JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 469 Slurry Sampling for the Determination of Lead in Marine Sediments by Electrothermal Atomic Absorption Spectrometry Using Palladium-Magnesium Nitrate as a Chemical Modifier* Pilar Bermejo-Barrera Carmen Barciela-Alonso Manuel Aboal-Somoza and Adela Bermejo-Barrera Department of Analytical Chemistry Nutrition and Bromatology Faculty of Chemistry University of Santiago de Compostela Spain A method for the determination of lead in slurries of marine sediment using palladium-magnesium nitrate as a chemical modifier has been optimized. To stabilize the marine sediment slurry different thickening agents were studied. The grinding time and the particle size were also studied. To achieve complete pyrolysis of the slurry sample two charring steps were used the first one at a low temperature 480°C and the second at 900 "C.The precision of the method was studied as within-run and within-batch precision; the relative standard deviations obtained in both cases were less than 3%. The relative standard deviation for the repeatability of the over-all procedure was 5.0%. The accuracy of the method was studied using a Certified Reference Material PACS-1 (Marine Sediment) from the National Research Council yf Canada which has a certified lead content of 404 20 mg kg-'; the result obtained was 41 8 & 11 mg kg- . The detection limit for lead was 0.22 pg I-'. Calibrations using aqueous standards and the reference material were compared. This method has been applied to the determination of lead in marine sediment samples from the Galicia Coast and the results were compared with a sample digestion method using nitric and hydrochloric acids in a high pressure poly(tetrafluoroethy1ene) bomb and microwave energy.No significant differences were found between the two procedures. Keywords Lead; sediment; slurry; elecfrothermal atomic absorption spectrometry In recent years in order to avoid problems related to conven- tional wet-oxidation and dry-ashing sample preparation pro- cedures for the determination of metals in sediment samples the direct analysis of solids and slurries by electrothermal atomic absorption spectrometry (ETAAS) has been extensively developed.' The direct analysis of solid samples offers several advantages over conventional procedures including (i) reduced sample preparation time; (ii) decreased possibility of loss of analyte through volatilization prior to analysis; (iii) reduced loss of analyte associated with retention by insoluble residues; and (iv) reduced possibility of sample contamination.Nevertheless there can be problems with the introduction of few milligrams of powdered sample into a graphite tube. The particular specimen selected could not be representative of the total sample the selection of adequate standards for calibration can be difficult and background absorption prob- lems can be present owing to the high levels of matrix present during atomization. Introduction of slurry samples combines the advantages of solid and liquid sampling,'4 and avoids many of the problems associated with direct solid sampling.The slurry technique gives better analytical performance than direct solid sampling because problems associated with weighing the sample and transferring the sample into the graphite tube are avoided. Moreover the concentration of the slurry can easily be changed so that the analyte concentration falls within the range of the calibration graph; however this can present two problems. If the slurry concentration is very high dilution of the slurry can only be carried out within a limited range because the precision is degraded for highly diluted slurries because only a small number of particles will remain in the slurry. If the analyte content in the original sample is low the concentration of the slurry can be increased although the injection precision can deteriorate if slurries are more concentrated.However there are still some problems related to the analysis of sediment samples in the form of slurries. Perhaps the two most important problems are the homogeneity of the slurry * Presented at the XXVIII Colloquium Spectroscopicum Internationale (CSI) Post-Symposium on Graphite Atomizer Techniques in Analytical Spectroscopy Durham UK July 4-7 1993. and calibration. Homogenization of the slurry can be achieved by mechanical or ultrasonic agitation of the sample powder in solution or by stabilization of the particles using a thixotropic thickening agent which increases the viscosity of the sus- pension. Different thickening agents have been used in the determination of lead by slurry sampling.Hoening and Van Hoeyweghen' proposed the use of a glycerol medium Littlejohn et aL6 proposed Viscalex HV-30 (acrylic acid polymers) and Bermejo-Barrera et aL7 the use of Triton X-100; Lynch and Littlejohn' also proposed the use of an antifoam agent. To avoid problems with the matrix interferences chemical modification is recommended. In the analysis of solid samples in many instances a chemical modifier added in the form of a liquid is ineffective for stabilization. Thus de Kersabiec and Benedetti' carried out a study of the effect of chemical modifiers in liquid and solid form such as phosphoric acid magnesium nitrate and nickel nitrate on the atomization of geological samples and in both cases double peaks were obtained. The use of chemical modifiers is not a problem for slurry sampling because the interaction between the chemical modifier and the particles of the solid sample is closer than for direct solid sampling.Among the chemical modifiers tried for lead Hinds and co-workerslO.'l proposed the use of palladium alone or a mixture of magnesium and palladium nitrate Lynch and Littlejohn' used only palladium Zongqiang et a1." investigated ammonium dihydrogen phosphate and Ebdon and Le~hotycki'~ proposed the use of ascorbic acid for slurries of various types of samples. In addition Hinds et ~ 2 . ' ~ proposed the use of a fast temperature programme without a chemical modifier. Moreover to avoid problems with the background absorp- tion Ebdon et a1.l' used an oxygen or air ashing step and Hoening et ~ 1 .' ~ proposed pre-digestion of the sediment with nitric acid. In this paper the direct determination of lead in slurries of marine sediment samples is studied using palladium-mag- nesium nitrate as a chemical modifier. Experimental Instrurnenta tion Measurements were carried out using a Perkin-Elmer Model 1 lOOB atomic absorption spectrometer equipped with an470 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Table 1 Graphite furnace programme Temperature/ Step "C Dry 1 100 Dry 2 200 Charring 1 480 Charring 2 900 Atomize 2500 Clean 2600 Ramp time/ 10 25 10 10 0 1 S Hold time/ 20 25 10 15 3 3 S Ar flow/ ml min- 300 300 300 300 stop 300 HGA-400 graphite furnace atomizer and an AS-40 auto- sampler. The source of radiation was a lead hollow cathode lamp operated at 10 mA which provided a 283.3 nm line with a spectral bandwidth of 0.7 nm.Deuterium lamp background correction was used. Pyrolytic graphite coated graphite tubes with L'vov platforms were used. For all measurements made during this study integrated absorbance with an integration time of 3 s was used. The temperature and time programmes for the atomizer are shown in Table 1. The volume injected was 20 pl. A Laser Coulter Series LS 100 Fraunhofer optical model from Coulter Electronics Hialeah FL USA was used to check the particle size distribution. An agitator (Vibromatic) and magnetic agitator (Agimatic) from Selecta (Barcelona Spain) were used in the preparation of the slurries. A Portland Model DMR-140 microwave oven and high- pressure poly( tetrafluoroethylene) (PTFE) bombs from Parr Moline IL USA were also used.Reagents Lead nitrate stock standard solution. A 1 mg ml-1 solution of lead (BDH Poole UK). Each test solution was prepared with ultrapure water immediately before use. Triton X-100. Triton X-100 polyethylene glycol mono(p- 1,1,3,3,-tetramethylbutylphenyl) ether for gas chromatogra- phy was obtained from Merck (Darmstadt Germany). Viscalex H V 30. This acrylic copolymer containing carboxyl groups was purchased from Allied Colloids (Bradford Yorkshire UK). Glycerol 99.5%. ACS-reagent grade glycerol was supplied by Sigma Chemicals (St. Louis MO USA). Palladium solution. Prepared by dissolving 300 mg of pal- ladium (99.999%) (Aldrich Chemical Milwaukee WI USA) in 1 ml of concentrated nitric acid and diluting to 100 ml with ultrapure water.If dissolution was incomplete 10 pl of hydro- chloric acid (Suprapur 35% with a maximum lead content of 0.001 mg l-' BDH) were added to the cold nitric acid and heated to gentle boiling in order to volatilize the excess of chloride. Magnesium nitrate. Suprapur Merck. Nitric acid. Suprapur (69.0-70.5%) maximum lead content Hydrochloric acid. Suprapur BDH. Reference material PACS-1 . Marine Sediment National Research Council Canada (NRCC) Ottawa Ontario Canada. Argon. N50 99.999% purity used as a sheath gas for the atomizer and to purge internally. Ultrapure water. Resistivity 18 M a cm-I obtained using a Milli-Q water-purification system (Millipore) All glassware was kept in 10% nitric acid for at least 48 h and washed three times with ultrapure water before use.0.002 mg 1 - ' BDH. Procedure for Slurry Preparation Lyophilized marine sediment samples were ground to reduce them to a particle size of less than 250ym. A portion of the sample 0.25 g was weighed and placed in a polyethylene vial and zirconia beads ( 5 g) and 3 ml of water were added. The vial was agitated in a flask shaker (Vibromatic) for 60 min and then the beads were separated using a sieve funnel (Haldenwanger Technische Keramic Dusseldorf Germany). Finally the slurry was adjusted to its final volume (10 ml) by the addition of water and the amount of Triton X-100 necessary to obtain a concentration of 0.1% Triton X-100. A portion of the slurry with an appropriate amount of chemical modifier was transferred into the autosampler cup and stirred magnetic- ally before being measured.Digestion Procedure For the acid digestion 0.1 g of sediment sample was placed in a PTFE digestion bomb with 2 ml of concentrated nitric acid and 0.5 ml of concentrated hydrochloric acid. The bombs were heated through use of microwave energy for 3 min. When the bombs were cool the samples were made up with ultrapure water to a final volume of 10 ml. Results and Discussion Optimization of the Graphite Furnace Programme Experiments were carried out to determine the optimum temperatures and times for the drying charring and atomiz- ation steps. To obtain an efficient pyrolysis two charring steps were used; the first at 480 "C and the second at 900 "C. These conditions were optimized by means of several measurements for a slurry sediment sample and for aqueous standards both in the presence of palladium-magnesium nitrate.The final ashing curves are shown in Fig. 1 and it can be seen that in both cases 900°C can be used as the optimum charring temperature. A possible problem with the use of slurries in the graphite furnace is incomplete ashing of the organic matrix. To avoid this problem Ebdon et aZ.15 used air or oxygen as an ashing aid thus the normal charring step is converted into an oxidative decomposition process. In previous work on the determination of lead in mussel slurries Bermejo-Barrera et aL7 used an ashing step with air obtaining a significant decrease in the background signal. To establish the effect of ashing sediment slurries in the presence of air or oxygen a study of various air and oxygen flow rates in the first thermal pre-treatment step was performed.The results obtained are shown in Fig. 2. There is no advantage for the background absorption when air or oxygen are used in the thermal pre-treatment step probably owing to the low amount of organic matter present in sediment slurries. For this reason and to avoid rapid degradation of the graphite tube the introduction of air or oxygen during the charring step was not used. Determination of the optimum atomization temperature was i 0.420 0.315 v - 7 0 0 500 700 900 1100 1300 1500 Temperatu re/"C Fig. 1 standard Charring curves for A slurry sediment; and B aqueousJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 47 1 I B I 8 8 c o 0 50 100 150 200 250 300 Air flow rate ml/rnin-' +? s 2 0.16 ( b ) 0.12 0.04 0 I I I I I I 0 50 100 150 200 250 300 Oxygen flow rate/ml min-' Fig.2 Effect of the use of (a) air and (b) oxygen in the first ashing step A background; B signal minus background carried out by studying different atomization temperatures between 1000 and 2600 "C.The choice of this temperature was not made from the atomization curve because double lead absorption peaks were observed at some atomization tempera- tures and hence selection of the optimum temperature was not possible. The effects of atomization temperature on the double peak formation in a sediment sample slurry in the presence of palladium are illustrated in Fig. 3. The double peaks appear at atomization temperatures between 1300 and 2200 "C and at 2800°C; a single absorption peak is observed for atomization temperatures of between 2300 and 2700°C.The optimum atomization temperature chosen was 2500 "C because using this temperature there are no double peaks and the maximum peak area is obtained. This double peak formation by lead in the presence of palladium has been observed previously and the DabekaI7 has explained this behaviour as being due to the formation of refractory lead species in the presence of pal- ladium. However double absorption peaks for lead in the absence of palladium have also been observed,17 and have been attributed to physical effects such as background prob- lems or partial occlusion of lead within the sample matrix or to the presence of two or more lead compounds (Pb PbO) 0 0.5 1.0 1.5 2.0 Time/s i Fig.3 Effect of atomization temperature on double peak formation in the presence of palladium A 2800; B 2700; C 2600; D 2500 E 2400; F 2300; G 2200; H 2100; I 1900; J 1700; K 1500; L 1300; and M 1000°C Time/s Fig.4 Influence of palladium on the absorption profile of a slurry sediment sample A without modifier added integrated absorbance 0.157 s; and B with palladium-magnesium added integrated absorbance 0.175 s having different volatilities. Moreover Qiao and Jackson" have recently observed that the distribution of analyte in the sediment sample contributes to the double peak formation because of two different forms of lead being present in the sediments lead adsorbed on the clay particles and lead on the organic carbon.These workers established that pal- ladium-magnesium nitrate as a chemical modifier was not necessary to determine lead in soil slurries because the slurry particle itself acts as a modifier to stabilize the analyte during pyrolysis. However experience of the present workers has shown that when the modifier was not used a small shoulder was observed (Fig. 4). To compare the atomization of lead from the sample slurry and from an aqueous standard the atomization temperature for lead aqueous standards with added palladium-magnesium nitrate was studied. In this case the double peak was not observed. The absorbance obtained when the atomization temperature was increased from 1100 to 2700°C is constant. Hence the optimum atomization temperature for the sediment slurry 2500"C can be used because at this temperature there are no problems with the aqueous standards.Amount of Chemical Modifier A series of measurements were carried out to determine the optimum concentration of both palladium and magnesium nitrate to be used by adding different amounts of each one of them to a series of slurries in the absence of the other. For the series containing only magnesium nitrate the appearance time is shortened when the amount of magnesium nitrate is increased but the integrated absorbance is very similar in all cases Fig. 5. When only palladium was added the appearance time of the signal was very similar but the peak height increased with the amount of palladium added and the peaks were narrower Fig. 5. The same effects were observed for aqueous standards of lead.The use of palladium with mag- nesium nitrate is recommended because although the peak shown in Fig. 4 (peak B) with palladium-magnesium nitrate and the peak in Fig. 5(a) with palladium only are similar when only palladium was used the absorbance peaks obtained from slurry samples are shifted in time compared with the aqueous standards. This behaviour has been observed previously by Hinds and Jack~on.'~ This shifting effect is avoided when the mixture is used. This is important because aqueous calibrations are to be used. Variations in the integrated absorbance for different amounts of palladium and magnesium nitrate for sediment slurries and aqueous standards are shown in Fig. 6. The optimum amounts chosen were 2mgl-I of palladium and 15mg1-' of mag- nesium nitrate.Grinding Time of the Slurry The optimum grinding time using the zirconia beads was determined by measuring the lead content in several slurries472 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 0.18 v) -. 8 0.17 ([I 2 9 0.16 U 4- 2 0.15 4- - 0.14 Y 0 0.5 1 .o 1.5 2.0 Time/s Fig.5 Effect of the amount of modifier on the absorbance peaks of lead. (a) Effect of different amounts of palladium. Integrated absorbances 0.154; 0.162; 0.157; 0.165; 0.162 for A 0; B 3; C 2; D 4; and E 6mg I-' of palladium. (b) Effect of different amounts of magnesium. Integrated absorbances 0.154; 0.154; 0.151; 0.146 and 0.147 s for A 0; B 5; C 10; and D 15 and 20 mg 1- ' of magnesium nitrate of marine sediment sample and a control sediment that had been ground for 30,45,60 and 90 min.The results are presented in Fig. 7 where it can be seen that a grinding time of 60 min is adequate for both types of samples because a longer grinding time does not give a significant improvement in the absorbance signal. Particle Size With the above grinding procedure a particle size of less than 5 pm was achieved which was checked by means of laser diffraction. The particle sizes were studied in a sediment sample and control sediment using different grinding times. The results obtained are given in Table 2. It can be observed that with a grinding time of between 30 and 90 min there is no significant change in the particle size. On the other hand in Fig. 7 can be seen the particle size distribution for both types of samples and these distributions are very similar in all cases 90.00% of the particles have a size of less than 0.5 pm.Effect of Predigestion of Sediment In a study on slurries of solid samples Hoening et all6 used pre-digestion of the sediment in acid medium which then mobilizes several elements into solution after preparation of slurry. Thus only a portion of the trace elements present remained in the solid phase which is nevertheless dispensed together with the enriched solution into the atomizer during sampling. To study the effect of acid pre-digestion on a sediment sample experiments were performed where 0.1 g of sediment sample was treated with 50 ml of concentrated nitric acid and 950 pl of water for different periods of time before the prep- aration of the slurry.For pre-digestion times of 0 5 10 15 and 20 h absorbances of 0.123 0.125 0.123 0.123 and 0.124 respectively were obtained (n=4). The use of the acid pre- digestion does not produce any improvement in the absorbance of lead. 0.19 / P L 0 2 0.17 U r L? 0.16 4- - 0.15 5 Fig.6 Effect of amounts of palladium and magnesium on the inte- grated absorbance signal for (a) slurry sediment and (b) aqueous standard 0.14 0.12 ~ 9 1 I 0.06 ' I I I I I I 30 40 50 60 70 80 90 Timelmi n Fig.7 control; and B sample Effect of grinding time on the absorbance of the slurry A Effect of Different Thickening Agents The effect of thickening agents was studied by preparing slurries that contained Triton X-100 Viscalex HV 30 or glycerol as the stabilizing agent. Several different concen- trations of Triton X-100 between 0.04 and 1 % were evaluatedJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 473 Table 2 Effect of grinding time on particle size Grinding time/ Mean size/ SD*/ Certified Reference Material PACS-1- rnin Pm Pm 30 0.775 0.630 45 0.784 0.603 60 0.768 0.564 90 0.773 0.540 30 0.838 0.808 45 0.843 0.783 60 0.845 0.773 90 0.820 0.704 Sediment sample- 95% Confidence limit/ Pm 0.65 1-0.898 0.666-0.902 0.658-0.879 0.667-0.879 0.680-0.997 0.689-0.996 0.693-0.996 0.68 2-0.958 * SD =standard deviation. to determine which concentration provided the most suitable visual dispersion of particles; 0.04% was found to be sufficient. Also the variation of absorbance of a sediment sample slurry with concentration of Triton X-100 was studied. The results obtained are presented in Table 3 and were blank corrected for each Triton concentration; a concentration of Triton X-100 of 0.1% is more than adequate higher concentrations produc- ing a reduction in absorbance.The same study was carried out with Viscalex HV-30 but it was necessary to increase the hold time during the second drying step for 60 s for this thickening agent. The results obtained blank corrected for each Viscalex concentration are given in Table 3 and again 0.1Y0 was an adequate concentration. Similarly slurries were prepared with glycerol concentrations of 20 30,40 50 and 60% m/v and the results blank corrected for each glycerol concentration obtained are also shown in Table 3. A concentration of 40% glycerol would be the opti- mum concentration but with the use of this thickening agent there were some problems.To dry the sample completely it was necessary to alter the drying steps a first drying step at 150°C with 5 s ramp time and 50 s hold time being necessary and then a second drying step at 350 "C with 80 s ramp time and 40 s hold time. Under these conditions the duration of the graphite furnace programme is considerably increased. Moreover when glycerol was used a heavy smoke appeared on the decomposition of glycerol at about 400°C. The addition of glycerol and Viscalex also caused problems with the reproducibility of the autosampler pipetting because the sample solution adhered to the outside of the autosampler capillary and would sometimes enter the tube through the injection hole impairing precision. This problem has been observed previously by Stephen et aL4 and Miller-Ihli.3 For all these reasons Triton X-100 was selected as the optimum thickening agent.Calibration and Standard Additions Graphs To obtain a calibration graph to standard aqueous solutions containing lead concentrations of between 0 and 7 pg 1-1 were added appropriate volumes of solutions of palladium mag- nesium nitrate and Triton X-100 to give concentrations of 2 mg 1-' 15 mg ml-' and 0.1% respectively. The standard 12 8 4 1 s o 0.5 1.0 2 3 4 6 10 a c 100 700 12 8 4 0 I Table 3 Variation of absorbance with various concentrations of thickening agents Particle diametedpm Fig. 8 Particle size distribution as a result of stirring the slurry with zirconia beads for 60 min (a) marine sediment sample; and (b) sediment control PACS-1 additions method was used over the same range of concen- trations using a sediment sample.The equations obtained were as follows calibration graph &=8.22 X 10-4+6.51 X w 3 C r=0.998 standard additions graphs QA=4.66 x l0-'+6.86~ 1 0 - 3 ~ r=0.996 where QA is integrated absorbance and c is the lead concen- tration in pg I-'. Both graphs are shown in Fig. 8 where they exhibit similar slopes. This means that aqueous calibration is a real possibility and this type of calibration was therefore used. Calibration Using a Certified Reference Material Calibration employing the Certified Reference Material PACS-1 (Marine Sediment) which had a certified lead content of 404f20 mg kg-l was carried out and compared with that using aqueous standards. The equation obtained was as follows calibration with PACS-1 Q ~ = -3.39x 10-3+5.5ox 10-3c This calibration graph is also shown in Fig.8. It was concluded that aqueous calibration is more suitable than calibration with PACS-1 for this analysis. Concentration of Triton X-100 (Yo v/v) Absorbance 0.04 0.122 0.08 0.130 0.10 0.132 0.50 0.110 1 .oo 0.115 Concentration of Viscalex HV 30 Concentration of (Yo v/v) Absorbance glycerol (% m/v) Absorbance 0.04 0.130 20 0.115 0.08 0.133 30 0.119 0.10 0.130 40 0.117 0.50 0.126 50 0.108 1 .oo 0.127 60 0.109474 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 0.10 0.08 cu m 2 0.06 2 0.04 d e 0.02 0 1 Fig. 9 Calibration graphs A calibration with aqueous standards; B standard additions; and C calibration with PACS-1 Sensitivity The limit of detection (LOD) the lowest concentration level that can be determined to be statistically different from a blank is defined as LOD=3SD/m (where m is the slope of the calibration graph) corresponding to a 99% confidence level.The limit of quantification (LOQ) is defined as the level above which quantitative results can be obtained with a specified degree of confidence. At the 99% confidence level the value recommended is LOQ=lOSD/m. In both cases SD is the within-run standard deviation of a single blank determination. The values obtained were 0.22pgl-' for the LOD and 0.74 pg 1-l for the LOQ based on ten replicate determinations of the blank. The characteristic mass mo is defined as the mass of analyte in picograms required to give a signal of 0.0044 s for inte- grated absorbance.The characteristic mass obtained was 11.4 & 0.33 pg. The characteristic mass obtained was lower than the literature values as Hinds et a1.I4 found 12.5 pg for lead without a chemical modifier and 18.2pg when they used palladium-magnesium nitrate. In both cases they used a fast furnace programme. Schlemmer and Welz20 found 12pg for 15 pg of palladium+ 1.6 pg of magnesium and Hinds et a2." obtained 14 pg for 0.6 pg of palladium+ 1.0 pg of magnesium. The LOD for the sediment sample using 0.25 g of sample diluting the slurry to 10 ml and taking 25 pl to prepare the final solution with the modifier (1000 pl) was 35 pg kg-'. This limit can be improved by taking a larger volume of the slurry sediment to prepare the final solution.However to obtain the lowest LODs using the slurry procedure it is necessary to increase the concentration of the sample suspended in the slurry. When the slurry concentration was increased an increase in the peak area was observed. This can cause some problems because an increase in sample concentration is related to an increase in viscosity which can impair the precision of the sample introduction. The effect of sample concentration on the precision of the lead signals was studied by preparing slurries that contained 1.0 2.5 5.0 and 10% m/v of sediment. Amounts of aqueous lead solutions were added to each type of slurry to obtain similar absorbances. Seven replicate injections of each slurry were performed and the relative standard deviation (RSD) values calculated.The results obtained are shown in Table 4 and it can be seen that Table 4 Effect of slurry concentration on precision; seven replicates Slurry concentration Coefficient of variation (YO m/v) Mean absorbance (%) 1 2.5 5.0 10.0 0.119 2.6 0.122 2.3 0.126 1.4 0.124 2.2 the slurry concentration does not affect the precision. Moreover the use of higher slurry concentrations does not cause any problems with the background signal as for 1% m/v the background signal was 0.026 s integrated absorbance and for 10% was 0.030s. The LOD using a slurry sediment sample of 10% m/v and taking 25 pl of this solution would be 8.8 pg kg-'. Precision The within-run precision (RSD) of the method (instrumental and matrix factors) obtained for ten replicate analyses of a single sample during the same run was 2.3% (for 2.2 pg I-' of lead).The within-batch precision of the method obtained for ten replicates of three samples with different concentrations of lead added was also investigated. To study the within-batch pre- cision three samples with 3 5 and 7 pg 1-1 of added lead were used and the results were 2.3 1.6 and 2.1% respectively. The repeatability of the over-all procedure was also studied by measuring ten different slurries of the same sample. The RSD obtained was 5.0%. Accuracy To study the accuracy of the method the Certified Reference Material PACS-1 with a certified lead content of 404+20 mg kg-' was used. The reference material and the blank were measured three times.The results obtained expressed as mean & SD was 418 & 11 mg kg-'. The recovery was also calculated and was found to vary between 97.3 94.6 and 99.0% for 3 5 and 7 pg 1-1 of added lead. Applications The method was applied to the determination of lead in marine sediment samples from the Galicia coast (north west Spain). Two sub-samples taken from each sediment sample were prepared in the form of slurries and two sub-samples from each sub-sample were subjected to AAS. The results obtained were compared with those achieved when the samples were digested with nitric and hydrochloric acids in high-pressure bombs using microwave energy. Again two sub-samples from each sediment sample were digested and then two sub-samples from each were subjected to AAS. To compare the results obtained by the two methods the paired t-test was applied.21 The results obtained are given in Table 5.As shown as the calculated t-value is smaller than that obtained from the t-distribution table hence the two methods do not give significantly different values for the mean concentration of lead. Table 5 Comparison of the results obtained in sediment samples from the Galicia coast (Spain) Sample no. 1 2 3 4 5 6 7 Slurry method/ Digestion method/ mg kg-' mg kg-' 8.4 6.2 17.5 17.5 37.8 35.6 46.9 46.9 51.4 49.2 46.9 44.7 78.6 78.6 Mean difference= 1.257 mg kg-' SD of mean difference= 1.176 mg kg-I t = 2.828 Critical value of t (P = 0.01) = 3.71JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 475 Conclusions The use of slurry sample introduction in ETAAS is a convenient method for the determination of lead in marine sediments. The use of palladium-magnesium nitrate as a chemical modifier in atomization of the slurry sediment provides good stabilization of lead up to 900 "C and avoids the formation of double peaks.With the proposed procedure a particle size of less than 5 pm was obtained and with the use of the Triton X-100 as a thickening agent the precision and accuracy of the method are good. Aqueous calibration is considered to be more appro- priate than calibration using a certified reference material thus simplifying the determination. It can be concluded that the slurry sampling procedure is more suited to the determination of lead in marine sediment samples than a wet-digestion procedure because complex dissolution of the marine sediment sample is avoided and thus sample handling is minimal and possible sources of contamination are reduced.References Bendicho C. and de Loos-Vollebregt M. T. C. J. Anal. At. Spectrom. 1991 6 353. Miller-Ihli N. J. Fresenius' J. Anal. Chem. 1990 337 271. Miller-Ihli N. J. J. Anal. At. Spectrom. 1988 3 73. Stephen S. C. Ottaway J. M. and Littlejohn D. Fresenius' 2. Anal. Chem. 1987 328 346. Hoening M. and Van Hoeyweghen P. Anal. Chem. 1986 58 2614. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Littlejohn D. Stephen S. C. and Ottaway J. M. presented at SAC '86-3rd BNASS Bristol UK July 1986 paper number PF13. Bermejo-Barrera P. Aboal-Somoza M. Soto-Ferrero R. and Dominguez-Gonzalez R. Analyst 1993 118 665. Lynch S. and Littlejohn D. J. Anal. At. Spectrom. 1989 4 157. de Kersabiec A. M. and Benedetti M. F. Fresenius' 2. Anal. Chem. 1987 328 342. Hinds M. W. Katyal M. and Jackson K. W. J. Anal. At. Spectrom. 1988 3 83. Hinds M. W. and Jackson K. W. J. Anal. At. Spectrom. 1990 5 199. Yu Z.-q. Vandecasteele C. Desmet B. and Dams R. Microchim. Acta 1990 I 41. Ebdon L. and Lechotycki A. Microchem. J. 1986 34 340. Hinds M. W. Latimer K. E. and Jackson K. W. J. Anal. At. Spectrom. 1991 6 473. Ebdon L. Fisher A. S. Parry H. G. M. and Brown A. A. J. Anal. At. Spectrorn. 1990 5 321. Hoening M. Regnier P. and Wollast R. J. Anal. At. Spectrom. 1989 4 631. Dabeka R. W. Anal. Chem. 1992,64 2419. Qiao H. and Jackson K. W. Spectrochim. Acta Part B 1992 47 1267. Hinds M. W. and Jackson K. W. J. Anal. Atom. Spectrom. 1988 3 997. Schlemmer G. and Welz B. Spectrochim. Acta Part B 1986 41 1157. Miller J. C. and Miller J. N. Statistics for Analytical Chemistry Ellis Horwood Chichester 1986. Paper 3104471 H Received July 27 1993 Accepted October 1 I 1993