Study On Applications And Information Extraction Of Mineral By Near Infrared Spectroscopy

Near-Infrared Spectroscopy (NIRS) would occur when there were absorption due to overtone and combination bands originating from fundamental molecules vibrations. The spectrum includes information mainly to C-H,N-H,S-H and O-H groups in compound. Modern NIRS analysis technology are composed of infrared spectrometer, Chemometrics software and applications model. Compared to traditional methods, NIRS technology has a lot of advantages. The advantages includes it need no pretreatment to the sample, no chemistry reagent.It can fulfill measurement in a few seconds, does no harm to the sample, makes no pollution to the sample and environment. The penetrability of the infrared spectrum is very strong, its energy is lower than visible lights, and thus it does no harm to human body. In a word, the NIRS analysis technology has the characteristic of high efficient, low expenditure, and no pollution. The majority of minerals have vibrant absorption in infrared band. Infrared spectrometer allows identification of minerals whose spectrum is alternative. So NIRS analytical instrument can fulfill distinguishment between most minerals in the open field, it is especially efficient for minerals that are rich in water. There is series of puzzles in NIRS analytic Technology. The key of the technology is that the near-Infrared spectrum, as the sources of information, have low effective information and to analyze the complex samples with NIRS is to extract slender information from complicated, overlapping, alternant spectrum. So extracting effective information from NIR spectrum and optimizing the original spectrum data are crucial to develop ideal mathematical model in NIRS. This paper, combined with Chinese geologic investigation item"The study of the field minerals NIRS analysis method and development of instrument", "The study of applications calcineed kaolin resource in coal measures strata in Jilin"(item number:3J101093401) and Jilin university innovation fund item"The study of the soil environment Near-Infrared detection information analysis system in situ",taking layer silicates minerals as the object of study, uses signals analysis method and data analysis theory to extract characteristic information from NIR spectrum. The aims of the study are to offer methods to solve several questions about mineral analysis and its application. The main methods and results of this paper are: 1.The NIR spectral features of major minerals, which have molecule of H2O, Al-OH and Mg-OH groups in the compound, were systematic summarized after the NIRS analytic principle, absorption principle of electronic processes and vibration processes were briefly addressed. Then drawn the conclusion that distinguishing the peak position of the spectrum absorption is an important method for identifying minerals. 2.Collecting representative spectrum of different content layer silicates samples with high distinguishable scanner. The original spectrum with abundant information is obtained. montronite, sericite, gaolinite and halloysite were used as the analytical samples for the model. NIRS quantitative model was set up by compounds comprise of montronite and sericite which were constituted by different percentages.NIR standard models of gaolinite and halloysite were set up by the results analyzed by IR and XRD methods. 3. Montronite and Sericite compound Samples that are mixed up of different percentages are used in the study. The NIR spectrum information extracted from the compound (include the position, depth, area and symmetrical degree of the peak spectrum) are used to optimize original data. By the characteristic parameter based on the spectrum information, this paper also advanced and established regression analysis,stepwise regression analysis mathematical model. The paper also addressed actual effect of the method. 4. Using correlation analysis technology to calculate forecasted correlation...