ISSN:0003-2654    

DOI:10.1039/AN96287FX029

出版商:RSC    

年代:1962

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96287BX031

出版商:RSC    

年代:1962

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN96287FP189

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

August, 1 %i2] THE ANALYSTOPPORTUNITY FOR AM E T A L L U R G I S Tinterested in SalesApplicants are invited to write in con-fidence giving full details of their careerand qualifications to the Group PersonnelAdviser, Hilger & Watts Ltd, 98 St.Pancras Way, London, N.W. 1 .WHIFFEN & SONS LIMITEDEXPERIENCEDANALYSTs v\vitCRESOL RED (Acid Range)THY MOL BLUE (Acid Range)BROMOCRESOL GREENMETHYL ORANGEBROMOPHENOL BLUEH E F F E R ' SOF CAMBRIDGEare always glad to buySCIENTIFICJOURNALSespecially complete sets andruns ofTHE ANALYSTalso scientific and technicallibraries, early books, bookson the history of science,etc.W. HEFFER & SONSLimitedPetty Cury, CambridgeEnglandNEUTRAL REDBROMOTHYMOL BLUEPHENOL REDBRILLIANT YELLOWTURMERICCRESOL RED (Alk.Range)ANAPHTHOLPHTHALEINTHYMOL BLUE (Alk. Range)PHENOLPHTHALEINTHY MOLPHTHALEINCLAYTON (Titan) YELLOWcOLoup pH PAYEE &- AND COLOUR C H M E L~~Y 2.6(Y) - 1.8(Pk) - 1.2(R)Y 3.6(Y) - 2.8(Pk) -1.qV)0YB 4.0(8) - 3.qPale Y) 3.qY)3.6(Y) - 4.0(6n) - 5.00)4.0(Y) - 3.6(0) - 2.B(R)R 4.4(R) - 3.6(6~) - 2&B)R 4.6(R) - 5.qPk) - 6.O(Y)Y 5.4(Y) - 6.O(Pk) - 6.6(R)Y 5.6(Y) - 6.4(6y-B) - 7.qP)BrY 6,0(Y)- 7,O(En) - 7 . 6 W6.qY) - 6,NPale Y) - 7.2(P)R 6 , q R ) - 6.8(0)-7.6(Y)YY6.6(Y) - 7.2(Pk) - 7.8(R)6.6(Y) - 7.4(Y-O) - 8.4(R)Y 6.8(Y)- 7.8(0) -8.NR-Br)C 7 3 C ) B.O(Pale 8) - 8.7(8)C 8.4(C) - 8.8(Pale Pk) - 9.6(R)C 10.5(C) - ll.O(Pale B) - ll.L?(B)Y 7.2(Y) ~ 8.O(P) ~ 8.MV)Y 7.8(Y) -- 8.8(6~) 9.4(R)Y 12.0(Y) - 12.5(Pk)- 13.O(R)LITMUS BOOKS RED, BLUE OR NEUTRAL (Bibulous and NonXIibulous)KEY TO COLOURSB Blue Br Brown C Colourless Gn Greon Gy Grey0 Orange P Purple Pk Pink R Red V Violet Y -YellowCOMPARATOR BOOKSNo.1035 for pH 1.0 to 3.5Six colour matches printed inside the cover of eachbook indicate the p H value in steps of 0.5 pH.No. 3651 for pH 3.6 to 5.1No. 5267 for pH 5.2 to 6.7No. 6883 for pH 6.8 to 8.3No. 8410 lor pH 8.4 to 10.0Six colour matches printed inside cover of eachbook with figures showing the pH value in stepsof 0.3 pH.UNIVERSAL BOOKS pH 1.0 to 11.0Eleven colour matches are printed in steps of 1.0 pHinside the covers of each book together with figuresshowing the pH value that each colour represents.OTHER TEST BOOKS AND PAPERSCALEDON YELLOW BOOKSCOBALT CHLORIDE BOOKS (Heat Sensitive)LEAD ACETATE BOOKSLEAD ACETATE PAPERS, 2" x 1" to Gas Referees' SpecificationLEAD ACETATE BOOKS, 2' x 1 ' ,, ,,MERCURIC CHLORIDE PAPERS, 2" x 2"POLE FINDER BOOKS, LARGE, 4 t ' x 3 i ', 24 leaves,Perforated 5 Strips,I I* ,, SMALL, as above but only 8 leavesI , I, 1.STANDARD SIZESTARCH AND POTASS. IODIDE BOOKSSTARCH BOOKSHENDON . LONDON . N.W.4 August, 19621 THE ANALYST xviiTHE S0CIlYI"F FOR ANALYTICAL CHEMISTRYFORMERLY THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTSFOCNDED 1874. IXCORPORXTED 1907.THE objects of the Society are to encourage, assist and extend the knowledge and study ofanalytical chemistry and of all questions relating to the analysis, nature and compositionof natural and manufactured materials by promoting lectures, demonstrations, discussionsand conferences and by publishing journals, reports and books.The Society includes members of the following classes :-(a) Ordinarj- Members whoare persons of not less than 21 years of age and who are or have been engaged in analytical,consulting or professional chemistry ; ( b ) Junior Members who are persons between the agesof 18 and 27 years and who are or have been engaged in analytical, consulting or professionalchemistry or bona fcde full-time or part-time students of chemistry.Each candidate forelection must be proposed by three Ordinary Members of the Society. If the Council intheir discretion think fit, such sponsorship may be dispensed with in the case of a candidatenot residing in the United Kingdom.Every application is placed before the Council andthe Council have the power in their absolute discretion to elect candidates or to suspend orreject any application. Subject to the approval of Council, any Junior Member above theage of 21 may become an Ordinary Member if he so wishes. A member ceases to be a JuniorMember on the 31st day of December in the year in which he attains the age of 27 years.Junior Members may attend all meetings, but are not entitled to vote.The Entrance Fee for Ordinary Members is L1 1s. and the Annual Subscription is L3 3s.Junior Members are not required to pay an Entrance Fee and their Annual Subscription is;51 1s.No Entrance Fee is payable by a Junior Member on transferring to Ordinary Nember-ship. The Entrance Fee (where applicable) and first year's Subscription must accompanythe completed Form of L4pplication for Membership. Subscriptions are due on January 1stof each year.Scientific Meetings of the Society are usually held in October, Nolrember, December,February, April and Ma\-, in London, but from time to time meetings are arranged in otherparts of the country.All members of the Society have the privilege of using the Library of The ChemicalSociety. Full details about this facility can be obtained from the Librarian, The ChemicalSociety, Burlington House, Piccadilly, London, W. 1.The Analyst, the official organ of the Society, which has a world-wide distribution, isissued monthly to all Ordinary and Junior Members, and contains original papers and notes,information about analvtical methods, Government reports, reviews of books and reportsof the proceedings of the Society.In addition, all Ordinary Members receil-e ,Jizal_vticaZAbstracts, providing a reliable index to the analytical literature of the world.Forms of application for membership of the Society may be obtained from the Secretary,The Society for Analytical Chemistry, 14 Belgrave Square, London, S.W. 1.Notices of all meetings are sent to members by post.LOCAL SECTIONS AND SUBJECT GROUPSTHE North of England, Scottish, Western and Midlands Sections were formed to promote theaims and interests of the Society among the members in those areas. The Microchemistry,Physical Methods and Biological Methods Groups have been formed within the Society tofurther the study of the application of microchemical, physical and biological methods ofanalvsis. All members of the Society are eligible for membership of the Groups.The Sections and Groups hold their own meetings from time to time in different places.There is no extra subscription for membership of a Section or Group. Application forregistration as a member should be made to the SecretaryNEW A N D FORTHCOMING TITLES:ADVANCES INI FLUORINE CHEMISTRY-3I MACROMOLECULAR CHEMISTRYSTANDARD METHODS FOR THEANALYSIS OF OILS, FATS & SOAPSB U T T E R W O R T H S4 & 5 Bell Yard, London, W.C.2

ISSN:0003-2654    

DOI:10.1039/AN96287BP201

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

AUGUST, 1962 THE ANALYST Vol. 87, No. 1037 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY NORTH OF ENGLAND AKD SCOTTISH SECTIONS A JOINT Meeting of the North of England and Scottish Sections was held on Thursday and Friday, June 28th and 29th, 1962, in the Department of Chemistry, Queen’s University, Belfast. The meeting took the form of a Symposium on “Instrument Techniques in Industry and Research.” Delegates were welcomed by Professor H. B. Henbest, Ph.D., D.Sc., A.R.C.S., D.I.C., F.R.I.C. There were three sessions of scientific papers, and Thursday afternoon was utilised for conducted tours of the Chemistry Department of Queen’s University and of the Department of Industrial and Forensic Science, Ministry of Commerce of the Government of Northern Ireland. On Thursday morning the Chair was taken by Professor C.L. Wilson, Ph.D,, D.Sc., F.R.I.C., F.I.C.I., and the following papers were presented and discussed: “Instrumental Methods of Continuous Analysis,” by G. Jessop, M.Sc., Ph.D. ; “Applications of Vapour-phase Infra-red Spectroscopy to the Functional Group Analysis of Propoxy and Butoxy Compounds by modified Zeisel Reactions,” by D. M. W. Anderson, B.Sc., Ph.D.; “Applications of Modern Techniques in Spectroscopy,” by R. A. C. Isbell, A.1nst.P. On Friday morning the Chair was taken by the Chairman of the Scottish Section, Mr. A. F. Williams, B,Sc., F.R.I.C., and the following papers were presented and discussed: “Investi- gations on the Determination of Noble Metals Using Oscillographic Polarography,” by I. Beattie and R.J. Magee, M.Sc., Ph.D., F.R.I.C., F.I.C.I. ; “Analytical Applications of the Flame Emission Spectra of Lead and Titanium,” by C. L. Chakrabarti, M.Sc., A.R.I.C., R. J. Magee, M.Sc., Ph.D., F.R.I.C., F.I.C.I., and Professor C. L. Wilson, Ph.D., D.Sc., F.R.I.C., F.I.C.I. ; “An Ultramicrospectrophotometric Method for the Determination of Complex Cyanides,” by I?. Haba and Professor C. L. Wilson, Ph.D., D.Sc., F.R.I.C., F.I.C.I. ; “Differential Cathode-ray Polarography,” by H. M. Davis, B.Sc., A.Inst.P., A.R.I.C. On Friday afternoon the Chair was taken by Dr. R. J. Magee, M.Sc., Ph.D., F.R.I.C., F.I.C.I., and the following papers were presented and discussed: “The Activation Analysis of High-purity Beryllium Using Penetrating Radiations,” by C. A. Baker; “Some Analytical Applications of Mass Spectrometry,” by A.Quayle, M.Sc., A.R.I.C. ; “A Modular Gas Chromato- graph System for the Analysis of Exit Streams from Reactors and for the Application Work Required for Process Analysers,” by C. W. Munday, B.Sc., A.R.I.C., and G. R. Primavesi, B.A. ; “The Methylene Insertion Reaction for the Identification of Hydrocarbons Using Gas Chromatography,” by E. S. Lane, B.Sc., Ph.D., F.R.I.C. This session ended with a General Discussion on “Quantitative Gas - Liquid Chromato- graphy in the Routine and Research Laboratories,” initiated by A. F. Williams, B.Sc., F.R.I.C., to which many delegates contributed. Many delegates enjoyed a ’bus tour of the City of Belfast on the Thursday evening and attended the Symposium dinner on the Friday evening, when Professor Wilson took the Chair. 609610 PROCEEDINGS [vol. 87 MICROCHEMISTRY GROUP THE thirty-fifth London Discussion Meeting of the Group was held at 6.30 p.m. on Wednesday, June 27th, 1962, at “The Feathers”, Tudor Street, London, E.C.4. The Chair was taken by the Chairman of the Group, Mr. C. Whalley, B.Sc., F.R.I.C. A discussion on “Ion Exchange in Microchemistry” was opened by J. Pilot, B.Sc., and J. A. R. Genge, M.Sc. BIOLOGICAL METHODS GROUP THE Summer Meeting of the Group was held on Thursday, June 7th, 1962, and took the form of a visit to the Laboratories of Smith, Kline & French Ltd., Welwyn Garden City, Herts. The production area, the pharmacological laboratories, the chemical laboratories and the animal rooms were all open to the visitors. Thirty-five members and friends attended and their thanks to the Company were proffered by Mr. W. A. Broom, B.Sc., F.R.I.C., Vice-chairman of the Group.

ISSN:0003-2654    

DOI:10.1039/AN9628700609

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

August, 19621 COCKBILL 61 I The Determination of Tantalum and Niobium A Review* BY M. H. COCKBILL (Research Department, London G. Scandinavian Metallurgical Co. Ltd., 39 Hill Road, London, S . W. 19) SUMMARY OF CONTENTS Introduction First approaches to the problem Methods depending on hydrolysis Gravimetric methods Chromatographic methods Ion-exchange methods Chlorination methods Liquid - liquid extraction methods Methods depending on reduction of niobium Colorimetric methods Spectrographic methods X-ray fluorescence methods Radioactivation methods Miscellaneous methods Conclusions THE separate determination of tantalum and niobium has been a difficult problem for many years. Tantalum and niobium never occur in the free state and always occur together in nature. They replace each other isomorphously in their minerals and are themselves some- times isomorphously replaced by elements of similar atomic radius, such as tin, titanium, antimony and bismuth.The best known minerals are the columbites and tantalites, which are an isomorphous series of niobates and tantalates of iron and manganese. The pre- dominating element gives the term by which the mineral is known, e.g., a mineral high in niobium is a columbite ; in stibio-columbites and stibio-tantalites, part of the tantalum and niobium has been isomorphously replaced by antimony. Tantalum and niobium have been found widely distributed in small amounts in cassiterites, in which tin has been isomorphously replaced, and also occur in association with other metals, such as yttrium, uranium, calcium and rare earths.Pyrochlore, now of some importance, is a complex niobate of calcium and sodium, with various amounts of many other elements; microlite is the tantalum equiva- lent of pyrochlore. Tantalum also occurs in some of the rarer minerals, such as fergusonite (a tantalate and niobate of yttrium, erbium and cerium) and samarskite, which also contains uranium. The usual methods of opening up these ores for analytical purposes are by fusion in alkaline peroxide or carbonate, by fusion in potassium hydrogen sulphate or by digestion with hydrofluoric acid. Leaching the melts or diluting the solutions with water then produces insoluble hydrated oxides of tantalum, niobium, titanium, etc. The analyst has then to contend with the extremely complex chemical nature of the solutions and precipitates so formed.Oxides and hydroxides of these elements co-precipitated in this way have been shown by thermal, X-ray and chemical analysis1 to form compounds of the type R(NbO,)., where R is tantalum, titanium, a rare-earth metal, etc. Because of the indefinite nature of the solutions and precipitates, the separation, and even identification, of mixed tantalum, niobium and titanium compounds is difficult. The techniques available in the past were so uncertain that the compounds of these three elements behaved almost as chemically identical substances when mixed, although not when finally separated. This phenomenon, which Crookes2 termed “loss of individuality,” is a result of the tendency of these elements to form occlusion complexes when precipitated in the presence of each other.The analysis of minerals and alloys for tantalum and niobium falls into two parts: the separation of tantalum and niobium together from all other substances, and their separation from each other. The latter separation poses a more difficult problem than any met with in the determination of most other elements. * Reprints of this paper will be available shortly. For details, please see p. 688.[Vol. 87 The earliest workers used fusion in potassium hydrogen sulphate to form soluble com- pounds of tantalum and niobium, and these procedures are still used. The original sample of mineral is either fused immediately in the potassium hydrogen sulphate or is first digested with hydrofluoric acid. This treatment with hydrofluoric acid serves for preliminary isolation of the tantalum and niobium.The digested sample is filtered, and the filtrate, which contains the tantalum and niobium, is evaporated to dl-yness to give a solid residue, which is then fused in potassium hydrogen sulphate. The second procedure effects an immediate and almost complete separation of the rare earths, alkaline earths and silica, but not all minerals are amenable to this treatment. The cooled melt from the fusion is extracted with dilute acid and sulphurous acid, and the tendency of tantalum and niobium compounds to be hydrolysed is utilised to separate these elements from other constituents ; the products of hydrolysis, however, are difficult to filter and strongly occlude impurities.This method has been applied to the determination of the combined tantalum and niobium contents of steels3 and can be carried out because the amount of precipitate produced by the hydrolysis is small. The first attempt at the separate determination of tantalum and niobium that met with any success was the classical work of Marigna~,~ whose method is completely summarised in “AnaZyse der MetaZZe.”5 No significant progress was made after the publication of Marignac’s results until Schoeller summarised the faulty methods previously used and laid down a basis for systematic work.6 Schoeller stated that “. . . A molecularly admixed constituent or complex precipitate cannot be extracted quantitatively by a process of selective solution . . . . it may be concluded that if we avoid the formation and manipulation of complex precipitates and adhere to the principle that all separation methods must proceed from a common solution of the constituents to be separated, the degree of accuracy thus obtained should be sub- stantially the same as that of the standard analytical methods for the commoner elements.” In Marignac’s method, the “common solution’’ to which Schoeller re€erred was a solution of the elements together with potassium fluoride in hydrofluoric acid.The method was based on the difference between the solubilities in hydrofluoric acid of potassium tantalum fluoride, K,TaF,, and potassium niobium oxyfluoride, K,NbOF,. The tantalum compound was the least soluble, and the two could be separated by repeated fractional crystallisation.It was possible to follow the progress of the separation by microscopic examination of the crystals, the flat tablets of the niobium oxyfhoride being easily distinguished from the long needles of the tantalum salt. A small specimen of the crystals could be removed on a platinum spatula and examined at a magnification of ~ 1 0 . Tantalum separated as clean rhombic needles of K,TaF,, but the presence of niobium altered this shape to that of toothed blades joined together in bunches. The niobium compound appeared later as square plates, usually with one incomplete corner. Any titanium that might be present could not be separated in this way, as the solubility of the corresponding fluoro salt was intermediate between that of the tantalum and the niobium compouiids, but nearer to that of the latter.Tantalum and niobium could not be completely separated by this method. Each oxide as a final product contained a small amount of the other, but, by rigidly standardising the procedure, these errors could be made to cancel each other out. Schoeller and his collaborators investigated the problem over a period of 17 years from 1919 onwards (see The Analyst for this period). They stressed the importance of tannin as an analytical reagent and applied it to the separation of tantalum from niobium. The formation of complexes with oxalic or tartaric acid was used to stabilise solutions of the metals, usually after fusion in potassium hydrogen sulphate. The use of such organic acids or, more generally, their ammonium salts, gave the “common solution of the constituents to be separated” and thereby avoided formation of the complex precipitates produced by hydrolysis that had previously defied attempts at quantitative resolution.Schoeller rejected the unsatisfactory method of hydrolysis previously used for the final separation of tantalum and niobium from impurities. He showed that, if a potassium hydro- gen sulphate melt was extracted into tartaric acid, the tantalum and niobium were held in solution, whereas other metals could be removed, e.g., by precipitation with hydrogen sulphide ; the tantalum and niobium salts were then hydrolysed by boiling with hydrochloric acid. The precipitation was not complete nor was the precipitate pure, but, after filtration and treatment of the filtrate to remove iron, the small amount of tantalum and niobium remaining in solution could be precipitated by tannin.‘Titanium was also precipitated almost completely 612 COCKBILL: THE ‘DETERMINATION OF FIRST APPROACHES TO THE PROBLEMAugust, 19621 TANTALUM AND NIOBIUM. A REVIEW 613 by the tannin, and tungsten was precipitated during the hydrolysis. Traces of other metals, such as zirconium, also appeared in the final combined precipitate. Tungsten could be removed by fusing the precipitates from the hydrolysis in potassium carbonate and then precipitating tantalum, niobium and titanium from an aqueous solution of the melt with a “magnesia mixture” defined by Schoeller as consisting of “1 g of magnesia sulphate, 2 g of ammonium chloride, 25 ml of water and 4 drops of ammonia.” The tungsten remained in solution, and titanium could be removed by utilising the solubility of its yellow complex with salicylic acid. The oxides of the three metals were fused in potassium hydrogen sulphate, and the melt was extracted with an oxalate solution.Sodium salicylate was added, and the oxalate was removed by adding calcium chloride solution. Tantalum and niobium were co-precipitated with the calcium oxalate, and the titanium remained in solution, These precipitations were never complete, and residues had to be recovered from various filtrates by treatment with tannin and repetition of the processes. However, the procedures were a great improvement on those previously used. The separation of tantalum from niobium depended on the different behaviour of their oxalo- and tartaro-complexes when dilute tannin solution was added.Tantalum needed a larger concentration of the complexing acid to keep it in solution, so that, when such a solution with tannin present was slowly diluted, the tantalum - tannin complex was precipitated first. The tantalum complex was lemon-yellow and the niobium complex orange-red, so that the progress of the separation could be followed by noting the colour of the precipitate. A clear-cut separation could not be accomplished in a single operation, as appreciable co-precipitation occurred, but three repetitions of precipitation, ignition, fusion and leaching before the final precipitation generally gave a sufficiently complete separation. However, the number of fractional precipitations needed depended on the ratio of tantalum to niobium present and also to a great extent on the experience of the analyst.Because, under similar conditions, tannin formed insoluble complexes with several other metals, such as titanium, zirconium and tungsten, it was necessary first to separate the tantalum and niobium from all other constituents of the original sample that could interfere with the tannin separation. Despite the discoveries of Schoeller and his collaborators, the complete analysis of a mineral was still highly complicated. The various separations were not well defined; filtrates and residues needed repeated treatment and several precipitates and solutions had to be handled at the same time or kept in reserve for various periods, so that cumulative personal errors and various indeterminate losses were usually unavoidable.In the last 30 years, numerous other methods for the separate determination of tantalum and niobium have been investigated, and the last 10 years in particular have been very fruitful, particularly since more powerful analytical tools have become available and more especially because of the increased importance of tadalum and niobium. The container material for uranium in an atomic reactor should have good thermal conductivity and a high melting-point, its thermal expansion should be similar to that of uranium, and its vapour pressure should be low a t high temperatures. Niobium compares well with other possible metals in these respects and is preferable to tantalum because of its much lower neutron-absorption cross- section.Kiobium for nuclear use should therefore contain as little tantalum as possible, and so a method of determining small amounts of tantalum in niobium was needed. Very low-grade ores and soils have been considered as possible sources of niobium, and this has led to a need for rapid field methods for detecting and determining very small amounts of the metal. Steels are now analysed on a routine basis for tantalum and niobium separately, and in the large amount of research on the industrial separation of these metals, rapid analytical methods must be used for following and assessing the new processes. This interest in the determination of tantalum and niobium has led to the development of many and varied methods; their determination no longer involves the difficult procedure it once did, and they can now be determined almost as simply as can many more commonplace elements.METHODS DEPENDING ON HYDROLYSIS Acid hydrolysis is a common method of separating tantalum and niobium from other metals, but not from each other. Attempts have nevertheless been made since Schoeller’s time to use acid hydrolysis for separating the two metals. Jaboulay’ made use of the dif- ferences in chemical behaviour between hydrogen peroxide and a solution of tantalum and niobium in sulphuric acid. Under these conditions niobium formed a stable perniobic acid,614 COCKBILL : THE DETERMINATION OF Pol. 87 but tantalum did not form an analogous compound. When such a solution of both metals was boiled, only tantalum salts were hydrolysed and tantalum was precipitated as the flocculent “tantalic acid’’ or hydrated oxide.Niobium was recovered from the filtrate by prolonged boiling with sodium sulphite. This method has been tested by many workers, but it has never been possible to attain complete separation of tantalum from niobium, Methods similar to those of previous workers were used for the preliminary separation of tantalum and niobium from other metals. If the hydrolysis is carried out by Jaboulay’s method, not all of the niobium is precipi- tated, and the last traces can be brought down by adding mercurous nitrate solution. This principle has been applied in works analysis for the recovery of niobium after separation by Marignac’s method (personal communication from Mr.G. M. Holmes), but results for niobium were sometimes high because of occlusion of potassium salts. This method is still applied to alloys when the final niobium pentoxide precipitate may contain titanium, which must be determined colorimetrically. Hydrochloric acid can be used instead of sulphuric acid in Jaboulay’s method, and the same method of separation can be applied with improved recoveries,s but the separation procedure must be repeated a t least once. Bagshawe and Elwell have discussed the efficiency of the hydrolysis method3 and have given details for determining niobium (and tantalum) in steels. A detailed discussion of the isolation of pure earth acids based on many years’ experience of a number of workers is available in two reports of the Chemists’ Committee of the Gesellschaft Deutscher Metall- hiitten- und Bergleute e.V.g GRAVIMETRIC METHODS Apart from the double-fluoride and tannin fractional precipitations, other gravimetric methods of separating tantalum from niobium have been proposed from time to time. The hydrolysis of acid solutions of the metals in the presence of hydrogen peroxide has already been mentioned.Hexamethylenetetramine has been used in determining tantalum or niobium, or both, in high chromium - nickel steels containing tungsten, vanadium and molybdenum and for separating niobium from titanium.10 Alimarin and Burovall have used sodium hypophosphite for precipitating tantalum from solution. The mixed oxides were fused in potassium hydrogen sulphate, and the melt was extracted with ammonium oxalate solution ; when sodium hypophosphite solution was added, the tantalum was precipitated as (Ta,O,)H(PO,H,). Hypophosphorous acid formed complexes with tantalum and niobium that were insoluble in tartaric or citric acid, but in oxalic acid only the tantalum compound was insoluble.These workers claimed that the reaction was specific for tantalum and would detect 1 p.p.m. of the metal, and also that no co-precipitation of niobium occurred. The precipitate was voluminous and granular,l2 which permitted easy washing and filtration, but the calcined precipitate was contaminated with phosphorous that could only be removed by further precipitation of the tantalum from potassium hydrogen sulphate - oxalate solution by tannin or cupferron.Tantalum has been precipitated from sulphuric acid - hydrogen peroxide solution by phenylarsonic acid,l3*l4 but the separation was not strictly quantitative and the final product was contaminated with arsenic ; re-precipitation was needed, as in the hypophosphite method. The use of n-propylarsonic acid has been advocated,15 but not more titanium than tantalum could be present and tannin had to be added if the amount of tantalum to be determined was less than 10mg. Zirconium was also precipitated by these reagents and therefore interfered in the determination of tantalum. The use of phenylarsonic acid as a precipitant for tantalum has been studied again recently. Majumdar and Mukherjee16 described its use for precipitating tantalum from oxalate solution and found that, when disodium ethylene- diaminetetra-acetate was used as masking agent, tantalum could be separated from numerous ions, zirconium being a notable exception.The precipitation of both tantalum and niobium by phenylarsonic acid is currently used in determining minor concentrations (as little as 0.001 per cent.) of these metals in steel; zirconium is used as carrier for the very small amount of earth acids to be precipitated, and, after ignition, the tantalum and niobium in the zirconium oxide are determined ~olorimetrically.~7 Contamination by arsenic is too slight to interfere and does not stain a platinum dish when one is used for the ignition. For determining tantalum in uranium - tantalum alloys, a turbidimetric procedure has been described in which phenylarsonic acid is used to precipitate the tantalum from sulphuric acid solution.18August, 19621 TANTALUM AND NIOBIUM.A REVIEW 615 Ferroin (o-phenanthroline - ferrous sulphate complex) was found to form precipitates with tantalum and niobium in hydrofluoric acid solution1g and was used to purify so-called “pure” compounds of these elements. The method was similgr to the tannin procedure in that complete separation could only be effected by fractional crystallisation. Morin and quercetin form red precipitates with niobium and orange precipitates with tantalum in strongly acid solution,20 and certain other flavone derivatives21 have been recom- mended as specific precipitants for tantalum in solution. A solution of morin or quercetin in methanol was added to the acidified solution of tantalum and niobium in tartaric or oxalic acid, and preliminary separation of the metals could be achieved in this way; zirconium was not precipitated under these conditions.The morin - niobium complex could be extracted into acetone, in which the tantalum complex was not soluble.22 Later reports have shown that attempts to separate tantalum from niobium with quercetin have not been successful.= Moshier and Schwarberg2* proposed the use of N-benzoyl-N-phenylhydroxylamine (tantalon) as a specific precipitant for tantalum from a solution of tantalum and niobium in hydrofluoric and sulphuric acids at pH 1. After the mixture had stood in a bath of cold water for 2+ hours, the precipitate was separated by filtration and then ignited to tantalum pentoxide. Titanium and zirconium did not interfere ; niobium interfered only when present to an extent greater than 20 per cent.of the tantalum, when repeated precipitations were necessary. Tungsten and molybdenum interfered, but could be removed by prior separation with cupferron at pH 3 to 4, when both these elements remained in solution. Majumdar and M~kherjee~~ described the use of the same reagent at pH 3.5 to 6.5 to precipitate niobium from tartaric acid solution ; after separation of the precipitate by filtration, tantalum could be precipitated from the filtrate by decreasing the pH. Disodium ethylenediaminetetra- acetate was used to keep most other metals in solution. The usual method of determining the tantalum and niobium was to ignite the precipitated organic complexes to oxides, but these workers later found26 that the niobium complex had a definite composition and could be weighed directly.A procedure in which tantalon was used to separate and determine niobium, tantalum and titanium has been described.27 Majumdar and Ray Chowdhury investigated the possibility of using cupferron to separate tantalum from niobium,28 but found it to be less effective than tantalon. Russian workers have described a means of determining niobium in the presence of tungsten with c~pferron.~~ Some research has been reported on the separation of tantalum, niobium and titanium in hydrochloric acid solution.30 The metals were precipitated separately from acid am- monium chloride solution by saturation with hydrogen chloride. Selenous acid31 p32 and a ~ r i d i n e ~ ~ have been suggested as precipitants for tantalum and niobium, but complete separation does not appear to be possible with either reagent.Malissa3* reported investigations of the formation of dithiocarbamate derivatives of niobium. He found that both sodium diethyldithiocarbamate and sodium pyrrolidine dithiocarbamate reacted with niobium, but not with tantalum, over the pH range 2 to 6; he showed that microgram amounts of niobium could be satisfactorily determined and that quantitative separation of tantalum from niobium was possible. 8-Hydroxyquinoline precipitates niobium from ammonium oxalate ~ o l u t i o n ~ ~ leaving titanium, tantalum, tin and antimony in the filtrate, from which tantalum can be recovered and determined by the tannin method.A small amount of tantalum finds its way into the niobium precipitate and vice versa, but the purity of the final oxides is better than 97-5 per cent. The separation of tantalum from niobium based on the preferential absorption of tantalum on to freshly precipitated ferric hydroxide has been de~cribed.~~ The disadvantage of gravimetric methods is that preliminary separation of the tantalum and niobium from other constituents of the sample is nearly always necessary. Some satis- factory gravimetric methods have been evolved specifically for alloys, but these are not generally suitable for minerals. CHROMATOGRAPHIC METHODS Some of the compounds of tantalum and niobium differ in their properties sufficiently to permit complete separation, but only by time-consuming fractionation procedures.Such methods have been made practicable for routine analysis by the introduction of ion exchange and chromatography. Chromatography involves the partition of compounds of tantalum and niobium between aqueous and organic solvents by adsorption of an aqueous solution616 COCKBILL THE 1)ETERMINATION OF LV-01. 87 of the compounds on an inert solid of large surface area and preferential removal of one and then the other with suitable solvents. This procetlure is equivalent to a large number of liquid - liquid extraction stages. Similarly, ion exchange involves the adsorption of the tantalum and niobium on a resin, and use is made of the difference in the equilibrium constants of the adsorption reactions between the two metals and the resin.Burstall, Swain, Williams and Wood3’ showed that it was possible to separate tantalum and niobium oxides from other metals by a chromatographic method,38~3g~40 and this was later applied to the separation of tantalum from niobium. A solution of the sample in hydrofluoric acid, without any previous separation process, was adsorbed on a small amount of cellulose and transferred to the top of a column of cellulose held in a polythene tube. (Ammonium fluoride added to the original fluoride solution suppressed any transfer of titanium in the subsequent elution operations.*l) The tantalum was first washed from the column with 250 ml of ethyl methyl ketone saturated with water. The column was then “conditioned” with 400 ml of ethyl methyl ketone containing 1 per cent.by volume of 40 per cent. w/v hydrofluoric acid. This procedure decreased the water content of the column and permitted niobium to be eluted by 500 ml of a 12.5 per cent. solution of hydrofluoric acid in ethyl methyl ketone. The great advantage of this method was that, in one operation, tantalum and niobium were each removed free from all the other metals that might be present, even the titanium remaining on the column. Only tungsten, molybdenum and potassium interfered, as they did not behave in a regular manner when subjected to chromatography under these condi- tions; a proportion of them usually appeared. in the niobium fraction, and a correction had to be made for their presence. This method can be used for determining tantalum and niobium in a wide range of materials, the only variations being in the solution of the sample or in the treatment of the final products. These variations depend on whether or not the sample contains large amounts of silica or phosphate, e.g., with low-grade phosphatic material (such as Sukulu soil), for which large initial samples must be taken, the final niobium fraction might contain phosphate ions.In this instance, precipitation by ammonia solution is used in the final determination of niobium. The “middle” fraction, i.e., the eluate from the intermediate “conditioning” stage, may occasionally contain a little niobium ; this happens if too large a sample is taken or if alkali metals are passed through the columns. Usually a milligram or so of iron is also found.The eluted metals are contained in polythene beakers, each as a solution in a large volume of ethyl methyl ketone also containing hydrofluoric acid. A time-consuming evapora- tion procedure under infrared lamps is used to concentrate the solutions to small volume, after which they are either ignited (with suitable precautions) in platinum dishes and then weighed as oxides or, when necessary, they are subjected to further chemical treatment, e,g., 10 ml of diluted sulphuric acid (1 + 1) are added to the small diluted volume of the niobitim fraction, and the mixture is heated to fumes to remove fluoride, diluted and boiled with ammonia solution. This treatment precipitates the niobium free from phosphate, molybdenum and potassium, but not entirely free from tungsten.However, a simple colorimetric correction for tungsten can be made. A more rapid chromatographic method for determining mixed tantalum and niobium in minerals and ores was later developed by Mercer and The column of cellulose used by these workers was only 3 inches long. A solution of the sample in hydrofluoric acid and containing ammonium fluoride was adsorbed on cellulose and transferred to the top crf the column, from which all the tantalum and niobium were eluted by 400 ml of ethyl methyl ketone containing 15 per cent. by volume of 40 per cent. w/v hydrofluoric acid. Again, tungsten was the only metal to interfere after the final hydrolysis. The disadvantage of these chromatographic methods lies in the preparation and handling of the reagents. The ethyl methyl ketone must be pure, and for each determination on the long column nearly 1200ml of eluate must be evaporated, due regard being paid to its inflammable, corrosive and toxic nature.It is essential to use the powdered cellulose specified, vix., Whatman’s standard grade for the long column, and a mixture of Whatman’s standard and coarse grades for the short column. A good deal of trouble and distrust of the methods has been caused by the use of impure ethyl methyl ketone or unsuitable types of powdered cellulose. The amount of eluate to be evaporated can be decreased if the “middle” fraction is ignored. By using both long and short columns concurrently, the time needed for a separate determination of tantalum and niobium can be greatly shortened. The tantalum can be eluted from the long column at the same time as the combined metals are eluted fromAugust, 19621 TANTALUM AND NIOBIUM.A REVIEW 617 the short one. Pu’iobium is determined by difference, and a check on the result for the niobium can in due course be obtained from the long column. Despite the length of time taken for a determination (1 to 2 days), the need to purify large amounts of ethyl methyl ketone and the handling of the hazardous eluates, this method can conveniently be used for routine laboratory determinations. A battery of columns, long and short in pairs, can be operated continuously, each pair being one stage behind the next, and an unbroken stream of deter- minations can be made fairly rapidly. The procedure must be strictly adhered to, but is no more tedious than the preparation of solutions for colorimetric work; it is chemically just as accurate, except for very small amounts.The colorimetric method devised by Hunt and Wells43 especially for analysing pure chromatographically produced mixed oxides (from the short column) can be used, and will decrease the time required for each determination. Specially purified ethyl methyl ketone that needs no treatment before use can be purchased. The partition chromatography of tantalum and niobium with ethyl methyl ketone has been used for their separation on a strip of upward diffusion in hydrofluoric acid being used for the determination of niobium in low-grade samples. The niobium was detected as a yellow band after the strip had been exposed to ammonia vapour and then sprayed with an aqueous solution of tannic acid and was determined by direct visual comparison of the band with those produced on standard strips.The method used to dissolve the sample depends on the material being analysed. Minerals may be directly dissolved in hydrofluoric acid, as for pyrochlore, or taken up after fusion in potassium hydrogen sulphate, as is necessary for refractory samples, e.g., ilmeno-rutile. Some samples of steel will dissolve in hydrofluoric acid with no troulle, but others need a mixture of hydrofluoric and nitric acids for their solution; if nitric acid is used, all nitrate must be subsequently destroyed, otherwise un- desirable oxidation of molybdenum occurs. The determination consists in placing spots of the fluoride solution of the sample on a strip of filter-paper and allowing a mixed solvent (ethyl methyl ketone, hydrofluoric acid and water) to diffuse up the strip.The spot of sample is near the bottom of the strip, but above the surface of the solvent. By cutting slots in a sheet of filter-paper, a number of such determinations can be carried out simultaneously; such specially cut paper strips are available (Whatman CRL/l). The movement of niobium from the test spot depends on the acidity of the solvent; when the acidity of the solvent is high, niobium is extracted in a narrow band in the solvent front. Thus, if high concentrations of acid are used, trace amounts of niobium can be detected, but the greater the amount of niobium present the more difficult becomes its determination by visual comparison with a standard.By decreasing the acidity of the solvent, the niobium is spread over a large area and visual comparison is easier. A suitable solvent for up to 20 pg of niobium pentoxide contains 2 per cent. v/v of water and 12 per cent. v/v of hydrofluoric acid (40 per cent. w/v). Titanium, zirconium, antimony and vana- dium move only slightly from the spot, but molybdenum, unless reduced, advances far enough to overlap the niobium; the trouble can be overcome by adding metallic iron. Although tantalum moves with the niobium, the pale yellow of its complex with tannin is difficult to see and renders its interference negligible. Anions do not interfere in the method. By adjusting the ratios of acid, ethyl methyl ketone and water in the solvent, tantalum can be separated from the niobium and subsequently detected with q~inalizarin~~ or phenyl- fluorone. For small amounts of tantalum and niobium, visual comparison of the developed strips can be replaced by a colorimetric finish.Tantalum and niobium can be separated from all interfering elements by using an eluting solvent consisting of 86 ml of ethyl methyl ketone, 12 ml of 40 per cent. hydrofluoric acid and 2 ml of water. The thiocyanate complex of niobium is soluble in acetone - water mixture and can be dissolved from the paper strip. If the ten strips on a piece of special paper include sufficient samples to permit the location of the earth acids by the development of one strip with tannin, the appropriate sections of the remaining strips can be cut off and placed in an aqueous solution of stannous chloride, ammonium thiocyanate and acetone.After the solution has been shaken and set aside for 15 minutes, its optical density can be measured with a spectrophotometer or an absorptiometer, e.g., a Spekker instrument, fitted with a mercury lamp and an Ilford No. 601 filter (personal communication from Mr. G. M. Holmes). A calibration graph is prepared by subjecting a series of pure niobium solutions containing from 1 to 10 pg of niobium pentoxide per 0.01 ml to the whole of the procedure described above and plotting the logarithm of the optical density against the weight of niobium present. The graph is not linear for concentrations of niobium greater than this range. In this method, the tantalum is collected with the niobium,618 COCKBILL THE DETERMINATION OF [Vol.87 but does not interfere with its determination.; in order to determine tantalum it must be completely separated from niobium. Tantalum and niobium are best separated from each other when there is a minimum of free hydrofluoric acid in the sample strip. To achieve this condition, the strips are prepared as for niobium, but are then allowed to dry for at least 2 hours over a saturated solution of potassium carbonate; they are then treated as in the determination of niobium. The tanta- lum, which advances in a narrow band in front of the wet area and about 0.5 cm in front of the niobium, is detected by spraying the strip with quinalizarin solution and exposing the sprayed strip to ammonia vapour.The section of the strip containing the tantalum is removed, and the tantalum is determined with phenylfluorone as described by Luke.46 In Luke's original method, the tantalum was isolated by liquid - liquid extraction, but use of the paper strip avoids this troublesome step, and only the colorimetric part of Luke's method is used. Other attempts a t the chromatographic separation of tantalum from niobium have been made. After fusion of the mixed oxides in potassium hydrogen sulphate, the melt was dissolved in ammonium oxalate - sulphuric acid solution. The pH of this solution determined whether tantalum or niobium would be adsorbed on the alumina, and, by suitable adjustment of pH, all the niobium and part of the tantalum were retained on the column when the solution was first passed through.Solutions of slightly different compositions were then used to elute tantalum and then niobium separately. Recoveries of the order of 98 per cent. were claimed, but no mention was made of the application of the method to anything other than pure mixed oxides. Different pH values had to be determined for different types of alumina, and the solutions used were extremely dilute. Bruninx, Eeckhout and G i l l i ~ ~ ~ studied the separation of tantalum from niobium on the micro scale by paper chromatography with several organic solvents. Good separation was achieved with a mixture of ethyl methyl ketone and hydrochloric acid, but the tantalum could not be separated from titanium ; these workers have also used electrophore- sis for this ~eparation.~~ Other chromatographic and electrophoretic methods have been described (see, for example, the paper by BBasius and Czekay5*), including one in which a two-phase eluting solvent of isobutyl methyl ketone and hydrofluoric acid was ~ s e d .~ 1 One method describes the use of a column of activated charcoal saturated with phenylarsonic acid in which tantalum is completely retained on the column. Although the niobium could easily be removed and determined, tantalum could not be recovered with greater than 60 per cent. An attempt to use a paper strip eluted with hydrochloric acid and ethyl methyl ketone has been made with limited success.53 Tikhomiroff4' used a column of activated alumina. ION-EXCHANGE METHODS Much investigation of ion exchange for the separate determination of tantalum and niobium has been carried out, particularly in America.Kraus and Moore54 showed that tantalum and niobium in hydrochloric - hydrofluoric acid solution could be separated by passing the solution through a column of Ilowex-1 anion-exchange resin. Tantalum and niobium radioactive tracers were used in a mixture of 9 M hydrochloric acid and 0.05 M hydrofluoric acid, and the elution was carried out with this solution. This resin has also been used for separating niobium from zirconium in oxalic and in hydrochloric - hydro- fluoric acid solution.56 It has also been shown that, like zirconium and hafnium, tantalum and niobium form complex chloro-ions in concentrated hydrochloric acid and that these complexes can be separated from each other on ion-exchange resins, such as D o ~ e x - 2 .5 ~ ~ ~ Gillis, Hoste, Cornand and Speecke used Dcbwex-2 in separating tantalum from niobium in oxalic acid solution.59 From a mixture of equal parts of tantalum and niobium, these workers, in one operation, recovered 95 per cent. of the niobium free from tantalum and 85 to 90 per cent. of the tantalum contaminated with 1 per cent. of niobium. The separation of small amounts of one metal present in the other has been carried out in hydrochloric-oxalic acid solution.60a61 Hague, Brown and BrightG2 studied the separation of tantalum, tungsten, molybdenum and niobium by anion exchange; they used a mixture of hydrochloric and hydrofluoric acids as eluting solution. Cabell and Milnere3 have also used anion exchange for separating tanta- lum from niobium.A solution of the two elements in 3 M hydrochloric - 0.1 M hydrofluoric acid mixture was put on a column of De-Acidite FF resin, and niobium was eluted by thisAugust, 19623 TANTALUM AND NIOBIUM. A REVIEW 619 solvent, the tantalum being recovered by elution with a solution containing ammonium fluoride and chloride. The last traces of tantalum could be removed from nearly pure niobium pentoxide by passing it, in hydrofluoric acid solution, through an anion-exchange column.64 A method was developed by Hague and M a ~ h l a n ~ ~ for the separation of tungsten, molybdenum, tantalum and niobium present in heat-resisting alloys. A solution of the elements in mixed hydrofluoric - hydrochloric acid was washed from a column of Dowex-1 or De- Acidite FF resin by a series of solvents.The tungsten (and titanium if present) was removed by a hydrochloric - hydrofluoric acid mixture and the molybdenum by hydrofluoric acid containing ammonium chloride. This latter solvent was neutralised to a pH of 4 to 6 and then used to elute first niobium and then tantalum. When all four metals were present, the elution took more than 9 hours, which is much longer than is needed for the chromato- graphic method, even although this entails the careful evaporation of large volumes of organic solvent. I t is clear that, as with the chromatographic method, this separation depends on the difference in behaviour of tantalum and niobium fluoride compounds in the presence of ammonium ions. Kraus and Nelson66 have summarised the elution behaviour of some twenty elements in hydrofluoric - hydrochloric acid mixtures 0-5 to 1.0 N in the former acid.Other methods have been described in which potassium bisulphate solution,67 oxalic acid solution68 and complexing agents69 are used. CHLORINATION METHODS The fact that tantalum and niobium form volatile chlorides has been used as a means of separating them from other constituents of a sample, and a procedure was proposed by Schafer70 for their separate determination. The mixed oxides, with tin and titanium as impurities, were chlorinated by heating with carbon tetrachloride in a closed tube a t 280" C. Tin and titanium chlorides were removed by distillation in V ~ C Z Z O a t 100" C; the penta- chlorides of tantalum and niobium were distilled in vacuo at 200" C and then weighed.After this operation, the pentachlorides were converted to oxides and re-weighed; the ratio of tantalum to niobium could be calculated from the change in weight on conversion from chloride to oxide. Although there is some doubt as to the true boiling-points of tantalum and niobium chlorides, both are in the range 240" to 250" C, which is sufficiently different from the boiling-points of the chlorides of those metals usually found with them to permit separation by fractional distillation of chlorides. The use of carbon tetrachloride as chlorina- ting agent necessitated a sealed high-pressure chlorination apparatus. I t proved difficult to open the bomb in which the reaction took place and then to transfer the contents to a weighing apparatus quantitatively and without hydrolysis of the reactive chlorides ; Schafer described a glass apparatus in which this might be accomplished.Octachloropropane has been used as a chlorinating agent at atmospheric pressure.71972 Prepared mixtures of the oxides of tantalum and niobium and related metals have been chlorinated in this way, but the method was not applicable to minerals unless great care was taken in the preparation of the mixed oxides for chlorination. Again, special apparatus had to be used. In general, chlorination methods for the separate determination of tantalum and niobium are less easily applied than other methods and are not applicable unless pure mixed oxides are prepared by other means, which may include preliminary chlorination and distillation, as described by Schafer.The ratio of tantalum to niobium must be such that the change in weight from chloride to oxide may be accurately determined; on 1 g of chlorides, the presence of 1 per cent. of one as impurity in the other will cause a difference from theory of about 1 mg in the weight of oxide. LIQUID - LIQUID EXTRACTION METHODS It is well known that tantalum, alone or with niobium, can be extracted from hydro- fluoric acid solutions into isobutyl methyl ketone. The degree of separation of tantalum from niobium depends on the concentration of hydrofluoric acid and on the presence of one or more other mineral acids. Work by the U.S. Bureau of Mines on the commercial separation of tantalum and niobium by these methods was described in detail by Werning, Higbie and their c o - w ~ r k e r s .~ ~ ~ ~ ~ 975 The effect of nitric and hydrochloric acids on the extraction of tantalum and niobium from hydrofluoric acid solution has been inve~tigated,~~ as has that of sulphuric a ~ i d . 7 ~ In620 COCKBILL: THE DETERMINATION OF [Vol. 87 the analysis of uranium alloys of tantalum and niobium, Milner and V C ’ O O ~ ~ ~ found that sulphuric acid assisted the extraction of tantalum into ethyl methyl ketone, and Milner and Edwards,79 in separating zirconium from tantalum and niobium, found that tantalum and niobium could be extracted almost completely into isobutyl methyl ketone from fluoride solutions. Leddicotte and Mooreso have shown that niobium almost free from tantalum can be extracted from concentrated hydrochloric acid into a solution of methyldioctylamine in xylene.The niobium was re-extracted from the organic phase into sulphuric, nitric or dilute hydrochloric acid. The use of tribenzylamine dissolved in chloroform or methylene chloride has also been investigateds1; this reagent was found to extract only niobium. Moore showed that niobium was extracted into di-jsobutyl carbinol from hydrofluoric - sulphuric acid solution82 and could thereby be separated from protactinium ; during its separation from zirconium,83 niobium has also been extracted into solutions of butyl phosphoric esters in di-n-butyl ether. Various other solvents investigated for the liquid - liquid extraction separation of tantalum and niobium from other elements or from each other include di-isopropyl ketone,83y8g,85 tributylphosphate,86,87 cyclohexanones8 and various solvents for extracting cupferrate~.~~ Some analytical procedures have been described in which tantalum is isolated by liquid - liquid extraction and then deter:mined, e.g., colorimetrically, when free from interfering elements.46 7 991 METHODS DEPENDING ON REDUCTION OF NIOBIUM Cunninghamg2 used cupferron to separate tantalum and niobium from other metals in the analysis of chromium - nickel and low-titanium steels by a modification of Schoeller’s tannin method, the niobium being finally determined by means of a Jones’ reductor. The addition of titanium dioxide was recommended to assist complete and uniform reduction ; the niobium was then re-oxidised with potassium permanganate.The determination of niobium by reducing the niobium in the mixed oxides obtained by preliminary separation from other metals has been investigated over many years and has not appeared to give satisfactory results. The difficulty has been to standardise a procedure that can then be applied with confidence. ‘The well known tendency of niobium salts to be hydrolysed to a colloidal hydrated oxide and the fact that the reduction never seems to reach the tervalent stage prevented the same degree of reduction from being attained in repli- cate experiments. The usual method was to titrate the reduced solution with a standard solution of potassium permanganate to correct for the reagent blank value, to subtract a figure for the titanium calculated from the amount found by a previous colorimetric deter- mination and then to use a conversion factor obtained by titrating a series of niobium solutions of known concentration that had been subjected to the selected pro~edure.~3 An exact procedure was described by Knowles and L ~ n d e l l , ~ ~ who gave instructions for the method of solution, the preparation of the reductor, the temperature to be used and the time to be allowed for reduction.Oka and Miyamotog5 have reported that a black precipitate was produced when a solution of pure tantalum was placed in contact with zinc amalgam. This precipitate consumed oxidant and was presumably a reduction product of tantalum. Oxidation by ferric sulphate changed the black precipitate into a gelatinous white one, but the presence of titanium prevented its formation altogether.The method finally recommended by these workers for determining niobium ensured that its reduction was effected a t about 60” C, which hastened the procedure; many previous workers had not completed the reduction because it was lengthy at room temperature. Despite all the work carried out on it, however, this method has never been really reliable, but it has had the advantage of being the most rapid method available for determining niobium in the presence of tantalum, zirconium and titanium. More recently, Headridge and Taylor96 have claimed success in reducing niobium in a Jones’ reductor in a fluoride solution. The reduced niobium was collected in a solution of ammonium ferric sulphate and the ferrous iron produced was determined by titration with potassium di- chromate.TomiCek and Spurnyg7 found that the reduction of niobium solutions by zinc amalgam was unsatisfactory, but that electro-reduction of niobium in a solution containing 26 per cent. of sulphuric acid at a mercury cathode was complete and reproducible. The tervalent niobium so formed was oxidised by ferric iron, and the ferrous iron produced was titrated with 0.01 N ceric sulphate, o-phenanthroline being used as indicator. If titanium was present a correction had to be made; the value of this correction was found by reducing the titaniumAugust , 19621 TANTALUM AND NIOBIUM. A REVIEW 621 salts with zinc amalgam in 25 per cent. sulphuric acid in a stream of hydrogen and then titrating the titanous ions potentiometrically. The fact that niobium tended to be reduced to the tervalent state led later investigators to apply polarographic methods to its determination.That both niobium and tantalum are reduced in oxalate and tartrate solution has been shown polarographically and spectro- metri~ally.~~ Dharg9 found that niobium produced a catalytic reduction wave that was depressed by chloride, perchlorate and sulphate ions, but enhanced by the presence of nitrate ions; he concluded that a complex nitro-niobate was formed. Tantalum, as a solution in hydrofluoric acid, in a supporting electrolyte saturated with sodium fluoride and at least 0-05 M in hydrofluoric acid has been found to give a polarographic wave of height proportional to the concentration of the metal; the half-wave potential was Polarographic methods have been described for determining niobium in stainless steel,lol 9 l o 2 in alloys used in nuclear W O ~ ~ ~ ~ ~ J ~ ~ ~ ~ ~ ~ and in titanium ores and pigments.lo6 The polaro- graphic behaviour of niobium in various media was thoroughly investigated by Ferrett and Milner,1°7 who used both conventional and square-wave polarography ; this investigation included a study of the polarography of niobium in solutions containing ethylenediaminetetra- acetic acid as did that of Kennedy later.los Polarographic methods are useful mainly for alloyslog and simple compounds of niobium or for niobium in solution with other metals, such as titanium.ll0 For complex minerals, preliminary separation of the tantalum and niobium is necessary.-0.95 volt .loo COLORIMETRIC METHODS It has been known for some time that niobium forms a yellow colour with hydrogen peroxide in concentrated sulphuric acid. In 1939 a method for determining tantalum, niobium and titanium in steel was described by Klinger and Koch,lll who made use of the fact that, in anhydrous sulphuric acid, the intensity of the titanium peroxide colour was a t a minimum, whereas that of the niobium peroxide colour was maximal. Thanheiser112 decreased the interference from titanium by adding phosphoric acid. Geld and Carrol1113 used this method for determining niobium in high-temperature alloys; they found that the presence of water increased the interference from titanium and therefore used dehydrated sulphuric acid.There was no interference from tantalum, which did not appear to form a complex with hydrogen peroxide under these conditions. Telep and used an ultraviolet spectro- photometer for investigating the absorption spectrum of perniobic acid ; they found that Beer’s law was obeyed at 324 mp in a sulphuric acid - phosphoric acid medium and that, although 150 p.p.m. of tantalum did not interfere, iron had to be removed. During an examination of the absorption spectra of the peroxy compounds of transition elements, Palilla, Adler and Hiskey115 y116 reported on a peroxy-tantalate having an absorption peak at 285mp in concentrated sulphuric acid. As the absorption peak of the peroxy-niobate is at 635 mp, a simultaneous determination of tantalum and niobium was possible.The absorption at 365mp can be measured with an absorptiometer fitted with a Wood’s glass filter and a mercury-vapour lamp. The peroxide method has been adapted for determining niobium in low-grade ores117y11* and has also been recommended for the determination of niobium in “pure” niobium.119 It was claimed that exceptional accuracy (& 0.1 per cent. at the 90 to 100 per cent. level) was attainable by measuring the optical density of the peroxy complex at 352 mp against standards of known composition. Hydroquinone120J21 and catecho1122 have been proposed as colorimetric reagents for nio- bium. By extracting the complexes with catechol under suitable conditions, niobium, tantalum and titanium can be separated from each other and from iron, zirconium, tungsten and aluminium.Niobium and tantalum form coloured complexes with catechol a t a pH of about 2.5 in tartrate or oxalate solutions in the presence of ethylenediaminetetra-acetic acid ; the yellow tantalum complex has an absorption maximum in the ultraviolet, and the niobium complex at 470 mp in the visible region. P a t r o v ~ k S ; ~ ~ ~ gave details of a method based on this reaction for determining both tantalum and niobium as mixed oxides in the presence of some impurities. Various workers have proposed methods involving use of the blue complex formed by niobium with m o l y b d o p h o ~ p h a t e , ~ ~ * ~ ~ ~ ~ ~ ~ 2 ~ and a method based on the formation of a yellow complex between tiron and tervalent niobium has also been described.127622 COCKBILL THE DETERMINATION OF [Vol.87 Kanzelmeyer and Freund128 reported the formation in hydrochloric acid solution of a niobium chloride complex having an absorption peak at 281 mp. Although several metals form complexes absorbing at this wavelength under these conditions, these workers evolved an analytical procedure for determining niobium. Tantalum forms a complex with hydroquinone, and, although niobium interferes, measure- ment of the absorption of the complex at 375 mp permits tantalum to be determined.129 The formation of an 8-hydroxyquinolinate of niobium permits its determination in ~teel.1~0 The steel is dissolved in perchloric acid, and the niobium is precipitated by hydrolysis with sulphurous acid, the precipitate being re-dissolved in ammonium citrate solution. The 8-hydroxyquinolinate is then extracted into chloroform, and the absorption is measured at 385 mp.Other coloured complexes formed by niobium and used for its determination have been ascorbic acid,lsl anthracene chrome violet132 and alizarin.133 Several workers have drawn attention to the coloured complex formed by niobium with the thiocyanate ion; tantalum forms no analogous complex, but titanium does and is liable to interfere. Norwitz, Codell and VerderamelM proposed the use of precipitation with tannin to remove nearly all of the titanium and subsequently measured the absorption of the yellow niobium - thiocyanate complex at 420 mp to decrease interference from titanium. In an adaption of the thiocyanate method for determining niobium in low-grade minerals,ls5 the colour was measured at 385 rnp in a solution reduced by stannous chloride; preliminary separation of niobium was necessary.Marzys136 described a similar method for low-grade ores containing 0.1 to 0.6 per cent. of niobium pentoxide. In this method, acetone was added to the solution to stabilise the complex, and the optical density was measured at 405 mp. If copper or uranium was present, the complex was extracted into ether before the measurement was made. Ward and Marran~inol~~ formed the niobium - thiocyanate complex in tartaric acid solution, concentrated it by extraction into ether and stabilised this solution with acetone, the optical density of the complex being measured a t 385mp; the method was intended for determining microgram amounts of niobium in rocks. For the determination of small amounts of niobium in “pure” tantalum and its oxide,138 the thiocyanate complex of niobium was extracted into ether and its absorption was measured at 386 mp.The niobium - thio- cyanate complex has also been separated from high purity tantalum oxide by an ion-exchange method.139 Crouthamel, Hjelte and Johnson140 devised a method for determining titanium, vanadium (tervalent) and niobium by measuring the optical densities of their complexes with thiocyanate ; the procedure included special pre-treatment of the samples to give correct development of colour. In order to avoid a time-consuming and often uncertain separation of titanium, Mundy141 proposed conditions permitting determination of both titanium and niobium from measurements at different wavelengths.Bacon and Milner142 investigated the conditions under which a stable niobium - thiocyanate complex could be formed so that its optical density was not critically dependent on changes in the concentrations of reagents. After investigating the extraction of the niobium - thiocyanate complex, with use of niobium-95 tracer, T r ~ i t s k i i l ~ ~ stated the optimum extraction conditions to be 0.08 to 0.10 g of potassium thiocyanate per ml, 0.10 to 0.12 g of hydrochloric acid per ml and 0.01 to 0.05 g of stannous chloride per ml; butanol was used as extracting agent, and 89 per cent. of the complex was extracted. Pyrogallol as a colorimetric reagent for tantalum appears to have been first mentioned by Platanov, Krivoshlikoff and Maraka’ye~l~~ and has since been investigated by several Russian workers, e.g., Alimarin and Frid.145 Hunt and Wells43 investigated this method and found that niobium and tantalum interfered with each other, but that the interference was linear in both instances and could be corrected for; the pH and the concentrations of salts in the solution of the complex had to be strictly controlled.The method was designed for determining tantalum and niobium in pure mixed oxides from a previous chromatographic separation. The oxides were fused in the minimum amount of potassium hydrogen sulphate, and the melt was dissolved in a little ammonium oxalate solution. For the determinations of tantalum and niobium, aliquots of this solution were made up to precisely defined concen- trations of potassium hydrogen sulphate and ammonium oxalate.The niobium colour was developed in an alkaline solution of pyrogallol by adding a solution of sodium sulphite in pyrogallol solution and the absorption was measured with a spectrophotometer a t 410 mp. The tantalum colour was developed in acid pyrogallol (pyrogallol solution plus sulphuric acid) and measured at 400 mp. Standard graphs were plotted from the results obtained by .usingAugust, 19621 TANTALUM AND NIOBIUM. A REVIEW 623 pure oxides of tantalum and niobium. The line for one metal, together with the correction for the other, was plotted on the same co-ordinates, and the corrections could be made by a succession of approximations, e.g., an apparent result for tantalum was obtained and sub- tracted from the known weight of mixed oxides present, thereby giving an approximate result for niobium pentoxide. The optical density corresponding to this amount of niobium was read from the “niobium-correction” curve, and the weight of tantalum pentoxide equiva- lent to this reading was then subtracted from the original apparent result for tantalum to give a truer figure; the correction could be repeated, but in practice this was unnecessary.This method is often used now, although several metals interfere, e.g., iron, tungsten, vanadium and especially titanium, one part of titanium corresponding to about four parts of tantalum. As the method is applicable to chromatographically prepared mixed oxides, tungsten is the only likely impurity; this can be determined colorimetrically in another portion of the sample, and allowance can be made for its presence in calculating the niobium and tantalum contents.In practice, the method is not suitable for determining less than about 5 per cent. of niobium in tantalum, as the correction to be made is large. The pyrogallol method has been used for determining tantalum in tantalum - titanium Separation with tannin can be used to remove titanium, but niobium interferes by causing some occlusion of titanium in the tantalum - tannin precipitate. As titanium also forms a yellow colour with pyrogallol, recovery is erroneously high. Tungsten interferes because it too forms a yellow colour with pyrogallol, but other metals normally encountered in such alloys do not interfere. D i n n i ~ ~ l ~ ~ described a method for determining tantalum in an acid pyrogallol solution.The absorption was measured at 385 mp, some corrections were necessary, and the method was better for pure mixed oxides than for impure material. Improved procedures have been suggested.148 A photometric method was proposed for determining niobium, tungsten and tantalum in stainless ; tantalum was determined by a pryrogallol method, and the optical densities developed by niobium and tungsten were measured at different wavelengths in a solution containing pyrogallol, stannous chloride, hydroquinone and phosphoric acid. A similar method has been described for determining tantalum via a yellow complex formed in an ethanolic solution of gallic acid.150 Marzys,151 in a procedure devised for application to ores, combined the thiocyanate - acetone method for determining niobium with a modified pyrogallol procedure for deter- mining tantalum in the same sample solution.The modifications to the pyrogallol procedure consisted in using tartaric acid as the complexing medium, to which was added hydrochloric acid and specified amounts of ammonium oxalate to give maximum absorption at 350 mp, which is available on an absorptiometer. Niobium did not interfere, but a correction had to be applied if titanium was present. No preliminary separation of tantalum and niobium was necessary except in rare instances. Tungsten, for example, rarely occurs in Sukulu soils, for which the method was devised, but a modification to include a correction for tungsten was described for use with samples in which it was present.Methods in which pyrogallol is used have been described for the determination of tantalum in zirconium alloys152 and in niobium ~ent0xide.l~~ Fluorone derivatives have been recommended as sensitive reagents for detecting and determining tantalum. Nazarenko and Shustovais4 described a method for determining tantalum in niobium metal, in which the tantalum was first separated by selective precipitation with 9-(~-dimethylaminophenyl)-2,3,7-trihydroxy-6-fluorone, known as “dimethyl fluorone.” The precipitate was ignited to give tantalum pentoxide, which was fused in potassium hydrogen sulphate, and then ammonium oxalate and water were added to dissolve the melt. The intensity of the colour developed by this solution with dimethyl fluorone was measured at 530 mp, These workers described the accurate determination of 0.002 per cent.of tantalum in 1-g samples of niobium or zirconium. More recently, Luke46 used phenylfluorone, (2,3,7- trihydroxy-9-phenyl-6-fluorone) in determining tantalum, measurements again being made at 530 mp. In this work, interference was avoided by extracting the tantalum into isobutyl methyl ketone and masking any extracted metals with ethylenediaminetetra-acetic acid ; with care, the method was suitable for determining small amounts of tantalum. SPECTROGRAPHIC METHODS Breckpot and Creffie~?~~ developed in 1937 a spectrographic method for determining tantalum and niobium, and procedures were described by S~hliessmanl~~ before the second World War. Methods for analysing very small samples have been described,157 and various624 COCKBILL THE DETERMINATION OF [Vol.87 procedures for determining the ratio of tantalum to niobium in mixed oxides have appeared from time to time.158J59J60 Most methods necessitate preliminary separation of the mixed oxides, and this part of the determination cart be lengthy. Corrections must sometimes be applied for titanium, and a series of standards is required. Spectrographic methods have been proposed for specialised purposes, such as the routine determination of niobium in large stocks of thin steel, uranium alloys, etc.,161 and of small amounts of niobium in ores.162 91G3 The disadvantage of optical spectrographic methods is the complicated nature of the lines produced.To a certain extent, this drawback can be overcome by using an instrument fitted with a grating instead of a prism to achieve a greater dispersion of the spectrum and by using photomultipliers to cobtained quantitative results. X-RAY FLUORESCENCE METHODS X-ray emission spectrography, or X-ray fluorescence, can be applied to the determination of small amounts of tantalum in niobium165 and of tantalum and niobium in mixed oxides.166~1~7 There is no need for preliminary separation of the elements from the other constituents of the sample, and, although tantalum and niobium are subject to serious line interference in the optical region, they are not so subject in the X-ray region. However, the nature of the chemical combination of the tantalum or niobium in an ore, as well as the physical state of the sample, will affect the intensity of its X-ray fluorescence spectrum.Absorption of fluorescent X-rays by a powder increases with its grain size,lS8 but this effect can be decreased by grinding the sample with an excess of an amorphous material, such as starch, and using the mixture as the target for bombardment. Other effects can be minimised by calibrating the instrument with a series of analysed samples as similar as possible (physically and chemically) to the unknown. Such a calibration is straightforward for simple compounds, such as alloys, but not for minerals, although a method involving arithmetical correction for inter-element effects has been described for the determination of tantalum, niobium, iron and titanium in complex mate1-ia1s.l~~ With a lithium chloride crystal, about 0.01 per cent.of tantalum in niobium should be recognisable above the background. A method for determining niobium in aqueous solution has been described for steels.l70 Provided that adequate standards are available, the separate determination of tantalum and niobium by X-ray fluorescence takes a inatter of minutes. RADIOACTIVATION METHODS Tantalum and niobium can be determined in the presence of each other by irradiating the sample with neutrons to produce radioactive isotopes of the metals, which can be measured separately. For the determination of tantalum mixed with niobium and titanium, Beydon and Fisherl7I suggested that the sample be irradiated with slow neutrons and that the gamma radiation be counted after 24 hours. Of the radioactive isotopes present, tantalum-182 had the longest half-life (111 days, compared with several minutes for the isotopes of niobium and titanium). Titanium could be determined colorimetrically, and niobium-95 tracer was used to determine the niobium by a procedure involving use of tannin.Kohn172 suggested a similar method for determining tantalum in ores and in ferro-niobium. The sample, with standards, was irradiated in a pile and then set aside for 7 days. All impurities had short-lived isotopes, but a correction had to be made for the activity due to iron. By irradiating a sample of tantalum mixed with niobium, iron, zirconium, tin, titanium, and silicon in a silica capsule for a week in the Harwell pile, Long173 achieved an accuracy of about 97 per cent. in mixtures containing more than 1 per cent.of tantalum. As this technique needs a source of neutrons its use is limited. Activation methods have been described for determining tantalum in niobium metal174 and in r0~ks.l'~ By using radioactive tracers, many of the less accurate separations may be corrected. The inclusion of niobium-95 in a separation with tannin will reveal any inaccuracies in the procedure. Milner and Smale~l'~ used radioactive niobium to correct for small amounts of niobium lost in the chemical treatment during its colorimetric determination in steels. Any chemical separation can be corrected by this method, provided that a suitable isotope is available. If a known trace amount of tantalum-182 is added during a determination so that it appears in the final mixed oxides, it is then necessary only to separate a small arbitrary amount of a pure tantalum compound, by any convenient method, to be able to measure the fraction of the total tantalum that has in fact been separated (personal communication from Mr.G. M. Holmes).August, 19621 TANTALUM AND NIOBIUM. A REVIEW 625 Boyd and Galan177 and Alimarin and Bilimovi~hl~~ have described a method in which radio-isotope tracers were used in the determination of small amounts of tantalum and niobium. A method for determining niobium in its alloys with, for example, chromium, has been des- cribed; this depends on the intensity of reflection of beta rays from the ~ a m p l e . l ~ ~ J ~ ~ MISCELLANEOUS METHODS Attempts have been made to use a thermobalance as a means of determining tantalum and niobium.Doan and Duvallsl reviewed such methods for determining niobium, but could find none that was entirely satisfactory. Tantalum catalyses the formation of iodine in the reaction between hydrogen peroxide and acid iodide solution, and this property has been usedls2 for determining concentrations of tantalum in the region of Several other ions have this,property, but niobium has not. The development of the starch - iodine colour was followed with a spectrophotometer, and the rate of reaction was computed. The rate of reaction of the same reagents was then measured after a known concentration of tantalum had been added, and the increase in the rate was used to determine the tantalum.M. CONCLUSIONS It is clear that the number of methods available for determining tantalum and niobium is large. Only if a series of determinations by each method was carried out would it be possible to recommend the best with confidence, but it is possible to assess the existing methods in the light of experience of some of them. First, it is nearly always necessary either to prepare pure mixed oxides from the assay sample or to make possibly complicated corrections for whatever interfering elements might be present. Methods based on Schoeller’s tannin or Marignac’s alkali-double-fluoride pro- cedures have the disadvantage of being time-consuming, but, in a laboratory equipped for the analysis of minerals, these methods may be suitable because they have been adapted to include the separation and determination of most elements likely to be found in association with tantalum and niobium.However, a great deal of experience in their practice is needed before these methods can yield reliable results; this standard of analytical ability is an art needing an apprenticeship. Gravimetric methods do not usually lead to the separation of tantalum from niobium cleanly in one step. In some special instances, a specific precipitant may isolate tantalum or niobium, but even so the compound obtained is not usually suitable for weighing directly. Previous isolation of the earth acids is always necessary. Methods depending on reduction of niobium to the tervalent state are unreliable. Niobium salts in solution are unstable and easily hydrolysed, especially in the presence of tantalum, so that the chances of attaining exactly the same conditions in replicate determinations are small. The nearest approach to a standard method for determining niobium in solution by reduction is one that gives precise instructions, so that the same state of reduction is always attained.Polarographic methods based on the formation of tervalent niobium are repro- ducible, but are not generally applicable to other than simple alloys and mixtures without lengthy preliminary separation of the earth acids. The only procedures available that can separate tantalum and niobium from all except traces of impurities, and are capable of so doing in one operation, are chromatography and ion-exchange methods. Liquid - liquid extraction methods are not as easily applied to analytical problems.Ion-exchange methods are lengthy and are more suited to the preparation of pure tantalum and niobium compounds than to their determination. Chromatography on cellulose also takes a long time for the complete separate determination of the metals, but may be adapted to extract both metals together in a much shorter time. This is probably the best available method for preliminary separation of tantalum and niobium together from all other constituents of a sample. For extremely small amounts of tantalum and niobium, a paper strip (which corresponds to a small-scale version of the cellulose column) is more conveniently used. Once isolated, the tantalum and niobium can be determined colorimetrically.Several precipitations may be necessary.626 COCKBILL: THE DETERMINATION OF [Vol. 87 The determination of tantalum and niobium has now reached a point at which it is Dissolve the sample by one of the recognised methods leading to a solution of the tantalum and niobium in hydrofluoric acid. Transfer this solution to a short column of cellulose, and isolate the tantalum and niobium together, free from all impurities, or, if the amount of tantalum and niobium is too small for convenient gravimetric handling, isolate the metals by paper-strip chromatography. possible to suggest a general method for all samples; such a procedure is outlined below. If the mixed oxides are obtained via a short column and contain- (i) a trace of niobium : determine this element by paper-strip chromatography, with subsequent spectrophotometric measurement of the thiocyanate complex ; (ii) a trace of tantalum : determine this element by paper-strip chromatography, with subsequent measurement of the phenylfluorone complex; (iii) all other ratios of the two elements: analyse for tantalum and niobium as described by Hunt and both metals being determined via their pyrogallol complexes. If the oxides are more conveniently obtained on a paper strip, complete the deter- mination colorirnetrically as in (i) and (ii) above.The procedure outlined above is not based on a complete exploration of all existing methods, and it might be more convenient to use other methods in special circumstances. For example, when determining small concentrations (0.001 per cent.) of tantalum in steel, it is more convenient to precipitate the tantalum, together with zirconium as carrier, by phenylarsonic acid from a solution of the steel and then to determine the tantalum in the oxides so obtained, as already described; this procedure serves for a preliminary concentration of the minute amount of tantalum present.If it is convenient to isolate and then to separate the tantalum and niobium by liquid - liquid extraction, it might prove easier to determine the niobium with phenylfluorone as described by Senise and Sant’Agostin0.18~ The advantage of this method is that hydrofluoric acid is not used and hence the extractions can be carried out in conventional glass apparatus. This review shows that the determination of tantalum and niobium has received more attention than is generally supposed and does not now present the difficulties once associated with it.I thank the Directors of the London & Scandinavian Metallurgical Co. Ltd. for permission to publish this review and Dr. W. Rosenblatt, chief chemist of that company, for allowing nearly 40 years’ experience of earth-acid chemistry to be freely drawn upon. Several of the more modern procedures, arrived at by judicious combinations of existing methods, are due to Mr. G. M. Holmes, works chemist of the above company, and Mr. P. J. McGloin, who have kindly made all their work available for this review. I also thank Mr. J. Magid, who tested some methods for routine use. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Morozov, I.S., Zhur. Neovg. Khim., 1956, 1, 791; Chem. Abstv., 1957, 51, 3340 I. Crookes, Sir W., “Select Methods of Chemic:al Analysis,” Fourth Edition, Longmans, Green & Co. Bagshawe, B., and Elwell, W. T., J . SOC. Chem. Ind., 1947, 66, 398. Marignac. J., Ann. Chim. Phys., 1866, 8, 5 and 60; 1866, 9, 249. Chemikerausschuss der Gesellschaft Deutscher Metallhutten- und Bergleute e.V., “Analyse dev Schoeller, W. R., “The Analytical Chemistry of Tantalum and Niobium,” Chapman and Hall Jaboulay, B. E., Rev. Metal., 1948, 45, 343. -, Chim. Anal., 1955, 37, 198; Anal. Abstr., 1955, 2, 2709. Wirtz, H., and Rothmann, H., 2. Erzbevgbau. u. Metallhiittenw., 1958, 11, 465; 1959, 12, 612. Traub, K. W., Ind. Eng. Chem., Anal. Ed., 1946, 18, 122. Alimarin, I. P., and Burova, T.A., Zhur. Prikl. Khim., 1945, 18, 289. Tougarinoff, B., Centenoive de 1’A. Congres, Sect. Met. Nonferreux, 1947, 134. Fucke, H., and Daublander, J., Tech. Mitt. Krupp. Forsch., 1939, 14, 174. Majurndar, A. K., and Mukherjee, A. K., Naturwissenschaften, 1958, 45, 239, Dupraw, W. A., paper presented at the Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, February, 1955; abstracted in Anal. Chem., 1955, 27, 309. Ltd., London, 1905. Metalle,” Springer-Verlag, Berlin, Gottingen and Heidelberg, 1949, Volume I, p. 324. Ltd., London, 1937.August, 19621 TANTALUM AND NIOBIUM. A REVIEW 627 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. Majumdar, A. K., and Mukherjee, A. K., Anal. Chim. Acta, 1959, 21, 330. Kidman, L., Darwent, C. L., and White, G., Metallurgia, 1960, 62, 125. Saint-James, R., and Lecomte, T., Anal. Chim. Acta, 1961, 24, 155. Gillis, J., Eeckhout, J., and Poma, K., Ver. Kon. Vlaam. Acad. Wetenschap. Belg., 1944, 6, 5. TomiCek, O., and Spurny, K., Chem. Listy, 1952, 46, 11. Popa, G., Negoin, D., and Barnlescu, G., 2. anal. Chem., 1954, 165, 16. Krishna Rao, B. S., Sarma, D. V. N., and RaghavaRao, B. S. V., Ibid., 1958, 160, 351. Chan, F. L., Talanta, 1961, 7, 253. Moshier, R. W., and Schwarberg, J. E., Anal. Chem., 1957, 29, 948. Majumdar, A. K., and Mukherjee, A. K., Naturwissenschaften, 1957, 44, 491.Langmyhr, F. J., and Hongslo, T., Ibid., 1960, 22, 301. Majumdar, A. K., and Ray Chowdhury, J . B., Ibid., 1958, 19, 18. Ponomarev, A. I., and Sheskol’skaya, A. Ya., Zhur. Anal. Khim., 1959, 14, 67. Wernet, J., 2. anorg. Chem., 1952, 267, 213. Alimarin, I . P., and Stepanyuk, E. I., Zavod. Lab., 1958, 24, 1064. Grimaldi, F. S., and Schnepfe, M. M., Anal. Chem., 1958, 30, 2046. Gibalo, I. M., and Malyarov, K. L., Metody Analiza Redkikh. i Tsvet. Metal., Sbornik, 1956, 105; Malissa, H., Mikrochim. Acta, 1958, 726. Belekar, G. K., and Athavale, V. T., Analyst, 1957, 82, 630; see also references listed in this paper. Reboul, R., I n d . Chim. Belge, 1959, supplement 1, 109. Burstall, F. H., Swain, P. J., Williams, A. F., and Wood, G. A., J . Chem.SOC., 1952, 1497. Williams, A. F., Ibid., 1952, 3155. Mercer, R. A., and Williams, A. F., Ibid., 1952, 3399. Burstall, F. H., and Williams, A. F., Analyst, 1952, 77, 983. Wood, G. A., National Chemical Laboratory Report CRL/AE 62, Teddington, Middlesex, 1950. Mercer, R. A., and Wells, R. A., Analyst, 1954, 79, 339. Hunt, E. C., and Wells, R. A., Analyst, 1954, 79, 345; National Chemical Laboratory Report CRL/AE 112, Teddington, Middlesex, 1953. -- , Analyst, 1954, 79, 351; National Chemical Laboratory Report CRLIAE 112, Tedding- to;, Middlesex, 1953. Hunt, E. C., North, A. A., and Wells, R. A., Analyst, 1955, 80, 172. Luke, C. L., Anal. Chem., 1959, 31, 904. Tikhomiroff, N., Compt. Rend., 1953, 236, 1263. Bruninx, E., Eeckhout, J., and Gillis, J., Mikrochim. Acta, 1956, 688.Blasius, E., and Czekay, A., 2. anal. Chem., 1957, 156, 81. Scott, I. A. P., and Magee, R. J., Talanta, 1958, 1, 329. Aleksandrova, L. S., and Chmutov, K. V., Izvest. Akad. Nauk SSSR, Otdel Khim. Nauk, 1960, Cabral, P. J. M., Rev. Port. Quim., 1959, 2, 51. Kraus, K. A., and Moore, G. E., J . Amer. Chem. SOC., 1949, 71, 3263 and 3855. Wacker, R. E., and Baldwin, W. H., U.S. Atomic Energy Commission Report ORNL-637, Oak Krauss, K. A., and Moore, G. E., J . Amer. Chem. SOC., 1951, 73, 9 and 2900. Huffman, E. H., Iddings, G. M., and Lilly, R. C., Ibid., 1951, 73, 4474. Huffman, E. H., and Iddings, G. M., Ibid., 1952, 74, 4714. Gillis, J., Hoste, J., Cornand, P., and Speecke, A., Med. Vlaam. Chem. Ver., 1953, 15, 63. Speecke, A., and Hoste, J., Ibid., 1957, 19, 190.Herrmann, M., I n d . Chim. Belge, 1958, 23, 123. Hague, J. L., Brown, E. D., and Bright, H. A., J . Res. Nut. Bur. Stand., 1954, 53, 261. Cabell, M. J., and Milner, I., Anal. Chim. A d a , 1955, 13, 258. -- , J . Appl. Chem., 1955, 5 , 482. HagAe, J. L., and Machlan, L. A,, J . Res. Nut. Bur. Stand., 1959, 62, 11 and 53. Kraus, K. A., and Nelson, I?., “Ion Exchange and Chromatography in Analytical Chemistry,” Special Publication No. 195, American Society for Testing Materials, Philadelphia, Pa., 1958. Spauszus, S., and Heimer, M., Chem. Tech., 1961, 13, 96. Speecke, &4., and Hoste, J., Talanta, 1959, 2, 332. Bandi, W. R., Buyok, E. G., Lewis, L. L., and Melnick, L. M., Anal. Chem., 1961, 33, 1275. Schafer, H., Angew. Chem., 1959, 71, 153; see also references listed in this paper.Atkinson, R. H., Steigman, J., and Hiskey, C. F., Anal. Chem., 1952, 24, 477. Hiskey, C. F., Newman, L., and Atkinson, R. H., Ibid., 1952, 24, 1988. Werning, J. R., Higbie, K. B., Grace, J. T., Spreece, B. F., and Gilbert, H. L., I n d . Eng. Chem., Werning, J. R., and Higbie, K. B., Ibid., 1954, 46, 2491. Higbie, K. B., and Werning, J. R., U.S. Bureau of Mines Report of Investigation No. 5239, Faye, G. H., and Inman, W. R., Canad. M i n . Met. Bull., 1957,50, 609; Canadian Department of Tews, J. L., and May, S. L., U.S. Bureau of Mines Report No. U-252, Washington, D.C., 1957. Milner, G. W. C., and Wood, A. J., U.K. Atomic Energy Research Establishment Report C/R-895, , , Anal. Chim. Acta, 1959, 21, 245. -- Chem. Abstr., 1959, 53, 1994 F., , Anal. Chim. Acta, 1956, 14, 74. ~ _ _ - 801; Chem. Abstr., 1961, 55, 141. Ridge, Tennessee, 1950. 1954, 46, 644. Washington, D.C., 1956. Mines and Technical Surveys Research Report No. MD-210, Ottawa, 1957. Harwell, 1952.628 79. 8F). 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. COCKBILL : THE DETERMINATION OF Milner, G. W. C., and Edwards, J . W., Anal. Chim. Acta, 1955, 13, 230. Leddicotte, G. W., and Moore, F. L., J . Anter. Chem. SOC., 1952, 74, 1618. Ellenburg, J. Y., Leddicotte, G.W., and Moore, F. L., Anal. Chem., 1954, 26, 1045. Moore, €7. L., Ibid., 1955, 27, 70. Scadden, E. M., and Ballou, N. E., Ibid., 1953, 25, 1602. Stevenson, P. C., and Hicks, H. G., Ibid., 1953, 25, 1517. U.S. Atomic Energy Commission, U.S. Patent No. 2,795,481, 1957. Fletcher, J . M., Morris, D. F. C., and Wain, A. G., Bull. Inst. M i n . Met., 1956, 65, 487 T. U.K. Atomic Energy Authority, British Patent No. 767,038, 1957. Chernikhov, Y. A., Tramm, R. S., and Pevzner, K. S., Zavod. Lab., 1956, 22, 637. Alimarin, I. P., and Gibalo, I. M., Dokl. Ahad. Nauk S S S R , 1956, 109, 1137. Theodore, M. L., Anal. Chem., 1958, 30, 463. Laudr, R. S., and Poludktov, N. S., Zavod. Lab., 1959, 25, 903. Cunningham, T. R., I n d . Eng. Chem., Anal Ed., 1938, 10, 233.Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis,” John Wiley & Sons Knowles, H. B., and Lundell, G. E. F., J . Bes. Nut. Bur. Stand., 1949, 42, 405. Oka, Y., and Miyamoto, M., Sci. Kep. Res. Inst., TGhoku Univ., A , 1949, 1, 115. Headridge, J. B., and Taylor, M. S., Anal:~st, 1962, 87, 43. TomiEek, O., and Spurny, K., Chem. Listy, 1952, 46, 6. Elson, R. E., J . Amer. Chem. SOC., 1953, 75, 4193. Dhar, S. K., Anal. Chivn. Acta, 1954, 11, 289. VlEek, A. A., Chem. Listy, 1955, 49, 260. Vivarelli, S., Chim. et Ind., 1955, 37, 1026. Brindley, D. J., Analyst, 1960, 85, 877. Elving, P. J., and Olson, E. C., Anal. Chem., 1956, 28, 338. Stricos, D. P., U.S. Atomic Energy Commission Report KAPL-M-DPS-3, 1960. Kennedy, J . H., J . Phys. Chenz., 1960, 64, 1590.Ralchin, L. A., and Williams, D. I., Analyst, 1960, 85, 503. Ferrett, D. J., and Milner, G. W. C., Nature, 1955, 175, 477; J. Chem. SOC., 1956, 1186. Kennedy, J. H., Anal. Chem., 1961, 33, 943. Mukhina, 2. S., and Tikhonova, A. A., Zavod. Lab., 1956, 22, 1154. Krylov, E. I., Kolevatova, V. S., and Samarina, V. A., Dokl. Akad. Nauk S S S R , 1954, 98, 593; Klinger, P., and Koch, W., Arch. Eisenhuttenw., 1939, 13, 127; Tech. Mitt. Krupfl Forsch., 1939, Thanheiser, G., Mitt. Kaiser Wilhelm Inst. Eisenforsch., 1940, 22, 255. Geld, I., and Carroll, J., Anal. Chem., 1949, 21, 1098. Telep, G., and Boltz, D. F., Ibid., 1952, 24, 163. Palilla, F. C., Adler, N., and Hiskey, C. F., Ibid., 1953, 25, 926. Hiskey, C. F., and Adler, N., J . Amer. Chem. SOC., 1958, 179, 1827.Pickup, R., Colon. Geol. Mineral Resour., 1953, 3, 358. -, Ibid., 1955, 5 , 174. Backer, R. O., Wiederkehr, V. R., and Goward, G. W., US. Atomic Energy Commission Report Johnson, C. M., Iron Age, 1946, 157, 66. McKaveney, J. P., Anal. Chem., 1961, 33, 744. Sarma, B., and Gupta, J., J . Indian Chena. Soc., 1955, 32, 285. Patrovsk9, V., Coll. Czech. Chem. Comm., 3958, 23, 1774. Davydova, A. L., Vaisberg, 2. M., and Burkser, L. E., Zavod. Lab., 1947, 13, 1038. Popel’, A. A., and Maksimova, L. P., Uch. Zap. Kazansk. Univ., 1956, 116, 86. Nonvitz, G., and Codell, M., Anal. Chem., 1954, 26, 1230. Flaschka, H., arid Lassner, E., Mikrochim. Acta, 1956, 778. Kanzelmeyer, J. H., and Freund, H., And. Chem., 1953, 25, 1807. Waterbury, G. R., and Bricker, C. E., Ibid., 1957, 29, 1474.Kassner, J. L., Garcia-Porrata, A., and Grove, E. L., Ibid., 1955, 27, 492. Janauer, G. E., and Korkisch, J., Anal. Chirn. Acta, 1961, 24, 270. Rarskaya, S. I., U.S.S.R. Patent No. 123,321. Mustafin, I. S., and RIolot, L. A., Ref. Zhuv., Khim., 1960, abstr. No. 84,453. Norwitz, G., Codell, M., and Verderame, F D., Anal. Chim. A d a , 1953, 9, 561. Milner, G. W. C., and Smales, A. A., Analyst, 1954, 79, 315. Marzys, A. E. O., Ibid., 1954, 79, 327. Ward, F. N., and Marranzino, A. P., Anal. Chem., 1955, 27, 1325. Hastings, J., and McClarity, T. A., Ibid., 1954, 26, 683. Bergstresser, K. S., Anal. Chern., 1959, 31, 1812. Crouthamel, C. E., Hjelte, B. E., and Johnson, C. E., Ibid., 1955, 27, 507. Mundv, R. J., Ibid., 1955, 27, 1408. Bacon, A., and Milner, G. W. C., Anal. Chim. Acta, 1956, 15, 129. Troitskii, K. V., Zhur. Anal. Khim., 1957, 12, 349. Platanov, M. S., Krivoshlikoff, N. F., and Maraka’yev, A. A., Zhur. Obshez’ Khim., 1936, 6, 1815. Alimarin, I. P., and Frid, B. I., Trudy Vses. Konf. Anal. Khirn., 1943, 2, 333. Nonvitz, G., Codell, M., and Mikula, J . J., Anal. Chim. Acta, 1954, 11, 173. Dinnin, J. I., Anal. Chem., 1953, 25, 1803. Inc., New York, 1929, p. 484. Anal. Abstr., 1955, 2, 3016. 14, 179. WAPD-204, Pittsburgh, Pa., 1958.August, 19621 TANTALUM AND NIOBIUM. A REVIEW 629 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161, 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. Dobkina, B. M., and Petrova, E. I., Zavod. Lab., 1951, 23, 421. Ikenberry, L., Martin, J. L., and Boyer, W. J., Anal. Chem., 1953, 25, 1340. Freund, H., Hammill, K. H., and Bissonnette, F. C., U.S. Bureau of Mines Report of Investigation Marzys, A. E. O., Analyst, 1955, 80, 194. Wood, D. F., and Scholes, I. R., Anal. Chim. Acta, 1959, 21, 121. Dobkina, B. M., and Petrova, E. I., Zavod. Lab., 1959, 25, 1064. Nazarenko, V. A., and Shustova, M. B., Ibid., 1957, 23, 1283. Breckpot, R., and Creffier, J., A n n . Soc. Sci. Bruxelles, 1937, 57, 290. Schliessman, O., Spectrochim. Acta, 1939, 1, 239; Tech. Mitt. Krupp Forsch., 1939, 14, 185; Herman, P., Spectrochim. Acta, 1948, 3, 389. Young, J. F., Iron Age, 1951, 162, 19. Peohlman, W. J., and Sarnowski, R. E., J . Opt. SOC. Amer., 1952, 42, 489. Cheadley, E. H., and Bowes, A. H., Murex Rev., 1952, 1, 274. Landis, F. P., and Pepkowitz, L. P., Anal. Chem., 1955,27, 141. Tarasevitch, N. I., Zheleznova, A. A., and Semenenko, K. A., Vestn. Moskov. Univ., 1957, No. 1, Nedler, V. V., Zavod. Lab., 1957,23, 1336. Moroshkina, T. M., and Prokof’ev, V. K., Vestn. Leningrad. Univ., Ser. Fiz. i. Khim., 1959, No. 2, 143; Chem. Abstr., 1959, 53, 1 6 8 2 4 ~ . Birks, L. S., and Brooks, E. J., Anal. Chem., 1950, 22, 1017. Campbell, W. J., and Carl, H. F., Anal. Chem., 1954,26, 800; Special Technical Publication No. 157, Mortimore, D. M., Romans, P. A., and Tews, J. L., Appl. Spectroscopy, 1954, 8, 24. Brown, F., Analyst, 1959, 84, 344. Mitchell, B. J., Anal. Chem., 1958, 30, 1894. Jones, R. W., and Ashley, R. W., Ibid., 1959, 31, 1629. Beydon, J., and Fisher, C., Anal. Chim. Acta, 1953, 8, 538. Kohn, A., Chim. et Ind., 1954, 71, 69. Long, J. V. P., Analyst, 1951, 76, 644. Halverson, G., and Shtasel, A., Anal. Chem., 1961, 33, 1627. Morris, D. F. C., and Olya, A., Talanta, 1960, 4, 194. Milner, G. W. C., and Smales, A. A., Analyst, 1954, 79, 425. Boyd, T. F., and Galan, M., Anal. Chem., 1953, 25, 1568. Alimarin, L. P., and Bilimovich, G. N., Coll. Czech. Chem. Comm., 1961, 26, 255. Bogdanov, N. A., and Funke, V. F., Zavod. Lab., 1955, 21, 181. Gaidadymov, V. B., and Il’ina, L. I., Trudg Komiss. Anal. Khim., 1958, 9, 240; Chem. Abstr., (This paper is one of a series dealing with various aspects of the chemistry Doan, U. M., and Duval, C., Anal. Chim. Acta, 1952, 6, 81. Yatsimirskii, K. B., Drobysheva, 0. M., and Rigin, V. I., Zhur. Anal. Khim., 1959, 14, 60; Chem. Senise, P., and Sant’Agostino, L., Anal. Chim. Acta, 1960, 22, 296. No. 5242, Washington, D.C., 1956. Jernkontorets Ann., 1939, 123, 58. 156. American Society for Testing Materials, Philadelphia, Pa., 1954. 1959, 53, 5009 D. of tantalum and niobium.) Abstr., 1959, 53, 9899 D. Received September 21st, 1961

ISSN:0003-2654    

DOI:10.1039/AN9628700611

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

630 WEST: THE SEARCH FOR NEW REAGENTS FOR [Vol. 87 The Search for New Reagents for Absorptiometry: Some Practical Considerations* BY T. S. WEST (Chemistry Department, University of Birmingham, Birmingham 15) The practical aspects of preliminary and final examination of potential spectrophotometric reagents is discussed with a view to the choice of reagent and the ions with which it is likely to react. Problems associated with sensi- tivity and lack of selectivity are dealt: with, and the discussion is illustrated by reference to some new spectrophotometric reagents for silver, copper, calcium, fluoride, chloride and sulphide. IT is important to realise that in the search for new absorptiometric reagents practical and theoretical considerations cannot be divorced. from each other.The theoretical approach to the choice of reagent must often take into account the experimental design within the framework of which the reagent must function, and the converse is also true. Accordingly, in this brief paper it is necessary to incorporate a certain amount of qualitative theory with the practical considerations. Then, as this paper was presented at a meeting entitled “New Analytical Reagents in Colorimetric Analysis,” it is appropriate to illustrate the various points with some examples of new reagents and thus to clothe the bare skeleton of practicality with some flesh of analytical interest. Briefly, these are to discuss the practical considerations that lead to the choice of a reagent and to outline the method of its examination from the initial through to the final stages.Since this can only be done sketchily in such a short paper it is appropriate to deal with matters under a series of headings. It is also appropriate to outline the t e r m of reference of this contribution. BASIS OF COLOUR FORMATION- The basis of colour formation is almost invariably complex formation, &., the formation of a co-ordination compound, Salt formation rarely if ever need be considered unless one classifies turbidimetric or nephelometric methods as true absorptiometric procedures. Colour can be developed by the reaction of two colourless ions, e.g., Bi3+ and I-, or more usually by allowing an already coloured molecule to react with a cation to produce a differently coloured complex. Most reagents are organic in nature and have a complex electronic structure.Colour production or change of colour can be attributed to deformation of, or interference with, the normal electronic structure of the reagent or of the ion itself. Most frequently we appear to be concerned with the former eventuality. Consequently, we will tend to look towards organic molecules containing conjugated bond systems or possessed of chromophoric groups capable of being affected directly or indirectly after co-ordination with the ion under test. NATURE OF THE REAGENT- It is in the nature of the reaction between a positively charged metal ion in solution and a co-ordinating agent that the latter should be neutral or anionic. Therefore, most reagents must possess such a character. If the metal ion is already present in solution in an anionic form, e.g., HgII as HgI,2-, then it is necessary for this to be broken down, or, more conveniently, one may use a positively charged molecule.Examples of this type of reaction are relatively rare, but a few recent ones may be noted in passing. For example, Cd2+ bound in solution as colourless CdI,2- reacts with the red coloured dipositive iris( 1,lO-phenanthro- line) - FeII ion to form an intensely coloured blue-red compound1 that can be extracted into chloroform. Similarly, antimohy, as SbCl,-, forms an extractable compound with the cationic dyestuff Brilliant green.2 There is no interference from 1000-fold amounts of arsenic, copper, nickel, zinc, cobalt, lead, etc., and the colours produced by other metals that form acido-complexes with chloride ion, viz., tin, mercury, cadmium and iron, are tran- sient, whereas the colour due to antimony remains unchanged for 24 hours.Similarly, * Presented a t the meeting of the Society on Thursday, February 8th, 1962.August, 19621 ABSORPTIOMETRY : SOME PRACTICAL CONSIDERATIONS 631 gallium, as the chlorogallate complex, can be determined with high precision in the presence of about 34 other metal ions3 The presence of metals such as aluminium actually favours the process. If two potential reagents are available, one of which can operate in an acidic medium and the other in an alkaline medium, then considerable attention must be paid to (a) the chemical characteristics of the metal ion and (b) the identity of other ions likely to be present in the test solution.Most absorptiometric reagents can be considered as weak acids that dissociate- HI? + H+ + R- and then react. [HI P I CHRI Dissociation constant, K = _____ Instability constant, K = W"+I CR-I" CMRXI Alternatively, one may consider the metal ion reacting with the reagent in acid solution and thus eliminating protons, i.e., Mn+ + xHR + MR,@-n)- + xH+. If the reagent - metal complex is reasonably stable, then the low value of the instability constant may best be taken advantage of in as acidic a medium as possible, when the ionisa- tion of the reagent is suppressed and the tendency for the weaker metal complexes to be formed is minimised. If the reagent is used in an alkaline solution, then the ready availability of the free ligand, R-, will result in less selective reaction, i.e., many other ions will also tend to react; against this must be reckoned the availability of a large number of masking agents that operate in alkaline solution. These may be used for direct masking or by a reversion technique depending on the chemical identities of the metal ions involved.The tendency of the reagent to ionise to form free ligand may be profoundly influenced by the substitution of suitably disposed electrophilic groups, etc., but this factor will not be dealt with here. NATURE OF THE COMPLEX FORMED- Logically, it may now be pointed out that the cadmium and antimony complexes mentioned above are neutral, i.e., they carry no charge and are insoluble in aqueous solution. So also are the metal complexes of 8-hydroxyquinoline, dimethylglyoxime, etc.With such reagents it is necessary to carry out an extraction in order to obtain a true solution for spectro- photometry. It is, of course, desirable to avoid extra operations, such as extraction, if there is no chemical reason for them. Consequently, it is advisable that the reagent should contain water solubilising groups, such as sulphonic acid substituents, which are not involved directly in the co-ordination reaction. Moreover, if it does become necessary to extract the metal complex for removal or concentration purposes this may still be done by forming an ion-asso- ciation complex with a suitable amine. The resulting neutral, and frequently unsolvated, complex usually passes readily across a phase boundary into an inert solvent such as carbon tetrachloride, toluene or chloroform.For example, 8-hydroxyquinoline-5-sulphonic acid does not form precipitates with calcium, magnesium, manganese, ironI1, ironII1, cadmium, cobalt, lead or ~ o p p e r , ~ but ultraviolet studies show that chelates are formed with aluminium, copper, lead, cobalt and n i ~ k e l . ~ The absorptions of these complexes can be measured in the ultraviolet region. The closely allied reagent 7-iodo-8-hydroxyquinoline-5-sulphonic acid was used by Yoe6 for the absorptiometric determination of ironII1. Recently, Ziegler and his co-workers' have shown that several advantages can be gained by extracting the complex into isopentanol with tri-n-butylamine as the counter ion. Thus it is possible to have the advantages of water solubility and extractability.I t is possible to learn quite a lot about the nature of the metal complex during the pre- liminary or exploratory experiments. For example, some idea of the state of the molecule can be obtained by attempts to extract the colour system into an inert solvent such as chloro- form OY in the presence of cationic or anionic counter ions or in the presence of oxygen-donor solvents such as isobutyl methyl ketone. Again, in the preliminary experiments, in the process of establishing the empirical constitution of the complex by a continuous variation procedure, e.g., Job's method, the curvature of the plot gives a good indication of the stability of the complex. If the curve is very rounded at the peak, a weak complex may be indicated,632 WEST: THE SEARCH FOR NEW REAGENTS FOR [Vol.87 and it is then a worthwhile proposition to try to bolster up the complex by addition of a water-miscible solvent such as acetone or by the addition of an indifferent electrolyte, which, for want of a better term, may be called a salting-out agent. BASIS OF REAGENT SELECTION- For a long time now analytical chemists have been obsessed by the idea of the functional group. Thus, for example, we know that most reagents containing the vicinal dionedioxime group -C-C- will react with Ni2+ in near neutral or alkaline solution to form a scarlet II II HON NOH precipitate, provided we do not supply a double bond characteristic to the -C-C- bond or attach -CF, groups to the carbon atoms.8 There are numerous other instances of such metal-selective groups.However, we do not know the real fundamental reasons behind such selectivity sufficiently well to design a new reactive grouping for a particular metal ion. Consequently, we can only be guided by experience and intelligent deduction from observed experimental phenomena with known or new systems and by an appreciation of known simi- larities in the ionic structure and periodic properties of ions. We can use such data to look hopefully for a selective reagent. Another alternative is to choose a multipurpose reagent and rely on masking action for selectivity. SEARCH FOR A PARTICULAR REAGENT- A short while ago the author and his co-worker, R. M. Dagnall, had occasion to search for a spectrophotometric reagent for silver ions present in only trace amounts.At such low levels, the standard method based on 9-dimethylaminobenzilidene rhodanine was completely unserviceable because of its extreme sensitivity to changes in acidity, etc., and the diphenyl- thiocarbazone (dithizone) method was insufficiently reliable because of the effects of variations in laboratory conditions, impurities in reagents and the extreme sensitivity of the silver extract towards light. One possibility was to try to modify the rhodanine reagent to obtain a more soluble derivative; another was to try to modify the dithizone molecule to obtain a derivative less sensitive to photo-decomposition, etc. Such reagents may be available in the dinaphthylthiocarbazone and 00’-ditolylthiocarbazone. However, we chose another avenue in the light of two facts.(1) Silver ion reacts with a wide range of anionic metallochromic reagents sufficiently well to “block” those reagents against other metals, and the complexes formed are usually not broken down readily. (2) Silver ion reacts rather weakly with the group of aminopolycarboxylic acids based on iminodiacetic acid (Le., the complexans), whereas most other metals react much more strongly-even the alkaline earths. With these facts in mind we examined a range of anionic metallochromic reagents known to be affected by silver ion and eventually obtained two reagents that proved to be satis- factory. The reagents, Pyrogallol red and Hromopyrogallol red, were subject to the series of tests outlined below- ( 1) Preliminary qualitative examination [Colour reaction/pH range/masking] I I I_ Suitable rkagents lJnszzitAble reagents (2) Preliminary measurement of absorption spectra.(3) Establishment of optimum pH for reaction (wavelength). (4) Establishment of optimum time of development (stability). (5) Continuous variation procedure (composition of complex). (6) Final absorption spectra. (‘7) Lambert - Beer law check. (8) Examination of interferences. (9) Elimination of interferences. IAugust , 19621 ABSORPTIOMETRY : SOME PRACTICAL CONSIDERATIONS 633 In the preliminary examination tests were carried out in small test-tubes with approximately 10-2 M solutions. A few drops of the silver solution were diluted to a reasonable volume and treated with reagent solution after adding, in one instance, mineral acid and, in others, various buffer solutions at pH 4, 7, 10 and 12.The colour reaction was then recorded in comparison with the reagent alone at the same pH. Simultaneously, the effect of adding an excess of ethylenediaminetetra-acetic acid (EDTA) was tested. Some reagents gave imperceptible reactions ; some produced colours of low intensity. Two proved worthy of further examination (see Fig. 1). As a result of the preliminary examination an idea of the approximate buffer region and an estimate of the time of colour development were known. The solution was then diluted to approximately lo4 M. A preliminary absorption spec- trum of the reagent alone and of the reagent in the presence of one, two and greater mole-ratios of silver ion next allowed us to select a wavelength for measurement of the complex and also furnished an idea of the reacting ratios (see Fig.2). OH Fig. 1. Pyrogallol red, R = H; Bromopyrogallol red, R = Br 0.20 - x .- $ 0.15- -0 Id - .- 'c 0.10- -0 0.05 - 0 300 400 500 600 I I 00 400 500 6 Wavelength, mp Fig. 2. Comparison of (a) pyrogallol red and (b) bromo- pyrogallol red under optimum conditions. Standard time of development, 90 minutes before dilution: curves A, 10 ml of reagent; curves B, 10 ml of reagent + 5 ml of M silver nitrate; curves C, 10 ml of reagent + 10 ml of 1 0 4 M silver nitrate; curves D, 10ml of reagent + 16ml of 1 0 - ' ~ silver nitrate; curves E, 10 ml of reagent + 20 ml of lo4 M silver nitrate The pH at which the colour reaction developed was now varied with fixed concentrations of reagent and silver, and changes in the absorption of the silver complex measured against the reagent were recorded, scanning the band area on either side of the selected wavelength to check on any marked shift in band maxima with change of pH.Now, at optimum pH and wavelength, the effect of time of colour development was measured with fixed concentrations of Ag+ ion and reagent over a period of time extending634 WEST: THE SEARCH FOR NEW REAGENTS FOR [Vol. 87 from 5 to 10 minutes upwards for several hours and overnight. This gave an idea of the minimum time necessary for optimum colour formation and a measure of the stability of the system. Armed with this knowledge the final steps, vix., constitution of the complex, recording of absorption spectra, Lambert - Beer law check, etc., were carried out on a routine basis.These particular reagents presented us with a rather more complicated path than usual because of a peculiar effect relating the speed of colour development to dilution. The colour develops rapidly in a small volume and is stable on dilution to volume, but, if the solution is diluted immediately the reagents are mixed, much less colour is formed and that only slowly. However, the use of both reagents provides reliable and extremely reproducible methods for determining ~ i l v e r . ~ The sensitivity in aqueous solution is not as high as that of dithizone, but, on the other hand, it is stable for more than 24 hours and is not affected by variations in laboratory conditions or impurities in reagents.The calibration curve is reproducible from day to day and from one laboratory to another. The interference of practically every ion except CuII can be eliminated by the addition of controlled amounts of EDTA as a masking agent, and the interference of CuII and other coloured metal ions by previous extraction of the silver as di-n-butylarnmonium silver salicylate into isobutyl methyl ketone and direct application of the colour reaction in the organic extract. By this procedure silver can be determined down to the 0-01 p.p.m. level.1° The reaction can also be applied to the indirect determination of chloride ion, once more providing a reliable and extremely sensitive met hod. l1 It is important that the preliminary qualitative tests should be carried out carefully and precisely.If this is not done it is possible to obtain misleading results and to miss completely colour reactions of extremely high sensitivity and great selectivity. For example, the alizarin complexan reaction for fluoride, the discovery of which was first announcedin December 1957,12 was completely overlooked by us when we first examined the effect of fluoride on the cerium111 and lanthanum complexes of alizarin c0mp1exan.l~ The reason was that we used a solution of fluoride sufficiently strong to prevent the formation of the blue ternary complex and sufficient to bleach the red lanthanon complexes with slow liberation of the free yellow reagent. This phenomenon has subsequently led others14 to argue that the method is limited to dilute solutions and cannot be used for amounts much above 50 pg or 10 p.p.m.In fact the method can be applied to >1600-pg amounts (>lo0 p.p.m.) without difficulty.16 The important point is that the reaction was only discovered when a much more dilute solution of fluoride ion was used and the test was observed more carefully over a longer period of time. Then, of course, the observer must have the intelligence and analytical background to appreciate the significance of such a colour reaction. It could quite easily have been missed or passed over without comment. This fluoride reaction has been dealt with at much greater length by and will not, therefore, be mentioned further in this discussion. Another example of the importance of precision and careful observation deserves to be cited.In studying the masking action of EDTA on sulphide precipitation, PSbiP reported that sulphide ion precipitates a dense black sulphide from alkaline iron solutions containing EDTA. Later on, however, when the experiment was repeated with proportionate amounts of sulphide ion, rather than a large excess, instead of a black precipitate, a soluble bright cherryred complex was obtained. The reaction is specific for sulphide ion or for FeIII. Subsequently, this led us to develop a similarly based method for the detennination of sulphide ion in the p.p.m. range.18 It is also appropriate to point out here that both these remarkably selective reactions are based on the formation of ternary rather than the more conventional binary complexes. This appears to be a significant trend in absorptiometry.Thus the ternary complex formed between ironII, 1,lO-phenanthroline and cyanide ion has recently been shown to provide an extremely sensitive method for the extractive determination of small amounts of cyanide ion,l and the 4,7-diphenyl-l , 10-phenanthroline also forms a similar complex that is extremely sensitive for ir0n.1~ The ternary complex with iodide ion, ironn and 1,lO-phenanthroline has been used similarly.20 We ourselves confirmed this result. SELECTIVE REAGENTS- Sometimes one wishes to examine a reagent as a potential absorptiometric reagent because of some peculiarity in its structure and because one feels that it may have a certain degreeAugust, 19621 ABSORPTIOMETRY : SOME PRACTICAL CONSIDERATIONS 635 of selectivity towards a group of metals.In the past, in conjunction with former co-workers, I have spent a considerable amount of time exploring and observing the reactions between 00'-dihydroxyazo dyes and metal ions with a view to obtaining metallochromic indicators and establishing relationships between structure and reactivity. As a result of this we have been able to synthesise some new reagents that have proved to be remarkably sensitive and selective. In the process, other existing compounds have also been found to be eminently suitable in a similar way. The preliminary net of examination cast over such a reagent must be considerably wider for, although one may have a strong indication among which group of metals complex forma- tion is to be expected, one cannot afford to neglect possible reactions with other metal ions.The procedure already indicated was followed for the reagent cyclotris-7- (l-azo-8- hydroxynaphthalene-3,6-disulphonic acid)21 (see Fig. 3). In this molecule the azonaphthalene rings carrying the hydroxyl groups have been cyclically linked together. Consequently, the reactive phenolic groups are closely packed in the centre in association with the azo-bonds. This provides a chelate cage into which only ions less than a certain size would be expected to fit. In an alkaline medium, about pH 12, the only ion to give a colour with the reagent is Ca2+. No colour is produced by Ba2+ or Sr2+. Fig. 3. Calcichrome The ionic radii of the alkaline earths are: Ca2+, 0.99 A; Sr2+, 1.12.A; Ba2+, 1.34 A. Other ions commensurate in size with Ca2+ are present in an anionic form, e.g., Zn2+, or are precipitated from solution, e.g., Mg2+.Fig. 4 shows the absorption spectrum of the reagent and its complex with Ca2+. The colour formed is stable for more than 24 hours and the method is free from interference by most metals except Cu2+. We have applied it successfully for the determination of calcium in the presence of 500-fold amounts of barium, strontium, zinc, etc." In our opinion this 10 Wavelength, mp Fig. 4. Absorption spectra at pH >12: curve A, metal-free reagent; curve B, calcium chelate636 WEST yol. 87 reagent is a considerable advance on glyoxal bis(2-hydroxyanil), the colour of which is stable for only 15 minutes after e~traction.~~ Further, the calibration is reproducible from day to day.Another extremely sensitive and selective reagent currently being developed by us is 2-naphthol-l- ( 3-sulpho-l-hydroxy-7-amino-2-azlonaphthalene)-8-sulphonic acid. This reagent is virtually specific for copper11 and can be used for its spectrophotometric determination at 510 mp down to 0.08 p.~.m.~* Similarly, the dyestuff Fast Sulphon black F25 (see Fig. 5 ) gives a highly sensitive and selective method for the spectrophotometric determination of copper11 down to 0.08 p.p.m. Preliminary indications are that only manganese11 and nickel are likely to cause serious interference.% The method works well down to 0.02 p.p.m. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 26. Fig. 5. Fast Sulphon black F REFERENCES Schilt, A.A., Anal. Chem., 1958, 30, 1409. Kristaleva, L. B., Zavod. Lab., 1959, 25, 1294. Kuznetsova, V. K., and Tananaev, N. A., Ref. Zhur. Khim., 1959, Abstr. No. 86,097. Albert, A., and Magrath, D., Biochem. J., 1947, 41, 534. Molland, J., Tidsskr. Kem. Berg., 1939, 19, 160. Yoe, J. H., J . Anzer. Chem. Soc., 1932, 54, 4143. Ziegler, M., Glemser, O., and Petri, N., 2. anal. Chem., 1956, 153, 415 and 1957, 154, 170; Angew Belcher, R., Sykes, A., and Tatlow, J. C., J . Chem. SOL, 1957, 2393. Dagnall, R. M., and West, T. S., Talanta, 1961, 8, 711. Betteridge, D., and West, T. S., Anal. Chim. Acta, 1962,26, 101; Dagnall, R. M., and West, T. S., Dagnall, R. M., and West, T. S., unpublished work. Belcher, R., Leonard, M. A., and West, T. S., quoted in West, T. S., “Recent Developments in Inorganic and Organic Analytical Chemistry,” Royal Institute of Chemistry Monograph No. 1, London, 1959, p. 21. Belcher, R., Leonard, M. A., and West, T. S., J . Chem. SOG., 1958, 2390. Bartkiewicz, S. A., and Robinson, J. W., .Anal. Chim. A d a , 1960, 22, 427. Belcher, R., and West, T. S., Talanta, 1961, 8, 863. Leonard, M. A., Communication at the meeting of the Society for Analytical Chemistry, London, PZbil, R., Coll. Czech. Chem. Comm., 1951, 16/17, 542. Sanderson, I. P., and West, T. S., unpublished work. Diehl, H., and Buchanan, E. B., Talanta, 1958, 1, 76. Vydra, F., and Pfibil, R., Ibid., 1959, 3, 72. Close, R, A., and West, T. S., Ibid., 1960, 5, 221. Herrero-Lancina, M., and West, T. S., Ibid., 1962, in the press. Williams, K. T., and Wilson, J. R., Anal. Chem., 1961, 33, 244. Cabrera, A. M., and West, T. S., unpublished work. Belcher, R., Close, R. A., and West, T. S., Chem. & Ind., 1957, 1647. Chem., 1957, 69, 174; Mikrochim. Acta, 1957, 215. Ibid., 1962, 27, 9. February Sth, 1962. Received March 21st, 1962

ISSN:0003-2654    

DOI:10.1039/AN9628700630

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

August, 19621 JONES AND NEWMAN 637 The Determination of Copper in some Foodstuffs with 2,9-Dimeth yl-l, 1 0-phenan throline BY P. D. JONES AND E. J. NEWMAN (Hopkin & WiZZiarns Ltd., Freshwater Road, ChadwelZ Heath, Essex) 2,9-Dirnethyl-l,lO-phenanthroline has been used for determining copper in foodstuffs, including those containing appreciable amounts of calcium. Results are compared with those obtained with diethylammonium diethyl- dithiocarbamate and zinc dibenzyldithiocarbamate. It is possible to use 2,9-dimethyl-l,lO-phenanthroline for the visual determination of copper and in the selective extraction of copper when the metal is finally determined with a more sensitive reagent. THE work described in this paper was part of a programme of investigations undertaken by the Metallic Impurities in Organic Matter Sub-committee of the Analytical Methods Committee of the Society for Analytical Chemistry into different reagents for determining copper.The Sub-committee is to recommend a method involving use of diethylammonium diethyldithiocarbamate, and the reasons for this recommendation will be given in its report. 2,9-Dimethyl-l,lO-phenanthroline (neocuproine) was first described as a reagent for copper by Smith and McCurdy.1 These and other workers2s3 have shown that the reagent is specific for copper. It has been used for determining small amounts of copper in metals and metal oxides,2 to l1 beverages and water,l2 fuel o i p biological fluids14 and fertilisers.16 No method has come to our notice in which the copper contents of foodstuffs have been determined with neocuproine, although at least one worker12 stated that the reagent should be suitable for this purpose.Excellent methods are available for determining copper in foodstuffs, and most of them involve use of reagents of the dithiocarbamate type, which are more sensitive than neo- cuproine (zinc dibenzyldithiocarbamate, for example , is twice as sensitive). However, in our opinion, the use of neocuproine merits more attention by virtue of its specificity. In this paper we indicate some ways in which it can usefully be applied. EXPERIMENTAL Neocuproine reacts with cuprous copper in the pH range 2 to 9 to form a yellow complex extractable into various organic solvents. We added citrate to the sample solutions and worked in the pH range 5 to 7 to prevent precipitation of calcium.Solutions containing fairly large amounts of calcium are liable to produce precipitates when the ethanolic solution of neocuproine is added, but these precipitates can be re-dissolved by adding more citrate. The cuprous - neocuproine complex was extracted into chloroform as described by Gahler: and the combined extracts were diluted with ethanol. Ethanol is necessary to stabi- lise the coloured complex, but, provided that at least 2 ml are present in a final volume of 25m1, the amount used is not important. (Absolute industrial methylated spirit can be used, but methylated spirit containing pyridine causes low results.) The reagents used in the extraction procedure can be easily purified by treatment with neocuproine and subsequent extraction of any copper complex produced into chloroform.Since the acids and ammonia solution cannot be purified in this way, we used the specially purified reagents available for determining lead in foodstuffs, as their copper contents are somewhat lower than those of analytical-reagent grade materials. By these means, reagent blank values were reduced to negligible figures. Optical-density measurements were made in 1-cm cells with a Beckman DU spectrophotometer. METHOD REAGENTS- all of the grade designated “low in lead.” Acids-Sulphuric acid, sp.gr. 1-84, nitric acid, sp.gr. 1-42, and perchloric acid, sp.gr. 1.54;638 JONES AND NEWMAN: THE DETERMINATION OF COPPER IN SOME [Vol. 87 Ammonia solution, sp.gr. 0~88-“Low in lead.” Ethanol-Absolute industrial methylated spirit, 74” O.P.Neocuproine sohtion-Prepare a 0.1 per cent. w/v solution of the reagent in the ethanol. Store in a copper-free-glass or polythene bottle. Chloroform-Analytical-reagent grade. Hydroxyammonium chloride - sodium citrate solution-Dissolve 25 g of hydroxyammonium chloride and 150 g of sodium citrate (both of analytical-reagent grade) in 500 ml of water in a large separating funnel. Add 10 ml of neocuproine solution, mix, and set aside for 10 minutes. Extract the copper by shaking vigorously with successive 10-ml portions of chloroform until the organic layer is colourless. Discard the chloroform extracts, and store the solution in a copper-free-glass or polythene bottle. Standard copper solution-Dissolve 0.393 g of analytical-reagent grade cupric sulphate in water containing 5ml of sulphuric acid, arid dilute with water to 1 litre. Dilute this solution, which contains 100pg of copper per ml, 10- or 100-fold with water immediately before use.PREPARATION OF CALIBRATION GFUPH- Dilute suitable portions of diluted standard copper solution, covering the range 0 to 150 pg of copper and contained in 100-ml separating funnels, to 20ml with water. To each add 10 ml of hydroxyammonium chloride - sodium citrate solution and 10 ml of neocuproine solution, mixing after each addition, and set aside for 15 minutes. Add 5 ml of chloroform, insert the stopper, shake vigorously for 30 seconds, and allow to separate. Dry the stem of the funnel with a roll of filter-paper, and run the chloroform layer through a small funnel fitted with a glass-wool plug into a 25-ml calibrated flask containing a few millilitres of ethanol; ensure that none of the aqueous layer is transferred to the flask, or turbidity will be produced.Extract with a further 5 ml of cliloroform by shaking for a few seconds, allow to separate, and add the chlorofonn layer to the first extract. Dilute the solution in the flask to the mark with ethanol, and mix. Measure the optical density of this solution at 457 mp in a 1-cm cell against a blank solution prepared by carrying out the procedure on 20ml of water. Construct a graph relating optical density to the number of micrograms of copper present; such a graph is linear and passes through the origin. When this procedure was used, we obtained a molar extinction coefficient for the cuprous -neocuproine complex of about 8500.The solution of the coloured complex is stable and can be kept in clear glassware on the bench for several hours without alteration in optical density. The optical measurement can also be made with an absorptiometer fitted with Ilford No. 601 (violet) or equivalent filters. To avoid transference of any water into the calibrated flask, we used short-stemmed separating funnels (Squibb type) and filled the bores of the stopcocks with chloroform before use. It is advisable to release pressure in the funnel by loosening the stopper and not by turning the stopcock. SUCROSE- We first applied the method to digests from the wet oxidation of sucrose, since the acid remaining after digestion would require neutralisation, and we were interested to find out whether or not blank values would be satisfactorily low and if our results would be adversely affected by the high concentration of salts.(Our attention had been drawn by Mr. J. A. Hawes, of Hopkin & Williams Ltd., to the fact that high concentrations of salts sometimes caused unreliable results in the closely similar determination of iron with 4,7-di- phenyl-1,lO-phenanthroline.) Digestion was carried out on 2-g samples of sucrose by a conventional procedure involving the use of sulphuric and nitric acids. Recovery experiments were made with known added amounts of copper, and in a parallel series of tests copper was determined with zinc dibenzyl- dithiocarbamate. In determining copper with neocuproine, the digest was transferred to a 100-ml separating funnel with water, 10 ml of hydroxyammonium chloride - sodium citrate solution were added, and the pH of the solution was adjusted to between 5 and 7 with ammonia solution.To the cooled solution were then added 1 O m l of neocuproine solution, and the determination was completed as described under “Preparation of calibration graph.’’ APPLICATION OF THE METHODAugust, 19621 FOODSTUFFS WITH 2,9-DIMETHYL-1,1O-PHENANTHROLINE 639 In determining copper with zinc dibenzyldithiocarbamate, the digest was transferred to a separating funnel with sufficient water to give a final concentration of sulphuric acid less than 1.8 N ; the solution was then cooled. Copper was extracted by shaking for 2 minutes with 10 ml of a 0.05 per cent.solution of zinc dibenzyldithiocarbamate in carbon tetrachloride, the optical density of the extract was measured at 435mp, and the copper content of the sample was ascertained by reference to a suitable calibration graph. This procedure is similar to that described by Martens and Githens,ls and we used their method to overcome interference from bismuth in samples to which we added other metals. No modification to the determination with neocuproine was made when other metals were added. The results are shown in Table I. TABLE 1 RESULTS FOR 2-g SAMPLES OF SUCROSE Copper found by- A f > zinc dibenzyl- neocuproine dithiocarbamate Sample method, method, Pg PLg Sucrose . . .. .. .. .. .. 0-8, 0.8 0.8, 1.0 Sucrose plus 25 pg of copper . ... . , 25.8, 25-1 25.9, 25-9 Sucrose plus 25 p g of copper plus other metals* 27.4, 25.8 mercury, lead and bismuth and 200 pg of iron. 24.8, 25.1 * The other metals added were 20 pg each of cobalt, nickel, manganese, arsenic, CALCIUM-CONTAINING FOODSTUFFS- Originally, we determined copper with neocuproine at pH 4 to 5, but, when this method was applied to digests containing appreciable amounts of calcium, a precipitate occasionally formed when the ethanolic solution of neocuproine was added; such precipitation led to low recoveries of added copper. We were able to overcome this difficulty in two ways: (i) by working at a slightly higher pH and adding more citrate and (ii) by using an aqueous solution of neocuproine. (The use of the aqueous neocuproine solution is described below.) The method was first applied to samples prepared by evaporating solutions of calcium hydrogen orthophosphate in dilute sulphuric acid until fumes were evolved and then treating the solutions as described for the determination of copper in digests from sucrose.For a solution derived from 0.215 g of CaHP0,.2H20 (equivalent to 50 mg of calcium), the results of recovery experiments were- Copper added, pg . . .. Nil Copper found, pg . . .. 0.0 and for a solution derived from 0.43 g of the phosphate were- Copper added, pg . . .. Nil 10 Copper found, pg . . .. 0.0 10.0 10 40 9.6; 40.0 (i.e., 100 mg of calcium), the results 20 40 40 19.5 39-0 40.0 Tests on similarly prepared solutions of calcium by the diethylammonium diethyldithio- carbamate method also showed the absence of copper from the calcium phosphate.The neocuproine method was then applied to the determination of copper in cheese, flour and powdered milk. The calcium contents of these substances were determined after dry ashing by Ince and Forster's method1' and found to be, respectively, 0.92, 0.23 and 1.12 per cent. The sample was heated in a Kjeldahl flask with 60 per cent. nitric acid until the initial reaction subsided. The liquid was allowed to cool, and the fat was removed by filtration. After evaporation of the filtrate and washings, sulphuric acid was added, and the digestion was continued in the usual way. Finally, a few millilitres of perchloric acid were added to destroy the last traces of organic matter. After the digest had been allowed to cool, about 20 ml of water were added, and the mixture was heated until fumes were evolved. Copper was then determined with neocuproine as described for samples of sucrose. The results, together with those obtained on flour by the diethylammonium diethyldithiocarbamate640 JONES AND NEWMAN: THE DETERMINATION OF COPPER I N SOME [Vol.87 method, are shown in Table 11. For determination with diethylammonium diethyldithio- carbarnate, copper was extracted from ammoniacal solution in presence of ethylenediamine- tetra-acetate and citrate by 15 ml of a 0.1 per cent. solution of the reagent in carbon tetra- chloride; the optical density of the extract was measured at 435 mp, and the copper content of the sample was calculated by reference to a suitable calibration graph.TABLE I1 RESULTS FOR COPPER IN POWDERED MILK, CHEESE AND FLOUR Copper found by- h I Weight of Copper ZneocuIjr oine diethylammonium diethyl- sample, added, method, dithiocarbamate method, g Pg CLg: Clg 3.5 43.5, 43.7, 42.8 - 11.3 - 4.5, 4.7 - 10.5, 9.0 9.0, 9-3 Samples of powdered milk- - 3 { 2 10 { 2 47-6, 51.6 - 10 {2 45.2, 45.1 - Samples of cheese- Samples of $ 0 ~ 4 - - 19.0, 18.8 48-5, 49.2, 47.6 48.0, 48.0 5 { Y; 40 The digests from milk and cheese gave a precipitate when diluted with water; for these samples, after the addition of hydroxyammonium chloride - sodium citrate solution, instead of making the usual adjustment of pH, ammonia solution was added dropwise, with shaking, until the solutions were clear. USE OF AQUEOUS NEOCUPROJNE REAGENT Neocuproine is soluble in dilute aqueous solutions of mineral acids.An aqueous solution of the reagent was prepared by dissolving 0.1 g' of neocuproine in 10 ml of 0.1 N hydrochloric acid and diluting to 100ml with distilled water. This solution was used for determining copper in solutions containing calcium sulphate and calcium hydrogen orthophosphate and also in digests from the foodstuffs. The procedure was otherwise identical to that described above, and the results are shown in Table 111. TABLE I11 COPPER FOUND WITH USE OF A~UEOUS NEOCUPROINE SOLUTION Sample Weight of sample, Copper added, g 20 P, 40 CaHP0,.2Hz0 . . 0.215 (equivalent to 50 mg of calcium) CaHP0,.2Hz0 . . 0.43 (equivalent to 100 mg of calcium) 50 mg of calcium) {Ti Ca,SO,.2HB0 . . . . 0.218 (equivalent to Flour ... . .. 6 Powdered milk . . Cheese .. .. 3 10 Copper found, CLg 0.0 9.1 20.0 40.6, 41.0 0.0 10.2 18.6 40.2, 41.0 0.0 25-0, 27.0 9.0 { Y; 45.3, 49.5 { 3.5 43.1, 43.5, 42.7 4.5, 4.5 24.6, 25.0 42.2, 43.4, 44-1 When an ethanolic solution of neocuproine is added to sample solutions a turbidity forms, It should therefore be possible but this is not observed when the aqueous reagent is used.August, 19621 FOODSTUFFS WITH ~,~-DIMETHYL-~,~~-PHENANTHROLINE 641 to use the aqueous reagent for determining copper, without extraction, by spectrophotometric measurement or by visual comparison with suitable standards. The smallest amount of copper that can be detected in 50 ml of solution in a Nessler cylinder is 0.5 pg. U S E OF NEOCUPROINE AS A SELECTIVE EXTRACTANT We have examined the possibility of using neocuproine and zinc dibenzyldithiocarbamate successively in determinations of copper to take advantage of the specificity of neocuproine and the sensitivity of zinc dibenzyldithiocarbamate.The procedure described below was used to determine the recovery of 25 pg of copper in presence of 100 pg each of cobalt, nickel, iron and bismuth. To about 20 ml of sample solution add 10 ml of hydroxyammonium chloride - sodium citrate solution, and adjust the pH to between 5 and 7 , if necessary, with ammonia solution. Add 10ml of neocuproine solution, mix, and set aside for 15 minutes. Add 5 ml of chloroform, and extract the copper by shaking for 30 seconds; then extract again by shaking for a few seconds with a further 5ml of chloroform.To the combined chloroform extracts add 50 ml of approximately N sulphuric acid and 5 or 6 drops of 0.1 N potassium permanganate. (This oxidises the cuprous copper in the complex to the cupric state and thereby destroys the complex.) Shake vigorously for 30 seconds, allow to separate, and discard the chloroform layer. To the aqueous layer add 10 ml of a 0.05 per cent. solution of zinc dibenzyldithiocarbamate in carbon tetrachloride, and shake vigorously for 2 minutes. Allow to separate, measure the optical density of the organic layer at 435 mp, and calculate the copper content of the sample by reference to the zinc dibenzyldithiocarbamate calibration graph. A determination by this procedure took less than 30 minutes, and the recoveries of 25 pg of copper in presence of other metals as listed above were 24.7,24-7 and 24.4 pg.Its feasibility having been demonstrated, we did not consider it necessary to investigate this technique further. INTERFERENCES Neocuproine is specific for copper; Smith and McCurdyl found that no cations other than cuprous copper formed an extractable coloured complex with the reagent. Luke and Campbell2 tested neocuproine on fifty-six metals, and Hibbits, Davis and Menke3 tested it on sixty-eight elements, of which only gold caused some interference. As well as the results reported above that were obtained in the presence of other metals, we have found that the optical density of a solution containing 40 pg of copper was unaffected by the presence in the solution of 400 pg each of ironI1, chromiumI1, nickelI1, bismuthII1, zincI1, tinII, mercuryI1, antimony111 and arsenicII1.The anions nitrate, chloride, perchlorate, sulphate, tetraborate, acetate and tartrate do not interfere with the test. Sulphide causes low results, and in presence of large amounts of phosphate or pyrophosphate two or three extractions may be necessary to remove all the copper from the aqueous solution. CONCLUSIONS It has been shown that 2,9-dimethyl-1 ,lo-phenanthroline is a satisfactory reagent for determining copper in digests from foodstuffs, including those containing sufficient calcium to produce a precipitate when the digest is diluted with water. Although the reagent is less sensitive than the dithiocarbamates, it has the advantage of being specific for copper, and this specificity can be used to separate copper before determining it by other methods.It is possible to develop visual methods for determining copper with neocuproine in aqueous solutions. We thank the Directors of Hopkin and Williams Ltd. and the Analytical Methods Committee of the Society for Analytical Chemistry for permission to publish this paper. REFEREMCES 1. 2. 3. 4. Gahler, A. R., Anal. Chem., 1954, 26, 577. Smith, G. F., and McCurdy, W. H., jun., Anal. Chem., 1952, 24, 371. Luke, C. L., and Campbell, M. E., Ibid., 1953, 25, 1588. Hibbits, J. O., Davis, W. F., and Menke, M. R., Talanta, 1960, 4, 101,642 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. JONES AND NEWMAN [Vol. 87 Crawley, R. H. A., Anal. Chinz. Acta, 1955, 13, 373. Fulton, J. W., and Hastings, J., Anal. Chem., 1956, 28, 174. “A.S.T.M. Methods of Chemical Analysis of Metals,” American Society for Testing Materials, Baba, H., Japan Analyst, 1956, 5, 631; Anal. Abstr., 1957, 4, 2929. Downey, T. A., Plating, 1957, 44, 383. Ishihara, Y., and Taguchi, Y., Japan Analyst, 1957, 6, 588; Anal. Abstr., 1958, 5, 2525. Frank, A. J., Goulston, A. B., and Deacutis, A. A., Anal. Chem., 1957, 29, 750. Sallavo, K. R., Suomen Kemistilehti, 1954, 27B, 46; Chem. Abstr., 1954, 48, 14008 G. Zall, D. M., McMichael, R. E., and Fisher, 1). W., Anal. Chem., 1957, 29, 88. Zak, B., and Ressler, N., Ibid., 1956, 28, 1158. Webb, H. J., and Nance, L. E., J . Ass. Off. Agric. Chem., 1960, 43, 506. Martens, R. I., and Githens, R. E., Anal. Chem., 1952, 24, 991. Ince, A. D., and Forster, W. A., Analyst, 1960, 85, 608. Philadelphia, 1956, p. 223. Received Februuvy 27th. 1962

ISSN:0003-2654    

DOI:10.1039/AN9628700637

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

August, 19621 BATES, ROWLANDS AND HARRIS 643 The Removal of Plant Extractives interfering in the Determination of Malathion Residues in Barley and Rice Bran BY ANGELA N. BATES, D. G. ROWLANDS AND A. H. HARRIS* (Agricultural Research Council, Pest Infestation Laboratory, Slough, Bucks.) Methods are described for the removal from extracts of barley and rice bran of substances interfering with the colorimetric determination of malathion residues on these commodities. The interfering materials are removed by chromatography of the extracts on acid-washed alumina (for barley) or fuller’s earth (for rice bran). The insecticide is eluted from both types of adsorbent by acetonitrile. THE method described by Norris, Easter, Fuller and Kuchar,l which has been submitted to collaborative study,2 for determining residues of malathion, S-1,Z-di(ethoxycarbony1)ethyl 00-dimethyl phosphorothiolothionate, has been found to give satisfactory results when applied to the determination of this insecticide in many kinds of plant material without purification of the extracts before their analysis.However, Norris and Kucha3 reported that an acidic substance, probably gossypol, was extracted from cottonseed and caused high blank values and low recoveries of insecticide when extracts of this commodity were analysed for malathion. These workers, following the example of Kolbezen and Reynolds: used chromato- graphy on acid-washed alumina to remove the interfering plant extractives. They eluted the insecticide from the column of adsorbent with acetonitrile and subjected this solution to the usual analytical procedure.In our investigation, analysis of samples of barley and rice bran treated with malathion indicated that both these commodities yielded extractives interfering with determination of the insecticide by Norris, Easter, Fuller and Kuchar’s method.1 Recoveries of malathion were erratic when extracts of barley were analysed without preliminary treatment ; chromato- graphy on acid-washed alumina and subsequent elution of malathion by acetonitrile removed the interfering barley extractives. With samples of rice bran, no malathion was detected when as much as 25 p.p.m. had been added, and separation of the insecticide from rice-bran extractives could not be effected on acid-washed alumina ; however, chromatography on fuller’s earth was found to be satisfactory.EXPERIMENTAL DETERMINATION OF MALATHION RESIDUES IN BARLEY- Samples of barley mixed with flour-based malathion dust to give concentrations of insecticide ranging from 4 to 32 p.p.m. were extracted with carbon tetrachloride, and the extracts were analysed, without preliminary purification, by Norris, Easter, Fuller and Kuchar’s method.1 (This method is based on alkaline hydrolysis of the insecticide to di- methyldithiophosphoric acid, which yields with cupric ions a yellow chelate compound soluble in carbon tetrachloride.) It was noted that the volume of 6 N hydrochloric acid required to neutralise the alkaline aqueous layer after the hydrolytic stage of the determina- tion was considerably smaller than that needed when barley extractives were not present, and, as shown in Table I, recoveries of malathion were variable.To eliminate possible losses of malathion during extraction, extracts of insecticide-free barley were prepared, and malathion dissolved in carbon tetrachloride was added to these immediately before the determination ; recoveries of malathion were again inconsistent (see Table I). The accuracy of the determinations was not improved by increasing the concentration or the volume of alkali used for hydrolysing the insecticide. As Norris and KucharS had reported the successful use of acid-washed alumina in removing cottonseed extractives that interfered with the determination of malathion, chromatography on this adsorbent was investigated as a means of eliminating the barley extractives inhibiting Road, London, W.C.1.* Present address: Tropical Stored Produce Liaison at Tropical Products Institute, 56-62 Grays Inn644 BATES, ROWLANDS AND HARRIS: REMOVAL OF PLANT EXTRACTIVES [Vol. 87 TABLE I RECOVERY OF MALATHION FROM BARLEY WITHOUT PURIFICATION OF EXTRACT BEFORE DETERMINATION The insecticide was added as a flour-based dust to the barley before extraction Weight of barley, g 10 20 5 10 20 20 Malathion added, ELg 80 80 160 160 160 160t Malathion found,* Recovery of insecticide, llg % 35, 27 44, 34 38, 89 48, 111 176, 83 110, 52 71,185 44,116 102, 44 64, 28 118, 77 74, 48 * Not corrected for the apparent malathion found in insecticide-free barley, as the optical densities derived from the untreated cereal were below the limit of the cali- bration graph.t Insecticide in solution was added to carbon tetrachloride extracts of the barley. hydrolysis of the insecticide. Hexane was used instead of carbon tetrachloride as extractant, and the extracts were applied to columns of acid-washed alumina containing 2.5 per cent. of water. No malathion was removed from alumina of this activity by hexane; diethyl ether and chloroform removed only small fractions of the amount of insecticide applied to the column, and neither ethanol nor methanol eluted all the malathion present. It was finally shown that the insecticide was completely eluted by acetonitrile, the interfering barley extractives being retained on the column. The acetonitrile extracts were subjected to the usual analytical procedure, and it was noted that the volume of 6 N hydrochloric acid required for neutralisation during each analysis was equal to the volume of 6 N alkali used to effect hydrolysis.The recoveries of malathion from samples of barley treated before extraction with a solution of the insecticide in hexane - carbon tetrachloride mixture (9 -+ 1, v/v) when pre- liminary purification on acid-washed alumina was used are shown in Table 11. When aliquots of a standard solution of malathion were subjected to the chromatographic procedure before hydrolysis and colour development, the resulting optical densities were approximately 8 per cent. lower than those produced when chromatography was omitted. The recovery figures given in Table I1 were derived from comparison with a standard graph plotted from the optical densities obtained after chromatographic treatment.TABLE I1 RECOVERY OF MALATHION FROM BARLEY AFTER CHROMATOGRAPHY OF EXTRACT ON ACID-WASHED ALUMINA Weight of barley, Malathion added, Malathion found,* Recovery of insecticide, g ELQ ELg % 25 49 45, 49, 48 92,100, 98 50 49 46, 49, 49 94,100,100 25 148 139,143, 145 94, 97, 98 50 246 224,230,233 91, 94, 95 25 246 211,220,225 86, 89, 91 * Not corrected for the apparent malathion found in insecticide-free barley, as the optical densities derived from the untreated cereal were below the limit of the cali- bration graph. DETERMINATION OF MALATHION RESIDUES IN RICE BRAN- Malathion dissolved in carbon tetrachloride was added to samples of rice bran, and each sample was immediately extracted with carbon tetrachloride in a Soxhlet apparatus ; the extracts were analysed as bef0re.l When malathion was added at levels up to 25 p.p.m., the optical densities derived from analysis of the carbon tetrachloride extracts did not exceed those obtained after analysis of extracts of insecticide-free rice bran.When malathion was added to the bran at the 50 to 100 p.p.m. level, small amounts of dimethyldithiophosphoric acid were detected, but these corresponded to less than one-tenth of the insecticide presentAugust , 19621 645 (see Table 111). In all these tests, the addition of phenolphthalein to the solution at the conclusion of the alkaline-hydrolysis stage produced no red colour, indicating that rice-bran extractives had neutralised all the alkali added.Some samples were then extracted at room temperature by tumbling a slurry of rice bran in distilled water with carbon tetrachloride for 4 hours. The interfering plant material was still extracted under these conditions, and recovery of malathion was again zero. Samples of rice bran freshly prepared from whole rice immediately before addition of insecticide, and rice bran that had been fumigated some weeks before treatment with malathion, were investigated, but on hot extraction with carbon tetrachloride these samples also gave solutions containing the interfering extractives. The results of these experiments are shown in Table 111. INTERFERING IN THE DETERMINATION OF MALATHION RESIDUES TABLE I11 RECOVERY OF MALATHION FROM RICE BRAN WITHOUT PURIFICATION OF EXTRACT BEFORE DETERMINATION Weight of rice bran, Malathion added, Malathion found, g lug lug 25 100 Nil 25* 100 Nil 25* 200 Nil 25 200 Nil 50 250 Nil 250 Nil 270 Nil 50t 100: 25 625 Nil 25 1250 20,20 25 1875 78 25 2500 55, 63 * Freshly prepared rice bran.t Rice bran fumigated before storage. : Extraction carried out a t room temperature. Attempts were then made to remove the interfering materials by chromatography on acid-washed alumina and subsequent elution with acetonitrile, which had proved successful with extracts of barley. As can be seen from Table IV, considerable improvement was effected by this procedure, but the recovery of malathion was still low and decreased with increase in the size of the rice-bran sample extracted. Chromatography of malathion in the absence of plant extractives on neutral alumina of activity Brockmann grade I1 suggested that some break-down of the insecticide occurred on the column (see Table IV).The behaviour of malathion on neutral alumina of low activity was then investigated, as Laws and Webley5 had used this adsorbent in determining residues of malathion on fruit and vegetables. Although the insecticide was stable on such neutral alumina, this adsorbent did not retain the interfering extractives from rice bran. The volume of 6 N hydrochloric acid required to neutralise the alkali added was considerably less than 1 ml for each eluate, and recoveries of malathion were poor (see Table IV). As it appeared that the extractives from rice bran preventing colorimetric determination of residues of malathion on this material could not be separated from the insecticide by chromatography on alumina, an alternative adsorbent of high activity was sought.A sample of fuller’s earth, purified for chromatography, was tested with standard solutions of dye in benzene - light petroleum mixture and found to show activity corresponding to grade I1 on the Brockmann scale for alumina. To shorten the time taken for chromatography, it was convenient to use only l o g of this very finely divided material in each column, although 2Og of adsorbent were used in each column of alumina; moreover, recovery of malathion from the column was improved by the reduction in the weight of fuller’s earth used. When malathion dissolved in hexane - carbon tetrachloride mixture was applied to columns of fuller’s earth, no insecticide was eluted by either of these solvents, and about 10 per cent.of the weight applied was not recovered by elution with ethanol. Acetonitrile, however, used in rather larger amount than had been found necessary for the alumina columns, removed virtually all the applied malathion from fuller’s earth. It was essential to regulate carefully the suction applied to the base of each column, as, if this was too vigorous, the lower part of the adsorbent was drawn into a tightly packed mass impenetrable to the eluting646 BATES, ROWLANDS AND HARRIS: REMOVAL OF PLANT EXTRACTIVES [VOl. 87 TABLE IV RECOVERY OF MALATHION FROM HEXANE EXTRACTS OF RICE BRAN AFTER CHROMATOGRAPHY ON ALUMINA Weight of rice bran, g 25 Zt 25: 50 Nil Nil Nil 25 Nl1 25 25t 25§ Malathion added, Pf? 100 100 200 245 245 245 245 245 245 245 245 245 245 Adsorbent used * Eluting agent J Hexane - ether (3 + 1, v/v) Acetonitrile C Hexane - ether (1 + 1, v/v) C Acetonitrile Hexane Hexane - ether (3 + 1, v/v) Carbon tetrachloride Hexane - ether (17 + 3, v/v) Hexane Hexane - ether (17 + 3, v/v) Hexane c { = { C D D Malathion found, Pg 68,63 75,85 142 132 85,100 11 198 195 245 128 39 97 229 157, 165 Nil, Nil, 172 Recovery of insecticide, 68,63 75,85 71 54 35,41 4 81 80 100 52 16 40 93 64,67 Nil, Nil 70 % * The adsorbents used were : A, acid alumina (Brockmann activity grade II/III) ; B, neutral alumina (activity grade 11); C, neutral alumina (activity grade IV); D, neutral alumina (activity grade V).t Hexane extract shaken with acetonitrile, and the latter solution placed on the column.Hexane extract shaken with aqueous methanol and !dried with anhydrous sodium sulphate before transfer to column. 4 Bran sample extracted with carbon tetrachloride instead of hexane. solvents. It was therefore not convenient to drain from each column the hexane with which the rice bran was extracted before the acetonitrile was introduced; as the presence of hexane in the eluate to be analysed was undesirable, carbon tetrachloride was passed through each column as soon as the level of the hexane extract had fallen to the top of the adsorbent. The acetonitrile was introduced when all the carbon tetrachloride had passed into the fuller’s earth. When hexane extracts of rice bran containing malathion were applied to columns of fuller’s earth, the extracts, which were rather viscous after concentration, passed through the columns of adsorbent extremely slowly.When the fuller’s earth was mixed with a fdter-aid material, such as kieselguhr or Celite 545, however, the rate at which the extract and the eluting solvents passed through the column was considerably faster, and the adsorptive TABLE V RECOVERY OF MALATHION FROM RICE BRAN AFTER CHROMATOGRAPHY OF EXTRACT ON FULLER’S EARTH - KIESELGUHR The insecticide in solution was added to the samples of rice bran immediately before their extraction with hexane in a Soxhlet apparatus Weight of rice bran, g 25 50 Et 2 50 25t 50 Malathion added, Icg 49 49 98 148 148 245 245 247 247 Malathion found,* Pg 54, 62 50 90, 95 148,146 125,124 258,233 205,209 250,247 200,206 Recovery of insecticide, % 110,104 102 93, 97 100, 99 86, 84 105, 95 84, 85 101,100 81, 84 * Not corrected for the apparent malathion found in insecticide-free rice bran, as the optical densities t Columns of fuller’s earth mixed with Celite 546 were used with extracts of these samples.derived from the untreated cereal were below the limit of the calibration graph.August, 19621 INTERFERING IN THE DETERMINATION OF MALATHION RESIDUES 647 properties of the fuller’s earth in relation to malathion or the rice-bran extractives were not altered. It appeared that the rice-bran extractives inhibiting hydrolysis of malathion were largely removed from the columns of fuller’s earth by the hexane and carbon tetrachloride, as these eluates completely neutralised all the alkali added during their analysis.The acetonitrile eluates, which contained the malathion, were consequently free from the inter- fering material; in analysing these, the usual amount of 6 N hydrochloric acid was required for neutralisation. The recovery of malathion from such eluates is shown in Table V. The optical density in each determination was referred to a standard graph constructed by carrying aliquots of a malathion solution through the chromatographic treatment before carrying out the usual analytical procedure. The slope of the graph so plotted was approximately 4 per cent. lower than that of a calibration graph constructed in the usual manner. METHOD APPARATUS- Soxhlet extraction apparatus. Chromatographic tubes-Glass tubes, 1 to 2 cm in diameter and 65 cm long, tapered at S+ectvo+hotometer or absor+tiometer-Capable of measuring optical densities at 41 8 mp.the lower end to retain a cotton-wool plug or fitted with a coarse sintered-glass plate. REAGENTS- Hexane. Acetonibrile-Fractionate, and collect the fraction boiling between 79” and 81” C. Carbon tetrachzoride-Analytical-reagent grade. Acid-washed alumina-(Material obtained from L. Light & Co. Ltd., Poyle Trading Estate, Colnbrook, Bucks., was used.) Add 2.5 per cent. w/w of water to the alumina con- tained in a screw-topped jar, shake well, and set aside overnight. The activity of the product is intermediate between Brockmann grades I1 and 111. FuZZer’s earth-Material graded as “purified M.F.C.” (obtainable from Hopkin and Wil- liams Ltd., Chadwell Heath, Essex) was used. The activity of this adsorbent corresponded to grade I1 on the Brockmann scale for alumina. KieseZguhr-Material graded as “purified M.F.C.” or Celite 545 was used. Other reagents-As listed by Norris, Easter, Fuller and Kuchar-l - PREPARATION OF EXTRACTS- Extract 25 to 50 g of barley or rice bran with hexane in a Soxhlet extractor for 8 hours; (grind barley before extraction). For recovery experiments, add a suitable amount of a solution of premium-grade (about 95 per cent.) malathion in hexane - carbon tetrachloride mixture (9 + 1, v/v) to the contents of the Soxhlet thimble before extraction. After the 8-hour period, adjust the volume of hexane in the flask to about 100 ml for samples of barley or about 30 ml for rice bran by distilling the excess of solvent up into the Soxhlet thimble or by concentrating the extract at 40” C, evaporation being assisted by a stream of dry air.PROCEDURE FOR DETERMINING MALATHION IN EXTRACTS OF BARLEY- Cover the surface of the alumina with 2 g of anhydrous sodium sulphate, and fill the column with hexane. Transfer the hexane extract of barley to the tube with a few millilitres of hexane, and apply suction to draw the solvent through the column at approximately 20 ml per minute. When all the hexane has been drained from the column, change the receiver, and elute with 26 ml of redistilled acetonitrile; draw this through the column by suction. (A fresh column of alumina should be used for each extract.) Transfer the acetonitrile eluate to a dry separat- ing funnel with 50ml of carbon tetrachloride used to rinse the receiver from the column.Add 75 ml of a 9 per cent. solution of sodium sulphate in water, 5 ml of concentrated hydro- chloric acid and 75 ml of distilled water to the contents of the funnel, and shake for I minute. Filter the carbon tetrachloride layer through a fluted filter-paper into a second dry separating funnel, and re-extract the aqueous layer in the first funnel with a second 50-ml portion of carbon tetrachloride by shaking the two phases for 1 minute. Filter the second carbon tetrachloride extract into the second separating funnel. (Do not permit any of the aqueous Prepare a column of 20 g of the acid-washed alumina in a chromatographic tube.648 BATES, ROWLANDS AND HARRIS: REMOVAL OF PLANT EXTRACTIVES [VOl.87 solution to run on to the filter-paper during either filtration.) Complete the determination as described by Norris, Easter, Fuller and Kuchar.1 Plot a standard graph from the results obtained by subjecting aliquots of a solution of premium-grade malathion, containing 25 to 250 pg of insecticide, to the chromatographic and analytical procedures described above. PROCEDURE FOR DETERMINING MALATHION IN EXTRACTS OF RICE BRAN- Mix 10 g of fuller’s earth with 5 g of kieselguhr or Celite 545, make into a slurry with hexane, and pour into the chromatographic tube. Drain the solvent from the column by applying suction so adjusted that the solvent passes through the column a t approximately 1 ml per 20 seconds. When the level of hexane has fallen almost to the top of the column, disconnect the suction, and cover the surface of the adsorbent with 2 g of anhydrous sodium sulphate.Transfer the hexane extract of rice bran to the tube with a few millilitres of hexane, and draw the solvent through the column by applying the same amount of suction as before. When the level of the liquid has fallen almost to the top of the adsorbent, add a further 25 ml of hexane, and draw this through the column. When all the hexane has passed into the adsorbent, add 25 ml of carbon tetrachloride, and, when the level of this solvent has reached the surface of the column, change the receiver, pour 40 ml of redistilled acetonitrile into the tube, and draw this through under suction. Transfer the acetonitrile eluate (con- taining some carbon tetrachloride) to a separating funnel with 50 ml of carbon tetrachloride, and determine its malathion content as described above for extracts of barley.Use a fresh column of adsorbent for each extract. Prepare a standard graph from the results obtained by subjecting aliquots of a solution of premium-grade malathion, containing 25 to 250 pg of insecticide, to this chromatographic procedure and then analysing them as described previous1y.l s3 CONCLUSIONS The results in Table I for the recovery of malathion from barley indicate that the colorimetric method is unreliable for determining residues on this cereal unless the interfering plant extractives are separated from the insecticide before the hydrolytic and colour-develop- ment stages of the determination.The fact that this separation can be achieved by chromato- graphy of the barley extracts on acid-washed alumina of high activity is shown by the results in Table 11. The insecticide is stable on alumina of activity grade I1 to I11 if the adsorbent is acid, although it appears to break clown on neutral alumina of similar activity. It can be totally eluted with acetonitrile, and the interfering plant extractives are absent from the eluate. Analysis, without preliminary purification, of rice-bran extracts containing known amounts of malathion, the results of which are shown in Table 111, suggests that only when milligram amounts of the insecticide are present is any dimethyldithiophosphoric acid produced by this method.Extractives from the rice bran obviously react with all the alkali added before any of the insecticide is hydrolysed. Only partial separation of the extractives and the insecticide was achieved by chromatography on various grades of both acid and neutral alumina, as can be seen from the low recoveries listed in Table IV. It appears that malathion is displaced from the alumina by the rice-bran extractives, as the insecticide appears in the hexane eluate when the plant material is present, whereas pure malathion is not eluted by hexane from this adsorbent even at low a ~ t i v i t y . ~ Adtonitrile solutions of the insecticide completely free from the bran extractives inhibiting hydrolysis of malathion can be obtained by chromatography of the extracts containing the insecticide on fuller’s earth. Recovery of malathion from 25-g samples of rice bran always exceeded 90 per cent.when the chromatographic technique was used (see Table V); the decreased recoveries (between 80 and 90 per cent.) for 50-g samples of bran was thought to indicate that extraction of insecticide from these larger samples was less efficient rather than that interfering plant material was present in the extracts, as the volume of 6 N hydrochloric acid required for neutralisation was always equal to the volume of added 6 N alkali. It was found to be essential to use freshly distilled acetonitrile for both types of chromato- graphic purification; an old sample of this solvent used in conjunction with a column of alumina led to extremely low recoveries of malathion from barley.No trouble has been observed with acetonitrile used up to 14 days after redistillation, but the solvent has not been used after a longer period than this without refractionation.August, 19621 INTERFERING IN THE DETERMINATION OF MALATHION RESIDUES 649 It appears from these observations on determinations of malathion on barley and rice bran by Norris, Easter, Fuller and Kuchar’sl method that an important indication of the presence in plant extracts of substances likely to invalidate the results is given by a decrease in the volume of 6 N hydrochloric acid needed for neutralisation of the solution for subsequent analysis. It is proposed to investigate a wider range of cereals and oil seeds, on which malathion might be used for insect control, for the presence of extractives inhibiting the hydrolysis of this insecticide and necessitating preliminary treatment of the extracts before colorimetric det emination of malathion. This investigation formed part of the research programme of the Pest Infestation Laboratory. We thank Dr. E. A. Parkin for his interest in this work and are grateful to Dr. D. M. Langbridge for recommending the use of acid-washed alumina and for supplying us with a sample of this material. Rice bran was obtained for us by Mr. J. A. McFarlane from J. Bibby & Sons Ltd. REFERENCES 1. 2. 3. 4. 5. Norris, M. V., Easter, E. W., Fuller, L. T., and Kuchar, E. J., J . Agric. Food Chem., 1958, 6, 111. Malathion Panel, Analyst, 1960, 85, 915. Norris. M. V., and Kuchar, E. J., J . Agric. Food Chem., 1959, 7, 488. Kolbezen, M. J., and Reynolds, H. T., Ibid., 1956, 4, 522. Laws, E. Q., and Webley, D. J., Analyst, 1961, 86, 249. Received February 15th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9628700643

出版商:RSC    

年代:1962

数据来源: RSC

摘要:

650 HOFFMAN SPECTROPHOTOMETRIC DETERMINATION [Vol. 87 Spectrophotometric Determination Fractions of Nickel in Wheat BY I. HOFFMAN (Analytical Chemistry Research Service, Canada Department of Agriculture, Ottawa, Ontario, Canada) A method is described for determining nickel in bran, feed and flour; solvent extraction is avoided, and the method is suitable for multiple deter- minations. Middleton and Stuckey’s digestion procedure is included, and the selectivity of dimethylglyoxime for nickel is utilised. A compact apparatus for filtration and the stability of the final colour contribute to the ease and reliability of the method. DESTRUCTIVE rust epidemics in grain-growing areas have occasioned interest in chemical means of control. New rust-resistant varieties of wheat are often attacked by new races of rust within a few years of their development. Wheat can be protected against infection from the seedling stage to maturity by frequent applications of such compounds as zineb (zinc ethylenebisdithiocarbamate) or sulphur, but this involves a commitment of considerable time and expense without proof of real necessity in a given crop year.Obvious advantage would be offered by a rust-eradicative treatment to be used only when rust is noticeable on the plants. In greenhouse and field tests, low concentrations of nickel salts displayed fungi- cidal activity against cereal rusts through both protective and eradicative action.l,2*3,4 Application of these salts to wheat increased acreage yields, and calculations showed that growers would receive greater net returns after deducting the cost of spraying6 These findings have stimulated the interest of both plant pathologists and chemical companies engaged in the manufacture and sale of fungicidal chemicals.However, accurate information in terms of the accumulation and distribution of nickel in the grain was required in order to afford a sound basis for judging any possible toxic hazard. It was foreseen that a con- siderable number of samples would be involved in such a survey; further, this number would be trebled by the necessity for milling the wheat into bran, feed and flour fractions for distribution studies . Of the analwcal procedures available it was considered that either Forster’s method6 or that described by Alexander, Godar and Linde’ might be adapted for our purpose.These procedures involve multiple solvent extraction of the nickel complex, and it was soon evident that the great amount of time spent in manipulating separating funnels would be a serious drawback to either method. However, preliminary work showed that Alexander, Godar and Linde’s method gave reliable results, whereas instability of the colour made Forster’s method unsatisfactory for determining nickel in wheat fractions. Other features detracting from the suitability of these methods were the potential danger introduced by the use of perchloric acid and the constant and close attention required during the preliminary wet digestion. A method described by Middleton and Stuckey* for destroying organic matter had im- portant advantages over other methods. The main reagent used was nitric acid, and only relatively small amounts were needed; the only attention required was the intermittent addition of fresh acid, and there were no critical stages in the procedure.It seemed likely that this method of preparing the sample could be extended to the determination of nickel. The procedure developed includes this more efficient digestion, retains the selectivity of dimethylglyoxime for nickel and offers the convenience of a stable spectrophotometric end-point . METHOD APPARATUS- Filtration a$fiaratus-The assembly shown in Fig. 1 was devised to facilitate the handling of multiple samples. The manifold at the back consists of a series of T-joints connected by tubing. Sfiectrofihotometer-A Beckman DU instrument equipped with matched 1-cm silica cells was used.Glass s+atzda-A glass handle was attached to a 0.5-inch section taken from the wall: of glass tubing that just fitted inside a 15-ml sinter funnel.650 HOFFMAN SPECTROPHOTOMETRIC DETERMINATION [Vol. 87 Spectrophotometric Determination Fractions of Nickel in Wheat BY I. HOFFMAN (Analytical Chemistry Research Service, Canada Department of Agriculture, Ottawa, Ontario, Canada) A method is described for determining nickel in bran, feed and flour; solvent extraction is avoided, and the method is suitable for multiple deter- minations. Middleton and Stuckey’s digestion procedure is included, and the selectivity of dimethylglyoxime for nickel is utilised. A compact apparatus for filtration and the stability of the final colour contribute to the ease and reliability of the method.DESTRUCTIVE rust epidemics in grain-growing areas have occasioned interest in chemical means of control. New rust-resistant varieties of wheat are often attacked by new races of rust within a few years of their development. Wheat can be protected against infection from the seedling stage to maturity by frequent applications of such compounds as zineb (zinc ethylenebisdithiocarbamate) or sulphur, but this involves a commitment of considerable time and expense without proof of real necessity in a given crop year. Obvious advantage would be offered by a rust-eradicative treatment to be used only when rust is noticeable on the plants. In greenhouse and field tests, low concentrations of nickel salts displayed fungi- cidal activity against cereal rusts through both protective and eradicative action.l,2*3,4 Application of these salts to wheat increased acreage yields, and calculations showed that growers would receive greater net returns after deducting the cost of spraying6 These findings have stimulated the interest of both plant pathologists and chemical companies engaged in the manufacture and sale of fungicidal chemicals.However, accurate information in terms of the accumulation and distribution of nickel in the grain was required in order to afford a sound basis for judging any possible toxic hazard. It was foreseen that a con- siderable number of samples would be involved in such a survey; further, this number would be trebled by the necessity for milling the wheat into bran, feed and flour fractions for distribution studies .Of the analwcal procedures available it was considered that either Forster’s method6 or that described by Alexander, Godar and Linde’ might be adapted for our purpose. These procedures involve multiple solvent extraction of the nickel complex, and it was soon evident that the great amount of time spent in manipulating separating funnels would be a serious drawback to either method. However, preliminary work showed that Alexander, Godar and Linde’s method gave reliable results, whereas instability of the colour made Forster’s method unsatisfactory for determining nickel in wheat fractions. Other features detracting from the suitability of these methods were the potential danger introduced by the use of perchloric acid and the constant and close attention required during the preliminary wet digestion.A method described by Middleton and Stuckey* for destroying organic matter had im- portant advantages over other methods. The main reagent used was nitric acid, and only relatively small amounts were needed; the only attention required was the intermittent addition of fresh acid, and there were no critical stages in the procedure. It seemed likely that this method of preparing the sample could be extended to the determination of nickel. The procedure developed includes this more efficient digestion, retains the selectivity of dimethylglyoxime for nickel and offers the convenience of a stable spectrophotometric end-point . METHOD APPARATUS- Filtration a$fiaratus-The assembly shown in Fig.1 was devised to facilitate the handling of multiple samples. The manifold at the back consists of a series of T-joints connected by tubing. Sfiectrofihotometer-A Beckman DU instrument equipped with matched 1-cm silica cells was used. Glass s+atzda-A glass handle was attached to a 0.5-inch section taken from the wall: of glass tubing that just fitted inside a 15-ml sinter funnel.August, 19621 OF NICKEL IN WHEAT FRACTIONS 651 REAGENTS- Nitric - sul@huric acid mixture-Mix 95 ml of nitric acid, sp.gr. 1.42, with 5 ml of sulphuric acid, sp.gr. 1-84. Nitric acid, fuming, sp.gr. 1.50. Sodium tartrate dihydrate solution, 20 per cent. w/v, aqaeous. Dimethylglyoxime solution 1 per cent. w/v, in ethanol. Ammonia solution, sp.gr.0.90. Bromine water, saturated. Potassium persulphate solution, saturated, aqueous. Standard nickel solution, 1 mg per md-Dissolve 6.7301 g of ammonium nickel sulphate, (NH,),S04-NiS04.6H,0, in water, and dilute to 1 litre. TREATMENT OF SAMPLES- Wheat kernels (var. Thatcher) were milled by the procedure described for the Ottawa micro flour mill,9 and the weights of bran, feed and flour were recorded; the yields were approximately 65 per cent. of flour, 27 per cent. of bran and 8 per cent. of feed. PROCEDURE- Place a weighed portion of the sample in a lipless 1-litre Berzelius beaker, moisten with 10 to 15 ml of water, add 10 ml of nitric - sulphuric acid mixture, and evaporate gently to dryness. Allow the beaker and charred contents to cool, add sufficient nitric acid, sp.gr.1-42, to moisten the residue, cover with a clock-glass, and place on a hot-plate at about 350" C (as registered by a surface thermometer). Evaporate to dryness, and continue heating for about 20 minutes. Remove from the hot-plate, allow to cool, and repeat the addition of nitric acid and subsequent heating until white patches appear in the residue. Substitute fuming nitric acid for the nitric acid, sp.gr. 1.42, when incandescence will not occur in the sample, and repeat the operation until a white residue is obtained. Remove the clock-glass, and continue heating until oxides of nitrogen have been expelled. Cool, add 15 ml of dilute sulphuric acid (1 + 4), cover the beaker, and heat at 350" C until fumes are evolved. Cool, rinse the sides of the beaker with 15 to 20 ml of water, and heat uncovered until fumes are evolved; do not evaporate to dryness.Repeat the rinsing and heating, then transfer the acid solution to a 200-ml Berzelius beaker; repeatedly rinse the 1-litre beaker, with water, adding the rinsings to the contents of the 200-ml beaker, until the final volume is about 100 ml. Evaporate to about 20 ml, allow to cool somewhat, add 5.0 ml of sodium tartrate solution, and heat to 80" C. Then add 4-0 ml of dimethylglyoxime solution, 2 drops of bromothymol blue indicator solution and ammonia solution, sp.gr. 0.90, dropwise, with constant stirring, until the solution becomes purplish in colour; the pH, which should be 8 to 9, can be checked externally by means of an indicator paper. Cover the beaker, and set aside, preferably overnight.With use of a glass rod having a rubber-covered tip and repeated washings with small portions of water, transfer all the precipitate to a fine-porosity 15-ml sinter funnel, and retain the beaker for the next step (see C , Fig. 1). Add 1.0 ml of saturated bromine water to the contents of the funnel, and bring the liquid into contact with the entire wall area by means of the glass spatula. After 5 minutes, apply suction, and collect the filtrate in the beaker retained previously. Continue to apply suction, and wash the sinter with hot water. (The inner beaker is covered with a polythene disc during this operation to exclude any rubber "crumbs" from the stopper-see C , Fig. 1 ; the stem of the sinter funnel passes through a hole bored in the disc.) Add a second 1-ml portion of bromine water as before, and wash well with hot water.Boil the filtrate for at least 15 minutes to expel bromine, keeping the volume constant by making occasional additions of water. (If desired, this is a convenient stage at which to interrupt the procedure.) Place the beaker and solution in cold water, and add, with stirring after each addition, 1 ml of ethanol, 3 ml of sodium tartrate solution, 6 ml of potas- sium persulphate solution, 5 ml of ammonia solution and 2 ml of dimethylglyoxime solution (in that order). Transfer the solution to a 50-ml calibrated flask, dilute to the mark with washings from the beaker, mix well, and measure the optical density at 465 mp. If a white precipitate forms, pass a portion of the solution through a sinter funnel before measuring the optical density.The final red-brown colour is stable for at least 3 hours.652 HOFFMAN [Vol. 87 STANDARD GRAPH- Place suitable portions of diluted standard nickel solution in a series of separate beakers, add 1 ml of dimethylglyoxime solution to each, and shake. Add 2 ml of saturated bromine water, boil as described above for the samples, and continue as above. A graph of optical density against weight of nickel present is linear (slope 0.0045) and passes through the origin. TEST FOR NICKEL IN FILTRATES- Place the filtrate in a separating funnel, and extract three times with 5-ml portions of chloroform. Combine the chloroform solutions, and extract with a 10-ml and then a 5-ml portion of 0.33 N hydrochloric acid.Place the combined acid extracts in a beaker, add 2 ml of saturated bromine water, and boil until chloroform has been completely expelled. Place the beaker and contents in cold water, and continue as described above for the sample. DISCUSSION OF ’THE METHOD Table I shows results for the recovery of nickel added to untreated wheat fractions, the weights taken being the same as those of the samples. The satisfactory results and the absence of nickel in any of the filtrates tested confirmed (a) that recovery of the nickel was complete and (b) that Middleton and Stuckey’:; method for destroying organic matter can be used in determining nickel. TABLE I RECOVERIES OF NICKEL FROM WHEAT FRACTIONS Fraction Weight of sample, Nickel added, Nickel recovered, g Pg Llg Bran .. 1~0000 10.0 7.4 30.0 26.8 60.0 56.0 Feed . . 1~0000 10.0 9.7 30.0 28.7 60.0 57.1 Flour . . 10~0000 10.0 10.2 30-0 32.3 30.0 31-6 60.0 60.0 60-0 59.0 The proposed procedure is comparatively ra:pid and extremely convenient for determining the nickel contents of a large number of wheat samples. An analyst’s daily average output of twelve to fifteen determinations can be increased by using additional sets of apparatus. Detailed information on the various field treatments and the resulting distribution and accumulation of nickel in the wheat fractions is given elsewhere.1° I thank Mr. J. W. Kemp for technical assistance. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Keil, H. L., Frohlich, H. P., and Van Hook, J. O., PhytopathoEogy, 1958, 48, 652. Peturson, B., Forsyth, F. R., and Lyon, C. B., Ibid., 1958, 48, 665. Keil, H. L., Frohlich, H. P., and Glassick, C. E., Ibid., 1958, 48, 690. Forsyth, F. R., and Peturson. B., Ibid., 1959, 49, 1. -- , Plant Disease Refity., 1960, 44, 208. Forsier, W. A., Analyst, 1953, 78, 560. Alexander, 0. R., Godar, E. M., and Linde, N. J., Ind. Eng. Chem., Anal. Ed., 1946, 18, 206. Middleton, G., and Stuckey, R. E., Analyst,, 1954, 79, 138. Kemp, J. G., Whiteside, A. G. O., MacDonald, D. C., and Miller, H., CereaE Chem., 1961, 38, 50. Hoffman, I., Carson, R. B., and Forsyth, F. R., J . Sci. Food Agric., 1962, 13, 423. Received February 19th, 1962

ISSN:0003-2654    

DOI:10.1039/AN9628700650

出版商:RSC    

年代:1962

数据来源: RSC