Laser-induced breakdown spectroscopy for solution sample analysis using porous electrospun ultrafine fibers as a solid-phase support

Abstract

The major application of the laser-induced breakdown spectroscopy (LIBS) technique had been in the analysis of solid samples because the measurement of LIBS for liquid samples experiences some experimental difficulties, such as splashing, a quenching effect, and a shorter plasma lifetime. In the present work, electrospun ultrafine fibers were explored and used for the first time as a solid-phase support to quantify chromium (Cr) and copper (Cu) in aqueous solutions by LIBS. The liquid sample was first transferred to an ultrafine fiber surface, which could minimize the drawbacks of liquid sample analysis with LIBS. Due to the special micro-porous structure, the electrospun ultrafine fibers could hold a larger liquid sample and also the liquid sample was easy to evaporate. On the other hand, as a polymer substrate, the porous electrospun ultrafine fibers contributed to the minimal blank since there was no other unwanted heavy metal matrix that affected the detection during the liquid LIBS analysis. Meanwhile, the large sampling spot to fiber diameter ratio will minimize the potential influence generated in the liquid sample distribution process. With this pre-treated sample technique, the sensitivities of LIBS for liquid samples are improved considerably and the detection limits for Cr and Cu reached 1.8 ppm and 1.9 ppm, respectively. Therefore, the present strategy definitely paves the way for a wider application of LIBS in liquid sample analysis.

---

1 Introduction

Laser-induced breakdown spectroscopy (LIBS) as a new analytical technique, is based on analyzing the emissions from transient laser-induced plasma for qualitative and quantitative determination of the elemental composition of a sample. In the past few decades, LIBS had been proven to be an extremely versatile analytical technique.^1–5 Several review papers have reported different aspects of LIBS, from basic principles to commercial instruments.^6–9 Because of the high sensitivity and minimal sample preparation of the technique, LIBS had been widely employed in a large range of industrial and scientific applications for the determination of trace components in different sample matrices, including aerosols,^10–13 gases,^14,15 and liquids.^16,17 Nevertheless, the main application of LIBS technology has been, up to now, the analysis of solid samples because the analysis of liquids using LIBS presents a number of experimental difficulties. First, the laser-induced plasma emission intensity could be significantly reduced since most of the plasma energy can be consumed in the vaporization process of the liquids.¹⁸ In either the bulk or the surface of liquids, the lifetime of the analyte excited states is relatively short compared to the plasmas generated in solid surfaces.^19,20 On the other hand, dissolved gases and particulate material in the liquid itself can lead to incorrect focusing of the laser beam and can also serve as breakdown sites prior to the laser beam focus.¹⁸ All these factors result in lowering the figures of merit, i.e., limits of detection, precision, accuracy, and sensitivity, with respect to solids or aerosols with LIBS analysis. As a consequence, different methodologies had been developed to resolve the difficulties in liquid sample analysis, including freezing the samples,²¹ using droplets,^22,23 or liquid jets,^24–26 and so on. Although these configurations have improved the plasma lifetime and sensitivity compared to direct LIBS analysis in liquid bulk,¹⁹ there has also been an increase in the complexity of the analytical equipment in all these methodologies and the experimental results still lack reproducibility. Other techniques, such as dual pulse LIBS ablation^27,28 or a combination of LIBS and laser-induced fluorescence,²⁹ have been performed to enhance the signal-to-noise ratio and lower the limit of detection (LODs). However, these two techniques require more complicated laser systems, making the experimental system even more complex and expensive.³⁰

The technique of liquid to solid conversion is a popular method for liquid LIBS analysis which could overcome the drawbacks of liquid analysis with LIBS and take advantage of solid targets. Therefore, a series of techniques related to liquid–solid conversion have been developed using various solid-phase sample supports before processing for LIBS analysis. So far, different solid-phase supports have been developed, such as a carbon planchet,¹⁸ bamboo charcoal,³⁰ metallic substrates,^19,31 filter papers,^32,33 calcium hydroxide pellets³⁴ and wood slices.^35,36 These materials have proven to be useful and applicable to improve the LODs in liquid LIBS analysis.

However, for the method with calcium hydroxide, the sample preparation is rather complicated, in which a filter paper is used as a solid-phase support and wrinkles form on the paper surface once it is dry, which make the paper too difficult to handle.³⁶ In addition, using metallic substrates will introduce apparent matrix effects, while using wood slices and bamboo charcoal substrates will seriously influence the measurement results since these substrates easily adsorb other materials. Porous electrospun ultrafine fibers are of potential importance to technological applications, such as catalytic carriers, energy storage systems, tissue engineering, wound dressing, sensors, drug delivery, etc.^37–39 Here, we are reporting the first exploration of using porous electrospun ultrafine fibers as a solid-phase support to quantify elements in aqueous solution by the LIBS technique. As they are a polymer material, there were no other unwanted heavy metal ions affecting the detection during the liquid LIBS analysis and they also minimized the blank contribution. Meanwhile, with a large surface area and special micro-porous characteristics, the porous electrospun ultrafine fibers could hold more samples. Therefore, the porous ultrafine fibers will be capable of fast and sensitive determination of elements in aqueous solutions using LIBS.

In this paper, we introduce novel porous electrospun ultrafine fibers as a solid-phase support for the first time to detect the concentration of metal elements in an aqueous solution utilizing the LIBS technique. The liquid sample was first transferred onto an ultrafine fiber surface. Due to the special micro-porous structure, the electrospun ultrafine fibers could hold a larger liquid sample and the loaded liquid sample was easy to evaporate. On the other hand, as a polymer substrate, the porous electrospun ultrafine fibers contribute to a minimal blank and heavy metal matrix, as there are almost no metal ions that affect the detection of heavy metals during the liquid LIBS analysis. Meanwhile, the crater effect on the plasma expansion and evolution of LIBS emission could be negligible due to the depth of the crater on the surface of the ultrafine fibers and the surface was always relatively flat after laser shots. Therefore, the present strategy definitely paves the way for a wider application of LIBS in liquid sample analysis. To increase the LODs, a compact Echelle spectrometer coupled with an EMCCD detector was used to detect the LIBS plasma emissions, which can provide higher resolution in a more compact size and cover a much wider spectral range than other conventional grating spectrometers. The laser energy and the time-delay dependence of the LIBS signals for Cr and Cu atomic lines were investigated in order to achieve optimal sensitivity. The calibration curves were built and the LODs for Cr and Cu were calculated respectively.

2 Experimental

2.1 LIBS instrumentation

A schematic diagram of the complete experimental system used in this study is shown in Fig. 1. A Q-switched Nd:YAG laser (LOTIS) operating at 1064 nm (pulse duration of 6 ns) with a maximal output pulse energy of 200 mJ and repetition rate of 10 Hz was used to irradiate the samples. The Nd:YAG laser beam was focused onto the target surface by a 50 mm focal length plano-convex lens at a 90° angle. The distance between the lens and the sample was set to approximately 40 mm to insure reproducible breakdown and produce an almost hemispherical plasma plume. The focal area on the target surface was about 2.0 × 10⁻³ cm². The energy per laser pulse was fixed to 40 mJ, which was measured at the target position using an energy-meter (Thorlabs ES220C). This laser system produced a typical irradiance of about 3.3 GW cm⁻². The plasma emissions were collected in side view geometry at a 50° angle relative to the spot where the laser beam was focused on the sample surface. A 1 m length fused-silica optical fiber mounted on a micro xyz-translation stage was used to collect the emission light from the plasma plume and feed it to a portable Echelle spectrometer with a 150 mm focal length (ARYELLE-UV-VIS, LTB150, Germany). The Aryelle 150 provides a constant spectral resolution (CSR) of 6000 over a wavelength range of 220–800 nm displayed in a single spectrum. An EMCCD camera (QImaging, UV enhanced, 1004 × 1002 pixels, USA), coupled to the spectrometer was used for the detection of the plasma emission. The overall linear dispersion of the spectrometer camera system ranges from 37 pm (at 220 nm) to 133 pm per pixel (at 800 nm). To prevent the EMCCD from detecting the early plasma continuum, a mechanical chopper was used in front of the entrance slit. The target was mounted on a two-dimensional rotating stage, which was used to provide a fresh surface after each laser pulse. To improve the spectral intensity, each recording was obtained by accumulating the signal of 100 ablation events on different sites.

Fig. 1 Schematic diagram of the experimental setup.

2.2 Preparation of the electrospun ultrafine fibers

The preparation of the porous electrospun ultrafine fibers was carried out as described in our previous report.³⁸ Polyether sulfone (PES) and polyethylene glycol (PEG) were dissolved in dimethyl sulfoxide (DMSO) at 60 °C until the solution became homogeneous. The solutions were still pellucid after 24 h. The concentration of PES was 20.3 wt%. The electrospun setup mainly consists of three major components (Fig. 2): a high-voltage power supply, a spinneret, and an electrically conductive collector. To electrospin the solutions, a direct current high voltage generator was used to provide a voltage of 15 kV. The solution was placed in a 50 mL syringe, to which a needle tip with a 0.5 mm inner diameter was attached. The positive electrode of the high-voltage power supply was connected to the needle. A grounded rotating drum was used as a collector and the speed of the drum was 0.6 m s⁻¹. The distance between the tip and the collector was maintained at 16 cm. The relative humidity was kept in the range of 71–76% without special instruction. The extraction of part of the PEG molecules from the PES–PEG blend ultrafine fiber matrix was carried out by washing them with water for several days at 90 °C.

Fig. 2 Schematic of a typical setup for the electrospinning.

2.3 Sample preparation

Stock solutions of Cr and Cu were prepared by dissolving known amounts of K₂Cr₂O₇ and CuSO₄·5H₂O in deionized water. Suitable dilutions were made from these stock solutions to prepare samples with Cr and Cu concentrations in the range of 0–800 ppm to establish calibration plots. A volumetric auto-pipette was used to transfer 100 μL of a solution drop onto the porous electrospun ultrafine fibers. The solution spread circularly over the porous electrospun ultrafine fibers with a radius of approximately 1 cm. The solution was evaporated to dryness with the help of a hot air blower. Subsequently, the porous electrospun ultrafine fibers were allowed to cool and were then mounted in the sample holder. The spatial distribution of the sample was marked on the porous electrospun ultrafine fibers so that the analysis could be performed over a region with a 1 cm diameter where both Cr and Cu were expected to be uniformly distributed.

3 Results and discussion

3.1 Characterization of the porous ultrafine fibers

In our experiment, porous PES-based ultrafine fibers with diameters ranging from 0.5 to 3 μm were prepared via an electrospinning technique. The average diameter of the porous ultrafine fibers was about 2.10 μm with a wider size distribution (see Fig. 3a). Meanwhile, removing part of the PEG molecules from the matrix of ultrafine fibers could bring about a much more porous structure and result in a higher porosity,³⁷ since the internal structure of all the ultrafine fibers is also porous (see Fig. 3b). On the other hand, part of the PES and PEG molecular chains might be entangledt, so there are still some PEG molecules in the matrix of ultrafine fibers and therefore, they can improve the hydrophilic properties for the ultrafine fibers. These two factors introduce a much better adsorption property to the porous electrospun ultrafine fibers.³⁷ In the present work, our approach consists of the evaporation of Cr and Cu solutions upon the porous electrospun ultrafine fiber substrate followed by LIBS analysis of the substrate surface. The SEM micrographs clearly indicate the presence of some new heavy metal loaded particles on the surface of the ultrafine fibers (see Fig. 3c), compared with the original ultrafine fibers (Fig. 3b). It is clearly shown that the porous ultrafine fibers have been fully filled, and there are no apparent porous holes existing on the surface of the fibers. The results of the SEM suggest that the porous electrospun ultrafine fibers have the capability to act as a solid-phase support for liquid LIBS analysis.

Fig. 3 SEM images of the porous PES-based ultrafine fibers (a and b, original ultrafine fibers; c, loaded with heavy metal).

As a polymer material, the original precursor solutions of the electrospinning contain a matrix polymer together with an alkoxide, water-soluble salt, or polymer precursor. In particular, all the polymer components of the ultrafine fibers are non-metal elements, including C(C(I) 247.86 nm), H(H(I) 656.43 nm), O(O(II) 388.22 nm, 435.84 nm) and Na(Na(I) 589.05 nm, 589.59 nm), which are normally induced by washing with water to extract the PEG molecules from the PES–PEG blend ultrafine fiber matrix (see Fig. 4). It is worth mentioning that S has a weak transition at a wavelength which cannot be detected in our experimental conditions. Therefore, there were no other unwanted heavy metal elements interfering with the detection of heavy metals during the liquid LIBS analysis, and thus the porous electrospun ultrafine fibers were selected as the substrate to minimize the blank contributions.

Fig. 4 Plasma emission signals of the porous electrospun ultrafine fibers.

In a particular case of solid analysis, however, an additional factor should be taken into account. The interaction of the laser beam with the sample surface can result in a crater formation. The creation of a crater on the solid-phase support surface could influence the nature of the plasma, and then consequently alter the accuracy of the experimental results.⁴⁰ The characteristics of an LIBS crater mainly depend on the physical properties of the sample, environmental conditions, and laser parameters.^41,42 When using the porous electrospun ultrafine fibers as a solid-phase substrate, there was also an apparent crater on the surface (see Fig. 5a). However, attributed to the laser and focusing parameters, the surface of crater was always relatively flat even after a series of laser shots at different positions in our experiment (see Fig. 5b). Therefore, the confinement effect on the influence of the plasma expansion and evolution of LIBS emission was almost negligible. As is shown in Fig. 5a, the diameter of the sampling spot formed by a laser pulse on the surface of fibers was about 756.3 μm, which is much larger than that of the average diameter of the fibers (2.1 μm, Fig. 3a). As a new solid-phase support, this large ratio of the sampling spot to the fiber diameter will be beneficial in minimizing the potential signal influence generated in the process of liquid sample distribution, which is an attractive characteristic for the porous ultrafine fibers.

Fig. 5 The sampling spots and craters on the surface of the porous electrospun ultrafine fibers after one laser shot (a) and serial shots (b).

3.2 Quantitative analysis and calibration curves

Quantitative analysis using the porous electrospun ultrafine fibers as a solid-phase substrate for determining the liquid-based analytes with LIBS technology was evaluated. 100 μL micro-droplets were placed on the porous electrospun ultrafine fiber substrate. To obtain the highest sensitivity, the micro-droplets were evaporated to dryness with the help of a hot air blower before analysis, and then a laser beam was focused on the dried micro-droplet spot to create the LIBS plasma. The analysis area was controlled to a radius of approximately 1 cm. The Cr and Cu plasma emission intensities were detected by a portable Echelle spectrometer. To obtain reliable and repeatable results, care should be taken to dry the liquid completely on the porous electrospun ultrafine fibers, since a small amount of moisture may drastically influence the emission intensity and may lead to erratic results.³²

Preliminary experiments were carried out to optimize various experimental parameters of the LIBS system in order to obtain the best SBR value. The dependence of the emission spectra on the laser pulse energy was studied. When the laser energy was higher than 40 mJ, slight charring of the porous electrospun ultrafine fibers was observed and hence, the laser energy was fixed at around 40 mJ for the current work. On the other hand, there was a strong background continuum emission at the beginning of the plasma formation due to bremsstrahlung and recombination events, which seriously interfered with the detection of atomic lines. Therefore, the spectral data acquisition was accomplished with a certain time delay after the initiation of the laser spark by obtaining an appropriate delay time through a pulse generator.^43,44 In general, the optimal acquisition delay time depends on the laser pulse energy, the kind of sample and the surrounding atmosphere. However, a single delay time could be chosen for all the analytical lines in practical applications.^10,45 As a compromise between the signal-to-noise ratio (SNR) and peak emission intensity, the optimum delay time was estimated to be 2.5 μs in the present study (see Fig. 6).

Fig. 6 Dependence of the LIBS signal-to-noise ratio on the delay time for the Cu(I) 324.75 nm emission line, using 40 mJ laser energy.

The spectral region (220–800 nm) was used to record the emission signals containing most of the strong emission lines for both Cr and Cu. Through careful comparison of these emission spectra, regions where the spectral interference from the porous electrospun ultrafine fibers were minimized, and suitable lines, two lines for Cu (Cu(I) 324.75 nm, Cu(I) 327.39 nm) and three atomic lines for Cr (Cr(I) 425.43 nm, Cr(I) 427.48 nm, Cr(I) 428.97 nm), were selected for quantitatively identifying Cu and Cr. There was no apparent spectrally interfering emission from the porous electrospun ultrafine fibers for the analytical lines (see Fig. 7). Apart from these analytical lines, Cu(I) 427.51 nm, Cu(III) 435.19 nm, Cr(I) 357.85 nm, Cr(I) 359.34 nm and Cr(I) 360.53 nm emission lines can also be used for calibration purposes but more sensitive lines were chosen for the analytical purpose in this work, which was especially useful for low concentration species. On the other hand, since the intensity of the analytical lines of Cr and Cu in the plasma plume was not strong due to their low concentration in the target solution, the self-absorption effect may be neglected within the experimental uncertainty. Therefore, these lines were considered to be good candidates for lowered LODs compared with other atomic emission lines.

Fig. 7 Emission spectra of Cu (a) and Cr (b) obtained by LIBS under identical analysis conditions.

The calibration curves for Cr and Cu were linear in the low concentration range and showed saturation at high concentrations due to self-absorption in the plasma. This is commonly observed and reported in laser-induced plasmas at atmospheric pressure.^33,34 For low concentrations, the calibration curves were obtained using the intensity of the selected emission lines of the elements versus the corresponding concentrations of Cr and Cu, and these results are shown in Fig. 8a and b, respectively. In these figures, each point corresponds to an average of ten measurements performed using fresh porous ultrafine fibers and the error bars are the standard deviations.

Fig. 8 Calibration curves obtained for the analytical lines of Cu(I) (a) and Cr(I) (b). All the elements were performed at a 2.5 μs delay time using a 40 mJ Nd:YAG laser operating at 1064 nm and a 6 ns pulse duration.

As shown in Fig. 8, the calibration curves of the elements in the ultrafine fiber matrix show a good linear fit (R² > 0.9) within the experimental uncertainty. In fact, these results give an expectation that the proposed porous ultrafine fibers would have the capability to be a novel substrate for good linearity of the calibration curves for liquid sample analysis. Therefore, the limits of quantitation (LOQs), meaning the lowest concentration that can be detected in the samples, were first investigated. The LOQs for Cr and Cu from the present work are estimated to be 10 ppm and 5 ppm (practical measurement), respectively. The signals obtained for 10 ppm of Cr and 5 ppm of Cu are shown in Fig. 9a and b, respectively. The same method for the LOQs measurement was also used by Gondal et al.⁴⁶

Fig. 9 Plasma emission signals obtained from the blank ultrafine fibers (red line) and a micro-droplet with a concentration of 5 ppm Cu (a) and 10 ppm Cr (b) (black line).

On the other hand, the LODs are both elemental and spectral line dependent. It is usually defined as the concentration that originates a net line-intensity equivalent to three times the standard deviation from the background, measured close to the line profile. The LOD for each element based on different analysis lines was calculated according to the IUPAC definition:

where, σ_B is the standard deviation of the background, and s is the sensitivity given by the slope of the corresponding calibration curve.

In this work, σ_B was determined from ten measurements of the background signals under the same experimental conditions, where the sample was the porous electrospun ultrafine fibers dipped in pure deionized water. Table 1 lists the slopes, 3σ_B values, and LODs determined in this work.

Table 1 Limits of detection for different elemental spectral lines

Element	Slope (mL μg^−1)	3σB	Detection limit (μg mL^−1)
Cr(I) 425.43 nm	30.637	55.46	1.8
Cr(I) 427.48 nm	24.346	47.64	2.0
Cr(I) 428.97 nm	17.810	56.22	3.2
Cu(I) 324.75 nm	28.482	53.83	1.9
Cu(I) 327.39 nm	11.809	38.89	3.3

---

The LODs of Cr and Cu were determined in different matrices (either liquids or solids) by different authors, using different LIBS experimental setups in different experimental conditions. The LODs obtained for Cr and Cu by several authors aere given in Table 2 for comparison purposes. The LODs obtained depend not only on the sample matrix but also on the LIBS equipment. Therefore, direct comparison of the LODs for solid and liquid matrices, or even comparison within the same kind of matrix is rather difficult to perform.¹⁹ However, as an organic polymer solid-phase support, no other unwanted heavy metal affected the detection during the liquid LIBS analysis with the porous electrospun ultrafine fibers. Moreover, with the help of a portable Echelle spectrometer coupled with an EMCCD camera, more attractive LODs were obtained without any pre-treatment for the liquid samples.

Table 2 A comparison of the detection limits (ppm) from the present work with those reported in the literature

Analytical line (nm)	Matrix	Sampling methodology	LODs (ppm)	Detection setup	Ref.
Cr(II) 283.50	Water	Liquid jet	30	Czerny Turner spectrometer-CCD	47
Cr(I) 520.45	Soils	Solid pellet	17	Czerny Turner spectrometer-ICCD	48
Cr(II) 267.72	Water	Ice	1.4	Czerny Turner spectrometer-ICCD	21
Cr(I) 425.43	Water	Liquid jet	39	Czerny Turner spectrometer-CCD	49
Cr(I) 425.43	Porous fibers	Liquid droplets	1.8	Echelle spectrograph-EMCCD	This work
Cu(I) 282.43	Soils	Solid pellet	61	Czerny Turner spectrometer-ICCD	48
Cu(I) 324.75	Water	Ice	2.3	Czerny Turner spectrometer-ICCD	21
Cu(I) 324.75	Alloy	Solid analysis	17	Echelle spectrograph-ICCD	50
Cu(I) 324.75	Alloy	Solid analysis	30	Echelle spectrograph-ICCD	51
Cu(I) 324.75	Porous fibers	Liquid droplets	1.9	Echelle spectrograph-EMCCD	This work

---

4 Conclusions

The results of this work demonstrated that porous electrospun ultrafine fibers were an excellent material as a solid-phase support to quantify elements in liquid samples with the LIBS technique. With this novel material as a solid support, the chromium (Cr) and copper (Cu) in an aqueous solution were transferred onto a solid surface for analysis. A portable Echelle spectrometer with an electron-multiplying CCD camera was used to improve the LIBS limits of detection for liquid sample analysis. The limits of detection for Cr and Cu from the present work were estimated to be 1.8 ppm and 1.9 ppm with the novel porous electrospun ultrafine fiber substrate. As a type of organic polymer solid-phase support, there were no heavy metals to interfere with the detection of liquid LIBS analysis with the porous electrospun ultrafine fibers. Therefore, the porous ultrafine fibers are easy to handle and provide a practical approach for fast and sensitive determination of elements in aqueous solutions using LIBS. The electrospun fiber material can be useful for various applications and will be interesting to the analytical community. However, in order to have a better understanding on the characteristics of this novel support material, further work concerning the adsorption capabilities of the porous electrospun ultrafine fibers and the laser–fiber interaction mechanism is currently being processed in our laboratory.

Acknowledgements

The authors are grateful for the financial support from the National Major Scientific Instruments and Equipment Development Special Funds (no. 2011YQ030113), the National Recruitment Program of Global Experts (NRPGE), the Hundred Talents Program of Sichuan Province (HTPSP), and the Startup Funding of Sichuan University for setting up the Research Center of Analytical Instrumentation.

References

1. D. W. Hahn, A. W. Miziolek and V. Palleschi, Appl. Opt., 2003, 42, 5937.
2. A. C. Samuels, F. C. Delucia Jr, K. L. McNesby and A. W. Miziolek, Appl. Opt., 2003, 42, 6205–6209.
3. M. R. Martelli, F. Brygo, A. Sadoudi, P. Delaporte and C. Barron, J. Agric. Food Chem., 2010, 58, 7126–7134.
4. T. Delgado, J. M. Vadillo and J. J. Laserna, J. Anal. At. Spectrom., 2013, 28, 1377–1384.
5. K. Rifai, S. Laville, F. Vidal, M. Sabsabi and M. Chaker, J. Anal. At. Spectrom., 2012, 27, 276–283.
6. A. K. Pathak, R. Kumar, V. K. Singh, R. Agrawal, S. Rai and A. K. Rai, Appl. Spectrosc. Rev., 2012, 47, 14–40.
7. K. Song, Y.-I. Lee and J. Sneddon, Appl. Spectrosc. Rev., 2002, 37, 89–117.
8. R. Yoshiie, Journal of the Spectroscopical Society of Japan, 2007, 56, 128–137.
9. Q. Y. Lin, G. h. Niu, Q. h. Wang, Q. l. Yu and Y. x. Duan, Appl. Spectrosc. Rev., 2013, 48, 487–508.
10. E. M. Cahoon and J. R. Almirall, Anal. Chem., 2012, 84, 2239–2244.
11. P. Diwakar, P. Kulkarni and M. E. Birch, Aerosol Sci. Technol., 2012, 46, 316–332.
12. M. A. Awan, S. H. Ahmed, M. R. Aslam, I. A. Qazi and M. A. Baig, Arabian J. Sci. Eng., 2013, 38, 1655–1661.
13. G. A. Lithgow, A. L. Robinson and S. G. Buckley, Atmos. Environ., 2004, 38, 3319–3328.
14. V. Kumar and R. K. Thareja, Laser Part. Beams, 1992, 10, 109–116.
15. R. Yoshiie, Materia, 2004, 43, 201–205.
16. F. A. Barreda, F. Trichard, S. Barbier, N. Gilon and L. Saint-Jalmes, Anal. Bioanal. Chem., 2012, 403, 2601–2610.
17. A. Matsumoto, A. Tamura, K. Fukami, Y. H. Ogata and T. Sakka, Anal. Chem., 2013, 85, 3807–3811.
18. R. L. V. Wal, T. M. Ticich, J. Joseph, R. West and P. A. Householder, Appl. Spectrosc., 1999, 53, 1226–1236.
19. M. A. Aguirre, S. Legnaioli, F. Almodóvar, M. Hidalgo, V. Palleschi and A. Canals, Spectrochim. Acta, Part B, 2013, 79–80, 88–93.
20. L. St-Onge, E. Kwong, M. Sabsabi and E. B. Vadas, J. Pharm. Biomed. Anal., 2004, 36, 277–284.
21. H. Sobral, R. Sanginés and A. Trujillo-Vázquez, Spectrochim. Acta, Part B, 2012, 78, 62–66.
22. C. Janzen, R. Fleige, R. Noll, H. Schwenke, W. Lahmann, J. Knoth, P. Beaven, E. Jantzen, A. Oest and P. Koke, Spectrochim. Acta, Part B, 2005, 60, 993–1001.
23. S. Groh, P. K. Diwakar, C. C. Garcia, A. Murtazin, D. W. Hahn and K. Niemax, Anal. Chem., 2010, 82, 2568–2573.
24. P. Yaroshchyk, R. J. Morrison, D. Body and B. L. Chadwick, Appl. Spectrosc., 2004, 58, 1353–1359.
25. N. K. Rai and A. K. Rai, J. Hazard. Mater., 2008, 150, 835–838.
26. C. B. Faye, T. Amodeo, E. Fréjafon, N. Delepine-Gilon and C. Dutouquet, Spectrochim. Acta, Part B, 2014, 91, 5–11.
27. A. Kumar, F. Y. Yueh and J. P. Singh, Appl. Opt., 2003, 42, 6047–6051.
28. S. Nakamura, Y. Ito and K. Sone, Anal. Chem., 1996, 68, 2981–2986.
29. S. L. Lui, Y. Godwal, M. T. Taschuk, Y. Y. Tsui and R. Fedosejevs, Anal. Chem., 2008, 80, 1995–2000.
30. D. H. Zhu, J. P. Chen, J. Lu and X. W. Ni, Anal. Methods, 2012, 4, 819–823.
31. Z. J. Chen, H. K. Li, F. Zhao and R. H. Li, J. Anal. At. Spectrom., 2008, 23, 871–875.
32. A. Sarkar, D. Alamelu and S. K. Aggarwal, Appl. Opt., 2008, 47, G58–G64.
33. D. Alamelu, A. Sarkar and S. K. Aggarwal, Talanta, 2008, 77, 256–261.
34. D. M. Díaz Pace, C. A. D'Angelo, D. Bertuccelli and G. Bertuccelli, Spectrochim. Acta, Part B, 2006, 61, 929–933.
35. Z. J. Chen, H. K. Li, M. Liu and R. H. Li, Spectrochim. Acta, Part B, 2008, 63, 64–68.
36. Z. J. Chen, G. Yogesh, T. Ying Yin and F. Robert, Appl. Opt., 2010, 49, C87–C94.
37. Z. M. Wei, Q. C. Zhang, L. H. Wang, M. L. Peng, X. J. Wang, S. R. long and J. Yang, J. Polym. Sci., Part B: Polym. Phys., 2012, 50, 1414–1420.
38. Z. M. Wei, Q. C. Zhang, L. H. Wang, X. J. Wang, S. R. Long and J. Yang, Colloid Polym. Sci., 2013, 291, 1293–1296.
39. J. W. Olesik, Spectroscopy, 1999, 14, 36–41.
40. M. Corsi, G. Cristoforetti, M. Hidalgo, D. Iriarte, S. Legnaioli, V. Palleschi, A. Salvetti and E. Tognoni, Appl. Spectrosc., 2005, 59, 853–860.
41. X. Z. Zeng, S. S. Mao, C. Y. Liu, X. L. Mao, R. Greif and R. E. Russo, Spectrochim. Acta, Part B, 2003, 58, 867–877.
42. R. E. Russo, X. L. Mao, C. Liu and J. Gonzalez, J. Anal. At. Spectrom., 2004, 19, 1084–1089.
43. Handbook of laser-induced breakdown spectroscopy, ed. D. A. Cremers and L. J. Radziemski, Wiley, Chichester, England, 2006.
44. A. Eslami Majd, A. S. Arabanian and R. Massudi, Opt. Lasers Eng., 2010, 48, 750–753.
45. R. Wisbrun, I. Schechter, R. Niessner, H. Schroeder and K. L. Kompa, Anal. Chem., 1994, 66, 2964–2975.
46. M. A. Gondal and T. Hussain, Talanta, 2007, 71, 73–80.
47. N. K. Rai and A. K. Rai, J. Hazard. Mater., 2008, 150, 835–838.
48. M. Dell'Aglio, R. Gaudiuso, G. S. Senesi, A. De Giacomo, C. Zaccone, T. M. Miano and O. De Pascale, J. Environ. Monit., 2011, 13, 1422–1426.
49. M. Y. Yao, J. L. Lin, M. H. Liu and Y. Xu, Appl. Opt., 2012, 51, 1552–1557.
50. W. T. Y. Mohamed, Opt. Laser Technol., 2008, 40, 30–38.
51. M. Sabsabi, V. Detalle, M. A. Harith, W. Tawfik and H. Imam, Appl. Opt., 2003, 42, 6094–6098.

---