摘要:

THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYEDITORAL ADVISORY BOARD"Chairman: H. J. Cluley (Wembley)"L. S . Bark (Salford)R. Belcher (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)L. R. P. Butler (South Africa)"R. M. Dagnall (Huntingdon)E. A. M. F. Dahmen (The Netherlands)A. C. Docherty (Billingham)D. Dyrssen (Sweden)J. Hoste (Belgium)H. M. N. H. Irving (Leeds)H. Kaiser (Germany)M. T. Kelley (U.S.A.)W. Kemula (Poland)"W. T. Elwell (Birmingham)"J. A. Hunter (Edinburgh)"G. F. Kirkbright (London)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)"J. M. Ottaway (Glasgow)"G. E. Penketh (Billingham)E. Pungor (Hungary)D. I. Rees (London)"R. Sawyer (London)P. H. Scholes (Sheffield)"W. H.C. Shaw (Greenford)S. Siggia (U.S.A.)A. A. Smales, O.B.E. (Harwell)A. Walsh (Australia)T. S. West (Aberdeen)A. L. Wilson (Medmenham)P. Zuman (U.S.A.)"A. Townshend (Birmingham)"Members of the Board serving on The Analyst Publications CommitteeREGIONAL ADVISORY EDITORSDr. J. Aggett, Department of Chemistry, University of Auckland, Private Bag, Auckland, NEWProfessor G. Ghersini, Laboratori CISE, Casella Postale 3986, 201 00 Milano, ITALY.Professor L. Gierst, Universite Libre de Bruxelles, Facult6 des Sciences, Avenue F.-D. Roosevelt 50.Professor R. Herrmann, Abteilung fur Med. Physik., 63 Giessen, Schlangenzahl 29, GERMANY.Professor W. E. A. McBryde, Dean of Faculty of Science, University of Waterloo, Waterioo, Ontario,Dr. W.Wayne Meinke, KMS Fusion Inc., 3941 Research Park Drive, P.O. Box 1567, Ann Arbor,Dr. 1. Rubeska, Geological Survey of Czechoslovakia, Kostelni 26, Praha 7, CZECHOSLOVAKIA.Dr. J. RGiiEka, Chemistry Department A, Technical University of Denmark, 2800 Lyngby. DENMARK.Professor K. Saito, Department of Chemistry, Tohoku University, Sendai, JAPAN.Dr. A. Strasheim, National Physical Research Laboratory, P.O. Box 395, Pretoria, SOUTH AFRICA.ZEALAND.Bruxelles, BELGIUM.CANADA.Mich. 48106, U.S.A.Published by The Chemical SocietyEditorial: The Director of Publications, The Chemical Society, Burlington House,London, W1 V OBN. Telephone 01 -734 9864. Telex No. 268001.Advertisements: J. Arthur Cook, 9 Lloyd Square, London, WC1 X 9BA. Telephone 01 -837 631 5.Subscriptions (non-members): The Chemical Society Publications Sales Office, Blackhorse Road,Lecthworth, Herts., SG6 1 HN.Volume 101 No 1201@ The Chemical Society 1976April 197

ISSN:0003-2654    

DOI:10.1039/AN97601FX013

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

ANALAO 101 (1201) 225-320 (1976)ISSN 0003-2654April 197622524425526026527227828629229830631 031 531 7THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYCONTENTSREVIEW PAPERThe Radioimmunoassay o f Drugs-J. Landon and A. C. MoffatORIGINAL PAPERSStudies on the Oxidation o f Amitriptyline-B, Caddy, F. Fish and J. TranterThe Determination o f Chromium in Plain Carbon Steel and Low-alloy Iron andSteel by Atomic-absorption Spectrophotometry-W. D. Cobb, W. W. Fosterand T. S. HarrisonThe Use o f Luminescence Spectroscopy in Aiding the Identification of Com-mercial Polymers-N. S. Allen, J. Homer and J. F. McKellarThe Use of a Spark as a Sampling - Nebulising Device f o r Solid Samples inAtomic-absorption, Atomic-fluorescence and Inductively Coupled PlasmaEmission Spectrometry-H.G. C. Human, R. H. Scott, A. R. Oakes and C. DWestDetermination of Submicrogram Amounts o f Mercury in Various Matrices byFlameless Atomic-fluorescence Spectrometry-P. Cavalli and G. RossiThe Determination o f Sodium in Water a t Ultra-trace Concentrations by Flame-less Atomic-absorption Spectrophotometry-D. J. Gardner, J. A. Pritchardand M. A. SadlerSpectrophotometric Micro-determination of Silicon-bonded Hydrogen withCarbenium-ion Reagents-Julian Chojnowski, Mieczyslaw Mazurek and LechWi lcze kA Critical Comparison of the Balling Procedure f o r the Measurement o fObscuration of Unsweetened Spirits with an Official French Method-P. J. WagstaffeThe Gas-chromatographic Determination o f Selenium(V1) and Total Seleniumin Milk, Milk Products and Albumin with 1,2-Diamino-4-nitrobenzene-Yasuaki ShimoishiThe Determination of Sorbitol in Foodstuffs Using Thermometric Titrimetry-L. S. Bark, D. Griffin and P. PrachuabpaibulDetermination o f Mercaptoaeetic Acid with Potassium Dichromate in thePresence of Potassium Iodide by Potentiometric and Visual Methods-N. Krishna Murty and K. Rama RaoCOM MU N l CATIONBackground Correction in Electrothermal Atomic-absorption SpectroscopyUsing the Zeeman Effect i n the Atomised Sample-J. B. Dawson, E. Grassam,D. J. Ellis and M. J. KeirBook ReviewsSummaries of Papers in this lssue-Pages iv, v, viii, x_.Printed by Heffers Printers Ltd, Cambridge, EnglandEntered as Second Class at New York. USA, Post Offic

ISSN:0003-2654    

DOI:10.1039/AN97601BX015

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

iv SUMMARIES OF PAPERS I N THIS ISSUE April, 1976Summaries of Papers in this IssueThe Radioimmunoassay of DrugsA ReviewSummary of ContentsIntroductionTerminologyBasis of radioimmunoassayRelated analytical techniquesRequirements for a drug radioimmunoassaySupplies of a suitable antiserumSupplies of labelled antigenMethod for separating the antibody-bound and free fractionsOther requirementsSensitivitySpecificityPrecision and accuracyPracticality and applicabilityDrug radioimmunoassay kitsAutomationUse of alternative analytical techniquesAdvantages and disadvantages of drug radioimmunoassayAnticipated trends in drug radioimmunoassayConclusionReprints of this Review paper can be obtained from the Publications SalesOfficer, The Chemical Society, Blackhorse Road, Letchworth, Herts, SG6 lHN,at L l per copy (with a 25% discount for six or more copies), post free.A remittance for the correct amount, made out to The Chemical Society,should accompany every order; these reprints are not available through TradeAgents.J.LANDONDepartment of Chemical Pathology, St. Bartholomew’s Hospital, West Smithfield,London, EClA 7BE.and A. C. MOFFATHome Office Central Research Establishment, Aldermaston, Berkshire, RG7 4PN.Analyst, 1976, 101, 225-243.Studies on the Oxidation of AmitriptylineThe quantitative determination of amitriptyline extracted from biologicalfluids has been achieved using an oxida.tion procedure. By use of selectiveoxidation methods a knowledge of the ;evels of 10-hydroxylated metabolitesof amitriptyline can be obtained.Some evidence is presented that dehydro-amitriptyline, or its demethylated derivatives, are natural metabolites ofamitriptyline in man.B. CADDY, F. FISH and J. TRANTERDivision of Pharmacognosy and Forensic Science, School of Pharmaceutical Sciences,University of Strathclyde, Glasgow, G1 IXW.Analyst, 1976, 101, 244-254April, 1976 SUMMARIES OF PAPERS I N THIS ISSUEThe Determination of Chromium in Plain Carbon Steel andLow-alloy Iron and Steel by Atomic- absorption SpectrophotometryCo-operative examination of published atomic-absorption procedures forseveral elements in steel showed good agreement between results apart fromthose for chromium. For this element the depressive effect of iron in theair - acetylene flame is eliminated when using the preferred nitrous oxide -acetylene flame.However, control of flame composition is essential asvanadium, molybdenum, aluminium and titanium in low-alloy synthetic testsolutions and samples increasingly enhance the chromium absorbance withincreased fuel richness. Copper, nickel and iron have no effect but, in anyevent, iron should be included in the calibration solutions. Hence a lean,oxidising flame must be used and the results for samples then agree closelywith the certificate values.W. D. COBB, W. W. FOSTER and T. S. HARRJSONBritish Steel Corporation, Scunthorpe and Lancashire Group, P.O. Box No. 1,Scunthorpe, South Humberside, DN16 1BP.Analyst, 1976, 101 , 255-259.VThe Use of Luminescence Spectroscopy in Aiding theIdentification of Commercial PolymersThe fluorescence and phosphorescence properties of a range of commercialpolymers have been examined.Each polymer tested exhibited characteristicexcitation and emission spectra. Phosphorescence lifetimes were also foundto vary significantly. The spectral data obtained are discussed with a view toapplying luminescence techniques as a rapid non-destructive method for thecharacterisation of commercial polymers.N. S. ALLEN, J. HOMER and J. F. McKELLARDepartment of Chemistry and Applied Chemistry, Salford University, Salford,M5 4WT.Analyst, 1976, 101, 260-264.The Use of a Spark as a Sampling - Nebulising Device for SolidSamples in Atomic- absorption, Atomic- fluorescence andInductively Coupled Plasma Emission SpectrometryA conventional high-voltage spark, operating at 50 Hz, is used for samplingand nebulising material from solid conducting samples. Gas, fed throughthe spark chamber, transports the metal particles into a flame or plasmafor analysis by atomic-absorption, atomic-fluorescence or plasma emissionspectrometry. Calibration graphs for the copper present in aluminium alloysand iron in brass are presented.H. G. C. HUMAN, R. H. SCOTT, A. R. OAKES and C. D. WESTNational Physical Research Laboratory, CSIR, P.O. Box 395, Pretoria 0001,South Africa.Analyst, 1976, 101, 265-271

ISSN:0003-2654    

DOI:10.1039/AN97601FP025

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

...Vlll SUMMARIES OF PAPERS I N THIS ISSUEDetermination of Submicrogram Amounts of Mercury in VariousMatrices by Flameless Atomic-fluorescence SpectrometryAn atomic-fluorescence spectrometric method is described for the dctermina-tion of mercury at the nanogram level. Solid samples are burnt in anoxygen stream, the combustion gases being condensed in a liquid nitrogencooled trap. Following the dissolution of the condensed matter, mercury isextracted by the reduction - aeration method and collected on gold, forming anamalgam. The combination of a high-efficiency light-gathering system witha gas-shielded windowless fluorescence cell, and the described procedure forthe release of mercury vapour from the amalgam, allows a detection limitof 0.03 ng to be obtained.A linear working graph covering the concentrationrange of mercury from 0.5 to 100 ng was established with aqueous standardsolutions of mercury.The whole procedure has been checked with two different NBS StandardReference Materials, excellent agreement of the measured values with thecertified values being found.P. CAVALLI and G. ROSS1Chemistry Division, Euratom-CCR, 21020-Ispra (Varese), Italy.April, 1976Analyst, 1976, 101, 272-277.The Determination of Sodium in Water at Ultra-traceConcentrations by Flameless Atomic-absorption SpectrophotometryA flameless atomic-absorption technique has been developed for the deter-mination of sodium in water at levels of 0.1 pg kg-1 and lower. A practicallimit of detection of 0.011 pg kg-l has been determined and the method isshown to be unaffected by the presence of ammonia at concentrations ofup to 8 mg kg-l.No anion interference has been found for the anions chloride,sulphate and hydroxide. In laboratory tests, a comparison of the atomic-absorption results with those obtained by using a sodium-selective glasselectrode indicated a positive bias in the electrode results. This bias was,however, found to be reduced during power-station plant trials when theglass electrode was presented with a low sodium water over a period of days.Thus the sluggish response of the electrode system may have contributedto the observed bias in the laboratory tests.D. J. GARDNER, J. A. PRITCHARD and M. A. SADLERCentral Electricity Generating Board, South Western Region, Scientific ServicesDepartment, Portishead, Bristol, BS20 9DH.Analyst, 1976, 101, 278-285.Spectrophotometric Micro - determination of Silicon- bondedHydrogen with Carbenium-ion ReagentsA sensitive spectrophotometric method for the micro-determination ofhydrogen bound to silicon has been developed. It is based on the reactioninvolved in hydride transfer from silicon to a carbenium ion.The methodmay be useful for the determination of trace amounts of the -Si-H group insilicone oils and in the siloxane oligomers used as intermediates in the siliconeindustry.JULIAN CHOJNOWSKI, MIECZYSLAW MAZUREK and LECH WILCZEKPolish Academy of Sciences, Centre of Molecular and Macromolecular Studies,90-362 E6di, Poland.Analyst, 1976, 101, 286-291X SUMMARIES OF PAPERS I N THIS ISSUE April, 1976A Critical Comparison of the Balling Procedure for theMeasurement of Obscuration of Unsweetened Spirits with anOfficial French MethodThe Balling procedure and an official French method for the determinationof obscuration and true alcoholic strengths have been examined.It is demon-strated that whereas the Balling procedure provides accurate values forobscuration, the French method yields consistently erroneous results. It istherefore concluded that the equation employed in the French method isnot applicable to determinations of alcoholic strength expressed in percentageby volume (20 OC/20 "C), having been derived for use *with measurementsof alcoholic strengths expressed according to the former French (15 "C/15 "C)percentage by volume scale.P.J. WAGSTAFFEDepartment of Industry, Laboratory of the Government Chemist, Cornwall House,Stamford Street, London, SE1 9NQ.Analyst, 1976, 101, 292-297.The Gas-chromatographic Determination of Selenium(V1) andTotal Selenium in Milk, Milk Products and Albumin with1,2 -Diamino - 4-nitrobenzeneSelenium (IV) reacts with lf2-diamino-4-nitrobenzene to form 5-nitropiaz-selenol, which is detected by means of a gas chromatograph equipped withan electron-capture detector. Trace amounts of selenium(V1) and totalselenium in milk, milk products and albumin were determined by thismethod with a practical detection limit of 0.005 pg. The amount ofselenium(V1) found in milk was about 60% of the total selenium content.YASUAKI SHIMOISHIDepartment of Chemistry, Faculty of Science, Okayama University, Tsushima,Okayama-shi 700, Japan.Analyst, 1976, 101, 298-305.The Determination of Sorbitol in Foodstuffs Using ThermometricTitrimetryA method is described for the direct thermometric titrimetric determinationof sorbitol in the presence of other carbohydrates and permitted food colour-ings. The sorbitol is made to react with sodium periodate and the heat of thereaction is used to determine the end-point of the titration.The method isaccurate to within 1.5% for the samples analysed. The results are comparedwith those obtained by using a previously reported method.L. S. BARK, Id. GRIFFIN and P. PRACHUABPAIBULRamage Laboratories, University of Salford, Salford, M5 4WT.Analyst, 1976, 101, 306-309.Determination of Mercaptoacetic Acid with Potassium Dichromatein the Presence of Potassium Iodide by Potentiometric andA potentiometric procedure for the assay of mercaptoacetic acid with potas-sium dichromate in the presence of potassium iodide is described. Theoxidation of mercaptoacetic acid proceeds via the liberation of iodine frompotassium iodide (added to the titration mixture) by the addition of potassiumdichromate. Oxalic acid is used as the catalyst to accelerate the reactionbetween potassium iodide and potassium dichromate. Mercaptoacetic acidis also determined visually with potassium dichromate, using starch ornaphthol blue - black as indicators.N. KRISHNA MURTY and K. RAMA RAODepartment of Chemistry, Andhra University, Waltair, India.Visual MethodsAnalyst, 1976, 101, 310-314

ISSN:0003-2654    

DOI:10.1039/AN97601BP029

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

APRIL 1976 The Analyst Vol. 101 No. 1201 The Radioimmunoassay of Drugs A Review* J. Landon Department of Chemical Pathology, St. Bartholomew's Hospital, West Smithfield, London, E C l A 7BE and A. C. Moffat Home Ofice Central Research Establishment, Aldermaston, Berkshire, RG7 4PN Summary of Contents Introduction Terminology Basis of radioimmunoassay Related analytical techniques Requirements for a drug radioimmunoassay Supplies of a suitable antiserum Supplies of labelled antigen Method for separating the antibody-bound and free fractions Other requirements Sensitivity Specificity Precision and accuracy Practicality and applicability Drug radioimmunoassay kits Automation Use of alternative analytical techniques Advantages and disadvantages of drug radioimmunoassay Anticipated trends in drug radioimmunoassay Conclusion Introduction The first radioimmunoassay was developed for insulin by Yalow and Bers0n.l Initially, this technique was applied, predominantly by endocrinologists, to the assay of circulating levels of peptide and protein hormones and was later extended to compounds, such as steroids and the thyroid hormones, which must be covalently linked to a carrier molecule in order to induce an antibody response.In 1968, Oliver et aL2 published the first radioimmunoassay for a drug, digitoxin, and since that time such assays have been established for a large number of drugs (Table I). These assays are proving of considerable value in pharmacodynamic and pliarmacokinetic studies and also in clinical practice, especially for drugs (such as digoxin and gentamicin, whose therapeutic levels approach those at which serious side-effects may occur).The purpose of this review is to discuss the basis, requirements, advantages and disad- vantages of radioimmunoassay with respect to drugs. The explosive growth of this subject enables only a superficial view to be given and it has been necessary to exclude drugs that have an endogenous counterpart (such as the vitamins, prostaglandins and thyroid hormones). Readers who require more detailed information about the general subject are referred t o books edited by Kirkham and Hunter113 and by Ode11 and Daughaday114 and to a recent issue of the British Medical BztZletin.ll5 * Reprints of this review will be available shortly. For details, see summaries in advertisement pages.225226 LANDON AND MOFFAT: THE RADIOIMMUNOASSAY Analyst, VoZ. 101 TABLE I DRUGS AND INSECTICIDES FOR WHICH RADIOIMMUNOASSAYS HAVE BEEN REPORTED Class of drug Analgesics Antibiotics Anticonvulsants Antidiabetic agents Antihypertensives Antineoplastic agents Barbiturates Cardiac depressants Cardiac glycosides Hallucinogens Insecticides Muscle relaxants Drug Aspirin Codeine E torphine Fentany 1 Hydromorphone Methadone Morphine Pentazocine Adriam ycin Amikacin Chloramphenicol Daunomycin Gentamicin Isoniazid Penicillins Sulphonamides Tetracycline Phenytoin Sulthiame Glibenclamide Saralasin Methotrexate Barbitone Pentobarbitone Phenobarbitone Others Procaineamide Digitoxin Digoxin Gi taloxin Gitoxin Ouabain Cannabinoids Lysergide Mescaline 2,5-Dimethoxy-4- methylamphetamine Aldrin DDT Dieldrin Malathion Tubocurarine Reference Butler et aL3*; Weiner et aL4*; Girard e t aL5* Spector and Parkers; Wainer et al.? Robinson et a1.8 Henderson et al.9 Wainer et al.? Liu and AdlerlO Spector and Parkera; Adler and Lid1; Spector12; Van Vunakis et al.13; Gershman et al.'4; Wainer et al.15; Spector et al.ls; Miller et al.17; Lindquist and Spratt'n; Catlin et al.lg; Berkowitz et a1.20; Ringle and Herndonzl; Berkowitz et a1.22; Capelle et aLa3; Koida et u Z .~ ~ ; Gross et aLZ5; Mu16 et aLZ6; Morris et al.Z7; Mu16 et al.28 Williams and Pittman20 Van Vunakis et al.30 Lewis et al.31 Hambergers2* Van Vunakis et aL30 Lewis et al.33; Mahon et al.34; Berk et aLS5; Mishew et al.3s Schwenk et al.37 Gleich and StankievicS8; Karchmer at aZ.SB; Hyslop et al.40; Wal and Kann41 W e d ~ r n ~ ~ * Queng et al.43 Tigelaar et al.44; Cook et aZ.46; Montgomery et al.dO; Robinson et al.47 Robinson et al.47 Glogner et al." Pettinger et aLp9 Jaton and Ungar-Waronso*; Bohuon et aL51; Levine and Powers5a; Raso and Schreiber5S Spector et al.ls Spector and F l ~ n n ~ ~ ; Flynn and Spectorss Chung et al.56; Satoh et ~ 1 .~ ~ ; Satoh and Kuroiwa68; Satoh et a1.6g Mu16 et a1.26; Mu16 et aLZ8 Russell and Ziffso* Oliver et aL2 and many others Smith et a1.61 and many others Lesnee2 Lesnes2 Seldon and SmithG3 Teale et al.64; Teale et ~ 1 . ~ ~ ; Teale et a1.6a; Gross et al.67; Teale et a1.68; Teale et aLa9; Marks et al.70 Van Vunakis et al.?l; Castro et al.72; Taunton-Rigby et al.73; Loeffler and Pierce74; Wingeleth et al.75 Van Vunakis et al.7s; Schnoll et aL77; Riceberg et al.78 Van Vunakis et al.7a; Riceberg et aL78; Kid0 et a1.79 Langone and Van Vunakis80 Centeno et a1.8l Langone and Van Vunakisso Centeno et a1.S1 Horowitz and Spector82 * Production of antibodies only was reported.[continuedApril, 1976 OF DRUGS. A REVIEW TABLE I--co;rztin.ued 227 Class of drug Drug Reference Morris et ~ 1 . 8 ~ Dumasia et aLe4; Lee and Ohsowaa6; Meikle et ~ 1 . 8 6 ; Hichens and Hoganse7 KunduS8; RaoaO; Rao et aZ.OO ColburnOl Cornette et aLg2; Royer et aZ.98 Kunduee; Raose; Kundus4 Colburn and Bullere5 Nygren et aZ.96 Cameron et d g 7 ; Warren and FotherbyQB Warren and FotherbyQa Colburng9 Cheng et uL10O; Mu16 et aZ.ZS Hooker and Boydl01 Langone et aZ.lo2; Langone and Van Vunakis103 Cernosek et uZ.'04; Langone et ~ 1 .~ 0 2 ; Haines et aZ.105; Matsukura et d 1 0 6 Langone et aZ.lo7 Steroids Stimulants Corticosteroids Dexamethasone Ethinyloestradiol Fluoxymestrone Medroxyprogesterone Mestranol Meth ylprednisolone Norethindrone Norethisterone Norgestrol Prednisone Amphetamine Strychnine Tobacco alkaloids Cotinine Nicotine Tranquillisers y- (3-Pyridyl) -y-0x0- N-me thylbu tyramide Chlordiazepoxide Chlorpromazine Diazepam Perphenazine Pimozide Dixon et aZ.108 Spectorlog; Shostek, quoted by Marks et ~1.110 Peskar and Spectorlll Shostek, quoted by Marks et ~1.110 Michiels et Terminology New disciplines tend to introduce their own terminology, which may vary from centre to centre or country to country. It may be helpful, therefore, to define in this section the more common terms that are used.Antigen (Ag) : a substance that will combine with a specific antibody. Iwzmztnogen : a substance that will provoke an immune response. Whereas most proteins are both immunogens and antigens, most drugs are antigens, as they can be bound by an antibody, but are not immunogenic unless first coupled to a larger molecule such as a natural protein (e.g., bovine serum albumin, BSA) or synthetic protein (e.g., polylysine). Hapten: an antigen that must be coupled to a larger molecule in order to provoke an anti- body response. Adjuvapzt : material added to an immunogen to enhance the antibody response. The most widely employed adjuvant is that introduced by Freund,l16 which consists of a neutral deter- gent (Arlacel A), paraffin oil and (in the "complete" form) killed mycobacteria.Antigenic determinant: the part (or parts) of the antigen molecule that combines with the binding site of an antibody. For a peptide or protein, each determinant comprises a sequence of about three to five amino-acids, and has a diameter of less than 1.5nm. With drugs, it comprises only part of the molecule, which is usually that most exposed on the conjugate used for immunisation. Antibody (Ab) : gamma-globulins, usually of the IgG class, which bind specifically to an antigen. Antibody binding site : that part of the gamma-globulin which binds to an antigen. Every IgG molecule has two combining sites, in its Fab portion, each comprising the N-terminal sequence of one light and one heavy chain.Antibody titre: that dilution of an antiserum that will bind a particular percentage (often chosen as 50%) of a fixed amount of labelled antigen in a fixed incubation volume. Antibody avidity: the energy of the reaction between the binding sites of an antibody and an antigen. It is essentially the same as the association constant (K) in physical chemistry, as discussed in the next section. Labelled antigen (*Ag) : an antigen that has been modified in order to enable its presence to be measured accurately. This modification most commonly involves the introduction into the molecule of a gamma- (1251, 1311 or 75Se) or beta- (14C, 3H) emitting isotope, but other labels (such as enzymes or fluorescent molecules) can be used.228 LANDON AND MOFFAT: THE RADIOIMMUNOASSAY Analyst, 'VOL?.101 Radioiodination : the process of incorporating 1251 or 1311 into antigen molecules, Isotope abundance: the percentage of the total element present as the isotope. For ex- ample, for NalSlI, only about 15-20% of the total iodine is 1311, whereas for Na1251, more than 80% is 1251. Yield : the percentage incorporation of label into antigen molecules. Specific activity : the amount of label incorporated into antigen molecules which, in the case of isotopes, is expressed as the radioactivity per unit mass of drug in curies per mole, where 1 Ci gives 2.22 x 10l2 disintegrations per minute." Assay specijcity: reflects the degree to which an assay is affected by substances other than that for which the assay was designed. Basis of Radioimmunoassay As recognised by Arrhenius in 1907, the reaction between the binding sites of an antibody and its specific antigen obeys the law of mass action: kl k2 Ag + Ab + Ag:Ab where Ag represents free antigen (the free fraction), Ag:Ab the antigen present in the anti- body-bound form (the bound fraction) and k, and k2 are the rates of the forward and back- ward reaction, respectively. The technique is based on determining the percentage of the total amount of Ag (free plus bound) that is present in the antibody-bound fraction (YoB). This percentage depends on only three factors: (a) it is directly related to the total amount of Ab present; (b) it is directly related to the avidity ( K ) with which the Ab binds the Ag, where K = k1/k2 = (Ag:Ab)/(Ab)(Ag) and (Ag:Ab), (Ab) and (Ag) refer to the molar concentrations of Ag : Ab complex, free antibody and free antigen, respectively, at equilibrium.(c) it is inversely related to the total amount of Ag present. In an immunoassay, the same concentration of the same Ab is present in each tube, i.e., factors (a) and (b) are kept constant so that the only factor that influences the percentage of the total Ag in the bound fraction is the total amount of Ag present. In order to deter- mine the percentage of the total Ag that has been bound, a constant amount of labelled antigen is added to each tube (to act as a tracer) and the bound and free fractions are separated by one of a number of techniques. The percentage of the total counts present in the bound fraction can then be determined and will accurately reflect, and be inversely related to, the total amount of antigen present.An unknown amount of the compound to be assayed can thus be quantified by comparing the distribution of the tracer with the distributions pro- duced by a series of standards. This distribution can be expressed in a number of ways, such as the percentage of total counts that are bound (%B) or free (%F) or the ratio of the counts in the two fractions (i.e., F/B or B/F) plotted against the concentration, or the logarithm of the concentration, of unlabelled antigen present in the system. Following immuni- sation of a rabbit with a digoxin:BSA conjugate, the dilution of the antiserum appropriate for assay purposes was established by means of a dilution curve, This involved the in- cubation of 250-pg amounts of [1251]digoxin with doubling dilutions of the antiserum, over the range 1:250 to 1:64000.After equilibrium had been attained, the bound and free fractions of the labelled digoxin were separated and the percentage present in the former was determined. I t can be seen (Fig. 1) that at serum dilutions of 1 : 250-1 : 2 000 most of the [125I]digoxin was bound, whereas at dilutions greater than 1 : 32 000 virtually none was in the bound fraction. Based on this information, it was then possible to set up a standard curve (Fig. 2) in which 250-pg amounts of [1251]digoxin and antiserum at a final dilution of 1 : 4 000 (which binds approximately 50% of the labelled drug in the absence of unlabelled digoxin) were incubated with various concentrations of unlabelled digoxin, over the range 0.25-8.0 ng ml-l.For an actual assay, a patient's sample replaces the standard digoxin and * The curie has recently been replaced by the becquerel as the unit of radionuclide activity in the SI A typical radioimmunoassay, for digoxin, was conducted as follows. system; I Ci = 3.7 x 1O1O Bq.April, 1976 O F DRUGS. A REVIEW 229 the result is read from the standard curve. For example, supposing that only 20% of the [1251]digoxin was bound, then the sample must have contained between 3.0 and 4.0 ng ml-l of digoxin. 1:250 1:l 000 1:4000 1:16000:1:64000 Antiserum dilution Fig. 1. An antiserum dilution curve. Doubling dilutions of serum from a rabbit immunised with a digoxin : BSA conjugate were incubated for 120min with 250pg of [12sI]digoxin. The total incubation volume was 500 p1 and the antibody-bound and free fractions were separated by the addition of dextran-coated charcoal, which adsorbs the latter.50 8 -- 40 2 30 C 3 0 .- 4-d z 20 a 2 0 10 I I I I 0.25 0.5 1 2 4 8 Concentration of unlabelled digoxin/ng mi-’ Fig. 2. Standard curve for various concentrations of unlabelled digoxin, over the range 0.25-8.0 ng ml-l, incubated with 250 pg of [1251]digoxin and rabbit anti- digoxin serum at a final dilution of 1 : 4 000. The total incubation volume was 500 p1 and the incubation time was 120min. The bound and free fractions were separated by the addition of dextran-coated charcoal. Related Analytical Techniques Radioimmunoassay is only one of several analytical procedures that depend upon the specific non-covalent binding of one reactant by another.Many people are confused by the diversity of terms employed to describe these techniques and, as several are already being employed for drug assay, a suggested classification is outlined in Fig. 3. The term “binding assay” is used to cover several techniques that are based on the same principles as radioimmunoassay. Such assays are subdivided primarily by reference to the binding protein employed, as this is the reactant’which imposes both specificity and sen- sitivity. Thus immunoassays employ specific antibodies, while circulating binding protein assays (CBPA) use a naturally occurring plasma protein. For example, it has not yet proved possible to obtain satisfactory antibodies against the major circulating form of vitamin D (25-hydroxycholecalciferol) and serum from rachitic rats is therefore usually employed as the binding protein for its assay.A third group, receptor binding protein assays, employ pro- teins that are usually extracted from the target organ of the drug. These assays can be further subdivided into three groups, namely those in which no labelled reactant is employed and those in which the assay depends upon observing the distribution of either the labelled drug or of the labelled binding protein after separation of the bound and free fractions. The term radioimmunoassay, and its acronym RIA, is excellent in that it signifies that a radioisotope is employed and that it is based on an immunological reaction. It must, how- ever, be restricted to assays in which the antigen is labelled to differentiate the technique from the immunoradiometric assays, introduced by Miles and Hales,ll7 in which labelled specific antibodies are used.Immunoradiometric assays, enzyme immunoassays, immunoenzymo- metric assays and immunoassays employing no labelled reactants are discussed later in this review. The last two groups are finally subdivided depending on the label.230 LANDON AND MOFFAT : THE RADIOIMMUNOASSAY AIzaZyst, Vd. 101 Different binding protein Different reactant label led Binding assays I I I (CBPA) immunoassays Circulating binding Receptor binding protein assays protein assays I I I reactant protein No labelled Labelled drug Label led binding - Isotope (i) Gamma-emitting (ii) Beta-emitting f F I uorescent Different label Enzyme Bacteriophage Particle Protein Fig.3. A classification of drug binding assays. Requirements for a Drug Radioimmunoassay Supplies of a Suitable Antiserum The production of a suitable antiserum is the most important step in developing an immuno- assay for a drug, as it is the antibody which determines sensitivity and specificity. The drug must first be linked covalently to a larger molecule to form a complex that is immuno- genic and the homogeneity of the immunogen assessed and, if necessary, a purification step introduced. The purified complex is then used to immunise appropriate animals according to a pre-determined dose and time schedule and, finally, the animals are bled and their sera assessed for the presence, titre, specificity and avidity of the resultant antibodies. Preearation of the immunogen Considerable attention must be given to this step in order to expose that part of the hapten molecule which differs from its main metabolites and natural or synthetic analogues, This aspect is important because the resultant antibodies are usually directed against the most exposed part of the molecule and it is essential to develop a relatively specific assay if the need for a complex and.time-consuming initial sample purification step is to be avoided. Production of the immunogen requires the availability of a suitable carrier molecule and conjugation reaction and of a reactive group on an appropriate part of the drug. A variety of carrier molecules have been employed, of which the most common have been human and bovine serum albumin (Table 11).Of the various active groups on proteins (Table 111), the most valuable, from the standpoint of conjugation, is the +amino group of lysine. Albumin has 59 such groups, together with a terminal amino group, although many are hindered sterically and are therefore unavailable. Direct conjugation of the drug to the carrier molecule can be accomplished, provided that the drug has a suitably reactive group, such as a carbonyl group (e.g., barbiturates, prosta- glandins and lysergic acid derivatives) or an amino group (e.g., amphetamine). Methods for TABLE I1 CARRIER MOLECULES USED IN DRUG RADIOIMMUNOASSAY Serum albumins, especially of bovine and human origin Egg albumin (ovalbumin) Gamma-globulins Haemocyanin, thyroglobulin, fibrinogen Synthetic polypeptides, especially poly-L-lysine and polyglutamic acid Synthetic polymers, such as polyvinylpyrrolidoneApril, 1976 OF DRUGS.A REVIEW TABLE I11 231 ACTIVE GROUPS OF PROTEIN CARRIERS (FOR CON JUGATION TO HAPTENS) €-Amino group of lysine Phenolic group of tyrosine Sulphydryl group of cysteine Carboxyl group of aspartic or glutamic acid Terminal amino and carboxyl groups Active positions on rings of tyrosine, tryptophan and histidine which Imidazole ring of histidine Hydroxyl group of serine } can be attached to haptens by diazonium salts Less reactive than the above direct conjugation together with a list of other functional groups that can be modified or permit the introduction of a carbonyl or amino group into the molecule are listed in Table IV.Particular problems are encountered with some drugs, including phenytoin (5,5-diphenyl- hydantoin), because they are chemically unreactive in their natural form. It may be possible, however, to use one of their major metabolites (e.g., P-hydroxyphenytoin in the case of phenytoin) as the hapten and, in this way, to obtain antibodies that will bind the parent molecule. TABLE IV METHODS OF COVALENTLY LINKING A DRUG TO A CARRIER PROTEIN See Erlanger118 and references therein for principles and methods of making drug : protein conjugates. Functional group Carboxyl on drug Aromatic amino Aliphatic amino Nitro H ydroxyl Carbonyl Active hydrogen Method of conjugation (i) Use of isobutyl chloroformate to form a mixed anhydride, which will react directly with protein amino groups to form peptide bonds.(ii) Use of carbodiimides (R-N=C=N-R’, where R and R’ are alkyl or aryl groups). l-Ethyl-3-(3-dimethylaminopropyl)carbodiimide or l-cyclohexyl-3-[2-morpholinyl- (4)_ethyl]carbodiimide can be used in aqueous solutions and dicyclohexylcarbodi- imide in organic solvents to form peptide bonds. (iii) Use of N-hydroxysuccinimide, in presence of dicyclohexylcarbodiimide, to form the N-hydroxysuccinimide ester, which may be more active than the free carboxyl group. Direct diazotisation to the active positions of tyrosine, tryptophan or histidine residues. (i) Use of carbodiimides (see above). (ii) Use of glutaraldehyde to link amino groups. (iii) Use of 2-xylylene diisocyanate or toluene-2,4-diisocyanate to link amino groups.(iv) Use of 4-nitrobenzoyl chloride to form 4-nitrobenzoylamide, which can then be reduced to the 4-aminobenzoyl derivative and linked by diazotisation as for an aromatic amino group. Reduce to a primary amino group and then as above. (i) Addition of equimolar proportions of anhydrides of dicarboxylic acids (e.g., succinic) to the drug gives half esters (e.g., hemisuccinate) and the free carboxyl group can then be linked as above. (ii) Introduction of a carboxyl group using w-halo-substituted carboxylic acids. (iii) Formation of a chlorocarbonate with phosgene and use of the Schotten - Baumann (iv) Use of oxidising agents to convert the alcohol into a carboxylic acid. reaction to link i t directly to amino groups on the carrier molecule.(i) Formation of the 0-(carboxymethy1)oxime using 0-(carboxymethy1)hydroxyl- (ii) Introduction of a carboxyl group using 4-hydrazinobenzoic acid. (iii) Schiff base formation directly with primary amino groups on the carrier molecule. (i) Direct conjugation using the Mannich reaction. (ii) Reaction with diazotised 4-aminobenzoic acid to give a derivative with a free amine and reaction of the free carboxyl group as above. carboxyl group.232 LANDON AND MOFFAT: THE RADIOIMMUNOASSAY Analyst, VoZ. IOI The degree of substitution can vary from about 2 to 100 hapten molecules per carrier molecule, but available evidence suggests that conjugates with a relatively low hapten to carrier ratio (less than 15: 1) are the most immunogenic. There is also evidence that a bridge of four carbon atoms in length increases the immunogenicity of the hapten, probably by preventing the carrier molecule from causing steric hindrance between the drug and antibody- producing sites.It is essential to appreciate, however, that animals immunised with such conjugates will produce populations of antibodies directed against the carrier molecule and the bridge as well as against the hapten. Indeed, the affinities of the first two populations are often greater than that of the hapten-directed antibodies, which is of critical importance in the choice of labelling technique, as discussed later. Immunisation The production of a suitable antiserum remains an art rather than a sciencellg and depends on such variables as the species of animal used, the dose and frequency of immunisation and, especially, on chance.Rabbits and guinea-pigs have been commonly employed although there is an increasing tendency to immunise larger mammals, such as goats or sheep. The response to a particular immunogen appears to be genetically determined, probably by a single gene transmitted as a Mendellian dominant. Hence it is best to immunise at least four, and preferably more, random bred animals. Relatively small amounts (about 100 pg) of the immunogen are dissolved in saline and used to prepare a stable water-in-oil emulsion with an adjuvant. The adjuvant is essential as it ensures that the immunogen is released slowly, facilitates its phagocytosis by macrophages and stimulates the reticulo-endothelial system with a resultant multiplication of macrophages and of immunologically competent lymphoid cells.The immunogen is injected into any of a variety of sites, with recent evidence favouring intradermal administration.120 After a delay of 7-8 days, circulating antibodies, predomi- nantly of the IgM class, are detectable and increase in concentration over several weeks. I t is usual to give a booster injection after 4 weeks or more and to take blood 10-14 days later. The response to the second injection differs from the primary response in being more rapid (about 4 days), increasing to a maximum in about 10 days and involving a greater production of antibodies, almost entirely of the IgG class. There is evidence that the most avid anti- bodies are produced by resting the animal for 3-6 months before giving a final boost.119 A summary of a suitable schedule for rabbits is given in Table V.TABLE V A SUGGESTED IMMUNISATION SCHEDULE FOR RABBITS" Step No. 1 2 3 4 5 6 7 8 Procedure Take six or more young rabbits that have had at least 1 week to adjust to their environment. Dissolve (or make a fine suspension of) the immunogen in saline or distilled water a t a concentration of 1 mg ml-1 and then make a stable water-in-oil emulsion of l-ml amounts with 3-ml amounts of Freund's complete adjuvant. Primary injections: inject 2 ml (500 pg) of the emulsion into each animal (a) in 0.25-ml amounts into the muscles of all four limbs and (b) in 0. l-ml amounts into approximately ten intradermal sites on the rabbit's back. Booster injection: after 6 weeks, inject 1 ml (250 pg) of a freshly prepared emulsion, in 0.25-ml amounts, into the muscles of all four limbs.10 and again 14 days after the booster injection, take approximately 40 ml of blood from an ear vein and pool the sera from the two bleeds. Assess each individual antiserum and dispose of those animals which have failed to produce reason- able antibody titres. 3-6 months later give a further booster injection to the remaining rabbits as in step 4, and bleed as in step 5. Continue to give booster injections as in step 7, after disposing of any rabbits that fail to produce high avidity antibodies. * It is appreciated that many other schedules have been employed and have produced satisfactory results. Assessment of antisera The presence of several different antigenic determinants in most immunogens results in the production of several antibody populations, each originating from a single clone of cells and directed against a single determinant.Aeril, 1976 OF DRUGS. A REVIEW 233 Every serum, including each bleed from an individual animal, must be assessed carefully.The titre is determined by means of an antiserum dilution curve as described earlier. From a practical standpoint, the avidity of the antiserum can be regarded as synonymous with the sensitivity that can be achieved in an assay and is determined by setting up a standard curve. The data from such an experiment can also be used to establish the equilibrium constant by means of a Scatchard plot,121 in which the ratio of bound to free antigen is plotted against the concentration of bound antigen, expressed in moles per litre. Finally, the speci- ficity can be determined (a) by constructing a series of standard curves using the drug, its major metabolites, structurally analogous drugs and other exogenous compounds and struc- turally related endogenous compounds, and (b) by assaying actual samples obtained from people not receiving and receiving the drug, to ensure that the results obtained are com- patible with a given situation.Supplies of Labelled Antigen Most drug immunoassays currently employ an isotopically labelled drug both to permit determination of the final distribution between the bound and free fractions and, when an initial purification step is included, to enable allowances for any losses to be made.The isotopes used fall into two groups, each with its advantages and disadvantages. One comprises the beta-emitters 3H and 14C, while the other includes the gamma-emitting iso- topes 1251, 1311 and ‘5e. When the number of samples that will require assay is small, there are many advantages in employing the drug labelled with 3H. Thus, replacement of lH by 3H does not cause marked steric or other changes in the molecule and seldom, therefore, influences its antigenicity; the resultant labelled drug is stable, which, together with the long half-life of 3H (12.3 years), helps to ensure a prolonged shelf-life; and there is minimal health hazard. Nonetheless, 3H-labelled drugs have a number of disadvantages for use in radioimmunoassays, including their low specific activity, the cost of liquid scintillant and counting vials and, because an tibody binding of the labelled antigen may significantly reduce the counts obtained, the need to dissociate the antigen: antibody complex if the bound fraction is to be counted.There is considerable merit in employing a drug labelled with a gamma-emitting isotope when the number of samples to be assayed is large. The samples can be counted directly in a gamma-counter, without the need for liquid scintillant or denaturing the antibody; no correction for quenching or chemiluminescence is required; and the counting time can be markedly reduced, thereby increasing sample throughput, because of the much higher specific activities that can be attained. Thus, it requires approximately 10000 atoms of 14C and 100 atoms of 3H to produce the same number of disintegrations per minute as one atom of lz5I.This has the additional advantage that a relatively inexpensive manual counter can be used for the short counting times required, until the number of samples justifies an automatic counter. 1251 has marked advantages compared with 1311 for labelling, including its greater isotope abundance, better counting efficiency, longer half-life and lesser health hazard (as it is less penetrating). 3H- and 14C-labelled drugs can be prepared by using a normal synthetic route, but this may be very expensive and a cheaper alternative, for 3H, is to hydrogenate the molecule or to exchange it with tritiated water. A variety of methods have also been introduced for radio- iodinating drugs without causing marked structural alterations in their antigenic determinant.These include the substitution of 1251 for 1271 in the small number of drugs (such as thyroxine) that contain iodine in their molecules, and the iodination of a suitable derivative of the drug made specifically for this purpose. Among the compounds used to prepare such derivatives are tyrosine, tyramine and histidine residues and the active acylating agent 3- (4-hydroxypheny1)propionic acid N-hydroxysuccinimide ester. Thus, iodine-labelled digi- toxin has been prepared by first linking it to tyrosine methyl ester and lysergic acid diethyl- amide has been radioiodinated after initial covalent linkage to a suitable synthetic polymer. The iodination can be performed after the derivative has been made or, as is essential with oestrogens whose A ring would be iodinated with consequent loss of antigenicity, a compound can be iodinated first and then linked to a drug.Another approach has been to radioiodinate the tyrosine and histidine residues of a protein covalently linked, by one of the reactions discussed earlier, t o the drug. Under no circum- stances should the conjugate employed as the immunogen be used for this purpose as the234 LANDON AND MOFFAT : THE RADIOIMMUNOASSAY Analyst, VoZ. 101 antiserum will usually contain antibody populations directed against the carrier molecule and/or bridge, which are of higher avidity than those directed against the drug. Thus, the addition of even large amounts of unlabelled drug will not compete with the labelled con- jugate.Success can be achieved provided that both the carrier protein and the reaction employed to produce the conjugate for labelling differ from those used for preparing the immunogen. This is illustrated in Fig. 4, in which a sensitive assay for three forms of genta- micin has been developed based on antibodies raised against a gentamicin : HSA conjugate prepared using carbodiimide, and a gentamicin : [1251]lysozyme conjugate linked by glutar- aldehyde. Labelled gentamicin: HSA could not, however, be used for assay purposes. I I I 0-32 2.5 20 320 Antibiotic concentration/pg ml- ’ 5 000 Fig. 4. A standard curve for gentamicin employing a rabbit antiserum raised against the drug covalently linked to HSA by a carbodiimide reaction and a gentamicin : [1251]lysozyme conjugate prepared using glutaraldehyde.All three components of the gentamicin preparation caused equivalent incubation of binding. Unlabelled gentamicin did not inhibit the binding of a gentamicin: [125I]HSA conjugate by this antibody. The actual radioiodination is usually performed by using chloramine T to oxidise iodide-125 to active iodine-125 and a small reaction volume so as to ensure a high yield.122 Sodium metabisulphite is then added in order to reduce any unreacted iodine-125 back into its in- active form. Exposure of the drug to strong oxidants and reducing agents may, on occasion, damage the antigen and enzymatic oxidation with lactoperoxidase or gaseous diffusion can be employed in order to prevent this damage.The reaction mixture will contain free, unreacted Na1251, unlabelled drug, undamaged labelled drug and labelled drug that has been damaged during the procedure by noxious products in the iodide, which varies between batches, and by the chloramine T. It is therefore essential to purify the various radioactive components and to assess the suitability of the labelled product in an assay. A variety of physicochemical techniques have been employed for purification, including gel and ad- sorption chromatography and electrophoresis. Assessment of the labelled drug involves studies of its binding in an excess of antibody (which should usually exceed 90%) and a check of the sensitivity of the standard curve obtained when it is used as the tracer. Regular checks of its radiochemical purity are also essential.Method for Separating the Antibody-bound and Free Fractions All radioimmunoassays require a separation procedure because the bound fraction does not precipitate spontaneously at the low concentrations employed. A wide variety of such procedures are available that exploit physicochemical or immunological differences between the two fractions, and some of these procedures are listed in Table VI. No single separation method is ideal and, in establishing a new assay, it is wise to compareApril, 1976 OF DRUGS. A REVIEW TABLE VI TECHNIQUES EMPLOYED FOR THE SEPARATION OF BOUND AND FREE ANTIGEN Basis of method Differential migration of the bound and free fractions, caused mainly by: Material or technique used (a) differences in charge Paper chromatography or chromatoelectrophoresis (b) differences in relative molecular Gel filtration mass Gel equilibration 235 Adsorption of the free fraction Precipitation of the bound fraction by: (a) organic solvents (b) salt (c) second antibody Solid-phase methods : (a) first antibody (b) second antibody Coated charcoal Ion-exchange resins Ethanol, dioxan, poly(ethy1ene glycol) Sodium or ammonium sulphate Coated on tubes or discs Covalently linked to Sephadex, paper discs, etc.Pol ymerised Covalently linked to solid matrix at least two, and preferably three, procedures that differ in the principles on which they operate. Those which depend on the differential migration of the two fractions, such as paper or column chromatography, are too complex and the number of samples that can be processed is too small for routine use.Gel equilibration is, however, a valuable adaptation of the latter, especially for use with low-avidity binding proteins, because it is virtually independent of time and temperature. An appropriate porous gel, which excludes the bound but not the free fraction, is added to the assay vials in accurately measured amounts and is present throughout the reaction. Once equilibrium has been reached, an aliquot of the supernatant solution, which contains the antibody-bound antigen, is counted. The Thyopac 4 kit for thyroxine, widely used in clinical chemistry, is based on this separation method, which has also been extended successfully to the assay of cortisol and digoxin. Table VII lists the separation methods that have been reported for just one drug, digoxin.TABLE VII SUMMARY OF SOME AVAILABLE RADIOIMMUNOASSAYS FOR DIGOXIN Isotope used for labelling S H SH SH 1261 1261 1261 Separation technique Adsorption of free fraction on to dextran-coated charcoal (Herbert et uZ.la7) Albumin-coated charcoal Somogyi precipitation (ZnSO, + NaOH) Poly(ethy1ene glycol) precipitation Second antibody precipitation (Hales and Randlelal) Dextran-coated charcoal Gel equilibration Reference Smith et ~ 1 . ~ ' ; Chamberlain et u Z . ~ ~ ~ ; Evered et u Z . ~ ~ ~ ; Oliver et ~ 1 . l ~ ~ ; Hoeschen and Proveda126 Lader et aZ.128 Meade and Kliest12# Barrett and Cohenlso Lanoxitest (Wellcome Reagents Limited) Horgan and RileylS2 Greenwood et ul.lss Adsorption techniques, based on the use of coated charcoal, are simple, rapid, applicable to large numbers of samples and are among the most popular separation methods for drug assay.Their use depends on the avidity with which the charcoal binds a particular drug and236 LANDON AND MOFFAT : THE RADIOIMMUNOASSAY AutdySt, Vd. 101 it is essential that the charcoal is pre-treated (for example, by the addition of serum) in order to prevent it also adsorbing gamma-globulins and, therefore, some of the bound fraction. Precipitation of the bound drug, by addition of neutral salts or organic solvents, also has much to recommend it, provided that the concentration of the precipitant used is carefully chosen so as to bring down all the bound and none of the free fraction. Two other separation methods must be mentioned because of their universal applicability.One, the second antibody technique, involves converting the bound fraction into an in- soluble micelle, by addition of an excess of antiserum raised against the IgG of the species used to produce the antibody employed in the actual assay. It is based on the finding that the antigenic determinants of IgG are separate from its binding sites, and final separation of the bound fraction involves centrifugation or filtration. The second antibody can be added before, during or after the first reaction. The other universal technique depends on the use of antibody attached covalently, or by adsorption, to an insoluble support with separation being achieved merely by pouring off the supernatant containing the free fraction.A variety of insoluble supports have been used, including paper discs, the walls of poly- styrene or polypropylene tubes and cellulose, Sepharose or Sephadex particles. Other Requirements Milligram amounts of the drug are required, in highly purified form, for use as a standard, and this requirement usually poses no problem. There are some drugs, however, which comprise a mixture of structurally related compounds, such as the three different forms of gentamicin (C,,, C, and C,) .present in pharmaceutical preparations. It is then essential to ensure that each behaves identically in the assay (as shown in Fig. 4). Ideally, supplies of the major metabolites of the drug and of structurally related drugs are also required for specificity studies. In some instances, when the antiserum used lacks specificity, it is necessary to undertake a preliminary sample purification step, using paper or thin-layer chromatography or some other procedure. In the longer term, attempts should be continued to improve assay specificity and thereby avoid this requirement.Advantages and Disadvantages of Drug Radioimmunoassay Sensitivity The ultimate sensitivity of any immunoassay depends on the avidity of the antiserum used. Radioimmunoassay is more sensitive than other immunological techniques (such as immuno- diffusion), because full advantage can be taken of the avidity of the antibody, owing to the efficiency of modern counters and the availability of labelled antigens of high specific activity. Many radioimmunoassays can determine circulating levels of a drug in the low picograms per millilitre range.This extreme sensitivity (several orders of magnitude greater than that obtainable by most other analytical techniques) is of great value for the assay of drugs, such as lysergic acid diethylamide, whose circulating levels are normally very low and also in situations when only small sample volumes are available, as in paediatric practice. It is also valuable when biological fluids, such as urine, which must normally be diluted several- fold in order to prevent any non-specific effects, require assay. Sensitivity can often be improved by a non-equilibrium technique in which the standard or sample and antiserum are pre-incubated prior to addition of the labelled Other means of increasing sen- sitivity are (a) to introduce an extraction and concentration step in sample handling, (b) to employ only the most avid antiserum, very small amounts of labelled drug of high specific activity and a prolonged incubation time and/or (c) to keep the volume of sample large in relation to the incubation volume.Extreme sensitivity is seldom necessary for the immu- noassay of drugs. Thus, for example, digoxin circulates at the nanograms per millilitre level, while serum levels of gentamicin are about 1 000 times higher. The range of an assay should be set so as to cover the concentration of the drug expected in the biological fluid to be examined. The time taken to reach equilibrium is related directly to the concentrations of the reactants, and the higher the levels to be determined, the greater are the concentrations of labelled drug and antiserum used and the shorter is the incubation period needed.An assay operating at the limits of its sensitivity is less precise than one that is not. Thus the digoxin assay referred to earlier (which is designed to cover the clinicallyApril, 1976 OF DRUGS, A REVIEW 237 important range, 0.5-8.0 ng ml-l) employs 250 pg of labelled antigen, equilibrium is reached within 1 h and the coefficient of variation for both within-assay and between-assay precision is less than 8%. In contrast, an assay designed to determine the minute concentrations of lysergic acid diethylamide found in urine, which must be diluted 10-fold so as to prevent non-specific interference, employs a much smaller amount of labelled drug and a prolonged period of incubation and is less precise.Specificity The specificity characteristic of radioimmunoassay is an inherent feature of all immunolo- gical procedures, owing to the close “fit” essential for antigen - antibody interactions. None- theless, a variety of factors may affect the reaction other than the concentration of the reac- tants and thereby lead to incorrect results : (i) Decreased antigenicity of the labelled drug. As radioimmunoassays are based on the use of labelled antigen to provide information concerning the distribution of the total antigen, it follows that spurious results may be produced if the tracer differs immunologically from the compound to be assayed. Most commonly, the labelled drug is less antigenic and therefore less is bound, with a consequent loss in assay sensitivity and precision.Factors that result in a loss in antigenicity (termed damage) include the introduction of isotope into the antigenic determinant, partial degradation or oxidation of the drug during labelling and progressive damage during storage. Finally, the labelled drug may be less stable than its unlabelled counterpart, and become degraded during incubation with plasma, serum or urine. This progressive loss in antigenicity, with a resultant decrease in binding, may be attributed incorrectly to high concentrations of unlabelled antigen, unless adequate controls are run and standard tubes contain equivalent amounts of the biological fluid. The problem of antibodies directed against the bridge or protein of a labelled conjugate has been discussed previous1 y .Problems arise if the labelled drug adsorbs non- specifically to the walls of the assay tube or to some component in the biological fluid being assayed. Thus, the adsorbed tracer is not available for binding to the antibody and falsely high or low results will be obtained, depending on the separation technique used. If the sample contains endogeneous binding proteins (such as transcortin or those which bind vitamin B,,), they must be excluded from taking part in the assay, either by introducing an initial purification step or by inactivating them. Many factors may impair antigen - antibody binding non-specifically, especially if the avidity of the antibody used is poor. These include performance of the assay at a non-optimum pH, and the presence in the incubate of high concentrations of protein, urea or electrolytes.Such factors invariably influence sensitivity and precision and will lead to inaccurate results if the pH and molality in the sample tubes differ from those of the standards. Non-specific interference is a particular problem with urine because of its high content of urea and sodium chloride. (iv) Specific inhibition of binding. As emphasised earlier, an antibody binds to the anti- genic determinant of the drug. Any compound, part of whose molecule is identical with or closely similar to this determinant, will compete for antibody binding sites and influence the results. VC’ith drug assays, specific interference may arise from the presence of an endogenous structural analogue (for example, cholesterol and cortisol could affect the results obtained in an assay for prednisolone) or, more commonly, from one or more metabolites of the drug or from another structurally related drug.Usually these cross-reactions can be predicted but, occasionally, totally unrelated drugs can affect an assay, for example salicylates, dia- zepam (7-chloro-2,3-dihydro-l-methyl-5-phenyl-lH-1,4-benzodiazepin-2-one) and phenytoin compete for binding sites on thyroxine-binding globulin and thereby affect competitive protein binding assays for the thyroid hormones. As discussed earlier, an antiserum raised against an immunogen usually binds to, and therefore shows specificity for, that part of the hapten molecule furthest from its site of conjugation to the carrier protein.In barbiturate radioimmunoassays, the protein is com- monly attached at the 5-position and resultant antibodies are selective only for the six- membered ring and are not affected by any change of groups in the 5-position. Thus, in one assay using [14C]pentobarbitone as the radioactive label, the antibody did not distinguish between barbitone (5,5-diethylbarbituric acid), phenobarbitone (5-ethyl-5-phenylbarbituric (ii) Competition for the labelled drug. (iii) Non-specific inhibition of binding.238 LANDON AND MOFFAT: THE RADIOIMMUNOASSAY Analyst, VoZ. 101 acid) or pentobarbitone [&ethyl-&( 1-methylbuty1)barbituric acid] and would not be expected to distinguish between pentobarbitone and its 5-alkyl hydroxylated metabolites. The antibody did, however, distinguish between barbiturates that differ in the 1-position, such as hexobarbitone [5-( l-cyclohexen-l-yl)-l,5-dimethylbarbituric acid], or molecules in which the oxygen atom in the 2-position is changed to a sulphur atom, such as thiopentone (the sulphur analogue of pentobarbitone). Such an assay is not significantly influenced by drugs with either a reduced carbonyl group in the 6-position (e.g., thymine) or those with five-membered rings (e.g., phenytoin).Similarly, antibodies raised against a phenytoin : protein conjugate are selective to drugs with five-membered rings and do not bind those which contain six- membered rings. The radioimmunoassay for morphine, reported by Spector and Parker,6s12 lacked speci- ficity, being influenced equally by morphine, codeine, normorphine and diamorphine, when [7,8-3H]dihydromorphine was used as the radioactive label.Nonetheless, this assay is very useful for screening samples and enables those samples which are positive to be investi- gated further, using a more specific assay. Other examples in which the carrier blocks part of the drug molecule with a resultant loss in specificity can be cited for the cardiac glycosides. The immunogen used to produce the antibodies employed in one diagnostic kit for digoxin was prepared by coupling the aglycone (digoxigenin) a t the 3-position via a molecule of succinic acid. Thus, the three sugar mole- cules attached at the 3-position in digoxin do not influence the assay, which can be used for both the aglycone and digoxin.Similarly, lanatoside C and deslanoside, which are two different derivatives of digoxigenin and which differ from digoxin only in the sugar residues, are also determined by the assay, whereas digitoxigenin and ouabagenin, which differ in the steroid part of the molecule, are much less effective. The specificity of a radioimmunoassay can be improved by prior extraction or chromato- graphy of the drug in order to separate it from its metabolites and other drugs or naturally occurring compounds that might interfere. Another technique to help ensure specificity is to employ two antisera each directed against different parts of the drug. This approach has been applied to the identification of lysergic acid diethylamide, where antibodies have been made by using immunogens formed from protein conjugates joined at the 1-position and at the primary amide of the d r ~ g , ~ ~ * ~ ~ both of which have very different selectivities for its various analogues.Precision and Accuracy The precision and accuracy of most radioimmunoassays do not approach those of a good chemical method but are comparable with those of chromatographic and better than those of biological procedures. Precision for a drug radioimmunoassay can usually be maintained such that the coefficient of variation both within assays and between assays is less than 10%. A value of 5% is possible, provided that (a) rigid quality control is exercised, (b) the assay is not run at its limits of sensitivity, (c) accurate automatic or semi-automatic sample dilutors and dispensers are employed, (d) several wash steps are included in the separation procedure, (e) counting errors are limited to 1% or less by obtaining at least 10 000 counts per tube and (f) an appropriately programmed desk calculator is used in determining the results.Practicality and Applicability It is very expensive to establish a radioimmunoassay group that is capable of developing new assays because of the need for animal facilities, an isotope laboratory and many costly items of equipment and reagents. In contrast to the high initial costs, however, the cost of assaying a sample can be very low and is dependent predominantly on sample throughput. The main advantage of radioimmunoassay, however, is its applicability to most drugs (and other compounds) over a wide range of concentrations. Once experience has been gained in establishing a radioimmunoassay for one drug, it is usually easy to develop assays for several other drugs using the same equipment and other facilities.Anticipated Trends in Drug Radioimmunoassay Drug Radioimmunoassay Kits A number of commercial kits for the radioimmunoassay of a variety of drugs are now avail-April, 1976 OF DRUGS. A REVIEW 239 able. The particular advantage of a good kit is that it enables a laboratory without ex- tensive facilities and staff experienced in radioimmunoassay to obtain accurate results. Such kits can play an important role in laboratories where the result of, for example, an assay to diagnose digoxin intoxication, is required quickly.The kit should fulfil a number of criteria. Thus, it should contain all necessary reagents, detailed instructions and, when appropriate, the normal range of concentrations of drug expected in the biological fluid to be examined because, as stressed earlier, the actual values obtained will depend on the particular antiserum used and the extent to which it cross-reacts with metabolites. Further, the kit should be relatively inexpensive, rapid and simple to use, require a minimum of equipment and, finally, have been subjected to rigid quality control. It must be realised that, in medical practice, the reporting of an erroneous result may lead to incorrect treatment with serious sequelae for the patient. It is mandatory, therefore, that strict quality control is enforced whether kits or the laboratory’s own reagents are being employed.In this context, we choose to define quality control as “the analytical and other steps which must be taken to ensure that results are of maximum clinical value’’ and have discussed the subject re~ent1y.l~~ Automation The three main reasons for automating drug radioimmunoassays are increasing demands for certain assays, the need to reduce costs to within acceptable limits and the improved precision that results. Nonetheless, many assays should not be automated either because the number of requests is small or because results may be required urgently, as for a patient with suspected digoxin intoxication. Partially automated systems are now widely available and the use of sample dilutors and repeating pipettes markedly improves precision and reduces the tedium of adding microlitre amounts of the various reactants to large numbers of tubes. The separation step, which usually involves centrifugation and aspiration of the supernatant solution, poses a much more difficult problem in the design of fully automated systems, as does the counting step.There are two approaches to full automation, namely continuous flow and discrete analy- sis.136 The former, based on the use of AutoAnalyzer equipment, often allows insufficient time for equilibrium to be reached and carry-over from samples that contain high drug concentrations may prove a problem. More success has been achieved with discrete systems and a fully automated procedure has recently been developed in the UK.13’ Use of Alternative Analytical Techniques Immzcnoradiometric assay This technique differs from radioimmunoassay in that the drug is measured directly by combination with an excess of specific isotopically labelled antibodies.138 The assay involves the preparation of a specific immunoadsorbent by covalently linking the drug to a solid support such as powdered cellulose.This immunoadsorbent is used to extract specific antibodies from the immune serum, which are radioiodinated while still bound to the solid phase and then eluted by decreasing the pH. Samples and standards are incubated for an appropriate time with a constant amount of labelled antibody, which is present inexcess. Unreacted antibody is then removed by adsorption on to an excess of immunoadsorbent and the higher the concentration of drug the greater is the number of counts in the supernatant.The advantages of immunoradiometric assays are the relative ease of labelling antibody molecules, the high specific activity that can be achieved owing to the many tyrosine residues present, the stability of the labelled product and the improved specificity that is possible with a “sandwich” technique.139 In this technique, the drug is incubated with immobilised anti- bodies directed against one part of the molecule and labelled antibodies directed against a second antigenic determinant are then added. This approach also offers the possibility of employing a universal label, the drug and unlabelled specific antibody first being reacted and a second labelled antibody then being added that is directed against IgG molecules of the species providing the initial antiserum.Alternative Labelling procedures The use of radioisotopes for labelling has a number of disadvantages, including the short half-life of A-ray emitting isotopes, the lack of availability of and cost of counting equipment240 LANDON AND MOFFAT: THE RADIOIMMUNOASSAY Analyst, Vol. 101 and the slight health hazard. Increasing attention is therefore being given to the use of alternative labels, such as fluorescent, enzyme or spin-labelled compounds. Recent reports have demonstrated the potential of enzyme labelling for immunoassay purposes. Thus assays have been developed for a variety of protein hormones, one of which has a sensitivity equal to that of radioimmunoassay. This approach has been applied to steroids and other haptens14*J41 and Rubenstein et al.142 have reported an assay for morphine employing a morphine: lysozyme conjugate.This assay offers the particular attribute of not requiring a separation step, as antibody binding of the labelled conjugate blocks its enzymic activity, presumably by steric hindrance of the enzyme - substrate reaction. Any of the conjugation reactions discussed earlier can be used, provided that they permit re- tention of the activity of the enzyme, and the ultimate sensitivity of the assay will depend on the catalytic constant of the enzyme and the sensitivity and accuracy with which its biological activity can be measured. Two commercial kits, for phenytoin and phenobarbitone, have recently become available that permit the assay of serum levels in a few minutes using simple equipment.Another, similar, approach involves the attachment to the drug of a molecule containing a free radical (“spin labelling”). In the unbound fraction, the spin-labelled hapten gives characteristic sharp peaks when subjected to electron spin resonance spectrometry, whereas these peaks are flattened once the molecule is bound by antibody.143 Again, separation of the bound and free fractions is unnecessary, but the technique requires expensive equipment, lacks sensitivity and is difficult to make quantitative in its present stage of development. Non-labelled immunoassay procedures A further approach, applicable to the measurement of haptens at the nanograms and micro- grams per millilitre levels, avoids the use of labelling while still taking advantage of the specificity of the antigen - antibody reaction.One line of research is based on the light- scattering effect, as determined by nephelometry, of antigen - antibody aggregates, which is dependent on the size rather than the number of particles. In one such assay, the antigen is made to react directly with antibody while in another the antigen is first made to react with an excess of antibody and the remaining “unused” antibody is then determined by the nephelometric effect of incubating it with the original immunogen.144 Conclusion There are indications that the unique combination of sensitivity, specificity and wide applicability that characterises radioimmunoassay will be employed to an increasing extent in the determination of drug levels in biological fluids.Great care is essential if meaningful results are to be obtained and each antiserum must be assessed in relation to its cross-reactivity with the appropriate drug, its stereoisomers and metabolites and exogenous and endogenous analogues. The labelled drug must also be assessed repeatedly by both physicochemical and immunological means. Because of the risk that erroneous results may be obtained by a radioimmunoassay, it is helpful to employ other physicochemical techniques for drug assay (such as mass spectrometry and gas - liquid chromatography) as a regular check. This is consistent with the need to compare the results obtained by a radioimmunoassay of a hormone with those obtained by an appropriate bio- assay.1. 2. 3. 4. 6. 6. 7. 8. 9. 10. References Yalow, R. S., and Berson, S. A,, J . Clin. Invest., 1960, 39, 1157. Oliver, G. C., Parker, B. M., Brasfield, D. L., and Parker, C . W., J . Clin. Invest., 1968, 47, 1035. Butler, G. C., Harington, C. R., and Yuill, M. E., Biochem. J., 1940, 34, 838. Weiner, L. M., Rosenblatt, M., and Howes, H. A., J . Immun., 1963, 90, 788. Girard, J. P., Hildebrant, F., and Favre, H., Helv. Med. Acta, 1969, 35, 86. Spector, S., and Parker, C . W., Science, N.Y., 1970, 168, 1347. Wainer, B. H., Fitch, F. W., Fried, J., and Rothberg, R. M., Clin. Immun. Immunopath., 1974, 3, Robinson, J. D., Morris, B. A., and Marks, V., Res. Commun. Chem. Path. Pharmac., 1975, 10, 1. Henderson, G.L., Frincke, J., Leung, C . Y., Torten, M., and Benjamini, E., J . Pharmac. Exp. Thar., Liu, C. T., and Adler, F. L., J . Immun., 1973, 111, 472. 155. 1976, 192, 489.April, 1976 OF DRUGS. A REVIEW 241 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. Adler, F. L., and Liu, C. T., J . Immun., 1971, 106, 1684. Spector, S., J . Pharmac. Exp. Ther., 1971, 178, 253. Van Vunakis, H., Wasserman, E., and Levine, L., J . Pharmac. Exp. Ther., 1972, 180, 514. Gershman, H., Powers, E., Levine, L., and Van Vunakis, H., Prostaglandins, 1972, 1, 407. Wainer, B. H., Fitch, F.W., Rothberg, R. M., and Schuster, C. R., Nature, Lond., 1973, 241, 537. Spector, S., Berkowitz, B., Flynn, E. J., and Peskar, B., Pharmac. Rev., 1973, 25, 281. Miller, C., Leung, C. Y., Nakamura, J., Winters, VV. D., and Benjamini, E., Proc. West. Pharmac.[Soc., Lindquist, C . A., and Spratt, J. L., Fedn Proc. Fedn Am. Socs Exp. Biol., 1973, 32, 719. Catlin, D., Cleeland, R., and Grunberg, E., Clin. Chem., 1973, 19, 216. Berkowitz, B. A., Cerreta, K. V., and Spector, S., Fedn Proc. Fedn A m . Socs Exp. Biol., 1973, 32, Ringle, D. A., and Herndon, Fedn Proc. Fedn A m . Socs Exp. Biol., 1973, 32, 319. Berkowitz, B. A., Cerreta, K. V., and Spector, S., J . Pharmac. Exp. Ther., 1974, 191, 527. Capelle, N., Fournier, E., Kuntz, M. O., and Treich, I., J . Eur. Toxic., 1974, 7, 147.Koida, M., Takahashi, M., and Kaneto, H., Jap. J . Pharmac., 1974, 24, 707. Gross, S. J., Grant, J. D., Wong, S. R., Schuster, R., Lomax, P., and Campbell, D. II., Immunochem- Mul6, S. J., Bastos, M. L., and Jukofsky, D., Clin. Chem., 1974, 20, 243. Morris, B. A., Robinson, J. D., Piall, E., Aherne, G. W., and Marks, V., J . Endocr., 1975, 64, 6p. Mul6, S. J., Whitlock, E., and Jukofsky, D., Clin. Chem., 1975, 21, 81. Williams, T. A., and Pittman, K. A., Res. Commun. Chem. Path. Pharmac., 1974, 7, 119. Van Vunakis, H., Langone, J . J., Riceberg, L. J., and Levine, L., Cancer Res., 1974, 34, 2546. Lewis, J. E., Nelson, J. C., and Elder, H. A., Antimicrob. Agents Chemother., 1975, 7, 42. Hamburger, R. N., Science, N . Y . , 1966, 152, 203. Lewis, J .E., Nelson, J. C., and Elder, H. A,, Nature, Lond., 1972, 239, 214. Mahon, W. A., Ezer, J., and Wilson, T. W., Antimicrob. Agents Chemother., 1973, 3, 585. Berk, L. S . , Lewis, J. L., and Nelson, J. C., Clin. Chem., 1974, 20, 1159. Minshew, B. H., Holmes, R. K., and Baxter, C. R., Antimicrob. Agents Chemother., 1975, 7, 107. Schwenk, R., Kelly, K., Tse, K. S., and Sehon, A. H., Clin. Chem., 1975, 21, 1059. Gleich, G. J., and Stankievic, R. R., Imnzunochemistry, 1969, 6, 85. Karchmer, A. W., Dismukes, W. E., de Riel, K., and Hyslop, N. E., Clin. Res., 1972, 20, 531. Hyslop, N. E., Calvert, J . F., Milligan, R. T., and Kern, I<., Fedn Proc. Fedn Am. Socs Ex$. Biol., Wal, J.-M., and Kann, G., C.R. Hebd. Se'anc. Acad. Sci., Paris, 1975, 280, 2405. Wedum, A. G., J . Infect.Dis., 1942, 70, 173. Queng, C. T., Duber, C. D., and McGovern, J., J . Allergy, 1965, 36, 505. Tigelaar, R. E., Rapport, R. L., Inman, J . K., and Kupferberg, H. J., Clin. Chim. Acta, 1973,43, 231. Cook, C. E., Kepler, J . A., and Christensen, H. D., Res. Commun. Chem. Path. Pharmac., 1973, 5, 767. Montgomery, M. R., Holtzman, J. L., Leute, R. K., Dewees, J. S., and Bolz, G., Clin. Chem., 1975, Robinson, J. D., Morris, B. A., Aherne, G. W., and Marks, V., J . Endocr., 1975, 64, 7p. Glogner, P., Bermeister, P., and Heni, N., Klin. Wochenschr., 1973, 51, 352. Pettinger, W. A., Keeton, K., and Tanaka, K., Clin. Pharmac. Ther., 1975, 17, 146. Jaton, J.-C., and Ungar-Waron, H., Arch. Biochem. Biophys., 1967, 122, 157. Bohuon, C., Duprey, F., and Boudene, C., Clin.Chim. Acta, 1974, 57, 263. Levine, L., and Powers, E., Res. Commun. Chem. Path. Pharmac., 1974, 9, 543. Raso, V., and Schreiber, R., Cancer Res., 1975, 35, 1407. Spector, S., and Flynn, E. J., Science, N.Y., 1971, 174, 1036. Flynn, E. J., and Spector, S., J . Pharmac. Exp. Ther., 1972, 181, 547. Chung, A., Kim, S. Y., Cheng, L. T., and Castro, A., Experientin, 1973, 29, 820. Satoh, H., Kuroiwa, Y., and Hamada, A., J . Biochem., Tokyo, 1973, 73, 1115. Satoh, H., and Kuroiwa, Y., J . Biochem., Tokyo, 1974, 76, 1293. Satoh, H., Kuriowa, Y., Hamada, A., and Uematsu, T., J . Biochenz., Tokyo, 1974, 75, 1301. Russell, A. S., and Ziff, M., Clin. Exp. Immzm., 1968, 3, 901. Smith, T. W., Butler, V. P., and Haber, E., New Engl. J . Med., 1969, 281, 1212. Lesne, M., Archs I n t .Pharmacodyn. Thbr., 1972, 199, 206. Seldon, R., and Smith, T. W., Circulation, 1972, 45, 1176. Teale, J . D., Forman, E. J., King, L. J., and Marks, V., Proc. SOC. Analyt. Chem., 1974, 11, 219. Teale, J. D., Forman, E. J., King, L. J., and Marks, V., Lancet, 1974, ii, 553. Teale, J. D., Forman, E. J., King, L. J., and Marks, V., Nature, Lond., 1974, 249, 164. Gross, S. J., Soares, J. R., Wong, S.-L. R., and Schuster, R. E., Nature, Lond., 1974, 252, 581. Teale, J. D., Forman, E. J., King, L. J., Piall, E. M., and Marks, V., J . Pharm. Pharmac., 1975, 27, Teale, J. D., Clough, J. M., Piall, E. M., King, L. J., and Marks, V., lies. Commun. Chem. Path. Marks, V., Teale, D., and Fry, D., BY. Med. J., 1975, iii, 348. Van Vunakis, H., Farrow, J. T., Gjika, H.B., and Levine, L., Proc. Natn. Acad. Sci. U.S.A., 1971, Castro, A., Grettie, D. P., Bartos, F., and Bartos, D., Res. Commun. Chem. Path. Pharmac., 1973, 6, Taunton-Rigby, A,, Sher, S. E., and Kelley, P. R., Science, N.Y., 1973, 181, 165. 1973, 16, 25. 719. istry, 1974, 11, 453. 1975, 34, 975. 21, 221. 465. Pharmac., 1975, 11, 339. 68, 1483. 879.242 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. LANDON AND MOFFAT: THE RADIOIMMUNOASSAY Analyst, Vd. 101 Loeffler, L. J., and Pierce, J. V., J . Pharm. Sci., 1973, 62, 1817.Wingeleth, D. C., Makowski, A., and Teitelbaum, D. T., Clin. Chem., 1974, 20, 870. Van Vunakis, H., Bradvica, H., Benda, P., and Levine, L., Biochern. Pharmac., 19G9, 18, 393. Schnoll, S. H., Vogel, W. H., and Odstrchel, G., Fedn PYOG. Fedn Am. Socs Exp. Biol., 1973, 32, Riceberg, L. J., Van Vunakis, H., and Levine, L., Anal. Biochem., 1974, 60, 551. Kido, Y., Nagamatsu, K., and Ishizeki, C., J . Pharm. Soc. Japan, 1974, 94, 1290. Langone, J . J.. and Van Vunakis, H., Res. Commun. Chem. Path. Pharmac., 1975, 10, 163. Centeno, E. R., Johnson, W. J., and Sehon, A. H., Int. Archs Allergy Appl. Immun., 1970, 37, 1. Horowitz, P. E., and Spector, S., J . Pharmac. Exp. Ther., 1973, 185, 94. Morris, H. G., de Roche, G., and Caro, C. M., Steroids, 1973, 22, 445. Dumasia, M.C., Chapman, D. I., Moss, M. S., and O’Connor, C., Biochem. J., 1973, 133, 401. Lee, Y . S., and Ohsowa, N., Endocr. Jap., 1974, 21, 481. Meikle, A. W., Lagerquist, L. G., and Tyler, F. H., Steroids, 1973, 22, 193. Hichens, M., and Hogans, A. F., Clin. Chem., 1974, 20, 266. Kundu, N., Steroids, 1974, 23, 155. Rao, P. N., Steroids, 1974, 23, 173. Rao, P. N., de la Pena, A., and Goldzieher, J. W., Steroids, 1974, 24, 803. Colburn, W. A., Steroids, 1975, 25, 43. Cornette, J. C., Kirton, K. T., and Duncan, G. W., J . Clin. Endocr., 1971,33,459. Royer, M. E., KO, H., Campbell, J . A., Murray, H. C., Evans, J. S., and Kaiser, D. G., Steroids, Kundu, N., Steroids, 1973, 22, 327. Colburn, W. A., and Buller, R. H., Steroids, 1973, 22, 687. Nygren, K. G., Lindberg, P., Martinsson Basu, W.T. K., and Johansson, E. D. B., Contraception, Cameron, E. H. D., Morris, S. E., and Nieuweboer, B.. J . Endocr., 1974, 61, 39. Warren, R. J., and Fotherby, K., J . Endocr., 1974, 62, 605. Colburn, W. A., Steroids, 1974, 24, 95. Cheng, L. T., Kim, S. Y., Chung, A., and Castro, A., Febs Lett., 1973, 36, 339. Hooker, S. Is., and Boyd, W. C., J . Immun., 1940, 38, 479. Langone, J , J., Gjika, H. B., and Van Vunakis, H., Biochemistry, N.Y., 1973, 12, 5025. Langone, J. J., and Van Vunakis, H., Res. Commun. Chem. Path. Pharmac., 1975, 10, 21. Cernosek, S. F., Langone, J. J., Gjika, H. B., and Van Vunakis, H., Fedsz Proc. Fedn Am. Socs Exp. Haines, C. F., Mahajan, D. H., Miljkovic, D., Miljkovic, hi., and Vesell, E. S., Clin. Pharm. Tlzer., Matsukura, S., Sakamoto, N., Imura, H., Matsuyama, H., Tamada, T., Ishiguro, T., and Muranaka, Langone, J .J., Franke, J., and Van Vunakis, H., Arch. Biochem. Biophys., 1974, 164, 536. Dixon, W. R., Earley, J., and Postma, E., J . Pharm. Sci., 1975, 64, 937. Spector, S., in Forrest, I. S., Editor, “The Phenothiazines and Structurally Related Drugs,” Raven Marks, V., Morris, B. A., and Teale, J. D., BY. Med. Bull., 1974, 30, 80. Peskar, B., and Spector, S., J . Plzarmac. Exp. Ther., 1973, 186, 167. Michiels, L. J. M., Heykants, J. J. P., Knaeps, A. G., and Janssen, P. A. J., Life Sci., 1975, 16, 937. Kirkham, K. E., and Hunter, W. M., Editors, “Radioimmunoassay Methods,” Churchill Livingstone, Odell, W. D., and Daughaday, W. H., Editors, “Principles of Competitive Protein-binding Assays,” BY. Med. Bull., 1974, 30, No. 1. Freund, J., Am. J . Clin. Path., 1951, 21, 645. Miles, L. E. M., and Hales, C. N., Biochem. J., 1968, 108, 611. Erlanger, B. F., Pharmac. Rev., 1973, 25, 271. Hurn, B. A., and Landon, J.. in Kirkham, K. E., and Hunter, W. M., Editors, “Radioirnmunoassay Vaitukaitis, J., Robbins, J. B., Nieschlag, E., and Ross, G. T., J . Clin. Endocr. Metab., 1971, 33, 988. Scatchard, G., Ann. N.Y. Acad. Sci., 1949, 51, 660. Hunter, W. M., and Greenwood, F. C., Nature, Lond., 1962, 194, 495. Chamberlain, D. A., White, R. J., Howard, M. R., and Smith, T. W., BY. Med. J . , 1970, iii, 429. Evered, D. C., Chapman, C., and Hayter, C. J., BY. Med. J . , 1970, iii, 427. Oliver, G. C., Parker, B. M., and Parker, C. W., Am. J . Med., 1971, 51, 186. Hoeschen, R. J., and Proveda, V., Can. Med. Ass. J., 1971, 105, 170. Herbert, V., Iau, K. S., Gottlieb, C. W., and Bleicher, S. J., J . Clin. Endocr. Metab., 1965, 25, 1375. Lader, S., Bye, A,, and Marsden, P., Eur. J . Clin. Pharmac., 1972, 5, 22. Meade, R. C., and Kliest. T. J., J . Lab. Clin. Med., 1972, 80, 748. Barrett, M. J., and Cohen, P. S., Clin. Chem., 1972, 18, 1339. Hales, C. N., and Randle, P. J., Biochem. J., 1963, 88, 137. Horgan, E. D., and Riley, W. J., Clin. Chem., 1973, 19, 187. Greenwood, H., Howard, M., and Landon, J., J . Clin. Path., 1974, 27, 490. Ekins, R. P., BY. Med. Bull., 1974, 30, 3. Challand, G., Goldie, D., and Landon, J.. BY. Mad. Bull., 1974, 30, 38. 719. 1974, 23, 713. 1974, 9, 265. Biol., 1973, 32, 511. 1974, 16, 1083. H., Biochem. Biophys. Res. Commun., 1975, 64, 574. Press, New York, 1974, p. 363. Edinburgh, 1971. Lippincott, Philadelphia, 1971. Methods,” Churchill Livingstone, Edinburgh, 1971, p. 121.April, 1976 OF DRUGS. A REVIEW 243 136. Landon, J., Rees, L., and Nye, L., Br. J . Hosp. Equip., 1975, 5, 29. 137. Butler, H. A., Bagshawe, K. D., Marchant, D., and Verdin, A., in Kirkham, K. E., and Hunter, W. M., 138. Woodhead, J. S., Addison, G. M., and Hales, C. N., BY. Med. Bull., 1974, 30, 44. 139. Addison, G. M., and Hales, C. N., Horm. Metab. Res., 1971, 3, 59. 140. Van Weemen, B. K., and Schuurs, A. H. W. M., Febs Lett., 1972, 24, 77. 141. Landon, J., Crookall, J., and McG;Segor, A., in Cameron, E. H. D., Hillier, S. G., and Griffiths, K., 142. Rubenstein, K. E., Schneider, R. S., and Ullman, E. F., Biochsm. Biophys. Res. Commun., 1972, 47, 143. Leute, R. K., Ullman, E. F., and Goldstein, A., J . Am. Med. Ass., 1972, 221, 1231. 144. Cambiaso, C. L., Biccomi, H. A., Masson, P. L., and Hersmans, J. F., J . Immun. Meth., 1974, 5, 293. Received October 24th, 1976 Accepted November 14th, 1976 Editors, “Radioimmunoassay Methods,” Churchill Livingstone, Edinburgh, 1971, p. 635. Editors, “Steroid Immunoassay, Alpha Omega, Cardiff, 1974, p. 183. 846.

ISSN:0003-2654    

DOI:10.1039/AN9760100225

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

244 Analyst, April, 1976, Vol. 101, pp. 244-254 Studies on the Oxidation of Amitriptyline B. Caddy, F. Fish and J. Tranter Division of Pharmacognosy and Forensic Science, School of Pharmaceutical Sciences, University of Strathclyde, Glasgow, G1 1XW The quantitative determination of amitriptyline extracted from biological fluids has been achieved using an oxidation procedure. By use of selective oxidation methods a knowledge of the levels of 10-hydroxylated metabolites of amitriptyline can be obtained. Some evidence is presented that dehydro- amitriptyline, or its demethylated derivatives, are natural metabolites of amitriptyline in man. Amitriptyline and its metabolite nortriptyline are both used as antidepressants and also as sympatholitics, antihistaminics and as mild tranquillisers.Both drugs are rapidly absorbed into the tissues and the serum concentration is never very high. The metabolism of both drugs is extensive and complex. Bio-transformations following nortriptyline administration appear to be well elucidated, the major metabolite reported in man, arising from normal therapeutic doses, being lO-hydr~xynortriptyline.~-~ de Leenheer and Heyndrickxl found other metabolites, but none corresponding to demethylnortriptyline, in contradiction to the work of Hammar et aL2 who reported finding this last compound in urine. The results obtained by Hammar et al. have since been supported by other investigations.394 Hucker and Porter5 reported that very little amitriptyline was excreted unchanged and they were able to isolate two metabolites which, they suggested, might be demethylated derivatives conjugated with glucuronic acid.From studies with rabbits, Corona and FacinoG identified 1 0-hydroxyamitript yline, nortript yline, 10-h ydroxynort ript yline, 10,ll -hydroxyamitript yline and 10,ll-hydroxynortriptyline in urine but they observed considerable fluctuations in the concentrations in different animals. Eschenhof and Rieder' obtained results similar to the above with rats and humans but, in addition, found the metabolites NN-didemethylamitriptyline and the N-oxide of amitripty- line. They also indicated that 10-hydroxy metabolites were not present in human urine. In rat liver perfusion studies, Braithwaite* has found that amitriptyline is rapidly converted into nortrip t yline, 10-h ydroxyamit ript yline and 10-h ydr oxynortript yline .He found that, apart from nortryptyline, the main metabolites of amitriptyline in man appeared to be the 10-hydroxynortriptyline isomers. In addition to the compounds reported by Braithwaite, J~rgensen~ has identified amitriptyline N-oxide and demethylnortriptyline in the urine of patients undergoing amit ript yline therapy . Most authors found considerable conjugation (usually as glucuronide) of the metabolites. The low levels of unchanged drug5J0J1 and the presence of many metabolites require that a sensitive and specific analytical procedure be used for the determination of amitriptyline extracted from biological fluids. For this reason many workers have used gas-chromatographic methods,ll@ although some advocate the use of spectrophotometric procedures.13-15 Both methods of analysis often require a preliminary clean-up, with all the problems and losses that such procedures entail. Wallace and DahP published an oxidative procedure for amitriptyline using alkaline permanganate as oxidant and oxidative assays for some tricyclic drugs have also been rep0rted.l' This paper concerns the use of selective oxidation for the quantitative and qualitative analysis of amitriptyline and its metabolites. The metabolism of amitriptyline is not so well understood as that of nortriptyline.Experimental Oxidation of Amitriptyline with Potassium Dichromate at Room Temperature18 To 1 ml of a solution of amitriptyline hydrochloride (containing 10-100 pg ml-1) in water, contained in a 40-ml glass-stoppered test-tube, add 4 ml of potassium dichromate solution (4% m/Y), 6 ml of 66% sulphuric acid and 5 ml of spectroscopic-grade hexane.Tilt-shake the tube for 60min a t room temperature, then remove an aliquot of the hexane layer andCADDY, FISH AND TRANTER 245 record its ultraviolet spectrum over the range 220400 nm in a cell of l-cm path length with hexane as reference. The above concentrations of reagents are those experimentally deter- mined (Tables I and 11) to give the best results. TABLE I DILUTIONS USED FOR CONTROLLED OXIDATIONS OF AMITRIPTYLINE WITH DICHROMATE AT DIFFERENT CONCENTRATIONS Drug solution*/ ml 1 0 1.07 1 .o 1 .o 1 .o 1 .o 1 .o Strong acid soh tiont/ ml 4.0 6.07 6.5 7.0 8.0 12.5 15.0 Water/ ml 3.0 1 0l-f 0.5 0 0 1.5 0 Dilute acid soh tion / ml 4.0 4.0 4.0 3.0 0 0 4.07 Dichromate solution§/ ml 4.0 4.07 4.0 4.0 4.0 5.0 5.0 Approximate total volume/ ml 16 ;? 16 16 20 21 Approximate normality of acid (4 9.07 6.0 9.8 10.5 12.0 15.0 17.0 * Two different concentrations of aqueous amitriptyline hydrochloride solution : 20 and 100 pg m1-1.t 66% V/V sulphuric acid. $ Approximately 1 N sulphuric acid. 9 4% mlV dichromate solution. 7 Data used for preparation of standard calibration graphs. In addition, the experiment that involved the use of conditions giving a final acid concen- tration of 9 N was repeated on a 20 pg ml-l solution of amitriptyline, the hexane volume being reduced to 2 ml and the ultraviolet spectrum being measured in a cell of 2-cm path length. TABLE I1 ABSORBANCE OF HEXANE LAYER FOLLOWING THE OXIDATION OF DIFFERENT CONCENTRATIONS OF AMITRIPTYLINE FOR 60 MIN AT ROOM TEMPERATURE WITH DICHROMATE SOLUTION USING CONTROLLED ACID CONCENTRATIONS Absorbance a t 264 nm r A \ Extracts containing 100 pg ml-l but Initial extracts corrected to be Sulphuric acid, Approximate containing equivalent to 66%*/ml acid normality 20 pg ml-l 20 pg ml-l 8.0 12.0 0.51t 0.49t 7.0 10.5 0.70t 0.70f 6.5 9.8 0.77 0.76 6.0 9.0 0.76 0.76 4.0 6.0 0.55 0.53 * Volume of strong acid used in 16 ml of reaction mixture.t Also exhibited a distinct maximum at 250 nm (anthraquinone). Preparation of standard calibration graphs The calibration graphs were prepared from data obtained as described above, using a final potassium dichromate solution concentration of 1% and an acid concentration of 9 N (Table I), on solutions of amitriptyline, nortriptyline and 10,l l-dehydroamitriptyline of various con- centrations (2, 5, 10, 15 and 20 pg ml-l).Preparation of Calibration Graphs for Drugs Extracted from Blood and Urine Add suitable volumes of an aqueous solution of the drug salt to 10-ml samples of drug-free urine and citrated whole blood such that the final concentrations of drug salt in the body fluid are in the range 1.0-10.0 pg ml-l. Place 10 ml of each urine sample in separate 500-ml reagent bottles together with 5 ml of 1 N sodium hydroxide solution and 100 ml of diethyl246 CADDY, FISH AND TRANTER: STUDIES ON Analyst, Vol. 101 ether. Roll the bottles on a roller shaker for 20 min at approximately 40 rev min-l, separate the ether layers, filter them through Whatman No.1 filter-papers, measure the volume and then transfer the ether layers to a second series of reagent bottles, each containing 5 ml of 1 N sulphuric acid. Roll this second series of bottles for 20 mh, separate the aqueous phases and evaporate 4 ml of each aqueous phase under reduced pressure at 50-60 "C for 3 min in order to remove the last traces of organic solvent. Oxidise the resulting acidic solution as described for the solutions of drug salts used in preparing calibration graphs for aqueous solutions. The extraction of drugs from blood is performed in a similar manner, except that the blood is made alkaline by the addition of 2 or 3 ml of concentrated ammonia solution. Oxidation of Amitriptyline with Potassium Permanganate at Room Temperature These oxidations were performed in a manner similar to those involving the use of potassium dichromate but using 100 pg of amitriptyline hydrochloride, 1 yo potassium permanganate in x N sodium hydroxide solution (see Table 111) and 5 ml of hexane.The ultraviolet spectrum of the hexane layer was recorded after 75 min in a cell of 1 cm path length. TABLE I11 DILUTIONS USED FOR THE OXIDATION OF AMITRIPTYLINE WITH PERMANGANATE "'g solution*/ ml Alkaline solutiont/ ml 1.0 15.0 1.0 7.5 1.0 5.0 1 .o 3.0 1.0 1.0 Potassium permanganate Water1 solution:/ ml ml 0 4.0 6.5 6.0 9.0 5.0 11.0 5.0 13.0 5.0 Approximate total volume/ ml 20 20 20 20 20 Approximate normality of the alkali 12.0 6.0 4.0 2.4 0.8 (4 * Aqueous, 100 pg m1-l solution of amitriptyline hydrochloride.t 16 N sodium hydroxide solution. $ 4% solution. High-temperature Oxidation of Amitriptyline with Potassium Dichromatelg Place a suitable amount of amitriptyline hydrochloride (1 ml of an aqueous solution containing 10-100 pg ml-l) in a 250-ml conical flask containing a 1-in, glass-coated magnetic follower. Add a mixture of 5 ml of 4% potassium dichromate solution and 15 ml of 12 N sulphuric acid, followed by 5 ml of spectroscopic-grade hexane. Next, heat the mixture at 60-70 "C under reflux, with constant stirring, for approximately 40 min. Cool the flask and, with the condenser still in position, rinse the apparatus with approximately 5ml of water, then transfer the contents of the conical flask to a 40-ml glass-stoppered test-tube. Place an aliquot of the hexane layer in a cell of l-cm path length and determine the ultra- violet spectrum between 225 and 400 nm, using hexane as the reference.High-temperature Oxidation of Amitriptyline with Potassium Permanganate This is carried out as described immediately above for potassium dichromate but using an oxidant prepared by mixing 5 ml of a 4% aqueous solution of potassium permanganate with 5 ml of an 8 N aqueous solution of sodium hydroxide. Amitriptyline Drug Regimen One half of a normal therapeutic dose of amitriptyline hydrochloride (two 25-mg Tryptizol tablets per day) was administered for two consecutive days to a subject known to be taking no other drugs. The total urine excreted over a period of 5 d, including the 2 d of drug administration, was collected and bulked at timed intervals (Table IV).A urine sample was collected before drug administration for the purposes of blank determinations. In a second experiment, three subjects (FD, SB, DR) were given one 25-mg TryptizolApril, 1976 THE OXIDATION OF AMITRIPTYLINE 247 tablet and a sample of their urine was collected several hours after administration. Subject FD was undergoing antihistamine therapy during the trial (two 50-mg Fabahistin tablets daily), but the other subjects were taking no other drugs. TABLE IV DATA FOR AMITRIPTYLINE ADMINISTRATION, URINE SAMPLE COLLECTION AND BULKING The first tablet was taken a t 10.00 a.m. and the urine was bulked a t about midnight each day, excepting the final day, when the last sample was collected at 5.00 p.m. Urine sample code Time/h 0 6 14 24 30 38 62 86 103 J1 J2 J3 J4 J5 Number of Volume of Tryptizol tablets bulked urine/ taken ml 1 1 1 1 1 640 1 930 1410 1380 1140 Extraction and Oxidative Assay of Urine Samples Following Oral Drug Adminis- tration The extraction and oxidation procedures used were those indicated above.Both non-hydrolysed and hydrolysed urine samples were used in 10-ml volumes, except in the oxidation of non-hydrolysed urine by cold dichromate, when 50 ml were employed. In all of the dichromate oxidations 2.0 ml of hexane were used to extract the ultraviolet absorbing material but 2.5 ml of hexane were used in the permanganate procedure. Hydrolysis of Urine Samples ing solution was heated on a boiling water bath for 1 h. concentration of the acid was approximately 20%.reflux for 1 h. (a). (b). The urine was adjusted to pH 1.0 with concentrated hydrochloric acid and the result- Sufficient concentrated hydrochloric acid was added to the urine sample such that the The resulting solution was heated under The conditions were as in (a) but the hydrolysis was carried out under reflux. (c). In each instance the hydrolysate was cooled, made alkaline with 12 N sodium hydroxide solution and extracted as described above. Chromatographic Analysis of Urine Samples Following Oral Drug Administration A number of hydrolysed and non-hydrolysed 50-ml urine samples were made alkaline, each was extracted with 50 ml of ether and the ether evaporated to dryness under reduced pressure at 50 "C. The residue was taken up in 0.5 ml of ether and this solution used for thin-layer chromatographic analysis.The ethereal solutions that remained after thin-layer chromatography had been carried out were evaporated to dryness, the residues taken up in 0.2 ml of hexane and these solutions used for gas - liquid chromatography. T hin-day er chromatography Polygram SIL N HR plates (0.2-mm layer thickness; 20 x 20 cm dimensions) were used as the adsorption phase, together with the development solvent acetone - toluene - ethanol - ammonia solution (sp. gr. 0.88) (40 + 40 + 12.5 + 2.5 V/V). Iodoplatinate spray was used as the detection agent.248 CADDY, FISH AND TRANTER: STUDIES ON Analyst, Vol. 101 Gas - liquid chromatography A Varian Aerograph 1400 gas chromatograph fitted with a flame-ionisation detector was used and two sets of conditions were as follows.A 6 foot x Qin i.d. glass column, packed with 10% Apiezon L on Chromosorb G (60-80 mesh), was prepared, and operated at 250 "C with a nitrogen flow-rate of 50 ml min-l, the injection block and detector block being maintained a t 270 "C. A 6 foot x Q in i.d. glass column, packed with 1% OV-25 on Chromosorb G (60-80 mesh), was prepared, and operated at 180 "C with a nitrogen flow-rate of 50 ml min-l, the injection block being maintained at 200 "C. 1. 2. Hydrolysis and Gas-chromatographic Analysis of Amitriptyline and Some of Its Metabolites This analysis was performed by subjecting aqueous solutions of amitriptyline and its major metabolites to the same hydrolysis, extraction and chromatographic processes as are detailed above. Results and Discussion Amitriptyline, nortriptyline and the hydroxyamitriptylines are all oxidised to anthra- quinone by alkaline permanganate while only those compounds which are unsaturated between the 10 and 11 positions are oxidised by acidic dichromate solution.17 In particular, no hexane-soluble material showing ultraviolet absorption is obtained from the oxidation of 10-hydroxyamitriptyline with acidic dichromate.Oxidation of Amitriptyline with Potassium Dichromate at Room Temperature Preliminary studies indicated that the nature of the product of these oxidations appeared to be dependent upon the acid concentration. Solutions of dichromate in approximately 17 N sulphuric acid yielded anthraquinone exclusively (adduced from ultraviolet spectral') but in much lower yield than hot alkaline permanganate (0.33 and 1.13 absorbance units, respec- tively, for concentrations equivalent to 20 pg ml-l of amitriptyline hydrochloride in hexane).At acid concentrations that were 15 N with respect to sulphuric acid, the spectrum indicated the formation of both 10,l l-dihydro- (5H)-dibenzo [a,d]cyclohepten-5-one (DiHDBCH) and anthraquinone, although the latter appeared to be present only in a low concentration. When using 9 N sulphuric acid, DiHDBCH was the sole product but the optimum reaction time was about 1 h (Table V). TABLE V ABSORBANCE OF HEXANE LAYER FOLLOWING THE OXIDATION OF AMITRIPTYLINE WITH DICHROMATE FOR VARYING TIMES AND USING VARIOUS ACID CONCENTRATIONS AT ROOM TEMPERATURE Approximate acid normality Timelmin 30 9 60 90 15 30 17 30 60 90 ~m,x./nm 264 264 264 250, 264 250 250 250 Absorbance* 0.62 0.76 0.77 0.42, 0.49 0.30 0.33 0.33 * Amitriptyline hydrochloride concentration equivalent to 20 pg ml-l in solution in hexane for each determination.It is apparent (from Table 11) that fairly rigorous control of acid concentration is necessary in order to achieve the minimum reaction time without the formation of anthraquinone. Because determinations of amitriptyline and its metabolites extracted from biological samples will necessitate the presentation of the drug for oxidation in 4.0 ml of dilute acid solution, this factor was taken into account when making dilutions of the strong acid for the remainder of the experiments. As a considerable reduction in volume occurs when mixing concentratedAfirzt, 1976 THE OXIDATION OF AMITRIPTYLINE 249 sulphuric acid and water, and undoubtedly some smaller reductions in volume will also occur when further dilutions are carried out, the normalities quoted are only approximations.However, following the procedure outlined, satisfactory reproducibility can be obtained (Table VI). To eliminate anthraquinone formation the adopted procedure thus utilised 6.0 ml of the strong acid, giving a final acid concentration of approximately 9 N in the reaction mixture. Using this acid concentration a reaction time of 1 h was sufficient to give the maximum DiHDBCH formation (Table 11). TA4BLE VI ABSORBANCES IN HEXANE OF VARIOUS CONCENTRATIONS OF AMITRIPTYLINE AND Drug Amitriptyline Nortriptyline Amitriptyline Nortriptyline Amitriptyline Nortriptyline Amitriptyline Amitriptyline Source Water Water Watcr Water Water Water Blood Urine NORTRIPTYLINE FOLLOWING OXIDATION Absorbance at Amax.of final drug concentration in hexanet f L 7 Oxidation 20 15 10 5 2 procedure* A,,,./nm pg ml-1 pg ml-1 pg ml-1 pg ml-f pg ml-' A 250 1.13 0.88 0.55 0.32 0.12 (57) (59) (55) (64) (60) A 250 1.13 0.86 0.56 0.31 0.11 (57) (57) (56) (62) (55) B 264 0.87 0.66 0.44 0.21 0.09 (44) (44) (44) (42) (45) B 264 0.91 0.67 0.45 0.22 0.10 (46) (45) (45) (44) (50) C 264 0.83 0.62 0.42 0.21 0.08 (42) (41) (42) (42) (40) C 264 0.84 0.63 0.42 0.22 0.09 (42) (42) (42) (44) (45) C 264 0.57 0.46 0.30 0.14 (29) (31) (30) (28) C 264 0.63 0.45 0.33 0.16 (32) (30) (33) ( 32) * Procedure A, alkaline permanganate at about 60 "C; B, acidic dichromate at about 60 "C and C, acidic t All values are means of at least two determinations. Values in parentheses = Concentration/pg ml-i dichromate a t room temperature. Absorbance x lo8 Because, following the administration of therapeutic doses, blood and urine levels of amitriptyline are apt to be low, the above procedure was carried out on both large (5.0 ml of hexane, 1-cm path length cells) and small (2.0 ml of hexane, 2-cm path length cells) scales.Oxidation of Amitriptyline with Potassium Permanganate at Room Temperature The oxidation of amitriptyline at room temperature by use of alkaline permanganate solutions (Table 111) produced higher yields of anthraquinone as the normality of the base was decreased from 12 to 2.4 (the absorbance values at 250nm at these normalities were 0.20 and 0.58, respectively).At a normality of 0.8 the yield was similar to that at 2.4. However, even after 75 min the maximum absorbance obtained from these procedures was low (0.58 absorption unit) compared with the absorbance following the hot oxidation pro- cedurels (1.13 absorbance units) and so these experiments were discontinued. Precision of the Oxidation Procedures When amitriptyline hydrochloride solutions (20 pg ml-1) were subjected to 15 replicate determinations by both hot permanganate and cold dichromate oxidation, the results exhibited coefficients of variation of 7.4 and 1.9%, respectively.250 CADDY, FISH AND TRANTER: STUDIES ON Analyst, VoE. 101 Calibration Graphs Obtained by the Oxidation of Solutions of Amitriptyline and Nortriptyline Hydrochlorides Oxidation with acidic dichromate Very good linearity over the range 2-20 pg ml-l was obtained when these drugs, in solution in water, were subjected to oxidation with dichromate, either a t room temperature or a t 60 "C (Table VI).Oxidation with alkalifie permanganate about 60 "C (Table VI). Recoveries from blood and urine In order to establish the effectiveness of an oxidative procedure for samples from biological sources, amitriptyline was extracted from blood and urine, to which it had been added, and then oxidised. The oxidation procedure used was the less sensitive but more precise, and more easily manipulated, cold dichromate method. For each drug a 20-min extraction on a roller bottle shaker at approximately 40 rev min-l proved to be as effective as extraction by vigorous hand shaking for 5 min.l9 The use of a roller shaker minimised emulsion formation, caused less operator fatigue and enabled several samples to be extracted concurrently.The extraction procedure detailed above led t o acceptable recoveries (76% from urine, 68% from blood), which were not improved by the use of larger volumes of ether or the use of stronger acid. As recoveries from urine appeared to be consistent, being 78, 77, 76, 75 and 7574, no further improvements were attempted. By use of this extraction procedure calibration graphs were produced for amitriptyline added to blood and urine (Table VI). In each instance the absorbance from a blank determina- tion at 264 nm was subtracted from that of the dosed sample.This technique gave acceptable linearity over the above range following oxidation at Determination of Amitriptyline in Urine Following the oral administration of amitriptyline, a preliminary assay was carried out on a sample of urine in order to determine the minimum volume that would be required for routine screening of the drug by the cold dichromate procedure following its extraction. The result indicated that the urine samples would have to be large if the final hexane phase was to give a reasonably large absorbance in the ultraviolet spectrum at 264 nm, and conse- quently 50-ml urine samples (Jl-J5 in Table VII) were assayed by use of this procedure. TABLE VII URINARY AMITRIPTYLINE LEVELS ASSAYED BY DIFFERENT OXIDATION FROCEDURES All urine samples coded J were collected from the same subject during the trial.Concentration of amitriptylinelpg ml-1 f A I Urine not hydrolysed t Urine hydrolysedt 1 AS B 0.80 0.63 14 0.09 0.41 38 0.16 0.89 1.19 1.43 J1 62 0.28 2.94 3.23 4.58 J2 86 0.17 1.93 3.20 2.68 J 3 103 0.15 1.44 2.78 2.04 J4 0.14 1.90 0.28 J5 J l 8 / l About240 0.04 SB 7 0.05 0.44 1.09 0.98 DR 7 0.11 0.52 2.60 1.11 FD 7 0.10 0.25 0.53 0.66 Sample code Time*/h ---7 A * After the initial administration of amitriptyline hydrochloride. t A, oxidation with potassium dichromate (1 %) in 9 N sulphuric acid a t room temperature; B, oxidation with potassium permanganate (1%) in 6 N sodium hydroxide solution at about 60 "C. 4 Concentration expressed as dehydroamitriptyline hydrochloride ; the others are expressed as amitriptyline hydrochloride.April, 1976 THE OXIDATION OF AMITRIPTYLINE 251 This analysis should measure only the level of amitriptyline and its demethylated metabolites, as the hydroxy metabolites give no absorbance reading in hexane after undergoing this pr0~edure.l~ Precise determination was not possible because, in many instances, a distinct maximum was observed at 250nm, indicative of anthraquinone formation, as well as the predicted maximum at 264 nm.Even when no distinct maximum was apparent at 250 nm the minimum between 225 nm and 264 nm was poorly defined, again indicating the presence of anthraquinone. Blank urine samples dosed with amitriptyline did not exhibit this phenomenon, thus establishing that any anthraquinone formed did not arise from poorly controlled oxidation conditions.An explanation of these results is the presence in urine of 10,l 1-dehydroamitriptyline, and/or the dehydro derivatives of demethylated metabolites, as these are the only compounds likely to be formed in urine which will yield anthraquinone by this oxidation procedure (10-hydroxy metabolites are known to be unaffected by this procedure17). Metabolites of this type have not yet been reported in man, perhaps because of their low concentration. The formation of these compounds from hydroxy metabolites in vitro would appear to be unlikely during storage at 4 "C. Because of the poor sensitivity of the cold dichromate procedure, assays were carried out using the hot alkaline permanganate method.18 A further increase in sensitivity was achieved by using 2.5ml of hexane for the extraction of the oxidation product.Smaller (10ml) urine samples (Jl-J5, Table VII), when assayed by use of this procedure, showed much higher levels of amitriptyline and metabolites than when assayed by the cold dichromate technique. This increase is hardly surprising as alkaline permanganate solution yields anthraquinone from the unchanged drug and all of its extracted metab0lites.l' Also, the oxidation yield was similar in each instance (46-50%) and so, bearing in mind the precision of oxidation yields from amitriptyline by the permanganate procedure, the results were thought to be reasonably accurate. The difference in levels between the two procedures will be approximately equal to the concentration of the unconjugated, extractable hydroxy metabolites and, in all instances, this difference is much greater than the value for unreacted drug and its demethylated metabolites.Hydrolysis of urine samples with hydrochloric acid, by any of the methods described above, gave an increase in the levels determined by use of both oxidation procedures (Table VII), indicating considerable conjugation of the metabolites; this is in agreement with the findings of other w0rkers.8~~Jl Following oxidation with dichromate at room temperature there was a well defined peak in the hexane layer at 250 nm. This peak must arise from the oxidation of 10,ll-dehydroamitriptyline and/or the dehydro derivatives of nortriptyline and demethylnortriptyline, which are formed by the dehydration of hydroxy metabolites (cis or tmns), as, if a solution of 10-hydroxyamitriptyline is subjected to hydrolysis and extracted, subsequent gas-chromatographic analysis shows only one peak, corresponding to 10,ll- dehydroamitriptyline.This finding is in agreement with the work of Braithwaite,s who has used this procedure in analysis; he found that the last named compound has better gas- chromatographic properties than the hydroxy compounds. The quantitative determination following oxidation at room temperature was effected by using a calibration graph prepared by oxidising a range of concentrations of 10,ll-dehydro- amitriptyline solutions (2-20 pg ml-1). Obviously, the presence of any amitriptyline or demethylated metabolites would cause inaccuracies as this oxidation procedure leads to the formation of DiHDBCH from these compounds.Such a disadvantage could not be overcome by using acidic dichromate solutions, which are more concentrated with respect to sulphuric acid, because, although their use would lead to the formation of anthraquinone, the yield would be very low. The assays of hydrolysed urine samples using alkaline permanganate, like those of non- hydrolysed samples, gave values that included all extractable metabolites. However, the levels found by these determinations were often lower than when using cold dichromate oxidation (Table VII) but, as anticipated, much higher than assays of unhydrolysed samples with alkaline permanganate. The use of alkaline permanganate oxidation of a urine extract following hydrolysis necessitated that the final hexane phase should be washed with dilute acid in order to obtain a clean spectrum.Triplicate analyses of some samples confirmed that the dichromate assay at room tempera- ture recovered higher levels of amitriptyline (assayed as dehydroamitriptyline hydrochloride)252 CADDY, FISH AND TRANTER: STUDIES ON Analyst, Vol. 101 than the hot alkaline permanganate procedure (assayed as amitriptyline hydrochloride). It is difficult to rationalise this discrepancy because the latter method would give similar yields of anthraquinone for amitriptyline and all of the metabolites considered, while the former would give high yields only with the dehydro metabolites; the contribution from amitriptyline and demethylated metabolites (which, in any case, do not yield anthraquinone) to the gross absorbance at 250nm from a 10-ml urine sample would be very small.One possibility is that other metabolites, inert to alkaline permanganate but oxidised by acidic dichromate, are extracted following hydrolysis, The N-oxide of amitriptyline is reported as being a m e t a b ~ l i t e , ~ ~ ~ but it is not extractable by the procedure adopted. With all methods of analysis, relatively high levels of amitriptyline (or 10,l l-dehydro- amitriptyline) were found in the urine sample J3 (Table VII), collected between 8 and 32 h after the final drug administration. By using an oxidative technique, Wallace and DahP found that maximum urinary levels, following the oral administration of a single 50-mg amitriptyline hydrochloride tablet, usually occurred between 12 and 24 h.This finding is in reasonable agreement with the present work. The increased elimination following the final drug administration (Tables IV and VIII) was not so great if the volumes of urine voided are considered (Table IV). Even 56 h after the final administration of drug the determined levels were high and higher than those found during the two days of drug administration. Even a urine sample collected 8 d following the final drug administration still had a measurable drug level. TABLE VIII TOTAL AMITRIPTYLINE IN URINE ASSAYED BY DIFFERENT OXIDATION PROCEDURES Amount of amitriptylinelmg r \ A Urine not hydrolysedt Urine hydrolysedt -7 AS B 1.13 1.04 14 0.15 0.67 38 0.31 1.72 2.30 2.76 62 0.39 4.15 4.55 6.45 86 0.23 2.66 4.41 3.70 103 0.17 1.64 3.17 2.33 Sample code Time*/h B) A J1 52 J3 54 55 * After the initial administration of amitriptyline hydrochloride.t A, oxidation with potassium dichromate (1 %) in 9 N sulphuric acid at room temperature: B, oxidation with potassium permanganate (1%) in 6 N sodium hydroxide solution a t about 60 "C. 1 Expressed as dehydroamitriptyline hydrochloride, the others as amitriptyline hydrochloride. Chromatographic Analysis of Urine Samples Following Oral Amitriptyline Ad- ministration In an attempt to rationalise some of the results obtained by oxidation, 50-ml urine samples were extracted and the extracts subjected to thin-layer chromatographic analysis. The extracts from urine samples that had not been hydrolysed contained four components that were not present in blank urine (Table IX).Two of these components gave weak responses to an iodoplatinate spray; the slower running compound had a similar migration to nortripty- line, while the faster moving compound could have been either amitriptyline or its dehydro derivative. The faster running of the other two components, both of which gave a strong reaction with the detection spray, was partially obscured by nicotine, but its RF value was similar to that of hydroxyamitriptyline metabolites. The remaining component had a smaller RF value than nortriptyline, suggesting that it was a more polar metabolite, such as h ydroxynor t rip t yline or hydrox ydeme t h ylnortrip t yline . Extracts from hydrolysed urine samples were subjected to thin-layer chromatography and yielded only three components, all of similar intensity.The faster running component had an RF value similar to that of amitriptyline and its dehydro derivative but was judged to be the latter, arising from the dehydration of the 10-hydroxy compound during hydrolysis. ThisApril, 1976 THE OXIDATION OF AMITRIPTYLINE 253 judgement was confirmed by gas - liquid chromatography. A second component, again partially obscured by nicotine, had a migration velocity similar to that of the hydroxy metabolites but was unlikely to be one of these, as they are dehydrated by the hydrolysis procedure. (It might have been a dehydration product of the more polar hydroxynortripty- line or hydroxydemethylnortriptyline.) that of nortriDtvline. The third component had an RB value similar relative to 10-hydroxy- amitriptyline (isomer 135) 1.35 1.25 1 .oo 0.91 0.66 0.97 I ./ TABLE IX THIN-LAYER CHROMATOGRAPHIC DATA FOR THE BASIC FRACTION EXTRACTED FROM HYDROLYSED AND UNHYDROLYSED URINE SAMPLES FOLLOWING ORAL ADMINISTRATION OF AMITRIPTYLINE Migration Standard Amitriptyline 10,l l-Dehydroamitriptyline 10-Hydroxyamitriptyline (1 35) 10-Hydroxyamitriptyline (1 34) Nortriptyline Nicotine to Urine sample J1 J 2 J3 J4 J5 Unhydrolysed 1.27+ 1.30+ 1.28+ 1.28+ 1.30+ 0.62+ 0.64+ 0.641 0.64 + 0.66 + 0.33+ + + 0.35+ + + 0.37+ + + 0.35+ + + 0.35+ + + Hydrolysed 1.28+ + + 1.25+ + + 1.25+ + + 1.28f + + 1.25+ + + 0.63+ + + 0.65+ + + 0.65+ + + 0.63+ + + 0.65+ + + *1.00+ + *1.00+ + *1.00++ *1.00+ + *1.00+ + *1.00+ + *l.OO+ + *1.00+ + *1.00+ + *1.00+ + Spot intensity: + weak, + + moderate, + + + strong. * Approximate assessment as spot corresponds to amitriptyline metabolite and nicotine in admixture.Gas - chromatographic analysis was inconclusive for the unhydrolysed urine but dehydro- amitriptyline was confirmed to be present in hydrolysed urine, although the peak could not be completely resolved from two other components of the urine extract. Conclusion Oxidation with alkaline potassium permanganate solution at 60 "C or with acidic potassium dichromate solution at room temperature as a technique for the assay of amitriptyline extracted from body fluids is satisfactory. The results from oxidative studies imply that dehydroamitriptyline or its demethylated derivatives may be natural metabolites of ami- triptyline in man; metabolites other than those already reported also appear to be present.By using the two oxidative procedures concomitantly, some knowledge of the levels of hydroxylated (at the 10-position) metabolites can be obtained. We thank Mr. Aksel Jgrgensen of H. Lundbeck and Co. for samples of lo-hydroxyamitripty- line. 1. 2. 3. 4. 5 . 6. 7. 8. References de Leenheer, A., and Heyndrickx, A., J . Pharm. Sci., 1971, 60, 1403. Hammar, C. G., Alexanderson, B., Holmstedt, B., and Sjoqvist, F., CZin. Pharmac. They., 1971, 12, Knapp, D. R., Gaffney, T. E., McMahon, R. E., and Kiplinger, G., J . Pharmac. Exp. Ticer., 1972, Alexanderson, B., and Borga, O., Eur. J , CEin. Pharmac., 1973, 5, 174. Hucker, H. B., and Porter, C. C., Fedn Proc. Fedn Am. Socs Exp. B i d , 1961, 20, 172. Corona, G. L., and Facino, R. M., Biochem. Pharmac., 1968, 17, 2045. Eschenhof, E., and Rieder, J., Arzneimittel-Forsch., 1969, 19, 959. Braithwaite, R. A., personal communication, 1973. 496. 170, 784.254 CADDY, FISH AND TRANTER Jorgensen, A., Personal Communication, 1973. Diamond, S., Curr. Ther. Res. Clin. Ex$., 1965, 7, 170. Braithwaite, R. A., and Whatley, J. A., J . Chromat., 1970, 49, 303. Braithwaite, R. A., and Widdop, B., Clin. Chim. Acta, 1971, 35, 461. Forbes, G., Weir, W. P., Smith, H., and Bogan, J., J . Forens. Sci. SOL, 1965, 5, 183. Amundson. M. E., and Manthey, T. A., 1. Pharm. Sci.. 1966, 55. 277. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Forrest, I. S., Rose, S. D., Brook&, L. G., Halpern, B., Bacon,-V. A., and Silberg, I., Agressologie, 1970, 11. 127. Wallace, J: E., and Dahl, E. V., J . Forens. Sci., 1967, 12, 484. Tranter, J., “Spectrophotometric Determination of Some Drugs Containing the Diphenylmethylidene Caddy, B., Fish, F., and Tranter, J., AnaZyst, 1975, 100, 563. Caddy, B., Fish, F., and Tranter, J., Analyst, 1974, 99, 555. Moiety Following Oxidation,” PhD Thesis, University of Strathclyde, 1973. Received May 23rd, 1975 Accepted Novembev l l l h , 1975

ISSN:0003-2654    

DOI:10.1039/AN9760100244

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Analyst, April, 1976, Vol. 101, $p. 255-259 255 The Determination of Chromium in Plain Carbon Steel and Low-alloy Iron and Steel by Atomic-a bsorption Spectrophotometry W. D. Cobb, W. W. Foster and T. S. Harrison British Steel DN16 1BP Corporation, Scuntlzorpe and Lancnshire Group, P.O. Box No. 1, Scunthorpe, South Humbersidc, Co-operative examination of published atomic-absorption procedures for several elements in steel showed good agreement between results apart from those for chromium. For this element the depressive effect of iron in the air - acetylene flame is eliminated when using the preferred nitrous oxide - acetylene flame. However, control of flame composition is essential as vanadium, molybdenum, aluminium and titanium in low-alloy synthetic test solutions and samples increasingly enhance the chromium absorbance with increased fuel richness.Copper, nickel and iron have no effect but, in any event, iron should be included in the calibration solutions. Hence a lean, oxidising flame must be used and the results for samples then agree closely with the certificate values. Procedures published by this laboratory1 for the determination of cobalt, copper, nickel, aluminium, lead, molybdenum, chromium and vanadium in iron and steel have been examined on a co-operative basis by nine laboratories, with the object of producing standard methods. Each laboratory analysed five samples each four times. When the results were evaluated statistically2 satisfactory agreements were obtained for these metals with the exception of chromium with which, in some instances, appreciable between-laboratory variations were observed, which could be associated with the variety of instruments used.Furthermore, the alloy steels showed a high bias for chromium with the recommended nitrous oxide - acetylene flame conditions. This hotter flame is preferred to the air - acetylene flame as the depressive chemical effect of iron, although controlled by the addition of 2% of ammonium chloride to samples and standards, is eliminated in the former.314 There is, however, a small loss of sensitivity. Advocates of the use of air - acetylene, on the other hand, find that absorption of chromium is much stronger if the flame is fuel rich but that the serious suppression by iron, and nickel, is much reduced in a hot, oxidising or lean flame in a three-slot b ~ r n e r .~ In a detailed study that culminated in the development of procedures for the determination of chromium and molybdenum in a complete range of steels, Thomerson and Price5 prefer the use of the nitrous oxide - acetylene flame as all interferences cannot be overcome in the other flame. Perchloric acid was considered to be the best solvent to use because, unlike sulphuric acid - phosphoric acid or nitric acid - hydrochloric acid mixtures, it did not interfere with the absorption response. Furthermore, it is essential to obtain both elements in the same oxidation states in both their sample and calibration solutions in order to achieve accurate results. They also found that the iron enhances the absorption of chromium in the nitrous oxide - acetylene flame, the effect increasing slightly with increase in iron content. Hence iron should still be added to the calibration solutions and in dilutions, when necessary, so as to maintain its concentration constant.The iron content of steel sample and calibration solutions should be maintained at the same level. For chromium the fuel flow was not critical and the optimum sensitivity was obtained for a red “feather” height of 0.5-1.0 cm. An observation height of 0.50 cm was preferred because at this height the level of flame noise was reduced. It is important to use pure chromium metal and not potassium dichromate for the standards because potassium reduces the ionisation of chromium in the nitrous oxide - acetylene flame in the standards only, thus giving low results.The method used gave rise to a rapid com- posite scheme of analysis for several elements in steel, which scheme was subsequently developed and published.’-?256 Analyst, VoZ. 101 Examination of more recent literature shows individual preferences for various flame compositions in order to suit the analysis being undertaken. In a comparative exercise Scholes7 reports the depression of chromium absorption by iron, nickel and molybdenum in the air - acetylene flame and a negligible enhancement in the nitrous oxide - acetylene flame. This depression by iron was overcome by Ottaway and Pradham,* who added quinolin-8-01 as a releasing agent and interference suppressor to a solution of steel in dilute hydrochloric acid. Below a 1% concentration of chromium the additions of iron to the standards was unnecessary.The purpose of the present investigation was to study the interferences of other elements on chromium, using various flame conditions for the published methodl and thus enabling the optimum conditions to be selected. COBB et al.: DETERMINATION OF CHROMIUM IN PLAIN Method The method is recommended for application to plain carbon steel and low-alloy iron and steel containing chromium in the range 0-0.12y0 (Note l), giving a reproducibility of &0.003%, and in the range 0-1.20%, giving a reproducibility of -,t0.015~0. The instrument conditions, using a Techtron AA-5 spectrophotometer, are tabulated below. Wavelength . . .. .. .. Lamp current. . .. .. .. Observation height . . .... Fuel setting . . . . .. .. Slit width . . .. .. * . Burner .. .. .. .. Support setting . . . . .. Damping . . .. .. . . Scale expansion . . . . .. 357.9 nm 15 mA 0.33 nm Nitrous oxide - acetylene 1 mm (Note 2) Acetylene cylinder pressure, 10 lb in-2 Fuel flow reading 6 to 9 (Note 3) Nitrous oxide cylinder pressure, 36 lb in-2 Gauge reading, 18 lb in-2 C x l Reagents Hydrochloric acid, sp. gr. 1.16-1.18. Nitric acid, sp. gr. 1.42. Stock iron solution. Transfer 50 g of pure iron to a 1.5-1 beaker. Add 300 ml of hydro- chloric acid (sp. gr. 1.16-1.18) and 25 ml of nitric acid (sp. gr. 1.42) andmark the level of acid. Digest until the iron is dissolved, replacing any loss of acid with hydrochloric acid. Complete the oxidation by the cautious addition of nitric acid and boil the solution in order to expel nitrous fumes.Add a further 200 ml of hydrochloric acid, cool the solution, dilute it to 1 1 and filter it. 20 ml of solution = 1 g of iron, 10 ml of hydrochloric acid (sp. gr. 1.16-1.18) and These concentrations ensure that calibration solutions contain the same amount of iron and acids as the sample solutions. Standard chromiwn solution. Accurately weigh a small amount of pure chromium shot, dissolve it in hydrochloric acid (50% V/V) and dilute to give a solution so that 1 ml of nitric acid (sp. gr. 1.42). 1 ml of solution = 100 pg of chromium. This procedure gives a chromium solution in the same oxidation state as the sample solutions, thus avoiding different excitation effects. Preparation of Sample Solution (50% V / V ) .then oxidise it with the minimum dropwise addition of nitric acid (sp. gr. 1.42). solution to expel nitrous fumes, cool and dilute it to 100 ml in a calibrated flask. necessary. Transfer 2.0 g of sample into a 150-ml conical beaker and add 40 ml of hydrochloric acid Cover the beaker with a clock-glass and digest until the sample has dissolved, Boil the Filter ifApril, 1976 CARBON STEEL AND LOW-ALLOY IRON AND STEEL BY AAS 257 For the range O-0.12yo of chromium, spray the 2% sample solution undiluted, and for the range 0-1.2%, transfer a 5-nil fraction of the sample solution into a 50-ml calibrated flask, dilute to the mark and mix for spraying. Preparation of Calibration Solutions For the range O-0.12yo of chromium Into each of a series of 50-ml calibrated flasks transfer 20ml of iron solution.Add chromium fractions according to the following table, dilute to the calibration mark and mix in each instance. Standard chromium Chromium, Chromium, O/ solution/ml p.p.m. /O 0 0 0 3 6 0.03 6 12 0.06 9 18 0.09 12 24 0.12 For the rangc O-1.2yo of chromium transfer 5 ml of the dilute iron solution. table, dilute to the calibration mark and mix in each instance. Dilute 20ml of iron solution to 50ml. Into each of a series of 50-ml calibrated flasks Add chromium fractions according to the following Standard chromium solu tionlml 0 2 4 6 8 10 12 Chromium, p.p.m. 0 4 8 12 16 20 24 Chromium, 0 0.2 0.4 0.6 0.8 1 .o 1.2 Yo Determination Spray the appropriate calibration solutions followed by the sample solutions. Spray water between each test and set the zero while spraying water each time.Plot the absorbance against element concentration. Construct a calibration graph and read off the percentage of chromium (Note 4) in each sample. Set the instrument according to the table of instrument conditions. (Repeat this process.) NOTES- 1. The range can be doubled by mixing equal volumes of the sample solution and the blank (nil calibration point) for spraying. 2 . In general the optimum position for the observation height is found by raising the burner until i t just begins to obstruct the light path. Then lower the burner 1 mm. 3. The flame condition that gives maximum absorbance while spraying a calibration solution, a red “feather” of approximately 7 mm, can give high results for steels containing titanium, vanadium, aluminium or molybdenum.The optimum setting is found by starting with a fuel-rich flame. Then reduce the acetylene flow to give a red “feather’ of approximately 2 mm in height (see below). 4. For the greatest accuracy, spray each sample solution between two calibration points for concen- trations of chromium above and below that present in the sample and relate the absorptions to per- centages of chromium. This point is indicated by a deflection of the meter. Discussion of Method Taking into account the bias shown by alloy steels and observations on the effects of flame conditions reported in the literature3y4 the authors decided to try the effects of three variations of the nitrous oxide - acetylene flame on synthetic solutions and samples as follows.258 Alternative Flame Conditions COBB et aZ.: DETERMINATION OF CHROMIUM IN PLAIN AnaZyst, VoZ.101 1. Lean Adjust the acetylene flow-rate to give a red “feather” approximately 2-3 mm high. This height is close to that which gives the leanest flame that can be safely supported on the burner - spray system of this particular instrument. 2. Medium Adjust the acetylene flow-rate while spraying a reference solution so as to obtain the maximum absorbance reading. This gives a red “feather” about 8-10 mm high. 3. Rich Increase the acetylene flow-rate almost until the flame becomes luminous, giving a red “feather” about 20 mm high. Tests on Synthetic Solutions Standard solutions of vanadium, molybdenum, copper , aluminium, nickel and titanium were prepared. Test solutioiis were then prepared by adding appropriate volumes of these solutions to a base solution containing 0.2% m/V of iron plus the equivalent of 0.60% of chromium.One test solution contained all six potential interferents but, in this instance, the base solution used contained 0.1% m/V of iron. Details of the composition of these test solutions are given in Table I. Spray these solutions using the above three flame conditions and relate the absorbances found for tests 2-6 in each instance to that for test solution 1 so as to give the equivalent percentage of chromium for each flame condition. (With our instrument the graph for chromium is linear and passes through the origin.) Samples Take a selection of samples of varied composition through the procedure detailed in the method.Spray each solution together with the appropriate calibration solution for the range concerned, usiiig the three flame conditions, and record the percentage of chromium found for each. (The solutions should be sprayed in small batches, particularly for condition 3, as a richer flame is more noisy and there is a possibility of carbon formation, especially with older types of burner.) TABLE I EFFECT OF VARYING THE FLAME CONDITIONS ON THE DETERMINATION OF CHROMIUM IN STEEL IN SYNTHETIC SOLUTIONS Flame conditions: height of red “feather”: lean, 3 mm; medium, 10 mm; and rich, 20 mm. Synthetic solution Chromium found for three different A r flame conditions, % Test Iron in test r No. solution, % Additions, % in a sample Lean Medium Rich 1 0.2 Chromium 0.60 (reference solution) 0.60 0.60 0.60 2 + vanadium 2.0 0.605, 0.605 0.63, 0.625 0.69, 0.70 3 + molybdenum 2.0 0.60, 0.60 0.61, 0.61 0.63, 0.615 4 + copper 1.0 0.60, 0.595 0.595, 0.60 0.60, 0.61 5 + aluminium 0.2 0.60, 0.60 0.615, 0.60 0.655, 0.635 6 + nickel 2.0 0.60, 0.605 0.60, 0.605 0.605, 0.60 7 + titanium 0.5 0.60, 0.60 0.63, 0.63 0.685, 0.685 8 0.1 Chromium 0.60 0.59, 0.60 0.59, 0.60 0.60, 0.60 9 + vanadium 2.0 1 A I + molybdenum 2.0 + copper i:: 0.60, 0.60 0.66, 0.64 0.72, 0.74 + aluminium + nickel 2.0 + titanium 0.5April, 1976 CARBON STEEL AND LOW-ALLOY IRON AND STEEL BY AAS Results 259 The results obtained for the tests on synthetic solutions are shown in Table I and those for the samples in Table 11.TABLE I1 EFFECT OF VARYING THE FLAME CONDITIONS ON THE DETERMINATION OF CHROMIUM I N STEEL SAMPLES Flame conditions: height of red “feather”: lean, 3 mm; medium, 10 mm; and rich, 20 mm.Chromium, yo Sample BCS 25/1 BCS 25211 BCS 25411 BCS 22512 BCS 404 BCS 432 BCS 433 BCS 435 BCS 321 BCS 324 6344611 63452/1 c Type Low alloy Low alloy Low alloy Low alloy Low alloy Plain carbon Plain carbon Plain carbon Mild Mild Plain carbon Plain carbon c Other elements known to be present, % Certifi- .L , cate c- V hlo cu Ti value 0.65 1.51 0.55 0.51 0.23 1.05 0.20 0.42 0.15 0.23 0.34 0.27 0.004 0.34 0.17 1.08 0.10 0.33 0.33 0.68 0.039 0.16 0.21 0.085 0.058 0.52 0.014 0.046 0.56 0.004 0.068 0.13 0.106 0.004 0.17 0.033 0.077 <0.01 0.03 0.020 (0.01 <0.01 0.03 0.024 <0.01 Found for three different flame conditions Lean Medium Kich 0.52, 0.53 0.53, 0.55 0.565, 0.5s 0.43, 0.43 0.43, 0.445 0.45, 0.45 0.28, 0.28 0.28, 0.29 0.29, 0.29 0.69, 0.70 0.705, 0.73 0.715, 0.75 0.215.0.225 0.22. 0.225 0.225. 0.225 A 7 1.08, 1.09 1.10, 1.10 1.12,1.12 0.50, 0.51 0.505, 0.51 0.53, 0.55 0.555, 0.555 0.56, 0.57 0.60, 0.615 0.109, 0.112 0.116, 0.118 0.128, 0.128 0.082, 0.083 O.OS7, 0.087 0.095. 0.096 0.064; 0.065 0.066; 0.067 0.071; 0.070 0.112, 0.114 0.115, 0.117 0.124, 0.123 The results given in Table I indicate that vanadium, molybdenum, aluminium and titanium enhance the chromium absorbance in a fuel-rich flame and this enhancement increases with increasing fuel richness. Conversely, the effect decreases to insignificance in the lean, oxidising flanie. Halving the amount of iron present has no significant effect in any flame while the combined effect of all the additions, although appreciable in the richer flames, is eliminated in the lean flame.The results for the samples, which include plain carbon and low-alloy material in each range of contents, show a similar pattern. Those for the lean flame are considered to be within experimental error of the certificate values. Scholes’ concludes that, whether the nature of the interference in the air - acetylene flame be dissociation, excitation or other phenomenon, the use of the higher temperature nitrous oxide - acetylene flame is an efficient means of eliminating the depressive effects observed. The above observations support and extend this view, the hotter, lean flame eliminating interferences from other elements that were iiot examined by Scholes.’ The presence of iron raises no problems in this work as it is standard practice to include it in the calibration solutions in all spectrophotometric analyses of iron and steel.Other laboratories confirm the advantage of the lean flame although the operating con- ditions vary with each instrument. Thus the 2-3-mm red “feather” is preferred for the Techtron AA5, Unicam SP90 and Perkin-Elmer 103 instru- ments, while either the 2-3 mm or 8-10-mm red “feather” can be used with the EEL 240 spectrophotometer. They all confirm that the more refractory the element the more inter- ference occurs with the richer flame condition. This is gratefully acknowledged. References 1. 2. RS 4237: 1967. 3. 4. 5 . 6. 7. Harrison, T. S., Foster, W. W., and Cobb, W. D., Metallurgia iWetaZ Forming, 1973, 40, 361. Slavin, W., “Atomic Absorption Spectroscopy,” Interscience Publishers, New York, 1968. Reynolds, R. J., Aldous, K., and Thompson, K. C., “Atomic Absorption Spectroscopy,” GriKin, Thomerson, D. R., and Price, W. J., Analyst, 1971, 96, 321. Thomerson, D. R., and Price, W. J., Analyst, 1971, 96, 825. Scholes, P. H., “Second Report of the Working Party on the Application of the Atomic Absorption Spectroscopic Technique : a Collaborative Study of Inter-Element Interference Effects,” BSC Open Report CDL/CAC/117/73. London, 1970. 8. Ottaway, J. M., and Pradharn, N. K., TaZanta, 1973, 20, 927. Received August 19th, 1975 Accepted October 27th, 1975

ISSN:0003-2654    

DOI:10.1039/AN9760100255

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

260 Analyst, April, 1976, Vol. 101, pp. 260-264 The Use of Luminescence Spectroscopy in Aiding the Identification of Commercial Polymers N. S. Allen, J. Homer and J. F. McKellar Department of Chemistry and Applied Chemistry, Salford University, Salford, M5 4WT The fluorescence and phosphorescence properties of a range of commercial polymers have been examined. Each polymer tested exhibited characteristic excitation and emission spectra. Phosphorescence lifetimes were also found to vary significantly. The spectral data obtained are discussed with a view to applying luminescence techniques as a rapid non-destructive method for the characterisation of commercial polymers. Since the studies carried out by Charlesby and his co-workers1v2 there has been considerable interest in the luminescence exhibited by commercial polymers, particularly with regard to the mechanisms of photo-degradation and stabilisation.Work in this field has shown that many commercial polymers exhibit their own characteristic fluorescence and phosphorescence emi~sions.l-~~ These emissions may originate either from chromophoric units that are basic to the polymer structure, or from impurities adventitiously incorporated during manufacture. Recently, Srnith2l has suggested that the luminescence properties of a polymer could be of use for its characterisation. In this paper we report on the fluorescence and phosphorescence properties of a range of commercial polymers, and the results obtained suggest that lumines- cence analysis should be valuable in the characterisation of a wide range of polymers, parti- cularly if it is used in conjunction with other techniques such as infrared spectroscopy and gas - liquid chromatography. Further, the proposed method requires no tedious sample preparation as it is possible to determine directly the luminescent properties of a polymer in a variety of physical forms such as powder, chip and film.Experimental Materials and Instrumentation The polymer materials listed in Table I were obtained from various manufacturers and contained no commercial additives. Corrected fluorescence and phosphorescence spectra were recorded by using a Hitachi Perkin-Elmer MPF-4 spectrofluorimeter with a double-grating monochromator (1200 lines per millimetre) and with phosphorescence attachments. Phosphorescence lifetimes were measured by coupling the sample intensity signal from the fluorimeter to a Tetronix DM-64 storage oscilloscope.The mean lifetime was taken as the time for the phosphorescence emission to decay to l / e of its initial intensity. All fluorescence measurements were carried out at 300 K and all phosphorescence measurements at 77 K. The wavelength scale of the instrument was checked by using anthracene in ethanol and phenanthrene in EPA solvent [diethyl ether - isopentane - ethanol (5 + 2 + Z ) ] at 77 K and was found to be correct to &0.5 nm. Method For fluorescence measurements, polymer samples in the form of granules and fibres were examined in a silica Dewar flask that contained no liquid nitrogen, without the chopper in position. Powder samples were placed in 5 mm 0.d.silica tubes before positioning the tubes in the Dewar flask. Film samples were analysed at an angle of 45" to the beam of excitation light in order to prevent any scattered light from entering the emission monochromator. Phosphorescence measurements on granules, fibres and powder materials were carried out as described above but with liquid nitrogen in the Dewar flask and with the chopper in position (use of the chopper eliminates any scatter). Strips of polymer film were positioned in the flask for these measurements.ALLEN, HOMER AND MCKELLAR 261 In our experience the luminescence properties of a particular polymer are consistent and are independent of the origin of manufacture. Indeed, these observations appear to be so from the literature where different workers have obtained similar spectra using the same polymer but supplied by different manufacturers.Results and Discussion Luminescence Properties of Commercial Polymers The use of luminescence spectroscopy as an analytical technique for polymer characterisation involves the measurement of the following properties : (a) The fluorescence emission spectrum, which is obtained by exciting the polymer with light of any wavelength that is capable of producing the emission spectrum. The most satis- factory spectrum for definitive purposes, however , is normally recorded at the wavelength maximum (Amax.) of the excitation spectrum. (b) The fluorescence excitation spectrum, which is obtained by recording the variation in intensity of the emission spectrum (Amax.) as a function of the excitation wavelength.The corrected excitation spectrum should closely match the absorption spectrum of the chromo- phore. (c) The phosphorescence emission spectrum, which is obtained by using the same method as was used for (a). (d) The phosphorescence excitation spectrum, which is obtained by using the same method as was used for ( b ) . (e) The phosphorescence lifetime, the value for which is obtained a s described under Experimental. Finally, if it is found that the wavelength maximum (Amax.) of either the fluorescence or phosphorescence spectrum of the polymer varies with the excitation wavelength, the presence of more than one chromophore is indicated and it is necessary to examine the polymer over a wide range of excitation wavelengths in order to achieve its complete characterisation.The above luminescence characteristics of a number of commercial polymers are shown in Table I. We have also included, when possible, the identities of the chromophores that are believed to be responsible for the various emissions. All of the polymers examined exhibited fluorescence emission in the range 300-450 nm and phosphorescence emission in the range 400-600 nm. Within each wavelength range , however, each polymer exhibited its own characteristic spectrum. Indeed, in certain instances the spectra observed were highly structured and knowledge of this structure assists further in the characterisation of a polymer. The clearly structured phosphorescence emission spectra of polyethylene terephthalate [Fig.1 ( a ) ] , wool and polystyrene [Fig. 1 ( b ) ] and polyethylene [Fig. l(c)] are good examples to illustrate this aspect. The fluorescence and phosphorescence excitation wavelength maxima for many polymers are different, thus indicating that the same light-absorbing species cannot be responsible for both emissions. A good example is that of polyethylene terephthalate [Fig. 1 (a)], with which distinct differences are evident in both the structure and maximum wavelength of the excitation spectra. Most of the polymers examined showed marked differences in their phosphorescence lifetimes. This property, although not so specific as the excitation and emission spectra, could, nevertheless, assist in characterising the polymer. However, a further feature of the results for the phosphorescence lifetime is that, in some instances, a change in the physical form of the polymer alters the emission lifetime.This effect occurs mainly with the highly crystalline synthetic polymers. For example, on pressing polyethylene powder into film it was found that the phosphorescence lifetime changes from 2.3 to 0.6 s. It is possible that these differences in phosphorescence lifetime are due to changes that occur in the morphological structure of the polymer.ll The advantages and disadvantages of luminescence spectroscopy for characterising polymers are given below. This feature may also be useful for characterising a particular polymer. Advantages 1. The technique is rapid and non-destructive.262 ALLEN et al.: USE OF LUMINESCENCE SPECTROSCOPY IN Analyst, VoZ.101 TABLE I LUMINESCENCE CHARACTERISTICS OF VARIOUS COMMERCIAL POLYMERS Fluorescence Ex’citation Polymer waveleng th/nm Polyethylene terephthalate chip film fibre Polyurethane - MDIt based film Nylon 6,6 chip . . . . . . fibre .. .. .. 320, 344, 357(s)* 344,357 344,357 372 357 357 . 6chip .. .. .. 335 6,lO chip . . . . . . 345, 355 11 chip . . . . . . 327(s), 340 12 chip . . .. . . 410 Wool fibre . . .. .. .. 370 Gelatin film . . . . . . . . 327 Cellulose film . . . . . . . . 320 Cellulose acetate film . . . . 305 Poly(viny1 chloride) film . . . . 290 Polytetrafluoroethylene film . . 328 Polyethylene (low density) powder.. film .. 273 Polyethylene - vinyl acetate film . . 265(s), 290 Polypropylene film . . . . .. 263 Poly-4-methylpent-1-ene film . . 285 Poly(viny1 alcohol) film . . . . 258(s), 295, 330 Polystyrene chip . . . . . . 318, 330 265(s), 300 (high density) film . . 265, 290 Polymer Polyethylene terephthalate chip film fibre Polyurethane - MDIt based film Nylon 6,6 chip . . . . fibre .. .. 6 chip .. .. 6,lO chip . . . . 11chip .. .. 12 chip .. . . Wool fibre . . . . . . Gelatin film . . . . .. Cellulose film . . . . . . Cellulose acetate film . . Poly(viny1 chloride) film . . Polytetrafluoroethylene film Polyethylene (low density) powder film (high density) film Polyethylene - vinyl acetate film Polypropylene film . . . . Poly-4-methylpent-1-ene film Poly(viny1 alcohol) film . . Polystyrenechip .. . . .. .. .. .. .. .. .. .. .. .. Emission wavelength/nm 370, 389,405 370, 389,405 370, 389,405 420 41 7 417 390 395.410 375(s), 385(s), 398 450 438 361(s), 390 355 416 a m ... 350 335(s), 350 29G(s), 310, 329(s), 354(s), 370(s) 295(s), 312(s), 330(s), 344(s), 358 312(s), 330, 344(s), 358(s) 309(s), 320 310(s), 330 360, 350(s) 336, 354, 368(s) Intensity (relative) Very strong Very strong Very strong Very strong Very strong Very strong Very strong Very strong Very strong Very strong Strong Medium Very weak Medium Very weak Very weak Medium Strong Medium Medium Strong Strong Very weak Very strong Phosphorescence , Excitation wavelength/nm 380, 318, 351 284, 310 284. 310 320 296 296 282 300 269(sj,- 273 268, 286(s) 296 284 285 284 273, 280 278,280 275 280, 327(s) 270, 290, 330 273, 315 260-280$ 290(s), 300, 336(s) 260-3201 260-2801 Emission wavelength/nm 425(s), 460 425(s), 477 425(s), 477 423,455,489 400 430 390(s), 420, 455(s) 430 423, 450(s) 363(s), 410 390.413.438, 463 430 440 445 440 450 367, 381, 391,405,416 420 450 455 420(s), 445, 480(s), 510(s) 430 436 398,428,456,492 Mean lifetimels 1.30 0.50 0.70 0.02 2.10 1.30 1.70, 1.60,l.lO 0.70 1.00 4.00 0.70 0.50 0.44 1.00, 0.88 0.30 0.40 2.30 0.60 0.35 0.35 0.50-1.20 0.86 0.40 0.008, 1.50 , Chromophore Polymers Polymers PolymerS Unknown Unknown Unknown Unknown Unknown Unknown Unknown Heterocyclic residuesz8 Heterocyclic residuesee Carbon ylzS CarbonyleJ Unknown Unknown Carbony1,l ‘I [h;&:::iknsa ’ Unknown trans- S tilbene”-lB Intensity (relative) Very strong Very strong Very strong Very strong Very strong Very strong Strong Strong Strong Strong Strong Weak Very weak Very strong Medium Very weak Strong Strong Weak Medium Medium Weak Very weak Very strong C hromophore Unknown Unknown Unknown Benzophenone type” CarbonyP1l CarbonyP-ll CarbonylQ-l1 Carbonyle-ll CarbonyP-ll Carbonylg-ll TryptophanDZ Unknown Uiiknown Unknown Unknown Carbon yll CarboIiyl1 Carbonyll Unknown Unknown Carbonyl1.%” Carbonyll 8w1l Unkiiown Acetophenone,ls-Ib benzaldeh yde * (s) = shoulder.t MDI = Diphenylmethane 4,4’-diisocyana te. 1 Broad and structureless spectrum. 2. Its use requires no tedious sample preparation. 3. It can be used to analyse polymers in powder, chip, film or fibre form. Normally, the amounts of polymer required are in the range 10-50 mg for powders and film, and up to 100 mg for fibres and chips. 4.It can be a highly sensitive technique (concentrations of about moll-1 of aromatic hydrocarbons in polyolefins have been reportedl~~,~~). 5. It can be used to characterise polymers within a particular class, e g . , the polyamides (see Table I). This advantage would be useful in conjunction with the use of other techniques such as infrared spectroscopy.April, 1976 AIDING THE IDENTIFICATION OF COMMERCIAL POLYMERS 263 Disadvantages 1. Use of the technique gives no information on those commercial polymers which are either not luminescent or only weakly so [e.g., poly(methy1 methacrylate) and polybutadiene26]. I 300 400 450 500 Wavelengthhm Fig. 1. (a), Fluorescence (broken line) and phos- phorescence (full line) excitation and emission spectra of polyethylene terephthalate chip ; (b) , phosphorescence excitation and emission spectra of polystyrene chip (full line) and wool fibre (broken line) ; and (c), phosphorescence excitation and emission spectra of polyethylene powder (full line) and film (broken line).However, this problem can sometimes be overcome by oxidising the polymer at a particular temperature in air for a given period of time. The non-volatile oxidation products are usually luminescent and are characteristic for that particular p 0 1 y m e r . ~ ~ ~ ~ ~ ~ ~ 2. Some light stabilisers quench the luminescence from polymers1618 and if their presence is suspected, they must be removed by solvent extraction. Similarly, other processing addi- tives such as antioxidants may also be present in the extract, some of which may be identified by luminescence spectroscopy.26264 ALLEN, HOMER AND MCKELLAR Conclusions We have examined the luminescence properties of about twenty different commercially important polymers using a corrected Perkin-Elmer MPF-4 spectrofluorimeter. As the number of polymers examined represents only a fraction of those used commercially, and because different instruments will give slightly different wavelength values, we recommend that the analyst construct a table of data (or spectra) by using his own instrument.Unless a corrected instrument is available the data given in Table I should be used only as a guide. When constructing such a table the form of the polymer should also be noted as this knowledge will help to eliminate any differences due to processing history and even morphology. Further, in addition to those given above, there may well be other advantages or disad- vantages in using luminescence analysis for characterisation purposes, which will obviously emerge with continued use of the technique in the study of polymer systems.The authors thank Dr. L. S. Bark (University of Salford) for helpful discussions and also the S.R.C. for financial support for one of us (J. H.). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. References Charlesby, A., and Partridge, R. H., Proc. R. Soc., Ser. A , 1965, 283, 312 and 329. Boustead, I., and Charlesby, A., Eur. Polym. J . , 1967, 3, 459. Pacifici, J. G., and Straley, J . M., J . Polym. Sci., Part B, 1969, 7, 7.Carlsson, D. J., and Wiles, D. M., J . Polym. Sci., Part B, 1973, 11, 759. Taylor, H. A., Tincher, W. C., and Hamner, W. F., J . Appl. Polym. Sci., 1970, 14, 141. Carlsson, D. J., Parnell, K. D., and Wiles, D. M., J . Polym. Sci., Part B, 1973, 11, 140. Allen, N. S., McKellar, J . F., and Phillips, G. O., Chemy Id., 1974, 300. Allen, N. S., McKellar, J. F., Phillips, G. O., and Wood, D. G. M., J . PoZym. Sci., Part A-I, 1974, Allen, N. S., McKellar, J. F., and Phillips, G. O., J . Polym. Sci., Part A-1, 1974, 12, 1233. Allen, N. S., McKellar, J. F., and Phillips, G. O., J . Polym. Sci., Part B , 1974, 12, 241. Allen, N. S., Homer, J., McKellar, J. F., and Phillips, G. O., BY. Polym. J., 1975, 7, 11. Allen, N. S., and McKellar, J . F., J . AppZ. Polym. Sci., in the press. George, G. A., J . APpl. Polym. Sci., 1974, 18, 419. Burchill, P. J., and George, G. A., J . Polym. Sci., Part B, 1974, 12, 497. KlBpffer, W., Eur. Polym. J . , 1975, 11, 203. Pivovarov, A. P., and Lukovnikov, A. R., Khim. Vys. Energ., 1968, 2(3), 220. Briggs, P. J., and McKellar, J. F., J . AppZ. Polym. Sci., 1968, 12, 1825. Harper, D. J., McKellar, J. F., and Turner, P. H., J . Appl. Polym. Sci., 1974, 18, 2805. George, G. A., J , Appl. Polym. Sci., 1974, 18, 117. Allen, N. S., McKellar, J. F., Phillips, G. O., and Chapman, C. B., J . Polym. Sci., Part B, 1974, 12, 723. Smith, H. F., in Craver, C. D., Editor, “Proceedings of the Symposium on Interdisciplinary Approaches on Polymer Characterisation 1970,” Plenum Press, New York, 1971, 249 ; Chem. Abstr., 1975, 82, 73556. Ghiggino, K. P., Nicholls, C. H., and Pailthorpe, M. T., J . Photochem., 1975, 4, 155. Jortner, J., J . Polym. Sci., Part A-1, 1959, 37, 199. Ranby, B., and Rabek, J. F., “Photodegradation, Photo-oxidation and Photostabilisation of Fox, R. B., and Cozzens, R. F., Macromolecules, 1968, 2, 181. Kirkbright, G. F., Narayanaswamy, R., and West, T. S., Analytica Chim. Acta, 1970, 52, 237. Beavan, S. W., Hackett, P. A., and Phillips, D., Eur. Polym. J., 1974, 10, 925. 12, 2647. Polymers,’’ Interscience Publishers, John Wiley & Sons, Inc., New York, 1975. Received September 18th, 1975 Accepted November 24th, 1975

ISSN:0003-2654    

DOI:10.1039/AN9760100260

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Analyst, April, 1976, Vol. 101, $9. 265-271 265 The Use of a Spark as a Sampling - Nebulising Device for Solid Samples in Atomic-absorption, Atomic- fluorescence and Inductively Coupled Plasma Emission Spectrometry H. G. C. Human, R. H. Scott, A. R. Oakes and C. D. West* National Physical Research Laboratory, CSIR, P.O. Box 395, Pretoria 0001, South Africa A conventional high-voltage spark, operating a t 50 Hz, is used for sampling and nebulising material from solid conducting samples. Gas, fed through the spark chamber, transports the metal particles into a flame or plasma for analysis by atomic-absorption, atomic-fluorescence or plasma emission spectrometry. Calibration graphs for the copper present in aluminium alloys and iron in brass are presented. As early as 1951 Monvoisin and Mavrodineanul used a spark source for removing material, in a finely divided form, from solid samples for analysis by flame emission spectrography. This technique does not appear to have been used in recent years, but considerable efforts have been expended in studying other methods of solid sampling. These methods have been reviewed by Mavrodineanu and Boiteux,2 and more recently by Kantor and P ~ n g o r .~ Most applications of solid sampling have been devoted to the transport of powdered samples into the analytical source. This requires the pre-treatment of the sample by grinding it to an approximately uniform fine particle size. Various methods of pre-treatment are also required by the more recent graphite-furnace techniques of solid-sample analy~is.~,~ When analysing conducting metal samples, however, it is decidedly advantageous to be able to avoid sample pre-treatment in an attempt to decrease the analytical time.The recent success of several worker^^^^,^ in the use of arcs for nebulisation purposes suggested that a re-investigation of the use of sparks for this purpose might be of value. I t is not the purpose of the present paper to assess the relative merits of d.c. arc and spark nebulisation. Rather, it will be shown that a conventional high-voltage spark can also be used successfully as sampler - nebuliser of solid conducting material for subsequent analysis by flame atomic-absorption or -fluorescence spectrometry, or inductively coupled plasma emission spectroscopy. If a spark is assumed to be simply an interrupted arc discharge, the amount of material nebulised by the spark will, by virtue of the lower duty cycle, be consider- ably less than that by a d.c. arc.However, this assumption neglects the possible influence of ion bombardment of the sample during the spark discharge on the ejection of material or cold cathode processes. Of fundamental interest is the possibility of using a separate flame or plasma as the excitation medium for the study of spark-off effects and sampling interferences in various spark discharges. Experiment a1 The spark source used was a voltage-stabilised, triggered source with the following charac- teristics: discharge voltage, 12 kV (constant) ; capacitance, variable in steps of 3 OOO pF from 3 000 to 12 000 pF; inductance, variable from 11 to 160 pH in 9 steps; repetition rate, 50 sparks s-l.The spark chamber (see Fig. 1) consists simply of a circular anode section, containing a 6.3 mm diameter thorium-treated tungsten rod as counter electrode, attached to one side of a hollow Perspex cylinder of 35 mm i.d. The other side of the cylinder was fitted with an O-ring to prevent gas leakage between the sample and chamber. The sample should have a flat surface of 35 mm minimum diameter and is pressed down by a metal pin (not shown), which is held by a spring and connected to the negative terminal of the source. The gas inlets and outlets are situated directly opposite each other. Tubing of length up to 70 cm and 6 mm i.d. conducts the gas, with nebulised particles, to the analytical zone.Atomic-absorption measurements were made with either a 100-mm slot burner supporting an air - acetylene flame or a 50-mm slot burner with a nitrous oxide - acetylene flame, while * On sabbatical leave from Occidental College, Los Angeles, Calif. 90041, USA.266 HUMAN et al.: USE OF A SPARK AS A SAMPLING - NEBULISING Analyst, Vol. 101 atomic-fluorescence measurements were made using a circular, Meker-type burner supporting a hydrogen - oxygen - argon flame. Part of the oxidant gases or argon could be directed to flow through the spark chamber. A 0.25-m grating monochromator (Jarrell-Ash, Model 82410) with an Ebert mounting, an aperture of f 1 : 3.5, dispersion of 1.6 nm mm-l and slit widths of 150 pm, giving a spectral band pass of 0.24 nm, was used for these experiments.The atomic-emission measurements were made using an inductively coupled plasma source operating at a stabilised power of 1 kW at 27 MHz. The plasma torch was constructed from fused silica and incorporated a tangential flow of coolant argon and a capillary injector.8~~ Coupling between the power source and plasma was achieved by means of a tunable capacitor and a water-cooled coil of 14 turns. More details of this source have been given el~ewhere.~J* A 0.5-m focal length grating monochromator (Jarrell-Ash, Model 82 000) with an Ebert mount- ing, having a dispersion of 1.6nm mm-l, was used for the emission measurements. Slits of 25 pm x 5 mm were used. The plasma was focused in a 1 : 1 ratio on to the entrance slit, using a 16 cm focal length silica lens, and the observation height in the plasma was 20 mm above the load coil.Radiofrequency interference by the spark source on the detection system and chart recorder used for registering signals was avoided by doubly screening the cable between the photo- multiplier and the detector, while the spark chamber was enclosed with the rest of the spark power source in a metal cabinet. Electrical interference from the spark was reduced to a negligible level in this way. /Sample , Perspex +Gas in Gas out +- Brass Anode Fig. 1. Schematic diagram of the spark chamber. Atomic-absorption Measurements The analytical working graph that is obtained for copper in aluminium, by using the air - acetylene flame for atomisation of the nebulised particles, is shown in Fig. 2.The spark- source parameters were C = 6 000 pF, L = 75 pH. Due to spark-off effects on the sample surface the signal was high initially and levelled off after approximately 1 min (see Fig. 3). For these preliminary measurements the signals were not integrated and the values were read off directly from a chart recorder. The detection limit, defined as that concentration giving a signal equal to twice the peak-to-peak fluctuations of the background, was found to be 0.1% of copper. In order to establish the precision of the method, repeated measurements of the absorption signal that was obtained by sparking sample NBS 603 (containing 0.29% of copper) were made. The sample was removed from the chamber for re-surfacing after each measurement. The relative standard deviation of the readings was 2.9%, while the standard deviation in concentration units was O.Ol~o.The readings, taken every 2 min, were averaged by eye for the last minute in order to obtain the value. Three sequential recordings of the absorption signal that illustrate the reproducibility of measurement are shown in Fig. 3. The spark-off effect was also found to be satisfactorily reproducible. Certain parameters were varied in an attempt to optimise the conditions but did not appear to influence the sensitivity to any large extent. As evidence existed (see the atomic-fluoresc- ence section) that the particles were not completely vaporised and atomised in the air - acetylene flame, both a hotter nitrous oxide - acetylene flame and a cooler hydrogen - oxygen - argon flame were used, with no significant effect on the size of the absorption signal.The gas atmosphere in the spark chamber also appeared to have very little influence on the A time constant of 1 s was used.April, 1976 DEVICE FOR SOLID SAMPLES IN ATOMIC SPECTROMETRY 267 absorption signal (approximately 1 1 min-l was passed through the chamber). In air (6-mm spark gap) the burn spots were small and the material in the centre appeared to have been melted, while argon (11-mm spark gap) produced a larger and more uniform burn spot without any signs of molten material. In spite of this visible difference the sensitivities obtained with both gases were almost equal. The spark parameters were varied by using capacitance values of 3 000, 6 000 and 9 000 pF and, for each of these, inductance values of 11, 75 and 160 pH.Again no significant change in sensitivity was noted. A tube, 70 cm long, between the spark and flame gave no lower sensitivity than a 10-cm tube, thereby indicating that there was no adherence of material to the tube walls. The polarity of the sample in the spark was also found to be unimportant, probably because of the oscillatory nature of the discharge. - Time Amount of copper in aluminium, % Fig. 3. Three sequential recordings of the Fig. 2. Atomic-absorption analytical copper absorption signal a t 324.8 nm. Time graph for copper in aluminium, using a scale, 1 min per division. Aluminium sample 10-cm air - acetylene flame. Spark para- containing 2.5% of copper. meters: C = 6 000 pF, L = 75 p H , V = 12 kV.Atomic-fluorescence Measurements A demountable, water-cooled, hollow-cathode lamp with boosted output,ll which was based on the design of Sullivan and Walsh,12 was used as a primary source for atomic-fluoresc- ence measurements. The flame used was a slightly fuel-rich hydrogen - oxygen - argon flame on a circular burner. The analytical graph obtained for copper in aluminium is shown in Fig. 4. In this instance a detection limit of 0.1% of copper was found (equal to that obtained with the atomic-absorption measurements). The working graph does not pass through the origin owing to a residual signal caused by reflection of the incident light by solid particles in the flame. These are probably nebulised particles, which are too large to be atomised in the flame.The noise of the background signal at a wavelength of 322.0nmJ which was adjacent to the analytical line, was a factor of 10 lower than the noise of the back- ground signal at 324.8 nm. From this it was inferred that a detection limit of 0.01 yo could be obtained in the absence of reflecting particles in the flame. Emission Spectrometry Using Inductively Coupled Plasma Excitation The simultaneous determination of several elements by emission spectrometry, as opposed to single-element determinations by atomic-absorption or atomic-fluorescence spectrometry, has distinct advantages for certain applications. In order to study the possibility of using the spark as a nebuliser in emission spectrometry, an inductively coupled plasma was ~ ~ e d . ~ This type of flame, with its high temperature, has proved to be a suitable source for many applications, and is generally used for the analysis of solutions.A study of alternative meth- cds of aerosol generation, Le., by hydride formation, thermal atomisation and remote solid aerosol generation using an arc source, has been made.15 In the present work, the low flow- rate pneumatic nebuliser that is normally used for solution analysis was simply removed and268 HUMAN et al.: USE OF A SPARK AS A SAMPLING - NEBULISING Analyst, VoZ. 101 Amount of copper in aluminium, % Fig. 4. Atomic-fluorescence analytical graph for copper in aluminium at 324.8 nm. Hydrogen - oxygen - argon flame cell, boos ted-ou tpu t hollow-ca thode lamp primary source. Spark parameters : C = 6000 pF, L = 75 pH, V = 12 kV.replaced by the spark chamber. A stabilised argon flow of 1 .O 1 min-l was passed through the spark chamber and this flow carried the nebulised material into the inductively coupled plasma via the injector tube. Previously analysed aluminium alloy and brass samples were used for this investigation. Spark-off graphs were recorded for a number of the elements in the aluminium alloy samples for various parameter settings of the spark source. A factorial design method was used to establish whether any significant trends occurred when the capacitance was varied between 3 000 and 12 000 pF, and the inductance between 11 and 160 pH. The spark gap was held constant at 11 mm. The elements studied were copper, magnesium and zinc. Typical spark-off graphs are shown in Fig.5. Except for the matrix element, aluminium, all of the graphs exhibited an initial decrease in intensity. With low inductance settings the graphs were monotonic, reaching an apparently stable level after 2-5 min, depending on the element and the capacitance. Higher inductance settings caused most of the curves to exhibit a distinct minimum after the initial decrease, followed by an increase to an apparently stable level. With high capacitance and low inductance settings, the initial decrease in intensity took place somewhat faster, resulting in apparently stable signals after 2-3 min. High capacitance and high inductance settings resulted in a longer time to reach stability. A comprehensive investigation into the effect of spark-source parameters on the emission characteristics of the plasma after stability had been reached was conducted.The signal to background ratio, as well as the signal noise (expressed as a percentage of the net signal), was measured for each element for various values of capacitance and inductance. The optimum spark-source parameters were established from the results obtained. The choice of capacitance setting was made difficult by the absence of definite trends for all the elements studied. For example, the signal to background noise ratio for magnesium (at a concentration of 0.1%) at 11 pH increased with increasing capacitance, while that for zinc and aluminium decreased. It was obvious, however, that an increase in the inductance value increased the signal to background noise ratio for low capacitance values.For high capacitance values the 75 pH inductance value generally appeared to produce the highest sensitivity. Possibly it could be reduced further by increasing the volume of the spark chamber. The signal to back- ground ratio generally appeared to increase with increasing inductance, except in the instance of magnesium at 0.1% when using a high capacitance. It was concluded from the above observations that for each element a particular set of spark- source parameters should be selected in order to obtain the most desirable analytical conditions. No significant change occurred with copper and magnesium (at 4.95%). In most instances the signal noise decreased when the inductance was increased.April, 1976 DEVICE FOR SOLID SAMPLES I N ATOMIC SPECTROMETRY 269 cu Mg (0.1%) Zn 3 000 pF 11 pH 3 000 pF 160 pH I 12 000 pF 160 pH t 12 000 pF 11 pH 0 2 4 6 0 2 4 6 0 2 4 6 Tirne/min - Fig.5 . excitation of sputtered material. 213.9 nm. Where a compromise is necessary for multi-element analysis, a capacitance of 9 000 pF, an inductance of 75 pH and a pre-spark time of 2 min appeared to be most suitable. The calibration graph that was obtained for copper in aluminium alloys is shown in Fig. 6. The plotted response was obtained by integrating the signal for 90 s following the 2-min pre- spark period. A sample with a high zinc content (4.3%) was included, and it can be seen that the copper emission intensity increased somewhat compared with a sample not containing zinc, thus giving an indication of a matrix effect that could have originated in the nebulising process.Spark-off curves for copper, magnesium and zinc, using an inductively coupled plasma for the The spectral lines used were: Cu, 327.4 nm; Mg, 279.6 nm; and Zn, 1 I I I I 0 0.1 0.2 0.3 0.4 C Amount of copper in aluminium, % Fig. 6. Inductively coupled plasma-emission analytical graph for the determination of copper in aluminium alloys. A 2-min pre-spark was used.270 HUMAN et al.: USE OF A SPARK AS A SAMPLING - NEBULISING Analyst, Val. 101 The precision of the method was determined by re-surfacing the samples a number of times and repeating the measurements. The relative standard deviation of the response for nine readings per sample varied between 4 and 6% for samples containing 0.5-0.05% of copper.From the calibration graph, the analytical precision was determined to be about 10% in the region 0.5-0.05~0 of copper in aluminium. A calibration graph obtained for iron in six brass samples is shown in Fig. 7. The iron line at 371.99 nm was used. I I I I I 0 0.05 0.1 0.1 5 0.2 0 Amount of iron in brass, % !5 Fig. 7. Inductively coupled plasma emission analytical graph for the determination of iron in brass samples. The detection limits obtained for a few elements in aluminium alloys are given in Table I. As a pure aluminium disc was not available, the background noise was measured adjacent to the spectral line of the element concerned. The relative standard deviation of the background was found to be less than that of the response at the wavelength of the analysis line for low concentrations of analyte.For example, the relative standard deviation of the background adjacent to 327.4 nm was o.4y0, whereas that of the response at 327.4 nm was 3.8% for a sample containing only 0.04% of copper. Because signal integration was used, the detection limits given in Table I were thus defined as the concentration necessary to produce a signal equal to 3 times the standard deviation of the background when measured adjacent to the spectral line used. TABLE I DETECTION LIMITS IN ALUMINIUM ALLOYS USING INDUCTIVELY COUPLED PLASMA EXCITATION Element Detection limit, p.p.m. Copper . . .. .. 3 Magnesium . . .. .. 2 Zinc . . .. .. .. 150 Discussion The results obtained in carrying out this investigation indicate that the amount of material sputtered from solid materials by a conventional spark discharge is adequate for their analysis by either flame (absorption or fluorescence) or plasma-emission spectrometry.The compara- tively large spot size of the discharge also ensures representative sampling from the surface. The use of the inductively coupled plasma offers high sensitivity and the possibility of simul- taneous multi-element analysis. An advantage of atomic-absorption and -fluorescence techniques, on the other hand, is their relative freedom from spectral interferences ; thus, a low-resolution, low-cost monochromator suffices. A disadvantage of these techniques is that normally only single element determina- tions can be made. This drawback may be overcome in atomic-fluorescence spectrometry by utilising a continuum source of excitation.l6April, 1976 27 1 A possible method of increasing the sensitivity of detection is to increase the spark repetition rate.A controlled spark repetition rate of up to 1 000 Hz is feasible,17 while the use of a free- running Enns-type spark may also be considered.18 The possibility of nebulising more material per spark also exists, for example, in the liquid-layer technique. Barnes and MalmstadtlS found that the volume of sample removed from an aluminium surface increased by a factor of 60 (approximately 10 for stainless steel) with this technique. A combination of these two techniques may lead to significant improvements in the detection limits. The time of analysis required with this system is lengthy, mainly because of the long pre- spark time that is necessary in order to obtain stability. The pre-spark time is decreased when a narrower spark gap is used; the ll-mm spark gap in argon requires a 2-min pre-spark time, while the 6-mm gap in air requires only 1 min of pre-spark time.The reason for this difference in time is that the smaller gap samples a smaller area, with the result that the surface layer is more rapidly prepared. The use of a high repetition rate spark will also overcome the disadvantage. This preliminary investigation, which was carried out originally in order to obtain qualita- tive information on the possible use of a spark as a sampling and nebulising agent for atomic spectrometry, has opened up some interesting possibilities. The results are encouraging, and further investigations should be made to develop this potential.A subject of particular interest is the particle size of various materials as a function of the matrix element, the spark parameters and the gas environment. The atomisation efficiency of these particles in the flame or plasma is also of considerable interest, while matrix and mutual interference effects should also be investigated. Evidence of such effects has already been noticed, e.g., an enhancement of the copper analytical signal when the zinc concentration is high, as observed in the plasma study (see Fig. 6), and a depression for high magnesium content, as observed in the atomic-absorption and -fluorescence studies ; the 0.5% copper analytical signal is below the analytical line in Figs.2 and 3 as a result of this sample containing 5% of magnesium, com- pared with 1.5% or less for the other samples. The premature curvature of the atomic- absorption and -fluorescence analytical graphs may also signify a matrix effect. Regarding the practical aspects of the technique, consideration may be given to the use of pressed discs made up of a mixture of a matrix element, e.g., copper powder, and powders of materials to be analysed, including non-conducting powders such as rocks, soils, etc. DEVICE FOR SOLID SAMPLES I N ATOMIC SPECTROMETRY 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. References Monvoisin, J., and Mavrodineanu, R., Spectrochim. Acta, 1951, 4, 396. Mavrodineanu, R., and Boiteux, H., “Flame Spectroscopy,” John Wiley & Sons, New York, Kantor, T., and Pungor, E., in “Colloquium Spectroscopicum Internationale XVII, Florence, 1973, Langmyhr, F.J., Thomassen, Y., and Massoumi, A., Analytica Chim. Acta, 1973, 67, 460. Fuller, C. W., Analytica Chim. Acta, 1972, 62, 261. Jones, J . L., Dahlquist, R. L., and Hoyt, E., Appl. Spectrosc., 1971, 25, 628. Winge, R. K., Fassel, V. A., and Kniseley, R. N., Appl. Spectrosc., 1971, 25, 636. Fassel, V. A., in “Colloquium Spectroscopicum Internationale XVI, Heidelberg, 1971, Proceedings,” Scott, R. H., and Kokot, M. L., Analytica Chinz. Acta, 1975, 75, 257. Scott, R. H., Fassel, V. A., Kniseley, R. N., and Nixon, D. E., Analyt. Chem., 1974, 46, 75. Human, H. G. C., Zeegers, P. J. Th., and van Elst, J. A., Spectrochim. Acta, 1974, 29B, 111. Sullivan, J. V., and Walsh, A,, Spectroclaim. Acta, 1965, 21B, 721. Greenfield, S., aild Smith, P. B., Analytica Chim. Acta, 1972, 59, 341. Nixon, D. E., Fassel, V. A,, and Kniseley, R. N., Analyt. Chem., 1974, 46, 210. Dahlquist, R. L., Knoll, J. W., and Hoyt, R. E., Reprint No. 7002 of paper presented a t Pittsburg Conference, 1975, Hassler Research Center, ARL, Goleta, Calif. 93017, USA. Johnson, D. J., Plankey, F. W., and Winefordner, J. D., Analyt. Chem., 1974, 46, 1898. Schroeder, W. W., Strasheim, A., and van Niekerk, J. J., Dev. Appl. Spectrosc., 1972, 10, 260. Enns, J. H., and Wolfe, R. A., J . Opt. SOC. Am., 1949, 39, 298. Barnes, R. M., and Rlalmstadt, H. V., Annlyt. Chem., 1974, 46, 66. 1965, pp. 108-110. Proceedings,” p. 83. Adam Hilger Ltd., London, 1972, p. 63. Received May 29th, 1975 Amended September 26th, 1975 Accepted November 28th, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100265

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

272 Analyst, April, 1976, Vol. 101, pp. 272-277 Determination of Submicrogram Amounts of Mercury in Various Matrices by Flameless Atomic- fluorescence Spectrometry P. Cavalli and G. Rossi Chemistyy Division, Euratom-CCR, 21020-Ispra ( Vurese), Italy An atomic-fluorescence spectrometric method is described for the determina- tion of mercury at the nanogram level. Solid samples are burnt in an oxygen stream, the combustion gases being condensed in a liquid nitrogen cooled trap. Following the dissolution of the condensed matter, mercury is extracted by the reduction - aeration method and collected on gold, forming an amalgam. The combination of a high-efficiency light-gathering system with a gas-shielded windowless fluorescence cell , and the described procedure for the release of mercury vapour from the amalgam, allows a detection limit of 0.03 ng to be obtained.A linear working graph covering the concentration range of mercury from 0.5 to 100 ng was established with aqueous standard solutions of mercury. The whole procedure has been checked with two different NBS Standard Reference Materials, excellent agreement of the measured values with the certified values being found. As a consequence of the increasing interest in assessing mercury contamination accurately, this laboratory has been faced with the problem of the determination of mercury in matrices of various natures and at concentrations very often below the parts per million level. Although flameless atomic-absorption spectrometry, in combination with the cold-vapour generation technique, is the most widely accepted and used method of carrying out these determinations,l atomic-fluorescence spectrometry shows some unique characteristics, which make this technique more attractive for routine applications.In the authors’ opinion, compared with atomic absorption, atomic fluorescence requires much simpler and less expensive instrumentation, exhibits a wider dynamic range because of the linear relationship between the fluorescence signals and the concentration of the fluo- rescent species, has a sensitivity that is not limited by the Lambert - Beer law but only by instrumental factors and exhibits less susceptibility to interferences from spurious vapours. Determinations of mercury by use of atomic fluorescence have been described by Wine- fordner and Staab2 and Vickers and MerrickJ3 using a flame as an atom reservoir, and by Muscat et al.,4 Muscat and Vickers5 and Thompson and Reynolds,B using a flameless cold- vapour technique.In these latter instances the results that have been presented indicate that further improvements could be achieved by a better designed and operated fluorescence cell and light-gathering system. Moreover, a mercury vapour generation system had to be constructed such that the fluorescence peak shapes would not be dependent on the nature of the sample (liquid or solid). In other words the mercury vapours to be introduced into the fluorescence excitation area should be generated and transferred in a specific and repro- ducible way. Experimental Apparatus Fig. 1 shows the layout of the instrument. The light emitted by a Philips 90-W low- pressure mercury vapour discharge lamp is focused by means of a short-focus silica lens in front of, and at a distance of 5 cm from, the entrance slit of a 10-cm focal length mono- chromator (Spex Micromate).In order to increase the excitation energy, a spherical mirror (MI), placed on the same optical axis, reflects the incident radiation back into the excitation area. An ellipsoidal mirror (M,) with two foci is placed in front of the entrance slit of the monochromator, the first focus being placed at the centre of the excitation area and the second one coinciding with the monochromator entrance slit. The dimensions and characteristicsCAVALLI AND ROSS1 273 of this optical system, together with details concerning its effectiveness in increasing the fluorescence signal, have been presented in a previous paper.The fluorescence intensity at 253.7 nm (isolated by the monochromator) is monitored with a photomultiplier (RCA 1Y28) and associated electronics (Keithley 414s d.c. amplifier and Hewlett-Packard 7127A strip-chart recorder). voltage amplifier Read-out Fig. 1. Schematic diagram of the instrument. A, mercury vapour discharge lamp ; B, windowless cell ; L, focusing lens ; MI, spherical mirror; and M,, ellipsoidal mirror. It should be pointed out that the use of a monochromator is not essential for the best results to be obtained; a non-dispersive system that is based on an interference filter that transmits at the wavelength of fluorescence is probably a more convenient basis for this type of measurement.In fact, slit width studies on the signal and on the signal to background ratios showed a behaviour similar to that described by Muscat and Vicker~,~ i.e., a linear increase in the signal intensity with increase in the slit width is obtained, while the signal to background ratio tends to stabilise at full slit width (2 000 pm). With the present equipment, a slit width of 1000 pm, corresponding to a band pass of 8 nm, has been selected. Under these conditions, a signal to background ratio of about 23 has been obtained on sweeping 20 ng of mercury into the excitation area. The Fluorescence Cell Muscat and Vickers5 used a Pyrex vapour cell with Vycor windows, within which the fluo- rescence excitation process took place; however, it was felt that the arrangement described by Thompson and Reynolds,6 consisting of a simple Pyrex tube connected to the mercury vapour generation system, would be more attractive because of its better suitability to the chosen optical system.Moreover, with the latter arrangement no possibilities exist for the excitation light beam to be scattered into the monochromator and no optical windows that can be fogged by any vapours that may be present are involved. The recorded fluorescence signal should have a more regular peak shape because of the shorter residence time of the mercury vapour in the excitation area. However, it is well known that air strongly de- presses the fluorescence signal, due to quenching of the excited mercury atoms, mainly by oxygen, but to a lesser extent by nitrogen molecules, so a greater efficiency might be ex- pected if mixing of the argon-entrained mercury vapours with air could be avoided in the excitation area. Therefore, following a series of tests with different geometries, the mercury vapour sprayer represented in Fig.2, incorporating a gas shielding system, has been utilised in this work. The item shown is conveniently made in black painted Perspex. With the geometry shown, laminar flow is obtained for both the shielding gas and the argon carrier, the mixing of the two gas streams taking place about 3 cm above the sprayer tip, provided that the flow of274 CAVALLI AND ROSS1 : DETERMINATION OF SUBMICROGRAM AMOUNTS OF An.aZyst, Vd. 101 the shielding gas is kept in the range from 5 to 7 1 min-l at an outlet pressure of 1.8-3.5 atm.The effectiveness of the gas shielding is clearly borne out by the working graphs obtained both without the gas shielding and with helium, nitrogen and argon as shielding gases (argon always being used as the mercury vapour carrier), which are shown in Fig. 3. With argon in use as the shielding gas, a 28-fold increase in the intensity of the fluorescence signal has been obtained. 26 r - - - l Fig. 2. Gas-shielded mercury vapour sprayer, manufactured from Perspex. Dimensions are given in millimetres The effects of the inlet pressure and the flow-rate of the argon carrier gas on the fluorescence signals have been carefully investigated in order to establish the most appropriate working conditions.Increased fluorescence signals have been obtained by increasing the argon flow-rate; however, steeper graphs are produced by decreasing the gas inlet pressure (Fig. 4). This behaviour could be related to the particular procedure used for transferring the mercury vapour to the fluorescence cell (to be discussed in the next section) and to the more con- centrated (undiluted with argon) cloud of mercury vapour submitted to the excitation pro- cess when lower argon pressures are used. An argon pressure of 0.1 atm at a flow-rate of 3 1 min-l has been chosen for all of the subsequent experiments. Mercury Vapour Generation When solid samples are to be analysed to determine the content of mercury, two basic procedures are generally employed, i.e., oxygen combustion of the samples and entrapment of mercury on a suitable amalgam-forming metal (gold or silver), or dissolution of the samples followed by use of a tin(I1) reduction method.The merits and disadvantages of both of these procedures have been exhaustively discussed in the literature.8-21 On the basis of extensive investigations that have been carried out in this laboratory, it could be concluded that by the first procedure interfering substances can be trapped on the amalgam-forming metal, causing incomplete recovery of mercury or spurious signals, parti- cularly when atomic absorption is used as the measurement technique, However, mercuryApril, 1976 MERCURY BY FLAMELESS ATOMIC-FLUORESCENCE SPECTROMETRY 275 0 1'2 20 30 40 50 f Amount of me:cury/ng Fig. 3. Effect of gas shielding on mercury fluorescence intensities.losses have been experienced during the dissolution of the samples, for which heating is required, and blank values exceeding by far the mercury concentrations sought are likely to be introduced by the rather large amounts of different reagents introduced during the whole of the chemical procedure. Moreover, it should be added that each of the procedures described in the literature has been specifically developed for a particular class of sample and cannot be considered as being generally applicable. The procedure that has been established in this laboratory represents a combination of both the combustion and the dissolution techniques, minimising the difficulties discussed above. A solid sample, placed in a platinum boat, is burnt in a stream of oxygen in a silica tube that is heated externally by a tubular oven (conditions: silica tube length, 25 cm; internal diameter, 2.5 cm; oxygen flow-rate, 3-5 1 min-l; combustion temperature, 800 "C). The combustion gases are passed over hot platinum wires, to help oxidation, and subsequently condensed in a glass trap cooled with liquid nitrogen in a similar manner to that described by ' O t 0.1 atm 6o t 3 atrn X- 40 3G 20 10 I 0 1 2 3 4 5 G 7 8 Carrier 33s fIow-rate/i min-' Variation of the fluorescence intensity of 10 ng of mercury with argon carrier gas flow-rate at different gas pressures.Fig. 4.276 CAVALLI AND ROSSI : DETERMINATION OF SUBMICROGRAM AMOUNTS OF Analyst, VoZ. 101 Aston and Riley.22 In addition to depending on the nature of the sample, combustion con- ditions must be varied in such a way as to ensure complete oxidation of the sample.For instance, a copper(I1) oxide section in the combustion train may be required for organic materials. The glass trap is brought up to room temperature and the condensed substances are dis- solved by adding 10ml of high-purity 10% nitric acid. Continuous stirring and heating (at 70 "C) of the solution for 10 min are recommended. The resulting solution is then sub- mitted to the tin(I1) reduction - aeration procedure in order to liberate the mercury vapour, argon being used as the carrier gas. The gas stream is dried by passing it through a U-tube packed with calcium chloride and magnesium perchlorate and the mercury vapour is collected on gold, as an amalgam, while other volatile compounds go to waste.No losses or memory effects are caused by the drier provided that it is changed frequently, i.e., after 15-20 deter- minations. Following the complete extraction of mercury from the solution (after about 4 min of argon bubbling), the amalgam is further flushed with argon for 1 rnin in order to com- plete amalgamation. The reaction flask is isolated from the gas line by means of a by-pass system, then the gas flow is stopped and the amalgam heated at 700 "C by use of the tubular oven for 2.5 min. The heating time has been found to be critical if the greatest reproduci- bility is to be obtained. The gas flow is then re-started and the mercury vapour cloud generated from the amalgam swept instantaneously into the windowless fluorescence cell.In this way minimum dilution of the mercury vapour with argon takes place, as is demon- strated by the sharpness of the peak recorded and the absence of any tailing. The mechanism of vapour transfer also explains the role played by the pressure and flow of the carrier gas as well as by the shielding gas. With the exception of the combustion step, the same procedure can be applied to the ana- lysis of liquid samples provided that they have been suitably treated. Results and Discussion Working graphs have been obtained by using aqueous mercury standard solutions freshly prepared from a 1000 pg ml-l mercury(I1) stock solution. Both the stock and diluted mercury solutions were acidified with a few drops of high-purity nitric acid.The aliquots (200 p1) of the standard solutions were directly introduced into the reduction - aeration flask, together with 50 ml of double-distilled water and 5 ml of a 10% m/V solution of tin(I1) chloride in 2 M hydrochloric acid in such a way as to cover a range of mercury content froin 0 to 100ng. Before the addition of the mercury standard solutions the tin(I1) chloride solution contained in the reaction flask was purged with argon for 4 min in order to eliminate blanks associated with the chemicals used. As an alternative, the same mercury standard solutions were placed in the platinum boats and carried through the same procedure as was used for the real samples. The two calibration procedures gave coincident values, resulting in a linear working graph.This fact indicates that the combustion and the vapour cold-trapping technique did not result in measurable mercury losses and that no depressive interferencell caused by volatilised platinum from the sample boat was found. In order to evaluate both the accuracy and the applicability of the proposed procedure, two standard reference materials from the National Bureau of Standards were analysed. These standards had certified mercury concentrations and completely different matrices. Five replicate determinations, using different amounts, were carried out on each sample, the appropriate results being shown in Table I. Sensitivity and Precision The limit of detection, expressed as three times the standard deviation of the noise asso- ciated with the signal, was equivalent to 0.03ng.In practice, mercury determinations at levels below 0.5 ng could not be carried out, this limit being set by a blank value that could not be eliminated rather than by instrumental limitations. This blank was found to be equivalent to 0.25 ng of mercury. The relative standard deviation calculated from a series of ten replicate determinations of 20 ng of mercury (in the form of an aqueous standard solution) was found to be 4.5%. However, it must be stressed that the reproducibility is almost independent of the concentration of mercury. In fact, equivalent signal variations were observed for concentrations of mercury both below and above 20 ng.April, 1976 MERCURY BY FLAMELESS ATOMIC-FLUORESCENCE SPECTROMETRY 277 Conclusions The method described in this paper has been applied for 2 years in this laboratory to the determination of mercury in a wide range of samples, including rocks, sediments, sewage waters, plants and vegetables, showing its applicability to routine analysis.Taking into account the very high sensitivity that has been obtained, precautions must be taken to operate the instrument under carefully controlled environmental conditions in order to minimise any possibility of its contamination by mercury. TABLE I COMPARISON OF FLAMELESS ATOMIC-FLUORESCENCE DATA WITH NBS VALUES FOR MERCURY DETERMINATION IN TWO NBS STANDARD REFERENCE MATERIALS Sample NBS SRM 1571 (orchard leaves) NBS SRM 1630 (trace mercury in coal) Concentration of mercury, p.p.m. Amount of r Replicate samplelmg Found NBS value 1 130.7 0.141 2 129.3 0.139 3 112.6 0.152 4 131.0 0.152 5 136.5 0.145 Mean value 0.146 Standard deviation 0.006 0.155 105.2 0.121 345.9 0.108 207.5 0.134 105.0 0.119 210.4 0.107 Mean value 0.118 Standard deviation 0.01 1 0.11* * New value recently set by NBS.2s 1.2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. References Ure, A. M., Analytica Chim. Acta., 1975, 76, 1. Winefordner, J. D., and Staab, R. A., Analyt. Chem., 1964, 36, 165. Vickers, T. J., and Merrick, S. P., Talanta, 1968, 15, 873. Muscat, V. I., Vickers, T. J., and Andren, A., Analyt. Chem., 1972, 44, 218. Muscat, V. I., and Vickers, T. J., Analytica Chim. Acta, 1971, 57, 23. Thompson, K. C., and Reynolds, G. D., Analyst, 1971, 96, 771. Benetti, P., Omenetto, N., and Rossi, G., Appl. Spectrosc., 1971, 25, 57. Iskandar, I. K., Syers, J. K., Jacobs, L. W., Keeney, D. R., and Gilmour, J. T., Analyst, 1972, 97, Skare, I., Analyst, 1972, 97, 148. Head, P. C., and Nicholson, R., Analyst, 1073, 98, 53. Ure, A. M., and Shand, C . A., Analytica Chim. Acta, 1974, 72, 63. Lidums, V., and Ulfvarson, U., Acta Chem. Scand., 1968, 22, 2150. Goleb, J. A,, Appl. Spectrosc., 1971, 25, 522. O’Gorman, J. V., Suhr, N. H., and Walker, P. L., jun., AppZ. Spectrosc., 1972, 26, 44. Henry, H. G., Stever, K. R., Barry, W. L., and Heady, H. H., Appl. Spectrosc., 1972, 26, 288. Thomas, R. J., Hagstrom, R. A., and Kuchar, E. J., Analyt. Chem., 1972, 44, 512. Rains, T. C., and Menis, O., J . Ass. Ofl. Analyt. Chem., 1972, 55, 1339. Willford, W. A., Hesselberg, R. J., and Bergman, H. L., J . Ass. Off. Analyt. Chem., 1973, 56, 1008. Agemian, H., and Chau, A. S. Y . , Analytica Chim. Acta, 1975, 75, 297. Joensuu, 0. I., Appl. Spectrosc., 1971, 25, 526. Bretthauer, E. W., Moghissi, A. A., Snyder, S. S.. and Mathews, N. W., Analyt. Chem., 1974, 46, 445. Aston, S. R., and Riley, J. P., Analytica Chim. Acta, 1972, 59, 349. National Bureau of Standards, personal communication. Received June 30th, 1976 Accepted October 20th, 1975 388.

ISSN:0003-2654    

DOI:10.1039/AN9760100272

出版商:RSC    

年代:1976

数据来源: RSC