A new method for accurate determination of 15 rare earth elements including Y-Lu in tea samples by inductively coupled plasma-mass spectrometry( ICP-MS) was proposed. Sample preparation was based on the modified oxygen flask combustion technique. By using quartz cloth coated with glycerol instead of filter paper as the ignition agent, a total amount of 0. 1 g sample could be completely burned in a 500-mL flask. Key factors affecting the sample preparation effect were systematically investigated, and it was found that the extraction efficiencies of over 90% for all the analytes could be realized with 5 mL of 4% HNO3+1% HF ( V/V) as the extractant for the combustion residue and the total extraction process could be finished in one minute under sonication, which resulted in a very fast sample preparation procedure that could be completed in less than 3 minutes. The relative standard deviations ( RSD) of six parallel determination values were between 2. 7% and 5. 5%, and the detection limits ranged from 0. 001 mg/kg to 0. 006 mg/kg. Three tea standard reference materials were analyzed with the method and the detection results agreed well with the standard values. The method was successfully applied to three real samples analysis.

A new method for determination of trace Cu, Cr, Mn, Co, Ni, Cd, Fe and Pb impurities in high pure tin oxide material was proposed. 0. 1 g of sample was successfully digested at 230℃ in 3 h when autoclaves with high pressure and high temperature were used. After that, over 99. 9% of the matrix Sn was removed by chloride volatilization at 150℃ for 1. 5 h, which eliminated the matrix effect and spectra interference in the following ICP-MS detection. The recoveries for all the analytes of the spiked test were 100%±5%, the RSDs for 6 runs of sample analysis were from 0. 09% to 3. 68%, and the limits of detection (LOD, 3σ) were from 0. 002 mg/kg to 0. 034 mg/kg, which indicated that the established method were of both accuracy and high sensitivity, and could be applied for the analysis of samples with purity up to 6N. Finally, three samples specified with 4N purity were analyzed by this method.

A method for the determination of impurities in high pure Nd_2O_3 employing inductively coupled plasma mass spectrometry(ICP-MS) was established, and emphasise was put on the elimination of matrix induced NdO~+ and NdOH~+ spectral interferences using cell collision technology(CCT). Using 10% O_2-10% Ar-80% He as the reaction/collision gas and with proper adjustment of the instrumental parameters, ~(159)Tb,~(165)Ho,~(163)Dy could be quantified at ~(175)TbO,~(179)DyO and ~(181)HoO respectively, while the matrix induced NdO~+ and NdOH~+ spectral interferences could be reduced by 25-70 folds. In addition, different elements were tested as internal standards for matrix effect correction under the CCT mode, and it was found that Re was best. Other impurities including La, Ce, Sm, Eu, Gd, Er, Tm, Yb and Lu were detected under conventional mode, except that Pr was detected under high instrumental resolution mode to eliminate peak tailing interference from Nd. In combination with mathematic correction, the established method can be applied for the determination of impurities in 99.999% Nd_2O_3 products with satisfactory RSDs and recoveries.

AnewmethodwasdevelopedforfastdeterminationofAlresidueinpolyetherketoneketone ( PEKK) special polymer material by graphite furnace atomic absorption spectrometry ( GF-AAS) . Instead of high temperature ashing, sample was dissolved in concentrated sulfuric acid and diluted by tetrahydrofuran, then was directly analyzed by GF-AAS. Systematical investigation was made on sample preparation, it was found that for 0. 2 g of sample weight, a clear sample solution could be obtained by heating at 220 ℃ with 2 mL of concentrated H2 SO4 for 4 min, and the obtained solution was ready for the subsequent mixing with tetrahydrofuran for sample dilution or dispersion before analysis. Redesign of the GF-AAS temperature program was also made, on the basis of the recommended program a pre-ashing step was added, which greatly helped increasing the analytical precision when the temperature was set at 400 ℃ and the total time was set at 40 s. The detection limits ( 3σ) were 38. 5 ng/g, the RSD for six parallel determinations of real samples was 2. 2%, and the recoveries for spiked tests were 99%-105%, respectively. In comparison with the traditional high-temperature ashing technique, the established method was more concise, faster and less exposed to contamination during sample preparation.

With the use of high pure HF and HNO3 reagents, and autoclaves made of high purity Rerfluoroalkoxy ( PFA ) material, a solution sample digestion technique effective for phosphate samples, subjected to high temperature fusion, was established. The whole procedure was concise, fast and of low blank value. Key factors such as the amount and ratios of the reagents, the digestion temperature and time, were systematically optimized, it was found that within 0. 5 h at 150 ℃, only 1. 7 mL of total reagent consumption could lead to a complete sample decomposition. Most importantly, the samples were not required to be ground to fine powder, which greatly reduced the risk of contamination. In addition, an effective liquid_liquid extraction procedure based on 5_nonylsalicylaldehyde oxime as the extractant was established for matrix separation and analyte preconcentration. Under the optimal extraction conditions of 5 mL of 15% extractant, 0. 5% HNO3 of extraction acidity and 20% HNO3 of back_extraction acidity, a matrix separation efficiency of over 99. 999% could be realized and a preconctration factor of 10 could be obtained, which resulted in complete elimination of the matrix_induced interference and great enhancement of the analytical sensitivity. After optimization of the operation parameters of ICP_MS, high signal to background detection of Cu in 20%HNO3 at 840 W of plasma power and low sample uptake rate were realized. The detection limits of 2. 5 ng/g, RSD of 3. 3% for six detections of parallel samples, and the recovery of 94. 3% for spike test were obtained, respectively. The method was finally applied to three real samples analysis, and the results agreed well with the data from laser adsorption loss experiment.