Open Access
Add to BibSonomy
Abstract
Laser-induced breakdown spectroscopy (LIBS) is carried out with compact 1064 nm laser and spectrometer components which are suitable for handheld applications. Bursts of ∼0.6 mJ, 5 ns laser pulses are generated by a passively Q-switched laser with a 1 kHz triggered pump diode. The miniature spectrometer with a set wavelength range of ∼188-251 nm has an instrumental broadening at the carbon analyte line, C I 193.09 nm, of less than 36 pm. Analytical calibration curves of C, as well as Cr, Ni, and Si are taken with certified reference samples of iron and steel in an argon purged setup. The net duration of the laser bursts is ∼0.7-1.4 s for a measurement, depending on the number of repetitions on the sample surface. The limit of detection (LOD) is determined to a mass fraction of 34 µg/g for C. High-alloy steels 1.4306 (0.01% C) and 1.4541 (0.035% C) are separated clearly by the LIBS measurement of carbon.
1. Introduction
Laser-induced breakdown spectroscopy uses atomic emission spectrometry (AES) to determine the elemental composition of solids, liquids or gases. A focused laser beam induces a luminous plasma of a small fraction of the material with excitation of neutral atoms and ions of the species to be analyzed. Similar to spark-AES, spectra of the plasma light exhibiting continuum radiation and spectral lines are recorded by a spectrometer. The spectral lines correspond to the atomic (ionic) transitions which can be assigned by tabulated data, e.g. [1]. The intensity of a spectral line is a measure of the number of the analyte species in the plasma and hence the analyte content, at least under certain conditions. The involved physical processes and the application of LIBS for a variety of materials are described in detail in the literature, e.g. [2–5].
Elemental analysis of steel is required in metallurgy and metalworking industry for quality assurance as well as material identification, e.g. to differentiate steel grades in production or for recycling. Carbon in steel varies from mass fractions of < 0.001% (or 10 µg/g) to > 2% and it plays a crucial role for properties of alloy steel and cast iron. Despite this fact, the spectrochemical determination of C at low levels is still challenging, especially for compact instruments. One reason is that the most sensitive lines are in the vacuum ultraviolet (VUV) wavelength range, which requires suitable detection systems for these wavelengths as well as vacuum or inert gas atmosphere, e.g. argon, to avoid strong absorption of the radiation. If an application allows short distances between sample and detection, the best compromise between wavelength region and sensitivity is the line C I 193.0905 nm [1], where the absorption by oxygen of ambient air is still moderate.
The preconditions for C detection can be realized quite well in stationary instruments. Equipped with high-resolution vacuum or inert gas filled spectrometer, argon flushed sample stand, laser pulse energies Q in the range of 50 mJ to several 100 mJ, and nanosecond pulse durations, limits of detection (3-sigma LOD, [6]) of < 10 µg/g are achieved by LIBS with multiple laser pulse bursts [7], see also [8,9] and Table C.2 in [3]. Lowest LOD’s are reported for VUV lines. Other, non-VUV lines at wavelengths (in air) of 247.856 nm, 833.514 nm [10], 909.483 nm, and 940.573 nm [11] are at least ten times less sensitive than the 193 nm line or they suffer from spectral interference [1]. Thus, the C I 247.856 nm line is completely interfered with the Fe II 247.857 nm line and therefore not useful for low-level C analysis of steel.
In the field of compact portable and handheld LIBS instruments, the size, weight, laser output energy, inert gas purge rate, and spectral resolution are restricted [12–24]. For handheld LIBS, calibration curves of C with mass fractions below 100 µg/g are not reported in peer-review papers according to our knowledge. One company announces in an application note on their website a LOD of 80 µg/g for their handheld instrument [25]. They report a measuring duration of 9-12 s using a 5.5 mJ, 50 Hz laser. This paper presents measurements with a representative sample set, especially including samples in the lower range near LOD. Furthermore, global calibration curves of Cr, Ni, and Si are added to assess the performance beyond C.
2. Experimental details
2.1 Experimental setup
The miniature spectrometer and the Nd:YAG laser were designed and assembled at Fraunhofer ILT with the target to be useful for handheld LIBS. Emphasis was placed on compact size and enabling low-level detection of C in steel. The outer dimensions of the non box-shaped spectrometer body are 110 × 80 × 21 mm3. The body with internal locating surfaces for fixing the mirrors, grating, and CMOS linear array sensor is a CNC machined aluminum alloy. The spectral range is fine-tuned to 187.6-251.7 nm by adjusting the 4320 grooves/mm grating. A small argon flow (nominal argon purity is 99.999 vol.-%) can be optionally switched on for enhanced VUV transmission. The 10 µm wide entrance slit of the uncrossed Czerny-Turner design is laser-cutted in-house. The 1064 nm laser has single-mode output with M2 ≈ 1.2 and the dimensions of the laser head in its final version are 142 × 40 × 20 mm3 with a resonator length of ∼70 mm. The 808-nm single-emitter pump diode is triggered with a repetition rate of 1 kHz and the pump pulses are fiber-coupled to the rear mirror of the laser with a Cr:YAG crystal-based passive Q-switch. The energy and duration of the laser pulses are Q ≈ 0.6 mJ, and τp ≈ 5 ns (FWHM), respectively. Depending on the pump parameters, one or two laser pulses per pump pulse are emitted. Bursts of laser pulses are generated by a corresponding sequence of N trigger pulses to the pump diode.
For this experimental validation, the laser focusing, plasma imaging optics to the spectrometer and the sample fixture were set-up in a compact way, too. However, during the study not all components are integrated in a handheld package. In a first phase, external supply units, such as diode driver, read-out electronics, argon gas, are used which were available in the laboratory. The working distance of the laser focusing optics is ∼27 mm. For measurement, the steel samples are manually hold in the focal plane by pushing them against a mechanical stop in front of the focusing optics. The focusing optics is followed by an argon flushed orifice outlet before the steel sample. The plasma light is directly imaged by two fused silica lenses to the entrance slit of the spectrometer. The front lens with 5 mm diameter is ∼9 mm distant from the LIBS plasma. Therefore the VUV transmission between plasma and spectrometer is not reduced by an optical fiber as in other setups, e.g. [13–16]. Furthermore, the higher grating frequency compared to 2400 or 3600 grooves/mm typically used in compact spectrometers as well as the stable, single-mode output of the laser are favorable for the C detection.
2.2 Samples
For the global calibration curve of C with mass fractions of up to 2.5% in section 3.2, a set A of 39 samples of pure iron, alloy steel, and one cast iron sample is used to cover the large mass fraction range. With exception of one setting-up sample (SUS), the samples of set A are certified reference materials (CRM) of well-known suppliers. The CRMs are accompanied by a certificate stating the mass fractions of the certified elements (not necessarily all constituents) and their estimates of uncertainty arising from interlaboratory round-robin analysis during certification. SUS are for spectrometric purposes in a routine operation, and these steels materials are selected, e.g. for homogeneity. Mostly, good quality reference values are provided for SUS by the suppliers but the values are not certified and no estimates of uncertainty are available.
Four samples of set A are used as a subset B for the instrumental broadening measurements in section 3.1. These samples are the low alloy steels 12X 15251 T of MBH Analytical Ltd (www.mbh.co.uk), SUS D/9 of the Bureau of Analysed Samples Ltd (BAS, www.basrid.co.uk, D/9 is the predecessor of the current D/11), SS-CRM 407/2 of BAS and the carbon steel SS-CRM 455/1 of BAS. Their C mass fractions are listed in the diagram legend in section 3.1.
Table 1. List of 15 CRM samples (set C) and 4 SUS samples (marked by “S”) with < 860 µg/g C and an excerpt of their compositions. For the CRMs, the mass fractions are taken from the certificates and in addition the estimates of uncertainty (“uncert.”) for C. Iron is calculated from [100%-(sum of certified mass fractions)]. For SUS, the guide values of the supplier reference sheets are listed.
For the low C measurements and the LOD determination of C in section 3.2, a higher number of samples, i.e. data points, in this range is advantageous for the significance of the measurement. Therefore low C samples from set A are added by further low C samples to a sample set C of 15 CRMs. Because of the emphasis to low C detection, these samples are listed in detail in Table 1. The set includes a pure iron sample with 5.1 µg/g as the lowest C content, approximating the matrix blank sample [6]. The two SUS samples BR 53 and Fe 130-35 in Table 1 are included in the measurements as further very low C samples. The use of the last two high-alloy SUS samples in Table 1 is explained at the end of section 3.2. The sets of samples were ground by hand with aluminum oxide abrasive (grain size 60) at paper-backed discs some days before the measurement.
3. Results and discussion
3.1 Instrumental broadening of spectrometer
The instrumental broadening, and therefore the spectral resolution, of large-size spectrometers such as used in stationary spark-AES is in the range of about 10-20 pm. This resolution is usually not achieved by miniature spectrometers. Therefore the observed spectral line widths are dominated by the instrumental broadening. The focus is on the wavelength range at 193 nm when regarding the C I 193.0905 nm analyte line. This wavelength in vacuum corresponds to 193.0268 nm in air [1], and 193.0275 nm in argon. The wavelength in argon is calculated by dividing the vacuum wavelength through the argon refractive index taken from [26]. The Doppler width of the line is ∼4 pm (calculated by Eq. T4.26 of [4] and 10 000 K) and the Stark width is ∼5.9 pm [27]. By neglecting these in a first order estimation, the measured spectral width of the C line of a LIBS plasma gives the instrumental spectral width winstr of the spectrometer, exactly at the wavelength of interest.
In Fig. 1, a Lorentzian profile of width wL is fitted to the measured C line for four steel samples. The abscissa wavelengths are in argon because it is calibrated to 23 spectral line wavelengths (in argon) of Fe, Al, C, Cr, Ni and Si excited in a LIBS plasma of different test samples. The mean central wavelength of 193.03 nm agrees well with the nominal value of 193.028 nm in argon. In a previous test, a Voigt profile was fitted by standard software [28], showing a dominant Lorentzian profile, as expected in most cases for the instrumental function [4]. The mean value and standard deviation (SD) of the four fits is wL = (36 ± 2) pm. In a second order estimation, the instrumental width at 193 nm is winstr ≈ (36-5.9) pm ≈ 31 pm, because the Stark broadening is Lorentzian and its width can be directly subtracted [4]. But as an upper limit value, the measured spectral width of 36 pm (FWHM) is taken. For a spectrometer of this size, it is regarded as a good resolution.
Fig. 1. Spectral width of measured C line of steel samples fitted by Lorentzian profiles with FWHM, wL. The C mass fractions of the four samples (set B, see section 2.2) are given in the legend.
3.2 LIBS measurements
Procedure and parameters
A typical LIBS procedure is to irradiate the sample with a sequence of a burst of pre-pulses and an immediately following burst of measuring pulses at one spot [7]. This is repeated for 3 to 10 times at different sample spots for averaging. The procedure is used in a similar way: To measure one sample spot, the argon purge in front of the laser focusing optics is switched on, and the sample is pushed against the mechanical stop of the laser focusing optics. Then, a trigger sequence to the pump diode of N = 140 pulses at a rate of 1 kHz is started by a button, generating a laser burst. The time-integrated sensor readout begins 20 ms delayed to the trigger sequence for stabilizing and surface cleaning by pre-pulses, hence the signal was integrated during 120 ms. For these measurements, the pump pulse parameters are adjusted in a way that two laser pulses per pump pulse were emitted at a time separation of ∼60 µs, therefore 240 laser pulses are integrated to one spectrum at a single sample spot within 0.14 s. The measurement is repeated at 5 or 10 spots after shifting the sample laterally (roughly ∼1 mm) at the mechanical stop. The net duration of the laser bursts is therefore 5 × 0.14 s = 0.7 s, or 10 × 0.14 s = 1.4 s, respectively. The total duration depends on how fast the laser focal spot or sample can be shifted in an automated setup. The resulting crater diameter of a laser burst is in the range of about 40 µm as measured by a microscope. Therefore the focal spot diameter 2rf is estimated to be less than 40 µm, leading to an irradiance of E = Q/(τp π rf2) > 9.6 GW/cm2, and a radiant exposure of H = E τp > 48 J/cm2 per laser pulse.
Analytical calibration curves of C, Ni, Cr, and Si
At first, a global calibration curve was taken for a large range from pure iron to > 2% C using sample set A, see section 2.2. For this, the spectra taken at five spots of a sample are numerically integrated over the respective analyte line and an internal standard line to determine these line intensity values for each sample. Numerical integration is carried out by standard software [28], or others. For normalization [5], the analyte line intensity is divided by the internal standard line intensity of Fe II 193.1845 nm (in air) for each sample spot. Then, the mean and SD of these five intensity ratios are calculated for each sample and given in Fig. 2.
Fig. 2. Global calibration curves for (a,b) C, (c,d) Ni, (e) Cr, and (f) Si. The normalized analyte line intensities are given vs. the mass fractions, or (f) the mass fraction ratios, respectively.
The evaluated analytical lines and their wavelengths (in air) are C I 193.0268 nm, Ni II 231.604 nm, Ni I 197.6246 nm, Cr II 206.1576 nm, and Si I 212.412 nm [1]. In the following, the spectral lines are denoted shortly by the element name and the wavelength divided by 0.1 nm.
The data are fitted by linear or 2nd order polynomial curves, and their equations and coefficients of determination are given in the graphs in the y(x) notation, where x stands for the mass fraction in %, and y for the intensity ratio. The units of the coefficients are omitted in these imprints for clarity. For C, the large range is given in Fig. 2(a). For better visibility of the medium range up to 0.6% C, the same data are drawn with extended axes of coordinates in Fig. 2(b). In both ranges, the calibration curves exhibit a good correlation with R2 > 0.99. To cover large ranges of mass fractions of an element, it is usual in AES to take emission lines with different sensitivity and saturation behavior for different mass fraction ranges. Therefore for Ni, the sensitive line Ni2316 is used for the lower range in Fig. 2(c). This line starts to saturate above 2% of Ni. In comparison, the less sensitive line Ni1976 in Fig. 2(d) does not work for low mass fractions of Ni, but increases linearly up to 18% Ni and is useful for analysis of high-alloy steels. Likewise, in Fig. 2(e) the linear response of the Cr2062 line is shown for high Cr mass fractions. The Si2124/Fe1932 intensity ratios correlate better to the mass fraction ratios Si/Fe, therefore these ratios are depicted in Fig. 2(f). In combination with the restricted spectral resolution, a reason is possibly that the Si2124 line is interfered stronger than the other lines. In AES, both types of calibration curves are used, i.e. intensity ratios versus mass fractions or mass fraction ratios. The curves of Fig. 2 demonstrate the good performance despite the compact dimensions of the spectrometer and laser.
Limit of detection (LOD) for C
For low C measurements and LOD determination, the samples of set C were used, see section 2.2. In Table 1 the mass fractions of the samples are listed together with the estimates of uncertainty for C as given in the certificates of the CRMs. In the mass fraction range of Fig. 2, these estimates of uncertainty are less than the size of the markers and therefore these errors of the mass fractions are neglected in Fig. 2 for clarity. But in the lower range, these are visible and they are added as error bars of the mass fraction values in Fig. 3.
Fig. 3. Analytical calibration curve for C in the low-level range up to 0.09% (a), and 0.016% (b) Note: There are no errors for the mass fractions available for the SUS samples and two of the CRMs (NBS 1165, NBS 1166).
The round markers in Fig. 3(a) are CRMs in the range up to ∼0.09% C showing a linear correlation. The procedure is that ten sample spots are measured, the resulting ten spectra are summated in groups of three to three sum spectra, and one spectrum is left out. The spectral line intensities (C1930)i and (Fe1932)i, of the sum spectra, i = 1-3, are determined and their intensity ratios are calculated, as described above. The mean and SD of these intensity ratios are shown in Fig. 3. The net duration of the laser bursts is therefore 1.4 s, see above.
Figure 3(b) is the enlarged range up to 150 µg/g from which the LOD is determined by Eq. 7 of [6] with k = 3 (3-sigma LOD)
Beside the CRMs, two SUS (BR 53 and Fe 130-35, see Table 1) are added in Fig. 3(b) to further validate the LOD value. The BR 53 is pure iron tested for good homogeneity by the supplier. For the mass fraction of C only an upper limit of 10 µg/g is provided and this is drawn in Fig. 3(b). It appears that the mass fraction is less and by taking this sample as blank sample the LOD would be ∼17 µg/g. Therefore, sample homogeneity may play a greater role than in spark-AES where sample areas of about 3-5 mm diameter are measured.
One important steel application for handheld LIBS is the separation of the widely used stainless steel grades 304 and 304L, also known as 1.4301 and 1.4307. Their compositions differ only by carbon with a standard specification of < 0.08% for 304, and < 0.03% for the low carbon version 304L, but their properties change significantly and they have to be clearly identified. Other speciality steels have actually less than 0.01% C ( = 100 µg/g C). In Fig. 3(a), the measured values of two SUS samples of stainless steel with 0.01% C and 0.035% C are added. The mass fractions are the guiding values by the SUS supplier. The standard specifications for the steel types 1.4306 and 1.4541 are < 0.03% C, and < 0.08% C, and therefore identical to 304L and 304. The samples are measured twice and the data show the clear separation of these stainless steels due to their low-level carbon content.
4. Conclusions
Carbon in steel can be measured by LIBS with components suitable for handheld instruments in a large range of mass fractions down to < 100 µg/g. The miniature spectrometer with an instrumental broadening of less than 36 pm in combination with a high laser repetition rate of 1 kHz enables a LOD of 34 µg/g for C which is validated with an adequate sample set in the low-level range.
To our knowledge, this LOD is the best value reported for a LIBS setup of this dimensions. It does not yet achieve the best performance of stationary instruments but is already useful for the differentiation of stainless steel grades such as 304 and 304L. In addition, analytical calibration curves of Cr, Ni and Si in alloy steel show good correlation with coefficients of determination of > 0.98 over a wide range of mass fractions. The net duration of the laser bursts is less than 1.4 s for the measurement of a sample.
Acknowledgments
Many thanks to Chao He and colleagues for laser cutting of spectrometer slits.
Disclosures
The authors declare no conflicts of interest.
References
1. A. Kramida, Yu. Ralchenko, J. Reader, and NIST ASD Team (2018). NIST Atomic Spectra Database (ver. 5.6.1), [Online]. Available: https://physics.nist.gov/asd [2019, July 25]. National Institute of Standards and Technology, Gaithersburg, MD.
2. A. W. Miziolek, V. Palleschi, and I. Schechter, Laser-induced breakdown spectroscopy (Cambridge University, 2006).
3. D. A. Cremers and L. J. Radziemski, Handbook of Laser-Induced Breakdown Spectroscopy (John Wiley, 2006).
4. D. W. Hahn and N. Omenetto, “Laser-Induced Breakdown Spectroscopy (LIBS), Part I: Review of Basic Diagnostics and Plasma—Particle Interactions: Still-Challenging Issues within the Analytical Plasma Community,” Appl. Spectrosc. 64(12), 335A–336A (2010). [CrossRef]
5. D. W. Hahn and N. Omenetto, “Laser-Induced Breakdown Spectroscopy (LIBS), Part II: Review of Instrumental and Methodological Approaches to Material Analysis and Applications to Different Fields,” Appl. Spectrosc. 66(4), 347–419 (2012). [CrossRef]
6. G. L. Long and J. D. Winefordner, “Limit of detection, a closer look at the IUPAC definition,” Anal. Chem. 55(7), 712A–724A (1983). [CrossRef]
7. V. Sturm, L. Peter, and R. Noll, “Steel analysis with laser-induced breakdown spectrometry in the vacuum ultraviolet,” Appl. Spectrosc. 54(9), 1275–1278 (2000). [CrossRef]
8. M. Khater, “Review: Laser-induced breakdown spectroscopy for light elements detection in steel: State of the art,” Spectrochim. Acta, Part B 81, 1–10 (2013). [CrossRef]
9. X. Jiang, P. Hayden, J. T. Costello, and E. T. Kennedy, “Double-pulse laser induced breakdown spectroscopy with ambient gas in the vacuum ultraviolet: Optimization of parameters for detection of carbon and sulfur in steel,” Spectrochim. Acta, Part B 101, 106–113 (2014). [CrossRef]
10. T. A. Labutin, S. M. Zaytsev, A. M. Popov, and N. B. Zorov, “Carbon determination in carbon-manganese steels under atmospheric conditions by laser-induced breakdown spectroscopy,” Opt. Express 22(19), 22382–22387 (2014). [CrossRef]
11. V. S. Burakov, M. V. Belkov, V. V. Kiris, K. Y. Catsalap, and N. V. Tarasenko, “Determination of carbon in low-alloy steels by laser-induced breakdown spectroscopy in air,” J. Appl. Spectrosc. 85(5), 864–869 (2018). [CrossRef]
12. K. Y. Yamamoto, D. A. Cremers, M. J. Ferris, and L. E. Foster, “Detection of metals in the environment using a portable laser-induced breakdown spectroscopy instrument,” Appl. Spectrosc. 50(2), 222–233 (1996). [CrossRef]
13. C. Lopez-Moreno, K. Amponsah-Manager, B. W. Smith, I. B. Gornushkin, N. Omenetto, S. Palanco, J. J. Laserna, and J. D. Winefordner, “Quantitative analysis of low-alloy steel by microchip laser induced breakdown spectroscopy,” J. Anal. At. Spectrom. 20(6), 552–556 (2005). [CrossRef]
14. Qingdong Zeng, Lianbo Guo, Xiangyou Li, Meng Shen, Yining Zhu, Jiaming Li, Xinyan Yang, Kuohu Li, Jun Duan, and Xiaoyan Zeng;, and Yongfeng Lu, “Quantitative analyses of Mn, V, and Si elements in steels using a portable laser-induced breakdown spectroscopy system based on a fiber laser,” J. Anal. At. Spectrom. 31(3), 767–772 (2016). [CrossRef]
15. A. Tortschanoff, M. Baumgart, and G. Kroupa, “Application of a compact diode pumped solid-state laser source for quantitative laser induced breakdown spectroscopy of steel,” Opt. Eng. 56(12), 1 (2017). [CrossRef]
16. V. N. Lednev, A. E. Dormidonov, P. A. Sdvizhenskii, M. Ya Grishin, A. N. Fedorov, A. D. Savvin, E. S. Safronova, and S. M. Pershin, “Compact diode-pumped Nd:YAG laser for remote analysis of low-alloy steels by laser-induced breakdown spectroscopy,” J. Anal. At. Spectrom. 33(2), 294–303 (2018). [CrossRef]
17. C. Alvarez-Llamas, C. Roux, and O. Musset, “A compact, high-efficiency, quasi-continuous wave mini-stack diode pumped, actively Q-switched laser source for laser-induced breakdown spectroscopy,” Spectrochim. Acta, Part B 148, 118–128 (2018). [CrossRef]
18. D. Day, B. Connors, M. Jennings, J. Egan, K. Derman, P. Soucy, S. Moller, and D. Sackett, “A full featured handheld LIBS analyzer with early results for defense and security,” Proc. SPIE 9482, 948206 (2015). [CrossRef]
19. Q. Li, J. Li, K. Bakeev, and S. Wang, “Developing mobile LIBS solutions for real world applications,” Proc. SPIE 9482, 94820F (2015). [CrossRef]
20. B. N. Bennet, M. Z. Martin, D. N. Leonard, and E. Garlea, “Calibration curves for commercial copper and aluminum alloys using handheld laser induced breakdown spectroscopy,” Appl. Phys. B: Lasers Opt. 124(3), 42 (2018). [CrossRef]
21. M. S. Afgan, Z. Hou, and Z. Wang, “Quantitative analysis of common elements in steel using a handheld-LIBS instrument,” J. Anal. At. Spectrom. 32(10), 1905–1915 (2017). [CrossRef]
22. J. Rakovský, P. Čermák, O. Musset, and P. Veis, “A review of the development of portable laser induced breakdown spectroscopy and its applications,” Spectrochim. Acta, Part B 101, 269–287 (2014). [CrossRef]
23. R. Noll, C. Fricke-Begemann, S. Connemann, C. Meinhardt, and V. Sturm, “LIBS analyses for industrial applications – an overview of developments from 2014 to 2018,” J. Anal. At. Spectrom. 33(6), 945–956 (2018). [CrossRef]
24. R. A. Crocombe, “Portable spectroscopy,” Appl. Spectrosc. 72(12), 1701–1751 (2018). [CrossRef]
25. SciAps ApNotes “Carbon analysis in stainless and carbon steels with handheld LIBS,” March 2018, https://www.sciaps.com/wp-content/uploads/2018/04/SciAps_ApNote_LIBSCarbonMARCH2018.pdf?x24702. [accessed Aug 1 2019].
26. A. Bideau-Mehu, Y. Guern, R. Abjean, and A. Johannin-Gilles, “Measurement of refractive indices of neon, argon, krypton and xenon in the 253.7-140.4 nm wavelength range. Dispersion relations and estimated oscillator strengths of the resonance lines,” J. Quant. Spectrosc. Radiat. Transfer 25(5), 395–402 (1981). [CrossRef]
27. N. Konjevic, A. Lesage, J. R. Fuhr, and W. L. Wiese, “Experimental stark widths and shifts for spectra lines of neutral and ionized atoms,” J. Phys. Chem. Ref. Data 31(3), 819–927 (2002). [CrossRef]
28. OriginPro2019, version 9.6.0.172, OriginLab Corporation, Northampton, MA 01060.
