaining fiber amplifier is more suitable for output power increasing due to its extremely low loss of fiber splicing. The highest output power of linearly polarized supercontinuum reaches 322.5 W.

4) Spectrum broadening. The short wavelength edge of linearly polarized supercontinuum generated in PM-PCF reaches 300 nm, and the long wavelength edge reaches 2300 nm. The spectrum of linearly polarized supercontinuum generated in PM-ZBLAN covers 350-4500 nm. The spectrum of linearly polarized supercontinuum generated in chalcogenide rib waveguide covers 2-10 μm.

It can be expected that in the near future, with the output power of the polarizing amplifier increasing and the use of large core nonlinear fiber, the output power of linearly polarized supercontinuum generated in polarization-maintaining fiber amplifier can exceed kilowatts. With the coupling efficiency between PM-PCF and pump laser increasing and better thermal management, the output power of the linearly polarized supercontinuum generated in PM-PCF is expected to exceed 200 W, and the spectrum is expected to reach the short wavelength and long wavelength limits of silica fiber.

The PER of a linearly polarized laser is usually measured by a rotating polarizer. This method is suitable for single-wavelength fiber lasers but cannot be used to measure the PER of each wavelength of a wide spectrum laser source such as supercontinuum. Two measurement methods suitable for the PER measurement of linearly polarized supercontinuum are introduced. By using these two methods, we can study the inherent mechanism of linearly polarized supercontinuum laser sources.

Rydberg atoms became increasingly crucial in the last decade because of their fascinating characteristics that distinguish them from conventional radio frequency (RF) sensors. First, the Rydberg atoms are self-calibrating thanks to the invariance of the atomic parameters, and their response is linked to Plank's constant. Second, atomic sensing systems break a key assumption behind the Chu limit of traditional electronic sensors by allowing a small vapor cell to operate over multiple octaves of frequencies from DC to THz. Third, instead of demodulated circuitry, Rydberg atoms can naturally extract the baseband signals from the carrier frequency. Fourth, Rydberg atoms may avoid internal thermal (Johnson) noise, even at room temperature. In recent years, the amazing introduction of the local oscillator (LO) RF field has assisted us in controlling ensembles of Rydberg atoms. However, most current reports on Rydberg atomic heterodyne sensors focus on measurements in the resonant region, which can only achieve highly sensitive detection at discrete frequencies due to the quantum nature of the atomic energy level. In this work, by extending the Rydberg atomic heterodyne technique from the resonant region to the off-resonant region, we experimentally validated the continuous broadband and high sensing sensitivity of Rydberg atoms.

When a strong LO field and a weak signal (SIG) field with frequency detuning on the order of kHz are irradiated to the atoms, the energy level will be modulated by the intermediate frequency (IF) in the resonant and off-resonant regions, which can be directly detected by optical electromagnetically induced transparency (EIT). At room temperature, a probe laser of 852 nm and a coupling laser of 509 nm propagate in opposite directions and overlap inside a 2 cm-long vapor cell containing cesium atoms, exciting the atoms to the Rydberg state for atomic sensing. In the resonant region, the LO frequency is set to 2.63 GHz, and the SIG frequency is set to 2.63 GHz+10 kHz. Both fields are illuminated into the vapor cell by a horn antenna 7 cm away from the optical path, and the polarization of the two RF fields is the same as that of the probe and coupling beams and propagates in a vertical direction to the laser beams. While in the off-resonant region, the frequencies of the LO and SIG fields are tuned to 300 MHz and 300 MHz+10 kHz, respectively. An aluminum parallel-plate waveguide serves as the microwave transmitter in the off-resonant region. The reflection coefficient (S₁₁) of the input port is below -20 dB from DC to 850 MHz (Fig. 2), indicating the excellent port matching performance of the parallel-plate waveguide.

In the resonant region, we calibrated the electric (E) field strength of the RF field using the Autler-Townes (AT) splitting effect. By adjusting the output power of the signal generator to satisfy the linear relationship between AT-splitting and RF field amplitude, we obtained the relationship between the square root of the signal generator output power and the E-field intensity calculated by AT-splitting (Fig. 3). The results show excellent linearity, and the weak RF E-field strength can be inferred from the fit line. Then, a spectrum analyzer was used to measure the intensity of the beat-note signal under Rydberg atomic heterodyne conditions. We measured a series of data points of the beat-note signal strength versus the applied SIG power (Fig. 4). The intensity of the received beat-note signal is approximately proportional to the strength of the applied SIG field with a linear dynamic range of over 45 dB. The minimum SIG output power is -85 dBm, which is limited by the background noise of the spectrum analyzer. By leveraging the gradient of the fit line, we can obtain the minimal detectable E-field of 220.94 nV/cm, with the corresponding sensing sensitivity of -131.9 (dBm/cm²)/Hz. Similarly, in the off-resonant region, through the relationship between the power injected in the parallel-plate waveguide and the E-field strength, we measured the minimum E-field strength of 19 μV/cm in the off-resonant region at 300 MHz, with a sensitivity of -93.2 (dBm/cm²)/Hz. Besides, we also measured the instantaneous bandwidth of the system in the off-resonant region (Fig. 5). By taking into account the negative detuning of the SIG and LO fields, the instantaneous bandwidth of 3 dB of the system reaches 90 kHz.

In the present study, two typical frequency points in the resonant and off-resonant regions were selected to experimentally verify the broadband and high sensitivity detection capability of Rydberg atomic sensors. During the measurement, a horn antenna and a parallel-plate waveguide were used as microwave transmitters in the resonant and off-resonant regions, respectively. Using the Rydberg atomic heterodyne technique, we successfully measured a minimum E-field strength of 220.94 nV/cm with a sensitivity of -131.9 (dBm/cm²)/Hz in the resonant region at 2.63 GHz and a minimum E-field strength of 19 μV/cm with a sensitivity of -93.2 (dBm/cm²)/Hz in the off-resonant region at 300 MHz, respectively. In principle, by adjusting the laser frequency to excite the alkali metal atoms to various Rydberg states and incorporating the distinct responses of Rydberg atoms to E-fields in the resonant and off-resonant regions, highly sensitive sensing of microwave E-fields can be achieved in the broadband continuous spectral range.

Coherent combining of fiber lasers by active phase control is an effective way to break through the power limit of a single fiber laser and achieve higher output power while maintaining good beam quality. Based on the research progress in China and abroad, this paper introduces the representative achievements in the past 20 years made by the coherent beam combination research group in National University of Defense Technology and presents the prospect of coherent beam combining (CBC) of fiber lasers.

We present our representative achievements in CBC of fiber lasers in this paper, which are organized as follows.

First, the high power key components for CBC were designed and manufactured. Various types of fiber amplifiers have achieved power breakthroughs. For example, a 500 W level single-frequency fiber amplifier, 7 kW level narrow line-width fiber amplifier, and 500 W femtosecond fiber amplifier were obtained. High-power phase modulators based on piezoelectric ceramics were developed. We also designed two kinds of high power adaptive fiber-optics collimators (AFOC), which were based on flexible hinges and piezoelectric bimorph actuators respectively.

Second, the active phase control of fiber lasers was studied. Various phase control methods were deeply researched, including the stochastic parallel gradient descent (SPGD) algorithm, dithering technique, heterodyne interference measurement technique, and deep learning algorithm. Some innovative phase control techniques were proposed to increase the control bandwidth, such as the single dithering technique, orthogonal dithering technique, and cascaded phase control technique.

Third, we also studied the high precision control of other optical parameters for CBC, including optical path difference control, tilt-tip control, and defocus aberration control. For example, we proposed an all-fiber optical path difference adaptive control method and simultaneously controlled phase and optical path in coherent combing of broadband light sources based on spectral filtering. In addition, a collimator was designed for defocus aberration compensation.

Fourth, beam combination techniques were demonstrated. Beam combination can be classified into tiled aperture and filled aperture. In the aspect of tiled aperture, a series of beam combination methods with high fill factor were designed and developed. For example, we proposed a coherent fiber-optics-array collimator that was mainly composed of a single unitary collimating lens and a prism. We also proposed a novel scheme of fiber collimator based on rod lens, which had good application prospects in the CBC of a large number of fiber lasers. In the aspect of filled aperture, we experimentally testified coherent polarization beam combining (CPBC) of eight low power fiber lasers, and 5.02 kW output power was obtained by CPBC of four fiber lasers with combining efficiency of 93.8% and beam quality of M²<1.3.

Fifth, based on the enabling technology mentioned above, a number of experimental systems were built. For high power fiber laser CBC systems, 1.08 kW output power was obtained by coherent combing of nine fiber lasers in 2011; CBC of a seven-channel fiber laser array with 7.1 kW overall output power was reported in 2020, and 21.6 kW was generated by CBC of 19 fiber lasers in 2021. For a large number of fiber laser CBC systems, phase locking of 32, 60, and 107 fiber lasers was realized by using the SPGD algorithm in 2014, 2019, and 2020 respectively. Based on the heterodyne interference measurement technique, efficient phase compensation of 397 and 1027 laser channels were realized in 2022 and 2023 respectively. For the pulsed fiber laser CBC system, 1.2 kW average power was generated by the coherent combining of seven nanosecond fiber amplifiers array in 2013; CPBC of two-femtosecond fiber lasers was realized with 313 W average power in 2018, and CPBC of two ultrafast laser channels was realized based on fiber stretcher and SPGD algorithm in 2022. For target-in-the-loop CBC systems, CBC of a fiber laser array with nine channels and 100 W level was reported in 2013, and atmospheric turbulence compensation was realized over a 1 km level propagation path for a six-channel fiber laser array based on target-in-the-loop CBC in 2018. In addition, CBC of fiber lasers with special wavelengths such as 1018 nm and 2 μm has also been achieved.

Sixth, the novel compact internal sensing phase locking techniques were presented. By using those techniques, the phase noises in the laser channels can be detected and compensated for before the lasers form the laser array. Based on spatial structure, internal phase locking of 12 fiber lasers was realized, and 1.5 kW output power was generated by CBC of three fiber lasers. Based on an all-fiber network, methods to compensate for π‑ambiguity between channels were proposed, and CBC of three fiber lasers was experimentally verified.

Seventh, CBC technique was employed for light field control, and special light fields such as vortex beams and vectorial beams were generated. For example, by CBC of six fiber lasers, a vortex beam with an output power of more than 1.5 kW has been generated.

Our group has researched CBC for nearly 20 years. Some representative results have been achieved. Artificial intelligence and light field control have been integrated with CBC. Some innovative breakthroughs have also been made in interdisciplinarity. The scientific research results have been continuously added to undergraduate and graduate courses such as Physical Optics and Advanced High Energy Laser Technology. A large number of graduate students have become the backbone force of scientific research. In the future, we will focus on the development of science and technology, student education, and talent cultivation integrally and make unremitting efforts to produce innovative results in this field.

The photonic lantern is a new type of photonic device that combines the advantages and characteristics of single-mode fiber and multi-mode fiber. It has important applications in the fields such as astronomical photonics, optical fiber communication mode division multiplexing, and optical fiber laser mode control.

This review introduces the structure and mode evolution theory of photonic lanterns, fabrication technology, the mode adaptive control based on photonic lanterns, their application in high-power fiber laser amplifiers to suppress the transverse mode instability, and utilization in large mode area fiber to excite special structural beams (such as orbital angular momentum modes). Through theoretical simulations and experimental exploration, the original design and fabrication criteria of photonic lanterns are improved. Meanwhile, two key design criteria for mode adaptive control are added: 1) optimizing the input fiber arrangement to improve the control bandwidth; 2) selecting the appropriate input core cladding ratio to expand the optional range of the output fiber. According to the above design requirements, N×1 photonic lanterns with excellent performance are prepared (N=3,5,6,7,…), as shown in Fig. 12. The phase of the input beams is actively modulated by the stochastic parallel gradient descent (SPGD) algorithm. The output beam of the optimized 3×1 photonic lantern with 30/125 μm output fiber is stable, and the M² factor is lower than 1.18 (Fig. 15). Orbital angular momentum modes (OAM₀¹ or OAM₀² modes) and higher-order linear polarization modes (LP₁₁or LP₂₁ modes) are obtained, and the corresponding modes purities are more than 0.85, as shown in Figs. 16 and 26(a). The mode adaptive control system based on photonic lanterns achieves stable fundamental mode output with M² factor ~1.4 in large mode area fiber with a core diameter of 50 μm. By adopting photonic lanterns, the transverse mode instability is suppressed in a fiber amplifier with a core diameter of 42 μm (Figs. 22 and 23). Finally, a possible technical solution is provided for further increasing the power of near-diffraction-limit fiber lasers with large mode areas and high brightness.

The mode adaptive control system based on photonic lanterns can effectively suppress TMI in the 42 μm core fiber amplifier. The results of selective amplification of high-order mode and OAM beams achieved by this technique has a wide application prospect in the fields requiring high power special beams. Further research will be focused on the design and fabrication of the photonic lanterns with more channels and better performance, as well as increasing the modulated parameters of the adaptive control system such as polarization and intensity.