FISCHER FISCHERSCOPE X-RAY 5000 Series, Руководство оператора

Страница: 48 / 188

Product Administration Base material in Def.MA
48 FISCHERSCOPE
®
X-RAY
Such features are substantial for WinFTM. They characterize this
program as a leading evaluation tool among other programs in the
X-ray fluorescence business.
Those arguments do not hold if the base material structure is
defined as "composition unknown". In this case the spectrum of the
base material is stored in the application as a spectrum file. Before
the spectrum of the coated sample can be evaluated, the base
material spectrum will be "subtracted" from that of the coated
sample. The base material may be undefined, but it has to be
measured during normalization and calibration. A potential
secondary excitation cannot be recognized and taken into account
because the nature of the base material fluorescence is unknown
to the program. Therefore, in case of a secondary excitation as in
the above mentioned examples, this choice of structure definition
must not be used.
While using "composition unknown" the base material can
implement a wide variety of species. A practical usage of this
feature is the case of unknown or unavailable substrates which can
vary unexpectedly during production or income and can be plated
with a potential intermediate layer of unknown thickness. As an
example we should look to a gold layer on a material from which we
only know that it consists of either brass or copper or zinc, but don't
know if the substrate is plated by a nickel layer or not. In this case
we have to define those elements which change under varying
composition as "interfering elements" (e.g. Ni and Zn) and at the
same time use brass of any kind as a base material during
normalization. With this application a normalization is obsolete if
brass is replaced by copper or zinc or if there is nickel of any
thickness underneath the gold layer.
Remark: While using "composition unknown" the base material of
the parallel application of the calibration standards will be
characterized as "unknown" as well and has to be measured during
normalization and calibration. Consequently the parallel application
calibration standards can be provided with a substrate that differs
extremely from that of the sample. But in case of "composition
unknown" the user has to decide if the substrate of the calibration
standards "fits" with that of the sample. For example, the practical
use of Ni/Fe standards in an Ni/Pd application will not be rejected
by the program but will produce a substantial difference to real
thickness during measurement because of critical excitation of Ni
by Pd-K fluorescence - while there is no such interaction of Ni with
Fe. In contrary a trial definition of the base material in the same way
but using the structure "consisting of one element" will lead to an
error text and will be rejected. This is because the fundamental
parameter method knows of that interaction, checks the base
material convergence and subsequently decides that in this case
Ni/Fe standards cannot be used for Ni/Pd because of secondary