JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 691 Flame Atomic Absorption Spectrometric Determination of Cadmium in Biological Samples Using a Preconcentration Flow System With an Activated Carbon Column and Dithizone as a Chelating Agent Yaneira Petit de Pefia Mercedes Gallego and Miguel Valcarcel* Department of Analytical Chemistry Faculty of Sciences University of Cordoba I4004 Cordoba Spain A combined flow atomic absorption spectrometric system was used to develop an efficient on-line preconcen- tration-solvent elution procedure for the determination of trace amounts of cadmium. The metal was preconcentrated as the dithizonate on a mini-column packed with activated carbon and eluted with isobutyl methyl ketone (4-methylpentan-2-one) into the nebulizer. A preconcentration factor of 40-1 30 equivalent to 3-12 ml of sample was achieved by using a time-based technique.The detection limit obtained (30) ranged between 0.3 and 1.3 ng ml-' and the relative standard deviation from 1.3 to 2.4% for a sample volume of 12 and 3 ml respectively. The results obtained for the determination of cadmium in reference materials testify to the applicability of the proposed procedure to the analysis of biological materials. Keywords On-line sorption-extraction; activated carbon mini-column; cadmium; biological material; flow injection flame atomic absorption spectrometry Knowledge of the detrimental effects of trace and ultra-trace amounts of metals in the environment and biological materials such as blood and animal tissues has greatly increased in recent years.Because of the extremely low concentrations of metals in these matrices a preliminary concentration step is usually necessary before their determination. There are cur- rently two major preconcentration methods used uiz. off- and on-line. The separation of metal ions from liquid samples in both methods can be obtained in several ways; however one of the most commonly used approaches involves allowing the sample to flow through a column packed with an active material. Preconcentration is accomplished by adjusting the sample pH to an appropriate value and using a chelating agent to interact with the metals of interest before they are retained or immobilized on the active material. A literature scan- search on enrichment of heavy metals on activated carbon by off-line procedures revealed that this is usually carried out after chelation with ammonium pyrrolidinedithiocarbamate (APDC ammonium pyrrolidin-1-yl dithioformate),' 8- hydroxyquinoline ( 8-HQ),2 potassium ethyl anth hate,^^^ dithi- zone,5 chrome azurol S6 or the ammonium salt of the dithi- ophosphoric acid 0,O-diethyl ester;7 also following desorption in a small volume of nitric acid the metal concentrations are measured by atomic absorption spectrometry (AAS) or (as a slurry) by inductively coupled plasma atomic emission spec- trometry (ICP-AES).Activated carbon has proved to be an excellent collector for ~admiurn~-~ after formation of neutral chelates. Dithizone has been used as a chelating agent for collection of trace amounts of cadmium on a thin layer of activated ~ a r b o n ; ~ the metal chelate is released from the activated carbon layer by digestion with 14 moll-' nitric acid then the nitric acid is evaporated and the released trace metal element is collected in a small volume of dilute acid.Recoveries above 90% can be obtained at hydrochloric and nitric acid concentrations below 0.005 and 0.002 moll- ' respectively; the detection limit is 0.3 ng ml-' of cadmium and the preconcen- tration factor is high (1 1 of water can be reduced to 1 ml). Two classes of column materials have been used for precon- centration of cadmium and matrix removal in flow injection analysis viz. ion exchanger^^'^ and sorbent materials such as C18;1s22 there is only one reference to the use of activated carbon impregnated with 8-HQ in continuous systems for preconcentration of cadmium by ICP-AES (detection limit 0.25 ng ml-' of cadmium).23 The present paper describes a new method whereby trace ~ * To whom correspondence should be addressed.amounts of cadmium are preconcentrated as a metal chelate using the chelating dye dithizone on an activated carbon mini- column included in the flow injection (FI) system. The chelate is eluted with a small volume of a water-immiscible solvent and the analyte does not disperse on transfer to the flame atomic absorption spectrometry (FAAS) instrument which increases the preconcentration factors. The system was used successfully for the determination of cadmium in biological reference materials. Experimental Apparatus A Perkin-Elmer 380 atomic absorption spectrometer equipped with a bead impact system in the burner chamber and a hollow cathode cadmium lamp was used.The wavelength and lamp current used were 229 nm and 4 mA respectively; deuterium arc background correction was employed throughout. The acetylene flow rate was 2.01 min-' and an air flow rate of 21.5 1 min-' was employed to obtain a clean blue flame. The flow system consisted of a Gilson-Minipuls-2 peristaltic pump furnished with poly(viny1 chloride) tubes two Rheodyne 5041 injection valves and a laboratory-made adsorption mini- column packed with 70mg of activated carbon. The mini- column (2.5 cm in length and 3 mm i.d) was made of poly (tetra- fluoroethylene) (PTFE) capillary and sealed at one end with a small glass-wool bead to prevent losses of material.The column was initially flushed with 0.1 moll-' nitric acid and the subsequent use of isobutyl methyl ketone (IBMK) as eluent in each operating cycle was sufficient to make it ready for re-use. Peak heights and areas were measured with a Merck-Hitachi D-2500 Chromato-Integrator. A Hetosicc Freeze Dryer Type CD-53-1 was also employed. Reagents and Standard Solutions All chemicals used were of analytical-reagent grade and the water was ultrapure (Milli-Q Water System Millipore Seville Spain). A 1000 mg 1-' cadmium stock solution was prepared by dissolving 1.000 g of the metal in a small volume of concentrated nitric acid and diluting to 1 1 with 1% (v/v) nitric acid. A saturated solution of dithizone (Riedel de Haen Hannover Germany) was prepared as follows 5 mg of dithi- zone were shaken electromagnetically in a 100 ml vessel con- taining 50 ml 0.4 moll-' ammonia solution for 3 min.The solution was then filtered and the filtrate diluted with water in a 100 ml calibrated flask. Isobutyl methyl ketone (Probus,692 JOURNAL OF ANAJLYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 Barcelona Spain) was also used. Darco 20-40 activated carbon (Aldrich Barcelona Spain) and polygosyl bonded silica reversed-phase sorbent with octadecyl functional groups (RP-C,,) 60-100 pm (Millipore) were employed as sorbents. Standard solutions (100 ml) containing 0.5-70 ng ml-' of cad- mium were all freshly prepared in 0.01 mol I-' sulfuric acid by appropriate dilution of the stock standard solution (1000 mg I-').Procedures Sample preparation The reference materials analysed were as follows Pig Kidney (Community Bureau of Reference BCR No. 186) City Waste Incineration Ash (BCR No. 176) Oyster Tissue [National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 1566a1 and Lobster Hepato- pancreas Marine (National Research Council Canada TORT-1). The reference materials were dried to constant mass by freeze drying at 6 Pa (0.04 mmHg) for 24 h after which an accurately weighed amount of 250-500 mg was digested24 with 10 ml of 65% nitric acid and 1 ml of 24.5% sulfuric acid in glass digestion tubes. The mixture was heated at 150-200 "C in a digestion block until the sample was completely dissolved and nitrogen fumes were given off. The tubes were allowed to cool for about 2min and the digestion was repeated (about six times) in the same way with multiple additions of 5 ml of nitric acid until a clear solution remained and appearance of nitrogen dioxide fumes had ceased.Once cold the solution was transferred quantitatively into calibrated flasks of 50 or 100ml capacity and made up to volume with ultrapure (Milli-Q) water. A reagent blank was prepared in parallel. Sub-samples of between 1 and 25 ml diluted to 50 or 100 ml and a pH of 2.0 adjusted with 2 moll-1 sulfuric acid were analysed immediately after preparation by introducing them into the manifold shown in Fig. 1. Continuous preconcentration-elution system The FI manifold for on-line preconcentration and elution of cadmium(I1) is illustrated in Fig.1 together with the optimized operating parameters. The sample or standard containing 0.5-70 ng ml-' of cadmium(11) in 0.01 moll-' sulfuric acid was continuously pumped through the manifold for 1-4 min and thoroughly mixed with the chelating reagent (saturated dithizone in 0.2 moll-' ammonia solution). After merging retention of the chelate took place on the activated carbon column placed in the loop of the injection valve and the sample matrix was sent to waste (W). During this step a water carrier was pumped to the instrument in order to flush the nebulizer after each measurement. The preconcentration step was termin- ated when the two injection valves were switched simul- taneously so elution of the adsorbed chelate took place when 150 pl of IBMK were passed through the adsorbed chelate to desorb it and sweep the cadmium to the detector. The duration of the elution step was set to 20 s.The peak height absorbance of the elution signal for quantification and a blank of 150 pl of IBMK injected prior to sample preconcentration were used (about 0.050 absorbance units). In this step the sample was also changed whereby the remainder of the previous sample in the pump tube was driven to waste and the next sample made ready for preconcentration. Results and Discussion Selection of Chelating Reagent and Eluting Solvent Activated carbon has been used as a trace collector for multi- element preconcentration by simply adjusting the pH to an appropriate value or using a chelating agent. Best results are obtained when the metals are complexed with organic chelating (AAS) Sample mI rnin-' 300cm Column )W2 H 2 0 4.0 Dithizone 0.3 Sample 3.0 Fig.1 Schematic diagram of the assembly used for on-line preconcen- tration of Cd. (a) and (b) adsorption and elution step respectively. Bold lines denote flows relevant to the individual stages. IV Injection valve; W waste; IBMK isobutyl methyl ketone agents prior to adsorption on the activated carbon. In order to test this approach a variety of chelating agents were assayed in an FI system similar to that shown in Fig. 1 namely 8-HQ dithizone 4-( 2-pyridylazo)resorcinol (PAR) ammonium diethyldithiocarbamate (NH,DDC) and APDC. The reagents at 0.1 % (m/v) concentration were prepared in ultrapure water except for dithizone (sparingly soluble in water) which was dissolved in ammonia solution (5 mg in 100 ml of 0.2 moll-' ammonia solution).Several calibration graphs were run for cadmium by using the above reagents; the sensitivity (slope of the calibration graph) achieved by using dithizone was about 20 40 60 and 100% higher than that obtained with APDC NH,DDC 8-HQ and PAR respectively. So dithizone was selected as the chelating agent. Elution of the adsorbed chelate from the column was investigated with different solutions. The solvents used for this purpose were 0.1 moll-' sulfuric acid ethanol acetone carbon tetrachloride chloroform and IBMK. By using the automatic configuration shown in Fig. 1 and 6 ml sample volumes containing 15 ng ml-' of cadmium(11) in 0.01 moll-' sulfuric acid and injecting 150 pl of the extractant the above mentioned eluting reagents were assayed in order to select the fastest.Solvent changeovers required flushing the column with IBMK in order to remove the residual adsorbed chelate. As can be seen in Fig. 2 the best results (difference between the sample and blank) were provided by IBMK and chloroform. The final choice was IBMK because it was less toxic than chloroform and released no hydrogen chloride in the flame; in addition it gave rise to a lower blank signal. Other extractants (sulfuric acid ethanol and acetone) provided worse results either because they were water miscible and the plug underwent dispersion or because the chelate could not be dissolved. Peak heights were selected as the analytical measurements because elution with IBMK was instantaneous.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL.9 693 il 0.100 I - 1 v * LJL ( a ) I I I----I 30 s Time - Fig. 2 FI peaks obtained with different eluting solutions. (a) and (b) blank and sample signal respectively. Sample concentration 15 ngml-' of Cd. Peaks (peak areas in FV s are given in parentheses for sample and blank respectively) 1 = 0.1 mol I-' H,SO (715; 120); 2 = ethanol (2740; 3719); 3 = acetone (7227; 4392); 4 =carbon tetra- chloride (3950; 2952); 5 = IBMK (8227; 4217); 6 = chloroform (7050; 2946) Optimization of Chemical Variables Initially an attempt was made to dissolve dithizone which is moderately soluble in dilute ammonia solution ethanol ace- tone chloroform and carbon tetra~hloride.~~ Water immiscible organic solvents were discarded since the aqueous sample and chelating agent should be miscible.The best results were obtained by dissolving 5 mg of dithizone in 100 ml of dilute ammonia solution or 20 + 80 (v/v) ethanol-water. The effect of the sample pH was studied by introducing 6ml (sample flow rate 3.0 ml min-'; pumping time 2 min) of a solution contain- ing 15 ng ml-' of cadmium into the system the pH being adjusted to 1.0-5.0 with dilute sulfuric acid. As can be seen in Fig. 3 the maximum chelate adsorption was achieved at pH 2.0 and 4.0 for dithizone in 0.2moll-' ammonia solution and 20% ethanol respectively since the ammoniacal medium allowed the reagent to be acidified to a greater extent even though the pH obtained on mixing the sample with the reagent was similar in both cases (pH about 2.5).Ammonia solution was therefore selected to dissolve dithizone because it enabled work to be carried out at a lower pH than the ethanol solution which is an obvious advantage in terms of selectivity with a view to future applications to real samples. The influence of various acids on the preconcentration reaction was also studied I 0.05 1 d I - 0 1 2 3 4 5 PH Fig.3 Effect of pH on Cd absorbance as measured after on-line preconcentration with dithizone in A 0.2 moll-' ammonia solution or B 20% v/v ethanol-water. Sample 15 ng ml-I of Cd; dithizone 5 mg in 100 ml of solvent by using samples containing 15 ng ml-I of cadmium at pH 2.0 adjusted with nitric sulfuric hydrochloric or perchloric acids. The absorbance decreased by 60% with nitric and hydrochloric acids and 80% with perchloric acid relative to that in sulfuric acid.The effect probably arose from the cadmium-dithizone chelate being protonated at an acid pH and sulfate ion being involved in the coordination sphere thereby favouring adsorp- tion of the chelate. Thus the aqueous samples were all prepared in 0.01 moll-' sulfuric acid. The effect of the dithizone concentration was studied in the range 4 x 10-5-4 x mol 1-' by dissolving different amounts (1-10 mg) in 100 ml of ammonia solution at various concentrations (0.1-0.4 moll-'). The signal increased with increasing dithizone concentrations up to 1.5 x moll-' (4 mg of dithizone in 100 ml) above which it remained constant because the solution was saturated. The signal also increased with increasing ammonia concentrations up to 0.2 mol 1-' at any of the dithizone concentrations assayed; however above 0.25 moll-' of ammonia the signal started to decrease as a result of the reaction being hindered by the increased alkalinity in the sample-dithizone reaction coil.A dithizone concen- tration of 2 x 1 0 - ~ mo1 1-1 ( 5 mg in 100 ml) in 0.2 mol I-' ammonia solution was finally chosen. The influence of temperature from 20 to 60°C on the preconcentration of 15 ng ml-' of cadmium was studied by thermostating the preconcentration reactor located before the column and the activated carbon column. The absorbance decreased slowly with increasing temperature above 30 "C (by 10 and 30% at 40 and 50°C respectively relative to room temperature) because probably the cadmium-dithizone com- plex decomposed so measurements were made at room tem- perature.Replacing the sample stream with 0.01 mol 1-1 sulfuric acid (blank) resulted in a similar absorbance in the eluting step to that obtained by successively injecting 150 pl of IBMK solvent before the preconcentration step; no blank (0.01 moll-' sulfuric acid) was therefore required. FI Conditions for Preconcentration-Elution The FI system was optimized by the univariate approach. The flow variables studied for a sample of 15 ng ml-' were the sample and reagent flow rate the length of the preconcen- tration coil and the volume of IBMK in addition to the IBMK flow rate for the preconcentration and elutions step. Changes in the flow rate of the sample (6.0 ml of a solution containing 15 ngml-' of cadmium) between 0.6 and 4.0mlmin-' at a constant reagent flow rate of 0.3 ml min-' resulted in increas- ing the analytical signals up to 1.5 mlmin-' because the sample was less markedly diluted at the higher flow rates (Fig.4). Very small variations were observed in the range 1.5-3.0 ml min-'. The signal decreased at flow rates above 3.0 ml min-' because the dispersion increased and also decreased the residence time. Once the dithizone solution was saturated increasing the flow rate was equivalent to increasing the dithizone concentration; however the sample was diluted 0 1 2 3 4 Sample flow rate/mt min-' Fig. 4 Dependence of the efficiency of preconcentration on the sample flow rate. Sample and dithizone concentrations as in Fig. 3694 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL.9 Table 1 Figures of merit of the determination of cadmium by preconcentration on activated carbon mini-column Detection limit/ Sampling frequency/ Enrichment factorf- Time/min Regression equation* Range/ng ml-I ng ml-' RSD (%) h-' 1 A = -0.002 + 0.006~ 3-70 1.3 2.4 40 40 2 A = 0.003 + 0.01 IX 1-40 0.6 1.7 25 80 2 A = 0.003 + 0.007~$ 2-60 1.1 2.2 25 50 4 A = 0.001 +0.019~ 0.5-22 0.3 1.3 113 130 * A absorbance and x cadmium concentration (ng ml-'). t Compared with conventional sample introduction of an aqueous solution (A = 0.006 + 1.42 x 10-4~). $ By using an RP-CI mini-column. simultaneously so the atomic absorption signals decreased with increase in the dithizone flow rate from 0.3 to 0.9 ml min-l. A reagent flow rate of 0.3 ml min-' and a sample flow rate of 3.0mlmin-' were chosen as a compromise.Maximum chelate adsorption was achieved at a length of the preconcentration coil (located before the activated carbon column in the loop of the injection valve) of 250cm as the chelate was completely formed at the resulting residence time (k 9 s). Longer lengths up to 430 cm gave rise to constant signals. With coils longer than 430 cm the chelate formed was carried along the system and probably adsorbed on the inner walls of the PTFE tubing (Le. spread over a large surface area) so it was incompletely dissolved by the injected IBMK and the analytical signal decreased as a result. A coil that was 300cm long and 0.5 mm in i.d. was selected for further experiments (residence time 11 s).The volume of IBMK played an important role in the chelate elution its effect being studied between 40 and 300 yl. The adsorbed chelate was found to be eluted throughout this range but some carryover was observed below 100 pl; at volumes above 150 pl the signal decreased through dilution of the analyte in the solvent. Accordingly experiments were carried out with 15 ngml-' of cadmium (the method can be applied to higher concentrations of cadmium) the volume of IBMK finally used being 150 pl in order to avoid carryover and allow usage of only one injection per sample. Since water acted as the carrier for injected IBMK it was essential to examine the effect of its flow rate. The peak area remained constant over the range 1-6mlmin-l but the peak height increased with increasing water flow rate up to 3.5 ml min-' because the nebulizer efficiency also increased in parallel.A water flow rate of 4.0 ml min-' was thus chosen. Determination of Cadmium Several calibration graphs for cadmium(11) were run by using sampling times of 1 2 and 4 min which is equivalent to using sample volumes of 3.0 6.0 and 12.0m1 respectively; higher enrichment factors were obtained at higher sample volumes. The correlation coefficients obtained ranged between 0.998 and 0.999 in all instances. The detection limit was calculated as three-fold the standard deviation of the peak height Table3 Results obtained for the determination of cadmium in reference materials. All values are given in pg g-' of cadmium Certified Found Reference material value (n=3) City Waste Incineration Ash (BCR No.176) 470 _+ 9 466 11 Pig Kidney (BCK No. 186) 2.71 k0.15 2.95f0.44 Oyster Tissue (SRM 1566a) 4.15k0.38 3.97L0.36 Lobster Hepatopancreas Marine (TORT-1) 26.3 & 2.1 28.6 & 2.1 absorbance for 15 injections of 150 pl of IBMK (blank). The precision of the method [expressed as the relative standard deviation (RSD)] was checked on 11 samples containing 5 or 20 ng ml-' of cadmium each at different time-based sampling of 2 and 4 or 1 min respectively. Preconcentration factors of up to 130 calculated as the ratio between the slopes of the calibration graphs provided by this method and by direct aspiration of cadmium(Ir) were achieved for a sample volume of 12 ml (sample flow rate 3.0 ml min-'; pumping time 4 min).A comparative study of the chelate retention on an activated carbon mini-column and a reversed-phase silica sorbent (RP-CI8) mini-column both of the same dimensions was carried out with the column placed after the 300 cm preconcen- tration PTFE coil in the loop of the injection valve. The characteristic parameters of the calibration graphs are listed in Table 1 from which the following conclusions can be drawn (a) at the same sample volume 6.0m1 the sensitivity (slope of the calibration graph) was 1.6 higher for activated carbon than for the RP-C18 sorbent; (b) the linear range was wider for the RP-Cl8 mini-column; and (c) such analytical features as the detection limit and precision were similar in both instances but the enrichment factor for the activated carbon mini-column was more favourable. Because of the non-specific character of the complexant dithizone the effect of the most common ions that react with it were investigated in order to identify potential interferences.The cations investigated included Cu2+ Ni2+ Co2+ Zn2+ Pb2+ Mn2+ Sn2+ Hg2+ F e 3+ B'3+ 1 and A13+ which were tested at concentrations up to 5 yg ml-l for a sample containing 5 ng ml-' of cadmium (sample loading time 2 min). Normally these cations are tolerated at high concen- trations owing to the high selectivity of the atomic spectro- Table 2 Tolerated concentrations of foreign cations in the determination of 5 ng ml-' of cadmium Ion cu2+ Bi3 + Sn2 + Fe3 Hg2+ Ni2 + Zn2 + co2+ Pb2+ Mn2+ ~ 1 3 + Metal concentration/ pg ml-I 0.30 0.50 0.50 0.75 0.75 1 .oo 1 .oo 1.50 2.50 5.00 5.00 Tolerated ratio [metal] [Cd] 60 100 100 150 150 200 200 300 500 1000 1000 Metal concentration/pg ml- ' 0.5 0.75 0.75 1 .o 1 .o 2.5 2.5 2.5 5.0 Signal supression (%) 56 14 55 10 16 56 12 39 27 -Table 4 Features of automatic-preconcentration methods for the determination of cadmium by atomic spectrometric techniques Detection FAAS FAAS FAAS FAAS FAAS FAAS FAAS ETAAS I/ ETAAS FAAS ETAAS ETAAS FAAS FAAS ETAAS ICP-AES ICP-AES ICP-AES ICP-AES Preconcentration method I-E* (Chelex-100) I-E (Resin-122) I-E (Chelex-100 S-HQ,? Resin-122) I-E (TriPEN)t$ I-E (Chelex-100) I-E (Chelex-100) I-E (Several resin) I-E (Amberlite XAD-2R Sorbent extraction (cis)** Sorbent extraction (c18)?? Sorbent extraction (c18)tt Sorbent extraction (C,,)$$ Sorbent extraction (CIS)?? Sorbent extraction@ Biosorption (Alga)? Co-precipitationv Co-precipitation 11 11 I-E (8-HQ)t I-E (IDAECE Sampling Eluting solvent Preconcentration factor frequency/h- RSD (%) Application Ref.2 moll-' HNO 2moll-' HNO 2 rnol I-' HN0 1 mol I-' HCl+0.1 moll-' HNOJ 6 mol I-' HNO 2moll-1 HNO 2 mol I-' HNO 2 rnol I-' HNO 0.2 moll-' HCl 2.0 mol I-' HCI+O.l mol 1-1 HNO Acetonitrile Ethanol Ethanol or methanol Methanol Ethanol 0.5 moll-' HCl 0.1 moll-' HCl IBMK IBMK 20 20-28 50-100 500 30 15 2-4 100 125 20 17 18-25 50 5-450 16-19 - 500 16 43-52 30-60 40 60 2 30 24 10 12 12 23 120 7 23 35 20 24 3-150 - - - 1.5-4.1 1.2-3.2 - - 3 < 1.7 0.8-1.2 1.7 5-10 4 3.3 1.4 2 2.0-3.3 1.1 1.5 1.5 5.9 Sea-water Water Tap water - - - Biological standard material Sea-water urine River water Antartic sea-water Sea-water Sea and river water SRM Sea and drinking water Sea-water SRM Sea-water and river water SRM Biological materials Biological samples Whole blood digests - - 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 26 27 28 * I-E ion exchange. t Immobilized on controlled pore glass.$ TriPEN N,N N'-tri( 2-pyridylmethy1)ethylene diamine. 9 IDAEC iminodiacetic-ethylcellulose. 7 XAD-2 resin functionalized with 1-( 2-thiazolylazo)-naphth-2-ol. 11 ETAAS electrothermal atomic absorption spectrometry. ** Chelation by pyrrolidin- 1-yl-dithioformate ( pyrrolidinedithiocarbamate) with subsequent adsorption. tt Chelation by DDC with subsequent adsorption. $$ Chelation by APDC with subsequent adsorption. & Active carbon impregnated with 8-HQ.M[ Coprecipitation on the iron@-hexahydroazepinium hexahydroazepine-1-dithiocarboxylate. )I /I Coprecipitation on the iron@)-hexahydroazepine-1-dithiocarboxylate.696 JOURNAL OF ANALYTICAL ATO'MIC SPECTROMETRY JUNE 1994 VOL. 9 metric technique (Table 2). Only copper@) interfered at concentration ratios below 100. In all instances the interferent decreased the cadmium signal because the dithizone concen- tration was inadequate for the chelates of both the interferent and the cadmiurn to be formed so formation of the foreign cation chelates at higher concentrations Was favoured or the IBMK volume used was inadequate for eluting all the chelates adsorbed. Manganese@) and aluminium(I1I) were both toler- ated at the highest concentrations assayed.Applications The accuracy of the proposed method for the analysis of biological samples was tested by determining cadmium in Pig Kidney Oyster Tissue Lobster Hepatopancreas Marine and City Waste Incineration Ash. Each sample was mineralized in duplicate as described previously (see Procedure) together with a similarly prepared blank. Each dissolved sample was analysed in triplicate. The blank absorbances corresponded to a cadmium concentration of less than 0.4 ng ml-I (this blank allowed the contribution of cadmium ion present in the reagents to the digestion sample used to be assessed). The results obtained are shown in Table 3. As can be seen consistent values with certified values were obtained. Conclusions The results obtained in this work testify to the applicability of on-line sorbent-extraction preconcentration to FAAS for the determination of cadmium in complex matrices such as biologi- cal materials. The proposed method is fairly fast and simple; also taking into account the lack of dithizone selectivity it could probably be applied to other cations such as lead zinc copper and mercury.The features of other automatic methods for the preconcentration and determination of cadmium by atomic spectrometry are summarized in Table 4. Different ion- exchange chelating resins have been used with a water-miscible eluent (dilute acids) which results in the analyte being dispersed on transfer to the detector. The preconcentration factors achieved range from 20 to 500; however an enrichment factor of 500 was achieved from 100 ml of sample by using a sampling time of 25 min'l but the immediate result of using longer sampling times was obviously a decreased sample throughput.The proposed system is clearly superior to existing continuous- flow alternatives using C18 mini-columns since (a) both the sample matrix and the eluent are sent to waste during the preconcentration step; (b) the preconcentration reactor and the column are located in the loop of the injection valve so the eluent sweeps any chelate potentially adsorbed by the reactor thereby avoiding any carry-over; and (c) the eluent (IBMK) is water immiscible and the analyte is not dispersed on transfer to the detector which results in higher preconcen- tration factors (about 20 for the methods shown in Table4) versus 40- 130 for the proposed method.The coprecipitation of cadmium with iron complexes provides lower preconcen- tration factors even though IBMK is also employed as the eluting solvent. The methods shown in Table4 have usually been applied to water samples (there are only four applications to biological materials). The Comision Interministerial de Ciencia y Tecnologia is acknowledged for financial support (Grant No. PB 93-0717). Y.P. de P. is also grateful to the University of Cordoba the University of Los Andes and Consejo Nacional de Investigaciones Cientificas y Tecnol6gicas (Venezuela) for additional financial support. Professor R. E. Santelli is also acknowledged for his invaluable suggestions. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 References Kimura M.and Kawanami K. Talanta 1979 26 901. Vanderborght B. M. and Van Grieken R. E. Anal. Chem. 1977 49 311. Kimura M. Talanta 1977 24 194. Devi P. R. and Naidu G. R. K. Analyst 1990 115 1469. Beinrohr E. Rojcek J. and Garaj J. Analyst 1988 113 1831. Ambrose A. J. Ebdon L. and Jones P. Anal. Proc. 1989,26,377. Monte V. L,. A. and Curtius A. J. J. Anal. At. Spectrom. 1990 5 21. Olsen S. Pessenda L. C. R. RiiiiEka J. and Hansen E. H. Analyst 1983 108 905. Fang Z. Xu S. and Zhang S. Anal. Chim. Acta 1984 164 41. Fang Z. RGiEka J. and Hansen E. H. Anal. Chim. Acta 1984 164 23. Malamas F. Bengtsson M. and Johansson G. Anal. Chim. Acta 1984 160 1. Bengtsson IM. Malamas F. Torstensson A. Regnell O. and Johansson G. Mikrochim. Acta 1985 3 209. Hartenstein S. D. REiEka J. and Christian G. D. Anal. Chem. 1985 57 21. Hirata S. Umezaki Y. and Ikeda M. Bunseki Kagaku 1986 35 106. Caroli S. Alimonti A. Petrucci F. and Horvath Z. Anal. Chim. Acta 1991 248 241. Purohit R. and Devi S. Analyst 1991 116 825. Porta V. Sarzanini C. Abollino O. Mentasti E. and Carlini E. J. Anal. At. Spectrom. 1992 7 19. Porta V. Abollino O. Mentasti E. and Sarzanini C. J. Anal. At. Spectrom. 1991 6 119. Sperling M. Yin X. and Welz B. J. Anal. At. Spectrom. 1991 6 295. Fang Z. Guo T. and Web B. Talanta 1991 38 613. Liu Z.-S. and Huang S.-D. Anal. Chim. Acta 1992 267 31. Welz B. Yin X. and Sperling M. Anal. Chim. Acta 1992 261 477. Okamoto Y. Murata A. and Kumamaru T. Anal. Sci. 1991 7 879. Adeloju S. B. Bond A. M. and Briggs M. H. Anal. Chem. 1984 56 2397. Burger K. Organic Reagents in Metal Analysis Pergamon Press Oxford 1973 pp. 118-124. Elmahadi H. A. M. and Greenway G. M. J. Anal. At. Spectrom. 1991 6 643. Welz B. Xu S. and Sperling M. Appl. Spectrosc. 1991 45 1433. Fang Z. and Dong L. J. Anal. At. Spectrom. 1992 7 439. Paper 3/05 704 F Received September 21 1993 Accepted February 14 1994