INSTRUMENTATION FOR TXRF AND GI-XRF - Total-Reflection X-Ray Fluorescence Analysis and Related Methods - page 165

In such an EDS, the X-ray photons emitted from the sample are directly
collected by a semiconductor detector. This special detector does not merely
count the individual photons but can also determine their different energies.
For any collected photon, it produces an individual voltage pulse, the amplitude
of which is proportional to the energy of this photon. All the detector pulses are
processed in an electronic measuring chain and finally sorted by a multichannel
analyzer. The content of this counter storage can be represented as the
particular X-ray spectrum and already observed during measurement. After-
ward it can be processed directly by a dedicated computer [42,43].
Reflectivity measurements may be taken by an additional detector, which
can simply be incorporated in the arrangement just described [44,45]. The
primary beam monochromatized by a first reflector is then reflected at the
layered or unlayered substrate. The intensity of this reflected beam can be
measured by a second simpler detector. It may be a photodiode or a scintilla-
tion detector—much simpler and cheaper than a semiconductor detector used
for TXRF. However, it must be ensured that this second detector is tilted at the
double angle 2 α when the layered sample is tilted at the single angle α —both
around the same axis a of Figure 3.13.

3.6.1TheSemiconductorDetector

The heart of any energy-dispersive spectrometer is a special solid-state detec-
tor—or rather a semiconductor detector [46-49]. It basically consists of a pure
silicon or germanium crystal. This crystal ought to be several millimeters wide
and thick and should be extremely resistive. Germanium can be purified by
zone refining to achieve the necessary ohmic resistance. In the last decade,
silicon wafers have also been produced with such a high degree of purity that
impurities of only several parts per billion (ppb) are left. The most common
impurity is boron, which modifies silicon to a p -type semiconductor with
decreased resistivity and increased conductivity. In order to suppress this
effect, another impurity can artificially be added to the crystal. Usually, the
boron“acceptors”are compensated or neutralized by lithium“donors.”
The lithium diffuses into the crystal at elevated temperature and“drifts”under
the influence of an electric field. In this way, a crystal with a high intrinsic
resistivity is produced with a thin p -type layer and n -type layer at the end planes
and a large intrinsic region between them. Such a crystal is termed a lithium-
drifted silicon crystal, or a Si(Li). There are also lithium-drifted germanium
crystals, or Ge(Li)s. However, these are being substituted by HPGe. Today,
both are largely replaced by high-purity silicon flats with a sideward voltage
drift—by so-called silicon drift detectors (SDDs).
As demonstrated in Figure 3.14 for a Si(Li), the frontal areas of the Si crystal
are coated with thin layers of gold serving as electrodes [47]. An inverse DC
voltage is applied, called a reverse bias ( p -type layer negative, n -type layer
grounded). It defines the direction of low conductivity, that is, of a small
leakage current in spite of the high voltage ( 500 to 1000 V). The total