R. Gonfiantini et al., ADSORPTION IN GAS MASS-SPECTROMETRY-II - EFFECTS ON THE MEASUREMENT OF ISOTOPE AMOUNT RATIOS, International journal of mass spectrometry and ion processes, 171(1-3), 1997, pp. 231-242
Citations number
9
Categorie Soggetti
Spectroscopy,"Physics, Atomic, Molecular & Chemical
Gas adsorption in the mass spectrometer inlet system (equipped with mo
lecular leak) causes a deviation from linearity of the measured InRi/l
data versus time, which affects the extrapolation to time t = 0 requi
red to obtain the isotope amount ratios of a sample. The model first p
resented in (R. Gonfiantini, S. Valkiers, P.D.P. Taylor, P. De Bievre,
Adsorption in gas mass spectrometry. I. Effects on the measurement of
individual isotopic species, Int. J. Mass Spectrom. Ion Proc. (1997)
in press) is further developed in order to show the effects of adsorpt
ion on the data obtained in long measurements of SiF4 samples and impr
ove the extrapolation of InRi/l values to t = 0. The model enables to
estimate the isotope fractionation factors for the processes of gas ef
fusion through the molecular leak, and for adsorption and desorption o
n the inlet system walls. The fractionation factors obtained for gas e
ffusion and adsorption are close to the expected value of (M-l/M-i)(1/
2), where M-l and M-i are the molar masses of the isotopic species con
sidered. The isotope fractionation factor for desorption, which can be
evaluated only indirectly from the data fitting, indicates that the h
eavy isotopes are preferentially retained in the adsorbate. Model extr
apolation of data obtained during the first 10 hours of measurement gi
ves values of R-i/l(0) which are slightly but significantly smaller th
an those given by linear extrapolation of the first two hour data. The
difference is, in relative terms, -(1.88 +/- 0.23) x 10(-4) for the S
i-29/Si-28 ratio and -(3.25 +/- 0.47) x 10(-4) for the Si-30/Si-28 rat
io. The resulting molar mass for silicon is (1.0 +/- 0.1) x 10(-6) par
ts lower than that obtained with the isotope amount ratios of the line
ar extrapolation. This effect is not negligible but it cancels by usin
g synthetic isotope mixtures for the measurement calibration. The mode
l does not match the data after 15 h of measurement. This may imply th
at more than one adsorption mode occurs. Effects related to molecule f
ragmentation and the space charge distribution in the ion source are d
iscussed. (C) 1997 Elsevier Science B.V.