Now, that S is in the "air" again, I decided to send a version of my notes
to the list members also, previously sent to other laboratories. At USGS,
Reston Isotope Laboratory we have been running S isotope of BaSO4, C and N
of organic compounds, N of soil samples etc. by hundreds now successfully.
Bruno Lavettre at Castech helped me troubleshoot the system, thanks for
that Bruno. Frank Trench was very helpful with the MS, and Peter Stow's
help at IsoMass was priceless with the Conflow interface.
Short summary of running S isotope of BaSO4 on the Carlo Erba EA 2500 -MS
Delta Plus system:
I. EA set up: "Flash point"
The fact that the melting point of the BaSO4 is 1580 C and the reactor
temperature is only 1020 C could create difficulties in the analysis of S
isotope of BaSO4. The "flash point" has a higher temperature, around 1800
C, (since the Tin oxidation reaction is exothermic); therefore it is
critical to use it well by creating a sharp, sudden, bright and single
flash. This should be done by coordinating: 1. He pressure and flow rate,
2. Oxygen pressure and flow rate, and 3. Sample start time. In fact, the
system is so sensitive, that, for example, even a one-second difference in
the sample dropping time can make a difference in the look of the flash,
peak shape and, consequently, the isotope result. The optimum set-up in
our system is: Cycle 100, Oxy 60 (O2 flow stops flowing into auto sampler),
Sample Start 10, and Sample stop: 40. The Helium regulator is set for 150
KPa, He flow Setup is 105 mL/min which measures 120 mL/minute at vent. The
Oxygen regulator is set for 150 KPa, and oxygen flow rate is 54 mL/minute
at Purge. Oxygen loop is 5 mL. Sample size is around 0.2 -0.3mg (BaSO4).
We DO NOT use V2O5 because it is toxic, and it is a catalyst for SO3
production and should be avoided. It might help stabilize the so called
oxygen effect by providing constant sours of oxygen in the combustion, but
I have never saw a study on that. Even then, it might help the accuracy
but not the precision of the result.
II. Leak in the system:
Even a tiny leak could interfere with the result; it could create doubled
peak shape, and unstable run etc. The high He flow rate makes the system
more sensitive to leaks (Bernoulli principal). Leaks could cause changes
in the flow rate in an uncontrollable manner, which could interfere in the
quantitative combustion, in the peak shape, and consequently in the isotope
III. Ashes in the reaction tube.
Ash accumulation is also critical. Although Tin oxide serves as a catalyst
in the reaction, it also partially "clogs" the system interfering with the
flow rate, though not necessarily in a measurable manner. This is,
however, a "controlled interference" causing a shift in delta values. By
running frequent standard materials, this shift can be corrected.
However, after around 50 samples (1 carousel) it is healthy to remove the
IV. Standard gas:
We use SO2 ANHY by Matheson, small cylinder around 15 lb. We do not
have any problem with this gas.
11 psi gives us 1.3 V signal in Cup 1. The length, ID and sometimes OD of
the capillary are critical, so does their location in the conflow. (We all
V. Mass Spectrometer:
Should be focused very carefully (I am not going into details here). It is
temperature sensitive, so it should be operating in a temperature
controlled environment with all the side panels on. We use resistor 1*10E9
for Cup1, and 2*10E10 for Cup2.
Internal precision check:
Zero Enrich method (standard gas on/ off 10 times) should give 0.1 per mill
or better standard deviation, and the Ratio Linearity test should give 0.12
per mill/Volt or better linearity.
External precision check:
Linearity test with same reference material should give standard deviation
0.2 per mill at n=16, between sample weight 0.15 - 0.500 mg BaSO4.
If this is not satisfied, one should try to run the same sample size only.
Precision should be for n=10 0.2 per mill.
If neither of the criteria has been met, the system is useless and the MS
should be refocused.
VI. Sample preparation.
It is recommended to dry the sample overnight at 110 C, and weigh it as
soon as it removed from the oven. Before weighing, careful homogenization
should be done in an agate mortar and pestle. We run 3 same standard
materials first, than alternate 10 unknowns and 2 standards. We use
standards NBS 127 and IAEA SO6, delta S= 20.91 and -32.85 respectively, and
correct our data for slope and intercept. If you have LIMS (Laboratory
Information Management System) for light stable isotopes created by
T.B.Coplen, it can do it for you.
Reference: A.Glesemann et al. Anal.Chem. 1994, 66 2816-2819.
T.B.Coplen, U.S.G.S. Open-File Report 98-284.
Chemist, Stable Isotope Laboratory
Reston, VA 20192