An Approach of Laser-Induced Breakdown Spectroscopy to Detect Toxic Metals in Crushed Ice Ball
This paper deals the application of laser-induced breakdown spectroscopy (LIBS) to toxic metals used as pigment in crushed ice-ball samples. The present work highlights the advantages of LIBS as in situ, real-time analytical tool for rapid detection of toxic or heavy metals like lead (Pb) and chromium (Cr) and non toxic elements like carbon (C), nitrogen (N), magnesium (Mg), calcium (Ca), sodium (Na), and potassium (K) in crushed ice-ball of different colors (red, green, yellow, pale yellow, and orange) collected from five different areas, with minimal sample preparation. For rapid surveillance of toxic metals we have used multivariate analysis, that is, principal component analysis (PCA) with the LIBS spectral data of ice-ball samples. This study suggests that LIBS coupled with PCA may be an instant diagnostic tool for identification and classification of adulterated and nonadulterated samples.
Many studies have shown that our acceptance of the food is difficult where the color of a food does not meet our expectations and aesthetics quality [1–4]. Thus, colors play very important role in our acceptance of food. Ice balls are prepared from crushed ice topped with sweetened colored syrup and served in the form of a ball in stick. Two types of food colors are frequently used in food and ice bar products: (i) natural or bio colors like carotenoids, flavnoids, anthocyanidins, chlorophyll, betalain, curcumin, and so forth, which are extracted from plants, and (ii) synthetic colors like Sunset Yellow FCF, Tartrazine, Ponceau 4R, Carmoisine, Erythrosine, Brilliant Blue FCF, Fast Green FCF, and Indigocarmine. The maximum permissible level of synthetic food colors that can be added either single or in mix proportion is 100 ppm . Instead of the previous permitted colors, some street vendor’s and small-scale ice-ball makers use low-grade nonpermitted colors like clothes color, copper sulphate, vermilion (mercury sulphide), lead chromate, lead sulphate, and so forth, for gaining undue profits which leads to serious health problems and potential dangers, like cancer, lead poisoning, embryo toxicity, teratogenicity, dermatitis, and eczema from repeated exposures [5–8].
Adulteration is the mixing of inferior quality material or inferior substance to the superior product, which reduces the nature, quality and originality in taste, color, odor, and nutritional value causing ill effects on the health of the consumers. As a result of these malpractices, the ultimate victim is a consumer, who innocently takes adulterated foods, and, leads to serious health problems such as allergy, gastrointestinal diseases [5, 9].
Thus, we require to assess the current scenario of usage of toxic food colorants in colored food commodities like crushed or uncrushed ice ball and ice cream bars. Therefore, there is a need for a technique, which has real-time, in-situ and quick detection capability, multielemental analysis of sample in any phase (solid, liquid, and gas) without or minimal sample preparation. LIBS is such an advanced technique having the said features and that can do qualitative and quantitative determination of toxic elements and differentiate from harmless elements by providing spectral signatures that enable the unique identification of adulterated foods [10–19].
Considering the previously mentioned features in the present study, we have used LIBS technique to detect and compare the toxic elements used in colored crushed ice ball. LIBS is an emission spectroscopy in which we compare the intensities of atomic lines of different species present in the sample under local thermodynamic equilibrium (LTE) . Therefore, before using the intensity of the spectral line for analysis purpose we have determined the plasma temperature and electron density in the laser-induced plasma to verify the existence of LTE in laser-induced plasma plume.
Nowadays, multivariate analysis of LIBS data is used for instant identification and classification of variety of samples [10, 11, 15]. Principal component analysis (PCA) is a linear technique used to map multidimensional data onto lower dimension with minimal loss of variance. We have also used PCA for rapid differentiation of adulterated and nonadulterated ice ball samples.
2. Material and Method
2.1. Sample Collection
Samples of crushed ice ball of different colors were collected from street vendors randomly from five local areas of Allahabad city. Samples from the area are coded as A, B, C, D, and E, respectively. From each area we have taken five different colored samples of crushed ice balls like dark yellow, pale yellow, orange, green, and red and are shown in Figure 1. The samples are kept in deep-freezer for 4 hours at −20°C in 100 mL glass containers that provides solid and hardened matrix. These colored ice samples were kept out from glass container and placed in glass petridish as shown in Figure 2. LIBS spectra of the sample placed in petridish are directly recorded to do analysis by LIBS technique for complete profile of elements present in the samples.
2.2. Experimental Setup for LIBS
The schematic diagram of experimental setup of the LIBS is shown in Figure 2. The Neodymium:Yttrium-Aluminum-Garnet (Nd:YAG) laser (Continuum Surelite III-10) was used for the present experiment having repetition rate of 1 Hz to 10 Hz, pulse width, that is, full width at half maximum of 4 ns and maximum energy of 425 mJ at 532 nm. Laser beam of 9 mm beam waist was focused on the ice bar sample kept on a moving sample stage using a planoconvex quartz lens of 30 cm focal length to get laser-induced plasma on the surface of the sample. The light emitted by the laser-induced plasma was collected by a collecting lens fitted at one end of the optical fiber which inclined at ~45° with respect to laser beam. The other end of the optical fiber is connected to the entrance slit of a grating spectrometer (Ocean Optics, LIBS 2000+) equipped with a charge-coupled device (CCD). The signals were analyzed using OOI LIBS 2000+ software. To avoid the formation of crater at the surface of the ice bar, the samples were placed on moving sample stage so that every laser pulse was targeted at a fresh location on the sample surface. The averaged LIBS spectra for 10 laser shots of different color of ice bar samples were recorded to get best signal to background ratio at 80 mJ energy, 10 Hz repetition rate, and 1.5 μs gate delay with a spectral resolution of 0.1 nm for 200–500 and a spectral resolution of 0.75 nm in the spectral range of 200–900 nm.
3. Results and Discussion
LIBS spectra of the different colored samples (ice balls) collected from different area were recorded in the spectral range of 200 nm–900 nm and the elements in LIBS spectra were identified using the NIST spectral database . A typical LIBS spectra of yellow, pale yellow and orange colored ice ball samples of all five region, that is, A, B, C, D, and E are shown in Figure 3. LIBS spectra of ice balls show the presence of the persistent spectral lines of carbon (C), nitrogen (N), magnesium (Mg), sodium (Na), potassium (K) calcium (Ca), lead (Pb), and chromium (Cr). Yellow, pale yellow, and orange colored ice ball samples basically made from yellow color and the presence of lead and chromium reflect that some of the greedy shopkeepers might have used lead chromate (toxic food color) which is cheap yellow color for providing yellow appearance to different food products. The LIBS spectra of red and green colored sample shown in Figure 4 clearly show the presence of spectral lines of major elements like carbon (C), nitrogen (N), magnesium (Mg), calcium (Ca), sodium (Na), and potassium (K), but the spectral lines Pb and Cr are absent.
We have also measured the intensities of spectral lines of lead and chromium in pale yellow, yellow, and orange colored ice ball samples and show them in terms of bar diagram in Figure 5. But before using the intensities of the spectral lines of the elements to correlate its concentration in the samples, we have to verify the assumptions of stoichiometric ablation, local thermal equilibrium, and optically thin plasma for laser induced plasma of different samples [13, 14, 17, 18]. Verification of these assumptions for the laser-induced plasma in the present experiment is discussed in the following sections.
3.1. Stoichiometric Ablation
To produce the laser-induced plasma on the surface of the ice-ball we have focused the laser beam on its surface using converging lens of 30 cm focal length. At the focal point of the sample, the spot size “” is calculated using the following formula:
Thus, the calculated power density at focal spot is ≈2 × 1012 W·cm−2 which satisfies the stoichiometric ablation condition .
3.2. Optically Thin Plasma
The condition of optically thin plasma means that the radiation emitted from an excited atom in the plasma should not be reabsorbed (self-absorption effect) by another atom in a lower energy state. Optically, thin plasma can be verified by comparing the intensity ratio of two interference free emission lines of a species, having close value of upper energy levels to the product of the ratio of their transition probabilities, their upper level degeneracies, and the inverse ratio of their wavelengths .
The intensity ratios of Ca 315.8/317.9 nm, Cr 357.8/359.3 nm, and Pb 363.9/368.3 nm are calculated from the LIBS spectra recorded for all five colors collected from all five regions and are tabulated in Table 1. The product of ratio of their transition probabilities, the ratio of their upper levels degeneracy, and the inverse ratio of their wavelengths are calculated using data from the literature and are also tabulated in Table 1. It is clear from Table 1 that is approximately equal to the corresponding , which clearly satisfy the condition of optically thin plasma.
3.3. Determination of Plasma Temperature and Electron Density for the Fulfillment of Local Thermal Equilibrium Condition
The plasma temperatures for different samples were determined using Boltzmann plot obtained from Boltzmann equation: where “” is the Boltzmann constant, “” is the partition function, “” is the transition probability, “” is the statistical weight for the upper level, “” is the excited level energy, “” is the temperature, and “” is a constant depending on experimental conditions. The Boltzmann plots were drawn using spectral lines of Ca I, Ca II, Cr I, and Pb I present in the LIBS spectra of all different colored samples collected from one area “” and are shown in Figure 6. It is clear from Figure 6 that the slopes of all lines are approximately parallel which show that the plasma temperatures are in LTE for all samples. Finally, the plasma temperatures for yellow ice ball collected from all five areas are calculated and are shown in Table 2. Similarly the Boltzmann plots for other samples collected from areas B, C, D, and E have been drawn and the slopes of these lines are also parallel. The condition of LTE was verified.
The other criteria for establishment of LTE condition is related to the electron density (Ne) given by where (cm−3), (K), and (eV) are electron density, plasma temperature, and the largest observed transition energy, respectively [24, 25].
The FWHM for Cr: 357.8 nm (yellow ice ball) and Ca: 422.6 nm (green ice ball) line is shown in Figure 7. where nm.
In the present paper, the Cr line at 357.8 nm and Ca lines at 422.6 nm are selected to measure the FWHM. A Lorentzian fit to the observed experimental data for these lines for yellow and green samples is presented in Figure 7. The calculated lower limit of electron density using (3) and experimentally estimated electron densities using (4) and (5) are tabulated in Table 2. It is clear from Table 2 that the plasma temperatures for yellow ice ball collected from all five areas are nearly the same and comes in the order of 104 Kelvin. The electron density (Ne) measured experimentally using (4) and (5) in plasma is greater to the lower limit of electron density calculated by (3) and this clearly reveals that plasma is in LTE.
After verifying the assumption of stoichiometric ablation, optically thin plasma, and LTE condition, one can use the intensities of spectral lines of the element to correlate its concentration in the samples. Figure 3 clearly shows the presence of atomic lines of Cr at (357.8, 359.3, 360.5) nm and Pb at (220.3, 363.8, 368.3) nm in the LIBS spectra of yellow, pale yellow, and orange samples collected from all five different area. We expect that the presence of these lines in the LIBS spectra of the previous samples may be due to the use of lead chromate (PbCrO4) for coloring of those food samples, but PbCrO4 is nonpermissible and banned toxic color and causes serious health hazards and may also cause cancer in the long run.
We have also measured the intensities of spectral lines of lead, chromium, and calcium in all samples and the results are shown in Figure 5. The intensity of the spectral line of Ca (317.9 nm) is almost the same in the LIBS spectra of all samples (Figure 5), which reflect that the concentration of Ca is almost the same in all samples. Therefore, this line is selected to get the spectral lines normalized intensity of Cr (359.9 nm) and Pb (368.3 nm) and the results are shown in Figure 5. It is clear from Figure 5 that the proportion of normalized intensity of lead and chromium and hence the proportion of concentration of Pb and Cr in yellow ice ball samples are equal which reflects that the same type of chemical (PbCrO4) might have been used in preparation of yellow ice ball samples. But in case of pale yellow and orange samples the proportion of Pb and Cr is not the same as the proportion of intensity of Pb and Cr and they are very much different in the LIBS spectra of these samples. Therefore, this result reveals that in addition to PbCrO4 some other salts of chromium may have been used for providing light yellow and orange color.
Recently, principal component analysis (PCA) of LIBS data has been used for rapid classification of adulterated and nonadulterated food materials [10, 11, 15]. Therefore, in the present paper, PCA has been also used for the identification and classification of different colored samples containing toxic pigments or elements. For the fulfillment of this purpose, we have prepared a dataset using LIBS spectra of yellow, pale yellow, orange, red, and green ice ball samples collected from five different places (A, B, C, D, and E) with 100 features (25 × 100 matrix). The library of LIBS data samples along with the Unscrambler PCA software (supplied by CAMO Software India Pvt. Ltd.) has been used in the present analysis [9, 10, 14]. When these LIBS data are used for classification we get the principal components PC1 (95%) and PC2 (4%) that explain the total variance (99%) among the dataset. Different plots are shown in Figure 8 that gives four parameters (i) scores: represents sample patterns and show sample differences and similarities (ii) loading: represents variable contribution and correlations, (iii) Influence plot: measures the distance from the projected sample, and (iv) variance plot: shows the percentage of total variance in the data. PCA score plot (Figure 8) clearly differentiates the samples containing lead and chromium in yellow, pale yellow, and orange ice ball samples of different regions from others. Thus, LIBS data with PCA enables us to rapidly classify and identify adulterated samples from nonadulterated samples.
The experimental results clearly reveal that the ice ball having yellow, pale yellow, and orange colors are adulterated by lead chromate, whereas red and green color ice ball samples are not adulterated. The results of the present study further demonstrate that LIBS coupled with PCA can be used as online powerful diagnostic tool for detection and classification of adulterated and nonadulterated food materials like ice balls, ice-cream, sweets, and so forth. In India, the consumption of ice balls by children and adult in summer season increases and thus calls for an appropriate fast technique to detect such adulteration which is met by LIBS.
Financial assistance from the BRNS, BARC, Mumbai (no. 2009/37/30/BRNS/2063), is gratefully acknowledged. Mr. Rahul Agrawal is grateful to Centre of Food Technology, University of Allahabad, for Financial Assistance. The authors are also thankful to Ms. Manju Tiwari, Centre of Food Technology, University of Allahabad, for providing help in sample collection.
M. Florian, H. Yamanaka, P. A. Carneiro, and M. V. B. Zanoni, “Determination of brilliant blue FCF in the presence and absence of erythrosine and quinoline yellow food colours by cathodic stripping voltammetry,” Food Additives and Contaminants, vol. 19, no. 9, pp. 803–809, 2002.View at: Publisher Site | Google Scholar
S. Dixit, R. C. Pandey, M. Das, and S. K. Khanna, “Food quality surveillance on colours in eatables sold in rural markets of Uttar Pradesh,” Journal of Food Science and Technology, vol. 32, pp. 373–376, 1995.View at: Google Scholar
M. B. Jacobs, “Coloring matters in foods,” in The Chemical Analysis of Food and Food Products, pp. 103–149, Robert E. Krieger, New York, NY, USA, 1973.View at: Google Scholar
Colouring matter, in Prevention of Food Adulteration Act of India-1954, pp. 002–116, Law Publishers, Allahabad, India, 2003.
S. Babu and I. S. Shenolikar, “Health and nutritional implications of food colours,” Indian Journal of Medical Research, vol. 102, pp. 245–249, 1995.View at: Google Scholar
G. Biswas, S. Sarkar, and T. K. Chatterjee, “Surveillance on artificial colours in food products in Calcutta and adjoining areas,” Journal of Food Science and Technology, vol. 31, pp. 66–67, 1994.View at: Google Scholar
R. V. Bhat and P. Mathur, “Changing scenario of food colours in India,” Current Science, vol. 74, no. 3, pp. 198–202, 1998.View at: Google Scholar
R. Agrawal, A. K. Pathak, G. K. Rai, and A. K. Rai, “Classification of milk of different origin using LIBS,” Asian Journal of Spectroscopy, special issue, pp. 141–146, 2010.View at: Google Scholar
A. Miziolek, W. V. Palleschi, and I. Schechter, Laser Induced Breakdown Spectroscopy: Fundamentals and Applications, Cambridge University Press, Cambridge, UK, 2006.
D. K. Chauhan, D. K. Tripathi, R. Agrawal, and A. K. Rai, “Detection of silicon in wheat (Triticum aestivum) using laser induced breakdown spectroscopy and phytolith analysis,” Journal of Research, SKUAST-J, vol. 10, no. 1, pp. 75–79, 2011.View at: Google Scholar
R. Kumar, A. K. Rai, D. Alamelu, and S. K. Aggarwal, “Monitoring of toxic elements present in sludge of industrial waste using CF-LIBS,” Environmental Monitoring and Assessment, vol. 185, no. 1, pp. 171–180, 2013.View at: Google Scholar
National Institute of Standards and Technology, “Electronic database,” http://physics.nist.gov/PhysRefData/ASD/lines_form.html.View at: Google Scholar
R. E. Russo, X. L. Mao, J. H. Yoo, and J. J. Gonzalez, “Laser ablation,” in Laser Induced Breakdown Spectroscopy, J. P. Singh and S. N. Thakur, Eds., pp. 49–82, Elsevier, Amsterdam, The Netherlands, 2007.View at: Google Scholar
H. R. Griem, Plasma Spectroscopy, McGraw-Hill, New York, NY, USA, 1964.
S. N. Thakur, “Atomic emission spectroscopy,” in Laser Induced Breakdown Spectroscopy, J. P. Singh and S. N. Thakur, Eds., pp. 23–48, Elsevier, Amsterdam, The Netherlands, 2007.View at: Google Scholar