Analyst, October, 1979, Vol. 104, $9. 937-943 937 Influence of Conditioning Agents on the Determination of Metallic Content of Sewage Sludge by Atomic-absorption Spectrophotometry with Electrothermal Atomisation M. J. T. Carrondo, R. Perry and J. N. Lester Public Health Engineering Laboratory, Imperial College of Science and Technology, London, S W7 2AZ Conditioning agents are often used to aid the de-watering of sewage sludges prior to disposal to agricultural land. These might interfere in the electro- thermal atomic-absorption spectrophotometric analysis of heavy metals. Possible interferences by inorganic and polyelectrolyte conditioners were studied at the higher rates of addition normally used in sewage practice. Analyses for cadmium, chromium, copper, nickel, lead and zinc were per- formed by flame and electrothermal atomisation methods in conditioned and unconditioned samples.The organic polyelectrolytes tested did not interfere, nor did most inorganic conditioners a t rates of addition consistent with normal sewage treatment practice. However, interferences occurred with aluminium chlorohydrate a t normal and very high addition rates, and with other inorganic conditioners a t very high addition levels. Keywords: Electrothermal atomic-absorption spectrophotornetry ; cadmium, chromium, copper, nickel, lead and zinc determination ; sewage sludge analysis ; interferences ; inorganic and polyelectrolyte conditioners During the process of wastewater treatment, heavy metals contained in the incoming sewage tend to be concentrated in the sludge pr0duced.l In the UK it is estimated2 that 40% of the sludges from all inland sewage treatment works are disposed to agricultural land.Liquid digested sludge is the most favoured form for land application, but because of handling problems many sludges are de-watered prior to application. To improve the de-watering process, conditioners are generally added.3 These include iron( 11) sulphate (copperas), iron(II1) chloride, aluminium sulphate, hydrated aluminium chloride (aluminium chloro- hydrate) and lime, or mixtures of t h e ~ e . ~ ? ~ More recently, organic compounds, mainly cationic polyelectrolytes, have been ~ s e d . ~ - ~ Inorganic conditioners must be added to the sludge in large amounts to achieve their effect (see Table I), whereas the organic conditioners will generally not exceed 1% of the dry matter present in the ~ l u d g e .~ - ~ Because of the toxic6 or phytotoxic7 nature of some of the heavy metals present, their concentrations must be determined prior to disposal to land to ensure that maximum permissible levels8 are not exceeded. The methods available for the determination of heavy metals in sewage sludges require some form of pre-treatment to destroy organic matter prior to analysis and this is time consuming. Recently, a rapid electrothermal atomic-absorption spectrophotometric method, utilising homogenisation as the only pre-treatment, has been successfully developed for the determination of cadmium, copper, chromium, nickel, lead and zinc in unconditioned sewage sludge.gJ0 Various causes have been suggested for the occurrence of interferences in graphite furnace atomic-absorption spectrophotometry.It has been suggested that interferences are due to physical effects, formation of molecular compounds in the gas phase resulting in molecular absorption, formation of refractory compounds and interaction between elements and gaseous compounds.ll The work of Sturgeon et a1.12 indicated that losses of some analytes might be due to the formation of the more volatile chlorides. Occlusion of the analyte in the matrix has also been proposed as a mechanism of interference.l3Sl4 The literature on cationic and anionic interferences occurring in electrothermal atomic-absorption spectro- photometry has been reviewed.15 Most of the work on interferences has involved the determination of a particular element in a particular matrix and the results are usually reported as an enhancement or depression of the signal when one or two additions of the938 CARRONDO et al.INFLUENCE OF CONDITI0N'"G AGENTS ON THE Analyst, Vol. 104 interferent are added to the analyte. This approach may lead to erroneous conclusions when attempting to predict interferences on different complex matrices with various inter- ferents and at different interferent to analyte ratios. Under these conditions interferences may either compensate or enhance each other or synergistic or antagonistic effects may occur.lG The work reported here was undertaken to determine if some of the common conditioning agents would cause interferences in the rapid electrothermal atomic-absorption spectro- photometric analysisg~10 of cadmium, chromium, copper , nickel, lead and zinc in sewage sludge.As interferences normally increase with an increase in the interferent to analyte ratio, the conditioners were used at two rates: the maximum addition normally found in sewage treatment practice and an additional rate 50% higher than that, to cover any abnormal situations (see Table I). Analysis of the unconditioned sludge by flame atomic- absorption spectrophotometry after digestion by a nitric acid - hydrogen peroxide procedure was also undertaken. TABLE I RATES AND METHODS FOR ADDITION OF INORGANIC CONDITIONERS Range of dosing rate in normal practice Inorganic conditioner (yo m/m dry solids) Method of addition Lime , ... . . . . 10-25 Added as 10% mlm slurry + copperas . . . . . . 5-20 Always added before lime as a 25% mlm solution Lime . . . . . . . . 10-15 Added as 10% mlm slurry + iron(II1) chloride . . . . 1-5 A.dded before lime as a 40% mlrn solution Aluminium chlorohydrate . . 0.5-3.0 Dose expressed as yo mlm aluminium oxide on dry solids Rates used in these experiments (% mlm dry solids) - Rate 1 Rate 2 25 37.5 20 30 15 22.5 5 7.5 3 4.5 Experimental Apparatus Flame atomic-absorption analysis was undertaken using a Perkin-Elmer, Model 603, atomic-absorption spectrophotometer equipped with deuterium background correction. The conditions for analysis were those recommended by the instrument manufacturer, with the exception of those for chromium for which slit 3 (0.2 nm) was used to reduce spectral inter- f erences .17 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 11. Aliquots of 20 or 50 p1 were injected into the electrothermal atomiser with an Eppendorf micropipette (Anderman & Co. Ltd., Surrey). Analysis was performed by direct comparison with standards in 1% V/V nitric acid. Nitric Acid - Hydrogen Peroxide Digestionls the volume was reduced to 10ml. peroxide were added repeatedly until the digestate turned a pale straw colour. A sample of 20 ml was digested on a thermostatic hot-plate with 30 ml of nitric acid until After cooling, 2ml each of nitric acid and hydrogen Homogenisation Approximately 250 ml of the sample, previously diluted 50-fold and acidified to 1% V/V with nitric acid, was homogenised in a 2-1 tall-form borosilicate beaker with an Ultra Turrax T45N homogeniser (Scientific Instrument Co.Ltd., London) for 5 min at 8000 rev min-l. To avoid contamination of the sample by chromium and nickel the original stainless-steelOctober, 1979 DETERMINATION OF METALLIC CONTENT OF SEWAGE SLUDGE BY AAS TABLE I1 OPERATING CONDITIONS FOR THE ATOMIC-ABSORPTION SPECTROPHOTOMETER AND ELECTROTHERMAL ATOMISER Conditions Wavelength/nm . . . . Slit width/nm . . . . . . Drying time/s . . . . .. Ashing time/s . . .. .. Atomisation time/s . . . . 20 $/mg 1-l . . . . . . 50 pl/mg 1-1 . . . .. . Drying temperature/"C . . Ashing temperature/"C . . Atomisation temperature/"C . . Working range using Working range using * Gas stop used. Metal Cd 228.8 0.7 100 30 250 40 2 100 5 0.005-0.03 0.002-0.02 Cr 357.9 0.2 100 30 1100 30 2 770 5 0.02-0.2 0.01-0.1 c u 324.8 0.7 100 30 700 30 2 770 4 0.05-0.4 0.02-0.2 Ni 232.0 0.2 100 30 800 30 2 770 5 0.1-1 .o 0.05-0.4 Pb 283.3 0.7 100 30 350 40 2 300 5 0.025-0.4 0.01-0.2 939 - Zn 307.6 0.7 100 30 450 35 2 500 5* 0.2-2.0 0.05-1.0 shaft, stator and rotor of the homogeniser were replaced with a replicate made of titanium, The suitability of this shaft for the homogenisation of samples to be analysed for cadmium, chromium, copper, nickel, lead and zinc has been reported previo~s1y.l~ Addition of Inorganic Conditioners Three commonly used inorganic conditioners were studied: lime and copperas, lime and iron(II1) chloride and aluminium chlorohydrate.The range of rates of addition of inorganic conditioners in normal practice, and also the dosing rates used in the experiment, are shown in Table I. Each conditioner was dosed at two rates, the maximum of the range used in normal practice and a rate 50% higher than this. Conditioning agents are added to sludge on the basis of its total solids content; the rate of addition is therefore expressed as a percentage by mass of the dry solids content of the sludge. Addition of Polyelectrolytes Three cationic polyelectrolytes were studied : Aquafloc 4051 (Dearborn Chemicals Ltd., Lancashire), Flocbel 170 (Float-Ore Ltd., Middlesex) and Zetag 94 (Allied Colloids Manu- facturing Co.Ltd., Yorkshire). Aquafloc 4051 and Flocbel 170 are solid polyelectrolytes and are 100~o active. Zetag 94 is a liquid polyelectrolyte containing 15% rn/m of active material. This again was added as a 0.1% m/m solution. The normal upper limit of polyelectrolyte addition in sewage treatment is 1% m/m of the dry solids content of the sludge. Each polyelectrolyte was added at two dose rates, 1.0 and 1.5% mlwz dry solids. They were applied as dilute 0.1% m/m solutions. Incorporation of Conditioners required amount of inorganic conditioner or polyelectrolyte solution. then added to make up the volume to 1250 ml and the solution was mixed by stirring. solution was then homogenised as described for unconditioned sludges. To 25 ml of mixed primary sludge in a 2-1 tall-form borosilicate beaker was added the Distilled water was The Reagents BDH Aristar-grade reagents were used for all analyses. Nitric acid, 70%, sp.gr. 1.42. Hydrogen peroxide, 100 volume. Standards were prepared by serial dilution of 1000 mg 1-1 metal stock solutions.940 CARRONDO et d. INFLUENCE OF CONDITIONING AGENTS ON THE Analyst, Vol. 104 Results and Discussion A sample of mixed primary sludge (total solids content 6.26%) was collected in a polythene container and acidified to 1% V/V nitric acid. For each of the pre-treatments (digestion and homogenisation of unconditioned and conditioned sludge) five replicates and two blanks were taken. The digested samples were analysed by flame atomic-absorption spectro- photometry and the homogenised samples by electrothermal atomic-absorption spectro- photometry. The results were treated statistically and the rnean values, within-group relative standard deviations (RSD) and the results of an analysis of variance by the F-test20 are presented.Tukey’s procedure20 was used to identify which means were different at the 0.05 probability level. The values obtained with the inorganically conditioned sludge are presented in Table 111 and there are highly significant effects for all metals, with the exception of zinc, for which no significant differences were found. The results obtained for cadmium, chromium and nickel in the presence of aluminium chlorohydrate, at both rates, are statistically different from the others; in all instances lower results were obtained. For these same elements with other conditioners, at both rates, no significant differences were found.The results obtained for copper with aluminium chlorohydrate were also significantly different from the others, but the higher rate of addition enhanced the results and sharply increased their scatter, as indicated by the RSI). Higher results were also obtained with the higher dose (rate 2) of lime and copperas (CaO + FeSO,). The results obtained in the presence of the other conditioners agree well with the values used for comparison, i.e., hydrogen peroxide digested samples analysed by flame atomic-absorption spectrophoto- metry and the homogenised samples analysed by electrothermal atomic-absorption spectro- photometry. Finally, for the determination of lead the high dosages (rate 2) of lime and copperas and lime and iron(II1) chloride yielded significantly higher results with a larger RSD.The presence of these conditioners at the maximum rate normally used in sewage practice (rate 1) or the presence of aluminium chlorohydrate did not seem to affect the results. As the presence of aluminium chlorohydrate, at both rates of addition, significantly affected the results obtained for cadmium, chromium, copper and nickel, the influence of this com- pound on the analysis of aqueous acidic standards was examined. If a similar effect for the standards as in the sludges had been observed, then the reproduction in the standards of the aluminium chlorohydrate content of the sludge would have allowed the use of a calibration method foI the analysis.However, the test showed that no effects occurred, i.e., the standaids yielded the same absorbance in the presence and absence of aluminium chlorohydrate. When polyelectrolyte conditioners were added no significant differences between the treated and untreated sludge were found for cadmium, copper, nickel and zinc (Table IV) but both rates of addition of Aquafloc 4051 caused an increase in the scatter of results for cadmium and lead. Zetag 44 appeared to affect the determination of lead (and to a lesser extent cadmium) in a similar mannner when added at rate 2. The determination of chromium and zinc with Flocbel 170 at the high dosage rate yielded results marginally lower than those obtained in some other instances.For both these elements, the results may be grouped into “high results” and “low results.” For chromium, the “high results” contain all but those obtained for Flocbel 170 dosed at rate 2 and the “low results” contain all but those for the digestion procedure. Thus, the evidence for rejecting the method for the determination of chromium and zinc, at least in the presence of polyelectrolytes equivalent to the maximum additions used in practice, is very weak. The addition of inorganic conditioners increased the amouhts of some ions, calcium, iron, aluminium, sulphate and chloride, for example, which are known to cause interferences in the assays of some or all of the six metals determined.13-15 At the maximum rates of addition of the conditioners used in sewage practice, ratios of analyte to interfering substances of 1 : 1000 or even greater will occur.I t would therefore be expected that interferences would occur in the presence of the conditioners. The results show that these conditioners did not interfere at either rate for zinc and that, apart from aluminium chlorohydrate, there were no interferences in the determination of cadmium, chromium or nickel. For copper and lead at some of the higher rates of addition, interferences did occur.October, 1979 DETERMINATION OF METALLIC CONTENT OF SEWAGE SLUDGE BY AAS TABLE I11 EFFECT OF DIFFERENT CONDITIONERS ON THE DETERMINATION OF CADMIUM, CHROMIUM, COPPER, NICKEL, LEAD AND ZINC IN SEWAGE SLUDGE BY ELECTROTHERMAL ATOMIC-ABSORPTION SPECTROPHOTOMETRY Metal Cd . . Cr . . c u .. Ni . . Pb . . Zn . . Pre-treatment * . . H202-HN03 Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeC1, (2) AlC1, (1) AlC1, (2) Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeC1, (2) AlC1, (1) AlCl, (2) Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeC1, (2) AlCl, (1) AlC1, (2) .. H20,-HN03 .. H202-HNO, .. H20, - HNO, Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeCl, (2) AlC1, (1) AlC1, (2) .. H202-HN03 Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeCl, (1) CaO + FeCl, (2) AlCl, (1) NC1, (2) .. H20a-HN03 Homog. CaO + FeSO, (1) CaO + FeSO, (2) CaO + FeC1, (1) CaO + FeC1, (2) MC1, (1) NC1, (2) F-test level of Mean concentration?/ significance mg 1-1 KSD,S % 0.01 0 . 2 6 ~ 4.3 0 . 2 5 ~ 5.5 0 . 2 6 ~ 10.2 0 .2 4 ~ 3.8 0 . 2 4 ~ 2.7 0 . 2 4 ~ 4.9 0.20b 6.5 0.18b 5.5 0.01 3 . 4 ~ 3.2ac 3.2ac 3 . 4 ~ 3.2ac 3 . 3 ~ 2.9bc 2.6b 0.01 1 6 . 9 ~ 1 7 . 2 ~ 1 7 . 5 ~ 2 1 .'Ob 1 7 . 7 ~ ~ 1 6 . 7 ~ 12.4d 20.3bc 0.01 6 . 5 ~ 6 . 2 ~ 6 . 4 ~ 6 . 5 ~ 6 . 2 ~ 5.7ab 5.lb 3 . 1 ~ 6.3 5.9 4.7 4.6 3.5 5.1 4.0 2.4 3.8 5.0 5.6 9.7 5.9 7.6 5.6 20.7 7.8 6.7 9.5 11.0 5.3 9.4 15.6 16.7 0.01 40a 7.8 43a 5.9 46a 8.4 57b 9.4 41a 5.9 58b 9.1 42a 7.3 42a 6.7 N.S.3 35a 4.8 34a 5.1 31a 5.2 33a 5.8 33a 5.7 35a 6.1 32a 6.1 34a 7.2 941 * H202 - HNO, = hydrogen peroxide - nitric acid digestion; Homog. = pre- treatment by homogenisation; CaO lime; FeSO, = copperas; FeC1, = iron(II1) chloride; AlCl, = aluminium chlorohydrate; (1) = rate 1 (Table I ) ; (2) = rate 2 (Table I). t Means not followed by a common letter are statistically different at the 0.05 significance level.RSD = relative standard deviation. 3 N.S. = not significant at the 0.05 significance level.942 Analyst, Vol. 104 The polyelectrolytes, which are organic in nature and added in much smaller amounts, would be expected to have much less effect, if any. It can be seen from the data presented that there was no effect on the results for cadmium, copper, nickel or lead and there was CARRONDO et al. : INFLUENCE OF CONDITIONING AGENTS ON THE TABLE IV EFFECT OF DIFFERENT POLYELECTROLYTES ON THE DETERMINATION OF CADMIUM, CHROMIUM, COPPER, NICKEL, LEAD AND ZINC I N SEWAGE SLUDGE BY ELECTROTHERMAL ATOMIC-ABSORPTION SPECTROPHOTOMETRY F-test level of Mean concentrationt / Metal Pre-treatment* significance mg 1-l RSD,$ % Cd Cr c u Ni Pb Zn ... . .. .. . . .. .. .. HZOZ - HNO, Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Homog. Flocbel (1) Flocbel (2) Aquafl (1) Aquafl (2) Zetag (1) Zetag (2) Hz02 - HNO, HZOZ - HNO, HSO, - HNO, H20z - HNO, HZOZ - HNO, N.S.$ 0.26a 0.25a 0.26a 0.26a 0.24a 0.24a 0.25a 0.24a 0.05 N.S. N.S. N.S. 0.05 3.4a 3.2ab 3.lab 3.0b 3.lab 3.2ab 3.4a 3.3ab 16.9a 17.2a 17.4a 16.7a 17.6a 16.4a 16.Oa 16.0a 6.5a 6.2a 6.0a 6.3a 6.5a 6.4a 6.3a 6.3a 40a 43a 41a 42a 45a 46a 46a 47a 35a 34ab 31ab 30b 32ab 31ab 32ab 33ab 4.3 5.5 6.6 7.9 12.8 10.3 6.9 8.6 6.3 5.9 8.2 6.6 4.1 7.7 8.6 6.3 3.8 5.0 8.6 5.3 4.3 3.3 5.9 6.0 7.8 6.7 5.9 5.1 4.5 6.3 2.3 5.7 7.8 5.9 7.7 7.2 11.0 10.2 7.9 12.5 4.8 5.1 6.5 6.3 4.6 4.5 5.9 4.9 * Flocbel = Flocbel 170; Aquafl = Aquafloc 4051; Zetag = Zetag 94.Other footnotes as in Table 111.October, 1979 DETERMINATION OF METALLIC CONTENT OF SEWAGE SLUDGE BY AAS 943 only weak evidence for an effect on the determination of chromium and zinc. Although only three polyelectrolytes were tested it is thought that similar results would have been obtained with others. Conclusion The rapid electrothermal atomic-absorption spectrophotometric procedure described here can be used for the determination of cadmium, chromium, copper, nickel, lead and zinc in sludges that have been conditioned with polyelectrolytes.When high rates of conditioners are used the method is not suitable for these analytes. However, the method could be used at rates of addition consistent with normal sludge treatment practice for all of the con- ditioners tested except aluminium chloroh ydrat e. We acknowledge financial support for this work from the Department of the Environment and the Department’s approval to publish these results. One of us (M. J. T. Carrondo) is also grateful to the Instituto Nacional De Investiga@o Cientifica, Lisboa, Portugal, for the award of a postgraduate scholarship. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.20. References Neufeld, R. O., Gutierrez, J., and Novak, R. A., J . Wat. Pollut. Control Fed., 1977, 49, 489. Ministry of Housing and Local Government, “Taken for Granted,” Report of the Working Party on Vesilind, P. A., “Treatment and Disposal of Sewage Sludges,” Ann Arbor Science Publishers, Ann Clark, E. I., and Fisk, B., Waf. Pollut. Control, 1978, 77, 509. Gale, R. S., and Baskerville, R. C., Filtr. Sep., 1970, 7, 37. Dulka, J. J., and Risby, T. H., Analyt. Chem., 1976, 48, 640A. Page, A. L., “Fate and Effects of Trace Elements in Sewage Sludge when Applied t o Agricultural Department of the Environment, “Report of the Working Party on the Disposal of Sewage Sludge Carrondo, M. J. T., Perry, R., and Lester, J . N., Analytica Chim. Acta, 1979, 106, 309. Stoveland, S., Perry, R., and Lester, J . N., Sci. Total Environ., in the press. Baudin, G., Chaput, M., and Feve, L., Spectrochim. Acta, 1971, 26B, 425. Sturgeon, R. E., Chakrabarti, C. L., and Langford, C. H., Analyt. Chem., 1976, 48, 1792. Smeyers-Verbeke, J., Michotte, Y., Van den Winkel, P., and Massart, D. L., Analyt. Chem., 1976, Cruz, R. B., and Van Loon, J . C., Analytica Chim. Acta, 1974, 72, 231. Fuller, C. W., “Electrothermal Atomization for Atomic Absorption Spectrophotometry,” Analytical Woodis, T. C., Hunter, G. B., and Johnson, F. J., Analytica Chim. Acta, 1977, 90, 127. Kirkbright, G. F., and Sargent, M., “Atomic Absorption and Fluorescence Spectroscopy,” Academic Geyer, D., Martin, P., and Adrian, P., Korrespondenz Abwasser, 1975, 22, 369. Stoveland, S., Astruc, M., Perry, R., and Lester, J . N., Sci. Total Environ., 1978, 9, 263. Sewage Disposal, HM Stationery Office, London, 1970. Arbor, Mich., 1974. Land-A Literature Review Study,” U.S. EPA-670/2-74-005, 1974. t o Land,” HM Stationery Office, London, 1977. 48, 125. Sciences Monograph No. 4, Chemical Society, London, 1977. Press, London, 1974. Bowker, A. H., and Lieberman, G. J., ‘ N.J., 1972. ‘Engineering Statistics,” Prentice-Hall, Englewood Cliffs, Received April 3rd, 1979 Accepted May 9th, 1979