Geoscience Reference
In-Depth Information
poised to become the prime method for
in situ
analysis of isotopic compositions of
carbon or oxygen.
Once the atoms from the sample are ionized, the ion beam is then accelerated at several
thousand volts and spread in mass in a magnetic prism. The beam corresponding to each
mass is normally collected in a Faraday cage, which often is a carbon-coated ceramic
box attached to a high-value resistor, typically
R
10
11
=
(ohms), and a voltmeter. By
using Ohm's law, the reading of the flux of ionic charges
I
, i.e. the electric current, can be
replaced by the reading of the potential difference
V
RI
, which is usually of the order
of one volt. When the ion beam is too weak, the analog current reading is replaced by the
counting of the individual arrivals of charged ions. Modern techniques allow this to be read
simultaneously from several beams corresponding to different masses (multiple collection)
and so make for more precise measurements.
Two remarks are called for so as to judge the relevance and quality of a method. First,
for all the methods discussed, atoms are in competition for ionization. A mixture of atoms
practically never gives an acceptable ionization yield. Except for ionic analysis, which
does not allow it, we seek to isolate the element to be analyzed from the other components
(matrix), while taking care to maintain a very high chemical separation yield. This is why
complex separation procedures must be carried out (extraction lines for gases and stable
isotopes, ion exchange on resins in solution for thermal ionization and plasma sources)
before performing mass spectrometric analysis.
This brings us to the second point, which is the requirement of precision. While a preci-
sion of 1% can give good-quality information for the isotopic analysis of argon, precision
better than 10
−
3
is required for meaningful lead and osmium isotopic data, and better than
10
−
4
for Hf, Nd, and Sr. How do we judge the required precision? A counting process is a
Poisson process, i.e. the probability of an event (arrival of an ion) occurring depends only
on the duration of measurement and not on the moment at which it is made. For this type of
process, the standard deviation of measurement (which yields the precision) is equal to the
square root of the number of events. If, say, we count a million ions, the standard deviation
of the measurement will be 1000 counts and the relative precision of counting therefore
will be one per 1000. Understandably then, this precision depends mainly on the number
of ions analyzed. The flux of particles cannot be increased indefinitely as the ion, electron,
and photon counting devices lose their linearity for a few hundred thousand strikes per
second and are totally saturated at a few million. At the extreme limit of analytical pos-
sibilities, secondary ion mass spectrometers (SIMS), such as the SHRIMP or the Cameca
IMS 1280, provide precise and accurate isotope compositions of Pb and U/Pb ratios on an
spot size of a few tens of micrometers. Given the current state of sensitivity of instruments,
it would be pointless, for example, to hope for a significant measurement on the same
impact of the isotopic composition of Nd, as the quantity of the element required would be
more than a hundred times greater than that required for lead. The elementary processes
of physics must be comprehended and measurement can only be refined by improving the
efficiency of ionization.
=
