Analyst, September, 1979, Vol. 104, pp. 831-836 83 1 Determination of Aluminium, Calcium, Iron and Magnesium in Sewages and Sewage Effluent by a Rapid Electrothermal Atomic-a bsorption Spectroscopic Method M. J. T. Carrondo, J. N. Lester and R. Perry* Public Health and Water Resource Engineering Section, Civil Engineering Department, Imperial College, of Science and Technology, London, SW7 2AZ The methods currently used for the determination of aluminium, calcium, iron and magnesium in sewages and final effluent are time consuming. A rapid electrothermal atomic-absorption spectroscopic procedure utilising homogenisation of samples as the only pre-treatment has been compared with digestion methods followed by flame atomic-absorption spectroscopic analysis in a statistically designed experiment.Low-sensitivity (secondary absorp- tion) lines were used for the electrothermal atomic-absorption analysis. The time saved by the use of this method is substantial and it could be used advantageously for routine analysis. Keywords : Electrothermal atomic-absorption spectroscopy ; aluminium, calcium, iron and magnesium detevmination ; sewages and sewage efluent ; pre-treatment by homogenisation ; low-sensitivity lines Atomic-absorption spectroscopy is probably the most commonly used method for the determination of aluminium, calcium, iron and magnesium in waste waters and effluents.1-3 Other methods, such as neutron-activation a n a l y ~ i s , ~ , ~ X-ray fluorescence6 and spectrophoto- metry,' have also been applied to the determination of these metals in waters, waste waters or sewage sludges.As samples of this nature contain organic and inorganic matrices, some form of treatment is required prior to flame atomic-absorption analysis. Of the methods availab1e,8s9 a limited number have achieved general use. Digestion with perchloric acid in conjunction with nitric acid is claimed to yield high recoveries.lO Procedures based on the use of sulphuric acid and nitric acid have been recommended' for waste waters, but some metals may be lost as insoluble ~ulphates.~ Digestions using hydrogen peroxide and nitric acid have been applied successfully to environmental samples.l19l2 The use of electrothermal atomisers allows pre-treatment to be minimised for samples with mainly organic matrices.l39l4 As electrothermal atomisers are more sensitive than flame atomisers, low-sensitivity (secondary absorption) lines must be used when the con- centration of the analyte is high, thus avoiding the need for excessive di1uti0n.l~ The use of a rapid electrothermal atomic-absorption method for the determination of cadmium, chromium, copper, lead, nickel and zinc in sewage sludges16-18 and sewages and effluents has been reported,Ig and its application to the determination of aluminium, calcium, iron and magnesium in sewage sludges has also been described.20 The results presented here were obtained from a statistically designed experiment to compare the rapid electrothermal atomic-absorption method of analysis using low-sensitivity lines with flame atomic- absorption analysis of acid-digested samples using high-sensitivity absorption lines.Two digestion procedures have been used, a sulphuric acid - nitric acid method and a nitric acid - hydrogen peroxide method. Experimental Apparatus spectrophotometer equipped with deuterium background correction. analysis and the working ranges used are presented in Table I. Flame analysis was undertaken using a Perkin-Elmer, Model 603, atomic-absorption The conditions for In order to remove inter- * To whom all correspondence should be addressed.832 CARRONDO et al. : DETERMINATION OF Al, Ca, Fe AND Mg IN Analyst, Vol. 104 TABLE I CONDITIONS FOR FLAME ATOMIC-ABSORPTION ANALYSIS Spectral Working Wavelength/ band width/ range/ Metal nm nm Flame type mg 1-l A1 . . . . 309.3 0.7 Nitrous oxide - acetylene, 1-50 reducing (rich, red) oxidising (lean, blue) oxidising (lean, blue) oxidising (lean, blue) Ca .. . . 422.7 0.7 Air - acetylene, 0.05-5.0 Fe . . . . 248.3 0.2 Air - acetylene, 0.05-5.0 Mg . . . . 285.2 0.7 Air - acetylene, 0.005-0.5 ferences or suppress ionisation, the samples and standards to be analysed for aluminium were made up to 2000mgl-1 in potassium chloride and those to be analysed for calcium and magnesium were made up to 0.5% in lanthanum.21 Electrothermal analyses were undertaken using the same spectrophotometer in conjunc- tion with a Perkin-Elmer HGA-76 heated graphite atomiser. The conditions and working ranges for electrothermal atomic-absorption analysis are presented in Table I1 ; the values reported were obtained with the less sensitive lines indicated.The atomisation programme used was identical for all metals and consisted, for the 20 x 10-6 1 samples used, in a drying stage at 100 "C for 30 s, a double-stage thermal decomposition with temperature ramping from 100 to 400 "C in 45 s (rate 2), followed by isothermal decomposition at 1200 "C for 30 s and atomisation at 2770 "C for 5 s for all metals except aluminium, for which an 8-s atomisation was used. The ramping stage during the thermal decomposition avoided spattering of the sample that would otherwise have occurred if the temperature had been suddenly increased from 100 to 1200 "C. TABLE I1 CONDITIONS FOR ELECTROTHERMAL ATOMIC-ABSORPTION ANALYSIS Wavelength/ Metal nm A1 . . . . 309.3 Ca . . . . 422.7 Fe . . . . 248.3 Mg . . . . 285.2 257.5* 239.9* 305.9* 202.6* Spectral band width/ nm 0.7 0.2 0.7 0.7 0.2 0.2 0.7 0.7 Sample volume Working range/ x 10-6/1 mg 1-1 20 0.05-1.0 20 0.20-4.0 20 0.01-0.08 20 1.00-20.0 20 0.05-0.5 20 0.20-5.0 20 0.002-0.04 20 0.02-0.50 * Spectral lines used in this work.Reagents Aristar-grade reagents were used for all analyses. Nitric acid, Toy0, sp. gr. 1.42. Sulphuric acid, 98%, sp. gr. 1.84. Hydrogen peroxide, 100 volume. Standard metal solutions. Standards were prepared by serial dilution of 1000 mg 1-1 All standard solutions were prepared so as to contain the same metal stock solutions. reagents as those added to the samples. Homogenisation Approximately 250 ml of undiluted raw sewage, settled sewage and final effluent samples were acidified to 1% V/V with nitric acid and homogenised in a 2-1 tall-form Pyrex beaker&?fitember, 1979 833 with an Ultra Turrax T45N homogeniser (Scientific Instrument Co.Ltd., London) for 5 min at 8000 rev min-l. This ultrasonic homogeniser works on the stator - rotor principle (centrifugal turbine) and produces a very finely homogenised suspension. Aliquots of 20 x 10-6 1 were injected into the electrothermal atomiser with an Eppendorf micropipette (Anderman & Co. Ltd., East Molesey, Surrey). Analysis was performed by direct compari- son with standards in 1% V/V nitric acid. SEWAGES AND SEWAGE EFFLUENT BY RAPID ELECTROTHERMAL AAS Sulphuric Acid - Nitric Acid Digestion7 A sample of 100 ml was digested in a 500-ml flask using an electric isomantle with 5 mI of 70% nitric acid until the volume was reduced to 10 ml; after cooling, 2.5 ml of 98% sulphuric acid were added and the digestate was heated until white fumes were evolved.If the digestion was not complete, a further 1 ml of 70% nitric acid was added until the digestate turned a pale straw colour. The digestate was filtered through a Whatman GF/C glass-fibre filter-paper. Nitric Acid - Hydrogen Peroxide Digestionll until the volume was reduced to 10 ml and then allowed to cool. of 70% nitric acid and hydrogen peroxide were added until the digestion was complete. A sample of 100 ml was digested on a thermostatic hot-plate with 5 ml of 70% nitric acid Subsequently, 2 ml each Results and Discussion Samples of raw sewage (387 mg 1-1 suspended solids), settled sewage (76 mg 1-1 suspended solids) and final effluent (17 mg 1-1 suspended solids) were collected in polythene containers, previously leached in 10% V/V nitric acid, and acidified to 1% V/V with nitric acid.For each of the pre-treatments (digestions and homogenisation) and for each sample type, five replicates and two blanks were examined. As the concentrations of aluminium in sewages and effluents were below the sensitivity of flame analysis (approximately 1 mg 1-1 for 1 yo absorption), these samples were analysed by electrothermal atomic-absorption spectroscopy. The homogenised samples were analysed by electrothermal atomic-absorption spectro- scopy. The results obtained were compared with those obtained by a method of standard additions and found to be in good agreement. The values reported are those.obtained by direct comparison with aqueous standards. The results were statistically evaluated ; the mean values, within-group relative standard deviation and the results of an analysis of variance by the F-test22 are reported in Table 111.Tukey’s test22 was used to identify which means were statistically different at the 0.05 significance level. The repeatability of the electrothermal analysis based on ten injections of the same sample (diluted, if necessary) is indicated in Table IV. No significant differences were found between the treatments for iron and aluminium in all samples. However, highly significant differences were obtained for calcium in all samples and for magnesium in settled sewage and final effluent samples. A comparison of the means by Tukey’s test indicated that the sulphuric acid - nitric acid digestion yielded lower results for the determination of calcium in all samples. Results obtained by flame atomic-absorption spectroscopy and the hydrogen peroxide - nitric acid digestion procedure and electrothermal atomic-absorption spectroscopy in conjunction with homogenisation were always in agreement.Moreover, these two methods yielded higher recoveries than the sulphuric acid - nitric acid digestion procedure. For the determination of magnesium in settled sewage and final effluent, the results obtained by the hydrogen peroxide-nitric acid digestion were comparable to those obtained by both of the other methods. However, a statistically significant difference exists between the sulphuric acid - nitric acid digestion followed by flame atomic-absorption spectroscopy and homogenisation followed by the electrothermal method, the former method yielding lower results than the latter.For unknown reasons, all values seem to be in good statistical agreement for the determination of magnesium in raw sewage. The lower results obtained for the determina- The digested samples were analysed by flame atomic-absorption spectroscopy.834 CARRONDO et aL. : DETERMINATION O F Al, Ca, Fe AND Mg IN AnaZyst, VOZ. 104 TABLE I11 COMPARISON O F ALUMINIUM, CALCIUM, IRON AND MAGNESIUM CONCENTRATIONS I N SEWAGES AND EFFLUENT USING SULPHURIC ACID - NITRIC ACID DIGESTION AND HYDROGEN PEROXIDE - NITRIC ACID DIGESTION FOLLOWED BY FLAME ATOMIC-ABSORPTION ANALYSIS WITH HOMOGENISED SAMPLES ANALYSED BY ELECTROTHERMAL ATOMIC-ABSORPTION ANALYSIS Metal Sample A1 . ... Raw sewage Settled sewage Final effluent Ca .. . . Raw sewage Settled sewage Final effluent Raw sewage Settled sewage Final effluent Mg . . . . Raw sewage Settled sewage Final effluent Fe . . Pre- treatment* H,02 - HNO, Homog. H,S04 - HNO, H202 - HNO, Homog. H2S04 - HNO, Homog. H2S0, - HNO, H,02 - HNO, Homog. H202 - HNO, Homog. H,02 - HNO, Homog. H202 - HNO, Homog. H2S04 - HNO, H,02 - HNO, Homog. H2S04 - HNO, Homog. H2S04 - HNO, Homog. H2S04 - HNO, Homog. H2S04 - HNO, H202 - HNO3 H2S04 - HNO, HZSO4 - HNO, H2S04 - HNO, H202 - HNO, H20, - HNO, HZO, - HNO, H2S04 - HNO, H202 - HNO, Homog. Mean F-test level of concentration/ Modet significance: mg I-'§ F F E E E E E E E F F E F F E F F E F F E F F E F F E F F E F F E F F E N.S.N.S. N.S. 0.01 0.01 0.01 N.S. N.S. N.S. N.S. 0.01 0.01 3.2a 3.la 3.2a 0.52a 0.53a 0.52a 0.14a 0.15a 0.15a 78a 90b 90b 75a 88b 91b 7 7a 88b 90b 1.9a 1.8a 1.8a 0.52a 0.50a 0.54a 0.15a 0.13a 0.14a 10. la 10.4a 11.0a 9. la 9.3ab 10.4b 8.6a 9.5ab 10.2b Relative standard deviation, yo 6.5 6.4 6.1 11.0 13.0 13.2 14.4 9.4 10.5 5.5 4.3 5.4 7.5 7.3 5.0 6.9 5.5 3.8 5.0 5.9 6.1 6.8 6.0 8.0 10.0 12.0 9.5 7.6 6.2 4.8 7.0 6.6 4.3 10.2 5.3 4.6 peroxide - nitric = hydrogen H2O2 - HNO, * H2S04 - HNO, = sulphuric acid - nitric acid digestion; acid digestion ; Homog. = pre-treatment by homogenisation. t F = flame atomic-absorption analysis ; E = electrothermal atomic-absorption analysis. : N.S. = not significant a t the 0.05 significancelevel.9 Means not followed by a common letter are statistically different at the 0.05 significance level. tion of calcium and to a lesser extent magnesium after digestion by the sulphuric acid - nitric acid method are probably due to the formation of insoluble sulphatesg that were retained in the filter or were not aspirated into the flame. The relative standard deviations obtained for the electrothermal atomic-absorption analysis of the homogenised samples compared well with the values obtained for flame analysis of digested samples. The repeatability of the electrothermal method reported in Table IV indicates that for the determination of aluminium in settled sewage and final effluent and iron in final effluent the relative standard deviations were high.This is in part the consequence of working close to or below the lower limit of the recommended working range [Table 11). For these two metals the repeatability of analysis in final effluent wasSeptember, 1979 835 also examined at the higher sensitivity (primary absorption) lines (Table 11) and the relative standard deviations were found to improve from 9.4% to 6.5% and from 8.0% to 4.8% for aluminium and iron, respectively. Thus, the homogenisation - electrothermal atomic- absorption method can yield more precise results if high-sensitivity lines are used for the determination of the lower concentrations. SEWAGES AND SEWAGE EFFLUENT BY RAPID ELECTROTHERMAL AAS TABLE I V REPEATABILITY OF ELECTROTHERMAL ATOMIC-ABSORPTION ANALYSIS OF RAW SEWAGE, SETTLED SEWAGE AND FINAL EFFLUENT SAMPLES (ONE OF EACH) INJECTED TEN TIMES Concentration/mg 1-1 Relative I standard A Metal Sample Mean Standard deviation deviation, yo Al .. . . Raw sewage 3.1 0.17 5.5 Settled sewage 0.54 0.05 9.3 Final effluent 0.16 0.015 9.4 Ca . . . . Raw sewage 9.1 0.35 3.9 Settled sewage 9.0 0.32 3.6 Final effluent 8.8 0.25 2.8 Fe . . . . Raw sewage 1.74 0.08 Settled sewage 0.52 0.034 Final effluent 0.15 0.012 4.6 6.5 8.0 Mg . . . . Raw sewage 0.43 0.013 3.0 Settled sewage 0.41 0.012 2.9 Final effluent 0.40 0.012 3.0 The repeatability of the results obtained for the analysis of aluminium varied slightly with time; this could have been due to variations in voltage that resulted in slight tempera- ture variations in the furnace at the time of atomisation.There was also an increase in sensitivity as the tubes became older; this type of interference effect can probably be explained by Fuller’s kinetic theory of at0misation.~3 Thus, a standard had to be injected more often than was required for other metals (every five as opposed to fifteen injections). Reported values for the electrothermal atomic-absorption analysis of biological samples quote relative standard deviations of 8% for the determination of aluminium in whole blood2* at 0.4 mg 1-’ and 10% in biological tissue13 at approximately 1 mg ml-l. The use of low-sensitivity lines for the determination of iron, calcium and magnesium by electrothermal atomic-absorption spectroscopy eIiminates, or substantially reduces, the extent of dilution that would be needed and so the amount of contamination that could occur if sensitive lines were used.The analysis is simple and there is no need to use the deuterium background corrector, which agrees with results reported previously for iron15 and calcium and magnesium.25 Conclusions The rapid electrothermal atomic-absorption method described here compares well with flame atomic-absorption spectroscopy in conjunction with the digestion methods tested. Homogenisation takes only 5 min as opposed to 3-6 h needed for digestion; this more than compensates for the additional time (2-3 min) required in the electrothermal as opposed to the flame method. The proposed method has the further advantage over flame atomic- absorption spectroscopy that it dispenses with the need to add interference removal agents to samples and standards prior to analysis. Deuterium background correction is not required for the determination of the metals indicated.The authors acknowledge the financial support for this work provided by the Department of the Environment and the Department’s approval for the publication of these results. One of us (ill. J. T. Carrondo) is also grateful to the Instituto Nacional De InvestigaCZo Clentifica, Lisboa, Portugal, for the award of a postgraduate scholarship.836 CARRONDO, LESTER AND PERRY References Welz, B., and Wiedeking, E., 2. Analyt. Chem., 1973, 264, 110. Department of the Environment/National Water Council, “Magnesium in Waters and Sewage Effluents by Atomic Absorption Spectrophotometry, 1977,” HM Stationery Office, London, 1978, Department of the Environment/National Water Council, “Calcium in Waters and Sewage Effluents by Atomic Absorption Spectrophotometry, 1977, Tentative Method,” HM Stationery Office, London, 1978, 12 pp.Lieser, K. H., Calmano, W., Heuss, E., and Neitzert, V., J . Radioanalyt. Chem., 1977, 37, 717. Nadkarni, R. A., and Morrison, G. H., Envir. Lett., 1974, 6, 273. Smits, J., and Van Grieken, R., Analytica Chim. Acta, 1977, 88, 97. Department of the Environment, “Analysis of Raw, Potable and Waste Waters,” HM Stationery Gorsuch, T. T., “The Destruction of Organic Matter,” Pergamon Press, New York, 1970. Hanson, N. W., Editor, “Official, Standardised and Recommended Methods of Analysis,” Second Agemian, H., and Chau, A. S. Y., Analyst, 1976, 101, 761. Geyer, D., Martin, P., and Adrian, P., Korresp. Abwass., 1975, 22, 369. Krishnamurty, K. V., Shpirt, E., and Reddy, M. M., Atom. Absorption Newsl., 1976, 15, 68. Krishnan, S. S., Quittkat, S., and Crapper, D. R., Can. J . Spectrosc., 1976, 21, 25. Lord, D. A., McLaren, J. W., and Wheeler, R. C., Analyt. Chem., 1977, 49, 257. Olsen, E. D., Jatlow, P. I., Fernandez, F. J., and Kahn, H. L., Clin. Chem., 1973, 19, 326. Lester, J. N., Harrison, R. M., and Perry, R., Sci. Total Envir., 1977, 8, 153. Stoveland, S., Astruc, M., Perry, R., and Lester, J . N., Sci. Total Envir., 1978, 9, 263. Carrondo, M. J. T., Perry, R., and Lester, J . N., Analytica Chim. Acta, 1979, 106, 309. Carrondo, M. J. T., Perry, R., and Lester, J. N., Sci. Total Envir., 1979, 12, 1. Carrondo, M. J . T., Lester, J . N., and Perry, R., Talanta, 1979, in the press. Kirkbright, G. F., and Sargent, M., “Atomic Absorption and Fluorescence Spectroscopy,” Academic 12 PP- Office, London, 1972. Edition, Society for Analytical Chemistry, London, 1973. Press. London. 1974. 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. Bowker; A. H., ‘and Lieberman, G. J ., “Engineering Statistics,” Prentice Hall, Englewood Cliffs, N.T.. 1972. Fullgr,‘C. W., Analyst, 1976, 101, 798. Langmyhr, F. J., and Tsalen, D. L., Analytica Chim. Acta, 1977, 92, 79. Pardhan, S. I., and Ottaway, J. M., Proc. Analyt. Div. Chem. Soc., 1975, 12, 291. Received February 28th, 1979 Accepted March 29th, 1979