摘要:

Investigation into the Kinetics of Selenium(vi) Reduction Using Hydride Generation Atomic Fluorescence Detection STEVE J. HILL LES PITTS AND PAUL WORSFOLD Department of Environmental Sciences University of Plymouth Drake Circus Plymouth UK PL4 8AA Utilizing the rapid response times that are possible with atomic fluorescence spectrometers the activation energy for the reduction of selenium(v~) to selenium(1v) has been calculated and found to be 90.4 kJ mol-' using 6 moll-' HCl. Results from the studies indicate that full reduction of selenium(v1) to selenium(1v) may be obtained in as short a time as 6 min at 70 "C. Keywords Activation energy; selenium; atomic fluorescence spectrometry; hydride generation; selenium reduction Since the analytical potential of hydride generation was first reported by Holak,' the technique has become an accepted method for the determination of elements which form volatile hydrides.These elements include antimony arsenic bismuth germanium lead selenium tin and tellurium. In essence the hydride is generated by chemical reduction of the sample and is then entrained in a current of inert gas (usually argon) and led into the observation zone where it is decomposed by heat (and in some cases free radical action) to form the atomic vapour. In practice the most commonly used technique relies on the action of acidified sodium tetrahydroborate on the analyte when the reaction proceeds as shown BH4- + H+ + 3H,O+H3BO3 + 8Ho (n + 2)H0+ Em+ +EH + H,(excess) where E represents the analyte and n and m may be equal.The use of sodium tetrahydroborate has generally super- ceeded the use of a metal-acid reaction originally reported by Holak because it is superior both in terms of reaction speed and reduction yield. It also provides less contamination of blanks and generates hydrides from all the hydride-forming elements. However one potential drawback of the method is that it requires the elements to be in a particular oxidation state before the hydride may be formed; the +4 state for selenium. Selenium is one of the elements most commonly determined using hydride generation. It exists naturally in both the + 6 and + 4 oxidation states in inorganic selenium compounds +4 and +2 as organo-selenium compounds 0 as elemental selenium and -2 as hydrogen selenide. The reduction of selenium(v1) to selenium(1v) is generally accomplished by heat- ing the analyte in the presence of hydrochloric acid usually at an acid strength of 6 moll-'.The reaction is as follows HSe0,- + 3H+ + 2C1- +H,SeO + Cl,(aq) + H,O Various workers report different conditions for this reduction but heating t o 70 "C for 30 min in an open-topped vessel is usually employed. The use of an open-topped vessel is import- Analytical Atomic Spectrometry ant as it allows the generated chlorine to escape and thus helps to prevent the back-oxidation reaction from taking place. It is still widely thought that heating above 70°C causes loss of 'volatile selenium compounds' in spite of the work by Krivan et ~ l . ~ who used radiotracers to disprove it. They showed that the apparent losses were in fact due to back oxidation of the reduced species.In an earlier p~blication,~ we reported that microwave energy could be used to aid the reduction as part of an on-line flow injection system. Using this technique the analyte-acid mixture flowed through the heating cell for a brief period (a matter of seconds) and was heated to a sub-boiling temperature. This fact prompted the authors to investigate the kinetics of the reaction as any reduction in the heating time would reduce the likelihood of contamination occurring. EXPERIMENTAL Since the introduction of hydride generation the determination of trace levels of selenium has generally relied on the use of a quartz furnace atomic absorption spectrometer (QFAAS) as the detection system. Using this technique with a continuous flow hydride generator a typical analysis would take of the order of 2 min as shown in Fig.1. As can be seen there is an initial delay time of typically 10-15 s in which the sample reaches the switching valve immediately prior to sample introduction into the gas-liquid separator. This is followed by a period in which the detector output reaches a steady state the rise time being typically 15-20 s. The analysis time is generally around 45-60 s during which the measurements are made. At the end of this period the switching valve ceases to supply analyte to the gas-liquid separator and the signal returns to the base line as the gas-liquid separator is cleared of any remaining analyte. The decay time is typically 30-45 s. The method relies on an equilibrium being reached in the gas- liquid separator and so it is difficult to follow anything but the slowest reactions owing to these inbuilt delays.0 50 100 Timds Fig. 1 absorption spectroscopy Typical response curve produced by hydride generation atomic Journal of Analytical Atomic Spectrometry May 1995 VoE. 10 40910 20 30 40 50 Tirne/s Fig. 2 Hydride generation atomic fluorescence peak obtained using a 50 ng g-' solution The use of atomic fluorescence spectrometers that are several orders of magnitude more sensitive than QFAAS offers a rapid response in the presence of an analyte. It is also possible to alter the conditions used for hydride generation and under computer control use very short sample injection times. The resulting peaks resemble flow injection peaks as shown in Fig. 2.Using this technique the equilibrium conditions in the gas-liquid separator are not upset during the very short period in which the sample is being introduced. As can be seen in Fig. 2 the whole analysis is complete within 45 s. As the sample injection only lasts for 2 s a 'snapshot' of the reaction may be obtained. If this is repeated at intervals of 1 min it is possible to build up a picture of the complete reaction process. Instrumentation The apparatus consisted of a hydride generator (Model 10.004 from PS Analytical Sevenoaks Kent) fitted with a Permapure drier tube (PS Analytical) and used in conjunction with an atomic fluorescence system (Excalibur PS Analytical). The instruments were under computer control using Touchstone software (PS Analytical).Reagents The reagents used were sodium tetrahydroborate 98% (Aldrich Gillingham Dorset) sodium selenate 99% (Aldrich) sodium selenite 99% (Aldrich) and hydrochloric acid (Anala Merck Lutterworth UK). Water was 18 Mohm from a Milli-Q system (Millipore Bedford MA USA). Stock solutions (100 pg ml-') of sodium selenate and sodium selenite were prepared. Working strength solutions of 50 ng ml-' were prepared daily from these. Sodium tetrahydro- borate solutions of 1.3% m/v in 0.1 moll-' NaOH were also prepared daily. Hydrochloric acid solutions (6 moll-') were prepared as required. Method A beaker containing 200 ml of 6 moll-' hydrochloric acid was placed in a water-bath and heated to the required temperature. Once thermal equilibrium had been achieved a small aliquot (100 pl) of the stock solution was introduced whilst main- taining continuous agitation.The hydride system was then operated to sample from the beaker every 60 s and the resulting selenium concentrations were stored by computer. This process was carried out in triplicate at each temperature (50 55 60 65 and 70 "C). The software was operated in the peak area mode as it was felt that this would provide greater precision than peak height bearing in mind the very short sampling times used and the sharp peaks obtained. 110 65 "C 60 "C 55 "C 1 100 90 80 70 60 50 40 30 20 10 X 0 5 10 15 20 25 30 Time/min Fig. 3 Graphical representation of the conversion of selenium(v1) to selenium(1v) at various temperatures 5 I 60 "C ' 50°C "C 0 5 10 15 20 25 30 Time/rnin Fig.4 Plots of ln(output maximum- output) against time RESULTS The results of the triplicate experiments were averaged and the detector output [related to selenium(1v) concentration] plotted against time for each temperature as shown in Fig. 3. The curves obtained are exponential in character and so were replotted using ln(output maximum - output) against time as shown in Fig. 4. The resulting straight lines indicate that the reaction is pseudo-first order in character. It can be seen from the Arrhenius equation k = A exp( - E,/RT) where k is the rate constant; A the pre-exponential factor; E the activation energy; R the gas constant and T temperature (K) that the slope of these lines may be calculated to obtain the rate constant. A plot of ln(k) against 1/T also produced a straight line as shown in Fig. 5 the slope of which is equal to E,/R.The result obtained gave a value for the activation energy of the reaction of 90.4kJmol-' using 6moll-' hydrochloric acid. DISCUSSION The activation energy calculated in this study is higher than that reported by Bye & Lund' but lower than the figure obtained by Pettersson and O h 6 (90.4 kJ mol-l c j . 83 kJ mol-' and 126 kJ mol-' respectively). However the experimental conditions with regard to the chemical composi- tion of the test solutions were not identical to those of Bye 41 0 Journal of Analytical Atomic Spectrometry May 1995 Vol. 10- 1 I 0.0029 0.00295 0.00300 0.00305 0.0031 1 T Fig. 5 Plot of k against 1/T and Lund since they were looking at a system containing perchloric acid in addition to hydrochloric acid in an attempt to more accurately duplicate conditions applicable to 'real' samples. In addition their value for the activation energy was carried out using 4 mol 1-1 hydrochloric acid as opposed to the 6mol1-' acid used in this study. The application of the more rapid atomic fluorescence system enabled many more data points to be obtained in our study than in that by Bye and Lund on which to base these calculations.Pettersson and Olin explained their much higher value for the activation energy as a function of uncertainties in tempera- ture since they did repeat the conditions used by Bye and Lund as regards the exact matrix used. They indicated that there was a time lag between the introduction of the flask to the water bath and the contents of that flask reaching a constant temperature and also that the contents of the flask did not reach the temperature of the bath.These obvious differences were noted by the authors of this study which was why the actual temperature of the test solution was measured. Petterson and Olin also indicated that full reduction took 30 min at 65 "C. As can be seen from Fig. 3 our results indicate that at this temperature full reduction is achieved within 10 min under our experimental conditions. From the plots of output against time in Fig. 3 it can be seen that the reaction goes to completion after 6 min at 70 "C. Prolonged heating for periods of 30min or more is therefore unnecessary. Shorter heating times may also reduce the poten- tial for contamination when using open-topped containers. As reported by Krivan et the solution may be boiled under reflux but it is often more practicable when dealing with large numbers of samples to have containers sitting in a water bath rather than have a large array of refluxing systems in operation. Provided the analysis is carried out within a very short time of the reduction taking place no back-oxidation should occur. L.P. would like to thank the Engineering and Physical Sciences Research Council and PS Analytical Ltd. for the provision of a CASE studentship. REFERENCES 1 Holak W. Anal. Chem. 1969 41 1712. 2 Sinemus H. W. Melcher M. and Welz B. At. Spectrosc. 1981 23 81. 3 Krivan V. Petrick K. Welz B. and Melcher M. Anal. Chem. 1985,57 1703. 4 Pitts L. Worsfold P. and Hill S. J. Analyst 1994 119 2785. 5 Bye R. and Lund W. Fresenius' Z . Anal. Chem. 1988 332 242. 6 Pettersson J. and O h A. Talanta 1991 38 413. Paper 4 f07585D Received December 13 1994 Accepted March 1 1994 Journal of Analytical Atomic Spectrometry May 1995 Vol. 10 41 1

ISSN:0267-9477    

DOI:10.1039/JA9951000409

出版商:RSC    

年代:1995

数据来源: RSC

摘要:

CUMULATIVE AUTHOR INDEX JANUARY-MAY 1995 Aboal-somoza Manuel 227 Adams Freddy C. 11 1 Allen Lloyd A. 267 Arbore Philippe 38 1 Arruda Marco A. Z. 55 Barciela-alonso M. C. 247 Barnes Barbara S. 177 Becker-Ross Helmut 61 127 Belazi Abd Ulhafid 233 Bermejo-barrera Adela 227 247 Bermejo-barrera Pilar 227 247 Betti Maria 381 Boonen Sylyie 81 Bordel-Garcia Nerea 3 11 Brueggemeyer Thomas W. 177 Brunetto M. R. 343 Bryant M. F. 295 Bulska Ewa 49 Burden Trevor J. 259 Burguera J. L. 343 Burguera M. 343 Caldwell Kathleen L. 367 Camara Carmen 321 Carrero P. 343 Caruso Joseph A. 7 castle Laurence 303 Cernohorskf TomaS 155 Cervera M. L. 353 Chakraborty Ruma 353 Chan W. T. 295 Cho Jung H. 335 Cho Kyu H. 335 Cimadevilla Enrique Alvarez- Cloud Jacques 287 Corns Warren T 287 Cossa Daniel 287 Crews Helen M.303 Crowe John B. 177 Curtius Adilson Jose 329 Dams Richard 81 Das Arabinda K. 353 Davidson Christine M 233 241 de Bievre Paul 395 de la Calle Guntinas M. Beatriz 111 321 de la Guardia Miguel 353 Ebdon Les 317 145 Cabal 149 Ebihara Mitsuru 25 Efstathiou Constantinos E. 221 Ek Paul 121 Entwistle Andrew 395 Fairman Ben 281 Fang Zheng 359 Fell Gordon S. 215 Ferron-novais M. 247 Florek Stefan 61 127 145 Fordham Peter J. 303 Furuta Naoki 25 Gallego Mercedes 55 Gallignani M. 343 Garcia Alonso J. Ignacio 381 Golloch Alfred 161 G6mez Gomez M. M. 89 Goodall Phillip 3 17 Gramshaw John W. 303 Grazhulene Svetlana S. 161 Greenfield Stanley 183 Grotti Marco 325 Halls David J. 169 Harnly James M. 187 197 Harrison Iain 215 Hayashi Yasuhisa 37 Heitkemper Douglas T.177 Hill Steve J. 317 409 Houk R. S. 267 Huang Meng-Fen 31 HuldCn Stig-Gorari 121 Hwang Chorng-Jev 31 Imai Shoji 37 Imbert Jean-Louis 93 Inoue Yoshinori 363 Ivaska Ari 121 Jackson Kenneth W. 43 Jantzen Eckard 105 Jedral Wojciech 49 Jiang Shiuh-Jen 3 1 Jones Phil 281 Kawabata Katsuhiko 363 Keating Gillian E 233 Kerrich Rob 99 Khvostikov Vladimir A. 161 Kim Ha S. 335 Kim Hyo J. 335 Kinard W. F. 295 Kirschner Stefan 161 Koch Lothar 381 Kotrlf Stanislav 155 Lajunen Lauri H. J. 117 Lee Gae H. 335 Lee Kee B. 335 Lerat Yannick 137 Littlejohn David 215 233 241 Eobinski Ryszard 11 1 Lund Walter 405 Lunzer Florian 311 Lyon Ian C. 273 Madrid Yolanda 321 Mahmood Tariq M. 43 Mao X. L. 295 Marawi Isam 7 Masera Eric 137 Massart D. Luc 207 Mauchien Patrick 137 Mazzucotelli Ambrogio 325 McCartney Martin 233 McCrindle Robert I.399 McCurdy Ed 303 McLaren J.W. 371 McLeod C. W. 89 Miyazaki Akira 1 Moens Luc 81 Monteiro Maria I n b C. 329 Moreda-piiieiro Antonio 227 Okuhara Kyoichi 37 Olson Lisa K. 7 Paama Lilli 11 7 Pang Ho-ming 267 Park Yang S. 335 Parry Susan J. 303 Paschal Daniel C. 367 Pasullean Benyamin 241 Penninckx Wim 207 Peramaki Paavo 1 17 Pereiro-Garcia Rosario 31 1 Perera Indral K. 273 Perez-corona M. Theresa 321 Perkins Charles V. 253 Petzold G. 371 Piiri Lindy 117 Pin Christian 93 Piperaki Efrosini A. 221 Pitts Les 409 Powell J. J. 259 Prange Andreas 105 Qiao Huancheng 43 Rademeyer Cornelius J. 399 253 Radziuk Bernard 127 197 Rivas C. 343 Rodel Giinther 127 Rondon C. 343 Russo R. E. 295 Sadler Daran A. 253 Saito Kengo 37 Sanjuan Jane 287 Sanz-Medel Alfredo 149 281 Schmecher Gisela R.61 Sena Fabrizio 381 Seubert A. 371 Smeyers-Verbeke Johanna 207 Smith Clare M. M. 187 Smith Monica M. 349 Stenz Herbert 127 Stockwell Peter B 287 317 Suzuki Yoshihito 363 Tanaka Toshiyuki 37 Tao Hiroaki 1 Taylor P. D. 259 Taylor Philip D. P. 395 Telgheder Ursula 161 Telouk Philippe 93 Thomaidis Nikolaos S. 221 Thompson Diana 303 Thompson K. Clive 317 Thompson R. P. H. 259 Ting Bill G. 367 Tischendorf Reinhard 61 Turner Grenville 273 Uchino Tomonori 25 Uggerud Hilde 405 Valcarcel Miguel 55 Vanhaecke Frank 8 1 Vankeerberghen Peter 207 Walder Andrew J. 395 Wang Jiansheng 7 Wang Yun-Zhou 359 Warren Arnold R. 267 White Mark A. 349 Wilson H. Kerr 349 Wolnik Karen A. 177 Worsfold Paul 409 Wrbbel Katarzyna 149 Xie Qianli 99 Zhang Ke 359 Zhang Yuan-Fu 359 311 Journal of Analytical Atomic Spectrometry May 1995 Vol. 10 41 3

ISSN:0267-9477    

DOI:10.1039/JA9951000413

出版商:RSC    

年代:1995

数据来源: RSC