As key filtering elements, guided-mode resonance (GMR) gratings have wide applications in optical communication. However, the transmission spectrum of a conventional GMR grating shows Lorentzian-type response, which limits the further application of GMR gratings in high-performance optical communication systems. Cascaded GMR gratings have been presented to realize flat-top filtering response, but the entire device has a large size and the fabrication process is complicated. In addition, it is difficult to achieve narrow-band flat-top filtering response with only a single-layer compound GMR grating. Hence, a cascaded double-layer compound grating is proposed to solve the problem in this work. The output spectra are analyzed by a combination of rigorous coupled wave analysis and eigenmode analysis, and the results show that the central wavelength of the filter is 1549.9 nm, and the line width of the flat-top spectrum is 0.5 nm.

We propose a simplified coherent detection system, which achieved phase diversity reception based on a single balanced photodetector (BPD) by alternately detecting the in-phase (I) and quadrature (Q) information components of a signal in the time domain. The simulation results show that a 25 Gbit/s non-return-to-zero signal has a receiving sensitivity of -39.97 dBm at a bit-error-rate threshold of 10 -3 after 25 km of standard single-mode fiber transmission. The Gram-Schmidt orthogonalization process algorithm is used to compensate the mismatch angle between the I and Q components caused by the frequency offset between the signal light and the local oscillator (LO) light. When the frequency offset is less than 1/5 of the symbol rate (±5 GHz), the sensitivity penalty is about 3.2 dB. In addition, a single PD can be used to replace the BPD if the LO optical power is high, and the sensitivity penalty is about 3 dB. The scheme provides a solution for low-cost coherent detection.

A method combining a back propagation (BP) neural network algorithm with the artificial bee colony algorithm is introduced, and the design of multi-pump Raman fiber amplifier is optimized by this method. The best learning model is determined by studying the numbers of hidden layers and neural nodes in the multilayer BP neural network, which can accurately reflect the mapping relationships of the pump wavelength and pump power with the distribution of Raman net gain, and can replace the traditional method for solving the Raman coupled wave equation. At the same time, in order to improve the flatness of the gain spectrum, the artificial bee colony algorithm is used to optimize the pump parameters and the optimal pump wavelength and pump power are obtained. The simulation results show that when the trained BP neural network model is added into the artificial bee colony algorithm, the desired gain performance of the studied Raman amplifier is achieved. Moreover, the maximum error between the target value and the predicted value is less than 0.29 dB. This design scheme provides a new method and idea for the study of Raman fiber amplifiers.

Aiming at the problem of large calculation errors of weight coefficients caused by factors, such as medium refractive index changes and lens distortion in light field particle image velocimetry (PIV) technology, a method for calculating light field PIV weight coefficients based on volumetric calibration tracking technology is proposed. First, the mapping relationship between the spatial object points and the microlens is established by the volume calibration method. Second, the ray tracing technology is used to calculate the weight coefficient of the discrete voxel to pixel in the control body. Finally, the weight coefficient calculated by this method is compared with the direct tracking method, and the spatial positions of the feature points of the calibration plate are reconstructed experimentally. The experimental results show that the weight coefficient calculated by the volumetric calibration tracking method is highly consistent with the calculation results of the direct tracking method. The position errors of reconstructed calibration plate feature points in transverse and depth directions are less than 0.002 mm and 0.250 mm, respectively.

In this paper, the imaging characteristics of multilayer micropatterns are studied. Micropatterns at micron scale with high light transmittance prepared via photolithography are attached to transparent glass substrates. The patterned glass substrates are in turn stacked together to around 50 layers and imaged by an optical microscope. The final image is clear on the upper layers and fuzzy on the lower layers. Through a light field analysis, the influences of crosstalk caused by adjacent layers and illumination on the imaging results are eliminated. The influence of aberrations is examined through thickness and number variations of the glass substrates. In addition, the object field distribution of interlayer reflection, the main noise in this imaging system, is quantitatively analyzed. Finally, an optimization solution is proposed, namely, to improve the signal-to-noise ratio and thus restore the unclear patterns of the lower layers by reducing interface reflection, enhancing noise absorption, and optimizing image processing. This research can facilitate the backup, imaging, and observation of the growing number of documents and thereby effectively enhance space utilization.

X-ray fluorescence computed tomography (CT) based on an X-ray tubes is affected by many factors, which thus results in poor image quality. Geometric parameter error is one of the important factors that restrict its high-quality image reconstruction. In this paper, the influence of the relative position deviation between the two-dimensional planar phantom and the detector on the projection data is analyzed, and the correction of geometric deviation parameters is realized with the local linear relationship between the original projection data and the reprojection data of the original reconstructed image. Geant4 is employed to simulate the fan-beam X-ray fluorescence CT system with deviation parameters, and simulated projection data are used to verify the correction method. The results show that the method can accurately calculate the geometric offset to a large extent and effectively eliminate the influence of geometric parameter error on the reconstructed image. Regarding the reconstructed image after correction, the information entropy is reduced and the average gradient and standard deviation are enhanced, with the image quality thus improved.

A method for displacement field measurement of digital speckle images using a convolutional neural network (CNN) is proposed. A series of digital speckle images with their exact displacement fields in multiple deformation modes are used to construct a dataset and CNN model for distinguish displacement field of digital speckle images are proposed. Verification experiments of simulated speckle images show that the proposed method is computationally efficient and achieves high test accuracy for random deformations, axial uniform deformations, shear deformations, and other deformation modes. Moreover, the uniaxial tensile test of silica gel shows that the proposed method accurately measures the displacement field of real speckle images and confirms its high computational efficiency. The proposed deep CNN can be used to efficiently and accurately test the displacement field of digital speckle images, thereby indicating good application prospects for material deformation testing.

The straight-through calibration approach is used in ellipsometry measurement to calibrate system model parameters for accurate sample measurement. The advanced parameter model of dual-rotating compensator Mueller matrix ellipsometer (DRC-MME) is complex, with several large parameters. Some advanced parameters are cascaded and coupled together. The independent relationship between parameters in the experiment must be studied to ensure the accuracy of the results of the calibration; thus, the degree of freedom of the DRC-MME through calibration experiments was theoretically analyzed in this study using singular value decomposition, Pauli matrix, and other tools. The “useless” parameters in data fitting methods are elucidated, and a simulation experiment supports the conclusion. Finally, the experimental platform is built, the straight-through calibration experiment is conducted, and the advanced parameter calibration results are obtained which are consistent with the actual physical significance. The thickness of the silicon dioxide film on the silicon substrate was measured using the calibrated advanced parameter model. The measurement result differs from the nominal value by 1.4 nm, and the repeated measurement accuracy is 10.2 pm. In this study, the proposed degree of freedom of the system analysis approach is applied to ellipsometer and provides an analysis idea for other multiparameter optimization processes.

To achieve high-precision detection of surface shapes of off-axis elliptical cylindrical mirrors, this paper proposed hybrid interferometry integrating the aberration-free point method and the computer-generated hologram (CGH) method. In view of the special surface shapes of off-axis elliptical cylindrical mirrors, the emergent cylindrical wave of a typical cylindrical (TC) mirror, a highly integrated element composed of a plane mirror and a CGH, was used as the detection light. The optical axis was oriented to be in alignment with the line connecting the elliptical focus on the incident side and the center of the off-axis elliptical cylindrical mirror for a smaller relative aperture of the measurement optical path. Then, interferometry was implemented through the pair of aberration-free conjugate points of the ellipse. With the off-axis elliptical cylindrical mirror regarded as a spatial rigid body with six degrees of freedom on the optical axis, an error separation matrix was deduced. The wavefront aberration theory was applied to derive the parameters of the alignment error brought by the alignment amount and to determine the alignment amount in interferometry. The experimental results show that this method offers an effective measurement of the surface shape of the off-axis elliptical cylindrical mirror, and the parameters of the alignment error can be deduced via an error separation matrix, which paves the way for further analysis and correction of the system error.

In order to meet the radiometric calibration requirements of space payloads for large aperture and high radiance ultraviolet (UV) integrating spheres, a large integrating sphere radiation source with an inner diameter of 2200 mm, an exit aperture diameter of 800mm and a fully pressed PTFE coating is developed by a new technology. The thickness of the PTFE coating is 25 mm. The integrating sphere has a 10 kW halogen lamp light source and two external 14 kW xenon lamp light sources with an electric aperture,which can achieve a 4 orders of magnitude of large dynamic range adjustment. The radiance of the integrating sphere at 250 nm is 0.54 μW·cm -2·nm -1·sr -1 and in the 279--400 nm ultraviolet band, the radiance is greater than one solar constant. The expanded uncertainty of radiance calibration is 5.28%. The performance of the UV integrating sphere is also tested, and the results show that the non-uniformity of the ultraviolet radiation source is 0.65% at 380 nm. The angular non-uniformity is better than 0.7% in the horizontal and vertical directions, and 1.05% in the ±45° directions when -20°≤θ≤20°. The influences of spectral reflectance on the surface uniformity and angular uniformity are studied. The results show that with the increase of spectral reflectance, the surface non-uniformity decreases by 0.17 percentage points and the angular non-uniformity decreases by 0.15 percentage points. In addition, the stability of the radiation source, the fluorescence effect, and the thermal balance of the system are also measured. The developed UV integrating sphere radiation source has the advantages of high brightness, large aperture, good uniformity, and high reliability in the UV band (250--400 nm). It is an ideal large-area Lamber surface source and can meet the radiation calibration requirements of aerospace payloads in UV band.

The calibration of the reconstruction matrix for polarization detection affects the accuracy of the polarization state restoration. In this paper, a detailed theoretical analysis of the reconstruction matrix calibration problem of a polarization detection system based on metalens is carried out. The selection of the calibration reference state is discussed, and two criteria of condition number and equal weighted variance are used to select the calibration reference state. Through the comparison of simulation and polarization detection reduction experiments, it is proved that selecting a regular octahedron reference state for calibration can obtain a more accurate reconstruction matrix and have better anti-noise ability.

A solar cell based on bottom-reflectivity-enhanced GaAs radial p-i-n junction nanowire array is proposed, and its spectral absorption and photovoltaic performance are studied by finite-different time-domain method and finite element method. The results show that replacing SiO2 between the polymer and the substrate with MgF2 dielectric layer with low refractive index not only significantly reduces the absorption loss of the substrate, but also significantly improves the optical absorptance of the nanowire array in the whole wavelength range. Moreover, the photoelectric conversion efficiency of the solar cell can be improved to 13.9% by optimizing the thickness of i region and the length of nanowire. This study provides a feasible way to realize low-cost and high-performance nano solar cells.

High frame rate infrared focal plane array detectors have important applications in infrared weapon systems, spectral imaging, and high-speed temperature measurement. Currently, in our country, lag in high frame rate infrared focal plane array technology limits the development of high performance infrared weapon, spectral imaging, and high-speed temperature measurement technologies. Targeting at high frame rate infrared imaging applications, a 384×288 array, 25-μm pitch digital readout integrated circuit is designed and fabricated, and then, hybridized to a long-wave HgCdTe detector array. A high frame rate 384×288 long-wave digital infrared detector assembly is implemented by assembling the hybrid focal plane array in a metal vacuum dewar and coupling with a Stirling cryocooler. Measurement results demonstrate that the maximum frame rate is 1012 Hz, noise equivalent temperature difference (NETD) is 16.8 mK, and dynamic range reaches 95.2 dB. An ignition of a lighter is successfully captured by the developed detector. In addition, the details of flame growth and spurting are clearly presented, and a good imaging effect is obtained.

The lateral growth of single-crystal diamond on a semi-open substrate holder was realized in this paper by microwave plasma chemical vapor deposition (MPCVD). The protrusion height (Δh) of the seed crystal above the substrate holder was adjusted to regulate the radicals distribution. The influence of protrusion height on the epitaxial lateral growth of single-crystal diamond was analyzed by combining optical emission spectrum with the Fourier transform infrared spectrum, Raman spectrum, white light interferometry results and optical morphology characterization outcomes. The analysis show that the relative concentration of C2 (516.08 nm) radicals in the central region (from -2 mm to 2 mm) of the plasma increased with increasing protrusion height. When the protrusion height was increased to 0.6 mm, the concentration of carbon-related radicals in the central region was relatively high, resulting in a vertical growth rate slightly faster than that in the surrounding area. The growth rate difference is favorable for the growth plane to automatically develop structures in an off-axis direction from the (100) crystal plane. The result of lateral growth of these structures is the formation of single-crystal diamonds without the polycrystalline diamond rim. These diamonds also have excellent infrared transmittance. The automatic development of off-axis structures on the top growth plane was the key to epitaxial lateral growth of single-crystal diamonds. However, the continued increase of protrusion height to 0.8 mm led to an unnecessarily high relative concentration of C2 (516.08 nm) radicals in the central region and formation of pyramidal hillocks, which is the disadvantage of the epitaxial growth of high-quality single-crystal diamonds.

Improving the nonlinear optical (NLO) response of two-dimensional nanomaterials and motivating the NLO applications in optoelectronics are still highly challenging at present. One of the solutions is to construct van der Waals heterojunctions with high quality and clarify the internal mechanism in NLO response. In this study, rhenium disulfide (ReS2)/graphene heterojunction films are successfully fabricated by liquid exfoliation combining vacuum filtration, with different mixing approaches adopted, including powder mixing, mixing after ultrasonic treatment, and mixing after centrifugation. A Raman spectrometer, a near-infrared spectrometer, and an atomic force microscope are utilized to characterize the optical properties of the heterojunctions. The NLO properties of these films are investigated by a self-developed open-aperture Z-scan system with a high repetition frequency and a narrow pulse width which is equipped with an 800 nm femtosecond laser. The imaginary part of the third-order nonlinear polarizability and figure of merit (FOM) of these films are obtained. The results show that the NLO response of ReS2/graphene films is significantly enhanced owing to the successful heterojunction construction. The FOM of ReS2/graphene films prepared by mixing after ultrasonic treatment is 3.5 times that of ReS2 film and 1.6 times that of graphene. The ReS2/graphene films in the case of mixing after ultrasonic treatment show the strongest reverse saturable absorption characteristics, and their FOM is 44% and 27% higher than that of films prepared by powder mixing and mixing after centrifugation, respectively. This paper provides a new idea for the fabrication of van der Waals heterojunctions and lays a theoretical foundation for the research of femtosecond optical limiters.

An updated light sheet fluorescence microscope (LSFM), combining the digital scanning light sheet fluorescence microscope (DSLM) and the line-scanning imaging (LSI) method, is designed to improve the imaging quality by decreasing sample scattering. Based on DSLM, a scanning galvanometer for descanning is added in the detection light path. Moreover, the scanning galvanometer in the illumination light path is synchronized with the descanning galvanometer to fix the evenly moving images at the same position in front of the camera during the control process, thereby realizing LSI. The linear array detector is simulated by the area array camera in the system for conveniently comparing with the traditional method. In addition, compared with the conventional LSFM, the improved system is more effective in suppressing the sample scattering problem in the imaging experiments of high-scattering fluorescent microsphere samples and zebrafish heart samples. Hence, the feasibility of LSFM based on LSI method can be verified.

As the most widely used wavefront sensors in adaptive optics, Shack-Hartmann sensors can measure not only the distortions caused by atmospheric turbulence, but also the aberrations brought by mirror position errors, which are usually introduced by wind, temperature change, and mechanical stress. In this paper, we established the functional relation between the subaperture slope and misalignment based on the Shack-Hartmann sensor, and then developed a computer-aided alignment method based on the centroid deviation of lattice spots before and after optical system misalignment. This method converts the misalignment calculation into multi-objective optimization, which can be solved by multi-objective intelligence optimization algorithms. Using a three-mirror optical system as an example, the co-simulation with Python and Zemax was conducted for alignment. The simulation results show that the misalignments are reduced to the micron level after three iterations, which can meet the practical alignment requirements. The simulation results demonstrated the correctness of the proposed method.

A reflective metasurface filter is proposed in this paper, which consists of a substrate, a nanograting with low refractive index, a Ag grating and a high-refractive-index dielectric SiNx layer. The simulation results show that with the incident angle of 45°, the reflection peak appears at the wavelength of 475 nm for the plane perpendicular to the grating line, while it is located at 550 nm for the plane parallel to the grating line. The electric field distribution characteristics demonstrate that for different incident planes, transverse electric field-polarized incidence excites guided-mode resonance in different areas, and the intensity of guided-mode resonance is different from that of surface plasmon resonance in the case of transverse magnetic field-polarized incidence. As a result, different colors are present in the two planes perpendicular to and parallel to the grating line, respectively. Accordingly, the fabricated metasurface sample has significantly optical variable color, which can be flexibly realized for large-area structures by nanoimprint lithography. The proposed reflective metasurface filter has a broad application prospect in the fields of anti-counterfeiting and information encoding.

A tunable broadband bandpass filter is proposed, which is based on vanadium dioxide (VO2) metamaterial. The simulation results show that the 3 dB bandwidth of this filter is 1.71 THz with the central frequency of 5.19 THz and the largest transmission rate of 0.77. Furthermore, this filter can maintain stable broadband transmission performances within the incidence angles of 0°~40°. The physical mechanism of this kind of filter performance is investigated based on the equivalent circuit of this filter and its surface current distributions at the resonance frequencies. Owing to the unique phase transition property of VO2, the bandwidth of the filter can accordingly change from 1.71 THz to 2.31 THz by tuning the conductivity of VO2. The proposed metamaterial filter possesses the advantages of structural simplicity, broadband performance, and tunable property, which makes it possible to find promising applications in THz communications, sensing and other emerging areas.

For better accuracy of mixture component identification and less complexity of modeling, a progressive detection system based on surface-enhanced Raman spectroscopy with rough classification followed by subdivision is constructed in this paper. First, a characteristic peak discrimination algorithm is used to extract features from all samples and establish a rough classification model, according to which substances are classified into single-component and multi-component ones. Then, automatic extraction of spectral characteristics is accomplished through normalization and principal component analysis. A subdivision model is developed upon a multi-output least squares support vector machine. Finally, the particle swarm optimization algorithm is employed to optimize the parameters so as to achieve an accurate prediction of the composition of multi-component samples. Experiments are carried out with Rhodamine 6G, Nile blue and crystal violet as probe molecules. The results show that the characteristic peak discrimination algorithm extracts sample features with an accuracy of 99.44%. The rough classification model correctly identifies all 90 samples in the blind test. Moreover, the correlation coefficient of the subdivision model for multi-component sample identification is not less than 0.995 and the root mean square error is not more than 2.67343%. Enabling both qualitative and quantitative detection of samples, the Raman detection system proposed in this paper provides an effective identification method for future detection of complex substances such as drugs.

The effect of diffractive optical elements (DOE) modulating the incident light field on the focusing field is studied. The 4π focusing properties of Bessel-Gaussian beams radially polarized after modulated by several different DOEs are analyzed via the Richards-Wolf vector diffraction theory. The numerical simulation results show that multiple spherical spots of potential applications form near the focal plane, and the number of spherical spots is related to that of the belts on DOE, while the DOE transmission function has little influence upon it. Moreover, the number of spherical spots increases with that of the belts on the DOE.

Cirrus clouds have a great impact on optical signal transmission in quantum communication. According to the cirrus cloud characteristics, the cirrus cloud ice particle distribution model is established, and the extinction properties of cirrus clouds are investigated. In addition, according to the scattering features, the quantitative relationships among the cirrus cloud ice water content, transmission distance, link extinction characteristics, channel capacity, and channel entanglement attenuation are analyzed. The change of teleportation fidelity under the influence of noise is studied and the simulation experiment is conducted. The experimental results show that when the transmission distance is 3 km and the cirrus cloud ice water content (volume mass)is 0.4 g·m -3, the corresponding link attenuation, channel capacity and channel entanglement are 0.05 dB·km -1, 0.94 dB·km -1 and 0.52, respectively. In contrast, they are 0.20 dB·km -1, 0.82 dB·km -1 and 0.07, respectively, when the cirrus cloud ice water content (volume mass) is 0.1 g·m -3. From these experimental results, one can see that the performance of the free-space quantum communication channel changes to different degrees under the influence of cirrus clouds. Therefore, the influence of cirrus clouds should be considered in quantum communication, and various performance parameters should be adjusted according to the change of cirrus clouds in order to ensure the communication quality.

Because Bell state Measurement is the prerequisite for quantum information processing, a non-destructive measurement scheme for Bell state is proposed. In this scheme, the parity gates are introduced as key components, and C-NOT gates and Toffoli gates are implemented using the property that parity gates do not change the states of the input photons. Then, Toffoli gates combined with Hadamard gates are used to measure the 4 Bell states deterministically. The proposed scheme can be used to measure two-photon entangled states and extended to the application scenarios of three-photon entangled state, providing a new idea for Bell state measurement and quantum information processing.

Lidar echo scene generation, based on three-dimensional scene construction and laser scattering principle, can be used to simulate the laser echo characteristics of real scenes and is an important means for testing the performance of laser array detector. To efficiently generate dynamic scene with high simulation accuracy, an adoption degree composed of the simulation similarity of laser echo scene and scene calculation time is constructed and used as the evaluation function. The spatial mesh number of laser echo scenes is thereby optimized. For an area array detector with 128×128 elements, the proposed method is used to optimize the number of spatial meshes of laser echos in two experimental scenes with different materials and distances. When subdivision number of the detector single pixel is 3×3 and the total number of spatial meshes is 384×384×L (L is subdivision number determined by distance accuracy), the adoption degree reaches the maximum values. The average generation time of each frame scene is less than 8 ms, and the dynamic scene generation with high frame rate (125 Hz) and high fidelity is realized. The theoretical and experimental results show that this method can balance the similarity of the generated echo scene and computing resource consumption when the computing resources are limited, and it is a practical optimization method for three-dimensional laser echo scene generation.

Retrieval of small-scale three-dimensional (3D) wind field based on a single radar is an important issue in various fields, including aviation meteorology and wind resource assessment. To obtain the 3D structure of small-scale wind field with high efficiency, this study proposes a variational retrieval method based on lidar detection. The proposed method comprises local and global variational methods. First, the local variational method is constructed locally in an analysis volume to preliminarily retrieve the 3D wind field, as the initial value. Second, the initial value is tuned using the global variational method in the entire lidar scanning volume. Simulation results show that the proposed method outperforms the traditional two-step variational method. The root mean square error of the algorithm reduces by 1.49 m/s and 1.19 m/s in the case of homogeneous and inhomogeneous wind fields, respectively. Simulation and field detection results both prove that the proposed algorithm can obtain accurate 3D wind field retrieval results with high efficiency when the lidar scanning volume covers a large range of the elevation angle. The proposed method can be applied in aviation safety, wind resource assessment and other fields.

Quartz-enhanced photoacoustic spectroscopy (QEPAS) is a gas detection technology that has developed rapidly in recent years. It has the advantages of high sensitivity, small equipment size, and immunity to environmental noise. We design an all-solid-state mid-infrared fiber-coupled QEPAS photoacoustic detection module. Under theories of gas thermodynamics and one-dimensional acoustic resonator, the sound pressure distribution and level of the detection module are simulated using COMSOL software. An optomechatronic detection module is then designed and processed. It integrates the acoustic resonator, photoacoustic cell, optical fiber module, and pre-amplification module for easy collimation, high stability, and strong anti-interference ability. With a high-power mid-infrared DFB laser at a central wavelength of 2 μm as the light source, CO2 detection is carried out via wavelength modulation technology. As a result, a detection limit of 3.7×10 -5 is obtained at an integration time of 1 s. Allan variance analysis shows that when the integration time is 1123 s, the detection limit of the system can reach 1.34×10 -6. The QEPAS system built on this module can be used to conduct real-time monitoring of indoor CO2 concentration.

The near-infrared laser with a wavelength of 1.064 μm is one of the main laser sources for laser ranging, free-space optical communication, and space optical remote sensing. Narrowband filters are one of the key components that suppress background light interference. At present, the full widths at half maximum of most filters are several nanometers. An ultra-narrow band-pass filter was designed and fabricated, which had a full width at half maximum of 0.19 nm, a peak transmittance of 70.2%, and a central wavelength of (1064±0.05) nm. Then it was annealed at 100 ℃, 200 ℃, and 300 ℃, respectively, and changes in its surface morphology and spectral characteristics were investigated. The experimental results indicate that the surface of the filter is smooth and annealing has little influence on it. In contrast, the transmission spectrum of the filter undergoes redshift as the annealing temperature rises. When annealed at 100 ℃ for 3 h, the filter experiences a spectral shift of 0.03 nm, which shows that the filter is suitable for space optical systems with limited temperature control.

Hf∶ZnO (HZO) films were prepared on sapphire substrate by radio frequency magnetron sputtering. The effects of oxygen flow rate and annealing temperature on the microstructure and photoelectric properties of the films were studied. For better crystallinity, the prepared films were annealed at 800 ℃ for 30 min and then naturally cooled to room temperature. The microstructure and photoelectric properties of annealed films were tested and analyzed. The results showed that as the oxygen flow rate increased from 0 to 0.6 mL/min, the microstructure became denser and the resistivity gradually decreased. However, when the oxygen flow rate increased to 0.8 mL/min, the crystallinity deteriorated and the resistivity suddenly increased. The HZO film prepared at an oxygen flow rate of 0.6 mL/min obtained the best photoelectric properties. The results of the Fourier transform infrared (FTIR) spectrometer manifested that the average transmittance of this film in the 3--5 μm waveband was 83.87% and the Hall effect test results demonstrated that its resistivity, carrier mobility, and carrier concentration were 1.66×10 -2 Ω·cm, 13.4 cm 2·V -1·s -1, and 2.82×10 19 cm -3, respectively. Meanwhile, the XRD results showed that the annealed films presented ZnO hexagonal wurtzite structure and grew in the (002) direction. The SEM results revealed the dense and uniform spherical particle structure on the surface of the films. In a nutshell, HZO films prepared at an oxygen flow rate of 0.6 mL/min and an annealing temperature of 800 ℃ can be used as the transparent window material for the 3--5 μm waveband.

For the purpose of achieving high-quality athermal imaging in the infrared detection system, an antireflection coating in the wave band of 3.7-4.8 μm was prepared on aspheric chalcogenide glass substrate. Adhesive layer material was selected in the experiment to improve the adhesion between the substrate and the coating. The finite element method and multi-physics simulation software were used to build a three-dimensional model that combined temperature field and thermal stress field. The stress distribution of aspheric film was analyzed. In view of the simulation results, the deposition process was optimized. The thermal stress of the chalcogenide glass substrate was reduced by temperature gradient heating and the stress of the deposited film was released by in-situ vacuum annealing so as to solve the film stripping problem of the aspheric mirror. The prepared film passed the tests of adhesion, humidity, and moderate friction of the MIL-C-48497A standard and the average transmittance in the wave band of 3.7-4.8 μm was 99.12%, which means the prepared film met the index requirements of infrared detection system.