X-Ray Elemental Mapping of Multi-Component Steels.
- PDF / 2,364,294 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 12 Downloads / 131 Views
This paper also deals with the problems associated with detecting such small precipitates in multi-component steels where there are overlaps between the characteristic Xray lines from the various elements. EXPERIMENT TEM specimens of a 3.5NiCrMoV low-carbon steel were made by manually polishing 3 mm discs to a nominal thickness of 40 pm. The discs were then ion beam thinned at an angle of 40, using a Gatan precision ion polishing system (PIPS) until a hole appeared. The analysis of the chemical composition was carried out using a VG HB603 FEGSTEM giving a probe size of 1.4 nm (FWTM) with a beam current of 0.5 nA. The STEM uses a windowless Si(Li) X-ray detector, which has a large solid angle of detection (0.3 sr). X-ray acquisition was carried out on an Oxford (Link) exl system, where elemental windows were defined over the K. lines of the following elements, C, 0, Al, Si, P, V, Cr, Mn, Fe, Ni, Cu, Mo and the La line of Mo. Two normalizing backgrounds were defined at 3.3 to 3.8 and 10.0 to 12.0 keV energies respectively. Elemental X-ray maps had an acquisition time of 100 ms per pixel with a 128x128 pixels resolution. Digital line-scans had a scan length of 64 nm with a total of 64 spectra being taken along the length of the scan. Each spectrum in the line-scan had an acquisition time of 5 seconds, and was normalized with respect to the background to remove the effects of thickness. RESULTS AND DISCUSSION The acquisition of elemental maps by the use of energy windows has been used for many years, and a typical use of this type of map is shown in Figure 1.
N.,;
Figure 1. Elemental X-ray maps for Ni, Mo, Cr and Mn, with corresponding BF image. The Ni and Mo maps show that the PAGB is not uniform in composition, which changes as the boundary plane changes. The four maps in Figure 1 are very effective in showing the precipitates that are not visible in the BF image, and that Mo and Ni have segregated to different parts of the PAGB. There are problems with this type of mapping, and it is these problems that are addressed in this paper. It must be noted at this stage that, modem developments in spectrum imaging, 88
especially "position tagged spectrometry" [6] will in the future overcome most of the problems. Unfortunately such techniques have been developed for use in the SEM and not for the high spatial requirements of STEM EDX microanalysis, in which X-ray counts are limited. The first problem is that there are no compositional look-up tables to these maps, i.e. the relationship between intensity and composition is not defined. Other work has shown that there is 3.2 ±0.3 wt%/o Ni in solid solution with the matrix Fe but the simple window map can not give this information. X-ray maps produced by this method can be successfully quantified as shown by Williams et al. [7]. Unfortunately, the method of Williams et al. requires the removal of the background at each pixel and no conflicts between the characteristic X-ray lines. This highlights the second problem with the X-ray maps shown in Figure 1. Comparing the Cr and
Data Loading...
