ISSN:0003-2654    

DOI:10.1039/AN96691FX037

出版商:RSC    

年代:1966

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96691BX039

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

iv SUMMARIES OF PAPERS I N THIS I S S U E [October, 1966Summaries of Papers in this IssueThe Application of the Oxygen-flask Combustion Technique tothe Determination of Trace Amounts of Chlorine and Sulphurin Organic CompoundsThe Schoniger oxygen-flask technique has been adapted to trace analysis.An electrically fired apparatus has been designed that is capable of com-busting up to 100 mg of material. As little as 0.1 pg of chloride can betitrated potentiometrically with silver nitrate solution.Sulphate is titrated with barium perchlorate with thorin as indicator.Visual and spectrophotometric detection of the end-point has been investi-gated and the latter is preferred for amounts down to 0.5 pg of sulphate.Cations which interfere can be readily removed by ion exchange.R.McGILLIVRAY and S. C. WOODGERICI Fibres Limited, Hookstone Road, Harrogate, Yorkshire.Analyst, 1966, 91, 611-620.An Assessment of Some Methods of Analysis for DimethoateResidues in Fruit and VegetablesThree methods for determining residues of dimethoate, 00-dimethylS-(N-methylcarbamoylmethyl) phosphorodithioate, in fruits and vegetableshave been investigated for the purpose of recommending a standard methodthat can be applied to a variety of temperate fruits grown in the UnitedKingdom. Laws and Webley’s general method for determining organo-phosphorus insecticide residues was satisfactory for residues of dimethoatein sprouts, lettuces and apples, but not in peas. Chilwell and Beecham’sprocedure was also satisfactory at the 1 or 2 p.p.m.level of added dimethoatein cabbages, lettuces, apples and peas. Giang and Schechter’s method,however, gave variable results.N. A. SMARTPlant Pathology Laboratory, Ministry of Agriculture, Fisheries and Food, Har-penden, Hertfordshire.Analyst, 1966, 91, 621-624.An Improved Method for Determining Residues of DiquatThe ion-exchange method for determining diquat residues in potatotubers1 has been modified to give a simpler operating procedure and increasedaccuracy. Experiments on untreated samples with 0.08 to 0.2 p.p.m. of diquatadded showed an average recovery of 76 per cent. with a standard deviationof 11 per cent. The method, with minor modifications, has been foundto be generally applicable to other food crops and to water. For a 250-gsample, the limit of determination is 0-01 p.p.m.A.CALDERBANK and S. H. YUENImperial Chemical Industries Ltd., Agricultural Division, Jealott’s Hill ResearchStation, Bracknell, Berkshire.Analyst, 1966, 91, 625-629.Improvement in Performance of a Simple Atomic Absorptiometerby Using Pre-heated Air and Town GasA simple atomic absorptiometer for the routine determination of mag-nesium, copper and zinc is described. A pre-heated air supply to the atomiserincreased the response efficiency of the absorptiometer 16-fold. An adjustableslot in the burner enabled a wide range of fuel gases to be used. A horizontalmonochromator slit gave improved light transmission and absorption.R. A. G. RAWSONRothamsted Experimental Station, Harpenden, Hertfordshire.Analyst, 1966, 91, 630-637vi SUMMARIES OF PAPERS IN THIS ISSUEActivation Analysis for Titanium and Niobium with Fast NeutronsFast-neutron activation methods have been developed for determiningtitanium and niobium. Scandium-47 formed by the (n,p) reaction on titan-ium-47 and yttrium-90 formed by the (n,cc) reaction on niobium-93 have beenused for determining these elements.Irradiations were carried out in theswimming-pool reactor “Apsara” a t the Atomic Energy EstablishmentTrombay, at a fission-flux of 3 x loll n per cm2 per second. l<adiochemicalprocedures were developed for the isolation of scandium and yttrium from theirradiated samples. The methods have been applied to the determination oftitanium in stabilised steels and the standard rocks G-1 and W-1 and ofniobium in stabilised steels.The advantages and limitations of the methodsare discussed.V. T. ATHAVALE, H. B. DESAI, S. GANGADHARAN, M. S. PENDHARKARand M. SANKAR DASAnalytical Division, Atomic Encrgy Establishment Trornbay, Bombay, India.[October, 1966Analyst, 1‘366, 91, 638-646.An Automatic, Modified Formaldoxime Method for Determining LowConcentrations of Manganese in Water Containing IronAn automatic method for determining manganese in fresh watcr withthe AutoAnalyzer is described, together with details of the analytical system.The method is based on a modification of the formaltloxime procedure pub-lished by Goto, Komatsu and Furukawa,l which consists in reacting man-ganese with formaldoxime in alkaline solution, and decomposing any ironformaldoxime formed with EDTA and hydroxylammoiiium chloride.Aluminium, zinc, copper, iron, calcium, magnesium, chloride and plios-phate do not interfere at concentrations above those normally found innatural water in Norway.The capacity of the method is 25 samples per hour, and the stanclsrddeviation is 5,ug of manganese per litre.A statistical comparison of the manual persulphatc oxidation mctlmdwith the automatic formaldoximc mcthod indicated that both mcthodsgive identical results.A. HENRIKSENNorwcgian Institiitc for Water Rcscarcli, (_>\lo 3 , Norway.Ancclyst, l!Mi, 91, 047-651.An Automatic Method for Determining Orthophosphate in Sewageand Highly Polluted WatersShort PuperA.HENRIKSENNorwegian Institutc for Water Research, ()slo 3, Norway.Analyst, IOCici, 91, 652-653.The Determination of 4-Aminobiphenyl in Aromatic AminesShort Paper0. NORMANStaveley Chemicals Ltcl., Stavcley Works, Chc.sterficlc1.and G. A. VAUGHANThc Coal Tar Rcscarch Association, Goincrsal, T,ccds.Analyst, 1066, 91, 653-654...Vlll SUMMARIES OF PAPEIZS I S THIS ISSUESpecific Spot Tests for Silver CyanideShoYt PapevF. FEIGLLabxatorio da Produ@o Mineral, 3linist6rio das Llinas e Encrgia, Rio de Janeiro.and A. CALDASEscola Nacional de Quimica, Univcrsidadc do B r a d .[Octuber, 1966Analyst, 19G6, 91, 634-636.The Determination of Benzylpenicillin Traces in PropyliodoneInjection B.P.Short PaperMrs. DOROTHY J .SIMMONS and J. P. JEFFERIESGlaxo Laboratories Lttl., Greenfortl, 3Iitltllcscx.Analyst, 1966, 91, 656-659.The Spectrophotometric Determination of TraceAmounts of TantalumShovt PapevJOHN H. HILLLawrence Radiation Laboratory, University of California, Livermor?, California.Analyst, 1966, 91, 659-662.The Organic-phase Spectrophotometric Determination of Ironwith ThiocyanateShort PapevE. CERRAI and G. GHERSINILaboratori C.E.S.E., Casella Postale 3086, Milano, Italy.Analyst, 1969, 91, 662-664.A Direct Photometric Procedure for the Determination ofBoron in NickelShovt PapevT. R. ANDREW and P. N. R. NICHOLSCentral Materials Laboratory, The Xullard Radio Valve Company, ?;ew Road,Rlitcham Junction, Surrcy.Analyst, 1966, 91, 664-667.A Modified Potentiostat for Controlled Potential AnalysisShort PaperJ. GRIMSHAW and R. K. QUIGGDepartment of Chemistry, Queen’s Univerqity, Belfast, Sorthcrn Ireland.Analyst, 1966, 91, 667-669.Determination of Specific Gravity of Glass Particles by aDensity Gradient MethodShort PaperS. S. KIND and L. SUMMERSCALESHome Office Forensic Science Laboratory, Haddon Lodge, 32 Rutland Drive,Harrogate, Yorkshire.Analyst, 1966, 91, 669-670

ISSN:0003-2654    

DOI:10.1039/AN96691FP199

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

October, 19661 THE ANALYST xiiiCLASSIFIED ADVERTISEMENTSSIR JOHN CASS COLLEGEMICROANALYSTrequired for the Department of Chemistry to be in chargeof the Microanalytical Laboratory with on; assistant,Applicants should be experienced i n organic elementalanalysis of materials containing hetero elements, and shouldbe capable of developing conventional and instrumentalmethods.Salary scale: k1260 x k35 to k1435.Applications should be made in writing, with the name ofa referee, to the Head of the Chemistry Department, SirJohn Cass College, Jewry Street, London, E.C.3.BRUNEL UNIVERSITYGAS CHROMATOGRAPHYA COMPREHENSIVE COURSEWednesdays 7-9 p.m.-Beginning October 12th, 1966.The Lectures will be given by specialists during the Autumnand Spring Terms and are supplemented by demonstrationsof equipment.Associated techniques will be considered.The course is suitable for graduates and others with a specialinterest in this field. I t is recommended for those proceedingto M.Tcch. and Ph.D. in this field.Fee &5 5s. Od.Application forms and further details from The AcademicRegistrar, Brunel University, Woodlands Avenue, Acton,W.3. Tel.: ACORN 6661.YOUNG CHEMIST/PHARMACIST required for QualityControl Department. Ability to handle physical instru-ments, e.g. a U.V. Spectrophotometer, pH meter, etc., someexperience in developing or assessing new pharmaceuticalformulations; and the preparation of appropriate Specifica-tion forms and Manufacturing Control forms, would be anadvantage.A four figure salary will be paid according to age andexperience.5-day week, company pension scheme.Apply to boxNo. 4088, The Analyst, 47, Gresham Street, London, E.C.2.CITY OF MANCHESTER HEALTH DEPARTMENTPublic Analyst’s Laboratory. Technician/assistantanalyst required. G.C.E. with definite scientific bias, orO.N.C. in chemistry. Excellent opportunity for youngperson with practical aptitude and ambition. The successfulapplicant would be encouraged to follow existing approvededucational courses, by which further qualifications up toA.R.I.C. are obtainable. Experience in modern techniquese.g. thin-layer or gas - liquid chromatography, etc., would b;an advantage. Salary according to age and qualification asfollows-with 3 “0” levels including chemistry--E440 a t age 20with 5 0” levels including chemistry-L510 a t age 20with O.N.C.in chemistry-L735 rising to @60.with H.N.C. in chemistry-Lg20 rising to @25.rising ,fo l725.rising to ,E725.5-day week. Applications to Medical Officer of Health,Town Hall, Manchestcr, 2.CITY OF LEICESTER requireDEPUTY CITY ANALYSTModern laboratory in ncw building. Minimum qualifica-tion Branch E Diploma of the K.I.C.Grade A.P.T. ‘‘B/C” E1610/&2110 p a . Superannuablepost. Apply giving details of qualifications and experienceand naming two referees to Medical Offcer of Health, CityHealth Department, 1~ Grey Friars, Leicester.GREEN’S PURE FILTER PAPERSfor all kinds of filtrationWrite for descriptive catalogue 43/6.65J. BARCHAM GREEN LTD., HAYLE MILL,MAIDSTONE, KENTRADIOCHEMICALANALYSISA senior chemist is required to act as Assistant Manager of the AnalyticalDepartment, responsible for day to day running of the department, and todeputise for the Manager in his absence. The Analytical Department checksthe purity of the Centre’s products comprising many radio-isotopes and labelledcompounds.Applicants must have had not less than three years post graduate experience andhave at least a second class honours degree in Chemistry or equivalent qualifi-cations.A sound knowledge of inorganic and organic analysis, of modernanalytical techniques and experience and proven ability in management areessential requirements. A knowledge of radiochemistry would be an advantage.Salary $2,410 p.a.rising by annual increments to $3,325 p.a.Assisted housing and superannuation schemes.Write for application form to the Personnel Officer (Ref: S.9558)THE RADIOCHEMICAL CENTREAMERSHAM * BUCKSOctober, 1966] THE ANALYSTS E C O N D E D I T I O NxviiK ~ r t RanderathThin-Layer ChromatographyReviews of the First EditionThis technique makes all the advantages of adsorption chromatography (rapidity, specific separations, applicabilityto lipophilic materials) available in the analytical field. I t is therefore hardly surprising that thin-layer chromato-graphy has rapidly achieved popularity and that there is already an extensive body of data illustrating the usefulnessof the method.This book will soon be indispensible to all chemists, biologists, and pharmacists who use chromatography and willalso help to opcn up further applications of thin-layer chromatography.(Angewandte Chemic - International Edition)Dr.Randerath has written a clear and concise monograph that provides a practical and theoretical discussion ofthin-layer chromatography (TLC) covering both quantitative and qualitative aspects of this technique. He has at-tempted to review pertinent references in the rapidly expanding list and has included an introductory chapter on thehistorical background that has led to the prcsent-day versatile application of this technique in all chemical andbiological disciplincs. Thc procedures and results for a wide variety of scparations are described in sufficient detailto permit duplication of methods recorded in the literature.Dr. Randerath’s book is intended for the research worker, but it will also be an excellent refercnce for college chem-istry courses such as organic qualitative analysis and general biochemistry.(Chemical Education)Translated from the German b~ D. D. Libman. 19GG. 2nd revised and enlarged edition. XIV, 285 pages with 9G figures,4 colored illustrations and 73 tables. Clothbound DM 38,- = $9.50.The German edition of the book is available under the title >>Diinnschicht-Chromatographiecc.V E R L A G C H E M I E - G M B HW E I N H E I M I B E R G S T R .ACADEMIC P R E S S I N C .N E W YORK & L O N D O

ISSN:0003-2654    

DOI:10.1039/AN96691BP209

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

OCTOBER, 1966 THE ANALYST Vol. 91, No. 1087 The Application of the Oxygen-flask Combustion Technique to the Determination of Trace Amounts of Chlorine and Sulphur in Organic Compounds BY R. McGILLIVRAY AND S. C. WOODGER ( I C I Fibres Limited, Hookstone Road, Harrogate, Yorkshire) The Schoniger oxygen-flask technique has been adapted to trace analysis. An electrically fired apparatus has been designed that is capable of com- busting up to 100 mg of material. As little as 0-1 pg of chloride can be titrated potentiometrically with silver nitrate solution. Sulphate is titrated with barium perchlorate with thorin as indicator. Visual and spectrophotometric detection of the end-point has been investi- gated and the latter is preferred for amounts down to 0.5 pg of sulphate. Cations which interfere can be readily removed by ion exchange.IN recent years the oxygen-flask combustion technique has become established for deter- mining chlorine, sulphur, phosphorus and other elements in organic compounds,l but its use for the determination of minor and trace amounts has received scant attention. Lehner2 used the technique to determine chlorine and sulphur in organic compounds at the 0.1 per cent. level, and Haslam and Squirrel13 used it to determine sulphur at a similar level in poly- (methyl methacrylate). It was thought that the technique could be applied to the deter- mination of microgram amounts of these elements in organic compounds and polymers. COMBUSTION OF THE SAMPLE- In trace analysis the reagents used should be free from the element being determined.The paper in which the sample is enclosed is a potential source of contamination. It was found that Whatman No. 42 paper contains about 300 p.p.m. of chlorine and about 150 p.p.m. of sulphur, which limit the sensitivity of the determination. No Whatman papers were found to be free from chlorine. The impurities could be partially washed out with water, but not enough to make the paper a practical proposition. EXPERIMENTAL I-litre flask with 834 neck Securi to hold tightly Heating element and quartz-wool pads Absorption solution Fig. 1. Combustion assembly 61 1612 MCGILLIVRAY AND WOODGER: APPLICATION OF THE [A?Zd$%t, VOl. 91 Other materials were tried. Polythene was not suitable because it tended to melt and drop into the absorption solution, taking with it a little of the sample.“Melinex” film washed with alcohol contained a negligible amount of chlorine and sulphur, but occasionally, molten drops of polymer fell from the basket, as with polythene. Juvet and Chiu4 enclosed the sample in glass-wool and ignited it by means of an electrically heated nickel chrome wire. This firing principle was adopted for the present work with two main alterations. Platinum was used for the heating filament to eliminate possible loss of chlorine and sulphur by reaction with the nickel chrome wire. Quartz-wool pads were used because the heat of combustion melted the glass-wool, which then coated the platinum heating filament. The apparatus shown in Fig. 1 consists of a 1-litre round-bottomed flask fitted with a B34 joint, into which fits the electrically fired combustion head (Fig.2). The current to the 1 -Platinum heating filament atinum wire 18s 9c m v4.g. wire joint Fig. 2. Electrically fired combustion head platinum heating element is supplied from a 25-volt Variac, connected in series to an ammeter. A current of 4 to 5 amps heats the elements to a bright red glow, sufficient to fire the sample. This combustion head is similar in principle to that used by H e m ~ e l . ~ The sample (about 100 mg) is placed on the centre of a fairly tightly compressed quartz- wool pad, approximately 30 mm long, 10 mm wide and 5 mm deep. A loose pad of quartz-wool, thin enough for the heat from the platinum filament to ignite the sample, is placed on top of the sample. The pads are fitted round the heating element, the thinner pad being next to the heating element, and are held in position by a folded platinum wire.A suitable absorption solution is placed in the 1-litre flask, the flask is filled withmedicinal oxygen, purified by passing it through a tube filled with soda asbestos, and the combustion head carrying the sample is inserted into the flask. The joints are secured with metal springs and the apparatus connected as shown in Fig. 1, the flask being inverted so that the absorption solution forms a seal with the joint. A current of 4 to 5 amps is passed through the heating element and is allowed to flow until combustion is complete (usually in 10 to 20 seconds) in order to prevent the formation of soot on the quartz-wool pads. The maximum amounts of sample that have been combusted smoothly and cleanly in the electrically fired apparatus have been about 100 to 150 mg, and it is felt that the technique could be modified to handle larger amounts, probably by using a larger combustion flask and enclosing the sample in sufficient quartz-wool to limit the supply of oxygen and slow down the combustion, making it less violent than when paper is used to wrap the sample.Non-volatile liquids can be added by micrometer syringe to a quartz-wool pad underneath the heating element, the pad then being folded round the heating element and the sample combusted as described above. The use of medicinal oxygen purified by passing through a U-tube filled with soda asbestos is necessary to obtain low “blanks” with the chlorine determinations.The quartz- wool used (type A, supplied by the British Thermal Syndicate Ltd.) contains no chlorideOctober, 19661 OXYGEN-FLASK COMBUSTION TECHNIQUE 613 or sulphate which can be washed out under the conditions used, but it is essential that it should not be handled with bare hands. DETERMINATION OF CHLORINE- For the determination of 1 p.p.m. of chlorine in a 100-mg sample, a method of deter- mining 0.1 pg of chloride was needed. The best method seemed to be the potentiometric argentometric titration carried out in a medium of relatively low dielectric constant. The titration assembly, which can deal with volumes of titration liquid of from 0.2 to 5 ml, consists of a vessel that accommodates a micro air-stirrer, a silver electrode, the tip of a micrometer syringe and one end of a salt bridge.The other end of the salt bridge and a calomel electrode dip into saturated potassium sulphate solution in a beaker. The silver and calomel electrodes are connected to an E.I.L. automatic titrimeter or a suitable potentio- meter. The smallest titration vessel used is 2 ml in capacity. The apparatus was used to determine chloride in the range 0.4 to 4 pg. The chloride was dissolved in 0.5 ml of 50 per cent. v/v aqueous ethanol, and 0.01 N silver nitrate was added in increments of 0.0002 ml from the micrometer syringe. The end-point was obtained by taking the mid-point of the steepest part of the curve of potential plotted against titrant volume. The results summarised below show that the method is satisfactory for the deter- mination of chloride in the range 0.4 to 4 pg, the end-points even with 0.4 and 0.7 pg of chloride being still sharp- Chloride, pg of C1- added .. 4.38 2-73 1.83 0-73 0.36 Chloride, pg of C1- found . . 4-42 2.70 1.88 0.73 0-38 The sensitivity of the method might be further increased by carrying out the titration in a total volume of 0.2m1, but with the present apparatus this is about the minimum amount that can be handled. When the above method was used to determine chloride in a total volume of 5ml of 50 per cent. v/v aqueous ethanol, it was found that below 2.5 pg of chloride the accuracy of the results fell sharply, and it is therefore recommended that the concentration of chloride be at least 0.5 pg per ml in the solution being titrated.METHOD APPARATUS- Combustion apparatus. Micrometer syringe-This is of 0 6 m l capacity, delivering 0.00005 ml as described in British Standard 1428, Part 6: 1955. Ultra-micro potentiometric titration assembly. REAGENTS- Hydrogen peroxide, 1 vol. Sodium hydroxide, 0.1 M. Ethanol, 75” O.P. Nitric acid, 0.1 M. Silver nitrate, 0.01 N. Potassium sulphate - agar solutiorlz-Heat together in a beaker 12 g of potassium sulphate, Oxygefi-Medicinal oxygen purified by passage through a tube filled with soda asbestos. 5 g of agar and 200 ml of water. PROCEDURE- Place 5 ml of hydrogen peroxide and 0-05 ml of sodium hydroxide solution in a 1-litre combustion flask. Accurately weigh 60 to 100mg of sample (see Note 1) on to the centre of a fairly tightly compressed quartz-wool pad.Assemble the combustion apparatus and combust the sample as described in the sub-section on “Combustion of Sample” (see Note 2). Liquids are delivered on to a pad from the micrometer syringe. After the combustion, turn the flask the correct way up, taking care that the liquid flows down the side of the flask and that it does not splash on to the quartz-wool pads. Leave it for 10 minutes to allow the combustion products to be absorbed. Remove the combustion head and wash the joint, adjoining tubes and quartz-wool with “metal-free” water, adding the washings to the 1-litre flask. Transfer the solution to a 50-ml separating funnel and wash614 MCGILLIVRAY AND WOODGER: APPLICATION OF THE [A%@@St, VOl. 91 the flask with water, adding the washings to the separating funnel.Place a suitably sized titration vessel (see Note 3) on a heated steam-bath, and half fill the vessel with liquid from the separating funnel. Allow the solution to evaporate, and, as it does so, add more solution from the funnel at a rate about equivalent to that of the evaporation. When the funnel has been drained of liquid, wash it with a few millilitres of water and add this to the titration flask. Evaporate the solution in the titration flask to the appropriate volume to give a final chloride concentration of at least 0.5 pg per ml and add an equal volume of ethanol. Add 0.1 ml of nitric acid. Assemble the potentiometric titration apparatus, switch on the stirrer and titrate with silver nitrate solution added from a micrometer syringe in increments of 0.0002 ml (see Note 4).Record the potential after each addition of silver nitrate and plot a graph of the silver nitrate added against potential. The mid-point of the steeply rising portion of the graph is the end-point. Carry out a blank combustion and titration by using the same apparatus and reagents as used in the determination, keeping the heating element at bright red heat for 5 minutes. NOTES- 1. If the sample contains more than 100 p.p.m. of chlorine, then a 60 to 70-mg sample is sufficient. If the sample contains less than 100 p.p.m., then it is desirable to take as large a sample as will combust smoothly, but not much more than 100 mg. If the chlorine content is between 1 and 5 p.p.m., a 70 to 100-mg sample can be combusted, the combustion products absorbed, the flask re-filled with oxygen and a further 70 to 100-mg sample combusted. The quartz-wool pad and the wire clip should not be handled with bare hands.Fingerstalls or thin rubber gloves should be worn. 2. The apparatus should be placed in a fume cupboard with the Variac located outside, and the cup- board kept closed during the combustion. 3. The size of the titration vessel and the necessary concentration of the liquid are dependent on the amount of chlorine present in the sample, Chlorine expected, Size of titration Volume of evaporated p.p.m. vessel, ml solution, ml More than 100 30 5 50 to 100 10 2.5 20 to 50 5 1 5 to 20 2 0.25 4. With a silver and a calomel electrode, the end-point is usually between 290 and 310mV. If a t the start the potential reading is below 250 mV, silver nitrate can safely be added until this reading is reached; thereafter, it should be added in increments of 0-0002 ml.Fig. 3. Titration cell for Unicam SP600. All measurements are in inchesOctober, 19661 OXYGEN-FLASK COMBUSTION TECHNIQUE . 615 ANALYSIS OF SAMPLES- The method was tested on samples containing known amounts of chlorine prepared by adding l-chloro-2,4-dinitrobenzene (M.A.S.) to purified benzoic acid that had been prepared from AnalaR material by two zone refinings followed by two crystallisations from water. Appropriate amounts of a solution of l-chloro-2,4-dinitrobenzene were added from a micro- meter syringe to approximately 50-mg samples of purified benzoic acid. The ethanol was allowed to evaporate before the combustion was carried out.Determination of the chlorine in the “purified” benzoic acid gave a titre of silver nitrate not significantly different from that of a blank combustion, viz., 0-0006ml of 0.01 N silver nitrate. Collected below are the results obtained with several samples containing from 5 to 500 p.p.m. of chlorine. The experiments with up to 50 p.p.m. were titrated in a total volume of 0.5 ml of 50 per cent. ethanol, while that with 500 p.p.m. was titrated in a total volume of 5 ml of 50 per cent. ethanol- Chlorine, p.p.m. of C1 addcd . . . . 510 51 12 6 Chlorine, p.p.m. of C1 found . . . . 506 50 11 8 of organic compounds and polymers, and some results are given in Table I. The method has been used to determine the chlorine content of a number of samples TABLE I DETERMINATION O F CHLORINE I N ORGANIC COMPOUNDS Sample Chlorine, p.p.m.of C1 Biphenyl and diphenyl ether . . . . . . 12, 13 Biphenyl and diphenyl ether . . . . . . 71, 71 Biphenyl and diphenyl ether . . .. . . 74, 82 Cyclohexanol and cyclohexanone . . . . 99, 108, 98, 103 Cyclohexanol and cyclohexanone . . . . 313, 326, 327, 340 Poly(ethy1cne terephthalate) . . . . . . 9, 10 DETERMINATION OF SULPHATE- To determine down to 1 p.p.m. of sulphur in a 100-mg sample of organic material, it is necessary to determine down to 0.1 pg of this element. The method of Fritz and Yamamura,6 consisting of the titration of sulphate with barium perchlorate in the presence of thorin - methylene blue mixed indicator, was scaled down. To the sulphate dissolved in 1 ml of water and 4 ml of ethanol, was added 0.05 ml of 0.05 per cent.ethanolic solution of thorin and 0-05 ml of 0.003 per cent. aqueous solution of methylene blue. The sulphate was then titrated with 0.01 N barium perchlorate with the micrometer syringe, the end-point being detected visually by the colour change from pale green to pale pink. The solution was stirred during the titration and the tip of the burette was kept just below the surface of the solution. Typical results of determining 20 to 200 pg of sulphate by this method were- Sulphate, pg of SO,2- added. . . . 216 43-5 23 Sulphate, pg of found . . . . 218, 216 42.0, 41.6 21, 22 These results are satisfactory, but with less than 10 pg of sulphate the end-point was less distinct, and it was necessary to reduce the volume of the titration medium. By reducing the volume of the titration to 0.5 ml, it was possible to titrate 1 to 10 pg of sulphate.A blank determination on the reagents used corresponded to about 0.5 pg of sulphate, which appeared to come from the methylene blue solution. Some results were- Sulphate, p g of SO,2- added. . . . 5.26 2.63 1.3 0.65 Sulphate, pg of found . . . . 5.2 2.7 1.4 0.7 These results show that the method is satisfactory for determining 1 to 500 pg of sulphate. A disadvantage is that the titrations have to be carried out in good daylight or in front of an illuminated screen approximating to daylight. SPECTROPHOTOMETRIC TITRATIONS- Menis, Manning and Ball7 overcame the difficulty of the poor end-point of the visual titration by carrying out the titration spectrophotometrically, with pentanol as solvent.They found methanol, butanol and isopropanol unsuitable as solvents. By titrating in a total volume of about 35 ml of pentanol they were able to determine down to 6 pg of sulphate.61 6 MCGILLIVRAY AND WOODGER: APPLICATION OF THE [A%@&St, VOl. 91 Pentanol is unsuitable for use in conjunction with the oxygen-flask combustion technique because it is preferable for the titration solvent to accommodate about 5 ml of water. For the visual end-point, a medium containing 80 per cent. of ethanol was used and the spectro- photometric titration of sulphate in this medium was investigated. A titration cell was made of glass with two clear glass windows about 4 to 4.5 cm apart. The rest of the vessel was coated with black paint.A vessel suitable for use with a Unicam SP600 is illustrated in Fig. 3. The sulphate was dissolved in 5 ml of water and placed in the titration cell, and 30 ml of ethanol and 1 ml of 0-01 per cent. ethanolic thorin solution were added. The solution was stirred by a stream of nitrogen, 0.001 M barium perchlorate was added in increments and the change in optical density at 520mp after each addition was noted. Optical density was plotted against volume of titrant added, and the end-point obtained from the intersection in the usual way. Some results obtained with 4 to 300 pg of sulphate were- Sulphate, pg of SO,2- added. . .. 298 149 37.3 5.3 3.7 149 37 5.2 3-8 148 36 4-8 3.8 . . . . Sulphate, p g of S0,Z- found { 22:; The minimum amount of solution that can be used in this titration cell is about 25 ml and in order to titrate amounts of sulphate smaller than 4 pg it is necessary to reduce the volume.A second titration cell was constructed 4 cm long and 1 cm wide, so that with 4 ml of titration solution the depth of the solution in the cell was 1 cm. The efficient stirring of this long narrow column of liquid presented some difficulties. Stirring by gas bubbles and a con- ventional rotary stirrer was unsatisfactory. A vacuum-operated vibratory stirrer such as that described by Stock and Fill8 with a paddle about 3-5 cm long and an amplitude of about 0.5 cm was found to be satisfactory. By using this cell it was found possible to titrate as little as 0.4 pg of sulphate with 0.0005 N barium perchlorate in a titration volume of 4 ml.INTERFERENCE DUE TO CATIONS- Cations which give a coloration with thorin interfere. They include the alkali metals, the alkaline earths, iron and cobalt. When 30 pg of calcium as a chloride solution was added to 144 pg of sulphate as sulphuric acid and the mixture titrated with 0-001 N barium perchlorate with a thorin - methylene blue indicator, only 88 per cent. of the sulphate was recovered. The cations could be removed by passing the solution through an ion-exchange resin column, but it is simpler to add resin and remove it when exchange has been completed. Amberlite IR-l20(H) resin was thoroughly washed with water until the washings gave no titre for sulphate when titrated with 0.001 N barium perchlorate. One scoopful (about 1 ml) of resin was added to the solution containing the sulphate and cation, the mixture was shaken well and allowed to stand for 10 minutes. The resin was removed by filtration, and the sulphate in the filtrate was titrated visually with barium perchlorate.TABLE I1 REMOVAL OF CATIONS BY AMBERLITE IR-l20(H) RESIN IN SULPHATE DETERMINATION Amount of cation present, Cation added as chloride mg Aluminium C a 1 c i u m Calcj uni Calcium Iron . . Sodium Strontium Zinc . , .. * . .. 1.1 .. .. .. 0.03 . . .. .. 0.12 .. .. . . 1.2 .. .. .. 0.96 .. .. .. 2.2 . . .. .. 1.0 .. .. .. 1-1 Sulphate, p g of SO,2- * added found 288 283 144 143 144 144 144 142 288 284 288 287 147 147 288 283 The results given in Table I1 show that the resin treatment is effective in removing up to about 1 mg of calcium, zinc, aluminium, iron and sodium.The treatment is probably effective for other cations, provided they are present in amounts of less than 1 mg.October, 19661 OXYGEN-FLASK COMBUSTION TECHNIQUE 617 The technique has proved generally useful for micro work. Samples received for sulphur determinations often contain small amounts of cations, the presence of which is indicated by the formation of a pink colour with the thorin indicator. Addition of Amberlite resin will remove the interfering cation, provided that not too much is present. The resin usually removes most of the thorin, probably as the cation - thorin complex, but more indicator can be added after removal of the resin. INTERFERENCE DUE TO ANIONS- Fritz and Yamamuras examined the effect of chloride, nitrate, fluoride, phosphate and sulphite ions on the barium perchlorate titration of sulphate and found that only phosphate and sulphite seriously interfere.The interference of sulphite can be neglected because the combustion products are absorbed in hydrogen peroxide which will convert any sulphite to sulphate. There is some confusion in the literature as to the extent of the phosphate inter- ference and consequently the effect of milligram amounts of phosphoric acid on the deter- mination of milligram amounts of sulphate was examined. The results shown in Table I11 are high in the presence of phosphate, the error increasing when more phosphate is present. The spectrophotometric titration gave slightly better results than the visual titration. TABLE I11 EFFECT OF PHOSPHATE ON THE DETERMINATION OF SULPHATE Visual titration with 0.01 N barium perchlorate- Sulphate added, mg .. . . 2.06 2-06 2-06 2-06 2.06 Phosphate added, mg . . 0 0.32 0-80 1.6 3-2 Sulphate found, mg . . . . 2.06 2.12 2-16 2.22 2-27 Percentage error . . . . + 3 +5 +8 + 10 Spectrophotometric titration with 0.01 N barium perchlorate- Sulphate added, mg . . . . 1.92 1-92 1.92 1.92 1.92 Phosphate added, mg . . 0 0.32 0.80 1-6 3.2 Sulphate found, mg . . . . 1-92 1.96 1-99 2-05 2.06 Percentage error . , . . +2 +4 +7 +7 For trace analysis, it is the effect of the phosphate on the determination of microgram amounts of sulphate that is important; 20, 52 and 98 pg of sulphate were determined in the presence of 100, 500 and 1000 pg of phosphate, the titrations being carried out spectro- photometrically with 0.001 N barium perchlorate, with thorin as indicator.TABLE I V EFFECT OF PHOSPHATE ON THE DETERMINATION OF MICROGRAM AMOUNTS OF SULPHATE Sulphateadded, pg 98 98 98 98 52 52 52 52 20 20 20 20 0 0 0 Phosphateadded, pg 0 100 500 1000 0 100 500 1000 0 100 500 1000 100 500 1000 Sulphatefound, pg 98 98 101 106 52 52 52 55 20 20 20 20 (1 <1 <1 Percentageerror 0 0 +3 + 8 0 0 0 +6 0 0 0 0 The results in Table IV show that up to 100 pg of sulphate can be determined in the presence of up to 500 pg of phosphate with reasonable accuracy. This means that if a 100-mg sample is combusted, then up to 300 p.p.m. of sulphur can be determined accurately in the presence of up to 0.15 per cent.of phosphorus. If the sample contains more phosphorus than this, then it is better to remove the bulk of it by precipitating it as the magnesium salt and removing it by filtration. Any magnesium that dissolves to form magnesium sulphate is then removed with an ion-exchange resin. This procedure has been described by Fritz and Yamamura,6 who used magnesium carbonate. The use of purified magnesium oxide is preferred because lower blanks can be obtained. The magnesium oxide (AnalaR) is purified by washing it well with hot water and drying. When 5 ml of solution containing 98 pg of sulphate (as sulphuric acid) and 30mg of phosphoric acid were treated with 100mg of magnesium oxide, the sulphate found in four separate spectrophotometric determinations was 93, 97, 101 and 98 pg, The precision is not as good as usual, but the results are satis- factory in view of the number of manipulations.618 MCGILLIVRAY AND WOODGER: APPLICATION OF THE [AfzaZyst, Vol.91 METHOD APPARAT u S- Combustion apparatus. Titration cells-(a) A cell for titrations in the range 25 to 35ml consisting of a black- painted glass vessel with two clear windows 4 to 4-5 cm apart, suitable for use with a Unicam SP600 is shown in Fig. 3. I t is provided with a top cover made from black Perspex, with holes for the burette and for a capillary tube, through which nitrogen is passed for stirring the solution. ( b ) A cell for titrations in the range 4 to 8 ml, similar to that described above but having a smaller volume between the two windows. Vibratory stirrer-This is vacuum-operated with a paddle about 3.5 cm long and a stroke of about 0-5cm.REAGEXTS- Water-All water must be “metal-free.” Hydrogen peroxide, I vol. Ethanol, 75” O.P. TJtorin solution, ethafzolic-Dissolve 25 mg of thorin in 5 ml of water and dilute to 50 ml Barium perchlorate, 0.001 and 0.0005 N-Prepare by dilution of 0.01 N barium perchlorate iwedicinal oxygen-Purify by passing through a U-tube filled with soda asbestos. Amberlite IR-l20(H) resin-Wash well with “metal-free” water before use and store with ethanol. and store in a polythene bottle. under water in a flask. PROCEDURE- Place 5 ml of hydrogen peroxide in the 1-litre combustion flask and weigh out accurately a 60 to 70-mg sample (see Note 5). Enclose the sample in quartz-wool and combust it as described in the chlorine determination, If the sample contains more than 50 p.p.m.proceed as follows- After the combustion, turn the flask the correct way up, taking care that the liquid flows down the side of the flask and that it does not splash on to the quartz-wool pads. Leave it for 10 minutes to allow the combustion products to be absorbed. Carefully remove the combustion head and wash the joint, the adjoining tubes and the quartz-wool with ethanol, adding the washings to the 1-litre flask (see Note 6). Carefully transfer this solution to the titration cell (a) and wash the flask carefully with ethanol, adding the washings to the titration cell; a total volume of 30 ml of ethanol is used for washing. Add 2 ml of thorin solution. Place the titration cell in a holder and set it in the Unicam so that the light passes through the solution between the two clear windows.Bubble nitrogen through the capillary tube to stir the solution. Stop the nitrogen flow and set the wavelength to 520 mp. Set the optical density to 0.300 and adjust the slit control so that the galvanometer pointer is at zero. Add 0.001 N barium perchlorate in increments of 0-04 ml. Stir after each addition by bubbling nitrogen through, and with the nitrogen supply off, read the optical density. Titrate until the optical density rises sharply and then continue for at least a further 0.2 ml. Plot optical density against volume of titrant added. The intersection of the base-line with the projection of the line forming the steeply rising portion of the curve gives the end-point.For this, fill the combustion flask with oxygen and keep the heating element at bright red heat for 5 minutes; use the same volumes of reagents as in the test and add the titrant in 0.02-ml increments. NOTES- Larger samples should be sub-divided, the flask re-filled with oxygen and the combustion repeated. The interference of calcium, aluminium, copper, iron, sodium and zinc can be overcome by removing the cations with hmberlite IK-l20(H) resin. To the liquor in the flask, add about 1 ml of Amberlite IR-lSO(H) resin, shake well and allow to stand for 10 minutes. Filter through a sintered-glass crucible (porosity 3), collecting the filtrale in the titration cell. Wash the combustion flask and crucible with 30 ml of ethanol, adding the washings to the filtrate.Carry out a blank combustion and titration. 5. The weight of sample taken should not appreciably exceed 100 mg. 6. Certain cations interfere and cause poor end-points and low results. Then add 2 ml of thorin solution and proceed with the titration.October, 19661 OXYGEN-FLASK COMBUSTION TECHNIQUE 619 If the sample contains less than 50 p.p.m. proceed as above to the stage where the combustion products are absorbed, then carefully remove the combustion head and wash the joint, the adjoining tubes and the quartz-wool with the minimum volume of ethanol, adding the washings to the l-litre flask. Carefully transfer this solution to a clean beaker and evaporate on a steam-bath until the volume is 1 ml. Transfer this solution to the smaller titration cell ( b ) and wash the beaker with a total of 4 ml of ethanol, adding the washings to the titration cell.Add 0.5 ml of thorin solution. Titrate as above with 0.0005 N barium perchlorate. ANALYSIS OF SAMPLES- The method was tested by preparing samples containing known amounts of sulphur by adding known volumes of an ethanolic solution of sulphonal (M.A.S.) to purified benzoic acid. The purified benzoic acid was found to contain less than 5 p.p.m. of sulphur. The following results by the spectrophotometric method agree well with the amounts added. By the visual titration method, a sample containing 10 p.p.m. of added sulphur gave a value of 9 p.p.m. of sulphur- Sulphur, p.p.m. of S addcd . . 999 350 56 Sulphur, p.p.m. of S found . . 996 346 58 The methods have been used to determine sulphur in a variety of compounds and poly- mers, and the results obtained with a selection of the products analysed are collected in Table V.In some of the samples the sulphur is present as an additive and in others as an impurity. TABLE V DETERMINATION OF SULPHUR IN ORGANIC COMPOUNDS AND POLYMERS Sample Sulphur found, p.p.m of S Terephthalic acid . . .. . . . . 20,lS Dimethyl terephthalate . . .. . . 47, 53 Dimethyl terephthalate . . .. .. 6, 8 Dimethyl terephthalate . . .. .. 3, 3 . . . . 5, 6 p-Xylene . . . . . . Tcrephthalic acid . . . . . . .. 78, 76 Dimethyl terephthalate . . .. . . 85, 80 Polystyrene . . . . . . .. .. 273, 275 Polystyrene . . .. . . . . . . 242, 233 Polypropylene . . . . . . . . 55, 62 Polypropylene . .. . .. . . 690, 700 Polypropylcne . . . . . . . . 320, 310 Poly(ethy1ene tcrephthalatc) . . .. 10,12 DISCUSSION The methods described have been in use for several years and have proved satisfactory. The electrically fired apparatus has been used to combust such widely different materials as 9-xylene, a fairly volatile liquid; dimethyl terephthalate, a fairly easily sublimable sub- stance ; polypropylene, an easily combustible polymer ; and poly(ethy1ene terephthalate) , a more difficultly combustible material. Doubts have been expressed about the efficiency of the combustion of materials that have been fired with a paper wick or by means of an electrically heated ~0il.1~9 The main objections are that the sample either melts, volatilises or sublimes before the combustion becomes established, or, with the heated coil, that a carbonaceous residue is left on the quartz-wool enclosing the sample. In the apparatus used, the heating element reaches its operating temperature almost immediately the current is switched on, and then the sample is fired within a second or two. The amount of volatilisation or sublimation of a 50 to 100-mg sample that takes place in this time must be small, and, if any does occur, it is more than likely that it condenses on the quartz-wool enclosing the sample. If the heating coil is kept at bright red heat during the combustion, no soot forms on the outside of the quartz-wool pad. Obviously not all compounds combust in the same way, but some measure of control can be gained by varying the thickness of the quartz-wool surrounding the sample. In this way the supply of oxygen to the sample can be varied.620 MCGILLIVRAY AND WOODGER These titration methods have been applied to the solutions obtained after combusting samples in oxygen to determine trace and minor amounts of chlorine and sulphur in semi- micro amounts of organic material. They have also been used for the ultra-micro deter- mination of major amounts in ultra-micro amounts of sample. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES McDonald, A. M. G., Analyst, 1961, 86, 3. Lehner, H., Chimia, 1959, 13, 248. Haslam, J., and Squirrell, D. C. M., J . Appl. Chem., 1961, 11, 244. Juvet, R. S., and Chiu, J., Analyt. Chem., 1960, 32, 130. Hempel, W., Angew. Chem., 1892, 5, 393. Fritz, J. S., and Yamamura, S. S., Analyt. Chem., 1955, 27, 1461. Menis, O., Manning, D. L., and Ball, R. G., Ibid., 1958, 30, 1772. Stock, J. T., and Fill, M. A., Analyst, 1944, 69, 212. Tomlinson, C., Talanta, 1962, 9, 1065. Received July 22nd, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100611

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

Analyst, October, 1966, Vol. 91, $$. 621-624 621 An Assessment of Some Methods of Analysis for Dimethoate Residues in Fruit and Vegetables BY N. A. SMART (Plant Pathology Laboratory, Ministry of Agriculture, Fisheries and Food, Harpenden, Hertfordshire) Three methods for determining residues of dimethoate, 00-dimethyl S-( N-methylcarbamoylmethyl) phosphorodithioate, in fruits and vegetables have been investigated for the purpose of recommending a standard method that can be applied to a variety of temperate fruits grown in the United Kingdom. Laws and Webley’s general method for determining organo- phosphorus insecticide residues was satisfactory for residues of dimethoate in sprouts, lettuces and apples, but not in peas. Chilwell and Beecham’s procedure was also satisfactory a t the 1 or 2 p.p.m.level of added dimethoate in cabbages, lettuces, apples and peas. Gang and Schechter’s method, however, gave variable results. METHODS of determining residues of dime t hoat e, 00-dimet h yl S- (N-me t h ylcarbamo ylme t h yl) phosphorcdithoate, in fruits and vegetables have recently been reviewed by de Pietri-Tonelli et aZ.l A method was required by this laboratory to serve as a basis for collaborative work with other laboratories, with the purpose of recommending a standard method for determining toxic dimethoate residues in a variety of temperate fruits and vegetables grown in the United Kingdom. The method should attain, in the hands of reasonably experienced workers, an accuracy of within a few tenths of a part per million of insecticide, and should be suitable for use in an analytical laboratory that is not equipped with highly specialised apparatus.Of the methods in the literature, some had only been applied to specific crops, and to adapt them for a range of temperate fruits and vegetables further development might be needed. Three generally applicable methods were selected and their performance was assessed over a range of analytical conditions. The method of Laws and Webley,2 which is reasonably rapid, was first investigated on account of its general applicability as a technique in organophosphorus residue analysis. Another method, that of Chilwell and Bee~ham,~ relying on total-phosphorus determination of the cleaned-up residue, was examined because it had been extensively used for dimethoate residue determinations in this country.A third method by Giang and S~hechter,~ which is specific for the thioglycollamide fragment of dimethoate and its oxygen analogue, was also investigated. The results from the assay of dimethoate by these methods are given below. EXPERIMENTAL The details of the methods of analysis chosen for investigation were carefully followed. The crops chosen for examination were some of those, apart from sugar beet, on which the insecticide is most likely to be used in this country, viz., apples, peas, cabbages, lettuces and Brussels sprouts. Assays of the dimethoate used were made. Pure dimethoate in chloroform or aqueous solution was added to, and recovered from, untreated fruits or vegetables at the 1 or 2 p.p.m. level.Under normal conditions of application the use of dimethoate is not expected to lead to a residue of greater than 2 p.p.m. in the treated crop; the accuracy of determinations at about this level is, therefore, important. The oxygen analogue of dimethoate was prepared by the method of Santi and de Pietri- Tonelli5 by using the column purification of Dauterman,6 and was recovered from a crop at the 2 p.p.m. level. In Chilwell and Beecham’s method, adjustment of the pH of the extract was made both with a pH meter and pH papers; close agreement was obtained. In addition to obtaining normal recovery results for lettuce with this method, the effects of (i) small variations in the pH of extraction, (ii) the temperature when distilling off the chloroform, (iii) the tempera- ture and pressure of micro distillation, (iv) procedures for rinsing the cold finger after micro distillation and also (v) the reaction time with perchloric acid on the recovery were investigated as described by Youden.’622 SMART: AN ASSESSMENT OF SOME METHODS OF AXALYSIS [AFZdJSt, VOl.91 When investigating Giang and Schechter’s method it was found necessary to purify the analytical grade sodium tungstate by Folin’s method,s otherwise high blank values were obtained. RESULTS LAWS AND WEBLEY’S METHOD- The slope of the straight line calibration graph obtained with potassium dihydrogen phosphate and the molybdate reagent was 16.3 pg of phosphorus per optical density unit. The slope of the line obtained by assaying different amounts of dimethoate by wet-ashing and colorimetric determination was 116 pg of dimethoate or 15.7 pg of phosphorus per optical density unit. Assays of 60 pg of dimethoate with, and without, the condenser system being attached to the wet-ashing flask were 97 and 94 per cent., respectively, (five determinations each) ; these values were not significantly different.The difficulties described by Brewertong were not observed. Reagent blanks were 2, 1, 1, 1 and 1 pg of apparent dimethoate. TABLE I APPARENT DIMETHOATE IN CROPS Crop* ,4pples . . Lettuces . . Cabbages . . Brussels sprouts Peas . . .. Laws and Webley - Mean, S.D.,*t CLg P.8 9 - . . 1 - . . 2 .. ti 1 3 . . Variable - L . . Chilwell and Beecham 7- 7- - Mean, S.D.,*t Mean, S.D.,*I. Giang and Schechter tLg PFLg Pbrr 6 &4 - tG . 3 1 i l 2 I t 1 Very high - - Very high 3 - - - 1 - * 50-g samples.t Standard deviation from 5 determinations. TABLE I1 RECOVERY OF ADDED DIMETHOATE FROM CROPS Recovery by methods of- Crop* Apples . . Lettuces . . Cabbages . . Brussels sprouts Peas . . . . 7 1 Laws and Webley Chilwell and Beecham Giang and Schechter Amount of ~-------*------, -7 - dimethoate Mean,? S.D.,? Mean, S.D.,? Mean, S.D.,? added f% I-G rug PLg I G PLg 46 1 3 . . 50 68 rt 17 100 70 1 4 48 f 3 . . 50 100 87 &2 93 &9 40 1 8 . . 50 . . 50 40 rt5 100 81 -+8 . . 50 1s: * 4’: 100 34:: f 3’: 74 i 4 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - * 50-g vegetable or fruit samples were used throughout. t Mean and standard deviations from 5 determinations. $ Mean and standard deviations from 4 determinations.Determinations on untreated crops are given in Table I and recovery results in Table 11. In these tables the reagent blank is subtracted from the crop blank and both are subtracted from the recovery results. Peas did not always give low blank values with the method: several samples of fresh peas gave satisfactory low blank values although one sample did not; one sample of frozen peas gave a low blank value but another sample of frozen peas gave very high figures (optical density >1). The number of chloroform extractions from methanol - water necessary to determine the amount of insecticide present was also investigated; recovery of dimethoate added to Brussels sprouts was not significantly different whether 2 or 3 chloroform extractions were used rather than the 4 advocated in the method. Recovery of the oxygen analogue of dimethoate added to lettuces at the 2 p.p.m. level was 25 and 27 per cent.The petroleum ether eluate from an alumina column used in the clean-up of a Brussels sprout blank contained a negligible amount of phosphorus.October, 19661 FOR DIMETHOATE RESIDUES IN FRUIT AND VEGETABLES 623 CHILWELL AND BEECHAM’S METHOD- The slope of the straight line calibration graph with potassium dihydrogen phosphate and the molybdate reagent was 16.4 pg of phosphorus per optical density unit. The slope of the line obtained by assaying different amounts of dimethoate by wet-ashing and colori- metric determination was 128 pg of dimethoate or 17.2 pg of phosphorus per optical density unit.Five assays of 100 pg of dimethoate averaged 96 per cent. purity with a standard deviation of & 1 per cent. Reagent blanks showed 3, 2’2’2 and 1 pg of apparent dimethoate. Determinations on untreated crops are given in Table I and recovery results in Table 11. A cabbage sample kept in deep freeze (-10” C) for 6 months showed little increase in its low apparent dimethoate content. The variations made in the special study of Chilwell and Beecham’s method are set out in Table IV (the key to which is Table 111), which gives the recoveries obtained. TABLE I11 FACTORS VARIED IN CHILWELL AND BEECHAM’S METHOD 10 per cent. acetic acid to pH 4 . . . . . . h to pH 3.2 .. .. .. .. B Distilling off chloroform a t 60” C . . . . . . C a t 70°C .. . . . . . . .. c Rinsing the whole of the micro distillation finger F the lower third of the finger .. . . f Normal factor- Variant factor- Sodium hydroxide solution to pH 7 . . . . B t o p H 8 .. . . . . .. .. b Micro distillation a t 1 mm . . . . . . D a t 3 t o 4 m m . . . . . . .. d Micro distillation a t 200” C . . . . . . E a t 1 7 0 ° C . . . . . . . . .. e Perchloric acid oxidation to white fumes . . G oxidation to less than white fumes . . g TABLE IV COMBINATION OF FACTORS VARIED IN CHILWELL AND BEECHAM’S METHOD AND RECOVERIES Combination f Factor 1 2 3 4 5 6 7 & A o r a , . . . A B o r c . . . . B C o r c . . . . c D o r d . . . . D E o r e . . . . E F or f . . . . F G o r g . . . . G Recovery, per cent.* 7 8 A A A a a B b b B B C C C C C D d d d d e E e e E f f F F f 80 81 83 91 74 * Corrected for blanks.g g G g G a a b b C C D D e E f F 79 80 G g The recovery of the oxygen analogue of dimethoate, added to lettuces at the 2 p.p.rn. level, was 21 and 33 per cent. GIANG AND SCHECHTER’S METHOD- The slope of the straight line calibration graph obtained by assaying different amounts of dimethoate by alkaline hydrolysis and colorimetric determination with purified Folin’s reagent was 215 pg of dimethoate per optical density unit. Reagent blanks were 25, 15, 1, 4 and 2 pg, (mean 9 pg) of apparent dimethoate. Determinations on untreated crops are given in Table I and recovery results for apples in Table 11. DISCUSSION No difficulty was experienced in obtaining satisfactory reagents both for Laws’ and Chilwell’s procedures. Laws’ method worked well with apples, lettuces and Brussels sprouts.However, it was not suitable for peas, whether fresh or frozen, when high blank vaIues were sometinies obtained; a modified clean-up procedure appears to be necessary. It was aIso surprising that when samples of peas gave low blank values, recoveries were very low, about 30 per cent. Chilwell’s method worked well for all of the samples examined. Blank vaIues were low and constant. When the effect on recovery of variation of experimental factors in the method624 SMART was investigated after the manner of Youden, it was found that a variation of 1 pH unit in extraction, of 10” C in the temperature of the distillation of the chloroform, of 3 mm and 30” C in the pressure and temperature of the micro distillation of the insecticide, of 200 per cent.in the amount of rinsing of the cold finger and, further, in the extent of wet-ashing, resulted in no significant difference in the recovery of added dimethoate. This indicated that Chilwell’s method is robust and should stand up well to collaborative study when variations of procedure can arise in different laboratories. Determinations with Giang and Schechter’s method were somewhat disappointing. The variability in results extended back to the blank assays of dimethoate in the absence of plant material. It partly arose from impurities in the phosphotungstic acid. A high blank value was found where Giang and Schechter reported a low one. The oxygen analogue of dimethoate, 00-dimethyl S-(N-methylcarbamoyl) phosphoro- thiolate, is present to a limited extent in the breakdown sequences of dimethoate and is the chief toxic metabolite; the oxygen analogue is more toxic than dimethoate itself.Residue methods for dimethoate should, therefore, be capable of determining the oxygen analogue, and Laws and Webley’s and Chilwell and Beecham’s methods were briefly investigated from this stand-point. Recovery results were, however, hardly satisfactory. The partition coefficient of the system water - chloroform is ~nfavourable.~ I thank Miss J. Gough for experimental assistance in this work. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES de Pietri-Tonelli, P., Bazzi, B., and Santi, R., ifi Gunther, F. A., Editor, “Residue Reviews,” Laws, E. Q., and Webley, D. J., Analyst, 1961, 86, 249. Chilwell, E. D., and Beecham, P. T., J . Sci. Fd Agric., 1960, 11, 400. Giang, P. A., and Schechter, M. S., J . Agric. Fa Chem., 1963, 11, 63. Santi, R., and de Pietri-Tonelli, P., “Ricerche sul Meccanismo D’azione della N-monometilammide Dauterman, W. C., Casida, J. E., Knaak, J . B., and Kowakczyk, T., J . Agric. Fd Chem., 1959, Youden, W. J., J . Ass. 08. Agric. Chem., 1963, 46, 55. Folin, O., J . Biol. Chern., 1930, 86, 179. Brewerton, R. V., N.Z. J . Sci., 1963, 6, 418. Springer-Verlag, Berlin, Gottingen and Heidelberg, 1965, Volume 11, p. 60. dell’acido 0,O-Dimethilditiofosforitacetico,” Montecatini, Milan, 1959. 7, 188. Received Jwly 7th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100621

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

Analyst, October, 1966, Vol. 91, @. 625-629 625 An Improved Method for Determining Residues of Diquat BY A. CALDERBANK AND S. H. YUEN (Imperial Chemical Industries Ltd., Agricultural Division, Jealott’s Hill Research Station, Bracknell, Berkshire) The ion-exchange method for determining diquat residues in potato tubers1 has been modified to give a simpler operating procedure and increased accuracy. Experiments on untreated samples with 0-08 to 0.2 p.p.m. of diquat added showed an average recovery of 76 per cent. with a standard deviation of 11 per cent. The method, with minor modifications, has been found to be generally applicable to other food crops and to water. For a 250-g sample, the limit of determination is 0-01 p.p.m. DIQUAT, or 1 ,l’-ethylene-2,2’-bipyridylium cation, which is manufactured in the form of its dibromide salt, is the active ingredient of Reglone and Preeglone herbicides.The method previously describedl relies on extracting diquat residues from potato tubers by boiling them with N sulphuric acid. The filtered hydrolysate is neutralised and passed through a cation-exchange resin column which retains the diquat together with some of the natural plant constituents. The diquat is then eluted with saturated sodium chloride solution and, after reduction, determined spectrophotometrically in the region of 379 mp. The main disadvantage of this method lies in the low average recovery (59 per cent.), and the rather tedious procedure involved in neutralisation and filtration. Since this method was developed, diquat has found increasing usage on a variety of crops, and the operating variables in the method have been subject to more critical examination.By modifying the conditions at the neutralisation, filtration and elution stages, increased recoveries have been obtained, and the revised method is suitable for routine analysis of large numbers of samples. Further, the method, with minor modifications, is now applicable to other crops, to animal tissue and excreta,2 milk2 and water. The general principles involved in the use of ion-exchange resins in residue analysis have already been discussed in some detail.3 The specific problems that are involved at the various stages of the method for determining diquat, and the modifications made, are described below. EXPERIMENTAL NEUTRALISATION OF ACID HYDROLYSATE- Traces of diquat are held quite firmly to starch, protein and leaf surfaces, and it has been found necessary to boil the macerated plant material with dilute sulphuric acid to release the diquat into solution.If the filtered acid hydrolysate is passed directly without neutralisa- tion through the cation-exchange resin, as described for the complementary bipyridylium herbicide, p a r a q ~ a t , ~ the recovery of diquat in the final effluent is low and the background absorption is high. This is because many of the hydrolysed plant constituents are basic and compete with diquat for adsorption sites on the resin under acid conditions. Further, the retained plant constituents are eluted along with diquat and increase the background absorption in the final effluent.Neutralisation of the hydrolysate has the advantage of giving better recoveries of diquat and considerably lower background absorption. This is because the ionisation of weakly basic materials in the hydrolysate is suppressed under neutral or alkaline conditions, and consequently these materials are then not readily retained by the resin column, leaving more sites available for diquat, the ionisation of which is not appreciably affected by change of pH. Neutralisation of the acid hydrolysis was previously carried out with calcium and sodium carbonates1 This is a tedious operation and has the disadvantage of causing losses of diquat by adsorption on to the precipitated calcium sulphate. The procedure is simplified con- siderably, and losses of diquat a t this stage are eliminated by neutralising with 10 N sodium hydroxide solution.The excess sodium ions in solution (about 0.5 N) do not interfere with the retention of diquat by the ion-exchange resin. Further, the end-point of alkali addition is easier to detect because the solution darkens appreciably and a fine precipitate usually forms.626 CALDERBANK AND YUEN: AN IMPROVED METHOD [Analyst, Vol. 91 The use of a filter aid is desirable to speed up filtration of the hydrolysate. It is now known that Hyflo Super-cel, which was previously used, removed an appreciable amount of diquat from solution by adsorption. Under comparable conditions, a coarser grade of a diatomaceous silicate, Celite 545, was found to retain less diquat, while allowing a quicker filtration.The coarsest grade of filter aid, Celite 560, gave a fast filtration rate, but did not completely remove the fine materials suspended in solution. ELUTION OF DIQUAT- Saturated sodium chloride solution was previously chosen for displacing diquat retained by the cation-exchange resin, Zeo-Karb 225 (containing 8 per cent. of divinylbenzene). Reduc- tion of diquat with sodium dithionite solution could be carried out in this eluant as effectively as in pure solution. Saturated sodium chloride solution, however, suffers from the dis- advantage that it produces tailing, so that recovery of diquat from the resin in 25ml of effluent is only 80 per cent. Hydrochloric acid (1 + 1) and saturated ammonium chloride solution (5 M) produce sharper elution peaks, the saturated ammonium chloride solution giving 95 per cent.recovery in 25 ml and quantitative recovery in 50 ml of effluent; both solutions introduce complications in the final assay. By modifying the conditions of the reduction it has been found possible to determine diquat accurately in saturated ammonium chloride solution, and consequently this solution is adopted as the preferred eluant. A 2.5 per cent. ammonium chloride solution (0.5 M) displaces diquat slowly from the selected cation-exchange resin, and it has been found that a 3.5-g bed of resin can be washed with up to 200 ml of 2.5 per cent. ammonium chloride solution before the adsorbed diquat starts to be eluted. In practice, the resin column is washed with 150 ml of this solution before the diquat is eluted.This procedure reduces the background absorption in the final effluent to a low level. REDUCTION AND DETERMINATION OF DIQUAT- When diquat is reduced in saturated ammonium chloride solution with 0.1 per cent. sodium dithionite in 3 N sodium hydroxide solution,l the colour fades rapidly. The stability of the colour is improved by decreasing the sodium hydroxide concentration, or by increasing the concentration of sodium dithionite. There is a limit to the concentration of sodium dithionite that can be used, because sodium dithionite absorbs strongly in the region of 379 mp, in which region reduced diquat has an absorption maximum. A reagent consisting of 0.2 per cent. sodium dithionite in 0-3 N sodium hydroxide gave a reduced diquat colour that was stable for about 3 minutes.A mixed reducing reagent containing 0.2 per cent. each of sodium dithionite and sodium metabisulphite in 0.3 N sodium hydroxide produced a reduced colour in saturated ammonium chloride solution that was stable for 30 minutes and this was stan- dardised. The conversion of diquat to its free radical is reversible, and if the dithionite is largely oxidised, the free radical will revert to diquat by losing an electron. The mixed reducing reagent may generate a buffered oxidation - reduction potential in which the free radical is more stable. In practice, the colour developed by adding the mixed reagent to ammonium chloride effluents from ion-exchange columns is rather less stable than the colour developed in pure ammonium chloride solution, presumably owing to the lower stability of the reduced back- ground absorption.However, the colour developed in the ion-exchange effluents is stable for 5 to 6 minutes, and thereafter fades slowly. For this reason, measurements of optical density should be completed within 5 minutes of adding the reagent. METHOD APPARATUS- Potato chipping machine, Macerator-A top-drive macerator, obtainable from Townson and Mercer, was used. BoiZing $asks---Round-bottomed flasks of 2-litre capacity fitted, by means of standard Heating mantles-300 watt and 1-litre capacity, with a maximum temperature of 450" C. Spectrophotometer-A Unicam SP600. ground-glass joints, with reflux condensers. An electrothermal heating unit containing six mantles is suitable.October, 19661 FOR DETERMINING RESIDUES OF DIQUAT 627 REAGENTS- Sulphuric acid, 18 N-Add cautiously with stirring 1 litre of sulphuric acid (sp.gr.1.84) to 1 litre of water and dilute the cooled solution to 2 litres with water. Octan-2-01. Celite 545-Obtainable from Johns-Manville Co. Ltd. Sodium hydroxide solution, 10 N. Indicator paper-eg., Universal test paper (Johnsons of Hendon Ltd.). Ethylenediaminetetra-acetic acid, disodium salt (EDTA) solution, 5 per cent. Cation-exchange resin-Permutit Zeo-Karb 225 (52 to 100 mesh), containing 8 per cent. of divinylbenzene in the sodium form. To prepare the column, wash 3-5 g of the resin with water into a 25-ml burette (i.d., 9 to 10 mm, and 50 cm long). Pass successively through the column at about 5ml per minute, 20ml of 2 N hydrochloric acid and 50ml of water.The column is then ready for use. Use a fresh column for each test. Ammonium chloride solution, saturated and 2.5 per cent. STANDARD SOLUTIONS OF DIQUAT- Stock solution, 250 9.p.m.-Dissolve 0.1229 g of pure diquat dibromide monohydrate, C,,H!,N,Br,.H,O (mol. wt., 362.1 ; 50.9 per cent. cation), in saturated ammonium chloride solution and make up to 250 ml with saturated ammonium chloride solution. Solution A, 10 p.p.m.-Dilute 10 ml of stock solution to 250 ml with saturated ammonium chloride solution. Also prepare standard solutions B, C, D and E (containing 1.5, 1.0, 0.5 and 0.25 p.p.m. of diquat) by diluting 15, 10,5 and 2.5 ml, respectively, of solution A to 100 ml with saturated ammonium chloride solution. These solutions are stable under normal laboratory conditions, but must not be exposed to sunlight. Reducing reagertt-Dissolve 0.2 g each of sodium dithionite and sodium metabisulphite in 100 ml of 0.3 N sodium hydroxide.This solution should be prepared immediately before use, and should on no account be used after keeping for more than 1 hour. It should be stored in 1-02 bottles (tightly covered with screwed-on lids) which are kept in a desiccator. Solid sodium dithionite is unstable in the presence of moisture. EXTRACTION AND CHROMATOGRAPHIC SEPARATION OF DIQUAT- Take about 1000 g of potato tubers at random from the sample provided, wash free from soil and remove surplus water with a dry cloth. Cut each tuber into four approximately equal segments and reject two that are opposite.Cut the remaining segments into small pieces and mix them thoroughly. Weigh out 250 g of tuber pieces into a macerator jar, add 200 ml of water and macerate for 3 minutes. Transfer the macerated material to a boiling flask. Rinse the jar with 50 ml of water, and add the rinsings to the contents of the flask, followed by 25 ml of 18 N sulphuric acid and 1 ml of octan-2-01. Mix well, attach a reflux condenser to the flask and heat to boiling on a heating mantle. Swirl the mixture occasionally, to prevent local overheating and charring, until the solution is boiling steadily. Boil under reflux for 5 hours and allow to cool (the solution can be left overnight at this stage). Prepare a Celite 545 filter as follows: moisten a Whatman No, 5 filter-paper with water under suction on a 16-cm Buchner funnel, supported on a 2-litre filter flask.Pour 150 ml of an aqueous suspension containing l o g of Celite 545 over the paper, suck dry, and discard the filtrate. Wash down the condenser with 50 ml of water, and filter the contents of the flask by moderate suction through the prepared filter. Suck the residue dry and wash with 150ml of water. Transfer the filtrate to a 2-litre beaker, and neutralise to pH 8 to 9 by adding, slowly with stirring, 10 N sodium hydroxide (about 45 ml) with a Universal test paper as external indicator (the colour deepens near the end-point and a fine precipitate usually forms). Add 50 ml of EDTA solution (the pH of the solution should now be 6 to 7) and re-filter the solution through a Celite 545 filter as before.Transfer the solution (900 to 1000 ml) to a 1-litre separating funnel suspended above the prepared resin column. Allow the solution to percolate through the column at a flow-rate of 5 to 10 ml per minute. Remove the funnel, and wash the column at 3 to 4 ml per minute628 CALDERBANK AKD YUEN: AN IMPROVED METHOD [Analyst, Vol. 91 with 150 ml of 2.5 per cent. ammonium chloride solution. Elute the diquat by passing saturated ammonium chloride solution through the resin column at about 1 ml per minute. Collect 50 ml of the effluent in a 50-ml calibrated flask, and mix. DETERMINATION OF DIQUAT- Transfer by pipette 10.0 ml of the effluent into a 15-ml glass-stoppered test-tube, add 2.0 ml of reducing reagent, and mix. Within 5 minutes, measure the optical density of the solution at 375, 379, 383 and 385 mp in a 4-cm optical cell, against a reference solution consisting of 10 ml of saturated ammonium chloride solution and 2 ml of reducing reagent.Call these E,,, and, respectively. Measure, concurrently with each series of analyses, the optical densities at 379mp of 10-0-ml aliquots of standards B, C, D and E, after adding 2-Om1 of reducing reagent. Construct a calibration graph relating optical densities of the standards to concentrations of diquat in p.p.m. CALCULATIOX- absorption by means of equations (1) and (2), and call the corrected optical densities and respectively. Correct the optical density at 379mp of the sample solution under test for irrelevant El379 == 3.79 E.379 - 2.28 E375 - 1.52 E385 . . .. * * (1) E”379 = 2.49 (2E37g - E375 - E.383) .. .. .. * (2) Use the mean of E’3,9 and Then the diquat content in parts per million to read off from the prepared calibration graph the concentration of diquat in the final effluent. Call this amount Y p.p.m. Y 100 5 percentage recovery - - - x RESULTS RECOVERY- To establish the accuracy of the improved method, 47 recovery experiments were carried out on 250-g portions of untreated potato tubers, with 0.08 to 0.2 p.p.m. of diquat added before hydrolysis. The recoveries (Table I) ranged from 51 to 97 per cent. with an over-all mean of 76 per cent. (standard deviation is 2 1 1 per cent.). TABLE I RECOVERY OF DIQUAT ADDED TO UNTREATED POTATO TUBERS Diquat added, Diquat recovery, per cent. Number of experiments p.p.m. mean standard deviation 22 0.08 7 5 f 8 19 0-1 76 & 13 6 0.2 79 & 14 By means of the two derived equations the background absorption contributed by natural plant constituents is eliminated, leaving only that contributed by diquat.The mean corrected optical densities at 379 mp, measured in a 4-cm optical cell and obtained on un- treated potato tubers, were usually found to be within 20.02, and this is equivalent to apparent diquat contents of +0-007 p.p.m. Taking into account the average apparent diquat content of untreated samples, the average recovery and the size of sample taken, the limit of determination for the improved method is considered to be 0.01 p.p.m. APPLICATION TO OTHER CROPS AND TO WATER- Because of its quick action and lack of residual activity in the soil, diquat has proved useful in a wide range of crops for pre-emergence and post-emergence weed control, pre- harvest desiccation and aquatic weed contr01.~ f6 With minor modifications, the improved method has been applied satisfactorily to a variety of other food crops and to water.For samples of high dry matter, such as seeds, and for those, such as grass and cereal straws, that have been in direct contact with the spray and are likely to contain higher residues,October, 19661 FOR DETERMINING RESIDUES OF DIQUAT 629 the amount taken for analysis can be conveniently reduced to 25 or 50 g. The diquat residues present in these samples can be quantitatively extracted into N sulphuric acid by boiling for 3 hours. Diquat is also an efficient aquatic herbicide at concentrations in the water of 1 p.p.m.or below. For analysing treated water, acid hydrolysis is not required, and traces of diquat that may be present are concentrated in the usual way by passing the sample solution through the resin column, followed by washing and elution. The results obtained by applying this ion-exchange method to various crops and to water are summarised in Table 11. Applica- tion of the method to the determination of residues of diquat in animal tissue, excreta and milk is described elsewhere.2 TABLE I1 SUMMARY OF THE RESULTS OBTAINED BY APPLYING THE IMPROVED METHOD TO VARIOUS CROPS AND TO WATER Period of boiling, Limit of determination, Recovery, Sample applied* Size of sample hours p.p.m. per cent. Potato tubers Fruits . . ::} 250g Cereal grains Peas and beans . . Cotton seeds . ’‘1 . Rape seeds . . Sunflower seeds , . J Grass . . Cereal straws : } Vegetables. . .. 50g Alfalfa and clover 25 g Silage . . .. 5 3 3 0 0.01 66 to 80 0.05 0.1 60 to 75 70 to 85 0.03 0.003 90 to 100 * These samples have been analysed with satisfactory results. We thank Mr. R. H. McKenna for assistance with the experimental work. REFERENCES 1. 2 . 3. 4. 5. 6. Calderbank, A., Morgan, C. B., and Yuen, S. H., Analyst, 1961, 86, 569. Black, W. G. M., Calderbank, A., Douglas, G., and McKenna, R. H., J . Sci. Fd Agvic., in the press. Calderbank, A., in Gunther, F. A., Editor, “Residue Reviews,” Springer-Verlag, Berlin, Gottingen and Heidelberg, 1966, Volume 12, p. 14. Calderbank, A., and Yuen, S. H., Analyst, 1965, 90, 99. Calderbank, A., and Crowdy, S. H., Ann. Rep. Appl. Chew., 1962, 47, 536. Springett, R. H., Outl. Agric., 1965, 4, 226. Received December 13th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100625

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

630 Analyst, October, 1966, Vol. 91, PP. 630-637 Improvement in Performance of a Simple Atomic Absorptiometer by Using Pre-heated Air and Town Gas BY R. A. G. RAWSON (Rothamsted Experimental Station, Harpenden, Hertfordshire) A simple atomic absorptiometer for the routine determination of mag- nesium, copper and zinc is described. A pre-heated air supply to the atomiser increased the response efficiency of the absorptiometer 16-fold. An adjustable slot in the burner enabled a wide range of fuel gases to be used. A horizontal monochromator slit gave improved light transmission and absorption. ALTHOUGH the conventional atomic-absorption method by Box and Walshl was adequate for determining magnesium, it was too insensitive for copper and zinc. Such insensitivity was probably caused by incomplete vaporisation of solids and was enhanced, for magnesium, by interfering elements that formed refractory compounds.Very hot flames, above 2000" C, generally improve the sensitivity, but these are costly, inconvenient to produce and have complex spectra in which there is more interference. Improved atomisation, which produced smaller solid particles and converted the solution to aerosol more efficiently, was a better alternative. This was achieved by pre-heating the air and gas supplies to the atomiser and spray chamber, respectively, and by using a newly designed burner. The need for such improvement was expressed by Allan2 and by Russell, Shelton and Walsh.3 INSTRUMENTATION APPARATUS- Source of emission-The conventional sealed-off hollow-cathode discharge tubes were made by Messrs.Hilger and Watts Ltd. Ancillary eq.uipment-The power pack for the hollow-cathode discharge tube and E.H.T. supplies for the photomultiplier valve (R.C.A. IP28) and the amplifier were as specified by Box and Wa1sh.l All electronic equipment was simple in design, cheap and reliable. Housing for bwner and lamp-The housing was made of 10-mm Asbestolite and the burner section was amply ventilated. The lamp and burner were in adjacent compartments separated by an Asbestolite partition with a hole of 25mm diameter in the appropriate - U .- - g : 2 8 0 5 36 38 40 41 40 3% 36 Extinction, per cent. Fig. 1. Town gas flame showing ad- vantage of horizontal monochromator slit over vertical slit with 2-mm burner slot widthRAWSON 631 position. A cowling and discharge duct above the burner is necessary in a laboratory with inadequate ventilation.Optics-The only optical system required in this unit was the monochromator, because the hollow-cathode lamp emits a near-parallel light beam. Focusing the beam, therefore, had little merit and did not warrant the quartz lenses that are needed for ultraviolet trans- mission. A horizontal monochromator slit was used to take advantage of the horizontal distribution of atomic vapour just above the blue cone of the flame (Fig. 1). With a vertical arrangement, a collimating slit placed after the flame improved the sensitivity by decreasing the unabsorbed emission transmitted in the upper part of the beam. After collimation, the smaller signal obtained from the detector had to be amplified more to obtain full-scale deflection.An optical bench was unnecessary because the components were rigidly constructed and easily aligned on the optical axis. E Fig. 2. Adjustable burner assembly. (Lettered parts of the apparatus are referred to in the text) The b.urn.er (the patent has been applied for)-This (Fig. 2) was designed for use with various fuel gases and was made from a stainless-steel tube of the following dimensions and standard wire gauge (s.w.g.) : main burner tube, A, and trunk tube, D: 25 mm over-all diameter (o.d.), 22 mm internal diameter (id.), 16"s.w.g.; tube, B: 22 mm o.d., 19 mm i.d., 16" s.w.g.; branch tubes, C: 19 mm o.d., 15 mm i.d., 14" s.w.g. The slot, 120 mm long, 3 mm deep, with width adjustable by arm, F, could be set between 0 and 2-25mm.The deep walls of the slot help laminar flow of the pre-mixed gases and also increase the quenching ability of the slot, thereby lessening the risk of "flashback." Narrow lips, E, were welded on the edges of the fixed tube, A, and the rotating tube, B, to maintain equal thickness in all positions, The inner walls of the slot were polished, straight and exactly parallel. The Y-shaped burner (each leg was equidistant from the centre and either end of the burner) was better than the T or "fish-tail" design, because these caused turbulent gas flow, and gases and solid particles would fractionate more. Another advantage was that each leg of the Y burner fed only a 60-mm length of slot. The main volume of gas carried by the trunk was equally split at the junction of the legs, ensuring even distribution to each half of the burner head. The long threaded rod, G, along the central axis of the burner head allowed the slot to be locked at any width (by using feeler gauges) and also acted as a baffle.Cooling the burner with water was unnecessary because, with pre-heated gases, a constant operating temperature was reached much more rapidly, contrary to previous observations by Clinton.* Pre-heating the incoming gases to 328" K increased the volume of gases above the blue cone only slightly, because this increased the normal flame temperature (2073" K) by only 40" K. The wide burner slot gave complete freedom from clogging and decreased flame noise caused by pressure fluctuations, because the back pressure in the spray chamber beneath the burner was only about 1 mm of water.Atomiser-The atomiser from the E.E.L. flame photometer (Design Xo. 866150), which has an adjustable capillary to control the pressure drop across the nozzle, was modified. The capillary was removed, the orifice of the nozzle drilled with a No. 58 drill, and the capillary replaced. Spray clinmnber-The chamber used in the E.E.L. flame photometer was modified by drilling a hole, of 8 mm diameter, near the end cap to hold a 0" to 250" C thermometer. A water seal on the drain tube maintained constant pressure; it consisted of a glass U-tube with the delivery leg bent over so that its contents, when displaced, dripped steadily from this632 [Analyst, Vol. 91 leg. A glass ring (19 mm i.d.) with 6 holes on its inner surface (to focus the gas stream on the spray jet) was fitted on the gas inlet inside the end cap so that the gas surrounded the spray (Fig.3). This improved the mixing of gases and the heating of spray and lessened flame turbulence. RAWSON: IMPROVEMENT IN PERFORMANCE OF A SIMPLE ATOMIC /Gas ring Fig. 3. End cap of spray chamber Pre-heaters (the patent has been applied for)-Tubular brass pre-heaters (5 mm id., 8 mm 0.d. and 2.5 mm apart) in the air and gas lines passed through the flame 25 mm above the burner head, pre-heating the air and gas before entering the spray chamber. The sections of tubing from the flame to the spray chamber were insulated with asbestos lagging. Both heaters were connected to the spray chamber with brazed screw sockets and asbestos washers, because soldered connections broke down at such high temperatures.Stainless-steel tubes would be better than brass. Pre-heaters, spray chamber and burner were rigidly fixed together to ensure alignment when the assembly was moved. Monochromator-A diffraction-grating monochromator (Model D292), supplied by Hilger and Watts Ltd., was mounted with the entry slit horizontal. The monochromator is sensitive to heat and was shielded from the burner housing by asbestos. A short length of 12.5 mm i.d. copper tubing (its interior surface painted with camera black), fitted to the entry slit, decreased stray light entering between the burner housing and the monochromator. PhotomuZti$&er-The photomultiplier valve (R.C.A. IP28), designed to respond to wave- lengths between 2150 and 5500A, also operated satisfactorily at the zinc emission line (2138-5A).The position of the valve, in relation to the exit slit of the monochromator, was critical. The signal was strongest when the monochromator slit was parallel to the photo-cathode of the valve, i.e., at 90” to the valve base. The amplifier meter reading was most stable when the E.H.T. supply to the photomultiplier valve was about 860 volts and the amplification control did not exceed 30 per cent. of its total movement. Pump ~ O Y town gas-A small diaphragm pump (Charles Austin Pumps Ltd.; capacity 5 litres per minute at 5 p.s.i.) controlled by a needle valve and flow-meter was used to ensure a constant supply of town gas. Air-compressor-A compressor, supplying clean dry air continuously at about 25 litres per minute at 50 p.s.i., was controlled by a diaphragm reduction valve, needle valve and flow-meter.RESULTS AND DISCUSSION HOLLOW-CATHODE LAMPS- The hollow-cathode lamps were run at the lowest possible currents (e.g., 4 to 5 mA for To prevent clean-up” of the filler gas, lamps were run for 8 hours continuously, once every month, magnesium) for minimum spectral line half-width and maximum sensitivity. when not in use. 4 < CHOICE OF FUEL- Propane and methane proved to be unsuitable because their burning velocities in air were too small (40 cm per second) and caused their flames to “blow off’’ unless the streaming velocity was kept slow, particularly when water vapour was introduced. A slow streaming velocity produced a weak flame that was vulnerable to draught and obstruction.AlthoughOctober, 19661 ABSORPTIOMETER BY USING PRE-HEATED AIR AND TOWN GAS 633 these gases had large B.T.U. values and produced more total heat than town gas, the hottest part of their flames was cooler, so that they were less able to vaporise solids. Town gas, which had a burning velocity of over 100cm per second because of its high hydrogen content, was used at a moderate flow-rate to give a firm flame free from interference. Town gas is not an ideal gas to use, because its composition varies. The only factors that affect reproducibility are its B.T.U. value, burning velocity and specific gravity, although suppliers of town gas rigidly control these factors for economic reasons. With hydrogen as a fuel (burning velocity in air 270 cm per second) the burner slot must be closed to 0.50 mm to prevent “flash back.” This increases back pressure in the spray chamber, increases the streaming velocity and narrows the flame, which gives unsatisfactory results with the horizontal monochromator slit.Increasing the streaming velocity increases flame turbulence and extends the flame front of the pre-mixed cone, thereby increasing the surface area of the reaction zone. This elongates the cone and, therefore, the contours of atomic vapour density, which then do not match the monochromator slit and so sensitivity is lost. FLAME TEMPERATURE- The flame temperature was increased by raising the carbon monoxide content of town gas from 20 to 50 per cent., but the hotter flame had no effect on magnesium absorption.Much hotter flames obtained with acetylene vaporised refractory compounds, but improved atomisation, which made the particles of such compounds smaller, was a better method. INTERFERING ABSORPTION LINES When zinc was determined a t 2138A, the flow of town gas through the burner, before ignition, caused complete extinction of the source beam. This was caused, partly by carbon monoxide in the town gas, which has a strong absorption line at this wavelength, but mostly by hydrogen sulphide which decomposes endothermically at this wavelength. The radiation responsible for this decomposition ranges5 in wavelength from 1800 to 2300 A. Accord- ing to the theory of the structure of flames, about half of the carbon monoxide and all of the hydrogen sulphide is oxidised in the pre-mixed cone on ignition. This was confirmed by a fall in extinction to between 40 and 50 per cent, on ignition.The amplifier gain was increased to restore 100 per cent. transmission, but this greatly decreased the sensitivity for zinc. Acetylene also had a strong absorption line at 2138 A that persisted after ignition, particularly in a fuel-rich flame, because of carbon monoxide in the interconal gases produced by combustion. Carbon dioxide, nitrogen, oxygen, methane and propane did not absorb at this wavelength. ATOMISATION- Increasing the diameter of the air nozzle of the atomiser lessened the air pressure behind the nozzle, which slowed heat transfer to the capillary. This permitted a hotter air supply to be used without vaporising the solution within the capillary, which would retard uptake of solution and cause the spray chamber to over-heat.The air was hot enough to warm the test solution as it passed through the capillary and to supply the latent heat required to vaporise it at the nozzle. Increasing the diameter of the nozzle also slowed the air passing out of it and allowed the vapour pressure of the solution to be increased without vaporising it within the capillary. ADJUSTMENT OF THE ATOMISER AND PRE-HEATERS- Great sensitivity was attained economically by feeding a small volume of hot air at low pressure, about 25 p.s.i., to the atomiser, but this gave a weak flame susceptible t o draught and pressure variation. A firm flame was, therefore, preferred with an air pressure of 30 to 35 p s i .A t this pressure, with the appropriate amount of town gas, the pre- heaters were at a red heat over about 2 inches of their length. The atomiser then took up solution rapidly and the spray chamber became correspondingly cold. The capillary of the atomiser was slowly adjusted until uptake was only about 2.5 ml per minute and then left to operate on distilled water for about 15 minutes to allow the spray chamber to reach constant temperature. The capillary was finally adjusted by rotating it in an anti-clockwise direction, about 5” at a time, until the run-off from the spray chamber stopped, indicating 100 per cent. atomisation.634 [Analyst, Vol. 91 To obtain the correct final balance, the air pressure to the atomiser was altered by 1 p s i at a time and the gas flow adjusted.This altered the heat output to the pre-heaters and made it possible to adjust the temperature in the spray chamber to 54" C After each run, the flame was extinguished and the spray chamber flushed out with dis- tilled water by passing cold air through the atomiser for 5 to 10 minutes to remove solids trapped by the spray chamber. The use of a secondary air supply, which did not pass through the atomiser and could therefore be heated without limit, achieved complete atomisation with ease but did not improve the sensitivity appreciably. Probably it did not contribute to the disintegration of the spray but evaporated the larger droplets that otherwise would have fallen out to the drain.Solid aerosol particles from such large droplets would not be easily vaporised by the flame. With pre-heated air, the hot droplets of spray were most probably split instantly into minute fragments on meeting the incoming air at about 200" C. The very fine particles of solid aerosol resulting were easily vaporised by the flame, giving a high density of atomic vapour. This physical change of state is possible only when the necessary latent heat of vaporisation is readily available. RAWSON: IMPROVEMENT IN PERFORMANCE OF A SIMPLE ATOMIC 1" C. USE O F ORGANIC SOLVENT- The use of an organic solvent in the test solution was said to increase sensitivity greatly under normal conditions, as reported by Allan.2 However, with pre-heating, 20 per cent. v/v of ethanol or propanol had no significant effect.Also, extracting and concentrating with an immiscible organic solvent was not necessary with copper because the sensitivity was adequate. ATOMISATION EFFICIENCY- With this system atomisation was efficient (Table I) when the spray chamber was kept at a constant temperature of 54" C during continuous uptake of distilled water or test solutions. TABLE I PERFORMANCE CHARACTERISTICS OF COMMERCIAL INSTRUMENTS AND OUR APPARATUS Unicam flame spectrophotometer SP900 r----A------, E.E.L. flame Our Cool flame Hot flame photometer instrument Air pressure, p.s.i. .. . . .. 22 29 12 34 Rate of uptake, ml in 10 minutes . . 39 43 24 21 Atomisation, efficiency per cent. . . 7 - 7 7 13.5 100 Fall-out to drain . . .. .. . . 36 40 21 ni 1 IMPROVED SENSITIVITY- An accurate comparison of this system with, and without, pre-heating is not possible because altering any one factor inevitably affects others.However, removing the pre- heaters, without manually altering anything else, decreased sensitivity to one-sixth, despite the solution being taken up faster (Fig. 4). This fraction becomes one-sixteenth when allow- ance is made for the increase in the rate of solution uptake (Table 11). TABLE I1 EFFICIENCY OF CONVERSION TO ATOMIC VAPOUR Soh tion, Percentage Volume p.p.m..of Uptake, of solution atomised, Optical magnesium nil per minute atomised ml per minute density A. With pre-heaters . . 0.065 1-78 100 1.78 0.152 B. Without prc-heaters 0-400 4-76 8 0.38 0.152 C . Ratio A/B . . . . 6.150 2.67 - 4-68 - To obtain the same optical density without, as with, pre-heating, 6.15 times the con- centration of magnesium and 2.67 times the amount of test solution were required, Despite the sensitivity increases of 6.15-fold with pre-heating, the volume of solution atomised in- creased only 4.68-fold.This figure does not allow for about 45 per cent. of solid aerosol trapped by the spray chamber. So there was a vast improvement in converting the solution to atomic vapour. Without pre-heating, only 27 per cent. of solid was trapped.October, 19661 ABSORPTIOMETER BY USING PRE-HEATED AIR AND TOWN GAS 635 Magnesium, p.p.m. Fig. 4. Absorption by magnesium atoms, with and without yre-heated air and gas supplies, in the presencc of 300 p.p.m. of strontium as strontium chloride a\. With pre-heaters: ratc of uptake = 1.78 ml per minute; spray chamber temperature = 55” C ; and atomisation efficiency = 100 per cent.B. Without pre-heaters: rate of uptake = 4.76 ml per minute; spray chamber temperature = 15” C; and atomisation efficiency = 8 per cent. THE FLAME- In normal operation (burner slot width 2 mm), the flame was 125 mm long, 11 mm wide and had a blue cone about 6 mm high when aspirating solution. Air, at room temperature and pressure, was supplied at 21.6 litres per minute. The volume of town gas used varied according to the type of flame required, but was usually about 5-4 litres per minute. Its average composition was hydrogen, 49 per cent. v/v; carbon monoxide, 20 per cent. v/v; methane, 10 per cent. v/v; nitrogen, 7 per cent. v/v; carbon dioxide, 5 per cent.v/v; un- specified hydrocarbons, 5 per cent. v/v; ethane, 4 per cent. v/v; and oxygen, 0.6 per cent. v/v: traces of sulphur compounds were always present. To find the vertical distribution of “ground-state” atoms in the flame, the monochromator slit was collimated to 2 x 0.20 mm and the extinction noted as the burner was lowered by 1 mm at a time (Fig. 1). This indicated the sensitivity that might be expected from a 0.2 x 7-mm vertical monochromator slit. The distribution of “ground-state” atoms in a cross-section of the flame was examined by superimposing a 1-mm vertical slit on the 0-2-mm horizontal slit. This showed that the intensity of the source beam fell rapidly as the vertical slit was moved away from its centre, and the amplifier gain was increased to restore full- scale deflection.This method proved unsatisfactory because the response curve of the amplifier had then to be taken into account. Instead, the instrument was set at full-scale deflection when the vertical slit was at the centre of the source beam. Then, without altering the amplifier gain, the extinction was expressed as the percentage of the incident light as the vertical slit was moved. ADVANTAGES OF THE HORIZONTAL MONOCHROMATOR SLIT- The monochromator slit gave better results when horizontal, with a broad flame, than when vertical, with broad or narrow flames. Fig. 1 shows that the density of atomic vapour decreases faster in the vertical direction than in the horizontal, giving the horizontal slit a distinct advantage. Collimating the horizontal slit to 4 or 5mm increased sensitivity slightly but was of little advantage because the increased amplifier gain required made the636 [A~zaZyst, Vol.91 meter less stable. Light transmission was a t a maximum because it was not necessary to collimate the source beam to the same extent as with a vertical slit. This was advantageous because the light transmitted by the diffraction grating monochromator was only one-third of that transmitted by a Unicam SP500 prism instrument under the same conditions. If less light was transmitted, full-scale deflection of the meter was obtained by passing more current through the hollow-cathode lamp (this broadened the spectral line-width of the source), by increasing the width of the monochromator slit (causing a loss of resolution), increasing the E.H.T.to the photomultiplier, or by increasing the amplifier gain (both of which caused unstable readings on the meter). RAWSON: IMPROVEMENT IN PERFORMANCE OF A SIMPLE ATOMIC 5 p.p.m. Fig. 5. Calibration lines for zinc, copper, manganese and calcium A = Zinc Monochromator slit = 0.30 mm Amplifier gain = 20 per cent. Lamp current = 10 mA Background noise = &On25 per cent. Flame emission = 0 per cent. Flame extinction = 42 per cent. Analytical limit (5 per cent. deflection) = 0.05 p.p.m. C = Manganese Monochromator slit = 0.2 mm Amplifier gain = 20 per cent. Lamp current = 10 mA Background noise = +Om25 per cent. Flame emission = 3 per cent. Analytical limit (5 per cent. deflection) = 0.25 p.p.m. B = Copper Monochromator slit = 0.10 mm Amplifier gain = 20 per cent.Lamp current = 5 mA Background noise = 0-75 per cent. over-all Flame emission = 1 per cent. Flame extinction = 3 per cent. Analytical limit (5 per cent. deflection) =0.125 p.p.m. D = Calcium (with 500 p.p.m. of lanthanum) Monochromator slit = 0.2 mm Amplifier gain = 23 per cent. Lamp current = 8 mA Background noise = f0.3 per cent. Flame emission = < 1 per cent. Analytical limit (5 per cent. deflection) = 0.25 p.p.m. COPPER, ZINC, MANGANESE AND CALCIUIIG- Fig. 5 shows calibration lines for copper, zinc, manganese and calcium. aids, e.g., optical system or scale expansion unit, were used at any time. No mechanical INTERFERENCE- Phosphorus, and particularly aluminium, interfered with the determination of mag- nesium, but all interference was eliminated by adding strontium as strontium chloride toOctober, 19661 637 the test solutions and standards to give 300 p.p.m. of strontium. The rich colour in the flame from strontium was also a good indicator of contamination. Other elements in concentrations commonly found in soil extracts and plant digests did not interfere when determining copper and zinc. Sensitivity for calcium was adequate but interferences were not examined. CONCLUSION ABSORPTIOMETER BY USING PRE-HEATED AIR AND TOWN GAS The instrument described is so improved in sensitivity that many elements can be directly determined without preliminary concentration. Some of the main requirements in atomic-absorption technique are : producing a very fine solid aerosol; converting solution to aerosol efficiently ; having the smallest practicable volume of interconal gases ; and having a long light path through the flame. Further detailed investigation is necessary to curtail the effects caused by the formation of refractory compounds. Devices that would make the particles of solid aerosol still smaller and so diminish chemical and physical interferences are needed. I thank Mr. B. Edwards for help with the burner design. REFERENCES 1. 2. 3. 4. Clinton, 0. E., Ibid., 1960, 16, 985. 5. Box, G. F., and Walsh, A., Spectrochiwz. Acta, 1960, 16, 255. Allan, J. E., Ibid., 1962, 18, 605. Russel, B. J., Shelton, J. P., and Walsh, A., Ibid., 1957, 8, 317. Mellor, J. W., “A Comprehensive Treatise on Inorganic and Theoretical Chemistry,” Longmans, Received January 7th, 1966 Green and Co., London, New York and Toronto, 1930, Volume 10, p. 128.

ISSN:0003-2654    

DOI:10.1039/AN9669100630

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

638 Analyst, October, 1966, Vol. 91, p p . 638-646 Activation Analysis for Titanium and Niobium with Fast Neutrons BY V. T. ATHAVALE, H. B. DESAI, S. GANGADHARAN, M. S. PENDHARKAR AND M. SANKAR DAS (Analytical Division, Atomic Energy Establishment Trombay, Bombay, India) Fast-neutron activation methods have been developed for determining titanium and niobium. Scandium-47 formed by the (n,p) reaction on titan- ium-47 and yttrium-90 formed by the (n,cc) reaction on niobium-93 have been used for determining these elements. Irradiations were carried out in the swimming-pool reactor “Apsara” a t the Atomic Energy Establishment Trombay, a t a fission-flux of 3 x 10l1 n per cm2 per second. Radiochemical procedures were developed for the isolation of scandium and yttrium from the irradiated samples.The methods have been applied t o the determination of titanium in stabilised steels and the standard rocks G-1 and W-1 and of niobium in stabilised steels. The advantages and limitations of the methods are discussed. A FEW studies have been reported in the literature on the determination of titanium and niobium by the activation-analysis technique. Brooksbank, Leddicotte and Dean1 determined titanium in aluminium-based alloys by using the reactions 46Ti (n,p) 46Sc and 48Ti (n,p) 48Sc. Gruverman and Henninger2 applied the reaction 48Ti (n,p) 48Sc to determine high concen- trations of titanium in corrosion residues. Kim and Meinke3 reported on the analysis of titanium in different materials by using the (n,y) reaction on titanium-50. Activation analysis of niobium by the (n,y) reaction was investigated by Brownlee,4 Kim5 and Kim and Meinke,‘j who used rapid radiochemical separation for the isolation of the niobium-94m (half-life 6.6 minutes) activity.Leddicotte ef a1.’ used the (n,y) reaction for the non-destructive determination of niobium in steels. So, except for the studies of Brooksbank et al. and Gruverman and Henninger, these investigators have used the short-lived radionuclides formed by the (n,y) reactions on these elements. A major difficulty in the use of short-lived nuclides for activation analysis is that the induced activity decays rapidly with time so that pneumatic facilities are required for irradi- ation and delivery of samples. Even with these facilities, direct y-spectrometric analyses with 51Ti and S4nzXb are difficult, as the energy of the y-rays associated with the decay of these nuclides is low, and therefore subject to serious interference from other nuclides in the sample.This necessitates chemical separations, which have to be rapid in view of the time factor. The difficulties in developing satisfactory procedures for this purpose have been discussed by Brownlee and arise from the complex solution chemistry of these element^.^ 99 In some recent studies, Sankar DasloY1l and Yule, Lukens and Guinn12 have shown the possibilities of using fast neutrons for activation analysis, a field of study that has not yet been fully explored. The application of the (n,p) reaction on titanium and the (n,cc) reaction on niobium for the activation analysis of these elements is described.FEASIBILITY OF THE FAST-NEUTRON ACTIVATION METHOD- The different (n,p) and (n,ct) reactions on titanium and niobium, together with their threshold energies,13 the cross sections for a spectrum of fission neutronslO and the half-lives and the characteristics of the radiations emitted by the product nuclides14 are given in Table I. Reactions giving product nuclides of short half-lives are not listed as they suffer basically from the same disadvantages as the short-lived (n,y) products. The very long-lived products are also not included on practical considerations of sensitivity. The yields from these reactions for the irradiation of 1 mg of the element for 40 hours at a fission-flux of 3 x 10l1 n per cm2 per second are given in the sixth column of Table I.The flux value is as monitored by 32S (n,p) 32P, with an assumed cfs of 60 mb,15 near the coreTABLE I THRESHOLD REACTIONS IN FAST-NEUTRON ACTIVATION ANALYSIS Fission flux = 3 x 10l1 n per cm2 per second: time of irradiation = 40 hours Cross section for a spectrum of Yield, ET. fission neutrons, d per minute Radiation and energy, Element Reaction MeV * 6 8 (mb) Half-life Per mg MeV 8- 0.367 (100%) y 0.885; 1.119 /3- 0.439 (66%), 6- 0.622 (34%) ; y 0.155 y 0.99; 1.04; 1.314 y 0-48; 0-83; 1.31 1.61 8.6 84 days 2.3 x 103 3.4 days 1 x 106 Titanium . . . * . . 46Ti (n,p) 4'3Sc 47Ti (n,p) 47Sc - 0.17 20.4 4sTi (n,p) 48Sc 3.28 0.28 44 hours 2.2 x 101 18- 0.64 (100%) 3.37 0.047 4.7 days 1.1 x 102 /?- 1.94 (17%); 0.66 (830/,) 50Ti (n,cc) 47Ca Niobium .. . . . . 93Nb (n,a) -5.12 0.077 64 hours 3.1 x 103 8- 2-27 (100%) No gamma S%e)( (EjtiE * " , where ET is the threshold energy, o(E) the excitation function and $(E) is given by the relationship13 +(E) = -Ee sinh 1 / 2 E Of8 = -- --- j %El a E 43 W CD640 ATHAVALE et al.: ACTIVATION ANALYSIS FOR [Analyst, Vol. 91 of the “Apsara” Reactor, Trombay. It may be seen that the yields are high enough to be of interest in activation analysis. The application of scandium-47 and yttrium-90 for the analysis of titanium and niobium are described here. EXPERIMENTAL TRRADIATION- The samples and standards (spectrographically pure oxides of titanium and niobium) were irradiated in 8-mm diameter polythene tubes for 20 to 40 hours in the A, position of the “Apsara” reactor at Trombay.In this position, the samples are irradiated a t a distance of 9 cm from the nearest “13 plate 46 per cent. enriched uranium fuel element.” The samples and standards were wrapped in 30-mil cadmium foils in order to cut down the induced (n,y) activities. After irradiation, the samples were cooled for 10 to 12 hours before they were processed for isolating scandium and yttrium activities. C r ~ ~ ~ f ~ ~ ~ ~ SEPARATION- Stevenson and Xervikl6 have reviewed the different procedures reported for the radio- chemical separation of scandium and yttrium. In all of these methods, preliminary separation from common impurities is achieved by precipitating scandium and yttrium as fluorides. Brooksbank, Leddicotte and Dean1 used lanthanum as a non-isotopic carrier for scandium.Because of the known differences in the solubilities of the fluorides and oxalates of scandium and rare earths1’ it was decided to carry out the radiochemical separation of scandium in the presence of 5 mg of scandium and 10 mg of lanthanum carriers, the lanthanum being used as a collector for the scandium fluoride. The separation of yttrium-90 was carried out with about 20 mg of yttrium carrier. The final purification of scandium from rare earths was achieved by selective extraction into 100 per cent. tributyl phosphate from 8 M hydrochloric acid. This procedurels was preferred to the thiocyanate extraction procedure described by Kemp and S m a l e ~ , ~ ~ because of the difficulties experienced in destroying large amounts of ammonium thiocyanate.The only difficulty in following the tributyl phosphate extraction method is the need to spin the solution in a centrifuge to obtain clear phase separation. Scandium was back-extracted into 2 M hydrochloric acid, precipitated as hydroxy-quinolinate,20 and dried at 105” C to determine the chemical yield. The purification of yttrium from other rare earths is generally achieved by solvent extraction or ion-exchange separation methods.16 In the present studies, the final purification of yttrium was achieved by the anion-exchange separation with a methanol - nitric acid mixture which has been reported by Desai, Krishnamoorthy Iyer and Sankar Das.21 Yttrium is finally precipitated as oxalate, dried and mounted for P--counting. After the counting the chemical yield is determined by igniting the oxalate to yttrium oxide.MEASUREMENT OF RADIOACTIVITY- The photo-peak activity of scandium-47 was measured with a 14 x 1-inch NaI(T1) crystal coupled to an E.M.I. 9536 photomultiplier and a single channel analyser. After locating the 0.155-MeV photo-peak, the window width was opened so that the entire photo- peak was counted. The /3--activity of yttrium-90 was measured with a conventional mica end-window G.M. set-up. The activity measurements were made through a 184-mg per cm2aluminium absorber to cut off any heavy rare-earth activity that accompanies the yttrium fraction in the ion-exchange procedure.21 METHOD REAGENTS- Scandium carrier solution-Dissolve 0-7661 g of ignited scandium oxide (purity 99-9 per cent.) in 5 ml of concentrated nitric acid and dilute the solution to 100 ml (1 ml = 5.0 mg of scandium).Yttrium carrier solution-Dissolve 1.27 g of Specpure yttrium oxide in dilute nitric acid and make the solution up to 100 ml (1 ml = 10 mg of yttrium). 8-Hydroxyquinoline solution, 5 per cent. w l v 8-hydroxyquinoline in 2 M acetic acid. Ammonium acetate, 2 RE.October, 19661 TITANIUM AND NIOBIUM WITH FAST NEUTRONS 64 1 TributyZ Phosphate-Wash Eastman Kodak analytical-reagent grade tributyl phosphate with M sodium carbonate solution and then twice with distilled water and pre-equilibrate it with 8 M hydrochloric acid before use. Methanol - nitric acid, (i) 2.5 per cent. : 7.0 M-Dilute 2.5 ml of 7.0 M nitric acid to 100 ml with distilled methanol; (ii) 10 per cent. : 1.0 M-Dilute 10 ml of 1.0 M nitric acid to 100 ml with distilled methanol. All other reagents are of recognised analytical purity.PROCEDURE FOR THE ISOLATION OF SCANDIUM FROM IRRADIATED SAMPLES- (a) TITANIUM STANDARDS- (1) Evaporate 1 ml of the scandium carrier in a 25-ml silica crucible and ignite gently. (2) Weigh the polythene tube containing the irradiated standard. Cut open the irradiated polythene tube and transfer the oxide into the silica crucible. Collect the cut pieces of the tube in a tared paper and weigh; the difference in the weights gives the weight of the standard (5 to 10mg of titanium dioxide) used for analysis. (3) Fuse the standard with about 2 g of potassium bisulphate. (4) Dissolve the melt in dilute acid and transfer the solution into a 40-ml lusteroid centrifuge tube, add 10 mg of lanthanum carrier, precipitate the hydroxide with ammonia solution, spin the solution in a centrifuge and discard the supernatant liquid.(5) Dissolve the hydroxides in 1 ml of concentrated nitric acid, dilute the solution to about 10 ml and add 1 ml of 40 per cent. hydrofluoric acid. Warm the tube in a water-bath swirling it occasionally to facilitate the precipitation of scandium fluoride. Cool, spin the solution in a centrifuge and discard the supernatant liquid. (6) Dissolve the precipitate by warming it with 1 ml of saturated boric acid and 1 ml of 8 M nitric acid, precipitate the hydroxides with ammonia solution, spin the solution in a centrifuge and discard the supernatant liquid. (7) Repeat steps (5) and (6) twice.(8) Dissolve the hydroxide in 10 ml of 8 M hydrochloric acid and transfer the solution to a stoppered centrifuge tube. Add 10ml of 100 per cent. tributyl phosphate and extract scandium into the organic phase. Let the mixture stand for a few minutes, then spin it in a centrifuge to separate the phases. (9) With a transfer pipette, remove most of the organic phase into a second centri- fuge tube. (10) Repeat the tributyl phosphate extraction of the aqueous phase and combine the organic layers. (11) Scrub the organic phase once with 10 ml of 8 M hydrochloric acid, spin the solution in a centrifuge and transfer the organic phase into a separating funnel. (12) Back-extract scandium with 2 M hydrochloric acid, collecting the aqueous phase in a second separating funnel.Repeat the back-extraction twice and combine the aqueous phases. (13) Introduce 10 ml of chloroform into the second separating funnel containing the scandium, and extract to remove any dissolved tributyl phosphate. Discard the chloroform. Repeat the chloroform extraction twice. (14) Transfer the aqueous phase into a 150-ml beaker and evaporate the solution nearly to dryness. Destroy any organic matter by evaporating twice with 1 ml of concentrated nitric acid. When the nitric acid is almost completely removed, dilute with distilled water to about 100ml. (15) Add 2 ml of 8-hydroxyquinoline solution followed by 10 ml of ammonium acetate solution. Add ammonium hydroxide drop-wise until the pH is about 8 (test with indicator paper). Bring the solution to the boil.Cool and filter through a de-mountable glass filter unit containing a 2-8-cm Whatman No. 40 filter-paper that has already been washed with water, dried and weighed. Wash the precipitate with warm distilled water, and dry it in air for about 5 minutes.642 ATHAVALE et al.: ACTIVATION ANALYSIS FOR [Anabst, Vol. 91 (16) Transfer the filter disc containing the scandium 8-hydroxyquinolinate to a weighed cover-glass kept in a Petri dish and dry it at 105" C for about 45 minutes. Cool the residue and weigh it. (17) Mount the precipitate on a 1/32-inch aluminium card, cover with Scotch Tape and count the y-activity of scandium-47. (18) Correct the count-rate for chemical yield. (b) GRANITE AND DIABASE SAMPLES- crucible containing the scandium carrier. sulphuric acid.(1) Transfer a known weight (150 to 200 mg) of the irradiated sample into a platinum (2) Evaporate twice with 5ml of 40 per cent. hydrofluoric acid and a few drops of Ignite the residue and fuse with 2 g of potassium bisulphate. (3) Dissolve the melt in dilute acid and transfer it into a 40-ml lusteroid centrifuge tube. (4) Proceed through steps (4) to (13) as for the titanium standard, repeating the fluoride precipitation cycle 3 times. (5) Transfer the aqueous phase into a 100-ml beaker, destroy any organic matter by repeated evaporation with nitric acid and evaporate to dryness. (6) Dissolve the residue in 10 ml of methanol - nitric acid (2.5 per cent. : 7.0 M). (7) Transfer the solution to an anion-exchange column. (30 cm x 8 mm of 50 to 100- mesh Dowex-1 x 8 resin in nitrate form and pre-equilibrated with the same solvent.) (8) Elute with 30ml of the same solvent, collecting the effluent in a 100-ml beaker, evaporate the solution to dryness and precipitate scandium as hydroxy-quinolinate as des- cribed for the titanium standard. (c) STEEL SAMPLES- (1) Dissolve a known weight (200 to 300mg) of the steel sample by warming it in a 50-ml beaker with concentrated hydrochloric acid and adding a few drops of 30 per cent.hydrogen peroxide to hasten the dissolution. Add the scandium carrier to the beaker before sample dissolution. (2) When the metallic pieces have dissolved completely, dilute the solution to about 20 ml and filter the solution, collecting the filtrate in the lusteroid tube. (3) Incinerate the filter-paper containing any black residue in a silica crucible, fuse it with potassium bisulphate, dissolve the fused cake in dilute acid and combine with the filtrate obtained in step (2).(4) Precipitate the hydroxides with ammonia solution and proceed in the manner described for the standard. . PROCEDURE FOR THE ISOLATION OF YTTRIUM FROM IRRADIATED SAMPLES- KIOBIUM STANDARD- (1) With 2 ml of the yttrium carrier (and no lanthanum) isolate the yttrium-90 together with any rare-earth impurities from the standard, following steps (1) to (7) as described for the isolation of scandium from irradiated titanium standard. (2) Dissolve the hydroxide in dilute nitric acid and evaporate the solution to dryness in a 100-ml beaker. (3) Dissolve the residue in 10 ml of methanol - nitric acid (2.5 per cent.: 7.0 M). (4) Transfer the solution to an anion-exchange column. (5) Elute with 30ml of the same solvent and discard the eluate. (6) Continue the elution with 50 ml of methanol - nitric acid (10 per cent. : M) and collect the eluate in a 100-ml beaker. (7) Evaporate the solution to dryness on a water-bath, dissolve the residue in water, dilute to about 20ml and precipitate yttrium as oxalate with 2 ml of saturated oxalic acid solution. (30 cm x 8 mm of 50 to 100- mesh Dowex-1 x 8 resin in nitrate form and pre-equilibrated with the same solvent.)October, 19661 TITANIUM AND NIOBIUM WITH FAST NEUTRONS 643 Filter the solution through a de-mountable glass filtration unit containing a 2.8-cm disc of Whatman No. 42 filter-paper, wash the residue with water and finally with acetone.(9) Remove the paper containing the yttrium oxalate, dry it under an infrared lamp, mount it on a 1/32-inch thick aluminium card with Scotch Tape and count the P-activity. (10) When counting is over, remove the precipitate (with the minimum of the Scotch Tape) into a tared crucible, ignite it to the oxide and weigh it as yttrium oxide. Correct the observed count-rate for chemical yield. STEEL SAMPLES- fluoride by the procedure described for isolating scandium from irradiated steels. yttrium isolated from the irradiated niobium standard. (8) Bring the solution to the boil to granulate the oxalate. (1) With 2 ml of the yttrium carrier instead of the scandium carrier, isolate the yttrium (2) Purify yttrium-90 by following the steps (2) to (10) described for purifying the RESULTS Two types of samples were analysed by this method.One type consisted of analysed samples of titanium and niobium stabilised steels, which are in wide use as structural com- ponents in reactor technology. Although chemical colorimetric methods have been reported for determining these elements, it was desirable to have an independent check on the analysis procedure. The “standard” rocks G-1 and W-1 are also included in this work to evaluate the general applicability of the method. The results obtained are given in Tables I1 and 111. TABLE I1 ANALYSIS OF SAMPLES FOR TITANIUM BY THE 47Ti (n,p) 47Sc REACTION Certified value, Present results, Sample per cent. per cent. Stablised steel-1 . . 0*50* 0.56; 0.49; 0-50; 0.48 0.53 t 0.51; 0.48; 0.53 Mean 0.51 f 0.03 (lo) Mean 0.37 & 0.01 (lo) Mean 0-170 & 0.012 (lo) Mean 0.62 BCS/235/1 .. * . 0.36 0.36; 0-35; 0.38; 0.37 Granite G-1 .. .. 0.156: 0.158; 0,177; 0.164; 0.183 Diabase W-l , . .. 0*64$ 0.62; 0.62 * Chemical colorimetric.2a $ Calculated from the percentage of titanium oxide values, Fleischer and Stevens (Ref. 23, see p. 626). BCS = British Chemical Standards. By 48Ti (n,p) 46Sc.10 TABLE rrr ANALYSIS OF SAMPLES FOR NIOBIUM BY THE 93Nb (n,or) 9OY REACTION Certified value, Present results, Sample per cent. per cent. Stabilised steel-2 . . 0-76* 0.85; 0-70; 0 . S O ; 0.82; 0.86; 0.76; 0.67; 0.84 Mean 0.79 & 0.07 ( l u ) Mean 0-84 f 0.04 (la) BCS/246 .. .. 0.82 0.88; 0.85; 0.79; 0-S4; * Chemical colorirnetri~.~~ BCS = British Chemical Standards.The results given in Tables I1 and I11 indicate satisfactory precision and accuracy for the methods. The values obtained for titanium in G-1 and W-1 compared well with the preferred values reported by Fleischer and Stevens.23 The only other results reported for titanium by the activation-analysis technique are 0.133 per cent. for G-1, and 0-54 per cent. for W-1 (as obtained by Kim and Meinke6), which are lower in comparison with our results.TABLE IV INTERFERENCE FROM NEUTRON-INDUCED REACTIONS FissirJn flux = 3 x 10l1 n per cm2 per second; thermal flux = 5 x loll n per cm2 per second; thickness of cadmium foil = 30 mils Reactions studied (n,r) interference Conflicting threshold reaction A t-- mP - 7 - 1 7 ------ L--- Abun- Abun- Abun- F,* dance, 0 dance, Cadmium ii,, dance, mb percent.Reaction b per cent. ratio Reaction R, mb percent. Rl - - 46Ti (n,p) 46Sc 8.6 7.99 46SC-46SC 24 100 34 47Ti (n,p) 47% 20.4 7.32 46Ca-47Ca--%7S~ 4-7d 0.25 0.0033 ND 60Ti(n,o()47Ca ;$47Sc 0,047 5.25 48Ti (n,p) 48Sc 0.28 73.99 - - - - soV (n,ol) 47Sc 2*1* 0.25 "Nb @,a) 0.077 100 8 S y - s O y 1-2 100 26 51V (n,cc) 48% 0.027 99-75 ?Zr (n,p) QoY 0*9P 51-46 ND: 47Sc activity not detected. * Estimated cross ~ecti0.rr.l~ Ratio of activities R, to R, (for equal weight of parent element) (1 : 300) (6 : 1) (1 : 8) b ? .. r-- bOctober, 19661 TITANIUM AND NIOBIUM WITH FAST NEUTRONS 645 RADIOCHEMICAL PURITY- The radiochemical purity of the scandium-47 separated from different samples and standard was checked by following the decay of the 0-155-MeV photo-peak for about 7 days.The observed half-life was between 80 and 82 hours. (Half-life reported14 for scandium-47 = 81.6 hours.) Scandium-46 and scandium-48 are also produced with scandium-47 during the irradiation. For an irradiation time of 40 hours their relative intensities are in the ratio 1 to 10 to 44. Scandium-46 and scandium-48 do not emit any low energy y-rays and therefore do not interfere directly in the measurement of the scandium-47 activity. The interactions of the high energy y-rays from these nuclides, however, enhance the measured photo-peak activity of scandium-47. The extent of this interference was not more than 2.5 per cent. of the photo-peak activity of scandium-47 in the first few days of measurement.The purity of yttrium-90 was checked by following the p--decay for about 1 week. The half-life obtained was between 65 and 66 hours. (Half-life of yttrium-90 reported14 = 64.8 hours.) NUCLEAR INTERFERENCE- One of the reasons for not using the threshold reactions in activation analysis is because often the same active nuclide is produced at a higher rate from an (n,y) reaction. The neutron-induced reactions that interfere are given in Table IV. It may be seen that scandium, if present in the samples, interferes in the analysis of titanium with the 46Ti (n,p) 46Sc reaction. By shielding the samples with cadmium foil, the activity from the (n,y) reaction could be reduced by a factor of 34. However, it should be pointed out that this factor depends on the relative intensities of the thermal and epithermal fluxes in the neutron spectrum and, therefore, will vary with the irradiation position even in the same reactor.Neither 47Ti (n,p) 47Sc nor 48Ti (n,p) 48Sc has any interference from a primary (n,y) reaction. However, during the irradiation of samples, scandium-47 may be produced from a number of other reactions that are listed in Table IV. Of these, the (n,a) reaction on titanium-50 causes no interference in comparative activation analysis. Formation of scandium-47 from calcium-46 is small because of the low abundance of this isotope, and the low cross section for the (n,y) reaction. It is further reduced in cadmium-shielded irradiations. The calculated yield from the (n,a) reaction on vanadium-50 is 300 times lower than that resulting from an equal weight of titanium, as indicated in the last column of Table IV.The yield is calculated on the basis of an estimated cross section of 2-1 mb for this particular reaction.l3 The reaction 48Ti (n,p) 48Sc has a positive interference from 51V (n,a) 48Sc, the yield from the latter being one-eighth of that obtained from the former, as indicated in Table IV. Therefore, both from the point of view of sensitivity and freedom from nuclear inter- ference, 47Ti (n,p) 47Sc is to be preferred to the other (n,p) reactions on this element. Determination of niobium in samples containing yttrium or zirconium will be in error. The induced activity from SgY (n,y) 99Y could be reduced by a factor of 26 by shielding the samples with cadmium foil. From equal weights of zirconium and niobium in the sample, the yield of yttrium-90 from the (n,p) reaction on zirconium-90 is estimated to be 6 times greater than that resulting from the (n,a) reaction on niobium.It should be possible to correct for the zirconium interference as follows. A specimen of zirconium is irradiated together with the sample, and the ratio of the activities of yttrium-90 to yttrium-92 (formed by the reaction g2Zr (n,p) g2Y ; &, estimated = 0-16 mbI3) determined, The activity of yttrium-92 from the sample is also measured, and by using the ratio of activities from the zirconium specimen, the activity of yttrium-90 corresponding to the zirconium content of the sample is calculated. The total yttrium-90 activity from the sample can therefore be corrected for the contribution from zirconium. ’- 3-4 hour CONCLUSION The method described shows the advantages in using the reaction 47Ti (n,p) 47Sc over the other (n,p) reactions on titanium for the activation analysis of this element.The limit of sensitivities for these reactions, defined as that weight of the element which, on irradiation646 ATHAVALE, DESAI, GANGADHARA4N, PENDHARKAR AND SANKAR DAS for 40 hours at a flux of 3 x loll n per cm2 per second, would give a measured activity equal to the background (the measurements having been made 24 hours after irradiation) can be calculated to be 1.7 pg of titanium and 54 pg of niobium. In addition, our studies indicate the possibilities of using fast-neutron activation where the thermal neutron-activation technique is associated with definite practical limitations.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. REFERENCES Brooksbanb, W. A., Leddicotte, G. W., and Dean, J. A., Analyt. Chem., 1958, 30, 1785. Gruverman, I. J., and Henninger, W. A., Ibid., 1962, 34, 1680. Kim, C. K., and Meinke, W. W., Talanta, 1963, 10, 83. Brownlee, J. L., jun., U.S. Atomic Energy Commission Report, T.1.D.-6311, 1960, 145. Kim, C. K., “Production and Use of Short-lived Radioisotopes from Reactors,” I.A.E.A., 1963, VOl. 11, 73. Kim, C. K., and Meinke, W. W., Analyt. Chem., 1963, 35, 2135. Leddicotte et al., U.S. Atomic Energy Commission Report, ORNL-2866, 1960, 26. Atkinson, R. H., Steigman, J., and Hiskey, C. F., Analyt. Chem., 1952, 24, 477. Sankar Das, M., Venkateswarlu, Ch., and Athavale, V. T., Analyst, 1956, 81, 239. Sankar Das, M., Ph.D. Thesis, University of Bombay, 1964. -, Paper presented a t the Study Group Meeting of the International Atomic Energy Agency Yule, H. P., Lukens, H. R., and Guinn, V. P., Nzccl. Instrum. Meth., 1965, 33, 277. Roy, J. C., and Hawton, J. J., Atomic Energy of Canada Ltd., Report AECL-1181, 1960. Strominger, D., Hollander, J. M., and Seaborg, G. T., Rev. Mod. Phys., 1958, 30, 585. Mellish, C. E., Nztcleonics, 1961, 19, (3), 114. Stevenson, P. C., and Nervik, W. E., “The Radiochemistry of the Rare Earths, Scandium, Yttrium and Actinium,” Nuclear Science Series : NA4S-NS-3021, National Academy of Sciences- National Research Council, Washington, D.C., 1961. on the Utilisation of Research Reactors, held in Bombay, 1964. Feibush, A. M., Rowley, K., and Gordon, L., Analyt. Chem., 1958, 30, 1605. Ishimori, T., Watanabe, K., and Nakamura, E., Bull. Chem. Soc. Japan, 1960, 33, 640. Kemp, D. M., and Smales, A. A., Analytica Chim. Acta, 1960, 23, 410. Pokras, L., and Bernays, P. M., Analyt. Chem., 1951, 23, 757. Desai, H. B., Krishnamoorthy Iyer, R., and Sankar Das, M., Talanta, 1964, 11, 1249. Athavale, V. T., Nadkarni, M. N., and Venkateswarlu, Ch., Analytica Chinz. Actu, 1960, 23, 440. Fleischer, M., and Stevens, R. E., Geochim. Cosmochim. Acta, 1962, 26, 525. Madhava Menon, V. P., Mahadevan, N., Srinivasulu, K., and Venkateswarlu, Ch., Scient. Ind. Res., Received May 24th, 1965 1962, 21B, 20.

ISSN:0003-2654    

DOI:10.1039/AN9669100638

出版商:RSC    

年代:1966

数据来源: RSC

摘要:

Analyst, October, 1966, Vol. 91, &5. 647-651 647 An Automatic, Modified Formaldoxime Method for Determining Low Concentrations of Manganese in Water Containing Iron BY A. HENRIKSEN" (Norwegian Institute for Water Research, Oslo 3, hrovwuy) An automatic method for determining manganese in fresh water with the AutoA4nalyzer is described, together with details of the analytical system. The method is based on a modification of the formaldoxime procedure pub- lished by Goto, Komatsu and Furukawa,l which consists in reacting man- ganese with formaldoxime in alkaline solution, and decomposing any iron formaldoxime formed with EDTA and hydroxylammonium chloride. Aluminium, zinc, copper, iron, calcium, magnesium, chloride and phos- phate do not interfere a t concentrations above those normally found in natural water in Norway.The capacity of the method is 25 samples per hour, and the standard deviation is 5 p g of manganese per litre. A statistical comparison of the manual persulphate oxidation method with the automatic formaldoxime method indicated that both methods give identical results. THE manganese content of fresh water is usually determined colorimetrically by oxidation with persulphate or periodate to form permanganate ions2 Compared with most colorimetric methods, the sensitivity is poor. As the concentration of manganese in natural water in Norway is normally low, a more sensitive method was needed that could be easily adapted for use with the Technicon AutoAnalyzer. Formaldoxime, which forms an orange - red colour with manganese in alkaline solutions, has been used for determining manganese in plant material by Sideris3y4 and Bradfield,5 and in water by Goto, Komatsu and Furukawa,l and Morgan and Stumm.6 The sensitivity of the formaldoxime method is estimated to be about five times that of the permanganate m e t h ~ d .~ A disadvantage of the formaldoxime method is that iron interferes by forming a violet - red coloured complex with the reagent. Sideris* recommended precipitation of iron as iron(II1) phosphate. Bradfields observed that precipitation of iron, calcium and mag- nesium in alkaline solutions was prevented by the addition of N-hydroxyethylethylene diaminetriacetic acid (HEEDTA), and the coloured iron complex was decomposed by heating the sample at 65" C for 2 hours.These observations were confirmed by Morgan and Stumm.6 Goto, Komatsu and Furukawal showed that the formaldoxime complex of iron was rapidly decomposed by adding EDTA and hydroxylammonium chloride, while the manganese formaldoxime was stable under these conditions. The procedure described by Goto et al. seemed to be the simplest for routine determina- tions of manganese on the AutoAnalyzer. EXPERIMENTAL Analytical-reagent grade chemicals were used. A Technicon AutoAnalyzer with a phototube colorimeter was used for the colorimetric determinations. All procedures described in the subsequent text were converted to automatic methods. The flow scheme of these is essentially the same as that shown in Fig. 1, except that one or more reagents were omitted. All values found are given in micrograms of manganese per litre, and are obtained by reading against a calibration graph obtained with manganese standards.INTERFERENCE FROM IRON- The concentration of iron in natural water in Norway usually varies up to a maximum The effect of this concentration range of iron in the following three of 1 mg per litre. formaldoxime procedures was investigated- * Present address : The Institute of Paper Chemistry, Appleton, Wisconsin, U.S.A.648 HENRIKSEN : AN AUTOMATIC, MODIFIED FORMALDOXIME METHOD [Analyst, Vol. 91 Sampler I I %0° Y 0. I00 A i r Sample Formaldoxime Iron (I I ) sulphate Ammonia Re-i ntrod uction A i r EDTA H y d roxy I am i n e Fig. 1. Flow diagram of apparatus used in an automatic method for determining manganese in water (i) Addition of formaldoxime and ammonia only.(ii) Addition of formaldoxime and ammonia, followed by addition of EDTA and hydroxylammonium chloride according to the procedure given by Goto, Komatsu and Furukawa. (iii) Addition of iron( 11) sulphate, formaldoxime and ammonia, followed by addition of EDTA and hydroxylammonium chloride. The results obtained are given in Table I. The interference from iron is high if no EDTA and hydroxylammonium chloride are added. It is considerably reduced if these reagents are added, but a significant interference still occurs at concentrations of iron below 1 mg per litre. These observations indicated that the interference from iron could be further reduced by adding a sufficient amount of iron as a reagent in the second procedure.The last column in Table I shows the results obtained when a solution of iron(I1) sulphate was added as a reagent, making the sample 2 p.p.m. with respect to iron. It is seen that the interference is then reduced to an insignificant level. TABLE I EFFECT OF IRON IN THE FORMALDOXIME PROCEDURE FOR DETERMINING MANGANESE Concentration, p g of iron(I1) per litre 60 120 250 500 1000 2500 Without hydroxyl- amine and EDTA, pg of manganese per litre 20 60 190 360 7 00 1600 Method of Goto et al., pg of manganese per litre 20 45 55 35 20 ( 5 Method of Goto et al. + 2 mg of iron(I1) per litre, p g of manganese per litre < 5 < 5 < 5 ( 5 (5 <5October, 19661 FOR DETERMINING MANGANESE IN WATER CONTAINING IRON 649 A manganese standard of 100 pg per litre was analysed in the presence of different concentrations of iron.The results, given in Table 11, indicate that iron does not influence the determination of manganese. TABLE I1 EFFECT OF IRON IN THE PRESENCE OF MANGANESE 1OOpg of manganese per litre Concentration, pg of iron(I1) per litre 0 50 100 250 500 1000 Without iron added, pg of manganese per litre 100 120 142 150 130 115 With 2 mg of iron per litre added, p g of manganese per litre 100 102 100 104 99 100 The rather surprising effect of the additional iron has not been further investigated. However, from the experimental evidence given in Table I, it is reasonable to assume that the EDTA and hydroxylammonium chloride do not decompose the iron - formaldoxime complex when this is present below a certain concentration.The beneficial effect of the additional iron may then be to raise the concentration of the complex to a level at which decomposition will occur. INTERFERENCE FROM COLOURED SUBSTANCES IN WATER- The coloured substances in water are potential interferences in the formaldoxime method. The standard manual method for determining them consists in measuring the absorption of the water a t 420mp in a filter photometer. However, a significant absorption is still observed at 480mp. Thirty samples from different localities in Norway were analysed for manganese with the manifold shown in Fig. 1. The samples were passed through the apparatus for a second time, replacing the iron(I1) sulphate and formaldoxime solutions with distilled water. The absorption arising from the colour of the samples measured under these experi- mental conditions was correlated to the absorption measured manually at 420 mp. The mean value of the “apparent” manganese content due to the coloured substances was found to be 35 pg of manganese per litre.The mean value of the colour was 37” H. However, no correlation was found between these two sets of results. Consequently, coloured water samples that are low in manganese content should be run through the apparatus for a second time, with the iron(I1) sulphate and formaldoxime solutions replaced by distilled water. The “apparent” manganese content caused by the natural colour of the water samples must be deducted to give the net concentration of manganese. EFFECT OF pH ON COLOUR FORMATION- Morgan and Stumm have shown that the intensity of colour of manganese formaldoxime depends upon the pH; it reaches its maximum at a pH of 9.2.This observation has been confirmed, and in the recommended procedure given above a sufficient amount of ammonia solution is added to give a pH of 9.0 to 9.4. EFFECT OF OTHER IONS- Morgan and Stumm also point out that large amounts of iron and alkaline earths may be precipitated in alkaline solutions (as phosphates, hydroxides and carbonates). The effect of these and other ions in concentrations above those normally found in natural water in our country has been studied. Solutions containing 100 pg of manganese per litre, and different concentrations of other ions were prepared according to Table 111. Calcium and magnesium (up to 100 mg per litre) alone, and in the presence of 3 p.p.m.of phosphate, do not have any effect on the determination of manganese, neither do concen- trations of iron up to 5 p.p.m. alone, or in the presence of 3 p.p.m. of phosphate. Phosphate, however, interferes at high concentrations in the presence of calcium. Nickel and cobalt also interfere. When equimolar concentrations of cobalt or nickel and manganese are present the interference is 7.5 and 3 per cent., respectively. If the sample is heated to 60” C during650 [Analyst, Vol. 91 the decomposition with EDTA and hydroxylammonium chloride, the interference from cobalt and nickel is reduced to 1 and 3 per cent., respectively. No interference was observed from aluminium, zinc, copper and chloride at the con- HENRIKSEN : AN AUTOMATIC, MODIFIED FORMALDOXIME METHOD centrations given in Table 111.TABLE I11 EFFECT OF VARIOUS IONS ON THE COLOUR FORMATION OF 100 pg of manganese per litre Ion Aluminium . . .. Zinc . . .. .. Chloride . . . . Copper . . .. Copper . . . . Copper . . .. Magnesium . . .. Magnesium . . .. Iron(I1) . . .. Iron(I1) . . . . Iron(I1) . . .. Iron(I1) . . .. Iron(I1) . . .. Iron(II1) . . .. Phosphate . . . . Phosphate . . . . Calcium . . .. Calcium . . . . Cobalt . . * . Nickel . . Phosphate and iron(I1) Phosphate and calcium Phosphate and calcium Phosphate and magnesium .. .. . . .. .. . I . . . . .. . . .. .. .. .. .. . . . . . . . . .. . . . . .. . . . . . . . . .. . . .. . . . . . . .. .. . . . . . . .. .. . . .. . . . . .. .. .. Concentration, mg per litre 4 2000 300 0.04 0.4 1-6 50 100 50 100 0.05 0.1 0.5 1.0 5 5 3 60 1 1 3 f 5 3 + 80 60 + 80 3 + 80 MANGANESE FORMALDOXIME Manganese found, p g per litre 100 103 100 100 104 102 102 102 98 100 100 102 98 100 102 104 100 98 130 175 100 98 20 102 COMPARISON OF THE MANUAL AND AUTOMATIC METHODS- The standard deviation of the automatic formaldoxime method was found to be 5 p g of manganese per litre in the range 20 to 500 pg of manganese per litre, determined from 20 samples. In order to compare this method with the manual persulphate oxidation method formerly used in our laboratory, 22 samples were analysed by both methods. The results were statistically compared by calculating their F-values and t-values. These were 1-16 and 0.126.Comparison of the t-value with a statistical table for t-distribution indicates a 90 per cent.probability that differences in the mean values are caused by random errors. REAGENTS- Iron(l1) sulphate-Dissolve 140 mg of ammonium iron(I1) sulphate, FeS04(NH4),S0,. 6H,O, in water containing 1 ml of concentrated sulphuric acid and dilute the solution to 1 litre. This solution contains 20 mg of iron per litre. Formaldoxime-Dissolve 20 g of hydroxylammonium chloride in 500 ml of distilled water. Add 10ml of formaldehyde solution (37 per cent. w/w). Ammonia solution (3 + 1)-Dilute 3 parts of concentrated ammonia solution with 1 part of water. EDTA, 0.1 M-Dissolve 37.2 g of ethylenediaminetetra-acetic acid disodium salt in distilled water and dilute the solution to 1 litre. Hydroxylammoni$m chloride-Dissolve 100 g of hydroxylammonium chloride in distilled water and dilute the solution to 1 litre. STANDARD SOLUTIONS- Dissolve 143.8 mg of potassium permanganate in 50 ml of water. Add 2 ml of concen- trated sulphuric acid.Add a 10 per cent. solution of sodium bisulphite drop-wise until the colour of the permanganate disappears. Boil the solution to expel excess of sulphur dioxide. Cool, and dilute to 1 litre. This solution contains 50 mg of manganese(I1) per litre. Dilute the solution to appropriate standards.October, 19661 FOR DETERMINING MANGANESE IN WATER CONTAINING IRON 651 PROCEDURE- The method finally adopted for determining manganese on the AutoAnalyzer is shown in Fig. 1. The sample is segmented with air and successively mixed with solutions of iron(I1) sulphate, formaldoxime and ammonia.After passing the solution through a quarter-length, time-delay coil, which allows for the colour to develop, the air bubbles are removed through an extended T-junction, the stream is re-introduced and again segmented with air. Solutions of EDTA and hydroxylamine are then added to destroy any iron formaldoxime formed. After passing the solution through a double-coil heating bath (25” C) and a half-length, time-delay coil, the colour is measured at 480 mp. The sampling rate is 25 per hour, lt minutes are allowed for both washing and sampling. The range expander is used at x 4 and x 10 expansions, giving an operative range of 0 to 0.8 mg of manganese per litre. A linear relationship between optical density and manganese concentration is obtained in this range. The shape of the AutoAnalyzer traces is nearly symmetrical. REFERENCES 1. 2. “Standard Methods for the Examination of Water, Sewage and Industrial Wastes,” Eleventh 3. 4. __ , Ibid., 1940, 12, 307. 5. 6. Goto, K., Komatsu, T., and Furukawa, T., Analytica Chim. Acta, 1962, 27, 331. Edition, American Public Health Association, Inc., New York, 1960, p. 155. Sideris, C. P., I d . Engng Chem. AnaZyt. Edn, 1937, 9, 445. Bradfield, E. G., Analyst, 1957, 82, 254. Morgan, J. J., and Stumm, W., J . Amer. Wat. Wks Ass., 1965, 57, 107. Received November 26th, 1965

ISSN:0003-2654    

DOI:10.1039/AN9669100647

出版商:RSC    

年代:1966

数据来源: RSC