Page 206 - Inorganic Mass Spectrometry - Fundamentals and Applications

I92 Cristy

Thus the SIMS analyst must be aware of the definition of RSF being used when
reading the literature.
An excellent way to create standards is ion implantation of the elements of
interest into the matrix. This works exceptionally well for se~conductors since
one can usually start with high-purity single-crystal materials that represent the
matrix of interest, Also the use of Eq. (4.8) is well suited for this purpose since ion
implanters usually quote doses in atoms per square centimeter. However7 Eq. (4.5)
serves just as well by converting the matrix concentration to atoms per cubic cen-
timeter. In this procedure, the implant profile is sputtered through, the implant ele-
are
ment secondary ions and the matrix element secondary ions each summed, and
the depth of the sputter profile is determined, usually by using a stylus profilome-
ter. The sensitivity factor is then calculated from

where 21i and 21paa are the sum of the counts of the implanted impurity ion and
matrix ion, respectively; d is the depth of the sputtered profile in centimeters; pm
is the atom density of the matrix element in atoms per cubic centimeter; and QI is
the
the implant fluence in atoms per square centimeter. Usually implant ion is iso-
topically pure, so no correction for its isotopic abundance is shown in this equa-
tion. Again the RSF, Si,m, is unitless. Note that one needs to know the atom density
of the matrix element in this application.
If one uses the RSF as defined in Eq. (4.8) for the ion implant calibration,
the expression becomes

(4.10)

where dl symbols are defined as earlier and RSF has units of atoms per cubic cen-
timeter.
In SMS, small changes in concentration have nonlinear effects on the rela-
tive sensitivities of the components; changes in vacuum environment or current
density of the primary ion beam also affect the relative sensitivities. Because it is
unreasonable to have precisely matching standards for every matrix and because
precise matching of instrument and vacuum parameters, although desirable, is
often difficult to achieve in practice, empirical methods to extend and adjust the
of
sensitivity factors (detemined from a reasonable number standards) are impor-
tant, ~cHugh suggested [64] that sensitivity factors would show a dependence on
E,, a parameter reflecting the electronic properties of the secondary ion-emitting
surface. Then, by plotting RSFs versus E,, derived from a few standard matrices,