In thermal ionization mass spectrometry instrumental fractionation produces bias which preclude direct measurements of the values of the isotopic ratios of the element in analysis. Cavazzini (2009; 2012) has shown that the empirical laws of fractionation, as power law and exponential law (Russel et al., 1978) can be interpreted in terms of a process of distillation in the ion source of the mass spectrometer, according to Rayleigh's (1902) distillation theory. This interpretation predicts linear paths of evolution in x vs. y and in x vs. x/z diagrams where x, y and z are measured values of isotopic ratios. Analytic representations of these paths of evolution show that the slope and the y-intercept of the linear paths are functions of the true values of the isotopic ratios involved, x0, y0, z0 and of a quantity, which we here call a, which is a function of the masses of the isotopes which define the isotopic ratios. Due to particular structure of slope and yintercept, if we know the values of the masses of the isotopes involved we can calculate the true values of the isotopic ratios. If a number of x vs. y distributions are obtained from different measurements of a sample, however, it is observed that the slope and the intercept of the straight line which best fits the distribution change. These changes can be interpreted as due to a not-exact equivalence of the fractionation factors of the isotopic ratios involved, which turns quantity a to parameter a' that does not depend only on the masses of the isotopes. This fact can be related to irreproducibility in loading the sample on the filament, to irreproducible behavior of the acceleration voltage, of voltages of the collimators in the ion source, of the current through the filament(s) etc. which determine the value of a' to change randomly. Since the values of the true isotopic ratios are sensible functions of these parameters, the calculated values may significantly differ, resulting in the impossibility to calculate the true values of the isotopic ratios by a single measurement of the sample. In this case, however, due to particular structure of slope and y-intercept of the theoretical lines which represents the distributions, it can be shown that parameter a' can be eliminated, and the slopes of distributions obtained from different measurements of a sample are expected to be linearly correlated with the respective intercept values, the slope and the intercept of this linear correlation being the true values of the isotopic ratios. The values of slope and intercept of the straight lines which best-fit the distributions we obtained from different measurements of the worldwide used NIST SRM 987 have been observed to be linearly correlated. The linear best-fit of this correlation suggests for this standard material 88Sr/86Sr = 8.29 ± 0.01 and 87Sr/86Sr = 0.70705 ± 0.00045 (1? errors), values which are significantly lower than those presently assumed in the literature. Cavazzini, G.( 2009): Rayleigh's distillation law and linear hypothesis of isotope fractionation fractionation in thermal ionization mass spectrometry. Int. J. Mass Spectr., 288, 84-91. Cavazzini, G. (2012): Distillation law and exponential model of isotope fractionation. Int. J. Mass Spectr., 309, 129-132. Rayleigh, J.W.S. (1902): On the distillation of binary mixtures. Philos. Mag., 42, 521-537. Russell, W.A., Papanastassiou, D.A., Tombrello, T.A. (1978): Ca isotope fractionation on the Earth and other solar system materials. Geochim. Cosmochim. Acta, 42, 1075-1090.
Inter-measure determination of the isotopic composition of strontium
2017
Abstract
In thermal ionization mass spectrometry instrumental fractionation produces bias which preclude direct measurements of the values of the isotopic ratios of the element in analysis. Cavazzini (2009; 2012) has shown that the empirical laws of fractionation, as power law and exponential law (Russel et al., 1978) can be interpreted in terms of a process of distillation in the ion source of the mass spectrometer, according to Rayleigh's (1902) distillation theory. This interpretation predicts linear paths of evolution in x vs. y and in x vs. x/z diagrams where x, y and z are measured values of isotopic ratios. Analytic representations of these paths of evolution show that the slope and the y-intercept of the linear paths are functions of the true values of the isotopic ratios involved, x0, y0, z0 and of a quantity, which we here call a, which is a function of the masses of the isotopes which define the isotopic ratios. Due to particular structure of slope and yintercept, if we know the values of the masses of the isotopes involved we can calculate the true values of the isotopic ratios. If a number of x vs. y distributions are obtained from different measurements of a sample, however, it is observed that the slope and the intercept of the straight line which best fits the distribution change. These changes can be interpreted as due to a not-exact equivalence of the fractionation factors of the isotopic ratios involved, which turns quantity a to parameter a' that does not depend only on the masses of the isotopes. This fact can be related to irreproducibility in loading the sample on the filament, to irreproducible behavior of the acceleration voltage, of voltages of the collimators in the ion source, of the current through the filament(s) etc. which determine the value of a' to change randomly. Since the values of the true isotopic ratios are sensible functions of these parameters, the calculated values may significantly differ, resulting in the impossibility to calculate the true values of the isotopic ratios by a single measurement of the sample. In this case, however, due to particular structure of slope and y-intercept of the theoretical lines which represents the distributions, it can be shown that parameter a' can be eliminated, and the slopes of distributions obtained from different measurements of a sample are expected to be linearly correlated with the respective intercept values, the slope and the intercept of this linear correlation being the true values of the isotopic ratios. The values of slope and intercept of the straight lines which best-fit the distributions we obtained from different measurements of the worldwide used NIST SRM 987 have been observed to be linearly correlated. The linear best-fit of this correlation suggests for this standard material 88Sr/86Sr = 8.29 ± 0.01 and 87Sr/86Sr = 0.70705 ± 0.00045 (1? errors), values which are significantly lower than those presently assumed in the literature. Cavazzini, G.( 2009): Rayleigh's distillation law and linear hypothesis of isotope fractionation fractionation in thermal ionization mass spectrometry. Int. J. Mass Spectr., 288, 84-91. Cavazzini, G. (2012): Distillation law and exponential model of isotope fractionation. Int. J. Mass Spectr., 309, 129-132. Rayleigh, J.W.S. (1902): On the distillation of binary mixtures. Philos. Mag., 42, 521-537. Russell, W.A., Papanastassiou, D.A., Tombrello, T.A. (1978): Ca isotope fractionation on the Earth and other solar system materials. Geochim. Cosmochim. Acta, 42, 1075-1090.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
