Stand-off analysis of moving targets using laser-induced breakdown spectroscopy

Abstract

This work investigates the capabilities of stand-off LIBS for the analysis of moving samples. The effects of sample speed, repetition rate and incidence angle of the laser beam on the sample have been studied. Stainless steel samples tilted from 0° to 45° at speeds up to 24 cm min⁻¹ have been analyzed at a distance of 10 m from the instrument. The spectral intensity of the spectra observed is related to the amount of sample surface overlapped by successive laser shots, and this can be controlled both by changing the laser repetition rate and the sample speed. As a proof of concept, a semi-quantitative analysis of stainless steel samples has been carried out under simulated field conditions—10 m from the laser source, incidence-collection angle of 40° and sample speed of 24 cm min⁻¹. Samples corresponding to five different stainless steel grades were sorted according to their content of Ni, Mo and Ti.

---

Introduction

Laser-induced breakdown spectroscopy (LIBS) currently represents one of the most versatile techniques for rapid analysis of materials. Spectral study of the light emitted by laser-induced plasmas provides valuable information for the characterization of the sample. The interaction between the laser beam and matter has been widely studied,^1,2 giving evidence that the efficiency of laser ablation widely depends on laser pulse and beam focal conditions,^3,4 incidence angle of the beam,⁵ laser-to-target distance,^6,7 sample physical properties,⁸ temperature,^9,10 morphology and roughness,^11,12 surrounding environment,^13,14etc. As expected, important parameters in a conventional laboratory-scale LIBS system, became critical when the system was used in the field. In many industrial environments, real time analysis of samples located on a moving production line should be performed. Several authors have studied the effect of sample movement on the LIBS signal. Castle et al.¹⁵ observed the variation of the emission signal with the sample velocity using a close-contact LIBS instrument. They found a decrease in the emission signal at speeds up to 100 μm s⁻¹, but for higher velocities the signal kept steadily growing. They concluded that this fact was related to the material redeposited around the crater which could be sampled at certain translational velocities.

Other authors have performed a sorting of irregular shape moving samples using LIBS.^16,17 However, besides the movement of the target, high sample temperatures add to the difficulties when working in production environments. Elemental analysis of steel-making products at a distance using LIBS has been previously reported.^18,19 This technique offers the solution for a wide range of industrial applications due to its rapid and in situ response at the industrial plant.^18–21

The purpose of this work is to investigate how the movement of the sample affects the emission signal at different angles of laser incidence and at relatively long distances from the target. The target distance was set at 10 m and the incident angle of the beam has been studied up to 60°. Samples were kept in movement at speeds between 6 cm min⁻¹ and 24 cm min⁻¹. The laser repetition rate and the speed of the sample are connected by the overlapped surface sampled in every situation. An assessment of this connection has been carried out in this work. As an example of the capabilities of this approach, several common steel samples in motion have been classified according to their compositions.

Experimental

Fig. 1 sketches the configuration of the experimental set-up. A Q-switched Nd:YAG laser, operating at 1064 nm with a pulse width and energy per pulse of approximately 10 ns and up to 1650 mJ, respectively, was used. The laser beam was led to the beam expander entrance using two folding mirrors and then focused at 10 m using a 3.2x beam expander integrated by a couple of 1064 nm antireflection-coated best-form lenses with effective focal lengths of −62.1 mm and 198.84 mm. The two elements are governed by a motorized linear stage and, by changing their separation, the laser beam can be tightly focused. This enables an even larger depth of focus, which is less sensitive to sample vibrations than that commonly found in stand-off LIBS setups and, at the same time, which produces a well defined crater that is of relevance for measurements of crater overlapping carried out in this work. The crater created under these conditions was approximately 4.3 mm in diameter, yielding a peak fluence of 11.5 J cm⁻². Given the flat beam profile and the dimensions of the focused spot, the craters obtained present a flat bottom. The plasma light returned was directly focused to an f/4 Newtonian telescope with a primary mirror 200 mm in diameter. The light is then focused to the entrance slit of a spectrograph fitted with a 1800 grooves mm⁻¹ grating. The dispersed light is finally detected by an intensified CCD detector. The spectral window of the spectrograph covered approximately 24 nm.

Fig. 1 Instrument setup. 1, Laser source; 2, folding mirror; 3, folding mirror; 4, focusing beam expander; 5, Newtonian telescope; 6, spectrograph; 7, intensified CCD; 8, computer.

Samples

The samples used in this work (stainless steel grades AISI 316, AISI 316Ti, AISI 430Nb, AISI 304, and AISI 430) were provided by Acerinox S.A. (Los Barrios, Cádiz, Spain). The moving samples were placed at 10 m from the instrument.

Results and discussion

A polished AISI 430 steel sample was fired at repetition rates of 0.625 Hz, 1.25 Hz, 2.5 Hz, 5 Hz and 10 Hz while it was moved perpendicularly to the laser beam at speeds of 6 cm min⁻¹, 12 cm min⁻¹ and 24 cm min⁻¹. The raw signal of Cr I at 540.979 nm was averaged over 20 pulses and is summarized in Table 1, together with the speed of the sample for the different repetition rates. As is shown, for a given repetition rate, the intensity decreases for increasing target speed, in good agreement with the findings of Castle et al.¹⁵ In addition, the overall intensity also decreases at lower repetition rates. When the sample moves, the laser hits a variable extent of fresh surface in addition to a portion of the previous crater, and thus a mixture of oxide and redeposited material is sampled. As the amount of fresh surface sampled increases, the intensity of emission decreases. This behaviour has been observed before on polished samples¹² and was attributed to the high reflectivity of the sample used in this experiment presenting a mirror-like surface finish. After the first pulse, the surface becomes rougher and darker. This darkening is owed, among other things, to redeposition and oxidation of the surface induced by the laser pulse. Such surface modifications improve the laser-to-target coupling, favouring a higher amount of ablated mass which, in turn, translates into an increase in signal intensity under dynamic conditions. To further prove this effect, the overlapping percentage was estimated for all combinations of speed and laser repetition rate used and their corresponding values are also summarized in Table 1. The overlapping percentage is defined as the percentage of crater area overlapped by the following laser pulses in a shot series when the target advances. Interestingly, those combinations of speed and laser repetition rate, with the same overlapping percentage, exhibit nearly the same intensity, which is indicative of the dependence of the signal on the proportion of modified surface hit by the laser irrespective of the sample speed. As a consequence, it is reasonably safe to assume that owing both to the low interaction time of the laser pulse—as compared with the sample speed—and to the wide field of view of the collection optics—as compared with plasma size—sample vibrations derived from its movement should not significantly affect stand-off LIBS measurements on a moving sample as long as the ratio of fresh to modified surface remains unaltered.

Table 1 Approximated overlapping percentage and intensity in function of the repetition rate and the speed

Sample speed/cm min^−1	Repetition rate/Hz	Overlapping (%)	Intensity (a.u.)
6	10	97	29 517
	5	94	27 987
	2.5	88	23 989
	1.25	76	18 831
	0.625	53	14 986
12	10	64	27 598
	5	88	24 291
	2.5	76	19 865
	1.25	53	15 942
	0.625	14	12 740
24	10	88	24 190
	5	76	19 131
	2.5	53	15 235
	1.25	14	12 356
	0.625	0	9582

---

Effect of incidence angle

The relative angle between the laser beam and the sample surface is a key parameter in real industrial analysis. Under laboratory conditions, this parameter can be easily controlled, but in the field the mobility of the sample could require analysis of the sample at incidence angles other than normal. In other cases, whilst the laser beam hits the sample at right angles, the target morphology could yield a different angle. The combined effect of both sample motion and focusing–collection angles has been evaluated.

An AISI 430 stainless steel sample was oriented at angles up to 60° to the laser beam. Simultaneously to firing the laser at 10 Hz, the sample moved along its surface plane at 6 cm min⁻¹ and 24 cm min⁻¹. Fig. 2 illustrates the intensity variation of the Cr I 540.979 nm emission line under the conditions of angle and target speed mentioned. As is shown, for the two speeds tested, the intensity drops with the angle although the decrease is not as significant as one might expect. There are three factors mainly responsible for this effect: overlapping, irradiance and reflectivity. Given the mirror-like finish of the sample, reflectivity should not increase substantially. On the other side, although the irradiance decreases with angle by a half—from 1.1 × 10¹⁰ W cm⁻² to 5.3 × 10⁹ W cm⁻²—it remains at a level well over the breakdown threshold. In addition, overlapping increases, as shown in the figure inset, mitigating the effect of the former two factors. As a result the signal decreases in magnitude with the angle less than expected but the data dispersion increases, probably due to the lower ablation rate at large angles. At angles below 40°, the results are promising and prove their analytical quality in terms of stability to changes of the angle of incidence.

Fig. 2 Variation of the Cr I 540.979 nm emission intensity for an AISI 430 sample with the angle of incidence of the laser beam (pulsing at 10 Hz). Sample speeds were 6 cm min⁻¹ (■) and 24 cm min⁻¹ (●). Distance of analysis was 10 m. The inset illustrates the variation of the overlapping percentage as a function of angle for both sample speeds.

Semi-quantitative analysis of moving samples

Under field conditions, analytical techniques demanding a tight control of the experimental parameters are of little, if any, use. To test this perception, a sorting of common steel grades samples was carried out. In order to mimic the field conditions in a steel factory, the experimental parameters were set as follows: the samples were kept moving at 24 cm min⁻¹ and 40° to the laser operating at 10 Hz. The distance from the system to the target was 10 m.

Fig. 3 illustrates the pattern recognition space with three variables: Ni I 547.691 nm, Ti I 499.107 nm and Mo I 550.649 nm signals. The results plotted were obtained for 10 measurements each of 20 laser shots averaged per sample. The composition of the steel samples measured by spark-OES is detailed in the graph legend. As observed, the identification of the different steel grades is feasible under the conditions described.

Fig. 3 Three dimensional space illustrating clustering of Ni, Mo and Ti raw signals to produce the discriminating pattern indicated by the shadowed areas. The sample speed was 24 cm min⁻¹, the laser angle was 40° and the distance of analysis was 10 m. The composition of the samples measured by spark-OES is detailed in the graph key.

Conclusions

Stand-off LIBS has been shown to be a useful analytical tool for difficult-to-access locations and hostile environments. The technique has demonstrated an excellent versatility when analyzing animated samples (up to 24 cm min⁻¹) in combination with different orientation angles and pulse repetition rates. It has been proved that the intensity of the spectrum is related to the amount of sample surface overlapped by successive laser shots, and this can be controlled by changing the repetition rate for a given sample speed. The technique has been demonstrated to be suitable for semi-quantitative measurements with a certain potential for quantitative analysis. The sorting out of steel grades has been easily achieved under field conditions. The maximum angle at which accurate sorting of samples could be achieved was 40°.

Acknowledgements

This work has been funded by the project CTQ2004-02966 of the Spanish Ministerio de Ciencia y Tecnología. The research group wants to specially thank Acerinox S. A. for the samples. One of the authors (C.L.M.) thanks the Spanish Ministerio de Ciencia y Tecnología for providing a research fellowship.

References

1. L. J. Radziemski, R. W. Solarz and J. A. Paisner in Laser Spectroscopy and its Applications, Marcel Dekker, New York, 1987.
2. S. A. Darke and J. F. Tyson, J. Anal. At. Spectrom., 1993, 8, 145–209.
3. L. M. Cabalín, D. Romero, J. M. Baena and J. J. Laserna, Fresenius’ J. Anal. Chem., 1999, 365, 404–408.
4. R. A. Multari, L. E. Foster, D. A. Cremer and J. M. Ferris, Appl. Spectrosc., 1996, 50, 1483–1499.
5. S. Palanco, J. M. Baena and J. J. Laserna, Spectrochim. Acta, Part B, 2002, 57, 591–599.
6. S. Palanco and J. J. Laserna, Rev. Sci. Instrum., 2004, 75, 2068–2074.
7. A. K. Knight, N. L. Scherbarth, D. A. Cremers and J. M. Ferris, Appl. Spectrosc., 2000, 54, 331–340.
8. M. A. Ismail, H. Imam, A. Elhassan, W. T. Youniss and M. A. Harith, J. Anal. At. Spectrom., 2004, 19, 489–494.
9. S. Palanco, L. M. Cabalín, D. Romero and J. J. Laserna, J. Anal. At. Spectrom., 1999, 14, 1883–1887.
10. C. López-Moreno, S. Palanco and J. J. Laserna, Spectrochim. Acta, Part B, 2005, 60, 1034–1039.
11. X. Mao, W. T. Chan and R. E. Russo, Appl. Spectrosc., 1997, 51, 1047–1054.
12. L. M. Cabalín, D. Romero, P. Lucena, J. M. Baena and J. J. Laserna in Proceedings of the International Conference on Process of Analytical Chemistry in Steel and Metals Industries, ed. R. Tomellini, Office of Officials Publications of the European Communities, L-2985 Luxembourg, 1999, pp. 336–344.
13. Y. I. Lee, T. L. Thiem, G. H. Kim, Y. Y. Teng and J. Sneddon, Appl. Spectrosc., 1992, 46, 1597–1604.
14. A. De Giacomo, M. Dell’Aglio, F. Colao and R. Fantoni, Spectrochim. Acta, Part B, 2004, 59, 1431–1438.
15. B. C. Castle, K. Talabardon, B. W. Smith and J. D. Winefordner, Appl. Spectrosc., 1998, 52, 649–657.
16. H. M. Solo-Gabriele, T. G. Townsend, D. W. Hahn, T. M. Moskal, N. Hosein and J. Jambeck, Waste Manage., 2004, 24, 413–424.
17. T. M. Moskal and D. W. Hahn, Appl. Spectrosc., 2002, 56, 1337–1344.
18. H. Bette, R. Noll, G. Muller, H. W. Jansen, C. Nazikkol and H. Mittelstadt, J. Laser Appl., 2005, 17, 183–190.
19. C. Lopez-Moreno, S. Palanco and J. J. Laserna, J. Anal. At. Spectrom., 2005, 20, 1275–1279.
20. A. I. Whitehouse, J. Young, I. M. Botheroyd, S. Lawson, C. P. Evans and J. Wright, Spectrochim. Acta, Part B, 2001, 56, 821–830.
21. D. Bulajic, G. Cristoforetti, M. Corsi, M. Hidalgo, S. Legnaoli, V. Palleschi, A. Salvetti, E. Tognoni, S. Green, D. Bates, A. Steiger, J. Fonseca, J. Martins, J. McKay, B. Tozer, D. Wells, R. Wells and M. A. Harith, Spectrochim. Acta, Part B, 2002, 57, 1181–1192.

---