The content of water vapor in atmosphere is very little and the volume ratio of moisture to air is about 0.1%～3%, but water vapor is the most active molecule in atmosphere. Due to advantages of the high resolution, wide range and highly automatic operation, the Raman lidar has become the useful newstyle tools to measure water vapor. The new mobile home Raman lidar developed firstly by Anhui Institute of Optics and Fine Mechanics, the Chinese Academy of Sciences in Oct. 2004, is introduced as well as its structure and specification. Process method of the lidar data is described. The practical and comparative experiments were made over Hefei city. The measurement results show that this lidar has ability to gain profiles of water vapor mixing ratio from surface to the height of about 8 km in nighttime. Meanwhile, the measurement of water vapor in daytime has been taken, and the profiles of water vapor mixing ratio in boundary layer have been detected.

Entrainment zone (EZ) at the top of convective boundary layer is a transition layer between the atmospheric convective boundary layer (ACBL) and free atmosphere, the turbulence structure of entrainment zone is one of the main properties. Because of the high level of entrainment zone, it is difficult to detect extensively, and little comprehension is obtained. Generation and evolution of the atmospheric convective boundary layer are simulated in laboratory convective water tank with the dimension of 150 cm×150 cm×60 cm. When the collision light beam propagates through the simulated atmospheric convective boundary layer, distribution of intensity of beam is fluctuating. Based on the theory of light wave propagation in turbulent media, the characteristic of temperature structure constant in entrainment zone can be given. In entrainment zone, the normalized temperature structure constant accords with rule. At the level 0.8zi, the temperature structure constant is minimum, then increases gradually with height, and gets maximum at the level zi. Over the level zi, the temperature structure decreases gradually with height, and gets to zero at the level 1.2zi. At the level zi of the boundary top, the normalized temperature structure constant varies within a certain range of 0.08～0.3, this could be contributed to the large-scale structure in entrainment zone.

In order to utilize the wavefront curvature sensor in the field of laser wavefront theoretical analysis for the signal of wavefront curvature sensor under the illumination of nonuniform intensity is made. By means of Fresnel diffraction, signal of wavefront curvature sensor is computed under the conditions that the first 10 Zernike polynomials phase profile, and Gaussian distribution or normal random distribution intensity profile, and then compared with the signal with the same phase profile but uniform intensity profile. Analysis indicates that nonuniform intensity introduces some error into wavefront curvature sensor signal. Intensity of Gaussian distribution impacts significantly the sensing signal of the defocus phase, in which the error percentage achieves 25%, but for other phase profile, the effect is ignorable. Illuminated by the normal random scintillation, there is a linear relation between the error percent and the root mean square of the normal random distribution. Under some given conditions, the wavefront curvature sensor also can be utilized in the field of illumination of nonuniform intensity.

Pointing error of transmitter and atmospheric turbulence cause signal fading for satellitetoground laser communication, and the antifading characteristics of distributed groundbased array receiver with optical preamplifiers are analyzed. It is shown that the distributed array receiver can decrease the influence of pointing error of transmitter, and the distance between the subapertures can be optimized according to the pointing error of transmitter and atmospheric turbulence. Compared with the conventional array receiver, optimal transmitting beam width decreases and transmitting power requirement is debased in the distributed array receiver. Another important advantage of distributed array is when the actual rms pointing error is larger than the designed value, the power penalty can be reduced greatly by changing the distance between the subapertures.

A novel scheme to realize multirate optical codedivision multipleaccess system (OCDMA) is proposed, with which the OCDMA system can transmit signals with different data rates and quality at the same time. The bits error ratio (BER) of the system is analyzed theoretically, and the probability to improve the performance by using the hardwarelimit is evaluated. Taking a system with two different rate signals as an example, the BERs of the proposed scheme and the multirate multilength OCDMA are simulated and compared. The results show that, with the rise of transmission rate, the former system obtained a better general performance than the latter, though there appears a slight decline for the performance of the address code words for it. Finally, relevant experiments are done to verify the performance of the code and the probability of multirate OCDMA.

Spatial holeburning caused by ununiform widening of erbiumdoped fiber barricades the fully gaincontrolled amplification of laser with single wavelength. A novel configuration of automatic gainclamped optical erbiumdoped fiber amplifier is presented, and a high-birefringence fiber Bragg grating is used to generate pumping lasers, peaks of which are written at 1549.3 nm and 1549.83 nm with a wavelength spacing of 0.53 nm. By adjusting a polarization controller, single or dual laser gain control is realized. The design has a gainclamped bandwidth of 40 nm (1530～1570 nm). Within the dynamic range of input power -40～-15 dBm, the average gain and noise figure of the dual laser gain controlled erbiumdoped fiber amplifier are approximately 22.22 dB and 8.69 dB, and their excursions are limited in 0.69 dB and 1.51 dB, respectively. The system performance testing shows that the more stable characteristic of dual laser controlled erbiumdoped fiber amplifier is an obvious advantage over the single laser controlling.

The evolution of the solitons from 10 GHz optical pulses with frequency chirp is experimentally investigated in standard single-mode fiber with different lengths by use of the pulse analyzer with the secondharmonic generation frequencyresolved optical gating. It is found that the pulses can evolve into the fundamental soliton when input power is higher than its the theoretical value. The temporal width, frequency chirp and timebandwidth product of the pulses which have propagated over 3.5 dispersion length in the fiber decrease with increase of input power. The pulses almost remain unchanged in width, but vary with the propagation distance in temporal shape, frequency chirp and timebandwidth product in the sequent propagation. The higher input power is, the narrower temporal width of the formed soliton is. When input power is lower than the predicted value from the soliton theory, the pulses cannot evolve into the fundamental soliton,the temporal width of output pulses increases and the frequency chirp decreases with increase of propagation distance.

The polarization properties of highindexellipticalcore Bragg fibers are investigated via fullvector finite element method with anisotropic perfectly matched layer boundary conditions. The dependence of mode birefringence and group velocity walkoff on the fiber structural parameters has been fully analyzed. It is shown that the birefringence of highindexellipticalcore Bragg fiber is on the magnitude order of 10-2, which is at least ten times that of the conventional polarizationmaintaining fiber, and the walkoff parameters exhibit quite unusual wavelength dependence different from that of conventional polarizationmaintaining fibers. These unusual properties allow the flexible combination of high birefringence and large or zero walkoff at suitable wavelengths,and it contains potential application value. And the influence on the polarization properties imposed by the refractiveindex variance in the lowindex region is also briefly discussed. The results is helpful for the design of microstructured polarization-maintaining fibers with high performance.

The spectral property of the uniform fiber Bragg grating in the high-birefringence photonic crystal fiber are studied. By using the compact supercell method, the transmission property of photonic crystal fiber is analyzed, the coupling rule of forward and backward transmission modes is studied, and properties of uniform fiber Bragg grating written in photonic crystal fiber are concluded. The dependence of Bragg wavelength λB, defined as the resonant wavelength of the fiber Bragg grating in photonic crystal fiber with C6v symmetry, on the fiber structural parameters is expressed. The spectrum of the fiber Bragg grating inscribed in a high-birefringent photonic crystal fiber is analyzed and the peaks of reflectivity with different polarization states are much separated for the high birefringent index. Finally, the spectrum of a common fiber Bragg grating in a birefringent twomode photonic crystal fiber are obtained, in which peaks of reflectivity corresponding to two polarization states of modes LP01 and LPe11 are separated due to high birefringence of the photonic crystal fiber.

The propagation of laser beam for intersatellite laser communications belongs to the farfield diffraction, but the optical testing and verification in the laboratory is on the condition of the nearfield diffraction. In terms of the Fresnel diffraction theory, the inherent difference on the optical tracking position errors between the farfield and nearfield effects is studied, their equivalent conditions are analyzed. And it is found that the use of beam scanning in the nearfield testing can accurataly simulate the mutual movement of satellites in the farfield and the trajectory formula for beam scanning is deduced. Investigation on the characteristics of focused spot and the effect from satellite vibration is presented. The results clarify the basic concepts necessary for the testing of optical tracking performance of intersatellite laser communication terminals, which would provide helps in the ground testing.

Based on interference effect at the airgap cementing layer of GlanTaylor prism and GlanFoucault prism, the effect of airgap polarizing prism on the transmitted light intensity distribution of single-mode Gaussian beam is analyzed. The result shows that, for certain wavelength and airgap thickness, the transmitted light intensity oscillates periodically with the change of incident angle, and for a certain incident angle, the transmitted light intensity oscillates periodically with the change of airgap thickness. The profile of transmitted Gaussian beam varies with the change of incident angle and airgap thickness. Considering the effect on light intensity and beam profile, GlanTaylor prism is more robust than GlanFoucault prism, indicating that the former has better general performance than the latter.

The reconstruction of optical tomography from measurement data is an illposed problem. For such an illposed problem, a multicriterion strategy is proposed, in which three criteria of squared error function, image entropy and local smoothness function are used for the reconstruction. The multicriterion reconstruction problem is transformed to a single criterion problem by using the vector optimization method. A dynamic weight coefficient solution is proposed to determine the weight coefficient of objective function. For the realization of optical tomography reconstruction, a gradient tree based algorithm is proposed for the gradient computation of the objective function with respect to optical parameters. Different results from the multicriterion reconstruction and single criterion reconstruction based on the squared error function are presented and compared. Experimental results show that this overcomes the shortcoming of conventional single creterion reconstruction only depending on the single objective and is valid for optical tomography reconstruction and the image quality is significantly improved.

A new implement algorithm is proposed to acquire the singularities of image based on subpixel multifractal method. The method reduces the errors of the edge measure simply generated by the gray level gradient value at the integer pixel position. The gray pixel level gradient distribution law at the position of subpixel is acquired by the algorithm combining CCD imaging mechanism. Utilizing the multifractal frame, the image is segmented into a series fractal sets of the different singularity exponents, corresponding to the image ranging from the edge to the texture. Moreover, the image edge imformation containing random details can be obtained through the method mentioned above. The influence degree of edge measure is simulated by varying unit distance of wavelet center projection. The results show that the method of subpixel dividing gradient can increase measurement robustness.Comparing most singular manifold extracted from 5×5 subpixel method and edge from Sobel operator (default threshold: 36.7920), the peak signaltonoise ratio is 9.3981 dB, when the algorithm is used in standard image segmentation. Applied on the surface cracks extraction of complex background, the results agree with the human visual reception better.

A new method to study wetetching process, infrared thermal image, is proposed. In wetetching process, the substrate of metal or semiconductor is dipped in chemical reagent, and the thermal energy and infrared emission are released because of the chemical energy. The signal of infrared emission is detected and sent to computer to be processed, the infrared thermal image is obtained, and process of wet etching can be analyzed by the signal. Theoretical analysis and experimental results show that the thermal diffusion can be monitored by the new infrared thermal image method directly, which is useful for monitoring and controlling wetetching. And it is also an expansion of infrared technology.

In order to solve the reconstruction of extremely incomplete data under the condition of limited viewing angles in optical computerized tomography (OCT), the orthogonal projection sampling method is proposed to reconstruct the threedimensional flow field containing shield by the least projection direction components with the prior knowledge algorithm based on modified algebra reconstruction technology (ART). With the numerical simulation of computer, the incomplate data reconstruction precision and meansquare error of threedimentional flow field containing shield are discussed under the condition of limited viewing angles. The results indicate that the orthogonal projection sampling method can greatly reduce the error and improve the precision in the reconstruction of extremely incomplete data for the threedimensional flow field containing shield within limited viewing angle, and this work provides an important reference for the practical measurement of incomplete data with OCT.

Through theoretical analysis and numerical simulation of many phasestepping algorithms, it is concluded that the detector nonlinearity would introduce systematic error with twice spatial frequencies of the interference pattern into the result phase. Relative error of nonstandard phasestepping algorithm will increase as the detector nonlinearity increases. An error compensating technique that the phase shift error of PZT and the nonlinear error of the optoelectronic detector counteract with each other is reported, and the practical errormatching criterion is presented. The simulation shows that this technique reduces the measurement error with near one order of magnitude. It is proved that the standard phasestepping algorithms have perfect zero error response to nonlinear detectors and no effect on the accuracy of the device exists, when the amplitude information of the wavefront is ignored.

The axial response of conventional confocal microscope is affected by the tilt angle of the surface to be measured. When the ultraprecisely machined surface is measured with pointing method based on the differential confocal microscope, the tilt angle of the surface to be measured influences the pointing signal, as well as the measuring accuracy and resolution of the treated surface. On the basis of the analysis of the theoretical model of the output signal of differential confocal microscope, it is concluded that the zero point is independent of the tilt angle of surface to be measured and the resolution of the pointing signal is not affected with the tilt angle in a certain range. The experimental results are also presented.

Phaseshifting phase measurement is studied according to the holographic principle and method, a function of synchronous superposition of object complex amplitude reconstructed from Nstep phaseshifting through one integral period is proposed, and Nstep phaseshifting function is for short. Meanwhile, a novel error analysis and maximum error evaluation method for phaseshifting phase measurement is presented. In Nstep phaseshifting phase measurement, the interferograms are seen as a series of inline holograms without diffraction and the reference beam is an ideal parallelplane wave. So the Nstep phaseshifting function can be obtained by summing after multiplying the interferogram by the original reference wave. In the ideal condition, the proposed method is a kind of synchronous superposition algorithm in which the complex amplitude is separated, measured and superposed. When error exists in measurement, the result from Nshifting function is the optimal expected value of the leastsquares fitting algorithm. In the above method, N+1step phaseshifting function can be obtained from Nstep phaseshifting function, indicating that the Nstep phaseshifting function can be separated into two parts: the ideal Nstep phaseshifting function and its errors. The phaseshifting errors in Nstep phaseshifting phase measurement can be treated as same as the errors of amplitude and intensity under understanding of N+1step phaseshifting function. The difficulties of the error evaluation in phaseshifting phase measurement are depressed by this error evaluation method. Meanwhile the maximum error evaluation method of phaseshifting phase measurement and its formula are proposed.

It's hard to measure the transient pulsed laser wavefront in inertial confinement fusion (ICF) system by the conventional methods for its short pulse width, large energy, high power and large distortion. An circular radial shearing interferometer based on spatial phase modulation is proposed to measure the transient near infrared laser pulse wavefront with high precision. Transient, highly precise measurement for near-infrared laser pulse wavefront, with pulse width of nanometer scale and central wavelength of 1064 nm, can be carried out with commonpath and no reference plane with round and rectangular aperture ranging from 30～150 mm. Theory of wavefront reconstruction has been validated by the computer simulation, and an error less than 1/1000λ is obtained. Compared with the results of ZYGO interferometer, an error less than 1/15λ for both peakvalley and root mean square value, is gained with good repeatability. The system has already been used in the ICF system to test the pulsed laser wavefront, and can also be applied to other visible and infrared lasers.

Polarized light backscattering spectra invertion model is established for epitheliumlike tissue phantom based on particles model of biological tissue and sensitivity of spectra to Mie scatters morphology. This inversion model is of multiple parameters, multiple extreme values and nonlinear, complicated trigonometric calculation is involved, and the model space is limited. Thus, floating genetic algorithms ( FGAs), which reserves the optimal unit for every epoch, are applied to invert polarized light backscattering spectra for epitheliumlike tissue phantom. The strategies for encode, fitness adjusting and selection are discussed. The results indicate that, with 70 iteration epochs the relative error for every parameter inclines to be stable, about 0.02% for the minimal and less than 3% for the maximal. The genetic algorithm based on real encode can obtain the morphological parameters of the upper layer simultaneously by inverting the polarized light backscattering spectra, with global convergence, high inversion precision and robustness.

Based on selfpumped phase conjugation experiment in two blocks of Cu∶KNSBN crystal, the influence of the incident angle and photorefractive crystal geometrical configuration on selfpumped phase conjugation properties is investigated. The maximum output of phase conjugation is found with variety in the incident angle for the same incident position. The different configurations determine different maximum output. The action of crystal geometrical configuration and incident angle in dual active regions selfpumped phase conjugation mechanism is analyzed theoretically. The relation between geometrical configuration and incident angle and selfpumped beam path and beamcoupling coefficient is built. The optimal incident angle for maximum output of phase conjugation light is given in selfpumped phase conjugation effect. Finally, the formula of reflectivity of phase conjugation beam is modified. The research lays foundation of theory and experiment for selecting the optimal incident angle in fact.

Modulation instability in fiber grating with nonlinearity management is studied based on the nonlinear coupledmode equations. The dispersion relation for modulation instability is obtained. Compared with normal nonlinear fiber grating, in the fiber grating with nonlinearity management the variance of Kerr nonlinearity alters the spectral width and amplitude of the gain spectrum for modulation instability, and leads to the occurrence of new ranges of instability. In the anomalous dispersion regime, the gain amplitude of two distinct sidelobes originally on either side of the zero wave number region diminishes to zero as the variance of Kerr nonlinearity increases, and a new single peak around zero wave number appears gradually after a region of zero gain; while in the normal dispersion regime, besides the original two gain regions, a new gain region occurs around the zero wave number region, and its gain amplitude increases as the nonlinearity management coefficient augments. As a result, the fiber grating with nonlinear management provides a larger space for the generation of modulation nonlinearity.

Organic film function material is accurately evaporated, with vacuum thermal evaporation method, under the monitoring of film thickness controller with high precision. A kind of multilayer structure highly bright red organic lightemitting device is developed, with the host and doping material evaporated at the same time in the same vacuum chamber as the emitting layer, the typical structure of which is ITO-CuPc(20 nm)-αNPD(60 nm)-Alq3 (40 nm)∶Rubrene(10%)∶DCJTB(1%)-Alq3(20 nm)-LiF(10 nm)-Al(100 nm). The results show that the brightness increases linearly with the driving voltage of 5～52 V. The electroluminescent spectrum of the device shifts to the blue region, the wavelength of the emission peak changes from 638 nm to 632 nm, and at the same time, the color coordinates CIE change accordingly. When the driving voltage is low (5 V), the luminous efficiency peaks with 0.497 lm·W-1, which declines slowly with the rise of driving voltage.

Variable optical attenuator (VOA) is one of the key devices in Dense Wavelength Division Multiplexing (DWDM) network. This paper proposes a method of designing a couplingbased VOA. The VOA uses three channel waveguides and utilizes the electrooptic effects of polymer. The Beam Propagation Method (BPM) analyzes and simulates the qualities of this device. The results show that the VOA achieves nearly 60 dB loss with a low insert loss (0.48 dB) and the drive voltage is 11.7 V. This design reduces the drive voltage effectively, and realizes a large loss.

Surface electron escape probability (P) in quantum efficiency formula of GaAs photocathode commonly is taken for a constant independent of incident photon wavelength (λ) within photocathode working wave band. Quantum efficiency formula is applied in data fitting of spectral response curves of reflectionmode GaAs photocathode grown by molecular beam epitaxy, in which the epitaxial layer thickness is 1.6 μm, and doping concentration is 1×1019 cm-3. Data fitting results show that theoretical curves cannot tally with experimental curves completely, especially with declined spectral response curves of photocathode in activation chamber. This deviation is caused by the relation of P and λ, and P is not a constant independent of λ. Based on the data fitting analysis of spectral response curves, it is found that the relation of P and λ for reflection-mode photocathode approximately satisfies exponential function, and P connects with λ through surface potential barrier factor (k). Photocathode surface potential barrier factors after high- and low-temperature activation process are 3.53 and 1.36 respectively.

Onedimensional photonic crystals with two CdS defect layers which had two defect modes of 762 nm and 800 nm was fabricated by vacuum deposition process. Two-photon absorption coefficient was investigated by pumpprobe measurement. The twophoton absorption coefficient was enhanced at two defect modes. 307 cm/GW twophoton absorption coefficient at defect mode of 800 nm was greater than 116 cm/GW at defect mode of 762 nm and they were 48 and 18 times as large as that of a single CdS thin film. The enhancement of twophoton absorption coefficients was due to the enhanced electric field intensity ascribed to light localization in the two CdS defect layers. The transfer matrix method was used to calculate the internal electric field intensity in the onedimensional photonic crystal sample. It was found that the electric field intensity in defect layers at defect mode of 800 nm was larger than that of 762 nm.

Nearfield subwavelength imaging based on negative refraction in wavelike twodimensional photonic crystals within infared band. The adopted rectangular isofrequency surface results in the inhomogereity of wavelike twodimensional photonic crystal unavoidably, so the imaging is confined in the nearfield region. For the lack of the circularshaped isofrequency surface in photonic crystal matching that of the incident medium, farfield imaging based on negative refraction is unavailable in wavelike twodimensional photonic crystal. The finitedifference timedomain method is used to simulate the nearfield imaging in wavelike twodimensional photonic crystal of different thickness. The resolution of 0.28λ is achieved for single point source when the thickness of photonic crystal is twice of lattice constant. The resolution will gradually degrade as the source moves beyond the nearfield domain. Meanwhile, the image is strongly confined in the nearfield region on the other side of photonic crystal in spite of the change of source position and photonic crystal thickness. In order to further demonstrate the subwavelength imaging ability of wavelike twodimensional photonic crystal, double sources are used. Resolution of 0.35λ is obtained, but the imaging quality depends on the thickness of photonic crystal greately.

The cavity field spectra of two coupling twolevel atoms interacting with twomode field through Raman process in Kerr medium cavity are studied. By solving the eigenequation, the cavity field spectrum expression is given out and some numerical results for the initial twomode field in pure number states are presented. It is found that the Kerr effect shifts the position of peaks rightward, and affects the height, frequency and number of the peak in the case of weak initial fields. An asymmetric spectral structure appears. When the Kerr effect is strong, the influence of the atoms coupling on the spectra is restrained. The spectral structure is dependent on the Kerr effect mostly. If the field is strong initially intensities of all the peaks are enhanced by atoms coupling interaction for the cavity without Kerr medium; along with accretion of Kerr effect, the main peak near the resonant frequency becomes higher and saturated soon, and the side peaks split and become lower.

The probe field gain and its relation with the population rule in high level in the quasiΛtype fourlevel atom system interacting with two light fields at low-light levels under stable state are numerically investigated. It is found that, even for low-light level, with the variation of probe field and the relevant frequency dteuning of interlevel transition, the system always displays three gain peaks for probe field simultaneously and two fixed positions where the electromagnetically induced transparency (EIT) is observed, when all phases of Rabi frequencies are zero. When the system displays EIT, the probe field related population in the high level is zero. When the phase of Rabi frequency is changed, gain and absorption for the probe field occur, EIT phenomenon also exists, but the phase change cannot influence the population in the high level alone. It is worth noting that all results are achieved without any approximation.

Highorder harmonic continuum in the cutoff region is demonstrated with an argon gas cell driven by 7 fs/0.4 mJ ultrashort intense laser pulses. It is found that the spectral structure, modulation depth and continuum bandwidth vary greatly when the carrier envelope phase (CEP) of driving laser pulse is stabilized at different value. At some CEP value, a smooth continuous spectrum with modulation depth less than 17% and 10 eV continuum bandwidth is achieved, supporting a transformlimited single 500 as pulse in time domain.

To get a small lattice mismatch substrate for GaN, highly ［001］ oriented LiGaO2 layers were successfully fabricated on (100) βGa2O3 single crystal substrates by vapor transport equilibration (VTE). Xray diffractions indicated that the asfabricated layers by VTE were singlephase. As revealed by scanning electron microscopy (SEM), the morphologies of the layers were influenced by the VTE temperature. It was found that the grain size of the layers enlarged with the increasing of the VTE temperature. Xray diffraction results showed that the layers changed from polycrystalline to single crystalline with the increase of VTE temperature. After annealing processing, color center was introduced into LiGaO2 layers as the absorption spectrum showed, kind of the color centers is determined by the annealing temperature. It was shown that (001) LiGaO2∥(100) βGa2O3 composite substrate for GaNbased epitaxial film could be fabricated by VTE technique at the temperature lower than the melting point of LiGaO2.

A Xray detector, based on the CsI(Tl) scintillating crystal and areaarray CCD, using fiber and fiber optic plate to couple and transmit light with fanbeam linear array scanning, has been put forth. The luminescent efficiency and Xray high-resolution detection are directly influenced by the dimension of CsI(Tl) scintillating crystal, and Monte Carlo method is adopted to simulate the influences, including the relations between the energy of Xray, the distance between Xray source and the scintillating crystal, and the thickness of CsI(Tl) crystal and the Xray energy distribution, and then the whole energy peak efficiency and the transition efficiency. The results show that while the Xray energy is of 120～450 keV, and the thickness of CsI(Tl) crystal is 0.3～1.5 cm, the whole energy peak is within the range of 31.34%～96.74%, and the transition efficiency is within the range of 12.8%～97.43%. So, the Xray energy and the dimension of CsI(Tl) scintillating crystal are the most important parameters determining the Xray high resolution detection. It has certain referenced values to optimize the dimension design of CsI(Tl) scintillating crystal for Xray high resolution detection.

The selfproduced grazingexit Xray fluorescence platform is proved with high repetition and stability, and the design of experimental arrangement is reasonable, by studying the repetition of grazingexit Xray fluorescence experiment on GaAs (100) polishing wafer under the same condition, and comparing with experimental results with conventional Xray source and synchrotron light source. The experimental curve fits quite well with the theoretical one, which indicates that the method using the critical angle of total reflection of monocrystal to calibrate the grazingexit angle is feasible. Practical divergent angle is evaluated by the derivative of the standard wafer's grazingexit Xray fluorescence curve.