摘要:

Analyst February 1983 VoZ. 108 pp 261-264 261 Determination of Lead in Soil by Graphite Furnace Atomic-absorption Spectrometry with the Direct Introduction of Slurries Kenneth W. Jackson and Alan P. Newman Department of Chemistry Shefield City Polytechnic Pond Street Shefield S1 1 WB Department of Geography and Environmental Studies Shefield City Polytechnic Wentworth Woodhouse, Rotherham S62 7TJ A simple procedure is described whereby 5-20mg of a powdered soil are mixed with 25 ml of water and while stirring 20-pl aliquots of the slurry are pipetted into a graphite furnace electrothermal atomiser. Calibration with aqueous standards gives the total lead concentration of the soil. This is much simpler than the complex hydrofluoric acid digestion procedures normally required to obtain complete recoveries from soil.The over-all precision of the analysis (8% relation standard deviation) is adequate for this application as is the sensitivity. A characteristic concentration ( A = 0.0044) of 0.5 pg 1-1 is equivalent to 0.62 pg g-1 soil in a typical slurry. Keywords Lead determination ; soil analysis ; solid-sample introduction ; electrothermal atomisation ; atomic-absorption spectrometry Commonly used procedures for determining the “total” concentration of a trace metal in soil include digestion with nitric a ~ i d l - ~ or with mixtures such as nitric - hydrochl~ric~-~ and nitric - perchloric acid^,^^^ followed by flame atomic-absorption spectrometric analysis (FAAS). These procedures can be rapid but lead is incompletely extracted unless some means of destroying the silicate matrix is employed.This was demonstrated3 when several digestion procedures were compared and a nitric - hydrofluoric acid treatment gave higher extraction efficiency than hydrochloric nitric or a nitric - hydrochloric acid mixture. A more complex procedure was employed by Stupar and Ajlec6 who digested the soil with nitric - perchloric - hydrofluoric acids and then fused the residues with sodium carbonate followed by dissolution in 1 + 1 hydrochloric acid. This procedure was hazardous and time consuming therefore they developed an alternative FAAS method using the direct nebulisation of soil slurries into the flame. This was much simpler and faster than the digestion procedure but variable accuracy suggested that complete recoveries were still not always achieved.Problems with slurry nebulisation include poor sample transport efficiency that prevents larger particles from reaching the flame. This can introduce a sampling problem because the analyte metal may not be distributed uniformly with particle size. Also many of the particles that reach the flame are too large to be atomised. Willis7 reported that when suspensions of geological samples were nebulised into a flame only particles less than 12 pm i.d. contributed significantly to the absorption signal. Fuller et aZ.8 were in close agreement when they found that during the analysis of titanium(1V) oxide and silicate rocks slurry particles need to be less than 10 pm i.d. if such losses are to be minimised. They also found difficulty in selecting appropriate calibration standards because the signal response was only 30-50% of the aqueous standard response.These workers reported that direct pipetting of their slurries into an electrothermal atomiser for atomic-absorption spectrometric analysis was much more successful as the particle size requirement was less than 25pm i.d. and the aqueous standards were satisfactory . A promising approach for achieving complete trace-metal recoveries from soil by atomic-absorption spectrometry appears to be slurry injection into an electrothermal atomiser (slurry - ETA - AAS). In this paper we describe such a procedure for determining lead in soil but the method of Fuller et aZ. was further simplified by omitting the thixotropic thickening agent they added to stabilise their slurries.Instead slurry homogeneity was maintained by continuous magnetic stirring throughout the pipetting operation 262 JACKSON AND NEWMAN DETERMINATION OF PB IN SOIL BY Analyst Vot. 108 Experimental Apparatus Soils were ground in a porcelain ball mill with magnesium silicate balls (Model 1 lR Pascal1 Engineering Co. Ltd.). Further milling for the slurry - ETA - AAS procedure used a miniature ball mill (Rotomill RM-100 Beckmann-RIIC Ltd.). For ETA - AAS an Instrumentation Laboratory Model 157 atomic-absorption spectro-meter fitted with a Model 555 CTF electrothermal atomiser was used. This atomiser incorporated a square cross-section graphite furnace with a removable graphite platform. The spectrometer was equipped with deuterium lamp background correction and the following operating conditions were used for lead determinations wavelength 283.3 nm ; slit width 1.0 nm; and hollow-cathode lamp current 5 mA.Absorption peak areas were measured over a 5-s integration time. Analyses by FAAS were performed on a Varian AA 1275 spectrometer with deuterium lamp background correction. Operating conditions for lead were flame air - acetylene; wavelength 283.3 nm; slit width 1.0 nm; and hollow-cathode lamp current 3 mA. Reagents and Standard Solutions Analytical-reagent grade nitric acid was used for the wet digestion of soils. A stock standard lead solution (1 000 mg 1-l) was prepared by dissolving 1.598 5 g of lead nitrate (analytical-reagent grade) in 5 ml of concentrated nitric acid diluted to 1 1 with distilled de-ionised water.Calibration standards were obtained by appropriate dilution of this stock standard solution. Distilled de-ionised water was used for preparing slurries. Procedure Soil $re-treatment Samples collected from coal measure strata areas of North Derbyshire and South Yorkshire, were of the brown-earth type except for sample 1 which was a heavy gley soil. Approxi-mately 2 kg were dried for 24 h at 105 "C and then ground by hand using a porcelain mortar and pestle to pass a 2-mm nylon sieve. Further grinding was for 24-48 h in the porcelain ball mill. The mill was cleaned between samples by grinding overnight with a small amount of dry silver sand and then washing thoroughly with distilled de-ionised water. For slurry - ETA - AAS however further grinding was required.Approximately 2 g were withdrawn and ground for 10 min in the Rotomill at a speed setting of 2. Samples prepared in this way were used directly for acid digestion prior to FAAS. Soil analysis de-ionised water were added. the maximum speed which did not permit air entrainment into the suspension. was being stirred. platform which was then inserted into the furnace without further treatment. programme outlined in Table I was then initiated. triplicate. The finely powdered soil was weighed (5-20 nig) into a 50-nil beaker and 25 r i d of distilled, Slurries were obtained by stirring magnetically for 3 min at Aliquots (20 p1) of the slurry were withdrawn by means of a micropipette while the suspension For each sample triplicate aliquots were pipetted directly on to the graphite The heating For calibration aqueous lead standards in the range 0-0.1 rng 1-1 were aliquoted (20 pl) in TABLE I FURNACE OPERATING PROGRAMME Stage r A \ Dry Ash 4toinise Time/s .. . . . . . . 5 15 35 5 0 10 Indicated temperature/"C . . 70 125 450 450 2000 200 February 1983 GRAPHITE FURNACE AAS AND DIRECT INTRODUCTION OF SLURRIES Results and Discussion 263 Slurry Homogeneity It was shown by Fuller et aZ.8 that errors in pipetting slurries into electrothermal atomisers were acceptable provided that the samples were ground to less than 25pm particle size. After grinding a soil in the Rotomill its particle size distribution was determined by placing a drop of the slurry on a microscope slide drying and examining under oil immersion at 1000 x magnification.Approximately 90% of the particles were <11 pm i.d. the remainder being <30 pm i.d. Fuller et al. used a thixotropic thickening agent to stabilise their slurries and in order to determine whether this would be necessary during our work 10 replicate 20-4 aliquots were pipetted from a stirred slurry and weighed. A relative standard deviation of 2.0% was obtained from the weighings. This is only slightly less precise than 10 replicate weighings of pure water aliquots (1.1% relative standard deviation) and so the simplified slurry technique allows a sufficiently representative sample to be transferred to the graphite platform. Determination of Lead in Soil Lead calibration graphs obtained at 283.3nm by pipetting 20-4 aliquots of aqueous standards into the electrothermal atomiser were linear to at least 0.1 mg 1-1.The charac-teristic concentration ( A = 0.0044) was 0.5pgl-1 or 1 x 10-l1 g which is equivalent to 0.63 pg g-l of soil in a typical slurry. Lead results for five soil samples analysed by the slurry - ETA - AAS procedure are shown in Table 11. For comparative purposes the ground samples were also analysed by a nitric acid digestion - FAAS procedure used routinely in our laboratory. This involved thoroughly digesting with hot 1 + 1 nitric acid until all evolution of the brown fumes had ceased and using aqueous calibration standards containing dilute nitric acid. It can be seen from Table I1 that the latter procedure gave results which were consistently lower as might be expected if nitric acid digestion gave incomplete lead extraction.This was verified by transfering the residues from the acid digestion to a filter-paper washing with distilled de-ionised water, drying and taking the residue through the slurry - ETA - AAS procedure. The final column of Table I1 shows the nitric acid digestion - FAAS results corrected in this way for the unliberated lead. Agreement with the slurry - ETA - AAS results is now much better and it is confirmed that nitric acid digestion - FAAS gives incomplete extraction of lead from soil. TABLE I1 MEASURED LEAD CONCENTRATIONS Measured lead concentration/mg kg-' of soil I Nitric acid FAAS corrected' Sample No. Slurry - ETA - AAS digestion - FAAS for residual lead 1 50 2 70 3 81 4 86 5 156 47 49 56 65 77 80 83 85 144 147 As a further test of the slurry- ETA - AAS technique a Certified Reference Material (Pond Sediment NIES Tsukuba Ibaraki Japan) was ground (0.5g) for 10min in the Rotomill and then analysed using aqueous standards.A result of 119 mg kg-l though slightly higher than the certificate value (105 & 6 mg kg-l) showed reasonable agreement. The precision of the slurry - ETA - AAS procedure was evaluated by preparing six slurries from separate weighings of soil sample 3. When each slurry was taken through the analysis procedure a relative standard deviation of 8% was obtained at a lead in soil concentration of 81 mg kg-l 264 JACKSON AND NEWMAN Conclusions When analysing soil great care must be taken to obtain a reasonably large sample which is typical of the land area to be studied.I t is then essential to reduce this to a small particle size and thoroughly mix before taking a sub-sample for analysis. This sub-sample should be as large as possible in order to minimise sampling errors. Therefore there is a greater possibility of errors being introduced in this slurry procedure which takes 5-20 mg of soil, compared with acid digestion procedures using typically 1 g of soil. Nevertheless the over-all precision of 8% relative standard deviation indicates an acceptable level of sub-sampling precision in this instance and the slurry technique is to be preferred to the alterna-tive solid-sampling - ETA - AAS procedure which involves weighing individual sub-milligram aliquots of solid material into the f ~ r n a c e .~ Besides being more error prone the latter technique is much slower if for example three replicate aliquots are taken for analysis. For the soils examined the slurry technique required only 10 min of additional grinding a single weighing and 3 min of mixing prior to pipetting aliquots on to the graphite platform. Thus it is even faster than the simple nitric acid digestion - FAAS procedure used in this work for comparative results. Certainly it is much faster than the complex hydrofluoric acid digestion procedures otherwise needed if total lead is to be determined. An additional advantage of the slurry - ETA - AAS procedure is the minimum risk of contamination because only water is used as a diluent and aqueous calibration standards are used.The soils studied are in our experience typical of North Derbyshire and South Yorkshire; the sensitivity and precision of the slurry - ETA - AAS procedure are adequate at these lead concentrations. The soils were mostly shale and further work will be undertaken to confirm that the technique is applicable to other soil types. I t is expected that the Rotomill grinding time will have to be extended for sandy soils. We are grateful to M. Clench and I. W. Eastwood for technical assistance and Instru-mentation Laboratory (UK) Ltd. for their interest and help. 1. 2. 3. 4. 5. 6. 7. 8. 9. References Seeley J. L. Dick R. H. Arvik J. H. Zimdahl R. L. and Skogerboe K. K. Appl. Spectvosc. 1972, Balraadjsing B. D. Commun. Soil Sci. Plant Anal. 1974 5 25. Harrison R. M. and Laxen D. P. H. Water Air Soil Pollut. 1977 8 387. Ritter C. W. Bergman S. C. Cothern C. It. and Zamierowski E. E. A t . Absovpt. Newsl. 1978, Markunas L. D. Barry E. F. Giuffre G. I?. and Litman K. J . Environ. Sci. Hpalth Pavt A 1979, Stupar J. and Ajlec R. Analyst 1982 107 144. Willis J. B. Anal. Chem. 1975 47 1752. Fuller C. W. Hutton It. C. and Preston B. Analyst 1981 106 913. Chakrabarti C. L. Wan C. C. and Li W. C. Spectrochim. Acta Part B 1980 35 93. 26 456. 17 70. 14 501. Received August 3rd 1982 Accepted September 17th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800261

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst February 1983 Vol. 108 pp. 265-276 265 Re-usable Sampling Tubes for Monitoring Airborne Mercury Vapour Concentrations Sample Collection and Analysis by Cold Vapour Atomic-absorption Spectrometry Ma rk Taylor Health and Sajety Executive London and Home Counties North Field Consultant Group 14 Cavdi=ff Road, Luton LU1 1PP A re-usable sampling tube with silvered alumina as the collection medium is used to monitor concentration levels of airborne mercury vapour. The device is ideally suited for personal sampling over long or short periods and its mercury blank is negligible. Analysis is by cold vapour atomic-absorption spectrometry using thermal desorption to release the mercury from the sanipling tube for subsequent measurement and a novel and convenient nieans has been devised for altering the measurement range of the analytical system.A detection limit of 3 ng equivalent to 0.5 pg III-~ for a 6-1 air sample has been achieved but this could easily be improved further by making small modifications to the analytical apparatus. The sampling and analytical procedures have been validated by comparing results obtained for samples from mercury standard atmospheres and for factory samples with those obtained by established procedures. Keywords A irborne mercury vapour ; silvered alumina ; re-usable sampling tube ; cold vapour atomic-absorption spectrometry A re-usable sampling tube has been devised which is used to monitor airborne mercury vapour concentrations. The device based on silvered alumina as the mercury collection medium is very easy to use and the analytical system which uses thermal desorption followed by cold vapour atomic-absorption spectrometry can be made highly sensitive.This paper describes the evaluation of the sampling tube and associated analytical system. The toxic effects of mercury have been known for a very long time and the Health and Safety Executive has long been concerned with measuring airborne concentration levels of mercury vapour in the workplace. Recently a personal sampling system based on a solid adsorbent hopcalite was adopted. Hopacalite is an excellent collection medium for mercury vapourl *2 and hopcalite-filled sampling tubes are commercially available but there are several drawbacks to its use. The tubes are manufactured in the USA are expensive and can only be used once.They are available in two sizes and the larger size which was found to be the most convenient requires a relatively heavy personal sampling pump to be used to achieve the optimum sampling air flow-rate of 1-3 1 min-l. Also hopcalite contains a small and variable contaminant level of mercury (about 0.4 pg g-1). The background level on a large tube is about 0.2 pg and if the concentration of mercury vapour being measured is around one tenth of the current threshold limit value (TLV)3 of 50 pg m-3 at least 100 1 of air should be sampled to be certain of obtaining a sample measurement of twice the back-ground level. However the background level has been found to vary markedly from tube to tube even within the same batch and consequently a considerably larger sample volume than 100 1 must be taken in this situation if the accuracy of the analytical measurement is to be guaranteed.Sample preparation steps prior to analysis are time consuming and the analytical step itself is relatively slow and tedious frequently involving sample dilution ; the hopcalite is dissolved in an acid mixture and made up to volume in a calibrated flask followed by conventional cold vapour atomic-absorption spectroscopic determination of mercury using tin(I1) chloride as the reducing agent. Mainly because of the drawbacks to using hopcalite sampling techniques based on amalgamation of mercury with noble metals followed by direct thermal desorption of samples into the analytical system were investigated. Several of these methods with silver4p5 or Crown Copyright 266 TAYLOR RE-USABLE SAMPLING TUBES FOR AIRBORNE HG Analyst Vol.108 golds9' as the collection medium have been described in the literature. Silver-wool even when new requires rigorous cleaning procedures prior to use4 and was found to be difficult and tedious to pack into small-bore sample tubes. A silvered substrate seemed more promising but the analytical apparatus described by Trujillo and Campbell8 for dealing with sample tubes containing silvered Chromosorb P seemed complex. Being designed to cope with organic mercury vapour samples as well as metallic mercury vapour their analytical system used an intricate two-stage thermal desorption unit made of Pyrex or quartz glass, which incorporated a tube loading mechanism cooling holes and jacket a copper(I1) oxide section to oxidise any organic vapours and holes for the electrical connecting wires to pass through.The sample was first desorbed from the sample tube and then recollected on gold granules mixed with sand in the second desorption section before the final desorption for sample measurement. Their apparatus also included two cold vapour cells of different path lengths connected in series each with its own Coleman Mercury Analyzer instrument, so that a sample that was above the measurement range of the first would be measurable using the second. The sampling and analytical system described here is much simpler and has several advantages over other methods. Sampling tubes are easy to prepare and can be re-used many times with the same packing a considerable economy.Silvered alumina has good collection efficiency for mercury vapour and the tubes can be adapted for sampling over periods of one or several hours by using different column lengths of packing. The easily cleaned tubes have no mercury blank no sample dissolution step is needed and they can be used with a small light-weight sampling pump. The analytical apparatus based on cold vapour atomic-absorption spectrometry is relatively simple and incorporates a novel means of changing the measurement range; ranges of 0-2.5 and 0-1Opg of mercury have been used so far but other ranges can easily be set up. A drawback of the system is that it is necessary to know the approximate airborne mercury concentrations under examination to set up the appropriate range.A battery powered mercury meter may be used at the sampling site to give a rough indication of levels. The sampling and analytical method has been validated by sampling a series of standard mercury in air concentrations with both silvered alumina and hopcalite tubes and com-paring the results. Further validation tests have involved sampling in a factory rnanu-facturing zinc amalgam batteries. Experimental Apparatus Silvering procedure a magnetic stirrer and glass-wool were used. A 1000-ml beaker 150-ml beakers a 1000-nil conical flask with stopper Yasteur pipettes, Sampling system The apparatus used was as follows sampling tubes of 10 cm over-all length 8 mm o.d., 5 mm i.d. with an S13 ball-joint a t each end and made to order from clear fused quartz glassware ; 22 SWG nichrome heating wire (Post Radio Supplies London) ; quartz-wool ; connecting tubing which was translucent silicone-rubber tubing of 9.0 mm bore 2.5 mm wall [Silescol tubing manufactured by Esco (Rubber) Ltd.Teddington Middlesex] ; plastic tubing adaptors for connecting different diameters of tubing; Parafilm sealing tissue (Gallen-kamp) required for capping the sampling tubes; and battery operated sampling pump, capable of maintaining a sampling air flow-rate of 100 ml min-1 over the sampling period, e.g. Rotheroe Mitchell C500. Analytical system The basic analytical system is shown diagrammatically in Fig. 1. An Instrument at ion Lab oratory 15 1 at oniic-absorption spec tropho tome t er fitted with Activion mercury hollow-cathode lamp and a standard 10 cm long quartz absorption cell for cold vapour mercury determination in a standard holder assembly was used with the follow-ing conditions wavelength 253.7 nm lamp current 3 111-4 slit width 320 pm 0.25 s aut February 1983 SAMPLE COLLECTION AND ANALYSIS BY COLD VAPOUR AAS Air in Quartz absorption cell 1=253.7 nm extraction Thermal desorption , power SUPPlY Variable 1 I regulation I clip 2nd sample tube for recollecting &J2tr.rnyEformer analysed sample (optional) hood 267 Fig.1. The Analytical System. integration mode and deuterium lamp background correction. A Servoscribe RE 541.20 pen recorder operated at 30 mm min-l chart speed was used to monitor the peak shape. Thermal desorption apparatus was used which consisted of a mains step-down transformer coupled to a variable auto transformer set to supply 190 W.The variable auto transformer was fitted with an appropriately rated anti-surge fuse at the mains input. Cup joint tubes of 10 cm over-all length 8 mm o.d. 5 mm i.d. with an S13 cup-joint at one end only were made to order from clear fused quartz glassware. Quickfit JC13 clips were used and a jointed receiver with vent and 7/16 Quickfit socket of over-all height 119 mm nominal bore 8 mm and fitted with matching cone with stem was made to order. Connecting tubing which was Silescol tubing of 5.0 mm bore 1.6 mm wall and 3.2 mm bore 1.0 mm wall and Y-shaped plastic tubing connectors were used. Air-flow restrictors made of rigid plastic of glass tubing 2.5 cm long of suitable external diameter to fit snugly into the connecting tubing at the appropriate position and giving a selection of internal diameters from 1.0 mm upwards would be ideai but here the actual restrictors used were improvised from items that were readily at hand.For the 0-2.5pg mercury range a 1.8-cm length of Thuscan Wall Plug (to take screw size No. 8 drill size No. 10) was used. This is made of rigid red plastic and has the cross-sectional shape as shown in Fig. 2. When it is fitted into 5 mm bore Silescol tubing air can pass along the fluted outside as well as through the centre hole (Woolworths or DIY shops. Suppliers are Thunder Screw Anchors Ltd. Horsham Sussex). For the 0-10 pg mercury range a Quickfit 5/13 cone with a 5-cm stem of approximately 2.5 mm bore made of Pyrex glass was used.V Fig. 2. Cross-section of Thuscan Wall Plug used as flow restrictor for 0-2.5 pg Hg range 268 TAYLOR RE-USABLE SAMPLING TUBES FOR AIRBORNE HG Analyst VoZ. 108 This flow restrictor was connected into the system by attaching 3.2 mm bore tubing at each end which was in turn connected to 5.0 mm bore tubing by means of straight plastic tubing adaptors of the appropriate size. The Quickfit cone was not an intentional design feature of the flow restrictor. An air flow-meter 0.1-3.4 1 min-l air (Jencons Meterate) a screw clip for regulating air flow and Diaphragm pump Charles Austin Dymax 2 output 3.251 min-l of free air and micropipettes including a 10-pl fixed volume micropipette used for dosing standard solutions on to sample tubes were used.Apparatus used in standard atmosphere tests The dynamic standard atmosphere system was made of borosilicate glass in which mercury-saturated air was appropriately diluted with further air to give the required mercury vapour concentrations. The system used two air dilution stages and was based on ref. 9 without the motor drive and syringe injection to the first dilution chamber Mercury vapour passed through tubing directly into this chamber from the source a glass vessel containing mercury-impregnated Kieselguhr which was wrapped with heating tape so that the rate of vaporisa-tion could be conveniently altered. The generated mercury standard atmosphere was passed into a cylindrical Pyrex glass vesse1,lO 40 cm long and 15 cm i.d. closed at both ends and fitted with an entry and an exit port for the air.The vessel had several screw-threaded sampling ports spaced along its length into which sampling devices could be inserted and fixed. One of these positions was connected by tubing to the inlet port of a Data Acquisition mercury meter. The other positions were reserved for the sampling tubes. The Data Acquisition mercury meter was a Model DA 1500 (Data Acquisition Ltd. Stockport Cheshire). The manifold sampling system used consisted of a dry vacuum pump to provide simul-taneous sampling air flows for up to six sampling tubes. The air inlet to the pump was split by suitable pipework to supply vacuum to six sampling nozzles. Each nozzle had its own controlling Hoke needle valve and rotameter so that each of the six air flows could be altered and accurately measured independently of the others.Two sizes of hopcalite tubes were used small tubes Cat. No. 226-17-1 and large tubes, Cat. No. 226-17-3. They are manufactured by SKC Inc. PA USA and supplied by MDA Scientific (UK) Ltd. Wimborne Dorset. Reagents All reagents used were of AnalaR grade unless otherwise stated. Distilled water was used. Silvering procedwe Silver nitrate. Ammonia soldion 0.88 g ml-l, Potassium hydroxide. n-Glucose (anhydrous). Aluminium oxide 4-8 mesh. Fisons laboratory reagent. Analytical system acid as supplied for atomic-absorption spectrophotometry by Hopkin and Williams. referred to in the text as mercury(I1) solution. Mercury standard solution. 1000 p.p.m. mjV mercury in approximately 0.1 ? perchloric It is Preparation of Sampling Tubes were then sieved.The silvering procedure is essentially the same as that used in ref. 11. The aluminium oxide granules were reduced further in size with a pestle and mortar and The fraction between 710 pm and 1.0 mm was retained for silvering. Caution-A face shield and rubber gloves should be worn when carrying o u t this procedure. Dissolve 20 g of silver nitrate in 300 ml of water in a 1000-ml beaker (solution A) and stir using a magnetic stirrer. Slowly add ammonia solution until a dark brown precipitate of silver(1) oxide forms. While stirring add more ammonia solution dropwise until th FebrGary 1983 SAMPLE COLLECTION AND ANALYSIS BY COLD VAPOUR AAS 269 solution just clears disregarding any small specks that may form.Dissolve 2 g of silver nitrate in 30 ml of water (solution B) and add dropwise to the mixture in the 1000-ml beaker, until the solution turns straw yellow. Slowly add 100ml of potassium hydroxide solution (14 g of potassium hydroxide in 100 ml of water) while constantly stirring. Add ammonia solution dropwise until the solution in the 1000-ml beaker just clears. Add more solution B dropwise until a thin straw-yellow precipitate forms again disregarding any small specks. Filter the solution (solution C) through glass-wool. Dissolve 7.8g of dextrose in 120ml of water in a 1000 ml conical flask and add 73 g of the alumina (710 pm-1.0 mm) to the dextrose solution. Swirl the contents of the conical flask to wet the alumina surfaces and then decant the excess of dextrose.Slowly add solution C to the alumina stopper the flask and shake it vigorously so that all the alumina surfaces come in contact with the silver solution. Rinse and decant the silvered alumina three times with distilled water and then dry it in an oven at 60 "C in an atmosphere free of sulphides. Repeat the silvering procedure on the same batch of alumina. Tubes for sampling periods of up to 1 h contain a 2-cm column of silvered alumina held in place by small plugs of quartz-wool; for longer sampling periods i.e. up to 6 h a 6-cm column of silvered alumina is used. Each tube has 35 turns of 22 SWG nichrome heating wire around it as the heating element for the thermal desorption stage. Tubes have a negligible mercury blank and can be re-used many times before the silvered alumina needs replacing.A tube is blanked by connecting it into the analytical system and thermally desorbing for a short time while air is drawn through simultaneously to remove mercury traces. The temperature within the tubes reaches 700 "C during desorption. Decant the excess of silver solution. A sampling tube is shown diagrammatically in Fig. 3. Silvered alumina (710 pm- 1.0 mm particle size) Ball joint I Quartz-woo' h ' ' Nichrome heating wire 10 cm 4 b Fig. 3. Sampling tube for mercury vapour. Description of the Analytical System The analytical system (Fig. 1) is largely constructed from clear fused quartz tubing and silicone-rubber tubing. Cup-joints on the fused quartz tubing enable the sample tubes to be clipped into place and thick walled silicone-rubber tubing is used to prevent accidental constriction which would alter the air flow-rate and affect signal peak shape.Air is drawn through the system by means of the diaphragm pump and the screw clip is used to set the required air flow-rate. The measurement range of the system is varied by means of a range of flow restrictors that force differing controlled fractions of the sample plug along the by-pass line away from the absorption cell. Typically only 5-10% of the plug is allowed to pass through the cell. The thermally desorbed sample can be recollected for a duplicate determination if desired, with 80-95% recollection efficiency by inserting a second sample tube into the system as shown in Fig. 1. Fig.4 shows typical signal peak shapes for the two measurement ranges that have been used. The sample peak is moni-tored using the pen recorder and should be of similar shape to the diagram; if this is so it indicates that the analytical result will be acceptable. A dissimilar peak shape may well indicate a constriction in the connecting tubing or some other fault. Leaks in the analytical A duplicate run is not usually required 270 TAYLOR RE-USABLE SAMPLING TUBES FOR AIRBORNE HG AnaZyyst VoZ. 108 3 2 1 0 4 3 2 1 0 Ti me/m i n Fig. 4. Signal peak shapes a t 30 mm min-' chart speed. (a) Lower calibration range (0-2.5 pg Hg) peak obtained for 1.5 pg of Hg; and (b) higher calibration range (0-10 pg Hg) peak obtained for 5 pg of Hg. system are minimised as it is under negative pressure.The jointed receiver vessel (Fig. 1) smooths out the pulses of the pump so that signal noise is minimised. Its shape and size are critical in controlling the shape of the signal peak. If it is omitted from the system the signal takes the form of a rapid succession of rather sharp peaks because of the pulsating air flow instead of the single relatively noise-free peak shown (Fig. 4). Calibration ranges of 0-2.5 and 0-10pg of mercury have been used so far; other ranges may be obtained by changing the flow restrictors. Calibration lines are basically linear but may curve towards the higher concentration end (Fig. 5 ) . The signal peaks shown (Fig. 4) are non-Gaussian the signal rising rapidly to begin with and tailing off more slowly after the maximum has been reached.The lower range gives sharper peaks than the higher one, presumably because larger amounts of mercury take longer to desorb so that their passage through the absorption cell is more protracted. Measurement of signal peak height was chosen in preference to measurement of integrated area because it was found to give adequate accuracy and precision and was a lot more convenient to use because the signal takes a relatively long time to return to the base line. Fig. 6 shows the modification to the system that is necessary before longer period samples taken on a 6-cm column length of substrate can be analysed. Direct desorption through the cold vapour cell of one of these sample tubes gives a very poor signal peak shape. The sample is therefore transferred from the 6-cm tube to a 2-cm tube before it is analysed.The analytical step takes 2-4 min depending on the calibration range being used and provided that the sample tube is unclipped from the system very shortly after the signal peak maxi-mum has been reached (the operator must wear effective heat resistant gloves to (lo this). Mercury remaining on the tube must then be flushed out at a later time. The step takes 8-10 min if all mercury is flushed from the tube at the time of analysis and the tube is allowed to cool to ambient temperature before removal and replacement with the nest sample. 0.800 0.700 0.600 0.500 0.400 2 0.300 Q 0.200 0.100 3) m 0 1 2 0 1 2 3 4 5 6 7 8 9 1 0 Amount of Hgiug Fig. 5. Calibration lines obtaiiicd for ranges (a) 0-2.8 and ( b ) 0-10 pg of Hg February 1983 Sampling tubing that will not easily constrict in use.SAMPLE COLLECTION AND ANALYSIS BY COLD VAPOUR AAS 27 1 Connect the blanked sample tube to the sampling pump using thick-walled silicone-rubber Plastic tubing adaptors can be used to join the Quartz absorption cell Fig. 6. Two-stage thermal desorption system for use with 6-cm silvered alumina tubes. different diameters of tubing that will be needed. I t may be necessary to fit a glass-fibre pre-filter in front of the sampling tube to prevent particulate matter from entering or a suitable adsorption tube to remove interfering vapours,4 while allowing mercury vapour to pass through. Switch on the sampling pump and allow adequate warm-up time before adjusting the sampling air flow-rate to 100ml min-l using a suitable air flow meter (if a Rotheroe Mitchell C500 pump is used its integral air flow meter should not be relied on for this purpose).At the end of the sampling period the air flow should again be checked and the mean of the two flow readings taken before and after sampling should be used to calculate the air volume sampled. Note which end of the sampling tube was remote from the sampling pump disconnect the tube and cap both ends with Parafilm sealing tissue, prior to analysis. Calibration Procedure The thermal desorption time should be as short as possible therefore pre-set the variable auto transformer of the thermal desorption apparatus so that the full power used (190 W) will be supplied immediately the on switch is activated but do not switch on yet.Make a series of calibration solutions by appropriate water dilution of the 1000 p.p.m. mercury standard solution such that a 10-pl aliquot of calibration solution will contain the calibration amount e.g. a 1-pg mercury calibration standard contains 1 pg of mercury(I1) in lop1 of solution. Place a 10-pl aliquot of a calibration solution on to one end of the silvered substrate column in a blanked 2-cm sample tube and insert this into the analytical system with the dosed end of the tube furthest from the absorption cell. Adjust the air flow-rate through the system to 1 lmin-l. Ensure that the spectrophotomer is correctly zeroed then thermally desorb the sample tube and note the maximum peak height reading. Plot a graph of maximum peak height versus mass of mercury.The unmodified analytical system (Fig. 1) is used. Analytical Procedure Short-term sampling tubes Remove the capping film from the ends of the used short-term sampling tube and connect the tube into the system (Fig. 1) with the correct orientation to maintain the same direction of air flow through the tube as during sampling. Adjust the air flow-rate through the system to 1 1 min-l. Ensure that the spectrophotometer is correctly zeroed then thermally desorb the sample tube and note the maximum peak height for comparison with the calibra-tion 272 TAYLOR RE-USABLE SAMPLING TUBES FOR AIRBORNE HG Analyst vd. 108 Long-term sampling tubes Remove the capping film from the ends of the used long-term sampling tube and insert the tube into the system in the “1st sample tube” position with the correct orientation to maintain the same direction of air flow through the tube as during the previous sampling.Connect a 2-cm tube into the “2nd sample tube’’ position. Adjust the air flow-rate through the system to 1 1 min-l and then thermally desorb tube 1 for 5 min. Allow the hot tube (1) to cool for about 3 min then re-adjust the air flow-rate to 1 1 min-l and connect the thermal desorption circuit to tube 2. Ensure that the spectrophotometer is correctly zeroed then desorb tube 2 through the absorption cell and note the maximum peak height reading for comparison with the cali-bration. Tests to Investigate the Mercury Collection and Thermal Desorption Efficiency of Silvered Alumina Several tests were carried out to investigate the collection and thermal desorption efficiency of the silvered alumina tubes for mercury vapour.In all the tests a similar experimental arrangement to that shown in Fig. 6 was used with a blanked 2-cm sample tube (2) in the “2nd sample tube” position and either a 2-cm or 6-cm sample tube (1) in the “1st sample tube” position. Tube 1 was always connected into position with the dosed end remote from the absorption cell. In the first test a 10-pl aliquot containing 1 pg of mercury(I1) was spotted on to a 2-cm tube (1). The air flow-rate was adjusted to 1 1 min-l and tube 1 w2s thermally desorbed for 5 min. The spectrophotometer reading indicated no mercury breaking through tube 2 during this time. After the thermal desorption period tube 1 was allowed to cool then a further 120 1 of air were pulled through the system again at the rate of 1 1 min-1.There was no indication of any mercury breaking through tube 2 during this 2-11 period which denion-strated the very good collection efficiency of silvered alumina for mercury vapour. In a second test tube 1 (2 cm) was dosed with 10-p1 aliquots containing up to 10 pg of mercury(I1). The air flow-rate through the system was adjusted to 1 lmin-l and tube 1 was thermally desorbed for 5 min. There was no indication of mercury breaking through tube 2 throughout the desorption period. Tube 1 was allowed to cool for about 3 min the thermal desorption circuit was connected to tube 2 and the air flow-rate was re-adjusted, where necessary to 11 min-l.Tube 2 was then thermally desorbed and the amount of mercury coming off was measured when this was within the measurement range of 0-2.5 pg of mercury in use at the time. Determination of analytical precision for 6-cm tubes The air flow-rate was set as in previous tests and tube 1 was thermally desorbed for 5 min to transfer the mercury to tube 2. As before there was no indication of mercury breaking through tube 2 during this period. Tube 2 was then thermally desorbed and the amount of mercury coming off was measured. The measurement range 0-10 pg of mercury was being used this time. Determination of analytical precision for 2-cm tidbes Analytical precision for the 2-cm tubes was determined as follows using the unmodified analytical system (Fig.1). For each of the calibration ranges used two mercury levels were selected one towards the lower and one towards the upper end of the range. For the determinations at each mercury level two sample tubes were used alternately so that one could be cooling while the other was in use. For each determination lop1 of the standard solution were aliquotted into a tube which was subsequently thermally desorbed and the amount of mercury coming off measured. Ten determinations were made at each of the four selected levels. Tests to Validate the Sampling and Aiialytical Technique The sampling and analytical procedures were validated by sampling mercury standard atmospheres generated within the laboratory and the atmosphere in a factory manufacturing zinc amalgam batteries.Set up the modified analytical system as shown in Fig. 6. Following the second test a 6-cm tube (I) was dosed with 5 pg of mercury(I1). This particular test was repeated 11 times with the same pair of tubes February 1983 Standard atmosphere tests Standard atmospheres of mercury vapour in air over the range 80-400 pg m-3 were sampled simultaneously with 2-cni silvered alumina tubes and large hopcalite tubes operated nominally at 100 ml min-l and 1 1 min-l respectively. The sampling period was 1 h in all instances. A manifold pumping system used to maintain sampling flow-rates kept the 1 1 min-l flow steady but the nominal 100mlmin-1 flow generally varied between 80 and 120mlmin-l. Small hopcalite tubes were connected behind the silvered alumina tubes to collect any break-through and the mercury levels of the standard atmospheres were continuously monitored with a Data Acquisition mercury meter.Air samples of 12 1 were taken from a similar range of standard atmospheres using 2-cm silvered alumina tubes operated at a sampling air flow-rate of 200 ml min-l. A mercury standard atmosphere of concentration 150 pgrnr3 as indicated by the mercury meter was sampled simultaneously over a 6-h period with 6-cm silvered alumina tubes backed by small hopcalite tubes and with small hopcalite tubes. Flow-rates were again set at 100 ml min-l and 1 1 min-I respectively for the two types of tube. The manifold pumping system was used with the large hopcalite tubes but individual sampling pumps were used to maintain the flows through the silvered alumina tubes with reasonable success.The mercury concentration of the standard atmosphere was again continuously monitored with the Data Acquisition meter. SAMPLE COLLECTION AND ANALYSIS BY COLD VAPOUR AAS 273 Long-term sampling was also tried. Samples taken within a factory The atmosphere within a factory manufacturing zinc amalgam batteries was sampled with 2-cm and 6-cm silvered alumina tubes operated at a nominal flow-rate of 100 mlmin-l. The sampling period for the 2-cm tubes was 1 11 while that for the 6-cm tubes was approxi-mately 5 h. For each sample taken on silvered alumina a duplicate sample was taken at the same position and over the same time period using a large hopcalite tube operated at a nominal flow-rate of 1 1 min-l. All sampling tubes were preceded by a 2.5-cm diameter glass-fibre pre-filter to prevent airborne mercury( 11) oxide dust from entering the tubes.Results and Discussion Results of Tests to Investigate the Mercury Collection and Thermal Desorption Efficiency of Silvered Alumina The results of the second test were as follows for masses of mercury(I1) dosed on tube 1 of 1 and 2.5 pg the corresponding masses of mercury measured from thermal desorption of tube 2 were 0.94 and 2.4p.g respectively. These results demonstrate the high collection efficiency of silvered alumina for mercury vapour and also the quantitative nature both of the thermal desorption process and of mercury transfer from one tube to another. Analytical precision for 6-em tubes The mean of the eleven mass determinations of mercury coming off tube 2 for 5 pg dosed on tube 1 was calculated to be 5.4 pg and the relative standard deviation based on ~CT,-~, was 12.7%.This test validates the analytical procedure used to deal with long-term samples and also demonstrates the re-usability of the sample tubes; neither of those used showed any visible physical deterioration after the tests. Analytical precision for 2-ern tubes The results of the test to determine analytical precision for the 2-cm tubes are given in Table I. Generally the relative standard deviation values are around 8-9%. However, 0.2 pg of mercury at the bottom end of the lower calibration range gave a value of 22.5%. The relative standard deviation figure corresponding to this level could be improved by increasing the sensitivity of the system sufficiently to bring 0.2 pg closer to the centre of the calibration range where precision is better because of the greater signal to noise ratio.The sample tubes are all slightly different because they are packed and wound with heating wire by hand. A more reproducible preparation method might improve precision further 274 TAYLOR RE-USABLE SAMPLING TUBES FOR AIRBORNE HG Analyst VoZ. 108 TABLE I ANALYTICAL PRECISION FOR 2-cm TUBES Relative standard deviation values based on 2un-1 for ten deterininations a t chosen mercury levels. Calibration range 0-2.5 pg of niercury-Relative standard Mercurylpg deviation yo 0.2 22.6 2.0 8.3 Calibration range 0-10 pg of mercury-Relative standard Mercurylpg deviation % 3.0 9.3 10.0 7.7 Limit of Detection The limit of detection realised in practice depends on the measurement range used and becomes higher as the fraction of sample passing through the absorption cell is decreased.When the by-pass line and flow restrictor were removed altogether so that all the sample had to pass through the absorption cell a limit of detection (20 noise) of 3 ng was achieved. This is equivalent to 0.5 pg m-3 for a 6-1 air sample (1/100 TLV). A lower detection limit was not required in these tests but could probably be achieved by using shorter lengths of tubing within tlie analytical system to reduce the amount of dead volume by increasing the length of the absorption cell and reducing its diameter to give increased concentration of mercury atoms in the light beam or by using a different cell shape12 to optimise this effect.I t is likely that the detection limit could also be lowered by varying the air flow-rate through tlie analytical system. Imprecision Arising from Sampling It is common for the sampling pump flow-rate to change by up to +loo/; during the sanipling period; occasionally the change is more than 3 loo//,. Such changes occur even when the precaution is taken of allowing the pump to warm up before adjusting it to give the required flow-rate and they occur for all sampling media. The method used liere of calculating the total sample air volume from the mean of the pre- and post-sampling flow-rate readings assumes that any change in flow-rate will have taken place in a constant and uniform manner over the whole sampling time span.This assumption is unlikely to be completely valid and consequently sampling is a very major cause of imprecision and inaccuracy in the analytical result obtained for a sample. Results of Tests to Validate the Sampling and Analytical Technique Standard atiizosplzcre tests No breaktlirougli was detected for the 2-cm tubes operated at a nominal 100 ml min-l for 1 h. The results obtained for these given in Table 11 correlate well with tlie other two sets of data for liopcalite and the Data Acquisition mercury meter witliin tlie constraints imposed Instrumental imprecision is smaller than imprecision arising from sampling. TABLE I1 RESULTS FROJI SILVERED ALVJIISA .4XD HOPCALITE TUBES FOR VARIOUS NERCVRl- ST.ISD.IRD ATMOSPHERES JIcrcury standard atmosphcrcs 1vc't-e sa,niplctl for 1 h using (a) 2-cni 3ilvcrctl alumina tubes opcratcd at 100 nil min-1 ant1 ( b ) lnrgc hopc;LlitC tubes operatctl at 1 1 inin-'.Data Acquisition mercury I%titiintctl riicrcury conccntration/pg iiirR -----L- ~~ --I iiietcr reading for niercury concentration/pg 111-~ (4 (b) 80 82 84 76 72 72 85 93 94 89 83 150 137 137 155 1.10 "50 2330 260 230 215 "!) 400 428 440 383 405 39 February 1983 SAMPLE COLLECTION AND ANALYSIS BY COLD VAPOUR AAS 275 by the sampling arrangement. For the 2-cm silvered alumina tubes operated at a nominal 200 ml min-l for 1 h there was slight mercury breakthrough between 2 and 574 indicating that a sampling rate of 100 ml min-l and a sampling time not exceeding 1 h are probably optimum for 2-cm tubes.Again the correlation of the silvered alumina results with the other values is perfectly acceptable within the experimental constraints imposed by the sampling arrangements. The results obtained for the long-term sampling tests are shown in Table 111. TABLE I11 LONG-TERM SAMPLING TEST RESULTS A mercury standard atmosphere of nominal concentration 150 pg m-3 was sampled for 6 h using (a) 6-cm silvered alumina tubes operated at 100 ml min-I and (b) large hopcalite tubes operated a t 1 1 min-I. Data Acquisition mercury Estimated mercury concentration/pg m-s meter reading for mercury r A \ concentrationlpg m-3 (4 (b) 150 138 125 160 159 161 Samples taken within a factory They show excellent agreement between the two sampling and analytical techniques taking into account unavoidable fluctuations of sampling flow-rate from sample to sample.The results obtained for the factory sampling exercise are shown in Table IV. TABLE IV RESULTS FOR SAMPLES FROM A FACTORY MANUFACTURING ZINC AMALGAM BATTERIES All samples were duplicated using (a) long or short term silvered alumina tubes operated a t 100 ml min-l and (b) long or short term large hopcalite tubes operated a t 1 1 inin-'. The short-term samples had a sampling time of 1 h whilst the long-term samples had a sampling time of about 5 h. Estimated mercury concentration/pg n1-3 Sample type Short term. . Long term . . (4 44 48 24 29 35 41 52 33 43 22 30 22 27 (4 39 46 20 26 32 38 62 39 41 27 31 22 26 Advantages and Disadvantages of the Proposed Method Compared with Hopcalite Several of the disadvantages of hopcalite have been mentioned in the Introduction namely expensive sample tubes that can only be used once a heavy personal sampling pump is needed for the larger sampling tube there is background level of mercury on hopcalite time-consuming .ample preparation prior to analysis and tedious analytical steps often involving sample dilution to keep within the measurement range.The proposed method is superior to hopcalite in respect of all of the above. Silvered alumina tubes are fairly costly to make though this is offset against the fact that each tube can be used repeatedly. Both 2-cm and 6-cm tubes are operated at an air flow-rate of 100 in1 niin-l so that the same light-weight sampling pump is adequate for use with each.They have no inherent background level o 276 TAYLOR mercury and there is no sample preparation prior to analysis. The analytical step for the silvered alumina tubes is much less involved than the corresponding one for hopcalite it is much easier to carry out and consequently is less prone to operator error. In addition the hopcalite procedure requires large amounts of calibrated glassware for large numbers of samples so that a sizeable post-analysis cleaning-up operation is required. The proposed method avoids this completely. Corrosive acids concentrated hydrochloric and nitric are used to effect dissolution of hopcalite. Corrosive reagents are completely avoided in the proposed procedure apart from being used in the silvering process.Also in the new analytical method fogging of the absorption cell walls by desorbed water vapour does not occur. I t frequently occurred with the recirculating system used for analysis of hopcalite tubes though could be remedied by occasionally heating the cell using a flow of warm air from a hair dryer. Analytical precision for the two methods is similar as is the over-all accuracy of the results obtained. However the proposed method is more flexible ; different measurement ranges can be used; the 0-10 pg range is several times wider than the calibration range attainable with the hopcalite analytical system or the range can be narrowed to give very high sensi-tivity. The limit of detection that can be achieved using the proposed method is also far superior as it is not limited by a background level on the sample tube.Disadvantages of the new method are that the analytical system is more complicated, though this is offset by the greater ease and speed of analysis cost of purchasing the com-ponent parts will be marginally greater and the fact that each sample can only be determined once unless a duplicate determination of lower accuracy is acceptable. Some prior know-ledge of the amount of mercury likely to be present in a sample is necessary so that an appropriate measurement range can be set up. Conclusion Silvered alumina is an effective sampling medium for airborne mercury vapour and can be I t offers considerable advantages over Sampling tubes are re-usable and could possibly be redesigned to fit an automatic used for both long- and short-term sampling. hopcalite. thermal desorption system with a view to automating the analysis. 1 . 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Rathje A. O. Marcero D. H. and Dattilo D. Am. Tnd. Hyg. Assoc. J . 1974 9 571. Hathje A. 0 and Marcero D. H. Am. Ind. Assoc. J . 1976 5 311. “Health and Safety ICxecutive Threshold Limit Values for 1980,” Guidance Note EH 15/80 HM Long S. J. Scott D. I<. and Thompson li. J. Anal. Chem. 1973 45 2227. Kneip T. J. Health Lab. Sci. 1975 12 158. Yoshida Z. and Motojiina K. Anal. Chim. Acta 1979 106 405. Dumarey I<. Heindryckx It. Dams I<. and Hoste J. Anal. Chil-n. Acta 1979 107 159. Trujillo 1’. I < . and Campbell E. E. Anal. Chem. 1975 47 1629. MDHS 3/81 “Generation of Test Atmospheres of Organic Vapours by the Syringe liijcction Technique. Portable Apparatus for Laboratory and Field Use,” HSE 1981 Health and Safety Executive Occupational Medicine and Hygiene Laboratories London. Stationery Office London 1980. Wright M. D. and Purnell C. J. unpublished work. “NIOSH Manual of Analytical Methods,” Volume 1 US Department of Health Education and Tuncel G. and Ataman 0. Y. A t . Spectrosc. 1980 1 126. Welfare Cincinnati OH April 1977 pp. 175-177 paras. 8.1 and 8.2. Received July 26th 1982 Accepted October 26th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800265

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst February 1983 Vol. 108 $9. 277-285 Extraction of Metals from Soils and Sewage Sludges by Refluxing with Aqua Regia 277 Michael L. Berrow and Winnie M. Stein Department of Spectrochemistry The Macaulay Institute for Soil Research Craigiebuckler Aberdeen, AB9 2QJ The analysis of soils for total metal content using various acid digestion procedures and flame atoniic-absorption spectroscopy has been investigated. Refluxing with aqua regia was more effective than digestion in an open vessel and produced results comparable to those obtained by bomb digestion the most vigorous method used. Refluxing with aqua regia extracted at least 80% of the total chromium copper lead and manganese from sewage sludges and sludge-treated soils. Analysis of the uncontaminated Canadian Reference soils showed that a similar proportion of the total cadmium iron and zinc was also extracted.Results for the analysis of six typical Scottish topsoils were in agreement with the conclusions obtained with the Canadian soils. Keywords Trace metal determination ; soils ; sewage sludges ; acid digestions ; atornic-absorption spectrophotornetry Some 45% of the 1.24 x lo6 dry tonnes of sewage sludge produced annually in England and Wales is utilised in agriculture1 and sludges from industrial areas can often contain large amounts of heavy metals.2 Excessive increases in the levels of metals in soils due to the dis-posal of metal-contaminated sludge on land can be harmful to plant and/or animal g r o ~ t h . ~ ? ~ The metals of major concern which are often present in sludges in amounts much greater than in soils are cadmium chromium copper lead molybdenum nickel and zinc.5 All of these elements can be satisfactorily determined in the finely ground ignited sludge or soil by cathode-layer arc emission spectrography with the exception of cadmium and zinc.These two elements produce weak spectral lines which result in poor sensitivity. Cadmium and zinc are, however the two elements that probably require the most careful monitoring when sewage sludge is disposed of on land.6 For this reason suitable acid-digestion procedures have been tested to establish the most efficient means for the dissolution of these and other elements in sludges and soils. Efficient dissolution is the essential first stage in the determination of metals by atomic-absorption spectrophotometry where the flame technique has the sensitivity required as well as the advantages of operational simplicity high specificity and relatively low cost.The analysis of soils for total metal content is also used in geochemical prospecting where it can be of assistance in the location of mineral or ore bodies. Metals of importance in this respect include Ag Cd Co Cr Cu Mn Ni Pb Sn and Zn.'s8 Recently atmospheric deposits and the disposal of a variety of wastes on land have been found to increase sometimes to harm-ful levels the amounts of some biologically important metals in soils.g Where availability for uptake by plants is important such as in agriculture and horticulture a determination of the mobile or readily extractable content is required.For pollution monitoring and ore prospect-ing however an analysis for total metal content is often necessary. Five mineral acids (hydrofluoric perchloric sulphuric nitric and hydrochloric) either separately or in a combination have been widely used for the simultaneous extraction of a large number of metals in s ~ i l s . ~ ~ l ~ Hydrogen fluoride is highly effective in breaking down most common silicates but is a dangerous chemical requiring careful handling. Perchloric acid is a very efficient oxidising agent but its use is also hazardous particularly for samples containing much organic matter. Procedures based on the use of sulphuric acid digestion are likely to produce low results for lead owing to precipitation of lead sulphate. Soils and sludges contain-ing much calcium may also produce low results for metals owing to precipitation or occlusion by insoluble calcium sulphate.ll Digestion procedures using hydrochloric acid nitric acid and aqua regia were therefore tested as these suffer from none of the hazards and drawbacks of the other three acids 278 BERROW AND STEIN EXTRACTION OF METALS FROM SOILS Analyst VoZ.108 Tests on four sewage sludge-treated soils using simple digestion of the ground soil with the three acids chosen above in a beaker on a steam-bath were carried out. Digestion in an open vessel was tested initially because the apparatus is simple easily cleaned and free from metals, which could cause contamination problems. A more vigorous digestion with boiling aqua regia for 2 h under reflux was then evaluated for the analysis of sludges sludge-treated soils and uncontaminated soils.The efficiency of extraction by the different procedures was tested by comparison with bomb digestions and by emission spectrographic analysis of the acid-insoluble residues and the original soils. This latter technique involves total consumption of the sample and provided an alternative measure of the total contents. The aim of the investigation was to establish a simple safe and rapid method using flame atomic absorption, for the determination of the total contents of metals in soils and sewage sludges. Experimental Apparatus Digestion apparatus All digestions in open vessels were carried out in 100-ml Pyrex beakers on a polypropylene steam-bath. Digestions with aqua regia under reflux were carried out in 100-ml round-bottomed flasks equipped with a 30-cni condenser on an Electrothermal Extraction apparatus (E30-220 Macfarlane Robson Limited Tyne and Wear).Bomb digestions were carried out in 25-ml capacity PTFE-lined general-purpose acid digestion bombs (Parr 4745 Parr Instrument Co. Moline IL USA). A tovtic-a bsorfit iou spectropJzoto.mrtry All determinations were made using an 11 751 or a I'E560 atomic-absorption spectrophoto-meter equipped with a single-slot burner and an air - acetylene flame except for aluminium, chromium and iron which were determined in a tfinitrogen oxide - acetylene flame. Hollow-cathode lamps (S. and J. Juniper and Co. Harlow Essex) and the following wavelengths were used A1 396.3 Cd 228.8 CAI 324.7 Cr 357.9 1;e 248.3 Pb 217.0 Mn 279.5 Ni 232.0 and Zn 213.9 nm.For all elements determined at a wavelength of less than 300 nm a deuterium lamp background correction was applied. For the determination of aluminium 1000 mg 1-' of sodium were added to all solutions as an ionisation suppressor. In some of the soil digests where the levels of cadmium were below the limit of detection of the normal flame technique the improved sensitivity of an atom-trapping device attached to a Techtron AAG instrument was used.12 E fit issio it sfitlct rogvafilt ic cifijmratzts Determinations of total trace element content by emission spectrograyh\~ were made using the cathode-layer enriclment effect in a 9-X t1.c. arc with a Hilger E492 large quartz spectro-graph according to tlie procedure described by J I i t ~ l ~ e l l .~ ~ First and second minute exposures were recorded over tlie wavelength range 480.0- 2'77.5 nni on Kodak 54 1 spectrograpliic plateb, developed in a D19 developer at 20 "C for 2 inin. The spectra were assessed by visual com-parison against those of standards made up in a syntlietic soil base. Good interpolations can be made to an accuracy of 520o/b of the actual amount present especially for contents at which the line is not too dense. Reagents AnalaR-grade acids were distilled in an all-Pyres distillation apparatus containing a piece of clean porous unglazed ceramic tile. Distillation of the acids used for digestion is rcconiniended so that very low blank levels can be maintained. It-nfcv. Water distilled from an all-glass apparatus was used for preparing tlie reagents and standard solutions.Hydvocldovic acid 6 11 20.2";; m/m sp. gv. (20 "C) 1.1 g nil-1. Prepared by diluting hydro-chloric acid sp. gr. (20 "C) 1 . l 8 g nil-1 with distillcd n-ater 4 + 3 followed by distillation in an all-Pyrex still rejecting the first and last portions of the distillate. Xitric acid 16 hi 707/ i i ~ / i i z sp. gr. (20 "C) 1.4 g nil-1. Prepared l q - diluting nitric acid, sp. gr. (20 "C.) 1.42 g with distilled water 10 + 1 followed by distillation in an all-Pyrex still rejecting tlie first and last portions of tlie distillate February 1983 Aqua regia. Hydrochloric acid 0.06 M. Nitric acid 2 M. Standard metal solutions. AND SEWAGE SLUDGES BY REFLUXING WITH AQUA REGIA 279 Prepared by diluting 10 ml of 6 M hydrochoric acid with water to 1 1 in a measuring cylinder and mixing thoroughly.Prepared by diluting 125 ml of 16 M nitric acid with water to 1 1 in a measuring cylinder and mixing thoroughly . Prepared by serial dilution of 1000 mg 1-1 metal stock solutions. All standard solutions were prepared containing tlie same reagents that were added to the samples. Prepared by mixing 6 M hydrochloric acid and 16 M nitric acid 3 + 1. Procedure Soil samples were dried in air at a temperature not exceeding 30 "C and sieved through a 2-mm sieve. The sieved fine soil was thoroughly mixed coned and quartered and about 30 g ground in an agate planetary ball mill (Fritsch Pulverisette No. 5 Alfred Fritscli OHG, Oberstein West Germany) for 30 min to <150 pm in size.Sewage sludges were dried at 105 "C thoroughly mixed and an aliquot was ground to a fine homogeneous powder. Ground soils and sludges were stored in glass bottles. Results Preliminary Experimental Work Various digestion procedures were tested on four sludge-treated soils using simple digestion in 100-ml Pyrex beakers on a steam-bath. Acids were added in 10-ml aliquots and samples allowed to digest to dryness between each addition. All treatments with nitric acid or aqua regia were finally treated with 10 ml of 6 M hydrochloric acid and taken to dryness to convert the soluble residues to the chloride form. All digests after finally taking to dryness were left on the water-bath for at least 0.5 h to dehydrate any silica in the residue.The dried residue was then dissolved in 0.06 M hydrochloric acid filtered and made up to 25 ml with 0.06 M hydro-chloric acid. A preliminary test involving ten replicate aqua regia digestions on one soil for cadmium copper manganese and zinc produced relative standard deviations of about 4%). The digestion procedures carried out on the four soils in duplicate included tlie following: 2 x 10 nzl of 6 M hydrochloric acid in an open vessel; 2 x 10 ml of 6 hi hydrochloric acid, digested and covered with a watch-glass for 1 11 after each addition; 2 x 10 ml of 16 M nitric acid in an open vessel; 2 x 10 ml of 16 M nitric acid digested and covered with a watch-glass for 1 h after each addition; 2 x 10 ml of aqua regia in an open vessel; 2 x 10 ml o f aqua regia, digested and covered with a watch-glass for 1 h after each addition; 4 x 10 ml of aqua regia in an open vessel; and 6 x 10 ml of aqua regia in an open vessel.The results for the digestion of one of these soils are shown in Table I from which the follow-ing were concluded digestion with 6 M hydrochloric acid 16 M nitric acid or aqua regia pre-pared from these produced rather similar results ; digestion with four or six successive 10-ml aliquots of aqua regia did not extract more of the metals than digestion with just two 10-ml aliquots ; and slightly greater amounts were generally extracted when the digest was covered with a watch-glass and allowed to reflux for part of the time. TABLE I PRELIMISARY ANALYSES MEAN VALUES OF DUPLICATE ANALYSES OF A SLCT1)GE-TREATED SOIL USING DIFFEREKT ACID DIGESTION PROCEDURES Mean exprcsscd as milligrams per kilogram in dry soil.Xlethod 1 J 3 4 J 6: 'I Proccdure Hydrochloric acid open vc.ssel Hydrochloric acid covrred Sitric acid opcm vcssc.l Sitric acid covcrcd Aqua rcgia open vt*sscl Aqua rrgia covered Aqua wgia x 2 open vc*ssel Aqua regia x 3 open vrsscl Total t1.c. arc Cadmium 4.7 .5.1 4.9 4.x 45 4.7 4.0 4.0 S.d. Chromium 5R 5!) 64 47 61 X.d.* K.d. 100 - a I Copper 53 54 64 i 4 *50 53 48 48 60 Lead Manganese Nickel Zinc 54 810 31 490 .i4 340 32 500 4X 320 29 490 46 320 30 490 46 290 29 470 45 300 30 460 X.d. 210 29 420 N.d. 220 28 410 80 330 40 N.d. * S o t determined 280 BERROW AND STEIN EXTRACTION OF METALS FROM SOILS AndYSt VOJ.108 In a similar test it was also found that the use of hydrogen peroxide to supplement the action Slightly greater of nitric acid did not extract more of the metals than nitric acid alone. amounts were extracted when the digest was allowed to stand overnight before filtering. Digestion Under Reflux Whilst considering the results from the preliminary investigations an open vessel digestion procedure (Method A) was compared with an extraction procedure that involves standing overnight in aqua regia followed by refluxing for a period of 2 h (Method B). The procedure is similar to that proposed by the EEC Community Bureau of Reference for sludge analysis. In order to assess the efficiency of extraction the total metal content was also determined by bomb digestion (Method C) and by dx.arc analysis. These three digestion procedures which were used to analyse the same four sludge-treated soils that were used in the preliminary tests are described below. Method A Place on a steam-bath and digest with four successive 10-ml aliquots of 6 M hydrochloric acid taking to dryness after each addition. Extract the residue with 0.06 M hydrochloric acid and filter washing with 0.06 M hydrochloric acid into a 25-ml calibrated flask. Weigh 1 g of sample into a 100-ml Pyrex beaker. Make up to the mark with 0.06 M hydrochloric acid. Method B Add 2-3 ml of water to obtain a slurry. Then add per gram of dry sample 7.5 ml of hydrochloric acid [sp. gr. (20 "C) 1.1 g ml-11 and 2.5 ml of nitric acid [sp.gr. (20 "C) 1.4 g ml-l]. Cover the flask and leave to digest at room temperature for 16 h. Fit a 30-cm reflux condenser on top of the flask and boil gently under reflux for 2 h on a temperature controlled electrothermal extraction apparatus. Filter the solution into a 100-ml calibrated flask. Rinse the filter-paper and residue five times with a few millilitres of warm (about 50 "C) 2 M nitric acid. Allow to cool and dilute to the mark with 2 M nitric acid. Weigh 2 g of sewage sludge or 3 g of soil into a 100-ml round-bottomed flask. Allow to cool and rinse the condenser with not more than 30 ml of water. Method C Bomb digestions of the ground soil with aqua regia gave rise to frothing problems when the bomb was opened so determinations were carried out using soil ignited overnight at 450 "C and 6 M hydrochloric acid as the digesting acid.Weigh 0.5 g of soil ash into a PTFE cup and add 10 ml of 6 M hydrochloric acid. Close the bomb carefully and place in an oven a t 120 "C for 6 h. Allow to cool transfer the digest into a 100-ml beaker and evaporate to dryness on a steam-bath to dehydrate the silica. Take up in 0.06 M hydrochloric acid and filter with wash-ing into a 25-ml calibrated flask. These show that aqua regia digestion under reflux improves the extraction over that in an open vessel by about 10-2070 for all the metals determined except cadmium for which the increase is smaller. The digestion under reflux produces results comparable to those obtained by bomb digestion. Re-extraction of the residues remaining after both open vessel and bomb digestion by further treatment with aqua regia shows some retention of metals.This is generally less for the bomb digestion residues than for those remaining after open vessel digestion. The retention may be related to the small volumes used in extracting and washing the residues to make up to a final volume of 25 ml. The residues analysed by d.c. arc following re-extraction in the open vessel and aqua regia reflux digestion (results adjusted for loss in mass during digestion) suggest some retention of chromium but most of the copper lead manganese and nickel have been extracted. I t can be concluded that refluxing with aqua regia for 2 h is a highly effective procedure for the extraction of metals from sludge-treated soils. An evaluation of Method B for the analysis of a range of samples (sewage sludges sludge-treated soils and uncontaminated soils) was therefore carried out.The results for one of the four sludge-treated soils tested are presented in Table 11 February 1983 AND SEWAGE SLUDGES BY REFLUXING WITH AQUA REGIA TABLE I1 281 COMPARISON OF DIGESTION PROCEDURES MEAN VALUES OF ANALYSES OF A SLUDGE-TREATED SOIL OBTAINED USING DIFFERENT DIGESTION PROCEDURES AND OF THE ACID-INSOLUBLE RESIDUES REMAINING Mean expressed as milligrams per kilogram in dry soil. Number of Digestion analyses Cadmium Chromium Copper Lead Manganese Method A 4 4.8 N.d.* 51 47 270 Final residue d.c. arc . . 2 N.d. 25 1 I 25 Re-extraction'df A I R i 4 <1.2 N.d. 6 3.3 16 Method B 5 5.0 88 61 85 300 Residue d.c.arc . . . . 2 N.d. 20 8 7 50 Method C 3 5.9 N.d. 56 50 320 Re-extraction'of AIR' 3 <2.3 N.d. 4.4 3.9 <9 Total d.c. arc . . . . 5 N.d. 100 60 80 350 * Not determined. t AIR acid-insoluble residues. Nickel 29 2.5 H 34 <n 34 <2.3 40 Zinc 490 11 N .d. 570 N.d. 600 4.8 N.d. Sewage Sludges Three sewage sludges were analysed ten times using 2-g samples and the mean and relative standard deviation (RSD) values are reported in Table 111. RSD values varied according to the element determined and the sample but averaged 3% for all seven elements in the three sludges. Determinations of the total contents of all the metals except cadmium were also obtained by four replicate analyses using d.c arc emission spectrography.Comparison of the values obtained by the two methods bearing in mind the semi-quantitative nature of the spectrographic technique showed that more than SOY0 of the total contents had been extracted. Sludge samples No. 2 and 3 Table 111 are from industrialised areas and because their zinc equivalents are 8270 and 12290 mg kg-1 dry matter respectively it is unlikely that they would be recycled to agricultural land. TABLE 111 ANALYSIS OF SEWAGE SLUDGES MEAN AND RELATIVE STANDARD DEVIATION VALUES FOR TEN ANALYSES OF THREE SEWAGE SLUDGES USING AQUA REGIA REFLUX DIGESTION Mean expressed as milligrams per kilogram in dry sludge and relative standard deviation (RSD) in %. r-A-- 4- r A - 7 7-7 -- r--\ * Sample Mean KSD Mean RSD Mean RSD Mean RSD Mean RG Mean RSD Mean RSD 1 15 4.7 92 4.1 400 2.6 300 3 .5 190 9.6 38 5.7 1030 2.8 2 71 3.6 860 2.3 880 1.1 1170 1.4 490 3.4 260 2.4 4430 2.8 3 25 2.8 560 1.7 670 1.4 430 0.7 380 3.4 940 0.8 3430 3.8 Cadmium Chromium Copper Lead Manganese Nickel Zinc Soils Seven soils (A-G) which included both uncontaminated (samples E and F) and metal-contaminated soils were each analysed five times and the results are reported in Table IV. RSD values again varied according to element determined and soil sample and tended to be better for copper (mean 2.2) and zinc (mean 2.6) than for the other elements but over-all averaged 3.9%. The results of analysis of the residues remaining after extraction and of the original soils by d.c. arc emission spectrography are reported in Table V. The aqua regia extractable values and the contents of the residues adjusted for the losses in mass during digestion should sum to the total content of the soil within the limits of error of the semi-quantitative method used which is generally so and indicates the degree of retention of metals by the siliceous residues.Expressing the aqua regia extractable values as percentages of the total contents shows that on average some 70% of the nickel 80% of the lead and 90% of the chromium copper and manganese have been extracted 282 BERROW AND STEIN EXTRACTION OF METALS FROM SOILS AIZfllySt VOl. 108 TABLE IV MEAN AND RELATIVE STANDARD DEVIATION VALUES FOR FIVE ANALYSES OF'SEVEN SOILS USING AQUA REGIA REFLUX DIGESTION Mcan exprcssed as milligrams per kilogram in dry soil and relative standard dcviation (IC3D) in :&.Cadrniiirn Chromium Copper r-h_- 7-7 r---Soil Mcan I S D Mean RSD Mean KSD A . . 6.8 3.4 94 1.1 i 1 1 . 3 I3 . rb.0 3.x xx 2.7 61 0.9 C . . <l.i - 40 X 5 24 1.5 D < 1 . 7 - 34 .;.!I 21 2.3 E <1.T - 3s 3.3 31 :i-b F . . <1.i - .50 1.2 26 3.> G . . 29 5.2 240 2.4 230 2.6 Manganrsc r--7 3lean KSD ::lo 3.x :I00 4.i 320 :;.9 310 6.3 440 .-).:: 470 2.s $40 2.7 Canadian Reference soils Because of the poor limits of detection for cadmium and zinc obtained using emission spectrography it has not been possible in the analyses reported so far to determine the amounts of these elements in either the soils or their aqua regia insoluble residues. Data on the effici-ency of extraction of these two elements have been obtained however by analysis of the Canadian Reference soils for which there are established recommended va1~es.l~ Table VI reports the mean RSD recommended and percentage solubility values for the four soils.The soils are as follows SO-1 a clay soil C-horizon SO-2 a podzol R-horizon SO-3 a calcareous soil and SO-4 a chernozem A-horizon. The recommended cadmium values are over-all means after rejection of outliers and are not yet certified values because of analytical variability. TABLE V METAL CONTENTS OF SOILS EXTRACTED BY REFLUXING WITH AQUA REGIA 19 ACID-INSOLVBLE RESIDUES AND 1s UNDIGESTED SOIL Contents expressed as milligrams per kilogram in dry soil. Soil content A-AR* AIR? Total: AR A1 R Total AR AIR Total AR AIR Total B-G-D-- 4 - - AR AIR Total AR AIR Total AR AIR Total F-G-Chromium 94 23 80 88 21 100 40 8 30 34 8 35 58 20 60 50 20 70 240 27 160 Copper 71 6 70 61 5 60 24 3 25 21 3 25 31 4 30 26 3 31 230 5 300 Lea?l 69 3 80 57 3 80 27 3 30 26 3 3 0 20 3 30 27 2 35 1250 5 1400 Manganese 310 37 350 300 50 350 320 35 400 310 35 400 440 40 600 470 50 500 840 60 S O 0 Nickel 30 <9 40 34 <8 40 16 <8 30 13 <x "5 .> 7 13 40 57 17 45 94 13 125 * AR aqua regia.t AIR acid-insoluble residues. $ Total undigested soil February 1983 AND SEWAGE SLUDGES BY REFLUXING WITH AQUA REGIA 283 The values are therefore reported in parentheses as is the value for nickel in SO-2.All values for cadmium are at the lower end of the range for cadmium in uncontaminated soils which ranges from about 0.02-2 mean 0.6 mg kg-l.15 The cadmium concentrations in the aqua regia digests were measured using the improved sensitivity of the atom-trapping technique in flame atomic absorption described by Khalighie et a1.12 The results show that in three of the soils all the zinc and in SO-4 90% of the total cadmium has been extracted. The recom-mended values for chromium copper lead and nickel in SO-2 and SO-3 are also all at the low end of the ranges for these elements in soils15 and the relatively poor RSD values for these elements in soil SO-2 in particular probably reflects this.The proportions of the recom-mended total contents extracted by aqua regia for the nine elements determined expressed as a mean percentage solubility are aluminium 35 cadmium 70 chromium 67 copper 78 iron 85, lead 46 manganese 64 nickel 58 and zinc 90%. The low percentage solubilities for aluminium and lead suggest that the bulk of the generally low lead contents are held in aluminosilicates such as the potassium feldspars which are moderately resistant to acid dissolution. TABLE VI ANALYSIS OF CANADIAN REFERENCE SOILS SO-1-SO-4 USING AQUA REGIA REFLUX DIGESTION Mean and recornmended values (RV) expressed in milligrams per kilogram (except for A1 and Fe yo) and relative standard deviation (1ZSD) and solubility values (Sol) in yo on an oven-dry basis (110 "C) yo.Soil so-1 . . s o - 2 . . SO-3 . . SO-4 . . Aluminium % ,-Ap- 7 Mean RSD R V Sol 4.38 5.0 9.38 47 2.39 2.3 8.07 30 0.99 3.7 3.05 33 1.70 2.6 5.46 31 Copper so-1 . . s o - 2 . . SO-3 . . SO-4 . . Mean RSD RV sol' 56 2.1 61 92 4.3 3.5 7 61 14 1.8 17 82 17 3.3 22 77 Manganese so-1 . . so-2 . . SO-3 . . SO-4 . . Mean RSD RV soi 676 2.1 890 76 189 1.9 720 26 370 3.6 520 71 494 1.3 600 82 Cadmium 7-v Mean RSD 1IV Sol 0.12 33 (0.15) 80 0.11 32 (0.18) 61 0.07 4.2 (0.14) 50 0.38 8.3 (0.42) 90 Iron yo Chromium 7-Mean IiSD IIV Sol 171 3.5 160 107 9.2 13.5 16 58 16.6 3.4 26 64 27 5.0 61 44 Lead r-Mean 5.55 3.58 1.44 2.11 7 Mean 74 <3.4 9.0 17 RSD RV 1.7 6.00 4.1 5.56 1.9 1.51 1.3 2.37 Nickel RSD RV 2.0 94 5.6 2.6 26 -L - V ? 1 Sol 93 64 95 89 7 Sol 79 (28 56 67 Mean RSD RV 9.8 5.8 21 4.2 18 21 8.7 7.2 14 9.0 1.7 16 Zinc r--L Mean lISD RV 157 2.9 146 75 3.9 124 52 1.1 52 98 1.6 94 soi 47 20 62 56 7 Sol 108 60 100 104 Scottish topsoils Analyses of six typical Scottish topsoils (0-25 cm) developed on tills derived from a range of different rock types were also carried out.These are uncontaminated topsoils developed on (1) granite and granitic gneiss (2) sands and gravels (3) basic igneous rocks (4) granite and granitic gneiss (5) andalusite schists and (6) slates and argillaceous schists.The results of five analyses of each soil are presented in Table VII. The total contents of chromium copper, lead and nickel were determined from ten replicate analyses by d.c. arc emission spectroscopy.13 Quantitative analyses were carried out for iron in five replicates of each soil following sodium carbonate fusion and dissolution by the colorimetric method of Scott16 and for aluminium and manganese by lithium metaborate fusions of the soils according to the procedure described by Verbeek et a1.I' The percentages of the total contents extracted by refluxing with aqua regia, expressed as means for all six soils were aluminium 37 chromium 61 copper 90 iron 88 lead '74 manganese 74 and nickel 44. The RSD values obtained ranged from 1.1 to 15.7 but over-all averaged less than 5%.The contents of cadmium were less than the limit of detection using conventional flame atomic-absorption analysis and these were determined using the atom-trapping technique referred to earlier. RSD values were often poorest for the soils with lowes 284 BERROW AND STEIN EXTRACTION OF METALS FROM SOILS Analyst Vol. 108 metal contents for example cadmium chromium copper iron lead and nickel in sands and gravels as might be expected. The RSD values were also generally lower for copper iron and zinc than for the other elements. The mean percentage solubility values are close to those obtained for the Canadian Reference soils except for lead for which the solubility is consider-ably greater in the Scottish soils. This is probably related to the fact that the Scottish soils are all topsoils and contain relatively larger amounts of organic matter with which much of the lead is closely associated.The masses of the acid-insoluble residues remaining after refluxing 3 g of each of the 17 soils examined with aqua regia ranged from 1.2 to 2.8 g mean 2.3 g. Most of the residues are pre-sumably silica or silicates because soils generally contain around 60% silicon expressed as 50,. The proportion of aluminium extracted from the soils suggests that about one third of the aluminosilicate components have been broken down. TABLE VII ANALYSIS OF SCOTTISH ARABLE TOPSOILS USIKG AQUA REGIA REFLUX DIGESTION Mean and total values expressed in milligrams per kilograni (except for A1 and Fc '$* and relative standard deviation (RSD) and solubility (Sol) in yo in dry soil.Soil 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 Aluminium % 1 Mean RSD Total Sol 1.51 1.1 7.00 22 1.22 3.2 4.56 27 4.97 2.6 8.02 62 2.02 7.1 6.35 32 2.70 2.7 6.68 40 2.63 3.3 6.58 40 Copper 7-v Mean RSD Total Sol 5.8 5.2 6 97 3.1 11.8 t 5 >62 20 2.9 25 79 20 3.3 20 100 26 2.0 25 104 16 3.6 15 107 Manganese Mean RSD Total soi 287 5.4 410 70 138 4.5 175 79 837 15.7 1650 51 495 7.3 690 72 215 7.3 260 83 585 7.8 570 103 Cadmium Chromium Mean RSD Total So; 0.22 5.2 - -0.072 20 -0.28 9.6 - -0.26 10 - -0.39 4.6 - -0.28 6.2 - -Mean 1tSD Total 1.96 1.1 2.21 0.96 2.9 1.02 6.56 1.1 8.66 2.50 2.5 2.85 2.57 2.1 2.73 3.38 2.9 3.84 Nickel -A_-Mean RSD Total 4.9 12.8 15 3.8 13.3 20 20 6.3 35 12 3.2 25 26 2.7 40 14 3.7 30 soi 89 94 76 88 94 88 7 Sol 33 19 56 47 64 46 Mean HSD Total soi 18 7.4 20 90 12 9.1 25 48 56 6.4 80 70 32 4.9 60 53 60 3.4 100 60 35 2.0 80 44 Lead LTEFGGE2 15 6.6 20 74 8.4 9.4 20 42 12 5.0 15 83 44 3.2 50 89 24 4.8 40 61 19 2.1 20 97 zinc --Ap- 7 Mean RSD Total Sol 41.6 2.2 -16.3 1.7 - .121 3.4 - -62 1.2 - -77 3.9 - - - -73 5.0 - -Discussion Some metals in sewage sludges are present in a highly soluble form. I t was found on average that about 50% of the total amounts of manganese nickel and zinc in 42 sludges from England and Wales was soluble in a mild extractant such as 2.5% acetic acid.2 I t has since been found that cadmium chromium copper lead nickel and zinc in typical sewage sludges can be efficiently extracted by refluxing in a glass tube with nitric acid,ls and the results presented in Table I11 are consistent with these findings.Digested sewage sludges contain on average about 50% organic matter and when applied to agricultural land much of this decomposes and the associated metals change their forms.lg Most of the metals added as sludge whether in organic or inorganic combination in the soil are extractable by boiling aqua regia. The analyses of sludge-treated soils reported in Table V show that 80% or more of the chromium copper lead and manganese are dissolved and thi February 1983 AND SEWAGE SLUDGES BY REFLUXING WITH AQUA REGIA 285 will also apply to cadmium iron and zinc as illustrated by the analysis of the Canadian Reference soils.Metals contained in the crystal lattices of unweathered primary minerals will not be as soluble and their dissolution will depend to some extent on the fineness of grinding of the soil. This is reflected in the lower proportions of the total contents of some metals extractable by boiling aqua regia in the uncontaminated Canadian and Scottish soils. For total decomposi-tion of silicates the use of hydrofluoric acid is essential as has been shown in the determination of metals in aquatic sediments by Agemian and Chau.20 These workers have found that the extracting efficiency for ten metals was in the decreasing order hydrofluoric - perchloric - nitric acid mixture boiling perchloric - nitric acid mixture boiling aqua regia and boiling nitric acid followed by several less vigorous extractants.Boiling aqua regia extracted 70% or more of the true total contents of copper iron lead and zinc from a lake sediment but extraction of cadmium manganese and nickel was relatively poor. For the purposes of determining the total metal contents of soils in order to decide upon permissible additions of sludge to land refluxing with aqua regia is a simple and efficient procedure. I t is suited to routine operation and does not involve the expense of acid digestion bombs or the special apparatus and facilities necessary for the use of the more vigorous hydro-fluoric or perchloric acids. Further the use of these hazardous reagents is avoided and the conventional flame atomic-absorption equipment required is widely available in Water Authority and soil analysis laboratories.In determining total contents of metals in soil for geochemical prospecting purposes however or relating soil contents to those of their parent rocks an analysis for true total contents may be necessary. We are grateful to C. M. Lau for the determination of cadmium in some of the soil digest solutions using the atom-trapping technique. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. References “Water Industry Review,” National Water Council London 1978. Rerrow M. I,. and Webber J . J. Sci. Food Agric. 1972 23 93. Patterson J . n. E. “Trace L<lements in Soils and Crops,” MAFF Tech.Bull. No. 21 HM Stationery Office London 1971 p. 193. “Application of Sewage Sludge to Cropland 44ppraisal of l’otential Hazards of the Heavy Metals to Plants and Animals,” Council for Agricultural Science and Technology Ames IA USA 1978, Report No. 64. “Report of the Working Party on the Disposal of Sewage Sludge to Land,” Department of the Environment/National Water Council London Report No. 5 1977. “Effects of Sewage Sludge on the Cadmium and Zinc Content of Crops,” Council for Agricultural Science and Technology Ames IA USA 1980 Report No. 83. Brotzen O. Kvalheim A. and Marmo V. in Kvalheim A. Editor “Geochemical Prospecting in Fennoscandia,” Interscience New York 1967 Chapter 8 p. 99. Brooks R. R. “Geobotany and Biogeochemistry in Mineral Exploration,” Harper and Row New York 1972. Rerrow M. L. and Rurridge J. C. Proc. Int. Conf. Management and Control of Heavy Metals in the Environment London CEP Consultants Ltd. Edinburgh 1979 p. 304. Hesse P. R. “A Textbook of Soil Chemical Analysis,” Murray London 1971. Gorsuch T. T. Analyst 1959 84 135. Khalighie J. Ure A. M. and West T. S. Anal. Chim. Acta 1981 131 27. Mitchell R. L. “The Spcctrochemical Analysis of Soils Plants and Related Materials,” Tech. Comm. No. 44A Commonwealth Bureau of Soils Harpenden Commonwealth Agricultural Bureau, Farnham Royal 1964. Bowman W. S. Faye G. H. Sutarno R. McKeague J. A. and Kodama H. Geostand. Newsl., 1979 3 109. Ure A. M. and Berrow M. L. in Bowen H. J. M. Editor “Environmental Chemistry,” Volume 2, Chapter 3 Royal Society of Chemistry London 1982 p. 94. Scott R. O. Analyst 1941 66 142. Verbeek A. A. Mitchell M. C. and Ure A. M. Anal. Chim. Acta 1982 135 215. Thompson K. C. and Wagstaff K. Analyst 1980 105 883. Berrow M. L. and Rurridge J. C. in “Inorganic Pollution and Agriculture,” MAFF/ADAS Ref. Agemian H. and Chau A. S. Y. Analyst 1976 101 761. Book No. 326 HM Stationery Office London 1980 p. 159. Received October llth 1982. Accepted November 24th 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800277

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

286 Analyst February 1983 Vol. 108 $9. 286-292 Application of Optical Emission Source Developments in Metallurgical Analysis Hugh Hughes Consultant 7 Rowan Avenue Guisborough Cleveland TS14 8DR Recent developments in metallurgical analysis which have potential for use in industry are described. The paper concentrates on three techniques firstly the use of pulse distribution analysis in spark-source optical emission spectro-scopy to determine soluble aluminium in steel separately from the aluminium that is present as inclusions in the steel matrix; secondly glow discharge spectroiiietry for the determination of alloying elements in steel a technique which has potential for both bulk and surface analysis of steel and lastly, inductively coupled plasma spectrometry. Data are presented together with reproducibility measurements which give the optimum conditions for niultielenient analysis of stccl using this technique.Keywords Pulse distvihution analysis ; spark-source optical emission spectro-scopy ; glow dischavge lamp ; inductively coupled plasma spectroiizetry ; metallurgical analysis Ohls in his paper on new aspects of atomic-spectroscopy presented at this Symposium made it clear that tlie application of optical emission spectroscopy in the metallurgical industry has been given new impetus in recent years by new developments in sources and their application. This paper is intended to expand on this in certain areas concentrating specifically on three subjects the use of pulse distribution analysis in spark-source optical emission spectroscopy; the use of tlie glow discharge lamp for both bulk and surface analysis of steel ; and tlie potential of inductively coupled plasma spectrometry in the metallurgical industry.Pulse Distribution Analysis This technique has been developed by the Japanese mainly to determine soluble aluminium in steel separately from the aluminium that is present as inclusions in the steel matrix, referred to as insoluble aluminium. In the past only the total aluminium content could be determined by optical emission. However a rapid determination of the soluble aluminium content is essential during production in order to adjust the content to the levels required by the specification being produced. Onodera ct a1.l first described the development at Nippon Insol.Al Discharge spots Discharge spot Including insol. A1 Excluding insol. Al Fig. 1 . Principle of pulse height analysis inethod HUGHES 287 Steel and Figs. 1 and 2 taken from their work describe the principle of the method. Briefly, the sizes of the individual light pulses from individual sparks vary depending on whether inclusions in the steel (insoluble aluminium) contribute light intensity to particular sparks (Fig. 1). As illustrated in Fig. 2 the analysis of the pulse-height distribution of the light I sol. AI Insoi. AI I r h-4 Light intensity Soluble Al chemical Insoluble Al chemical Pig. 2. Principle of pulsc height analysis method. Im Distribution mcdian; N total number of pulses. intensity from individual sparks separates the contribution of insoluble and soluble aluminium and allows the conversion of the data into their respective contents.There was much interest in the British steel industry when the method was first described and samples were submitted for analysis. killings from the same samples were at tlie same time chemically analysed by several laboratories. Fig. 3 illustrates the comparison of chemical and spectro-0.14 1 s * 0.12 .-L c 0.10 2 0.08 2 0.06 2 0.04 a -< 0.02 -0 cn ’ / 0 0.02 0.040.060.080.100.120.1~ Soluble Al (chemical) O/O Fig. 3 ilnalysis of UK steels using J ~ p ~ n e i v calibration (pulse height analyis method). 5000 v) 4- .- 5 4000 2 3000 + .-2000 c .-v) C a, E 1000 -0 5.0 10.0 15.0 20.0 25.0 Co ba It concentration O/O 1;ig.4 . Glow discharge for cobalt (345.3 nm) I3inary steel calibration with corrccted intensities. metric soluble aluminium contents. The agreement i\ riot ,atisfactory particularly at tlie higher contents. Furthcr investigation showed that the particle size of the inclusions in the British samples \\as much coarser than that in the calibration samples. This u.ould seem to indicate tliat the sanipling in particular the cooling rate of samples for anal>vsis must be matched very closely to that of the calibration samples as this was thc only difference between the t\vo 5ets of samples. The Institut de Recherches de la Siderurgie Franqaise in Metz 1s at present investigating the topic further as part of the European Coal and Steel Community (ECSC) research programme and this should clarify the position prior to its more general introduction in European steelworks.Glow Discharge Spectrometry The glow discharge lanip wa\ first developed by Grinsm2 and the principle of its operation has been described b c f o r ~ . ~ . ~ In the case of bulk analysis Kutterworth7 illustrated the simplicity of the technique for alloy steel analyiis compared with .park-iource optical einiiiion when glow discharge conditions are properly selected. Fig. 4 illustrates a basic calibration graph for cobalt based on binary I t has potential for both bulk5 and surface analysis. 288 HUGHES APPLICATION OF OPTICAL EMISSION SOURCE AnaZyst VoZ. 108 TABLE I GLOW DISCHARGE DETERMINATION OF COBALT IN HIGH-SPEED STEELS Certificate Sample value % SS 481 * .. . 0.21 SS 482 0.24 ss 483 1.94 SS 484 . . 10.20 ss 485 5.06 SS 486 0.13 Glow discharge value % 0.20 0.25 2.00 10.30 5.10 0.11 95% Confidence limit 20 0.01 0.01 0.06 0.14 0.09 0.01 standards and Table I shows the results obtained for the determination of cobalt in high-speed steels using this calibration graph. Compared with spark-source optical emission where complex procedures are used to correct for interference effects a simple procedure is used. This is based on correcting the light intensities in the steels to be analysed by the factors with which the iron line intensities are enhanced or reduced in the steels compared with the line intensities for the same iron contents in the binary calibration standards.Apart from the application by Radmacher and de S ~ a r d t ~ however it still remains for the lamp to be widely introduced into the steel industry for routine bulk analysis production control. This probably reflects a degree of conservation on the part of the steelwork’s chemist. A parallel can be drawn from the relatively slow introduction of X-ray fluorescence into the industry because of the entrenched position of spark-source optical emission X-ray fluorescence is now however fully established and integrated into the analytical scheme. Much more interest has latterly been shown in the application of the glow discharge lamp to surface analysis. The work of Berneron and Charbonnier6 and Le Roy8 typifies application to steel surface problems particularly Le Roy’s work on the contamination of strip steel surfaces.Work by Jowitt and Hughes at the Teesside Laboratories of the Hritish Steel Corporation (BSC) has confirmed the potential of the glow discharge lamp for application in another area namely the problem of de-carburisation of some sheet and round steel products during their production. I t is important to be able to measure this to ensure that product performance is not affected by excessive de-carburisation. At present de-carburisation is determined by a metallographic method that is labour intensive and highly subjective in the operator’s judgement. In this work samples were examined using an RSV glow dis-charge lamp. The operating conditions selected lOOOV and 140mA gave a surface removal rate of 4 pm min-l. This is satis- Fig.5 shows the profile of the crater formed. Fig. 5. Profile of glow discharge crater. 0 5 10 15 Depth of penetrationlym Carbon profile of standard steel. Fig. 6. factorily uniform with some build-up of debris around the edges. Fig. 6 shows a continuous profile of the carbon content of a standard steel which is not de-carburised and Fig. 7 illustrates the profile obtained in the case of a production sample of sheet steel. This shows a de-carburised layer extending approximately 15 pm into the steel. Fig. 8 illustrates a calibration graph for carbon using BCS standards. This is satisfactory even though the 193.1 nm line which is used because of spectrometer limitations is not the best carbon lin February 1983 DEVELOPMENTS I N METALLURGICAL ANALYSIS 289 Depth of penetrationlym Fig.7. Carbon profile of de-car-burised steel. I I 0.2 0.4 0.6 0.8 1.0 1.2 0 Carbon O/O Fig. 8. Glow discharge calibration for carbon in low-alloy steel. for glow discharge analysis owing to its high background The profile of the de-carburisa-tion can therefore be quantified as well as the depth. The glow discharge lamp based on the RSV design is now available from instrument manufacturers such as Philips and ARI,. I t now remains for it to become established and accepted as a tool in metallurgical analysis, both for the surface analysis application just described and in bulk analysis for production control. Inductively Coupled Plasma Spectrometry With the general availability of commercial instruments the metallurgical industry has shown much interest in plasma sources.The steel industry in fact illustrated this interest by the funding by the European Coal and Steel Community of a major project to investigate the potential of the inductively coupled plasma. German French Belgian and British laboratories participated in this joint project. My former laboratory at Teesside was responsible for investigating the performance of a high power plasma using a Kontron system capable of an output of up to 10 kW at the induction coil. One of the objectives was to optimise operating conditions to allow simultaneous analysis using an E 1000 dual grating direct -reading spectrometer . Table I1 summarises the variations in operating parameters which were used these being TABLE I1 VARIATION OF PLASMA OPERATING PARAMETERS Power .. . . Fixed 10 k W Coolant gas (N,) . . . . Varied 19.5-33 1 min-l Plasma gas (Ar) . . . . . . . . Varied 10.5-17 1 min-l Sample uptake rate (carrier gas Ar) . . Varied 1.9-4 nil min-l 2.4-4.5 1 min-' Observation height . . . . Varied 2-26 mm (above the coil) varied systematically but independently. Fig. 9 illustrates a set of data obtained with fixed coolant gas and uptake rate (24 1 min-l and 2.8 ml min-l respectively). Line to back-ground ratios are plotted against observation height for four plasma gas flow-rates. Fig. 10 is a similar figure with variable coolant gas rate for fixed plasma gas and uptake rate (17 1 min-l and 3.6 ml min-1). These data together with reproducibility measurements, gave the optimum conditions for multi-element analysis.These were as follows coolant gas flow-rate 33 1 min-l; plasma gas flow-rate 15 1 min-1; sample uptake rate 3.5 ml min-l; and observation height 12 mm above the coil. A compromise observation height has to be used in simultaneous multi-element determina-tion but Table I11 shows that such a compromise does not seriously reduce performance this being on average over 90% of best performance. Using these conditions multi-element solutions were used to determine the linear ranges of calibration graphs the usable ranges of calibration graphs allowing some curvature within the limits of detection. Table IV illustrates a selection of the results obtained and Fig. 11 the calibration graph obtained for aluminium 290 HUGHES APPLICATION OF OPTICAL EMISSION SOURCE A~zaZyst VoZ.108 I I I I I 20 16 12 8 4 5 20 16 12 8 4 20 16 12 8 4 20 16 12 8 4 Height/mm above coil Fig. 9. Characteristics of plasma with variable plasma gas flow-rate (a) Coolant gas 24 1 min-'; uptake 10.5; (b) 12.5; (c) 15.0; and (d) 17.0 1 min-l. rate 2.8 ml min-l. TABLE I11 LINE TO BACKGROUND RATIOS AT COMPROMISE HEIGHT (12 mm) AS A PERCENTAGE OF THE VALUE AT OPTIMUM HEIGHT Optimum A1 12 0 100 As . . 12 0 100 Ce . . 20 8 83 Mo . . 14 2 99 V . . . . 10 -2 97 Zn . . 10 -2 55 Element heightlmm AH February 1983 DEVELOPMENTS I N METALLURGICAL ANALYSIS TABLE IV TYPICAL DETECTION LIMITS AND LINEAR AND USABLE DYNAMIC RANGES Element pg ml-l rangelpg ml-1 range/ pg ml-l Detection limit/ Linear dynamic Usable dynamic A1 .. 0.2 Cd . . 0.02 Fe . . 0.01 Mn . . 0.002 Ni . . 0.1 V 0.05 2 x 103 >6 x lo3 5 x 103 105 104 4 x 105 2 x 104 >2 x 106 104 5 x 104 > 104 > 104 E-0 5 20 16 12 8 4 - 20 16 12 8 4 I I I 20 16 12 8 4 20 16 12 8 4 Height/mm above coil 3500 F . 3000 } J 2 2500 -F *= 2000 e P 1000 --5 1500 -w .-+. 500 - -I I 0 2000 4000 Concentration/Fg m I- ' Fig. 11. Aluminium cali-bration graph. t 291 Fig. 10. Characteristics of plasma with variable coolant gas flow-rate (a) 19.5; (b) 24.0; (c) 28.5; and (d) 33.0 1 min-'. Plasma gas 17 1 min-l; uptake rate, 3.6 ml min-l. These illustrate a capability to determine from micrograms per millilitre to percentage contents. Limits of detection for elements of interest in steel are listed in Table V and have been supplemented by values obtained using a Philips P V 8210 and data reported by Wagner and Petins using an ARL 34000.The results are not directly comparable the first two are for dilute acid or water solutions but the data of Wagner and Petin are for solutions con-taining 1 g of steel per 100 ml. If 1 0 times the limit of detection is taken as the lowest quantitative determinable concentration (LQDC) and a dissolution procedure based on 1 g of steel in 100 ml is used the LQDC in percentage steel content is one tenth of the numerica 292 Element A1 . . As . . . . co . . . . Cr . . c u . . . . Mn . . Mo . . . . Ni . . . . P Si . . V 7 h/nm 308.2 193.7 228.6 267.7 327.4 257.6 287.1 231.6 178.2 251.6 311.0 HUGHES TABLE V LIMITS OF DETECTION FOR ELEMENTS IN STEEL E 1 ooo* -T-+--7 Limit of detection/ pg ml-l 0.2 0.3 0.1 0.02 0.1 0.002 0.1 0.1 1.5 0.09 0.05 7-h/nm 308.2 193.7 228.6 267.7 324.7 257.6 202.0 341.4 214.9 251.6 290.8 PV82 1 O* 7 Limit of detection/ pg ml-l 0.04 0.007 0.003 0.002 0.002 0.0005 0.005 0.015 0.04 0.01 0.003 7-A/nm 396.1 197.3 345.3 267.7 327.4 257.6 281.6 341.4 178.2 288.1 311.0 340007 - - i 7 Limit of detection/ pg ml-l 0.013 0.15 0.02 0.019 0.004 0.006 0.016 0.01 0.12 0.13 0.015 Dilute solutions.t 1 g steel per 100 ml solutions. value of the limit of detection listed.With the first two sets o f data some allowance must be made for the effect of iron in solution on the limits of detection but for most purposes the technique can cover steel analysis requirements. The same characteristics namely good limits of detection and wide dynamic ranges make the technique suitable for a wide range of other applications in the metallurgical industry such as oxide and non-ferrous alloy analysis. 1. 2. 3. 4. 5. 6. 7. 8. 9. References Onodera M. Mishizaka M. Saeki M. Salrata T. hrippon Steel Tech. Rep. Ovrvsras 1!)77 No. 9, Grimni W. Spectvochim. Acta Pavt €3 1968 23 443. Boumans 1’. W. J. M. A n a l . Chern. 1972 44 1219. Hughes H. “Proceedings of the 29th Chemists conference Scarborough,” HISPA Lo~ldoa 1976, Radmacher H. W. and de Swardt M. C. Spectvochim. Acta Pavt H 1975 30 353. Berneron Ii. and Charbonnier J . C. A n a l . Proc. 1980 17 488. Butterworth A. ESC Open Report T/CS/552/3/78/C. Le Roy V. “Proceedings of the 34th Chemists Conference Scarborough,” BSC Teesside Laboratories, Wagner A. and Petin J. “Proceedings of CETAS conference Salzburg,” ClIM Liege Belgium, 73-77. pp. 30-38. Grangetown Middlesbrough 1983. 1979. Received August 26th 1982 Accepted October 27th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800286

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst February 1983 Book Reviews 293 AN INTRODUCTION TO ATOMIC ABSORPTION SPECTROSCOPY. A SELF-TEACHING APPROACH. By L. EBDON. Pp. xiv + 138. Heyden. 1982. Price k9. ISBN 0 85501 714 7. The success of his self-teaching course on atomic-absorption spectroscopy has encouraged Dr. Ebdon to publish an introductory text for those new to this important and widely used technique. The primary aim of the “course” is “to introduce and develop a knowledge of the theory instru-mentation and practice of atomic absorption emission and fluorescence spectroscopy in flame and electrothermal atom cells.” The main emphasis of the book is on flame spectroscopy with major sections on the nature of flames and flame atomisation interferences and errors in flame spec-troscopy and advice on the general practice of flame spectroscopy.Although most chapters deal with atomic absorption (hence the title of the book) Dr. Ebdon rightly introduces the reader to atomic-emission and atomic-fluorescence spectroscopy and explains the relative merits of each technique. The brief but informative sections on emission sources and emission theory will help those learning about AAS to follow developments in other complementary techniques such as ICP - OES. The chapters on modified flanie cells mercury determination and hydride generation are short and merely describe the concepts involved. A rather longer (and more important) chapter provides up-to-date information on electrothermal atomisation. Throughout the text Dr. Ebdon provides keywords to assist the self-learning process and questions are interspersed between sections to test understanding.In addition twenty revision questions and three practical exercises are given at the end of the book together with six definitive experiments on key aspects of the course. The author also provides an appendix on safety pro-cedures an important topic to cover in an introductory programme. Dr. Ebdon specifically states that his book is intended to be read in conjunction with supple-mentary texts. More detailed information could have been given in some sections particularly those on applications. However this is a minor criticism and satisfactory reference is made to other sources of knowledge where appropriate. In general this is an excellent and much needed book on the basic concepts of analytical atomic spectroscopy.Dr. Ebdon’s contribution should find extensive use in many undergraduate and postgraduate courses and will be of great value to those seeking an easily read but authoritative introduction to the technique. It is therefore not a replacement for existing publications on the subject. L). LITTLEJOHN METAL VAPOURS I N FLAMES. c. T H . J. ALKEMAIIE TJ. HOLLANDER w. SNELLEMAN and 1’. J. TH. ZEEGEKS. 1033. Pergamon Press. 1982. Price L50; $95. ISBN 0 08 018061 2. International Series in Natural Philosophy Volume 103. Pp. xxii + Perhaps no book in the history of spectroscopy has been written with such breadth and depth as this book; a t least this reviewer is aware of no other. In addition i t would be dishonest of this reviewer to imply that he has read the entire book.However various sections of this book have been read in depth and others only in passing. This book is concerned with the properties ancl behaviour of metal atoms and their compounds in the vapour phase of laboratory combustion flames. The book has considerable use not only for those interested in combustion research and diagnostics but also those interested in analytical flame spectroscopy. Of course this approach by Alkemade and his group has always been of great personal satisfaction to this reviewer. Alkemade et al. cover all of the basic processes in flames including atomisation excitation, dissociation diffusion and radiation. The authors bring together detailed discussions of principles of phenomena including equilibria kinetics radiational processes (absorption luminescence and emission) broadening processes intensities of racliative transitions (including the effect of laser excitation) the structure of flames the calculation of flame equilibrium composition and adiabatic temperatures deviations from flame equilibrium physical processes leading to excitation and de-excitation chemical processes leading to excitation and cle-excitation non-thermal radiation, ionisation in flames (seedecl and not seeded) and diffusion.Alkemade et al. discuss not only the concepts and principles of the above but also the experimental measurement of parameters such a 294 BOOK REVIEWS Analyst Vol. 108 flame temperatures dissociation energies ionisation energies quenching rate constants rate con-stan ts for dissociation ancl ionisation diffusion coefficients line profiles antl broadening and shift parameters and of course concentrations.An excellent discussion of instrumentation (burners, nebulisers gas flow the optical train photodetectors noise the measurement system as well as light sources for AAS and AFS) and the introduction of aerosols into flames is given. Much of the book is applicable to non-flame systems such as furnaces microwave plasmas liF plasmas and areas. The book is incredibly well referenced and written for both the physicist interested in the physics of flames and for the applied scientist interested in using flames for analytical purposes. This book is a nzust for every analytical spectroscopist physical spectroscopist and engineer interested in flames.I t is also a book that should become a classic such as Mitchell and Zemansky’s “lieson-ance Radiation and Excited Atoms,” because i t will be in niany ways completely dateless. I t is a book in which the authors should have great pride and for which this reviewer is thankful. J. I). WINEFORDNER THE THEORY OF VIBRATIONAL SPECTROSCOPY AND ITS APPLICATIONS TO POLYMERIC R~ATERIALS. Pp. s + 530. John By PAUL C. PAINTER MICHAEL M. COLEMAN and JACK L. KOKNIG. Wley. 1982. Price k44.50. ISBN 0 471 09346 7. %‘hen as in the present instance authors of books provide a clear statement of their aims and intentions reviewers have a yarclstick by whicli to assess their efforts. In their preface Professor Koenig and his former students Dr. Coleman and llr. Painter who have established reputations as vibrational spectroscopists in their own right state that wliat is needed is a monograph that contains all the tools essential for understantling the vibrational spectra of polymers.Their boldness in suggesting they believe that they have provided that compendinni is not misplaced as this book surely is destined to become a classic tlie bible of serious wn-kers in the field. l’he first nine chapters accounting for about 40° of the total provide a systematic ancl lucid account of the theory of vibrational spectroscopy that will also be of considerable value to anyone not specifically concerned with polymers who wishes to become acquainted with the subject. The subject is developed from first principles and tlie authors then move on to internal co-ordinates symmetry syninietry co-ordinates and niolecular force fields.I<acli stage is illustratetl by reference to specific esamples such as tlie vibration of the carbon clioside molecule i n one and two dimensions water and the niolecular type AU for tlie determination of tlie nuniber of active modes and SH to illustrate symmetry co-ordinates. These various concepts are brought together in a chapter devoted to tlie vibrational analysis of the benzene molecule. The background material concludes with an account of the factors that determine infrared and liaman intensities, with SO as an example. The application of tliese techniques to polymeric materials begins with a chapter on lattice dynamics continues with one on the symmetry analysis of polymer chains antl is folloivetl by a third on the normal co-orclinate analysis of isolated polymer chains.’l‘liis paves the way for an account of the normal vibrational analysis of polyethylene in the single chain approsiniation. ’ h e vibrational analysis of polymer crystals is then considered largely by reference to poly-ethylene but transition dipole coupling in polypeptides is also discussed. The influence of defects arid disorder is then outlined sufficientlJ- for the serious reader to be able to follow the detailed publications in this rather involved field. l’he book concludes with a substantial chapter of 155 pages in which selected but \vide-ranging examples are given of the application of normal co-ordinate anal)-sis to polymers. ’I’hese esaniples include polypropylene poly(butene-1) poly(viriy1 chloride) poly(vin~-lidene cliloride) poly-(vinylidene fluoride) polytetrafluoroetliylene polybutacliene E~oIycliloroI)rene polyisoprene, polystyrene and poly(etliy1ene tereplitlialate) .‘fliere are also sections on polyaniicles pol?-peptides and proteins in which polyglycine is considered in some detail. I t is a pity to haye to criticise such a meritorious volume but tlie index falls bclo\v the high standard of the test in that major subject headings not referenced include internal co-ordinates February 1983 BOOK REVIEWS 295 polybuta-l,2-diene and polytetrafluoroethylene. Finally the authors may lay fair claim to the title of polymaths by virtue of the wide range of apt literary quotations that grace their intro-ductory comments and the beginnings of the chapters.W. I?. MAUUAMS RECENT ADVANCES IN ANALYTICAL CHEMISTRY. PROCEEDINGS OF A ROYAL SOCIETY I~SCUSSION MEETING HELD ON ~ T H AND ~ O T H DECEMBER 1981. Edited by J. M. THOMAS R. BHLCHER and T. S. WEST. Pp. vi + 219. The Royal Society. 1982. Price k28.30. ISBN 0 85403 191 x. This volume reprinted from Phil. Trans. R. SOG. London Ser. A 1982 305 469-689 is doubly welcome in its own right and as it marks a renaissance of interest in analytical chemistry by the Royal Society. Many of its early Fellows from Robert Boyle on were keenly interested in the techniques and results of analytical chemistry but regrettably this desirable state did not continue into more recent times. Prof. J. M. Thomas in his courteous “Introduction,” gives a perceptive view of analytical chemistry as the grammar of experimental science and notes some of the import-ant techniques not dealt with in the meeting such as FABMS and applications e.g.forgeries in the visual arts and in the forensic sciences. “The Role of Analytical Chemistry in Academic and Industrial Chemistry” is discussed by Prof. R. Belcher in what was his last public lecture. This broad review contains much new information about the historical development of Chairs in the subject particularly in London. Sir Alan Walsh in describing “Atomic Absorption and Atomic Fluorescence Methods of Analysis; their Merits and Limitations,” draws attention to some of the less common modes of atomisation and their application. Highlights of atomic emission by use of an ICP are outlined by 13.M. Barnes under “Recent Advances in Analytical Atomic Radiofrequency Emission Spectroscopy.” Special emphasis is paid by F. Adams to developments in methodology and instrumentation in “Recent Advances in Analytical Spark Source Mass Spectrometry.” The second section of papers concerns aspects of elemental electronic and molecular spectroscopy. In “Electron Energy-loss Spectroscopy for Elemental Analysis ’’ It. F. Egerton outlines the basic physics involved and a range of applications including the measurement of radiation damage in organic compounds. 1). A. Jefferson discussed the applicability of “X-ray Emission Analysis in the Electron Microscope” to the simultaneous structural and compositional investigation including crystals containing fewer than 1010 atoms.Aspects of electron photon and ion excitation are reviewed in detail in “Surface-specific Analytical Techniques” by J . C. liivikre. C. A. Fyfe et al. show by a variety of examples of “Analytical Applications of High-resolution Suclear Magnetic Resonance Spectroscopy of Solids” that the spectra obtained by cross-polar isation - magic-angle spinning produce data in many ways complementary to X-ray diffraction. The review of “Recent Advances i n Analytical Infrared Spectroscopy” by V. C. Farmer was interestingly based on dis-cussion and solutions of the instrumental interpretative and manipulative difficulties of the tech-nique. I<. Kalvoda draws attention to “Recent Advances in Polarographic Analysis” and T. Fujinaga “Recent Ad-vances in Analytical I’otentiometry with Ion-selective Electrodes.” Under the heading “Flow Injection Methods a Sew Tool for Instrumental Analysis,” J. IKldiCka reviews the gradient and solution handling aspects of the flow-injection technique and their applications. J . H. Purnell notes four areas of activity in “The Current Chromatographic Scene,” namely in HPLC solid surface modification open- tu be m icro-colu ni ns and to t a 1 sys tem s optimisation . J . 1C. Moody gives a critical review of work carried out by the SBS on “The Sampling Handling and Storage of Materials for Trace Analysis.” As befits the computer age the final paper is “liecent Advances in Microprocessors for Analytical Chemistry,” in which G. Horlick reviews work to date and outlines the capabilities role seen for and likely impact of “consumer market” computers in analytical instrumentation and laboratories.It is regretted that no record has been included of the poster sessions that supplemented and diversified the content of the over-all meeting. The reviewer looks forward to future Royal Society Discussion Meetings on analytical cheniistry a t the same high standard of presentation and reporting on other areas which time or circumstances excluded. The next three papers relate to analytical atomic spectroscopy. ’ 1 . 1 ~ third section deals with electroanalytical chemistry and chromatography. The book concludes with two papers on sample handling and microprocessors in analysis. This wcrli is an excellently produced and scholarly record of an important meeting.D. THORBURN BURN 296 BOOK REVIEWS Analyst Vol. 108 GENERAL HANDBOOK OF ON-LINE PROCESS ANALYSERS. By D. J. HUSKINS. Ellis Horwood Series in Analytical Chemistry. Pp. 239. Ellis Horwood. 1981. Price k30. ISBN 0 85312 329 2 (Ellis Horwood); 0 470 27292 9 (Halsted Press). This is the first of a series of five books all by the same author on the subject of on-line process analysis and analysers and in some respects it is a little difficult to review a part without seeing the whole of the series. The volume described as a general handbook is intended to serve as a necessary introduction and support to the other titles which are as follows “Quality Measuring Instruments in On-Line Process Analysis,” “Optical Methods in On-Line Process Analysis,” a ‘Electrical and Magnetic Methods in On-Line Process Analysis” and “Gas Chromatographs as On-Line Process Analysers.” The book lists the graphic symbols and abbreviations used in the texts and provides information on the units and symbols appropriate to the subject.A chapter on the concentration of com-ponents in gas mixtures and liquid mixtures or solutions should help to avoid later errors in under-standing or calculation and a separate chapter deals in some detail with the various types of errors calibration correlation and control checks necessary in on-line analysis. Following chapters go in more detail into the calibration or standardisation of gas analysers including the preparation and storage of precision gas mixtures together with supply gas purification. Chapter 8 deals fairly comprehensively with sample handling and conditioning for gas and liquid streams covering such aspects as sampling probes sampling lines sample take-off and the conditions required for turbulent and laminar flow.Various forms of filters separators coalescers scrubbers and strippers are well described together with the important aspects of stream switching and multiple sample systems. This chapter also includes methods of obtaining a sample at low pressures and some descriptions of commercially available gas sampling probes. The importance of correct housing for analysers and such factors as temperature control and ventilation are covered in Chapter 9 together with fire toxic and radioactivity hazard monitoring and alarm systems. Very short chapters deal briefly with signals and recording the use of digital systems and some control problems.In the chapter on maintenance and availability I was pleased to read the useful rule “Make one person responsible for the equipment and thus reduce knob twiddling,” and to see some treatment of the calculation and recording of analyser availa-bility reliability and maintainability. Brief descriptions are given of some 22 applications of on-line process analysers all to gaseous or liquid systems. The final chapter in the book presents annotated contents lists for Books 2-5 cross-referred applications and techniques lists a methods list for some important materials produced in industry and concludes with considerations necessary in calculating the economic justification for installing an analyser. The areas covered within the author’s direct experiences are marked by the completeness of their treatment. Other areas are not covered quite as fully and there are some omissions. For example I could find no mention of methods for automated sampling and transfer of solids or suspensions or of dealing with the important area of effluent sampling. Perhaps these will be covered in the context of specific analyser descriptions in the other volumes. This handbook with its companion volumes should help greatly to clarify the advance thinking and practical actions taken by readers in the design, development purchase commissioning use and maintenance of on-line process analysers in the plants for which they have responsibility. This book is well produced with clear print and diagrams. D. C. M. SQUIRREL

ISSN:0003-2654    

DOI:10.1039/AN9830800293

出版商:RSC    

年代:1983

数据来源: RSC