摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 249 SHORT PAPERS Determination of Dissolved Sulphide in Groundwaters by Inductively Coupled Plasma Atomic Emission Spectrometry* Kathryn Lewin and J. Nicholas Walsh Geology Department, Royal Holloway and Bedford New College (University of London), Egham Hill, Egham, Surrey W 2 0 OEX, UK Douglas L. Milest Hydrogeology Research Group, British Geological Survey, Wallingford, Oxfordshire OX 10 BBB, UK A rapid and simple inductively coupled plasma atomic emission spectrometric (ICP-AES) technique for the determination of dissolved sulphide at the parts per billion level in groundwater is described. Sulphide is stabilised on sampling by the addition of KOH, converted into H2S under acidic conditions and stripped from solution in a gas - liquid separator by a stream of argon.Sulphur emission intensity is monitored at 180.7 nm. Dissolved sulphate does not interfere and spectral interference from calcium is avoided. Keywords: Sulphide determination; hydrogen sulphide; water; inductively coupled plasma atomic emission spectrometry Information on the redox conditions prevailing in ground- water systems is important because of their influence on the release and mobility of many elements, including several that are potentially toxic.1.2 While EH values measured with a platinum electrode provide a useful semi-quantitative guide,2J the direct determination of dissolved reduced species such as sulphide is highly desirable. Several methods for the determination of sulphide in waters have been described, including colorimetry,4 titrimetry,5 atomic fluorescence spec- trometry,6 ion chromatography7 and the sulphide ion-selec- tive electrode,g all of which have limitations in terms of sensitivity, speed or interference effects. Direct nebulisation of water samples into an ICP and the monitoring of sulphur emission intensity at 180.73 nm has been used successfully to determine sulphate in oxidising groundwaters, provided that a correction is made for spectral interference from a nearby calcium line.9 During the applica- tion of this technique to a wide variety of groundwaters, very high apparent concentrations of sulphur were occasionally detected during the analysis of waters from low EH environ- ments.Enhanced sensitivity arising from the liberation of hydrogen sulphide from solution was suspected as the cause of these erroneous values. Initial experiments suggested that this phenomenon could be used as the basis of a method for the determination of trace levels of dissolved sulphide by induc- tively coupled plasma atomic emission spectrometry (ICP- AES).Experimental Apparatus Plasma emission measurements were made using an Applied Research Laboratories 34000C system; operating conditions are given in Table 1. Initial experiments were carried out using a sealed gas - liquid separator based on a Dreschel bottle. To overcome problems of plasma instability during sample changing and to reduce the volume of the reaction vessel, a continuous flow separator was later constructed (Fig. 1). Argon flow-rates were controlled by a GEC-Philips rotameter.A Watson- Marlow 502s variable speed peristaltic pump was used to control liquid flow-rates. Reagents Reagents were of analytical-reagent grade or better. A stock solution of sulphide (ca. 0.1 M) was prepared by dissolving 12 g of Na2S.9H20 in 500 ml of de-ionised water; dilute working solutions were prepared daily and standardised against 0.1 M AgN03 solution. Sulphide standard solutions and samples were stabilised by the addition of KOH to give a 1 g 1-1 solution. Sampling Groundwater samples were taken via polyethylene tubing attached to suitable taps on the borehole headworks and filtered under well-head pressure through 0.45-pm pore size Table 1. Plasma operating conditions Spectrometer . . . . ARL 34000C Quantovac; grating ruled Hamamatsu R306 photomultiplier tube 1080 lines mm-l Wavelength .. . . 180.73 nm, 3rd order Frequency . . . . 27.12MHz Forward power . . 1200 W Reflectedpower . . 5 W Argon flow-rates . . Coolant 11, plasma 1.2, Read-out . . . . Mean of three 5-s integrations external light path purge 1.5 1 min-l 1% HCI Sample + I, Argon ICP /_4 Waste * Presented at the Third Biennial National Atomic Spectroscopy t To whom correspondence should be addressed. Symposium (BNASS), Bristol, UK, 23rd-25th July, 1986. Fig. 1. Gas - liquid separator250 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 5 - > E 4 - 2 .- ; 3 - C .- C In .- .z 2 E u1 1 - - 1 I I I 1 0 20 40 60 80 100 HCI plus sample flow-rate/mI min-’ Fig. 2. Effect of liquid flow-rate on the emission intensity of 2 pg 1-1 of sulphide at an argon flow-rate of 0.35 1 min-1 0 0.25 0.50 0.75 Argon flow-rate/l min-’ Fig.3. Effect of argon flow-rate on the emission intensity of 5 pg 1-’ of sulphide at a liquid flow-rate of 25 ml min-1 for both HC1 and sulphide solutions 20 f i O l 2 3 4 5 min I 3.9 wcl I - ’ I Time -+ 5 Fig. 4. Recorder trace showing emission intensity at 180.7 nm during the nebulisation of three groundwater samples containing trace amounts of sulphide membrane filters. Samples were stabilised by the addition of KOH to give a 1 g 1-1 solution. Results and Discussion Gas and liquid flow-rates through the separator greatly influenced emission intensity. Increased sensitivity could be obtained at high liquid flow-rates (Fig. 2) but, to keep sample and reagent consumption down to a reasonable level, flow- rates of 25 ml min-1 for both the sample and hydrochloric acid streams were used throughout.At low gas flow-rates, slow mixing of solutions and inefficient stripping of H2S resulted in low emission intensities, whereas high flow-rates forced liquid out of the reaction vessel prematurely. An argon flow-rate of approximately 0.35 1 min-1 was found to be optimum (Fig. 3). Typical signals obtained from three groundwater samples containing low concentrations of sulphide are shown in Fig. 4. The calibration is linear up to at least 100 pg 1-1 of sulphide. During the analysis of groundwaters by this technique, it is clearly vital that no interference from sulphate-sulphur occurs, e.g., by carryover of aerosol into the ICP.The insensitivity of the method to sulphate was investigated by analysing sulphide-free solutions containing 1000 mg 1-1 of sulphate; no increase in emission intensity above background was recorded. The 180.73-nm sulphur line suffers from slight interference from a partially overlapping calcium line.9 Although the intensity of this calcium line is weak compared with that of the 180.73-nm sulphur line, serious interference could occur during the determination of trace amounts of sulphide in samples containing typical environmental concen- trations of calcium if sample matrix reaches the ICP. To test the efficiency of the separator , sulphide-free solutions con- taining 1000 mg 1-1 of calcium were analysed; no increase in emission intensity above background at 180.73 nm was detected.Simultaneous monitoring of the calcium 319.9-nm line also showed no increase in signal. Potential interference in the liberation of HZS by several metals commonly present in groundwaters at low concentra- tions was investigated by spiking 10 pg 1-1 sulphide standard solutions with FeII, Fe”1, Zn, Cu and A1 in the range 0.1-100 mg 1-1. Not surprisingly, copper added at the higher concen- trations caused immediate precipitation of sulphide as CuS. However, in natural groundwaters in equilibrium with any sulphide minerals present, free dissolved sulphide would still be liberated. Aluminium did not interfere significantly below 1 mg I-l, which is approximately 1000 times greater than its typical concentration in groundwater, and no interference was observed from iron or zinc.A 3a detection limit calculated from the background noise was 0.2 pg 1-1 of sulphide. A within-batch relative standard deviation of 0.45% was obtained from 20 mea:,rements at 10.5 pg 1-1 of sulphide. The analysis time was typically 3 min per sample. Comparable performance was achieved when the separator was used in conjunction with a Philips Model PV8490 ICP operating at 50 MHz and a SOPRA Model F400 (68 Rue Pierre Joigneaux, F92270 Bois-Colombes, France) vacuum scanning monochro- mator. The technique has been used successfully to determine trace amounts of dissolved sulphide in numerous ground- waters from major UK aquifers. Financial support for this work from the Natural Environment Research Council is gratefully acknowledged. This paper is published with the permission of the Director, British Geological Survey (NERC). 1. 2. 3. 4. 5. 6. 7. 8. 9. References Edmunds, W. M., Miles, D. L., and Cook, J. M., in Eriksson, E., Editor, “Hydrochemical Balances of Freshwater Systems,” IAHS-AISH Publication No. 150, Berlings, Arlov, 1984, p. 55. Champ, D. R., Gulens, J., and Jackson, R. E., Can. J . Earth Sci., 1979, 16, 12. Whitfield, M., Limnol. Oceanogr., 1974, 19, 857. Rees, T. D., Gyllenspetz, A. B . , and Docherty, A. C., Analyst, 1971,96, 201. Department of the Environment, “Sulphide in Waters and Effluents 1983, Tentative Methods,” HMSO, London, 1983, p. 6. Shahwan, G. J., and Heithmar, E. M., Spectrosc. Lett., 1984, 17, 377. Rocklin, R. D., and Johnson, E. L., Anal. Chem., 1983,55,4. Gulens, J., Water Res., 1985, 19, 201. Miles, D. L., and Cook, J. M., Anal. Chim. A m , 1982, 141, 207. Paper J6/82 Received September 9th, 1986 Accepted September 26th, 1986

ISSN:0267-9477    

DOI:10.1039/JA9870200249

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 25 1 Excitation of Gallium, Indium, Selenium, Tellurium, Arsenic and Antimony in a Helium Microwave-induced Plasma Katherine J . Tim m i ns Directorate of Quality Assuranceflechnical Support, Materials Centre, Royal Arsenal East, Woolwich, London SE 18 6TD, UK The radial distributions and detection limits for the excitation of the most sensitive atom lines of Ga, In, Se, Te, As and Sb in an atmospheric pressure microwave-induced helium plasma have been ascertained. Keywords: Microwave-induced plasma; atomic emission spectrometry; radial distribution; detection limits; volatile elem en ts Spatial studies on the inductively coupled plasma (ICP) have been extensively investigated (see, for example, reference 1), but less research has been carried out on microwave-induced plasmas (MIPs), probably owing to their much smaller dimensions.Most of the work that has been reported concerns axial distributions, which have been shown to be highly dependent on gas, flow-rate, power,* the interaction with the quartz capillary wall,3 the element4 and whether the line observed is due to an atomic or ionic species.5 Very few studies have dealt with radial distributions,697 including the three-filament and toroidal Ar MIPs of Kollotzek et al.,g in which the plasma is centred in the silica tube. The excitation of Ga introduced into a He MIP at atmospheric pressures by means of a heated Ta filament9 (electrothermal vapourisation, ETV) was reported10 as being optimum (for 417.206 nm) in the position shown in Fig.1, i.e., at the outer edge of the plasma, but approximately central in the plasma tube. The elements Ga, In, Se, Te, As and Sb have been similarly investigated as a preliminary study f o r the determination of these volatile elements in nickel-base alloys, and the results are presented here. Experimental and Results The equipment and parameters used were as follows: mono- chromator, Hilger Monospek 1000, 1200 lines mm-1 grating; slits, 10 pm; photomultiplier tube, Hamamatsu R955; 1 : 1 image of MIP on slit; readout, Gay - Milano tracking voltmeter, 30 ms f.s.d., or Tekman TE200 recorder, 0.3 s f.s.d.; cavity, Beenakkerll TMolO; generator, EMS Microtron 200 Mark 3; microwave power, forward, 75 W; gas, helium, 1.1 1 min-l; plasma tubes, silica, inner 5 mm o.d., 2.5 mm i.d., outer (for support) 5 mm i.d.; sample device (see Fig. 2), filament, tantalum, 0.5 mm diameter; power supply, Lambda LES-F-02-OV-VI; d.c.current, dry, 4 A, volatilise, Ga, In 24 A, Te, Sb and As 22 A and Se 21 A; sample size, 5 pl. The elements were introduced as solutions in dilute nitric acid (1 + 49). No study was made of the effect of possible matrix interferences on these results or of a change in the filament material, at this stage. Initially, the system was set up using a recorder but, in the course of the study on Ga, it was found that a faster method of data capture was necessary, so all final data have since been recorded as voltages on a tracking voltmeter. The current used to volatilise the elements was the only experimental parameter (apart from the wavelength) that was altered.The minimum current necessary to obtain the maximum stable voltage reading was chosen for each individual element. * Presented at the Third Biennial National Atomic Spectroscopy Copyright Controller HMSO, London, 1986. Symposium (BNASS), Bristol, UK 23rd-25th July, 1986. Fig. 1. the optimum position for the analysis of gallium Photograph of the image of the plasma on the slit, showing ( a ) 5 9 8 7 A 6 8 I n 2 F: 1 1 I / / \L/ I -To MIP \ I \ \ '9 i '6 '9 Fig. 2. Schematic diagram of the sample introduction unit (ETV). (a) Unit insert, from top; ( b ) assembled unit, from side. 1, Gas in; 2, gas out; 3, sample injection port; 4, stopper; 5 , tungsten rods; 6 , wire filament; 7, silica sleeves; 8, screw joint; and 9, borosilicate glass body The investigation was carried out, for comparison purposes, under fixed conditions of gas flow, microwave power, etc., which have been found9 to be suitable for the ETV-MIP system.Under these conditions, it was found that In and As behave as Ga, and give a maximum emission intensity at approximately the same position. However, Se, Te and Sb have a broad profile over the whole width of the plasma tube252 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 -1.0 -0.5 0 0.5 1 .o Fig. 3. Typical curves showin the signal to noise ratios across the plasma tube for: (A) Ga, In an$ As; and (B) Se, Sb and Te when the plasma is attached to the left-hand wall of the tube (as in Fig. 1) Table 1. Detection limits (3a) for some sensitive lines of Ga, In, As, Sb, Se and Te Distance across tube/mm Wavelength/ Element nm Ga .. 417.206 403.298 294.364 287.424 In . . 451.132 410.177 303.936 As , , 278.020 274.499 234.984 228.812 Sb . . 259.806 231.147 217.926 217.589 206.838 Se . . 206.279 203.985 196.090 Te . . 238.576 238.325 225.904 214.275 Detection limit/ ng ml- 1.2 1.7 0.6 0.9 0.7 0.7 0.9 70 167 48 49 24 4.2 42 14.6 24 78 190 90 53 104 25 100 Lowest previously published detection limit/ ng ml-1 (unless indicated otherwise) 3 2.7 X 10 g s-’ - 100 40 100 - 2 16 1.9 40 - 10 000 20 - 200 Reference 8 14 15 16 2 - - - - 17 18 17 19 - - - - 20 19 - - - 20 (see Fig. 3). The spectrum of each element was scanned using a glass frit nebuliser with desolvationl2 to introduce the sample.Detection limits for the most sensitive emission lines for each element were determined using ETV sample intro- duction and are given in Table 1. At the position at which the plasma excites Ga, As and In, the helium background emission is weak and so is that due to molecular species (OH, NH, N, etc.). However, although 228.8 nm is a sensitive line for As, it lies on the NO-yl3 molecular band, which is still sufficiently intense, even at this position, to reduce the signal to background ratio, thus leading to a poorer detection limit than might have been expected. For Se, Te and Sb, on the other hand, one may observe the plasma tube image at any radial position and so avoid the detrimental effect of background emission. In most instances where a previously published best detection limit has been given, a direct comparison with an electrothermal technique is not available.Other ~orkers2~15,16,20 used nebulisation as the method of sample introduction. Argon MIPS or a nitrogen capacitatively coupled plasma15 were employed, so again the comparison is not equal but, nevertheless, the value of removing the water before the sample is atomised is striking, particularly for In and the 203.985-nm line of Se. Fricke et aZ.19 used a carbon cup, but values for a tantalum strip were also given: 120 ng ml-1 for Se at 196.090 nm and 100 ng ml-1 for Sb at 231.147 nm, so an investigation into the filament material may well be of value for certain elements. Atsuya et al.ls used potassium to enhance the intensity of the As 228.8-nm line, but it was not thought appropriate to do this here as the matrix was to be a nickel base alloy.Dagnall et aZ.14 used a gas chromatograph - MIP system and Mulligan et al.17 used hydride generation. Conclusion The results have demonstrated that, for the emission lines cited, some elements (Ga, In and As) have an optimum radial position in the helium MIP, whereas others (Se, Te and Sb) have a broad profile over the whole plasma. It is advisable, therefore, to determine the best position for individual elements, as one may be able to select the radial part of the plasma tube image, which avoids any interfering molecular bands, and thus improve the signal to background ratio. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. References Furuta, N., and Horlick, G., Spectrochim.Acta, Part B , 1982, 37,53. Kawaguchi, H., Hasegawa, M., and Mizuike, A., Spectrochirn. Acta, Part B , 1972, 27,205. Wandro, R. F., and Friedrich, H. B., Anal. Chern., 1984,56, 2727. Matousek, J. P., Orr, B. J., and Selby, M., Appl. Spectrosc., 1984, 38, 231. Estes, S. A., Uden, P. C., and Barnes, R. M., Anal. Chem., 1981, 53, 1829. Freeman, J. E., and Hieftje, G. M., Spectrochim. Acta, Part B , 1985,40, 653. Michlewicz, K. G., and Carnahan, J. W., Anal. Chem., 1985, 57, 1092. Kollotsek, D., Tshopel, P., and Tolg, G., Spectrochim. Acta, Part B , 1984,39, 625. Brooks, E. I., and Timmins, K. J., Analyst, 1985, 110, 557. Timmins, K. J., “Proceedings CSI XXIV,” 1985, Paper No. B028, p. 220. Beenakker, C. I. M . , Spectrochim. Acta, Part B , 1976,31,483. Stahl, R. G., and Timmins, K. J., “Proceedings of the 1987 Winter Conference on Plasma and Laser Spectrochemistry, January 12-16, 1987, Lyon, France, C25.” Pearse, R. W. B., and Gaydon, A. G., “The Identification of Molecular Spectra,” Fourth Edition, Chapman and Hall, London, 1976, p. 237. Dagnall, R. M., West, T. S . , and Whitehead, P.,Analyst, 1973, 98, 647. Dahmen, J., ICP Inf. Newsl., 1981,6,576. Fallgatter, K., Svoboda, V., and Winefordner, J. D., Appl. Spectrosc., 1971, 25, 347. Mulligan, K. J., Hahn, M. H., Caruso, J. A., and Fricke, F. L., Anal. Chem., 1979, 51, 1935. Atsuya, I., Alter, G. M., Veillon, C., and Vallee, B. L., Anal. Biochem., 1977,79, 202. Fricke, F. L., Rose, O., Jr., and Caruso, J. A., Anal. Chem., 1975, 47, 2018. Hingle, D. N., Kirkbright, G. F., and Bailey, R. M., Talanta, 1969, 16, 1223. Paper J6l8.5 Received September 17th, I986 Accepted December 8th, I986

ISSN:0267-9477    

DOI:10.1039/JA9870200251

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 253 Use of 1,5-Bis(di-2-pyridylmethylene)thiocarbonohydrazide as an Extracting Reagent for the Determination of some Transition Metal Ions by AiomicAbsorption Spectrometry* A. Bustos, F. Sanchez-Rojas, C. Bosch Ojeda, A. Garcia de Torres and J. M. Can0 Pavon Department of Analytical Chemistry, Faculty of Sciences, University of Malaga, 2907 I Malaga, Spain 1,5-Bis(di-2-pyridylmethylene)thiocarbonohydrazide, DPTH, has been studied as an extracting reagent for Coil, Ni", Cull, Fell, Znll and Cd". An atomic absorption spectrometric method for the determination of trace amounts of cobalt in solution with DPTH is described. This compound reacts with cobalt in an acidic medium to produce a complex that can be extracted into chloroform.The sensitivity of the method is 0.4 pg ml-l. The method has been applied satisfactorily to the determination of cobalt in several steel samples. Keywords : Co ba I t de te rm in a ti0 n ; a tom ic absorption spec t ro m e try; extraction ; I , 5- bis (di-2-p yrid ylm e th yl- en e) thioca rbon o h ydrazide In recent years several atomic absorption spectrometric methods, based on the separation and pre-concentration of cobalt by liquid - liquid extraction, have been repor- ted. 175-Bis(di-2-pyridylmethylene) thiocarbonohydrazide (DPTH) is a good chelating agent for metal ions. Most of the complexes formed are coloured and readily extractable into a range of organic solvents, particularly chloroform. Prepara- tive and analytical details of DPTHl and its application to the spectrophotometric determination of trace amounts of cobalt and nickel have reported previously.2 This paper decribes a method for the determination of trace amounts of cobalt by atomic absorption spectrometry after extraction using DPTH and chloroform.The method has been applied to the determination of cobalt in steel samples, and offers appreciable advantages in selectivity in comparison with other extraction reagents (e.g. , dithiocarbamates and nitroso- naphthols) . 3 The application of the reagent to the determination of other metal ions (e.g., zinc, cadmium, iron and copper) by liquid - liquid extraction combined with atomic absorption spec- trometry or ultraviolet - visible spectrophotometry is currently being investigated.Experimental Apparatus Absorbance measurements were made using a Shimadzu 240 spectrophotometer with 1 .O-cm silica cells. A Perkin-Elmer 2380 atomic absorption spectrometer was used for AAS measurements at the 240.7-nm cobalt line with an air - acetylene flame. Reagents Analytical-reagent grade chemicals were used throughout. DPTH solution. The ligand was synthesised as described previously.1 Stock solutions of 0.05-0.1% mIVDPTH in DMF and chloroform were prepared. These solutions are stable for at least a week. Standard solutions. Cobalt(II), copper(II), nickel(II), zinc(II), cadmium(I1) and iron(I1) standard solutions were prepared from their commercial salts (nitrate or sulphate) and standardised titrimetrically or gravimetrically . Working stan- dard solutions were prepared by suitable dilution of the standard solutions.* Presented at the Third Biennial National Atomic Spectroscopy Symposium (BNASS), Bristol, UK, 23rd-25th July, 1986. Procedures Liquid - liquid extraction of metal ions The extraction procedure was as follows. A volume of sample solution containing 10 pg of the metal ion was placed in a 100-ml separating funnel. The pH was adjusted by dropwise addition of sodium hydroxide or perchloric acid solutions. Sufficient sodium perchlorate solution was added to maintain an ionic strength of 0.1. The solution was diluted to 10 ml with distilled water and 10 ml of 0.05% mlV DPTH solution in chloroform were added. The funnel was agitated for 5 min on a mechanical shaker. After settling, the aqueous phase was collected and after filtration the equilibrium pH measured.To determine the amount of metal ion in the organic phase, a 5-ml portion was pippeted into a 10-ml calibrated flask and diluted to the mark with ethanol. The absorbance was measured by UV - visible spectrophotometry at the maximum wavelengths against a reagent blank. Determination of cobalt with DPTH A solution containing 10-160 pg of cobalt was taken in a 100-ml separating funnel; 16 ml of 0.1 M sodium perchlorate, 2 ml of 1.2 N perchloric acid and sufficient distilled water to maintain the total volume of the aqueous phase of 20 ml were added. An 8-ml volume of 0.1% DPTH solution in chloroform was added and the mixture agitated on a mechanical shaker for 5 min. The phases were allowed to separate and 5 ml of the organic phase were pipetted into a 10-ml calibrated flask and diluted to the mark with ethanol.This solution was aspirated into an oxidising air - acetylene flame and the atomic absorption was measured at 240.7 nm. The calibration graph was constructed with standard solutions treated in the same way. Determination of cobalt in steels Dissolve an accurately weighed sample (0.3-1.0 g) in ca. 30 ml of hydrochloric acid (1 + 1) and heat gently until dissolution is complete. Add 10 ml of nitric acid (1 + 1) and continue boiling until vapours are no longer evolved. Place an aliquot of this solution containing 10-160 pg of cobalt in a 100-ml separating funnel. The analysis is completed as described above. Results and Discussion Extraction of Metal Ions Experiments involving extraction of diverse metal ions with DPTH at different pH values have shown that the extraction is254 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL.2 0 2 4 6 8 0 2 4 6 8 PH Fig. 1. Extraction of metal ions into chloroform versus pH using DPTH as extracting reagent. Concentration of each ion, 3 pg ml-*; and volumes of aqueous and organic phase, 10 ml in each instance a 2 4 6 8 0.4 0.8 1.2 1 2 3 4 5 0.3 v,,, Vorg [NaC104]/10-* M [DPTH]/10-3 M Fig. 2. Influence of experimental variables on the extraction of Co" - DPTH complex into chloroform. (a) Influence of perchlorate concentration; ( b ) influence of reagent concentration (organic phase); and ( c ) effect of phase-volume ratio. Cobalt concentration 10 pg in each instance influenced strongly by the pH (Fig.1). At pH G 1.5, cobalt and copper complexes are quantitatively extracted, whereas the amounts of nickel(II), iron@), cadmium(I1) and zinc(I1) extracted are small; moreover lead(II), mercury(I1) and palladium(I1) are not extracted at all in very acidic media. Liquid - Liquid Extraction of Cobalt followed by Atomic Absorption Spectrometry Characteristics of the cobalt complex: choice of extracting solvent In order to avoid the large number of interferences in the homogeneous medium, the yellow 1 : 2 cobalt - DPTH complex was solvent extracted. The complex is soluble in some oxygen containing organic solvents, such in alcohols and ketones, but the reagent is not soluble in these. Chloroform was chosen as the extracting solvent because it gave the best separation of the two phases and the reagent is soluble in this medium.This extractive method has been used satisfactorily for the determination of trace amounts of cobalt by flame atomic absorption spectrometry. Effect of p H The extraction behaviour of the cobalt - DPTH complex was studied in the pH range 0.5-10.0. Results show that a quantitative extraction is produced between pH 0.8-4.0, hence all extraction studies were carried out at pH 1.0-1.5. At Hammett u function values of less than 0.8, an emulsion was produced. At higher pH values, incomplete extraction of the complex was observed. Results obtained for cobalt, as well as other metal ions are shown in Fig. 1. Extraction efficiency and stability The recovery factors for the extraction of cobalt were calculated by means of a series of experiments in which the atomic absorption of cobalt in the organic phase was com- pared with that of a standard prepared in water-saturated chloroform.In all instances, cobalt in the range 10-160 yg was extracted completely from the aqueous solution by a single extraction with 8 ml of 0.1% mlV DPTH solution in chloroform. A second extraction with another 8 ml of reagent solution showed that the cobalt content remaining in the aqueous phase was negligible. The atomic absorption of the cobalt complex remained constant when solutions were kept at room temperature for at least one week. Effect of reagent concentration, shaking time and ionic strength The effect of the initial concentration of DPTH in chloroform on the absorbance of the organic phase was studied by extracting 10 pg of cobalt from a solution at pH 1.3 in a single extraction.A DPTH concentration of 1.0 X 10-3 M was found to be adequate for the complete extraction and a concentra- tion of 1.3 x 10-3 M (8 ml of 0.1% mlV) was chosen to ensure a slight excess of the reagent. The extraction of the cobalt complex is rapid under the conditions recommended in the procedure. A shaking time of 5-6 min is sufficient for the complete extraction of the cobalt. Extraction of the cobalt(I1) - DPTH complex in an acidic medium was examined by varying the sodium perchlorate concentration [Fig. 2(a)]. A slight increase in the extraction was observed when the sodium perchlorate concentration in the aqueous phase was increased; this effect is not significant and can be attributed to a salting out effect. Effect of phase-volume ratio The volume of the aqueous phase was varied in the range 8-40 ml and the volume of the organic phase was kept constant at 8 ml (0.1% mlV DPTH), giving a phase-volume ratio range of from 1 to 5.As seen in Fig. 2(c), the extraction decreases with a phase-volume ratio of greater than two. Atomic absorption determination of cobalt Under the optimum conditions used a linear calibration graph was obtained for 0.5-8.0 pg ml-1 of cobalt in the aqueous phase. The sensitivity is 0.4 yg ml-1 (organic phase), so that the relative sensitivity according to the phase-volume ratio is 0.3 pg ml-1, when the volume of aqueous phase is 20 ml (extraction with 8 ml of organic phase). Ten determinations of standard solutions containing 1.8 yg ml-1 gave a relative standard deviation of 2.5% (P = 0.05).Study of Interferences The effect of diverse ions on the determination of cobalt was examined under the optimum conditions of the procedure. The tolerance limit was taken as that concentration which does not cause more than a 3% change in the atomic absorption. For these studies different amounts of the ionic species were added to a solution of 2.2 pg ml-1 of cobalt. The starting point was a 2000-fold mlm ratio of interferent to cobalt, and if interference occurred, the ratio was progressively reduced until interference ceased. The tolerance limits (Table 1) show that cobalt can be determined in the presence of a large number of diverse ions including most of those that are commonly associated with cobalt in natural and synthetic mixtures.Applications To evaluate the effectiveness of the method, a series of recovery experiments were carried out for several steel samples. The steel samples, supplied by ACERINOX, Algeci- ras, Spain, had the following given concentrations (as deter- mined by XRF, with calibration lines obtained using NBSJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 255 Table 1. Tolerance of foreign ions in the determination of 2.2 pg ml-1 of cobalt by the method proposed Tolerancelpg ml- Ion or species 4500 2300 700 500 250 200 50 7 I-, SCN-, As02-, hod3-, citrate, tartrate, Br- , C1-, C2042-, EDTA, P043-, B40,2--, Cr3+, Pb2+, Zn2+, Mn2+, U022+, S 2 0 3 * - , K+, Sr2+, Ba2+, Mg2+, Li+ Fe3+, F- , Br03- Cd2+ Sb3+ Ni2+, Mo6+, Ce4+, W6+ Hg+, Hg2+ cu2+ Ag+ ~ ~ Table 2.Determination of cobalt in samples of steels, n = 6 Cobalt content, % Sample Reported Found A . . . . . . . . 0.208 0.208 k 0.003 B . . . . . . . . 0.208 0.208 2 0.003 c . . . . . . . . 0.222 0.222 k 0.005 D . . . . . . . . 4.90 5.09 k 0.07 standard steels): steel A, C 0.064, Si 0.62, Mn 1.50, Sn 0.024, Ni 8.71, Cu 0.287, Cr 18.33, P 0.032, S 0.005, Mo 0.385, Ti 0.663, Nb 0.021 and Co 0.208%; steel B, C 0.019, Si 0.29, Mn 1.58, Sn 0.022, Ni 9.42, Cu 0.493, Cr 18.75, P 0.037, S 0.006, Mo 0.492, Ti 0.002, Nb 0.011 and Co 0.208%; steel C, C 0.035, Si 0.29, Mn 1.27, Sn 0.021, Ni 8.67, Cu 0.424, Cr 17.31, P 0.035, S 0.002, Mo 0.466, Ti 0.003, Nb 0.010 and Co 0.222%; and steel D, C 0.755, Mn 0.290, Si 0.316, Cr 4.223, P 0.016, S 0.009, V 1.077, W 18.63, Mo 0.0953 and Co 4.90%. The results of the analyses of steels A-D are given in Table 2. References 1. 2. 3. Bonilla Abascal, J. R., Garcia de Torres, A., and Can0 Pavbn, J. M., Microchem. J . , 1981,26, 55. Can0 Pavbn, J. M., Garcia de Torres, A., and Bosch Ojeda, C., Analyst, 1985, 110, 1137. Borggaard, 0. K., Christensen, H. E. M., Nielsen, T. K., and Willems, M., Analyst, 1982, 107, 1479. Paper J6l69 Received July 30th, 1986 Accepted January 21st, 1987

ISSN:0267-9477    

DOI:10.1039/JA9870200253

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 257 Graphite Furnace Atomic Absorption Spectrometric Screening Method for the Determination of Aluminium in Haemodialysis Concentrates Jan Rud Andersen Royal Danish School of Pharmacy, Department of Chemistry AD, 2 Universitetsparken, DK-2100 Copen hagen, Denmark A procedure is described for the determination of aluminium in haemodialysis concentrates, in which the samples are diluted 1 + 3 prior to analysis. The method is based on Zeeman-corrected atomic absorption spectrometry employing stablised temperature platform furnace conditions (STPF), and the method of standard additions is used for the quantitation. The limit of detection is 3.0 pg 1-1, and the precision is of the order of 10% (relative standard deviation) below 50 pg 1-1.The accuracy was assessed by recovery experiments and was found to be acceptable. The procedure is recommended for routine screenings. Keywords: Aluminium determination; Zeeman-effect background correction; graphite furnace; atomic absorption spectrometry; haemodialysis concentrate The interest in effects of aluminium on the human body have recently increased considerably. In 1976, aluminium was implicated as the cause of the clinical conditions dialysis encephalopathy and dialysis osteomalacia observed in renal failure patients who are undergoing long-term haemodialysis. 1 In these patients, aluminium accumulates in the body, causing the above-mentioned disorders in the brain and in the skeleton. One of the sources of the aluminium is the haemodialysis concentrates used in preparing the dialysis baths.The water used for dilution can be, and is in fact most often, effectively purified by reverse osmosis or ion exchange. Consequently, there is a need for a reliable analytical procedure for the determination of aluminium in haemo- dialysis concentrates in order to minimise the patients exposure to the element. The haemodialysis concentrates are, however, not easily analysed. Their aluminium contents should ideally be very low, i.e., in the low pg 1-1 range, hence the sensitivity of graphite furnace atomic absorption spectrometry (GFAAS) is required. However, the salt content of the concentrates is very high, of the order of 30%, and such high salt contents are often incompatible with GFAAS. Two ways of overcoming this problem have been devised, one relying on very high dilution of the concentrate2.3 and the other on addition of 40% nitric acid to the concentrate, three parts acid to five parts concentrate.4 Both approaches suffer from serious drawbacks.The former by the decreased sensitivity caused by the dilution and the latter by high blanks compromising the detection limit and by short graphite tube life. In this paper we present an improved procedure where only four-fold dilution is necess- ary. Zeeman-effect background correction is used, and the stablilised temperature platform furnace (STPF) concepts is employed. Sufficient sensitivity and precision is achieved, the blank level is low and the L'vov platform secures relatively long graphite tube life, i.e., normally more than 200 firings.Experimental Instrumentation A Perkin-Elmer Zeeman 5000 atomic absorption spec- trometer equipped with a Perkin-Elmer AS-40 autosampler * Presented at the Third Biennial National Atomic Spectroscopy Symposium (BNASS), Bristol, UK, 23rd-25th July, 1986. was used. The atomisation signals were displayed on a Perkin-Elmer R1000-A recorder and their integrated absorb- ances (A s) printed out on a Perkin-Elmer PRS-10 printer. Pyrolytically coated graphite tubes with platforms of solid pyrolytic graphite inserted were used throughout. The instrumental conditions are given in Table 1. Reagents The nitric acid was purified by sub-boiling distillation in an all-quartz apparatus (Hans Kiirner, Rosenheim, FRG). The Triton X-100 was of scintillation grade, purchased from E.Merck (Darmstadt, FRG). A certified 1 g 1-1 aluminium reference solution (Titrisol, E. Merck) was used, and aliquots of this were diluted with a solution containing nitric acid and Triton X-100 (see Sample Pre-treatment) to yield working standards. Milli-Q water, which is a type I ultrapure water prepared using a Milli-Q de-ionisation unit (Millipore, Bed- ford, MA, USA) was used throughout. Table 1. Instrumental conditions for the determination of aluminium in haemodialysis concentrates Wavelength . . . . 396.2nm Spectralbandpass . . 0.7nm Lampcurrent . . . . 20mA Samplevolume . . . . 2 0 ~ 1 Graphite furnace programme* Temper- Step ature/'C Ramp/s Hold/s Dry1 . . . . . . . . . . 120 30 20 Dry11 . . . . . . . . 300 50 20 Char .. . . . . . . . . 1600 40 30 Atomise . . . . . . . . 2500 O t 7 Clean . . . . . . . . . . 2700 1 2 Cool . . . . . . . . . . 20 3 5 * The internal argon gas flow was stopped and the Zeeman-effect background correction was on during atomisation. As temperatures may vary slightly between instruments for a given setting, they should be regarded as approximate values only. t Maximum power heating.258 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 0.1 18 0.132 0.066 0.070 0.018 -Time Fig. 1. Recorder traces showing the integrated absorbance values (A s) of the atomisation signals from a haemodialysis concentrate. Duplicate injections of sample, and sample with additions of 12.5 and 25 pg 1-1 of aluminium are shown. The sample contains 14.6 k 3.9 pg 1-1.Note that although the peak heights (absorbance) are rather different, the A s values are better reproduced (most pronounced for the smallest signals) Contamination Control Contamination is a serious problem when dealing with samples containing low concentrations of aluminium because of the ubiquitous nature of the element. Therefore, all utensils, e.g., sample cups, pipette tips and sample containers, were carefully decontaminated before use by a nitric acid wash as previously described.6 Sample Pre-treatment The haemodialysis concentrates are analysed directly after dilution 1 + 3 with a solution that is 2% in nitric acid and 0.1% with respect to Triton X-100. Three separate aliquots are made of each sample, containing no, one and two standard additions, respectively, and the aluminium content is calcu- lated from these by linear regression.Some concentrates contain hydrogen carbonate, and special attention is required when mixing these due to the liberation of carbon dioxide. Discussion and Results The direct injection of undiluted or slightly diluted haemo- dialysis concentrates into the graphite tube has previously been generally avoided. It was argued that the results obtained in this fashion were erratic4 and presumably imprecise. However, with the present furnace programme (Table 1), and by using the four-fold dilution described above, we are able to obtain reasonably precise and, to the best of our knowledge, accurate results. Fig. 1 shows a typical set of recorder traces and the absorbance seconds values of the atomisation signals from a haemodialysis concentrate treated as described.We found that a long, two-stage drying process is mandatory. If more rapid drying is attempted, erratic results are indeed obtained. This is probably due to the high salt content of the samples. (A typical haemodialysis concentrate contains: potassium chloride, 4.62; calcium chloride, 6.84; magnesium chloride, 4.73; sodium acetate, 132.9; and sodium chloride 178.6 g 1-1.) When faster heating rates are applied during drying, small “flakes” may virtually be observed to “jump off” the platform and deposit on the tube walls, and a cracking sound can be heard from the tube. The background absorbance at the 396.2-nm line is not very high; it seldom exceeds 0.1 A s and is easily corrected for by the Zeeman system.However, the chemical interference from the matrix is severe, but the utilisation of the combination of the L’vov platform, maximum power atomisation, gas stop etc., known as the STPF concept,S reduces this interference considerably. There remains, however, some signal de- pression, and a characteristic mass of ca. 20 pg per 0.0044 A s is found. For less complicated sample types, the characteristic mass of aluminium is 12-14 pg.6>7 The rationale for choosing the 396.2-nm line for Zeeman corrected AAS has been given previously,6 but it should be mentioned that the addition of a matrix modifier to the samples, in this instance is unnecessary. Matrix modification is incorporated in order to stabilise the analyte, thereby allowing higher char temperatures to be utilised, and it is an integral part of the STPF concept.However, Mg(NO& is the recommended matrix modifier for aluminium determinations, and Mg ions are a constituent of the haemodialysis concentrate (see above), and nitrate ions are added with the diluent containing nitric acid. Hence, the diluted samples have inherent matrix modification. There is no residue build-up on the platform with the present furnace programme. When the tube fails to perform satisfactorily, either because of physical breakdown or by lack of sensitivity, the platform still looks new and unpitched. More than 200 firings with good sensitivity are normally obtained, but it should be emphasised that the conditioning of the tube is critical.6 Analytical Performance From an analytical point of view three things are important in assessing the performance of a method.Firstly, the limit of detection, which may be defined as the blank value plus three times the standard deviation of the blank value.8 We find a value for the limit of detection of 3.0 yg 1-1 for a 20-yl injection diluted 1 + 3 as described above. This value was calculated from 20 determinations of the blank, which was pure diluent solution. Secondly, the precision, which for this method proves to be sufficient for screening purposes. For aluminium concentra- tions above 50 pg 1-1, relative standard deviations (RSDs) of better than 5-7% are achieved. It is our experience, however, that such high contents are very seldom present in haemo- dialysis concentrates. Below 50 pg 1-1 the RSDs increase with decreasing content.The analytical results for three different, randomly chosen preparations of concentrates may serve as examples of the precision below 20 pg 1-1 as follows: preparation I, 4.0 k 1.9 pg 1-1; preparation 11, 12.2 k 1.5 yg 1-1; preparation 111, 17.4 k 1.4 pg 1-1 ( n = 3). Finally, the accuracy, and for this procedure this is difficult to assess as no suitable reference material exists. However, the recoveries of 12.5 pg 1-1 of aluminium added to haemodialysis concentrates ranged from 93 to 108% with a mean value of 100% ( n = 5) indicating an acceptable accuracy. In conclusion, we have presented a screening method for the determination of aluminium in haemodialysis concentrates that is fast, sensitive, sufficiently precise and accurate, and due to its simplicity not prone to contamination. The technical assistance of Susanne Reimert and the help of Hanne Blzehr in providing the samples, is gratefully acknowl- edged. References 1. Alfrey, A. C., LeGendre, G. R., and Kaehny, W. D., N. Engl. J . Med., 1976, 294, 184. 2. Halls, D. J . , and Fell, G. S., Analyst, 1985, 110, 243.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 259 3. Starkey, B. J., Taylor, A . P., and Walker, A. W., in Taylor, A . , Editor, “Aluminium and other Trace Elements in Renal Disease,” Baillikre Tindall, London, 1986, p. 177. 4. Allain, P., Mauras, Y., and Der Khatchadourian, F., Anal. Chern., 1984,56, 1196. 5 . Slavin, W., Manning, D. C., and Carnrick, G. R., At. Spectrosc., 1981, 2, 137. 6. Andersen, J. R., and Reimert, S., Analyst, 1986, 111, 657. 7. 8. Slavin, W., and Carnrick, G. R., Spectrochim. Acta, Part B , 1984,39,271. American Chemical Society Committee on Environmental Improvement, Anal. Chern., 1980,52,2242. Paper J6l61 Received July 24th, 1986 Accepted August 20th, I986

ISSN:0267-9477    

DOI:10.1039/JA9870200257

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 261 Direct Determination of Cobalt in Acetic Acid Extracts of Soils by Graphite Furnace Atomic Absorption Spectrometry Margaret C. Mitchell, Michael L. Berrow and Charles A. Shand The Macaulay Institute for Soil Research, Craigiebuckler, Aberdeen AB9 ZQJ, UK A method is described for the direct determination of cobalt in 0.43 M acetic acid extracts of soils by atomic absorption spectrometry using a graphite furnace. The sample is nebulised and introduced to the furance as an aerosol. With this technique and the use of optimised operating conditions, the influence of major concomitants is negligible and the determination is virtually free from interference of extraneous elements. A comparison of analyses by this method with those obtained by conventional flame atomic absorption spectrometry following pre-concentration of the extract and with direct current arc atomic emission spectrometry following chemical concentration shows good agreement. The method has been applied to the determination of cobalt in a large number of Scottish topsoils derived from a wide variety of parent materials.Keywords: Cobalt determination; graphite furnace; atomic absorption spectrometry; soils; acetic acid extracts As a trace element, cobalt is essential for animal health and growth and a lack of the element in the ruminant diet causes a wasting disease, “pining.” In materials such as soil extracts, plants, biological tissues and sea water, cobalt frequently is present only at low, pg kg-1, levels and for this reason graphite furnace atomic absorption spectrometry (GFAAS) has been combined with a pre-concentration technique for its determination. 1 For the assessment of soil cobalt status and its availability to herbage plants, the determination of 0.43 M acetic acid (HOAc) extractable cobalt is the method used in Scotland for farm advisory purposes.The deficiency criterion, in freely drained topsoils, is an HOAc-extractable soil content of <0.75 mg kg-1.2 The determination of cobalt in HOAc extracts by flame atomic absorption spectrometry (FAAS), using an air - butane flame, suffers from interference by other elements present in the extract, e.g., calcium and aluminium.3 In order to obtain a concentration of cobalt in the extract that can be accurately determined (minimum 0.1 mg 1-I), the 400-ml HOAc filtrate must be evaporated, oxidised and taken up in a final volume of 10 ml, thereby increasing the cobalt concentra- tion 40-fold, but at the same time concentrating the interfering elements. Interference by these elements, however, can be overcome by appropriate matching of the standards.With the alternative procedure using chemical concentration followed by d.c. arc atomic emission spectrometry, the cobalt is collected as a precipitate and determined in a powder form, in the same matrix as the powder standards used for ~alibration.~ Both methods are labour intensive. This paper describes the direct determination of cobalt in acetic acid extracts of soils by graphite furnace atomic absorption spectrometry.Experimental Apparatus An Instrumentation Laboratory (UK) Ltd. (IL) 751 atomic absorption spectrometer, with a graphite furnace atomiser (IL555) interfaced to an autosampler (IL254) (furnace injec- tion system) and a Linseis pen recorder were used. Pyrolytic graphite coated tubes were employed throughout. A FAAS method with an air - acetylene flame was also used and background correction with a deuterium continuum source was employed for both of these methods. The chemical *Presented at the Third Biennial National Atomic Spectroscopy Symposium (BNASS), Bristol, UK, 23rd-25th July, 1986. concentration procedure, with analysis of the concentrate by d.c. arc atomic emission spectrometry, followed the method of Mitchell and Scott.4 Reagents A 1000 mg 1-1 Co stock solution was prepared by dissolving cobalt metal (Johnson Matthey and Co.Ltd.) in the minimum amount of re-distilled analytical-reagent grade nitric acid. This stock solution was used to prepare a range of standards containing 2-50 pg 1-1 Co in 0.43 M HOAc (re-distilled analytical-reagent grade). Sample Preparation A 20-g sample of air-dried soil (<2 mm) was shaken with 400 ml of 0.43 M HOAc on an end-over-end shaker for 16 h. The mixture was filtered (Whatman No. 542, 18.5 cm) without washing and 10 ml of the filtrate were taken for analysis by GFAAS. The remainder was analysed by FAAS following the method of Ure and Mitchell3 with an air - C2H2 flame. Extracts prepared in a similar way were used for analysis by the chemical concentration procedure followed by d.c.arc atomic emission spectrometry. 0 10 20 30 40 50 Cobalt con cent rat ion/pg I Fig. 1. Calibration graph for determination of cobalt Table 1. Furnace programme Stage Parameter Drying Ashing Atomisation TemperaturePC . . . . . . 175 650 950 2500 5 25 25 5 Time/s . . . . . . . .262 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, MARCH 1987, VOL. 2 Table 2. Extractable major elements* in soils. Results expressed as mg kg-1 in air-dry soil Ca A1 Fe P Mg K Group? Range Mean Range Mean Range Mean Range Mean Range Mean Range Mean A . . <10&2200 1200 95-1200 480 68-720 220 8.3-160 39 22-130 53 24-140 89 A' . . 1400-6600 3000 420-1700 970 350-3200 1200 7.4-94 33 NDS NDS 68-150 100 * Ca, Mg, P and K in soil groups A and A' were extracted with 0.43 M acetic acid and A1 and Fe with normal ammonium acetate solution.t Group A consists of topsoils and group A' peaty soils. $ ND = not determined. 0 0.2 0.4 0.6 0.8 1 .o Graphite fu rnace/mg kg - 1 Fig. 2. Comparison between GFAAS and FAAS for the determina- tion of cobalt in 59 topsoils Table 3. Comparison of results for cobalt determinations in ten Scottish topsoils obtained by three different methods. Results expressed as mg kg-1 in air-dry soil Chemical concentration - d.c. Soil No. GFAAS* FAAS* arc? 1 2 3 4 5 6 7 8 9 10 0.18 0.62 0.74 0.74 1.88 0.90 0.33 0.35 0.92 0.55 0.205 0.23 0.60 0.74 0.73 1.983 1.88 0.91 0.43 0.33 0.92 0.415 0.43 0.29,0.29 0.69,0.69 0.82,0.79 0.85,0.72 2.43,2.20 0.85,0.85 0.38,0.38 0.37,0.41 0.93,0.87 0.55,0.56 * Aliquot taken from same extraction of 20 g in 400 ml of HOAc.t Separate aliquot, from extraction of 20 g in 800 ml of HOAc. 5 Results obtained by the standard additions procedure. Determinations were carried out in duplicate. Operating Conditions Sample introduction with this system was by time-controlled nebulisation and deposition in the furnace tube. Various deposition times on the IL254 were examined and it was found that a 5-s deposition time (for ca. 5 pl) of Co standard solutions in the range 0-35 pg 1-1 gave a linear calibration graph with slight curvature towards 50 pg 1-1, enabling the majority of the soil extracts to be analysed without dilution (Fig. 1). A wavelength of 240.7 nm and a band width of 0.30 nm were used for GFAAS throughout. The integrated (1 s) peak-height measurements were recorded.Interferences It has been suggested by Conley et al.5 that sample introduc- tion by nebulisation into the furnace reduces interference effects in GFAAS. No interference was in fact encountered, using the furnace programme given in Table 1, for the following elements: calcium, 70-150; aluminium, 0.50-100; iron, 0.50-150; magnesium, 2.4-9.6; potassium, 0.50-100; phosphorus, 0.50-100; silicon, 100 mg 1-1; and mixtures containing all the above elements, except for silicon, at their highest and lowest levels, on standards containing 0 , 5 and 20 pg 1-1 of cobalt. The above concentrations of interferents cover the maxima that might be expected in an acetic acid extract of a topsoil (Table 2). Results and Discussion A GFAAS method was chosen for the determination of cobalt in acetic acid extracts of soils because of the ease and speed with which the results can be obtained compared with FAAS or chemical concentration.The comparison between the GFAAS and FAAS results using 59 topsoils ( r = 0.9711) is illustrated (Fig. 2). The results obtained from the analyses of ten Scottish topsoils by GFAAS, FAAS and chemical concentration with d.c. arc atomic emission spectrometry, illustrated in Table 3, show good agreement. In the chemical concentration - d.c. arc atomic emission procedure, the use of a 20 g to 800 ml soil to solution ratio rather than 20 g to 400 ml is historical and makes little difference to the extractable cobalt values. In a comparison of the results for 25 soils, the 1 + 20 extraction ratio produced values that were on average 3.2% lower than the 1 + 40 values. The range of 0.43 M acetic acid extractable cobalt in a number of Scottish topsoils, derived from widely differing parent materials, was 0.08- 1.98 mg kg-1, and storage of the extracts for up to 3 months at 4°C had no effect on the results.The precision of the analytical procedure is good; replicate analyses of a solution containing 5 pg 1-1 provides an RSD of 6.1%. The detection limit is 0.02 mg kg-1 of extractable cobalt in the soil, which is well below the advisory criterion for deficient soils of 0.75 mg kg-1 of cobalt. Conclusion Cobalt can be determined directly in 0.43 M acetic acid extracts of soils by graphite furnace atomic absorption spectrometry. The results obtained are in good agreement with those obtained by two other spectrochemical methods. The pro- posed method is rapid and saves considerable analytical time, laboratory space and sample preparation time compared with other procedures. References 1. 2. Iu, K. L., Pulford, I. D., and Duncan, H. J., Anal. Chim. Actu, 1979, 106, 319. Bulletin No. 1, Macaulay Institute for Soil Research - Scottish Agricultural Colleges, MISR-COSAC Liason Group, The Macaulay Institute for Soil Research, Aberdeen, 1985, 13 pp. 3. Ure, A. M., and Mitchell, R. L., Spectrochim. Actu, Part B, 1967, 23, 79. 4. Mitchell, R. L., and Scott, R. O., Spectrochim. Actu, 1948, 3 , 367. 5. Conley, M. K., Sotera, J. J., and Kahn, H. L., AID Report No. 149, Instrumentation Laboratory Inc., Analytical Instrument Division, Wilmington, MA, USA, 1981. Paper J6l75 Received August 15th, 1986 Accepted September 5th, 1986

ISSN:0267-9477    

DOI:10.1039/JA9870200261

出版商:RSC    

年代:1987

数据来源: RSC