ISSN:0003-2654    

DOI:10.1039/AN97196FX037

出版商:RSC    

年代:1971

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97196BX039

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

iv THE ANALYST [October, 1971THE ANALYSTEDITORIAL ADVISORY BOARDChairman; H. J. Cluley (Wernbley)*T. Allen (Bradford)*L. S. Bark (Salford)M. T. Kelley (U.S.A.)W. Kemula (Poland)*G. F. Kirkbright (London)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)*G. Nickless (Bristol)*J. M. Ottaway (Glasgow)*G. E. Penketh (Billingham)R. Belcher (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)*R. C. Chirnside (Wembley)A. C. Docherty (Billingham)D. Dyrssen (Sweden)*W. T. Elwell (Birmingham)*D. C. Garratt (London)*R. Goulden (Sittingbourne)S. A. Price (rodworth)D. I. Rees (London)E. B. Sandell (U.S.A.)A. A. Smales, O.B.E. (Harwell)H. E. Stagg (Manchester)E. Stahl (Germany)A. Walsh (Australia)T. S.West (Lcrtdon)P. Zuman (U.S.A.)j. Hoste (Belgium)D. N. Hume (U.S.A.)*J. A. Hunter (Edinburgh)H. M. N. H. Irving (Leeds)A. G. Jones (Welwyn Garden City)*Members of the Board serving on the Executive Committee.NOTICE TO SUBSCRIBERSSubscriptions for The Analyst, Analytical Abstracts and Proceedings should be(Other than members of the Society)sent through a subscription agent or direct to:The Chemical Society, Publications Sales Ofice,Blackhorse Road, Letchworth, Herts.Rates for 1972(a) The Analyst, Analytical Abstracts, and Proceedings, with indexes . . ..index), and Proceedings . . .. .. . . . . .. ..index), and Proceedings . . .. .. .. .. .. * .(d) The Analyst and Analytical Abstracts, with indexes , . .. .. ..index) .. . . . . . . .. .. .. .. .. ..index) .. .. .. .. .. ..(b) The Analyst, Analytical Abstracts printed on one side of the paper (without(c) The Analyst, Analytical Abstracts printed on one side of the paper (withThe Analyst and Analytical Abstracts without Proceedings-(e) The Analyst and Analytical Abstracts printed on one side of the paper {without(f) The Analyst and Analytical Abstracts printed on one side of the paper (with . . . . . . . .f 33.50 $80.40€3450 $82.80f40-5.0 $97.20€ 3 I -00 $74.40f 32.00 $76.80f38.00 $9 I .20(Subscriptions are NOT accepted for The Analyst and/or for Proceedings alone)Membcrs should send their subscriptions to the Hon. TreasureSUMMARIES OF PAPERS I N THIS ISSUE.Summaries of Papers in this IssueLiquid Scintillation Counting as an Analytical ToolA ReviewSUMMARY OF CONTENTSIntroductionBasic conceptsInstrumentationSolvents, scintillators and additivesTechniquesSample preparation for homogeneous systemsSample preparation for heterogeneous systemsSample preparation for Cerenkov countingStandardisationData handlingConclusionsAppendixIsotopes measured in a liquid scintillation counter[October, 1971REPRINTS of this Review paper will soon be available from The Society forAnalytical Chemistry, Rook Department, 9/10 Savile Row, London, W 1X 1 AF,a t 25p per copy, post free.A remittance for the correct amount, made out to The Society forAnalytical Chemistry, MUST accompany every order; these reprints are notavailable through Trade Agents.J. A.B. GIBSON and A. E. LALLYHealth Physics and Medical Division, A.E.R.E., Harwell, Didcot, Berkshire.Analyst, 1971, 96, 681-688.A Method for the Determination of Disodium Cromoglycateand Other ChromonesThe hydrolysis of disodium cromoglycate in alkaline solution has beenstudied and a method for its determination is presented. The method isbased on spectrophotometric measurement of the change in absorptiona t 310 nm, which takes place on opening the y-pyrone ring of the chromonenucleus. The technique is applicable, with suitable minor modifications, tothe determination of other compounds containing the chromoiie nucleus,such as flavones.J. TILLMAN and D. W. WHYMARKFisons Limited, Pharmaceutical Division, Research and Development Laboratories,Bakewell Road, Loughborough, Leics.Analyst, 1971, 96, 689-698.The Flame-photometric Determination of Alkalis inCeramic MaterialsThe flame-photometric method in genera1 use for the determination ofalkalis in the ceramic industries was originally devised for the EEL, Model 100,flame photometer, with a coal gas flame, which is not now generally available.Current supplies of town gas, propane and methane (natural gas) flames arecompared and interferences evaluated.Perchloric and hydrochloric acidsare found to have a depressant effect; sodium is enhanced by potassium andis subject to spectral interference by calcium.Propane is chosen as the preferred fuel. The effect of chlorine-containingacids is eliminated by the use of sulphuric and nitric acids for the initialdecomposition and the spectral interference from calcium by the additionof aluminium sulphate ; sodium - potassium inter-element effects are eliminatedby the use of a caesium buffer, Although the procedure is principally devisedto give optimum results with aluminosilicates, its extension to high-limematerials is also considered.R.P. EARDLEY and R. A. REEDBritish Ceramic Research Association, Queens Road, Penkhull, Stoke-on-Trent,ST4 7LQ.Andyst, 1971, 96, 699-711October, 19711 THE ANALYST vi i new liquid assetsTwo new liquid scintillators NE 232 and NE 235 have been added to the world's mostextensive range of solutions. I f your special requirements are still not covered by thescintillators listed below, our research team of chemists and physicists with overtwenty years experience in the field would welcome the challenge of helping you.NE 232: is based on deuterated cyclohexane (C6D12), and is useful for neutron detection wherea scintillator with a high D : C ratio is required.NE 235: This new low cost scintillator has a mineral oil base and is of particular interest for largevolume liquid scintillation counting.It has a high flash point, excellent light transmission proper-ties and does not attack 'Perspex', 'Lucite' or 'Plexiglas'. Light output 40% anthracene.Standard RangeFor Table of Physical Constants, see 1970 Catalogue, page 4. * i.e. Perspex, Lucite or Plexiglas. t D/CRatio.Copies of Bulletin 53 on liquid Scintiilators and Chemicals for Internal Countingare available free on request, along with twelve-page full-colour Brochure50 on Automatic Sample Changer Systems.Sighthill, Edinburgh EH11 4EY, Scotland.Tel031-443 4060 Cables: Nuclear, Edinburgh Telex: 72333...Vlll SUMMARIES OF PAPERS I N THIS ISSUE [October, 1971A Potentiometric Procedure for the Assay of IsonicotinicAcid Hydrazide (Isoniazid)A potentiometric procedure for the assay of isonicotinic acid hydrazide(isoniazid) with vanadium(V) a t room temperature is described. The reductionof vanadium(V) to vanadium(1V) by isoniazid in an acidic medium is cata-lysed by osmium tetroxide, and the application of this method to the assay ofisoniazid in pharmaceutical preparations is considered. Oxalic acid interferesin the determination, although commonly used excipients such as starch,dextrin, sucrose, glucose, lactose and gum acacia do not interfere.P.V. KRISHNA RAO and G. BALA BHASKARA RAOChemistry Department, Andhra University, Waltair, India.Auzalyst, 1971, 96, 712-715.The Determination of Nikethamide and Other Compounds inPharmaceutical Dosage Forms by Thin-layer ChromatographyThe qualitative examination and quantitative assay of pharmaceuticaldosage forms containing nikethamide by thin-layer chromatography isdescribed. When applicable, the determination of accompanying activecompounds such as adenosine, caffeine, strychnine and theophylline by thesame method is also described, Individual quantitative determinations ofthe eluted drugs are performed by ultraviolet spectrophotometry.W.M. CARMICHAELAnalytical Division, Ciba-Geigy Limited, 4000 Base1 2, Switzerland.Analyst, 1971, 96, 716-720.The Gas-chromatographic Determination of 2,3,7,8-Tetrachloro-dibenzo-p- dioxin in 2,4,5- Trichlorophenoxyacetic Acid (“2,4,5-T”),2,4,5-T Ethylhexyl Ester, Formulations of 2,4,5-T Esters and2,4,5 - TrichlorophenolA gas-chromatographic method for the determination of trace amounts ofa toxic impurity, 2,3,7,8-tetrachlorodibenzo-j~-dioxin, is described. A purifiedextract of the sample was subjected to gas chromatography on a columncontaining either 2 per cent. of OV-17 on Diatomite CQ or 1 per cent. ofHi-Eff 8 BP on Gas-Chrom 2 with electron-capture detection. 2,4,5-Tri-chlorophenoxyacetic acid and 2,4,5-trichlorophenol were purified by chromato-graphy of an ether extract of the sample on a column of alumina, followed byshaking with sulphuric acid.For 2,4,5-trichlorophenoxyacetic acid esters andformulations, saponification and chromatography on a Celite - sulphuric acidcolumn, followed by chromatography on a column of alumina, were necessary.Recoveries of 2,3,7,8-tetrachlorodibenzo-~-dioxin ranged from 89 to 98per cent., and the standard deviation of the method at a level of 0.3 p.p.m.was 0.03 p.p.m. The limit of detection was about 0.05 p.p.m.D. A. ELVIDGEQuality Control, Analytical Research, Boots Pure Drug Co. Ltd., Pennyfoot Street,Not t ingham.Analyst, 1971, 96, 721-727.Determination of Total Hydrolysable Nitrogen in AcidicAqueous Solutions of Nitriles Containing CyanideA method is described for the determination of total hydrolysablenitrogen in acidic aqueous solutions containing acrylonitrile, acetonitrile,hydrogen cyanide and ammonia. A recovery of about 97 per cent. is obtainedfor all the components examined. The method is based on the classicalRadziszewski reaction, and involves reaction with 30 per cent. w/v hydrogenperoxide followed by alkaline hydrolysis to ammonia. The important featureof the method is that it enables hydrogen cyanide to be determined togetherwith the other components.D. C. WHITEB.P. Chemicals International Limited, Research and Development Department,Great Burgh, Yew Tree Bottom Road, Epsom, Surrey.Analyst, 1971, 96, 728-733

ISSN:0003-2654    

DOI:10.1039/AN97196FP149

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

xii THE SOCIETY FOR ANALYTICAL CHEMISTRYREPORTS OF THE ANALYTICAL METHODS COM-MITTEEThe Reports of the Analytical Methods Committee listed below may be obtained direct from TheSociety for Analytical Chemistry, Book Department, 9/10 Savile Row, London, WlX 1AF (not through TradeAgents), a t the price of lop. to members of thc Society and 15p. to non-members. Remittances mustaccompany orders and be made payable to “Society for Analytical Chemistry.”Additives in Animal Feeding Stuffs Sub-committee :Certain Reports published before 1046 have been omitted from this list, but are still available.Report of the Antibiotics Panel : The Determination of Penicillin, Chlortetracycline and OxytetracyclineReport of the Hormones Panel : The Determination of Stilboestrol and Hexoestrol in CompoundReport of the Prophylactics Panel : The Determination of Nitrofurazone in Compound Feeding Stuffs.Report of the Vitamins (Water-soluble) Panel : The Determination of Water-soluble Vitamins inReport of the Vitamins (Fat-soluble) Panel : The Determiration of Fat-soluble Vitamins in DietReport of the Prophylactics in Animal Feeds Sub-committee: The Determination of Amprolium inReport of the Prophylactics in Animal Feeds Sub-committee : The Determination of Sulphaquinoxaline.Report of the Prophylactics in Animal Feeds Sub-committee : The Determination of Acinitrazole.Report of the Prophylactics in Animal Feeds Sub-committee : The Determination of Ethopabate inReport of the Prophylactics in Animal Feeds Sub-committee: The Determination of FurazolidoneReport of the Prophylactics in Animal Feeds Sub-committee : The Determination of Dimetridazole inReport of the Prophylactics in Animal Feeds Sub-committee : The Determination of DinitolmideReport of the Prophylactics in Animal Feeds Sub-committee : The Determination of Amproliuni,in Diet Supplements and Compound Feeding Stuffs.Feeding Stuffs.Compound Feeding Stuffs.Supplements and Compound Feeding Stuffs.Animal Feeding Stuffs.Feeds.in Feeds.Animal Feeds.(Zoalene) in Animal Feeds.Sulphaquinoxaline and Ethopabate when Present Together in Animal Feeds.Analytical Standards Sub-committee :Sodium Carbonate as a Primary Standard in Acid - Base Titrimetry.Sulphamic Acid as a Primary Standard in Acid - Base Titrimetry.Report No.15.Application of Gas - Liquid Chromatography to Essential-oil Analysis : Interim Report on theSpectral Characteristics of Eugenol.Fish Products Sub-committee :Nitrogen Factor for Cod Flesh.Meat Products Sub-committee (formerly Meat Extract Sub-committee) :Analysis of Meat Extract.Determination of Gelatin in Meat Extract and Meat Stocks: Interim Report.Nitrogen Factors for Pork and Nitrogen Content of Rusk Filler (as one reprint).Nitrogen Factors for Beef.Nitrogen Factors for Chicken.Nitrogen Factors for Liver.Nitrogen Factor for Veal.Nitrogen Factors for Turkey.Nitrogen Content of Rusk Filler.Nitrogen Factor for Kidney.Nitrogen Factor for Tongue.Nitrogen Factor for Barley.Nitrogen Factor for Blood.Metallic Impurities in Foodstuffs Sub-committee :Determination of Lead in Foodstuffs : Tentative Method.Metallic Impurities in Organic Matter Sub-committee :Methods for the Destruction of Organic Matter.Notes on Perchloric Acid and its Handling in Analytical Work.The Determination of Lead.The Determination of Small Amounts of Arsenic in Organic Matter.The Determination of Small Amounts of Copper in Organic Matter.The Determination of Small Amounts of Mercury in Organic Matter.The Determination of Small Amounts of Tin in Organic Matter.The Determination of Small Amounts of Zinc in Organic Matter.The Use of 50 per cent.Hydrogen Peroxide for the Destruction of Organic Matter.The Determination of Small Amounts of Tin in Organic Matter.The Determination of Small Amounts of Cadmium in Organic Matter.Essential Oils Sub-committee :Determination of Linalol in Essential Oils.Determination of Citronellol in Admixture with Geraniol.Part 1.Amounts of Tin u p to 30 p gPart 2. Amounts of Tin from30 to 150 pgxiv SUMMARIES OF PAPERS I N THIS ISSUE [October, 1971The Theoretical and Practical Aspects of Electronic Timingas One Method of Increasing the Analysis Rateon the Technicon AutoAnalyzer IThe need for an increase in the speed of analysis of soil and plant samplesis discussed with reference to the use of the Technicon AutoAnalyzer I. Areliable electronic timer that controls the operation of the Technicon Sampleris described. A circuit diagram of the timer is shown, with a list of componentsused.J.A. VARLEY and K. F. BAKERTropical Soils Analysis Unit, Land Resources Division, Overseas DevelopmentAdministration, Coley Park, Reading, Berkshire.Analyst, 1971, 96, 734-738.The pH Meter as a Hydrogen-ion Concentration Probe:A PostscriptMeasurements of the ratio between the quantity H (= 10-PH) and thetrue hydrogen-ion concentration in solutions of fixed ionic composition havebeen made in alkaline solutions, and are in agreement with values previouslydetermined in acidic solutions.W. A. E. McBRYDEDepartment of Chemistry, University of Waterloo, Waterloo, Ontario, Canada.Analyst, 1971, 96, 739-740.The Determination of Small Amounts of Copper in OrganicMatter by Atomic-absorption SpectroscopyReport prepared by the Metallic Impurities in Organic Matter Sub-Committee.ANALYTICAL METHODS COMMITTEE9/10 Savile Row, London, W1X 1AF.Analyst, 1971, 96, 741-743.Nitrogen Factor for Coal FishReport prepared by the Fish Products Sub-committee.ANALYTICAL METHODS COMMITTEE9/10 Savile Row, London, W1X 1AF.Analyst, 1971, 96, 744-745.The Determination of Dimetridazole in Animal Feeds :Revised MethodReport prepared by the Prophylactics in Animal Feeds Sub-committee.ANALYTICAL METHODS COMMITTEE9/10 Savile Row, London, WlX IAF.Analyst, 1971, 96, 746-749.Cracking of Combustion Tubes when using a Semi- automaticFlash- combustion ApparatusCommunicationM. A. LEONARD and W. J. SWINDALLDepartment of Analytical Chemistry, David Keir Building, The Queen's Universityof Belfast, Belfast, BT9 5AG, Northern Ireland.Analyst, 1971, 96, 749

ISSN:0003-2654    

DOI:10.1039/AN97196BP157

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

OCTOBER, 1971 THE ANALYST Vol. 96, No. 1147 Liquid Scintillation Counting as an Analytical Tool .A Review* BY J. A. B. GIBSON AND A. E. LALLY (Health Physics and Medical Division, A .E.R.E., Harwell, Didcot, Berkshire) SUMMARY OF CONTENTS Introduction Basic concepts Instrumentation Solvents, scintillators and additives Techniques Sample preparation for homogeneous systems Sample preparation for heterogeneous systems Sample preparation for Cerenkov counting Standardisation Data handling Conclusions Appendix Isotopes measured in a liquid scintillation counter INTRODUCTION LIQUID scintillation counting is commonly used for counting a wide range of p- and cc-emitting radioisotopes in many chemical forms. The technique provides for a good detection efficiency (up to about 100 per cent,), and normally involves a minimum of chemical preparation.The relatively high backgrounds obtained in this method compared with other detection systems, e.g., proportional counters, limit the sensitivity for some isotopes, particularly a-emitters. A second, more serious, limitation is the variable efficiency caused by a reduction in the light output (quenching) in the presence of certain chemical impurities. Other impurities may introduce chemilurninescence, which gives an unknown and variable background. These effects are most important for low-energy 8-emitters such as 3H and 14C. Reduction of quenching and chemilurninescence have been widely discussed throughout the literature, and many methods of measuring the efficiency have been devised. There already exist several reviews dating from the proceedings of a symposium in 1958, edited by Bell and Hayes,l and culminating in a second symposium (edited by Branksome2) in 1970.The proceedings of the latter will provide much up-to-date information for the experienced user of the technique. This more limited review cannot compete with the depth and coverage of such a symposium. We therefore aim to provide a critical introduction to the methods and materials used in liquid scintillation counting. The review is divided into four general sections, the basic concepts of the technique, instrumentation, the choice of scintillators and additives, and the techniques used in sample preparation and standardisation and analysis of the subsequent data. The information is finally summarised as a table in the Appendix, which gives an indication of the range of isotopes that can be counted with a liquid scintillation counter.The techniques that will not be discussed include the use of liquid scintillators for neutron measurements and as anti-coincidence shields. These and other techniques are discussed in detail by B i r k ~ . ~ BASIC CONCEPTS The basic organic liquid scintillator consists of a solution of one or more fluorescent aromatic solutes dissolved in an aromatic solvent (usually toluene or dioxan). Other com- pounds (additives) can be added to this basic solution so as to incorporate the various radioactive samples into the scintillator. * Reprints of this paper will be available shortly. For details see summaries in advertisement pages.0 SAC and the authors.682 GIBSON AND LALLY: LIQUID SCINTILLATION COUNTING [Artalyst, Vol. 96 #I * Fast coincidence < 41 P 1 . \ Photomultiplier Liquid scint iI lator 'Photomultiplier J. Pulse amplifier f Single channel analyser $ Gate* 4 = * 1 Solute - - - electrons ----- ,Photocathode , I * Pulseadder 4 \ I . J I I J. Pulse amplifier Pulse amplifier Y * Single Single $. I Gate* Gate* channel analyser channel a na I yser electrons ___f \ Dynode \ \ \ j'%k Electrons $. Scaler- readout channel 1 4 $. Scaler- readout Scaler- readout channel 2 channel 3 i * Gates only open when pulses are observed from both photomultiplier tubes., Fig. 1. Schematic diagram of the scintillation process and electronics Alpha and beta radiations from radioisotopes in the scintillator deposit energy in the solvent.This energy is transferred through the solvent until it reaches a solute molecule which converts the energy to a light quantum (Fig. 1). Light quanta from several solute molecules are detected by the photomultiplier tubes (usually two), which convert about 25 per cent. of the light quanta into electrons. The over-all quantum efficiency from the original particle to the production of electrons is about 0.2 per cent., and the deposition of 6 keV of electron energy (approximate mean energy for 3H /3-particles) results in only one or two electrons being produced in the photocathode of the photomultiplier tube. Because of the statistical nature of the proce~s,~ some events will not be observed at all, and therefore the maximum theoretical counting efficiency is less than 100 per cent.5 (typically 60 per cent.for SH with two tubes in coincidence). The electrons are accelerated through a series of dynode stages that produce between two and six electrons for each electron input. This multiplication may amount to a 105-fold increase over eleven stages. Spontaneous emission from the photocathode results in a background noise output from the photomultiplier tube. By using two photomultipliers in coincidence, the single-noise event in one tube only can be rejected, but excessive noise will result in accidental coincidences that can be reduced by cooling to below ambient temperature. The external background of gamma and cosmic radiation will produce events in the scintillator that can be reducedOctober, 19711 AS AN ANALYTICAL TOOL 683 by using:an anti-coincidence shield.Added materials may be chemiluminescent and produce single photons, but this effect is largely eliminated by using the coincidence system. Chemical impurities may interfere with the transfer of energy from solvent to solute to produce “chemical quenching,” or they may absorb the light emitted from the solute molecules to produce “colour quenching.” Both effects will reduce the light output of the scintillator6 and thus reduce the counting efficiency for the sample. The two types of quenching have different effects on the output spectrum of the counter, and chemiluminescence will give yet a third change in the spectrum. It is these spectrum changes and shifts that form the basis of most techniques for the determination of counting efficiency and the detection of chemi- luminescence.Chemical quenching can be avoided with high-energy /I-emitters (/Imax. > 0.15 MeV) by the use of Cerenkov counting. The sample is counted directly without a scintillator and the light output is produced when the speed of the /I-particles exceeds the speed of light in the medium.’ This technique, although simpler, is still subject to colour quenching and chemi- luminescence. These techniques are discussed later. INSTRUMENTATION A large number of complete instrument systems are commercially available for use with liquid scintillators. They vary in complexity from single-sample instruments to systems handling hundreds of samples with an output of efficiencies, corrections and the disintegration rates for up to three isotopes.The choice depends upon the application and the variety of isotopes encountered in a particular laboratory. The basic components of a typical coincidence system are shown in Fig. 1. The photo- multiplier tubes are normally contained in a temperature-controlled box, which also contains the sample changer. The choice of the temperature is a compromise between the lower backgrounds obtainable at 0 “C and the miscibility of some samples at reduced temperatures. At present, the majority of new systems operate at temperatures sljghtlybelow ambient to ensure that the sample is homogeneous. The system may contain a number of independent channels (usually three), each with a separate amplifier and channel-width controls.This enables two or perhaps three isotopes to be counted simultaneously. Some systems incorporate an auto- matic external source of y-radioactivity. Following an initial sample count, the source is automatically transferred from a shielded container to a position close to the sample and a second count in the three channels is obtained. This second count with the external source enables an estimate to be made of any quenching present; details of the method are discussed later. The output data are normally printed on to a paper tape and, if necessary, can be punched on to tape for computer analysis. In some machines the computer is built into the system and the processed results are produced directly. SOLVENTS, SCINTILLATORS AND ADDITIVES The purpose of the solvent is (a) to provide a medium for containing the sample, (b) to transport the energy from the source of radiation and (c) to contain the solutes and allow for the emission of light.The scintillation solute transforms the energy into light, but if the wavelength is unsuitable for the photomultiplier tube then a secondary solute can be used as a wavelength shifter. The two basic solvents used are alkylbenzenes, such as toluene and xylene, and aliphqtic ethers, the most common being 1 + 4 dioxan - water. Although the dioxan base produces a scintillator with a lower scintillation efficiency than that of the alkylbenzene bases, it has the advantage of being miscible with water. All solvents must be of the highest purity to avoid quenching effects which can be caused by trace amounts of impurities.Typical primary solutes are 2,5-diphenyloxazole (PPO) and 2-(4‘-t-butylphenyl)-5-(4”-bi- phenyl)-lJ3,4-oxadiazole (Butyl-PBD) . The two common secondary solutes (wavelength shifters) are 1,4-bis- (5-phenyloxazol-2-yl) benzene (POPOP) and 1,4-bis- (4-methyl-5-phenyl- oxazol-2-y1)benzene (DM-POPOP). The solutes must be soluble at the operating temperature of the system and also must not be precipitated by the addition of the sample, e.g., water. Radioisotopes are most commonly prepared for counting in aqueous solutions, so that water miscibility is an essential requirement in many scintillators. Dioxan-based systems are therefore most suitable for this purpose, but naphthalene is usually added to increase684 GIBSON AND LALLY: LIQUID SCINTILLATION COUNTING [A%a&Si!, Vol.96 the energy transfer from the solvent to solute and hence to increase the counting efficiency. Many popular dioxan-based scintillators originate from the Brays solution , comprising naphthalene, PPO, POPOP, methanol and ethylene glycol in dioxan. Ethanol and mono- methyl and monoethyl ethers of ethylene glycol can be used to reduce the freezing-point of dioxan scintillat 0 ~ s . ~ $lo Toluene-based scintillators can be diluted with various polar solvents, such as ethanol and methanol, to increase water miscibility. This system was designed mainly for organic solvents, but now that solubilisers are freely available, a wide variety of sample materials, e.g., blood, urine and biological tissue, can be counted in toluene-based systems.Typical solubilisers are hyamine hydroxide, NCS (Nuclear Chicago) and the “Bio-Solv” range (Beck- man Instruments). Other additives such as Triton X-100 enable emulsion counting to be performed in toluene systems.ll Gel counting is a technique to be used when the sample is insoluble in the scintillator or in any convenient solvent. Normal dioxan-based scintillators can be converted into gels by stirring in finely divided silica to provide a thixotropic phase in which the insoluble material is suspended. The choice between buying ready-made scintillators and preparing one’s own depends upon the number and variety of samples to be a n a 1 y ~ e d . l ~ ~ ~ ~ Shelf-life is an important consideration. Dioxan-based systems should be kept under nitrogen to prevent oxygen absorption and in tightly stoppered vessels to prevent loss of solvent and the subsequent crystallision of naphthalene, both of which reduce the counting efficiency.Toluene-based scintillators are not affected by either of these phenomena. TECHNIQUES This section covers the factors involved in selecting the technique, sample preparation (homogeneous, heterogeneous or for Cerenkov counting) , standardisation and data handling. Initially, when a new isotope or material is to be counted, various factors must be considered: (a) the energies and types of radiation emitted by the sample, e.g., a-, B- or y-radiation; (b) the chemical form of the sample and whether chemical preparation could improve the effectiveness of the method; (c) the choice of the scintillator system and the possible inclusion of additives to obtain miscibility, reduce quenching and improve the counting efficiency; (d) instrumental conditions to ensure adequate sensitivity and an accurate assessment of the efficiency and background; and (e) the method of data analysis should be included in the selection of the technique.After choosing the technique, it is finally necessary to ensure that conditions remain constant by regular checks with calibrated standards. SAMPLE PREPARATION FOR HOMOGENEOUS SYSTEMS- Sample preparation should normally be kept to a minimum so as to avoid losses from incomplete chemical recovery and isotope effects caused by the different properties of the natural and radioactive isotopes. Chemical impurities introduced by processing may also introduce unknown quenching effects.Aqueous systems can be added directly to dioxan-based scintillators and to toluene- based scintillators when a solubiliser is added.14J5J6 Biological materials such as blood, urine, salts, sugars and the alkali digests of plasma and tissue samples can be incorporated into toluene-based scintillators with Bio-Solv (BBSB and BBS3, Beckman Instruments Ltd.). BBS2 is an acid solubiliser for alkaline tissue digests and aqueous solutions, and it must be neutralised before counting. BBS3 is a general-purpose solubiliser for all types of aqueous samples and for blood and plasma.16 NCS (Nuclear Chicago Corporation) and Hyamine 1OX can also be used for tissues and purified biological material.17 However, the quaternary ammonium base of hyamine hydroxide is strongly chemi1uminescentl8 and should be used with caution.It is necessary always to use some met hod to check for chemical and colour quenching and chemiluminescence. Materials that are strongly luminescent or produce severe quenching require further treatment. Colour quenching can often be eliminated, or at least reduced, by digestion withOctober, 19711 AS AN ANALYTICAL TOOL 685 hydrogen peroxide and perchloric acid. This method has been used successfully for soft tissues, solid biological materialsfQ (e.g., teeth and bones) and for the determination of radio- activity on filter-papers.20 Isotopes investigated include S2P, S5SJ 45Ca, 55Fe and 57Co. Alternatively for 3H and 14C, complete combustion of the sample to water or to a soluble carbonate produces a simple counting method.The Schoniger oxygen-flask method21 was the forerunner of this technique and many modifications have been made.22 An automatic version of the instrument is now available commercially as a “Tritium Oxidiser” (Packard Instruments Ltd.). Recovery experiments to investigate the chemical yield and any isotope effect are an essential part of any combustion experiment. SAMPLE PREPARATION FOR HETEROGENEOUS SYSTEMS- Insoluble materials and other samples that cannot be processed chemically can be measured as suspensions in gels or as emulsions. If necessary, a solid support such as a filter-paper can be used. A reduction in counting efficiency may occur through self-absorption in particles, supports, etc., and standardisation may be difficult.Suspension counting in gels is useful for incorporating reasonable amounts of precipitates that are otherwise insoluble in liquid scintillators. A transparent gelling agent such as Cab-0-Sil, a finely divided silica powder, is mixed to give a thixotropic gel that is fluid when shaken but firm when at rest. This technique has been used for barium and strontium carbonates, perchlorates, etc., and for an iron ferriphosphate complex for the determination of 55Fe and 59Fe (Eakins and B~owII~~) and 239Pu and 241Pu (Eakins and L a l l ~ ~ ~ ) . Emulsion counting can be used for incorporating large volumes of aqueous samples into scintillators. An example of this technique is the use of Triton X-100 detergent with a toluene-based scintillator.12 This technique is very sensitive to pH, temperature and salt concentrations.SAMPLE PREPARATION FOR CERENKOV COUNTING- In the simplest form no sample preparation is required and the solution is placed directly into a counting phial. The efficiency is improved by effecting an increase in the refractive index (to reduce the energy threshold) and this may be necessary for p-emitters of lower energy. The introduction of a wavelength shifter will further increase the efficiency.’ Chemical quenching is eliminated in this method, but colour quenching and chemiluminescence are still important. Quenching can be reduced by using the decolorising techniques discussed above. Standardisation is carried out either by adding an internal standard or by using a high-energy y-emitter to produce photoelectrons in the solution.The major advantage of this technique is the high sensitivity with large volumes, or for flow monitoring without changing the liquid passing through the detector. STANDARDISATION- The counting efficiency for a high-energy /3-emitter can be nearly 100 per cent., and the effects of quenching are then small. If all samples in a particular experiment have the same composition, then a simple standardisation technique is all that is necessary. However, for low-energy /3-emitters in a wide range of materials, either chemical or colour quenching, or both, will normally be present. The three methods most commonly used are the use of an added internal standard, an automatic external standard25 (usually a long-lived y-ray emitter), and a channels ratio method.26 The advantages and disadvantages are summarised in Table I.The normal procedure is to choose the method, e.g., channels ratio, in which the efficiency is plotted against the ratio for a series of standards with different amounts of a quenching material. The ratio obtained during an experiment can then be converted into an efficiency by the use of this graph. Such a calibration curve is necessary for each isotope, scintillator and instrument setting. A similar technique is used with the external standard method, but the internal standard will give the efficiency directly for each sample. The background of the counter is also affected by quenching, and for low-level counting it is necessary to know the background for various values of the channels ratio.The presence of chemiluminescence introduces an increased variable background and can lead to erroneous results if it is significant compared with the radioactivity of the sample.686 GIBSON AND LALLY: LIQUID SCINTILLATION COUNTING [Analyst, Vol. 96 DATA HANDLING- The output from manual instruments can normally be analysed with a desk calculator. Automatic systems that process hundreds of samples per day can produce five or more items of information per sample, and manual processing becomes tedious. The simplest technique normally involves the use of a small desk-top computer,27 which can be programmed to take information about the sample (perhaps counts in three channels), the external standard (three more counts) and the time of the measurement, and produce the mean disintegrationrate for that sample.The operator of the computer will have prepared a suitable polynomial fit for the efficiency versus ratio curve, and the data can be transferred by hand or by paper tape. The use of more sophisticated computers may be necessary for larger outputs or for a wide variety of samples.28 If each sample is different in radioisotope, scintillator or type of quenching, then manual methods are usually the most efficient. The use of both the external standard and the channels ratio methods for each sample will normally reveal the type of quenching or the presence of chemiluminescence, and it is good practice to compare the efficiencies determined by these methods by using statistical tests.TABLE I METHODS OF DETERMINING COUNTING EFFICIENCY Method Internal standard Automatic external standard Channels ratio Advantages Gives individual results Best method for highly quenched samples Only reliable method when solid support Corrects for both colour and chemical material is present quenching Automatic with no handling problems Only short repeat count needed Composition of sample unchanged Only one count needed No handling of the sample required Composition of the sample unchanged Independent of sample volume Independent of inhomogeneity in the sample { CONCLUSIONS Disadvantages Possible errors when measuring small amounts of standard solution with a pipette Second count needed, i.e., time consuming with large numbers of samples Sample cannot be re-counted Not suitable for dual-label samples Dependent upon sample volume and the Sample must be homogeneous Poor accuracy with highly quenched Instrumental costs and maintenance accurate positioning of the source samples Long counting time needed for accuracy with low-activity samples Poor accuracy with highly quenched samples Needs a t least two channels Liquid scintillation counting is the accepted technique for many radioisotopes in a wide range of chemical forms.It lacks sensitivity for very low levels of a-activity and cannot compete with internal gas counters used for natural tritium levels. This leaves a wide field of analytical application in chemistry, biochemistry and medicine. The presence of quenching agents in most samples can normally be detected, and their effect either reduced or corrected for by the choice of suitable standardisation methods.Similarly, chemiluminescence can be detected and reduced by using a different technique. Improvements in the future may come from increased photocathode efficiency and from improved chemical techniques, with the best use of scintillators and solubilisers to reduce unwanted effects. Use of the technique is essentially a practical problem, which presents new facets with each new type of sample. Appendix ISOTOPES MEASURED IN A LIQUID SCINTILLATION COUNTER This appendix is intended as a preliminary guide to the versatility of the technique and gives some idea of the sensitivity of the method. The information in Table I1 includes a range of isotopes and the matrix from which they were extracted.The preparation tech- nique can be obtained from the references in the final column, but brief details of the additives are given, together with the scintillator used. The efficiency of counting is given only as aOctober, 19711 AS AN ANALYTICAL TOOL 687 guide, but it can be used to give an approximate indication of sensitivity if the background is taken as typically 10 to 25 counts per minute for most systems. The reference list is fairly limited considering the vast literature on this subject, and represents methods, techniques and theoretical information that we have found to be useful in the theory and application of liquid scintillation counting. TABLE I1 ISOTOPES MEASURED IN A LIQUID SCINTILLATION COUNTER Isotope Matrix 3H 3H 3H 3H 3H 14c 14c 3 S P 3% 35s 36~1 4sCa ssFe S°Fe 63Ni OOSr O O Y 1311 147Pm aloPb pu (4 34lPu 33spu 1.2. 3. 4. 5. 6. 7. 8. 9. Water Water Water Blood, plasma, Plasma, urine urine 14C-toluene l’C-fructose Organic Various compounds Vegetation Various Various Various Counting form Water Water emulsion Water emulsion Blood, plasma, urine Plasma, urine, emulsion 14C-toluene 14C-fructose Aqueous solutions Aqueous solutions BaSO, precipi- tate in a gel H2S04 on glass- fibre disc Aqueous solution of NaCl CaCl, in dibutyl Blood phosphate Ferrjphosphate complex in gel Aqueous solution Tetrapyridine- nickel dithio- cyanate Various 2-Eth ylhexanoic acid solution of SrCO, Plasma Plasma in gel Urine Di-2-ethylhexyl phosphate complex Aqueous solutions Aqueous concen- trate Urine, faeces, Ferriphosphate blood complex in gel Bone, liver, Acidic solution spleen, urine after in-vial oxidation N.S.Not stated. Scintillator PPO - p-bis(o-methyl- styryl) benzene in dioxan Triton N-101 in p-xylene Triton X-100 in toluene BBS3 solubiliser in toluene Triton X-100 Hyamine 1OX in toluene BBSl solubiliser in Triton X-100 in toluene toluene None +. wavelength shifter Dioxan 3 g I-1 of p-terphenyl in Dioxan toluene Toluene Dioxan Dioxan Toluene Toluene Toluene Dioxan Dioxan 0.4 g 1-1 of PPO in Toluene ethanol Counter tempera- ture/OC 0 to 25 17 to 25 4 2 0 N.S. 0 N.S. N.S. N.S. 12 N.S. -5 4 0 - 4 -2 N.S. N.S. 4 N.S. Efficiency, per cent. 23 24 27 37 30 20 88 75 25 50 65 78 85 85 19.4 33.4 65 83 95 85 95 97 86 21 85 Reference 14 15 11 16 29 16 11 7 7 30 31 32 33 23 34 35 36 37 38 24 39 REFERENCES Bell, C.G., and Hayes, F. N., Editors, “Liquid Scintillation Counting,” Pergamon Press, Oxford Branksome, E. D., Editor, “The Current Status of Liquid Scintillation Counting, ” Grune and Birks, J. B., “The Theory and Practice of Scintillation Counting,” Pergamon Press, Oxford and Gale, H. J., and Gibson, J. A. B., J . Scient. Instrum., 1966, 43, 224. Gibson, J. A. B., and Gale, H. J., J . Phys. E., Series 2, 1968, 1, 99. -- , IM. J . A$$. Radiat. Isotopes, 1967, 18, 681. Elriik, R. H., and Parker, R. P., Ibid.. 1968, 19, 263. Bray, G. A., Analyt. Biochem., 1960, 1, 279. Lerch, P., and Cosandey, M., Adv. Tracer Meth., 1966, 3, 107. and New York, 1958. Stratton, New York and London, 1970.New York, 1964.688 10. 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. GIBSON AND LALLY Polesky, H. F., and Seligson, D., Analyt. Biochem., 1965, 10, 347. Turner, J. C., Int. J . Appl. Rudiat. Isotopes, 1969, 20, 499. White, D. R., Ibid., 1968, 19, 49. Burke, C. W., Humphrey, K., and Beardwell, C. G., Ibid., 1968, 19, 666. Moghissi, A. A., Kelley, H. L., Regnier, J. E., and Carter, M. W., Ibid., 1969, 20, 145. Lieberman, R., and Moghissi, A. A., Ibid., 1970, 21, 319. “Beckman International Newsletter,” 1968, SH-68-1. Hansen, D. L., and Bush, E. T., Analyt. Biochem., 1967, 18, 320. Horrocks, D. L., Int. J . Appl. Radiat. Isotopes, 1968, 19, 859. Herberg, R. J., Analyt. Chem., 1960, 32, 42. Mahin, D. T., and Lofberg, R. T., in Branksome, E. D., Editor, op. cit., p. 212. Schoniger, W., Mikrochim. Acta, 1955, 123. Oliverio, V. T., Denham, C., and Davidson, J. D., Analyt. Biochem., 1962, 4, 188. Eakins, J. D., and Brown, D. A., Int. J . A$@. Radiat. Isotopes, 1966, 17, 391. Eakins, J. D., and Lally, A. E., Rep. U.K. Atom. Energy Autlz., AERE-R 6640, H.M. Stationery Takahashi, I. T., and Blanchard, F. A., Analyt. Biochem., 1970, 35, 411. Glass, D. S., Int. J . Aflfll. Radiat. Isotopes, 1970, 21, 631. Williams, M. A., Cope, G. H., Jackson, J. L., and Hill, P., Biochem. J., 1970, 118, 379. Figdor, S. K., Comp. Biomed. Res., 1970, 3, 201. Whyman, A. E., Int. J . Appl. Radiat. Isotopes, 1970, 21, 81. Willis, C. P., Olson, D. G., and Sill, C. W., Analyt. Chem., 1970, 42, 124. Lloyd, R. A., and Rees-Evans, D. B., Int. J . Appl. Radiat. Isotopes, 1965, 16, 393. Moghissi, A. A., in Branksome, E. D., Editor, op. cit., p. 86. Hardcastle, J. E., Hannapel, R. J., and Fuller, W. H., Int. J . A@pl. Radiat. Isotopes, 1967, 18, 193. Harvey, B. R., and Sutton, G. A., Ibid., 1970, 21, 519. Uyesugi, G. S., and Greenberg, A. E., Ibid., 1965, 16, 581. Bell, T. K., J . Clin. Path., 1967, 20, 629. Ludwick, J. D., Analyt. Chem., 1964, 36, 1104. Fairman, W. D., and Sedlet, J., Ibid., 1968, 40, 2004. Lindenbaum, A., and Lund, C. J., Radiat. Res., 1969, 37, 131. Office, London, 1970. Received April Sth, 1971 Accepted June 25th, 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600681

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

AIzaZyst, October, 1971, Vol. 96, $$. 689-698 689 A Method for the Determination of Disodium Cromoglycate and Other Chromones BY J. TILLMAN AND D. W. WHYMARK (Fisons Limited, Pharmaceutical Division, Research and Develofiment Labwatmies, Bakewell Road, Loughborough, Leics.) The hydrolysis of disodium cromoglycate in alkaline solution has been studied and a method for its determination is presented. The method is based on spectrophotometric measurement of the change in absorption at 310 nm, which takes place on opening the y-pyrone ring of the chromone nucleus. The technique is applicable, with suitable minor modifications, to the determination of other compounds containing the chromone nucleus, such as flavones. DISODIUM cromoglycate (DSCG), the disodium salt of 1,3-bis(2-carboxychromone-5-yloxy)- 2-hydroxypropane, is a new therapeutic substance1 of value in the treatment of allergic asthma.In the course of the commercial development of disodium cromoglycate (I) it became necessary to devise a method of analysis suitable for the examination of aged material. The most likely type of degradation was considered to be hydrolytic, analogous to the reaction of flavones in alkaline solution.2 Such a degradation would involve the opening of the ypyrone ring of the chromone nucleus, followed by the loss of a two-carbon fragment as oxalate, and the formation of a bisacetophenone derivative (111). The open-ring inter- mediate (11) had not previously been isolated and was thought to be unstable. The bisacetophenone derivative has phenolic properties and can be determined spectro- photometrically by coupling with diazotised amine (e.g., +-nitroaniline) .3 It was considered that the determination of this compound before and after quantitative hydrolysis should HOCH CH2.O 2 OH I1 aqueous NaOH / I give an accurate value for the amount of unchanged chromone present in a formulation.The hydrolysis of disodium cromoglycate in alkaline solution was therefore studied in detail. It was soon shown that the open-ring intermediate (11) was reasonably stable in strongly alkaline solution and that in this enol form its ultraviolet absorption spectrum was different from those of DSCG and the bisacetophenone. This difference was such that it was feasible to base a method of analysis on the formation of the intermediate..As it is possible that the open-ring compound (11) might be present in aged material it is essential that any method 0 SAC and the authors.690 TILLMAN AND WHYMARK: A METHOD FOR THE DETERMINATION [AutaZyst, Vol. 96 I 200 225 250 275 300 325 350 375 Wavelength/nrn Fig. 1. Ultraviolet spectra recorded during the hydrolysis of DSCG in 0.1 M sodium hydroxide a t 37 OC. DSCG concentration 1.8 mg per 100 ml in 10-mm cell. Time after preparation of solution: 1, 2 minutes; 2, 19 minutes; 3, 36 minutes; 4, 53 minutes; 5, 70 minutes; 6, 86 minutes; 7, 100 minutes; 8, 141 minutes: 9, 170 minutes: 10, 189 minutes; 11, 222 minutes; and 12, 263 minutes of analysis used in stability testing should permit discrimination between this compound and DSCG. A method based on the formation of this intermediate is therefore potentially very specific for the chromone nucleus. THE HYDROLYSIS OF DSCG IN ALKALINE SOLUTION AND THE NATURE OF THE HYDROLYSIS The hydrolysis of disodium cromoglycate in 0.1 M sodium hydroxide solution at 96 "C was followed by a thin-layer chromatographic determination.The bisacetophenone derivative was shown to be the ultimate product of the reaction, but the presence of at least one inter- PRODUCT- 0 . 4 1 0.2 4 0.4 I 0.2 \ I 0.0 I I I I I 225 250 275 300 325 350375 'Wavelengthhm Fig. 2. Ultraviolet spectrum of an alkaline solution of the bisacetophenone derivative of DSCG a t a concentration of 2.2 mg per 100 ml in 10-mm cellOctober, 19711 OF DISODIUM CROMOGLYCATE AND OTHER CHROMONES 691 mediate was indicated.The course of the hydrolysis of a more dilute solution at 37 "C was followed spectrophotometrically (Fig. 1). Graph 1 in this figure corresponds to DSCG itself, with an absorption maximum at a wavelength of 326nm. The course of the reaction is indicated by a hypsochromic shift to 310nm, accompanied by an increase in absorbance (graph 12). A maximum is reached after 4 hours and the system remains stable for about 30 minutes before further changes occur which eventually result in the spectrum of the bisacetophenone derivative (Fig. 2). It can be inferred from the isosbestic point at 277 nm (Fig. 1) that under the given conditions there are only two species present during the initial stages of the hydrolysis. These experiments, therefore, confirmed the formation of an inter- mediate during the hydrolysis of DSCG to its bisacetophenone derivative.At 25 "C in 0.1 M sodium hydroxide solution complete production of the intermediate takes 16 hours and further hydrolysis to the bisacetophenone is slow. A 1 per cent. solution of DSCG that had been hydrolysed under controlled conditions produced on acidification a yellow precipitate that was extracted into solvent ether. Separa- tion and evaporation of the ether layer yielded a yellow solid. The parent acid of DSCG is white and completely insoluble in solvent ether, and the infrared spectrum of the yellow solid differs from that of cromoglycic acid although it does indicate that the compound is a carboxylic acid. The yellow solid was soluble in dilute alkali and gave an ultraviolet spectrum identical with that of the intermediate formed in the hydrolysis of DSCG.It was also readily soluble in ethanol from which it deposited, either on standing or, more rapidly, on acidification and heating, a white precipitate with an infrared spectrum identical with that of cromoglycic acid. It was apparent that no carbon atoms had been lost during the hydrolysis and subsequent reactions and it was therefore concluded that the intermediate in the hydrolysis reaction was the acyclic sodium salt (11), the parent acid of which is yellow and soluble in ether and which readily cyclises in acid - ethanol solution. The ester (IV) is known.4 IV A solution of (IV) in chloroform was extracted with aqueous alkali and the ultraviolet absorption spectrum of this extract was found to be identical with that of the intermediate in the hydrolysis of DSCG.This is further evidence for the acyclic nature of the intermediate. An attempt was made to isolate the acyclic sodium salt by freeze-drying a hydrolysate. Two experiments were performed. In one the hydrolysate was freeze-dried directly and in the other the hydrolysate was carefully neutralised to pH 7.5 with dilute hydrochloric acid before drying. These procedures produced two apparently different residues, that obtained from the alkaline solution being bright yellow and that from the neutral solution being white. Their infrared spectra were different and also different from that of DSCG. Both salts re- dissolved in water, but on acidification the salt from the alkaline solution gave a yellow precipitate while that from the neutral solution gave a white precipitate.The infrared spectra of these two precipitates were different from each other and from that of cromoglycic acid (V, Fig. 3). However, both could be converted into cromoglycic acid by digestion in acidified ethanol. This evidence confirms that the two sodium salts were the keto and enol forms of the acyclic compound, from which the chromone structure could readily be regenerated. Their proton nuclear magnetic resonance spectra also support this in that the band at about 3.5 7, which is caused by the proton in the %-position in the chromone ring, is not observed. Further evidence that supports the tautomeric structures theory was obtained by examination of the spectral changes that occurred when the pH of the hydrolysate was adjusted from 13 to 6.5 (Fig.4). The strong band at 310 nm disappears and is replaced by two weaker bands at 268 nm and 333 nm. On making the solution alkaline again the original spectrum can be regenerated. This is consistent with a tautomeric system involving labile692 I aqueous NaOH TILLMAN AND WHYMARK: A METHOD FOR THE DETERMINATION [Analyst, Vol. 96 H+ ? I OH It H+ pH<4 + H+ pH<4 'f MCH2- O-CH2- I - H0OC.C H -HOOC-C HO II $5 0 ~ - c H '-H+& I H50H HOOC. OH V Fig. 3. Summary of the relationships between DSCG and its hydrolysis products protons, and is explained by shifts in the equilibrium between I1 and VI (Fig. 3). The fore- going evidence is summarised in the reaction plan (Fig.3). The alkaline hydrolysis of DSCG proceeds via ring opening to form a /3-diketo compound, which is accompanied by a change in the ultraviolet absorption spectrum. The proposed method makes use of this reaction and the associated increase in absorbance at 310 nm. The @-diketo compound exists in the diketo form in neutral and acidic solutions, and in the enol form in alkaline solution. Further hydrolysis results in the formation of a bis- acetophenone derivative by loss of a two-carbon atom fragment as oxalate. 225 250 275 300 325 350375 Wavelength/nm Fig. 4. Ultraviolet spectra of the initial hydrolysis product of DSCG at: 1, pH 13; and 2, pH 6.5. Concentration of 1.2 mg per 1OOml in 10-mm cellOctober, 19711 REAGENTS- OF DISODIUM CROMOGLYCATE AND OTHER CHROMONES EXPERIMENTAL Sodium hydroxide solutions, 0.10 and 0.50 M.693 PROCEDURE- Weigh accurately an amount of sample containing about 250 mg of DSCG and transfer it to a 500-ml calibrated flask. Add about 200 ml of distilled water, shake to dissolve the sample and dilute to the mark with distilled water. Mix thoroughly. Transfer a 2-ml aliquot with a bulb pipette to a 100-ml calibrated flask containing 10 ml of 0-1 M sodium hydroxide solution and 70ml of distilled water. Dilute to the mark with distilled water and mix thoroughly. Measure the absorbance of solution A at 310 nm in 1-cm silica cells, with distilled water in the reference cell. Also prepare a reagent blank consisting of 10 ml of 0.1 M sodium hydroxide diluted to 100 ml with distilled water. Read the absorbance of this solution against the distilled water reference at 310 nm.Subtract the reading from the sample reading, (A3&. All sample measurements must be completed within 10 minutes of adding the sample aliquot to the dilute sodium hydroxide solution. To a second 100-ml calibrated flask transfer by pipette 20.0 ml of 0-5 M sodium hydroxide solution and approximately 60ml of distilled water. Swirl to mix. To this flask transfer 2 ml of the sample solution by using a bulb pipette, again swirl to mix and immediately dilute to the mark with distilled water. Mix thoroughly. (This is solution B.) Store solution B at 25 & 045°C in a thermostatically controlled water-bath for at least 16 hours and not longer than 19 hours. Measure the absorbance of the solution at 310nm in a 1-cm silica cell, with distilled water in the reference cell.Prepare a reagent blank by diluting 20 ml of 0.5 M sodium hydroxide solution to 100 ml with distilled water. Read the absorbance of this solution against the distilled water reference at 310 nm. Subtract this reading from the sample reading, (Aslo),+ It is important to adhere to the order of addition of reagents and sample in the preparation of solution B, and to the storage temperature and time of 16 to 19 hours before measurement of absorbance. (This is solution A.) CALIBRATION GRAPH- Weigh accurately about 1 6 g (equivalent of dry powder) of pure DSCG into a 500-ml calibrated flask. Dissolve, make up to volume with distilled water and mix (solution C). Transfer aliquots of 5, 10, 15, 20 and 25 ml by pipette into 100-ml calibrated flasks, dilute them to the mark with distilled water and mix thoroughly (solutions D).Carry out the procedure for each of the solutions D and plot a graph of AA against the concentration of anhydrous DSCG (in mg per 100 ml) in the final 100 ml of solution. The latter is calculated from the expression- (100 - M ) 100 DSCG (mg per 100 ml) = 0.04 VW, where W3 is the weight (in g) of pure DSCG taken for the stock solution C, M the moisture content of the pure DSCG determined by drying under vacuum over phosphorus pentoxide at 105 "C, and V the volume (in ml) of the aliquot taken to prepare solution D. The points obtained should form a straight line passing through the origin. CALCULATION- Read from the calibration graph the weight of DSCG in mg per 100 ml of solution corresponding to the experimental value of AA, say W, mg.The increase in absorbance caused by hydrolysis = AA = (A31o)B : (A310)A. Then the percentage of DSCG in the sample =w2 x 25 000 Wl where Wl is the weight of sample in mg.694 RESULTS AND DISCUSSION RELATIONSHIP BETWEEN DSCG CONCENTRATION AND AA- It was confirmed that the increase in absorbance at 310nm, AA, was proportional to the DSCG concentration in the range from 0.1 to 1.5 mg per 100 ml when the procedure was applied to samples of the pure compound. The calibration graph was found to be reproducible and to pass through the origin. EFFECT OF SODIUM HYDROXIDE CONCENTRATION ON THE RECOVERY OF DSCG- The procedure was applied to identical aliquots of the stock solution.The amount of sodium hydroxide used to effect the hydrolysis was varied to give different final concen- trations of alkali but all other conditions were as stated under Procedure. The results are shown in Table I. It is evident from these results that a 5 per cent. variation in the concen- tration of alkali has no effect on the result but that insufficient alkali leads to low recoveries caused by incomplete reaction. TILLMAN AND WHYMARK: A METHOD FOR THE DETERMINATION [Analyst, Vol. 96 TABLE I EFFECT OF VARIATION OF SODIUM HYDROXIDE CONCENTRATION ON THE RECOVERY OF DSCG Sodium hydroxide concentration/M 0.105 0.105 0.100 0.100 0.095 0.095 0.090 0.080 0.070 0.060 0.050 AA 0-395 0.395 0.396 0.394 0.395 0-398 0.385 0.368 0.347 0.325 0.302 Recovery of DSCG, per cent.100.0 100.0 100.3 99.7 100.0 100.8 97.5 93.2 87.8 82-3 76.5 EFFECT OF TEMPERATURE OF HYDROLYSIS ON THE RECOVERY OF DSCG- The procedure was applied to identical aliquots of the stock solution but the temperature at which hydrolysis was carried out was varied. The results are shown in Table 11. Variation of the temperature has a marked effect as low results are obtained at temperatures both higher and lower than 25 "C. A temperature of 25 & 0.5 "C was therefore adopted for the routine operation of the method. At lower temperatures the hydrolysis is slow, and incomplete reaction leads to low results. At higher temperatures the decomposition of the open-ring compound, 11, to the bisacetophenone derivative, 111, occurs at a significant rate and also leads to low results.TABLE I1 EFFECT OF TEMPERATURE ON THE RECOVERY OF DSCG Temperaturel'C . . .. 21 21 25 25 30 30 Recovery of DSCG, per cent. . . 94.8 93.4 100.2 99.8 96.6 96.4 EFFECT OF TIME OF HYDROLYSIS ON THE RECOVERY OF DSCG- An alkaline solution of DSCG was prepared according to the procedure andits absorbance at a wavelength of 310 nm measured as a function of time. The results are shown in Table 111. A maximum is reached and maintained for about 3 hours. On the basis of these results a reaction time of 17.5 & 1 6 hours was adopted for the routine operation of the method. TABLE I11 EFFECT OF TIME OF HYDROLYSIS ON THE RECOVERY OF DSCG Time/hours . . .. .. 2 4 7 11 16.6 17.5 18.6 19.6 21.8 Recovery of DSCG, per cent.. . 38.0 62.0 82.3 94.2 99.6 100.0 99.5 98.7 94.9October, 19711 OF DISODIUM CROMOGLYCATE AND OTHER CHROMONES 695 INTERFERENCES- Lactose-DSCG is normally formulated for clinical use as a blend with lactose; therefore the effect of lactose on the recovery of DSCG by the proposed procedure was investigated. Two solutions containing identical concentrations of DSCG (approximately 50 mg per 100 ml) were prepared.One solution also contained lactose at a concentration of 50 mg per 100 ml. Aliquots (2ml) of each of the solutions were treated as under Procedure. The results are shown in Table IV and confirm that lactose has no effect on the determination of DSCG by the proposed method. TABLE IV EFFECT OF 1 mg PER 100 ml OF LACTOSE ON THE RECOVERY OF DSCG Lactose absent . . . . 100.3 100.5 99.5 100.0 100.0 Mean 100.1 Lactose present .. . . 100.5 99.7 100.5 100.5 100.0 Mean 100.2 Bisacetophenone derivative-The bisacetophenone derivative, 111, is a possible degradation product of DSCG, and it was considered necessary to show that its presence would not interfere in the proposed method. It was shown that no change in the ultraviolet spectrum of a solution of the bisacetophenone derivative in 0.1 M sodium hydroxide occurred during a period of 16 hours at a temperature of 25 "C. In particular, the absorbance at a wavelength of 310 nm of an alkaline solution containing 2 mg per 100 ml of the bisaceto- phenone derivative was unchanged over this period. The presence of this compound in the sample will not, therefore, interfere in the determination of DSCG by the procedure outlined in this paper.Open-ring compoztnd, II-If the degradation of DSCG in a storage sample occurs by a hydrolytic process, the open-ring compound, 11, may also be present. This compound does not contribute to the increase in absorbance at 310nm and therefore it will not interfere in the determination of DSCG. However, it can be seen from Fig. 3 that its absorption spectrum is pH-dependent. It is essential that the initial absorbance be measured in an alkaline solution with a pH greater than 11 so that I1 is in its enol form, which has an 0.0 I I I I I I I 1 225 250 275 300 325 350 375 Wavelengthhm Fig. 5. Ultraviolet spectra recorded during the hydroly- sis of compound VII in 0.3 M sodium hydroxide a t 25 "C. Concentration 2.0 mg per 100 ml in 10-mm cell.Time after preparation of solution: 1, 6 minutes; 2, 18 minutes; 3,29 minutes; 4, 43 minutes; 5, 60 minutes; and 6, 77 minutes696 TILLMAN AND WHYMARK: A METHOD FOR THE DETERMINATION [Analyst, VOl. 96 absorption maximum at 310 nm. If the initial measurement is made while using 0.1 M sodium hydroxide solution, then the rate of hydrolysis is such that the absorbance increases too rapidly for an accurate initial reading to be made. At a sodium hydroxide concentration of 0.01 M, I1 is in its desired form and the rate of hydrolysis is such that there is no measurable change in absorbance for up to 10 minutes after preparing the solution. ACCURACY AND PRECISION- The DSCG content of a fresh 50 per cent. blend with lactose was determined by ultra- violet spectrophotometry at a wavelength of 326nm, at which the absorption maximum occurs.The mean of twenty determinations was 51.4 per cent. w/w and this value was taken to be the true DSCG content of the blend. The standard deviation was 1-00 per cent., which includes assay variation and slight blend heterogeneity. The DSCG content of the blend was then determined twenty times by using the procedure given above. The mean of the twenty results was 52-0 per cent. w/w and the standard deviation was 1.23 per cent. The mean recovery obtained with the proposed method, with respect to the direct spectrophotometric method, was 101.2 per cent. The accuracy and precision of the proposed procedure are therefore comparable with those of the direct method. APPLICATION OF THE HYDROLYSIS APPROACH TO OTHER CHROMONES- The reaction upon which the procedure is based is general for compounds with the chromone structure.With suitable modifications it was possible to devise methods of analysis for other compounds with a basic chromone structure. Three further compounds were examined : lJ3-bis(2-carboxychromone-7-yloxy)-2-hydroxypropane disodium salt (VII) , which is the 7,7‘ isomer of DSCG ; cyclohexano [g]chromone-2-carboxylic acid, sodium salt (VIII) ; and morin (3,5,7,2’,4’-pentahydroxyflavone) (IX) . HOCH {cH*oQ&cooNj 2 dCOONa VI I Vlll IX OH It was found that by using suitable conditions of alkalinity and temperature, slow changes in the ultraviolet spectra of solutions of these compounds could be effected. In each case a steady state was reached. The changes in the spectra are shown in Figs.5, 6 and 7. The TABLE V SUMMARY OF CONDITIONS SUITABLE FOR THE DETERMINATION OF VARIOUS COMPOUNDS CONTAINING THE CHROMONE NUCLEUS Sodium hydroxide Tem- Optimum Wave- Compound concentration/M perature/”C hydrolysis time length/nm I 0.1 25 17 f 1.5 hours 310 VII 0-3 25 110 f 10 minutes 354 VIII 0.2 10 4 f 0-5 hours 354 IX 0.1 37 3.5 f 0.5 hours 415 318 Change in absorbance for 1 mg per 100 ml + 0.42 + 0.67 +0*31 +0.61 - 0.68October, 19711 OF DISODIUM CROMOGLYCATE AND OTHER CHROMONES 1.4 1.2 I I I I - I - O.*1 0.6 0.4 0.2 ' ' 7 I /- I I - 0.oc I I I I I - 225 250 275 300 325 350 ~~ 400 450- 50 697 WaveIength/nm Fig. 6. Ultraviolet spectra recorded during the hydrolysis of compound VIII in 0.2 M sodium hydroxide a t 10 "C.Concentration 1.9mg per 100ml in 10-mm cell. Time after preparation of solution: 1, 1 minute; 2, 6 minutes; 3, 43 minutes; 4, 69 minutes; 5, 120 minutes; 6, 177 minutes; 7, 240 minutes. increase (or decrease) in absorbance at a suitable wavelength was shown to be directly pro- portional to the concentration of the individual compound. The conditions appropriate for the determination of the four compounds are summarised in Table V. The existence of isosbestic points in the spectra in Figs. 1, 5, 6 and 7 supports the contention that after a steady state has been achieved there exists in solution an equilibrium between the chromone t I I 11 I Wavelengthlnm Fig. 7. Absorption spectra recorded during the hydrolysis of morin, compound IX, in 0.1 M sodium hydroxide a t 37 "C.Concentration 1.1 mg per 100 ml in 10-mm cell. Time after preparation of solution: 1, 2 minutes; 2, 10 minutes; 3, 20 minutes: 4, 35 minutes; 6, 65 minutes; 6, 85 minutes; 7, 105 minutes; 8, 120 minutes; 9, 135 minutes; 10, 150 minutes; and 11, 180 minutes698 TILLMAN AND WHYMARK structure and the open-ring compound, probably in its enol form. It may be that the equilibrium is such that all or nearly all of the original chromone has been converted into the open-ring species and that the spectrum of a pure compound is observed. Experiments carried out during these investigations showed that the two most important parameters were the alkalinity and the temperature at which hydrolysis was carried out. In the reaction sequence- Chromone + open-ring compound -+ acetophenone derivative the first reaction is favoured by high alkali concentrations and the second by higher tem- peratures.In order to obtain maximum effect, the extent of the second reaction should be minimised, while allowing the first reaction to proceed more rapidly. A compromise must be made between alkali concentration and temperature so that a steady state is main- tained for a sufficient period to enable accurate final measurements to be made. CONCLUSIONS This approach to the determination of DSCG provides the high degree of specificity required in testing the stability of formulations. It is sufficiently accurate and the precision is of the order that would be expected from a spectrophotometric method. Its successful application to three other chromones, including a flavone, suggests that it may have wide applicability for the assay and determination of purity of such compounds. We thank Mr. H. Cairns for helpful discussions and Mr. M. L. Ray-Johnson for technical assistance. We also thank Dr. J. S. G. Cox and Fisons Limited, Pharmaceutical Division, for permission to publish this paper. REFERENCES 1. Pepys, J., and Frankland, A. W., Editors, “Disodium Cromoglycate in Allergic Airways Disease,” a Symposium held a t the Royal Society of Medicine, London, March 5th, 1969, Butterworths, London, 1970. Finar, I. L., “Organic Chemistry,” Fourth Edition, Longmans, London, 1968, Volume 2, p. 691. Moss, G. F., Jones, K. M., Ritchie, J. T., and Cox, J. S. G., “Toxicology and Applied Pharma- cology,” Academic Press, London, 197 1. Cox, J. S. G., Beach, J. E., Blair, A. M. J. N., Clarke, A. J., King, J., Lee, T. B., Loveday, D. E. E., Moss, G. F., Orr, T. S. C., Ritchie, J. T., and Sheard, P., in Harper, N. J., and Simmonds, A. B., Editors, “Advances in Drug Research,” Academic Press, London, 1970, Volume 6, p. 116. Received Mumk 16th, 1971 Accepted May 24th, 1971 2. 3. 4.

ISSN:0003-2654    

DOI:10.1039/AN9719600689

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Autalyst, October, 1971, Vol. 96, $9. 699-711 699 The Flame-photometric Determination of Alkalis in Ceramic Materials BY R. P. EARDLEY AND R. A. REED (British Ceramic Research Association, Queens Road, Penkhull, Stoke-on-Trent, ST4 7LQ) The flame-photometric method in general use for the determination of alkalis in the ceramic industries was originally devised for the EEL, Model 100, flame photometer, with a coal gas flame, which is not now generally available. Current supplies of town gas, propane and methane (natural gas) flames are compared and interferences evaluated. Perchloric and hydrochloric acids are found to have a depressant effect; sodium is enhanced by potassium and is subject to spectral interference by calcium. Propane is chosen as the preferred fuel. The effect of chlorine-containing acids is eliminated by the use of sulphuric and nitric acids for the initial decomposition and the spectral interference from calcium by the addition of aluminium sulphate ; sodium - potassium inter-element effects are eliminated by the use of a caesium buffer.Although the procedure is principally devised to give optimum results with aluminosilicates, its extension to high-lime materials is also considered. THE method for the determination of alkalis in ceramic materials is based on work carried out 14 years ago1 with the Evans Electroselenium, Model 100, flame photometer. This work showed that it was legitimate to compare simple solutions of the alkali sulphates with sample solutions (prepared by decomposition with hydrofluoric acid) in which the acid used for dissolution was hydrochloric acid.In the intervening years small changes in the apparatus, e.g., in the construction of nebulisers, and considerable changes in the composition of town gas have occurred, and during the past year or so occasional discordant results obtained have suggested that these changes might have altered interferences both qualitatively and quantitatively . At the time of the original work, town gas consisted mainly of coal gas. At the present time town gas may take many forms ranging from natural gas containing 84 to 90 per cent. of methane to a mixture of this with low pressure gas and high pressure gas. This composition may vary from day to day and even from hour to hour. Flame temperature is therefore likely to vary (Table I) and this variation, together with the impending introduction of North Sea gas, created interest in the effect of the fuel gas on the determination of alkalis, particularly from the point of view of the possible inter- ference effects. The ultimate objective was to ensure that the method for their determination TABLE I COMPARISON OF THE COMPOSITION OF COAL GAS, LOW AND HIGH PRESSURE GAS AND TWO SAMPLES OF TOWN GAS Composition, per cent.v/v Constituent Hydrogen . . .. .. .. Methane .. .. .. .. Ethane .. .. .. .. Carbon monoxide . . .. .. Carbon dioxide . . .. .. Nitrogen .. .. .. .. Unsaturated hydrocarbons . . Calculated flame temperaturel'c Coal gasS 49.8 26-8 1-5 11.8 2.4 4.2 4.6 1916 Low pres- sure gas 40.6 29.8 8.4 7.6 11.2 1.4 1810 - High pres- sure gas Town gass Town gas* 47-1 60 44.9 34'3 } 23.6 } 19.3 1-8 2.7 9.1 16.8 12.3 2.7 - 8.0 1892 - 3.0 5.4 1970 1870 1880 - The calculated flame temperature for propane is 1925 "C and for methane 1880 "C.0 SAC and the authors.700 EARDLEY AND REED : THE FLAME-PHOTOMETRIC [ArtaZyjst, Vol. 96 yielded correct results by evaluating the respective interference effects for various gases, selecting the best of the gases for further experiments and attempting to overcome the interferences with this gas. Three gases were chosen for study, town gas, propane, which is available commercially in a reasonable degree of purity and is already used in several other commercial instruments, and methane (the major constituent of North Sea and Saharan gases). Andrew and Nichols2 drew attention to the variability of town gas and of commercial butane, and for this reason butane was not included in our experiments.It should be made clear that during the period of experimental work the town gas used was from the normal supply. The results presented for use of town gas may therefore have been subject to any variation that could arise from fluctuations in gas composition such as those illustrated in Table I. EXPERIMENTAL MODIFICATIONS TO THE INSTRUMENT- Preliminary experiments indicated that neither propane nor methane could be burnt satisfactorily on the normal EEL burner head. As soon as a satisfactory flame had been produced, as judged by the appearance of separate blue cones, the flame "lifted off." This was caused by the low burning velocity of the two gases.A satisfactorily stable flame could be obtained by the use of a Mdker-type head but a serious loss of sensitivity occurred, probably resulting from the obstruction of the aerosol flow. At the suggestion of Mr. V. R. Williamson of Evans Electroselenium Ltd., the thickness of the standard burner head was increased to about half an inch and a stable flame of good sensitivity was obtained. The dimensions of the burner head are shown in Fig. 1. The head used in these studies was 0.660 H J- *Ten holes - 118 Dia. drilled '1 I through 0.906 Dia. p6 B.S. 19.16 +O-0013 +0-0008 - 14-40.984 Dia. Assembly a' (Note relative posiiion to holes) of slot H 0.906 Dia. +0-0008 H7 B.S. 1916 Fig. 1. Design of burner for EEL, Model 100, flame photometer to burn propane and natural gas.Material is of lS/S Cr - Ni steel (all measurements are in inches)October, 19711 DETERMINATION OF ALKALIS I N CERAMIC MATERIALS 701 constructed of brass and performed satisfactorily during the 4 months in which it was in use. However, for prolonged use its construction in stainless steel would seem preferable. During operation of the instrument, difficulty was often found in setting the full-scale reading, which was caused by the coarseness of the sensitivity control. This problem was overcome by replacing the carbon potentiometer with a wire-wound multi-turn helical potentio- meter (20 k a ) and the sensitivity knob with a geared multi-turn control knob. Even with the new arrangement the second-to-second stability of the readings was still not satisfactory.By trial and error it was found that optimum damping was obtained by attaching a 250-pF electrolytic capacitor across the terminals of the photocell. With the bottled gases the pressure of the gas supplied to the instrument was controlled at 0.432 p.s.i. (12 inches w.g.) by the use of a Calor propane regulator. For propane this can be attached directly to the cylinder; for methane, because of the higher cylinder pressure, an adaptor to a British Oxygen Co. hydrogen regulator was made to take the Calor regulator. STANDARDISATION- The procedure for the determination of alkalis commonly used in the ceramic industries is described in several standard method^,^ and is based on decomposition of the sample with hydrofluoric acid. Sulphuric acid was originally chosen to eliminate the fluorides, so that it was logical to standardise by using solutions of the alkali sulphates. In later years it was found that nitric and perchloric acids could be used to remove fluorides and that hydro- chloric acid could be used for the dissolution of the residue without altering the validity of sulphate standards. To test the current validity of this basis of standardisation, solutions containing 10 and 5 p.p.m.of K20, 5 and 2.5 pap.m. of Na20 and 20 and 10 p.p.m. of Li,O were prepared from their sulphates and chlorides. The instrument was adjusted in each instance to give full-scale deflection on the highest sulphate standard and zero on distilled water, and the other solutions were sprayed. The results observed for town gas, propane and methane are shown in Table 11.TABLE I1 COMPARISON OF SULPHATE AND CHLORIDE STANDARDS Solution concentration, p.p.m. Sodium- Na,O (sulphate) 6 . . .. .. (chloride) 5 . . .. .. (sulphate) 2.5 . . .. .. (chloride) 2.5 . . .. .. K,O (sulphate) 10 . . .. .. (chloride) 10 . . .. .. (sulphate) 5 . . .. .. (chloride) 6 . . .. .. Li,O (sulphate) 20 . . .. .. (chloride) 20 . . .. .. (sulphate) 10 . . .. .. (chloride) 10 . . .. .. Potassium- Lithium- Mean reading Town gas Propane Methane r A I 100 100 100 100 100 99.5 54 54 54.5 53.5 54 54.5 100 100 100 100 100 100 62 54 64 62 54 54 100 - - 98 51 49.5 - - From the results in Table I1 it seems that the temperatures of the flames are adequate to overcome the differences between the boiling-points of the alkali sulphates and chlorides.With lithium the slight difference may arise from uncertainty in the assay of the lithium carbonate used for preparation of the chloride standard. It was decided at this stage to restrict attention to effects on sodium and potassium only, as they are the most important industrially, and in any event the usual problem with the determination of lithium in ceramic materials lies in determining the significance of a deflection of one or two divisions. EFFECT OF ACIDS ON THE DISSOLUTION OF THE SAMPLE- Although it is usual to use hydrochloric acid for the final dissolution of the residue, depending on the nature of the material the fluorides can be removed with sulphuric or perchloric acid. It was, therefore, necessary to investigate the effect of each of the three702 EARDLEY AND REED THE FLAME-PHOTOMETRIC [Analyst, Vol.96 acids. Hydrochloric acid will be present in fixed amounts, i.e., 20ml of the dilute acid (1 + 19) used to dissolve the residue. On the other hand, sulphate and perchlorate concen- trations will depend on the composition of the residue, most of each anion being present as a salt, and possibly a little free acid remaining trapped in the residue. From calculations based on the possible composition of the residue, 3 ml of 1 + 4 perchloric acid (sp.gr. 1-54) and 5 ml of sulphuric acid (1 + 9) would yield equivalent amounts of perchlorate and sulphate. To test the effect of these anions at these concentrations, standard solutions prepared from alkali sulphates and chlorides to yield combined solutions equivalent to (i) 10 p.p.m.of K,O and 5 p.p.m. of Na,O and (ii) 5 p.p.m. of K,O and 2.5 p.p.rn. of Na,O with appropriate additions of acid were compared with similar solutions without addition of acid, setting full scale on the latter and the zero on distilled water. The comparisons were carried out by using town gas, propane and methane (Table 111). TABLE I11 EFFECTS OF ADDITIONS OF ACID ON SULPHATE AND CHLORIDE STANDARDS WITH THE DIFFERENT FUELS Mean readings less blanks r A 1 Solution Sulphate standards Chloride standards concentration, Acid r A 'I I A 5 p.p.m. added Town gas Propane Methane Town gas Propane Methane Potassiwm- 10 Nil 100 100 100 100 100 100 10 HCl 96-7 96.3 93.5 95.0 95.5 95.5 99.0 99.5 99.5 98.0 100.0 99.5 97.0 95.8 95.8 97.0 97.0 10 HW4 10 HClO, 97.3 6 Nil 51.5 55.0 55.0 51-0 54.5 65.3 6 HCl 48-7 53.8 53.3 48.0 53.0 52.8 50.0 55.5 55.3 50-0 55.0 55.5 49.0 54.0 54.0 49.0 53.5 64.0 5 %SO, 5 HC10, 5 Nil 100 100 100 100 100 100 5 HC1 97.0 98.5 96-0 96.0 98.3 97.8 5 H304 97.5 99.5 99.3 97.0 99.5 99.8 5 HC104 98.0 98.5 97.8 97.2 97.8 98.3 2.5 Nil 52-0 54.0 54.5 51-3 54.5 55-0 2.5 HCl 50.3 53-0 53.0 49-3 53.5 54.0 2.5 HW4 51.3 54.0 54.0 51.5 55.5 56.0 2.5 HClO, 51.0 54.0 53.5 50.8 54.5 54.8 Sodium- The fact that there was little difference between the behaviour of sodium and potassium from either sulphate or chloride standards indicated that there was little merit in altering the basis of calibration.The investigation of chloride standards was therefore discontinued at this juncture.It can be seen that both perchloric and hydrochloric acids produce significant negative errors with all gases. The depression with sulphuric acid is smaller and, with the exception of town gas, of negligible proportion. The effect of perchlorate ion is particularly disturbing in that the concentration in the solution can be so variable. It was decided to check if the hydrochloric acid addition could be increased to reach a saturation point beyond which the perchloric acid might have a negligible effect. The results of these experiments are shown in Table IV. Plateaux were not attained in the range of 10 to 50 ml of dilute hydrochloric acid (1 + 19) per 250 ml of solution. An increase to the equivalent of 20 ml of hydrochloric acid (1 + 1) led to a depression more or less proportional to the acid content.Experiments were also carried out to ascertain if the depression noted for perchloric acid was at saturation level, by increasing the addition of dilute acid (1 + 4) from 3 to 10 ml, and also if the depressions for the two acids were additive. The results are also shown in Table IV. For town gas and methane, the perchloric acid depressions were not saturation values and the effects of the two acids were in fact additive. It was decided to return to the use of sulphuric and hydrofluoric acids as the decomposition medium, but to use nitric acid for the dissolution so as to avoid depressing the solubility of calcium sulphate (such as would result from the decomposition of a bone ash or a building clay) by the presence of a large excess of sulphuric acid.October, 19711 DETERMINATION OF ALKALIS IN CERAMIC MATERIALS 703 TABLE IV FOR 5 P.P.M.OF Na20 AND 10 P.P.M. OF K20 EFFECTS OF VARIOUS ADDITIONS OF ACID ON THE INTENSITIES OBSERVED Mean readings for 5 p.p.m. of Na20 Mean readings for 10 p.p.m. of K20 P r Control .. . . .. 100 100 100 100 100 100 Hci (1 + i 9 ) ' i o . . .. . . 98.5 99.5 99.5 97.5 97.5 97-0 (1 + 19) 30 .. .. . . 96.0 98.0 97.0 93.3 93.3 91-5 (1 + 19) 50 .. .. . . 94.0 96-5 95.8 89-3 90.0 88.3 HCl (1 + 1) 20 . . .. . . 83.5 86.0 84.8 68.5 70.0 67.5 HC104 (1 + 4) 3 . . .. . . 98.0 98.5 97.8 97.3 97.0 95.8 (1 + 4 ) 10 .. 95.5 98.5 95.8 91-8 93.8 90.3 H C ~ O ~ (1 + 4) lopius Hci ii+iq'io 94-0 98.3 93.8 89.5 91.5 87.8 Addition/ml Town gas Propane Methane Town gas Propane Methane (1 + 19) 20 .... . . 97.0 99.0 97.8 95.3 95.3 93-5 (1 + 19) 40 .. .. . . 96.0 97.5 97.8 91-8 91.3 89.8 HNO, (1 + 19) 20 . . .. .. 100 100 100 99.5 100 99.5 Tests also showed (Table IV) that the effect of nitric acid is negligible, thus permitting the addition of nitric acid to the decbmposition mixture. This addition assists the complete destruction of organic matter and the removal of excess of sulphuric acid as nitrosulphonic acid. In samples with relatively high calcium contents it is desirable to reduce the sulphate-ion concentration. Tests on decomposition with sulphuric acid - nitric acid and hydrofluoric acid revealed no interferences in the complexometric determination of calcium or magnesium. Table IV also shows that the effects of addition of acid were most serious with methane as fuel, only a little less serious with town gas and relatively less important with propane.EFFECTS OF ONE ALKALI ON THE INTENSITY OF ANOTHER- In the original work on the determination of alkalis, inter-element effects between sodium and potassium were found to be negligible, but because of the changes in interference caused by anions the validity of this finding was checked with the three fuels. Solutions containing 4 p.p.m. of Na20 were sprayed alone and with additions of 8 and 100 p.p.m. of K,O and compared with the usual 5 p.p.m. of Na20 - 10 p.p.m. of K20 standard. At the same time blanks were carried out on 8 and 100 p.p.m. of K20. These solutions would represent, respectively, samples of high sodium-to-potassium ratio such as borax frit, soda feldspar and alumina, the ratio normally used as a combined standard solution and low sodium-to-potassium ratios, such as in potash feldspar.Similarly, solutions containing 8 p.p.m. of K20 were prepared with additions of 0, 4 and 80 p.p.m. of Na,O and compared with the normal 10 p.p.m. of K,O - 5 p.p.m. of Na20 standard, and blanks were measured. The results are shown in Table V. TABLE V INTER-ELEMENT EFFECTS BETWEEN SODIUM AND POTASSIUM FOR VARIOUS RATIOS Mean readings less blanks Solution r A > concentration, p.p.m. Town gas Propane Methane Sodium jilter- Na20 4 K,O 0 .. 79.2 79.3 79-0 4 8 .. 81.8 82.5 82.6 4 100 .. 81-5 82.5 82.7 K20 8 Na20 0 .. 81.7 81.3 81-5 8 4 .. 81.7 81.5 81.5 8 80 .. 82.5 82.3 82-0 Potassium jilter- The effect of potassium was most marked, and appeared to reach a maximum with 8 p.p.m., or with a ratio of K20 to Na20 of 2: 1. Further experiments with smaller increments of potassium were carried out at the 4 p.p.m.and 2 p.p.m. levels of Na20 (Table VI), by using propane alone. At this stage of the investigation it was decided to use propane as fuel gas in the remaining work. This gas had the advantage of relatively constant composition (at least it is known when variation can occur, e.g., when a new cylinder is used) and it also yielded lower levels of interference from additions of acid.704 EARDLEY AND REED : THE FLAME-PHOTOMETRIC [Artalyst, Vol. 96 TABLE VI EFFECT OF INCREASING AMOUNTS OF POTASSIUM ON THE LIGHT EMITTED BY Na20 K,O added, p.p.m.0 0.5 1 2 4 8 100 I A Town gas Propane 79.2 79.3 79.8 78.8 80.5 79.3 81.3 79.8 80.7 80.5 81.8 82.5 81-5 82.5 (Full scale set on 5 p.p.m. of Na,O in the presence of 10 p.p.m. of K,O) Test solution 4 p.p.m. of Na,O Mean readings less blank - I Methane 79.0 78.8 79.3 79.8 81.2 82.0 82.7 (Full-scale set on 2.5 p.p.m. of Na,O in the presence of 5 p.p.m. of K,O) Test solution 2 p.p.m. of Na,O Mean readings less blank Propane 77.8 78.4 78.6 80.0 80.8 80-5 - Although the results for the lower sodium concentration were less reproducible, because of increased “scale expansion,” both experiments indicated that a steady figure was reached when the K20 content reached 8 p.p.m. EFFECT OF OTHER ELEMENTS- In normal aluminosilicates, the maximum concentrations of other elements that would be expected to be present are: calcium oxide 50 p.p.m.(5 per cent.), aluminium oxide 400 p.p.m. (40 per cent.), magnesium oxide 50 p.p.m. (5 per cent.) and iron(II1) oxide 100 p.p.m. (10 per cent.). The same levels apply to aluminous materials (containing more than 45 per cent. of A1,0,) as these are not completely dissolved, so that only the soluble alumina needs to be considered. To determine the possible effects of these constituents, solutions of 4 p.p.m. of Na20 and 8 p.p.m. of K,O were sprayed in the presence and absence of calcium oxide 50 p.p.m., aluminium oxide 400 p.p.m., magnesium oxide 50 p.p.m. and iron(II1) oxide 100 p.p.m. The magnesium and calcium solutions were prepared from Specpure materials and the iron and aluminium solutions from analytical-reagent grade metals.Blanks were also sprayed to evaluate either impurity levels or the possibility of spectral interference. The results obtained are shown in Table VII. TABLE VII EFFECTS OF CALCIUM, MAGNESIUM, ALUMINIUM AND IRON ON THE DETERMINATION OF SODIUM AND POTASSIUM Full-scale deflection set on 5 p.p.m. of Na20 or 10 p.p.m. of K 2 0 Solution concentration, p .p. m. Na,O 4 + K,O 8 . . .. .. Na,O 4 + K,O 8 + MgO 50 Na,O 4 + K,O 8 + A1,0,400 . . .. Na,O 4 + K,O 8 + Fe,O, 100 Blank on CaO (50). . .. .. .. Blank on MgO (50) .. .. .. Blank on A1,0, (400) . . .. .. Blank on Fe,O, (100) . . .. .. Na,O 4 + K,O 8 + CaO 50 . . .. . . .. . . .. 7 Mean 85.5 83.0 82-8 81.5 4 2 0.8 0 - Sodium Potassium --h---7 & Mean less blank Mean Mean less blank 82.0 - 82-0 (81.5) 82.3 82.3 81.0 82.0 82.0 82.0 81.6 81.6 81.5 82.0 82.0 0 0 0 0 - - - - - - - - All of the interferences are within experimental error; with calcium, however, the blank reading is not wholly due to the presence of Na,O, as spectral interference also occurs.In this case, therefore, it is not legitimate merely to deduct the blank. ELIMINATION OF INTERFERENCES- Three forms of interference had been found; that arising from the use of perchloric and hydrochloric acids in the dissolution of the sample had already been eliminated by modification of the decomposition procedure. The spectral interference of calcium oxide band spectra on sodium can be suppressed by the addition of aluminium, preferably as the sulphate, so that the calcium is complexed by both aluminium and sulphate ions.It was necessary thatOctober, 197 13 DETERMINATION OF ALKALIS IN CERAMIC MATERIALS 705 the addition should increase neither the blank errors of the determination nor the salt content of the solution to an undesirable extent and must therefore be kept to a minimum. (The purity of AnalaR aluminium sulphate is not adequate for this purpose.) The effect of potas- sium on sodium probably results from the ionisation equilibria in the flame. Once the temperature of the flame is adequate to cause the dissociation of the molecules of alkali salts into atoms (from which the characteristic radiation is emitted), an equilibrium exists between atoms, ions and electrons, i.e., Na (atom) + Na+ (ion) + e K (atom) + K+ (ion) + e As the ease of ionisation increases in the order of lithium, sodium, potassium, rubidium and caesium, it follows that potassium will ionise to a greater extent than sodium and hence produce a greater concentration of electrons.If sodium is also present the effect would be to displace the sodium equilibrium to the left, increase the concentration of sodium atoms and thus the amount of light passing through the sodium filter. For the same reasons it is unlikely that sodium could affect the potassium intensity, and these predictions are borne out by the facts observed. It is clear that this effect could be overcome either by matching the Na20 - K,O of the standards to that of the sample or by adding an excess of potassium to ensure maximum light emission. Neither of these solutions was attractive as both would complicate the determination of the two main alkalis.Caesium had already been used as an ionisation buffer by Sanui and Pace.5 They used a concentration of 550p.p.m. of caesium, but as their work was carried out with an air- acetylene flame in which the ionisation could be expected to be greater as a result of the higher temperature, it seemed reasonable to expect that less caesium would be required for the lower temperature of the air - propane flame. The addition of both aluminium sulphate and an ionisation buffer would need to be made on an aliquot of the main alkali solution, as EDTA end-points for lime and magnesia, which were also determined on this solution, would deteriorate in the presence of a large excess of aluminium.A 1 + 1 dilution would appear to be convenient and it would then follow that by using a standard at 10 p.p.m. of K,O and 5 p.p.m. of Na20, full-scale deflection would be equivalent to 2 per cent. of K20 and 1 per cent. of Na20 (0.25 g of sample in 250 ml of stock solution). As this would obviate the need for dilution in most sodium determinations, it was decided to propose the use of standards containing 20 p.p.m. of K,O, thus eliminating the need for any dilution with most samples. Calibration graphs indicated that the degree of curvature at the higher concentration was acceptable. To determine the amount of aluminium required to overcome the effect of up to 10 per cent. of calcium oxide in the sample (50 p.p.m. as presented to the instrument), a solution of this concentration was sprayed with the addition of 0, 100, 200, 300, 400 and 500 p.p.m.of A1203. None of the normal sources of aluminium sulphate was found to be of sufficient purity; ultimately, a satisfactory solution was prepared from aluminium metal. The metal is dissolved in nitric and sulphuric acids, and the solution is evaporated until all nitrous fumes have been removed and the sulphuric acid fumes strongly; the residue is then dissolved in water. TABLE VIII Blanks were also measured. The mean readings are shown in Table VIII. EFFECT O F ALUMINIUM ADDED AS SULPHATE ON THE SPECTRAL INTERFERENCE O F LIME I N THE SODIUM DETERMINATION Full scale set on 5 p.p.m. of Na,O - 10 p.p.m. of K20; 50 p.p.m. of CaO Also, added, p.p.m. . . .. . . 0 100 200 300 400 500 Mean reading less blank .. . . 10.75 1 0.25 0.25 0.25 <0*25 The presence of 200 p.p.m. of A1,0, appears to be adequate to suppress the calcium emission. This concentration of alumina had no interfering effect on the intensity of the alkali emission. After establishing the desired Al,03 concentration it was now necessary to determine the amount of caesium required to overcome ionisation effects. Additions of 10, 20 and 30p.p.m. of caesium were made to solutions containing 200 p.p.m. of A120, and various amounts of Na,O and K20 (Table IX).706 EARDLEY AND REED : THE FLAME-PHOTOMETRIC [Artalyst, Vol. 96 TABLE IX EFFECT OF INCREASING CAESIUM ADDITION ON THE INTENSITY OBSERVED 200 P.P.M. OF AI20, Full scale set on 5 p.p.m. of Na20 - 20 p.p.m. of K20 FROM 4 P.P.M.OF Na20 AND 16 P.P.M. OF K20 IN THE PRESENCE OF Solution concentration, p.p.m. Sodium reading Potassium reading Na20 K2O A1203 cs Mean Mean less blank Mean Mean less blank r A \ & & 4 4 4 4 4 4 0 0 0 0 0 0 16 16 16 0 0 0 16 16 16 0 0 0 200 200 200 200 200 200 200 200 200 200 200 200 10 20 30 10 20 30 10 20 30 10 20 30 82.3 83.3 83.4 81.8 82.8 83.2 - - 0.8 1.7 2.0 81.5 81.6 81.4 81.0 81.1 81.2 - 87.8 91-3 94-0 - - 87.8 91.0 94-4 1.9 4.3 7.0 85.9 87.0 87.0 - The results indicate that 30 p.p.m. of caesium are required to ensure freedom from ionisation effects for sodium and potassium. The efficiency of the addition in the presence of various alumina contents up to a maximum of 450 p.p.m. of A120, was examined (i.e., addition of 200 p.p.m. to overcome interference by lime plus 250 p.p.m., for example, from the 50 per cent.contributed by raw bauxite). Although the effect of alumina variation was negligible there seemed to be a slight indication of an effect on ,the potassium intensity (see Table X). It therefore seemed advisable to calibrate the instrument in the presence of the median alumina content of, say, 300 p.p.m. of A120,. TABLE X EFFECTS OF VARIOUS ALUMINA CONTENTS ON THE INTENSITY FROM Full scale set on 5 p.p.m. of Na20 - 20 p.p.m. of K20 Na20 K2O A1203 cs r S o d i u m m 4 16 200 30 81.8 86.7 4 16 320 30 81.8 86.4 4 16 450 30 81-6 85.9 4 P.P.M. OF Na20 AND 16 P.P.M. OF K20 IN THE PRESENCE OF 30 P.P.M. OF CS Solution concentration, p.p.m. Mean reading less blank < A I CALIBRATION FOR SODIUM, POTASSIUM AND LITHIUM- A propane flow-rate of 400 ml minute-l was reached as the mean figure obtained when several analysts set the flame to give “well defined blue cones.” The variation in gas setting was considerable, but although the sensitivity varied the shape of the calibration graph was unchanged.Lithiam-The sensitivity of the instrument was inadequate to permit calibration with less than 20 p.p.m. of Li20 standard and for this reason it was decided to continue the practice of determining this element on the undiluted stock alkali solution. As potassium is known to interfere spectrally with lithium a graph was constructed of this interference (Fig. 2) to enable it to be corrected for. 0.02 0.0 1 0 1 2 3 4 5 6 7 Equivalent percentage of K2O Fig. 2. Spectral interference of potassium on lithium (1 per cent.E 10 p.p.m.)October, 19711 DETERMINATION OF ALKALIS I N CERAMIC MATERIALS 707 Sodium and $otassium-Sodium was then calibrated for up to 1 per cent. (5 p.p.m.) of Na20 and potassium for 4 per cent. (20 p.p.m.) of K20 on the assumption that the stock solution would be at a concentration of 1 g 1-1 of sample and that this solution would be diluted 1 + 1. Calibration solutions, in addition to their appropriate alkali contents, also contained 30 p.p.m. of caesium, 300 p.p.m. of A120, and nitric acid so as to be as nearly as possible equivalent to the diluted sample solution. The curvature of the potassium calibration graph is slightly greater at 20 p.p.m. than at 10 p.p.m. full-scale deflection, but is still acceptable.Each scale division of the top 10 per cent. of the graph is now equal to 0.055 per cent. of K,O whereas, on the previous basis (10 p.p.m. of K20) and allowing for the 4-fold dilution needed, it was 0.044 per cent. TRIALS ON SYNTHETIC SAMPLES- As most of the available standards have been analysed by the old procedure the accuracy of the alkali determinations must be in doubt. Synthetic sample solutions were therefore prepared covering a wide range of composition, with and without the appropriate alkali content, the latter solutions acting as blanks. The composition of these standards and the results for alkalis are shown in Table XI. The solutions were made at a sample concentration of 1 g 1-1 except for the high-silica materials, which were made at 4 g 1-1 as required by the British Standard method.TABLE XI COMPOSITION OF SYNTHETIC STANDARDS AND RESULTS FOR ALKALIS Oxide content, per cent. Synthetic samples CAS3 .... .. CAS6 .. .. CAS7 .. .. B.C.S. 269 . . .. B.C.S. 316 . . .. B.C.S. 309 . . .. B.C.S. 313 . . .. B.C.S. 314 . . .. B.C.S. 267 . . .. us 102 . . .. 2CAS 11 .. .. 2CAS 13 .. .. US 70a . . .. US 99a . . .. AN 30 . . .. Building clay Clay Earthenware body Firebrick Firebrick Sillimanite High purity silica Silica brick Silica brick Silica brick Potash feldspar Soda feldspar Potash feldspar Feldspar Borax frit Al,O, 16 30 18 32 40 40* 0.2 0.8 0.8 2 18 16 18 20 8 Fe,% 6 1 0.4 3-2 3.2 1.6 0.04 0-5 0.8 0.65 0.4 0.1 0.1 0.1 0.2 CaO 2.4 0.2 0.8 0.2 0.4 0.4 0.02 1.8 1.8 2-3 0.4 0.25 0.1 0.6 14 MgO 2 0.4 0.4 1 0.4 0.2 0 0.05 0.05 0.15 0.1 0 0 0 0.2 K,O K,O Na,O added found added 2.60 2.46 0.40 2.50 2.46 0.40 1-26 1.27 0.60 2.60 2.49 0.50 0.50 0.51 0-10 0-50 0.48 0.30 0.025 0.023 0.013 0-063 0.063 0-060 0.125 0.128 0.075 0.313 0.313 0.025 10.0 10.0 3.00 0.75 0.77 8.00 12.5 12.5 2-50 6.00 4-98 6.00 1.25 1.27 9.00 Na,O found 0.49 0.49 0.49 0.48 0.10 0.29 0.013 0-049 0.073 0-025 2.95 7.95 2.48 6.00 9.00 Li,O added 0.02 0.04 0.01 0.13 0.09 0.01 0 0.01 0 0 0 0 0 0 0.10 Li,O found 0.03 0.04 0.0 1 0.13 0.09 0.01 0 0.015 0.006 0.006 0 0.01 0 0 0.13 CAS, 2 CAS and AN samples are analysed samples used by the British Ceramic Research Association * To represent the proportion of alumina dissolved during decomposition.for co-operative work.[Autalyst, Vol. 96 The blank solutions were made for a dual purpose, firstly to evaluate the sodium and potassium blanks arising from the additions of other constituents and also to evaluate the potential interferences in the determination of lithium.Positive blank errors for lithium were found in samples equivalent to B.C.S. 314, B.C.S. 267, US 102 and AN 30. All of these samples had a low Al,O,-to-CaO ratio. Although the errors are small their occurrence would suggest that caution must be applied in the interpretation of results for lithium if the A1,0,- to-CaO ratio in the stock solution is less than that required to overcome the calcium inter- ference, i.e., 4: 1. The remainder of the results were satisfactory and so several standard samples were re-analysed to determine the alkali content. ANALYSIS OF ACTUAL SAMPLES- The results (Table XII) indicate, as might be expected, little difference between the old and new figures when the alkali content is low.With the feldspars the results show the expected movements in that the potassium results are generally higher following removal of the depressant effect of chlorine-containing acids, and the sodium results are higher because of removal of ionisation errors. TABLE XI1 ALKALI DETERMINATIONS ON ANALYSED AND STANDARD SAMPLES 708 EARDLEY AND REED : THE FLAME-PHOTOMETRIC K,O, per cent. Na,O, per cent. Li,O, per cent. Previous Sample type Code Found value Found value Found value - Building clay . . CAS 3 2.44 2.41 0.58 0.63 0.02 0.02 Clay . . .. CAS5 2-57 2-65 0.37 0.40 0.02 0.04 Earthenware body CAS 7 1-51 1.45 0.72 0.74 0-03 0.01 Firebrick .. .. B.C.S. 269 2.63 2-62 0.38 0.37 0-13 0-13 Firebrick . . . . B.C.S. 315 0.56 0.52 0.12 0.13 0.10 0.09 Sillimanite . . . . B.C.S. 309 0.48 0.46 0.31 0.34 0.02 0.01 High purity silica . . B.C.S. 313 0.04 0.04 0.007 <0*01 <0.01 <0*01 Silica brick . . B.C.S. 314 0.09 0.09 0.04 0.05 <0-01 0.01 Silica brick . . B.C.S. 267 0.13 0.14 0.05 0.06 t 0 . 0 1 - Silica brick .. us 102 0.33 0.32 0.015 0.015 <0-01 <0*01 Feldspar . . . . 2 CAS 11 10.61 10.35 2.91 2.88 <Om01 - Feldspar . . .. 2 CAS 13 0.65 0.62 8-28 8.02 0.0 1 - Feldspar . . . . US 70a 11.78 11.8* 2.48 2*65* 0.01 - Feldspar . . .. US 99a 5.34 5.2* 6.26 6.2* 0.01 - Borax frit . . .. AN 30 1-14 1.13 8.49 8*2t to 8.8: 0.05 0.10 Previous - Previous - * United States Bureau of Standards (provisional figures).t Determined by the old flame-photometric method; closer simulation of standards gave 8.44. Gravimetric method, with the zinc uranyl acetate precipitate dried at room temperature. Comparison with the borax frit is difficult because of the wide range of results previously obtained. The original flame-photometric results were low, as might be expected from the ionisation effects, and most errors in the gravimetric method (giving sodium zinc uranyl acetate) used tend to be positive. ALKALIS IN HIGH-LIME MATERIALS- For samples of high-lime content such as bone ash, dolomites, cements and calcium aluminates, the addition of 200 p.p.m. of Al,O, was not expected to remove the interference in the determination of sodium. To allow for the effects of 60 per cent.of calcium oxide, it was found that the addition of alumina needed to be increased to 1500 p.p.m. To test the accuracy of the modification, standards were synthesised representing the top, bottom and mid-points of the calibration graph for sodium and potassium. The degree of bowing was identical with that with the lower Al,O, content. Two high-lime samples were synthesised representing a bone ash and an ignited dolomite in both a “test” and blank form. The results for the determination of alkalis in these samples are shown in Table XIII, together with the composition of the synthetic solutions. “Blanks” of 0.03 and 0.07 per cent. of Li,O were found for the bone ash and dolomite, respectively. In view of the low alkali contents of this type of sample, it may be better to treat the whole of the sample solution with caesium and Al,O, and then determine Na,O, K,O and Li,O without dilution.It is not usual to determine calcium and magnesium oxides on the solution prepared for the alkali determination with these types of material. However, it would seem that atomic-absorption spectroscopy offers a better solution to the problem of the analysis of these materials for alkalis.October, 19711 DETERMINATION OF ALKALIS I N CERAMIC MATERIALS TABLE XI11 COMPOSITION AND RESULTS FOR SYNTHETIC SAMPLES WITH HIGH LIME CONTENT Oxide content, per cent. 709 K20 K20 Na20 Na20 Li,O Li,b Sample A1,0, Fe20, CaO MgO P20, added found added found added found Bone ash 0.1 0.1 56 1 40 0.05 0.04 0.30 0.30 0.05 0.08 Dolomite 0.2 0.2 60 40 - 0.05 0.05 0.10 0.10 0-05 0.12 CONCLUSIONS Town gas is now too variable in composition to form the basis of an instrumental method for the flame-photometric determination of alkalis, and methane has been found to be less satisfactory than propane with regard to depressive effects and to flame stability.This would suggest that North Sea gas would be equally unsuitable. Bottled propane would seem to offer the best source gas of constant composition and is less subject to depressive effects. The flame-photometric method, although free from interference from anions as originally given, has now been found to be subject to depressive effects because of the chlorine-containing acids used for the decomposition, to give low results for some sodium determinations because of ionisation effects and to give some high results because of spectral interference of calcium on sodium and lithium.The depressive effects caused by the presence of hydrochloric and perchloric acids can be removed by the use of sulphuric and nitric acids for the decomposition of the sample. The original method showed no discernible interference of one alkali on another, but it was now found that potassium interfered in the determination of sodium. The tendency to obtain low results for sodium when the K,O-to-Na,O ratio is less than 2: 1 can be overcome by the addition of a caesium sulphate ionisation buffer to give a final concentration of 30 p.p.m. of caesium. The spectral interference of calcium on the sodium determination can be excluded by the addition of aluminium sulphate solution (containing 200 p.p.m.of A1,0,) to an aliquot of the main alkali stock solution. The interference on lithium is not serious unless the alumina-to- lime ratio in the sample is less than 4: 1. When this is so, more accurate results would be obtained either by measurement on the aliquot diluted for the sodium and potassium deter- mination, with a 2-fold loss in sensitivity, or by addition of caesium and aluminium to the whole of the stock alkali solution. If accurate results for lithium oxide are required the additions of caesium and an increased amount of aluminium must be made to the whole of the 250 ml of stock alkali solution. METHOD APPARATUS- The modification essential to the use of propane gas is shown in Fig.1. Optional modi- fications that aid the reproducible operation of the instrument are replacement of the carbon potentiometer used in the sensitivity control by a 2O-kQ wire-wound helical potentiometer and the addition of a 250-pF electrolytic condenser to improve the damping of the galvano- meter. REAGENTS- Unless otherwise stated, all reagents should be of analytical-reagent grade when available and distilled water should be used throughout the analysis. Aluminium sulphate solution equivalent to approximately 10 mg ml-l of Al,O,-Clean analytical-reagent grade aluminium metal by washing it with hydrochloric acid, ethanol and ether. Weigh 10.58 g of the clean, dry metal, add 40 ml of sulphuric acid (sp.gr. 1-84), 120 ml of nitric acid (sp.gr. 1.42) and about 50ml of water. Allow to react in the cold, then raise the temperature gradually until dissolution of the metal is complete and heat the solution until all of the nitric oxides are expelled and the sulphuric acid fumes strongly.Cool, dissolve the crystalline melt in distilled water and dilute to 2 litres. Aluminium sulphate solution equivalent to approximately 2000 9.p.m. of A120,-Dilute 400 ml of the above aluminium sulphate solution to 2 litres.710 EARDLEY AND REED : THE FLAME-PHOTOMETRIC [Analyst, Vol. 96 Caesium sulphate solution equivalent to approximately 300 p.p.m. of Cs-Dissolve 0-41 g of Cs2S0, (Johnson Matthey Specpure) in 1 litre of water. Hydro$uoric acid, 40 per cent. Dilute nitric acid (1 + 19). Sulpkuric acid - nitric acid mixture-To 650 ml of water, add 100 ml of dilute sulphuric acid (1 + 1) and 250 ml of nitric acid (sp.gr.1.42). STANDARD SOLUTIONS- Lithium solution A equivalent to 400 +.p.m. of Li20-Dehydrate lithium sulphate mono- hydrate by heating it at 150 "C for 24 hours. Dissolve 1.4719 g of the anhydrous lithium sulphate in 1 litre of water. Lithium top standard equivalent to 20 $.p.m. of Li,O-Dilute 25eOml of the lithium solution A plus 40 ml of dilute nitric acid (1 + 19) to 500 ml. Potassium - sodium solution A equivalent to 400 p.p.m. of K20 and 100 p.p.m. of Na20- Dissolve 0.7400 g of anhydrous potassium sulphate (dried at 150 "C) and 0.2292 g of an- hydrous sodium sulphate (dried at 150 "C) in 1 litre of water. Potassium - sodium solution B equivalent to 40 p.9.m. of K20 and 10 P.9.m.of Na20- Dilute 50.0 ml of the potassium - sodium solution A to 500 ml. Potassium - sodium top standard equivalent to 20 p.p.m. of K20 and 5 $.p.m. of Na20- To a 1-litre calibrated flask, add 40 ml of dilute nitric acid (1 + 19), 100 ml of the caesium sulphate solution, 30 ml of aluminium sulphate solution (equivalent to approximately 10 mg ml-l of A12.03) and 50.0 ml of the potassium - sodium solution A; dilute to 1 litre. Potasszum - sodium zero standard-This solution is used as a diluent for alkalis in excess of 4 per cent. of K20 and 1 per cent. of Na,O. To a 1-litre calibrated flask, add 40ml of dilute nitric acid (1 + 19), 100 ml of the caesium sulphate solution and 30 ml of aluminium sulphate solution (equivalent to approximately 10 mg ml-l of A1203) ; dilute to 1 litre.Check the blanks for Na20 and K20 by spraying with the zero set on distilled water and the full scale set on the top standard. Typical values are deflections of two divisions for Na20 and seven for K20. Readings greatly in excess of these invalidate the use of the buffer, and the origin of the blank must be investigated and rectified. INTERMEDIATE CALIBRATION SOLUTIONS- Lithiztm-To four 50-ml calibrated flasks, add 4ml of dilute nitric acid (1 + 19) and 10, 20, 30 and 40ml of the lithium top standard; dilute to 50ml. These concentrations are equivalent to 4, 8, 12 and 16 p.p.m. of Li20 or 0.4, 0.8, 1.2 and 1.6 per cent. of Li20. Potassium - sodium-To seven 200-ml calibrated flasks, add 8 ml of dilute nitric acid (1 + 19), 20ml of the caesium sulphate solution, 6 ml of aluminium sulphate solution (equivalent to 10 mg ml-l of Al,03), and 5, 10, 20, 40, 50, 60 and 80 ml of the potassium- sodium solution B to give 1,2,4,8, 10,12 and 16 p.p.m. of K20 and 0*25,0*5,1,2,26, 3 and 4 p.p.m.of Na20, respectively, equivalent at the dilution used to 0.2, 0.4, 0.8, 1.6, 2.0, 2.4 and 3-2 per cent. of K20 and 0.05, 0.1, 0.2, 0.4, 0.5, 0-6 and 0.8 per cent. of Na20, respectively. CALIBRATION- and full-scale deflection on the appropriate top standard. Construct calibration graphs by setting the zero of the instrument on the zero standard DECOMPOSITION OF THE SAMPLE- Weigh 0.250 g (1 g for high-silica materials) of the sample dried at 110 "C into a platinum dish and ignite gently to remove organic matter. To the cool dish, add 10 ml of sulphuric acid- nitric acid mixture and 10 ml of hydrofluoric acid, and evaporate to dryness on a sand-bath in a fume cupboard, taking care to prevent spurting.Cool the dish, add 10 ml of the sulphuric acid-nitric acid mixture and rinse down the sides of the dish with water. Evaporate carefully to dryness. To the cool dry residue, add 20 ml of dilute nitric acid (1 + 19) and warm to dissolve. Cool, filter the solution if necessary through a No. 42 Whatman filter-paper and wash the dish and filter-paper with cold water. Dilute the filtrate and washings with water to 250 ml to give stock solution A.October, 19711 DETERMINATION OF ALKALIS IN CERAMIC MATERIALS 711 DETERMINATION OF ALKALIS- Potassium and sodium-Transfer 25.0 ml of the stock solution A to a 50-ml calibrated flask containing 5 ml each of the caesium sulphate solution and aluminium sulphate solution (equivalent to 2000 p.p.m.of A1203) (Note). Dilute to 50 ml to give solution B. Set up the instrument, insert the appropriate filter and spray the sample solution B, setting full scale on the potassium - sodium top standard and the zero with the potassium - sodium zero standard. If the alkali content is in excess of the normal range, readings can be brought on to scale, either by dilution of solution B with zero standard or by appropriate dilution of an aliquot of solution A before addition of caesium and aluminium sulphates. It is essential to maintain, in the solutions presented to the instrument, concentrations equivalent to additions of 5ml of the caesium sulphate solution and 5ml of aluminium sulphate solution (equivalent to 2000 p.p.m.of A1203) in 50ml of solution. NOTE- Attempts to simplify the procedure by making a mixed solution of caesium and aluminium sulphates are not advised, as serious loss of caesium will occur because of the slow precipitation of sparingly sol- uble caesium aluminium sulphate. Lithium-Insert the lithium filter and spray the stock solution A, setting full scale on 20 p.p.m. of Li20 and zero on distilled water. Lithium is subject to positive errors caused by potassium light passing through the filter. The magnitude of the error must be ascertained by setting up the instrument as for the lithium calibration and spraying solutions of 20, 50 and 100 p.p.m. of K20, equivalent to sample contents of 2,5 and 10 per cent. of K20, and a correction graph prepared. Note the reading and apply the appropriate correction for the K20 content determined, by reference to the correction graph. EXTENSION OF THE GENERAL METHOD TO THE ANALYSIS OF HIGH-LIME MATERIALS Only the modification for sodium and potassium will be described. The determination of lithium is rarely required in these materials and, because of the low sensitivity and suscepti- bility to interference, it is of doubtful accuracy. SPECIAL SOLUTIONS REQUIRED- High-lime potassium - sodium zero standard-To a l-litre calibrated flask, add 40 ml of dilute nitric acid (1 + 19), 100 ml of the caesium sulphate solution and 150 ml of aluminium sulphate solution (equivalent to approximately 10 mg ml-1 of A1203) ; dilute to 1 litre. High-lime potassium - sodium to$ standard equivalent to 20 p.$.m of K,O and 5 p.9.m. of Na20-To a 1-litre calibrated flask, add 40 ml of dilute nitric acid (1 + 19), 100 ml of the caesium sulphate solution, 150 ml of aluminium sulphate solution (equivalent to approxi- mately 10mgml-1 of A1203) and 50.0ml of the potassium -sodium solution A; dilute to 1 litre. PROCEDURE FOR THE DETERMINATION OF SODIUM AND POTASSIUM- Transfer 25.0 ml of the stock solution A to a 50-ml calibrated flask containing 5 ml of the caesium sulphate solution and 7.5 ml of aluminium sulphate solution (equivalent to about 10 mg ml-1 of A1203). Dilute to 50 ml. Spray the solution against the top and zero standards synthesised above. We thank Dr. N. F. Astbury, Director of Research of the British Ceramic Research Association, for permission to publish this paper. REFERENCES 1. 2. 3. 4. 6. Bennett, H., Eardley, R. P., and Hawley, W. G., Trans. Brit. Caram. SOL, 1956, 57, 1. Andrew, T. R., and Nichols, P. N. R., Analyst, 1967, 92, 156. Spiers, H. M., “Technical Data on Fuel,” Sixth Edition, British National Committee, World Power British Standard 1902 : Part 2A : 1964. Sanui, H., and Pace, N., AppZ. Spectrosc., 1966, 20, 135. Conference, London, 1961. Received July 29th. 1970 Accepted March 22nd, 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600699

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

712 Analyst, October, 1971, Vol. 96, pfi. 712-715 A Potentiometric Procedure for the Assay of Isonicotinic Acid Hydrazide (Isoniazid) BY P. V. KRISHNA RAO AND G. BALA BHASKARA RAO (Chemistry Defiartment, Andhra University, Waltair, India) A potentiometric procedure for the assay of isonicotinic acid hydrazide (isoniazid) with vanadium(V) at room temperature is described. The reduction of vanadium(V) to vanadium(1V) by isoniazid in an acidic medium is cata- lysed by osmium tetroxide, and the application of this method to the assay of isoniazid in pharmaceutical preparations is considered. Oxalic acid interferes in the determination, although commonly used excipients such as starch, dextrin, sucrose, glucose, lactose and gum acacia do not interfere. SEVERAL c~lorimetricl~~s~s~ and titrimetric5 to 17 procedures have been reported for the assay of isoniazid in pharmaceutical preparations.The hydrazino group in the compound is sus- ceptible to oxidation and many of the cited titrimetric procedures depend on this property. The potentiometric titration of isoniazid with potassium bromate7 in an 8 to 12 per cent. hydrochloric acid medium and in the presence of potassium bromide is considered to be the best method, although Kuhni, Jacob and Grossglauser13 state that the titration of isoniazid with 0.05 N potassium bromate gave results that were 0.5 per cent. too high. The redox reaction between isoniazid and quinquivalent vanadium does not appear to have been considered as the basis of a quantitative titrimetric method. Recently Krych and Lipiec18 have reported a spectrophotometric method for the determination of vanadium(V) with isoniazid that involves measuring at a wavelength of 420nm the absorbance of the orange-red complex formed between the reactants at a pH of 1.98. These authors state that the complex is not stable for longer than 10 to 15 minutes and that the results are reproducible only to within 4 per cent.Gowda and Gopala Raolg proposed a method for the assay of isoniazid that involves treating an aliquot of an aqueous solution of the compound with an excess of 0.05 M sodium vanadate solution in a medium of 4 M sulphuric acid, allowing the reaction mixture to stand for one minute and titrating the unreacted vanadate with a standard solution of ammonium iron(I1) sulphate. N-Phenylanthranilic acid is the redox indicator used in this case.These authors claimed definite advantages for their method over iodimetric and bromate titration procedures because the common excipients such as lactose, glucose and starch did not interfere in the determination. We have undertaken a detailed study of the reaction and have succeeded in developing an accurate potentiometric method for the assay of the compound in pharmaceutical preparations. EXPERIMENTAL REAGENTS- Sodium vanadate solution, 0.1 N-A standard vanadium(V) solution was prepared (by dissolving sodium orthovanadate in water) and standardised against a standard solution of potassium &chromate, which was in turn standardised against a solution of ammonium iron( 11) sulphate, the end-points in both the titrations being located potentiometrically.Isonicotinic acid hydraxide-The isoniazid used in this investigation was of U.S.P. grade. A 0.05 M solution in water was prepared from a sample that had been twice recrystallised from aqueous ethanol and dried at 110 "C for 2 hours. The aqueous solution thus prepared was standardised against a standard solution of potassium bromate following the potentio- metric method described by Vulterin and Zyka.7 Osmium tetroxide-A 0-1 per cent. solution in 0.1 N sulphuric acid was prepared and stored in an amber glass bottle (sample supplied by Johnson Matthey Chemicals Ltd., London). 0 SAC and the authors.RAO AND RAO 713 This reagent (in 0.2-ml amounts) was used as a catalyst in all the experiments described in this paper.Orthophosphoric acid-E. Merck’s “Pro Analysi” grade orthophosphoric acid (85 per cent .) was used throughout this investigation (any analytical-reagent grade acid can be used). SuZphuric acid-Analytical-reagent grade nitrogen-free sulphuric acid was used without further purification. All other reagents and chemicals used were of analytical-reagent grade. Nydrazid and Isonex tablets, manufactured by Squibb Pharmaceuticals and Dumex Pharmaceuticals (India), respectively, were extracted with water and the aqueous extracts analysed to determine the active constituent (isoniazid). APPARATUS- The potentiometric titration assembly used consists of a Pye potentiometer graduated in millivolts, a galvanometer (Cambridge Instrument Co., London), a saturated calomel reference electrode, a bright-platinum rod (0.2 mm in diameter) as indicator electrode and a porous-plate salt-bridge filled with a saturated solution of potassium chloride.The titration mixture is stirred with an electromagnetic stirrer. Preliminary experiments showed that a direct titration of isoniazid with sodium vanadate in an acidic medium containing any mineral acid would not be possible because the red complex formed on the addition of sodium vanadate decomposed slowly and the potentials were not stable. The use of several catalysts such as orthophosphoric acid, iodine mono- chloride, osmium tetroxide and cepper(I1) sulphate did not improve the situation. The reverse titration, i.e., that of vanadate with isoniazid, is also slow at both room and elevated temperatures in media containing various concentrations of sulphuric, perchloric and phos- phoric acids (ranging from 0-25 to 4.0 M).However, the reaction was markedly catalysed by osmium tetroxide, although in a 0-5 to 2 . 0 ~ sulphuric acid medium the potentials did not attain stable values. Nevertheless, further experiments showed that the addition of 1.0 ml of orthophosphoric acid gave stable potentials. We have therefore carried out experi- ments to ascertain the optimum concentrations of the catalyst and phosphoric or sulphuric acid that will give a satisfactory potentiometric titration, each time keeping one of the parameters constant. Efect of varying the catalyst coutcentration-Experiments on the variation of the osmium tetroxide catalyst concentration in the range 0.05 to 2.0rnl of a 0.1 per cent. solution in 50 ml of the titration mixture (containing the optimum concentration of either phosphoric acid or sulphuric acid for this reaction) have shown no significant deviations in the accuracy of the method.However, we noticed that when the catalyst solution was present in volumes below 0.2 ml, the potentials were not stable near the equivalence point, thus causing consider- able delays with each determination. Efect of varying phosphoric or sdphuric acid concentration--In these experiments 0.2 ml of osmium tetroxide was added to the titration mixture and the acid concentration was varied. Experiments showed that variation of the phosphoric acid volume from 0.5 ml to 20.0 ml (in a total volume of 50ml of titration mixture) did not affect either the accuracy or the speed of the titration, while greater concentrations of phosphoric acid resulted in a higher consumption of isoniazid and an abnormal drift in the potentials near the equivalence point.Moreover, the potential break near the equivalence point was not sharp but evenly distributed between successive additions of the titrant, thus leading to substantial errors. In the case of titrations in media 0.25 to 4.0 M in sulphuric acid, we found that without the addition of at least 1.0 ml of phosphoric acid the potentials were not stable and that the break in potential at the equivalence point could not be located accurately. At higher concentrations of sulphuric acid the results were always several per cent. lower, even if phos- phoric acid and osmium tetroxide were added.Because of these findings, we recommend the following procedure for the potentiometric assay of isoniazid with sodium vanadate a t room temperature. RECOMMENDED PROCEDURE- A suitable volume of the standard vanadium(V) solution is transferred to a 150-ml titration vessel and 1 to 20 ml of 85 per cent. orthophosphoric acid (or a mixture of 2.5 to 10.0 ml of sulphuric acid (1 + 1) and 1.0 ml of 85 per cent. orthophosphoric acid) 0.2 ml714 RAO AND RAO: A POTENTIOMETRIC PROCEDURE [Analyst, Vol. 96 of 0.1 per cent. osmium tetroxide solution are added. The resulting mixture is diluted to 50 ml and titrated potentiometrically with 0.05 M isoniazid solution, the potentials being noted 1 minute after the addition of each portion of the titrant.A typical potential v e ~ s u s volume graph is given in Fig. 1. 2 Isoniazid/ml Fig. 1. Potentiometric titration of vanadium(V) with isoniazid The potential break at the equivalence point is about 100 to 120 mV per drop (approxi- mately 0.04 ml) of 0.05 M isoniazid. A large number of determinations of isoniazid have been carried out according to the recommended procedure and the results compared with those obtained by the B.P. method.20 Some typical results are given in Table I and show that the proposed method yields results that agree with those by the standard B.P. method with an average deviation of 0.45 per cent. TABLE I COMPARATIVE ASSAYS OF ISONIAZID Isoniazid/mg By B.P. method . . 1.04 3.19 6.04 9-34 12-2 16.3 22-4 25.9' By proposed method .. 1.05 3.21 6.03 9.29 12.1 16.4 22.4 26.1 Deviation, per cent. . . 0.96 0.63 0.17 0.54 0.82 0.61 0.00 0.71 Mean deviation = 0.56 per cent. APPLICATION OF THE PROPOSED METHOD TO THE ASSAY OF ISONIAZID IN PHARMACEUTICAL PRE- Two tablets are dissolved in 50 ml of water, the resulting mixture is filtered through an IG4 sintered-glass crucible and the filtrate is made up to 100 ml. This solution is then trans- ferred to a microburette and the isoniazid content ascertained by potentiometric titration against a standard solution of vanadium(V) according to the recommended procedure. The results thus obtained were compared with those obtained by using the standard B.P. method,20 and are given in Table 11. The relative deviation for each method is also shown. Interfereutces-Sucrose, glucose, lactose, starch, dextrin and gum acacia, which are usually used as excipients in pharmaceutical preparations, do not interfere in this deter- mination.However, oxalic acid interferes at all concentrations and uranium(V1) and chromium(II1) also interfere. In all the interference studies the excipients were added in amounts up to a 50-fold excess relative to the calculated amount of isoniazid at the end of each titration. PARATION S-October, 19711 FOR THE ASSAY OF ISONICOTINIC ACID HYDRAZIDE (ISONIAZID) TABLE I1 ASSAY OF ISONIAZID IN TABLETS 715 Amount of isoniazid found with the B.P. Deviations found with the proposed Deviations Amount of isoniazid Tablet assayed method/mg from average method/mg from average Nydrazid .... 94.9 94.7 94.9 95.3 95.4 95.4 Mean 95-1 Isonex . . . . .. 101.2 101.1 101.1 101.2 100.9 101.0 Mean 101.1 - 0.20 - 0.40 - 0.20 + 0.20 + 0.30 + 0.30 s.d. 0.30 + 0.10 0.00 0.00 + O * l O - 0.20 - 0.10 s.d. 0.12 95.4 94.9 95.4 94.9 95.0 95.5 Mean 95.2 101.4 101,o 100.9 101.0 101.1 101.3 Mean 101.1 + 0.20 - 0.30 4- 0.20 - 0.30 - 0.20 + 0.30 s.d 0.28 + 0.30 -0.10 - 0.20 -0.10 + 0.00 + 0.20 s.d. 0.20 DISCUSSION Under the experimental conditions prescribed, 4 moles of vanadium(V) are reduced per mole of isoniazid oxidised, according to the reaction- C,H,NCONH.NH, + 4 V(V) + H,O -+ C,H,NCOOH + 4 V(1V) + N, + 4H+ Whereas an orange-red 1: 1 complex formed between vanadium(V) and isoniazid at a pH of 1.98 has been reported by Krych and Lipiec,l* under our experimental conditions isoniazid is quantitatively oxidised to nitrogen and isonicotinic acid.(When a drop of osmium tetroxide solution is added to the vanadium(V) - isoniazid complex at pH 2, copious evolution of nitrogen occurs and the solution turns blue, thus showing that the complex undergoes internal oxidation - reduction under the catalytic influence of octavalent osmium.) Further work on the elucidation of the reaction mechanism is in progress and will be reported separately. One of us (G.B.B.R.) thanks the Council of Scientific and Industrial Research, India, for the award of a Junior Research Fellowship. We also thank Albert David (Private) Ltd., Calcutta, India, for the generous gift of a sample of isoniazid, and Professor G. Gopala Rao for his helpful suggestions and keen interest in this investigation.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. REFERENCES Naito, Takio, Shirai, H., and Oda, N., Bull. Nagoya City Univ. Pharm. Sch., 1955, 3, 34. Brettoni, B., Archs Ital. Sci. Farmac., 1955, 2, 227. Machek, G., Scientia Pharm., 1956, 24, 11. Erlenmeyer, H., and Fallab, S., Exfierientia, 1952, 8, 298. Haugas, E. A., and Mitchell, B. W., J. Pharm. Pharmac., 1952, 4, 687. Vulterin, J., Colln Czech. Chem. Commun., 1963, 28, 1391. Vulterin, J., and Zyka, J., Chemickk Listy, 1954, 48, 1745. Zyka, J., Ibid., 1954, 48, 1754. Vulterin, J., Thesis, Charles University, Prague, 1961. Laszlovszky, J . , Acta Pharm. Hung., 1960, 30, 101. Spacu, P., and Teodorescu, G., Revue Chim. Buc., 1957, 8, 42. Jancik, F., Cinkova, O., and Korbl, J., Colln Czech. Chem. Commun., 1959,24, 2695. Kiihni, E., Jacob, M., and Grossglauser, H., Pharm. Acta Helv., 1954, 29, 233. Berka, A., and Zyka, J., Chemickk Listy. 1956, 50, 314. Barakat, M. Z., and Shaker, M., Analyst, 1966, 91, 466. Van Pinxteren, J. .A. C., and Verloop, M. E., Pharm. Weekbl. Ned., 1964, 99, 1125. Devani, M. B., and Shishoo, C. J., J . Pharm. Sci., 1970, 59, 90. Krych, Z., and Lipiec, T., Chemia Analit, 1967. 12, 535. Gowda, H. S., and Rao, G. G., 2. analyt. Chem., 1959, 165, 36. “The British Pharmacopoeia, ” Pharmaceutical Press, London, 1963, p. 429. Received July 27th, 1970 Accepted January 12th, 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600712

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

716 Analyst, October, 1971, Vol. 96, pp. 716-720 The Determination of Nikethamide and Other Compounds in Pharmaceutical Dosage Forms by Thin-layer Chromatography BY W. M. CARMICHAEL* (Analytical Division, Ciba-Geigy Limited, 4000 Base1 2, Switzerland) The qualitative examination and quantitative assay of pharmaceutical dosage forms containing nikethamide by thin-layer chromatography is described. When applicable, the determination of accompanying active compounds such as adenosine, caffeine, strychnine and theophylline by the same method is also described, Individual quantitative determinations of the eluted drugs are performed by ultraviolet spectrophotometry. NN-DIETHYLNICOTINAMIDE (Coramine, Ciba ; generic name nikethamide) is a widely used respiratory stimulant for oral and parenteral use, which is formulated alone and in multi- component pharmaceutical dosage forms.It can be synthesised by treating nicotinic acid with thionyl chloride and allowing the resultant acid chloride to react with diethylamine hydroch1oride.l The liquid amide is purified by vacuum distillation, during which a very minor fraction consisting of N-ethylnicotinamide is occasionally obtained (Candolfi, E., and Hurzeler, H., personal communication) as a result of the presence of monoethylamine impurity in the diethylamine used for the synthesis. The two solvent systems used in the quantitative thin-layer chromatographic determination of nikethamide (see Table 11) enable the two compounds to be adequately separated. Nikethamide is a markedly stable compound, and although the amide linkage can be hydrolysed by refluxing it with acid or, more readily, with a base, hydrolytic or oxidative decomposition has not been observed in accelerated stability studies on dosage forms con- taining the compound.2 In the course of the present work, no evidence could be found in vitro of oxidative N-oxide formation or ring cleavage, although N-oxide formation is a known metabolic pathway for nikethamide in mammalian tissues.3 Copious analytical literature exists on the assay of nicotinic acid amides, including nikethamide. Dosage control of nikethamide has been traditionally carried out by non- aqueous titrimetry with perchloric and acetic a c i d ~ , ~ the method having been applied to both the free base and its calcium chloride and thiocyanate salts.Colorimetric methods have involved the use of the hydroxamic acid reaction for amides and subsequent complexa- tion with iron(III),5 ion-pair formation with bromophenol blue6 and the’ Zincke - Konig reaction,7,8 in which cleavage of the pyridine ring is followed by condensation of the resulting aldehyde with a suitable amine.g,lO,ll The last method suffers from the disadvantage that it requires the use of the highly toxic reagent cyanogen bromide. Gas-liquid chromato- graphic assay with a range of stationary phases has been described for nikethamide alone and in the presence of nicotinic acid.12J3 Several paper and thin-layer chromatographic methods for detecting nikethamide have been p ~ b 1 i s h e d . l ~ ~ ~ ~ ~ ~ ~ Spectrophotometric determination of the drug following its isolation by thin-layer chromatography has also been described17 in connection with the control of doping of horses.The present paper describes the identification and quantitative deter- mination of nikethamide in solid and liquid pharmaceutical dosage forms by thin-layer chromatography on silica gel. In the quantitative assays, the evaluation, with the exception of ephedrine (see below), is performed by ultraviolet spectrophotometry (Ama,. for nikethamide in ethanol = 261 nm; E = approximately 3400 a t a concentration of 0.02 mg ml-l and A = about 0.3). Absorption maxima and end dilution concentrations for the other four active compounds determined by this method are given in Table I. * Present address : Cilag-Chemie A.-G., 8201 Schaffhausen, Switzerland. 0 SAC and the author.CARMICHAEL EXPERIMENTAL REAGENTS- 717 All reagents used were of analytical-reagent grade.Standard substances , which were obtained commercially, were of pharmaceutical purity [e.g:, B.P., U.S.P. or DAB (Deutsche Arzneibuch) grades]. The nikethamide used was a typical production batch of Coramine.* The calcium chloride and thiocyanate salts of nikethamide were obtained from the same source. For thin-layer chromatography, plates pre-coated with a 0.25-mm layer of silica gel HF,,,? were used without further treatment. DEVELOPING SOLVENT SYSTEMS- 1. Chloroform - ethanol (95 per cent.) (10 + 1 v/v). 2. Ethyl methyl ketone - ammonia solution (sp.gr. 0.9) (10 + 1 v/v) or isopropyl alcohol - 3. Chloroform - ethanol (95 per cent.) - ammonia solution (sp.gr.0.9) (100 + 20 + 1 v/v). 4. Chloroform - methanol - glacial acetic acid (25 + 65 + 10 v/v). 5. Isopropyl alcohol - formic acid - water (70 + 20 + 10 v/v). 6. Ethyl acetate - ethanol (95 per cent.) (90 + 10 v/v). 7. Cyclohexane - chloroform - diethylamine (5 + 4 + 1 v/v). 8. n-Butyl alcohol - glacial acetic acid - diisopropyl ether - water (9 + 6 + 3 + 1 v/v), 9. Chloroform - ethanol (95 per cent.) - diethyl ether (5 + 5 + 2 v/v), plates developed ammonia solution (spgr. 0.9) (10 + 1 v/v). without chamber saturation. three times with intermediate drying in cool air stream. CHROMATOGRAPHIC PROCEDURE- The standard techniques of ascending thin-layer chromatography were used18 ; when not otherwise indicated, plates were developed over a distance of 10 to 15cm in a saturated chamber, saturation being achieved by lining the tank with strips of chromatographic paper soaked in developing solvent.At least half an hour was allowed for the tank atmosphere to reach equilibrium before plates were developed at room temperature. VIEWING OF SPOTS- Nikethamide was made visible by its fluorescence-quenching effect when developed plates were viewed under an ultraviolet lamp (254 nm) or by the Munier and Macheboeuf variation of the Dragendorff reagent (iodobismuthate, reagent No. 92 in reference 19). N-Ethyl- nicotinamide was made visible by fluorescence quenching or with the Reindel-Hoppe reagent (chlorine followed by o-tolidine, reagent No. 45 in reference 19), which also gave a faint coloration with nikethamide.Other active compounds accompanying nikethamide in dosage forms were also made visible by one of these methods or by charring with methanolic sulphuric acid (1 + 1 v/v) at 120 "C. QUALITATIVE EXAMINATION- For the identity test, active compounds were extracted from powdered solid dosage forms with water, ethanol or ethanol - water (1 + 1 v/v). Dilutions were normally adjusted to give a nikethamide concentration of 20 mg ml-l. Standard solutions of each active com- pound were prepared at the same concentration as for the test solution and with the same solvent ; 5 pl of standard and test solutions were chromatographed. Developed plates were dried in a stream of warm air, and the compounds were made visible. The identity of each active compound was confirmed when corresponding standard and test spots behaved in a similar way and exhibited the same Rp value after being made visible.QUANTITATIVE ASSAY- Dosage forms were extracted (by using an ultrasonic bath and temperatures of up to 50 to 60 "C, if necessary) or diluted with water or ethanol - water (1 + 1 v/v) to give a nikethamide concentration of 2 mg ml-l. A standard solution at the same concentration * Obtained from the Pharmaceutical Chemical Production Division, Ciba-Geigy (Klybeck) Limited, Basel. t Merck, Darmstadt.718 CARMICHAEL : THE DETERMINATION OF NIKETHAMIDE AND [Analyst, VOl. 96 was prepared with the same solvent. By using a 1OO-pI microsyringe (e.g., Terumo, Japan), 100 pf of standard and test solutions were applied as separate 5 to 7-cm long narrow bands on the starting line of a 20 x 20-cm pre-coated thin-layer chromatographic plate.Experience has shown that manual sample application can be carried out with a reproducibility of better than & 2 per cent. Before and after development of the plates, excess of solvent was removed with a stream of cool air. Standard and test zones of nikethamide were identified while viewing the developed plates under ultraviolet light (254 nm). After marking them, the relevant zones of adsorbent were scraped off the plate and each was placed in a 25-ml conical flask fitted with a ground-glass stopper. Standard and test blanks were prepared from a clean area of adsorbent of equal size. With a pipette, 10.0 ml of ethanol - water (1 + 1 v/v) were introduced into each flask and the suspensions were thoroughly shaken for about 5 minutes, either by hand or by mechanical shaker, before being centrifuged for 5 minutes at about 3000 r.p.m.The clear solutions were evaluated by ultraviolet spectrophotometry in 1-cm quartz cuvettes, the absorbance of each standard and test solution being measured at 261 nm against the corresponding blank. In the multi-component dosage forms listed in Table 11, caffeine, theophylline, adenosine and strychnine were determined in an analogous way. The caffeine and theophylline deter- minations were performed by using the plate used for the nikethamide assay, to which corre- sponding mixed standards were applied with the appropriate test solution. Higher relative concentrations were required for the adenosine and strychnine determinations, which were performed with separate plates.Each of these four compounds was determined at a wave- length and dilution normally adopted for its determination by ultraviolet spectrophotometry.20 These parameters for all five compounds determined in this way are presented in Table I. TABLE I END DILUTION CONCENTRATIONS AND ABSORPTION MAXIMA FOR ACTIVE COMPOUNDS DETERMINED BY THIN-LAYER CHROMATOGRAPHY, ELUTION AND ULTRAVIOLET SPECTROPHOTOMETRY Concentration a t Compound hmax./nm final dilutionlmg ml-l Nikethamide . . .. .. 261 0.02 Adenosine . . .. .. 260 0.005 Caffeine . . .. .. 272 0.005 Strychnine . . .. .. 254 0.02 Theoph ylline .. .. 270 0.008 Solvent: ethanol - water (1 + 1 v/v) RESULTS AND DISCUSSION The calcium chloride and thiocyanate salts of nikethamide are frequently incorporated in solid dosage forms, rather than the free amide, which is a liquid.As a result of the weak ion-exchange effect of the adsorbent, these salts behave like nikethamide itself when chro- matographed on silica gel, even when neutral developing solvents are used. Details of the dosage forms investigated and the thin-layer chromatographic results accumulated are presented in Table 11. All nine solvent systems are suitable for the identification of niketh- amide. Quantitative determinations have been performed by using the solvent systems 1 and 7, system 7 being particularly adaptable to nikethamide - caffeine combinations. For the quantitative determination of strychnine in the latter combinations, continuous develop- ment with methanol for 1 to 18 hours was found to be the best method of separation.While the strychnine travelled 2 to 3 cm from the starting line, nikethamide and caffeine migrated to the top of the plate. System 1 was adopted for the determination of each active compound in the nikethamide - adenosine combinations. Two developments were required to ensure that adenosine travelled sufficiently far from the point of application and was adequately separated from adenine, a potential contaminant arising from the acid-catalysed hydrolysis of adenosine. Ephedrine in nikethamide - ephedrine combinations was chromatographed with solvent systems 3 and 4 consecutively. After elution with methanol, ephedrine was determined by a previously published method involving colorimetry of the copper( 11) complex of the dithiocarbamate formed by reacting ephedrine with carbon disulphide in alkaline solution.=October, 19711 OTHER COMPOUNDS I N PHARMACEUTICAL DOSAGE FORMS 719 Quantitative thin-layer chromatography with solvent system 1 or 7, and for ephedrine systems 3 plus 4, serves as both a dosage control and stability assay procedure for nikethamide, adenosine , caffeine , ephedrine , theophylline and strychnine.Each compound is separated from adjuvants and from known decomposition products, should degradation occur. Adsorp- tion effects on adjuvants or on the chromatographic adsorbent were overcome by the high polarity of the extraction solvents used, and by the use of ultrasonics and elevated tempera- tures for extracting the active compounds from solid dosage forms.The latter were found t o be most efficiently pulverised by using a domestic electric coffee grinder prior to extraction. The absorbance of the blanks was found in every case to be below 0.05 absorbance units, unless a faint turbidity was present arising from inadequate centrifugation. This turbidity was readily removed from the standard, test and blank solutions by adding Celite analytical filter aid (Johns-Manville) and re-centrifuging the mixtures after briefly shaking them. In all instances, recoveries of active compounds from the adsorbent were found to exceed 95 per cent. when compared with parallel, unchromatographed solutions. In the absence of detectable decomposition, results obtained with the quantitative thin-layer chromatographic methods normally fell within &4 per cent.of the declared dosages. A spread of this order is to be expected for determinations of this type.22 TABLE I1 THIN-LAYER CHROMATOGRAPHIC RESULTS FOR NIKETHAMIDE DOSAGE FORMS Combination Nikethamide . . .. .. .. .. Nikethamide - glucose. . .. .. .. Nikethamide - glucose - ascorbic acid Nikethamide - theophylline - adenosine . . Nikethamide - caffeine - strychnine sulphatef Nikethamide - caffeine - strychnine sulphatef - . . sodium salicylate . . .. .. .. Nikethamide : ephedrine hydrochlorideg . . Dosage form* i, ii, iii iii, iv, v iv i, ii ... 111 i ... 111 Dosage ratio (mg ml-1 or Solvent mg per unit) system 100 to 250 1 125 : 1500 1 8 1 8 200:68:1 1 7 125 : 1500 : 550 100 : 34 : 0.5 65 : 50 : 0.25 - 100: 50: 0.25 : 50 7 9 130:15 1 2 3 4 5 6 Ratio of 100 RF values (approximate) 50a 50:O 60 : 50 50:O:O 60:50:80b 55: 40: 07C*d - 70 : 40: 45e 60: 70: 45 : 0 90: 80: 30: 50 to 70 50:O 50 : 20 67: 17 80 : 70 50: 65 55 : 30 a N-Ethylnicotinamide, 100 R g = 35; b dehydroascorbic acid, 100 R g = 90; C for adenosine determination, develop twice; d adenine, 100 RF = 15; e N-ethylnicotinamide, 100 Rp = 55; f quantitative determination of strychnine by continuous development (1+ hours) with methanol ; and g quantitative determination of ephedrine by development with systems 3 plus 4.* Dosage forms: (i) aqueous solution capable of injection; (ii) gelatine capsule; (iii) tablet (may contain nikethamide as its calcium chloride or thiocyanate salt) ; (iv) effervescent tablet; and (v) caramel sweet.Solvent systems that were found to be suitable for the qualitative examination of particular dosage form combinations are given in Table 11. For nikethamide, system 1 was most widely applicable. The chromatographic behaviour of the compounds was good, with the exception of ephedrine, which produced tailing effects with systems other than 3 and 4. Slight aerial oxidation of ascorbic acid occurred during chromatography and gave rise to a satellite spot (dehydroascorbic acid). In most instances, quantitative assay of each component in nikethamide combination preparations was possible by thin-layer chromatography. Two important exceptions were glucose and ascorbic acid, which were more conveniently determined by established titrimetric methods23 involving oxidation with hexacyanoferrate(II1) or iodine reagents for glucose, and with iodine for ascorbic acid.For the determination of glucose in combinations containing ascorbic acid, interference by the latter was eliminated by retaining it on Amberlite IRA-410 (OH- form) anion-exchange resin (Rohm & Haas) , the glucose being eluted with warm water.720 CARMICHAEL CONCLUSIONS Thin-layer chromatography is a rapid, sensitive and selective method for the identification and quantitative assay of preparations of nikethamide combinations that compares favourably with other methods currently available. While the quantitative determinations described above involve the elution of the active compounds from the adsorbent, preliminary studies have shown that with the possible exception of ephedrine, the ideal behaviour of these compounds during thin-layer chromatography as regards spot symmetry and the absence of tailing effects should permit the application of direct densitometric evaluation of chromato- grams (fluorescence-quenching mode) as a routine procedure.The author thanks Mrs. P. Brunschwiler and Miss R. Matthies for experimental assistance, and Mr. E. Candolfi for the sample of N-ethylnicotinamide. The management of Ciba-Geigy is thanked for permission to publish this paper. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. 17. 18. 19. 20. 21. 22. 23. REFERENCES Hartman, M., and Seiberth, M., US. Patent 1 403 117, 1922; Ciba AG, Basel, Swiss Patent 90 807, Giiven, K.C., and Kanber, S., Eczacilik Biilt., 1969, 11, 170. Bickel, M. H., Pharmac. Rev., 1969, 21, 325. Spengler, H., and Kaelin, A., Centenaire Soc. Suisse Pharm., 1943, 542. Bergmann, F., Analyt. Chem., 1952, 24, 1367. Delaville, J., Ann. Biol. Clin., 1962, 20, 479. Eilhauer, H. D., Fahrig, G., and Krautschick, G., 2. analyb. Chem., 1967, 228, 276. Fuentes-Duchemin, J., and Casassas, E., Analytica Chim. Acta, 1969, 44, 462. Wallen, O., Farmaceutisk Revy, 1947, 46, 233. Deltombe, J., Annls Pharm. Belg., 1953, 8, 59. DOUZOU, P., and LeClerc, A. M., Analytica Chim. Acta, 1956, 12, 239. Vessman, J., and Schill, G., Svensk. Farm. Tidskr., 1962, 66, 601. Prosser, A. R., and Sheppard. A. J., J . Pharm. Sci., 1968, 57, 1004. Macek, K., Hais, I. M., Kopeckf, J., and GaspariC, J., Editors, Journal of Chromatography, Supplementary Volume “Bibliography of Paper and Thin-Layer Chromatography 1961-1965,” Elsevier, Amsterdam, 1968. Guven, K. C., and Pekin, o., Eczacilik Bult., 1966, 8, 115. Haywood, P. E., and Moss, M. S., Analyst, 1968, 93, 737. Baumler, J., Brault, A. L., and Im Obersteg. J., Schweizer. Arch. Tierheilk., 1964, 106, 346. Stahl, E., Editor, “Thin-Layer Chromatography,’’ Second Edition, George Allen & Unwin, London, “Anfurbereagenzien fur Diinnschicht- und Papier-Chromatographie,” E. Merck AG, Darmstadt, 1966. Higuchi, T., and Brochmann-Hanssen, E., Editors, “Pharmaceutical Analysis,” Interscience Vrombaut, R., Thomas, L., and Cooreman, H., Pharm. Tijdschr. Belg., 1969, 46, 197. Shellard, E. J., Editor, “Quantitative Paper and Thin-Layer Chromatography,” Academic Press, Ashworth, M. R. F., “Titrimetric Organic Analysis, Part I1 : Indirect Methods,” Interscience Received January 15th, 1971 Accepted Apvil 26th, 1971 1923. 1969. Publishers, New York, 1961. London, 1968, Chapter 3. Publishers, London, 1965.

ISSN:0003-2654    

DOI:10.1039/AN9719600716

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, October, 1971, Vol. 96, $$I. 721-727 721 The Gas-chromatographic Determination of 2,3,7,8- Tetrachlorodibenzo- p-dioxin in 2,4,5- Trichlorophenoxyacetic Acid (“2,4,5-T”), 2,4,5-T, Ethylhexyl Ester, Formulations of 2,4,5-T Esters and 2,4,5=Trichlorophenol BY D. A. ELVIDGE (Quality Control, Analytical Research, Boots Pure Drug Co. Ltd., Pennyfoot Street, Nottingham) A gas-chromatographic method for the determination of trace amounts of a toxic impurity, 2,3,7,8-tetrachlorodibenzo-~-dioxin, is described. A purified extract of the sample was subjected to gas chromatography on a column containing either 2 per cent. of OV-17 on Diatomite CQ or 1 per cent. of Hi-Eff 8 BP on Gas-Chrom 2 with electron-capture detection. 2,4,6-Tri- chlorophenoxyacetic acid and 2,4,5-trichlorophenol were purified by chromato- graphy of an ether extract of the sample on a column of alumina, followed by shaking with sulphuric acid.For 2,4,5-trichlorophenoxyacetic acid esters and formulations, saponification and chromatography on a Celite - sulphuric acid column, followed by chromatography on a column of alumina, were necessary. Recoveries of 2,3,7,8-tetrachlorodibenzo-~-dioxin ranged from 89 to 98 per cent., and the standard deviation of the method at a level of 0-3 p.p.m. was 0.03 p.p.m. The limit of detection was about 0-05 p.p.m. 2,3,7,8-TETRACHLORODIBENZO-fi-DIOXIN is known to cause chloracne and liver damage at very low levels of concentration in rabbits,l and is believed to possess teratogenic properties. Under certain conditions it is formed during the manufacture of 2,4,5-trichlorophenol and of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) from 2,4,5-trichlorophenol.c’aoNa + c’ac’ 180°C ~ “nIDc’ NaOH CI ‘ CI NaO ’ CI +methanol CI CI One method of producing 2,4,5-trichlorophenol commercially is by hydrolysis of 1,2,4,5- tetrachlorobenzene with methanolic sodium hydroxide solution. If other chlorinated benzenes are present, then other polychlorinated phenols could be formed, which would in turn give rise to chlorinated dioxins, the level of which would be expected to be far lower than that of the 2,3,7,8-tetrachlorodibenzo-~-dioxin itself. In the United States, work at the Bionetics Research Laboratories2 has shown that one sample of 2,4,5-T examined contained about 27 p.p.m. of 2,3,7,8-tetrachlorodibenzo- $-dioxin, whereas the normal level is about 1 p.p.m.or less. During the last few years it has been shown that 2,3,7,8-tetrachlorodibenzo-p-dioxin is a component of the chick oedema factor that causes hydropericardium in chickens.3 The more highly chlorinated dioxin 1,2,3,7,8,9-hexachlorodibenzo-p-dioxin has also been isolated from toxic fats and identified by gas chromatography3 and X-ray ~rystallography.~ Quali- tative identification of the chick oedema factor in oils and fats by gas chromatography after suitable clean-up has been rep~rted,~,~,’ and bio-assay methods have also been used.*s9 How- ever, no methods for the quantitative determination of 2,3,7,8-tetrachlorodenzo-~-dioxin or any of the other individual components of chick oedema factor have been published.This paper describes a gas-chromatographic procedure for the quantitative determination of 2,3,7,8-tetrachlorodibenzo-~-dioxin in 2,4,5-T, 2,4,5-T ethylhexyl ester, formulations of 2,4,5-T esters and 2,4,5-trichlorophenol. Confirmation of the presence of 2,3,7,8-tetrachloro- dibenzo-$-dioxin is obtained by gas chromatography on a second stationary phase. 0 SAC and the author,722 ELVIDGE : THE GAS-CHROMATOGRAPHIC DETERMINATION [AIzalySt, VOl. 96 EXPERIMENTAL CAUTION-Because of the highly toxic nature of 2,3,7,8-tetrachlorodibenzo-p-dioxin extreme care must be taken when handling solutions and materials containing this compound. REAGENTS- Hexane, distilled-Two litres of hexane were distilled through a column, 8 inches x 1 inch, of Fenske helices, 4 mm in diameter, and the first 1800 ml of distillate were collected.The suitability of the distilled material for use in the subsequent procedures was checked as follows: 50 ml were evaporated to 1 ml and ZO-pl aliquots were subjected to gas chromato- graphy on columns containing the stationary phases OV-17 and Hi-Eff 8 BP, under the conditions described later. Hexane that showed peaks with the same retention times as 2,3,7,8-tetrachlorodibenzo-p-dioxin was rejected. The Puriss grade supplied by Koch-Light was found to be satisfactory after distillation. Benzene-Analytical-reagent grade. Diethyl ether-Anaesthetic grade B.P., distilled. Two litres of diethyl ether were distilled through a column, 8 inches x 1 inch, containing Fenske helices, 4mm in diameter, and the first 1800 ml of distillate were collected.Ethanol, 95 per cent.-B.P. grade. Methanol-Analytical-reagent grade. Diethyl ether - hexane (1 + 9 v / v ) and (1 + 4 v/v). Methylene chloride-Analytical-reagent grade. Sulphuric acid, concentrated-Analytical-reagent grade. Sodium sdphate, anhydrous-Analytical-reagent grade. Potassium hydroxide, M. Hydrochloric acid, concentrated-Analytical-reagent grade. Alumina-Chromatographic grade, neutral, Brockmann activity I, 100 to 240 mesh. Camag material (Hopkin and Williams Ltd.) was found to be suitable. Celite 535, acid-washed-Celite 535 (Johns-Manville) was soaked in concentrated hydro- chloric acid overnight, washed with water until neutral and the "fines" were removed by decantation. The Celite was then filtered off, washed with methanol, then with methylene chloride and finally dried overnight at 50 "C.2,3,7,8-Tetrachlorodibenzo-p-dioxin, reference material-This was prepared by the following method: 98.75 g (0.5 mole) of 2,4,5-trichloropheno1 were dissolved in 250 ml of toluene, the solution being stirred and heated on a steam-bath; 33 g (0.5 mole) of potassium hydroxide pellets were added, which dissolved with precipitation of the potassium phenoxide. Water was removed azeotropically and l o g of copper powder (activated by iodinelO) and 1 g of copper(I1) acetate were added. About 150ml of toluene were distilled off, 500ml of dry nitrobenzene were added and distillation was continued until an internal temperature exceeding 210°C was reached. The mixture was gently refluxed in an atmosphere of nitrogen for 18 hours, cooled below 5 "C and filtered under suction.The solid on the filter was washed with nitrobenzene (50 ml) and then with about 75 ml of methanol. After sucking it as dry as possible the solid (80 g) was slurried with 400 ml of N sodium hydroxide solution, filtered, washed with water and air dried. The resultant olive green - brown coloured solid (28 g) was recrystallised from anisole to give 12.5 g of a pinkish white crystalline powder with melting- point 302 to 304 "C (literature value 295 O W ) . The infrared spectrum, as a paraffin mull, was identical with the published spectrum.ll Gas chromatography of the compound on OV-17 under the conditions described gave one major well defined peak and a small amount of impurity with a relative retention time of 0.55.2,3,7,8-Tetrachlorodibenzo-p-dioxin standard solution-Ten milligrams of 2,3,7,8-tetra- chlorodibenzo-p-dioxin, accurately weighed, were dissolved, with warming, in 5 ml of benzene contained in a 100-ml calibrated flask and the solution was adjusted to volume with hexane; 5.0ml of this solution were diluted to 50ml with hexane and a further dilution of 5-0ml of this solution to 50 ml was made with hexane (1.0 ml of the final solution contained 1.0 pg of 2,3,7,8-tetrachlorodibenzo-p-dioxin). APPAFUTUS- water, dried and rinsed with hexane. All glassware was cleaned in chromic - sulphuric acid mixture, washed thoroughly withOctober, 19711 OF 2,3,7,8-TETRACHLORODIBENZO-~-DIOXIN 723 A Perkin-Elmer F11 gas chromatograph, equipped with an electron-capture detector (with a 500-mCi tritium source) and a I-mV Leeds and Northrup Speedomax W recorder, was used. CoZzcmns-(a) A 1.8-m glass column of 3-mm i.d.was packed with 2 per cent. of OV-17 (Supelco Inc.) on 80 to 100-mesh Diatomite CQ (Pye Unicam Ltd.) and conditioned at 250 "C for 48 hours without the detector attached. (b) A 1.8-m glass column of 3-mm i.d. was packed with 1 per cent. of cyclohexane dimethanol succinate (Hi-Eff 8 BP) (Applied Science Laboratories Inc.) on 100 to 120-mesh Gas-Chrom 2 (Applied Science Laboratories Inc.) and conditioned at 230 "C for 48 hours without the detector attached. The following operating conditions were used : detector voltage 2 V, detector temperature 205 "C, recorder chart speed 12 inches hour-l and carrier gas nitrogen (oxygen-free); for the OV-17 column: column temperature 200 "C, injection port temperature 210 "C and carrier gas flow-rate 60ml min-l; and for the Hi-Eff 8 BP column: column temperature 200 "C, injection port temperature 210 "C and carrier gas flow-rate 50 ml min-l. Attenuation settings were selected so that an injection of 20 ng of 2,3,7,8-tetrachloro- dibenzo-p-dioxin produced a peak height of at least 30 per cent.of the full-scale deflection. PROCEDURE FOR PRELIMINARY TREATMENT OF THE SAMPLE FOR 2,4,5-T AND 2,4,5-TRICHLORO- PHENOL- Extractiort--About 1.0 g of sample, accurately weighed, was transferred to a dry 500-ml separating funnel and dissolved in 50 ml of ethanol; 100 ml of water were added, followed by 10 ml of M potassium hydroxide.The alkaline solution was extracted with three 50-ml portions of diethyl ether, the ether extracts were combined and the separating funnel was rinsed with 10 ml of diethyl ether, which were added to the main ether extracts. The combined ether extract was washed with two 20-ml portions of M potassium hydroxide and then with three 25-ml portions of water, after which it was dried by shaking it with 10 g of anhydrous sodium sulphate and filtered through a cotton-wool plug into a 500-ml round-bottomed flask; the separating funnel and cotton-wool were rinsed with two 15-ml portions of diethyl ether and the rinsings added to the flask. The ether was removed in a rotary evaporator at 40 "C, 5 ml of methanol were added to the residue and the solution was again evaporated to dryness.The residue was then dissolved in 2 ml of hexane. AZumina column chromatography-A column was prepared in a glass tube, 150 x 10 mm, which was fitted with a tap and plugged with a small wad of cotton-wool, by filling the tube with hexane, adding 5 g of alumina and allowing the alumina to settle while gently tapping the tube to release any air bubbles. The hexane solution from the extraction step was transferred with a Pasteur pipette to the top of the alumina column. The 500-ml flask was rinsed with three successive 2-ml portions of hexane, and the washings were added to the alumina column. The column was washed with 50 ml of hexane and then with 50 ml of diethyl ether - hexane (1 + 9 v/v). The 2,3,7,8-tetrachlorodibenzo-p-dioxin was then eluted with 50 ml of diethyl ether - hexane (1 + 4 v/v), the eluate being collected in a 100-ml flask and the solvent removed in a rotary evaporator at 40 "C.The residue was dissolved in 1 ml of hexane, which was subjected to treatment with sulphuric acid as follows. Szdphuric acid treatment-The hexane solution was transferred with a Pasteur pipette to a 10-ml stoppered cylinder and the flask was rinsed with two 1-ml portions of hexane, the washings being added to the cylinder; 3 ml of sulphuric acid were added and the cylinder was shaken rapidly for 3 minutes on a mechanical shaker. After allowing the layers to separate the hexane layer was transferred to a second cylinder containing 3 ml of sulphuric acid, the first cylinder being washed with two 0-5-ml portions of hexane and the washings added to the second cylinder.After shaking the contents of this cylinder for 3 minutes the layers were allowed to separate and the hexane layer was transferred to a 100-ml flask via a funnel containing 2 g of sodium hydrogen carbonate supported on a plug of cotton-wool. The sulphuric acid layer in the cylinder was washed with three 2-ml portions of hexane, each washing being transferred via the sodium hydrogen carbonate layer to the 100-ml flask. The solvent was removed in a rotary evaporator at 40 "C and the final residue dissolved in 1.0 ml of hexane.724 ELVIDGE : THE GAS-CHROMATOGRAPHIC DETERMINATION [Analyst, Vol. 96 PROCEDURE FOR PRELIMINARY TREATMENT OF THE SAMPLE FOR 2,4,5-T ESTERS AND 2,4,5-T ESTER FORMULATIONS- SaponiJication and extraction-A sufficient amount of 2,4,5-T ester or ester formulation to contain the equivalent of about 1.Og of 2,4,5-T was accurately weighed and transferred to a 250-ml flask; it was then dissolved in 50 ml of ethanol, 5 ml of M potassium hydroxide (or sufficient to provide a 1-ml excess over the amount required for saponification) and 2 ml of water were added, and the mixture was refluxed on a steam-bath for 10 minutes.After cooling, the solution was transferred with 50 ml of water to a separating funnel containing 50 ml of water. The extraction procedure described above was then followed, commencing a t “The alkaline solution was extracted . . .,” except that the final residue was dissolved in 10 ml of hexane. Celite - sulphuric acid colunzn chromatography-A Celite - sulphuric acid column was pre- pared by intimately mixing 6 g of Celite with 4 ml of concentrated sulpliuric acid and packing this mixture, in portions, into a glass column, 200 x 18 mm, fitted with a tap and plugged with a small wad of cotton-wool, tamping each portion firmly with a glass rod.The hexane solution from the saponification and extraction step was transferred to the Celite column and the flask was rinsed with three 5-ml portions of hexane, which were added to the column. A further 50 ml of hexane were added to the column and the total eluate was collected in a 250-ml flask and evaporated to dryness in a rotary evaporator at 40 “C. The residue was dissolved in 2 ml of hexane and treated as described below. Alumina column chromatography-The hexane solution was chromatographed on a column of alumina as described for 2,4,5-T and 2,4,5-trichlorophenol. The residue from the diethyl ether - hexane (1 + 4 v/v) fraction was dissolved in 1.0 ml of hexane.PROCEDURE FOR THE GAS-CHROMATOGRAPHIC DETERMINATION OF 2,3,7,8-TETRACHLORODI- BENZO-$-DIOXIN- Aliquots of 20 p1 of the 2,3,7,8-tetrachlorodibenzo-~-dioxin standard solution and the sample solution from the above final preliminary treatments were chromatographed on the OV-17 and Hi-Eff 8 BP columns under the conditions already described. On either column, the 2,3,7,8-tetrachlorodibenzo-fi-dioxin was eluted after about 15 minutes; its concentration in the samples was calculated from the following formula- hE 1.0 2,3,7,8-Tetrachlorodibenzo-~-dioxin in sample, p.p.m.= - hs w where hE is the height of its peak in the sample chromatogram, lzs is the height of its peak in the standard chromatogram and W is the weight, in grams, of the sample taken. RESULTS AND DISCUSSION The use of the electron-capture detector had the advantage of high sensitivity, and thus avoided the necessity of handling relatively large amounts of sample. However, such a choice required a vigorous clean-up procedure to prevent gross contamination of the detector by other chlorinated compounds present and to remove interfering substances with the same retention times as 2,3,7,8-tetrachlorodibenzo-fi-dioxin. No single clean-up procedure was found to be adequate, but while a combination of extraction, alumina column chromatography and sulphuric acid treatment reduced the background to a satisfactory level for 2,4,5-T and 2,4,5-trichlorophenol, for 2,4,5-T esters and their formulations a preliminary saponification, extraction, Celite - sulphuric acid column chromatography and alumina column chromato- graphy were required. It was necessary to carry out the Celite - sulphuric acid column treat- ment of the unsaponifiable fraction from the esters and their formulations before chromato- graphy on alumina to remove compounds that otherwise would have been detrimental to the latter procedure.The same clean-up technique could probably have been applied to 2,4,5-T and 2,4,5-trichlorophenol with equal success, but as this had no apparent advantage the simpler procedure of shaking with sulphuric acid after alumina column chromatography was adopted for these substances. Similar clean-up procedures have been used for the detection of chick oedema fact or.p6 9 Most commercial material contained a significant amount of an impurity that had the same retention time as that of 2,3,7,8-tetrachlorodibenzo-p-dioxin and efforts to remove it were mainly unsuccessful. Hexane of satisfactory purity proved difficult to obtain.October, 19711 OF 2,3,7,8-TETRACHLORODIBENZO-$-DIOXIN 725 Treatment with fuming sulphuric acid or potassium permanganate, or refluxing with sodium or alkaline silver nitrate showed little improvement. Eventually distillation of carefully selected hexane, initially of low impurity content, provided a satisfactory reagent. RECOVERY OF 2,3,7,8-TETRACHLORODIBENZO-fi-DIOXIN- Amounts of 0.5 and 1.0 ml of the 2,3,7,8-tetrachlorodibenzo-p-dioxin standard solution were added to 1.0-g portions of 2,4,5-T that were free from the impurity (less than 0.05 p.p.m.), dissolved in 50ml of ethanol, and the samples were analysed by the method described.Recoveries of 98 and 95 per cent., respectively, were obtained. Also, 1.0 ml of the 2,3,7,8-tetra- chlorodibenzo-+dioxin standard solution was added to 1.5 g of 2,4,5-T ethylhexyl ester dissolved in 50ml of ethanol and the mixture was analysed by the method described. A recovery of 89 per cent. was obtained. PRECISION- The precision of the gas-chromatographic procedure was determined by replicate injec- tions of 20 pg of the standard 2,3,7,8-tetrachlorodibenzo-$-dioxin solution.The standard deviation of the 2,3,7,8-tetrachlorodibenzo-~-dioxin injected was 0.6 ng. The precision of the whole procedure was determined by analysing five 1.0-g portions of one of the 2,4,5-T samples, and the results ranged from 0.25 to 0.31 p.p.m., with a mean value of 0.27 p.p.m. and a standard deviation of 0.03 p.p.m. LINEARITY OF RESPONSE OF THE ELECTRON-CAPTURE DETECTOR- Portions of 5-0, 10.0, 15.0 and 20.0 p1 of the 2,3,7,8-tetrachlorodibenzo-~-dioxin standard solution were injected on to the OV-17 column. The graph of peak height versus nanograms of 2,3,7,8-tetrachlorodibenzo-fi-dioxin was rectilinear over the range examined and passed through the origin. LIMITS OF DETECTION- Under the conditions of the method the lowest level of 2,3,7,8-tetrachlorodibenzo-fi-dioxin detectable (twice the noise level) was found to be 0.05ng per pl of final solution.For a sample concentration equivalent to 1.0 g ml-l, the limit of detection was, therefore, 0.05 p.p.m. Table I shows the results obtained on several samples of 2,4,5-T from four manufacturers. The determinations were carried out in duplicate on separately weighed amounts of sample. TABLE I 2,3,7,8-TETRACHLORODIBENZO-fi-DIOXIN CONTENT OF 2,4,5-T 2,3,7,8-Tetrachlorodibenzo-fi-dioxin, p.p.m. r t Sample OV-17 Hi-Eff 8 BP A A, 0.30, 0.27 0.27 0.17, 0.12 0-16 0.60 Aa 0.60, 0.42 0.29, 0.31 - 0.27, 0.28 0.33 C D <0*05, <0-05 < 0.05 2 TABLE I1 2,3,7,8-TETRACHLORODIBENZO-~-DIOXIN CONTENT OF 2,4,5-T ESTER AND FORMULATIONS 2,3,7,8-Tetrachlorodibenzo-fi-dioxin, p.p.m.f A t Sample OV- 17 Hi-Eff 8 BP 2,4,5-T ethylhexyl estey- E 0.26, 0.28 F 0.27, 0.24 G 0.19, 0.17 H 0.09, 0.13 2,4,5-T ester formulation- 0.20 0.22 0.16 0.09726 ELVIDGE : THE GAS-CHROMATOGRAPHIC DETERMINATION [A uta&St, VOl. 96 All of the samples contained less than 0.5 p.p.m. of 2,3,7,8-tetrachlorodbenzo-p-diorrin. Sample B contained an impurity that interfered in the determination with Hi-Eff 8 BP as the stationary phase. This impurity was not present in any other sample. I I I I I 75 60 45 30 15 Time/m inu tes Fig. 1. Chromatogram of an extract of 2,4,6-T on 2 per cent. OV-17 on Diatomite CQ A typical chromatogram of an extract from 2,4,5-T containing about 0-5p.p.m. of 2,3,7,8-tetrachlorodibenzo-$-dioxin, with OV-17 as the stationary phase, is shown in Fig.1. Table I1 shows the results on samples of 2,4,5-T ethylhexyl ester and ester formulations. Again the contents of 2,3,7,8-tetrachlorodibenzo-$-dioxin were less than 0.5 p.p.m. A typical chromatogram of an extract from 2,4,5-T ethylhexyl ester containing about 0.2 p.p.m. of 2,3,7,8-tetrachlorodibenzo-fi-dioxin, with Hi-Eff 8 BP as the stationary phase, is shown in Fig. 2. Time/m inu tes Fig. 2. Chromatogram of an extract of 2,4,5-T ethylhexyl ester on 1 per cent. Hi-Eff 8 BP on Gas-Chrom 2 Finally the results on two samples of 2,4,5-trichloropheno1 are shown in Table 111. Neither sample contained appreciable amounts of 2,3,7,8-tetrachlorodibenzo-~-dioxin. In nearly all of the samples of 2,4,5-T and 2,4,5-T esters a large impurity peak was observed with a relative retention time of about 3.5 on OV-17 and 4.2 on Hi-Eff 8 BP (2,3,7,8-tetrachlorodibenzo-~-dioxin = 1.00).This impurity was not removed even by furtherOctober, 19711 OF 2,3,7,8-TETFUCHLORODIBENZO-@-DIOXIN 727 treatment with sulphuric acid, but as it did not interfere in the determination its elimination was of no consequence. Evidence from its mass spectrum suggests that the impurity is not a polychlorinated dioxin but possibly bis-(2,4,5-trichlorophenoxy)methane. All chromatograms were run for 1Q to 2 hours before injection of further samples to clear the columns of long-running impurities. TABLE I11 2,3,7,8-TETRACHLORODIBENZO-~-DIOXIN CONTENT OF 2,4,5-TRICHLOROPHENOL 2,3,7,8-Tetrachlorodibenzo-p-dioxin, p.p.m. , Sample OV- 17 Hi-Eff 8 BPI 0.32, 0-32 0.25 <0.06, <0.05 < 0.05 J K CONCLUSIONS The highly toxic impurity 2,3,7,8-tetrachlorodibenzo-@-dioxin can be detected, and its concentration determined, in 2,4,5-T, 2,4,5-T esters, formulations and 2,4,5-trichlorophenol by means of gas chromatography with either OV-17 or Hi-Eff 8 BP as stationary phase after suitable clean-up procedures have been applied.The lower limit of detection, taking the equivalent of 1 g of 2,4,5-T in the initial sample, is about 0.05 p.p.m. The standard deviation of the procedure at a level of 0.3 p.p.m. is 0-03 p.p.m. The author thanks Mr. W. H. Stephenson for his interest and encouragement, and Mr. R. Banks for preparing the 2,3,7,8-tetrach3orodibenzo-~-dioxin. 1. 2. 3. 4. 5. 6 . 7 . 8. 9. 10. 11. REFERENCES Bauer, H., Schultz, K. H., and Spiegelberg, U., Arch. Gewerbepath. Gewerbehyg., 1961, 18, 538. Press release from the United States Department of Agriculture, February 6th, 1970; Nature, Higginbotham, G. R., Huang, A., Firestone, D., Verrett, J., Ress, J., and Campbell, A. D., Nature, Cantrell, J. S., Webb, N. C., and Mabis, A. J., Chem. Engng News, 1967, 45, 10. Firestone, D., Ibrahaim, W., and Horwitz, W., J . Ass. 08. Agric. Chern., 1963, 46, 384. Huang, A., Firestone, D., and Campbell, A. D., Ibid., 1967, 50, 16. Higginbotham, G. R., Firestone, D., Chavez, L., and Campbell, A. D., Ibid., 1967, 50, 874. Douglass, C. D., and Flick, D. F., Ibid., 1961, 44, 449. Flick, D. F., Gallo, L., Winbush, J., Douglass, C. D., and Friedman, L., Ibid., 1962, 45, 231. Vogel, A. I., “A Textbook of Practical Organic Chemistry, including Qualitative Analysis,” Long- Tomita, M., Ueda, S., and Narisada, M., J . Pharm. SOC. Japan, 1959, 79, 186. Load., 1970, 226, 309. Lond., 1968, 220, 702. mans, Green & Co. Ltd., London, 1948, p. 185. Received December l l t h , 1970 Accepted April 15th, 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600721

出版商:RSC    

年代:1971

数据来源: RSC