224 Analyst, March, 1979, Vol. 104, pp. 224-231 Determination of Chromium in Natural Waters and Sewage Effluents by Atomic-absorption Spectrophotometry Using an Air = Acetylene Flame K. C. Thompson and K. Wagstaff Severn-Trent Water Authority, Malvern Regional LaboGvatory, 141 Church Street, Malvern Worcestershire WR14 2AN A simple method for the determination of chromium in natural waters and sewage final effluents by atomic-absorption spectrophotometry using an air - acetylene flame is described. The sample is concentrated by evapora- tion by a factor of five. Interference effects were minimised by the addition of ammonium perchlorate and were further reduced by working with a flame on the verge of luminosity rather than a distinctly luminous flame. Keywovds : Chromium determination ; atomic-abs orption spectrophotometry ; aiv - acetylene pame ; .Pzat.ural waters and sewage efluents The World Health Organization European Standard1 quotes a limit for chromium(V1) in potable waters of 0.05 pg ml-1, and a European Economic Community Directive,2 con- cerning the quality required of surface waters intended for abstraction of drinking water, quotes a total chromium limit of 0.05 pg ml-l.There is a requirement for a routine atomic- absorption spectrophotometric method suitable for the determination of chromium at these levels in a wide range of natural waters and sewage effluents. Ideally, the method should have a sample preparation stage that will allow subsequent analysis for other toxic metals of interest. The proposed method should have a detection limit of about 0.005 pg ml-l of chromium.Most manufacturers of atomic-absorption spectrophotometers quote detection limits for chromium in pure solution, when calculated as 4.65 times the within-batch standard deviation of the blank,3 of between 0.007 and 0.01 pg ml-l in the luminous air - acetylene flame. Under these flame conditions, inter-element effects for chromium are severe:s5 and it appears that some form of pre-concentration and also some method of minimising potential inter-element effects are necessary. The commonly used solvent-extraction technique using ammonium t et rame t hylenedit hiocarbam a te - 4-met hylpent an-2-one is very sensitive but suffers from several disadvantage^.^,^ A pre-treatment step utilising acid digestion is required to break down any insoluble or or,ganically bound chromium.A potassium permanganate oxidation step is necessary to convert chromium( 111) into chromium(V1). Reduction of the excess of permanganate and careful pH adjustment must be made before the extraction is performed. The pH adjustment step is critical if co-extraction of manganese with subsequent emulsion formation is to be avoided. Overnight standing of the extract prior to nebulisation is recommended.s This paper describes a simple atomic-absorption spectrophotometric method that utilises a concentration by evaporation technique, in which ammonium perchlorate is incorporated into the sample solution in order to minimise inter-element effects from the sample matrix. The determination is carried out in an air - acetylene flame.Experimental Apparatus Atomic-absorptiort spectrophotometer. A Varian Techtron 1200 fitted with a standard high-solids air - acetylene burner (titanium burner grid) and a corrosion-resistant nebuliser was used. Borosilicate glass beakem and test-tGbes. Tall-form 100-ml beakers with a spout, and engraved at the 5 ml level. These beakers were initially boiled in 50% V/V hydrochloric acid (36% m/m) and were reserved for this work. Calibrated (0.1 m1) 10-ml tubes with ground-glass stoppers (Exelo Ltd.), which were initially soaked in 50% V/V hydrochloric acid and regularly cleaned using laboratory detergent, were used.THOMPSON AND WAGSTAFF 225 Reagents Hydrochloric acid, 25% V/V. Dilute 250 ml (&2 ml) of hydrochloric acid (36% m/m) (analytical-reagent grade) to 1 1 ( 5 2 ml) with de-ionised water.Ammonium perchlorate solution, 10% m/V. Dissolve 50 g (&O.l g) of ammonium perchlorate (Fisons Ltd.) in about 450 ml of de-ionised water and dilute to 500 ml (&l ml) with de-ionised water. Caution-Ammonium perchlorate is a potentially hazardous chemical and any solution spillage should be dealt with immediately in order to avoid any subsequent fire risk. Hydrogen peroxide, 6% mfm. Aluminium oxide anti-bumping granules. BDH Chemicals. These granules were boiled with nitric acid, washed with de-ionised water and dried prior to use. Standard chromizm(lI1) chloride solution (1 000 pg ml-l of chromium). BDH Chemicals. Standard potassium dichromate solution (1 000 pg ml-l of chromium).Hopkin and Williams. BDH Chemicals, analytical-reagent grade (20 volume). Optimisation of the Method Choice of flame and flame conditions It is well known that chromium determinations in the air - acetylene flame are prone to inter-element effect^^^^^*-^^ and that the chromium sensitivity is dependent on the oxidation state of the c h r ~ m i u m . ~ J ~ J ~ These problems can be avoided by using the dinitrogen oxide - acetylene flame but the use of this flame results in a decrease in the detection limit of approximately 4-&fold compared with the luminous air - acetylene flame. Many labora- tories prefer to avoid the routine use of the dinitrogen oxide-acetylene flame for their standard toxic metal analyses and for these reasons a method utilising the air - acetylene flame was developed.Inter-element effects in the air - acetylene flame can be minimised by setting the acetylene flow so that a non-luminous flame is 0btained.~,~J0 However, this results in a significantly decreased detection limit compared with the luminous-flame conditions that are normally used for this determination. Previous work14 has shown that under luminous-flame con- ditions chromium calibration graphs over the range 0-20 pg ml-l exhibited inflexions and well defined maxima. The effect was especially pronounced for chromium(II1) solutions; it appeared to depend upon the age of the solutions and was also observed in solutions con- taining 10% V/V nitric acid (70% m/m). For this study, the acetylene flow was set so that a flame on the verge of luminosity was obtained.The chromium characteristic concentration under these conditions (0.10 pg ml-l) was just over twice that observed in the luminous flame (0.043 pg ml-l). Choice of interference suppressor Various reagents have been recommended for minimising inter-element effects in the atomic-absorption spectrophotometric determination of chromium in the air - acetylene flame. These include lanthanum chloride,15 sodium sulphate,s ammonium ~ h l o r i d e , ~ ~ ~ ammonium bifluoride4~lo and quinolin-8-01.~~~~ The incorporation of 5 000 pg ml-l of lanthanum (as the chloride) in all the solutions was found not to overcome many inter- element effects and resulted in a significant background-absorption signal. The addition of 5000 pg ml-l of sodium sulphate actually enhanced the suppression caused by 1000 pg ml-l of calcium and magnesium on a 10 pg ml-l chromium solution.* Recently, the addition of ammonium perchlorate has been recommended for minimising inter-element effects of elements other than chr~mium.l~-~~ It was decided, therefore, to compare the addition of ammonium chloride with the addition of ammonium perchlorate to various synthetic solutions so that the final solutions contained 2% m/V of the added salt, as recommended by BarnesQ for ammonium chloride.Ammonium salts, as expected, were found not to increase significantly the background-absorption signal. Table I shows that an interference suppressant is essential and that ammonium perchlorate is better than ammonium chloride. All the results in Table I were obtained using a flame on the verge of luminosity.If the acetylene flow was increased so as to obtain the maximum chromium response (a distinctly luminous flame), inter-element effects were significantly increased in all instances. Ammonium perchlorate, at a concentration of 2% m/V in the final nebulised solution, was used in all further work. Increasing the ammonium perchlorate concentration226 THOMPSON AND WAGSTAFF: DETERM~NATION OF CHROMIUM IN Analyst, VoZ. 104 COMPARISON OF AMMONIUM CHLORIDE AND AMMONIUM PERCHLORATE TABLE 1 AS INTERFERENCE SUPPRESSORS OF INTER-ELEMENT EFFECTS All solutions contained 2 pg ml-l of chromium(1 [I) and 10% V / B hydrochloric acid (36 yo nz/m). Sequential background correction was applied. Relative signal with 2% m/ V of added suppressant r \ A Interfering substance Concentration* / Signal with no Ammonium Ammonium added p g ml-l suppressant added chloride perchlorate None .... .. - 100 100 100 :e} 74 92.5 95.3 Ca (as C1) .. SO, (as H,SO,)' ' .. Na (as Cl) .. .. 480 79.5 85.9 90.1 4 000 1000) ca (as Cl) . . .. Mg (as Cl) . . .. 2 000 . . . . 200} 85 Ca (asCl) .. .. Fe (as Cl) . Mg (as Cl) . . . . 200 89.9 98.4 M g (as Cl) . . .. 1000 66.2 92.7 100.8 should be divided by 6 when related to the analyte solution. * The concentrations shown represent the final concentrations in the nebulised solution and to 3% m/V did not appear to offer any further significant reduction in the inter-element effects. Instrzcmental operating conditions Table I1 gives the optimised instrumental operating conditions.Automatic background correction at the 357.9-nm chromium line is not to be recommended, as balancing the hydrogen and chromium lamp intensities at this wavelength is not easy and a severely degraded chromium detection limit is normally observed. Sequential background correction using the lead 357.3-nm non-resonance line wits used and found to be satisfactory. Table I11 gives the typical background-absorption signals (expressed as a chromium concentration) from the main matrix elements and from some typical samples. It can be seen that the presence of sulphate significantly enhances thie background absorption from calcium and magnesium. However, this table shows that the background-absorption signals for most natural-water and sewage-effluent samples are irelatively small.TABL:E I1 OPTIMISED INSTRUMENTAL OPERATING CONDITIONS Wavelength . . . . . . . . . . . . 357.9 nm Background correction wavelength . . . . 357.3 nm (Pb) Slit width . . .. . . . . .. . . 0.5nm Airflow .. .. .. . . . . . . As recommended in handbook Acetylene flow . . . . .. .. . . Flame on verge of luminosity (no yellow luminosity visible) Distance from top of burner grid to pos:ition where the grid just intercepted the light beam (0.01 absorbance) . . .. .. . . . . 3.5mm Integration period . . . . . . . . . . 3 s Wash solution . . . . . . . . . . 3% V / V hydrochloric acid (36% 44March, 1979 WATERS AND EFFLUENTS BY AAS USING AN AIR - ACETYLENE FLAME TABLE I11 TYPICAL BACKGROUND-ABSORPTION SIGNALS USING THE 357.3-nm LEAD NON-RESONANCE LINE All solutions contained 2% m/V ammonium perchlorate and 10% V / V hydrochloric acid (36% m/m).227 Substance added or Concentration/ sample p g ml-l Ca (as C1) . . .. 10000 .. 10 000 9 600) Ca (as C1) SO, (as H2So4)' * .. Na (as C1) . . .. 10000 SO, (as 'H2S04) .. Mg (as C1) .. .. 10000 Mg (as Cl) . . . . 10 000 .. 10000 9 GOO} Na (as C1) . . SO, (as H,S04) .. 9 600) SO, (as H2S04) .. 19 200 Tap water 1* . . River water 2* River water 3* .. Sewage effluent 6* . . ' '} See Table V Background-absorption signal/pg ml-1 of chromium 0.09 0.22 0.03 0.035 0.17 0.26 <0.01 <0.002 0.007 0 0.009 0 0.006 0 * The background-absorption signals were measured after the samples were concentrated by evaporating to one fifth of volume. The observed signals were then divided by 5.Comerratvation techniqzle A five-fold concentration step of 50ml to 10 ml was found to result in an acceptable detection limit (less than 0.005 pg rnl-l), tolerable inter-element effects and very low back- ground absorption. The time for the evaporation step was approximately 1.5 h. Hydrochloric acid was used rather than nitric acid, as previous work had shown that the background-absorption signal at 357.3 nm from 10000 pg ml-l of calcium (as the chloride) was increased approximately four-fold in the presence of 10% V/V nitric acid (70% m/m), but was unaffected by the presence of 10% V/V hydrochloric acid (36% m/m). Also, a small, but significant, negative background-absorption signal had been observed with strong nitric acid solutions at 357.9 nm. However, nitric acid is the preferred acid for other types of samples (e.g., sewage sludge) .20 Method NOTE- This method should not be used with trade wastes or any unknown samples for safety reasons.alternative method is given below (see Alternative Method). An Volumes (50 & 0.5ml) of the samples, standards and blanks were placed in 100-ml borosilicate glass beakers and 4 ml (&O.l ml) of 25% V/V hydrochloric acid, 2 ml (50.05 ml) of 10% m/V ammonium perchlorate solution and some aluminium oxide anti-bumping granules were added. The beakers were then placed on a hot-plate with the temperature set such that gentle simmering occurred. The evaporation step was carried out in a fume cupboard and all normal safety precautions were observed. When the volume of the solution had decreased to 20 ml (-&5 ml), 0.5 ml (&0.05 ml) of hydrogen peroxide (6% m/m) was then added. This ensured that any chromium(V1) would be converted into chromium(III).21 The evaporation was then continued until the final volume was about 5 ml (rf 1 ml) and this took approximately 1.5 h.The solutions were allowed to cool and the contents transferred into the 10-ml calibrated borosilicate glass tubes. The beakers were228 THOMPSON AND WAGSTAFF: DETERMINATION OF CHROMIUM IN Analyst, VoZ, 104 carefully washed out using three approximately 1.5-ml washes with de-ionised water from a wash-bottle with a very fine nozzle. The contents of the tube were then diluted to volume and shaken and any suspended matter was allowed to settle prior to nebulisation.The final acid concentration during the evaporation step, in conjunction with the addition of hydrogen peroxide, should ensure adequate digestion of particulate and organically bound chromium in natural waters and effluents. The solutions in the beakers do not boil dry if the temperature of the hot-plate is carefully set because the presence of the ammonium perchlorate significantly increases the boiling-point of the liquid, and hence decreases the rate of evaporation as the solution approaches dryness. It is not normal practice to heat solutions of perchlorate in the presence of organic matter, but it should be stressed that each beaker contains only 200 mg of ammonium perchlorate (equivalent to 170 mg of perchloric acid). The method is applicable only to river samples, potable waters and sewage final efluents and the solution in the beaker should not boil dry.In order to attempt to evaluate the risk. of explosion, 10 ml of an industrial digested sludge containing 6% of dry solids with levels of over 3000 pg g-l of copper and zinc were added to three beakers, diluted to 50ml and1 carried through the procedure. When the evaporation stage was nearly completed the temperature of the hot-plate was increased to the maximum and the solutions in the beakers were allowed to boil dry. Although violent spitting and some deflagration were observed, no explosive reaction occurred. Numerous similar tests with 50 ml of 5000 and 10000 pg ml-l glucose solutions and 50 ml of a 1000 pg ml-l glucose solution containing 0.5 nil of vegetable oil yielded similar results.It is generally agreed that carbohydrates, and vegetable oils and fats constitute a particular hazard in the presence of perchloric acid.22 The use of 10-ml calibrated stoppered tubes rather than 10-ml calibrated flasks was considered accurate enough for this type of analysis and facilitated solution transfer and storage of large numbers of samples. Some laboratories will prefer to avoid boiling down a solution containing ammonium perchlorate and an alternative method was tested. This involved using a larger initial volume so that the ammonium perchlorate could be incorporated after the evaporation stage and allow efficient washing of the beaker used for the evaporation. Caution-During the whole course of this study no violent reactions were observed, but if the technique was scaled up the risk would almost certainly increase.Alternative Method Volumes (100 -+ 1 ml) of the samples, standards and blanks were placed in 150-ml boro- silicate beakers and 8 ml (rfi0.2 ml) of 25% V/V hydrochloric acid and some aluminium oxide anti-bumping granules were added. The beakers were then placed on a hot-plate and gently simmered until the solution volume had decreased to 40 ml (&lo ml), then 1 ml (kO.1 ml) of hydrogen peroxide (6% m/m) was added. The evaporation was continued until the final volume was about 8 ml (&2 ml). The solutions were allowed to cool, 4 ml (kO.1 ml) of 10% m/V ammonium perchlorate then added and the contents transferred into 20-ml calibrated borosilicate glass tubes.The beakers were carefully washed using four approximately 2-ml washes with de-ionised water from a wash-bottle with a very fine nozzle. The contents of the tube were then diluted to volume and shaken and any suspended matter was allowed to settle prior to nebulisation. All the results quoted in this paper were obtained by using the original method except for some precision measurements. No significant blanks were observed but prior to this work base-line noise and drift were observed during chromium determinations on humid days. Water droplets could be seen in the tube connecting the air supply from the air compressor water trap to the atomic- absorption spectrophotometer, This probleni was overcome by inserting an additional water trap in the air line. Inter-element Effects Table IV (in addition to Table I) shows the effect of various substances on the chromium response.It can be seen that for typical levels of the main matrix elements in natural waters and sewage effluents, the method would appear to be satisfactory. Aliquots of tap water, sample 1 (see Table V), were spiked with equal amounts (0.04 and 0.4pgml-1) ofMarch, 1979 WATERS AND EFFLUENTS BY AAS USING AN AIR - ACETYLENE FLAME TABLE IV EFFECT OF VARIOUS SUBSTANCES ON THE CHROMIUM RESPONSE All solutions contained 2% m/V ammonium perchlorate and 10% V/V hydrochloric acid (36% m/m). Instrumental conditions as in Table 11. Concentration of Chromium interfering substance? / concentration*/ Interfering substance* pg ml-l pg ml-l Relative signal 2 100.0 2 98.7 2 000 200} 2 96.6 100 None - Ca Mg Ca Mg PO, (as P) Ca 100 1000 Mg c u Zn Ca Mg Fe 200 None 1 100.0 500 1 100.0 Ca 100 Mg SiO, 500 100) 1 96.1 Ca 50 Mg Detergent (Mannoxol) 500 100) 1 99.6 Ca 500 Mg so* 500 loo} 1 100.8 Ca Mg PO, (as P) 50 Ca 5001 1000 loo} 2 95.0 loo} 2 97.4 229 Mg NH, (as N) NO, (as N) Ca SiO, Ca Zn Mn Mg Mg c 1 99.6 500 100 1000 100 99.7 97.1 * All cations were added as chlorides.SO, was added as H,SO,, SiO, as sodium silicate, t The concentrations shown represent the final concentrations in the nebulised solution phosphates as (NH,),HPO, and NH, and NO, as NH,NO,. and should be divided by 5 when related to the analyte solution. chromium( 111) and chromium(V1) and taken through the procedure, and no significant difference in response for the two oxidation states was detected.Results The technique was initially tested using a sample of tap water with a moderate total230 THOMPSON AND WAGSTAFF : DETERMINATION OF CHROMIUM IN Analyst, Vol. 104 ANALYTICAL RESULTS ON SAMPLES USED, PRIOR TO CONCENTRATION BY EVAPORATION Sample No. 1 2 3 4 5 6 7 Total hardness Source (CaCO,) Tap water . . . . . . 238 River water . . .. . . 516 River water . . .. . . 538 River water . . .. .. 327 Sewage works final effluent . . 563 Sewage works final effluent . . 593 Contentlpg ml-l - h Total Mg Na Fe 12 11 (0.03 43 35 0.38 51 460 0.20 15 25 1.7 24 44 0.65 27 78 0.41 1 Conductivity/ SO,a- C1- pScm-l 50 18 405 302 34 986 403 800 2 900 119 37 587 348 39 1093 296 94 1431 hardness value (see Table V). A large sample of the tap water was acidified with hydro- chloric acid to pH 2.5 and split into four aliquots, three of which were then spiked with 0.02, 0.2 and 0.4 pg ml-l of chromium(II1). The resulting samples were then analysed in duplicate over nine days and Table VI gives the results of these analyses.The alternative method, in which the ammonium perchlorate was added after the evaporation stage, showed within- batch standard deviations comparable to those in the original method (see Table VII). TYPICAL Sample Unknown standard TABLE VI PERFORMANCE DATA FOR TAP WATER, SAMPLE NO. 1, ANALYSED IN DUPLICATE OVER NINE DAYS Added chromium/ Mean chromium Standard deviation/ pg ml-l result/pg ml-l pg ml-l I (0.2 pg ml-l) . . .. Tap water .. .. Tap water . . .. 0.02 Tap water .. ..0.2 - Tap water .. .. 0.4 Blank . . .. .. - 0.197 0.001 3 0.0204 0.200 0.396 - 0.004 94 0.000 93 0.001 2 0.004 37 0.01 11 0.000 92* * The within-batch standard deviation (s) of the blank was calculated using the nine sets of paired blank r e s ~ l t s . ~ Hence, the 4.65 s detection limit is 0.0043 pg ml-l. Recovery tests using the original method were then carried out using three river-water samples and two sewage works final effluents. These samples were selected as they exhibited high conductivities and high sulphate levels and would be expected to give a good indication of maximum potential inter-element effects. Table V gives the main matrix element concentrations for these samples. Table VIII shows the recoveries obtained after the addition of 0.04 and 0.4pgml-l of chromium(II1) to these samples, and also shows some recoveries obtained in the absence of ammonium perchlorate.The natural chromium level in the samples, except for sample 2, was below the detection limit. Sample 2 was found to contain 0.015 pug ml-l of chromium. Using the electrothermal atomisation technique utilising the standard additions method of calibration, a chromium level of 0.016 pg m1-l was found in this sample. It can also be TABLE VII COMPARISON OF ORIGINAL AND ALTERNATIVE METHODS Sample Standard . . * . Tap water . . .. Within-batch standard deviationlpg ml-l (9 degrees of freedom) p g ml-L method method 0.2 0.002 0 0.001 8 0.2 0.002 2 0.001 9 - Added chromium/ Origmal AlternativeMarch, 1979 WATERS AND EFFLUENTS BY AAS USING AN AIR - ACETYLENE FLAME 231 seen from Table VIII that the proposed method would appear to be satisfactory for natural water and sewage final effluent analysis and that in the absence of ammonium perchlorate poor recoveries were observed, especially if the flame conditions were set for maximum response (Le., a distinctly luminous flame).TABLE VIII RECOVERY TEST RESULTS Chromium recoveredl pg ml-1 f A Flame on verge of luminosity (Table 11) Sample Sample No. River water . . 2 River water . . 3 River water . . 4 Sewage final effluent . . 5 Sewage final effluent .. 6 Chromium added (0.04 pg ml;l) plus ammonium perchlorate 0.040 0.037 1 0.039 6 0.0386 0.038 2 Chromium added (0.4 pg ml-l) plus ammonium perchlorate 0.380 0.379 0.404 0.387 0.370 7 ~- Chromium added (0.4 pg ml-l) and no ammonium perchlorate 0.293 0.296 0.332 0.312 0.306 Luminous flame with chromium added (0.4 p g ml-l) and no ammonium perchlorate 0.188 0.184 0.230 0.243 0.246 Conclusions The concentration by evaporation technique with the addition of ammonium perchlorate and hydrochloric acid would appear to be a rapid pre-concentration technique for the atomic-absorption spectrophotometric analysis of total chromium in natural waters and sewage final effluents using the air - acetylene flame.A detection limit of 0.0043 pg ml-l (4.65 s) was obtained and inter-element effects were considered acceptable. The proposed technique has the additional advantage that other elements can also be determined. The authors thank Mr. W. F. Lester, Director of Scientific Services, Severn-Trent Water Authority, for permission to publish this work.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. References “World Health Organization European Standards for Drinking Water,” Second Edition, 1970. EEC Directive, Oflcial Journal of the European Communities, 75/440/EEC, July 1975. Wilson, A. L., and Cheeseman, R. V., “Manual on Analytical Quality Control for the Water Industry,” Water Research Centre Technical Report TR66, Water Research Centre, Medmenham, 1978. Rawa, J. A., and Henn, E. L., Am. Lab., 1977, 9 (8), 31. Thompson, K. C., and Reynolds, R. J ., “Atomic Absorption, Fluorescence and Flame Emission Spectroscopy, A Practical Approach,” Griffin, London, 1978. Midgett, M. R., and Fishman, M. J., Atom. Absorption Newsl., 1967, 6, 128. Gilbert, T. R., and Clay, A. M., Analytica Chim. Acta, 1973, 67, 289. Hurlbut, J. A., and Chriswell, C. D., Analyt. Chem., 1971, 43, 465. Barnes, L., Analyt. Chem., 1966, 38, 1083. Purushottam, A., Naidu, P. P., and Lal, S. S., Talanta, 1973, 20, 631. Ottaway, J. M., and Pradhan, N. K., Talanta, 1973, 20, 927. Kodama, M., Shimizu, S., Sato, M., and Tominaga, T., Analyt. Lett., 1977, 10, 591. Kraft, G., Lindenberger, D., and Beck, H., 2. Analyt. Chem., 1976, 282, 119. Thompson, K. C., Analyst, 1978, 103, 1258. Tenny, A. M., Instrum. News, 1967, 18, 14. Burke, K. E., and Albright, C. H., Dev. Appl. Spectrosc., 1970, 8, 33. Oguro, H., Nippon Kagaku Kaishi, 1976, 7 , 1409. Oguro, H., Nippon Kagaku Kaishi, 1977, 8, 225. Oguro, H., Nippon Kagaku Kaishi, 1977, 8, 218. Thompson, K. C., and Wagstaff, K., “Development of a Technique for the Analysis of Certain Toxic Metals in Sewage Sludges,” Internal Report No. TP116/ML19, Severn-Trent Water Authority , Malvern . Vogel, A. I., “A Text Book of Macro and Semimicro Qualitative Inorganic Analysis,” Longmans, Green, London, 1955. Gorsuch, T. T., “The Destruction of Organic Matter,” Pergamon Press, Oxford, 1970. Received September 7th, 1978 Accepted October 17th, 1978