Figure 4. Relative abundance of m/z 17-21, together with the
results from the fit from one injection with 18O
water. On this scale, the small error bars (around
10-4) are not visible. On top, residuals from the true
signal with respect to the fitted signal of QMS (m/z 17-21) are shown.
The error bars are the standard deviations from 14 separate injections.
As a proof-of-method the enriched
H218O water was diluted
with (approximately) 1 and 2% water with 18O at
natural abundance. The expected differences in 18O
abundances between these dilutions and the not diluted enriched water
based on weights, are shown in Table 1. The measured (fitted)18O abundances and the expected 18O
abundances are within 0.03% of each other. We concluded that real
(small) differences in abundances are correctly measured.
<Table 1>