JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 82 1 Sensitive Method for Determination of Lead by Potassium Dichromate-Lactic Acid Hydride Generation Inductively Coupled Plasma Atomic Emission Spectrometry* M. C. Valdes-Hevia y Temprano M. R. Fernandez de la Campa and Alfredo Sanz-Medelt Department of Physical and Analytical Chemistry Faculty of Chemistry University of Oviedo Julian Claveria s/n 33006 Oviedo Spain Continuous flow plumbane generation for sample introduction into an inductively coupled plasma (ICP) and further determination of lead by atomic emission spectrometry (AES) has been investigated in order to increase the detection limits of conventional nebulization ICP-AES. Continuous hydride generation from different media is discussed and the generation of plumbane using potassium dichromate with lactic acid has been selected.Sensitivity selectivity and accuracy of the corresponding determination of lead by ICP-AES are investigated. The proposed method (detection limit 2 ng ml-I precision 1.3% at the 50 ng ml-' level) has been applied to the determination of low levels of lead in soft drinks sediments and lichens. The results obtained show good agreement with certified or expected values. Keywords Lead; inductively coupled plasma atomic emission spectrometry; continuous hydride generation; potassium dichromate-lactic acid There is currently much concern regarding the adverse effects of lead on health' and the considerable amount of organic and inorganic lead pollutants emitted into the atmosphere.Thus in recent years many countries have given special attention to reducing the lead content in gasolines replac- ing tetraalkyllead species additives in petrols by other chemicals and enforcing adequate control of lead concen- tration in the en~ironment.~~~ Therefore the present inter- est in developing more sensitive selective and reliable analytical methods to determine traces and ultratraces of this element in samples of biological and environmental interest is understandable. Analytical atomic spectrometry techniques particularly inductively coupled plasma atomic emission spectrometry (ICP-AES) have been widely used for this purpose. However the low levels to be analysed are demanding the use of more sensitive methods than those provided by conventional nebulization ICP-AES. The introduction of lead into the atomizer as a vapour by plumbane generation has been widely applied to enhance the sensitivity of lead determinations by atomic absorption spectrometry (AAS).4-23 It is recognized today that the efficient generation of lead hydride the least known of the hydrides of the Group 14 elements because of its instabil- it^,^ requires the simultaneous presence of an oxidant and a complexing agent of metastable lead(Iv) before final reduction to PbH4 with NaBH,.In fact many oxidants and complexing agents have been reported to increase the sensitivity of the lead hydride based techniques i.e. potassium dichromate with tartaric acid malic acid5v6 or lactic acid;7 hydrogen peroxide with citric acid,8 nitric acid6y7v9 or hydrochloric a ~ i d ; ~ .~ ~ ammonium persulfate with nitric a ~ i d ; ~ - ~ J ~ - l ~ potassium hexacyanoferrate(~~r),l~ am- monium cerium(II1) nitrate with oxalic acid,Is or nitroso R- salt.23 The generation of plumbane has also been achieved from organic medial6-** and from 'ordered media' (micelles and vesicles).26 Madrid et aL7 compared different systems and methodologies for plumbane generation by AAS tech- niques. They concluded that to their knowledge the best oxidant system was potassium dichromate-lactic acid.7 * Presented at the 1993 European Winter Conference on Plasma t To whom correspondence should be addressed. Spectrochemistry Granada Spain January 10-1 5 1993. Therefore the purpose of the present work was to examine the ability of the potassium dichromate-lactic acid system to improve the analytical performance characteristics of continuous hydride generation (HG) ICP-AES techniques for the determination of lead.As a result a sensitive and selective analytical method for the determination of lead by continuous HG-ICP-AES has been developed using the dichromate-lactic acid reaction medium. The validity of this method has also been demonstrated by the successful determination of lead at trace levels in some environmental and food samples. Experimental Instrumentation An ICP Philips Model PU7000 was used for ICP-AES detection. Other details of the experimental flow system and the ICP-AES set-up used are given in Table 1 and Fig. 1. Reagents A 1000 pg ml-l lead@) stock standard solution was stabilized in 0.5 moll-' HN03 (Merck).Working solutions were freshly prepared daily by diluting appropriate aliquots of the stock solution with ultrapure water. Sodium tetra- hydroborate(II1) solutions were prepared by dissolving NaBH (Carlo Erba) in ultrapure water (Milli-Q) and stabilized in a 0.1% m/v sodium hydroxide solution. Solutions were prepared weekly and filtered before use. Potassium dichromate (Merck) and lactic acid (Merck 90%) solutions were prepared by appropriate dissolution of the required amounts of the reagents in ultrapure water. All mineral acids and metal salts used were of analytical- reagent grade and ultrapure water (Milli-Q) was used throughout. General HG-ICP-AES Procedure Continuous plumbane generation In the flow system described here (Fig. I) the sample a 0.3% m/v potassium dichromate and 3% v/v lactic acid solutions were passed through a four channel union cross by a peristaltic pump at a rate of 0.75 ml min-l each.The resulting mixture (see Fig. 1) was merged with a 5% m/v822 Tetrahydroborate- Acid Oxidant Sample (lead) - JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 0.75 ml min-' 0.75 ml min-' Table 1 Instrumentation ICP Spectrometer Gas-liquid separation interface Peristaltic pump (four channels) Microwave oven for sediment and lichen samples digestion Philips Model PU7000 equipped with 40 MHz source unit Grid type nebulizer provided with the instrument Gilson Minipuls 2 Milestone. Digestion module MLS- 1200; exhaust module EM-5; automatic capping module AC- 100; closed vessels SV- 1401 I0 Table 2 Optimum conditions for lead hydride generation by ICP-AES Plasma experimental conditions Wavelengthhm Radio frequency forward powerlkW Nebulizer pressure/lb in-* Coolant gas flow rate/] min-' Auxiliary gas flow ratell min-l Final sample flow ratelm1 min-' Drain flow rate/ml min-' Integration ti me/s Chemical parameters Potassium dichromate Lactic acid Sodium tetrahydroborate 220.353 0.7 30 13 0 3 0.3% m/v (flow rate 0.75 ml min-') 3% v/v (flow rate 0.75 ml rn1n-l) 5% m/v in NaOH 0.1% m/v (flow rate 0.75 ml min-I) sodium tetrahydroborate solution (flow rate 0.75 ml min-l) via an ordinary T-piece producing a final flow rate of 3 ml min-l.This solution feeds the grid nebulizer of the ICP detuned in order to separate the volatile species. Thus PbH goes to the plasma and is separated from the liquid phase going to waste.Lead was measured at the 220.353 nm emission line under the conditions given in Table 2. Background correction at 220.371 and 220.339 nm was carried out. Dissolution of Sediments and Lichens A 0.200 g sample was weighed directly into a poly(tetra- fluoroethylene) (PTFE) vessel and 1.5 ml of HN03 0.5 ml of HClO and 0.5 ml of HF were added. After tightly capping the vessels using the Milestone capping station the sample carousel was placed in the microwave oven using the heating programme shown in Table 3. After cooling the vessels were uncapped and 3.5 ml of a solution of 6% m/v H3B03 were added for complexing the excess HF; then the PTFE vessels were placed again in the microwave oven and heated for 5 min at 60 W power.The sample solutions were then filtered and the contents transferred into 100 ml calibrated flasks and diluted with ultrapure water (Milli-Q). This solution was the sample (lead) used for PbH generation ICP-AES. Separate solid ICP (grid nebulizer) portions (1.000 g each) from each sample were dryed at 105-1 10 "C and used to determine the water content of the sample. Juice samples were directly analysed as received without any sample pre-treatment following the continuous flow plumbane generation described above. Results and Discussion Optimization of Instrumental and Chemical Parameters Using continuous gas-liquid separation and ICP detection the effect of ICP and chemical generation variables such as nebulizer gas pressure forward r.f.power coolant gas flow sample flow rates etc. were studied by following the general HG-ICP-AES procedure and a univariant-type experimen- tal search. The optimum ICP instrumental values obtained are summarized in Table 2 and were selected to study optimum chemical parameters (concentration of reactants flows etc.) for continuous PbH generation and introduc- tion into the plasma. Maximum signal-to-background ratio was always the optimization criterion. The results observed have been plotted in Fig. 2 for potassium dichromate lactic acid and NaBH and show that the optimum concentrations are 0.3% m/v 3% v/v and 5% m/v respectively. The effect of sample flow rate has been studied from 1 to 10 ml min-I and as expected the signal increased almost linearly with the sample flow rate as shown in Fig.3 in which the observed results have been plotted. A flow rate of 3 ml min-' was selected as a compromise in order to avoid flooding of the spray chamber in the commercial set- Table 3 Microwave oven heating programme Step Time/min PowerIW I 5 60 2 5 120 3 10 180 4 5 0 5 10 120 Fig. 1 Schematic diagram of the continuous HG flow system usedJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 823 Dichromate (% m/v) 0 0.,2 0;4 0.,6 0. j 2 4 6 8 10- Lactic acid (% v/v) or tetrahydroborate (YO m/v) Fig. 2 Optimization of A tetrahydroborate; B lactic acid; and C dichromate concentrations % .- E 1001 I I I I w l 3 5 7 9 Sample flow rate/mI min-' Fig. 3 Effect of sample flow rate on lead signal A maximum intensity; and B background intensity.[Pb]=100 ng ml-l and drain flow rate= I 1 ml min-l. Other conditions as in Table 2 up but the results in Fig. 3 show a way of increasing the detection power if necessary; wider diameter drain tubes and a higher sample flow can be used without flooding. Analytical Performance Characteristics Using the experimental conditions given in Table 2 the analytical parameters defining the performance of the proposed method were evaluated. The calibration graph was linear up to 0.5 pg ml-I of Pb and the detection limit for the element ( 3 ~ ~ ) was 2 ng ml-I. The within-run precision evaluated by analysing ten replicates containing 50 ng ml-I of Pb turned out t o be 2 I .3%. It should be stressed however that working at a sample flow rate of 9 ml min-' (see Fig.3) and draining with adequate tubing able to drain out at 1 1 ml min-l the detection limit observed was 0.7 ng ml-1 with a within-run precision evaluated (over 10 ng ml-1 of Pb) of around f 3%. Effect of the Nature of the Acid Used for PbH4 Generation The mineral acids tested were HCl HC104 HN03 H2S04 and HF (with boric acid) because they are currently used in many sample digestions. All these mineral acids except sulfuric acid which produced maximum intensity at a concentration of 0.05 mol 1 - I and at a level of 0.2 mol 1 - l caused the emission signal of Pb to disappear showed similar behaviour to that illustrated for nitric acid in Fig. 4 curve A. As citric acid is normally present in the soft drinks to be analysed its addition to the HG medium was also investigated in more detail.Results showed that inadequate HG of lead from dichromate-lactic acid was obtained; the lead recovery decreased almost exponentially with increas- ing citric acid concentration under the conditions of the proposed procedure as shown in Fig. 4 curve B. Lead(rv) metastable complex formation with lactic acid in the proposed procedure which seems to act as an intermediate favouring plumbane generati~n,~ could be hindered by citric acid acting as a competitive ligand. For comparison purposes Fig. 4 curve C shows the observed effect of citric acid using a different PbH4 generation system namely ammonium persulfate-nitric acid-cetyltrimethylammonium bromide (CTAB).26 It can be seen that the effect of citric on Pb signals is much less in this latter PbH4 generation medium.Kinetic Studies In order to further investigate the nature of the observed enhancement using potassium dichromate-lactic acid,7 batch AAS signals were studied by comparing the peak shape of the transient AAS signal obtained when generating PbH4 from the oxidant systems2? potassium dichromate- lactic acid (0.3% m/v K2CrZ0 2% v/v lactic acid); ammonium persulfate-nitric acid-CTAB; and ammonium persulfate-nitric acid [3% m/v (NH4)2SZOs 2% v/v HN03 and 1 x mol 1 - I CTAB]. The results are shown in Fig. 5 and indicate that the presence of micelles and vesicles seems to accelerate the HG kinetics in the ammonium persulfate-nitric acid system but the potassium dichro- matic-lactic acid system provided higher and sharper lead peaks.Interference Studies Five possible sources of lead interferences were investigated for selective PbH4 generation (i) hydride forming elements; (ii) transition metals; (iii) alkali metals; (iv) alkaline earth metals; and ( v ) anions. All the elements tested and the level of tolerance observed in the determination of 0.1 ppm of lead by the proposed method are summarized in Table 4. [Nitric acidl/mol I-' 0 0.15 0.30 0.45 0.60 0.75 I I I I I n t 7 2 1 1 I I I I I 0 0.05 0.10 0.15 0.20 0.25 [Citric acidl/mol I-' Fig. 4 Effect of A nitric acid; and B citric acid on the ICP-AES signal of lead with HG from potassium dichromate-lactic acid medium; and C effect of citric acid on the ICP-AES signal of lead with HG from ammonium persulfate-nitric acid-CTAB medium824 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 Table 4 Interference studies on 100 ng m1-I of Pb Amount of interferenv Pb interferent Recovery* Interferent pg ml-I mass ratio (O/O) 50 10 50 10 50 10 10 10 50 50 50 50 50 50 200 1000 1000 1000 1000 1000 1000 1000 5000 5000 1000 1000 1000 0.5 1:500 1:lOO 1300 1:lOO 1:500 1:lOO 1:lOO 1:lOO 1:500 1:500 1:5 1500 1:500 1:500 1:500 1:2000 1 10000 1 10000 1 10000 1:10000 1 10000 1 I0000 1 10000 1 :50000 1 :50000 1 10000 1 10000 1 10000 82 77 103 78 82 93 84 78 123 102 94 99 99 100 93 104 81 103 103 100 100 100 I00 102 100 94 102 90 * The errors (precision) observed in each case (expressed as relative standard deviation) were within f 3%. i- I Time -t Fig. 5 Effect of the different oxidant systems on the lead batch AAS peak profile A potassium dichromate-lactic acid absor- bance 0.447 5-fold reduced signal; B ammonium persulfate-nitric acid-CTAB absorbance 0.082; and C ammonium persulfate- nitric acid absorbance 0.039 The observed interferences due to hydride forming elements can be explained by competitive reactions in which these elements successfully compete with lead for the NaBH to form their corresponding hydrides at the expense of plumbane production.Copper always interfered in this determination at the levels studied. This interference could be caused by a reduced rate of evolution.27 High levels of alkali alkaline earth metals or common anions were found not to affect lead hydride generation (see Table 4). Analysis of Real Samples The method established for lead determination by continu- ous HG-ICP-AES from a mixture of potassium dichrom- ate-lactic acid was applied to the determination of low levels of lead in certified sediments and in lichens. Certified reference materials (CRMs) sediments and lichens from the Community Bureau of Reference (BCR) were analysed according to the general recommended procedure by reference to an aqueous calibration line and also by standard additions techniques (to check for matrix interferences).Background correction (at 220.37 I and 220.339 nm) was employed. The results obtained in each case corrected for dry mass in pg g-* can be seen in Table 5 for sediments and the certified lichen. Commercial fruit juices were analysed directly without any sample pre-treatment by HG-ICP-AES and the results obtained by the proposed method were compared with the Table 5 Analysis of real samples Obtained value/ Certified value*/ Sample lug g-' Pg g-l BCR CRM 227 Estuarine Sediment Calibration line 128.3 f 2.8 146f3 Standard additions 144.7 f 3.4 146k3 BCR CRM 320 River Sediment Calibration line 41.3+ 1.3 42.3k 1.6 Standard additions 42.9 2 2.1 42.3 k 1.6 BCR RM Lichens TP-24 Calibration line 5.84 k 0.0 I 5.7 1 f 1.30 HG-ICP-AESf/ ETGAAS/ ng ml-1 ng m1-l Apple 66.4 f 7.3 65.3 f 1.9 Orange 13.0 +- 5.2 11.3 +.1.0 Pineapple 32.6 f 2.7 33.0 k 3.7 * Recommended value given for Lichens T2-240. f Determination using standard additions.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 825 results for lead obtained by electrothermal atomic absorp- tion spectrometry (ETAAS) for the same samples. The lead content in the three commercial juices analysed was evaluated by using the technique of standard additions because of the interference of citric acid (see Fig.4 curve B) an important component in these commercial juices usually present at a concentration level of 0.1 mol 1-* (2% m/v). The validation of these latter results was carried out by ETAAS. The results obtained by both methods can be seen comparatively in Table 5 and show very good agreement between the expected and the observed values of lead content. It should be noted that the precision obtained by HG-ICP-AES was worse than that with ETAAS prob- ably because of matrix interferences which indicates the necessity of using the standard additions technique.Conclusions The results demonstrate that a mixture of potassium dichromate and lactic acid provides improved analytical performance characteristics (higher sensitivity good preci- sion and low interference level) for the determination of lead by ICP with plumbane generation. This is probably due to enhanced kinetics and efficiency of PbH4 generation (see Fig. 5). The detection limit of the ICP-AES determina- tion of lead in water with HG using the recommended procedure is 2 ppb which compares favourably with 13 ppb by HG-ICP-AES using ammonium persulfate-nitric acid and with 9 ppb adding CTAB surfactant to the latter medium.26 Conventional nebulization of lead by ICP-AES observed in this set-up was 20 ppb. The method seems to be more selective than the more common method used for plumbane generation ( i e .ammonium persulfate-nitric acid ~ y s t e m ~ - ~ ~ ' ~ - ~ ~ ) and has proved to be adequate for the determination of low levels of lead in environmental samples of varied matrices ( i e . sediments plant material and fruit juices). We gratefully acknowledge Fundacion para el Foment0 en Asturias de la Investigacion Cientifica Aplicada y la Technologia (FICYT) and Comision Interministerial de Ciencia y Technologia (CICYT) for financial support and a grant to M. C. V. H. y T. The loan of the ICP instrument from Unicam Analytical Systems (Cambridge) is also gratefully acknowledged. References 1 Beockx R. Anal. Chem. 1986 58 274A. 2 Taylor A. Clin. Endocrinol. Metab. 1985 14 658. 3 Off. J. Eur.Comm. 1980 23 11. 4 Fleming H. D. and Ide R. G. Anal. Chim. Acta 1976,83 67. 5 Brindle I. D. and Le X. Anal. Chem. 1989 61 1175. 6 Jin K. and Taga M. Anal. Chim. Acta 1982 143 229. 7 Madrid Y. Meseguer J. Bonilla M. and Camara C. Anal. Chim. Acta 1990 237 181. 8 Vijan P. E. and Wood G. R. Analyst 1976 101 966. 9 Aroza I. Bonilla M. Madrid Y. and Camara C. J. Anal. At. Spectrom. 1989 4 163. 10 Jin K. Taga H. Yoshida H. and Himike S. Bunseki Kagaku 1978 27 759. 1 1 Madrid Y. Bonilla M. and Camara C. J. Anal. At. Spectrom. 1988 3 1097. 12 Madrid Y. Bonilla M. and Camara C. J. Anal. At. Spectrom. 1989 4 167. 13 Sanz J. Basterra P. Galban J. and Castillo J. R. Mikro- chim. Acta 1989 1 271. 14 Tao R. and Zhou H. G. Fenxi Huaxue 1985 13 253. 15 Li J. Liu Y. and Lin T. Anal. Chim. Acta 1990 231 151. 16 Aznarez J. Palacios F. Vidal J. C. and Galban J. Analyst 1984 109 7 13. 17 Aznarez J. Vidal J. C. and Carnicer R. J. Anal. At. Spectrom. 1987 2 55. 18 Nerin C. Olavide S. and Cacho J. Anal. Chem. 1987 59 1918. 19 Bonilla M. Rodriguez L. and Camara C. J. Anal. At. Spectrom. 1987 2 157. 20 Thompson K. C. and Thomerson D. R. Analyst 1974 99 595. 21 Madrid Y. Bonilla M. and Camara C. Analyst 1990 115 563. 22 Yan X.-p. and Ni Z.-m. J. Anal. At. Spectrom. 1991 6,483. 23 Zhang S.-z. Han H.-b. and Ni Z.-m. Anal. Chim. Acta 1989 221 85. 24 Paneth F. and Rabinowitsch E. Ber 1925 58 1138. 25 Ikeda M. Nishibej J. Mamada S. and Tujino R. Anal. Chim. Acta 198 1 125 109. 26 Valdes-Hevia y Temprano M. C. Aizpun Fernandez B. Fernandez de la Campa M. R. and Sanz-Medel A. Anal. Chim. Acta 1993 in the press. 27 Nakahara T. in Sample Introduction in Atomic Spectroscopy ed. Sneddon J. Elsevier Amsterdam 1990 ch. 10. Paper 3/00284E Received January I I993 Accepted March 16 1993