Dear Ben,

I found that in a non-continuous flow system (i.e. preparation line
and traps not being permanently flushed with e.g. He) the dimensions
of the trap and, hence, the amount of MgClO4 being used can
negatively affect accuracy of d13C measurements. However, this was
caused by a memory effect and not by fractionation. Reducing the net
volume of the trap by two thirds eliminated the memory effect (18" x
1/4" down to 6" x 1/4").

The influence of different drying agents on d13C of CO2 has been
published in "Isotopes Environ. Health Stud." 30, 349-358 (1994).

If you like I can send you a copy (please mail your address directly
to me and not over the list).


Cheers,
                    Wolfram




*****************************************************
Dr. W. Meier-Augenstein
Lecturer/Research Fellow

University of Dundee, Dept. of Anatomy & Physiology,
Small's Wynd, DUNDEE  DD1 4HN, United Kingdom

Tel.: +44-(0)1382-34/5124, /4968
      +44-(0)468 -314563
Fax:  +44-(0)1382-34/5514

e-mail: [log in to unmask]

URL: http://www.dundee.ac.uk/AnatPhys/wma/wolfram.htm
*****************************************************