Application of Inductively Coupled Plasma Atomic Emission Spectrometry in Forensic Science MARKO LALCHEVa, IONTCHO IONOVa AND NONKA DASKALOVAb aResearch Institute of Forensic Sciences and Criminology–Ministry of Interior, BG-1000 Sofia, P.O.Box 934, Bulgaria bInstitute of General and Inorganic Chemistry, Bulgarian Academy of Sciences, BG-1113 Sofia, Bulgaria Table 1 Specifications of the spectrometer, ICP source and ICP-AES was applied to the determination of trace amounts operating conditions of toxic elements in animal tissues, and also a large number of trace elements in bullet lead and silver.The analysis of these JY 38 (Jobin-Yvon) Monochromator— materials is important in forensic science applications. Mounting Czerny–Turner, focal length 1 m Grating Holographic, 2400 grooves mm-1 Keywords: Inductively coupled plasma atomic emission Wavelength range 170–700 nm (1st order) spectrometry; forensic science; toxic elements; animal tissues; Dispersion 0.38 nm mm-1 bullet lead; silver Entrance slit 0.02 mm Exit slit 0.04 mm Resultant spectral slit 15.2 pm The Research Institute of Forensic Science and Criminology Practical spectral band-width 15.6 pm at the Ministry of Interior of Bulgaria maintains a great Photomultiplier Hamamatsu TV, R 446 HA interest in techniques for the multi-element analysis of different Rf generator— PlasmaTherm, Model HFP 1500 D materials according to the requirements of modern forensic Frequency 27.12 MHz (±0.05%) science examination. Recently, a project was initiated for the Oscillator Crystal-controlled at 13.56 MHz determination of a number of elements in animal tissues, bullet Power output 0.5–1.5 kW lead and silver for the following reasons: the determination of toxic trace elements in animal tissues is interrelated with cases Nebulizer— Meinhard, concentric glass of acute toxic poisoning of people and animals; the elemental Pump— Peristaltic, ten-roller, analysis of bullet lead is important in connection with the Gilson Minipuls II (Gilson investigation of cases of murder or physical injury of people Medical Electronics, France) and animals with firearms; the wide application of silver and its increasing price on the world market has led to illegal Operating conditions— operations involving this metal.Incident power 1.0 kW Reflected power 10 W There are several papers in the literature describing the Outer argon flow rate 15 l min-1 application of various techniques for trace element determi- Carrier flow rate 0.5 l min -1 nations in materials of concern to forensic scientists.1–6 For Liquid uptake rate 1.3 l min-1 the solution of the above-mentioned problems, techniques are Transport efficiency of ICP 3% needed which combine the following features: (i) trace analyt- system ical methodology with multi-element capability; and (ii) the possibility of element determinations in a broad concentration range including detection limit levels.ICP-AES is a powerful Sample Digestion Procedure analytical technique combining all these features.7 The merits of ICP-AES for the analysis of biological and/or clinical Animal tissues materials have been discussed8 on the basis of real practical A 10 g sample was treated with HNO3–H2SO4 (30 ml) by a analytical tasks, whereby ICP-AES has been selected as the wet decomposition procedure. The wet digestion was carried method of choice. out in a closed system by using an autoclave with Teflon The purpose of this work was to show the possibilities of vessels [Perkin-Elmer (Norwalk, CT, USA), N3; working ICP-AES in the determination of trace elements in tissues, volume=0.12 l, pressure=50 bar and maximum temperature= bullet lead and pure silver. 160°C].A notable advantage of digestions carried out in closed systems is that volatilization losses can be minimized.8 EXPERIMENTAL The final volume of the sample solution was 100 ml.The solvent blank had a concentration of 183 mg ml-1 H2SO4. Instrumentation The matrix blank was a 10 g wet sample of the tissue in 100 ml The experiments were performed with the Jobin-Yvon of solution. The tissues contained the ‘normal values’ of element (Longjumeau, France) equipment specified in Table 1. concentrations. Bullet lead Reagents and Reference Solutions All reagents were of analytical-reagent grade (Merck, A 0.250 g sample was dissolved under heating with 1.5 ml of HNO3 (1+3) in the presence of tartaric acid (0.250 g).The Darmstadt, Germany) and doubly distilled water was used throughout. The stock solutions of the elements (1 mg ml-1) final volume of the sample solution was 25 ml. The solvent blank had a final concentration of 10 mg ml-1 tartaric acid were prepared from Merck Titrisol solutions. The reference solutions for the determination of the analytes were prepared and 27 mg ml-1 HNO3 . The matrix blank contained 10 mg ml-1 lead and the solvent blank.by precise matching of sample acidity and matrix content. Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12 (21–24) 21Silver understanding the ambivalentnature (i.e., toxic versus essential) of many elements by developing methods that determine A 0.250 g sample was dissolved under heating with 2.0 ml of minimum absolute amounts of trace elements in these materials HNO3 (1+1). The final volume of the sample was 25 ml. The in the most precise and accurate way.Detection limits and solvent blank was a 60 mg ml-1 solution of HNO3. The matrix accuracy of ICP-AES depend on both multiplicative and blank contained 10 mg ml-1 silver and the solvent blank. spectral interferences.8 The multiplicative and spectral interferences on the prominent lines19 were investigated. The net analyte signals in the RESULTS AND DISCUSSION presence of 183 mg ml-1 H2SO4 decreased by 30% in compari- ICP-AES is not free from matrix effects9–14 including acid son with the corresponding values in pure aqueous solution.matrix interferences.15–17 The latter was of significance since These results were in accordance with those reported by the samples were prepared for analysis by acid dissolution (see Chudinov et al.15 A 10 g biological sample digest at the same Sample Digestion Procedure). The multiplicative (non-spectral) acidity did not change the slope of the calibration graphs. The and additive (spectral) interference effects in the presence of influence of H2SO4 and the biological matrix was studied at a the above-mentioned matrices were studied.concentration level equal to that in the final sample solution. Multiplicative interferences are related to sensitivity changes No spectral interferences in the trace element determination in the analyte signals so that the signal-to-background ratio is were registered. Hence, the detection limits in solution were modified, i.e., the slope of the calibration graph is affected and calculated from the equation for pure aqueous solutions10 the accuracy may be reduced.This effect is relatively small, i.e., it hardly affects the detection limits. In trace analysis, CL=2Ó2×0.01×RSDBL×BEC (1) spectral interferences may deteriorate detection limits considerably and endanger accuracy. The signal-to-background ratio where CL is the detection limit in solution, RSDBL is the decreases as a result of enhanced background caused by stray relative standard deviation of the blank in % and BEC is the light, line wing or direct line overlap.If the presence of an background equivalent concentration. interfering line is not recognized, the result of the analysis will Table 2 compares the detection limits with respect to the be inaccurate. Spectral interferences affect the intercept of a dissolved sample of wet tissue and the range of ‘normal values’ calibration graph plotted on a linear scale.7 The type and of trace element concentrations in human liver, compiled from magnitude of both multiplicative and spectral interferences are various sources.20–22 The detection limits were measured by specific for the sample type.The effect of the sample matrix using eqn. (1), RSD=2%. It should be noted that the ‘normal on the accuracy and detection limits cannot be predicted in levels’ of trace element concentrations in tissues vary for the general terms. Only detailed experimental data can reveal the different countries.The difference in these levels might be due situation for each analytical line of the analyte in the presence to eating habits, types of food, geographical conditions or of a given matrix and acidity. The influence of matrix type anthropogenic sources of pollution. and acid concentration on the detection limits and accuracy In our toxicological work, the analytical task required the was investigated. determination of lethal concentration levels of the abovementioned toxic elements.These levels are higher than the ‘normal values’ by a factor of 10–100 and are therefore readily Animal Tissues measurable by direct ICP-AES analysis. The mean values were determined from the analysis of three separate dissolutions of Recently,18 the developments in atomic spectrometric techniques for the analysis of clinical and biological materials were each sample. We have no tissues with ‘certified values’ available and hence reviewed.It was noted that the analytical chemist can help in Table 2 Comparison between the detection limits with respect to the dissolved sample of wet tissue (in ng g-1) obtained in this work by using ICP-AES and ranges of ‘normal values’ or mean values of trace element concentrations in human liver of clinical interest, compiled from various sources This work (ICP-AES) Concentration range of ‘normal values’ in accordance with reference data/ng g-1 Selected analytical Detection limit/ line/pm ng g-1 ref. 20 ref. 21 ref. 22 Ag I 328 068 75 —* — — As I 234 984 1600 5–15 33–70 10 Ba II 455 403 5.2 — — 1300 Be I 313 042 5.0 — — — Bi I 223061 430 — 12–56 5 Cd II 214 438 200 300 1100–2300 1000 Cr II 205 559 110 8 12–230 50 Cu I 324 754 26 500–800 2100–2300 6000 Hg I 253650 560 30–150 160–1300 100 Mn II 257610 15 1000–2000 450–2100 1000 Pb II 220 353 1700 350–550 160–1000 1000 Sb I 206 833 330 — <10–70 5–10 Se I 203985 220 250–400 — 300 Sn I 233 484 880 — — <400 Tl I 276787 940 — — — Zn I 213 856 30 40000–60 000 21000–82 000 — Co II 238 892 120 6 13–62 — Fe II 239562 94 150000–250 000 — 150000 Mo II 202030 140 3600 200–1200 — Ni II 221 647 200 5–13 28–220 30 V II 309 311 120 — — 2–40 * No data available. 22 Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12cannot confirm the accuracy of the procedure in this way. The tration, CA) in the presence of lead or lead–antimony alloy. In these cases, QI(la)=0; QW(Dla)=SW(Dla)/SA, where SW(Dla) absence of losses was established as follows: three separate dissolutions of a sample of human liver (10 g) with ‘normal is the sensitivity associated with wing background (signal per unit matrix concentration, CI).The use of the symbol Dla values’ were made in the presence of As, Bi, Cd, Hg, Pb, Sb, Se and Sn (volatile elements). The analytes were added at instead of la expresses that SW(Dla) refers to the overall background level in the spectral window (Dla) viewed and not concentration levels equal to five times the detection limits (Table 2).The reference solutions contained the matrix blank specifically to la as happens with the partial line sensitivity. The background level is defined as the minimum background and the above-mentioned analytes at concentrations ranging from the detection limits to 10 times the detection limits. No in the ‘smoothed’ scan in a given spectral window (Dla). No line and wing background interferences were registered volatilization losses were observed in the decomposition procedure.in the presence of 10 mg ml-1 silver as a matrix. In this case, QI (la)=0 and QW(Dla)=0. Detection limits were calculated In cases of chronic poisoning with toxic elements, their concentrations are several times higher than the ‘normal using eqn. (1). The multiplicative interferences in the presence of the above- values’. In this work, the ‘normal levels’ of the toxic elements were used as reference values, i.e., in all cases tissues from a mentioned matrices were investigated.The net analyte signals decreased by 15% (bullet lead) and 20% (silver). The effect is normal body and corpse were analysed for evidence of poisoning. However, for As, Bi, Cr, Hg, Sb, Sn and V, the detection negative in both cases. A sensitivity change of this magnitude is not negligible in analysis because the results can be inaccur- limits obtained in this work by direct introduction of the sample solutions into the ICP with conventional pneumatic ate.For example, the content of silver in a certified reference material (E3–4, NIPKI, Plovdiv, Bulgaria) was obtained from nebulization were higher than the corresponding ‘normal levels’ (Table 2). Hence, analysis at the level of ‘normal values’ calibration graphs constructed from standards prepared in the solvent blank and in the solvent blank+10 mg ml-1 lead in requires analyte preconcentration.These possibilities will be discussed in a separate paper. solution (Table 3). Using Student’s criterion, a statistical difference between the two sets of data was found. Reviewing the data presented in Table 3, it can be concluded that accurate Bullet Lead and Silver results for the analyte content can be obtained if the calibration standards contain the same matrix concentration as the sample. The spectral interferences were studied in the presence of The sensitivity change depends on the type and concentration 10 mg ml-1 lead, 10 mg ml-1 lead–antimony alloy (9+1) and of the matrix. 10 mg ml-1 silver, respectively. Information on the interfering matrix lines was derived from the wavelength scans centred around the analytical lines. Details of how the results were Table 3 Content of silver in the lead certified reference material E3–4, obtained and quantified are given elsewhere.12 Fig. 1 shows an obtained by using calibration procedures in the solvent blank and in example of wing background interference by Pb 257 726 pm the solvent blank+10 mg ml-1 lead in solution.The mean values were on Mn II 257 610 pm. XW is the net wing background signal. determined from the analysis of three replicates of E3–4 The detection limits in the presence of 10 mg ml-1 lead or Readings from 10 mg ml-1 lead–antimony alloy (9+1) were calculated from calibration graph the following equation:23,24 Readings from in prepared calibration graph in solvent blank CL,conv=2Ó2×0.01×RSDBL×[BEC+QI(la)CI+QW(Dla)] prepared solvent +10 mg ml-1 (2) blank (%) lead (%) Certified where CL,conv is the conventional detection limit; QI(la)=SI /SA, No.value (%) Xi X� Xi X� where SI is the partial sensitivity of an interfering line, defined as the signal per unit matrix concentration (CI), produced by 1 0.00140 0.00118 0.00119 0.00138 0.00139 2 0.00140 0.00120 0.00142 the line at the peak wavelength of an analytical line and SA is 3 0.00140 0.00119 0.00139 the sensitivity of the analytical line (signal per analyte concen- Table 4 Detection limits with respect to the dissolved solid sample in the solution (in %) for solid concentrations of 10 mg ml-1 Pb (column 1) and 10 mg ml-1 Pb–Sb alloy (Pb5Sb=951) (column 2) Detection limit (%) Analytical line/pm 1 2 Al I 396152 2.9×10-4 3.4×10-4 Ag I 328068 7.0×10-5 7.0×10-5 As I 234984* 1.8×10-3 1.0×10-3 Bi I 223 060 4.6×10-4 4.6×10-4 Cd I 214 438 4.0×10-5 4.0×10-5 Cu I 324 754 4.7×10-5 4.7×10-5 Fe II 238204 6.0×10-5 6.0×10-5 Mg II 279 553 1.2×10-5 1.2×10-5 Mn II 257610 2.0×10-5 2.0×10-5 Sb I 206838 4.5×10-4 —† Sn I 235484 1.2×10-3 1.2×10-3 Tl I 276 781 9.8×10-4 9.8×10-4 Z213856 3.2×10-5 3.2×10-5 * The most sensitive arsenic line, As I 193696,20 could not be used Fig. 1 Example of a spectral scan over a spectral region ±200 pm since it lies outside the wavelength range of the ICP equipment (see Table 1). around the analytical line, Mn II 257 610 pm.Interferent: 10 mg ml-1 lead. † Matrix element. Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12 23Table 5 Detection limits with respect to the dissolved solid sample The financial support from the National Fund for scientific in the solution (in %) for solid concentrations of 10 mg ml-1 Ag research of the Ministry of Science, Education and Technology obtained in this work (column 1), ref. 25 (column 2) and ref. 26 of Bulgaria under registration No.X-499A is gratefully (column 3) acknowledged. Detection limit (%) Analytical line/pm 1 2* 3† REFERENCES Cu I 324 754 1.0×10-5 —‡ 1.0×10-5 1 Krasnobaeva, N., Nedyalkova, N., and Lalchev, M., T heses of the Zn I 213 856 5.0×10-5 — 5.0×10-5 Institute of Forensic Science and Criminology–Ministry of Interior, Mn II 257610 1.0×10-5 5.0×10-5 2.0×10-5 1980, 9, 111. Mg II 279553 2.6×10-6 — 2.0×10-5 2 Locke, J., Anal. Chim. Acta, 1980, 113, 3. Na I 588995 5.5×10-5 — — 3 Lalchev, M., Ionov, I., and Daskalova, N., Anal.L ab., 1992, 1, 77. Fe II 359940 5.0×10-5 5.0×10-5 5.0×10-5 4 Lalchev, M., Ionov, I., and Daskalova, N., Presented at the 29th Se I 196096 2.3×10-3 — 5.5×10-4 Colloquium Spectroscopicum Internationale, Leipzig, 1995. Ni II 231 604 5.0×10-5 5.0×10-5 ×10-4 5 Peters, A., and Koons, R., ICP Inf. Newsl., 1988, 13, 751. Cd I 214 438 2.0×10-4 — 1.0×10-4 6 Powell, G. L., Robinson, R. R., Cocks, B., and Wright, M., Te II 214281 1.0×10-3 — 8.0×10-4 J.Forensic Sci., 1978, 23, 712. Pb II 283 307 5.0×10-5 5.0×10-5 ×10-4 7 Boumans, P. W. J. M., in Inductively Coupled Plasma Emission Ca II 393 366 1.6×10-5 — 2.0×10-5 Spectroscopy, Part 1, Methodology, Instrumentation and Perfor- Pd II 340 458 2.2×10-4 — — mance, ed. Boumans, P. W. J. M., Wiley, New York, 1987, p.100. Sb I 206 833 3.3×10-4 — 4.5×10-4 8 Maessen, F. J. M. J., in Inductively Coupled Plasma Emission Bi I 223061 4.0×10-4 — 5.0×10-4 Spectroscopy, Part 2, Applications and Fundamentals, ed.Boumans, Hg I 253652 1.0×10-4 — 1.5×10-4 P.W. J. M., Wiley, New York, 1987, p. 100. 9 Boumans, P. W. J. M., Fresenius’ Z. Anal. Chem., 1986, 324, 397. * 56 MHz ICP was used. 10 Boumans, P. W. J. M., and Vrakking, J. J. A. M., J. Anal. At. † 40.68 MHz ICP was used. Spectrom., 1987, 2, 513. ‡ No data available. 11 Boumans, P. W. J. M., in Inductively Coupled Plasma Emission Spectroscopy, Part 1, Methodology, Instrumentation and Performance, ed. Boumans, P.W. J. M., Wiley, New York, Tables 4 and 5 summarize the detection limits with respect 1987, p. 358. to the dissolved solid sample in the solutions: 10 mg ml-1 12 Daskalova, N., Velichkov, S., Krasnobaeva, N., and Slavova, P., lead (Table 4, column 1), 10 mg ml-1 lead–antimony alloy Spectrochim. Acta, Part B, 1992, 47, E1595. (Pb5Sb=951) (Table 4, column 2) and 10 mg ml-1 silver 13 Velichkov, S., Daskalova, N., and Slavova, P., Spectrochim. Acta, (Table 5). The RSD was 2%.Table 5 shows for comparison Part B, 1993, 48, E1743. the literature data.25,26 It should be noted that under the 14 Daskalova, N., Velichkov, S., and Slavova, P., Spectrochim. Acta, Part B, 1996, 51, 733. influence of the strong irradiation by the ICP, partial reduction 15 Chudinov, E. G., Ostroukhova, I. I., and Varvanina, G. V., of silver to the metallic state was observed. The deposition of Fresenius’ Z. Anal. Chem., 1989, 335, 25. the silver metal on the nozzle and capillary of the concentric 16 Brenner, I.B., Mermet, J. M., Segal, I., and Long, G. L., glass nebulizer leads to changes in the nebulization efficiency Spectrochim. Acta, Part B, 1995, 50, 323. and sample uptake rate. This effect was eliminated by using a 17 Brenner, I. B., Segal, I., Mermet, M., and Mermet, J. M., concentric metallic nebulizer with a Pt-capillary (Jobin-Yvon). Spectrochim. Acta, Part B, 1995, 50, 333. 18 Taylor, A., Branch, S., Crews, H. M., Halls, D. J., and White, M., The good analytical characteristics of the proposed methods J.Anal. At. Spectrom., 1994, 9, 87R. satisfy the requirements for identification and classification of 19 Boumans, P. W. J. M., L ine Coincidence T ables for Inductively materials in forensic examinations. Bullet lead or silver was Coupled Plasma Atomic Emission Spectrometry, Pergamon Press, identified on the basis of their trace element content. This Oxford, 1980, p. 1984. information and other evidence can be used to answer the 20 Iyengar, G.V., Concentrations of 15 T race Elements in some question: what is the likely origin of these materials? Selected Adult Human T issues and Body Fluids of Clinical Interest from Several Countries: Results from a Pilot Study for the Establishment of Reference Values, Report No. 1974 of the Institute of Medicine, Ju�lich Nuclear Research Center, 1985. CONCLUSIONS 21 Sumino, K,, Hayakawa, K., Shibata, T., and Kitamura, S., Arch. Environ. Health, 1975, 30 487. ICP-AES as a multi-element plasma source was applied to the 22 Popov, T. A., Zaprianov, Z. Z., Bentchev, I. B., and Georgiev, elemental analysis of materials relevant to forensic science G. K., Atlas of T oxicokinetics, ‘Medicina i fizkultura’, Sofia, 1984. 23 Boumans, P. W. J. M., and Vrakking, J. J. A. M., Spectrochim. applications (tissues, bullet lead and pure silver). The loss of Acta, Part B, 1988, 43, 69. accuracy caused by systematic errors in the forensic examin- 24 Boumans, P. W. J. M., Tielrooy, J. A., and Maessen, F. J. M. J., ation is a serious problem for material identification and Spectrochim. Acta, Part B, 1988, 43, 173. classification. In order to achieve the required accuracy, cali- 25 Vankatasubramanian, R., Biswas, S. S., and Murty, P. S., Indian bration by precise matching of matrix content and sample J. T echnol., 1991, 29, 605. acidity, as well as essential information on the type of spectral 26 Jobin-Yvon Division d’Instruments, Applications P009, Longjumeau, 1985. interferences, is required. In this work, only a few wing background interferences were registered in the determination Paper 6/03869G of trace elements in bullet lead. These results indicate that Received June 4, 1996 ICP-AES is a powerful tool for the analysis of materials of Accepted October 2, 1996 interest to forensic scientists. 24 Journal of Analytical Atomic Spectrometry, January 1997, Vol.