Design, construction and assessment of a field-deployable laser-induced breakdown spectrometer for remote elemental sensing

Spectrochimica Acta Part B 61 (2006) 88 95 www.elsevier.com/locate/sab Design, construction and assessment of a field-deployable laser-induced breakdown spectrometer for remote elemental sensing Santiago Palanco, Cristina López-Moreno, J. Javier Laserna * Department of Analytical Chemistry, Faculty of Sciences, University of Málaga, E29071 Málaga, Spain Received 10 August 2005; accepted 4 December 2005 Available online 20 January 2006 Abstract A field-deployable laser-induced breakdown spectrometer for measurements in the hundreds of meters range has been presented. The system is capable of elemental analysis with no previous preparation and in near real time, with the only requirement of a free line-of-sight between the instrument and the sample. Main factors influencing LIBS performance at stand-off distances are outlined. LIBS signal is shown to depend on range of analysis, peak power, beam quality, laser wavelength and optics dimensions. A careful control of focusing conditions has been shown to be of importance to avoid interferences from air breakdown by the stand-off focused beam. D 2005 Elsevier B.V. All rights reserved. Keywords: Laser; Breakdown; Plasma; Spectrometry; Field; Portable; Remote 1. Introduction The most significant advantage of remote optical measurement techniques over more conventional techniques lies in the fact that they can perform real time, in situ analysis along an open path. These techniques have been widely used in a variety of applications, astrochemistry being chronologically the first. Nowadays, the most relevant application is environmental monitoring. In the past, this science has relied on methods such as differential optical absorption spectrometry (DOAS), as introduced by U. Platt and D. Perner in the late seventies [1]. Since then, it has been used in both terrestrial and satellite based systems [2]. Another method extensively used in earth monitoring is LIDAR [3], an acronym that stands for light detection and ranging, which represents the optical analogue of RADAR. Nevertheless, these techniques are restricted to the analysis of molecular species in gases, aerosols or liquids but there are situations in which the elemental analysis of solid samples is required. Laser-induced breakdown spectroscopy (LIBS) has been employed in a wide variety of applications and has kept in a constant growth since its introduction in the * Corresponding author. E-mail address: laserna@uma.es (J.J. Laserna). early eighties [4,5]. Among them, it is worth mentioning those performing analysis of both pure liquids [6] and colloidal solutions [7], industrial applications in the field of semiconductors [8,9] and steel production control [10 12] for which high-demanding applications like the real-time analysis of molten steel has been demonstrated [13,14]. In 1987, a paper was published describing the analysis of metal samples at distances ranging between 0.5 to 2.4 m carried out by using focusing lenses of different focal length, and collecting the light emitted by the laser-induced plasma by means of a fiber optic cable [5]. The same basic configuration was stretched up to 24 m in 1993 [15], and two years later a similar singleaxis system was reported [16]. Nonetheless, to perform analysis in dangerous or even hostile environments it would be desirable to develop a system capable of data acquisition at stand-off distances from the sample. Eye-catching examples of this trend are current projects featuring LIBS as one of the analytical techniques to be included in the next generation of planetary exploration vehicles [17 19] or its application to monitoring in hostile industrial environments [14,20,21] or to environmental monitoring [22]. Recently, results have been reported exploring the feasibility of LIBS for distances beyond 100 m [23,24]. To further exploit the inherent advantages of the technology, a mobile LIBS-based system capable of performing stand-off measurements in the hundreds of meters range has been built. 0584-8547/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.sab.2005.12.004

S. Palanco et al. / Spectrochimica Acta Part B 61 (2006) 88 95 89 The system is capable of elemental analysis with no previous preparation and almost in real time, with the only requirement of a free line-of-sight between the instrument and the sample. In the present paper, the design and construction considerations of the instrument are reported and its performance is discussed on the basis of the spectral response, the stand-off irradiance achieved upon the range of analysis and its influence on plasma properties, as well as the influence of the focusing accuracy to the stand-off LIBS signal. 2. Experimental The layout of a typical bench-top LIBS instrument is integrated by a pulsed laser, optical components for beam conditioning, guiding and focusing, a sample holder and optical components for collecting and guiding the plasma light to a spectrograph. The dispersed light is usually detected by an intensified CCD detector used in conjunction with a delay generator to synchronize the detection gate to the time interval when atomic emission occurs in the plasma. Further data processing is often carried out on a computer. Fig. 1A shows the setup of the present remote LIBS instrument. The main physical differences with a bench-top LIBS system lie in the optics used for sending the laser beam and gathering of the plasma light. While simple optics can be used in a bench-top setup, focusing and gathering must be carefully studied when it comes to remote operation. In previous works by this group, commercial components such as beam expanders or astronomy Fig. 1. (A) Schematic of the instrument showing the optical module (top view) and the electronics module: 1, laser source; 2, folding mirror; 3, lenses; 4, dichroic mirror; 5, focusing mirror; 6, primary mirror; 7, fiber optic cable; 8, Mechelle spectrograph; 9, detector; 10, PC; 11, laser power supply. (B) Snapshot of the instrument showing the optical module (top) including the optical bench and the spectrometer and the electronic module (bottom) illustrating the rack with the PC and the step-motor drivers, the delay generator and the laser power supply.

90 S. Palanco et al. / Spectrochimica Acta Part B 61 (2006) 88 95 telescopes have sufficed for the task. However, these components were installed to laboratory instruments which were mostly used indoor or as much under controlled outdoor climate conditions. The aim of the present work was to build a field-deployable instrument capable of approaching real-life distances of measurement in the hundreds of meters. Therefore, a completely new concept had to be conceived. The laser source was a Quantel Brilliant, a Q-switched Nd:YAG laser pulsing at 350 mj and 20 Hz in the fundamental wavelength. Owing to the radially variable reflectance of the output coupler, the spatial energy distribution of the multimode beam is Gaussian-like. There are two main reasons for choosing such a laser: (i) the proven reliability and ruggedness of solid-state lasers which are paramount for a field instrument and (ii) the short pulse length of this particular model (4.6 ns) which, from a relatively compact laser head, provides high-power pulses of up to 78 GW with a divergence of 4.5 10 4 rad. An additional component focuses the beam to a sufficiently small spot in the remote sample surface in order to provide the necessary irradiance to produce a plasma (10 7 10 8 Wcm 2 for most solids) without inducing breakdown of the air in the sample proximity. Air breakdown would prevent laser energy from reaching sample surface and, although the required irradiance level in clean air (10 4 particle cm 3 ) is in the order of 10 10 W cm 2, in the presence of seeding particles, the breakdown threshold may decrease by at least one order of magnitude. In order to reduce the spot size produced at a large distance of the focusing optics, the laser beam must be significantly expanded. In the coaxial setup previously built by our group [23], the diameter of the expanded beam was limited since focusing optics caused an obstruction to light gathering. Overcoming this problem without oversizing the collection optics requires the design of a single optical arrangement in charge for both focusing and gathering. A feasible system is an off-axis telescope where the light reaching the primary mirror is transmitted in absence of any obstruction by the secondary mirror. Known as a Herschelian telescope, this design is common in collimation and target projection systems. However, owing to the own off-axis nature of these systems, the mirrors must be corrected for optical aberrations which increases their cost. The following points have been considered at the design stage: (i) the instrument is not required to perform well over a wide field-of-view. Moreover, in the scope direction, the UV radiation of plasma about 3 mm in diameter can be considered as a point source at distances beyond 25 m. (ii) Although the beam expander is required to perform well at only one wavelength, the returning plasma light spans over the whole UV VIS range. (iii) Limitations of manufacturing capabilities make it impossible to achieve the maximum theoretical performance of a paper design in real hardware and, in practice, a system built from a less-optimum design based on a small number of simple surfaces can perform as well if not better, while keeping costs down. Optical requirements have been accordingly downgraded and commercial components have been used in preference to custom-built. The primary mirror is a reflection-enhanced aluminum mirror (diameter 203.2 mm, f/10). In order to prevent spherical aberration, the curvature of the mirror is parabolic. However, using a 200 mm curved mirror under an off-axis configuration will produce appreciable astigmatism and comma for aperture ratios faster than approximately f/40. Thus, only an off-axis section of the mirror is used to transmit the laser beam. The diverging element features a pair of bestform fused-silica 1064-nm-coated lenses (L1: diameter 25.4 mm, effective focal length 189.0 mm, L2: diameter 25.4 mm, effective focal length 62.1 mm). This lens pair is shifted and tilted with respect to the laser beam axis generating an aberration which is similar and opposite to that induced by the primary mirror. Such aberration partially corrects that produced by the off-axis parabolic mirror. Although the corrected image is not completely free from astigmatism and comma, the theoretical spot produced is diffraction limited. A 1064 nm dichroic mirror placed between the lens pair and the primary mirror is used to fold the beam towards the telescope axis, this mirror being the optical element in charge for the separation of the returning plasma light. The fact that light gathering does not employ refracting optics frees the return path from chromatic aberration. Eventually, the plasma light enters a fiber optic cable (FOC) with an aperture ratio close to the f/10 of the primary mirror. For the sake of construction simplicity, a FOC (fused silica, 600 Am core, 2 m length) is used instead of placing the spectrograph slit right at the back focus of the system. Losses related to configurations with and without FOC have been discussed elsewhere [39]. Two spectrometers have been used in this work: an Andor Mechelle ME5000 (195 mm focal length, F/7, k /Dk 5000) fitted with an Andor istar DH734 intensified CCD (1024 1024 pixels, 13.6 Am 2 pixel, intensifier diameter 18 mm). The whole 200 975 nm range can be acquired at once with an acquisition rate below 2 Hz. The second spectrometer is an Oriel MS125 (125 mm focal length, F/3.8, 3600 line mm 1 grating, 25 Am slit) fitted with an Andor Instaspec DH501 intensified CCD detector (1024128 pixels, 26 Am 2 pixel, intensifier diameter 25 mm). The wavelength range of this system is a 37 40 nm window manually set from 200 to 405 nm. The maximum acquisition rate of this spectrometer is 55 Hz although the maximum pulse repetition rate of the laser is 20 Hz. The system focus is adjusted by altering the distance between the diverging lenses and the primary mirror. This is achieved by inserting a motorized flat aluminum mirror (diameter: 76.2 mm, k / 10) between the mentioned optical elements. The position of this mirror has been chosen attending to several criteria such as aluminum damage threshold, compactness and minimizing of the primary-mirror off-axis angle. A further role of this mirror is to fold the telescope axis, making it more compact. In addition, a much simpler instrument operation is obtained with such design since the focus of both light paths can be controlled by a single element. Still, with the described configuration, the system would be only optimized for a given focusing distance, its performance being lower at a different one. To minimize this effect, additional motorization has been provided for the remaining optics which enables some optical correction along the whole

S. Palanco et al. / Spectrochimica Acta Part B 61 (2006) 88 95 91 Fig. 2. Comparison of local and remote LIB spectra both obtained with the instrument (see text for acquisition details). The grayed spectra correspond to the transmission of the dichroic mirror in the same wavelength range. working range of the instrument. Nevertheless, this feature has not been used for this work and all data shown here has been obtained with the system optimized for a 120 m range. Fig 1B shows a snapshot of the system. 3. Results and discussion The performance of the system in terms of useable wavelength range was evaluated in the first place. A Ti standard sample was chosen owing to the highly populated spectra exhibited by this element. The LIB spectrum was acquired at 30 m and then compared to a LIB spectrum acquired locally at approximately the same laser irradiance and using the FOC as the only collection element. Fig. 2 illustrates that both spectra look similar in the VIS range. Conversely, the performance of the system seems to decrease in the UV, with zero transmission at 350 370 nm and below 300 nm. Given that the FOC was used for both spectra, the source for this issue was traced to the remaining optical components and in particular, to the fused-silica dichroic mirror. The overall transmittance spectrum (i.e. substrate plus both coatings) of this element has been plotted in Fig. 2. As shown, the areas of lower transmittance of the mirror match the regions where the remote LIB spectrum is attenuated. Particularly, the dip at 355 nm coincides with the third harmonic of the fundamental Nd:YAG radiation for which reflectance of the mirror is designed to be a maximum. This effect is inherent to reflective dielectric coatings and cannot be avoided. 3.1. Range dependence of LIBS signal The performance of the system was evaluated as a function of the analysis range. An aluminum sample was employed for this task and the LIBS signal of the 396.152 nm Al(I) line was averaged over 100 laser shots. Fig. 3 illustrates the behavior of the referred line versus the distance from the system to the Fig. 3. Behavior of the LIBS signal registered for the 396.152 nm Al(I) line with the measuring range (left axis). The fitting line corresponds to an inverse fifth power of the working range. On the right axis the full width at half maximum of the same atomic line is plotted.

92 S. Palanco et al. / Spectrochimica Acta Part B 61 (2006) 88 95 target. It must be remarked that owing to detector saturation at distances closer than 35 m, the maximum microchanel plate (MCP) gain setting (450 counts photoelectron 1 ) was not used being necessary to lower this setting to about 20 counts photoelectron 1 for the whole experiment. As shown, the signal follows a strong decay with the range. This effect was attributed to several factors which can be grouped in those related to the formation of the plasma and those related to light gathering. In a former work [23], the gathered LIBS signal, S LIBS was shown to be proportional to the solid angle of collection, X and thus, to decay with the inverse square range, r for identical plasma conditions: X S LIBS n D ¼ n E 4p ¼ n A M E 4pr 2 ð1þ where n D is the number of detected photons of a given wavelength, n E is the number of emitted photons of the same wavelength, and A M is the area illuminated in the primary mirror. However, the data in Fig. 3 does not fit to an inverse square function, evidencing that plasma conditions are changing with range. This fact is confirmed by the right-axis plot in Fig. 3 which shows a progressive broadening of the 396.152 nm Al(I) line with the working range. Further, since the number of emitted photons of a given wavelength is proportional to the number of emitting species in the plasma, the signal is expected to decrease as irradiance lowers with range. To confirm this point, the average ablation rate (AAR) was measured by firing a known number of pulses to small samples of known weight. The weight loss divided by the number of pulses was taken as the AAR. The results are plotted as a function of the range in Fig. 4 along with a fitting line corresponding to an inverse third power function of the range. Additionally, the top axis shows the irradiance, I achieved by the system at the target which is a function of the pulse power, P and of the radius of the focused Gaussian laser spot, w: w ¼ 2kr pd M 2 ð2þ where k is the laser wavelength, D is the beam diameter at the focusing optics and M 2 is the beam quality parameter. Thus, I ¼ ppd2 4k 2 ð3þ M 4 r 2 Given the single-axis design of this system, D coincides with the diameter of the section of the primary mirror illuminated by the laser beam, A M and Eq. (3) can be rewritten as: I ¼ PA M k 2 ð4þ M 4 r 2 This equation fits the observed inverse square dependence of the irradiance with the range and therefore the empirical 3/2 power dependence of AAR with the irradiance as shown in Fig. 4. AAR I 3=2 ¼ P3=2 A 3=2 M k 3 ð5þ M 6 r 3 Assuming the number of emitted photons is proportional to the AAR, Eq. (5) could be substituted in Eq. (1): S LIBS P3=2 A 5=2 M 4pk 3 ð6þ M 6 r 5 Eq. (6) does not intend to be a general equation for stand-off LIBS since approximations have been done to its formulation. However, the authors believe it is a working model of the main factors influencing the LIBS signal such as its inverse fifth Fig. 4. Average ablation rates measured at different ranges. The fitting line corresponds to an inverse third power function of the range. At the top axis, the laser irradiance is illustrated, showing a 3/ 2 power dependence of AAR with this parameter.

S. Palanco et al. / Spectrochimica Acta Part B 61 (2006) 88 95 93 power decay with range which is in good agreement with experimental results in Fig. 3. Besides pulse power, Eq. (6) reveals the critical importance of laser wavelength, beam divergence and dimensions of the optical element in order to achieve a stand-off plasma leading to a high signal of analytical quality. On the other side, there are aspects missing from this approximation such as spectral response of the system, losses associated to light transmission in the air at long distances, or factors related to the properties of the laser-induced plasma. 3.2. Acquisition timing Fig. 5 evidences the strong dependence of plasma properties on the range. The behavior of the 396.152 nm Al(I) line in terms of FWHM has been plotted as a function of the acquisition delay and range. The instant when the laser pulse is fired is taken as the reference for acquisition timing (delay = 0). As shown at the right side of the figure, the FWHM seems to decrease with the delay for a given distance. This is the expected behavior for plasma decay after the end of the laser pulse. However as illustrated by the plot in the top of Fig. 5, the Al line seemingly broadens with distance even though the data in Fig. 4 shows a great decrease in irradiance and mass removal with increasing distance. The plateau in linewidth at ranges above about 60 m seems to coincide with the leveling off in mass removal but the linewidth data follows an apparently abnormal behavior: one would expect a narrower line in what should be a cooler, less dense plasma as the range increases. A possible explanation to the latter was attributed to the resonant nature of the Al line used for the study in conjunction with the lower temperature of plasmas produced at long ranges. Under such conditions, the number of non-excited Al atoms in the ground level is high which favors self-absorption and thus, an increase in the FWHM measured. This assumption is supported by the strong self-reversal exhibited by the Al line at long ranges and short acquisition delays as shown at the upper-right spectra in Fig. 5. An additional factor to be taken into account is the extra delay introduced by the stand-off measurement (note that the reference for acquisition is the firing of the laser pulse): the laser pulse and the plasma light take an additional 6.6 ns per every meter from the system to the target. At 80 m, the extra delay amounts to approximately 530 ns what would lower the measured FWHM from 1.5 nm to approximately 0.6 nm attending to the top plot. The mentioned effect can be also noticed in the upper-right figure just by comparing the spectra corresponding to 500 and 1000 ns delay. 3.3. Influence of plasma surrounding ambient Another factor to be held in account is the interaction of the plasma with the surrounding air favored by the long depth of focus, DOF related to stand-off configurations. The DOF, by definition, is twice as long as the Rayleigh length which, in turn is defined as the distance between pffiffi the positions in the optic path where the beam radius is 2 times larger than it is at the beam waist: DOF ¼ 8k r 2 M 2 : ð7þ p D According to Eq. (7), an ideal laser beam (M 2 =1) expanded to 120 mm and then focused at 100 m would produce a DOF of about 2 m which is 2500 times longer than that produced by a 6 Fig. 5. Behavior of the 396.152 nm Al(I) line width versus acquisition delay and working range. On the top-right, the line profiles at different acquisition delays for a 80-m range have been plotted.

94 S. Palanco et al. / Spectrochimica Acta Part B 61 (2006) 88 95 mm beam of similar characteristics focused at 100 mm. In practice, plasma formation capability extends beyond that number owing partly to the higher M 2 used in commercial lasers. This is an advantage when focusing over a remote target of unknown topography. However, interaction of the beam with dust particles in the sample surroundings which at high pulsing rates are often sample particles previously ablated may result in an undesired breakdown and reduction of the laser energy reaching the sample. Another factor to be considered is the interaction of the plasma with the surrounding atmosphere leading to beam shielding by the optically thick path. Fig. 6 illustrates the LIBS signal for Ti (Ti(I) 625.810, 625.810, 626.110 nm), H(I) (656.272, 656.285 nm), N(I) 746.831 nm and O(I) (777.194, 777.417, 777.539 nm) normalized to background versus the distance from the Ti sample to the focus of the instrument (DTF) which was set at 80 m. Negative values represent sample positions closer to the system. Each point is the average of 100 laser shots. As shown by the upper plot, Ti and H emission follow a similar trend with a maximum at the system focus and nearly symmetric branches at each side of this point. In fact, emission corresponding to these elements has been found to show a good correlation of 0.966. On the other side, emission corresponding to O and N follow a common behavior (R = 0.976) which is substantially different from that shown by Ti and H. There are two possible sources for H: ambient water vapor and H molecules adsorbed in the metal surface. The fact that H emission follows the metal trend instead of that of O and N leaves space to speculation about the metal surface as a possible source. However, from the experience gained from recent stand-off experiments we have observed that time progression of the H emission is usually different from that of O and N. Taking into account that light gathering spatially integrates the whole plasma emission, it seems feasible that the H behavior in Fig. 6 could be attributed to plasma heterogeneities in temperature and electron density. Given the difficulties to setup a dry air chamber in the field, this point could not be addressed and will be done in a further work. Oppositely, O and N emission is clearly sourced in molecules of the ambient air being dissociated when sample is positioned close and beyond the system focus, i.e. when the irradiance is a maximum on the plasma front-irradiance is below the breakdown threshold for clean air at 80 m. Additionally, the laser prints on thermal paper for each position have been included with Fig. 6. Besides the astigmatism affecting the instrument at an 80 m range, it can be noticed like the spots obtained at positive values of DTF are not as sharp as those obtained for negative values of DTF. This finding is indicative of the laser beam being somehow defocused by the heated/ignited air and is in good agreement with the higher emission intensity found for O and N at positive DTFs. Fig. 6. Signal to background ratios at an 80-m range for Ti (Ti(I) 625.810, 625.810, 626.110 nm), H(I) (656.272, 656.285 nm), N(I) 746.831 nm and O(I) (777.194, 777.417, 777.539 nm) versus the sample distance to the instrument focus. ( ) Ti, (?) H,(r) O, and (*) N. The laser prints in a thermal paper are also illustrated for each position. The acquisition delay was 1000 ns.

S. Palanco et al. / Spectrochimica Acta Part B 61 (2006) 88 95 95 4. Conclusions A field-deployable LIBS-based system for stand-off measurements in the hundreds of m range has been presented. The system is capable of elemental analysis with no previous sample preparation and almost in real time, with the only requirement of a free line-of-sight between the instrument and the sample. The instrument performance and, generally, LIBS performance at stand-off distances, has been shown to depend on several parameters such as range, peak power, beam quality, laser wavelength and optics dimensions. No influence of the atmosphere to the stand-off LIBS signal has been observed in the studied distance range. However, focusing conditions have been shown to be critical in order to avoid interferences from air breakdown by the stand-off focused beam. On-going work will be directed to extend the current working range of the instrument as well as the applications already affordable with the current range of analysis. Acknowledgements The authors thank Andor Technology for the loan of the Mechelle 5000 spectrogrpah used in part of the work and specially to Dr. Olivier Bernard for his efforts and interesting suggestions. This work was partially supported by Project CTQ2004-1854 of the Spanish Ministerio de Educación y Ciencia. One of the authors (CLM) thanks the Spanish Ministerio de Ciencia y Tecnología for a research fellowship. References [1] U. Platt, H. Pätz, Simultaneous measurement of atmospheric CH 2 O, O 3, and NO 2 by differential optical absorption, J. Geophys. Res. 84 (1979) 6329 6335. [2] U. Platt, in: M.W. Sigrist (Ed.), Air Monitoring by Spectroscopic Techniques, John Wiley and Sons, Inc., New York, 1994. [3] J. Allen, C. Platt, Lidar for multiple backscattering and depolarization observations, Appl. Opt. 16 (1977) 3193 3199. [4] D.A. Cremers, L.J. Radziemski, Detection of chlorine and fluorine in air by laser-induced breakdown spectroscopy, Anal. Chem. 55 (1983) 1246 1252. [5] D.A. Cremers, The analysis of metals at a distance using laser-induced breakdown spectroscopy, Appl. Spectrosc. 41 (1987) 572 578. [6] C.A. Sacchi, Laser-induced breakdown in water, J. Opt. Soc. Am., B 8 (1991) 337 345. [7] X.Y. Pu, W.Y. Ma, N.H. Cheung, Sensitive elemental analysis of aqueous colloids by laser-induced plasma spectroscopy, Appl. Phys. Lett. 83 (2003) 3416 3418. [8] M. Hidalgo, F. Martín, J.J. Laserna, Laser-induced breakdown spectrometry of titanium dioxide antireflection coatings in photovoltaic cells, Anal. Chem. 68 (1996) 1095 1100. [9] M. Milán, P. Lucena, L.M. Cabalín, J.J. Laserna, Depth profiling of phosphorus in photonic-grade silicon using laser-induced breakdown spectrometry, Appl. Spectrosc. 52 (1998) 444 448. [10] J.A. Aguilera, C. Aragón, J. Campos, Determination of carbon content in steel using laser-induced breakdown spectroscopy, Appl. Spectrosc. 46 (1992) 1382 1387. [11] L.M. Cabalín, M.D. Romero, C.C. García, J.M. Baena, J.J. Laserna, Timeresolved laser-induced plasma spectrometry for determination of minor elements in steelmaking process samples, Anal. Bioanal. Chem. 372 (2002) 352 359. [12] L. Peter, V. Sturm, R. Noll, Liquid steel analysis with laser-induced breakdown spectrometry in the vacuum ultraviolet, Appl. Opt. 42 (2003) 6199 6205. [13] J. Gruber, J. Heitz, N. Arnold, D. Bäuerle, N. Ramaseder, W. Meyer, J. Hochörtler, F. Koch, In situ analysis of metal melts in metallurgic vacuum devices by laser-induced breakdown spectroscopy, Appl. Spectrosc. 58 (2004) 457 462. [14] S. Palanco, S. Conesa, J.J. Laserna, Analytical control of liquid steel in an induction melting furnace using a remote laser induced plasma spectrometer, J. Anal. At. Spectrom. 4 (2004) 462 467. [15] J. Blacic, D. Pettit, D.A. Cremers, N. Roessler, Laser-induced breakdown spectroscopy instrument for elemental analysis of planetary surfaces, International Symposium on Spectral Sensing Research, Science and Technology Corporation, Hampton, Virginia, 1993, p. 302. [16] D.A. Cremers, A. Koskelo, J. Barefield, Remote elemental analysis by laser-induced breakdown spectroscopy using a fiber-optic cable, Appl. Spectrosc. 49 (1995) 857 859. [17] A.K. Knight, N.L. Scherbarth, D.A. Cremers, M.J. Ferris, Characterization of laser-induced breakdown spectroscopy (LIBS) for application to space exploration, Appl. Spectrosc. 54 (2000) 331 340. [18] F. Colao, R. Fantoni, V. Lazic, A. Paolini, G. Ori, L. Marinangeli, A. Baliva, Investigation of LIBS feasibility for in situ planetary exploration: an analysis on Martian rock analogues, Plant. Space Sci. 52 (2004) 117 123. [19] B. Sallé, J.L. Lacour, E. Vors, P. Fichet, S. Maurice, D.A. Cremers, R.C. Wiens, Laser-induced breakdown spectroscopy for mars surface analysis: capabilities at stand-off distances and detection of chlorine and sulfur elements, Spectrochim. Acta Part B 59 (2004) 1413 1422. [20] S. Palanco, J.M. Baena, J.J. Laserna, Open-path laser-induced plasma spectrometry for remote analytical measurements on solid surfaces, Spectrochim. Acta Part B 57 (2002) 591 599. [21] J.M. Vadillo, P.L. Garcia, S. Palanco, M.D. Romero, J.M. Baena, J.J. Laserna, Remote, real-time, on-line monitoring of high-temperature samples by noninvasive open-path laser plasma spectrometry, Anal. Bioanal. Chem. 375 (2003) 1144 1148. [22] C. Lopez-Moreno, S. Palanco, J.J. Laserna, Remote laser-induced plasma spectrometry for elemental analysis of samples of environmental interest, J. Anal. At. Spectrom. 19 (2004) 1479 1484. [23] S. Palanco, J.J. Laserna, Remote sensing instrument for solid samples based on open-path atomic emission spectrometry, Rev. Sci. Inst. 75 (2004) 2068 2074. [24] K. Stelmaszczyk, P. Rohwetter, G. Mejean, J. Yu, E. Salmon, J. Kasparian, R. Ackermann, J.P. Wolf, L. Woeste, Long-distance remote laser-induced breakdown spectroscopy using filamentation in air, Appl. Phys. Lett. 85 (2004) 3977 3979.