Study to reduce laser-induced breakdown spectroscopy measurement uncertainty using plasma characteristic parameters

Using standard brass alloy samples, an approach to reduce the laser-induced breakdown spectroscopy measurement uncertainty was tested and proved. Two important parameters for plasma characterization, the plasma temperature and the electron density, were applied to minimize the signal uncertainties due to uncontrollable experimental parameter variations. Results show that for the pulse-to-pulse analysis, the signal fluctuations can be significantly reduced by utilizing the plasma characteristic information. The major source for the single pulse fluctuations is the redistribution of the characteristic line at different temperatures according to the Boltzmann distribution under LTE. The change of the degree of ionization also contributes to the signal fluctuations. For the multi-pulse analysis, due to the nonlinear relationship between the plasma temperature and the line intensity, it is not applicable to utilize the Boltzmann distribution to reduce the influences of the plasma properties. However, normalization with the combination of the whole spectrum area and the ratio between the ion and atom number density of the same element can further increase the measurement accuracy.     

Line broadening mechanisms in the low pressure laser-induced plasma

The spectral profiles of Ca and Rb lines have been studied in a laser induced plasma as a function of pressure (1–10 torr) and delay time with respect to the plasma initiation (1–10 μs). Measurements were made in a plasma induced by the 1064-nm output of a Nd:YAG laser on a calcium carbonate matrix, doped with Rb. Spectral profiles were measured in absorption using a narrow-band cw Ti:Sapphire laser. It was shown that in the case of a trace element (Rb in a CaCO₃ matrix), the broadening mechanism was Doppler-dominant, whereas for a major matrix component (Ca), resonance broadening was the main contributor to the line shape. The plasma was shown to be non-equilibrium provided by the difference between the kinetic (3000 K) and the excitation (8000 K) temperatures. The electron number density at delay times of 5–10 μs and pressures of 1–10 torr was estimated not to exceed 10¹⁵ cm⁻³. The number densities of Ca atoms in the ground and the excited (23 652 cm⁻¹) states were evaluated by measuring line width and peak absorption at 732.6 nm. They were found to be in the range of (1.5–2.2)×10¹⁷ cm⁻³ for the ground state and (1.5–33)×10¹¹ cm⁻³ for the excited state.     

Application of calibration-free laser-induced breakdown spectroscopy to radially resolved spectra from a copper-based alloy laser-induced plasma

In this work, the Calibration-Free approach for Laser-Induced Breakdown Spectroscopy (CF-LIBS) was applied for the first time to radially resolved spectra emitted by a laser-induced plasma. The radial profiles of plasma temperature and electron number density were used to calculate the local relative concentration of the elements of interest. We analyzed a set of profiles of the local spectral emission coefficient obtained previously by means of spatial deconvolution of the spectra from a copper-based alloy (Cu 93, Fe 5, Mn 1, Ni 1 wt.%) laser-induced plasma. A spatially integrated spectrum of the same plasma was also analyzed for comparison purpose. The relative abundance of the minor components Fe, Mn and Ni was calculated. The results obtained from the central region of the plasma were closer to the nominal concentrations than those obtained from the spatially integrated spectrum. However, an increasing deviation was observed towards the plasma edge. It is proposed that this deviation could be the result of a gradual departure from Local Thermodynamic Equilibrium (LTE) conditions due the significant decrease of the electron density at the external shells of the plasma.     

Radiative models of laser-induced plasma and pump-probe diagnostics relevant to laser-induced breakdown spectroscopy

The paper describes past and present efforts in modeling of laser-induced plasma and overviews plasma diagnostics carried out by pump-probe techniques. Besides general information on existing plasma models, the emphasis is given to models relevant to spectrochemical analysis, i.e. models of radiating plasma. Special attention is paid to collisional-radiative (CR) and collisional-dominated (CD) plasma models where radiative processes play an important role. Also, calibration-free (CF) models are considered which may endow with the possibility for standardless spectroscopic analysis. In the diagnostic part, only methods based on the use of additional diagnostic tools (auxiliary lasers, optics, and probes) are described omitting those based on plasma own radiation. A short review is provided on image-based diagnostics (shadowgraphy, schlieren, and interferometry), absorption and fluorescence, Langmuir probe, and less frequently used cavity ringdown and Thomson scattering methods.     

Curve of growth methodology applied to laser-induced plasma emission spectroscopy

The curve-of-growth (COG) method was applied to a laser-induced plasma. The plasma was produced by a Nd:YAG laser on the surface of steel samples containing 0.007–1.3% of Cr. The emission was collected from the top of the plasma by means of a 45° pierced mirror and aligned onto an intensified charge-coupled device (ICCD) with a gate width of 1 μs and a variable delay time. The resonance 425.4 nm Cr line was used for construction of the COG. The temperature of the plasma (∼8000 K at 5-μs delay) was determined from a Boltzmann plot. The damping constant a, proportional to the ratio of the Lorentzian to the Doppler line widths, was found from the best fit of a series of calculated COG to the experimental data points and was 0.20±0.05. The number density of neutral Cr atoms which corresponded to the transition between low and high optical densities, was ≈6.5·10¹² cm⁻³. The cross-section for broadening collisions of Cr atoms with atmospheric species (presumably N₂) was calculated to be (66±16) Å. The shape of the 425.4-nm Cr line was additionally checked by scanning an ultra-narrow cw Ti:Sapphire laser across the atomic transition and found to be in agreement with preliminary estimates. The potential of the COG method for laser breakdown spectroscopy is discussed.     

Study of early laser-induced plasma dynamics: Transient electron density gradients via Thomson scattering and Stark Broadening,and the implications on laser-induced breakdown spectroscopy measurements

To further develop laser-induced breakdown spectroscopy (LIBS) as an analytical technique, it is necessary to better understand the fundamental processes and mechanisms taking place during the plasma evolution. This paper addresses the very early plasma dynamics (first 100 ns) using direct plasma imaging, light scattering, and transmission measurements from a synchronized 532-nm probe laser pulse. During the first 50 ns following breakdown, significant Thomson scattering was observed while the probe laser interacted with the laser-induced plasma. The Thomson scattering was observed to peak 15–25 ns following plasma initiation and then decay rapidly, thereby revealing the highly transient nature of the free electron density and plasma equilibrium immediately following breakdown. Such an intense free electron density gradient is suggestive of a non-equilibrium, free electron wave generated by the initial breakdown and growth processes. Additional probe beam transmission measurements and electron density measurements via Stark broadening of the 500.1-nm nitrogen ion line corroborate the Thomson scattering observations. In concert, the data support the finding of a highly transient plasma that deviates from local thermodynamic equilibrium (LTE) conditions during the first tens of nanoseconds of plasma lifetime. The implications of this early plasma transient behavior are discussed in the context of plasma–analyte interactions and the role on LIBS measurements.     

Analysis of paper by laser-induced plasma spectroscopy

Material distributions in paper and paper coatings are important factors which determine the characteristics and quality of paper. We have used laser-induced plasma spectroscopy (LIPS) to determine various properties of coated and uncoated papers. Depth profiles of double coated papers were determined. Microscopic variations of pigments in a double-coated paper were analyzed. It was observed, that the two coating layers can be distinguished, if they differ enough in composition. Filler distributions of uncoated copier papers were determined. The surface of the paper was studied for five samples containing different amounts of binder in the coating. Lateral contamination in paper analysis i.e. splash over effect of ablation was studied by varying the sampling distance and number of laser pulses used for ablation. Influence of splash over effect on lateral analysis is briefly discussed.     

Experimental determination of the Stark widths of Pb I spectral lines in a laser-induced plasma

Stark widths of 34 spectral lines of Pb I have been measured in a Laser-Induced-Plasma (LIP). The optical emission spectroscopy from a LIP generated by a 10 640 Å radiation, with an irradiance of 1.4 × 10¹⁰ W cm^− 2 on a Sn–Pb target in an atmosphere of argon was analyzed between 1900 and 7000 Å. The Local Thermodynamic Equilibrium (LTE) conditions and plasma homogeneity have been checked. The 34 spectral lines measured in this paper correspond to the transitions n(n = 7, 8)s→6p², n(n = 6, 7)d→6p². The population levels distribution and the corresponding temperatures were obtained using Boltzmann plots. The plasma electron densities were determined using well-known Stark broadening parameters of spectral lines. Special attention was dedicated to the possible self-absorption of the different transitions. Stark broadening parameters of the spectral lines were measured at 2.5 µs after each laser light pulse, where the electron temperature was close to 11 200 K and the electron density to 10¹⁶ cm^− 3. The experimental results obtained have been compared with the experimental values given by other authors.     

The first measured Ag I, Ag II and Ag III Stark broadening parameters

The Stark widths (W) and shifts (d) of two neutral (520.908 and 546.550 nm), eleven singly (211.382, 224.643, 224.874, 232.029, 232.468, 233.140, 241.141, 241.323, 243.781, 244.793 and 276.754 nm) and three doubly (216.189, 231.004 and 239.569 nm) ionized silver (Ag I, Ag II and Ag III, respectively) spectral lines have been measured in nitrogen plasma at about 18,000 K electron temperature and electron density ranged between 0.65 × 10²³ and 1.15 × 10²³ m^− 3. They are the first measured W and d values while those of the Ag II and Ag III lines are the first published data in these spectra. The modified version of the linear, low-pressure, pulsed arc was used as a plasma source operated in nitrogen with silver atoms, as impurities, evaporated from silver cylindrical plates located in the homogeneous part of the discharge. No theoretical predictions exist for W and d values of above mentioned spectral lines. Besides, we have checked the transition probability ratio of two investigated Ag I lines. An agreement with theoretical predictions was found.     

On the Stark broadening in the Au I and Au II spectra from a helium plasma

The Stark FWHM (Full-Width at Half of the Maximal line intensity, W) of 5 neutral and 26 singly ionized gold (Au I and Au II, respectively) spectral lines have been measured in laboratory helium plasma at approximately 16,600 K electron temperature and 7.4 × 10²² m^− 3 electron density. Five Au I and ten Au II W values are reported for the first time. The Au II W values are compared with recent theoretical data, calculated based on a modified semi-empirical approach, and also with existing experimental W values. Our normalized Stark widths are six times higher than those measured in a laser-produced plasma. Possible explanation of this is recommended here. An agreement (within the accuracy of the experiment and uncertainties of the theoretical approach used) with the recently calculated W data was found in the 6p–7s Au II transition. The calculated hyperfine splitting for the five Au II lines in the 6s–6p transition is also presented. At the stated helium plasma conditions, Stark broadening has been found to be the dominant mechanism in the Au I and Au II line shape formation. A modified version of the linear low-pressure pulsed arc was used as a plasma source operated in helium, with gold atoms as impurities evaporated from the thin gold cylindrical plates located in the homogeneous part of the discharge, providing conditions free of self-absorption. This plasma source ensures good conditions for generation of excited gold ions due to Penning and charge exchange effects.     

Asymmetric Stark broadening of the Fe I 538.34 nm emission line in a laser induced plasma

In this work asymmetries along with shifts in the line profiles of neutral iron emission lines coming from a laser induced plasma have been detected. The plasma was produced in air at atmospheric pressure on a 50% Fe–Ni alloy and the emission was collected at a temporal window of (2.5, 3) μs. To avoid the effect of spatial inhomogeneity on the profiles, a deconvolution procedure was applied to obtain the spatially resolved emissivity. Asymmetric theoretical Stark profiles, which take into account the effect of static ions, were used to be fitted to the experimental data of the emission profile of the line Fe I 538.34 nm. The fitting of the theoretical profile to the experimental data was carried out by means of the least squares method using genetic algorithms to automatically solve the optimization problem. The correlation coefficient was higher for the asymmetric fits than for the symmetric ones. From the fit, the quasistatic ion broadening parameter α, the electron broadening parameter w_e, and the total shift of the maximum of the line d_t, were obtained. The ion parameter α varied in a range (0.2–0.3) for an electron density between (4–15) × 10¹⁶ cm^− 3. The ion influence on the total broadening was of 15–20%. The total shift varied in the range (0.01–0.06) nm and it was mainly given by the ion shift, the electron shift being negligible. For the electron density range in this work, approximated linear behaviors of the total width and shift with electron density have been obtained.     

The Stark broadening of the He I 667.8 nm line

The half-width and asymmetry parameter of the He I 2¹ P–3¹ D λ=667.8 nm line have been measured in a laser produced He plasma for electron densitiesN _e between 7×10²² m^?3 and 1.7×10²³ m^?3. Results, obtained for the half-width, agree very well with the theoretical results forN _e ?10²³ m^?3 and the agreement becomes somewhat worse forN _e >10²³ m^?3. We have measured the red asymmetry which is considerable lower than the theoretical one. We have also established good agreement with theoreticalj _A,R (x) profile for this line.     

Stark shift and broadening of FI and ClI lines

We report measured Stark shifts and widths of neutral flourine and chlorine lines. Wall stabilized arc is used as a plasma source. Electron densities 2–4×10²² m^?3 are determined from the width of theH _β line and electron temperatures 9500–10 000 K from plasma composition data. Experimental results for FI and ClI Stark widths and FI Stark shifts agree within 10% with semiclassical calculations. ClI Stark shifts are systematically smaller for about 20% than theoretical data with the only exception of the line from multiplet no. 15 where the discrepancy goes up to 49%. Results of investigation of similarities and regularities of Stark widths are in agreement with the study of Wiese and Konjevi?. Comparison of experimental Stark shifts shows certain types of regularities.     

Investigation of optimized experimental parameters including laser wavelength for boron measurement in photovoltaic grade silicon using laser-induced breakdown spectroscopy

The quantification of boron and other impurities in photovoltaic grade silicon was investigated using the LIBS technique with attention to the laser wavelength employed, temporal parameters, and the nature of the ambient gas. The laser wavelength was found to have a moderate effect on the performance of the process, while the type of purge gas and temporal parameters had a strong effect on the signal-to-background ratio (SBR) of the boron spectral emission, which was used to determine the boron concentration in silicon. The three parameters are not independent, meaning that for each different purge gas, different optimal temporal parameters are observed. Electron density was also calculated from Stark broadening of the 390.5 nm silicon emission line in order to better understand the different performances observed when using different gases and gating parameters. Calibration curves were made for boron measurement in silicon using certified standards with different purge gases while using the temporal parameters which had been optimized for that gas. By comparing the calibration curves, it was determined that argon is superior to helium or air for use as the analysis chamber purge gas with an UV laser.     

Main parameters influencing the double-pulse laser-induced breakdown spectroscopy in the collinear beam geometry

Two Nd:YAG lasers emitting at 532 nm were combined in the same direction (collinear beam geometry) for double-pulse laser-induced breakdown spectroscopy studies on aluminum samples at atmospheric pressure in air. The influence of the delay between the two laser pulses was investigated for the background emission, for lines detected in aluminum samples and for atmospheric lines with different detection systems (photomultiplier tube, Czerny–Turner spectrometer and echelle spectrometers). The optimization of the delay between the two laser pulses depended on the excitation energy levels of the emission lines: two optima of interpulse delays were observed in the collinear geometry. Different regimes of laser–plasma interactions were discussed depending on the interpulse delay for aluminum samples. Furthermore, the effect of the sampling geometry, in terms of lens-to-sample distance, focal length of the focusing lens, was studied to determine parameters influencing the single- and double-pulse scheme. Besides, the analytical performance of the system was evaluated to characterize the advantages of the double-pulse laser-induced breakdown spectroscopy in terms of improvement of sensitivity and reduction of self-absorption effect for aluminum samples.     

Analysis of heavy metal aerosols on filters by laser-induced plasma spectroscopy

Particulate heavy metals can lead to severe toxic and carcinogenic effects in humans when inhaled in higher concentrations. For the development of a quasi-continuous emission monitor based on automatic filter sampling on a filter band, laser-induced plasma spectroscopy (LIPS) was studied for analysis of heavy metal aerosols on quartz fiber filters. The system consists of a 19-inch laser and detector module connected to a miniaturized sensor head through fiber optics, allowing maximum flexibility of the set-up. Parameters for optimum time-resolved analysis, i.e. detection wavelength, timing and filter material, were established. The LIPS investigations were accompanied by a rigorous reference analysis based on total reflection X-ray fluorescence (TXRF) analysis. The detection limits for heavy metals (Cd, Ni, As, Co, Mn, Sb, Cr, Tl, Sn, V, Cu and Pb) on filters varied between 0.01 and approximately 0.91 μg cm⁻², corresponding to volume detection limits of 0.02–2.73 μg m⁻³. Analysis of filter samples from waste incineration demonstrated the potential of the LIPS approach. In combination with an echelle spectrometer, ambient samples from environmental monitoring could be characterized in much better detail, due to the improved detection limits and the superior spectral resolution, and spectral range of the echelle.     

Time-resolved plasma properties for double pulsed laser-induced breakdown spectroscopy of silicon

A systematic measurement of plasma properties (temperature, electron number density, pressure) was performed during LIBS of silicon with two nanosecond pulsed lasers operating at 1064 nm. The spectral characteristics of the plasmas were measured to determine the plasma properties as delay time between the laser pulses was changed from 0 to 10 ms. The plasma properties and crater dimensions increased abruptly from 100 to 200 ns. The crater depth increased from 2 to 10 μm (volume increased about 5 times) per pair of double pulses. Enhanced mass removal was indicative of a phase explosion mechanism. Spatial images of plasma emission were measured to study the dynamics of plasma expansion.     

Stark broadening and shift of multiply ionized carbon spectral lines

Stark widths and shifts of sixCII, threeCIII and twoCIV spectral lines have been measured in a linear pinch discharge plasma and compared with available experimental and theoretical data. Electron density, (0.86?1.64)×10²³ m^?3, was determined by single wavelength laser interferometry using the visible 632.8 nm transition of He-Ne laser. The electron temperature of 38000 K was derived from the Boltzmann slope of severalCII spectral lines, and ratios of severalCII toCIII spectral lines. The stark widths (w) dependence on:(i) the upper-level ionization potential (I) of corresponding lines;(ii) net charge (z) of the emitter core seen by the optical electron undergoing transition, and(iii) electron temperature (T) was found to be of the form:w=az ² T ^?1/2 I ^?b . However, it should be noticed that the essential role in the obtained trends belongs to the energy of the emitter core. The established overall trend is used to predict Stark widths of uninvestigated spectral lines originating from the given transition arrays.     

Correlation between properties of a solid sample and laser-induced plasma parameters

To investigate the influence of mechanical properties of a solid sample on the laser-induced plasma parameters (temperature and mass ablated) a number of aluminum–lithium alloys and lithium ferrites with different microstructure and composition have been studied. The specific approach to estimate excitation temperature for low-resolution emission spectra has been developed. The main limitations of this approach were discussed on the basis of comparison with the energy width of several multiplets. Overall uncertainties for temperature calculation were evaluated by taking into account the accuracy of Einstein's coefficients and errors of the proposed multiplets method. The temporal evolution of laser plasma during the evaporation of these materials was studied. Extremely high value of the Li I excitation temperature has been estimated to be T > 10⁵ K for the annealed ferrite ablation, in contrast to the temperature T  1.5 · 10⁴ K for non-annealed ferrite. Only for ablation of annealed ferrite the Li II emission line at 548.4 nm was observed. It means that this laser-induced plasma was the hottest. In the case of alloys, the temperature calculated by using Li I transitions was proportional to the microhardness of the solid samples. The negative correlation between crater volume/opto-acoustic signal and alloy microhardness was observed. At the first pulses the mechanical properties of the alloys didn't correlate with the ablated mass, while the maximal correlation coefficients were observed after ablation by 10 or 50 consequent laser pulses.     

Stark broadening of Mg I and Mg II spectral lines and Debye shielding effect in laser induced plasma

We report Stark broadening parameters for three Mg I lines and one Mg II line in the electron number density range (0.67–1.09) · 10¹⁷ cm^− 3 and electron temperature interval (6200–6500) K. The electron density is determined from the half width of hydrogen impurity line, the H_α, while the electron temperature is measured from relative intensities of Mg I or Al II lines using Boltzmann plot technique. The plasma source was induced by Nd:YAG laser radiation at 1.06 μm having pulse width 15 ns and pulse energy 50 mJ. Laser induced plasma is generated in front of a solid state surface. High speed photography is used to determine time of plasma decay with good homogeneity and then applied line self-absorption test and Abel inversion procedure. The details of data acquisition and data processing are described and illustrated with typical examples. The experimental results are compared with two sets of semiclassical calculations and the results of this comparison for Mg I lines are not unambiguous while for Mg II 448.1 nm line, the results of Dimitrijević and Sahal-Bréchot calculations agree well with our and other experimental results in the temperature range (5000–12,000) K and these theoretical results are recommended for plasma diagnostic purposes. The study of line shapes within Mg I 383.53 nm multiplet shows that the use of Debye shielding correction improves the agreement between theoretical and experimental Stark broadening parameters.