ISSN:0003-2654    

DOI:10.1039/AN97499FX025

出版商:RSC    

年代:1974

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97499BX027

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

iV THE ANALYST [July, 1974THE ANALYSTEDITORIAL ADVISORY BOARDChairman: H. J. Cluley (Wembley)*L. S. Bark (Solford)R. Belcher (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)'R. M. Dagnall (Hunfingdon)E. A. M. F. Dahmen (The Netherlands)*J. B. Dawson (Leeds)A. C. Docherty (Billingham)D. Dyrssen (Sweden)*W. T. Elwell (Birmingham)*D. C. Garratt (London)J. H a t e (Belgium)D. N. Hume (U.S.A.)H. M. N. H. Irving (Leeds)M. T. Kelley (U.S.A.)*J. A. Hunter (Edinburgh)W. Kemula (Poland)*G. F. Kirkbright (London)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)*J. M. Ottaway (Glasgow)*G. E. Penketh (Billingham)S. A. Price (Tadworth)D. I. Rees (London)E. 6. Sandell (U.S.A.)*R. Sawyer (London)A. A.Smales, O.B.E. (Harwell)H. E. Stagg (Manchester)E. Stahl (Germany)A. Walsh (Australia)T. S. West (London)P. Zuman (U.S.A.)*A. Townshend (Birmingham)* Members of the Board serving on the Executive Committee.NOTICE TO .SUBSCRIBERSSubscriptions for The Analyst, Analytical Abstracts and Proceedings should beThe Chemical Society, Publications Sales Offlce,Blackhorse Road, Letchworth, Herts.Rates for 1974(other than Members of the Society)sent to:(a) The Analyst, Analytical Abstracts, and Proceedings, with indexes . . . . €37.00(b) The Analyst, Analytical Abstracts printed on one side of the paper (withoutindex), and Proceedings . . . . . . . . . . . . . . f38.00(c) The Analyst, Analytical Abstracts printed on one side of the paper (withindex), and Proceedings .. . . . . . . . . . . . . €45.00The Analyst and Analytical Abstracts without Proceedings-(d) The Analyst and Analytical Abstracts, with indexes . . . . . . . . €34.00(e) The Analyst, and Analytical Abstracts printed on one side of the paper (withoutindex) . . . . . . . . . . . . . . . . . . €35.00(f) The Analyst, and Analytical Abstracts printed on one side of the paper (withindex) . . . . . . . . . . . . . . . . . . L42.00(Subscriptions are NOT accepted for The Analyst and/or for Proceedings aloneJuly, 19741 SUMMARIES OF PAPERS I N THIS ISSUE VSummaries of Papers in this IssueGas - Liquid Chromatographic Determination of Vitamin D, inFortified Full-cream Dried MilkA gas - liquid chromatographic method is described for the determinationof vitamin D, (ergocalciferol) in full-cream dried milk that contained 10 pg ofadded vitamin per 100 g.The method involves extraction of the vitaminplus fat with a mixture of ethanol, sodium hexametaphosphate and a surface-active agent; saponification of the fat ; and extraction of the unsaponifiablematter containing the vitamin with a mixture of light petroleum and diethylether. Cholesterol and other interfering substances are removed in a novelway by dry-column chromatography on two columns of aluminium oxide andthe final determination of ergocalciferol is made by gas - liquid chromato-graphy, using vitamin D, (cholecalciferol) as an internal standard.The method, which can be completed in 2 days, is also applicable to driedmilk that has been fortified with cholecalciferol by using ergocalciferol as theinternal standard.Results obtained by the proposed method are given forthe ergocalciferol content of National Dried Milk and of some proprietarydried milk powders.J. G. BELL and A. A. CHRISTIEDepartment of Trade and Industry, Laboratory of the Government Chemist, Corn-wall House, Stamford Street, London, SEl 9NQ.Analyst, 1974, 99, 385-396.A Simplified Colorimetric Method for the Determination of AscorbicAcid in Pure Solutions and in Pharmaceutical PreparationsA precise and specific method for the determination of L-ascorbic acidin the presence of dehydro-L-ascorbic acid and 2, S-diketo-~-gulonic acid hasbeen developed. The method is based on the reaction of phenylhydraziniumchloride with ascorbic acid in 0.1 N hydrochloric acid at a temperatureof 50 f 2 "C.The stable yellow colour produced is measured spectrophoto-metrically at 395nm. The results obtained by this method and the B.P.method are compared. There is no interference from other vitamins, minerals,glucose, sucrose or the most commonly used excipients with this methodand i t has been successfully applied to pharmaceutical preparations. Thesimplicity of the procedure permits rapid analysis, which is necessary forroutine control work.NAG1 WAHBA, DAWOUD A. YASSA and RAMZY S. LABIBBiochemistry Department, Faculty of Medicine, Ain-Shams University, Cairo,Egypt.Analyst, 1974, 99, 397-402.Chemical Interferences in the Determination of Manganese in PlantMaterial by Atomic-absorption SpectroscopyCombinations of calcium $Zus magnesium with sulphate may cause lowresults in the determination of manganese by atomic-absorption spectroscopy.Maximum depression of manganese absorbance is noted when the mole ratio ofcalcium plus magnesium to sulphur is 1 : 1. When plant samples are digestedwith a mixture of nitric and perchloric acids, low results are obtained only forthose materials which are low in calcium plus magnesium and high in sulphur,e.g., grain ; correct values are obtained by incorporating lanthanum chloridein the solution. If sulphuric acid is used in the digestion, very low values areobtained and addition of lanthanum chloride effects" little improvement.E. G. BRADFIELDUniversity of Bristol, Department of Agriculture and Horticulture, Long AshtonResearch Station, Bristol, BS18 9AF.Analyst, 1974, 99, 403-407

ISSN:0003-2654    

DOI:10.1039/AN97499FP073

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

July, 19741 THE ANALYST ixCLASSIFIED ADVERTISEMENTSThe rate for classi$ed advertisments is 35p a line (or spaceequivalent of a line) with a n extra charge of lop for theuse of a Box Number. Semi-displayed classifiedadvertisements are L4 for single-column inch.Copy for classified advertisements required not later thanthe 18th of the month Preceding date of publication whichis on the 16th of each month. Advertisements should beaddressed to J . Arthur Cook, 9 Lloyd Square, London,W C l X 9RA. Tel: 01-837 6315ANALYST Vols. 80-93, ANALYTICAL ABSTRACTSVols. 1-15 Bound in Blue. Offers. Box No. 235 c/oJ. Arthur Cook, 9 Lloyd Square, London WClX 9BA.APPOINTMENTS VACANTANALYTICAL CHEMISTWith an expanding rcsearch programme, the Paint Research Associ-ation has a vacancy for an analyst capable of applying modeminstrumental techniques to the solution of practical problems.Theappointment will be made a t Senior Research Officer level (presentsalary range L2,280-L3,460). Applicants should have a good firstdegree and research or industrial experience with chromatographicand spectroscopic techniques. Please apply, giving details ofqualifications and previous experience to: The Head of ChemistryDivision, The Paint Research Association, Waldegrave Road,Teddington, Middlesex T W l l 8LD.Bureau ofAnalysedSamples Ltd.supply a range of Standard ReferenceMaterials for the MetallurgicalChemist and Spectrographer.Full particulars are given in listNo. 448 which will be supplied freeon request.Newham Hall, Newby,Middlesbrough, Teesside TS8 9EATelephone: 0642 37216Deve I o p me n t An a I ystLondonf 2200-f 2600Our Central Laboratories are situated in Londonwhere they will be moved into new,purpose-built, accommodation before the endof the year.Currently employing some 100 people thelaboratories, under the Director of Researchembrace the fields of Chemical Analysis,Food Technology, Biochemistry, Microbiology,Packaging, Hygiene and Home Economics.A vacancy has arisen for a DevelopmentAnalyst to be involved with the developmentof new analytical methods and theapplication of instrumental techniques to foodanalysis.Applications are invited from those with aminimum of an Honours Degree in Chemistry.Conditions of service are excellent andinclude:a 5-day week, 18 working days annualholiday (21 days after one year's service),a first class staff restaurant, pension and lifeassurance schemes and sports and socialf ac i I it ies.Please write or telephone for an applicationform:M.H. Wells (LB14/DA), J. Sainsbury Ltd,Stamford House, Stamford Street,London SE1 9LL.Tel.: 01 -928 3355, Ext. 2723X THE ANALYST [July, 1974APPOINTMENTS VACANTNATIONAL COAL BOARDScottish AreaIll INSTITUTE OF OCCUPATIONALMEDICINEENVIRONMENTAL BRANCHThe following are vacant posts at the Institute of Occupational Medicine, Edinburgh, for work on research projects supportedby the "Health in Mines" Research Programme of the Commission of the European Communities and the National CoalBoard. Some of these appointments are initially for two or three years but it is anticipated that they may be extended.M5/6 CHEMIST/ANALYSTInstitute of Occupational Medicine EdinburghM5 SALARY SCALE $3255 to $3920M6 SALARY SCALE $2845 x H 0 5 t o S3420Experienced Chemist/Analyst for research on analytical problems concerned with inhaled gases and dusts and to superviseroutine analytical work.Candidates should have a good honours degree or high professional qualifications and post graduateexperience in gas analysis, gas absorption kinetics or related topics and in general inorganic analysis.M6/7 MINING ENGINEER (RESEARCH)Institute of Occupational Medicine EdinburghM6 SALARY SCALE $2845 x $105 t o $3420M7 SALARY SCALE $2345 x 290 t o $2885Mining Engineer (Research) to assist in the planning, organising and analysis of field research in coal mines into the infuenceof mining and geological factors on airborne dust production and other environmental hazards.Should have a goodHonours degree in mining; postgraduate research experience in mine ventilation or rock mechanics would be advantageous.M6/7 MINING INVESTIGATOR(Outstat ioned in Nott ing hamsh i re)INSTITUTE OF OCCUPATIONAL MEDICINEM6 SALARY SCALE $2845 x $105 t o $3420M7 SALARY SCALE $2345 x $90 to $2885Mining Investigator, outstationed in Nottinghamshire, to participate in the field research programme of the Institute ofOccupational Medicine especially studies of airborne dust surveillance procedures and of factors affecting dust formation.Should have a degree or HNC in engineering or science and experience in mining.PHYSICS GRADUATEM7 SALARY SCALE $2345 x $90 to $2885Physics graduate for research on the measurement of airborne particles and related problems. The post will suit a new(Honours) graduate.It may be possible for a suitable candidate to register for a higher degree.Applications in writing to Area Staff ManagerBecretary,NATIONAL GOAL BOARDScottish AreaGreenparkGreenend,EDINBURGH EH17 7Pxii SUMMARIES OF PAPERS IN THIS ISSUEA Method for the Determination of Copper in Biological MaterialsInvolving the Use of Sodium DiethyldithiocarbamateThe method described for determining copper is based on a complexationreaction effected in acetic acid solution.The acetic acid not only effects com-plexation but also dissolves the complex and no intermediate extraction istherefore necessary. The sensitivity is higher than that for other methodsbased on diethyldithiocarbamate. Various metal ions interfere in the deter-mination of copper using diethyldithiocarbamate and the elimination of theseinterferences is discussed.The application of the method to the determination of trace amounts ofcopper in various biological materials, particularly foodstuffs, is discussed.M. E. M. S . de SILVAGovernment Analyst's Laboratory, Colombo, Sri Lanka.[July, 1974Analyst, 1974, 99, 408-412.The Absorptiometric Determination of Aluminium in Water.A Comparison of Some Chromogenic Reagents and the Developmentof an Improved MethodThe properties of chromogenic reagents used for the absorptiometricdetermination of aluminium in water are compared, and an experimentalcomparison of catechol violet, Eriochrome cyanine R and stilbazo has beenmade.Catechol violet is considered to be the most suitable and a methodinvolving the use of this reagent has been developed. When using this method,the standard deviation of analytical results varied from 0.004 to 0.008 mg 1-1of aluminium for concentrations of 0.05 and 0.3 mg 1-1 of aluminium, respec-tively. With the exception of fluoride, substances normally present in treatedwaters did not cause important interference. The effect of fluoride (1 mg 1-1or less) is tolerable for most purposes, but appreciably greater concentrationsrequire preliminary removal of fluoride or correction for its effect.Satis-factory results have been obtained with the method in six other laboratories.The method is simple and rapid ; ten samples can be analysed in approximately16 hours. The method has advantages over other commonly used methods,and is recommended for use in water analysis laboratories.W. K. DOUGAN and A. L. WILSONWater Research Centre, Medmenham Laboratory, Medmenham, Marlow, Buckingham-shire, SL7 2HD.Analyst, 1974, 99, 413-430.Spectrophotometric Determinatioa of Organophosphorus Pesticideswith 4- (4-Nitrobenzy1)pyridineA spectrophotometric method has been developed for the determinationof certain organophosphorus pesticides. It involves the reaction of 4- (4-nitro-benzy1)pyridine with the pesticide in an acetone - water - ethanol mixturea t 100 "C to produce the dye precursor.The colour is produced after theaddition of tetraethylenepentamine.C. R. TURNERCentre for Overseas Pest Research, Csllege House, Wrights Lane, London, W8 6s J.Analyst, 1974, 99, 431-434July, 19741 THE ANALYST xiiicrRing 0843 24261 Ext. 28'5for a 48 hourdelivery serviceItBOOKSMONOGRAPHSREPRINTSReliabilityManufactured on an advanced produc-tion plant from materials carefullyselected for their high purity. Duringprocessing lamps are baked at a pres-sure of 1094 Torr to provide maxi-mum outgassing giving rise to lowbackground and clean line spectrum.UniversalDesigned to fit not only Rank HilgerH1170 and H1550 but almost everyAtomic Absorption Spectrophoto-meter , being suitable for 3 modeoperation: D.C..Pulsed or Modulated.Being at the lower end of the price7Valueorders for all publications ofthe Society (except journals)should be sent direct or througha bookseller to-THE SOCIETY FORANALYTICAL CHEMISTRYBook Department9/10 Savile Row,London, W I X IAFRANK HILGERHOLLOWCATHODELong LifeReliabi1.t yCompletely sealed lamp with windowfused to the glass envelope, preciselyfilled to the appropriate gas pressure.E!! HE:ggZmps achieve strong emis-SELECTED ANNUAL REVIEWSof theANALYTICAL SCIENCESVolume 2 - 1972CONTENTSThe Techniques and Theory of Thermal Analysis Applied to Studies on InorganicMaterials with Particular Reference to Dehydration and Single Oxide Systems -D.DollimoreDevelopments in Ion Exchange - F. VernonThermometric and Enthalpimetric Titrimetry - L. S. Bark, P. Bate and J. K. GrimeObtainable from-Pp. vi + 149 f5.00; U.S. $13.00 ISBN 0 85990 202 1The Society for Analytical Chemistry, Book Department,9/10 Savile Row, London W I X I A FMembers of The Chemical Society may buy personal copies a t the special price of f3.00; U.S. $8.0SUMMARIES OF PAPERS I N THIS ISSUEEnthalpimetric Determination of Hydroxyl Values of Glycerol -Alkylene Oxide Polyethers and Butane- 1,4-diol - Adipic AcidPolyesters[July, 1974Direct-injection enthalpimetry has been used to measure the hydroxylvalues of some polyethers and polyesters.A determination in duplicate canbe carried out in a few minutes, and it requires relatively little operationalskill. The technique clearly has great potential as a method for industrialplant control. The accuracy of determinations is f 1 to 2 per cent., which isadequate for commercial purposes, although the method is not as precise assome more lengthy procedures.I. I. KADUJI and J. H. REESLankro Chemicals Limited, Eccles, Manchester, M30 OBH.Analyst, 1974, 99, 435-438.Polarographic Determination of Thallium, Palladium, Nickel, Zincand Indium with DiethylenetriamineThe polarographic determination of the metal ions thallium(1) , palladium-(11), nickel(II), zinc(I1) and indium(II1) is carried out in a 0.4 M solution ofdiethylenetriamine. Well defined and irreversible waves are obtained, whichare suitable for the determination of small amounts (1 to 7 mg) of thesemetals.The method gives accurate and reproducible results.A. L. J. RAO and ASHOK KUMARDepartment of Chemistry, Punjabi University, Patiala, India.Analyst, 1974, 99, 439-441.The Zone Electrophoresis of Lubricating Oil AdditivesElectrophoresis can provide useful separations of lubricating oil additives.The feasibility studies were based on thin-layer electrophoresis with silicagel G plates in non-aqueous (mainly pyridine - acetic acid) and mixed aqueoussolvent (ethanol - water) buffer media as well as with agarose and polyacryl-amide gel support systems.The migration behaviour of various polymer, sulphonate, salicylate,phenate and dialkyldithiophosphate types of additives is discussed. Generally,the dialkyldithiophosphate additive anions showed the greatest mobilities,followed by the sulphonates, salicylates and phenates, while the polymer-typereaction products of polyisobutenylacrylic acid or polyisobutenylsuccinicanhydride with tetraethylenepentamine were immobile.Thin-layer electrophoresis based on ethanol - water - boric acid - sodiumacetate buffer systems, coupled with a chromatographic stage in the seconddimension, merits further attention. Gel electrophoresis is also promisingand the disc techniques might be developed as a finger-printing method forthe characterisation of additives.D. LEIGHTON, G. J. MOODY and J. D. R. THOMASChemistry Department, University of Wales Institute of Science and Technology,Cardiff, CF1 3NU, Wales.Analyst, 1974, 99, 442-452

ISSN:0003-2654    

DOI:10.1039/AN97499BP079

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

JULY, 1974 Vol. 99, No. 11 80 THE ANALYST Gas - Liquid Chromatographic Determination of Vitamin D2 in Fortified Full-cream Dried Milk BY J. G. BELL AND A. A. CHRISTIE (Departntent of Trade and Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SE 1 9NQ) A gas - liquid chromatographic method is described for the determination o f vitamin D, (ergocalciferol) in full-cream dried milk that contained 10 pg of added vitamin per 100 g. The method involves extraction of the vitamin plus fat with a mixture of ethanol, sodium hexametaphosphate and a surface- active agent; saponification of the fat; and extraction of the unsaponifiable matter containing the vitamin with a mixture of light petroleum and diethyl ether. Cholesterol and other interfering substances are removed in a novel way by dry-column Chromatography on two columns of aluminium oxide and the final determination of ergocalciferol is made by gas - liquid chroma- tography, using vitamin D, (cholecalciferol) as an internal standard.The method, which can be completed in 2 days, is also applicable to dried milk that has been fortified with cholecalciferol by using ergocalciferol as the internal standard. Results obtained by the proposed method are given for the ergocalciferol content of National Dried Milk and of some proprietary dried milk powders. THE biological assay methods of determining vitamin D involving the use of rats or chickens are time consuming, expensive and imprecise but are still the only methods that are sensitive enough to cope with the low levels present in unfortified food.For many years there has been a need for a satisfactory chemical method and various chemical and physico- chemical methods have been proposed, including colorimetry, ultraviolet spectroscopy and infrared spectrophotometry, but none of these methods has been found to be entirely satisfactory even for fortified foods. However, in 1960, the application of gas - liquid chromatography to the determination of vitamin D gave the first indication of the possibilities of this technique1 and, in 1966, a major development occurred when Murray, Day and Kodicek2 showed that, by converting ergocalciferol and cholecalciferol into their isovitamins, only one peak instead of two was obtained for each vitamin on the chromatogram.Since then, there has been an increasing interest in the use of gas - liquid chromatography for the deter- mination of vitamin D in food, pharmaceutical and natural products. Avioli and Lee3 investigated the determination of vitamin D in plasma by gas - liquid chromatography and claimed to detect as little as 5 ng of the vitamin. In 1968, a procedure was developed for the assay of vitamin D in multivitamin preparations by this method with a detection limit of 0.2 Sheppard, LaCroix and Prosser5 isomerised pure specimens of ergocalciferol and cholecalciferol with acetyl chloride and 1,2-dichloroethane before measuring their responses on a gas chromatograph, but this technique has not been applied to natural products. In 1969, the use of electron-capture detectors instead of flame-ionisation detectors was reported by Wilson, Lawson and KodicekG who, by forming the heptafluorobutyrate esters of the vitamins, were able to detect down to long of the isovitamins.During 1970 and 1971, several papers appeared on the determination of vitamin D in such products as fortified non-fat dried milk,’ multivitamin tablets8 and ultraviolet irradiated milk,g using gas - liquid chromatographs fitted with flame-ionisation detectors. This laboratory needed a method to check the vitamin D content of fortified full-cream roller-dried milk (National Dried Milk) and of vitamin tablets and preparations that contained vitamins A, C and D. In a recent paper by the present authors,1° the gas - liquid chromato- graphic determination of vitamin D in cod-liver oil was described. The present paper describes the development of a method for determining vitamin D in National Dried Milk by extracting the vitamin with the fat; saponification of the fat and extraction of the un- saponifiable matter; removal of retinol (vitamin A) and sterols in two stages by dry-column 385 @ SAC; Crown Copyright Reserved.386 BELL AND CHRISTIE : GAS - LIQUID CHROMATOGRAPHIC DETERMINATION [A?Za&St, VOl.99 chromatography on aluminium oxide; formation of the trimethylsilyl derivatives of the iso- vitamins; and gas - liquid chromatography with flame-ionisation detection. The aim has been to develop a method that is capable of giving a reliable result within 2 days and that can be used with confidence by less experienced staff.EXPERIMENTAL EXTRACTION AND SAPONIFICATION OF THE LIPIDS FROM DRIED MILK- The conventional procedure for isolating an oil-soluble vitamin, such as vitamin D, is to saponify the material with ethanolic potassium hydroxide solution and to extract the unsaponifiable matter with solvents. For dried milk containing 10 pg of vitamin D per 100 g, saponification of at least 100 g of dried milk was required in order to obtain sufficient vitamin D for quantitative measurement. The saponification of more than a few grams of dried milk is, however, a tedious and cumbersome procedure, so it was decided to extract the lipid fraction containing vitamin D from the dried milk before saponification. Attempts to separate the lipids from the milk by solvent extraction were unsuccessful but a reagent that allows the release of the lipid fraction and prevents the precipitation of calcium salts was developed and found to be most satisfactory.This reagent, which consisted of a mixture of ethanol, a surface-active agent and sodium hexametaphosphate, allowed the lipid fraction containing the vitamin to be extracted from the milk in a single stage with a mixture of equal volumes of light petroleum and diethyl ether. I t was also found that, after removal of the solvents, gently heating the lipid residue with ethanolic potassium hydroxide solution in a water-bath at 70 to 80 "C for 15 minutes allowed it to be readily saponified with the minimum destruction of vitamin D. DRY-COLUMN CHROMATOGRAPHY- The unsaponifiable matter contains, in addition to vitamin D, various substances such as retinol and sterols, which cause interference in the determination of the vitamin by gas - liquid chromatography.It is therefore essential to remove such substances and many procedures have been suggested for this purpose. Adsorption columns of aluminium oxide, magnesium oxide or silicic acid are commonly employed and, more recently, Celite impreg- nated with poly(ethy1ene g1ycol)ll and Fluoropak 8012 have also been used. Sterols are often removed by precipitation with digitonin , and thin-layer chromatography has been proposed by several workers as an additional clean-up procedure. The use of a combination of column chromatography, thin-layer chromatography and selective precipitation of the sterols is not uncommon but even with such an elaborate clean-up procedure the gas chromatogram shows that some interfering substances remain.The procedure developed by Bell and Christielo for the clean-up of the unsaponifiable matter of cod-liver oil involved partition chromatography on Sephadex LH20, precipitation of sterols with digitonin and adsorption chromatography on a Florisil column. The technique was well suited to the analysis of cod-liver oil containing about 100 pg of vitamin D in 50 g, as, in spite of losses at some stages of the procedure, sufficient of the vitamin remained to provide a satisfactory chromatogram. However, when the technique was applied to dried milk containing 1Opg of vitamin D per lOOg, losses were excessive and the amount of vitamin D that remained was insufficient for quantitative measurements.A method for the removal of interfering substances, particularly the sterols, which are especially troublesome during gas - liquid chromatography, was therefore developed. Dry-column chromatography, first described in two papers by Loev and co-workers ,13J4 is an improved chromatographic technique by means of which separations comparable with those obtained by thin-layer chromatography can be carried out rapidly in a column on a preparative scale. This technique was successfully applied by Dead5 to the separation of dimeric tocopherol products from vitamin E and appeared to be worth investigating as a means of separating vitamin D from sterols. First, it was shown that vitamin D and cholesterol could be cleanly separated in pure solution and in extracts of dried milk on thin-layer plates of silica gel.In carrying out this separation by dry-column chromatography in order to obtain sufficient vitamin D for its detection on the gas chromatograph, the important observation was made that by converting the vitamin into its isovitamin form complete separation from cholesterol could be achieved (Fig. 1).July, 19741 OF VITAMIN D2 I N FULL-CREAM DRIED MILK 387 Various aspects of this separation were investigated, including the length and diameter of the column, the type and activity of the adsorbent and the choice of eluting agent for development of the chromatogram. Fig. 2 illustrates the effect of the activity of the silica gel on the eluting characteristics of vitamin D and isovitamin D; Fig.3 gives similar informa- tion for aluminium oxide. The system that was finally selected for the isolation of vitamin D involved the use of two glass columns (40 x 1 cm), each containing aluminium oxide de- activated with 8 per cent. of water, as the adsorbent, with chloroform as the eluting agent. The first dry column separated a fraction from the unsaponifiable matter of dried milk containing vitamin D, retinol, cholesterol and other sterols. This fraction was then treated with antimony trichloride in order to convert vitamin D into isovitamin D before application to the second dry column of aluminium oxide. Vitamin A, as its anhydro derivative, eluted with the solvent front and was followed by isovitamin D, well separated from cholesterol and other sterols. In this way, isovitamin D was substantially freed from sterols without recourse to precipitation with digitonin, and an extract was obtained that was suitable for the preparation of the trimethylsilyl ethers of the vitamins and their subsequent determination by gas - liquid chromatography.0.9 08 0.7 2 0.6 .fl 0.5 a 0.4 0.3 0.2 0.1 ru a a - - - - - - - - - - 1 10 20 30 40 50 60 70 80 90 100 Volume of eluate/m I Fig. 1. Separation of isovitamin D and cholesterol by dry-column chromatography : peak 1, isovitamin D; peak 2, vitamin D; and peak 3, cholesterol. Column size, 1 x 30 cm; material, silica gel; and eluting agent, chloroform Some modifications of the procedure of dry-column chromatography described by Loev and c o - w o r k e r ~ ~ ~ ~ 1 ~ were found to be advantageous. Instead of stopping the chromatogram when the solvent front reached the bottom of the column, the elution was allowed to continue until the required fraction was obtained.The technique was also greatly facilitated by the use of two marker dyes, 2,4-diaminoazobenzene and aminoazobenzene, which have the same eluting characteristics as those of vitamin D and isovitamin D, respectively. FORMATION OF TRIMETHYLSILYL ETHERS- Murray, Day and Kodicek2 showed that the conversion of ergocalciferol and chole- calciferol into their isovitamins by treatment with antimony trichloride solution produced a single peak for each vitamin on the gas chromatogram instead of two for each of the un- modified vitamins. Their technique, which has the advantages of distinguishing clearly between ergocalciferol and cholecalciferol and of allowing one vitamin to be used as internal standard for the other, has been further improved.Formation of the trimethylsilyl ethers of the isovitamins before injection on to the gas chromatograph gave derivatives with shorter retention times and eliminated the tailing effect on the peaks. The reagents (hexamethyl- disilazane and trimethylchlorosilane) that were initially available for formation of these derivatives had the disadvantages that they required a reaction medium such as pyridine, which resulted in a large solvent front on the chromatogram, and produced inconvenient solid reaction products.388 BELL AND CHRISTIE : GAS - LIQUID CHROMATOGRAPHIC DETERMINATION [Aaalyst, Vol. 99 Developments in silylation techniques have led to the production of some improved reagents, one of which, NO-bis(trimethylsilyl)acetamide, has been found to be effective and easy to use without the need of an additional reaction medium or the production of insoluble reaction products.This compound has therefore been used throughout this work for the preparation of trimethylsilyl ether derivatives. 0.5 a) g s o - ' .-----: ---- i ---------------- dJ ............ ...i.. ............................................... ........... 0 - .-*. . . ,'*'I : '. ; \ ; ; \ ; I I \ : : \ * . (6) I \ . \ : \ : \: c 3 I I - I \ i : .................................................... .......................... I I / ----- ,--,-,/ .' \,,,,A ---------- I I ' " I I 8 1.0 2 0.5 .*..: -. : : r'\ - . (Cl f ' . r ' : : : I \ ; \ I \ ; I \ - \ ; \ . \ : \ I I - \ . \: I I I / 5 ----- &- ---- -----. ,,,,,-----) ........................................... ............................. Y 0.5 .*'. : : : : /-': : : (d) I' \ I I \ i { 1 ; i \ - . \ : : \ : : - I I \ i I \ * I \ i -,--,,c' %---.Ac--,- -----------. ...................................... .................................... I I I I I Volume of eluate/ml Fig. 2. Effect of activity of silica gel on the eluting charac- teristics of vitamin D (dotted line) and isovitamin I) (broken line). Water added to silica gel, per cent.: (a), 14; (b), 16; (c), 18; and (d), 20. Column size, 1 x 30 cm; and eluting agent, chloroform CONDITIONS FOR GAS - LIQUID CHROMATOGRAPHY- In recent years, several high-purity, thermally stable stationary phases, which are far superior to the silicone oils previously used, have become available for gas - liquid chromato- graphy.These substances are silicones with polarities that range from the non-polar OV-1July, 19741 OF VITAMIN D, I N FULL-CREAM DRIED MILK 389 - - 1 I 0.5 0 I " 0.5 0 1.0 I 0 20 40 60 80 100 120 Volume of eluate/ml Fig. 3. Effect of activity of aluminium oxide on the eluting characteristics of vitamin D (dotted line) and isovitamin D (broken line). Water added to aluminium oxide, per cent.: (a), 6 ; (b), 8; and (G), 10. Column size, 1.2 x 25 cm; and eluting agent, chloroform through the slightly polar OV-17 to the most polar phenylmethylsilicone OV-25 and when they are applied as 3 per cent. coatings on a specially treated and silanised diatomaceous earth (Gas-Chrom Q), they provide excellent stationary phases for the gas - liquid chromato- graphy of steroids and related substances, such as ergocalciferol and cholecalciferol.The choice of stationary phase depends on the nature of the separation required and on the various conditions of gas - liquid chromatography such as column length, oven temperature and the rate of flow of the carrier gas. An example of the effect of an increase in polarity of the stationary phase on the retention time and resolution of the trimethylsilyl ethers of cholesterol and the isovitamins of ergo- calciferol and cholecalciferol is given in Fig. 4, from which it can be seen that OV-17 provided the best resolution with the minimum distortion of the peaks.Some workers have suggested that temperature programming of the chromatograph offers some advantages but in our experience no significant improvement in the resolution of the peaks was obtained by this technique and reproducibility was difficult to achieve. Isothermal conditions were therefore used throughout this work. INTERNAL STANDARD FOR GAS - LIQUID CHROMATOGRAPHY- One of the advantages claimed by Murray, Day and Kodicek2 for chromatographing the isovitamins of ergocalciferol and cholecalciferol, instead of the unmodified vitamins, was390 BELL AND CHRISTIE : GAS - LIQUID CHROMATOGRAPHIC DETERMINATIOK [A?ZdySt, VOl. 99 that one vitamin could be used as an internal standard for the other, thus accounting for the losses of vitamin that occur throughout the procedure.However, Sheppard, LaCroix and Prosser5 found that the isotachysterols of ergocalciferol and cholecalciferol, prepared by treatment of the vitamins with acetyl chloride, exhibited different gas - liquid chromato- graphic responses and they concluded that to use one compound as an internal standard for the other would lead to errors of accuracy and precision. 0.11 I ji I n 1 0 50 60 30 40 10 20 0 Retention timefminutes Fig. 4. Effect of polarity of stationary phase on retention time and resolution of the trimethylsilyl ethers of cholesterol, ergo- calciferol isovitamin and cholecalciferol isovitamin by gas - liquid chromatography. Peak 1, 5cc-cholestane ; peak 2, cholesterol tri- methylsilyl ether ; peak 3, cholecalciferol isovitamin trimethylsilyl ether ; and peak 4, ergocalciferol isovitamin trimethylsilyl ether.Column size, 2.1 m x 6 mm; column packing (on Gas-Chrom Q): (a), 3% OV-1; (b), 3% OV-61; (G), 3% OV-17; and (a), 3% OV-25; and oven temperature, 230 "CJuly, 19741 OF VITAMIN D, I N FULL-CREAM DRIED MILK 391 In our laboratory the linearity of the detector response is determined by injecting known amounts of ergocalciferol and cholecalciferol as trimethylsilyl derivatives of the isovitamins and measuring the areas of the resulting peaks with a disc integrator. Fig. 5 illustrates that there is a slight difference in the response of the ergocalciferol and cholecalciferol derivatives in the range 0 to 6 pg, which should be taken into account in the calculation. Our procedure has therefore been to calibrate, at regular intervals, the detector response of the gas chromato- graph by using known amounts of ergocalciferol and cholecalciferol and to determine the relationship of the peak areas for unit masses of the vitamins.The ratio of the peak areas is then applied in the calculation to correct for the slightly different detector responses of the two vitamins, thus allowing one vitamin to be used as an internal standard for the other. This factor (R) is of the order 1.02 to 1.09 when cholecalciferol has been added as an internal standard and the reciprocal of this value when ergocalciferol is the internal standard. Ergocalciferol and cholecalciferoVpg Fig. 5. Calibration of the gas - liquid chromatograph. Detector response of trimethylsilyl ethers of ergocalciferol isovitamin (B) and cholecalciferol isovitamin (A) PROCEDURE REAGENTS AND MATERIALS- All reagents should be of analytical-reagent quality unless otherwise stated.Ethanol, absolute. Sodium hexametaphosphate, pure. Triton X-100. Diethyl ether. Light petroleum (boiling range 40 to 60 "C)-Distol or gas - liquid chromatographic grade. Potassium hydroxide, pellets. Ergocalciferol (vitamin D2), pure, 40 x lo6 i.u. g l . Cholecalciferol (vitamin D3), pure, 40 x lo6 i.u. gl. Chloroform-Distol or gas - liquid chromatographic grade. Antimony trichloride solution, 25 per cent. m/V in chloroform. Tartaric acid solutiou, 40 per cent. m/V. 5~~Cholestane. NO-Bis (trimethylsilyl) acetamide-Supplied by Phase Separations, Queensferry, Flintshire. Aminoazobenzene solution, 1 mg ml-1 an chloroform.2,4-Diaminoazo benzene solution, 1 mg ml-1 in chloro form. Alumiizium oxide, gae.uti/al, Byockmann grade I-Supplied by Woelm, Eschwege, Germany.392 BELL AND CHRISTIE : GAS - LIQUID CHROMATOGRAPHIC DETERMINATION [Analyst, Vol. 99 Deactivate the aluminium oxide with 8 per cent. of water as follows. From a sealed 1000-g canister of the aluminium oxide transfer 500 g of the material into a similar, empty canister. Add, by pipette, to tlie aluminium oxide in each canister 40ml of distilled water and im- mediately screw the caps on tightly. Shake the canisters vigorously for at least 5 minutes, during which time they will become hot. Place them in cold water and when they have reached room temperature mix the contents again by placing the canisters on a roller mixer for 2 to 3 hours.Allow the canisters to stand overnight, then combine the contents of the two canisters and mix again for 1 hour. The aluminium oxide has now been deactivated with 8 per cent. of water and is ready for use. Provided that the material is stored in a screw-capped canister it does not need to be renewed for at least 6 months. F;nt extraction solution-Weigh 200 g of sodium hexametaphosphate into a 1-litre beaker and 100 g of Triton X-100 in a second 1-litre beaker. Add 800 ml of water to eachbeaker and dissolve the solids by warming. Mix the solutions in a suitable bottle and dilute the volume to 2 litres with water. Shake the solution well and allow it to stand overnight; shake it well the following day before use.APPARATUS- water at its boiling-point and to accommodate four 1-litre conical flasks. not require the use of grease on the rotating parts is suitable. Water-bath-This bath should be of sufficient size and heating capacity to maintain the Rotary film evaporator-A Buchii, Model R/B, type was used but any type that does Hand vibrator. Glass columns for clzrowzatogra$hy-These columns, of dimensions 1 x 40 cm, were fitted with a sintered-glass disc of porosity 1 at the bottom and a Quickfit B19 socket at the top. Dry block heater-Tecam Dri-block. Gas - liquid chromatograph-The Pye 104 series, Model 64, fitted with 2-1 m x 3 mm i.d. glass columns packed with 3 per cent. OV-17 on Gas-Chrom Q, 100 to 120 mesh, was used. Other models fitted with a flame-ionisation detector can be used if they are capable of accommodating glass columns and of providing comparable sensitivity.Recorder and integrator-Honeywell Electronik 194, 1-mV input, fitted with a disc integrator. All operations should be carried out in artificial light. EXTRACTION OF FAT AND VITAMIN D- Weigh 100 g of dried milk into a 1-litre conical flask, add 250 ml of water at a temperature of 80 to 85 "C and mix by swirling the flask until a creamy paste free from lumps is formed. Place the flask in a boiling water bath and, after 5 minutes, add 100 ml of fat extraction solution. Mix well and leave the flask in the water-bath for a further 10 minutes with occasional swirling. Cool the solution for 5 minutes by placing the flask in an ice-cooled water-bath.Add 200 snl of ethanol, mix well and continue cooling until the solution reaches at least room temperature. Transfer the contents of the flask into a 1-litre conical separating funnel fitted with a PTFE key and add 400 ml of a 1 + 1 mixture of diethyl ether and light petroleum. Shake the mixture vigorously for 1 minute and leave it to stand for at least 15 minutes so as to allow the two phases to separate. Discard the aqueous layer. Wash the sides of the separating funnel with a few millilitres of water, swirl it very gently and when the phases have separated, discard the aqueous layer. Transfer the solution into a rotary evaporating flask of suitable capacity and reduce the volume of the solution to about half. Add 50 ml of absolute ethanol to the flask and continue the evaporation until the removal of solvent is complete, i.e., until the residue suddenly becomes clear.SAPONIFICATION AND EXTRACTION OF THE UNSAPONIFIABLE MATTER- Weigh 20 g of potassium hydroxide pellets into a 250-ml conical flask. Add about 10 ml of water followed by 140 ml of absolute ethanol, shake the mixture and heat it at 70 to 80 "C until the solid is completely dissolved. Pour the hot ethanolic potassium hydroxide solution into the flask containing the fat extracted from the milk and swirl the flask gently. Dissolve tlie fat, pour the solution back into the 250-ml conical flask, and, as internal standard, add 10 In1 of absolute ethanol containing 10 pg of cholecalciferol. Heat the flask in a water- bath at 70 to 80 "C for 15 minutes, then add 50 ml of water and cool the flask quickly to room temperature.July, 19741 OF VITAMIN D2 I N FULL-CREAM DRIED MILK 393 Transfer the contents into a 1-litre conical separating funnel containing 400ml of a 1 + 1 mixture of diethyl ether and light petroleum, rinsing the flask by filling it with water and adding the rinsings to the separating funnel.Shake the funnel vigorously for 1 minute and allow the phases to separate completely, usually for 10 minutes. Discard the aqueous phase and wash the solvent phase with four 250-ml portions of water, taking care to avoid the formation of emulsions, which may occur if the funnel is shaken too vigorously; it is therefore advisable to increase the thoroughness of shaking from a very gentle action in the first wash to a vigorous action in the last wash.If emulsions do form, they can be dispersed by the addition of a few millilitres of ethanol. Transfer the solvent extract into an evaporating flask and reduce the volunie of the liquid to about half, then add 50 ml of absolute ethanol and continue the evaporation until the removal of the solvent is complete. Transfer the residue of unsaponifiable matter containing the vitamin D into a small evaporating flask with absolute ethanol and evaporate to dryness. Finally, add a few millilitres of chloroform, evaporate again to dryness and dissolve the residue in 2 ~ n l of chloroform. DRY-COLUMN CHROMATOGRAPHY OF THE UNSAPOSIFIABLE MATTER- Set up a chromatographic column in a vertical position with an electrical vibrator tliat just touches its side.Weigh 35 g of the deactivated aluminium oxide into a 100-ml conical flask and, while the vibrator is operating, pour the material through a funnel into the column. As soon as all the aluminium oxide has passed into the column, switch off the vibrator. The height of the aluminium oxide should be about 2 to 3 cm from the lower edge of the socket but, if not, it should be adjusted to this height by modifying the amount of aluminium oxide taken. The column is now ready for use. To the flask containing the solution of unsaponifiable matter in chloroform, add 2 g of deactivated aluminium oxide and 100 p1 of diaminoazobenzene solution as a marker for column chromatography. Adsorb the unsaponifiable matter and the dye on the aluminium oxide by carefully removing the solvent on a rotary evaporator, continuing the evaporation until the aluminium oxide is dry and free flowing.Carefully pour the aluminium oxide on to the top of the prepared column, disregarding the small amount of material that may adhere to the sides of the flask, then tap the column gently to settle the added aluminium oxide and cover it with a thin layer of glass beads (80 mesh). By using a Pasteur pipette, wet the beads with chloroform, taking care to avoid the entrapment of air bubbles, and then add sufficient chloroform to provide a head of 2 to 3 cm. Maintain this level of liquid by any convenient method, for example, by allowing chloroform to drip into it from a stoppered separating funnel. Allow the eluate to run to waste until the yellow band containing the dye is within 1 cm of the sinter at the bottom of the column, then wash the outlet stem of tlie column with chloroform.By inspection in ultraviolet light, tlie fluorescence of retinol will be observed just above the dye. Collect the eluate until all of the retinol has just passed through the column. The volume collected should be about 5 ml and with this size of column will begin to elute after about 40 ml of liquid have passed through the column. Evaporate the collected fraction to dryness on a rotary evaporator and dissolve the residue in 1 ml of chloroform. This fraction contains ergocalciferol, cholecalciferol, retinol, sterols and the marker dye. PREPARATION OF THE ISOVITAMINS- To the extract of the vitamins collected from the first dry column, add 4 ml of antimony trichloride solution, shake the mixture and allow it to stand for exactly 1 minute.Add 6 ml of tartaric acid solution, shake the mixture thoroughly and transfer it into a 25-ml separating funnel. Rinse the flask with 10 ml of light petroleum, add the rinsings to the separating funnel and shake it well. Allow the phases to separate and discard the coloured aqueous phase. \Vasli the light petroleum layer three times with an equal volume of water, discarding the washings, then dry the stem of the separating funnel with filter-paper and filter the light petroleum through a small filter-paper into a rotary evaporating flask. Evaporate the solution to dryness and clissolve the residue in 2 ml of chloroform. DRY-COLUXN CHROMATOGRAPHY FOR REMOVAL OF RETIKOL AND STEROLS- Prepare another dry column of aluminium oxide as described previously. Add 1OOpl of the aminoazobenzene solution as a marker to the residue containing the isomerised vitamins394 BELL AND CHRISTIE GAS - LIQUID CHROMATOGRXPHIC DETERMINATION [A?Zh!bSt, VOl.99 and transfer the mixture on to the top of the prepared column in the same way as before. Develop the chromatogram with chlorofonn and when the leading edge of the dye reaches the bottom of the column, usually after 20 ml have eluted, collect 15 ml of eluate. Remove the solvent on a rotary evaporator and transfer the residue, by extracting it into small portions of light petroleum, into a small specimen tube (46 x 13 mm). Place the tube on a dry-block heater set at approximately 50 "C and evaporate off the solvent in a stream of nitrogen.The residue contains ergocalciferol and cholecalciferol and is substantially free from other vitamins and sterols. FORMATION OF TRIMETHYLSILYL ETHERS AND GAS - LIQUID CHROMATOGRAPHIC PROCEDURE- To the residue in the specimen tube add 100 pl of NO-bis(trimethylsilyl)acetamide, stopper the tube and allow the mixture to stand at room temperature for 10 to 15 minutes. Add loop1 of light petroleum and mix well. The solution is now ready for injection into the chromatograph. Draw 2 p1 of light petroleum containing 10 pg ml-l of 5cc-cholestane (marker) into a 5-p1 syringe, remove the needle of the syringe from the solvent and draw in 1 pl of air. Now introduce 2 pl of sample solution into the syringe and draw the plunger further back so that the sample is contained within the barrel of the syringe.Discharge the contents of the syringe into the gas chromatograph, allowing the solvent to flush out the last traces of sample from the needle. The operating conditions of the gas chromatograph were : column oven temperature 235 "C; detector oven temperature 260 "C; injection block temperature 300 "C; carrier gas, nitrogen at the rate of 50 ml min-l, hydrogen at the rate of 50 ml min-l and air at the rate of 500 ml min-1; and amplifier attenuator x 10 = full-scale deflection 1 x 10-11 A. CALIBRATION OF THE DETECTOR RESPONSE OF THE GAS CHROMATOGRAPH- Inject into the gas chromatograph 0, 1, 3, 4, 5 and 6pg of ergocalciferol and of chole- calciferol, as trimethylsilyl ethers of the isovitamins, and measure the areas of the peaks obtained.Construct calibration graphs for both compounds relating mass and peak areas and confirm that a linear relationship exists between these parameters for each vitamin (Fig. 5). From the same data derive a factor (R) equivalent to the ratio of the peak area of cholecalciferol to that of ergocalciferol for equal masses of these substances and check that this factor is a constant that is independent of the mass of the compound injected, as shown in Table I. Use R for correcting the slightly different detector responses of the two vitamins in the calculation of the result as follows. Ergocalciferol (vitamin D2)/pg g-l = Peak area of ergocalciferol derivative x cholecalciferol added (pg) x R Peak area of cholecalciferol derivative x mass of sample (g) where Peak area of unit mass of cholecalciferol Peak area of unit mass of ergocalciferol R = TABLE I DERIVATION OF A FACTOR FOR RELATING THE DETECTOR RESPONSES OF ERGOCALCIFEROL (D2) AND CHOLECALCIFEROL (D3) ON THE GAS - LIQUID CHROMATOGRAPH Trimethylsilyl ethers Peak areas Peak areas per of isovitamins of of microgram of Factor R- - & (pA-, Peak area per microgram of D, D,//lg D,/M D2 =, D2 Da Peak area per microgram of D, 1.0 0-5 1.0 1.0 1.0 1.0 1.25 1.50 1-75 1.0 1.0 0.5 1.25 1.50 1.75 1.0 1.0 1.0 665 334 676 765 615 675 840 1010 1180 705 718 363 1025 1000 1255 720 715 717 665 668 676 765 615 675 672 673 674 705 718 726 820 667 717 720 715 717 1.06 1.08 1-07 1-07 1.08 1.06 1.07 1-06 1.06July, 19741 OF VITAMIN D2 I N FULL-CREAM DRIED MILK 395 RESULTS Samples of full-cream, roller-dried milk (National Dried Milk) were examined at intervals over a period of several months for their content of added ergocalciferol. Table I1 shows the results obtained for six samples, each of which was examined in duplicate on the same occasion and a further set of six samples, each of which was examined in triplicate on separate occasions.All the results conform to the specification for National Dried Milk, which requires that the vitamin D content should be between 0-044 and 0-132 pg g-l. TABLE I1 ERGOCALCIFEROL CONTENT OF NATIONAL DRIED MILK triplicate on separate occasions Samples Nos. 1 to 6 examined in duplicate on same occasion and Nos. 7 to 12 in Sample No. 1 2 3 4 5 6 Ergocalcif eroll tLg g-l Difference Mean 0.104, 0.093 0.01 1 0.099 0.100, 0.089 0.01 1 0.095 0.101, 0.080 0.02 1 0.091 0.073, 0.067 0.006 0.070 0.098, 0.086 0.012 0.092 0.058, 0.062 0.004 0.060 Sample No.7 8 9 10 11 12 Ergocalciferol/ PLg 8-l Mean 0.048, 0.054, 0.060 0.054 0.075, 0.065, 0.064 0.068 0.094, 0.086, 0.096 0.092 0.087, 0.085, 0.082 0-085 0.058, 0.062, 0.063 0.061 0.120, 0-123, 0.130 0.124 For recovery experiments a known amount of ergocalciferol was added with the internal standard cholecalciferol to milk powder that was unfortified with vitamin D. A recovery of 97 per cent. was obtained relative to cholecalciferol, and losses of both vitamins were 20 to 30 per cent. when taken through the entire procedure. The proposed method was then applied to several brands of proprietary milk powder and the results are shown in Table 111.TABLE I11 ERGOCALCIFEROL CONTENT (pg g-l) OF PROPRIETARY BRANDS OF FULL-CREAM MILK POWDER Manufacturer Replicates Mean Declaration on label 0-093 0.088 0.086 0.088 (average) I 0.0S8 A 0.09'7 0-092 0.09 1 B 0.082 0.090 0-086 0,099 0.104 } o'lol C DISCUSSION The proposed method has been designed for routine use in monitoring the levels of ergocalciferol added to full-cream dried milk. A determination can be completed within 2 days and the method has been used successfully by less experienced staff in our laboratory for 12 months. Dried milk is usually fortified with ergocalciferol, but if cholecalciferol were to be added instead of ergocalciferol, it would be detected and the method could readily be adapted to measure the concentration of cholecalciferol present + However, in the unlikely event of fortification of the milk with both ergocalciferol and cholecalciferol, the proposed method would not be directly applicable for quantitative measurements.As in most other foods the vitamin D content of liquid milk is low. Average values calculated from published figures are equivalent on a dry mass basis to 0.003 pgg-l for winter milk and 0.006 pgg-l for summer milk. These amounts are much less than the average difference between replicates shown in Table I1 and represent at most 13 per cent. of the vitamin D content of National Dried Milk, which is required to contain between 0.044 and 0.132 pg 8-1. Consequently, the error introduced in the determination by neglecting the natural levels of ergocalciferol and cholecalciferol, both of which may be present in liquid milk, is fairly small.396 BELL AND CHRISTIE A novel feature of the proposed method is the removal of sterols and other interfering substances by a modified system of dry-column chromatography. The technique has the advantages that precipitation of sterols as their digitonides is avoided, losses of vitamins by adsorption on the voluminous digitonin precipitates are eliminated and background interference on the gas chromatogram is reduced.The beneficial effect of using two dry columns of aluminium oxide is shown in Fig. 6, which is a typical chromatogram of the ergocalciferol content of National Dried Milk and shows the response obtained for 0.1 pg of the trimethylsilyl ethers of the isovitamins of ergocalciferol and cholecalciferol. The 1 Fig.6. Gas - liquid chromatogram of ergocalci- ferol in dried milk: peak 1, added 5a-cholestane; peak 2, cholesterol trimethylsilyl ether; peak 3, added cholecalciferol isovitamin trimethylsilyl ether ; and peak 4, ergocalciferol isovitamin trimethylsilyl ether proposed method is therefore capable of detecting the levels of vitamin D naturally present in some foods, for example margarine (0.08 pg g-l) and dried egg (0.05 pg g-l), but is insuffi- ciently sensitive to detect the levels present in butter, liver and cheese (0.01, 0.005 and 0.003 pg g-l, respectively). The ten-fold increase in sensitivity required to reach such low levels may well be obtained by a combination of further improvements in procedures for the removal of interfering substances and, in particular, by the formation of halogenated derivatives of the vitamins for measurement on a gas - liquid chromatograph fitted with an elec tron-capture detect or. This paper is published with the permission of the Government Chemist. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Ziffer, H., Vanden Heuvel, W. J., Hashti, E. 0. A., and Horning, E. C., J . Amer. Chem. Soc., 1960, Murray, T. K., Day, K. C., and Kodicek, E., Biochem. J., 1966, 98, 293. Avioli, L. V., and Lee, S . W., Analyt. Biochern., 1966, 16, 193. Murray, T. K., Erdody, P., and Panalaks, T., J . Ass. Off. AnaZyt. Chem., 1968, 51, 839. Sheppard, A. J.> LaCroix, D. E., and Prosser, A. R., Ibid., 1968, 51, 834. Wilson, P. W., Lawson, D. E. M., and Kodicek, E., J . Chromat., 1969, 39, 75. Panalaks, T., Anal st, 1970, 95, 862. Edlund, D. O., anxAnfinsen, J. R., J . Ass. Off. Analyt. Chem., 1970, 53, 287. Janecke, H., and Brendal, R., Naturwissenschaften, 1971, 58, 54. Bell, J. G., and Christie, A. A., AnaZyst, 1973, 98, 268. Theivagt, J. G., and Campbell, D. J., Analyt. Chem., 1959, 31, 1375. Chen, P. S., Terepka, A. R., and Remsen, N., Ibid., 1963, 35, 2030. Loev, B., and Snader, K. M., Chern. & I n d . , 1965, 15. Loev, B., and Goodman, M. M., Ibid., 1967, 2026. Dean, A. C., I b i d . , 1971, 677. 82, 641 1. Received December 7th, 1973 Accepted January 8t12, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900385

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

AnaZyst, July, 1974, Vol. 99, $9. 397-402 397 A Simplified Colorimetric Method for the Determination of Ascorbic Acid in Pure Solutions and in Pharmaceutical Preparations BY NAG1 WAHBA, DAWOUD A. YASSA AND RAMZY S. LABIR (Biochemistry Department, Faculty of Medicine, A in-Shams University, Cairo, Egypt) A precise and specific method for the determination of L-ascorbic acid in the presence of dehydro-L-ascorbic acid and 2,3-diketo-~-gulonic acid has been developed. The method is based on the reaction of phenylhydrazinium chloride with ascorbic acid in 0.1 N hydrochloric acid at a temperature of 50 f 2 "C. The stable yellow colour produced is measured spectrophoto- metrically at 395nm. The results obtained by this method and the B.P. method are compared. There is no interference from other vitamins, minerals, glucose, sucrose or the most commonly used excipients with this method and it has been successfully applied to pharmaceutical preparations.The simplicity of the procedure permits rapid analysis, which is necessary for routine control work. ASCORBIC acid can be determined by titration with ammonium cerium( IV) sulphate, according to the method of the B.P.l or with 2,6-dichlorophenolindophenol, according to that of the U.S.P.2 Barakat, Abd El-Wahab and El-Sad9 used N-bromosuccinimide for the determina- tion of ascorbic acid in pure solution, pharmaceutical preparations and biological fluids. In all of the above methods interference from other reducing substances, such as organic and inorganic iron(I1) compounds, phenolic compounds, sodium metabisulphite (usually included in pharmaceutical preparations as an antioxidant) and reduced forms of nicotinic acid derivatives, occurs.In the Roe and Kuether rneth~d,~ ascorbic acid is oxidised to dehydroascorbic acid, and the latter is coupled with 2,4-dinitrophenylhydrazine. The defect of this method is that dehydroascorbic and 2,3-diketogulonic acids react in the same way. 2,3-Diketogulonic acid is biologically inactive and is usually present as a decomposition product of ascorbic acid in pharmaceutical preparations that have been stored. Roe, Mills, Oesterling and Damron5 have modified this method in order to render it applicable to the determination of ascorbic acid, dehydroascorbic acid and 2,3-diketogulonic acid in the presence of each other.However, the modification is time consuming as it requires three separate determinations. In the determination of ascorbic acid by the Schmall, Pifer and Wollish6 method, which involves the use of diazotised 4-methoxy-2-nitroaniline, the colour fades within 10 minutes. The use of phenylhydrazinium chloride as a colorimetric reagent for the determination of glucose in blood and lactose in milk has been reported by Wahba and co-worke~s.~~~ During the studies on the rate of reaction of phenylhydrazine with glucose, fructose, lactose and ascorbic acid at different temperatures, it was noticed that the rate of colour development with the above-mentioned sugars was very slow in comparison with ascorbic acid. It was decided, therefore, to investigate the applicability of this colour reaction to the determination of ascorbic acid.In this paper the resulting simple, precise and specific method for the accurate deter- mination of ascorbic acid in pure solutions and in pharmaceutical preparations, without interference from other compounds usually present in the preparations, is reported. EXPERIMENTAL REAGENTS- PhenyZhydraziniuw chloride solution-A 1 per cent. m/ V solution of phenylhydrazinium chloride (recently recrystallised from 95 per cent. ethanol) was made up in 0.1 N hydro- chloric acid. Standard solutions of ascorbic acid-These solutions were made up to concentrations of 25, 50 and 100 pg ml-1 in distilled water. 8 SAC and the authors. (The pH was found to range from 1-1 to 1.3.)398 [Analyst, Vol. 99 T a t solutions, containing 50 to 100 pg ml-l of ascorbic acid-Tablets and capsules were extracted by crushing and stirring in distilled water.Aqueous solutions for injection and syrups were diluted with distilled water and used directly. Oxalic acid (2 per cent.) or metaphosphoric acid (5 per cent.) can be used in both standard and test solutions to protect ascorbic acid from oxidation, especially if metal ions (e.g., Cu2+) are present in the samples to be analysed. PROCEDURE- Transfer by pipette 1.0 ml of each of the standard and test solutions into 10-ml stoppered test-tubes or cylinders, then add accurately 5-0 ml of the phenylhydrazinium chloride solution to each. Allow them to stand for 1 hour in an incubator or water-bath maintained at 50 & 2 "C. Cool, and measure the absorbance of each solution at 395 nm against a blank carried out simultaneously by using the same reagent but with 1.0ml of distilled water instead of the test solution.WAHRA et al.: COLORIMETRIC DETERMINATION OF ASCORBIC ACID Calculate the amount of ascorbic acid from the standard calibration graph. RESULTS AND DISCUSSION EFFECT OF REDUCING SUGARS- The effect of glucose, fructose and lactose on the determination of ascorbic acid by use of the proposed procedure was studied at temperatures from 40 to 55 "C. It was found that, within the temperature range investigated, concentrations of such sugars as much as 50 to 100 times greater than that of ascorbic acid did not cause any interference. EFFECT OF DEHYDROASCORBIC AND 2,3-DIKETOGULONIC ACIDS- Dehydroascorbic acid was prepared according to the method of Roe and Kuether4 by use of active charcoal and bromine oxidation.It was also prepared by oxidation with the theoretical amount of 0.1 N iodine solution, as described by Penny and 2ilva.Q 2,3-Diketo- gulonic acid was prepared by allowing a solution of dehydroascorbic acid to mutarotate for 18 days at room temperature. In some experiments it was desirable to remove hydrobromic or hydriodic acid that resulted during the preparation of dehydroascorbic acid. This was done by adding 0.1 N silver nitrate solution in order to precipitate the halide, followed by 0.1 N hydrochloric acid to precipitate the excess of silver ions. The precipitates were removed by filtration and the final volume was adjusted with 0.1 N hydrochloric acid to the required concentration.The determination of ascorbic acid, dehydroascorbic acid and 2,3-diketogulonic acid was carried out as described above under Procedure, and also by Roe and Kuether's method.* The results are shown in Table I. Each result in the table represents the mean of four separate determinations. TABLE I DETERMINATION OF ASCORBIC, DEHYDROASCORBIC AND 2,3-DIKETOGULONIC ACIDS BY THE PROPOSED PROCEDURE AND THE ROE AND KUETHER METHOD4 Recovery by- Proposed procedure, Roe and Kuether Analysed material per cent. method, per cent. A I \ Ascorbic acid . . .. . . .. . . 101.1 103.2 Dehydroascorbic acid prepared by use of- Active charcoal . . .. .. .. 00.0 98.6 Bromine oxidation . . . . .. . . 102.3 104-3 Bromine oxidation, HBr removed . .. . 00.0 97.2 Iodine oxidation . . .. . . . . 100.6 100.9 Iodine oxidation, HI removed . . . . 00.0 97.8 2,3-Diketogulonic acid . . .. .. .. 25.6 115-6 00.0 - 2,3-Diketoguloiiic acid, HI removed . . .. From these results, it is clear that whereas ascorbic acid was recovered equaUy well by both methods, dehydroascorbic acid prepared by use of charcoal did not react at all in the proposed procedure. On the other hand, dehydroascorbic acid prepared by bromine or iodine oxidation yielded complete recovery in the same procedure. However, when hydro- bromic and hydriodic acid were removed, no reaction was obtained. Thus, it was concludedJuly, 19741 I N PURE SOLUTIONS AND IN PHARMACEUTICAL PREPARATIONS 399 that the complete recovery in the presence of hydrobromic or hydriodic acid is due to the reversibility of the ascorbic acid oxidation- -2H + H,O +2H Ascorbic acid e = t Dehydroascorbic acid - 2,3-Diketogulonic acid During the long incubation time (1 hour) at 50 "C, removal of ascorbic acid by reaction with phenylhydrazine allowed further reduction of dehydroascorbic acid, until the conversion was complete.This finding suggests also the possibility of the application of the proposed pro- cedure to the determination of total active vitamin C, i.e., ascorbic and dehydroascorbic acids, by use of reducing agents, such as hydrogen ~ulphide.~ The recovery obtained for 2,3-diketogulonic acid was 25.6 per cent., but again, none was recovered when hydriodic acid was removed. It is concluded, therefore, that this recovery reflects the content of unconverted dehydroascorbic acid, which reverted to ascorbic acid during the incubat ion with phenylhydrazine.Thus, neither dehydroascorbic acid nor 2,3-diketogulonic acid reacts with phenyl- hydrazinium chloride under the conditions used. No interference from dehydroascorbic acid (prepared by using charcoal) or 2,3-diketogulonic acid (without hydriodic acid) was observed when the temperature of incubation in the proposed procedure was varied in the range from 30 to 60 "C. EFFECT OF VITAMINS, HORMONES, MINERALS AND SUCROSE- The proposed method has been applied to the determination of ascorbic acid in pharnia- ceutical preparations containing other vitamins, minerals, glucose, sucrose and some excipients. The composition of these preparations (as stated) is given in Table 11, and the per- centage recoveries by the proposed and B.P.methods are shown in Table 111. It can be seen from these results that the method has been successful. In all samples, the recovery of added ascorbic acid by the proposed method ranged between 98.1 and 102.1 per cent. As expected, the recovery of the original ascorbic acid content was very low (60 to 80 per cent.) in the incubated samples (I and 11), due to deterioration by oxidation. It is remarkable that in both samples, recovery was lower by the proposed method than by the B.P. method (82 t o 89 per cent.). TABLE I1 FORMULATIONS OF ANALYSED SAMPLES Sample A \ Constituent I Ascorbic acid/mg .. . . . . 500 - Thiamine hydrochloride/mg . . . . Riboflavinelmg . . .. .. .. Nicotinamidelmg .... .. Pyridoxine hydrochloride/mg . . .. Vitamin A/i.u. . . .. . . . . Vitamin D,/i.u. . . .. . . . . Vitamin E/mg . . .. . . ,. Calcium pantothenatelmg . . . . Glucoselg . . . . .. .. .. Calcium gluconogalactogluconatelg . . - Inositol/mg . . . . .. .. Choline chloride/mg . . .. .. Iron(l1) sulphate/mg . . .. .. Copper(I1) sulphatelmg . . .. .. Potassium iodatelmg . . . . .. Ethinyl oestradiol/mg . . .. .. Methyltestosterone/mg . . .. .. Water for injection/ml . . .. .. Syruplml .. .. . . .. . . Lactose (capsules)/mg . . .. .. - - - - - - - - - - - - - - - - I Excipients (tablets) : starch, and magnesium stearatelmg . , to 680 VI viI 75 75 1 5 1.2 2.5 10 6 2 0.5 4000 5000 400 500 2 5 4 - - - - - - 50 c 50 99 - 1-26 - 0-1 - 0.01 - 2.5 - - - to 250 to 320400 [Artalyst, Vol.99 In fresh samples, comparison of the percentage recovery of the ascorbic acid originally present in the preparations shows that it is slightly lower (3 per cent.) in samples I and 11, but markedly higher in samples I11 to VII, when determined by the B.P. method as against the proposed procedure; in all instances of the use of the B.P. method it is not less than 100 per cent. As shown in Table 11, these latter samples (I11 to VII) were those containing substances that were expected to interfere in the B.P. method, thus giving falsely high results. The difference was very marked (16 per cent.) in sample VII, in which there was a large amount of iron(I1) sulphate, other vitamins and hormones. Thus, it is concluded that vitamins, minerals, glucose, sucrose and most excipients usually found in pharmaceutical preparations either do not interfere in the proposed method, or their interference, if any, is much less than with other methods, such as that of the B.P.However, in the presence of excessive amounts of riboflavin, it is advisable to add about 0.5 g of talc to the test solution, shake, and filter it in order to eliminate the slight interference of the yellow colour of ribo- flavin, according to Wahba and Fahmy.lo WAHBA et al. : COLORIMETRIC DETERMINATION OF ASCORBIC ACID TABLE I11 COMPARISON BETWEEN THE RESULTS OF THE PROPOSED AND B.P. METHODS Recovery by- I Stated/ Added/ Proposed method, B.P. method, Sample' I (fresh) . . .. .. .. .. I (incubated at 45 "C for 12 months) . . I1 (fresh) . ... .. . . .. I1 (incubated at 45 "C for 12 months) . . 111 . . . . . . . . . . .. IV . , . . .. .. .. . . v .. . . .. .. . . VT .. .. .. .. . . .. VII . . .. .. .. .. . I mgt - 100 200 100 200 100 200 100 200 100 200 200 500 10 20 20 40 10 20 - - - - - - - - per cent.: 108.2 (h0.7) 100.1 ( 5 0 . 5 ) 99.6 (50.7) 60.2 (&l-1) 99.1 (h0.4) 98.6 (50.9) 109.3 (f0-8) 99.8 (&O-7) 101.2 (&O'S) 80.6 (f 1.1) 98.7 (f 1.2) 101-7 (A0.8) 100.8 ( f 0.75) 100.1 (fO.6) 99.8 (f0.59) 106.4 (& 1.1) 100.9 ( f 0.65) 100.3 ( f 0-85) 107.1 ( k l - 2 ) 99.6 (f0.3) 101.1 (f0-5) 108.2 ($I 1.1) 100.1 ( f 0.41) 101-7 (f0.39) 112.2 ( f 0.98) 102.1 (f0.6) 100.8 (f0.3) * For sample compositions, see Table 11. t Added to the sample in the form of an aqueous solution containing 100 mg ml-1 of ascorbic $ Mean of four experiments ; values in parentheses are standard deviations of individual acid.results. EFFECT OF CATALYSTS AND PRESERVATIVES- We found that the presence of metal salts, such as copper(I1) sulphate or iron(II1) chloride, neither inhibited nor accelerated the colour formation. On the other hand, the addition of stronger oxidising agents, such as hydrogen peroxide and potassium permanganate solutions, lowered the extinction at the maximum, which nevertheless maintained its position. Acids used during the extraction to preserve ascorbic acid in its reduced state, such as 2 per cent. oxalic acid or 5 per cent. metaphosphoric acid, did not show any interference in the proposed method.July, 19741 IN PURE SOLUTIONS AND IN PHARMACEUTICAL PREPARATIONS 401 EFFECT OF pH- The colour reaction was studied at other pH values by using acetate buffers (B.P.standard). Irregular results were obtained, the maximum absorption was shifted to a higher wavelength at pH 3, 4 and 5 and the sensitivity to micro-amounts (25 and 50 pg) decreased. PROPERTIES OF THE COLOUR- The absorption spectrum (Fig. 1) of the colour produced in the proposed procedure was scanned in a l-O-cm cell in a Carl Zeiss PMQ I1 spectrophotometer. Fig. 1 shows that the yellow colour obtained gives a maximum absorbance at 395 nm. The standard graph plotted in Fig. 2 shows that the Beer - Lambert law is obeyed in the range from 25 to 100 pg of ascorbic acid. The colour obtained after 1 hour at 50 "C remains stable for at least 24 hours.Wave lengt h/n m Fig. 1. Absorption spectrum (at 50 "C) of ascorbic acid chromogen The structure of the isolated compound is unknown. It is a highly stable product under the conditions used, and microanalysis showed that it was not a simple osazone or hydrazone, as in the instance of sugars. Further investigations are still being carried on by the authors. The method is liable to an error of &4 per cent. and the standard deviation was found to be between 5 0 . 5 and 5 1 . 7 per cent. CONCLUSION The proposed method offers several advantages over the commonly used procedures. It has a high degree of specificity that may be due to the enediol grouping of ascorbic 0 10 20 30 40 50 60 70 80 90 100 110 Concentrat ion/pg ml-1 Fig. 2. Relationship between concentra- tion and absorbance a t 50 "C402 VC‘AHBA, YASSA AND LABIB acid, as dehydroascorbic and 2,3-diketogulonic acids do not react, in contrast with the methods involving the use of 2,4-dinitrophenylhydra~ine,~*~ with which only dehydro- ascorbic and 2,3-diketogulonic acids react and which theref ore require an oxidation step.Other methods based on the reducing power of ascorbic a ~ i d l - ~ suffer from interference from other reducing substances. With the proposed procedure, reducing substances do not interfere, nor do vitamins, glucose, sucrose, minerals and common excipients in amounts usually encountered in multivitamin pharmaceutical preparations. The proposed procedure is rapid, simple, sensitive (down to 25 pg ml-l), and suitable for routine analysis (especially so, because of the stability of the colour produced). It can be carried out directly, requiring only extraction with distilled water for capsules or tablets. The method shows good precision and its accuracy compares favourably with con- ventional procedures. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. “The British Pharmacopoeia 1968,” The Pharmaceutical Press, London, 1968, p. 65. “The United States Pharmacopeia,” XVIIIth Revision, Mack Co., Easton, Pa., 1970, p. 52. Barakat, M. Z., Abd El-Wahab, M. F., and El-Sadr, M. M., Analyt. Chem., 1955, 27, 536. Roe, J. H., and Kuether, C. A., J. Biol. Chem., 1943, 147, 399. Roe, J. H., Mills, M. B., Oesterling, M. J., and Damron, C. M., Ibid., 1948, 174, 201. Schmall, M., Pifer, C. W., and Wollish, E. G., Analyt. Chem., 1953, 25, 1486. Wahba, N., Hanna, S., and El-Sadr, M. M., Analyst, 1956, 81, 430. Wahba, N., Ibid., 1965, 90, 432. Penny, G. R., and Zilva, S. S., Biochem. J., 1943, 37, 39. Wahba, N., and Fahmy, E., J. Pharm. Pharmac., 1965, 17, 489. Received J u N e 18th. 1973 Amended November 22nd, 1973 Accepted January 29th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900397

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, July, 1974, Vol. 99, $9. 403407 403 Chemical Interferences in the Determination of Manganese in Plant Material by Atomic-absorption Spectroscopy BY E. G. BRADFIELD ( Uvtiversity of Bristol, Department of Agriculture and Horticulture, Long Ashton Research Station, Bristol, BS18 9AF) Combinations of calcium plus magnesium with sulphate may cause low results in the determination of manganese by atomic-absorption spectroscopy. Maximum depression of manganese absorbance is noted when the mole ratio of calcium plus magnesium to sulphur is 1 : 1. When plant samples are digested with a mixture of nitric and perchloric acids, low results are obtained only for those materials which are low in calcium p2us magnesium and high in sulphur, e.g., grain ; correct values are obtained by incorporating lanthanum chloride in the solution.If sulphuric acid is used in the digestion, very low values are obtained and addition of lanthanum chloride effects little improvement. ALL AN^ first reported the use of atomic-absorption spectroscopy for the determination of manganese in plant material. He found no interference from any of the other elements present in the samples examined and his findings were subsequently confirmed by David2 and Buchanan and M~raoka.~ More recently, Ebdon, Kirkbright and West4 noted that magnesium caused a serious depression of both fluorescence and absorbance of manganese by a chemical effect but that calcium had a negligible effect. However, in those instances when inter- ferences from cations have been investigated, mention is seldom made of the accompanying anions although these may have a marked effect on the extent of any depression of absorbance in the flame.For example, in a study of the determination of zinc in agricultural analysis by atomic-absorption spectroscopy, Bradfield and Spincer5 noted a depression of zinc absorbance in the presence of magnesium sulphate, but no depression was observed if magnesium and sulphate were tested as chloride and sulphuric acid, respectively. The depressive effect of magnesium sulphate was noticeable only in the cool propane - air flame and disappeared in the hotter acetylene - air flame. Therefore, it seemed desirable that chemical interferences in the determination of mangan- ese by atomic-absorption spectroscopy should be examined further. This paper is concerned with the influence of magnesium and calcium on manganese absorbance in the acetylene - air flame and the implication of the results for the determination of manganese in plant material.EXPERIMENTAL AND RESULTS APPARATUS AND INSTRUMENT CONDITIONS- An EEL 240 atomic-absorption spectrophotometer (reciprocal dispersion 6.6 nm mm-l) fitted with an air - acetylene burner (slot dimensions 100 x 0-62 mm) and operated under the conditions shown below was used. These conditions were maintained constant throughout the experiments. Wavelength Slit setting Lamp current 279-4 nm No. 3 (0.14 mm) 10 mA The diameter of the lamp beam at its focus at the centre of the burner was 4 mm. The air flow was set to 3.2 1 min-1 and the acetylene flow adjusted until no yellow colour was visible above the cone of the flame (0.6 1 min-l).Under these conditions the flame was about 30 mm in height and about 7 mm in width and absorbance readings of solutions containing 3 pg ml-l of manganese were at a maximum. The absorbance varied little with the height of the burner but the most stable readings were obtained with a setting of 10. In this position the centre of the beam from the hollow-cathode lamp was about 36mm above the burner top when measured at its focus midway along the burner length. STUDY OF INTERFERENCES- Efect of anions-The absorbance of solutions of manganese sulphate and manganese chloride were identical over the range 1 to 5 pg ml-l of manganese and additions of sulphate, chloride and phosphate as their respective acids, in the range 1 to 200 pg ml-l, produced no change in the measured value.@ SAC and the authors.404 [A~talyst, Vol. 99 Efect of cations-Iron (10 pg ml-l), potassium (500 pg ml-l), magnesium (100 pg ml-1) and calcium (250 pg ml-l) were added as their chlorides and sulphates to solutions containing 3 pg ml-l of manganese and the absorbances were recorded (Table I). No interference was observed when the above solutions were added as their chlorides but a marked decrease in absorbance occurred when magnesium and calcium were added as sulphates. TABLE I ABSORBANCE OF MANGANESE (3 pg ml-l) IN THE PRESENCE OF OTHER CATIONS Results are the means of three determinations BRADFIELD : CHEMICAL INTERFERENCES IN THE DETERMINATION OF AS CHLORIDE AND SULPHATE Manganese Manganese - Added as sulphate Chloride Sulphate * Added as chloride Chloride Sulphate Iron (10 p g ml-l) .. . . 0.265 0.265 Iron (10 pg ml-l) . . . . 0.265 0.265 Potassium (500 p g ml-l) . . 0.270 0.270 Potassium (500 p g ml-l) . . 0.270 0.270 Magnesium (100 pg ml-l) . . 0.260 0.270 Magnesium (100 pg ml-l) . . 0.125 0.130 Calcium (250 p g ml-l) . . 0.270 0.270 Calcium (250 pg ml-l) . . 0.075 0.090 The addition of combinations of magnesium and sulphate was therefore studied in more detail: by examining the effect of increasing amounts of magnesium (as sulphate) on the absorbance of 3 pg ml-1 of manganese (as sulphate) (Fig. 1) ; and by examining the effect of in- creasing amounts of magnesium (as chloride) in combination with increasing amounts of sul- phate as sulphuric acid on the absorbance of 3 pg ml-I of manganese (as chloride) (Fig.2). Fig. 1 shows that the absorbance of manganese in the flame was progressively reduced by increasing concentrations of magnesium sulphate in the solution. When magnesium was present as the chloride, the mole ratio of magnesium to sulphur at which a decrease in absorb- ance of manganese commenced was dependent on the magnesium concentration ; maximum depression occurred when the mole ratio was 1 :1 (Fig. 2). Similar results were obtained when calcium was substituted for magnesium. 100 200 300 400 Mg (as MgSO,)/pg ml-l Fig. 1. Effect of magnesium sulphate on absorbance of manganese a t different heights in the air - acctylene flame. Height in flamelmm: A, 10; and B, 3.5 At the levels used in the above experiment, the effect of combinations of magnesium and sulphate on the absorbance of manganese was completely eliminated by addition of 2000 pg ml-l of lanthanum (as chloride). The observed effect of magnesium (as sulphate) on the absorbance of manganese is an example of a reduction in the vaporisation rate of the solute by occlusion of the analyte in a matrix, which is not completely volatilised in the flame (Type IIa condensed-phase effect on the classification of Alkemade6). The degree of interference becomes less when the absorbance is measured at a higher Interference is less as the droplet size and hence the size of the solid particle in the flame is Characteristics of this effect are as follows: position in the flame (Table I1 and Fig.1). decreased, e.g., by addition of an alcohol (Table 11).July, 19741 MANGANESE I N PLANT MATERIAL BY ATOMIC ABSORPTION 405 1 2 3 4 Mole ratio, cation t o sulphur I ! ! 1 2 3 Mole ratio, cation to sulphur Fig. 2. Effect of increasing concentrations of sulphate (as sulphuric acid) in the presence of calcium (a) and magnesium (b) as chlorides on the absorbance of manganese. Concentration of calcium or magnesiumlpg ml-l: A, 25; B, 50; C, 100; and D, 200 Interference is reduced or eliminated by addition of easily volatile solutes, e g . , sodium chloride or lanthanum chloride to the solution (Table 111). The degree of interference should be dependent on flame temperature, being less at higher temperatures. The absorbance of 3 pg ml-l of manganese in the presence of increas- ing amounts of magnesium sulphate was measured in flames with different ratios of acetylene to air (Table IV). Although the degree of interference is reduced in the acetylene-lean flame, the over-all sensitivity to manganese is also reduced.It is therefore concluded that the best way of overcoming interference from alkaline earth sulphates in the acetylene - air flame is by addition of an easily volatile solute to the solution. TABLE IT. EFFECT OF MAGNESIUM SULPHATE ON THE ABSORBANCE OF 3pgml-l OF MANGANESE AT DIFFEREIU'T HEIGHTS (mm) IN THE FLAME WHEN VAPORISED FROM AQUEOUS AND 20 PER CENT. PROPAN-8-OL SOLUTIONS Height when vaporised from aqueous solution/mm Height when vaporised from 20 per cent. propan-2-01 solution/mm A A r \ I > Magnesiurnlpg ml-' 2.5 5.5 8.5 10.0 2.5 5.5 8.5 10.0 0 0.260 0.315 0.285 0.255 0.400 0.405 0.365 0.345 20 0.200 0.285 0.285 0.270 0.345 0.385 0-370 0.345 100 0.098 0.185 0.235 0.245 0.200 0.300 0.360 0.350 400 0.040 0.095 0.145 0.165 0.130 0.145 0.255 0.270 EFFECT OF HYDROCHLORIC ABSORBANCE OF 3 pg ml- Chloride (hydrochloric Magnesium1 acid) 1 0 1000 pg ml-l pg ml-l 20 100 400 TABLE I11 ACID, SODIUM CHLORIDE AND LANTHANUM CHLORIDE ON THE Chloride Chloride (sodium (lanthanum chloride) 1 chloride) / Absorbance pg ml-l Absorbance pg ml-l Absorbance 0.275 1000 0.285 1000 0.275 0.232 0.285 0-270 0.150 0.270 0.275 0.088 0.280 0.275 -1 OF MANGANESE I N THE PRESENCE OF MAGNESIUM SULPHATE406 BRADFIELD : CHEMICAL INTERFERENCES IN THE DETERMINATION OF [Analyst, Vol.99 TABLE IV EFFECT OF MAGNESIUM SULPHATE ON THE ABSORBANCE OF 3 pg ml-l OF MANGANESE IN FLAMES WITH DIFFERENT ACETYLENE - AIR RATIOS Air flow-rate 3.2 1 min-l and acetylene flow-rate/l min-l r A > Magnesiumlpg ml-1 0.62 0.58 0.45 0 0.245 0-260 0.190 0-160 100 0.078 0.082 0.098 400 0.036 0.042 0.062 20 0.175 0.182 PLANT ANALYSIS- Different types of plant material were analysed by six different procedures.Concentra- tions and mole ratios of calcium plus magnesium to sulphur in the solution resulting from digestion of these materials are shown in Table V, and the results obtained for manganese (p.p.m.) are shown in Table VI. PROCEDURES- (A) With nitric, perchloric and sulphuric acids. Weigh a suitable amount of plant material (previously oven-dried at 105 "C for 1 hour) into a Pyrex 50-ml conical flask and ash it overnight in a muffle furnace at 450 "C.To the cooled ash, add 2 ml of 16 M nitric - 60 per cent. perchloric acid mixture (3 + 1) and 0.5 ml of 18 M sulphuric acid. Place a reflux funnel in the flask and digest the mixture on a hot-plate until dense white fumes of perchloric acid are evolved. Remove the reflux funnel and evaporate the contents of the flask to incipient dry- ness. Replace the reflux funnel and boil the mixture for 5 minutes. Transfer the digest to a Pyrex glass tube calibrated at 20 ml and dilute it to the mark. Determine the manganese by atomic-absorption spectroscopy. (B) As A, but add lanthanum chloride to the final solution to give a concentration of 2000 pg ml-l of lanthanum. (C) As A, but determine manganese by the colorimetric formaldoxime procedure (Bradfield').(D) As A, but omit sulphuric acid from the digestion mixture. This is the standard diges- tion procedure used at Long Ashton for the trace-metal analysis of plant material. The advantage of using a combined dry ashing - wet digestion procedure has previously been discussed.' The six procedures used are as follows. To the cooled flask, add 10 ml of 0.5 M hydrochloric acid. (E) As D, but add lanthanum chloride to the final solution. (F) As D, but determine manganese with formaldoxime. DISCUSSION The reduction in measured absorbance of manganese which is observed in the presence of magnesium or calcium sulphate, or both, seems to be adequately explained as a condensed- phase effect (Alkemade6).The rate of production of manganese atoms from the mixed phase consisting of manganese, magnesium and calcium sulphates is less than that from a mixed phase consisting of the corresponding chlorides during the short time in which the particle is in TABLE V CONCENTRATIONS OF CALCIUM, MAGNESIUM AND SULPHUR AND MOLE RATIO OF CALCIUM ~ Z U S MAGNESIUM TO SULPHUR I N DIGESTS OF PLANT MATERIAL Nitric and perchloric acids Mag- Mole ratio of calcium r A > Calcium/ nesium/ Sulphur/ plus magnesium t o Nitric, perchloric and pg ml-1 pg ml-l pg ml-l sulphur sulphuric acids Lucerne . . . . 820 75 85 8.9 All cations are present as Oat grain . . . . 70 60 65 2.1 sulphate; mole ratio 1 Tomato leaf . . 640 60 240 2.5 Kale 1 . . .. 1630 65 500 2-6 Kale 2 . . . . 4140 160 1600 2-2July, 19741 MANGANESE I N PLANT MATERIAL BY ATOMIC ABSORPTION TABLE V I ANALYSIS OF PLANT MATERIALS BY DIFFERENT PROCEDURES Results are the means of six determinations 407 Procedure A B C D E F P = 0.05 P = 0.001 7 Lucerne 29-5 29.4 36.5 37.8 37-0 37.4 3.6 6.5 Manganese, p.p.m. A Oat grain Tomato leaf 45.3 325 55.4 332 52.5 423 50.0 420 58.6 450 59-9 443 5.0 24 9.0 43 Kale 1 26.4 24.9 36.0 35.3 37.7 39.2 2-6 4.6 7 Kale 2 10-3 8-9 13.3 13.1 15.2 13.9 0.7 1.4 the flame. Consequently, the degree of interference from magnesium as sulphate becomes less as the height (and hence the time at which the absorbance is measured in the flame) is increased (Table 11). Also, as vaporisation and atom production take place from the surface of the solid phase, the rate of this vaporisation should increase as the particle size decreases.In the presence of an alcohol the droplet size of the spray entering the flame is reduced, the surface area is increased and vaporisation and absorbance of manganese is increased (Table 11). Further, incorporation of a large excess of an easily vaporised salt, e.g., sodium or lanthanum chloride, into the solution results in a greater dispersion of the involatile sulphate mixed phase and an increase in its rate of volatilisation, as discussed by Baker and Garton.8 Analysis of plant material shows that highly significant differences are obtained when manganese is determined by atomic-absorption spectroscopy on solutions from different digestion procedures (Table VI). Fig. 2 indicates that the mole ratio of calcium or magnesium t o sulphur at which a reduction in manganese absorbance starts to occur depends to some ex- tent on the absolute amounts of calcium and magnesium present in thesolution.However, use of this figure in conjunction with Table V indicates that low results for the determination of manganese in a nitric acid - perchloric acid digest might be expected for oat grain only. Table VI shows that when nitric and perchloric acids are used for the digestion, oat grain (P = 0-OOl), kale 1 (P = 0.01) and kale 2 (P = 0-05) give low results for manganese (cf. procedures D and F) and that, for these samples, correct values are obtained by incorporating lanthanum chloride in the solution (cf. procedures E and F). If sulphuric acid is incorporated in the digesting acid (as is sometimes advised for the destruction of organic matter when using nitric and perchloric acids) the mole ratio of calcium plus magnesium to sulphur becomes 1 :1 and low results for manganese would be expected for all samples examined.The results show that when sulphuric acid is employed for the digestion, very low values are obtained for manganese by atomic-absorption spectroscopy (cf. procedures A and C) and, in these circumstances, little improvement is achieved by addition of lanthanum chloride except for oat grain (cf. procedures A and B). The use of sulphuric acid should therefore be avoided whenever possible in sample prepara- tion, not only for the reasons discussed above but also because of the potential danger of precipitation of alkaline earth sulphates and loss of manganese by adsorption. (Compare the slightly lower figures given by procedure C with those given by procedure F.) 1. 2. 3. 4. 5. 6. 7. 8. REFERENCES Allan, J. E., Spectrochim. Ada, 1959, 10, 800. David, D. J., Atom. Absorption Newsl., 1962, No. 9, 1. Buchanan, J. R., and Muraoka, T. T., Ibid., 1964, No. 24, 1. Ebdon, I,., Kirkbright, G. F., and West, T. S., Talanta, 1970, 17, 965. Bradfield, E. G., and Spincer, D., J . Sci. Fd Agric., 1965, 16, 33. Alkemade, C. Th. J., Analyt. Chem., 1966, 38, 1252. Bradfield, E. G., J . Sci. Fd Agric., 1964, 15, 469. Baker, C. A., and Garton, F. W. J., Rep. U.K. Atom. Energy Auth., AERE-R4390, 1961. Received December 14th, 1973 Accepted February 1 Ith, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900403

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

408 Analyst, July, 1974, Vol. 99, pp. 408-412 A Method for the Determination of Copper in Biological Materials Involving the Use of Sodium Diethyldithiocarbamate BY M. E. M. s. DE SILVA (Government Analyst’s Laboratory, Colombo, Sri Lanka) The method described for determining copper is based on a complexation reaction effected in acetic acid solution. The acetic acid not only effects com- plexation but also dissolves the complex and no intermediate extraction is therefore necessary. The sensitivity is higher than that for other methods based on diethyldithiocarbamate. Various metal ions interfere in the deter- mination of copper using diethyldithiocarbamate and the elimination of these interferences is discussed. The application of the method to the determination of trace amounts of copper in various biological materials, particularly foodstuffs, is discussed.MANY colorimetric methods have been described for the determination of trace amounts of copper, and those based on bat hocuproine (2,g-dimet hyl-4,7-diphenyl- 1,lO-phenanthroline)l and neocuproine (2,9-dimethyl-1, lO-phenanthr~line)~ are considered to be specific, Extraction with dithizone (diphenylthiocarbazone) in acidic solution is extremely sensitive but not specific. The most widely used methods, however, are still based on dithio~arbamates,~-5 including zinc dibenzyldithiocarbamate.6 Diethyldithiocarbamate reacts with copper to give a yellow - brown suspension in slightly acidic or alkaline solution and, as this suspension is not suitable for use in direct determina- tions, it has been customary to form the copper complex in aqueous solution and to extract this complex with an immiscible solvent (chloroform and carbon tetrachloride are con- sidered to be the most satisfactory solvents for this purpose). Zinc, cadmium, indium, magnesium, silver, lead, iron, cobalt and nickel form white or coloured complexes with diethyldithiocarbamate, and methods have been devised for the elimination of these interferences by using preferential complexation with EDTA,5 itr rate,^ etc.The procedure described in this paper is based on the use of glacial acetic acid to effect complexation and to effect dissolution of the copper complex, giving clear and strongly coloured solutions. The molar extinction coefficient varies with the concentration of the acetic acid in solution, but the determination is carried out when the molar extinction coefficient has reached its optimum value.The drawbacks associated with the use of aqueous suspensions are thereby overcome and an alternative to the usual procedure in which solvent extraction is necessary has been developed. The sensitivity is higher than for other methods based on diethyldithiocarbamate and the reproducibility is considerably enhanced. Inter- ference from various metal ions (at reasonable concentrations) is eliminated by the use of EDTA. EXPERIMENTAL REAGENTS- All reagents were of analytical-reagent grade. Glacial acetic acid. Sodium diethyldithiocarbamate, ( C2H5) ,NCSSNa.3H20-The reagent is added in the solid form. It dissolves in the acetic acid solution and eventually decomposes, but this decom- position does not interfere in the reaction. Ethyle~~ediami~~etetraacetic acid, C,,H,,0,N,.2H20-This reagent is added in the solid form.Stock standard copper solution-A 0-3929-g amount of copper(I1) sulphate (CuS0,.5H20) was dissolved in water and the volume made up to 1 litre in a standard flask. This solution contains 100 pg ml-l of copper. 0 SAC and the author.DE SILVA 409 Working standard copper solzttiouts-Working standards containing 2.5, 5.0, 7-5 and 10.0 pg ml-l of copper were prepared by serially diluting the stock standard solution when required. These working standard solutions are, however, stable. ABSORBANCE MEASUREMENTS- The absorbance measurements were made with a Beckmann, Model DB, spectro- photometer using l-cm cuvettes.The reference cuvette was filled with copper-free distilled water in every instance. METHOD PROCEDURE- Transfer a 10-m1 aliquot of the copper solution (free from mineral acid) by pipette into a 50-rnl calibrated flask and add 2.5 ml of copper-free distilled water from a burette, then add 30 ml of glacial acetic acid and swirl the contents of the flask in order to mix them. Then add about 10mg of sodium diethyldithiocarbamate and again swirl the solution. Allow the flask to stand for 30 minutes after making the volume of the liquid up to the mark with glacial acetic acid. Read the absorbance at 430 nm in a l-cm cuvette in the Beckmann spectrophotometer. REAGENT BLANKS- acetic acid medium. described under Reproducibility. Reagent blanks presented little difficulty; diethyldithiocarbamate dissolves readily in the The reproducibility of the reagent blank was studied together with the other experiments The results that are relevant are as follows : Average absorbance of Over-all Degrees in l-cm cuvctte absorbance units freedom reagent blank standard deviation, of 0.013 0-0008, 19 The over-all standard deviation is reported as there was no significant difference between the between-batch and the within-batch standard deviations at the 95 per cent.confidence level. REPRODUCIBILITY- Aliquots of 10 ml of each of the working standards containing 2.5,5.0,7.5 and 10.0 pg ml-1 of copper were taken in quadruplicate and subjected to the treatment described under Procedure. Reagent blanks were also prepared in quadruplicate, 10.0 ml of copper-free distilled water being used in place of the copper solutions.Absorbance measurements were made and the results of analytical interest were obtained by subtracting the mean absorbance of the reagent blank from the absorbance of the copper solution. The procedure was repeated on five different occasions and the results were analysed statistically. The between-batch standard deviations were not significantly larger (at the 95 per cent. confidence level) than the within-batch standard deviations. The results were pooled so as to obtain the over-all standard deviations reported in Table I. TABLE I REPRODUCIBILITY OF THE METHOD Mean absorbance reading corrected for Over-all standard Amount of copper absorbance of deviation, absorbance Degrees in 50-ml flask/pg reagent blank units of freedom 25 0.150 0.0013, 10 50 0.298 0.0017, 19 75 0.446 0.0022, 19 100 0.596 0*0028, 19 The mean absorbance of the blank solutions was 0-013 with a standard deviation of 0.0008p410 CALIBRATION GRAPH- DE SILVA: DETERMINATION OF COPPER I N BIOLOGICAL MATERIALS [Analyst, VOl.99 In studying the reproducibility, it was shown that the between-batch standard deviation was not significantly larger than the within-batch standard deviation. The mean absorb- ance of the twenty readings obtained at each level (corrected for the absorbance of the reagent blank) was therefore used in order to plot the calibration graph. The calibration graph was linear and passed through the origin, and can be repre- sented by the equation y = 0.596 x x where y is the absorbance reading obtained under the conditions described under Procedure and x is the amount of copper in micrograms in the 50-ml flask.INTERFERENCES- Several metal ions interfere in the determination of copper using diethylditliiocarbamate. However, the following modification of the method, involving preferential complexation of the interfering ions with EDTA, eliminates the interfering effects of iron, nickel, cobalt, manganese and mercury. Transfer by pipette a 10-ml aliquot of the copper solution containing the interfering metal ion and free from mineral acid into a 50-ml flask, then add 2.5 ml of copper-free dis- tilled water from a burette followed by 25 ml of glacial acetic acid. Add 50 mg of EDTA and swirl the flask in order to mix the contents.Allow the flask to stand for 15 minutes, then add 10 mg of sodium diethyldithiocarbamate, swirl the flask and allow it to stand for 15 minutes. Filter the suspension, if any, into a 50-ml calibrated flask through a small Whatman No. 41 filter-paper and carefully wash the flask and the residue on the paper with small volumes of glacial acetic acid. Finally, make the solution up to the mark with glacial acetic acid and allow it to stand for 15 minutes before reading the absorbance. The results are reported in Table 11. TABLE I1 INTERFERENCE OF METAL IONS I N THE DETERMINATION O F 25pg OF COPPER Copper foundlpg I A 1 Metal ion (100 pg) In the absence of EDTA Fe2+ 36.4 Fez+ 34.2 Nie+ 32.6 Co2 + 35-8 Mn2+ 25.6 Pb2+ 25.8 Hg2+ 26.2 In the presence of EDTA 25-1 25.3 24.8 24.8 25.0 25.2 25.4 In addition to the interfering metal ions mentioned above, certain other common constituents of food and other biological materials were also studied and the results are reported in Table 111.TABLE I11 EFFECT OF CONSTITUENTS OF BIOLOGICAL MATERIALS ON THE DETERMINATION OF 25 pg OF COPPER Substance added (50 mg) Copper found/pg CaCO, 24.8 24.6 24.6 A1,(SO4)2. 24.9 (NH4)2S04 25.0 NaCl 25.1 MgSO4 Ca,(PO4), BEHAVIOUR OF THE COPPER - DIETHYLDITHIOCARBAMATE COMPLEX I N ACETIC ACID SOLUTION AND DETERMINATION O F THE OPTIMUM CONDITIONS FOR THE DETERMINATION OF COPPER Preliminary studies were carried out on the behaviour of the complex in the acetic acid medium. In these experiments, a standard solution containing 10 pg ml-1 of copperJuly, 19741 USING SODIUM DIETHYLDITHIOCARBAMATE 41 1 in 90 per cent.V/V glacial acetic acid was used, which was prepared by dilution of the stock standard solution with glacial acetic acid. A 10-ml volume of the solution containing 100 pg of copper was transferred by pipette into a 50-ml calibrated flask and a sufficient amount of copper-free distilled water was added so as to adjust the final strength of the acetic acid to the required value. This solution was then examined as described under Procedure. The absorbance readings obtained at each acetic acid concentration, corrected for the absorbance of the reagent blank, were used to calculate the molar extinction coefficients. RESULTS- The molar extinction coefficient of the copper - diethyldithiocarbamate complex varied with the concentration of glacial acetic acid in the solution.It was observed that at glacial acetic acid concentrations up to about 50 per cent. V/V, the solutions were turbid owing to the poor solubility of the complex in the medium, and that the molar extinction coefficient increased for glacial acetic acid concentrations in the range 60 to 78 per cent. V/V, reaching a broad peak with a maximum at 75 per cent. V/V, corresponding to a value of 18 935 1 mol-l cm-l. The variation of molar extinction coefficient for glacial acetic acid concentrations in the range 72 to 78 per cent. V/V was very small. The molar extinction coefficient then decreased, reaching a fairly constant value at glacial acetic acid concentrations in the range 90 to 96 per cent.V/V. Details of these results are given in Table IV. TABLE IV VARIATION OF MOLAR EXTINCTION COEFFICIENTS WITH GLACIAL ACETIC ACID CONCENTRATION Glacial acetic Corrected absorbance reading per cent. V / V acid in solution, at 430 nm corresponding to Calculated molar extinction coefficient 100 pg of copper in 50 ml 60 0.480 15 253 62 0.506 16 080 64 0.530 16 840 66 0.554 17 605 68 0.570 18 073 70 0.584 18 555 72 0.592 18 810 74 0-596 18 935 76 0-596 18 935 78 0.588 18 685 80 0.578 18 365 82 0.560 17 790 84 0.534 16 965 86 0.510 16 206 88 0.485 15 410 90 0.473 15 030 92 0.473 15 030 94 0.479 15 220 96 0.473 15 030 On the basis of the results, it was decided to carry out the determination of copper The absorption curve under optimum conditions, i.e., in 75 per cent.V/V glacial acetic at a glacial acetic acid concentration of 75 per cent. V/V. acid, showed a maximum at 430 nm and a minimum at 372 nm. APPLICATION OF THE METHOD TO FOOD AND OTHER BIOLOGICAL MATERIALS Treatment of liquid samples, e.g., water and spirituous Liquors, containing 0.1 p.p,m, or more of cof@er-For colourless liquid samples, evaporate 50 ml of the sample and adjust the final volume t o 10ml with copper-free distilled water. For coloured liquid samples, evaporate 50 ml of the liquid to dryness in a platinum dish and oxidise the residue with a few drops of concentrated nitric acid, heating the dish to complete the oxidation. Finally, evaporate the oxidised liquid to dryness and take up the residue in 10 ml of copper-free distilled water.412 DE SILVA Treatment of solid samples containing 5 p.p.m.OY more of copper-Ash 1 g of well sampled material in a platinum dish, dissolve the residual ash in dilute hydrochloric acid and evaporate the solution to dryness. If the residue is yellow in colour, dissolve it in 2 ml of water and evaporate the solution to dryness, then repeat this step. This should remove any colour. Take up the residue in 10ml of copper-free distilled water. Alternatively, wet oxidise 1 g of well sampled material using Middleton and Stuckey's method,' remove any excess of nitric acid, and add 10 rnl of copper-free distilled water. Determination of copper-To the final 10-ml volume of the sample treated as described above, add 2-5 ml of copper-free distilled water and 25 ml of glacial acetic acid and proceed as described under Interferences. The author thanks the Government Analyst, Sri Lanka, for encouragement in carrying out this work. REFERENCES 1. 2 . 3. 4. 5. 6. 7. Smith, G. F., and Wilkin, D. H., Analyt. Chem., 1953, 25, 510. Smith, G. F., and McCurdy, W. H., Ibid., 1952, 24, 371. Callan, T., and Henderson, J. A. R., Analyst, 1929, 54, 650. Haddock, L. A, and Evers, N., Ibid., 1932, 57, 495. Sedivec, V., and Vasak, V., Colllz Czech. Chem. Commun., 1950, 15, 260. Abbott, D. C., and Polhill, D. A., Analyst, 1954, 79, 547. Middleton, G., and Stuckey, R. E., Ibid., 1954, 79, 138. Received June 4th, 1973 Amended February 4th, 1974 Accepted February 12th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900408

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, July, 1974, Vol. 99, $9. 413-430 413 The Absorptiometric Determination of Aluminium in Reagents and the Development of an Improved Method Water. A Comparison of Some Chromogenic BY W. I(. DOUGAN AND A. L. WILSON ( Water Research Centre, Medmenham Laboratory, Medmenham, iWavlow, Buclzinghanzshire, SL7 2HD) The properties of chromogenic reagents used for the absorptiometric determination of aluminium in water are compared, and an experimental comparison of catechol violet, Eriochrome cyanine R and stilbazo has been made. Catechol violet is considered to be the most suitable and a method involving the use of this reagent has been developed. When using this method, the standard deviation of the analytical results varied from 0.004 to 0.008 mg 1-1 of aluminium for concentrations of 0.05 and 0-3 mg 1-1 of aluminium, respec- tively.With the exception of fluoride, substances normally present in treated waters did not cause important interference. The effect of fluoride (1 mg 1-1 or less) is tolerable for most purposes, but appreciably greater concentrations require preliminary removal of fluoride or correction for its effect. Satis- factory results have been obtained with the method in six other laboratories. The method is simple and rapid ; ten samples can be analysed in approximately l& hours, The method has advantages over other commonly used methods, and is recommended for use in water analysis laboratories. THE concentration of aluminium in potable water is often an important factor in assessing both the quality of the water and also the performance of the associated treatment plant.Concentrations of less than 0.1 mg 1-1 of aluminium are the common aim, and we required an analytical method capable of the following performance : Concentration range . . . . 0 to 0.3 nig 1-1 of aluminium Standard deviation >0.005 mg 1-1 of aluminium or 5 per cent. of the aluminium Bias .. .. .. . . > O - O l mg 1-1 of aluminium or 10 per cent. of the aluminium . . . . concentration (whichever is the greater) concentration (whichever is the greater) Several analytical techniques have been used for determining such aluminium concen- trations, but we restricted our attention to absorptiometric methods (measuring in the visible region) because the necessary instrumentation is usually available in laboratories concerned with the analysis of potable water.When this work was started, aluminon was the most commonly used chromogenic reagent in this country. A number of standard methods using this reagent1-3 are based largely on the work of Packham,* who showed that aluminon has the following disadvantages: (i) the coloured product is a lake, and thus requires very careful control of all experi- mental conditions as well as the addition of a colloid stabiliser [gum arabic or poly( vinyl alcohol)] ; (ii) in order to achieve proper colour development, the sample must be heated, which is inconvenient and increases the analytical time; (iii) the calibration graph is not linear; (iv) close control of the temperature of the processed sample is desirable, the optical ( v ) fluoride causes very large negative errors; for example, 1.0 mg 1-1 of fluoride causes (vi) different batches of aluminon may give markedly different calibration graphs. density changing by approximately 0.6 per cent."C-l at 20 "C; an error of -0.07 mg 1-1 of aluminium when 0-1 mg 1-1 of aluminium is present; @ SAC and the authors.414 [Analyst, Vol. 99 In view of these disadvantages, the literature to the end of 1969 was surveyed in order to determine whether more suitable chromogenic reagents had become available since Pack- ham's work. Particular attention was paid to papers concerned either with methods for analysing water or with comparison of chromogenic reagents. Methods that require sample concentration (e.g., by solvent extraction) were excluded as they are usually slower and more difficult to carry out.COMPARISON OF CHROMOGENIC REAGENTS DOUGAK AND WILSON : THE ABSORPTIOMETRIC DETERMINATION Several workers have reported experimental comparisons of different chromogenic reagents. For example, Packham4 compared aluminon, alizarin red S and haematoxylin for the analysis of water. The last reagent was regarded as the least satisfactory because the coloured product had poor stability, and the calibration graph was markedly non-linear. The sensitivity obtained with alizarin was only half that with aluminon, and the optical density of the blank determination using the former reagent was undesirably large. Packham concluded that aluminon was the most satisfactory reagent. GiebleIS studied the use of aluminon, alizarin red S, Eriochrome cyanine R and haema- toxylin for the analysis of water, and also concluded that aluminon was best because it was the least subject to interference.Vezzi6 also compared aluminon, Eriochrome cyanine R and haematoxylin, and concluded that aluminon was best, although the effect of fluoride was large. Pakalns7 described a method involving the use of chrome azurol S, and compared it with procedures with aluminon and Eriochrome cyanine R. The last two reagents were especially liable to interference from chromium and cobalt ; Eriochrome was also seriously affected by phosphate and, in addition, required strict control of the pH of the solution. Although Eriochrome gave greater sensitivity than chrome azurol S, the latter was preferred on the grounds of simplicity, selectivity and reliability.Corbett and Guerins selected alizarin red S, arsenazo, Eriochrome cyanine R, 8-hydroxy- quinoline and stilbazo from a survey of methods for determining aluminium in steel. Non- linear calibration graphs were reported for Eriochrome and stilbazo, close control of experi- mental conditions was necessary for Eriochrome and the stability of its coloured product with aluminium was poor. Alizarin red S and arsenazo were considered to be the most suitable reagents. Aginskaya and Petrashen9 compared aluminon and stilbazo. The latter gave twice the sensitivity of the former although the stability of the coloured product was better for the former. Methods involving the use of aluminon and Eriochrome cyanine R for the analysis of water were compared in an inter-laboratory study,1° which showed that both reagents were liable to interference from fluoride and polyphosphate, that Eriochrome gave less biased results when other metals were present and that neither reagent gave good accuracy.The greater speed and simplicity of the method with Eriochrome were considered sufficient grounds for preferring this reagent, and the method has been included in a recent compilation of standard methods.ll In another inter-laboratory study,12 no marked differences in errors were observed for the reagents aluminon, alizarin red S, Eriochrome cyanine R and haematoxylin. These comparative studies did not allow definite conclusions to be drawn on the best reagent of those investigated. However, aluminon, alizarin red S, arsenazo, chrome azurol S, Eriochrome cyanine R and stilbazo were selected for more detailed consideration. In addition, catechol violet * (3,3',4'-trihydroxyfuchsone-2"-sulphonic acid) was also included because it appeared to have favourable properties.Although no method involving the use of catechol violet for the determination of aluminium in water has been published, several workers have described the use of this reagent for other materials. Thus, Anton13 first proposed its use for the absorptiometric determination of aluminium, and Wilson and Sergeant14 applied it to mineral analysis. Tanaka and Yamayoshil5 have reported on the optimum experimental conditions for determining aluminium, and Meyrowitz16 has also investigated the effect of different experimental conditions and applied the reagent to the analysis of minerals.Some of the analytically important properties of the selected reagents for determining aluminium in pure solution are summarised in Table I. In a few instances, when relevant results were not reported in the literature, approximate values were measured. * Catechol violet is also often called pyrocatechol violet.July, 19741 OF ALUMINIUM I N WATER 415 Of the reagents listed in Table I, only aluminon requires a heating period for colour development, and this reagent was therefore rejected. Arsenazo was also not further con- sidered because the molar extinction coefficient was too small (for our purpose, a value of approximately 2 x lo4 or greater was considered desirable).Alizarin red S and chrome azurol S were also discarded because the optical densities of blank determinations were large, and reagents giving greater sensitivity produced smaller blank values. Accordingly, stilbazo, Eriochrome cyanine R and catechol violet were selected for more detailed experimental study. TABLE I COMPARISON OF SELECTED CHROMOGENIC REAGENTS Molar extinction coefficient/ Reagent 1 mol-l cm-1 6.7 x lo4 (ref. 8) (mean of values in refs. 14-16) Eriochrome cyanine R , . Catechol violet . . 6.3 x 1 0 4 Stilbazo . . . . 3.8 x 1 0 4 Aluminon . . . . -2.0 x 104 Chrome azurol S . . 2.2 x lo4 Alizarin red S . . . . 1.0 x 10'1 Arsenazo .. . . 1.2 x 104 Wavelength of maximum absorption/ nm 535 580 520 525 568 480 580 Optical density of blank (10-mm cuvettes) 0.1 0.08 0.15* 0.06 0.5 0-3 0.04" Adherence to Beer's law References Slight 8,17 deviation Yes 14, 16, 16 Slight 8 Slight 4 Slight 4 deviation deviation Yes 7 deviation Yes 8 * Values measured by the present authors.7 Addition of calcium as well as alizarin allows a value of approximately 1.8 x lo4 to be achieved.8 EXPERIMENTAL REAGENTS, APPARATUS AND TECHNIQUE- Analytical-reagent grade materials were used whenever available, and de-ionised water was used throughout ; the aluminium content of this water was 0.003 mg 1-1 or less. Standard aluminium solutions were prepared by dissolution of aluminium wire (purity at least 99.9 per cent.) in hydrochloric acid, and dilution to appropriate volumes; all standard solutions were 0-1 N with respect to hydrochloric acid. Optical densities were measured with a Hilger and Watts Uvispek spectrophotometer using 10 or 20-mm cuvettes, the reference cuvette being filled with water. Whenever appropriate, the order of analysis of a batch of determinations was random, and each result given is the mean of two independent determinations, except when stated otherwise.During the work, the temperature of the laboratory varied between 16 and 30 "C, usually being in the range 18 to 22 "C. For tests with catechol violet, all conditions were as described under Method, except when stated otherwise. The conditions used for tests with stilbazo and Eriochrome are described in the following section. COMPARATIVE TESTS ON CHROMOGENIC REAGENTS- Stilbazo-For solutions that contained only aluminium, we found that the following procedure was suitable.To 35 ml of sample (0.1 N with respect to hydrochloric acid*) were added 3.0 ml of a 0.02 per cent. m/V aqueous solution of stilbazo and 10.0 ml of a 30 per cent. m/V aqueous solution of hexamine. The optical density was measured 10 minutes later at 500nm using 20-mm cuvettes. Tests showed that the optimum pH was 5.6, and that no additional colour developed if larger amounts of stilbazo were used. Colour develop- ment was virtually complete within 3 minutes, but the optical density then slowly increased over the the next few hours although the rate of increase was not large enough to cause important errors. * It was decided that samples should be collected into hydrochloric acid (to give a final acidity of 0.1 N) in order to ensure sample stability.416 DOUGAN AND WILSON: THE ABSORPTIOMETRIC DETERMINATION [Analyst, Vol.99 Eriochrome cyanine R-The method of Shull and Guthanl' was used without modification, and preliminary tests showed that their recommendations corresponded closely to the optimum experimental conditions. This method is the same as that tentatively recommended by the American Public Health Association.ll The colour developed within a few minutes, was essentially stable between about 5 and 15 minutes after adding the Eriochrome reagent, and then slowly decreased. All subsequent measurements were made in 10-mm cuvettes, 7 minutes after the addition of Eriochrome. Using each of these reagents and also catechol violet, duplicate analyses of solutions containing 0, 0.02, 0.04, 0-08, 0.12 and 0.20 mg 1-1 of aluminium were made in each of three batches (one batch per day).The effects of orthophosphate and fluoride were also determined. The results from these tests are summarised in Tables 11, I11 and IV. COMPARISON OF Wave- length of maximum Chromogenic absorption/ reagent nm Stilbazo 500 Eriochrome 535 cyanine R Catechol violet 585 TABLE I1 STILBAZO, ERIOCHROME CYANINE R AND CATECHOL VIOLET Optical density of blank (10-mm cuvettes) 0.15 0.07 0.06 Linearity Molar of extinction Standard calibration coefficient/ deviation*/ 1 mol-1 cm-l mg 1-1 of A1 graph Linear in the 1.74 x 10' 0.007 range 0- 0.2 mg 1-1 of A1 deviations from linearity in the range 0- 0.2 mg 1-1 of A1 Linear in the 6-94 x lo4 0.005 range 0- 0.2 mg 1-1 of A1 Small 6.2 x 104 0.009 Stability Effect of of variation coloured of pH,t product,: per cent.per cent. 1.5 <1 4.0 <1 2.0 < I * Pooled value for all concentrations in the range 0 to 0.2 mg 1-1 of aluminium (at least 25 degrees t Percentage change in optical density caused by a change of h0.1 pH unit in a processed $Percentage change in optical density caused by a change of 5 5 minutes in the time a t which of freedom). sample containing 0.2 mg 1-1 of aluminium. the optical density is measured for a sample containing 0.2 mg 1-1 of aluminium. SELECTION OF OPTIMUM REAGENT- Table I1 shows that each of the three reagents can be used for the satisfactory deter- mination of aluminium in solutions that do not contain other impurities. The procedures involving these reagents are all simpler and quicker than the methods involving aluminon ; in addition, Eriochrome cyanine R and catechol violet give much greater sensitivity than aluminon.The molar extinction coefficient found for the aluminium - stilbazo product is much smaller than that reported by Corbett and Guerh8 The difference is possibly caused by the different pH values used by them (6.8) and us (5.6). This point was not investigated further in view of the susceptibility of the stilbazo procedure to interference from phosphate and fluoride (see Tables I11 and IV). TABLE I11 EFFECT OF ORTHOPHOSPHATE Error,* mg 1-l of Al, in the determination of 0.13 mg 1-1 of A1 A I \ Reagent 0.00 mg 1-1 of A1 0.30 mg 1-1 of A1 Catechol violet . . .. o*ooo -0*001 - 0.002 Eriochrome cyanine R .. o*ooo - 0.020 - 0.079 Stilbazo .. .. .. - 0.003 -0.087 -0.219 * The concentration of orthophosphate in samples taken for analysis was 1-7 mg 1-1 of phosphorus. The values given in the table are the differences observed between solutions analysed with and without phosphate. The results are intended primarily to indicate the relative effects of phosphate for each reagent; approximate 95 per cent. confidence limits for each value given are 0.015 mg 1-1 of aluminium.July, 19741 OF ALUMINIUM I N WATER 41 7 Catechol violet has the useful advantages of having the greatest sensitivity and highest wavelength of maximum absorption ; the latter characteristic decreases errors caused by the natural colour and turbidity of certain waters. Catechol violet is also less critically dependent than Eriochrome on the pH of the solution, and the former reagent also gave linear calibration graphs.Our results for catechol violet agree well with those of other workers, which is a good indication of the reliability of the reagent. Thus, Tanaka and Yamayoshi,15 Wilson and Sergeant,l* MeyrowitP and the present authors all obtained linear calibration graphs with molar extinction coefficients of 6.8 x lo4 (580nm), 6.2 x lo4 (580nm), 6.0 x lo4 (580 nm) and 6.9 x lo4 (585 nm), respectively. Further, the optimum pH values found were 6.0, 5.9 to 6.2, 6-1 to 6.2 and 6.1 to 6.3, respectively. TABLE IV EFFECT OF FLUORIDE Concentration in samples taken for analysis Alummiurn/ FluoridE, mg I-' of A1 nig 1-1 of F 0-043 0-5 0.043 1.0 0.13 0-5 0.13 1.0 0.30 0.5 0.30 1.0 Error," mg 1-1 of R1, caused by fluoride 7 A -i Catechol Eriochromet - - 0.002 - 0.022 N.M.- - 0.009 - 0.027 N.M. - 0.035 - 0.002 - 0.034 N.M. - 0.090 - 0.01 1 - 0.069 N.M. - 0.079 - 0.006 - 0.068 -0.136 -0.218 -0.179 - 0.023 - 0.122 Aluminon7 violet cyanine R Stilbazos * The values quoted in the table are the differences observed between solutions analysed with and without fluoride. The results are intended primarily to indicate the relative effects of fluoride for each reagent; approximate 95 per cent. confidence limits for each value quoted are 0.015 mg 1-1 of aluminium. t Calculated from results quoted by Packham." $ N.M. = not measured. Calculated from results quoted by Shull and Guthan." One of the most important sources of error in the absorptiometric determination of aluminium is the interference caused by phosphates and fluoride.The magnitudes of these errors for the three reagents and, for comparison, for aluminon, are given in Tables I11 and IV. These results clearly show that the catechol violet procedure was the least subject to inter- ference. These tests were made only for orthophosphate, and it is well knownl7JB that con- densed inorganic phosphates cause greater interference than orthophosphate when deter- mining aluminium. However, catechol violet is less subject to errors caused by orthophosphate and as condensed inorganic phosphates are easily hydrolysed to orth~phosphate,~J~ it seemed that problems caused by the condensed phosphates would not be insuperable if catechol violet were used.Our results on the effect of orthophosphate when Eriochrome is used are not in agreement with those of Shull and Guthad' or I<night,l* who reported much smaller interferences. However, we obtained essentially identical results by either of those two procedures. In addition, the effect we observed for orthophosphate was not changed by using Eriochrome cyanine R from different suppliers, including one of those used by Shull and Guthan. It is worth noting that Pakalns7 also reported that orthophosphate caused marked interference when Eriochrome was used. On the basis of the above results, it was concluded that catechol violet has useful advan- tages over the other reagents commonly used for the absorptiometric determination of aluminium in water.The detailed development and testing of a method involving the use of catechol violet is described below. DEVELOPMENT OF CATECHOL VIOLET METHOD- Preliminary choice of experimental conditions-In order to ensure stability of the alu- minium content of samples between sampling and analysis, it was decided that samples would be collected into hydrochloric acid so that their final acidity was 0.1 N. It was therefore necessary to add a buffer solution to acidified samples in order to achieve the optimum pH (5.9 to 6-2) reported14-16 for the formation of the aluminium - catechol violet complex. A418 [Analyst, Vol. 99 buffer based on hexamethylenetetramine (hexamine) was chosen , and tests showed that it gave a satisfactory control of pH. The proportions of sample and buffer volumes specified under Method were chosen so as to obtain maximum sensitivity.Samples and standard solutions were analysed in small polyethylene bottles in order to reduce the possibility of contamination from glassware. The method was intended for use in many laboratories, and calculations indicated that, when certain absorptiometers were used, samples containing aluminium at concentrations greater than about 0.3mg1-1 would require dilution before analysis in order to achieve adequate accuracy in reading the absorptiometer scale. Accordingly, that concentration was used as the upper limit in subsequent work. Tests showed that the absorption peak of the aluminium - catechol violet complex was 585 nm (see Fig. l), and that wavelength was used throughout. Efect of $H-Solutions containing either 0 or 0.3 mg 1-1 of aluminium were analysed as described under Method, except that hexamine and various amounts of ammonia solution were added separately so as to achieve different pH values in the final solutions.The effects. of pH on the optical densities of the blank and the standard are indicated in Fig. 2, which shows that a pH of 6-1 to 6.2 was suitable; this value was used in all subsequent work. DOUGAN AND WILSON : THE ABSORPTIOMETRIC DETERMINATION 01 I I I 1 I 1 500 520 540 560 580 600 620 Wavelengthhm Fig. 1. Absorption spectra: A, 0.3 rng 1-I of aluminium; and B, blank containing no aluminium. Solutions contained in 10-mm cuvettes 0.02 1 5.4 5.6 5.8 6.0 6.2 6.4 6.6 pH value Fig. 2. Effect of pH value: A, 0.3 mg 1-1 of aluminium; and B, blank containing no alu- minium.Solutions contained in 10-mm cuvettes Efect of amouptt of catechol violet-Solutions containing either 0 or 0.3 mg 1-1 of aluminium were analysed as described under Method, except that various amounts of catechol violet were added. The results are given in Table V, which shows that 3 ml of 0.0255 per cent. m/V reagent were adequate for determining up to 0.3 mg 1-1 of aluminium. This amount of reagent was therefore used in all subsequent work, but, for convenience, it was added as 2 ml of a 0.038 per cent. m/V solution. Other tests indicated that, with this amount, the calibration graph was linear up to approximately 0.4 mg 1-1 of aluminium. TABLE V EFFECT OF THE AMOUNT OF CATECHOL VIOLET ON OPTICAL DENSITY Volume of 0.0255 per cent.Aluminium concentration/mg 1-' of A1 m/ V catechol violet A f 3 solution added/ml 0.0 0.3 0-3* 2.0 0-046 0.510 0.464 3.0 0-068 0.584 0-5 16 4.0 0-074 0.586 0.512 * Values corrected for the blank.July, 19741 OF ALUMINIUM I N WATER 419 Variations in the shape of the calibration graph when different batches of reagent are used have been reported for chromogenic reagents used for aluminium. In order to check this effect for catechol violet, several batches of the reagent were used in analyses of solutions containing 0 or 0.3 mg 1-1 of aluminium. The results are given in Table VI, which shows that the reagent has a good shelf-life, and that there was little important variation among the different batches. TABLE VI OPTICAL DENSITY RESULTS FROM DIFFERENT BATCHES OF CATECHOL VIOLET Date of manufacture of batch June, 1964 .... April, 1967 . . .. May, 1967 . . . . May, 1968 . . .. December, 1964 . . February, 1971 .. Aluminium concentration/mg 1-1 of A1 -- 0.0 0.3 0-3* . . 0.076 0.606 0.530 . . 0.083 0.623 0.540 . . 0.082 0.609 0.527 . . 0.063 0.581 0.518 . . 0-075 0-605 0.530 . . 0.074 0.598 0.524 * Values corrected for the blank. Rate of formation of aluminiwn - catechol violet complex-Solutions containing 0, 0.1 and 0.3 mg 1-1 of aluminium were analysed as described under Method, but the optical densities of the processed solutions were measured at different times after adding the buffer. The results are given in Fig. 3, which shows that, although the optical density of each solution increased with time, the rate of increase was so small that it did not constitute a problem in measure- ment.In all subsequent work, optical densities were measured between 10 and 20 minutes after addition of buffer, but other times could equally well be used. 0.59 I 1 Time of colour development/hours Fig. 3. Rate of colour development: A, 0.3 mg 1-1 of aluminium; B, 0.1 mg 1-1 of aluminium; and C, blank con- taining no aluminium. Solutions contained in 1 0-mm cuvettes Efects of temperature and sunlight-Solutions containing 0 and 0-3 mg 1-1 of aluminium were analysed at 5, 22 and 40 "C and in diffuse daylight and bright October sunshine. For the blank, no effect greater than 0.004 optical density unit was found, the effect of sunlight was not significant, and the optical density of the aluminium - catechol violet complex increased by about 0.2 per cent."C-l. Efect of ironTWilson and Sergeant14 found that iron interfered seriously and we con- firmed this result. They overcame this effect by adding hydroxylammonium chloride to samples, in order to reduce iron to the iron(I1) state, followed by 1,lO-phenanthroline, in420 [Analyst, Vol. 99 order to forrn the stable iron(I1) chelate. Calculations indicated that 1 ml of a 0-1 per cent. m/V solution of 1,lO-phenanthroline would be sufficient for 2 mg 1-1 of iron in a sample, and it was also decidedlg to use 1 ml of a 10 per cent. m/V solution of hydroxylammonium chloride; for simplicity, these two reagents were combined in one solution. The effect of different amounts of the combined reagent on the determination of aluminium in the absence of iron was first tested, with the results given in Table VII.DOUGAN AND WILSON : THE ABSORPTIOMETRIC DETERMINATIOK TABLE VII EFFECT OF A MIXTURE OF ~,~O-PHENANTHROLINE AND HYDROXYLAMMONIUM CHLORIDE ON OPTICAL DENSITY Aluminiuni concentration/mg 1-1 of A1 7- - adcled/ml 0.0 0.3 0.3" 0 0.104 0.658 0.554 0.6 0.069 0.609 0-540 1.0 0-064 0.600 0.536 2.0 0-061 0.57 1 0.510 Volume of mixed reagent * Values corrected for the blank. From the results in Table VII, it was decided that the addition of 1 ml of the combined reagent was suitable. Tests were made of the effect of the reagent on the interference by iron, with the results given in Table X. These results indicated that the effect of iron was satisfactorily small for our purposes, and the use of tlie mixed reagent was adopted in subse- quent work, Eflect of time periods between additions of reageizts-Tests showed that no errors greater than 0.005 mg 1-1 of aluminium were caused by variations in the time period between the addition of the 1,lO-phenanthroline and the catechol violet reagents from 20 s to 12 to 14 minutes.When the period between the addition of the catechol violet and buffer reagents was varied between 20 s and 2.5 hours, the largest error was 0.003 mg 1-1 of aluminium. It was concluded that the timing of the addition of reagents was not critical. Stability of reagents and standard solzttions-A set of all reagent solutions was prepared as described under Method and stored on tlie laboratory bench.Ten days later, another set was freshly prepared, and both sets were used in order to analyse duplicate portions of solutions containing 0 and 0-3 mg 1-1 of aluminium. This experiment was repeated after 5 and 11 weeks using the original and freshly prepared reagents. The results (Table VIII) indicate that the reagents were adequately stable for at least 11 weeks. TABLE VIII OPTICAL DENSITY RESULTS SHOWING STABILITY OF REAGENT SOLUTIONS Aluminium concentration/nig 1-1 of A1 I A -l Reagent 0.0 0.3 0-3* Freshly prepared . . . . 0.040 0.570 0-530 10 days old . , .. . . 0.042 0.560 0.5 18 Freshly prepared . . . . 0.059 0.596 0.537 5 weeks old . . .. . . 0.065 0.602 0.537 Freshly prepared . . . . 0.038 0.567 0.529 11 weeks old . . . . . . 0.054 0.574 0-520 * Values corrected for the blank.Similar tests were made in order to compare old and freshly prepared standard aluminium solutions. The results showed that solutions containing 500 and 2.5 mg 1-1 of aluminium were stable to within 1 per cent, for periods of at least 2 years and 6 months, respectively.July, 19741 OF ALUMINIURII IN WATER 42 1 DETERMINATION OF ALUMINIUM IN THE WATER USED FOR BLANK DETERMINATIONS- The procedure described under Method requires a blank determination, the result for which is subtracted from the result for a sample in order to obtain the concentration of aluminium in the sample. If the water used for the blank determination contains aluminium, the blank correction will be too large, and results for samples falsely low. It is necessary, therefore, to be able to determine the concentration of aluminium in the water used for the blank.Several procedures were examined as the basis for this determination. The only approach found to be suitable was to evaporate a portion of the blank water, in order to concentrate any aluminium present, and to analyse two blanks, one using the evaporated water and the other using unevaporated water. The difference between the two results can then be used to calculate the aluminium content of the original water. This procedure has the disadvantage that the water being evaporated may become contaminated, but our tests showed that any such error could be adequately controlled for our purposes. NEED FOR PRE-TREATMEXT OF SAMPLES- Samples may contain forms of aluminium that do not react with catechol violet, e.g., complexed or undissolved species.Thus, when the total aluminium content is of interest, preliminary treatment of samples may be needed in order to convert all forms of aluminium into forms that react with the reagent. The need for pre-treatment was therefore investigated. Several treatments of different severities (see below) were applied to different surface waters that had been treated with aluminium sulphate. Two types of water were collected from most plants: (l), alum-treated water after coagulation and sedimentation; (2), the same as (1) but after filtration in the water treatment plant. All samples were collected into hydrochloric acid as described under Method, and certain samples were diluted with water after the treatment procedures had been applied so that the aluminium concentrations in the solutions analysed were less than 0.3 mgl-l.Treatment 1-The sample was analysed without further treatment as described under Method as soon as possible after collection. Treatment 2-A 35-ml volume of the sample was heated just to boiling in a covered 100-ml Pyrex-glass beaker. The sample was then cooled and analysed as described under Method. Treatment 3-A 35-ml volume of the sample was placed in a polyethylene bottle, which was then loosely capped and heated in a water-bath at 80 "C for 30 minutes. The sample was cooled and analysed as described under Method. Treatment 4-This was the same as treatment 2, but the sample was evaporated to approximately 10ml and then, after cooling, diluted to 35ml with water.Treatment 5-A 35-ml volume of the sample was placed in a 100-ml silica beaker, and 2 ml of sulphuric acid (sp. gr. 1.84) were added. The beaker was heated on a hot-plate until white fumes were evolved, then 2 ml of nitric acid (sp. gr. 1.42) were added cautiously in small portions, each portion being added when brown fumes were no longer observed. The beaker was then cooled, its contents were cautiously diluted with water to 20 ml, and three drops of 0.02 per cent. mlV m-cresol purple indicator solution added. The solution was neutralised by adding 9 ml of dilute ammonia solution (1 + l), cooling and then cautiously adding ammonia solution (1 + 3) until the pH was 3.0 (a pH meter and magnetic stirrer were used). The neutralised solution was transferred into a 60-ml polyethylene bottle, diluted with water to 35 ml and analysed as described under Method, except that 5 ml of 12 per cent.m/V hexamine solution were added in place of the 10 ml of hexamine buffer solution. Tests showed that the optical density obtained for a given concentration of aluminium was 12 per cent. less than that for the other four procedures. This effect was caused by the large concentration of sulphate in the treated sample, and was corrected for by using an appropriate calibration graph. Before analysing samples, blanks and standard solutions were analysed by each of the above treatments so as to ensure that adequate accuracy could be achieved. Treatments 2 , 4 and 5 gave rather worse precision than 1 and 3.Therefore, three portions of each sample and blank were analysed by procedures 2, 4 and 5 and duplicate analyses were made for procedures 1 and 3. The results of these tests are given in Table IX, and two points should be noted. (i) The 95 per cent. confidence limits for each result in the table depend on the treatment method,422 [Artalyst, Vol. 99 the aluminium concentration and the degree of dilution (if any) of the sample before analysis. Approximate values for these limits can be summarised for simplicity as follows: DOUGAN AND WILSON : THE ABSORPTIOMETRIC DETERMINATION > O . O l mg 1-1 of A1 >Om02 mg 1-1 of A1 >0.04 mg 1-1 of A1 . . . . . . . . . . . . For all treatments for all filtered waters and for the settled For all treatments for the settled water from the River Severn For all treatments for the settled waters from plants 1 and 2 on the River Thames and for treatments 1, 2, 3 for the settled water from plant 3 on the River Thames For treatments 4 and 6 for the settled water from plant 3 on the River Thames.(ii) The samples from plant 1 on the River Thames were re-analysed 4 days after collection using treatment 1. The results for the settled and filtered waters were 0.545 and 0.099 mg 1-1 of aluminium, respectively. water from the River Churnet >0.06 mg 1-1 of -41 . . . . TABLE IX EFFECT OF PRELIMINARY TREATMENT METHODS ON THE APPARENT ALUMINIUM CONCENTRATIONS OF SAMPLES Apparent concentration of aluminiumlmg 1-l of A1 Treatment Treatment Treatment Treatment Treatment r A 1 Sample 1* 2 3 4 6 River Thames, plant It- Settled water .. . . . . 0-478 0.577 0.530 0.558 0.551 Filtered water . . . . . . 0.094 0.100 0.093 0.103 0.093 Settled water . . . . . . 0.324 0.333 0.305 0.256 0.330 Settled water . . . . . . 1.066 0.988 1.027 1.086 1-174 Filtered water . . . . . . 0.161 0.161 0-162 0.167 0-163 Settled water . . . . . . 0.260 0.240 - - 0-258 Filtered water . . . . . . 0.043 0.040 c - 0.030 - - 0.204 Settled water . . . . . . 0.186 - - 0.039 Filtered water . . . . . . 0.045 * These analyses were made at the following times after sample collection: 18, &, 4, 5 and t Magnafloc LT25 coagulant aid also used in the plant. $ Magnafloc LT22 coagulant aid also used in the plant. River Thames, plant 2- River Thames, plant 37- River Severn- River Clawnet$-- - - 18 hours (in the order of sampling locations in the table). The results in Table IX indicate that preliminary treatment (other than collection of the sample into acid) is not generally required.Accordingly, the recommended method of analysis described in the following section does not include any specific pre-treatment pro- cedure. METHOD PRE-TREATMENT OF SAMPLES- The method specifies collection of samples into hydrochloric acid, and this acidification alone may well suffice to convert all forms of aluminium into those which react with catechol violet. This method should be used whenever possible because of its simplicity, rapidity and ability to give good precision. However, certain samples may contain “non-reactive” forms of aluminium (e.g., undissolved or complexed species) so that, when the total aluminium content is of interest, some preliminary treatment of samples is needed in order to convert all of the aluminium into “reactive” forms.Each analyst should ensure that a pre-treatment appropriate to his samples is used. If a pre-treatment is used, care is needed to ensure that the pH values of the processed samples are between 6.0 and 6.2. REAGENTS- Use analytical-reagent grade materials whenever possible. Water-Use distilled or de-ionised water for blank determinations and for preparing standard and reagent solutions. The water should preferably have an aluminium content that is negligible in comparison with the smallest concentration to be mcasurcd in samples.July, 19741 OF ALUMINIUM IN WATER 423 De-ionised water was used in this work and contained not more than 0.003 mg 1-1 of aluminium.The aluminium concentration in the water can be determined as described under Procedure. Hydrochloric acid, 5 N (& 0.02 N). Hydrochloric acid, 0.1 N (* 0.001 N)-Prepare this acid by dilution of 5 N hydrochloric acid with water, 1,lO-Phenanthroline solution, 0.1 per cent. m/V-Dissolve 50 g (& 0.5 g) of hydroxylam- monium chloride in approximately400 ml of water, and dissolve in that solution 0.5 g (& 0.005 g) of 1,lO-phenanthroline hydrate. A slight pink coloration of the resulting solution is un- important. Transfer the solution into a 500-ml calibrated flask, dilute to the mark with water and mix. When the solution was stored in a polyethylene bottle, it was adequately stable for at least 11 weeks.Catechol violet solution, 0.0375 per cent. m/V-Dissolve 0.075 g (& 0.001 g) of catechol violet in approximately 15 ml of water. Transfer the solution into a 200-ml calibrated flask, dilute to the mark with water and mix. When the solution was stored in a Pyrex-glass flask, it was adequately stable for at least 11 weeks. On standing, spore growth was observed in this solution; the spores were avoided when dispensing the solution by pipette. The growth of spores was less in distilled than in de-ionised water, and could also be decreased by storing the solution in a refrigerator. The use of chemical inhibitors was not investigated. Hexamine bufer solution, 30 per cent. m/V-Dissolve 150 g (& 0-5 g) of hexamine in approxi- mately 350 ml of water and filter the solution through a Whatman GF/C filter if undissolved foreign matter is present.Transfer the solution into a 500-ml calibrated flask, add 8.4 ml (& 0.1 ml) of ammonia solution (sp. gr. O.SSO), dilute to the mark with water and mix. (The ammonia solution should be taken from a fresh stock.) When this solution was stored in a polyethylene bottle, the cap of which was replaced directly after use, the solution was adequately stable for at least 11 weeks. Standard aluminium solution A-Dissolve 0-420 g (& 0.001 g) of aluminium wire (99.9 per cent. purity is suitable) in 20 ml of hydrochloric acid (sp. gr. 1.18). Transfer the solution into a 1-litre calibrated flask, dilute to the mark with water and mix. When this solution is stored in a polyethylene bottle, it should be adequately stable for at least 2 years.Standard aluminiwz solution B-Transfer by pipette 5.0 ml of standard solution A into a 1-litre flask, add 20 ml (& 0.1 ml) of 5 N hydrochloric acid, dilute to the mark with water and mix. When this solution is stored in a polyethylene bottle, it should be adequately stable for at least 6 months. APPARATUS- For the highest accuracy, it is preferable to reserve the necessary apparatus solely for aluminium determinations. Detergents and chromic acid should not be used for cleaning the apparatus; dilute hydrochloric acid (1 + 9) was found to be satisfactory. Before analysing samples, it is also recommended that the cleanliness of apparatus be checked by carrying out a set of blank determinations (see Procedure).Polyethylene bottles-Two types are required, both of which should have caps made solely of polyethylene. (a) Thick-walled, rigid bottles of any convenient size are used for collecting samples. Make a mark on the side of each sample bottle to indicate the volume to be collected. (b) Small bottles are used for analysis of samples; a capacity of approximately 60 ml is suitable. 1 ml of solution = 0.42 mg of aluminium. 1 ml of solution = 2.1 pg of aluminium. SAMPLE COLLECTION- when the sample has been collected, the final acidity is 0.1 N (& 0.002 N). collection, take care not to distort the bottle while holding it. PROCEDURE- Analysis of samPle-Add 35 ml (& 0.5 m1) of the sample to a small polyethylene bottle previously washed and shaken so as to remove as much water as possible from it.Add 1 ml (& 0.1 ml) of 1,lO-phenanthroline solution and mix by swirling, add 2 ml (& 0-05 ml) of catechol violet solution and mix by swirling, then add 10 ml (& 0.1 ml) of the hexamine buffer Add sufficient 5 N hydrochloric acid to a sample bottle, with a mark on its side, so that, During sample424 DOUGAN AND WILSON : THE ABSORPTIOMETRIC DETERMINATION [Analyst, Vol. 99 solution, replace the cap on the bottle and shake it well. (In a batch of samples, each reagent can be added to all samples before adding the next reagent. After addition of the buffer, the pH of the solution should be in the range 6.0 to 6.2; check that this pH is achieved each time a fresh buffer solution is prepared.) Between 10 and 20 minutes after adding the hexamine buffer solution, measure the optical density of the solution at 585 nm using 10-mm cuvettes, the reference cuvette being filled with water.Let the optical density be A,. Blank determination-Add 35 ml (5 0.5 ml) of 0.1 N hydrochloric acid to a small poly- ethylene bottle and treat this solution as though it were a sample. The final blank solution should have a pale yellow colour; a blue or green colour indicates either an incorrect pH or contamination by aluminium. Let the optical density be A , . Correction foy colour and turbidity-If samples are appreciably coloured, turbid, or both, a correction is required. This correction should seldom be necessary for treated waters, but details are given for completeness.Analyse a second 35-ml portion of sample as described under Analysis of sample, but add 3 ml of water instead of the 1,lO-phenanthroline and catechol violet solutions. Let the optical density be A , . Determination of aluminium in the water used for the blank determination-This deter- mination is not necessary if the aluminium content of the water used for the blank is either known or negligible. Add 1 .O ml (& 0.02 ml) of 5 N hydrochloric acid to a 100-ml calibrated Pyrex-glass beaker followed by 50 ml (& 1 ml) of the water under test. Cover the beaker with a Pyrex-glass watch- glass and evaporate the mixture on a hot-plate until the volume is reduced to approximately 20 ml. Add further 50-ml(& 1 ml) portions of water, evaporating the mixture to 20 ml after each addition, until a total of 250 ml of water has been added and the volume in the beaker is 50 ml (& 2 ml). Cool the beaker and analyse a 35-ml portion of its contents as described under Analysis of sampls.Calculation. of reszclts-Calculate the optical density, A,, due to aluminium in the water used for the blank from the equation Let the optical density be A , . A, = ( A , - A,)/4. Calculate the aluminium content, C,, of the water from A , and the calibration graph. Calculate the apparent optical density, A a , due to aluminium in the sample from the equation A , = A , - A , or, when a correction for colour, turbidity, or both, is necessary, from the equation A , = A , - A , - A , + A , where A,, is the optical density of the sample cuvette measured against the reference cuvette when both are filled with water.Calculate the apparent aluminium content, C,, of the sample from A a and the calibration graph. The true aluminium content, Ct, of the sample is given by Ct = Ca + C w NOTES- 1. If the sample contains condensed inorganic phosphates, place the small polyethylene bottle (with the cap fitting loosely) containing the 35-ml portion of sample in a boiling water bath for 2 hours. Cool, and then proceed with reagent additions as described under Aizalysis of sample. 2. Fluoride interferes although, for concentrations of 1 mg 1-1 of fluoride and less, the effect will usually be negligible (see Results). If i t is required to correct for the effect of fluoride and the fluoride content of samples is essentially constant, the appropriate amount of fluoride can be added to the blank and standard solutions used to prepare the calibration graph. Otherwise, the fluoride must be removed4 by adding sulphuric acid to the sample, evaporating the mixture to dryness, dis- solving the residue in 35 ml (& 0-5 ml) of 0.1 N hydrochloric acid and then analysing the solution as described under Analysis of sample.3. The method has been thoroughly tested for the range 0 to 0.3 mg 1-1 of aluminium. Therefore, although the calibration graph is linear up to approximately 0.4 mg 1-1 of aluminium, it is recom- mended that smaller portions of sample, V ml, be used for analysis when its concentration is likely appreciably to exceed 0.3 mg 1-1 of aluminium. When this is done, the portion of sample should be diluted to 35 ml (f 0.5 ml) with 0.1 N hydrochloric acid before analysis.The aluminium concentra- tion in the original sample, C,, is then calculated from the equation where C, is the concentration of aluminium in the diluted sample. c, = (35/V)C, + c,July, 19741 OF ALUMINIUM IN WATER 425 PREPARATION OF CALIBRATION GRAPH- To a series of small polyethylene bottles, add 30, 31, 32, 33, 34 and 35 ml (all *0.5 ml) of 0.1 N hydrochloric acid and then add to these bottles 5.0, 4.0, 3.0, 2.0, 1.0 and O-Oml, respectively, of standard aluminium solution B. Analyse each of these solutions as described under Analysis of sa.rptpZe. These determinations should be repeated at least once on another day, and then again as necessary until the calibration graph is defined with the required accuracy.Normally, two batches of determinations should suffice. Subtract the average optical density of the blank from the average optical densities of the other solutions and plot the corrected results against the concentration of aluminium. The above solutions are equivalent to 0.306, 0-245, 0.184, 0.122, 0.061 and 0.000 mg 1-1 of aluminium, respectively; these concentrations allow for the dilution of samples by the acid into which they are collected. For measurements at 585nm, the calibration graph is linear up to approximately 0.4 mg 1-1 of aluminium. When absorptiometers and Ilford No. 606 colour filters are used, the graph is also linear, but the optical densities for a given concentration of aluminium are approximately 83 per cent.of those measured at 585 nm. For measurements at 585 nm, the slope of the calibration graph increases by approximately 0.2 per cent. for an increase in temperature of 1 "C. RESULTS A number of tests of the performance of the method were made, and are described below. EFFECT OF OTHER SUBSTANCES- Solutions containing either 0 or 0.3 mg 1-1 of aluminium and 0.1 N with respect to hydrochloric acid were analysed as described under Method in the presence and absence of other impurities. The organic coagulant aids tested were chosen to represent the main classes of compounds in use. The results of these tests are given in Table X. Preliminary tests showed that condensed inorganic phosphates resulted in markedly low results, presumably owing to formation of complexes with aluminium.For example, in determining 0-3 mg 1-1 of aluminium, pyrophosphate in concentrations of 0-2, 0-4 and 1-0 mg 1-1 of phosphorus caused negative errors of approximately 0.08, 0-15 and 0.3 mg 1-1 of aluminium, respectively. It was thought that this effect would be readily overcome by heating the acidified sample before analysis so that the condensed phosphates were hydrolysed to orthophosphate, whose effect is very much smaller (see Table X). Tests showed that hydrolysis was more rapid at 100 "C than 80 "C, and that a heating time of 2 hours should usually suffice. Using this time and the higher temperature, the effects of different phosphates were determined, with the results given in Table XI. The slight effects caused by Calgon S can be attributed to the orthophosphate that remains after hydrolysis.Fluoride commonly interferes in absorptiometric methods for aluminium, and the results in Table XI1 show the magnitude of this error for the catechol violet method. Other tests showed that the effect of 1 mg 1-1 of fluoride on the determination of 0-3 mg 1-1 of aluminium was not reduced by using twice the normal amount of catechol violet. An exploratory test indicated that boric acid might be useful in reducing the fluoride effect. For example, when 1 ml of a 1 per cent. m/V solution of boric acid was added to samples, the observed effects of 1 mg 1-1 of fluoride on the determination of 0-0 and 0.3 mg 1-1 of aluminium were +0-002 and -0.007 mg 1-1 of aluminium, respectively. This amount of boric acid had no substantial effect on the interference caused by 4 mg 1-1 of fluoride.The use of boric acid was not further investigated as, for our purposes, the effect of fluoride was considered to be sufficiently small not to warrant the complication of an additional reagent. PRECISION AND CALIBRATION GRAPH- On each of 10 days, duplicate analyses were made of the following standards and samples, all of which were 0.1 N with respect to hydrochloric acid: (i) de-ionised water, i.e., the blank; (ii) de-ionised water + 0.05 mg 1-1 of aluminium; (iii) de-ionised water + 0.15 mg 1-1 of aluminium; (iv) de-ionised water + 0.30 mg 1-1 of aluminium;426 DOUGAN AND WILSON : THE ABSORPTIOMETRIC DETERMINATION [Analyst. Vol . 99 (v) sample of treated water derived from the River Thames; (vi) as (v). + 0.1 mg 1-1 of aluminium .TABLE X EFFECTS OF OTHER SUBSTANCES ON THE DETERMINATION OF ALUMINIUM Other substance Ca2+ . . Ca2+ . . Mg2+ . . Na+ .. K+ . . Alkalinity Alkalinity SO,,. .. NO,. . . NO,. . . ~ 0 ~ 3 - . . ~ 0 ~ 3 - . . Po4,. .. SiOSa- . . Humic acid Fulvic acid Detergentst Chlorine . . Chlorine .. Ammonia Co*+ . . Ni*+ . . Cda+ . . cu*+ . . Cu*+ . . Zn*+ . . Pb*+ . . PbB+ . . Fe8+ . . Fe8+ . . Mn*+ . . Mn*+ . . CrS+ . . Cr3+ . . Cra+ . . Snz+ . . Sn2+ . . .. . . . . .. .. .. .. .. .. .. . . . . .. .. . . .. .. .. . . .. .. . . .. .. . . .. .. .. .. . . . . .. .. . . . . .. .. Coagulant aids- B.T.I. A100 . . B.T.I. A150 . . Magnafloc LT24 Wisprofloc 20 Wisprofloc P .. . . .. .. .. .. . . . . .. . . . . .. . . .. . . .. . . . . .. .. . . .. . . .. .. .. . . . . . . . . . . .. .. .. .. . . . . .. .. .. . . . . .. .. .. . . I . . . .. .. . . .. . . . . .. . . . . .. .. .. .. . . .. . . . . .. .. . . .. .. . . . . . . . . . . .. .. . . . . . . . . . . Concentration of other substance/ mg 1-1 500 100 100 100 50 300 (as CaCO,) 200 (as CaCO,) 100 so 10 8.2 (as P) 4.1 (as P) 1-7 (as P) 10 5 5 50 10 (as } 5 0.5 (as N) } 2 2 2 2 1 2 2 1 1 0.3 2 1 0.5 0.25 0.025 2 1 0.5 0.5 0-5 n .. 3 .. 3 Effect" of other substance. mg 1-1 of Al. at an aluminium concentration of 7 0-000 mg 1- + 0.006 + 0.003 + 0-004 + 0.002 + 0.005 + 0-006 0-000 + 0.001 + 0.005 + 0.002 - 0.005 - 0.005 0.000 + 0.009 + 0.003 + 0.004 + 0-006 + 0. 004 + 0.003 + 0.007 - 0.002 0.000 + 0.003 + 0.005 - 0.050 - 0.038 - 0-002 + 0.005 + 0.00 1 .0.004 . 0.003 +0.001 . 0.042 0.000 + 0.001 0~000 . 0.002 . 0.002 . 0.003 + 0.034 0-000 o*ooo + 0.052 + 0.004 + 0.01 1 + 0.006 +0.012 0.000 + 0.004 + 0.002 + 0.005 + 0.002 + 0.002 + 0.005 + 0.002 + 0.002 + 0.002 + 0.003 . 0.009 . 0.006 . 0.003 + 0.006 + 0.003 +0.010 + 0.007 o*ooo +0*014 + 0.004 + 0.006 - 0.006 - 0.028 -0.015 - 0.002 -0.016; - 0.007 + 0.003 + 0.002 +0*001 + 0.003 + 0.003 * If the other substances had no effect. results would be expected to lie (95 per cent . confidence limits) within the following ranges : 0.000 5 0.003 for 0.000 mg 1-l of A1 0.000 & 0.006 for 0.300 mg 1-1 of A1 t Daz. Dreft. Onio. Quix. Surf and Tide (equal proportions by mass were used) . The results for (ii) to (vi) were corrected for the blank. and then analysed statistically20 in order to obtain estimates of the standard deviations corresponding to within-batch (s.) and between-batch ( s b ) variations . These results are summarised in Table XIII. which also shows that the calibration graph is linear . Other tests showed a 7 per cent . deviation from linearity at a concentration of 0-6 mg 1-l of aluminium .Jul~7, 19741 OF ALUMINIUM IN WATER 427 TABLE XI EFFECTS OF CONDENSED INORGANIC PHOSPHATES AFTER HEATING ACIDIFIED SAMPLES Substancc Concentration of substance/ mg 1-1 Pyrophosphate . . . . 1 (asp) Hexametaphosphate . . . . 1 (as P) Tripolyphosphate . . . . 1 (asp) Calgon St . . .. - . 10 Calgon S .. . . . . so: Effect* of substance, mg 1-1 of Al, a t an aluminium concentration of 0.000 mg 1-1 0.300 mg 1-1 A I \ + 0.003 - 0.005 + 0.001 - 0.002 0~000 + 0.004 + 0.00 1 - 0.008 + 0.001 - 0.037 * 95 per cent.confidence limits for the results are the same as those in Table X. Calgon S is a glassy sodium polyphosphate used in water treatment; 10 mg 1-l of Calgon S is approximately equivalent to 4.1 mg 1-1 of phosphorus. These tests were made over a period of 40 days; in that time, neither of the two water The stability $ Heating period, 3 hours. samples showed any systematic change in their aluminium concentrations. of acidified samples, therefore, is satisfactory. TABLE XI1 EFFECT OF FLUORIDE Concentration Effect" of fluoride, mgl-1 of Al, a t an aluminium concentration of of fluoride/ f A \ mg 1-1 0.000 mg 1-I 0.043 mg 1-1 0.130 mg 1-1 0-300 mg 1-1 - 0.01 1 0.5 - 0.002 - 0.002 - 0.002 1.0 - 0.006 - 0.009 -0.011 - 0.023 4.0 - 0.006 - - - 0.192 * 95 per cent.confidence limits for the results are the same as those in Table X. ROBUSTNESS OF THE METHOD- As a check on the ability of the method to withstand variations in the amounts of reagents added to samples, a factorial experiment was carried out using solutions containing 0 or 0.3 mg 1-1 of aluminium. The results are given in Table XIV. TABLE XI11 PRECISION OF ANALYTICAL RESULTS Standard deviation*/ Mean mg 1-l of A1 Concentration optical r found/ Recovery, Solution density S W s b s t t mg 1-1 of A1 per cent. Standard solutionslmg I-' of Al- 0.00 0.059, 0.0011 0.0062: 0.0063 - - 0.05 0-151, 0.0021 0.00359 0.0040 - - 0.15 0.336, 0.0029 0*0054§ 0.0061 - - 0.30 0.599, 0.0029 0.0073$ 0.0079 - - Treated River Thames 0.244, 0.0028 0.003511 0.0044 0-102 - Same + 0.1 mg 1-1 of A1 0.419, 0.0023 0*0050$ 0.0055 0.199 99-77 Samples- * The degrees of freedom for s w and s b are 10 and 9, respectively.t st is the standard deviation of any one result in any one batch of analyses. 2 Signifies significance a t 0.001 probability level. Fj Signifies significance a t 0.01 probability level. 11 Signifies significance at 0.05 probability level. 7 Allowance has been made for the dilution of the sample caused by the addition of the standard aluminium solution : correction approximately 2.8 per cent. The results show that satisfactorily small errors were caused by the variations in the amounts of reagent tested. Other tests, in which all reagent volumes except the acid added to samples were closely controlled, showed no significant variations when the acidity of samples varied between 0 .0 9 5 and 0 . 1 0 5 ~ .428 DOUGAN AND WILSON THE ABSORPTIOMETRIC DETERMIK\;ATION [A?ZdySt, 1'01. 99 TABLE XIV EFFECTS OK OPTICAL DENSITY" OF VARIATIONS IN THE AMOUNTS OF REAGENTS ADDED TO SAMPLES Sample 0.098 N in acid r A .l 0.8 ml of 1 , l O - 1.2 ml of 1,lO- phenanthroline phenanthroline solution solution w -------7 1.9 ml of 2.1 ml of 1.9 ml of 2.1 ml of Volume of catechol catechol catechol catechol buffer/ml violet violet violet violet 9.8 0.533 0.546 0.540 0.537 10.2 0.555 0.556 0.538 0.541 (0.048) (0-048) (0.045) (0.044) (0,047) (0.049) (0.048) (0.048) Sample 0-102 N in acid r-- A -l 0.8 ml of 1 , l O - 1.2 ml of 1 , l O - phenanthroline phenanthroline solution solution 7- 1-9 ml of 2.1 ml of 1.9 ml of 2.1 ml of catechol catechol catechol catechol violet violet violet violet 0.542 0.554 0.541 0.541 (0-046) (0.046) (0.044) (0.045) 0.550 0.549 0.540 0.542 (0.051) (0.050) (0.045) (0-044) * Single determinations were made for each condition tested.The values in parentheses are for the blank determinations; the other values are for the 0.3 mg 1-1 of aluminium solution after blank correction. For comparison, the corresponding optical densities for 0.1 N, 1.0 ml of 1,lO-phenanthroline, 2.0 ml of catechol violet and 10.0 ml of buffer were 0.545 and (0.046). INTER-LABORATORY TESTS- A number of members of the Water Research Association collaborated with the authors in a trial of the catechol violet method. After preliminary tests in which each laboratory became familiar with the method, prepared calibration graphs and checked precision, two samples were prepared and portions sent to each laboratory: tap water: spiked with 0-125 mg 1-1 of aluminium and 0.1 N with respect to same as sample 1, but also spiked with 0.2 mg 1-1 of fluoride.Sample 1 . . Sample 2 . . . . . . hydrochloric acid ; Each laboratory analysed each of these samples and made a blank determination once on each of five days. None of the laboratories except our own knew the amounts of aluminium added to samples 1 and 2. The results of these tests are given in Table XV, which indicates a satisfactory performance of the method in six independent laboratories. All of these laboratories adopted the method for their normal analyses.TABLE XV RESULTS OF INTER-LABORATORY TESTS Mean concentration found*/ mg 1-1 of A1 Total standard deviationt/ mg 1-1 of h l Laboratory 1 2 3 4 5 6 Authors.. . . Over-all mean . . Expected value Sample 1 0.117 f 0.015 0.126 & 0.012 0-130 f 0.002 0.116 f 0.005 0-144 f 0.004 0.133 f 0.002 0.132 f 0.003 0.128 0.128 -7 Sample 2 0.119 & 0.011 0.121 f 0.011 0-123 & 0-001 0.118 f 0.002 0.141 f 0-007 0.134 -& 0.006 0.124 & 0.005 0.126 0.128 Sample 1 0.012 0.010 0.002 0.004 0.003 0.002 0.002 - 7 Sample 2 0.009 0.009 0.002 0.002 0.006 0.005 0.004 * 95 per cent. confidence limits are given after each mean. t These estimates each have four degrees of freedom. DISCUSSION NEED FOR PRE-TREATMENT OF SAMPLES- It is simple, in principle, to check the relative efficiencies of different pre-treatment procedures for converting all forms of aluminium into forms that react with catechol violet.This is a hard, bore-hole water. The concentration of aluminium in the tap water was determined to be 0.003 mg 1-*.July, 19741 O F ALUMINIUM I N WATER 429 However, it is more difficult to establish their absolute efficiencies because the true aluminium content of samples is not known. In our work, it was assumed that a wet-oxidation procedure was 100 per cent. efficient, and other treatments were judged in relation to that procedure. This assumption is considered reasonable in view of the vigour of the wet oxidation and the likely forms of aluminium in samples. The collection of samples into hydrochloric acid (as recommended in the method) is itself a form of pre-treatment, and the results in Table IX indicate that this acidification alone was sufficient for all of the filtered waters examined.It is possible that lower results would have been obtained by direct analysis of these samples immediately after collection. However, the samples from plants 1 and 2 on the River Thames were analysed within 16 and 4 hours of collection, respectively, and correspond to the situation that is likely to occur in routine analysis. On this basis, it appears that filtered waters usually require little or no pre-treatment provided that the recommended method of sample collection is used. The results for the settled waters in Table IX are less clear. Direct analysis of the acidified samples sufficed for the samples from plant 2 on the River Thames and from the Rivers Severn and Churnet.The result by direct analysis for plant 1 on the River Thames is significantly smaller than those obtained by the other treatments, and there is a suggestion of the same effect for plant 3 on the River Thames. It is concluded, therefore, that some settled waters may require pre-treatment but it probably need not be very vigorous. The results suggest that the need for pre-treatment may increase as the aluminium concentration in the sample increases and as the period between sampling and analysis decreases. PERFORMANCE OF THE METHOD- The precision achieved (Table XIII) is satisfactory for our purposes, and was essentially the same for standard soluticm and samples of treated water.The results show that for concentrations between 0-05 and 0.3 mg 1-1 of aluminium, the main source of random error was variations in results from one batch of analyses to another. Thus, the precision could be further improved by establishing the calibration appropriate for each batch of analyses. However, the use of one calibration over a number of batches will be adequate for most purposes. The reasons for the significant between-batch variability were not determined, but it seems likely to have been caused mainly by variations in the pH values of processed samples and in the times after processing at which optical density measurements were made. The “criterion of detection”21 (95 per cent. confidence level) was approximately 0.003 mg 1-1 of aluminium.The method is reasonably selective. From the results in Tables X, XI and XI1 and the concentrations of other substances likely to be present in treated waters, only condensed inorganic phosphates and fluoride cause practically important interferences. The effect of those phosphates is almost completely eliminated by heating the sample before analysis ; this procedure is simple and had no deleterious effect on precision. For concentrations of fluoride not greater than 1 mg l-l, the interference will often be negligible, but greater concentrations of fluoride cause much larger errors. No simple procedure for eliminating the effect of fluoride was established, although the use of boric acid (for preferential complexing of the fluoride) may be useful.Thus, if samples contain sufficiently high concentrations of fluoride to cause unacceptable errors, its removal by evaporating to fumes with sulphuric acid4 is necessary. Essentially complete recovery of aluminium added to a treated water was achieved (Table XIII), which again indicates that the accuracy of the method is satisfactory. The results obtained independently by a number of other laboratories (Table XV) also provide good evidence of the suitability of the method. Finally, the tests described under Robustness of the method show that very strict control of the amounts of the 1,lO-phenanthroline and catechol violet is not required. As in other methods for determining aluminium, careful control of the pH of processed samples is neces- sary, but the tolerances allowable on the amounts of acid and buffer are readily achieved.Control of temperature and the time of measurement is not critically important. Experience with the method over a period of approximately 14 years has confirmed its reliability. COMPARISON WITH OTHER REC~OMMENDED METHODS- standard methods for use in the determination of aluminium are: At present, the chromogenic reagents recommended in authoritative collections of430 DOUGAN AND WILSON Department of thc Environment1 . . .. .. .. .. .. . . aluminon British Standards Institution2 . . .. .. .. .. .. .. . . aluminon American Public Health Association et aE.11 . . . . .. .. .. . . Eriochrome cyanine R Our comparison of those reagents with catechol violet indicates that the latter has several advantages. The detailed tests of the performance of the catechol violet method have confirmed the suitability of this reagent, and we would recommend the method in preference to those involving the use of aluminon and Eriochrome cyanine R. We acknowledge the help of Mr. M. J. Beckett, who carried out much of the literature survey. We thank Messrs. J. H. Clarke, D. B. Ford, C. E. Harris, J. Jeffery, B. Priest and W. A. Smith for arranging for the tests of the method in their laboratories. Thanks are also due to the Technical Services Department of BDH Chemicals Ltd. for providing several batches of catechol violet used in this work. We thank Dr. R. G. Allen, Director of the Water Research Centre, for permission to publish this paper, which is based on the Water Research Association’s Technical Paper TP. 103. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. REFERENCES Department of the Environment, “The Analysis of Raw, Potable and Waste Waters,” H.M B.S. 2690 : Part 4 : 1967, British Standards Institution, London, 1967, pp. 5-10. American Public Health Association, American Water Works Association and Water Pollution Control Federation, “Standard Methods for the Examination of Water and Wastewater,” Twelfth Edition, American Public Health Association, New York, 1965, pp. 53-56. Stationery Office, London, 1972, pp. 69-70. Packham, R. F., PFOC. SOC. Wat. Treat. Exam., 1958, 7, 202. Giebler, G., 2. analyt. Chem., 1961, 184, 401. Vezzi, S., Boll. Laboratori Chim. Prov., 1969, 20, 402. Pakalns, P., Analytica Chim. Acta, 1965, 32, 57. Corbett, J. A., and Guerin, B. D., Analyst, 1966, 91, 490. Aginskaya, N. A., and Petrashen, V. I., Trudy Novocherk. Politekh. Inst., 1958, 72, 13. Public Health Service, U.S. Department of Health, Education and Welfare, “Water Metals No. 5,” Study No. 34, Analytical Reference Service, Publication No. 1910, Public Health Service, Cincinnati, 1969. 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, pp. 57-62. Tech. Pap. Wat. Res. Ass., No. TP.73, Water Research Association, Medmenham, 1970. Anton, A., Analyt. Chem., 1960, 32, 725. Wilson, A. D., and Sergeant, G. A., Analyst, 1963, 88, 109. Tanaka, K., and Yamayoshi, K., Bunseki Kagaku, 1964, 13, 540. Meyrowitz, R., Prof. Pnp. U.S. Geol. Surv., No. 700-D, 1970, pp. D225-229. Shull, K. E., and Guthan, G. R., J . Amev. Wut. Whs Ass., 1967, 59, 1456. Knight, A. G., PYOC. SOC. Wat. Treat. Exam., 1960, 9, 72. 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, pp. 189-192. Wilson, A. L., Talanta, 1970, 17, 31. Roos, J. B , Analyst, 1962, 87, 832. Received June 18th, 1973 Amended February 6th, 1974 Accepted Febvuury 12th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900413

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, July, 1974, Vol. 99, pp. 431434 431 Spectrophotometric Determination of Organophosphorus Pesticides with 4- (4- Nitrobenzy1)pyridine BY C. R. TURNER* (Centre for Overseas Pest Research, College House, Wrights Lane, London, W8 5s J ) A spectrophotometric method has been developed for the determination of certain organophosphorus pesticides. It involves the reaction of 4-(4-nitro- benzy1)pyridine with the pesticide in an acetone - water - ethanol mixture at 100 "C to produce the dye precursor. The colour is produced after the addition of tetraethylenepentamine. ~-(~-NITROBENZYL)PYRIDINE has been used as a reagent for the determination of a variety of alkylating agents1-* Preussmann, Schneider and Epple5 demonstrated the alkylating properties of organophosphorus insecticides by their reaction with 4-(4-nitrobenzyl)pyridine at 100 "C.Bedford and Robinson,6 in a review paper, made a critical study of the 4-(4-nitro- benzy1)pyridine colour reaction while Getz and Watts7 reported a rapid colorimetric deter- mination of organophosphate pesticides by use of the reaction with 4-(4-nitrobenzyl)pyridine in a slightly alkaline solution at 175 to 180 "C. This temperature is inconveniently high for samples containing low-boiling solvents, such as ethanol or acetone, and may even result in the loss of the organophosphorus esters themselves if they have high vapour pressures. A modification to the method was therefore developed that involved the use of a lower reaction temperature. The modified method is sensitive, reproducible and suitable for many organophosphorus pesticides.METHOD APPARATUS- Hilger Spekker absorptiometer. Bath-A bath the temperature of which is controllable to &l "C at 100 "C, for example, the TE-4/CTB made by Techne (Cambridge) Ltd., must be used. REAGENTS- crystallised from cyclohexane; melting-point, 70 to 71 "C) in acetone. 4-(kNitrobertzyl)pyridine-A 5 per cent. m/V solution of 4-(4-nitrobenzyl)pyridine (re- Tetraethylene$entnmine, technical-A 10 per cent. V/V solution in acetone. Ethanol, 95 per cent. PROCEDURE DETERMINATION OF THE OPTIMUM REACTION TIME- To a series of 18 x 1.5-cm test-tubes are added 1 ml of 95 per cent. ethanol containing 50 pg of an organophosphorus pesticide, 1 ml of the 4-(4-nitrobenzyl)pyridine reagent and 2 ml of glass-distilled water. The solutions are mixed and heated at an accurately controlled temperature of 100 "C in an ethylene glycol bath.The tubes are arranged so that they stand upright with 10 cm of each immersed in the bath liquid. Under these conditions the mixture refluxes gently. The tubes are removed at intervals and cooled in water at a temperature of about 16 "C. At this stage in the procedure the mixture can be left for as long as 30 minutes without affecting the final absorption reading. Next, 5 ml of the 10 per cent. tetraethylene- pentarnine solution are added, carefully washing the sides of the tubes, the solutions mixed well and then read without delay at a wavelength of 580 nm, with an Ilford spectrum filter 606, in 1-cm cells against distilled water. The absorbances obtained after different heating times are shown in Table I.* Present address : Centre for Overseas Pest Research, Division of Chemical Control, Porton Down, @ SAC; Crown Copyright Reserved. Salisbury, Wiltshire, SP4 0 JQ.432 TURNER : SPECTROPHOTOMETRIC DETERMINATION OF ORGANOPHOSPHORUS [Analyst, Vol. 99 TABLE 1 ABSORBANCE OF THE COLOURS GIVEN BY 50 pg OF DIFFERENT PESTICIDES, HEATED FOR VARIOUS PERIODS OF TIME Heating timelminutes r 5 10 15 20 30 40 chosen for Absorbance calibration/ A \ Time Pesticide \ minutes f h Malathion .. . . 0.345 0.705 0.720 0.710 0.695 0.670 15 Dichlorvos . . . . 0.900 0.985 1.020 1.030 1.000 0.950 17 Tetrachlorvinphos . . 0.575 0.700 0-710 0.690 0.640 0.530 12 Fenchlorphos . . . . 0.750 0.790 0.790 0.760 0-735 0.685 12 PREPARATION OF STANDARD GRAPH- Once the optimum heating time has been established the procedure is repeated with the tubes containing 0 to 50pg of the organophosphorus pesticide.A reagent blank is included and its absorbance value is subtracted from the values obtained from the other samples. The standard graph is constructed by plotting the net absorbance against the concentration of pesticide on arithmetic graph paper. The stability of the colour varies according to the pesticide and it is important to measure the absorbance as soon as possible. The results given in Table I1 show that the absorbance decreases with time in samples exposed to daylight from a northerly direction in unstoppered test-tubes. TABLE I1 STABILITY OF THE COLOUR FROM 50 pg OF DIFFERENT PESTICIDES IN DAYLIGHT FROM A NORTHERLY DIRECTION Pesticide Time/minutes 0 30 20 30 Absorbance r A -l r A \ Malathion .. . . . . . . 0.720 0.690 0.640 0-590 Dichlorvos . . . . . . . . 1-030 0-970 0.860 0.770 Tetrachlorvinphos . . . . . . 0.710 0.688 0.640 0-600 Fenchlorphos . . .. . . . . 0.790 0.750 0.700 0.620 Provided that care is taken to avoid bright sunlight and that the described procedure is carefully followed, the reproducibility of absorbance is good, as shown for malathion in Table 111. TABLE I11 REPLICATION OF MALATHION DETERMINATIONS Malathion/pg Mean absorbance Number of determinations Coefficient of variation, per cent. 25 0.351 10 1.1 50 0.712 10 0-5 RESULTS AND DISCUSSION Table IV lists the organophosphorus pesticides used and Fig. 1 is the absorption graph obtained for 20 pg of malathion.The absorbance values for 50 pg of the various compounds are given to show the variation in sensitivity of the reaction with structure. The absorbance decreased with increasing size of the alkyl group, as shown by the series dimethyl, ethyl methyl and diethyl 2,5-dichloro-4-iodophenyl phosyhorothionates. Some esters, 2,4,5-trichlorophenyl isopropyl phosphoroamidothionate and 4-bromo- 2,5-dichlorophenyl isopropyl methyl phosphorothionate, produced no colour. There was also no reaction with examples of hydrolytic decomposition products of some of the esters, such as desmethyl-bromophos, desmethyl-fenchlorvinphos and ammonium dimethyl phos- phorothionate, and so a valuable feature of the method is that it determines only the original triester in the presence of such decomposition products.4 E Y TABLE IV ABSORBANCE VALUES FROM 50 ,ug OF VARIOUS ORGANOPHOSPHORUS PESTICIDES Compound S-[ 1,2-Di(ethoxycarbonyl)cthyl] dimethyl phosphorothiolothionate 2,2-Dichlorovinyl dimethyl phosphate .. . . . . . . . . Dimethyl 3-methyl-4-methylthiophenyl phosphorothionate , . a-Cyanobenzylideneamino diethyl phosphorothionate . . . . Dimethyl 2,4,5-trichlorophenyl phosphorothionate . . . . . * 2,5-Dichloro-4-iodophenyl dimethyl phosphorothionate . . . . 2,5-Dichloro-4-iodophenyl diethyl phosphorothionate . . . . 2,5-Dichloro-4-iodophenyl ethyl methyl phosphorothionate .. 2,5-Dichloro-4-iodophenyl dimethyl phosphate . . . . . . 2-Chloro- 1 - (2,5-dichlorophenyl) vinyl dimethyl phosphate . . . . 2-Chloro-l-(2,4-dichlorophenyl)vinyl dimethyl phosphate .. . . Isopropyl (2,4,5-trichlorophenyl) phosphoramidothionate . . . . Methyl (2,4,5-trichlorophenyl)ethyl phosphoramidothionate . . Methyl (2,4,5-trichlorophenyl) phosphoramidothionate . . . . 4-Bromo-2,5-Dichlorophenyl dimethyl phosphorothionate . . . . trans-2-Chloro- 1 - (2,4,5- trichlorophenyl) vinyl dimethyl phosphate . . (2-Chloro-a-cyanobenzylideneamino) diethyl phosphorothionate . . S-a-Ethoxycarbonylbenzyl dimethyl phosphorothiolothionate . . 4-Bromo-2,5-dichlorophenyl isopropyl methyl phosphorothionate . . 2-Diethylamino-6-methylpyrimidin-4-yl dimethyl phosphorothionate 2-Ethyl-6-ethoxypyrimidin-4-yl dimethyl phosphorothionate . . 2-Isopropyl-6-methoxypyrimidin-4-yl dimethyl phosphorothionate 6-Ethoxy-2-isopropyl pyrimidin-4-yl dimethyl phosphorothionate S-(N-Ethylcarbamoylmethyl) dimethyl phosphorothiolothionate .. Dimethyl S- (N-methylcarbamoylmethyl) phosphorothiolothionate .. . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . a . . . . . . . * . Melting-point/"C Common name Liquid (technical 95 per cent.) Liquid (technical 99 per cent.) Liquid (technical 95 per cent.) Malathion Dichlorvos Fenthion 97 to 98 Tetrachlorvinphos 68 Chlorphoxim Liquid (technical 95 per cent.) Liquid (technical 90 to 92 per cent.) Phenthoate 40.5 to 41.5 Fenchlorphos 74 to 75 Iodofenphos 49 to 50 - 38 to 41 - 64 t o 65 - 101 to 103 - 67 to 69 - 62 to 63 I 42 to 43 - 66 to 67 - 50 to 50-5 - 56 to 57 Bromophos Phoxim Liquid (technical 94 per cent.) Pirimiphos-methyl Liquid (technical 99 per cent.) Liquid (technical 99 per cent.) Liquid (technical 99 per cent.) - - - 65 to 66 Ethoate-methyl 51 to 52 Dimethoate Absorbance 0.720 1.030 0.815 0.710 0-560 0.580 0.750 0.780 0-610 0.375 0.580 0.670 0.750 0.780 0.000 0.550 0.730 0.810 0.500 0.000 0.810 0.750 0.750 0.980 1.01434 TURNER 400 500 600 Wave I en gt h /n m 700 Fig.1. Absorption curve obtained for 20 pg of malathion The mechanism of the formation of the chromogen from the reaction of 4-(4-nitrobenzyl)- pyridine with a variety of alkylating agents has been de~cribed.l~~~* Kramer and Gamsong postulate that the reaction between 4-(4-nitrobenzyl)pyridine and organophosphorus pesti- cides is a phosphorylation and not an alkylation, but the results given in Table IV, although limited in number, give most support to the suggestion that alkylation also occurs with these compounds.This support is illustrated by the higher sensitivities for the dimethyl compounds compared with the diethyl or monomethyl esters and by the similar responses of phosphates and phosphorothionates. It was also found that, although triethyl phosphate was inactive, the trimethyl ester did react S~OW~Y, the optimum heating time being 60 minutes and the absorbance for 50 pg being 0.450. Although the procedure has proved mainly satisfactory, occasionally erratic results were obtained. These results are due to the rapid volatilisation of the acetone in the reaction mixture, and if this effect is troublesome, it can be avoided by substituting 95 per cent. ethanol for acetone as the solvent for the 4-(4-nitrobenzyl) pyridine. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Epstein, J., Rosenthal, R. W., and Ess, R. J., Analyt. Chem., 1955, 27, 1435. Dixon, B. E., and Hands, G. C., Analyst, 1959, 84, 463. Friedman, 0. M., and Bodger, E., AnaZyt. Chem., 1961, 33, 906. Klatt, Q., Griffin, A. C., and Stehlin, J. S., Proc. SOG. Exp. Bid. Mad., 1960, 104, 629. Preussmann, R., Schneider, H., and Epple, F., A rzneimzttel-Forsch, 1969, 19, 1059. Bedford, C. T., and Robinson, J,, Xenobiotica, 1972, 2, 307. Getz, M. E., and Watts, R. R., J . Ass. 08. Agric. Chem., 1964, 47, 1094. Sawiki, E., Bender, D. F., Hauser, T. R., Wilson, R. N., and Meeker, J. E., A?iaZyt. Chem., 1963, Kramer, D. N,, and Gamson, R. M., Ibid., 1957. 29, 21A. 35, 1479. Received October 24th, 1973 Accepted February 1 lth, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900431

出版商:RSC    

年代:1974

数据来源: RSC