Method for measuring contents of various components in covering agent by using fluorescence spectrophotometer

Abstract

The invention discloses a method for measuring contents of various components in a covering agent by using a fluorescence spectrometer. The method comprises the following steps: 1) preparing a pre-melting sample; 2) manufacturing a fuse piece; and (3) carrying out spectrum detection on the fuse piece prepared in the step (2) by utilizing a fluorescence spectrophotometer, wherein in the step 1), sodium nitrate and a mixed flux composed of lithium tetraborate and lithium metaborate are used for carrying out pre-melting treatment on a sample, and a pre-melted sample is obtained; in the step 2), the pre-molten sample is subjected to melting treatment again by using a mixed flux, and the fuse piece is obtained. According to the method disclosed by the invention, the rapid analysis of Al2O3, SiO2, CaO, MnO, MgO, TiO2 and Fe2O3 in the covering agent can be completed at the same time, and the labor intensity is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of raw material analysis, and particularly relates to a method for determining the contents of various components in a covering agent by using a fluorescence spectrometer, and more particularly to a method for determining the contents of Al2O3, SiO2, CaO, MnO, MgO, TiO2, and Fe2O3 in a covering agent by using an XRF-1800 fluorescence spectrometer. Background Art

[0002] During continuous casting, the molten steel is transferred into the tundish. Due to the too large exposed surface of the molten steel, the heat loss increases, resulting in too large a temperature drop of the molten steel. The covering agent covers the surface of the molten steel and plays a role in heat insulation and reduction of heat loss on the molten steel surface. With the continuous development of iron and steel production technology, the requirements for the cleanliness of steel are continuously improved. It is required that the covering agent not only has the function of heat insulation, but also has various metallurgical functions such as preventing secondary oxidation of molten steel, adsorbing floating inclusions on the molten steel surface, and participating in the refining of molten steel to reduce or remove harmful elements in the molten steel. The covering agent is divided into 3 types: acidic, neutral, and alkaline according to different chemical compositions. The content differences of each component in different types of covering agents are relatively large, which brings difficulties to quantitative analysis. Summary of the Invention

[0003] Objective: The chemical wet method is usually used to determine the chemical components in the covering agent. However, this type of method requires measuring the content of each component one by one, with complex operations and a long detection cycle, and cannot meet the requirements of rapid detection. As a modern detection method, X-ray fluorescence spectrometry has the characteristics of simple operation, fast analysis speed, wide analysis range, high precision, and can simultaneously determine multiple elements, and is widely used in the detection of various materials in the metallurgical industry. Since there is no standard substance for the covering agent, the present invention selects standard substances such as electric furnace slag, blast furnace slag, converter slag, open hearth furnace slag, and basic blast furnace slag according to the content of each component in the sample to establish a calibration curve, uses melting sample preparation to eliminate the mineral effect and matrix effect of the sample, and uses a method combining the theoretical influence coefficient method and the empirical coefficient method to correct the absorption enhancement effect between coexisting elements, and establishes an analytical method for simultaneously determining the main and minor components in the covering agent by melting sample preparation-X-ray fluorescence spectrometry. This method can simultaneously complete the rapid analysis of Al2O3, SiO2, CaO, MnO, MgO, TiO2, and Fe2O3 in the covering agent, reduce the labor intensity, improve the work efficiency, and save chemical reagents; the final result can provide rapid and powerful data support for the research on the law of the covering agent and the process research of the entire steelmaking process.

[0004] The present invention is specifically realized through the following technical solutions.

[0005] The present invention provides a method for determining the contents of various components in a covering agent by using a fluorescence spectrometer, which includes the following steps:

[0006] 1) High-temperature calcination and preparation of a pre-molten sample;

[0007] 2) Prepare a fused bead;

[0008] 3) Use a fluorescence spectrometer to perform spectral detection on the fused bead obtained in step 2);

[0009] Wherein:

[0010] In step 1), after the sample is calcined, the sample is pre-fused with sodium nitrate and a mixed flux composed of lithium tetraborate and lithium metaborate to obtain the pre-fused sample;

[0011] In step 2), the pre-fused sample is fused again with the mixed flux to obtain the fused bead.

[0012] In some embodiments, the multiple components include Al2O3, SiO2, CaO, MnO, MgO, TiO2, Fe2O3.

[0013] In some embodiments, in step 1), the pre-fusion treatment is carried out in a corundum crucible with a graphite reaction substrate.

[0014] In some embodiments, the corundum crucible is a corundum crucible with a volume of 80 mL.

[0015] In some embodiments, the ratio of lithium tetraborate to lithium metaborate in the mixed flux is 2:1.

[0016] In some embodiments, the operation of step 1) is as follows: Take a constant-weight porcelain boat, weigh it, weigh about 2 g of the sample and place it in the porcelain boat, weigh it again, calcine the porcelain boat at 800 °C for 40 min, cool it and weigh it, accurately weigh 0.4000 g of the calcined sample, add 0.2500 g of sodium nitrate and 3.0000 g of the mixed flux, mix well, transfer it to a corundum crucible with a graphite reaction substrate, place it in a muffle furnace and calcine it at 1050 °C for 30 min, cool it, take out the fused mass, clean the graphite powder attached to the surface, and obtain the pre-fused sample.

[0017] In some embodiments, the operation of step 2) is as follows: Weigh 4.0000 g of the mixed solvent and spread it evenly in a platinum crucible, place the pre-fused sample in the platinum crucible, add 5 drops of saturated ammonium iodide solution, place the crucible on a refractory lining board, place it in a muffle furnace that has been heated to 1050 °C, melt it at high temperature for 10 min, take it out, shake it quickly, then place it on the refractory lining board again, melt it at 1050 °C for 5 min, take it out, shake it again, and slowly cool it to room temperature. After cooling, pour out the glass sheet to obtain the fused bead.

[0018] In some embodiments, the operation in step 3) is: using the side of the fused sheet that is close to the bottom surface of the crucible as the irradiation surface, and performing spectral detection using the fluorescence spectrometer.

[0019] In some embodiments, the fluorescence spectrometer is an 1800 type X-ray fluorescence spectrometer.

[0020] The present invention has the following beneficial effects:

[0021] 1. The present invention proposes a method for determining the contents of Al2O3, SiO2, CaO, MnO, MgO, TiO2, and Fe2O3 in the covering agent by the melting method. For the existing methods to detect Al2O3, SiO2, CaO, MnO, MgO, TiO2, and Fe2O3 in the covering agent, they are mainly analyzed by methods such as gravimetry, inductively coupled plasma atomic emission spectrometry, and volumetric method. The process of analyzing the covering agent by gravimetry is cumbersome and has a long cycle. A large amount of chemical reagents are required during the analysis experiment, which is not conducive to rapid batch analysis. Inductively coupled plasma atomic emission spectrometry can simultaneously determine elements such as Al2O3 and CaO, but its pretreatment procedure is still relatively cumbersome and the reagent consumption is large. This project can achieve obtaining the content data of Al2O3, SiO2, CaO, MnO, MgO, TiO2, and Fe2O3 by a single staff member in a short time with less reagent consumption and a one-time detection;

[0022] 2. The present invention uses the pre-oxidation melting method to pretreat the sample in advance to avoid the sample corroding the platinum crucible. The covering agent contains reducing substances, which will corrode the platinum crucible during high-temperature melting. After the crucible is corroded, it will affect the preparation of the sample, thereby affecting the accuracy of the test. This method completely oxidizes the sample through pre-oxidation treatment, and the platinum crucible will no longer be corroded during the melting of the platinum crucible;

[0023] 3. The present invention uses the melting method to eliminate the mineral effect and particle size effect during the sample testing process, improve the accuracy of the method test, and expand the scope of use of the method. Description of the Drawings

[0024] Figures 1 - 5 Respectively show the standard working curves of MgO, Fe2O3, SiO2, Al2O3, and CaO.

[0025] Figure 6 Is a schematic diagram of the final reaction device.

[0026] Figure 7 And Figure 8 Respectively are the results of the values of the Mandel h and k consistency statistics obtained statistically.

[0027] Figure 9 Shows the repeatability limit and reproducibility limit iteration curves of aluminum oxide.

[0028] Figure 10 and Figure 11 are the results of the values ​​of Mandel's h and k consistency statistics respectively.

[0029] Figure 12 The repeatability and reproducibility limit iteration curves for silica are shown.

[0030] Figure 13 and Figure 14 Results are given for the values ​​of Mandel's h and k agreement statistics obtained by grouping the laboratories.

[0031] Figure 15 The repeatability limit and reproducibility limit iteration curves for magnesium oxide are shown.

[0032] Figure 16 and Figure 17 The values ​​of Mandel's h and k agreement statistics obtained by grouping the laboratories are given.

[0033] Figure 18 The iteration curves of the repeatability limit and reproducibility limit of ferric oxide are shown.

[0034] Figure 19 and Figure 20 The values ​​of Mandel's h and k agreement statistics obtained by grouping the laboratories are given.

[0035] Figure 21 The repeatability limit and reproducibility limit iteration curves for calcium oxide are shown. DETAILED DESCRIPTION

[0036] The research contents of the present invention include: research on flux ratio of molten glass sheets, research on oxidation method selection, research on oxidant selection and dosage, optimization of instrument test conditions, synthesis and configuration of standard substances, calibration curve and drawing of calibration curve, research on method accuracy and precision and method comparison.

[0037] Technical route: Determine the appropriate amount of pre-oxidant, amount of pre-oxidation flux, melting time, and pre-oxidation conditions. The pre-oxidized sample is melted at high temperature in a platinum crucible, and the conditions for melting the sample are determined by different flux amounts, melting times, and melting temperatures. After determining the method for melting the sample, the curve is established after the covering agent sample is set, and the test conditions such as the test voltage, current, and test spectrum of the instrument during the test are determined by standard substances with different contents. After determining the optimal test conditions, the curve is drawn and the sample is tested, and the precision and accuracy experiments of the method are carried out. Finally, the repeatability limit and reproducibility limit of the method are calculated.

[0038] The content of the present invention will be described in detail below through specific embodiments. The embodiments are intended to facilitate the understanding of the present invention rather than to limit the content of the present invention.

[0039] 1 Main instruments and reagents

[0040] Model 1800 X-ray fluorescence spectrometer (Shimadzu Corporation, Japan) (operating parameters are shown in Table 1);

[0041] Muffle furnace, platinum-gold crucible, 80 mL corundum crucible;

[0042] Mixed flux: lithium tetraborate and lithium metaborate (66.67:33.33);

[0043] Release agent: ammonium iodide (saturated solution);

[0044] Oxidizing agent: sodium nitrate;

[0045] Reaction lining material: high-purity graphite;

[0046] Table 1: Recommended table of determination parameters

[0047] Element Analysis line 2θ angle (degree) KV / mA Time S Detector Crystal Al Kα 144.71 40 / 70 20 FPC PET Si Kα 108.98 40 / 70 20 FPC PET Mg Kα 45.08 40 / 70 20 FPC TAP Ca Kα 113.08 40 / 30 20 FP LiF Fe Kα 57.520 40 / 70 20 SC LiF Mn Kα 86.17 40 / 70 20 SC LiF Ti Kα 62.97 40 / 70 20 SC LiF

[0048] 2 Experimental methods

[0049] 2.1 Preparation of samples

[0050] The covering agent sample is passed through a 200-mesh sieve and pre-dried at 105 - 110 °C for 1 h.

[0051] 2.2 Burning and pre-melting of samples

[0052] Take a constant-weight porcelain boat, weigh it, weigh about 2 g of the sample and place it in the porcelain boat, weigh it again. Burn the porcelain boat at 800 °C for 40 min, cool it and weigh it. Accurately weigh 0.4000 g of the burned sample, add 0.2500 g of sodium nitrate and 3.0000 g of the mixed flux, mix well, transfer it to a corundum crucible with graphite as the reaction substrate, place it in a muffle furnace and burn it at 1050 °C for 30 min. After cooling, take out the melt block, clean the graphite powder adhering to the surface, and obtain the pre-melted sample.

[0053] 2.3 Making melt slices

[0054] Weigh 4.0000g of mixed solvent and spread it evenly in a platinum crucible, place the glass frit formed by pre-oxidation and melting in the platinum crucible, add 5 drops of saturated ammonium iodide solution, place the crucible on a refractory lining, and place it in a muffle furnace that has been heated to 1050°C. Melt at high temperature for 10 minutes. After taking it out, shake it quickly, then put it on a refractory lining, and melt it again at 1050°C for 5 minutes. Take it out, shake it again, and slowly cool it to room temperature. After cooling, pour out the glass piece (molten glass piece) to obtain a molten piece, use the side close to the bottom of the crucible as the irradiation surface, and perform spectral detection.

[0055] 2.3 Preparation of standard working curve

[0056] As there is no standard covering agent material for sale in the market, 9 selected covering agent samples were calibrated and the calibrated covering agent samples were used as standard samples to establish the analysis curve.

[0057] We used chemical methods to determine the contents of Al2O3, SiO2, CaO, MgO, Fe2O3, MnO, and TiO2, and took the average value after multiple measurements as the true content of each oxide in the covering agent sample. The results are shown in Table 2.

[0058] Table 2: Content of each element in the covering agent

[0059]

[0060]

[0061] 2.3 Drawing of standard working curve

[0062] According to the steps 3.2.2 and 3.2.3, the standard covering material with the set value in Table 3 is used to make a fused sheet, and then the intensity is tested on the fluorescence spectrometer. A working curve is established based on the covering agent sample content (according to the result after correction for loss on ignition) and the tested fluorescence intensity. The content of the analyzed element can be calculated through the curve by measuring the intensity of the unknown sample. We divided each curve into sections according to the actual test situation. Each element is divided into two sections, high and low, to draw the curve. After segmentation, the deviation of the low-content test result is smaller and the test value is more accurate. The specific situation of each curve is as follows Figures 1 - 5 As shown, Figures 1 - 5 The curves of MgO, Fe2O3, SiO2, Al2O3 and CaO are shown as examples respectively.

[0063] 3 Results and discussion

[0064] 3.1 Determination of pre-oxidation melting scheme

[0065] For the covering agent sample, each component mainly exists in the form of oxides, but there are still a small amount of elemental substances. If it is directly melted in a platinum crucible, obvious corrosion will occur after long-term use, resulting in spots inside the platinum crucible, the surface of the sample is no longer smooth, and the sample is prone to cracking during cooling. Therefore, to avoid the corrosion of the platinum crucible during melting, after pre-oxidation melting, it is then melted in a platinum crucible to prepare a glass fusion slice.

[0066] 3.1.1 Selection of pre-oxidation vessels and lining materials

[0067] At present, the method of using pre-melting oxidation to prepare samples for testing the components in slag by X-ray fluorescence spectrometer has been reported in the literature. The main methods are: burning the sample at high temperature to convert the reducing substances in the sample into oxide form, thereby eliminating the corrosion of the platinum crucible by the sample; another is the method that has emerged in recent years of forming a protective film by the flux adhering to the wall, and then putting the sample and the oxidant into the protective film for melting together. This method has strict requirements for the melting equipment on the one hand, and on the other hand, the operator needs to have quite high skills when making the protective film, and this method is difficult to promote. Through experimental verification of the above two methods, the method of high-temperature burning cannot completely oxidize the reducing substances, and obvious corrosion of the platinum crucible still occurs after long-term use. For the method of forming a protective film by adhering to the wall, since a high-frequency melting furnace is used for heating, when melting, the protective film is first melted, and then the sample is melted, which cannot play a protective role. Therefore, it is crucial to find a simple and feasible melting method to completely solve the corrosion phenomenon.

[0068] We take a reaction vessel with a reaction lining, let the sample and the oxidant carry out high-temperature melting oxidation, fully oxidize the reducing substances in the sample, and then transfer it to a platinum crucible for high-temperature melting to prepare a glass sheet. Through the heat resistance of various material vessels and the high-temperature volatilization performance of related materials, it is determined to use a corundum crucible as the reaction vessel and high-purity graphite as the reaction lining material.

[0069] Experiments were carried out on corundum crucibles with volumes of 35 mL, 50 mL, 80 mL, and 100 mL. When the volume is too small, the flux is easy to contact the crucible wall during melting, and the sample will be lost. When the volume is too large, when melting in a muffle furnace, due to the size limitation of the furnace chamber, only 2 to 3 samples can be melted at a time, and the working efficiency is too low. Considering the melting effect and working efficiency, it is finally determined to use an 80 mL corundum crucible as the reaction vessel.

[0070] The materials used as reaction linings are required to have the following characteristics: 1. They should not contain the elements to be measured; 2. They should not enter the test sample or can be easily removed if they do enter the test sample; 3. They should provide good isolation during high-temperature reactions to prevent the flux from contacting the crucible wall; 4. The frit should be easily separable from the lining material. Through screening, graphite is selected as the reaction lining material. The main component of graphite is carbon, which can be oxidized into carbon dioxide gas and removed at high temperatures. Under the condition of being covered by the flux, the volatilization rate of graphite is very slow, ensuring that it can play a good role in isolating the flux and the crucible during the entire reaction process. Moreover, the frit floats on the surface of the graphite, and after melting is completed, the sample frit is easily separable from the graphite. After being compacted, graphite forms a solid and smooth surface to ensure that the sample will not fall into the interior of the graphite during the melting process, thus avoiding sample loss.

[0071] After determining the lining material and the reaction vessel, the graphite is compacted and the surface in contact with the flux is made smooth. A special tool is used to compact it to form a pit for easy placement and melting of the sample. The final reaction device is as shown in Figure 6 a and b in the figure.

[0072] 3.1.2 Types and Dosages of Oxidants

[0073] Sodium nitrate begins to decompose at 380 °C, releases a large amount of oxygen at 400 °C - 600 °C, and releases nitric oxide and nitrogen dioxide during decomposition at 700 °C - 865 °C. Moreover, the decomposition rate of sodium nitrate is relatively fast. Sodium nitrate has a strong oxidation ability. Finally, the oxidant is 0.2000 g of sodium nitrate.

[0074] 3.1.3 Pre-oxidation Melting Temperature and Flux Dosage

[0075] Based on the sample weighing amount and the dosage of the oxidant, in order to achieve better melting and dispersion of the sample, while ensuring that the sample can be well oxidized and melted, and there will be no loss of the sample due to flux splashing during the melting reaction, the final determined flux addition amount during pre-oxidation melting is 3.0000 g. A 1:2 type flux of lithium metaborate and lithium tetraborate is used. The general melting temperature of this kind of flux is 1050 °C. Therefore, the melting temperature is determined to be 1050 °C.

[0076] Through the above experiments, the final melting scheme was determined as follows: Weigh 0.4000 g of the sample, add 0.2000 g of sodium nitrate and 3.0000 g of the mixed flux. After mixing evenly, place it in a corundum crucible with a graphite reaction inner lining, and place it in a muffle furnace to melt at 1050 °C for 30 min. Cool down, take out the melt block, sweep away the graphite powder on the surface, and obtain the pre-melted sample (also called the pre-oxidized melted sample). Place it in a platinum crucible that has been pre-weighed with 4.0000 g of the mixed flux, add 5 drops of ammonium iodide demolding agent, place it in a muffle furnace to melt at 1050 °C for 15 min. During this period, take it out to expel air bubbles, shake well and then melt again. Finally, take it out and cool down to obtain a melt sheet. After demolding, use an X-ray fluorescence spectrometer for determination. The measured value needs to be corrected according to the loss on ignition during high-temperature calcination, and the corrected result is the final result.

[0077] 3.2 Accuracy Experiment of the Testing Method

[0078] Determine the contents of Al2O3, SiO2, CaO, MnO, MgO, TiO2, and Fe2O3 in the covering agents Y3-159, Y4-289, Y3-99, Y3-160, Y4-290, Y5-38, Y5-39, Y3-100, Y3-158, and Y3-101 by chemical methods. At the same time, prepare glass sheets by melting according to the steps in 3.2.2 and 3.2.3, and detect Al2O3, SiO2, CaO, MnO, MgO, TiO2, and Fe2O3 according to the set method. Compare the test values of the two methods. The specific test situation is shown in Table 3.

[0079] Table 3: Comparison of Test Values of 7 Elements by Two Methods (%)

[0080]

[0081]

[0082] It can be seen from Table 3 that when the values of Al2O3, SiO2, CaO, MnO, MgO, TiO2, and Fe2O3 in the covering agent are tested by this method and compared with the values determined by the chemical method, the difference between the two is small, indicating that this method has high accuracy.

[0083] 3.3 Precision Experiment

[0084] According to the method of preparing glass sheets by melting in the steps of 3.2.2 and 3.2.3, conduct precision experiments with standard substances: Y9-84, Y9-75, Y10-25, Y10-22, Y9-77, Y10-39, Y9-73, Y9-231, and Y9-232 respectively. Each sample is measured 10 times, and the average value, standard deviation, and relative standard deviation of the 10 measurements are calculated. The specific data is shown in Tables 4-12.

[0085] Table 4: Precision Data of Y9-84

[0086]

[0087] Table 5: Precision Data of Y9-75

[0088]

[0089] Table 6: Precision Data of Y10-25

[0090]

[0091]

[0092] Table 7: Precision Data of Y10-22

[0093]

[0094] Table 8: Precision Data of Y9-77

[0095]

[0096] Table 9: Precision Data of Y10-39

[0097]

[0098]

[0099] Table 10: Precision Data of Y9-73

[0100]

[0101] Table 11: Precision Data of Y9-231

[0102]

[0103] Table 12: Precision Data of Y9-232

[0104]

[0105]

[0106] It can be seen from the above data that the relative standard deviations of the precision test data of the nine specimens are all less than 10%, indicating that this method has high precision.

[0107] 3.4 Repeatability Limit and Reproducibility Limit of the Method

[0108] Nine groups of specimens were melted to prepare glass chips according to this method, and precision experiments were carried out at different levels of content. This experiment requires that under repeatability conditions, each laboratory independently measures the element content at each level 4 times, that is, in a short period of time, by the same experimenter, using the same instrument, the same experimental conditions, and the same calibrated instrument, for measurement.

[0109] 4 Summary and Analysis of Original Data

[0110] 4.1 Aluminum Oxide

[0111] 4.1.1 Summary of Precision Original Data

[0112] This precision was measured 4 times for aluminum oxide at 9 levels by different personnel, and the measurement results are shown in Table 13.

[0113] Table 13: Original Data of Precision of Aluminum Oxide (%)

[0114]

[0115] 4.1.2 Unit Mean

[0116] The unit mean and unit variance of the experimental results are shown in Tables 14 and 15.

[0117] Table 14: Unit Mean

[0118]

[0119] Table 15: Unit Variance

[0120]

[0121] 4.1.3 Consistency Check

[0122] Figure 7 and Figure 8 give the statistically obtained values of Mandel's h and k consistency statistics.

[0123] Judging from the graphs of Mandel's statistics h and statistics k, the data consistency is good.

[0124] 4.1.4 Test for Outliers

[0125] 4.1.4.1 Test for Unit Variance

[0126] The Cochran test was applied to obtain the results in Table 16 below:

[0127] Table 16: Results of Cochran Test

[0128] Horizontal Test value Cochran's test statistic 5% probability level 1% probability level 1 4.3039 0.391 0.532 0.626 2 19.7926 0.277 0.532 0.626 3 30.2983 0.384 0.532 0.626 4 21.3690 0.290 0.532 0.626 5 31.5495 0.513 0.532 0.626 6 4.1724 0.584 0.532 0.626 7 22.7010 0.414 0.532 0.626 8 21.6482 0.260 0.532 0.626 9 12.7909 0.527 0.532 0.626

[0129] N = 4, P = 6, the 5% critical value of the Cochran test is 0.532; the 1% critical value is 0.626. The Cochran test did not find any discrepant values or outliers.

[0130] 4.1.4.2 Test unit mean

[0131] The Grubbs test was applied to the unit means, and the results are shown in Tables 17 and 18 below:

[0132] Table 17: Grubbs test results

[0133]

[0134]

[0135] Table 18: Two-point Grubbs test results

[0136] Level 1 2 3 4 5 6 7 8 9 Two low values 0.730 0.335 0.440 0.543 0.051 0.066 0.477 0.209 0.385 Two high values 0.136 0.391 0.260 0.084 0.545 0.517 0.157 0.493 0.248 G(1％) 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 G(5％) 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 Conclusion Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified

[0137] The single-point Grubbs test was performed on the unit means respectively. When P = 6, the critical value of the Grubbs test at the 1% significance level is 1.973, and at the 5% significance level is 1.887; no discrepant values or outliers were found through the single-extreme Grubbs test. No discrepant values or outliers were found in the two-point test.

[0138] 4.1.5 Statistical results of m, r, R and coefficient of variation CV(R)

[0139] The results of the mean m, repeatability limit r, reproducibility limit R and coefficient of variation CV(R) for each level are shown in Table 19. The coefficient of variation CV(R) for all levels did not exceed MAXCV(R).

[0140] Table 19: Statistical results of each level of aluminum oxide

[0141]

[0142] Among them, the coefficient of variation CV(R) = 100×σ(R) / m, R = 2.8×σ(R),

[0143] Target coefficient of variation AIMCV(R): lgCVR = -0.3466lgm + lg1.47721

[0144] Maximum coefficient of variation, when m > 0.001%, MAXCV(R): lgCVR = -0.3466lgm + lg3.24670,

[0145] When m ≤ 0.001%, MAXCV(R) = 35.71.

[0146] 4.1.6 Establish the regression equation between r, R and m

[0147] 4.1.6.1 Case of \(r^ = a + bm\)

[0148] The values of \(a\) and \(b\) are obtained by weighted cubic iterative regression

[0149] a = 0.09535

[0150] b = 0.0077

[0151] r^ = 0.09535 + 0.0077m

[0152] 4.1.6.2 Case of \(R^ = a + bm\)

[0153] The values of \(a\) and \(b\) are obtained by weighted cubic iterative regression

[0154] a = 0.1298

[0155] b = 0.0112

[0156] R^ = 0.1298 + 0.0112m

[0157] Figure 9 The \(a\) and \(b\) show the repeatability limit and reproducibility limit iterative curves of aluminum oxide

[0158] 4.1.7 Coefficient of variation

[0159] The coefficient of variation of the smoothed statistical results is shown in Table 20

[0160] Table 20: Coefficient of variation of smoothed aluminum oxide at each level

[0161]

[0162] The coefficient of variation at each point is less than MAXCV(R). Therefore, the determination range of aluminum oxide is recommended to be 0.020% - 5.00%

[0163] 4.2 Summary of raw data on the precision of silicon dioxide

[0164] 4.2.1 Raw data on the precision of silicon dioxide

[0165] This precision is determined 4 times for 9 levels of silicon dioxide by different personnel, and the determination results are shown in Table 21

[0166] Table 21: Raw data on the precision of silicon dioxide %

[0167]

[0168]

[0169] 4.2.2 Unit Mean Value

[0170] The unit mean value and unit variance of the experimental results are shown in Tables 22 and 23.

[0171] Table 22 Unit Mean Value

[0172]

[0173] Table 23 Unit Variance

[0174]

[0175] 4.2.3 Consistency Check

[0176] Figure 10 and Figure 11 The statistically obtained values of Mandel h and k consistency statistics are given, where Figure 10 is the between-laboratory consistency Mandel statistic h grouped by laboratory, Figure 11 is the between-laboratory consistency Mandel statistic k grouped by laboratory.

[0177] Judging from the graphs of Mandel statistic h and statistic k, the data consistency is good.

[0178] 4.2.4 Outlier Test

[0179] 4.2.4.1 Test of Unit Variance

[0180] The following results are obtained by applying the Cochran test:

[0181] Table 24: Results of Cochran Test

[0182]

[0183]

[0184] N = 4, P = 6, the 5% critical value of the Cochran test is 0.532; the 1% critical value is 0.626. No discrepant values or outliers are found in the Cochran test.

[0185] 4.2.4.2 Test of Unit Mean Value

[0186] The Grubbs test is applied to the unit mean value, and the results are as follows:

[0187] Table 25: Results of Grubbs Test

[0188] Level 1 2 3 4 5 6 7 8 9 Total average 27.8469 6.6293 4.3353 7.6711 5.2778 27.9712 7.8527 4.4507 15.9046 Standard deviation of the mean 1.579E-01 1.008E-01 6.557E-02 9.939E-02 8.373E-02 1.458E-01 8.678E-02 6.055E-02 1.120E-01 Maximum of the mean 28.1462 6.7989 4.4227 7.7811 5.4398 28.2586 7.9950 4.4879 16.0530 Minimum of the mean 27.6942 6.4871 4.2210 7.5250 5.2292 27.8616 7.7921 4.3304 15.7643 Gmax 1.895 1.683 1.333 1.106 1.936 1.972 1.639 0.614 1.324 Gmin 0.967 1.411 1.743 1.470 0.580 0.752 0.698 1.986 1.254 Gn 1.895 1.683 1.743 1.470 1.936 1.972 1.639 1.986 1.324

[0189] Table 26: Results of Two-Point Grubbs Test

[0190] Level 1 2 3 4 5 6 7 8 9 Two low values 14.738 0.037 3.531 0.378 1.798 16.370 1.501 0.267 25.492 Two high values 18.413 3.144 0.003 0.392 0.001 0.032 0.004 0.009 4.123 G(1％) 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 G(5％) 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 Conclusion Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified

[0191] Perform single-point Grubbs tests on the unit averages. When P = 6, the critical value of the Grubbs test at the 1% significance level is 1.973, and at the 5% significance level is 1.887. No outliers are found through the Grubbs single extreme value test, and no discrepant values or outliers are found in the two-point test.

[0192] Statistical results of 4.2.5m, r, R and coefficient of variation CV(R)

[0193] The results of the average value m, repeatability limit r, reproducibility limit R and coefficient of variation CV(R) at each level are shown in Table 27, and the coefficient of variation CV(R) at all levels does not exceed MAXCV(R).

[0194] Table 27 Statistical results of each level of silica

[0195]

[0196] Among them, the coefficient of variation CV(R)=100×σ(R) / m, R = 2.8×σ(R),

[0197] Target coefficient of variation AIMCV(R): lgCVR=-0.3466lgm + lg1.47721

[0198] Maximum coefficient of variation, when m>0.001%, MAXCV(R): lgCVR=-0.3466lgm + lg3.24670,

[0199] When m≤0.001%, MAXCV(R)=35.71.

[0200] 4.2.6 Establish the regression equation between r, R and m

[0201] 4.2.6.1 The case of r^ = a + bm

[0202] Use weighted cubic iterative regression to find the values of a and b

[0203] a = 0.1310

[0204] b = 0.0079

[0205] r^ = 0.1310 + 0.0079m

[0206] 4.2.6.2 The case of R^ = a + bm

[0207] Use weighted cubic iterative regression to find the values of a and b

[0208] a = 0.02050

[0209] b = 0.0118

[0210] R^ = 0.02050 + 0.0118m

[0211] Figure 12 In it, a and b show the repeatability limit and reproducibility limit iterative curves of silica.

[0212] 4.2.7 Coefficient of Variation

[0213] The coefficient of variation after smoothing the statistical results is shown in Table 28.

[0214] Table 28 Coefficient of Variation after Smoothing of Each Level of Silica

[0215]

[0216] The coefficient of variation at each location is less than MAXCV(R). Therefore, the recommended determination range of silica is 0.020% - 5.00%. 4.3 Summary of Original Data on Precision of Magnesium Oxide

[0217] 4.3.1 Original Data on Precision of Magnesium Oxide

[0218] This precision was determined 4 times for 9 levels of magnesium oxide by different personnel, and the determination results are shown in Table 29.

[0219] Table 29 Original Data on Precision of Magnesium Oxide %

[0220]

[0221]

[0222] 4.3.2 Unit Mean

[0223] The unit mean and unit variance of the experimental results are shown in Tables 30 and 31.

[0224] Table 30 Unit Mean

[0225]

[0226] Table 31 Unit Variance

[0227]

[0228] 4.3.3 Consistency Check

[0229] Figure 13 and Figure 14 give the values of Mandel h and k consistency statistics obtained by grouping and statistically analyzing by laboratory. From the graphs of Mandel statistic h and statistic k, the data consistency is good.

[0230] 4.3.4 Test for Outliers

[0231] 4.3.4.1 Test for unit variance

[0232] Applying the Cochran's test gives the following results:

[0233] Table 32 Cochran's test results

[0234] Level Test value Cochran's test statistic 5% probability level 1% probability level 1 9.4020 0.581 0.532 0.626 2 15.7794 0.328 0.532 0.626 3 8.5770 0.326 0.532 0.626 4 15.7110 0.308 0.532 0.626 5 8.5108 0.357 0.532 0.626 6 9.7259 0.448 0.532 0.626 7 16.7031 0.513 0.532 0.626 8 13.7719 0.271 0.532 0.626 9 11.5802 0.271 0.532 0.626

[0235] N = 4, P = 6, the 5% critical value of Cochran's test is 0.532; the 1% critical value is 0.626. No discrepant values and outliers are found in the Cochran's test.

[0236] 4.3.4.2 Test for unit mean

[0237] Applying the Grubbs' test to the unit mean gives the following results:

[0238] Table 33 Grubbs' test results

[0239] Level 1 2 3 4 5 6 7 8 Total average 9.4020 15.7794 8.5770 15.7110 8.5108 9.7259 16.7031 13.7719 Standard deviation of the mean 9.452E-02 1.017E-01 9.930E-02 8.394E-02 9.581E-02 9.182E-02 9.862E-02 9.211E-02 Maximum of the mean 9.5179 15.8659 8.7182 15.8599 8.6598 9.8485 16.8253 13.8449 Minimum of the mean 9.2862 15.6202 8.4466 15.6219 8.4280 9.5939 16.5840 13.5906 Gmax 1.226 0.850 1.422 1.774 1.555 1.334 1.239 0.793 Gmin 1.224 1.566 1.313 1.061 0.864 1.438 1.208 1.968 Gn 1.226 1.566 1.422 1.774 1.555 1.438 1.239 1.968

[0240] Table 34 Two-point Grubbs' test results

[0241] Level 1 2 3 4 5 6 7 8 Two low values 0.318 0.027 0.427 0.561 0.592 0.323 0.139 0.034 Two high values 0.197 0.625 0.153 0.088 0.127 0.233 0.475 0.718 G(1％) 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 G(5％) 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 Conclusion Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified

[0242] Performing the single-point Grubbs' test on the unit mean respectively, when P = 6, the critical value of the Grubbs' test at the 1% significance level is 1.973, and at the 5% significance level is 1.887; no discrepant values and outliers are found through the single-extreme-value Grubbs' test, and no discrepant values and outliers are found in the two-point test.

[0243] 4.3.5 Statistical results of m, r, R and coefficient of variation CV(R)

[0244] The results of the mean m, repeatability limit r, reproducibility limit R and coefficient of variation CV(R) for each level are shown in Table 35, and the coefficient of variation CV(R) for all levels does not exceed MAXCV(R).

[0245] Table 35 Statistical results of each oxidation level

[0246]

[0247]

[0248] Among them, the coefficient of variation CV(R) = 100×σ(R) / m, R = 2.8×σ(R),

[0249] Target coefficient of variation AIMCV(R): lgCVR = -0.3466lgm + lg1.47721

[0250] Maximum coefficient of variation. When m > 0.001%, MAXCV(R): lgCVR = -0.3466lgm + lg3.24670,

[0251] When m ≤ 0.001%, MAXCV(R) = 35.71.

[0252] 4.3.6 Establish the regression equations between r, R and m

[0253] 4.3.6.1 The case of r^ = a + bm

[0254] Use weighted cubic iterative regression to find the values of a and b

[0255] a = 0.0058

[0256] b = 0.0151

[0257] r^ = 0.0058 + 0.0151m

[0258] 4.3.6.2 The case of R^ = a + bm

[0259] Use weighted cubic iterative regression to find the values of a and b

[0260] a = 0.2378

[0261] b = 0.0062

[0262] R^ = 0.2378 + 0.0062m

[0263] Figure 15 a and b in it show the iterative curves of the repeatability limit and reproducibility limit of magnesium oxide.

[0264] 4.3.7 Coefficient of variation

[0265] The coefficient of variation after smoothing the statistical results is shown in Table 36.

[0266] Table 36 Coefficient of variation after smoothing for each level of magnesium oxide

[0267]

[0268] The coefficient of variation at each place is less than MAXCV(R). Therefore, it is recommended that the determination range of magnesium oxide be 0.020% - 5.00%. 4.4 Summary of the original data on the precision of iron(III) oxide

[0269] 4.4.1 Original data on the precision of iron(III) oxide

[0270] This precision was determined 4 times for 9 levels of iron(III) oxide by different personnel. The determination results are shown in Table 37.

[0271] Table 37 Precision raw data of iron(III) oxide %

[0272]

[0273] 4.4.2 Unit average value

[0274] The unit average value and unit variance of the experimental results are shown in Table 38 and Table 39.

[0275] Table 38 Unit average value

[0276]

[0277] Table 39 Unit variance

[0278]

[0279]

[0280] 4.4.3 Consistency check

[0281] Figure 16 and Figure 17 give the values of Mandel h and k consistency statistics obtained by grouping and statistics according to laboratories. From the graphs of Mandel statistics h and statistics k, the data consistency is good.

[0282] 4.4.4 Test for outliers

[0283] 4.4.4.1 Test for unit variance

[0284] Applying Cochran's test gives the following results:

[0285] Table 40 Cochran's test results

[0286] Level Test value Cochran's test statistic 5% probability level 1% probability level 1 0.8950 0.277 0.532 0.626 2 0.7979 0.519 0.532 0.626 3 0.7796 0.343 0.532 0.626 4 3.7129 0.263 0.532 0.626 5 0.7923 0.471 0.532 0.626 6 0.8876 0.547 0.532 0.626 7 2.0722 0.511 0.532 0.626 8 1.3833 0.246 0.532 0.626 9 1.1322 0.245 0.532 0.626

[0287] N = 4, P = 6, the 5% critical value of Cochran's test is 0.0.532; the 1% critical value is 0.626. Cochran's test did not find any discrepant values or outliers.

[0288] 4.4.4.2 Test for unit average value

[0289] Applying Grubbs' test to the unit average value gives the following results:

[0290] Table 41 Grubbs' test results

[0291] Level 1 2 3 4 5 6 7 8 9 Total average 0.8950 0.7979 0.7796 3.7129 0.7923 0.8876 2.0722 1.3833 1.1322 Standard deviation of the mean 1.972E-02 3.180E-02 1.468E-02 5.790E-02 2.236E-02 1.966E-02 3.997E-02 1.552E-02 1.989E-02 Maximum of the mean 0.9191 0.8389 0.7943 3.7969 0.8096 0.9114 2.1098 1.4121 1.1718 Minimum of the mean 0.8706 0.7759 0.7546 3.6532 0.7589 0.8682 2.0190 1.3660 1.1189 Gmax 1.220 1.290 0.999 1.450 0.772 1.211 0.939 1.855 1.991 Gmin 1.237 0.691 1.706 1.031 1.491 0.990 1.333 1.114 0.668 Gn 1.237 1.290 1.706 1.450 1.491 1.211 1.333 1.855 1.991

[0292] Table 42 Two-point Grubbs' test results

[0293] Level 1 2 3 4 5 6 7 8 9 Two low values 0.502 0.714 0.196 0.518 0.016 0.412 0.089 0.583 0.732 Two high values 0.106 0.002 0.402 0.067 0.642 0.120 0.577 0.111 0.039 G(1％) 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 G(5％) 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 Conclusion Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified

[0294] Perform single-point Grubbs tests on the unit averages respectively. When P = 6, the critical value of the Grubbs test at the 1% significance level is 1.973, and at the 5% significance level is 1.887; no outliers or extreme values are found through the Grubbs single extreme value test. No outliers or extreme values are found in the two-point test.

[0295] 4.4.5 Statistical results of m, r, R and coefficient of variation CV(R)

[0296] The results of the average value m, repeatability limit r, reproducibility limit R and coefficient of variation CV(R) at each level are shown in Table 43, and the coefficient of variation CV(R) at all levels does not exceed MAXCV(R).

[0297] Table 43 Statistical results of each level of iron(III) oxide

[0298]

[0299] Among them, the coefficient of variation CV(R) = 100×σ(R) / m, R = 2.8×σ(R),

[0300] Target coefficient of variation AIMCV(R): lgCVR = -0.3466lgm + lg1.47721

[0301] Maximum coefficient of variation, when m > 0.001%, MAXCV(R): lgCVR = -0.3466lgm + lg3.24670,

[0302] When m ≤ 0.001%, MAXCV(R) = 35.71.

[0303] 4.4.6 Establish the regression equation between r, R and m

[0304] 4.4.6.1 Case of r^ = a + bm

[0305] Use weighted cubic iterative regression to find the values of a and b

[0306] a = -0.0465

[0307] b = 0.0833

[0308] r^ = -0.0465 + 0.0833m

[0309] 4.4.6.2 Case of R^ = a + bm

[0310] Use weighted cubic iterative regression to find the values of a and b

[0311] a = 0.0161

[0312] b = 0.0570

[0313] R^ = 0.0161 + 0.0570m

[0314] Figure 18 In it, a and b show the repeatability limit and reproducibility limit iteration curves of iron(III) oxide.

[0315] 4.4.7 Coefficient of variation

[0316] The coefficient of variation after smoothing the statistical results is shown in Table 44.

[0317] Table 44 Coefficient of variation after smoothing for each level of iron(III) oxide

[0318]

[0319] The coefficient of variation at each location is less than MAXCV(R). Therefore, it is recommended that the determination range of iron(III) oxide be 0.050% - 5.00%. 4.5 Summary of original data on the precision of calcium oxide

[0320] 4.5.1 Original data on the precision of calcium oxide

[0321] This precision was determined 4 times for 9 levels of calcium oxide by different personnel. The determination results are shown in Table 45.

[0322] Table 45 Original data on the precision of calcium oxide %

[0323]

[0324] 4.5.2 Unit average value

[0325] The unit average value and unit variance of the experimental results are shown in Table 46 and Table 47.

[0326] Table 46 Unit average value

[0327]

[0328]

[0329] Table 47 Unit variance

[0330]

[0331] 4.5.3 Consistency check

[0332] Figure 19 and Figure 20 The values of Mandel's h and k consistency statistics obtained by grouping and statistical analysis by laboratory are given. From the graphs of Mandel's statistic h and statistic k, the data consistency is good.

[0333] 4.5.4 Detection of Outliers

[0334] 4.5.4.1 Testing the Variance of the Unit

[0335] Applying Cochran's test gives the following results:

[0336] Table 48 Results of Cochran's Test

[0337] Level Test value Cochran's test statistic 5% probability level 1% probability level 1 21.3503 0.332 0.532 0.626 2 21.4985 0.546 0.532 0.626 3 23.7740 0.594 0.532 0.626 4 24.5140 0.359 0.532 0.626 5 23.9352 0.405 0.532 0.626 6 28.3171 0.488 0.532 0.626 7 28.5276 0.343 0.532 0.626 8 48.6206 0.317 0.532 0.626 9 50.5204 0.416 0.532 0.626

[0338] N = 4, P = 6, the 5% critical value of Cochran's test is 0.532; the 1% critical value is 0.626. No discrepant values and outliers are found. 4.5.4.2 Testing the Mean of the Unit

[0339] Applying Grubbs' test to the mean of the unit gives the following results:

[0340] Table 49 Results of Grubbs' Test

[0341]

[0342]

[0343] Table 50 Results of Two-Point Grubbs' Test

[0344] Level 1 2 3 4 5 6 7 8 9 Two low values 0.465 0.430 0.307 0.435 0.150 0.761 0.262 0.018 0.358 Two high values 0.171 0.209 0.344 0.335 0.530 0.402 0.273 0.702 0.242 G(1％) 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 0.0116 G(5％) 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 0.0349 Conclusion Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified Qualified

[0345] Performing a single-point Grubbs' test on the mean of the unit respectively, when P = 6, the critical value of Grubbs' test at the 1% significance level is 1.973, and at the 5% significance level is 1.887; through the Grubbs' single extreme value test, no discrepant values and outliers are found, and no discrepant values and outliers are found in the two-point test.

[0346] 4.5.5 Statistical Results of m, r, R and Coefficient of Variation CV(R)

[0347] The results of the mean m, repeatability limit r, reproducibility limit R and coefficient of variation CV(R) at each level are shown in Table 51, and the coefficient of variation CV(R) at all levels does not exceed MAXCV(R).

[0348] Table 51 Statistical Results of Calcium Oxide at Each Level

[0349]

[0350] Among them, the coefficient of variation CV(R) = 100×σ(R) / m, R = 2.8×σ(R),

[0351] Target Coefficient of Variation AIMCV(R): lgCVR = -0.3466lgm + lg1.47721

[0352] Maximum coefficient of variation. When m > 0.001%, MAXCV(R): lgCVR = -0.3466lgm + lg3.24670,

[0353] When m ≤ 0.001%, MAXCV(R) = 35.71.

[0354] 4.5.6 Establish the regression equations between r, R and m

[0355] 4.5.6.1 The case of r^ = a + bm

[0356] Use weighted cubic iterative regression to find the values of a and b

[0357] a = 0.0637

[0358] b = 0.0077

[0359] r^ = 0.0637 + 0.0077m

[0360] 4.5.6.2 The case of R^ = a + bm

[0361] Use weighted cubic iterative regression to find the values of a and b

[0362] a = 0.110

[0363] b = 0.0007

[0364] R^ = 0.110 + 0.0117m

[0365] Figure 21 a and b in it show the iterative curves of the repeatability limit and reproducibility limit of calcium oxide.

[0366] 4.5.7 Coefficient of variation

[0367] The coefficient of variation after smoothing the statistical results is shown in Table 52.

[0368] Table 52 Coefficient of variation after smoothing for each level of calcium oxide

[0369]

[0370] The coefficient of variation at each point is less than MAXCV(R). Therefore, it is recommended that the determination range of calcium oxide be 0.020% - 5.00%. 4.6 Repeatability and reproducibility equations for each oxide

[0371] Table 53 Summary table of repeatability and reproducibility for each oxide of aluminum, silicon, calcium, magnesium, manganese, titanium, and iron

[0372] Element Level range Repeatability r Reproducibility R <![CDATA[Al2O3]]> 0.020％～5.00％ r＝0.09535+0.0077m R = 0.1298 + 0.0112m <![CDATA[SiO2]]> 0.020％～5.00％ r＝0.1310+0.0079m R = 0.0205 + 0.0118m CaO 0.020％～5.00％ r＝0.0637+0.0077m R = 0.110 + 0.0117m MgO 0.050％～5.00％ r＝0.0058+0.0151m R = 0.2378 + 0.0062m MnO 0.020％～5.00％ r＝0.0029+0.0214m R = 0.0036 + 0.0295m <![CDATA[TiO2]]> 0.020％～5.00％ r＝-0.00023+0.0508m R = 0.0035 + 0.0608m <![CDATA[Fe2O3]]> r＝-0.0465+0.0833m R = 0.0161 + 0.0570m

[0373] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for determining the contents of multiple components in a covering agent by using a fluorescence spectrometer, characterized in that, The method includes the following steps: 1) High-temperature burning of the sample and preparation of a pre-molten sample; 2) Preparation of a fused slice; 3) Using a fluorescence spectrometer to perform spectral detection on the fused slice obtained in step 2); Wherein: In step 1), sodium nitrate and a mixed flux composed of lithium tetraborate and lithium metaborate are used to perform pre-melting treatment on the sample to obtain the pre-molten sample; In step 2), the pre-molten sample is melted again using the mixed flux to obtain the fused slice.

2. The method according to claim 1, characterized in that, In step 1), the pre-melting treatment is carried out in a corundum crucible with graphite as the reaction substrate.

3. The method according to claim 2, wherein The corundum crucible is a corundum crucible with a volume of 80 mL.

4. The method according to any one of claims 1 to 3, characterized in that, The ratio of lithium tetraborate to lithium metaborate in the mixed flux is 2:

1.

5. The method according to any one of claims 1-4, characterized in that, The operation of step 1) is as follows: Take a porcelain boat of constant weight, weigh it, weigh about 2 g of the sample and place it in the porcelain boat, weigh it again, burn the porcelain boat at 800 °C for 40 min, weigh it after cooling, accurately weigh 0.4000 g of the burned sample, add 0.2500 g of sodium nitrate and 3.0000 g of the mixed flux, mix well, transfer it to a corundum crucible with graphite as the reaction substrate, place it in a muffle furnace and burn it at 1050 °C for 30 min, after cooling, take out the fused block, clean the graphite powder adhering to the surface, and obtain the pre-molten sample.

6. The method according to any one of claims 1-5, characterized in that, The operation of step 2) is as follows: Weigh 4.0000 g of the mixed solvent and spread it evenly in a platinum crucible, place the pre-molten sample in the platinum crucible, add 5 drops of saturated ammonium iodide solution, place the crucible on a refractory lining board, place it in a muffle furnace that has been heated to 1050 °C, melt it at high temperature for 10 min, take it out, shake it quickly, then place it on the refractory lining board again and melt it at 1050 °C for 5 min, take it out, shake it again after taking it out, and slowly cool it to room temperature. After cooling, pour out the glass sheet to obtain the fused slice.

7. The method according to any one of claims 1-6, characterized in that, The operation of step 3) is as follows: Using the side of the fused slice that is close to the bottom surface of the crucible as the irradiation surface, perform spectral detection using the fluorescence spectrometer.

8. The method according to any one of claims 1-7, characterized in that The fluorescence spectrometer is a 1800-type X-ray fluorescence spectrometer.

Images