Open Access
Add to BibSonomy
Abstract
Three-dimensional multi-elemental mapping of composite wear-resistant coatings by laser-induced breakdown spectroscopy has been demonstrated for the first time, to the best of our knowledge. Individual clads of 1560 nickel alloy reinforced with tungsten carbide were synthesized by a co-axial laser cladding technique. Electron energy dispersive -ray spectroscopy revealed elemental maps for major elements (W, Ni, Co, Cr, Fe) but failed to measure silicon and carbon. Laser-induced breakdown spectroscopy was utilized for elemental mapping of carbon and all other elements of interest. It was demonstrated that three-dimensional elemental profiling for a few tens of micrometers requires substantial laser spot overlapping during the scanning procedure in order to achieve good accuracy of depth measurements. Elemental maps for nickel, iron, chromium, silicon, tungsten, and carbon were quantified for volume with 30 μm lateral and 4 μm depth resolution in the case of tungsten carbide particles in nickel alloy.
© 2017 Optical Society of America
1. INTRODUCTION
Fast growth of additive technologies provided new capabilities for producing complex composite materials with desired properties. Among numerous types of additive technologies, the laser cladding technique is one of the most powerful due to a unique combination of several advantages such as 3D structures production, capability to synthesize composite materials, and fast cladding process (few per second) [1,2]. However, the laser cladding technique is still under development due to complexity of numerous parameters’ influence on produced sample quality. Knowledge of the chemical composition and elemental mapping is of fundamental importance for such composite materials. Elemental maps can be used for predicting composite material properties or/and improving technology. For example, wear-resistant coating synthesis by the laser cladding technique is highly required in the mining industry and machinery. Typical wear-resistant coatings are based on a nickel alloy matrix reinforced with tungsten carbide particles. Coating hardness can be improved by increasing tungsten carbide particles concentration, but this will also trigger crack formation and tungsten carbide particles dissolution. In order to control coating quality, the express and reliable analytical techniques for elemental mapping are highly required. Typically, composite material is studied by an electron microscope equipped with an energy dispersive -ray (EDX) spectrometer [1]. The EDX technique provides good lateral resolution (1–3 µm, depending on the matrix) but suffers from low sensitivity for light elements (carbon, boron, etc.). It should be noted that carbon is a key element for tungsten carbide coating quality due to the possible tungsten carbide (WC) particles melting, which results in chemical interaction with binding matrix elements and formation of complex carbides. These carbides can dramatically change the coating microstructure and, consequently, decrease the coating hardness. Elemental mapping of carbon is of primary interest for tungsten carbide coatings due to the strong influence of WC grains dissolution and secondary carbides formation [3].
Recently, laser-induced breakdown spectroscopy (LIBS) was recognized as a powerful tool for elemental mapping of laser cladded coatings [4,5]. LIBS is a laser-based analytical technique capable of quantitatively analyzing and obtaining elemental maps for most of the elements, including light elements such as hydrogen, boron, carbon, etc. [6–8]. In recent years, many research groups successfully utilized LIBS for elemental mapping of different analytes in numerous applications. For example, Laserna and co-workers [9,10] performed a systematic study on 3D mapping of heavy metals in automobile catalytic converters. Though spatial resolution was low (0.2–1 mm), areas were mapped by LIBS for the layers cut from catalysts. Gornushkin et al. [11,12] developed a compact LIBS system for rapid identification of solids and particles with spatial resolution of 20 μm. Noll et al. [13] designed a micro-LIBS spectrometer capable of measuring 20 elements simultaneously with a spatial resolution of 5 μm. Later they developed a very fast mapping LIBS system based on 1 kHz laser with 20 μm resolution providing a image acquisition less than 11 min [14]. A large area () mapping by LIBS with 50 μm lateral resolution was demonstrated by Boue-Bigne [15], competing with the EDX technique. Cravetchi et al. [16] demonstrated multi-element microanalysis LIBS for chemical mapping of 10 μm precipitate particles. Menut et al. [17] performed a quantitative analysis for micro-LIBS mapping for both conductive and nonconductive materials. A cerium in uranium ceramic matrix was mapped with 3 μm spatial resolution. Authors achieved 2% single-shot reproducibility with estimated cerium limit of detection at 1% wt. Femtosecond LIBS was demonstrated to provide better spatial resolution for mapping applications due to the absence of a heat affected zone [18–20]. Feasibility of femtosecond LIBS with 450 nm spatial resolution was demonstrated by Zorba et al. [21]. 3D LIBS elemental mapping became increasingly popular for both nanosecond [22] and femtosecond LIBS [23,24] during the last decade.
In this study, we carried out 3D LIBS elemental mapping for composite tungsten carbide wear-resistant coatings produced by the laser cladding technique. To the best of our knowledge, 3D LIBS mapping has not been carried out so far for depth beyond a few micrometers in the case of metal samples. Depth profiling for a metal matrix is more challenging compared to ceramics [22] or polymer samples [23] due to the strong laser crater influence on laser ablation. Additionally, composite materials with a metal binding matrix have not been studied with 3D LIBS mapping. In the first part of the paper, we compare 2D elemental maps obtained by LIBS and EDX techniques. Then we optimize 3D LIBS profiling for depths beyond tens of micrometers. In the final part, we obtain 3D LIBS elemental maps of tungsten carbide particles in nickel alloy.
2. EXPERIMENTAL
A. Samples Description
Individual clads of tungsten carbide wear-resistant coatings were produced by a coaxial laser cladding technique [3,25]. A laser beam of a continuous-wave ytterbium-doped fiber laser (1064 nm, 2 kW, YLS-5 by IPG Photonics) was focused on a 2 mm spot in the coaxial cladding nozzle in argon atmosphere. 1560 nickel alloy and tungsten carbide powders (see Table 1) were mixed in the feeder (PF-2/2, GTV) and then flushed by argon gas to the coaxial laser cladding head (YC-50, Precitec), which was installed on an industrial six-axis robot arm (IRB-2600, ABB) manipulated with 100 μm precision. Individual clads were deposited on low alloy steel substrate (Fe37-3FN) at 1 kW laser power, flow 6 (1560 nickel alloy) (WC) g/s, and cladding head movement speed of 7 mm/s. Clads were cut, glued in epoxy resign, and then polished to a 1 μm finish in order to reveal their cross sections. Scanning electron microscopy (SEM) was carried out using a TESCA VEGA LMH microscope equipped with an EDX microanalysis system (AZtecEnergy, Oxford Instruments). The EDX detector was calibrated using standard samples of Ni (99.99%), Mn (99.99%) and Fe (99.999%).
Table 1. Chemical Composition (wt. %) of the Steel Substrate and Powders Used for Laser Cladding
B. Experimental Setup
Elemental mapping by LIBS was carried out in an experimental setup described previously in detail [26,27]. Briefly, a laser beam of a pulsed Nd:YAG laser (1064 nm, 10 ns, , 10 mJ/pulse, 5 Hz) was vertically focused onto the sample surface by a microscope objective (, focal length 8 mm). Estimated focal spot diameter at level was 10 μm, but a 110 μm diameter crater was produced by 10 mJ pulses ablation. To improve spatial resolution for LIBS mapping, we decreased the laser pulse energy to 0.4 mJ/pulse; thus, 50 μm diameter craters were obtained. The laser beam fluence at the sample surface equals , so plasma emission was bright enough for acquiring spectra with good signal-to-noise ratio. In the case of smaller pulse energies (0.2 mJ), the crater diameter was 30 μm, providing better spatial resolution. However, in the last case, we failed to reliably quantify the carbon line C I 193.09 in the spectrum with good signal-to-noise ratio. Individual clads were placed in a sample holder providing constant lens-to-sample distance. The sample holder was installed on a two-coordinate motorized translational stage (8MT173, Standa LTD), which can be moved with 0.5 μm precision. Laser plasma emission was collected by a quartz lens () and transferred to the spectrometer entrance slit with 1:1 magnification. The spectrometer (2400-grooves/mm, , Shamrock 303i, Andor Inc.) was equipped with an intensified CCD camera (iStar, Andor Inc.) for time-resolved measurements. The motorized stage provided synchronization electric pulses that triggered the laser and the gated camera acquisition. The surface scanning was performed by recording the plasma emission spectrum for each sampling spot. Overall control and data handling were carried out by a custom software developed in the LabVIEW environment. All the LIBS measurements were carried out in air at ambient pressure and temperature.
3. RESULTS AND DISCUSSION
A. EDX Elemental Maps
Elemental mapping was carried out by electron EDX spectroscopy before any LIBS measurements, and results are presented in Fig. 1. A semispherical clad profile was formed with good quality weld junction to the substrate. Tungsten carbide grains distribution was rather uniform. Some of the tungsten carbide particles had irregular shapes indicating partial tungsten carbide particle dissolution during synthesis. Heavy elements for the matrix alloy (Ni, Cr, Fe) and tungsten carbide particles (W, Co) were clearly observed on the maps. However, carbon was not quantified by the EDX spectrometer due to the low intensity peak spectrally interfering with major elements. A silicon map was obtained by the EDX spectrometer, but the brightest areas corresponded to tungsten carbide (WC) particles. A closer examination of a single tungsten carbide area and nearby area revealed that WC particles are free from silicon (see spots S3 and S6 in Fig. 2). Spectral interference of silicon and tungsten peaks is the main reason for the false silicon elemental map.
Fig. 1. SEM image and EDX elemental maps for substrate (Fe), 1560 alloy (Ni, Cr, Si, and Fe) reinforced with tungsten carbide particles (W, C, Co).
Fig. 2. Detailed SEM image (a) of WC/Co particle in 1560 alloy matrix. EDX spectra for 1560 alloy (spot S3, b), and WC/Co particle (spot S6, c).
B. LIBS Spectra
The optimal choice of LIBS spectral lines depends on a number of factors, including detector sensitivity, line intensity, and spectral interference. In this study, we were focused on light elements mapping by LIBS, including carbon, silicon, and boron. The choice of a spectral line for carbon analysis in iron-containing materials is a challenge due to a strong spectral interference. For example, carbon lines in UV (C I 247.86) and near IR (C I 833.51) spectral regions strongly interfere with iron and tungsten lines [7,28]. Carbon line C I 193.09 is an optimal choice, since this spectral region is moderately absorbed by air, while our detector was sensitive down to 189 nm.
An example of the laser plasma spectrum in 190–210 nm range is presented in Fig. 3. The atomic line for silicon (Si I 190.13) had low intensity but could be reliably quantified in the spectrum. Unfortunately, we cannot choose spectra lines for boron. Some sensitive boron lines in the deep UV region (B I 181.80 and B I 182.62 nm) cannot be measured, since this spectral region is not covered by our camera sensor. Line B I 208.95 spectrally interferes with major element lines (nickel, chromium, and iron) and cannot be extracted even with multiple peak fitting due to the very low intensity of the boron line. All other atomic/ionic lines for nickel (Ni I 197.69) chromium (Cr I 199.99), iron (Fe I 193.45), tungsten (W II 207.91), and cobalt (Co II 194.13) were chosen to be free of spectral interference and to have sufficiently high intensity in the spectrum. The tungsten carbide spectrum contained lines for nickel, chromium, and iron due to possible partial ablation of matrix material by the outer part of the laser beam. Spectral line constants presented in Table 2 were extracted from the NIST database.
Fig. 3. Laser plasma spectra for matrix 1560 Ni-Cr-Fe-B-Si alloy (black color) and tungsten carbide particle (red color). Spectra were acquired with 1 μs gate and 0.1 μs delay. Atomic/ionic lines selected for LIBS mapping are marked underlined bold.
It is well known that gating plasma spectra improves sensitivity of LIBS analysis by rejection of the continuum background in the first moments of plasma expansion. It is desirable to skip high electron density during the early stages of plasma emission, since Stark broadening results in widening of atom/ion lines, and hence a spectral interference is possible. With electronically gating plasma spectra acquisition, one can obtain spectra with high signal-to-noise ratio and with spectrally resolved lines. The optimum delay is known to depend on the chosen atomic/ionic line characteristic (energy of upper level, Stark broadening coefficient) thus we considered the optimal choice for major elements (Ni, Cr, Fe, W, Co, Si, and C) with special attention to carbon. The atomic/ionic signals were estimated as corresponding line integrals with background correction. Line broadening was estimated as full width at half-maximum (FWHM). Evolution of major element signals and line widths for plasmas induced on the matrix alloy and at the WC particle are presented in Fig. 4. Plasma emission duration was limited to few microseconds, so in this study we chose a delay time of 0.1 μs and integration time of 1 μs in order to acquire spectra with resolved lines and sufficiently high intensity.
Fig. 4. Temporal dependence of atomic intensities (open circles) and full widths at half-maximum (open triangles) for nickel Ni I 197.69 (a) and carbon C I 193.09 (b) lines.
C. 2D Elemental Mapping by LIBS
Typically, LIBS mapping is performed with spot-by-spot sampling without any spots overlapping in order to suppress possible influence of craters formed by successive pulses. In our case, five laser pulses created a crater with 50 μm diameter (Fig. 5). For lower energy pulses, the crater diameter was 20 μm but plasma emission was too weak, and we could not quantify plasma spectra in the 190–210 nm range with meaningful signal-to-noise ratio. The typical diameter of a tungsten carbide particle was 70–100 μm so we acquired an elemental map with a 40 μm step between locations, which resulted in 10% area overlapping of sampling spots.
Fig. 5. Laser crater profiles with a step of 100 μm (a) and 40 μm (c) and corresponding crater track cross sections (b). The cross section for craters with 100 μm steps is shifted (b) by 4 μm for better view.
LIBS mapping was performed by point-to-point scanning with every spot sampled by five laser pulses to improve repeatability. Comparing contour maps in Fig. 6, it is clearly seen that locations of tungsten carbide particles can be easily identified with carbon, tungsten, and cobalt emission maximums. Nickel and chromium are correlated with each other and inversely correlated with carbon and tungsten. Summarizing Fig. 6, LIBS mapping provided results for light elements mapping, such as carbon and silicon, which were not possible to quantify by the EDX spectrometer.
Fig. 6. Single laser clad elemental map for iron, nickel, chromium, silicon, cobalt, tungsten and carbon, as revealed by LIBS.
D. Optimizing 3D LIBS Elemental Mapping
3D LIBS elemental mapping for non-metal samples was carried out in several studies with layer-by-layer ablation [22,23,30]. However, tens of micrometer depth profiling for metal alloy samples is a challenge due to laser crater formation, which influences laser ablation and plasma emission, so sampling strategy should be optimized. For example, multiple scanning through the same sampling spots will result in preferential ablation of the same spot; thus, plasma will be influenced by the crater walls’ geometry. It is well known that deep and narrow craters impact the effective laser beam irradiance on a crater surface; thus, plasma temperature and electron density are also affected, resulting in depth dependence of atomic/ionic line intensities [31]. Additionally, crater wall material will be ablated or evaporated by expanding plasma, thus, depth resolution becomes poorer for deeper craters [32]. In order to study crater formation influence on plasma emission under our conditions, a series of single-shot spectra were acquired for the same sampling spot. Nickel atomic line Ni I 361.95 intensity was measured for every laser shot, and results are presented in Fig. 7. We also measured plasma temperature in parallel with Ni I 361.95 line intensity. 1560 nickel alloy contains 5% wt. of iron, so plasma temperature can be measured according to the Boltzmann plot method with non-resonant iron lines (listed in Table 2).
Fig. 7. Atomic nickel line (Ni I 361.95) intensity (black) and plasma temperature (violet) dependence (a) during deep crater formation. 3D crater profiles and corresponding cross sections are shown in (b) after ablation with 1, 5, 10, and 30 pulses.
Laser crater profiles formed after different numbers of laser pulses were quantified with a white-light interferometry microscope (NewView 6200, Zygo). Large rims were formed for 10 and 30 pulse craters with estimated rims volume equal to of the crater volume. Rims were formed during melt displacement by expanding laser plasma; thus, laser mixing during successive pulses ablation will influence depth elemental profiling. As the pulse number increased, both Ni I 361.95 atomic lines intensity and plasma temperature were continuously decreasing. A good correlation between Ni I 361.95 line intensity and plasma temperature was observed, except for the plasma generated by first few shots. The last feature was attributed to the influence of surface contaminations (oxidized material, polishing particles, etc.), which can strongly affect ablation and plasma properties. The Ni I 361.95 line intensity fluctuated near the same average value during the first 10 laser pulses but then slowly decreased with the pulse number increasing. For example, the initial nickel signal drops 4-fold for the 30 μm deep crater produced with 30 laser pulses.
Generally, 2D LIBS mapping is performed by sampling laser spots as close as possible to each other but without spots overlapping in order to prevent influence of the previously formed crater. However, if laser spots were not overlapped during LIBS mapping, then the crater cavity will influence the plasma emission, providing poor results for 3D elemental LIBS maps. In order to estimate such influence, we produced a series of rectangular areas by successive scanning with one-shot sampling and 0% laser spots overlapping. We specifically chose areas of 1560 Ni-alloy, which were free of tungsten carbide particles in order to ablate homogeneous material. The results of surface profiling for areas after 1, 5, 10, and 30 scans by one-shot sampling are shown in Fig. 8. The cavity bottom surface had the specific pattern with well-resolved locations of laser craters. Moreover, huge re-solidified rims formed during ablation (for example, see 5th and 10th scans in Fig. 8), resulting in mixing of layers from different depths that can strongly affect the 3D LIBS mapping results. The depth profiling precision was defined as an arithmetical mean roughness value, which was defined according to EN ISO 4287. The large curvature of the cavity area bottom decreased precision of depth profiling to 7 μm. The ablation rate defined as cavity area depth per single laser pulse was estimated as 0.8 μm/scan.
Fig. 8. 1560 Ni-alloy areas ablated with 1, 5, 10, and 30 scans (a) of single-shot sampling and corresponding cross section (b) for sampling scheme with 0% spot overlapping (50 μm step) (c). Crater depth as a function of ablating pulses number was fitted with linear function (d).
Fig. 9. Nickel atomic line Ni I 361.95 average intensity and standard deviation (a) for corresponding intensity maps (b) in the case of successive sampling with single-shot pulses and 0% laser spots overlapping.
Intensity of the Ni I 369.95 line were quantified for the 1st, 5th, 10th, and 30th scan in order to study the influence of depth profiling on elemental mapping (Fig. 9). The Ni I 369.95 line intensity maps and corresponding signal reproducibility (standard deviation for all points in the map) revealed that successive sampling decreased nickel line intensity twice and worsened reproducibility 2.5 times. These results are in agreement with a previous study [33] in which increased fluctuation of atomic line intensity was observed during 10 consecutive single-shot ablating scans.
In order to improve depth resolution and suppress influence of “surface cratering,” we changed the sampling strategy and produced several rectangular maps with a 30 μm step (step equals to crater radius). Greater spots overlapping (38% per area) significantly improved the ablated area roughness (arithmetical mean roughness value, EN ISO 4287) (Fig. 10). The number of ablating pulses squared, but ablated volume increased only two times with an estimated ablating rate of 1.7 μm/scan. Lower roughness of the ablated area bottom () resulted in two-fold improved absolute accuracy (four-fold for relative) of depth estimation. Ni I 361.95 line intensity maps (Fig. 11) comparison revealed a moderate improvement of signal reproducibility for depths below 40 μm compared to the first scan reproducibility. Interestingly, when the ablated depth reached 60 μm, the Ni I 361.95 line intensity map decreased near the edges of the ablated area compared to the central part. This can be attributed to a decrease of laser beam influence on inner walls, as well as wall influence on plasma expansion. However, this impact was located near sampling area edges and can be neglected for inner areas during 3D mapping.
Fig. 10. 1560-Ni alloy ablated areas () with 1, 5, 10, and 30 scans (a) and corresponding cross section (b) for sampling scheme with 38% spot overlapping (30 μm step) (c). Ablation rate nonlinearly depended on pulse number; thus, depth as a function of ablating pulses was fitted by a quadratic function (d).
Fig. 11. Nickel atomic line Ni I 361.95 standard deviation (a) for corresponding maps (b) in the case of successive sampling with single-shot pulses and 38% laser spots overlapping.
Finally, 3D elemental maps were acquired for a small area of tungsten carbide individual clads with 38% laser spot overlapping (30 μm step) according to previously discussed 3D mapping optimization. Layer-by-layer contour maps for nickel, tungsten, and carbon in a composite coating sample are presented in Fig. 12. Each layer represents a area. Depletion of atomic/ionic line intensities during depth profiling was corrected according to Fig. 11(a) in order to obtain better elemental maps. According to the presented elemental contour maps, a 3D structure of tungsten carbide particles in a nickel matrix can be observed. For example, the single tungsten carbide particle appearance according to 3D LIBS mapping is marked with a violet circle for better view.
Fig. 12. Layer-by-layer contour elemental maps for nickel (green), tungsten (red), and carbon (magenta). Violet circle indicates an example of tungsten carbide particle (100 μm in diameter) appearance during layer-by-layer LIBS mapping.
4. CONCLUSIONS
3D multi-elemental mapping of composite wear-resistant coatings by LIBS has been demonstrated for the first time. Individual clads of 1560 nickel alloy reinforced with tungsten carbide particles were synthesized by a co-axial laser cladding technique. Electron EDX spectroscopy revealed elemental maps for major elements (W, Ni, Co, Cr, Fe) but failed to measure carbon and silicon. LIBS was utilized for elemental mapping of carbon and silicon, as well as for all other elements of interest. It was demonstrated that 3D mapping of metal samples for depths beyond a few micrometers requires a change of sampling strategy. For example, substantial laser spot overlap during mapping (38% by area) decreased roughness of the ablated area bottom and hence improved accuracy of depth profiling. Elemental maps for nickel, tungsten, and carbon were quantified for volume with 30 μm lateral and 4 μm depth resolution in the case of tungsten carbide particles in nickel alloy.
Funding
Russian Science Foundation (RSF) (16-19-10656).
REFERENCES
1. A. J. Pinkerton, “[INVITED] Lasers in additive manufacturing,” Opt. Laser Technol. 78, 25–32 (2016). [CrossRef]
2. E. M. Birger, G. V. Moskvitin, A. N. Polyakov, and V. E. Arkhipov, “Industrial laser cladding: current state and future,” Weld. Int. 25, 234–243 (2011). [CrossRef]
3. A. G. Grigoryants, A. Y. Stavertiy, K. O. Bazaleeva, T. Y. Yudina, N. A. Smirnova, R. S. Tretyakov, and A. I. Misyurov, “Laser surfacing of nickel-based composite war-resisting coatings reinforced with tungsten carbide,” Weld. Int. 31, 52–57 (2016). [CrossRef]
4. J. A. Varela, J. M. M. Amado, M. J. J. Tobar, M. P. P. Mateo, A. Yañez, and G. Nicolas, “Characterization of hard coatings produced by laser cladding using laser-induced breakdown spectroscopy technique,” Appl. Surf. Sci. 336, 396–400 (2015). [CrossRef]
5. V. N. Lednev, P. A. Sdvizhenskii, M. N. Filippov, M. Y. Grishin, V. A. Filichkina, A. Y. Stavertiy, R. S. Tretyakov, A. F. Bunkin, and S. M. Pershin, “Elemental profiling of laser cladded multilayer coatings by laser induced breakdown spectroscopy and energy dispersive X-ray spectroscopy,” Appl. Surf. Sci. 416, 302–307 (2017). [CrossRef]
6. D. W. Hahn and N. Omenetto, “Laser-induced breakdown spectroscopy (LIBS), Part II: review of instrumental and methodological approaches to material analysis and applications to different fields,” Appl. Spectrosc. 66, 347–419 (2012). [CrossRef]
7. M. A. Khater, “Laser-induced breakdown spectroscopy for light elements detection in steel: state of the art,” Spectrochim. Acta B 81, 1–10 (2013). [CrossRef]
8. K. Sugiyama, T. Fujii, T. Matsumura, Y. Shiogama, M. Yamaguchi, and K. Nemoto, “Detection of chlorine with concentration of 0.18 kg/m3 in concrete by laser-induced breakdown spectroscopy,” Appl. Opt. 49, C181–C190 (2010). [CrossRef]
9. P. Lucena and J. J. Laserna, “Three-dimensional distribution analysis of platinum, palladium and rhodium in auto catalytic converters using imaging-mode laser-induced breakdown spectrometry,” Spectrochim. Acta B 56, 177–185 (2001). [CrossRef]
10. P. Lucena, J. M. Vadillo, and J. J. Laserna, “Spatial distribution of catalytically active elements and deactivants in diesel-engine automobile converters by laser-induced plasma spectrometry,” J. Anal. At. Spectrom. 17, 548–551 (2002). [CrossRef]
11. I. B. Gornushkin, B. W. Smith, H. Nasajpour, and J. D. Winefordner, “Identification of solid materials by correlation analysis using a microscopic laser-induced plasma spectrometer,” Anal. Chem. 71, 5157–5164 (1999). [CrossRef]
12. I. B. Gornushkin, A. Ruiz-Medina, J. M. Anzano, B. W. Smith, and J. D. Winefordner, “Identification of particulate materials by correlation analysis using a microscopic laser induced breakdown spectrometer,” J. Anal. At. Spectrom. 15, 581–586 (2000). [CrossRef]
13. R. Noll, H. Bette, A. Brysch, M. Kraushaar, I. Mönch, L. Peter, and V. Sturm, “Laser-induced breakdown spectrometry—applications for production control and quality assurance in the steel industry,” Spectrochim. Acta B 56, 637–649 (2001). [CrossRef]
14. H. Bette and R. Noll, “High speed laser-induced breakdown spectrometry for scanning microanalysis,” J. Phys. D 37, 1281–1288 (2004). [CrossRef]
15. F. Boué-Bigne, “Laser-induced breakdown spectroscopy applications in the steel industry: rapid analysis of segregation and decarburization,” Spectrochim. Acta B 63, 1122–1129 (2008). [CrossRef]
16. I. V. Cravetchi, M. Taschuk, G. W. Rieger, Y. Y. Tsui, and R. Fedosejevs, “Spectrochemical microanalysis of aluminum alloys by laser-induced breakdown spectroscopy: identification of precipitates,” Appl. Opt. 42, 6138–6147 (2003). [CrossRef]
17. D. Menut, P. Fichet, J.-L. Lacour, A. Rivoallan, and P. Mauchien, “Micro-laser-induced breakdown spectroscopy technique: a powerful method for performing quantitative surface mapping on conductive and nonconductive samples,” Appl. Opt. 42, 6063–6071 (2003). [CrossRef]
18. S. P. Banerjee, Z. Chen, and R. Fedosejevs, “High resolution scanning microanalysis on material surfaces using UV femtosecond laser induced breakdown spectroscopy,” Opt. Lasers Eng. 68, 1–6 (2015). [CrossRef]
19. A. Assion, M. Wollenhaupt, L. Haag, F. Mayorov, C. Sarpe-Tudoran, M. Winter, U. Kutschera, and T. Baumert, “Femtosecond laser-induced-breakdown spectrometry for Ca2+ analysis of biological samples with high spatial resolution,” Appl. Phys. B 77, 391–397 (2003). [CrossRef]
20. Y. Lu, V. Zorba, X. Mao, R. Zheng, and R. E. Russo, “UV fs-ns double-pulse laser induced breakdown spectroscopy for high spatial resolution chemical analysis,” J. Anal. At. Spectrom. 28, 743–748 (2013). [CrossRef]
21. V. Zorba, X. Mao, and R. E. Russo, “Ultrafast laser induced breakdown spectroscopy for high spatial resolution chemical analysis,” Spectrochim. Acta B 66, 189–192 (2011). [CrossRef]
22. G. Nicolas, M. P. Mateo, and V. Pinon, “3D chemical maps of non-flat surfaces by laser-induced breakdown spectroscopy,” J. Anal. At. Spectrom. 22, 1244–1249 (2007). [CrossRef]
23. H. Hou, L. Cheng, T. Richardson, G. Chen, M. Doeff, R. Zheng, R. Russo, and V. Zorba, “Three-dimensional elemental imaging of Li-ion solid-state electrolytes using fs-laser induced breakdown spectroscopy (LIBS),” J. Anal. At. Spectrom. 30, 2295–2302 (2015). [CrossRef]
24. J. R. Chirinos, D. D. Oropeza, J. J. Gonzalez, H. Hou, M. Morey, V. Zorba, and R. E. Russo, “Simultaneous 3-dimensional elemental imaging with LIBS and LA-ICP-MS,” J. Anal. At. Spectrom. 29, 1292–1298 (2014). [CrossRef]
25. A. G. Grigoryants, R. S. Tretyakov, I. N. Shiganov, and A. Y. Stavertiy, “Optimization of the shape of nozzles for coaxial laser cladding,” Weld. Int. 29, 639–642 (2015). [CrossRef]
26. V. Lednev, S. M. Pershin, and A. F. Bunkin, “Laser beam profile influence on LIBS analytical capabilities: single vs. multimode beam,” J. Anal. At. Spectrom. 25, 1745–1757 (2010). [CrossRef]
27. V. N. Lednev, S. M. Pershin, A. F. Bunkin, A. A. Samokhvalov, V. P. Veiko, S. I. Kudryashov, and A. A. Ionin, “Double pulse laser induced breakdown spectroscopy with Gaussian and multimode beams,” Spectrochim. Acta B 124, 47–55 (2016). [CrossRef]
28. T. A. Labutin, S. M. Zaytsev, A. M. Popov, and N. B. Zorov, “Carbon determination in carbon-manganese steels under atmospheric conditions by laser-induced breakdown spectroscopy,” Opt. Express 22, 22382–22387 (2014). [CrossRef]
29. A. Kramida, Y. Ralchenko, J. Reader, and NIST ASD Team, “NIST Atomic Spectra Database (version 5.5.1), [Online],” 2017, available: https://physics.nist.gov/asd.
30. V. Piñon, M. P. Mateo, and G. Nicolas, “Laser-induced breakdown spectroscopy for chemical mapping of materials,” Appl. Spectrosc. Rev. 48, 357–383 (2013). [CrossRef]
31. S. S. Mao, X. Zeng, X. Mao, and R. E. Russo, “Laser-induced breakdown spectroscopy: flat surface vs. cavity structures,” J. Anal. At. Spectrom. 19, 495–498 (2004). [CrossRef]
32. S. M. Pershin, R. H. Tabares, and R. A. Nunes, “Correction of the spectral lines of a laser plasma for measurement of the depth profile of layered materials in an ablation crater,” Quantum Electron. 29, 862–864 (1999). [CrossRef]
33. I. Lopez-Quintas, V. Pinon, M. P. Mateo, and G. Nicolas, “Effect of surface topography in the generation of chemical maps by laser-induced plasma spectroscopy,” Appl. Surf. Sci. 258, 9432–9436 (2012). [CrossRef]
