|
|
|
|
Beryllium (Be, Z=4) is the lightest element that
can presently be analyzed on an electron microprobe. In the future, it may be possible to
measure lithium (Z=3). Presently, however, the
detector's windows absorb too many of the
ultra low-energy lithium X-rays for it to be practical.
Hydrogen (Z=1) and helium (Z=2) will never be
measured using an electron microprobe because
each has only one electron shell and therefore
they do not produce characteristic X-rays when bombarded with
electrons. One of the biggest problems associated with
doing any kind of beryllium analyses is that the
beryllium X-ray line can undergo an enormous
amount of energy shifting or shape changing.
The plot to the right shows the measured Be Ka
line from three materials, pure Be metal, Be-copper
(with 1.75 wt.% Be), and beryl (Be3Al2Si6O18 ~13 wt.% BeO). |
|
|
The shift is a result of changes in the energy of the outer electron shell of the Be atom due to differences in
bonding. The Be X-ray line in beryl is shifted off about 4 eV from the X-ray line produced in Be metal. In Be-copper, the X-ray
line is more complicated. The peak has not shifted in one direction, but has bifurcated, producing a higher energy peak and a
lower energy peak. These multiple peaks probably occur due to the multiple phases in the alloy, and the fact that the Be in
these multiple phases has different bonding energies. As a result, they produce a Be X-ray line with slightly different energy. |
|
Because of the peak shifts that can occur
with Be, great care must be taken with any
type of analyses. For map or line analyses
it is a matter of making sure that the spectrometer has been peaked for maximum
counts. However with quantitative analyses,
it is important that the peak shape is modeled.
The different phases in the Be-copper alloy
can be seen in the backscattered electron
image to the left. The alloy has segregated
into three phases. There is a light-gray phase
that forms the framework of the alloy; there
is a medium-gray phase which forms most
of the interstitial volume, and there are
inclusions of a dark phase that fills the rest
of the interstitial volume. Since the intensity of the backscattered
electron signal is controlled by the average
atomic number of phase being imaged, the
dark phase must have a lower average atomic
number then the light- or medium-gray
phases. Therefore the various phases of the |
|
|
Be-copper alloy must have significant differences in their element composition. An example line profile across the various phases of the alloy is shown to the right. The profile shows a relationship between the backscattered electron intensity and the abundance of Be and Fe.
Below are quantitative electron microprobe analyses of the different phases. As expected, the dark phase is much higher in Be, which ranges from a low of 1.2 weight percent in
the light-gray phase to almost 6 weight percent in the dark phase. Fe and Co also vary between the phases, but to a much smaller extent. Also, listed in the table below is the certified bulk analyses from the National Institute of Standards and Technology (NIST).
When doing quantitative analyses on a beryllium-bearing phase, care must be used in selecting the matrix correction routine. The different correction routines can produce significantly varied results. These correction routines play a large role in the Be analyses because of the high absorption
of the low energy Be X-rays. The best results will be obtained when the effects of the correction routines are minimized. This can be done by using standards that are close to the unknown's composition. |
|
|
|
|
|
|