Hello Daniel,

Here in Lyon we have tried to go down in sample sizes for 13C pollen
analyses with our Eurovector EA-IsoPrime system. We managed to get
reliable numbers (SD = 0.2permil) for sample sizes between 2 and 45
micrograms of pollen which contains about 40%C. The system seems to be 
able to handle that amount of material for pollen analyses.
The problem you'll be facing with very small samples is homogeneity of
your aliquote. We did another funny experiment with our EA-IRMS
system analysing wood sample which we did not griund very finely. We 
first measured five 100 micrograms aliquots of our wood for 13C and got 
SD=0.44permil which is rather high for that kind of experiment. We then 
took the same sample and we weighed five aliquotes of 1.2 milligrams, 
measured them using our diluter system and then got SD=0.07permil. 
Obviously there, "small" was not "beautiful".
I also presented results at the last JESIUM meeting dealing with that 
aspect for small, single fossil carbonate analyses where I showed that 
going down in sample sizes, the increase in SD was due to 
heterogeneities of the sample rather than analytical deterioration. So 
my advice is to be careful with the interpretation of your data from 
very small aliquotes. A suggestion, if it is possible, would be to first 
test the homogeneity of your sample with amphipods which are not 
endangered where you have plenty of material to play with.
Regarding the reference material sizes, you'll be facing the same
problem depending on what kind of reference material you use. Of course 
running calibrated material the same way as samples is always good 
analytical practice.Nevertheless our experience with both our IsoPrime 
systems is that they can cope rather well with linearity variations both 
in DI or CF mode even when running small beam
sizes down to 0.5-1nA. That could give you a bit of latitude for the 
reference material amounts.
You can contact me off the list for further discussions.
Hope this helps.

Best regards,
-----------------------------------------------------
François FOUREL

  UMR CNRS 5125 PEPS
  Université Claude Bernard Lyon 1
  2 rue Raphaël Dubois - Bât. Géode
  F-69622 VILLEURBANNE Cedex - FRANCE
  Tél : +33 (0) 4 72 44 62 42
  FAX : +33 (0) 4 72 43 35 88
  Email : [log in to unmask]
  Web: http://webpeps.univ-lyon1.fr/
-----------------------------------------------------


Daniel Nelson a écrit :
> Hi
> I'm working with an endangered amphipod.  I would like to do some isotope 
> analysis (C&N) to look at diet but I can't sacrafice wild individuals. However, I 
> do have a permit to remove walking legs for genetic analysis.  One large 
> amphipod (dried) weighs approximately 4.0 mg and one dried leg weighs 
> approximately 0.185 mg.  What is the smallest sample weight that can be 
> measured accurately?  Can I analyze individual legs, 3 or 4 legs, or doI need 
> to do the whole individual?  I realize the whole individual would probably be 
> best but again, I can't sacrafice individuals.  Thanks a lot.
> 
> Dan Nelson
> MS Student
> University of Idaho
>