Hi Paul,


The time shift you are observing is perfectly normal and I can only say
"I told you so". Not you personally of course but the stable isotope
community incl manufacturers at large be it at conferences or be it in
publications (J Chromatogr A, Anal Chim Acta, and, most recently in
Pier's "Handbook of SIA Techniques").

The moment 2H compound specific isotope analysis was announced as the
next best thing to sliced bread, I warned about the problems arising
from the chromatographic isotope effect. Every analytical lab carrying
out quantitative GC-MS analysis by the Stable Isotope Dilution
technique using poly- and per-deuterated internal standards is
exploiting this effect on a daily basis.

The degree of time displacement depends on the nature of the compound
and on chromatographic parameters such as polarity of the stationary
phase, column temperature, and carrier gas flow. So, by choosing "bad"
GC conditions, one can suppress this effect to a certain degree (usually
at the expense of good peak resolution elsewhere) but to misquote
Scotty: "Ye cannae change the law of physics, captain".

The chromatographic isotope effect is not caused by a vapour pressure
effect but is the result of different solute/stationary phase
interactions that are dominated by Van der Waals dispersion forces
leading to an earlier elution of the heavier isotopomer (hence, "inverse
isotope effect" as it seemingly defies conventional wisdom). This
difference in chromatographic solute/stationary phase interaction is
caused by the lower molar volume of the heavier isotopomer as compared
to its more abundant lighter analogue. The reason for the decrease in
molar volume of deuterated molecules is the increased bond strength and
thus shortened bond length between 12C-D and 13C-D as compared to 12C-H
and 13C-H. Since most Hs (and Ds) are sitting on the outside of an
organic molecule, the volume contraction for poly- and per-deuterated
compounds is much more pronounced than for 13C-labelled compounds.

Bond length (and bond strength) between two atoms is a function of their
reduced mass. While the reduced mass for 12C-2H is almost twice that of
12C-1H (85.76% higher), the reduced mass of 12C-13C is only 4% higher
than that of 12C-12C.

At natural abundance, the resulting chromatographic isotope effect (due
to partial deuteration) is more noticeable for small hydrocarbons such
as those you are investigating.

If you would have chosen to analyse longer chain HCs (e.g. >C16), you
probably wouldn't have noticed, i.e. the effect would have been of the
same order as we are used to from 13C analysis of molecules with a
molecular weight of about 150 Dalton (give or take).

Theoretical considerations (back of the envelope stuff) show that one
should be able analyse 2H labelled compounds by CSIA provided they have
molecular weight of >250 Dalton and don't carry more than 2-3
deuterium's, as this should result in a time-shift of 1s or less.


I don't really have a solution for this (sorry). I'd imagine one could
export the peak data into Excel and do the ratio calculation "by hand"
as it were. Have a chat with Ian Begley at Iso-Analytical who I think
has some experience in doing that sort of thing (Ian Begley
<[log in to unmask]>).


Cheers,

Wolfram



> -----Original Message-----
> From: Stable Isotope Geochemistry
> [mailto:[log in to unmask]] On Behalf Of Paul Eby
> Sent: 17 November 2004 18:02
> To: [log in to unmask]
> Subject: HD by GC-IRMS
>
>
> List members,
>
> Our lab has been working on hydrogen isotope measurements by
> GC. Specifically, we have been measuring hydrocarbons
> anywhere from methane up to C10, on a Delta Plus XL.
>
> The problem is that extreme time shifts can be found between
> the mass 2 and mass 3 signals: upwards of 2 seconds. This is
> using the same columns and injection methods that we would
> for d13C, for which the time shift is almost negligible.
>
> I've been informed by Finnigan though that Isodat isn't
> capable of handling time shifts that large. In order to stop
> the software from confusing adjacent peaks as a time shift,
> they have imposed an artificial limit on the time shift - it
> can only be as large as the integration time (default is 0.25
> seconds).
>
> Finnigan is apparently working on a fix for this, but in the
> meantime I am looking for other solutions. Increasing the
> integration time, for instance, has limited usefulness as it
> also changes the integration and at 2 seconds misses most of the peak.
>
> Have others experienced this large time shift? Is it simply
> related to chromatographic efficiency, or are there other
> factors that are involved? Why do these large shifts appear
> when measuring hydrogen but not carbon?
>
> Paul Eby
>
> [log in to unmask]
> Biogeochemistry Lab (E-Hut, room 102)
> School of Earth and Ocean Science
> University of Victoria
>
> PO Box 3055
> Victoria, B.C.
> Canada
> V8W 3P6
>
> phone: 250-721-6183
> fax:   250-472-4620
>