2013 Volume 53 Issue 11 Pages 1953-1957
A confocal micro-XRF method combined with two individual polycapillary lenses was applied to steel sheets coated with anti-corrosive paint in order to nondestructively observe 3D elemental distribution of paint steels and corroded paint-coated steels. Nondestructive depth analysis and 3D elemental mapping of the painted steel sheets were demonstrated under the confocal XRF configuration. Three different painted steel sheets were prepared by cation electrodeposition coating for automotive onto flat steel sheets modified with a zinc phosphate conversion coating. These painted sheets were then caused to corrode by means of accelerated exposure to a salt bath (5 mass% NaCl) at 55°C for 240 hours. Depth elemental profiles of Ti, Zn, and Fe obtained by confocal micro-XRF measurements were in excellent agreement with that of the prepared sample. Elemental depth profiles and maps of the corroded painted sheets showed some blisters caused by crevice corrosion, which started from the site of a scratch. The distributions of Ti and Fe were approximately homogeneous in both the paint layer and the steel substrate, while the distributions of Zn, Mn, Ca, and Cl were heterogeneous.
Chemical conversion coating is an essential process for improving the corrosion resistance of steel and the interfacial adhesive properties to paint. Numerous coating techniques for protecting steel from corrosion have been developed. In particular, a zinc phosphating process has been widely used in the automobile industry as a method of pretreating steel surfaces.1,2,3) Some elements, such as Mn, Ca, Si, and Ni, are added to zinc phosphate coatings in order to raise the quality of the phosphate conversion coatings for corrosion resistance. Corrosion beneath a paint layer is a special form of corrosion that causes blistering, cracking, and filiform corrosion. Elemental analysis in a micro region is very important for detecting and evaluating corrosion underneath paint. SEM-EDS and EPMA3,4,5) are often used in the analysis of corroded specimens. However, these methods require destructive pretreatment that peels the paint layer from the corroded steel for measurement because the corrosion beneath the paint is not visible.
Micro x-ray fluorescence (XRF) analysis offers the great advantage of providing nondestructive elemental analysis at an ambient pressure. A polycapillary x-ray (focusing) lens is very useful for obtaining a micro x-ray beam in combination with a laboratory x-ray source. Recently developed polycapillary x-ray lenses enable 2D elemental imaging with high spatial resolution in the order of ten micrometers. Another recent trend of micro-XRF using polycapillary x-ray lenses is 3D elemental analysis. A confocal micro-XRF method proposed by Gibson and Kumakhov6) with two individual polycapillary lenses provides 3D elemental images. Figure 1 shows a schematic diagram of a conventional micro-XRF configuration (Fig. 1(a)) and a confocal micro-XRF configuration (Fig. 1(b)). In conventional micro-XRF, a micro x-ray beam produced by a polycapillary x-ray lens irradiates the sample, and the excited fluorescent x-rays from the sample are detected by an x-ray detector. In this configuration, it is not possible to obtain a 3D elemental image because the irradiated micro X-ray beam penetrates deep into the sample. In a confocal micro-XRF setup, a second polycapillary (half) lens is positioned in front of the x-ray detector. The foci between excitation and detection are adjusted to be at the same point, which is called the “confocal point,” as shown in Fig. 1(b). Thus, fluorescent x-rays are focused by the polycapillary half lens into a small focal volume. Under this confocal configuration, it is possible to obtain XRF intensity of the analyte in a small region leading to nondestructive 3D elemental mapping images.
Schematic diagrams of conventional micro-XRF configuration (a) and confocal micro-XRF configuration (b).
We have developed a laboratory-made confocal micro-XRF instrument using a low power x-ray tube and applied it to many samples.7,8,9,10) In addition, our research group recently developed a new confocal micro-XRF instrument using advanced polycapillary x-ray lenses with a depth resolution of 14 mm at an energy of 11.4 keV (AuLβ).11) Nondestructive 3D elemental analysis using the new confocal micro-XRF instrument was demonstrated on multilayered automotive paint fragments for forensic sample12) and industrial sample such as a micro SD card13) at Osaka City Univ. in Japan.
In this paper, we applied the confocal micro-XRF method for painted steel sheets and painted steel sheets that were corroded. Nondestructive depth analysis and 3D elemental mapping of the painted steel sheets is demonstrated under confocal XRF configuration. In addition, in order to clarify the mechanism of corrosion growth and deterioration of painted steel sheets, we propose a novel method for analyzing painted steel that is corroded.
Three kinds of painted steel sheets labeled “C-10,” “C-15,” and “C-20” having dimensions of 70 mm × 50 mm × 1 mm were measured. A uniform paint of epoxy type coating was applied by cation electrodeposition coating for automotive to the steel sheets modified with zinc phosphate conversion coating. The thickness of each paint coat on the steel sheets was 10 μm for C-10, 15 μm for C-15, and 20 μm for C-20. After being painted, the coated steel sheets were scratched three times in a linear pattern with a line width of about 200 μm. The sheets were then caused to corrode by accelerated exposure to a salt bath (5 mass% NaCl) at 55°C for 240 hours. Figure 2 shows photographs of the corroded painted steel sheets of C-15 and C-20.
Photographs of corroded painted steel sheets of C-15(a) and C-20(b). These sheets were corroded by accelerated exposure to a salt bath (5 mass% NaCl) at 55°C for 240 hours after being scratched three times in a linear pattern with a line width of about 200 mm.
A laboratory-based confocal micro-XRF system was constructed at Osaka City University. Figure 3 is a schematic diagram of the confocal micro-XRF setup. A metal ceramic-type 50W x-ray tube with a Mo anode [MCBM 65B-50, rtw, Germany] was operated at 50 kV and 0.6 mA. The size of the focal spot at the anode of the x-ray tube was 50 μm × 50 μm. Primary x-rays were focused to 10 μm by a polycapillary x-ray full lens attached to the x-ray tube. For the detection side, a polycapillary half lens was attached to an air-cooled silicon drift detector (SDD) [Vortex EX-50, SII Nano Technology, Japan] (sensitive area: 50 mm2, energy resolution <130 eV at 5.9 keV). Both polycapillary lenses were manufactured by X-ray Optical Systems in the USA. The output focal distances of the lenses were 2.5 mm for the full lens and 3.0 mm for the half lens. The size of the focal spot at the focal point of the half lens was experimentally evaluated to be 10 μm at an x-ray energy of 17.4 keV (MoKα). Both polycapillary x-ray lenses were set in the optimum confocal geometry with the angle between the incident and fluorescent x-rays adjusted to 90°. The focal spots of both lenses were precisely adjusted to be at one common point by using an X-Y-Z stage (YA07A-R1, Kohzu Precision Co., Ltd, Kawasaki, Japan) with a precision of 1 μm and a range of translation of ±10 mm. Motor drivers and a motor controller (NT2400, Laboratory Equipment Co., Japan) were used to control the translation stages by using SCAN-2 software (Laboratory Equipment Co., Japan). Fluorescent x-ray signals from the SDD were analyzed by a multichannel analyzer (NT2400/MCA, Laboratory Equipment Co., Japan). The depth resolution of the confocal 3D-XRF instrument evaluated by a thin-foil scanning method was varied from 23 μm to 14 μm for an energy range from 4.51 keV (TiKα) to 11.4 keV (AuLβ). The detailed data are reported elsewhere.11)
Experimental setup of confocal 3D micro-XRF instrument.
To confirm the analytical performance of the developed confocal micro-XRF instrument, we performed nondestructive elemental depth analysis of Ti, Zn, and Fe in the non-corroded C-10 painted steel sheet at a single point. These elements in the sample are contained as TiO2 (white pigment with the rutile structure) for major component of the paint, zinc phosphate for conversion coating, and major component of steel substrate, respectively. Figure 4 shows elemental depth profiles of Ti, Zn, and Fe at a single point of the non-corroded C-10 painted steel sheet measured by the confocal micro-XRF instrument. The sample was scanned perpendicularly to its surface in 2-μm steps; the measurement time was 1000 s per step. Each elemental profile was normalized by the maximum XRF intensities. As shown in Fig. 4, elemental depth profiles of Ti, Zn, and Fe in the C-10 painted steel sheet showed three profiles having a peak at different depths. As compared with the top positions of each profile, the difference between the paint layer (Ti) and the steel substrate (Fe) was about 10 μm. This result was in excellent agreement with that of the prepared sample. In addition, it was found that the Zn conversion coated layer was present on the Fe substrate. However, it was difficult to distinguish the differences of both peaks because the depth resolution of our confocal micro-XRF spectrometer ranged from 14 μm to 23 μm.
Elemental depth profiles of Ti, Zn, and Fe on C-10 painted steel sheet measured by confocal micro-XRF.
First, we performed elemental depth analysis of the corroded painted steel sheet at a single point in the uncorroded region. Figure 5 shows the depth profiles of six elements (Fe, Ti, Zn, Mn, Ca, and Cl) of the C-15 and C-20 corroded painted steel sheets at the uncorroded region measured by the confocal micro-XRF instrument. The measurement condition for the depth analysis was the same to that of the C-10 sheet. Respective XRF intensities of the analytes were normalized with the maximum intensity. Each layer was clearly distinguished in the depth intensity profiles of each element. As shown in Fig. 5, the depth profiles of the six elements of both samples were divided into three layers, as was the case with the result of C-10. The differences of the paint layer (Ti) and the steel substrate (Fe) were 16 μm for the C-15 and 20 μm for the C-20. These results were in good agreement with those of the non-corroded painted steel sheets. Depth profiles of Ca and Mn were almost same as those of Ti and Zn. These results suggested that Ca was added to the paint as a white pigment (CaCO3) and Mn used for the chemical conversion coating of the steel. Depth profile of Cl was similar to that of Ti and Ca. However, the depth profile of Cl on the surface layer was not clear compared to that of other depth profiles.
Elemental depth profiles of Fe, Ti, Zn, Mn, Ca, and Cl at uncorroded point of C-15(a) and C-20(b) corroded painted steel sheets measured by confocal micro-XRF.
Next, elemental depth imaging of the corroded painted steel sheets was performed to observe the elemental distributions in the different layers of the sample. Figure 6 shows nondestructive elemental depth images of the C-15 and C-20 corroded painted steel sheets at the site of the scratch. The minimum step size was 3 × 20 μm, and the measurement time was 10 s per pixel. As shown in Fig. 6, some blisters could be clearly observed and they originated from the scratch. In addition, distributions of Ti and Fe were approximately homogeneous in the paint layer and the steel substrate, while the distributions of Zn, Mn, Ca, and Cl were heterogeneous. In particular, components of the conversion coatings (Zn and Mn) were localized at specific points in the sample.
Elemental depth images of C-15(a) and C-20(b) corroded painted steel sheets at scratch site measured by confocal micro-XRF.
To confirm more detailed elemental distribution, elemental depth analysis of the C-15 and C-20 corroded painted steel sheets at different specific points was performed. Figures 7 and 8 show the elemental depth profiles of C-15 (Fig. 7) and C-20 (Fig. 8) corroded painted steel sheets at two different points. As shown in Figs. 7 and 8, depth profiles of Ti, Ca, and Cl were a single peak, while those of Zn and Mn were two peaks. At the blister point, a significant amount of Zn had moved into the paint layer in comparison with that at the site of the scratch. On the other hand, Mn remained in the substrate. In addition, it seems that the depth profiles of Fe were also a single peak as were those of Ti, Ca, and Cl. However, by obtaining a reconstructed image after adjusting the intensity scale of Fe in the Fig. 6(a) image and the Fig. 7 depth profile, it was found that a slight amount of Fe had moved into the paint layer from the steel substrate (Fig. 9).
Elemental depth profiles of C-15 corroded painted steel sheets at scratch site (a) and at blister site (b).
Elemental depth profiles of C-20 corroded painted steel sheets at scratch site (a) and at blister site (b).
Nondestructive 3D-XRF analysis using a confocal configuration was demonstrated on painted steel and corroded painted steel samples. SEM-EDS (EPMA) is often used for the analysis of corroded specimens. However, these methods require destructive pretreatment to expose each layer for measurement. In contrast, confocal micro-XRF has a significant advantage in 3D elemental analysis in that it enables nondestructive depth analysis and nondestructive elemental mapping. Therefore, we believe that the confocal micro-XRF method will be effective for in-situ analysis of corroded steel.
The results obtained using confocal 3D-XRF give elemental depth information of corroded samples nondestructively. The nondestructive elemental depth analysis of the uncorroded painted steel sheet was at a single point. As compared with each intensity profile, the results obtained from confocal micro-XRF measurements were in excellent agreement with those of prepared samples. To confirm the elemental distributions in the corroded samples, nondestructive depth imaging of C-15 and C-20 corroded steel sheets was performed around the site of the scratch. Depth images of Fe showed approximately uniform distribution within the substrate. However, depth images of the components of the conversion coatings (Zn and Mn) were localized at specific points in the sample. In particular, at the blister site, a significant amount of Zn had moved into the paint layer in comparison with that at the site of the scratch. On the other hand, more Mn remained in the substrate.
In the future, a fundamental approach to this technique will be to improve the spatial resolution of the confocal volume in order to obtain more detailed layer information. In addition, P and Si of the samples could not be detected in our results because our confocal setup did not operate under a vacuum so that fluorescent radiation was absorbed by air before detection by the SDD. Recently, our research group developed a new 3D-XRF instrument equipped with a vacuum chamber to obtain the distribution of light elements that cannot be detected by measurement in air.14)
This work was supported by a JSPS (Japan Society for the Promotion of Science) Grant-in-Aid for Scientific Research (B), a Bilateral Program of the JSPS and FWF in Austria, and an ISIJ (The Iron and Steel Institute of Japan) Research Promotion Grant.
