2022 Volume 62 Issue 5 Pages 811-820
This paper reviews three emission spectrometric methods that can be applicable to the inclusion analysis in steel materials, which provides information on the number, the size and distribution, as well as the chemical composition of inclusion particles. Cathodoluminescence (CL), X-ray-excited optical luminescence (XEOL), and beam-scanning laser-induced breakdown spectrometry (LIBS) are employed for the imaging measurement of the inclusions. As a typical specimen, alumina inclusions are evaluated to compare their analytical performance. The easy handling and rapid response of them are significant features for an application to the on-site/in-line analysis in the production site of steel.
It is the most important information for determining the chemical composition of a sample that the radiation in a wavelength range from visible to X-ray is emitted from condensed phases as well as atomized species of the analyte elements. Atomic emission emanated from their excited states, which are derived from the outer electron orbitals, is observed and analyzed in most of the cases, because it enables the quantitative analysis to be conducted more accurately and precisely. Plasma emission spectrometry, such as radio-frequency inductively-coupled plasma - optical emission spectroscopy (ICP-OES), is a typical analytical tool for the determination of alloyed and impurity elements in steel except for gaseous ones over a wide concentration range from several ten % down to a few ppm.1,2) ICP-OES has contributed to the development of advanced products in the steel-making industry.3,4) On the other hand, plasma emission spectrometry using spark discharge plasma (SD-OES), which is frequently named as QuantVac (QV),5) has been employed for on-site/in-line analysis in the steel production, and it advances specially for the quality and process control for steel products.6) While ICP-OES generally requires the pretreatment of samples including the acid decomposition and aqueous dissolution, SD-OES can be adopted to the direct determination of elements in solid steel samples, which is a major reason why this analytical method has played an important role in the steelmaking process.6)
The steel production needs a variety of the analytical information which is obtained by not only the elemental (qualitative/quantitative) analysis but also the state analysis.7) The state analysis provides chemical, physical, or metallurgical information to clarify how constituent elements and compounds contain in the matrix of steel materials.8) The result of the state analysis sometimes gives a determining solution in the material design, since the physical/chemical forms of containing elements as well as their amounts in a steel material may play significant roles in the function of the material. For instance, an oxide layer formed on the surface of steel determines the corrosion resistance, and oxide inclusions precipitated in a steel material determines the mechanical strength. Due to various kinds of the information that should be obtained, different analytical methods are applied for evaluating such properties. Generally, ICP-OES provides poor information on the state of samples, because such information may be lost through the acid decomposition of them. However, emission spectrometry using other plasma sources is now employed in the state analysis of actual steel products. The steelmaking process now employs the pulse discrimination analysis (PDA) in SD-OES,9,10) in order to obtain the distribution in size and amount of inclusion particles, such as alumina, embedded in the metallurgical structure of steel materials, which may become a key factor for determining the performance of the final product. The PDA method is evaluated from a variation in the intensity of an analytical emission line per each spark shot, based on a statistical model.9) Another case that atomic emission spectroscopy is utilized in the field of steel state analysis is depth profiling in glow discharge plasma - optical emission spectrometry (GD-OES). GD-OES has the characteristics suitable for direct analysis of solid samples, enabling the rapid, simultaneous, and multi-elemental determination.11) Moreover, since the sample atoms are introduced into the plasma by cathode sputtering, there is an analytical feature that variations in the chemical composition can be measured in the depth direction, which is called a depth profile.12) The depth profile of surface coatings, such as oxide scales and electroplated layers on steel materials, is measured in-site using GD-OES for the quality and process control.13,14)
Luminescence spectroscopy provides unique information in steel analysis, which differs from that in atomic emission spectroscopy such as SD-OES and GD-OES. Luminescent emission induced by irradiation with a high-energy electron beam is observed the most commonly, called cathodeluminescence (CL), and a similar phenomenon is also caused by irradiation with an X-ray beam, called X-ray-excited optical luminescence (XEOL).15) Such luminescent effects are attributed to electron excitation/de-excitations causing between the valence and conduction bands of insulator-like compounds, such as oxide, carbide, and nitride.16) The actual mechanisms are much complicated depending on the kind of emitting compounds, which relate to defect levels and impurity energy levels located in the band gap.17) Therefore, the wavelength of the emitted radiation is peculiar to the compounds which can be applicable to their qualitative/quantitative analyses. Their characteristics also enable the size, shape, and chemical components of the heterogeneous structure in a sample to be analyzed using the imaging measurement of the luminescent radiation. Inclusion particles in steel materials are a suitable analytical target for CL and XEOL observations.
When a solid surface is irradiated with a high-energy laser beam, a part of the surrounding gas atoms/molecules is ionized and then becomes a plasma state, which is called laser breakdown followed by laser-induced breakdown plasma.18) At the same time, the thermal evaporation of sample atoms occurs on the sample surface due to the incident energy of the laser beam, and eventually sample atoms are ejected towards the plasma. This phenomenon is called laser ablation and is used for sampling of analyte atoms in the spectrometric method. Laser-induced breakdown - optical emission spectroscopy (LIBS), which is based on the excitation/de-excitations of analye atoms in the plasma, can be adopted to the qualitative/quantitative analysis of a solid sample, like SD-OES and GD-OES. LIBS has several features suitable for the direct analysis of solid samples, such as on-site/in-line analysis in the steel production, in which an as-received specimen can be analyzed with little pre-treatment without any direct contact with it. Accordingly, the analytical procedure can be conducted easily and rapidly. Furthermore, LIBS can provide the spatial distribution of elements in a sample by scanning the laser beam on the sample surface and synchronously recording the intensity of an emission line,19,20) which is analogous to the image observation by CL and XEOL. Thus, inclusion particles in steel materials may become a suitable analytical target also in LIBS.
This article represents a brief review on recent advance of the imaging observation in emission spectrometry, especially for the application to steel analysis. In this topic, several studies on the inclusion analysis of steel materials are introduced when CL and XOEL spectrometry, and a scanning LIBS method are employed as the analytical tool.
Quantitative and morphological evaluation of inclusion particles, which are embedded in the metallurgical structure of steel, is an important issue in the analysis of steel products, because they generate various types of defects in them and thus degrade the quality of the final product.21,22,23,24) For example, non-metallic inclusions may reduce the fatigue life as well as the stress-strain characteristics at elevated temperatures. Non-metallic inclusions in steels mainly exist in the form of oxides (e.g., Al2O3, SiO2, MnO, CaO, MgAl2O4), sulfides (e.g., MnS, CaS, FeS, TiS), and nitrides (e.g., AlN, TiN, BN).25,26) Such inclusions may be formed during the de-oxidation process of molten steel by the added elements or through entrainment of slag and refractory materials into the molten steel. So-called ‘hard’ inclusions, such as coarse Al2O3 particles, may cause any mechanical fatigue of structural steel materials and sometimes leave any scratches and creases on the surface of the final product, depending on the number density and size of the inclusions.27,28) Particularly, we should note large inclusions having a size of more than 100 μm, because they seriously deteriorate the mechanical properties of steel products. Thus, the allowable size of inclusion particles is regulated in several kinds of the steel products; for instance, the maximum inclusion sizes in specialized steel for automobile production are 100 μm for the sheeting, 5 μm for the lead frame, 15 μm for the ball bearing, and 20 μm for the wire.22) Furthermore, the negative effects of inclusions also depend on their shapes and the distribution in the iron matrix, owing to differences in their mechanical properties such as elasticity coefficient, adhesion with the steel, and thermal expansion coefficient. Therefore, it is an important task in steel analysis that various properties of inclusions in steel products are clarified and that such information can be acquired easily and rapidly, especially in the on-site/in-line analysis.29) For this purpose, inclusions in steel materials should be investigated in more detail by developing noble analytical methods.
2.2. Evaluation of Inclusion ParticlesFigure 1 shows a schematic presentation of the analytical methods that are now employed for the inclusion analysis of steel materials. Direct observation of inclusion particles is conducted by viewing with an optical microscope, which is a referee method in the Japanese Industrial Standard (see Fig. 1(a)).30) The size and spatial distribution of inclusion particles having μm-order and the larger diameters can be measured precisely; however, this procedure accompanies very time-consuming works, including manual counting of the inclusion number, such that it cannot be applied to the on-site analysis in steel production.
Explanation on the analytical methods for inclusion particles in steel: (a) microscopic observation, (b) SEM-EDX (WDX), (c) CL and XEOL, and (d) scanning-beam LIBS. (Online version in color.)
The size and shape of the inclusions, together with their chemical compositions, are usually identified using scanning electron microscope (SEM) equipped with an energy dispersive X-ray detector (EDX) or with a wavelength dispersion X-ray detector (WDX).7) This technique is based on the phenomena caused by irradiating a focused electron beam onto a sample surface. It provides an image of secondary/reflected electrons for an inclusion particle as well as a characteristic X-ray signal of the constituent elements which contributes to the elemental analysis, as presented in Fig. 1(b). SEM-EDX (WDX) has become a common analytical technique using commercially-available apparatuses owing to wider applications of SEM, and electron probe microanlyser (EPMA), which works under a similar operation principle to SEM-EDX, is also employed for the inclusion analysis.31) Inclusions of more than several micrometers in size are counted from the SEM image, and then each marked inclusion is assigned to a particular kind of compound from the characteristic X-ray. These analytical procedures are also time-consuming because a target particle has to be found in a microscopic view of the SEM image; therefore, it is difficult to apply them to the on-site/in-line analysis.
The PDA method in SD-OES is now carried out for rapid analysis of inclusions in steel, since the size distribution of a particular inclusion kind over a wide observation area can be estimated from the pulse-like emission signals using a statistical model.32,33) However, this method cannot be available for a precise identification of the size and shape of each inclusion because the emission signal does not necessarily correspond to an individual particle.
CL spectrometry provides the emission image of inclusion particles in steel. It is a major advantage beyond the other analytical methods that an image of inclusions in a certain sampling area can be obtained like an instant photograph using a digital camera, enabling to identify the size and shape of the inclusion particles as the whole figure, as illustrated in Fig. 1(c).16) Also, XOEL becomes a rapid analytical tool for the inclusion analysis in steel; in addition, the X-ray irradiation does not require a vacuum atmosphere for the measurement, applicable to the on-site analysis in the steelmaking process.15)
In LIBS associated with a scanning laser beam, the resultant micro plasma is sequentially produced with an interval of several 10 μm by irradiation of a pulsed laser (the repetition frequency of 10 to 103 Hz), which allows an elemental map of inclusions to be obtained, as illustrated in Fig. 1(d). This observation is attributed to the fact that the sampling area for each laser shot is within the micro plasma, in which constituent elements of an inclusion particle are excited/de-excited to emit the characteristic atomic spectrum. The analytical procedure can be conducted simply and rapidly, where an as-received solid sample can be analyzed with little pre-treatment and without any direct contact with it. LIBS has several features suitable for the on-site/in-line analysis in the steel production.20)
Table 1 indicates a comparison in several characteristics between SEM-DEX, CL, XOEL, and LIBS, with respect to the analytical performance.
| Methods/Characteristics* | SEM-EDX | CL | XOEL | LIBS |
|---|---|---|---|---|
| Analytical sensitivity | ○ | ○ | △ | △ |
| Analytical precision | △ | △ | △ | ○ |
| Analytical accuracy | △ | × | × | ○ |
| Multi-elemental determination | ○ | △ | △ | ◎ |
| Lateral resolution | ◎ | ◎ | ○ | △ |
| In-depth resolution | × | × | × | △ |
| Without vacuum condition | × | × | ○ | ○ |
| Handling of equipment | △ | ○ | ◎ | ○ |
| Portability of equipment | × | ○ | ◎ | ○ |
| Robustness | △ | △ | ○ | △ |
We can understand that the luminescent radiation is derived from an electron transition between the valence and conduction bands that are reconstructed from the outermost electron orbitals of a solid-phase material. While wavelength of the luminescent radiation is principally determined by the band gap, the actual mechanism is much more complicated due to a strong dependence on the type of the emitting material. There appears no luminescent radiation in metal and metallic compounds, because overlapping or very small gap of the energy bands results in few channels for the electron transitions. This effect can yield an important advantage in the measurement of the luminescent radiation when inclusion particles are detected in a metal matrix such as steel, because the emitting radiation can be observed with very low background of the matrix material. In semiconductor materials, there are direct transition paths between the valence and conduction bands, for which the energy gap corresponds to the photon energy in the visible/ultraviolet wavelength region. However, a major de-excitation process occurs not with light emission but with heat emission, in the band transitions of semiconductor materials except for several compound semiconductors.
An important source for the luminescent radiation in materials analysis is various electron transitions within the band gap of insulator-like materials. The direct transition between the energy bands is a minor transition path for the luminescent radiation due to the wider band gap of an insulator; nevertheless, luminescent emission can be observed in many insulator materials, which may be available for the inclusion analysis in CL and XEOL spectrometry. Some insulator materials have various types of intermediate energy levels between the valence and conduction bands, which play a role for the initial and final states of the luminescent radiation especially in the visible wavelength region (their energy gaps of less than 3 eV). Such radiation may provide important information for the inclusion analysis in steel materials. These intermediate energy levels are originated from complicated phenomena occurring in the crystalline/metallurgical structure as well as the chemical composition of the insulator material. Figure 2 illustrates an energy level diagram in the band structure of aluminum oxide (Al2O3) schematically and simply, which represents typical excitation/de-excitation mechanisms for the luminescence. Several defect energy levels, such as an oxygen vacancy in the crystalline structure of aluminum oxide, are in electron transition channels of the luminescent radiation having a characteristic wavelength. In addition, several impurity elements such as trivalent chromium ion, which are included by replacing the crystalline position of aluminum ions in the matrix oxide, also produce the energy levels for the luminescent radiation having a different characteristic wavelength. The luminescence derived from the impurity energy level generally has higher emission intensity than that from the others, because the impurity element behaves like an atomized free atom in the crystal field of the host substance, and thus conserves discrete energy levels to which the excitation might be concentrated. In fact, the luminescent spectrum and the intensity may varies depending on the crystalline state of the matrix substance including defects and impurity elements; therefore, we should further investigate the excitation mechanism to apply the luminescent phenomenon to the inclusion analysis more accurately.
Model of excitation/de-excitation paths for CL and XEOL phenomena in the band structure of alumina.
When a solid surface is irradiated with a laser beam having a high energy density, a large number of energetic electrons are supplied to the surrounding gas near the surface and then electron collisions occur as a chain reaction, producing a plasma state above the irradiated surface. This phenomenon is called laser-induced breakdown plasma (LIP). At the same time, atoms are ejected from the surface due to the incident energy of the laser beam, and eventually the atoms are introduced into the plasma. This phenomenon is called laser ablation and is used for sampling of analyte atoms in the spectrometric method. LIBS measures the atomic emission from the plasma, which is the same excitation mechanism as other plasma sources for atomic emission spectrometry. Figure 3 schematically illustrates excitation/de-excitation channels occurring for emission of aluminum, where the atomization, excitation, and ionization are simultaneously caused in the plasma by the irradiation energy of laser. Differing from stable and continuous plasma like ICP, LIP is produced as a transient phenomenon in the plasma due to a pulsed operation of the laser. Therefore, the emission spectrum varies spatially and temporally following the creation/extinction of the pulsed plasma. The emission intensity of the sample atom changes with the progress of the plasma and generally reaches a maximum value immediately after the plasma generation. This is due to the fact that a large number of high-energy particles are generated in the early stage of the plasma and actively repeat excitation collisions with sample atoms. However, since there remain gas particles that cause the excitation collision even during the expansion period of the plasma, the emission spectrum of the sample atom is still observed.34,35) In the LIBS method, it does not necessarily provide a good analytical result that the time-dependent emission intensity is integrated totally over the entire transient region. This is because, immediately after the breakdown, a large number of high-speed electrons cause bremsstrahlung when they collide with gas particles. This effect contributes to high background emission that interferes with the measurement of the emission intensity from sample atoms. Further, in such a high-density plasma body, it may be problems that the emission spectrum is degraded by the pressure broadening of the line profile and the self-absorption phenomenon. While the time-integrated measurement is usually carried out in atomic emission spectrometry, a time-resolved measurement method is recommended in the LIBS method. The delay time and gate width for such a measurement should be optimized to improve the analytical performance.36,37)
Schematic energy level diagram for the atomic emission of aluminum atom excited in laser-induced plasma.
Cathodoluminescence (CL) analysis, which is based on the phenomenon of light emission from materials as a result of electron bombardment, was applied to the identification of inclusions in steels38,39) and phases in materials related to steelmaking such as refractories40,41,42,43) and nozzle clogs.44) This analytical method can be used to simultaneously identify the size, shape, and chemical components of inclusions in steels by capturing their CL images using an optical camera more quickly than the conventional methods, even though CL analysis requires vacuum conditions for generating the electron beam.
Kaushik et al. published a research report regarding the inclusion characterization that was actually carried out in a steelmaking plant.38) Cathodoluminescence microscope, in which inclusion particles were observed as an emission image, was applied to evaluate remelted buttons of the molten steel and nozzle clog materials in the casting process in the steel production, indicating that the kinds of the inclusions could be identified from the emission color and that their size and amounts could be estimated from the emission image. Yin et al. reported that inclusions which caused sliver defects in steels, such as Mg–Al spinel and CaO–MgO–Al2O3 inclusions, were determined from luminescence colors of CL images for the steel surface.39) Therefore, the CL measurement provided useful knowledge on a rapid screening for the inclusion analysis for the process control.
Imashuku et al. previously applied CL analysis to the identification of oxide particles embedded in metal powders such as iron and copper, using their synthesized samples.45) However, a few research papers have not been found with respect to the practical identification of inclusions in steels. A noticeable point in such an analysis is that the CL colors of some inclusions in steels may be changed by heat-treatment or the presence of impurities. For example, the CL color of alumina (Al2O3) inclusions was reported to change from red to blue or green when the Al2O3 inclusions contain impurities.38,39) However, such changes in the CL colors of inclusions in steel have not been well explained; therefore, the emission mechanism for the CL color of the Al2O3 inclusions should be investigated in more detail. The following section will introduce a fundamental study about CL color from Al2O3 oxide inclusions in an iron matrix, which was carried out by Imashuku.29)
The CL analysis of the inclusion sample was conducted using a custom scanning electron microscope - cathodoluminescence (SEM-CL) system,45) as shown in Figs. 4(a) and 4(b). A quartz viewport or a flange used to introduce an optical fiber into the SEM chamber was attached to a commercial SEM instrument. The CL images were captured through the quartz viewport using a digital single-lens reflex camera equipped with two close-up lenses. The emission spectra were acquired using an optical spectrometer by connecting an optical fiber to the optical spectrometer through another flange of the SEM chamber. The exposure times for obtaining the CL spectra and images were set from 1 to 30 s. In addition, SEM observation and EDX point analysis were also carried out for the same sample as in the CL analysis by placing it perpendicular to the electron beam as shown in Fig. 4(c). The acceleration voltage of the SEM-CL and SEM-EDX systems was set to 17 kV. The beam current bombarding the samples was 19 nA. An iron specimen containing Al2O3 inclusions was prepared using a mixture of an electrolytic iron powder and 1 mass% of an aluminum powder, and it was then heated in an alumina crucible at 1.56×103°C for 10 min in argon atmosphere. A CL image of the polished surface of the inclusion sample is shown in Fig. 5(a), with the corresponding SEM image shown in Fig. 5(b). The illuminated area in Fig. 5(a) corresponded to inclusions observed in the SEM image (Fig. 5(b)). All the inclusions were confirmed to be Al2O3 by EDX point analysis. Al2O3 inclusion particles located in the lower portion of Fig. 5(a), which are labeled as 3, 4, 5, and 6, yield blue luminescence, while blue and red-violet luminescence is observed from different inclusion particles, labeled as 1 and 2, in the upper portion. As observed in Fig. 5(c), the difference in luminesce color was clearly found in a CL image for the same sample area, which was captured with a digital camera whose build-in infrared filter was removed so that the infrared light having wavelengths of more than 680 nm could be freely transmitted onto the CCD detector of the camera. This modification resulted in an increased sensitivity of the camera in the infrared wavelength region. We also collected CL spectra of two Al2O3 inclusion particles, having mixed blue and red-violet luminescence (inclusion 1 in Fig. 5(a)) and blue luminescence (inclusion 5 in Fig. 5(a)), as shown in Fig. 6. Both of the CL spectra comprised several peaks at the same wavelengths. The broad peaks at around 335 and 480 nm were attributed to oxygen vacancies trapping one and two electrons, respectively.46,47,48,49,50) The sharp peak at 697 nm and the broad peak at around 750 nm originated from trivalent chromium ions (Cr3+) and trivalent titanium ions (Ti3+), the trace-level amounts of which randomly substitute the octahedrally-coordinated site of trivalent aluminum ions (Al3+) in the host crystalline lattice, respectively.46,47,48,49,50) The difference in the relative intensities of each peak would explain the variation in the CL color of the inclusions, even though these belong to the same kind of compound (Al2O3), It should be considered that the emission images in Fig. 5 are taken with the digital camera having a sensitivity range from 420 to 680 nm. For the inclusion 1 in Fig. 5(a), the corresponding spectrum of Fig. 6(a) indicates that the integrated intensity of the blue region (445–500 nm) is slightly greater than that of the red region (620–680 nm), which is narrower than the defined red region (620–740 nm) owing to the infrared filter in the digital single-lens camera. Thus, the Al2O3 of inclusion 1 was seen to be mixed blue and red-violet luminescence. In contrast, in Fig. 6(b), the integrated intensity of the blue region is much greater than that of the red region. Thus, the Al2O3 of inclusion 5 in Fig. 5(a) had blue luminescence. The digital camera quipped with the build-in infrared filter cannot detect the Cr3+ peak at 695 nm but can detect the short-wavelength tailing of this peak below 680 nm. Therefore, the difference in the CL colors can be attributed to the concentrations of Cr and/or Ti, and oxygen vacancies in the Al2O3 inclusions, which may be dependent on starting materials as well as the heating conditions when the sample alumina oxide is prepared. Different CL colors of Al2O3 inclusions in steel were reported in previous studies.38,39) However, the origin of this CL color difference has not been understood fully; therefore, studies on the emission mechanism should be continued.
Schematic illustrations of SEM-CL and SEM-EDX systems: (a) SEM-CL system for obtaining CL spectra, (b) SEM-CL system for capturing CL images, and (c) SEM-EDX system. This article is reprinted from Ref. 29 under the permission of Elsevier.
(a) CL and (b) SEM images of the polished surface of a test specimen of steel. (c) CL image of the same area when they were captured using a digital camera from which the installed infrared filter was removed. This article is reprinted from Ref. 29 under the permission of Elsevier. (Online version in color.)
The group of Imashuku also reported the CL phenomenon of nitride inclusions in steel, indicating that blue‐violet and blue emissions were observed in boron nitride and aluminum nitride, respectively, whereas titanium nitride had no luminescence.51)
For identification of inclusion particles in a metal matrix, we can utilize a similar phenomenon to the CL image and spectrum by irradiating an X-ray beam, abbreviated as XEOL.15) The XEOL image provides analytical information of the size, shape, and composition of nonmetallic inclusions in a steel material. Additionally, the X-ray excitation can be performed in an ambient atmosphere without any vacuum apparatuses; as a result, the XEOL technique becomes a promising approach for the on-site/in-line analysis of nonmetallic inclusions in the steel production. The following section introduces a study by Imashuku et al., regarding how Al2O3 inclusions in a steel sample emit the luminescent light under X-ray irradiation.52)
A test specimen of steel, which followed an actual de-oxidation process with aluminum in the industrial steel production, was prepared using a mixture of 96 mass% of an electrolytic Fe powder, 1 mass% of an Al powder, and 3 mass% of an MgO powder, and it was heated in an alumina crucible under an argon atmosphere at 1.55×103°C for 30 min. An XEOL image of the test steel sample was recorded using an experimental setup shown in Fig. 7. An X‐ray tube with a rhodium target was operated at 20 kV and 200 mA. A digital single‐lens reflex camera equipped with a zoom lens, whose sensitivity ranged from 420 to 680 nm, was used for acquisition of XEOL images. In addition, an elemental analysis of the steel sample was conducted using a scanning electron microscope equipped with a silicon drift energy‐dispersive X‐ray detector (SEM-EDX). Figure 8 shows an XEOL image of the steel sample which includes several emitting portions of blue, red, and green. It can be understood from the previous results of CL analysis27,50) that the portions of blue and red luminescence in Fig. 8 are assigned to the same kind of Al2O3 particles, whereas their luminescent colors are determined by the luminescent species of oxygen vacancies and impurity elements of Cr3+ or Ti3+ in the Al2O3 phase45,46,47,48,49,50) As described in the previous section, because they have each characteristic emission peak at different wavelengths, their concentrations in the Al2O3 particles and thus their relative intensities would affect the CL color. Namely, Al2O3 inclusions with lower content of Cr and Ti emit blue luminescence, whereas those with the higher content emit red luminescence. On the other hand, the particles emitting green luminescence in Fig. 8 comprise MgO·Al2O3 spinel inclusions because of the similarity in the luminescent spectrum. A major reason for the green color is that there appears an intense peak in the wavelength region of green in the spinel compound, which originates from divalent manganese ion (Mn2+) substituting the lattice site of divalent magnesium ions (Mg2+).53,54) An elemental analysis by SEM-EDX confirmed the assignment for the luminescent species in XEOL analysis. Namely, only Al was detected in the portions emitting blue and red luminescence (Al2O3 inclusion), whereas Al and Mg were detected in the portions emitting green luminescence (MgO·Al2O3 spinel inclusion). In addition, it was possible to distinguish between a mixed oxide of MgO and Al2O3 and the spinel compound from the luminescent color.
Schematic illustration of the setup for acquisition of X‐ray‐excited optical luminescence image. This article is reprinted from Ref. 52 under the permission of Wiley.
X‐ray‐excited optical luminescence image of a test specimen of steel. The exposure time was 30 s. This article is reprinted from Ref. 52 under the permission of Wiley. (Online version in color.)
XEOL spectrometry can provide an analytical information for inclusion particles in steel materials, such as the size, shape, and distribution, as similar to CL spectrometry. Its wider applications to the inclusion analysis are further expected because of several merits beyond SEM-EDX and CL methods.
The non-contacted and remote-sensing measurement can be conducted for solid samples, because LIBS is based on the probe and detection of visible/ultraviolet radiation under an ambient atmosphere.55,56,57,58,59,60) Therefore, LIBS can be applicable to determine the lateral and the depth (three dimensional) distribution of an element by moving the laser beam on a sample surface. In detection of inclusion particles in steel materials, several papers have reported on the distribution analysis of them by using LIBS.19,20,61,62,63,64,65,66)
Bouê-Bigne suggested a calibration method in LIBS analysis for obtaining the chemical composition as well as the size of non-metallic inclusion particles in a steel sample, in order to estimate the cleanness of the metallurgical structure.61) He compared the analytical result with that obtained with an SEM-EDX method, and pointed out the fast analysis of LIBS.
Kuss et al. published a research work that a scanning LIBS was employed in inclusion analysis of a steel material, to distinguish between different inclusion types based on their chemical compositions. It indicated that coincidences of high intensity peaks for a constituent element at the same sampling positions could be connected to a specific inclusion type.62)
Nakahata et al. reported on a quantitative map of alumina inclusions in a commercial stainless steel not only on the surface but in the depth direction by using scanning LIBS in a single-shot mode.20) The intensity ratio of Al I 396.152 nm to Cr I 396.368 nm was measured each for the single shot, while the irradiation positions were step-wise moved in the planar direction and then the same sampling area was repeatedly irradiated by subsequent laser shots in the depth direction, by using a precisely-driven X-Y-Z sample stage. The number of alumina particles was mapped from the intensity ratio of Al/Cr each for the irradiation points in both the lateral and in-depth directions.
Aimoto et al. applied an LIBS method to evaluation of surface defects in a Zn-coated steel plate and segregated portions of an iron slab by using a Paschen-Runge-type monochromator for the multi-elemental analysis.65) They suggested that this method could be applicable to inspect several types of defects in steel materials and that elemental maps of the constituent elements, Si, Al, Ca, and so on, in the defects could be obtained easily and rapidly.
Matsuda et al. evaluated a two-dimensional distribution of alumina inclusion particles in ferritic stainless steels from variations in the emission intensity of an Al I line, when a 1-kHz Q-switched Nd:YAG laser was scanned on a sample surface in LIBS.66) Their research will be introduced in more detail. Figure 9 shows a block diagram of the experimental system.20) The laser used in the experiment was a Q-switched Nd:YAG laser having an oscillation wavelength of 532 nm (the second harmonic of the fundamental frequency). The laser beam was focused onto a sample surface using a planoconvex lens, and an average of the irradiation energy was measured with a thermal sensor. The laser had an ability to be operated at repetition frequencies up to 1 kHz with a pulse width of less than 8 ns. The emission signal from the plasma was measured on a spectrometer, comprising a grating monochromator and a photomultiplier, whose effective spectral band-path was ca. 0.1 nm. The emission intensity was detected and averaged through a current/voltage converter, a pre-amplifier and a low-pass filter circuit, and it was finally recorded with an analog-to-digital converter on a personal computer. A cut-off frequency of the low-pass filter needed to be optimized for an average/smoothing of the emission intensity with an efficient precision, when scanning the laser beam on the sample surface appropriately. Three kinds of ferritic stainless steel were prepared as the test sample, which had different aluminum contents of soluble (Sol.) and insoluble (Insol.) aluminum in an acid decomposition pretreatment. The Insol. aluminum corresponded to alumina inclusions in the steel matrix. A resonance atomic line, Al I 396.152 nm, was selected as the analytical line in the LIBS measurement. A variation in of the Al I emission intensity was observed during a line scan of the laser beam for two different samples, as shown in Fig. 10. There appeared several sharp peaks in the emission signal together with a continuum emission which would be mainly originated from the emission intensity of Sol. aluminum in addition to a background emission. After subtracting the component of the continuum emission, emission peaks of the alumina inclusions could be extracted from the emission signal, from which the number and size of the particles could be estimated. A typical result of the two-dimensional map of alumina inclusions, based on the emission peaks of the Al I line after the background correction, is illustrated in Fig. 11. Here, the color tone of red ranked an increment in the emission intensity from pale red to dark red, while the color of white expressed irradiated portions having the Al I intensities below the background level, in which alumina particles could not be detected. The analysis time was approximately 20 min for the sample area of 4.5 mm2 (3.00 × 1.50 mm) to obtain the distribution of alumina inclusions. This LIBS method could provide the map of inclusions more rapidly, compared with a conventional method where each inclusion particle was counted with an optical microscope. Figure 11 indicates that the number of alumina particles is counted to be 35, 27, and 47 for three different samples, respectively. This result well corresponded to the amount of the Insol. aluminum as a result of the chemical analysis. Therefore, the two-dimensional map of the Al I emission intensity would roughly represent the actual positions of alumina inclusions in the sampling area, even though they had a variety of the size and smaller ones might not be detected.
Block diagram of the experimental apparatuses. This article is reprinted from Ref. 66 under the permission of Elsevier. (Online version in color.)
Variations in the Al I emission intensity along with scanning the laser beam, in stainless steel samples having different amounts of Al2O3 inclusions. This article is reprinted from Ref. 66 under the permission of Elsevier.
Two-dimensional maps of the Al I emission intensity for three stainless steel samples having different amounts of Al2O3 inclusions. This article is reprinted from Ref. 66 under the permission of Elsevier. (Online version in color.)
This paper demonstrates that inclusion particles in a steel material can be evaluated via the imaging measurement in CL, XOEL, and scanning LIBS, which provides the analytical information on their size and distribution as well as the chemical composition. The scanning LIBS is conducted through the simplest procedure with little pretreatment of the sample and without a vacuum atmosphere; therefore, it may be suitable for the actual application in the on-site/in-line steelmaking processes. On the other hand, the luminescence spectrometry can provide detailed information on the state of inclusions in an iron matrix, such as the kind of compounds as well as elements. In addition, the lateral resolution of the CL imaging is superior to that of scanning LIBS, due to the well-focused probe of the electron beam. It is an advantage of the XEOL imaging beyond the CL that it can be operated under an ambient atmosphere, which is available for the on-site analysis. As considering their analytical characteristics, we can obtain the emission image of inclusions appropriately.
