摘要:

THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYEDITORIAL ADVISORY BOARD"Chairman: H. J. Cluley (Wernbley)"L. S. Bark (Salford)R. Belcher (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)L. R. P. Butler (South Africa)E. A. M. F. Dahrnen (The Netherlands)A. C. Docherty (Billingham)D. Dyrssen (Sweden)J. Hoste (Belgium)H. M. N. H. Irving (feeds)H. Kaiser (Germany)M. T. Kelley (U.S.A.)W. Kemula (Poland)"W. T. Elwell (Birmingham)"J. A. Hunter (Edinburgh)"G. F. Kirkbright (London)G. W. C. Milner (Harwell)G . H. Morrison (U.S.A.)"J. M. Ottaway (Glasgow)"G. E. Penketh (Billingham)"T. B. Pierce (Harwell)E. Pungor (Hungary)D. I. Rees (London)"R. Sawyer (London)P. H. Scholes (Sheffield)"W. H.C. Shaw (Greenford)S. Siggia (U.S.A.)A. A. Smales, O.B.E. (Harwell)A. Walsh (Australia)T. S. West (Aberdeen)A. L. Wilson (Medmenham)P. Zuman (U.S.A.)*A. Towns hend (Birmingham)"Members of the Board serving on The Analyst Publications CommitteeREGIONAL ADVISORY EDITORSDr. J. Aggett, Department of Chemistry, Universily of Auckland, Private Bag, Auckland, NEWProfessor G. Ghersini, Laboratori CISE, Casella Postale 3986, 201 00 Milano, ITALY.Professor L. Gierst, Universit6 Libre de Bruxelles, Facult6 des Sciences, Avenue F.-D. Roosevelt 50,Professor R. Herrmann, Abteilung fur Med. Physik., 63 Giessen, Schlangenzahl 29, GERMANY.Professor W. E. A. McBryde, Dean of Faculty of Science, University of Waterloo,Waterloo, Ontario,Dr.W . Wayne Meinke, KMS Fusion Inc., 3941 Research Park Drive, P.O. Box 1567, Ann Arbor,Dr. I. Rubeika, Geological Survey of Czechoslovakia, Kostelni 26, Praha 7, CZECHOSLOVAKIA.Dr. J. Rb%ka, Chemistry Department A, Technical University of Denmark, 2800 Lyngby, DENMARK.Professor K. Saito, Department of Chemistry, Tohoku University, Sendai, JAPAN.Dr. A. Strasheim, National Physical Research Laboratory, P.O. Box 395, Pretoria, SOUTH AFRICA.ZEALAND.Bruxelles, BELGlU M.CANADA.Mich. 481 06, U.S.A.Published by The Chemical SocietyEditorial: The Director of Publications, The Chemical Society, Burlington House,London, WIV OBN. Telephone 01 -734 9864. Telex No. 268001.Advertisements: J. Arthur Cook, 9 Lloyd Square, London, WC1 X 9BA. Telephone 01 -837 631 5.Subscriptions (non-members): The Chemical Society Publications Sales Office, Blackhorse Road,Letchworth, Herts., SG6 1 HN.Volume 101 No 1205@ The Chemical Society 1976August 197

ISSN:0003-2654    

DOI:10.1039/AN97601FX029

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

ANALAO 101 (1205) 593-688 (1976)ISSN 0003-2654August 197659360161 161 6622634639644652657661666672678683685THE ANALYSTTHE ANALYTICALREVIEW PAPERFibre Identification andORIGINAL PAPERSJOURNAL OF THE CHEMICAL SOCIETYCONTENTSAnalysis of Fibre Blends-H. M. AppleyardColorimetric Determination of Thiamine Propyl Disulphide and ThiamineDisulphide in Their Respective Pharmaceutical Preparations-M. V. SivaramaKrishnan, S. N. Mahajan and G. Ramana RaoDetermination of Nortriptyline i n Sugar-coated Tablets by High-performanceLiquid Chromatography-J. R. Salmon and P. R. WoodSpecific Determination of Testosterone in Oily Injections-Emil Fahmy, DawoudA. Yassa and Nagi WahbaComputer-aided Identification of Powdered Vegetable Drugs-Georgina H.Jolliffe and G.0. JolliffePerformance Characteristics for the Spectrophotometric Determination ofTotal Iron i n Freshwater Using Hydrochloric Acid-W. Davison and E. RiggDetermination of Trace Amounts of Chlorophenols by Gas - Liquid Chromato-graphy-D. S. Farrington and J. W. MundayGas-chromatographic Determination of Selenium i n Pure Elemental Arsenicand Arsenic( 111) Oxide w i t h 4-Nitro-1.2-diaminobenzene-M. Akiba, Y.Shimoishi and K. TBeiDetermination o f Copper by Anodic-stripping Pulse Voltammetry in ThiocyanateMedia-A. W. Mann and R. L. DeutscherDetermination of Nanomole Amounts of Aluminium by Use of a Fluoridelon-selective Electrode-Nj. Radi6Separation of Rhodium and Iridium Using Silicone Rubber Foam Treated w i t hTri-n-octylamine-A.Baghai and H. J. M. BowenDetermination o f Nanogram Amounts of Praseodymium and Terbium by Meansof Candoluminescence Emission-R. Belcher, K. P. Ranjitkar and Alan Towns-hendAtomic-absorption Determination o f Bismuth in Complex Nickel-base Alloysby Generation of its Covalent Hydride-J. E. DrinkwaterUse of Sodium Borohydride for Cold-vapour Atomic-absorption Determinationo f Trace Amounts of Inorganic Mercury-R. C. RooneyCOM M U NlCATlONDetermination o f Volatile Elements by Carbon Furnace Atomic-emissionBook ReviewsSummaries o f Papers in this lssue-Pages iv, v. viii, xSpectrometry-J. M. Ottaway and R. C. HuttonPrinted by Heffers Printers Ltd, Cambridge, EnglandEntered as Second Class at New York, USA, Post OfficSUMMARIES OF PAPERS I N THIS ISSUE August, 1976Atomic- absorption Determination of Bismuth in Complex Nickel- baseAlloys by Generation of its Covalent Hydriderequirement has arisen in the aero-engine industry for a method that iscapable of determining trace amounts of bismuth in complex nickel-basealloys. This paper deals with the determination of bismuth a t the 0-1 p.p.m.level in nickel-base alloys after dissolution of the alloy followed by generationof bismuth hydride and its subsequent measurement by atomic-absorptionspectrophotometry. Interference effects, due to the complex matrix, have beenovercome and a procedure has been developed that is rapid and sensitive a tthese very low levels.The method is applicable to a wide range of nickel-basealloys commonly used in the aero-engine industry.J.E. DRINKWATERRolls-Royce (1871) Limited, Bristol Engine Division, P.O. Box 3, Filton, Bristol,ns12 7qE.Analyst, 1976, 101, 672-677.Use of Sodium Borohydride for the Cold-vapour Atomic-absorptionDetermination of Trace Amounts of Inorganic MercuryMercury in acidic solution is reduced to the metallic state with sodium boro-hydride. The reduction is rapid, gives enhanced sensitivity over that obtainedby use of more conventional reductants, and is conveniently carried out byusing the same apparatus as is used for the determination of arsenic, selenium,antimony, etc., by the hydride generation technique.R. C. ROONEYRooney R: Ward Ltd., Blackwater Station Estate, Camberley, Surrey.Analyst, 1976, 101, 678-682.Determination of Volatile Elements by Carbon FurnaceAtomic-emission SpectrometryCommunicationJ. M. OTTAWAY and R. C. HUTTONDepartment of Pure and Applied Chemistry, University of Strathclydc, CathedralStreet, Glasgow, G1 1SL.Analyst, 1976, 101, 683-684

ISSN:0003-2654    

DOI:10.1039/AN97601BX031

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

iv SUMMARIES OF PAPERS I N THIS ISSUE August, 1976Summaries of Papers in this IssueFibre Identification and Analysis of Fibre BlendsA ReviewSummary of ContentsIntroductionIdentification of Natural FibresAnimal FibresVegetable FibresSilkIdentification of Man-made FibresQuantitative AnalysisReprints of this Review paper can be obtained from The Chemical Society,Publications Sales Officer, Blackhorse Road, Letchworth, Herts., SG6 lHN, atL1 per copy (with a 25% discount for six or more copies), post free.A remittance for the correct amount, made out to The Chemical Society,should accompany every order; these reprints are not available throughTrade Agents.H. M. APPLEYARDIntcrnational Wool Secretariat, Technical Centre, Valley Drive, Ilkley, Yorkshire,LS29 8PB.Analyst, 1976, 101, 593-600.Colorimetric Determination of Thiamine Propyl Disulphide andThiamine Disulphide in Their Respective Pharmaceutical PreparationsA rapid and sensitive colorimetric method for the determination of thiaminepropyl disulphide and thiamine disulphide, as pure drugs and in their respec-tive dosage forms, has been developed, .which is based on a colour reactionwith 2,6-dichloro-p-benzoquinone-4-chlorimine in the presence of methanolichydrobromic acid in a chloroform medium.The colours produced fromthiamine propyl disulphide and thiamine disulphide have absorbance maximaa t 445 and 460 nm, respectively. The concentration range for Beer’s law com-pliance is 0-5.75 p g ml-l for thiamine propyl disulphide and 0-12 p g ml-1for thiamine disulphide.The method can be applied to the analytical controlof thiamine propyl disulphide and thiamine disulphide in dosage forms withsimple matrices, but when other compounds carrying an amino group arepresent a separation is necessary as the proposed colour reaction is a generalone for all such compounds. The removal of riboflavine and analgin, presentin some of the currently available commercial formulations of thiamine propyldisulphide and thiamine disulphide, was found to be necessary and is described.M. V. SIVARAMA KRISHNAN, S. N. MAHAJAN and G. RAMANA RAOQuality Control Laboratories, Indian Drugs and Pharmaceuticals Limited, Hydera-bad-500037, India.Analyst, 1976, 101, 601-610.Determination of Nortriptyline in Sugar-coated Tablets byHigh-performance Liquid ChromatographyA method is described for the determination of nortriptyline in sugar-coatedtablets.Nortriptyline is extracted into dilute aqueous hydrochloric acidcontaining triflupromazine hydrochloride as an internal standard. Theacidic extract is then chromatographed by a reversed-phase separation usinga basic solvent system in conjunction with a phenyl Corasil column.The main advantages of the method are the simple extraction, the over-allspeed of analysis and the degree of precision obtained. The method hasbeen used to monitor the stability of experimental formulations and couldreadily be modified for the assay of single tablets.J. R. SALMON and P. R. WOODE. R. Squibb and Sons Limited, Reeds Lane, Moreton, Merseyside.Analyst, 1976, 101, 611-615August, 1976 SUMMARIES OF PAPERS I N THIS ISSUESpecific Determination of Testosterone in Oily InjectionsMulti-stage extraction from a solution of testosterone oily injection inheptane with 85% ethanol yields a hormone extract in which testosteronecan be specifically determined by applying the sulphuric acid - picric acidreaction. A parallel analysis of standard samples, which are dissolved indifferent oily media, and of identical blank media shows that the oily residuesand pharmaceutical adjuvants do not interfere a t a wavelength of 640 nmand that the recoveries, calculated by comparison with a standard or directlyfrom the value of for the ester concerned, are within the range98.4-101.6y0.Testosterone is also determined with similarly high accuracy in complexinjections containing oestradiol or progesterone although these hormonespass quantitatively into the hormone extract.The possibility of determining both testosterone and either progesteroneor oestradiol in the same extract.EMIL FAHMY, DAWOUD A.YASSAResearch Department, Socikti: Misr pour 1'Industric Pharmaceutique, 92 El MatariaStreet, Post El Zeitoun, Cairo, Egypt.and NAG1 WAHBABiochemistry Department, Faculty of Medicine, A4in Shams University, Cairo, Egypt.Analyst, 1976, 101, 616-621.The precision, for assays of 25 mg ml-l solutions, is & 1%.Computer - aided Identification of Powdered Vegetable DrugsThe identification of 174 powdered crude drugs of organised structure isaided by computer analysis of the recorded observations of eleven simplehistological characters, thus enabling a relatively inexperienced microscopistto reach a positive conclusion with comparative ease.GEORGINA H.JOLLIFFE and G. 0. JOLLIFFEDepartment of Pharmacy, Chelsea College, University of London, Manresa Road,London, SW3 6LX.Analyst, 1976, 101, 622-633.Performance Characteristics for the SpectrophotometricDetermination of Total Iron in Freshwater UsingHydrochloric AcidA method is described for the determination of total iron in freshwater usinga wet-oxidation procedure followed by spectropliotometric measurement in5.93 f 0.07 M hydrochloric acid.It has been used successfully in routine analysis for several years andrequires only standard equipment and inexpensive reagents. The criterionof detection is 0.009 mgl-1 for the 95% confidence level. The relativestandard deviation a t 0.2 mg 1-' is 4.476 and a t 3 mg 1-1 is 1.5% for 19 degreesof freedom. Comprehensive interference tests show that the method isadequately selective for use with freshwater. Copper is the only substanceto interfere significantly and below 0.1 mg I-' it has a negligible effect.W. DAVISON and E. RIGGFreshwater Biological Association, Windcrmere Laboratory, hmblcside, Cumbria,LA22 OLP.Analyst, 1976, 101, 634-638.

ISSN:0003-2654    

DOI:10.1039/AN97601FP061

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

...Vlll SUMMARIES OF PAPERS I N THIS ISSUE August, 1976Determination of Trace Amounts of Chlorophenols byGas - Liquid ChromatographySolutions of trichlorophenols, tetrachlorophenols and pentachlorophenol aremade to react with 2,4-dinitro- l-fluorobenzene in the presence of pyridineas a catalyst. The derivatives are suitable for determination by gas - liquidchromatography using an electron-capt ure detector. The method has beenapplied to the determination of chlorophenols in chicken flesh.D. S. FARRINGTON and J. W. MUNDAYDepartment of Industry, Laboratory of the Government Chemist, Cornwall House,Stamford Street, London, SE1 9KQ.Analyst, 1976, 101, 639-643.Gas - chromatographic Determination of Selenium in PureElemental Arsenic and Arsenic(II1) oxide with4-Nitro- 1,2- diaminobenzeneA simple and practical method for the determination of very small amountsof selenium in elemental arsenic and arsenic(II1) oxide is described.A mixtureof 20 ml of fuming nitric acid and 5 in1 of orthophosphoric acid dissolvesup to 1 g of the arsenic sample and also oxidises the selenium quantitativelyto the quadrivalent state. As the reaction of selenium(1V) with 4-nitro-1,2-diaminobenzene is scarcely interfered with by other elements, chemicalseparation from the matrix element is not necessary. The 5-nitropiaselenolformed in the reaction is extracted into toluene and determined by means ofa gas chromatograph equipped with a sensitive electron-capture detector.No loss of selenium is observed following this treatment.M. AKIBA, Y.SHIMOISHI and K. TOE1Department of Chemistry, Faculty of Science, Okayama University, Tsushima,Okayama-shi, Japan.,4nalyst, 1976, 101, 644-651.Determination of Copper by Anodic- stripping PulseThe determination of copper in natural waters, particularly those containinghigh concentrations of halide ions, by anodic-stripping voltammetry (ASV) isfacilitated by the addition of thiocyanate; a solution with a thiocyanate con-centration of 1.0 x M permits copper to be determined in the presenceof lead. The copper peak, at this thiocyanate concentration, occurs a tapproximately - 0.20 V and is resolved clearly from the steeply rising portionof the ASV curve caused by the oxidation of mercury. The height of thecopper peak is shown to be proportional to the copper concentration for concen-trations in the range 3-100 pg1-l.A.W. MANN and R. L. DEUTSCHERCSIRO, Division of iMineralogy, Private Mailbag, P.O. Wembley, WesternAustralia, 6014.Analyst, 1976, 101, 652-656.Voltammetry in Thiocyanate MediAugust, 1976 THE ANALYST ixReprints of Review PapersReprints of the following Review Papers published in The Analyst since 1967 are available fromthe Publications Sales Officer, The Chemical Society, Blackhorse Road, Letchworth, Herts., SG61HN (not through Trade Agents).The price per reprint is k1; orders for six or more reprints of the same or different Reviewsare subject to a discount of 26%. The appropriate remittance, made out to The ChemicalSociety, should accompany any order.“Activation Analysis,” by R.F. Coleman and T. B. Pierce (January, 1967).“Techniques in Gas Chromatography. Part I. Choice of Solid Supports,” by F. J. Palframan“Heterocyclic Azo Dyestuffs in Analytical Chemistry,” by R. G. Anderson and G. Nickless“Determination of Residues of Organophosphorus Pesticides in Food,’’ by D. C. Abbott and“Radioactive Tracer Methods in Inorganic Trace Analysis: Recent Advances, ” by J . W.“Gamma-activation Analysis,” by C. A. Baker (October, 1967).“Precipitation from Homogeneous Solution,” by P. F. S. Cartwright, E. J . Newman and“Industrial Gas Analysis,” by (the late) H. N. Wilson and G. M. S. Duff (December, 1967).“The Application of Atomic-absorption Spectrophotometry to the Analysis of Iron and“Inorganic Ion Exchange in Organic and Aqueous - Organic Solvents,” by G.J. Moody and“Radiometric Methods for the Determination of Fluorine,” by J. K. Foreman (June, 1969).“Techniques in Gas Chromatography. Part 11. Developments in the van Deemter RateTheory of Column Performance,” by E. A. Walker and J. F. Palframan (August, 1969).“Techniques in Gas Chromatography. Part 111. Choice of Detectors,” by T. A. Gough andE. A. Walker (January, 1970).“Laser Raman Spectroscopy,” by P. J. Hendra and C. J . Vear (April, 1970).“Ion-selective Membrane Electrodes,” by Ern0 Pungor and Kl&ra T6th (July, 1970).“X-ray Fluorescence Analysis,” by K. G. Carr-Brion and I<. W. Payne (December, 1970).“Mass Spectrometry for the Analysis of Organic Compounds,” by A.E. Williams and H. E“The Application of Non-flame Atom Cells in Atomic-absorption and Atomic-fluorescence“Liquid Scintillation Counting as an Analytical Tool,” by J. A. B. Gibson and A. E. Lally“The Determination of Some 1,4-Benzodiazepines and Their Metabolites in Body Fluids, ”“Atomic-fluorescence Spectrometry as an Analytical Technique, ” by R. F. Browner(October, 1974).“The Use of Precipitate Based Silicone Rubber Ion-selective Electrodes and Silicone RubberBased Graphite Voltammetric Electrodes in Continuous Analysis,’’ by 2s. FCher,G. Nagy, K. T6th and E. Pungor (November, 1974).“The Examination of Meat Products with Special Reference to the Assessment of the MeatContent,” by D. Pearson (February, 1975).“Chemiluminescence in Gas Analysis and Flame-emission Spectrophotometry, ” by J .H.Glover (July, 1975).“The Analytical Role of Ion-selective and Gas-sensing Electrodes in Enzymology, ” byG. J. Moody and J. D. R. Thomas (September, 1975).“Thiazolylazo Dyes and Their Applications in Analytical Chemistry,” by Havard R. Hovind(November, 1975).“Sample Preparation in the Micro-determination of Organic Compounds in Plasma or Urine, ”by Eric Reid (January, 1976).“Recent Advances in the Ring Oven Technique,” by Herbert Weisz (March, 1976).“The Radioimmunoassay of Drugs,” by J. Landon and A. C. Moffat (April, 1976).“Analysis and Assay of Polyene Antifungal Antibiotics,” by A. H. Thomas (May, 1976).“Properties and Uses of the Colorimetric Reagents 2-Nitroso-5-dimethylaminophenol and2-Nitroso-5-diethylaminophenol for Cobalt,” by Kyoji T6ei and Shoji Motomizu (July,1976).and E.A. Walker (February, 1967).(April, 1967).H. Egan (August, 1967).McMillan (September, 1967).D. W. Wilson (November, 1967).Steel,” by P. H. Scholes (April, 1968).J. D. R. Thomas (September, 1968).Stagg (January, 1971).Spectroscopy,” by G. F. Kirkbright (September, 1971).(October, 197 1).by J . M. Clifford and W. Franklin Smyth (May, 1974).“Fibre Identification and Analysis of Fibre Blends,” by H. M. Appleyard (August, 1976)X SUMMARIES OF PAPERS I N THIS ISSUEDetermination of Nanomole Amounts of Aluminium by Use ofa Fluoride Ion-selective ElectrodeInvestigation of the rate and mechanism of the reaction between alu-minium(II1) ions and fluoride ions in buffered aqueous solution showed thepossibility of determining nanomole amounts of aluminium.Potential - timecurves recorded during the reaction, using a lanthanum fluoride electrode inconjunction with a reference electrode, constitute the primary data in thisstudy. The initial rates of decrease of the concentration of free fluoride ionwere calculated and shown to be proportional to the amount of aluminiumin solution. A procedure is presented, based on these observations, thatprovides for the determination of aluminium in the range 8-300 nmol.Nj. RADICDepartment of Analytical Chemistry, Faculty of Chemical Technology, Universityof Split, Split, Yugoslavia.Analyst, 1976, 101, 657-660.August, 1976Separation of Rhodium and Iridium Using Silicone Rubber FoamTreated with Tri -n-octylamineMilligram amounts of rhodium and iridium can be separated from solutionsin hydrochloric acid containing free chlorine by batch or column absorptionof iridium on to silicone rubber foam loaded with tri-n-octylamine. Undersuitable conditions, more than 99% of the rhodium remains in the aqueousphase while 98.5 f 0.9% of the iridium is retained by the foam.The iridiumcan be recovered quantitatively from the foam by elution with a small amountof ethanol, and much of the tri-n-octylamine can also be recovered, as canthe foam support.A. BAGHAI and H. J. M. BOWENDepartment of Chemistry, IJniversity of Reading, Whiteknights, Reading, RG6 2AD.Analyst, 1976, 101, 661-665.Determination of Nanogram Amounts of Praseodymium and Terbiumby Means of Candoluminescence EmissionPraseodymium and terbium give rise to intense red and green candolumines-cence, respectively, in a calcium oxide - calcium sulphate matrix, placed ina hydrogen - nitrogen - air flame. The emissions are used to determine0.05-1 ng of praseodymium and 1-25 ng of terbium in 1 pl of solution. Mostother lanthanoids have little effect on the determination of praseodymium,but cerium, scandium, europium and praseodymium seriously interfere inthe determination of terbium.R. BELCHER, K. P. RANJITKAR and ALAN TOWNSHENDDepartment of Chemistry, Birmingham IJniversity, P.O. Box 363, Birmingham,B15 2TT.Analyst, 1976, 101, 666-671

ISSN:0003-2654    

DOI:10.1039/AN97601BP065

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

AUGUST 1976 The Analyst Vol. 101 No. 1205 Fibre Identification and Analysis of Fibre Blends A Review* H. M. Appleyard International Wool Secvetarzat, Technical Centye, Valley Dvive, Ilkley, Yorkshire, LS29 8PB Summary of Contents Introduction Identification of Natural Fibres Animal Fibres Vegetable Fibres Silk Identification of Man-made Fibres Quantitative Analysis Introduction Although there has been an academic interest in identifying natural fibres for many years, it was largely due to the introduction of labelling laws that fibre identification and analysis of blends began to assume a much greater importance. The first labelling law for textile products came into force in the USA in 1939. Since then, similar laws have come into force in many other countries; such labelling has frequently been extended in recent years to care-labelling .The original US law placed the onus upon the manufacturer to state not only the types of fibres contained in a fabric but to state also the percentage of each type of fibre present. Such laws assume that there are people who are capable of both identifying the different types of fibres and of analysing blends. Identification of the natural fibres mainly involves the recognition of a number of diagnostic features when the fibre is viewed under a microscope. Quantitative analysis of blends of animal fibres is therefore carried out on the basis of subjective examination and consequently there are no official standards as there are for blends of man-made fibres, which can be examined by chemical means. Variation in results of chemical analysis of man-made fibre blends should be largely dependent on variation within the blend, i.e., within the yarn or fabric. International standards usually quote a precision of the order of 5 1 % at 95% confidence limits on analyses of man-made fibres.In contrast to this precision, analyses of blends of animal fibres would generally be carried out to an accuracy of kayo, or *2% if many more fibres are measured for diameter and are counted. This problem will be dealt with more fully later. Accuracy in the analysis of some blends may be an extremely difficult factor to define and sampling methods become important. It is therefore imperative to state that a quantitative analysis refers only to the sample received; this is particularly important when dealing with some types of fabric.Examples of this variation have been seen in blends of fur fibres with wool; these fibres are frequently short and fly easily during processing; there may also be differential up-take of certain fibres in the carding process. Ideally then, samples should be taken at a number of places. However, this is not always possible because, in practice, samples submitted for analysis are frequently small, sometimes barely large enough to allow a full quantitative analysis to be made. Methods of sampling are fully described in specifications issued by the British Standards Institution1 and the International Standards Organization.2 Identification of Natural Fibres Animal Fibres Identification of animal fibres can be achieved only by making use of microscopical tech- * Reprints of this paper will be available shortly.For details see summaries in advertisement pages.594 APPLEYARD : FIBRE IDENTIFICATION AND Analyst, Vol. 101 niques because staining and solubility techniques are of no value as these fibres have the same chemical composition. However, many animal fibres have very characteristic physical features that have long been recognised by microscopists. As early as 1914, Solaro3 published a book on the microscopical appearance of many fibres including natural and the limited number of man-made fibres that were in existence at that time. Later workers in the field of fibre identification began to look in greater detail into the differences in the physical features of animal hairs.For example, Hausman4 published a dissertation on the structure of mammalian hair in which he set out a descriptive scheme and nomenclature to be applied to the patterns formed by the margins of the cuticular scales. The examination of animal fibres in whole mount using a mountant of suitable refractive index, e.g., liquid paraffin, is of prime importance because it will give not only the general appearance of the fibre but also an indication of what other preparations are needed in order to show up the characteristic features. The following questions should be asked when identifying animal fibres. (i) What is the general profile of a fibre-is it regular in thickness along the length of the fibre? (ii) What is the scale pattern and does it change from root to tip? (iii) Is there a medulla and if SO of what type is i t ? (iv) Is pigment present and if so how is it distributed? (v) What is the cross-sectional shape of the fibre and of the medulla? (vi) How thick is the cuticle? (vii) What is the pigment distribution as seen in the cross-section? Scale casts Because it is not always possible to see the detail of the scale pattern in whole mount owing to the presence of a medulla, or of pigment or dye, it is necessary to make a cast of a fibre in a suitable medium.Several methods of making casts have been described; for example, Manby5 in 1933 suggested a method of making casts in celluloid. Hardy and Plitt6 later introduced a method of making casts in a plastic film and later still Auber and Appleyard' reported a method of making casts in a layer of a thermoplastic of low melting-point on a microscope slide.In this method, a 20% solution of poly(viny1 acetate) in benzene is used; approximately 0.33 ml of the solution, when spread evenly over a 75 x 25 mm slide, pro- duces a layer that is ideal for making a cast of a fibre of approximately 22 pm in diameter. There are advantages in this last method in that fine fibres, which may be difficult to remove from the plastic, can be dissolved out with sodium hydroxide solution without spoiling the cast. Coarse fibres having wide medullae are flattened by the applied load, thus producing a cast of a greater area of fibre surface. A very comprehensive system for describing scale patterns was introduced by Wildman,s which takes into consideration the over-all pattern, the form taken by the scale margins and the distance between margins.By this system, fibres can be categorised and compared on the basis of scale structure. The distance between scale margins can also be used to in- dicate the total thickness of the cuticular scale layer. Extreme differences in this thickness were shown by Appleyard and Gre~ille,~ who illustrated the thin cuticular layer on wool fibres, which is usually only one scale thick, and the thick cuticular scale layer on human hair, which consists of several closely overlapping scales whose margins are consequently very close together. Early workers on animal fibres suggested that scales on fine wool fibres were coronal; the fallacy of this suggestion was shown by Wildman,s who devised a method of examining the scale pattern round the complete circumference of a fibre by making what he termed rolled impressions.The method was later developed by Molgaardlo for examining the surface of man-made fibres. It is important to examine the scale pattern on as long a portion of fibre as possible because frequently the pattern may change from root to tip. In some instances there may be several changes in pattern over the length of the fibre, while in others the only change may be at the tip of the fibre or near the root portion. There may also be a transitional stage between two patterns, which is clearly seen on many goat hair fibres. At first it was thought that goat hair was the only fibre that showed such a transitional stage, but Appleyard and Perkin11 showed that the same pattern could be found on the short, straight fibres from the face and legs of sheep. Hairs from the same body regions on other animals were examined but no others showed this characteristic transitional change.More recently, Brunner and Coman12 described their work on fibres from some Australian animals, in which they reverted to the erroneous description of coronal scales.August, 1976 ANALYSIS OF FIBRE BLENDS 595 The scanning electron microscope has become a most useful instrument for the examination of the surface structure of fibres and no review would be complete without a reference to the work of Sikorski and Hepworth,13 who showed differences in the scale structure of some animal hairs by this means.However, although this method of examination is useful it is limited to qualitative examination and cannot be used for quantitative analysis. Cross-sectional appearance The cross-sectional appearance of animal fibres frequently offers the clearest features for identification purposes. Although the plate method, first described by Preston14 and later by Ford and Simmens,15 is very useful for cutting sections of man-made fibres where the fibre contour is the most important feature, it is far less useful for animal fibres, where much thinner sections are needed than those produced by the plate method. A very convenient microtome for producing sections of approximately 25 pm thickness is the Hardy microtome or one of the various modifications of it. A later and very useful method of obtaining sections of fibres of all types is the grinding technique described by Lomas and Sirnmens.l6 Thin sections that show clearly defined detail of the structure of the fibres can be obtained in this way. Cross-sections are most useful for the examination of natural fibres.This applies particularly to animal fibres where the cross-sectional appearance often contains the most diagnostic features; not only is the contour of the fibre important but other features such as the presence and contour of the medulla, distribution of pigments and the thickness of the cuticular layer must be noted. Such features were described by Appleyard,17 who showed photomicrographs of various preparations of fibres used for textile and other purposes. Cortical cell size Some work has been carried out in an attempt to identify animal fibres by the dimensions of the cortical cells.Two methods have been tried, the first by Satlow,18J9 who broke fibres down into cortical cells and subsequently measured their length and diameters. The second method, by Appleyard,20 involves measurement of the cortical cell diameter only. This method consists in mounting cross-sections of fibres in 2-chlorophenol, which is a swelling agent, but before the swelling action starts cortical cell boundaries are shown up very clearly. By counting the number of cortical cells in the section and relating this number to the cross- sectional area of the fibre cortex, the mean diameter of the cells can be calculated. It was found from this work that not only was there a greater number of cells in coarse fibres but that the cells were also greater in diameter than those in the finer fibres.Identi- fication of fibres by this method is limited because it was found that if the mean cortical cell diameter was plotted against mean fibre diameter the results were grouped around two almost parallel lines, with the mean diameter of cortical cells of camel hair grouped around the lower line and that of other fibres examined grouped around the upper line. Later, some fur fibres were examined in this way and the results were mostly grouped around a line about midway between the lines for camel hair and other fibres. This method of mounting sections of fibres also provides interesting and useful information about the arrangement of the cortical cells in the fibre.Vegetable Fibres Descriptions and methods of identifying vegetable fibres have been well documented by Mauersberger21 and by the Textile Institute.22 In addition to the microscopical examination of these fibres in whole mount and in cross-section, it is sometimes useful to take measurements of the ultimates. The plate and grinding methods of preparing sections are usually adequate for the purpose of fibre identification where it is necessary to observe the over-all cross- sectional shape of the bundles of ultimates as well as shape and size of the ultimates and of the lumen. Staining techniques and other tests such as rotation on drying, ashing and photomicro- graphs of the whole mounts and cross-sections of a number of vegetable fibres are given in the book by the Textile Institute.22 L ~ n i a k ~ ~ also gives much useful information on vegetable fibres.A very useful recent addition to the literature on the microscopy of cotton fibres, and in particular of the fine structure of the fibres, is given by De Gruy et aL2*596 APPLEYARD : FIBRE IDENTIFICATION AND Analyst, Vd. 101 Silk Some of the references already quoted also consider silk fibres; most workers on silk appear to refer back to Howitt,2512s who described the features that are characteristic of the different types of silk, and where the most clear indication of the type of silk is shown in cross-sections. Identification of silk may be slightly complicated by the practice of tin weighting which, although it may not alter its microscopical appearance, may affect its behaviour chemically.Tin weighting can be distinguished by the white ash skeleton which becomes incandescent in the flame of a Bunsen burner. Because silk is soluble in the solvents for wool, care must be taken in analysing blends that contain these two fibres; the usual method of determining the proportions of wool and silk in a mixture is to remove the silk by dissolving it in 75% m/m sulphuric acid. Identification of Man-made Fibres Although microscopy in various forms is useful for identifying man-made fibres, chemical methods are equally important, and because of the complex chemical structure of many man-made fibres some of the more sophisticated analytical equipment is invaluable for qualitative work.At one time stress was placed upon the value of the cross-sectional shape of fibres, but in more recent years it has become evident that cross-sectional shape cannot be relied upon as a feature to be used in fibre identification. The reason for this is that manu- facturers of man-made fibres can now produce fibres of different contours to suit specific purposes, for example tri-lobal nylon is produced for fabrics where a high degree of lustre is needed. At the opposite end of the scale, de-lustring agents are introduced to give fibres varying degrees of matt finish. Some man-made fibres can be identified by using a normal light microscope, while more positive identification can be made by using a polarising microscope, either by a simple method of noting the interference colours produced by the fibres in the parallel and the perpendicular positions with a first-order red plate inserted above the objective, or by a much more comprehensive method of measuring the birefringence of the fibres using a suitable compensator.These methods have been known for a long time but they have not been as widely used as they could have been because many workers have used chemical and staining techniques either in preference or because of the lack of the appropriate equipment. Some work was published on the use of the polarising microscope for textiles in the late 1930s, and since then other workers have described studies either on the principles of polarising microscopy or on its application. Hallimond2' produced an excellent work describing the principles and use of the polarising microscope and its application in the field of textiles was described by the Textile Institute.22 There are more sophisticated methods for identifying man-made fibres and possibly the most common is infrared spectroscopy, which has been well documented over the years.Although there is much documentation on infrared spectra, it is usually recommended that individual workers should also build up their own library of spectra. Pyrolysis gas chromato- graphy can also be used but it is considered better to use this method as a confirmatory test rather than as an initial diagnostic test. Because of inherent problems with this method, it is advisable to check the pyrogram of the unknown fibre with pyrograms of known fibres run on the same day.Because the more elaborate and sophisticated analytical apparatus is not available to many workers in the field of textile fibres, methods of fibre identification using chemical tests have been developed; the Textile Institute's publication22 gives a very comprehensive scheme of analysis that incorporates staining, solubility and melting-point techniques. Sometimes it is necessary to be able to recognise fibres from the same group but which have been made by different manufacturers. Nylons and polyesters, for example, can some- times be recognised by a trace element that has been introduced as a marker by the manu- facturer, but this can usually be done only by the larger laboratories where there is access to special equipment. On the other hand, some of the different acrylic fibres can be identified by their solubility in nitromethane at different temperatures.22 For acrylic fibres, it may be necessary to distinguish between the different types of fibres because there may be slight differences in dyeing or processing properties between fibres produced by different manu- f acturers.AztgzLst, 1976 ANALYSIS OF FIBRE BLENDS 597 Although viscose fibres of different types will dissolve in the same solvents, the appearance under the microscope of some of them may be misleading.These fibres are usually asso- ciated with an irregular cross-sectional shape, the irregularities having the appearance of striations when the fibre is examined in whole mount. In contrast, viscose fibres of high wet modulus are almost circular in cross-section and consequently when viewed under the normal light microscope they are indistinguishable from other fibres of similar cross-sectional shape ; they can, however, be distinguished by measuring their birefringence.Quantitative Analysis Quantitative analysis of fibre blends has become the subject of national and international standards and has been made necessary by the enforcement of stricter labelling laws designed to give consumer protection in addition to assisting in care labelling. At present, these standards have been limited to blends of fibres that can be separated by chemical means and they have been developed from published methods or from interlaboratory trials organised by working groups of various national and international commit tees.The accuracy of any analysis of fibre blends depends in the first instance upon (i) the extent to which the sample is representative of the material and (ii) whether or not any added matter is present on the sample. In the first place, sampling must be considered to be of prime importance; unfortunately, however, the samples received by the analyst are often small and may not be truly representative of the bulk, and this fact must be included in any report. Details of random sampling are available from the British Standards Insti- tution= and the International Standards Organi~ation.~~ The methods described range from sampling of loose fibres through various stages of processing up to the finished fabric. Wildmans also described methods of sampling blends of animal fibres in preparation for making quantitative analyses of such blends by microscopical methods.Complications arise in sampling patterned fabrics because it is important to make sure that there is a proper representation of yarns, which may consist of fibres that are different from those that are present in other parts of the pattern. In most processes, it is assumed that in intimate mixtures the fibres will have been evenly blended during the processing stages, but it cannot always be assumed that this is so with blends of loose fibres and sometimes it has been found necessary to ensure that adequate blending has taken place by hand-carding the sample before the test specimen is taken. Even after taking these precautions, it is necessary to avoid bias in sampling due to the presence of some fibres that are much coarser than the bulk or because of wide variations in fibre length.The next point to consider in the analysis is whether or not there is any added matter on the fibres. This is important because of the wide range of substances that may be added to textile fibres either to assist in processing or in the form of special finishes. Again this subject has been considered in depth by BSI and IS0 committees; a paper dealing with methods of removing added matter is included in the BSI literatures0 and it is recommended that this information should be used in combination with methods of analysis. It seems safe to assume that this question will become more difficult in the future if more resin finishes are introduced; already difficulties have been encountered with blends of resin-treated wool and man-made fibres.It has been found that some resin finishes are completely soluble in sodium hypo- chlorite solution and others only partially so. It may be found that some of these partially soluble finishes will dissolve in sodium hydroxide solution while those which are soluble in sodium hypochlorite solution may be only partially soluble in sodium hydroxide solution. When a suitable random and prepared sample has been obtained, quantitative analysis can be carried out by sequential dissolution of the component fibres in the blend in the appropriate solvent. It is not possible in a paper of this type to go into all the details of quantitative analysis of man-made fibre blends by chemical means; such details can be obtained by reference to various published standards, e.g., B631, IS032 and the EEC directive.33 Some examples of the solvents to be used for specific blends are given in Table I.Although this table is by no means a complete list of either fibres or solvents, it does serve to give some indication of the types of fibres that can be found either in blends or, as sometimes occurs, as contaminants. Where cellulose fibres are referred to in the table, this term in- cludes cotton, bast and leaf fibres and various forms of viscose fibres. The test methods described in various standards referred to earlier are separated into those required for binary and for ternary blends. Alternative methods are given for some ternary598 APPLEYARD : FIBRE IDENTIFICATION AND Analyst, Vol.101 TABLE I EXAMPLES OF SOLVENTS USED IN ANALYSIS OF FIBRE BLENDS Fibre mixture Fibre being removed Solvent Secondary cellulose acetate with Secondary cellulose acetate Acetone other fibres except modacrylic Animal fibres with other fibres Animal fibres except acetate, silk, poly(viny1 alcohol) or regenerated protein fibres Viscose with cotton Viscose Alkaline sodium hypochlorite solution* Sodium zincate solution or formic acid and zinc chloride? Secondary cellulose acetate with cellulose triacetate benzyl alcohol Cellulose triacetate with some Cellulose triacetate Dichloromethane other fibres Cellulose fibres with polyester Cellulose fibres 75% m/m sulphuric acid Acrylics, some modacrylics or Acrylics, modacrylics and NN-Dimethylformamide Secondary cellulose acetate 70% V/V aqueous acetone or chlorofibres, with other fibres chlorofibres Chlorofibres with some other Chlorofibres Silk with animal fibres Silk Cellulose fibres and animal fibres Nylon with other fibres Nylon fibres Cellulose fibres Azeotropic mixture of carbon disulphide and acetone 75% m/m sulphuric acid 70% m/m sulphuric acid 80% m/m formic acid * Approximately 1 M solution (33-37 g 1-1 of active chlorine) to which 5 g 1-1 of sodium hydroxide t Known as Marschall's method.34 are added.blends because of the effects on some fibres of the solvent used to remove one or other of the components of the blend. In order to avoid some of these difficulties, four variants to the test method for ternary mixtures are given in the BS,31 IS032 and EEC33 documents.These variants include, for example, removing two components successively from a single specimen, or using two specimens and first removing one component from one sample and another component from the second sample, when the mass of the third component can be found by difference. A list of fibre mixtures with the names of the solvents and the variants to be used is given in BS 4407.31 The composition of these fibre blends is calculated on the dry masses of the component fibres and subsequently that result is corrected for moisture regain by using the figures published by various countries, e.g., Trade Descriptions Act 1973,35 or in an EEC directive. The term regain is used to denote the absorption of water by the fibre and has been defined as the mass of moisture present in a textile material expressed as a percentage of the oven-dry mass.Because of the variations in standard regains adopted by different countries, the whole question of allowances for regains is at present under consideration. A microscopical method of making a quantitative analysis of mixtures of some man-made fibres or man-made fibres and wool has been devised by lo ma^.^^ This is known as the point counting method and is based upon a system of counting the numbers of different fibres in a cross-section. A specially designed eyepiece graticule is used and points are scored where the marks on the graticule fall on the fibre images. An equation for calculating the composition of the blend is given in the paper.Analyses of animal fibre blends are subjective because they are dependent upon the ability of the analyst to recognise fibres of different origins. Because of the character of animal fibres, analyses can be carried out only by using microscopical techniques. After making a preliminary examination of the fibres in the blend, using the techniques described under Identification of Natural Fibres, the next step is to prepare slides following the procedure for measuring fibre diameter by the projection microscope method described by the IWT0.37 The sampling techniques for blends in different forms, described by Wildman,s should be followed, and the slides made up from the random specimen of the bulk. The normal con- ventions for diameter measurement are followed and mean diameters, the standard deviationsAugust, 1976 ANALYSIS OF FIBRE BLENDS 599 and the numbers of fibres of each type on the slide are calculated.amount ( A ) of the component fibres can be calculated from the equation From these figures the where %?A = number of pieces measured and counted of A fibres, dA = mean diameter of A fibres and UA = standard deviation of fibre diameters of A fibres. Corresponding values for B fibres are %?B, d~ and GB. If the fibres are non-medullated the specific gravity can be ignored, but if any of the fibres are medullated then this fact must be taken into account by measuring the diameter of the medulla, where this is possible, and introducing the mean diameter, standard deviation and number of medullae into the equation.When the medulla cannot be measured satis- factorily, as in many rodent hairs, then the specific gravity of the fibres must be taken into account. Values for the specific gravity of a number of fibres are given in the publication by the Textile Institute.22 The accuracy of this method from a statistical view-point depends upon the number of fibres measured and counted. Bray38 produced a series of tables for calculating the number of fibres to be measured and counted based upon mean diameters, standard deviations and coefficients of variation of the component fibres obtained from preliminary measurements. The tables give the numbers to be measured and counted in order to give an accuracy of &4y0 at 95% confidence limits. In general, this accuracy has proved to be sufficiently accurate because of the possible variations from the original blend during processing; how- ever, provision was also made by Bray to work to an accuracy of &2y0 if necessary.Although several workers have quoted equations for calculating the mass percentage of animal fibres in a blend, the equation quoted by Wildman,8 based on Bray’s unpublished papers, has proved to be highly satisfactory and is now well established. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. References British Standards Institution, “Methods of Obtaining Laboratory Test Specimens of Textile International Standards Organization, “Preparation of Laboratory Test Samples and Test Specimens Solaro, A., “Studio Microscopico e Chemico pel Reconoscimento della Fibre Vegetali, Lave, Peli, Hausman, L.A., Am. Nut., 1920, 54, 496. Manby, J., Jl R. Microsc. SOG .,,, 1933, 53, 9. Hardy, J. I., and Plitt, T. M., An Improved Method for Revealing the Surface Structure of Fur Auber, L., and Appleyard, H. M., Nature, Lond., 1951, 168, 763. Wildman, A. B., “The Microscopy of Animal Textile Fibres,” Wool Industries Research Association, Appleyard, H. M., and Greville, C. M., Nature, Lond., 1950, 166, 1031. Molgaard, J., Nature, Lond., 1959, 184, 264. Appleyard, H. M., and Perkin, M. E. A., J . Text. Inst., 1965, 56, 1. Brunner, H., and Coman, B. J., “Identification of Mammalian Hair,” Inkata Press, Victoria, 1974. Sikorski, J., and Hepworth, A., “The Place of the SEM in Textile Research and Industry,” Pro- Preston, J.M., “Modern Textile Microscopy,” Emmott and Co. Ltd., London, 1933. Ford, J. E., and Simmens, S. C . , J . Text. Inst., 1959, 50, 148. Lomas, B., and Simmens, S. C., J . Microsc., 1970, 92, 37. Appleyard, H. M., “Guide to the Identification of Animal Fibres,” Wool Industries Research Satlow, G., 2. Vey. Dt. Ing., 1944, 88, 328. Satlow, G., Faserforsch. Text. Tech., 1963, 16, 143. Appleyard, H. M., J l R. Microsc. Soc., 1967, 87, 1. Mauersberger, H. R., “Matthews Textile Fibres,” Sixth Edition, John Wiley & Sons, New York and “Identification of Textile Materials,” Seventh Edition, Textile Institute, Manchester, 1976. Luniak, B., “The Identification of Textile Fibres,” Pitman, London, 1953. deGruy, I. V., Carra, J. H., and Goynes, W. R., “Fine Structure of Cotton-An Atlas of Cotton Howitt, F. O., “Silk,” Textile Institute, Manchester, 1948. Howitt, F. O., “Bibliography of the Technical Literature on Silk,” Hutchinson, London, 1946. Hallimond, A. J ,, “The Polarising Microscope,” Third Edition, Vickers Limited, York, 1970. Materials for Chemical Testing,” BS 4658 : 1970. for Chemical Testing,” ISO/DIS 5089. Pelliccie, Seti Naturali, Seti Artificiale,” Milan, 1914. Fibres,” US Department of the Interior, Wildlife Circular No. 7, 1940. Leeds, 1964. ceedings of the Second Annual SEM Symposium, Chicago, April, 1969. Association, Leeds, 1960. London, 1954. Microscopy,” Marcel Dekker Inc., New York, 1973.600 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. APPLEYARD British Standards Institution, “Methods of Fibre Sampling for Testing,’’ BS 2645 : 1965. International Standards Organization, “Textiles-Fibres-Methods of Sampling for Test,” ISO/DIS British Standards Institution, “Methods of Test for Textiles,’’ Handbook No. 11, British Standards British Stazdards Institution, “Methods of Test for Quantitative Chemical Analysis of Fibre International Standards Organization, “Textiles, Binary Fibre Mixtures. Quantitative Chemical EEC Council Directive, July 1972 (72/276/EEC). Marschall, A., Kunstseide Zellwolle, 1940, 22, 215. Department of Industry, “Explanatory Guidelines on the Textile Products (Indicating Fibre Lomas, B., J . Text. Inst., 1975, 66, 301. IWTO, “Measurement of Fibre Diameter-Projection Microscope Method,” IWTO-8-61. Bray, R. J., unpublished work. 1130. Institution, London, 1974. Mixtures, BS 4407 : 1969. Analysis,” ISO/R 1833 : 1971. Content) Regulations 1973,” HM Stationery Office, London, 1974. Received January 26th, 1976 Accepted March loth, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100593

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Analyst, August, 1976, Vol. 101, @. 601-610 601 Colorimetric Determination of Thiamine Propyl Disulphide and Thiamine Disulphide in Their Respective Pharmaceutical Preparations M. V. Sivarama Krishnan, S. N. Mahajan and G. Ramana Rao Quality Control Laboratories, Indian Drugs and Pharmaceuticals Limited, Hyderabad-500037, India A rapid and sensitive colorimetric method for the determination of thiamine propyl disulphide and thiamine disulphide, as pure drugs and in their respec- tive dosage forms, has been developed, which is based on a colour reaction with 2,6-dichloro-p-benzoquinone-4-chlorimine in the presence of methanolic hydrobromic acid in a chloroform medium. The colours produced from thiamine propyl disulphide and thiamine disulphide have absorbance maxima a t 445 and 460 nm, respectively.The concentration range for Beer’s law com- pliance is 0-5.75 pgml-1 for thiamine propyl disulphide and 0-12 pgml-l for thiamine disulphide. The method can be applied to the analytical control of thiamine propyl disulphide and thiamine disulphide in dosage forms with simple matrices, but when other compounds carrying an amino group are present a separation is necessary as the proposed colour reaction is a general one for all such compounds. The removal of riboflavine and analgin, present in some of the currently available commercial formulations of thiamine propyl disulphide and thiamine disulphide, was found to be necessary and is described. Thiamine propyl disulphide and thiamine disulphide are more easily absorbed and have longer retention times in the body compared with conventional forms of vitamin B, and are now used in various pharmaceutical formulations.Zima and Hotovyl have shown thiamine disulphide to be a metabolic product of thiamine in animals and to possess a higher and more sustained activity than thiamine. Zima et al. also state that thiamine disulphide is reduced to thiamine under physiological conditions by cysteine or glutathione.2 Thiamine propyl disulphide and thiamine disulphide are not official in any pharmacopoeia and the methods available in the literature for their determination are tedious and time consuming. As the disulphides of thiamine are not directly oxidisable to thiochrome, they are first reduced to thiamine with cysteine under appropriate conditions ; the free thiamine is then isolated by chromatography and oxidised to thiochrome in alkaline solution with potassium hexacyanoferrate(II1) for fluorimetric mea~urement.~ Fujiwara and Matsui4 have suggested cyanogen bromide in place of potassium hexacyanoferrate(II1) for the oxidation of thiamine to thiochrome while Patrick and Wright5 used mercury(I1) oxide for the same purpose.Methods based on these principles have been described by various for the determination of disulphide forms of thiamine in biological materials. A new fluorimetric determination of thiamine propyl disulphide with the copper(1) ion has been reported by Yamane et a1.l0 A polarographic determination of thiamine propyl disulphide in combination with other vitamins has been worked out by Asahill Thiamine and thiamine propyl disulphide have been separated by paper ionophoresis and determined by ultraviolet spectrophotometry at a wavelength of 245 nm.12 The only colorimetric method13 published for the determination of thiamine propyl disulphide in tablets is that involving the use of bromocresol green, which has an error of &6%. 2,6-Dichloro-~-benzoquinone-4-chlorirnine is a well known colorimetric reagent for pyridoxine and other phenolic compounds.Recently it has also been used for the deter- mination of piperazine and its salts14 under different experimental conditions. Its bromo analogue has been reported to give a colour reaction with disulphides in alcoholic solution at pH 2.0.15 However, when this reaction was extended to the former reagent for colour development with thiamine propyl disulphide and thiamine disulphide, negative results were obtained. These investigations led to the development of the method reported here.602 Analyst, VoZ.101 The colorimetric procedure described below , using 2,6-dichloro-~-benzoquinone-4-chlorimine as a colorimetric reagent, is simple, sensitive and accurate and is applicable to the analysis of the thiamine compounds when pure and in their respective dosage forms with simple matrices. KRISHNAN et al. : COLORIMETRIC DETERMINATION OF THIAMINE Experimental Apparatus Colorimeter. A Spectronic-20 (Bausch Pr Lomb Inc., Rochester, N.Y.) was used. Reagents and Materials The high-purity thiamine propyl disulphide and thiamine disulphide used in the preparation of the standard graphs were obtained from Takeda Chemical Industries Limited, Japan, and E.Merck, Darmstadt, Germany, respectively. The purity of both of the samples was determined by titration with perchloric acid. Other chemicals and reagents employed were of analytical-reagent grade. 2,6-Dichloro-p- benzoquinone-4-chlorimine (N-chlor0-2~6-dichloro-p-benzoquinone monoimine) solution, 0.05 and 0.2% in methanol. Use freshly prepared reagent. Hydrobromic acid solution in methanol, 0.01 and 0.02 M. Prepare an approximately 0.1 M solution of hydrobromic acid by diluting 6.5 ml of hydrobromic acid (sp. gr. 1.48) to 500 ml with methanol. After ascertaining the exact concentration by titration, dilute with methanol to give 0.01 and 0.02 M solutions. Hydrochloric acid solution, 2 M.Florisil, 100-200 US mesh (chromatographic grade). Amberlite ion-exchange resin, IR-4B (OH). Preparation of Standard Graphs Thiamine propyl disulphide Weigh accurately about 90mg of thiamine propyl disulphide and dissolve it in sufficient methanol to produce 100 ml of solution. Dilute 5 ml of this solution to 50 ml with methanol. Place 0.2-, 0.6-, 1.0-, 1.2- and 1.6-ml portions of the standard solution in separate 25-ml calibrated flasks and add successively 10 ml of chloroform, 1.0 ml of 0.05% 2,6-dichloro- p-benzoquinone-4-chlorimine reagent and 0.5 ml of 0.01 M methanolic hydrobromic acid solu- tion to each flask. Prepare a corresponding blank by substituting methanol for the standard solution. Mix the solutions and allow them to stand for 20 min. Make the solutions up t o volume with chloroform and measure the absorbances at 445nm against the blank.Thiamine disulphide Weigh accurately about 180mg of thiamine disulphide and dissolve it in methanol to produce 100ml of solution. Dilute 5ml of this solution to 50ml with the same solvent. Proceed as described under Thiamine Propyl disulpphide, up to and including the addition of chloroform. Then add successively 1 .O ml of 0.2% 2,6-dichloro-~-benzoquinone-4-chlor- imine reagent and 0.5 ml of 0.02 M methanolic hydrobromic acid solution. Mix the solutions and allow them to stand for 40 min. Make the solutions up to volume with chloroform and measure the absorbance at 460nm against the blank. Prepare a blank as described under Thiamine propyl disulphide. Preparation of Sample Solution methanol as given under Preparation of Standard Graphs.Prepare solutions of the thiamine propyl disulphide and thiamine disulphide samples in Tablets and Capsules Transfer an accurately weighed portion of the powder, equivalent to about 18 mg of thiamine propyl disulphide, into a 200-ml calibrated flask. Next add 100ml of methanol and digest the contents of the flask on a water-bath for 10 min, adding methanol intermittently to replace the lost solvent. Cool and make up to volume Thiamine firopyZ disulphide.August, 1976 PROPYL DISULPHIDE AND THIAMINE DISULPHIDE IN PHARMACEUTICALS 603 with chloroform. Shake the flask thoroughly and filter the contents through a dry Whatman No. 1 filter-paper. Reject the first 30ml of filtrate and collect the subsequent portion in a glass-stoppered Erlenmeyer flask.Thiamine disulphide. Take an amount of powder equivalent to about 36 mg of thiamine disulphide and proceed as described under Thiamine Propyl disulphide. Thiamine propyl disulphide with ribojavine. The procedure described above yields a solution containing a small amount of riboflavine with these samples. Although the riboflavine does not interfere with colour development, slightly higher spectrophotometric results are obtained owing to the contribution of the riboflavine’s absorbance at the wavelength of absorbance measurement. Riboflavine was therefore eliminated from the sample solution by adsorption on to a Florisil column, through which the thiamine propyl disulphide or thiamine disulphide passes unadsorbed.The following procedure, based on that described by Higuchi and Brochmann-Hanssen,16 was adopted. Prepare a slurry of 4 g of Florisil in methanol and transfer it to a clean, dry, 25-ml burette (internal diameter, 12 mm) plugged with a pad of fine glass-wool. Wash the resulting column with two 10-ml portions of methanol. Transfer an amount of capsule powder, representing about 45 mg of thiamine propyl disulphide, into a 100-ml calibrated flask, add about 80 ml of methanol and digest the mixture on a water-bath for 10 min, adding methanol occasionally in order to maintain the volume. Cool the flask and dilute to volume with methanol, then filter the contents through a dry Whatinan No. 1 filter-paper, rejecting the first 30 ml; pipette out 10 ml from the subsequent portion on to the Florisil column.Allow the solution to pass through the column at a rate of 0.3 ml min-1 and collect it in a 50-ml calibrated flask. Next, elute the column with several 5-ml portions of methanol at the same rate, collecting the eluate in the same flask up to the 50-ml mark. Thiamine disulphide with ribojavine. Take an appropriate amount of powder, equivalent to about 45mg of thiamine disulphide, and proceed as described above under Thiamine propyl disulphide with ribojavine. Thiamine $ropy1 disulphide with analgin (dipyrone) . The presence of analgin inhibits colour development with the reagent and therefore it was found to be necessary to separate the two drug entities. This was accomplished by the use of an anion-exchange resin, which retains analgin, allowing the thiamine propyl disulphide to pass through the column unadsorbed.The procedure given below was followed for the sample preparation and ion-exchange separation. Place about 20 g of Amberlite resin in a 250-ml beaker, add 100 ml of 2 M hydrochloric acid and allow the mixture to stand for 30 min. Decant the acid solution and wash the resin with de-ionised water until the washings are free from chloride ions. Transfer the resin slurry to a 25-ml burette (internal diameter, 12 mm) containing a plug of glass-wool to give a column about 18 cm in height. Wash the resin column eight times with suc- cessive 10-ml volumes of methanol in order to replace the water. As the resin shrinks in methanol the final height of the column is reduced to about 8.5 cm. Carry out the extraction procedure as described under Thiamine proPy1 disulphide with ribojnvine and transfer 20 ml of the collected filtrate to the prepared ion-exchange column.Adjust the flow-rate to 0.3 ml min-1 and collect the solution in a 200-ml calibrated flask. As soon as the solution comes to a level slightly above the resin, elute the column with suc- cessive 10-ml portions of methanol passing at a rate of 0.5 ml min-l, and collect the eluate in the same flask as the solution until the liquid reaches the 200-ml mark. Development and Measurement of the Colour Use a 1.0-ml aliquot of the sample solution for all formulations except for thiamine disulphide with riboflavine and thiamine propyl disulphide with analgin. For these last formulations, take a 2.0-ml aliquot of the sample solution in a 25-ml calibrated flask and reduce the volume to about 1.0 ml with a current of nitrogen or by gently heating.In all instances follow the corresponding procedure described under Preparation of Standard Graphs, beginning at the addition of 10 ml of chloroform. Calculate the concentrations of thiamine propyl disulphide and thiamine disulphide from their standard graphs.604 Selection of Reaction and Extraction Media Several experiments making use of commonly available laboratory solvents were carried out by use of the procedure outlined above in order to find a suitable medium for the colour reaction of thiamine propyl disulphide or thiamine disulphide with 2,6-dichloro-+-benzo- quinone-4-chlorimine and hydrobromic acid to take place. The results of these experiments are summarised in Table I.KRISHNAN et al. : COLORIMETRIC DETERMINATION OF THIAMINE Analyst, Vol. 101 Results and Discussion TABLE I DEVELOPMENT AND DECAY OF COLOUR IN DIFFERENT SOLVENTS Solvent Thiamine propyl disulphide- Water . . .. .. Butan-1-01 . . .. Ethanol .. .. Methanol . . .. Carbon tetrachloride . . Dichloroethane . . .. Chloroform . . .. Water . . .. f . Butan-1-01 . . .. Ethanol .. .. Methanol . . .. Carbon tetrachloride . . Dichloroethane . . .. Chloroform . .. Thiamine disulphide- .. .. .. .. .. .. .. .. . . .. . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Maximum absorbance attained* No colour No colour 0.121 0.267 0.475 0.515 0.621 No colour No colour 0.221 0.372 0.561 0.692 0.692 Time taken to reach maximum absorbancelmin - 60 180 5 10 20 - 60 90 6 15 40 Time to start of decay after maximum absorbancelmin - 30 90 26 110 340 - 60 210 36 105 440 * Absorbance measured a t 445 nm for thiamine propyl disulphide and 460 nm for thiamine disulphide. On the basis of the above results, chloroform was chosen for our experiments although dichloroethane would also be suitable.The substitution of hydrochloric acid or bromine dissolved in carbon tetrachloride for hydrobromic acid resulted in the formation of a similar colour. However, the greatest intensity and stability of colour was achieved by the use of hydrobromic acid, as is shown by the results given in Table 11. TABLE I1 COLOUR FORMATION WITH SOME HALOGEN-CONTAINING REAGENTS Reagent* Time to Maximum Time taken to start of decay absorbance reach maximum after maximum attainedt absorbancelmin absorbancelmin Thiamine $ropy1 disulphide- Hydrochloric acid .. .. .. 0.323 Bromine in carbon tetrachloride . . 0.436 Hydrobromic acid . . .. .. 0.521 360 60 20 Thiamine disulphide- Hydrochloric acid . . .. . . No colour - Bromine in carbon tetrachloride . . 0.311 120 Hydrobromic acid . . .. f . 0.592 40 60 30 340 - 230 440 * 0.01 M solutions were used for thiamine propyl disulphide and 0.02 M solutions for thiamine Absorbance measured a t 445 nm for thiamine propyl disulphide and 460 nm for thiamine disulphide. disulphide. Although they were found to be suitable media for colour development, chloroform and dichloroethane could not be employed for the extraction of thiamine propyl disulphide andAugzcst, 1976 PROPYL DISULPHIDE AND THIAMINE DISULPHIDE IN PHARMACEUTICALS 605 thiamine disulphide from their formulations owing to the low recoveries of these drugs obtained from simulated mixtures.Methanol was found to be the most suitable solvent in this respect. However, the presence of an excess of this solvent slowed down the colour formation. Thus, when a 2.0-ml, or greater, aliquot of the methanolic extract of the sample was taken, the excess of solvent was removed by evaporation prior to colour development. Concentration of Reagents The intensity and stability of the colour also depend upon the concentration of hydrobromic acid solution, as is shown in Figs. 1 and 2 for thiamine propyl disulphide and thiamine disulphide, respectively.It is evident that 0.5 ml of 0.01 M and 0.5 ml of 0.02 M hydrobromic I I I I 5 15 25 35 45 55 Time/min Fig. 1. Variation in the intensity and stability of the colour formed with thiamine propyl disulphide (3.85 pg ml-I; wavelength, 445 nm) on addition of 0.5-ml amounts of methanolic hydrobromic acid : A, 0.01 ; B, 0.02; C, 0.05; and D, 0 . 1 ~ . acid solution represent the optimum amounts to give maximum intensity and stability of the colour for thiamine propyl disulphide and thiamine disulphide, respectively. The optimum concentrations of 2,6-dichloro-~-benzoquinone-4-chlorimine solutions employed were 0.05% for thiamine propyl disulphide and 0.2% for thiamine disulphide, using 1 ml of the reagent in each instance (Table 111). 0.6 0.5 a 0.4 z 2 0.3 a 0.2 0.1 r n Dl Time/min Fig.2. Variation in the intensity and stability of the colour formed with thiamine disulphide (7.85 pg ml-I; wavelength, 460 nm) on addition of 0.5-ml amounts of methanolic hydrobromic acid: A, 0.01; B, 0.02; C, 0.05; and D, 0.1 M. Absorption Spectrum The absorption spectrum of the colour showed a maximum at 445 nm for thiamine propyl disulphide and at 460nm for thiamine disulphide (Fig. 3). Beer’s law is obeyed in the concentration range 0-5.75 pgml-1 for thiamine propyl disulphide and 0-12.0 pg ml-1 for thiamine disulphide. The slopes of the absorbance graphs were found to be reproducible (Fig. 4).606 KRISHNAN et aZ. : COLORIMETRIC DETERMINATION OF THIAMINE Analyst, VoZ. 101 TABLE I11 OPTIMISATION OF CHLORIMINE REAGENT CONCENTRATION Absorbance Concentration, Thiamine propyl Thiamihe % disulphide disulphide 0.025 0.05 0.1 0.2 0.3 0.495 0.465 0.496 0.502 0.496 0.503 0.496 0.504 0.496 0.504 I l l 1 400 420 440 460 480 500 520 Wavelength/nrn Fig. 3.Absorption spectrum of the colour formed from A, thiamine propyl disulphide and B, thiamine di- sulphide. 0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.011.012.0 Concentration/pg ml-' Fig. 4. Reproducibility of absorbance veysus con- centration graphs for the colour formed with A, thia- mine propyl disulphide (at 446nm) and B, thiamine disulphide (at 460 nm) with 2,6-dichloro-p-benzo- quinone-4-chlorimine. The different symbols represent results from experiments carried out in triplicate. Stability of the Colour With application of the experimental conditions described, the colour reached its full intensity in 20 and 40 min and remained stable thereafter for 340 and 440 min for thiamine propyl disulphide and thiamine disulphide, respectively (Fig.5). I I! 480 720 960 Time/min Fig. 5. Absorbance of the final coloured solution 'ueisus time after the addition of hydrobromic acid solution : A, thiamine propyl disulphide (at 445 nm) ; and B, thiamine disulphide (at 460 nm). Sensitivity The colour formation responds to a concentration as low as 0.5 pg ml-l for both compounds. The method gives highly reproducible results even at this low concentration, as can be seen from the observed standard deviations of 0.003 and 0.004 for absorbance readings on five 1-ml aliquots of the same solutions of thiamine propyl disulphide and thiamine disulphide, respectively (Table IV).August, 1976 PROPYL DISULPHIDE AND THIAMINE DISULPHIDE IN PHARMACEUTICALS 607 TABLE IV REPRODUCIBILITY OF ABSORBANCE AT LOW CONCENTRATIONS Absorbance Experiment Thiamine propyl number disulphide 1 0.065 2 0.061 3 0.069 4 0.063 5 0.065 Mean .. . . .. 0.065 Standard deviation . . 0.003 Thiamine disulphide 0.039 0.031 0.032 0.029 0.029 0.032 0.004 Interferences Starch, lactose and talc, used as excipients in the formulations analysed, did not interfere, as can be seen from the recovery data summarised in Table VII. As was mentioned under Experimental, thiamine propyl disulphide failed to yield any colour by the proposed method in the presence of analgin, which is present in some of its market formulations. Separation of the two compounds was therefore necessary.As analgin is highly soluble in water and thiamine propyl disulphide sparingly so, the removal of analgin by washing with cold water (5-10 "C) was attempted, but the recovery of thiamine propyl disulphide was low, ranging from 75 to 85%. Soxhlet extraction of the propyl disulphide with chloroform, in which analgin is insoluble, was also attempted, but the recovery was again only 88-96%. Compared with these low and erratic results, the removal of analgin on an anion exchanger gave a recovery of thiamine propyl disulphide of more than 98%. The interference of riboflavine in formulations was overcome by adsorbing it on a column of Florisil through which thiamine propyl disulphide and thiamine disulphide passed unaltered. The reproducibility of the proposed method was checked by repeated determinations on several aliquots of the sample solution and also by taking different amounts of the same sample for determination.The results given in Table V, obtained by two analysts and on different days, show that the method has good reproducibility. TABLE V REPRODUCIBILITY STUDIES ON THE PROPOSED METHOD Reproducibility with solutions made from different amounts of the same sample Reproducibility with different aliquots of the same solution I A I I A \ Thiamine propyl Thiamine Experiment Thiamine propyl Thiamine number disulphide, yo disulphide, yo disulphide, % disulphide, % 1 99.83 99.78 100.30 99.95 2 99.83 99.62 99.83 99.80 3 99.83 99.78 99.78 100.20 4 99.75 99.92 100.40 100.50 Mean .. . . 99.81 Standard deviation 0.04 99.78 0.12 100.08 100.11 0.32 0.30 Owing to the non-existence of official assay procedures for the two drugs, which would allow comparative evaluation, the method was applied to simulated mixtures to establish its accuracy and the results are recorded in Table VI. Samples of thiamine propyl disulphide and thiamine disulphide, as well as their commercially available formulations, were assayed by the proposed method. As an additional check on accuracy, recovery experiments were performed on all of the forrnulations analysed. For this purpose, known amounts of the drug were added, by way of aliquots of the standard solution, at the extraction stage. The results are summarised in Table VII. As the recoveries of the added drug are close to the theoretical value in all instances, it is to be concluded that high assay values obtained for formulations are due to the presence of more active ingredient than the manufacturer's label claim suggests (overage).608 KRISHNAN et aZ.: COLORIMETRIC DETERMINATION OF THIAMINE Analyst, VoZ. 101 TABLE VI ANALYSIS OF SIMULATED MIXTURES Sample* Amount takenlmg Amount recovered/mg Recovery, yo A 23.52 23.65 100.60 B 21.24 21.12 99.45 C 6.20 5.13 98.65 D 5.38 5.30 98.51 E 44.90 45.24 100.80 F 45.08 44.94 99.70 G 48.50 47.92 98.81 H 49.40 48.90 99.00 I 43.20 43.10 99.77 44.10 44.30 100.50 45.40 45.62 100.50 46.90 46.70 99.59 JK L * A to D: thiamine propyl disulphide weighed and mixed with 250 mg of lactose and 50 mg of starch. E and F: 5 mg of riboflavine additional to the composition of A to D.G and H: 250 mg of analgin additional to the composition of A to D. I and J : thiamine disulphide weighed and mixed with 250 mg of lactose and K and L : 6 mg of riboflavine additional to the composition of I and J. 50mg of starch. TABLE VII DETERMINATION OF THIAMINE PROPYL DISULPHIDE AND THIAMINE DISULPHIDE BY USE OF THE PROPOSED METHOD Comparison with label claim by the Sample* proposed method, % A Thiamine propyl disulphide 1 99.8 2 99.4 B Thiamine disulphide 1 99.9 2 100.1 C Beneuront capsules 1 114.5 2 117.9 D Beneuron Forte? capsules 1 111.4 2 118.9 E Benalgist capsules 1 114.6 2 117.5 F Thiamine disulphide tablets 1 109.2 2 107.9 Standard added to samplelmg 4.89 4.83 20.96 21.32 24.75 25.80 9.91 9.91 Standard recovered by the proposed method, yo I 99.3 98.6 99.0 101.0 98.1 99.9 101.9 99.4 * ,4 and B : pure samples ; assay expressed as % .C: each capsule contains 5 mg of thiamine propyl disulphide. D: each capsule contains 50 mg of thiamine propyl disulphide and 5 mg of riboflavine. E: each capsule contains 50 mg of thiamine propyl disulphide and 250 mg of analgin. F: each tablet contains 35mg of thiamine disulphide. 7 Franco-Indian Pharmaceuticals Private Limited (India) are the registered proprietors of the Trade Marks Beneuron, Beneuron Forte and Benalgis. A uniformity of content test for two batches of thiamine propyl disulphide capsules was carried out and the batches were found to comply with the USP17 requirements (Table VIII). Under the conditions described for colour development, the compounds listed in Table IX produced colours with 2,6-dichloro-fi-benzoquinone-4-chlorimine reagent.Typically, 5 mg of the compound were dissolved in the minimum volume of methanol and 5 ml of chloroform, and 1 .O ml of a 0.4% solution of 2,6-dichloro-fi-benzoquinone-4-chlorimine in methanol and 1.0 ml of 0.05 M methanolic hydrobromic acid were added sequentially; the colours obtained are indicated in Table IX. It can be seen that all the compounds listed have an amine group in common. However, no colour was developed with compounds such as phenacetin, acetanilide and phthalyl sulphathiazole, wherein the amino group is acylated. Thiamine hydrochloride gave onlyAupst, 1976 PROPYL DISULPHIDE AND THIAMINE DISULPHIDE IN PHARMACEUTICALS 609 TABLE VIII VERIFICATION OF UNIFORMITY OF CONTENT IN CAPSULES OF THIAMINE PROPYL DISULPHIDE BY USE OF THE PROPOSED METHOD Sample 1 2 3 4 5 6 7 8 9 10 Amount of sample A per capsule*/mg 5.20 4.88 5.39 5.92 6.27 5.38 6.42 5.94 5.72 6.38 Amount of sample B per capsule*/mg 5.31 5.40 6.41 5.62 5.85 5.68 5.98 6.39 6.12 5.88 *Both of the commercial batches were labelled as con- taining 5 mg of thiamine propyl disulphide per capsule, but the average assay values inclusive of overages were 5.72 mg per capsule for A and 5.89 mg per capsule for B.a slight yellow colour on standing for 1 h but when the solution was brought to pH 7.0 by the addition of methanolic sodium hydroxide solution and subjected to the above colour reaction, an intense orange - red colour developed.The colour reaction can thus be presumed to be caused by involvement of the amino group with the colorimetric reagent under the experi- mental conditions described. TABLE IX APPLICATION OF 2,6-DICHLORO-jb-BENZOQUINONE-4-CHLORIMINE AS CHROMOGENIC AGENT FOR OTHER COMPOUNDS Compound Colour produced Sulphamerazine* Violet Sulphadiazine Violet Sulphaphenazole Violet Sulphadimidine* Violet Sulphathiazole Violet Sodium 4-aminosalicylate* Violet Amidop yrine Bluish violet changing Methyl dopa Orange - red to reddish brown Compound p-Bromoaniline p-Chloroaniline 2,4,6-Tribromoaniline p - Anisidine p-Toluidine NN-Dimethylaniline p-Phenetidine p-Aminobenzoic acid Colour produced Orange - red Orange - red Violet Orange - red Reddish brown Bluish green Reddish brown Violet turning to reddish brown *These compounds gave colours on adding a more concentrated solution of hydrobromic acid (1.0 ml of 0.1 M) instead of 0.05 M solution.The authors thank Mr. K. Venkateswara Rao for technical assistance and Takeda Chemical Industries, Japan, and E. Merck, Germany, for the generous donation of high-purity samples of thiamine propyl disulphide and thiamine disulphide, respectively. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. References Zima, O., and Hotovy, R., Arzneimittel-Fo~sch., 1956, 6, 623; Chem. Abstr., 1957, 51, 2181i. Zima, O., Ritsert, K., and Moll, Th., 2. Physiol. Chem., 1941, 267, 210; Chem. Abstr., 1941, 35, 40456. Jansen, B. C. P., Red Trav. Chtim. Pays-Bas Belg., 1936, 55, 1046; Chem. Abstr., 1937, 31, 14438. Fujiwara, M., and Matsui, K., Analyt. Chem., 1953, 25, 810. Patrick, R., and Wright, J. F. H., Analyst, 1949, 74, 303. Ventura, U., Giacolone, E., and Sciorelli, G., Int. 2. VitamForsch., 1966, 36, 286; Chem. Abstv., 1967, Itada, N., J . Vitam., 1959, 5, 61; Chem. Abstr., 1959, 53, 20151~. Rindi, R., and Perri, V., Int. 2. V'itamForsch., 1962, 32, 398; Chem. Abstr., 1964, 61, 6033d. Yano, M., Vitamins, Kyoto, 1958, 15, 640; Chern. Abstr., 1962, 56, 14695. Yamane, Y., Miyasaki, M., and Tanaka, T., Vitamins, Kyoto, 1971, 43, 307; Chem. Abstv., 1971, Asahi, Y., J . Vitam., 1958, 4, 118; Chem. Abstr., 1958, 52, 20892~. 66, 416c. 75. 675265.610 KRISHNAN, MAHAJAN AND RAO 12. 13. 14. 15. 16. 17. Hoshino, M., Nakamura, S., Kuriyama, M., and Iwata, T., Takeda Kenkyusho Nempo, 1957, 16, 10; Grzegorzewicz, W., Pawlak, E., Chrzaszcz, W., and Surowiecki, J., Acta Pol. Pkarm., 1972, 29, 601; Baggi, T. R., Mahajan, S. N., and Ramana Rao, G., J . Ass. 08. Analyt. Chem., 1974, 57, 1144. Guzzoni, M. T., and Pietra Lissi, T., Farmaco, Ed. Sci., 1964, 19, 771; Chem. Abstr., 1965, 62, 3408e. Higuchi, T., and Brochmann-Hanssen, E., “Pharmaceutical Analysis,” Interscience Publishers Inc., “United States Pharmacopoeia,” XVIIIth Revision, Mack Co., Easton, Pa., USA, 1970, p. 930. Chem. Abstr., 1958, 52, 10265h. Analyt. Abstr., 1973, 25, 431. New York, 1961, p. 666. Received May 29th, 1975 Amended February 20th, 1976 Accepted Murch 19tk, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100601

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

Artalyst, AGguust, 1976, Vol. 101, pp. 611-615 61 1 Determination of Nortriptyline in Sugar-coated Tablets by High-performance Liquid Chromatography* J. R. Salmon and P. R. Wood E. R. Squibb and Sons Limited. Reeds Lane, Moreton, Merseyside A method is described for the determination of nortriptyline in sugar-coated tablets. Nortriptyline is extracted into dilute aqueous hydrochloric acid containing trifiupromazine hydrochloride as an internal standard. The acidic extract is then chromatographed by a reversed-phase separation using a basic solvent system in conjunction with a phenyl Corasil column. The main advantages of the method are the simple extraction, the over-all speed of analysis and the degree of precision obtained. The method has been used to monitor the stability of experimental formulations and could readily be modified for the assay of single tablets.Nortriptyline is used in combination with fluphenazine for the treatment of mild to moderate anxiety, tension or agitation. The development of a new nortriptyline - fluphenazine com- bination product containing 30 mg of nortriptyline base and 1.5 mg of fluphenazine hydro- chloride per tablet required methods for monitoring the stability of experimental formulations. The BP assay for nortriptyline in tablets is based on ultraviolet absorbance measurements following extraction of the drug from basic solution into diethyl ether, followed by back- extraction into dilute acid. As any degradation products of nortriptyline would be expected to exhibit ultraviolet absorbance, this procedure will not be indicative of stability.Similarly, the non-aqueous titration described for the determination of nortriptyline in the raw material and capsules cannot be considered to indicate stability. The level of 1,2,4,5-dibenzocyclohepta- 1,4-diene-3-one (dibenzsuberone), described as ketone in the BP, an impurity and a potential degradation product of nortriptyline, is determined by quantitative thin-layer chromatography. Several gas-chromatographic have been described for the determination of nortriptyline in biological fluids. The low levels of nortriptyline in these fluids have led to attempts to increase sensitivity by the formation of derivatives suitable for electron-capture detecti0n.~-8 For the determination of nortriptyline at relatively high levels in phanna- ceutical preparations the sensitivities achieved by derivative formation and electron-capture detection are not required.A gas-chromatographic method involving the chromatography of nortriptyline as its base was developed in our laboratories. This method involved the aqueous extraction of nortriptyline hydrochloride, followed by the addition of sodium hydroxide solution and extraction of the base with cyclohexane. Although this procedure was shown to be suitable for the assay of uncoated tablet cores, the presence of a sugar coating on the finished product gave rise to severe interference. This interference was shown to result from the need to extract the active ingredient from the basic solution. Recently, two high-performance liquid chromatographic methods have been described for the separationg and determinationlo of nortriptyline. These methods are based on absorption chromatography and would therefore require the extraction of nortriptyline from an aqueous extract of the tablets. Thus, the problems associated with the gas-chromatographic assay can be anticipated in any attempt to apply these procedures to sugar-coated tablets.In order to overcome this difficulty the reversed-phase chromatographic method described in this paper was developed. Experiment a1 Reagents Acetonitrile. Reagent grade. Methanol. Analytical-reagent grade. Aminoniuum carbonate. Analytical-reagent grade. * Presented in part a t the 35th International Congress of Pharmaceutical Sciences (FIP), Dublin, 1975.612 SALMON AND WOOD : DETERMINATION OF NORTRIPTYLINE Analyst, Vol.101 Hydrochloric acid, 1 moll-l. Tri$upromaxine hydrochloride solution (internal standard). A 1.25 g 1-1 solution in distilled Nortriptyline hydrochloride (reference standard). Eluting solvent. Methanol - acetonitrile - 0.25% m/V aqueous ammonium carbonate solu- water. tion (40 + 40 + 20 V/V). Apparatus The high-performance liquid chromatography apparatus was constructed from the following equipment : a 500-ml capacity, high-pressure chromatography pump (Metering Pumps Limited, London) ; a variable-wavelength, single-beam detector (Cecil 212 Variable Wave- length UV Monitor) ; and a septum injection head (Perkin-Elmer, Part No. 0087-3015). Chromatograms were recorded on a strip-chart recorder. A stainless-steel column (1 000 x 2.1 mm), packed with phenyl Corasil (Waters Associates Inc.) by the modified tap-fill procedure described by Kirkland,ll was used.After use, the columns were flushed with methanol; several of these columns have been in use for over 1 year with no loss in performance. The use of other types of high-performance liquid chromatography equipment, for example the Chromatronix 3100 liquid chromatograph, in conjunction with the described procedure has proved satisfactory. The liquid chromatograph was operated under the following conditions: flow-rate 1.0 ml min-1, detector wavelength 254 nm, sensitivity 0.2 at full-scale deflection. The detector was operated at 254 nm rather than at the wavelength of maximum absorbance for nortriptyline (239 nm) so that the method could readily be transferred to laboratories using fixed-wavelength (254 nm) detectors.Chromatograms were recorded using a chart speed of 0.5 cm min-1. A 10-pl syringe (SGE 10 BLR) was used for sample injection. Other apparatus required included an ultrasonic bath and general laboratory glassware. Procedure Transfer three tablets (each containing an amount of nortriptyline hydrochloride equivalent to 30 mg of the base) to a 100-ml stoppered calibrated flask. Add to the tablets about 50 ml of distilled water and, by agitation in an ultrasonic bath, ensure their complete disintegration. Add 20.0 ml of triflupromazine hydrochloride solution (internal standard) and 2.5 ml of hydrochloric acid (1 mol F), then dilute the suspension to 100 ml with distilled water.Thoroughly mix the suspension. Filter an aliquot of the suspension on a filter-paper (What- man No. 42). Discard the first 10-15ml and then collect 10ml of the filtrate. Solutions prepared in this way should be stored in stoppered vessels in the absence of light if any delay in chromatographing the solution is necessary. Prepare a standard solution of nortriptyline hydrochloride in the following manner, Weigh accurately 100 & 5 mg of nortriptyline hydrochloride reference standard into a 100-ml calibrated flask. Add approximately 50 ml of distilled water and agitate the mixture in order to dissolve the nortriptyline hydrochloride. Next add 20.0 ml of triflupromazine solution and 2.5 ml of hydrochloric acid (1 moll-'). Dilute the mixture to 100 ml with distilled water and thoroughly mix the solution. With the chromatograph operating under the conditions described, inject into it 5 p l of the standard and sample solutions.Calculation The nortriptyline base content of a single tablet (w), measured in milligrams, is given by R, x 3 x 1.139 R x W, w = where R is the peak height ratio of nortriptyline to triflupromazine for the sample, R, the peak height ratio of nortriptyline to triflupromazine for the standard, W, the mass (in mg) of nortriptyline hydrochloride used for preparation of the standard solution and 1.139 is a conversion factor from nortriptyline hydrochloride to nortriptyline base.A upst, 1976 I N SUGAR-COATED TABLETS BY HPLC Investigation of Experimental Variables Solvent System 613 Over the last 2-3 years we have used reversed-phase high-performance liquid chromato- graphy to solve several analytical problems.During this period we used an empirical approach in establishing suitable solvent systems, which was based on the initial use of a 1 + 1 V/V mixture of water - methanol or water - acetonitrile, followed by modification of the ratio of the solvents or, for example, addition of acetonitrile to methanol - water mixtures. By use of the above approach the solvent system methanol - acetonitrile - 0.25% m/V ammonium carbonate solution (40 + 40 + 20 V / V ) was developed. Initially, the ammonium carbonate was added to ensure that nortriptyline, triflupromazine and fluphenazine underwent chromatography as their bases. However, the concentration of ammonium carbonate in the solvent system was shown to have a considerable influence on the chromatography.This effect is illustrated in Figs. 1 and 2. 3 600 3 500 p 400 300 E C .- a c, E 200 0 0.1 0.2 0.3 0.4 Ammonium carbonate solution, % m/V Fig. 2. Effect of concentration of Fig. 1. Effect of ammonium carbonate con- ammonium carbonate solution on the centrvation on the chromatography of fluphen- azine, triflupromazine and nortriptyline : a, in- jection; b, solvent; c, fluphenazine; d, triflu- promazine ; and e, nortriptyline. Solvent system: 40 ml of methanol + 40 ml of aceto- nitrile + 20ml of 1, 0.1; 2, 0.2; 3, 0.25; 4, 0.3; and 5, 0.4% m/ V ammonium carbonate solution. chromatography of fluphenazine (1)’ triflupromazine (2) and nortriptyline (3). . Twenty millilitres of the am- monium carbonate solution were added to methanol - acetonitrile (40 ml of each).Increasing the concentration of ammonium carbonate results in small changes in the retention time of triflupromazine and fluphenazine but, over the concentration range examined, causes a large decrease in the retention time of nortriptyline. On the basis of these observa- tions the concentration of ammonium carbonate selected for use in subsequent work was 0.25% m/V, the chromatograms obtained being consistent with the requirements of resolution and analysis time. The effect of ammonium carbonate on the reversed-phase chromatography of antibiotics has been discussed12 and several reasons for its effect on separations suggested. It has been demonstrated that small adjustments to the ammonium carbonate concentration may be necessary in order to maintain uniformity of assay from laboratory to laboratory.A study of the effect of the amount of hydrochloric acid that was injected on to the column was effected by varying the amount of hydrochloric acid solution (1 moll-l) added to placebo tablets that had been spiked by the addition of known amounts of nortriptyline hydrochloride. Over the range 1-6 ml no changes in retention time or peak height ratio were observed. Linearity The graph of response to variation in the nortriptyline concentration over the range 0.1-1.0 mg ml-1 Linearity of response was confirmed by chromatographing standard solutions.614 SALMON AND WOOD : DETERMINATION OF NORTRIPTYLINE Analyst, VoZ. 101 with a constant triflupromazine concentration (0.25 mg ml-l) was linear and passed through the origin.Similarly, the graph of response to variation in the triflupromazine concentration over the range 0.05-0.25 mg ml-l with a constant nortriptyline concentration (1.0 mg ml-1) was linear and passed through the origin. Effect of Temperature In the range 2342 "C only slight changes in the retention times of the various com- ponents were observed. However, significant changes in the peak height ratios occurred. This effect is illustrated in Fig. 3. ~ Y 0.8 I I 20 30 40 2 Temperature/'C Fig. 3. Effect of temperature on the peak height ratio of nortriptyline hydrochloride to triflupromazine hydrochloride. At normal ambient temperatures the effect of a 1 "C change corresponded to a 1.5% change in peak height ratio.Consequently, it was thought to be advisable to control the temperature of the chromatographic column thermostatically. To minimise errors caused by temperature fluctuation, injections were carried out in the sequence standard, sample 1, sample 2, standard, repeating the sequence for further samples. By using this procedure, precision was maintained and the need to control the column thermostatically was eliminated. Results and Discussion The recovery of the assay was established by the addition of aqueous solutions of nor- Following disintegration of the tablets the solutions were triptyline to placebo tablets. assayed by use of the recommended procedure. The results are given in Table I. TABLE I DETERMINATION OF NORTRIPTYLINE I N SPIKED PLACEBO EXTRACTS Mass of nortriptyline hydrochloride/mg 1 30.2 31.13 103.1 30.2 30.26 100.1 30.2 30.29 100.3 31.7 32.00 101.1 31.7 32.14 101.4 31.7 31.83 100.4 Added Found Recovery, yo From the results given in Table I the mean recovery was calculated to be lOl.l%, with a coefficient of variation of &1.17'.August, 1976 IN SUGAR-COATED TABLETS BY HPLC 615 The results obtained for tablets (Table 11) support the recovery and precision data obtained during the development of the assay.TABLE I1 DETERMINATION OF NORTRIPTYLINE IN EXPERIMENTAL FORMULATIONS Nortriptyline base contentlmg per tablet Assay number Bktch 1 Batch 2 Batih 3 1 31.5 30.1 30.0 2 31.3 30.4 30.4 3 31.0 30.4 29.7 4 31.0 30.4 29.1 5 30.3 30.3 29.2 Mean . . . . 31.0 30.3 29.7 The assay procedure has been used over a period of 18 months to evaluate the stability of several experimental formulations.During this period the assay has proved to be satis- factory. An assay of placebo tablets showed that the average contribution of the tablet excipients was +0.75%. Our main aim in developing this procedure was to evaluate the stability of experimental formulations. It was thus necessary to show the stability-indicating nature of the method. Solutions of dibenzsuberone and ally1 suberol, both potential degradation products of nor- triptyline, were chromatographed. As the product also contains fluphenazine it was necessary to establish the occurrence of any possible interferences from fluphenazine or its degradation products. The retention times given in Table I11 for the major components and their possible degradation products showed the assay to be stability indicating.TABLE I11 RETENTION TIMES FOR NORTRIPTYLINE, FLUPHENAZINE AND THEIR POSSIBLE DEGRADATION PRODUCTS Compound Retention time/s Nortriptyline . . .. .. 360 Triflupromazine . , .. .. 180 Fluphenazine . . .. .. 90 Fluphenazine sulphoxide . . 90 Dibenzsuberone . . .. .. 84 Ally1 suberol . . .. .. 96 The method described compares favourably with the BP assay method, particularly because of its ability to indicate stability. The time of operation for a single assay is somewhat greater than that required for the BP assay. However, for multiple assays the operation time is significantly reduced and is comparable with that of the BP method. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. References Hucker, H. B., and Stauffer, S . C., J. Pharm. Sci., 1974, 63, 296. Norheim, G., J. Chromat., 1974, 88, 403. Borga, O., and Garle, M., J. Chromat., 1972, 68, 77. Gaffney, T. E., Hammer, C. C., Holmstedt, B., and McMahon, R. E., Analyt. Chem., 1971, 43, 307. Walle, T., and Ehrsson, H., Acta Pharm. Suec., 1971, 8. 27. Braithwaite, R. A., and Widdop, B., Clin. Chim. A d a , 1971, 35, 461. Borga, O., Palmer, L., Linnarsson, A., and Holmstedt, B., Analyt. Lett., 1971, 4, 837. Wallace, J. E., Hamilton, H. E., Groggin, L. K., and Blum, K., Analyt. Chem., 1975, 47, 1616. Knox, J. H., and Jurand, J., J. Chromat., 1975, 103, 311. Watson, I. D., and Stewart, M. J., J . Chromat., 1975, 110, 389. Kirkland, J. J., J . Chromat. Sci., 1972, 10, 129. White, E. R., Carroll, M. A., and Zarembo, J. E., J . Antibiot., 1975, 28, 205. Received February 25th. 1976 Accepted April 2nd, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100611

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

616 Analyst, August, 1976, Vol. 101, pp. 616-621 Specific Determination of Testosterone in Oily Injections Emil Fahmy, Dawoud A. Yassa Research Department, Socikttk Misr pour l’lndustrie Pharmaceutique, 92 El Mataria Street, Post El Zeitoun, Cairo, Egypt and Nagi Wahba Biochemistry Department, Faculty of Medicine, A in Shams University, Cairo, Egypt Multi-stage extraction from a solution of testosterone oily injection in heptane with 85% ethanol yields a hormone extract in which testosterone can be specifically determined by applying the sulphuric acid - picric acid reaction. A parallel analysis of standard samples, which are dissolved in different oily media, and of identical blank media shows that the oily residues and pharmaceutical adjuvants do not interfere at a wavelength of 640nm and that the recoveries, calculated by comparison with a standard or directly from the value of EiE for the ester concerned, are within the range 98.4-101.6~0.The precision, for assays of 25 mg ml-1 solutions, is f 1%. Testosterone is also determined with similarly high accuracy in complex injections containing oestradiol or progesterone although these hormones pass quantitatively into the hormone extract. The possibility of determining both testosterone and either progesterone or oestradiol in the same extract, without further manipulation, constitutes a useful saving in time and labour. The procedures that have been used so far for the assay of testosterone in injections are not specific and are insufficiently accurate. Those favoured by recent editions of most pharmac~poeiasl-~ are based on the colorimetric determination of its isonicotinyl hydrazone, a reaction given by any A4,3-ketosteroid, by substances present in the solvent oils and, after a long period of standing, by androsterone and oestrone.* As regards accuracy, it may be significant that the limits of quantitative determination permitted by both the British and United States pharmacopoeias form the wide range of 90-110~o, although the former applies the reaction directly to a solution of the oily injection in chloroform while the latter requires prior chromatographic separation of the steroid on a double-layered column of silanised and purified siliceous earths.A new colour reaction for testosterone has recently been r e p ~ r t e d , ~ involving the addition of 1% picric acid solution after allowing the dry hormone to react with concentrated sulphuric acid.This addition produces a stable green colour with its wavelength of maximum absor- bance (Amax.) at 640 nm and a molar absorptivity ( E ) of 2.1 x lo4 1 mol-l cm-l. No colour was given by the majority of the other steroids tested, including progesterone and oestradiol, while of the few steroids that gave appreciable absorbance values at 640 nrn only hydroxy- progesterone was likely to be present with testosterone in a complex dosage form. This paper describes a procedure, based on the sulphuric acid - picric acid reaction, which is applicable to oily injections containing testosterone esters alone or in association with progesterone or oestradiol esters.Experimental Because of the specificity of the sulphuric acid - picric acid colour reaction, extraction of the hormone was necessary mainly to exclude the oily medium and other materials contained in it, which might undergo carbonisation on heating with concentrated sulphuric acid or which might otherwise interfere with the result by absorption of light in the region of 640 nm. The oily media in general use for preparing hormone injection solutions are arachis oil and other vegetable oils with similar physical characteristics, ethyl oleate or a castor oil- benzyl benzoate (7 + 3) mixture. In order to find a general-purpose extraction procedure,FAHMY, YASSA AND WAHBA 617 providing a non-interfering residue, blank formulations were prepared containing, in each medium, 0.1% of propyl gallate, 0.1% of citric acid and 2% of benzyl alcohol.Identical authentic samples were also prepared by adding testosterone propionate to each medium to form solutions containing 25 mg ml-l. This concentration of testosterone pro- pionate, being the lowest commonly available, was used in order to simulate the worst conditions of interference from the oily medium. A method of solvent extraction was contemplated in which ethanol was used to extract the hormone from a solution of the oily injection in an alcohol-immiscible solvent. Both blank and sample formulations were extracted in this manner and portions of the extracts assayed. The procedure was considered successful if, after evaporating the ethanol, the residue from the blanks failed to show an appreciable absorption at A,,,.640 nm while that from the sample gave an almost lOOyo recovery. At first, a simple extraction of a solution of 1 ml of sample in 40 ml of light petroleum (boiling range 40-60 "C) with eight 20-ml portions of 90% ethanol was tried. The combined ethanolic extracts were diluted to exactly 200ml with ethanol and the test was run after evaporating 1-ml aliquots of the diluted extracts. Satisfactory results, as judged by the above-mentioned criteria, were obtained when the experiment was applied to formulations in arachis oil. However, this procedure was soon abandoned as it proved inapplicable to formulations in ethyl oleate and castor oil; the dried extracts obtained charred on treatment with concentrated sulphuric acid, even if 85% ethanol was used, while lower concentrations of ethanol resulted in low hormone recoveries.A multi-stage extraction procedure was then investigated (see below). The use of equili- brated heptane and absolute ethanol, as recommended by the Pharmacopoeia Nordica 1963, gave satisfactory results with formulations in arachis oil but again the residues from ethyl oleate and castor oil formulations charred on being treated with concentrated sulphuric acid. In an attempt to obtain a non-carbonisable residue and in view of the solubility in ethanol of both ethyl oleate and castor oil, ethanol of a lower concentration was tried. It was noticed that the degree of carbonisation diminished as the concentration was reduced to 95 and 90%. When 85% ethanol was used the residue gave a clear solution with concentrated sulphuric acid and on subsequent treatment with a 1% solution of picric acid and measurement at 640 nm it gave almost lOOyo recovery (Table I).This concentration was critical, as when 80% ethanol was employed the recovery decreased to 91.16% (mean of three experiments). The efficacy of solvents other than heptane was tested in order to facilitate substitution in case of the unavailability of heptane; hexane and light petroleum (boiling range 40-60 "C) were equally effective. Reagents and Equipment All materials were of analytical-reagent grade. Heptane. Ethanol, 85%. Sulphuric acid, B.P. specijcation (97% mlm). Sulphuric acid, 50% VlV solution in water. Picric acid, 1% vn1V solution in water. Carl Zeiss PMQ I1 spectrophotometer.Place 85ml of absolute ethanol in a 100-ml calibrated flask, dilute to volume with water and mix. Absorbance measurements were made on this instrument using 1-cm cells. Extraction Solvents Equilibrate about 60 ml of heptane with about 80 ml of 85% ethanol by shaking them in a 250-ml separating funnel and separating the lower phase, consisting of ethanol saturated with heptane, from the upper phase of heptane saturated with ethanol. Assay Preparation Transfer an accurately measured volume of the injection equivalent to about 100-150 mg of the testosterone ester into a 10-ml calibrated flask, add heptane to the mark and mix.618 FAHMY, YASSA AND WAHBA: SPECIFIC DETERMINATION Analyst, VoZ. 101 Multi-stage Extraction Prepare three 50-ml separators, numbered I, I1 and 111, and a 100-ml calibrated collecting flask.Pipette 15 ml of heptane saturated with ethanol into each separator; to separator I add 1 ml, accurately measured, of the assay preparation and mix. Next pipette 10 ml of ethanol saturated with heptane into the separator, shake it and leave the layers to separate; run the lower, ethanolic, phase into separator 11, shake the separator and run the separated ethanolic phase into separator 111. Again shake the separator and transfer the lower phase into the collecting flask. Repeat this operation three times, each time using a fresh 10-ml portion of ethanol saturated with heptane. Thereafter discard the contents of separator I. Pipette 10ml of ethanol saturated with heptane into separator 11, shake it and use the separated ethanolic phase to extract the contents of separator 111; collect the separated lower phase in the same collecting flask.Then discard the contents of separator 11. Finally, pipette 10 ml of ethanol saturated with heptane into separator 111, shake the separator and again collect the lower phase in the same collecting flask. Hormone Extract flask now contains the hormone extract. Dilute the contents of the calibrated collecting flask to volume with absolute ethanol; this Colour Production Place exactly 1 ml of the hormone extract into a long test-tube (1.2 x 20 cm) and dip the bottom of the tube into a boiling water-bath in order to evaporate most of the ethanol, the last trace being blown off by a current of nitrogen. To the dry residue, still dipped in the water-bath, add exactly 2 ml of concentrated sulphuric acid and continue the heating with occasional shaking for 10 min; then cool the acidic solution in ice.Gradually add exactly 2 ml of picric acid solution, with shaking, and heat the mixture in a boiling water-bath for 3 min before cooling it to room temperature. Finally, add exactly 10 ml of 50% sulphuric acid to the green-coloured solution and mix; determine the absorbance of the solution in a l-cm cell. Calculation Calculate the content of testosterone (ester) in the solution, either from the absorbance obtained by repeating the operation of colour production but using a solution in ethanol of a standard sample of the appropriate testosterone ester at a concentration of 100-150 pg rn!-l, comparable with that of the hormone extract, or directly from the absorbance found using Etzk values of 616 for testosterone propionate, 526 for the enanthate, 515 for the cypionate and 504 for the phenylpropionate.Recovery Table I is a summary of the results obtained from seven standard solutions that were TABLE I RECOVERY* OF 25 mg OF TESTOSTERONE PROPIONATE ADDED TO VARIOUS MEDIA Medium Arachis oil . . .. Arachis oil . . .. Ethyl oleate . . .. Ethyl oleate . . .. Ethyl oleate . . .. Average .. .. Castor oil - benzyl benzoate Castor oil - benzyl benzoate Standard deviation . . Coefficient of variation, yo Amount foundtlmg ml-1 Amount found$/mg ml-1 .. 25.15 26.16 .. 24.60 24.60 .. 25.16 25.40 . . 25.25 25.26 24.60 .. 24.65 .. 24.65 24.80 24.80 ..24.60 .. 24.86 24.94 . . k0.301 f 0.322 .. 1.21 1.29 * Mean of three experiments. $ By comparison with a standard. Using E$& = 616.August, 1976 OF TESTOSTERONE I N OILY INJECTIONS 619 prepared by adding testosterone propionate to the different media to give concentrations of 25 mg ml-1. The results are calculated, for comparison purposes, by use of both methods described under Calculation above. In all the subsequent tables the results were calculated by using the value of E$& of the ester concerned. Precision The precision of the method was assessed by determining the testosterone propionate content of the hormone extracts obtained from ten aliquots of an ethyl oleate standard sample that contained 25 mg ml-l of testosterone propionate (Table 11).TABLE I1 RECOVERY OF 25 mg OF TESTOSTERONE PROPIONATE ADDED TO AN ETHYL OLEATE BLANK FORMULATION MEDIUM Aliquot number . . .. 1 2 3 4 5 6 7 8 9 10 Amount found/mgml-1 . . 24.8 24.9 25.2 25.2 24.6 25.3 25 25.15 25.25 24.6 Mean recovery + standard deviation = 25.0 f 0.263 mg ml-l. Complex Formulations The analysis of complex formulations was studied by using combinations of 25 mg of testosterone propionate and either 1 mg of oestradiol monobenzoate (Table 111) or 10 mg of progesterone (Table IV). Eight samples of each combination were prepared, using a common medium of ethyl oleate containing 0.1% of propyl gallate, 0.1% of citric acid and 2% of benzyl alcohol. TABLE I11 RECOVERY* OF 25 mg OF TESTOSTERONE PROPIONATE (REPORTED METHOD AND ISONIAZID METHOD) AND 1 mg OF OESTRADIOL MONOBENZOATE (B.P.1968 METHOD) FROM 1 ml OF ETHYL OLEATE Testosterone propionate found, % Reported W i d method method 100.8 104.4 100.5 103.6 99.6 106.8 100.8 106.6 100.2 104.8 101.3 105.9 100.4 103.3 100.4 105.6 Mean . . .. .. 100.5 105.1 Standard deviation . . & 0.498 & 1.315 * Eight determinations. Oestradiol monobenzoate found by B.P. method, yo 102.0 101.5 98.9 102.6 100.2 99.8 101.1 100.8 100.8 - + 1.21 It was considered desirable to determine the amount of the female hormone that was transferred into the hormone extract by the modified multi-stage extraction. In order to determine the progesterone present the total A4,3-ketosteroids content is first assayed by use of the isoniazid reaction. Evaporate 2 ml of the hormone extract to dryness, add 5 m l of chloroform and shake to dissolve the residue.Complete as described in the British Pharmacopoeia 1968 under the assay of Nandrolone Decanoate Injection B.P., com- mencing at the words “add 10 ml of the reagent.’’ Calculate the total steroids content from the absorbance obtained by repeating the operation with an ethanolic solution of standard testosterone propionate and standard progesterone in amounts equivalent to those theoretically present in the hormone extract. The progesterone content is calculated by subtracting the testosterone propionate content, as assayed by the reported method, from the total A4,3-ketosteroids content.620 FAHMY, YASSA AND WAHBA: SPECIFIC DETERMINATION Analyst, VoZ. 101 TABLE IV RECOVERY* OF 25 mg OF TESTOSTERONE PROPIONATE AND 10 mg OF PROGESTERONE FROM 1 ml OF ETHYL OLEATE BY PROPOSED AND ISONIAZID METHODS Mean .. .. .. Standard deviation . . * Eight samples. Testosterone propionate/mg (sulphuric acid - picric acid reaction) 25.1 25.05 25.1 25.1 25.4 25.2 25.15 25.02 25.14 kO.119 Total testosterone propionate + progesteronelmg (isoniazid reaction) 35.15 35.15 35.1 35.6 35.4 35.5 35.25 35.22 36.29 Progesteronelmg (by subtraction) 10.05 10.1 10.0 10.5 10.0 10.3 10.1 10.2 10.15 Oestradiol was assayed by means of the British Pharmacopoeia method, using the dried residue from 4 ml of the hormone extract. Complete as described under the assay of Oestradiol Injection B.P., commencing at the words “Add 1 ml of iron - phenol reagent.” Results and Discussion The extracts obtained from the blank media, prepared from various oils, gave no colour and the absorption at 640 nm was practically nil. However, the recoveries from the various authentic samples (Table I) were within the range 9S.4-101.6~0.Accuracies of this order provided an assurance that the extraction was efficient, that interference from the various oil residues and pharmaceutical adjuvants was absent and, hence, that the method could be applied to commercial preparations with a medium of unknown composition (Table V). TABLE V ANALYSIS OF TESTOSTERONE ESTERS IN COMMERCIAL OILY INJECTIONS BY USE OF THE REPORTED METHOD Active ingredient hmount declared/mg ml-1 Amount found/mg ml-1 Testosterone propionate . . .. 25 25.05 25.10 25.15 Testosterone enanthate . . .. 50 50.1 50.4 50.15 Testosterone propionate .. .. 25 25.4 Oestradiol monobenzoate .. 1 25.10 25.15 Testosterone enanthate . . .. 65 66.3 Oestradiol valerate . . .. 4 66.2 66.6 Testosterone propionate . . .. 15 15.10 15.35 Progesterone . . .. .. 10 16.90 Testosterone propionate . . .. 25 26.1 Progesterone . . .. .. 10 25.4 25.2 The good agreement of the results, whether calculated by comparison with a standard or directly by using the value of E:&, is accounted for by the reproducibility of the reaction and stability of the colour. While we employed the latter method of calculation exclusively throughout the rest of this work, it should be noted that this is only possible when analytical- quality reagents are used (this is especially true of the sulphuric acid) and when the detailed extraction and reaction procedures are adhered to.August, 1976 OF TESTOSTERONE I N OILY INJECTIONS 621 The results of measurements of precision (Table 11) showed that the average value obtained as a result of ten individual assays of an authentic sample was 25mg ml-l and that no individual assay differed from the theoretical value by more than 1.6% ; the standard deviation of k0.263 mg ml-1 indicated that the assay could be repeated with good precision.Tables 1x1 and IV show that both oestradiol monobenzoate and progesterone accompanied testosterone propionate quantitatively into the hormone extract. As expectedJ5 the presence of either hormone did not vitiate the results of testosterone assays carried out by using the sulphuric acid - picric acid reaction.In contrast, when the isoniazid reaction was applied to hormone extracts containing testosterone and oestradiol monobenzoate, the results of the testosterone assays tended to be rather high, while with extracts containing testosterone and progesterone, the results represented the total hormone content. Advantage could not be taken of the ratio between the absorbance intensities at 640 and 470 nm, which has been shown to be valuable in confirming the identity of te~tosterone,~ as the absorbance at 470 nm was unduly high. This interference at 470 nm was traced, while working with blank formulations, to contamination of the hormone extract with oil residues. Conclusion In view of the non-specificity of the reactions currently available for the determination of testosterone, a number of workers handling oily injections of this steroid attempted to reduce the interference from extraneous matter by prior chromatographic separation of the steroid, e.g., Umberger,* Roberts and Florey6 and Smith.’ Few of these methods have achieved general acceptance, probably because of their intricacy or lack of reproducibility.However, Smith’s method has been adopted, after modification, by the US Pharmacopeia.2 The introduction of oily injections including mixtures of sex hormones has created further analytical problems. Although these mixtures have not achieved official recognition so far, such preparations have been in common use for many years. Nevertheless, no satisfactory methods for their analysis have, to our knowledge, been reported.The phenolic ring characteristic of oestradiol affords a means of separating it from other sex hormones and using specific reactions for its determination. However, differentiation between testosterone and progesterone presents serious difficulties because of their structural similarity. A more fruitful approach, in our view, was to concentrate on attempting to find selective reactions rather than to achieve absolute separation of the constituent con- cerned by involved means. As is reported above, the application of the sulphuric acid - picric acid reaction to a hormone extract obtained by a simple multi-stage liquid - liquid extraction procedure has afforded a convenient solution to a number of problems associated with the determination of testo- sterone in simple and complex formulations, especially those combining testosterone with progesterone. The feasibility of assaying progesterone or oestradiol together with testosterone in the same hormone extract, without further manipulation, constitutes a useful saving of labour and time. References 1. 2. 3. 4. 5. 6. 7. “British Pharmacopoeia 1973,” The Pharmaceutical Press, London, 1973, p. 463. “United States Pharmacopeia,” XVIIIth Revision, Mack Publishing Company, Easton, Pa., 1970, “Pharmacopoeia Nordica Edito Norvegica,” Aas and Wahls Boktrykkeri, Oslo, Volume 111, 1963, Umberger, E. J., Analyt. Chem., 1955, 27, 768. Fahmy, E., Yassa, D. A., and Wahba, N., Analyst, 1974, 99, 759. Roberts, H. R., and Florey, K., J . Pharm. Sci., 1962, 51, 794. Smith, E., J . Phrarm. Sci., 1967, 56, 630. p. 713. p. 207. Received September 9th, 1975 Accepted December 30th, 1975

ISSN:0003-2654    

DOI:10.1039/AN9760100616

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

622 Anabst, August, 1976, VoL 101, pp. 622-633 Computer-aided ldentif ication of Powdered Vegetable Drugs Georgina H. Jolliffe and G. 0. Jolliffe Department of Pharmacy, Chelsea College, University of London, Manresa Road, London, S W3 6LX The identification of 174 powdered crude drugs of organised structure is aided by computer analysis of the recorded observations of eleven simple histological characters, thus enabling a relatively inexperienced microscopist to reach a positive conclusion with comparative ease. The analyst may be called upon to identify a powdered crude drug and this problem can be approached by chemical and microscopical examination. Simple chemical analysis provides information about the active constituents (e.g. , alkaloids) present , semi-specific or specific chemical tests (e.g., Vitali Morin test for tropane alkaloids) elucidate the nature of the com- pound(s) present, while chromatographic and spectroscopic analysis will provide the profile of the compounds (e.g., alkaloids such as hyoscine and hyoscyamine) present. Nevertheless, the correct identity (e.g., Atropa belladonna L., Datura stramonium L., Hyoscyamus muticzGs L., Hyoscyamus nzger L.) of the powdered crude drug necessitates a detailed microscopic examina- tion of the sample followed by reference to suitable for the identification of powdered crude drugs and/or powder atlases7-11 or the use of the punched-card system pro- posed by Nelson and Fish in 1969 and published in manual form by Nelson12 in 1972. This approach, of necessity, is laborious and considerable experience is required to enable the analyst to identify the powder by microscopy.The use of reference atlases, while a valuable aid in confirming identity, does not initiate a systematic approach to the problem of identification of powdered vegetable drugs. Tables of histological features are sometimes difficult for the inexperienced analyst to interpret as the description, particularly of “inflorescences and floral members” and “seeds and fruits,” may apply to the powdered drug or, sometimes, it is more appropriate to the sectional appearance of the drug. Before an atlas or table can be used the drug must be referred to its morphological group; error at this stage is time consuming and frequently results in incorrect identification. The punched-card system requires the assessment of over 100 characters , some of which are difficult for the inexperienced analyst to interpret, and entails micro-scale measure- ments.Hence, it was pertinent to devise a scheme whereby an analyst relatively inexperi- enced in microscopy could rapidly obtain the correct identity of a powdered organised crude drug from as comprehensive a list as possible, not necessarily limited to those in current commercial demand. Therefore, a computer program has been developed based upon 11 main characters (Fig. 1). In the program is a data bank of coded information corresponding to the characteristics of each drug listed (Table I). One point is allocated for each agreement, giving a maximum score of 30 points and the three highest scores are printed out (more if identical points have been scored by a number of drugs).The program also includes characteristics of items that are present but may be difficult to identify-in these circumstances a score of one is given, regardless of the input data; this procedure makes some allowance for lack of observation by the inexperi- enced microscopist and for differences due to natural variation. The computer printout (Fig. 2) indicates the drug(s) that agree with the input data (Le., 30 points scored) and when the input data does not entirely agree with any drug (i.e., the maxi- mum of 30 points has not been scored) the computer printout indicates for the highest scores the number(s) of the characters that should be checked. Final confirmation of identity of the powder by the proposed method should, in common with other aids to identification, be ob- tained wherever possible by comparison with an authentic reference standard of the named drug.or Computer Program In the program, written in BASIC, the 11 main (30 specific) characters for 174 powdered crudeJOLLIFFE AND JOLLIFFE CHARACTERS - Enter + if present, - if absent 7 623 NO. +/- 1 * * a ... POLLEN . . . . . . . . . . . . . . . . . . . . . . . . . Prisms Needles Sphenoids ClusterdRosettes Microcrystals . . . . . . CALCIUM OXALATE 26 27 28 29 30 Crystal Layer 1 Crystal Sheatha ALEURONE~ . . . . . . . . . . . . . . . . . . CORK . . . . . . . . . . . . . . . . . . . . . LIGNIFIED PARENCHYMA . . . . . . . . . . . . STOMATA ... TRICHOMES ... . . . . . . raminaceous overing Unicellular Mu lticel lu lar Glandular Stalk Unicellular Stalk Mu lticellu tar Head U nice1 I u lar I ead Multicellular .. . . . . VESSE LS/TR ACH E I DS ... rignified ... b o n- I ig n if ied STONE CELLS . . . . . . . . . . . . . . . . . . FIBRES . . . . . . . . . . . . . . . . . . . . . STARCH . . . . . . . . . . . . . . . . . . . . . ... Fl Fl Fl ... 9 ... 10 14 19 20 24 25 aCrystal sheath or calcium oxalate localised in the region of the main veins For the purpose of this observation aleurone includes protein as grains or amorphous protein Fig. 1. Analytical data sheet. drugs of organised structure are coded in DATA statements in such a form that 0 represents a character definitely absent, 1 a character definitely present and 2 a character that is present but sometimes difficult for the less skilled microscopist to confirm.For convenience, the 30 specific characters are arranged in groups of five digits with the leading zeros suppressed. The DATA statement also includes coded information on the morphological group of the drug. The flow diagram of the program is shown in Fig. 3. The drugs are, in general, referred to by their common names but the corresponding botanical names are given in the list of drugs in Table I. The program calls for a reference to be submitted in the first instance, followed by suitable input data. The input data are required in a slightly different form from the DATA statements in the program: the numeral 1 represents an absent character and 2 an observed character. The numerals are input at the terminal in six blocks of five characters and the program is designed to check that only the numerals 1 and 2 are typed in to the computer and also that six blocks of five digits only are transmitted.The necessity for the difference in data presenta- tion is to accommodate the check in the program to obviate accidental false information being accepted for processing. As an analyst may find it difficult to interpret the data as stored in the program (ie., with the leading zeros suppressed) the information is presented in Table INo. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Name of drug in program Aconite leaf Aconite root Ailanthus Almond Alstonia American veratrum Angustura Anise Areca Arnica flower Aspidosperma Bearberry Belladonna herb Belladonna leaf Belladonna root Betony Blackhaw Black hellebore Black pepper Boldo Broom tops Bryony root Buchu Cacao Calumba Canella Capsicum fruit Caraway Cardamom Cascara Cascarilla Cassia Catnep Celery Chamomile Chicory Cinchona Cinnamon Cloves Coca TABLE I LIST OF POWDERED ORGANISED CRUDE DRUGS Latin name A conitum napellus L.Aconitum napellus L. Ailanthus glandulosa Desf. Prunus communis Arcang var. dulcis Schneider Alstonia scholaris (L.) R. Br. Veratrum viride Aiton Galipea oficinalis Hancock Pimpinella anisum L. Areca catechu L. Arnica montana L. A spidosperma quebrachoblanco Schlect. Arctostaphylos uva-ursi Sprengl A tropa belladonna L. Atropa belladonna L. A tropa belladonna L.Betonica oficinalis L. Viburnum pruni folium L. Helleborus niger L. Piper nigrum L. Peumus boldus Mol. Cytisus scoparius Link. Bryonia dioica Jacquin Barosma spp. Theobroma cacao L. Jateorhiza palmata Lamarck Canella alba Murray Capsicum minimum Roxb. Carum carvi L. Elettaria cardamomum Manton var. minuscula Rhamnus purshiana DC. Croton eleuteria J . J. Bennett Cinnamomum cassia Blume Nepeta cataria L. Apium graveolens L. Chamaemelum nobile L. Cichorium intybus L. Cinchona spp. Cinnamomum zeylanicum Nees Syzgium aromaticum (L.) Merril et L. M. Perry Erythroxylum coca Lamarck and Burkill E. truxillense Rusby 11111 11111 11121 21121 21111 12111 22111 11122 11111 11111 21111 21111 11212 11212 11212 11111 21121 11111 21111 11111 11111 11111 11121 11121 21111 11121 11212 11122 21111 21121 21121 12112 11111 11122 11121 11111 21112 12112 11121 21121 Stored data for characters 1-30 (see Fig.1)* 11111 11111 12111 11211 11121 11111 11121 1121 1 11213 11113 12121 11111 11312 11111 11122 11111 11121 11113 11212 11111 11113 11123 11111 11211 11121 11131 11211 11211 11211 12121 11121 11121 11111 11213 11113 11123 11121 11121 11113 21111 11111 21111 11111 11111 11111 11111 21111 11111 21111 11111 21111 12111 121 11 11111 11121 11111 11111 21111 12111 21111 11111 21111 11111 11111 11111 12111 21111 11111 11111 11111 11111 11 121 21111 21111 11111 11111 11111 21111 12211 11111 12231 11111 11111 11111 11111 12211 11111 12222 11111 12213 12122 12122 11111 12122 11111 11111 11111 12221 12211 11111 12211 12121 11111 11111 12122 31111 11111 11111 12121 11111 12222 12211 12122 11111 11111 11111 11111 11112 11112 11112 11113 11111 11112 11111 11112 11113 12122 11113 13132 22222 22222 11112 21122 11111 11112 11112 11112 11112 11112 11112 11113 11112 11111 21122 11112 11113 11111 11111 11111 21122 11112 12122 11112 11111 11111 11112 12111 11211 11111 11112 121 12 12221 11311 12111 12221 12221 12221 12211 12211 11112 12221 11211 12222 11113 11221 11111 12121 12121 12121 1121 1 11223 12121 11211 12221 12221 12321 12211 12211 12121 12221 11221 12221 11112 12211 12112 1121 1 11221 12221 12212 11211 Morpho- logical groupt 5 8 5 6 1 8 1 3 6 2 1 5 4 5 8 5 1 8 3 5 4 8 5 6 8 1 3 3 6 1 1 1 4 3 2 8 1 1 2 5 Discrimi- nation index 100 94 99 98 82 87 98 90 86 100 72 100 100 99 100 100 86 93 99 100 99 96 98 100 91 82 100 97 89 93 100 s 9 100 90 99 87 100 89 100 100TABLE I-continued b 41 42 43 44 46 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 Cocillana Coffee Colchicum corm Colchicum seed Colocynth fruit Coltsfoot leaf Condurango Coriander fruit Cotton root bark Couch grass Cummin Damiana Dandelion root Datura innoxia herb Datura mete1 herb Derris Digitalis Digitalis lanata Dill Dioscorea Duboisia leaf Egyptian henbane herb Elecampane Eriodictyon Ergot Eucalyptus Eupatorium Fennel Fenugreek Frangula Galangal rhizome Garlic Gelsemium Gentian Ginger Ginseng Grass Grindelia Guaiacum Hemlock fruit Hemlock leaf Henbane herb Henbane leaf Henna Hops Horseradish Hydras tis Indian hemp Guarea rusbyi (Britton) Rusby Coffea arabica L.Colchicum autumnale L. Colchicum autumnale L. Citrullus colocynthis Schrader Tussilago farfava L. Gonolobus condurango Triana Coriandrum sativum L. Gossypium herbaceum L. and other spp. Agropyron repens Beauvais Cuminium cyminum L. Turnera diffusa Willd. Taraxacum oficinale Wiggers Datura innoxia Miller Datura fastuosa var. alba Nees Derris elliptica (Roxb.) Benth. and Digitalis purpurea L. Digitalis lanata Ehrh. Anethum graveolens L. Dioscorea villosa L. Duboisia myoporoides R. Br. Hyoscyamus muticus L. Inula helenium L. Eriodictyon californicum Greene Claviceps purpurea TuIasne Eucalyptus globulus Labillardihre Eupatorium perfoliatum L. Foeniculum vulgare Miller Trigonella foenum-graecum L. Frangula alnus Miller A l p h a oficarum Hance Allzum sativum L.Gelsemium nitidum Michaux Gentiana lutea L. Zingiber oflcinale Roscoe Panax schinseng Nees Lolium, Dactylis, Festuca, Poa and Phleum spp. Grindelia camporum Greene Guaiacum oficinale L. and G. sanctum L. Conium maculatum L. Conium maculatum L. Hyoscyamus niger L. Hyoscyamus niger L. Lawsonia alba Lamarck Humulus lupulus L. Cochlcaria armoracia L. Hydrastis canadensis L. Cannabis sativa L. D. malaccensis Prain 21111 11111 11111 11111 11111 11111 21121 21122 11121 11111 11122 11121 11111 21222 21222 21111 11111 11111 11122 12111 11212 21212 11111 11121 11111 21121 11111 11122 11111 21121 11111 21111 21111 12112 11111 11122 11111 11 123 21111 11122 11111 21222 21131 11121 11121 11111 11111 11121 12121 11211 11111 1121 1 11212 11111 11121 11211 11121 11112 11211 11112 11121 21312 21312 12122 11111 11111 11212 11121 11111 11312 11123 11111 11111 11121 11111 11212 11211 12121 11111 11111 11122 11 121 11121 11112 11312 11112 11112 11211 11111 21312 21111 11111 11212 11121 11 122 11311 11111 11111 11111 11111 21111 21111 11111 31111 11111 11111 21111 11211 11111 12111 12111 11111 21111 21111 31111 11111 12111 12111 11111 21111 11111 21111 21111 31111 11111 11111 11111 11111 11111 11111 11111 11111 11111 21111 11111 31111 21111 121 11 121 11 21111 21111 11111 11111 21111 11111 11111 31111 11111 11111 12121 11111 11111 11111 11111 12121 12213 12121 12122 12122 11111 12122 11112 11111 11111 11112 11112 11111 12212 11111 11111 12222 11111 11111 11111 11111 21111 11111 11111 11111 11111 22211 11112 11111 11111 11111 12122 12122 11111 12212 11111 11111 12212 11111 12221 11113 12121 11112 11121 11113 11121 11112 12111 11112 11112 11111 12221 11113 12311 11111 11221 11112 12211 11112 12211 31122 11212 11112 11111 22222 12222 21122 12232 11112 12221 22222 11111 21122 11111 11112 12211 11112 12221 21122 11111 12222 12222 11112 11211 21122 11221 11111 11111 11112 11111 21322 11222 11112 11211 11111 11111 11111 11221 11112 11221 11112 12111 11112 12221 11112 11121 11111 21221 11112 11121 11112 11132 21122 11222 11112 11211 11112 11211 11112 11111 22122 12222 22122 11113 11112 11111 12122 11212 11112 11121 11112 11221 22122 12211 1 6 8 6 3 6 1 3 1 8 3 5 8 4 4 8 5 5 3 8 5 4 8 6 4 5 4 3 6 1 8 7 8 8 8 8 4 4 9 3 5 4 5 5 4 8 8 4 x 76 2 92 3 92 * 76 2 v 99 0, 100 83 98 98 99 100 100 100 99 89 # z 85 2 98 96 5 100 !z 100 0 100 0 87 99 3 100 98 93 97 U 86 $ 84 100 100 100 100 100 95 98 97 97 99 89 100 96 100 100 g 100 g 93 8No.89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 126 126 127 128 129 130 131 132 133 Name of drug in program Indian podophyllum Ipecacuanha Ipomoea Ispaghula Jaborandi Juniper berries Kousso Lavender Lily of the valley herb Linseed Liquorice Lobelia Logwood Lonchocarpus Lucerne Mace Male fern Marigold Marshmallow Mat6 Matricaria Mustard Nutmeg Nux vomica Oak Orange leaf Orris Jalap Pareira Pellitory Peppermint Phytolacca leaf Pimento Pine Podophyllum Poke root Pomegranate Poppy petal Pyrethrum Quassia Quillaia Raspberry leaf Rauwolfia serpentina Rauwolfia vomitoria Red-rose petal TABLE I-continued Latin name Podophyllum hexandrum Royle Cephaelis ipecacuanha (Brot.) A.Rich Ipomoea orizabensis (Pellet.) Ledanois Plantago ovata Forskal Pilocarpus microphyllus Stapf. Ipomoea fiurga Hayne Juniperus communis L. var depressa Pursh Brayera anthelmintica Kunth Lavendula oficinalis Chaix Convallaria majalis L. Linum usitatissimum L. Glycyrrhiza glabra L. and other spp. of Glabra Lobelia inflata L. Haematoxylon campechianum L. Lonchocarpus nicou DC. Medicago sativa L. Myristica fragvans Van Houtten Dryopteris filix-mas (L.), Schott Calendula ojicinalis L. Althea oficinalis L. Ilex paraguensis St. Hilaire Matricaria chamomilla L. Brassica nigra Koch and B. alba Boissier Myristica fragrans Van Houtten Strychnos nux-vomica L.Quercus robur L. Cirus aurantium var. Bigaradia, Hook. f. Iris germanica L., I . pallida Lamarck, I . porentina L. Chondodendron tomentosum Ruiz et Pavon Anacyclus pyrethrum DC. Menlha pifierita L. Phytolacca decandra L. Pimento ojicinalis Lindley Pinus sylvestris L. Podophyllum peltatum L. Phytolacca decandra L. Punica granatum L. Papaver rhoeas L. Chrysanthemum cinerarifolium Vis. Picroena excelsa (SW.) Lindley QuilliZja saponaria Molina Rubus idaeus L. Rauwolfia serpentina Benth. Rauwolfia vomitoria Afz. Rosa gallica L. 11121 12111 21121 11111 11121 21121 21111 21121 21111 22111 11111 21111 11111 21111 21111 21111 11111 11111 21121 11121 31121 11121 11111 11111 11111 21121 21111 21111 11111 11111 11111 12111 21121 11111 11121 12111 31121 11111 21121 21111 21111 11121 21111 21111 11121 Stored data for characte 11121 11111 11111 11122 11111 11111 11121 11111 11111 11211 11111 11111 11112 21111 12211 11121 11111 11111 11212 11111 11111 11113 21111 12212 11113 31131 12222 11111 11111 21111 11211 11111 11111 12122 11111 11111 11312 21111 12211 11112 11111 11111 12122 11111 11111 12113 23111 12122 11111 11111 11111 11111 11111 11111 11113 21111 12122 11112 11111 11111 11112 21111 12211 11113 21111 12212 11211 11111 11111 11211 11111 11111 11211 11111 12211 12122 11111 11111 12111 21111 11111 (see Fig.1)* 11131 11111 11122 11111 11121 11111 11112 11121 11111 21111 11211 21111 11112 11111 11121 11111 11123 11111 11121 11111 11111 21111 11113 21111 11132 11111 11131 11111 11112 21111 11122 11111 11122 11111 11111 11111 11111 11111 11111 12222 11111 12211 11111 11111 11111 11111 11111 12222 11111 11111 12213 11111 11111 12211 :rs 1-30 11112 12221 11112 12221 11112 12221 11111 11121 11112 11211 11112 12121 11112 12211 12222 11112 22222 11112 11112 11222 11113 12111 11112 11221 11112 12222 11112 11221 11112 12221 21222 11232 11112 12111 11112 11221 12122 12112 11112 11221 11112 12211 12122 12112 11113 12111 11113 11121 11111 13111 11111 12221 11112 11211 11112 11121 11112 12221 11112 12211 22222 11222 11112 11111 11112 12221 11112 11121 11112 12221 11111 11221 11111 12121 11111 11112 12122 12112 11112 11221 11111 12221 31132 11212 11112 11221 11112 12221 11112 11112 Morpho- logical group t 8 8 8 6 5 8 3 2 2 4 6 8 4 9 8 4 6 8 2 8 6 2 6 6 6 1 5 8 8 8 4 5 3 9 8 8 1 2 2 9 1 5 8 8 2 Discrimi- nation index 76 97 83 78 97 96 98 100 100 100 84 95 99 57 85 100 97 86 92 99 99 90 84 76 99 99 100 98 93 97 100 98 100 96 76 100 76 100 99 45 82 99 83 84 100TABLE I-continued 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 Red sanders Rhatany Rhubarb Rue Saffron Sage Salep Sandal wood Santonica Sappan Sarsaparilla Sassafras bark Sassafras wood Savin tops Senega Senna fruit Senna leaf Serpentary Simaru ba Slippery elm Soy bean Star anise Stavesacre Stramonium herb Squill Stramonium leaf Strophanthus Sweet flag Tea Tobacco Turmeric Turpeth Valerian Vanilla Wahoo White hellebore Wild cherry Witch-hazel bark Witch-hazel leaf Xanthium Zedoary Pterocarpus santalinus L.f . Krameria triandra Ruiz et Pavon Rheum spp. Ruta graveolens L. Crocus sativus L. Salvia oficinalis L. Orchis spp. Santalum album L. Artemisia cina Berg Caesalpinia sappan L. Smilax ornata Hook. f . Sassafras variifolium (Salisbury) 0. Kuntz Sassafras oficinale Nees and Eberm. Juniperus sabina L. Polygala senega L. Cassia acutifolia Delile and C . angusdifolia Vahl Cassia acutifolia Delile and C . angustifolia Vahl Aristolochia serpentaria L. Simaruba oficinalis DC. and S. amara Aubl. Ulmus fulva Michaux Glycine soja Sieb. et Zucc. Urginea maritima (L.) Baker Illicium verum Hook. f. Delphinium staphisagria L. Datura stramonium L., D. stramonium var. tatula (L.) Torr. Datura stramonium L., D. stramonium var.tatula (L.) Torr. Strophanthus kombe' Oliver Acorus calamus L. Thea sinensis L. Nicotiana tabacum L. Curcuma domestica Valeton. Ipomoea turpethum R. Brown Valerian oficinalis L. Vanilla planifolia Andrews Euonymus atropzdrpureus Jacquin Veratrum album L. Prunus serotina Ehrhart Hamamelks virginiana L. Hamamelis virginiana L. Xanthium strumarium L. Curcuma zedoaria Roscoe 21111 21111 11121 11121 11111 11111 121 11 21111 11121 21111 12111 12112 11111 21 112 11111 21121 21121 11111 21111 21111 21111 12111 21111 11111 21222 31121 31121 21121 11121 11212 11111 11121 11111 22111 11121 12111 21121 21111 21121 11111 11111 11112 12121 11131 11111 11113 11111 11111 11112 11113 11112 11112 11122 11112 11112 11122 12211 12111 11121 11121 12131 1121 1 11111 11213 11211 21312 21111 1121 1 13131 11111 11111 11121 11 123 11112 11212 11 121 11111 11 121 12121 12112 11111 11121 11111 11111 11111 21111 11111 11 121 11111 11111 21111 11111 11111 11111 11111 11111 31111 23111 11211 11111 11111 11111 11111 21111 21111 11111 12111 12111 11111 11111 11112 12111 21111 11111 11111 21111 11111 11111 11111 11111 11211 21111 11111 11111 11111 11111 11111 12211 12122 12211 11111 12212 11111 11111 11111 11111 21111 12211 12211 12211 11111 11111 11111 11111 11111 11111 11111 12122 12122 12211 11111 12211 12122 12211 11111 12211 12211 11111 11111 11111 11111 12121 12122 12211 11112 11112 11111 11112 11112 22221 11112 11112 12122 11112 11112 11111 11112 11112 11112 11112 11112 11112 11111 11111 11113 11112 11112 11112 21122 21122 11113 11112 11112 22222 11113 11112 11112 11112 11111 11112 11113 11113 11112 12122 11112 11221 11221 21121 11111 11212 21111 11121 11221 12112 11221 13321 12221 11221 11211 12111 12211 11211 13221 12211 11221 12131 11121 12311 11111 12222 11113 13111 11221 12211 11211 11121 1222 1 1222 1 12211 11121 12221 12221 12231 12211 11111 11221 9 8 8 6 2 6 8 9 2 9 8 1 9 4 8 3 5 8 1 1 6 7 3 6 4 6 6 8 5 6 8 8 8 3 1 8 1 1 6 5 8 57 96 99 89 100 100 100 57 90 57 94 98 89 100 100 100 100 85 97 95 85 99 96 95 89 99 99 94 100 99 100 75 100 100 96 87 68 71 100 100 100 * 1 = Absent; 2 = present; and 3 = not always present or difficult to identify.t 1 = Barks; 2 r= flowers; 3 = fruits; 4 = herbs (or entire organisms); 5 = leaves: 6 = seeds; 7 = underground leaves; 8 = underground roots and stems; and 9 = woods.4628 JOLLIFFE AND JOLLIFFE : COMPUTER-AIDED 76/02/20. 10.12.58. PROGRAM DRUGID DRUCID - DRUG ID/ENTIFICATION BY I(IICROSC0PY - 10 OEC 1975 THIS PRKRAR CONTAINS 174 PWDERED ORCANISED DRKS FCU 20 1976 DO YOU WISH TO INSERT DATA? YE5 NArE/REF: ? TEST DATA 1 DATA 7 l1121,11211,21111~12212,22l22~l2211 ........................................................ NAmi/RiFERENCE - TEST DATA 1 DATE - FEB 70 19715 DATA CHECK - 11121 11211 21111 12212 22122 12211 POINTS HAVE BEEN ALLOCATED TO A AAXIlrlUirl OF 30 INPUT DATA AGREES ENTIAELY UIITH I INDIAN HEiW - HERO (OR ENTIRE UGANISI'I) 26 ERIODICTYW - LEAF 26 HOPS 9.. CHECK REF. NOS. - B 22 27 29 - HLRd (OR ENTIRE ORGANISm) CHECK REF. NOS. - 10 21 27 30 26 MATHICARIA - .FLRJER *** CHECK REF. NOS. - 8 21 28 30 26 SANTONICL - FLWER CHECK REF.NOS. - 8 21 28 30 ........................................................ a** CONFIRM BY CWIPARISW WITH AUTHENTIC SA;rtPLC *** ........................................................ lr(ORE DATA 7 Y NARL/RCf i ? TEST DATA 2 DATA 7 11121,12111J11211,11111J11112J11211 ........................................................ NAIPE/REFEHENCE - TEST DATA 2 DATA CHECK - 11121 12111 11211 11111 11112 11211 DATE - FEE 20 1976 POINTS HAVE aEEN ALLCCATED TO A AAXIirIUN OF XI 29 COCA - LEAF e.8 CHECK REF. NJS. - 1 27 SENNA LEAF - LEAF *.- WEU HEF. Nas. - 1 17 i a 27 SWEET FLAG - UNDERGROUND CHECK REF. NOS. - 1 13 29 ........................................................ '** CONFIRrrl aY CWPARISLh WITH AUTHENTIC SAMPLE **. ........................................................irIORE OATA ? Y NAtIE/t?EF: ? TEST DATA 3 DATA ? 11121.11121.11111,11111,11111,12121 . . ~ S ~ ~ . ~ . ~ . . * . ~ . . * . I ~ ~ M ~ O ~ ~ ~ ~ M O ~ ~ ~ ~ * ~ ~ ~ . * ~ ~ ~ ~ ~ ~ ~ ~ . ~ ~ ~ . ~ ~ NAiIE/iiCFERENCE - TEST DATA 3 DATE - FCB 20 197fi DATA CHECK - 11121 11121 1 1 1 1 1 11111 11111 12121 PUIiiTS HAVE BEEN ALLJCATED Til A NAXInlUM Q 30 INPUT DATA AGREES ENTIAELY WITH a CANELLA - BARK I P(MLCRANAT'E - dARK 23 BLACKHAW - BAilK *** CtlLCK REF. NUS. - 1 29 WAHJO - DARK *** CHECK REF. NOS. - 27 ........................................................ ........................................................ APRE DATA ? N a** cmmn a y CONPARISON WITH AUTHENTIC SAWLE *** CP 7.941 SCCS. RUN CWWETE. Analyst, Vol, 101 Fig. 2.Computer printout from DRUGID.August, 1976 IDENTIFICATION OF POWDERED DRUGS 629 in the same format as the input data, with the addition that the numeral 3 (not a valid number for input information) represents a character not always present or difficult to identify. The output information includes a copy of the input data for immediate checking and also a list of the more likely identities based upon a scoring system, with a maximum of 30 points. The method of scoring is as follows: (a) the input blocks of data are added numerically to the corresponding data blocks in the program for each drug; (b) each drug is given an initial score of 30 points; (c) each block of the added data is examined for a numeral 2 and one point is deducted for each numeral 2 present (see Table II), which corresponds with a disagreement between the input and output data; + List of drugs in stated Full list Title dimensions 1 ' Read: Name of drug Morphological group Analytical data Date Fig. 3.Flow diagram of program DRUGID. i Examine each addition for a figure 2 and subtract number found from 30630 JOLLIFFE AND JOLLIFFE : COMPUTER-AIDED TABLE I1 TRUTH TABLE Analyst, Vol. 101 Corresponding coded information Characters present (+) or absent (-) * - - + + +I- +I- Program: 0 0 1 1 2 2 - + - + - 4 Input: 1 2 1 2 1 2 J X X J J + Sum= 1 2 2 3 3 4 I:: J = Agreement, corresponding to numbers 1, 3, 4; x = disagreement, corresponding to number 2 only. (a) the printout comprises the drugs, their corresponding morphological group together with the highest scores, subject to a minimum of three drugs and a maximum of three different scores ; (e) when less than 30 points are obtained, the printout indicates which character number(s) (subject to a maximum of five) is at variance; (f) when more than one drug receives 30 points, complete identity can be achieved by reference to Table 111; and (g) identity is finally confirmed by comparison with an authentic reference sample.This program will also print the complete list of drugs in alphabetical order or, if preferred, in morphological groups. TABLE I11 Name of drug Alstonia Quillaia American veratrum White hellebore Anise Celery Aspidosperma Cocillana Witch-hazel bark Blackhaw Pomegranate Canella Pomegranate Cardamom Soy bean Cassia Cinnamon Chicory Elecampane Colchicum seed Ispaghula Nutmeg Condurango Wild cherry Datura mete1 herb Stramonium herb FEATURES TO DISTINGUISH DRUG COMBINATIONS SCORING 30 POINTS FOR CORRECT INPUT DATA Feature Latex tubes present Latex tubes absent; powder froths when shaken with water Calcium oxalate needles up to 100 pm Calcium oxalate needles rarely exceed 60 pm Covering trichomes frequently detached in powder, conical, slightly curved, Covering trichomes occur as small protruberances on fragments of epicarp in Powder reddish brown, no odour, taste slightly aromatic and bitter; abundant Powder light brown, spicy odour and slightly pungent and astringent taste; Powder pale pinkish buff, no odour and slightly bitter and astringent taste; Stone cells numerous, in large groups Stone cells few, isolated Cork, occasional fragments, non-lignified Cork, abundant fragments, lignified Calcium oxalate, one or two small prisms embedded in starch of the endosperm Calcium oxalate prisms in cells of the cotyledons, not embedded in starch Starch frequently more than 10 pm diameter; fibres up to 40 p m in diameter Starch rarely over 10 pm in diameter; fibres up to 30 pm in diameter Branching laticiferous vessels present Laticiferous vessels absent Testa epidermis devoid of mucilage Testa epidermis transparent cells filled with mucilage Fat present crystallises from chloral hydrate solution as feathery masses Fibres non-lignified ; latex cells filled with granular substance Fibres rare, lignified Basal cell of covering trichomes rarely exceeds 35 pm in diameter Basal cell of covering tichomes frequently measures up to 80 pm in diameter warty walls lateral view starch starch not abundant starch infrequent cellsAugust, 1976 Name of drug Derris Lonchocarpus Galangal rhizome Male fern Gelsemium Rauwolfia vomitoria Henna Rue Indian podophyllum Podophyllum Turpeth Ipomoea Wild cherry Linseed Mustard Logwood* Quassia* Rauwolfia serpentina Red sanders* Sandal wood* Sappan* Matricaria Santonica Blackh aw Canella Pomegranate? Quassia Rauwolfia serpentina Condurango Ipomoea Wild cherry IDENTIFICATION OF POWDERED DRUGS TABLE III-continued Feature Calcium oxalate prisms up to 20 pm Calcium oxalate prisms 20-30 pm Starch 18-30 pm in length, 7-15 pm in width Starch 18-25 pm in diameter Stone cells, walls heavily thickened so that lumen is sometimes occluded Stone cells, lumen greater than wall thickness Lower epidermis, cuticle striated Lower epidermis, cuticle not striated 631 Cluster crystals, few, 30-40 pm and rarely over 60 pm in diameter Cluster crystals, 20-100 p m and often over 60 p m in diameter Cluster crystals, abundant, 10-35 p m in diameter Numerous masses of resin which stain a deep yellowish brown with 0.02 N iodine solution Resin absent Sclerenchyma layer, cells long and narrow with irregularly thickened, pitted walls Sclerenchyma layer, cells with radial and inner tangential walls thickened Black mustard : Seedcoat fragments brown White mustard : Seedcoat fragments yellow Colour soluble in water, coloured purple by lime water and black with iron(II1) Aqueous extract, no colour with lime water or iron(II1) chloride solution Powder not completely lignified Red resin in all elements, colour insoluble in water, soluble in alcohol forming a deep red solution coloured violet with alkali Oil present in all elements, oil soluble in alcohol Colour soluble in water, coloured crimson by lime water, no change with iron(II1) chloride solution Covering trichomes occur as papillae on stigma surface and inner epidermis of ligulate and tubular florets Covering trichomes, usually detached in powder, unicellular, thin walled and found in groups of loosely felted masses Oil cells absent Oil cells present Oil cells absent All elements lignified All elements not lignified Fibres non-lignified ; latex cells filled with granular substance Numerous spherical masses of resin which stain deep yellowish-brown with Fibres rare, lignified forming “beaker cells” chloride solution 0.02 N iodine solution * Powders completely lignified.t These may be distinguished by their stone cells, see above. Discrimination Index The limited input data make it impossible to guarantee a unique identification and, depend- ing on the sample examined, a degree of confidence in the output can be achieved which varies with the discrimination index, I , defined by the equation I = 100 -lO(n, -1) -n2, where n, is the number of drugs listed with 30 points and n2 is the number of drugs listed with 29 points when data for a named drug are input. Thus, the discrimination index would be 100 if only one drug achieved 30 points with none scoring 29 points and zero if there were 10 drugs with 30 points and 10 with 29 points.A modified computer program was used to examine the input data for each of the drugs in order to check the validity of the program and so calculate the corresponding values for the discrimination index (Table I). The success of the program is indicated by the fact that the discrimination index is 100 for 57 drugs (about 33%), 90-99 for 71 drugs (about 41%) and only 16 drugs (about 9%) have an index less than 80, the lowest being 45. The five lowest values for the discrimination index are shown by five of the woods in the list of drugs (Table I), indicating that the chosen char- acters are not ideal for the resolution of this group of drugs. However, the program does632 JOLLIFFE AND JOLLIFFE : COMPUTER-AIDED Analyst, Vol. 101 separate them successfully from the other powders in the list and reference to Table I11 shows that each can be differentiated by application of a simple test.Work is in progress for the computer-aided identification of powdered timbers. When nl 2 2 (see the example given in Fig. 2), simple distinguishing features are shown in Table 111. However, in all instances final confirmation by comparison with an authentic sample is advisable. Computer Hardware and Software A suitable teleprinter or visual display unit was linked via an acoustic coupler and Post Office telephone line to the Kronos CDC 6600 computer, situated at Imperial College, Univer- sity of London. A copy of the program DRUGID (drug identification by microscopy) can be obtained, on request, from the authors. Users of the above computer can also be given direct access to the program, if desired.Experimental The powder is examined microscopically in the following reagents and the appropriate entry made on the analytical data sheet (Fig. 1). Chloral hydrate solution. Chloral hydrate (50 g) dissolved in distilled water (20 ml). Used for calcium oxalate, cork, stomata, trichomes, fibres and pollen. The powder should be “boiled” in the presence of the mountant in order to remove unwanted cell content. Picric acid solution. A saturated solution of picric acid in water made by dissolving 1 g of picric acid in 95ml of distilled water. Used for protein in the form of aleurone grains or amorphous protein. As many seeds and fruits contain abundant fixed oil, it is advisable to de-fat the powder by immersion in light petroleum (boiling range 40-60 “C) for a few minutes before staining.In addition, the background colour can be removed, after staining, by irrigating the mount with water, when the yellow colour given by protein will be clearly visible. Phloroglucinol solution. Phloroglucinol (1 g) plus ethanol (95%) to 100 ml. This solution is used, in the presence of concentrated hydrochloric acid, for lignified parenchyma, vessels/ tracheids and stone cells. If lignified, these structures will acquire a red coloration. Iodine water. Mix 1 volume of a weak solution of iodine BP13 with 9 volumes of distilled water. Used for starch granules, which yield a characteristic blue - black colour.Textbooks such as those by Wallis4 and Trease and Evans5 will provide useful background botanical information. The usual format of +/- indicating present/absent is used to record the assessed characters on the data sheet. This information is converted into numerals (- = 1; + = 2) to convert into suitable input data for computer analysis. Conclusion The advantage of the method described is that even a relatively inexperienced microscopist (albeit a good analyst) can, with confidence, arrive at the correct identity of any one of the 174 powdered crude drugs of organised structure listed. Therefore, identification of such powders, hitherto a problem for the specialist analyst, can be brought within the province of the general analyst. References 1. 2. 3. 4. 5. 6. Kramer, H., “Scientific and Applied Pharmacognosy,” Second Edition, Chapman and Hall, London, Claus, E. P. , “Gathercoal and Wirth Pharmacognosy,” Third Edition, Henry Kimpton, London, “Tables for the Identification of Powdered Drugs of Organised Structure, ” Pharmaceutical Society Wallis, T. E., “Textbook of Pharmacognosy,” Fourth Edition, J. and A. Churchill, London, 1960, Trease, G. E., and Evans, W. C., “Pharmacognosy,” Tenth Edition, Baillibre Tindall, London, 1972, Brain, K. R., and Turner, T. D., “The Practical Evaluation of Phytopharmaceuticals,” Wright- 1920, p. 713. 1956, p. 687. of Great Britain, London, 1963. p. 577. p. 674. Scientechnica, Bristol, 1976, p. 166.August, 1976 IDENTIFICATION OF POWDERED DRUGS 633 7. 8. 9. 10. 11. 12. 13. Greenish, H. G., and Collin, E., “An Anatomical Atlas on Vegetable Powders,” J. and A. Churchill, Fluck, H., Schlumpf, E., and Siegfried, K., “Pharmakognostischer Atlas zur Pharmacopoea Hel- Deryng, J ., “Atlas Sproszkowanych RoSlinnych Surowc6w Leczniczych,” Pahstwowy Zaklad Eschrich, W., “Pulver-Atlas der Drogen des Deutschen Arzneibuches,” Gustav Fischer Verlag, Jackson, B. P., and Snowdon, D. W., “Powdered Vegetable Drugs,” J. and A. Churchill, London, “Analytical Microscopy of Vegetable Materials,” published privately by Nelson, P. F., 114 Corsebar British Pharmacopoeia 1973,” HM Stationery Office, London, 1973, p. 251. London, 1904. vetica,” Kommissionsverlag Wepf & Co., Basle, 1935. Wydawnictw Lekarskich, Warsaw, 1961. Stuttgart, 1966. 1968. Drive, Paisley, Scotland, 1972. Received February 27th, 1976 Accepted April lst, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100622

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

634 Analyst, August, 1976, Vol. 101, $9. 634-638 Performance Characteristics for the Spectrophoto- metric Determination of Total Iron in Freshwater Using Hydrochloric Acid W. Davison and E. Rigg Freshwater Biological Association, Windermere Laboratory, Ambleside, Cumbria, LA 22 OLP A method is described for the determination of total iron in freshwater using a wet-oxidation procedure followed by spectrophotometric measurement in 5.93 f 0.07 M hydrochloric acid. It has been used successfully in routine analysis for several years and requires only standard equipment and inexpensive reagents. The criterion of detection is 0.009 mgl-1 for the 95% confidence level. The relative standard deviation at 0.2 mg 1-1 is 4.4% and at 3 mg 1-1 is 1.5% for 19 degrees of freedom. Comprehensive interference tests show that the method is adequately selective for use with freshwater.Copper is the only substance to interfere significantly and below 0.1 mg 1-1 it has a negligible effect. Measurements of the total amount of iron in freshwater are required by the water treatment industry and are of great interest to ecologists. A method for total iron content that measures particulate as well as soluble forms is necessary because the iron(I1) state is thermodynam- ically unstable in the presence of dissolved oxygen, and solubility products for the iron(II1) hydroxide, carbonate, sulphide and phosphate are readily exceeded1 to produce colloidal suspensions. Particulate organic matter may also contribute to the total iron. Therefore, measurements of soluble fractions of iron can be misleading if their contribution to the total amount is unknown.Two methods are recommended widely for the determination of iron. Atomic-absorption spectroscopy is quick and selective but lacks sensitivity.2 The alternative procedure that is usually employed is sensitive and involves the reduction of iron(II1) to iron(I1) followed by complexation with a chromogenic reagent specific for the iron(I1) ~ t a t e . ~ , ~ In order to deter- mine total iron, both methods usually require pre-treatment of the sample and it has been shown4 that only a wet-oxidation procedure gives complete digestion. Determination of iron by means of hydrochloric acid is very well known and was reported as long ago as 1914.6 One of us (E.R.) has used the method, with small modifications, for the past 11 years to determine total iron in various waters in the English Lake District having iron concentrations in the range 0.05-10.0 mg 1-l.The method has worked satis- factorily during this time with no special problems, except for the need, common to all low-level iron determinations, for meticulous cleanliness in order to prevent contamination. In view of its successful use, it was thought worthwhile to provide a full evaluation of its performance characteristics so that comparison with the currently recommended methods could be made. Although numerous publications concerning the method have appeared,G no detailed precision results are available in the literature. The suitability of the method for determinations of iron in freshwater has not been sufficiently appreciated.It has already been stressed that a wet digestion of the sample is necessary and this provides an ideal preliminary to a measurement performed in a strongly acidic medium. The nature of the absorbing tetrachloroferrate( 111) anion has received rigorous investigation' and the spectra of other metal chloride complexes have been well characterised,8 so providing a good theoretical background to the understanding of the method. Experimental Apparatus beam spectrophotometer fitted with a 1-cm flow-through cell. except the 120-ml conical fiasks, which were either Pyrex or silica. The absorbance of solutions was measured at 360 nm using a Beckman, Model 25, double- All glassware was Pyrex, It was cleaned withDAVISON AND RIGG 635 5.9 M hydrochloric acid and then thoroughly washed with distilled water.A Radisil infrared heater (700 W) was positioned with the element approximately 20 cm above a hot-plate fitted with an aluminium alloy top. Reagents 1 All reagents should be of AnalaR grade. Water. Single-distilled water that has been checked for freedom from iron. Hydrochloric acid, 5.93 & 0.07 M. 000 ml. Hydrochloric acid, 1.1 &- 0.1 M. Nitric acid, sp. gr. 1.42. Perchloric acid. sB. PY. 1.70. Dilute 500 ml of hydrochloric acid (sp. gr. 1.18) to Standard iron sohdron, 1 000 pg mkl. Dissolve 0.5 & 0.000 1 g of iron wire (99.998%) in 25 ml of 8 M nitric acid and make up to 500 ml with water. Procedure Thoroughly shake the sample, pipette 100ml into a 120-ml conical flask and evaporate to dryness by gently heating overnight on a hot-plate positioned under an infrared heater. When cool, add 0.5 -J= 0.01 ml of perchloric acid and 1 &- 0.01 ml of nitric acid and wet any residue on the walls by rotating the flask.Place on a hot-plate and heat until fuming. When the yellow - brown coloration due to organic matter, as distinct from pale yellow due to iron, has been destroyed, fume off most of the residual acids. Allow the flask to cool and add 3 & 0.1 ml of 1.1 M hydrochloric acid, rotate the flask to wash down the walls and then fume off the acid on a hot-plate. When cool, dissolve the residue in the flask in 10 & 0.05 ml of 5.93 M hydrochloric acid, taking care to dissolve any residue on the walls. Transfer the solution into a 15-ml centrifuge tube and centrifuge at 300 rev min-1 for 3 min to remove any suspended matter (e.g., silica) as sediment.Measure the absorbance in a 1-cm cell at 360 nm, using distilled water as the reference solution. The total analytical time for 10 samples is approximately 3 h 40 min and the operator time is approximately 70 min. This does not include the time taken for the 100-ml sample to evaporate to dryness but a minimum of handling is involved at this stage. Selection of Measuring Conditions The method was used exclusively for total iron measurements when it was desired to include both soluble and suspended forms. Any pre-treatment that was adopted had to be capable of destroying organic material completely and the combination of perchloric and nitric acids seemed suitable as it is a recommended2s3 rigorous digestion procedure.Evapora- tion of samples by using a combination of gentle heating on a hot-plate and infrared heating from above was chosen as comparison with evaporation in an oven or solely on a hot-plate showed that the combined technique introduced the least contamination. It also required a minimum of operator time as samples could be left without attention, which enabled them to be evaporated overnight. Desesa and Rogerss have shown absorption spectra for several metal ions in hydrochloric acid. The spectrum for iron(II1) is almost flat over the range 310-370 nm with a maximum at 360nm, while the copper spectrum has a peak at 280nm. The chosen wavelength of 360nm has, therefore, the advantages of maximum sensitivity and of being well removed from the potentially interfering copper peak.According to Desesa and Rogers, sensitivity to changes in hydrochloric acid concentration is least in the range 5.5-6.5 M. Dilution of 500 ml of concentrated hydrochloric acid to 1 000 ml produces a 5.93 -J= 0.07 M solution. This range results in a theoretical maximum error in absorbance of about 1%, and is derived from the maximum limits of percentage composition of the commercial acid. Acid from any one bottle will not exhibit this range and in the authors' experience no between-bottle variation has been noticed. All measurements were made at room temperature (21-24 "C). The absorption is known to be virtually independent of temperature in the range 10-25 O C .1 0 Colour development is instantaneous and has been reported to be stable for at least 48 h,s so the timing of the absorbance measurement was not critical.636 DAVISON AND RIGG : SPECTROPHOTOMETRIC DETERMINATION Analyst, VoZ. 101 Calibration Graph Separately prepared calibration graphs showed no significant variation and were linear from 0 to 3.5 mg 1-l. An absorbance of 1.000 corresponded to 1.88 mg 1-1 of iron. Precision Tests The experimental design for the precision test was based on defining a batch as a set of analyses performed in one day. Ten analyses were made in each batch to provide duplicate determinations for the following five samples : (a) distilled water, (b) distilled water containing 0.200 mg 1-1 of iron, (c) distilled water containing 3.00 mg 1-1 of iron, (d) a bulk sample of Windermere lake water that had been filtered through a GF/C filter, and (e) same as (d) but with 0.700mg1-1 of iron added.The complete analytical procedure was applied to all samples. The concentrations of the standards were chosen so that they represented both the low and high extremes of the cali- bration graph. Two spectrophotometric readings were taken for each solution, making a total of 20 measurements for each batch. The order of the spectrophotometric measurements was randomised and results were obtained for a total of 10 batches. The analysis of variance technique as described by Wilsonll was used to evaluate the results. This enabled estimates of the standard deviation within-batches (sw), between- batches (sb) and the total standard deviation (st) to be obtained.A summary of the statistical results is given in Table I. It was also possible to estimate the contribution of the spectro- photometer to the standard deviation of the blank values. At 10 degrees of freedom it corresponded to 0.002mg1-1 of iron. TABLE I PRECISION OF RESULTS FOR THE DETERMINATION OF IRON AFTER WET-OXIDATION OF SAMPLES Standard deviation*/mg 1-1 Mean r A \ Solution absorbance SW Sb St (a) 0.00 mg 1-1 0.010 0.004(10) - I (b) 0.20 mg 1-1 0.105 O.OOS(l0) N.S. (9) 0.009 (d) Windermere 0.164 0.007(10) N.S. (9) 0.009 (c) 3.00 mg 1-l 1.598 0.038(10) 0.023 (9) 0.044 (e) Windermere + 0.543 0.021(10) N.S. (9) 0.027 0.70 mg 1-1 Concentra- tion found/ Recovery, mg 1-1 % - - 0.308 - 1.021 101.9 f 1.7t * The degrees of freedom for the within- (sw) and the between-batch (sb) standard deviations are t 95% confidence limits also given.given in parentheses. N.S. = not statistically significant a t the 6% probability level. Interference Tests Known concentrations of possible interfering substances were added to standard solutions containing 0.00, 0.20 and 3.00 mg 1-1 of iron and the complete analytical procedure applied, These concentrations were chosen for compatibility with the precision tests and because they covered the full concentration range. The results are given in Table 11, the values being the mean of duplicate determinations. Confidence limits were calculated using the sw values in Table I. Most substances did not cause statistically significant effects, but, as is to be expected, the over-all trend was to produce an increase in absorbance. Part of the enhance- ment produced by the major cations can probably be attributed to the reagents used for supplying these cations contributing a significant iron impurity.Support for this supposition was gained by repeating the tests at half of the concentration of added ions, which had the effect of halving the interference, bringing it to an acceptable level. These reduced cation concentrations are still large compared with those in most freshwaters, so interference from this source is improbable. Copper(I1) ion was the only significant interfering species. Measurements at different copper concentrations are included in Table I1 and show that below 0.1 mg 1-1 the interference becomes negligible.August, 1976 OF TOTAL IRON IN FRESHWATER USING HYDROCHLORIC ACID TABLE I1 EFFECT OF OTHER SUBSTANCES ON THE DETERMINATION OF IRON Other substance Ca2+ .... Mg2+ . . . . Na+ .. .. K+ . . .. . . Ca2+ .. . . Mg2+ .. .. Na+ .. .. K+ . . .. .. SO,2- .. . . c1- . . .. .. NO,- .. .. HCO,- . . .. so,2- .. .. c1- . . .. . . NO3- . . . . HCO,- . . .. Concentration of other substance*/ mg 1-l .. .. . . 50 .. 250 .. . . . . 25 .. 50 .. .. .. .. .. .. . . 25 .. 40 . . 5 . . 1 .. 50 .. 2 SiO, .. .. F- .. .. .. NO,- . . .. Orthophosphate (as P) Pyrophosphate (as P) Hexametaphosphate (as P) 1 Calgonj . . .. .. 20 Alkalinity . . . . . . 300 as CaCO, cua+ . . . . . . 2 CU2f .. . . . . 1 cu2+ . . . . . . 0.1 Mn2+ . . . . . . 5 Ni2+ .. . . . . 20 coa+ ... . . . 1 Sna+ . . .. . . 1 Zn2+ . . . . .. 2 Cda+ . . . . . . 2 Pb2+ . . . . .. 10 Cr3+ . . . . .. 2 Effectt of other substancelmg 1-1 of iron r A \ [Fe] = 0.00 mg 1-1 0.025 0.013 0.017 0.006 0.011 0.006 0.000 0.015 0.004 0.000 0.009 0.002 0.218 0.107 0.017 0.002 0.003 0.004 0.017 0.006 -0.004 0.004 0.002 637 [Fe] = 0.20 mg 1-l [Fe] = 3.00 mg 1-1 0.053 0.024 0.027 0.01 1 0.054 0.007 - 0.002 0.01 1 0.013 0.019 0.017 0.009 0.233 0.126 0.01 1 0.002 0.006 0.021 - 0.006 0.004 0.008 0.008 0.004 0.068 0.030 -0.028 0.060 0.047 0.037 -0.037 0.019 0.071 0.060 0.051 -0.012 0.252 0.139 0.010 0.002 0.049 0.038 0.009 0.002 0.006 0.036 -0.036 * The braces denote the simultaneous presence of the indicated substances in the test solutions. t If the other substances had no effect, the results would be expected (95% confidence) to lie within the ranges 0.000 & 0.008, 0.200 f 0.018 and 3.000 f 0.084 mg 1-l, respectively.j Calgon is the registered trade mark of a glassy sodium hexametaphosphate. Discussion The inappreciable contribution of between-batch variation to the precision of the deter- minations clearly demonstrates that the method does not require a separate calibration graph for each batch. Further evidence for the invariance of measurements was provided by examination of the results of the routine analyses performed in the past year. A 0.5 mg 1-1 standard had been included in each set of analyses and for the 105 measurements made the relative standard deviation was 3.2%. This good precision figure must include any effects due to the possible differences in the percentage composition of acids coming from different bottles, and from differing levels of iron impurity in the reagents used.The within-batch precision achieved for the lake water sample is as good as that of the standards, which provides authority to apply precision figures obtained from measurements on standards to routine measurements. The precision of the method ranges from a total relative standard deviation of 4.4% at 0.2 mg 1-1 to 1.5% at 3 mg 1-1 for 19 degrees of freedom. These values correspond well with those found by Dougan and Wilson4 for total iron deter- mination using 2,4,6-tri-(2-pyridyl)-1,3,5-triazine (TPTZ) after a wet digestion. The use of638 DAVISON AND RIGG TPTZ without digestion gave better precision figures, which suggests that contamination introduced in the digestion step was the major source of error.Routine atomic-absorption spectroscopy would be expected to have poorer precision. The within-batch standard devi- ation of the blank can be used to calculate the criterion of detection12 using the expression 2/2ts, where t is Student’s t. A value of 0.009 mg 1-1 is obtained for the 95% confidence level. The mean blank value of 0.019 mg 1-1 can be accounted for to a large extent by contamination introduced by reagents. Using the maximum permissible iron contents quoted by the manu- facturer for AnalaR acids, the expected blank would be 0.024 mg 1-l. Therefore, this source probably makes a considerable and virtually constant contribution to the blank, which helps to rationalise the comparatively low standard deviation of 0.004 mg 1-l.Efforts to reduce the blank level by using purer reagents would not be worthwhile because the expected iron contribution from Aristar acids would still be 0.012 mg l-l, and this at a greatly increased cost. The interference studies show the method to be adequately selective for determinations of iron in freshwater. Copper is the only substance to make a significant contribution to absorbance, but below 0.1 mg 1-1 it can be neglected. Most freshwaters have a much lower copper content and so no problem would occur. If analyses of effluents or mine waters having appreciable copper content were required, a shift to higher wavelengths would further improve the selectivity, but with a slight reduction in the sensitivity of the method.The measured performance characteristics show that the useful working range of the method should be 0.020 - 3 mg l-l, although the precision achieved at this lower figure would be poor and perhaps 0.050 mg 1-1 would be a more realistic limit for routine work. Extension of the range to higher levels can be performed readily by taking a smaller initial volume of water which, if less than 10 ml, eliminates the need for a preliminary evaporation step. Comparison of the within-batch standard deviation of the blank of 0.004 mg 1-1 with that due to the spectrophotometer of 0.002 mg 1-1 shows that a cuvette of increased path length would produce only marginal improvement in precision for the lower concentration samples.Conclusions The method described has a precision that clearly rivals that of other methods used for total iron determination. It is adequately selective for measurements in freshwater and has proved successful in routine analysis over a number of years. Considering that it includes a wet oxidation step it is economical on operator time and it uses only equipment and in- expensive reagents that are readily available. While not advocating that the method replace the usual determinations of iron by complexation of the iron( 11) state, the sensitivity, precision and selectivity demonstrated here show that it should not be precluded as an analytical technique. We thank Dr. P. A. Cranwell and Mr. J. Heron for reading the manuscript and the late Mr. F. J. H. Mackereth for introducing the method into this laboratory and so stimulating this study. We are grateful to BDH Chemicals Ltd. for providing quality control information for their AnalaR hydrochloric acid. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Stumm, W., and Lee, G. F., Schweiz. 2. Hydrol., 1960, 22, 295. American Public Health Association, American Water Works Association and Water Pollution Control Federation, “Standard Methods for the Examination of Water and Wastewater,” Thirteenth Edition, American Public Health Association, New York, 1971. Department of the Environment, “Analysis of Raw, Potable and Waste Waters,” H.M. Stationery Office, London, 1972. Dougan, W. K., and Wilson, A. L., Wat. Treat. Exam., 1973, 22, 100. Huttner, C., 2. Anorg. Chem., 1914, 86, 341. Wolf, R. H. H., and Orhanovic, M., 2. Analyt. Chem., 1966, 216, 405. Standley, C. L., and Kruh, R. F., J. Chem. Phys., 1961, 34, 1450. Glasner, A., and Avinur, P., Talanta, 1964, 11, 761. Desesa, M. A., and Rogers, L. B., Analytica Chim. Acta, 1952, 6, 534. Ishibashi, M., Shigematsu, T., Yamamoto, Y., Tabushi, M., and Kitagawa, T., Bull. Chem. SOC. Wilson, A. L., Talanta, 1970, 17, 31. Wilson, A. L., Talanta, 1973, 20, 725. Japan, 1957, 30, 433. Received February 6th, 1976 Accepted March 17th, 1976

ISSN:0003-2654    

DOI:10.1039/AN9760100634

出版商:RSC    

年代:1976

数据来源: RSC