dear list_members,

I am trying to measure Oxygen isotope ratios (d17O and d18O)
in oxygen gas using Europa GEO 20-20 mass spectrometer. I see
that GEO gives correct d18O at high enrichment but  does not
give correct d17O. The details of the problem are as follows:

We know that in O2 gas abundance of mass 32 is  99.52%
mass 33 is 0.08% and mass 34 is 0.4%. The mass spectrometer
gives two ratios 33/32(0.0008) and 34/32(0.004) in sample and
reference gases and finally d33O2 and d34O2 correspond to d17O
and d18O respectively.

We noted that during measurements, the mass spectrometer gives
the correct zero enrichment values when the same O2 gas is put
in both reference and sample sides.

It also gives correct delta value(d17O and d18O near zero)  if
one aliquot of reference gas is measured as sample.

However, when we put enriched O2 sample gas of known composition
it gives correct d18O value, but d17O is around 10-15 permil more
than what is expected, i.e.,measured d17O value-real d17O value~ 10
to 15 permil.

The background for three amplifiers are:
amp 1(major 32 beam)=1.44 *e-09
amp 2(minor 33 beam)=2.68 *e-12
amp 3(minor 34 beam)=2.68 *e-12

The backgrounds are subtracted during analysis.

Typical signals during the measuremnts are:

beam1 =1.6 *e-08, beam 2(mass 33)= 1.13*e-11, beam 3(mass 34)=7.13*e-11


The mismatch in mass 33 is independent of signal strength.

It is a big puzzle to me i.e., why it doesn't give correct
delta 17O in high enrichment(around 100 per mil) region.
I do not have other oxygen gas yet to test the mismatch in 20
to 30 per mil level.

I suspect it could be a nonlinearity problem with amplifier 2
which corresponds to mass 33.

Any suggestion would be highly appreciated.

regards,
antra.


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ms.Antra
Research Fellow
Stable Isotope Laboratory
Planetary and Geosciences Division
Physical Research Laboratory
Navrangpura
Ahmedabad-308 009
India

Phone: Office:   91-079-6302129-4270
       Residence:91-079-6300505
       Fax:      91-079-6301502
email: [log in to unmask]
       [log in to unmask]
URL:http://www.prl.ernet.in/~antra
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~