Vincenzo Palleschi

The physics and chemistry of the phenomenon have been well known for many years, and this knowledge can tell us how self-absorption can be not only “corrected,” but also tuned to our advantage in analytical applications of LIBS.

In a recent article published in this journal (1), I discussed seven common errors that may affect the results of laser-induced breakdown spectroscopy (LIBS) analysis. Among these errors, one of the recurring ones is related to misunderstanding the process of self-absorption in LIBS plasmas. Because many people working in LIBS still consider self-absorption as a problem yet to be solved, I am dedicating a full article to this topic. The physics and chemistry of the phenomenon have been well known for many years, and this knowledge can tell us how self-absorption can be not only “corrected,” but also tuned to our advantage in analytical applications of LIBS.

Self-absorption occurs in spectrochemistry when part of the radiation emitted by an extended source is re-absorbed before exiting the source and eventually reaching the detector. In LIBS quantitative analysis, the measurement of the radiation intensity determines the number concentration of the emitting analyte. When the effects of self-absorption are negligible, a linear relationship is typically obtained between the measured intensity and the analyte number concentration. Dealing with linear relations makes life easier for the analyst because it allows for the building of (linear) calibration curves from which limits of detection (LOD) and limits of quantification (LOQ) of the analyte being studied can be determined.

In most spectro-analytical techniques, self-absorption effects can be considered negligible; however, in LIBS analysis, we often make measurements involving large concentrations of the analyte (percent level). In this case, the probability for a photon emitted by an atom or ion in the plasma to be reabsorbed by another atom or ion before exiting the plasma could be quite high, and definitely higher than that for the same analyte diluted at sub-ppm concentration in, for example, inductively coupled plasma–optical emission spectroscopy (ICP-OES) analysis.

The result of a strong self-absorption in LIBS is a nonlinear saturation effect, which is observed in the calibration curve; when the number concentration of the analyte increases, the measured intensity of the emission line does not grow proportionally. At large concentrations, the calibration curve saturates; it is well-known in the literature that in a log–log scale, the slope of the calibration curve changes from 1 to 1⁄2 with the increase of the analyte concentration and the effect of self-absorption.

Don’t Mistake Self-Reversal for Self-Absorption

Two observations should be made at this point, both correlated to frequent analytical errors. The first error that is often made in dealing with self-absorption in plasmas is mistaking this effect with the self-reversal phenomenon. Self-reversal manifests as a narrow dip on top of the emission line. Its presence is related to self-absorption, but self-reversal shows up only in the presence of spatial gradients of the plasma temperature and electron number density, which usually occur at the periphery of the plasma. Therefore, although the presence of self-reversal is an indication that the self-absorption phenomena in the emitting source are not negligible, as well as the spatial inhomogeneities in the plasma, its absence does not mean that self-absorption can be considered negligible.

Understand What Is Really Happening in a “Correction”

The second error that is, in my opinion, more subtle and substantial is associated to the need that some people working in LIBS feel to “correct” the effects of self-absorption. Although the term “correction” can be used at a colloquial level, it is nevertheless important to know that this term does not correspond to the true nature of the phenomenon.

An emitting source is also an absorbing source. There is no problem to be solved unless we insist on using an unrealistic treatment of the emission process in which each emitter is considered independent from the others. Instead, what we could do is take into account the physical processes occurring during the propagation of the radiation in the plasma (spontaneous emission, absorption, and stimulated emission, with the latter being in general negligible). In the approximation of homogeneous plasma in local thermal equilibrium, the number of photons emitted by the laser-induced plasma at a given wavelength along the line of sight of the detector can be easily determined by solving a simple differential equation:

In the above equation, the ε coefficient is proportional to the number of emitters and to the spontaneous emission coefficient 

 of the transition (see Figure 1). Following the Einstein treatment of the propagation of radiation in a two-level atomic system (the same which is used to model the laser effect), it’s easy to demonstrate that the absorption coefficient, 

(λ) is the Planck black body radiation function at the plasma temperature (

). The radiation propagation equation can be easily solved, at each wavelength, along the line of sight of the detector.

FIGURE 1: Schematic representation of the phenomena occurring within an emitting plasma.

The shape of the lines emitted by an optically thin plasma (

 <<1) is usu- ally well approximated by a Lorentzian function, peaked at the wavelength of the transition. However, because the absorption coefficient is proportional to the emission coefficient, the effect of self-absorption is more evident where the emission is stronger (at the peak of the line). This deformation of the shape of the emission lines makes the calculation of their integral intensity slightly more complex, but it can be demonstrated that, at least for low to moderate self-absorption, the integral intensity of the line would be given by:

) is the absorption coefficient calculated at the peak of the emission line, and 

 is the integrated intensity (total number of photons emitted) that would have been measured on the detector if the plasma was optically thin.

It can also be demonstrated that the full width at half maximum (FWHM) of the self-absorbed line will increase with respect to the corresponding FWHM in optically thin plasma conditions following the same square root behavior:

Therefore, if we can estimate the FWHM of the line in the optically thin limit, it would be possible, by measuring the FWHM of the self-absorbed line, to estimate the value of the factor:

Substituting it in the relation for the line intensity, the simple relation is obtained:

A more precise calculation leads to the formula 

, which is valid also in case of strong self-absorption. 

 is, by definition, proportional to the number of emitting atoms or ions in the plasma; therefore, the above equation allows recovering a linear relation between the measured intensity, 

, and the concentration of the analyte (see Figure 2).

In conclusion, we can safely say that the self-absorption effect in LIBS is not a problem anymore. On the contrary, the simple treatment of the plasma as a homogeneous emitter at a temperature 

 makes it possible to recover the linear dependence between signal intensity and number concentration of the analyte, which characterizes an optically thin plasma. Moreover, knowledge of the physics of the phenomenon contributes to improving the precision of the determination of fundamental plasma parameters as transition probabilities and Stark coefficients of non-optically thin emission lines (2). The numerical simulation of the radiation propagation equation has also demonstrated that the approximation of homogeneous plasma is not critical for the determination of the Stark coefficients of self-absorbed LIBS emission lines (3).

(2) S.M. Aberkane, A. Safi, A. Botto, B. Campanella, S. Legnaioli, F. Poggialini, S. Raneri, F. Rezaei, and V. Palleschi. 

https://doi.org/10.1016/j.jqsrt.2021.107714

 is a Senior Researcher at the National Research Council of Italy in Rome, Italy, and a professor with Pisa University in Pisa, Italy. Direct correspondence to: 

Articles

Avoiding Misunderstanding Self-Absorption in Laser-Induced Breakdown Spectroscopy (LIBS) Analysis

	Laser-induced breakdown spectroscopy (LIBS) is an "easy" technique. Setting up a LIBS experiment may take only a few hours; the equipment required is often quite standard (a laser, a couple of lenses, a spectrometer, and a computer) and relatively cheap, with respect to other spectro-analytical techniques. However, planning a sound LIBS experiment and correctly analyzing the resulting spectra is not a trivial task. In my 30 years of experience in the field (my first paper on LIBS was published in 1991), I've seen several recurrent errors that can irreparably plague the results of LIBS measurements.

	Here is my personal compilation of the top seven errors that one should avoid in LIBS analysis.

	As strange as it might seem, many LIBS experiments fail because of a misidentification of the spectral lines of one or more elements. The LIBS technique can detect practically all the elements in the periodic table (~100 x 2, considering neutral and ionized species), and some of these elements may show hundreds of spectral lines. It is, therefore, very easy to make a mistake, and assign a spectral line to the wrong element. A minimal shift of the LIBS spectrum might "transform" lines of common elements (calcium, magnesium, sodium) into other (sometimes very dangerous or exotic) elements. This is why the identification of the emitting species in the LIBS spectrum should never be done based on a single line; instead, the multiplicity of information coming from the different emission lines of the elements should be exploited!

	The LIBS technique is not particularly sensitive for most elements of interest. For this reason, many different methods have been developed, aimed to improve the analytical performance of the technique. One of the most important figures of merit of an analytical technique is the limit of detection (LOD), which is the minimum quantity of the element that can be detected in a given measurement (the lower, the better). Under certain conditions, the LOD of a given element can be derived from the corresponding calibration curve using the well-known formula 3

 is the standard deviation of at least 10 independent measures of the analyte on a suitable blank, and 

 is the slope of the corresponding calibration curve. One should keep in mind that in the approximation of an ideal calibration curve with an intercept equal to 0, 

*LOD represents the minimum signal that can be distinguished from the fluctuations of the signal of the blank with a limit of confidence of 3

	Therefore, any indetermination on the intercept, coming from the uncertainty in the linear fit of the calibration curve, for example, would contribute to the 

 of the blank. This effect can be much higher than the measured noise, and therefore it should be considered in the calculation of the LOD.

	Moreover, being able to detect an element does not mean to be able to quantify it. The limit of quantification (LOQ) of an element is conventionally taken as 3–4 times its LOD. For determining the LOD, it is appropriate to have as the last point before the blank at least one sample with concentration close to the LOQ.

	A common error in LIBS analysis is the determination of the calibration curves using only a few standards (typically <10), without measuring the blank and with the lowest point corresponding to concentrations hundreds of times higher than the LOD calculated with the 3

	The local thermal equilibrium (LTE) approximation is often used for modeling LIBS plasmas. In this approximation, the laser-induced plasma is described as a system where all the processes are at equilibrium, except the radiation. In the framework of LTE, it is possible to define a plasma temperature that describes the system, and is the same for all the processes at equilibrium. The condition of having a negligible loss of energy by radiation in the time needed by the system to reach the equilibrium of the other distributions is expressed by the McWhirter criterion:

	which is, however, necessary and sufficient only for a stationary and homogeneous plasma. In case of non-stationary or non-homogeneous plasmas (for example, in typical LIBS plasmas), two other criteria must be fulfilled for having an LTE plasma:

		In case of non-stationary homogeneous plasma, the time needed for the establishment of excitation and ionization equilibria must be much shorter than the time of variation of thermodynamic parameters.

		In case of non-stationary, non-homogeneous plasma, the diffusion length of atoms or ions, during a time period of the order of the relaxation time to the equilibrium, must be shorter than the variation length of temperature and electron number density in the plasma.

	LIBS plasmas are highly dynamic systems, characterized by fast expansion and corresponding changes in electron temperature and number density. These parameters vary in time; therefore, their determination must be done using time-resolved spectrometers, with gate time typically lower than 1 µs. As we have seen, the assessment of the validity of LTE approximation requires the determination of plasma electron temperature and number density. A common error in LIBS experiments is the use of time-integrated (or time-resolved with long gate times) spectrometers for determining the fulfillment of LTE conditions in the plasma or for determining the composition of samples applying the calibration-free LIBS (CF-LIBS) algorithm (which is based on the LTE approximation).

	Several recent works in LIBS describe plasma self-absorption as the major open problem in LIBS analysis. In fact, there are currently many methods for evaluating self-absorption in LIBS plasmas, as well as using this effect for improving the performance of LIBS as an analytical method. It is, therefore, inappropriate to present self-absorption as a problem in a LIBS work without at least mentioning the main strategies that can be used for evaluating (and thus compensating for) the importance of these effects. Not considering them is one of the most common errors in LIBS experiments. Moreover, one should not confuse self-absorption (a phenomenon that is always present in plasmas, being intrinsic in the interaction of radiation with plasmas) with the self-reversal phenomenon, which manifests with a narrow dip at the center of the spectral line. Self-reversal occurs only when the plasma is not homogeneous (colder at the borders).

	The introduction of double-pulse LIBS about 20 years ago has been one of the major breakthroughs in LIBS research. Enhancements of the signal up to two orders of magnitude have been demonstrated, using double-pulse LIBS (typically in collinear mode, with two nanosecond pulses delayed by several hundreds of nanoseconds). Orthogonal (reheating) DP-LIBS is also used, although the experimental setup is more complex and the enhancements relatively modest.

	A complete modeling of collinear dual-pulsed LIBS (DP-LIBS) has existed for at least 15 years (and the basic mechanism of enhancement was already explained by Pershin in 1989). The explanation of the enhancement is related to the shock wave produced by the first laser pulse, which creates a favorable low-density environment in which the second laser pulse can create the analytical plasma.

	As for self-absorption, the theoretical and practical knowledge about DP-LIBS is well established; however, from time to time new "explanations" of the DP-LIBS effect are proposed.

	Discussing alternative theories is always legitimate, although the experimental confirmations of the fundamental role of the first shock wave are overwhelming. Therefore, any possible alternative theory about the DP-LIBS enhancement mechanism must account for the phenomena that are already successfully described by the current theoretical treatment.

	LIBS spectral analysis can be quite complex. In these last years, several chemometric methods have been proposed that can be effectively used for improving the performance of LIBS, for simplification, classification, or quantitative analysis of the samples.

	However, using these methods without any verification of the results or knowledge of the principles at the basis of the algorithms used can be very dangerous. The most common error in LIBS chemometric analysis is the poor comparison of the results of the algorithms with the standard tools that have been used for decades, before the advent of artificial neural networks (ANNs), machine learning, and similar chemometric tools. In LIBS analysis, one should demonstrate that the chemometric methods are effectively better with respect to simpler approaches. Before using a neural network for quantitative LIBS analysis, for example, one should demonstrate that classical univariate calibration and linear multivariate analysis (for example, partial least square [PLS] analysis) do not work well.

	The number of samples must be high enough to guarantee the statistical validity and significance of the results obtained, and these results must be correctly validated on external data-not used for the training of the algorithm.

	When chemometrics is used for classification, care must be taken to avoid systematic effects; the samples must be prepared in the same way and analyzed in random order. One must also take care of not mistaking causes and effects; in a LIBS analysis for the screening of cancerous tissues, for example, a discrimination between normal and cancerous tissue could occur if the cancer patients have been treated with specific drugs that accumulate in the tissues, while the healthy patients were not.

	The algorithms used in chemometrics are very powerful, but-just like Mickey Mouse with the enchanted brooms in the movie "Fantasia"-one should never lose control of them!

 is with the Applied and Laser Spectroscopy Laboratory at ICCOM/CNR, in Pisa, Italy. Direct correspondence to: 

Seven Common Errors to Avoid in LIBS Analysis

The monitoring of the processes in the steel and metal industry calls for techniques that are capable of measuring the composition of metallic alloys at a distance and on moving conveyor belts. In many cases, such as in the recycling of automotive scrap, the geometry of the objects to be analyzed can vary, and surface coatings can be present. In this column installment, we discuss the application of laser-induced breakdown spectroscopy (LIBS) to two industrial projects to illustrate how the above-mentioned problems can be faced and successfully resolved.

	
	
	Laser-induced breakdown spectroscopy (LIBS) (1) is an analytical technique that has recently been applied to determining the composition of materials of interest in many fields, such as industry (2â5), environment (6â8), biomedicine (9,10), forensic science (11â13), cultural heritage (14â16), and archaeology (17,18).
	
	LIBS is an elemental technique, and therefore it is particularly suited for the analysis of metals (2â4). Its relatively poor analytical performances (19,20) with respect to the typical laboratory techniques used in the field, in conjunction, on the other hand, with its capability of operating in situ, remotely, and without any treatment of the samples, suggests that the ideal application of LIBS is in the field of in situ analysis and, in particular, industrial processes monitoring.

	The Applied and Laser Spectroscopy Laboratory in Pisa, Italy, has participated in a number of projects for the development of prototypes and procedures for efficient monitoring of industrial processes. Some examples include the analysis of steel pipes at high temperatures (CE Project LIBSGRAIN [21]), precious alloys in the recycling industry (22), and exhaust gases from electrical power plants (23).

	Our laboratory is currently involved in two important projects that have several points in common: Both of them are related to the analysis of metals in an industrial environment by LIBS. The first project is called laser-based continuous monitoring and resolution of steel grades in sequence casting machines (LACOMORE) started last year with the funding of the European Commission in the framework of the Research Fund for Coal and Steel. The purpose of the LACOMORE project is to accurately model the sequential cast process (24) and experimentally determine the variation of the steel composition during the process using LIBS technology. The second project, called selective recovery of nonferrous metal automotive shredder by combined electromagnetic tensor spectroscopy and laser-induced plasma spectroscopy (SHREDDERSORT), was founded by the European Commission in the 7th Research Framework. The purpose of this project is aimed at developing a new dry sorting technology for nonferrous automotive shredders.

	In this column installment, we compare the specific requirements of the LIBS experimental apparatus for these two cases and outline the strategies that are similar in the two applications, as well as the different implementation of the LIBS technology in the each case.

Comparison of the Experimental Strategies

Experimental Strategy for Remote Measurements


	
	
	In both of the two industrial applications of the LIBS technology, the target of the measurement is moving at a steady speed. However, in the LACOMORE case the geometry of the measurement is fixed (the height and lateral dimensions of the slab or billet are always the same), while in the SHREDDERSORT situation the size and shape of the scraps vary greatly on the conveyor belt, which moves much faster (4 m/s) with respect to the LACOMORE case (1.5 m/min).

	Thus, the measurements strategies are different. In both cases, the experimental constraints impose the need for remote measurements; for the LACOMORE application, a single laser system is used, given the fixed geometry of the analysis. The LIBS system is a variation of the stand-off LIBS instrument realized by the Laserna group at the University of Malaga (25), which has been successfully tested in the typical conditions of the steel plants where the prototype will be installed (5-m distance between the laser and the target, with a target temperature around 900 Â°C). In the Malaga LIBS instrument, the laser beam is focused using a high-power beam expander on the steel slab or billet surface. The laser beam energy is 400 mJ in 5-ns pulses (full width at half maximum [FWHM]), with a repetition rate of 10 Hz.

	In the SHREDDERSORT case, on the other hand, the automotive scraps do not always pass the measurement point in the same position, so the experimental strategy in this case is based on the use of a battery of several low-cost lasers that fire only when the scrap passes in the corresponding position of the conveyor belt (see Figure 1). The lasers provide an energy of 1 mJ in 2-ns FWHM pulses, with a repetition rate of 10 Hz.

	The distance between the laser and the sample is 70 cm on average, and the laser beams are focused with long-focal-length lenses to maximize the depth of field. In fact, the variation in the heights of the samples and, consequently, in the distance between the lens and the sample surface, does not allow for a tight focusing of the laser beam.

Light Collection and Spectral Separation


	
	The LIBS instrument that is used for the LACOMORE application collects the LIBS signal using a Cassegrain telescope (25). In the SHREDDERSORT application, on the other hand, the light is collected by a set of lenses (one per laser) and sent, through a bundle of optical fibers, to the spectrometer. Both in the LACOMORE and SHREDDERSORT applications, the emission lines of the main elements of interest span all the spectrum from near ultraviolet (UV) to near infrared (NIR); therefore the use of a broadband spectrometer is necessary. To guarantee the robustness of the equipment and reduce its costs, the spectral resolution of the spectrometer is not particularly high (0.1 nm between 190 and 450 nm, 0.3 nm from 450 to 900 nm) without intensification, but it is possible to trigger the spectral acquisition at a given delay after the laser pulse to optimize the signal-to-noise ratio (26). 
	
	

Representativeness of the Signal and Its Enhancement


	
	The maximum repetition rate of the laser used for both the SHREDDERSORT and LACOMORE applications is 10 Hz. While in the LACOMORE case the steel slab or billet moves at the typical speeds of a few meters per minute, which allows a very accurate sampling of the steel composition with a spatial resolution of a few hundreds of micrometers (higher than the one allowed by the dimension of the laser spot on the sample), the analysis and sorting of automotive scraps in SHREDDERSORT must necessarily be based on the result of a single LIBS spectrum. In fact, the distance traveled on the conveyor belt by a scrap between two consecutive measurements would be 20 cm (assuming a speed of the conveyor belt of 2 m/s) to be compared with the average size of the scraps, which is on the order of 10 cm. The problem of the representativeness of the LIBS signal, therefore, is a major concern, especially in view of the presence of dirt or coating at the surface of the scraps. Measurements performed at our laboratory demonstrated that in some cases the LIBS signal does not stabilize before 50â60 shots (in the same spot). This corresponds to a depth of the coating layer of about 100 mm. Moreover, also the signal fluctuations can be very high (see Figure 2).

	For enhancing the signal and background ratio and, at the same time, increasing the depth of analysis of a single laser measurement, the lasers used in the SHREDDERSORT application are operated in double-pulse mode (27). The use of a collinear double-pulse configuration offers a number of advantages in terms of signal quality and stability; the comparison between two spectra, acquired in single- and double-pulse mode, is shown in Figure 3.

	Typical signal enhancements of a factor of three for aluminum and five for magnesium, are obtained in double-pulse configuration using the optimum delay between first and second laser pulse, which is around 2 Âµs, as shown in Figure 4 for magnesium.

	At delays longer than the optimum, the signal enhancement remains important up to 30 Âµs before decreasing slowly up to reach the level of 0 delay (single pulse) at an interpulse delay around 75 Âµs.

	The signal enhancement and stabilization given by the use of the double-pulse LIBS setup suggested its use also in the LACOMORE application. Also in the case of steel, the signal enhancement is maximum at an interpulse delay of about 1â2 Âµs (28). At these interpulse delays, the movement of the sample is minimal between the first and second pulse, even in the SHREDDERSORT application, where the conveyor belt moves at a speed of 2 m/s. It has been demonstrated that the double-pulse analysis, in these conditions, is as effective as in the case of a steady sample (29).

	An additional benefit of the double-pulse strategy is that the first laser pulse has the effect of cleaning the surface that is analyzed by the second laser pulse, and this is particularly useful in the presence of surface dirt (SHREDDERSORT) or oxide scales (LACOMORE).

Quantitative Analysis of the LIBS Spectra


	
	The number of LIBS spectra to be analyzed in a typical LIBS application to the on-line analysis of metallic samples is huge. In 1 h of measurements, 36,000 spectra-each typically containing information about more than 5000 wavelengths-are accumulated, at a laser repetition rate of 10 Hz.

	Moreover, the LIBS spectra are, by their nature, extremely variable in response to changes in the physical parameters of the plasma-that is, ablated mass, electron number density, and electron temperature-which in an industrial environment reflect the unavoidable variations in the laser energy delivered to the sample because of laser energy fluctuations or changes in the laser focusing on the surface. The use of analytical procedures that are at the same time precise, robust, and fast is, thus, of paramount importance.

	A method that fulfills all of these requirements is the LIBS spectral analysis based on the use of artificial neural networks (ANN) (30â32). This method has been proposed recently in several research papers (33â38) because of the quickness and robustness of the algorithms and the availability of specialized commercial and open-source software. The idea at the basis of most of the ANN applications to the analysis of LIBS spectra is, in fact, the construction of a nonlinear combination of a given number of inputs (LIBS spectral intensities), optimized for obtaining the best representation of a set of outputs (elemental concentrations) on a number of known samples (training). The coefficients of this multivariate, nonlinear combination, after a proper validation and testing, can be used for calculating the desired outputs from a set of inputs obtained on unknown samples (prediction).

	In the LACOMORE application, the use of ANN has allowed the precise reconstruction of the variation of the concentration of some key elements in the transition layer between two consecutive grades (see Figure 6).

	The ANN method allows the statistical analysis of the LIBS spectra one by one, providing the additional information that the steel composition in the transition layer is highly inhomogeneous. This suggests the possibility that the mixing of the two grades, in the transition layer, is not perfect. A model of the mixing process should thus take into account of this effect.

	In the SHREDDERSORT application, on the other hand, the determination of the precise composition of the scraps is not very important, while it is important to perform a quick sorting of the nonferrous portion of the scraps (about 8% of the total weight in the automotive shredder) separating, for example, wrought from cast aluminum. It is worth noting that none of the current sorting technologies are able to successfully sort the light fraction of the metals (Al and Mg) into individual alloys, which consequently must be downgraded to produce cast aluminum.

	Figure 7 shows the results of the sorting of 21 aluminum scraps of known characteristics (wrought or cast) on the basis of the LIBSâANN analysis. The experimental discriminant between cast and wrought alloys is the concentration of silicon (wrought alloys should contain silicon at a considerably lower concentration). 

	The samples analyzed in Figure 7 are real scraps (dirty and sometimes coated); therefore, some of the wrought aluminum samples are misclassified as cast. The samples of wrought alloys misclassified as cast are not particularly problematic, since this misclassification only affects the efficiency of the process, without compromising the quality of the recovered wrought material. In fact, the presence of a surface contamination in general would increase the concentration of silicon, which is also present in some dirt, at the sample surface. Consequently, it is highly improbable that a cast alloy would be classified as wrought.

	In this column installment, we have shown that LIBS can be successfully applied to the monitoring of processes in the metal industry. The cases of steel production and automotive scrap sorting and recovery were presented. The measurement strategies can be different according to the specific situation; however, the recurring need for remote analysis of moving targets favors the use of methods that improve the signal and background ratio in the LIBS spectra and for fast and robust analytical techniques capable of determining the sample composition on the basis of a single LIBS spectrum. These requirements strongly suggest the use of a collinear double-pulse LIBS configuration and ANN analysis of the obtained spectra. The results presented demonstrate that this approach is appropriate for the scope of the analysis.

	The research reported in this column was partially funded by the European Commission through the LACOMORE (laser-based continuous monitoring and resolution of steel grades in sequence casting machines) project in the framework of the RFCS Research Fund for Coal and Steel and the SHREDDERSORT (selective recovery of non-ferrous metal automotive shredder by combined electromagnetic tensor spectroscopy and laser-induced plasma spectroscopy) project in the 7th Framework Programme.

	(1) A. Miziolek, V. Palleschi, and I. Schechter, 

y (Cambridge University Press, Cambridge, UK, 2006).
	(2) R. Noll, H. Bette, A. Brysch, M. Kraushaar, I. MÃ¶nch, L. Peter, and V. Sturm, 

(6), 637â649 (2001).
	(3) J. Gurell, A. Bengtson, M. FalkenstrÃ¶m, and B.A.M. Hansson, 

 46â50 (2012).
	(4) P. Werheit, C. Fricke-Begemann, M. Gesing, and R. Noll, 

(11), 2166â2174 (2011).
	(5) Ã. Aydin, J. Makowe, and R. Noll, 

 401â412, 471 (2004).
	(6) M. Casini, M.A. Harith, V. Palleschi, A. Salvetti, D.P. Singh, and M. Vaselli, 

(2), 633â639 (1991).
	(7) R.T. Wainner, R.S. Harmon, A.W. Miziolek, K.L. McNesby, and P.D. French, 

(6), 777â793 (2001).
	(8) S. Pandhija, N.K. Rai, A.K. Rai, and S.N. Thakur, 

(1), 231â241 (2010).
	(9) V.K. Singh and A.K. Rai, 

(5), 673â687 (2011).
	(10) V. Motto-Ros, L. Sancey, X.C. Wang, Q.L. Ma, F. Lux, X.S. Bai, G. Panczer, O. Tillement, and J. Yu, 

 168â174 (2013).
	(11) M.B. Rosenberg and C.R. Dockery, 

(11), 1238â1241 (2008).
	(12) M. Tofanelli, L. Pardini, M. Borrini, F. Bartoli, A. Bacci, A. D’Ulivo, E. Pitzalis, M.C. Mascherpa, S. Legnaioli, G. Lorenzetti, S. Pagnotta, G.H. Cavalcanti, M. Lezzerini, and V. Palleschi, 

 70â75 (2014).
	(13) E.M. Rodriguez-Celis, I.B. Gornushkin, U.M. Heitmann, J.R. Almirall, B.W. Smith, J.D. Winefordner, and N. Omenetto, 

(5), 1961â1968 (2008).
	(14) D. Anglos and V. Detalle, Springer Series in 

531â554 (2014). 
	(15) A. Nevin, G. Spoto, and D. Anglos, 

(2), 339â361 (2012).
	(16) M. Castillejo, M. MartÃ­n, D. Silva, T. Stratoudaki, D. Anglos, L. Burgio, and R.J.H. Clark, 

 191â198 (2000).
	(17) M. Ferretti, G. Cristoforetti, S. Legnaioli, V. Palleschi, A. Salvetti, E. Tognoni, E. Console, and P. Palaia, 

(12), 1512â1518 (2007). 
	(18) A. Giakoumaki, K. Melessanaki, and D. Anglos, 

(3), 749â760 (2007).
	(19) E.H. Evans, J.A. Day, W.J. Price, C.M.M. Smith, K. Sutton, and J.F. Tyson, 

(7), 808â833 (2003).
	(20) D.W. Hahn and N. Omenetto, 

 347â419 (2012).
	(21) D. Bulajic, G. Cristoforetti, M. Corsi, M.S. Hidalgo, S. Legnaioli, 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, 

(7),1181â1192 (2002).
	(22) S. Legnaioli, G. Lorenzetti, L. Pardini, V. Palleschi, D.M.D. Pace, F.A. Garcia, R. Grassi, F. Sorrentino, G. Carelli, M. Francesconi, F. Francesconi, and R. Borgogni, 

 123â126 (2012).
	(23) M. Corsi, V. Palleschi, A. Salvetti, and E. Tognoni, 

(1), 69â79 (2002).
	(24) J. Campbell, 

Complete Casting H, and Book: Metal Casting Processes, Techniques, and Design

 (Elsevier Butterworth-Heinemann, UK, 2011).
	(25) I. Gaona, J. Moros, and J.J. Laserna, 

(11), 1750â1759 (2013).
	(26) E. Grifoni, S. Legnaioli, M. Lezzerini, G. Lorenzetti, S. Pagnotta, and V. Palleschi, 

doi:10.1155/2014/849310 (2014).
	(27) V.I. Babushok, F.C. DeLucia Jr., J.L. Gottfried, C.A. Munson, and A.W. Miziolek, 

(9), 999â1014 (2006).
	(28) S. Legnaioli, G. Lorenzetti, L. Pardini, G.H. Cavalcanti, and V. Palleschi, 

 117â141 (2014).
	(29) G. Cristoforetti, S. Legnaioli, V. Palleschi, E. Tognoni, and P.A. Benedetti, 

(1), 219â225 (2010).
	(30) S.S. Haykin, 

Neural Networks: A Comprehensive Foundation, Second Edition

 (Prentice-Hall, New Jersey, 1999).
	(31) L.I. Kuncheva, 

Combining Pattern Classifiers: Methods, and Algorithms

 (Wiley Interscience, New Jersey, 2004).
	(32) R. Mikut, and M. Reischl, 

Data Mining Tools, WIREs Data Mining, and Knowledge Discovery 

(John Wiley & Sons, Inc., 2011).
	(33) J.O. Caceres, S. Moncayo, J.D. Rosales, F.J.M. De Villena, F.C. Alvira, and G.M. Bilmes, 

(9),1064â1072 (2013).
	(34) J. El Haddad, M. Villot-Kadri, A. IsmaÃ«l, G. Gallou, K. Michel, D. BruyÃ¨re, V. Laperche, L. Canioni, and B. Bousquet, 

 51â57 (2013).
	(35) J. El Haddad, D. BruyÃ¨re, A. IsmaÃ«l, G. Gallou, V. Laperche, K. Michel, L. Canioni, and B. Bousquet, 

(1), 57â64 (2014).
	(36) G. VÃ­tkovÃ¡, K. NovotnÃ½, L. ProkeÅ¡, A. Hrdlicka, J. Kaiser, J. NovotnÃ½, R. Malina, and D. Prochazka, 

 has a Master degree in Science for Cultural Heritage from the University of Padoa, with a thesis about diagnostic analysis on Beato Angelico’s painting. Since February 2013, she has been a PhD student in the Science of Cultural Heritage Conservation at the University of Florence, in collaboration with ICCOM-CNR in Pisa. Her research interests include XRF spectroscopy, multispectral imaging, and other techniques applied to Cultural Heritage.

 is with the Applied and Laser Spectroscopy Laboratory (ICCOM-

CNR) in Pisa, Italy and the National Interuniversity Consortium of Materials Science and Technology (INSTM). He graduated in Physics with distinction in 1984 at Pisa University. Since September 1988 he has worked as a researcher at Italian National Research Council (CNR) in Pisa, at Institute of Atomic and Molecular Physics, then at Institute for Chemical-Physical Processes and since September 2010 at Institute of Chemistry of Organometallic Compounds. Currently, he is responsible of the activities of the applied and laser spectroscopy laboratory. He has published more than 160 papers in the field of laser spectroscopy that received more than 3700 citations (H-index 34) and authored the PCT/EP Patent #99/01842. He is chief editor of the 

Journal of Applied and Laser Spectroscopy

 and member of the editorial boards of the journals 

 He is a member of the Scientific Committee of the Center for Ancient Mediterranean and Near Eastern Studies.

 has a Masters degree and a PhD in Science for Cultural Heritage from the University of Florence, with a thesis on confocal Raman spectroscopy. Since 2012 she has been a research fellow at ICCOM-CNR. Her research includes plasma (LIBS) and Raman spectroscopy techniques applied to cultural heritage. She is also contributing to the development of an algorithm for the automatic element detection in LIBS spectra.

 has a bachelor degree in Conservation of Cultural Heritage from Suor Orsola Benincasa University and a master degree in archaeology from Pisa University. Since 2015, he has had a scholarship at the Institute of chemistry of organometallic compounds ICCOM-CNR Research Area of Pisa, in spectroscopic methods for the analysis of matrix materials inhomogeneous. His main interests are applied laser spectroscopy, X-ray fluorescence, multispectral imaging and colorimetry, archaeometry and pottery analysis.

 graduated in physics and obtained a PhD in the Science of Cultural Heritage Conservation (chemical sciences) at the University of Florence. In 2011, he received a Master degree in Innovation and Knowledge Transfer (MIT) at the School of Management MIP, Politecnico of Milano. He was enrolled as Researcher at the Institute of Chemistry of Organo-Metallic Compound (ICCOM-CNR) in Pisa, working in several research activities of the applied laser spectroscopy group, particularly in the field of spectroscopic techniques applied to cultural heritage (pigment analysis, analysis of ceramics, archeo-metallurgy, and more), environmental, bio-medicine, and material characterization such as monitoring of industrial processes.

	In the version of this article that appeared in print, incorrect units were specified in several sentences on page 24. The corrected paragraphs are shown below: 

The monitoring of the processes in steel and metal industry calls for techniques that are capable of measuring the composition of metallic alloys at a distance and on moving conveyor belts. In many cases, such as in the recycling of automotive scrap, the geometry of the objects to be analyzed can vary, and surface coatings can be present. In this paper, we discuss the application of LIBS to two industrial projects to illustrate how the above-mentioned problems can be faced and successfully resolved.

Vincenzo Palleschi

Articles

Avoiding Misunderstanding Self-Absorption in Laser-Induced Breakdown Spectroscopy (LIBS) Analysis

Seven Common Errors to Avoid in LIBS Analysis

Applying LIBS to Metals Processing