蛍光X線による全岩化学組成の分析 : その1.主要元素
スポンサーリンク
概要
- 論文の詳細を見る
X-ray fluorescence analysis (XRF) is one of the easiest and most rapid analytical methods that can be used for major and trace element analyses of rock samples. This paper presents an analytical technique for major elements in rock samples by XRF. Grinding experiments were done by the planetary ball mill Fritsch P-5,and granodiorite and andesite were used as hard and less hard samples. First, samples were crushed into small pieces ( < 4 mm in diameter). Approximately 20-25 g of samples were set in containers made of tungsten carbide (WC) and agate. A rotation speed of 200-300 rpm and a grinding time of 20 min yielded rock powders with an average grain size of 3.7-8.4 μm and a dispersion of 11.7-122 μm^2 for WC containers, and an average grain size of 5.9-6.2 μm and a dispersion of 34.8-69.2 μm^2 for agate containers. However, agate containers could not grind granodiorite samples completely. Rock powder was fused into a glass bead by a glass-bead sampler Philips THG-5. The rock powder was put into an oven and dried overnight at 110 'C. The mixing ratio of the sample : flux was chosen to be I : 10 to reduce matrix effects and 0.5000 g of the sample and 5.000 g of Li_2B_4O_7 were weighed and put into a Pt cup. After trial and error, the most suitable conditions of fusion were determined. To correct for variations in X-ray intensity, three corrections were made : monitor and drift corrections to adjust the temporal variations for long and short periods, and a bead correction to cancel variations arising from the roughness of the bottom of the Pt cup. Among these, the drift correction was found unnecessary because of the stabile X-ray intensity during the measurement. The bead correction is expected to be useful, but its effect could not be confirmed. 36 samples ranging in composition from basalt to dacite were analysed by XRF ma-chines of Shizuoka University and the Institute of Earthquake Research, University of Tokyo. Both gave consistent results with relative errors of ±5 % for SiO_2,TiO_2,Al_2O_3,Fe_2O_3,MgO and CaO. The relative errors are less than ±10 % for MnO (> 0.1 wt%), Na_2O (> 2 wt%) K_2O (> 0.5 wt%) and P_2O_5 (>0.2 wt%).
- 静岡大学の論文