The analysis of Mo5SiB2 in the SEM with the use of EDS and WDS

Abstract

Refractory metals and their alloys show potential for high temperature applications due to their increased melting point and creep resistance. Mo‐Si‐B ternary alloys consisting of the phases Mo ss (molybdenum‐based solid solution)‐Mo 3 Si (A15)‐Mo 5 SiB 2 (T2), with melting points over 2000 °C, are particularly favorable for new high‐temperature materials. However these alloys show a lack of oxidation resistance in the intermediate temperature range, 650‐750 °C, and possess a relatively high density (9.6 g/cm 3 ) compared to Nickel‐Based Superalloys. The characterization of the Mo 5 SiB 2 phase with an SEM with X‐ray microanalysis analytical capabilities presents a real challenge. The B‐Kα X‐ray has an energy of 183.3 eV, but might shift slightly due to its bonding with Mo. This X‐ray line energy is very close to the Mo‐Mζ line which is at 192.6 eV. In addition, the K‐shell absorption edge of boron is at 192 eV, just below the Mo‐Mζ line. Finally the absorption coefficient of B‐Kα in Mo is very high, and is even more extreme in Si. As a result, EDS microanalysis proved to be insufficient. The EDS spectrum of the Mo 5 SiB 2 phase shows a very small peak at the B‐Kα position, but without any separation from the Mo‐Mζ line. The elemental distribution maps even showed a strong artifact: due to the higher Mo content in the Mo ss phase and the absence of a boron absorption edge the Mo‐Mζ peak is considerably higher in that phase, resulting in an incorrect increase of intensity of boron in the Mo ss phase. See figure 1. Alternatively, WDS spectrometers have a much better energy resolution and are capable of separating to a large extend the B‐Kα and Mo‐Mζ peaks. Modern parallel‐beam WDS spectromers are also very sensitive to the low‐energy part of the spectrum, and can detect small amounts of boron with very high efficiency. A careful energy scan over the B‐Kα and Mo‐Mζ peaks can be seen in figure 2. And of course creating an element distribution image for boron now indeed showed the correct boron distribution, as can be seen in figure 3. The final challenge lies with the quantitative analysis: what exactly is the weight percentage of boron in the supposed Mo 5 SiB 2 phase? The B‐Kα and Mo‐Mζ peaks still partially overlap even with WDS, and one has to be very careful in the selection of the background support points to correctly subtract the background X‐ray intensity. Using a stoichiometric Mo 5 SiB 2 standard this can still be done, and an accurate quantification can be performed. This case shows how vital it can be for certain applications to widen the range of available microanalysis tools with a parallel‐beam WDS spectrometer to perform analyses beyond the performance limit of EDS.