摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 119 Determination of Ultra-trace Levels of Metal Ions in Sea-water With On-line Pre-concentration and Electrothermal Atomic Absorption Spectrometry* V. Porta 0. Abollino E. Mentasti and C. Sananini Department of Analytical Chemistry University of Torino Via P. Giuria 5 70725 Turin ltaly An on-line pre-concentration system for electrothermal atomic absorption spectrometry was developed. A minia- ture silica C18 column was inserted at the tip of the autosampler arm. A modification of the tubing line of the auto- sampler allowed either the flow of the sample through the column or the operation of the autosampler in the normal mode.The retention of the metal ions in the form of complexes on the microcolumn was achieved by using pyrrolidin-l-yl dithioformate as the complexing agent; acetonitrile was then used for the elution.The direct injection of the eluate into the graphite furnace gave high pre-concentration factors ranging between 20 and 225 which are sufficient for the determination of Cd Pb Cu Ni Co and Fe in Antarctic sea-water. The blank level was very low and the detection limits ranged from 0.4 ng I-' (Cd) to 25 ng I-' (Fe). Keywords Ultra-trace determination enrichment; me fa/; sea- water; electrothermal atomic absorption spectro- metry The determination of metal ions in samples from uncontami- nated areas requires very powerful analytical techniques to detect extremely low amounts of analytes. Very few tech- niques namely electrothermal atomic absorption spectrometry (ETAAS) inductively coupled plasma atomic emission spectrometry (ICP-AES) and ICP mass spectrometry (ICP- MS) have sufficient sensitivity and often the analysis is hin- dered by severe matrix effects.Pre-concentration and separation procedures are the solution to these problem^.^-^^ Off-line pre-concentration systems'- require very large sample volumes (500-1000 ml) since the final volume usually necessary is at least 5 mi of which only a small part is used for analysis by ETAAS. In order to reduce the sample consumption and the possibili- ty of contamination flow injection (on-line) pre-concentration systems have been introd~ced.~-'~ These manifolds have mainly been coupled to flame atomic absorption spectrome- ters ICP atomic emission spectrometers and ICP mass spec- trometers with no attention having been given to graphite furnace (GF) atomic absorption spectrometers.This is prob- ably because GFAA spectrometers have a non-continuous de- tection mode which does not allow the use of apparatus optimized for different spectrometric techniques and requires the development of new manifolds that are more complex than those which work on continuous instruments. One example which is not a true on-line pre-concentration system has been reported in the 1iterat~re.I~ These workers op- timized a pre-concentration system using a microcolumn of silica to which 8-hydroxyquinoline was chemically bound. This system differed from a normal off-line system in having an elution step that was performed directly in an autosampler cup on a final volume of 0.3-1.6 ml.A simple on-line pre- concentration system for ETAAS has been developed by in- serting a small column of solid substrate at the tip of the auto- sampler arm. This assembly allows the autosampler of the spectrometer to be used as part of the pre-concentration appa- ratus. The loading step is performed by a peristaltic pump and the elution and the injection of the eluate into the furnace are performed by the autosampler. Experimental Reagents and Apparatus High-purity water was obtained from a MilliQ purification * Presented in part at the Fifth Biennial National Atomic Spectroscopy Symposium (BNASS) Loughborough. UK 18th-20th July. 1990. system (Millipore) supplied with de-ionized water. All acids and solvents were purified by sub-boiling distillation in a quartz still starting from laboratory-reagent grade chemicals.A concentrated ammonia solution was prepared by isothermal distillation of laboratory-reagent grade ammonia. Silica C (4540 mesh Merck) and Amberlite XAD 2 and XAD 7 (50-100 pm Serva) were used as received. The ligand solutions were prepared from analytical-reagent grade ammo- nium pyrrolidinedithiocarbamate (ammonium pyrrolidin-l-yl dithioformate APDC Merck) and purified on-line see below. The ligand (0.06 g) was dissolved in 250 ml of a 0.2 mol dm-3 solution of ammonium acetate at pH 9.2 in order to preserve the reagent. The desired pH of 4.5 k 0.3 was attained by mixing the ligand solution with the acidified samples. The Antarctic sea-water sample used for the evaluation of the method was collected during the 1989-1990 Italian expedi- tion to Antarctica.The sea-water was sampled in the Ross Sea in the area of the Italian base (Terranova Bay 74" 41' 42" South 164" 07' 23" East). The water was sampled at a depth of 0.5 m and immediately filtered on a 0.45 pm membrane filter. The sample was kept frozen during all transportations and storage in Turin. After melting it was acidified (pH 1.7) with ultrapure HNO,. The atomic absorption spectrometer was a Perkin-Elmer 5 100 with Zeeman-effect background correction. The 5 100 spectrometer was equipped with an HGA 600 graphite furnace and an AS-60 autosampler. The system was driven by a com- puter and fully automated. Pyrolytic graphite coated graphite tubes were used. All instrumental parameters were as recom- mended by the manufacturer Procedure All the sample preparations and manipulations were conducted under a laminar flow clean bench (class 100).The pre-concentration column was inserted at the tip of the autosampler arm (Fig. 1). The column (Fig. 2) was prepared from a piece of Tygon tubing (2.1 mm i.d. 2.0 cm long). The silica C or XAD 2 bed was about 4 mm high corresponding to a total volume of 14 pl. For the XAD 7 a shorter column (3 mm) was used because of the high back pressure observed during the clean-up step. The resin was retained in the column by two plugs of Teflon net (75 pm). The polytetrafluoroethylene (PTFE) tubing (1.5 mm i.d.) and the column were connected by two additional .pieces of Tygon tubing with an 0.d. of 2.0 mm which matched the i.d.of the column and an i.d. of 1.3 mm which fitted with the PTFE tubes. For injection into the furnace the usual capillary of the auto-120 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 199 1 VOL. 6 Purification column Ligand 0- Sample - 3.0 mi min-' - Enrichment column Sampler arm Fig. 1 Schematic diagram of the pre-concentration manifold 3 7 1 1 Fig. 2 Section of the enrichment column. 1 PTFE tube (0.15 mm i.d.); 2 Tygon tube (1.3 mm i.d. 2.0 mm 0.d.); 3 Tygon tube (2.1 mm id.); 4 PTFE; and 5 resin sampler was replaced by a short piece of 0.5 mm i.d. PTFE tubing which reduced the void volume to 5-10 p1 during elution with the organic solvent. All the tubing of the autosampler beyond the sample aspira- tion pump was replaced by PTFE tubing (1.5 mm i.d.) and a four-way PTFE valve (Rheodyne) was introduced into the line to allow switching between the sample and the autosampler lines (Fig.1). The sample was driven by a peristaltic pump (Gilson Minipuls 2) at a flow-rate of 3.0 ml min-' and mixed in a T-junction with the ligand solution (0.8 ml min-I). The pH of the final solution was 4.5 f 0.3 and the concentration of the ligand solution was about 3.5 x lo-' mol dm-3. The ligand so- lution was purified on-line by passing through a column filled with XAD 7. Different pre-concentration times were adopted for the de- termination of each ion from a minimum of 30 s for Fe equiv- alent to 1.5 ml of sample; up to 6 min for Co equivalent to 18 ml of sample. After the loading step the valve was switched and the normal analysis process of the Zeeman 5100 was started. The line was first flushed with 1.4 ml of a solution taken from the reservoir. (The acidified water usually used for the flushing of the autosampler was replaced by a diluted buffer solution of ammonium acetate at pH 4.5.) The washing step is necessary in order to remove any sample still in the line and to clean any sea-water matrix from the resin.After the clean-up step 80 pl of acetonitrile from a sample cup were aspirated into the column stripping the metal complexes from the silica C, and the eluate was injected into the furnace. During the first experi- ments it appeared that one elution cycle was insufficient to recover all the metal ions from the column so a double aspira- tion-injection cycle with an intermediate evaporation step was adopted.The total analysis cycle took only 2-3 min except for very long pre-concentration times since the loading of a sample can be performed during the atomization of the previous sample. A typical temperature-time cycle of the graphite furnace for drying pyrolysing and atomizing a sample contain- ing Cd was 90 "C for 45 s; 600 "C for 10 s; and 1800 "C for 4 s respectively. A calibration graph was obtained by the standard additions method but external calibration graphs could be used as no matrix effect was observed during the analysis. The pre-concentration manifold was easily removed from the autosampler and the normal analysis system restored in less than 10 min. The pre-concentration columns were used for at least 15-20 enrichment cycles before being re-packed al- though the lifetime of the columns is much longer.The packing materials have been described in detail previously.4.13.1H Optimization A widely used chemical method was chosen for the pre- concentration ~ t e p ~ . ' ~ . ' ~ and only the parameters dependent on the pre-concentration apparatus were investigated. However the on-line apparatus described in this paper did not require a lengthy optimization of the parameters. It was found that the pre- concentration efficiency was not sensitive to the parameters by which it is usually affected such as flow-rate and contact time. The uptake by the microcolumn is quantitative for the metal ions investigated namely Cd Pb Ni Cd and Fe with sample flow-rates ranging from 0.5 to 3.0 ml min-I.No experiment was conducted at higher flow-rates. The contact time between the ligand and metal ion which is proportional to the inverse of the flow-rate and to the length of the tube between the T-junction and the column had no effect on the recovery of the metal ions considered. Three different solid substrates were used silica C, XAD 7 (a polyacrylate resin) and XAD 2 (a styrene-p -divinylbenzene copolymer). All three solids while behaving differently gave very good results. The double elution with acetonitrile gave a total re- covery of Cd" from the three substrates of 97 103 and 99% respectively. With a mixture of acids (1.0 mol dm-> HCI + 0.1 mol dm-> HNO,) an 80% yield was obtained using XAD 7 and 27% using XAD 2. This is probably due to the high apolarity of the latter.In addition the precision of the results was found to be much higher with the organic solvent than with the acid eluent. The acid eluent was not used on the silica C, because the bonds between the silica and the alkyl chains are hydro- lysed by the acid with a consequent loss of performance. Some elutions were performed with methanol but this organic solvent proved to be less efficient than acetonitrile. The main drawback of the method was the incomplete release of the complexes with only one aspiration-injection cycle While Cd could be released from XAD 7 and silica C with only one injection of the eluent Cu and Pb required at least two injections (see Table 1). The XAD 2 substrate re- quired a minimum of two successive elutions with acetonitrile for all three reference metal ions.By considering the higher efficiency of the elution step and the precision of the results for the chelating substrate silica C was used for the determination of trace metal ions in sea-water. Results and Discussion Recovery Using the optimized conditions the recovery was almost com- plete for the six metal ions Table 2. The only exception being Table 1 XAD 7 and XAD 2 with one or two successive elutions Recovery of different metal ions in spiked sea-water from C Recovery (%I Metal Number of elutions C X A D 7 X A D 2 Cd 1 90 94 70 2 loo 105 97 Pb I 78 85 52 2 103 99 98 c u I 50 62 58 2 100 9.5 88JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 121 Table 2 obtained from totally independent quadruplicate measurements Recovery of ions present in spiked sea-water; determinations Metal Recovery (%) Cd Pb cu Ni c o Fe 99 f 5 9 2 f 10 1 0 4 f 7 93 f 8 80f 15 106k 12 Co for which an 80% yield was obtained.The recovery yield was calculated from the results of the analysis of spikes added to Antarctic sea-water where the concentration of the analytes is so low that very small absorbance signals were obtained and hence did not contribute significantly to the concentration of spike added. In the analysis of real samples spike concen- trations comparable to the analyte levels expected were used Co 20 ng 1-I; Cd and Pb 50 ng I-'; Cu and Ni 500 ng I-'; and Fe lo00 ng 1-I. Pre-concentration of Mn" and Cr"' was carried out but these elements were not recovered at levels which could be detected by this procedure.An anomalous blank be- haviour was observed for Zn so that evaluation of the efficiency of the method could not be performed. Absolute Blanks The absolute blanks and the detection limits of the method are reported in Table 3. These values refer to pre-concentration of a 3.0 ml aliquot of the sample. By passing the ligand solution through an XAD 7 purification column the blank level was reduced by at least 50% and any problem from indeterminate contamination during the preparation of the ligand solution was avoided. An increase in injection volume did not give a proportional increase in the blank signal. This means that the blank signal was mainly due to the second and third steps of the pre-concentration procedure i.e.the washing and the elution. The determination of the blank level of Cd with a 45 s pre-concentration gave an integrated absorbance value of 16 A s while a simple analysis procedure with the AAS appara- tus without pre-concentration gave a value of 10 A s. As a result the detection limit can be lowered by increasing the sample volume. Analysis of Sea-water The pre-concentration system was used for the analysis of Ant- arctic sea-water. The concentrations found are shown in Table 4. In Fig. 3 the absorbance peaks for Cd" present in the blank sea-water and spiked sea-water are reported. Different volumes of sample were used for the determination of each metal. The pre-concentration factors using an injection volume of 80 yl into the furnace were in the range from 20 (Cd) to 225 (Co).Table 3 Absolute blank values and detection limits (DL) obtained using the proposed method. The values are calculated for a pre-concentration volume of 3.0 ml; determinations obtained from totally independent quad- ruplicate measurements Metal Blank/ Cd 6.0 f 0.4 Pb 1 7 f 4 cu 140f 10 Ni 5 0 f 9 c o 2 * 1 Fe 322 * 25 Pi? DL*/ ng I-' 0.4 4 10 9 1 25 * Expressed as three times the standard deviation of the blank. Table 4 Concentration of some trace metals in Antarctic sea-water; the determinations were obtained from totally independent quadruplicate measurements Metal Concentration/ ng I-' Cd 7.6 f 0.4 Pb 1 8 + 4 cu 148f 16 Ni 230 ? 23 co 3 + 1 Fe 810 f 40 Concentration range*/ ng I-' 14-20 34 125-250 200-490 3 200-427 * Samples collected from the same area in the Antarctic at a different time.0.322 C n Timeis Fig. 3 Antarctic sea-water; and C. spiked Antarctic sea-water (19.6 ng I - ' ) Absorbance peaks after pre-concentration of Cd. A Blank; B Unfortunately data obtained using a different method are not available therefore the accuracy of the method cannot be evalu- ated. However these values are within the same range of con- centration as the values found in samples collected from the same area during other expeditions (Table 4).1h.'Y-22 The precision of the determinations is very high considering that the values are the mean of results of four replicate analy- ses performed independently. The precision of a single analy- sis set was about 4% for Cd 3% for Cu and 5% for Pb which required a long pre-concentration time.The precision is com- parable to that obtained for analytes present at higher concen- trations by using an AA instrument without a pre- concentration step. Conclusions A simple on-line pre-concentration manifold for a graphite furnace atomic absorption spectrometer has been presented which permits the fast and precise determination of metal ions at very low concentrations in water samples. This tech- nique is a simple method giving high pre-concentration factors with low sample consumption. The blank level can be easily controlled and the ratio of the blank to the sample signal can be reduced using long pre-concentration times. Other ligands or solid substrates can be used for the retention of the metal ions from water samples.Work is in progress in this direction. The authors gratefully acknowledge the financial support from C.N.R. (Rome) and Minister0 della Pubblica Istruzione (Rome).122 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 1 2 3 4 5 6 7 8 9 10 11 12 13 References Willie S. N.. Sturgeon R. E.. and Berman S. S. Anal. Chim. Acta 1983 149 59. Nakashima S. Sturgeon R. E. Willie S. 'N. and Berman S. S. Anal. Chim. Acta 1988,207,297. Weisel. C. W. Duce R. A. and Fasching J. L.. Anal. Chem. 1984 56 1050. McLeod C. W. Otsuki A. Okamoto. K. Haraguchi H.. and Fuwa K.. Analyst 1981 106.419. Pai. S. C. Whung. P. Y.. and Lai R. L. Anal. Chim. Acta 1988 211 257. Pai S. C. Anal. Chim. Acta 1988,211,271. Sarzanini C.. Mentasti E. Gennaro M. C. and Marengo E. Anal. Chem.198557 1960. Abollino O. Mentasti E. Porta V. and Sarzanini C. Anal. Chem. 1990,62,2 1. Olsen S. Pessenda L. C. R. RilZiaa J. and Hansen E. H. Analyst 1983,108,905. Fang Z. RiIhEka J. and Hansen E. H. Anal. Chim. Ada 1984,164 23. Fang Z. Xu S. and Zhang S. Anal. Chim. Acta 1984,164,41. Fang Z. Xu S. and Zhang S. Anal. Chim. Acta 1987,200,35 RilZi&a J. and Amdal A. Anal. Chim. Acta 1989,216,243. 14 15 16 17 18 19 20 21 22 Beauchemin D. and Berman S. S. Anal. Chem. 1989,61 1857. Porta V. Sarzanini C. and Mentasti E. Mikrochim. Acta 1989 3 247. Porta V. Sarzanini C. Abollino O. and Mentasti E. Anal. Chem. in the press. Porta V. Mentasti E. Abollino O. and Sarzanini C. in prepara- tion. Nakashima S. Sturgeon R. E. Willie S. N. and Berman S. S. Fre- senius 2. Anal. Chem. 1988,330,592. Stary J. Solvent E.vtrac.tion of Metal Chelates Pergamon Oxford 1964. Mentasti E. Porta V. Abollino O. and Sarzanini C. Ann. Chim. (Rome) 1989,79,629. Mentasti E.. Porta V. Abollino 0.. and Sarzanini C. Proceedings of the Conference on 'Emironmental Impact in Antarctica,' Rome 8-9 June 1990 pp. 31-36. Burton J. D. and Statham P. J. in Heavy Metals in the Marine Envi- ronment ed. Fumess R. W. CRC Press Boca Raton FL 1990. Paper 010461 7E Received October- 15th 1990 Accepted December 12th 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600119

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 123 Determination of Ammonium Acetate Extractable Molybdenum in Soil and Aqua Regia (Hydrochloric Acid and Nitric Acid 3+1) Soluble Molybdenum in Soil and Sewage Sludge by Electrothermal Atomic Absorption Spectrometry* William H. Rowbottom t Central Analytical laboratory The Scottish Agricultural College (Edinburgh) West Mains Road Edinburgh EH9 3JG UK Methods were developed for the direct electrothermal atomic absorption spectrometric determination of ammoni- um acetate extractable molybdenum in soil and aqua regia (hydrochloric acid + nitric acid 3+1) soluble molybde- num in soil and sewage sludge. An atomization temperature of 2900 "C provided maximum sensitivity without reducing the lifetime of the furnace. Multiple injections with an individual drying and ashing stage were used to provide the sensitivity required for the extractable determinations.Satisfactory accuracy was obtained and charac- teristic mass precision and detection limits were determined. Molybdenum values are reported for a range of Community Bureau of Reference [( BCR) Belgium] Certified Reference Materials. Keywords Molybdenum determination; electrothermal atomic absorption spectrometry; soil and sewage sludge analysis; ammonium acetate extraction; acid digestion Following legislation governing the use of sewage sludge in agriculture the determination of Mo in soil and sludge has become increasingly important. Molybdenum has an essential role in plant nitrogen metabolism and cauliflowers in particu- lar display deficiency symptoms when grown on soils with low Mo content.' High levels of Mo in soil however raise the pos- sibility of enhanced entry of Mo into food chains and may also lead to Mo toxicity in cattle fed on herbage grown on the soil.' Where the availability of Mo for uptake by plants is impor- tant a method for the determination of the mobile or readily extractable Mo is required For pollution monitoring however a method for determining the total content of Mo in the soil is necessary.The extractable Mo content of soil can be assessed using 1 rnol dm-' ammonium acetate solution (pH 7.0) while digestion with aqua regia (hydrochloric acid + nitric acid 3+1) is used to give an indication of total Mo levels in soil and sewage sludge. Extractable Mo levels in soil vary from less than 0.01 to over 0.7 mg kg-' and the total Mo content is nor- mally in the range 0.2-5.0 mg kg-'.The total Mo content of sewage sludge can vary widely depending upon the source of the sludge but is generally around 3.0 mg kg-I for sludges from works receiving little industrial effluent. The analytical methods most frequently employed for the determination of Mo in soil and sewage sludge are spectropho- tometry3 and flame atomic absorption spectrometry (FAAS).4 Both methods however lack the sensitivity required for soils of low Mo status unless time-consuming refinements are incor- porated. A polarographic method with the required sensitivity has been reported' and tried in this laboratory but it was slow and prone to interference problems.To achieve the sensitivity required while avoiding the need for sample pre-concentration or purification electrothermal AAS (ETAAS) has become the method of choice. The most commonly reported problem when using ETAAS for the determination of Mo is the formation of thermally stable carbides by reaction of Mo with the atomizer material. This reduces atomization efficiency and sensitivity and may also result in 'carry-over' a slow accumulation of the element on the furnace which effects repeatability. Various elemental interference effects usually overcome by solvent extraction * Presented at the Fifth Biennial National Atomic Spectroscopy Sympo- + Present address Clatto Laboratory Tayside Regional Council Water sium (BNASS) Loughborough UK 18th-20th.July 1990. Services Department Dalmahoy Drive Dundee DD3 3RP UK. and a reduced furnace lifetime have also been noted. In the direct determination of ammonium acetate extractable Mo in soil by ETAAS Baucells et ~ 1 . ~ found interference from Al Fe and Mg. Using 10% vlv nitric acid as a finishing solvent following evaporation of the extract they experienced corrosion of the pyrolytic graphite layer of the graphite furnace and a 50% reduction in sensitivity after 30 firings. A reduced atomization temperature of 2400 OC frequent re- calibration and a computer program were used to overcome the problem. Lechotycki' reported interference from Al Fe Ca and V in the direct determination of aqua regia soluble Mo in soil by ETAAS but found that only V produced a significant interfering effect.In both of the papers mentioned above it was concluded that the respective fractions of soil Mo could be determined satisfactorily by ETAAS without solvent extrac- tion prior to measurement. The objective of this work was to develop a direct ETAAS method for the determination of both extractable Mo in soil and aqua regia soluble Mo in soil and sewage sludge. Experimental Apparatus A Varian AA 1475 atomic absorption spectrometer with a GTA-95 furnace auto-sampler and Hewlett-Packard printer were used for the investigations. The instrument was equipped with a deuterium lamp background correction system which allowed the simultaneous display of both atomic and back- ground absorption peaks. Molybdenum was determined using the 313.3 nm line and a bandpass of 0.5 nm.Pyrolytic graphite coated graphite furnace tubes were used with sample injection directly onto the tube wall. The peak area mode was chosen partly because of the broad peaks asso- ciated with Mo and partly because the use of integrated absor- bance can compensate for the difference in rates of atomization that may occur between samples and standards.x Reagents All chemicals were of Aristar or SpectrosoL grade. The I rnol dm-3 ammonium acetate extracting solution was prepared by mixing together 2.5 I volumes of 4 rnol dm-3 acetic acid and 4 rnol dm-2 ammonia solution adjusting the pH to 7.0 using 1 rnol dm-3 acetic acid or 1 rnol dm-3 ammonia solution and di- luting to 10 1 with distilled water. The ayua i-e,qia digestion124 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 199 1 VOL. 6 acid was prepared by mixing one volume of concentrated ( 1.42 g cm-I) nitric acid with three volumes of 6 mol dm-3 hydro- chloric acid.Working standards were prepared daily from a Mo stock solution [ I mg ml-I (BDH)] by dilution with 1 mol dm-3 ammonium acetate solution for the extractable de- terminations or by dilution with a solution of 1.6% v/v diges- tion acid in 1 mol dm-3 ammonium acetate for the aqua r-egia soluble determinations. Sample Preparation For the determination of extractable Mo eight soil samples chosen to represent the range of soil types received in this laboratory were air dried ground so that they could pass through a 2 mm sieve and extracted for 16 h with 1 mol dm-3 amonium acetate solution; 7.5 g of soil and 120 ml of extrac- tant were used. Following centrifugation Mo was determined directly in the extract.For the determination of aqua rcgia soluble Mo six standard reference materials obtained from the Bureau of Analysed Samples Middlesbrough UK were used. Five were Communi- ty Bureau of Reference (BCR) Belgium Certified Reference Materials two soils and three sewage sludges while the sixth was Harbour Marine Sediment (PACS- 1) obtained from the National Research Council Canada. Only the marine sediment had a certified Mo value. The materials were digested by reflux boiling in 16 ml of aqua regia for 2 h; 2 g of sewage sludge or marine sediment or 5 g of soil were taken. Following digestion the solutions were made up to 100ml with distilled water.A further 1+9 dilution was then carried out on a small aliquot using 1 mol dm-3 ammonium acetate solution and Mo was de- termined directly in this solution. The moisture content of the reference materials was determined separately and the results corrected to 100% dry matter. Results and Discussion Instrument Operating Parameters The furnace atomization temperature recommended for Mo varies with the manufacturer but is normally within the range 2600-2700 "C. In a paper that described the possible atomiza- tion mechanisms of Mo Chakrabarti et al.9 reported a marked increase in sensitivity with increasing atomization tempera- tures above 2600 "C. Fig. 1 shows the variation of peak area with atomization temperature found in the present work for a Mo standard solution.A more than 3-fold increase in peak area occurs on increasing the temperature from 2600 to 2900 "C. This significant change is also illustrated in Fig. 2. 0.9 ' v) s 0.6 - a -Y a 2 0.3 - I 1 1 1 I 2600 2700 2800 2900 Atomization temperaturePC Fig. 1 pg ml-' Mo standard solution injection volume 35 pl Variation of peak area with atomization temperature for 0.05 3000 0.5 2 0 1 .o 2.0 3.0 3000 0.5 E (b) a / I I I I 1 I I I I 0 -..,... L- -- / -' L. J Tirne/s 0 1 .o 2.0 3.0 Fig. 2 Atomization peaks for 0.05 pg mi-' Mo standard solution for (a) atomization temperature 2600 "C and (h) atomization temperature 2900 "C. Injection volume 35 pl The use of an atomization graph such as that shown in Fig. 1 is important in establishing optimum conditions. With some elements a loss of sensitivity will occur if the atomization temperature is too high owing to more rapid diffusion of the analyte atoms within the furnace.'O This however is not the situation with Mo.The formation of two separate non-volatile Mo carbides has been demonstrated by Sneddon et al." and Chakrabarti et al.9 Using X-ray diffraction to study the internal surface of graphite atomizers they identified both Mo,C and MoC. At temperatures around 1800 "C MozC is converted into MoC which remains until atomization. It is likely that with high atomization temperatures such as 2900 "C thermal dissociation of MoC occurs more readily with a consequent increase in sensitivity. The use of pyrolytic graphite coated graphite furnaces mini- mizes carbide formation and it has also been noted that in the determination of V by ETAAS" carbide formation was reduced when the sample was injected into a pre-heated furnace at 150 "C.Preliminary work with hot injection in this instance indicated a slight reduction in sensitivity when a tem- perature of 150 "C was used. This was probably owing to boiling of the sample at the moment of injection as the injec- tion rate could not be controlled. An injection temperature of 100 "C was chosen to prevent boiling but allow a more rapid drying of the sample than could be achieved with injection at ambient temperature. Owing to the refractory nature of Mo high ashing tempera- tures may be used without loss of the element and Varian rec- ommend a maximum ashing temperature of 1700 "C. Chakrabarti et al.9 and Monte and CurtiusI3 reported a dip in the ashing temperature graph for Mo from 1000 to 1600 OC however an ashing graph prepared during initial work on the present method did not show a dip and indicated that little is to be gained by using an ashing temperature above 1000 "C.Background absorption could be reduced to minimal levels in all of the samples examined by ashing at 1000 "C for 10 s. Table 1 gives the furnace operating parameters used throughout the investigation. In the interest of economy nitro- gen was used as the purge gas during the drying and cooling steps with argon which provides greater sensitivity ,I4 replac- ing the nitrogen during ashing and immediately prior to atomi- zation. The sample injection volume was 35 pl. For the determination of ammonium acetate extractable Mo a multipleJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL.6 125 Table 1 Furnace operating parameters Step Temperature/ Time/ Gas Gas Read No. "C s flow/ls-' type command 1 2 3 4 5 6 7 8 9 10 100 200 200 lo00 lo00 lo00 2900 2900 2900 90 3.0 30 25 10 10 I .0 1 .0 2.0 1 .0 13 3 .0 3 .0 3.0 3.0 3.0 0.0 0.c 0.0 3.0 3.0 Nitrogen Nitrogen Nitrogen Nitrogen Argon Argon Argon Argon Nitrogen Nitrogen injection technique was used to achieve the required sensitivi- ty. This involved the injection of three separate 35 pl aliquots into the furnace each being individually dried and ashed before the final atomization step. Analysis using ETAAS in general requires more care than does analysis by FAAS and it is advisable to spend some time optimizing instrument condi- tions before analysis in order to avoid problems later.Correct positioning of the injection probe within the furnace at the time of sample injection is important if accurate and precise injec- tions are to be achieved. This is particularly true when large in- jection volumes such as 35 PI are being used. The optimum final probe position for the equipment used in this work was 1- 1.5 mm above the furnace wall. It is also important that the outside surface of the injection probe remains clean. If the probe becomes dirty or greasy the surface tension of the sample droplet will break during expulsion and sample liquid will adhere to the outside of the probe. In order to prevent this from happening a 2.5% v/v solution of Decon 90 in distilled water was used in the sample rinse bottle.Furnace Lifetime The prob!em of reduced furnace lifetime and rapidly diminish- ing sensitivity reported by Baucells et al.h was thought to arise from a combination of sample solution acidity and the atomi- zation temperature used (2400 "C). The final acid strength of the aqua regia digest solutions in this work was approximately 1.6% v/v compared with the 10% v/v nitric acid used by Bau- cells et a1.6 but the atomization temperature at 2900 "C was considerably higher. Preliminary work involving the injection of ammonium acetate soil extract solutions into the furnace gave no indication of reduced furnace lifetime despite the high atomization temperature. In view of this it was decided to carry out dilutions of the total Mo aqua regia digest solu- tions with 1 mol dm-7 ammonium acetate in an effort to buffer the acidity. Fig.3 shows the variation of peak area with the number of furnace firings for a Mo standard solution prepared both in 1.6% v/v aqua 1-egia and 1.6% vlv aqua i-egia in 1 mol dm-' ammonium acetate. In both instances the overall reduction in sensitivity was surprisingly small. Buffering with ammonium acetate however did appear to extend the lifetime of the furnace. After 150 firings the conventional acidified standard displayed a 6.2% reduction in sensitivity compared with 4.2% for the ammonium acetate buffered standard and after 300 firings the reductions in sensitivity were 19 and 12.8% respec- tively. During actual analysis the small decrease in analytical signal could easily be controlled by periodic re-calibration.Initial work undertaken to establish the optimum atomization time at 2900 "C indicated a marked reduction in furnace life- time when atomization without purge gas flow was extended beyond 2 s. As can be seen from Fig. 2 the re-introduction of gas flow after 2 s resulted in the loss of the tail of the peak. This however had only a minimal effect on analytical preci- sion which at the start of the study based on four replicate v) 3 t! B Y (0 0.4 - L A 0 150 300 Number of furnace firings Fig.3 Variation of peak area with number of furnace firings. A Mo standard solution in 1.6% ayua regia; and B Mo standard solution in 1.6% ayira regia in 1 mol dm-j ammonium acetate solution firings and expressed as relative standard deviation (RSD) was 2.1 % for the acidified standard and 1.8% for the buffered standard.After 300 firings these figures had become 4.2 and 4.7% respectively. Frequent checks involving the firing of blank solutions immediately after samples or standards con- taining a high concentration of Mo showed no evidence of 'carry-over'. Ammonium Acetate Extractable Molybdenum Table 2 shows a comparison of ammonium acetate extractable Mo results obtained using both a calibration graph and the method of standard additions. The eight soil samples chosen to provide a range of matrix types were extracted and analysed on four separate occasions using both approaches. The individual results shown are the mean values for the replicate analyses. For all but the sandy silt loam the method of standard additions gave the higher result.It is possible that matrix interference produced a small signal suppression which could only be compensated for by standard additions. However the recovery experiment results given in Table 3 which were calculated using the calibra- tion graph suggest an alternative explanation. As was previous- ly observed the re-introduction of a purge gas flow following a 2 s atomization time results in the loss of the tail of the peak. Where a direct comparison is made between standards and samples using the calibration graph this does not matter provid- ing the peak shapes are similar. Using the method of standard additions however the proportion of the peak area not meas- ured at high concentrations of Mo gives a calibration with a smaller slope and thus high results. Use of the calibration graph provides sufficient accuracy for agricultural advisory purposes where soil pH also needs to be taken into account in the interpre- tation of extractable Mo values.1s Table 2 tained by calibration graph and the method of standard additions Comparison of ammonium acetate extractable Mo results ob- Mo found*/mg kg-' Sample Calibration graph Method of standard additions Loamy sand Sandy loam Sandy silt loam Sandy clay loam Clay loam Organic loam Peaty loam Peat 0.030 (f 0.0004) 0.068 (f 0.00 1 5 ) 0.02s (+ 0.00 I 2 ) 0.042 (k 0.0019) 0.008 (k0.0oO6) 0.03s (f 0.00 1 5 ) 0.098 (f 0.0044) 0.056 (f 0.0012) 0.034 (k 0.0018) 0.077 (+ 0.0017) 0.024 (f 0.00 10) 0.046 (f 0.0013) 0.009 (f 0.o008) 0.042 (f 0.0013) 0.1 10 (k 0.0066) 0.069 (k 0.0066) *Results are the mean values f standard error of the mean of four values.126 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH I99 I VOL.6 Table 3 tracts; n=3 Recovery of Mo added to ammonium acetate solution soil ex- Table 4 bration graph and the method of standard additions Comparison of uyitu r.eg:iu soluble Mo results obtained by cali- Sample Before Amount Average amount Mean addition/ng added/ng recoveredhg recovery( X Loamy sand 5.6 Sandy loam 12.8 Sandy silt loam 4.7 Sandy clay loam 7.9 Organic loam 6.6 Peaty loam 18.3 Peat 10.5 Clay loam 1.5 4.0 8.0 4.0 8.0 4.0 8.0 8.0 8.0 4.13 8.83 4.06 8.60 3.96 8.26 8.90 8.47 103.3 110.4 101.6 107.5 99.2 103.3 I 1 1.3 105.8 The recovery experiment in which aliquots of the soil ex- tracts were spiked with Mo was carried out on three separate occasions with the mean recoveries shown (Table 3).The precision of the determination including sample weigh- ing and extraction calculated on the four replicate analyses of each sample and expressed as % RSD varied between samples and ranged from 2.7 for the loamy sand to 15 for the clay loam. Analytical precision based on four replicate firings of a standard solution was 2.4% RSD. The limit of detection in solution was 0.125 ng ml-I equivalent to 0.002 mg kg-' in air dried soil giving a practical determination range of 0.010-0.20 mg kg-I. The limit of detection was calculated as the mean plus twice the standard deviation of ten blank readings. The characteristic mass as measured was 7.6 pg.This compares well with the in- strument manufacturers quoted figure of 8.0 pg. Aqua Regiu Soluble Molybdenum There are few if any soil or sewage sludge reference materials listed in the literature of suppliers as having certified or quoted values of Mo content. The Harbour Marine Sediment PACS- 1 (National Research Council Canada) was the only Mo certified material of a similar nature found. As the verification of ana- lytical methods is greatly assisted by access to such materials it was decided to use the BCR reference materials listed in Table 4 in the development of the method for aqua r-egia soluble Mo. As far as this worker is aware Mo values have not previously been reported for these samples and it was hoped the results would be of use to others engaged in similar work.Table 4 shows a comparison of aqua regia soluble Mo results obtained using both a calibration graph and the method of standard additions. The six reference materials were digest- ed and analysed on four separate occasions using both ap- proaches. The individual results shown are the mean values for the replicate analyses. The results suggest that interference effects are minimal. This is confirmed by the recoveries of added Mo given in Table 5 and the values obtained for the marine sediment PACS-1 which has a certified total Mo value of 12.3 k 0.9 mg kg-l. Metals contained in the crystal lattices of unweath- ered primary minerals are not completely dissolved by diges- tion with aqua r-egia and it is usual depending on the nature of the sample to solubilize and recover only 60-90% of the element.For the marine sediment however the presence of more easily dissolved secondary minerals would explain the agreement between the aqua I-egia soluble value and the certified total. The recovery experiment values given in Table 5 are the means of individual recoveries obtained on three separate oc- casions. The results were calculated using the calibration graph. The use of aqua i-egia as a digestion acid is now well established,Ih and as the objective of the experiment was to demonstrate the accuracy of the graphite furnace analytical finish Mo was added to the samples after digestion rather than before. Precision for the full determination calculated on the four replicate analyses of each sample ranged from 0.42 to Mo found*/mg kg-' Sample Light sandy soil Over-fertilized soil Sewage sludge - domestic BCR 142 BCR I43 BCR 144 BCR 145 Sewage sludge Sewage sludge - industrial BCR 146 PACS- I Harbour Marine Sediment Calibration graph Method of standard additions 0.22 (+ 0.0082) 0.22 (f 0.0132) 6.2 (+ 0.I03 1 ) 6.6 (k 0.1250) 4.6 (+ 0.1040) 4.6 (+ 0.1440) 5.3 (k 0.1 190) 5.0 (k 0. I 190) 13.3 ( f O . 1 190) 13.9 (k0.1110) 12.0 (k 0.0250) 12.6 (k 0.1780) *Results are the mean values f standard error of the mean of four values. Table 5 Recovery of Mo added to reference material digests; n=3 Sample Before Amount Average amount Mean additionhg added/ng recovered/ng recovery*( %) BCR 142 Light sandy soil 3.3 12.0 12.4 103.3 BCR 143 Over-fertilised soil 93.0 50.0 55.5 I 11.0 Sewage sludge - domestic 27.6 50.0 54.I 108.2 Sewage sludge BCR144 BCR I45 31.8 50.0 53.1 106.2 BCR I46 Sewage sludge- PACS- I Harbour industrial 79.8 100.0 106.2 106.2 Marine Sediment 72.0 50.0 50.3 100.5 7.4% RSD. Analytical precision based on four replicate firings of a standard solution was 1.8% RSD. The limit of detection in solution was 0.43 ng ml-l equivalent to 0.086 mg kg-' in dry soil or 0.22 mg kg-' in dry sewage sludge. These limits could be improved upon by reducing the final dilution factor or by using the multiple injection approach employed in the extract- able determinations. The practical determination range for soil was 0.4-7.0 mg kg-' and for sewage sludge 1 .&I 8.0 mg kg-I. The characteristic mass as measured was 7.6 pg.Conclusions For the samples examined Mo can be determined with accept- able accuracy and precision without resorting to the method of standard additions or solvent extraction prior to measurement. 'The sensitivity of the proposed method is more than adequate for the determination of total Mo for pollution monitoring pur- poses and may be further enhanced by multiple sample injec- tion to allow the determination of extractable Mo. Use of an atomization temperature of 2900 "C overcomes the problems associated with carbide formation and provides maximum sen- sitivity without reducing furnace lifetime. Complex matrix matching of standards is not required.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 127 The author thanks Dr. P. Crooks Central Analytical Depart- ment Scottish Agricultural College for his encouragement support and constructive criticism throughout this work.7 Lechotycki A. J. Anal. At. Spectrom.. 1990,525. 8 Heav Metals in Soils ed. Alloway. B. J. Blackie Glasgow 1990 p. 60. 9 Chakrabarti C. L. Shaole W. Marcantoni F. and Headrick K. L. References Sauchelli V. Trace Elements in Agi.iculture Van Nostrand Reinhold New York. 1969. p. 141. Underwood E. J. Truce Elements in Human wid Animal Nutrition Academic Press London 4th edn. 1977 p. 122. Standing Committee of Analysts Methods for the Examination of Waters and Associated Materials Molybdenum (especially in sludges and soils) by Spectrophotometry 1982 SCA No. 70 HM Stationery Office London 1983. Standing Committee of Analysts Methods for the Examination of Waters and Associated Materials Methods for the Determination of Metals in Soils Sediments and Sewage Sludge and Plants by Hydro- chloric-Nitric Arid Digestions with a note on the Determination of the Insoluble Metal Contents 1986 SCA No. 98 HM Stationery Office London 1987. Edmonds T. E. Commun. Soil Sci. Plant Anal. 1982 13 1. Baucells M. Lacort. G. and Roura M. Analyst 1985,110 1423. Fresenius 2. Anal. Chem. 1986,323,730. 10 Slavin W. Manning D. C. and Camrick G. R. Talanta 1989 36 171. 1 1 Sneddon J. Ottaway J. M. and Rowston W. B. Analyst 1978 103 776. 12 Apostoli P. Allesio L. Dal Farra M. and Fabbri P. L. J . Anal. At. Spertrom. 1988,3,47 1. 13 Monte V. L. A. and Curtius A. J.. J. Anal. At. Spectrom. 1990 5 21. 14 Analytical Methods for Graphite Tube Atomisers ed. Rothery E. Varian Techtron Victoria Australia 1982 p. 19. 15 Macaulay Institute for Soil Research and Scottish Agricultural Col- leges Liaison Group Ad1Pisor-y Soil Analysis and Interpretation Bulle- tin I 1985. 16 Berrow M. L. and Stein W. M. Analyst 1983 108,277. Paper 0104305B Received September 24th 1990 Accepted December 10th. 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600123

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 129 Pre-concentration by Coprecipitation Part 1. Rapid Method for the Determination of Ultra-trace Amounts of Germanium in Natural Waters by Hydride Generation-Atomic Emission Spectrometry* Ian D. Brindle Mary E. Brindle and Xiao-chun Let Chemistry Department Brock University St. Catha fines Ontario L2S 3A 1 Canada Hengwu Chen Department of Chemistry Hangzhou University Hangzhou Zhejiang 3 7 0028 People 3 Republic of China A rapid method for the determination of ultra-trace amounts of germanium in sea-water surface water and ground- water has been developed. This method initially involved coprecipitation of gallium and magnesium with hydroxide ions. Subsequent investigations have shown that gallium is unnecessary for quantitative recovery if calcium and carbonate ions are added with the magnesium and the precipitation is carried out at a high pH.The formation of a rather coarse precipitate aids in the filtration step by reducing the filtration time from hours to minutes without any loss of analyte. The control of pH is crucial to the successful use of this method. Concentrations of germanium in waters have been determined by hydride generation atomic emission spectrometry with a detection limit for the method of 0.6 pg ml-' (30) for a 1 I sample. Determinations of germanium concentrations in ground sea and surface waters range from 0.5 to 17 pg mi-' (3 I sample). Initial experiments into the application of the method to samples spiked with 20 ng ml-1 concentrations of copper colbalt nickel and zinc have proved to be effective.Keywords Pre -concen tra tion; coprecipita tion; germanium determination; hydride genera tion; natural water Analytical methods used for the determination of trace levels of germanium have included spectrophotometry1-3 and atomic spectrometry."I3 Among the spectrophotometric determina- tions of germanium a recently reported method3 showed the best detection limit to be approximately 48 ng ml-I of germa- nium. Hydride generation coupled with atomic spectro- metry,I@l3 however has generally provided a further improvement in the detection limit for the determination of germanium. Andreae9 trapped germane in a liquid nitrogen U- tube before releasing it to an atomic absorption spectrometer for detection obtaining a detection limit of about 200 ng ml-I.Zhang et a/.' I recently demonstrated in situ pre-concentration of germanium in a palladium coated graphite furnace. Upon the increase of temperature of the graphite furnace the trapped germanium was released and then determined by atomic ab- sorption spectrometry. With this in situ pre-concentration sen- sitivities expressed in terms of characteristic mass for arsenic antimony and selenium of 10.0 13.1 and 14.7 pg per 0.0044 A respectively were reported. A gas flow batch type hydride generation system for the determination of hydride-forming elements in the presence of L-cysteine has been developed.12-15 With this method high sensitivity was obtained without pre-trapping. A detection limit of 12 pg ml-I was obtained for germanium.'3 This is the best detection limit reported to date.However the method was not sensitive enough for the direct determination of germani- um in surface-water samples as concentrations of germanium in these samples were at sub-pg ml-l levels. A pre- concentration method was therefore required in order to bring the concentrations within the measurable range. Coprecipitation by hydroxides of metals such as Al Cd Fe Ga La and Mg has been reported for the pre-concentration of trace Although these pre-concentration systems have been widely used Akagi et al." claimed a superior copre- cipitation system for the determination of trace metals by induc- tively coupled plasma atomic emission spectrometry (ICP- * Presented st the Fifth Biennial National Atomic Spectroscopy Symp- t Present address Department of Chemistry.University of British sium (BNASS). Loughborough UK 18th-20th July. 1990. Columbia. Vancouver. British Columbia V6T I WS. Canada AES). They reported that precipitation of Ga(OH) in the pres- ence of magnesium gave good recoveries for Al Co Cr Fe La Mn Ni Pb Ti V Y and Zn. With this system these workers noted that spectral interferences from gallium in the determina- tion by ICP-AES were negligible and the contamination was minimal. Akagi and Harapchi*> have recently reported a modification of the method that included a centrifugation step. Another method for the determination of elements in water which involves a reductive precipitation has been described. Tetrahydroborate( HI) was used to reduce palladium and iron24 or palladium alone2s which allowed efficient collection of trace concentrations of various elements from water.Since iron con- tains germanium as a ubiquitous impurity and as palladium in- terferes with the generation of hydrides," this method cannot be used for the determination of those elements by hydride generation AES. The combination of gallium and magnesium was thought to be an appropriate medium for the collection of ultra-trace concen- trations of elements in surface waters where magnesium would be added to the water to aid in the coprecipitation process. By using the coprecipitation approach for the pre-concentration of germanium the gallium-magnesium system was investigated. Experimental Instrumentation Data were acquired with a Spectraspan V d.c.plasma atomic emission spectrometer equipped with a Beckman hydride gen- erator modified as described elsewhere.14 Signals were record- ed on a Fisher Recordall Series 5000 chart recorder. A Brinkmann variable volume Macro-Transferpettor was used for all analyte injections with the volume fixed at 5.0 ml. Determinations of pH were performed with a Radiometer pH meter (PHM 82 standard pH meter). Reagents The germanium standard L-cysteine and sodium tetrahydro- borate( 111) were prepared as previously described." l 7 Gallium metal (Aesar Toronto Ontario Canada) was dissolved in con-130 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 centrated nitric acid and diluted with distilled water to the re- quired concentration. Magnesium (Aesar) was dissolved in 1+1 nitric or hydrochloric acid.Calcium was prepared from calcium carbonate (Aesar) by dissolving in 1 mol dm-3 hydro- chloric acid. Sodium hydrogen carbonate [BDH (AnalaR) Toronto Ontario Canada] was dissolved in distilled water. Solutions of metals were prepared from dilutions of atomic ab- sorption standards or from the elements which were of analyti- cal-reagent grade or better. Purification of the reagents was performed by adding freshly prepared magnesium hydroxide to a solution of the reagent at the appropriate pH to prevent dissolution. The solution was stirred and heated and the precipitate allowed to settle after which the supematant solution was decanted off; thus the blank values for germanium were reduced by approximately half.This prwedure was described by Doroshkov and Chuiko26 for the purification of magnesium sulphate. Precipitation Procedure Method one To 1 1 of water were added magnesium and gallium solutions to give final concentrations of 500 and 25 pg ml-I respective- ly. The sample was then treated by the same method as used by Akagi et a1.22 For the determination of germanium the so- lution was prepared as described in previous work.I2 Method two Solutions of magnesium calcium and sodium hydrogen car- bonate (all 1% m/v) were added to a 1 1 volume of sample to give final concentrations of 200 60 and 200 pg ml-' re- spectively. The pH was then raised to 9.75 * 0.25 with sodium hydroxide solution ( 1 mol dm-3) and the solution heated to near boiling point and then allowed to cool over- night.After cooling the solution was filtered through a 5 pm filter membrane. The membrane was removed from the appa- ratus and placed in a beaker. The precipitate was dissolved in 1 mol dm-' nitric acid and the solution treated appropri- ately for the determination. For the determination of germa- nium the solution was prepared as described in previous work.'* Results and Discussion Preliminary Studies Method One For the coprecipitation of trace amounts of germanium similar conditions to those reported by Akagi et al." were used. A satisfactory recovery of 95% from a solution containing 0.1 ng ml-1 of germanium was obtained. However the procedure was very time consuming with the filtration step requiring up to 8 h to recover the precipitate from 1 1 of solution.In order to overcome this problem attempts were made to reduce the filtration time whilst retaining the good recovery of ger- manium. Overnight digestion produced a precipitate that was more tractable. The supernatant solution (approximately 900 ml) passed through the filter rapidly and slow filtration occurred only as the last 100 ml of the mixture of precipitate and solu- tion from the bottom of the beaker were filtered. Thus the total filtration time was shortened from 8 to approximately 3 h. A recovery of 97% from 0.1 ng ml-I of germanium in 1 .O 1 of so- lution was obtained by use of this technique. Various digestion procedures were applied in an attempt to reduce the filtration time further. Filter membranes of various pore sizes were also investigated.The results are shown in Table 1. The results in Table 1 also indicate that slight changes in the concentrations of magnesium and gallium did not significantly affect the recovery of germanium. With a higher concentration of magnesium and gallium however the time required for filtration was also longer owing to the larger amount of precipitate formed. Obviously variation of sample volume changed the filtration time as this depends on the amounts of sample and precipitate. Since the 5 pm filter mem- brane gave similar results to 0.4 and 1.0 pm filters 5 pm filter membranes were used in further studies. Coprecipitation of Germanium in the Presence of Mg2+ Ga3+,Ca2+ and HC03- Based on the preliminary studies described above coprecipita- tion and determination of germanium in water samples were performed following method one as stated under Experimen- tal.During the analysis of a particular ground-water sample the filtration time for the sample precipitate was much shorter than that for the precipitate formed in a blank. A visibly coarser precipitate was obtained from the sample than from the blank which contained distilled water and the reagents for the coprecipitation. The sample and the blank differed only in their matrix. The ground-water sample contained a high level of calcium (approximately 200 pg ml-I). The formation of CaC0 was thought to be responsible for the larger particle size of the precipitate which eased the filtration. Therefore further investigations involving the addition of reagents to allow the formation of CaC03 were carried out.Instead of the direct addition of CaCO to the sample solu- tion the addition of Ca2+ [in the form of CaC& or Ca(N03)2] Table 1 Recovery of Ge and filtration time under different experimental conditions Concentration of Mg/ Concentration of Ga/ Overnight Sample pg ml-' pg ml-' digestion 1 .0 I of distilled water 0.1 ng ml-' of Ge in 1 .0 1 of 0.1 ng ml-' of Ge in 1 .0 I of 0.1 ngml-'ofGein l.0lof 0.02 ng ml-' of Ge in 0.5 1 of 0.1 ng ml-' of Ge in 1 .0 1 of 0. I ng ml-I of Ge in I .0 I of 0.05 ng mf-' of Ge in I .O 1 of 0.01 ng ml-I of Ge in 1.0 I of distilled water distilled water distilled water distilled water distilled water distilled water distilled water distilled water 500 500 550 480 5 10 640 560 500 500 25 25 30 25 40 38 25 25 25 No Yes No Yes Yes Yes Yes Yes Yes Heating digestion No No Yes Yes Yes Yes Yes Yes Yes Membrane size/pm 1 1 1 1 1 0.4 0.4 5 5 Filtration time/h 8 3 3 1 0.8 3 2 0.8 0.7 Recovery of Ge (96) - 97 97 98 93 91 94 99 97131 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL.6 g 120 c I (a) A 1 .- 5 100 E C 80 & 60 r $ 40 20 0 a - 0 100 200 300 400 500 600 700 Filtration time/min Fig. 1 (a) Recovery of germanium as a function of magnesium at two concentrations of calcium. ( h ) Filtration time as a function of magnesium concentration at two concentrations of calcium. A 60; and B 180 pg ml-l of calcium a 0 g 7 7.5 8 8.5 9 9.5 10 10.5 11 PH Fig. 2 Recovery of germanium as a function of pH and NaHC03 solutions to the sample was chosen. Upon the addition of NaOH used to adjust the pH of the sample solu- tion (e.g.to pH lo) CaC03 Mg(OH)? and Ga(OH)3 were formed. This precipitate quantitatively collected the analyte and formed larger precipitate particles. With this modification an initial set of two experiments gave recoveries of 98-104% for 100 ng of germanium spiked in 1.6 1 of distilled water and less than 10 min were required for the filtration. Optimizations were carried out with this modified coprecip- itation system by varying the concentration of Mg2+ at calcium concentrations of 60 and 180 pg ml-i in order to maximize the recovery of germanium and minimize the filtration time while the concentrations of Ga3+ and NaHC03 were kept constant at 25 and 150 pg ml-l and the pH of the solution was kept at 9.Fig. l(a) shows the recovery of germa- nium and Fig. l(h) the filtration time as a function of the con- centration of Mg2+ in the presence of 60 and 180 pg ml-I of Ca”. As can be seen from Fig. 1 at a Ca2+ concentration of 60 pg mki complete recovery of germanium was obtained when the Mg?+ concentration reached approximately 30 pg ml-I. Below this level germanium was not completely recov- ered. Within the range of Mg2+ concentrations 30-380 pg ml- I good recoveries of germanium were obtained. With a higher concentration of Mg” a longer filtration time was also re- quired as indicated in Fig. I(h). probably because of the for- mation of a larger amount of Mg(OH) precipitate. When a concentration of Ca2+ of 180 pg ml-i was used a higher minimum concentration of Mg’+ was also required to provide good recovery of germanium.As shown in Fig. I(a) the re- covery of germanium was incomplete at Mg2+ concentrations below 180 pg ml-i in the presence of 180 pg m1-I of Ca2+. However trace amounts of germanium were completely re- covered when the Mg2+ concentration was greater than 250 pg ml-I in the solution containing 180 pg ml-i of Ca2+. The filtration time was generally shorter when more Ca2+ was present in the solution as illustrated by the two curves in Fig. 1. The results shown in Fig. I suggest an interdependence between the concentrations of Mg2+ and Ca2+. This interdepen- dence can be ascribed to competitive precipitation and copre- cipitation among the various species forming the precipitate. The presence of Ca2+ did not enhance the recovery of germa- nium probably because CaC03 could not efficiently collect the germanium; but its presence allowed the precipitate to be more easily removed hence a more rapid filtration time was achieved.The basicity of the solution was also optimized as it is another important factor which influences the quality of copre- cipitation. Fig. 2 demonstrates the effect of pH on the recovery of germanium. As shown in Fig. 2 at a pH value of 8. I only 22% of the germanium was recovered. At this pH little precip- itate was observed. Over the pH range of 9.0-10.5 germanium was quantitatively recorded. Thus a pH of between 9.5 and 10.0 was chosen for the precipitation. Coprecipitation of Germanium With Mg and Ca Method Two Suprisingly when the Ga3+ concentration was varied from 1.5 to 60 pg ml-I complete recovery of germanium was obtained at all of the gallium concentrations studied.Therefore further studies were carried out with a coprecipitation system in the absence of gallium. It was found that in the absence of gallium trace amounts ( 100 ng) of germanium were quantitatively recovered from a 1.6 1 solution containing 60 180 and 600 pg ml-I of Ca2+ Mg and NaHC03 respectively after the pH had been adjusted to 10. For comparison a recovery of 9 1 % was also obtained from 100 ng of germanium in 1.6 1 of a solution containing only 120 pg ml-i of Mg2+ after the pH had been adjusted to 10. These results suggest that Ga(OH) is not necessary as a coprecipita- tion carrier. It appears that Mg(OH)2 is the primary carrier in the coprecipitation of trace amounts of germanium. The proposed coprecipitation system developed in this work consists of Mg” Ca2+ and HC03-.The optimum condi- tions for this system were briefly re-examined based on those previously studied where gallium was present. It was found that the concentrations of Mg2+ and Ca2+ showed the same in- terdependence and their optimum concentrations were similar to those shown in Fig. 1. However with the modified system without gallium the filtration time was further reduced to between 1 and 5 min for a sample of approximately 2 1. Quan- titative recoveries of germanium were obtained at a similar pH range as shown in Fig. 2. Thus at pH values between 9.0 and 10.5 recoveries of 90-1 10% were obtained and hence a pH of 9.75 f 0.25 was used.Samples prepared by the three methods were investigated by electron microscopy. The precipitate formed from gallium and magnesium hydroxides verges on being amorphous. The addition of calcium and carbonate to the initial solution causes the formation of dendritic crystals which probably act as a filter medium to support the relatively amorphous Ga-Mg precipitate. Crystals of possibly cubic habit are also observed covered with the amorphous precipitate. For the precipitate formed with magnesium and calcium rosettes of dendrites and framboidal twinned crystals possibly of cubic habit are apparent with little evidence of an amorphous pre- cipitate. Once the optimum concentrations of Mg2+ and Ca2+ had been chosen the concentration of NaHC03 was varied from 25 to 750 pg ml-I and showed little effect on germanium re- covery.An NaHCO? concentration in the range of 100-200 pg ml-I was chosen.129 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991. VOL. 6 Table 2 Recoveries of spiked germanium from water samples Concentration of Ge determined/ Ge spiked/ Ge found/ Ge recovery Sample pg ml-I ng ng (%) 500.0 ml of sea-water 4.5 0 10.0 ml of 1 .O ng ml-I of Ge spiked into 500.0 ml of sea-water 24.0 - - 10.0 9.75 - 0 98 3.0 I of Smith- 10.0 ml of 2.5 ville well water 15.7 ng ml-I of Ge spiked into 3.0 1 of Smithville well water 23.6 25.0 23.7 95 Table 3 Concentration of germanium in water samples Concentration of Ge/pg ml-l ~ Source of water sample Method one Method two A well in Smithville. Ontario 17 17 ? 1.5 (n=3) Welland River Ontario 1.5 1.4 Port Colborne Lake Erie 1.3 Port Weller Lake Ontario 0.7 Queenston Spring water Niagara Falls Ontario 0.5 Sea- water 4.5 - - - - Recovery of Germanium Spiked Into Water Samples Recoveries of the germanium spiked into real water samples were also studied the results are listed in Table 2.The recov- eries of germanium spiked into a sea-water sample was obtained by method one where gallium was present. Recovery results for 25 ng of germanium spiked into a Smithville (Ontario Canada) well-water sample was obtained by method two in the absence of gallium. As can be seen from Table 2 satisfactory recoveries of germanium were obtained in both in- stances indicating that the use of gallium is unnecessary. Determination of Germanium in Water Samples A number of water samples were analysed for germanium ini- tially by method one; the results are summarized in Table 3.For comparison two of the samples were also analysed by method two and the results are in good agreement with those obtained by method one. Germanium concentrations in all the water samples analysed are at low pg ml-I levels. One groundwater sample (Smithville well water) contained a higher concentra- tion of germanium. The reason for this higher germanium con- centration in the well water is currently being investigated. The concentration of germanium in the reagent blank a 3.0 1 solution containing 170 65 and 130 pg ml-I of Mg Ca and NaHC03 respectively at a pH of 10 was found to be approxi- mately 1 pg ml-I. Blank values for germanium from solutions of 6% NaBH 0.4% L-cysteine and 0.01 mol dm-.> HNO? that had not undergone pre-concentration were not detectable.When the reagents were pre-purified simply by stirring with Mg(OH) the blank was reduced to approximately half of the original amount as indicated under Experimental. Pre-concentration of Transition Elements A preliminary study of transition elements (Cu Co Ni and Zn) spiked into distilled water at 20 ng ml-I gave effective re- coveries using method two but higher pH values were required than for the collection of germanium. Essentially quantitative recoveries ( 100 f 10%) of the four elements tested were obtained in the pH range 10-10.5. At the lower pH values smaller amounts of precipitate were obtained. Thus at pH 9 recoveries of 20-30% were obtained.The detection limits for the elements are given by the manufacturer as Cu 0.002 pg ml-I at 324.754 nm; Co not given at 341.474 nm; Ni 0.006 pg ml-I at 305.082 nm; and Zn 0.006 pg ml-1 at 202.548 nm. Conclusion A rapid quantitative method for the pre-concentration of ele- ments from various waters has been developed. Results for the pre-concentration of germanium are excellent. Further work on the pre-concentration of other elements from surface and ground waters will be undertaken in order to develop a useful method for the determination of ultra-trace concentrations of elements in waters. The authors thank the Ontario Ministry of the Environment for funding this research (Grant 4346). H.C. gratefully ac- knowledges the assistance of Hangzhou University for provid- ing funds for a visiting scholarship.G. Hooper of Norton Advanced Ceramics is thanked for the electron microscope work. I 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 References Hemandis V. Macia L. and Sala J. V. Analyst 1987 112 1007. Shen H. Wang Z. and Xu G.. Analyst 1987 112,887. Nukatsuka I. Takahashi K. Ohzeki K. and lshida R. Analyst 1989,114 1473. Eckhoff M. A. McCarthy J. P. and Caruso J. A. A d . Chrm. 1982,54 165. Thompson M. and Pahlavanpour B. Anal. Chim. A m 1979 109 251. Fazakas J. Tulunta 1984,31 573. Nakata F. Sunahara H. Fujimoto H. Yamamoto M.. and Kumena- ru T.. J. Anal. At. Specwom. 1988 3 579. Hambrick G. A.. Ill Froelich P. N. Jr. Andreae M. O. and Lewis B. L. Anal. Chem. 1984,56,421 Andreae M. O. and Froelich P. N. Jr. Anal. Cheni. 1981 53 287. Halicz L. Analvst 1985 110,943. Zhang L. Ni Z. M. and Shan X.-q. Spwtwdiirn. Acta. Part B 1989,44,339. Brindle 1. D. Le X.-c. and Li X.-f. J . Anal. A?. Spectt.om. 1989. 4 227. Brindle 1. D. and Le X.-c. Anal. Chern. 1989.61 1175. Brindle 1. D. and Le X.-c. Anulwt 1988. 113 1377. Brindle 1. D. and Le X.-c. Anal. Chim. A m . 1990 229 239. Bruninx E.. P h i l i p J. Res. 1979,33,264. Novikov. A. I. and Shchekoturova E. K. Radiokhimiw 1972 14 152. Andrianov A. M. and Poladyan V. E. Zh. Anul. Khini.. 1975 30 1622. Sarkisov. E. S. Lidin R. A.. and Krymskaya. E. B.. I:).. Akucl. Nard. S.S.R.R. Neor<q. Muter .. 1970,6 281. Tsuyama A. and Nakashima S. Birnseki Kagakrr 1980.29 8 I . Shigelomi. Y. Birnseki K a p k i r 1975 24 699. Akagi. T.. Fuwa K.. and Haraguchi H. A n d . Chin?. A m . 1985. 177 139. Akagi. T. and Haraguchi H. A w l . Ckeni.. 1990 62 8 1. Skogerboe. R. K. Hanagan. W. A. and Taylor H. E.. A n d . Chem. 1985.57.28 15. Nakashima. S. Sturgeon R. E. Willie. S. N.. and Berman S. S.. Anal. Chin?. A r w 1988. 207 29 1. Doroshkov V. Ya.. and Chuiko V. T.. /:I-. Khini. Khini. TrA.hnol.. 1967. 10 1087. Puper- 0/03345F Rec.ei\*ed July 24th. 1990 Accepted No\vmher 12th. 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600129

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

133 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH I99 I VOL. 6 Elimination of Copper Interference by Continuous Flow Matrix Isolation in the Determination of Selenium by Flow Injection Hydride Generation Atomic Absorption Spectrometry* Stephen G. Offley and Nichola J. Seare Department of Chemistry Loughborough University of Technology Loughborough Leicestershire L E 1 1 3TU UK Julian F. Tysont Department of Chemistry University of Massachusetts Amherst MA 0 1003 USA Helen A. B. Kibble Development Department Philips Scientific Analytical Division Cambridge CB 1 ZPX UK A flow system incorporating a microcolumn of strongly acidic cation-exchange resin (Dowex 50W) to achieve continuous flow matrix isolation was used to eliminate copper interference in the determination of selenium by flow injection hydride generation atomic absorption spectrometry. The microcolumn manifold used for the selective retention of the copper interferent was interfaced with the hydride generation manifold through a flow injection sample injection valve.The two manifolds were made independent of each other in order to achieve optimum performance characteristics for both the matrix isolation and hydride generation. Following removal of the copper a 400 PI sample was injected into a water carrier stream. This was merged with hydrochloric acid and subsequently with sodium tetrahydroborate soution. After introduction of argon the selenium hydride was separated by a glass U-tube separator and atomized by a tube-in-flame atomizer. The intermittent regeneration of the column with 1 mol dm-3 HCI enabled repeated matrix isolation without any loss in resin efficiency or the need for column repacking and gives the option for total automation.The procedure was validated through accurate analyses of two copper metal reference materials National Institute of Standards and Technology Standard Reference Material 454 Unalloyed Copper XI and Bundesanstalt fur Materialforschung und -prufung Germany Certified Reference Material 361 Copper containing 479 and 36 pg g-1 of Se'V respectively. The system was found to have a characteristic concentration of 1 .O ng mi-' limit of detection of 2.1 ng ml-l relative standard deviation of 1.5% (10 ng mi-l SeIv n = 12) and sample throughput of 51 h-'. Keywords Selenium determination; flow injection hydride generation atomic absorption spectrometry; copper interference removal; continuous flow matrix isolation Copper has been cited as a serious interfering element in the determination of selenium by hydride generation atomic ab- sorption spectrometry (AAS).I-5 Many attempts have been made to remove or reduce this interference with varying degrees of success.Interference reduction has been reported through the optimization of both acid and sodium tetrahydro- borate ~oncentrations,3.~ and the application of a variety of complexing agents."' I Various manual matrix isolation procedures have been applied to hydride generation AAS including coprecipitationI* and ion exchange13-I6 but they have the disadvantage of being tedious labour intensive and to some extent dependent on op- erator skill.A variety of resins have been used for matrix iso- lation including Chelex lOo,I3-l6 Dianion SKIBIS and Dowex 50W-X16.16 Hershey and Keliher'6 successfully applied Dowex 5OW-X 16 cation-exchange resin for the isolation of both arsenic and selenium from a wide range of interfering matrix species including Cu Ni Co and Ag. The resin was re- ported to be more beneficial than the Chelex 100 resin espe- cially when dealing with environmental samples as the Chelex 100 resin gave poor analyte recovery for some samples. With the introduction of continuous flow hydride generation to replace discrete batch methods a reduction in interference has been reported.l4.l7 A further development in continuous flow methodology that of the yse of the flow injection (FI) format was first reported by Astrom.I8 The use of FI tech- niques has been reported for the determination of hydride- forming elements giving accuracy precision low sample and * Presented at the Fifth Biennial National Atomic Spectroscopy Sympo- t To whom correspondence should be addressed.sium (BNASS) Loughborough UK 18th-20th July 1990. reagent consumption and superior sample throughput capabili- ties in comparison with both continuous flow and discrete batch methods. 18-26 Flow injection hydride generation AAS has been reported to reduce interference effects including that of copper in the determination of s e l e n i ~ m . ' ~ . 2 ~ . ~ ~ A variety of procedures for the automation of sample pre- treatment in AAS have developed with FI methodology.Systems capable of both pre concentration and matrix isolation employing liquid-liquid extraction,27-** ion ex~hange'~J~ and precipitation31J2 have been reported. To date little attempt has been made to apply such systems to the removal of interference effects in hydride generation AAS. Ikeda33 reported a system incorporating a microcolumn of chelating resin for the removal of copper and nickel in the determination of selenium. Although matrix isolation was carried out on-line the system relied on manual sample pipet- ting had limited precision and no facility for column regenera- tion. have recently reported on a system for the de- termination of arsenic in a nickel-based alloy by using continu- ous flow hydride generation AAS incorporating on-line matrix removal ltia a microcolumn of strong cation-exchange material (SCX).Although successful for the application some degree of compromise had to be made between the perfor- mance characteristics of the matrix isolation system and the hydride generation manifold itself. Column regeneration was carried out after every four determinations by pumping 1 mol dm-3 HCl through the column to waste after disconnect- ing the sample line. This paper reports an FI system for the determination of se- lenium by hydride generation AAS. Elimination of the inter- ference from copper is achieved by a continuous flow matrix Riby et134 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 n I AAS A R I " P Argon Fig. 1 Schematic diagram of the FI manifold with a continuous flow matrix isolation unit P peristaltic pump S sample; H water; A HCI; R NaBH,; W waste; V I switching valve; V2 sample-injection valve; C microcolumn; and G gas-liquid separator isolation procedure based on a microcolumn of strongly acidic cation-exchange resin (Dowex 50W). The manifold design lends itself to full automation and overcomes the limitations of previously reported Experimental Apparatus A Philips Scientific SP9 atomic absorption spectrometer equipped with a Philips data coded selenium hollow cathode lamp operated at 7.5 mA was used for all determinations.A spectral bandpass of 1.0 nm was used with the 196.0 nm sele- nium spectral line. The signals were recorded on a Tekman TE 200 chart recorder (2-10 mV range) all measurements being expressed as peak-height absorbance.A 50 mm Philips Scientific universal burner was used to support the air- acetylene flame-heated T-shaped silica cell (air flow setting 28; acetylene flow setting 15). The FI hydride generation manifold shown in Fig. I was developed from the gas-liquid separator and hydride atomization systems of a Philips Scientific PU9360 continuous flow vapour system. Two peri- staltic pumps were employed. A Gilson Minipuls 3 was used for the hydride generation manifold and a Gilson Minipuls 2 for the matrix isolation unit. Control of the flow-rates was achieved through application of different bore standard mani- fold tubing (Altec). All manifold tubing consisted of 0.8 mm i.d. polytetrafluoroethylene (PTFE) tubing (Anachem). Mani- fold channels were connected through three-way connector T- pieces (Anachem) which aided reagent mixing.A micro-bore glass column [50 x 3.0 mm i.d. (Anachem)] fitted with porous 25 pm PTFE frits was incorporated into the external sample loop of a rotary sample injection valve (Anachem). Sample in- jection was acheived using a Rheodyne Model 5020 fixed volume loop injector valve operated by an electrically activat- ed universal valve switching unit (Anachem). Sample loops of various volumes were prepared using PTFE tubing (0.5-1.5 mm i.d.) cut to appropriate lengths. Reagents Analytical-reagent grade water produced by a LiquiPure RG system (reverse osmosis followed by ion exchange) was used for all solutions and as a carrier stream. A sodium tetrahydro- borate solution ( 1 % m/v in a 0.1% m/v NaOH solution) was prepared using sodium tetrahydroborate pellets (SpectrosoL BDH) and filtered through a Whatman 541 filter-paper. With refrigeration this solution was usable for up to 3 d.The hydro- chloric acid (6 mol dm-3) reagent solution was of SpectrosoL grade (BDH). All selenium(iv) standard solutions were pre- pared by dilution of a standard solution of selenous acid (Spec- trosol BDH) containing IOOOpgml-' of SelV. For the interference investigation work copper(i1) sulphate pentahy- Table 1 tinuous flow matrix isolation. Injection volume 409 yl Optimized variables for hydride generation AAS with con- Ij-vdride generation AAS- Reagent Concentration Flow -ratelm I min-' H20 carrier HCI IVaBH .Ar - 6.0 6 rnol drn-3 4.2 1.0% m/v 3.2 - 600 Continuous flow matrix isolation- Reagent Concentration Flow-rate/ml min-' Sample < lo00 pg ml-' Cu 2.0 HCI (column regenerant) 1.2 mol dm-A 2.0 drate (AnalaR BDH) was used to prepare interferent stan- dards.Conditioning of the T-shaped silica atomization cell was carried out using a 5% v/v solution of hydrofluoric acid (AnalaR BDH). High-purity argon was used as the purge gas (99.998% Ar BOC). The digestion of the copper metal refe- rence materials was carried out using nitric acid (Aristar BDH) and hydrochloric acid (Aristar BDH). The cation- exchange resin was Dowex 50W-X8 (Drymesh 100-200 hy- drogen form 8% cross linkage Sigma). Two copper metal reference materials National Institute of Standards and Tech- nology (NIST) Standard Reference Material (SRM) 454 Unal- loyed Copper XI and Bundesanstalt fur Materialforschung und -priifung (BAM) Certified Reference Material (CRM) 361 Copper were obtained from the Bureau of Analysed Samples (Middlesbrough UK).The microcolumn was packed under suction with an aqueous slurry of Dowex 50W-X8-200 resin (250 mg of dry resin). Study of Operational Variables Operating conditions for analysis procedure Prior to use the resin was conditioned by pumping a solution of hydrochloric acid (1.2 mol dm-3) through the column for ap- proximately 5 min at a flow-rate of 2.0 ml min-I. The hydride-generation manifold and matrix-isolation unit were operated according to the optimized variables shown in Table 1. The optimized variables were obtained using the same optimization procedure as Riby et al.s4 In obtaining the optimum variables sensitivity was a prime concern but con- sideration was also given to precision interference tolerance and reliability of the system.A more rigorous approach to op- timization could have been carried out but it was decided that it was not necessary to prove the benefits of the manifold design. Standard solutions requiring no matrix isolation were pumped continuously through the injection valve bypassing the microcolumn. The injection valve was activated intermit- tently (load time 20 s; and injection time 10 s) thereby sam- pling the standard solution and introducing it into the hydride- generation manifold. Following the introduction of the standards the microcol- umn was switched into the sample line and analytical-reagent grade water was continuously pumped through it to remove the hydrochloric acid.After washing the column the sample was also introduced by continuous pumping. In order to fill the void volume of the column and pump tubing the sample was pumped continuous- ly for 60 s prior to sampling of the column eluent as described for standard solutions. After triplicate injections the microcol- umn was regenerated (using HCI) and the sample line washed with analytical-reagent grade water. [Hydrochloric acid ( 1.2 rnol dm-3) was pumped continuously through the microcolumn in the opposite direction to that of the sample flow (2 ml min-1 30 s).] Following the short period of regeneration the column was switched back in-line and the sampling procedure re-JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL.6 n a 0.2 0.1 135 - - Table 2 Summary of the sampling procedure for matrix isolation including timing sequences Timing sequence ( I ) / S Operation 0 20 80 100 I10 170 200 Vs* HzO pumped ria column V sample pumped ria column V,f activation (load position) V activation (inject position) V activation (load position) injection procedure repeated to give triplicate injections V,$. HCI regenerant pumped iiu column H,O pumped iiu sample line Vs H20 pumped ria column procedure repeated I = 0 *Vs Valve containing column in sampling configuration. tV Sample injection valve. SV Valve containing column in bypass configuration. peated. During the transference of the sample uptake tube from sample to analytical-reagent grade water the sample pump was stopped in order to prevent the introduction of air into the column.The hydride-generation manifold was em- ployed continuously throughout the whole procedure indepen- dently of the matrix-isolation unit. A summary of the sampling procedure including timing sequences is shown in Table 2. Digestion procedure The methodj-7 used for the digestion of the two copper metal reference materials NIST SRM 454 and BAM 361 was as follows. A sample of copper metal (0.5 g) was accurately weighed into a clean Pyrex beaker (50-100 ml). To the beaker were added 10 ml of 8 rnol dm-3 HNOj and the beaker was covered with a watch-glass The beaker was placed on a hotplate and the contents were heated to near dryness then removed from the hotplate. Once cool 10 ml of 6 mol dm-3 HC1 were added and the beaker was placed in a steam-bath.After approximate- ly 15 min of heating in order to dissolve the sample residue the resultant solution was allowed to cool. The cool digest so- lution was transferred into a calibrated flask (100 ml) and diluted to volume with analytical-reagent grade water. Prior to analysis the NIST SRM 454 and BAM 361 digests were further diluted to produce working samples containing 75 and 1000 pg ml-1 of copper respectively. Results and Discussion Optimization of FI Hydride-generation Method The effect of the argon carrier gas flow-rate over the range 200-600 ml min-1 was studied. Increasing the argon flow-rate caused a significant increase in peak-height absorbance and rate of transport of the hydride to the atomization cell.The performance of the manifold was significantly affected by the condition of the silica atomization T-cell. Performance characteristics varied from one cell to another depending on age and analytical history. Previously unused T-cells showed very poor performance characteristics. This problem was over- come to some extent by conditioning the surface of the silica The silica cells were soaked in a solution of hydrofluo- ric acid (5% v/v) for a period of 4 h in order to etch the surface and therefore aid hydride atomization. For the new silica cells in particular an improvement in performance was observed with repeated use and after processing relatively high concen- trations of hydride. At the start of each analysis a standard so- lution of lo00 ng ml-l of Se"' was injected two or three times.This procedure enhanced both the precision and sensitivity of the system for all subsequent determinations. 0 2 4 6 8 10 12 14 16 Flow-rate/mI rnin-' Fig. 2 Effect of the flow-rate of NaBH4 solution on the absorbance of 100 ng ml-' of SeIV. Concentration of NaBH 1 .O% m/v; concentration of HCl 3.6 mol drn-j. All other variables as in Table 1 0.4 - $ 0.3 - C Q e s 2 0.2 - 0.1 - 0 1 2 3 4 5 NaBH concentration (% m/v) Fig. 3 Effect of the concentration of NaBH4 solution on the absorbance of 100 ng ml-I of Se'". Flow-rate of NaBH4 3.2 ml min-I; concentration of HCl. 3.6 rnol dm-3. All other variables as in Table 1 The effect of the hydrochloric acid flow-rate on the sensitiv- ity for SeIV was observed to be negligible over the range 2-14 ml min-I.This observation was unexpected and to some extent unexplained. The increase in the hydrochloric acid flow-rate which was predicted to reduce signal response by dilution appears to be offset by an increased rate of transport through the gas-liquid separator. The hydrochloric acid concentration was also observed to have little or no effect on the selenium signal response over the range of 1.2-7.2 mol dm-3. The optimum hydrochloric acid concentration was decided after consideration of the tolerance of the system to the copper in- terference which is reported later. Both the flow-rate and concentration of sodium tetrahydro- borate were observed to have a significant effect on the sensi- tivity of the system as shown in Figs. 2 and 3 respectively.The relationship between the flow-rate of the aqueous carrier solution and the resulting peak-height absorbance is shown in Fig. 4. It can be seen that to achieve maximum sensi- tivity the carrier flow-rate should be kept as high as possible. The optimum carrier flow-rate was decided upon after con- sideration of the tolerance of the system to copper interfer- ence which is reported later. The effect of the sample injection volume on the sensitivity of SeIv is shown in Fig. 5. Increasing the injection volume gave rise to an increase in sensitivity until the steady-state limit was reached at approximately 700 pi. Although high in-136 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 0.4 - 2 0.3 - C e 2 2 0.2 - O-l I ~~~ 0 2 4 6 8 1 0 1 2 1 4 Flow-rate/ml min-' Fig.4 Effect of the flow-rate of the water carrier on the absorbance of 100 ng ml-I of SeIV. Concentration of HCI 3.6 mol dm-3. All other varia- bles as in Table 1 0.02 I I I I I J 0 400 800 1200 1600 Sample injection volume/pl Fig. 5 Effect of the sample injection volume on the absorbance of 20 ng ml-I of SeIV. Concentration of HCI 3.6 mol dm-j. All other variables as in Table 1 Table 3 Manifold performance data for determination of selenium H20 carrier Characteristic flow -rate/ Injection volume/ concentration/ ml min-' PI ng i d - ' RSD* (%) 6.2 409.3 0.96 (0.39)' 1.3 13.0 409.3 0.79 (0.32)" I .s 13.0 922.3 0.50 (0.46)' 1.6 * Relative standard deviation (n=9). + Characteristic concentration in absolute terms (ng of Se"). jection volumes favoured high sensitivity reducing the injec- tion volume reduced the sample volume requirement and im- proved the throughput.An injection volume of 409yl was used for all further investigative work and this volume pro- duced 90% of the steady state absorbance signal. Performance data for combinations of aqueous carrier flow-rate and injec- tion volume are given in Table 3. Prior to the investigation into matrix isolation. a study was made of the tolerance of the hydride-generation manifold to copper interference. In agreement with the work of Welz and Melcher3 and Welz and Schubert-Jacobs,' the tolerance could be significantly improved by employing high hydrochloric acid concentrations as shown in Fig. 6. An optimum hydro- chloric acid concentration of 6.0 mol dm-3 was chosen for the analysis after consideration of the interference tolerance the I I I 1 1 2 4 6 8 1 0 1 2 cu concentration/pg ml-' Fig.6 Influence of the HCI concentration on the interference of Cu in the determination of 20 ng ml-' of Se'" A 2.4; B 6.0; and C 9.6 mol dm-3 HCI. All other variables as in Table 1 hydrochloric acid consumption and the difficulty in pumping high concentration acids.3-3 The water carrier flow-rate was kept at 6.0 ml min-I in order to reduce the dilution of hydro- chloric acid in the hydride-generation system and to optimize the precision whilst retaining an adequate sensitivity during the proposed analyses. Because of the manifold design the in- jected sample was merged with the hydrochloric acid reagent prior to the sodium tetrahydroborate solution.This order of reagent addition was the reverse of that reported in other systems.33.34 The hydrochloric acid was added prior to the sodium tetrahydroborate solution in order to reduce the time interval between the addition of the reductant and the hydride separation in which the reduction of the interferent could occur. Controlling the reaction time in such a manner as to favour the main hydride-gener!tion reaction enhances interfer- ence tolerance as reported by Astrom.Ix Optimization of the Continuous Flow Matrix Isolation Procedure The Dowex 50W cation-exchange resin was chosen for use in this work because it has been previously reported to be appli- cable to matrix isolation in hydride-generation AAS.16 An investigation was made into the efficiency of the resin incorporated into the continuous flow matrix isolation unit against the variables of sample flow-rate and sample pH.The copper capacity of the column was assessed for the two varia- bles by pumping a 1000 pg ml-I copper standard solution through the column and continuously monitoring the copper concentration of the column eluent by flame atomic absorption spectrometry (FAAS). Breakthrough of the column was judged to have occurred when the copper concentration of the column eluent exceeded 1.0 yg ml-I. The efficiency of the resin improved with reduced sample flow-rate (sample pH 4.0) owing to increased contact time between the copper ions and the active sites of the resin (Fig. 7). Interfacing the matrix-isolation manifold and the hydride- generation manifold with a sample injection valve as report- ed by Nord and KarlbergZx for liquid-liquid extraction with FAAS proved extremely beneficial to the systems performance.As the matrix-isolation unit was independent of the hydride- generation manifold the sample flow-rate through the column could be kept low in order to keep the column efficiency high without having to compromise the sensitivity of the hydride-generation manifold as reported for the system of Riby et U I . ~ ~ in which the microcolumn was situ- ated on-line. By pumping the sample at low flow-rates prob-I37 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991. VOL. 6 a E 4 - a 12 1 I I I I I 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Flow-rate/ml min-’ Fig. 7 Retention of Cu as a function of sample flow-rate through the microcolumn.Sample solution pH 4.0 2t ~~ 0 1 2 3 4 5 Sample solution pH Retention of Cu on the microcolumn as a function of sample Fig. 8 solution pH. Sample flow-rate 2.0 ml min-’ lems with resin compaction in the column and build-up of high back-pressures were eliminated. On the introduction of copper (1000 pg ml-I) to the column resin high concentra- tions of hydrogen ions were eluted from the resin by the process of cation exchange. The drop in pH associated with this exchange process produced a slight shrinkage of the resin but this caused no detrimental effect to the system as the shrinkage was reversed on regeneration of the resin. A sample flow-rate of 2.0 ml min-I through the column was decided upon after consideration of both the column efficiency and sampling rate.At a sample flow-rate of 2.0 ml min-I the 409 p1 sample loop of the injection valve could be filled in 20 s. As shown in Fig. 8 from the relationship between the reten- tion of copper and the sample pH (sample flow-rate 2.0 ml min-I) the efficiency of the resin reached an optimum above pH 2.0. The retention of copper at pH values above 5.0 was not investigated because the addition of ammonia solu- tion in order to obtain such pH values caused turbidity of the samples. Such turbidity was assessed to be a source of analyte loss through coprecipitation.16 The incorporation of the microcolumn within the sample loop of an injection valve permitted the intermittent regene- ration of the column without the need to dismantle the hydride-generation manifold or replace the column as re- ported for previous systems.33.34 By directing the regenerant solution to flow in the reverse direction to that of the sample the possibility of column resin compaction was Table 4 confidence interval) Results of analyses of copper metal reference materials (f95% Certified value for SeIV/ Se‘” found/ Sample clt? g-‘ I4 g-‘ NlST SRM 454 BAM 361 36 f 0.6 479 f 8 476.0 rtr 7.2 37.1 f 0.7 eliminated.During the regeneration procedure water was pumped through the sample line and injection valve to waste in order to wash out the previous sample. During the column regeneration process it was possible to analyse stan- dard solutions which bypassed the column to check for sensitivity drift. After 3 months of continued use and re- generation the performance of the matrix-isolation column was similar to that obtained when the column was initially packed with resin.Analysis of Copper Metal Reference Materials NIST SRM 454 and BAM 361 In order to validate the matrix isolation system analyses of two copper reference materials NIST SRM 454 Unalloyed Copper and BAM 361 Copper was attempted. Prior to analysis the metal digest solutions were adjusted to pH 4.00 with the addition of ammonia solution. This produced some degree of precipitation and significantly reduced the capacity of the column. The reduction in capacity was identified as owing to a competition between copper and am- monium ions for the active sites on the resin. In order to over- come this problem the pH adjustment step was removed from the sample preparation. The pH values of the two analysed samples NIST SRM 454 and BAM 36 I were 2.1 and 1.I re- spectively therefore even without pH adjustment the column capacity was still sufficient to permit triplicate injections of each sample solution before column regeneration (column capacity >5.0 mg of copper). The removal of the pH adjust- ment step had the further advantage of reducing (i) sample preparation time; (ii) the chance of sample contamination; and (iii) the chance of precipitation. Thus the loss of analyte through coprecipitationIh was eliminated. The results of the analyses of the reference materials are given in Table 4. The system was calibrated using aqueous standards (5-50 ng ml-I). Calibration was linear up to at least 50 ng ml-1 of SeIV having a slope of 4.35 x A ng-I ml an in- tercept of 5.1 x The precision of the system based on injections of an aqueous 10 ng ml-I SeIV standard was 1.5% relative standard deviation (RSD) ( n = 12).A characteristic concentration of 1 .O ng ml-I SeIV was routinely achieved. The detection limit of 2.1 ng ml-1 was calculated from the resultant calibration graph? A sample throughput of 17 h-I was achieved with trip- licate injections. The sensitivity of the system compared well with that of the conventional operation of the Philips PU9360 continuous flow vapour system. For steady-state analysis (sample volume 5.0 ml) a characteristic concentration of 0.91 ng ml-I SeIv was re- ported.” For less difficult samples the characteristic concen- tration could be improved by increasing the carrier flow-rate and injection volume (see Table 3).One of the major advantages of this system design in com- parison with others is the high precision obtained. For the system of Riby et a1.34 and Ikeda3j the RSDs quoted for the de- terminations of 10 ng ml-I of As111 and 10 ng ml-I of Selv were 3.0 and 3.8% respectively. These figures are clearly inferior to those quoted in this work. Such poor precision can be attri- buted to crude sample introduction procedures into the two manifolds. In the system of Ikedaz3 a 500 pl sample was intro- duced manually from a Pipetman P-1000. Riby ef al.jJ em- ployed a timed injection procedure pumping the sample A and a correlation coefficient of 0.9992.138 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL.6 continuously into the system over a short period of time (10 s). The timing was carried out manually using a Seiko analogue stop watch which would make the precision very susceptible to operator performance and slight fluctuations in the flow-rate. In the system of Riby et ~ 1 . ~ 3 ~ the incorporation of the matrix isolation on-line required a compromise in the performance characteristics for the matrix isolation and hydride generation. Optimum sensitivity was achieved at a carrier flow-rate of 9.0 ml min-1 and yet the optimum matrix isolation characteristics were achieved at 2-4 ml min-I. Even at a compromise of 6.0 ml min-1 the system was susceptible to high back-pressure and column compaction problems requiring intermittent re- placement of the column.In the proposed system no problem with back-pressure or column compaction was observed as the sample flow-rate through the column was kept low (2.0 ml min-1) thereby optimizing matrix isolation without com- promising the hydride-generation performance. By incorporating the matrix-isolation column within the sample loop of an injection valve the rapid intermittent re- generation of the column took place without interrupting the operation of the hydride-generation manifold. Ikeda”3 reported no column regeneration facility in the system used. The column was simply replaced after every 25 determinations (column capacity 50 mg of Cu). By employing a relatively large column in order to increase the column capacity the sample throughput was restricted to 30 h-I.Column regenera- tion was reported in the system of Riby et but was under- taken manually. (Prior to column regeneration the water carrier line was disconnected from the rest of the hydride- generation manifold and 1 mol dm-3 HCl was pumped through the column to waste.) No indication of sample throughput ca- pabili ties were reported. Another benefit of the proposed manifold design is the potential for automation. By using existing technology full automation of the system is feasible.38 Automation of previous system^^^"^ would be difficult without major modifications being made. By employing the flow injection valve as an interface the matrix-isolation procedure could be applied to other flow in- jection hydride-generation manifolds irrespective of their oper- ating variables gas-liquid separator design or atomization procedures.Conclusion The use of a flow injection valve as an interface between a continuous flow matrix-removal manifold and a continuous flow hydride-generation manifold allows separate optimization of each procedure so that an interference of up to 1 mg ml-I of copper in the determination of selenium could be removed. Previous work16 indicates that with only minor modifications the same system could be used for the determination of other hydride-forming elements in the presence of a variety of inter- fering species such as Ni Fe Co and Ag. Financial support for S.G.O. by the Science and Engineering Research Council (SERC) and both provision of equipment and financial support from Philips Scientific (Cambridge UK) is gratefully acknowledged.1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 References Smith A. E. Analyst 1975 100,300. Meyer A. Hofer C. Tolg G. Raptis S. and Knapp G. Fresenius Z. Anal. Chem. 1919,296,337. Welz B. and Melcher M. Analvst 1984 109,569. Welz B.. and Schubert-Jacobs M. J. Anal. At. Specrrom. 1986 1 23. Bax D. Agterdenbos J. Worrel. E. and Kolmer. J. B. Spectrochim. Acta. Part B 1988,43 1349. Boarnpong C. Brindle I. D. Le X.-c. Pidwerbesky L. and Cecca- relli Ponzoni C. M. Anal. Chem 1988,60 1185. Kirkbright G. F. and Taddia M. Anal. Chim. Acta 1978 100 145. Yamamoto M. Shohji T. Kumamaru T. and Yamamoto Y. Fre- senius 2. Anal. Chem. 198 1,305 1 1.Belcher R. Bogdanski S. L. Henden E. and Townshend A. Analyst 1975,100,522. Peacock C. J. and Singh S. C. Analyst 1981 106,931. Bye R. Engvik L. and Lund W. Anal. Chem. 1983,55,2457. Bedard M. and Kerbyson J. D. Can. J . Spectrosc. 1976,21,64. Jones J. W. Capar S. G. and O’Haver. T. C. Analvst 1982 107 353. Narasaki H. and Ikeda M. Anal. Chem. 1984,56,2059. Narasaki H. Anal. Sci. 1986,2 141. Hershey J. W. and Keliher P. N. Speibtroc-him. Acts. Part B 44,329. Pierce F. D. and Brown H. R.,Anal. Chem. 1977,49 1417. Astrom O. Anal. Chem.. 1982,54 190. Yamamoto M. Yasuda M. and Yamamoto Y.. Anal. Chem. 57 1382. Chan C. C. Y. Anal. Chem. 1985,57 1482. Ikeda M. Anal. Chim. Acta 1985. 167,289. 989 985 Pacey G. E. Straka M. R. and Cord J. R. Anal. Chem. 1986 58 502. Fang Z. Xu S. Wang X. and Zhang S. Anal. Chim. Acta 1986 179,325. Yamamoto M. Takada K. Kumamaru T. Yasuda M. Yokoyarna S. and Yamarnoto Y. Anal. Chem. 1987,59,2446. Pettersson J. Hansson L. and O h A. Talanta 1986,33,249. McLaughlin K. Dadgar D. Smyth M. R. and McMaster D. Analyst 1990 115,275. Sweileh J. A. and Cantwell F. F. Anal. Chem. 1985,57,420. Nord L. and Karlberg B. Anal. Chim. Acta I98 1 125 199. Olsen S. Pessenda L. C. R. R&&a J. and Hansen E. H. Analyst. 1983,108,905. Kamson 0. F. and Townshend A. Anal. Chim. Acta 1983 155 253. Martinez-Jimenez P. Gallego M. and Valcarcel M. Analyst. 1987 112 1233. Adeeyinwo C. E. and Tyson J. F. Anal. Chim. Acra 1988 214 339. Ikeda M.,Anal. Chim. Acta 1985 170,217. Riby P. G. Haswell S. J. and Grzeskowiak R. J . Anal. At. Spec- rrom. 1989,4 181. Welz B. and Melcher M. Analwt 1983 108,2 13. Miller J. C. and Miller J. N. in Statistics for Analytical Chemistry. Ellis Horwood Chichester 1984 p. 96. Philips Scientific Atomic Absorption Data Book. Philips. Cambridge 5th edn. 1988 p. 53. Bysouth S. R.. Tyson J. F. and Stockwell P. B. J . Airtom. Chem.. 1989 11,36. Paper Oi03451 G Recei\*ed July 30th I990 Accepted Septemhei- 19th 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600133

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 139 Effect of Long-chain Surfactants on Drop Size Distribution Transport Efficiency and Sensitivity in Flame Atomic Absorption Spectrometry With Pneumatic Nebulization* Juan Mora Antonio Canals and Vicente Hernandis Division de Quimica Analitica Universidad de Alicante 030 71 Alicante Spain The variations of the droplet size distribution transport efficiency and absorbance which arise as a consequence of the addition of long-chain surfactants to aqueous solutions of Mnll were studied. The results show that the drop size distribution of the primary aerosol does not change on increasing the surfactant concentration irrespective of whether cationic or anionic long-chain surfactants are used. Concomitantly neither transport efficiency nor absorbance change.A mechanism is also suggested in order to explain these results based on the following points ( i ) surfactant molecules require a certain period of time in order to redistribute on the new surface that is being generated during the nebulization step; and (ii) this time increases as the length of the hydrophobic chain increases. Keywords Surfactant; flame atomic absorption spectrometry; drop size distribution; transport efficiency The most common method of sample introduction in atomic spectrometry is through the pneumatic nebulization of sample solutions. In the nebulization step a primary aerosol is generat- ed the characteristics of which will have a great influence on the signal intensity (absorbance) and the degree of some inter- ferences.Obviously in order to obtain a high transport efficiency and a decrease in the interferences it is very important that the primary aerosol be as fine as possible for a given set of gas and liquid flow-rates. Among the physical properties of the solution surface tension is probably the most influential on the characteristics of the primary aerosol. Owing to the high value for the surface tension of water many workers have suggested the addition of surfactants to the aqueous solutions in order to decrease their surface tension and hence. to improve transport efficiency and/or atomization efficiency. Some workers have found noticeable sensitivity im- provements,'-I however others have found little or none.I2-I4 It is clear that there is a certain controversy about this matter.Several mechanisms sometimes contradictory have been proposed in order to explain the sensitizing effect of the sur- factants in flame atomic absorption spectrometry (FAAS). Kodama and Miyagawa7 observed that an increase in surfac- tant concentration yields finer aerosols until the critical micel- lization concentration (CMC) is reached. These finer aerosols yield finer solid particles in the flame thus giving rise to an in- crease both in transport and in atomization efficiencies. Ac- cording to these workers results there is no influence from the surfactant or analyte charge [sodium dodecyl sulphate (SDS) or dodecyltrimethylammonium chloride Cr"' or CrVIJ. proposed a model which is an enlarge- ment of the aerosol ionic redistribution model previously intro- duced by Borowiec ef a1.,Is to explain the improvement of the sensitivity they observed on adding increasing amounts of SDS to solutions of Cu". They claimed to find sensitivity im- provements of up to 44% for SDS concentrations slightly below the CMC.Similar results were obtained for other anionic surfactants. The mechanism they proposed assigned a great influence to the surfactant charge. Thus with the analyte used neither cat ionic (hexadec y 1 trime t h y lammon i um bromide CTAB) nor non-ionic (Triton X- 100) surfactants modify the analytical signal although the surface tension values reached in each instance are very similar. The mechanism they pro- posed to account for this signal improvement is based on the migration of surfactant molecules to the droplet surface.If the Komahrens e f * Presented at the Fifth Biennial National Atomic Spectroscopy Sympo- sium (BNASS) Loughborough UK I8th-20th July 1990. ionic (hydrophilic) ends of the surfactant molecules are oppo- site in charge to the analyte ions then the analyte ions will tend to associate with the surfactant molecules and the surface of the large droplets will become analyte enriched. When these large droplets break apart the smallest droplets formed will be analyte enriched. As these small droplets are more effectively sampled a signal enhancement is observed. Venable and Ballads explained the signal variations in the presence of surfactants by the ability of some of them to keep the analyte in solution under conditions where otherwise it could precipitate. Recently Yan and Zhangii have proposed a mechanism based on the formation of reverse micelles during the nebuliza- tion process.If the hydrophilic chain of the surfactant is oppo- site in charge to the analyte then the latter will be attached to the hydrophilic end of the surfactant within the droplets. The micro-environment created by the surfactant molecules forming the micelle around the analyte ion will cause the atomization efficiency to increase because of the reduction processes involving the products of the decomposition of those molecules thus giving an improvement in the sensitivity. In this model the charges of the surfactant and analyte contribute to the improvement. In general cationic surfactants give rise to improvements in the signal owing to anionic analytes and \,ice wi-sa whereas non-ionic surfactants have no effect.These workers state that the surface tension decrease is of no significance in the improvement. Farino and Browner,' on comparing the behaviour of organic solutions with that of aqueous solutions of surfactants concluded that in FAAS the sensitizing effect depends on the nebulizer-spray chamber design more than on the surfactant charge. For instance using a paddle spray chamber they found signal enhancements of between 0 and 8% whereas with an impact bead spray chamber the enhancements were between 0 and 20%. From all these examples it appears that there is a certain controversy among the published results not only on the mag- nitude of the sensitizing effect but also on the significance of the surfactant and analyte charges or on the enhancement mechanism (transport or atomization). Another point to be noted is that in many of these papers some relevant informa- tion is missing and one or several important parameters have not been properly controlled.Therefore this paper aims to clarify the mechanism of the surfactant effect in FAAS. To this end the drop size distribu- tions of the primary aerosols generated in the nebulization step the transport efficiencies for analyte and solvent and the analytical signal have been measured.140 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 Manganese(l1) was chosen as the analyte because in FAAS the sensitivity of this cation is only slightly altered by the com- position of the flame.Thus minor variations in the nature of the flame caused by the presence of the surfactant molecules will not significantly affect the efficiency of the analyte atomi- zation. The surfactants employed are two typical long-chain surfac- tants one anionic (SDS) and the other cationic (CTAB). Experimental Aqueous solutions of Mn" 2 pg g-' in 1% v/v HCl were em- ployed throughout. The surfactants were SDS and CTAB both from Fluka with a degree of purity of about 98%. An adjustable concentric nebulizer (Perkin-Elmer) locked at the position of maximum interaction between gas and liquid was used in all the experiments. Paddles were used as impact surfaces within the spray chamber. Liquid flow was controlled by means of a peristaltic pump (Gilson Minipuls 2) and kept constant at 4.5 ml min-I throughout the experiments.The addition of surfactants to aqueous solutions might modify their viscosity which in turn might modify the uptake rate. The use of a peristaltic pump avoids this potential problem. It was considered that it might be the lack of control of the uptake rate that may effectively modify the experimental results. However the pumping rate chosen lies close to the natural (normal) uptake rate of the neb- ulizer operated at 5.65 1 min-' with air. When evaluating the effect on the nebulization step of increasing surfactant concen- tration it was necessary to ensure that the liquid and gas flow- rates were kept constant. The air flow for nebulization was kept constant at 5.65 I min-I by means of a precision flow meter (Cole-Pamer).Similarly total air flow and gas flow were kept constant throughout the experiments. Transport efficiencies the percentage of solvent (E,) and analyte (q,) taken up that reaches the flame were measured by an indirect method (in the batch version). A drawing of the device employed is shown in Fig. I in the drain collec- tion position. Firstly the solution K is sprayed for 2-3 min in order to condition the spray chamber A. During this time the valve C is placed in the drain waste position. In this po- J r l Fig. 1 Device employed for the measurement of the transport efticiency by the indirect method (shown in drain collection position). A. Spray chamber (front view) B. nebulizer C . drainage outlet of the spray chamber D. drainage waste outlet E.drainage collection outlet; F. tube to allow the loop to continue being the closure system of the spray chamber during the drainage collection period; G. valve; H. tube for drainage col- lection; 1. loop J . waste K. solution and L. peristaltic pump sition the drainage flows through tube C to tube D then to the loop I and finally on to waste J. After this the valve is turned through 90" in the clockwise direction to the drain col- lection position and the drainage is collected in a previously weighed tube H. When in this position the drainage flows by gravity through tubes C and E to the drain collection tube. The tube F serves to connect the spray chamber to the loop which then continues to act as the closure system of the spray chamber and thereby damps possible pressure variations in the spray chamber.With this system the spray chamber works the same way irrespective of the valve position. The drainage is collected for 10-15 min after which the valve is switched back through 90" in the counterclockwise direction tlo the waste position again. The collection tube is disconnect- ed and re-weighed. From these results the values for E and the total solvent transport rate (Stol) i.e. the amount of solvent reaching the flame (in ml min-I) are calculated. The c value and the total analyte transport rate (Wtot) value i.e. the amount of analyte reaching the flame (pg min-I) are ob- tained by comparing the absorbances of the drained with the original solutions. Drop size distributions for the primary aerosols were meas- ured 28 mm from the nebulizer tip by means of a laser Fraun- hofer diffraction system (Malvern Instruments Model 2600~) the measurement range of which was 1.9-188 pm.The calcu- lations were performed with a model-independent algorithm using software version M5.4.I6 A Perkin-Elmer 373 atomic absorption spectrometer equipped with a hollow cathode lamp was used. The instru- mental conditions used are shown in Table 1. Results and Discussion Drop Size Distributions for the Primary Aerosols 'The drop size distribution (in volume) of the primary aerosols obtained with water and with the highest surfactant concentra- tions above their respective CMC are shown in Figs. 2 and 3. From these results it is apparent that under the experimen- tal conditions employed the distributions are not significantly different from one another.This means that the surfactants em- ployed are ineffective in modifying the aerosol drop size distri- butions. The same conclusions can be drawn from Table 2 which shows the most significant distribution parameters of the primary aerosols obtained with these solutions although there are important decreases in their surface tension values as the surfactant concentration increases. Obviously the surface tension values measured under surface equilibrium conditions (as determined here) are ineffective for the generation of a greater surface area. However in Table 2 some slight though unequivocal ten- dencies can be observed. There are slow increases in droplet size and obscuration with increasing surfactant concentration for both SDS and CTAB.The explanation for these behavi- ours cannot be linked to the surface tension decrease (as then the size tendency would be just the opposite) or the small viscosity increase as a coarser distribution would give rise to lower obscuration values (the amount of light diffracted Table 1 Instrumental conditions used for the FAAS determinations Parameter Setting Wavelength Slit-width Lamp intensity Height above burner Acetylene flow-rate Air flow-rate (total) Air flow-rate (nebulizer) Integration time 279.5 nm (Mn I ) 0.2 nm 35.0 mA 8.0 mm 2.7 I min-' 19.6 I min-' 5.65 1 min-' 5.0 sJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 141 0 1 2 3 4 5 Ln ( i/p m I 0 1 2 3 4 5 Ln ( d/p m I Fig. 2 tions of variable concentration.A 0 mmol dm3; and B 1 .S mmol dm-j Drop size d distribution of the primary aerosol for CTAB solu- by a given volume of sprayed solution increases as the mean drop size of the distribution decreases). In our opinion the variations are a result of the fact that once a droplet has been generated the surfactant molecules cover its surface almost completely (depending on the surfactant concentra- tion). This causes the solvent evaporation rate to diminish as the surfactant concentration increases. This is related to the hydrodynamic effect of the surf act ant^.'^ Therefore if the be- haviour of two solutions are compared one free of surfac- tants and the other with a given surfactant concentration both would have the same distribution at the outlet of the nebulizer nozzle but then the droplets of the first aerosol will undergo a relatively rapid initial evaporationIx that leads to a small size reduction on their way towards the laser beam (placed 28 mm from the nebulizer tip) whereas the droplets of the second aerosol would hardly undergo any variation in their size because their evaporation rate is much slower.Fig. 3 of variable concentrations. A 0 mmol dm-? and B 8.5 mmol d m 3 Drop size d distribution of the primary aerosol for SDS solutions Surface Tension Effect Surface tension values for aqueous solutions of surfactants are static values when obtained under surface equilibrium condi- tions. These values might not apply in fast processes such as pneumatic nebulization which take place in milliseconds or less.For this reason it might be advisable in these instances to refer to the ‘dynamic’ surface tension values.I9 In surfactant solutions surface tension is a property in which a certain period of time is required for the surfactant molecule to migrate from the bulk of the liquid or from the previously ex- isting surface to the surface that is now being generated and to distribute and orientate on it.?() Thomas and Potter” state that SDS requires about 10 ms to migrate to the new surface and to distribute on it in order to modify the surface tension of the solution effectively. WestI9 found that an increase in Triton X- 100 concentration does not noticeably affect the emission signal of Ca when using a direct injection burner Table 2 Parameters of the drop size distributions for the primary aerosols of the surfactant solutions Surface cm*/ tension/ DW+/ D4.# D3.N Distribution mmol d d N m-’ x lo3 Pm Pm Pm span1 Obscuration I1 CTAB- 0.1880 0 70.43 14.2 17.6 8.2 I .8 0.2 0.4 0.8 1.5 SDS- 0 2.0 4.0 7.5 8.5 40.93 35.95 34.22 29.86 70.43 5 1.68 44.39 35.92 34.23 4.4 4.5 4.7 4.8 4.2 4.3 4.6 4.9 5 .o 18.4 18.6 18.4 19.5 17.6 17.3 18.0 19.2 19.8 8.3 8.2 8.3 8.4 8.2 8.I 8.3 8.4 8.5 1.9 1.9 1.8 1.9 1 .8 1.8 I .8 I .9 3.0 0. I985 0.1974 0.20 I6 0.2024 0.1880 0.2037 0.207 I 0.2245 0.2272 * (’” = Surfactant concentration. i- Dso = Droplet distribution diameter below which 50% of the cumulative aerosol volume is found. Hence. Dy( and D l o = 90 and 10%. respectively. $ DJ,3 = Statistical diameter the diameter of the droplet whose mass is equal to the average mass of the distribution (mass mean diameter).8 DJ.? = Statistical diameter. the diameter of the droplet whose surface is equal to the average surface of the distribution (surface mean diameter). also 1 Distribution span [(D% - DIo)/Ds(,]. A measurement of the distribution spread. II This value is based on the measured laser intensity and indicates the proportion of light which is being scattered out of the beam by the sample. I t will known as Sauter mean diameter. depend on the amount of sample added (see reference 16).142 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 Table 3 Values of the transpon parameters and absorbance for the surfactant solutions CTAB- 0 70.43 0.4 1 9. I 0.4 1 4.6 0.1 so 0.8 34.22 0.39 8.7 0.37 4.1 0.153 I .s 29.86 0.40 8.9 0.39 4.3 0.IS2 0.2 40.93 0.42 9.3 0.43 4.8 0.149 0.4 35.95 0.4 1 9.1 0.4 I 4.6 0.149 SDS- 0 70.43 0.4 1 9. I 0.4 1 4.6 0.1 50 2.0 5 1.68 0.43 9.5 0.45 5.0 0. I47 4.0 44.39 0.44 9.7 0.47 5.2 0. IS6 7.5 35.92 0.4 1 9.1 0.4 1 4.5 0. I53 8.5 34.23 0.39 8.7 0.37 4.1 0.149 although the surface tension values show a net decrease. The explanation being that the surfactant molecules have insufficient time to reach equilibrium on the new surface during the time that the primary aerosol is formed. Addison,’* working with Cs-Cx aliphatic alcohols found that the time necessary to reach a given equilibrium surface tension in- creases with increasing chain length and with decreasing con- centration. The process of pneumatic droplet generation in AAS is clearly too rapid for the molecules of the most common long- chain surfactants to move to the surface being generated and they will not be able to modify the droplet formation process.This is again related to the ‘hydrodynamic effect’ of the sur- factant solutions,~7 which causes a local transient increase in the surface tension when a volume of liquid in contact with the surface is replaced with liquid coming from the bulk of the so- lution. The local surface tension at this moment in time is about the same as for the solvent alone. Hence it is reasonable to assume that the drop size distribu- tion of the primary aerosols for water and for aqeuous solu- tions of long-chain surfactants should be similar except in relation to the evaporation rates as discussed above.Transport The amount of analyte that will finally reach the atomization cell is dependent on the primary aerosol characteristics for a given spray chamber configuration. Of the whole aerosol only the smallest droplets go through the entire spray chamber carried by the gas stream the remainder of the aerosol will be unable to follow the gas flow-lines and will go to waste. It seems clear that if under the same experimental conditions two solutions of equal analyte concentration give rise to similar primary aerosols they will undergo the same type and extent of liquid losses and hence they will transport the same amounts of analyte and solvent to the flame. Table 3 shows that within the poor precision limits inherent in the in- direct method employed for the transport measurement this is true.Thus it can be seen that transport rates do not improve at least to a significant extent when surface tension goes from 0.070 N m-I for water to 0.030 N m-I for 1.5 mmol dm-3 of CTAB. Signal Given that Slol and W, values do not change in the presence of long-chain surfactants (of cationic and anionic nature) at con- centrations of greater and less than their CMC and given that the characteristics of the tertiary aerosol will also remain fun- damentally unchanged,lx the absorbance is unlikely to depend on the surfactant concentration of the solutions. Table 3 shows the results of the absorbance of the solutions at 279.5 nm (,Mn I). As can be seen the absorbance values remain nearly constant irrespective of the nature or concentration of the surfactant.Although the experimental conditions are not in general strictly comparable the results seem to contradict those of other workers,G‘x.l I which have shown noticeable signal im- !provements brought about by the use of long-chain surfac- tants. However some 9.12-14 workers have claimed not to have observed any significant improvement. The results shown in this work agree with those of the second group of workers and suggest a possible explanation for this behav- iour. The use of spray chambers in which re-nebulization is significant could favour the action of surfactants given that the re-nebulization (i.e. at the impact bead surface) can be much slower than the primary nebulization. This could explain the results obtained by Farino and Browner.9 On the other hand it could also be possible for short-chain surfactants in more concentrated solutions to have sufficient time to redistribute on the surface being generated as their equilibration time is shorter; thus they could have a tensioactive effect in the nebulization step.” Although presently embarking on a systematic study on the effect of short-chain surfactants in aqueous solutions we can say that the additon of 1 % m/m of pentanoic acid to an aqueous solution of MnlI yields a net decrease in the mean drop size of the primary distribution and as expected parallel improvements (40-50%) in transport and signal. Conclusions The presence of long-chain surfactants of different charge in variable concentrations decreases the surface tension values of their aqueous solutions but in spite of this they do not modify the characteristics of the pneumatically generated aero- sols.This is attributable to the ‘hydrodynamic effect’ of the surfactants which takes place during the very short period of time required for the nebulization. Surface-active molecules require some time to migrate to the surface. This time increas- es with increasing chain length and with decreasing concentra- tion. With long-chain surfactants this time is so great that they do not influence nebulization. Similar primary aerosols should lead to similar S,, and W,, values and also to similar absor- bance values. The last effect will be true only if the surfactant does not modify the atomization efficiency which seems to be the situation for Mn”.Short-chain surfactants require a higher concentration than long-chain surfactants to reach a given surface tension value,JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991. VOL. 6 143 but their surface equilibration time seems to be shorter. Hence it is possible that short-chain surfactants do really modify primary aerosol characteristics thus influencing transport variables and signal. This point is currently under investigation in our laboratories. The CICYT (Spain) is acknowledged for financial support (Grant No. PB88-0288). J.M. expresses his appreciation to the Instituto de Estudios Alicantinos Juan Gil-Albert (Diputacion de Alicante Spain) for his scholarship. References Nukiyama. S. and Tanasawa Y.. Experiments on Atomi:arion of Liquids in an Air Stream Defence Research Board Department of National Defence Ottawa Canada 1950.Browner. R. F. Boom A. W.. and Smith D. D. Anal. Chem. 1982 54 1411. Gustavsson. A. Anal. Chem. 1983.5594. Canals A. Wagner J. Browner R. F. and Hemandis V. Specfro- chim. Acta. Part B 1988,43 132 1 . Venable R. L. and Ballad R. V. Anal. Chem. 1974,46 13 1 . Kcdama M. Shimizu M. Sato M.. and Tominaga T. Anal. Letf. 1977. 10,591. Kodama M. and Miyagawa S. Anal. Chem. 1980,52,2358. 8 9 10 1 1 12 13 14 15 16 17 18 19 20 21 22 Komahrens H. Cook K. D. and Armstrong. D. W. Anal. Chem. 1982,54 1325. Farino J. and Browner R. F.. Anal. Chem. 1984,56,2709. Ward T. J. Armstrong D. W. Czech B. P.. Koszuk. J. F. and Bartsch R. A. Anal. Chim. Acra 1986 188 301. Yan Z. Y. and Zhang W.. J. Anal. At. Spectrom. 1989.4 797. Pungor E. and Mahr M.. Talanra. 1963 10,537. Stupar J. and Dawson J. B.. AppI. Opt. 1968 7 I35 1 . Lockyer R. Scott J. E. and Slade S.. Nature (London) 1961 189 830. Borowiec J. A. Boom A. W. Dillard J. H.. Cresser M. S.. Browner R. F.. and Matteson M. J. Anal. Chem. 1980.52 1054. 2600 Particle Sizer User Manual Malvem Instruments Malvem UK 1987. Mans C.. Llorens. J. and Costa J. Invest. Cienc. (Span. Tranl. Sci. Am.) 1988,136,78. Canals A. Hemandis V. and Browner R. F. Specrimhim. Acta. Part B 1990,45,59 1 . West A. C. Anal. Chem. 1964,36,310. Sharp B. L. J. Anal. At. Specworn. 1988,3,613. Thomas W. D. E. and Potter L. J. Colloid Interface S c i . . 1975 50 397. Addison C. C. J. Chem. Soc. 1945,98. Paper Oi02707C Received June 18th 1990 Accepted October I6th 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600139

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 145 Determination of Trace Elements in Solid Plastic Materials by Laser Ablation-Inductively Coupled Plasma Mass Spectrometry* John Marshall and Jeff Franks ICI plc Wilton Materials Research Centre P. 0. Box 90 Wilton Middlesbrough TS6 8JE UK Ian Abell and Chris Tye VG Elemental Ion Path Road Three Winsford Cheshire CW 7 3BX UK The application of laser ablation inductively coupled plasma mass spectrometry (ICP-MS) to the direct determina- tion of trace elements in solid plastic materials has been studied. A neodymium:yttrium aluminium garnet (Nd:YAG) laser was used to ablate the samples and the particulate matter thus produced was transported to the plasma torch of the ICP-MS system for measurement. The system was applied to the examination of polypropy- lene polyester poly(viny1 chloride) nylon and polyethylene containing a variety of fillers and other additives. It was found that by using carbon-13 as an internal standard in order to adjust for variations in ablation and trans- port of the different sample types semi-quantitative analysis could be achieved with an accuracy that was within a factor of 2 of the known value for most of the elements investigated.Quantitative measurements made using laser ablation-ICP-MS and with calibration standards of matrix composition similar to that of the materials being ana- lysed showed good agreement with known values. The application of the technique to the investigation of the spatial distribution of elements is briefly explored.Keywords Inductively coupled plasma mass spectrometry; laser ablation; plastics; solid sampling; trace element determination The inductively coupled plasma (ICP) is an extremely versa- tile spectrochemical source providing efficient atomization excitation and ionization and thereby facilitating trace element analysis using optical' and mass spectrometric techniques.'.' Nevertheless despite the widespread use of the ICP for the analysis of liquid samples no generally applicable procedure has been devised for the introduction of solids to the dis- charge. It is possible to volatilize solid samples by using directed laser en erg^.^ Several laser ablation sampling systems have been designed for the purpose of optical emission measure- ment either or in conjunction with an auxiliary ex- citation source such as a spark,' direct current plasma,x hollow cathode' or an ICP.1(hi3 Some earlier work was carried out with ruby laser^,^ and these continue to be employed,I4 but ablation studies have also been reported involving the use of nitro- gen,Is neodymium:yttrium aluminium garnet (Nd:YAG)".I6 and continuous wave Ar 1ase~s.l~ In recent years ICP mass spectrometry (ICP-MS) has been employed increasingly for trace element analysis.' The tech- nique offers better sensitivity than does ICP optical emission spectrometry (ICP-OES) with attendant spectral simplicity and these advantages are particularly relevant in solid Sam- pling applications.However despite initial studies which in- dicated the potential of the laser ablation-ICP-MS combination for the direct analysis of relatively few papers have been published on this subject. The analysis of polymer based composite materials by atomic spectrometry can involve rather lengthy sample prepa- ration procedures." Thus although ICP-MS offers analytical advantages (e.g.sensitivity range of elements covered and semi-quantitative scanning capability for the identification of unknowns) in this type of application limitations exist because of the requirement to present the sample to the instrument in liquid form. This problem is overcome to some extent by the use of rapid wet oxidation procedures carried out by means of microwave digestion." However this approach cannot be applied for all elements and the efficiency may vary for some * Presented in part at the Fifth Biennial National Atomic Spectroscopy Symposium (BNASS).Loughborough. UK. I8th-20th July. 1990. analytes depending on the nature of the acids used and the con- stituents of the sample matrix. Consequently other traditional procedures such as dry or wet ashing or fusion may have to be adopted for particular elements and this effectively diminishes the advantage of the multi-element analysis capability provided by ICP-MS. X-ray fluorescence spectrometry may be used for the direct examination of polymeric materiakZ3 However sen- sitivity limitations particularly for elements with a low atomic number and a lack of reference standards for calibration neces- sitate the use of solution based determinations (e.g. by flame atomic absorption spectrometry ICP-OES or ICP-MS).The use of laser solid sampling in conjunction with a multi- element detection facility such as that provided by ICP-MS would appear to have potential for the analysis of polymeric composite materials more commonly known as plastics. In the present work the application of a Nd:YAG laser ablation- ICP-MS instrument to the qualitative and quantitative determi- nation of trace elements in a number of common plastic mate- rials is described. Experimental Instrumentation A commercially availably VG Laserlab system was used to ablate samples. This system consisted of a Xe flash tube pumped Nd:YAG laser with an output at the fundamental wavelength of 1064 nm and associated optics and electronics. The laser could be operated in either fixed Q or Q-switched modes.In the present work all measurements were made under Q-switched operation with a laser energy of 0.2 J per shot. Other laser instrumental parameters are given in Table 1 . Radiation from the laser was folded through 90" using mirror optics onto a 75 mm focal length lens and thence through the angled top (45" to minimize reflection) of the quartz ablation cell onto the surface of the sample (see Fig. I ). The cell had an inlet at the base to allow the carrier gas argon to transport the ablated material out via a vent at the top to the injector tube of the standard ICP torch. A solenoid valve located in the gas line to the ICP was used to re-direct the carrier gas to the vent to allow access to the ablation cell for the removal and/or in-146 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL.6 - Table 1 Laser ablation parameters for sampling of plastics \ I Nd:YAG laser r\ Mirror Laser mode Flash lamp voltage Beam focus position Laser energy Raster positions Raster area Shots per position Aerosol path length Carrier gas flow Q-switched 800 v 10 mm de-focused 0.2 J per shot 6 x 6 8 x 8 m m 5 2.5 m I .O I min-' To video Camera monitor I To vent TO ICP-MS TO ICP-MS Fig. 1 Schematic diagram of laser ablation system stallation of samples. Samples were mounted on a holder clamped to the polytetrafluoroethylene base of the ablation cell. It was normal practice to purge the cell for at least 1 min following sample placement in order to remove any atmo- spheric contamination. The position of ablation on the sample was adjusted by means of computer controlled stepper motor driven translational x-y stages on which the cell was mounted. The laser and the ablation chamber were totally enclosed and fully interlocked to prevent any possibility of operator expo- sure to laser radiation.A Sony charge coupled device (CCD) camera was mounted above the ablation cell to permit the op- erator to view the sample ria a video monitor. A shutter system was used to prevent the CCD camera being exposed to high intensity radiation during ablation events. The hardware could be controlled manually via an electronic console or re- motely via the PlasmaQuad software. A VG PlasmaQuad 2+ ICP-MS instrument was used for all measurements. The conditions used are given in Table 2. The PlasmaQuad instrument software has a laser data acquisition module which allows the user to define the measurement pro- cedure in terms of number and spacing of ablation positions number of repeat shots at each position element isotopes ac- quisition delay time and dwell and spread times for a given in- tegration period.Measurements were made in the scanning mode with post-run data manipulation being handled lia the calculations software module. Sample Presentation and Calibration The plastic material was moulded in a hot press and presented for analysis as discs of 40 mm in diameter and approximately 5 mm in thickness. The ICP-MS ion lenses were tuned to give a good elemental response for the ablation of a solid glass standard Standard Reference Material (SRM) 6 12 obtained Table 2 ICP-MS operating parameters for laser ablation of plastics ICP radiofrequency forward power Carrier gas flow Auxiliary gas flow Coolant gas flow No.of channels No. of scan sweeps Dwell time Detector mode 13O W 1 .O I min-' 0.6 1 min-' 13 1 min-' 2048 100 160 ps Pulse from the National Institute of Standards and Technology (NIST) and this was stored for use in the semi-quantitative rneasurement mode. Adjustment of the laser parameters was used to optimize the intensity of the ion response for plastic samples. Signals resulting from a 6 x 6 position computer con- trolled raster (covering a surface area of about 8 x 8 mm) with five ablations per position (see Fig. 2) were integrated over a 33 s period to give an overall ICP-MS intensity value for that sample.For semi-quantitative work carbon- 13 was used as an internal standard as this was considered to be representative of ablation of the bulk matrix. The laser ablation element re- sponse graph stored on a computer disk in the instrument was then used to generate estimated semi-quantitative concentra- tion data. For quantitative measurements solid samples of known content were ablated in turn and the responses ob- tained were passed into the quantitative calibration software module of the PlasmaQuad. The content of unknown samples was then estimated against the calibration function thus derived. Results and Discussion Sample Ablation It was possible by suitable selection of laser power and focus i;o ablate a range of plastic materials in a controlled manner. Photographs of the ablation sites produced by the laser in a glass filled nylon sample are presented in Fig.2(a and h). A ;raster of ablations was carried out on the sample at 6 x 6 posi- tions and five shots were made at each position. Circular evenly spaced ablation craters were generated in a grid pattern Ion the surface of the sample as shown in Fig. 2(a). This photo- ,graph was taken at a 20-fold magnification. A further magnified photograph at one of these sites is presented in Fig. 2(h). The scale is indicated by lines at the base of the photo- graph each line corresponding to a distance of 0.1 mm. The width of each crater is of the order of 0.7 to 0.8 mm in size. Since the focus of the system was set to give a laser spot size of about 0.2 mm it is clear that a significant amount of ma- terial around the site has also been volatilized. However the area surrounding the craters is relatively free from condensate indicating that the sample is removed fairly cleanly from the ablation site.Straight lengths of glass fibre coated with melted polymer can be observed in the crater itself. For some matrices it was found that the polymer melted and diffused away from the centre of the ablation site leading to the formation of wells in the sample surface. Under extreme conditions charring of the polymer matrix could take place. Such problems were more noticeable for unfilled materials. However these difficulties could be eliminated simply by re- ducing the laser power. Visual inspection of the sample was found to be a reliable means of ascertaining to a first approxi- mation the suitability of the conditions employed. Some experiments were carried out in an effort to determine the amount of material volatilized from a plastic matrix during a laser ablation event typically employed for bulk analysis (See Table 1).A sample of nylon was ablated in a 6 x 6 square raster using five points per shot. The sample was weighedJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 I 147 P A Fig. 2 (a) Photograph of ablation sites in glass filled nylon and ( h ) magnified version of ( u ) . See text for further details before and after the ablation process had taken place. It was found that 300 pg of material were removed from the sample when the laser beam was focused on the sample surface.When the beam was de-focused by about 10 mm (the condition used for the analyses) the amount of material rerioved was greater than about 720 pg. A simple calculation indicates that about 4 pg of material are ablated per shot under the conditions used. Given the sensitivity of the ICP-MS technique for liquid sample introduction the amount of sample being produced by the laser should be sufficient to achieve detection limits in the ng g-' range for the direct determination of trace elements in plastics. Semi-quantitative Analysis One of the main advantages of ICP-MS for the measurement of solutions is that rapid survey analyses can be performed on a semi-quantitative basis simply using the element response graph stored in the instrument computer. It was of interest to investigate the potential of laser ablation-ICP-MS for the anal- agous mode of analysis with respect to solids using a single glass reference material (NIST SRM 612) to calibrate the re- sponse.A survey scan of a poly(viny1 chloride) (PVC) sample introduced into the ICP-MS instrument by laser ablation was performed and the spectral information gathered is presented in Fig. 3. The spectrum has been split into three mass ranges to allow presentation with appropriate ion intensity scales being used for the various analytes. A range of elements generated directly from the solid sample can be identified clearly in this instrument spectrum. Whilst this qualitative approach is obvi- ously useful in the preliminary analytical investigations of a problem the extent to which such signals can be quantified is of greater importance.A sample of polyester containing known amounts of a range of trace elements was examined using laser ablation-ICP-MS. The sample was one of a range of calibration solids used for X- ray fluorescence (XRF) analyses. The ICP-MS instrument was operated in the semi-quantitative mode using carbon- 13 as an internal standard for calibration purposes. The results are pre- sented in Table 3. All of the laser ablation-ICP-MS data were within a factor of 2 or better of the nominal data for the sample. Several elements were also identified which were not known to be present or could not be measured by XRF indicating the utility of the proposed method. In a similar experiment a poly- propylene sample was examined for trace metal content. The results are presented in Table 4.Almost all of the laser abla- tion-ICP-MS results are again within a factor of 2 of the nominal values; the results for the two outliers Al and P are t 75 80 85 90 95 100 105 110 115 120 125 130 135140 145 150 L 155 160 165 170 175 180 185 190 195 200 205 210 215 220 miz Fig. 3 Laser ablation-ICP-MS spectral scan of PVC material within a factor of 3 of the nominal value. The extent to which the results differ is likely to be a function of the suitability of the internal standard selected. Carbon- I3 was used as the internal standard primarily because carbon is a major element present in all polymer-based samples. It may be possible to improve the accuracy by selecting a metal in the mid-mass range as the inter- nal standard but this is not an approach that is generally appli- cable since it infers prior knowledge about the sample.The semi-quantitative results reported here are acceptable in terms of accuracy in these 'survey' applications given the difficulty of obtaining solid polymer standards for a range of elements.148 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 Table 3 Semi-quantitative analysis of polyester by laser ablation-ICP- MS (LA-ICP-MS) Element LA-ICP-MS Nominal result/pg g-I value/pg g-' Li Al P Mn c o Zn Ge Mo Sn Sb Pb * NA = not available. 0.2 590 90 76 25 30 2 2.3 I8 160 3.2 NA* 350 105 73 37 50 4 NA NA 170 NA Table 4 ICP-MS Semi-quantitative analysis of polyproplylene by laser ablation- Element LA-ICP-MS Nominal result& g-' value/pg g-' Li P Al Ti Mn Fe c o cu Zn Zr Cd Sn Ba Pb * NA = not available.0.2 18 240 I20 30 30 12 20 300 41 33 340 2.7 6.5 NA* 45 100 60 NA 20 NA 18 250 NA NA NA 430 NA The relative amounts of elements at a given position in a solid sample may be of equal importance to their quantification. Using the laser elemental measurements can be made with a degree of lateral resolution limited only by the laser spot size. The system can be operated in such a way as to drive the s- or y-stages in a straight line across the material making a series of sequential ablations which are temporally and hence in terms of distance physically resolved. An example of this procedure is shown in Fig. 4. A composite ma- terial was examined for sodium and iron levels in terms of their lateral distribution in the sample. Since the time resolved measurement process is rapid laterally resolved data can be acquired on a multi-element basis.The profile of sodium inten- sity [Fig. 4(a)] shows a dip in the centre of the lateral scan in- dicating a change in concentration. This was found to correlate with measurements suggesting a reduction in the loading of the polymer at this position in the composite. A rather differ- ent profile is found for iron [Fig. 4(h)] indicating particulate contamination rather than a definite trend in bulk concentra- tion. This type of information would not be easy to obtain by the conventional sample preparation-liquid analysis approach and indicates one way in which the laser ablation technique can be used to broaden the application of atomic spectrometric measurement. Since the laser removes material from the surface of the sample and gradually penetrates into the bulk by repeated abla- tion it might be expected that some information relating to element concentration with respect to sampling depth could be obtained.A nylon sample was examined by observing in turn t 2 [I I 5 12 19 26 33 40 47 54 61 68 Time/s Fig. 4 Lateral scan of composite material using laser ablation-ICP-MS with time-resolved acquisition for (0) data acquisition for sodium at mass 23 and (h) data acquisition for iron at mass 54 the signals for five separate laser shots at a single .v-y- coordinate position. The cumulative intensities for copper-63 and -65 isotopes and for carbon- 13 obtained with respect to the number of laser shots are presented in Fig. 5.The carbon- 13 re- sponse shows a linear increase with respect to the number of shots and is indicative of the general matrix ablation. It is evident that the intensity responses for the two copper isotopes are non-linear initially and follow a similar growth which even- tually approximates to a linear increase. This would appear to indicate that there was either surface contamination of the sample with copper or possibly a dependence of copper additive concentration on depth. Alternatively this feature could be an artefact of the ablation process resulting from the selective vol- atilization of copper. This has not been observed in the exami- nation of other samples however and would indicate that some type of depth related information can be obtained.Quantitative Analysis !Samples of polyester and polypropylene were analysed on a quantitative basis using standards of each material to yield a combined calibration function for several elements. Sufficient data were available to provide calibration for Al Si P Co Zn and Sb. The quantitative data obtained for each sample are given in Table 5 . The results presented in Table 5 show rea- sonably good agreement with the known values particularly in view of the fact that the calibration was based on more than one type of polymer. Cobalt was detected at a level of 7 ppm in the polypropylene standard no reference value was avail- able for this element. However no antimony could be detect- ed. Elements such as Si and P suffer from spectral !interferences in ICP-MS due to the presence of molecular species and this can be exacerbated by concomitants in solu- I.ion based measurements.It is encouraging to note that the ilaser sampling approach gives results that are in reasonable agreement with the values obtained by ICP-MS for these ele- ment s . It was recognized that the use of solid standards with identi- cal matrix matching would be the best method of assessing the performance of the laser ablation-ICP-MS technique. Since polymeric standards for trace element analysis are not avail- able from the usual agencies it is normal practice to use previ- ~ x ~ s l y analysed samples or nominal formulation data for initial calibration. In most instances the elemental composition in !:he material will be restricted according to its product function149 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL.6 50 I 0 1 2 3 4 5 6 Fig. 5 Cumulative ion intensity for copper and carbon for successive laser ablation shots at an unfilled nylon sample A "Cu; B. "CU; and C Laser shots I3c and for particular matrices single element calibration stan- dards are therefore more easily obtained. Samples of pigment- ed polyethylene were analysed for the phosphorus content by laser ablation-ICP-MS using XRF solid calibration standards for quantification purposes. The results are presented in Table 6. The results obtained using laser ablation-ICP-MS show good agreement with the reference method particularly given that the XRF calibration is based on nominal values. The same five samples contained Ni and data were acquired during the ablations of the polyethylene for this element also.The results of the analyses are given in Table 7. Again good agreement was obtained between the two sets of results indicating the validity of the laser ablation method. It should be noted that laser ablation offers higher sensitivity than XRF since Ni was detected in sample E (Table 7) at a level of 1 pg g-' while it could not be detected using XRF. This facet is demonstrated in the application of the technique to the determination of trace amounts of Mg in polypropylene. It is difficult to determine Mg by XRF at the low pg g-I level. Materials have previously been characterized by the use of neutron activation analysis (NAA) or by wet chemistry-atomic spectrometric procedures.Data obtained for the analysis of five such samples by laser ab- lation-ICP-MS using solid calibration standards for Mg are given in Table 8. Very good agreement is achieved between the laser ablation results and the values determined by NAA at the low ppm level. Conclusions The results presented here suggest that laser ablation-ICP-MS is a viable technique for the determination of trace elements in plastics. In addition to the analytical advantages of achieving good sensitivity at the low pg g-' level and wide coverage on a multi-element basis the technique would appear to have po- tential for generating information relating to the spatial distri- bution of analytes in plastic materials. The level of precision which was obtained by laser ablation-ICP-MS in the present work was typically of the order of just under 10% relative standard deviation although better results have been observed in specific instances.This is presently perceived as a limitation of the technique and is currently the subject of further study.?' The authors thank ICI plc for permission to submit this paper for publication. Thanks are due to J. Dale and P. Waterton Table 5 Quantitative analysis of plastics by laser ablation-ICP-MS Polyester Polypropylene LA-ICP-MS result/ Element pg g-' Al 357 Si 720 P 93 c o 32 Zn 45 Sb 155 * NA = not available. + ND = not detectable. Nominal value/ PE b-l 350 770 105 37 50 I70 LA-ICP-MS result/ I05 600 60 7 205 NDi Pg P-' Nominal value/ PE !K' 100 750 45 NA* 200 ND Table 6 laser ablation-ICP-MS Quantitative determination of P in pigmented polyethylene by P content/pg g-' LA-ICP-MS XRF A 297 282 B 3 I4 295 c 139 101 D 227 227 E 36 I 370 Sample Table 7 laser ablation-ICP-MS Quantitative determination of Ni in pigmented polyethylene by Ni content/@ g-' LA-ICP-MS Sample A 204 B 162 C 210 D 80 E 1 * ND = not detectable.XRF 204 I74 208 84 ND* Table 8 lat ion-ICP-MS Quantitative determination of Mg in polypropylene by laser ab- Mg contentipg g-l LA-ICP-MS NAA Sample A 1 1 13 B 6.3 6.1 C 23 24 D 17 13 E 29 31 who produced the photographs of the laser ablation sites and to P. L. Warren for the provision of plastic calibration standard materials. References 1 Iiidurtivrly Coupled Plusniu Eniissioii S p i ~ ~ r i .o . ~ i o i ~ ~ Purr 11. Applicw- (ions and Fundanieiituls. ed. Boumans. P. H. J. M. Wiley. New York I987.p. 1. 2 Houk R. S.. Aiiul. Chmi.. 19x7. 58. 97A. 3 Applicarioiis of Irrdirc~ti~~i~l~ Coiipld Plusniu Muss Spcc~ri.onictr:\.. eds. Date. A. R.. and Gray. A. L.. Blackie. Glasgow. 1989. p. 169. 4 Runge. R. F.. Minck. R. W.. and Bryan. F. R. Spectrochim Acta. 1964.20.733.150 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 5 6 7 8 9 10 I I 12 13 14 15 Millard J. Dalling. R. H. and Radziemski L. J. Appl. Spec7rosc.. 1986,40,49 1. Dimitrov G. and Zheleva Ts. Specrrochim. Acta Part B . 1984 39 1209. Van Deijke W.. Balke J. and Maessen.. F. J. M. J. Spectim.him. A m Purr B 1979.34. 359. Mitchell P. G. Sneddon J. and Radziemski L. J. Appl. Spectroosc. 1986.40 274. Iida Y.. Spectrochint. A m Part B. 1990. 45 427. Ishizuka T. and Uwamino Y.. Spectrochim. A m . Purr B 1983 38 519. Thompson M. Goulter J. E. and Sieper F. Analysr 1981 106 32. Su G. and Lin S . . J . Anal. Af. Specfmm. 1988,3 841. Carr. J . W. and Horlick G.. Spcc.ti.oc.him. A m Part B 1982 37 1. Thompson M. Chenery S. and Brett L. J. Anal. At. Spec'rrom. 1989.4 1 1 . Kagawa K.. Matsuda Y. Yokio S. and Nakajima S. J. Anal. Ar. Spewom.. 1988 3 4 15. 16 1 '7 1 :3 1 '9 20 21 22 23 24 Darke S. A. Long S. E. Pickford C. J. and Tyson J. F.. J . Anal. At. Spewom. 1989,4,7 15. Gagne J. M. Pianarosa P. Lafleur F. and Perrault G. J. Anal. At. Spectrom. 1988,3,683. Gray A. L. Analyst 1985 110 551. Arrowsmith P. Anal. Chem. 1987,59 1437. Arrowsmith P. Appl. Spemosc. 1988 42 1988. Marshall J. Carroll J. and Sparkes S. T.. J. Anal. At. Spectrom. 1989.4,25 1 R. Marshall J. and Franks J. Anal. Pro(.. 1990,27 240. Warren P. L. Anal Pi.oc. 1990,27 186. Franks J. Marshall J. and Abell I. unpublished work. Paper- Oi03900D Received August 28th 1990 Accepted Decemher- loth 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600145

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 151 Determination of Arsenic in Samples With High Chloride Content by Inductively Coupled Plasma Mass Spectrometry* Simon Branch Les Ebdon Mick Ford Mike Foulkes and Peter O’Neilf Plymouth Analytical Chemistry Research Unit Department of Environmental Sciences Polytechnic South West Drake Circus Plymouth Devon PL4 8AA UK The arsenic content of samples with a high chloride content has been determined using inductively coupled plasma mass spectrometry (ICP-MS) with the addition of nitrogen to the carrier gas. With as little as 0.032 I min-1 nitrogen in the argon carrier gas the level of the polyatomic ion 75ArCI+ which interferes in the determination of monoisotopic arsenic is reduced to negligible levels. This modification is shown to be effective for the analysis of real samples even for solutions which contain greater than 1% chloride.The presence of peaks at masses 49 and 51 when nitrogen is added is attributed to 14N3%I+ and 14N37CI+ respectively. The arsenic content of reference materials determined by nitrogen addition ICP-MS are in agreement with certificate values. Conventional argon-only ICP-MS values show an apparent 30% increase in arsenic owing to chloride interference. The nitrogen addition study has been extended to include a comparison of the total arsenic content of urine samples obtained from diet-controlled volunteers with the cumulative contribution from arsenic species determined by alternative coupled techniques. Keywords Inductively coupled plasma mass spectrometry; arsenic determination; nitrogen addition; chloride interference; urine and seaweed Since the development and commercial introduction’.’ of in- ductively coupled plasma mass spectrometry (ICP-MS) one of the limiting features has been the occurrence of polyatomic ion interference^^.^ usually in the region of the spectrum below 80 u.The polyatomic ion 40Ar35C1+ is a particular problem because it precludes the determination of monoiso- topic arsenic 7sAs in samples with a high chloride content including urine samples and marine prod~cts.3-5-~ A number of approaches to overcome this interference have been pro- posed. These include mathematical correction based on the 40Ar37C1+ signal,’ the coprecipitation of the chloride with silver,x chromatographic separation of the arsenic and chlo- ride9.I0 and hydride generation ICP-MSI I in which the arsenic and chloride are separated using a membrane gas- liquid separator. However none of these methods are partic- ularly convenient therefore arsenic in biological samples is still usually determined by hydride generation or electrother- mal atomic absorption spectrometry (ETAAS).These methods also have disadvantages. For example marine samples can contain arsenobetaine which has been identified in a wide range of marine products1’ and is both non- reducible13 and as with other organo-arsenics resistant to normal digestion methods.I4 Biological samples that have high chloride levels are also likely to contain phosphate a compound known to interfere in the determination of arsenic by ETAAS.Is Recently Evans and EbdonI6 described two methods to over- come polyatomic ion interferences.In the first method small volumes of an organic solvent such as propan-2-01 were added to samples and standards. The second approach involved bleeding small flows of a molecular gas either oxygen or ni- trogen into the argon nebulizer gas. They found that these methods reduced polyatomic interferences on 75As and 77Se to a substantial extent. In the present paper the successful determination of arsenic in National Institute of Environmental Standards (NIES) (Ibaraki Japan) Certified Reference Material (CRM) No. 9 Sargasso Seaweed National Institute of Standards and Technology (NIST) (Gaithersburg MD USA) Standard Reference Materials (SRMs) 8413 Mixed Diet and 1573 Tomato Leaves (National Research Council Canada * Presented at the Fifth Biennial National Atomic Spectroscopy Sympo- sium (BNASS) Loughborough UK.I8th-20th July. 1990. (NRCC) Ottawa Ontario Canada DORM 1 Dogfish Muscle Nycomed AS Oslo Norway. Seronorm reference urine and urine from volunteers who had eaten fish known to contain non-toxic arsenobetaine prior to urine collection is reported. Nitrogen addition was used during the determina- tion of arsenic as it is the simplest method and avoids the problems associated with the introduction of organic solvents to the ICP-MS.17 To confirm further the integrity of the results the total arsenic determined by nitrogen addition ICP- MS was compared with the sum of the individual arsenic species found by high-performance liquid chromatography (HPLC) ICP-MS9 and hydride generation cryogenic trapping atomic absorption spectrometry (HG-CT-AAS).IX Experimental Instrumentation The ICP-MS results were obtained using a VG Elemental Plas- maQuad 2 (VG Elemental Winsford Cheshire UK) which is designed to admit low flows of a second gas into the nebulizer gas.The instrument was fitted with an Ebdon V-groove high solids nebulizer (PSA Sevenoaks Kent UK). Operating con- ditions are given in Table 1. The HPLC-ICP-MS measure- ments were obtained using the conditions and instrumentation previously described.’ The HG-CT-AAS analyses were performed using a continu- ous flow hydride generator (PSA) and a glass U-tube packed with silanized 40-mesh glass beads which was immersed in liquid nitrogen as a gas trap.The gaseous hydrides were dried prior to trapping by passage through a water trap and two sodium hydroxide traps all of which were maintained at 0 “C. Detection was by atomic absorption spectrometry using an SP9 spectrometer (Philips Analytical Cambridge UK). This Table 1 Inductively coupled plasma mass spectrometer operating conditions. Indium ( I 0 0 pg I-I) was used as the internal standard Parameter Setting Forward power/” 1650 Reflected power/W 20 Ar nebulizer gas flow-rate/l min-I N2 nebulizer gas Row-rate/l min-I Auxiliary gas flow-raten min-1 0.82 0.03 I .0 Coolant gas flow-rate/l min-’ 14152 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 I Parameter Setting ; 5 O O o - Delay ti me/s 5 .O $3000 C 3 Hydride generution- 0.5 s 4 0 o o - Sodium tetrahydroborate concentration (% m/v) Hydrochloric acid concentration/mol dm-3 Rise time/s 5 .O Analysis time/s I80 1 .O - 0 3 2 2000 Memory time/s 10 lo00 Atomic absorption spectrometer- W avelengthhm 193.7 - 80 .r i U -60 8 -40 2 - C 8 . !I20 s t - CI .- II - .- - C I I I arrangement has been described previously.lg Instrumental conditions are detailed in Table 2.Materials and Chemicals Two reference materials were first analysed Seronorm refe- rence urine and the IES CRM No. 9. In addition urine samples were collected from research students who had eaten a fish- based meal which was known to contain 6.5 mg kg-' of arsenic principally as arsenobetaine. Arsenic recovery from solutions of Aristar sodium chloride (BDH Poole Dorset UK) was also determined.All standards were freshly prepared from 1000 mg I-' As stock solutions of sodium arsenite (AnalaR grade BDH) dimethylarsinic acid (DMAA) (Sigma Poole Dorset UK) monomethylarsonic acid (MMAA) (donated as disodium methanearsonate by Dr. A. Howard Southampton University Southampton Hampshire UK) and arsenobetaine (AB) (donated by Professor K. I. Irgolic of Texas A and M University TX USA). All standards and samples were spiked with indium (Aldrich Milwaukee WI USA) to give a final concentration of 100 pg I-' and made up to volume with 2% Aristar nitric acid (BDH). Sample Preparation Urine samples were collected from volunteers directly into measuring cylinders. After recording the volume sub-samples were decanted into dry acid-washed polypropylene bottles.The samples were then filtered 1 ml was removed spiked with indium and made up to 10 ml with 2% nitric acid. This solution was then analysed by nitrogen addition ICP-MS. Samples of the filtrate were also analysed by HPLC-ICP-MS and HG-CT-AAS. The Seronorm urine was similarly prepared by 1 + 9 dilution with 2% nitric acid and spiking with indium to a final concentration of 100 pg I-'. The seaweed samples were prepared by microwave diges- tion. Approximately 0.2 g of the NIES CRM No. 9 was accu- rately weighed into a microwave digestion bomb (Savillex Minnetonka MN USA). Nitric acid (Aristar 3 ml) was added to the bomb which was partly sealed and left overnight to ini- tiate digestion of the contents. The sealed vessel was placed in a domestic microwave oven [Toshiba ER-76 1 E Toshiba (UK) Camberley Surrey UK] and heated using a two-stage process.Firstly a low power setting (2-3 maximum setting = lo) was used for 5 min the bomb was then cooled for 15 min in a refri- gerator at 4 "C and the pressure was slowly released from the vessel. Secondly after re-sealing the microwave process was repeated using a medium to high power setting (5-7) for 3 min. The vessel was again cooled in a refrigerator for 15 min and the pressure slowly released. The digested contents were quantitatively transferred into a 100 ml calibrated flask and made up to the mark with 2% nitric acid. A dilution of 1 + 9 using 2% nitric acid was performed to bring the arsenic con- Fig. 1 Effect of increasing chloride concentration on arsenic and indium signals in ICP-MS with the addition of nitrogen to the carrier gas.Actual concentrations arsenic 100 pg 1-I; and indium 100 pg I-'. A Arsenic concentration determined in pg 1 - I ; B indium signal in counts s-I; and C arsenic signal in counts s-I centration below 50 pg 1-I. This final solution also contained 100 pg 1-' of indium as internal standard. Total arsenic analy- ses were performed by nitrogen addition ICP-MS. Results and Discussion Synthetic Sodium Choride Solutions In order to determine the validity of the experimental protocol investigations began by using nitrogen addition ICP-MS to determine the apparent concentration of 100 pg 1-I arsenic spikes in solutions containing 0 10 100 loo0 and 10 OOO mg 1-1 of sodium chloride. The sodium chloride solutions were prepared so that they could be diluted with 2% nitric acid in the same manner as the urine samples.The results are shown in Fig. 1. The arsenic concentration was determined using the ICP-MS quantitative analysis software whilst simultaneously the raw counts for arsenic and indium were recorded to assess whether there was any enhancement suppression or transport effects. At sodium chloride concentrations of 0 10 I 0 0 and 1000 mg I-' arsenic quantification was within 5% of the theo- retical value (Fig. I line A). Only at a concentration of 10 000 mg I-' NaCl was a significant interference observed. This was a result of an enhanced indium signal. These results indicate effective removal by the nitrogen addition of 40Ar35C1. This was confirmed by the counts at 77 u which arise from 40Ar37C1 not exceeding 20 counts s-l effectively a blank level for any of the analysed solutions.These results indicate that urine samples containing up to 10000 mg kg-I sodium chloride can be analysed for As with confidence provided they are diluted (10-fold). Even a solution containing 100 000 mg kg-I sodium chloride or lo% should yield results within 15% of the correct value. However this last option is not feasible by direct nebulization as the ICP-MS instrument will not tolerate such a high percentage of dissolved solids on a continuous basis. Flow injection methods can however be used in such cases of ultra-high salt concentration.20 Reference Materials The NIES CRM No. 9 and urine samples were analysed using normal operating procedures. The lenses nebilizer gas flow and torch position were adjusted to give maximum response for "As.This was usually found to lie in the region 200 000-300 000 counts s-' per ppm As. The results are given in Table 3. This table also compares the values obtained for total As in NIES CRM No. 9 with and without nitrogen addi- tion to the carrier gas. It is apparent that the 5% indicated chlo-153 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY. MARCH 1991. VOL. 6 Table 3 Determination of arsenic in reference materials with and without nit rogen addition Certified Concentration Concentration concentration of As with of As without Sample of As nitrogen addition; nitrogen addition NIES CRM NO. 9 * - 1 115*9 106f11 1 1 1 f 12 2 103 f 8 3 121 f 1 1 4 - - - Mean 115k9 I 10 f 3.3$ Seronor-m iuine$- 1 20011 201 f 6 2 - 214f 17 3 - 219k 16 195k 10 4 Mean 200 207 k 4.34 - * Arsenic content in mg kg-I.t Range is one standard deviation (1?=3). $ Standard error of the mean value. 0 Arsenic content in pg I-'. fl Range is 176-208. It ND = Not determined. 484 10 SO+ 19 46 f 8 48 * 9 48 f 3.14 ND II ND ND ND - Table 4 addition ICP-MS. Arsenic content is given in mg kg-' Determination of arsenic in reference materials using nitrogen Reference material Certified value* Nitrogen addition ICP-MS value* NIST 843 1 Mixed Diet NIST 1573 Tomato Leaves NRCC DORM 1 Dogfish Musclet 0.924 f 0.344 0.27 f 0.05 17.7 k 2. I 0.947 k 0.1 1 0.24 -t 0.0 1 17.8 k 0.7 * Range is one standard deviation. 1?=3. t Chloride content given as I .134. Table 5 All results are given in pg of arsenic Total arsenic in urine compared with arsenic species in urine. Sample source Total*? AB?$ DMAAS MMAAS As,$ Volunteer. A- Sample 1 58.5 k3.9 56.8 f 1.0 <0.4 <0.8 1.3 Sample 2 34.8 45.5 35.5 f 2.4 < 0.5 <0.9 1.2 Volunteer- L L Sample1 17.8f3.5 18.9f2.6 0.4 <0.1 ~ 0 . 1 Sample 2 84.8f4.0 88.5 f5.0 ~ 0 . 4 ~ 0 . 8 ~ 0 . 8 * Determined by nitrogen addition ICP-MS. t Range is one standard deviation ~ = 3 . $ Determined by HPLC-ICP-MS. 5 Determined by HG-CT-AAS. ride content of these samples produced a high bias (30% in- crease above the certified arsenic value) when nitrogen gas was absent. When the nitrogen addition was used good agree- ment between the certified and experimental values was ob- tained.The chloride level in the urine is not certified but reported 24 h values lie in the range 4.4-6.5 g." Again using nitrogen addition excellent agreement between the certified and experimental results was observed. Recently Mulligan et al.' described a similar sample preparation procedure for the determination of trace elements in urine. The determination of arsenic was compromised by 4'JAr35Cl which produced a posi- tive interference. The results of the present study clearly indi- cate that if nitrogen addition is used these problems can be overcome. Table 4 illustrates the results obtained for the determination of arsenic in NIST SRMS 8431 Mixed Diet and 1573 Tomato t - - Q 52 56 60 64 68 72 76 m/z Fig. 2 nitrogen ICP-MS spectra of urine ( u ) without and ( h ) with the addition of Leaves and DORM 1 Dogfish Muscle using nitrogen addition ICP-MS.Sample preparation for these materials was as de- scribed for NIES CRM No. 9 (see under Sample Preparation). The nitrogen addition ICP-MS results are again in good agree- ment with the certified values for these samples which are known to contain high levels of chloride. Results for Urine from Human Volunteers Total arsenic in urine from volunteers who had eaten fish meals was known to be composed of contributions from AB DMAA MMAA and inorganic arsenic AS,).?^ Methods have been developed in this laboratory to determine these arsenic species.y.'x These methods were used to compare arsenic species in urine collected from volunteers with the value for total arsenic determined by nitrogen addition ICP- MS.The results are shown in Table 5. All the results have been corrected for the volume of the urine sample taken i.e. the arsenic output is given as pg of As in the sample. The good agreement between these independent techniques confirms the promise of nitrogen addition ICP-MS in allow- ing the routine determination of arsenic in urine. The full results of the speciation study of urine will form part of a 1 at er pub1 icat ion. The mechanism that removes the 40Ar35CI+ ion is not fully understood. Fig. 2 shows the extent to which the nitrogen addi- tion modifies the mass spectrum in the range 48-80 u for the same urine sample. The elimination of 40Ar35C1+ is confirmed by the absence of a peak at 77 u in the lower trace and leaves the peak at mass 75 due to arsenic alone at a concentration of =20 pg I-'.The increase in the peak seen at mass 49 can be at- tributed to the presence of 14NT1+ which suggests that the peak at mass 51 might in part be due to 14N37C1+. Certainly the nitrogen added decreases the C10+ and ArO+ peaks (the former interfering with vanadium at mass 51 the latter with iron at mass 56) with the subsequent large increase at mass 54 attri- buted to ArN+. Conclusions The addition of small flows of nitrogen to the argon carrier gas has allowed arsenic to be determined successfully by ICP-MS in samples with a high chloride concentration. Nitrogen addi- tion removes the polyatomic ion ArCl+ which interferes with monoisotopic arsenic (75 u). Experiments showed this154 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL.6 modification to be effective for solutions containing in excess of 1% chloride. When conventional argon-only carrier gas is used a value for arsenic 30% higher than that certified is found in NIES CRM No. 9. Close agreement with the certified values for arsenic in NIES CRM No. 9 and Seronorm refe- rence urine is obtained when as little as 0.03 1 min-' of nitro- gen is present in the carrier gas. Nitrogen addition ICP-MS has proved successful for the determination of arsenic in urine samples. The total arsenic content of urine samples obtained from volunteers fed on a controlled diet agree with the cumulative contribution from the arsenic species AB DMAA MMAA and Asi as deter- mined by HPLC-ICP-MS and hydride generation cyogenic trapping AAS.This is further confirmation of the promise that nitrogen addition ICP-MS has to offer in overcoming proble- matic arsenic determinations in biological samples of high chloride content. References Gray A. L. in Applications of Inductively Coupled Plasma Mass Spectrometry eds. Date A. R. and Gray A. L. Blackie and Son Glasgow 1989 p. 1. Gray A. L. and Date A. R. Analyst 1983,108 1033. Munro S. Ebdon L. and McWeeney D. J. J. Anal. At. Specworn. 1986 1,211. Tan S. H. and Horlick G. Appl. Spectrosc. 1986,4,445 Rideout P. S. Jones H. R.. and Williams J. G. Analyst 1988 113 1383. Heitkemper D. Creed J. Caruso J. and Fricke F. L. J. Anal. At. 7 8 9 1 0 I I 12 13 I4 I5 16 17 18 19 20 21 Specwont.. 1989.4 219. Mulligan K. J.. Davidson. T. M.. and Caruso J.A.. J. Artul. At. Spec- tront.. 1990,5 30 1. Lyon T. D. B.. Fell. G. S.. Hutton. R. C.. and Eaton A. N.. J. A m / . Ar. Spec-tr(int. 1988,3,601 Branch. S.. Bancroft. K. C. C. Ebdon. L.. and O'Neill. P.. Awl. Proc.. 1989. 26 73. Lyon. T. D. B. Fell. G. S. Hutton R. C. and Eaton A. N.. J. Artal. At. Specworn.. 1988.3 265. Branch S.. Corns. W. T.. Ebdon L. Hill S.. and O'Neill. P..J. Attul. At. Specwom.. 199 I. in the press. Lawrence J. F.. Michalik P.. Tam G. and Conacher. H. B. S.. J. Agric.. Food Chem. I986,34,3 15. Howard. A. G.. and Comber S. D. W.. Appl. Orgartomet. Chem.. 1989,3,509. Maher W. A.. Microchent. J. 1989.40 132. Chakraborti D.. Irgolic K. J. and Adams F.. Ittt. J . Ettviron. Aital. Chem. 1984,17,24 1. Evans E. H. and Ebdon. L. J. Anal. At. Specworn. 1989,4,299. Hutton R. C. J. Anal. At. Specworn.. 1986 1,259. Howard. A. G.. and Arbab-Zavar M. H. Anal~st I98 I 106,2 13. Ebdon. L.. Hill S. J.. Walton. A. P. and Ward R. W. Anu/yst 1988 113 1159. Dean J. R. Ebdon L. Crews H. M.. and Massey R. C. J . Anal. At. Specworn. 1988,3,349. Iyengar G. V. Kollmer W. E.. and Bowen H. J. M. The Elemental Composition of Human Tissues and Body Fluids. Verlag Chemie Weinheim 1978 p. 127. Paper 010397214 Received September 3rd I990 Accepted October 25th. I990

ISSN:0267-9477    

DOI:10.1039/JA9910600151

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 155 Determination of Arsenic by Hydride Generation Inductively Coupled Plasma - Mass Spectrometry Using a Tubular Membrane Gas-Liquid Separator* Simon Branch Warren T. Corns Les Ebdon Steve Hill and Peter O’Neill Plymouth Anavtical Chemistry Research Unit Department of Environmental Sciences Polytechnic South West Drake Circus Plymouth Devon PL4 8AA UK A silicone rubber membrane gas-liquid separator (MGLS) has been used to determine arsenic by hydride genera- tion inductively coupled plasma mass spectrometry (ICP-MS). The use of a membrane separator eliminated the polyatomic ion interference of ArCI+ on As+ which is a significant interference in pneumatic nebulization ICP-MS and a residual problem with hydride generation using a conventional U-shaped gas-liquid separator.Optimum op- erating conditions for the MGLS were identified as 0.1% m/v sodium tetrahydroborate 4 mol dm-3 hydrochloric acid 0.9 I min-’ argon purge gas flow 0.8 I min-l auxiliary gas flow and 1.8 kW forward power. Use of the MGLS enables the determination of arsenic in solutions of hydrochloric acid by ICP-MS. The arsenic content (25 ng ml-l) of a reference water sample was successfully determined. Keywords Hydride generation; inductively coupled plasma mass spectrometry; membrane gas-liquid separator; arsenic determination Since the advent of inductively coupled plasma mass spectro- metry (ICP-MS) in the early 1980s it has become widely rec- ognized that polyatomic ions represent one of the major sources of Such interferences are particularly troublesome for monoisotopic elements one prime example of which is arsenic.Arsenic has a mass:charge ratio of 75 which coincides with that of 40Ar35C1+. Obviously a chloride based in- terference severely limits the determination of arsenic in bio- logical and environmental samples and in these instances methods such as mathematical c~rrection,’.~ chromatographic separation of arsenic from chloride,s or the addition of an organic solvent (e.g. propanol) or a molecular gas (e.,g. nitro- gen or oxygen) have to be employed.6 The generation of gaseous covalent hydrides has long been known as an effective way of improving sensitivity in atomic spe~trometry.~ More recently hydride generation (HG)ICP- MSsv9 has attracted attention as a way of overcoming the poor nebulization and transport efficiency (typically of the order of 1-2%) of ICP-MS and thus has the potential of improving sen- sitivity by two orders of magnitude.Powell et al.R have de- scribed the optimum operating conditions using this technique for the determination of As Bi Hg Sb Se and Te in environ- mental samples. In addition Dean et aL9 have studied the effi- ciency of hydride generation for a number of organic and inorganic selenium species and Janghorbani and TingIo have compared the performance of continuous hydride generation for arsenic with pneumatic nebulization. The determination of arsenic by HG-ICP-MS is limited by the transport of chloride as sodium chloride in the fine spray formed by the bubbling of the purge gas through the conventional U-shaped gas-liquid separator.This problem can be overcome by using nitric rather than hydrochloric acid but ideally a more effective gas-liquid separator is required. Recently tubular gas-liquid separators have been described for flow injection (FI)ll and FI-HG atomic absorption spectrometry (AAS).’* Both of these last mentioned papers described gas-liquid separators based on the diffusion of gas through a tubular microporous membrane constructed of polytetrafluoroethylene. A similar design has also been used by Wang et al.I3 to determine lead in a variety of reference ma- terials and galena samples by isotope dilution HG-ICP-MS. Cave and Greenl4*l5 have detailed the use of a silicone rubber tubular gas-liquid separator to determine As HCOc I- Se * Presented in part at the Fifth Biennial National Atomic Spectroscopy Symposium (BNASS) Loughborough UK 18th-20th July 1990.Sni4 and reduced sulphur in groundwater by ICP optical emis- sion spectrometry.Is In this paper the application of a membrane gas-liquid sep- arator (MGLS) based on the design of Cave and Green,l4.Is to the determination of arsenic by ICP-MS is described its effi- ciency in excluding chloride from the plasma assessed and an- alytical application explored. Experimental Instrumentation A continuous flow hydride generator (PSA 10,003 PS Analyt- ical Sevenoaks Kent UK) was used fitted with both the sup- plied U-shaped gas-liquid separator and the MGLS described by Cave and Green.I4.Is A short transfer line of silicone rubber tubing was fitted directly to the base of a Fassel plasma torch in the ICP-MS (PlasmaQuad 2 VG Elemental Winsford Che- shire UK).Reagent flow-rates were set at 7.5 ml min-I for hy- drochloric acid and sample solutions and 3.0 ml min-l for sodium tetrahydroborate solution. The hydride generator delay time was set at 5 s rise time at 25 s analysis time at 30 s and memory time at 1 min. The MGLS is shown in Fig. 1. The sodium tetrahydroborate and hydrochloric acid concen- trations carrier gas flow-rate ICP-MS forward power and aux- iliary gas flow-rate were all optimized in order to obtain a maximum arsenic signal using a 100 ng ml-’ solution. Prior to each variation in one of these parameters the ion lenses were tuned for maximum response. The conditions used are sum- marized in Table I.Data were collected using the ICP mass spectrometer time resolved software. This enabled a full standard range to be ac- NaBH Argon in To ICP-MS torch 1‘ Acidified sampleblank Fig. 1 Membrane gas-liquid separator156 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 199 1 VOL. 6 Table 1 Optimal operating conditions Hydride generation- Sodium tetrahydroborate solution concentration Sodium tetrahydroborate solution flow-rate Hydrochloric acid concentration Hydrochloric acid/sample flow-rate Carrier gas flow-rate 0. I % m/v 3.5 ml min-’ 4 mol dm- 7.5 ml min- 0.9 1 min-‘ IC P - MS- Forward power Reflected power Coolant gas flow Auxiliary gas flow Data collection mode 1.8 kW ?Y 14 1 min 0.8 I rnin-‘ 72-79 u time resolved acquistion quired in one run whilst the chloride interference was moni- tored simultaneously by collecting data at 77 u which corre- sponds to 40Ar37C1.Peak heights were recorded by manual measurement of the hard copy output produced for each run. Reagents Unless specified otherwise all reagents were of AnalaR or Aristar grade (BDH Poole Dorset UK). De-ionized water was used throughout (Milli-R04 Millipore MA USA). Sodium tetrahydroborate solutions were prepared by dissolv- ing sodium tetrahydroborate (SpectrosoL BDH) in 0.1 mol dm-3 sodium hydroxide solution (stabilizer). Fresh solu- tions were prepared daily. Standard solutions were prepared by appropriate dilution of stock 1000 mg I-’ arsenic(l1i) chloride solution (SpectrosoL BDH) using hydrochloric acid solutions. Fresh solutions were prepared daily.Potassium iodide ( 1 % m/v in 4 mol dm-2 hydrochloric acid) was used to reduce arsenate to arsenite. In order to determine the accuracy of the method arsenic was determined in a water certified reference material (Interna- tional Atomic Energy Authority Simulated Fresh Water IAEA/ W-4). All glassware was soaked in 10% nitric acid for 24 h prior to use and then rinsed five times with distilled water. Results and Discussions Optimization Five parameters critically influence the magnitude of the HG- ICP-MS signal. These are the sodium tetrahydroborate and acid concentration forward power carrier and auxiliary gas flows. The order in which the various experimental conditions were optimized was not important however prior to each vari- ation the ion lenses were tuned for maximum response.The significance of each of these variables was investigated by uni- variate searches the results of which are shown in Figs. 2-6. The variation in net signal given as peak height with the concentration of sodium tetrahydroborate (Fig. 2) shows a re- markably sharp maximum unlike those previously demon- strated for either HG-ICP-MSX or HG-AAS.lh Cave and GreenJd have reported a similarly shaped profile although the maximum was not as pronounced giving an optimum sodium tetrahydroborate concentration of 0.4% m/v. The optimum sodium tetrahydroborate concentration found in this study was 0.1% m/v. This value demonstrates one particular advantage of the MGLS lowering the concentration of sodium tetrahy- droborate below the levels normally used (l-3%) reduces transition metal interferences.” The decay observed in the signal at higher concentrations in this work is the product of a number of factors.The excess of hydrogen produced dilutes the sample and may also affect the flow of the gases through the membrane. ‘I Furthermore ongoing research in our labora- 0 0.2 0.4 0.6 0.8 1.0 1.2 Sodium tetrahydroborate concentration (% m/v) Fig. 2 Variation in signal with sodium tetrahydroborate concentration l o t I I I I I I I I 1 0 1 2 3 4 5 6 7 8 Hydrochloric acid concentration/mol dm-3 Fig. 3 Variation in signal with hydrochloric acid concentration tories has shown that the introduction of hydrogen gas into the ICP-MS instrument generally causes a reduction in sensi- tivity. The hydrochloric acid concentration optimization plot (Fig.3) while similar in shape to that previously reported by Cave and Green,I4 identified a much higher optimum concentration of between 6 and 7 mol dm-’ in contrast to the 0.56 rnol dm-’ in the earlier paper. This higher level is fortuitous in over- coming transition metal interferences” but unfortunately results in damage to the silicone rubber membrane. There- fore a compromise concentration of 4 rnol dm-’ HCl was se- lec ted. As expected the net signal increased with increasing forward power Fig. 4. A similar result was found by Powell et ~ 1 1 . ~ although they reported a plateau above 1200 W a feature not seen in the present study. These workers also reported a higher background due to photon noise at high power settings but we found plots of signal to background ratio had an identi- cal appearance to those shown in Fig.4. This variance may merely reflect the different instrumentation used. Fig. 5 shows the variation of signal with carrier gas flow. The carrier flow acts in the manner of the injector gas flow and the signal rises rapidly as the plasma is punched reaches a maximum and then degrades again as the argon dilutes the sample. Similarly the auxiliary gas flow produces a sharp maximum followed by a rapid decay as the argon again dilutes the sample. Furthermore carrier and auxiliary gas flows affect the position of the plasma relative to the sampler cone. The degree of ionization varies throughout the analytical region and thus these variations in the sampling position will result in changes in sensitivity.The optimum operating conditions are summarized in Table 1 .157 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 100 0' I 1 I I I I 1 I 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Forward power/kW Fig. 4 Variation in signal with forward power 0.5 70 E E 60 s r UJ Q r -Y .- 50 L AO 0.6 0.7 0.8 0.9 1.0 1.1 Carrier gas fIow-rate/t min-' Fig. 5 Variation in signal with carrier gas flow-rate 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Auxiliary gas flow-rate/l min-' Fig. 6 Variation in signal with auxiliary gas flow-rate Chloride Exclusion Chloride interference can be monitored at 77 u which corre- sponds to 40Ar'7C1. Table 2 shows the magnitude of the inter- ference using various methods of sample introduction. Using pneumatic nebulization a smaller arsenic signal (75 u ) was seen relative to that observed using other methods.The signal at 77 u arises from instrumental noise and the arsenic(ii1) chloride used as a standard. When the standard was made up in 4 rnol dm-'' HCI a massive argon chloride contribution at 75 and 77 u was seen. The relative abundance of the '5CI:'7C1 iso- topes is 3 I . Given the signal at 77 u approximately SO 000 counts at 75 u were derived from argon chloride. As expected when using hydride generation the arsenic signal increased and the argon chloride signal decreased. Table 2 Relative signal and chloride interference using different sample introduction methods Sample introduction method Counts from 500 ng ml-' standard 77 u 72 75 u Pneumatic nebulization 13095 (As in 2% HNO3) Pneumatic nebulization 64 725 (As in 4 rnol dm-j HCI) Conventional gas-liquid separator (As in 3 rnol dm-' HCI) 84 835 Membrane gas-liquid separator (As in 4 rnol dm-j HCI) 235 127 7 529 I785 56 However with the conventional gas-liquid separator there is still a significant contribution from argon chloride to the arsenic signal owing to sodium chloride passing into the plasma from the aerosol formed in the gas-liquid separator.Some of this spray condenses in the transfer line and this may lead to large droplets passing to the plasma producing plasma instability and noisy signals. The MGLS was highly effective at gas-liquid separation and no moisture was observed in the transfer line. The effective exclusion of chloride from the plasma can be seen from the results obtained.The potential improvement in sensitivity using the MGLS is also apparent from the results although it was not possible to realize this po- tential fully here owing to the high background levels resulting from the contamination encountered in the reagents used. Thus the current practical limit of detection (30 on the blank) is quoted as 0.5 ng ml-' although the problem of reagent purity is currently being addressed. Determination of Arsenic in Water Reference Material In order to evaluate the use of the membrane on a realistic sample arsenic was determined in a water certified reference material. Since arsenite and arsenate have different hydride generation kinetics potassium iodide (1% m/v in 4 rnol dm-3 HCI) was used to reduce arsenate to arsenite.This solution was used for all sample and standard preparations. The reference material IAEA/W-4 was diluted with the blank 4 mol dm-' hydrocholoric acid solution and analysed under the conditions described in Table 1. The results obtained (23.4 k 0.6 ng ml-' for lo and n=6 compared with the certified value of 25 k 1.25 ng ml-I) clearly demonstrate the accuracy and precision of HG-ICP-MS using the MGLS. Furthermore the results again show the effectiveness of the MGLS in excluding chloride from the ICP-MS instrument the presence of which would have lead to a large positive bias in the analysis. IAEAIW -4 Conclusions The use of an MGLS with continuous flow hydride generation has been shown to be a highly effective method for overcom- ing argon chloride interference when determining arsenic by ICP-MS.By using the proposed configuration linear calibra- tion plots were obtained from 1 to over 100 ng ml-' of arsenic. A further advantage of the MGLS is the ability to dampen pump noise from the HG system which results in good repro- ducibility typically 2-3%. A detection limit of 0.5 ng ml-' was obtained and it is believed this could be improved dramatically with use of reagents of higher purity. The system described offers great potential for the determination of arsenic and sele-158 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 nium in samples with high chloride content such as urine and saline waters. The authors thank PS Analytical and the Science and Engi- neering Research Council for the award of an SERC CASE studentship to W.T.C.and the Ministry of Agriculture Fisher- ies and Food for the award of a grant to S.B. References Gray. A. L.. in Application of Inducti\vIy Coirpled Plasma Mass Specrromety eds. Date A. R. and Gray A. L.. Blackie Glasgow Munro S.. Ebdon L. and McWeeny D. J. J. Anal. At. Spectrom. 1986 1 21 1. Vaughan M. A. and Horlick G. Appl. Specrrosc.. 1986.40 434. Ridout P. S. Jones H. R. and Williams J. G. Ana/.vst 1988 113 1383. Lyon T. D. B. Fell G. S. Hutton R. C. and Eaton. A. N. J. Anal. At. Specworn. 1988,3 60 1. 1989 pp. 29-32. 6 7 a 9 10 I I 'I 2 13 14 15 16 17 Evans E. H.. and Ebdon. L. J. Anal. At. Spectrnm. 1989,4.299. Godden R. G. and Thomerson D. R. Analvst 1980,105 1 137. Powell M. J. Boomer D. W. and McVicars R. J. Anal. Chem. 1986,58,2684. Dean J. R. Parry H. G. M. Massey R. C. and Ebdon L. ICP lnf. Newisl. 1990 15 569. Janghorbani M. and Ting B. T. G. Anal. Chem. 1989,61,701. Montomizu S. Kyoji. T. Kuwaki T. and Oshima M. Anal. Chem. 1987,59,2930. Yamamoto M. Takala K. Kumamaru T. Yasuda M.. Yokayama S. and Yamamoto Y. Anal. Chem. 1987,59,2446. Wang X . Viczian. M. Lasztity A. and Barnes R. M. J. Anal. Ar. Specworn. 1988 3 82 1. Cave M. R. and Green K. A. At. Spemmc. 1988 9 149. Cave M .R. and Green K. A. J. Anal. At Spectrom. 1989,4,223. Ebdon L. Wilkinson J. R. and Jackson K. W. Anal. Chim. A m 1982,136 191. Welz B. and Schubert-Jacobs M. J. Anal. At. Specatinm.. 1986. 1 23. Paper 010084 7H Received July 2nd I990 Accepted November 23rd 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600155

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 159 Correction of Mineral Acid Interferences in Optical Emission Spectrometry on Copper Standardization Louise M. Garden* ICI Advanced Materials Engineering Compounds and Polymers 8JE UK Inductively Coupled Plasma and Manganese Using Internal Group Wilton Middlesbrough Cleveland TS6 John Marshall ICI plc Wilton Materials Research Centre Wilton Middlesbrough Cleveland TS6 8JE UK David Littlejohn Department of Pure and Applied Chemistry University of Strathclyde Cathedral Street Glasgow G 1 IXL UK A study has been made of mineral acid matrix interferences in trace element determinations by inductively coupled plasma optical emission spectrometry. It has been shown that signal suppressions of up to 40% can be observed under normal operating conditions.The interferences can largely be compensated for by using an internal stan- dardization procedure. Precision is also improved several fold when scandium is used as a real-time internal stan- dard. This is attributed to the reduction of noise by simultaneous measurement of emission signals. Keywords Inductively coupled plasma optical emission spectrometry; Myers-Tracy signal compensation; internal standardization; acid interference effect Although interferences found in inductively coupled plasma optical emission spectrometry (ICP-OES) are in general less severe than in competitive techniques such as flame atomic ab- sorption or electrothermal atomic absorption spectrometry significant matrix effects have been reported. l 4 Inorganic and organic acids are commonly used for sample preparation e.g.extraction digestion or dissolution and consequently are major components of the sample matrix in many routine ana- lytical situations. One of the first reports of interferences caused by acids was published by Greenfield et The inter- ferences found were largely attributed to differences in the transport of the acid-containing solutions compared with those of aqueous solutions into the plasma. Farino et d.,6 in a more extensive study of the effects of mineral acids suggested that the observed suppression of the analyte emission was caused by differences in the mass of analyte entering the plasma. The change in analyte mass transport into the plasma was attri- buted to differences in the viscosity between the acid and the aqueous solutions.Viscosity is known to affect directly primary droplet formation by the nebulizer as indicated by the viscosity term which appears in the Nukiyama and Tanasawa equation.' Viscosity also affects droplet formation by changing the solution uptake rate of the nebulizer. The effects of organic acids have also been and have generally been found to cause an increase in the analyte emission intensity. This effect has been attributed to increased nebulization efficiency and increased plasma temperature.' There are several methods of compensating for or removing the effects of acids during an analysis. The most successful and most widely utilized approaches for routine analysis involve the use of matrix matched calibration standards and the standard additions method.I0 Generic protocols of this type can be applied to most problems associated with analyses in acid matrices.However alternatives such as measurement of the hydrogen line (HP) at 486.133 nm to compensate for the effects of varying acid concentrations on the introduction of the sample into the plasma" and the use of mathematical models based on regression analysis have also been explored." Internal standardization has been used to improve the per- formance of ICP-OES by (indirectly) reducing plasma noise." This method has also been used to reduce matrix interferences. Two specific developments which have been used successfully for this purpose are the generalized internal reference methodr4 and the parameter related internal standard method.l5.lh Both of these methods use two or more internal standards to correlate the analyte emission with various plasma fluctuations.However other workers have shown that the use of a single in- ternal standard might be applicable in certain situations. Dal- quist and Knoll" used cobalt as an internal standard to compensate for acid interferences. Uchida et al." used yttrium as an internal standard with a micro-sampling technique to compensate for changes in the amount of sample fed to the plasma. The choice of the internal standard for use during analysis has been the subject of extensive study. Barnett et ~ l . l ' . ~ ~ ' pro- posed guidelines for matching the physical properties of the analyte and reference element so that the ratio of the emission intensity from the analyte to the emission intensity from the in- ternal standard was insensitive to slight fluctuations in the plasma parameters.Sedcole et al." found that no one element was suitable for correcting interferences on all analytes. The importance of the purity and solubility of the internal standard was highlighted in a study by Wallace,'! who also studied pos- sible spectral interferences by the internal standard on the analyte wavelengths. In investigating real-time internal standardization (i-e. si- multaneous measurement of the analyte and internal standard intensity) Belchamber and Horlick23 found that it was possible to correlate the emission from the internal standard with the emission from the analyte element. Scandium was shown to be a promising internal standard (i.e. there was good correlation between the scandium emission and the emission from several of the analytes) compared with the other elements studied.It was found that measurement precision was generally improved by a factor of two using real-time internal standardization. In a similar correlation study Myers and Tracy" used manganese as an internal standard while studying the noise behaviour of 20 analytically important elements. Carrier gas flow-rate and observation height were found to change significantly the degree of correlation between the emission from the analyte element and the internal standard. Hence by altering plasma conditions a single internal standard could be used to improve analytical performance. The plasma background emission in the range 200-700 nm was also studied.It was found that the160 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 plasma background showed sensitivity to the fluctuations in the r.f. power and the carrier gas flow-rate. However fluctuations in the analyte emission and the background emis- sion were generally uncorrelated. Schmidt and SlavinI3 pursued the reaktime internal stan- dardization approach using scandium as an internal standard. In addition to sampling the analyte and internal standard emis- sion in real time the background intensity was also continous- ly measured. Prior to calculating the internal standard ratio the background emission intensity was subtracted from both the analyte and internal standard emission intensities.The net in- tensities were then used to determine the internal standard ratio. This procedure has been referred to as Myers-Tracy signal compensation (MTSC)zS-z6 and has been used to com- pensate successfully for the interference effects caused when solutions of NaCII3 and KCIzs with a high content of dissolved solids have been nebulized. It has also been shownzs that by using MTSC there can be considerable improvements in ana- lytical precision for elements which exhibit different behavi- ours in the ICP. It would appear from a survey of the literature that the in- ternal standardization approach may offer the option of an aqueous standard calibration strategy for the examination of samples containing acids in conjunction with improved meas- urement precision.The application of the MTSC procedure to the ICP-OES determination of copper and manganese in acidic media is described in the present paper. Experimental Instrumentation A Perkin-Elmer Plasma I1 sequential ICP-OES system was employed for all measurements. The spectrometer has a focal length of 1 .O m and spectral resolution of 0.018 nm. The stan- dard operating parameters for the spectrometer can be found elsewhere.2S The plasma conditions used throughout unless otherwise stated are given in Table 1. The de-mountable plasma torch unitz7 was fitted with a 2 mm i.d. acid resistant alumina injector. The sample introduction system consisted of a Ryton crossflow nebulizer and associated double-pass spray chamber. Samples were introduced into the nebulizer by means of the manufacturer’s peristaltic pump.An integration time of 1 s was employed for all measurements. Precision data were obtained from ten replicate integrations. The mode of op- eration of the MTSC assembly has been described in detail previ~usly.~~ All calculations were performed on a Perkin- Elmer 7500 Series computer. Reagents Manganese and copper standard solutions were freshly prepared as required from stock solutions containing 1000 pg ml-1 of the appropriate element (SpectrosoL BDH Poole Dorset UK). A 1000 pg ml-I stock solution was prepared by dissolving the ap- propriate amount of scandium oxide (BDH laboratory-reagent grade 99.9%) in hydrochloric acid. Aristar grade acids were used in the preparation of all ‘interferent’ solutions. Deionized distilled water was used as a diluent and blanks of all the rea- gents were monitored during the study for contamination.Table 1 Standard plasma conditions Plasma parameter Value 1 .oo 1 .oo Plasma gas flow-ratefi min-’ 15.0 Observation height (above induction coil)/mm 15 Carrier gas flow-rate/l min-’ Intermediate gas flow-rate/l min-l Applied power/kW 1 .o Table 2 Optimum plasma conditions for manganese and copper Plasma parameter Mn (257.6 nm) Cu (324.8 nm) Carrier gas flow-ratel1 min-’ 1 S O 1.30 Plasma gas flow-ratefl min-’ 15.0 15.0 Observation height (above induc- tion coil)/mrn 12 18 Applied power/kW 0.80 0.80 Intermediate gas flow-rate/l min-’ 1 .oo 1 .oo Internal Standard A concentration of 20 pg ml-l of scandium was used in all solu- tions because this level of internal standard had been previously shown to be the minimum required to achieve adequate preci- sion and to avoid spectral interferences at analyte wavelengths.” Results and Discussion Aqueous Solutions It was of interest initially to determine whether the MTSC pro- cedure improved the performance of the Perkin-Elmer Plasma I1 instrument for determinations in aqueous solutions. Calibra- tion graphs for manganese and copper were obtained while using the optimum conditions for each element both with and without signal compensation.The optimum conditions for both the manganese ion line at 257.6 nm and the copper atom line at 324.8 nm were obtained by univariate optimization. They are given in Table 2. Four solutions were used to obtain the calibration graphs for manganese.These contained 0 1.0 5.0 and 10.0 pg ml-’ of manganese and 20 pg ml-I of scandium (internal standard). The solutions were used for both calibration graphs i.e. with and without MTSC. This ensured that any differences ob- served in the calibration graphs were not caused by using dif- ferent solution concentrations but resulted from the different signal processing procedures. Background correction was em- ployed. Calibration graphs were constructed for manganese from data obtained with and without the use of MTSC. The results of the calibration for manganese including values of the relative standard deviation (RSD) of the manga- nese intensities and the manganese ratios are shown in Table 3. Linear regression was used to fit the best line through the points on the calibration graphs.The values of the regression coefficient (Y) standard error of the slope and standard error of the intercept for each line are given in Table 3. It was noticeable that when MTSC was used the scatter of data points about the regression line was significantly reduced in comparison with conventional measurement. It was also noted that the measurement precision improved by more than a factor of two when MTSC was used. It was important to verify that the observed improvement in instrumental performance using the MTSC procedure for man- ganese was not only true of this particular element but would also be observed for analytes with different excitation charac- teristics. Copper I (324.8 nm) is considered a typical ‘soft’ line and as such requires different excitation conditions from man- ganese.’H Cali bration graphs were therefore obtained for copper using a similar procedure to that described for manganese. The emission intensities emission ratios and RSDs for each meas- urement are presented in Table 4.The precision of the measure- ments was again found to be improved generally by a factor of two when MTSC was employed. Acid Solutions Manganese and copper (wavelengths 257.6 and 324.8 nm re- spectively) were chosen as representative analyte lines for the study of acid interferences. The effects of hydrochloric nitricJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 161 Table 3 Calibration data for manganese with and without internal stan- Table 5 Effect of hydrochloric acid on recovery of manganese with and without internal standardization dardizat ion Manganese Emission concentration/ intensity pg ml-' 0 0 I .00 1 4 902.37 5.00 70204.77 10.00 127488.35 Line of best fit- A.From intensities Intercept Slope of line Standard error (slope) Standard error (intercept) I' RSD Emission RSD (intensity) ratio (ratio) (%I (%) 0 2.5 12.34 1 . 1 2.8 6 1.45 0.7 1.5 125.83 0.7 - - B. From ratios 2482.13 Intercept -0.77 12748.84 Slope of line 12.62 0.9986 I' 0.9999 473.7 I Standard error (slope) 0. I70 Standard error 2 658.7 1 (intercept) 1.102 Table 4 Calibration data for copper with and without internal standardi- zation Copper Emission RSD Emission RSD concentration/ intensity (intensity) ratio (ratio) pg ml-' (%) (%) - - 0 0 0 I .00 7 339.69 1.8 6.36 0.7 5 .OO 36091.13 2.5 3 1.90 0.4 10.00 7 1 894.25 3.6 62.64 0.4 50.00 352403.43 3.0 3 18.76 1.6 Line of best fit- A.From intensities B. From ratios Intercept 2 132.56 Intercept -0.35 Slope of line 7039.49 Slope of line 6.248 I' 0.99999 r 0.99994 Standard error (slope) 16.14 Standard error (slope) 0.0171 Standard error Standard error (intercept) 369.86 (intercept) 0.4384 and sulphuric acids were investigated. Standard plasma condi- tions were used. Hydrochloric acid A number of solutions were used containing 10 pg ml-' of manganese 10 pg ml-' of copper and 20 pg ml-' of scandium with hydrochloric acid concentrations of from 0 to 20% (v/v). Analyte signal recovery was calculated as a percentage of the response obtained for aqueous analyte solution.The results of the study for manganese are presented in Table 5. From these results it can be seen that when MTSC was not used the recovery of manganese from hydrochloric acid was about 90% of the total manganese content. If MTSC was used then the manganese recovery was increased to ap- proximately 98% of the total an improvement in recovery of 8%. It was also apparent that without MTSC the precision of the measurements was approximately 2% whereas with MTSC the precision was improved by a factor of five to give values of 0.4%. It is clear that the procedure involving inter- nal standardization resulted in a considerable improvement in both manganese signal recovery and in measurement preci- sion. However it was noted that although values of recovery increased to 98% of the expected value there was still a deficit of 2%.This deficit may be caused by some additional factor which cannot be compensated for by using the internal standardization method or alternatively may be a reflection of Without MTSC With MTSC Acid concentration Recovery RSD* Recovery RSD* (% v/v) (%) (%) (%) 0 100.0 2.2 100.0 0.5 5 91.0 2 .o 97.3 0.4 10 89.2 2.3 97.7 0.3 15 90.4 2.3 97.8 0.4 20 90.5 2.4 98.6 0.5 * The RSD quoted is that of the manganese emission intensity not the RSD of the recovery. ~ ~~~~ the measurement error in the internal standardization proce- dure. A similar series of measurements was also made for copper. The measurement procedure and the solutions used were the same as those in the previous section. The results of the inves- tigation are tabulated in Table 6.As observed for manganese without MTSC there was a considerable decrease in the copper signal recovery when the acid solutions were nebulized. With copper however it was observed that as the acid concentra- tion increased the percentage recovery decreased unlike man- ganese where there was no noticeable trend in recovery with increasing acid concentration. It was observed that the copper signal recovery was significantly decreased when even small acid concentrations were added to the aqueous solution and that the addition of larger acid concentrations did not cause further significant decreases in copper recovery. When signal compensation was used the copper recovery was approximate- ly 100%. Thus MTSC compensated fully for the interference effect of hydrochloric acid on copper.The measurement preci- sion for copper was improved by about a factor of four when using MTSC. Nitric acid The effect of nitric acid on manganese emission was evaluated using a similar procedure to that described above. The nitric acid was added in concentrations ranging from 0 to 20% (v/v). Manganese copper and scandium concentrations were 10 10 and 20 pg ml-I respectively. The results obtained with and without signal compensation are shown in Table 7. The interference effect caused by nitric acid on manganese emission was of approximately the same order of magnitude as that of hydrochloric acid. When internal standardization was used the recovery of manganese from nitric acid was increased by 8% relative to the recovery obtained without MTSC.This improvement was similar to that obtained when MTSC was applied to the measurement of signal recovery for manganese in hydrochloric acid solutions. Precision of measurement was im- proved by about a factor of six using internal standardization. ~~~ Table 6 without internal standardization Effect of hydrochloric acid on recovery of copper with and Without MTSC With MTSC Acid concentration Recovery RSD* Recovery RSD* (% v/v) ( % I (%I (%) (% 1 0 100.0 I .s 100.0 0.4 5 92.6 I .4 100.4 0.4 10 89.9 1.2 101.3 0.4 IS 89.4 1.4 100.9 0.3 20 89.2 I .0 101.3 0.3 * The RSD quoted is that of the copper emission intensity. not the RSD of the recovery.162 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 199 1 VOL. 6 Table 7 internal standardization Effect of nitric acid on recovery of manganese with and without Without MTSC With MTSC Acid concentration Recovery RSD* Recovery RSD* (9% v/v) (%I (%I (%) 0 100.0 1.8 100.0 0.7 5 90.8 1.4 98.8 0.3 10 90.1 2.5 97.9 0.3 15 89.5 2.1 97.2 0.4 20 87.7 1.8 96.4 0.3 * The RSD quoted is that of the manganese emission intensity not the RSD of the recovery.Table 8 Effect of nitric acid on recovery of copper with and without in- ternal standardization Without MTSC With MTSC Acid concentration Recovery RSD* Recovery RSD* (% v/v) (S) (%) (%) (%I 0 100.0 1.3 100.0 0.2 5 90.4 1.3 99.0 0.3 10 88.9 1.4 98.6 0.4 15 90.2 2.0 98.6 0.4 20 87.4 3.0 97.7 0.3 * The RSD quoted is that of the copper emission intensity not the RSD of the recovery. The interference effect of nitric acid on copper emission was investigated using a similar experimental procedure to that de- scribed above.The results of this study are summarized in Table 8 and show that the recovery of copper from nitric acid solutions was reduced compared with the copper recovery from hydrochloric acid. When internal standardization was not used the recovery of the copper signal for copper in 20% (v/v) nitric acid was 87.4% whereas under similar conditions the copper recovery from hydrochloric acid [20% (v/v)] was 89.2%. When MTSC was used the copper recovery from nitric acid [20% (v/v)] was 97.7% and the recovery from 20% hydrochloric acid was 101.3%. These results suggest that the interference effect caused by nitric acid is somewhat greater than that produced by hydrochloric acid. As found previously the precision of measurement was improved considerably (up to a factor of six) by using the internal standardization method.This is attributed to removal of the variation in rate of sample uptake with the less viscous acid solution. Thus it would appear that source flicker noises in the plasma derived from sample transport effects may be eliminated by correlating emission intensities in real time using an intensity ratio meas- urement. Sulphuric acid A similar but extended study of the interference effects of sul- phuric acid was carried out. The concentrations used were 0 0.5 1.0,2.0,5.0 10.0 15.0 and 20.0% (v/v). The experimental procedure used was identical to that described previously. The results are shown in Tables 9 and 10.Considering initially the results obtained for manganese without using signal compensation it can be seen that the in- terference effects of sulphuric acid are considerably more severe than those of hydrochloric or nitric acids. The manga- nese emission recorded from the 20% (v/v) acid solution was only 60% of the expected value. When MTSC was applied the manganese recovery was increased by up to 33% and the minimum manganese recovery was 94.2%. There was a trend Table 9 Effect of sulphuric acid on recovery of manganese with and without internal standardization Without MTSC With MTSC Acid concentration Recovery RSD* Recovery RSD* (% v/v) (%Jo) (%I (%) (%b) 0.0 I 0.5 1 .o 2.0 5 .O 10.0 15.0 20.0 100.0 95.7 95.7 92.9 82.4 73.6 66.0 61.0 I .4 2.2 1.8 2.8 1.8 1.4 1.6 2.3 100.0 98.9 98.0 98.7 96.9 95.6 95.0 94.2 0.1 0.2 0.2 0.2 0.3 0.4 0.6 0.4 * The RSD quoted is that of the manganese emission intensity not the RSD of the recovery.Table 10 without internal standardization Effect of sulphuric acid on recovery of copper with and Without MTSC With MTSC Acid concentration Recovery RSD* Recovery RSD* (% VIV) (%o) (%'o) 0.0 0.5 1 .o 2.0 5.0 10.0 15.0 20.0 100.0 96.8 95.8 93.7 82.3 72.9 66.7 61.6 1.8 1.4 2.4 1.2 1 .o 0.7 0.8 1.5 100.0 98.1 96.2 97.8 95.4 94.4 93.9 92.3 0.3 0.3 0.2 0.4 0.4 0.5 0.5 0.6 * The RSD quoted is that of the copper emission intensity not the RSD of the recovery. of decreasing signal recovery with increasing acid concentra- tion which was most noticeable when MTSC was not used. The precision was improved by up to ten times when internal standardization was used.However the improvement in meas- urement precision was not as consistent as that observed for hydrochloric and nitric acids (when an improvement in preci- sion by a factor of 5-6 was observed at each acid concentra- tion). Nevertheless the benefits of MTSC for analysis of a sulphuric acid matrix are obvious. The recovery of copper in the presence of sulphuric acid so- lutions was also substantially lower than for the other acids (see Table 10). The magnitude of the interference effect of the sulphuric acid on copper was similar to that observed for man- ganese. There was again a significant improvement in signal re- covery when the signal compensation procedure was used [from 61.6 to 92.3% at 20% (v/v) sulphuric acid].As expected measurement precision was dramatically improved using MTSC but the values of RSD obtained were less consistent than those for similar solutions of hydrochloric or nitric acid. There is a trend towards poorer precision at higher acid concen- trations using MTSC but even using 20% (v/v) acid the RSD is still better than that achieved without signal compensation. Conclusions During this study it was found that when MTSC was applied to the measurement of copper and manganese emis- sion intensities from aqueous solutions the measurement precision increased by a factor of two. It was also apparent that in constructing calibration graphs the MTSC procedure may reduce the scatter of data points about the regression line. In determining the analyte signal recovery from hydro-JOURNAL O F ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL.6 163 chloric nitric and sulphuric acids it was shown that the MTSC procedure enhanced the analyte signal recoveries thereby reducing or removing totally (for nitric acid) the suppressive effects of the acids on analyte emission. In general an improvement in measurement precision was also observed. In view of the consistently high signal recoveries obtained using MTSC which were largely independent of acid concen- tration it is clear from this work that it should be possible to use aqueous calibration standards for ICP-OES analysis of acid solutions. The advantage of this approach is that the inter- nal standard can be monitored simultaneously to check for bias and drift problems and thereby offers an in-built quality check for the analysis.Although signal recoveries were not always 100% the use of an internal standard ensures consistency of response which is not necessarily true when matrix matching of standards are used. Obviously the two procedures are not mutually exclusive and there may still be an advantage in using internal standardization in conjunction with the matrix matching. Internal standardization was not found to compensate totally for the effects of the mineral acids used. However it may be possible to remove residual interferences by optimizing the plasma parameters. In a multi-element situation it may be necessary to optimize the plasma conditions sequentially in order to give improved correlation of the analyte emission with that of the internal standard.This possibility is presently being examined. The authors gratefully acknowledge the support of the SERC (for L.M.G.) and ICI Wilton Materials Research Centre (Dr. W. C. Campbell) for providing funding for this work. They also thank ICI plc for permission to submit this paper for publication. References 1 Blades M. W. and Horlick G. SpecnmAim. Acta. Part B. 198 I 36. 881. 2 Koirtyohann S. R. Jones J. S.. Jester C. P.. and Yates D. A.. Spec- trochim. Acta Part B I98 I 36,49. 3 Moore G. L. Humphries-Cuff. P. J. and Watson A. E.. Specfro- chim. Acta Part B 1984,39. 9 15. 4 Bamiro F. O. Littlejohn D. and Marshall J.. J. Anal. At. Spec~trnni.. 1988.3.279. 8 9 10 I I 12 13 14 IS 16 17 18 19 20 21 22 23 24 25 26 27 28 Greenfield S.McGeachin H. McD. and Smith P. B.. Anal. Chim. Acta I976.84,67. Farino J. Miller J. R.. Smith D. D. and Browner R. F. Anal. Chem. 1987 59,2303. Nukiyama S.. and Tanasawa Y.. in Experiments on the Atomisation of Liquids in an Air Stream. translator Hope E.. Defence Research Board Department of National Defence Ottawa Ontario Canada 1950. Xu. J. Kawaguchi H.. and Mizuike A. Anal. Chim. Acts 1983 152 133. Hettipathirana T. D. Wade A. P. and Blades M. W. Spactroc.him. Acta Part B. 1990.45 27 I . Kalivas J . H. and Kowalski B. R.. Anal. Chem. 1982,54 560. Botto R. I. Spectrochim. Acta. Part B 1985 40 397. Delijska A.. and Vouchkov M. Fresenius Z. Anal. Chem.. 1985 321,448. Schmidt. G. J. and Slavin W. Anal. Chem.. 1982,54 2491. Lorber A. Goldbart 2.. Harel A.. Sharvit E. and Eldan. M. Spec- rroc.him. A m Purr B . 1986,41 105. Ramsey. M. H. and Thompson M. Analyst 1985 110. 5 19. Thompson M.. and Ramsey M. H.. AnalJst. 1985. 110. 1413. Dahlquist. R. L. and Knoll I. W.. Appl. Spe(,trosc. 1978 32. I . Uchida H. Nojiri Y. Haraguchi H.. and Fuwa K. Anal. Chim. Acta 198 1. 123 57. Bamett W. B. Fassel V. A. Kniseley R. N. Spectrochim. Acta Part B 1968,23,643. Barnett W. 9.. Fassel V. A.. Kniseley R. N.. Spertrochim. Acta Part B 1970,25 139. Sedcole J. R. Lee J.. and Pritchard M. W.. Spertr-ochim. Actu Part B 1986.41.227. Wallace G. F. At. Specrrosc. 1984 5 5. Belchamber R. M. and Horlick. G.. Specri.ochin1. Acta Part B. 1982 37. 1037. Myers S. A. and Tracy D. H.. Spectrochim. A m Part B. 1983 38. 1227. Marshall J.. Rodgers. G.. and Campbell W. C.. J. A~zul. At. Spec- rrom. 1988 3 241. Salit M. L.. Pruszkowski. E.. Yates. D. A.. and Collins J. 9.. paper presented at the Pittsburgh Conference and Exposition New Orleans. LA. USA 1988 Paper 1320. Wallace G. F.. Pirc. V. V.. and Angel A. A. Appl. SpectImc.. 1985. 5. 195. Boumans. P. W. J. M.. and Lux-Steiner. M. Ch. Spwrrochim Acta Part B . 1982.37,97. Paper- 0/01962C Receiited May 2nd. I990 Accepted December l o t h 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600159

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1991 VOL. 6 I65 Matrix Interferences Observed With a Thermospray Sample Introduction System for Inductively Coupled Plasma Atomic Emission Spectrometry Margaretha T. C. de Loos-Vollebregt Runzhong Peng* and Johan J. Tiggelman Laboratory of Analytical Chemistry Delft University of Technology De Vries van Heystplantsoen 2 2628 RZ Delft The Netherlands Matrix interferences due to the presence of metallic ions were studied for a thermospray sample introduction system used with inductively coupled plasma atomic emission spectrometry. Corresponding results obtained with a V-groove nebulizer indicated that the interferences were more pronounced in the thermospray system. In the presence of mineral acids (HCI HC104 and H,PO,) similar interferences were observed for the thermospray sample introduction system and the V-groove nebulizer.The excitation temperature of the plasma was about 500 K lower when the sample was introduced in the form of a thermospray. The electron density was also lower when using the thermospray system. Keywords Inductively coupled plasma atomic emission spectrometry; thermospray sample introduction; matrix interference ; V-groo ve nebulizer Several workers have reported on the use of a thermospray sample introduction system as an alternative to pneumatic nebulizers for sample introduction used with inductively coupled plasma atomic emission spectrometry (ICP-AES). I-R The thermospray sample introduction system produces a higher signal to background (SD) ratio and the detection limits are lower.The reduction of detection limits has been at- tributed to the higher analyte transport efficiency of the ther- mospray vaporizer.s Vermeiren et al.2J reported signal depressions for different elements in a matrix of 1 g I-' of NaCl and in a matrix of 1.7 g 1-' of NaNO for a conventional plasma with side-on observation. They concluded that the mass transport efficiency for Mn and Zn decreased in the presence of a matrix of NaN03 using a thermospray vaporizer in combination with a conical flask spray chamber a heated tube and a condenser.K Significant matrix interferences from metallic ions and H3P04 were also reported by de Loos-Vollebregt et ~ 1 . ~ for thermospray sample introduction into a water cooled low gas flow plasma with end-on observation. In the present study the effects of different matrix salts and mineral acids on measure- ments of Cd Mg and Mn using conventional ICP-AES are in- vestigated using a thermospray sample introduction system with a water cooled spray chamber.Experimental The thermospray sample introduction system was described in detail in a previous p~blication.~ The diameter of the fused silica capillary was 50 pm. The operating conditions for the thermospray sample introduction system are presented in Table 1. The aerosol was introduced directly into a 125 ml water cooled spray chamber in order to remove part of the water vapour from the spray before introduction into the plasma. The cooling water for the spray chamber was kept at a temperature of 287 K. For comparison the measurements were repeated with a V- groove nebulizer.The V-groove nebulizer was constructed in this laboratory and had a gas introduction hole of i.d. 0.2 mm and a sample introduction channel of 1 mm. The V-groove nebulizer was used with a 125 ml Scott-type spray chamber without a cooling system. A Perkin-Elmer ICP Plasma I1 spectrometer with a holo- graphic grating of 1800 lines mm-I and a Hamamatsu R 787 * On leave from the Center of Geological Analysis Jiangxi Bureau of Geology and Mineral Resources Nanchang People's Republic of China. Table 1 Operating conditions for the V-groove nebulizer and the thermo- spray system Parameter Setting Generator power Carrier argon flow-rate V-groove nebulizer Thermos pray Outer argon flow-rate Auxiliary argon flow-rate Observation height V-groove nebulizer Thermos pray Sample uptake rate V-groove nebulizer Thermos pray Thermospray vaporizer temperature 1.04 kW 1.14 1 min-' 1 .OX I min-' 15.0 1 min-' 1.0 I min-' 17 mm 19 mm 1.0 ml min-' 0.3 ml min-' 610 K photomultiplier tube were used in all experiments.The crystal controlled radiofrequency (r.f.) generator operates at 27.12 MHz with a maximum power of 1.8 kW. Data acquisition and processing were performed with a Perkin-Elmer 7500 Profes- sional Computer and Plasma I1 applications software version 9.6. The integration time was 0.1 s in all experiments. The standard solutions were prepared from 1000 mg I-' stock solutions (Merck) by dilution with de-ionized water. Results and Discussion The optimum conditions obtained for the S/B ratio of Mn I1 257.610 nm according to the approach of Boumans and Lux- Steinerg for thermospray sample introduction and for the V- groove nebulizer are presented in Table 1.Influence of a Matrix of Metallic Ions The influence of the matrix of KCI NaCl and CaCI on the in- tensities of the ion and atom lines of Cd Mg and Mn were in- vestigated. For these measurements the optimum instrumental conditions shown in Table 1 were used for both sample intro- duction systems. The analyte concentration was constant at 0.2 mg 1-I while the matrix concentration varied between 0 and 200 mg 1-I for the K Na and Ca ions. The relative intensity of the Cd I 228.802 nm line measured in the matrix is presented in Fig. I(a) for the thermospray sample introduction system.An enhancement of 25% was found in each matrix for the highest concentration of the inter- ferent. The corresponding results obtained with the V-groove166 1.3 1.2 1.1 1 0.9 .- 3 v) 0.8 .f 1.3 c .- c - al a 1.2 1.1 (a) Oa9 t I I I I 0 0.2 2 20 200 0.8 ' Ion concentration/mg I-' Fig. 1 The influence of increasing concentrations of A Na; B Ca; and C K on the intensity of the Cd 1228.802 nm emission line. (a) Thermo- spray sample introduction and (h) V-groove nebulizer JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH I99 I VOL. 6 1.3 1.2 1.1 1 0.9 t 0.9 I 1 J 0 0.2 2 20 200 0.8 Ion concentration/mg I-' Fig. 2 The influence of increasing concentrations of A Na; B K; and C. Ca on the intensity of the Cd I1 214.448 nm emission line. (a) Thermo- spray sample introduction and (h) V-groove nebulizer 1.3 1.2 1.1 (a) 0.6 0 0.2 2 4 8 Acid concentration (%I Fig.3 The influence of increasing concentrations of A HCI; B HCIO,; and C H3P04 on the intensity of the Cd I 228.802 nm emission line. (a) Thermospray sample introduction and (h) V-groove nebulizer 1.6 3 1.4 .- v) Q) c .- 0) 1.2 > .- c - I2 1 c! A B 0.8 1 1.2 1.4 1.6 Rf powerikw Fig. 4 The influence of increasing rf power on the intensity of the Cd I 228.802 nm emission line measured in a matrix of 20 mg I-' of Na at dif- ferent observation heights A 17; B 19; and C 21 mm using thermospray sample introduction nebulizer are shown in Fig. l(h). The matrix effect was less than 5% with the V-groove nebulizer for concentrations of up to 200 mg I-' of the ions studied. Similar results were found for the Cd I1 214.448 nm line.In the thermospray sample introduction system the matrix effect was 15 20 and 25% for Ca K and Na respectively [Fig. 2(a)] whereas again no significant influence of the matrix was ob- served with the V-groove nebulizer [Fig. 2(h)]. The influence of the matrix was similar for the Mg I1 279.553 Mg I 285.213 and Mn I1 257.610 nm lines. Effect of Mineral Acid The matrix effect of the mineral acids HCI HCIO and H,POJ was measured. The concentration of the acids was variedJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 199 1 VOL. 6 167 0.8 I I 0.8 1 1.2 1.4 1.6 Carrier g a s flow-rate/i min-' Fig. 5 The influence of the carrier gas flow-rate on the intensity of the Cd I 228.802 nm emission line measured in a matrix of 20 mg I-' of Na at different observation heights A 17; B 19; and C 21 mm using thermo- spray sample introduction 5m:2 16 118 ;O ;2 :4 Observation height/mm Fig.6 Excitation temperature measured with A. thermospray sample in- troduction; and B the V-groove nebulizer at different observation heights in the plasma 3 le8 t s 5 1.6 - L 0 1.4 - .- 3 s 1.2 - g 1 - cn 0 C - 2 0.8 - w I I 1 0.6 12 14 16 18 20 22 Observation heightlmm Fig. 7 Electron density measured with A thermospray sample introduc- tion; and B. the V-groove nebulizer at different observation heights in the plasma between 0 and 8% (m/m). The elements and experimental conditions were the same as those used in the study of the influence of metallic ions. The results for Cd I 228.802 nm are shown in Fig.3(a) for the thermospray sample introduction system and in Fig. 3(h) for the V-groove nebulizer. No matrix effect was observed for HCI with the thermospray sample in- troduction system and for the V-groove nebulizer. The effect of HCIOl is somewhat larger for the V-groove nebulizer i.e. -25% compared with -1 I % in the thermospray sample intro- duction system. The results found for the Cd I1 line at 214.438 nm were similar to those found for Cd I 228.802 nm in Fig. 3 (a) and (h). A severe matrix effect of up to -35% was ob- served for H3P04 in the thermospray system and -25% for the V-groove nebulizer. Optimization The influence of the generator power on the observed interfer- ences was studied for a matrix of NaCI. The matrix effect of 20 mg 1-I of Na which is in the middle of the range studied so far was measured for Cd Mg and Mn ion and atom lines at four different power levels (0.94 1.14 1.34 and 1.54 hW). The results for Cd I 228.802 nm are shown in Fig.4 for obser- vation heights of 17 19 and 21 mm. From Fig. 4 it was con- cluded that the matrix effect does not change significantly with the generator power for the different observation heights. Similar results were found using the V-groove nebulizer. The influence of the carrier argon flow-rate on the matrix effect was also studied for Cd Mg and Mn in the presence of 20 mg I-' of Na. The flow-rate was varied between 0.9 and 1.5 1 min-I for both the thermospray system and the V-groove neb- ulizer. For the thermospray sample introduction system the results for Cd I 228.802 nm measured at observation heights of 17 19 and 2 I mm are shown in Fig.5. A slightly increasing matrix effect up to about +lo% was measured at observation heights of 17 and 21 mm whereas at an observation height of 19 mm a signal enhancement of 30% was found. Results similar to those observed for the Cd atom emission line were obtained for the Cd ion emission line and also for the atom and ion emission lines of the other elements studied. With the V-groove nebulizer the matrix effect was within 10% at observation heights of 17-21 mm for all the emission lines studied. Excitation Temperature and Electron Number Density The excitation temperature of the plasma was measured from the emission intensities of the Fe lines reported by Blades and Caughlin.1" The temperature was calculated from the Boltz- mann plots of ln(Ih"-tf-1) versus En measured for 1 g I-' of Fe where I is the line intensity h is the wavelength of the Fe I transition g is the degeneration of the upper level,fis the os- cillator strength of the transition and En is the excitation energy.The temperature was measured at different observa- tion heights of between 13 and 23 mm above the load coil. The results are presented in Fig. 6. The temperature is almost con- stant for all the observation heights. The temperature measured with the thermospray sample introduction system is about 5700 K whereas in the experiments with the V-groove nebu- lizer the measured excitation temperature is about 6150 K. The decrease in the excitation temperature is in agreement with a higher solvent load of the plasma.The electron number density was measured at observation heights of between 13 and 23 mm above the load coil using the width of the HP 486.1 nm line." The results are presented in Fig. 7. The electron density decreases with the observation height for the thermospray system and for the V-groove nebu- lizer. The electron density is lower when the thermospray system is used compared with the V-groove nebulizer. The aerosol of the thermospray system causes a higher S/B ratio owing to the greater amount of smaller droplets that reach the plasma.' Unfortunately the higher solvent load of the plasma decreases the temperature of the plasma and the electron density and results in increased matrix interferences. Conclusions The thermospray sample introduction system for ICP-AES de- scribed in this paper provides on average about three times better detection limits compared with the V-groove nebulizer. A significant enhancement of interferences is observed forI68 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 199 I VOL.6 thermospray sample introduction. It can be concluded that al- though the thermospray system offers better sensitivity it can only be successfully applied for sample solutions of low salt concentrations. The performance of the thermospray system is similar to that of the ultrasonic nebulizer.I2 The sample introduction system is expensive because a non- metallic high-performance liquid chromatography (HPLC) pump is required. Moreover the washing and stabilization time is longer compared with pneumatic nebulizers when changing from one sample solution to another in the thermo- spray system.Future applications of the thermospray system are antici- pated in flow injection and interfacing the HPLC with ICP- AES for speciation of organometallic compounds. Financial support from Perkin-Elmer Norwalk CT USA is gratefully acknowledged. R. P. thanks the Chinese State Educa- tion Committee and the Dutch Ministry of Education for a grant. References 1 Elgersma J . W. Maessen F. J. M. J. and Niessen W. M. A Spec- trochim. Actu Part B 1986,41 1217. 2 Vermeiren K. A. Taylor P. D. P. and Dams R. J . Anal. At. Spec- trom. 1987,2,383. 3 Koropchak J. A. and Winn D. A. Appl. Spectr-osc. 1987,41 13 1 1. 4 Koropchak J. A. and Winn D. H. Trends Anal. Chem. 1987 6 171. 5 Koropchak J. A. Aryamanya-Muchisha H. and Winn D. H. J. Anal. At. Spectrom. 1988,3,799. 6 de Loos-Vollebregt M. T. C. Tiggelman J. J. Bank P. C. and De- graeuwe C. J. Anal. At. Spectrom. 1989,4213. 7 Peng R.-z. Tiggelman J. J. and de Loos-Vollebregt M. T. C. Spec- tsochim. Acta Part B 1990,45 189. 8 Vermeiren K. A. Taylor P. D. P. and Dams R. J. Anal. At. Spec- trom. 1988,3,57 I 9 Boumans P. W. J. M. and Lux-Steiner M. Ch. Spectrochim. Acta. Past B 1982,37,97. 10 Blades M. W. and Caughlin B. L. Spectsochim. Actu. Part B 1985,40,579. 1 I Hasegawa T. and Haraguchi H. in Inductively Coupled Plasma in Analytical Atomic Spectrometry eds. Montaser A. and Golightly D. W. VCH New York 1987 ch. 8. 12 Fassel V. A. and Bear R. Spectrochim. Acta Part B 1986 42 1089. Paper OJ03079A Received July 9th 1990 Accepted November Sth 1990

ISSN:0267-9477    

DOI:10.1039/JA9910600165

出版商:RSC    

年代:1991

数据来源: RSC