Hi John,
See my responses below.
"Furey, John S
ERDC-EL-MS Contractor" To: [log in to unmask]
<[log in to unmask] cc:
E.ARMY.MIL> Subject: Re: preservation of DIC samples
Sent by: Stable Isotope
Geochemistry
<[log in to unmask]
U>
01/26/2005 02:44 PM
Please respond to
Stable Isotope
Geochemistry
JF: I note that you find .1% CuSO4 works. Is that after filtering? How does
just filtering (and refrigerating) perform for you? Since Cu complexes with
carbonate etc even if they don't precipitate at some concentration/pH, can
the chemistry affect fractionation down the road? I'm trying to make this
isotope relevant.
DD: I see a 1-2 per mil shift in d13C-DIC between poisoned and non-poisoned
filtered samples after a holding time of about 2 weeks. The non-poisoned
samples show consistently lower d13C-DIC, and higher DIC concentrations of
several ppm C. These water samples have significant DOC concentrations
(2-8 ppm C), thus plenty of potential to create CO2 in solution via
bacterial activity. Copper complexes in solution will be digested to CO2
along with all other forms of inorganic C in the sample upon acidification.
We react 5 mL sample aliquots with 2uL of 5% phosphoric acid at 90 C for 3
minutes, and sparge the samples. The CO2 peak is quantified on an IR
detector, cleaned of other gases, and carried into the mass spec.
JF: The dirt is my interest, the water less so. But in characterizing
what's left behind in the water, the handling is important. For instance
hypothetically you could do reverse osmosis and declare by fiat that what
makes it through the membrane is what is DIC, operationally defined, for
you. You would still have to characterize the recovery and fractionation,
regardless of whether you defined dissolved as passing RO, or .45 um or
whatever.
DD: Yes, DIC is operationally defined by the filter pore size, which
determines particle sizes that end up in the sample. However, there is
no other isotopic fractionation of DIC due to filtering unless significant
CO2 gas loss/exchange occurs during collection and storage. We do our best
to minimize this.
What seems to work for us, recovery with standards anyway, is centrifuging
the VOA vials at 10000 rpm for 10 min, then syringe filtering a clean clear
middle aliquot. Even then, it's usually good thumb pressure (what is that?
20 psi?), discard first mL etc. Of course this is lab handling, not field
handling. But how do I know that this relatively minor handling didn't
affect samples? I don't know. The only possible way, as far as I know, is
to add recovery standards at collection of replicate samples, and there are
still potential issues with partitioning and aging... I digress. But not
really. For instance, we also do heavy-labeled large molecules and have
mm-separate isotope bands. It isn't far-fetched, is it?, that 10 cm of
centrifuged water will exhibit spatial fractionation.
DD: I have heard of this method, but in order to achieve isotopic
separation I believe you need to have a labeled compound as well as
relatively high concentrations of the labeled compound of interest. For
the natural abundance work I am doing, I doubt that ultra-centrifuging
would have any effect on discriminating between isotopic composition of the
carbonate species. I have not tested this.
My point being that "dissolved" is operationally defined. If you dissolve
say NBS19, add some as a minor perturbation to an actual grab sample right
before screwing down the lid, can there be any detection of significant
fractionation due to handling?
DD: My opinion is that fractionation of DIC will be due either to bacterial
activity or CO2 exchange with the atmosphere during handling.
------------------------------------------------------------------------------------------*
|
