P. Xu et al., The methodology study of trace element in-situ analyses using Laser Ablation Inductively Coupled Plasma Mass Spectrometer., ACTA PETR S, 16(2), 2000, pp. 291-304
National Institute of Standards and Technology (NIST) Standard Reference Ma
terial (SRM) 610, 612, 614 glasses have been measured by Laser Ablation Ind
uctively Coupled Plasma Mass Spectrometry (LA-ICP-MS). The experiments are
designed elaborately to optimal UV laser probe (266 nm) and ICP-MS operatin
g conditions. The measurement results of 78 experiments of 25 groups in 5 d
ays show that 30 trace elements in NIST612 have the statistic precision of
about 4% when the Ca-44 was chosen as internal standard, meanwhile those el
ements of same NIST612 standard glass have the statistic precision of about
6% when the Si-29 was chosen as the internal standard. Good measurement ac
curacy of less than 10% which compared with certified data can also be acqu
ired routinely when we measure the NIST610, NIST612 and NIST614 glasses usi
ng the NIST612 as the calibration reference material. The total relative st
atistic deviation of measurement can be deduced by two individual test stat
istic deviations. Choosing NIST612 as calibration reference material, we ge
t total relative statistic deviation value of less than 10% of most trace e
lements both in NIST610 (450 mu g/g) and in NIST612 (40 mu g/g) when we cho
ose Ca-44 as internal standard. By the contrast, we can get the total relat
ive statistic deviation value of 15% to most of trace elements in NIST612 i
f we choose 29Si as internal standard. It suggests that Ca-44 is better tha
n Si-29 as the internal standard mass. Except to Ni, Sc, Rb, Zr, Sm, Gd and
Pb, that most of trace elements have the total relative statistic deviatio
n value of less than 20% in NIST614(0. 8 mu g/g) which is a little higher t
han in NIST610 and NIST612 show that the measurement precision will deterio
rate with the decreasing of concentration of measured element. High blank a
ccounts for the bad Sc value and the interference of CaO may account for th
e bad Ni value. The bad precision of Rb, Zr, Sm, Gd and Ph may result from
the memory effect of whole system which will be displayed more obviously wh
en we analysis the low concentration elements.