To counter atmospheric wind-field detection system based on Fabry-Perot interferometer the model of retrieval of wind velocity and temperature is deduced. Using Zemax software, the configuration of the system is numerical simulated, and interference fringes corresponding to two wavelengths are produced. Combined with the method of the least squares fitting a round image, data processing and the wind velocity inversion are completed. And by the comparison of the theoretical values, the accuracy of fringe movement is obtained, and wind velocity error is less than 4.2%. According to temperature detection theory, the reasonable specturm files are built and ideal simulated interference rings are obtained. Atmospheric temperture can be retrieved, and the error is less than 7.5%. Results show that simulation model, retrieval theory, data analysis and processing method are feasible.

An algorithm based on eigenfunctions was proposed for phase retrieval by solving transport-of-intensity equation. The phase could be expanded as a series of Laplacian′s eigenfunctions, and the coefficients of eigenfunction could be obtained by the integral of eigenfunction and intensity derivative along the optical axis, then the phase was reconstructed. The analytic expressions of Laplacian′s eigenfunctions on unit circle and rectangular domain were achieved. Phase retrieval by eigenfunctions on both domains was simulated, and an accurate reconstruction result could be obtained by the algorithm.

A method for real-time, quick calibration of four-quadrant photodetector (QD) signals in optical tweezers system is provided. The experimental results indicate that the method is feasible and precise. Its instability is less than 1%. In addition, the effect of the bead depth on calibration factor is investigated experimentally, and it is found that the calibration factor decreases when the bead depth increases. The method can be used to monitor the stability of the optical tweezers system.

The metal-semiconductor-metal (MSM) structured Mg0.2Zn0.8O photodetectors with different electrode spacings are made by radio frequency magnetron sputtering, conventional ultraviolet exposure and wet etching methods. The relationship between electrode spacing and dark current or responsivity is studied. Before achieving the punch-through voltage, the dark current and responsivity both decrease with the electrode spacing increasing. Furthermore, the specific mechanisms are also investigated.

The grating used for chirped pulse compressor should have broad waveband and high diffraction efficiency. A metal-multilayer-dielectric grating (MMDG) with broad waveband and high diffraction efficiency is proposed to stretch and compress chirped pulse. It consists of three parts as substrate, metal and multi-layer dielectric mirror and surface relief structure. Rigorous coupled-wave analysis is employed to analyze the diffraction property of MMDG to extend working waveband and improve the diffraction efficiency. Simulation results show that when the groove depth, residual thickness, duty cycle and incident angle are 272 nm, 10 nm, 0.23 and 54° respectively, the average diffraction efficiency is higher than 97% for TE polarized light over the 150 nm bandwidth centered at 800 nm as the wavelength changes from 732 to 886 nm.

Diffractive optical element (DOE) which maintains beam shape and intensity distribution precisely with high efficiency of the source is used as the beam shaping and smoothing element for the off-axis illumination (OAI) of lithographic system. A vector analyzing model of DOE for lithography is established based on vector angular-spectrum method, and the performance of DOE designed by using scalar algorithm such as uniformity and beam shaping is analyzed. The results show that the scalar approximation cannot fit the requirement when the diffraction angle of DOE is large and the dimension of DOE is reduced to the working wavelength, or large error of caculations will be caused.

The precision of the groove density influence the dispersion and the diffraction of gratings and the configuration of the spectrograph. In order to improve the precision of the groove density of the plane holographic gratings, a method of adjusting the groove density spatial frequency multiplication of plane holographic grating is proposed. A reference grating with a specific groove density is set in the exposure area. Adjusting the interference angle, different orders of the two diffraction wavefronts match and form Moiré pattern at the screen in a special orientation. The groove desity multiple of the reference gratings can be obtained. The relation between the period of the reference grating and the interference fringe pattern is proved in theory. The equation regulation of the reference grating for groove density alignment is pointed out. The precision relation between the groove density of the interference fringe pattern and the width of the reference grating is also analyzed. The results indicate that the precision of the groove density of adjusting is less than 1 nm at the 100 mm width of the reference grating for 300 line/mm groove density alignment; the error decreases with increasing the width of the reference grating. The method of adjusting the groove density spatial frequency multiplication of plane holographic grating based on the groove density of the reference grating satisfies the fabrication requirement.

The model of spectral clipping due to the tiled-grating gap of parallel grating pair compressor is established for the circular input beam, and the signal-to-noise ratio caused by the spectral clipping is simulated. Compared with sech2 pulse, the impact of tiled-grating gap on the signal-to-noise ratio of Gaussian pulse is less serious. With the tiled grating gap width decreasing and the beam diameter increasing, the signal-to-noise ratio of the compressed pulse can be improved. For a certain tiled-grating compressor and beam diameter, the highest signal-to-noise ratio is available with the central wavelength of the pulse equal to the central wavelength of the spectral clipping (the central wavelength of the wave band clipped by the tiled-grating gap).

There are errors between the theoretical values and the actual values for the parameters of type-Ⅳ concave holographic grating during its recording and using process, which will affect the spectral performance of the concave holographic grating. A simple grating optimization function which can be used to describe the spectral width is derived from the Fermat′s principle and a comparison to the root-mean-square (RMS) optimization function which is complex and disable to describe the spectral width is made to justify its validity. The errors are divided into two categories by their location relative to the reference plane and a quantitative analysis of their effects on the spectral performance of type-Ⅳ concave holographic grating is made. The results indicate that the error of one recording parameter or using parameter in the reference plane can be compensated by adjusting other recording parameters or other using parameters, and the recording parameter errors can be compensated by adjusting the using parameters when the grating constant is not changed; the compensation for the parameter error outside the reference plane can only be made by adjusting other parameter′s distance from the reference plane.

A guided-mode resonant filter (GMRF) based on the structure of anti-reflection (AR) thin film waveguide is proposed to realize three-primary-color filter with narrow-band. The proposed grating consists of three parts: a sub-wave grating, a waveguide and a substrate. The effects of sub-wave grating and thickness of waveguide on the GMRF reflection spectrum are investigated in detail with the method of rigorous coupled-wave analysis (RCWA). Simulation results show that the structure of λ/4-λ/2-λ/4 can improve the performance of GRMF effectively. When the grating layer is located near the substrate, the cut-off band of the optimized filter for red, green and blue primary colors are 447, 349 and 279 nm, respectively. The reflectance is above 95% and reflection bands are all narrower than 0.2 nm. The optimized GMRF shows excellent performance with narrow band and can be useful for color synthesis and modulation.

The conical array on the surface of ultra-light-trapping black silicon, which is divided into five layers, is designed using the equivalent medium theory. It′s considered that the refractive index of silicon varying with the wavelength. The elements which affect the reflectivity are analyzed systematically using the rigorous coupled-wave theory, such as incident angle, the height of the conical array and the duty factor. According to the simulation, the reflectivity is always less than 5% when the incident wavelength is from 300 to 1200 nm and the incident angle is from 0° to 60° with the conical array′s period of 135 nm and its height of 480 nm.

A high resolution vacuum ultraviolet (VUV) beamline covering 5~40 eV photon energy range for study of angle-resolved photoemission spectroscopy (ARPES) is constructed newly at the National Synchrotron Radiation Laboratory (NSRL). There are three pieces of gratings, with groove densities of 300, 600 and 1200 line/mm respectively in the spherical grating monochromator (SGM). The laminar grating with groove density of 1200 line/mm, duty cycle of 0.35, groove depth of 35 nm, and active area larger than 120 mm×20 mm is fabricated using holographic-ion beam etching on silicon substrate. The spectral calibration and performance test are performed during the ARPES beamline early adjustment and test. The undulator radiation is used directly as light source for the wavelength calibration and spectroscopic test, and absorption peaks of known gas as a standard for calibration. The results show that the energy resolution can reach 2.6 meV around 29.2 eV in absorption spectrum of Ar using 1200 line/mm grating and slit opening of 30 μm, namely energy resolving power of 11000, which meets the demand of design specifications.

The difference of amplitude profile between signal and local oscillator wave will affect the heterodyne efficiency of the inter-satellite coherent optical communication system. Based on the assumption that signal and local oscillator wave have the same polarization and geometric center, and that they are phase matching, the detailed derivation of heterodyne efficiency is given when the signal has uniform or Gauss distributions, and the local oscillator wave has Airy, uniform or Gaussan distributions. The accurate analysis formulas of heterodyne efficiency are presented for some different signal and local oscillator wave distributions. The performance of heterodyne photomixing for different signal and local oscillator wave distributions are completely given by using the numerical calculations and comparative analysis method. The results indicate that the Airy+Airy and Gauss+Gauss cases have the optimal heterodyne performance, the heterodyne efficiency can reach 100%; the values of heterodyne efficiency for the Airy+Uniform and Gauss+Uniform cases are 72% and 82%, respectively, which are similar. The Airy+Gauss case is restricted by the detector radius and the Gaussian beam′s waist radius at the same time, and it is usually used in the actual system. These provide some theoretic principles for designing the inter-satellite coherent optical communication system.

There is a trade-off between spatial resolution and sensing range of conventional Brillouin optical time domain reflectometry (BOTDR) system. Correlation coded pulses encoding technique has been used to solve this problem. However, when the sampling duration is smaller than the unit pulse width, the restored traces might be distorted by using direct correlation decoded method, thus it will decrease the spatial resolution and accuracy of the system. The influence of the direct correlation based decoded method on the spatial resolution of the system is studied, and a re-sampling based correlation technique is proposed to solve the problem introduced by the direct correlation method. 2 m spatial resolution is achieved over 20 km fiber by using 512 bit 20 ns Golay complementary codes and the new decoded method. The experimental result shows that the proposed decoding method is an effective technique for data processing of coded pulse based BOTDR system.

Based on the generalized nonlinear Schrdinger equation, the spectral evolution and the time-frequency characteristics of the soliton in photonic crystal fiber with three zero-dispersion wavelengths are numerically simulated. The results show that not only both blue- and red-shifted dispersive waves are generated in the two normal dispersion regions of the photonic crystal fiber, which is in good agreement with the phase-matching condition, but also high-intensity dispersive wave-soliton is generated in the low-frequency anomalous dispersion region due to the existence of the second anomalous dispersion region. Furthermore, it is found that in high-frequency normal dispersion region the blue-shifed dispersive wave is emitted by the dispersion wave-soliton.

The goal of wavelength conversion technology is low cost, low-power dissipation and integratability. An all-optical wavelength conversion scheme based on Fabry-Pérot laser diode (FP-LD) mode competition is proposed. The wavelength conversion employing dual-mode competition of the FP-LD is realized by injecting probe light and signal light into different modes of the FP-LD. The principle of wavelength conversion is analyzed. The signal-to-noise ratio (SNR), extinction ration (ER) and chirp characteristics of the signal after wavelength conversion are studied experimentally. The wavelength conversion and transmission experiments of 10 Gb/s non-return-to zero (NRZ) signals at different wavelengths are conducted. The results of bit error rate test show that the power penalty after wavelength conversion is less than 1 dB, and the power penalties of different wavelengths after 20 km transmission are all less than 1 dB.

A new model of liquid axicon for producing a single bottle beam with tunable size is proposed. In the aspect of geometrical optics, it analyses the principle that how the liquid axicon produces the bottle beam is analysed, and meanwhile, the diffracting theory is used to simulate and demonstrate. It is concluded that liquid axicon can produce single bottle beam when the liquid refractive index of the injected liquid is higher than that of the axicon material. Through optics software TracePro the model of liquid axicon is established and the properties of the transmission conversion of the light beam are simulated. The influence upon the size, location of the bottle beam with different refractive index is calcalated and simulated, and the conclusion shows that, the lower the liquid refractive index is, the smaller the size of bottle beam is, and the theoretical analysis basically corresponds with the simulation result.

Based on the previous analysis of spatial distribution of the ghost images in high-power laser system, interference among the ghost image beams reflected by front edge and the trailing edge of the pulse which overlapped together were further considered. According to the Shenguang (SG)Ⅱ facility, the maximal focal length of minimal aberration lens which induced ghost images interference were calculated corresponding to the different laser pulse width range from 3 to 10 ns. Several situations of the interferences existing with the ghost images were simply simulated with incident plane wave. The results indicate that high density modulations existed on the focal plane of ghost images which greatly enhanced the peak power of the ghost images. Near the ghost focal plane, there were also dramatically spatial interference modulations which increased ghost beam power greatly especially for the higher order ghost images, which should be emphasized greatly in practice.

To detect dim and small target in infrared image with low signal-to-noise ratio, a novel algorithm based on multi-scale gradient of gray level is presented. Firstly, for all points in an image, the algorithm selects up, down, left and right reference points around the current point according to the size of target, and identifies potential targets based on the maximum gradients between current point and reference points. Then, final targets are identified with the fusion of potential target detection results of the three sequential images. The algorithm does not need background prediction, and achieves not only higher detection rate but also lower false alarm rate in infrared image with low signal-to-noise ratio and complex background. Experimental results show the validity of the algorithm.

Considering the influence of acceleration and the Gaussian envelope on laser Doppler velocimeter (LDV), the parameters estimation of Doppler signal with Gaussian envelope is investigated based on introducing the acceleration. According to the theory of mathematics statistics, the Cramer-Rao lower bounds (CRLB) of Doppler circular frequency and its first order rate of change are analyzed, formulas of CRLB are given, and the power spectrum estimation with adjustment is discussed. The theoretical analysis and simulation results show that the CRLB of the variance of estimated parameters are related to the observation data, signal-to-noise ratio (SNR), the width of Gaussian envelope, and it can be decreased by increasing the length of observation data and improving the SNR. When the length of observation data is fixed, the larger the acceleration is, the narrower the Gaussian envelope is, and the CRLB of Doppler circular frequency and its first order rate are also larger. The gap between the variances of the measuring results and the CRLB narrows when the SNR of the signal is improved, almost eliminates when the SNR is higher than 6 dB. The acquired result is in harmony with those acquired by other researchers.

Ultrahigh-order guide modes are excited by employing the free-space coupling technique in a symmetrical metal-cladding optical waveguide fabricated with a submillimeter-scale transparent PMN-PT ceramics. The voltage-dependent variation of the reflectivity of the light intensity is obtained according to the shift of the attenuated total reflection (ATR) resonance dip and the quadratic electro-optic coefficient of the transparent PMN-PT ceramics is then calculated.

Based on the multi-beam interference principle, the Gaussian beam transmitted-intensity expression on a non-parallel angle-tuned thin-film filter in oblique incidence has been derived. According to the result, the influence of the transmitted-intensity distribution with different incident angles, especially with different wedge angles of the non-paralleled thin-film filter has been analyzed theoretically. The calculation and experimental results show that the transmitted-intensity distribution, transmitted-peak and the rectangle degree are greatly influenced by the wedge angle, the incident angle and the polarity of the wedge angle. In order to get stable transmitted-characteristics and high rectangle degree, the parallelism of the angle-tuned thin-film filter should be improved in fabrication.

Based on the ultraviolet resist (ZPU) and polysulfone polymer materials, the structure parameters of the polymer ring resonator such as waveguide cross section, bending radius, waveguide gap in the coupling region are designed and optimized, and the filtering characteristic is analyzed. The polymer micro-ring resonator is fabricated using the traditional micro-fabrication techniques including contact lithography and reactive iron etching. The experimental results agree well with the simulation results. Results show that around the telecom wavelength of 1550 nm, the free spectral range, 3 dB bandwidth, insertion loss, extinction ratio and quality factor of the polymer micro-ring resonator are 0.21 nm, 0.04 nm, 26 dB, 11 dB, 3.87×104, respectively. The polymer micro-ring resonator is very useful for telecom and integrated sensor chips.

The principle of Kerr noise is analyzed in the integrated optical resonator gyroscope (IORG), and the mathematical model is built by the theory of guided-wave optics, the relation of the fundamental detection limit caused by the Kerr noise Ωk is formulated. Furthermore, the curve of resonator length with Ωk under different spectrum line widths of laser is simulated. As to the prominent reason of Kerr noise in the IORG which is the unbalance splitting ratio of integrated optical modulation, the calculation and software simulation are carried out. After that, the suppression methods to eliminate the Kerr noise are put forward. Finally, the experimental setup is built up and by adjusting the adjustable attenuator the equal optical intensity is realized, which shows that the Kerr noise in the IORG is effectively suppressed simultaneously.

A novel fiber laserquantum-dot-doped fiber laser (QDFL) is presented. A two-lever system is established based on the observed single-peak photoluminescence of the CdSe/ZnS quantum dot (QD) that is used as the active medium. The optimal parameters (doping concentration, fiber length, reflectivity of output mirror, pump wavelength) of QDFL are determined by solving the rate equations and the power propagation equations in the two-level system, using a genetic algorithm and taking the laser output power as an objective function. Compared with the conventional Nd3+-doped fiber laser, QDFL shows a lower saturated doping concentration, a shorter saturated fiber length and a higher pump efficiency. Upon a pump power of 2 W, the simulated laser power of QDFL can reach 1.5 W. Further, such a single-wavelength QDFL can be developed into a novel multi-wavelength laser in multiple sized doped cases.

Nonpolar ZnO films of high crystallization have been prepared by metal organic chemical vapor phase deposition on LiGaO2 (100) substrates. Effect of substrate temperature on the crystallization of ZnO films is investigated. X-ray diffraction (XRD) results indicated that high quality nonpolar ZnO films are obtained at 500 ℃. The surface morphologic image and crystal size are observed by atomic force microscope (AFM). The ZnO film deposited at 500 ℃ shows a smoother and uniform morphologic image, the root mean square (RMS) value roughness is 0.626 nm. According to the photoluminoscence (PL) spectra, the near band emission peaks of the ZnO films are located at 375 nm. However, the yellow emission is stronger when the film is deposited at 400 ℃, inferring that the crystallization property of ZnO film is poor at lower temperature, and there are more defects in the film.

A new nanometer twisted conjugated gammadion chiral structure is proposed, and the twisted angle of upper and lower layers of the structure is optimized. The transmission and reflection spectra for the incident wave with different frequencies are obtained from numerical simulations. The resonance dips of the transmission spectra at the frequencies of 145 and 300 THz are observed. The circular dichroism, polarization azimuth rotation angle are calculated from simulated transmission and reflection spectra. The results show that the larger circular dichroism and exceptionally strong optical activity are found at the resonance frequency region. The maximum polarization azimuth rotation angle θ can reach 95°. Then the effective parameters of the structure, including relative permittivity, relative permeability, chiral parameter and refractive indexes, are retrieved from simulated transmission and reflection spectra. It is found that the negative refractive indexes of the right-handed circularly polarized and left-handed circularly polarized waves of the structure can be realized at the resonance frequency region. The numerical results demonstrate that the negative refractive index of the chiral metamaterial is due to the large chiral parameter.

Ga-assisted desorption of native surface oxide on GaAs substrate is investigated at low substrate temperature. The progress is monitored by reflection high energy electron diffraction (RHEED), and time dependence of RHEED intensity curves is recorded by a CCD camera. By fitting the curves, the oxide desorption rates are deduced, which depend strongly on the substrate temperature and can be fitted well by Arrhenius relationship. Furthermore, the absorption coefficients of the oxide layer to the high energy electron are estimated. The feasibility of using Ga-assisted desorption for overgrowth is confirmed by growth of high quality low-density InAs quantum dot samples. The result suggests that Ga-assisted desorption is extremely promising for the overgrowth, where a smooth interface is needed to be well preserved.

Scanning transmission electron microscopy (STEM) and temperature dependent photoluminescence (PL) measurement are used to study the influence of In fraction on the optical properties of InGaN/ GaN blue light-emitting diode (LED). STEM results reveal that both of two samples have the same quantum-well structure. Low-temperature dependence of PL shows that the peak energy of one sample with lower In fraction exhibites a classical S type (Red-Blue-Red) with increasing temperature. Currently it is recommended that the blue shift of peak energy (with increasing temperature) is mainly due to exciton recombination, which is caused by inhomogeneous In distribution. However, the sample with higher In fraction doesn′t show any blue shift about the peak energy. This unnormal phenomenon can be mainly attributed to the large thermal barrier caused by potential fluctuation of high In composition, which prohibits carriers transition from strong localized state into weak localized state (this process can cause blue shift of energy). Meanwhile, band filling process of carriers becomes prominent in the role of energy blue shift from 160 K to 300 K, this can be attributed to the more obvious quantum confined effect in higher In fraction sample, compared with lower In fraction, resulting in smaller average red shift of peak energy.

In order to study the temporal behavior of the low-amplitude dark and gray spatial solitons in biased centrosymmetric photorefractive crystals, the expressions of time-dependent space-charge field and dynamical evolution equation are obtained. The temporal behavior of the intensity profiles and the intensity full width at half maximum (FWHM) of dark and gray solitons are obtained by numerical method. The results indicate that a broad soliton is generated at the beginning, and as time evolves, the intensity width of photovoltaic spatial solitons decreases monotonously to a minimum value toward steady state. Within the same propagation time, the higher the ratio of soliton peak intensity to the dark irradiation intensity is, the shorter the FWHM of photovoltaic solitons is. Dark and gray solitons have the similar temporal properties.

A novel design of the gas Cherenkov detector (GCD) using three pieces of 90° off-axis parabolic mirror and a turning flat mirror will take place the conventional GCD using Cassegrain reflector optics. Light collection is more effective by the using of the off-axis optical system and positioning the optical detector at the exit pupil of the optical system. Optical ray tracing demonstrates that how light can be collected by the optical detector from different object surface. An integration design of the whole GCD is given and the details are analyzed to enhance light collection efficiency and offer better radiation shield.

Liquid crystal tunable filter (LCTF) is a powerful dispersive element in multi-spectral imaging. Its physical size, however, exceeds the flange focal distance of general commercial off-the-shelf photographic lenses, which makes it difficult to combine with commercial lenses and cameras directly for multi-spectral imaging applications. A relay lens design working at unit magnification is proposed to accommodate the interface. The spectral range is from 400 nm to 720 nm, and the working F number is 4.5. With the symmetrical design strategy, coma, distortion and lateral color are canceled out automatically. The result satisfies the requirements of multi-spectral imaging. This system is compact and low-cost. It is applicable to the design of a portable multi-spectral imager based on the liquid crystal tunable filter.

A design method of elliptical window optical system is proposed and it is based on the fixed aspheric corrector. The algebraic expression of the fixed aspheric corrector is given combining the generalized Coddington equation and the theory of geometric optics. In addition, eliminating astigmatism and sine condition are taken as evaluating parameters, aspheric surface equation is fit by least square method to eliminate astigmatism and coma. Zernike polynomials is used to replace traditional merit function in order to overcome slow convergence in traditional design method of elliptical window optical system. The athermalization of elliptical window optical system is realized by comparing matches of different materials. A complete design example of elliptical window optical system is demonstrated and the results show that the MTF of optical system approaches the diffraction limit across the entire scanning field.

Optical system of a space fisheye camera based on 1/3 inch (1 inch=25.4 mm) CCD is designed considering the space environment. The focal length of the fisheye optical system is 2.1 mm, the F number is 4, and the field of view is 210°. The required working temperature is between -40 ℃-+60 ℃. After careful studying the influence of space environment such as the space γ radiation and ultraviolet radiation, temperature change and vacuum on the optical system, the technique for keeping image quality in space environment is put forward. The fisheye lens can reach excellent imaging quality in such an austere environment, and provide important guarantee for information capture with extreme wide-angle view on the orbit.

It′s very difficult to calculate the calibration coefficients precisely using least mean squares (LMS) and extended Kalman filters (EKF) algorithm for wide field of view star sensor because of large optical aberration. And the error factors that affect measuring precision of star sensors are analyzed, which indicate that it is important to calibrate optical distortion before focal length calibration by comparing simulation data with and without distortion and noise. At the same time, four kinds of optical distortion methods are introduced. In order to improve calibration precision, a new method is proposed, in which main point is calibrated firstly, optical distortion parameter secondly and focal length finally. Simulation results demonstrate that the focal length calibration deviation reaches 3.1 μm and 2.2 μm using LMS and EKF algorithm after optical aberration correction. The statistical error of star angular distance is about 1/10~1/8 of the traditional on-orbit calibration method.

SU-8 epoxy resin is a kind of polymer material of nonlinear optical properties. It is ultraviolet sensitive and its refractive index increases because of the photopolymerization. Finite-difference time-domain method is employed to simulate the process of photopolymerization, and the inner distribution of refractive index of SU-8 film is obtained. Both the process in the plane thin film and the thin film with a microlens are simulated. It is concluded that self-focusing effect which results from the refractive index change leads to the formation of self-written waveguide. The self-written waveguide consists of two continuous parts, one is cone and the other is cylinder. The length of the self-written waveguide increases with the time of exposure. Given the same exposure time, cone part in the SU-8 film with a microlens is more convergent than that without a microlens, and the length of the self-written waveguide in the SU-8 film with a microlens is longer than that without a microlens. The length of the self-written waveguide decreases with the increase of the ratio of the height to the basal diameter, given the same exposure time and a constant basal diameter of the microlens. The length of the self-written waveguide with the ratio 0.08 is 12.2% longer than that with the ratio 0.24, and the cone part with the ratio 0.08 is 19.2% longer than that with the ratio 0.24.

In order to realize the electro-optic tunable filter with high rectangular degree, low sidelobe, electro-optic tuning reverse derivation and passband width tuning, the N-stage interdigital electrodes distributed periodically are provided with independent voltages V on the x-cut, y-propagation Ti-diffused LiNbO3. The induced periodic electrical field could achieve conversion of the quasi-TE mode and the quasi-TM mode and also the wavelength filtering. The finite impulse response (FIR) digital filter network is composed of this optical path. And there is a one to one correspondence between the transmission function H(z) and the voltage matrix V. Using z-transform and method of undetermined coefficients, the voltage matrix V is derived to realize the electro-optic tuning reverse derivation. The tunable passband width in the free spectrum range (FSR) can be realized with different voltage matrices V. This method is proved to be feasible through the simulation. When N is equal to 17, side mode suppression ratio (SMSR) could reach 25 dB. Also, the rectangular degree of the filtering increases with the increase of cascading stage N.

The dispersion relation for coaxial dielectric-metal-dielectric structure by making an explicit connection with the planar dielectric-metal-dielectric structure is obtained. Each mode splits into a high effective refractive index branch and a low effective refractive index branch. An intuitive picture that allows for a qualitative understanding is provided. The effective refractive indexes and propagation losses of the modes are calculated by using the analytical dispersion expression. They fit well with the numerical simulation results obtained by using finite element method in the range of 0.5~1.5 μm. The results the applicability of the dispersion equation are also analyed.

Various shapes of three-layers composite material nanoparticles is manufactured by using high-resolution electron beam lithography method and the influence of shape on extinction properties for the nanoparticles was studied. The results show that if the direction of light source polarization is parallel to the short axis, the resonance peak gets a blue shift with the increasing aspect ratio; if the direction of light source polarization is parallel to the long axis, the resonance peak gets a red shift with the increasing aspect ratio. The extinction properties of nanoparticles is also calculated by using finite-difference time-domain algorithm and Lorentz model of surface plasmon, the curve of extinction spectra and the position of resonance peaks are basically consistent with the experimental results. Besides, the influence of thickness on extinction properties for the body material is studied, and it is found that if the thickness changed from 20 to 90 nm, the resonance peak gets a blue shift of 3~115 nm.

As the actual light intensity has nonuniform spatial distribution, the surface plasmon polariton (SPP) standing wave field induced by two interference Gaussian beams is simulated using angular spectrum representation on the basis of the Kretschmann model. The Gaussian beams-induced SPP interference fringes have complex amplitude distribution, which is distinctly different from those induced by ideal plane wave. This indicates that the spatial non-uniformity of light intensity can seriously affect the distribution of exposure depth. The effects of the thickness, loss of metal films and the photoresist dielectric constant on the standing wave field are also investigated, and the results show that inappropriate metal thickness and tiny absorption loss will greatly weaken field intensity of the standing wave. Besides, the surface plasmon resonance will not be generated if the photoresist dielectric constant is too large.

Vectorial properties of finite electromagnetic beams are described by a global characteristic unit vector and the generalized Jones vector. A parallel characteristic unit vector to the propagation axis corresponds to the cylindrical vector beam. The cylindrical vector beams in the transmission are described by the representation theory in the paraxial approximation. It is found that the characteristic unit vector changes in the transmission, which is neither parallel nor perpendicular to the propagation axis. The change of the characteristic unit vector causes a displacement perpendicular to the incidence plane, and the displacement depends on the polarization ellipticity. The differently circularly (left and right) polarized light beams undergo opposite displacement. Similar to the previously observed so-called spin Hall effect of uniformly polarized beam, spin Hall effect of cylindrical vector beam could also be generated in the transmission.

Spontaneous four-wave mixing originated from third-order (Kerr) nonlinear effect in fiber, is an efficient way to produce quantum correlated photon pairs. If the commercially available high-nonlinear fiber can be used as the Kerr nonlinear medium, it will be beneficial for simplifying the experimental setup, because the length of fiber and power of pump can be reduced. Near-degenerate photon pairs in telecom band are produced in 20 m-long high nonlinear fiber by pulsed laser. The influence of weak-birefringence of high nonlinear fiber upon the production efficiency of photon pairs is experimentally investigated and analyzed. When the polarization of pump is along the fast or slow axis of the high nonlinear fiber, the production efficiency of photon pairs is maximal, and photon pairs are co-polarized with the pump; when the polarization of pump is along the 45° in respect to the fast axis, the production efficiency of photon pairs is minimal, which is about 80% of maximum.

In a synthetic aperture imaging ladar (SAIL), the imaging quality is greatly degraded from the laser speckle effect of diffused target. A systematic solution to reduce speckle is given, and thus the theoretic basis is founded for the design of SAIL structure and operating mode. The speckle characteristics relating to the scale of target resolution element, the wavelength change from laser chirp, the correlation of target surface and the uncertain distribution of intensity at receiver plane are investigated. A complex coherence function of integrated speckle over antenna aperture is defined, it is a convolution between the correlation of antenna profile and the complex correlation factor of the speckle from a resolution element and its width is the possible length for aperture synthesizing. A design principle for transmitter aperture, receiver aperture and used length for synthesizing is given to achieve a long synthesizing length. And a fluctuation phenomenon of beat signal due to the speckle wandering caused by the chirp is also discovered and analyzed. Then a sliding spotlight mode which has a magnification from the SAIL movement to the beam scan is considered to effectively utilize the limited length of aperture synthesizing. A multi-channel transmitter/receiver construction is proposed to enhance the detectability of speckled echo.

To research the influence of the defocusing of CCD on the measurement accuracy when using the joint transform correlator (JTC) to measure the image motion of space camera, the quadratic phase factor induced by defocusing is deduced based on the theory of Fresnel diffraction. Then the measurement error caused by defocusing is studied by emulation and experiment respectively for the JTC which is processed by using the method of power spectrum subtraction and binarization with threshold of zero. The results show that, the measurement error is not more than 10% when the defocusing amount is within 5% of the focal length of the Fourier lens, and the robustness is good; if the defocusing amount is increasing, the measurement error will nonlinearly increase observably. The results of the research can supply provide useful reference for the real-time and online measurement of image motion.

In order to investigate the polarized light scattering between films and substrates, first-order vector diffraction theory is applied to derive the polarized bidirectional reflectance distribution function (PBRDF) of substrate. The analysis results are compared with the experimental ones, and the influence of the thickness of SiO2 and TiO2 films on the PBRDF is numarically simulated, at interfaces roughness perfectly correlated and completely uncorrelated models, respectively. And the relation between dielectric constant and the PBRDF is discussed. The results show that with the increment of thickness of SiO2 films the PBRDF is coincident little by little at interfaces roughness perfectly correlated and completely uncorrelated models. And film can perfectly replicate surface profile of substrate at this time. With the increment of thickness of TiO2 films, the PBRDF becomes larger and larger, which illustrates that TiO2 film has smoothing effect on the roughness of the substrate. The real part and imaginary part of dielectric constant have opposite influences on the PBRDF. The optical information of the substrate can be measured and calculated by the PBRDF of different films.

Based on the principle that both scalar potential and vector potential are independent of the dielectric tensor, by using the relation between electric dipoles and their potentials, the analytical expression of scattering field for the anisotropic medium target is developed. The transformation of dielectric tensor is researched. The expression of a tensor in spherical coordinate system is presented in detail. The electric fields inside and outside the target are expanded in series, and then the primary scattering field from the anisotropic gypsum cone is presented. The theoretical result coincides with those in literatures and its validity is tested. When the wavelength is near to the size of the nanoparticle, some simulations are carried out, the result shows that the scattering of anisotropic cones has the characteristic of dipole radiation. The method used is simple and universal. The result has provided a theory for studying the laser scattering from a complex anisotropic target.

The laser diode self-mixing interferometry can be employed to measure nano-particles and nano-fluids. The analytical expression of power density of the feedback self-mixing signal from a flowing Brownian motion system is obtained theoretically. It reveals that the power density is of the Voigt profile, which is a convolution of the Lorentzian and Gaussian line profiles. The characteristics of the self-mixing spectrum are studied numerically. It is found that the Doppler shift and the Gaussian term in the power density are determined by flow velocity while the Lorentzian term is determined by the Brownian motion of particles (i.e., the particle size). The study provides fundamentals for the measurement of nano-fluids.

The quantitative models of cholesterol and triglyceride analysis in whole blood are established by using near-infrared transmission spectroscopy combined with chemo-metrics methods. First the transmission spectra of whole blood in wavelength range of near infrared full spectrum (800~2500 nm) is acquired, and then the calibration models for cholesterol and triglyceride analysis are optimized by using interval partial least squares (iPLS). The results show that the optimal wave band of calibration models for cholesterol and triglyceride lies 1650~1730 nm and 2260~2340 nm respectively, and the prediction coefficient of optimal analysis model is 0.79 and 0.865 respectively, the root mean square errors of prediction of related prediction model are 0.5 mol/kL and 0.28 mol/kL respectively. The study demonstrates that compared with whole spectra data model, the iPLS model can not only improve precision, but also simplify the model. Near-infrared spectroscopy combined with iPLS can be used successfully to determine the content of cholesterol and triglyceride in whole blood.

Plasma of steel sample is produced by single and reheating double pulses laser-induced breakdown spectroscopy in air, and the spectrum excited by two ways is studied. Two lasers are used in reheating double pulses laser-induced breakdown spectroscopy. One arm is used to ablate samples to produce plasma; while the other arm is reheat to the produced plasma. Through comparison of reheating double pulses and single laser on steel sample show that intensity of emission spectrum is enhanced; signal of continuous spectrum is amplified; Signal repeatability is improved using reheating double pulses excitation, the measured signal relative standard deviation is decreased from 5.0% to 2.0% by ten times. In addition, the factor affecting signal enhancement is presented and discussed. The signal enhancement level with the delay between two pulses and the acquisition delay are also investigated. At the same time, an additional research issue to be tackled is the effect of the increasing influenced by excitation energy levels. The higher the excitation energy levels are, the more the enhancements are.

Tantalum ethoxide [Ta(OC2H5)5] and water vapor (H2O) are used as chemical precursors to deposit Ta2O5 thin films on K9 and JGS1 substrates by atomic layer deposition (ALD) at the temperature of 250 ℃ and 300 ℃ respectively. Characteristics of the films such as optical properties, microstructure and surface morphological image are investigated by spectrophotometer, X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), scanning electron microscope (SEM) and atomic force microscopy (AFM). The results show that both the as-deposited and annealed Ta2O5 films are amorphous. The film deposited at the temperature of 250 ℃ possesses excellent microstructure with good optical properties in the spectral region from mid-ultraviolet to near-infrared, and it can be well used for optical coatings as a kind of high-refractive-index materials.

A theoretical analysis of X-ray microscopy based on the fractional Talbot effect of phase grating and two different imaging formulas when the grating is placed in different locations is given. A cell model of water and protein is established. According to the imaging formulas and the cell model, a set of system parameters are designed and a conclusion which can be made is that mid-energy X-ray between 1 and 2 keV will be the optimal choice considering long focal depth and high image contrast. Then the imaging results with 1.5 keV X-ray is simulated and the phase distribution of the model is recoveried by the phase stepping method. Simulation results show that the phase contrast image with mid-energy X-ray is much better than the absorption contrast one in contrast and resolution under the same signal-to-noise ratio.

A program for the ray-tracing of cylindrically Laue bent crystal is developed based on the lamellar model of X-ray geometrical diffraction. The program is compiled by C and runs under the computing environment VC++. It is used to simulate different projects of the direction and energy band of the diffracted rays. The results show that the program is exactly correct on the trace of rays according to the excellent agreement between the simulations and theoretical calculation (or experimental results). Study of the X-rays′ path (in and out) and energy band which can be diffracted by making use of this ray-tracing program, we obtain the optical properties of Laue bent crystal such as focusing, collimation and dispersion. It is useful for the research and application of Laue bent crystal.