抄録
Preliminary experiments on the accuracy of inclusion gas analysis with a quadrupole mass spectrometer (QMS) were carried out. Quantitative analysis of gases with a QMS is made by utilizing the two kinds of data; the intensity of the major peak for each gas and a factor called sensitivity of QMS (=major peak intensity/pressure of gas introduced in the QMS). The sensitivity is instrument-dependent. Since the amount of gas released from a natural sample may vary from sample to sample, the following instrumental factors must be examined prior to applying a QMS to inclusion gas analysis. (1) Cracking pattern for each gas does not change with pressure. (2) Sensitivity of the QMS does not vary with pressure. It was observed that the pattern coefficients of CH4, N2, CO2, and C2H6 do not change in the pressure range from 1×10−3 to 1×10−5 Pa. The slight variation in sensitivity with decreasing pressure was observed for Ar, N2, CO2, CH4, C2H6 and H2. However, the relative sensitivity (i.e., sensitivity for each gas/that for N2) for each gas is almost constant in the pressure range from 5×10−4 to 2×10−5 Pa, with the exception for H2. The multi-component working standard gas was analyzed by using the pattern and sensitivity coefficients obtained in this study. The results are in good agreement with the data obtained by gas chromatographic analysis except for H2. However, it was observed that the concentration of H2 of the multi-component standard gas increases whereas that of CO2 decreases with the progress of successive analyses. This phenomenon is due to the results of mass discrimination during the introduction and evacuation of the gas.
