Experiments are conducted to systematically study the single ionization and non-sequential double ionization (NSDI) of rare gas atoms (Xe and Ar) and diatomic molecules (O2 and N2 with different symmetry of molecular structure) in near-infrared intense femtosecond fields with various wavelengths of 800 nm, 1250 nm, and 1500 nm. It is found that,compared with Xe, Ar, and N2, O2 has a higher yield of NSDI as the slope of the flat region in the changing curve of light intensity increases. The wavelength scaling of the yield of NSDI in the flat region for all gases deviates from the prediction of the semiclassical "three-step" model of ionization. Furthermore, we find that the yield of NSDI of O2 is significantly suppressed compared with that of its companion atom Xe in all wavelengths, and the suppression becomes more pronounced as the wavelength increases. For the yield of NSDI of N2, however, it is almost the same as that of its companion atom Ar at a short wavelength of 800 nm, but it is also suppressed at long wavelengths. These experimental findings can be used to improve the wavelength scaling of molecular NSDI and identify the influence of molecular structure on NSDI of molecules, which indicates that the ultrafast control of molecular NSDI involving multi-electron correlation effects can be achieved by macroscopically tuning the wavelength.

Bulk heterojunction (BHJ) is one of the most meaningful structures of organic photodetectors (OPDs). It can be achieved by mixing donor and acceptor in a solvent with the aid of a sol-gel method. Furthermore, BHJ has been proven to be a very effective method to adjust the diffusion length of organic semiconductor excitons. The preparation of OPDs with this structure is a very important work. The conjugate polymer donor poly(3-hexylthiophene) (P3HT) and fullerene derivative acceptor [6,6]-benzene C61-butyric acid methyl ester (PC61BM) are used to form the active layer to prepare a binary BHJ OPD, so as to improve the performance of MgZnO ultraviolet photodetector. The results show that the responsivity of the OPD can reach 0.721 A/W, which is about 3.9 times higher than that of MgZnO ultraviolet photodetector. In addition, the OPD exhibits excellent detection ability, better responsivity and high light to dark current ratio.

The serious chromatic dispersion of traditional diffractive optical elements (DOEs) limits their application in wide-band imaging. In this paper, a design of achromatic diffractive lens is proposed. After coding the microstructure height on the whole diffractive lens, the microstructure height is optimized by the particle swarm optimization algorithm to balance the focusing effect of each wavelength in the visible spectrum at a specific focal length, so that the point spread function (PSF) of each wavelength in the visible spectrum becomes almost the same, achieving the goal of achromatism in the continuous visible light band from 400 nm to 700 nm. The simulation and imaging test of traditional DOE, zone achromatic DOE, and coding optimized achromatic DOE proposed in this paper are carried out. Compared with the other two DOEs, the PSF of coding optimized DOE at each wavelength is more consistent, and the cut-off frequency under visible light is increased to 8.6 lp/mm. Therefore, compared with traditional DOE and zone achromatic DOE, the coding optimized DOE proposed in this paper has better achromatic ability and higher resolution in visible light.

This paper proposes a fast pixelated mask optimization method for extreme ultraviolet lithography. An improved fast pixelated thick mask model is utilized in mask optimization. The point pulses on the edge pixels are set according to the mask pixel size. On the basis of the dual edge evolution strategy, the optimization variables are adaptively initialized in each epoch of optimization according to the difference between the current resist pattern contour and the target pattern contour. The optimization efficiency is improved by generating the initial individuals and population using priori information. One-dimensional line-space patterns and two-dimensional complex patterns are used for simulation. Simulation results show that the imaging simulation accuracy is effectively improved. In addition, the optimization efficiency of the two complex patterns is effectively increased by the proposed method.

Indoor light-emitting diode (LED) light sources are with non-flat light intensity distribution, and the conventional Lambert model fails to take indirect channels, noise and interference in the environment, obstruction, interior borders, and irregular room layouts into account. To address these problems, this paper proposes an optimal light source layout scheme based on a parallel fully connected convolutional neural network (PFCNN) model for indoor visible light positioning (VLP). The datasets in the fingerprint database are constructed by collecting light source information, such as the coordinates, power, and orientation angle of the light source, and the corresponding light intensity distribution on the receiving plane. The parameter characterizing the flatness of light intensity distribution is measured by a Monte Carlo method, and a fully connected neural network and a parallel fully connected neural network are utilized to build a visible light channel model. A prediction model for light intensity flatness is then developed by the proposed PFCNN model, and an optimal light source layout is achieved by the momentum particle swarm optimization K-Means++ (Mot-PSO-K-Means++) algorithm. Simulation analysis shows that parallel fully connected neural networks improve accuracy by 84.69% compared with that of fully connected neural networks. In the 5 m×5 m×3 m indoor space, light intensity flatness reaches 92.00% under the 4-LED layout, and light intensity ranges from 340 lx to 440 lx. Those under the 12-LED layout are, respectively, 92.00% and 980-1120 lx. Therefore, the proposed scheme, with higher flatness and applicability, can be applied to actual indoor VLP scenes, and it can provide theoretical support for in-depth research of indoor VLP.

An automatic synchronous capture scheme is designed for a fast and large-range frequency hopping receiver based on integrated signal processing and digitization of microwave photons. The designed synchronization scheme is based on direct power detection and serial search strategy. A field-programmable gate array is used in the experiments to control and adjust the timing of locally received frequency-hopping patterns. The automatic synchronous capture receiving for frequency hopping signals with a switching time of ~4 ms and a hopping frequency up to 33 GHz is realized in the experiments.

Image plane digital holographic detection is an important application of coherent light imaging. However, due to the spatial filtering of coherent transfer function, charge coupled device (CCD) cannot detect the complete image light field when the exit pupil size of the imaging system is small. Moreover, using the current popular theory that the amplitude distribution of the image light field can only be calculated, the image light field with amplitude and phase distribution which is convenient for optical detection can not be reconstructed.A method of designing specific illumination light is proposed based on the analysis of coherent light imaging theory. Using the proposed method, the object plane detection area can be complete imaging by optical system. By combining the hologram of object illumination light with the proposed method, the ideal image light field which is convenient for practical detection can be reconstructed.

Photonic integrated interferometric imaging is an emerging imaging technology. It uses photonic integrated circuits to obtain the Fourier spectrum of the target. At present, energy loss and noise interference are the key factors restricting the development of this technology. To analyze their impact on photonic integrated interferometric imaging systems, this paper studies the relationship between the baseline length and the interference signal energy and the effect of noise on imaging quality. The results show that the energy of the output signal is at the nanowatt level. The amplitude and visibility of the interference signal decay sharply as the baseline length increases. The amplitude of the interference signal under the long baseline is extremely small. In the case of insufficient detector accuracy, the visibility of the interference signal cannot be obtained. Overall, the photonic integrated interferometric imaging systems are suitable for collecting low-frequency signals of the target. In addition, the noise controlled at 10-3 of the interference signal strength and the phase error controlled at 1/40λ will not greatly affect the imaging result.

Due to the unevenness of the corrosion layer and effect of the coatings, it is difficult to accurately detect the corrosion thickness of steel plates by the existing nondestructive testing (NDT) approaches. Based on the properties of the high transmission to non-polar materials and reflection on the surface of polar metal materials of terahertz (THz) wave, a nondestructive terahertz time-domain spectroscopy (THz-TDS) is utilized to detect the corrosion thickness under coatings in this paper. The experimental results show that in the effective frequency range of 0.2-1.2 THz, the refractive index of corrosion products, epoxy resin, rubber and cement paste is 2.80, 1.94, 2.18 and 2.04, respectively. The delay time difference between the sample and the reference signals in transmission THz-TDS exhibits a linear relationship with the refractive index. The reflected THz signal can identify the corrosion layer on steel plate surface, and accurately measure the corrosion layer with thickness greater than 40 μm. The corrosion of steel plates under coatings can be judged by the number and attenuation of signal amplitude in reflection THz-TDS. The thicknesses of the coatings and corrosion layer are determined with accuracy greater than 90% by the delay time difference corresponding to the signal amplitudes. The application and accuracy of THz-TDS for NDT corrosion thickness detection of steel plates under coatings are therefore proved.

Vector network analyzer is an important measurement instrument in the fields of radio and microwave. In this paper, we design and implement a microwave photonic vector network analyzer (MPVNA) based on optical sampling, which uses ultrastable optical pulse trains, generated by mode-locked laser, to directly sample single tone signals via a Mach-Zehnder modulator, and then processes the sampling data with digital signal process to obtain the scattering parameters of the device under test. The experimental result shows that MPVNA bandwidth is 20 GHz when using an electro-optic modulator with bandwidth of 20 GHz in the system and a larger system bandwidth can be achieved by using an electro-optic modulator with larger bandwidth. The system dynamic range is about 60 dB and the minimum system frequency resolution is 11.92 Hz. Then we use this system to test the scattering (S) parameters of a bandpass filter with a center frequency of 10 GHz and compare the test result with that of a commercial vector network analyzer. The comparison results show that the average amplitude deviation of S21 is 0.1241 dB and the average phase deviation of S21 is 3.6356° in the passband of the bandpass filter, which proves that the test results of the system is well consistent with that of a commercial vector network analyzer.

The circular sinusoidal fringe has a constant phase (encoded as zero phase) at the circular center, and its center can be used as a reference point for phase unwrapping to obtain absolute phase for calculating the height information of the measured object. As the carrier phase of the circular fringe is a non-linear function, the existing Fourier transform method based on circular fringe projection needs to solve a quadratic equation, judge roots, and fit to obtain the pixel displacement corresponding to height information of object surface. The robustness of the algorithm is not good in fact. In this paper, an improved circular fringe projection profilometry based on Fourier transform is proposed and investigated. By projecting another fringe with horizontal movement, the calculation of pixel displacement in circular fringe profilometry is simplified, and the calculation of pixel displacement is changed from solving a first-order equation to solving a quadratic equation. The robustness and surface reconstruction accuracy of circular fringe profilometry based on Fourier transform are improved. Computer simulation and experiments validate the effectiveness of proposed method in 3D measurement, which is especially suitable for out-of-plane measurement of the whole plane.

To reduce the effect of manufacturing error of mold inserts on the optical performance of molding diffractive optical elements (DOEs), single point diamond turning process of microcrystalline aluminum alloy mold inserts with a curved substrate diffractive microstructure surface is studied. Based on the shadowing effect and scattering effect, the influence of mold insert manufacturing error on the diffraction efficiency of DOEs is analyzed, a mathematical relationship model among the tool radius, tool deflection angle, and diffraction efficiency is established, and an optimization method for improving machining accuracy is proposed. The turning experiment of diffractive optical mold inserts is carried out under the guidance of the method, and the expected diffraction efficiency is compared with the experimental results. The results show that using process parameters obtained by the proposed method can reduce the influence of mold insert manufacturing error on diffraction efficiency and improve the optical performance of DOEs, which provides a reference for SPDT diffraction optical mold insert manufacturing.

In order to solve the problem that the accuracies of dual-view three-dimensional (3D) video extensometer are different in two directions, and to avoid the problems of camera synchronization and high cost caused by using multiple cameras, a biaxial three-dimensional video extensometer based on single-camera four-view imaging is proposed. A rectangular pyramid prism is installed in front of a single camera, and the area to be measured can be observed from four views by a single camera. The 3D reconstruction of the measured points before and after deformation is carried out by using the strong constraint equation of multiple views, and the linear strain between the measured points is calculated. In order to avoid the influence of prism refraction on imaging, narrow-band monochromatic light is used for illumination. The measuring accuracy of the proposed biaxial video extensometer is verified by stainless steel tensile experiment. The experimental results show that the 3D biaxial video extensometer can accurately measure the linear strains in X and Y directions with high precision. The absolute error of strain is within 30 με after the average de-noising of linear strains in two directions. Comparing the measured mechanical parameters with the measured results of strain gauge, the relative error of elastic modulus and Poisson ratio is 0.56% and 1.8%, respectively.

A hybrid neural network model based on 3D-2D convolution tandem is proposed as the spatial feature extractor to overcome the problem of low accuracy of conventional iteration reconstruction algorithm in the case of limited optical windows and projection views in practical flame reconstruction. In this model, 3D convolution is utilized to extract spatial features from multi-view projections simultaneously, and 2D convolution is used to further accelerate the training speed and reduce computational consumption. Compared with conventional iteration reconstruction algorithm and reconstruction algorithms based on residual networks, the proposed model has the advantages of high reconstruction accuracy and low time consumption. It shows potential in flame on-line monitoring and rapid reconstruction.

In this paper, nitrogen-vacancy centers in diamonds are created by high-energy electron irradiation and vacuum annealing. The macroscopic color changes and internal defect transformations in those diamonds before and after irradiation and annealing are investigated. The effects of irradiation dose and vacuum annealing temperature on the yield of nitrogen vacancies are discussed. The results show that the concentration of nitrogen-vacancy centers rises and then drops as irradiation dose increases. This is because under a higher irradiation dose, vacancies are more likely to form vacancy clusters, which do not constitute nitrogen-vacancy centers during the annealing process. The concentration of nitrogen-vacancy centers increases with the increasing annealing temperature and then saturates in the temperature range of 800-900 ℃. As the temperature further increases, the concentration declines due to the interaction of nitrogen vacancies with interstitial atoms generated during the irradiation process under a high temperature.

On the basis of the variable-coefficient fractional Schr?dinger equation with potential barriers, we study the influence of chirp parameters and potential barrier functions on propagation dynamics of Gaussian beams by the combination of the numerical simulation and analytical method. The results reveal that the linear chirp will weaken the intensity of a sub-beam after splitting, while the quadratic chirp will change the respiratory amplitude and width of beams, and thus the beams no longer oscillate strictly according to the sinusoidal law, which leads to a large and a small respiratory trajectories eventually. After the introduction of the barrier effect, the beams at the barriers deform due to reflection and transmission. The reflected and transmitted beams are still modulated by the diffraction coefficient and are superimposed at the barriers before transmitting again. Under an appropriate barrier depth, the beams are completely reflected when they encounter the barriers, and all the energy of the beams is trapped between the two barriers. Thus, in the fractional system, the regulation and management of beam transmission can be achieved through potential barriers and chirp parameters.

The laser beam expander, which can expand the beam diameter and compress the divergence angle, is widely used in laser ranging, atmospheric detection, and other fields. In order to compress the external dimensions of the laser beam expander and reduce the distance between the output beam and the input beam, a multi-stage reflection laser beam expander composed of multiple confocal off-axis parabolic mirrors is proposed in the paper. According to the design requirements, the system parameters of the multi-stage reflection laser beam expanders are calculated. A single-group beam expander and two multi-stage beam expanders are simulated, and their tolerances are analyzed by ZEMAX optical design software. The dimensions and tolerances of the single-group beam expander and the multi-stage beam expanders are compared. The results show that the multi-stage beam expander can achieve high-magnification laser beam expansion in a shorter distance within a relatively loose tolerance range.

The advantages of small aberration and high transmittance exist in Dyson prism imaging spectrometers studied at home and abroad. The purpose of this paper is to improve the dispersion uniformity of the Dyson prism imaging spectrometer. By analyzing the dispersion law of prisms made of different materials, this paper chooses fused silica and CaF2 materials to form achromatic prism group, and uses super-ring reflecting surface and lens group to correct other aberrations. Through optimization, a Dyson prism imaging spectrometer with a slit length of 80 mm and uniform dispersion is obtained. At the same time, the structure of double-objective lens and the means of splicing slits are chosen to meet the overall requirement of 200 km swath. The results show that the resolution is 10 nm in 400-2500 nm band and the slit length is 80 mm. The design value of the optical system is under the 25 lp/mm cut-off frequency, and the modulation transfer function is larger than 0.6. According to the spectral position of the image surface after dispersion of each wavelength, we obtian that in the range of 400-1200 nm, the fitting accuracy between the dispersion curve and the fitting line increases from 0.88 in the case of a double material prism to 0.94 in the case of a single material prism, which greatly improves the dispersion uniformity.

The optical wedge is a prism with a small apex angle, which is often used to control the deflection angle of the light beam or compensate for the deviation of the small angle of the light beam. A driving method for a kind of liquid crystal device driven by four electrodes with rectangular aperture to realize an optical wedge is proposed. The advantage of the liquid crystal wedge lies in its small size and light weight. The wedge angle can be controlled by voltage and the optical wedge can be rotated without a mechanical device. It is expected to be used in some fields where the deflection direction and angle of the beam need to be controlled, such as optical tweezers, lidar and so on.

Erbium-doped lithium niobate on insulator (Er∶LNOI) has attracted much attention because of its high gain. The Er∶LNOI waveguide amplifier is modeled and the corresponding energy-level rate equation is established. The gain performance of Er∶LNOI waveguide is simulated by using the model and compared with experiments. The internal net gains of the amplifier at the signal wavelength of 1531.5 nm and 1550.0 nm are studied at 980 nm and 1484 nm pump wavelengths. The waveguide gains with different signal powers and pump powers are compared by experiment and simulation. In addition, the effect of waveguide length on amplifier gain is also investigated.

A widely tuned and potentially integrated optoelectronic oscillator (OEO) is constructed based on direct modulation optical signal injection semiconductor laser. The tuning range of 9.73-21.30 GHz and the phase noise lower than -105 dBc/Hz@10 kHZ are realized. The structure of the experiment is simple and compact. If the master laser and the slave laser are integrated on the same chip, the system can be further miniaturized. If the master laser with larger modulation bandwidth is used, the tuning range of OEO can be further expanded.

In this paper, a dual-core mode converter based on directional couplers is presented, which can achieve the conversion from the fundamental-order mode (LP01 mode) to the high-order mode (LP11 mode). The theoretical analysis model of the dual-core mode converter is built by the finite element and beam propagation methods to explore the influence of optical fiber parameters on the conversion performance. Numerical simulations reveal that the bandwidth of this mode converter can reach 320 nm when the extinction ratio is greater than 20 dB, and its conversion rate is 99.7% when the bandwidth reaches 400 nm, and the extinction ratio is greater than 10 dB. Compared with the traditional mode converter, the proposed converter features a simple structure, a high conversion rate, and wide bandwidth. Therefore, it is suitable for applications in optical communication systems.

In this paper, a graphene-coated hybrid dielectric nano-parallel wire waveguide is designed. This waveguide consists of two cylindrical and one elliptical cylindrical dielectric nano-parallel wires. The five lowest-order modes are classified, the effects of working wavelength, Fermi energy of graphene, and structural parameters on the real part of effective refractive index, propagation length, and quality factor of the five lowest-order modes are investigated by the finite element method. The results show that the five lowest-order modes can be synthesized from the low-order modes of cylindrical nanowires and elliptical cylindrical nanowires. The transmission characteristics of the five modes can be effectively adjusted when the operating wavelength and Fermi energy of the graphene are adjusted. When the structural parameters are changed, the transmission characteristics of the first two modes change significantly, while the transmission characteristics of the other three modes do not change significantly. Compared with a waveguide composed of two elliptical cylinders and a cylindrical dielectric nano-parallel wire, the waveguide structure designed in this paper has a longer propagation length and a higher figure of merit. The theoretical research in this work is helpful to the design, fabrication, and application of graphene coated hybrid dielectric nano-parallel wire waveguides.

A method that can generate ultra-wide and flat noise in the whole millimeter-wave band is proposed. Based the vernier effect, the frequency intervals and incoherent light widths of the main ruler and the subordinate ruler are designed by using two incoherent optical frequency combs with different frequency intervals. Simulation results show that the proposed method can generate ultra-wide band and flat full-band millimeter-wave white noise with a frequency range of 30-300 GHz. Experimentally, the amplified spontaneous emission noise is filtered out of two comb-shaped lights with a vernier frequency interval through a programmable filter, and two millimeter-wave white noises are achieved based on uni-traveling-carrier photodetector with two different band widths. The frequency ranges of the two white noises are 130-170 GHz and 280-380 GHz, and the corresponding flatness are ±2.25 dB and ±3.10 dB, which verifies the correctness of the proposed theory.

The vortex half-wave plate (VHP) is a polarized optical element whose fast axis orientation changes in a specific angular direction in the spatial distribution. By cascading two or more low-order VHPs, any high-order cylindrical vector beam (CVB) can be generated. Based on the Jones matrix of a single VHP, the equivalent Jones matrix of the VHP cascaded is derived, and the change of the equivalent order after the VHP cascaded is theoretically explained. In the experiment, the two VHPs of m=-1 and m=2, are used to cascade to generate the one-order and three-order radial polarization and angular polarization CVBs, to calculate the Stokes parameters and draw the vector beam polarization distribution diagram, and to be compared with one-order CVBs that directly generated by the VHP of m=1, which proves the reliability of the cascade method. When the cascade produces the same high-order CVB, the different cascading sequence of the VHPs can affect the degree of beam distortion caused by the center misalignment. After experimental comparison and analysis, it is obtained that high-quality and high-order CVBs can be stably produced in the cascade application. By cascading multiple VHPs, the CVBs within 100 orders are generated.

Based on Richards-Wolf vector diffraction integral theory, we investigate the focusing characteristics of Laguerre-Gaussian-distributed cylindrical vector (CV) vortex beams coaxially incident on a 4Pi focusing system. The focusing system is composed of two identical high-numerical-aperture lenses, with diffractive optical elements (DOEs) inserted. The numerical simulation results show that the phase modulation of DOE makes the focal field exhibit a spherical structure, and more optical spheres can be obtained by increasing the number of DOE rings. Applying six-ring DOE, we obtain six optical spheres arranged axially in one row at the spacing of 0.83λ (λ is wavelength) with a full width at half-maximum (FWHM) of 0.42λ after the focusing of the radially polarized beams under the conditions of in-phase incident beams on both sides and the topological charge m=0. The focusing of the azimuthally polarized beams under the same conditions generates a double-row structure with 5 optical spheres in each row and the row spacing of 0.75λ. In such a case, the size and longitudinal spacing are consistent with the radial focusing results. An optical chain can be formed by the focusing of CV beams with a polarization angle of 52°. The optical chain structure can also be obtained by the focusing of CV vortex beams with m=±1. The focal field intensity distribution of the focused CV vortex beams transforms into a dark channel structure when m=2. Furthermore, the focal field can be controlled to move in the longitudinal direction by adjusting the phase difference of the incident beams on both sides of the 4Pi focusing system. The moving distance has a linear relationship with the phase difference, and the moving speed is linearly related to the change rate of the phase difference. These results have potential applications in microparticle trapping and manipulation.

Optomechanical systems based on the silicon nitride (SiN) membrane and Fabry-Perot cavity have important application value in quantum physics and precision measurement. The tunability of the resonant frequencies of the SiN membrane is significant for controlling the interaction between the light field and the membrane. In this paper, a method based on high-frequency nonharmonic excitation to adjust the resonant frequencies of the SiN membrane is proposed and demonstrated experimentally. First, we establish the theoretical frequency-response model of the SiN membrane resonator with a high-frequency nonharmonic excitation. Then, a fiber-optic Fabry-Perot interferometer is constructed to measure the vibration of the SiN membrane. Experimental results show the resonant frequencies of the SiN membrane can be adjusted by controlling the exciting voltage. The frequency shift of fundamental mode is more than that of high-order modes. Further, the proposed method is used to stabilize the resonant frequency of the SiN membrane, and the corresponding frequency shift rate is 1/200 of that without excitation. This paper provides a simple and robust method for stabilizing the resonant frequencies of SiN membranes and for controlling the linear and nonlinear coupling between mechanical modes and multi-mode cavity optical force interaction.

Gas detection plays an important role in many fields. Due to its advantages of high sensitivity and fast response, photoacoustic spectroscopy readily meets various gas detection needs. Based on the photoacoustic spectroscopy, this paper uses a mid-infrared LED with a central wavelength of 4300 nm instead of using a laser as the excitation light source of carbon dioxide gas. The technologies of long optical paths and acoustic resonance are exploited to develop a T-type photoacoustic cell that has an absorption cell with a gold-plated inner wall and an acoustic resonance tube coupled to its body. The first-order longitudinal resonance mode and resonance frequency of the photoacoustic cell are obtained through finite-element simulation. For a modulated output from the LED light source, a hardware drive circuit is designed. On the basis of the above work, a photoacoustic detection setup for carbon dioxide gas with an automated process is built. Experimental results reveal favorable linearity between photoacoustic signals and sample concentrations, and the noise-equivalent concentration (volume fraction) is 1.24×10-4. The Allan deviation is used to evaluate the stability of the setup. When the average time is 200 s, the detection sensitivity of the setup is 1.8×10-5.

A series of transparent conductive silver nanowire films with different concentration parameters and sheet resistances are prepared by the spin coating method. By the adjustment of the process parameters, the prepared films have high optical transmittance and good uniformity in the optical spectrum range of 400-10,500 nm. The influence of concentration parameters on the photoelectric properties of films is analyzed. The analysis indicates that when the concentration of silver nanowires is greater, the networks built between the silver nanowires are denser, and the effective conductive paths of the films increase, which improves the conductivity of the films. When the mass concentration of silver nanowires is 1.0 mg/mL, the film has high optical transmittance in visible and infrared spectrum bands, with the sheet resistance of 106.1 Ω/sq. To reduce the sheet resistance of the transparent conductive silver nanowire films, this paper performs thermal annealing treatment on the films, and the nodes between silver nanowires are welded by increasing the temperature, which reduces the contact resistance between silver nanowires and improves the conductivity of the films. Through the thermal annealing treatment of the films, the sheet resistance of the films is reduced from 106.1 Ω/sq to 49.5 Ω/sq when the optical transmittance of the films changes slightly.

Ultrathin silicon oxide (a-SiOx) and titanium nitride (TiN) composite film with high electron concentration has been prepared by wet chemical oxidation and direct current magnetron sputtering, and its passivation effect on the crystalline silicon surface in semiconductor-quasi-insulator-semiconductor (SQIS) photovoltaic devices is investigated. The experimental results show that the power conversion efficiency of the photovoltaic device is relatively increased by 20.72% compared with that of the aluminum back field electrode. The reason for the increase of open-circuit voltage is analyzed by combining the minority carrier lifetime measurement and AFORS-HET simulation software. Ultrathin a-SiOx/TiN and n-Si have a valence band off of 2.28 eV, which can slow down the recombination of holes on the rear interface and reduce the recombination rate to 1/10 of the original value, thereby effectively increases the open-circuit voltage and achieves the purpose of improving the power conversion efficiency of heterojunction devices. Therefore, the rear passivation contact composite material with simple process and low cost is beneficial for SQIS photovoltaic devices.