摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 23 Mechanisms of Transition Metal Interferences in Hydride Generation Atomic Absorption Spectrometry Part 4.* Influence of Acid and Tetrahydroborate Concentrations on Interferences in Arsenic and Selenium Determinations Bernhard Welz and Marianne Schubert-Jacobs Department of Applied Research, Bodenseewerk Perkin-Elmer & Co GmbH, 0-7770 Uberlingen, FRG An increase in the hydrochloric acid concentration from 0.5 to 5 mol 1-1 improves the range of interference-free determination for arsenic and selenium in the presence of cobalt, copper or nickel by factors of 5100. Decreasing the sodium tetrahydroborate(ll1) concentration from 3 to 0.5’/0 mlVincreases the range of interference-free determination for both elements further by factors of 4-50.The only exception is the interference of copper on selenium, which is slightly more pronounced with the less concentrated tetrahydroborate(ll1) solution. The proposed mechanism is that both in higher acid and with lower tetrahydroborate(ll1) concentrations the interferent is reduced to a lesser extent to the metal (the interfering species) owing to ( i ) better solubility of the metal in the more concentrated acid, (ii) the formation of chloro complexes, thus reducing the concentration of free ions, and (iii) a larger percentage of the tetrahydrobor- ate(lll) being consumed by the acid. Copper, however, interferes with selenium in the ionic form. Keywords: Hydride generation atomic absorption spectrometry; arsenic determination; selenium determination; interference mechanisms; sodium tetrah ydroborate(ll1) concentration It is well documented that a number of transition metals, mainly those of Groups VIII and IB, can cause severe signal depressions in hydride generation atomic absorption spec- trometry.Smith1 made the first systematic study on the effect of 48 elements on the determination of hydride-forming elements and found that many of the interfering elements formed precipitates after the addition of sodium tetrahydro- borate(II1). He proposed that preferential reduction of the metal ion interferent in solution to a different oxidation state or to the free metal can cause precipitation of that species, which can then either coprecipitate the analyte element or adsorb the volatile hydride formed and catalytically decom- pose it.Kirkbright and Taddia2 also noticed that, in the presence of elements such as nickel, palladium or platinum, after the addition of the reductant a finely dispersed black precipitate was formed. For the determination of arsenic virtually complete suppression of the signal was observed on addition of nickel powder. The authors pointed out that nickel and other Group VIII elements are hydrogenation catalysts and can adsorb hydrogen in large amounts. Hence capture and decomposition of the hydride by the finely dispersed metal can occur. Meyer et al. 3 reported numerous interferences from transi- tion metals on the determination of selenium but did not mention an interferent precipitation. They proposed that the selenium hydride, after its generation, forms insoluble sele- nides or stable complexes with the free ions of the interfering elements in a secondary reaction when it is transported through the sample solution by the carrier gas and the hydrogen.In Part 1 of this series,4 we investigated transition metal interferences on the determination of selenium in a system in which the selenium hydride is generated in pure acid solution and comes into contact with the metal ions only in a second flask. It could be shown that capture and decomposition of the selenium hydride by the finely dispersed metal is the most * For Part 3 of this series, see reference 5. likely mechanism of interference whenever precipitation of the metallic species occurs. It could also be shown that several orders of magnitude higher concentrations of the interfering metal can be tolerated when its reduction and precipitation are avoided or delayed, e .g . , by the addition of a buffer ion which is more easily reduced.5 One of the simplest and most frequently applied methods of reducing transition metal interferences is to increase the acidity of the reaction mediurn.3.4>6-* The explanation most frequently used for this phenomenon is the increased solubil- ity of the reduced metal and/or of the compound formed between the analyte element and the interferent in the strong a ~ i d . 3 ? ~ Another possible explanation would be that more of the tetrahydroborate(II1) is used up by the higher acid concentration and converted into hydrogen so that less is available for the reduction of the interfering element to the metal.If this is true, a reduction of the tetrahydroborate(II1) concentration should have the same effect. However, there is no indication in the literature that would support this assumption. Sodium tetrahydroborate(II1) concentrations reported in the literature vary between 0.3 and 10% mlV, but most workers have used concentrations between 1 and 3% mlV. The only criterion applied to the selection of the optimum tetrahydroborate concentrations is typically the best sensitiv- ity that can be obtained for reference solutions.9-13 Thompson and Thomersonl4 even found that the tetrahydroborate concentration should be increased in order to ensure adequate reduction of the analyte element when analysing samples with high concentrations of transition metal ions.Evans et al. 15 also felt that it is desirable to maximise the amount of tetrahydro- borate(II1) used to overcome consumption of this reagent by other elemental species present. Dittrich et al. ,I6 on the other hand, found that only a very small percentage (<0.1%) of the tetrahydroborate(lI1) is used to generate the hydride, and that the consumption of reductant by competitive reactions cannot be the only reason for the observed interferences. In this work, we have systematically investigated the influence of different sodium tetrahydroborate(II1) concen- trations (0.5, 1.0 and 3.0% mlV) on some typical transition metal interferences in the determination of arsenic and24 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 ~ Table 1. Operating parameters for the MHS-20 hydride system Cell tempera- Element Purge I/s Reaction/s Purge II/s ture/OC 10 40 900 8 35 900 As . . . . . . 35 Se . . . . . . 35 0.5 mol 1-1 HCI 0'4 1 3% I I 1 %o 5.0 mot I-' HCI 0.3 0.5% 5: 0.2 - 0.1 - I I 0 5 10 Ti me/m i n 15 Fig. 1. Influence of hydrochloric acid concentration (0.5 and 5 moll-1) and sodium tetrahydroborate(II1) concentration (0.5, 1 and 3% m/V) on the signal of 50 ng of As(V) in 10 ml of solution selenium in low (0.5 moll-1) and high (5 moll-1) hydrochloric acid concentrations. The purpose of the investigation was to shed more light on the sometimes contradictory reports about the influence of the tetrahydroborate(II1) concentration on interferences, and to obtain a greater insight into the interference mechanisms in hydride generation AAS.Experimental Apparatus A Perkin-Elmer Model 4000 atomic absorption spectrometer equipped with electrodeless discharge lamps, operated at 8 W for arsenic and at 6 W for selenium from an external power supply, was used for all determinations. A spectral band pass of 0.7 nm was selected to isolate the 193.7-nm arsenic line, and 2 nm was used for the determination of selenium at the 196.0-nm line. The signals were recorded on a Perkin-Elmer Model 56 recorder set at the 10-mV range. All measurements are expressed as peak height unless stated otherwise. The Perkin-Elmer Model MHS-20 hydride system used has been described in detail elsewhere17; the instrumental settings used for the determination of arsenic and selenium are summarised in Table 1.Reagents All reagents used, except for the sodium tetrahydrobor- ate( 111) powder, were of analytical-reagent grade or higher purity. Hydrochloric acid was further purified by sub-boiling distillation. Sodium tetrahydroborate(III) solution, 0.5, 1 and 3% mlV. Prepared by dissolving sodium tetrahydroborate( 111) powder (Riedel-de-Haen) in de-ionised, distilled water and stabilising with 1% mlV sodium hydroxide solution. The solutions were prepared freshly every day and filtered before use. Arsenic( V) and selenium(IV) stock standard solutions, 1000 mg 1-1. Prepared by diluting Titrisol solutions (Merck) containing 1000 mg of the element to 1 1 with de-ionised, distilled water. Aliquots were diluted with 0.5 moll-1 HCl to obtain appropriate working reference solutions.0.5 0.4 0) 0 0 0.3 5: a 11 0.2 0.1 r 1 O/O 0.5% 5.0 mol 1-1 HCI 1 . . I 0 5 10 15 Ti me/m i n Fig. 2. Influence of hydrochloric acid concentration (0.5 and 5 moll-l) and sodium tetrahydroborate I11 concentration (0.5, 1 and 3% mlV) on the signal of 100 ng of Se[IV] in 10 ml of solution Results and Discussion The influence of three transition metals, cobalt, nickel and copper, on the absorbance signal of arsenic and selenium was investigated in 0.5 and 5.0 moll-1 hydrochloric acid solution using 3, 1 and 0.5% mlV sodium tetrahydroborate(II1) solution as the reducing agent. Both the hydrochloric acid and the tetrahydroborate(II1) concentrations have an influence on the peak-height sensitivity and signal shape, as shown in Fig.1 for arsenic. A higher acid concentration reduces the signal height and width. A reduction in tetrahydroborate(II1) concentration decreases the signal height but increases its width. The apparently small peak recorded in 5 moll-1 HC1 with 3% mlVNaBH4 is an artifact because the recorder cannot follow the very rapid signal accurately. The behaviour of selenium and the signal shapes recorded are similar to those of arsenic, and are shown in Fig. 2. These differences in absorbance and in peak shape are not taken into consideration when transition metal interferences are investigated because the acid and tetrahydroborate(II1) concentrations are typically kept constant during a set of experiments. The interferences are therefore expressed as relative sensitivity (Yo), and the absorbance (peak height) of the pure analyte solution in the same acid and with the same tetrahydroborate(II1) concentration is set to 100%.A typical graph of relative sensitivity against interferent concentration is shown in Fig. 3 for arsenic in the presence of nickel. It is obvious that an increase in the acid concentration and a decrease in the tetrahydroborate(II1) concentration both result in a substantial increase in the range of interference-free determination of arsenic, which means that higher nickel concentrations can be tolerated without influencing the arsenic absorbance. In most instances the interference starts at a well defined transition metal concentration and becomes very pronounced within an order of magnitude increase in the interferent concentration.The only exception from this otherwise uniform behaviour was found for 5 mol 1-1 HC1 and 0.5% mlV NaBH4. The moderate signal depression caused by nickel concentrations of 200 mg 1-1 or more becomes progressively more pronounced with every determination and irreparably deteriorates the sensitivity of the whole system. Regular cleaning of the quartz tube is essential under these conditions. We believe that aJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 25 0 0.1 1 .o 10 100 1000 Ni concentration/mg I-’ Fig. 3. Influence of sodium tetrahydroborate(II1) concentration (0.5, 1 and 3% mlv) on the interference of nickel on the determina- tion of arsenic in 10 ml of 0.5 and 5 moll-1 HCl Table 2. Range of interference-free determination of 50 ng of As(V) in 10 ml of 0.5 and 5.0 moll-’ hydrochloric acid using 0.5,l and 3% m/V NaBH, solution.Values given are highest concentrations of interfer- .ent in mg 1-1 that can be tolerated without affecting the peak-height sensitivity for arsenic 0.5 moll-’ HCl 5 moll-’ HC1 3% 1% 0.5% 3% 1% 0.5% Interferent NaBH, NaBH4 NaBH4 NaBH, NaBH4 NaBH, Co(I1) . . 1 1 5 50 100 200 Cu(I1) . . 10 100 100 50 500 500 Ni(I1) . . 0.1 0.2 3 3 5 100 ~ ~~~~ ~ ~ Table 3. Range of interference-free determination of 100 ng of Se(1V) in 10 ml of 0.5 and 5.0 mol 1-1 hydrochloric acid using 0.5, 1 and 3% m/V NaBH, solution. Values given are highest concentrations of interferent in mg 1-1 that can be tolerated without affecting the peak-height sensitivity for selenium 0.5 mol 1-I HCl 5 moll-’ HCl 3% 1% 0.5% 3% 1% 0.5% Interferent NaBH, NaBH4 NaBH, NaBH, NaBH, NaBH4 Co(I1) .. 1 5 10 100 200 400 Ni(I1) . . 0.2 3 10 10 50 200 Cu(I1) . . 0.2 0.2 0.1 10 5 5 small amount of nickel salt is carried into the heated quartz cell as an aerosol where it is deposited. We have observed a similar effect for metals or metal salts placed in the quartz cell.l* As in this earlier experiment, the sensitivity could be restored by cleaning the quartz cell with hydrofluoric acid. The influence of acid and tetrahydroborate(II1) concentra- tion on the interference of all three investigated transition metals in the determination of arsenic and selenium is summarised in Tables 2 and 3, respectively. We have chosen the “range of interference-free determination” to describe the effect because the highest transition metal concentration that has no influence on the analyte signal is usually fairly well defined and reproducible. When the data in Tables 2 and 3 are considered, a very consistent pattern appears for all analyte - interferent combi- nations except for the influence of copper on selenium.This pair will therefore be discussed separately. For all the other combinations, however, a substantial increase in the range of interference-free determination is observed at higher acid and lower tetrahydroborate(II1) concentrations. However, there is more than this qualitative trend in the data in Tables 2 and 3. Table 4 lists the improvement in the range of interference-free determination of arsenic and selenium when the hydrochloric acid concentration is increased from 0.5 to 5 mol 1-1.With the exception of the copper - selenium pair, which will be discussed later, there is an obvious decrease in the improvement factor with increas- Table 4. Improvement in the range of interference-free determination of arsenic and selenium with an increase in hydrochloric acid concentration from 0.5 to 5 mol 1-1 Improvement factor NaBH, concentration, % m/V Analyte Interferent 3 1 0.5 Average As . . . . Co 50 100 40 63 Ni 30 25 33 29 c u 5 5 5 5 Se . . . . Co 100 40 40 60 Ni 50 17 20 29 c u 50 25 50 42 Table 5. Improvement in the range of interference-free determination of arsenic and selenium with a decrease in tetrahydroborate(II1) concentration from 3 to 0.5% mlV Improvement factor HC1 concentration/mol 1- 1 Analyte Interferent 0.5 5 Average As .. . . Co 5 4 4.5 Ni 30 33 32 c u 10 10 10 Se . . . . Co 10 4 7 Ni 50 20 35 c u 0.5 0.5 0.5 ingly positive electrochemical potential (Co*+ < Ni2+ < Cu*+). The average improvement factors for cobalt (63 and 60) and nickel (29 and 29) are essentially identical for the two analyte elements,_ arsenic and selenium. A similar correlation can be found for the improvement in the range of interference-free determination with tetrahydro- borate(II1)’ concentration decreasing from 3 to 0.5% mlV (Table 5). The average improvement factors for cobalt (4.5 and 7) and nickel (32 and 35) are again very close for arsenic and selenium. This indicates that the changes in acid and tetrahydroborate(II1) concentrations act predominantly on the interferent and not on the analyte.Compound formation between the analyte element and the interferent as the primary reaction step and the reason for the interference, as proposed for the interference from some Main Group elements,16 is unlikely. The arsenides and selenides of cobalt and nickel are all insoluble in hydrochloric acid. It would therefore be difficult to explain a two to three orders of magnitude better solubility by a ten-fold increase in acid concentration and a six-fold decrease in tetrahydroborate( 111) concentration. A competitive reaction of the type where some of the sodium tetrahydroborate(II1) is used up by matrix elements, and less is therefore available for reduction of the element of interest to the hydride and a lower signal is obtained, has been proposed by several workers as the mechanism of transition metal interferences.1914,15 The signals and sensitivities obtained for arsenic and selenium at different concentrations of hydrochloric acid could be interpreted in the same way by the more highly concentrated acid consuming more of the tetrahydroborate(II1) so that less is available for hydride formation. It has been shown, however, with radioactively labelled selenium that the efficiency of hydride formation is 95 k 3% even under the conditions where the smallest integrated signal (peak area) is obtained, that is, 5 moll-’ HCl and 3% mlV NaBH4.19 The difference in the signals obtained for arsenic and selenium is therefore not due to insufficient hydride generation but to incomplete atomisation of the hydrides formed. A competition for the sodium tetrahydro-26 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 borate(II1) as an explanation for the interferences becomes totally unlikely when the observations in Table 5 are con- sidered. A lower tetrahydroborate(II1) concentration results in all instances, with the exception of the copper - selenium interference, in a substantial increase in the range of inter- ference-free determination. A very small proportion (<0.1%) of the tetrahydroborate(II1) is required for hydride generation anyway,l6 so that a depletion is unlikely a priori. By the method of exclusion we therefore end up with a mechanism that was already proposed earlier for transition metal interferences, namely preferential reduction of the interferent, and capture and decomposition of the hydride formed at the finely dispersed metal precipitate .43 This mechanism allows all the observed phenomena to be explained.Increasing hydrochloric acid concentration can result in a smaller concentration of bivalent ions of the interfering metal to be present owing to the formation of chloro complexes. Most transition metals show a higher solubility in more highly concentrated acids. Both effects will result in a shift of the metal precipitation to higher concentra- tions. Although a competition for the sodium tetrahydrobor- ate(II1) between the analyte and the interferent was excluded because of the very small amount of reductant required for hydride formation, there may well be a competition between the acid and the interfering metal ions.Both do consume relatively large amounts of tetrahydroborate(II1) so that high acid concentrations (e.g., 5 mol 1-1) can lead to a noticeable depletion, particularly if lower concentrations of the reductant (e.g., 0.5% mlv) are used. This again is in agreement with the experimental observations. The strongest argument for the preferential reduction mechanism, however, is the finding that the factors for improvement in the range of interference-free determination (Tables 4 and 5 ) , although very different for different interfering elements, are essentially identical for arsenic and selenium determinations in the presence of the same interfer- ent. This can only be explained by a direct interaction of the acid and/or tetrahydroborate(II1) concentration with the formation of the interfering metal species, e.g., its precipita- tion. Copper was excluded from the discussion of the results up to now because its behaviour is clearly different in some instances. The influence of copper on the determination of arsenic still fits into the general pattern discussed for the other transition metals. The increase, by a factor of 5, in the range of interference-free determination of arsenic caused by an increase of hydrochloric acid concentration from 0.5 to 5 mol 1-1 is not as impressive as for the other elements. This, however, is within expectations because with increasingly positive electrochemical potential the interfering element will be reduced more easily and the acid concentration cannot influence this process very much. Perhaps the hydrochloric 100 >= 80 'G 60 8 c .- > .- 4- (u $ 40 .- c - g 20 I I-' HCI \?\ \, 5.0 mol I- \$\'\ HCI rnl V NaBH4 0 0.1 1 .o 10 100 1000 Cu concentrationlmg I-' Fig.4. Influence of sodium tetrahydroborate(II1) concentration (0.5, 1 and 3% mlV) on the interference of copper on the determination of selenium in 10 ml of 0.5 and 5 moll-' HC1 acid acts through the formation of chloro complexes, reducing the number of copper ions available for reduction. The influence of copper on the selenium determination is significantly different from all other analyte - interferent combinations discussed here. In comparison with the arsenic determination, an increase in the hydrochloric acid concentra- tion has a substantially more pronounced influence on the range of interference-free determination of selenium in the presence of copper (Fig. 4).A decrease in the tetrahydrobor- ate(II1) concentration, however, has very little influence or even an adverse effect, which means that greater freedom from interferences is observed with higher tetrahydrobor- ate(II1) concentrations. This behaviour cannot be explained by a preferential reduction of the interfering metal ion and an interaction of the reduced species with the hydride formed. In earlier work we found evidence that the copper interference on selenium, at least in 5 mol 1-1 HCl, is probably caused by the copper(I1) ion.4 The most likely mechanism is then a reduction of the selenium to the hydride, which, after its generation, reacts with the free copper ions to form relatively insoluble copper selenide, as proposed by Meyer et al.3 Copper selenide is slightly soluble in hydrochloric acid, so that the strong influence of the higher acid concentration can at least in part be explained by a better solubility of the selenide under these conditions. The formation of chloro complexes in the more highly concentrated acid, which has already been discussed for arsenic, could have an additional effect by reducing the number of free copper ions available for reaction. Such a gas - liquid reaction would depend mainly on the speed of diffusion of the selenium hydride to the gas - liquid interface, the selenium hydride concentration in the gas bubble and the residence time of the gas bubble in the solution.With higher concentrations of acid and tetrahydro- borate(II1) the reaction becomes more violent and the amount of hydrogen formed increases substantially. Under these conditions both the concentration of selenium hydride in the gas bubble and the residence time of the gas bubble in the solution may be reduced, which is in agreement with the observations. Finally, a change in the colour of the solution to reddish is observed on addition of tetrahydroborate(II1). This is pro- bably due to a reduction of Cu(I1) to Cu(I), which means a depletion in the interfering ions. This depletion should be more pronounced and the related interferences less pronoun- ced with more highly concentrated tetrahydroborate(II1) solutions, which was also confirmed by the experiment.Conclusion The interferences of cobalt, copper and nickel on the determination of arsenic, and the interference of cobalt and nickel on selenium, are caused by a preferential reduction of the metal ion to a lower oxidation state or to the metal. The finely dispersed precipitate then captures and decomposes the hydride formed in a secondary reaction. The interference of copper on selenium is caused by a reaction of selenium hydride with free copper(I1) ions to form fairly insoluble copper selenide. One way to extend the range of interference- free determination of arsenic and selenium in the presence of transition metals is to increase the acid concentration, and 5 moll-1 HC1 has been found to be favourable.An additional improvement is obtained when the concentration of the sodium tetrahydroborate(II1) solution is reduced to 0.5% mlV when the interference is caused by a reduced species, e.g., the precipitated metal. Interferent concentrations of 100 mg 1-1 or more can be tolerated for all analyte - interferent combinations investi- gated, except for copper in the determination of selenium, when the proposed acid and tetrahydroborate(II1) concentra-JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 27 tions are used. The only disadvantage appears to be that the sensitivity for arsenic and selenium is lower under these conditions than at lower acid and higher tetrahydroborate(II1) concentrations. This lower sensitivity, however, is not due to less efficient hydride formation but to less efficient atomisa- tion of the hydrides. A higher atomisation efficiency would further increase the usefulness of the proposed measures. References 1. 2. 3. Smith, A. E., Analyst, 1975, 100, 300. Kirkbright, G. F., and Taddia, M., Anal. Chim. Acta, 1978, 100, 145. Meyer, A., Hofer, Ch., Tolg, G., Raptis, S., and Knapp, G., Fresenius Z . Anal. Chem., 1979, 296, 337. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. McDaniel, M., Shendrikar, A. D., Reiszner, K. D., and West, P. W., Anal. Chem., 1976, 48, 2240. Vijan, P. N., and Wood, G. R., Talanta, 1976, 23, 89. Verlinden, M., Baart, J . , and Deelstra, H., Talanta, 1980, 27, 633. Sturman, B. T., Appl. Spectrosc., 1985, 39, 48. Branch, C. H., and Hutchison, D., Analyst, 1985, 110, 163. Thompson, K. C., and Thomerson, D. R., Analyst, 1974, 99, 595. Evans, W. H., Jackson, F. J., and Dellar, D., Analyst, 1979, 104, 16. Dittrich, K . , Vorberg, B., and Wolthers, H., Talanta, 1979,26, 747. Welz, B., and Melcher, M., Anal. Chim. Acta, 1981, 131, 17. Welz, B., and Melcher, M., Analyst, 1983, 108, 213. Krivan, V., Petrick, K., Welz, B., and Melcher, M., Anal. Chem., 1985, 57, 1703. 4. Welz, B., and Melcher, M., Analyst, 1984, 109, 569. NOTE-References 4,5 and 18 are to Parts 2, 3 and 1 of this series, 5. Welz, B., and Melcher, M., Analyst, 1984, 109, 577. 6. Welz, B., and Melcher, M., Spectrochim. Acta, Part B, 1981, 36, 439. Paper 35/21 7. Welz, B., and Melcher, M., Vom Wasser, 1982, 59, 407. Received August 7th, 1985 8. Welz, B., and Melcher, M., Vom Wasser, 1984, 62, 137. Accepted August 23rd, 1985 respectively.

ISSN:0267-9477    

DOI:10.1039/JA9860100023

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 29 Low-resolution Monochromator System for Electrothermal Atomisation Atomic Emission Spectrometry with Probe Atomisation Douglas C. Baxter, lain S. Duncan, David Littlejohn, John Marshall and John M. Ottaway Department of Pure and Applied Chemistry, University of Strathclyde, Cathedral Street, Glasgow G7 7x1, UK Gordon S. Fell Department of Clinical Biochemistry, Royal Infirmary, Glasgow G4 OSF, UK and 0. Yavuz Ataman Middle East Technical University, lnonu Bulvari, Ankara, Turkey A 174-mm focal length Ebert monochromator has been used to construct a low-cost spectrometer for ETA-AES. An oscillating quartz refractor plate was positioned in front of the exit slit inside the monochromator, to achieve rapid automatic background correction by wavelength modulation.The spectrometer was used to measure Cd, Cr, Cu, Mn, Mo, Ni, Pb and V atomic emission from a Perkin-Elmer HGA-72 atomiser modified for probe atomisation and equipped with totally pyrolytic graphite tubes. The 1-19 I-' detection limits obtained for Cd, Cr, Cu, Mn and Pb were a factor of 4-13 poorer than those achieved previously with an echelle spectrometer and similar atomisation system. The variation in detection limits can be explained partly by difference in the optical characteristics of the spectrometer systems. The results obtained suggest that satisfactory analytical performance can be achieved in ETA-AES by use of an inexpensive low-resolution monochromator. Keywords: Atomic emission; electrothermal probe atomisation; low-resolution monochromator; wavelength modulation In electrothermal atomisation atomic emission spectrometry (ETA-AES) , scattered black-body radiation1 and broad-band molecular emission spectra2 are the main sources of back- ground radiation that can degrade the sensitivity and accuracy of the technique in practical analysis.To compensate for these effects, wavelength-modulation procedures have been devel- oped which allow automatic background correction of ETA- AES signals.3.4 Recently, a wavelength-modulated ETA-AES system was described that consisted of an kchelle monochro- mator equipped with a rotating quartz chopper for square- wave wavelength modulation and a Perkin-Elmer HGA-72 electrothermal atomiser .4-6 By application of either platform7 or probe8.9 atomisation, sub-p.p.b. detection limits have been obtained for many elements with this instrument. Although the availability of wavelength-modulation background correc- tion was a distinct advantage in practical analysis,5.6 the excellent detection limits obtained with this system were also due in part to the high resolution and dispersion provided by the kchelle grating. However, not all features of the spec- trometer design are ideally suited for ETA-AES. In par- ticular, the small slit height (0.5 mm) and long focal length reduce the light throughput. In wavelength-modulation ETA-AES, the detection limit is a function of the signal to noise ratio (SNR) and it is possible to predict the SNR, and hence detection limit, from a consideration of the contributing instrumental para- meters.lOJ1 For the detection of an emission line in the presence of a background continuum, the signal to noise ratio at the limit of detection (where the line intensity is much less than the background intensity) can be expressed as equation (1) if the photomultiplier tube is shot noise limited,12 .. Zs2 H W ( A / P ) E( A) Q (A) ~ Z B AA 2eAf (SNR)2= - - where Z, (W sr-1 mm-2) and ZB (W sr-1 mm-2 nm-1) are the analyte line and background intensities, respectively, Hand W are the slit height and width (mm), respectively, A is the effective area of the grating illuminated by the incident radiation (mmz), F is the focal length (mm), Ah is the band pass (nm), E(h) is the transmission efficiency of the grating and optical components, Q(A) is the photomultiplier tube cathode sensitivity (A W-I), e is the electronic charge and Af is the noise band width of the output circuit (Hz).The temperature of the atomiser tube determines the values of I, and ZB. The influence of all the other parameters given in equation (1) depends on the design features of the spec- trometer, As W = DF(Ah), where D is the angular dispersion (rad nm-I), the influence of the spectrometer parameters on the SNR can be expressed as equation (2) where K1 is a constant covering all additional terms in equation (1). In wavelength-modulation ETA-AES, it is normal to define the detection limit as the analyte mass or concentration giving a signal equal to the peak to peak background noise (i.e., SNR = 1). On this basis, Harnlyll used the version of equation (2) developed by Snelleman'o to express the detection limit, Nmin, as equation (3) HDAt -4 Nmin= K2 (7) .. . . * * (3) where the spectral efficiency parameters in equation (2) are replaced by a combined term t and K2 is a constant covering all non-monochromator parameters. For wavelength-modulation ETA-AES, the best detection limit is achieved when the spectrometer monochromator has a large slit height and effective grating area, a short focal length, a large angular dispersion and high spectral efficiency in terms of optical transmission and photomultiplier tube sensitivity. Low detection limits are obtained in wavelength-modulation ETA-AES with the echelle spectrometer system because the area of the grating is large and the angular dispersion is high.However, equation (3) indicates that a short focal length monochromator operated with a high slit height may also give reasonable ETA-AES detection limits. Although the use of a short focal length monochromator requires acceptance of a30 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 Table 1. Instrumentation Electrothermal atomiser Lens . . . . . . . . Monochromator . . . . Detector . . . . . . Wavelength-modulation system . . . . . . Recorder . . . . . . Oscilloscope . . . . . . Perkin-Elmer HGA-72 equipped with totally pyrolytic graphite tubes 53 mm long, 9 mm i.d., supplied by Dr. B. Lersmacher , Philips Research Labs., Aachen, FRG. Pyrolytic graphite probes cut from TPG tubes or prepared directly by vapour deposition (see Fig.2). . . . . F = 170 mm; diameter = 28 mm Ebert configuration, focal length of 174 mm. Grating 1200 lines mm-l blazed at 250 nm; angular dispersion 0.0012 rad nm-1; area 20 x 20 mm. Band pass 0.2,0.5, 1.0,2.0 and 10 nm. Slit height in this work, 3 mm . . EM1 9558 QB; EM1 PM 25B power supply . . Oscillating refractor plate, 15 mm long x 5 mm wide x 3 mm thick. MSE R4160 VSS motor. Labora- tory constructed function generator (sinusoidal waveform). Ortec Brookdeal 9503-SC lock-in amplifier with Ortec Brookdeal 5002 pre-amplifier metric recorder oscilloscope . . . . Servoscribe RE 541.20 potentio- Telequipment DM 64 storage + I 4 , I I I I ox;;;- h [ Current 11 I pre-amplifier recorder amplifier generator Reference Fig. 1. low-resolution monochromator Schematic representation of ETA-AES system based on a poorer angular dispersion than that of the Cchelle spec- trometer, this may not necessarily be a disadvantage as ETA atomic emission spectra have relatively few lines and spectral interferences are rarely encountered.So, as predicted by Harnly,ll it should be possible to use a low- or medium- resolution monochromator for ETA-AES measurements and achieve detection limits comparable to those obtained with an Cchelle spectrometer. To assess this possibility an ETA atomic emission spec- trometer has been constructed that incorporates a 174-mm focal length Ebert monochromator. The monochromator is similar to that used in the Pye Unicam SP-9 atomic absorption spectrometer and was modified for wavelength modulation by introduction of a quartz refractor plate at the exit slit.The instrument was used to measure atomic emission intensities generated by a Perkin-Elmer HGA-72 atomiser modified for graphite probe atomisation.gJ3 The pg 1-1 detection limits obtained were on average an order of magnitude poorer than values reported previously for the kchelle spectrometer - ( a ) a*Trfg mm internal diameter 6 mm Fig. 2. Dimensions of HGA-72 totally pyrolytic graphite tube and probes: (a) TPG tube with probe slot; (b) laboratory made TPG probe; and (c) TPG probe prepared by vapour deposition (see reference 14) HGA-72 combination, but this is explained partly by the optical differences between the two instruments. Besides providing detection limits that are suitable for many applica- tions in trace element analysis, the use of a low-resolution monochromator instead of an Cchelle spectrometer has the additional advantage that it is then relatively inexpensive to construct an instrument system for ETA atomic emission spectrometry.Experimental The low-resolution ETA-AES instrument was constructed from the components given in Table 1. The system has three main sections designed to generate atomic emission intensi- ties, isolate the analyte wavelength of interest and produce background corrected atomic emission signals (Fig. 1). The operational characteristics of each section of the instrument are described below. Probe Atomisation A Perkin-Elmer HGA-72 atomiser was modified for probe atomisation as described by Giri et uZ.gJ3 A 9 mm diameter hole was cut in the front of the HGA-72 atomiser housing to allow the introduction of the graphite probe into the tube through a 6 mm long by 2 mm wide slot cut in the wall at the centre of the tube (Fig.2). Atomiser tubes made of pure pyrolytic graphite were used for all experiments. l4 The probe was held in a small clamp that was attached to an optical bench mount and was moved manually into and out of the furnace by sliding the mount along the optical bench, which was positioned perpendicular to the furnace workhead. Initially, probes were prepared in the laboratory by cutting and filing totally pyrolytic graphite (TPG) cuvettes into the required shape (Fig. 2), but later results were obtained with TPG probes prepared directly by a pyrolysis vapour deposition process.14 The height of the probe was adjusted so that on entering the atomiser, the probe head was close to, but not touching the bottom surface of the graphite tube.The procedure used to dry and atomise solution droplets injected on to the probe head has been described in previous publications.8J3 Volumes of either 20 or 50 p1 were injected on to the head of the graphite probe. The probe was then positioned close to the atomiser tube, which was heated to a temperature of 800-1000 "C to evaporate the droplets in a period of between 40 and 60 s. In this study ETA-AES signals were obtained only for aqueous standard solutions. Hence, when drying was complete, the probe head was moved from the furnace assembly and the furnace tube heated to the required atomisation temperature (2600-2750 "C) without theJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 31 Urine Table 2. Influence of refractor plate frequency on the ETA-AES signal to noise ratio for chromium "0 Refractor Peak height frequency/ for Cr, background noise, Signal to noise Hz arbitrary units" arbitrary units ratio 30 52 3 17.3 65 52 3 17.3 120 90 5 18.0 140 116 6 19.3 165 126 6 21.0 180 124 6 20.6 plate intensity Peak to peak * Signal for 20 pl, 5 pg 1-1 of Cr. Aqueous blank 165 Hz 90 Hz 30 Hz 165 Hz 90 Hz 30 Hz Fig. 3. Effect of refractor plate oscillation frequency on signals obtained at 425.43 nm for 2 0 4 volumes of water and a urine sample known to contain 0.3 pg I-' of Cr inclusion of an intermediate ash stage. The HGA-72 required a period of 8-10 s to establish an approximateIy constant tube temperature.Accordingly, a total atomisation time of 13 s was selected and the probe was introduced directly into the hot atomiser tube after exactly 10 s. If the gas stop mode was required, the inert gas flow was interrupted a few seconds before probe introduction to limit the entrainment of atmos- pheric oxygen. Low-resolution Spectrometer The optical characteristics of the Ebert monochromator used in the construction of the low-resolution ETA-AES instru- ment are given in Table 1. A hole was drilled in the base of the monochromator to allow introduction of a 3 mm thick quartz plate that was attached to an MSE R4160 VSS motor. The plate, 15 mm long by 5 mm wide, was positioned as close as possible to the exit slit.Oscillation of the refractor plate was controlled by means of a laboratory constructed function generator. A sinusoidal waveform was used for all measure- ments, although the function generator could also provide a square wave at low frequencies. The refractor plate could be oscillated at frequencies up to 180 Hz. Previous wavelength- modulation studies15J6 have indicated that in continuum source AAS no significant improvement in signal to noise ratio is achieved at refractor plate frequencies above 40-60 Hz, for a sinusoidal waveform. A similar observation was made for ETA-AES in this study as indicated by the results for Cr given in Table 2. Although an increase in the analyte emission intensity was observed as the plate frequency was increased from 30 to 180 Hz, a corresponding increase in the peak to peak noise caused SNR values to be similar at all frequencies investigated.However, it was observed that the correction of urine matrix background emission signals was more efficient at refractor plate frequencies greater than 90 Hz. This is illustrated in Fig. 3 by the background emission signals obtained following probe atomisation of 20 pl of a urine sample containing an almost undetectable Cr concentration (0.3 pg 1-1). At refractor plate frequencies greater than 90 Hz, the observed signal was less susceptible to spurious peaks and base-line shift, features which are attributed to the urine ~~ ~~ Table 3. ETA-AES detection limits obtained with the low-resolution monochromator Optimum Wavelength/ tempera- atomisation Detection limitt/pg 1-1 Element nm ture*/"C Method l$ Method 2§ Cd .. . . Cr . . . . c u . . . . Mn . . . . Mo . . . . Ni . . . . Pb . . . . v . . . . 326.11 425.43 324.75 403.08 379.83 352.45 405.78 437.92 2700 2700 2600 2700 2750 2700 2600 2700 37 0.1 1 .o 0.2 4.0 3.0 6.0 1.5 8 0.1 1 .0 0.1 5.0 2.0 4.0 2.0 * N2 gas flow stopped during atomisation. t Based on 50-pi injection volumes, except for Cu and Mn when 20 $ Detection limit calculated as the concentration giving a signal 9 Detection limit based on twice the standard deviation of replicated pl were used. equal to the peak to peak background noise. signals with an SNR of ca. 10 ( n = 7). matrix background emission. As the analyte intensity recovered was greater at higher frequencies (as shown in Table 2), and background correction was more efficient (as illustrated in Fig.3), most ETA-AES signals measured with the low-resolution spectrometer system were obtained with a refractor plate frequency of 165 Hz. The monochromator was equipped with an EM1 9558 QB end-window photomultiplier tube, which had S20 (bialkali Sb : Cs) response, covering the wavelength range 180-850 nm. The photomultiplier tube voltage was supplied by an EM1 PM25B variable power supply. Most emission signals were measured with a voltage setting of 0.9 kV. The combined monochromator - photomultiplier tube assembly was enclosed fully in a metal casing to prevent stray light entering the monochromator. The housing was mounted on a 1.5-m optical bench, such that the centre of the monochromator entrance slit was directly above the optical bench rail in line with the HGA-72 atomiser.A 28 mm diameter quartz lens of 170-mm focal length was used to form a 1 : 1 image of the atomiser tube on the entrance slit. Signal Processing Equipment An Ortec Brookdeal 9503-SC lock-in amplifier equipped with a 5002 current pre-amplifier was used to measure the modulated photomultiplier output current. The amplifier was tuned to allow detection at twice the refractor plate frequency (2f mode) in order to discriminate more effectively between the atomic line intensity and the sloping background con- tinuum intensity. The function generator provided the refer- ence waveform. With the present system a Servoscribe RE 541.20 potentiometric strip-chart recorder was used to record the background corrected atomic emission intensities.As the response time of the chart recorder was ca. 0.3 s (full-scale deflection), the lock-in amplifier output time constant was set to 0.3 s, which was identical to the time constant used for signal measurement with the echelle spectrometer ETA-AES system described in previous papers.&9 Atomic Emission Measurements The analytical performance of the low-resolution ETA-AES system was evaluated by measurement of probe atomic emission signals for a range of elements of different volatility and atom excitation energy. The elements selected were Cd, Cr, Cu, Mn, Mo, Ni, Pb and V. Aqueous standard solutions in the appropriate concentration range for each analyte were obtained by dilution of stock solutions prepared from salts of the highest available purity.32 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 ~~ ~~ Table 4. Comparison of probe atomisation ETA-AES detection limits obtained with low-resolution and Cchelle spectrometers Detection limit (DL)lpg l-l* Low-resolution 6chelle Detection limit spectrometer spectrometer ratio Cd . . . . . . 37 3.0 12.3 Cr . . . . . . 0.1 0.013 7.7 Cu . . . . . . 0.4T O . l $ 4.0 Mn . . . . . . 0.08§ 0.008 10.0 Pb . . . . . . 6.0 0.45 13.3 * Calculated as the concentration giving a signal equal to the peak to peak background noise; 5O-pl injection volume; Cchelle values from reference 8. t Estimated for a 50-pl injection volume, based on a calculated detection limit of 1.0 pg 1-1 obtained with a 20-pl volume.$ Obtained with end-entry probe format (reference 9). 0 Estimated for a 5O-pl injection volume, based on a calculated detection limit of 0.2 pg 1-1 obtained with a 20-1.11 volume. Element (DLd (DL) (DL* D L ) Table 5. Comparison of the optical characteristics of the low-resolution Ebert monochromator and Cchelle spectrometer instruments Low-resolution Parameter Ebert 6chelle Focallength(F)lmm . . . . . . 174 750 Angular dispersion (D)/rad nm-1 0.0012* 0.014-0.011t Illuminated grating area (A)lmm2 161 2005 Entrance slit height (H)lmm . . 3 0.5 r+)-*. . . . . . . . . . 17.3 7.3-8.2 * Reference 18. t Angular dispersion values for Cchelle grating in wavelength range 325-425 nm, respectively. An Ircon 1100 optical pyrometer was used to measure the atomiser tube temperature that gave the optimum signal to noise ratio for each element. The pyrometer was focused on the inside wall of the tube as viewed through the injection hole and temperatures were calculated on the assumption that the emissivity of graphite was 1.0 under these conditions.Results and Discussion The optimum atomisation temperatures derived for each element on the basis of SNR calculations are given in Table 3. Although the elements studied differ widely in atom appear- ance temperature and excitation energy, the temperatures at which the optimum SNR value was obtained were very similar. Detection limits were calculated in two ways. In method 1, the detection limit was defined as the concentration of analyte solution that gave an atomic emission signal equal to the peak to peak background noise.In method 2, the detection limit calculation was based on twice the standard deviation of replicate signals obtained for analyte solution concentrations that gave an SNR value of about ten. With the exception of Cd, the detection limit values calculated by both procedures agree within a factor of two. Of the elements included in this study, five (Cd, Cr, Cu, Mn and Pb) were also measured during the evaluation of the probe - Cchelle spectrometer ETA-AES system described in a previous paper.8 As the atomisation conditions and external collection optics used in both investigations were similar, a comparison of the detection limits obtained for the five elements with the low-resolution and Cchelle spectrometers is given in Table 4. The detection limit values were calculated as the concentration of analyte solution that gave a signal equal to the peak to peak background noise on atomisation of 50-p.1 sample volumes.As the original low-resolution spectrometer detection limit values for Cu and Mn were obtained on the basis of 2 0 4 injection volumes, the values included in Table 4 for these elements have been modified to give estimated detection limits that correspond to a volume of 50 pl. With the exception of Cu the detection limits obtained with the echelle spectrometer system are about one order of magnitude better than those obtained with the low-resolution instrument. It should be noted that the figure quoted as the kchelle - probe detection limit for Cu was obtained with an end-entry probe system9 rather than the side-of-tube format used for all other measurements.Hence, the Cu detection limits given in Table 4 may not be directly comparable. As illustrated by the expression given in equation (3), the detection limit obtained in wavelength-modulation ETA-AES depends upon the optical characteristics of the spectrometer. The important parameters are the focal length (F), angular dispersion (D), the effective grating area ( A ) , the entrance slit height (H) and the spectral efficiency of the instrument (t). The values of F, D, A and H for the low-resolution and echelle spectrometer instruments are given in Table 5. The angular dispersion values for the Cchelle grating in the wavelength range 325425 nm were calculated from figures quoted by Harnly.11 The effective area of illumination of the Cchelle grating was determined as the product of the area of the grating and cos 8, where 8 is assumed to be the blaze angle, which in this instance is 63".Hence, the effective area of illumination was computed as being 96 X 46 X 0.454 mm2. As mentioned previously, the same optical system was used to focus radiation from the electrothermal atomiser on to the entrance slit of the low-resolution and echelle monochroma- tors. With the 28 mm diameter, 170-mm focal length lens placed 340 mm from the entrance slit, the grating of the low-resolution monochromator was not completely filled and the effective area of illumination was calculated as 161 mm2. With the Cchelle spectrometer the same optical parameters caused the grating to be slightly overfilled, which resulted in a slight reduction in the transmission efficiency. It was not possible to obtain accurate values of t , the combined spectral efficiency factor.Consequently, in calculations based on the expression given in equation (3), the value of t at each analyte wavelength has been assumed to be similar for both spec- trometer systems. On the basis of information published recently on the spectral efficiency of a Spectraspan 111 Cchelle grating spectrometer,17 the spectral efficiency values for the 174-mm Ebert monochromator18 and the response charac- teristics of the photomultiplier tubes used for the ETA-AES measurements, it is probable that this assumption is not unreasonable. In addition, a factor of 2 for the difference in the t values at any wavelength will only introduce an error of v 2 in the detection limit comparison. From the information given in Table 5 , it is expected that under the experimental conditions described, the Cchelle spectrometer should have a detection limit advantage over the low-resolution instrument by a factor of 2.1-2.4. However, an additional factor that influences the detection limit compari- son has still to be considered.The wavelength modulation system employed in the Cchelle spectrometer study8 was based on the rotation of a sectored wheel that provided a square- wave modulation waveform.4-7 It is known19JO that square- wave modulation has an inherent signal to noise advantage over sine-wave modulation by a factor of ca. 1.8.Considering all factors, therefore, the detection limit advantage of the Cchelle spectrometer system should range from a factor of 4.3 for Cd and Cu (i.e., at ca. 325 nm) to a factor of 3.8 for Cr (425 nm) . With the exception of Cu, the actual detection limits obtained with the low-resolution spectrometer system were 7.7-13.3 times poorer than those obtained with the Cchelle spectrometer (Table 4), a factor of 2-3 times poorer than predicted. This may be due to a combination of factors, suchJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 33 as a difference in the spectral response characteristics of the spectrometers, slight anomalies in the HGA-72 atomiser modifications or variations in the atomisation conditions. Although each individual factor may only introduce a small error, the combined effect could easily account for the discrepancy between the experimental and theoretical detec- tion limit comparisons.Clearly it will be necessary in future work to optimise the collection optics to ensure that the grating of the low- resolution monochromator is completely filled. At the time of this study no alternative lenses were available. Although the optical system was not ideal, the general performance of the low-resolution spectrometer was satisfactory, and the results obtained confirm in part, the theoretical predictions made by Harnly .I1 In practical analysis, the low-resolution monochro- mator system has been applied successfully to the direct determination of Cr in urine, and this work is described in a subsequent paper .21 Replacing the echelle spectrometer with an inexpensive low-resolution monochromator has allowed the construction of a comparatively low-cost ETA-AES system that is suf- ficiently sensitive for many applications in, for example, environmental, biological and food analysis, and covering a wide range of elements.7 The instrumental concept may therefore be of significant interest in locations where financial resources represent a serious problem or limitation and could be of value in many developing countries.On the basis of the findings of this study, other low-resolution monochromator systems are currently under construction in our laboratory. This second generation of instruments will incorporate the microcomputer control and data acquisition facilities devel- oped recently in our laboratory for ETA-AES and continuum source AAS .22 The authors thank Pye Unicam Ltd., Cambridge, UK, for the gift of the low-resolution monochromator and Philips Research Laboratories, Aachen, FRG, for the provision of totally pyrolytic graphite materials. Financial support from the Pye Foundation (D.L.), the SERC (J. M.) and the Scottish Home and Health Department (D. C. B.) is gratefully acknowledged. Participation of 0. Y. A. in the project was made possible through the provision of NATO Grant No. 123-82. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20 * 21. 22. References Littlejohn, D., and Ottaway, J. M., Analyst, 1977, 102, 553. Hutton, R. C., Ottaway, J. M., Epstein, M. S., and Rains, T.C., Analyst, 1977, 102, 658. Epstein, M. S., Rains, T. C., and O’Haver, T. C., Appl. Spectrosc., 1976, 30, 324. Bezur, L., Marshall, J., and Ottaway, J. M., Spectrochim. Acta, Part B , 1984,39, 787. Ottaway, J. M., Bezur, L., and Marshall, J., Analyst, 1980, 105, 1130. Marshall, J., and Ottaway, J. M., Tulanta, 1983, 30, 571. Bezur, L., Marshall, J., Ottaway, J. M., and Fakhrul-Aldeen, R., Analyst, 1983, 108, 553. Giri, S. K., Littlejohn, D., and Ottaway, J. M., Analyst, 1982, 107, 1095. Marshall, J., Giri, S. K., Littlejohn, D., and Ottaway, J. M., Anal. Chim. Acta, 1983, 147, 173. Snelleman, W., Spectrochim. Acta, Part B , 1968, 23,403. Harnly, J. M., Anal. Chem., 1984, 56, 895. Sharpe, M. R., Pye Unicam Ltd., Cambridge, UK, personal communication, 1980. Giri, S. K., Shields, C. K., Littlejohn, D., and Ottaway, J. M., Analyst, 1983, 108, 244. Van den Brekel, C. H. J., and Lersmacher, B., in Bloem, J., Verspui, G., and Wolff, L. R., Editors, “Proceedings of the 4th European Conference on Chemical Vapour Deposition ,” Philips, Eindhoven, 1983, p. 321. Harnly, J. M., and O’Haver, T. C., Anal. Chem., 1981, 53, 1291. Harnly, J. M., Anal. Chem., 1982, 54, 876. Zander, A. T., Miller, M. H., Hendrick, M. S., and Eastwood, D., Appl. Spectrosc., 1985,39, 1. Pye Unicam Ltd., York Street, Cambridge, UK, personal communication, 1984. O’Haver, T. C., Epstein, M. S., and Zander, A. T., Anal. Chem., 1977,49, 458. Koirtyohann, S. R., Glass, E. D., Yates, D. A., Hinderberger, E. J., and Lichte, F. E., Anal. Chem., 1977, 49, 1121. Baxter, D. C., Littlejohn, D., Ottaway, J. M., and Fell, G. S., J . Anal. At. Spectrom., 1986, 1, 35. Marshall, J., Carroll, J., Littlejohn, D., Ottaway, J. M., O’Haver, T. C., and Harnly, J. M., Anal. Proc., 1985, 22, 67. Paper J5lI Received April lst, 1985 Accepted June 3rd, I985

ISSN:0267-9477    

DOI:10.1039/JA9860100029

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 35 Determination of Chromium in Urine by Probe Electrothermal Atomisation Atomic Emission Spectrometry Using a Low-resolution Monochromator Douglas C. Baxter, David Littlejohn and John M. Ottaway Department of Pure and Applied Chemistry, University of Strathclyde, Cathedral Street, Glasgow G7 7XL, UK Gordon S. Fell and David J. Halls Department of Clinical Biochemistry, Royal Infirmary, Glasgow G4 OSF, UK A low-resolution monochromator incorporating wavelength modulation background correction has been used to determine chromium in urine by ETA-AES. Chromium atomic emission was generated in a Perkin-Elmer HGA-72 graphite furnace equipped with a totally pyrolytic graphite tube and modified for probe atomisation. No significant spectral or chemical interferences from the urine matrix were encountered and samples were analysed against aqueous standards either directly or after 1 + 1 dilution with water.The accuracy of the procedure was confirmed by comparative analysis with two different ETA-AAS methods. The ETA-AES technique was used to measure chromium in urine down to a level of 0.3-0.5 ng ml-I. It is predicted that a lower detection limit can be achieved by optimisation of the spectrometer optical system. Keywords: Atomic emission spectrometry; atomic absorption spectrometry; probe electrothermal atomisation; urine chromium; low-resolution monochromator Several elements are known to be biologically significant owing to their presence, in trace amounts, in essential enzymes and enzyme co-factors.The metabolic function of chromium arises due to its presence as a constituent of glucose tolerance factor (GTF).1 The structure of GTF is at present uncertain, but it is postulated to be a complex containing trivalent chromium, nicotinic acid and various amino acids, which acts as an insulin co-factor specifically participating in the binding of insulin to its receptors.' Hence, the major interest in the biological role of chromium is in its relationship to diabetes. Historically, the measurement of chromium in biological materials has proven extremely difficult, largely due to the limitations of available instrumentation and the very low levels of chromium present in biological materials.* A substantial amount of analytical data has been accumulated since the biological significance of chromium was first recognised, but confusion remains as to the actual concentra- tion present in healthy individuals. Widely divergent values have been reported for urine and serum, varying from 185 ng ml-l as measured by Monacelli et al.in 19563 to 0.14 ng ml-1 by Kayne et al. in 19784 for serum chromium concentrations. A critical appraisal of the literature was published in 1980 by Versieck and Cornelis,s which clearly demonstrated the downward trend in reported chromium levels up to the present day. This trend reflects the improve- ments in instrumental techniques and analytical methodology achieved in recent years, and also agreater understanding of the problems involved in any biological analysis.It is probable that many of the early values published for chromium were influenced by the effects of contamination particularly from syringe needles and storage containers. Such is the ubiquity of chromium in the environment that stringent precautions must be taken to avoid extraneous additions. Electrothermal atomisation atomic absorption spec- trometry (ETA-AAS) is currently the most widely used technique for the determination of trace metals in clinical analysis. However, ETA-AAS has been criticised on two counts with respect to the determination of chromium in biological materials. Guthrie et a1.6 concluded that the commonly used deuterium arc background correction pro- cedure was inadequate for the determination of chromium in urine. This prompted Kayne et al.4 to modify a Perkin-Elmer 603 atomic absorption spectrometer by adding a high-intensity tungsten - halogen lamp for background correction.Improved detection limits for elements with wavelengths in the near UV and visible wavelength regions were reported when this lamp was used for background correction, and these workers reported the measurement of chromium in human serum and urine at concentrations of less than 0.1 ng ml-1. However, recent work738 has demonstrated that the deuterium arc can in fact provide adequate background correction at the 357.9-nm chromium wavelength provided the atomisation temperature is carefully selected. An additional problem associated with the determination of chromium in biological samples by ETA-AAS is that the detection limit for chromium is similar to the levels present in, for example, normal serum and urine samples.It is now recognised that normal urinary chromium excretion is in the range 0.1-1.0 ng ml-1, whilst serum levels are typically 0.1-0.2 ng ml-1.9 It is evident that an alternative, more sensitive method than ETA-AAS is required for the determination of biological chromium levels. The availability of an additional analytical technique of the required sensitivity would, at the very least, provide a method that could be used to check the accuracy of concentrations determined by It is now well established that most commercial electro- thermal atornisers can be used as an excitation source for analysis by atomic emission spectrometry, and ETA-AES has proved to be an extremely sensitive technique for the determination of many elements, including chromium.When a Perkin-Elmer HGA-72 electrothermal atomiser modified for either platformlo or probell atomisation was used in conjunction with an echelle spectrometer equipped with wavelength-modulation background correction, detection limits for chromium in aqueous solution were less than 0.05 ng ml-1.10311 It has been suggested12 that detection limits comparable to those obtained with the kchelle spectrometer can be achieved in ETA-AES by use of a low- or medium- resolution monochromator. Replacing the Cchelle spec- trometer with a comparatively low-cost monochromator has the advantage that it would be relatively inexpensive to construct an instrument system for ETA-AES that could be used for the determination of chromium and other metals in ETA-AAS.36 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 Table 1. Electrothermal atomiser programmes for the HGA-72 and HG A-500 Atomic emission (425.43 nm), HGA-72 Atomiser Stage temperature/'C Time/s Conditions* Dry . . . . . . 800-1000 60 Probe outside tube A s h . . . , . . 1250 50 Probe inside tube Atomise . . , . 2700 15 Probe in after 13 s Atomic absorption (357.89 nm), HGA-500 Atomiser Ramp Hold Conditions/ Stage temperature/'C time/s time/s ml min-* Ar Dry? . . . . 120 7 20 300 A s h . . . . . . 1000 7 20 300 Atomise . . . . 2400 0 5 30 Clean . . . . 2700 1 2 300 * N2 gas flow at all stages; flow-rate in modified atomiser head probably less than normal due to hole in casing; 50 p1 for aqueous studies, 20 pl for urine analysis.(internal flow) t 2 0 4 volume. clinical analysis. In a previous paper13 the construction of an ETA-AES instrument based on a 174-mm focal length Ebert monochromator was described. An oscillating quartz refractor plate was positioned in front of the exit slit inside the monochromator, to achieve rapid automatic background correction by wavelength modulation. The spectrometer was used to measure the atomic emission of Cd, Cr, Cu, Mn, Mo, Ni, Pb and V produced by probe atomisation in a modified Perkin-Elmer HGA-72 atomiser. The system has been applied to the determination of chromium in urine. The urine matrix caused no significant chemical or spectral interferences on the measurement of chromium, and levels down to 0.3-0.5 ng ml-1 in urine could be determined with reasonable accuracy and precision, as comparative studies with ETA- AAS indicate.Experimental Electrothermal Atomisation Atomic Emission Spectrometry The low-resolution ETA-AES instrument used for the analy- sis of urine samples has been described in detail in a previous paper.13 A Perkin-Elmer HGA-72 atomiser modified for probe atomisationl3>14 was used to generate chromium atomic emission. The probes used for most of this work were made of pure pyrolytic graphite and had probe-head dimensions of either 5 x 5 or 4 X 4 mm, both of 300 pm thickness.15 The probe was moved manually into and out of the totally pyrolytic graphite tube via a slot cut in the tube wall at the centre of the atomiser. Aqueous and urine sample volumes were injected on to the probe head with a micropipette, and the probe was then moved close to the atomiser tube heated to between 800 and 1000 "C.With aqueous solutions, a 50-yl sample volume could be dried in approximately 60 s. A 2 0 4 aliquot was generally selected for urine to avoid sample spreading problems and to maintain a reasonably short drying time. Once the drying phase was complete, the probe head was moved into the atomiser to ash the sample. For chromium, the maximum ash temperature that could be used without loss of analyte was 1250 "C. After an ashing period of 50 s, the probe was removed from the atomiser tube, which was then heated to the required atomisation temperature. Once the atomiser had achieved an approximately constant temperature (8-10 s) the probe head was moved rapidly into the hot tube.For chromium, atomisation temperatures in the range 2650-2840 "C gave a similar atomic emission intensity at the chromium 425.43-nm wavelength. However, the optimum signal to noise ratio was obtained at 2700 "C and this temperature was selected for all urine analyses. Gas flow was maintained during the atomisation stage to assist removal of residual urine matrix vapours. With probe atomisation, the use of gas flow only results in a 50% reduction in the Cr atomic emission intensity. A summary of the atomisation programme used for chromium determinations by ETA-AES is given in Table 1. The Ebert monochromator was modified for wavelength modulation as described previously. 13 The signal to back- ground ratio for chromium and the correction of urine background signals were optimum at a refractor plate oscilla- tion frequency of 165 Hz.Spectral measurements were made at the Cr 425.43-nm wavelength with a spectrometer band pass of 0.2 nm and a reduced slit height of 3 mm. An Ortec Brookdeal 9503-SC lock-in amplifier equipped with a 5002 current pre-amplifier was used to measure the modulated photomultiplier output signal. The amplifier was tuned to allow detection at twice the refractor plate frequency (2f mode) in order to discriminate more effectively between the atomic line intensity and the sloping background continuum intensity. The function generator used to control the motion of the refractor plate also provided the reference waveform for the lock-in amplifier. The lock-in amplifier output time constant was set to 0.3 s, which was similar to the full-scale response time of the Servoscribe RE 541.20 strip-chart recorder used to record the background corrected atomic emission intensities. Electrothermal Atomisation Atomic Absorption Spectrometry A Perkin-Elmer 2280 atomic absorption spectrometer and an HGA-500 electrothermal atomiser were used for atomic absorption analysis of urine samples.The method applied was that developed by Halls and Fell.8 Aqueous and urine samples (20 p1) were dried, ashed and atomised using a pyrolytically coated tube and the programme given in Table 1. During atomisation, an internal gas flow of 30 ml min-1 was maintained. Atomic absorption signals were measured at the 357.89-nm chromium wavelength with a spectrometer band pass of 0.7 nm.The chromium hollow-cathode lamp current was reduced to 11-12 mA to match the intensity of the D2-arc background correction lamp intensity at the analyte wavelength. A spectrometer scale expansion factor of ~7 was required when analysing samples that contained a low chromium concentration. Inter-laboratory Comparative Analysis of Urine Samples Urine samples containing elevated chromium levels were analysed by the above methods and then sent to the US Department of Agriculture, Beltsville Human Nutrition Research Center, MD, USA for analysis by Veillon and colleagues. The method used at the USDA involved an ETA-AAS procedure slightly different to that employed at the Glasgow Royal Infirmary. Details of the method, which involved calibration by standard additions, have been publi- shed previously.16 Sample Collection and Preparation All samples analysed in this work were collected at Glasgow Royal Infirmary.Some were from diabetics on a chromium supplementation trial and others from patients on total parenteral nutrition (TPN). These patients, especially the group on TPN, tend to have urine chromium concentrations higher than normal. Urine was voided into a plastic bottle and transferred into an acid-washed 2-1 container at various times during a 24-h collection period. The sample was then acidified with 5 ml of glacial acetic acid. Aliquots were then transferred into 25-ml Sterilin plastic containers. Alternatively, aliquots of the 24-h urine samples were transferred immediately into 25-ml Sterilin containers prior to acidification and 200 pl of concentrated sulphuric acid added.The Sterilin containersJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 37 used in this work had previously been shown to be free of chromium contamination. Urine samples were analysed either directly or after 1 + 1 dilution with distilled water for aqueous calibration measure- ments, or with chromium aqueous solutions for standard- additions analysis. Chromium standard solutions were pre- pared by serial dilution of a 1000 pg ml-1 chromium solution supplied by BDH Chemicals Ltd. and prepared from Cr(N03)3.9H20 dissolved in dilute HN03. In the preparation of aqueous chromium standards, no further addition of acid was made.Results and Discussion Comparison of the Precision and Sensitivity of ETA-AAS and ETA-AES Methods for the Determination of Chromium The determination of chromium in biological samples at Glasgow Royal Infirmary is currently performed by the ETA-AAS method developed by Halls and Fell.8 Although the optical characteristics of the low-resolution ETA-AES instrument were not completely optimised,l3 it was decided to assess the system for the determination of urine chromium by a comparison with the AAS procedure. Ten aqueous chro- mium standards in the concentration range 0.1-300 ng ml-1 were prepared and each solution analysed seven times by both ETA-AAS and ETA-AES. This allowed a comparison of the instrument precision for both systems, without the influence of matrix related factors.A summary of the results obtained under the optimised experimental conditions defined previ- ously is given in Table 2. The percentage RSD values quoted illustrate the signal precision obtained with the absorption and emission instruments. It is evident that the ETA-AAS technique provides a method of greater precision below 3 ng ml-1 for the measurement of chromium aqueous solutions. At the 0.1 ng ml-l chromium level, the ETA-AES system was unable to detect any discernible signal under gas flow conditions, and as the percentage RSD measured at 0.2 ng ml-l was high, this concentration is obviously very close to the detection limit. The sensitivity performance of the emission system would undoubtedly be improved through more prudent selection of the optical parameters.13 In addition, the precision of the emission measurements suffered in comparison with the absorption signals owing to the lack of an autosampling facility and autoprobe device.The AS-1 autosampler used in conjunction with the HGA-500 atomiser clearly contributed to the reasonable precision of the atomic absorption measurements. Although the low-resolution ETA-AE instrument in its present form was apparently inferior to the Perkin-Elmer 2280/HGA-500 atomic absorption combination for the analy- sis of aqueous solutions, it is known that the presence of a complex matrix, such as urine, can degrade the characteristics of ETA-AAS signals. To assess the influence of urine on the measurement of chromium by ETA-AES, a series of 1 + 1 diluted urine samples were analysed and the precision of the signals compared with that of aqueous solutions covering the same chromium concentration range.The urine solutions were prepared by the addition of aqueous chromium stan- dards to a urine sample that had a chromium concentration of less than 0.2 ng ml-1. Each urine and aqueous solution was analysed seven times and percentage RSD values obtained Table 2. Comparison of the precision of ETA-AAS and ETA-AES signals obtained for aqueous chromium solutions Table 4. ETA-AES analysis of 1 + 1 diluted urine samples by standard additions and aqueous calibration procedures in comparison with ETA- A AS ETA-AES Chromium ETA- AAS Relative peak concentration/ height ngml-l Absorbance* RSD, YO intensityt RSD, Yo 0.0 0.002 0.1 0.003 18.9 0.2 0.006 8.6 1 .o 93.9 0.5 0.007 8.4 2.7 20.4 1 .o 0.018 6.0 7.7 14.7 3.0 0.049 3.9 27.1 4.9 10.0 0.146 4.2 54.5 3.8 30.0 0.338 2.7 140.2 1.7 100.0 0.580 2.2 317.6 2.4 300.0 1.016 4.1 498.5 3.3 * Measurements made using x 7 scale expansion; results quoted are the mean of 7 signals divided by the expansion factor; 20-1.11 volume dispensed by AS-1 autosampler.t Measurements based on 2 0 4 volumes dispensed manually; mean of 7 signals. Table 3. Influence of urine matrix on the magnitude and precision of chromium atomic emission intensity Aqueous solutions Diluted urine (1 + 1) Chromium Relative peak Relative peak concentration/ height height ng ml-1 intensity* RSD, Yo intensity* RSD, YO 1.0 1 .o 19.7 1.2 23.2 3.0 3.4 6.2 3.5 7.4 10.0 10.9 3.5 10.2 3.1 30.0 31.5 2.5 31.4 4.6 100.0 68.5 2.3 69.2 5.3 300.0 125.9 1.1 127.2 1.3 * 2 0 4 volumes dispensed manually; mean of 7 signals.Chromium concentratiodng ml-I ETA-AES ETA-AAS Relative Standard Aqueous aqueous Sample slope* additions calibration calibration 1 1.06 0.25 - 0.22 2 1.03 0.67 0.56 0.58 3 1.02 0.62 0.64 0.54 4 0.94 1.05 1.06 - 5 1.01 1 S O 1.30 - 6 1.07 1.40 1.37 - 7 1.05 2.45 2.03 - 8 0.97 0.40 0.43 - * Relative slope calculated as the ratio of the standard additions calibration slope to the aqueous calibration slope. Table 5. Direct analysis of undiluted urine samples by ETA-AES in comparison with an ETA-AAS method; results are chromium concentrations in ng ml-1 First analysis period Second analysis period Sample ETA-AES* ETA-AAST ETA-AES* ETA-AASt 1 0.6 2 0.2 0.6 1.5 k 0.2 1.4 2 1.4 2 0.3 0.6 1.6 k 0.2 1.3$ 3 0.3 k 0.2 0.4 1.6 k 0.2 1.1 4 0.5 k 0.2 0.4 4.2 k 0.3 3.6 0.4 0.6 2 0.2 0.5 5 * Samples analysed without dilution; direct aqueous calibration.t Samples analysed after 1 + 1 dilution with distilled water; direct aqueous calibration; RSD (Yo) values for AAS measurements in the range 5-20Y0. 0.7 -t- 0.2 $ Sample analysed after 1 + 3 dilution with distilled water.38 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 from the intensity measurements. The relative peak height intensities and percentage RSD values obtained are given in Table 3. The urine matrix had no significant influence on the magnitude of chromium atomic emission signals and caused only a slight degradation in signal precision.Analysis of Urine Samples by ETA-AES Using the Method of Standard Additions Most of the ETA-AAS methods developed for urine chro- mium determination have employed standard addition calib- ration procedures to overcome matrix chemical interference effects.*JG18 Similarly, continuum-source AAS9 and ETA- AES10J9 methods based on platform atomisation have been reported that also required standard-additions calibration. The results presented in Table 3 suggested that with probe atomisation, the effect of urine matrix on chromium atomic emission measurements was minimum. To investigate this further, a series of urine samples, that varied considerably in colour, salt content and chromium concentration, were analysed by standard additions and direct aqueous calibration procedures following 1 + 1 dilution with aqueous chromium standards and distilled water, respectively.Three of the samples were also analysed by the ETA-AAS method developed by Halls and Fe11.8 The results obtained are summarised in Table 4. For each set of standard-additions solutions, the slope of the calibration graph was compared with that obtained for aqueous calibration. As shown in Table 4, none of the standard addition calibration slopes varied significantly from that of the aqueous calibration graph. For most of the samples in the concentration range 0.25-2.45 ng ml-l of Cr, the ETA-AES values obtained by standard additions and aqueous calibration were in close agreement. In addition, the concentrations determined by the ETA-AAS method for three samples were similar to the values deter- mined using the low-resolution emission instrument.The results presented in Table 4 confirm that no severe chemical matrix interference effects are encountered in the determina- tion of chromium in urine by ETA-AES with probe atomisa- tion. This is contrary to the findings of many ETA-AAS studies. Hinderberger et al.18 reported slopes of 0.89 for tube-wall atomisation and 1.24 for platform atomisation with a matrix modifier for urine chromium determination. Two factors probably contributed to the lack of matrix effects in the current emission study. The ash temperature used was 1250 "C, which greatly reduced the sample matrix present during atomisation and hence minimised the concentration of inter- fering species. Also, the use of probe atomisation assisted the prevention of chemical vapour phase interference effects.The fast heating rate of the probe caused rapid volatilisation of the sample, which reduced the influence of the urine matrix on the rate of atomisation. In addition, the high temperature of the furnace vapour at the time of volatilisation assisted the Table 6. Inter-laboratory comparative analysis of urine samples; results are chromium concentrations in ng ml-l ETA-AAS* Sample ETA-AES GRI1 GRI 2 USDA 1 2.5 k 0.3 2.6 2.4 2.2 2 2.5 k 0.3 2.6 2.4 2.4 3 2.8 _+ 0.2 3.0 2.8 2.6 4 3.5 k 0.3 3.0 2.8 3.0 5 3.7 k 0.3 3.6 3.2 2.4 6 5.3 -t 0.3 6.0 5.0 2.6 * Samples analysed by ETA-AAS on two separate occasions at Glasgow Royal Infirmary (GRI) and once at the US Department of Agriculture, Beltsville, MD, USA (USDA); RSD (Yo) values for AAS measurements in the range 5-10%.dissociation of molecules and reduced the effect of chloride salts on chromium atom production. The method used to obtain the ETA-AAS results given in Table 4 was also based on peak-height measurements and direct aqueous calibration. Halls and Fell8 have reported that by judicious selection of atomiser conditions, the peak-height mode of measurement can be used without interference from the urine matrix. Direct Determination of Chromium in Undiluted Urine by A set of urine samples from patients at Glasgow Royal Infirmary were analysed by ETA-AES with the low-resolution spectrometer system, without dilution and using aqueous calibration standards. The samples were also analysed by the ETA-AAS procedure described previously,g after 1 + 1 dilution with distilled water.The results obtained are given in Table 5, columns 2 and 3. Reasonable agreement was obtained between the two techniques for all samples except number 2. The atomic absorption analyses were performed first , and it was suspected that contamination of sample 2 had occurred before the atomic emission analysis. Such is the ubiquity of chromium in the environment that even the most stringent precautions sometimes cannot avoid contamination of samples. The problem is illustrated by the second set of analytical results given in Table 5, columns 4 and 5. The urine samples were re-analysed by both techniques , approximately 1 week after the first set of determinations. Again, reasonable agreement was obtained between the atomic emission and atomic absorption measurements, but the results in Table 5 indicate clearly that samples 1, 3 and 4 were contaminated after the initial atomic emission analysis and that sample 2 had been contaminated after the first atomic absorption analysis.This comparative study illustrates succinctly the contamina- tion related problems that can occur in chromium determina- tion at the ng ml-1 level. ETA-AES Inter-laboratory Analysis Six pooled urine samples containing elevated chromium concentrations were analysed by ETA-AAS at Glasgow Royal Infirmary8 and then by ETA-AES using the low-resolution monochromator system and probe atomisation. Owing to the contamination problems described above, the samples were then reanalysed at the Glasgow Royal Infirmary by ETA- AAS (Table 6, column 4) prior to transportation to the US Department of Agriculture , Human Nutrition Research Center, Beltsville, MD for analysis by an alternative ETA- AAS procedure.16 The USDA method involved a standard- additions procedure, samples were diluted 1 + 1 with distilled water and analysed against aqueous standards at Glasgow Royal Infirmary, and direct analysis of the urine samples without dilution was employed at the University of Strathclyde for the emission method.The chromium concen- trations determined by the various procedures are given in Table 6. With the exception of sample 6, excellent agreement was obtained between the atomic absorption procedures and the ETA-AES method. Conclusions The low-resolution monochromator ETA-AES system des- cribed previously13 is capable of determining urine chromium concentrations down to 0.3-0.5 ng ml-1 with acceptable accuracy.The application of probe atomisation minimised chemical interferences, and the wavelength modulation tech- nique provided adequate correction of the urine background spectrum. It was possible therefore, to analyse samples directly against aqueous standards. Comparative analysis with two ETA-AAS procedures, applied in different laboratories, revealed that the AES method gave acceptable accuracy in theJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 39 concentration range studied (0.5-2.5 ng ml-1 of Cr). The present investigation also confirmed that with careful selec- tion of temperature conditions, the direct analysis of diluted urine samples can be achieved by ETA-AAS with wall atomisation.The current ETA-AES instrument does not provide better sensitivity than established ETA-AAS procedures, but it could be used routinely to provide a quick and reliable check on chromium concentrations determined by ETA-AAS. To improve the AES detection limit, it will be necessary to optimise the optical characteristics of the instrument as discussed elsewhere,13 and this is presently under investiga- tion. It should then be possible to measure chromium concentrations at the 0.1 ng ml-1 level or below with adequate precision and accuracy. The authors wish to thank Pye Unicam Ltd, Cambridge, UK for the gift of the low-resolution monochromator and Philips Research Laboratories, Aachen, FRG for the provision of total pyrolytic graphite materials.Financial support from the Pye Foundation (for D. L.) and the Scottish Home and Health Department (for D. C. B.) is gratefully acknowledged. The cooperation of Mrs P. M. Dunbar, Department of Clinical Biochemistry, Glasgow Royal Infirmary and C. Veillon and colleagues at the US Department of Agriculture, Human Nutrition Research Center, Beltsville, MD, USA, is also greatly appreciated. 1. 2. References Barret, J., O’Brien, P. O., and Pedora de Jesus, J., Poly- hedron, 1985,4, 1. Anderson, R. A., Polansky, M. M., Bryen, N. A., Roginski, E. E., Patterson, K. Y., and Reamer, D. C., Diabetes, 1982, 31, 212. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Monacelli, R., Tanaka, H., and Yoe, J. M., Clin. Chim. Acta, 1956, 1, 577. Kayne, F. J., Komar, G., Laboda, H., and Vanderlinde, R. E., Clin. Chem., 1978, 24, 2151. Versieck, J., and Cornelis, R., Anal. Chim. Acta, 1980, 116, 217. Guthrie, B. E . , Wolf, W. R., and Veillon, C., Anal. Chem., 1978,50, 1900. Routh, M. W., Anal. Chem., 1980, 52, 182. Halls, D. J., and Fell, G. S . , in Bratter, P., and Schramel, P., Editors, “Trace Element Analytical Chemistry in Medicine and Biology, Volume 2,” De Gruyter, Berlin, 1983, p. 667. Versieck, J., Hoste, J., Barbier, F., Steyaert, H., De Rudder, J., and Michels, H., Clin. Chem., 1978, 24, 303. Bezur, L., Marshall, J., Ottaway, J. M., and Fakhrul-Aldeen, R., Analyst, 1983, 108, 553. Giri, S. K., Littlejohn, D., and Ottaway, J . M., Analyst, 1982, 107, 1095. Harnly, J. M., Anal. Chem., 1984, 56, 895. Baxter, D. C., Duncan, I. S., Littlejohn, D., Marshall, J . , Ottaway, J . M., Fell, G. S . , and Ataman, 0. Y., 1. Anal. At. Spectrom., 1986, 1, 29. Giri, S. K., Shields, C. K., Littlejohn, D., and Ottaway, J. M., Analyst, 1983, 108, 244. Van der Brekel, C. H. J., and Lersmacher, B . , in Bloem, J., Verspai, G . , and Wolff, L. R., Editors, “Proceedings of the 4th European Conference on Chemical Vapour Deposition,” Philips, Eindhoven, 1983, p. 321. Veillon, C., Patterson, K. Y., and Bryden, N. A., Anal. Chim. Acta, 1982, 136,233. Veillon, C., Patterson, K. Y., and Bryden, N. A., Clin. Chem., 1982, 28,2309. Hinderberger, E. J., Kaiser, M. L., and Koirtyohann, S . R., At. Spectrosc., 1981, 2, 1. Harnly, J. M., Patterson, K. Y., Veillon, C., Wolf, W. R., Marshall, J., Littlejohn, D., Ottaway, J. M., Miller-Ihli, N. J., and O’Haver, T. C., Anal. Chem., 1983, 55, 1417. Paper J512 Received May 22nd, 1985 Accepted September 11 th, 1985

ISSN:0267-9477    

DOI:10.1039/JA9860100035

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 41s AUTHOR INDEX Abd El-Haleem, S. H., Abdallah, A. M., S/491, S/508 Abdelrahman, A. N., S/129 Abdullah, M., 3939, S/1048 Abdullah, M. H., S/129 Adams, F., S/1064 Adams, F. C., S/115 Adams, L. A., S/C734, Adeberg, V., S/85 Ageev, M. N., S/198, S/791 Aikens, R. S., S/C768 Akrida-Demertzi, K., S/814 Aladyshkina, A. E., S/135, Albasel, N., S11210 Albers, D., S/C779 Albert, R., S/673 Albert, R. H., S/861 Albini, A., S/C549 Alcock, N. W., S/104 Alden, R., S/C310 Aldrighetti, F., S/480 Aleksanadrova, T. P., SO33 Aleksandrov, S., S/1120 Alekseeva, N. N., 970 Aleksenko, A. N., S/36 Aleshko-Ozhevskii, Yu. P., Alexandrov, S., S/C959, Allah, P., SIC417, SI495, Allavena, G., S/C549 Allen, G. M., S/126 Allen, S., S/174 Allison, J., S/150 Almeida, M.C., S/C249, Al-Samaraire, A. T. A., Alter, G., 9153 Amankwah, S. A., S1836 Amine, N. E., S/492 Analytical Methods Committee, S/926 Anan’eva, L. V., S/56 Andela, S., S/908 Andersen, A,, S/82 Andersen, I . , S/982 Anderson, H., S/C1173, Anderson, J . , S/1229 Anderson, R. A., S/927 Anderson, R. D., S/199, S/245 Anderson, R. J. M., S/716 Anderson, T. A., S/C289, S/C347, S/C764 Andersson, K., S/978 Andersson, P., S/481 Andrae, M. O., S/1021 Andreae, M. O., S/1035, Aneva, Z., 9866 Angerer, J., S/C525, S/989, Anigbogu, V. C., SIC318, Anikanov, A. M., S/57 Anikov, A. M., S/56 Anselmo, V. C., S/C281 Anttila, R., Sl84 Aparisi Querada, L., S/785 Apel, C. T., 9653, S1664 9492 s/c749 Sl136 S/787 S/1116 S/497, S/965 S/C270 SIC698 S/C1183 S/1051 s/994 SIC324 Apel, M., S/166 Appleton, J.M. H., S/808 Araghi, H. G., S/1213 Archibold, 0. W., S/831 Archuleta, F. L., SIC1127 Arellano, S. D., SlC317, Aristarkhova, G. G., S/791 Aronson, J . K., S/857 Arpadzhyan, S., S/1120 Arrowsmith, P., S/C291 Ash, K. O., 947 Asher, C. J . , S/1042 Astruc, A., 9117, S/587 Astruc, M., S/117, S/587 Atano, T., S/580 Aten, C. F., S/C737 Atinmo, T., S/50, S/790 Atnashev, Y. B., S/26 Auld, J. W., S/C783 Aulis, R., S/1227 Aurand, K., S11016 Avigour, A., S/C712 Avni, R., S/C704 Axner, O., 9636 Azegami, T., Sl58 Aziz-Alrahman, A. M., S/648 Baasner, J., S/C546 Bacetti, L. B., S/40 Bachmann, K., 9161, SIC569 Bacon, J. R., S/1075 Bahreyni-Toosi, M. H., S/184, 9644, SI999 Bai, L., S/1208 Bakalyar, D. M., 9406 Balicki, A., S/172 Balicki, M. R., S/833, S/1065, Balke, J., S/lOO, S/C423, Ballantine, D.S., S/941 Bamiro, F. O., S/C477 Bancroft, K. C. C., S/794 Bansho, K., S11216 Bao, C., 9839 Baoying, T., S/C675, SIC681 Baranova, L. L., S/173, Barbaro, M., S/C445 Barkowski, T., S/C314 Barna, P., S/915 Barnes, R. M., S/44, S/C421, S/588, S/638, S/C703, SIC706, S/C711, S/C719, S/881, S/974, S/984, S/1096, S/C1147, 91202, S/1218 Barnes, R. M., (ed.), S/1091 Barnett, N. W., Sic944 Barnett, W. B., S/C267, S/C272, SIC571 Baroni, U., S/858 Barrett, P., S/C755 Bartlett, H. E., S/C1191 Bar-Ziv, E., S/405 Bass, D. A., SIC253 Basta, N. T., 9792 Bastiaans, G. J . , SIC359 Batalov, V. I., S/26 Batifol, F. M., S/598 Baucells, M., S/C442 Bauch, F., SIC560 Bauer, C. F., S/C1146 Baurner, H. -P., S/C562 Bauslaugh, J., S/496 Baussand, P., S/809 S/C759, S/C760 S/C1171 S/983 S/1113 Bayer, W., S/C528 Beard, J., S/C734, S/C749 Beard, M., S/C734 Beaty, J ., S/C394 Beaty, J. S., S/731 Beck, R., S/C749 Beckwith, P. M., S/C783, Bellevance, T., S/C778 Belliveau, J., S/C723, S/C724 Belmore, R., S/C394, S/731 Belyaev, V. N., S/847 Ben, Y., S/C388, S/1221 Benabdallah, M. Z., S/117, Benassi, C. A., S/177 Bencze, K., S/C515, S/C529 Benga, G., S/1086 Bengert, G. A., S/1018 Bengtsson, M., S/1057 Bennett, P. A., S/C338, Benson, J. M., S/C1136 Bentley, G. E., S/C761, Berger, H., S/C567, S/C568 Bergert, K. D., S/977 Berggren, P, O., S/975 Bergner, K. G., S/925 Bergue, J. M., S/619 Berliner, L. D., S/173 Berman, S. S., S/201, S/C419, SjI037, S/1093, S11095 Bernal Melchor, A. M., S/15 Berndt, H., S/C546, 571032, Berneike, W., S/C560, S/C563 Bernier, J., S/C371 Berry, W.F., S/903 Bertenyi, I . , S/C451 Bertholf, R. L., S/985 Berthoud, T., S/1099 Bertolaccini, M. A., S/1083 Bertram, H. P., S1156 Besler, W., S/C559 Besson, J., S/809 Bettero, A., Sl177 Bettinelli, M., S/858 Bezhiashvili, G. N., S/4 Bezuidenhout, E. M., S/1004 Bhattacharya, S. K., S/645 Biederman, H., S/1212 Bieniewski, T. M., S/653, Biggs, W. R., 9131, S/C770 Bigois, M., S/C474 Bilba, D., S/35 Bilhorn, R., S/C765 Bilhorn, R. B., S/C370, Black, S. S., SIC251 Blades, M. W., S/45, S/C261, S/C307, S/C393, S/C414 Blanchard, J . , S/C452 Blinova, E. S., S/l111 Bloom, N. S., S/1012 Bogen D. C., S/C348 Bohler, W., SIC272 Bohmer, R. B., S/C1181 Bois, N., S/124, S/628 Bolibrzuch, E., S/1066 Bol’shakova, L. I., 9607 Bolton, A., Sl1227 Bolton, D., Y190 Bombelka, E., S/1060 S/1226 S/587 SIC682 SIC763 S/1062 S/664 SIC768 Bondareva, N.V., S/74 Boniforti, R., S/1043 Bonner Denton, M., S/C1126 Bonnin, E., S/C383 . Boon, N. A., Y857 Boorn, A., SIC771 Boorn, A. W., S/C291, S/C294, S/C295, S/C464, S/C745, S/C953, S/C1172 Boothe, E. D., S/210, S1485, S1641 Borg, H., S/481 Borggaard, 0. K., S/593 Bornstein, A. A,, 9854 Borsier, M., S/C427 BorszCki, J., S/C466 Boss, C. B., S/C381 Boulos, M. I., S/1218 Boumans, P. W. J. M., S/C411, S/1077, S/1078, S/1079 Boumans, P. W. J. M. (ed.), S/17 Bourcier, D. R., SIC313 Bourke, J. B., S/C737 Boutron, C. F., Y598 Bouzanne, M., S/1030 Bowen, H. J. M., S/197 Boyer, K. W., S/673 Boyle, J. R., S/1112 Bozic , J . , S/C364 Brackenbury, K . , SlCll90 Bradley, P., SIC368 Bragg, E., S/C728 Brajter, K., S/823 Branch, C.H., S/1107 Brandt, F., S/C566 Brandt, N., S/C738 Brandt, P. J . , SIC1 185 Braselton, W. E., S/812 Brashnarova, A., S/C951, Braude, A. Yu., S/818, S/900 Brech, F., SIC316 Brechmann, M. J., S/C758 Breder, C. V., S/C328 Bregadze, V. G., S/4 Brenner, I. B., S/222, S/C415, SIC707, Sl834, SIC1177 Brewer, S. W., S1637 British Standards Institution, S/242, S/899 Brocas, J., S/C453 Brock, J. C., S11002 Brodie, K., S/C695 Brodie, K. G., SIC349 Broekaert, J. A. C., S/C305, Bromble, S., S/C733 Brooks, K. A., S/C692 Brossier, P., S/102 Brovko, I. A. , S/218 Brown, A. A., S/883, S/893, Brown, P. G., S/C385 Brown, R. J., S/131 Brown, R. K . , S/C341, SIC729, s/c777 Brown, R. M., SIC462 Brown, S., S/985 Brown, T. M., SIC327 Browne, R.M., S/874 Browner, R. F., S/121, S/132, Bruce, M., S/C730 91025 S/1050 Sl918, Sl996 SIC28642s JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 Brumbaugh, W. G., S/C247 Brunner, W., S/C518 Bruns, H., S/C520 Bryan, S. R., S/674 Bryden, N. A., S/616, S/927 Brzezinska, A., S/172, S/1065 Brzezinska-Paudyn, A., S/833, Bubert, H., S/1081 Buch, B., S/924 Buchner, H. W., S/149 Buddin, N. P., S/1230 Bujupi, I., S/79 Bukhbinder, G. L., S/640 Bulska, E., S/887 Buratti, M., S/973 Burgener, P., Sic343 Burnett, R. T., 91201 Burns, I. G., S/1209 Bursey, J. T., S/C397 Burstenbinder, J., S/C1178 Burylev, B. P., S/75 Burylev, V. P., S/1123 Burylin, M. Yu., S/1123 Bushee, D. S., S/C334, S/1028 Bushway, A. A., S/1073 Bushway, R. J., S/1073 Butte, W., S/C562 Bye, R., S/175, S/826, S/1100, Bykhovskii, M.Ya., S/818, Byrd, J. T., S/1035 Cacho, J., 9202 Cadet, J. L., S/C443 Cadwell, L. , S/C723, SIC724 Callander, M. E., S/1204 Calzaferri , G . , S/973 Campbell, A. D., S/1029 Canepa J. A., S/838 Cantle, J. E., S/C746, S/C752, Cantwell, F. F., S/579 Caravelli , G . , S/973 Carleer, R., S/C446 Carmi, U . , S/C704 Carnahan, J. W., S/C386, S/849, S/1109 Carnrick, G. R., S/C267, S/C272, S/C571 Carter, D. E., S/976 Carter, R. D., S/C367 Cartwright, B., S/C701 Caruso , J . , SIC730, SIC956 Caruso, J. A,, S/C383, S/C385, S/C744, S/1224, S/1231 Casetta, B., S/177, S/480, S/909 Caughlin, B. L., S/C261, S/C414 Cavalli, P., S/995, S/lOOO, S/1099 Celesk, E. M., Y112 Chakrabarti, C. L., S/C1148 Chakraborti, D., S/115, Chambers, B., S/209, S/1230 Chan, C.C. Y., S/1101 Chan, S. K., S/C345 Chan, W., S/862 Chandler, H. A., S/643 Chandola, L. C., S/232 Chang, X., S/484 Channon, S. M., S/810 Chapman, A. H., S/1206 Chappuis, P., S/5 Chatelier, D., S/619 Chatt, A., SIC676 S/C1171 S/1105 s/900 S/C1142 S/1064 Chau, A. S. Y., Silo34 Chau, Y. K., S/1018 Cheam, V., S/1034 Chen, C. J., S/506 Chen, D., 9578 Chen, F., S/32 Chen, J., S/62 Chen, P. Y., S/806 Chen, W., S/106 Chen, X., S/C360 Chen, Y., S/48, S/839 Chen, Z., W989 Cheng, X., 9839 Cheung, Y., SIC463 Cheung, Y. Y., S/C1186 Chiang, S., W621 Childers, A. G., S/C292, S/C346, S/C398 Ching, L. C., S/C682, S/C683 Chiou, K. Y., S/113 Chmielnicka, J., S/811 Christensen, H. E. M., S/593 Christian, G. D., S/213, Christiansen, J., S/924 Chu, J., S/208 Chung, C. H., S/122 Chvany, C.L., S/C368 Claase, C., S/C1179 Clark, G. D., SIC756 Clark, J. R., S/188 Clarke, P. A., S/C469 Clay, D. E., S/54 Coates, P., SIC680 Cochran, J. L., S/C355 Coleman, D. M., S/126 Coleman, G. N., S/C318, S/C324, S/C355, S/C380 Coleman, J. T., S/137 Coleman, K., S/C723, SIC724 Coleman, M. F. M., S/864 Coles, S. M., SIC430 Colinet, E., S/401, S/402, Collecchi, P., S/1033 Collins, J . , SIC75 1, SIC774 Collins, J. B., S/C389, S/C392, S/C426, S/C429, S/906 S/C287 S/403, S/482 Colombi, A., S/973 Colombo, A., S1200 Comaford, D. J., S/C303, Combs, P. A., S/805 Condo, D. P., 9903 Cook, I. G., S/1117, 91234 Cook, K. K., S/938 Cook, W. P., S/1073 Cordos, E., S/25 Corr, S. P., S/C477 Corsini, A., S/621 Cortas, N. K., S/992 Cosma, M., S/25 Costantini, S., S/479, S/480 Cottenie, A., S1162, SI1210 Cox, A.G., 91234 Cox, J. A., S/C363 Coyle, F. T., S/C741 Cram, L. E., SIC399, S/875 Crawford, A. J., S/645 Crawford, R., S/C394 Creculius, E. A., S/1012 Cremers, D. A., SIC1127 Criaud, A., S/872 Crock, J. G., S/901 Crouch, S. R., S/C781, Cui, X., S/114 Cummings, P. M., S/1031 S/C760 S/C782 Currey, N. A., S/C690 Cutaj ar , J., S/C693 Czech, N., S/101 Czyz, J., S/214 Dabeka, R. W., S/861, S/940 Dabritz, J. L., S/C759, S1C760 Dadgar, D., S/183 Dai, L., S/583 Dalager, P., S/1121 Dalager, P. D., S/C303, Dalman, N., S/834 Damnjanovic, N., S/79 Dams, R., S/C413, S/871, Danen, W. C., SIC1143 Dannecker, W., S/C544, S/C567, SIC568 Danzer, K., S/151 Date, A. R., S/C461, SlC463, S/C1182, S/C1186 Daum, P. H., S/lOO3 Daus, R., S/C561 Davey, D. E., S/C699 Davidowski, L.J., S/854 Davidson, T. M., S/663 Davidson, W. R., S/C769, Davies, J. E., S/1117 Davirov, A. D., S/218 Davis, A. E., S/C1186 Davis, L. I. Jr, S/406 Dawson, J. B., 9184, 9644, De Ambrosis, A., S O 4 9 de Aza, S., S/C447 De Beer, B., S/C1172 De Benzo, Z. A., S/41 De Borger, R., SO07 De Broe, M. E., S/642 De Doncker, K., S/1072 de Galan, L., S/C420, S/C685, De Groot, G., 9159 De Groot, H. J., S/967 De Haas, E. J. M., S/967 De Jonghe, W. R. A., S/115 De la Guardia Cirugeda, M., De La Guardia, M., S/997 De Lecuw, I., S/789 De Mora, S. J., S/612, S/890 De Wolff, F. A., S/642, S/967 Deelstra, H., S/243 Deelstra, H. A., S/789 Degre, J. P., S/95 Dellien, I., S/1063 Deloye, F. X., S1619 Delves, H. T., S/103, S/895 Demarin, V. T., S/76 Demenna, G. J. DeMenna, G.J., S/C311, Demers, D. R., S/C279, Deming, R. L., S/C331 Demko, P. R., S/1028 Deng, C., S/799 Denton, M. B., S/C370, Department of the SIC760 S/1072 S/C776 s/999 S/1076 S/785 S/C735, S/C742 SlC344, S11217 SIC391, S/C765, S/C768 Environment and National Water Council, S/1040 Der Khatchadourian, F., S/965 Derie, R., 3851 Derler, R. A., S/C311 Desaulniers, J. A. H., S/1037 Deutsch, R. D., S/665 Deutsch, Y., SIC713 Dewalle, F. B., S1670 D’Haese, P., S/967 D’Haese, P. C., 9642 Di Pasquale, G., S/909 Diallo, A., SIC421 Diamantatos, A., S/C1187 Dibbs, H., S/13 Dick, W. A., S/929 Didorenko, T. O., S/667 Diehl, K. -H., S/167 Dietl, F., S/1038 Dietz, M. L., S/907 Dilworth, H. C., S/C1127 Ding, D., S/130 D’Innocenzio, F., Sl1083 Dinsmore, W. W., S/1204 Dittrich, K., S/C519, S/1108 Dmitriev, M.T., Sl900 Doerger, J. U., S/832 Doering, R. F., WOO3 Doerr, J. A., S/C758 Doherty, W., S/1227 Domyan, S., S/C774 Donard, 0. F. X., S/C276 Dong Song, L., Sl1044 Dorado Lbpez, M. T., S/C454 Doudou, T., Sl788 Douglas, D., S/C291, S/C294, S/C295, S/C743, S/C771, S/1065 Douglas, D. J., S/C464, SIC745, S/C953, S/C1172 Downey, S. W., Sl205, S/221 Drago, M., SIC549 Drazniowsky, M., S/810 Drews, M., S/1016 Drobyshev, A. I., S/27 Droessler, M. S., S/C271 Drude de Lacerda, L., Sl624 Drummer, 0. H., S/986 Druon, M., 9170 D’Silva, A. P., S/C288, S/C753, S/891 du Boer, J. L. M., S/908 Du, W., S/226 Dubois, D., S/C452 Duce, R. A., S/1011 Duennbier, V., S/856 Duffield, R. J., S/184, S/644, D’Ulivo, A., SI885 Dulude, G. R., S/C249, Dumarey, R., S/871, S/1072 Dungs, K., SIC512, S/998 Dunn, J.L., S/1018 Dymott, T., SIC1163 Dymott, T. C., 9822 Dymova, M. S., S/70 Dziesinski, W., S/29 Eastwood, I. W., S/3 Ebdon, L., S/163, S/C472, S/919, YC942, SIC958 Eberz, U., SIC573 Eckhoff, M., S/C730, S/C956 Ediger, R., S/C751, S/C755, Ediger, R. D., S/C302, s/999 S/C270 S/C1141 S/C392, S/C426, S/C429, SlC774, Sl906 Edmonstone, G., S/28 Egranova, I. G., S/134 Ehman, D., S/C281 Eid, M. A., S/129 Eiermann, R., S/C546 Ekanem, E. J., S/7 El Naggar, R. M., 949 Elbanowska, H., 9650JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 435 Eldad, H., S/C415, S/C707, Eldan, M., S1798 El-Defrawy, M. M., S/491, Elder, M. L. , Sl190 Elej alde , C., S193 Elin, R. J., S/922 Ellis, D. J . , S/184, S1644 El-Sayed, A. B., S/492 El-Shahat, M. F., S1492 Elsholz, O ., SIC535 Emel’yanov, B. V., S/900 Endo, M., S/1110 Engelmann , U . , SIC522 Engel’sht, V. S., S/1222 Engleman, R. Jr., SIC1140 Engvik, L., S/826 Enriques, J., SIC956 Eom, T. Y., S/80 E’Ottavio, T., 91003 Epstein, M. S., S1575, S1608 Eremenko, A. M., S1119 Erler , W . , S/C269 Erlich, S., S1222, S1834 Ermakova, N. V., S1660 Erzinger, J., SIC541 Esposito, M., S11033 Esser, P., SIC538 Essien, M., SlC340, S/C1132 Evans, J. C. , SIC736 Evans, K. L., S/C289, SIC764 Evans, S. J., SlC279, S1931 Evens, F. M., S/C306 Fadeeva, L. A., S/78, S1660 Fagioli, F., SI118 Faires, L. M., S/C264, Sl653, S/664, SIC1 140 Falk, G., 9236 Falk, H., S1234, Sl235 Fan, F., S/139 Fan, J., S/34 Fang, X., S11017 Fang, Y., S/817 Fang, Y. -Z., S/1007 Fang, Z., S/585, S1629, S1651, S11067, S/1114 Farino, J., S1121 Farmer, A.J. D., SIC399 Fartum, P., S/82 Fasching, J. L., S1836, S11011 Fassel, V. A., 9164, SIC288, S/597, SlC753, S1891 Fateley, W. G., SIC263 Faure, D., SIC459, S/C460 Faure, P. K., S/C1167 Feitsma, K. G., S1963 Fell, G. S., S/493 Feng, D., S/655 Feng, X., S/33, S/635 Fengdi, F., Sl1 Ferraroli, R., S/1043 Ferreira, N. P., S/C1196 Fidel’man, B. M., S/74 Fiehn, W., S/C530 Fietkau, R., S/C246 Finlayson, R. J., SIC678 Finton, D., S1592 Fiszman, M., S1624 Fitzgerald, E. A., S1854 Flanagan, E. B . , SIC350 Fleitz, P. A., SlC382, S/C385 Florian, K., S/C952 Fodor, A., S125 Fodor, P., SIC451 Fogg, T. R., S/C284 Fonseca Ruano, M. J., S/15 Fontana, F., S1858 Foote, J. W., Sl103 S/834 S/508 Formanek, Z . , S/238 Forsyth, D. S . , Sl1124 Foster, J .E., S/C741 Foster, P., S1809 Fouillac, C., S1872 Fox, J.B . , S1793 Fox, M. R. S., S1112 Fox, R. L., S/1225 Fox, S., SIC452 Frache, R., S11033 Fraile, V. R., S/41 Francois, J. P., S/C44G Frank, A., SIC428 Franke, J. P., S/963 Frary, B., S/C349, SIC695 Freimann, P., SO65 French, J. B . , SIC291 Fricke, F. L., S/112, SlC387, SlC722, SlC744, Sl928, s1937 Friedrich, H. B., S/138 Frigerio, I. J., S/C693 Frigieri, P., S11043 Fritz, J. S., S1848 Froelich, P. N., S/116, S11021 Fry, R. C., S/C246, SlC260, S/C262, SlC263, S/C298, SIC299, S/C376 Fu, X., S/888 Fuavao, V., S/912 Fuavao, V. A., S/1054 Fudagawa, N., S11056 Fujino, O., S165, Sl580, Fukuzaki, N., S/1023 Fulford, J . , SlC464, S/C745, SIC953 Fulford, J. E., S/C294, SIC295, S/C1172 Fuoco, R., Sl885 Furstenau, G., SIC532 Furunushi, Y., S/204 Furuya, H., S/622 Fuwa, K., S1208, S1486, S1921, 9936, S/939, S/979, S/1048, S11052, Sl1085 S11027 Gagne , P.H., S/C395 Gainford, A. R., S1244 Galiano, F., S1177 Gallego, M., S/840 Gallimore, D. L., SIC1137 Gan, S . , S1224 Gandjar, I. G., S/117 Gano, J. T., S/131 Garbarino, J. R., S/C954, Garber, R. W., Silo03 Garcia Vior, L. O., S171 Gardiner, P. E., SIC551 Gardner, J., SIC368 Garner, T. J., SIC758 Garnica, A., S/202 Gaston, C. M., 9112, S1928, Gawne, I(. M., S/55 Gedeon, A. Z . , S/C688 Geesey, G. G., 91205 Geiger, L. J., S154 Gelagutashvili, E. S . , S14 Gentry, J . S., SIC381 George, T., SIC734 Geraldes, M. da S . , S/C1165 Gerlach, W., SIC531 Gershman, L. L., S/1015 Gierczak, C. A., Sll50 Gieseke, U., Sic562 Gijbels, R.H. H., S/717 Gil’bert, E. N., 9640 Gillson, G., SIC390 SIC1 145 s1937 Giordano, R., S/479, 9480 Gips, C. H., S1646 Girvin, D. C., SIC736 Goddard, P. G., SIC752 Goddard, P. J., S/C437, S/C465, S/C746, SIC962 Godlevskii, A. P., S11013 Goedde, M., S/158 Goesta, L., S/2 Goldbart, Z . , S/C424, SlC457, S/C470, S/C705, S/C708, S/C714, S/798, S11106 Goldsmith, J. E. M., S1716 Gomez Coedo, A., S/C447, SlC454, 9666 Gomez, J. J., S171 Gomiscek, S., Sl231 Gonchakov, A. S., S177 Gonner, W., SIC514 Gonska, E., S/401 Gonska, H., S110, S1200, S/402, S/482, S/897 Gonzalez Perez, C., S/15 Goode, S. R., S11230 Goodgame, D. M. L. , S1187 Goodwin, T. G., S1645 Gorbacheva, 0. V., 9791 Gorbauch, H., S/153 Gordeeva, A. N., S/660 Gorlova, M. N., S/81, S/661 Gorny, G. R., SIC322 Gorovoi, B.M., 9818 Goshi, Y . , S1504 Goulter, J., S/1121 Goulter, J. E., S/C300, SlC303, S/C317, SK356, SlC759, S/C1197 Graeser, K., S11020 Grahame-Smith, D. G., S1857 Graner, C. A. F., S140 Grases, F., S187, 9902 Grasse, F. R., 9796 Gray, A. L., S/C418 Grazuliene, S., Sl77 Green, R. B., S1206, S1576 Green, R. J., S11042 Greenway, G. M., SIC944 Griebenow, W., S/1207 Griepink, B., S/10, S1200, S1401, S1402, S/403, S/482, S/897, S/1045 Griessel, J . F., S/C1160 Griffin, H., S/C721, S/C723, S/C724, S/C740, SIC778 Griffis, D. P., S/674 Grobecker, K. -H., S1182, S/C517, S/C548 Grobenski, Z., S/C269, SlC273, S/C556, SlC961, S1969, S/1047, S/C1161 Grosdaillon, P., SIC1177 Grosser, J . , S1193, S/878 Grote, B . , S/C444, SIC1175 Grote, M., SIC1169 Groves, W. L., S/C306 Gruzdeva, T.M., S/853 Gu, G., S119 Guan, S., S1601, S/870 Gucci, P. M. B . , S/1083 Guillard, O., S1192 Guletskii, N. N., S/36 Gunkel, G., SIC565 Gunshin, H., S/788 Gunter, W. H., SIC439 Guo, C., S/406 Guo, T., S/577 Guo, X., S1139 Haas, D., SIC730 Haas, D. L., SIC385, S1586, Sl1224, 91231 Haberland, K., SIC521 Haines, J., SIC1164 Hale, M., S/C1001 Hall, C. R., SlC769, SIC776 Hall, J. E., S1206, S1576 Hall, W. D., SIC775 Hallam, C., S/917, S/C943 Hallbauer, D. K., SIC1165 Hallis, K. F., S/857 Halls, D. J . , Sl993, S/493 Halsing, G. J., SlC756, Halverson, T. L., S1199, S/245 Ham, N. S., S/C400 Hamada, M., S/68 Hamada, S., S/37, S/67, S/215 Hamaki, M., S146 Hambrick, G. A., S/1021 Hamilton, E. P., SIC676 Hamilton, V. T., S/C761, Hampel, G., S/C523 Han, H., S/1019 Hanamura, S., S/212, S1627 Handy, R.W., SIC397 Hannaker, P., S/609 Hannon, R., S/C756 Hansen, E. H., S/629, S/1114 Hansen, R. J., SIC727 Hanson, R. L., S/C1136 Hara, H., S1504 Hara, L. Y., S/803 Hara, S., S/501 Haraguchi, H., S/208, S/486, S/506, S/876, S/939, S11048 Haraldsen, L. C., S/C1180 Harel, A., SIC457 Hareland, W. A., S/C1133 Haring, R., SIC250 Harms, U., S/C557 Harnly, J. M., S/194, SlC248, 3626, S/933, S1934, S1935, S1987, S/1014 Harriott, M., SI183 Harris, A. M., S/C1188 Harris, J. M., S/1114, S/C1128 Harrison, R. M., S/612, S1890, Hartenstein, S. D., S/C287 Hartenstein, S. T., 9213 Harzdorf, C., S/C526 Hasegawa, S., SI505 Hasegawa, T., S/486, 9506, Hassett D. F., SIC351 Hassett, D. J., S/C351 Haswell, S. J., SM94 Hauser, P., S/45 Hauser, P. C., SE307 Hausknecht, K.A., S/C738 Hausler, D. W., S1865 Havezov, I., S/824 Hayman, P. B . , S/187 Hays, R. G., SIC764 He, J . , S1584 He, Y., S11017 Hecko, Z . , S/217 Hee, S . S. Q., S/1112 Heindorf, M. A., SI150 Heinig, W., 9886 Heininger, P., S/856, S/1122 Heinrich, R., S/C525, S/994 Heinschild, H. -J., S/C541 Heller, S., SIC565 Heltai, D., S11043 Henderson, T. R., S/C1136 S/C766 SIC763 S/911, S/1199 S/876, S/104844s JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 Hendrick, M. S . , S/C330, Henrion, G., S/856, S/1122 Hentschel, W. M., S/47 Heonig, M., W241 Herak, M., S/C455, S/C471 Hergenreder, R., SIC1141 Hernandez, H. A . , SIC333 Hernandez Mendez, J., S/15 Hershey, J. W., SIC747 Hertz, R. K., SlC734, S/C749 Hess, C. F., S/408 Hewitt, C. N., S/612, S/890, Heyndrickx, A , , S/967 Hieftje, G.M., 9125, S/C257, S/C282, S/C292, S/C336, S/C346, S/C375, S/C377, S/C384, SIC398, S/C409, YC438, S/581, S/625, S/665, S/C1125, S/C1144, S/C1149, 91219 S/Cll52 S/911 Hill, H. H. Jr, S/C332 Hill, S. J., S/919, S/C942 Hinton, E. R., S/C350 Hiraide, M., S/500 Hirasawa, K., S/970 Hirate, N., S/C369 Hiratsuka, H., S/83 Hirokawa, K., S/123, S/220 Hirose, F., S/505 Hitachi, Ltd., S/237 Hoegetveit, A. C., S/982 Hoenig, M., S/487, S/591 Hoey, L. D., S/44 Hoffman, M. K., SIC298 Hoffmann, E., S11006 Hoffmann, E. W., S/807 Hoffmann, H. -J., S/C559 Hojyo, T., S/155 Holak, W., S11088 Holcombe, J. A . , S/C251, S/C252, SlC253, S/C27 1 Holding, S. T., S/C458 Holler, U., Sl977 Holz, E., 9930 Hoobin, D. L., S/C683, Hood, W. H., S/C373 Horiuchi, K., S/795 Horlick , G ., S/C293, S/C304, S/C315, S/C390, S/1049 Hornik, Y., S/C704 Horovitz, C. T., SIC709 Horvath, J. J., S/190 Horvhth, Z., S/C421 Horwitz, E. P., S/837 Horwitz, W., S/673 Hosseini, J. M., 9922 Hoste, J., S/C413, S/1072 Hou , J . , S/655 Hou, Q. L., S/609 Houk, R. S., S/C1150 Howe, A. M., S/1026 Howes, L., S/986 Hoyle, W. C., S/C727, S/C739 Hrudey, S. E., 91074 Hu, S., S/656 Hu, X., S/589, S/589 Hu, Y., S1105 Hua, B., S/C326 Huang , B . , S/603, S/870, Huang, F . , 9229 Huang, H., S/634 Huang, M., S/C257, SIC438 Huang, W., S/23 Hudnik, V., S/863 SIC689 HSO, W-D., S/C318 91092 Hue, S-F., S/C254 Huff, E. A . , S/837 Hughes, H., S1819 Hughes, S. K., S/C308, S/C756, SIC766 Huizenga, J. R., S/646 Hulanicki, A., S/887 Hull, D. R., S/C290, S/C309, SI1049 Human, H.G. C., S/995, SI1000, S/C1167 Hunt, J. L., S/244 Hunziker, H. E., SIC308 Hutchinson, E., S/972 Hutchison, D., S/1107, Hutchison, D. G., SIC461 Hutsby, W., S/1209 Hutton, R. C., S/C465, Huwel, L., YC723, S/C724 Hwang, D. J. D., SIC289 Iancheva, M., S/866 Ibrahim, M., SIC383, S/C730, Ichihashi, M., SI1068 Ichikawa, Y., S/1023 Ichinose, N., SI178 Ide, K., S/503 Ide, R. G., S/C702 Iiri, S., S/C369 Ikeda, M., S/910 Ilsoee, C., S/593 Imaeda, K. , S/195 Imai, S., SIC679 Imai, T., S/970 Imasaka, T., S/898, Sl1102 Imbert, J. L., S/96, S/C432, SIC434, s/c474 Inoue, M., S/970 Inoue, T., NO68 Inspektor, A . , S/C704 Inui, T., S/178 Ionashiro, M., 940 Irgolic, K. J., S/1064 Irsch, B., S/786 Ishibashi, N., S/1102 Ishizuka, T., S/66, S/216 Israel, Y., S/C711 Ivanfy, A. B., S/C1196 Ivanov, Y.V., S/1013 Ivanova, E., S/824 Iwanuma, K., S/882 Iwao, S., S/160 Izumi, I., S/58 Jackman, D. C., S/815 Jackman, S., S/C310 Jackson, C. J., S/1026 Jackson, K. W., S/3 Jackson, R., S/896, S/1022 Jaensch, P., 9149 Jaeschke, W., S/C561 Jain, M. P., S/C361 Jakubowski, M., S/811 James, J. V., S/406 Janrjen, E., SIC552 Janousek, I., S/658 Janssen, A., S/825 Janssens, E., SIC412 Jarrell, R. F., S/C368 Jassie, L. B . , S/C329 Jaxa-Bykowski, W., S/C448 Jay, P. C., S/892 Jennische , P . , S/C945 Jensen, A . , S/14 Jewell, K. E . , S/929 Ji, W., S/1059 Jiang, B., Sl23 Jiang, G . , 9657 SIC1 182 S/C700 S/1231 Jiang, H., S/1067 Jiang, M., S/128 Jiang, S., S1595 Jiang, Z., S/108 Jin, Q., S/130, SIC388, SI623, Jin, S., S/130 Jin, Y., S/233 Jin, Z., S/846 Jing, S., S/60 Jocobi, H., S/C562 Joergensen, S. E ., S/14 Johansson, G., S/1057 Johnson, J. L., S/845 Johnson, T., S/873 Jones, C., S/C301 Jones, J. B., SIC784 Jones, L., S/C1190 Jones, P., S/919, SIC942 Jordan, D. E . , SIC757 Jordanov, N., SI824 Jordon, D., S/86 Jorgesen, B., SIC778 Judelevic, J. G., S/1108 Junlong, Q., S/1 Junzhuo, L., SIC681 Kabil, M. A., S/1115 Kable, E. J. D., S/C1165 Kacsir , J . , SIC957 Kacsir, J. M., S/C319, S/C323, Kai-Jing, H., S/509 Kaiser, E. W., S/407 Kaiser, G., S/672, S/1069 Kaiser, H., S/1097, S/C1173, Kajimoto, M., 9970 Kakos, S., S/406 Kallmann, S., SIC1189 Kamberova, Ts., S/30 Kamla, G. J., S/C372 Kane, J. S., 31014 Kaneene, J. B . , S/812 Kaniewski, A., S/C679 Kanipayor, R., S/850 Kaplan, B. Y., S/173 Kaplan, B. Ya., S/1113 Kapoor, S.K., 9232 Karabegov, M.A., S/127 Karadjova, I., S/1116 Karadzhova, I., S/1120 Karasawa, I., S/913 Karl, B., SIC374 Karpenko, L. I., SI78, S/660 Kashio, Y., S/37, S/67, S/215 Kasimova, 0. G., SI607 Kassir, Z. M., S/C698 Kassovicz, J., Sl143 Katalevskii, N. I., S/56, S/57 Kato, N., W788 Katyal, M., S/611 Kawaguchi, H., S/152 Kawai, T., S/1085 Kazaryan, S. A., S/867 Keane, J. M., S/C262 Kehry , G., 9925 Keil, R., 9179 Keliher, P. N., S/C747 Keller, H. -G., S/C573 Kemp, G. J., S/615 Kerr, D. N. S., S/810 Kester, D. R., 91214 Kettrup, A., SIC1169 Khathing, D. T., S/171 Khattab, F. I., S/49 Kheng, L .C., SIC678 Kiboku, M., S/501 Kikui, N., S/889 Kimina, V. P., S/647 s/1221 S/C726 S/C1183 Kimura, A . , S/882 King, A. D., S16.52, S/855 King, E. E . , S/C260 King, J.N., S/848 Kingston, H. M., SIC329 Kinsey, W. J., S/C319, S/C323, S/C433, S/C473, S/C726, S/C732, SIC742, SIC957 Kirkbright, G. F., SIC944 Kirsch, B., S/212 Kirsch, D., S/85 Kisslak, G. E . , S/C389 Kitazume, E . , S/C285 Kizu, R., S/46 Kleff, U., S/C524 Kleijburg, M. R., S/1089 Klein, E., S/C733 Klemm, D., S/151 Klevay, L. M., S/111 Klick, D., S/407 Klim, M., SIC335 Klurjendorf, B . , Sic548 Klueppel , R . J . , S/93 1 Kluge, W., S/97 Klunder , G. L., S/C352 Knapp, G., S/92, S/C466, S/C514 Knutti, R., S/C547 KO, R., S/88 Kobayashi, J., S/614 Kobayashi, T., S/503, S/505 Koch, K. R., S/C1190 Kodama, Y., S/160 Koenig, R., S/235 Kohimeier, M., S/C530 Kohri, M., S/154 Koide, M., S/1044 Kolb, A., SIC511 Kollotzek, D., S/672, S/1009, Kolonina, L. N., S/847 Kolosova, L.P . , S/135, SI136 Kolysh, A. V., S/73 Kometani, T. Y., S/C275 Kometani, Y., S/65 Kompiang, S., SIC680 Kondo, K., Y882 Kong, Y., S/229 Koopman, B. J., S/646 Kooterin, S. A., S/647 Kop, F. R., SIC467 Kopfler, F. C., S/832 Koppel, P., S/715 Kopytin, Y. D., S/1013 Korallus, U., S/C526 Korennoi, E. P., S/36 Korepanov, V. E., S/26 Korol’kov, V. A., S/1013 Korshunov, M. B., S/1228 Koscielniak, P., S/21, S/498 Kosizky, V. P., SlC378, Kostrzewska, B., S/9 Kovacic, N., S/C703 Kovacs, Z., S/842 Kovalev, G. G., S/847 Kozak, E . , S/863 Kozusnikova, J., S/582 Krakovska, E., 999, S/C949 Kralicekova, E . , S/968 Krasil’shchik, V. Z., S/801 Krasnova, S. G., S/39 Kreeftenberg, H. G., S/646 Kretov, A. I., S/1123 Kreuning, G., S/C423 Kreuzer, W., S/167, SIC555 Krishnan, B., S/860 Kritsotakis, K., S/669 S/1069 SIC379JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 455 Krivan, V., S/191, S/C531 Krivozubova, G.V., S/133 Krull, I. S., S/C334, S/1015, Kubo, M., S/500 Kubon, Z., S/217 Kubota, T., 9502 Kuehn, J. E., S/137 Kuhns, D. W., S/718 Kujirai, O., S/154 Kulish, N. G., S/75 Kullmer, G., S/880, S/883 Kumamaru, T., S/501 Kumar De, A., S/C677 Kummerow, F. A., S/1086 Kumpulainen , J . , S/94 Kunselman, G. , S/C371 Kurfiirst, U., S/182, SIC517 Kurosawa, M., Silo48 Kurosawa, Y., S/1068 Kuss, H. M., S/181 Kuwagaki, Y., S/195 Kuz’michev, M. V., S/72 Kuznetsov, L. B., S/847 Kuzyukov, A. N., S/662 Labarraque , G . , S/42, S/C427, Labuda, J., S/13 Lacock, R., S/C258 Lacort, G., S/C442 Lahtonen, R., S/920 Laing, S., S/105 Lajunen, L. H. J., S/84, S/904 Lam, B ., S/C266, S/C268 Lam, J. H., S/C1171 LaMontagne, R. , S/C723 Lancione, R. L., S/142, Landheer, F., S/13 Langmuir, C., SIC733 Langmyhr, F. J., S/489 Lapteva, T. N., S/830 Laqua, K., SIC305 Larkins, P. L., 9654 Laverlochere, J., S/809 Law, R., S/C1139 Lawler, M. R., S/111 Laws, M. J., S/C1191 Lazarova, V., S/217 Lazure, R., S/C325 Le, T., S/610 Le, X., S/108 Leblondel, G., S/497 Ledent, G., S/107 Lee, C. H., S/80 Lee, G. H., S/C258 Lee, J., S/488, S/613, S/1041 Lee, M., S/893 Legere, G., S/C343 Lehmann, R., S/C273, S/1047 Leighty, D. A., S/C274, S/C290, S/C748, S/C750, S/C773, S/1119 S/1028 SIC449 SIC296 Leis, F., S/C305 Leonard, L. P., S/C738 Leong, M., S/C288 Lepla, K., S/C315 Leppin, S., S/193 Levine, D. Z., 9813 Levy, G. M., S/C325 Lewalter, J., S/C526 Lewis, B.L., S/1021 Lewis, S. A., S/194, 9987 Lewis, W. J., S/986 Li, A., S/971 Li, D., S/19 Li, K-P., SIC255 Li, S., S/33, S1635, S/657 Li, X . , 962, S/114, Sl846 Li, Y., S/623 Li, Z., S/1046 Liao, Y., S/971 Liao, Z., 9108 Lichte, F. E., S/C720, S/901 Liddell, P. R., S/C338, S/C682, S/C683, S/C689 Lienemann, P., S/C475 Liese, Th., S/C542 Lieser, K. H., S/120 Likhareva, N., S/30 Lillie, C. H., S/C332 Lin, H. M., S/806 Lin, J. L., S/1110 Lin, S., S/634 Lin, Y., S/602 Lindgren, I., S/636 Linton, R. W., S/674 Lippert, K. R., S/C396 Lips, M. J., S/176, S/990 Liqing, T., S/C681 Lister, B., S/C1193 Littlejohn, D., S/C477 Liu, C., S/601 Liu, C. Y., S/806 Liu, F., S/604 Liu, J., S/128, S/C388 Liu, S . , S/610 Locatelli, C., S/118 Long, G. L., S/C352 Lontsikh, S. V., S/219 Lorber , A ., S/C424, S/C457, S/C470, S/C705, S/C708, S/798, S/1106 Lord, D.W., S/894 Lorenzen, W., S/180 Lou, Z., S/16 Louvrier, J., SI95, S/124, Love, A. H. G., S/1204 Lovelace, R. R., S/639 Lu, H., S/128 Lucht, R. P., S/C1129 Luck, J., S/C1162, S/C1178 Luederitz, P., S/193, S/878 Luedke, C., S/1006 Lui, J., S/223 Lumas, B. K., S/C267, S/C571 Luo, P., S/31 Luo, X., 9484 Luque de Castro, M. D., L’vov, B. V., S/574, S/820, Lynch, A. T., S/C381 Lynskey, D. J., S/244 Lyon, R. S., S/C256, S/C327, Lyons, D. J., 9980, S/1087 Ma, Y., S/33, 960, 9635 Ma, Z., S/584 Maas, A. H. J., 9859 McAllister, T., S/C400 McAughey , J. J., S/966 McBridge, K., S/183 McCormack, J. D., S/C278 McCurdy, D. L., S/C299, McCurdy, R. F., S/620, MacDonald, T. J. , S/1112 McGann, D. F., SO073 McGuire, J.A., SIC258 Machado, I. J., S/232 McKee, C. M., S/1204 McKenzie, A. D., Sl861 McLaren, J. W . , SlC419, S/170, S/628 S/840 S/827 SIC762 S/C376 S/1053 S/1080, S11093 McLeod , C . W . , S/C463, S/C469, S/1117, S/1234 McMaster, D., SI1204 McNeal, T. P., S/C328 McNulty, R., SIC778 Macpherson, M. T., S/C437 Maessen, F. J. M. J., S/100, S/176, S/C423, S/C450, S/983, S/990 Magnuson, I., S/636 Magyar, B., S/C475, S/C534 Maier, D., SIC564, S/1024 Makhova, N. N., S/787 Makino, T., S/1203 Maksimov, D. E., S/72 Maksimov, G. A., S/39 Malamas, F., S/1057 Mallett, R. C., S/C1151 Malyutina, T. M., S/173 Manabe, R., S/C368 Mandry, R., S/1108 Manning, T. E., S/C380 Manoliu, C., S/884 Manuel, 0. K., S/113 March, J. G., S/87, S/902 Marchandise, H., 9482 Marinescu, M., SIC948 Marinov, M.I., S/8 Markuszewski, R., 91213 Marold, M., S/868 Marquardt, D. , S/193 , S/878 Marshall , J . , S/C477 Marshall, K. A., S/C257, S/C336, S/C438 Marshall, M. A., S/631 Marshall, W. D., S/1124 Martens, W. M., S/908 Martin, M. L., S/151 Martinsen, I., S/1070 Martinu, L., S/1212 Masal’tseva, L. V., 9827 Maskery, D., S/C364 Masna, 0. I., S/662 Masters, R. A., S/1220 Masuno, K., S/882 Matherny, M., S/99, S/C947, S/C952, S/C960 Mathre, 0. B., SIC725 Matousek, J. P., S/189, Matsuda, S., S/C679 Matsueda, T., S/69 Matsui, M., S/65, S/580, Matsumoto, K., S/979 Matsuno, K., S/160 Matsuo, H., S/501 Matsuo, K., S/1068 Matsusaki, K., S/916 Matsushita, Y., S/83 Matusiewicz, H. , S/588, S/C719, S/974, S/1096, s11202 Mauersberger, K. , S/886 Mauras, Y., S/C417, S/495, S/497, S/965 Mausert, T. A., S/137 May, K., S/157 May, T.W., S/845 Mayr, P., S/C570 Mazzucotelli, A., S/1033 Mbofung, C. M. F., S/50, Meckel, L., S/C539 Meddings, B., S/1097, SlC1173, S/C1183 Medina, R., S/141 Meeus-Verdinne, K . , S1241, SIC691 91027 S/790 9487 Mehra, M. C., S/C679 Meinhrat, A. O., 9809 Melcher, M., S/494, S/590, Melton, J. R., S/C784 Meranger, J. C., S/620, S/1053, S/1201 Mermet, J. M., S/96, SlC432, S/C434, S/C478, S/C543, S/1080, S/C1198 S/1024 Merrill, R. M., S/C277 Merz, W . , Sl90 Messman, J. D., S/575 Metz, G. H., S/C699 Meyberg, F., S/C567, SIC568 Meyer, G. A., S/873, SO232 Mianshi, Z., S/984 Miao, X., S/601, S/870 Michaelis, W. , S/C560, S/C563 Michel, R. G., S/C330, S/841, Michlewicz, K. G., S/849 Mierzwa, J., 929 Mikhailova, M., S/61, S/64 Miller, D ., SIC722 Miller, D. C., S/C382, S/C387, S/663 Miller, D. G., S/C385 Miller, M. H., S/C312 Miller, R. G., S/832 Miller-Ihli, N. J. , S/C366, Milley, J. E. , S/C676 Mills, J. C., S/C696, S/C702 Milne, D. B., S/617 Milner, B. A., S/918 Minderhoud, A., SO039 Mingjun, F., S/1 Mirtskhulava, N. A., S/843 Miskar’yanta, V. G., S/1111 Mitchell, J. C., S/C398 Mitchell, P. G., S/C320, Mitchell, T. E., S/55 Mittelman, M. W., S/1205 Miyazaki, M., S/46 Mizuike, A., S/152, SI500 Mo, D. M., S/C283 Modreski, P. J., S/C735 Mohl, C., S/12, S/158, S/C551, Moise, C., S/102 Moisov, L. P., S/1123 Montaser, A., S/C345 Moore, C. B., S/838 Moore, G. L., S/C1174 Morganthaler, L., S/C301 Mori, K., S/53 Morita, M., S/921, S/1052 Morris, B. W., S/615 Morris, P. T., S/406 Morrisroe, P.J., S/C395 Mostafa, M. A., S/491, S/508, Mostaghimi, J., S/1218 Mothes, W., S/1108 Motooka, J. M., S/600 Mottola, H. A., S/631 Mousty, F., S/988 Mowthorpe, D. J., S/C469 Miiller, N., S/C1170 Muller-Vogt, G. , S/C511 Mullins, T. L., S/490 Muntau, H., S/10, Sl200, S/C1152 S/626 SIC1134 SIC553 S/1115 S/401, 9402, S/403, S/482, S/897 Munter, R. C., 51199, 51245 Munz, H., Sl180 Murayarna, T., S149946s JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 Murphy, L. C., S/C249, Murzina, 0. I., S/853 Musikas, N., S/95 Muszynski, M., SIC449 Muzgin, V. N., S/26 Myasoedova, G. V., S/607 Mykytiuk, A. P., S/C419, Nabrzyski. M., S19 Nacu, A., Sl35 Nag Mo, S., S1155 Nagahiro, T., S/611 Nagashima, K., Sl68 Nagourney, S. J., SIC348 Nagy, S., SIC335 Naifeng, Z., 911 Nair, K.P. P., S/51 Nakahara, T., S/889 Nakamura, S., S/879, S/1068 Nakanishi, K., S/1102 Nakashima, S., S/483 Nakata, F., S/501 Nakayama, K., S/622 Naranjit, D. A., S/654, S/850 Narres, H. -D., S/12, S/C553 Nash, D. L., S/C354 Nasser, T. A. K., S/C698 Natarajan, S., S/905 Natishan, J. J., 9903 Natschke, D. F., SIC397 Nazarova, M. G., S/173, Nedler, V. V., S/1111 Neidhart, B., S/C512, S/998 Nel, I., SIC1190 Nerin, C., S/202 Nkve, J., S1494 Neve, N., S/C464 Nevissi, A. E., 9670 Newman, R., S/C250 Newman, W. J. Jr, S/C352 Newsome, D., S/C735 Ng, K. C., S/C758 Ng, R., S/C1173, S/C1183 Ng, R. C., S/1097 Ni, Z., S/802, S/1019 Nickel, H., S/C952, SlC960, Nicolas, D. J., S/C1191, Nielsen, J. S., S/1074 Niemczyk, T. M., S/C297, S/653, SI664, S/C1138 Niemzyck, T. M., S/C373 Nihon Jareru Attsushu, K.Nikdel, S., S/C367, S/1002 Nilsson, C. A., S/978 Nilsson, T., S/975 Nimjee, M. C., S/1065 Nisamaneepong, W., S/C730, S/1224, S/1231 Nisbet, J. A., S/972 Niu, W., S/C250 Niwa, M., S/970 Nogar, N. S., S/205, S/221 Nohe, J. D., SIC278 Nojiri, Y., 9208, S/506, Noller, B. N., SIC690 Nomura, T., S/913 Norman, E. A., Sit327 Norton, G. A., S/1213 Novak, L., SIC551 Novatskaya, N. V., S/135, S/C270 S11037 S/1113 S/C1153 S/C1192 K., S/146 S11085 S1136 Nygaard, D. D., S/C274, S/C290, SlC748, SIC750, S/C773, S/1119 Nygren, O., S/978 Odashima, T., S/59 Oechsle, D., 9672, 91069 Oefele, Ch., SIC523 Oesten, E., S/2 Ogaram, D. A,, S/185, S/632 Ogugbuaja, V. O., S/671 Ohasawa, K., S/195 Ohashi, Y., S/204 O’Haver, T. C., S/194, SJ575, S/626, S/987 Ohls, K., S/1061 Ohls, K.D., SIC416, S/C1154 Ohmori, Y., S/C679 Ohta, K., 9211 Ohtani, Y., S/195 Okada, A., SIC369 Okamoto, K., S/936 Okochi, H., S/154, S/505 Okutani, T., 9502 Okuyama, Y., Sl37, S/67, Olear , G., SIC740 Olear, G. G. 11, S/C721 O’Leary, M., S/188 Olesik, J. W., SIC259 Olle, M., S/C474 Olsen, K. B., SIC736 Omenetto, N., S/C410, S/988, S/215 Sl995, S/lOOO, S/1082, S/1099 Omololu, A. S/50 Ondasova, M., S/C947 O’Neill, P., SM94 Ono, A., S/67 Onoue, T., S/499 Orr, B. J., S/189 Orren, M. J., S/C1180 Osheim, D. L., S/923 Osipova, V. A., 9661 Ostapczuk, P., S/158 Osten, D. E., S/C337 Oster, O., SlC527, S/C550 Ostroumenko, P. P., S/119 Otsuki, A., S11085 Ottaway, J. M., S/C477, S/493, S/C5 10 Ozerskaya, G. A., S/791 Paakki, M., S/94 Pacheco, J. L., SIC359 Paetzold, H.J., Sl235 Page, J. R., S/929 Pak, Y., S/1219 Paksy, L., SIC946 Palmer, B. A., S/C754, Palmisano, F., S/1058 Panaro, K. W., S/1015 Pang, S., S1594 Papoff, P., S1885 Papp, L., 9842, S/1036 Parczewski, A., S1498 Parisi, A. F., S/C282 Park, C. T., Sic291 Park, D. A., S/C280 Parker, L. R. Jr., S/638 Parkinson, I. S., S/810 Parlow, A., S/C532 Parson, M. L., S/C280 Parsons, M. L., S/C256, SjC289, S/C327, S/C347, SIC762, SlC767, SI803 Passariello, B., S/C445 Patterson, D., S/C687 Patterson, K. Y., S/616 Pautova, L. F., Sl900 S/C1140 Pavlov, V. A., 9198, S/791 Peeler, J. T., Sl112 Peggs, A., S/197 Pelizzetti, E., S/649 Perkins, C. M., S/857 Perkins, S. L., S/877, S/914 Persson, L., 91063 Peter, F., S/1200 Peterson, E. J., S/C761, Petersson, L. R., SIC428 Petrov, A. S., S/1123 Pfeiffer, W.C., S/624 Phillips, Q. T., SIC766 Pickford, C. J., S/171, SIC462 Piepmeier, E. H., S/C258, SlC314, S/1220 Pietra, R., S/988 Pijpers, F. W., S/1089 Pilipenko, A. T., S/668 Pilipenko, E. P., S/26 Pillai, C. K., S/905 Pimenov, V. G., S/39 Pindar, A. G., S/1206 Pinel, R., S/117, S/587 Pinstock, H., S/1223 Piriou, A., S1192 Pivonka, D. E., S/C263 Platzer, B., SIC514 Pleban, P., S/C310 Pliesovska, N., S/C952 Plyusnina, E. E., S/830 Poisetti, P., S/858 Pokrovskii, V. A., S/73 Polansky, M. M., S/927 Poll, K. G., S/807 Poluektov, N. S., S/667 Pop, V. I., S/1086 Popescu, A., S/25 Popescu, O., S/1086 Porutiu, D., S/1086 Potapova, V. G., S/74 Potin-Gautier, M., 9587 Potter, N. M., S/639 Pougnet, M. A. B., S/C1180 Poulos, T., Sic371 Poussel, E., S/C478, S/C1198 Pramauro, E., S/649 Prellwitz, W., SIC550 Pressouyre , B ., S/C443 ‘Prewett, W. G., S/C1166 Price, W. J., S/822 Pritchard, M. W., S/1041 Pronchatov, A. N., S/39 Prosbova, M., S/968 Proudfoot, A., S/986 Proulx, P., S/1218 Pruszkowska, E., S/C302, S/C392, S/C426, S/C429, S/C751, S/906, S/C1141 Puchyr, R. F., S/C362 Pudill, R., S/C521 Puri, B. K., S1611, S/C679, Puschel, P., S/238 Qaan, Y., S/1017 Qi, W., S/1017 Qi, W. Q., SO96 Qin, F., S/19 Qiu, G., S/594 Qiu, T. Y., SIC441 Quaglia, A,, SIC325 Quan, E., SIC771 Quan, E. S. K., S/C294, SIC295, S/C464, SlC745, SlC9.53, S/C1172 Queay, J., Sl11 17 Queirazza, G., S/1043 Qun, T., SIC675 S/C763 s/1110 Raaijmakers, I. J. M. M., Radchenko, E. D., S/867 Rademacher, P., SIC560 Rademeyer, C. J., S/C1167 Rader, J. I., S/112, S/928 Radojevic, M., S/1199 Radziemski, L.J., S/C320, S/C440 S/C340, S/C1131, S/C1132, SIC1 134 S/1047 Radziuk , B . , S/C273, S/496, Radziuk, B. H., S/654, S1850 Ragazzi, E., S/177 Rains, T. C., S/C283, S/596, S/608, S/C1155 Ralston, N. V. C., S/617 Ramsey, J. M., SIC342 Ramsey, M. H., S/186 Ranger, C. B., S/C358, Rao, A. S., S/860 Rapsornanikis , S . , SIC276 Rawlins, L. K., SIC350 Ray, R. A., SIC306 Rayson, G. D., S/C336, Razumova, L. S., S/173, Rcheulishvili, A. M., S/843 Reamer, D. C., SIC744 Rechenberg, W., S/C537 Reiss, D., S/192 Ren, J., S/203 Ren, W., S/971 Rettberg, T. M., S/C252 Reyniers , K . , SIC446 Reynolds, J. G., Sic770 Rezaaiyaan, R., S/C336, S/581 Rezchikov, V. G., S/853 Rice, G. W., S/C753, S/891 Rice, T. D., SIC697 Richardson, L., S/C755 Richter, F.W., S/1060 Riddle, G. O., S/C720 Riddle, M. R., SIC251 Riepe, W., SIC1156 Ries, H., Sl1060 Rigg , J . C., S/404 Rigin, V. I., 51829 Riley, C., S/964 Ring, E. J., Sic1194 Rinne, D., S/599 Rios, J., S/C281 Rish, A. M., S/27 Rish, M. A., S1218 Riviello, J. M., SIC348 Robberecht, H., S/243 Robberecht, H. J., S/789 Robbers, J., S/156 Roberson, P., S/832 Robinson, J. W., S/210, S/C353, 9485, S/641 Robinson, M. D., S/C687 Rocks, B. F., S/964 Rodgers, A. L., S/981 Rodionova, T. V., S/835 Rodionova, V. N., S/140 Roe, K. K., Y116 Roeck, J. S., S/873, 91232 Rogers, E: R., SIC350 Rogge, M., S/599 Rohl, R., SIC559 Romero, F., S/93 Ronksley, B., SIC453 Roofayel, R. L., S1980 Rose, C. A., SIC293 Rosenblatt, G., S/C771, S/C365 SIC398 S/1113 S/C776JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 47s Rosick, U., SIC532 Rosopulo, A., S1167, S1182, Ross, P.F., S1923 Rossi, G., Sl995, SIlOOO, Roth, M., Sl158 Rotin, V. A., S1818, S1900 Rougnet, M. A. B., S1981 Roura, M., SIC442 Rousselet, F., S15 Routh, M. W., SlC300, SIC555 S11099 SlC317, SIC356, SlC476, SlC759, SlC760, SlC772, Sl1098, SIC1 197 Sl636 Rubinsztein-Dunlop, H., Ruchkin, E. D., S1816 Rudnevskii, A. N., S172 Ruggles, J. A., SIC1134 Rump, H. H., Sl153 RGliEka, J., S1213, SlC287, Rybina, T. F., S170 Rylaarsdam, J. C., Sl654 Ryzhova, R. I., Sl136 Sabbatini, L., 91094 Sabbioni, E., S1988 Sacks, R., SIC779 Sacks, R. D., Sl637, S11104 Sadler, M. F., SIC1191 Saeed, K., 9496 Sager, M., Sl828 Saichenko, A. N., 91222 Saichenko, L. A., Sl1222 Saikazu, M., S1882 Saini, G., Sl649 Sakai, T., Sl627 Sakla, A.B., S1508 Salisbury, C. D., 9862 Salit, M. L., SlC347, SIC767 Salopek, M. A., SIC1143 Samchuk, A. I., S1668 Samiullah, Y., Sl1118 Sanchez, L., SIC758 Sandron, B., SIC460 Sangerlaub, G., SIC562 Sangster, B., S1159 Sano, M., S1195 Sansoni, B., SIC518 Sarkissian, L. L., Sl804 Sarudi, I., Jr., S1110 Sarx, B., SIC569 Satake, M., Sl611, S11110 Sato, C., S1109 Sato, M., Sl504 Satoh, K., S159 Satzer, D., SIC383 Satzger, R. D., SIC744 Sauls, F. C., Sl903 Savinykh, V. N., 976 Savolainen, A., SlC721, SlC740, SIC778 Savory, J., Sl985 Scanavini, L., S1118 Schaefer, K., S1786 Schaller, K. H., SIC523 Scheeline, A., SlC372, S1718, Scheeren, P. J. H., S/915 Schenkelaars, H. J. W., Schikarski, M., SIC520 Schindler, E., Sl1084 Schinkel, H., 998 Schleicher, R.G., SlC309, SlC334, SIC775 Schleisman, A. J. J., SIC263 S1629, S11114 SIC1130 SIC440 Schlemmer, G., SlC516, Schliecher, R. G., SIC290 Schlieckmann, F., S11071 Schmid, W., Sl191 Schmidt, D., S1165 Schmidt, P. F., S1674 Schmidt, R., S1156 Schmidt, W., S11038 Schmiedel, G., SIC550 Schmitt-Henco, C. H., S1153 Schmitz, K. A., SIC327 Schmitz, K. H., Sl149 Schoenberger, E., Sl143 Schonburg, M., SlC560, SIC563 Schrader , W., SlC456, SlC545, SlC1161, SIC1175 Schram, D. C., SIC440 Schramel, P., Sl22, SlC554, Schreiber , B . , SIC466 Schroeder, W. H., Sl896, Schulman, S. G., S1797 Schulz Hendrick, M., Sl841 Schulze, G., SIC535 Schulze, H., S118 Schumann, P., S1151 Schutyser, P., SIC413 Schwarz, B., S11207 Schwarzer, R. R., S1671 Schwedt, G., Sl932 Scokart, P. O., Sl241, 9487, Scott, M., S1643 Sedykh, E.M., Sl607 Seeley, R. C., SlC284, SIC725 Seifert, B., S11016 Seiko Denshi Kogyo, K. K., Sekiguchio, H., S1882 Sekimizu, M., Sl614 Selby, M., S1189 Seliskar, C. J., SlC382, SIC1161 Sl852 Sl1022 S1591 S1145 SlC385, SlC387, S1630, Sl663, SIC722 SIC1152 Seltzer, M. D., S1841, Semenenko, K. A., S1661 Semenov, A. D., S156, Sl57 Semov, M. P., S11005 Semova, A. Y., S11005 Sen Gupta, J. G., S/507 Sermin, D. F., SlC425, SlC710, SlC1157, SlC1184, SIC1 197 Shabanova, L. N., Sl640 Shaburova, V. P., S1816 Shah, P., SIC1131 Shan, X., S1802 Shao, Y-B., SIC304 Sharicz, K. E., SlC341, SlC729, SIC777 Sharma, K. D., SIC361 Sharp, B. L., SIC422 Shatskaya, S. S., Sl133 Shaw, C. J., SlC465, SlC700, SlC746, SIC962 Shaw, G. J., SIC752 Shcheplyagin, E.M., 939 Shcherbinina, N. I., 9607 Sheikh, M. S., S1835 Shekiro, J. M. Jr., Sl605 Shen, D., 9610 Shen, H., Sl578 Shenhar, A., S1143 Sheppard, N., Sl404 Sherwood, R. A., S1964 Shevchenko, L. D., Sl78 Shevyakova, L. V., 9787 Shi, J., S116 Shibasaki, H., Sl882 Shibata, T., Sl37, Sl67, S1215 Shibata, Y., S11068 Shields, J. P., SIC258 Shimazaki, A., S183 Shimizu, H., Sl499 Shimozawa, M., S146 Shiowatana, J., SIC691 Shiryaeva, 0. A., Sl847 Shishov, V. N., S139 Shope, J. R., SIC741 Shore, L., Sl20 Shrader, D. E., S/C338, SIC955 Shreedhara Murthy, R. S., SIC421 Si, M., Sl19 Si, Z., S11059 Sidorovskii, A. I., 9219 Siemer, D. D. U., S1147 Siems, W. F., SIC332 Siggaard-Andersen, O., S1859 Silin, A. V., Sl867 Simensen, C. J., Sl82 Simeonov, V., Sl1116 Simon, F. J., S1825 Simoneaux, J., SIC374 Simonova, L.N., Sl835 Sims, G., SIC765 Sims, G. R., SlC370, SIC768 Sinemus, H. W., Sl169, SlC564, Sl1024 Siroki, M., SlC455, SIC471 Skibinski, L., S143 Sklarew, D. S., SIC736 Skrabak, J. W., SIC296 Slanker, M., Sl812 Slatkavitz, K. J., S144, SIC331 Slavin, W., Sl91, SlC272, SlC684, SIC686 Slickers, K., Sl148 Slonawska, K., Sl823 Slusarczuk, G., SIC738 Smirnova, E. V., S1219 Smit, H. C., Sl915 Smith, B. R., Sl592 Smith, B. W., Sl627, S11082 Smith, C. L., Sl600 Smith, D. D., Sl132, SIC286 Smith, J. B., Sl47 Smith, N. J., Sl966 Smith, R., Sl1004, Sl1008 Smith, S. B., SlC309, SlC334, Smith, S. B. Jr., SIC775 Smith, T. R., SIC391 Sneddon, J., SlC320, SlC339, S1805, Sl912, S11054, S11055, SlC1134, SIC1135 Snook, R. D., S1185, S1187, S1632 Soffientini, A., Sl909 Solin, T., Sl979 Sommer, R., SIC570 Song, W., S1233 Soriano Cubells, M.J., Sl785 Soriano, M. J., Sl997 Sotera, J. J., SlC249, SIC270 Soudiere, J., SIC452 Soupet, P., SIC459 Spann, K. P., S11087 Sparkes, S. T., SIC472 Sperling, K. -R., Sl168, Sperling, M., SIC544 Sprenger, K., SIC531 S11028 SIC513 Sprowell, W. L., Sl47 Stahlberg, R., Sl97 Staiger, K., S11020 Stallard, M. O., S11044 Stanislavova, L., Sl1025 Steel, A. W., Sl125, SlC375, Steele, R. J., Sl991 Steglich, F., Sl97 Stehle, J. L., SIC431 Stein, A., S1151 Steiner, J. D., SlC455, Stephens, R., Sl577 Stevenson, C. D., Sl796 Sthapit, P. R., 9493 Stober, J., S1832 Stoeppler, M., Sl12, S1157, Sl158, Sl166, SlC517, SIC551, SIC553 SIC377 SlC468, SlC471, S11098 Stone, G. J., SIC766 Stowe, H. D., Sl812 Strasheim, A., SIC1181 Strauss, J.A., SIC1196 Strelow, F. W. E., S1228, Strohl, A. N., Sl1211 Struckmann, I., Sl85 Stunzi, H., SIC558 Sturgeon, R. E., Sl201, s11037, s11095 Sturman, B., SIC695 Sturman, B. T., SlC357, SlC694, Sl800 Su, W., Sl33, Sl635 Subramanian, K. S., Sl620, s11053, s11201 Suddendorf, R. F., 9938 Sudo, E., S1503 Sugiyama, M., Sl65, Sl580, Suh, M. Y., S180 Sukhov, L. T., Sl821 Sulik, P. L., Sl603 Sun, D., Sl33, Sl635 Sun, L., Sl1067 Sun, Y., Sl601 Sun, Z., Sl846 Sunaga, H., S16 Sung, J. F.C., Sl670 Sutton, M. M., Sl488, S1618 Suzuki, K., S1614 Suzuki, K. T., S16 Suzuki, M., Sl211 Swan, J. M., Sl1104 Sweeney, V., SIC692 Sweileh, J. A., Sl579 Swenters, K. M. E., S1717 Swindall, W. J., 9183 Sztraka, A., SIC565 Tabatabai, M. A., Sl792 Tafforeau, C., Sl497 Taga, M., Sl109 Takagi, Y., SIC679 Takeuchi, T., SIC679 Talapova, D.M., S/133 Tamm, R., SIC536 Tamura, H., Sl178 Tamura, K., Sl195 Tan, S. H., SIC293 Tanaka, N., SIC363 Tanaka, T., Sl152 Tanner, R. L., Sl1003 Tao, C., Sl846 Tapia, T. A., Sl805 Tasker, D., 91121 Tasker, D. B., SK300, Taylor, A., Sl918, Sl996 s11229 S11027 SIC35648s JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 Taylor, H. E., SlC954, Taylor, P., SIC413 Temerdashev, Z. A . , 91123 Temmerman, E., S1871 Terada, S . , S1178 Terashima, S., S138, S1227 Tessari, G., S11094 Thiemann, E., S1149 Thomassen, Y., S!496, S1850, Thompson, K. C., 9917, Thompson, M., S1186, Thorburn Burns, D. , S1183 Tian, L., S1583 Tian, R., S116 Tian, S., Sl1208 Tielrooij, J. A., SIC450 Tikkanen, M. W., SlC297, Tilch, J., S1235, S11006 Timmins, K.J., SIC436 Tioh, N. H., S1638 Tiphaneau, K., 9192 Tisack, M., SIC779 Tobias, R. S., S1874 Tobschall, H. J . , S1669 Toelg, G., S1672, S11009, Tolmachova, M. T., S11228 Tominaga, M., S11216 Tondello, G., S11090 Toray Industries, Inc., S1239 Torgil, B . , S12 Torsi, G., S11058 Trassy, C., SIC441 Trost, R., SIC566 Tsalev, D., S11120 Tschoepel, P., S1672, S11009, Tsuge, A., S166, S1216 Tsutsumi, T,, 9195 Tumanova, A. N., S173 Tunni, M., Sl52 Turkin, Y. I., S127, S136 Turlakiewicz, Z., S1811 Tursunov, A. S., S1218 Tyas, M. J . , S1874 Tyson, J. F., S1207, S1804, Tzouwara-Karayanni, S. M., Uden, P. C., S144, SIC331 Ueda, J., SIC283 Ueda, T., S1502 Uehira, P., Sl1052 Uehiro, T., S1921 Ullmann, R., SIC540 Umezaki, Y., S11216 Umland, F., S11071, S11223 Ure, A. M., SlC950, S11075 Uren, B.I., SIC1188 Urh, J. J., SlC386, Sl1109 Ushakova, T. G., S1830 Uwamino, Y., S166, S1216 Vaeth, E., S1930 Valcarcel , M . , 9840 Vall, G. A., S1134 Valle, F. J . , SIC447 Valle Fuentes, F. J., S1666 Van Cleuvenbergen, R. J. A., Van de Vyver, F. L., S1642 Van Deijck, W., S1176, S1990 Van Der Linden, J., SIC1193 Van Der Plas, P. S . C., SIC1 145 Sl 1 070 SlC943, S11031 S/C1001 SIC1 138 S11069 91069 S1808 9814 S415 SlC420, Sl1076 Van der Walt, T. N., S1228, Van Diejck, W., S1100, S1983 Van Doren, J . B., S1592 Van Heerden, A. M., S11004 Van Hoeyweghen, P., S1591 Van Loon, J., S1172 Van Loon, J. C., S128, S1654, S11229 S1833, S1850, S11065, SlC1158, SIC1171 Van Loon, L. R., SIC1171 Van Maarseveen, I., SIC1196 Van Mol, W. E., S1115 Van Oeveren, R., SIC1179 Van Schoor, O . , S1243 Van Schoor, 0. E., S1789 Van Vilsteren, T., S1646 Vander Voet, A., S11227 Vanhentenrijk, S . , S1107 Vaughan, M. A., SIC293 Veber, M., S1231 Veillon, C., S1616 Veller, N. D., Sl81 Venkateswarlu, Ch., S1905 Veollkopf, U., SIC273 Verbeek, A. A., S11010 Verlinden, J. A. A., S1717 Verloo, M., S1162 Vernillo, I., S1479, S1480 Victor, A. H., SIC1195 Vierkorn-Rudolph, B., S1161 Vilsmeier, K., S1141 Vincent, W. R., S1797 Vindel Maeso, A., SIC454 Vinokurov, L. K . , S152 Vinson, J. A., SIC679 Vladimirskaya, I. N., S1847 Voellkopf, U . , SlC269, SlC961, S1969, S11047, SIC1161 Vogel , W., SIC468 Voight, H., 9977 Voinovitch, I. A., S195, S1124, S1170, S1619, S1628, S11030 Vokt, B., S/C555 Volk, P. R., SlC341, SlC729, Volkmer, M., SIC533 Volkova, E. A., S1667 Vollkopf, U., SlC556, SIC571 Vollmer, J.W., SlC392, SlC426, SlC429, S1906 Voropaev, E. I., S1801 Vos, G., S11215 Voth, L. M., SlC349, SIC955 Voudouris, E., S1814 Vracheva, N., S1824 Vrakking, J . J. A. M., SIC777 SlC411, S11077, S11078, S11079 Vrzgula, L., S1968 Vucko, H., S1863 Vuj itit, G . , SlC455, SIC471 VukiEeviC, S., SIC455 Vuori, E., 994 Waejten, U., S11060 Wagatsuma, K., 9123, S1220 Wagner, H. A., SIC522 Waidmann, E., S1158 Wakid, N. W., 9992 Wallace, G. F., 96.52, S1855 Wallwork, J . C., S1617 Walsh, J . N., SIC1163 Walters, P. E., SIC439 Walton, J. C., SIC737 Wan, C. C., S1621 Wandro, R. F., S1138 Wandt, M. A. E., S1981 Wang, C., S/105 Wang, C. C., S1406 Wang, F., S1623 Wang, G., S1799 Wang, J., S1229 Wang, N., S116 Wang, Q., S116 Wang, R., S160 Wang, S.T., S11200 Wang, X., S116, S132, S1230 Wang, Z., S132 Wangen, L. E., SIC761 Ward, M. K., S1810 Warner, D. K., S1630 Wassal, M. P., S1822 Waszczylo, Z., SIC364 Watling, R. J., SIC1168 Watson, A. E., SIC1176 Watson, P. S., SIC321 Watters, R. L. Jr., SlC283, Watterson, J. I. W., SIC1159 Weaver, S., SIC733 Webb, D. R., S1976 Weber, J. H., SIC276 Wedgewood, R. S., SlC760, Weers, C. A., SlC265, SIC572 Wei, F., S1196, S11017 Wei, G., S1583 Weidemann, H., SIC526 Weili, Z., S111 Weisberg, H. F., S1859 Weisel, C. P., S11011 Weissman, S. H., SIC277 Weitkamp, C., SlC560, SIC563 Welz, B . , S1494, SlC516, S1590, S11024 Wendl, W., SIC511 Wendt, H. R., SIC308 Werthmann, B . , S11207 Weunsch, G., S1101 Wheatley, M. R., SIC1001 Whiteside, P.J., S1822, S1880, Whitten, W. B., SIC342 Wholers, C., SIC394 Wibetoe, G., S1489 Wichman, M. D., SlC298, Wiedegreen, A. M., SIC278 Wiegand, P. M., SIC782 Wigfield, D. C., S1877, S1914 Wildeman, T. R., S1901 Wilhelm, E., SIC523 Wilken, M., S1924 Willay, G., S1869 Willey, D., S1592 Williams, J. C., 9137 Williams, R. R., S147 Williams, T. R., S1592 Willie, S . N., S1201 Willis, H. A., S1404 Willmann, K. H., S1825 Willoughby, R. C., SIC286 Wills, M. R., S1985 Willson, W. R., S1600 Wilson, B. L., S1671 Wilson, D. A., SIC384 Wilson, S. J . , S1874 Winefordner, J. D., S1190, Winge, R. K., S1597 Winkel, P., 9924 Wirsz, D., SIC393 Wittmann, A , , S1869 Wittmann, A. A., SIC467 Wittmer, C., S/C302 Wohlers. C. C., SIC309 S1608 SIC776 S1883 SIC376 S1212, 9627, S11082 Wolf, W.R., SlC366, S1933, s1934, s1935 Wolnik, K. A., 9112, SlC387, SlC722, Sl928, S1937 Wolthers, B. G., Sl646 Wong, P., S11227 Wong, P. T. S., S11018 Worsfold, P. J., S11117 Wrembel, H. Z., SIC435 Wu, B . , S116 Wu, J. C., SIC353 Wu, S., S124 Wu, T., S1223 Wu, X., S1225 Wu, Z., S133, S1635 Wunsch, G., S1181, SIC533 Wysocka-Lisek, J., S129 Xia, L., S11208 Xianhong, H., SIC675 Xianyang, Z., SIC675 Xiao-Quan, S., S1509, S11046 Xin, R., S1240 Xisheng, L., S11 Xiuhuan, Y., SlC675, SIC681 Xu, B., S1817 Xu, L., Sl31, S1594 Xu, L. Q., S122 Xu, S . , S1203, S1585, S1651 Xu, T., S1817 Xu, X., S1114 Xueping, W., SIC681 Yagi, M., S1483 Yajima, T., S16 Yakshin, V. V., S11228 Yamada, K., S1154 Yamamoto, M., S1659, S11103 Yamamoto, Y., S1659, S1916, Yamashige, T., S1659 Yan, Y., S160, S1130 Yang, C., S1589 Yang, G . , SlC388, 91221 Yang, H. C., SIC373 Yang, M., S1139 Yang, M. H., S1806 Yang, Q., S116 Yano, Y., S168 Yao, J . , S1223, S123 Yasuda, M., S11103 Yates, D., SlC751, SIC1141 Yates, D. A., SlC302, SlC392, SlC426, SlC429, SlC774, S1906 XU, B. -X., S11007 XU, T. -M., S11007 Sl1103 Yatsenko, L. F., S1574 Ye, G., S1633 Ye, L., Sl578 Ye, R., S1128 Ying, P., S1224 Yirts, J . G., Sl188 Yokota, A., S1195 Yokoyama, T., S1195 Yoshii, S., S183 Yoshikawa, M., S1788 Yoshino, T., S1916 Yoss, N. L., S11086 Young, C. P., SIC738 Young, K. F., SIC766 Youngberg, C., S1592 Yu, A., SlC388, S1623, S11221 Yu, B., Sl203 Yu, C., S160 Yu, S., S1633, S1888 Yuan, Z., S1802 Yuhin, Y. M., S1133 Zago, A., S11090 Zakharieva, I., S130 Zander, A. T., SIC312JAAS REFERENCE SUPPLEMENT, FEBRUARY 1986 495 Zarcinas, B. A., SIC701 Zauke, G. -P., SIC562 Zeeman, P. B., S/C439, Zeeuw, R. A., S1963 Zelyukova, Yu. V., S1667 Zeng, L., S/1208 Zeng, X., S1601, S1870 Zeng, Y., S1484 Zerbe, J., S1650 Zerezghi, M., S1606 Zhan, G., S1484 Zhang, B-C., SIC326 Zhang, D., S160, S1226 Zhang, G., S1578 S/1050 Zhang, H., S148, SlC388, 9623, S11221 Zhang, J., S1139 Zhang, L-X., SIC254 Zhang, M., S1105 Zhang, Q., S1623 Zhang, S., Sl585, S1651, Zhang, Y., S1484 Zhang, Z., S133, S1130, S1224, S1635, S11092 Zhanxia, Z., SlC675, SIC681 Zhao, C., S1106 Zhe-Ming, N., S11046 Zheng, J., S116 S11067 Zheng, Y . , S189, S1144 Zhizhuang, H., SIC345 Zhong, Y., S1105 Zhou, D., S1583 Zhou, H., S1604 Zhou, J., S1634 Zhou, T., S1584 Zhou, W., S163 Zhou, Z., S1589 Zhu, C., S134 Zhu, H., S1589 Zhu, J., S131 Zhu, Q., S1583 Zhuang, H. Z., S/881 Zhuang, M., SIC360 Zihlmann, J., SIC534 Zijlstra, W. G., S1859 Zil’bershtein, Kh. I., S11005, Zimmerman, R. L. Jr, S/C374 Zinger, M., SlC266, S/C268, Zizak, G., S1190 Zoller, W. H., S1941 Zolotovitskaya, E. S., S174 Zou, M., S1633 Zuehlke, R. W., S11214 Zul’figarov, 0. S., S1668 S11233 S/844Electronically typeset and printed by Heffers Printers Ltd, Cambridge, England

ISSN:0267-9477    

DOI:10.1039/JA986010041S

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 45 Direct Current Plasma as an Excitation Source for Flame Atomic Fluorescence Spectrometry-Some Applications Martha S. Hendrick, Philip A. Goliber and Robert G. Michel Department of Chemistry, University of Connecticut, Storrs, CT 06268, USA The concentrations of manganese, chromium, copper and cobalt were accurately determined in various standard steel samples, as were manganese, copper, zinc and iron in tomato leaves and bovine liver (National Bureau of Standards reference materials). Scatter signals were evaluated by means of the two-line method. Various potential spectral interferences were studied including the effect of 5000 pg ml-1 iron solutions on manganese and chromium determinations. The plasma behaved as would be expected of a narrow line source.The two-electrode version of the direct current plasma that was used, limited the precision of the measurements to 10% relative standard deviation. Keywords: Atomic fluorescence spectrometry; direct current plasma; scatter correction; spectral interferences; steel analysis The use of an inductively coupled plasma as a radiation source for excitation of atomic fluorescence in a flame has been studied in some detail by Winefordner and co-workers.1.2 A high concentration of any particular metal (typically 20 000 pg ml-l) was nebulised into the ICP and the resulting radiation was used to excite atomic fluorescence of the same metal as the analyte in a flame. The ICP proved to be an excellent narrow line source because of its freedom from self-reversal and its high stability.Atomic fluorescence detection limits were in the low ng ml-1 range for some elements but were poor (pg ml-1) for others. Omenetto and ~o-workers3~4 applied this approach to the determination of palladium in nuclear-waste samples3 and corrected for scatter4 by aspirating into the plasma a solution containing both the analyte and an element used to monitor the scattering signal. The correction was achieved simultaneously by subtracting the signals obtained with two monochromators, each of which observed the same area of the flame. This paper reports applications of the use of a direct current plasma (DCP) as a radiation source for the excitation of atomic fluorescence in a flame. We have found5 that the spectral output of the DCP, when a high concentration of a metal is aspirated, can be a narrow line only under carefully controlled experimental conditions.The way that the required high concentration is nebulised into the DCP is critical. The aerosol delivery tube (ADT) , which introduces the solution aerosol from underneath the DCP, has to be positioned carefully in order to ensure that atoms in front of the plasma do not cause either self-absorption, with subsequent broaden- ing of the line, or self-reversal, which results in a loss of intensity of radiation from the DCP and therefore a loss in atomic fluorescence signal. These considerations have been discussed in detail in another publication.5 Detection limits for DCP excited atomic fluorescence in a flame5 are typically in the 1-30 ng ml-1 range, which is comparable to detection limits obtained using other radiation sources, such as the ICPl or the xenon arc.6 As with the ICP, a few elements gave poor detection limits (pg ml-1).Literature detection limits obtained when microwave excited electrode- less discharge lamps (EDLs) have been used remain the best atomic fluorescence detection limits of all the non-laser approaches .7 These applications of DCP excited atomic fluorescence in a flame were developed in order to evaluate the potential seriousness of scatter interferences and to demonstrate that the DCP approach could be used to determine the concentra- tions of metals in samples. Scatter signals were investigated for various metals in several matrices by use of a two-line scatter correction technique using analyte ion and atom line emission from the DCP.Manganese was the metal that was primarily tested for scatter interferences. Theory of Two-line Scatter Correction In the conventional experimental arrangement for atomic fluorescence, the source intensity is modulated with a mechan- ical chopper or by electronic means. Fluorescence and scatter signals are independent processes and both occur when the source is illuminating the flame. The total signal may be expressed as the sum of the fluorescence and the scatter at the analytical wavelength. At the intensity of conventional sources (non-laser), both fluorescence and scatter are linearly related to the intensity of radiation at the wavelength of interest. In the two-line method of background correcti0n,~28 measurements of fluorescence and scatter at two different wavelengths allow the scatter to be estimated.This can be done by measuring the scatter signal from a second line generated by a second element that is aspirated into the DCP4 or, a non-fluorescing or fluorescing line of the analyte element at a second wavelength can be used to provide the estimate. The derivation of two expressions for the calculation of scatter correction factors using the two-line approach is presented here. The first expression is a general one that was modelled on that developed by Haarsma et aZ.9 They used their expression for scatter correction of atomic fluorescence when they were using EDLs as the excitation source at two different EDL operating temperatures.The expression devel- oped here allows a scatter correction to be made when both the analyte line and the second, scatter correction, line both fluoresce. The second expression is a simplified version of the first and is applicable where fluorescence occurs only at the analyte wavelength and not at the second wavelength. General Expression for Two-line Scatter Correction Let c = the unknown concentration (pg ml-1) of the analyte in the sample solution; c, = concentration of analyte; a = the known concentration (pg ml-1) of the analyte metal added to the sample solution (standard-additions method); hl = the wavelength of one analyte spectral line; h2 = the wavelength of a second spectral line for the same analyte; I0 = intensity of the plasma source; If = fluorescence intensity; Is = scatter intensity; It = sum of If and I,; and rn (1) = the slope of the calibration graph (fluorescence intensity versm concentra- tion).If the scatter signal is independent of analyte concent- ration (true for analyte concentrations below about 10L10346 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 pg ml-l), the slope can be determined from the increase in intensity measured with and without standard additions. (1) P = m (h2)lrn (hi) . . . . - * (2) q = I0 (h2)/10 (A,) * (3) It (c + a, hl) - 4 (c, hl) . . . . a m (hi) = We define two quantities, p and q , such that which is the ratio of the slopes of the analytical graphs at the two wavelengths and which is the ratio of the plasma source intensities at wavelengths hl and h2.Scatter intensity is proportional to the lamp intensity and independent of the analyte concentration, if the difference in wavelength between the two transitions is small: . . . . The fluorescence at a given concentration may be calculated from the slope of the calibration graph and the concentration: If (cx, hi) = m (hl) c, . . . . . . (5) We define a scatter-equivalent concentration, c,, as that concentration of analyte which produces a fluorescence signal intensity equal to the scatter intensity. This scatter signal would introduce systematic errors into the analysis if it were not corrected for by scatter correction techniques. The total intensity at a given concentration and wavelength is expressed as It (c, hi) = m (A,) c + I, (hi) .. . . (7) which rearranges to m (hi) c = It (c, hl) - I, (11) . . . . (8) The expression for the second wavelength is developed in cs (A,) = 1, (hl)/rn (hl) * - * * (6) the same way: Substituting I,(h2) from equation (4) and dividing equation (9) by equation (8) yields (h2) c = It (c, hz) - I, (hz) . . . . (9) (10) m (hz) c - It (c, h2) - 41s (hl) m ( h d c It (c, hl) - I s (hl) -- . . . . Incorporating the definition of p from equation (2) into equation (10) and solving for I, gives Substitution into equation (6) and solving for c, gives If a scattering solution, which does not contain analyte (c = 0), and hence does not fluoresce, is aspirated into the flame, expression (7) simplifies to Similarly, By dividing equation (14) by equation (13) and substituting into equation (4) we obtain It (C = 0, hi) = I, (hi) .. . . (13) It (C = 0, h2) = I, (h2) . . . . (14) Substitution of this expression into equation (3) gives . . a It (c = 0, h2) = It (c = 0, A,) . . (16) This can be determined from measurements of the scattering solution. The ratio p is then calculated from the slopes of the calibration graphs using equation (2). These are obtained from standard additions of known concentrations of analyte, as indicated in equation (1). Measurements of the total signal observed at each wavelength at any concentration c will provide the necessary data to solve equation (12). This allows a scatter correction to be made when fluorescence occurs at both the analyte line and at the second, scatter correction line. Expression for Scatter when Fluorescence Occurs at Only One Wavelength If fluorescence occurs at one wavelength, and not at the other wavelength, the expression for the scatter constant is simpler.If m (A2) = 0, then from equation (7) we obtain It (c, h2) = Is (h2) . . . . . . (17) Table 1. Instrumentation and experimental conditions Model number Apparatus (manufacturer) Parameter Conditions . . . . . . . . . . Plasma Spectrajet I1 Current 6 A (Spectrametrics Inc., Gas flow: anode and cathode 1.3 1 min-1 2.4 1 min-1 Andover, MA) Gas flow: nebuliser Monochromator . . . . . Photomultiplier tube (PMT) . PMT housing . . . . . . . Solution flow 1.4 1 min-1 Entrance and exit slits 0.2.5 mm Double monochromator Model DH-20A 0.2 m Middle slit 2.0 mm (Instruments SA, Band pass 0.5 nm Metuchen, NJ) 9893QB/3.50 (Thorn EMI-Gencom Inc., Plainview, NY) PR1400RF (Products for Research, Danvers, MA) Photoncounter .. . . . . Model1112 with Model 1120 discriminator (Princeton Appl. Res., Princeton, NJ) Microcomputer . . . . . . PetModel4106 (Commodore, Wayne, PA)JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 47 Equation (4) may be rearranged to give Substituting equation (17) into equation (18) gives Incorporating equation (19) into equation (6) yields 1, (A,) = 1, (h2)lq . . . . * . (18) I, (A,) = 1, (c, h2)lq . . . . . . (19) Therefore the scatter-equivalent concentration at one wavelength may be determined from the ratio of the line intensities, the slope of the calibration graph at the first wavelength and the total signal observed at the non- fluorescing wavelength for any analyte concentration.The scatter-equivalent concentration may be used as a correction factor for the systematic error introduced by scatter. If the correction is small, relative to the concentration of the analyte, it may be neglected. In the work described here it was attempted to use this general approach to two-line scatter correction by measuring the analyte signal at its atom line wavelength and measuring the scatter at a non-fluorescing or fluorescing ion line of the same analyte. It turned out that the scatter signals that were encountered were so small that it was not possible to evaluate truly this general scatter correction approach. The theory is presented here for future experimental evaluation.Experimental The plasma type and conditions and other instrumental settings are given in Table 1 and the experimental arrange- ment for the DCP excited flame atomic fluorescence system is shown in Fig. 1. The instrument has been described fully in an earlier paper.5 The aerosol delivery tube (ADT) was removed from the Spectrajet I1 DCP plasma apparatus and mounted separately to allow the ADT to be placed 3 mm behind the centre line of the DCPS in a direction away from the detection system. This prevents high concentrations of atoms from causing self-absorption and self-reversal in the region in front of the plasma through which the fluorescence was observed. The plasma was positioned with the plane of the electrodes perpendicular to the excitation optical path.The image of the DCP was focused on to a mechanical chopper that was operating at 80 Hz. The image from the chopper was then refocused and magnified at the flame position. It was magnified in order to ensure complete illumination of the flame within the field of view of the monochromator. The resulting image was about 15 mm in diameter and the magnification was optimised by moving the magnifying lens to and fro between the flame and the chopper to maximise the atomic fluorescence signal. Intense background emission from the tips of the electrode sleeves was excluded from the Chopper Separated DCP . I flame _ _ t w Reference Spherical mirror Fig. 1. Experimental arrangement for the DCP excited flame atomic fluorescence sytem excitation image by use of optical baffles.A nitrogen-shielded flame was used to reduce the flame background. The signal from the photomultiplier tube was set in phase with a reference signal from the chopper by use of a laboratory constructed phase shifter. The signal was then processed by the photon counter, which was operated in the chop mode (lock-in amplifier mode). The hardware and software for collecting the data from the photon counter has been described elsewhere. 10 Sample Preparation Stock solutions. Stock solutions destined for introduction of high concentrations into the DCP were prepared by dissolving analytical-reagent grade chemicals in the minimum amount of hydrochloric acid followed by dilution with de-ionised water to a concentration of 20000 1.18 ml-1. Standard solutions for use in the flame were prepared from 1000 pg ml-1 solutions diluted from the higher concentration stock solutions.Steel samples. Steel samples (0.5 g) (National Bureau of Standards, Standard Reference Materials SRM 126c and SRM l l h ) were placed in beakers; 10 ml of concentrated hydro- chloric acid (37%) and 5 ml of concentrated nitric acid (70%) were aded to each sample. Known amounts of the 1000 pg ml-1 aqueous standards were then added to these solutions to allow for the use of the standard-additions technique. A blank solution containing only the acids was prepared in the same manner. The solutions were warmed on a hot-plate for 1 h or until clear, without reducing to dryness, then diluted with de-ionised water, filtered through ashless filter-paper and further diluted to 200 ml.For manganese determinations a 10-ml aliquot of each solution was further diluted to 100 ml. Tomato leaves. Samples of tomato leaves SRM 1573 (0.25-0.5 g) were spiked with standard additions and digested with 10 ml of nitric acid followed by addition of aqua regia as described by McCaffrey et ~ 1 . 1 1 In addition, a variety of dissolution techniques were evaluated for the determination of iron because preliminary data for iron determinations gave low results. Ultimately, the most accurate iron results were obtained by use of a total dissolution technique that is used at the National Bureau of Standards.12 Each sample was digested in a PTFE beaker with 25 ml of concentrated nitric acid and 25 ml of distilled water for 1 h; 10 ml of perchloric acid and 5 ml of hydrofluoric acid were added, the beaker was covered and the sample digested for 2 h.The cover was then removed and the sample evaporated until it was nearly dry; 2 ml of nitric acid and 30 ml of water were then added and the mixture heated at low temperature (<60 "C) until the sample had dissolved. A clear colourless solution resulted, which was diluted to 100 ml. Bovine liver. Samples (0.5 g) of dried SRM 1577a were placed in large heavy walled test-tubes and standard additions were made. Ten ml of concentrated nitric acid were added to each sample and the tubes heated for 3 h in a hot water-bath. The solution was then cooled and diluted with 0.04 M hydrochloric acid to 100 ml. Iron was determined using a modification of the procedure for tomato leaves by omitting the hydrofluoric acid and by reducing to 5 ml the amount of perchloric acid that was used.Results and Discussion Magnitude of Scatter Signals Synthetic scattering solutions In order to obtain information about the size of scatter signals manganese was used as an example element and scatter was measured at the analytical atom line (279.48 nm) and at an ion line (259.37 nm). The ion line is sufficiently close in wavelength to the atom line to justify an evaluation of its use as the second line in two-line background correction. The48 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 Table 2. Signal sizes from various scattering solutions Average signalkounts s- Scattering solution Mn I MnII Lanthanum (10000 yg ml-1) 232 1190 Aluminium (1000 pg ml-1) .. 316 1840 Calcium(1000ygml-1) . . 74 223 Vanadium (1000 yg ml-1) . . 45 215 Iron(5000pgml-1) . . . . 88 188 De-ionised water (0.04 M HCl) 20 200 Signal ratio 5.1 5.8 3.0 4.8 2.1 10.0 scatter signals obtained from various scattering solutions at each of the two lines are listed in Table 2. The data indicate firstly, that the ion line is significantly more intense than the atom line because the ion line scatter signals are by far the biggest and secondly, that scatter is a potential interference in DCP excited atomic fluorescence spectrometry. The ratios between the ion line and atom line scatter signals for each matrix are also given in Table 2. These ratios differ by a factor of between two and five. The ratios cannot be expected to be the same for all matrices because of the complicated relationships that exist between the size of the scatter signals and the particle sizes that exist in the flame, the wavelength of the incident radiation and the concentrations of the scattering solutions.These complications are discussed at length in papers on scattering in atomic fluorescence by Larkins and Willis13 and by Doolan and Smythe.14 It is not important that these complications exist provided that measurements are made that account for these effects. This is implicit when two-line scatter correction is used and can be carried out by appropriate use of the equations developed above. The equations allow for possible fluorescence at both of the lines that are of concern. Fluorescence at the manganese ion line was observed and was about one fifth that of the atom line when a manganese solution in de-ionised water was being aspirated into the flame. Real Samples Despite the above observation of fluorescence at the ion line, fluorescence was not detected at the ion line when measure- ments were made of signals produced from tomato leaf samples to which standard additions had been made.This was probably because the tomato leaves contain 4.5% potassium, 3% calcium and 1% magnesium on a dry-mass basis. These would certainly act as ionisation suppressants in the flame and thus prevent there being a significant enough population of manganese ions to cause fluorescence at the ion line. Therefore, for tomato leaves the experiments described here were simplified because no account had to be taken of fluorescence at the second line.At the manganese ion line there was no significant difference in signal size between the blank solutions (245 k 28 counts s-1) and the tomato leaf sample solutions (261 k 30 counts s-1). The atom line was less intense than the ion line (Table 2), which indicated that when making measurements at the atom line no difference was to be expected in scatter signals between the blank and tomato leaf samples. Accord- ingly, the acid blank was, in general, subtracted from the sample signals at the analyte wavelength without further scatter correction. This applied to both tomato leaves and to the determination of manganese in bovine liver for which no fluorescence at the ion line, nor differences in scatter were observed between the acid and blank samples.Measurements of scatter signals for the determination of manganese in steel showed that fluorescence could be detected at the ion line. However, at the concentration of manganese in steel (0.5%), the ratio between scatter and fluorescence was very small and lay between 0.01 and 0.02. The precision of the measurements for all samples was typically 10% relative standard deviation. Hence, it could not be said that the measurements of the ratios were precise enough to use the ratios for any calculations of scatter correction factors. Interpretation of the Scatter Signal Sizes The reason for these small scatter signals is likely to be because the DCP is not a particularly intense radiation source for atomic fluorescence.Microwave excited electrodeless discharge lamps (EDLs) are much more intense sources and real sample scatter signals are small but significant and need to be corrected.15 The DCP gives a detection limit (Table 3) for flame atomic fluorescence of manganese that is about a factor of 10 worse than with manganese EDLs.16 This can be extrapolated, approximately, to mean that a factor of 10 smaller scatter signals can be expected when the DCP is used rather than the EDL. We have not measured real sample scatter signals for manganese EDLs, but for cadmium EDL excitation, which results in a similar detection limit15 (0.1 ng ml-1) to manganese16 (0.2 ng ml-I), real sample scatter signals are typically 130-230 counts s-1.13 Identical instrumen- tation and experimental conditions were used in references 15 and 16 and in the present work.If this figure of 130-230 is reduced by a factor of 10 to account for the smaller intensity and poorer detection limit (2 ng ml-1) characteristic of the DCP, then it is clear that the scatter signals in the DCP can be expected to be small and probably need not be corrected. Other considerations such as the realtive widths of the lines from these two sources would change the sizes of scatter signals slightly if there were significant differences in the widths of these two line sources. However, the intensity effect is likely to be the most significant factor. In support of these arguments, all our attempts to measure scatter signals in all real sample matrices revealed scatter signals that were indistinguishable from the noise. The data reported later in this paper were obtained without measurements of scatter signals after checking at a second non-fluorescing line to make sure that scatter signals were not measurable.This does not mean that scatter signals are always going to be too small to cause a problem in DCP excited flame atomic fluorescence. The size of the scatter signal varies greatly with the type of sample and with the amount of sample dissolved in solution. It is only true that the samples tested here did not give significant scatter signals and that with a more intense light source the same samples would probably have caused larger scatter signals. Spectral Interferences Spectral interferences are rare when a narrow line source is used for the excitation of atomic fluorescence in flames.We have found5 that the DCP does behave like a narrow line source, but the possibility exists of the emission line in the plasma being broad enough to excite fluorescence of species other than the analyte. Table 3 lists analyte lines and potential interferences that we have studied previously11 in the context of wavelength modulated continuum source excited atomic fluorescence spectrometry in a flame. These were significant interferences for the latter technique and were therefore the most likely to produce an interference here for DCP excited flame atomic fluorescence. The interferents listed in Table 3 were studied at the high concentrations indicated and the degree of interference at the analyte line expressed as the concentration of the analyte that would give the same size of signal.In all instances the interferences were insignificant because the interfering signals were at levels that were smaller than the detection limits for these elements. The most serious interference was that of iron on the determination of chromium; 5000 pg ml-1 of iron pro- duced a signal at the analyte line equivalent to 0.02 pg ml-1 ofJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 49 Table 3. Potential spectral interferences for DCP excited atomic fluorescence in a flame Interferent concentratiordyg ml- 1 ~ ~~ ~ DCP Analy te * Analyte detection limit Wavelengthhm Interferent Actual equivalent Cr , . . . . . 11 357.9 Fe 357.9 and 358.6 5000 - Ni 359.8 1000 - Cr .. . . . , 11 359.3 Fe 358.1 and 358.7 5000 0.02 Ni 359.8 1000 - Mn . . . . . . 2 279.5 Mg 279.6 and 280.3 1000 - Pb 280.2 1000 - * Analyte equivalent concentration is defined as the amount of analyte that would produce the same signal as the interferent scatter signal. Table 4. Analysis of Standard Reference Materials by DCP excited atomic fluorescence in a flame SRM High-nickel steel (SRM 126c) Steel, 0.2% carbon (SRM l l h ) Tomato leaves (SRM 1573) . . Bovine liver (SRM 1577a) . . Element . . Mn c u Cr c o c o c o . . Mn c u Cr . . Mn c u Zn Fe . . Mn c u Fe Zn * Uncertainty 2 1 in last significant figure. t Standard deviation of three replicate samples. $ 95% Confidence limits. W avelengthhm 279.48 324.7 240.7 241.1 241.4 279.48 324.7 357.8 279.48 324.7 213.8 248.3 279.48 324.7 248.3 213.8 357.8 Standard value 0.468% * 0.040% * 0.062% * 0.008% * 0.510% * O.O61%* 0.025 O/o * 238 f 7 yg g- l$ 62+6pgg-1$ 690 k 25 pg g-l$ 9.9 f 0.8 yg g-1$ 158 -+ 7 pg g-1$ 194 f 20 pgg-l$ 123 + 8 pg g-1$ 11 t 1 ygg-1$ Determined value 0.44 k 0.4%? 0.043 2 O.OO5YoT 0.016 f 0.008Yot 0.010 -+ 0.003°/ot 0.49 f O.O46%T 0.060 k 0.008°/ot 0.021 k 0.003°/oT 241 f 15 pgg-lt 14+4ygg-'f 57k5pgg-'t 732 2 70 yg g-lt 9.5 f 1.5 pgg-lt 164k 10pgg-lt 196-+29pgg-'t 113 k 7 yg g- lt analyte.The detection limit for the determination of chro- mium is 0.01 yg ml-1. This also cannot be considered a significant interference. In addition, it is not clear that the latter small interference is indeed a spectral interference. It could, for example, be due to scatter rather than fluorescence but it is too small for meaningful measurements to be made in order to demonstrate the source of this inteference. In a final experiment concerned with spectral interferences manganese was aspirated into the plasma in order to carry out a manganese determination in the flame. In this situation a standard solution of high-purity iron in the flame (5000 pg ml-l) was found not to cause any interfering signals at the manganese 279.48-nm line.Such an interference was plausible due to the presence of the iron 279.50-nm line. Standard Samples Analyses The results of the determination of various metals in some NBS Standard Reference Materials are given in Table 4. For the analysis of basic open-hearth steel (SRM l l h ) and high-nickel steel (SRM 126c), the slopes of the aqueous calibration standards were equal to the slopes of the standard- additions graphs.Direct analysis of the unspiked sample by comparison with aqueous standards gave good results for all elements except cobalt, which was present at quite low concentrations, and this analysis required a standard- additions approach. The analysis of steel samples using aqueous standard reference solutions has been reported for hollow-cathode lamp excited atomic fluorescence in an ICP,17 which together with the data presented here further demon- strates the high spectral selectivity of atomic fluorescence with two different light sources. ICP and DCP emission techniques require background correction techniques because of a complex, line-rich background generated by the iron matrix ,I8 and matrix matching is required.Results for tomato leaves and bovine liver (Table 4) are in good agreement with the standard values although the precision of the analyses were poor. A two-electrode jet DCP, which is not a very stable plasma, was used in this work. Its more modern successor, the three-electrode jet, is more stable and if it were to be used for excitation of atomic fluorescence would probably lead to better precision at the normal concentrations used in analyses. It is unlikely that such an improvement in precision would be paralleled by an improve- ment in detection limit, as radiation source noise is not usually the limiting noise in atomic fluorescence measurements. Source noise is a fixed proportion of the total fluorescence signal and hence decreases with the signal size.Here the limiting noise was flame background noise. Conclusions This work has demonstrated, for real samples analyses, that DCP excited atomic fluorescence in a flame appears to be free from both large scatter interferences and spectral interfer- ences. However, scatter is a potential interference if particu- larly highly scattering matrices are to be analysed. The reason for the generally small size of the scatter signals encountered in this work most likely lies in the low intensity of the DCP as a line source for atomic fluorescence. This is indicated by detection limits that are only in the 1-30 ng ml-1 range.5 The freedom from spectral interferences is a natural consequence of the use of line source excitation.50 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 The most likely application of DCP excited atomic fluores- cence is as an adjunct to DCP emission spectrometry at times when spectral interferences in the DCP can be conveniently solved by use of the DCP excited atomic fluorescence approach.This has also been suggested3 as the most likely application of ICP excited atomic fluorescence in a flame. In many ways the DCP - flame fluorescence approach can be compared to continuum source excited atomic fluorescence in a flame where sequential multi-element analysis is a significant advantage because the radiation source can excite most elements in the flame.7 For DCP excitation, only the high concentration of solution that is introduced into the DCP needs to be changed in order to change the element to be determined.Further, with continuum source excitation spec- tral interferences cause a few problems.6Jl These are non- existent when the DCP is used as a line source for excitation. Disadvantages of this approach are similar to those for continuum source excited atomic fluorescence in that it is only a sequential rather than simultaneous multi-element tech- nique and sensitivity is not an improvement over any currently used techniques. Acknowledgment is made to the United States Coast Guard, Research and Development Center, Groton, CT, for the loan of the DCP, the donors of the Petroleum Research Fund administered by the American Chemical Society, Research Corporation and the University of Connecticut Research Foundation, for partial support of this research.R. G. M. was supported by a Research Career Development Award from the National Institutes of Environmental Health Sciences under grant number ES00130. References 1. Epstein, M. S., Nikdel, S., Omenetto, N., Reeves, R., Bradshaw, J., and Winefordner, J. D., Anal. Chem., 1979,51, 2071. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Omenetto, N., Nikdel, S . , Bradshaw, J. D., Epstein, M. S., Reeves, R. D., and Winefordner, J. D., Anal. Chem., 1979,51, 1521. Cavalli, P., Rossi, G., and Omenetto, N., Analyst, 1983, 108, 297. Omenetto, N., Crabi, G., Nesti, A., Cavalli, P., and Rossi, G., Spectrochim. Acta, Part B , 1983, 38, 549. Goliber, P. A., Hendrick, M. S . , and Michel, R. G., Anal. Chem., 1985, 57, 2520. See also Hendrick, M. S., and Michel, R. G., 9th Annual FACSS Meeting, Philadelphia, PA, 1982, September 19-24, paper number 196; and Hendrick, M. S . , Goliber, P. A., and Michel, R. G., 13th ACS Northeast Regional Meeting, 1983, June 26-29, paper number 23. Johnson, D. J., Plankey, F. W., and Winefordner, J. D., Anal. Chem., 1975,47, 1739. Wu, M.-L. W., and Michel, R. G., Analyst, 1985, 110, 937. Doolan, K. J., and Smythe, L. E., Spectrochim. Acta, Part B , 1977,32, 115. Haarsma, J. P. S . , Vlogtman, J., and Agterdenbos, J., Spectrochim. Acta, Part B, 1976, 31, 129. Wu, M.-L. W., and Michel, R. G., Anal. Instrum., 1984, 13, 117. McCaffrey, J. T., Wu, M.-L. W., and Michel, R. G., Analyst, 1983, 108, 1195. Epstein, M. S . , personal communication. Larkins, P. L., and Willis, J. B., Spectrochim. Acta, Part B, 1974, 29, 319. Doolan, K. J., and Smythe, L. E., Spectrochim. Acta, Part B , 1979, 34, 187. Michel, R. G., Hall, M. L., Ottaway, J. M., and Fell, G. S., Analyst, 1979, 104,491. Seltzer, M. D., and Michel, R. G., Anal. Chem., 1983, 55, 1817. Demers, D. R., Busch, D. A., and Allemand, C. D., A m . Lab., 1982, March, 167. Michaud, E., and Mermet, J. M., Spectrochim. Acta, Part B, 1982, 37, 145. Paper J5l22 Received August 6th, 1985 Accepted September l l t h , 1985

ISSN:0267-9477    

DOI:10.1039/JA9860100045

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 51 Evaluation of a Continuously Variable Impactor for Use with Flame Atomisation Clare E. O'Grady, lain L. Marr and Malcolm S. Cresser Departments of Chemistry and Soil Science, University of Aberdeen, Meston Walk, Old Aberdeen AB9 ZUE, UK A critical assessment has been made of the extent to which an impact cup or bead may be used to regulate sensitivity in flame atomic absorption and emission spectrometry, particularly with a view to increasing the useful working ranges of calibration graphs. The impactors were located on adjustable arms in front of the nebuliser capillary tip. The changes in sensitivity due to impactor position were examined over a range of capillary tip to impactor distances for manganese and copper for both absorption and emission.The results show that the two impactors affect the absorption and emission sensitivities to different extents. This observation may be explained in terms of measured changes in flame temperature and in the amount of water reaching the flame. Aerosol droplet size distributions were measured and related to the interference effects at different impactor distances. Keywords: Atomic absorption; atomic emission; impact bead; impact cup; flame temperature A major drawback of atomic absorption spectrometry is the limited working range of calibration graphs. The main methods used to increase working range include burner rotation, to shorten the atom cell path length, and the use of an alternative less sensitive wavelength. 1 One main disadvan- tage of these methods compared with dilution is that, although the sensitivity is decreased, droplets with relatively high analyte and concomitant element concentrations lead to larger solid particulates after solvent evaporation, and thus to increased incidence and extent of incomplete volatilisation interference effects.Further, lines with suitable sensitivity are not always available and burner rotation, even through 90°, only extends the working range by about 8-fold. The impact cup has been suggested as an alternative way to increase useful working range.2 So far its use has been advocated at a fixed distance from the nebuliser, to give only a single equivalent dilution factor. The purpose of this work was primarily to evaluate how readily the impact cup and bead could be used to obtain a continuous range of equivalent dilution factors from normal, optimised sensitivity up to about 25-fold equivalent dilution, thereby greatly extending the normal linear working concentration range. It was expected that this should be achievable with no increased tendency towards incomplete volatilisation interferences, because of the small droplet size distribution of aerosol reaching the flame.2 Experimental A Baird A3400 atomic absorption spectrometer was used for all the experimental work.In addition to the standard bead support arm, which was about 20 mm long, a second, cup-support arm was inserted that was 190 mm long but with similar geometry. The cup position could be adjusted easily by pushing the arm in or out.The arm was marked with a millimetre scale so that the distance from the nebuliser to the cup could be read off. For all experiments the acetylene flow was kept at 2.6 1 min-1 and the air flow at 9.2 1 min-1. Measurement of Equivalent Dilution Factors The equivalent dilution factors (defined as the ratio of absorbance or emission signal with no impactor present to the measured signal) were measured for nebuliser capillary tip to impactor distances from 0 to 70 mm for the cup and from 0 to 12 mm for the bead. The bead diameter was 9.53 mm and the cup design was as shown in reference 2. Absorbance readings were taken at each distance for 1 yg ml-1 magnesium and 5 yg mll manganese solutions, and emission readings for a 1 pg ml-1 sodium solution for both the cup and bead.Measurement of Changes in Sensitivity Calibration graphs of copper and manganese for both absorp- tion and emission were obtained for the cup and bead over the range 0-70 mm. The slope was calculated over the linear portion of each calibration graph and this parameter was then plotted against distance. The fuel and air flow were as described above and the height of measurement above the burner was 4-9 mm for the absorbance measurements and 3-15 mm for the emission readings. Measurement of Interference Effects The extent of suppression of absorbance from 20 yg ml-1 of calcium by 20 pg ml-1 of phosphate in an air - acetylene flame was measured at impactor distances varying from 0 to 70 mm for the cup. Similarly, the measurements were made for the bead but with 10 yg ml-1 of calcium and 10 pg ml-1 of phosphate.The results were expressed in terms of the ratio of absorbance values in the presence and absence of phosphate and this ratio, converted to a percentage value (relative absorbance, YO), was plotted against impactor distance. Measurement of Droplet Size Distributions Droplet size distributions were measured using a Flow Sensor cascade impactor using the method as proposed by Cresser and Browner,3 aspirating 10 mg ml-1 sodium solution. The aerosol measured was that leaving the nebuliser - spray chamber - burner system from the Baird A3400 atomic absorption spectrometer. The set-up used was as described by Smith and Browner4 and is shown in Fig. 1. The sodium on each plate was washed off into 100-ml flasks and the concentration of sodium measured by flame emission spec- trometry using a Pye Unicam SP9 spectrometer.The droplet size distributions were measured at impactor distances cover- ing the range 3-70 mm for both the bead and the cup. The results were plotted as the sodium on each plate expressed as a percentage of the total sodium collected against cut-off diameter.52 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 Results and Discussion Equivalent Dilution Factors From Fig. 2, it can be seen that the cup gives a large range of equivalent dilution factors for all three elements, ranging from no dilution up to a relative dilution factor of 32-fold for sodium. At cup distances above about 15 mm, the graphs flatten out and increasing the distance to 70 mm causes only a slight further reduction in equivalent dilution factor.The positioning of the cup must be accurate because an error of 4 1 mm can, for small distances, reduce or increase the dilution factor by about 25%. The bead, on the other hand, only gives an equivalent dilution factor when it is positioned very close to the nebuliser, i.e., about 1 mm, but it gives an increase in sensitivity of around 2-fold when positioned further away. It is possible, therefore, to go from a 20-30-fold equivalent dilution to an increase in sensitivity of around 2-fold when both bead and cup are fitted. This was achieved in practice by having the impact cup on the long arm and the impact bead on the shorter arm and interchanging the two as required, without any need to extinguish the flame.The equivalent dilution factor from optimised sensitivity conditions is thus 40-60-fold. Changes in Sensitivity Fig. 3 shows how the signal per unit concentration changes for copper and manganese with the cup and bead for both emission and absorption. These graphs show that the influ- ence of bead or cup position is different in the two techniques, impactor Fig. 1. Aerosol measurement system using cascade impactor suggesting a possible flame temperature effect in emission. The results are summarised in Table 1. In typical analytical flames of temperature 2000-3300 K, the number of excited atoms varies exponentially with tempera- ture whilst the number of ground-state atoms remains essentially constant .5 Therefore, any slight temperature fluc- tuation will have a greater effect in flame emission than in atomic absorption.If the ratio of emission signal to absor- bance is plotted as a function of bead or cup distance, temperature effects should show up more clearly. This has been done for the two impactors in Fig. 4. It was decided to attempt to measure flame temperatures to see if changes could be detected to explain the difference between trends in absorption and emission sensitivity with impactor distance. Also, an experiment was conducted to measure the amount of water reaching the flame at various impactor distances. The flame temperatures were measured using the two-line ratio method with 100 pg ml-1 of iron at 344.06 and 371.99 nm. The intensities were measured eight times from spectral scans at bead distances of 3 , 5 , 10,25,40 and 70 mm and the mean temperature was calculated for each distance.The emission was measured 5-15 mm above the burner and a triangular collimator was used to restrict the area of flame examined to the interconal and lower secondary diffusion zones. Fuel and support gas flows were as described previously. To measure the volume of water reaching the flame, a rectangular container was made to give a tight fit around the burner head. The container was filled with silica gel. A second tube containing silica gel was used in series with the container to collect any further water. The silica gel was weighed before and after collection of the distilled water aerosol and the difference in mass gave the amount of water collected.The aspiration rate during these measurements was 0.0489 k 0.00014 ml s-1. Measurements were repeated for bead distances of 3, 5, 10,25, 40 and 70 mm. Fig. 5 shows how the amount of water varied with impactor - bead distance. The amount of water reaching the flame reaches a maximum at a bead distance of 5-10 mm and then decreases for distances up to 70 mm. From these results and those in Fig. 4, it might therefore be expected that the temperature would be at a minimum at a bead distance of 5-10 mm. The graph of flame temperature against bead distance (Fig. 5) suggests that the flame temperature does indeed reach a minimum at a bead distance of about 10 mm. However, as the large error bars in Fig. 5(b) indicate, the precision of the two-line flame temperature measurement is poor and the temperature changes are relatively small.The results never- theless appear to support the hypothesis that a temperature 0 10 20 30 40 50 60 70 5 t h Distance of cup from nebuliser/mrn Distance of bead from nebuliserhm Fig. 2. Mg; and C, Na Equivalent dilution factors for magnesium and manganese by AAS and sodium by AES for the cup (a) and for the bead (b). A, Mn; B,JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 0.9 0.8 5 0.7 'v, 0.6 .- > .- c C 0.5 & 0.4 n 0.3 a, 53 - - - - - - - , I , 0.040 0.032 0.024 0 '= 0.016 2 0.008 c C .- .I- 0 15 30 45 60 75 x 0.06 m C 0.05 0, * 0.04 0.03 C 3 Distance of CUD from L - .- C 0 .- .I- F .I- C a, C s .I- .- C 3 a, Q (0 c 0, v) L - .- l i 4 I Distance of bead from nebu liserimm 0 15 30 45 60 75 0 15 30 45 60 75 Distance of cup from ne buliserim m Distance of bead from nebu I iserim m Fig.3. Changes in sensitivity with impactor distance for copper (a, b) and manganese (c, d ) by AAS and AES for the cup (a, c) and bead (b, d). A, Absorbance; B, emission Table 1. Comparison of the ratios of the highest to the lowest slope of copper and manganese calibration graphs obtained with the cup or the bead by emission or absorption. EHT was kept constant throughout each set of emission readings with changing cup or bead distance for each element Bead (decrease) Cup (decrease) Element Emission Absorbance Emission Absorbance Mn . . . . 2.3 3.6 11.5 15 c u . . . * 1.4 2.5 8.3 13 0 Distanceimm +- 0 Distanceim m c 0 15 30 45 60 75 0 15 30 45 60 75 Distance/mm Distance/mm Fig.4. Changes in the ratio of emission intensity to absorbance with impactor distance for manganese (a, c) and copper (b, d ) for the cup (a, b ) and bead (c, d ) g 2100 E $ 1950 1900 0 0.10 > 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Distance of bead from nebuliserimm Fig. 5. Volume of water reaching the flame at any particular bead distance (a) and flame temperature at any particular bead distance (b) 100 8 G u 95 2 90 s .; 80 m 85 a m - U 15 JU 45 bU 15 0 15 30 45 60 75 Distance of cup from ne bu I i serim rn Distance of bead from nebuliserimm Fig. 6. Interference effect for the cup (a) and bead ( b ) of phosphate on calcium. Relative absorbance is the ratio of absorbance in the presence of phosphate to absorbance in the absence of phosphate expressed as a percentage shift caused by water loading of the flame may explain the observed trends.No temperature measurements were made with the impact cup because of the relatively low atomic emission intensity. However, the same mechanism could be important for the cup in that as the cup - nebuliser distance is increased and more water reaches the flame, there is a relative decrease in emission sensitivity coincident with a decrease in temperature. Fig. 4 suggests that small amounts of water may apparently enhance the flame temperature. Absorbance is less tempera- ture dependent than emission and therefore is not signifi- cantly affected. Interference Effect and Droplet Size Distributions Fig. 6 shows that the maximum suppression of calcium absorbance by phosphate occurs at a bead distance of around 5-10 mm, which is the distance at which most water reaches the flame.If the droplet size distributions for the bead are also considered (Fig. 7), it is notable that many large droplets are produced at a bead distance of 10 mm. As the bead is moved further away than 10 mm from the nebuliser, there is less interference, but the level of interference increases again towards a bead distance of 70 mm. Again, the droplet size distributions would help to explain this pattern, in that fewer large droplets were produced at 25 and 40 mm, but more large droplets were found at 70 mm. It is also probable that at distances beyond that at which the water maximum occurs (5-10 mm) the temperature increases and the extent of interference is reduced, except when the occurrence of larger droplets in the flame increases the probability of interference.The impact cup is very successful in reducing the extent of interference at cup distances up to 9-10 mm, in that 100% of the interferent-free absorbance was observed. After that distance the interference increased steadily until it was similar to the interference observed with the bead at 70 mm. Again, this can be related directly to droplet size distributions as shown in Fig. 7, in that as the distance increased the contribution from larger droplets also increased.54 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY. FEBRUARY 1986, VOL. 1 60 50 40 8 $ 30 * 20 .- -0 10 0 50 6o 30 .- 73 * 20 10 0 0 2 4 6 8 10 0 2 4 6 8 10 Droplet diameter/pm Fig.7. Droplet size distributions for the cup (a) and bead ( b ) at distances of 5 (e), 10 (0), 25 (U) and 70 (0) mm Conclusions By careful use of the bead and cup it is possible to achieve a range of sensitivities from a doubling of the signal to a 20-30-fold equivalent dilution compared with no impactor present. A more accurate procedure would be possible with a micrometer screw drive so that the distance could be measured to within k0.2 mm. However, simply sliding the cup to give the required signal is simple, inexpensive and adequate for most situations. The difference in change of sensitivities with distance for absorption and emission is probably due to flame temperature effects caused by changes in the amount of water reaching the flame.Although the method for measuring flame temperature has been reported to be inaccurate, with the precision only being 50-200 K,6 temperature measurements tend to support the hypothesis that large amounts of water cool the flame more. It is possible to remove interference effects completely with the impact cup, but at the cost of reduced sensitivity. Even at a cup distance of 25 mm, where relatively more water is reaching the flame, the extent of interference is reduced considerably compared with the situation when the bead is used. The extent of interference can be related to the droplet size distributions, flame temperature and the amount of water reaching the flame. By use of the cup and bead together, the tedium of diluting or concentrating samples could be largely eliminated. The authors are indebted to the Science and Engineering Research Council for financial support, and to Professor John Ottaway for valuable discussion of the flame temperature measurements. References 1. Price, W. J . , “Analytical Atomic Absorption Spectrometry,” Heyden, London, 1972, Chapter 4. 2. Cresser, M. S . , Analyst, 1979, 104, 792. 3. Cresser, M. S., and Browner, R. F., Spectrochim. Acta, Part B , 1980, 35, 73. 4. Smith, D. D., and Browner, R. F., Anal. Chem., 1982,54,533. 5 . Thompson, K. C., and Reynolds, R. J . , “Atomic Absorption, Fluorescence and Flame Emission Spectroscopy. A Practical Approach,” Second Edition, Griffin, London, 1978, Chapter 8. Elder, M. L., and Winefordner, J. D., Prog. Anal. At. Spectrosc., 1983, 6 , 293. 6. Paper J5l9 Received June 19th, 1985 Accepted August 30th, 1985

ISSN:0267-9477    

DOI:10.1039/JA9860100051

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 55 Improvement of a Pneumatic Nebuliser for Atomic Absorption Spectrometry* Barry T. Sturman Varian Techtron Pty. Limited, 679 Springvale Road, Mulgrave, Victoria 3 7 70, Australia The importance of the liquid transport characteristics of atomic absorption nebulisers is discussed and a simple method for determining these characteristics reported. This method was used to optimise the pumping characteristics of a commercial pneumatic nebuliser. Compared with the original nebuliser, the optimised version is significantly less sensitive to variations in the level of sample liquid (approximately a factor of five improvement). The effects of the nebulising gas pressure and sample uptake rate are presented and the improved analytical performance of the new nebuliser is shown.Keywords: Nebulisers; flame atomic absorption; pumping characteristics The precision and accuracy of flame atomic absorption spectrometry depend critically on the performance of the nebuliser and spray chamber. A fundamental requirement is that the rate of supply of solution to the nebuliser remains constant for all the standards and samples in an analysis. In commercial flame atomic absorption instruments it is current practice for the pneumatic nebuliser to also act as a pump, transporting liquid from the sample vessel through a length of flexible sampling capillary tube and the nebuliser capillary into the venturi where the primary nebulisation takes place. If the nebuliser does not perform adequately as a pump, the uptake rate may vary with the level of sample liquid relative to the nebuliser.As the absorbance typically varies with the uptake rate, this can result in solutions of the same concentra- tion giving different absorbances, depending on the level of liquid in the vessel. While much work has been reported on the characterisation of aerosols produced by atomic absorption nebulisers,l the pumping characteristics of these nebulisers have received less attention.2-4 O’Grady et a1.2 evaluated three methods for the measurement of “nebuliser suction ,” and found that under typical operating conditions the suction (the pressure differ- ence between the end of the capillary in the venturi and the atmosphere) varied with the liquid uptake rate permitted by the dimensions of the sample capillary tube.This paper reports a simple and convenient procedure for determining the variation of the pressure difference deve- loped by the nebuliser with uptake rate. This procedure has been used to optimise the pumping characteristics of a commercial pneumatic nebuliser , resulting in significant improvements in analytical performance. Theory It is well established that at the flow-rates typical of atomic absorption nebulisers, the relationship between the rate of liquid flow in a narrow tube and the pressure difference between the ends of the tube is given by Poiseuille’s Equation xR4P 800qL * . * ‘ * Q=- * (1) where P = pressure difference in Pa; q = viscosity in poise; R = radius in mm; L = length in mm; and Q = liquid flow-rate in rnl s-1. As shown by O’Grady et ul.,* it cannot be assumed that the pressure difference P across the capillary of a pneumatic * Presented in part at the 36th Pittsburgh Conference and Exposition on Analytical Chemistry and Applied Spectroscopy, New Orleans, LA, USA, March 1985. nebuliser is independent of the uptake rate. To optimise the pumping characteristics of atomic absorption nebulisers it was necessary to determine the relationship between the pressure difference developed by the nebuliser and the uptake rate. For a given capillary tube and a sample of viscosity q it is convenient to define . . . * (2) 8OOq L a=- nR4 * * ’ ’ so from equation (1) P Q=- . . . . . . * (3) a The quantity a is the resistance to flow resulting from the tube dimensions and sample viscosity, and is subsequently referred to as the hydrodynamic resistance.To calculate the hydrodynamic resistance for a nebuliser, it is necessary to know the viscosity of the sample liquid and the dimensions of both the capillary in the nebuliser and the sampling capillary. With the capillary diameters used in this study, pressure losses at the joint of the two capillaries were insignificant and the total hydrodynamic resistance was calculated as the sum of the resistance of each of the two capillaries. The viscosity of the distilled water samples was obtained from the temperature of the samples using standard published tables.5 Having determined the hydrodynamic resistance and measured the uptake rate, one can calculate the pressure difference developed by the nebuliser from equation (3).Repeating the procedure with a number of sample capillaries of different dimensions then reveals the relationship between uptake rate and pressure drop for the nebuliser. Experimental Nebulisers The design and construction of the Varian nebulisers used in this work has been discussed previously.6 The nebulisers are available as an adjustable uptake model, in which the position of the nebuliser capillary in the venturi can be adjusted by the user, and as a fixed uptake model in which the position of the nebuliser capillary in the venturi is not adjustable.6 The improvements in performance reported here were obtained by optimising the position of the capillary in the venturi, as will be explained later. Liquid Uptake Rate Measurements To measure uptake rates unaffected by hydrostatic head effects we used the apparatus shown in Fig.1. As indicated,56 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 Burette- R calibrated Water 7 1 I absorption c( NehtiliSer Ca pi1 la ry Fig. 1. Apparatus for measuring the uptake rate of the nebuliser 60 I I 0 I 2 3 4 5 6 7 8 9 1 0 Uptake rate/rnl min-’ Fig. 2. Relationship between uptake rate and pressure difference developed by the nebuliser for two nebuliser settings: A, original setting for the “fixed uptake” version6; and B, new maximum pressure drop setting the modified calibrated flask was located so that the gradua- tion mark was at the same height as the platinum - iridium capillary in the nebuliser. For each measurement, the flask was initially filled with distilled water to a level above the graduation mark.Timing commenced as soon as the water level reached the graduation mark and water was added from the burette to keep the level constant during the timed period. At the end of the timed period the burette was turned off and the volume of liquid added was read. Uptake rates were varied by using different lengths of 0.6 mm i.d. and 0.4 mm i.d. polyethylene tubing. The mean diameters of these capillary tubes were calculated from the mass of water contained in a known length. The pressure difference developed by the nebuliser at zero uptake rate was measured by connecting a U-tube mercury manometer directly to the platinum - iridium capillary. Atomic Absorption Measurement Analytical studies were carried out using a Varian SepctrAA 40 atomic absorption spectrometer equipped with a Varian PSC-56 sample changer. Operating conditions were as set by the instrument software.The glass impact bead position and burner orientation were adjusted to obtain maximum absor- bance for each element. Results and Discussion The original capillary setting for the Varian fixed uptake rate nebuljser is shown by Howarth et al. in Figure 2,A of their paper.6 The plot of the pressure difference developed by the nebuliser versus uptake rate for this nebuliser is shown in Fig. 2. The pressure difference developed by this nebuliser did not vary greatly with uptake rate. At the recommended uptake rate of 4.5-5 ml min-1 the pressure difference developed by the nebuliser was close to 9.5 kPa.The pressure difference resulting from a hydrostatic head change of 10 cm of water is 0.98 kPa, which is a significant fraction of the total pressure difference developed by the nebuliser. Clearly, it would be advantageous to increase the pressure difference at normal uptake rates so that the pressure differences resulting from changes in the liquid level in sample vessels were a smaller fraction of the total pressure drop across the nebuliser capillary. The position of the capillary in the venturi throat was adjusted until the pressure drop was at a maximum. This setting is shown by Howarth et al. in Figure 2,B of their paper.6 The relationship between pressure drop and uptake rate for this modified nebuliser is shown in Fig. 2.The pressure drop developed at uptake rates in the range 3-5 mi min-1 was at least a factor of three greater than with the former setting of the nebuliser, and the pressure drop decreased very much more rapidly with increasing uptake rate. It was found that the relationship between pressure drop P and uptake rate Q shown in Fig. 2 could be described by where fi and y are constants for a given nebuliser driven at a specified nebulising gas pressure and Po is the pressure drop at zero uptake rate. For a given hydrodynamic resistance a, by definition from equation (3), so from equation (4) P=P,-cjQ-yQ2 . . . . . . (4) P = aQ (5) or Note that when P and y = 0 equations ( 5 ) and (6) are equivalent to Poiseuille’s equation. From equation (6), ignoring the physically meaningless negative solution, Q = ye’+ (P+ a)Q-P()=O .. . . (6) . . . . ( 7 ) [(P + a> + 4Ypo1°.5 - (fi + a) 2Y Equations (5) and (7) proved to be very useful for establishing nebuliser parameters. Equation ( 5 ) was used to find the hydrodynamic resistance required to obtain a specified uptake rate, and hence to select the required dimensions for the sampling capillary for the nebuliser, given the dimensions of the platinum - iridium capillary in the nebuliser. Equation (5) can also be used to predict the effect of changes in sample viscosity on uptake rate. As indicated in equation (2), in a given system the hydrodynamic resistance cx is proportional to the sample viscosity 7 . By differentiating equation ( 5 ) it is easy to show that the fractional change in uptake rate for a given fractional change in resistance is If the pressure drop is independent of uptake rate, fi and y = 0 and d& da .. * . . . . . -- _ - - Q a (9) When S and y are positive, as they are for the modified nebuliser (Fig. 2, line B) <1 Po - PQ - YQ’ Po + and from equations (8) and (9) it is clear that the fractional change in uptake rate for a given fractional change in sample viscosity is less than it would be if the pressure drop developed by the nebuliser were independent of uptake rate.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986. VOL. 1 57 Table 1. Hydrostatic head effects in the analysis of a 5 mg 1-1 copper solution, as explained in the text Average error (Yo) = (mean result for 5-mi tubes - mean result for full tubes) x 100 mean result for full tubes Nebuliser Average error, Yo O1d“fixed” .. . . . . -5.4 0ld“variable” . . . . . . -2.4 New . . . . . . . . -0.6 Table 2. Analysis of 5 mg I-’ copper in 5% sodium chloride solution, as explained in the text mean reported result - 5 mg 1-1) 5 mg 1-1 Error = ( Relative Samples standard Nebuliser analysed deviation, YO Error, Yo Old “fixed” . . 15 1.1 -0.42 New . . . . 26 0.7 -0.24 Equation (7) can be used to predict the effect of hydrostatic head changes on uptake rate. The pressure drop at zero uptake rate is given by Po = PA - P N where PA is the external pressure (i.e., atmospheric pressure) and PN is the pressure in the nebuliser venturi at the exit of the capillary tube (at zero uptake rate). From equation (lo), it is obvious that any change in atmospheric pressure PA will result in the same change in Po.For the purposes of calculation, hydrostatic head effects may be treated as changes in atmospheric pressure and hence as changes in Po. For a given system the hydrodynamic resistance a is constant and and y are assumed to be constant for small changes in Po. With constant a, B and y, and differentiating with respect to Po, . . . . . . (10) Calculations of the expected hydrostatic head effect with the new nebuliser setting indicated a significant improvement over the old setting. To investigate this experimentally we took ten test-tubes filled with a 5 mg 1-1 copper solution and placed them in a Varian PSC-56 sample changer alternated with ten test-tubes containing only 5 ml of the same test solution.Copper was used as the test element because the absorbance (at uptake rates near 4 ml min-1) was known to be significantly altered by changes in the uptake rate. The hydrostatic head difference 1 .o 0.9 0.8 0.7 & 0.6 9 2 0.5 n 0.4 0.3 0.2 0.1 0 0 1 2 3 4 5 6 7 8 9 1 0 Uptake ratelm1 min-’ Fig. 4. Effect of uptake rate on the absorbance in the air - acetylene flame for: A, copper ( 5 mg 1-1); B, chromium ( 5 mg I - I ) ; and C, magnesium (0.2 mg 1-1). Uptake rates were altered by using different lengths or diameters of polyethylene sampling capillary 0.6 I 0.5 8 0.4 C ((1 + 2 a 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 1 0 Uptake rate/ml min-’ Fig. 5. Effect of uptake rate on the absorbance in the nitrous oxide - acetylene flame for: A, calcium (1.5 m 1-1 + 1 g 1-1 of potassium); B, aluminium (50 mg 1-1); and C, silicon 800 mg 1-I).Uptake rates were altered by using different lengths or diameters of polyethylene sampling capillary between a full tube and one with only 5 ml of sample was 12.5 cm of water (1.23 kPa). This was very close to the maximum hydrostatic head difference that might occur in the use of this sample changer. The spectrophotometer was calibrated using a single standard of 5 mg 1-1 of copper in a full test-tube and the 20 “samples” were analysed automatically. The test was carried out three times, with the same nebuliser set up first as the fixed uptake version as indicated by Howarth et al. in Figure 2,A of their paper,6 then as the adjustable uptake version (Figure 2,C in reference 6) and finally with the new maximum pressure drop setting.The uptake rates of all three configurations were very similar (3.8-3.9 ml min-1). This was achieved either by choice of appropriate capillary dimensions or (with the variable nebuliser) by adjusting the uptake rate control. Results of these tests are shown in Table 1. A worthwhile improvement in the accuracy of the analysis was achieved with the new nebuliser setting. A potential probelm with any alteration of a nebuliser is a possible increase in the tendency of the nebuliser to clog when solutions containing high levels of dissolved solids are analysed. To investigate this, we carried out an automated analysis of a test solution of 5% mlV sodium chloride solution58 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 containing 5 mg 1-1 of copper, using the nebuliser in the old fixed uptake configuration and with the new maximum pressure drop setting. The variable uptake version is not recommended for samples with high levels of dissolved solids. The single calibration standard was the 5 mg 1-1 of copper in 5% sodium chloride test solution in a full test-tube. In each instance the automated analysis was carried out using a series of test-tubes filled with the solution, eliminating any hydros- tatic head effects. The sample number at which there was any visible disruption of the flame by clogging of the burner was noted. Analysis was continued until the reported result was below the lowest result reported before the flame was visibly disrupted by clogging.The number of samples analysed and the standard deviation of the results are shown in Table 2, together with the average percentage error. In both the old and the new configuration the nebuliser did not clog. The duration of the test was limited by the clogging of the burner slot with salt deposits. It was pleasing to find that the new nebuliser version not only failed to clog but also processed more samples with marginally better precision and accuracy than was obtained with the nebuliser set in the old configuration. Having found useful improvements in analytical perfor- mance with the new nebuliser configuration, we decided to develop a commercial version. We carried out detailed studies of the effect of oxidant gas pressure and of solution uptake rates on the performance of the new nebuliser.The effect of oxidant gas pressure on pressure drop and uptake rate with a given hydrodynamic resistance is shown in Fig. 3. A maximum occurred close to 200 kPa oxidant gas pressure. Analytical studies showed that maximum signal to noise ratio was also obtained at ca. 200 kPa oxidant gas pressure, and this pressure was used for all subsequent work. The uptake rate of the new nebuliser is controlled by the hydrodynamic resistance of the capillary system. The effect of varying the uptake rate (by using different capillaries) on the absorbance of some elements measured in the air - acetylene flame is shown in Fig. 4. While the absorbance of the relatively easily atomised element copper continued to increase with increasing uptake up to at least 8 ml min-1, the more refractory element chromium showed no increase in response above 5 ml min-1 and a decrease in response above 7 ml min-1.Use of high uptake rates is undesirable because of increased clogging of the burner and increased risk of interferences, and, as shown here, may not even lead to an increase in the absorbance signal. The levelling out and eventual decrease of absorbance at higher uptake rates could perhaps be the result of overloading the flame with water and so decreasing the efficiency of atomisation. Another contri- buting factor could be a change in the droplet size distribution of the aerosol at higher uptake rates, resulting in a greater proportion of large droplets reaching the flame. As shown in Fig. 5 , the effect was even more severe with elements measured in the nitrous oxide - acetylene flame.With the three elements shown, the effect was least for calcium , which is relatively easily atomised and less sensitive to flame conditions. The effect was very pronounced for silicon, which is difficult to atomise and extremely sensitive to flame conditions. These tests, and others investigating the effect of uptake rate on precision, led to the conclusion that an uptake rate of between 4 and 5 ml min-1 was the most appropriate for general analytical use. This uptake rate was achieved by using a 238 mm x 0.38 mm i.d. polyethylene sampling capillary. Other uptake rates could of course be obtained by using a longer (or narrower) capillary for a decreased uptake rate and a shorter (or wider) capillary for an increased uptake rate. For the reasons stated previously, we do not recommend the use of uptake rates above 5 ml min-1. At the recommended uptake rate (4-5 ml min-1) the sensitivity and precision of the modified nebuliser were very similar to those reported for the original “fixed” version.6 Conclusion We have developed a simple and convenient procedure for characterising the liquid transport or pumping properties of atomic absorption nebulisers. Use of this procedure has led to the modification of a commercial pneumatic nebuliser, resulting in worthwhile improvements in analytical perfor- mance. The author wishes to thank Ted Rothery for suggesting this approach to measuring the pumping characteristics of nebu- lisers and for helpful discussions. References 1. 2. 3. 4. 5. 6. Browner, R. F., and Boorn, A. W., Anal. Chern., 1984, 56, 786A, 87.5A, and references therein. O’Grady, C., Marr, I. L., and Cresser, M. S . , Analyst, 1984, 109, 1085. O’Grady, C. E., Marr, I. L., and Cresser, M. S . , Analyst, 1984, 109, 1183. O’Grady, C. E., Marr, I. L., and Cresser, M. S . , Analyst, 198.5, 110, 431. Weast, R. C., Editor, “CRC Handbook of Chemistry and Physics,” CRC Press, Boca Raton, FL, 1982, p. F40. Howarth, H., McKenzie, T. N., and Routh, M. W., Appl. Spectrosc., 1981,359, 164. Paper J5/8 Received June 17th, 1985 Accepted August 15th, 1985

ISSN:0267-9477    

DOI:10.1039/JA9860100055

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 59 A Comparison of Cloud Chambers for Use in Inductively Coupled Plasma Nebulisation Systems Leslie S. Dale and Stephen J. Buchanan CSIRO, Division of Energy Chemistry, Lucas Heights Research Laboratories, New lllawarra Road, Lucas Heights, NSW, 2234, Australia A comparison of the analytical performance of five cloud chambers of different geometry used in conjunction with a commercial concentric glass nebuliser is reported. The comparison was based on measurements of transport efficiency, background equivalent concentration, analyte emission stability, detection limit, equilibration time and memory effect. Significant variations in performance indicated the importance of cloud chamber geometry on the nebulisation system.Best results were obtained with a new design based on a cylindrical chamber with a central tangential inlet. Droplet size distribution measurements revealed that although the primary distribution of the nebuliser aerosol contained a substantial proportion of small droplets of similar size to those in the ultimate distributions of the nebulisation systems, a large number of droplets are removed from the aerosol by separation processes occurring in the chambers. Keywords: Cloud chambers; nebulisation system; inductively coupled plasma; atomic emission spec- trometry The major advantages of inductively coupled plasma atomic emission spectrometry (ICP-AES) are the high analytical precision achieved as a result of the emission stability of the source and the high detection capability for a wide range of elements as a result of its high temperature.However, these properties are influenced by the performance of the nebulisa- tion system. Analytical precision is determined by the stability of analyte emission, and hence by the stability of aerosol transport rates, and sensitivity is determined by the transport efficiency. Therefore, the desired characteristics of the nebulisation system, consisting of a nebuliser and cloud chamber, are the production of a stable aerosol stream of small droplets with a narrow size distribution and a high transport rate. Commercially available nebulisers, including the concentric glass, cross-flow and those based on the Babington principle, produce aerosols with broad droplet size distributions; it is necessary to remove the larger droplets using a cloud chamber as a separator or filtering device.Consequently, the cloud chamber must be treated as an important component when assessing the performance of the nebulisation system. Although there have been a number of comparisons of the Performance of nebulisers used in ICP-AES,1-4 the effect of cloud chamber geometry in nebulisation systems has received less attention. Ebdon and Cave5 compared a Scott double- pass cylinder6 and a cyclone chamber using a commercial concentric glass nebuliser and a cross-flow nebuliser of their own design. The signal to noise ratio for aluminium was up to 30% higher with the cyclone chamber although condensation occurred in the injector tube of the torch. Olsen et al.7 carried out an extensive investigation of nebulisers and showed how the droplet size distributions were altered by different cloud chambers (cyclone, Scott single and double-pass cylinders).They concluded that chamber geometry was an important factor in achieving high signal to noise ratios. Novak and Browner8 compared a chamber used in atomic absorption spectrometry and a concentric tube chamber with a variety of nebulisers. Their measurements of droplet size distributions of a number of nebuliser - cloud chamber combinations demonstrated that the cloud chamber effectively reduced the mass distribution of the aerosol, and that chamber geometry was a critical factor in achieving high signal to background and signal to noise ratios. Because cloud chamber geometry has such a significant effect on the performance of the nebulisation system, a more comprehensive comparison of different cloud chambers was undertaken to examine, in detail, its effect on analytical performance.This was achieved by studying the characteris- tics of a number of chambers when used with the same nebuliser operating under identical conditions. Five cloud chambers were selected: the Scott double-pass cylinder, a conical chamber with an impact bead, two cyclones of 250- and 500-ml capacity and a cylinder of 180-ml capacity with a central tangential inlet. Selection was made on the basis of published performance data, use in commercial instruments or personal laboratory experience. The chambers had basic differences in their geometries with respect to shape, volume, impact surface and mode of aerosol separation.Although the effect of cloud chamber geometry on analyte emission stability and sensitivity are the most important aspects in comparing analytical performance , equilibration time and memory effect are also worthy of investigation. Rapid equilibration times and short washout times maximise the sample throughput rate and the former leads to reduced sample consumption, which may be of significance in sequen- tial instruments. The assessment of analytical performance was therefore based on measurements of transport efficiency, background equivalent concentration, analyte emission stab- ility, detection limit, equilibration time and memory effect. Droplet size distributions were also measured to assess the performance of each nebulisation system in terms of their ultimate mass distributions.Experimental All measurements were made on an ICP spectrometer based on a Labtest Plasma 2000 generator and matchbox. Light from the plasma (2 mm aperture, 16 mm above the load coil) was focused on to the entrance slit of a 0.5-m Ebert monochroma- tor fitted with a 2160 grooves mm-1 grating blazed at 260 nm. The reciprocal linear dispersion was 0.8 nm mm-1. Emission signals from the photomultiplier (EMI6256S) were measured using a locally built STD bus-based microcomputer with a Motorola 6809 microprocessor. The operating power level was 1 kW. The TR-30-A3 concentric glass nebuliser was obtained from J . E. Meinhard Associates, Santa Am, CA, USA. It was operated at a gas flow-rate of 1 1 min-' and 207 kPa (30 lb in-2) pressure.The solution uptake rate was 2.7 ml min-1. The nebulisation system was located outside the60 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 torch chamber and connected to the base of the torch by a 60-cm length of polythene tube. Cloud Chambers All cloud chambers were built in the laboratory, with the exception of the conical chamber, which was taken from an ARL instrument. The cyclone chambers were constructed from Quickfit Erlenmyer flasks; they were scaled-down versions of the 1-1 capacity chamber originally used in Unicam SP900 flame spectrophotometers. The smaller volumes were chosen on the basis of previous experience with them in flame spectrometry. The design concept for the 180-ml cylinder emerged during the course of this work; details are shown in Fig.1. Measurement of Transport Emciency A direct aerosol collection procedure similar, in principle, to that described by Smith and Browner9 was used. A fibre filter plug was used to collect the aerosol at the end of the delivery tube. Triplicate measurements of each nebulisation system were made, the nebuliser uptake rate being checked before each measurement. Other Measurements Background equivalent concentrations, emission stabilities and detection limits were measured for chromium (202.5 nm), iron (238.2 nm) , manganese (257.6 nm) , aluminium (308.2 nm), zirconium (343.8 nm) and strontium (407.8 nm). The background equivalent concentrations, expressed as the concentration of the element giving rise to a signal equal to the background level, were determined using signals from solu- tions with concentrations three to five times higher than the background level.Emission stabilities were obtained by n f Drain Fig. 1. Schematic diagram of cylindrical cloud chamber. All dimensions in millimetres calculating the relative standard deviation of ten 5-s integra- tions of signals from solutions with analyte concentrations three to five times higher than the background level. Detection limits were based on twice the standard deviation of the background (ten 5-s integrations) and solution concen- trations that were approximately 50 times higher than the detection limit. Equilibration times were determined by measuring, on a chart recorder, the time taken for a signal from a 0.05 pg ml-1 Sr solution to reach 95% of its steady value.From these chart records the delay times, that is, the time elapsed before any response was detected, were calcu- lated. The memory effect was determined by measuring the time taken for the signal from a 5 pg ml-1 Sr solution, run for 5 min, to decay to a signal equivalent to 0.005 pg ml-1 Sr ( i . e . , 0.1% of the original). Equilibration and memory effect measurements included the nebuliser solution uptake time. The above quantities were measured randomly in triplicate under identical operating conditions and then averaged. Droplet Size Distributions These were carried out on a Malvern Type 3300 particle sizer, which is based on the laser diffraction principle. The aerosol was fed through polythene tubing to a windowless cell situated in the optical path of the instrument.The mass distribution of the nebuliser alone was measured by directing the aerosol across the optical path at the cell position. The distribution was measured at a number of locations in the aerosol cloud at distances of 3-10 cm from the nebuliser nozzle. Results Data for transport efficiencies, background equivalent con- centrations, analyte emission stabilities and detection limits are shown in Table 1. The transport efficiencies could be reproduced to about 5%. The efficiency of the filter as an aerosol particle collector was checked by monitoring the plasma emission, with the filter in the delivery line, while aspirating the test solution. Some of the aerosol passed through the filter but the amount lost was insignificant (ca.1%). The procedure was therefore satisfactory for determin- ing transport efficiencies and allowed useful comparisons to be made. To facilitate comparison with the other data listed, average values for the six elements relative to the 180-ml cylinder are shown. These provided a good estimate of the over-all performance of each nebulisation system. Table 2 shows the equilibration and delay times obtained. Washout times are shown in Table 3, together with the equivalent number of volume changes that took place while the signal decayed to the specified level. The results of the droplet size distribution measurements are shown in Table 4. The mass median diameter is the droplet size below which 50% of the aerosol mass occurs.Data for the mass fractions of droplets below 1.2, 5.0 and 10.5 km give an indication of the shape of the distributions. The droplet size distribution of the nebuliser alone had an average mass Table 1. Transport efficiencies and average relative background equivalent concentrations, analyte emission stabilities and detection limits Transport Background Analyte efficiency, equivalent emission Detection Cloud chamber % concentration* stability* limit* Cylinder (180 ml) . . . . . . 1.26 1 .o 1 .o 1 .0 Cyclone (250 ml) . . . . . . 1.21 1.1 1.4 1.9 Cyclone (500 ml) . . . . . . 0.91 1.5 2.4 2.3 Cone (50 ml) . . . . . . . . 0.52 2.1 2.3 3.6 Double-pass cylinder (250 ml) . . 0.39 2.9 2.9 4.1 * Values are relative to those for the 180-ml cylindrical chamber.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 61 median diameter of 18.4 pm with 11.7% m/m of the droplets being below 10.5 pm. Discussion Transport Efficiency The observation that cloud chamber geometry changed the transport efficiency by as much as a factor of three is indicative of the importance of this factor to the nebulisation system. To explain these variations in transport efficiency, reference is made to the aerosol transport model of Browner et al. , l o who described the aerosol in terms of its primary, secondary and tertiary distributions. The primary droplet distribution of the nebuliser is modified by impingement of the aerosol on an impact surface where larger droplets are removed, small droplets follow the gas stream and intermediate droplets of sufficiently high velocity may shatter and generate smaller droplets.The aerosol then assumes its secondary droplet size distribution. Beyond the impact surface, further separation occurs through processes of impaction, turbulence, centrifugal loss, gravitational settling and evaporation. This is the tertiary distribution that is transported to the plasma. Based on this model, differences in transport efficiencies may be explained. In the cylinder and cyclone chambers, for which the highest values were obtained, the droplets which remain after initial impaction with the chamber walls follow the circular path of the gas stream when larger droplets are removed by centrifugal loss. There is little visual turbulence in these chambers.In contrast, the conical chamber and double- pass cylinder, which yielded lower efficiencies, have different impact surfaces from the other chambers. In the conical chamber impaction of the aerosol on the bead, located in a direct line with the nebuliser nozzle, produces considerable condensation on its surface. Some aerosol turbulence was visible in the region between the nozzle and the bead, which suggests further separation by this process. In the double-pass cylinder, the impact surface is the relatively large area of the inner cylinder, and the end of the outer cylinder. Considerable Table 2. Equilibration times for various cloud chambers Equilibration Cloud chamber timels Delayls Cylinder . . . . . . 11 4 Cyclone (250 ml) . . . . 19 3 Cyclone (500 ml) .. . . 25 4 Cone . . . . . . . . 9 2 Double-pass cylinder . . 28 9 Table 3. Washout times for various cloud chambers No. of volume Cloud chamber Volume/ml Timels changes Cylinder . . . . . . 180 33 3.1 Cyclone . . . . . . 500 110 3.6 Cone . . . . . . 50 38 12.7 Double-pass cylinder . . 250 63 4.2 Cyclone . . . . . . 250 53 3.5 condensation was visible up to half way along the inner cylinder. Turbulence was also obvious in this region. The probability of further separation by secondary impaction on the chamber walls is high, owing to the large surface to volume ratio. Variations in transport efficiencies may therefore be attri- buted mainly to the influence of the cloud chamber geometry on secondary and tertiary mechanisms induced by the shape and position of the impact surface.The major mechanisms that limit the transport efficiency in the conical chamber and double-pass cylinder appear to be condensation and, to a lesser extent, turbulence. In the cylinder and cyclone cham- bers, the higher efficiencies are due to the lack of turbulence, and the predominant tertiary mechanism appears to be centrifugal loss. Condensation on the impact surface must be lower as there is considerable drainage of waste solution on the chamber walls as a result of this centrifugal loss. Background Equivalent Concentration, Analyte Emission Stability and Detection Limit The 180-ml cylinder gave the best over-all performance. The background equivalent concentrations for this chamber and the 250-ml cyclone are only marginally different.This would be expected from their similar transport efficiencies, as this quantity depends on the amount of aerosol reaching the plasma. However, the significant differences in analyte emission stability and detection limit indicates a more stable aerosol mass transport rate with the 180-ml cylinder. For the other chambers, the background equivalent concentrations are in line with their lower transport efficiencies and their analyte emission stabilities result from less stable aerosol mass transport rates. This combined with lower background equi- valent concentrations, is responsible for the decline in detection limits. Equilibration Time Although the conical chamber had the fastest equilibration time, it was only marginally better than that of the 180-ml cylinder.However, allowing for the delay period, both chambers had about the same response time. The cyclone chambers had a relatively short delay time but their equilibra- tion times were long, A possible explanation for this is that because of the close proximity of the impact surface to the top outlet, some of the aerosol has a short residence time in the chamber. This portion of the aerosol therefore circumvents the system by entering the gas stream near this outlet. This explanation is supported by the fact that the 180-ml cylinder produced a longer delay time than the 250-ml cyclone. With the cylinder, the impact surface is the centre of the chamber wall and the aerosol paths to the outlets are longer, thus minimising this effect. The double-pass cylinder has a long delay time and a long equilibration time.This is probably due to the long aerosol path length and substantial mixing within the chamber. This effect is exemplified by the similarity of the times for this chamber and the 500-ml cyclone, although the volume of the cylinder was only half that of the cyclone. Table 4. Droplet size distribution data for various cloud chambers Mass median % mlm below diameter/ Cloud chamber Pm 1.2 pm 5.0 pm 10.5 pm Cylinder . . . . . . . . 5.43 4.3 28.4 83.9 Cyclone (250 ml) . . . . . . 6.78 2.7 16.1 67.2 Cone . . . . . . . . . . 3.91 5.1 49.4 90.4 Double-passcylinder . . . . 4.72 4.7 37.3 90.8 Cyclone (500 ml) . . . . . . 6.22 1.4 29.0 75.762 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL, 1 Memory Effect The 180-ml cylinder and the cone gave similar washout times, which were substantially better than the others.With the exception of the cone, the number of volume changes required to reach the 0.1% level was between three and four. On the basis of chamber volume alone, and taking the average number of changes of 3.6 obtained for the other chambers, the washout time for the conical chamber should have been about 10 s instead of the 38 s obtained. This suggests that the washout time for this chamber is prolonged by inefficient mixing. Droplet Size Distributions The measurements of droplet size distributions show that although both the mass median diameter and the size distribution were altered by the cloud chamber geometry, there was no correlation with any of the other parameters measured.Both the cone and the double-pass cylinder had lower mass median diameters and narrower distributions, but also they had the lowest transport efficiencies and poorer emission stabilities. The higher mass median diameters and broader distributions obtained with the cyclones are pro- bably due to the larger droplets contained in that part of the aerosol that circumvents the system. It can be seen, from the droplet size distribution data for the 180-ml cylinder, that placement of the nebuliser at the mid-point prevents this occurrence. Of most significance is the comparison between the droplet size distributions obtained with the chambers and that of the nebuliser alone. The latter distribution contained 11.7% mlm of the droplets below 10.5 pm whereas the cloud chambers had distributions in which 67-90% mlm of the droplets were below this value.As the highest transport efficiency obtained was only about 1.2%, up to 90% of droplets suitable for transport to the plasma were removed by separation mechanisms in the chambers. The cloud chambers therefore not only perform their desired function of filtering the larger droplets from the primary distribution, but also considerably reduce the popula- tion of those droplets suitable for transport to the plasma. Performance of Cylindrical Chamber It has been shown that the cylindrical chamber generally provides the best performance. This may be attributed to certain features of its design. The tangential inlet results in little turbulence, and separation of the aerosol particles is achieved largely by centrifugal loss.This leads to high transport efficiency and stability. The favourable equilibration and memory characteristics are due, in part, to its small volume and the central location of the nebuliser inlet. With this arrangement the two outlets reduce the aerosol residence time, leading to rapid response and short washout times. The droplet size distribution produced by this chamber is compar- able to those of the cone and double-pass cylinder while maintaining a higher transport efficiency. Prevention of aerosol particles from circumventing the system contributes to the narrow size distribution. Conclusion Cloud chamber geometry is an important consideration when optimising the analytical performance of a nebulisation system.Of the chambers studied, those based on centrifugal loss for separation provided best performance. The cylindrical chamber with a central tangential inlet gave the best over-all performance. The consistently high performance level of this chamber has been demonstrated by its continuous and routine use in these laboratories for over 1 year. A limiting factor in achieving higher transport efficiencies is the loss of a large proportion of droplets of suitable size range produced by the nebuliser . The authors gratefully acknowledge the skill of W. T. Williams, Scientific Glassblower, who fabricated the cloud chambers. They are also indebted to D. Shirlaw for carrying out the droplet size distribution measurements. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Gustavsson, A., Spectrochim. Acta, Part B , 1984, 39, 743. Wohlers, C. C., ZCP Znf, Newsl., 1977, 3, 2. Greenfield, S., McGeachin, H. McD., and Chambers, F. A., ICP Znf, Newsl., 1977, 4, 117. Garbarino, J. R., and Taylor, H. E . , Appl. Spectrosc., 1980, 34, 584. Ebdon, L., and Cave, M. R., Analyst, 1982, 107, 172. Scott, R. H., ZCPZnf. Newsl., 1978, 3, 425. Olsen, S. D., Strasheim, A., and Perry, A., paper presented at the 9th International Conference on Atomic Spectroscopy and the XXIII Colloquium Spectroscopium Internationale, September 4-8, 1981, Tokyo, Abstract No. 5All. Novak, J. W., and Browner, R. F., Anal. Chem., 1980,52,792. Smith, D. D., and Browner, R. F., Anal. Chem., 1982,54,533. Browner, R. F., Boorn, A. W., and Smith, D. D., Anal. Chem., 1982,54, 1411. Paper J5l15 Received July 15th, I985 Accepted August 2 7th, 1985

ISSN:0267-9477    

DOI:10.1039/JA9860100059

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 63 Flow Injection Atomic Absorption Spectrometry: The Kinetics of lnstru ment Response John M. H. Appleton and Julian F. Tyson Department of Chemistry, University of Technology, Loughborough, Leicestershire LEI 1 3TU, UK The concept of dispersion coefficient is discussed with particular reference to flow injection atomic absorption spectrometry where the detector contributes appreciably to the analytical signal characteristics. Single- and parallel-tank models of instrument response are developed and critically examined. The progress made to date by investigators of nebuliser performance is briefly reviewed prior to developing a semi-empirical extended-tank model of instrument response. The capabilities of this model are explored by deriving a set of equations for instrument response, and comparing the predictions with experimental results.Agreement is generally good. Advantages of the modelling approach are discussed. Keywords: Flow injection; atomic absorption; kinetics; instrument response Flow injection (FI)1 is an elegant and versatile technique that continues to attract increasing attention from analysts world- wide. Of fundamental importance in FI is an understanding of the process of dispersion of a sample in a carrier stream under conditions of laminar flow. By controlling dispersion, the analyst may manipulate small volumes of samples and reagents with speed, simplicity and precision to obtain analytical results more efficiently in terms of time, labour and materials consumed.The extent to which dispersion occurs in an FI manifold is usually quantified by the dispersion coefficient D, as defined by equation (1) - * (1) . . . . . . D=-- C m CP where C, is the original analyte concentration and C, is “the analyte concentration at the maximum of the peak.”2 (All symbols used are explained in Table 1.) More recently, in connection with the miniaturisation of FI apparatus, RfiiiEka and Hansen3 have introduced the concept of dispersion factor, defined as the volume required to give D = 2 divided by the Table 1. List of symbols used in text c . . . . h,h‘ . . k . . . . m . . . . p . . . . q . . . . Q . . . . t , t p . . . . t’,t’, * . u . . * . Urnax. * * v . . . . V’ . . . . vi . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . Absorbance; corresponding to Cm and C,, respectively Concentration (in a flowing steam, integrated across the entire stream normal to the flow) Instantaneous concentration input; injected and peak concentrations of stream Detector, manifold and response dispersion coefficients Volumetric fraction of stream flowing via path i Gradients of graph when t = 0 and 03 Detector responses corresponding to C, and Cp, respectively Value of ulV, ulV’ for hypothetical and real tanks, respectively Constant relating absorbance and concentration Analyte mass (in tank) In the model, the fraction of the sample stream contributing to the analytical signal Diluent flow through the hypothetical tank Total flow through hypothetical tank Time and time to reach peak response Peak width at constant height and lln of peak Volumetric flow-rate of carrier, sample Flow-rate giving highest peak for given Vi Volume of hypothetical tank forming basis of Volume of real mixing tank Sample volume injected height model volume of the flow line.Whichever parameter is used to quantify dispersion, Cp has to be known. However, the transient concentration Cp can only be determined indirectly. For an ideal detector (one which accurately reproduces the concentration profile of the deter- minand entering the detector) equation (1) may be extended to yield where Hm and Hp are the instrument responses corresponding to C, and Cp, respectively. Thus, although the dispersion coefficient, an important FI parameter, has been defined by the concentration ratio C,/C,, it can only be determined from the response ratio HmIHp.For many FI techniques, this constitutes a valid and convenient method of determining dispersion coefficient. However, despite an increasing number of papers involving flow injection atomic absorption spectrometry (FI-AAS) , little has been said about the error of extending the practice to this field. This application of the concept of dispersion coefficient therefore requires clarification , particularly for the benefit of newcomers to this promising extension of atomic absorption spectrometry (AAS). Certain difficulties arise when detector response is non-ideal, as is frequently so in Detectors for FI are usually flow-through types, designed to cause minimum disturbance of the flowing stream, so that detector contribution to the over-all signal is negligible (i.e., the dispersion coefficient due to the detector, Dd, is unity). The performance of such detectors is close to ideal when the detector response is linear with respect to concentration and rapid. Used as an FI detector, the atomic absorption spectrometer has neither of these qualities. Orderly sample flow is totally disrupted during nebulisation to create an aerosol suitable for flame atomisation. The resulting analy- tical signal, H p , relates to peak concentration of aerosol entering the flame rather than to Cp. The process of aerosol generation and conditioning takes time, so that response is not instantaneous. In addition, spectroscopic limitations restrict the linear range of AA instruments, causing calibration plots to deviate towards the concentration axis.As a result of these peculiarities, the AA spectrometer behaves as a non-ideal FI detector. If a discrete sample plug is placed in the carrier stream close to an ideal detector so that no appreciable manifold (all flow regions excluding those subject to detector effects) dispersion occurs, the resulting manifold dispersion coefficient will be unity. As will be shown later, results of this experiment for FI-AAS confirm that the AA spectrometer is not an ideal FI-AAS.64 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 detector. Clearly, the ratio A,/A, does not reflect manifold dispersion, but some apparent dispersion, due solely to the detector.In FI-AAS work, therefore, it is important to distinguish between this apparent dispersion due to the response characteristics of the spectrometer and real disper- sion occurring in the flow manifold preceding the detector. In conventional use, AA spectrometers furnish steady-state absorbance signals from which concentrations are derived. In this context, “concentration” refers to the concentration of the sample entering the nebuliser, i.e., the spectrometer is regarded as a “black box” detector that enables the concentra- tion of a sample input to be determined. What happens to the sample inside the box is of secondary importance to the majority of users. Adopting the same approach in FI-AAS we shall define C, as the maximum concentration of determinand in the stream immediately prior to entering the nebuliser. Defined in this way, the ratio C,/C, is clearly a property of the FI manifold and will be termed the manifold dispersion coefficient, D,, where Cm CP D,=- .. . . . . * * (3) In conventional atomic absorption work, the sample must be aspirated for several seconds before a steady-state readout is obtained. The delay caused by the response time is usually acceptable so long as sufficient sample is available to produce the steady-state absorbance. The initial dynamic response of the instrument is explicitly ignored and no allowance has to be made for it. Only steady-state absorbance is read, and this relates directly to input concentration. With transient concen- tration profiles such as those existing in the flowing streams in FI-AAS, steady-state conditions are seldom attained, so that the effects of instrument response cannot be ignored.The signal is no longer a simple function of the sample concentra- tion C, (as in conventional work), but also depends upon the injected volume, Vi and the sample flow-rate, u. To distinguish D, from the response ratio A,/A,, the latter will be termed the response dispersion coefficient, D,, so that . . . . . . . * (4) Am D, = - A , Dr=f(D,,Dd) . . . . * * ( 5 ) Thus D, includes contributions from both D , and Dd, i.e., In particular, for the ideal detector (Dd = 1) we have Dr = D,; whilst, in situations where manifold dispersion is negligible D, = Dd. The detector dispersion coefficient D d is thus confirmed as the value of A,/A, when manifold dispersion is negligible.These concepts of dispersion coefficients might be usefully applied to other non-ideal FI detectors (e.g., slow-response potentiometric detectors). Equation (5) presents a challenge, as it embodies the idea of compounding dispersion coefficients. Although the topic was broached by RfiiiEka and Hansen2 more than six years ago, little quantitative progress appears to have been made. Yet the subject has considerable appeal. An understanding of it might enable manifold dispersion coefficients to be measured by FI-AAS methods. On a more general basis, it would be very convenient if one could plan an FI system by theoretically combining the dispersion coefficients of the various com- ponents prior to experimental trial. The question of environ- ment also needs to be resolved: is the dispersion coefficient of a single component constant, or does it depend upon the position of that component in combination with others? The existing theory [equations (6 and 7)] suggests that the dispersion coefficient of a 10 cm length of FI tubing depends upon whether it constitutes the first 10 cm or the last 10 cm of a 20 cm length of similar tubing.RfiiiCka and Hansen,2 on the basis of the definition of dispersion coefficient, proposed that a series of n components, having individual dispersion coefficients D , , D2, . . . D,, combined to produce an over-all dispersion coefficient, D , where D = D1 x D2 x . . . D, . . . . (6) However, this proposal is not consistent with their “Rule 5,” which states that the dispersion coefficient of the sample zone is proportional to the square root of the distance travelled.They have recently proposed3 that the effect of changing the length of the flow channel may be better described on the basis of dispersion factor rather than dispersion coefficient. Our experimental results support Rule 5 (but only when the sample volume is small), which may be expressed as equation (7) where K is a constant. It should be noted that the equation has to have the property of D = 1 when L = 0. Thus for tubes of equal diameters and lengths Ll, L2, . . . L, . . ( D - l ) = K L J . . . . . . (7) (Dl - 1)2 = K2 L, (02 - 1)2 = K2 L2 (D, - l)2 = K2 i, and applying equation (7) to the series combination yields i.e., It is suggested that this “tubes in series” equation might form a suitable basis for a quantitative approach to combining dispersions coefficients.Applying equation (8) to the combi- nation of an FI manifold of dispersion coefficient D , and a non-ideal detector of dispersion coefficient Dd enables equa- tion (5) to be re-written as which is also currently under investigation. The present work, however, is concerned exclusively with detector response; that is, it seeks to account for the detector’s contribution to the signals observed in FI-AAS. ( D - 1)’ = K2 (L1 + L2 + . . . L,) ( D - 1)2 = (Dl - 1)2 + (02 - 1)’ + . . . ( D , - 1)2 . . (8) ( D , - 1)2 = ( D , - 1)2 + (D, - 1)2 . . (9) Instrument Response Theory Use of Physical Models The scope of this investigation spans the areas of dispersion in FI systems and instrument response in atomic absorption.Both subjects have attracted the earnest attention of numer- ous researchers in recent years, yet have resisted accurate mathematical description. The difficulties centre around the complex and multivariate processes involved. Some of these are not well understood, except perhaps in terms of ideal behaviour, to which the real system seldom conforms. Even this treatment frequently produces equations that can only be solved by making further restrictive assumptions. In conse- quence, the solutions are often partial or conditional approxi- mations, and may involve parameters not readily related to the experimental variables. Examples of various treatments of the problems encountered in dispersion studies may be found in the work of R6iiEka and Hansen,2J Betteridge,4 Taylor,s Tijssen,6 van den Berg et al.,7 Reijn et ~ 1 .~ 8 3 9 Vanderslice et af.10 and Gomez-Nieto et uZ.11. Despite the dedicated efforts of these and other workers it is still not possible to write an expression for dispersion coefficients in terms of sample properties, tube dimensions and operating conditions. The extensive work on AA response falls within two broad categories: aerosol generation, which includes investigations by Castleman,12 Nukiyama and Tanasawa,13 Lane,14 Mugele and Evans,ls Bitron,16 Hrubecky17 and Mercer et af.18; and over-all nebuliser action, investigated by Stupar and Daw- ~011~19 Willis,20 Cresser and Browner ,Z1-23 Cresser and co- workers ,24-26 Browner and co-worker~27~28 and Gustavs- son.29-30 These excellent studies have furnished a wealth of information about aerosol quality and nebuliser performance,JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986.VOL. 1 65 I 1 I 0 /----- I , 2 0 t, Time - Fig. 1. Absorbance - time profile €or discrete sample introduction according to the single-tank model quantified by an array of specially defined parameters. Such information is invaluable in improving instrument perfor- mance. Nevertheless, it appears unlikely that absorbance will ever be directly expressed as a function of sample properties and nebulisation conditions. The intermediate steps are too many and too complex. In such situations a simple physical model may find useful application. Such models, developed by trial, error and modification, are based on simulation and simplification rather than rigorous mathematical treatment. They provide a tangible summary of the real system at operational level, a set of parameters readily identified with the experimental vari- ables, a basis for explanation, experiment, optimisation and forecast, and open the way to further theoretical develop- ment.It is for these reasons that a simple modelling approach is adopted as a means of circumventing the present impasse in this field. Single-tank Model Initial adjustment of an AA spectrometer involves selecting the values of various operating parameters to produce an acceptable instrument response. Thereafter the majority of these values are usually maintained constant whilst those selected for investigation are varied in turn.In the following discussion it is assumed that the instrument has been optimised and that the only experimental variables are to be sample concentration and volume injected. Within its linear range, the response of an AA spectrometer to a step change in concentration (see Fig. 1) has been shown3l to resemble an exponential growth curve of the type A = kC, (1 - e-Ul/V) . . . . . . (10) i.e., the instrument behaves as if the concentration step were modified by passage through a hypothetical well-stirred tank of volume V , prior to detection by an ideal detector. IfA = Am when t = 00, then A, = kC, and equation (10) may be re-written as A = A m ( l -e-Ur/V) . , . . . . (11) or l n ( & ) = V . . ut . . . . Equation (12) may be used to test experimental data; a linear plot of ln[A,I(A, - A)] against t indicates that the data conforms to the single-tank model.31 Compound Exponential Model: Tanks in Parallel As will be shown later, the response of the detector used in our studies deviated somewhat from that predicted by the single-tank model.If the flow of sample through the nebuliser does not correspond to ideally mixed flow, but rather to arbitrary flow, which appears likely considering the geometry of the typical spray chamber, then a spread of residence times occurs. Part of the sample passes almost unimpeded to the flame, whilst A Fig. 2. Basis of the parallel-tanks model. The flow is considered to split at point A, flow through a number of well-stirred tanks and recombine at point B other parts are held up for varying lengths of time. This type of flow may be modelled by a number of tanks in parallel, each accommodating a portion of the sample flow.The model illustrated in Fig. 2, consists of n tanks of volumes Vl-V,, connected in parallel. In this model, the fractions of the sample flowing through tanks 1, 2, . . . n are fl, f2 . . . fn, respectively, so that fl + f 2 . . . fn = 1. The flow-rate through tank i is ui where ui =fiu and for tank i the effluent concentration is Ci where ci = c, [ 1 - exp (-h 31 As mass is conserved at point B, then UlCl + u2c2 + . . . u,c, = u c therefore U l C l + u2c2 + . . . u,c, C= U n i . e . , = c ficj i = 1 The corresponding decay curve on switching back from the steady state to water is (14) The simplest parallel-tanks model consists of just two tanks in parallel, so that equation (13) reduces to The parallel-tanks approach provides an interesting and realistic model of the system response to step changes in concentration (see later).However, when applied to predict responses to more complex inputs, it produces complicated mathematical equations that are difficult to solve. In contrast, the single-tank model retains its essential simplicity and enables the determination of useful response equations for various inputs particularly where the response is observed soon after the input as in FI-AAS. Where the over-all response dispersion coefficient D, is dominated by an appreciable manifold dispersion coefficient D,, as may be so in FI-AAS, and is an essential condition for continuous dilution calibration32 work, the detector may be regarded as ideal as its contribution to the signal is negligible.For example, assuming that equation (9) is valid, for a system having D, = 10 and Dd = 1.5, then D, = 10.01.66 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 Prediction of System Responses Using the Single-tank Model The basic postulate of the single-tank model is that the system responds as though the sample stream leaving the manifold passes through a small well-stirred tank prior to detection by an ideal detector. The analyte mass content of the model’s hypothetical tank is given by the equation dm dt - = u (Ci - C) where Ci is the instantaneous concentration entering the tank and C is the concentration in the tank, at time t.Division by V yields However, within the linear range of the system response, A = kC, then The general equation (16) is the fundamental single-tank model prediction of the system response to an instantaneous concentration input Ci to the nebuliser. Conventional aspiration, i .e., step concentration input When a sample of concentration C, is aspirated, then Ci = C, = constant and the absorbance - time relationship follows equation (11). This predicts that the step concentration change generates an exponential increase in absorbance and that the maximum signal A , is recorded only after an infinite time. However, for appropriate values of u and V , 0.999 A, is reached after 2.5 s. Discrete sample introduction The single-tank model predicts that concentration may be determined using absorbance values recorded before the attainment of the steady state, provided that the response is observed after a fixed interval of steady sample flow.This principle is the basis of discrete sample nebulisation for which calibration is valid for both peak-height and peak-area modes. In accordance with equation (11) the peak absorbance A,, attained when a sample of concentration C, is aspirated continuously for time tp at flow-rate u, is given by A,= kC, [ 1 - exp (- 31 . . . . As tp is constant, then A, 0~ C,. The peak area is given by the sum of the two areas designated a and b in Fig. 1, + b = kC,[s,’p (1 - e-uf/V) dt + 1“ (1 - e-Ufp/V) e-u‘lvdt] but as tp is constant, peak area KC,. Thus, the model predicts that both peak height and peak area are valid measures of sample concentration.Beyond the linear range of instrument response calibration graphs are curved, as for conventional aspiration. ‘P Continuous dilution calibration This technique employs a real mixing tank of volume V’ to produce an exponential standard concentration - time profile. In response to a step change from zero to C, in the concentration input to the tank, the effluent concentration has been shown32 to be C = C,(1 - e-h’f) . . . . . . (18) where h‘ = ulv’. The detector response to an instantaneous input Ci is described by equation (16). Writing u/V = h, and substituting Ci = C from equation (18) yields dA - = h[kC, (1 - e-h’f) - A ] dt therefore dA - + hA = hkC, (1 -e-h’f) dt Integration and imposition of the condition A = 0 when t = 0 yields [h (1 - e-h’f) - h’ (1 - e-hf)] .. (19) A=- kCm (h - h’) as the system response to the exponential concentration - time input. If the volume of the real tank is large then, h’ + 0 so that the response approximates to that for a single tank [equation (lo)]. Flow injection In FI-AAS, other conditions being constant, the peak signal A, is always less than the steady-state signal Am owing to real dispersion in the manifold and apparent dispersion due to the detector. In the absence of any manifold dispersion, the peak absorbance is given by equation (17). The time to reach the peak is t, = Vi/u, thus A,=A,(l - e-vi’? . . . . . . (20) From this equation, the volume injected (Vi)99 to produce a peak signal equal to 99% of the steady-state absorbance, (Vi)99, is calculated to be Vln(100).For a typical value of V for the instrument used in these studies (40 pl) the value of (Vi)99 is 180 pl, i.e., a volume of at least 180 pl must be injected to produce an A, value equal to 99% of A,. If volumes of less than 180 pl are injected, then, the recorded peak would be reduced, due to the failure to attain the steady-state response. The injection of small sample volumes will give rise to high apparent dispersion coefficients due solely to detector response characteristics. By way of example, a 5-pl sample injected close to the nebuliser would yield a peak measuring about 12% of A,, i.e., showing an “apparent” or “detector” dispersion coefficient of 8. Despite this, FI-AAS is a valid technique because, for a fixed injection volume, a constant fraction, (1 - e-Vi’V), of the pulse is recorded, and this remains proportional to the sample concentration, as explained in the discussion of discrete nebulisation.Defects of the single-tank model Equation (20) predicts that A , is independent of the sample fow-rate u. In practice this is not so, thus, although the single- tank model accounts for the effects of varying sample volume and concentration, it does not predict the results of changing flow-rate. Before describing the further refinement of the model, a brief account of the contributions of other workers towards rationalising the complexities of pneumatic nebulisa- tion is presented. Pneumatic Ne buliser A recent report by Browner and Boorn27 has emphasised the current interest in nebuliser studies as a possible means of improving instrument performance.The nebulisation process is illustrated in Fig. 3. Air at a pressure PI, maintained by a suitable compressor, is allowed to escape via a venturi nozzle N. As the gas accelerates towards sonic velocity at the venturi throat, its pressure falls isentropically (adiabatically and reversibly) to a value P2 in accordance with Bernoulli’s principle. Beyond the throat, the airstream broadens, its velocity falls and kinetic energy is transformed into potential energy, compressing the air to a pressure P3, close to atmospheric pressure. Mathematical treatments of these processes are given in standard texts of fluid mechanics.33J4 By comparison, the mechanism of aerosol formation is notJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986.VOL. 1 67 Sample Air Waste Fig. 3. Basic mode of action of the concentric pneumatic nebuliser. N is a venturi nozzle, I an impact bead and C a centrifugal spoiler. For detailed explanation see text 0 10 20 30 Droplet diameterlym Fig. 4. Droplet size distribution according to the Nukiyama and Tanasawa equation. N, is the number of droplets with diameters between ( x - Ad2) and (x + Ax/2). When Q$Q, > 5000 and V, > 180 m s-l, p and q are re orted to be 2 and 1, respectively. The values of A and B are 0.9 and (2, respectively well understood and no exact theoretical treatment is avail- able. One can only be guided by the results of various empirical investigations and the theoretical explanations that have been proposed.In a typical concentric nebuliser, P2 is below atmospheric pressure, so that sample is aspirated into the air stream through the nebuliser capillary, whose tip is axially positioned in the venturi tube. The resultant diverging jet of primary aerosol Al strikes the impact bead I and shatters to yield a secondary aerosol A*, which progresses through a centrifugal spoiler (not present in all commercial designs) before entering the flame as the tertiary aerosol A3. As fine droplets improve instrument sensitivity and linearity of response whilst reducing interference effe~tsl9>~’ the supreme purpose of the nebuliser assembly is to introduce into the flame a large mass of sample in the finest possible form. Much effort has been directed towards attaining this goal, both in nebuliser design and in aerosol re~earch.25.2~ Despite the extent of the work, understanding of the processes involved is still rather limited.Many sources17~26~27~35 quote the empirical equations of Nukiyama and Tanasawa [equations (21a and b)] as a starting point in describing primary aerosol formation in concentric nebulisers. The distribution of droplets may be described by an equation of the form (see Fig. 4) N, =A~~exp(-Bxq) . . . . (21a) where x is droplet diameter and A , B, p and q , are constants and N, is the number of droplets with diameter between ( x - Ad2) and ( x + Ax/2). The Sauter mean diameter of the droplets (diameter of droplet whose volume to surface area ratio is the mean of the distribution) is given by xo where 585 where cr is surface tension<dyn cm-I), p is liquid density (g ml-I), 17 is liquid viscosity [poise (p)] , V1 and Vg are the linear velocities of liquid and gas flows (m s-I), respectively, and Q, and Qg are the volumetric flow-rates of liquid and gas, respectively (ml s-1).These useful equations26 are based on several hundred tests under different conditions using sub- sonic gas flows.35 Bitronl6 has shown that equation (21b) applies equally well to supersonic flow. However, it should be noted that the equation is dimensionally inconsistent and is also said to predict too high a proportion of large droplets.15 Its value, nevertheless, unlike the proposed mathematical alternatives, is that its parameters are readily identified with the operational variables of nebuliser systems.Equation (21a) indicates a wide range of droplet diameters in the primary aerosol (see Fig. 4) whilst (21b) implies that the mean droplet diameter is reduced by employing high volume- tric gas flow at high throat velocity, and low volumetric liquid flow emerging at high linear velocity, i.e., a fine nebuliser capillary is best. These results are supported by the findings of Lane14 who investigated the critical velocity at which droplets shattered in air. He found 612 d = - (v - u)* where d is the droplet diameter in mm, v is the critical air velocity and u is the droplet velocity on bursting in m s-l. Thus, increasing u stabilises larger droplets, whilst increasing v destabilises larger droplets in favour of smaller ones.Cresser25 has observed that increased volumetric flow of liquid increases the pressure P2 at the capillary tip. It might be envisaged that increased liquid flow results in loss of kinetic energy of the air stream due to collision with an increasing mass of liquid, which in turn removes the seat of the pressure depression. At high liquid flow-rates the mass flow of liquid (5-10 ml min-1) may approach that of the gas (8 1 min-1). As aerosol production and transport are powered by the energy of the air stream, it is likely that these processes will be influenced by liquid flow-rate. However, the bulk of the energy expended in forming droplets and accelerating them to the speed of the air stream is redistributed as thermal energy as ca. 95% of the aerosol mass strikes the walls of the spray chamber and drains to waste.Evaporation of water from the droplets and from the walls of the spray chamber constitutes the main loss of energy carried by the air stream.18 A typical Sauter mean droplet diameter may be calculated using data for distilled water at 20 “C namely (7 = 72.6 dyn cm-1, p = 0.998 g ml-1 and 17 = 0.010 p. Substituting these values in equation (216) yields xo = (:::) - + 28.7 ( y) ’” For a typical nebuliser, V, = 1 m s-1, therefore Vg - V, = Vg = 330 m s-1 Ql 1000 x - = = 1 QP 8 1 min-1 1000 X 8 ml min-1 4990 330 L xo = - + 28.7 = 15.0 + 28.7 = 43.7 pm i. e. , the mean droplet diameter of the primary aerosol is 43.7 pm. Three sets of results calculated in this way are illustrated in Fig. 5. After investigating the effect of nebuliser geometry under fixed operating conditions, Hrubeckyl7 confirmed that maxi- mum nebulisation occurs when the liquid is injected along the central axis of the air stream at the point of maximum velocity.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 68 300 200 . 5. IX 100 0 I I I 500 1000 1501 Q,/Qi Fig. 5. Mean droplet diameter as a function of Q$Q,. The values of V, - V1 are: A, 100; B, 200; and C, 300 m s-1 0 20 40 60 Droplet diameterlym Fig. 6. Approximate relative distributions of droplets in primary (horizontal shading), secondary and tertiary (vertical shading) aero- sols of a typical AAS nebuliser Detector I Effluent flow-rate, 0 = (4 + P U ) t Sample stream u, j I pu ~ i ;iluent flow-rate, instantaneous concentration Ci ( I-Pb Fig.7. Basis of the extended-tank model In this way, the maximum area of liquid surface comes in contact with the air stream with little disturbance of the flow to impair the formation of ligaments/droplets at the surface, after the mechanism proposed by Castleman.12 Impactor surfaces, placed in the path of the primary aerosol jet, are widely employed to increase the volume of aerosol generated. Experimental evidence indicates that they also raise the mean diameter of the droplet distribution.21 Forma- tion of larger droplets might result from fragmentation of the liquid film deposited on the surface of the bead25 and from disruption of the airstream both by the bead and by the shattering of the primary liquid droplets.17 Thus the imme- diate result is probably an increased volume of secondary aerosol of inferior quality.Fortunately, many of the larger droplets are short-lived, as the great bulk are removed by gravitation and by collision with the walls of the spray chamber, a process promoted by such devices as centrifugal spoilers, so that the tertiary aerosol that finally enters the flame is much finer. These modifications of the aerosol, from generation to entry into the flame, are illustrated in Fig. 6. In view of the complexity of the nebulisation process, the probability of deriving theoretical expressions linking absor- bance and nebulisation rate appears remote. Clearly u, Ql and Q, contribute to the character of the aerosol entering the flame, which in turn determines the absorbance signal, but, apart from qualitative guidelines, little more can be said to link these entities.Under these circumstances, a possible solution is to attempt to extend the single-tank model to take account of the effects of flow-rate. Extension of the Single-tank Model to Accommodate the Effects of Flow-rate The single-tank model was extended as shown in Fig. 7. The well-stirred hypothetical tank of volume V has two inputs: a constant stream of diluent, flow-rate q , and sample stream, flow-rate u. A useful fraction, p, of the sample is dispersed throughout the tank whilst the remainder flows to waste. The observed system response is proportional to concentration and rate of introduction to the detector. Incompressible flow is assumed throughout and p is a function of flow-rate.Con- sideration of the analyte mass balance of the tank yields (2) tank = (2) in - ( 2 ) O”t =PUG - QC Division by. V yields dC PUC~ QC . . . . (224 ----- - dt V V ” The model for the detector response is that the absorbance is proportional to the rate at which analyte enters the detector, i.e., A = kQC. Substitution for QC in equation (22a) yields - V d A - + A = kpuCi . . . . (22b) Q dt Equations (22a and b) are mathematical expressions of the extended-tank model of the system response. Examination of the Extended-tank Model An acceptable model must account for the experimentally observed responses of the system. System response to a step change in concentration from zero to C , at a steady flow-rate u Substitution of Ci = C, into equation (22a) yields dC QC puC, -+-=- dt V V Integration and writing C = 0 when t = 0 gives the solution C=-Cm(l P U -e-@’v) .. Q . . (23a) and A = kpuC, (1 - e-Qr/v) . . . . (23b) i.e., the model predicts the exponential growth of absorbance to a steady-state value A , equal to kpUCm, the systemJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 69 response time constant being determined by the total flow Q through the system, rather than by the sample flow u. Thus, more rapid response is predicted by increasing the diluent flow, q. The greater dilution of the sample is compensated by increased volume flow-rate to the ideal detector so that absorbance is unaffected if p remains constant. At first sight, it might be thought possible to equatep in the model with the transport efficiency of the real nebuliser, but that would be gross oversimplification as there is an empirical function relating absorbance and sample flow-rate for a given set of operating conditions.In a real system, absorbance depends on a number of interconnected variables. It is not possible to vary sample flow-rate whilst all the others remain constant (e.g., nebulisation efficiency changes with flow-rate). Browner et a1.28 have developed the concept of useful aerosol mass (W,) as a single figure of merit for nebuliser perfor- mance. Whilst useful in nebuliser studies and in development work, the concept is of limited operational application as it is not readily evaluated in terms of the experimental variables. In FI-AAS work, interest centres around the useful exploitation of flow-rate while all other operating parameters remain fixed.The model parameter p may find application in this context. It is emphasised, however, that p , like W,, is a complex function of over-all nebuliser performance, and which, even for a single system, probably does not have day to day reprclducibility. Nevertheless, it does represent a convenient means of making quantitative allowance for nebulisation effects within a physical model, which otherwise totally ingores that problem area, the aerosol. Results obtained using real AA detectors reveal that the gradient of the graph of absorbance against flow-rate de- creases with increasing flow-rate, eventually becoming zero or negative under the influence of such factors as lower nebuliser efficiency, changes in droplet size distribution and increased solvent loading to the flame.These effects would be modelled using a p-function that decreases with sample flow-rate. It is hoped that a further study of the way in which p varies might provide information useful in optimising the system. If the function p is known in terms of u , equation (27) may be solved to give a value for urnax., the sample flow-rate giving maximum FI-AAS response using an injected volume Vi. Peak- width measurements When the single-tank model was adopted in preference to the parallel-tanks model, it was accepted that observations should be limited to the initial part of the absorbance - time response graph which approximates to a simple exponential function. Thus it is expected that the single-tank model will lose its quantitative validity as observation times are extended up towards steady-state conditions.This limitation should be borne in mind when considering measurements of peak width. Some error might appear inevitable, nor can it be minimised by measuring close to the base line, because although this corresponds to a short observation time for the growth graph, it represents a long observation time for the decay graph. It may be that the best accuracy is achieved by measuring at half the peak height, as is done in chromatography, as it has been shown (see later) that the initial part of the response graph approximates closely to a single exponential function. Tyson36 has used peak-width measurement at a constant height above the base line as a means of extending the calibration range of an atomic absorption spectrometer over several orders of magnitude.Signal growth is described by equation (23b), which may be rearranged to give ‘=--In(&) V . . . . Q A , - A Whilst for the corresponding signal decay, the time measured from the peak maximum is given by (29) FI-AAS predictions Assuming no manifold dispersion (i.e., plug flow), the response predicted by the extended-tank model to a step change from 0 to C, is given by equation (23b). If a discrete volume Vi is injected at a constant flow-rate u, then tp = Vi/U and A, = kpu C, (1 - e-Qvi/”V) . . . . (24) When u and Vi are constant, equation (24) simplifies to A , 0~ C,, i.e., peak height is a valid measure of injected concentration.When Ci and Vi are constant, then A, varies with u according to the equation Thus the peak height varies with flow-rate in a similar manner to the steady-state value. Note however that both the steady-state maximum A , and the fraction recorded as the pulse A, are functions of flow-rate, so that the response dispersion coefficient, too, varies with flow-rate, i.e., from equation (25), A,=A,(l-e-QVi/uv) . . . . (25) 9-l . . . . (26) Am A , D, = - = (1 - e-QVi/u As u and C, are constant, equation (25) predicts that A , increases exponentially with Vi up to a maximum of A, when vi = 00. Experiments (see later) show that A , passes through a maximum value (AJmax. with increasing u , and thereafter decreases. Substituting Q = q + p u into equation (24), followed by differentiation with respect to u and setting d(A,)ldu = 0 yields the equation Fig.8 shows that the total peak width at the absorbance Substituting the appropriate expressions from equations level A l is given by t’ where t’ = t, - tl + t2. (28) and (29) yields the equation Substituting from equation (26) and rearranging gives At constant flow-rate both Q and D, are constant and absorbance is proportional to the concentration of analyte in the hypothetical tank, so that equation (30) reduces to the form t’ = m ln(C, - C,) + c . . . . (31) i.e., a plot of peak width against ln(C, - Cl) is a straight line of gradient VlQ. Substitution for A , and D, in equation (30) yields the following expression for the variation of peak width with flow-rate. t’ = In [,,QVi/uV- 1) (y - l)] .. (32) Peak width-at constant height is therefore dependent upon both injected concentration and flow-rate. If peak width is measured at a constant fraction of peak height, Adn, then A , _ n A , = n D , - A1 A ,70 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 t 8 5 $ 2 m A Fig. 9. Effect of increasing Ci on peak width (Vi is constant). The peak width at half maximum height remains constant and the width at constant height increases Substituting this value of A,/A1 into equation (30) yields the equation (a) Water 0.25 mg I-’ Mg I I 0 t2 Time (decay) Fig. 8. Basis for the derivation of an expression for peak width from the extended-tank model t‘,=-ln V (-) nD,-1 . . Q Dr-1 . . (33) Thus, in contrast with the last result, the peak width at a constant fraction of peak height is independent of the injected concentration.Substitution from equation (26) yields the equation t’, = 1 n ln [(n - l)eQvi/vu + 11 . . . . (34) If the peak width is measured at half the peak height then n = 2 and t’* =I In (eQVilVu + 1) . . . . (35) Q This is illustrated in Fig. 9. Experimental Apparatus The following apparatus was used: Shandon Southern A3400 atomic absorption spectrometer, Tekmann TE200 chart recorder, Gilson Minipuls 2 peristaltic pump, Rheodyne Model RH50313-way PTFE rotary valve, Altex Model 201-25 eight-port injection valve and Marriott bottles (constant-head reservoirs). r - l Recorder I 1 I 0 5 10 Ti me/s Fig. 10. (a) Experimental set-up and (b) results for study of absorbance versus time for a step change in concentration from 0 to 0.25 mg 1-1 of magnesium Standard Solution A 1000 mg 1-1 solution of magnesium (BDH Chemicals Ltd.) was used.Procedures Investigation of single-tank model of system response Marriott bottles containing triply distilled water and 0.25 mg 1-1 magnesium solution were connected to alternate inlets of the three-way valve. The common outlet was connected directly to the nebuliser by means of a 4.5 cm length of 0.58 mm i.d. PTFE tube (Fig. 10). No pump was employed. Switching the valve smartly changed the input from water to magnesium solution. The growth of absorbance with time was monitored by the chart recorder operating at its maximum speed of 1 cm s-1. The procedure was repeated four times and the five sets of results were pooled to obtain the mean absorbance - time growth profile.The solution flow-rate was determined by measuring the time taken for the aspiration of 20 ml of solution, delivered by pipette to the intake of the appropriate supply line without changing the hydrostatic head. Effect of flow-rate on instrument response A 0.5 mg 1-1 magnesium solution was used as the test sample as the AA sensitivity for magnesium is high (0.5 mg 1-1 givesJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 71 Table 2. Variation of A,/A, with volume injected and flow-rate ulml min-1 vi/pl 1 2 3 4 5 6 7 8 9 30 1.90 2.13 2.29 2.54 2.80 3.08 3.39 3.69 3.91 70 1.29 1.29 1.42 1.54 1.67 1.79 1.92 2.03 2.13 100 1.16 1.15 1.21 1.29 1.38 1.46 1.54 1.61 1.62 150 1.03 1.05 1.07 1.11 1.17 1.23 1.28 1.30 1.31 I I 5 10 Timeis Fig.11. Plot of In[A,l(A, - A ) ] versus t for results shown in Fig. 10 0 5 10 Ti meis Fig. 12. Computer generated curves on the basis of the parallel- tanks model for different values off,: A, 0.95; B, 0.9; C, 0.8; D, 0.7; and E, 0.6. The values of u, V , and V, are 0.127 ml s-,, O.O44f1 and 1.98(1 - fi), respectively 0.44 absorbance) and the calibration graph shows no signifi- cant curvature over the concentration range 0-0.5 mg 1-l. Using the Altex valve, 30 p1 of the magnesium solution were placed in a flowing stream of distilled water immediately before it entered the nebuliser. Triplicate peaks were recor- ded at various pumping rates over the range 1-9 ml min-1. The procedure was repeated for 70-, 100- and 15O-pl injec- tions.The corresponding steady-state signals (Vi = a) were A, 0.22 al c m f $ a n 1 I 5 10 Timeis Fig. 13. Comparison of single- and parallel-tanks models of the absorbance - time function for a step concentration change from zero to 0.5 mg 1-1 Mg. A, Best fit for single-tank model; B, best fit for parallel-tanks model; and C, experimental curve 0.7 0.6 0.5 al c $ 0.4 s n 0.3 (1) a 0.2 0.1 \ / A /Ij/x-x-x\ x‘x ‘X \ ‘x B ‘x. ‘ C D ‘x. E 1 0 5 10 Flow-rate/m I m in Fig. 14. Effect of flow-rate on steady-state absorbance and on peak absorbance for the injection of a variety of sample volumes. A, Steady state; B, 150; C, 100; D, 70; and E, 30 p1 obtained by pumping a continuous stream of 0.5 mg 1-1 magnesium solution to the nebuliser at the same flow-rates and noting the absorbance values.In order to minimise manifold dispersion, the injection valve and nebuliser were connected by a short length (4.5 cm) of 0.58 mm i.d. PTFE tubing. The performance of the pump was checked by72 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. I Table 3. Variation of peak absorbance with flow-rate and volume injected Injected volume (V)lpl Flow- rate 30 70 100 150 00 QIV Pearson value correlation indicated coefficient (u)l ml min-1 Peak absorbance (A,) 1 0.105 0.155 0.172 0.195 0.200 0.396 0.996 2 0.158 0.260 0.293 0.320 0.336 0.824 1.000 3 0.204 0.330 0.385 0.435 0.467 0.846 0.999 4 0.224 0.369 0.440 0.512 0.567 1.01 0.999 5 0.226 0.379 0.457 0.537 0.632 1.01 1 .ooo 6 0.214 0.369 0.452 0.535 0.660 1.06 1 .ooo 7 0.196 0.346 0.431 0.518 0.665 1.13 1 .ooo 8 0.176 0.317 0.404 0.496 0.649 1.26 1 .ooo 9 0.157 0.288 0.380 0.467 0.614 1.43 0.999 ~~ Table 4.Values of extended-model parameters V , b and q Pearson Least-squares fit to correla- y = A + Bu + Cu* tion Volume coeffi- (ViYP1 A B C cient 30 0.4331 12.18 -35.38 0.996 70 0.3368 12.51 -39.84 0.989 100 0.3715 11.05 -28.78 0.975 150 0.3488 11.13 -29.29 0.966 Mean values . . . . . . . . Values implied for model VI bl q1 pl sml-’ PIS-’ 82 2.9 35.6 80 3.2 26.9 90 2.6 33.6 90 2.6 31.3 85.5 2.8 31.9 0 5 10 15 Flow-rate/ml min-’ Fig. 15. Variation of response dispersion coefficient (Or) with flow-rate for several injection volumes: A, 150; B, 100; C, 70; and D, 30 pl. ( a ) , Curves predicted and (b), experimental curves Table 5.Variation of urn,,. with Vi umax,Iml min-1 V11pI Model Experimental 30 6.9 4.5 70 7.6 5.0 100 8.0 5.4 150 8.6 6.8 measuring the solution flow-rate through the system at pump settings of 200, 400, 600, 800 and 1000, using the method described above. The free aspiration rate of the nebuliser via the 4.5 cm inlet tube was measured. Results and Discussion The mean absorbance - time profile is shown in Fig. 10, and the variation of detector dispersion coefficient with volume injected and flow-rate in Table 2. These results clearly show the non-ideal behaviour of the atomic absorption detector. If the single-tank model is applicable then a plot of ln(A,/A, - A) against t [equation (12)] would be linear. As can be seen from Fig. 11 the plot consists of two linear portions. The initial gradient, GO is 2.86 s-1 and the final gradient, G,, is 0.064 s-1.The aspiration rate was calculated to be 0.127 ml s-1. Although the single-tank model is clearly not applicable throughout the entire timescale of the experi- ment, the initial portion of the plot, occupying about 0.7 s indicates that the model would be applicable up to about 80% A,. Were the failure of the model due to inadequate amplifier or recorder response time, then this would be most apparent when the signal was increasing rapidly. However, this was not observed. The two apparently linear sections of the graph of ln(A,/A, - A) against t suggest that the system response might conform to a compound exponential function contain- ing two exponential terms with very different time constants, corresponding to the gradients Go and G,.This would be the parallel-tanks model with two tanks (Fig. 1) for which the basic response is given by equation (15), in which ufilV1 is Go (small tank with rapid response) and uf2/V2 is G, (large tank with slow response). From Fig. 11, Go was found to be 2.86 s-1 and G,, 0.064 s-1 thus V1 = 0.44fi and V, = 1.98 (1 - f l ) . A computer was used to generate response curves according to equation (15), for various values of fi. These are shown in Fig. 12. Of the values tested,fi = 0.9 ( V , = 0.04 ml, V2 = 0.2 ml) gave a good fit. No doubt an even better fit could be obtained by carefully adjusting the parameters of this very flexible model. The effect of the second tank in the model can clearly be seen from Fig.13, which also shows the curve for a single tank of volume 0.044 ml (the best-fit volume for this model). Extended-tank Model FI-AAS response is described by this model in equation (25), which may be rearranged to give i.e. , within the linear range of instrument response, QlV may be measured as the gradient of a plot of ln(A,/A, - AP) against Vilu. Data for such plots taken from the results shown in Fig. 14 are given in Table 3 and show that QlV is not constant but increases with u, as predicted by the model. As Q = q + pu (Fig. 7), equation (25) can be re-arranged to give Examination of the steady-state absorbances in Table 3 (also shown in Fig. 14) suggesteh that the relationship between p and u might be p = u(1 - bu) where b is a constant. Substitution for p in equation (36) gives (37) Values of the left-hand term of equation (37), at constant Vi, were tabulated against the nine values of flow-rate in Table 3,JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 73 10 v) % 0 3 m 5 Y Q 0 0.1 0.2 0.3 Flow-rate/mI s-1 Fig. 16. Solid lines, variation of peak width, t’, with flow-rate as predicted by the extended-tank model, equation (32) for various volumes injected. C, is 0.5 mg 1-1; A l is absorbance 0.1; and t, k, q and V are 2.8 s m1-l- 22.9 A s ml-1 m r l 1.0.032 ml s-’ and 0.085 ml. respectively. Broken lines indicate t& experimental results. A, 150; B, 100; C, 70; and D, 30 p1 and the best least-squares fit to the general quadratic y = A + Bu + Cu2 was computed. Comparing the empirical results with the model gives A = q/V; B = l / V ; and C = -b/V.The results of the calculation are shown in Table 4. The mean values of the model parameters are V = 0.085 ml, b = 2.8 s ml-1 and q = 0.032 ml s-1. Some of the model’s predictions are examined, using these values. urnax,, flow-rate giving maximum peak height u,,,, varies with Vi, as indicated by equation (27). The model parameters obtained givep = (1 - 2.8u), dpldu = -2.8. After substituting these values into equation (27), the equation was solved for each value of Vi by an iterative procedure using a microcomputer. The resulting values of urnax. are compared with the experimental values in Table 5. The model predicts the observed upwards trend of u,,,, with increasing Vi. Response dispersion The variation of D, with flow-rate is decribed by equation (26).Values calculated using this equation are compared with the experimental values in Fig. 15. Peak width Peak widths computed using the model [equation (32)] are shown in Fig. 16, together with the experimental values. It can be seen that there is good agreement between the experi- mental values and the calculated values. The interesting result arises that, whilst increasing the flow-rate leads to an increased detector dispersion coefficient it nevertheless results in narrower peaks owing to the more rapid change of absorbance with time. Conclusions The mode of operation of the nebuliser of an atomic absorption spectrometer confers important detector charac- teristics that must be taken into account when experimental results are interpreted.The instrument smooths the sample input, and the response always lags behind it: points that are easily overlooked when the instrument is operated in the conventional steady-state mode. In a kinetic situation the consequences become more significant as the concentration - time gradient of the input increases. Thus in FI-AAS the detector’s response kinetics may exert a considerable influence over the peak absorbance, so that the ratio of steady state to peak absorbance cannot be routinely employed as an index of sample dilution in the manifold. The extended-tank model for nebuliser behaviour produces good agreement between experimental and predicted behavi- our for the variation of dispersion coefficient with flow-rate and volume injected and of peak width as a function of flow-rate and volume injected.Although the model predicts that the flow-rate at which the maximum peak height will occur as flow-rate is varied increases as the flow increases, there is not particularly good agreement between the numer- ical values. The results show, however, that a simple model is a valuable aid to understanding the behaviour of a complex system involving several variables. Where exact mathematical treatment fails, such models allow predictions of system behaviour and the continuation of theoretical development. Semi-empirical models, which allow characteristics of the real system to be incorporated into the basic model, may provide an accurate account of the response of the real system. The equations developed should prove useful in the interpretation of results obtained in FI-AAS work.Financial support for J. M. H. A. by the SERC is gratefully acknowledged. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22 * 23. 24. 25. References RGiiEka, J., and Hansen, E. H., “Flow Injection Analysis,” Wiley, New York, 1981. R8iitka, J., and Hansen, E. H., Anal. Chim. Acta, 1978, 99, 37. RciiEka, J., and Hansen, E. H., Anal. Chim. Acta, 1984, 161, 1. Betteridge, D., Anal. Chem., 1978, 50, 832A. Taylor, G., Proc. R. SOC. London, A , 1953,219, 186. Tijssen, R., Anal. Chim. Acta, 1980, 114, 71. van der Berg, J. M. H., Deelder, R. S . , and Egbrink, H. G. M., Anal. Chim. Acta, 1980, 114,91. Reijn, J. M., van der Linden, W. E., and Poppe, H., Anal. Chim. Acta, 1980, 114, 105. Reijn, J. M., and Poppe, H., Anal. Chim. Acra, 1983,145,59. Vanderslice, J. T., Stewart, K. K., Rosenfield, A. G., and Higgs, D. J., Talanta, 1981, 28, 11. Gomez-Nieto, M. A., Luque de Castro, M. D., Martin, A., and Valcarcel, M., Talanta, 1985, 32, 319. Castleman, R., J. Res. Nut. Bur. Stand., 1931, 6, 369. Nukiyama, S., and Tanasawa, Y., translated by Hope, E . , “Experiments on Atomisation of Liquids in an Air Stream,” Defense Research Board, Department of National Defense, Ottawa, Canada, 1950. Lane, W. R., Ind. Eng. Chem., 1951, 43, 1312. Mugele, R. A., and Evans, H. D., Znd. Eng. Chem., 1951,43, 317. Bitron, M. D., Ind. Eng. Chem., 1955, 43, 23. Hrubecky, H., J. Appl. Phys., 1958, 29, 572. Mercer, T. T., Tillery, M. I . , and Chow, H. Y . , Am. Znd. Hyg. Assoc. J . , 1968, 29, 66. Stupar, J., and Dawson, J. B., Appl. Opt., 1968, 7, 1351. Willis, J. B., Spectrochim. Acta, 1967, 23, 811. Cresser, M. S., and Browner, R. F., Appl. Spectrosc., 1980,34, 365. Cresser, M. S., and Browner, R. F., Spectrochim. Acra, Part B, 1980, 35, 73. Cresser, M. S., and Browner, R. F., Anal. Chim. Acta, 1980, 113, 33. O’Grady, C., Marr, I. L., and Cresser, M. S . , Analyst, 1984, 109, 1085. Cresser, M. S., Anal. Proc., 1985, 22, 65.74 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 26. Green, H. L., and Lane, W. R., “Particulate Clouds,” Spon, London, 1957. 27. Browner, R. F., and Boorn, A. W., Anal. Chem., 1984, 56, 780,875A. 28. Browner, R. F., Boorn, A. W., and Smith, D. D., Anal. Chem., 1982,54, 1411. 29. Gustavsson, A., Anal. Chem., 1984,56,817. 30. Gustavsson, A., Anal. Chem., 1983, 55, 94. 31. Tyson, J. F., Appleton, J. M. H., and Idris, A. B . , Anal. Chim. Acta, 1983, 145, 159. 32. Tyson, J. F., and Appleton, J. M. H., Tulanta, 1984, 31, 9. 33. Daugherty, R. L., and Franzini, J. F., “Fluid Mechanics with Engineering Applications,” Sixth Edition, McGraw-Hill, New York, 1965. 34. Walshaw, A. C., and Jobson, D. A., “Mechanics of Fluids,” Second Edition, Longmans, London, 1972. 35. Hidy, G. M., and Brock, H. R., “The Dynamics of Aerocolloi- dal Systems,” Volume 1, Pergamon, Oxford, 1970. 36. Tyson, J. F., Analyst, 1984, 109, 319. Paper 5.516 Received May 31st, 1985 Accepted August 2nd, 1985

ISSN:0267-9477    

DOI:10.1039/JA9860100063

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 75 Signal to Noise Ratios for Flow Injection Atomic Absorption Spectrometry James M. Harnly and Gary R. Beecher US Department of Agriculture, Agricultural Research Service, Beltsville Human Nutrition Research Center, Nutrient Composition Laboratory, Belts ville, M D 20705, USA Signal to noise ratios for flow injection coupled with flame atomic absorption spectrometry (FI-AAS) were compared with conventional nebulisation. Peak heights and areas were determined for a series of sample volumes (20, 100 and 500 pl) and flow-rates (3.2, 1.6,0.8 and 0.4 ml min-l). The reduced sample flow-rates (with a constant nebuliser gas flow-rate) improved nebulisation efficiencies by factors of 4-1 2. Peak-area signal to noise ratios were better than those for peak heights in every instance.The signal to noise ratios for FI-AAS peak height and area approached but never exceeded the ratios for conventional nebulisation. Keywords: Flow injection; flame atomic absorption spectrometry; signal to noise ratio Flow injection (FI), combined with flame atomic absorption spectrometry (FAAS), is a versatile tool.' The attractive features of FI-AAS are: (a) small sample volumes; (b) rapid sample throughput; ( c ) discrete nebulisation (tolerance of high salt concentrations, minimisation of viscosity effects and reduction of chemical interference effects); ( d ) alternative calibration possibilites; and (e) matrix dilution (addition of reagents for suppressing interferences). A disadvantage of FI-AAS is the inherently poorer signal to noise ratios arising from the smaller sample volumes.2 A feature of FI-AAS that has remained largely unexplored is the increased nebulisation efficiency, or signal to volume ratio, obtained using reduced sample pumping rates while the nebuliser gas flow is held constant.s6 Without a flow injector, a nebuliser will take up liquid at a rate that is determined by the gas flow-rate and the length and bore of the capillary tubing.The analyst adjusts the gas flow or (in the case of a variable nebuliser) the aspiration rate to give maximum sensitivity. The optimum sample uptake rate is usually 6-10 ml min-1. Wolf and Stewart5 showed that reducing the sample uptake rate, without changing any other parameters, reduces the signal but increases the atomisation efficiency.The net result is a 33% decrease in the peak height but a four-fold increase in the peak area (provided that all of the signals fall in the linear range). The increased nebulisation efficiency was also accompanied by an increase in the duration of the analytical signal and an increase in the area noise. As a result, peak-area signal to noise ratio increased by a factor of 1.6. Thus using reduced sample flow-rates and peak integration, improved signal to noise ratios can be obtained. It has been demonstrated that higher peak heights can be obtained for FI-AAS if the sample pumping rate exceeds the optimum aspiration rate of the nebuliser.7 This increase in the peak height, however, was achieved through the increase in the sample delivery not an increase in the nebulisation efficiency.The nebulisation efficiency decreases as can be seen by the reported less than linear response of the sample signal to the sample flow-rate.' To date, increased efficiencies for conventional nebulisers have only been achieved by reducing the sample pumping rate. This paper examines the peak-height and -area signal to noise ratios that are achieved for a series of reduced sample pumping rates while the nebuliser gas flow-rate is held constant. The signal to noise ratios are also examined as a function of the sample volume and are compared with conventional nebulisation. The trade-offs in the choice of FI-AAS parameters are considered. Experimental The FI system has been described previously.5 A standard straight-line configuration with one six-port rotary sampling valve (Model AH60, Valco Instrument Co., Houston, TX) was used with sample loops of 20 (0.05 cm i.d.), 100 (0.08 cm i.d.) and 500 p1 (0.04 cm i.d.).A 15 cm long (0.06 cm i.d.) section of tubing connected the sampling valve to a conven- tional pneumatic nebuliser. The sample pumping rate was determined by a variable rate positive displacement minipump (Milton Roy Co., Riveria Beach, FL) while the nebuliser gas flow was held constant. Flow-rates of 3.2, 1.6, 0.8, and 0.4 ml min-1 were used. Data acquisition was accomplished with a simultaneous multi-element atomic absorption continuum source spec- trometer (SIMAAC) , which has been described previously.8~9 With SIMAAC, data can be acquired for up to 30 s and the analytical signal can be integrated over any pre-set interval to the nearest 1/56 of a second.Data acquisition was triggered at the time the sample loop was switched into the main stream. Optimum integration intervals were determined for each sample volume and flow-rate by measuring the appearance time and ending time of the peaks from strip-chart recorder Table 1. Integration interval (seconds) Sample Sample size/pl ml min-1 20 100 500 3.2 4.3 7.5 15.5 1.6 6.4 10.7 26.9 0.8 10.7 20.0 0.4 17 pumping rate/ - - - Table 2. FI-AAS peak heights (absorbance)* Sample Sample size/yl ml min-1 20 100 500 pumping rate! 3.2 0.0585 0.118 0.148 1.6 0.0509 0.0947 0.123 0.8 0.0389 0.0685 0.4 - - 0.0232 - * Conventional nebulisation (9.0 ml min-1) = 0.1572 A.76 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 Table 3. FI-AAS peak areas (absorbance seconds)* ~ Table 4. FI-AAS peak-height coefficients of variation* Sample Sample size/pl pumping rate/ ml min-1 20 100 500 3.2 0.0588 0.313 1.18 1.6 0.0966 0.481 1.70 0.8 0.136 0.679 0.4 0.170 - - - * Conventional nebulisation (9.0 ml min-I) for 5 s = 0.780 A s. 0.20 0.16 cn 5 0.12 (0 2 Y a 0.08 0.04 I I I 1 .o 2.0 3.0 Sample pumping rate/ml min-’ Fig. 1. Peak area versus the sample pumping rate for a 20-1.1.1 sample volume. Nebuliser gas was held at a constant flow, which provided a conventional sample nebulisation rate of 9.0 ml min-l tracings. The intergration intervals used in this study are shown in Table 1. No data were collected at 100 pl for a sample pumping rate of 0.4 ml min-1 or at 500 pl for sample pumping rates of 0.8 or 0.4 ml min-1 as the peak lasted longer than 30 s, the integration limit of the computer program.For this study, data were only acquired for Cu, 324.7 nm, in an air - acetylene flame. Peak heights and areas were determined for each of the sample loop sizes and pumping rates listed above. Ten measurements were made for each set of conditions for a standard and a blank. Data for conventional nebulisation were obtained without the FI, using the same gas flow-rate and a 16.5 cm (0.06 cm i.d.) piece of polyethylene capillary tubing attached to the nebuliser. Results and Discussion The FI-AAS peak heights and areas for 10 pg ml-1 of Cu are listed in Tables 2 and 3, respectively, as functions of the sample volume and pumping rate.The largest peak heights were produced, as expected, by the largest sample volumes and the fastest pumping rates. The best peak-height signals for FI, however, were still less than that for conventional nebulisation, which gave a peak height of 0.157 at an uptake rate of 9.0 ml min-1. Larger peak heights can be obtained for FI-AAS with faster sample pumping rates or larger sample sizes, but the maximum peak height can only be equivalent to that for conventional nebulisation.5J0J1 The largest peak areas were obtained for the largest sample volumes but at the lowest pumping rates, and consequently, the longest integration periods (Table 3). For the smaller sample volumes (20 and 100 pl), the peak areas were proportional to the volume.At 500 pl, the areas were no longer proportional due to a much larger fraction of the sample peak exceeding the linear absorbance range (greater Sample Sample size/pl pumping ratel ml min-1 20 100 500 3.2 4.4 1 .o 0.5 1.6 1.5 0.5 0.9 0.8 2.7 1.3 - 0.4 9.6 - - * Conventional nebulisation (9.0 ml min-1); coefficient of variation = 0.003. Table 5. FI-AAS peak-area coefficients of variation* Sample Sample size/pl pumping ratel ml min-1 20 100 500 3.2 4.6 0.2 0.3 1.6 1.8 1 .o 0.3 0.8 2.1 0.8 0.4 2.6 - - - * Conventional nebulisation (9.0 ml min-I); coefficient of variation = 0.008. than 0.1; the linear range ends at a lower absorbance for SIMAAC than for conventional line source AAS). Reducing the sample pumping rate produced an accelerat- ing increase in the nebulisation efficiency.This is illustrated by the graph of the peak area versus pumping rate in Fig. 1. Injection of 100 or 500 1-11 at pumping rates of between 0.8 and 3.2 ml min-1 gave peak areas ranging from approximately a factor of two worse to approximately a factor of two better than the value of 0.78 A s obtained for conventional nebulisation for 5 s (750 pl). An integration period of 5 s was chosen for conventional nebulisation as this interval is used in our laboratory for all routine determinations. The areas of the three undetermined values in Table 3 (100 pl at 0.4 ml min-1 and 500 yl at 0.4 and 0.8 ml min-1) will all exceed the value for conventional nebulisation (0.78 A s). Unfortunately the required integration period (35-80 s) exceeded the limits of our data acquisition program.The data in Table 3 cannot be used to compare the nebulisation efficiency, or signal to volume ratios, of FI and conventional nebulisation as some of the absorbances fell in the non-linear region. Nebulisation of a standard of a lower concentration (in the linear range) permitted the signal to volume ratios to be compared. The FI efficiencies were 4.1, 6.8,9.5 and 11.9 times greater than conventional nebulisation at sample pumping rates of 3.2, 1.6, 0.8 and 0.4 ml min-1, respectively. The cost of these increased efficiencies is increased integration time (Table 1) and reduced sample throughput. Signal to noise ratios were computed for the data in Tables 2 and 3 in two ways. The first method employed ten repeat determinations of the analytical signals and used the com- puted means and standard deviations to compute the coeffi- cients of variation.The standard deviations of these measure- ments reflect the base-line noise contribution (arising from the light source)and the analytical signal and matrix background fluctuation noises. As the detected limit is approached the sample noise components diminish in significance until the base-line noise is limiting. Accordingly, the second method computed signal to noise ratios using the analytical signals and the base-line noises. Tables 4 and 5 give the coefficients of variation, based on precision of the analytical signals, shown in Tables 2 and 3. For 100- and 500-yl sample volumes, the relative precisions forJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL.1 77 Table 6. FI-AAS peak-height signal to base line noise ratios* Sample Sample size/yl pumping rate/ ml min-1 20 100 500 3.2 39 79 99 1.6 34 63 82 0.8 26 46 0.4 15 - - - * Conventional nebulisation (9.0 ml min-l) = 105. Table 7. FI-AAS peak-area signal to base line noise ratios* Sample Sample size/yl pumping rate/ ml min-1 20 100 500 3.2 79 277 77 1 1.6 98 398 842 0.8 112 402 - 0.4 102 - - * Conventional nebulisation (9.0 ml min- l ) = 904. Table 8. FI-AAS area base-line standard deviation (absorbance seconds) * Sample Sample size/p1 ml min-1 20 100 500 pumping rate/ 3.2 0.007 0.001 1 0.0015 1.6 0.0010 0.0012 0.0020 0.8 0.0012 0.0017 - 0.4 0.0017 - - * Convention nebulisation (9.0 ml min-1) for 5 s, area base-line standard deviation = 0.0009 A s.lo-* ~ , j 10 100 10-4 1 .o In teg ration t ime/s Fig. 2. Area base-line standard deviation versus integration time both height and area were close to 1.0% or better. Sample volumes of 20 pl yielded consistently poorer coefficients of variation than 100 or 500 1-11, ranging from 2 to 10% for peak height and 2 to 5% for peak area. The peak-height and -area signal to noise ratios, based on the base-line noise levels, are shown in Tables 6 and 7. These values are inversely proportional to the detection limit. The base-line noise level for the peak-height measurements was based on the standard deviation of the base-line absorbance. The standard deviation of the base-line absorbance is indepen- dent of the number of values used for the computation; the more values that are used, the greater the confidence of the computed standard deviation.For this study, a standard deviation of 40.0015, the average of several computations, was used for determining the peak-height signal to noise ratios. The base-line noise levels for peak area were based on the standard deviation of ten or more repeat integrations for the atomisation of a blank for the time intervals listed in Table 1. The area base-line standard deviations (Table 8) are propor- tional to the square root of the integration time. Fig. 2 illustrates this relationship. A first-order least-squares fit to the logarithms of the data gave a slope of 0.4998. This means the area base-line noise is shot-noise limited: i.e. , limited by the uncertainty of the arrival of photons at the photomultiplier tube cathode.The values listed in Table 8 were used in computing the peak-area signal to noise ratios. In general, the peak-height signal to noise ratios showed the same pattern as the peak-height data (Table 2) while the area signal to noise ratios showed less of an increase than might be expected from the area data (Table 3). This reflects the constant nature of the base-line absorbance noise compared with the time dependent base-line area noise. None of the FI-AAS height or area signal to noise ratios exceeded the signal to noise ratios of conventional nebulisation for height, 105, or area, 904. The data in Tables 6 and 7 illustrate the compromises involved in the selection of operating parameters for FI-AAS. The most consistent generalisation to be found in the data is that the signal to noise ratios for peak area are better than those for peak height.Selection of other operating parameters depends on the needs of the analyst. For samples requiring the best detection limits and having no volume limitations, conventional nebulisation offers the best detection limits. FI-AAS, using 100-p1 sample volumes, offers signal to noise ratios comparable to 5 s of conventional nebulisation. Still larger sample volumes and integration intervals would offer better signal to noise ratios for both methods. However, FI-AAS attains comparable detection limits primarily by imitating conventional nebulisation (large sample volumes and high flow-rates) and thus sacrificing some of the more desirable features of FI.However, where sample volume, not detection limits, is critical, FI-AAS offers high precision for very small sample volumes. In this study, 20 p1, the smallest sample volume injected, gave relative peak-area precisions of 2% (Table 5). Smaller sample volumes, with larger relative precisions, could be used depending on the requirements of the determination. Using these volumes sample throughput would be quite high. If both the detection limits and sample volume are critical, FI-AAS again offers useful compromises. Conventional nebu- lisation requires 750 1.11 for a 5-s data acquisition (at a sample uptake rate of 9.0 ml min-1) and yields a signal to noise ratio of 900. Using 100 1-11 of sample and a pumping rate of 1.6 ml min-l, FI-AAS with peak-area measurement can achieve a signal to noise ratio of 400, i .e . , a detection limit only a factor of two worse than conventional nebulisation, but requiring just 13% of the sample volume. Finally, with no detection limit or sample size restrictions, FI-AAS and conventional nebulisation, with an autosampler, are quite similar. FI-AAS has the advantage of increased precision in sample handling, fewer physical interferences, a higher rate of determination and general versatility. While the data reported in this study are highly specific to the experimental set-up employed, the general trends can be applied to FI-AAS in general. The integration times and peak heights and areas (Tables 1, 2 and 3) are dependent on the nebuliser and the FI configuration. However, reducing the sample pumping rate (while holding the gas flow constant) can be expected to reduce the signal and increase the atomisation efficiency of any pneumatic nebuliser .The exact noise levels will depend on the AAS, but almost all spectrometers employing hollow-cathode lamps are shot-noise limited (the integrated base-line noise is proportional to the square root of78 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY, FEBRUARY 1986, VOL. 1 the integration time). Consequently, while it may be useful for the analyst to characterise individual spectrometers, the conclusions of this study should be helpful in predicting signals and signal to noise ratios for FI-AAS. 6. 7. 8. 9. References 10. 1. 2. 3. 4. 5. Tyson, J. F., Analyst, 1985, 110, 419. Tyson, J. F., and Sarkissian, L. L., Anal. Proc., 1985,22, 19. Jones, D. R., Tong, H. C., and Manahan, S. E., Anal. Chem., 1976, 48, 7. Szivos, K . , Polos, L., and Pungor, E., Spectrochim. Acta, Part B , 1976,31, 289. Wolf, W. R., and Stewart, K. K., Anal. Chem., 1979,51,1201. Koropchak, J. A., and Coleman, G. N., Anal. Chem., 1980, 52, 1252. Brown, M. W., and RGiiEka, J., Analyst, 1984, 109, 1091. Harnly, J. M., O’Haver, T. C., Golden, B., and Wolf, W. R., Anal. Chem., 1979, 51, 2007. Harnly, J. M., Miller-Ihli, N. J., and O’Haver, T. C., J . Autom. Chem., 1982, 4, 54. Treit, J., Nielsen, J. S., Kratochvil, B., and Cantwell, F. F., Anal. Chem., 1983, 55, 1650. Olsen, S., Pessenda, L. C. R., RGiiEka, J., and Hansen, E. H., Analyst, 1983, 108, 905. Paper J.5126 Received August 14th, 1985 Accepted September 18th, 1985

ISSN:0267-9477    

DOI:10.1039/JA9860100075

出版商:RSC    

年代:1986

数据来源: RSC