Highly Refined Acquisition Of High Temperature Raman Spectra And The Fundamental Research Of Quantitively Analysis

The development of metallurgy from experienced to scientific is emphatically marked by understanding the properties of molten slag along with the knowledge of its micro structure. The interconnection between micro structure and properties is an important background facilitating new technology and guiding production.In the past two decades, several approaches such as high temperature X-ray and high temperature Raman spectroscopy have been developed to research the micro structure of molten slag,. Raman spectroscopy provides abundant information of vibration modes of molecular or elementary micro structure and so is beneficial to the study on thermodynamic and kinetic behaviors of metallurgical slag. So Raman spectroscopy was selected by the present author as the major instrument in the study.So far there is no popularly accepted understating on the Raman spectra of molten silicate. On one hand it is considerably difficult to precisely record a Raman spectrum under high temperature. On the other hand the decovolution of the high temperature Raman spectra is diversified. Referring the vibration mode of anion clusters, especially the effect of complex anion as Al, Fe on which more research works are expected to progressively clarify the interconnection between micro structure and properties.This dissertation is intended to discuss the factors determining the precision of the spectra recorded by the high temperature Raman spectroscope developed in Shanghai University and the way to make an improvement. At first, the fundamental of Raman spectroscopy and its new progress are reviewed. Based on the illustration of diverse approaches in suppressing thermo emission under high temperature two sets of high temperature Raman spectroscope have been put forward. According to accumulated time-resolution principle in modification of the U1000 scope provided by Jobin Yvon company to form SU-HTRS. It is the first one in China which can work stably under 2000 K. The second set named as SU-HTRS(T/S) was developed by introducing the idea of coupling accumulated time resolution and spatial resolution into LabRAM HR800. In addition it is planned to introduce a pulsed UV laser into SU-HTRS(T/S), which will helpful during the investigation of surface behavior of some samples with a strong UV absorbency. These characteristics make both of HTRS developed in Shanghai University are able to stand up on the frontier in global.The advantages and disadvantages of SU-HTRS(T/S) have been reviewed as the main contribution in this thesis. The author points out:1) When use a confocal microscopy to record a spectrum the cost of improving the spatial resolution is the degradation of Raman intensity. It was found in experiment if the background thermal emission is raised by 30% meanwhile the Raman scattering intensity can be elevated by 6-fold. So during recording HTRS the principle of adjusting diaphragm claims enough scattering light should be input in the monochrometer, and the increase of thermal emission is limited. Therefore the coupling of time-resolution and spatial resolution approaches provides the capability to adjust recording operation in a wider range.2) The ability to suppress thermal emission in SU-HTRS(T/S) is reduced by almost two magnitudes in comparison with that in SU-HTRS. This hints up to now the optimized coupling of two resolution approaches does not be realized in SU-HTRS(T/S).3). The first measure in optimization of SU-HTRS(T/S) should be the introduction of coincidence circuit to control the information transfer from MCP to CCD inside ICCD.The gist in operation mode of SU-HTRS(T/S) is according to the characteristics of sample to adjust the contribution ratio of accumulated time resolution and spatial resolution. It is wisdom to record melt spectra mainly rely on accumulated time resolution. By selecting the size of the confocal hole properly we can get rid of the high temperature background at most, but keep the Raman signal immovable.In order to match the in situ micro Raman under high temperature a new high temperature heating stage was developed by the author after analyzing the stages published in literature. The working temperature of it stably reaches 1923K.The premise to deduce reliable structural information from a melt spectrum is the spectrum deconvolution should be correct. For this purpose, it is obligated to enhance the signal to noise ratio of the spectrum by mathematic methods. Several kinds of methods have been compared and the results show that wavelet is the best one.In other aspect, new illustrations of the spectra of sodium silicate, calcium silicate and alumina silicate have been proposed to clarifying the misunderstanding and ambiguous concepts.1). Although a lot of experiment research combining simulation show that the structure of binary silicate could be resolved into five units as Q₀～Q₄, but in principle the partial peak of Q4 could not be involved in the high wavenumber region, because no O_nb is included in a Q₄ and its scattering intensity is much lower than others. In the sample with high content of SiO₂ such as Na₂O-4SiO₂, the right shoulder at 1094 cm^-1 peak is suggested as the adjacent enhanced phenomenon of Q₃ and Q₄.Because of the difference of cations in the deconvolution of the spectra of CaSiO3 and Na2SiO3 glasses one must consider that the kinds of structural unit involved in them can be different. This is the equilibrium of the conversion among Q_i may be diverse and the actual progressing degree adjacent to equilibrium can also be in discrepancy. Due to that partial structural feature under high temperature and the heating history is contained in a glass, the glass spectra are always impossible to show a good corresponding relation with the relevant melt spectra.2). People found that several properties in ternary aluminosilicates do not exhibit a sudden turning over at the ratio Al/M=1, and when Al/M＜1 the existence of O_nb has been proved. These results do not coincide with the popular view point. According to the chemical characteristics of Al the author proposed a new concept that every Al-O tetrahedron contains one coordinated bond. The major part is of Al-O_nb＜_Al^na, and a few of Al-O_b＜_Al^Al. The former is composed of two 2(s/p) electrons of O_nb in the ordinary Na-O_nb-Si bond and the empty 3p_z orbital of A1, it belongs to the category ofÏ€bond. The later is composed of two 2(s/p) electrons of O_b in the ordinary Si-O_b-Al bond and two 3p_z empty orbital of two Al cations. It belongs to the category of delocalizedÏ€bond. The oxygen anion included in the coordinated bond is of third valence.3). The ab initio calculation of ternary aluminosilicate was carried out. The results is in accord with the experiment results, which indicates that when introducing Al into binary silicate the peaks in high wavenumber region exhibit a red-shift.4). The spectra of Al₂SiO₅ polymorph (kyanite, andalusite and silimanite) have been compared with the spectra ofζ-Na₂Si₂O₅ published by Fleet. The comparison indicates the spectra of two sample series are in accordance. As Fleet claimed that both Ob and Onb are coordinated to (6)Si. So it is reasonable to infer that in three Al₂2SiO₅ minerals both O_b and O_nb are coordinated to ⁽⁶⁾Al. In addition, it is supposed that Si-O＜_(6)Al^(6)Al is one of the micro structural units according to the three-dimensional configuration of the Al₂SiO₅ crystal. So the criterion of the assignment of a peak should not be the coordination characteristics but the bonding behavior.5). The room temperature spectra of jadeite (Na₂O·Al₂O₃·4SiO₂) and albite (Na₂O·Al₂O₃·6SiO₂) have been compared with the spectra of Na₂Si(Si₂O₇) and Na₆Si₃(Si₉O₂₇). The spectra of two sample series are similar. And it is found that several kinds of tetrahedra are coexisted in both series of samples.