IRMS data 
25 
C22 C23 C24 C25 C26 C27 C28 C29 C30 C31 C32 C33 
Fig. 15: d13C-means of 
sample 1, 10 from IAEA with 
standard deviation as error 
margin. IAEA measured in 
triplicates (n=3). 
IAEA performed analysis In triplicates and used a more elaborated methodol 
ogy compared to the NFI, with following steps Involved: 
• Preparation of a linear chained hydrocarbon fraction of the paraffin wax mix 
ture 
• Reference C0 2 , calibrated against PDB, was Introduced five times at the 
beginning, three times at the end of every Isotopic GC determination. 
• Calibration standards of n-alkanes were Introduced every 4 samples. 
For the duplicate analysis, the results of both laboratories show that the 8 13 C- 
content of each Individual n-alkane Is within the range of 1 %o (Fig. 15 and 
Fig. 14). In case of IAEA’s duplicate analysis (Fig. 15), sample 10 shows unsta 
ble 13 C-contents along the series of n-alkanes. Especially the sudden Increase 
(< 1 per mil) from C28 to C29 does not correspond to the average composition 
of the neighbouring n-alkanes. 
However, there was an outlier In the NFI’s 
duplicate run of the first sample, resulting In 
Increased RSD (repeatability) and very low 
matching-ratios (< 10% for comparison of 
set 1 “1 on 2” and “1 on 3” and >59% for “1 
on 10”). This was very obvious even without 
the application of any statistical test (e. g. 
Grubbs, etc.). After eliminating the outlier by 
Ignoring the whole run and using just the 
data from the second analysis (Instead of the 
mean, only values from the second run were 
considered, n=1), the comparison turned out 
to be very well In accordance (> 83% for 
comparison of set 1 “1 on 2” and “1 on 3” 
and 100% for “1 on 10”). 
IAEA 
NFI 
<1 
<1 
Set_l 1 on 2 
100% 
83% 
1 on 3 
100% 
100% 
2 on 3 
100% 
100% 
Set_2 4 on 5 
8% 
31% 
4 on 6 
0% 
8% 
5 on 6 
0% 
0% 
Set_3 
Dup 
7 on 8 
7 on 9 
8 on 9 
Ion 10 
92% 
77% 
85% 
92% 
92% 
92% 
83% 
100% 
Tab. 5: IRMS-Conclusion matrix