Detection of trace elements in nondegradable organicspent clay waste using optimized dual-pulsed laser

induced breakdown spectrometer

Ahmed Asaad I. Khalil,1,2 Mohammed A. Gondal,3,* and Mohamed A. Dastageer3

1National Institute of Laser Enhanced Sciences (NILES), Cairo University, Giza 12613, Egypt2Physics Department, Faculty of Science for Girls, Dammam University, Box 838, Dammam 31113, Saudi Arabia

3Physics Department and Center of Excellence in Nanotechnology, King Fahd University of Petroleum & Minerals (KFUPM),Box 5047, Dhahran 31261, Saudi Arabia

*Corresponding author: magondal@kfupm.edu.sa

Received 9 September 2013; revised 24 December 2013; accepted 2 January 2014;posted 4 February 2014 (Doc. ID 197310); published 10 March 2014

The detection of trace elements present in nondegradable organic spent clay waste has been carried outusing an optimized dual-pulsed laser induced breakdown spectrometer. The two laser pulses at 1064 and266 nm were collinearly collimated and focused on the sample surface in order to enhance the signalintensity. The atomic transition lines at 568.8 nm (Na-I), 504.2 nm (Pb-II), 405.8 nm (Pb –I),443.56 nm (Ca-I), 469.41 nm (S-I), 520.8 nm (Cr-I), 643 nm (Cd-I), and 928.1 nm (Cl-I) were used asmarker wavelengths, and the concentrations of 688, 300, 204, 460, and 2440 ppm of Pb, S, Cd, Cr,and Cl, respectively, were detected in the 5% spent clay in the binder. The limits of detection of Pb,S, Cd, Cr, and Cl were estimated to be 6.7, 17.2, 6.5, 5.1, and 14.8 ppm, respectively, from the calibrationcurve for each element. In order to confirm the reliability of our system, the concentrations of the reportedelements detected using our systemwere compared to the ones obtained with inductively coupled plasmaemission spectroscopy and found to be in good agreement. © 2014 Optical Society of AmericaOCIS codes: (300.0300) Spectroscopy; (300.6210) Spectroscopy, atomic; (300.6360) Spectroscopy,

laser; (300.6365) Spectroscopy, laser induced breakdown.http://dx.doi.org/10.1364/AO.53.001709

1. Introduction

Industrial wastes in solid, liquid, and gaseous formsare the major contributors to environmental pollu-tion. Heavy metals are highly dense elements, andeven a low concentration of them is quite toxic andthey can easily accumulate in the living organismand contaminate our ecosystem [1]. Lead, mercury,arsenic, and cadmium are the most common heavymetals found in the industrial effluents dischargedmainly from the metallurgical, petrochemical, paint,and tanning industries. The chemical composition of

spent clay (used clay) is as follows: 44.2% of SiO2,39.7% of Al2O3, 1.39% of TiO2, 0.13% of Fe2O3,0.08% of S, 0.002% of CrO, 0.03% of Cl, 0.013%of Na2O, 0.05% of CdO, 0.013% of Pb, and 0.034% ofP2O5. The detection and precise quantification of thetoxic elements present in organic waste are amongthe most sought-after propositions in the areas ofhealth and environmental protection. There aremany established analytical tools, such as induc-tively coupled plasma-atomic emission spectroscopy(ICP-AES), atomic absorption spectroscopy (AAS),microwave induced plasma (MIP) [2–5], and, ofcourse, x-ray photoelectron spectroscopy (XPS)and x-ray fluorescence spectroscopy (XRF). Most ofthese techniques involve bulky and expensive

1559-128X/14/081709-09$15.00/0© 2014 Optical Society of America

10 March 2014 / Vol. 53, No. 8 / APPLIED OPTICS 1709

instrumentation, and time-consuming and cumber-some sample preparation. Besides these inherentlimitations, these methods are not rapid, and are lessselective and less sensitive in nature [6]. Laser in-duced breakdown spectroscopy (LIBS) is based onthe atomic emission spectroscopy of the laser inducedplasma and with modern laser and spectroscopictools. LIBS outsmarts the traditional techniques ofelemental analysis by catching up from their pointof limitations in terms of prolonged sample prepara-tion methods and the use of various chemicals.

In LIBS a high-power pulsed laser beam of a fewnanoseconds pulse duration is focused on the samplesurface; the highly intense laser pulse momentarilycreates a plasma on the sample surface through mul-tiphoton and collisional processes. The initial tem-perature of the multicolored white plasma reachesas high as 25 kilo Kelvin and rapidly falls, whichwe technically call cooling of plasma [7–10]. Theplasma emission is directed on to a monochromator/spectrograph for wavelength dispersion, which deci-phers the presence of any element based on its char-acteristic atomic emission wavelengths. The timingof the detection after the laser excitation is a keyparameter to deal with. A few hundred nanosecondsafter excitation, the plasma remains very hot and thewhite plume gives no meaningful information as ityields only a broad featureless continuum spectrum.As time passes by up until a few microseconds, thecomponents of the plasma recapture electrons andbecome excited neutral atoms, and after few micro-seconds, these atoms transfer from the excited statesto lower states by emitting the characteristic wave-lengths inherent to the elements present in the sam-ple [11]. In addition to the emission from the neutralatoms, we can also detect emission from the singlyionized atomic species provided we keep our detec-tion time window to an earlier time. The detectiontime gate is adjusted by tuning the delay betweenthe laser and the detector.

Improving the signal-to-noise ratio (S/N) and thelimit of detection (LOD) is the principal goal inany analytical system and so it is in the case of LIBS.Many variants of LIBS systems for the elementalanalysis have been investigated, and among thesethe dual-pulsed LIBS system showed significant im-provement for sensitivity, S/N, and LOD [12–15]. Inthe case of a dual-pulsed LIBS system, two laserbeams (266 and 1064 nm) are collimated to be collin-ear with each other and spatially coincide on thesample surface with proper beam steering optics[16–18]. A suitable delay is introduced between thetwo laser beams (order of nanoseconds to microsec-onds between the two beams) to optimize the LIBSsignal. In addition to the delay between the two la-sers, there should be a delay between the laser andthe detection system with the help of another delaygenerator [19–24]. The intuitive reason for the LIBSsignal improvement could be the additional energyfrom the second laser. However, this was ruled out bythe study of Sattman et al. [25]. Corsi et al. [26,27],

studying plume evolution, attributed the signal en-hancement in the dual-pulsed LIBS system to thefaster plume expansion in double-pulse laser abla-tion using blast wave theory. Hohreiter et al. [28] car-ried out a detailed study on the plasma created bythe single and dual pulse and suggested inconclu-sively that some kind of coupling between the secondlaser and the preablated plasma due to shock in-duced reduced pressure, or the thermal lensingeffect, could occur through preablation laser spark.Although there have been many works [29,30] at-tempting to explain this process, so far no satisfac-tory explanation has been given to explain theunderlying physical process for the LIBS signal en-hancement, and there is still a good scope for theo-retical work. Whatever the physical process of thisenhancement, one aspect that is proven beyonddoubt is that the signal enhancement is accompaniedby improved ablation on the sample surface.

In this work, the dual-pulsed LIBS has been opti-mized in terms of beam geometry, the time delaybetween the two lasers, the sampling gate width,the excitation/signal acquisition delay, and the exci-tation sources. The optimized DP-LIBS is applied fordetection of trace metals present in the nondegrad-able organic spent clay. In the DP-LIBS, the wave-length combination of 1064 and 266 nm pulsedlaser beams has shown the best signal level. Thesetwo beams were collinearly collimated and focusedon the sample surface, rather than making a 45°angle between them, and this geometry has shownbetter enhancement of the signal intensity. Theatomic transition lines at 568.8 nm (Na-I), 504.2 nm(Pb-II), 405.8 nm (Pb-I), 443.56 nm (Ca-I), 469.41 nm(S-I), 520.8 nm (Cr-I), 643 nm (Cd-I), and 928.1 nm(Cl-I) were used as marker wavelengths. Concentra-tions of 688, 300, 204, 460, and 2440 ppm of Pb, S, Cd,Cr, and Cl, respectively, were detected in the 5%spent clay in the binder. The limits of detection ofPb, S, Cd, Na, Ca, Cr, and Cl were estimated to be6.7, 17.2, 6.5, 2.8, 1.8, 5.1, and 14.9 ppm, respectively,from the calibration curve for each element. The cal-ibration curve for each element was made by exter-nally adding the respective metallic compounds tothe same matrix (organic spent clay). In order to con-firm the reliability of our system, the concentrationsof the reported elements detected using our systemwere compared to the ones obtained with inductivelycoupled plasma emission spectroscopy and found tobe in good agreement.

2. Experimental Setup

A. DP-LIBS/ICP Setup

Figure 1(a) depicts the schematic diagram of our DP-LIBS system applied for the analysis of spent claysamples. Two Q-switched pulsed Nd:YAG lasers(Big Sky lasers) are used in the DP-LIBS system,of which the laser that delivers the fourth harmonic(266 nm) serves as a laser 1 and the one that deliversfundamental (1064 nm) serves as laser 2, with the

1710 APPLIED OPTICS / Vol. 53, No. 8 / 10 March 2014

former triggering the latter. The pulse duration ofboth lasers is around 6 ns, and the pulse repletionrate is 20 Hz. The energy of laser 1 is 20 mJ∕pulse,and that of laser 2 is 100 mJ∕pulse. The two laserbeams were made to overlap and focused on the sam-ple surface, in such a way that both the laser spotswere on top of the other. The laser pulse energy wasmeasured with an energy meter (Ophir Model 300,Ophir Optics, Inc., North Andover, Massachusetts,USA). Once the LIBS signal was optimized withoptical alignments, the delay between the laserswas adjusted at submicrometer level to optimizethe DP-LIBS signal.

The DP-LIBS signal was collected by the fiber bun-dle assembly (100 μm core), and fed into the sevenchannels of the DP-LIBS spectrometer (Ocean OpticsLIBS 2500�). The distance between the head of thefiber optic cable and the plasma emission was kept at10 mm. A 10 mm focal length lens was installed atthe front of the fiber optic cable to collect the plasmaemission. Each channel is a spectrograph CCD cam-era assembly optimized for a particular wavelengthregion. The flash lamps and the Q-switches of bothlasers were externally controlled by the DP-LIBSsystem in the sequence depicted in Fig. 1(a). Laser1 produces plasma sparks at the target surface,and after a suitable time delay, the second laser pulse

is fired to ionize the neutral and the detached elec-trons generated by the first laser, which could be ac-celerated by the electric field of the second laserbeam to ionize further neutrals. We observed thatby irradiance of the second laser beam that passesthrough the plume and impinges on the targetsurface enhances the DP-LIBS-signal many fold. Itwas noticed that the DP-LIBS emission was strongerif it was observed around a 45° angle to the laserpulse. The average spectra due to 10 laser pulseswere registered for each data point. It was noted thataveraging the 10 laser pulse spectra tended to reducethe noise to a great extent when compared with thesingle laser pulse spectrum of the sample; i.e., theaveraged spectrum should have

p10 ∼ 3 times less

noise than the nonaveraged spectrum. In our work,the concentrations of different chemical elementspresent in spent clay samples were also measuredwith a calibrated ICP spectrometer to verify the val-idity of the results achieved with our DP-LIBS sys-tem. In our work, we applied the ICP spectrometermodel ICP-OES (spectro-ciros provided by SpectroAnalytical, Germany). Clay is dried and crushed inpowder form. Thus we injected the powder quantitiesin the center of the ICP in the presence of Ar gas.

During the optimization process of the dual-pulsedLIBS system, we tried many pairs of fundamental

Fig. 1. (a) Schematic diagram of dual-pulsed LIBS system. (b) Photographs of the insoluble organic powder, and pressed and sinteredpellets. The pellets have a diameter of 35 mm.

10 March 2014 / Vol. 53, No. 8 / APPLIED OPTICS 1711

and harmonics of Nd:YAG for laser 1 and laser 2 andwe found 266 and 1064 nm yielded the best S/N. Thedelay between the excitation and detection wasoptimized, and it was 700 ns. The delay betweenthe two lasers was optimized, and it was between 9.8and 10.2 ms. The laser pulse energy we used was20 mJ∕pulse for laser 1 (266 nm) and 100 mJ∕pulsewas for laser 2 (1064 nm), and this also shows an op-timum value. In our previous works on a sample thatgives out lot of debris, we used a vacuum nozzle toclear off the debris from the sample surface. In thesecases, with or without a vacuum nozzle made a lot ofdifference to the LIBS signal. But in this clay sample,the use of a vacuum nozzle to clear up the debris didnot make any difference to the signal level. Also thetarget was rotated to obtain a fresh surface and keepthe plume continuously clear.

B. Preparation of Samples and Calibration

The chemical composition of the spent clay is as fol-lows: 44.2% of SiO2, 39.7% of Al2O3:, 1.39% of TiO2,0.13% of Fe2O3, 0.08% of S, 0.002% of CrO, 0.03% ofCl, 0.013% of Na2O, 0.05% of CdO, 0.013% of Pb, and0.034% of P2O5. Samples of spent clay were collectedand pressed in the form of pellets by adding a certainamount of sodium chloride (free from other impu-rities) as a binder material, and all the spent clay pel-lets were of the same thickness as shown in Fig. 1(b).The die filled with clay was then annealed in theoven at a temperature of 60°C to make the targetsample hard to minimize pitting during the laserablation process. The pellets were then polished toobtain a smooth surface necessary for operation ofuniform plasma sparks for LIBS analysis. In orderto avoid the crater building on the sample surfaceby repeated high-energy laser beams, sampleswere kept on the X–Y translator system that wascontrolled by computer software during the data col-lection. In order to calibrate our DP-LIBS system forquantitative analysis, the compounds of the ele-ments (Pb, Na, Ca, S, Cr, and Cd—all from SigmaAldrich) were externally added to the sample matrix(spent clay) in various known concentrations (2000,5000, 12,000, 15,000, and 50,000 ppm), and the pelletwas made for each concentration as mentionedbefore.

3. Results and Discussion

A. Optical Emission Spectra

Figure 2 depicts the DP-LIBS spectrum of the or-ganic waste samples recorded in the 200–620 nmand 630–970 nm wavelength regions, where we cansee the presence of sulphur, chromium, cadmium,lead, sodium, calcium, potassium, and chlorine in ad-dition to other minor additives and impurities. Ourexperience shows that appearance/disappearanceor any variation from the NIST standard databaseis vastly specimen dependent. For example, for a par-ticular specimen, a particular line may be quite high,but the same line for some other specimen may be

completely absent irrespective of the intensities re-ported in the NIST database. This is because ofthe self-absorption of the atomic emission lines bythe specimen matrix itself. In our case, spent clayis very complex, and naturally self-absorption of eventhe prominent lines of Na II and Cl II, for example, isabsent. All the major atomic emission lines chosenfrom Fig. 2 for the quantification are listed in Table 1along with their atomic transitions, the correlationfactor (R2) values, slopes, the standard deviation(SD) of their respective calibration curve, the LODof DP-LIBS, and the maximum permissible safe limit(MPSL) according to the Saudi Arabia regulation. Inaddition to the atomic emission lines listed in Table 1,Si I, Sn I, Zn I, Fe I, Cr I, and Ba I transactions arealso recorded in Fig. 2. We also notice some of theprominent lines of the observed elements are absentdue the self-absorption effect explained before. InDP-LIBS, a second laser after an optimum time delaycan enhance the DP-LIBS signal, which helped us todetect the weak emission lines of lead, carbon, chro-mium, and sulphur in the spent clay organic sam-ples. It is worth mentioning that the line emissionintensity in DP-LIBS depends on the total ablatedmass by dual-pulsed lasers, the number density ofthe emitting species, the plasma temperature, and

Fig. 2. DP-LIBS spectra showing different chemical elementspresent in the organic sample in the 200–620 nm and 630–970 nmwavelength regions.

1712 APPLIED OPTICS / Vol. 53, No. 8 / 10 March 2014

the degree of ionization of plasma. These parametersdepend on the laser wavelength and the energy of theincident laser [16]. The selection of the marker wave-length of a particular element is based on two crite-ria: first, it should have a measurable intensity, withlinear increase of the intensity corresponding withthe concentration in the same spent clay matrixfor a reliable calibration, and second, the transitionlines should be reasonably isolated and have mini-mal interference from other strong emissions forquantitative analysis.

B. Calibration of our DP-LIBS System

The typical evolution of the DP-LIBS signal with in-crease in concentration of the chemical element (Pb,Na, and Cd) present in the calibrated sample is de-picted in Figs. 3(a)–3(c) for various concentrations,i.e., 2000, 5000, 12,000, 15,000, and 50,000 ppm. Themarkerwavelengths for theabove elements are504.2,568.8, and 643.8 nm, respectively. These lines wereselected for calibration and analysis as they are freefrom overlapping and interference from other chemi-cal elements and have a minimal self-absorption ef-fect. These factors can affect the overall accuracy ofour DP-LIBS system. In order to reduce the residualnoise due to fluctuation in laser power, the spectrawere averaged for 10 laser shots for each data point.

The calibration curves for lead, sodium, calcium,sulfur, chromium, cadmium, and chlorine are de-picted in Fig. 4. As is clear from Fig. 4, with the in-crease in the concentration of Pb, Na, Ca, S, Cr, Cd,and Cl, the DP-LIBS signal level increases and theDP-LIBS signal intensity shows a linear dependenceon the concentration of each chemical elementpresent in the samples. The error bars shown in Fig. 4denote the variation of DP-LIBS signal intensity andare based on the replicate of 10 measurements. A lin-ear regression approach was used to fit the appropri-ate trend lines to DP-LIBS data recorded. Allcalibration curves for the data point demonstratedgood linearity fitting with correlation coefficients,R2 values that are listed in Table 1 and depictedin Fig. 4. The slopes of the above-mentioned plotswere also listed in Table 1. The atomic emission linesat 469.41, 643.8, 568.8, 443.56, 520.8, and 928.8 nm

have maximum intensities of 300, 204, 1140, 336,460, and 2440 arbitrary units for 50,000 ppm S,Cd, Na, Ca, Cr, and Cl concentration, respectively,while the ionic emission line at 504.2 nm has maxi-mum intensity of 688 arbitrary units for 50,000 ppmPb concentration as depicted in Fig. 1. In Fig. 4, wefocused on the low concentration values of differentelements to obtain the linearity with DP-LIBS sig-nals, which is very important, rather than at highconcentrations. This emission line shows a weakerDP-LIBS signal due to the reabsorption by otherchemical elements present in the spent clay samples.After establishing the calibration curves as depictedin Fig 4, the concentration of each chemical elementin our samples was estimated, which is presented inTable 2. The five specified concentrations are used toshow the increase of LIBS signal with the spiked upconcentrations of respective elements in Fig. 3. Infact we have done 15 spiked up samples from 50to 50,000 ppm of each element and used these datafor the calibration curve.

C. Limit of Detection

The LOD of a detection system is the lowest concen-tration that can be detected with the system and isalso used to determine the figure of merit of any ana-lytical instrument. The LOD basically indicates thecapability of each individual DP-LIBS system andcan be estimated using the following equation [6]:

LOD � 3σB∕S; (1)

where σB is the SD of the noise level and S is the cal-ibration sensitivity (slope of the calibration curve),which is the ratio of the line intensity to the concen-tration. The LOD, as defined by Eq. (1), depends onthe residual noise. The SD of the mean value was cal-culated from the DP-LIBS measurements data. TheLOD is then determined by using the SD of this dataand the slope of the calibration curve. The LOD isabout 0.00067%, which is equivalent to 6.7 ppm forPb, which is much higher than the safe permissiblelimit of Pb in solid materials. By using the signal-to-noise method, the peak-to-peak noise aroundthe analyte retention time is measured, and

Table 1. Elements Detected in Spent Clay Collected from Saudi Companies and their Limit of Detection of DP-LIBS and MaximumPermissible Safe Limit

ElementsDetectedin SpentClay

Wavelength(nm) Configuration

EinsteinCoefficientAji �s−1� R2 Slope

StandardDeviation (SD)

DP-LIBS

Limit ofDetection(LOD) ofDP-LIBS(ppm)

MaximumPermissibleSafe Limit(MPSL)

(ppm) [24]

Pb II 504.2 6s27p2P01∕2 → 6s27d2D3∕2 — 0.993 0.011 0.032 5.8 0.5

Pb I 405.8 6s26p23P2 − 6s26p�2P01∕2�7s3P0 8.9 × 107 0.989 0.0098 0.033 6.7 0.5

S I 469.41 3s23p3�4S0�4s − 3s23p3�4S0�5p 8.37 × 105 0.994 0.0057 0.049 17.2 800 sulfateCd I 643.8 4d105s5p − 4d105s5d 5.9 × 107 0.996 0.0095 0.031 6.5 —

Na I 568.8 2p63p − 2p64d 1.21 × 107 0.992 0.0181 0.026 2.8 1000Ca I 443.56 3p64s2 − 3p64s4p 2.18 × 108 0.996 0.032 0.029 1.8 —

Cr I 520.8 3d5�6S�4s − 3d4�5D�4s4p�3P0� 1.48 × 108 0.991 0.0107 0.027 5.1 1Cl I 928.8 3s23p4�3P�4s − 3s23p4�3P�4p 3.9 × 106 0.990 0.0043 0.032 14.9 0.1

10 March 2014 / Vol. 53, No. 8 / APPLIED OPTICS 1713

subsequently, the concentration of the analyte thatwould yield a signal equal to a certain value of S/Nis estimated. The noise magnitude is measured by anauto-integrator of the instrument. A S/N is generallyaccepted for estimating LOD, and S/N of 10 laserpulses is used for estimating LOD. This method iscommonly applied to analytical methods that exhibitbaseline noise.

D. Precision of DP-LIBS Results

We averaged the laser pulses to stabilize the vari-ance of DP-LIBS signal intensity, and the relativestandard deviation (RSD) was decreased by increas-ing the number of laser pulses. The precision ofan analytical instrument is usually estimated withthe RSD as follows [2]:

RSD% � �Standard deviation∕mean� � 100: (2)

For spent clay samples, the precision of DP-LIBSmeasurements depends on the reproducibility oflaser pulses and the complexity and homogeneityof the sample. Typical values of RSD for DP-LIBSare in the range of 1%–5% [6]. The same methodwas applied for materials analysis reported inRef. [31], and the precision was found to be 2%–4%.It is worth mentioning that the RSDs were 3.2% forPb, 2.7% for Cr, 3.1% for Cd, and 4.9% for S. In thisstudy, it was found that the concentrations of some ofthe toxic chemical elements detected, such as sulfurand chlorine, exceed the MPSL set by the Occupa-tional Safety and Saudi Environmental RegulatoryAgency and Health Administration (OSHA) [31].The concentrations of sulfur and chlorine measuredwith our DP-LIBS system are (300� 15) and�2440� 122� ppm, respectively, which are higheras compared with the standard permissible limit of1 ppm for Cr and 0.1 ppm for Cl set by the Environ-mental Protection Authority (EPA) and other regula-tory authorities. Meanwhile, the concentrationvalues of cadmium, chromium, and lead are higherthan the permissible safe limit. The limits of detec-tion of toxic metals such as Cr, Cd, S, Pb, and Cl were5.0, 6.5, 17.2, 6.7, and 14.8 ppm.

E. Accuracy of DP-LIBS Results

In order to verify the accuracy of our DP-LIBS sys-tem, a standard ICP spectrometer was also appliedto detect sulphur, cadmium, chromium, and lead inour samples. Comparison of the values obtained bythe two methods—our DP-LIBS setup and ICPresults—is presented in Table 2, which indicatesgood agreement within the 0.02% to 0.74% errorlimit. These samples have different concentrationsof trace elements, which could be due to the raw ma-terials being used for production of spent clay. Theaccuracy of an analytical method is defined as themeasured experimental values that are true or closeto the accepted “actual” values. In general, the accu-racy is the percentage difference between the mea-sured experimental values with DP-LIBS and themeasured actual values with a standard methodsuch as ICP. We present the accuracy of the DP-LIBSsystem relative to the standard ICP measurementsin terms of relative accuracy (RA). It is a measureof the closeness of our measurements with respectto the standard (here ICP measurements). The RAis calculated as follows [32]:

Fig. 3. DP-LIBS spectrum of the insoluble organic substance atfive different concentrations of (a) lead at 504.2 nm ionic transi-tion, (b) sodium at 568.8 nm atomic transition, and (c) cadmiumat 643.8 nm atomic transition in the organic substance as markers2000, 5000, 12,000, 15,000, and 50,000 ppm of Pb, Na, and Cd,respectively, by weight in organic substances.

1714 APPLIED OPTICS / Vol. 53, No. 8 / 10 March 2014

RAjdj � SD × t0.975

���

np

M× 100; (3)

where d is the difference between the DP-LIBSmeasurement and the ICP (standard method). SD

is the standard deviation of the DP-LIBS measure-ment, M is the measurement from the standardmethod, n is the number of measurements, andt0.975 is the t value at 2.5% error confidence. (The tvalue is defined as a student’s t-test as hypothesis

Fig. 4. Calibration curves for the detection of lead, sodium, calcium, sulphur, chromium, cadmium, and chlorine using ionic and atomictransition lines at Pb II 504.2 nm, Pb I 405.8 nm, Na I 568.8 nm, Ca I 443.56 nm, S I 469.41 nm, Cr I 520.8 nm, and Cd I 643.8 nm and928.8 nm, respectively, as markers.

10 March 2014 / Vol. 53, No. 8 / APPLIED OPTICS 1715

tailed tests. In statistics, the t-statistic is a ratio ofthe departure of an estimated parameter from its no-tional value and its standard error. It is used in hy-pothesis testing, for example, in the student’s t-test.)In order to achieve good accuracy and precision in theDP-LIBS measurements, the following precautionswere undertaken. Few laser shots were fired to cleanthe sample surface prior to actual measurements.Fluctuation of the laser pulse was also consideredduring the performance of the experiment for better-ment of precision and accuracy. In order to overcomethis problem, sufficient warm-up time was givenfor the laser to stabilize. Also the DP-LIBS spectrawere averaged for 10 laser pulses. The variationin DP-LIBS signal intensity was stabilized by aver-aging for different laser shots, while the RSD de-creases with an increasing of number of lasershots. The RA of different elements present in thespent clay for each spiked up concentration is listedin Table 2 along with the concentrations of DP-LIBSand ICP.

The continuous support by the Laser ResearchGroups at NILES and Physics Department, KingFahd University of Petroleum and Minerals(KFUPM), through DSR project no. RG1201-1/2 isgratefully acknowledged.

References1. J. O. Duruibe, M. O. C. Ogwuegbu, and J. N. Egwurugwu,

“Heavy metal pollution and human biotoxic effects,” J. Phys.Sci. 2, 112–118 (2007).

2. R. Fantoni, L. Caneve, F. Colao, L. Fornarini, and V. Lazic,“Methodologies for laboratory laser induced breakdown spec-troscopy semi-quantitative and quantitative analysis—areview,” Spectrochim. Acta B 63, 1097–1108 (2008).

3. E. H. Evans, J. A. Day, C. Palmer, and C. M. Smith, “Advancesin atomic spectrometry and related techniques,” J. Anal. At.Spectrom. 25, 760–784 (2010).

4. O. T. Butler, W. R. L. Cairns, J. M. Cook, and C. M. Davidson,“Atomic spectrometry update. Environmental analysis,”J. Anal. At. Spectrom. 25, 103–141 (2010).

5. A. Ferrero and J. J. Laserna, “A theoretical study of atmos-pheric propagation of laser and return light for stand-off laserinduced breakdown spectroscopy purposes,” Spectrochim.Acta B 63, 305–311 (2008).

6. A. Miziolek, V. Palleschi, and I. Schechter, Laser InducedBreakdown Spectroscopy (LIBS): Fundamental and Applica-tions (Cambridge University, 2006).

7. M. Ciucci, M. Corsi, V. Palleschi, S. Rastelli, A. Salvetti, and E.Tognoni, “New procedure for quantitative elemental analysisby laser-induced plasma spectroscopy,” Appl. Spectrosc. 53,960–964 (1999).

8. J. Cunat, F. J. Fortes, and J. J. Lasagna, “Real time and in situdetermination of lead in road sediments using a man-portablelaser-induced breakdown spectroscopy analyzer,” Anal. Chim.Acta 633, 38–42 (2009).

9. M. N. Shaikh, S. Hafeez, and M. A. Mohammed, “Comparisonof zinc and plasma parameters produced by laser-ablation,”Spectrochim. Acta B 62, 1311–1320 (2007).

10. S. Laville, M. Sabsabi, and F. R. Doucet, “Multi-elementalanalysis of solidified mineral melt samples by laser-inducedbreakdown spectroscopy (LIBS) coupled with a linear multi-variate calibration,” Spectrochim. Acta B 62, 1557–1566(2007).

11. NIST Atomic spectra database, http://www.nist.gov/physlab/data/asd.cfm.

12. E. Tognoni, V. Palleschi, M. Corsi, and G. Cristoforetti,“Quantitative micro-analysis by laser-induced breakdown

Tab

le2.

Conce

ntrationofC

hem

ical

Elemen

tsDetec

tedin

Inso

luble

Organ

icSubstan

cesColle

cted

from

Loca

lSau

diIndustries

andComparisonofD

P-LIBSwithStandardConve

ntional

Tec

hnique(ICP)a

Hostof

Was

teMaterials

Sulphu

rChlorine

Cad

mium

Chr

omium

Lea

d

Spe

ntClay(SC)+

Chem

ical

Elemen

tDP-LIB

S(ppm

)IC

P(ppm

)RA

(%)

DP-LIB

S(ppm

)IC

P(ppm

)RA

(%)

DP-LIB

S(ppm

)IC

P(ppm

)RA

(%)

DP-LIB

S(ppm

)IC

P(ppm

)RA

(%)

DP-LIB

S(ppm

)IC

P(ppm

)RA

(%)

SC�

2000

ppm

9.5

150.08

5.5

70.15

3.6

50.03

712

0.46

2028

0.54

SC�

5000

ppm

2127

0.04

810

0.02

911

0.04

2531

0.67

5259

0.43

SC�

12;000

ppm

5061

0.02

1821

0.63

2125

0.02

5556

0.59

142

148

0.32

SC�

15;000

ppm

6573

0.03

2325

0.17

2731

0.03

7882

0.48

180

191

0.25

SC�

50;000

ppm

300

302

0.05

2440

2453

0.74

204

207

0.05

460

472

0.31

688

695

0.31

a RA,relative

accu

racy;SC,sp

entclay.

1716 APPLIED OPTICS / Vol. 53, No. 8 / 10 March 2014

spectroscopy: a review of the experimental approaches,” Spec-trochim. Acta B 57, 1115–1130 (2002).

13. J. Uebbing, J. Brust, W. Sdorra, F. Leis, and K. Niemax,“Reheating of a laser-produced plasma by a second pulselaser,” Appl. Spectrosc. 45, 1419–1423 (1991).

14. C. Gautier, P. Fichet, D. Menut, J.-L. Lacour, D. L’Hermite, andJ. Dubessy, “Quantification of the intensity enhancements forthe double-pulse laser-induced breakdown spectroscopy in theorthogonal beam geometry,” Spectrochim. Acta B 60, 265–276(2005).

15. D. N. Stratis, K. L. Eland, and S. M. Angel, “Effect of pulsedelay time on a pre-ablation dual-pulse LIBS plasma,” Appl.Spectrosc. 55, 1297–1303 (2001).

16. R. E. Neuhauser, U. Pannev, R. Niessner, G. A. Petrucci, P.Cavalli, and N. Omenetto, “On line and in-situ detection oflead aerosols by plasma-spectroscopy and laser-excited atomicfluorescence spectroscopy,” Anal. Chim. Acta 346, 37–48(1997).

17. G. S. Senesi, M. Dell’Aglio, R. Gaudiuso, A. De Giacomo,C. Zaccone, O. De Pascale, T. M. Miano, and M. Capitelli,“Heavy metal concentrations in soils as determined bylaser-induced breakdown spectroscopy (LIBS), with specialemphasis on chromium,” Environ. Res. 109, 413–420(2009).

18. F. Capitelli, F. Colao, M. R. Provenzano, R. Fantoni, G.Brunetti, and N. Senesi, “Determination of heavy metals insoils by laser induced breakdown spectroscopy,” Geoderma106, 45–62 (2002).

19. M. A. Gondal, Z. Ahmed, M. M. Nasr, and Z. H. Yamani,“Determination of trace elements in volcanic rock samplescollected from cenozoic lava eruption sites using LIBS,”J. Environ. Sci. Health A 44, 528–535 (2009).

20. A. F. M. Y. Haider, M. A. Rony, R. S. Lubna, and K. M. Abedin,“Detection of multiple elements in coal samples from Bangla-desh by laser-induced breakdown spectroscopy,” Opt. LaserTechnol. 43, 1405–1410 (2011).

21. L. Peng, D. Sun, M. Su, J. Han, and C. Dong, “Rapid analysison the heavy metal content of spent zinc-manganese batteriesby laser induced breakdown spectroscopy,” Opt. LaserTechnol. 44, 2469–2475 (2012).

22. A. A. I. Khalil, “Spectroscopic studies of UV lead plasmasproduced by single and double-pulse laser excitation,” LaserPhys. 23, 015701 (2013).

23. B. Salle, J. Lacour, E. Vors, P. Fichet, S. Maurice, D. A.Cremers, and R. C. Wiens, “Laser-induced breakdown spec-troscopy for Mars surface analysis: capabilities at stand-offdistances and detection of chlorine and sulfur elements,” Spec-trochim. Acta B 59, 1413–1422 (2004).

24. M. A. Gondal, A. Dastageer, M. Maslehuddin, A. J. Alnehmi,and O. S. B. Al-amoudi, “Detection of sulphur in the reinforcedconcrete structures using a dual pulsed LIBS system,” Opt.Laser Technol. 44, 566–571 (2012).

25. R. Sattmann, V. Sturm, and R. Noll, “Laser-induced break-down spectroscopy of steel samples using multiple Q-switchNd: YAG laser pulses,” J. Phys. D 28, 2181–2187 (1995).

26. M. Corsi, G. Cristoforetti, M. Giuffrida, M. Hidalgo, S.Legnaioli, V. Palleschi, A. Salvetti, E. Tognoni, and C.Vallebona, “Three-dimensional analysis of laser induced plas-mas in single and double pulse configuration,” Spectrochim.Acta B 59, 723–735 (2004).

27. F. Colao, V. Lazic, R. Fantoni, and S. Pershin, “A comparison ofsingle and double pulse laser-induced breakdown spectros-copy of aluminum samples,” Spectrochim. Acta B 57, 1167–1179 (2002).

28. V. Hohreiter, J. E. Carranza, and D. W. Hahn, “Temporalanalysis of laser-induced plasma properties as related tolaser-induced breakdown spectroscopy,” Spectrochim. ActaB 59, 327–333 (2004).

29. D. L. Wiggins, C. T. Raynor, and J. A. Johnson III, “Evidence ofinverse Bremsstrahlung in laser enhanced laser-inducedplasma,” Phys. Plasmas 17, 103303 (2010).

30. L. Torrisi, S. Gammino, A. Picciotto, D. Margarone, L. Laska,J. Krasa, K. Rohlena, and J. Wolowski, “Method for the calcu-lation of electrical field in laser-generated plasma for ionstream production,” Rev. Sci. Instrum. 77, 03B708 (2006).

31. Environmental Regulation Standards for Saudi Industries setby Royal Commission Al-Jubail, Saudi Arabia, http://www.rcjy.gov.sa.

32. M. G. Natrellla, Experimental Statistics, NBS Handbook91 (National Institute of Standards and Technology, 1963).

10 March 2014 / Vol. 53, No. 8 / APPLIED OPTICS 1717