Analyst, May, 1978, Vol. 103, @p. 447-451 447 Influence of Ascorbic Acid on the Matrix Interferences Observed During the Carbon Furnace Atomic-absorption Spectrophotometric Determination of Lead in Some Drinking Waters J. G. T. Regan and J. Warren Department of Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, Lon don, SE19NQ Nine samples of drinking water taken from a range of locations in England and Scotland have been analysed for lead by using carbon furnace atomic- absorption spectrophotometry. Spiking experiments have been carried out in order to determine the severity of the matrix interference. The suppression of the lead signals ranged from 22 t o 84%. No relationship was found to exist between the hardness of a water sample and its suppression effect.Further spiking experiments carried out in the presence of 1% m / V of ascorbic acid showed that the suppression effect of eight of the water samples was reduced to a level of less than 5%. The remaining water sample gave a suppression of 18%. This water was not the hardest examined, nor did it give the highest suppression in the previous experiment, The natural lead contents of the nine waters were determined both by carbon furnace atomic-absorption spectrophotometry in the presence of ascorbic acid and by a method that involves solvent extraction - flame atomic absorption. Statistical analysis, using a t-test, indicated that there was no significant difference (at the 95% confidence level) in the results obtained by using the two techniques.Keywords : Lead determination ; drinking water ; carbon furnace atomic- absorption spectrophotometry ; matrix interferences ; ascorbic acid The analysis of drinking waters for lead by carbon furnace atomic-absorption spectrophoto- metry is made difficult by severe suppression of the lead signal caused by sample matrix constituents. The use of the method of standard additions appears to be an obvious way of surmounting the difficulty but this procedure is time consuming and, with signal suppressions as high as 84%, may still not yield a result. The addition of sulphuric acid or phosphoric acid partially overcame the effects due to magnesium but increased those due to calcium. Campbell and Ottawayl suggested that analysis in the presence of 10% V/V nitric acid overcomes the suppression of lead by calcium.It was found, however, that while this suppression was reduced to the 10-15% level, a severe loss of sensitivity, attack on atomiser components and spreading on drying, with resultant loss of precision, also occurred. Thompson et al.2 have proposed the use of lanthanum-treated tubes for the analysis of water for lead and cadmium and reported a reduction of suppression effects to the & 7% level. Our earlier paper3 on the use of ascorbic acid showed that its addition at the 1% m/V level overcomes numerous single-element interferences on the lead signal. Some further work has been carried out on the use of ascorbic acid in the determination of lead in some drinking waters in order to study its effectiveness in overcoming multiple interferences arising from a natural, rather than a synthetic, matrix.Various attempts to overcome these matrix effects were made, but without success. Experimental Apparatus A Perkin-Elmer HGA-72 heated graphite tube atomiser installed in a Perkin-Elmer 403 atomic-absorption spectrophotometer fitted with the optical modification was used together Crown Copyright.448 REGAN AND WARREN: INFLUENCE OF' ASCORBIC ACID ON MATRIX Analyst, VoZ. 103 with a Telsec chart recorder of 10 mV f.s.d. and a response time of 0.3 s for full-scale deflec- tion. Reagents Lead nitrate. Nitric acid (sp. gr. 1.42). Ascorbic acid. Specpure (Johnson Matthey Chemicals Ltd.). Aristar (BDH). Standard Laboratory Reagent (Fisons) . Operating Conditions All lead peak-height determinations were made on 50-pl volumes and the mean (coefficient of variation better than 5%) of at least four replicate injections was taken for each solution.Measurements were made using the lead absorbing line at 283.3 nm and a spectral band width of 0.7 nm (slit 4). The thermal programme used was: drying at 100 "C for 45 s; thermal destruction at 450 "C for 30 s; atomisation at 2 080 "C for 7.5 s; and maximum temperature burn-out for 5 s; all temperatures as shown on the instrument indicator. Auto- matic background correction was used throughout. The gas-stop facility was not used. All samples were made 0.015 M with respect to nitric acid on collection. Effect of Sample Matrix on the Lead Signal In order to determine the effect of the sample matrix on the lead signal a spiking experi- ment was carried out.Three millilitres of distilled, de-ionised water were added to an acid-washed, dry, 50-ml calibrated flask and sufficient sample was added to make up to the mark. The spiked sample was prepared by adding 2 ml of distilled, de-ionised water and 1 ml of a 2.5 pg ml-1 solution of lead (as nitrate) to the 50-ml flask prior to the addition of sample. Pasteur pipettes were used to make accurate adjustments of the final volume. The above procedure produced a solution containing 0.050 pg ml-l of added lead. The solutions obtained by this procedure were measured for apparent lead concentration by comparison with lead standard solutions in 0.015 M nitric acid. Some of the matrix constituents of the samples were determined to ascertain whether or not a simple relationship between these and suppression of the lead signal existed.Effect of Addition of Ascorbic Acid To ascertain the effectiveness, if any, of ascorbic acid in overcoming matrix interferences, the experiment was repeated in the presence of 1% m/V of ascorbic acid in samples, spiked solutions and standard solutions. The ascorbic acid concentration of 1% m/V was achieved by replacing 2 ml of the water added to the calibrated flasks with 2 ml of 25% m/V ascorbic acid solution. The resulting solutions were measured for apparent lead concentration by comparison with lead standard solutions in 1% m/V ascorbic acid and 0.015 M nitric acid. No reagent blanks for lead were found during this study. However, other batches of ascorbic acid, even from the same manufacturer, have been found to contain lead.Determination of Lead by Solvent Extraction - Flame Atomic-absorption Spectro- p ho t ome t ry The lead content of the samples was also determined by a method, based on that proposed by the Department of the Environment and the :National Water Council Standing Committee of Analysts, currently used in this Laboratory. The method consists in extraction of the lead as its ammonium tetramethylenedithiocarbamate complex into 4-methylpentan-Z-one, followed by measurement by means of flame atomic-absorption spectrophotometry. Results and Discussion Effect of Sample Matrix on the Lead Signal Table I, exhibit a range of signal suppressions varying from 22 to 84%. The results of the spiking experiment without the addition of ascorbic acid, given in Table I1 givesMay, 1978 INTERFERENCES I N THE AAS DETERMINATION OF LEAD I N WATER 449 some data concerning the matrix constituents of the waters examined, together with the conductivity of each sample.It can be seen that no simple correlation exists between the matrix constituents deter- mined and the degree of observed signal suppression. For example, sample D, with calcium, magnesium, chloride and sulphate concentrations of 98,29, 180 and 170 pg ml-l, respectively, caused a suppression of 44%, whereas sample E, with calcium, magnesium, chloride and sulphate concentrations of only 54, 14, 53 and 76 pgml-l, respectively, gave the highest suppression (84%) found in this study. TABLE I SUPPRESSION OF LEAD PEAK HEIGHT BY DRINKING-WATER SAMPLES Apparent concentration of lead r Sample A B C D E F G H I Sample/pg ml-1 Not.detected 0.044 0.067 Not detected Not detected Not detected Not detected 0.002 8 0.006 4 Sample containing 0.050 pg ml-l of added leadlpg ml-1 0.027 0.081 0.106 0.028 0.008 0.038 0.014 0.0205 0.028 0 Apparent amount of lead added/ pg ml-1 Suppression, % 0.027 46 0.037 26 0.039 22 0.028 44 0.008 84 0.038 24 0.014 72 0.017 7 65 0.021 6 57 Further, although samples B and F gave similar suppressions of 26 and 24%, respectively, they possessed widely different matrices, which could indicate that the effect of a single interferent is modified by other species present in the matrix. TABLE I1 SOME MATRIX CONSTITUENTS (pg ml-l) AND THE CONDUCTIVITY VALUES OF THE DRINKING-WATER SAMPLES EXAMINED Albumi- Sample N nitrogen NH, as noid A 0.03 0.05 B 0.01 0.03 c 0.01 0.01 D 0.05 0.27 E 0.03 0.07 F 0.01 0.01 G 0.01 0.03 H 0.01 0.01 I 0.02 0.03 Total Free residual residual Sample chlorine chlorine Oxygen absorbed Nitrite Nitrate from Chloride Alkalinity Hardness as N as N KRZnO,* as C1 as CaCO, as CaCO, Iron Zinc Copper 0.001 1.0 1.5 7 25 40 0.04 0.1 0.3 0,001 1.0 1.4 4 15 20 0.04 0.1 0.17 0.001 5.8 0.2 11 230 260 0.04 3.5 0.12 0.001 7.0 1.5 180 130 360 0.04 0.05 0.01 0.001 3.4 0.5 53 120 180 0.04 0.1 0.005 0.002 1.0 0.2 22 130 140 0.3 0.6 0.005 0.002 8.8 0.1 25 250 320 0.005 6.5 0.01 0.001 4.0 0.1 47 150 240 0.04 0.05 0.039 0.002 4.6 0.1 27 60 140 0.10 3.2 0.005 Conductivity/ Cadmium Fluoride Calcium Magnesium Sulphate Sodium Potassium pS cm-l A 0.05 0.05 0.001 0.1 10 1.6 1 5 4.4 0.5 120 B 0.05 0.05 0.001 0.1 3.4 1.0 9 12.4 0.7 80 C 0.05 0.05 0.001 0.1 98 2.1 60 18.0 1.8 480 D 3.2 2.0 0.002 0.1 98 29 170 14.9 4.4 790 E 0.05 0.05 0.001 0.60 54 14 76 39.5 3.4 540 F 0.05 0.05 0.001 0.1 52 2.6 34 11.9 2.7 320 G 0.05 0.05 0.001 0.80 112 2.8 24 13.8 2.1 600 H 0.05 0.05 0.001 0.135 48 26 80 47.8 3.6 530 I 0.05 0.05 0.001 0.4 45 2.5 6 12.8 2.2 310 * Oxygen absorbed in 4 h at 26.7 "C from ~ / 8 0 KMnO, solution (an empirical measurement of organic materials).Effect of Addition of Ascorbic Acid The results of the spiking experiment in the presence of 1% m/V of ascorbic acid, given in Table 111, indicate the effectiveness of ascorbic acid in overcoming the observed suppression effects of drinking-water matrices on the lead signal.It can be seen that with the exception of sample H the suppression effects have been reduced to a level of 5% or less.450 REGAN AND WARREN: INFLUENCE 01; ASCORBIC ACID ON MATRIX Analyst, VoZ. 103 The complexity of the matrix interference system is again indicated by the fact that the interference associated with sample D, which exhibited the highest suppression of 84%, was completely eliminated, whereas that associated with sample H, which had a smaller suppression of 65%, could only be reduced to the 18% level by the addition of ascorbic acid, TABLE I11 SUPPRESSION OF LEAD PEAK HEIGHT B Y DRINKING-WATER SAMPLES AFTER ADDITION O F ASCORBIC ACID I Sample A B C D E F G H I Apparent concentration of lead A 7 Sample containing 0.050 pg ml-l of Sample/pg ml-l added lead/pg ml-I Not detected 0.049 0.061* 0.111 0.081* 0.129 Not detected 0.048 Not detected 0.051 Not detected 0.050 0.0202* 0.0702 0.008 7" 0.049 5 0.012 6" 0.060 0 Apparent amount of lead added/ pg ml-1 0.049 0.050 0.048 0.048 0.051 0.050 0.050 0.040 8 0.047 4 Suppression, yo 2 0 4 4 -2 0 0 18 5 * These results should be multiplied by the dilution factor of 50/47 to give natural lead levels based on the original water sample.Accuracy A measure of the accuracy of the results obtained for lead by means of the carbon furnace atomic-absorption spectrophotometric analysis of samples treated with ascorbic acid can be gained from the results of the spiking experiment shown in Table 111. A further measure of the accuracy of these results was gained by comparison with results obtained by using the method involving solvent extraction followed by measurement by means of flame atomic- absorption spectrophotometry .It can be seen from Table IV that the difference in the values obtained by means of the two techniques was less than 10% relative. Statistical analysis of the data, using a t-test, indicated that there was no significant difference (at the 95% confidence level) in the results obtained by using the two different methods. TABLE IV COMPARISON OF RESULTS FOR LEAD IN DRINKING WATER OBTAINED BY THE ASCORBIC ACID METHOD AND THE SOLVENT EXTRACTION - FLAME ATOMIC-ABSORPTION SPECTROPHOTOMETRIC METHOD Concentration of lead + 2a*/pg ml-I A I -l Sample Ascorbic acid method. Solvent extraction - AAS method A Not detected B 0.063 & 0.003 C 0.086 f 0.003 D Not detected E Not detected F Not detected G 0.0215 f 0.0002 H 0.0096 f 0.0007 I 0.0133 4 0.0006 Not detected 0.058 & 0.003 0.092 f 0.002 Not detected Not detected Not detected 0.021 8 f 0.000 6 0.0098 f 0.0006 0.0147 f 0.0038 * 2a values are from a limited number (3-5) of independent determinations carried out at different times and using fresh sample aliquots and standards.Limit of Determination The limit of determination for lead in drinking water using the addition of ascorbic acid is 0.002 pg ml-l with a 5O-pl sample volume a:nd measurement at the 283.3-nm line. This limit is of the same order as that obtained with the solvent extraction-flame atomic-May, 1978 451 absorption spectrophotometric procedure.Further improvement could be expected on using a larger sample volume and the lead absorbing line at 217 nm. The use of the gas-stop facility may also prove beneficial. INTERFERENCES I N THE AAS DETERMINATION OF LEAD IN WATER Conclusions The suppressions of the lead signal exhibited by the nine waters examined, which were gathered from a wide range of geographical locations, bore no simple relationship to the hardness of the water or to the concentration of any of the individual matrix constituents studied. The complexity of this situation demonstrates that suppression effects on lead measured in artificial, single-interferent systems cannot be extrapolated to real, multi- interferent matrices. Also, the effectiveness of ascorbic acid in combating the suppression of the lead signal was not found to be related to the composition of the matrix or the actual degree of suppression caused by the matrix. Although the use of ascorbic acid was not completely successful in overcoming interference in the determination of lead, the suppression effect was greatly diminished, to less than 5% for eight of the water samples and to 18% for the remaining sample. Where suppression by the matrix reduces the lead signal to below the detection limit, the addition of ascorbic acid may be beneficial owing to (1) the reduction of the suppressive interference and (2) the enhancement of the lead signal, as previously reported.3 The authors thank the Government Chemist for permission to publish this paper and also the Water Section of the Laboratory of the Government Chemist for providing most of the matrix constituent data shown in Table 11. References 1 . 2. 3. Campbell, W. C., and Ottaway, J. M., Talanta, 1975, 22, 729. Thompson, K. C., Wagstaff, K., and Wheatstone, K. C., Analyst, 1977, 102, 310. Regan. J . G. T., and Warren, J., Analyst, 19’76, 101, 220. Received September 30th, 1977 Accepted December 6th, 1977