Hi Peter,
Did you run your lab standard samples and check their std? Or you may need to
run “zero-delta” test for your masspec.  If your lab standard samples have
good std (normally less than 0.01 for C-13, less than 0.02 for O-18), it means
your masspec is running well.  So, your big std for your soil samples may  be
caused by following factors:
  1)Make sure your CO2 gas is pure enough. If you didn’t purify your CO2 gas
completely, such as H2O exist, it may affect on your masspec measurement
obviously.
  2) Make sure your soil samples have very pure calcite, because if any other
carbonate species exist in your soil samples, such as dolomite, ankerite etc,
may change your sample results largely if a certain reaction temperature
applied.
 3)Please try to run Peak Scanning for your soil sample CO2 gas, see what
types of gas species in your samples gas, by comparing with your lab reference
CO2 scanning results.

Hope these help.

Dr Zheng-Hua Li
Stable Isotope Lab
Dept of Geological Sciences
Uni of Tennessee, Knoxville, TN37996
Tel: 865-974-9622
Fax: 865-974-2368
USA







>===== Original Message From Stable Isotope Geochemistry
<[log in to unmask]> =====
>Dear listmembers,
>
>I would like some comment on the following:
>
>During our 13C in soil analyses we have noticed that the standard deviations
>of some samples were better than others. In the beginning I thought this was
>probably caused by contamination during weighing, so we analysed them again.
>And again the same samples had a better standard deviation than others. All
>samples have +/- the same carbon-content and were measured in the same
>batch.
>
>Of course I know that there is such a thing as a matrix-effect, but can it
>really influence the results that much, even if the weighed amount and the
>'type' is the same (type = all soils, just different fractions and different
>locations)? I am talking about variations in standard deviation (n=3) of
>about 0.1 (sdevs around 0.15 permil for 'worst' samples and about 0.05
>permil for 'best' samples).
>
>Thank you very much!
>
>Cheers,
>
>Peter
>
>
>---
>Peter Vervloedt
>Laboratory of Applied Physical Chemistry
>Faculty of Agricultural and Applied Biological sciences
>Ghent University
>Coupure, 653
>9000 Gent
>Belgium
>
>tel.: +32 (0)9 264 6048
>fax.: +32 (0)9 264 6230
>http://fltbwww.rug.ac.be/isofys
>
>"I have not failed. I've just found 10000 ways that won't work"
>- Thomas Alva Edison (1847-1931)
>---