摘要:

1The PC'1 Version of the ~~ ~ ~NIST/EPA/MSDCMass SpectralDatabaseNational Institute of Standard5 and Technology USAEnvironment Prolection Agency USAMass Spectrometry Data Centre UKThis unique database features:The mass spectra of 53,9940 retrieval in seconds@ both tabular and graphic display0 fully interactive searching by:compounds- individual peaks- abundances of 10 major peaks- molecular weight- molecular formula- compound name- CAS Registry Number- identification numberplus a complete sequential searchof the entire databasechemical structure displays for morethan 96% of compoundsa facility to create your own databasealongside the NIST/EPA/MSDCDatabase - can search eachindividually or both togetherand much more/I A free demonstration disk * * is available on requestSEND FOR FURTHERDETAILS NOW!Please contact:The Mass Spectrometry Data CentreRoyal Society of ChemistryThomas Graham HouseScience Park, Milton RoadI ROYALSOCIETY OFCHEMISTRY IBOOKS FROM W I L E YHigh Performance LiquidChromatography in BiotechnologyEdited by W.S.HANCOCK, Gennentech Inc, California, USAThis is an extremely useful analytical book and it is particularlyso in all areas of the life sciences and biotechnology. This is dueto the fact that this system provides the potential for faster morereliable separations and many important changes anddevelopments have occurred in these practical applicationsrecently. The book is designed to meet the high performance liquidchromatographers daily needs and should save a good deal of timeand effort.0471825840 576pp 1990 574.35/$112.25Unified Separation ScienceJ.C.GIDDINGS. University of Utah, UKThis is an advanced text by one of the most foremost experts inseparation science. It unifies the complex welter of techniques usedfor chemical separations by clearly formulating the concepts thatare common to them. The mass transport phenomena underlyingall separation processes are developed in a simple physical-mathematical form. The limitations and optimum performance ofalternative separation techniques and the factors enhancing andlimiting separation power can thus be described and explored.0471520896 approx 368pp due 1991 approx 542.15/$63.20An Introduction to LaboratoryAutomationV. CERDA, University of the Balearic Islands, Spain, and G.RAMIS.University of Valencia, SpainIntroduces the reader to the basic principles and practicaltechniques of automating the chemical laboratory. It also includestechniques for connecting instruments to a computer, dataacquisition, communication protocols, laboratory robotics andexamples from analytical chemistry. Both control and measurementtechniques are included.Chemical Analysis Series0471618187 336pp 1990 555.15B83.45Concepts and Applications ofMolecular SimilarityEdited by M.A. JOHSSON and G.M. MAGGIORA. both of theGpjohn Company, Michigan, USAThis volume offers authoritative overviews of topics related to thedefinition, computation, and application of molecular similarityand emphasizes current research trends with molecular similarityas the unifying concept.0471621757 416pp 1990 551.35B77.75Gas Chromatography: Biochemical,Biomedical and ClinicalApplicationsEdited by R.E.CLEMENT, Ontario Ministry of the Environment,CanadaThis book describes the techniques of gas chromatography as theyare applied to biochemical, biomedical and clinical studies. Thesubject matter can be divided into techniques and applications.The first half of the book is largely devoted to the instrumentationand equipment used for gas chromatography analysis, while thelatter half describes specific applications.Volume 111 in Chemical Analysis: A Series of Monographs onAnalytical Chemistry and Its Applications0471010480 406pp 1991 f70.50/$106.50iley books are available through your bookseller W Alternatively order direct from Wiley (payment to JohnWiley 61.Sons Ltd). Credit card orders accepted by telephorw -(0243) 829121 or FREEFONE 3477. Please note that pricesquoted here apply to UK and Europe only.WILEYPublisks Since 1807JOHN WILEY & SONS LTDBAFFINS LANE . CHICHESTERWEST SUSSEX PO19 1UDI Cambridge CB4 4WFUnited KingdomTelephone: (International) +44 (0)223 420066Fax: (International) +44 (0)223 423623Telex: 81 8293 ROYAL Serviceslnformat ionCircle 001 for further informationCircle 002 for further informatioPUBLISH IN THE ANALYST-THE ANALYSTWELCOMES PAPERS ON:00Biochemical analysis EPR and ESRC hemometricsMass spectrometry ChromatographyVibrational spectroscopy ElectrochemistryAtomic & molecular absorption spectroscopyEditor, The Analyst, Harpal S.Minhas,The Koyal Society o f Chemistry.Thomas Graham House. Science Park.Milton Road, Cambridge CB4 4WF. UKTel: 0223 420066. Fax: 0223 423623.Tlx: XlX2Y.3 ROYALUS Associate Editor, Dr. J. F. Tyson,Department o f Chemistry.University o f Massachusetts. Amherst,MA 01003, USATel: 413 545 01%. Fax: 413 545 44W@a- *& ROYALSOCIETY OFCHEMISTRYInformationServicesEditorial Manager, Analytical Journals: Judith EganCircle 003 for further informatio...111The XXVII Colloquium Spectroscopicum InternationaleXXVII CSI bwill be held inGrieg Hall, Bergen, NorwayJune 9-14 1991 IUPAC199 1NORWAYThis traditional biennial conference in analytical spectroscopy will once again provide a forum for atomic, nucicar andmolecular spectroscopists worldwide to encourage personal contact and the exchange of experience.Participants are invited to submit papers for presentation at the XXVII CSI, dealing with the following topics:Basic theory and instrwnentaion of-Atomic spectroscopy (emission, absorption, fluorescence)Molecular spectroscopy (UV, VIS and IR)X-ray spectroscopyGamma spectroscopyMass spectrometry (inorganic and organic)Electron spectroscopyRaman spectroscopyMoss bauer spectroscopyNuclear magnetic resonance spectrometryMethods of surface analysis and depth profilingPhotoacoustic spectroscopyMetals and alloysGeological materialsIndustrial productsBiological samplesFood and agricultural productsApplication of spectroscopy in the analysis of-Special emphasis will be given to trace analysis, environmental pollutants and standard reference materials.The scientific programme will consist of both plenary lectures and parallel sessions of oral presentation. Specific times will bereserved for poster sessions.PRE- AND POST-SYMPOSIAIn connection with the XXVII CSI the following symposia will be organised:Pre-sy m p o s i eI.GRAPHITE ATOMISER TECHNIQUES IN ANALYTICAL SPECTROSCOPYJune 6-8, 1991, Hotel Ullensvang, Lofthus, Norway.II. CHARACTERISATION OF OIL COMPONENTS USING SPECTROSCOPIC METHODSJune 6-8, 1991, Hotel Hardangerfjord, e)ystese, Norway.111.MEASUREMENT OF RADIO-NUCLIDES AFTER THE CHERMOBYL ACCIDENTJune 6-8, 1991, Hotel Solstrand, Bergen, Norway.Post-syrnposiurr+IV. SPECIATION OF ELEMENTS IN ENVIRONMENTAL AND BIOLOGICAL SCIENCESJune 17-19, 1991, Hotel Alexandra, Loen, Norway.For further information contact:THE SECRETARIATXXVll CSIHSD Congress-ConferenceP.O. Box 1721 NordnesN-5024 Bergen, Noway.Tel. 47-5-31 841 4, Telex 42607 hsd n, Telefax 47-5-3245553rcle 005 for further informatioTHE ANALYST READER ENQUIRY SERVICE FEB'91For further information about any of the products featured in the advertisements in this issue, please writethe appropriate number in one of the boxes below.Postage paid if posted in the British Isles but overseas readers must affix a stamD.PLEASE USE BLOCK CAPITALS LEAVING A SPACE BETWEEN WORDS Valid 12 months1 NAME__ p~ 2 COMPANVPLEASE GIVE YOUR BUSINESS ADDRESS IF POSSIBLE.IF NOT, PLEASE TICK HERE4 TOWN5 COUNTYPOST CODE6 COUNTRY7 DEPARTMENTDIVISION8 YOUR JOB TITLEPOSITION9 TELEPHONE NOFOLD HERE1I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIPostagewill bepaid byL i censeeDo not affix Postage Stamps if posted in Gt. Britain,Channel Islands, N. Ireland or the Isle of ManBUSINESS REPLY SERVICELicence No. WD 106Reader Enquiry ServiceThe AnalystThe Royal Society of ChemistryBurlington House, PiccadillyLONDONW1E 6WFEnglandIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII!IIIIIIIIIIIIIIIIIIIIIIIIIIIIII1IIIIII

ISSN:0003-2654    

DOI:10.1039/AN99116BP003

出版商:RSC    

年代:1991

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN99116FX005

出版商:RSC    

年代:1991

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN99116BX007

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1991. VOL. 116 117 Construction and Evaluation of a Regenerable Fluoroimmunochemical-based Fibre Optic Biosensor James R. Bowyer, Jean Pierre Alarie and Michael J. Sepaniak" The University of Tennessee, Department of Chemistry, Knoxville, TN 37996- 1600, USA Tuan Vo-Dinh Health and Safety Research Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6101, USA Robert 0. Thompson Department of Chemistry, Oberlin College, Oberlin, OH 44074, USA A microscale fibre optic biosensor that is capable of in situ regeneration is described and characterized. By combining recently developed fibre optic sensing technology with a capillary column reagent delivery system, it is possible t o perform a variety of bench-top affinity assay procedures both repetitively and remotely.The configuration of the sensing chamber at the terminus of the fibre is an important design feature. The construction and operation of the sensor is described and the results of evaluations of the sensor using an antibody-antigen system are presented. Affinity assay steps such as the delivery of solid phase affinity reagents, secondary reagents and rinse solutions are demonstrated. Sampling is accomplished by mild aspiration. Relative standard deviations (RSDs) for these steps are all less than 10%. The capability of selectively measuring fluorescently labelled anti-rabbit immunoglobulin G (IgG), in the presence of a similar protein, by utilizing its immunospecific interaction with rabbit IgG immobilized on silica beads is demonstrated, and exhibits an RSD of 6.2%.A near linear calibration graph is presented over a concentration range of between 0.01 1 (approximately the limit of detection) and 0.1 1 mg ml-7. Keywords: Fibre optic; biosensor; fluoroimmunoassa y; laser; immunoglobulin G The ability to monitor chemical concentrations spectrally both at a distance and in situ has been enhanced with the advent of small diameter optical fibres that transmit light efficiently over long distances and wide spectral regions. Such measurements offer advantages over traditional approaches that involve sampling, then transporting the sample to the laboratory for subsequent analysis. In particular, chemical concentrations can be measured directly in hostile or not easily accessible environments without the chemical or physical alterations of sample composition commonly associated with traditional approaches. Ideally, the measurements are performed con- tinuously without removing the fibre optic 'sensor' from the remotely located sample.Unfortunately, in situ measure- ments can be complicated and generally do not exhibit the analytical sensitivity and selectivity associated with conven- tional laboratory techniques. These problems have been addressed with the development of fibre optic chemical sensors (FOCSs).1-3 The FOCS signals are a result of the interaction of an analyte with a reagent phase. The chemical and physical specificity afforded by the interaction contributes to the selectivity of the measurement, and high sensitivity is achieved if the interaction results in a signal that can be detected by laser excited fluorimetry.The reagent phase is immobilized at the sensing terminus of the optical fibre in a variety of ways including direct covalent attachment,l entrapment in a gel5 and containment in an analyte permeable chamber.6 Traditionally, FOCSs have been used to measure small molecules or ions.7-11 The application of FOCSs to the measurement of large molecules has been accomplished recently by using bioaffinity reagent phases.3 Among the analyte-reagent phase combinations that have been employed are lectin-carbohydrate, 12.13 enzyme-sub- strate,14-16 and antibody-antigen or -hapten. 17-23 Recent work has focused on the use of immunochemical reagent phases to perform remote measurements by competitive- binding17 and direct4.6 assay procedures. The specificity of the immunochemical recognition (antibody-antigen association constant, K,, typically in the range 108-1012) resulted in excellent sensitivity and selectivity in the previous work.* To whom correspondence should be addressed. In order for a sensor to be used in a continuous fashion one of the following criteria must be met: one, the interaction of the analyte and reagent phase must be rapid and reversible, thereby permitting a competitive equilibrium binding opera- tion;24 two, the reagent phase reservoir must not be appre- ciably depleted during the measurement period; or three, the sensor must be capable of in situ regeneration. Many of the sensors that have been described in the literature do not adequately satisfy any of these criteria and are more correctly termed probes.In general, our previous immunochemical- based FOCSs fall into this category. Two approaches to continuous monitoring with immunochemical-based FOCSs have been reported. Anderson and Miller21 utilized a competi- tive equilibrium binding approach to monitor continuously the drug phenytoin with an antibody-based FOCS. Sensor opera- tion was based on homogeneous fluoroimmunoassay (FIA) principles and involved competition between labelled and unlabelled drug for the antibody. We have demonstrated the feasibility of using a microscale regenerable biosensor (MRB) to perform rapid, repetitive (i. e., pseudo-continuous) measurements of a benzo[a]pyrene metabolite .25.26 Operation of this sensor is based on heterogeneous FIA procedures.Each approach has its advantages and disadvantages. The competitive equilibrium binding approach requires a compro- mise in K , value, as rapid response favours small values, whereas sensitivity and selectivity are enhanced with large values.3 The MRB is versatile and can exploit the advantages of antibodies with large values of K,, but its operation can be relatively complicated. 'The MRB combines typical FOCS instrumentation with a reagent delivery system that employs capillary columns. In principle, the many FIAs that are used extensively in clinical laboratories27 can be performed remotely, in situ, and in very small samples with the MRB. The appropriate FIA protocol for a given analysis depends largely on the nature of the analyte.Natural fluorophores can be measured using a direct assay procedure that does not involve the complication of labelling reagents with fluorophores. The determination of non-fluorescent analytes can be accomplished by either competitive binding or sandwich assay techniques.3 The former is used for small analytes and involves competition118 ANALYST, FEBRUARY 1991, VOL. 116 between labelled and unlabelled analyte for a limited amount of antibody. The antibody is usually immobilized on a support to facilitate the removal of any unbound reagents. Dose- response graphs have a negative slope and exhibit short dynamic ranges. Labelled and unlabelled materials are generally mixed simultaneously, a process that does not easily lend itself to in situ analysis. Nevertheless, we used an immunochemical-based FOCS to perform remote, single- measurement, competitive-binding assays.17 Sandwich assays are generally performed using a solid-phase antibody (immo- bilized on a support) that binds the analyte, and a second antibody that is fluorescently labelled and which ‘tags’ the analyte for measurement. Dose-response graphs have a positive slope and a moderate dynamic range. However, the analyte must be fairly large (relative molecular mass >lO3) so that the two antibodies can recognize different epitopes on the analyte.3 As the labelled reagent can be added subsequent to the initial incubation, the sandwich assay procedure is more easily adapted to continuous sensing. We are currently exploring the use of the MRB for performing sandwich-type FIAs.As described above, assay protocols can vary greatly, but generally employ affinity reagents immobilized on solid supports (immunobeads in this work) and involve the addition of secondary reagents and rinse solutions. The analytical significance of the MRB can be seen by comparing its operation and characteristics with that of one of the most popular instruments for performing FIAs. In the early 1980s, Pandex developed a versatile method of using immunobeads to perform FIAs.28 With the Pandex method, the immuno- beads are mixed with the sample and various reagents and placed in funnel-shaped microtitre plate wells that contain a frit at the bottom. The reagents and rinse solutions are drawn through the frit and the signal emanating from the trapped immunobeads is measured by front-surface fluorimetry .The microtitre wells are arranged on a plate that permits many samples to be analysed in a short period of time. It is very significant that the operational protocols for the MRB resemble those of the instrument used with the Pandex method. Immunobeads, secondary reagents, and rinse solu- tions can be added in different orders with the MRB in order to perform a variety of FIAs. Although large numbers of samples cannot be analysed, repetitive measurements can be performed remotely in very small volume samples. The principal aim of the work presented herein is to demonstrate the feasibility and to evaluate the repeatability of performing isolated and combined affinity assay operations with the MRB. The general design of the MRB is also described.Calibration graphs were obtained for a rabbit immunoglobulin G-anti-rabbit immunoglobulin G (rab IgG- anti-rab IgG) immunological system. These graphs were used to evaluate the response range and sensitivity of the MRB. By rinsing to expel non-specific interferents (compounds that do not bind to the immobilized affinity reagent), the specificity of the affinity-analyte interaction is exploited in order to measure the analyte selectively. Another advantage of the MRB is that reproducible sampling of the analyte can be accomplished by an aspiration procedure. Previously de- scribed sensors3 sampled the analyte by passive diffusion, a procedure that has limited applicability as many analytes exhibit low diffusivities either because of their large size or a lack of sensor permeability.Future prospects for the MRB are also briefly discussed in this paper. Experimental Materials used in this work were of 400 pm core diameter, plastic-clad fused silica fibres (QSF-400), supplied by General Fiber Optics, Cedar Grove, NJ, USA. The epoxy used was a 5 minute Epoxy manufactured by Devcon, Danvers, MA, USA. Syringe needles were of a standard size 16-gauge Luer lock variety. A fibre optic column bundle template was constructed in-house from Lexan. Stainless-steel frits ( 5 pm porosity) were obtained from Newmet Krebsoge, Terryville, CN, USA. A Model 7010 Rheodyne high-performance liquid chromatography (HPLC) injection valve and Luer lock syringe valves (Mininert syringe valves, Catalogue No.654051) were purchased from Alltech, Deerfield, IL, USA. The injection valve, fitted with a 50 pl injection loop, was used for bead injection. Plexiglas T-connectors were fashioned in-house. The syringe pump, for sample and reagent introduc- tion, was obtained from Sage Instruments, Cambridge, MA, USA (Model 341A). Phosphate buffered saline (PBS), pH 7.4; fluorescein isothiocyanate (FITC); anti-rabbit immuno- globulin G-fluorescein isothiocyanate (anti-rab IgG-FITC), of fluorophore to protein (F: P) ratio of approximately 3; rabbit immunoglobulin G-fluorescein is0 t hiocy ana te (rab IgG-FITC), F : P ratio of approximately 4; rabbit immunoglob- ulin G (rab IgG) (lyophilized); and human immunoglobulin G-fluorescein isothiocyanate (human IgG-FITC), F : P ratio of approximately 4, were obtained from Sigma, St.Louis, MO, USA. Human blood serum was donated by a single subject. Immunobeads were prepared by immobilizing the immunospecific reagent rab IgG on silica beads ( 5 pm diameter) using a previously reported procedure.29 Fluor- escently labelled latex beads used in the bead delivery studies were Fluoresbrite plain microspheres (6.29 pm diameter; excitation wavelength, 488 nm; emission wavelength, 525 nm) and were obtained from Polysciences, Warrington, PA, USA. Instrumentation The instrumentation consisted of an argon ion laser operated at 488 nm (Model 2001SL, CyonicsKJniphase, San Jose, CA, USA), a photomultiplier tube (PMT) (R943-02, Hamamatsu, Middlesex, NJ, USA) and housing (Thorn EMI, Fairfield, NJ, USA), a chopper (Model 9479, EG&G Ortec, Oak Ridge, TN, USA), various lenses and optical filters (Corion, Hollis- ton, MA, USA), a 25 mm diameter mirror with a 2 mm hole bored through the centre (prepared in-house), a monochro- mator (Model H-10, Instrument SA, Metuchen, NJ, USA), a strip-chart recorder (Cole-Parmer Instruments, Chicago, IL, USA), PMT-HV power supply (Model 556, EG&G Ortec, Oak Ridge, TN, USA), and a picoammeter (Model 485, Keithley, Cleveland, OH, USA).The arrangement of these components is shown in Fig. 1 and is similar to a previously described arrangement .6 n E A - v B C> -D I F H I capillary columns used in the MRB Were of various sizes ranging from 2oo to 520 pm i-d- and were purchased from Polymicro Technologies, Phoenix, AZ, USA. The fibre optics Fig.1 Schematic diagram of the optical configuration used with an MRB. A, Argon ion laser (488 nm); B , chopper; C, mirror with hole; D, focusing optics; E, fibre positioner; F, MRB; G, monochromator and PMT; H, picoammeter; and I , strip-chart recorderANALYST, FEBRUARY 1991. VOL. 116 119 Sensor Conditioning Procedure Before using the MRB, all tubes were connected to their respective reservoirs and filled with water to remove any trapped air. It is important that air be removed to avoid problems with ‘signal altering voids’ in the system. Water was also forced through the frit to remove air trapped in the pores. In order to ensure there were no leaks, water was aspirated through the frit and any air bubbles observed in the capillaries were an indication of a leak in the system.Throughout this procedure reagent delivery and rinsing steps are performed with the outlet columns sealed using the syringe ordoff valves, thereby restricting flow through the frit; flushing steps are performed with the outlets opened. Bead sa Aspii mple Frit ?ation and introduction Fibre optic- U Frit Aspiration T-connector injection Jk loop \ pro ‘Bead inlet .Rinse and flush inlets ‘Outlets Jction ‘Aspiration R i nse/f I ‘us h i n I ets -Outlets 1 Syringe on/off valve- Fig. 2 chamber and (b) reagent delivery system Diagram of ( a ) frit and column configuration at MRB sensing Construction Construction of the MRB (see Fig. 2) is begun by affixing six short sections of a 200 pm i.d. capillary column and the fibre optic into a template such that they protrude through the template by approximately f to 3 in.A reasonably symmetrical tip is formed by smoothing the applied epoxy along the columns, thereby forming a bundle with the periphery defined by the columns (see Fig. 2). As the epoxy hardens, the bundle is drawn out of the template slightly and more epoxy is added in the same manner. After carrying out this procedure, the tip is sufficiently long to allow grinding and polishing without losing the symmetrical arrangement. In order to ensure that the bundle fits tightly in the stainless-steel frit (shown in Fig. 2), the template is constructed with the hole arrangement slightly smaller than the frit opening. Once the epoxy has set, the bundle is removed from the template and ground to a flat polished surface using a grinding wheel with fine lapping film; during this process, the tip is sized using the frit.In order to keep the columns from becoming plugged with debris during the grinding, water is pumped through the columns. In order to facilitate the operation of the MRB, different diameter capillary columns are connected to the 200 pm i.d. columns just beyond the epoxy; two 520 pm i.d. columns are used as outlets, three 320 pm i.d. columns are used as inlets (two are used as rinsing inlets and one is used for introduction of beads), and one of the original 200 ym i.d. columns is used to aspirate a given volume of sample (see below). Outlet and inlet columns are secured in 16-gauge needles with epoxy and terminated with a syringe ordoff valve. The capillary column for bead injection is secured in a short length of tubing and, by use of a suitable ferrule and fitting, is attached to the HPLC injection valve.Column lengths are arbitrary except for the aspiration column, which has a length dictated by the desired aspiration volume and is attached to an aspiration T-connec- tor. Construction of the MRB is completed by affixing the hollowed frit on the polished capillary column-fibre optic bundle such that a small volume (approximately 1 pI) sensing chamber is formed (see Fig. 2). The operation of the MRB for each of the main evaluation steps is described below. Reagent Phase Delivery The first evaluation of the MRB involved determining the repeatability of reagent delivery to the sensing chamber. The ability to rinse and then flush the chamber was also evaluated.In this study, bead slurry and liquid reagent solutions are individually placed in the sensing chamber via their respective delivery capillaries. The Fluoresbrite beads (50 pl, 8 mg ml-1) are introduced from the HPLC injection valve into the chamber. Once a stable fluorescence signal is obtained and recorded the chamber is rinsed with approximately 25 yl of solvent. When performing actual assays, rinsing is necessary to remove unreacted material and possible interferents. After the rinse is completed, a second signal is recorded; the tip is then cleared by flushing, and the process repeated. The repeatability of delivering liquid reagents was similarly evaluated. The FITC solution (1 x 10-4 mol dm-3), used in the evaluation, is introduced from a reservoir and not as an aspirated sample.This operation mimics the introduction, and subsequent removal, of an excess of secondary reagents, such as a labelled second antibody in an MRB-implemented sandwich assay. Sampling by Aspiration Samples are collected by connecting a 30 cm length of one of the 200 pm capillary columns (volume, approximately 10 pl) to the T-connector configuration shown in Fig. 2, and then aspirating with a 5 ml syringe while the sensing tip is in the sample. Once the sample has been drawn past the junction in the T-connector, the tip is rinsed and a background signal recorded. The sample is then delivered via the aspiration tube from the junction in the T-connector. This operation reduces variations in the sample volume resulting from experimental variables such as frit permeability. The sample and rinse (approximately 50 pl) are collected in 1 ml calibrated vials and diluted to the mark prior to spectrophotometric measurement using an ultraviolet/visible diode array spectrophotometer (Hewlett-Packard, Model HP 8452, Palo Alto, CA, USA).By comparing the absorbances with a calibration graph, sampled volumes are determined. Combined Affinity Steps After demonstrating the feasibility and determining the repeatability of performing isolated affinity steps, the steps were combined to measure fluorescently labelled anti-rab IgG and a possible interferent, human IgG-FITC. In this study, the injection loop is filled with immunobeads (10 mg ml-I), the sample is collected, any excess of sample is rinsed from the sensing chamber and the immunobeads are delivered to the sensing chamber.Once the beads are in place, a syringe pump is used to deliver slowly (flow-rate approximately 10 p1 min-1) the aspirated sample, followed by the rinse solution, to the immunobeads in the sensing chamber. At several points during the rinsing step, the flow is stopped and the signal from the immunobead-bound anti-rab IgG-FITC is recorded. The120 ANALYST, FEBRUARY 1991, VOL. 116 beads are then flushed from the chamber and the background signal is recorded. This procedure is also used to obtain a dose-response graph for anti-rab IgG-FITC. In this and other studies, the incident radiation is chopped (duty cycle about 6%) in order to avoid any photo-decomposition of the reagents.Results and Discussion The analytical attributes afforded by fluoroimmunoassays (e.g., high sensitivity and selectivity) are all potentially available with the MRB. The purpose of this work is to demonstrate that the steps common to FIAs can be performed individually and in combination in a repeatable fashion with the MRB, thereby illustrating its versatility as an analytical tool. Detectability, calibration capability, and the selectivity afforded by the specificity of immune reactions, are also demonstrated. Future reports will emphasize the utilization of the MRB for specific assays. Isolated Affinity Assay Steps The performance of FIAs with the MRB requires that: ( i ) controlled and adjustable amounts of immunobeads can be delivered to the sensing chamber; (ii) secondary reagents can be reproducibly delivered to the chamber; (iii) the chamber can be rinsed to remove excess of liquid-phase reagents and impurities that do not bind to the immunobeads; (iv) sample solutions can be accurately and precisely aspirated, then delivered to the immunobeads; and ( v ) the contents of the chamber can be flushed completely so that the process can be repeated.The results of an evaluation of these isolated affinity assay steps are presented in Table 1. Fluorescently labelled beads (Fluoresbrite) were delivered to the chamber five times with a relative standard deviation (RSD) Of 7.7% (first column in the table). Following each delivery, the beads were rinsed without being removed from the chamber (third column), and were then flushed from the sensor.The complete flushing of the beads sometimes required mild aspiration to remove beads lodged in the sensing chamber. Delivery of solutions to the beads without removing them from the field of view of the fibre optic was critical to the operation of the MRB. In order to improve precision in an actual assay, the beads can be labelled with a spectrally distinct fluorophore to provide some normalization for the amount of beads introduced. The second column in the table illustrates the result of delivering an FITC solution to the chamber five times. The precision of this operation was very good (RSD = 0.6%) and, as the fourth column indicates, the FITC solution was efficiently removed from the chamber by rinsing.Reproducible sample volume collection is also critical to the operation of the MRB, particularly as there is no convenient means of normalizing for the volume of sample taken in situ. Whereas previous sensors have collected the analyte via slow passive diffusion through permeable materials, the MRB has a unique analyte aspiration function which is rapid, and independent of analyte diffusivity and permeability variations. In fact, the capability of ‘cleaning-out’ the MRB frit in situ, by rinsing, represents a significant advantage of this sensor. The last three columns of Table 1 illustrate the precision of aspirating small and large volumes of analytes in water (series 1 and 2) and in blood serum (series 3). The RSDs for five aspirations were always less than 8.0% with the biological matrix exhibiting the lowest RSD.The differences in sampled volumes and the RSD are probably due to the size and adhesion characteristics of the IgG, as compared with the FITC, and the viscosity of the matrix in series 3. Combined Assay Steps (Measurement of Anti-rab IgG-FITC) Having verified that the MRB can accomplish commonly encountered steps in FIA protocols, experiments were con- ducted to demonstrate that the operations can be combined to conduct an actual assay. In this experiment, rab TgG was covalently bound to silica beads, using a 1 ,l’-carbonyldiimid- azole linkage that has been shown to bind about 12 mg of IgG per gram of beads,29 and anti-rab IgG was measured. Although not studied in this work, proteins are normally measured by using either a competitive-binding or a sandwich assay procedure.Alternatively, IgG can be measured using the native fluorescence of the protein; however, the quantum efficiency is low and the maximum excitation wavelength (approximately 280 nm) is not convenient. Thus, FITC- labelled antibody was chosen as the analyte (anti-rab IgG- FITC, at approximately 0.1 mg ml-1) in this measurement, which represents the direct assay of a large molecule. A similar protein (human IgG-FITC, at approximately 0.1 mg ml-1) was chosen as an interferent. The advantages of employing this system include: the availability and low cost; the compatibility with the argon ion laser source; the established immobilization procedure;29 and the availability of structurally similar potential interferents (see below).Moreover, the sensing of IgG is clinically significant30 and the measurement of a large protein with an FOCS is novel and demonstrates the aforementioned advantages of sampling by aspiration. Table 1 Signal and reproducibility data for performing isolated affinity assay steps with the MRB (for details see under Experimental) Delivery of reagents Rinsing of reagents (signal/nA) (signal/nA)* Aspiration volume/pI? Measurement Solid phase$ Liquid phase3 Solid phase Liquid phase Series 1 Series 2 Series 3 - - 0 1 510 25.2 2 560 25.2 3 470 25.3 4 550 25.2 5 470 25.3 .f 510 25.3 RSD 7.7 0.6 - - - - 3.78 510 3.70 12.3 10.5 13.4 560 3.83 13.4 12.9 13.4 470 3.78 11.7 11.5 13.9 550 3.73 11.9 11.4 13.2 470 3.78 12.1 11.4 13.0 510 3.77 12.0 11.9 13.0 7.7 1.2 5.4 7.5 2.5 * A 20-30 pl volume of water delivered to sensing chamber via inlet capillary columns (bead signal unchanged, FITC expelled).Between t Aspiration of 1 x mol dm-’ (series 2); and $ A 50 yl volume of 8 mg ml-*, 6 ym, Fluoresbrite beads delivered to sensing chamber. § A 20 p1 volume of 1 X measurements, a 50-60 yl volume of water delivered to sensing chamber with outlet capillaries open to flush the sensor. 5 mg ml-1 rab IgG-FITC in human blood serum (series 3). mol dm-3 FITC (series 1); 5 mg ml-1 rab IgG-FITC, FITC concentration = 1.4 X mol dm-3 FITC delivered to sensing chamber via inlet capillary columns with unlabelled beads in chamber.ANALYST, FEBRUARY 1991. VOL. 116 121 Table 2 Demonstration of combined affinity assay steps with the MRB (for details see under Experimental) Rabbit IgG affinity beads Non-affinity beads Anti-rab IgG- Human IgG- Anti-rab IgG- FITC FITC FITC Measurement A" B t C$ A B C A B C 1 13 90 21 16 140 16 15 100 16 2 14 91 23 15 138 16 14 98 15 3 13 92 20 15 138 15 14 100 14 4 13 96 20 15 136 15 14 98 14 5 12 100 22 16 138 17 14 96 14 X 13 94 22 15 138 16 14 98 15 RSDg 5.4 4.3 6.2 3.6 1.0 5.3 3.1 1.7 6.1 * A, Background signal.-t B, Aspiration signal. measured using MRB following aspiration (approximately 10 PI) and before delivering beads to the sensing chamber. $ C, Post delivery signal, measured after delivering beads, aspirated sample and rinse solution to the sensing chamber. 5 Values for RSD are given as percentages. 150 Q 100 C 0 v) Q) .- .- w 4 50 U 0 B 6 ' 1 rnin ' 2 min '10 min'20 rnin' Begin new run Fig.3 Temporal representation of the signals obtained in the assay of anti-rab IgG-FITC (see Table 2 for further details). A, Anti-rab IgG-FITC reacted with immunobeads; B, human IgG-FITC reacted with immunobeads; and C, anti-rab IgG-FITC reacted with silica beads. Arrows indicate: 1, sample introduction followed by rinse; and 2, tip flush The procedure described in the Experimental section was used to obtain the data in Table 2. The signal levels shown in the table represent three points in the analysis and are recorded for the test system and two immunologically non-specific systems. The three measurements are: the background level; the signal after the aspiration step; and the signal after immunobeads, sample and rinse solution are delivered to the chamber.Recall that the aspiration first fills (or partly fills depending on flow patterns) the sensing chamber, then fills the aspiration capillary. As the immuno- beads are delivered, the sample that occupies the sensing chamber is expelled and the signal drops to the background level. When the aspirated sample is delivered to the chamber, the signal increases dramatically then drops during rinsing as the excess of sample analyte is removed from the chamber. The change in signal during operation is described later with reference to Fig. 3. Table 2 provides encouraging data, and an indication that refinements in the sensor chamber configura- tion and optimization of the operating procedures are necessary. An important point illustrated in Table 2 is that the precision is actually better for the combined affinity steps than for the isolated aspiration step (the RSD for immune specific assay is 6.2%).This indicates that the combined variance in a particular assay is not necessarily an additive function of those for the isolated steps. This is not surprising as the evaluation of the isolated steps does not perfectly mimic the steps in every assay. In particular, the effects of differences in the amount of anti-rab IgG-FITC solution aspirated in the present study are reduced, because, as the sampled volume passes through the immunobeads, the most readily observed beads are saturated with the analyte and, subsequently, beads that are less easily viewed by the fibre optic are involved in the immune reaction. Nevertheless, the wide variety of available assay protocols validates the previous evaluation of precision of the isolated assay steps, as it can provide some insight into the expected assay precision.A second notable advantage of the MRB in this assay is the selectivity afforded by the specificity of the immune reaction. After rinsing, the anti-rab IgG-FITC signal is appreciable, while the assay of the similar protein, human IgG-FITC, is statistically negligible. This indicates that the immunobeads can be rinsed to remove non-specific interferents while retaining the analyte. The blank signal for anti-rab IgG-FITC, obtained using non-affinity beads, is also negligible. This excellent selectivity is particularly important in sensing applications as, unlike conventional analyses, bench-top isolation of analyte via extractions, chromatographic separa- tions, etc., is difficult or impossible.Hence, samples are inherently complex and analyses susceptible to interference. Fig. 3 readily illustrates several points that are not apparent from Table 2. The first is that the signal obtained from the sampled anti-rab IgG-FITC, when initially delivered to the sensing chamber that is filled with immunobeads, is higher than the aspiration signal. This is probably due to the reaction and concentration of the analyte on the beads closest to the fibre before the beads redistribute during the rinse. The non-specific human IgG-FITC and the non-affinity beads do not exhibit this signal pattern. Variation in the initial heights of the aspiration signals for the two types of IgGs is due to the variation in the labelling ratios. The second is the decrease in the signal as the rinsing continues.This occurs despite the amount of immunobead-bound antibody in the sensing chamber being several times the amount of anti-rab IgG- FITC in the 10 ~1 sample. The decrease is associated with removal of the unbound protein, removal of the specifically bound protein and redistribution of the beads. Having removed the unbound protein in the first 10 min of the rinse (see Fig. 3), removal of specifically bound protein does not seem to be a major contributor to the signal decrease. This can be seen in Fig. 3, where the rinse continues for an additional 10 rnin with negligible signal reduction.The decrease in signal due to immunobead redistribution is most problematic, and probably occurs when the rinsing mixes the reacted and unreacted beads and pushes the beads to the outside walls of the sensing chamber, out of the fibre optic field of view. Movement out of the field of view is because the side-walls of the frit are thinner than the end-wall (see Fig. 2). This is supported by the observation that during rinsing, the flow is most prominent through the side-walls and not the end of the frit. Studies with a new frit design, having approximately equal thicknesses for the end-wall and the side-walls, are in progress. Currently, work is also progressing on a delivery system consisting of three inlets and three outlets (symmetric- ally arranged), through which all of the reagents will be introduced and removed.It is hoped that this arrangement will permit more symmetrical flow patterns within the sensing chamber, thereby minimizing the effects of redistribution. Other operational and design parameters which were not optimized in these preliminary experiments are bead concen- tration, reagent phase flow-rates, chamber geometry and frit permeability. A dose-response graph was constructed over the rangeANALYST, FEBRUARY 1991, VOL. 116 122 40 0 0.02 0.04 0.06 0.08 0.10 0.12 Concentration of anti-rabbit IgG-FITC/mg ml-1 Fig. 4 Dose-response IgG-FITC assay (i.e., calibration) for anti-rab 0.011-0.11 mg ml-1 of anti-rab IgG-FITC (see Fig. 4). Excluding the highest concentration data point, the correla- tion coefficient for the linear portion of the graph was 0.9969. The non-linear portion of the graph is possibly due to the saturation of those beads within the field of view, before rinsing removes the sample and redistributes the beads.The limit of detection (LOD) (signal to noise ratio of 2) is approximately 8 x 10-3 mg ml-1 (5 X 10-8 mol dm-3) but could possibly be improved by increasing the volume of aspirated sample. By increasing the sample volume, the sample would come into contact with the beads for a longer time before rinsing and redistribution influences the signal. The absolute LOD is approximately 8 x 10-5 mg (5 x 10-13 mol). Although the linear dynamic range was limited for this MRB, it is not believed that this is an inherent difficulty with the design but rather because the operational and design parameters have yet to be fully developed.Both the dynamic range and the LOD are expected to improve when the changes discussed above are implemented. Future work will be necessary in order to realize the full potential of this system for performing long-term FIA sensing in real matrices using sandwich, competitive-binding and enzyme assay protocols. This research was supported by the National Science Founda- tion under contract number CHE-8708581 and the Division of Chemical Sciences, Office of Basic Energy Research, US Department of Energy, under contract number DE-FGO5- 86ER13613 with the University of Tennessee, Knoxville, TN, USA. References 1 2 3 Seitz, W. R., CRC Crit. Rev. Anal. Chem., 1988, 19, 135. Wolfbeis, 0.S . , TrAC, Trends Anal. Chem., Pers. Ed., 1985,4, 184. Sepaniak, M. J . , Tromberg, B. J., and Vo-Dinh, T., Prog. Anal. At. Spectrosc., 1988, 11, 481. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18. 19 20 21 22 23 24 25 26 27 28 29 30 Vo-Dinh, T., Tromberg, B. J., Griffin, G. D., Ambrose, K. R., Sepaniak, M. J . , and Gardenhire, E. M., Appl. Spectrosc., 1987,41, 735. Kulp, T. J . , Camins, I., Angel, S. M., Munkholm, C., and Walt, D. R., Anal. Chem., 1987, 59, 2849. Tromberg, B. J . , Sepaniak, M. J., Alarie, J . P., Vo-Dinh, T., and Santella, R. M.. Anal. Chem., 1988, 60, 1901. Saari, L. A., and Seitz, W. R., Anal. Chem.. 1982, 54, 821. Munkholm, C., Walt, D. R., Milanovich, F. P., and Klainer, S. M., Anal. Chem., 1986, 58, 1427. Zhujun, Z., Mullin, J . L., and Seitz, W.R., Anal. Chim. Acta, 1986, 184, 251. Wyatt, W. A., Bright, F. V., and Hieftje. G. M., Anal. Chem., 1987,59,2272. Zhujun, Z . , and Seitz, W. R., Anal. Chim. Acta, 1984,160,305. Schultz, J. S . , Mansouri, S . , and Goldstein, I. J . , Diabetes Care. 1982, 5, 245. Meadows, D., and Schultz, J. S . , Talanta, 1988,35, 145. Arnold, M. A., Anal. Chem., 1985, 57, 565. Fuh, M. R. S., Burgess, L. W., and Christian, G. D., Anal. Chem., 1988, 60,433. Wangsa, J., and Arnold, M. A., Anal. Chem., 1988, 60, 1080. Tromberg, B. J . , Sepaniak, M. J., Vo-Dinh, T., and Griffin, G. D., Anal. Chem., 1987, 59, 1226. Petrea, R. D., Sepaniak, M. J . , and Vo-Dinh, T., Talanta, 1988, 35, 139. Sutherland, R., Dahne, C., Place, J. F., and Ringrose, A. S . , Clin. Chem. (Winston-Salem, NC), 1984, 30, 1533. Andrade, J. D., Vanwagenen, R. A., Gregonis, D. E., Newby, K., and Lin, J. N., IEEE Trans. Electron Devices, 1985, ED-32, 1175. Anderson, F. P., and Miller, W. G., Clin. Chem. (Winston- Salem, NC), 1988,34, 1417. Vo-Dinh, T., Tromberg, B. J., Sepaniak, M. J . , Griffin, G. D., Ambrose, K. R., and Santella, R. M., in Optical Fibers in Medicine IZI, ed. Katzir, A., Proceedings of SPIE 910-18, Los Angeles, 1988. Tromberg, B. J., Sepaniak, M. J . , and Vo-Dinh, T., in Optical Fibers in Medicine I I I , ed. Katzir, A., Proceedings of SPIE 906-06, Los Angeles, 1988. Liu, B. L., and Schultz, J . S., IEEE Trans. Biomed. Eng., 1986, Alarie, J . P., Bowyer, J . R., Sepaniak, M. J . , Hoyt, A. M., and Vo-Dinh, T., Anal. Chim. Acta, 1990,236, 237. Sepaniak, M. J., Tromberg, B. J . , Alarie, J . P., Bowyer, J. R., Hoyt, A. M., and Vo-Dinh, T., in 196th National Meeting of the American Chemical Society (Los Angeles, CA), eds. Murray, R. W., Dessy, R. E., Heineman, W. R., Janata, J., and Seitz, W. R., American Chemical Society, Washington, 1989, pp. Sepaniak, M. J., Clin. Chem. (Winston-Salem, NC), 1985, 31, 671. Joky, M. E., Pandex Research Report, Lit. No. 4001/1.5M, July, 1983. Alarie, J. P., Sepaniak, M. J., and Vo-Dinh, T., Anal. Chim. Acta, 1990, 229, 169. Papadea, C., and Check, I. J., CRC Crit. Rev. Clin. Lab. Sci., 1989, 27, 27. BME-33, 133. 3 18-330. Paper 0103429 K Received July 27th, 1990 Accepted October 4th, 1990

ISSN:0003-2654    

DOI:10.1039/AN9911600117

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1991, VOL. 116 123 Voltammetric Behaviour of Screen-printed Carbon Electrodes, Chemically Modified With Selected Mediators, and Their Application as Sensors for the Determination of Reduced Glutathione Stephen A. Wring and John P. Hart* Science Department, Bristol Polytechnic, Coldharbour Lane, Frencha y, Bristol BS16 lax UK Brian J. Birch Unilever Research, Sensors Group, Colworth Laboratory, Colworth House, Sharnbrook, Bedford MK44 1 LO, UK The evaluation of screen-printed carbon electrodes, chemically modified with selected ferrocene, phthalo- cyanine and hexacyanoferrate(iii) derivative mediators for the determination of reduced glutathione (GSH), is described. Cyclic voltammetry was used t o investigate the effect of pH on the electrochemical behaviour of these mediators incorporated in the disposable electrodes.Values of the electron-transfer coefficient (an,) were calculated for the oxidation of the mediators in phosphate buffer solution and solutions containing 0.48 mmol dm-3 GSH. Amperometry in stirred solutions was used t o construct hydrodynamic voltammograms for each of the modified electrodes; these voltammograms were used t o elucidate their steady-state behaviour. Both electrochemical techniques were used t o calculate the reduction in overpotential for the oxidation of GSH; these calculations were performed for all of the chemically modified electrodes. Amperometry in stirred solutions was used as the technique for quantitative measurements, t o determine the calibration response factors (PA mmol-1 dm3) and limits of detection for selected mediators towards GSH.The appropriate applied potentials were selected by reference t o the hydrodynamic waves; a range of values were investigated t o find the potential that gave the maximum sensitivity. The most promising mediators were cobalt phthalocyanine, iron phthalocyanine and ferrocenecarbaldehyde. Keywords: Screen-printed carbon electrodes; chemically modified ferrocene, phthalocyanine and hexacyano- ferrate(ii1) mediators; reduced glutathione; cyclic voltammetry; amperometry Recently, there has been considerable interest in the use of disposable, chemically modified electrodes for the determina- tion of various biomolecules. 1-7 Indeed, careful selection of suitable electron mediators can significantly reduce the over- potential necessary for the determination of some analytes and enhance the selectivity of electroanalytical sens0rs.J-7 Frew and co-workers597 described this enhancement in a device for plasma glucose which uses the electrochemically generated ferricinium ion to act as an electron acceptor from the reduced flavoenzyme, glucose oxidase.Batchelor et aZ.6 also demonstrated that 4-methyl-o-benzoquinone could be incorporated in a disposable device for the determination of the ketone body 3-hydroxybutyrate. In a recent investigation,g we described a method of producing screen-printed graphite electrodes chemically modified with the important electrocatalyst cobalt phthalo- cyanine (CoPC); these devices successfully reduced the overpotential necessary for the determination of reduced glutathione (GSH), ascorbic acid and coenzyme A at plain graphite electrodes.This electrocatalyst has also been incor- porated in carbon-paste9.*" and re-usable carbon-poxy resin electrodes,lO which were used for the determination of GSH in human whole blood9 and plasma" by high-performance liquid chromatography with electrochemical detection. Our chemically modified screen-printed electrodes (SPEs) previously described8 were easily and reliably fabricated and could be employed, using differential-pulse voltammetry or in the amperometric mode with stirred solutions, for quantita- tive determinations in simple sample matrices. However, before analyses could be performed on complex biological samples their selectivity would need to be further enhanced.") Therefore, it was considered that the study of other mediators that could permit improved selectivity, through application of even lower applied potentials,'? was worthy of * To whom correspondence should be addressed.investigation. We set out to examine the possibility of using a variety of organometallic compounds as mediators for the determination of GSH. To our knowledge, apart from CoPC, none of these compounds has previously been used for this application with SPEs. This investigation involved three studies. The purpose of the first was to use cyclic voltammetry to examine the electrochemical behaviour of the selected mediators in plain phosphate buffer solutions, and those containing GSH, over a range of pH values. In the second part, hydrodynamic voltammograms were obtained by using simple amperometry in stirred solutions for each of the modified SPEs.Finally, amperometry in stirred solutions was used to compare the selectivity and sensitivity of the most promising electrodes for the determination of GSH. In these studies, GSH was used as the analyte of choice, because it is a very important cofactor in many physiological processes and also plays a key role in the detoxification of some common drugs;'3,14 in addition, changes in its circulating concentration can be used as a marker for certain dis- orders. 15-17 Experimental Chemicals and Reagents The CoPC was purchased from Kodak, all other mediators studied were obtained from Aldrich. Graphite (Ultra Carbon Ultra 'F' grade UCP-1M) and GSH were supplied by Johnson Matthey and Sigma, respectively.All the materials used for the production and cleaning of the screen-printing template were obtained from Sericol. The inert support used for the electrodes was semi-rigid, white poly(viny1 chloride) (PVC) marketed under the trade name Pentawhite and was obtained from ADP. The cellulose acetate and solvents used to prepare the graphite suspension for printing on to the Pentawhite were obtained from Aldrich.124 ANALYST, FEBRUARY 1991, VOL. 116 The supporting electrolyte used throughout all the investi- gations was phosphate buffer solution, prepared from 0.5 rnol dm-3 stock solutions of sodium dihydrogen orthophos- phate, disodium hydrogen orthophosphate and orthophos- phoric acid. These were mixed to yield buffer solutions of the required pH values (a pH meter was used).These were subsequently diluted to provide working solutions of 0.05 mol dm-3. All solutions were prepared with purified water (> 18 MQ cm) obtained using a Millipore Milli-Q purification system. All GSH solutions were prepared in the appropriate working buffer immediately prior to use and were protected from light during all investigations. Purified nitrogen was obtained from BOC. Apparatus Cyclic voltammetric and amperometric measurements were obtained using a Metrohm E612 VA-scanner in conjunction with a Metrohm E611-detector; these were used with a JJ Instruments PN4 x-y plotter to record voltammograms and amperograms. A three-electrode cell was used, incorporating the SPEs with a saturated calomel reference electrode (Russell Electrodes) and a laboratory-constructed platinum- wire counter electrode.Electrical contact to the SPEs was facilitated with a spade connector glued into a piece of glass tubing (15 x 0.3 cm i.d.) to form an electrode holder. For amperometric measurements in stirred solutions, a small circular stirring disc of 14 mm diameter (BDH) was placed in the bottom of the cell and rotated at a fixed rate by a Whatman Mini-MR stirrer. Elect rode Construct ion The SPEs consisted of a circular 3 mm working area with a 25 x 1 mm connecting strip. These were printed in parallel groups of six electrodes separated by a 7 mm space.* The SPEs were prepared by the method and template described previously.8 In brief, for the unmodified electrode this involved preparing a 1.5% m/m solution of cellulose acetate in a 1 + 1 (v/v) mixture of cyclohexanone and acetone; this solvent system permits the graphite film to adhere to the PVC support.Immediately prior to use, 1.1 g of this solution were added to 0.5 g of graphite in a small glass vial. These components were mixed to form an even suspension, which could then be printed through the screen on to the PVC support, which had previously been cleaned with ethanol. After use, the screen template was cleaned in a commercial thinner solution (Sericol XG) . For the modified electrodes, 5% m/m of the required mediator was added to the graphite. Once printed, the electrodes were left in the fume cupboard overnight to allow the solvents to evaporate. Immediately prior to use individual electrodes were cut from the piece of PVC, and the connecting strip was trimmed to 15 mm and covered with insulating tape (RS Components), leaving the 3 mm circular working area exposed, in addition to a 6 mm length at the opposite end to allow electrical contact with the spade connector in the electrode holder.Voltammetric Procedures Using SPEs Cyclic voltammetry Cyclic voltammetric measurements were obtained for blank solutions of 0.05 mol dm-3 phosphate buffer (pH 3, 5 and 7) and then for the same solutions containing 0.48 mmol dm-3 GSH, using both the unmodified and 5% m/m modified SPEs in order to investigate the effects of pH. Higher pH values were not studied because GSH is particularly unstable in alkaline media.18 The mediators used for this study were split into two groups; the first group consisted of molecules based on ferrocene, namely, ferrocene itself, dimethylferrocene, ferrocenedicarboxylic acid, ferrocenecarboxylic acid, ferro- cenecarbaldehyde and dimethylferrocenedicarboxylic acid.The second group consisted of CoPC, iron phthalocyanine (FePC), Prussian Blue and potassium hexacyanoferrate(n1). The voltammetric conditions were as follows: initial poten- tial, -0.5 V; scan rate, 20 mV s-1; and final potential, 1.2 V. All experiments were performed in triplicate, using a fresh electrode for each run; all the results quoted in later sections are the mean values for each of the parameters studied. Prior to each experiment the supporting electrolyte solution was de-aerated with purified nitrogen to eliminate the oxygen reduction waves.Hydrodynamic voltammetry Hydrodynamic voltammograms were obtained for both modi- fied and unmodified SPEs by amperometry in stirred solutions of 0.05 rnol dm-3 phosphate buffer (20 ml) and in similar buffer solutions containing 0.48 mmol dm-3 GSH. The applied potentials of the working electrodes were increased in steps; the resulting steady-state anodic-current responses were measured for each plateau and plotted versus applied poten- tial. Each experiment was performed in triplicate with fresh electrodes and the results quoted represent the mean current values. Calibration, Sensitivity and Selectivity By use of amperometry, in stirred solutions of 0.05 rnol dm-3 phosphate buffer (pH 7), the magnitude of the anodic current responses following additions of small volumes of stock GSH solutions was recorded over the final concentration range 1.48 x 10-7-2 x 10-3 mol dm-3 GSH (i.e., for solutions containing 1.48 X 10-7, 8.67 X 10-7, 4.76 x 10-6, 4.97 X 10-5, 4.76 X 10-4 and 2 x 10-3 rnol dm-3 GSH). In each instance, the stock GSH solution was added to 20 ml of plain buffer solution in the voltammetric cell, and the difference in the recorded current was measured; a fresh electrode was used for each individual determination .* The applied potential values selected for these investigations were taken, where possible, at different points along the hydrodynamic wave for each mediator studied. The sensitivity of the calibration response (calculated from the slope of the calibration graph) was then determined at every applied potential for each of the selected mediators; these results were plotted versus applied potential and could be used to establish the relationship between sensitivity and selectivity for each electrode type.Results and Discussion The voltammetric evaluation of the different mediators for the electrocatalytic determination of GSH at the modified screen- printed electrodes is described below. Cyclic Voltammetry Cyclic voltammograms were recorded, for both groups of the modified SPEs, in 0.05 rnol dm-3 phosphate buffer (pH 3, 5 and 7) and in similar solutions containing 0.48 mmol dm-3 GSH. Cyclic voltammetric studies on the ferrocene group of mediators For all but one mediator, the cyclic voltammograms obtained in plain buffer solutions, at the SPEs modified with the ferrocene compounds, showed a single quasi-reversible redox couple; the anodic (Ep,) and cathodic (Epc) peak potential values, and their separation (6Ep), are given in Table 1.The voltammograms obtained at the electrodes modified with dimethylferrocenedicarboxylic acid revealed only a singleANALYST, FEBRUARY 1991, VOL. 116 0.2 125 - +- ------ +------+ I I I I I Table 1 Values of peak potential recorded, using cyclic voltammetry in plain phosphate buffer solution, ferrocene group of mediators Buffer Mediator PH Ferrocene 3 5 7 Ferrocene- carbaldehyde 3 Dimethylferrocene 3 5 7 5 7 Ferrocene- carboxylic acid 3 5 7 Ferrocene- dicarboxylic acid 3 5 7 Dimethylferrocene- dicarboxylic acid 3 5 7 at the SPEs-modified EpJ V 0.330 0.347 0.372 0.540 0.540 0.530 0.200 0.203 0.223 0.428 0.315 0.315 0.640 0.460 0.450 0.810 0.810 0.800 EpJ V 0.146 0.137 0.150 0.250 0.240 0.200 0.063 0.063 0.070 0.310 0.237 0.233 0.490 0.370 0.360 - - - with -the 6Epl V 0.184 0.210 0.222 0.290 0.300 0.330 0.137 0.140 0.153 0.118 0.078 0.082 0.150 0.090 0.090 - - - -1 2 1 ' I I 1 I anodic wave; hence, the oxidation appeared to be irreversible over the potential range studied.For the determination of GSH involving mediated processes, anodic responses are of more importance than the cathodic responses because the magnitude of the former current is expected to increase during electrocatalysis. Bearing this in mind, the behaviour of the oxidation waves was investigated during the remainder of this study.The graph of E,, versus pH for the modified electrodes in plain buffer solution [Fig. l ( a ) ] indicates that the anodic wave 30 f 2o \ . -! 10 0 3 4 5 6 7 pH of the 0.05 mol dm-3 phosphate buffer Fig. 2 Effect of pH on ip, for the ferrocene group of mediators using ( a ) 0.05 mol dm-3 phosphate buffer and ( b ) solutions of the same buffer containing 0.48 mmol dm-3 GSH. Symbols as in Fig. 1 shifts to more negative potentials with increasing pH only for the electrodes modified with ferrocenecarboxylic and fer- rocenedicarboxylic acids; for the remainder of the mediators the electrochemical behaviour was independent of pH over the range studied. For the first two SPEs, a break in the graph line occurs at pH 5; this suggests the presence of a pK, (or pK') for the carboxylic acid groups at this approximate value.Similar redox behaviour was observed in the presence of 0.48 mmol dm-3 GSH [Fig. 1(b)]; however, in solutions of pH 7 an extra anodic wave (Epa, 0.65 V) was seen for the dimethylferrocenedicarboxylic acid-modified electrodes, and the peak for ferrocenedicarboxylic acid-modified devices had moved to a more positive potential. All of the electrodes, except those modified with dimethyl- ferrocene, afforded an enhanced anodic current response in the presence of GSH. Fig. 2(b) shows the difference in current measured between the plain buffer solution [Fig. 2(a)] and solutions containing GSH; a negative value indicates that the anodic current measured in the presence of GSH was less than that recorded in the plain buffer solution.Interestingly, the i,, versus pH graph reveals that the mediators with the most positive E,, values afforded the most enhanced current responses. Values of the electron-transfer coefficient (an,) were calculated,19 where possible, for the anodic waves obtained with plain buffer solution and with solutions containing 0.48 mmol dm-3 GSH (Fig. 3). This graph indicates clearly that the an, values determined at the modified SPEs, which demon- strate an electrocatalytic response for GSH, are reduced in the presence of the analyte. This observation is especially apparent for the SPEs modified with ferrocenedicarboxylic acid and ferrocenecarbaldehyde in pH 7 solutions. These electrodes were amongst those that exhibited the largest increase in current response in the presence of GSH.(Unfortunately, the an, values could not be determined for GSH at the other promising SPE containing dimethylferro- cenedicarboxylic acid, because the resulting new wave at 0.65 V was not sufficiently resolved.) The decrease in the observed an, values implies that, in the presence of GSH electrocataly- sis, the oxidation of the mediator becomes more irreversible.126 0.8 0.7 c" 0.6 0.5 ANALYST, FEBRUARY 1991, VOL. 116 - - - - 0.4 - 3 4 5 6 7 pH of the 0.05 mol dm-3 phosphate buffer Fig. 3 Effect of pH on the values of the electron transfer coefficient (an,), determined for the ferrocene group of mediators, in 0.05 mol dm-3 phosphate buffer (broken lines) and solutions of the same buffer containing 0.48 mmol d ~ r - ~ GSH (solid lines).Symbols as in Fig. 1 3 - However, as enhanced current responses were recorded at these devices, the reduction in the an, values could be reflecting the increased time taken during the ECE mechan- ism5 involved in the mediated oxidation of GSH compared with the simple E mechanism occurring in the plain buffer solutions. Therefore, the results suggest that the intermediate chemical reaction is the rate-determining step in the over-all oxidation process. Approximate values for the apparent experimental rate constant ( k , ) were calculated from the 6Ep values (at scan rates when 6Ep was greater than 200 mV), by the method of Laviron,20321 for two of the most promising mediators (ferro- cenecarbaldehyde and ferrocenedicarboxylic acid) in pH 7 buffer solutions.The mean value of k, for the ferrocene- dicarboxylic acid-modified electrodes, for scan rates between 150 and 250 mV s-1, was 0.357 s-1 [n = 3, relative standard deviation (RSD) = 13.9%]. This value agrees well with those determined by Laviron and Roullier21 for ferrocene-modified polymer electrodes. However, the calculated mean value of k , for the ferrocenecarbaldehyde-modified electrodes, under the same experimental conditions, was only 0.049 s-1 ( n = 3, RSD = 8.5%). This suggests that electron-transfer rates for the latter mediator in the present SPE matrix can appear very slow, particularly at high scan rates; this observation is confirmed by the large 6Ep values (Table 1) observed with this compound. . 0. ( a ) Cyclic voltammetry with the phthalocyanine and hexacyano- ferrate( 111) derivative mediators CoPC.The electrochemical behaviour of CoPC-modified electrodes has been described previously,g,10 and the observa- tions from the present study confirm these findings. In plain phosphate buffer solution, two irreversible anodic waves were recorded; e.g., in the pH 3 buffer solution these occurred at 0.48 and 0.80 V, respectively. Using the notation published previously,l" these were designated waves 2 and 3, respec- tively. When GSH was added to the supporting electrolyte solution, an additional irreversible anodic wave (wave 1) was observed at less positive potentials than waves 2 and 3 [Fig. 4(b)]. As peak 1 has been used successfully for the determina- tion of GSH in biological samples, it was again studied as the peak of interest for the remainder of this investigation.The cyclic voltammograms recorded for the FePC- modified SPEs in plain phosphate buffer solution (pH 3 ) revealed two anodic waves [Fig. 4(a)] and one broad cathodic (Epcl, -0.15 V) wave; the position of the second anodic wave was found to be dependent on the pH of the supporting electrolyte. In the pH 5 and 7 buffer solutions the broad cathodic wave was resolved into two smaller waves (pH 5: Epcl, -0.08V; E,,2, -0.23 V, and pH 7: Ep,l, -0.13 V; Epc2, -0.30 V). This suggests that the electrode reactions for FePC are reversible. FePC. I 0 . . . . . . . . o . . . . . . . o.2 t --------- 0 3 4 5 6 7 Fig. 4 Effect of pH on E,, for the phthalocyanine and hexacyanofer- rate(w) group of mediators using (a) 0.05 mol dm-3 phosphate buffer and ( b ) solutions of the same buffer containin 0.48 mmol dm-3 GSH.0, Prussian blue; A , FePC wave 1; #. FePC wave 2; 0, hexacyanoferrate(u1); and +, CoPC wave 1 pH of the 0.05 mol dm-3 phosphate buffer 1 1 I 1 2 t'b' -1 3 4 5 6 7 pH of the 0.05 mol dm-3 phosphate buffer Fig. 5 Effect of pH on ipa for the phthalocyanine and hexacyanofer- rate(m) group of mediators using ( a ) 0.05 mol dm-3 phosphate buffer and ( 6 ) solutions of the same buffer containing 0.48 mmol dm-3 GSH. Symbols as in Fig. 4 Similar voltammetric behaviour was observed for the electrodes in the buffer solutions containing 0.48 mmol dm-3 GSH [Fig. 4(b)]; however, an additional, poorly resolved, anodic wave was recorded at about 0.7 V for each of t h e pH values studied (not shown). The magnitude of the anodic waves 1 and 2 was also found to increase in the presence of GSH in pH 7 phosphate buffer solution [Fig.5(a) and (b)]. As the electrocatalytic response of waves 1 and 2 was the most promising in terms of sensor selectivity, their behaviour wasANALYST, FEBRUARY 1991, VOL. 116 f 1.0 0.8 . * I 2? 2 0.6 0 5 0.4 0 2 0.2 I 9 0 - 127 - - - - - investigated during the remainder of this study, and is summarized in Figs. 4 and 5 . HexucyanOferrate(iii). For each buffer pH studied, two anodic waves (Epal, about 0.30 V; Epa2, about 0.45 V) and one cathodic wave were recorded; the behaviour of the first anodic wave is summarized in Fig. 4(a) and ( 6 ) . The second anodic wave observed for this usually model redox system probably arises owing to some interaction of the hexacyanoferrate(111) with the electrode matrix or solvent system.This behaviour has been described previously for other modified polymer-based electrodes,2* and this process is considered probable because only a single, small and tailing cathodic wave was observed; in addition, at applied potentials below 0.0 V the current response became very erratic and noisy. Fig. 5(u) and (b) clearly indicates that, despite the presence of these matrix interactions, the magnitude of the first anodic wave increases in the presence of GSH, with the maximum response occurring in pH 3 buffer solutions. Prussian Blue [iron( 111) hexacyanoferrute(rr~)] . The cyclic voltammograms of the Prussian Blue-modified electrodes, obtained at each buffer pH studied in the absence of GSH, reveal an ideal quasi-reversible redox couple with single anodic and cathodic waves.As seen in Fig. 4(u) and ( b ) the position of the peak potential (Epa) is independent of pH over the range studied. In buffer solutions containing 0.48 mmol dm-3 GSH, the anodic current increases to give the maximum mediated response at pH 5 [Fig. 5(u) and ( b ) ] . The values of an, were calculated for the anodic responses shown in Fig. 5(a) and (6) for each of the second group of modified SPEs (Fig. 6). This graph confirms the observations made by using the ferrocene group of electrodes, where the electron-transfer coefficient is decreased in the presence of a mediated response to GSH. The mechanism for this process is undoubtedly similar to that mentioned previously and involves an ECE mechanism as the Fe2+ moiety is electro- chemically oxidized to Fe3+, whereupon it is chemically reduced by the GSH and subsequently re-oxidized electro- chemically.For the CoPC-modified SPEs, values of an, could only be determined in the presence of GSH because wave 1 is absent in plain buffer solution; the mechanistic behaviour of this mediator has been described previously.10 Hydrodynamic Voltammetry For biomedical sensor applications with use of enzymes to enhance selectivity, buffer solutions usually need to be at, or near, physiological pH values. In the second part of our current investigation, we studied the anodic current response of each of the modified SPEs to GSH in pH 7 phosphate buffer solution, using amperometric detection in stirred solutions.0.6 - 2 0.4 - 0.2 - I I 1 I I I 1 3 4 5 6 7 pH of the 0.05 mot dm-3 phosphate buffer Fig. 6 Effect of pH on the values of the electron transfer coefficient (an,), determined for the phthalocyanine and hexacyanoferrate(li1) group of mediators, in 0.05 mol dm-3 phosphate buffer (broken lines) and solutions of the same buffer containing 0.48 mmol dm-3 GSH (solid lines). 0, Prussian Blue; A , FePC wave 1; +, CoPC wave 1; and 0, hexacyanoferrate(iri) wave 1 This simple detection technique is particularly important for sensor applications and has been used successfully for the analysis of selected biomolecules in our work and that of other workers. 10,2244 Hydrodynamic voltummetry with the ferrocene group of mediators The hydrodynamic voltammograms [Fig.7 ( u ) ] indicate that, for solutions of GSH prepared in phosphate buffer solution (pH 7), enhanced anodic current responses are observed only at the SPEs modified with ferrocenedicarboxylic acid, ferro- cenecarbaldehyde and dimethylferrocenedicarboxylic acid. These findings confirm the observations carried out by cyclic voltammetry in the same buffer solutions. These voltammo- grams were constructed by plotting the difference in current measured between the response in plain phosphate buffer solution (pH 7) and the same solution containing 0.48 mmol dm-3 GSH; the current at the initial applied potential, under steady-state conditions, was used as the zero reference point. Hence, the voltammograms illustrate the current arising solely from the mediated oxidation of GSH at the electrode surface.1.2 1 .o 0.8 0.6 0.4 0.2 0 1 0 100 200 300 400 $00 600 700 800 900 1'2 I (b) - P- -0.2 I I i I I 1 I -200 0 200 400 600 800 25 -200 0 200 400 600 800 1000 Applied potential/mV versus SCE Fig. 7 Hydrodynamic voltammograms for ( a ) the fcrrocene group of mediators; (b) CoPC and the hexacyanoferrate(ii1) mediators; and ( c ) FePC. The anodic current values represent the difference in the measured current between 0.05 mol dm-3 phosphate buffer (pH 7) solutions and those containing 0.48 mmol dm-3 GSH, and hence illustrate the current arising from thc mediated oxidation of GSH at the modified electrodes. (a): x , Ferrocene; A, ferrocenecarb- aldehyde; + , dimethylferrocene; V , ferrocenecarboxylic acid; 0, ferrocenedicarboxylic acid; and 0, dimethylferrocenedicarboxylic acid.(6): 0, Prussian Blue; 0, hexacyanoferrate(ii1); V, CoPC; and A , unmodified128 1 x 10-3 ANALYST, FEBRUARY 1991, VOL. 116 - + I 1 I Hydrodynamic voltammetry with the phthalocyanine and hexacyanoferrate(m) group of mediators The hydrodynamic voltammograms for the four mediators confirm their electrocatalytic behaviour in stirred solutions [Fig. 7(b) and (c)]. The response for the CoPC-modified SPEs clearly reveals a ‘shoulder’ at 0.4 V on the main voltammo- gram owing to the first analytical wave (peak l), which is seen only in the presence of GSH. Fig. 7(b) also shows the current response recorded at the unmodified SPEs; as expected, negligible current values were recorded at the applied potentials, corresponding to the maximum steady-state currents for the mediated electrodes.Reduction of Overpotential With the Different Mediators The results from both of the voltammetric studies indicate clearly that the overpotential necessary for the electrochem- ical detection of GSH can be decreased by using suitable electron mediators. Final selection of the optimum mediator for a particular application will depend on its efficiency at reducing overpotential and its sensitivity, i.e., its current response factor (PA mmol-* dm3) during calibration. The reductions in overpotential for the electrocatalytic oxidation of GSH are given in Table 2 for each of the modified electrodes that demonstrated a favourable response. The values given are the differences between the E,, values for the oxidation of GSH at the unmodified SPEs, obtained by cyclic voltammetry (Epa, 1.26 V) and hydrodynamic voltammetry (>1.4 V, a plateau was not reached within the usable potential window of the electrodes), and the E,, values at the modified devices.From those data, and the magnitude of the amper- ometric current responses seen in Fig. 7(a-c), the most promising mediators for the selective determination of GSH at pH 7 are: CoPC, FePC, ferrocenedicarboxylic acid, ferro- cenecarbaldehyde and Prussian Blue. Large increases in the current response were also recorded for dimethylferrocene- dicarboxylic acid; however, for the last mediator the applied potentials necessary for oxidation were deemed to be too positive for practical sensor applications.Calibration In the final part of this investigation, amperograms were recorded, by using the method described previously,g for the most promising modified electrodes. The concentrations studied covered the range 1.48 x 10-7-2 x 10-3 mol dm-3 GSH, and the applied potential values were selected from different points on their hydrodynamic waves. The current Table 2 Comparison of the reduction in measured overpotential for the oxidation of GSH between the modified and unmodified SPEs in 0.05 mol dm-3 phosphate buffer (pH 7); where: AT]CV = ECVunmodrfied - ECVrnodified and AT]AMP = EAMPunrnodifled - EAMPrnodified Mediator Ferrocene Ferrocene- carbalde h yde Dimethylferrocene- dicarboxylic acid Ferrocene- dicarboxylic acid Prussian Blue FePC (wave 1) FePC (wave 2) Hexacyanoferrate(n1) CoPC (wave 1) Cyclic voltammetry 889 ( AT]cv)/mV * Indicates no significant response.730 610 720 1080 1180 1010 960 960 Amperometry (A%4MP)ImV * > 880 >670 > 1000 >1200 >lo10 >950 > 1000 * response factors were calculated from the slope of the calibration graph for each of the mediators by using the different applied potentials; these values were plotted versus potential to yield the current response profiles [Fig. 8(a) and (b)]. Fig. 8(a) indicates that FePC-modified electrodes offer the best selectivity during calibration owing to the low applied potentials required. However, for optimum sensitivity, the ferrocenecarbaldehyde-modified SPEs become the devices of choice, although for concentrations of <4.76 x 10-6 mol dm-3 GSH, high background currents and noise restric- ted the practical detection limits to this concentration when the applied potential was 400 mV and to 8.67 x 10-7 rnol dm-3 for Eapplied = 525 mV.By using the most sensitive potential (450 mV) the limit of detection was 1.48 x 10-7 mol dm-3 GSH as determined for the other devices. Fig. 8(a) also indicates that the optimum sensitivity cannot be obtained on the plateau of the hydrodynamic wave, but at a potential slightly greater than E0.5. Fig. 8(b) illustrates the current response factors plotted versus the position of the applied potential on the hydrodynamic wave. For the ferro- cenecarbaldehyde- and CoPC-modified electrodes the opti- mum response was achieved at approximately the E0.65 point.Unfortunately, owing to excessive baseline noise, calibration was not possible for the full range of potentials with the FePC- and ferrocenedicarboxylic acid-modified electrodes; however, the slopes of their response curves do increase towards the mid-point of the hydrodynamic wave. A possible explanation for this observation involves the redox cycling mechanism that was proposed earlier. At lower applied potentials the energy barrier for the chemical reduction of the oxidized electron mediator by the GSH will be at a minimum, allowing this reaction to proceed at a greater rate. This process would in turn replenish the reduced form of the mediator for further electro-oxidation and hence the generation of larger electro- catalytic currents.ANALYST.FEBRUARY 1991, VOL. 116 129 t c t ?! 3 u -0.4 0 0.4 0.8 -0.4 0 0.4 0.8 PotentialN versus SCE I I 1 -0.4 0 0.4 0.8 Fig. 9 GSH (solid lines) using: ( a ) ferrocenecarbaldehyde: ( b ) FePC; and ( c ) ferrocenedicarboxylic acid modified SPEs. Scan rate, 20 mV s-1 Cyclic voltammograms recorded in 0.05 mol dm-3 phosphate buffer (broken lines) and the same solutions containing 0.48 mmol dm-3 For comparison, Fig. 9(a)-(c) shows the cyclic voltammetric behaviour of these iron-containing mediators in 0.05 mol dm-3 phosphate buffer (pH 7) and for the same solutions containing 0.48 mmol dm-3 GSH; the cyclic voltammograms for CoPC have been described previously.8.") No response was observed with use of amperometry at constant potential for the Prussian Blue-modified electrodes during calibration.(The experiments were repeated and the results confirmed; at present we are unable to explain this phenomenon.) The highest current responses were obtained with the phthalocyanine-based sensors. This suggests that they should offer better sensitivity and selectivity when compared to other mediators. Conclusion The voltammetric data and calibration results reported here suggest that the screen-printed carbon electrodes can be used as a substrate suitable for modification with a variety of electron mediators. The systematic evaluation of these elec- trocatalysts has shown that, with careful selection of the potential, the oxidation process for GSH can be controlled; therefore, the selectivity and sensitivity can be altered to suit the particular application.This study has shown that, of the mediators selected initially, the most promising for the determination of GSH are: CoPC, FePC and ferrocenecarbaldehyde . The most selective media- tor, requiring the least positive applied potential, was FePC, while the most sensitive was ferrocenecarbaldehyde. However, both of these responses were superimposed on anodic background currents arising from an initial electrochemical oxidation of the Fez+ moieties. Conversely, the anodic current measured at the CoPC devices arises solely from the presence of the GSH in solution, which could permit faster response times for some applications. It is envisaged that the screen-printed carbon electrodes chemically modified with these mediators, and used in conjunction with selective enzymes, could form the basis of selective biosensors for the determination of GSH in biolog- ical matrices.The authors thank the National Advisory Board for financial support. They are also grateful to Apple Litho (Bristol) for their help and advice regarding the screen-printing of the carbon electrodes. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 References Turner, A . P. F., Senb. Actuar., 1989, 17, 433. Scheller, F., Schubert, F., Pfeiffer, D., Hintsche, R., Drans- feld, I., Renneberg, R.. Wollenberger, U., Riedel, K., Pavlova, M.. Kuhn, M., Muller, H.-G., Tan, P. M., Hoffmann. W., and Moritz, W.. Analyst, 1989. 114, 653. Frew, J . E., Bayliff, S. W., Gibbs, P. N . B., and Green, M. J . , Anal. Clzim. Actu, 1989,224, 39. Patriarche. G . J . , Kauffmann, J.-M., and Vire, J.-C., Redox Chem. Interfacial Behav. Biol. Mol., 1987, 3. 479. Frew, J . E., and Hill, H . A. O., Anal. Chem.. 1987, 59. 933A. Batchelor, M. J . , Green, M. J., and Sketch, C. L . . Anal. Clrim. Acra, 1989, 221, 289. Frew, J . E., and Green, M. J., Anal. Proc., 1988, 25, 276. Wring, S. A . , Hart, J. P., Bracey, L.. and Birch, B. J., Anal. Clrim. Acta, 1990, 231, 203. Halbert, M. K.. and Baldwin. R. P., J . Chromarogr., Biomed. Appl., 1985. 345, 43. Wring. S. A., Hart, J . P.. and Birch, B. J., Analyst, 1989, 114, 1563. Wring, S. A . , Hart, J. P., and Birch, B. J., Analyst, 1989, 114, 1571. Imisides, M. D., Wallace, G . G . , and Wilke, E. A . , TrAC, Trends Anal. Chem., Pers. Ed.. 1988, 7, 143. Meistcr, A., J . Biol. Cliem., 1988, 263, 17205. Black. M.. Ann. Rev. Med.. 1984, 35. 577. Curello. S., Ceconi, C.. Cargnoni, A , , Cornacchiari. A , , Ferrari, R., and Albertinia, A., Clin. Clzem., 1987, 33, 1448. Hall. R., and Malia, R. G., Medical Laboratory Haematology. Butterworths, London, p. 294. Eastman, K. D., Clinical Haematology. Wright, Bristol, 6th edn., 1984. p. 126. Perrett, D., and Rudge, S. R.. J. Plrarm. Biomed. Anal., 1985, 3, 3. Galus. Z., Fundamentals of Electrochemical Analysis. Ellis Horwood, Chichester. 1976, p. 237. Laviron, E., J. Electroanal. Chem., 1979, 101, 19. Laviron, E., and Roullier, L., J . Electrounal. Chem., 1980, 115, 65. Swain, A.. Int. Ind. Biotechnol.. 1988, 8. 11. Bennetto, H. P., Dekeyzer. D . R.. Delaney, G. M., Koshy, A., Mason, J . R., Razack, L. A . , Stirling, J . L., and Thurston, C. F., Int. Analyst, 1987, 8, 22. Wring, S. A., Hart, J . P.. and Birch, B. J., Anal. Chim. Acta, 1990, 229, 63. Paper 0103295F Received July 23rd, I990 Accepted September 27th, 1990

ISSN:0003-2654    

DOI:10.1039/AN9911600123

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1991, VOL. 116 131 Accumulation Voltammetry of Copper(i1) Using a Carbon Paste Electrode Modified With Di-8-quinolyl Disulphide Kazuharu Sugawara, Shunitz Tanaka and Mitsuhiko Taga* Department of Chemistry, Faculty of Science, Hokkaido University, Nishi-8, Kita- 10, Kita-ku, Sapporo-shi, Hokkaido 060, Japan The accumulation voltammetry of copper(\\) was investigated with a carbon paste electrode modified with di-8-quinolyl disulphide (DQDS). The DQDS was reduced to quinoline-8-thiol by applying a suitable potential. Copper(ii) was accumulated on the electrode as the copper(ii)-quinoline-8-thiol complex at a constant potential in 0.1 mol dm-3 acetate buffer. The reduction peak of the copper(1i) complex was then observed at -0.30 V by scanning the potential in a negative direction using the differential-pulse mode.The calibration graph for copper(l1) was linear over the range 3 x 10-9-2 x 10-6 mol dm-3 with accumulation for 5 min at -0.05 V. As copper(ii) was selectively accumulated on the electrode, the influence of concomitant ions was negligible. The method was applied to the determination of copper(ii) in Geological Survey of Japan rock reference materials. Keywords: Accumulation voltammetry; copper(//) determination; di-8-quinolyl disulphide; quinoline-8-thiol Voltammetric methods involving an accumulation process, such as stripping voltammetry and adsorptive voltammetry, arc generally very sensitive for the determination of trace metal ions. However, their application to practical samples is limited by the interferences from the sample matrix.In order to solve this problem, it is necessary to accumulate the analyte on an electrode selectively. For this purpose, chemically modified electrodes (CMEs) have been used; for example, a glassy carbon electrode the surface of which was modified by a ligandl.2 and a carbon paste electrode (CPE) prepared by mixing a ligand and graphite powder. The CPE is easily prepared and gives a stable electrode response. The CPEs modified with dimethylglyoxime3 and diethylenetriamine4 were applied to the selective accumulation of nickel(i1) and silver( I ) , respectively, and the determination of trace metal ions in natural water was achieved. In addition, the CPE modified with the iron(rr)-pyridine-4-carbaldehyde complex was able to accumulate organic compounds such as primary amines by the formation of a Schiff's base.5 We have also reported on the accumulation behaviour and stripping voltam- metry of silver(1) with a CPE modified with thia-crown compounds.6 In this paper, the accumulation and voltammetric behav- iour of copper(I1) using a CPE modified with di-8-quinolyl disulphide (DQDS) is described.For the accumulation of copper(ii) on an electrode, CMEs modified with 2,9-dimethyl- l,lO-phenanthroIine7 and diethyldithiocarbamateg have been reported. However, the pre-treatment of the electrode modified with 2,9-dimethyI-l,lO-phenanthroline is trouble- some and the electrode modified with diethyldithiocarbamate lacks selectivity. On the other hand, DQDS has been used for the selective solvent extraction of copper(1i) from acidic medium .9 The reagent is stable and insoluble in water, and forms quinoline- 8-thiol on reduction.The voltammetric behaviour of metal- quinoline-8-thiol complexes has also been studied. 1(&12 Hence it should be possible to develop a highly sensitive and selective method for the determination of copper(ri) by using a CPE modified with DQDS. The influence of concomitant ions on the method was found to be negligible. The method was applied to the determination of copper(i1) in Geological Survey of Japan rock reference materials. * To whom corrcspondence should be addressed. Experimental Apparatus A polarographic analyser [Princeton Applied Research (PAR) 174Al coupled to a PAR 315 electroanalytical con- troller and an Omnigraphic Model 2000H x-y recorder were used.A glassy carbon rod was used as the counter electrode, and a saturated calomel electrode (SCE) with a diaphragm tube containing 1 rnol dm-3 potassium nitrate was used as the reference electrode. All potentials were measured against the SCE. The measurements were carried out at 25 k 0.1 "C. Preparation of the CPE Graphite powder (0.098 g) was added to DQDS (0.002 g) and about 0.06 ml of liquid paraffin and mixed well in a mortar with a pestle to obtain a paste. A portion of the paste was taken and used to fill the top of a glass tube; the surface of the electrode was then smoothed with the end of a spatula. The active surface area of the electrode used was 0.07 cm2. Reagents Di-8-quinolyl disulphide was purchased from Dojindo Chem- ical Laboratories and used without further purification.A 1 X 10-2 mol dm-3 stock solution of copper(I1) was prepared by dissolving copper metal (99.999%, Mitsuwa Chemicals) in 5% HN03. The supporting electrolyte was acetic acid-sodium acetate buffer (pH 4.5). High-quality nitrogen was used for de-aeration. All other reagents were of analytical-reagent grade. Procedure After de-aerating the supporting electrolyte containing cop- per(ii), the latter was accumulated on the electrode for 5 min at -0.05 V. After a rest time of 15 s, the potential was scanned in the negative direction and voltammograms were recorded by differential-pulse polarography (pulse amplitude, 50 mV; pulse interval, 1 s; and scan rate, 5 mV s-1).132 ANALYST, FEBRUARY 1991, VOL.116 1 I 0 -0.2 -0.4 -0.6 EIV versus SCE Fig. 1 Accumulation voltammograms of the copper(i1)-DQDS complex. Accumulation on a CPE containing 2% DQDS at -0.05 V for S min in 0.1 rnol dm-3 acetate buffer. A, Blank; B, S x 10-8; C, 1 x and E, 4 x 10-7 rnol dm-3 copper(i1). Pulse amplitude, SO mV; pulse interval, 1 s; and scan rate, 5 mV s-l D, 2 x Results and Discussion The voltammograms obtained with the CPE modified with DQDS are shown in Fig. 1. In the supporting electrolyte solution, no peak appeared in the potential range between -0.05 and -0.55 V and the residual current was stable in this range. When the accumulation process was carried out for 5 min at -0.05 V in a solution containing 5.0 x 10-8, 1.0 x 10-7, 2.0 x 10-7 or 4.0 x 10-7 rnol dm-3 copper(iI), the reduction wave appeared at about -0.30 V.The peak current was proportional to the concentration of copper(r1) and the accumulation time. At an unmodified CPE, the reduction peak of copper(1r) was not observed. It is thought that the accumulation of copper(i1) in solution occurs because of the formation of a complex between copper(r1) and the reagent (DQDS) on the electrode. Relationship Between the Accumulation Potential and pH The reduction peak current of copper(I1) obtained with the CPE modified with DQDS depends on the accumulation potential and on the pH of the supporting electrolyte. Therefore, the relationship between the peak current and the accumulation potential was investigated at various pH values using acetate (pH 4 . 9 , phosphate (pH 7.0) and ammonia (pH 9.0) buffers.As shown in Fig. 2, the reduction current of copper(r1) could be observed over a wide range, viz., +0.2 to -0.20 V, in the acetate buffer (Fig. 2, A), demonstrating that copper(r1) could be accumulated on the electrode over this range of accumulation potential. The largest and most reproducible peak current was obtained at -0.05 V (optimum potential). In phosphate (Fig. 2, B) and ammonia (Fig. 2, C) buffers, the same relationship as that for the acetate buffer was obtained, but the potential was shifted in a negative direction. However, no significant current was obtained in hydrochloric acid (pH 1.0) or sodium hydroxide (pH 13) solution. When the accumulation was carried out at a more positive potential than the optimum potential, the peak current decreased.The accumulation potential is also related to the reduction of DQDS to quinoline-8-thiol. It appeared that the amount of copper(1r)-quinoline-8-thiol which was accumulated on the electrode was very small because the reduction of DQDS to quinoline-8-thiol was difficult at more positive potentials. At a more negative potential than the +0.2 +0.1 0 -0.1 -0.2 -0.3 E,,,N versus SCE Fig. 2 Dependence of the peak current (i,) on the accumulation potential (Eacc). Accumulation on a CPE containing 2% DQDS for 5 min in various supporting electrolytes of concentration 0.1 mol dm-3 in the presence of 4 X lo-’ mol dm-3 copper(i1). A, Acetate buffer (pH 4.5); B, phosphate buffer (pH 7.0); and C, ammonia buffer (pH 9.0) 9 8 I 7 0. 6 5 +0.10 +0.05 0 -0.05 -0.10 -0.15 EIV versus SCE Fig.3 Relationship between reagent peak and maximum of the accumulation potential. A, Potential of reagent peak at which DQDS was reduced to quinoline-8-thiol; and B, optimum potential in various supporting electrolytes. 0, Acetate buffer (pH 4.5); A, phosphate buffer (pH 7.0); and 0, ammonia buffer (pH 9.0) optimum potential, the peak current decreased sharply because the reduction of the complex had already occurred at the optimum potential. When copper(r1) was accumulated at each optimum potential, the peak current was independent of the nature of the supporting electrolyte and the same peak currents were observed. However, because the residual current increased with an increase in pH, acetate buffer, in which the residual current was relatively small, was used.Mechanism of the Electrode Reaction As mentioned above, a dependence of the peak current on the accumulation potential was observed, showing that a potential must be applied to the electrode in order to accumulate copper(r1). Further, the optimum accumulation potential shifted to a more negative value as the pH of the solution increased (Fig. 3). The shift of optimum accumulation potential was analogous to the shift of peak potential at which DQDS was reduced to quinoline-8-thiol. Consequently, it appears that an accumulation potential is necessary to reduce DQDS to quinoline-8-thiol. Therefore, the mechanism of the electrode reaction could be as follows: DQDS is first reduced to quinoline-8-thiol [reaction (l)]; copper(rr) is then accumulated on the electrode as the copper(i1)-quinoline-8-thiol complex by the reaction between copper(r1) and the quinoline-8-thiol now present on the electrode [reaction (2)].ANALYST, FEBRUARY 1991, VOL.116 133 s-s SH Cull+2 -c @ g + 2 H + SH \ /N N\ / The copper(i1)-quinoline-8-thiol complex on the electrode is reduced at -0.30 V to give copper metal and 'free' quinoline-8-thiol [reaction (3)]. Effect of the Accumulation Time The relationship between the peak current and the accumu- lation time is shown in Fig. 4. The peak current increased linearly over a period of 10 min on increasing the accumu- lation time in both 1 X 10-7 and 5 X 10-7 rnol dm-3 copper(r1) solutions. However, the accumulation time was kept at a constant value greater than 10 rnin for the 5 x 10-7 rnol dm-3 solution because the amount of reagent on the surface of the electrode was saturated for copper(u) ions.5.0 7 0 5 10 t,,,/min 15 Fig. 4 Dependence of the peak current (ip) on the accumulation time (fact). Accumulation on a CPE containing 2% DQDS at -0.05 V in 0.1 rnol dm-3 acetate buffer containing A. 1 x 10-7; and B, 5 x mol dm-3 copper(i1) 3.0 a sa 2.0 .- 1.c I I I 0 10 20 DQDS:(graphite + DQDS) (%) Fig. 5 Effect of amount of DQDS. Accumulation at -0.05 V for 5 min in 0.1 mol dm-3 acetate buffer containing A, 2 x 10-7; and B, 4 X lo-' rnol dm-3 copper(rr) Effect of the Amount of DQDS The effect of altering the ratio of DQDS to graphite powder in the paste on the peak current was investigated (Fig. 5 ) . The peak current increased with increasing amounts of DQDS up to 1% and decreased above 3%.This might be related to the destruction of the mechanical integrity of the paste. It is thought that electron transfer at the CPE is adversely affected by increasing the amount of DQDS. The ratio of DQDS to graphite in the paste was fixed at 2% so that the peak current was constant. Calibration Graph When accumulation was carried out for 5 min at -0.05 V in 0.1 rnol dm-3 acetate buffer followed by a potential scan using differential-pulse polarography , the calibration graph was linear over the range 3 x 10-9-2 x 10-6 rnol dm-3 copper(i1) with a correlation coefficient of 0.997. The detection limit was 8 x 10-10 rnol dm-3 at an accumulation time of 5 min. The relative standard deviation for five determinations of 5 x 10-8 and 5 x 10-7 rnol dm-3 copper(1r) was 3.5 and 2.5%, respectively.Influence of Concomitant Ions The influence of concomitant ions is shown in Table 1. It is known that Ag', B P , Co", Fe"', PtIV, RulI1, Sblll, SeIV and Zn" can form complexes with quinoline-8-thiol. Measurements were carried out for concomitant ions at concentrations of 50 and 100 times that of copper(1r). The values in Table 1 show the ratio of the peak current in the presence of the Table 1 Influence of concomitant ions on the peak current of copper(1i). Accumulation on a CPE containing 2% DQDS at -0.05 V for 5 min in 0.1 rnol dm-3 acetate buffer (pH 4.5) containing 4 x lop7 rnol dm-3 coppcr(ii) Concentration of concomitant ion/ pmol dm-3 20 40 20 40 20 40 20 40 20 40 20 40 20 40 20 40 20 40 20 40 Signal ratio (YO) 100 99 95 89 103 99 102 101 98 99 102 97 100 100 102 102 97 98 103 100 Table 2 Determination of copper(i1) in Geological Survey of Japan rock standard reference materials Concentration of copper(rI)/pg g-' Proposed Recommended Sample method t SD" value JA-1 (Andesite) 204 t 7 203 JB-1 (Basalt) 58 2 2 56 JG-1 (Granite) 1.6 t 0.1 1.5 * n = 12.134 ANALYST, FEBRUARY 1991, VOL. 116 concomitant ions to that in their absence (as a percentage).When Bi"' was added in a 100-fold excess it interfered in the determination of copper(r1). As bismuth ions form a complex with the reagent, the reduction peak for the copper(i1) complex may decrease. However, the reduction peak for bismuth was not observed. In anodic stripping voltammetry, the presence of an equivalent amount of Bill' caused an interference in the determination of copper(i1).In the proposed method, the peak current was not affected even by the addition of a 50-fold excess of Bi"' to the solution. Also, other ions in Table 1 did not interfere up to 40 pmol dm-3. This is because the accumulation of copper(1r) on the electrode occurs selectively. Application of the Method to Rock Standard Reference Materials The method was applied to the determination of copper(i1) in rock reference materials prepared by the Geological Survey of Japan. The sample was decomposed with a mixture of HN03, HC104 and HF under pressure at 150 "C using a PTFE bomb. After the solution had been evaporated to dryness, the residue obtained was dissolved in dilute HN03.An aliquot of this solution was taken for the determination. The results are shown in Table 2. The values obtained using the proposed method agree with the recommended values; hence the proposed method is suitable for the determination of cop- per(I1) in rocks. Conclusion Di-8-quinolyl disulphide was reduced to quinoline-8-thiol by applying a suitable potential. By using the reaction between copper(1i) and the quinoline-8-thiol produced, the former could be selectively accumulated on the electrode as the copper(u)-quinoline-8-thiol complex. A highly sensitive and selective method for measuring copper(I1) was developed using accumulation voltammetry. The method was applied to the determination of copper(i1) in rock samples. References 1 2 3 4 5 6 7 8 9 10 11 12 Isutsu, K., Nakamura, T . , Takizawa, R., and Hanawa, H., Anal. Chim. Acta, 1983, 149, 147. Chastel, O., Kauffman, J.-M., Patriarche, G. J., and Christian, G . D., Anal. Chem., 1989, 61, 170. Baldwin, R. P., Christensen, J. K., and Kryger, L., Anal. Chem., 1986,523, 1790. Cheek, G . T., and Nelson, R. F., Anal. Lett., 1978, A l l , 393. Guadalupe, A. R., Jhaveri, S. S . , Liu, K. E., and Abruna, H. D., Anal. Chem., 1987, 59, 2436. Tanaka, S., and Yoshida, H., Talanta, 1989, 36, 1044. Prabhu, S. V., and Baldwin, R. P.,Anal. Chem., 1987,59,1074. Guadalupe, A. R., and Abruna, H. D., Anal. Chem., 1985,57, 142. Bankovsky, Yu. A., Ievins, A. F., Luksha, E. O., and Bochkans, P. Ya., Zh. Anal. Khim., 1961, 16, 150. Toropova, V. G., Budnikov, K., and Zhiyangulova, F. G., Zh. Obshch. Khim., 1976, 46, 1125. Toropova, V. G., Budnikov, K., and Zhiyangulova, F. G., Zh. Obshch. Khim., 1977,47, 1148. Nakabayashi, Y., Masuda, Y., and Sekido, E., J. Electroanal. Chem., 1986, 205, 209. Paper 0102862 B Received June 26th, 1990 Accepted September 21st, 1990

ISSN:0003-2654    

DOI:10.1039/AN9911600131

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1991, VOL. 116 135 Comparative Performance of 14-Crown-4 Derivatives as Lit hiu m-selective Electrodes Ritu Kataky, Patrick E. Nicholson and David Parker* Department of Chemistry, University of Durham, South Road, Durham DHI 3LE, UK Arthur K. Covington Department of Chemistry, University of Newcastle, Newcastle-upon-Tyne, UK A series of neutral ionophore-based lithium-selective liquid-membrane electrodes have been prepared and the electrode performance compared with similar electrodes based on the lithium ionophores ETH 1810-ortho-nitrophenyl octyl ether (oNPOE) and ETH 2137-bis(l-butylpentyl) adipate (BBPA). By using a diamide substituted 14-crown-4 macrocycle, selectivities for Li+ in the presence of Na+ of log q;ha = -3.25 and -2.92 were obtained for diisobutylamide-oNPOE and di-n-butylamide-oNPOE derivatives.The di-n-butylamide-oNPOE based electrode functioned satisfactorily in serum, exhibiting a fast response time (10-15 s), an acceptable lifetime of 50 d and minimal protein interference. Keywords: Lithium; ion-selective electrode; selectivity; crown ether; serum An ideal Li+ ionophore for use in monitoring the concentra- tion of Li+ ions in blood during manic depressive psychosis therapy has yet to be found. The search is imperative as a close monitoring of lithium concentration during treatment is required in order to secure a therapeutic effect and to avoid an overdose of lithium which could lead to fatal poisoning.' There is a narrow gap between therapeutic and toxic levels (Fig. 1). A series of chiral 14-crown-4 derivatives have been systematically synthesized and characterized (Table 1).Their performance, in serum, has been compared with the perfor- mance of commercial Li sensors and with results obtained by atomic absorption measurements. A diisobutylamide and a di-n-butylamide 14-crown-4 based sensor exhibit high selectiv- ity for Li+ over Na+ with log kc>tNa values of -3.25 and -2.92, respectively. These are superior to the best Li+ in Na+ selectivities reported previously (ETH 1810, log kEtN, = -2.45).'-11 The diisobutylamide derivative, however, behaves poorly in serum. The di-n-butylamide based elec- trode functions satisfactorily in serum, exhibits a fast response time (about 10-15 s) and has an acceptable lifetime. Table 1 14-Crown-4 derivatives studied R R 2.14C4Thio 4. 14Cd Dibenz rH2OCH2Q] 3. 14C4 Monobenz (CH20CH2Ph) 5. 14C4 01 7. 14c4 Est (CHzOH) (CH2C02Me) 9. 14C4 Butam (CH~CONBU~) ( CH ;OCH zP h ) 6. 14c4 Diol (CH20H) 8. 14c4 Diest (CH2C02Me) 10. 14C4 Dibutam 11. 14c4 Diibutam (CH~CONBU~) [CHZCON(~BU)~] * To whom correspondence should be addressed. Experimental Design of the Ionophores A 14-crown-4 skeleton has been the basis of previous Li+ ionophores, as it has an optimum cavity size for incorporating Li+ ions.4-8 The substituents in these earlier sensors were chosen for their effect in improving the lipophilicity of the sensor rather than their effect in enhancing Li selectivity in m I E 2 1.5 E E . - .- -I Y 1 .o 0.5 0 2 4 6 8 10 Consecutive controls Fig. 1 Monitoring Li levels during therapy.The first two controls were made once a week after patient A, 0, was started on 800 mg d-I of Li2C03. Side effects were detected and the patient was taken off Li therapy for 3 months. Therapy was restarted at 400 mg d-l (third control) and gradually increased to 600 mg d-l (fourth control). The fifth control showed a high Li level of 1.2 mmol dm--7 corresponding to a dosage of 1200 mg d- l . The following controls correspond to oncc every month while the dosage was maintained at 800 mg d-'. Patient B, 0, was started on 250 mg d-I of Li2C03 (first control) which was increased to 750 mg d-I (second control). The toxic level at control 3 corresponds to administration of a diuretic. The treatment was stopped and restarted after 3 months at 250 mg d-1 (fourth control), gradually increased to 500 mg d-l (controls 5 and 6) and then maintained at 800 mg d-I (controls 7-12). (Blood samples were taken 12 h after administering Li2C03 tablets.The instrument used was a Corning 405 flame photometer)136 ANALYST, FEBRUARY 1991, VOL. 116 binding. Improving the selectivity for complexing Li+ in the presence of excess of Na+ has been the central aim of our recent work. The Li+ ion may exhibit octahedral coordination. This suggested the possibility that by incorporating additional ‘axial’ donor sites on the 14-crown-4 ring, 1 : 1 complexation with Li+ would be favoured, while the formation of 2 : 1 complexes (Li+ : M+) would be suppressed. This tendency to form 2 : 1 complexes is particularly important for Na+ and K+, and is suppressed with the sterically demanding ‘axial’ substituents incorporated in the 14-crown-4 ring.The length of the side chains was chosen bearing in mind, not only purely geometric considerations for optimizing amide oxygen-Li+ interactions, but also the fact that a chelate-ring size of six intrinsically favours Li+ complexation whereas a five-mem- bered ring favours binding by the larger K+ and Na+ ions. Furthermore, Li+ is a small, ‘hard’, polarizing ion, hence ‘hard’ o-donors with large dipole moments, such as amide or phosphonate groups, should enhance Li+ in Na+ selectiv- ity. 10.11 Taking these factors into account, the 14-crown-4 derivatives shown in Table 1 were synthesized according to published procedures. 11 Cl Reagents and Chemicals Chloride salts of lithium, sodium, potassium and magnesium (BDH AnalaR) were dried at 50 “C and stored over silica gel.Calcium chloride solution (BDH AnalaR, 1 mol dm-3) was used. All standard solutions were prepared in de-ionized water and their cation concentrations checked by atomic absorption spectrometry. The 3-morpholinopropanesul- phonic acid (MOPS) and 4-(2-hydroxyethyl)piperazine-l- ethanesulphonic acid (HEPES) buffers12 and their sodium salts were obtained from Sigma. The materials for the electroactive membranes were: high relative molecular mass poly(viny1 chloride) (PVC) ; ortho- nitrophenyl octyl ether (oNPOE); bis( 1-butylpentyl) adipate (BBPA); potassium tetrakisb-chloropheny1)borate (K- TpCIPB); and lithium ionophore IV (ETH 2137), all obtained from Fluka; and a Philips Li ionophore (IS, 561).Tetrahydro- furan (THF) was of spectroscopic grade, distilled from sodium benzophenone ketyl. Serum was collected from healthy volunteers and was not stabilized by addition of heparin, but was used immediately. Apparatus Calibration and selectivity measurements A Philips IS (561) electrode body was used to mount the electroactive membranes. The filling solution was 1 x 10-3 mol dm-3 LiCI. The electrodes were fitted in a constant- volume cell made from a water-jacketed glass tube with a B19 ground-glass joint. The cell was incorporated in a flow system (Fig. 2). The reference electrode (porous plug, calomel, filled with 3.5 mol dm-3 KCI, RE1 Petiacourt) was placed, downstream, in a reference cell made from a water-jacketed glass tube from which a short capillary fitted with a ceramic plug had been drawn out.This formed a constrained diffusion junction with the sample. The temperature of the system was maintained at 37 “C by using a Techne Tempette Junior TE-85 thermostatically controlled bath. The solution was drawn at a constant rate (about 3 ml min-1) using an RS 330-812 peristaltic pump. The ion-selective and reference electrodes were connected to a digital multimeter (Keithley 197, Autor- anging Microvolt DMM) via a buffer amplifier. A flat-bed Linseis 17100 chart recorder, provided with back-off facilities, was used for monitoring potential differences. A suitable capacitance was connected across the input of the chart recorder to smooth out residual noise. I l l Initial Diluent Constant solution solution volume cell mstant volume cell Reference cell - Waste Fig.2 Constant volume cell and reference cell ( a ) Flow system in which the cell was incorporated. (b) Blood plasma studies The electroactive membranes were mounted on poly- carbonate stems, according to published procedures. 13 The cells were connected to a multiple switch (Fig. 3). The solution was injected into the system by use of disposable 1 ml syringes. The signal monitoring system was similar to that described above. Measurements were made at ambient temperature. Atomic absorption measurements were carried out on a Perkin-Elmer (5000) instrument. A Corning 654 (Na+, K+, Li+) analyser was used for comparative studies. The determinations of refractive indices of serum for plasma water were made on a Goldberg AO, TS meter and concentrimeter.Membrane preparation The oNPOE-based membranes were composed of 1.2% ionophore, 65.6% oNPOE, 32.8% PVC and 0.4% KTpClPB in 6 cm3 of THF. The ETH 2137 and a 14-C-4 di-n-butylamide sensor were also made into membranes consisting of 2.0% ionophore, 65.6% BBPA and 32.4% PVC in 6 cm3 of THF. The membranes were cast according to published proce- dures. 14 Procedure Calibration and selectivity measurements The ion-selective electrodes (ISEs) were calibrated using a constant dilution technique.15 A fixed interference method was used for selectivity coefficient measurements. In order toANALYST, FEBRUARY 1991, VOL. 116 300 280 260 240 220 200 $180 € ail60 140 120 100 80 137 - - - - - - - - - - - - Chart recorder Syringe Reference I I Waste Flow diagram showing the cells connected to the multiple Fig.3 switch establish the behaviour of the electrodes in the ionic concen- trations present in blood, a background interferent solution of 150 mmol dm-3 NaCl, 4.3 mmol dm-3 KCI and 1.26 mmol dm-3 LiCI was used and all of the measurements were made at 37°C. Blood plasmu studies Initial studies were carried out using three different dilutions of serum (serum + diluent: 1 + 1, 1 + 4 and 1 + 9) each containing a range of LiCl concentrations, 0.25,0.50,0.75 and 1.00 mmol dm-3. The diluent was 145.0 mmol dm-3 NaCl, 5.0 mmol dm-3 NaMOPS, 6.7 mmol dm-3 MOPS and 4.0 mmol dm-3 KCl (pH = 6.86). The Corning 654 (Na+, K+, Li+) analyser was used, according to the manufacturer’s instructions, for comparative studies.16 Further tests were carried out with the best ionophore and ETH 2137 (the ionophore used in the Corning analyser). An aqueous solution of 140.0 mmol dm-3 NaCl, 4.0 mmol dm-3 KCl, 1.2 mmol dm-3 CaCI2 (20 ml) and a serum sample (20 ml) were dosed with 0.1 mol dm-3 LiCl solution to give Li+ concentrations ranging from 0.5-5.0 mmol dm-3 (AQ1-7 and PLl-7, Tablc 4). The calibration solutions used were: Cal 1 [ 135.0 mmol dm-3 NaCl, 3.6 mmol dm-3 KCI, 1 .0 mmol dm-3 LiCI, 5.0 mmol dm-3 NaMOPS, and 6.7 mmol dm-3 MOPS, pH = 6.86, ionic strength ( I ) = 144.61 and Cal2 (135.0 mmol dm-3 NaCl, 3.6 mmol dm-3 KCI, 2.5 mmol dm-3 LiCI, 5.0 mmol dm-3 NaMOPS, and 6.7 mmol dm-3 MOPS, pH = 6.86,I = 146.1). The cells were flushed with deionized water and air and calibrated with Cal 1 and Cal 2 prior to each sample injection.Atomic absorption measurements were performed in parallel. Results and Discussion The results of the calibrations of the various 14-crown-4 based ionophores in comparison with those used by Philips and Corning both in aqueous LiCl and in LiCl in the presence of interferents (150.0 mmol dm-3 Na+, 4.3 mmol dm-3 K + , and Table 2 Characteristics of Li ionophores. Theoretical slope at 37 “C = 61.54 mV decade-’ LiCl in a solution of LiCl in (ISONaCI, 4.3 KCI, de-ionizcd 1.26 CaC12)/ water mmol dm-3 Philips 14-crown-4-thio I4-crown-4-monobenz 14-crown-4-dibenz 14-crown-4-01 14-crown-4-diol 14-crown-4-est 14-crown-4-diest 14-crown-4-nbutam 14-crown-4-dinbut am 14-crown-4-diibutam ETH 2137 Slope 62.0 60.0 53.1 60.0 37.3 32.0 54.5 62.0 56.0 60.0 50.0 62.0 * LD = Log (limit of detection).LD* -4.5 -5.1 -4.9 -4.6 -4.6 -3.2 -3.9 -5.0 -5.0 -4.4 Slope LD” Logky’ 47.0 -2.6 -1.89 44.0 -2.0 -1.14 45.0 -2.2 -1.35 62.0 -2.6 -1.77 - - - - 25.0 - 36.0 - 61.0 -3.8 -2.92 61.0 -4.1 -3.25 60.0 -2.7 -1.90 - 6o i Therapeutic range 1.2-0.7 mmol dm-3 0 1.0 2 0 3.0 4.0 5.0 6.0 -Log [Li] Fig. 4 Li ISE slopes in the presence of interferents: Na+, 150; K + . 4.3; and Gal+, 1.26 mmol dm-3. A, Ideal; B, diisobutylamide C. di-n-butylamide; D, ETH 2137; E, dibenzyl: and F, Philips 1.26 mmol dm-3 Ca’+) are given, (Table 2). These prelimi- nary tests revealed that the alcohol derivatives behaved very poorly. The di-substituted derivatives were better behaved than their mono-substituted analogues.The dibenzyl, di-n- butylamide and diisobutylamide derivatives performed well even in aqueous solutions containing serum levels of sodium, potassium and calcium, the last two showing a significant improvement over the ETH 2137 and Philips (IS 561) Li+ sensors (Fig. 4). The selectivity coefficients of the most promising iono- phores were determined in interferent concentrations of 0.1 mol dm-3. The only exception was proton interference, in which a 1 X 10-3 mol dm-3 HCl solution was used. The values obtained are shown in Table 3. The target selectivity coefficients were calculated using the Nikolsky-Eisenman equation based on a contribution of less than 1%, by the activity of the interferent ion, in comparison with the activity of the primary ion.” The values for the ‘best’ lithium ionophore, to date, (ETH 1810) are included for comparison.The diisobutylamide and the di-n-butylamide ionophores meet the target selectivities for K+, Ca’+ and Mg2+. As the pH of plasma is 7.3-7.4, the proton interference measured at pH 3 is not important.The selectivity observed with respect to Na+ is superior to ETH 1810 and ETH 2137.138 ANALYST, FEBRUARY 1991, VOL. 116 Preliminary tests with the two amide ionophores exhibiting the best selectivities in vitro, were performed with diluted serum to give an indication of their stability in biological media. The results were compared with those obtained using a Corning 654 (Na+, K+, Li+) analyser in both the concentra- Table 3 Selectivity coefficients of Li ionophores in 0.1 mol dm-3 interferent background Log k r t Na K Ca Mg H* Target values (-4.3 <-2.8 <-3.0 (-3.5 <-2.1 Philips -1.3 -2.15 -3.22 -3.7 -1.0 14-crown-4-dibenz -1.4 -2.3 -4.5 -5.8 -3.5 14-crown-4-dinbutam -3.0 -3.5 -4.2 -5.7 -0.9 14-crown-4-diibutam -2.9 -4.3 -4.3 -5.3 1.1 ETH 1810? -2.45 -2.6 -2.7 -4.0 -1.0 ETH2137 (Corning) -1.9 - - - - * Proton interference in 0.001 mol dm-3 HCl (pH = 3).? Values from the Fluka Selectophore catalogue. Table 4 Protein interference. Error ( Y o ) = { [C(actual) - C(expected)]/ [ c ( , , ~ ~ ~ ~ ~ ~ J } ~ 1 0 0 . kYt = 1.2 x 10-2 Corning, 1.2 x 10-3 di-n- butylamide, 6.3 X 10-4 diisobutylamide Error (YO) Plasma : water [Li]/ ratio mol dm-3 Corning* Corning? 1 : 10 0.22 0 -58.8 0.40 0 -6.7 0.60 -1.3 -6.0 0.80 -2.1 -5.0 1 : 5 0.20 0 -46.7 0.35 -0.3 -11.1 0.54 -2.8 -8.9 0.72 -4.8 -5.2 1 : 2 0.125 -100.0 -52.6 0.22 -17.4 -9.1 0.34 -18.1 -9.1 0.42 -18.9 -7.9 * Concentration mode.? Millivolt mode. Di-n- butyl- amide 0 0 0 -1.2 0 0 -3.0 -4.2 -44.0 -6.8 -7.8 -7.4 Diiso- amide butyl- -30.6 -24.0 +5.4 +12.9 -100.0 - 100.0 -65.0 -30.7 - 100.0 -91.3 -82.8 -52.2 tion and millivolt mode. Voltage readings were converted to concentrations using the equation where E,, cLi(u) and cNa(,) are the voltage reading, Li+ concentration and sodium concentration, respectively, in the unknown sample and E c a l 1 , ~ ( ~ ~ 1 1 ) and C N a ( C a l 1 ) are the 5.0 - ( a ) 4.5 - 4.0 - 3.5 - 3.0 - 2.5 - 2.0 - 4.0 3.5 2.5 3.0 i *.O 1 & 0 1.5 1.0 1 !i 0.51 1 , , , , , 0 1 2 3 4 5 6 7 Sample no. Fig. 5 Comparison of atomic absorption spectrometry, ETH 2137- BBPA and di-n-butylamide-oNPOE ISEs in (a) a ueous solution and (b) plasma.0, ETH 2137; ., di-n-butylamide; $. flame; A , ETH 2137; and 0, di-n-butylamide. 0 and H, Using concentrations; and A , and 0, using activities Table 5 Comparisonsf flame photometry, ETH 2137-BBPA ISE and di-n-butylamide-oNPOE ISE, using concentrations directly. Concentration units, mmol dm-3; cun, uncorrected concentrations; cCrr concentrations corrected for plasma water; refractive index for plasma sample = 1.354; plasma water = 91.5%; “a] in plasma determined by atomic absorption spectrometry = 148 mmol dm-3; error (%) = (cISE - cFlame)/ CFlame ETH 2137* Di-n-butylamide” Flame? Solution C Error (YO) C Error(%) c AQ1 0.40 +1.1 0.40 +1.1 0.39 AQ2 0.86 +6.2 0.78 -3.7 0.81 AQ3 1.24 -13.9 1.39 -3.5 1.44 AQ4 1.58 -9.2 1.70 -2.3 1.74 AQ5 2.19 -14.0 2.41 -5.5 2.55 AQ6 3.28 -15.0 3.67 -4.9 3.86 AQ7 3.79 -14.4 4.22 -4.7 4.43 Cun PL1 0.57 PL2 0.97 PL3 1.43 PL4 1.69 PL5 2.40 PL6 3.34 PL7 4.29 CC, 0.62 1.06 1.56 1.85 2.62 3.65 4.69 Error (% ) + 19.2 +0.9 +3.3 +0.5 -7.1 -6.6 -6.6 Cun 0.46 0.99 1.37 1.65 2.43 3.39 4.37 CC, 0.50 1.08 1.50 1.81 2.66 3.71 4.78 Error (YO) -3.8 0.52 +2.8 1.05 -0.7 1.51 -1.6 1.84 -5.6 2.82 -5.1 3.91 -4.8 5.02 * Error on ISE values, about 1% (equivalent to k0.3 mV).? Error on flame values, about 3%.ANALYST, FEBRUARY 1991, VOL. 116 139 corresponding values in calibration solution, Cal 1. kC.tNa is the appropriate selectivity coefficient and S is the slope of the electrode at ambient temperature.Concentrations were cor- rected18 for 93% plasma water. Expected values were based on atomic absorption measurements of Li+ in the diluents (Table 4). The diisobutylamide based derivative behaved very poorly, whereas the performance of the di-n-butylamide based electrode is better than that of the Corning ETH 2137 electrode. Errors are larger in solutions with higher serum and lower lithium concentrations. The approximate 18% error (Table 4) observed in the Corning concentration mode, which is the commonly used mode, is disconcerting. To substantiate these results, further tests were performed on the di-n-butylamide-oNPOE based electrode and an ETH 2137-BBPA based electrode made from the parent iono- phore. (Calibrations with electrodes based on the opposite 0.005 0.004 0.003 0.002 r t 0.001 0 28 32 36 40 44 48 52 Days Lifetime of a 14-crown-4 di-n-butylamide ISE used in serum Fig.6 for 4 weeks combination, viz., di-n-butylamide-BBPA and ETH 2137- oNPOE in aqueous solutions containing Na+, K+, Ca2+ and Li+ at levels normally present in blood, showed that these combinations were unsuitable for use as clinical sensors.) Aqueous samples AQ1-AQ7 and plasma samples PL1-PL7 were injected into the flow cells and readings were taken as before. Atomic absorption measurements, on each sample, were performed in parallel. The results, shown in Table 5 , were corrected for the presence of protein using refractive index measurements. The di-n-butylamide electrodes, again, appeared to be superior to the ETH 2137 electrode. The errors are higher for plasma samples containing >2.0 mmol dm-3 Li+.Activity coefficients for Cal 1, Cal 2 and solutions AQ1- AQ7 were calculated using the Pitzer equation19 (Table 6). Concentrations were calculated using E, - E c a l 1 = S log{[cLi(u) yLi(u) + kc:Na C N ~ ( U ) ~ ~ a ( u ) l / [cLi(Call) YLi(Cal1) -k kE,t~a CNa(Ca1 1 ) YNa(Cal I)]> (2) where cy = a , 'a' being the activity of the appropriate ion and y the corresponding activity coefficient. The other symbols have the same significance as in equation (1). The results given in Table 7 show marked improvement when activity coefficient corrections are made for the solutions of higher lithium concentrations (AQ3-AQ7 and PL4-PL7). The behaviour of the di-n-butylamide electrode is significantly improved by using activities [Fig.5(a) and ( b ) ] . The stability of a di-n-butylamide-oNPOE based ionophore used for 4 weeks in serum was monitored for a further 3 weeks with Cal 1 and Cal 2. The results are shown in Fig. 6. The selectivity coefficient was 1.26 x 10-3 for the fresh membrane. After 4 weeks it decreased to 2.0 x 10-3 and after a further 2.5 weeks to 3.96 X 10-3. The fresh ETH 2237-BBPA membrane has a selectivity coefficient of 1.26 x 10-2, much higher than that of the di-n-butylamide sensor. Table 6 Activity coefficients based on the Pitzer equation. Solutions AQI-AQ7 contain 0.140 rnol dm-3 N a f . 0.004 rnol dm-3 K+ and 0.0012 rnol dm-3 Ca7+; Cal 1 and Cal2 contain 0.140 rnol dm-3 Na+ and 0.004 rnol dm-3 K+ [Li]/ Solution rnol dm-3 I YLi YN a AQ 1 AQ2 AQ3 AQ4 A 0 5 AQ6 AQ7 Cal I Cal2 0.0005 0.001 0.0015 0.002 0.003 0.004 0.005 0.001 0.0025 0.1457 0.1462 0.1467 0.1472 0.1482 0.1492 0.1502 0.1 440 0.1455 0.7758 0.7757 0.7756 0.7754 0.7752 0.7749 0.7747 0.7765 0.7761 0.7542 0.7540 0.7537 0.7535 0.7533 0.7529 0.7526 0.775 1 0.7546 Conclusion The best clinically relevant Li ionophore reported, to date, is ETH 1810-oNPOE, with a log ky2y3tNa = -2.45; well below the target value of -4.3 required for less than 1% interference. This ionophore is reported to have a slow response and limited lifetime when used in serum.20.21 The preferred ionophore, ETH 2137-BBPA has log kEtNa = -1.9.The target ionophore with kE:Na = 5.0 X 10-5 has a potential difference ( A E ) of 24.5 mV at 37°C when trans- ferred from Cal 1 to Cal 2.The ETH 2137-BBPA electrode has a AE of 11.9 mV (kstNa = 1.2 x 10-2) corresponding to a correction factor of 12.6' mV whereas the di-n-butylamide- Table 7 Comparison of flame photometry. ETH 2137-BBPA ISE and di-n-butylamide-oNPOE ISE, making activity coefficient corrections ETH 21 37 Di-n-butylamide Flame Solution AQ I AQ2 AQ3 AQ4 AQ5 AQ6 AQ7 c u n PL 1 0.63 PL2 1.04 PL3 1.52 PL4 I .80 PL5 2.54 PL6 3.52 PL7 4.51 c 0.45 0.92 1.33 1.69 2.30 3.45 3.97 Error (YO ) +15.4 + 13.6 -7.6 -2.9 -9.8 - 10.6 - 10.4 c,, Error (YO) 0.69 +32.7 1.14 +8.6 1.66 +9.9 1.97 +7.1 2.77 -1.8 3.85 - 1.5 4.93 -1.8 C 0.38 0.78 1.42 1.75 2.50 3.83 4.41 C" n 0.44 1 .oo 1.40 1.70 2.52 3.53 4.56 c,, 0.48 1.09 1.53 1.85 2.75 3.86 4.99 Error(%) c -2.6 0.39 -3.7 0.81 -1.4 1.44 +0.6 1.74 -2.0 2.55 -0.8 3.86 -0.5 4.43 Error (YO) -7.7 0.52 +3.8 1 .05 -1.3 1.51 +0.5 1.84 -2.5 2.82 -1.3 3.91 -0.6 5.02140 ANALYST, FEBRUARY 1991, VOL.116 oNPOE electrode has a AE of 18.5 mV (kg:Na = 2.4 X 10-3) requiring a correction factor of 6.0 mV. The sensor is stable in serum, has a short response time (about 10-15 s) and a lifetime of at least 50 d. These factors demonstrate that this ionophore is reliable for the assay of Li. We thank SERC for support, Dr. C. T. G. Flear (Royal Victoria Infirmary, Newcastle-upon-Tyne) and Dr. C. J. Fischer (County Hospital, Durham) for useful discussions and Robert Coult for carrying out the atomic absorption measure- ments. References Amdisen, A., in Handbook of Lithium Therapy, ed. Johnson, F. N., MTP Press, Lancaster, 1986, ch.2. Metzger, E . , Dohner, R., and Simon, W., Anal. Chem., 1987, 59. 1600. Oggenfuss, P., Mod, W. E., Oesch, U . , Ammann, D., Pretsch, E., and Simon, W., Anal. Chim. Acta, 1986, 180, 299. Kimura, K., Kitazawa, S . , and Shono, T., Chem. Lett., 1984, 639. Metzger, E., Aeschmann, R., Egli, M., Suter, G., Dohner, R., Ammann, D., Dobber, M., and Simon, W., Helv. Chim. Acta, 1986, 69, 1821. Attiyat, A. S . , Christian, G. D., Xie, R. Y . , Wen, X . , and Bartsch, R. A., Anal. Chem., 1988, 60, 2561. Gadzekpo, V. P. Y., Moody, G. J . , and Thomas, J . D. R., Analyst, 1986, 111, 567. Kimura, K., Yano, H., Kitazawa, S . , and Shono, T., J. Chem. SOC., Perkin Trans. 2, 1986, 1945. Metzger, E., Ammann, D., Schefer, U., Pretsch, E . , Simon, W., Chimia, 1984, 38, 440. 10 11 12 13 14 15 16 17 18 19 20 21 Kataky, R., Nicholson, P. E., and Parker, D., Tetrahedron Lett., 1989, 30,4559. Kataky, R., Nicholson, P. E., and Parkcr, D., J . Chem. SOC., Perkin Trans. 2, 1990, 321. Good, N. E., Winget, G. D., Winter, W., Connolly, T. N., Izawa. A.. and Singh, R. M. M., Biochemistry, 1966. 5 , 467. Covington, A. K., Kelly, P. M., and Maas, A. H. J . , in Methodology and Clinical Applications of ion-Selective Elec- trodes. eds. Maas, A. H. J . , Buckley, B. M., Manzoni, A . , Moran, R. F., Siggaard Andersen, O., and Sprokholt, R., Davies Printing Co., Rochester, 1989, vol. 10, p. 4. Craggs, A., Moody, G. J . , and Thomas, J . D. R., J. Chem. Educ., 1974,51, 541, Horvai, G., Toth, K., and Pungor, E., Anal. Chim. Acta, 1976, 82, 45. Corning 654 (Naf, K f , Lif) Analyser Reference Manual, Corning, 1988. Ammann, D., Anker, P., Metzger, E., Oesch, U., and Simon, W., in Ion Measurements in Physiology and Medicine, eds. Kessler, M., Harrison, D. K., and Hoper, J . , Springer Verlag, Berlin, 1985, p. 102. Czaban, J . D., and Legg, K. D., in Proceedings of the Workshop on Direct Potentiometric Measurements in Blood, ed. Koch, W. F., Gaithersburg, MD, 1983, p. 63. Covington, A. K., and Ferra, M. I. A., Scand. J. Clin. Lab. Invest., 1989, 49, 667. Metzger, E., Dohner, R., Simon. W., Voncherschmitt, D. J., and Gautschi, K., Anal. Chem., 1987, 59, 1600. Gadzekpo, V. P. Y., Hungerford, J . M., Kadry, A. M., Ibrahim, Y. A., Xie, R. Y., and Christian. G. D., Anal. Chem., 1986. 58, 1948. Paper 0l03204B Received July I7th, 1990 Accepted August 20th, 1990

ISSN:0003-2654    

DOI:10.1039/AN9911600135

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1991, VOL. 116 141 Magnesium as a Modifier for the Determination of Barium in Offshore Oil-well Waters by Direct Current Plasma Atomic Emission Spectrometry and Flame Atomic Absorption Spectrometry Mohammad Jerrow and lain Marr Department of Chemistry, University of Aberdeen, Meston Walk, Old Aberdeen, AB9 ZUE, UK Malcolm Cresser Department of Plant and Soil Science, University of Aberdeen, Meston Walk, Old Aberdeen, AB9 ZUE, UK It is shown that the addition of magnesium (5 m g ml-1) to samples for the determination of barium by d.c. plasma atomic emission spectrometry enhances the sensitivity of the analysis and dramatically reduces interference from calcium and strontium at both atomic and ionic emission wavelengths. In the presence of high concentrations of sodium, magnesium is a slightly less effective plasma modifier, but still allows the determination of barium with a precision that is adequate for most practical purposes.Magnesium (in both the presence and absence of sodium) also reduces the interference of alkaline earth concomitants in the determination of barium by atomic absorption spectrometry using a fuel-rich dinitrogen oxide-acetylene flame. Keywords: Barium; oil-well water; direct current plasma atomic emission spectrometry; flame atomic absorption spectrometry; magnesium The determination of barium in offshore oil-well waters is important because of the risk of precipitation of barium sulphate scale as a consequence of mixing with sea-water pumped downhole to maintain hydrostatic pressure and assist oil production.Thus, there is a need for a reliable procedure for the measurement of barium in the mg 1-1 range in the presence of large excesses of sodium and magnesium and variable amounts of calcium and strontium. Although it is widely accepted that the determination of barium by flame atomic absorption spectrometry (FAAS) requires the use of a high temperature, dinitrogen oxide- acetylene flame to minimize interferences and give the required sensitivity, the determination is not straightforward. In the presence of high concentrations of alkali or alkaline earth elements, at least five potential mechanisms exist for interference. Barium has a low ionization potential, and its determination at low concentrations is thus susceptible to ionization interfer- ence from concomitant elements with low ionization poten- tials.Generally, the lower the ionization potential of the concomitant element at any given molar excess, the greater the ionization suppression and enhancement of atomic ab- sorbance (or suppression of ionic absorbance). This effect has been extensively investigated by Maruta et af. ,1 although these workers did not report the tolerance limits for concomitants in the presence of ionization buffers. Capacho-Delgado and Sprague2 and Koirtyohann and Pickett3 investigated the strong molecular absorbance of CaOH in air-acetylene flames at the main barium resonance line, 553.6 nm, and showed that the absorption spectrum of CaOH closely resembled the emission spectrum of barium in this vicinity. In the dinitrogen oxide-acetylene flame, the formation of free atoms is increased, and that of polyatomic oxgyen-containing species reduced, especially in the fuel-rich flame.Thus, for example, using potassium naphthenate as an ionization buffer, Holding and Rowson4 were able to deter- mine barium in used lubricating oils in the presence of a modest excess of calcium, following dilution of samples with 2-methylpropan-2-01 plus toluene (3 + 2) containing small amounts of water. Concentrations of molecular species such as CaOH tend to be greatest at the flame edges.5 The extent of molecular absorption interference therefore depends critically upon flame stoichiometry,4 flame geometry and the optical path of the hollow cathode lamp beam through the flame. It further depends upon the performance of the nebulizer and spray chamber used.Thus, often different results are obtained in spectral interference studies on different instruments, or even on the same instrument, operated under slightly changed conditions. Rooney and Woolley6 suggested that much of the spectral interference attributed to CaOH could be an instrumental artefact, reflecting the inadequate ability of detectors to discriminate between a.c. and d.c. signals when the latter were excessively large. They suggested that the calcium-barium system could be a valuable test of this aspect of instrument performance. However, they did not apparently consider the point outlined above; namely the variation between instru- ments with regard to the flame width effectively used and the flame geometry.This could undoubtedly be reflected in the differences between performance in the series of instruments which they studied. The other potential contributing factors to net interference, when using high salt matrices, are aerosol ionic redistribution, the AIR effect,7 and other transport interferences related to changes in aspiration rate or to primary aerosol generation by the nebulizer. To date, AIR effects have apparently not been studied for ternary cation mixtures, but their possible occur- rence for samples with a concentrated sodium chloride matrix certainly warrants investigation. When Cioni et af. ,8 in 1976, reviewed interference effects in the determination of barium in silicates by AAS, they did not consider all the mechanisms outlined briefly above, but they nevertheless concluded that the separation of barium from concomitant elements was an essential prerequisite to rcliable analysis.Numerous other workers have reached the same conclusion, and separation techniques suggested include ion exchange"10 and co-precipitation. 11 Solvent extraction has not been used much for barium, but its use, prior to furnace AAS determination, has been recommended. 12 The low excitation potential for barium suggests that the determination of the element by atomic emission spec- trometry (AES) should be highly sensitive if a high-tempera- ture excitation source is employed. In practice, a detection limit of 0.3 pg 1-1 has been reported for the analysis of hard water by inductively coupled plasma atomic emission spec- trometry (ICP-AES) after 10-fold evaporative preconcentra- tion, with no spectral interference from up to 200 mg 1-1 of magnesium .I 3 Other inter-element effects were not con-142 ANALYST, FEBRUARY 1991, VOL. 116 sidered. The effects of easily ionizable elements (EIEs) in ICP-AES are more complex than in FAAS. Blades and Horlick14 showed that the EIEs enhanced emission of both atomic and ionic lines in the lower regions of the plasma, but depressed emission for both species in the upper regions. The mechanism in the lower region was attributed to enhanced collisional excitation as a result of an increased number of electrons with sufficient energy to cause excitation of atomic or ionic species. The enhancement increased slightly as the concomitant element ionization potential decreased.A similar conclusion was drawn by Gunter et a1.15 Thompson and Ramseyl6 used a twin-nebulizer technique to demonstrate that calcium matrix effects were a consequence of something happening to the plasma itself, and not to a transport, volatilization or atomization phenomenon. They compared three techniques for minimizing calcium matrix effects, and found interactive matrix matching to be the most useful. Correction by the parameter-related internal standard method (PRISM) was partially successful only for some determinands. They suggested that calcium might prove a useful excitation buffer for other matrix elements. Maessen et al. 17 studied the separate and combined matrix effects of selected alkali and alkaline earth metals on net ICP-AES emission line and background intensities for a series of elements, and showed that composite matrix effects were less than the sums of individual matrix effects.These results show that great care is needed if reliable results for barium determination by ICP-AES in complex and variable alkali and alkaline earth element matrices are to be obtained. Collins18 advocated the use of a d.c. plasma (DCP) as early as 1967 for the determination of five elements, including barium, in oilfield waters. Although sodium, magnesium and calcium all interfered in the determination of barium, he found that, by appropriate, approximate matrix matching using a synthetic brine and adding propionic acid, matrix effects could be kept within acceptable limits (&lo%).Johnson et al.19 showed that phosphate did not interfere significantly in the determination of barium by DCP-AES. 1 1 I I I I I I I 1 , L I C I I t t I t 0 1 2 3 4 5 0 1 2 3 4 5 lnterferent concentrat ion/mg m I - Fig. 1 Effects of increasing concentrations of 0, Ca; 0, Sr; and A, Mg on the determination of barium (emission from 1 pg ml-I) at ( a ) ( c ) and ( e ) 553.6 nm and ( b ) , (d) and cf) 455.4 nm in the absence of other cations [ ( a ) and (b)] and in the presence of 3 g 1-1 Na [ ( c ) and (d)] or 5 g 1 Mg [ ( e ) and cf)] The above consideration suggested that further investiga- tion of the potential of DCP-AES for the determination of barium in oilfield waters should prove worthwhile, in order to see if a suitable excitation buffer could be found to allow simple, rapid and reliable determination.Experimental Apparatus Atomic and ionic absorption measurements were made at 553.6 and 455.4 nm, respegtively, by using a Philips SP9 atomic absorption spectrometer with a fuel-rich dinitrogen oxide-acetylene flame at a spectral bandwidth of 0.2 nm. Emission measurements for the same transitions were made using a Spectraspan I11 d.c. plasma echelle optical emission spectrometer, with 200 nm high, 100 nm wide slits. Reagents Stock solutions of sodium, magnesium, strontium and barium were made from analytical-reagent grade salts and a calcium stock solution from calcium carbonate, following dissolution in the minimum possible volume of dilute hydrochloric acid. The stock solutions, other than that for barium, were analysed by AES by scanning through the emission from the DCP, to check that they did not contain significant amounts of barium.Results and Discussion The graphs in Fig. l(a) and (6) show the effects of increasing amounts of calcium, strontium and magnesium upon the apparent barium emission signals when solutions containing these cations plus 1 pg ml-1 of barium were nebulized into the DCP. The enhancements in emission signal are broadly comparable at both the atomic line [Fig l ( a ) , (c) and ( e ) ] and the ionic line [Fig. l(b), (d), and 01. This confirms that the effect is not significantly attributable to the suppression of the ionization by the concomitant elements. Thus, the DCP in this respect behaves much as the ICP, as discussed earlier. The atomic and ionic absorption results, on the other hand [Fig.2(a)] show that as the atomic absorption signal increased with increasing concomitant concentration and increasing ioniza- a, c m 0 I I I 0 1 2 3 4 5 lnterferent concentration/mg ml-I Fig. 2 Effects of 0, Ca; a. Sr; and A, Mg on the atomic (solid line) and ionic (broken line) absorbance from barium at 1 pg ml-1, ( a ) in the absence of other cations and (b) in the presence of Na at 3 g I-*ANALYST, FEBRUARY 1991. VOL. 116 L 143 tion suppression, the ionic absorption signal decreases. The decrease is not pro ram, however, because of the spectral interference from the CaOH species. Thus, while the order of suppression of ionization is Sr > Ca > Mg, as would be expected from the respective ionization potentials, the order of apparent atomic absorption enhancement is Ca > Sr > Mg.The close relationship between the graphs in Fig. l(a), ( h ) , ( c ) and (d) suggests in fact that ionization suppression by concomitant elements is of minor importance in the DCP, and that the effects of calcium, strontium and magnesium are probably via changes induced in the plasma and enhanced collisional excitation. The effects of magnesium and strontium are similar, but calcium has a more substantial impact, which is not the trend that would be expected for an ionization suppression effect. Another possible contributing factor could be ionic redistribu- tion which sometimes occurs upon aerosol generation, the AIR effect. This seems unlikely to be a major factor here, however, because hitherto AIR effects have only been observed at higher alkali metal concentrations, and because of the disparate behaviour in DCP-AES and FAAS.In FAAS, the presence of an excess of sodium, as expected, largely suppresses the ionization interference of magnesium, calcium and strontium on the determination of barium, although the calcium spectral interference from CaOH absorbance is still discernible [Fig. 2(6)]. Background compensation is clearly essential if FAAS is to be used for the determination of barium in the presence of variable amounts of calcium on the instrument used here. However, the output from continuum sources at this wavelength is so low on some instruments that the background correction is inoperable. Sodium at a concentration of 3 g 1-1 also substantially reduces the extent of interference from magnesium, calcium and strontium in the determination of barium by DCP-AES [Fig.l(c) and ( d ) ] . At the wavelength of the barium atomic emission, 553.6 nm, magnesium still has a significant effect, and barium could not be reliably determined by DCP-AES if only the sodium concentrations were matrix matched, and magnesium concen- tration was variable between samples. Thus, although an excess of sodium apparently stabilizes the excitation condi- tions, the effect is inadequate for reliable routine work. The results given above suggest that magnesium has a greater effect on plasma excitation conditions than does sodium, which in turn suggests that magnesium might be a more useful plasma modifier for the DCP than sodium.The graphs in Fig. l(e) and U, confirm that magnesium may indeed be used to prevent interference from calcium and strontium in barium determinations. At the ion line, however, calcium still causes a small enhancement in barium emission [Fig. 1 0 1 . As most samples of interest are likely to contain sodium at a fairly high concentration, the effect of magnesium ( 5 g 1- l ) was also studied in the presence of sodium (3 g 1-1). The results (Fig. 3) show that the mixture is less efficient than magnesium alone in stabilizing the plasma against the effects of calcium and strontium. However, they confirm that a concentration of up to 2 g 1-1 of calcium or strontium can be tolerated, and that if samples contain a few hundred or a few thousand mg I-' of calcium and/or strontium, the subsequent analysis will be more accurate if a sodium plus magnesium modifier is used than if sodium alone is used.If magnesium is added as a modifier in FAAS, the effects of calcium and strontium on barium absorbance are much reduced (compare Fig. 4 with Fig. 2). Surprisingly, in the presence of 5 g 1-1 of magnesium, low concentrations ( 4 0 0 mg 1 - 1 ) of calcium and strontium depress the atomic absorp- tion signal of barium [Fig. 4(a)], an effect not echoed in the ionic absorbance graphs. This is possibly an incomplete volatilization interference, observed because it is low in the red feather zone. In the presence of sodium (3 g 1 - 1 ) plus magnesium ( 5 g 1 - I ) , high concentrations of calcium and strontium both significantly enhance the barium atomic absorbance [Fig.4(b)]. In order to confirm that the beneficial effects observed for magnesium in DCP-AES were not confined to the specific set of operational parameters used, observations were made at different heights in the plasma analytical zone and at different aspiration rates. Fig. 5(a) shows that increasing the aspiration rate from 1.45 to 2.60 ml min-1 did not adversely influence the enhancement of barium emission caused by magnesium. It is notable that this increase in aspiration rate had a very small effect upon the size of the barium signal, almost certainly I I I I I I 0 1 2 3 4 5 lnterferent concentration/rng rnl-' Effects of increasing concentrations of 0. Ca; and a, Sr on the emission from barium at 1 yg ml-' at ( a ) 553.6 nm and (h) 455.4 nm, in the presence of Mg at 5 g 1 - 1 plus Na at 3 g 1-' 1 I I I I 0 1 2 3 4 5 lnterferent concentration/rng ml-1 Fig. 4 Effects of increasing concentrations of 0, Ca; and a, Sr on the atomic (solid line) and ionic (broken line) absorbance from barium at 1 pg ml-1 in the presence of (a) Mg at 5 g 1-1 and (b) Mg at 5 g I-' plus Na at 3 g 1-1144 w ANALYST, FEBRUARY 1991, VOL.116 (6) 0 0.3 0.6 0.9 1.2 1.5 Calcium/mg mi-' Fig. 5 Effects of aspiration rate on interferences in the signals from barium at 1 pg ml-1; solid lines = 1.45 ml min-1, broken lines = 2.60 ml min-1. ( a ) Effect of increasing amounts of Mg; (b) effects of increasing amounts of Ca 0, in the absence; and H, presence of Mg at 5 g I-' because of the less favourable aerosol size distribution produced.Magnesium was effective in dramatically reducing interference from increasing amounts of calcium at both aspiration rates [Fig. 5(b)]. It is interesting to note that low concentrations of calcium depressed the barium emission at 455.4 nm at the higher aspiration rate, but not at the lower rate, even when magnesium was present. This, plus the relative shapes of the graphs in Fig. 5(a), indicate that the effect of magnesium as a plasma modifier or excitation buffer is greater when the water loading to the plasma (or posssibly the aerosol size distribution reaching the plasma) is lower. Changing the height of observation in the plasma had very little effect upon the benefits of using magnesium as an excitation buffer. It is a common misconception that ionization buffers in FAAS totally suppress the ionization of the determinand, but this is not always so.Fig. 2, for example, shows that 3 g 1-1 of sodium is approximately as effective as 3 g 1-1 of strontium. Adding calcium or strontium at high concentrations to solutions containing sodium and magnesium may further enhance atomic absorbance signals. Great care is therefore needed when selecting ionization buffers for use in FAAS for samples such as oilfield brines. Conclusions Magnesium is an excellent excitation buffer for the determina- tion of low concentrations of barium at 553.6 nm by DCP-AES, eliminating the effects of up to 4 g 1-1 of both calcium and strontium. However, tolerance of these elements is reduced to about 2 g 1-1 in the presence of sodium at a concentration of 3 g 1-1, and, even at only 2 g 1-1 of a concomitant, a 5% change in signal is still observed.Clearly, therefore, great care is needed when determining barium by DCP-AES to ensure that whatever excitation buffer is used can cope with the conceivable range of concomitant elements present. Magnesium has the further advantage that it is less expensive than the more commonly used high-purity lithium salts for DCP-AES. Care is also needed when determining barium in offshore oilfield waters by flame AAS. It should not be presupposed that the sodium will be totally adequate as an ionization buffer if other easily ionized concomitants are present. The presence of magnesium may drastically alter the effects of calcium and strontium on the determination of barium by FAAS, and it appears that incorporation of both sodium and magnesium into standards is required, as well as background compensa- tion for CaOH absorbance.M. Jerrow is grateful to the Iraq Ministry of Higher Education for financial support for this work. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 References Maruta. T., Takeuchi, T.. and Suzuki, M., Anal. Chim. Acta, 1972, 58, 452. Capacho-Delgado, L., and Sprague, S.. At. Absorpt. Newsl., 1965,4, 363. Koirtyohann, S. R., and Pickett, E. E., Anal. Chem., 1966,38, 585. Holding, S. T., and Rowson, J. J., Analyst, 1975, 100, 465. Cresser, M. S., Keliher, P. N., and Kirkbright, G. F., Selected Ann. Rev. Anal. Sci., 1973, 3, 139. Rooney, R. C., and Woolley, J . F., Analyst, 1978, 103, 1100. Borowiec, J., Boorn, A. W.. Dillard, J . H., Cresser, M. S., Browner, R. F., and Matteson, M. J., Anal. Chem., 1980, 52, 1059. Cioni, R., Mazzucotelli, A., and Ottonello, G., Analyst, 1976, 101, 956. Sixta, V., Miksovsky, M., and Sulek, Z., Fresenius Z. Anal. Chem., 1975, 273, 193. Strasheim, A., Strelow, F. W. A., and Norval, E., J . Chem. South Afr. Chem. Inst., 1967, 20, 25. Bano. F. J., Analyst, 1973, 98, 655. Sugiyama, M., Fujino, 0.. and Matsui, M., Bunseki Kagaku, 1984, 33, E123. Thompson, M.. Ramsey, M. H.. and Pahlavanpour, B., Analyst. 1982, 107, 1330. Blades, M. W., and Horlick, G.. Spectrochim. Acta, Part B , 1981,36, 881. Gunter, W., Visser, K., and Zeeman, P. B., Spectrochim. Acta, Part B , 1985, 40, 617. Thompson, M., and Ramsey, M. H . , Analyst, 1985. 110, 1413. Maessen, F. J . M. J., Balke, J., and de Boer, J . L. M., Spectrochim. Acta, Part B , 1982,37, 517. Collins, A. G., Appl. Spectrosc., 1967, 21, 16. Johnson, G. W., Taylor, H. E., and Skogerboe. R. K., Anal. Chem.. 1979.51, 2403. Paper Of031 17H Received July l l t h , 1990 Accepted October 15th, 1990

ISSN:0003-2654    

DOI:10.1039/AN9911600141

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST. FEBRUARY 1991, VOL. 116 14.5 Proposed Mechanism for the Action of Palladium and Nickel Modifiers in Electrothermal Atomic Absorption Spectrometry Anatoly Volynsky* V.I. Vernadsky Institute Qf Geochemistry and Analytical Chemistry, USSR Academ y of Sciences, 79 Kosygin Street, I 7 7975 Moscow, USSR Sergei Tikhomirov and Anatoly Elagin All-Union Research Institute of Organic Synthesis, 72 Radio Street, 107005 Moscow, USSR By the use of Fourier transform infrared spectrometry it was found that palladium chloride decreased the temperature of the reduction of PbO and Ga203 with graphite; nickel chloride only catalysed the reduction of Ga203. A hypothesis is proposed that nickel and palladium salts promote the reduction of compounds (in atomic absorption measurements) at relatively low ashing temperatures.The resultant free elements form intermetallic compounds or solid solutions with metallic Pd and Ni, thus reducing or almost eliminating loss of analyte due to sublimation of halides, oxides, dimers and other compounds. The high efficiency and universal action of Pd modifiers are because the Pd metal can be easily formed from its compounds and also by the unique catalytic properties of metallic Pd. Keywords: Electrothermal atomic absorption spectrometry; palladium modifier; nickel modifier; mechanism of action; Fourier transform infrared spectrometry The compounds of nickel' and palladium and platinum2 were introduced as modifiers in 197.5 and 1979, respectively. At present, nickel compounds, platinum group metals (PGMs) and various different mixtures based on these are widely used for the determination of a great number of elements in various Palladium and nickel are known to form inter- metallic compounds and solid solutions with the determined elements in the graphite furnaces.G9 The formation of such compounds causes an increase in the maximum permissible ashing temperature during the determination of elements of high- and mid-volatility.However, the mechanism of forma- tion of the intermetallic compounds and solid solutions in graphite atomizers, when using such modifiers, is still rather vague. I t is clear that Pd interacts in its metallic form. According to Morikawa rt al. , I 0 activated carbon partially reduces palladium chloride to the metal at room temperature. However, virtually all other elements detectable by elec- trothermal atomic absorption spectrometry (ETAAS) (with the exception of the noble metals), exist in the graphite atomizer in the form of oxides, chlorides" or other salts, within the temperature range 300-800"C.It is still not clear why Pd is the most efficient modifier in the majority of instances .s. 12 The supposition has been made,".l3 that Ni and PGM modifiers catalyse some processes that occur in graphite atomizers for ETAAS. It might be that in the first stage of the process of the formation of solid solutions and intermetallic compounds, the graphite of the atomizer catalytically reduces the oxide analytes at low temperatures.14 Nickel and PGMs Table 1 Initial (T,,,) and final ( Tf) temperatures of the reduction of Pb and Ga oxides with graphite PbO Gal03 Experimental conditions T,,I"C T,I"C T,,l0C T+l°C Without catalyst 430 740 735 990 present 520 780 600 840 NiCI2.6H20 PdClz present 340 500 360 720 * Present address: Laboratory of Organic Analysis. Department of Chemistry.Moscow State University, 119899 Moscow, USSR. are known as efficient catalysts for the reduction of the oxides of Mo, V, Cu, Sn, Re, Pb, W, Fe and Ni (other oxides have not yet been studied) with hydrogen, carbon monoxide and some hydrocarbons. By the use of X-ray photoelectron spectrometry,17 it has been found that lead chloride is thermostable in a graphite atomizer at up to 600°C; at higher temperatures it sublimes, whereas, in the presence of palladium chloride, metallic Pb is already apparent at 200°C.This paper describes the investiga- tion of the reduction of Pb and Ga oxides with graphite in the temperature range 100-1000 "C using Fourier transform infrared (FTIR) spectrometry. Experimental Apparatus For the identification of the gaseous products of the reactions, a Bruker FTIR spectrometer, Model IFS-113~. with a gas chromatographic interface was used. A JEOL pyrolyser, Model PL-722, was connected to the interface and the carrier gas flow (Ar, 30 ml min - 1 ) was controlled by a Carlo Erba Fractovap 4200 chromatographic block. The carrier gas was additionally purified, for traces of water vapour and free oxygen, with the aid of a Supelco high capacity gas purifier. The temperature of the pyrolytic oven was controlled by a thermocouple and registered on an analogue recorder with a 10 mV scale.Procedure Electrographite was pounded in a vibration mill. Just before the experiment, the graphite powder was heated in an Ar atmosphere for 40 min at 800°C to remove adsorbed gases. The pre-heated graphite powder was mixed with the oxides (PbO, 3.5 mg; Ga203, 1.6 mg) with a mass ratio of approximately 5 : 1. The catalyst (PdC12 or NiC12-6H20) was added to the reaction mixture in an atom ratio of carbon to metal of approximately 2 5 : 1 . The maximum mass of the reaction mixture (about 25 mg) was limited by the size of the quartz crucible. The oxide mass in the mixture resulted in absorbance values of up to 0.3 during the measurements. The identification of the gaseous products of the reaction was performed within the following spectral windows: 2200-2 100 cm- 1 for carbon monoxide; 2380-2300 cm- 1 for carbon146 ANALYST, FEBRUARY 1991, VOL.116 dioxide; 1700-1500 cm-1 for water; and 1860-1800 cm-1 for phosgene. The sensitivity for carbon dioxide is about 1.5-fold higher than that for carbon monoxide. Results As can be seen from Table 1, the reduction of lead monoxide in the absence of a catalyst starts at 430°C. Below this temperature carbon dioxide is formed [Fig. 1(a)] due to the decomposition of the traces of PbC03 (tdec = 315"C).18 The preliminary heating of PbO (10 min at 400°C in an Ar atmosphere) , although significantly reducing the area of this peak, fails to eliminate it completely. The interaction of Ga203 with graphite starts when the temperature is raised to 735 "C (Table 1).The erratic baseline shown in Fig. l(b) is due to the temperature fluctuation of the light-pipe.19 The effect appears to be pronounced in this particular instance because of the relatively high expansion of the ordinate axis. The addition of nickel chloride substantially decreases the temperature of the reduction of Ga203 with graphite (Table 1) and changes the mechanism of the process towards the formation of carbon dioxide (Fig. 2). Nickel chloride does not produce a catalytic effect on the reduction of lead monoxide with graphite (Table 1). No traces of phosgene or water vapour are registered within the temperature interval exam- ined (Fig. 3). Obviously, chlorine from nickel chloride evolves in the form of HC1,2" but absorption bands due to HCI (3000-2800 cm-1) lie beyond the limits of the interval examined.Water contained in NiCI2.6H20 apparently ad- sorbed on to the graphite starts to interact (at 80OoC) forming carbon dioxide (Fig. 2). Palladium chloride sharply decreases the temperature of the reduction of the metal oxides with the graphite (Table 1). It can be seen that the C02 evolution is entirely complete in 3 min [Fig. 4(a)]. Carbon monoxide formation during the reduction of Ga203 becomes negligible [Fig. 4(b)]. A special 'blank' experiment was performed by heating the powdered - 40 4- .- 8 3 ' 30 h 2 ; 20 2 2 10 a 4- .- 0 m 800 0 600 2 I a 400 $ F % 200 0 6 12 18 24 Time/mi n 0 8 16 24 32 Ti rne/m i n graphite with palladium chloride in the absence of both lead and gallium oxide and neither CO2 nor CO formation was detected.Discussion The results obtained corroborate the theory of the catalytic action of some modifiers on the reduction of the oxides of certain metals by graphite. It is necessary to note that these results have been obtained using electrographite powder as the reductant. Pyrolytic graphite is a much more inert material than electrographite, therefore, the reduction with pyrolytic graphite proceeds significantly slower. Hence, in the most popular graphite tubes with a pyrolytic graphite coating the catalytic effect should be more discernible. The direct experimental verification of this proposition is impossible by the procedure used. The main peculiarities of the pyrolytic graphite are the high degree of crystalline order and low concentration of active sites on its surface.21 After severe pounding these peculiarities are lost.Another difference of our experimental conditions from those used for ETAAS is the amount of lead and gallium oxides used, i.e. , mg instead of the pg and ng levels which are typical for ETAAS. The chemical and physical properties of clusters usually differ significantly from the properties of the bulk materials. This is a typical problem in the study of the processes that occur in the graphite tubes used in ETAAS, excluding the instances of usage of mass spectrometry or radioactive isotopic analysis. Thus, the temperatures of the reduction in the graphite tubes might differ from those obtained in our experiments. We considered the principal processes occurring in graphite atomizers in the presence of the PGM modifiers with palladium chloride as an example.When a reductant (e.g., the graphite of the atomizer) is present, palladium chloride 800 u B 600 2 h 4- 400 al I- 200 I I 1 I 8 16 24 32 Interaction of graphite with Ga203 in the presence of Timehi n Fig. 2 NiC12.6H20. 1, CO2: and 2, CO 0.12 800 0 B 2 20.08 2 600 2 Q) a m 400; 2 200 <0.04 0 2300 21 00 1900 1700 Wavenurn berlcrn-' Fig. 1 Interaction of graphite with ( a ) PbO and ( b ) GazO3. 1, CO,; and 2, CO Fig. 3 Ty ical IR spectrum of a mixture of Ga203, graphite and NiC12-6H28; T = 680 "CANALYST, FEBRUARY 1991. VOL. 116 147 Table 2 Thermal properties of the most thermostable chlorides and oxides of the PGMs and nickells v) C c .- al 0 L m 2 0 2o 'E z 0 6 12 18 24 Ti me/mi n A v) C 3 w .- 2 45 E a - k 30 c .- a, C m 15 5: 0 s 800 0 % 600 2 E 400 2 al I- 200 800 600 $ E 400 9 .: I- 200 8 16 24 32 Tim e/m i n Fig. 4 presence of PdCI,. 1, CO?; and 2, CO Interaction of graphite with ( a ) PbO and ( b ) Ga203 in the decomposes at the relatively low temperatures used during the ashing stage. The metallic Pd produced catalyses the reduction of the lead and gallium oxides and the lead chloride17 with the graphite. The reduced metals dissolve in the Pd forming intermetallic compounds7-9 or solid solutions.6 Such a process is promoted in the graphite tube due to the amount of the modifier being 100-1000-fold higher than the amount of analyte. The increase in the sensitivity of the determination in the presence of such modifiers is caused by the formation of compounds of low volatility at the relatively low ashing temperatures.This formation decreases or eliminates any losses of analyte due to the sublimation of volatile chlorides, 1722 oxides,7,'3 dimers,7 hydrides'4 and other com- pounds. Sometimes the catalyst not only decreases the temperature of the reduction for the compounds being determined, but also simultaneously changes the mechanism of the reduction. Without the catalyst, the products of Ga203 reduction are CO [Fig. l(h)] and possibly Ga202s or Ga0.26 In the presence of NiCI? and especially PdCI2 the main gaseous product of the reaction is C 0 2 [Figs. 2 and 4(b)]. The marked increase in the maximum ashing temperature and sensitivity for gallium26 suggests that in the presence of Pd and Ni modifiers, Ga203 is reduced to non-volatile products, e.g., to the free metal.This supposition is verified by the absence of GaO in the electrothermal atomizer gas phase at the ashing stage in the presence of Ni compounds.26 Nickel forms compounds that are more thermostable than those of the PGMs (Table 2). The lack of elemental Ni in the reaction mixture at 300400°C is probably the main reason for the absence of the catalytic effect of nickel chloride on the reduction of PbO. According to thermodynamic calculations, nickel chloride is stable in an argon atmosphere in the graphite atomizer in the temperature range 230-630"C.24 Free hydrogen can reduce it to the metal at these temperatures; however, below 1000°C the reaction between graphite and water, which produces free hydrogen, proceeds slowly with- out a catalyst.Nickel in the form of its compounds does not act as a catalyst for this p r o ~ e s s . ' ~ At 600°C thermohydrolysis results in transformation of the nickel chloride into the oxide .?O Chloride TI"C Oxide TPC PdCI? dec. * 500 PdO m.p. 870 PtCl, dec. 581 PtO dec. 550 dec. 1100 RuCl3 dec. >SO0 IrOz IrC12 dec. 773 Rhz03 dec. 1100-1150 m.p. 1984 RhCl3 dec. 450-500 NiO NiC12 subl. t 973-987 * Dec. = decomposes. ? Subl. = sublimes. Data from the literature, on the conditions for metallic Ni formation from the nitrate and oxide in the graphite atomizer, seem to be contradictory. According to thermodynamic calculations, nickel oxide is stable in the atomizer at tempera- tures between 230 and 630°C under a partial pressure of free oxygen of 1 x 10-6 bar.24 By the use of mass spectrometry it has been shown that metallic Ni is formed in the graphite atomizer in substantial amounts only at about 15OO0C.7.23 It has been found by thermal and X-ray diffraction analysis that graphite reduces Ni(N03)2.6H20 to Ni metal in an argon atmosphere at 1 15OoC.'8 Catalysts significantly decrease the reduction temperature (down to 630°C for CUCI).'~ At the same time less than 1% of the nickel nitrate is reduced to nickel dicarbide at 330 "C.23 Nickel nitrate is partially reduced to the free metal at relatively low ashing temperatures; thus it is necessary to introduce 1000 times as much Ni into the graphite atomizer as Pd in order to create a concentration sufficient for the effective catalysis of the reduction of the analyte.12 In our experiments it would appear that significant amounts of catalytically effective metallic Ni were formed from NiCl2.6H20 at temperatures between 520 and 600 "C (Table 1). Nickel metal catalyses the reduction of Ga203 with graphite to free Ga, and the oxidation of graphite with water vapour. It is known that graphite readily reacts with water at 800°C in the presence of metallic Ni.27 Nickel metal and nickel oxide are also efficient catalysts for the oxidation of carbon monoxide to carbon dioxide; hence, carbon dioxide is the final product of graphite oxidation (Fig. 2). Mixed oxide formation is a possible mechanism for the stabilization of highly volatile elements in the presence of nickel nitrate,.at ashing temperatures below 600°C. Such compounds are known for Ga, TI, Ge, As, Sb, Bi, Se and Te; ZnO and NiO form solid solutions.30 At elevated tempera- tures these compounds can be reduced to the corresponding intermetallic compounds. Evidently such a mechanism for stabilization is marginally possible using nickel chloride as the modifier. The high efficiency and universality of Pd modifiers can be explained not only by the ease of formation of Pd metal from its compounds but also by the unique catalytic properties of Pd. In particular the catalytic properties of Pd are only slightly dependent on the particle size, because Pd electron configura- tion depends weakly on cluster size.3' The electron configura- tion of Pd2 is 4d95s'. This configuration approaches 4d9.45~0." at eight to ten atoms where only minor changes take place with increasing particle size.The configuration determined from experimental measurements on bulk Pd is 4d9.65s0.4. Sample matrix and the parameters of determination (especially the temperature and heating rate of the atomizer during the drying stage) can significantly affect the size and the structure of the catalytically effective particles, thus restricting the field of efficient use of the PGM modifiers (except Pd). Excellent results have been reported by Dahl et a1.32 on the use of a mixture of Pd, Rh and Ir compounds for Sb determination in different matrices. They can probably be explained by the successful combination of the unique catalytic properties of Pd with the high melting points of Rh148 (m.p.= 1960°C) and Ir (m.p. = 2450°C). This prevents the loss of Sb and enables ashing at extremely high temperatures. The authors thank E. M. Sedykh and G. N. Takhtarova for their valuable assistance. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 References Ediger, R. D., At. Absorpt. Newsl., 1975, 14, 127. Shan, X.-q., and Ni, Z.-m., Huaxue Xuebao, 1979, 37, 261; Chem. Abstr., 1980, 92, 220474~. Voth-Beach, L. M., and Shrader, D. E., J. Anal. At. Spectrom., 1987, 2, 45. Schlemmer, G., and Welz, B., Spectrochim. Acta, Part B , 1986, 41, 1157. Ni, Z.-m. , and Shan, X.-q., Spectrochim. Acta, Part B, 1987,42, 937. Shan, X.-q., and Wang, D.-x., Anal. Chim. Acta, 1985, 173, 315. Styris, D. L., Fresenius Z. Anal. Chem., 1986, 323, 710.Teague-Nishimura, J. E., Tominaga, T., Katsura, T., and Matsumoto, K., Anal. Chem., 1987, 59, 1647. Wendl, W., and Miiller-Vogt, G., J. Anal. At. Spectrom., 1988, 3, 63. Morikawa, K.. Shirasaki, T., and Okada, M., in Advances in Catalysis and Related Subjects, ed. Eley, D. D . , Academic Press, New York, 1969, vol. 20, p. 98. Frech, W., Lundberg, E., and Cedergren, A., Prog. Anal. At. Spectrosc., 1985, 8, 257. Brzezinska-Paudyn, A , , and Van Loon, J. C., Fresenius Z . Anal. Chem., 1988, 331, 707. Rettberg, T. M., and Beach, L. M., J. Anal. At. Spectrom., 1989, 4, 427. Volynsky, A. B., XXVI Colloquium Spectroscopicurn Interna- tionale, Sofia, 1989, Abstracts Volume I, p. 95. Il’chenko, N. I., Usp. Khim., 1972, 41, 84. Charcosset, H., and Delmon, B., Ind. Chim.Belge., 1973, 38, 481. Sakurada, O., Takahashi, H., and Taga, M., Bunseki Kagaku, 1989, 38, 407. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 ANALYST, FEBRUARY 1991, VOL. 116 CRC Handbook of Chemistry and Physics, ed. Weast, R. C., CRC Press, Boca Raton, 68th edn., 1987. White, R., ChromatographyIFourier Transform Infrared Spec- troscopy and its Applications, Marcel Dekker, New York, 1990, pp. 57 and 58. Welz, B., Akman, S . , and Schlemmer, G., Analyst, 1985, 110, 459. Huettner, W., and Busche, C., Fresenius Z . Anal. Chem., 1986. 323, 674. Sedykh, E . M., and Belyaev, Yu. I., Prog. Anal. At. Spectrosc., 1984, 7, 373. Droessler, M. S . , and Holcombe, J. A., Spectrochim. Acta, Part B, 1987, 42, 981. Dedina, J., Frech, W., Cedergren, A.. Lindberg, I., and Lundberg, E., J. Anal. At. Spectrom., 1987, 2, 435. McAllister, T., XXVI Colloquium Spectroscopicum Interna- tionale, Sofia, 1989, Abstracts Volume IV, p. 55. Shan, X.-q., Yuan, Z.-n., and Ni, Z.-m., Anal. Chem., 1985, 57, 857. McKee, D. W., in Chemistry and Physics of Carbon, eds. Walker, P. L., Jr., and Thrower, P. A., Marcel Dekker, New York, 1981, vol. 16, p. 1. Richardson, R. T., and Rowston, W. B., Proceedings of the Second European Symposium on Thermal Analysis, Aberdeen, 1981, p. 355. Pushkarev, V. A., in Physical Chemistry of Oxides, ed. Men, A. N., Nauka, Moscow, 1971, p. 87 (in Russian). Landolt-Bornstein. Zahlenwerte und Funktionen aus Natur- wissenschaften und Technik, Neue Serie, ed. Hellwege, K.-H., Springer-Verlag, Berlin, 1975, Gesamtherausgabe, Gruppe 111, Band 7, Teil a-f. Hamilton, J. F., and Baetzold, R. C., Science, 1979, 205, 1213. Dahl, K., Martinsen, I., Salbu, B., Radziuk, B., and Thomas- sen, Y., XXVI Colloquium Spectroscopicum Internationale, Sofia, 1989, Abstracts Volume I, p. 91. Paper 0101 009J Received March 6th, 1990 Accepted July 23rd, 1990

ISSN:0003-2654    

DOI:10.1039/AN9911600145

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1991, VOL. 116 149 Determination of Nickel Tetracarbonyl by Gas Chromatography Alexander Harper AEA Technology, Harwell Laboratory, Oxfordshire OX7 7 ORA, UK Capillary gas chromatography, with use of an electron-capture detector, has been assessed as a detection technique for nickel tetracarbonyl in gas samples. Detection limits as low as 1 part in 1011 can be achieved under the appropriate conditions. Keywords: Gas chroma tograph y; electron-capture detector; nickel tetracarbon yl Nickel tetracarbonyl is a volatile, thermally unstable material, which may be formed by the direct reaction of CO with nickel or nickel-containing alloys. Interest in the analysis of gases for Ni(C0)4 at low concentrations arises both from the toxicity of the material and from its role in the transport of catalytically active nickel in gas circuits. The carcinogenic behaviour associated with the chronic inhalation of Ni(C0)4 is reflected in the exposure guidelines recommended for this material, 100 vppb (parts per lo9 by volume) short term exposure limit (STEL) (10 min).Even more stringent controls have been applied in the past; the American Conference of Industrial Hygienists set a value of 1 vppb in 1959, although this was subsequently increased to 0.05 mg m-3 of nickel [equivalent to 20 vppb of Ni(C0)4].2 These low values clearly point to the need for a reliable method for the determination of Ni(C0)4 vapour at vppb levels for environmental monitoring. The transport of catalytic nickel in gas circuits involves the formation of Ni(C0)4 in relatively cool portions of the system and its subsequent thermal decomposition in regions of the circuit at higher temperatures.The catalytic nickel thereby formed can lead to the decomposition of hydrocarbon materials with the resulting formation of carbonaceous deposits in undesirable places. For example, in the primary cooling circuits of Advanced Gas-Cooled Reactors, where CO is maintained in the C 0 2 coolant to inhibit corrosion of the graphite moderator, the phenomenon could lead to the formation of layers of material of relatively low thermal conductivity on heat-transfer surfaces. In systems such as this, where the rate of gas flow is high, even trace concentrations of Ni(C0)4 in the gas phase can result in the transport of significant amounts of catalyst. Analysis of gases for Ni(C0)4 has been accomplished by two techniques: adsorption of the component of interest on to a substrate and subsequent assay of the substrate for nickel,3 and by exploitation of the chemiluminescent reaction between Ni(C0)4 and ozone .4-7 Adsorption techniques possess the advantage that they mag be calibrated without resource to gas standards containing Ni(C0)4. Determinations by such methods are, however, time consuming and require large samples of gas, especially at low concentrations.Further, the ultimate sensitivity is limited by the inevitable nickel contami- nation of suitable adsorbents. Chemiluminescence possesses the merit of sensitivity, but commercially available equipment is expensive, and the method can be subject to interference from the chemiluminescent reactions of other components of the sample matrix.6 Gas chromatography, coupled with a suitable detection system, offers the potential for good selectivity and sensitivity combined with a modest sample volume and analysis time.This paper describes the evaluation of gas chromatography, with electron-capture detection, for the determination of Ni(C0)4 in CO-C02 gas mixtures. Experimental Gas Chromatograph The gas chromatograph used in the present study was a Hewlett-Packard Model 5890, fitted with a six-port gas- sampling valve and a constant-current electron-capture detec- tor (ECD). A facility was available-to allow operation of the oven at sub-ambient temperatures. Data acquisition, storage and analysis were effected by using a Hewlett-Packard Model 59970 workstation.All the experiments described were performed with the use of a 10 m x 0.53 mm capillary column, coated with a 2.65 pm film of 5% diphenyl-95% dimethyl polysiloxane gum (Hew- lett-Packard). The use of a capillary column restricts carrier gas flow-rates to values too low to allow operation of the ECD on carrier gas alone. A subsidiary gas feed (make-up gas) was therefore supplied to the detector. This has the advantage that carrier gas composition is not limited to those gases suitable for ECD operation. The analytical conditions finally adopted for the assessment of sensitivity are summarized in Table 1. Materials Nitrogen (high purity, oxygen free) was obtained from Air Products. Carbon monoxide (99.5%) was obtained from the same source and passed over active charcoal to remove traces of Ni(C0)4 and Fe(CO)5 before use. Gaseous C 0 2 was obtained from converters charged with solid carbon dioxide (Distillers).Nickel tetracarbonyl (>97%) was obtained in liquid form from Pfaltz and Bauer, and was used without further purification. Table 1 Summary of chromatographic conditions Non-isothermal 10 m x 0.53 mm coated 10 m x 0.53 mm coated with 2.65 ym film of 5% with 2.65 p n film of 5% diphenyl-95% dimethyl diphenyl-95% dimethyl polysiloxane gum polysiloxane gum Parameter Isothermal Sample volume 0.25 ml 3.0 ml Column Injector temperature 30 "C 30 "C Column temperature 30 "C -30 "C for 1 min then ramp at 40 "C min- I to 10 "C Make-up gas Nitrogen Nitrogen Make-up gas flow-rate 60 ml min- 1 Carrier gas flow-rate 2 ml min-1 2 ml min-1 Detector temperature 60 "C 60 "C Carrier gas co co 60 ml min- 1150 ANALYST, FEBRUARY 1991, VOL.116 Standard Gases Stock mixtures of Ni(C0)4 in CO-C02 were prepared by transferring between 0.8 and 1.0 kPa of Ni(C0)4 vapour into an evacuated aluminium cylinder. The cylinder was then pressurized, with as little delay as possible, to 4 MPa with a gas mixture of 2% v/v CO in C02. Carbon monoxide was maintained in the mixture to suppress the decomposition of Ni(C0)4 to metallic nickel and CO. Although this mixture was not analysed directly, dilutions made from it suggest an Ni(C0)4 concentration in the range 150-200 parts per million by volume (vppm). Standards in the range 1-5 vppm were prepared by transferring about 60 kPa of the stock mixture into an evacuated cylinder and pressurizing to 4 MPa as before.Subsequent dilutions of this standard provided mixtures with nominal Ni(C0)4 concentrations as low as 5 vppb. Cylinders with nominal Ni(C0)4 concentPations greater than 0.5 vppm were standardized by passing a known volume of gas over active charcoal and assaying the charcoal for nickel. The method has been described in detail elsewhere.3 This technique allows the standardization of gas mixtures independently of any Ni(C0)4 source. Gas mixtures contain- ing 4.5 and 1.2 vppm of Ni(C0)4 were analysed again after 3700 and 2850 h, respectively, at laboratory temperature (20 k 2 "C). No significant change in Ni(C0)4 concentration was observed, indicating the stability of these gas mixtures.Gas mixtures analysed in this way were, therefore, regarded as primary standards. Analysis of mixtures with nominal concentrations below 0.5 vppm was effected by gas chromatography. A calibration graph was prepared by dynamic dilution of a primary standard with C02 by using commercially available electronic mass flow meters (Brooks Instruments Model 5850). This graph was used to standardize mixtures with concentrations down to 50 vppb. Dilution of this mixture allowed the standardization of more dilute mixtures in the same manner. Dynamic dilution and a series of standards allowed the production of a continuous range of concentrations, covering several orders of magnitude, for the assessment of instrument performance. Results Column Performance Experiments involving the use of a 0.25 ml gas sample with a helium carrier gas flow-rate of 2 ml min-1 showed that, on the column described above at 30 "C, the major components of the mixture (CO and C02) were essentially not retained.Nickel tetracarbonyl was clearly separated from these materials, and was eluted as a clean, Gaussian peak. Experiments with carbon monoxide samples, which were known to contain both Ni(C0)4 and Fe(CO)S, showed Fe(CO)S to be eluted much later than Ni(C0)4, as might be expected from their relative boiling points [Ni(CO), 43 "C; Fe(CO)S 102.8 " C ] . ~ ~ ~ The column, therefore, provides adequate resolution of the two carbonyl compounds formed by the direct reaction of CO with a metallic substrate. Similar results were obtained with a CO carrier.Although the studies described in this paper were per- formed with CO-C02 as the sample matrix, the same chromatographic conditions should provide satisfactory resol- ution for the investigation of gaseous environmental samples, as the major components of air are not retained on the type of column used here. Carrier Gas Composition The significance of this parameter lies in the possibility of premature decomposition of Ni(C0)4 on the column or in the detector. This phenomenon could be reduced by the addition of CO to the carrier gas. The response of the detector to a 10 vppb gas standard was measured by using pure helium and CO as carrier gases. With the exception of the detector temperature, which was 105 "C, the other conditions were those shown in Table 1 for the isothermal example.With helium as the carrier gas the peak area obtained was 90 arbitrary units, whereas with carbon monoxide the peak area was 890 arbitrary units. These results indicate that detector response can be substantially enhanced by using CO as the carrier gas. Detector Temperature The sensitivity of the ECD is known to be a function of temperature. 1" The significance of detector temperature in these analyses was assessed from the response of the detector to a 0.25 ml sample of a 19 vppb Ni(C0)4 standard. The carrier gas was pure carbon monoxide and the make-up gas pure nitrogen. Peak area is shown as a function of detector temperature in Fig. 1. Detector response decreases steadily with tempera- ture. It is therefore clear that, for optimum response, the detector should be operated at a relatively low temperature, but that the temperature should be constant to ensure reproducibility. The latter criterion requires operation at temperatures significantly above ambient; a value of 60 "C was adopted.Oven Temperature The most straightforward mode of operation is to maintain a constant column temperature throughout the analysis. In the present situation, adequate temperature control could be maintained at an oven temperature of 30 "C. The use of an isothermal column, however, limits the maximum sample size that can usefully be employed. The need to maintain satisfactory resolution limits carrier gas flow-rate, and, hence, the rate at which the sample loop can be purged. An excessive sample volume results in malformed peaks and poor resolu- tion.Under isothermal conditions, a 0.25 ml sample was the largest that could be used reliably. One strategy for increasing sample size is to operate the column at low temperature during the flushing of the sample loop. Under these conditions, Ni(C0)4 progresses extremely slowly through the column and is concentrated in a narrow band at the start of the column. Raising the temperature then allows elution of this material as a sharp peak. Fig. 2 shows the effect of this strategy on a 1.0 ml gas sample containing 19 vppb of Ni(C0)4. In this example, the column temperature was kept at -20 "C for 1 min, then ramped at 40 "C min-1 to a loo0 1 900 m m Y m a $' 700 600 0 0 0 3 500 50 100 150 200 Detector temperature/"C Fig.1 Effect of detector temperatureANALYST, FEBRUARY 1991, VOL. 116 h v) C c .- = 1000 2 F 2 & 100. c .- - m Y m a 10 15 1 - . 175 loooo - IJY c .- 5150 - 2 F $125 - + .- Y a, c ;loo - ? 8 75 n L c 0 - + 1.0 1.5 2.0 2.5 3.0 3.5 tlm i n Fig. 2 Effect of temperature programme on peak shape. A, Temperature programmed; and B, not temperature programmed Table 2 Calibration data at low concentration Isothermal Non-isothermal Slope 87.72 2046 Standard error of slope 0.90 56 In terce p t 5.5 17 Correlation coefficient 1 .OO 0.99 Number of data 9 20 Standard error of intercept 3.1 12 Detection limit (vppb) 0.08 0.01 constant temperature of 10 "C. A maximum sample size of 3 ml could be accommodated by using the temperature programme given in Table 1.Sensitivity Calibration graphs prepared under both isothermal and non-isothermal conditions, as defined in Table 2, are shown in Fig. 3. Over the range of concentrations considered the graphs are significantly non-linear. The relationships between peak area (A) and concentration ( c ) in vppb are given by: A = 105c0.87, for the isothermal example; and A = 1158c'JXl, for the non-isothermal example. In order to establish the detection limit, only data with peak areas of less than about 300 arbitrary units were considered. Over this restricted range the linearity was excellent in both instances. Unweighted linear regression by least squares was used to determine the slope and intercept of the best line through the data. These values, together with the associated standard errors and correlation coefficients, are given in Table 2.Also shown in Table 2 are the values for the detection limits. These have been defined as those concentrations at which the lower 95% confidence limit becomes equal to the calculated y-intercept. 1 1 The results of this exercise are summarized in Table 2. The sensitivity of the method is clear from the detection limits quoted. That the difference in detection limit between isothermal and non-isothermal exam- ples is not pro ruta with sample size reflects the greater degree of scatter in the non-isothermal data. 0.001 0.01 0.1 1 10 100 Ni(C0h (vppb) Calibration graphs. A, 0.25 ml loop size (isothermal); and B , Fig. 3 3 ml loop size (non-isothermal) Conclusions Gas chromatography with an ECD has been shown to be an effective technique for the determination of Ni(CO)4 in CO-C02 gas mixtures.Appropriate choice of carrier gas and chromatographic conditions allows detection limits as low as 0.01 vppb to be obtained. Although the method has been devised for the analysis of CO-C02 gas mixtures, the technique should also be appropriate for the analysis of air samples. In the latter instance, however, consideration would have to be given to the stability of the gas sample if analysis and sampling took place at different locations. This work was jointly funded by the United Kingdom Atomic Energy Authority and Nuclear Electric plc under the Thermal Reactor Agreement. The author is grateful to these bodies for permission to publish this paper. 1 2 3 4 5 6 7 8 9 10 11 References Occupational Exposure Limits, HSE Document 40/90, HM Stationery Office, London, 1986. American Conference of Governmental Industrial Hygienists List, 1989-90. Eller. P. M., Appl. tnd. Hyg., 1986, 1, 115. Stedman, D. H.. Tammaro, D. A., Branch, D. K . , and Pearson, R., Anal. Chem., 1979, 51, 2340. Stedman, D. H., and Tammaro, D. A., Anal. Lett., 1976,9,81. Houpt, P. M . , van der Waal. A., and Langeweg, F., Anal. Chim. Acta. 1982, 136, 421. Hikade, D. A., Stedman, D. H., and Walega, J. G., Anal. Chem.. 1984, 56, 1629. Handbook of Chemistry and Physics, ed. Weast. R. C., CRC Press, Boca Raton, FL, 60th edn., 1980, B-101. Handbook of Chemistry and Physics. ed. Weast, R. C., CRC Press. Boca Raton, FL, 60th edn., 1980, B-86. Grob, R. L.. Modern Practice of Gus Chromatography, Wiley-Interscience, New York, 1985, p. 262. Sharaf, M. A.. Illman, D. L., and Kowalski, B. R., Chemo- metrics, Wiley-Interscience, New York. 1986, p. 128. Paper 0/02273J Received May 22nd, I990 Accepted September 27th, I990

ISSN:0003-2654    

DOI:10.1039/AN9911600149

出版商:RSC    

年代:1991

数据来源: RSC