Origin of life and evolution, early life on Earth and the global climate changes are among the major concerns because of their close relations with human beings. X-ray spectrometry (X-ray fluorescence and X-ray absorption) plays an important role in making scientific discoveries by determination of the elements and their species. In the review, we demonstrate the application of XRS to the studies in the origin of life, the identification of life traces on the early Earth and the global climate changes. Major reviews are on (1) how the analyses of C, Fe and S species are used in interpreting their roles in the hydrothermal vents and the RNA world; (2) what are the difficulties in the identification of early life and can be done by the determination of stromatolite, minerals with disproportionation reaction and organic microfossils in the identification of biogenic and abiogenic process; (3) what are the correlations among Fe sources, the species, organic matter and the bioavailability in the carbon circles in the oceans and the inland water systems.

The concentration, spatial distribution and speciation of elements in environmental matrices are the key to understanding their biological functions and environmental behaviors. This paper aims to review the recent applications and challenges of X-ray spectrometry in biology and ecological environment. It is demonstrated that X-ray fluorescence spectroscopy analysis can provide quantitative data on the translocation and distribution of elements in living plants. Micro-X-ray fluorescence and X-ray absorption spectroscopy are unique in providing in situ information, and both help to understand the interaction between organisms and elements, especially the uptake, transport, accumulation and detoxification mechanism of elements in organisms. Meanwhile, both are also used to reveal the environmental behaviors such as the source, evolution and fate of elements in typical environmental samples. However, due to the complexity and diversity of biological and environmental matrices, there are still some technical difficulties and challenges, such as overcoming the self-absorption effect of X-ray fluorescence, accurately identifying the low-abundance (5%~10%) elemental species, and rapidly arresting the transient redox reactions of elements in living cells.

In recent years, with the development of X-ray source, X-ray monochromatic-focusing and detector technology and calculation methods, X-ray fluorescence (XRF) and X-ray Absorption Spectroscopy (XAS) is used to obtain better two-dimensional/three-dimensional element spatial distribution characteristics and element morphology and coordination information in the lab and widely applied in geology, environment, biology, materials, medicine, art, cultural heritage and industry. However, more requirements are put forward to analyze spatial element distribution and species characteristics in the lab, such as analyzing unigue samples with complex matrices, obtaining better detection limits and shorter detection time, and higher spatial resolution, etc. Laboratory-type XRF element spatial distribution and imaging technology are divided into focus scanning type 2D/3D XRF element spatial distribution imaging technology (focus scanning 2D/3D XRF), full field-micro X-ray fluorescence (FF-MXRF) and XRF computer Tomography (XRF-CT). The advancement of 2D/3D XRF hardware technology includes: liquid metal jet source and pyroelectric X-ray generator and other new laboratory X-ray source have improved the excitation efficiency closed-loop feedback systems overcome problemes caused by rough sample surfaces. X-ray monochromatic and vacuum systems play an important role in reducing background and light elements interferences. Pn-charge coupled device and micro pore optic have promoted the progress of FF-MXRF; The development of monochromatic focusing systems such as toroidal curved crystal, spherical curved crystal and columnar curved crystal have enabled laboratory XAS technology development. The continuous improvement of calculation methods has greatly promoted laboratory XRF element spatial distribution imaging technology and laboratory XAS technology, especially the development of 2D/3D XRF and XRF-CT. Exploring laboratory X-ray source systems, efficient monochromatic and focusing optical systems, and promoting dynamic X-ray film shooting technology are essential in the future.

Since the ultrafine reference materials were reported firstly by the “Geoanalysis2003” international conference and the standard SRM2703 for ultrafine marine sediments were published firstly by NIST, the research and development of ultra-fine reference materials and analysis of ultrafine samples had made gratifying progress in China. The research background of ultra-fine reference materials in China was introduced briefly. It was emphasized that the historical evolution of sample size and analytical accuracy promoted alternately in geological analysis and that the analysis of X-ray fluorescence spectrometry had played an important role on finding that sample size had become the key to make further improvement on the analytical accuracy. And the research work on the application of ultra-fine reference materials was reviewed in China, such as carbonate, marine sediments, Platinum group elements in seamount cobalt rich crust, platinum group elements, Gulf estuarine sediments, a variety of minerals. And the research of ultra-fine samples was reviewed by X-ray fluorescence technology and Inductively plasma spectroscopy/mass spectrometry in recent years. The technology and the research work related to the preparation of ultra-fine samples were discussed. The significance of the work and its recent and far-reaching influence on the analytical laboratory were also pointed out.

In-situ X-ray fluorescence analysis technology is an instrumental analysis technology for rapid qualitative and quantitative analysis of elements in the measured object under in-situ working conditions. It is widely used in some fields where large analytical instruments and chemical analysis methods cannot be directly operated. This paper reviews the research progress of in-situ X-ray fluorescence analysis technology in China in the past 20 years. The research progress and main technical characteristics of in-situ X-ray spectrometer are reviewed from the perspectives of in-situ analysis and in-situ sampling analysis. The vital technical problems of in-situ analysis data processing of X-ray spectrum are discussed. The representative applications of X-ray spectrum analysis in the geological survey, environmental pollution investigation, identification of cultural relics and alloy analysis are introduced. The research status and progress of in-situ X-ray spectrometers in the world are evaluated. The potential research directions put forward to make great progress in more application fields.

The fitting software QMXRS (Quantitative analysis of Micro-Energy Dispersive X-ray fluorescence spectra) which is used for fitting micro-energy X-ray fluorescence spectrum focused by poly-capillary optics is developed in our Lab with Python language. The signal de-noising of the wavelet transform, background subtraction, energy calibration and characteristic peak fitting of spectrum is carried out by QMXRS. The ploy-capillary X-ray optics used here can improve the intensity of X-ray and change the distributions of X-ray fluorescence spectra. The background prediction model of QMXRS can effectively retain peak information and improve the accuracy of data processing of energy spectra. On the other hand, the constraint of the full width at half maximum(FWHM) of the Gaussian peak is added to the conventional nonlinear least-squares fitting. The Micro-Energy dispersive X-ray fluorescence spectra of the NIST 610 standard sample is used to demonstrate the performances of QMXRS. The results show that our QMXRS is better than PyMca and QXAS.

High temperature vulcanized (HTV) silicone rubber composite insulators are widely used in UHV transmission lines, and its ultraviolet aging resistance has been paid more and more attention. In this paper, the samples of high temperature vulcanized silicone rubber come from two manufacturers, X-ray full spectrum scanning and narrow zone spectrum analysis were carried out on the radiated samples to analyze the influence of ultraviolet radiation on the surface chemical properties of high temperature vulcanized silicone rubber, and then to determine the aging mechanism of high temperature vulcanized silicone rubber. The results are as follows: the main elements of high temperature vulcanized silicone rubber are O, C and Si, among which O1S mainly exist in the form of O—Si—O (532.4 eV). After ultraviolet radiation, the COOH peak of 534 eV was increased, and the integral area increased with the extension of irradiation time. C1S was C—H, C—C (284.8 eV) or C—O (286.3 eV). With the extension of radiation time, the integral area of C—H and C—C binding energy peak decreased, while that of C—O binding energy peak increased slightly. Si2p was Si—C (102.39 eV), the binding energy peak of SiOx (x=3～4) was increased after ultraviolet radiation (103.6 eV), and the integral area increased with the extension of radiation time. The mechanism of accelerated aging caused by UV radiation is that part of Si—C and C—C bonds with a low bond energy of silicone rubber are cut off by ultraviolet radiation. After cleavage, free radicals can combine with each other and cross-link to form SiOx (x=3～4); Free radicals are oxidized by ozone to COOH. X-ray photoelectron spectroscopy (XPS) plays an important role in the study of aging mechanism of high temperature vulcanized silicone rubber.

The segregation of elements and the original particle boundary of powder metallurgy superalloy are the important factors affecting the material properties. Because the particle size is usually tens of microns, the traditional analysis method of composition distribution can not realize the fine characterization of the composition distribution at the original particle boundary of powder metallurgy superalloy. Microbeam X-ray fluorescence spectrometry (μ-XRF) is a non-destructive micro area composition distribution analysis technology developed in recent years. It can realize the rapid and high-resolution distribution analysis of elements in a wide range of materials. At present, it has been widely used in geology, archaeology, biology and other fields. However, there are still some difficulties in the quantitative distribution and characterization of complex block metal composition, which has not yet been applied in the powder metallurgy industry. In this study, the fluorescence spectrum behavior of each the element in superalloy was studied. The quantitative model of element was corrected by the type matching bulk standard sample of superalloy. The quantitative analysis method of composition distribution of Superalloy based on μ-XRF was established, which met the needs of powder metallurgy industry for fine quantitative characterization of powder boundary composition distribution on a large scale. The SPS powder superalloy samples treated by high purity cobalt alloying were taken as the research object. The quantitative statistical distribution of Ni, Co, Cr, Mo, W, Ta, Ti and Al in the powder sintered samples after different milling time were analyzed. The influence of different milling time on the composition distribution of the sintered samples was discussed. The results show that there are a large number of original particle boundaries, and the composition distribution is not uniform. The pure Co powder particles added by ball milling only exist in the outer layer of superalloy particles, resulting in the Co content at the edge of particles is significantly higher than that in the particle center. When the milling time is short, there are many Co enrichment areas at the boundary of the original particles. When the milling time is increased to 24 hours, due to the alloying of ultra-fine cobalt powder and superalloy in the process of mechanical mixing, the composition distribution uniformity of sintered samples is greatly improved, and the content of Co at the boundary of original particles decreases significantly, while the content of other elements increases. The results show that the diffusion of each element in this sample is obvious, which is helpful to the improvement of element segregation. Therefore, the preparation process of the powder metallurgy superalloy is improved. This method can also be used to characterize the composition distribution of other P/M industrial products and provide data support for the optimization of P/M process and the improvement of product quality.

The pollution of heavy metals in the soil will affect the quality of agricultural products, which could further influence human health. Multiple heavy metals are usually detected with chemical methods in soil, during the process, strong oxidizing materials would be used to digest the samples in the laboratory, and finally, the dissolved solutions are tested. X-ray fluorescence spectrometry could realize the rapid detection of multiple heavy metals in soil, but compared with chemical methods, which has a higher detection limit. For mercury, the national pollution limit is lower than other metals, which makes it difficult to detect rapidly with X-ray fluorescence spectrometry in low-content samples. In this paper, an enrichment device was designed to realize the enrichment of mercury in soil, after testing with the X-ray spectrometer, the rapid detection of mercury in soil could be realized, which met the requirement of the actual test. The soil samples that had been weighed accurately would be heated first, and in this process, the mercury would be desorbed, and at the same time, the filter membrane was used to adsorb it, so as to realize the enrichment of mercury. A mercury generator was used to provide the air with a certain amount of mercury, and different kinds of membranes would be used to study the effects of adsorption. The result found that carbon fiber filter membranes have a good effect on adsorption and could enrich mercury in the air. When different flow velocity was adopted with the same weight of the samples, and the desorption temperature was set up at 800 ℃, the adsorption behavior of two membranes was studied. The results showed that, with the increase of the flow velocity, the intensity of the first membrane decreased, but the intensity of the second membrane increased, which meant that lower flow velocity was a benefit for the adsorption of membranes. when the different amount of mercury contained in solution was added in high-purity silicon dioxide, after enriching and testing by the designed device, the working curve could be obtained with the linear correlation coefficient to be 0.998 5. And the detection limit and quantification limit could be calculated as 7.52 and 25.06 ng respectively when multiple high-purity silicon samples were tested. It meant that if the weight of the sample was 0.3 g, the quantification limit would be 0.083 mg·kg-1 in the soil. The relative deviations were no more than 11.1% for the national standard samples except one sample that was below the quantitative limit, which indicates that this method could realize the rapid detection of mercury in the soil for agricultural land.

As the most important items in the field of toy safety, 8 harmful elements (chromium, arsenic, selenium, cadmium, antimony, barium, mercury and lead) were determined by the wet chemical method of sample preparation and instrumental analysis in current most of standard methods, but the pre-treatment process is complicated, time-consuming and costly. EDXRF has the advantages of simple pre-processing, non-destructive, and high efficiency. It is often used widely for the screening of harmful elements in electronic and electrical products. However, due to the variety of elements and toys materials, the lower limits, and the overlapping interference between target elements, it is difficult to use EDXRF to test accurately 8 certain elements in toys up to now. A new rapid method for measuring the content of eight elements in toy plastics by EDXRF was established. The measurement conditions of target elements, such as analytical line, tube voltage, analytical time and sample thickness, were investigated. According to the characteristics of the target elements, 3 combinations of filter and tube voltage were selected finally. This method can complete the test of 8 elements within 135 s totally for one daily sample. The interference from spectra overlap was corrected by the interference coefficient method. The matrix effect was corrected by experience coefficients method, using rhodium scattered radiation as internal standard simultaneously. Polyethylene (PE) and polyvinyl chloride (PVC) were selected as the standard samples to make the working curve according to the characteristics of the common plastic sample matrix of toys. The calibration curve method was effective. For the target elements, the detection limits of this method were between 0.5 and 37 mg·kg-1. It can meet the screening limits requirements of most of the national standards of safety of toys. The accuracy and precision of the method were evaluated by several materials (PP, PE, ABS, PVC). The relative error (n=6) is within 25%, which indicates that the method has good accuracy. The RSD (n=6) is within 6%, except for some RSDs of individual elements in individual samples within 9%~15%, with good precision. If the test time is extended from the 30 s to 2 min, all RSDs can be reduced to less than 5%. The method can be applied to rapid screening detection of the eight harmful elements in all kinds of toy plastics.

Fishmeal is a kind of high protein feed material which plays an important role in aquaculture. There is a great market demand for fishmeal in China, but the quality of fishmeal from different places is different. In order to ensure the quality and safety of fishmeal, it is very important to establish a traceability system of fishmeal origin. The energy dispersion X-ray fluorescence spectrum is able to detect the type and content of mineral elements in the sample depending on the energy of the element’s radiation X-ray fluorescent photons. The types and contents of mineral elements contained in fishmeal may vary depending on the origin of fishmeal, so this paper proposes for the first time to use the energy dispersion X-ray fluorescence spectroscopy (EDXRF) method to scan the fishmeal to obtain the element information of fishmeal elements. After preprocessing the original spectrum, whale optimization algorithm is used to improve the adaptive net analyte signal weight K_lacal hyperplane method can identify the spectrum vector of fishmeal samples, and then identify the origin of fishmeal samples. Firstly, 51 fishmeal samples from Liaoning and Zhejiang were pressed, and different filters were set in the detection program of EDXRF spectrometer, 51 groups (6 spectra in each group) of spectra were obtained. Then the spectrum is preprocessed, and the baseline is corrected based on the adaptive iterative reweighted penalty least squares algorithm (airPLS), so as to eliminate the impact of baseline drift and improve the accuracy. Wavelet transform is used to smooth the spectrum and remove the high-frequency noise of the spectrum curve. The 16-dimensional vector representing the element content of each fishmeal sample was obtained by calculating the peak area of six effective spectral regions. Finally, whale optimization algorithm is used to select the key parameters (neighbor number, principal component fraction, adjustment parameter) of the adaptive net analyte signal weight K_local hyperplane (ANWKH) method, and then the adaptive net analyte signal weight K_local hyperplane model is established by using the found optimal parameters. 70% of the fishmeal samples from each place of origin are selected as the training set, 30% as the test set to identify the fishmeal place of origin. Fishmeal samples are from Liaoning and Zhejiang provinces. The accuracy of the prediction model is 94.3% and 100% respectively. The total accuracy is 97.3%, which is higher than the accuracy of the adaptive net analyte signal weight K_local hyperplane classification. The results show that the method based on energy dispersive X-ray fluorescence spectrum can accurately realize the origin traceability of fishmeal, and the adaptive net analyte signal weight K_local hyperplane method improved by whale optimization algorithm can find the optimal parameters and establish a model with higher classification accuracy. This paper provides a reference for more detailed origin traceability of fishmeal at home and abroad in the future.

-ray fluorescence spectrum micro-area analysis method has the characteristics of rapid, simple and non-destructive testing, and can detect the distribution of elements on the surface of health food. Inductively coupled plasma mass spectrometry has the advantages of low detection limit, wide linear range and multi-element simultaneous determination. Methods: A health food was analyzed by X-ray fluorescence spectrum micro-region analysis technique, and the elemental content was semi-quantitatively analyzed to determine the elements of calcium (Ca), iron (Fe), ruthenium (Ru) and molybdenum (Mo) in health products. The main elements of Ca and Fe were quantitatively analyzed by microwave digestion-inductively coupled plasma mass spectrometry (ICP-MS). The average value of Ca element was 6.23%, the relative standard deviation was 1.78%, the average value of the Fe element was 3.82%, and the relative standard deviation was 2.14%. The results were consistent with the conclusion that the content of Ca in the sample with X-ray fluorescence spectrum micro-analysis was between 6.0% and 10.0%, and the content of Fe in the sample was between 2.0% and 4.0%. The distribution of the elements on the surface of health food was not uniform by X-ray fluorescence spectrum analysis. The X-ray fluorescence spectrum micro-area analysis method cannot only measure the elemental semi-quantitative content in health food by using the trace sample, but also measure the elemental distribution in health food. Combined with inductively coupled plasma mass spectrometry, the content of the elements of interest in the sample can be further detected, and the quantitative analysis results can be obtained.

Homogeneity is one of the three characteristics of reference materials. X-Ray Fluorescence Spectrometry has lots of strengths, like simplicity, high precision, and multiple elements simultaneous analysis. So X-Ray Fluorescence Spectrometry is one of the important methods to test the homogeneity of reference materials. At present, the application of X-Ray Fluorescence spectrometry to test the homogeneity of reference materials is still controversial. The homogeneity test requires that the sample weight is the minimum, and the general geochemical reference materials minimum sample weight is 0.1 g. The sample weight is about 4 g when the XRF is used for the homogeneity test; the obtained result is theoretically insufficient to support whether the sample is homogeneous under the condition of the minimum sampling weight. Based on the optimization of instrument parameters, this study changed the size of the previous pelleting mold to make the sampling weight 0.1 g, using pressed powder pellets, selected three soil (GBW07425, GBW07428, GBW07388) and three stream sediment (GBW07375, GBW07378, GBW07379) certified reference materials, each reference material taken from 15 bottles, sampling 2 from each bottle, the number of the total samples is 30, choosing SiO2, Al2O3, Fe2O3, MgO, CaO, Na2O, K2O, Mn, Ti and P 10 components to test. According to the single factor analysis of variance of F values and measured values of the standard deviation (S) and the relative standard deviation (RSD) comprehensive to evaluate the homogeneity of reference materials. Through theoretical calculation showed that the minimum sample weight of 10 main components was less than 0.1 g under the condition that the radiation radius of the sample was 5 mm. The results with 0.1 g pressed powder pellets showed that the accuracy of the method in this study was high, the relative errors were less than 16%, the precision was high, the relative standard deviations were no more than 4.3%, the F values were less than the critical values, and the homogeneity of soil and sediment certified reference materials was good. X-Ray Fluorescence Spectrometry is applied for determination under the condition that the minimum sampling amount is 0.1 g, which can not only solve the long-standing dispute in the homogeneity test of geological reference materials but also provide technical support for the application of X-Ray Fluorescence Spectrometry in other fields.

Under the background of low counting rate, the high-precision measurement of X-ray spectrum is affected by the statistical fluctuation of X-ray flow, which determines the theoretical limit of given detector energy resolution, while the influence of other factors can be reduced by appropriate noise filtering and electronic technology. Previous studies on energy resolution mostly use spectral deconvolution to post-process the energy spectrum, so as to reduce the full width at half maxima (FWHM) of the characteristic peak. These post-processing methods are based on modeling the obtained energy spectrum as two random variables, i. e. input energy spectrum and detector response function, which is often computationally expensive and inefficient. A multi-pulse local average (MPLA) algorithm is proposed to optimize the X-ray spectrum data processing platform, which is an online real-time spectrum acquisition method. This method averages the pulse amplitude value in the dynamic window. MPLA algorithm involves two variable parameters; one is the average window size r, the other is the average pulse amplitude number n. The implementation process of the algorithm includes the following four steps: step 1, read the first pulse amplitude and locate an average window, update the current average window amplitude and pulse number after reading successfully; step 2, read the next pulse amplitude, judge the number of pulses in the average window after each update, and continue the third step when it is less than the preset parameter n, otherwise. Then perform step 4; step 3, continue to read the next pulse amplitude; step 4, average the pulse amplitude in the corresponding average window, and the average is the channel address to be updated and counted, and then clear the pulse amplitude and pulse number in the average window. In the part of theoretical derivation, this paper studies the transformation of the original probability density function (PDF) in the application of MPLA process, deduces the analytic expression of the probability density function obtained after the application of MPLA and proves the following characteristics after the transformation of MPLA probability density: (1) symmetrical distribution, MPLA retains the mean value and symmetry. (2) For single peak symmetrical distribution, MPLA reduces variance and sharpens distribution peak. In the experiment, the iron ore sample is used as the measurement object, and the results processed by the MPLA algorithm are compared with the results obtained by the traditional spectral method. The results show that in the typical case of the spectrum peak with normal distribution PDF, even if only two pulse heights are averaged, the FWHM of the transformed peak is narrowed.

In the measurement of the X-ray fluorescence spectrum, when the interval between the two cases is short, there is a rising or falling edge pulse pile-up. If the pulse resolution of the spectrometer is not enough, the accidental coincidence effect occurs when the event interval is less than the pulse resolution of the spectrometer. When the pulse pile-up occurs on the rising edge of the signal, the next stage electronics is hard to identify the pile-up pulses and regards them as a single pulse, leading to the coincidence peaks and incorrect spectrum measurement. When the pulse pile-up appears at the falling edge, and the pile-up pulse interval is shorter than the digital shaping time of the multichannel pulse amplitude analyzer, the pile-up pulse will be discarded and lead to the reduction of spectrum count and cause a low throughput rate, which has a negative effect on the precision of radioactivity measurement. In this paper, increasing the signal-to-noise ratio of the analog circuit, reducing the false trigger and shortening the forming time of the fast forming channel, so as to improve the pulse resolution ability of the spectrometer and reduce the coincidence effect, a digital spectrometer of the fast and slow dual forming channel with low accidental coincidence effect is developed. The energy spectrometer has designed a fast channel with high pulse resolution. Based on the symmetrical zero area trapezoid forming algorithm, it can effectively eliminate the shortcomings caused by the narrowing of fast channel time, combined with the judgment of the trapezoidal flat top, the low-frequency noise suppression and error reduction are realized trigger probability. At the same time, the analog circuit with high signal-to-noise ratio and low noise is designed to reduce the probability of false noise triggering and an accidental coincidence of fast channel. In this paper, firstly, the fast channel time enhancement ability is verified by simulation, and then Cu samples were excited through the X-ray tube created by Moxtek to obtain characteristic X-rays, and the signal is detected by the high-resolution SDD detector of KETEK. The tube current of the X-ray tube was adjusted to acquire X-ray fluorescence spectra with a count rate ranging from 13 to 103 kcps and determine the relationship between the probability of accidental coincidence and count rate. Then, the influence of the shaping time change of the fast channel on accidental coincidence was analyzed. Experiments showed that a short fast-channel shaping time resulted in high pulse pair resolution and low accidental coincidence effect. At a 103 kcps count rate, under the fast-channel shaping time of 150 ns, the accidental coincidence probabilities of the Kα, Kβ, and Kα+Kβ combination peaks of Cu were 1.568%, 0.265%, and 0.403%, respectively. Under the same fast-channel shaping time, the accidental coincidence probability of the proposed digitized digital spectrometer was 60% lower than that of the DP-5 digital spectrometer produced by Amptek.

Current evaluation methods for spectral similarity are mainly based on the shapes and amplitudes of spectra, but these two can only reflect the outline information of spectra, and cannot well reflect the fingerprint characteristics of spectra of ground objects. In order to better embody the application of spectral characteristics in the evaluation, it is proposed herein a method for evaluating spectral similarity based on first-order gradient information. Firstly, it is proposed an MSAM similarity evaluation method, further, a modified gradient spectral angle matching (MGSAM) method by adjusting the traditional spectral angle similarity evaluation method SAM. MGSAM compares the gradient angle matching degree of the two spectral curves. The gradient information of the spectral curves can highlight the existence of “fingerprint” characteristics such as spectral absorption peaks, so MGSAM can fully reflect the similarity of the spectral characteristics of the two contrast curves. By analyzing the influence of offset information and spectral depth on MSAM and MGSAM, it is pointed out that MGSAM has stronger robustness to offset information, and can objectively reflect the difference in spectral depth, so as to directly reflect the fidelity of spectral features in the photoelectric systems or related algorithms. By applying MGSAM as the evaluation method to the evaluation of compressed sensing imaging system, the simulation results showed that as the change of sampling rate, the MSAM values ranged between 0.998～1, while MGSAM values ranged between 0.72～1, with obvious change and great difference. It objectively reflects the fidelity ability of the compressed sensing system for spectral features and has a stronger differentiation ability, thereby providing a more objective evaluation method for such systems. By applying MGSAM to the classification of ground objects based on spectral similarity, and selecting Salinas, Pavia and Indian Pines for the test data, the results showed that the average classification accuracy based on MSAM was 0.86, while that based on MGSAM was 0.93. This shows that MGSAM can highlight the role of spectral features in classification and greatly improve the classification accuracy.

Infrared spectrum analysis plays an important role in many fields such as natural science, engineering technology, and so on. With the continuous development of computer and artificial intelligence technology, higher requirements have been imposed on infrared/near-infrared spectral analysis. Based on artificial neural networks, the deep learning algorithm performs representation learning by extracting hierarchical features from data layer by layer. It has unique advantages in analyzing the details features of data. It has been successfully applied in many fields such as computer vision, speech recognition, and disease diagnosis. Although deep learning has achieved good results in the analysis of images, audio, and text data, its application in infrared/near-infrared spectral analysis is still very limited. A deep learning convolution operation method for infrared spectroscopic analysis is presented. Firstly, one-dimensional Fourier Transform Infrared Spectroscopy (FTIR) data are transformed into two-dimensional RGB color image data through Symmetrized Dot Patterns (SDP), and then, the transformed SDP color image data is fed into the VGG (Oxford Visual Geometry Group) deep convolutional neural network for deep learning to establish a classification and recognition model. By SDP transformation, the infrared spectra of sevensingle-component gases of different concentrations, including methane (CH4), ethane (C2H6), propane (C3H8), n-butane (C4H10), iso-butane (iso-C4H10), n-pentane (C5H12), iso-pentane (iso-C5H12), and its mixtures convert to 224×224 color images. The SDP transformed images show a significant difference in the distribution of the pattern points and are more in line with the data format of the VGG convolution operation. The SDP-VGG method is used to identify the methane concentration range in gas logging: the gas logging gas is a mixture of the above seven components of alkanes, and the concentration ranges of methane are divided into five categories: <20%, 20%～40%, 40%～60%, 60%～80%, and 80%～100%. The infrared spectra of different seven-component alkane mixed gas samples are collected by the infrared spectrometer in the wavenumber range of 4 000～400 cm-1 and scanning interval 12 nm. Without special pre-processing and feature extraction, 4 500 samples are used to establish the identification model of various methane concentration ranges by the SDP-VGG method. The recognition accuracy of the SDP-VGG model reached 91.2%, which is better than the recognition accuracy of 88.7% and 86.2% of the Support Vector Machine (SVM) and Random Forest (RF) models established by the same infrared spectral data. The research shows that SDP combined with deep learning can accurately extract the key features of infrared spectra. It is a more effective infrared spectral analysis method, which improves the recognition accuracy of the infrared spectrum and has broad application prospects.

Phase measurement sum-frequency vibration spectroscopy (SFG) can obtain molecular orientation information of the material surface, but there are still some key issues that remain unresolved, including experimental repeatability, experimental design, and interface analysis. Phase error can cause spectral changes and mislead interface structure analysis. Therefore, analyzing and accurately controlling errors is the key technique for phase measurement SFG. We used z-cut quartz as the phase standard, measured the sum frequency vibration spectrum of octadecyltrichlorosilane (OTS) modified on the fused silica substrate in the C—H vibration band, analyzed the phase spectrum of OTS. The results show that in the OTS imaginary spectrum, the two positive peaks at 2 878 and 2 936 cm-1 are the symmetrical vibration (CH3ss) and Fermi resonance (CH3FR) of the terminal CH3, and the negative peak at 2 960 cm-1 is anti-symmetric stretching vibration of CH3 (CH3as), and the spectral characteristics and designation of these three peaks are consistent with the literature. The negative peak near 2 910 cm-1 is CH2 anti-symmetrical stretching (CH2as). Compared with the literature, there is an offset of about 20 cm-1, and a negative peak is also observed near 2 850 cm-1, which belongs to CH2 symmetrical stretching (CH2ss). We think that the difference from the literature may be due to the effect of sample preparation time on the molecular arrangement of OTS. By establishing the relationship between the imaginary part spectrum of OTS and the orientation angle of CH3, it is found that the angle between the c-axis and the surface normal of the three vibration modes of CH3 is less than 90°, and its H is more oriented upward and arranged in order, indicating that the phase measurement can obtain richer surface information, compared with the intensity measurement. At the same time, the influence of the inconsistency of the position of the test sample and the reference sample on the phase measurement accuracy is discussed. By measuring the imaginary part spectrum of OTS at three different positions (12.1, 12.3, and 12.73 mm), and compared with simulated phase error on the imaginary part spectrum, indicate that the 2.5 μm displacement between the measurement position of the test sample and the reference sample corresponds to 1 ° phase error the phase shift of 20° will cause the zero position to move about 6 cm-1, which causes changes in the position and sign of the vibration peaks, leading to incorrect interpretation of the spectrum. Therefore, in order to obtain stable and reliable phase information of the interfacial molecules, it is necessary to strictly control that the measurement positions of the two samples are consistent.The results of this experimental study provide guidance for improving the accuracy of phase measurements and provide an effective means for the detection and analysis of surface states of molecular molecules, including the detection of small signals.

To identify the types of binder used in the theater color painting of Jiayuguan pass, the Fourier infrared spectra were collected, which of the reference samples consist of leather glue, fish glue, egg white, egg yolk, and casein prepared by traditional technology and three cultural relics samples. The factor analysis combined with the linear discriminant analysis (PCA-LDA) was used to construct the mathematics model and determine the kind of the binder of relics samples. It was found that there was abundant information about the molecular structure in the range of 1 800~1 000 cm-1 of the reference samples infrared spectra. The common infrared spectrum characteristics of protein compounds in this range are the peak of CO bond stretching vibration around 1 650 cm-1, the peak of C—N bond stretching vibration and N—H bond bending vibration around 1 542 cm-1, and the peak of C—N bond stretching vibration around 1 240 cm-1. Besides, due to the fatty substances contained in yolk, skin glue and milk reference samples, there is also a stretching vibration peak of saturated fatty acid ester carbonyl CO bond near 1 745 cm-1. Through the analysis of factor score scatter diagram, it could be seen that there are differences among different reference samples. According to this, factor scores of the reference sample infrared spectrum were used as the training database to calculate the discriminant function by LDA. The group center figures, and cross-validation of the functions were performed and the accuracy of the discriminant equation is 93.3%. Base on the spectrum of three relics samples, which still have the characteristics of protein binder, the spectrum difference between them and reference can be seen due to degradation. PCA-LDA analysis model was used to identify the kinds of the binder for cultural relics, and the results were all leather glue. According to the analysis above, it could be concluded: A stable and effective infrared spectrum discrimination PCA-LDA model can be established for the discrimination of binder. The theater color painting binder used in the Jiayuguan pass was identified as leather glue by this model.

Tension wood is produced on the upper sides of the inclined trunk or branch of hardwood when its orientation is shifted from the vertical. Unlike the opposite wood formed on the lower sides, tension wood fiber is characterized by the presence of a specific layer, called the gelatinous layer which displays various physical and chemical properties. In this study, TEM imaging was used to reveal the variation in the cell wall layering structure between the Populus nigra tension wood and opposite wood. Furthermore, confocal Raman microscopy with 532 nm exciting laser (spatial resolution is about 0.5 μm) was used to visualize the variation in the distribution of fiber wall components and porosity between tension wood and opposite wood, meanwhile the topochemical correlation was innovatively revealed by Raman overlaid image. When integrating over the Raman band at 2 942 cm-1 (Cellulose, hemicelluloses and lignin C—H stretching vibration), the sublayers of tension wood and opposite wood were successfully distinguished. By integrating over the band at 1 094, 1 598 and 904 cm-1 in the normalized average Raman spectra, the distribution of cellulose, lignin and xylan was visualized. Lignin-cellulose and lignin-xylan overlaid Raman images displayed that for the tension wood cellulose mainly existed in the gelatinous layer, and the concentration of cellulose and xylan within the cell wall regions was higher than that of opposite wood. Specifically, the secondary wall of tension wood had an increased lignin concentration. Double-wall line scan showed that the distribution of lignin, cellulose and xylan was highly regional dependence and displayed gradient changes along the adjacent cell wall. Moreover, line scan analysis revealed the variation in D2O concentration along the adjacent fiber secondary wall and confirmed the more abundant porosity distribution of the gelatinous layer compared with secondary wall and middle lamella. The above results are helpful to understand the forming mechanism of unique physical and chemical properties of tension wood and expand the application of micro Raman-spectroscopy technology in the field of plant cell wall pore structure research.

β-gallate type compounds are promising solid-state ionic conductor, which has important application value in the field of energy storage. These compounds exhibit complex lattice dynamics due to the conducting layer tends to have an excess of alkali metal ions, which makes it difficult to further understand its conductive mechanism. Both pressure and temperature can affect the structure of materials by changing the spacing between atoms, and it has great application value in studying the dynamic process of materials, especially the diffusion process of ions. So far, the temperature dependence of vibrational properties has received less attention, and the high-pressure behavior of β-gallate type compounds has not been reported. Due to the unique advantage of laser Raman scattering technique in studying the lattice dynamics of matter, especially the pressure and temperature-dependent Raman spectroscopy, it is an effective experimental method for studying the lattice dynamics of the β-gallate type compounds. In this work, a novel β-gallate type K0.294Ga1.969O3 (KGO) crystal was successfully synthesized by using large-volume-press technology. The crystal was characterized by a scanning electron microscope, energy spectrum. The crystal structure of KGO is analyzed by single-crystal X-ray diffraction and compared with the crystal structure of β-Ga2O3. The lattice dynamics of disordered alkali metal ions in the KGO conducting layer was studied by pressure and temperature dependent Raman spectroscopy. We found that the β-gallate type KGO crystal structure formed by alternatively stacked-layer spinel-blocks and the loose conducting plane remains stable at the pressure up to 23.3 GPa. The significant difference in the pressure coefficients between high and low-frequency Raman modes are derived from different types of vibration. It is evidenced that the presence of thermally activated processes K+ ions in KGO at approximately 300 ℃, it’s embodied in the intensity of low-frequency Raman mode related to alkali metal K+ motion increases rapidly, while that of high-frequency vibration mode related to Ga-O polyhedron increases slowly. And the mobile K+ ions undergo disorder diffusion process along the conduction plane. Our results will contribute to a deeper understanding of the conductive mechanism of β-gallate type compounds, and it is also very important to achieve accurate compositional control and doping of β-gallate type compounds.

In order to identify the ingredient in Melaleuca alternifolia oil cells at room temperature, and avoid sample pretreatment and extractions, which can be labour-intensive. Oil Cells Distribution on Different Parts of Melaleuca alternifolia and the principal component in oil cells have been studied by Micro-Raman Spectrometer. It has been found that there are barely any oil cells on the soft branches. There are more oil cells on new leaves compared to mature leaves. In Raman spectroscopy of mature leaf of oil cells, 1 675/726 cm-1(CC stretching/ring deformation) are a characteristic key band of terpinene-4-ol, 1 700/754 cm-1(CC stretching/ ring deformation) for γ-terpinene, 1 609 cm-1(CC stretching)for α- terpinene and 1 522/1 156/1 011 cm-1(CC stretching/C—C stretching/C—C in-plane rocking)for β-carotene. In Raman spectroscopy of new leaves oil cells, 745 cm-1 (ring deformation) is a characteristic key band of cis-sabinene hydrate, 1 609 cm-1 for α- terpinene and 1 522/1 160/1 008 cm-1 for β-carotene. The principal components are different in oil cells of new and mature leaves. It is the first report that there are cis-sabinene hydrate and β-carotene in oil cells of Melaleuca alternifolia. This method can be used for quality control and developmental research for Melaleuca alternifolia plant essential oil extraction.

For marine diesel engines, lubricating oil is often contaminated by the coolant, resulting in the deterioration of lubricating oil, further leading to its functional failure. The main components of the coolant are water, ethylene glycol, and a small number of additives such as anti-corrosion, anti-cavitation, and defoaming. The application of Raman spectrum to detect the concentration of coolant contaminating lubricating oil is a kind of Raman spectrum detection problem for complex mixtures. The quantitative analysis method of single Raman peak strength cannot meet the quantitative detection of concentration. Therefore, Raman spectral analysis and LSTM neural network data mining are applied to lubricant coolant contamination. Under laboratory conditions, diesel oil samples with coolant contamination concentrations of 2%, 1.5%, 1%, 0.5%, 0.25% and 0% were prepared. Each oil sample was analyzed by Raman spectroscopy for 50 times, and a total of 300 Raman spectral data were obtained. 80% of the data were randomly selected as neural network training samples, and the remaining data were taken as test samples. The wavenumber of Raman spectral sample data was 300～2 000 cm-1. Data preprocessing, including sampling, fitting, discrete point average gradient estimation. The training sample set was constructed, and the LSTM neural network was combined with multi-layer full connection layer (FC) to establish four different neural network model structures, including FCs, LSTM-FCs-1, LSTM-FCs-2, andLSTM-FCs-3. The average error curves and detection accuracy curves of the four networks on the training set and test set are obtained. The results showed that the accuracy of FCs, LSTM-FCs-1, LSTM-FCs-2, and LSTM-FCs-3 neural network models was 96.7%, 93.3%, 98.3% and 83.3%, respectively. In order to study the robustness of the four models, the detection accuracy of the four neural network models was analyzed by selecting any wavenumber of 1% and adding noise whose amplitude changed by 1% randomly. The results were 88.3%, 90.0%, 96.7% and 78.3%, respectively. It can be seen that compared with the other three neural network structural models, LSTM-FCs-2 model is more suitable for quantitative estimation of lubricant coolant contamination, and its highest accuracy can still reach 96.7% after adding noise, and its robustness is better than the other three models. Raman spectroscopy combined with the LSTM-FCs-2 model in the LSTM network was applied to the sample of lubricating oil in use with 0.2% and 0.4% coolant contamination concentrations, respectively, with relative errors of 5.0% and 7.5%. It shows that this method can be used to detect the concentration of used lubricating oil contaminated by the coolant.

Oil-polluted water is becoming more and more widespread, which poses a serious threat to the human and ecological environment. Rapid, accurate and reliable monitoring of oil pollution in water is essential for understanding its environmental behavior and assessing human exposure risk. Petroleum components often have better spectral responses, but there are few reports on the rapid monitoring of water contaminated by petroleum hydrocarbons based on spectral technology. This study focuses on the groundwater and surface water of a typical petroleum-contaminated site. Use standard methods to obtain sample conductivity, total organic carbon and Cl-, NO-3, SO2-4, Na+, K+, Mg2+, Ca2+, NH+4, volatile organic compounds and petroleum hydrocarbons C6-C9, C10-C14, C15-C28, C29-C40 concentration, and the samples were characterized by ultraviolet-visible spectroscopy, synchronous fluorescence spectroscopy and three-dimensional fluorescence spectroscopy, and use multivariate data analysis methods to evaluate the possibility of rapid identification and application of spectroscopy technology in petroleum contaminated sites. The results showed that: ① Ultraviolet-visible parameters and synchronous fluorescence parameters indicate that the molecular structure of organic substances in contaminated submerged water is complex, contains a large number of aromatic compounds and contains a large number of substituents such as hydroxyl, carbonyl, carboxyl and esters on the organic substances. Three-dimensional fluorescence parameters indicate that the organic substances in contaminated submerged water have undergone a long period of biological transformation, indicating that the organic substances in contaminated submerged water have strong stability and poor. ② The three-dimensional fluorescence spectra of samples contaminated by benzene series showed obvious fluorescence peaks in areas Ⅰ and Ⅳ, and shoulder peaks in area Ⅴ. The three-dimensional fluorescence spectra of samples contaminated mainly by naphthalene series showed obvious fluorescence peaks in areas Ⅱ and Ⅳ, and shoulder peaks in area Ⅰ. The three-dimensional fluorescence spectra of samples contaminated mainly by naphthalene series and phenanthrene series showed the highest fluorescence peak in zone II, and shoulder peaks existed in Ⅰ, Ⅲ, Ⅳ and Ⅴ; ③ The concentration of C6-C9 components of petroleum hydrocarbons can be rapidly indicated by the ultraviolet-visible parameters S308～363, SUVA254 and the volume of zone Ⅰ in the three-dimensional fluorescence spectrum, the concentration of C10-C14 and total petroleum hydrocarbons (TPH) can be rapidly indicated by the volume of zone Ⅳ in the three-dimensional fluorescence spectrum, and the concentration of C15-C28 and C29-C40 can be rapidly indicated by the volume of zone Ⅰ in the three-dimensional fluorescence spectrum. Spectral parameters combined with multivariate data analysis can be used to quickly identify petroleum-contaminated water bodies, providing a new fast on-line monitoring and analysis method for groundwater petroleum pollution monitoring and remediation.

Grapevine downy mildew is the most serious grape disease worldwide. Early detection of this disease can achieve early control so that quality and yield are improved. A test method based on multicolor fluorescence images (MFI) on grape leaves and a Support Vector Machine (SVM) model was proposed in the current study. Multicolor fluorescence imaging was performed on 145 inoculated leaves and 145 healthy leaves from the backside at six consecutive DPI (Days Post Innoculation). 16 fluorescence parameters (F440, F520, F690, F740 and their respective ratios) were obtained. Based on the image variation of four independent fluorescence wavelengths as DPI proceeding, single-factor ANOVA and correlation analysis were conducted. Four wavelengths of F520, F690, F440/F740 and F690/F740 were best selected with stronger detection ability of early infection and low correlation. For better detection, an SVM model was constructed with all four features. The results showed that the four basic bands F440, F520, F690, F740 and their ratios had the ability to detect early infection of grapevine downy mildew. F440 and F520 were more sensitive to the infection than F690 and F740. Start from 2 DPI, the area of the lesion could be highlighted in the fluorescence images of F440 and F520, at which the fluorescence intensity of the inoculated leaves was significantly higher than that of healthy leaves (p<0.01), and the difference increased with the increase of DPI (p<0.000 1). At F690 and F740 bands, the fluorescence intensity of inoculated leaves decreased gradually with the increase of DPI, and there was no significant difference between inoculated and healthy leaves from 1DPI to 3DPI. At 4DPI, inoculated leaves’ fluorescence intensity was significantly lower than that of heathy leaves (p<0.05) and the difference increased at 5DPI and 6DPI(p<0.01). The fluorescence parameters of healthy leaves changed little. F440 was the most susceptible to interference with the maximum coefficient of variation among the four bands, while F520 was more stable with the least coefficient of variation. With the increase of DPI, the detection accuracy of SVM model for distinguishing healthy and inoculated leaves was gradually improved, at 1DPI, the accuracy of SVM with multi-features was 65.6%, the accuracy of 3DPI achieved 82.2%, and the average accuracy was 84.6% in the whole experimental period (6 d), which was better than the best threshold method (F520 with 61.1% at 1DPI, 78.9% at 3DPI and 80.0% in the whole experimental period). In conclusion, the MFI technology with SVM model can achieve the early detection of downy mildew before the onset of symptoms, which provides a theoretical basis and proof for the development of portable equipment for early diagnosis of grape downy mildew.

Terahertz time-domain spectroscopy (THz-TDS) has been applied in the detection of skin cancer, skin burn, scar treatment, and the THz spectral parameters in the time domain and frequency domain are used to discriminate the different tissues. In the general reflection THz in vivo measurement, the skin should be placed on the top surface of a medium window, resulting in water content change in skin surface because of occlusion, and finally, interfere the accuracy of the measurement. THz biomedical application is transferring from ex vivo to in vivo, the THz spectral parameters changing should be analyzed when measuring the occluded skin. In this paper, the occlusion process is measured using THz reflection system, and 13 feature parameters of measured THz signals such as peak to a peak value and Full width at half maximum (FWHM) are proposed and analyzed. Results show that the peak to the peak value of time-domain signals and transfer function and fitting slope of maximum and minimum value decay exponentially over occluding time while fitting slope and time distance of maximum and minimum value of transfer function increase exponentially over occluding time. The FWHM and log spectrum remain stable along with the occluding time. Afterwards, the double Debye model is used to describe the dielectric constant of skin in 0.2~1 THz frequency, and the combined genetic algorithm and Levenberg-Marquardt optimization method are used to extract the Debye parameters at a different occluding time. Results show that ε∞ and εs both increase exponentially with an increase of 27.8% and 12.5% respectively in 5 minutes, while the ε2, τ1 and τ2 remain stable over the occluding time. Next, the skin is taken regard as a stratified medium, based on the Bruggeman effective medium theory, the previous optimization algorithm which takes the measured reflectivity and calculated reflectivity as the objective function is also used to extract the skin water content along with occluding time. Results show that water content in stratum corneum grows exponentially with occluding time and increases by 23.8% in 5 minutes. Consequently, the THz spectral change results from occlusion of the skin due to the contact of the medium window should be carefully considered when applying the THz-TDs in clinical application. Our research could improve the accuracy of THz in vivo detection and promote its clinical application.

Solution cathode glow discharge-atomic emission spectroscopy is a novel technology for metal ion detection in the aqueous solution that has emerged in recent years. It has the remarkable characteristics of rapid, real-time and low-cost detection. In this paper, an industrial injection pump was used to realize the flow injection analysis for the solution cathode glow discharge excitation source, and the narrow-band filter was used to extract the metal element spectral signals of Na, K, Ca, Li, Sr, and Cs, and the photomultiplier and ammeter were employed to obtain optical signals to electric signals, to realize the detection of metal elements in aqueous solution. The effects of injection volumes of 100 and 166 μL on the signal intensity of 1 mg·L-1 Na element was analyzed. The relative standard deviations (RSD) of the signal intensity are 4.64% and 1.95%, respectively, indicating that both injections had good repeatability. In order to obtain better analysis performance, the influence of parameters such as DC discharge voltage, slit width and photomultiplier tube supply voltage on signal strength were explored. The experimental results show that a high signal-to-back ratio is obtained when the DC discharge voltage is 1 000 V, the slit width is 70 μm, and the photomultiplier tube supply voltage is -800 V. the detection limits of six metal elements Na, K, Ca, Li, Sr and Cs were measured when the equipment works at flow injection mode, which was 2.78, 4.23, 589, 9.45, 981 and 83.6 μg·L-1. Na and K elements in the mixed solution were quantitatively analyzed and measured, the measurement errors were 7.5% and 6.67%, and the precisions were 1.24% and 0.89%, respectively, indicating that the homemade equipment can detect metal elements with high accuracy.

In order to control the content of trans fatty acids (TFAs) in the process of oil deodorization, this paper presents a fast method for detecting trans fatty acids (TFAs) content in soybean oil based on near-infrared spectroscopy. First, we prepared 100 soybean oil samples with different TFAs content, and detected precisely the values of TFAs contents by gas chromatography. Then, the near-infrared spectrum of oil samples was scanned and denoised by various methods, and it is found that the denoising effect of MSC was the best. In order to study the characteristic absorption of TFAs in near-infrared region, we used a variety of iPLS methods to select the characteristic band of the spectral data, and the characteristic absorption band of TFAs is selected as 7 258～7 443/6 502～6 691/6 120～6 309 cm-1. On this basis, the Kalman filtering algorithm is used to select the characteristic wavelength variables, and 27 TFAs characteristic wavelength variables are optimized. The deep belief network (DBN) is adopted to construct the correction model, and we found that the performance of the DBN model is the best adopting 3 hidden layers and 50-35-90 hidden layer nodes. Finally, the DBN model with this parameter is compared with the regression model of trans fatty acid content established by PLS. The results show that: when we used the whole denoised spectrum to construct model, the prediction effect of DBN is better than that of PLS, R2 is 0.879 4, RMSEP is 0.060 3 and RSD is 2.18%. When we used the selected characteristic band to model, the prediction effect of the PLS model is better than that of the DBN model. Using the 27 optimized characteristic wavelength variables to construct model, DBN has a good prediction effect, R2 is 0.958 4, RMSEP is 0.035 0 and RSD is 1.31%. It shows that the generalization ability of DBN is better, which achieved better prediction results by using a small number of wavelength variables. The proposed method in this paper can meet the practical needs, and provide technical support for online detecting and regulating TFAs content and producing low/zero TFAs oil products.

The spectral irradiance (radiance) responsivity of the detector is one of the most important parameters. Traditional spectral calibration is realized by using a broadband lamp and monochromator. The newly built facility using a laser and detector to calibrate the spectral responsivity, with the measurement uncertainty dramatically reduced. First, the facility couples tunable laser into an integrating sphere to generate a laser Lambert source. Then using standard trap detectors which can be traced back to cryogenic radiometer and an aperture with the known area, the spectral irradiance responsivity from 400 to 900 nm can be calibrated. The research is focused on four aspects: (1) Since the standard detector is calibrated at separate laser wavelengths by the cryogenic radiometer, the spectral responsivity at any other wavelengths must be interpolated. Compared with the direct measurement method using the spectral responsivity facility, the results show that the quantum efficiency difference using numerical interpolation is less than 0.074% from 400 to 900 nm. (2) In order to reduce the influence due to the laser power drift, the integrated charge method and a monitor detector are used. The charges of the standard detector and the detector under test are measured with the standard deviation of the repetition smaller than 0.1%. (3) The laser power when the standard detector is traced to cryogenic radiometer is totally different from that when the standard detector is used to transfer the spectral irradiance responsivity. Beam addition method is adopted to test the linearity of the detectors. Measurement results show that the nonlinearity correction is less than 0.025% when the current of the detector is varied from 0.2 mA to 3 nA. (4) The responsivity uniformity of the detector is measured to evaluate the uncertainty due to the difference between irradiance mode and power mode. The nonuniformity in an area of 5 mm in diameter is less than 0.03%. The uncertainty of the spectral irradiance responsivity facility from 400 to 900 nm is 0.14%~0.074% (k=1). The spectral irradiance responsivity of the detector measured by the new facility and lamp-monochromator facility is compared. Results show that the two facilities show good agreement from 400 to 900 nm. The difference between the two methods is within the uncertainty of the lamp-monochromator facility.

Lake Qiandao has low primary productivity and high Secchi disk depth. It is of great importance to explore the bio-availability of chromophoric dissolved organic matter (CDOM) in Lake Qiandao to unravel the carbon cycling in the lake. By comparing the changes of CDOM absorption and fluorescence pre- and post-28 days of bio-incubation, we aimed to reveal the bio-availability characteristics of CDOM in Lake Qiandao. The results showed that the mean values of a254 and S275-295 decreased, and the mean value of humification index (HIX) increased after 28 days of biodegradation. This indicated that laboratory biodegradation resulted in a decreased CDOM concentration and increased aromaticity of CDOM molecules. CDOM absorption a254 decreased by a mean of 14.3%±4.8% and with a range from 4.3% to 23.6% after 28 days of microbial degradation across the sixty sampling sites. Three fluorescent components were obtained by coupling excitation-emission matrices (EEMs) and parallel factor (PARAFAC) analysis, including a terrestrial humic-like C1, and tryptophan-like C2 and C3. Tryptophan-like C2 and C3 decreased by 54.1%±18.2% and 53.2%±14.3%, followed by C1 (28.2%±9.1%), suggesting a relatively high CDOM bio-availability in Lake Qiandao. After 28 days of biodegradation, and the major fluorescence peak changed from tryptophan-like C2 to terrestrial humic-like C1, indicating that the bio-availability of the tryptophan-like component was higher than that of humic-like component, and the T peak was degraded and the A peak was retained during the 28 days of laboratory bio-incubation. High values of the difference of absorption coefficient between pre- and post-incubation, representing the bio-availability of CDOM, i. e. Δa254, were found in the downstream southeastern lake regions, and is similar to that of tryptophan-like C2 prior to incubation. This suggested that CDOM bio-availability was the highest in the southeastern lake areas. High values of C2 and C3 were found in the northwestern inflowing lake regions post 28 days of incubation, similar to that of C1 and a254 post 28 days of incubation, suggesting that there might be freshly production of protein-like substances during the bio-incubation. There were close correlations between C1 and a254 pre- and post-bio-incubation, indicating that terrestrial humic-like substances were bio-stable. After 28 days of bio-incubation, the high values of tryptophan-like components disappeared from the lake center and southeastern lake areas, indicating that microbial metabolism has an influence on the application of tryptophan-like fluorophores in point-source-pollution identification due to the longer water residence time of the lake area compared to the remaining lake regions.

The laboratory visible-near infrared (VIS-NIR) spectroscopy has been frequently used in quantifying soil components because it is effective, fast and nondestructive etc. The higher spectral resolution is the richer soil information we could obtain. However, hyperspectral data are red undant and should be preprocessed. The study of the effects of different spectral resolutions on the modeling of soil components is relatively inadequate. Taking advantage of the European Land Use/Cover Area Frame Statistical Survey (LUCAS) dataset having 19 036 soil samples, we investigate the effects of different spectral resolutions on modeling soil components: total soil nitrogen (N), organic carbon (OC), calcium carbonate (CaCO3), and clay. To achieve this, we took the partial least squares regression (PLS) method as the evaluation model and randomly chose 30% samples for independent verification. Firstly, the spectral data which have 4 200 bands with 0.5 nm spectral resolution were resampled to 2, 4, 8, …, 1 024 nm respectively using average reflection value by of uniform interval sampling. The results are as follows: (1) when the spectral resolution was decreased, the inversion accuracy of soil components showed a downward trend; (2) when the spectral resolution was higher than 64 nm, higher model validation accuracies were obtained for estimating the four selected soil components (R2>0.65, RPD>1.7); (3) the accuracy for CaCO3 and clay components was significantly reduced when the spectral resolution was lower than 128 nm; (4) of the four soil components, CaCO3 was the most sensitive to spectral resolution. It has higher accuracy (R2>0.86, RPD>2.72) at high spectral resolutions, but the accuracy reduced most rapidly as the spectral resolution decreases. Secondly, based on the spectral response functions for a group of common satellite sensors, the inversion performances of using GF2, S3A, L8, Aster, S3OLCI, and Modis spectral bands are summarized as follows: (1) all sensors achieved higher accuracy for soil N and OC even if GF2 has 4 different bands (R2=0.56; RPD=1.51); (2) a low accuracy was obtained for CaCO3 and clay; (3) besides the number of spectral bands, the band positions are also important and the sensors (S3A, L8, Aster, and MODIS) having bands in the spectral range 1 100～2 500 nm showed a stronger performance than the sensor (e. g. S3OLCI) without the corresponding bands. The results from this study provide a guiding reference for preprocessing hyperspectral data of soil, selecting suitable satellite data sources and designing new optical sensors for soil Vis-NIR spectroscopy.

The study on the valence state of selenium in soil and stream sediment contributes to understanding the migration and transformation of selenium (Se). At present, there are many methods on the research of valence state of Se extracted partly from soil and stream sediment, but how to determine the valence state of all Se is still a difficult problem and the difficulty lies in how to digest Se in soil and stream sediment completely with valence state unchanged. The research showed that Se(Ⅵ) could be reduced to Se(Ⅳ) by 6.0 mol·L-1 HCl and the valence state of Se(Ⅳ) and Se(Ⅵ) stayed stable in 1.2 mol·L-1 HCl solution for 48 h at room temperature. Se, in soil and stream sediment, is digested with HNO3+HF+HClO4 completely and the valence state of Se(Ⅳ) and Se(Ⅵ) was unchanged when the HClO4 heated to white smoke appeared, but after the HClO4 heated to dryness, Se(Ⅳ) will all be oxidized to Se(Ⅵ). The method to determine Se(Ⅳ) and Se(Ⅵ) in soil and stream sediment by hydride generation-atomic fluorescence spectroscopy(HG-AFS) was developed based on this research, and the samples were digested with HNO3+HF+HClO4 and heated to white smoke appeared, then stopped heating (to avoid local heating to dry). After the digestion, the samples which cool down to room temperature were dissolved with 1.2 mol·L-1 HCl, and Se(Ⅳ) can be detected by HG-AFS. The Se(Ⅵ) will all be reduced to Se(Ⅳ) by being heated in 6.0 mol·L-1 HCl solvent, and then total Se can be detected by HG-AFS, from the total quantities minus the Se(Ⅳ) concentration we obtained the Se(Ⅵ) concentration. The results show that Se in soil and stream sediment is digested completely and the valence state of Se(Ⅳ) and Se(Ⅵ) stayed stable. The detection limits of Se(Ⅳ) and total Se were 4.5 and 5.1 ng·g-1, respectively. The recovery rate of Se(Ⅳ) and Se(Ⅵ) was 102%～108% and 94%～104%.

In order to improve the spectral intensity and the signal-to-back ratio of the characteristic spectral lines, promote the application of LIBS technology in the detection of trace heavy metals in soil. The experimental parameters in the process of soil analysis were optimized, and the element of Cr was analyzed. The Nd∶YAG laser with an output wavelength of 1 064 nm, the pulse width of 10 ns and pulse frequency of 1~10 Hz was used as the light source to focus the pulse laser on the surface of soil samples to generate laser plasma. Experimental parameters such as laser excitation energy, sample distance from lens and spectrometer collection delay were optimized. Firstly, the spectral intensity and the signal-to-back ratio of the laser energy from 60 to 110 mJ were compared. It was found that the plasma radiation intensity rises first and decreases, and the best experimental results can be obtained when 90 mJ excitation energy is selected. Secondly, the variation of spectral intensity from 5 mm before coke to 5 mm after coke is compared. It was found that when the distance between the sample and the lens was 1 mm after the focus (i. e. the focus position was 121 mm), the characteristic spectral lines and the information to back ratio of Cr elements reached the best. Finally, the influence of the acquisition delay of the spectrometer on the spectral line strength and the signal-to-back ratio were analyzed. The results show that the influence trend of energy on plasma radiation intensity is roughly the same, and the experiment result is best when the collection delay is 1 000 ns. Under the optimum experimental conditions (that is, the laser energy 90 mJ, focus position 121 mm, the acquisition delay 1 000 ns), 12 soil samples containing heavy metal Cr were detected by spectroscopy. Meanwhile, in order to reduce the interference of the external environment, the average values of the spectra obtained from 10 laser ablation positions of the same sample were pretreated. Chromium (Ⅰ) 357.86 nm, chromium (Ⅰ) 425.44 nm and chromium (Ⅰ) 427.49 nm were selected as characteristic lines. The calibration curves of doping concentration and spectral intensity were established. The detection limits of the three lines were 74.62, 64.07 and 67.49 mg·kg-1, respectively. The goodness-of-fit values R2 were 0.98, 0.97 and 0.99, respectively. RMSE was 0.41, 0.33 and 0.35, respectively. At the same time, partial least square method and support vector machine algorithm are introduced to improve the accuracy of calibration model further. The results show that the optimization of experimental parameters improves the quantitative detection parameters of trace elements by LIBS technology. The optimal spectral intensity and signal-to-back ratio are obtained. Good experimental results are obtained by the Lorenz fitting calculation of calibration curve, which has important reference significance for the detection of trace heavy metal elements by LIBS technology.

Combined with X-ray fluorescence spectroscopy, a prediction model based on deep convolutional neural network regression is proposed to predict the content of heavy metal element Zn in soil. Related pretreatment of the original soil, and soil compaction by powder compaction method, and the soil spectrum was obtained by X-Ray-fluorescence (XRF). Compared with traditional detection methods, the XRF method has the advantages of fast detection speed, high accuracy, simple operation, non-destructive sample properties, and simultaneous detection of multiple heavy metal elements. Therefore, XRF is combined with deep convolutional neural networks to achieve Precise prediction of heavy metal element Zn content in soil. In the experiment, box plots were used to eliminate abnormal data in the X-ray fluorescence spectrum. Entropy weight method and multiple scattering correction were used to correct the sample box data. The Savitzky-Golay smooth denoising method and linear background method are used to preprocess the spectral data, which can effectively solve the problems of noise and baseline drift caused by the external environment and human factors. The obtained one-dimensional spectral data vector is processed by constructing a spectral data matrix, this method converts 5 sets of parallel spectral data vectors at the same concentration and the same water content into a two-dimensional spectral information matrix and uses this matrix as the input of the deep convolutional neural network prediction model to meet the operational requirements of the convolutional layer. The learning ability of the deep convolutional neural network prediction model is improved, and the training difficulty of the model is reduced. The deep convolutional neural network prediction model is built with 3 layers of convolutional layers and activated using the RELU activation function. The maximum pooling method is used to reduce the dimensionality of the data and increase the Dropout layer to prevent overfitting. The ADAM optimizer is used to optimize the prediction model. The prediction model uses the mean relative error (MRE), loss function (LOSS), and mean absolute error (MAE) to determine the optimal learning rate of the model is and the optimal number of iterations is 3 000. The prediction model of the deep convolutional neural network is compared with the BP prediction model, the ELM prediction model, and the PLS prediction model, Analyze and compare the prediction model with the mean square error (MSE), root mean square error (RMSE), and fitting coefficient (R2), The results show that in predicting the content of heavy metals in soil, the prediction model based on deep convolutional neural network is superior to the three prediction models of BP, ELM, and PLS, which improves the prediction accuracy.

Heavy metal (like Pb, Cd, Cu, et al) in soil has affected the human health for a long time. So, the detection and prevention of soil Pb content have been a hot topic at home and abroad. Traditional methods of soil heavy metal detection, such as atomic absorption spectrometry (AAS), X fluorescence spectrometry (XRFS), are high cost, complicated, time-consuming, cannot meet the requirements of rapid analysis, and is easy to form secondary pollution of samples. Laser-induced breakdown spectroscopy (LIBS), a typical atomic emission spectrum, is a combination of laser technology and spectroscopic technology. It is based on the analysis of characteristic spectral line information that is excited to emit atoms and ions in a substance, and then compositions of the substance were studied. LIBS technology can rapidly detect the composition and content of material elements in any state (solid, liquid and gaseous). It is regarded as an emerging technology in the field of future chemical detection and rapid green analysis. LIBS technology has the advantages of simple pre-processing (or no processing required) for samples, multi-element simultaneous analysis, long-distance measurement, and wide applicability. Based on those advantages, it is widely used in various fields and viewed as one of the research hotspots. Under the background of agricultural informatization, the elements of Pb in the soil will be considered as the research carriers. And laser-induced breakdown spectroscopy (LIBS) technique combined with theoretical analysis and mathematical modeling will be employed to accurately detect the contents of Pb content. Then, the univariate calibration curve methods were built to predict heavy metal Pb content. Firstly, 15 soil samples with known Pb concentration gradient were selected for analysis. Soil LIBS spectral data were pretreated with different pre-processing methods. Three models based on LIBS intensity, peak areas, Lorentz fit intensity after normalized corresponding Pb content was established and fitted to analyze Pb content in soil quantitatively. The results show that the R2 of soil Pb content prediction based on three calibration curve models are 0.918 0, 0.910 1 and 0.914 3, respectively. The results of the three calibration curve analysis methods are good. It indicates that LIBS combined with univariate calibration curve method showed high reliability in detecting Pb in soil. The research results provide the theoretical foundation for developing the diagnosis and prevention technology of heavy metal contamination in soil and offer technical support for scientific spraying and precision management in agricultural production.

The soil pH is the key factor affecting the transformation of soil nutrients and the soil fertility. The detection of pH value of soil by near-infrared spectroscopy can provide an important basis for the development and utilization of soil resources. As a typical algorithm of deep learning in artificial intelligence, the convolutional neural network can not only extract the characteristics of complex spectral data but also reduce the training parameters of the network and improve the efficiency of network operation, because its structure has the ability of “local perception, weight sharing”. In this paper, the convolution neural network is applied to the modeling and analysis of the near-infrared spectrum, and a soil pH prediction method based on convolution neural network and the near-infrared spectrum is proposed. The network is built by Python calling Tensor Flow toolkit, and its structure is composed of the input layer, convolution layer, pooling layer and full connection layer. The spectral sample dataset of mineral soils, collected from the Statistical Survey of Land Use and Coverage conducted by the European Statistical Office in 2008—2012, was employed as an object of study. In order to eliminate the baseline drift in the spectrum and improve the signal-to-noise ratio, the first derivative and Savitzky-Golay smoothing of the original visible near-infrared spectrum (400~2 500 nm) were carried out. In the model training process, 15 000 samples are randomly selected as the training set, and the remaining 2 272 samples are selected as the test set. The effects of the number of convolution layers and training iterations on the model performance are discussed. The ReLU activation function and Adam optimizer are used to prevent the gradient disappearance of the model and improve the stability of the model. Then, the goodness of fit of the model is analyzed and calculated, and finally, the network model is compared with the traditional BP and PLSR models. The experimental results show that when the number of iterations of the model is 2 500, and the number of convolution layers is 4, the model reaches the best performance, and the mean square error of the training set is reduced from 1.898 to 0.097; the goodness of fit of the test set is 0.909, which is 0.117 and 0.218 higher than BP and PLSR models respectively. The results indicate that convolution neural network can extract the internal characteristic information of soil near-infrared spectrum, so as to realize efficient and accurate prediction of soil pH on a large scale. CNNR model can provide guidance for crop planting and precision fertilization to achieve the goal of soil structure stability and sustainable development. The convolution neural network-based NIRS regression method can also be applied to other soil information research.

Rice blast is a worldwide destructive rice disease. It is of great significance for rice disease control and precision spraying to detect rice blast early and identify the severity of the disease. Based on field experiment and natural infection of rice blast, infected leaves and healthy leaves were collected in the early stage of leaf blast. Hyperspectral images in the spectral range of 400~1 000 nm were captured and the spectral data were extracted. Rice leaves will not immediately show lesions at the beginning of the disease, so it is impossible to identify and collect samples of infected leaves without lesions. In order to realize the early detection of infected leaves without visible lesion, this study proposed to take hyperspectral data of lesion-free areas adjacent to the lesioned areas on the infected leaves as level 1 samples. According to the area of the lesion, the samples were divided into four levels: level 0 (109 pieces) for healthy leaves, level 1 (116 pieces) for infected leaves without visible lesion, level 2 (107 pieces) for leaves with lesion area <10%, and level 3 (101 pieces) for leaves with lesion area <25%. Principal component analysis (PCA) and competitive adaptive reweighting sampling (CARS) were used to extract feature variables; PCA algorithm was used to reduce further the dimension of the bands extracted by CARS. The support vector machine (SVM), PCA4-SVM, PCA8-SVM, CARS-SVM and CARS-PCA-SVM models for early detection of rice blast were build based on the full spectral variables and extracted feature variables, respectively. In this study, all models had high detection accuracy for all levels of samples. Level 1 had good detection accuracy, similar to other levels. All models had an overall accuracy rate above 94.6%. The highest was the CARS-SVM model at 97.29%, and the CARS-PCA-SVM model at 96.61% was slightly lower, but its number of input variables was only 6, which was 71.43% less than that of 21 in the CARS-SVM model. It further reduced the complexity of CARS-SVM model and improved the operation speed. So, the comprehensive evaluation of CARS-PCA-SVM model was optimal, with the identification accuracy of 97.30%, 94.87%, 94.29% and 100.00% for each level, respectively. Therefore, it is feasible to use hyperspectral imaging technology to detect the early stage of rice blast. The results presented in this paper can provide new ideas for the detection of infected leaves without lesions at the beginning of rice blast, and provide a theoretical basis for the early control of rice blast, precision spraying of pesticide and the development of detection instruments.

Above-ground biomass (AGB) is an important index to evaluate crop growth and yield estimation, and plays an important role in guiding agricultural management. Therefore, the rapid and accurate acquisition of biomass information is of great significance for monitoring the growth status of potato and improving the yield. The hyperspectral images, measured plant height (H), above-ground biomass and three-dimensional coordinates of ground control point (GCP) were obtained in budding potato period, tuber formation period, tuber growth period, starch accumulation period and mature period. Firstly, based on UAV hyperspectral image and GCP to generate the DSM of the experimental field, the plant height (Hdsm) of potato was extracted by DSM. Then the first-order differential spectrum, vegetation index and green edge parameters are calculated using UAV hyperspectral images. Furthermore, the correlation between hyperspectral characteristic parameter (HCPs), green edge parameter (GEPs) and potato AGB was analyzed. The first seven hyperspectral characteristic parameters and the optimal green edge parameter (OGEPs) with good correlation with AGB were selected for each growth period. Finally, the AGB of different growth period was estimated by partial least square regression (PLSR) and random forest (RF) based on the combination of HCPs, HCPs and OGEPs, HCPs and OGEPs and Hdsm. The results show that: (1) the Hdsm is highly fitted to H (R2=0.84, RMSE=6.85 cm, NRMSE=15.67%). (2) The optimal green edge parameters obtained in each growth period are not completely the same. The OGEPs of the budding period, the tuber growth period and the starch accumulation period are Rsum, and the OGEPs of the tuber formation period and the mature period are Drmin and SDr, respectively. (3) Compared with HCPs, the accuracy of AGB estimation could be improved by adding OGEPs to HCPs, OGEPs and Hdsm to HCPs at different growth period of potato, and the latter improved the accuracy more greatly. (4) The R2 of AGB modeling and verification estimated by PLSR and RF showed an upward trend from budding period to tuber growth period and then began to decrease. On the whole, R2 decreased after increased. The estimation of AGB by PLSR is better than RF in each growth period, among which the AGB estimation of tuber growth period was the best. Therefore, the estimation accuracy of potato AGB can be improved by combining the OGEPs and plant height in HCPs and using PLSR method.

The contents of arachin, conarachin and subunits significantly affect the gel properties and solubility of peanut proteins, and then affect its application in meat products and beverage. In this study, we collected 178 peanut varieties, measured arachin, conarachin, 23.5 and 37.5 kDa subunits contents by chemical methods. On the basis of peanut sample spectrum scan by near-infrared spectrum technology, we used Partial Least Squares Regression (PLSR) stoichiometry to build a mathematical model with the chemical data. By comparing single and composite spectral pretreatments, model correlation coefficient and errors to value the performance of the models. The best pretreatment method for arachin model was determined as 2nd-der with Detrend, the correlation coefficient of correction (Rc) set was 0.92, and the standard error of calibration (SEC) was 1.41; the best pretreatment method of conarachin model was detrended with 1st-der, the Rc and SEC were 0.85 and 1.46; the best pretreatment method for the 23.5 kDa subunit model was Normalization with 2nd-der, the Rc and SEC were 0.91 and 0.53; Detrend with Baseline was the best pretreatment method for the 37.5 kDa model, the Rc and SEC was 0.91 and 0.53. External validation results showed the Square Errors of Prediction (SEP) of arachin and conarachin were 1.25 and 0.73, respectively. The SEP of 23.5 kDa model and 37.5 kDa model were 0.47 and 0.75 separately. In this study, the contents of arachin, conarachin, 23.5 and 37.5kDa subunits in the whole peanut were detected simultaneously, rapidly and non-destructively based on NIRS. It’s important for the breeding specialist to select special varieties and raw materials for the protein processing industry.

Hyperspectral imaging is a new non-destructive testing technology which combines imaging and spectrum. It is an indirect analysis method. The establishment of the analytical model is critical, which needs to comprehensively consider the interaction among various modeling factors. This paper aimed to investigate the optimization of visible/near-infrared hyperspectral quantitative detection model for protein content in chilled Tan mutton based on the Box-Behnken design. The hyperspectral images of meat samples were collected by the visible/near-infrared hyperspectral imaging system. The reflectance spectral characteristics of chilled Tan mutton were analyzed. The protein contents were regarded as an external disturbance. The dynamic change of spectral signal was studied by two-dimensional correlation spectra under disturbance conditions. The synchronization spectra and autocorrelation spectra were analyzed to find the sensitive variables related to protein contents. Multiplicative scatter correction (MSC) and standard normalized variate (SNV) were used to extract useful signal and optimize the spectral quality of selected characteristic bands. In order to achieve data dimensionality reduction and reduce the burden of processing a large number of spectral data, competitive adaptive reweighted sampling (CARS) and variable combination population analysis (VCPA) were used to perform secondary extracted feature wavelengths. Extraction method, spectral pretreatment and multivariate calibration methods were factors, and each factor had 3 different levels. The response surface experimental design was used to build an optimal detection system for protein content analysis of chilled Tan mutton. The results indicated that there were strong autocorrelation peaks at 473, 679, 734 and 814 nm. The feature bands in the range of 473~814 nm were a sensitive area of protein detection in mutton. MSC and SNV could effectively eliminate the interference of scattering. Sixteen and nine characteristic wavelengths were selected by CARS and VCPA from 2DCOS, respectively. The factors in descending order affecting the predictive performance of the model were detection band, preprocessing method and modeling method. The 2DCOS-SNV-LSSVM model was selected with a high operating rate and prediction capability (Rc=0.858 8, RMSEC=0.005 8; Rp=0.860 4, RMSEP=0.005 7). The results showed that the application of the box-behnken method in the optimization of visible/near-infrared hyperspectral (400~1 000 nm) modeling parameters could effectively realize the intelligent monitoring and fast non-destructive analysis of Tan mutton quality. It could also provide a theoretical reference for the optimization of the model and improving prediction accuracy.

Surface-enhanced Raman spectroscopy (SERS) of chicken was collected by DXRTM micro-Raman spectrometer with gold colloid as an active substrate and NaCl solution as the active agent. Rapid identification of sulfadimidine (SM-2) and sulfadiazine (SPD) residues in chicken was achieved. Raman peaks at 937 and 1 188 cm-1 were used to determine whether there are SM-2 and SPD in chicken or not. The Single-factor experiment method was used to optimize the experimental conditions, and the optimum experimental conditions were obtained: the addition amount of Gold glue was 500 L, the addition amount of NaCl solutionwas 100 L and the adsorption time was 5 minutes. The original Raman spectra were pre-processed by adaptive iterative penalty least squares (air-PLS), normalization and second derivative. Then the eigenvectors were extracted by principal component analysis (PCA). Finally, the first four PCA scores were used as input values of the support vector machine (SVM) classification model, and the SVM classification model based on C-SVC type was established. The optimal penalty parameter c was 0.01, and the kernel parameter g was 0.1. The overall classification accuracy of the model was 93.23%, the sensitivity of chicken containing SM-2+SPD was 100%, and the specificity of chicken containing SPD was 99.02%. The results showed that this method had good identification effects. It could be used to detect and identify SM-2 and SPD antibiotic residues in chicken quickly.

The freshness of lettuce is one of the most important factors affecting the lettuce quality, and it depends on the storage time. Therefore, it has important research value to identify the lettuce samples with different storage time accurately. Because the near-infrared reflectance (NIR) spectra of lettuce with different storage time have different characteristics, it is feasible to use NIR technology to identify lettuce with different storage time. Gath-Geva allied fuzzy c-means (GGAFCM) clustering was proposed to replacing the Euclidean distance in allied fuzzy c-means (AFCM) clustering with the exponential distance. By iterative computations, GGAFCM can produce fuzzy membership and typical values, which combine with near-infrared reflectance spectroscopy (NIRS) to achieve the classification of the lettuce samples with different storage time accurately. The experiment was conducted on fresh samples of lettuce, which were collected with Antaris Ⅱ spectrometer every 12 hours. The spectral wavenumber ranges from 10 000 to 4 000 cm-1. At first, by principal component analysis (PCA), the 1 557-dimensional spectra of lettuce samples were compressed to the 22-dimensional data whose discriminant information was extracted by fuzzy linear discriminant analysis (FLDA). As a result, the 22-dimensional data were transformed into the two-dimensional data by FLDA with two discriminant vectors. At last, the cluster centers of fuzzy c-means (FCM) clustering acted as the initial cluster centers of both GGAFCM and AFCM, and lettuce samples with different storage time were identified by FCM, GGAFCM and AFCM whose clustering accuracies, fuzzy membership values and iterative times were analyzed. Experimental results indicated that with the same initialization conditions, the GGAFCM algorithm adopted in this study has higher discrimination accuracy than FCM and AFCM. In the case of m=2, the discrimination accuracy of GGAFCM reached 95.56%, while the clustering accuracy of FCM and AFCM was 91.11%. GGAFCM converged after 4 iterations, while both AFCM and FCM needed 8 iterations to reach convergence. Based on NIRS, GGAFCM combined with PCA and FLDA can efficiently, quickly and nondestructively complete the accurate identification of lettuce samples with different storage time. It provides the experimental foundation and reference method for accurate and rapid identification of lettuce storage time and has certain practical application value.

Tea polyphenols (TP) is one of the important ingredients of yellow tea, which has health and medicinal effects. Moreover, accurate estimation of tea polyphenol content is of great significance for tea quality identification and quantitative analysis. Previous scholars have used E-nose, E-tongue, hyperspectral and near-infrared techniques to conduct research on the estimation of tea polyphenols, and they have achieved good results. However, due to the lack of spatial characteristics, it is difficult to meet the accuracy requirements for the comprehensive judgment of the internal and external quality of tea. With the development of hyperspectral imaging system (HIS), although the estimation of tea texture based on the gray level co-occurrence matrix (GLCM) has made progress, there are still some obstacles in practical application. On the one hand, if the resolution is low, there will be no significant difference in the texture features of the image, and fewer features will not be able to fully interpret the image, resulting in lower model accuracy. On the other hand, if the resolution is high, the increase of features will make the model more complicated. Therefore, on the premise of retaining the original information of the hyperspectral image, it is necessary to explore further the potential features of hyperspectral images, especially the details of the texture. Consequently, a method of combining spectral and spatial features is proposed to improve the accuracy of tea polyphenol estimation. First, the wavelet coefficients are extracted using continuous wavelet transform based on the spectral information obtained from the hyperspectral image. Second, the wavelet coefficient features are extracted based on the wavelet coefficients, including 959 and 1 561 nm at the 4th scale, 1 321, 1 520 and 1 540 nm at the 5th scale, and 1 202 and 1 228 nm at the 6th scale. Furthermore, two characteristic wavelengths are preferred based on the sum of the energy of wavelet coefficients, which are 1 102 and 1 309 nm, respectively. Then, the gray level co-occurrence matrix and wavelet texture are extracted according to the hyperspectral image corresponding to the characteristic wavelength. Finally, the wavelet coefficient features, co-occurrence matrix, wavelet texture, and their combinations were used to construct an estimation model for the content of polyphenols in yellow tea. By comparing different regression methods based on different characteristics, including partial least squares regression (PLSR), support vector regression (SVR), and random forest (RF), five types of yellow tea were analyzed and verified. The experimental results show that SVR model based on the fusion of wavelet coefficient features, co-occurrence matrix texture, and wavelet texture achieves the best results with R2 of 0.933 0 for calibration set and 0.823 8 for the validation set. Therefore, the proposed model can effectively improve the prediction accuracy of tea polyphenol content, which also provide a technical basis for predicting other components of tea.

The grade quality of flue-cured tobacco plays an important role in the formulation design and the stability of the cigarette industry. In this paper, 768 tobacco samples from 40 prefecture-level cities in China, 2018 are selected for the experiment. The samples were classified and graded by traditional industrial grading method, including 7 different grade grades of tobacco leaves. The way to establish a tobacco quality grade prediction model by near-infrared spectroscopy and the near-infrared absorption spectrum characteristics of chemical groups and related components in different grades of tobacco are studied. The results show that the national tobacco grades prediction model is established in the non-segregated area, and the prediction standard deviation between the modeling set and the test set is not more than 1.35. After the samples are divided into five major production areas, models of each production area are established, and the prediction standard deviation of the models built in each production area after the division is found to be lower than that of the national model. The model in the Southeast region, the Southwest region and the Huanghuai region increased greatly, and the standard deviation of the test set was no more than 1. 07. After the standard normal transform (SNV) pretreatment of the average spectrum of different quality grades tobacco samples, the analysis is performed based on the information of the organic groups and related substances absorbed by the near-infrared light in different frequency ranges. It is found that tobacco with better quality grades has the characteristics of lower cellulose content and higher sugar content such as starch. The tobacco with lower quality grades has the characteristics of higher cellulose content and lower sugar content such as starch. At the same time, the worst quality grade (the upper and lower) tobacco has the characteristics of higher protein content. The results show that the application of near-infrared spectroscopy can realize the rapid prediction of the quality level of tobacco leaves. The prediction deviation is basically between adjacent levels, which meets the actual application requirements, and the prediction accuracy can be further improved by establishing prediction models of different production areas. At the same time, different grades of tobacco have different absorption characteristics of groups mainly composed of cellulose, starch, sugars, and proteins, which is also the information basis for applying near-infrared spectroscopy to achieve rapid detection of tobacco quality grades. This has important practical significance for improving the tobacco leaf grading evaluation system, further optimizing the grading scheme, and providing more scientific method guidance and technical support for product quality and maintenance.

Objects on the ground or in the atmosphere have unique polarization characteristics when they reflect, scatter, transmit, and radiate, which can be used to solve problems that some traditional optical detectors cannot, and increase the information richness. Polarization characteristics of targets have been widely used in military, environment, agriculture, medicine and many other fields. In geology, previous pieces of literature about polarization mainly focused on petrology study, in which the polarization characteristics of rocks or rock-forming minerals were used to distinguish different rock types or to explain the mechanism of differences in rock polarization properties. Though the achievements provide important references for lithological interpretation in geological mapping, no focuses on economic geology, another important branch of geology. Since ore deposits are usually associated with alterations, it is of great scientific value and economic interest to study the polarization characteristics of alteration minerals, so that the polarization spectral technology can be used to guide ore prospecting work. In this study, we carried out polarization spectral measurement and data processing for nine porphyry-related, three that cannot be distinguished by shortwave infrared, and two that cannot be distinguished by visible light alteration minerals using instrument and software developed by Beijing Research Institute of Uranium Geology in an attempt to fill the blank on polarization characteristics of alteration minerals. The results show that the polarization spectra of alteration minerals in the potassic, propylitization, sericite and advanced argillic zones of porphyry deposits are significantly different. The differences in spectral shapes and absorption intensities can be used to identify different alteration minerals, thus divide the alteration zones. The results also suggest that the polarization characteristics of three shortwave infrared blind minerals of quartz, fluorite, and potash feldspar, and two visible-light blind minerals of calcite and dolomite are significantly different, implying that a spectrometer with polarization system can identify more kinds of alteration minerals, which further implying a new generation of micro spectrometers with polarization system could be very useful in mineral exploration. Consequently, it is necessary to carry out more further and systematic research and practice on polarization characteristics of alteration minerals.

Due to the shortage of high-quality coal resources, lignite has become the main coal used in our country. Lignite has a low degree of coalification, which produce a lot of black ash and carbon dioxide when burning. The metal ions contained in the soot harm human health, so it is very meaningful to carry out the research on lignite soot. Laser-induced breakdown spectroscopy (LIBS) is a fast and multi-element method, which is suitable for in situ online detection of soot. Three lignite samples (O, H, L) with different lead concentrations were prepared in this paper, where O was the original lead-free sample. Laser Induced-breakdown Spectroscopy (LIBS) was used for insitu online detections of lignite and soot. The experimental instruments are mainly composed of laser, spectrometer, reflector, focusing lens, trigger device, carrier platform and analysis system. First, the wavelength drift of the experiment was calibrated using a high-purity lead block, and the elemental composition of the lignite samples O, H, and L was analyzed. It was found that lignite O contained C, Si, Fe, Mg, Al, Ca, Sr, Na and other elements, while N, O, Hα, Hβ and other air elements were detected. In addition, there were 8 more spectral lines of lead in the lead-containing lignite spectrum. Finally, a spectrum identification table of the main elements in lignite was given. Then the lignite was ignited with 447 nm continuous light, and 1 064 nm pulse light was focused on the soot for in situ online detectings. The qualitative analysis of the soot spectrum found that the soot contained metal ions such as Mg, Ca, Al, Sr, and Pb, indicating that some metal ions in the lignite would be discharged into the air with the soot and endanger human health. By comparing the spectrum of lignite and soot, it was found that the signal-to-noise ratio of soot was bad, and the spectral line strength of all elements was much weaker than that of lignite. In addition, it was found that the relative intensity of the carbon atom spectral line in soot was the highest among all elements (no open fire), which proved the effectiveness of LIBS for detecting CO2. In addition, the CN molecular spectrum in the experiment was analyzed, and the specific wavelength of the CN molecule was given. The rotation temperature of the CN molecule was 6 780 K and the vibration temperature was 7 520 K using the software LIFBASE fitting. At last, the lead concentration in the soot of samples H and L were analyzed, and the reference line (Ca Ⅱ 363.846 nm) was selected to normalize and compare the relative intensities of lead elements at 363.956, 368.346 and 405.780 nm. It was found that the relative intensities of the three characteristic spectral lines had a good linear relationship with their actual lead concentration, which indicates that the LIBS technology is feasible for semi-quantitative analysis of heavy metal elements in lignite soot.

Jadeite is a kind of precious jade. The value of different grades of jadeite varies greatly. Jadeite is treated by filling, dyeing, etc. to improve the appearance quality, and it is pretended to be natural jadeite. It is very necessary to identify jadeite jade. In this paper, the samples of A, B, C, B+C jadeite with different colors on the market are collected comprehensively. Based on the description of conventional gemological characteristics, the three-dimensional fluorescence spectra are tested. Three-dimensional fluorescence spectroscopy is a new fluorescence analysis technology developed in recent years. It has not been widely used in gemology. At present, the non-destructive testing of the gem filled with glue mainly depends on the infrared spectrum. The test results will be affected by the polishing degree of the sample surface and the transparency of the sample. The three-dimensional fluorescence spectroscopy technology has no high requirements on the polishing degree and transparency of the sample, to a certain extent, it can avoid the influence of the polishing degree and transparency of infrared spectrum on the test results, three-dimensional fluorescence spectroscopy technology is used to analyze the three-dimensional fluorescence spectrum characteristics of different types of jadeite samples on the market, Except that A jadeite has no fluorescence reaction, the fluorescence center of B jadeite is mostly concentrated at 380 nm(λex)/440 nm(λem). It has moderate, strong blue-white fluorescence under long-wave of the ultraviolet lamp, The fluorescence center of C jadeite is concentrated at 365 nm(λex)/443 nm(λem). It shows weak purple fluorescence under long and short-wavelength ultraviolet light. The fluorescence center of B+C purple jadeite is concentrated at 365(λex)/443 nm(λem), It has blue-violet fluorescence under long-wavelength ultraviolet light, the fluorescence peak of B+C green jadeite is mainly concentrated at 290(λex)/308 nm(λem), It has weak blue-white fluorescence under short wave ultraviolet light, the fluorescence peak of B+C yellow jadeite is concentrated at 335(λex)/377 nm( λem), Weak green fluorescence appears under long-wavelength ultraviolet light, the fluorescence peak of B+C red jadeite is 290(λex)/308 nm(λem). Weak green fluorescence under long-wavelength ultraviolet light. Under the 255 nm excitation light source, the luminescent range of jadeite jade with different treatment types is concentrated in the purple-blue region, the center wavelength of different processing types of jade is B+C green jadeite >B jadeite >C jadeite, Under the 365 nm excitation light source, the fluorescence intensity of jadeite samples is obviously stronger than that of short wave, and the luminescence range of jadeite jade with different treatment types is concentrated in the purple-green region, the center wavelength of different processing types of jadeite is B+C yellow jadeite >B+C green jadeite >B+C purple jadeite >C jadeite >B jadeite. Three-dimensional fluorescence spectroscopy is helpful to characterize resin, organic dyes, metal dyes. It can quickly and effectively identify different types of treated jadeite.

Guangyuan Thousand-Buddha Grottoes arethe largest group of cliff figures and temple grottoes in Sichuan Province. The grottoes are of great value, studying of which contributes to a better understanding of ancient politics, economics and culture. However, after a long period of weathering, they are suffering from various kinds of diseases and in urgent need of protection. We conducted phase and component analysis to elucidate the pigments and technologies of painting layers and provide evidence for conservation, with polychrome fragments that fallen from eight different grottoes as samples. The X-ray diffraction (XRD) analysis identified pigments such as iron oxide red (Fe2O3), cinnabar (HgS), minium (Pb3O4) and carbon black, as well as lead white (PbSO4), gypsum (CaSO4·2H2O), feldspar and quartz in preparation layer and support rocks. This result confirmed that deterioration of the rocks is the main reason of the pigment layer fallen. In analyses of microscopic examination on cross-sections and scanning electron microscopy with an attached energy-dispersive spectrometer (SEM-EDX), two different painting technics, i. e. preparation layer of lead white and painting without preparation layers were recognized. Because of its low concentration, the green-blue pigment could only be recognized using laser confocal Raman Microspectroscopy, and the result was langite (Cu4SO4(OH)6·2H2O). Langite is seldomly reported as a pigment. As an uncommon but widespread mineral, it is not likely to be separated from other copper minerals in ancient pigments. Manufactured basic copper sulphate pigment (Bremen green) is synthesized in the eighteenth and nineteenth century. As a result, the painting may have been redrawn in modern times if pure basic copper sulphate pigment is discovered. These results have provided scientific evidence for related conservation research and provided a new case of using langite as pigment．In addition, langite may also be helpful in studying the preservation history of grottoes.

Ethanol gasoline for motor vehicles(E10) is a new and clean automobile fuel. In the combustion process, trace heavy metal impurities have a crucial impact on the driving and maintenance of the car. Some combustion products may pollute the environment and threaten people’s health. Therefore, it is necessary to control the trace elements. The method with the direct dilution of isooctane was used to realize the quantitative determination of sodium and zinc in ethanol gasoline, The spectral lines of sodium and zinc were respectively 589.592 and 213.857 nm, the effects of atomization gas flow rate and peristaltic pump speed on the signal to back ratio were optimized. The semiconductor refrigeration atomization system (the temperature of the atomization chamber was set to -10 ℃) was applied to reduce the volatility of ethanol gasoline. At the same time, the effects of dilution ratio, standard internal elements and diluent types were studied. The results show that: (1)When kerosene was used as the diluent, the recoveries of sodium and zinc were above 120%, and the ratio of Y internal standard was above 1.20, When isooctane was used as the diluent, the recovery rate and standard internal ratio meet the requirements, which may be caused by the high density and viscosity of kerosene at low temperature, it was concluded that isooctane is more suitable as diluent ; (2)When Co was selected as the standard internal element, the standard internal ratio of some samples was more than 120%, and when Y was the internal standard, the standard internal ratio meets the requirements, it may be due to the poor stability of Co or the presence of Co in the sample, Y was suitable. (3)The detection limits of sodium and zinc are 0.013 and 0.005 mg·kg-1 respectively. The method showed lower detection limits, and the recoveries ranged from 85.1%～106.0%. The relative standard deviation (RSD) was between 1.0% and 4.8%.(4)Compared with the microwave digestion method, the result of this method is close to the theoretical value, which makes up the disadvantage of the microwave digestion method. The results show that the method was rapid, sensitive and accurate, and it was suitable for the determination of sodium and zinc in ethanol gasoline.

Due to the requirement of continuity and spectral separability, hyperspectral technology has the ability to distinguish different types of the same ground object, and the spectral data acquisition speed is fast, and the operation is simple. Spectral analysis has made outstanding achievements in monitoring water distribution and water indicators. Biochemical oxygen demandis one of the important indicators to evaluate water pollution, the current conventional measuring method for 5 culture method, and this method consumes reagent, complicated operation, more interference factors, determination of time is long, can not reflect the water quality changes in time, can’t early warning of emergent water pollution events in a timely and effective manner, in view of the traditional methods of faults, explore the content of water, BOD estimation based on the technology of hyperspectral and inversion for water quality assessment is of great significance. This test three surface water in xi ’an area as the research area, a total of 60 sites, each site repeat 10 times spectra and the BOD value, average as an original spectrum and the BOD value, Person correlation coefficient method is used to filter the spectrum and the BOD value of sensitive wavebands, and principal component analysis and least square method are used to eliminate spectral index of multicollinearity, BOD water quality index of the multivariate linear regression model and partial least squares regression model. The results were as follows: (1) the BOD sensitive bands were generally distributed at 600～900 nm, and a total of 35 original spectral indicators with significant correlation were screened out, of which the absolute value of the correlation coefficient of 758 nm was the highest (0.418). (2) the accuracy of multiple linear regression model of Z1, Z2 and BOD indexes obtained by principal component analysis (R2=0.565, RMSE=0.007) is good, and the BOD concentration of 0～0.2 and 0.4～0.6 mol·L-1 can be clearly distinguished in the principal component analysis. (3) partial least-squares regression between spectral index and BOD index shows that the model accuracy R2 of the partial least-squares regression model is up to 0.896, RMSEP=0.746 9 (root mean square error with one crossing method). By jack test, it is found that 628 nm has a very significant influence on the BOD content of inversion water body, and the bands of 889 and 893 nm have a significant influence on it. (4) according to the model fitting accuracy, the selected optimal BOD inversion model is the partial least squares regression model, and the accuracy of the partial least squares model is verified to be good (R2=0.81). Based on the above test results, an inversion method based on partial least squares hyperspectral BOD parameters of water quality is proposed, which provides a new method for dynamic detection of water quality BOD parameters.

Selecting samples in the transfer set is also important in calibration transfer. The purpose of selecting samples in the transfer set is selecting standard samples of both primary and secondary spectra with the same concentrations. After that, the transfer model between primary and secondary spectra can be generated. Finally, the prediction set of secondary spectra can be corrected by transfer model and estimated by the model generated by primary spectra. The commonly used sample selection methods include Kennard-Stone (KS), SPXY and SPXYE methods. Based on the features of those methods, a new sample selection method called weighted SPXYE (WSPXYE) was proposed and applied in transfer set selection. The WSPXYE defines the distance between each paired samples in advance, which is composed of the normalized distances between spectra (dxs), concentration (dys) and errors (des). The weighted sum of the former three distances can set as the WSPXYE distance: dwspxye=αdxs+βdys+(1-α-β)des. After obtaining dwspxye, the samples with large values of dwspxye, can be selected as transfer set. WSPXYE is the generalization on KS, SPXY and SPXYE methods, while KS, SPXY and SPXYE methods are special cases of WSPXYE with the weights of α and β set as 1 and 0; 0.5 and 0.5 and 0.333 and 0.333, respectively. Two calibration transfer methods, including direct standardization (DS) and canonical correlation analysis combined with informative component extraction (CCA-ICE) has been applied to testing the transfer set selected by WSPXYE. Results showed that WSPXYE could choose proper weights to select good transfer samples to achieve low errors in both validation and prediction sets.

Semiconductor nanocrystals (NCs) have been widely researched and reported in the past few years due to their excellent light stability, wide emission persistence and high extinction coefficient. Among them, CdSe NCs are widely used in electronic lighting, solar power generation, photoelectric sensing and other fields. However, the electrical, thermodynamic and photophysical properties of CdSe NCs have a size dependence, crystal surface defects and dangling bonds, and serious biological and environmental toxicity are prone to occur in traditional preparation methods applications, which limit their direct application. To realize the application of quantum dots in various fields, the emission wavelength, size distribution and fluorescence properties of CdSe NCs must be strictly controlled. In this study, monodisperse colloidal luminescent CdSe quantum dots were synthesized by high-temperature thermal injection method, and CdSe NCs were modified with surface ligands, and the effects of ligands with different alkyl chain lengths on the size distribution and fluorescence properties of CdSe NCs were studied. In addition, the spinning solution was prepared by changing the solvent ratio and hybridized with polyvinylpyrrolidone (PVP) to prepare PVP/CdSe QDs hybrid fibers. The results show that the CdSe NCs modified by surface ligands have good stability in the organic solution due to the decrease of intermolecular adsorption the modification of surface ligands, as well as adjustable solubility, which compensate for defects and dangling bonds caused by the decline in fluorescence performance, and play an important regulatory role in the formation of CdSe crystal structure. More importantly, this study combines surface ligand modification and hybridization to improve the adhesion of surface ligands and avoid direct contact between cadmium selenide nanocrystals and the polymer matrix during the preparation of hybrid materials. The fluorophore provides a good microenvironment and ensures the fluorescence performance of CdSe NCs, and the hybrid fiber also has stable fluorescence performance. The introduction of PVP has reduced the biotoxicity and environmental toxicity of CdSe NCs, made the material more environmentally friendly has better biocompatibility, and greatly increased the material’s application range. The experimental results show that PVP/CdSe QDs hybrid microfibers have good hybrid compatibility and dispersion, excellent fluorescence performance and material formability, simple synthesis ways and low cost, and applied to solution processing, optical lighting, Electrode materials, and biological imaging and other fields.