Linearity problem seems to be solved. It looks like the culprit was the
reduction column. Over the past five or six years we have packed our
reduction column with copper turnings residing in the "hot" zone of the
furnace. To do this we added quartz chips (about 3 or 4 inches) on the
top and bottom of the reduction column during packing. This was
initially done just to conserve copper. However, this September we ran
out of quartz chips and decided to pack the column with all copper.
When this was done it seems that combusted sample "attached" itself to
the cooler copper turnings (maybe?) in the reduction column and carried
over to the next sample. Once we started packing the reduction column
as before (using quartz chips) we were able to get our standard accuracy
back to its target of 245 permil.
I know as we changed reduction columns several times using all copper
that the carry over varied a bit and it was a lot less noticeable (of
course) at natural abundance levels which are fairly close together in
value (-11.2 permil and -26.3 permil vs -11.2 permil and 245 permil).
Since my background is not chemistry if anyone has more detail as to why
this occurs feel free to contact me off list.
Thanks for your inputs.
Director of Instrumentation and Information Systems
Advanced Breath Diagnostics, LLC
105 Westpark Drive, Ste 150
Brentwood, TN 37027
e-mail: [log in to unmask]
On 9/12/2014 3:39 AM, Robert Van Hale wrote:
> Hi, what preparation system is connected to your ms and what phase are your sample materials: solid, liquid or gas? I would bet on prep system effects beating any contribution from "changing linearity." Think carry-over here.
> What do you use as standards? Natural adundance standards would easily provide an extrapolation error of that size. What do you get if you multiply the standard error of your calibration slope by the separation between highest standard and your QC's?
>> -----Original Message-----
>> From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On
>> Behalf Of Shane Crabtree
>> Sent: Friday, 12 September 2014 10:27 a.m.
>> To: [log in to unmask]
>> Subject: [ISOGEOCHEM] enriched 13C linearity
>> Does anyone have experience running enriched carbon samples and have the
>> linearity of their mass spec system move over time? If so, what could be
>> the problem.
>> Currently I have to QC (we made in-house) that are 245 permil and 314.9
>> permil. The lower QC is a mixture but the higher one is a pure
>> substance. The values coming out of the system for the higher reference
>> are coming out around 313.9 permil. The values for the 245 permil QC
>> jump around quite bit more but are around 243.1 permil.
>> The values of my natural abundance samples are spot on.
>> So any ideas what would cause the values to systematically move down at
>> the upper end? Dirty source? New filament (filament is less than a year